Earth’s atmosphere certainly has a propensity to turn iron into its oxidised form, rust – Fe2O3.nH2O. But long before steel was made and cars were built, the atmosphere was apparently much less oxidising, instead containing methane and other gases. The textbook view is that a global irreversible event, the Great Oxygenation Event, marks the dramatic change from an anoxic atmosphere to one that contained free oxygen and rusted minerals and cars alike. As explained in a recent On The Rocks blog post, this textbook view is being questioned by new geochemical evidence arguing for a much earlier, more protracted and episodic build-up of free oxygen. Here I review some of the reasons why the idea of a Great Event of Oxygenation came to be in the first place.
Our journey takes us to Lake Huron in Ontario, Canada. Along its north shore and extending a couple of hundred kilometers inland, exquisite exposures of sedimentary rocks are found containing straight-forward, no-nonsense geological evidence for a dramatic and sharp transition of Earth’s surface conditions. The exposures in question represent a pile of sedimentary rocks – the Huronian Supergroup – originally nearly 10 km in thickness. They were deposited in what might have been a failed rift basin, where rapid subsidence of the ground permitted fast accumulation of sediments. The general picture of sedimentation is that of repeating cycles.
At the base of the pile, we find the bedrocks of the Superior Province, consisting of old (Archaean) granites and greenstones. Into this flat ancient landscape rivers incised narrow, E-W trending valleys and deposited sediment and basalt into topographic lows: initially conglomerates interlayered with lavas, then sandstones, before water depth increased and mud settled in calmer conditions (see Figure 1). The quartz and feldspar-rich sandstones of the Matinenda Formation don’t differ much from younger equivalents other than by the fact that in their pebble beds, they contain detrital pyrite (see Figure 2). Tumbled pebbles of pyrite do not exist in modern rivers as the mineral readily decomposes, literally within days, under an oxidising atmosphere. Thus, the presence of these tumbled pyrites shows that both the atmosphere and the river water in which they were transported were devoid of free oxygen. Topping the cycle are rocks of the Ramsey Lake Formation, spectacularly unsorted clastic detritus deposited below ice, providing evidence for a substantial glaciation event – possibly the first pervasive in Earth history.
And then we are back in muddy conditions before the sandstones of the Mississagi Formation were deposited. These again contain pyrite grains but also another unusual mineral –uraninite. Like pyrite, it too is unstable under oxidising conditions. It is best demonstrated with a Geiger counter, which detects ionising radiation such as alpha and beta particles and gamma rays. Listen here for a comparison between the radiation emitted by Mississagi sandstone and a much younger Devonian sandstone. In the context of the Huronian Supergroup, the presence of uraninite thus further supports conditions without free oxygen. Further up in the stratigraphy, we find two more cycles of spectacular glacial sediment, represented by the Bruce and Gowganda Formations. Earth’s climate was thus very unstable, possibly reflecting the on-going changes in atmospheric composition and a concomitant adjustment of the greenhouse.
The final piece in the puzzle is only found by the time the last muddy sediment, the Gordon Lake Formation, was deposited. Even from a distance, its outcrops look very different, because they are distinctive dark red, representing the oldest known red beds on Earth (Figure 3).
In the case of the Gordon Lake Formation, the red beds consist of silt- to mud-sized detritus with ferric oxides. The presence of these ferric oxides is strong evidence for free oxygen but a complicating factor is that geologists distinguish between primary, secondary and diagenetic red beds. Only the first represent deposits of originally red soil sediment, whereas the oxidation in diagenetic red beds could have occurred a long time after the original sediment was deposited. Here is where it pays to look at the rocks in some more detail.
On several of the bedding surfaces the muddy beds, the trained eye recognises the polygonal traces of cracks (check out the brighter areas above the pencil in Figure 4). These could be desiccation cracks formed when the tidal flat in which the mud was deposited fell dry and the mud contracted.
The alternative would be cracks that form below water due to changes in salinity and associated contracting of clay minerals and expulsion of water from the sediment. Regardless, the fact that the iron in the sediment around the cracks is reduced is strong evidence that the red colouration of the mud is a primary feature and direct evidence for oxygenation of the atmosphere. As seen in cross section (Figure 5), the cracks are clearly filled with sandy material that was deposited above the mudflat, suggesting that the cracks probably formed by desiccation.
In summary, none of the newly reported geochemical or isotopic lines of evidence for oxygenation of the atmosphere before the Great Oxidation Event are accompanied by sedimentological observations as compelling as those preserved in the Huronian Supergroup. Until similarly convincing observations are presented in older sedimentary rocks, interpretations of geochemical and isotopic datasets, in my view, will remain equivocal.
By Prof. Balz Kamber, Chair of Geology & Mineralogy, Trinity College Dublin.