info:doi/10.1038/440426a 2006-03-23T00:00:00Z 2006-03-23T00:00:00Z Don E. Canfield &copy; 2006 Nature Publishing Group Nature 440 7083 426 427 0028-0836 1476-4679 (?-mix:\s+\W+|\W+\s+|\s+|\/) methane inclusions primary years not billion probably ago biosphere have quartz life 3.5 dykes which organisms been Formation ancient some bubbles microbial Dresser Ueno composition isotopic rocks but silica et cyanobacteria ecosystems evidence cannot found oxygen-producing within al formed secondary these known subsurface most us oxygen Earth's 13 surface carbon 1 geological rather growth record remains more can Fig But active strongly Earth results page many samples anoxygenic associated deep origin much phototrophs nature Western place present methanogens slightly has crust therefore all C-depleted authors consistent one together sulphate about identified into series non-biological basalts sedimentary younger viewed tell These issue Australia C fluid production early today organic than know source their do at before terminate processes thermally compelling interpretation vast lavas housing lake originated possible incomplete still contained combined bubbly produce while Other Many maybe structures emerging compositions further complexity believed had introduced sediment sequence proof zoom stromatolites use time Another unknown now gas paper Might insights 516 presence evolution 517 normal different history population kilometre controversial same methanogenesis anything Between fractured concentrations scattered activities around adds 3.45 however 3.42-billion-year-old stromatolite reported those certain usually South extracted like piece return indicates deposition during crushed complex biological Evidence marine accomplishes oldest-known so whether dominant estimates sulphate-reducing distinct producing higher quite existed also value thermal setting So involved picture type investigations interpreted cyanobacterial teeming magnitude confident possibly minerals face randomly resealed coastal Unfortunately them Apex few puzzle argument 3.8 massive note numerous colleagues photosynthesis contain dioxide measured contrast anaerobic take don't aerobic existence stratigraphic alone pillow very identify think enquire Carbon gases pieces concentrated 2 reducers forward including margins water along point I being explore energy others microbes sediments 4 extra missing commonly environments below mats let mainly layers Here components see photosynthetic All evolved CH story affiliation Chert our data earliest Small broadly intruded scarce suggest crucial depleted thermogenic where million biologically isotopes Piecing after morphology they normally putting mat-forming majority exists biogenic modern decided understanding However bacteria produced accept means here specific backdrop jigsaw first anticipate chemistry least intriguing report align connect aspect Africa Still research 300 lack good its precipitated deposited structure represents microorganisms. hydrocarbons Fossil uncovered (?-mix:\.\s+(?=[A-Z])) A deep subsurface biosphere exists today, and by some estimates it is as massive as the one on the Earth's surface A few of the pieces connect, but many of them don't, and most are missing All in all, I take the data at face value, and accept that the authors have probably uncovered the oldest-known samples of biologically produced gas Another intriguing aspect to Ueno and colleagues' research is its geological setting Between 3.5 billion and 3.45 billion years ago, a series of lavas known as pillow basalts, and sedimentary rocks, were deposited in what is now the Dresser Formation But the presence of life within Earth's crust, combined with good evidence for life in surface environments, suggest that by 3.5 billion years ago the Earth was teeming with microorganisms. But we can be certain that an active biosphere was in place By contrast, the methane from the primary inclusions was much more strongly depleted in 13 C, and is consistent with a microbial origin Carbon isotopes of this type, however, cannot tell us which organisms were present Evidence for the activities of sulphate-reducing bacteria, which use sulphate rather than oxygen in energy production, has been reported from the Dresser Formation For further insights into early Earth ecosystems, we zoom forward around 300 million years to the Dresser Formation in Western Australia Fossil structures known as stromatolites, which are probably of biological origin, have been found in slightly younger rocks from the same stratigraphic sequence Here, a series of organic remains has been interpreted as possibly being those of cyanobacteria However, microbial mats have been identified in 3.42-billion-year-old rocks from South Africa , and the chemistry of the rocks indicates that anoxygenic phototrophs (photosynthetic organisms that do not produce oxygen) were probably the dominant mat-forming organisms In their investigations, Ueno et al . crushed samples of quartz ( Fig. 1 ) and measured the concentrations of the extracted gases and the isotopic composition of the methane carbon It is from here that Ueno et al . ( page 516 of this issue) report evidence for the existence of an ancient population of methane- producing microbes, commonly known as methanogens It was probably quite complex, including many of the components of modern anaerobic ecosystems Many of the bubbles align along the growth margins of the quartz, and so are believed to have formed as the quartz was precipitated: they are viewed as primary inclusions Might all of the methane have a non-biological source? The authors anticipate this argument and note that their primary fluid inclusions lack the higher (C 2+ ) hydrocarbons that are normally associated with the most strongly 13 C-depleted thermogenic methane Of these, we can identify anoxygenic phototrophs, sulphate reducers and methanogens Other bubbles are scattered randomly within the quartz (and are probably primary), while still others are concentrated where the quartz was fractured and resealed Piecing together the early history of life from the geological record is like putting together an incomplete jigsaw puzzle Small bubbles (fluid inclusions) are found within quartz minerals in the dykes, and these bubbles contain mainly carbon dioxide and water, but also some methane (CH 4 ) So whether or not cyanobacteria were present is crucial to our understanding of the nature of the biosphere 3.5 billion years ago Still, the isotopic composition of the methane from the primary inclusions is not very different from that of the most 13 C-depleted, thermally formed methane The basalts were intruded by numerous silica dykes, some of which terminate in the sedimentary layers (see Fig. 1 of the paper on page 517) The evidence, found in the isotopic composition of organic carbon , is rather scarce, but it is compelling The organisms involved in stromatolite growth are unknown The primary and secondary inclusions contained methane with distinct isotopic compositions The results of Ueno et al . do not tell us anything about the magnitude of the ancient subsurface biosphere The secondary inclusions had a composition consistent with (but not proof of) a non-biological, thermal source of methane that was probably introduced some time after the silica dykes were first in place The silica dykes housing the methane originated from deep in Earth's crust — maybe as much as 1 kilometre below the surface The story is this These are viewed as secondary inclusions These results, therefore, point to an active subsurface biosphere This is not the normal bubbly lake or coastal marine sediment usually associated with methanogenesis This means that the dykes are ancient and were formed during the deposition of the sediments This represents one of the earliest specific microbial processes to have been identified in the geological record: it adds an extra piece to the emerging picture of the structure and complexity of ancient microbial ecosystems To explore this issue, we return to Western Australia and to the Apex Chert, which is slightly younger than the Dresser Formation Unfortunately, a cyanobacterial affiliation cannot be decided by morphology alone, and the interpretation of these remains as biogenic is controversial We cannot, therefore, be confident that oxygen-producing organisms existed 3.5 billion years ago We know — or at least we think we know — that life evolved by 3.8 billion years ago What about oxygen? An aerobic biosphere was not possible before the evolution of oxygen-producing cyanobacteria, and oxygen-producing photosynthesis accomplishes the vast majority of the primary production today With this backdrop, let us enquire more broadly into the nature of life 3.5 billion years ago Inclusive evidence. This optical photomicrograph shows quartz from the Dresser Formation in cross-polarized light. The coloured crystals are coarse-grained quartz, which contain the tiny fluid inclusions analysed by Ueno et al . . (Image courtesy of Yuichiro Ueno; width is about 3 mm.) info:doi/10.1126/science.283.5402.674 info:doi/10.1038/nature04584 info:doi/10.1038/35065071 info:doi/10.1130/0016-7606(1999)111<1256:OOGCSI>2.3.CO;2 info:doi/10.1038/nature02888 info:doi/10.1038/416076a info:doi/10.1016/S0016-7037(03)00462-9 info:doi/10.1073/pnas.95.12.6578 1