Credit: D.E. CANFIELD

Spectacular layered sedimentary rocks, such as those pictured here from the Hamersley Basin of Western Australia, are a widespread yet episodic feature of the Precambrian geological landscape. These ‘banded iron formations’ — the colouring is largely due to variable iron content — were laid down from an iron-rich ocean. Periods during which they accumulated are generally thought to have been ended by oxygenation of sub-surface ocean waters, forcing the oxidation and almost complete precipitation and deposition of dissolved iron from hitherto anoxic waters. However, as D. E. Canfield reports elsewhere in this issue (Nature 396, 450; 453; 1998), such a deep-reaching breath of fresh air may not, after all, have been needed to relieve the ocean of its ferrous burden.

Canfield bases his case on a simple model of cycling of oceanic oxygen and nutrients, along with measured sulphur-isotope signatures of sedimentary sulphide deposits and evidence of the emergence of various life forms. He argues that the ocean could have lost its iron from bottom waters that were akin to an oxygen-free sulphide ‘soup’. Focusing on the Proterozoic Eon (from about 2.5 billion years ago to the end of the Precambrian around 540 million years ago), he proposes that known episodes of oxidation of the Earth's surface were insufficient to make the deep ocean aerobic until at least one billion years ago. If so, the end of the last great deposition of banded iron formations, some 1.8 billion years ago, stemmed not from the oxygenation of the deep — as generally held — but from some other process.

The sulphur-isotope data implicate a shift in the marine sulphur cycle as an alternative explanation. Canfield hypothesizes that a known oxidation of the Earth's surface about two billion years ago increased seawater sulphate concentrations by oxidizing continental mineral sulphides. Such an increase could have relieved existing limitations on the oceanic rate of sulphate reduction to sulphide, so allowing sufficient buildup of sulphide to precipitate out the ocean's dissolved iron as iron sulphides. On this analysis, the great oxygenation of the ocean interior did not occur for about another billion years, following another Earth-oxidation event which swept the deep waters clean of sulphide and, perhaps literally, breathed new life into the ocean.