Tải bản đầy đủ - 0 (trang)
Chapter 7.6 The Marine Carbonate and Chert Isotope Records and Their Implications for Tectonics, Life and Climate on the Early Earth

Chapter 7.6 The Marine Carbonate and Chert Isotope Records and Their Implications for Tectonics, Life and Climate on the Early Earth

Tải bản đầy đủ - 0trang


Chapter 7.6: The Marine Carbonate and Chert Isotope Records

Fig. 7.6-1. Sr isotope evolution of seawater and ocean fluxes (A) and its implications for the relative

influence of continental weathering on ocean composition (B). The seawater 87 Sr/86 Sr curve4 in (A)

has been constructed from the lowermost ratios for each time interval, e.g., lowest 87 Sr/86 Sr ratios

of ca. 2.7 Ga limestones of southern Africa are 0.7011–37 . The river runoff (RR) curve has been

determined by assuming a modern-like 9:11 relationship between Sr input from marine carbonate

weathering (RRC) and silicate weathering (RRS), respectively. (Continued on next page.)

7.6-2. The Sr Isotope Composition of the Early Ocean


Fig. 7.6-1. (Continued.) The RRC curve assumes that carbonate rocks undergoing weathering preserve the isotopic composition of the seawater from which they precipitated and have a skewed age

distribution, with an average age of 250 Ma3 , and so lags seawater 87 Sr/86 Sr. The RRS curve is an

idealistic representation based on predicted crustal evolution (O’Nions et al., 1979); other authors

assume much earlier crustal Rb/Sr differentiation with minimal isotopic evolution5 . Ocean crust

alteration provides less radiogenic Sr to the oceans (MI). The curve in (B) assumes that seawater

87 Sr/86 Sr results from simple binary mixing between RR and MI, and shows that the influence of

continental weathering was less prior to 2.5 Ga, and negligible prior to 3.0 Ga. Literature sources are:

1 Bickle (1994); 2 McCulloch (1994); 3 Peucker-Ehrenbrink and Miller (2006); 4 Shields and Veizer

(2002), Veizer et al. (1999); 5 Kamber and Webb (2001); 6 McCulloch (1994); 7,8 Veizer et al. (1989);

Zachariah (1998).


The Sr isotope composition of modern seawater (87 Sr/86 Srsw ) is known very precisely

at just under 0.7092 (Fig. 7.6-1(A)). Because of the high concentration of Sr (∼8 ppm)

and its long residence time (∼4–8 Myr) in seawater (Holland, 1984), the global ocean is

isotopically homogeneous with respect to Sr, while secular trends in 87 Sr/86 Srsw through

geological time can be attributed to globally integrated changes in the relative influence

of two major Sr sources to the ocean; river runoff (RR) and submarine hydrothermal

exchange of Sr (MI) during ocean crust alteration by seawater (Veizer, 1989). Because

the 87 Sr/86 Sr ratios of these two fluxes are known for modern Earth – ∼0.7124 (Palmer

and Edmond, 1989; Peucker-Ehrenbrink and Miller, 2006) and ∼0.703 (Hofmann, 1997),

respectively – we know that the influence of rivers dominates over that of ocean crustseawater interactions by a ratio of about 2:1. In order to estimate trends in the relative

influence of ‘riverine Sr’ versus ‘mantle Sr’ back through time in the same fashion, we

need to reconstruct not only the 87 Sr/86 Srsw of past oceans, but also the long-term isotopic

evolution of these two compositional end-members, which is easier said than done. The

87 Sr/86 Sr

sw curve is continually being improved (Veizer et al., 1999; Shields and Veizer,

2002), while the isotopic evolution of the upper mantle, for the purpose of the present

discussion, is relatively uncontroversial. However, the riverine Sr flux is more difficult to

reconstruct due to the incongruent leaching behaviour of Rb versus Sr in carbonate and

silicate minerals of the upper crust (Goldstein, 1988), uncertainties regarding the effect of

crustal evolution on continental freeboard and on the 87 Rb/86 Sr of silicate minerals, and the

likelihood of unsystematic changes in the age, and therefore 87 Sr/86 Sr, of rocks undergoing

weathering through time.

Because of the tendency of carbonates to incorporate radiogenic 87 Sr, e.g., from Rb-rich

clay minerals, during post-depositional recrystallisation, lowermost 87 Sr/86 Sr ratios are

most likely to best represent the isotopic ratio of the contemporaneous ocean. In most cases,

this assumption can be confirmed by the systematic correlation between low 87 Sr/86 Sr

ratios and high Sr contents (e.g., Veizer et al., 1989). The Archean 87 Sr/86 Srsw record

(Shields and Veizer, 2002) is based mainly on marine carbonate rocks that were deposited


Chapter 7.6: The Marine Carbonate and Chert Isotope Records

in greenstone belts, and barites that have ambiguous origins. Marine carbonate rocks deposited before about 2.0 Ga exhibit lowermost 87 Sr/86 Sr ratios that approach the isotopic

evolutionary curve of the upper mantle (Fig. 7.6-1(A)), while prior to the Neoarchean, the

two are virtually indistinguishable at current resolution. For example, marine carbonates

from ca. 2.7 Ga greenstone belts in Zimbabwe and Canada yield 87 Sr/86 Sr values as low

as 0.7011–0.7013, respectively (Veizer et al., 1989), which is very similar to 87 Sr/86 Sr estimates for the contemporaneous, partially depleted upper mantle of close to, or slightly

below, 0.7010 (Machado et al., 1986).

All older carbonate rocks so far analysed, with the exception of ca. 3 Ga marbles from

India (Zachariah, 1998) (Fig. 7.6-1(A)), yield consistently radiogenic and highly variable

87 Sr/86 Sr ratios >0.702, most likely indicative of isotopic alteration, while ca. 3.49 Ga

barites from the Warrawoona Group of the Pilbara Craton in Australia are interpreted to

provide maximum 87 Sr/86 Sr constraints for the upper mantle, and minimum 87 Sr/86 Sr constraints for contemporaneous seawater, at 0.7005 (McCulloch, 1994). Despite the scarcity

of firm constraints, it seems likely that 87 Sr/86 Srsw closely paralleled the isotopic evolution of the upper mantle before about 3 Ga, and that the oceans were “mantle buffered”

(Veizer et al., 1982) throughout the Archean due to vigorous circulation of seawater via

submarine hydrothermal systems (McCulloch, 1994; Van Kranendonk, 2006). Although

protocontinents appear to have existed prior to this time (Campbell, 2003), the 87 Sr/86 Sr

isotopic evolution of Rb-enriched crustal materials appears to have had negligible influence

on 87 Sr/86 Srsw until after 2.5 Ga.

With the exponential decline of internal heat dissipation on a cooling Earth, the vigor

of the hydrothermal system also declined, while at the same time the flux of radiogenic

strontium from growing continents brought in by rivers started to assert itself (Veizer et al.,

1982). This tectonically controlled transition from “mantle-” to “river-buffered” oceans

across the Archaean/Proterozoic transition (Fig. 7.6-1(B)) is a first order feature of terrestrial evolution, with consequences for other isotope systematics and for the redox state of

the ocean/atmosphere system (Goddéris and Veizer, 2000; Veizer and Mackenzie, 2003).


The two dominant exogenic reservoirs of carbon are carbonate rock, formed either as

a marine precipitate or during alteration of seafloor ocean crust, and organic matter in

sediments. Carbonate and reduced carbon are linked in the carbon cycle via atmospheric

CO2 and carbon dioxide species dissolved in the hydrosphere. Carbon deposition is balanced by the mantle carbon flux with δ 13 C of about −5❤PDB (Hayes and Waldbauer,

2006) [light stable isotope ratios are commonly expressed in per mil (❤) deviations relative to international standards such as PeeDee Belemnite (PVB or V-PDB) for carbonates

and standard mean ocean water (SMOW or V-SMOW) for waters and minerals]. In the

absence of autotrophic photosynthesis, this would also be the isotopic composition of seawater (Broecker, 1970). However, δ 13 C for the total dissolved carbon (TDC) in modern

seawater is ∼1 ± 0.5❤ (Kroopnick, 1980; Tan, 1988) because of a kinetic isotope effect

7.6-3. The C Isotope Composition of Early Marine Carbonate Rocks


that serves to enrich organic matter in the lighter isotope, 12 C. Most of this enrichment

derives from the isotope-discriminating properties of RuBP carboxylase, the key enzyme

of the Calvin cycle (Schidlowski, 2001). Buried organic matter is consequently enriched

in 12 C by ∼28❤ with respect to buried carbonate rock, which results from a combination

of a 30❤ fractionation between dissolved inorganic carbon (DIC) in surface seawater and

sedimentary organic carbon, and a 2❤ fractionation between DIC and precipitated carbonate (Hayes and Waldbauer, 2006). The former isotopic discrimination has varied through

time between globally integrated mean values of 25❤ and 50❤ (Des Marais et al., 1992),

largely because of changing atmospheric CO2 levels and the changing relative importance

of methanotrophic metabolisms, which are associated with greater fractionations. The latter isotopic fractionation of 2❤ between DIC and carbonate minerals seems likely to hold

for all kinds of marine carbonate minerals as far back as 3.45 Ga (Hayes and Waldbauer,


Because the source and sink C-isotopic fluxes need to be balanced on geological timescales, the average δ 13 Ccarb of ∼0❤ (Schidlowski et al., 1983; Shields and Veizer, 2002)

(Fig. 7.6-2) and δ 13 Corg of ∼ −28❤ imply that carbon has generally been removed from

the exogenic system into the crust as an approximately 9:2 mixture of carbonate minerals

to organic matter, according to the mass balance:

δ 13 Cinflux = forg .δ 13 Corg + (1 − forg ).δ 13 Ccarb


where forg = Corg (Corg + Ccarb ). The marine carbonate and


therefore, strong support for the existence of a global microbial ecosystem throughout the

marine sedimentary record and perhaps back as far as 3.8 Ga (Schidlowski et al., 1983;

Schidlowski, 2001). Archean kerogen shows increasing evidence for metamorphic overprint that, when accounted for by extrapolating back to near primary H/C ratios, indicates

that initial carbonate-kerogen isotopic discrimination was perhaps as high as 50❤ during

the Neoarchean (Fig. 7.6-2). Such low δ 13 C values for kerogen suggest the influence of

methanogenic bacteria (e.g., Ueno et al., 2006), and further imply that the 9:2 ratio was

not constant through time and may even have been as high as 10:1 during parts of the

Neoarchean (Des Marais et al., 1992). How far back in Earth history can carbon isotopes

be used to support the existence of life on Earth?

The oldest, arguably marine, carbonate isotope data derive from the Isua supracrustal

belt (ISB), SW Greenland, which contains the oldest rocks on Earth interpreted to be of

initially sedimentary origin. However, the origin of both carbonate and graphite in the ISB

is the subject of considerable controversy (Schidlowski, 2001; van Zuilen et al., 2003).

For a generation, the δ 13 C values of carbonates and graphites from Isua have been considered to represent the earliest evidence for life on Earth (Schidlowski et al., 1979). These

original interpretations considered that the high and unusually variable δ 13 C values of ISB

graphite, in particular, were related to its complex amphibolite-grade metamorphic history, while biogenic interpretations were based on the assumption that the graphite-bearing

metacarbonate rocks of the ISB were metamorphosed sediments. By contrast, other detailed studies, recently exemplified by van Zuilen et al. (2003), argue that “most of the

graphite in the ISB occurs in carbonate-rich metasomatic rocks (metacarbonates), while

kerogen δ 13 C records


Chapter 7.6: The Marine Carbonate and Chert Isotope Records

Fig. 7.6-2. Plot of age versus δ 13 C of marine carbonates (after Shields and Veizer, 2002) and δ 13 C

of kerogen corrected for thermal alteration (after Des Marais, 2001). Boxes on right-hand axis refer

to Ediacaran and Phanerozoic ranges. δ 13 C carb has averaged about 5❤ higher than mantle carbon

δ 13 C since ca. 3.4 Ga, indicating the continual existence of a significant autotrophic microbial community since that time. Older δ 13 C records from Greenland seem likely to be of non-marine origin

and are shown as open circles (see text for further explanation).

sedimentary units, including banded iron formations and metacherts, have exceedingly low

graphite concentrations”. These observations, together with isotopic arguments, support a

metasomatic origin for “most, if not all” carbonate minerals (calcites, ferroan dolomites

and siderites) from Isua, while all the graphite is now interpreted to derive from the thermal

decomposition of secondary, metasomatic siderite (van Zuilen et al., 2003). Nevertheless,

metasedimentary rocks, such as banded iron formations and metacherts, of the ISB also

contain small amounts of carbonate minerals (<5%) that are not necessarily of metasomatic origin (van Zuilen et al., 2003). δ 13 C compositions of these carbonates are also

highly variable, but range up to normal marine values (−9❤ to +1❤), whereas associated metacherts exhibit high δ 18 O values that are only slightly lower than nonmetamorphic

Archean cherts (Perry et al., 1978; Perry and Lefticariu, 2003).

7.6-3. The C Isotope Composition of Early Marine Carbonate Rocks


Fig. 7.6-3. Crossplot showing stable carbon and oxygen isotopic compositions in ❤PDB of carbonates from the ca. 3.4 Ga Strelley Pool Chert at two sites in W. Australia (after Lindsay et al., 2005).

Three outlying samples are interpreted to have undergone isotopic exchange during replacement; all

other samples show positive correlation that could represent a diagenetic trend, implying that primary

carbonate values were close to +2/+3❤ and −12/−11❤ for δ 13 C and δ 18 O, respectively.

The oldest, undisputed marine sediments derive from the ca. 3.4 Ga Strelley Pool Chert

of the Kelly Group (Van Kranendonk et al., 2003, 2006a) that exhibit δ 13 Ccarb values between −8❤ and +4❤ with an average of +1.0 ± 2.2❤ (Lindsay et al., 2005; Veizer

et al., 1989). Interestingly, δ 13 C data are similar to those from the slightly younger, middle Archean Barberton Greenstone Belt that range between −4❤ and +3❤, averaging

0.2 ± 1.6❤ (Schidlowski et al., 1975; Veizer et al., 1989). The Kelly Group has only been

subjected to lower greenschist-grade metamorphism (Van Kranendonk, 2000); however,

only few studies have reported detailed mineralogical and petrographic information for

samples used in isotope studies. In this regard, data for carbonates in the Strelley Pool Chert

stand out as an exception (Lindsay et al., 2005). In that study, a clear correlation in least

altered samples between high δ 13 C and δ 18 O values could be demonstrated (Fig. 7.6-3),

which could reflect isotopic exchange during fluid overprinting (Hofmann et al., 1999),

possibly at elevated temperatures, or mixing with expulsed C-rich fluids derived from the


Chapter 7.6: The Marine Carbonate and Chert Isotope Records

degassing of CO2 at depth (Lindsay et al., 2005). Whichever scenario is preferred, stratified

ferroan dolostones of the c. 3.40 Ga Strelley Pool Chert appear to be systematically better

preserved, isotopically speaking, than vein-related carbonates, with least altered, presumably near marine values of +2 to +3❤ for δ 13 C and −12 to −11❤ for δ 18 O. The high

δ 13 C ratios of carbonates from the Strelley Pool Chert are consistent with all other Archean

marine carbonate rocks (Fig. 7.6-2), while their close association with stromatolite reef

complexes (Hofmann et al., 1999; Van Kranendonk et al., 2003; Allwood et al., 2006a)

provides strong support for the existence of a microbial community of global extent as

far back as 3.40 Ga, with additional supporting evidence from the Mesoarchean Barberton

Greenstone Belt (Tice and Lowe, 2004). Although available isotopic evidence is consistent

with oxygenic photosynthesis, it does not exclude the possibility that only methanogenic

and anoxygenic photoautotrophic bacteria contributed to the primary microbial biomass at

this time (Schidlowski et al., 1983; Des Marais, 2001).

Existing constraints on primary δ 13 C of middle Archean kerogen average ∼ −36❤ (see

compilations in Hayes et al., 1983; Strauss and Moore, 1992; Hayes and Waldbauer, 2006)

and derive mainly from cherts and shales of the Fig Tree and Onvervacht Groups of the

Barberton Greenstone Belt, which have been metamorphosed to greenschist facies. This

value is similar to the lowest, and presumably least altered, δ 13 C values from cherts of the

Warrawoona Group (Hayes et al., 1983), although similarly depleted carbonaceous material from cherts of the Pilbara Supergroup are interpreted as abiogenic (Lindsay et al., 2005;

Marshall, this volume). Carbonate associated with weathered basalts in the Warrawoona

Group exhibit δ 13 Ccarb values of −0.3❤ (Nakamura and Kato, 2004), similar to contemporaneous, sedimentary marine carbonates. Assuming that these isotopic constraints are

representative, Eq. (1) tells us that about 14% of the carbon being delivered by outgassing

of mantle CO2 was being buried as reduced, organic carbon by 3.4 Ga (Hayes and Waldbauer, 2006).



Oxygen isotopes, 18 O, 17 O and 16 O, undergo fractionation with respect to each other

during many surface processes, such as evaporation, condensation, precipitation and clay

mineral formation. Lower temperatures lead to larger isotopic fractionations, thus allowing the 18 O/16 O ratio of minerals to be used as a paleothermometer, as well as a tracer

of surface processes within the hydrological cycle. However, water is almost ubiquitous

in the surface environment, and so isotopic exchange with fluids after burial is a constant

source of concern when interpreting the δ 18 O values of ancient rocks. The hydrological cycle favours the retention of 16 O in clouds, leading to negative δ 18 O values in rainwater and

snow compared with seawater. Therefore, isotopic exchange with meteoric groundwater

and/or at elevated temperatures will lower the δ 18 O values of marine authigenic precipitates after burial. Conversely, precipitation in a restricted marine environment affected by

elevated evaporation rates leads to enrichment of 18 O in carbonate or chert minerals.

7.6-4. The O Isotope Composition of Early Marine Sedimentary Rocks


The δ 18 O values of marine carbonates and cherts depends on the isotopic composition

of porewaters, and the temperature-dependent fractionation during initial precipitation and

possible recrystallisation to a thermodynamically more stable mineral phase, i.e., calcite

or dolomite in the case of calcium carbonate minerals, or quartz in the case of silica minerals. Other factors such as pH are of only secondary importance, while metabolic or vital

effects can be neglected for the Precambrian. If the isotopic composition of seawater is

known, then the δ 18 O values of well preserved carbonate minerals and chert can be used to

determine the approximate temperature of formation. However, if seawater δ 18 O is unconstrained and/or the pristine, open marine nature of the mineral phase cannot be guaranteed,

then paleotemperature estimates will be ambiguous. This is the basis of the long-standing

controversy regarding the 18 O-depletion of ancient carbonate and chert minerals (Muehlenbachs, 1998; Veizer et al., 1999). One school of thought maintains that seawater δ 18 O has

remained fixed throughout Earth history (Gregory, 1991; Muehlenbachs, 1998) and that

low δ 18 O values must relate to later alteration or higher surface temperatures in the past

(Knauth and Lowe, 2003). The opposing viewpoint (Veizer et al., 2000; Kasting et al.,

2006) considers that seawater δ 18 O can change through time, implying that δ 18 O-based

paleoclimate studies (Knauth and Epstein, 1976; Knauth and Lowe, 1978, 2003; Robert

and Chaussidon, 2006) systematically overestimate past ocean temperatures (Wallmann,


It has long been noted that early Precambrian cherts yield anomalously low δ 18 OSMOW

values compared with today (Fig. 7.6-4). This has been interpreted in terms of higher ocean

temperatures of 75 ◦ C or more during the Archean (Knauth and Epstein, 1976; Knauth and

Lowe, 1978, 2003). Most of these data derived from the ca. 3.4 Ga Onvervacht Group of the

Barberton Greenstone Belt, South Africa (Fig. 7.6-4), while coverage through the rest of

the Precambrian was, and still is, sparse (Perry and Lefticariu, 2003; Robert and Chaussidon, 2006). New data (Knauth and Lowe, 2003) from the Swaziland Supergroup are

consistent with previous results (Perry, 1967; Knauth and Lowe, 1978), in that the highest δ 18 O value found in these cherts (+22❤SMOW ) is about 8❤ lower than the lowest

values of “representative” shallow marine cherts of the Devonian, while Barberton Greenstone Belt cherts are on average depleted in 18 O by about 10❤ compared with recent cherts

(Perry and Lefticariu, 2003). By comparison, metacherts of the ca. 3.8 Ga Isua supracrustal

belt show even lower maximal δ 18 O values, of +20❤SMOW (Perry et al., 1978) and are at

higher metamorphic grade than their Barberton (and Pilbara) counterparts. If seawater δ 18 O

during the early-mid Archean was comparable to today (Muehlenbachs, 1998), and these

maximal δ 18 O values represent an open marine isotopic signature, then such low δ 18 O

implies that early oceans were hot, with ambient temperatures of between 55 and 90 ◦ C

(Perry and Lefticariu, 2003).

Additional support for hot early oceans has come from a silicon isotope study of marine cherts (Fig. 7.6-4). In their study, Robert and Chaussidon (2006) report a positive

correlation between δ 18 O and δ 30 Si in Precambrian cherts. Unlike δ 18 O, δ 30 Si values in

silica do not so much depend on temperature as on the isotopic composition of the medium

from which the chert precipitated. The authors argue that seawater δ 30 Si is controlled by

the difference between the temperature of the oceans and that of hydrothermal fluids. The


Chapter 7.6: The Marine Carbonate and Chert Isotope Records

Fig. 7.6-4. Paleotemperature estimates based on the assumption that seawater δ 18 O has remained

unchanged throughout Earth history. Maximum chert δ 18 O line is from Knauth and Lowe (1978)

and Robert and Chaussidon (2006); mean carbonate δ 18 O line is from data in Shields and Veizer

(2002) and Jaffrés (2005); δ 30 Si constraints are from Robert and Chaussidon (2006). Mid-Archean

marine cherts and carbonates fit the long-term trend of decreasing lower δ 18 O with age, and imply

implausibly hot climates before 450 Ma (see text for explanation).

observed trend towards higher δ 30 Si values between 3.5 and 0.8 Ga could therefore be interpreted as reflecting a progressive decrease in ocean temperature of approximately the

same magnitude as that discerned from the δ 18 O record. Conversely, this correlation could

instead reflect a progressive increase in the globally integrated mean temperature of hydrothermal alteration (Shields and Kasting, 2007), in which case low chert δ 18 O values

could simply reflect lower seawater δ 18 O.

The use of cherts to reconstruct past ocean temperature is controversial as the open marine environment of formation of many Precambrian cherts has been questioned (de Wit et

al., 1982; Perry and Lefticariu, 2003). In this regard, it may not be meaningful to compare

directly the isotopic compositions of (1) greenstone-associated early diagenetic cherts of

the early and middle Archean, some of which have been interpreted to be of replacive,

hydrothermal origin (e.g., Van Kranendonk, 2006), (2) iron formation-associated early diagenetic chert bands formed between 2.5 and 1.8 Ga, or (3) peritidal, evaporite-associated

cherts of the Meso- and Neoproterozoic, with predominantly biomediated pelagic cherts

7.6-4. The O Isotope Composition of Early Marine Sedimentary Rocks


of the Phanerozoic. Despite such reservations, the general 10❤ depletion of Archean

cherts is mirrored by an equivalent isotopic depletion in contemporaneous and most later

Precambrian marine carbonates (Shields and Veizer, 2002), suggesting that neither later

alteration nor restricted settings was the primary factor controlling the controversial δ 18 O

trend (Kasting and Ono, 2006).

Like cherts, carbonates commonly show a wide range of δ 18 O values (Fig. 7.6-5), which

relates to later alteration by meteoric fluids, isotopic exchange at high temperatures, as well

as precipitation in 18 O-enriched evaporitic settings. However, marine carbonates are less

ambiguous than cherts in terms of their depositional environment and possible hydrothermal overprint, because open marine signatures can be confirmed from Sr and C isotope

studies on the same samples. In addition, some early diagenetic, low-Mg calcite cement

types, such as molar-tooth structure, can be traced through two billion years back into the

Archean (Shields, 2002), thus providing confidence that isotopic trends are meaningful. In

several cases it can be demonstrated that evaporite-related carbonates are systematically

enriched in 18 O (e.g., Kah, 2000; Bau and Alexander, 2006), indicating that environmental

trends are preserved, and that high, outlying δ 18 Ocarb values are not necessarily representative of the global ocean. As with the chert record, ancient marine carbonates are known

to be anomalously depleted with respect to their more recent counterparts; however, the

δ 18 O trends of these two mineral groups are quite different (Fig. 7.6-4). In the carbonate record, low δ 18 O values are found throughout the period before 450 Ma (Shields and

Veizer, 2002), while the chert record shows modern day-like δ 18 O and δ 30 Si values already

during the Mesoproterozoic (Knauth and Lowe, 1978), and Paleoproterozoic (Robert and

Chaussidon, 2006).

Well preserved, low-Mg calcite brachiopod shells (Veizer et al., 1999) and early diagenetic, low-Mg calcite cements (Johnson and Goldstein, 1993) of the early Phanerozoic

exhibit the lowest δ 18 O values of the Phanerozoic. Highest δ 18 O values of these demonstrably well-preserved marine minerals are generally between 0–5❤ 18 O-enriched relative to

highest values for early marine cements (molar-tooth structure) of the Archean and Proterozoic (Fig. 7.6-5). Because mean ocean temperatures could not realistically have exceeded

35 ◦ C since the evolution of vertebrates, low carbonate δ 18 O during the early Phanerozoic has been interpreted as primarily due to increasing seawater δ 18 O from −5❤SMOW

since 500 Ma, with temperature a secondary factor only (Veizer et al., 2000). Some wellpreserved Archean limestone units (Abell et al., 1985a, 1985b; Bishop et al., 2006) even

exhibit δ 18 O values (Fig. 7.6-5) that are mostly within the normal range of well-preserved,

early Ordovician brachiopod shells (Shields et al., 2003), which is inconsistent with any

major decrease in surface temperatures since at least 2.9 Ga. If low seawater δ 18 O persisted throughout the entire Precambrian, then temperatures are unlikely to have reached

far outside the range of Phanerozoic oceans. An apparent maximum depletion of 5❤ from

the Neoarchean to the Cambrian/Ordovician (cf. Shields et al., 2003; Bishop et al., 2006)

implies a maximal temperature difference of ∼20 ◦ C. In other words, ocean temperatures

higher than about 55 ◦ C seem to be highly unlikely during the late Archean. But what about

the pre-3.0 Ga world?


Chapter 7.6: The Marine Carbonate and Chert Isotope Records

Fig. 7.6-5. Marine limestone and dolostone δ 18 O plotted as mean values of data from stratigraphic

units with sample size >2 (Shields and Veizer, 2002; Jaffrés, 2005). 98 sample sets characterized

as clearly altered or from clearly evaporitic settings have been removed from a total of 785. Boxes

with filled circles refer to the range and mean δ 18 O values, respectively, of early marine, originally

low-Mg-calcite cements (Johnson and Goldstein, 1993; Bishop et al., 2006; Shields, unpubl. data);

open circles are from representative isotopic studies of Archean carbonate rocks (Abell et al., 1985a;

Lindsay et al., 2005). With the exception of intervals of cool climate, calcites and dolomites of the

Precambrian and early Phanerozoic are up to 5–10❤ depleted in 18 O with respect to later Phanerozoic and modern equivalents (Veizer et al., 1999; Shields and Veizer, 2002; Shields et al., 2003). The

persistence of low δ 18 O into the early Ordovician implies however that changing seawater 18 O/16 O

and not temperature is likely to have been the major controlling factor over this trend. This implies

that surface temperatures were not as extreme on the early Earth as implied in Fig. 7.6-4, and most

likely lower than 55 ◦ C since 3.5 Ga.

The existing early-mid Archean δ 18 O record is unfortunately sparse and is rendered

less useful by the frequent absence of mineralogical and petrographic details in publications. Nevertheless, if we consider the previously discussed constraints on pristine

marine dolomite δ 18 O from the ca. 3.4 Ga Strelley Pool Chert (Lindsay et al., 2005) of

Tài liệu bạn tìm kiếm đã sẵn sàng tải về

Chapter 7.6 The Marine Carbonate and Chert Isotope Records and Their Implications for Tectonics, Life and Climate on the Early Earth

Tải bản đầy đủ ngay(0 tr)