Insight into gold's origins could help prospectors.
Three billion years ago there was no life on land and no oxygen in the atmosphere. But the rivers ran with gold.
The world's biggest gold deposits washed up at their South African resting place in little bits, say geologists - possibly settling a century-old debate. Understanding the origins of the Witwatersrand Basin deposits could help prospectors to recognize the rock features that point to gold.
"The gold was transported into the basin by streams and rivers," says geologist Jason Kirk of the University of Arizona, Tucson. The gold was formed about three billion years ago, his team has found1. But the rocks on top are about 250 million years younger.
The arguments about how South Africa's Witwatersrand Basin became laden with gold have raged ever since the deposits were discovered. More gold has come from these 7,000 square kilometres than from any continent - about 50,000 tonnes over 120 years, nearly half of all the gold ever mined.
Some geologists think that water eroded the gold from rocks elsewhere and deposited it as sediment in the basin where it was then buried by other sediments - miners descend more than 2 kilometres to retrieve it. The new findings support this scenario.
Others argue that the basin's sedimentary rocks formed first, and gold was injected into them later when tectonic movements pushed superheated water containing dissolved gold deep underground. They think the case for erosion remains unproven.
If gold did form as a sediment, as Kirk's team suspects, prospectors should look for different rock structures than those that might trap fluid gold, says geologist Bruce Yardley of the University of Leeds, UK.
Kirk's team dates the South African gold by measuring the amounts of the elements rhenium and osmium mixed with it. Rhenium decays into osmium, so the proportions of these elements record when gold ore rose up into the Earth's crust from the mantle below.
Some of the gold probably washed in, agrees geologist Graham Pearson of the University of Durham, UK. But under the microscope, much of Witwatersrand's gold looks as if it crystallized within previously buried rocks, he says.
Gold may have accumulated through erosion, then dissolved in hot water and reset, he suggests. This is possible, Kirk agrees. As well as earthquakes and other movements of the crust, a meteorite struck the basin about 2 billion years ago.
The dates are unlikely to settle the debate, says Neil Phillips, head of exploration and mining at the Commonwealth Scientific and Industrial Research Organization in Melbourne, Australia. He suggests that some flaw in rhenium-osmium dating may have misled Kirk's team.
All the rocks surrounding the gold fields show clear signs of having been deformed by temperature and pressure, says Phillips. If the gold wasn't forced into the rocks, he says, this begs the question of what vast and ancient deposit was worn down to provide Witwatersrand's gold. "No one has suggested a viable explanation" for where this earlier deposit might have come from he says.
Kirk, J. et al. A major Archean, gold- and crust-forming event in the Kaapvaal Craton, South Africa. Science, 297, 1856 - 1858, (2002).