Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Cooper, A. K., P. J. Barrett, H. Stagg, B. Storey, E. Stump, W. Wise, and the 10th ISAES editorial team, eds. (2008). Antarctica: A Keystone in a Changing World. Proceedings of the 10th International Symposium on Antarctic Earth Sciences. Washington, DC: The National Academies Press. A Pan-Precambrian Link Between Deglaciation and Environmental Oxidation T. D. Raub and J. L. Kirschvink1 ABSTRACT were to suddenly shift to even a 5-10 percent lower luminos- ity, our oceans would rapidly freeze over. We infer that this Despite a continuous increase in solar luminosity to the pres- climatic regulation is due in large part to a combination of ent, Earthâs glacial record appears to become more frequent, greenhouse gassesâprincipally H2O, CO2, and CH4âwhich though less severe, over geological time. At least two of the have varied over time. For one of these, CO2, there is a clear three major Precambrian glacial intervals were exceptionally inorganic feedback mechanism helping regulate climate intense, with solid evidence for widespread sea ice on or near (Walker et al., 1981), as CO2 removal by silicate weathering the equator, well within a âSnowball Earthâ zone produced increases with temperature, a process that can act on a 106- to by ice-albedo runaway in energy-balance models. The end 107-year timescale. of the ï¬rst unambiguously low-latitude glaciation, the early Geologists observe that a major shift in redox state of Paleoproterozoic Makganyene event, is associated intimately Earthâs atmosphere happened sometime between 2.45 and with the ï¬rst solid evidence for global oxygenation, includ- 2.22 Ga ago, as signaled by the loss of a mass-independent ing the worldâs largest sedimentary manganese deposit. fractionation signal in sulfur isotopes, the disappearance of Subsequent low-latitude deglaciations during the Cryogenian common detrital pyrite and uraninite from stream deposits, interval of the Neoproterozoic Era are also associated with and the appearance of true continental redbeds, documented progressive oxidation, and these young Precambrian ice ages by a reworked paleosol that cements together coherent hema- coincide with the time when basal animal phyla were diver- titic chips magnetized in random directions (Evans et al., sifying. However, speciï¬cally testing hypotheses of cause 2001). The sedimentary sulfate minerals barite and gypsum and effect between Earthâs Neoproterozoic biosphere and also become more prevalent in evaporative environments glaciation is complicated because large and rapid True Polar post ~2.3 Ga, as seen in the Barr River Formation of the Wander events appear to punctuate Neoproterozoic time and Huronian Supergroup of Ontario (see Figure 1). may have episodically dominated earlier and later intervals The reappearance of sedimentary sulfates after the Gow- as well, rendering geographic reconstruction and age corre- ganda and Makganyene Glaciations at about 2.2 Ga follows lation challenging except for an exceptionally well-deï¬ned a nearly 800 myr absence in the rock record (Huston and global paleomagnetic database. Logan, 2004), arguing that enough oxygen was then present in the atmosphere to oxidize pyrite to sulfate in quantities that INTRODUCTION sulfate-reducing organisms could not completely destroy. Numerous hints in the rock record suggest a general Despite a 30 percent increase in solar luminosity during the relationship between changes in atmospheric redox state past 4.6 billion years, we have solid geological evidence that and severe glaciation. Most dramatically, the sedimentary liquid water was usually present on the surface. If the sun package deposited immediately after the Paleoproterozoic low-latitude Makganyene glaciation in South Africa contains 1 Division of Geological and Planetary Sciences, California Institute of a banded iron formation-hosted manganese deposit that is the Technology, Pasadena, CA 91125, USA. richest economic unit of this mineral known on Earth; Mn 83
84 ANTARCTICA: A KEYSTONE IN A CHANGING WORLD FIGURE 1 Gypsum casts, mud cracks, and ripples from the Barr River Formation north of Elliot Lake, Ontario, Canada. can only be precipitated from seawater by molecular oxy- peculiarities of low-latitude tillites, BIFs, abrupt and broadly gen (Kirschvink et al., 2000; Kopp et al., 2005). Similarly, synchronous glacial onset and termination, and many other Neoproterozoic glacial events are associated with apparent features of these events (Evans, 2000; Hoffman, 2007; Hoff- bursts of oxygenation and may have stimulated evolutionary man and Schrag, 2002; Hoffman et al., 1998). No alternative innovations like the Ediacara fauna and the rise of Metazoa. hypothesis even attempts to explain as many diverse features We argue here that Precambrian glaciations are generally of the Precambrian glacial record. followed by ï¬uctuations in apparent redox parameters, con- Initially, the most fundamental result driving the Snow- sistent with a postulate by Liang et al. (2006) that signiï¬cant ball Earth hypothesis was a soft-sediment fold test on a quantities of peroxide-generated oxidants are formed and varvite-like member of the ~635 Ma Marinoan-age Elatina released through glacial processes. formation in South Australia, which implied incursion of sea ice into subtropical latitudes (Figure 2) (Sumner et al., 1987). A few years later, Evans et al. (1997) demonstrated LOW-LATITUDE GLACIATION AS A SNOWBALL EARTH similarly robust results from the ~2.22 Ga Makganyene gla- Despite assertions to the contrary (Lovelock, 2006), climatic ciation in South Africa, indicating that at least two intervals regulatory mechanisms have not always maintained large of geological time, separated by more than a billion years, open areas of water on Earthâs surface. Substantial evidence experienced low-latitude glaciation. Comparison of less exists that large-scale continental ice sheets extended well robust paleomagnetic data for all Precambrian glaciations into the tropics, yielding sea ice at the equator (Embleton with well-documented paleolatitudes for Phanerozoic gla- and Williams, 1986; Evans et al., 1997; Sohl et al., 1999; cial deposits yields an interesting schism. With the possible Sumner et al., 1987). The deposition of banded iron oxide exception of the Archean Pongola event, there is a total formations (BIFs) associated with glacial sediments implies absence of evidence for polar or subpolar glaciation through- both sealing off of air-sea exchange and curtailing the input out the Precambrian, while marine glacial sedimentation of sulfate to the oceans, which otherwise would be reduced never breaches the tropics through the Phanerozoic (Evans, biologically to sulï¬de, raining out Fe as pyrite. The Snow- 2003). While the counterintuitive Precambrian polar glacial ball Earth hypothesis (Kirschvink, 1992) accounts for the gap must be largely an artifact of the paleogeographic and
RAUB AND KIRSCHVINK 85 A B N W E Before Fold Correction S N C W E After Fold Correction S FIGURE 2 Soft-sediment paleomagnetic fold test on the rhythmite member of the Elatina Formation, Pichi-Richi Pass, Australia. The initial paleomagnetic study on this member by Embleton and Williams (1986) displayed a nearly equatorial remanent magnetization held in hematite, but lacked a geological ï¬eld test to verify that the characteristic magnetization was acquired at or near the time of deposition. As part of the Precambrian Paleobiology Research Group (PPRG) at the University of California, Los Angeles, in 1986, Bruce Runnegar provided J. L. K. with an oriented block sample of this unit (Figure 2A), which displayed an apparent soft-sediment deformation feature. Careful subsampling and demagnetization of this block by then undergraduate student Dawn Sumner (now at the University of California, Davis) revealed a horizontally aligned, elliptical distribution of directions consistent with the earlier result (Figure 2B). However, correction for the bedding deformation signiï¬cantly tightened the distribution, making it Fisherian and passing the McElhinny (1964) fold test at P <0.05. This result, along with an equally interesting result from a layer deformed by a glacial drop stone in the Rapitan Banded Iron Forma- tion of Canada, was published as an American Geophysical Union abstract (Sumner et al., 1987); this led directly to the Snowball Earth Hypothesis (Kirschvink, 1992), and had the desired effect of stimulating further studies conï¬rming the primary, low-latitude nature of the Elatina glacial event (Sohl et al., 1999; Williams et al., 1995). rock preservation records (Evans, 2006), the data consensus intervals dominated by dispersal of cratonic fragments points to an anomalously severe glacial mode in Proterozoic from previous supercontinents (Kenorland and Rodinia, time relative to the Phanerozoic Era. respectively), all Phanerozoic glacial events appear related Evans (2003) suggests that this shift in Earthâs glacial to episodes of continental amalgamation. (Possible Ordovi- mode reflects the evolution of macroscopic continental cian glaciation could mark the formation of Gondwanaland; life, especially of lichen and fungi through the Ediacaran- Carboniferous-Permian glaciation marks the assembly of Cambrian transition (see also Peterson et al., 2005). Such Pangea; and the Miocene-present glacial epoch arguably organisms might modulate the silicate-weathering feedback presages the formation of a future supercontinent termed to disfavor climate extremes, although the specifics of âSuperAsiaâ after the likely centroid of amalgamation.) whether endolithic organisms promote or hinder physical The characteristic length-scale of each supercontinent and chemical weathering is surprisingly still ambiguous (see was centered at the âequatorâ and spread, as a yellow Beerling and Berner, 2005). box, over the lifespan of that supercontinent. Blue wax- This fundamental Precambrian-Phanerozoic shift in ing triangles indicate intervals of dominant supercontinent Earthâs glacial mode also appears to manifest itself in the amalgamation, and red waning triangles indicate intervals relation of glacial events to a plate-tectonic supercontinent of dominant supercontinent fragmentation and dispersal. A cycle. Figure 3 relates a simpliï¬ed compilation of Earthâs purple zone between the Paleoproterozoic supercontinent, glacial record to a schematic representation of Earthâs super- Nuna, and the Mesoproterozoic-Neoproterozoic supercon- continents through time. Whereas the Paleoproterozoic and tinent, Rodinia, indicates basic uncertainty as to whether Neoproterozoic low-paleolatitude glacial events occupied Nuna broke apart and reassembled into Rodinia, or whether
86 ANTARCTICA: A KEYSTONE IN A CHANGING WORLD Supercontinents and Glaciations Gp El o Po SP-H M Gb Ed Gk Go Pe Pl 90 Length-Scale of Continents o 60 o 30 Super Asia P Gondwana FIGURE 3 Character of glaciations and Kenorland a Rodinia Nuna plate tectonics versus Earth history, mea- o n 0 g sured in geons (100-million-year blocks of e geological time). Surface areas for each of o a the demonstrated or likely supercontinents 30 in Earth history were estimated, and a characteristic length scale for each super- 60 o continent was deï¬ned as the square root of its surface area, converted from kilometers o Past Future to degrees of arc. The vertical axis repre- 90 sents a characteristic meridian on Earth, 30 25 20 15 10 5 0 -5 running from 90 degrees north latitude to 8 90 degrees south. Time x 10 years (Geons) a single supercontinent simply grew monotonically over that cal box model called GEOCARBSULF (Berner, 2006) pre- interval. The future supercontinent SuperAsia is predicted to dicts monotonic increases in atmospheric oxygen concentra- begin its formal lifespan ~250 million years from now, when tion spanning the late Ordovician, Carboniferous-Permian, the oceanic lithosphere at the edge of the Atlantic ocean will and Miocene-present intervals of geologic time (for a recent have reached a foundering density and produced subduc- discussion of the Paleozoic data, see Huey and Ward, 2005), tion zones for enough time to reunite South America with with precipitous declines at the end-Permian. We suggest southern Africa and North America with northern Africa and that a recent model for ice-based peroxide formation (Liang Eurasia. Presumably Australia will have long since crumpled et al., 2006) contributes signiï¬cantly to this Phanerozoic a neo-Himalayan orogenic belt still higher between its north- glaciation-oxygenation association, and extends even more ern margin and southeast Asia-eastern India. signiï¬cantly through the more severe Precambrian glacial Maximum equatorward extents of ice ages were esti- episodes as well. mated from the paleomagnetic database (dark icicle ï¬ll) or using artistic license (light icicle ï¬ll) where paleomagnetic PONGOLA: EARTHâS OLDEST KNOWN GLACIATION data do not yet exist (e.g., for the Neoproterozoic Ghubrah event). An icicle was dropped from the North Pole to that The middle Archean Pongola Supergroup exposed in Swa- maximum equatorward latitude, with thickness approximat- ziland and parts of South Africa contains massive diamictite ing a plausible duration for each glacial event. Precambrian of the Klipwal and Mpatheni Members of the Delfkom For- glaciations are abbreviated as follows: Po = Pongola; H-SP = mation of the Mozaan Group (Young et al., 1998), which is Huronian and Snowy Pass Supergroups (at least two glacia- constrained to be younger than underlying volcanics of the tions not correlative to South African Makganyene glacia- Nsuze group dated at 2985 Â± 1 (Hegner et al., 1994) and tion); M = Makganyene; Gp = Gariep; Gb = Ghubrah; Ed = older than a 2837 Â± 5 Ma quartz porphyry sill (Gutzmer et al., Edwardsburg; El = Elatina-Ghaup; Gk = Gaskiers-Egan; Go 1999). The diamictites contain a diverse clast composition = Gondwana; Pe = Permian; Pl = Pleistocene. with striated and faceted pebbles, and occasional dropstones Whereas Precambrian glaciations appear restricted to that attest to a glacial origin. intervals of supercontinent fragmentation and dispersal, Although all sedimentary redox indicators throughout Phanerozoic glaciations appear more generally associated the Pongola Supergroup argue for widespread anoxia, stud- with supercontinent amalgamation and intervals of orogen- ies of sulfur isotopes that indicate that mass-independent esis. All glaciations are plausibly connected with minor or fractionation (MIF) decreases during and/or after the glacial major episodes of environmental oxidation or atmospheric intervals have been interpreted to support the presence of oxygenation. atmospheric oxygen (Bekker et al., 2005; Ohmoto et al., For most of Phanerozoic time, an integrated geochemi- 2006). Although this is the conventional interpretation, senso
RAUB AND KIRSCHVINK 87 stricto this is not required. The presence of signiï¬cant MIF mentation is not known, as all paleomagnetic components argues for O2 levels below that needed to form an ozone-UV identiï¬ed so far have failed ï¬eld stability tests (Hilburn et shield, whereas the absence of MIF could indicate either a al., 2005). volcanic sulfur source or increased ocean and atmosphere Glaciogenic units in southern Africaâs Transvaal Super- mixing. In fact, before the studies were done, Kopp et al. group include the Duitschland, Timeball Hill, and Makg- (2005) predicted that a drop in sulfur MIF would be present anyene formations. Hannah et al. (2004) obtained a Re-Os in the Pongola sediments simply from increased ocean and pyrite isochron from the Timeball Hill Formation yielding atmosphere mixing expected for a time of glaciation com- an age of 2.32 Ga for the unit, while Cornell et al. (1996) pared with an ice-free world. Nonetheless, relative oxidation obtained a Pb-Pb isochron indicating ~2.22 Ga for the age of the oceans that could also draw down atmospheric SO2 of Ongeluk Formation volcanics that interï¬nger with the levels, even if unassociated with molecular oxygenation, top of the Makganyene diamictite. The youngest detrital remains a viable explanation for the geochemical blips asso- zircons from thin sedimentary interbeds between ï¬ows of ciated with Pongola glaciation. the Ongeluk volcanics are ~2.23 Ga (Dorland, 2004), cor- Nhelko (2004) studied the paleomagnetism of the Pon- roborating the Pb-Pb isochron for the volcanics themselves. gola diamictite and found an unusually strong and stable Paleomagnetic data from the Ongeluk volcanics indicate magnetization held in detrital magnetite, presumably derived that the Makganyene is a low-latitude Snowball Earth event from pulverizing basaltic-composition clasts present in the (Evans et al., 1997; Kirschvink et al., 2000). Sedimentary diamictite. He estimated the paleolatitude of deposition redox indicators in the Duitschland and most of the Timeball at ~48Âº, with positive fold and conglomerate tests on the Hill imply reducing conditions, but the uppermost units of characteristic, two-polarity magnetization. As other Snow- the Timeball Hill formation contain a hematitic oolitic unit, ball Earth lithostratigraphic markers such as cap carbonates which if primary, hints again that the redox potential of the and carbonate clasts are generally absent, there is as yet no atmosphere and ocean system reached the ferrous/ferric suggestion that Earthâs oldest glaciation might have been a transition (which is energetically only halfway between the low-latitude, global event. hydrogen and oxygen redox potentials). In Canada a paleosol at Ville St. Marie, Quebec (Rainbird et al., 1990) contains granule and pebble clasts with reddened PALEOPROTEROZIC GLACIAL INTERVALS rims at approximately the stratigraphic level of the Lorrain At least three (and potentially many more) discrete intervals Formation. In South Africa the ï¬nal pulse of the glaciogenic of glacial activity punctuate the geological record between succession records ice-rafted dropstones in the basal units about 2.45 and 2.22 Ga (e.g., Hambrey and Harland, 1981). of the massive banded iron and sedimentary manganese in Of these, the best-known and best-preserved belong to the the Hotazel Formation. Together with the superjacent, ran- Huronian Supergroup of Canada and the Transvaal Super- domly magnetized hematitic breccia paleosol (Evans et al., group of southern Africa. 2001) (see âBackgroundâ), there is unequivocal evidence In Canada the classic Huronian succession includes for signiï¬cant oceanic oxidation as well as atmospheric the Ramsey Lake, Bruce, and Gowganda diamictites, sepa- oxygenation in the immediate aftermath of low-paleolatitude rated from one another by thick successions of interbedded Snowball Earth glaciation. marine and ï¬uvial sediments. A single carbonate unit (the Espanola formation) overlies the middle, Bruce Formation STURTIAN AND MARINOAN glacial horizon, with a gradual (not abrupt) transition from the diamictite to carbonate in the Elliot Lake region (abrupt After at least a ~1-billion-year absence through late Paleo- transitions are seen elsewhere but could represent post- proterozoic time, all of the Mesoproterozoic Era, and the ï¬rst glacial transgressions or unconformities). Basal volcanics half of the Neoproterozoic Era, BIFs reappear at <720 Ma, have been U-Pb dated at ~2.45 Ga, and the entire glacial intimately associated with early glacial deposits of the succession is cut by dikes and sills of the Nipissing swarm, âCryogenianâ interval (Klein and Beukes, 1993). At least providing an upper age constraint of ~2.22 Ga. Sedimentary three discrete glaciations punctuate the latter half of Neopro- indicators of a generally reducing surface environment terozoic time (Evans, 2000), and current correlation schemes are common in and around the Ramsey Lake and Bruce appear to permit ï¬ve or more distinct events. The older diamictites, but the ï¬rst appearance of continental redbeds among these tend to be associated with hematite-enriched appears just after the Gowganda event. This is either strong BIFs interrupting otherwise suboxic-to-anoxic, organic- evidence for surface redox conditions reaching the ferrous- rich sediments, again suggesting penetration of oxidants to ferric transition, or else the evolution of terrestrial iron- anomalous water depths accompanying deglaciation (Klein oxidizing organisms. As with the Archean Pongola event, and Beukes, 1993). MIF range of sulfur isotopes is diminished brieï¬y after each The younger two of the Neoproterozoic deglaciations glacial unit, hinting at but not proving transient oxidation occupy the newly deï¬ned Ediacaran Period (Knoll et al., events. Unfortunately, the paleolatitude of Huronian sedi- 2006), at its base (~635 Ma, Condon et al., 2005; Hoffmann
88 ANTARCTICA: A KEYSTONE IN A CHANGING WORLD et al., 2004) and approximately its middle (~580 Ma, CORRELATION CAVEATS FOR EDIACARAN-CAMBRIAN Bowring et al., 2003). At ~635 Ma the basal Ediacaran EVENTS âMarinoanâ low-latitude event is only rarely associated In a comprehensive study of inorganic and organic carbon, with banded iron and sedimentary manganese formation (in and sulï¬de as well as carbonate-associated-sulfate sulfur Brazilâs Urucum province), but it is frequently associated isotopes nearly spanning the Ediacaran Period, Fike et al. with reddened carbonate and shale dominating immediately (2006) infer at least 25 million years of increasing bacte- postglacial sea-level transgression (e.g., Halverson et al., rial sulfate reduction in the oceans following the Marinoan 2004). âSnowballâ deglaciation. A sudden event known in Oman Patterns of sulfur isotopic fractionation in carbonate- as the Shuram anomaly then quickly oxidized a previously associated sulfate change across the Marinoan glaciation, isolated dissolved organic carbon reservoir, and the remain- such that seawater sulfate concentration was minimal during der of the Ediacaran Period experienced increasing levels of and after early Cryogenian glaciations, but signiï¬cant fol- sulfur dissimilation reactions, permitted by enhanced oxygen lowing Marinoan glaciation (Hurtgen et al., 2005). Consis- concentrations (Fike et al., 2006). tent with this trend, the postglacial transgressive sequences Although the Shuram anomaly might correlate to the containing reddened carbonate and shale immediately after Gaskiers glacial event, in line with the general deglaciation- Marinoan deglaciation eventually culminate in black shale oxygenation association sketched in this paper, its age is horizons with microbialaminate textures and isotopic sig- strictly underconstrained, with widely varying estimates natures consistent with sulfate-reducing bacterial mat com- (e.g., see Condon et al., 2005, and Le Guerroue et al., 2006). munities (e.g., Calver and Walter, 2000; Calver et al., 2004; Because the Ediacaran Period is ubiquitously punctuated see also Hoffman et al., 2007). with paleomagnetic anomalies suggesting multiple, rapid true polar wander events (Evans, 1998; Evans, 2003; Raub et MID-EDIACARAN EGAN/GASKIERS GLACIATION al., 2007) which might also oxidize vast quantities of organic carbon (Kirschvink and Raub, 2003; Raub et al., 2007), While the basal Ediacaran deglaciation marks the end of an glaciations are not the only available and attractive correla- unambiguously low latitude, likely Snowball Earth event, tion targets for major isotopic excursions. In fact, decreased the middle interval of Ediacaran successions in northwest generation time and increased frequency of mutation ï¬xa- Australia and in Newfoundland is punctuated by a glacial tion accompanying niche isolation and global warming in event of uncertain severity. Correlation between the Egan the aftermath of rapid true polar wander bursts has been glaciation in Australiaâs Ediacaran carbonate belt (Corkeron, proposed as an explicit mechanism linking true polar wander 2007) and the Gaskiers glaciation in Newfoundlandâs Avalon to the evolution of Ediacara and Metazoa (Kirschvink and terrane (Bowring et al., 2003) is not established, however Raub, 2003). In that respect, even the direct link between the both glacial events are younger than the Marinoan glaciation, ï¬nal Precambrian, âGaskiersâ deglaciation and the evolution and both are associated with anomalous carbonate facies in of animal phyla must be regarded as still hypothesized more otherwise siliciclastic-dominated successions (Corkeron, than proven. 2007; Myrow and Kaufman, 1999). As with the basal-Ediacaran Marinoan deglaciation, the mid-Ediacaran Gaskiers deglaciation is associated with THE PEROXIDE PUMP: postglacial reddening, culminating in pyrite-rich black shale A MECHANISM FOR DEGLACIAL OXYGENATION at a presumed maximum ï¬ooding level. Silicate-hosted iron Many glaciologists have noted a semiregular oscillation in increases from pre-glacial to postglacial time, suggesting a the quantity of hydrogen peroxide contained in Antarctic step-function increase in atmospheric oxygen (Canï¬eld et and Greenland ice cores, with concentrations increasing al., 2006). dramatically during the interval of enhanced ozone hole due Because the megascopic Ediacara fauna appear in the to anthropogenic emissions (Frey et al., 2005, 2006; Hutterli thick turbidite deposits following the Gaskiers deglaciation, et al., 2001, 2004). Similar peroxide peaks are inferred for back-of-the-envelope calculations suggest that the aftermath the polar regions of Mars and the ice sheet encasing Jupiterâs of the last Precambrian glaciation marked the ï¬rst moment moon, Europa (Carlson et al., 1999). in Earth history when atmospheric oxygen levels exceeded Liang et al. (2006) generalize the phenomenon of per- ~15 percent of the present atmospheric level (Canï¬eld et oxide snow produced by photolysis of water vapor above al., 2006). However, the Ediacara fauna have not yet been a cold ice sheet and applied 1-D mass-continuity models found in Newfoundland in the same, continuous stratigraphic of peroxide production to hypothetical glacial scenarios, section as the Gaskiers deglaciation, so the precise cause including Snowball Earths. and effect of postglacial oxygenation and the evolution of With modern volcanic outgassing and dry adiabatic complex life remains ambiguous. lapse rates, and at modern atmospheric pressure and UV inci-
RAUB AND KIRSCHVINK 89 dence, a ~10-million-year-long Snowball glacial event easily Corkeron, M. 2007. âCap carbonatesâ and Neoproterozoic glacigenic suc- might rain out and capture in ice ~0.1 to 1.0 bar of molecular cessions from the Kimberley region, north-west Australia. Sedimentol- ogy 54:871-903. oxygen-equivalent hydrogen peroxide. The sensitivity of this Cornell, D. H., S. S. Schutte, and B. L. Eglington. 1996. The Ongeluk basal- astonishing result trends toward higher peroxide production tic andesite formation in Griqualand West, South-Africaâsubmarine for a depressed hydrologic cycle and lower global mean alteration in a 2222 Ma Proterozoic sea. Precambrian Research temperature, both plausible in a Snowball Earth scenario. 79(1-2):101-123. UV-depletion of stratospheric ozone and enhanced molecu- Dorland, H. 2004. Provenance, Ages, and Timing of Sedimentation of Selected Neoarchean and Paleoproterozoic Successions on the Kaapvaal lar hydrogen escape to space (both correlated, among other Craton. D.Phil. thesis. Rand Afrikaans University, Johannesburg. factors, to decreased geomagnetic ï¬eld intensity) would also Embleton, B. J. J., and G. E. Williams. 1986. Low paleolatitude of depo- increase peroxide mixing rates at Earthâs surface. sition for late Precambrian periglacial varvites in South Australiaâ We suggest that the model and mechanism of Liang implications for Paleoclimatology. Earth and Planetary Science Letters et al. (2006) can explain a pan-Precambrian association in 79(3-4):419-430. Evans, D. A. 1998. True polar wander, a supercontinental legacy. Earth and the geologic record of deglaciation with trace or signiï¬cant Planetary Science Letters 157(1-2):1-8. environmental oxidation and, during the aftermath of at least Evans, D. A. D. 2000. Stratigraphic, geochronological, and paleomagnetic the two most unambiguous Snowball Earth events, atmo- constraints upon the Neoproterozoic climatic paradox. American Jour- spheric oxygenation. We note that the Phanerozoic record nal of Science 300:347-433. of relative atmospheric oxygen concentration inferred by Evans, D. A. D. 2003. A fundamental Precambrian-Phanerozoic shift in Earthâs glacial style? Tectonophysics 375(1-4):353-385. the GEOCARBSULF model is also consistent with mono- Evans, D. A. D. 2006. Proterozoic low orbital obliquity and axial- tonic oxygen production during and immediately following dipolar geomagnetic field from evaporite palaeolatitudes. Nature glaciation. 444(7115):51-55. Evans, D. A., N. J. Beukes, and J. L. Kirschvink. 1997. Low-latitude glacia- . tion in the Paleoproterozoic. Nature 386(6622):262-266. ACKNOWLEDGMENTS Evans, D. A. D., J. Gutzmer, N. J. Beukes, and J. L. Kirschvink. 2001. . Paleomagnetic constraints on ages of mineralization in the Kalahari T. D. R. was supported by a National Science Foundation manganese ï¬eld, South Africa. Economic Geology 96:621-631. Graduate Fellowship, and we gratefully acknowledge sup- Fike, D. A., J. P. Grotzinger, L. M. Pratt, and R. E. Summons. 2006. Oxida- . port from the Agouron Institute and the National Aeronautics tion of the Ediacaran Ocean. Nature 444(7120):744-747. and Space Administration Exobiology program. Frey, M. M., R. W. Stewart, J. R. McConnell, and R. C. Bales. 2005. Atmo- spheric hydroperoxides in West Antarctica: Links to stratospheric ozone and atmospheric oxidation capacity. Journal of Geophysical Research, REFERENCES Atmospheres 110(D23). Frey, M. M., R. C. Bales, and J. R. McConnell. 2006. Climate sensitivity Beerling, D. J., and R. A. Berner. 2005. Feedbacks and the coevolution of of the century-scale hydrogen peroxide (H2O2) record preserved in plants and atmospheric CO2. Proceedings of the National Academy of 23 ice cores from West Antarctica. Journal of Geophysical Research, Sciences U.S.A. 102(5):1302-1305. Atmospheres 111(D21301). Bekker, A., S. Ono, and D. Rumble. 2005. Low atmospheric pO2 in the after- Gutzmer, J. N. Nhleko, N. J. Beukes, A. Pickard, and M. E. Barley. 1999. math of the oldest Paleoproterozoic glaciation. Astrobiology 5(2):244. Geochemistry and ion microprobe (SHRIMP) age of a quartz porphyry Berner, R. A. 2006. GEOCARBSULF: A combined model for Phanero- sill in the Mozaan Group of the Pongola Supergroup: Implications for zoic atmospheric O2 and CO2. Geochimica et Cosmochimica Acta the Pongola and Witwatersrand supergroups. South African Journal of 70(23):5653-5664. Geology 102(2):139-146. Bowring, S., P. Myrow, E. Landing, and J. Ramezani. 2003. Geochrono- Halverson, G. P., A. C. Maloof, and P. F. Hoffman. 2004. The Marinoan . logical constraints on terminal Neoproterozoic events and the rise of glaciation (Neoproterozoic) in northeast Svalbard. Basin Research metazoans. Geophysical Research Abstracts 5:13219. 16(3):297-324. Calver, C. R., and M. R. Walter. 2000. The late Neoproterozoic Grassy Hambrey, M. J., and W. B. Harland. 1981. Earthâs Pre-Pleistocene Glacial Group of King Island, Tasmania: Correlation and palaeogeographic Record. Cambridge: Cambridge University Press. signiï¬cance. Precambrian Research 100(1-3):299-312. Hannah, J. L., A. Bekker, H. J. Stein, R. J. Markey, and H. D. Holland. . Calver, C. R., L. P. Black, J. L. Everard, and D. B. Seymour. 2004. U-Pb 2004. Primitive Os and 2316 Ma age for marine shale: Implications zircon age constraints on late Neoproterozoic glaciation in Tasmania. for Paleoproterozoic glacial events and the rise of atmospheric oxygen. Geology 32(10):893-896. Earth and Planetary Science Letters 225(1-2):43-52. Canï¬eld, D. E., S. W. Poulton, and G. M. Narbonne. 2006. Late-Neopro- Hegner, E., A. KrÃ¶ner, and P. Hunt. 1994. A precise U-Pb zircon age for the terozoic deep-ocean oxygenation and the rise of animal life. Science Archean Pongola Supergroup volcanics in Swaziland. Journal of African 315(5808):92-95. Earth Sciences 18:339-341. Carlson, R. W., M. S. Anderson, R. E. Johnson, W. D. Smythe, A. R. Hilburn, I. A., J. L. Kirschvink, E. Tajika, R. Tada, Y. Hamano, and S. Hendrix, C. A. Barth, L. A. Soderblom, G. B. Hansen, T. B. McCord, Yamamoto. 2005. A negative fold test on the Lorrain Formation of the J. B. Dalton, R. N. Clark, J. H. Shirley, A. C. Ocampo, and D. L. Huronian Supergroup: Uncertainty on the paleolatitude of the Paleopro- Matson. 1999. Hydrogen peroxide on the surface of Europa. Science terozoic Gowganda glaciation and implications for the great oxygen- 283(5410):2062-2064. ation event. Earth and Planetary Science Letters 232:315-332. Condon, D., M. Zhu, S. A. Bowring, W. Wang, A. Yang, and Y. Jin. 2005. Hoffman, P. F. 2007. Comment on âSnowball Earth on Trial.â EOS, Transac- U-Pb ages from the neoproterozoic Doushantuo Formation, China. Sci- tions of the American Geophysical Union 88(9):110. ence 308(5718):95-98. Hoffman, P. F., and D. P. Schrag. 2002. The Snowball Earth hypothesis: Testing the limits of global change. Terra Nova 14(3):129-155.
90 ANTARCTICA: A KEYSTONE IN A CHANGING WORLD Hoffman, P. F., A. J. Kaufman, G. P. Halverson, and D. P. Schrag. 1998. A . Le Guerroue, E., P. A. Allen, A. Cozzi, J. L. Etienne, and M. Fanning. 2006. Neoproterozoic Snowball Earth. Science 281(5381):1342-1346. 50 Myr recovery from the largest negative delta C-13 excursion in the Hoffman, P. F., G. P. Halverson, E. W. Domack, J. M. Husson, J. A.Higgins, Ediacaran ocean. Terra Nova 18(2):147-153. and D. P. Schrag. 2007. Are basal Ediacaran (635 Ma) post-glacial Liang, M. C., H. Hartman, R. E. Kopp, J. L. Kirschvink, and Y. L. Yung. , . âcap dolostonesâ diachronous? Earth and Planetary Science Letters 2006. Production of hydrogen peroxide in the atmosphere of a Snowball 258(1-2):114-131. Earth and the origin of oxygenic photosynthesis. Proceedings of the Hoffmann, K. H., D. Condon, S. A. Bowring, and J. L. Crowley. 2004. U- National Academy of Sciences U.S.A. 103(50):18896-18899. Pb zircon date from the Neoproterozoic Ghaub Formation, Namibia: Lovelock, J. E. 2006. The Revenge of Gaia: Earthâs Climate Crisis and the Constraints on Marinoan glaciation. Geology 32(9):817-820. Fate of Humanity. New York: BasicBooks. Huey, R. B., and P. D. Ward. 2005. Hypoxia, global warming, and terrestrial McElhinny, M. W. 1964. Statistical signiï¬cance of the fold test in paleo- Late Permian extinctions. Science 308:398-401. magnetism. Geophysical Journal of the Royal Astronomical Society Hurtgen, M. T., M. A. Arthur, and G. P. Halverson. 2005. Neoproterozoic . 8:338-340. sulfur isotopes, the evolution of microbial sulfur species, and the burial Myrow, P. M., and A. J. Kaufman. 1999. A newly discovered cap carbonate efï¬ciency of sulï¬de as sedimentary pyrite. Geology 33(1):41-44. above Varanger-age glacial deposits in Newfoundland, Canada. Journal Huston, D. L., and G. A. Logan. 2004. Barite, BIFs and bugs: Evidence of Sedimentary Research 69(3):784-793. for the evolution of the Earthâs early hydrosphere. Earth and Planetary Nhelko, N. 2004. The Pongola Supergroup in Swaziland. D.Phil. thesis. Science Letters 220(1-2):41-55. Rand Afrikaans University, Johannesburg. Hutterli, M. A., J. R. McConnell, R. W. Stewart, H. W. Jacobi, and R. C. Ohmoto, H., Y. Watanabe, H. Ikemi, S. R. Poulson, and B. E. Taylor. 2006. Bales. 2001. Impact of temperature-driven cycling of hydrogen peroxide Sulphur isotope evidence for an oxic Archaean atmosphere. Nature (H2O2) between air and snow on the planetary boundary layer. Journal 442(7105):908-911. of Geophysical Research, Atmospheres 106(D14):15395-15404. Peterson, K. J., M. A. McPeek, and D. A. D. Evans. 2005. Tempo and mode Hutterli, M. A., J. R. McConnell, G. Chen, R. C. Bales, D. D. Davis, of early animal evolution: Inferences from rocks, Hox, and molecular and D. H. Lenschow. 2004. Formaldehyde and hydrogen peroxide in clocks. Paleobiology 31(2):36-55. air, snow and interstitial air at South Pole. Atmospheric Environment Rainbird, R. H., M. A. Hamilton, and G. M. Young. 1990. Formation and 38(32):5439-5450. diagenesis of a sub-Huronian saprolithâcomparison with a modern Kirschvink, J. L. 1992. Late Proterozoic low-latitude global glaciation: weathering proï¬le. Journal of Geology 98(6):801-822. The Snowball Earth. Section 2.3 in The Proterozoic Biosphere: A Raub, T. D., J. L. Kirschvink, and D. A. D. Evans. 2007. True polar wan- Multidisciplinary Study, eds. J. W. Schopf et al., pp. 51-52. Cambridge: der: Linking deep and shallow geodynamics to hydro- and bio-spheric Cambridge University Press. hypotheses. In Treatise on Geophysics, vol. 5, ch. 14, eds. M. Kono and Kirschvink, J. L., and T. D. Raub. 2003. A methane fuse for the Cambrian G. Schubert. Washington, D.C.: American Geophysical Union. explosion: Carbon cycles and true polar wander. Comptes Rendus Geo- Sohl, L. E., D. V. Kent, and N. Christie-Blick. 1999. Paleomagnetic polarity science 335(1):65-78. reversals in Marinoan (ca 600 Ma) glacial deposits of Australia: Implica- Kirschvink, J. L., E. J. Gaidos, L. E. Bertani, N. J. Beukes, J. Gutzmer, L. N. tions for the duration of low-latitude glaciation in neoproterozoic time. Maepa, and R. E. Steinberger. 2000. Paleoproterozoic Snowball Earth: . Geological Society of America Bulletin 111(8):1120-1139. Extreme climatic and geochemical global change and its biological Sumner, D. Y., J. L. Kirschvink, and B. N. Runnegar. 1987. Soft-sediment consequences. Proceedings of the National Academy of Sciences U.S.A. paleomagnetic ï¬eld tests of late Precambrian Glaciogenic Sediments. 97(4):1400-1405. Transactions of the American Geophysical Union 68:1251. Klein, C., and N. J. Beukes. 1993. Sedimentology and geochemistry of Walker, J. C. G., P. B. Hays, and J. F. Kasting. 1981. A negative feedback the glaciogenic Late Proterozoic Rapitan iron-formation in Canada. mechanism for the long-term stabilization of Earthâs surface tem- Economic Geology and the Bulletin of the Society of Economic Geolo- perature. Journal of Geophysical Research, Oceans and Atmospheres gists 88(3):542-565. 86(NC10):9776-9782. Knoll, A. H., M. R. Walter, G. M. Narbonne, and N. Christie-Blick. 2006. Williams, G. E., P. W. Schmidt, and B. J. J. Embleton. 1995. The Neopro- The Ediacaran Period: A new addition to the geologic time scale. Lethaia terozoic (1000-540 Ma) Glacial Intervalsâno more Snowball Earth. 39(1):13-30. Comment. Earth and Planetary Science Letters 131(1-2):115-122. Kopp, R. E., J. L. Kirschvink, I. A. Hilburn, and C. Z. Nash. 2005. The Young, G. M., D. G. F. Young, C. M. Fredo, and H. W. Nesbitt. 1998. Earthâs paleoproterozoic snowball Earth: A climate disaster triggered by the oldest reported glaciation: Physical and chemical evidence from the evolution of oxygenic photosynthesis. Proceedings of the National Archean Mozaan Group (similar to 2.9 Ga) of South Africa. Journal of Academy of Sciences U.S.A. 102:11131-11136. Geology 106(5):523-538.