COUNTLESS passers-by have trampled the crime scene. Outlines of bodies lie smeared and jumbled. Witnesses are petrified into silence. Fingerprints have been washed with boiling water and other evidence tampered with almost beyond recognition. It sounds like a case for a fictional detective. But instead it’s a headache for geologists pursuing signs of early life in the oldest rocks on Earth.
Two years ago the case appeared clear-cut. The earliest signs of life were chemical fingerprints – telltale carbon signatures in 3.8 billion-year-old rocks from Greenland, and in 3.5 billion-year-old microfossils from Western Australia. That means life emerged no more than half a billion years after the young Earth became habitable. But now the evidence is being challenged. “The whole context of the Earth’s early history has, we think, been misunderstood,” says Martin Brasier from the University of Oxford, one of the geologists who have re-opened the case. Suddenly the earliest signs of life that investigators can agree are more than a billion years younger – nearly one quarter the age of the Earth.
With enough twists and turns to rival a courtroom drama, exhibits previously admitted as evidence for early life could instead be non-living remnants of the primeval soup. And without hard evidence of the event itself, could life have been slower to get started than everyone thought? The geologists still seem to believe that it was quick off the mark. But even if absence of evidence does not amount to evidence of absence, the ruckus has far-reaching implications. In less than a year, two separate missions will land on Mars to hunt for signs of ancient life. Until the case is closed in our own backyard, they have little chance of solving the Martian question.
Advertisement
There are only a few places where remnants of the Earth’s ancient crust have been cordoned off from the cycles of plate tectonics. The Apex Chert is one such site, an outcrop of 3.5 billion-year-old rocks at the Chinaman Creek in Pilbara, Western Australia. In the late 1980s and early 1990s, Bill Schopf of the University of California at Los Angeles found structures resembling single-celled microbes inside the flint-like chert particles. The chert was thought to have settled in a shallow lagoon or river mouth, which could have offered a comfortable home for early life. Schopf described the structures as microfossils of 11 species of photosynthetic microbes similar to cyanobacteria, and sent them to the Natural History Museum in London as examples of the Earth’s oldest fossils.
Brasier visited the museum in 1999 to photograph the specimens for a geology textbook. But when he examined the sections of rock, he was stunned by what he saw. Besides the cell-like shapes that Schopf described, there were also complex branching and folding structures unlike simple microbes. “I felt ill,” he recalls. “So much depends on these: the assumption about early life, the assumption about what we’ll find on Mars.”
Reeling from the shock, Brasier collected samples of his own from the Apex Chert and spent hours tracing the structures inside the rocks with his son Alexander, a geology student at the University of Edinburgh. “We began to see that the ‘microfossils’ were part of a wide spectrum of odd-looking structures, most of which were far too chaotic to be called fossils,” Brasier says. Rather than being relics of early life, he thinks the rock was once a glassy gel that recrystallised as it cooled. “This causes all sorts of pretty patterns,” Brasier explains. “People show pictures of little blobs and say ‘oh, look, these look a bit like bacteria’, but there are dozens of processes out there which simulate little blobs.”
The same objection applies to early stromatolites, mushroom-shaped rocks containing intricate textures thought to have been created by colonies of microbes. “We can generate very complex and sophisticated stromatolites completely abiogenically,” says Brasier, “It’s like ripples or snowflakes: a self-organising structure.” He points out that these structures form particularly easily around hydrothermal vents – volcanic springs on the sea floor. And, pointing to satellite images and detailed geological maps, Brasier and other geologists now argue that the Apex Chert bed is actually the upper part of a dyke or volcanic pipe. So instead of being a blue lagoon, the site may have been a seething hydrothermal vent and volcanic pipe extending perhaps 2 kilometres into the Earth’s crust. “That immediately raises questions,” says Brasier, “How could [photosynthetic] cyanobacteria survive down there?”
Schopf is indifferent to this re-evaluation of the evidence. “I’m agnostic – either way, life was there,” he says. He has prepared further evidence supporting his view, using a technique called laser Raman spectroscopy. This can measure the degree of ordering of carbon atoms, distinguishing highly ordered crystalline graphite, for example, from less well-ordered graphite that might once have been messy living material. Schopf’s microfossil structures contain a higher concentration of jumbled carbon than the surrounding rock matrix, suggesting that they could be remnants of early life. The journal Nature published Brasier and Schopf’s diametrically opposed papers back-to-back in March 2002 (vol 416, p 73).
Since then, other researchers have argued that Raman spectroscopy alone cannot provide evidence that material was once alive. Meanwhile, another group has offered testimony supporting Schopf’s case, describing a similar range of structures associated with much younger microbial fossils that have been transformed by intense volcanic heating.
“For every interpretation there is an equal and opposite interpretation,” comments Stephen Mojzsis of the University of Colorado at Boulder. He argues that while certain shapes may seem to provide evidence of early life, they are unreliable. He says the more compelling evidence lies in chemical fingerprints: carbon isotope ratios. And this has convinced him that life on Earth is truly ancient.
Just under 99 per cent of all Earthly carbon exists as the stable carbon-12 isotope, with just over 1 per cent as stable carbon-13. Living processes such as photosynthesis discriminate against the heavier carbon-13 isotope, so living or formerly living material contains a lower ratio of carbon-13 to carbon-12 than inorganic carbon. Geologists measure this in parts per thousand (per mil) using a scale known as δ13C to compare the ratio of carbon-13 to carbon-12 in their sample with that of an agreed standard. Biogenic material, with its higher carbon-12 content, is isotopically light so it usually has a telltale negative δ13C value of at least −20 per mil.
Schopf’s microfossil structures have δ13C values of around −26 per mil, which tallies with early life. But the ratio proves its worth in rocks older than 3.5 billion years, where carbon isotopes are all geologists have to go on in their search for the earliest signs of life. All rocks older than this are metamorphic: transformed from their original sedimentary or volcanic forms by high temperatures and pressures. This transformation tampers with evidence for life, obliterating any microfossils. It also alters carbon isotope signatures, but these changes are predictable and it is still possible to calculate the original carbon isotope values.
In the 1970s, geologists analysed the carbon isotopes in 3.8 billion-year-old rocks from Isua in west Greenland. Although heavily metamorphosed, the rocks looked like they were once sedimentary, originally laid down in what might have been a cosy underwater habitat. Their carbon isotope values pointed to the presence of photosynthesising early life. But the geology of Isua is very complicated and it turns out that many of the rocks may be volcanic in origin rather than sedimentary, making them unlikely cradles for life. What’s more, new research suggests that the isotopically light carbon previously reported in Isua rocks is a recent contamination, rather than the ghost of early life.
Geologists measure carbon isotopes by cooking a rock sample until carbon becomes carbon dioxide, which is then analysed by a mass spectrometer. Ancient carbon should be preserved in the Isua rocks as graphite, which combusts at 700 °C. But when Mark Van Zuilen of Scripps Institute in La Jolla, California, increased the temperature gradually during analysis, he found that the light carbon was given off at 450 °C. Rather than coming from ancient graphite, this biogenic carbon may have been introduced later by groundwater.
Other rocks from Isua that seemingly started out as sedimentary do contain graphite with carbon isotope values indicative of early life, once allowances are made for metamorphic changes. But even this evidence has been challenged. Ronny Schoenberg of the University of Queensland in St Lucia, Australia, has suggested it might be caused by contamination from an unlikely quarter. He has uncovered patterns of tungsten isotopes that suggest the metamorphosed sediments contain material from meteorites. These heavenly trespassers could have delivered carbon with isotope values of −18 per mil – similar to those found in the metamorphosed rock, which scientists have taken as transformed signs of life. Crucially, in meteorites, such values are not indicative of life but simply reflect different carbon ratios from non-living, extraterrestrial sources.
But meteorite bombardment should have left more widespread evidence, according to Mojzsis. “The most carbon-rich meteorites are only about 5 per cent carbon by weight, so what mechanism exists that you would have massive contamination of sediments by meteorites without any other physical or chemical evidence?” he asks. Having failed to find any other signs of meteorite impacts, Mojzsis argues that the tungsten anomalies come from other rocks. Yet another battle is being waged in the scientific literature.
Mojzsis believes he still has one star witness in the case for early life at Isua. He has found particles of graphite encased in a mineral called apatite, which would have shielded the graphite from the ravages of metamorphism and preserved their original carbon isotope values. And at around −30 to −50 per mil, those values point to early life. “The carbon we found in these tiny inclusions is very isotopically light,” says Mojzsis, “All perfectly consistent with early life.” But even this is not enough to convince sceptics like Van Zuilen, who thinks the light carbon may be the result of a small fraction of graphite failing to reach normal isotopic equilibrium during formation.
Fossilised soup
In another twist to the story, other researchers argue that other evidence of early life cited by Mojzsis may actually be the non-living remnants of the primordial soup – the broth of raw materials from which life emerged. Mojzsis has uncovered other isotopically light graphites in 3.8 billion-year-old rocks from Akilia Island on the west coast of Greenland. But the origin of the Akilia rocks has also been challenged. Chris Fedo of George Washington University in Washington DC and Martin Whitehouse from the Swedish Museum of Natural History in Stockholm showed that patterns of rare earth elements in the Akilia rocks match those found in rocks with volcanic rather than sedimentary origins. Mojzsis and others are contesting this interpretation (Science, vol 298, p 917A, 2002).
If the Akilia rocks were originally volcanic and later modified by interactions with hot fluids, then the isotopically light graphite could be the result of something called the Fischer-Tropsch synthesis. At temperatures of around 400 °C and hundreds of atmospheres of pressure, these reactions can turn non-living carbon monoxide and hydrogen into organic molecules such as hydrocarbons and possibly amino acids, using metal catalysts that would have been common on the early Earth. And laboratory experiments by Juske Horita of Oak Ridge National Laboratory in Tennessee and Michael Berndt from the University of Minnesota at Minneapolis show that Fischer-Tropsch reactions can produce abiogenic methane with an isotopic fractionation of −30 to −50 per mil – indistinguishable from supposed signs of life.
“On the pre-life world, organic matter must have been made by entirely non-biological means to get life started in the first place,” says Schopf. “Oddly, no one has yet identified remnants of the primordial soup in the early geologic record.” Brasier suggests that Fischer-Tropsch synthesis could have helped to make the primordial soup, and fingers it as a suspect in the case of the Apex Chert, which would explain why Schopf’s microfossils contain a high proportion of light carbon. If he’s right, the remnants of the primordial soup may have been under geologists’ noses all along, except that they were mistaken for signs of early life.
So what are the conclusive signs of ancient life itself that the geologists can now all agree on? There are chemical fingerprints in rocks from 2.7 billion years ago that seem to place cyanobacteria on the scene by that time. “I think people agree one of the more firm points is 2.7 billion years ago,” says Van Zuilen. But this is only one line of evidence and may yet be disputed. According to Brasier, the first really convincing microfossils do not appear until the Gunflint Chert from Ontario, Canada, 1.9 billion years ago. When he made that statement at a NASA astrobiology conference in April 2002, it caused a gasp of horror.
That’s because, despite all the wrangling, experts still cling to the idea that life began early in the Earth’s 4.5 billion-year history. This stance looks like an act of faith, given that the origins of life remain a mystery. Even so, among geologists the verdict seems unanimous. Mojzsis proposes that the prerequisites for the emergence of life – liquid water, raw organic materials and abundant sources of energy – were all in place 4.3 billion years ago. “By the time you have sedimentary rocks and a clearly retained hydrosphere, I suspect life can’t be far behind,” adds Fedo. Even Brasier suspects that life was probably under way by 3.5 billion years ago.
Although this debate hasn’t changed opinions about when life emerged, it has persuaded some geologists to rethink how they interpret evidence. Scientists sometimes follow the principle of Occam’s razor to draw conclusions, where the simplest explanation is regarded as the best. “The simplest interpretation of these [carbon isotope] data are as the remains of past life,” Mojzsis asserts. But Brasier argues that geologists should abandon this notion and instead consider that the evidence only points to the presence of life if this can be proven beyond reasonable doubt.
The prosecution and defence continue to argue, but both sides agree that the case needs to be resolved quickly. “This is just the kind of debate that needs to go on right now,” says Mojzsis, “because these rocks and the chemical information contained therein can be repeatedly tested in Earthly laboratories, unlike when we go to an ancient Martian surface and try to do these things remotely, without a map, without knowing what the rocks are or having any concept of their age.”
The debate over carbon isotopes casts doubt on whether the space probes poised for launch can really glean firm answers about early life elsewhere. “That could include robotic tests to be conducted by the forthcoming Beagle mission to Mars,” Brasier notes. When the Beagle 2 probe lands on the Martian surface at the end of this year, it will look for water, organic matter and carbon isotope fractionation in an investigation of past and present life. The Beagle team are confident of their approach. “This is a very full geological investigation and no results will be considered independently,” says team member Judith Pillinger of the Open University in Milton Keynes. “Interpretation will have to fit with the entire suite of data, with consideration of the full context.”
Beagle 2 or NASA’s Mars Exploration Rover Mission, which is due to land at the beginning of 2004, could still find evidence of past life on Mars, if clear signs exist. “There’s no way to get an abiotic dinosaur,” quips Fedo. But if the only evidence for ancient Martian life comes from carbon isotopes or cell-like shapes, the jury is likely to remain out until the case of early life on Earth is solved.