The relationship between Earth’s complement of free oxygen and life seems to have begun in the Archaean, but it presented a series of paradoxes: produced by photosynthetic organisms oxygen would have been toxic to most other Archaean life forms; its presence drew an important micronutrient, dissolved iron-2, from sea water by precipitation of iron-3 oxides; though produced in seawater there is no evidence until about 2.4 Ga for its presence in the air. It has long been thought that the paradoxes may have been resolved by oxygen being produced in isolated patches, or ‘oases’ on the Archaean sea floor, where early blue-green bacteria evolved and thrived.
A stratigraphic clue to the former presence of such oxygen factories is itself quite convoluted. The precipitation of calcium carbonates and therefore the presence of limestones in sedimentary sequences are suppressed by dissolved iron-2: the presence of Fe2+ ions would favour the removal of bicarbonate ions from seawater by formation of ferrous carbonate that is less soluble than calcium carbonate. Canadian and US geochemists studied one of the thickest Archaean limestone sequences, dated at around 2.8 Ga, in the wonderfully named Wabigoon Subprovince of the Canadian Shield which is full of stromatolites, bulbous laminated masses probably formed from bacterial biofilms in shallow water (Riding, R. et al. 2014. Identification of an Archean marine oxygen oasis. Precambrian Research, v. 251, p. 232-237).
Limestones from the sequence that stable isotope analyses show to remain unaltered all have abnormally low cerium concentrations relative to the other rare-earth elements. Unaltered limestones from stromatolite-free, deep water limestones show no such negative Ce anomaly. Cerium is the only rare-earth element that has a possible 4+ valence state as well one with lower positive charge. So in the presence of oxygen cerium can form an insoluble oxide and thus be removed from solution. So cerium independently shows that the shallow water limestones formed in seawater that contained free oxygen. Nor was it an ephemeral condition, for the anomalies persist through half a kilometer of limestone.
The study shows that anomalous oxygenated patches existed on the Archaean sea floor, probably shallow-water basins or shelves isolated by the build up of stromatolite reef barriers. For most prokaryote cells they would have harboured toxic conditions, presenting them with severe chemical stress. Possibly these were the first places where oxygen defence measures evolved, that eventually led to more complex eukaryote cells that not only survive oxygen stress but thrive on its presence. That conjecture is unlikely to be fully proved, since the first undoubted fossils of eukaryote cells, known as acritarchs, occur in rocks that are more than 800 Ma years younger.