d C depth proﬁles from paleosols across the …...Thus, paleosol d13C variation with soil depth in...

1459

GSA Bulletin; September 2000; v. 112; no. 9; p. 1459–1472; 12 figures; 1 table.

d13C depth profiles from paleosols across the Permian-Triassicboundary: Evidence for methane release

Evelyn S. Krull*Gregory J. RetallackDepartment of Geological Sciences, University of Oregon, Eugene, Oregon 97403-1272, USA

ABSTRACT

Stable carbon isotopic analyses of organ-ic carbon (d13C) in individual paleosol pro-files from Permian–Triassic sequences ofAntarctica reveal systematic isotopic vari-ations with profile depth. These variationsare in many cases analogous to those inmodern soils, which are functions of redoxconditions, soil development, and degreeand type of microbial decay. In modernsoils, these isotopic depth functions developindependently from vegetation changes (C3

versus C4 vegetation) and can be diagnosticof soil orders. This study shows that soil-intrinsic functions can be preserved in thed13C values of paleosols as old as 260 Maand constitute valuable data for paleoeco-logical interpretations.

A large carbon isotopic offset of as muchas 10‰ in whole paleosol profiles across thePermian-Triassic boundary indicates signif-icant changes in the soil biogeochemistryand the soil-atmosphere system. Early Tri-assic paleosols are distinctive in their ex-tremely low d13C values (to 242‰) and of-ten show an anomalous d13C depth distri-bution compared to both Permian paleosolsand modern soils. Highly depleted d13C val-ues, as the ones in Early Triassic paleosols,are suggested to be associated with micro-bial methane oxidation (methanotrophy).This hypothesis implies increased methaneconcentrations in the Early Triassic soil-at-mosphere system. Increased atmosphericmethane was probably partly responsiblefor the global carbon isotopic shift docu-mented in marine and terrestrial sedimentsacross the Permian–Triassic boundary.

*Present address: School of Chemistry, Physics,and Earth Sciences, Flinders University, Adelaide,SA 5001, Australia; e-mail: [email protected].

Keywords: boundary, carbon-13, methane,paleosols, Permian, Triassic.

INTRODUCTION

Soil organic matter constitutes the largestsurficial pool of organic carbon on continentsand has long been recognized as an importantcomponent of the global carbon cycle (Kernand Schlesinger, 1992). The depth distributionof stable and radiogenic carbon isotopes fromsoil organic matter reflects soil-forming pro-cesses that influence soil organic matter. Re-sults from such studies have been used (1) toquantify rates of soil carbon turnover and stor-age, (2) to study climate-related vegetationchanges, (3) to analyze isotopic effects of de-composition and humus formation, and (4) toclassify isotopic trends with soil development(O’Brien and Stout, 1978; Stout and Rafter,1978; Becker-Heidmann and Scharpenseel,1990, 1992). Detailed sampling of the entiresoil profile is crucial to any of these interpre-tations because d13C depth profiles of soil or-ganic matter reflect systematic fractionationprocesses (Becker-Heidmann and Scharpen-seel, 1986). With a few exceptions, analysesof paleosol organic matter for d13C depth pro-files are rare in Tertiary and older paleosols(Cerling et al., 1991; Bestland and Krull,1999). Most d13C analyses of organic matterfrom pre-Quaternary paleosols reported in theliterature usually consist of one sample perprofile. Commonly, these samples are taken inconjunction with pedogenic carbonate fromBk or Ck horizons, but it is often unclearwhich horizon was sampled (Kingston et al.,1994; Mora et al., 1996). As a result, d13Cvalues display more scatter in these studiescompared with the database from more dense-ly sampled pedogenic carbonates (Quade etal., 1995; Mora et al., 1996).

Previous studies have used carbon isotopicvalues from organic matter mostly as an ac-

cessory tool to assess diagenetic effects of as-sociated carbonates or as an additional vari-able in modeling past atmospheric pCO2 orevaluating vegetation changes (Cerling et al.,1989; Quade et al., 1995; Mora et al., 1996).Thus, paleosol d13C variation with soil depthin the absence of C3/C4 vegetation changes re-mains little explored. As our results show,pedogenic d13C trends similar to those of mod-ern soils can be preserved and recognized inpaleosols and can be important indicators ofbiogeochemical processes in terrestrial pa-leoenvironments.

METHODS

Stratigraphic sections bracketing the Perm-ian–Triassic boundary were measured in theAllan Hills and at Graphite Peak, Antarctica(Fig. 1; Retallack et al., 1995, 1996a; Krull,1998). Each paleosol type (in the sense of Re-tallack, 1988) was characterized in detail fromfield observations, petrographic analyses, andbulk-rock geochemical analyses (Table 1).Field observations included identification ofcolor according to the Munsell color chart, re-action with dilute HCl acid, and degree of de-velopment of paleosols, based on criterion ofRetallack (1988). Thin sections of represen-tative profiles were counted for 500 points us-ing a Swift automatic point counter for grain-size analysis and for analysis of constituentminerals with accuracy of ;2%. Fresh rocksamples from paleosols were collected fromoutcrops for analysis of major and trace ele-ments and for d13C analysis. For d13C analysisof the long stratigraphic sections, we refrainedfrom analyzing coals to ensure adequate com-parison among pedotypes and to avoid isotop-ic artifacts due to the very different mode offormation of Histosols. Instead, deeply rootedunderclays, which are less susceptible to wa-terlogging and accumulation of organic mat-

1460 Geological Society of America Bulletin, September 2000

KRULL AND RETALLACK

Figure 1. Locality map of Graphite Peak and Allan Hills in Antarctica.

TABLE 1. LATE PERMIAN AND EARLY TRIASSIC PALEOSOLS OF ANTARCTICA

U.S. Taxonomy* Diagnosis Former ecotype Time to form

Histosol Thick clean coal (0), on clay or silt,deeply penetrating Vertebrariaroots.

Seasonally drained, swampydepressions

20–50 k.y.

Entisol Gray-green cherty siltstone withsimple (unchambered) root traces.

Streamside swales of flood plains 5–100 yr

Gleyed Inceptisol Gray-green silty or sandy surface (A)over large berthierine nodules(Bg).

Seasonally wet flood plain 1–2 k.y.

Ultisol Gray silty fissured surface (A) overthick clayey subsurface (Bt).

Higher parts of flood plain 40–60 k.y.

Alfisol Gray-green surface (A) over red(10R) clayey subsurface (Bt).

Well-drained flood plain or terrace 40–50 yr

Gleyed Entisol Dark gray shaly surface (A) overgreen-gray subsurface.

Streamside swales and flood plaindepressions

5–100 yr

*Soil Survey Staff (1997).

ter, were sampled. For d13C analysis of indi-vidual paleosol profiles, coals were included.

Chemical analyses for major elements(SiO2, TiO2, Al2O3, Fe2O3, MnO, MgO, CaO,Na2O, K2O, and P2O5) were obtained by in-ductively coupled plasma-fusion spectroscopyand loss on ignition from 4 h at 1000 8C. Stan-dard deviations were 0.23 wt% (SiO2), ,0.01wt% (TiO2), 0.13 wt% (Al2O3), 0.12 wt%(Fe2O3), 0.005 wt% (MnO), 0.03 wt% (MgO),0.16 wt% (CaO), 0.02 wt% (Na2O), 0.16 wt%(K2O), and 0.004 wt% (P2O5). Analysis forferrous iron was from ammonium sulfate ti-tration (standard deviation was 0.15 wt%).Analyses of barium and strontium were by X-ray fluorescence. All geochemical analyseswere done by Bondar Clegg Inc., Vancouver,British Columbia.

In preparation for isotopic analysis, sampleswere powdered and sieved through a 0.210mm screen, then treated for 1 h with hot 2 NHCl to remove all carbonates and other acid-soluble minerals. Acid-insoluble residues werewashed until neutral and oven-dried for atleast 12 h. The dried and weighed residueswere loaded into 9 mm quartz tubes with 1 gCuO, 50 mg Ag foil, and 1 g elemental Cueach, following the procedure of Boutton(1991). The tubes were evacuated, sealed, andcombusted at 860 8C for 6 h and then slowlycooled to 25 8C over 20 h. CO2 was collectedby cryogenic distillation and total organic car-bon (TOC) content was determined by mea-suring the volume of CO2 and converting it tototal amount of carbon (Boutton, 1991). Thed13C analysis was done using the FinniganDelta-E triple collector isotope ratio massspectrometer at the Illinois State GeologicalSurvey. Standard deviation from replicates av-eraged 0.32‰. Isotope results are reported inthe conventional notation as per mil deviationfrom the Peedee belemnite standard (Craig,1957).

CHRONOSTRATIGRAPHICFRAMEWORK

Permian–Triassic strata investigated herebelong to the Victoria Group and consist ofcyclothems of clastic alluvial sediments, coalmeasures, and paleosols, similar to other well-studied Gondwanan sequences (Retallack etal., 1996b). Two localities, Graphite Peak inthe central Transantarctic Mountains and Al-lan Hills in southern Victoria Land, were se-lected for detailed stratigraphic, geochemical,and carbon isotopic studies of paleosols (Fig.2, A and B).

Permian strata in both localities are com-posed of fining-upward sequences of sand-

stone, siltstone, underclay, and coal. CoalyHistosols are the most abundant paleosol typesin these sequences, next to weakly to moder-ately developed Inceptisols. Early Triassic pa-leosols lack coaly Histosols and paleosol typesconsist of Inceptisols, Entisols, and a few Al-fisols (Retallack and Krull, 1999).

The Permian–Triassic boundary in similarGondwanan sequences has been establishedby correlating high-precision 206Pb/238U radio-metric dates of detrital zircons in marine se-quences of China (Claoue-Long et al., 1991)with radiometric dates on tuffs in coal mea-sures of New South Wales (Veevers et al.,1994). These data, together with the d13C ex-cursion in terrestrial deposits in Australia,identify the boundary at the lithologicalchange from coal measures to sandstone andsiltstone (Retallack, 1995; Morante, 1996). InAustralian basins the negative d13C excursionoccurs at or near the base of the palynologicalProtohaploxypinus microcorpus zone, whichcoincides with the cessation of Late Permian

coal measures and the onset of the Early Tri-assic coal gap (Morante, 1996; Retallack etal., 1996b). Thus, the boundary in terrestrialsections is placed between the last Late Perm-ian coal seam, containing Glossopteris leavesand Vertebraria roots, and sandy-silty depos-its yielding the seed fern Dicroidium callip-teroides (Retallack, 1995).

Graphite Peak

At Graphite Peak (Fig. 2A) the paleonto-logically established Permian–Triassic bound-ary is between the last Glossopteris-bearingcoal of the Late Permian Buckley Formationand strata that yield the earliest Triassic rep-tiles Lystrosaurus, Thrinaxodon, and Prola-certa of the Fremouw Formation (Hammer,1989, 1990). Fossil remnants of Lystrosaurus(at 291 m), Dicroidium zuberii (at 369 m), andDicroidium odontopteroides (at 558 m) indi-cate early Scythian, late Scythian, and Anisianages, respectively. Small amounts of shocked

Geological Society of America Bulletin, September 2000 1461

STABLE CARBON ISOTOPE DEPTH PROFILES—EVIDENCE FOR METHANE RELEASE

Figure 2. Stratigraphic sections of Graphite Peak (A) and Allan Hills (B) showing chang-es in paleosols, d13C (solid circles; PDB—Peedee belemnite), and total organic carbon(TOC—open circles) values. Note that for the long stratigraphic sections, we did notanalyze coals to ensure adequate comparison among Permian and Triassic pedotypes.Instead, we sampled deeply rooted underclays, which are comparable to gleyed Incep-tisols.

quartz have been reported from a claystonebreccia immediately above the last coal at 256m, and a faint iridium enrichment occurs atthe base of the last coal (Retallack et al.,1998). A pronounced shift toward more neg-ative d13C values occurs within the last coalseam of the Late Permian Buckley Formationat 256 m (Fig. 2A), and low isotopic valuescontinue into the overlying Early Triassic Fre-mouw Formation (average is 231.4‰). Thisshift toward negative d13C values in the ter-minal Permian, the persistently low isotopicvalues, and the lower TOC values in the EarlyTriassic, are consistent with results from otherisotopic studies (e.g., Magaritz et al., 1992;Morante, 1996) and suggest that this excur-sion at the Buckley-Fremouw contact marksthe Permian-Triassic boundary.

An earlier negative excursion occurs at thebase of the Graphite Peak section. These lowvalues may record an earlier Late Permiannegative isotopic event, as reported in marineand terrestrial sections by Baud et al. (1989),Magaritz et al. (1992), Grotzinger and Knoll(1995), and MacLeod et al. (1997). This neg-ative excursion may correlate with a LatePermian (end Guadalupian) marine extinctionevent (Stanley and Yang, 1994) and a LatePermian (but not latest Permian) extinctionevent in the terrestrial Karoo basin, South Af-rica (King, 1991).

Allan Hills

In the Allan Hills (Fig. 2B), the Weller CoalMeasures include mid-Permian, but not latestPermian, fossil plants and pollen (Kyle, 1977;Kyle and Schopf, 1982). The overlying Feath-er Conglomerate in the Allan Hills has yieldedno fossils of established biostratigraphic util-ity, but in other parts of Victoria Land, lateEarly Triassic to early Middle Triassic (Spa-thian to early Anisian) palynomorphs werefound in the upper part of the Feather Con-glomerate (Kyle, 1977). In the Allan Hills, thelowermost Lashly Formation has fossil plantsof Middle to Late Triassic (Anisian–Ladinian)ages (Retallack and Alonso-Zarza, 1998).

The d13C values of the mid-Permian WellerCoal Measures Formation remain at an aver-age value of 223.5‰ until 17 m above thelast mid-Permian coal seam (Fig. 2B). In thelower Feather Conglomerate, d13C values in-crease (average is 217.2‰); maximum valuesreach 216.6‰. By comparison, stratigraphi-cally higher in the Feather Conglomerate, d13Cvalues are lower, but range widely, from246.2‰ to 217.2‰. A critical hiatus isthought to be present at the base or within theFeather Conglomerate, obscuring the Perm-


KRULL AND RETALLACK

Figure 2. (Continued ).

ian-Triassic boundary. Such stratigraphic in-completeness has long been recognized in theAllan Hills (Collinson et al., 1994) and hin-ders exact placement of the boundary. Thebasal Lashly Formation in the Allan Hills isMiddle Triassic (Retallack and Alonso-Zarza,1998) and d13C values are about 6.5‰ lowercompared with data from the Permian WellerCoal Measures (–29.8‰).

PEDOGENESIS AND DIAGENESIS OFORGANIC MATTER

Carbon isotopes of organic matter generallyresist fractionation during burial diagenesis(Hayes et al., 1983). At conditions belowgreenschist facies (,350 8C) and at H/C ratios.0.2, diagenetic alteration of organic carbonisotopes is negligible (Hayes et al., 1983).Diagenetically affected high-rank kerogen isalso indicated by a dry ash-free carbon content(DAF) .92% (McKirdy and Powell, 1974).

Both localities of this study were affectedto different degrees by regional and localmetamorphism from Jurassic dolerite intru-sions (Coates et al., 1990). The H/C values forAllan Hills coals average 0.44 (Schopf andLong, 1966) and 0.46 for Graphite Peak(Coates et al., 1990) and are thus above theisotopically critical values. The DAF carboncontents for both localities are below the crit-ical value, being 84% for the Allan Hills and67% for Graphite Peak (Schopf and Long,1966; Coates et al., 1990). In addition, authi-genic mineral assemblages in the Allan Hillsand the central Transantarctic Mountains in-clude zeolites (Ballance, 1977), which have anupper stability limit of 300 8C and 3 kbar, wellbelow isotopically critical conditions ofgreenschist facies.

In addition, to ensure that our isotopic dataare not the result of contamination or ther-mally driven loss of hydrocarbons duringdiagenesis, we compared TOC contentsagainst d13C values of each formation (Fig.3, A and B). No significant statistical corre-lation between TOC and d13C values was rec-ognizable; R2 values are consistently below0.1. From these data, we conclude that ourd13C values can be interpreted as an originalisotopic signature.

d13C AND SOIL ORGANIC MATTER INMODERN SOILS

Organic carbon dynamics in modern soilsneed to be considered in order to evaluate an-cient d13C and TOC depth trends in paleosols.During humus formation in soils, a large frac-tion of organic matter is lost due to decom-

position processes, and only the resistant andrefractory organic constituents are incorporat-ed in the long-term organic carbon pool (Bal-esdent and Mariotti, 1996). The TOC contentin modern nonpeaty soils decreases abruptlyfrom the litter and O horizon (10%–40%) tothe A horizon (1%–5%) and B horizon (;1%)and approaches a constant value in the C ho-rizon (,1%) (Stevenson, 1969; Huang et al.,1996). Exceptions are Histosols (soils devel-oped from peat, containing .50% by weightorganic matter), where decomposition of or-ganic matter has been impeded and much ofthe original carbon content is preserved (Ste-venson, 1969).

In modern soils systematic d13C depthtrends have been suggested to be governed ei-ther by vegetation change of C3 versus C4

plants or by different rooting depths under amixed C3 and C4 vegetation (Kelly et al.,1991). However, isotopic data from soils de-veloped under pure C3 or C4 vegetation alsoshow systematic fractionation trends in d13Cvalues with depth (Stout and Rafter, 1978; Na-delhoffer and Fry, 1988; Becker-Heidmannand Scharpenseel, 1986). Thus, d13C depthdistributions are not necessarily a function ofvegetation change, but are significantly con-trolled by decompositional processes. Theseinclude selective preservation of more resis-



Figure 3. Cross plots of carbon isotopic values (d13C ‰; PDB—Peedee belemnite) and total organic carbon contents (TOC%) for (A)the Permian Buckley Formation and Triassic Fremouw Formation (Graphite Peak), and (B) the Permian Buckley Formation and TriassicLashly Formation (Allan Hills).

Figure 4. Examples of typical d13C (PDB—Peedee belemnite) depth profiles from modernsoils. (A) A modern Histosol from New Zealand (modified from Stout and Rafter, 1978).(B) A modern gleyed Alfisol from northern Germany (modified from Becker-Heidmannet al., 1996). (C) A modern well-drained Alfisol from northern Germany (modified fromBecker-Heidmann and Scharpenseel, 1986). (D) A modern rice-land soil at Los Banos,Philippines (modified from Neue et al., 1990).

tant macromolecules, microbial fractionationof organic matter, and addition of microbialbiomass to the carbon pool. In addition, highclay content may enhance selective adsorptionof the heavier isotope (Becker-Heidmann andScharpenseel, 1986, 1992; Wedin et al., 1995;Balesdent and Mariotti, 1996; Huang et al.,1996). Isotopic depth trends in modern soilsthat are independent of vegetation changes canbe divided into three distinct categories: (1) ad13C depth trend toward slightly lower values,(2) a uniform d13C depth trend, or (3) a d13Cdepth trend toward higher values.

Slight decreases in d13C values with soildepth are common for soils that are constantlywaterlogged, such as peat-producing Histosols(Fig. 4A) (Stout and Rafter, 1978). The an-aerobic conditions apparently produce thesmall variations in d13C values with depth andcause little microbial fractionation of originalorganic matter (Stout and Rafter, 1978; De-ines, 1980; Stout et al., 1981). The occasionalslight decrease in d13C values with depth inthe order of 2‰–3‰ in peaty soils may arise


KRULL AND RETALLACK

Figure 5. Carbon isotopic depth profile of a Permian Histosol, showing a typical uniformdistribution of d13C values (PDB—Peedee belemnite) and decreasing total organic carbon(TOC) content with depth.

from preservation of isotopically more nega-tive lignin compared with hemicellulose orcellulose (Deines, 1980; Benner et al., 1987;Boutton, 1996).

A uniform depth trend is characteristic forrelatively young and poorly drained soils thatundergo more fluctuating, less waterloggedconditions than Histosols (Stout and Rafter,1978) (Fig. 4B). In these soils, the uniformd13C depth profiles are mainly the result oflittle time for soil formation and therefore lim-ited decompositional fractionation, which isfurther impeded be partial waterlogging.

Pronounced d13C increases of as much as5‰ with soil depth are the norm in mature,well-drained modern soils (Fig. 4C; Stout etal., 1981; Becker-Heidmann and Scharpen-seel, 1986; Balesdent and Mariotti, 1996). In-creases of as much as 7‰ are exceptional andhave been found only in riceland soils, wherecarbon turnover rates are high and methano-genesis is pronounced (Fig. 4D; Neue et al.,1990). In soils other than rice-land soils, thissystematic d13C increase from the A to the Bhorizon is attributed to a combination of bio-logical and physical factors (Nadelhoffer andFry, 1988; Becker-Heidmann and Scharpen-seel, 1990; Balesdent and Mariotti, 1996).Aerobic conditions within the soil favor selec-tive loss of 12CO2 accompanying microbialdegradation and fractionation, particularly inthe subsurface horizon around the nutrient-rich rhizosphere and in the presence of clays(Martin and Haider, 1986). The abundant claysin the subsurface are also known to selectivelybind 13C (Mortland, 1986; Becker-Heidmannand Scharpenseel, 1986).

In many modern soils, a 1‰–3‰ return tomore negative values in the lower B and Chorizons has been observed. This d13C trendis mostly a function of the translocation of sol-uble, relatively young, and undecomposed or-ganic substances down the profile (Becker-Heidmann and Scharpenseel, 1992; Balesdentand Mariotti, 1996).

d13C DEPTH DISTRIBUTION INPALEOSOLS

Studies of depth distribution of stable car-bon isotopes from paleosol organic mattermust consider the possibility of vegetationchanges. Fossil evidence for C4 plants beforethe Tertiary is equivocal and it is consideredunlikely that C4 plants thrived at the high pa-leolatitudes of Antarctica with insufficientyear-round illumination, low temperatures,and high precipitation (Retallack and Dilcher,1988; Ehleringer et al., 1991). Therefore, mi-crobial fractionation and clay content are con-

sidered the most important fractionation pro-cesses in Permian and Triassic paleosols.

In addition, interpretation of isotopic depthtrends from paleosols has to consider possibleisotopic changes shortly after burial. In non-peaty paleosols, TOC contents typically donot exceed 1 wt% and are often below 0.1%;however, a trend toward decreasing carboncontent with depth analogous to modern soilsis usually preserved (Krull, 1998; Bestlandand Krull, 1999). Decomposition of organicmatter after soil burial may influence d13C val-ues (Balesdent and Mariotti, 1996). Duringthis process, fractionation by anaerobic bac-teria, increased contributions of bacterial cellsat the expense of soil humus, and new for-mation of organic macromolecules is possible(Gong and Hollander, 1997; Harvey andMacko, 1997; Lichtfouse et al., 1998). Takinginto account these potential postburial chang-es, several Permian and Triassic paleosol typesare described and discussed with respect totheir individual d13C depth trends. A summaryof the studied Permian and Triassic paleosolsis given in Table 1.

Late Permian Coaly Histosol

Histosols are organic-rich soils with thickpeaty horizons that typically form in low-ly-ing, swampy, and boggy areas with a shallowgroundwater table. Permian Histosols containabundant chambered roots of Vertebraria,fragments of the plant Glossopteris, and athick coaly O horizon (Fig. 5; Table 1).

The d13C distribution in the type PermianHistosol (at 250 m, Fig. 2A) is relatively uni-form (2.2‰ total variation); the average valueis 224.8‰, typical for C3 vegetation. Fromthe O horizon to the A and Bw horizons, thed13C values increase from 225.5‰ to223.7‰. The TOC contents decline withdepth, from 25.6% to 0.3%.

These trends in d13C and TOC are similarto modern peaty soils. The relatively uniformisotopic values are most likely associated witha permanently high water table where decom-position by anaerobic bacteria is slow. Con-sequently, organic matter undergoes only mi-nor compositional changes, preserving muchof its original d13C value (Stout et al., 1975;



Figure 6. Carbon isotopic depth profile and bulk-rock geochemical data from a gleyedPermian Entisol, showing small d13C variation with depth. Gleying and low weatheringrates are indicated by low ratios of FeIII1/FeII1 and alumina/bases, respectively. PDB—Peedee belemnite. TOC—total organic carbon.

Figure 7. Carbon isotopic depth profile and bulk-rock geochemical data from a gleyedTriassic Inceptisol, showing decreasing values with depth. The relatively high d13C valuesmay indicate the presence of CAM (Crassulacean acid metabolism) plants. PDB—Peedeebelemnite. TOC—total organic carbon.

Nadelhoffer and Fry, 1988). The slight in-crease from the O to the Bw horizon is inter-preted as a result of temporarily better drainedconditions (Fig. 5). Periods of better drainage

would favor faster degradation by aerobic bac-teria and consequentially would result ingreater fractionation and enrichment in d13C.Better drained conditions are also indicated by

deeply rooted Vertebraria roots, representingbetter aerated conditions that predated signif-icant peat accumulation.

The significant TOC decrease with depthhere is typical in modern soils due to decom-positional processes and is expected to be pre-served in paleosols as well (Stevenson, 1986).It is interesting that the d13C values do notfollow this trend. The strong decline in TOCcontent and the uniformity of d13C values, asseen in the depth profile and cross-plot in Fig-ure 5, suggest that d13C distributions with pa-leosol depth are not primarily controlled byTOC content.

Late Permian Entisols

Entisols are very weakly developed soils,typically developed on flood plains beneathplant communities early in ecological succes-sion (Buol et al., 1989; Soil Survey Staff,1997).

The type Permian Entisol (at 75 m, Fig. 2A)contains fossil plants of the horsetails Phyl-lotheca and Paracalamites and is composedof thinly bedded, gray-green, chlorite-bearingtuffaceous siltstones (Krull, 1998) (Fig. 6; Ta-ble 1). Ratios of alumina to other base cations(Al2O3/CaO 1 MgO 1 Na2O 1 K2O) andFeIII1/FeII1 are low (Fig. 6). Carbon isotopicvalues average 224.8‰ and display minimalvariation down the profile (0.6‰). The TOCcontent increases down the profile from 0.08%in the uppermost A horizon to 0.11% in theupper C horizon; there is no apparent corre-lation (R2 5 0.36) between d13C and TOCvalues.

As in modern soils, greenish or grayish col-ors are indicative of the presence of reducediron and suggest anaerobic conditions downthe profile from temporary waterlogging (glei-zation) (Buol et al., 1989; Moore et al., 1992).Weathering and oxidation in gleyed soils isnot pronounced, as indicated by low ratios ofalumina/base cations and FeIII1/FeII1. The uni-form d13C values of this paleosol are typicalfor gleyed, weakly developed soils (Becker-Heidmann et al., 1996). Anaerobic conditionsfavor preservation of organic matter, whichexplains the slightly higher TOC contents withdepth.

Triassic Inceptisol

Inceptisols are moderately developed soilsthat have modest clay accumulations in the Bhorizon (Birkeland, 1984; Buol et al., 1989).The degree of weathering, indicated by the ra-tio of alumina to base cations, is generally low(Birkeland, 1984). The type fossil Inceptisol


KRULL AND RETALLACK

Figure 8. Carbon isotopic depth profile and bulk-rock geochemical data from a TriassicUltisol, showing pronounced and systematic isotopic variation with depth. The d13C trendshows the classic isotopic curve of modern well-developed soils with 13C-enrichment in theBt horizon. The return to more negative d13C values in the C horizon is most likely dueto translocation of relatively undecomposed carbon down the profile, analogous to trans-location processes in modern soils. PDB—Peedee belemnite. TOC—total organic carbon.

(at 70 m, Fig. 2B) is characterized by largechamosite (iron chlorite) nodules in the Cghorizon and abundant reduced iron species(FeII1) with depth (Fig. 7 and Table 1). Thed13C values steadily decrease from the A tothe Cg horizon, from 216.8‰ to 220.3‰.The TOC content increases slightly down theprofile. An apparent correlation between d13Cand TOC is shown by an R2 value of 0.7 (Fig.7). However, this correlation is not thought tobe meaningful, considering the scale of vari-ation in TOC (0.07%) compared to the muchlarger variation of d13C values (3.5‰).

As in the previous Entisol, poorly aeratedsoil conditions may have resulted from water-logging, which impeded weathering and de-composition of organic matter. Nonetheless,increased d13C values with depth indicate frac-tionation by bacterial decomposition and per-haps seasonally better drained conditions.Bacterial decomposition alone does not sufficeto explain the relatively high isotopic valuesthat average 219‰. These values are too highfor C3 plants (modern average is 227‰) andtoo low to account for a flora dominated byC4 plants with isotopic values ranging from217‰ to 29‰ (average is 213‰) (Deines,1980; Boutton, 1996). Although the observedvalue of 219‰ can be interpreted as a mixof C3 and C4 plants (Sikes, 1994), it is unlikelythat C4 plants existed at that time or at suchhigh paleolatitudes. By comparison, CAM(Crassulacean acid metabolism) plants haved13C values intermediate between C3 and C4

plants (the range is 228‰ to 210‰) andcould account for these values (Boutton,1996).

Triassic Ultisol

Ultisols are well-developed, deeply weath-ered soils that commonly form below forestvegetation. These soils are characterized by aclay-enriched (Bt) horizon, strong weatheringprofiles, and base-depleted clays (Buol et al.,1989). The type fossil Ultisol (at 195 m, Fig.2B) has abundant, thick, and deeply reachingroot traces (Fig. 8; Table 1). Values for ratiosof alumina/bases and Ba/Sr are relatively high.Clay accumulation is pronounced in the sub-surface horizon, as estimated from field ob-servation and grain-size analysis (Fig. 8).

The type Ultisol paleosol profile has d13Cvalues that average 225.8‰. The d13C valuesincrease markedly from 228‰ in the A ho-rizon to 220‰ in the upper Bt and 219.4‰in the lower Bt horizon. This isotopic trendcorresponds with an increase in clay contentand a decrease in organic carbon content.From the B to the C horizon, d13C values de-

crease from 221‰ to 233‰ and TOC valuescontinue to decline. The overall isotopic var-iation within this profile is .10‰. There isno significant correlation (R2 5 0.1) betweend13C values and TOC content (Fig. 8).

Overall base depletion, high Ba/Sr ratios,and clay enrichment in the B horizon supportthe interpretation of this paleosol as a well-weathered Ultisol (Buol et al., 1989). The pro-nounced d13C trend toward higher valuesdown the profile is comparable to d13C trendsof well-developed, well-drained modern soils(Becker-Heidmann and Scharpenseel, 1990,1992). Such an isotopic trend is interpreted asa combination of adsorption of the heavierisotope on clays and microbial fractionation

during decomposition (Balesdent and Mariot-ti, 1996). More negative values in the lower-most part of the paleosol may represent trans-location of relatively unfractionated soil or-ganic matter, analogous to modern soils(Becker-Heidmann and Scharpenseel, 1986,1992).

Permian-Triassic Boundary Paleosols

The Permian-Triassic boundary at GraphitePeak has been previously identified paleonto-logically and has been confirmed in this studyby both changes in paleosol types across theboundary and d13C chemostratigraphy (Retal-lack and Krull, 1999; Krull, 1998). The



Figure 9. Carbon isotopic depth profiles and bulk-rock geochemical data from the Perm-ian-Triassic boundary section, showing the last Late Permian Histosol, overlain by anEarly Triassic Inceptisol. Unique in this profile is the rather large negative shift in the Chorizon, already indicating changing organic carbon dynamics. The Early Triassic Incep-tisol is characterized by strong 13C depletion in the B horizon, unlike most d13C profilesin modern soils. PDB—Peedee belemnite. TOC—total organic carbon.

boundary sequence at Graphite Peak consistsof a Late Permian coaly Histosol, directlyoverlain by an Early Triassic Inceptisol (at 256m, Fig. 2A). No coaly Histosols are found inthe Early Triassic, in agreement with the glob-al coal gap of the Early Triassic (Retallack etal., 1996b). Similarly, the iron- and chlorite-rich Inceptisols, which are diagnostic of EarlyTriassic paleosol types, are absent in Permianstrata (Retallack and Krull, 1999; Krull,1998). The following description and discus-sion of carbon isotopic depth profiles from theboundary sequence illustrate the rapidlychanging carbon dynamics from the LatePermian to the Early Triassic.

The last Late Permian paleosol is a fossilHistosol (Fig. 9). Little d13C variation (0.4‰)

occurs within the coaly O horizon. From theA to the C horizon d13C values decrease to237.1‰. This decrease of 10‰ is remarkablefor Histosols, which typically have much lesspronounced isotopic depth profiles (Stout etal., 1975). Such low d13C values are outsidethe range known for C3 plants (Deines, 1980).This paleosol is further characterized by mark-edly declining TOC values from the O(20.8%) to the A (0.08%) and C horizons(0.04%), and generally low values of weath-ering ratios of alumina/bases and Ba/Sr. Nosignificant correlation (R2 5 0.3) was ob-served between d13C values and TOC content(Fig. 9).

The strong isotopic variation and the verylow d13C values are unusual for modern and

fossil Histosols and contrast with the domi-nantly anaerobic conditions in Histosols thatgenerally impede strong fractionation. Low ra-tios of alumina/bases and Ba/Sr indicate thatweathering and leaching were not substantialin this paleosol. Overall, this paleosol displaysa d13C depth profile that is not only unchar-acteristic for Histosols, but also does not cor-respond with any documented d13C depth pro-file of modern soils.

The Early Triassic paleosol that directlyoverlies the Histosol has been classified as afossil Inceptisol (Table 1; Fig. 9). This paleo-sol is characterized by a grayish-greenish Ahorizon and a gleyed B (Bg) horizon withiron-chlorite rich nodules. Carbon isotopicvalues are highly variable and have exception-ally low d13C values. An abrupt d13C decreaseoccurs from the A horizon (–26.5‰) to theiron- and chlorite-rich Bg horizon (–42.5‰).In the Bw horizon d13C values return to highervalues of on average 230.2‰ and decreaseagain in the C horizon to 239.2‰. The TOCcontents are low throughout the profile butshow a steady decrease from 0.1% in the Ahorizon to 0.07% in the B and 0.03% in theC horizons. No significant correlation (R2 50.2) was observed between d13C values andTOC content (Fig. 9).

Insufficient drainage and a fluctuating watertable are indicated by abundant minerals con-taining FeII1 in nodules of the Bg horizon andlow weathering ratios of alumina/bases. Underthese gleyed conditions a relatively homoge-neous d13C trend would be expected. By com-parison, the observed d13C profile displays alarge spread of .10‰. Only well-developedmodern soils show pronounced isotopic vari-ation with 13C enrichment with depth. How-ever, the depth trend in this Early TriassicInceptisol is associated with lower d13C (13Cdepleted) values in the B horizon instead ofthe expected 13C enrichment. Furthermore,these low values of 242‰ in the Bg and Chorizons clearly exceed the isotopic range thatis commonly known for C3 plants. These iso-topic features have not been observed in mod-ern soils; thus no modern analogues have beendocumented that could explain this d13C trend.

Other Early Triassic Paleosols

The distinct isotopic characteristics presentin this Early Triassic Inceptisol of (1) largeintraprofile variation of d13C values, (2) verylow d13C values in the subsurface (B horizon),and (3) anomalously low d13C values to240‰, are found in other Early Triassic pa-leosol types as well (Table 1). For example,the isotopic depth profile in the Early Triassic


KRULL AND RETALLACK

Figure 10. Carbon isotopic depth profile and bulk-rock geochemical data from an EarlyTriassic Alfisol, showing strong 13C depletion in the B horizon. The distinctly negative d13Cvalues and the reversed isotopic depth trend (of more negative values in the B horizon)point to significant diffusion of methane into the profile and fixation by methane-oxidizingbacteria. PDB—Peedee belemnite. TOC—total organic carbon.

type Alfisol (at 380 m, Fig. 2A) is character-ized by very low d13C values (average236.5‰) and high variability of .10‰ (Fig.10). From the A horizon to the Bt horizon, val-ues decrease from 238.7‰ to an average of240‰. In the C horizon, d13C values increaseto 230.6‰. A weak correlation between d13Cvalues and TOC content is suggested by R2 50.6. Although, it is questionable whether a var-iation of 0.02% in TOC values can be consid-ered significant (Fig. 10).

The Early Triassic type Entisol (at 260 m,Fig. 2A) is another example of these distinctEarly Triassic d13C depth trends. In this paleo-sol, d13C values average 236.4‰ and showlarge intraprofile variations (.10‰) (Fig. 11).Isotopic values decrease from 235.3‰ in theA horizon to 239.0‰ in the Bw horizon. Inthe C horizon, d13C values increase to228.0‰. No significant correlation was ob-served between d13C values and TOC content(R2 5 0.3) (Fig. 11).

ISOTOPIC ENIGMA

Large Isotopic Variation

Several paleosols investigated in this studyshow isotopic depth variations that are.10‰. In modern soils, isotopic variationsseldom exceed 5‰ (Becker-Heidmann et al.,1996), except in rice-paddy soils (Fig. 4D;Neue et al., 1990) or when associated with aC3/C4 vegetation change. Such vegetationchanges are known to impart large and abruptisotopic depth trends on soils (Ambrose andSikes, 1991; Kelly et al., 1991; Pessenda etal., 1996), but C4-like photosynthesis is un-documented for the late Paleozoic–early Me-sozoic. Considering the unlikelihood of con-tribution from C4 plants, most intraprofilevariations are believed to be governed by fac-tors other than vegetation change.

Many of the isotopic depth profiles pre-sented here preserve gradual changes withdepth, analogous to modern soils, but with agreater isotopic spread. It is interesting thatPermian Histosols and weakly developedPermian Inceptisols do not display majorshifts in d13C values with depth, but better de-veloped paleosols do.

The more pronounced isotopic depth trendsin paleosols compared with modern soils sug-gest that either soil fractionation processeswere different then or that postburial fraction-ation processes had affected the original iso-topic values. Possible causes for such larged13C variations are examined in the followingdiscussion of the type fossil Ultisol (Fig. 8).This paleosol has a trend similar to that of

modern well-developed soils, with increasingd13C values from the A to the B horizon andreturn to lower values in the C horizon. How-ever, the large intraprofile d13C variation hereis very different from the common d13C vari-ation observed in modern soils.

The most negative d13C values in this pa-leosol occur in the A horizon (–28.9‰) andC horizon (–33.3‰) and the highest values arefound in the lower Bt horizon (–19.3‰). Thelarge increase of d13C values in the B horizonmight be due to postburial fractionation. Post-burial microbial decomposition within the Bhorizon may have resulted in increasingly de-graded organic matter with high d13C values.With continued degradation, organic carbon inburied soils tends to become increasingly en-riched in bacterial cells (necromass) which,due to their high resistance to further degra-dation, are selectively preserved, especially inthe presence of clays (Lichtfouse et al., 1995a,1995b; Gong and Hollander, 1997; Harveyand Macko, 1997; Hedges and Oades, 1997).The observed d13C trend suggests that B ho-rizons that were subject to significant frac-tionation processes during pedogenesis areparticularly susceptible to further postburial

fractionation. This is consistent with the ob-servation that paleosols with little pedogenicmodification, such as Histosols or weakly de-veloped Permian Entisols, have small isoto-pic variations.

13C-Enriched Values

The d13C values higher than typical for C3

plants occur in several paleosols of the FeatherConglomerate of Middle Triassic age (Figs. 2and 7). These gleyed Inceptisols probablyformed under partially waterlogged condi-tions. The relict bedding suggests only weakcolonization by plants (Table 1). The d13C val-ues with depth are fairly uniform and average217.9‰. This isotopic value is well outsidethe average of pure C3 (–27‰) and pure C4

(–13‰) plants, but within the range of CAMplants (Deines, 1980).

The 13C enrichment in this paleosol couldbe produced by the colonization of aquaticCAM plants, such as Isoetes. Evidence thatthe CAM photosynthetic pathway was well es-tablished during the late Mesozoic has beendocumented isotopically and paleobotanicallyfrom fossil plants of the latest Cretaceous



Figure 11. Carbon isotopic depth profile and bulk-rock geochemical data from an EarlyTriassic Entisol, displaying the anomalous isotopic depth trend and very negative valuescommon to other Early Triassic paleosols of this study. PDB—Peedee belemnite. TOC—total organic carbon.

(Bocherens et al., 1993). Living aquatic plantssuch as the primitive spore-bearing Isoetes, aswell as some monocots (Poaceae) and dicots(Crassulaceae and Plantaginaceae) are knownto use the CAM photosynthetic pathway (Kee-ley, 1990). Fossil species of Isoetes and alliedplants have been reported from numerous lo-calities in Early Triassic paleosols of Austra-lia, South Africa, Argentina, and India (Retal-lack, 1997).

13C-Depleted Values

Early Triassic paleosols show anomalouslylarge isotopic variation with extremely de-pleted values, usually in the subsurface hori-zons. Isotopic values as low as 235‰ havebeen observed in leaves of tropical rainforests(Jackson et al., 1993). This occurs when CO2

is assimilated from 13C-depleted soil respira-tion. Biochemical fractions such as lipids canbe depleted in d13C by as much as 8‰ (av-erage depletion is 5‰) relative to the wholeplant (Deines, 1980). Individual lipid com-pounds (biomarkers) may be depleted by 9‰or more relative to bulk soil carbon (Lichtfou-se et al., 1995b; Huang et al., 1996). Bio-markers of bacterial origin (hopanoids) with

d13C values lower than 240‰ have been re-ported from Eocene lake sediments and havebeen associated with methane recycling bymethanotrophic bacteria (Hayes et al., 1987;Freeman et al., 1990). Unfortunately, com-pound-specific isotopic analysis of biomarkersdoes not provide conclusive evidence for pre-dominance of certain bacterial populations informer sedimentary organic matter. Damsteand Schouten (1997) demonstrated that inmost of these biomarker studies, bulk d13Cvalues are not significantly influenced by thepresence of isotopically depleted biomarkers.Instead, the bulk d13C values of these studiesshow values typical of C3 plants, suggestingthat the contribution of individual bacterialcompounds to the bulk isotopic signature issmall. Therefore, only in cases where the d13Cvalues of bulk organic matter show significantdepletion can it be concluded that a large frac-tion of the organic matter is derived frommethanotrophic processes (Perry Burhan etal., 1995; Holland, 1997).

Global occurrence of such highly depletedcarbon isotopic values in Earth’s history israre. The only other time in Earth’s historywhen d13C values as low as the ones in theseEarly Triassic paleosols are detected globally

was during a brief interval of the Precambrian(Hayes et al., 1983; Hayes, 1994; Holland,1997). Pronounced excursions with d13C val-ues as low as 260‰ occurred worldwide atthe Archean-Proterozoic transition, ca. 2.75–2.5 Ga. These low d13C excursions during theArchean-Proterozoic transition are interpretedas marking the onset of widespread methan-otrophy (methane oxidation). Methanogenesisis thought to have been ubiquitously presentduring the Archean. However, prior to onsetof methanotrophy, the isotopic signature ofmethanogenic fractionation could not be ef-fectively recorded due to the lack of conver-sion from gaseous to solid matter (Hayes,1994). Thus, the evolution of methanotrophicbacteria allowed for oxidation of methane andincorporation of the 13C-depleted carbon frommethanogenesis in the geologic organic car-bon isotopic record (Hayes, 1994).

Today, methanotrophic bacteria are abun-dant in a variety of ecosystems, occurringwherever oxygen and methane occur in closeproximity such as in soils, oxic and anoxicsediments, freshwater and marine systems,and plant tissues (King, 1992; Hanson andHanson, 1996). Most methanotrophs are aer-obic or facultative anaerobic bacteria that ox-idize atmospheric methane within oxic-anoxicinterfaces. Due to their ubiquity, they play animportant role in uptake of atmospheric meth-ane and in the biosphere-climatic feedbacksystem (Prieme and Christensen, 1997; Saariet al., 1997). Methanotrophic bacteria are par-ticularly abundant in mid-latitudinal forestsoils. Here, the source of methane is not frombelow-ground methanogenesis, but from dif-fusion of atmospheric methane into the soil,where it is oxidized in the subsurface. At suf-ficiently high methane concentrations, frac-tionation against the heavier 13C isotope canbe as much as 20‰ (Coleman et al., 1981;Alperin et al., 1988; Tyler et al., 1994). Thus,the residual methane compared with thesource methane is 13C enriched, whereas therespired CO2 and bacterial cells of methano-trophs become extremely depleted in 13C,down to d13C values of 260‰ (Borowski etal., 1997; Boschker et al., 1998; Jahnke et al.,1999). Although widespread in modern soils,methanotrophy is not dominant enough to sig-nificantly alter the d13C signature of bulk soilorganic matter. Consequently, to preserve amethanotrophic signature in Early Triassic pa-leosol organic matter, two processes are nec-essary. First, atmospheric methane levels musthave been high enough to sustain a large pop-ulation of methanotrophic bacteria. Second,postburial decomposition degraded much ofthe soil organic matter and preserved the more


KRULL AND RETALLACK

Figure 12. Isotopic depth profiles of paleo-sols from Graphite Peak, showing an av-erage offset of ;7‰ in d13C values betweenPermian and Triassic paleosols. PDB—Pee-dee belemnite.

refractory bacterial organic matter, such thatits contribution dominates the bulk d13Crecord.

IMPLICATIONS FOR THE PERMIAN–TRIASSIC BOUNDARY

The negative d13C shift from the Late Perm-ian to the Early Triassic has been interpretedas a chemostratigraphic marker of the Perm-ian–Triassic boundary (Erwin, 1993; Morante,1996). In our Antarctic sections d13C valuesdecrease on average 7‰ from Late Permianto Early Triassic paleosols (Fig. 12). Compa-rably large isotopic shifts have also beenreported from terrestrial sediments from Aus-tralia and South Africa (Morante, 1996;MacLeod et al., 1997). Due to continuous car-bon exchange between the atmosphere-hydro-sphere-biosphere system, such low valueshave been interpreted as evidence of a 13C-depleted Early Triassic atmosphere (Erwin,1993).

An atmospheric isotopic shift of this mag-nitude is most likely explained by an in-creased flux of 13C-depleted methane (Erwin,1993). Its preservation in the organic carbonisotopic record was probably achieved by bac-terial methane oxidation. Previous hypothesesof large methane releases across the Permian-Triassic boundary involved destabilization ofclathrates, possibly associated with a meteor-ite impact, or methane release from continen-tal shelf deposits during the Late Permian re-gression (Erwin, 1993; Bowring et al., 1998).

The duration of the isotopic shift across thePermian–Triassic boundary, which is neces-sary to model the volume of carbon involved,is more difficult to estimate. Earlier studiesbased on paleontological evidence estimatethe excursion to have lasted 1–3 m.y. (Holseret al., 1989; Magaritz, 1989), suggesting anisotopic recovery lag throughout the Scythian,analogous to the recovery lag in the fossil re-cord (Erwin, 1993). More recent studies basedon sedimentation rates (Wang et al., 1994) andU/Pb zircon geochronology in China (Bow-ring et al., 1998) estimate a much shorter du-ration of the isotopic excursion, lasting only50–150 k.y. The estimates by Bowring et al.(1998) proved to be controversial, as follow-up studies produced different results, suggest-ing a significantly longer duration (Mundil etal., 1999). Thus, estimates of the duration ofthe Permian-Triassic boundary are equivocaland await further documentation.

The Permian–Triassic d13C shift is compa-rable to a similar isotopic event in the latePaleocene, for which a catastrophic, short-term methane release has been hypothesized

(Dickens et al., 1997; Bains et al., 1999; Katzet al., 1999). For an estimated duration of 10k.y. and an isotopic shift of 2‰–3‰, theamount of methane necessary is ;1.4–2.8 31018 g. If the methane was derived from clath-rates, this amount would correspond to ;14%of the present-day reservoir (using estimatesof Kvenvolden, 1993).

In the case of the Permian–Triassic bound-ary, a more sustained or repeated methane fluxwould be required, even when using the con-servative estimates by Bowring et al. (1998).Evidence for a prolonged (but possibly highlyvariable) methane release is supported by theunusual d13C depth distribution in Early Tri-assic paleosols (13C depletion in the subsur-face horizon and very low values). A short-term (hundreds to thousands of years)methane release would result in rapid oxida-tion to 13C-depleted CO2 due to the brief res-idence time of methane (10 yr; Khalil andRasmussen, 1990). Therefore, a short-term(tens of thousands of years) release would notproduce the evidence for increased methano-trophy as seen in several paleosols (Fig. 2A).Instead, the preservation of these anomalousdepth trends and occurrence in numerous pa-leosols throughout the Early Triassic suggeststhat methane release must have been pro-nounced enough to support increased meth-anotrophic activity over a long period of time,and must have occurred as multiple events,because paleosol profiles with normal valuesoccur between the 13C-depleted events (Fig.2A).

An explanation for such a long-term releaseof methane has to involve an efficient triggermechanism as well as a positive feedback sys-tem. The marine regression during the Chang-xingian stage would have substantiallylowered hydrostatic pressure in shelf environ-ments and could have led to initial clathratedestabilization and methane release. Thiswould have led to the first pulse of 13C-de-pleted methane and the onset of an increase ingreenhouse gases (CH4 and CO2). Additionalgreenhouse gases such as CO2 and H2O fromthe eruption of the Siberian Traps during thelatest Changxingian stage (250 6 0.2 Ma)may have furthered paleoclimatic warming(Erwin, 1993; Renne et al., 1995; Bowring etal., 1998). In response to this warming trend,clathrates in the polar regions could have be-come increasingly unstable, resulting in a sec-ond pulse of 13C-depleted methane (Krull,1999; Krull et al., 2000). This additional re-lease of methane and CO2 would further addto greenhouse warming and generate a posi-tive feedback cycle of clathrate destabilizationand global warming. This continued, but prob-

ably highly variable, supply of methane to theEarly Triassic atmosphere probably accountsfor the low d13C values of paleosol organicmatter observed in this study.

CONCLUSIONS

From our study of d13C trends of paleosolorganic matter across the Permian-Triassicboundary, the following conclusions can bedrawn.1. Organic carbon isotopic values from Perm-ian and Triassic paleosol depth profiles canpreserve an original pedogenic signature anal-ogous to modern soils.2. Paleosols associated with severe pertur-bations of the global carbon cycle (as in theEarly Triassic) yield isotopic depth profilesof greater d13C variation, more extreme val-ues, and depth trends without modernanalogs.3. The extreme negative isotopic values andthe 13C depletion in the B horizon of EarlyTriassic paleosols are interpreted as a result ofmethanotrophic activity during a prolongedmethane release during the Early Triassic.



ACKNOWLEDGMENTS

We thank M. Alperin, E.A. Bestland, M.Bird, T. Hoehler, E. Pendall, and J. Quade forthorough reviews, and Chao-li Jack Liu, KeithHackley, and Wang Hong for much useful dis-cussion. Isotopic analyses were carried out atthe Illinois State Geological Survey IsotopicLaboratory, and their help and support aregreatly acknowledged. Work was funded byNational Science Foundation grant OPP-9315228 and a Student Research Grant fromthe Department of Geological Sciences, Uni-versity of Oregon.

REFERENCES CITED

Alperin, M.J., Reeburgh, W.S., and Whiticar, M.J., 1988,Carbon and hydrogen isotope fractionation resultingfrom methane oxidation: Global Biogeochemical Cy-cles, v. 2, p. 279–288.

Ambrose, S.H., and Sikes, N.E., 1991, Soil carbon isotopeevidence for Holocene habitat change in the KenyaRift Valley: Science, v. 253, p. 1402–1405.

Bains, S., Corfield, R.M., and Norris, R.D., 1999, Mecha-nisms of climate warming at the end of the Paleocene:Science, v. 285, p. 724–727.

Balesdent, J., and Mariotti, A., 1996, Measurement of soilorganic matter turnover using 13C natural abundance,in Boutton, T.W., and Yamasaki, S., eds., Mass spec-trometry of soils: New York, Marcel-Dekker, Inc., p.83–111.

Ballance, P.F., 1977, The Beacon supergroup in the AllanHills, central Victoria Land, Antarctica: New ZealandJournal of Geology and Geophysics, v. 20, p. 1003–1016.

Baud, A., Magaritz, M., and Holser, W.T., 1989, Permian-Triassic of the Tethys: Carbon isotope studies: Geo-logische Rundschau, v. 78, p. 649–677.

Becker-Heidmann, P., and Scharpenseel, H.-W., 1986, Thinlayer d13C and d14C monitoring of ‘‘lessive’’ soil pro-files: Radiocarbon, v. 28, p. 383–390.

Becker-Heidmann, P., and Scharpenseel, H.-W., 1990, Car-bon isotope dynamics in some tropical soils: Radio-carbon, v. 31, p. 672–679.

Becker-Heidmann, P., and Scharpenseel, H.-W., 1992, Theuse of natural 14C and 13C in soils for studies on globalclimate change: Radiocarbon, v. 34, p. 535–540.

Becker-Heidmann, P., Scharpenseel, H.-W., and Wiech-mann, H., 1996, Hamburg radiocarbon thin layer soilsdatabase: Radiocarbon, v. 38, p. 295–345.

Benner, R., Fogel, M.L., Sprague, E.K., and Hodson, R.E.,1987, Depletion of 13C in lignin and its implicationsfor stable carbon isotope studies: Nature, v. 329, p.708–710.

Bestland, E.A., and Krull, E.S., 1999, Paleoenvironmentsof early Miocene Kisingiri volcano Proconsul sites:Evidence from carbon isotopes, paleosols, and hy-dromagmatic deposits: Geological Society of LondonJournal, v. 156, p. 965–976.

Birkeland, P.W., 1984, Soils and geomorphology: NewYork, Oxford University Press, 372 p.

Bocherens, H., Friis, E.M., Mariotti, A., and Pedersen,K.R., 1993, Carbon isotopic abundances in Mesozoicand Cenozoic fossil plants: Paleoecological implica-tions: Lethaia, v. 26, p. 347–358.

Borowski, W.S., Paull, C.K., and Ussler, W., 1997, Carboncycling within the upper methanogenic zone of con-tinental rise sediments: An example from the methane-rich sediments overlying the Blake Ridge gas hydratedeposit: Marine Chemistry, v. 57, p. 299–311.

Boschker, H.T.S., Nold, S.C., Wellsbury, P., Bos, D., deGraaf, W., Pel, R., Parkes, R.J., and Cappenberg, T.E.,1998, Direct linking of microbial populations to spe-cific biogeochemical processes by 13C-labelling of bio-markers: Nature, v. 392, p. 801–805.

Boutton, T.W., 1991, Stable carbon isotopic ratios of naturalmaterials: I. Sample preparation and mass spectro-metric analysis, in Coleman, D.C., and Fry, B., eds.,Carbon isotope techniques: San Diego, California, Ac-ademic Press, p. 155–171.

Boutton, T.W., 1996, Stable carbon isotope ratios of soilorganic matter and their use as indicators of vegetationand climate change, in Boutton, T.W., and Yamasaki,S.I., eds., Mass spectrometry of soils: New York, Mar-cel Dekker, p. 47–81.

Bowring, S.A., Erwin, D.H., Jin, Y.G., Martin, M.W., Dav-idek, K., and Wang, W., 1998, U/Pb zircon geochro-nology and tempo of the end-Permian mass extinction:Science, v. 280, p. 1039–1045.

Buol, S.W., Hole, F.D., and McCracken, 1989, Soil genesisand classification (third edition): Ames, Iowa StateUniversity Press, 446 p.

Cerling, T.E., Quade, J., Wang, Y., and Bowman, J.R.,1989, Carbon isotopes in soils and paleosols as eco-logic and paleoecologic indicators: Nature, v. 341, p.138–139.

Cerling, T.E., Quade, J., Ambrose, S.H., and Sikes, N.E.,1991, Fossil soils, grasses, and carbon isotopes fromFort Ternan, Kenya: Grassland or woodland?: Journalof Human Evolution, v. 21, p. 295–306.

Claoue-Long, J.C., Zhang, Z.C., Ma, G-G., and Du, S.-H.,1991, The age of the Permo-Triassic boundary: Earthand Planetary Science Letters, v. 105, p. 182–190.

Coates, D.A., Stricker, G.D., and Landis, E.R., 1990, Coalgeology, coal quality, and coal resources in Permianrocks of the Beacon supergroup, Transantarctic Moun-tains, Antarctica: Mineral resources potential of Ant-arctica: Antarctic Research Series, v. 51, p. 133–162.

Coleman, D.D., Risatti, J.B., and Schoell, M., 1981, Frac-tionation of carbon and hydrogen isotopes by meth-ane-oxidizing bacteria: Geochimica et CosmochimicaActa, v. 45, p. 1033–1037.

Collinson, J.W., Isbell, J.L., Elliot, D.H., Miller, M.F., andMiller, J.M.G., 1994, in Veevers, J.J., and Powell,C.M., eds., Permian-Triassic Pangean basins and foldbelts along the Panthalassan margin of Gondwana-land: Geological Society of America Memoir 184, p.173–222.

Craig, H., 1957, Isotopic standards for carbon and oxygenand correction factors for mass spectrometric analysisfor carbon dioxide: Geochimica et CosmochimicaActa, v. 12, p. 133–149.

Damste, J.S., and Schouten, S., 1997, Is there evidence fora substantial contribution of prokaryotic biomass toorganic carbon in Phanerozoic carbonaceous sedi-ments?: Organic Geochemistry, v. 26, p. 517–530.

Deines, P., 1980, The isotopic composition of reduced or-ganic carbon, in Fritz, P., and Fontes, J.C., eds., Hand-book of environmental isotope geochemistry, Volume1, The terrestrial environment, A: Amsterdam, Elsev-ier, p. 329–406.

Dickens, G.R., Castillo, M.M., and Walker, J.C.G., 1997,A blast of gas in the latest Paleocene: Simulating first-order effects of massive dissociation of oceanic meth-ane hydrate: Geology, v. 25, p. 259–262.

Ehleringer, J.R., Sage, R.F., Flanagan, L.B., and Pearcey,R.W., 1991, Climate change and the evolution of C4

photosynthesis: Trends in Ecology and Evolution, v.6, p. 95–99.

Erwin, D.H., 1993, The great Paleozoic crisis—Life anddeath in the Permian: New York, Columbia UniversityPress, 327 p.

Freeman, K.H., Hayes, J.M., Trendel, J.-M., and Albrecht,P., 1990, Evidence from carbon isotope measurementsfor diverse origins of sedimentary hydrocarbons: Na-ture, v. 343, p. 254–256.

Gong, G., and Hollander, D.J., 1997, Differential contri-bution to sedimentary organic matter in oxic and an-oxic environments, Santa Monica Basin, California:Organic Geochemistry, v. 26, p. 545–563.

Grotzinger, J.P., and Knoll, A.H., 1995, Anomalous carbon-ate precipitates: Is the Precambrian the key to thePermian?: Palaios, v. 10, p. 578–596.

Hammer, W.R., 1989, Lystrosaurus zone (Triassic) verte-brates from the Beardmore Glacier region, Antarctica:U.S. Antarctic Journal, v. 24, p. 40–41.

Hammer, W.R., 1990, Thrinaxodon from Graphite Peak,

central Transantarctic Mountains, Antarctica: U.S.Antarctic Journal, v. 25, p. 37.

Hanson, R.S., and Hanson, T.E., 1996, Methanotrophic bac-teria: Microbiological Review, v. 60, p. 439–471.

Harvey, H.R., and Macko, S.A., 1997, Catalysts or contrib-utors? Tracking bacterial mediation of early diagenesisin the marine water column: Organic Geochemistry,v. 26, p. 531–544.

Hayes, J.M., 1994, Global methanotrophy at the Archean-Proterozoic transition, in Bengtsen, S., ed., Early lifeon Earth: Sweden, Nobel Symposium Proceedings 84,New York, Columbia University Press, 630 p.

Hayes, J.M., Kaplan, I.R., and Wedeking, K.W., 1983, Pre-cambrian organic geochemistry, in Schopf, J.W., ed.,Earth’s earliest biosphere; its origin and evolution:Princeton, New Jersey, Princeton University Press, p.93–132.

Hayes, J.M., Takigiku, R., Ocammpo, R., Callot, H.J., andAlbrecht, P., 1987, Isotopic compositions and probableorigins of organic molecules in the Eocene Messelshale: Nature, v. 329, p. 48–51.

Hedges, J.I., and Oades, J.M., 1997, Comparative organicgeochemistries of soils and marine sediments: OrganicGeochemistry, v. 27, p. 319–361.

Holland, H.D., 1997, Evidence for life on Earth more than3850 million years ago: Science, v. 275, p. 38–39.

Holser, W.T., Schonlaub, H.P., Attrep, M., Jr., Boeckelmann,K., Klein, P., Magaritz, M., Orth, C.J., Fenninger, A.,Jenny, C., Kralik, M., Mauritsch, H., Pak, E.,Schramm, J.M., Stattegger, K., and Schmoller, R.,1989, An unique geochemical record at the Permian/Triassic boundary: Nature, v. 337, p. 39–44.

Huang, Y., Bol, R., Harkness, D.D., Ineson, P., and Eginton,G., 1996, Postglacial variations in distributions, 13Cand 14C contents of aliphatic hydrocarbons and bulkorganic matter in three types of British upland soils:Organic Geochemistry, v. 24, p. 273–287.

Jackson, P.C., Meinzer, F.C., Goldstein, G., Holbrook,N.M., Cavelier, J., and Rada, F., 1993, Environmentaland physiological influences on carbon isotope com-position of gap and understory plants in a lowlandtropical forest, in Ehleringer, J.R., et al., eds., Stableisotopes and plant carbon-water relations: San Diego,Academic Press, 555 p.

Jahnke, L.L., Summons, R.E., Hope, J.M., and Des Marais,D.J., 1999, Carbon isotope fractionation in lipids frommethanotrophic bacteria II: The effects of physiologyand environmental parameters on the biosynthesis andisotopic signature of biomarkers: Geochimica et Cos-mochimica Acta, v. 63, p. 79–93.

Katz, M.E., Pak, D.K., Dickens, G.R., and Miller, K.G.,1999, The source and fate of massive carbon inputduring the latest Paleocene thermal maximum: Sci-ence, v. 286, p. 1531–1533.

Keeley, J.E., 1990, Photosynthetic pathways in freshwateraquatic plants: Trends in Ecology and Evolution, v. 5,p. 330–333.

Kelly, E.F., Amundson, R.G., Marino, B.D., and DeNiro,M.J., 1991, Stable isotope ratios of carbon in phyto-liths as a quantitative method of monitoring vegetationand climate change: Quaternary Research, v. 35, p.222–233.

Kern, R.A., and Schlesinger, W.H., 1992, Carbon stores invegetation: Nature, v. 357, p. 447–448.

Khalil, M.A.K., and Rasmussen, R.A., 1990, Constraints onthe global sources of methane and an analysis of re-cent budgets: Tellus, v. 42B, p. 554–559.

King, G.M., 1991, Terrestrial tetrapods and the end Permianevent: A comparison of analysis: Historical Biology,v. 5, p. 239–255.

King, G.M., 1992, Ecological aspects of methane oxidation,a key determinant of global methane dynamics: Ad-vances in Microbial Ecology, v. 12, p. 431–468.

Kingston, J.D., Marino, B.D., and Hill, A., 1994, Isotopicevidence for Neogene Hominid paleoenvironments inthe Kenya rift valley: Science, v. 264, p. 955–959.

Krull, E.S., 1998, Paleoenvironmental and carbon isotopicstudies (d13Corg) from terrestrial and marine strataacross the Permian-Triassic boundary in Antarcticaand New Zealand [Ph.D. dissert.]: Eugene, Universityof Oregon, 208 p.

Krull, E.S., 1999, Latitudinal influence on the Permian-Tri-


KRULL AND RETALLACK

assic carbon isotopic excursion: Evidence for methanerelease in high latitudes: Geological Society of Amer-ica Abstracts with Programs, v. 31, no. 7, p. 359.

Krull, E.S., Retallack, G.J., Campbell, H.J., and Lyon, G.L.,2000, d13Corg chemostratigraphy of the Permian-Tri-assic boundary in the Maitai Group, New Zealand:Evidence for high-latitudinal methane release: NewZealand Journal of Geology and Geophysics, v. 43, p.21–32..

Kvenvolden, K.A., 1993, Gas hydrates: Geological per-spective and global change: Reviews in Geophysics,v. 31, p. 173–187.

Kyle, R.A., 1977, Palynostratigraphy of the Victoria Groupof South Victoria Land, Antarctica: New ZealandJournal of Geology and Geophysics, v. 20, p. 1081–1102.

Kyle, R.A., and Schopf, J.M., 1982, Permian and Triassicstratigraphy of the Victoria Group, TransantarcticMountains, in Craddock, C., ed., Antarctic geoscience:Madison, University of Wisconsin Press, p. 649–659.

Lichtfouse, E., Berthier, G., Houot, S., Barriuso, E., Ber-gheaud, V., and Vallaeys, T., 1995a, Stable carbon iso-tope evidence for the microbial origin of C14-C18 n-alkanoic acids in soils: Organic Geochemistry, v. 23,p. 849–852.

Lichtfouse, E., Dou, S., Girardin, C., Grably, M., Balesdent,J., Behar, F., and Vandenbroucke, M., 1995b, Unex-pected 13C-enrichment of organic components fromwheat crop soils: Evidence for in situ origin of soilorganic matter: Organic Geochemistry, v. 23, p. 865–868.

Lichtfouse, E., Chenu, C., Baudin, F., Leblond, C., Da Sil-va, M., Behar, F., Derenne, S., Largeau, C., Wehrung,P., and Albrecht, P., 1998, A novel pathway of soilorganic matter formation by selective preservation ofresistant straight-chain biopolymers: Chemical andisotope evidence: Organic Geochemistry, v. 28, p.411–415.

MacLeod, K.G., Smith, R.M.H., Koch, P.L., and Ward, P.D.,1997, Multiple carbon isotopic excursions in the Per-mo-Triassic boundary interval of South Africa: Geo-logical Society of America Abstracts with Programs,v. 29, no. 6, p. 462.

Magaritz, M., 1989, 13C minima follow extinction events:A clue to faunal radiation: Geology, v. 17, p. 337–340.

Magaritz, M., Krishnamurty, R.V., and Holser, W.T., 1992,Parallel trends in organic and inorganic carbon iso-topes across the Permian/Triassic boundary: AmericanJournal of Science, v. 292, p. 727–739.

Martin, A., and Haider, K., 1986, Influence of mineral col-loids on the turnover rates of soil organic carbon, inHuang, P.M.. and Schnitzer, M., eds., Interactions ofsoil minerals with natural organics and soil microbes:Soil Science Society of America Special Publication17, p. 283–304.

McKirdy, D.M., and Powell, T.G., 1974, Metamorphic al-teration of carbon isotopic composition in ancient sed-imentary organic matter: New evidence from Australiaand South Africa: Geology, v. 2, p. 591–595.

Moore, S.E., Ferrell, R.E., and Aharon, P., 1992, Diageneticsiderite and other ferroan carbonates in a modern sub-siding marsh sequence: Journal of Sedimentary Pe-trology, v. 62, p. 289–310.

Mora, C.I., Driese, S.G., and Colarusso, L.A., 1996, Middleto late Paleozoic atmospheric CO2 levels from soil car-bonate and organic matter: Science, v. 271, p. 1105–1107.

Morante, R., 1996, Permian and Early Triassic isotopic rec-ords of carbon and strontium in Australia and a sce-nario of events about the Permian-Triassic boundary:Historical Biology, v. 11, p. 289–310.

Mortland, M.M., 1986, Mechanisms of adsorption of non-humic organic species by clays, in Huang, P.M., andSchnitzer, M., eds., Interactions of soil minerals with

natural organics and microbes: Soil Science Societyof America Special Publication 17, p. 59–76.

Mundil, R., Ludwig, K.R., and Renne, P.R., 1999, Highresolution U/Pb data for the Permian-Triassic transi-tion at Meishan, China: Geological Society of Amer-ica Abstracts with Programs, v. 31, no. 7, p. 289.

Nadelhoffer, K.J., and Fry, B., 1988, Controls on naturalnitrogen–15 and carbon–13 abundances in forest soilorganic matter: Soil Science Society of America Jour-nal, v. 52, p. 1633–1640.

Neue, H.U., Becker- Heidmann, P., and Scharpenseel, H.W.,1990, Organic matter dynamics, soil properties, andcultural practices in rice lands and their relationshipto methane production, in Bouwman, A.F., ed., Soilsand the Greenhouse effect: New York, John Wiley andSons, p. 457–466.

O’Brien, B.J., and Stout, J.D., 1978, Movement and turn-over of soil organic matter as indicated by carbon iso-tope measurements: Soil Biology and Biochemistry,v. 10, p. 309–317.

Perry Burhan, R.Y., Adam, P., Trendel, J.M., Albrecht, P.,and Nissenbaum, A., 1995, Paleoenvironmental stud-ies of the Be’eri sulfur deposit: Structural and carbonisotopic aspects, in Grimalt, J.A., and Dorronsoro, C.,eds., Organic geochemistry: Developments and appli-cations to energy, climate, environment and humanhistory: Donostia–San Sebastian, Italy, A.I.G.O.A.(Iberian Association of Environmental and OrganicGeochemistry), p. 67–69.

Pessenda, L.C.R., Aravena, R., Melfi, A.J., Telles, C.C.,Boiulet, R., Valencia, E.P.E., and Tomazello, M.,1996, The use of carbon isotopes (13C, 14C) in soil toevaluate vegetation changes during the Holocene incentral Brazil: Radiocarbon, v. 38, p. 191–201.

Prieme, A., and Christensen, S., 1997, Seasonal and spatialvariation of methane oxidation in a Danish spruce for-est: Soil Biology and Biochemistry, v. 29, p. 1165–1172.

Quade, J., Cater, J.M.L., Ojha, T.P., Adam, J., and Harrison,T.M., 1995, Late Miocene environmental change inNepal and the northern Indian subcontinent: Stableisotopic evidence from paleosols: Geological Societyof America Bulletin, v. 107, p. 1381–1397.

Renne, P.R., Zichao, Z., Richards, M.A., Black, M.T., andBasu, A.R., 1995, Synchrony and causal relations be-tween Permian-Triassic boundary crisis and Siberianflood basalt volcanism: Science, v. 269, p. 1413–1416.

Retallack, G.J., 1988, Field recognition of paleosols, inReinhardt, J., and Sigleo, W.R., eds., Paleosols andweathering through geologic time: Principles and ap-plications: Geological Society of America Special Pa-per 216, p. 1–20.

Retallack, G.J., 1995, Permian-Triassic life crisis on land:Science, v. 267, p. 77–80.

Retallack, G.J., 1997, Earliest Triassic origin of Isoetes andquillwort evolutionary radiation: Journal of Paleontol-ogy, v. 71, p. 500–521.

Retallack, G.J., and Alonso-Zarza, A.M., 1998, Middle Tri-assic paleosols and paleoclimate of Antarctica: Journalof Sedimentary Research, v. 68, p. 169–184.

Retallack, G.J., and Dilcher, D.L., 1988, Reconstruction ofselected seed ferns: Missouri Botanical Gardens An-nals, v. 75, p. 1010–1057.

Retallack, G.J., and Krull, E.S., 1999, Landscape ecologicalshift at the Permian-Triassic boundary in Antarctica:Australian Journal of Earth Sciences, v. 46, p. 785–812.

Retallack, G.J., Krull, E.S., and Robinson, S.E., 1995,Permian and Triassic paleosols and paleoenvironmentsof southern Victoria Land, Antarctica: U.S. AntarcticJournal, v. 30, p. 33–36.

Retallack, G.J., Krull, E.S., and Robinson, S.E., 1996a,Permian and Triassic paleosols and paleoenvironmentsof the central Transantarctic Mountains, Antarctica:U.S. Antarctic Journal, v. 31, p. 29–32.

Retallack, G.J., Veevers, J.J., and Morante, R., 1996b,

Global coal gap between Permian-Triassic extinctionand Middle Triassic recovery of peat-forming plants:Geological Society of America Bulletin, v. 108, p.195–207.

Retallack, G.J., Seyedolali, A., Krull, E.S., Holser, W.T.,and Ambers, C.P., 1998, Search for evidence of impactat the Permian-Triassic boundary in Antarctica andAustralia: Geology, v. 26, p. 979–982.

Saari, A., Martikainen, P.J., Ferm, A., Ruuskanen, J., deBoer, W., Troelstra, S.R., and Laanbroek, H.J., 1997,Methane oxidation in soil profiles of Dutch and Finn-ish coniferous forests with different soil texture andatmospheric nitrogen deposition: Soil Biology andBiochemistry, v. 29, p. 1625–1632.

Schopf, J.M., and Long, W.E., 1966, Coal metamorphismand igneous associations in Antarctica, in Gould, R.F.,ed., Coal science: American Chemical Society Ad-vances in Chemistry Series 55, p. 156–195.

Sikes, N.E., 1994, Early hominid habitat preferences in EastAfrica: Paleosol carbon isotopic evidence: Journal ofHuman Evolution, v. 27, p. 25–45.

Soil Survey Staff, 1997, Keys to soil taxonomy: PocahontasPress, Blacksburg, Virginia, 544 p.

Stanley, S.M., and Yang, X., 1994, A double mass extinc-tion at the end of the Paleozoic Era: Science, v. 266,p. 1340–1344.

Stevenson, F.J., 1969, Pedohumus: Accumulation and dy-namics during the Quaternary: Soil Science, v. 107,p. 470–479.

Stevenson, F.J., ed., 1986, Cycles of soil: New York, JohnWiley & Sons, 380 p.

Stout, J.D., and Rafter, T.A., 1978, The 13C/12C isotopic ra-tios of some New Zealand tussock grassland soils:Stable isotopes in the Earth Sciences: New ZealandDepartment of Scientific and Industrial Research,DSIR (Department of Scientific and Industrial Re-search) Bulletin 220, INS (Institute of Nuclear Sci-ences) Contribution no. 806, 75–83.

Stout, J.D., Rafter, T.A., and Troughton, J.H., 1975, Possi-ble significance of isotopic ratios in paleoecology, inSuggate, R.P., and Cresswell, M.J., eds., Quaternarystudies: Selected papers from the IX INQUA Con-gress, Christchurch, New Zealand: Royal Society ofNew Zealand Bulletin 13, p. 279–286.

Stout, J.D., Goh, K.M., and Rafter, T.A., 1981, Chemistryand turnover of naturally occurring resistant organiccompounds in soil, in Paul, E.A., and Ladd, J.N., eds.,Soil biochemistry: New York, Marcel Dekker, p. 19–24.

Strauss, H., Des Marais, D.J., Hayes, J.H., and Summons,R.E., 1992, The carbon-isotope record, in Schopf,J.W., and Klein, C., eds., The Proterozoic biosphere:Cambridge, Cambridge University Press, p. 117–128.

Tyler, S.C., Crill, P.M., and Brailsford, G.W., 1994, 13C/12Cfractionation of methane during oxidation in a tem-perate forested soil: Geochimica et CosmochimicaActa, v. 58, p. 1625–1633.

Veevers, J.J., Conaghan, P.J., and Shaw, S.E., 1994, Turningpoint in Pangean environment at the P/Tr boundary,in Klein, G.D., ed., Pangea: Paleoclimate, tectonics,and sedimentation during accretion, zenith, and break-up of a supercontinent: Geological Society of AmericaSpecial Paper 288, p. 187–196.

Wang, K., Geldsetzer, H.H.J., and Krouse, H.R., 1994,Permian-Triassic extinction: Organic d13C evidencefrom British Columbia, Canada: Geology, v. 22, p.580–584.

Wedin, D.A., Tieszen, L.L., Dewey, B., and Pastor, J., 1995,Carbon isotope dynamics during grass decompositionand soil organic matter formation: Ecology, v. 76, p.1383–1392.

MANUSCRIPT RECEIVED BY THE SOCIETY JUNE 29, 1998REVISED MANUSCRIPT RECEIVED NOVEMBER 29, 1999MANUSCRIPT ACCCEPTED JANUARY 26, 2000

Printed in the USA

d C depth proﬁles from paleosols across the …...Thus, paleosol d13C variation with soil depth in...

Documents

Transcript of d C depth proﬁles from paleosols across the …...Thus, paleosol d13C variation with soil depth in...