Remnant cores and bands containing filaments.
a, Silica-rich colloform cores containing filaments, surrounded by replacive pyrite (gold). Combined transmitted and reflected light (TL and RL) photographic montage. b, Thin banded layer containing filaments within paragenetically early chert (TL and RL). c, Detail of a showing the partial replacement of quartz-rich colloform structures displaying fine-scale lamination (TL and RL). Scale bar is 1 mm for a, b and 0.1 mm for c.
Western Australia's Pilbara region continues to yield tantalising clues on the origin of Earth's life forms.
Oxygen-producing photosynthesis may not have been around as long as previous studies suggest say Australian researchers, but not all are convinced.
Dr Birger Rasmussen of Curtin University of Technology in Perth and colleagues report their findings in today's issue of the journal Nature.
"Without oxygen we wouldn't have the development of more complex aerobic life - multicellular life. So [its production is] a really important event in earth's history," Dr Rasmussen said.
Nearly a decade ago, co-author Dr Jochen Brocks of the Australian National University in Canberra and colleagues reported evidence of photosynthesis in 2.7 billion-year-old rocks from Western Australia's Pilbara region in the journal Science.
They reported finding "biomarkers", molecular fossils, including hydrocarbons such as hopane and sterane, believed to come from the membranes of photosynthetic cyanobacteria.
This landmark study pushed evidence for oxygen-producing cyanobacteria back by 300 million years and presented scientists with a conundrum, said Dr Rasmussen.
He says it was not until 2.45 to 2.32 billion years that the first major rise in atmospheric oxygen occurred, as recorded in the Great Oxidation Event - a mass oxidation of minerals in the Earth.
"If you had oxygen-producing cyanobacteria at 2.7 [billion years ago] churning out oxygen, why did it take so long - why did it take 300 million years - before we see the results of that?" he said.
The biomarker study also set back the age of the rise of eucaryotes by a billion years - long before the Great Oxidation Event.
All in all it created "a yawning palaeontological divide", says Assistant Professor Woodward Fischer of California Institute of Technology, in a commentary accompanying the new study by Rasmussen and team.
Scientists have spent a lot of time developing models to explain this time lag but now Dr Rasmussen and colleagues say the earlier research got it wrong.
"You may not need an explanation [for the time lag]. It may be simpler than that," he said.
Dr Rasmussen and colleagues have carried out another analysis of Pilbara rocks, including samples used in the biomarker study.
They say that carbon-13 to carbon-12 isotope ratios associated with the biomarkers indicate they are younger than organic residues found in the rocks themselves, and they must have become associated with the rock at a later date.
"The results we've got suggest that the biomarkers were probably not indigenous to the rock they were extracted from," Dr Rasmussen said.
He says about 2.2 billion years ago the rock was heated as high as 200 degrees Celsius and the biomarkers must be no older than this because they are unlikely to have survived the heat.
"They could have entered the rock after that peak heating event, possibly infiltrated from younger sedimentary rock, or even during the drilling process [to obtain rock samples]," he said.
The original team's method involved crushing the rock and dissolving it in solvents, allowing for greater potential for contamination of the sample, whereas the new study analysed organic carbon residues in situ by examining ultra thin slices of the rock.
Not the last word
Geologist Professor Malcolm Walter, who heads up the Australian Centre for Astrobiology at Macquarie University in Sydney, says the latest study has been much anticipated as experts knew there were limitations to the earlier study.
But he says "it's definitely not the last word".
"I'm keeping an open mind on it at this stage," he said.
Professor Walter says there is independent evidence to suggest cyanobacteria was around earlier, including 2.7 to 2. 8 billion-year-old stromatolites found in the same rocks that yielded the biomarkers.
"Those particular stromatolites have characteristics that suggest they were built by cyanobacteria," he said.