PP31D-01 INVITED
Carbon Isotope Record of Ancient Oceans
Commencing with the 80's it became generally accepted that δ13CDIC of the oceans is characterised not only be spatial but also temporal variability. This eventually led to the concept of carbon isotope stratigraphy. The Phanerozoic δ 13C trend, originally defined by whole rock data on carbonates, was subsequently confirmed by measurements based on marine low-Mg calcitic shells (Veizer et al., 1999, Chemical Geology, 161, 59-88). Yet even the shell-based data define only a band of about 3 - 4 ‰ width that reflects both the physical properties of the environment as well as the so-called vital effects of the shell secretion processes. This alone, amplified by uncertainties of dating and correlations, should restrict the utility of carbon isotope stratigraphy as a global (as opposed to local or regional) correlation tool to spikes in excess of the band itself. In the Precambrian, the Neo- and Paleoproterozoic time spans contain carbonate rocks with a large spread, but mostly highly positive, δ 13C values, setting them apart from the rest of the record with "normal" marine signature of ~ 0 ‰. These were interpreted as reflecting the global "Snowball Earth" and/or the "Great Oxidation Event" scenarios. Yet, the Paleoproterozoic carbonates of South Africa (Pretoria-Postmasburg Group) contain 13C-enriched isotope values only in near-shore basins, changing regionally from + 10 ‰ to - 2.5 ± 2.5 ‰ (Frauenstein, 2005, Ph.D. Thesis, Ruhr Universitaet, Bochum). These trends are therefore local/regional rather than global phenomena. Carbonates from the Duitschland formation trend to lighter δ 18O values, from - 3 ‰ to - 20 ‰, towards the Bushveld complex. Concurrently, their δ 13C values decrease from +10 ‰ to +5 ‰, or increase from - 2.5 ‰ to + 5 ‰, also in the direction of the Bushveld complex. Modelling suggests that the converging isotope trends could have resulted from alteration in a rock buffered system at a temperature of about 80°C.
PP31D-02
Carbon Isotope Chemostratigraphy, the Baby and the Bathwater
Secular variations in the carbon isotopic values of carbonate sediments and rocks and their individual components have been applied successfully to problems of stratigraphic correlation and for interpretation of past changes in the global carbon cycle. However, this methodology is not without problems. A major tenet of stable isotope chemostratigraphy involves sampling and analyzing multiple, widely separated sequences, and, if possible, multiple carbon-bearing components (e.g., carbonate and organic carbon) in order to demonstrate a global signal. In some cases, this methodology has been short-circuited in the zeal to reveal a new event or excursion, particularly for time intervals for which adequate sequences are somewhat rare. Likewise, although most carbonate researchers are quite aware of the possible importance of diagenesis, particularly in organic-carbon rich sequences or in shoal-water carbonate sequences with longer-term subaerial exposure events, such overprints commonly go unrecognized or are considered of minor impact. Studies of stable isotope variations in carbonate sequences should always employ textural and geochemical methodologies for detecting and even quantifying diagenesis, if possible. Although some diagenetically overprinted or misinterpreted geochemical data have undoubtedly appeared in the literature, there are many excellent examples of global carbon isotope variations in records expressed in pelagic biogenic carbonate, marine organic carbon, platform carbonates, and terrestrial organic matter. Arguably, one of the best-documented examples is the Cenomanian-Turonian (ca. 93 Ma) positive carbon isotope excursion. The amplitude of the Cenomanian-Turonian carbon isotope excursion is similar among all types of records, but there are subtle pattern differences that arise from differences in sedimentation rate among and within sequences. Organic carbon and carbonate carbon isotope signals also may differ in phasing and amplitude for certain events, which may provide information regarding changes in atmospheric- oceanic carbon isotope variations versus changes in the partial pressure of carbon dioxide in the atmosphere. The early Aptian carbon isotope excursions serve as a key example of this. The platform-basin carbonate carbon isotope correspondence undoubtedly works best during times of "calcite seas," such as during the Cretaceous when the complication of greater dominance of more 13-C enriched aragonite in platform settings, such as during the late Neogene, is generally absent. The baby remains robust. Indeed, carbon isotopes still have great utility for reconstructing water-column isotope gradients, global changes in carbon cycling, and for pattern-based long-distance correlation. However, we do need to be judicious in our choice of samples and to temper our enthusiasm for seeking out and interpreting extreme signals.
PP31D-03
Phanerozoic and Neoproterozoic Negative Carbon Isotope Excursions, Diagenesis and Terrestrialization
Comprehensive data sets of Phanerozoic and late Precambrian carbon isotope data derived from carbonate rocks show a similar positive relation when cross-plotted with oxygen isotope values. The range and slope between the time periods is identical and the processes responsible for the relation have been well documented in Quaternary sediments. These processes include the stabilization of isotope values to ambient meteoric water values during shallow burial and flushing of carbonate sediments. Both data sets show strongly depleted carbon (-9 per mil PDB) and oxygen isotope values that retain seemingly systematic stratigraphic patterns with the Quaternary and Phanerozoic examples that demonstrably record meteroric water values. Similar values and patterns in the Precambrian are interpreted as primary marine in origin with significant implications for an ocean carbon mass balance not possible in the Phanerozoic carbon cycle. A similar compilation of carbonates older than one billion years do not show a relation between carbon and oxygen isotopes, lacking the negative carbon values evident in the younger record. We hypothesize that this difference records the onset of significant organic carbon on the land surface and the alteration of meteoric waters toward Phanerozoic values. We demonstrate the meteoric affinities of Neoproterozoic carbonates containing prominent negative isotope excursions recorded in the Shuram and Wonoka Formations of Oman and South Australia commonly attributed to whole ocean isotope variation. The conspicuous absence of negative carbon isotope values with normal marine oxygenisotope values in the Phanerozoic and Neoproterozic identifies a consistent relation between these time intervals and suggests that, as well accepted in the Phanerozoic, negative carbon isotope excursions less than -3 per mil are not a record of marine processes, but rather the later terrestrial biotic influence on meteoric water values.
PP31D-04 INVITED
Carbon isotope gradients on early Paleozoic platforms
Large positive δ13C excursions occur during sea level lowstands in the Ordovician and Silurian. In some cases, the sedimentary units hosting these excursions are stratigraphically equivalent to known glacial deposits. There has been considerable speculation on the significance of this linkage. Some workers view the inferred carbon cycle perturbations as evidence for a biotic response to the deteriorating climate, which may have invigorated thermohaline circulation and nutrient upwelling, thus increasing ocean productivity and organic carbon burial. Others believe that the excursions were driven by a positive shift in the isotope value of the global carbon weathering flux, reflecting increased weathering of calcareous sediments exposed by eustatic regressions. A closer examination of one of the excursion intervals, the Hirnantian, with examples from three basins on two paleocontinents, reveals a recurrent pattern: excursions in shallower nearshore settings are larger than deeper offshore ones. This implies shore-to-basin gradients in sedimentary δ13C, with values higher in nearshore settings. What gradients signify is a question that bears importantly on interpretations of carbon isotope records in ancient marine settings. If shelf sedimentary gradients reflect original seawater gradients in δ13C values, then local scale carbon cycling is implied, which profoundly affects the way that we 'read' carbon isotope signals in epeiric carbonate successions. In this talk, the evidence for carbon isotope gradients is reviewed, with particular emphasis on the Hirnantian sea level lowstand. A sea level driven, local scale, carbonate weathering model is presented that is consistent with shelf gradients. The main feature of the model is that sea level fluctuations, globally, cause adjustments in the isotope value of the carbonate weathering flux, locally, which in turn causes seawater δ13C values to increase in restricted nearshore settings of epeiric seas. If this interpretation of shelf gradients is correct, the model predicts that: (1) ocean δ13C changes are smaller than the largest excursions recorded in epeiric sea deposits, (2) δ13C excursions may behave like chemostratigraphic sea level curves, and (3) the global secular δ13C curve of seawater is overprinted by local scale carbon cycling processes in epeiric seas, and is unlikely to be a true representation of the ocean's secular evolution during the Paleozoic.
PP31D-05
The Oxidant Budget of Dissolved Organic Carbon Driven Isotope Excursions
Negative carbon isotope values, falling below the mantle average of about -5 per mil, in carbonate phases of Ediacaran age sedimentary rocks are widely regarded as reflecting negative excursions in the carbon isotopic composition of seawater lasting millions of years. These isotopic signals form the basis of chemostratigraphic correlations between Ediacaran aged sections in different parts of the world, and have been used to track the oxidation of the biosphere. However, these isotopic values are difficult to accommodate within limits prescribed by the current understanding of the carbon cycle, and a hypothetical Precambrian ocean dissolved organic carbon (DOC) pool 100 to 1000 times the size of the modern provides a potential source of depleted carbon not considered in Phanerozoic carbon cycle budgets. We present box model results that show the remineralization of such a DOC pool to drive an isotope excursion of the magnitude observed in the geological record exhausts global budgets of free oxygen and sulfate in 800 k.y. These results are incompatible with the estimated duration of late Ediacaran isotope excursions of more than 10 m.y., as well as geochemical and biological indicators that oceanic sulfate and oxygen levels were maintained or even increased at the same time. Therefore the carbon isotope record is probably not a useful tool for monitoring oxygen levels in the atmosphere and ocean. Covariation between the carbon and oxygen isotope records is often observed during negative excursions and is indicative of local processes or diagenetic overprinting.
PP31D-06
Were carbon isotopic gradients in post-snowball oceans inverted?
In the Otavi Group of Namibia, 900 m of platform carbonate strata above the 635-Ma post-glacial cap
dolostone are consistently lighter in δ13C by up to 2.5 per mil than the adjacent 180-320 m of
correlative, >0.6-km-deep, foreslope strat. Assuming the foreslope carbonate was equilibrated with deeper
water, the normal isotopic gradient dynamically sustained by the biological 'pump' was inverted.
Geochronology of isotopically correlative Doushantuo strata in South China suggests that the inverted
gradient lasted for ~2 Myr after the glacial termination. Thereafter, the inverted gradient disappears. A
similar transient inverse gradient follows the older Cryogenian glaciation in Arctic Alaska. We hypothesize
that the inverse gradients reflect high pCO2 in the glacial aftermaths. This had two consequences.
First, the size of the DIC pool was enlarged, reducing the isotopic effect of the biological pump. Second,
isotopic fractionation was strongly temperature-dependent due to the large fractionation between CO2
and CO32- coupled with the dominance of CO2(gas) among carbon species at pH<7.2. A
difference in SST of ~25° between the areas of air-sea equilibration is required to account for a
2.5 per mil gradient in δ13C. This is realized in non-upwelling zones of the southern hemisphere
today, where warm subtropical surface waters are underlain by Antarctic Intermediate Water. For our
hypothesis to be valid, a large meridional temperature gradient must have coexisted with strong CO2
radiative forcing.
http://www.snowballearth.org
PP31D-07
Extreme Carbon Isotope Anomalies of the Proterozoic Eon: Fact or Fiction?
Post-Paleozoic carbon isotope variations constructed from time-series analyses of calcareous microfossils generally pale in comparison with the extreme variations recorded in Paleozoic and Proterozoic aged successions. The latter are primarily preserved in fine-grained inorganic carbonates, which due to their great antiquity and potential for diagenetic alteration have been viewed by some as imperfect recorders of seawater chemistry. In part, this bias stems from the study of Modern carbonate platforms, where significant carbon isotope variations result from vital effects, bioturbation, and sea grass aeration of sediments. In addition, the open framework of biotic reefs allows for the infiltration of diagenetic fluids through lithified platforms. None of these biological issues, however, applies to Proterozoic carbonate accumulations, which are generally fine-grained and pervasively cemented. While geochemical tests of diagenetic alteration have been used for Proterozoic samples (e.g., δ18O, Δδ13C, and Mn/Sr), they have been variably applied and appear as moving targets from basin to basin, if reported at all. Greatest confidence in the validity of carbon isotope trends come from stratigraphic measurements that reveal smooth δ13C variations, and the comparative analysis of multiple, broadly equivalent basins. For both the Neoproterozoic and the Paleoproterozoic these show profound changes in carbon isotope distribution with time, especially related to widespread paleoclimatic and biotic events. High resolution stratigraphic and geochemical studies of carbonate and co-existing organic matter in post-glacial Neoproterozoic cap carbonates from Namibia and Brazil, and of Ediacaran aged successions in the western USA, South China, and elsewhere provide new insight into extremes of the ancient carbon cycle. On the one hand, the standard model of proportional fluxes is consistent with detailed observations of the cap carbonates. However, there is also strong evidence for carbon limitation to photoautotrophs in the immediate glacial aftermath, and of an unusually strong surface-to-deep δ13C gradient. On the other hand, the standard model is difficult to reconcile with the > -10‰ carbon isotope excursions in the Ediacaran Period. These require an additional source of 13C depleted alkalinity, which based on the invariance of the organic carbon isotope record support the idea that Proterozoic seawater was buffered by a proportionally larger DOC pool until very near the end of the eon.
PP31D-08
Alternative Explanations for Variations in the Patterns of Carbon Isotopes in Organic and Inorganic Records
Recent work has shown that variations in the patterns of stable carbon isotopes obtained from platform derived carbonates at locations in the Atlantic Ocean, Indian Ocean, and the Pacific Ocean all show synchronous patterns over the past 10 to 25 Myrs. These patterns however are not related to the global patterns of carbon isotopic variations seen in open oceanic sediments over the same period. This absence of agreement between these two records is believed to be caused by simple changes in the input of materials with differing carbon isotopic compositions originating on shallow-water carbonate platforms. Such a phenomenon is a well established for controlling the carbon isotopic composition of bulk organic material (i.e. terrestrial vs. marine), but has been largely ignored for carbonate materials. In this presentation alternative explanations for variations in the carbon isotopic composition (other than changes in the amount of buried organic carbon) of carbonates are explored. These include differences in the origin of the sediments, diagenesis, changes in the predominant mineralogy precipitated in the oceans throughout geological time, and variations in the pCO2 of the atmosphere.