Download New insights into the burial history of organic carbon on the early Earth

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Anoxic event wikipedia , lookup

Transcript
Geochemistry
Geophysics
Geosystems
3
Article
G
Volume 5, Number 8
5 August 2004
Q08001, doi:10.1029/2004GC000713
AN ELECTRONIC JOURNAL OF THE EARTH SCIENCES
Published by AGU and the Geochemical Society
ISSN: 1525-2027
New insights into the burial history of organic carbon on the
early Earth
Christian J. Bjerrum
Geological Institute, University of Copenhagen, Øster Voldgade 10, DK-1350 Copenhagen K, Denmark
([email protected])
Donald E. Canfield
Danish Center for Earth system Science (DCESS) and Institute of Biology, University of Southern Denmark, Campusvej
55, DK-5230 Odense, Denmark ([email protected])
[1] The isotope record of organic matter and calcium carbonate is often used to infer the burial history of
organic carbon through time. As organic carbon burial is widely held to control long-term oxygen
production, the isotope record also relates to the production rates of oxygen on Earth. Current
interpretations of the record suggest a long-term consistency in the proportion of total carbon buried as
organic carbon ( f ratio), with some important periods of much higher burial proportions. The isotope
record is analyzed here with a new carbon isotope mass balance model, which considers submarine
hydrothermal weathering of ocean crust as a significant removal pathway of inorganic carbon. With this
model the f ratio is considerably reduced if isotopically depleted inorganic carbon is precipitated during
hydrothermal weathering and if hydrothermal weathering dominates inorganic carbon removal from the
surface environment. In contrast to previous calculations, our analysis of the carbon isotope record shows
that organic carbon burial in the Archean accounted for only between 0% and 10% of the total carbon
burial. These low burial proportions would have contributed to a slow accumulation of atmospheric
oxygen in the Archean.
Components: 5711 words, 4 figures.
Keywords: Archean; carbon isotopes; mass balance; model; Precambrian.
Index Terms: 1010 Geochemistry: Chemical evolution; 1040 Geochemistry: Isotopic composition/chemistry; 4805
Oceanography: Biological and Chemical: Biogeochemical cycles (1615)
Received 6 February 2004; Revised 25 April 2004; Accepted 25 June 2004; Published 5 August 2004.
Bjerrum, C. J., and D. E. Canfield (2004), New insights into the burial history of organic carbon on the early Earth, Geochem.
Geophys. Geosyst., 5, Q08001, doi:10.1029/2004GC000713.
1. Introduction
[2] The history of organic carbon burial is usually
deduced from steady state carbon isotope mass
balance models. In constructing these models it is
assumed that inorganic carbon and organic carbon
are removed from the surface mixed layer of
the oceans into underlying sediments [Hayes,
1983; Hayes et al., 1999; Kump, 1991; Des Marais
et al., 1992; Des Marais, 2001]. This type of
model, applied to the Precambrian carbon isotope
Copyright 2004 by the American Geophysical Union
record, has lead to the prevailing view of a long
term consistency through time in the proportion
of total carbon buried as organic matter [e.g.,
Schidlowski, 1988; Watanabe et al., 1997; Catling
et al., 2001; Kump et al., 2001; Holland, 2002].
This does not, however, mean that the burial flux
of organic carbon has remained the same. Higher
heat flow in the Archean most likely was accompanied by a higher flux of CO2 from the mantle
[Des Marais, 1985, 1997; Sleep and Zahnle,
2001]. Greater CO2 degassing would have been
1 of 9
Geochemistry
Geophysics
Geosystems
3
G
bjerrum and canfield: burial history of organic carbon
accompanied by greater total carbon burial. Therefore associated with the long term consistency in
burial proportions of organic carbon is the prospect
that early Archean O2 production rates were even
higher than today, which is paradoxical with
evidence for low atmospheric oxygen concentrations [Holland, 2002]. We question, however, an
important basic premise of this model. In particular, we relax the assumption that inorganic carbon
is only removed from the surface ocean reservoir.
This assumption is not strictly valid today, as
bicarbonate is extracted from the deep ocean
during hydrothermal alteration of ocean crust
(ocean crust carbonatization) [Brady and Gislason,
1997; Wallmann, 2001], and most likely it was not
true early in Earth history [Walker, 1990].
[3] Thus, on the very early Earth, before significant continental crust formed, nearly all buffering
of volcanic CO2 degassing must have been through
ocean crust carbonatization (OCC) [Walker, 1990].
Even after significant amounts of continental crust
had formed, the presence of ubiquitous hydrothermal carbonates in Archean greenstone belts suggests that ocean crust carbonatization was still an
important CO2 sink [Veizer et al., 1989b; Nakamura
and Kato, 2002]. Model studies of the early carbon
cycle furthermore indicate that through the Archean
OCC probably accounted for between 30% and 90%
of the inorganic carbon removal [Sleep and Zahnle,
2001]. Ocean crust carbonatization influences
carbon isotope mass balance when the carbonates
so precipitated are depleted in 13C relative to the
surface waters. This occurs, for example, when there
is an isotope gradient from the surface to the deep
ocean. Today, the decay of 13C-depleted organic
matter in the deep ocean results in a 2% surfaceto-deep difference in the isotopic composition of
DIC (dissolved inorganic carbon) [Broecker and
Peng, 1982]. Therefore any carbonate precipitated
in isotopic equilibrium [Alt, 1999] with waters
at depth due to OCC will be on average 2%
d13C-depleted relative to carbonates precipitated
in the surface ocean [Des Marais, 2001]. Other
processes discussed below also could have contributed to 13C depletion during OCC.
[4] In Archean and early Proterozoic rocks an
isotope contrast is observed between carbonates
precipitated from surface and deep water. The
ancient carbonate rocks can be divided into two
broad groups with different isotopic compositions
depending on where they were precipitated. Inorganic carbon precipitated from the surface ocean
has d13C values around 0%, whereas carbonates of
10.1029/2004GC000713
hydrothermal origin, formed from massive carbonatization, have more 13C-depleted mean values of
1.3% [Des Marais, 2001; Veizer et al., 1989b].
Furthermore, typical micritic siderites in deepwater BIF (banded iron formation) facies have
d13C values around 3% to 5% indicating
deeper marine waters 13 C-depleted compared
to surface water [Beukes et al., 1990; Kaufman et
al., 1990]. The observations, then, point to two
isotopically distinct burial sinks for inorganic
carbon in the Archean, surface-derived carbonates
with a d13C of around 0% and deep water-derived
carbonates with relatively 13C-depleted values
compared to the surface ocean. In what follows
we will focus our discussion on OCC as it probably
represented the most important deep-water sink for
inorganic carbon [Walker, 1990; Sleep and Zahnle,
2001; Nakamura and Kato, 2002]. We recognize
that early diagenetic carbonate formation in deep
water sediments, fuelled by the decomposition of
organic matter, could also remove isotopically
depleted DIC, and that this process would also
contribute to isotope balance in the same manner as
we will explore below.
[5] Thus deep ocean carbonate removal likely
represented a sink for 13C-depleted carbon in early
Earth history. This sink influences isotope mass
balance, and importantly, the isotopic composition
of surface DIC, independent of organic carbon
burial. We develop a new carbon-isotope mass
balance model including deep-ocean carbonate
precipitation by way of OCC. We then calculate
the impact of this on organic carbon burial as
deduced from the Precambrian carbon isotope
record [Des Marais, 2001; Shields and Veizer,
2002].
2. Three Sink Carbon Isotope Mass
Balance Model
[6] Steady state mass balance requires that the flux
of carbon into the ocean-atmosphere system (Fin)
balances the total flux of inorganic carbon (Fb) and
organic carbon (Fbo ) out of the system [e.g., Hayes,
1983; Des Marais, 2001]:
Fin Fbo þ Fb ¼ 0:
ð1Þ
Total inorganic carbon removal is the sum of two
district fluxes; Fb = Fb,sc + Fb,oc, where Fb,sc is the
burial flux of sedimentary carbonate carbon and
Fb,oc is the removal flux of OCC carbon (Figure 1).
The carbon removed by OCC can be written as a
2 of 9
Geochemistry
Geophysics
Geosystems
3
G
bjerrum and canfield: burial history of organic carbon
10.1029/2004GC000713
isotopic differences the following isotopic mass
balance is obtained:
din dsc Db f lDs ð1 f Þ ¼ 0
Figure 1. New carbon isotope mass balance model.
Removal of inorganic carbon (Fb) occurs as sedimentary
stromatolithic carbon and inorganic precipitates in the
surface reservoir, but also in the deep ocean through
ocean crust carbonatization (OCC). The proportion of
DIC removed by OCC is given by l. Fob is the flux of
organic carbon buried, and Fin is the flux of dissolved
inorganic carbon (DIC) into the oceans. In the Archean,
significant OCC probably resulted in a greater deep
ocean inorganic carbon sink than today.
fraction (l) of the total inorganic carbon removed,
so Fb = (1 l)Fb + lFb, and the isotopic mass
balance of equation (1) may then be written:
Fin din Fbo dorg þ ð1 lÞFb dsc þ lFb doc ¼ 0;
ð2Þ
where doc is the isotopic composition of inorganic
carbon removed by OCC and dsc is the isotopic
composition of inorganic carbon precipitated and
buried in shallow water. din is the isotopic
composition of carbon entering the ocean-atmosphere system, and dorg is the isotopic composition
of buried organic carbon. Dividing equation (2)
through by Fin leads to
din f dorg ð1 lÞð1 f Þdsc lð1 f Þdoc ¼ 0;
ð3Þ
where f = Fob (Fin)1 is the fraction of total carbon
buried as organic matter ( f ratio), and following
from equation (1), (1 f ) = Fb (Fin)1.
[7] In principle, the isotopic difference between the
inorganic carbon removed by OCC and the carbonate precipitated from the surface ocean can be
constrained from observations, and this difference
is given as Ds = doc dsc. Likewise, the isotopic
difference between organic carbon and inorganic
carbon removed from the surface ocean can be
constrained from observations as is given by Db =
dorg dsc. [e.g., Hayes, 1983; Schidlowski, 1988;
Kump, 1991; Des Marais et al., 1992]. With these
ð4Þ
The last term on the left-hand side in equation (4)
is new compared to previous isotope mass balance
equations [e.g., Hayes, 1983; Kump, 1991; Des
Marais, 2001]. This term designates the importance of OCC as an inorganic carbon sink (l),
whose significance depends on the magnitude of l
and the isotopic difference between inorganic
carbon removed in the surface and deep ocean (Ds).
[8] If l = 0, then all inorganic carbon is removed
from the surface reservoir, and equation (4) is
identical to previous isotope models [e.g., Hayes,
1983; Kump, 1991; Des Marais, 2001]. In this limit
the isotopic composition (dsc) of sedimentary inorganic carbonate removed from the surface ocean is
given only by din, Db, and f, and it is independent of
the internal processing of carbon in the ocean
[Kump, 1991]. This is similarly true if there is no
13
C gradient in the ocean, meaning Ds = 0%.
[9] For a constant value of f, and assuming a Ds of
2%, as today, the isotopic composition of DIC
removed from the surface ocean (dsc) increases by
around 2% as l increases from 0 to 1 (Figure 2a).
This increase is independent of organic carbon
burial. The increase in dsc results from mass balance considerations and occurs as 13C-depleted doc
(compared to dsc) starts to dominate the inorganic
sinks (as l increases; Figure 2). Therefore the
isotopic composition of carbonate removed from
surface oceans is not a unique indicator of the
burial proportion of organic carbon, f, as long as
l > 0. An even larger effect on dsc is observed, as a
function of l, with a larger Ds (Figure 2b). With
Ds = 5%, as has been suggested for the Archean
and early Proterozoic (see below), increasing l
from 0 to 1 increases dsc by around 4%. Such a
significant influence of l on dsc shows that previous calculations of the f ratio in the Precambrian
are potentially very uncertain, and a reevaluation of
the carbon isotope record is thus required.
3. Precambrian Carbon Burial
[10] The history of the fraction of carbon buried as
organic carbon ( f ) over geologic time can be
analyzed by rearranging equation (4):
f ¼ ðdin dsc lDs ÞðDb lDs Þ1 :
ð5Þ
In using this equation, dsc and Db are obtained
directly from the isotopic record, Ds is potentially
3 of 9
Geochemistry
Geophysics
Geosystems
3
G
bjerrum and canfield: burial history of organic carbon
10.1029/2004GC000713
floor weathering [Walker, 1990; Nakamura and
Kato, 2002]. Furthermore, recent carbon cycle
models indicate significant early Archean OCC,
with l values ranging between 0.35 and 0.95, and
decreasing to 0.25 by the end of the Proterozoic
Figure 2. The isotopic composition of carbonate
carbon precipitated in the surface ocean (dsc) contoured
as a function of l, the relative importance of deep ocean
DIC removal, and f, the fraction of total carbon buried as
organic matter. (a) Contoured using equation (4) with Db
= 30 %, the Phanerozoic mean (equation (2)) and with
Ds = 2 %, as today. (b) The same as in Figure 2a, but
with Ds = 5 %. Increasing l significantly decreases
the f ratio associated with a given isotopic composition
of the surface carbonates (Figure 2b).
constrained from observations, and l is constrained
from independent carbon cycle modeling. As in
previous f ratio calculations, we need not consider
what happens to the carbon after it is buried
[Hayes, 1983; Schidlowski, 1988; Des Marais et
al., 1992; Hayes et al., 1999]. In other words, we
assume that the isotopic composition of volcanic
and weathered lithosphere carbon (din) remains
constant over time, with a value of 5%
[Haggerty, 1999]. To obtain values for dca and
Db, the isotope database record of Des Marais
[2001] and Shields and Veizer [2002] was filtered
with a Butterworth filter with a half power point
frequency cutoff of (0.3Gyr)1, after the d13C
record was first averaged in 10 Myr intervals and
linearly interpolated when no data were present in a
time interval (Figures 3b and 4b).
[11] Archean submarine basalts are extensively
carbonatized indicating probable significant sea-
Figure 3. Modeled Archean-Proterozoic carbon cycle
(3.5 Giga-annum before present (Ga) to 0.7 Ga).
(a) Conservatively calculated decrease in l the fraction
of DIC removed though ocean crust carbonatization
(OCC), with methane as an important greenhouse gas in
the early Precambrian [Sleep and Zahnle, 2001]. (b) The
isotopic composition of inorganic carbon (top, dsc, from
Shields and Veizer [2002]) and organic carbon (bottom,
from Des Marais [2001]) are shown through the
Precambrian. Solid curves are the filtered d13C record.
(c) Different possible values of Ds, the isotope
difference between carbonate removed from the surface
and deep ocean reservoirs. The solid and dash-dot
curves are extrapolations of Ds estimated from isotope
composition of Archean rock (see text). The dashed line
is the present mean ocean Ds. (d) Calculated fraction
of total carbon buried as organic matter ( f ) from
equation (5), where Db is the difference between the
lower and upper curve in Figure 3a, din = 5%, and l is
from Figure 3b. Ds is a free parameter, where the three
curves are as in Figure 3c (dashed, solid, and dash-dot).
The fine dotted line is for l = 0, as in equation (4). In
all cases the calculated f ratio is lower than previously
interpreted (thin dashed line, f 0.2).
4 of 9
Geochemistry
Geophysics
Geosystems
3
G
bjerrum and canfield: burial history of organic carbon
10.1029/2004GC000713
[Sleep and Zahnle, 2001]. The lower estimates of l
were calculated assuming that methane was an
important Archean greenhouse gas, and that consequently, atmospheric CO2 concentrations were
relatively low (Figure 3a) [e.g., Pavlov et al., 2000;
Sleep and Zahnle, 2001]. These models assume a
constant and modern continental crust size. The
mass of the continental crust, and its availability to
remove CO2 by weathering, will influence the
evolution of l. For this reason we have recalculated from the equations of Sleep and Zahnle
[2001] the evolution of l for the methane greenhouse scenario (with low atmospheric CO2) and
assuming continental growth. To do this, the silicate weathering expression of Sleep and Zahnle
[2001] was multiplied by the normalized continental crust volume reconstruction of Kamber et al.
[2003] where the crust reached 70% of its present
volume by 2 Ga (Figure 4a). (The free parameters
of Sleep and Zahnle [2001] were set to Fridge =
2.4 1012 mol/yr, a = 0.9, and thydro = 3 106 year.)
In this example very high l values of near 1 are
observed in the early Archean, decreasing to
around 0.39 at the end of the Precambrian. The
model results all point to significant Archean OCC
and the prospect of an important second sink for
DIC from the ocean-atmosphere system. If the
isotopic composition of the OCC sink differs from
the surface reservoir sink there will be, as we have
seen above, an error in the calculation of the f ratio
if the OCC sink is not considered.
Figure 4. Modeled Archean-Proterozoic carbon cycle
with continental growth. (a) Recalculated decrease in l,
the fraction of DIC removed though ocean crust
carbonatization (OCC) with methane as an important
greenhouse gas and continental growth up to 2 Ga
(calculation after Sleep and Zahnle [2001]). (b) As in
Figure 3b, solid curves are the filtered d13C record. (c)
Different possible values of Ds, the isotope difference
between carbonate removed from the surface and deep
ocean reservoirs, are presented, where the solid and
dash-dot curves are extrapolations of Ds estimated from
isotope composition of Archean rocks (see text). The
dashed line is the present mean ocean Ds. (d) Calculated
fraction of total carbon buried as organic matter (f) from
equation (5), where Db is the difference between the
lower and upper curves in Figure 4a, din = 5 %, and l
is from Figure 4b. Ds is a free parameter, where the three
curves in Figure 4c (dashed, solid, and dash-dot) are
used to calculate three possible f ratio curves. Fine
dotted line is for l = 0, as in equation (4). The solid line
represents our best estimate for the f ratios in the
Archean and early Proterozoic for the given l.
[12] To ascertain the isotopic composition of the
OCC sink, we must evaluate Ds. As mentioned
above, Archean carbonatized rocks of various ages,
which are considered part of the regional lower
temperature carbonatization halo, have a mean
d13C value of 1.3 ± 1.6%, and an end-member
seawater(?) value of +2.2% [Veizer et al., 1989b].
The regional halo is sedimentologically and geochemically distinct from carbonate in sepentinized
shear zone and conduit rocks with d13C = 4 ±
2%, which probably had strong affinity to a deepseated mantle source. Veizer et al. [1989a] also
analyzed sedimentary carbonates in Archean
greenstone belts and found mean d13C value of
+1.5 ± 1.5% [Veizer et al., 1989a]. The carbonates
formed from OCC, and the sedimentary carbonates, are isotopically different at the 99.9% level.
These values imply therefore a Ds 2.8%. The
diagenetic alteration of Precambrian carbonates
studied by Veizer et al. [1989a, 1989b] is thought
to be small on the basis of trace element and
strontium isotope considerations. However, progressive recrystallization during hydrothermal
5 of 9
Geochemistry
Geophysics
Geosystems
3
G
bjerrum and canfield: burial history of organic carbon
circulation has led to continues loss of Sr and a
small increase in the d13C values in OCC carbonates. The sedimentary carbonates have also apparently experienced a small decrease in d13C
values upon alteration. Thus the estimated Ds is
conservative and originally could have been larger
(i.e., more negative).
[13] Typical micritic siderites in deep-water BIF
(banded iron formation) facies are also isotopically
depleted with d13C values around 5%, ranging up
to around 3% [Beukes et al., 1990; Kaufman et al.,
1990]. In contrast, contemporaneous shallow-marine carbonates have d13C values mostly around
0%, suggesting Ds values of between 3 to 5%
at this time, approximating the Ds value indicated
from averages of carbonatized Archean rocks.
[14] Various processes could have contributed to a
large Ds in the Archean and early Proterozoic. A
large Ds could have resulted from an equally large
vertical isotope gradient in marine DIC. The factors
establishing such a large vertical isotope gradient
are not well understood but could include reduced
ocean circulation and/or a more efficient biological
isotope-organic pump. These would act to increase
the accumulation of carbon oxidation products in
the deep ocean reservoir leading to a larger vertical
isotope DIC gradient [Broecker and Peng, 1982].
For example, today, DIC in deep waters of the
stratified Black Sea and Kyllaren Fjord are, respectively, depleted by 6% and 15% relative to
surface waters [Fry et al., 1991; Smittenberg et al.,
2004]. The oxidation of 13C-depleted organic matter from chemoautotrophic carbon fixation at the
chemocline [Hayes et al., 1999], or from the
oxidation of methane across the chemocline
[Hayes, 1994], could also help establish a larger
vertical isotope gradient in DIC. The processes
acting to establish a large vertical isotope gradient
in DIC in the Archean would, however, need to
have been quite pronounced. This is because
higher marine DIC concentrations would have
been associated with higher atmospheric CO2 concentrations, and this would decrease the importance of the organic pump relative to DIC mixing
[Kump and Arthur, 1999]. Other aspects of the
carbon cycle, admittedly speculative, may potentially have contributed to the removal of 13Cdepleted DIC during OCC.
[15] Thus Rothman et al. [2003] proposed that
Precambrian anoxic waters may have accumulated
very high concentrations of dissolved organic carbon (DOC) compared to present-day oxic marine
bottom water. They argued that the dissolved
10.1029/2004GC000713
fermentation products of organic matter settling
into an anoxic deep ocean would degrade slowly
allowing DOC to accumulate to higher concentrations. We speculate that elevated concentrations of
DOC would circulate through ocean crust where
microbial fermentation, and high temperature
decarboxylation of organic compounds with magnetite, would result in the liberation of CO2, CH4,
and other organic decarboxylation products
[McCollom and Seewald, 2003]. Elevated CO2
could precipitate as calcite or siderite depending
on fluid chemistry and represent a sink for carbon.
The prospect would have existed, therefore, for the
introduction of 13C-depleted DIC into altered
ocean crust, although the magnitude of this process
is difficult to constrain.
[16] The direct extraction of mantle CO2 into
carbonatized basalts during OCC is expected to
be minimal, as in more recent ocean crust such
carbonate generally have precipitated in equilibrium
with seawater DIC [Alt, 1999; Alt and Teagle, 2003].
By contrast, in recent basalts organic carbon oxidation can contribute 13C-depleted carbon to hydrothermal carbonates when these are intercalated
with sediments [Alt and Teagle, 2003]. A similar
process could have introduced 13C-depleted carbon
into ancient OCC carbonates. However, this still
amounts to the introduction of 13C-depleted carbonate into the crust and it would influence the isotope
systematics of the ocean system in the same manner
as outlined above.
[17] We obviously have an incomplete understanding of what controlled Ds early in Earth history.
Processes such as deep ocean siderite precipitation,
and in particular OCC, acted as probably significant removal pathways of DIC. Also, direct isotope
observations suggests that the deep carbonates
were isotopically distinct from surface water carbonates with a likely isotope difference, Ds, of 3
to 5% in the Archean and early Proterozoic
oceans. This isotope difference would have generated a sink with the possibility of influencing the
isotopic composition of DIC in the surface reservoir independently of organic carbon burial. In
what follows we explore the evolution of the f
ratio through the Precambrian with a variety of Ds
values ranging from today’s value of 2%, to
5%, and finally to an absolute maximum value
of 7.5% (Figures 3c and 4c).
[18] We begin with the most conservative scenario
yielding the smallest isotope effect on the f ratio.
This is the scenario assuming a methane-rich
atmosphere and constant present-day continental
6 of 9
Geochemistry
Geophysics
Geosystems
3
G
bjerrum and canfield: burial history of organic carbon
size, giving the lowest values for l in the Archean
(Figure 3b). In this case f ratios straddle 0.07 for
most of the Archean with a Ds of 5%, are as low
as 0.05 with Ds of 7.5%, and are around 0.1 for a
Ds of 2% (Figure 3d). If we assume continental
growth through early Earth history much higher
Archean values for l are generated, yielding much
lower f ratios (Figure 4d). In this case f ratios are
around zero in the Archean for Ds < 5%, and are
still substantially reduced, compared to the case
without OCC, with modern Ds = 2%. Importantly, for all l and Ds values explored, the f ratios
are much lower during the Archean and early
Proterozoic than at present ( f 0.2), implying
low proportions of total carbon buried as organic
carbon.
[19] There is a pronounced positive excursion in
the f ratio between 2.4 and 2.0 Ga, as has been
previously noted [Des Marais et al., 1992; Knoll et
al., 1986; Karhu and Holland, 1996]. Our results,
however, reduce the magnitude of the excursion
compared to what is calculated in the absence of
deep ocean carbonate removal (l = 0; upper
dashed curve, Figures 3d and 4d) by amounts
depending on Ds and the evolution of l. Relatively
high f ratios for the Neoproterozoic are also
preserved with the absolute value depending,
again, on the model scenario.
4. Perspective
[20] There is a long-standing opinion that the f ratio
through Precambrian was generally similar to
today [Schidlowski, 1988; Catling et al., 2001;
Watanabe et al., 1997; Holland, 2002]. This is
not supported by our results [see also Des Marais
et al., 1992]. However, significant removal of
isotopically depleted DIC from deeper ocean
waters can have a large influence on the f ratio,
and it is likely that the f ratio was less than half the
present ratio during nearly all of the Archean, and
could have been as low as zero for some of the
time. Low f ratios mean less organic carbon burial,
and imply relatively lower rates of oxygen release
to the surface environment [Holland, 1984; Lenton
and Watson, 2000]. Low rates of organic carbon
burial could have acted alone, or more likely
together, with high fluxes of mantle-derived
reduced gases [Kump et al., 2001; Holland,
2002] to explain the low concentrations of O2
in the Archean atmosphere [Holland, 1999;
Canfield et al., 2000; Farquhar et al., 2000]. While
including OCC into our model reveals lower
Archean f ratios than previously recognized,
10.1029/2004GC000713
marine phosphorus limitation could have been
one of the factors contributing to low f ratios
[Bjerrum and Canfield, 2002].
[21] As pointed out by Des Marais [2000] and
Towe [1990], even modest organic carbon burial
depends on the activities of oxygen-producing
organisms and a carbon cycle involving aerobic
respiration. Thus, while our analysis involves no
assumptions about the presence of oxygenproducing or oxygen-utilizing organisms, even
f ratios as high as 0.05 would apparently require
their presence [Des Marais, 2001]. By contrast,
our less conservative f ratio results, with values
straddling zero (Figure 4d), make no predictions
about the presence of cyanobacteria in the early
Archean environment, and such low f ratios could
conceivably be produced independent of oxygenic
photosynthesis.
[22] We view this contribution as a starting point.
Clearly the consideration of deep ocean DIC
precipitation, and in particular OCC, in global
carbon cycle modeling has important consequences for our interpretation of f ratios in the past.
This becomes extremely important in the early
Precambrian when OCC was likely most significant. Our model results depend critically on Ds
(the isotope difference between OCC carbon and
shallow water derived-inorganic carbon), and the
history of l (the proportion of total carbon
removed by OCC). These values will be better
constrained with focused efforts to better identify
the magnitudes of OCC preserved in ancient
ocean crust, and the magnitude of past differences between OCC and shallow sedimentary
carbonate. Likewise such studies will help identify the processes responsible for the isotopic
composition of OCC carbonates.
Notation
OCC ocean crust carbonatization.
f fraction of total carbon buried as organic
carbon.
d13C isotopic composition of carbon reported
relative to the PDB standard. d13C =
(Rsample/RPDB 1) 1000, where R =
13 12
C/ C is the isotopic abundance ratios.
13
d d C in abbreviated form.
DIC dissolved inorganic carbon.
l fraction of total inorganic carbon removed
by OCC.
7 of 9
Geochemistry
Geophysics
Geosystems
3
G
bjerrum and canfield: burial history of organic carbon
Ds isotopic difference between inorganic carbon removed by OCC and carbonate
precipitated from the surface ocean water,
%.
Ga giga-annum before present.
Acknowledgments
[23] We thank D. J. Des Marais for allowing us use his d13C
database, N. H. Sleep for sharing unpublished model results on
ocean crust carbonatization though time, and T. M. McCollom
and M. T. Rosing for insight in decarboxylation, serpentinization, and hydrothermal systems. We furthermore thank two
reviewers for very helpful comments on an earlier version of
the manuscript and two journal reviewers for helpful suggestive comments. This work was funded by the Danish National
Research Foundation (Danmarks Grundforskningsfond) and
Danish National Science Foundation (SNF).
References
Alt, J. C. (1999), Very low-grade hydrothermal metamorphism
of basic igneous rocks, in Low-Grade Metamorphism, edited
by M. Frey and D. Robinson, pp. 169 – 201, Blackwell, Malden, Mass.
Alt, J. C., and D. A. H. Teagle (2003), Hydrothermal alteration
of upper oceanic crust formed at a fast-spreading ridge:
Mineral, chemical and isotopic evidence from ODP Site
801, Chem. Geol., 201, 191 – 211.
Beukes, N. J., C. Klein, A. J. Kaufman, and J. M. Hayes
(1990), Carbonate petrography, kerogen distribution, and
carbon and oxygen isotope variations in an early Proterozoic
transition from limestone to iron-formation deposition:
Transvaal Supergroup, South Africa, Econ. Geol., 85,
663 – 690.
Bjerrum, C. J., and D. E. Canfield (2002), Ocean productivity
before about 1.9 Gyr ago limited by phosphorus adsorption
onto iron oxides, Nature, 417, 159 – 162.
Brady, P. V., and S. R. Gislason (1997), Seafloor weathering
controls on atmospheric CO2 and global climate, Geochim.
Cosmochim. Acta, 61, 965 – 973.
Broecker, W. S., and T.-H. Peng (1982), Tracers in the Sea,
Lamont-Doherty Earth Observatory, Palisades, N. Y.
Canfield, D. E., K. S. Habicht, and B. Thamdrup (2000), The
Archean sulfur cycle and the early history of atmospheric
oxygen, Science, 288, 658 – 661.
Catling, D. C., K. J. Zahnle, and C. P. McKay (2001), Biogenic
methane, hydrogen escape, and the irreversible oxidation of
early Earth, Science, 293, 839 – 843.
Des Marais, D. J. (1985), Carbon exchange between the
mantle and crust and its effect upon the atmosphere:
Today compared to Archean time, in The Carbon Cycle
and Atmospheric CO2: Natural Variations Archean to
Present, Geophys. Monogr. Ser., vol. 30, edited by E. T.
Sundquist and W. S. Broecker, pp. 602 – 611, AGU,
Washington, D. C.
Des Marais, D. J. (1997), Isotopic evolution of the biogeochemical carbon cycle during the Proterozoic Eon, Org. Geochem., 27, 185 – 193.
Des Marais, D. J. (2000), When did photosynthesis emerge on
Earth?, Science, 289, 1703 – 1705.
10.1029/2004GC000713
Des Marais, D. J. (2001), Isotopic evolution of the biochemical
carbon cycle during the Precambrian, in Stable Isotope Geochemistry, Rev. Mineral. Geochem., vol. 43, edited by J. W.
Valley and D. R. Cole, pp. 556 – 578, Mineral. Soc. of Am.,
Washington, D. C.
Des Marais, D. J., H. Strauss, R. E. Summons, and J. M. Hayes
(1992), Carbon isotope evidence from the stepwise oxidation
of the Proterozoic environment, Nature, 359, 605 – 609.
Farquhar, J., H. Bao, and M. H. Thiemens (2000), Atmospheric influence on Earth’s earliest sulfur cycle, Science,
289, 756 – 759.
Fry, B., H. W. Jannasch, S. J. Molyneaux, C. O. Wirsen, J. A.
Muramoto, and S. King (1991), Stable isotope studies of the
carbon, nitrogen and sulfur cycles in the Black Sea and
Cariaco trench, Deep Sea Res., Part A, 38 suppl. 2, 1003 –
1019.
Haggerty, S. E. (1999), A Diamond Trilogy: Superplumes,
supercontinents, and supernovae, Science, 285, 851 – 860.
Hayes, J. M. (1983), Geochemical evidence bearing on the origin of aerobiosis: A speculative hypothesis, in Earth’s Earliest
Biosphere: Its Origin and Evolution, edited by J. W. Schopf,
pp. 291 – 301, Princeton Univ. Press, Princeton, N. J.
Hayes, J. M. (1994), Global methanotrophy at the ArcheanProterozoic transition, in Early Life on Earth, Nobel Symp.,
vol. 84, edited by S. Bengtson, pp. 221 – 236, Columbia
Univ. Press, New York.
Hayes, J. M., H. Strauss, and A. J. Kaufman (1999), The
abundance of 13C in marine organic matter and isotopic
fractionation in the global biogeochemical cycle of carbon
during the past 800 Ma, Chem. Geol., 161, 103 – 125.
Holland, H. D. (1984), The Chemical Evolution of the Atmosphere and Oceans, Princeton Univ. Press, Princeton, N. J.
Holland, H. D. (1999), When did the Earth’s atmosphere become oxic? A reply, Geochem. News, 100, 20 – 22.
Holland, H. D. (2002), Volcanic gases, black smokers, and the
Great Oxidation Event, Geochim. Cosmochim. Acta, 66,
3811 – 3826.
Kamber, B. S., A. Greig, R. Schoenberg, and K. D. Collerson
(2003), A refined solution to Earth’s hidden niobium: Implications for evolution of continental crust and mode of core
formation, Precambrian Res., 126, 289 – 308.
Karhu, J. A., and H. D. Holland (1996), Carbon isotopes and
the rise of atmospheric oxygen, Geology, 24, 867 – 870.
Kaufman, A. J., J. M. Hayes, and C. Klein (1990), Primary and
diagenetic controls of isotopic compositions of iron-formation
carbonates, Geochim. Cosmochim. Acta, 54, 3461 – 3473.
Knoll, A. H., J. M. Hayes, A. J. Kaufman, K. Swett, and I. B.
Lambert (1986), Secular variations in carbon isotope ratios
from Upper Proterozoic successions of Svalbard and East
Greenland, Nature, 321, 832 – 838.
Kump, L. R. (1991), Interpreting carbon-isotope excursions—
Strangelove oceans, Geology, 19, 299 – 302.
Kump, L. R., and M. A. Arthur (1999), Interpreting carbonisotope excursions: Carbonates and organic, Chem. Geol.,
161, 181 – 198.
Kump, L., J. Kasting, and M. Barley (2001), Rise of atmospheric oxygen and the ‘‘upside-down’’ Archean mantle,
Geochem. Geophys. Geosyst., 2(1), Paper number
2000GC000144.
Lenton, T. M., and A. J. Watson (2000), Redfield revisited:
2. What regulates the oxygen content of the atmosphere?,
Global Biogeochem. Cycles, 14, 249 – 268.
McCollom, T. M., and J. S. Seewald (2003), Experimental
constraints on the hydrothermal reactivity of organic acids
and acid anions: II. Acetic acid, acetate and valeric acid,
Geochim. Cosmochim. Acta, 67, 3645 – 3664.
8 of 9
Geochemistry
Geophysics
Geosystems
3
G
bjerrum and canfield: burial history of organic carbon
Nakamura, K., and Y. Kato (2002), Carbonate minerals in the
Warrawoona Group, Pilbara Craton: Implications for continental crust, life, and global carbon cycling in the early
Archean, Resour. Geol., 52, 91 – 100.
Pavlov, A. A., J. F. Kasting, L. L. Brown, K. A. Rages, and
K. R. Freedman (2000), Greenhouse warming by CH4 in the
atmosphere of the early Earth, J. Geophys. Res., 105,
11,981 – 11,990.
Rothman, D. H., J. M. Hayes, and R. E. Summons (2003),
Dynamics of the Neoproterozoic carbon cycle, Proc. Natl.
Acad. Sci. U.S.A., 100, 8124 – 8129.
Schidlowski, M. (1988), A 3,800-million-year isotopic record
of life from carbon in sedimentary-rocks, Nature, 333, 313 –
318.
Shields, G., and J. Veizer (2002), Precambrian marine carbonate isotope database: Version 1.1, Geochem. Geophys. Geosyst., 3(6), 1031, doi:10.1029/2001GC000266.
Sleep, N. H., and K. Zahnle (2001), Carbon dioxide cycling
and implications for climate on ancient Earth, J. Geophys.
Res., 106, 1373 – 1399.
Smittenberg, R. H., R. D. Pancost, E. C. Hopmans, M. Paetzel,
and J. S. Sinninghe Damste (2004), A 400-year record of
environmental change in an euxinic fjord as revealed by the
10.1029/2004GC000713
sedimentary biomarker record, Palaeogeogr. Palaeoclimatol.
Palaeoecol., 202, 331 – 351.
Towe, K. M. (1990), Aerobic respiration in the Archean?,
Nature, 358, 54 – 56.
Veizer, J., J. Hoefs, D. R. Lowe, and P. C. Thurston (1989a),
Geochemistry of Precambrian carbonates: II. Archean greenstone belts and Archean sea water, Geochim. Cosmochim.
Acta, 53, 859 – 871.
Veizer, J., J. Hoefs, R. H. Ridler, L. S. Jensen, and D. R. Lowe
(1989b), Geochemistry of Precambrian carbonates: I. Archean
hydrothermal systems, Geochim. Cosmochim. Acta, 53,
845 – 857.
Walker, J. C. G. (1990), Precambrian evolution of the climate
system, Palaeogeogr. Palaeoclimatol. Palaeoecol., 82, 261 –
289.
Wallmann, K. (2001), Controls on the Cretaceous and Cenozoic evolution of seawater composition, atmospheric CO2
and climate, Geochim. Cosmochim. Acta, 65, 3005 – 3025.
Watanabe, K., H. Naraoka, D. J. Wronkiewicz, K. C. Condie,
and H. Ohmoto (1997), Carbon, nitrogen, and sulphur geochemistry of Archean and Proterozoic shales from the Kaapvaal Craton, South Africa, Geochim. Cosmochim. Acta, 61,
3441 – 3459.
9 of 9