Oxygen
Page 6
34 • SILENCE OF THE AEONS
total). This means that the overall ratio of carbon-12 to carbon-13 on or in the Earth is constant (99.89 to 1.11). In other words, if we add up the total amount of carbon in plants, animals, fungi and bacteria, and buried as coal, oil and gas, and present in the air as carbon dioxide, and dissolved in the oceans and swamps as carbonates, and petrified as carbonate rocks (such as limestone), then we will find that the overall ratio of stable carbon isotopes is 99.89 to 1.11.
Despite this fixed ratio, there are still some small but definite variations in the ratio of carbon-12 to carbon-13 in the carbon buried in the rocks. These variations are brought about by living things, and so far as we know, only by living things. The reason is that photosynthetic cells using carbon dioxide from the air or sea to make organic matter prefer to use carbon-12. This is because the lighter carbon-12 atoms have a slightly greater vibrational energy, which means that a smaller input of energy (activational energy) is needed for a reaction to take place. The reactions of the carbon-12 isotope are therefore catalyzed more quickly by enzymes than those of the heavier (less vibrational) carbon-13 isotope. The faster rate at which carbon-12 bonds are cracked means that organic matter becomes enriched in carbon-12 relative to carbon-13. In fact, the ratio of carbon-12 to carbon-13 is skewed towards carbon-12 by an average of 2 or 3 per cent compared with the unadulterated background ratio.
When the remains of plants, algae or cyanobacteria are buried in sediments, their extra carbon-12 is buried with them. Because the buried organic matter is enriched in carbon-12, it is impoverished in carbon-13.
This means that more carbon-13 is left behind as carbonates in the oceans or rocks, or as carbon dioxide in the air. This is called the principle of mass balance – which simply says that what is buried below the ground cannot be found above the ground. The implications of this elementary idea have a surprisingly long reach. Both carbon-12 and carbon-13 are incorporated into carbonate rocks (such as limestone) in a ratio that reflects their relative concentration in the oceans. As more carbon-12 is buried as part of organic matter, more carbon-13 is left behind in the oceans, and so the carbonate rocks have a relatively high content of carbon-13. Thus biological activity is betrayed in two different ways: by an enrichment of carbon-12 in buried organic matter, such as coal, or by an enrichment of carbon-13 in carbonate rocks such as limestone.
Geological periods conducive to carbon burial, such as the Carboniferous (about 300 million years ago) with its huge, low-lying swamps and
Three Billion Years of Microbial Evolution • 35
massive coal seams, leave robust carbon-12 signatures in organic inclusions such as coal in the rocks. The farther back in time we go, the harder it is to read carbon signatures, if only because less and less organic matter survives intact. Eventually, the samples shrink to the size of grains and require sophisticated equipment to read them. With this in mind, Steven Mojzsis and colleagues set about studying the ancient Greenland rocks, determined to think small. Their approach brought swift rewards: they found minute carbon residues trapped inside grains of a calcium phosphate mineral called apatite. Apatite can be secreted by microorganisms, but can also crystallize inorganically from the oceans, so the association of carbon with apatite is, in itself, no more than suggestive of life. When the Scripps team examined the carbon-isotope ratios, however, the results were startling. The carbon inclusions were enriched in carbon-12 by as much as 3 per cent over the normal background ratio. As a leading geochemist, Heinrich Holland, remarked in the journal Science: “the most reasonable interpretation of the data is surely that life existed on earth more than 3.85 billion years ago.” Not only this, but life may even have discovered the trick of photosynthesis, which is, after all, the main source of carbon signatures today.
Is this credible? Other pieces of evidence fit the same story. Moving forward a mere 300 million years, to the 3.5-billion-year-old rocks in Warrawoona in Western Australia, we find microscopic fossils that resemble modern cyanobacteria. Throughout the Precambrian period, most cyanobacteria lived in communal structures called stromatolites: great domes of living rock, which grew to heights of metres. A few living stromatolites are still found today in the right conditions — in Shark Bay in Western Australia, for example. Nearby, shapes resembling modern stromatolites are imprinted in rocks 3.5 billion years old. There is little evidence of geothermal activity, past or present, in these bays, so it seems likely that the microbes living in these ancient stromatolites gained their energy from photosynthesis, just as they do today. While none of these findings is conclusive on its own, when taken together, the carbon signatures, microfossils and fossil stromatolites do make it look as if photosynthetic bacteria were already colonizing the early Earth at least 3.5
billion years ago.
The earliest definitive evidence for the existence of cyanobacteria must wait another 800 million years. We are now 2.7 billion years before the present, floating in the shallows of an ocean that was soon to precipitate some of the largest iron-ore formations in the world. Today we can
36 • SILENCE OF THE AEONS
visit these iron formations in the Hamersley Range, near Wittenoom in Western Australia. For such old rocks, they have suffered relatively little chemical and physical change, called metamorphosis by geologists. Heat and pressure, the twin forces of metamorphosis, tend to destroy flimsy biological molecules. Because the Hamersley Range had suffered so little metamorphosis, Jochen Brocks and his colleagues at the Australian Geological Survey and University of Sydney, held out hope that a few ancient molecules — characteristic biological fingerprints called biomarkers —
might have survived intact in the shales underlying the iron formations.
After conducting a painstaking series of extractions and laboratory tests to eliminate the possibility of contamination with more recent molecules, their hopes were rewarded in full when they discovered a rich mixture of recognizable biomarkers. Their work was promptly published in Science in August 1999, with a flurry of commentary. Not only had the Australian surveyors found fingerprints diagnostic of cyanobacteria — that is, molecules found only in cyanobacteria — they also found a large number of complex steranes, a family of molecules derived from sterols such as cholesterol, which have only ever been found in the cell membranes of our own direct ancestors, the single-celled eukaryotes.
The finding was a double whammy: proof that oxygen-producing cyanobacteria and the first representatives of our own eukaryotic ancestors coexisted not less than 2.7 billion years ago. The earliest known fossils of eukaryotic cells date to about 2.1 billion years ago, so Brocks and his colleagues had pushed back the evolution of the eukaryotes 600
million years. This is significant in terms of the environment that must have existed to support these cells. Apart from anything else, the biosynthesis of sterols is an oxygen-dependent process, requiring more than a trace of oxygen in the atmosphere. Modern eukaryotes can only synthesize sterols if given at least 0.2 to 1 per cent of the present atmospheric levels of oxygen, and there is no reason to suppose that their ancestors were any different. If the cyanobacteria had indeed evolved between 3.5
and 3.85 billion years ago, as was suggested by the fossil evidence in the Warrawoona rocks and the carbon signatures, it is quite plausible that some free oxygen could have accumulated in the atmosphere by this time. But did this increase in oxygen correspond exactly in time to the evolution of the eukaryotes? And if so, did the rise in oxygen in fact stimulate their evolution?
Three Billion Years of Microbial Evolution • 37
Trends in carbon-isotope ratios can be used, in principle, to calculate changes in atmospheric oxygen. This is because the burial of organic matter prevents the complete oxidation, by respiration, of the carbon produced by photosynthesis. As photosynthesis and respiration are essentially reverse reactions, the one generating and the other consuming oxygen, any increase in the amount of carbon buried should lead to an equivalent increase
in the amount of free oxygen left over in the air. If we know exactly how much carbon was buried at any one time, then, in principle, we can calculate how much oxygen must have been left in the air. In practice, however, unless we can be certain that the rate of oxygen removal by volcanic gases or the erosion of land masses remains constant, all we can say is that there was a qualitative increase in oxygen. During recent geological history, the younger rocks preserve a detailed history of environmental change, and we are sufficiently familiar with most of the important parameters to calculate oxygen levels on the basis of carbon burial, as we shall see in Chapter 5. Unfortunately, this approach is unreliable when dealing with the very ancient Precambrian period — there are so many uncertainties that, at best, we only get a sense of the direction of change. For a more quantitative estimate, we must employ other methods.
One clue to oxygen levels during this period is to be found in the very same iron formations that overlie the shales of the Hamersley Range.
Massive sedimentary iron formations were deposited here and around the world in alternating bands of red or black ironstone (haematite and magnetite, respectively), and sediment, typically flint or quartz. The individual bands range in depth from millimetres to metres, while the formations themselves can be up to 600 metres [approximately 2000 feet] thick.
Most of these formations were deposited between 2.6 and 1.8 billion years ago, but sporadic outcrops range in age from 3.8 billion to 800 million years.
Today, after the exhaustion of most premium ore deposits, the banded iron formations are by far the world’s richest source of low-grade iron ore. According to the US Geological Survey, world iron-ore resources still exceed 800 billion tons of crude ore, containing more than 230 billion tons of iron, much of which comes from Australia, Brazil and China.
Of this total, at least 640 billion tons were laid down between 2.6 and 1.8
billion years ago. The Hamersley formation alone contains 20 billion tons of iron ore, with 55 per cent iron content.
Exactly how these iron formations came into being, or why they should be banded, is a mystery. Or rather, there are so many possible
38 • SILENCE OF THE AEONS
explanations, and so little evidence to support one theory over another, that few geologists would be bold enough to attempt a categorical explanation. There have nonetheless been some imaginative attempts. Ancient superstition held that large deposits of haematite (from the Greek ‘blood-like’) formed from the streams of blood that flowed into the ground after great battles. More scientifically, the banding of ironstones has been attributed to cyclical extinctions of algal populations, overcome by their own toxic oxygen waste. Neither theory has much credence. In fact, there is no reason to suppose that all the formations were produced in the same way, especially those separated by deep gulfs of time. But some general principles do apply to them all, and these reveal something of the conditions under which they must have formed. Most importantly, no banded iron formations have been deposited since atmospheric oxygen approached modern levels. Because iron does not dissolve in the presence of oxygen, the immediate implication is that the oceans were oxygen-free before the deposition of the banded iron formations, and too well aerated to support their formation in later times. To tease the truth out of this implication, we will need to look at the behaviour of iron in a little more detail.
Only the Earth’s core, and meteorites, contain pure iron. Tools made from meteoritic iron are an expensive curiosity. All iron in ores from the Earth’s crust is oxidized to some extent, although we shall see that iron in the oxidized state does not always imply the presence of oxygen. There are two main forms of iron in nature, ferrous iron (Fe2+), which tends to be soluble, and the more highly oxidized ferric iron (Fe3+), best known in the guise of rust (ferric oxides), which is insoluble.2 In the presence of oxygen, soluble ferrous iron is oxidized to insoluble rust. Not surprisingly, there is very little iron dissolved in today’s well-ventilated oceans, as oxygen snatches electrons from dissolved iron and the ferric oxide compound precipitates out as rust before any iron build-up can occur. One exception is the poorly ventilated floor of the Red Sea, where dissolved iron is enriched to 5000 times normal levels, and only bacteria can survive. The early Precambrian oceans must have been similar in this respect: in the absence of oxygen, dissolved iron from volcanic emissions and erosion could have accumulated to very high levels.
2 The two forms differ in their degree of oxidation, ferric iron (Fe3+) being more oxidized than ferrous iron (Fe2+).
Three Billion Years of Microbial Evolution • 39
A second modern example gives an idea of what might have happened next. The Black Sea is the largest body of poorly oxygenated water in the world, and is stratified into two layers. The surface waters are well oxygenated to depths of about 200 metres [656 feet], and if not fished to oblivion support a teeming ecosystem, including the famous caviar sturgeon. In contrast, the deeper waters, which account for 87 per cent by volume of the Black Sea, are stagnant and cannot support animal life (with the sole exception, it seems, of nematode worms, the only known animal that can complete its life cycle in the absence of oxygen).
The current state of the Black Sea seems to have developed about 7500
years ago, several thousand years after the end of the last ice age, in an event that has been linked to Noah’s Flood by the marine geologists William Ryan and Walter Pitman of Columbia University. As the great land glaciers melted, the sea level around the world rose by several hundred feet. The Black Sea, however, was isolated in its own basin by a land bridge across the Bosphorus, and the glacial meltwater did not affect its depth as much as that of the surrounding seas. The basin was left low and dry, so to speak, well below sea level, as is the Dead Sea today.
Whether as the consequence of an earthquake, or stormy weather, or the pressure of the rising Mediterranean, the land bridge spanning the Bosphorus finally collapsed with a roar that must have sounded like the wrath of God. This, say Ryan and Pitman, was the reality of Noah’s flood.
Salt water poured into the low Black Sea basin at an estimated rate of 10
cubic miles [42 million cubic metres] per day — a cascade 130 times greater than the Niagara Falls. The villages clinging to the shores were drowned beneath the Mediterranean waters, they say, in a catastrophe whose memory reverberated around the ancient world. An area the size of Florida was added to the existing lake.
Since biblical times, the shallow, tideless straits of the Bosphorus have impeded mixing of the brackish Black Sea water with the saline water of the Mediterranean. The denser saline sinks to the bottom, and the undisturbed bottom waters rarely come into contact with the air. The only living things that thrive in these depths are anaerobic (oxygen-hating) bacteria. Many of these are sulphate-reducing bacteria, which generate the noxious gas hydrogen sulphide as a waste product. Because hydrogen sulphide reacts with any oxygen percolating down, the depths remain anoxic and the stratified system, once established, sustains itself. The build-up of hydrogen sulphide makes the deep waters of the Black Sea stink of rotten eggs, and stains the mud on the bottom black, giving the
40 • SILENCE OF THE AEONS
sea its modern name. Its ancient name, the Euxine, lends itself to the term euxinic, which refers to any foul-smelling sulphidic body of water, lacking oxygen, movement and animal life in the depths.
The Black Sea, although the largest, is not the only euxinic body of water on the planet. Similar conditions occur in some Norwegian fjords that are separated from the open ocean by shallow glacial sills. Even the oceans occasionally develop euxinic conditions. Climatic conditions sometimes conspire to cause an upwelling of nutrient-rich bottom waters to the surface. Here, the combination of plentiful nutrients and bright sunlight stimulates an algal bloom, leading to a massive but transient increase in biomass. As the nutrients are exhausted, the algae die and sink to the bottom. Their decay consumes oxygen faster than it can be replenished by currents or diffusion fr
om the oxygen-rich surface waters. These oxygen-poor conditions stimulate a second bloom, this time of oxygen-hating sulphate-reducing bacteria, which release hydrogen sulphide as they break down the organic matter. Stagnant conditions may set in for periods of months until the supply of decaying organic matter is exhausted. Occasionally, the stagnant waters well up to the surface, releasing hydrogen sulphide gas into the atmosphere. One such upwelling occurred in St Helena Bay near Cape Town in South Africa in 1998, provoking furious and misguided complaints about the smell of rotten sewage in the air.
Such combinations of circumstances may explain the genesis of banded iron formations. Back in Precambrian times, the low levels of atmospheric oxygen must have kept the oceans permanently euxinic. The surface waters, however, were home to photosynthetic bacteria at least 2.7 billion years ago, and perhaps as long as 3.8 billion years ago. As happens today, there must have been frequent upwellings of the bottom waters, bringing dissolved nutrients and iron into contact with the photosynthetic bacteria living in the surface layers. If these bacteria were cyanobacteria, as suggested by the biomarkers in the Hamersley Range, then they would have been producing oxygen as a waste product of photosynthesis. In such oxygen-rich waters, dissolved iron welling up to the surface would have precipitated out as rust, and sunk to the bottom of the ocean to form beds of red haematite and black magnetite.
If this was the case, the banding of ironstones with flint or quartz could have been produced by seasonal influences, such as higher rates of photosynthesis (and therefore oxygen production) in the summer than in the winter, or seasonal upwellings according to climatic variations. The seasonal fluctuations in iron deposition would have been set against a
Three Billion Years of Microbial Evolution • 41
steady precipitation of silica. This could not happen today. There is little dissolved silica in the modern oceans: it is extracted by some algae and lower organisms for use in their ‘skeletons’. However, in the days when bacteria ruled the waves, silica was not used in this way, and so must have continuously exceeded its solubility limit of about 14–20 parts per million. It would have precipitated in a steady rain to form thick beds of flint or quartz, alternating with seasonal beds of ironstone.