The story of O:
How did oxygen come to be the second most abundant gas on Earth?
Paul G. Falkowski1 and Yukio Isozaki2
1
Department of Earth and Planetary Sciences and Institute of Marine and Coastal
Sciences, Rutgers University, 71 Dudley Road, New Brunswick, NJ 088901, USA
falko@imcs.rutgers.edu
2
Department of Earth Science & Astronomy, The University of Tokyo, 3-8-1
Komaba, Meguro, Tokyo 153-8902, Japan
isozaki@ea.c.u-tokyo.ac.jp
The gas composition on a planet reflects (bio)geochemical processes on its
surface. Two gases overwhelmingly dominate Earth’s atmosphere: N2 and O2. The
former is primordial and its presence and abundance are not driven by biological
processes. Indeed, N2 is virtually inert and has an atmospheric lifetime on the order
of ~ 1 billion years {Berner, 2006 #12694}. In contrast, O2 is continuously produced
biologically via the oxidation of H2O driven by energy from the sun. The gas almost
certainly was virtually nonexistent in Earth’s atmosphere when the planet was
formed, is highly reactive, and has an atmospheric lifetime of ~ 4 million years
{Keeling, 1993 #12695}. Yet, despite the ~250 ratio in atmospheric lifetimes, O2
came to comprise between ~ 10 and 30% of the atmospheric volume for the past
several 500 million years {Berner, 2004 #102; Falkowski, 2005 #109}. How did O2, a
gas critical to the evolution of animal life, become the second most abundant gas on
Earth? The story isn’t as simple as it might first appear {Catling, 2005 #12141;
Kump, 2008 #13008}. Here we briefly examine the story of O.
How the Earth got water
O is produced as an element (not the gas, O2) via the so-called “main line”
reaction sequence from successive 4He fusion reactions in hot stars {Ziurys, 2007
#13014}. It is delivered to planets chemically bound to other elements. Through
successive cycles of heating and cooling O reacts with Si and C to form two of the
major anions that, together with metal cations, comprise the fundamental minerals in
mantle and crust (ca. 80% of Earth’s volume), and with H to form water {Holland,
1984 #4786}. Additional water was also delivered to the planetary surface via
meteriorites and possibly comets; however, the proportion of the three sources is not
well constrained by H/D ratios in the contemporary ocean {Robert, 2001 #9125}.
Regardless, isotopic data suggest that Earth’s surface contained liquid water within
~200 million years following the accretion of the planet {Mojzsis, 2001 #7367}.
Liquid water is a necessary condition for life as we know it, but it is not a sufficient
condition for the biological production of O2.
How water is split.
Although H2O can be oxidized to its component elements by high energy
(UV) photons, this reaction can produce only extremely small concentrations of O2
because of strong negative feedbacks {Kasting, 1981 #13007}. By far, the
overwhelming source of oxygen is the biological oxidation of water, driven by a
single microbially-derived process, the evolution of which remains poorly understood
{Blankenship #1336; Falkowski, 2007 #12961}. It appears to have arisen once in a
single clade of Bacteria, and was appropriated via a single primary endosymbiotic
event to become the progenator of all photosynthetic eukaryotes including algae and
higher plants {Falkowski, 2007 #12961}. The core of the oxidation machinery is
found in Photosystem II, a multimeric protein complex containing 4 Mn atoms which
are successively oxidized by single, photocatalyzed reactions that create electron
holes upstream. This is only known 4 electron transfer reaction in biology, and critical
details on how it works are still unknown. Regardless, O2, produced as a waste
product, via the reaction:
2H2O ---- 4 e + 4 H+ + O2
This reaction is basically a biogeochemical half-cell; the protons and electrons
generated are primarily used to reduce CO2 to allow the growth of new cells (and
hence the formation of organic matter):
CO2 + 4 e + 4 H+ ------ (CH2O) + H2O
On time scales of years to millennia, these reactions are closely coupled to the reverse
process of respiration, such that net production of O2 is virtually nil. That is, without
burial of organic matter in Earth’s rocks, there would be very little free O2 in the
atmosphere. Hence, the evolution of oxygenic photosynthesis, the most complex
energy transduction process in biology, was a necessary, but not sufficient condition
to oxidize Earth’s atmosphere.
The improbability of tectonics
Net oxidation of the atmosphere requires long-term storage of the reductants,
primarily as organic carbon. By far, the major reservoir is the crust. The major
mechanism for burial of organic matter is sedimentation and accretion onto cratons
(stabilized continents), and to a lesser extent subduction deep into the mantle. These
processes are driven by plate tectonics; a process that is presently unique to Earth
among the planets in our solar system. Radiogenic heat in Earth’s interior drives
mantle convection allowing continents to collide and separate to form new ocean
basins {Wilson, 1966 #215; Worsley, 1986 #152}. These collisions and separations
are cyclical; the oceanic lithosphere (plate) moves away from mid-oceanic ridge and
eventually subducts under the continent to close up the ocean once again. At the
subduction margin, a fraction of the organic matter buried in the sediments
{Premuzic, 1982 #8641} is tectonically added to continents, forming coastal
mountain belts, i.e. increasing continental mass. Indeed, unless organic matter stored
in marine sediments is stabilzed in cratons, it will be subducted, heated, and in part
recharged to the atmosphere via volcanism, where it would become reoxidized, or
subducted much deeper into the mantle {Tappert, 2005 #13004}. Thus, burial of
organic matter, which contains reducing equivalents derived from the biological
oxidation of water, implies a net oxidation of the atmosphere. The basic concept is
that oxygen in the atmosphere requires an imbalance between oxygenic
photosynthesis and aerobic respiration on time scales of millions of years; hence to
generate an oxidized atmosphere more organic matter must be buried than respired.
Although the oxidation of the atmosphere initially appears to have occurred about 2.2
billion years ago, probably by the rapid radiation of cyanobacteria and stabilization of
the nitrogen cycle {Fennel, 2005 #12096}, the atmospheric concentrations were
probably ~ 1% by volume or less, and the deep ocean was likely still anoxic
{Canfield, 1998 #1859}. Large increases in atmospheric oxygen appear to have
occurred much later.
The oxidation of the atmosphere may also have gotten a boost in the latest
Precambrian age about 750 million years ago, when tectonic processes accelerated the
burial of organic matter. Metamorphic rock records through history suggest that heatflow from the Earth’s interior decreased steadily as the planet gradually cooled for
nearly 4 billion years, and the geothermal gradient reached a threshold where hydrous
(OH-containing) minerals were subducted deep into the mantle at depths of 410-660
km {Maruyama, 2005 #13013}. This process would have led to a massive transfer of
surface water into mantle, which, in turn, slowly lowered sea-level (by up to 600 m,
25) and potentially accelerated extensive emergence of continents. The geotherminduced first major sea level drop in history would lead to extensive continental
erosion and the concomitant voluminous mass wasting along continental margins,
which would, in turn, accelerate the burial of organic matter. The burial of large
amounts of organic carbon over the past 750 million years is mirrored in a significant
rise in atmospheric oxygen, which is often blamed for triggering the Cambrian
explosion of metazoans {Knoll, 1999 #12208; Maruyama, 2005 #13013}. Further
increases in burial efficiency were potentially accelerated by the evolution of large
eukaryotic algae, in the Proterozoic which sink rapidly and become incorporated into
sediments as organic matter, and much later by the rise of land plants, especially trees,
in the Carboniferous. The former are major sources of petroleum and gas while the
latter are sources of coal.
How well do we know the history of oxygen on Earth? Surprisingly perhaps,
not very well. Most of inferences come from isotopic analyses, especially of carbon
and sulfur {Berner, 2004 #102}. Photosynthetic carbon fixation strongly
discriminates against 13CO2, such that the resulting organic matter is 20 to 25 parts per
thousand enriched in 12CO2 relative to the source carbon. In contrast, carbonates,
formed by the precipitation of HCO3 with Ca or Mg, retain the isotopic signal of the
source carbon (e.g., HCO3 in seawaters). If photosynthesis exceeds respiration
(implying burial of organic carbon), 12C is removed from the mobile pool of carbon in
the atmosphere and oceans, leaving 13C enriched as a source of carbon for biological
and geochemical processes. Thus, the isotopic composition of carbon in carbonates
can be used to track the extent to which organic matter was buried and was returned
to the mobile pools over geological time {Canfield, 2005 #13011}. This record
reveals intervals of strong isotopic excursions and secular trends spanning hundreds
of millions of years. One of these excursions occurred about 350 million years ago
and corresponds to the rise of land plants in the Carboniferous. Models suggest that
this period of Earth’s history witnessed O2 concentrations as high as about 30%
{Berner, 2004 #102}. By the end of the Triassic extinction, O2 concentrations appear
to have been significantly lower {Isozaki, 1997 #5139}, perhaps even as low as 12 to
15% {Falkowski, 2005 #109; Berner, 2006 #13010; Belcher, 2008 #13009}, but rose
secularly in the ensuing 200 million years, to the current level of 21% {Falkowski,
2005 #109}. However, it is very difficult to observe a long-term secular increase in
C in carbonates over the past 2 billion years – implying that the reservoir of
13
inorganic carbon in Earth’s mantle is extremely large relative to the fraction of
organic carbon buried {Hayes, 2006 #12241}, and that the burial of organic carbon is
roughly balanced by oxidation and weathering {Berner, 2004 #102}.
This very brief exploration of how oxygen came to be the second most
abundant gas on this planet clearly shows the incredible contingencies required.
Biological, geochemical and geophysical processes conspired to produce the
planetary atmosphere that allows complex animal life to have evolved. Perhaps
ironically, although it has been known for over 200 years that oxygenic
photosynthesis is responsible for the production of oxygen, we still do not understand
the basic mechanism responsible for water splitting, nor do we understand what
controls the concentration of the gas in our planetary atmosphere {Berner, 2004
#102}. The former issue almost certainly will be resolved within a decade, as
increasingly higher resolution structures of the photosystems become available,
together with increasingly sophisticated biophysical approaches to measuring electron
transfer reactions {Barber, 2008 #13006}. The latter issue will be more difficult to
constrain, but a better understanding will emerge from more complete (not
necessarily, complex) models coupled with better integrated biogeochemical
measurements {Hayes, 2006 #12241}. We do know however, that the gas is not
present in abundance on any other planet or their moons in our solar system. Hence,
we left to search for the existence of life on those objects based on a compromise
between our technological capabilities and our unfulfilled dreams of searching for
proxy evidence. As technological capabilities develop, we can search for the existence
of oxygen on planets outside of our solar system {Lawson, 2007 #13012}. While the
identity of the gas on distant planets is difficult to confirm, it is a critical component
in understanding how complex life can exist on Earth and how rare this planet really
is.
References
Figure 1. Legend goes here.