Oxygen

by Nick Lane

  This dramatic picture sounds plausible. If ice really did cover the whole Earth, then it is quite probable that only a few cells or tiny animals would survive, scratching a living in hot springs, or beneath translucent or thin ice, through which sunlight could penetrate.4 No doubt life was 4 Life would have survived comfortably in the hot springs and black smokers at the bottom of the oceans. Some say this is where life originated in any case; others say that the world was repopulated after the last snowball Earth by hydrothermal bacteria. It is therefore possible that evidence we think dates back to the origins of the life actually only dates back to the repopulation of the world following the snowball Earth bottleneck. I doubt that this view is true: the cyanobacteria are too distinct from the hydrothermal bacteria to have evolved from them so recently; and there is evidence of cyanobacteria from before all the snowball Earths. Somehow, then, the cyanobacteria survived all the snowball Earths, probably under thin ice near the Equator or in thermal springs on the Earth’s surface.

  64 • FUSE TO THE CAMBRIAN EXPLOSION

  lucky to keep a tenuous grip in the hellish acid bath that followed. No wonder there was so little burial of organic matter. After scourging itself so thoroughly, the Earth regained a climatic equilibrium. Now the survivors had a whole planet to themselves. They must have multiplied like mad. In this they were aided by high levels of minerals and nutrients, eroded by glaciers worldwide and swept into the oceans by the deluge. All these nutrients, all this empty space, must have stimulated the greatest Ediacaran

  radiation

  +8

  +6

  +4

  +2

  0

  (‰ PDB)

  δ13C

  –2

  Fe Fe

  Fe

  ?

  ?

  Late Precambrian

  1000

  900

  800

  700

  600

  500

  Time (million years ago)

  Figure 3: Changes in carbon isotope ratios during the late Precambrian and early Cambrian periods. The changes are given in parts per thousand (0/00) relative to the PDB standard. (The PDB standard is the carbon-13 level found in belemnite from the Pee Dee formation in South Carolina; belemnite is a form of limestone made from the calcification of an extinct order of molluscs related to squid, which were widespread in the Jurassic and Cretaceous periods.) The asterisk at the left-hand side shows the average present-day carbon-13 value. The peaks of carbon-13 (positive anomalies) indicate a substantial increase in the amount of organic-carbon burial (and therefore probably rising oxygen levels), whereas the troughs (negative anomalies) indicate virtually no organic-carbon burial. The negative troughs correspond to possible ice-ages or snowball Earths, of which the two most important were the Sturtian (750–730 million years ago) and the Varanger (610–590 million years ago). Fe indicates the presence of banded-iron formations. The cross marks a major extinction of microplankton that immediately predated the appearance of the Ediacaran fauna — the cushioned Vendobionts and the first worms. Adapted with permission from Knoll and Holland, and the National Academy of Sciences.

  Snowball Earth, Environmental Change and the First Animals • 65

  blooming of cyanobacteria and algae the world has ever seen: a world of blue-green ocean. These blooms must have produced prodigious amounts of free oxygen in a short period, oxygenating the surface oceans and air between each of the ice ages.

  All this extra oxygen could only persist in the air if it was not consumed by the respiration of other bacteria, or by reaction with rocks, minerals and gases. When oxidized, a single atom of iron loses one electron to oxygen to form rust. In contrast, each atom of organic carbon gives up as many as four electrons to form carbon dioxide. A single atom of organic carbon therefore consumes four times as much oxygen from the air as does a single atom of iron. By far the best way of preventing the complete re-uptake of atmospheric oxygen is to prevent it from reacting with organic matter; and the easiest way of doing that is to bury the organic matter rapidly.

  The essential difference between conditions today, in comparison with those that followed hard on the heels of the snowball Earths, is the rate of rock erosion, which is slower today than it was then. Slow erosion normally equates with slow burial of organic matter, as it takes longer for organic matter settling on the ocean floor to be buried under new sediments originating from rock erosion and organic matter. This leaves time for bacteria to break down the organic matter produced, for example, by algal blooms, consuming oxygen in the process. In present-day conditions, therefore, bacteria breaking down organic matter maintain the status quo. In contrast, high rates of erosion in the wake of glaciation lead to high rates of sedimentation and burial. Some organic carbon inevitably gets mixed up in this overall flux. In the aftermath of a snowball Earth, then, high rates of erosion ought to have led to high rates of carbon burial, and so to persistent oxygenation.

  The theory makes sense, but is there any evidence that the rate of erosion was high after a snowball Earth? Did this really lead to a rise in free oxygen? Think about that for a moment. How do we even start to answer questions like these? How can we possibly know what the rate of erosion was, 590 million years ago? Where should we look for evidence that free oxygen increased at this time? This is the very stuff of science, and the conclusions that can be drawn from clever reasoning backed up by precise measurements never cease to amaze me. There is indeed evidence that the rate of erosion increased after the snowball Earth, and that this was coupled to an accumulation of free oxygen.

  Each piece of evidence, by itself, leaves room for doubt, but taken

  66 • FUSE TO THE CAMBRIAN EXPLOSION

  together, I find the central assertion convincing — there was a rise in free oxygen in the immediate aftermath of the snowball Earth. This rise corresponded in time with the evolution of the first large animals — the Vendobionts. Here’s a brief resumé of the evidence so that you can make up your own mind (or marvel at the ingenuity of the human mind).

  We start with another set of isotope signatures. The rate of erosion in the distant past can be estimated by measuring the ratio of strontium isotopes in marine carbonates. Two stable isotopes of strontium — strontium-86 and strontium-87 — differ in their distribution between the Earth’s crust and the mantle underneath it. The mantle is rich in strontium-86, whereas the crust is more richly endowed with strontium-87. The major source of strontium-86 in the oceans is the igneous rock basalt. This rock is extruded continuously from the mantle at the mid-ocean ridges, from where it spreads slowly across the ocean floor before diving back into the mantle beneath the ocean trenches. A little strontium dissolves from the basalt into seawater. The speed of dissolution is more or less constant. The gradual build-up of dissolved strontium-86 in the oceans is balanced by a steady uptake of strontium by marine carbonates, such as limestone (calcium carbonate). This is because strontium can displace its sister element, calcium, in the crystalline structure of limestone. As each of these processes takes place at a steady rate, we would not expect the relative amount of strontium-86 in limestone to fluctuate a great deal. In fact it varies quite a lot. Strontium-87 is to blame.

  The quantity of strontium-87 in the oceans depends on the rate of erosion of the continental crust. Periods of glaciation and mountain-building intensify erosion and run-off into the rivers, delivering strontium-87 to the oceans. Like strontium-86, strontium-87 is incorporated into marine limestones. The ratio of strontium-86 to strontium-87

  incorporated depends on their relative concentration in sea water. In periods of high continental erosion, more strontium-87 gets into the oceans, so more is trapped in marine carbonates than in periods of low erosion. The ratio of the two strontium isotopes in limestone rocks of a certain age therefore gives an indication of the rate of erosion at the time that they were formed. According to Alan Kaufman of the University of Maryland and his Harvard University colleagues Stein Jacobs
en and Andrew Knoll, the ratio of strontium-87 to strontium-86 in marine carbonates rose steadily in the aftermath of the snowball Earth, indicating a high rate of rock erosion. Not only this, but the correlation between

  Snowball Earth, Environmental Change and the First Animals • 67

  carbon-isotope ratios (more carbon-12 buried) and strontium-isotope ratios (more strontium-87 in the rocks) implies that a high rate of erosion did indeed go hand in hand with a fast rate of carbon burial. This should have led to a rise in free oxygen.

  Two independent methods corroborate such a rise in oxygen. The first method hinges on the ratio of sulphur isotopes in iron pyrites, which are iron sulphides (FeS2), and was reported by Donald Canfield in Nature in 1996. We first met Canfield in Chapter 3, along with his ingeniously tangential conclusions based on the behaviour of sulphate-reducing bacteria, which reduce sulphate to hydrogen sulphide. Here he does it again for a later period. Working this time with Andreas Teske, at the Max Planck Institute for Marine Microbiology in Bremen, Canfield demonstrated an ecological turning point in the way that bacteria handled sulphur, starting soon after the meltdown of the last snowball Earth. For a full 2 billion years before this, the sulphide-producing activities of sulphate-reducing bacteria caused the sedimentary iron sulphides to become enriched in sulphur-32 by about 3 per cent compared with background ratios. Then, suddenly, after the last snowball Earth, around 590 million years ago, the sulphides in sediments became enriched by about 5 per cent. They have been enriched by about 5 per cent ever since, so this figure is virtually diagnostic of modern ecosystems. What had happened?

  The figure of 3 per cent is easy to explain. Sulphate-reducing bacteria rely on a one-step conversion of sulphate into hydrogen sulphide. This simple process enriches the sulphur-32 in hydrogen sulphide by about 3

  per cent. The enriched hydrogen sulphide is then free to react with iron to make iron pyrites. The trouble is, a 5 per cent enrichment cannot be achieved by a one-step bacterial process. It can only happen in an ecosystem that recycles its raw materials, in the same way that that we can concentrate carbon dioxide from our breath by repeatedly breathing in and out of a plastic bag.

  In the case of hydrogen sulphide, the recycling needs oxygen. The system works as follows. Sulphate-reducing bacteria thrive in stagnant muds at the bottom of the sea. The hydrogen sulphide they produce percolates up the water column and reacts with oxygen filtering down. A mixing zone develops between the stagnant conditions deep down and the aerobic conditions higher up. Today, this zone is inhabited by numerous inventive bacteria that live on sulphur. Some of these oxidize hydrogen sulphide, producing the element sulphur, while others convert the

  68 • FUSE TO THE CAMBRIAN EXPLOSION

  elemental sulphur back into a mixture of sulphate and hydrogen sulphide. Because this sulphate is regenerated biologically, it is itself enriched in sulphur-32. The sulphate-reducing bacteria take this biologically generated sulphate and convert it back to hydrogen sulphide. Each cycle enriches the sulphates and the sulphides a little bit more in sulphur-32.

  Finally, an average of about 5 per cent enrichment is achieved. This value of 5 per cent is just an equilibrium point, at which hydrogen sulphide is likely to react with iron to produce iron pyrites. Once formed, the heavy iron pyrites settle out into the bottom sediments, preserving the equilibrium for posterity.

  What Canfield and Teske propose, then, is that ‘modern’ types of ecosystems, requiring modern levels of oxygen, began to develop soon after the end of the last snowball Earth. They back their conclusions by molecular-clock calculations, which confirm an increase in the number of species of sulphur-metabolizing bacteria at this time. Thus, Canfield and Teske project a rise in atmospheric oxygen to nearly modern levels in the final years of the Precambrian.

  The second method that points to a rise in free oxygen is the pattern of so-called rare-earth elements. The relative amounts of these trace elements, such as cerium, in marine carbonates depends on their abundance in sea water at the time; and this depends on their solubility. The solubility of many elements differs according to the level of oxygen. We have already seen that iron becomes less soluble in the presence of oxygen, whereas uranium becomes more soluble. If we see a shift in the relative concentrations of different elements in the rocks (some becoming more abundant, some less so) we get an indication of the degree of oxygenation of the oceans at the time of their formation. According to Graham Shields, at the University of Ottawa in Canada, and Martin Brazier, at the University of Oxford, the marine carbonates that formed in Western Mongolia during and after the snowball Earth period record a shift in the pattern of rare-Earth elements, indicating a rise in the oxygenation of the oceans.

  Uniquely in the history of our planet, all these factors — the carbon isotopes, sulphur isotopes, strontium isotopes and rare-earth elements —

  simultaneously point to a rise in free oxygen. Indeed, the wild swings in environmental conditions during the 160-million-year snowball Earth period may have pushed atmospheric oxygen up to nearly modern levels.

  At the same time, however, there was a re-emergence of banded-iron

  Snowball Earth, Environmental Change and the First Animals • 69

  formation, after a hiatus of nearly a billion years, suggesting that the deep oceans still contained large amounts of dissolved iron. If so, there can have been little oxygen in the ocean depths.

  We emerge blinking, then, from the great Varanger ice age — the last snowball Earth — which ended some 590 million years ago, into a world in which the surface oceans and the air are well oxygenated — well enough for us to breathe — but the deep oceans are still stagnant, like the Black Sea today, saturated in hydrogen sulphide. Then suddenly, within a few million years of the dawn of this new and better world, we find the first large animals, the strange bags of protoplasm known as Vendobionts, floating in the shallow waters, and worms wending their way through the muddy bottoms of the continental shelves, all across the world. This was an age bursting with potential. Strangely, the fulfilment of the potential may have spelled its early demise.

  The philosopher Nietzsche once remarked that mankind will never mistake himself for a god so long as he retains an alimentary tract; the need to defecate is our most unflattering quality. In a clever paper published in Nature in 1995, Graham Logan and his colleagues, then at Indiana University, contradicted Nietzsche, arguing, in effect, that we owe our most god-like qualities, indeed our very existence, to the primal need for defecation. Faecal pellets from the first large animals, they say, cleansed the oceans, paving the way for the Cambrian explosion. Few theories of environmental change in the terminal Precambrian are quite so down to earth (or at least, seafloor).

  Basing their arguments on a detailed study of carbon isotopes in molecular fossils, Logan’s group found that virtually all the organic carbon produced during the long period of environmental stasis from 1.8 billion to 750 million years ago was not buried in sediments, but was instead broken down again and reused by bacteria lower in the water column. The dead remains of the tiny, almost weightless, bacteria sank very slowly into deeper waters, giving consumers plenty of time to reuse any available carbon. As most carbon was reused, the rate of carbon burial was low. Because oxygen only accumulates when carbon is buried, there can have been very little long-term accumulation of oxygen in the air, and little stimulus for change. Worse than this, any oxygen percolating down the water column was neutralized by hydrogen sulphide rising up

  — a situation that could sustain the stagnant conditions indefinitely. In the wake of the very first snowball Earth (2.3 billion years ago), high rates of erosion and carbon burial brought about a sudden change, but the

  70 • FUSE TO THE CAMBRIAN EXPLOSION

  debris was ultimately depleted and the status quo — tortuously slow organic burial — was restored. The restoration of this status quo after the snowball Earth seems to account for the fact that oxygen levels did not rise above 5 to 18 per cent for
a billion years. Left to their own devices, then, bacteria may never have broken out of an endless equilibrium.

  Logan argues that the status quo was finally broken for ever by the evolution of animals with guts — a leap that could only be achieved in shallow waters with the aid of oxygen (only oxygen-requiring respiration is efficient enough to support the evolution of multicellular animals with guts). The relatively heavy faecal pellets of these animals must have sunk rapidly to the ocean bed, cutting swathes through the heaving population of oxygen-hating sulphate-reducing bacteria. Peppering the bottom sediments with nutrients, the faecal pellets were buried in turn under more sediments, so depriving the sulphate-reducing bacteria of their organic nutrients and, through their burial, contributing to the oxygenation of overlying waters. The lean pickings for the sulphate-reducing bacteria, combined with better oxygenation of deep waters, must have hastened the retreat of these oxygen-haters to the anoxic bottom sediments.

  Suddenly, at a time when the huge genetic potential implicit in the segmented bilateral structure of the Cambrian animals was poised, just waiting for an opportunity, a vast, new, well-oxygenated ecosystem opened up like the promised land. The expansion of these motile, predatory, genetically loaded animals into the vacant eco-space must have given the gentle floating Vendobionts no chance. They must have been shredded like swollen plastic bags in a combine harvester.

  The fall of the Vendobionts to predators is speculative, but there can be no doubt that rising atmospheric oxygen did correlate with radiations in biological diversity in the Precambrian era. In Chapter 3, we noted a link between oxygen and the rise of the eukaryotes, and following that, a link between oxygen and the first stirrings of multicellular life. Now we must consider the links between oxygen and the appearance of the first animals of any size, the Vendobionts, and the bilateral segmented Cambrian animals in turn (see Figure 4). While these links are beyond question in themselves, it is an embarrassingly common mistake to confound a correlation for a causal relationship. You may recall that Preston Cloud’s third criterion for linking oxygen to evolution was a good biological basis for