PALEOCEANOGRAPHY, VOL. 16, NO. 2, PAGES 155-163, APRIL 2001 An osmium isotopeexcursionassociatedwith the late Paleocene thermal maximum: Evidenceof intensifiedchemicalweathering G. Ravizza,R. N. Norris,andJ. Blusztajn Department of GeologyandGeophysics, WoodsHoleOceanographic Institute,WoodsHole,Massachusetts M.-P. Aubry InstitutdesSciencesde l'Evolution,Universit6MontpellierII, Montpellier,France Abstract.In thelatestPaleocene anabruptshifttomorenegative •5•3C valueshasbeendocumented atnumerous marineand terrestrialsites[Braloweret al., 1997; Crameret al., 1999;Kaiho et al., 1996;Kennettand Stott, 1991; Koch et al., 1992; Stottetal., 1996;Thomas andShackleton, 1996;Zachos etal., 1993].Thiscarbon isotope event(CIE) iscoincident withoxygenisotopedatathatindicatewarmingof surfacewatersat highlatitudesof nearly4ø-6øC [Kennettand Stott,1991]and moremoderate warmingin thesubtropics [Thomas etal., 1999].Herewereport•87Os/•88Os isotope records fromtheNorth Atlantic and Indian Oceans which demonstratea >10% increase in the •87Os/•88Osratio of seawater coincident with the late Paleocene CIE.Thisexcursion tohigher•87Os/•88Os ratiosisconsistent witha globalincrease in weathering rates.Theinferenceof increased chemical weathering duringthisintervalof unusual warmthis significant because it provides empirical evidencesupporting theoperationof a feedbackbetweenchemicalweathering ratesandwarmglobalclimate,whichactsto stabilizeEarth's climate [Walker et al., 1981]. Estimatesof the durationof late PaleoceneCIE [Bainset al., 1999; Bralower et al., 1997;NorrisandROhl,1999;R•hl et al., 2000]in conjunction withtheOsisotopedataimplythatintensified chemical weathering inresponse towarm,humidclimates canoccurontimescales of 104-105 years.Thisinterpretation requires thatthe latePaleocene thermalmaximumOs isotopeexcursion be producedmainlyby increased Os flux to the oceanratherthana transient excursion tohigher•87Os/•88Os ratiosinriverrunoff.Although weargue thattheformerismorelikelythanthelatter,we cannotruleoutsignificant changes in the•87Os/•88Os ratioof rivers. thetemperature teedback on silicateweathering rateis largely unimportant in regulating globalclimate[Edmond andHuh,1997]. The interplaybetweenglobaltemperature andchemicalweather- Inresponse, others havenoted thatintheabsence ofaWalker-style ing ratesis a criticalpartof theongoingdebateovertheunderlying weathering feedback, thereis no mechanism to preventextreme 1. Introduction mechanisms thatbothperturbandstabilizeEarth'sclimatesystem. The hypothesis that a negativefeedbackbetweenglobaltemperatureandsilicateweatheringratesactsto stabilizethe long-termclimateof the Earth [Walkeret al., 1981] is appealingbecauseof its simplicitybut is extremelydifficultto test.The feedbackoperates suchthat rising temperatureacceleratessilicate-weathering rates (schematically represented by the Urey reaction,CaSiO3+ CO2-• SiOn_ + CaCO3),causinga drawdownof atmospheric CO2.This in turn leadsto coolertemperatures via diminishedgreenhouse warming. The idea that chemicalweatheringplaysan importantrole in regulatingEarth's climatehistoryledto thedevelopment of a series of modelswhichhindcastlong-term(i.e., Phanerozoic) changesin atmosphericCO2 content based on geochemicalmass balance variationsin the CO2 contentof the ocean-atmosphere system [Bickleet al., 1998;BroeckerandSanyal,1998]. ThelatePaleocene thermalmaximum [Zachos etal., 1993]representsa uniqueopportunity to examinetheresponse of the Earth systemto unusualwarmth.This eventoccurred-55millionyears ago [Norris and ROhl, 1999]. It is identified in the marine record by an abruptnegativecarbonisotopeexcursioncoincidentwiththe extinction of 35-50%of deep-sea foraminifera [Aubryetal., 1998; Thomasand Shackleton, 1996].Associated oxygenisotopedata indicateseasurfacetemperatures at highlatitudes increased by as muchas4ø-6øC [Kennettand Stott,1991]duringthecarbonisotope event (CIE) with more modestwarmingat lower latitudes [Thomas et al., 1999].Terrestrial records indicateunusually warm calculations[Berner, 1994, and referencestherein]. Interest in the and humidconditionsas well [Schroeder,1992].The durationof the LPTM is difficult to constrain.Interpolationbetweenstratipothesisthat orogenesiscould acceleratethe above reaction and graphicdatums[Braloweret al., 1997],modelsof thecarboncycle overwhelmthesupposed negativetemperature feedback,givingrise whichattributetheCIE to methanehydratedecomposition [Bains to glaciation[Raymoand Ruddiman,1992].Recentstudiesof mod- et al., 1999;Dickenset at., 1997],andastronomically tunedestiern riversshowthatCO2consumption duringchemicalweathering matesbasedon cyclostratigraphy[Norris and R6hl, 1999; ROhl is stronglydependent uponlithologyandtectonicsettingswith lit- eta!., 2000] suggests the durationof theCIE is roughly2 x 10s tle or nodependence on temperature, leadingto theconclusion that years.The high-resolution carbonisotopedataandcyclostratigraphy resultsalsoindicatethat the initial excursionto light carbon historyof silicateweatheringwas furtherheightenedby the hy- Copyright 2001by theA•nerican Geophysical Union. isotoperatiosoccurred in <104years.If theweathering feedback proposed by Walkeret al. [1981]playsan importantrolein regu- Papernumber 2000PA000541. 0883-8305/01/2000PA000541512.00 the late Paleoceneshouldbe accompanied by a net increasein lating Earth's climate on this timescale,then the unusualwarmth of globalsilicateweathering. 155 156 RAVIZZA ET AL.' LPTM Os ISOTOPE EXCURSION The marineosmium(Os) isotoperecordis theonlypaleoceanographictool currentlyavailablethat hasthe potentialto constrain changesin chemicalweatheringassociated with brief climaticexcursionssuchas the LPTM. Both pelagicclaysand metalliferous carbonates preservea recordof pastvariationsin the Os isotopic compositionof seawater[Oxburgh, 1998; Pegram eta!., 1992; Pegram and Turekian, 1999; Peucker-Ehrenbrinket al., 1995; Ravizza, 1993; Reuschet al., 1998]. At any point in time, the •a7Os/•aaOs ratio of seawaterreflectsthe balancebetweensources of Os with low ratios,suchas hydrothermaland extraterrestrial sources,andcontinentalsourceswith muchhigher•a7Os/•aaOs ratios.In thisrespect,themarineOsisotoperecordis analogous to the marineSr isotoperecord,a widelyexploitedrecordof thegeologic historyof chemicalweathering, withbothsystems providingglobally integrated recordsof dissolved inputto theoceanthroughtime. The marineresidence timeof Sr is approximately -2 x 106 years [PalmerandEdmond,1992],toolongto recordshort-duration perturbationsin oceanchemistryrelatedto intensifiedsilicateweathering. The marineresidencetime of Os is 2-3 ordersof magnitude 0.5 1.0 1.5 2.0 2.5 143 i .... I DSDP 213 144--.. 145 shorterthanthat of Sr [Levasseur et al., 1999;Oxburgh,1998; Peucker-Ehrenbrink and Rav&za, 1996;Sharmaet al., 1999].Consequently,it is possiblethat the marineOs isotoperecordcould preserveevidence of transientperturbationsin global chemical weathering rates.Givenourcurrentunderstanding of themarineOs cycle,a pulseof intensified chemicalweathering is expected to pro- duceanexcursion to higher•aVOs/•SaOs ratiosfollowedbya recovery to lower ratios.This studywas designedto seekevidenceof suchan Os isotopeexcursionin the late Paleocene. 147 2. Samples In order to construct a record of 187Os/•a8Os variations acrossthe LPTM eventwe haveanalyzed sediments recovered fromtwoDeep Sea Drilling Project (DSDP) sites. One is locatedin the Indian Ocean (Site 213), and the other is locatedin the North Atlantic I 148 l........ -- ....., , I , , , •__l , , , , -2 -1.5 -1 -0.5 (Site 549). Site 213 wasselectedfor initial analysisbecause it is lithologicallysimilarto basalmetalliferouscarbonates whichaccumulate on recently formed oceaniccrust and have been demon- strated torecorddetailedlaVOs/•gaOs variations of Neogene seawater [Oxburgh, 1998; Peucker-Ehrenbrinket al., 1995; Ravi:•a, 1993; Reuschet al., 1998].Nannofossil [AttbryandSanfilippo,1999]and foraminifera[BerggrenandNorris, 1997]biostratigraphy at DSDP 213 are consistent with a late Paleoceneage.Bulk carbonate 8•-•C datafromSite213 showthatthesesediments capturethelatterpart of theCIE butnottheonsetof theevent(Figure1).• Sediments from DSDP Site549 wereselected for furtherinvestigation because previouswork demonstrates this sequence recordsthe CIE associated withtheLPTM [Attbry,1998;Stottet al., 1996).Lithologically, the sediments fromDSDP Site549 areredto buffcoloredclaybearing nannofossil oozes,suggesting a significant component of hydrogenousFe oxides.Thesesedimentsare not, however,analogousto basalmetalliferoussedimentswhichhavebeenusedin previousOs isotopestudies because theyarelocatedfar abovebasement andpresumablyaccumulated far removedfrommid-oceanridgehydrothermal activity. •Supporting datafor FigureI is availableelectronically at World Data Center-A for Paleoclimatology,NOAA/NGDC, 325 Broadway, Boulder, CO 80303 (e-mail: [email protected]; URL: http://www.ngdc. noaa.gov/paleo). • (•180 %0 Figure1. Carbonandoxygenisotope datafromanalyses of bulkcarbonate fi'omDeepSeaDrilling Project(DSDP) Site213. The low 15•Cratiosat the bottomof thisrecordcorrespond to thelatterhalfof thecarbon isotope event(CIE), whichdefinesthe latePaleocene thermalmaximum(LPTM). Depthis reportedin metersbelowseafloor(mbsf). 3. Methodology Osisotope analyses of bulksediments wereperformed bymagneticsector inductively coupled plasma source-mass spectrometry (ICP-MS).In thisstudy,cellulose filterpapers containing theOs preconcentrated by NiS fire assayweredigestedin closedTeflon vesselswith 1 mL of nitric acid and then diluted fivefold with water. Oswasseparated fromthissolution andintroduced directly intotheplasma asOsO4 bybubbling theAr sample gasthrough the solution [Hassleretal., 2000].Themainadvantage of thismethod relative to morecommon thermal ionization methods is thespeed of analysis. An individual Osisotope analysis ontheICPMSrequires-15 minof instrument time.Thepotential of thismethodto enhance sample throughput is important forhigh-resolution paleoceanographic studies. Analyses wereperformed overa periodof 6 months onfourseparate days.Onthesedaysa totalof 11analyses RAVIZZA ET AL.' LPTM Os ISOTOPE EXCURSION 157 Table la. Re-Osand Bulk CarbonateC andO IsotopeData for Site213• Sample 16R-I, 102-106 16R-I, 141-144 16R-2 36-40 16R-2 100-104 16R-2 140-143 16R-2 140-143 16R-3 41-44 16R-3. 100-101 16R-3, 140-143 16R-4. 36-39 16R-4. 73-75 16R-4. 93-95 16R-4, 101-103 16R-4, 122-125 Depth, Re, Os, mbsf ppt ppt 143.02 42 143.4l 143.86 144.50 144.90 144.90 145.41 146.00 146.40 146.86 147.23 147.43 147.51 147.72 33 47 46 34 25 31 136 52 38 83 192 233 354 126 42 132 295 167 ••?Re/• aaOs 13.6 2.5 3.4 6.3 2.4 ••?Os/• •Os b•aC, b•80, % % 0.4089 _+0.0063 2.13 - 1.101 0.3494 0.3416 0.3530 0.3465 0.3443 0.3578 0.3628 0.3417 0.3494 0.3748 0.4017 0.4124 0.4015 2.156 2.23 2.201 2.17 -1.014 -0.94 -0.776 -0.769 2.032 2.038 -0.905 -0.666 1.917 1.724 1.419 1.096 0.92 -0.781 _+0.0023 _+0.0028 _+0.0057 _+0.0039 _+0.012 _+0.004 _+0.0044 _+0.0026 _+0.0022 _+0.0036 _+0.0025 _+0.0119 _ 0.0033 -0.784 - 1.109 -1.539 - 1.887 "Repol'ted uncertainties in •87Os/•8•Os ratiosarebasedon2{5in-runcounting statistics. Notethatthecomplete bulkcarbonate C andO isotope datasetis not tabulatedbut is plottedin Figure 1. Here mbsfis metersbelowseafloorandppt is partsper thousand. of between400 and 1250 pg of Os standardwereanalyzed.After discardingoneoutlier,the mean•87Os/•8•Os ratioaftercorrection for instrumentalmassbias was 0.1740 +_0.0006 (1 standarddeviation). This valuecompareswell with the acceptedvalueestablished by repeated thermal ionization analysis of the same solution (0.1741 +_0.0003;1 standarddeviation)[Hassleret al., 2000]. Mea- sured•9øOs/•Osratiosfor thestandard analyses (1.9864+_0.0035; I standarddeviation)werealsoindistinguishable from theaccepted value (1.9840), indicatingthat thereis no detectablecarryoverof Os betweensampleand standardanalyses.If carryoverdid occur, elevated •9øOs/•88Osratios should be measured in standards be- causeall samplesare spikedwith •9øOsin orderto determineOs concentrations by isotopedilution. Between2 and5 g of sedimentwerefusedfor eachanalysis.The reagentsused in the NiS fire assaydetermineproceduralblanks. The magnitudeof the fusion blank correctionis variable because the massratio of sedimentto reagentsvaried amongthe sample suite. In all samplesfrom Site 549 and in those samplesfrom Site 213 representative of the LPTM, proceduralblank corrections were 2.2% of the total Os concentration or less. In the low Os con- centrationsamplesfrom Site 213, blank correctionwere aslargeas 12%. Uncertainties 4. Results Measured 187Os/188Os ratiosat thetwo sites(Tablesla andlb) rangefrom0.34to0.44.Thesevaluesaresimilartothosepreviously reportedfor Pacific metalliferoussediments[Peucker-Ehrenbrink et al., 1995]andPacificpelagicclays[PegramandTurekian,1999] of Paleoceneto Eoceneage. Os concentrations of bulk sediment fromSites213 and549 aresimilarto thosepreviously measured in metalliferous carbonates whichhavebeenshownto recordpast changes in theOs isotopiccomposition of seawater[Oxburgh,1998; Pegrain and Turekian, 1999; Peucker-Ehrenbrinketa!., 1995; Ravizza, 1993;Reuschet al., 1998]. MeasuredRe/Os ratiosare low enoughthatcorrections for in situdecayof •8?Re areunimportant. The 187Os/188Os variationsare well correlatedwith variationsin stablecarbonisotopes.At boththe North Atlantic andIndianOcean sitesthe highest187Os/•88Os ratiosarecloselyassociated withthe lowest813Cvalues(Figure2). As81qCshiftstowardhighervalues nearthe topof thecarbonisotopeexcursion, •87Os/•88Os ratiosdecline to lower values.The amplitudeof the excursioncapturedat bothsitesis -0.06 187Os/188Os units,an increaseof >10% relativeto post-LPTM ratios. The North Atlantic record (Site 549), which associated with fusion blank corrections are the spans thecarbonisotope anomaly, showsthat187Os/•88Os isotope ra- singlemostimportantfactor limiting the precisionand accuracyof theselow-concentrationsamples. tiosrecoverto valuessimilarto pre-LPTM (Figure2b) values,indicating a transientperturbationof the marine Os cycle. There is a Table lb. Re-Os and Bulk CarbonateC and O IsotopeData for Site 549• Sample Depth, Re, Os, mbsf ppt ppt •513C, •8?Re/• 88Os •7Os/•Os % •180, % 16R-4, 68-72 16R-4, 98-102 336.68 336.98 35 41 145 129 1.2 1.6 0.3764 _+0.0031 1.457 -0.53 0.3773 _+0.0016 1.388 -0.527 16R-5, 16R-5, 16R-5, 16R-5, 16R-5, 16R-6, 16R-6, 16R-6, 16R-6, 16R-6, 337.29 337.47 337.85 338.17 338.47 338.73 338.96 339.33 339.66 339.97 34 138 119 132 141 159 199 288 243 172 259 1.2 0.3839 0.3779 0.3856 0.4212 0.4350 _+0.0013 _+0.0013 _+0.0017 _+0.0021 _+0.0012 1.192 -0.846 1.147 -0.881 O.637 -1.025 0.4387 _+0.0018 -2.816 0.4000 _+0.0023 -0.473 0.3883 _+0.0021 1.879 -1.003 0.3666 _+0.0073 2.293 -0.651 0.3633 _+0.0014 2.191 -0.425 11-15 29-32 67-71 99-103 129-133 5-8 28-33 65-69 98-102 129-131 69 44 77 50 49 72 65 2.6 1.6 2.4 1.3 0.8 1.5 1.9 "Reported uncertainties in 187Os/188Os ratiosarebasedon2c5in-runcounting statistics. -0.056 -1.764 -0.721 -2.022 0.627 -1.169 158 RAVIZZA ET AL.' LPTM 54.3 54.4 Os ISOTOPE EXCURSION AGE (Ma) (a) 54.5 2'41'[[ 54.6 [ 54.7 ] 54.8 ] ]t __•_,•0..0• •;)•!180½lC]•t.t•BuIk Carbonate 0.44 0.42 (• 1.6!' 0.40 '- 0.38 o r,,. 0.36 0.8 : ..... • , , ,O 143 144 145 146 0.34 148 147 Depth (mbsf) AGE (Ma) 54.6 54.7 54.8 54.9 , , , , (b) 55.0 55.1 • M. subbotinae 'l•Et•'•I;• / CO-/tEl' ©• - 0.44 0.42 4-.m B.4a --E•. • o 2 ---BulkCarbonate /" O •• •, \ O _ \ 0 - 187 Os/188 Os -• / -2-' ,, 336 ,, • •, 337 • \/ , , • , , ,•, 338 oo - o - 0.38 r,,. _ (20 - 0.36 '- _ • .... 339 • .... 340 0.34 341 Depth (mbsf) Figure2. Comparison of bulksediment [•?Os/•Osvariations to carbonisotopevariations in (a) Site213 and(b) Site549.Carbon isotopedatafor theplanktonicforaminiferM. subbotinae at Site549 arefi'omStottet al. [1996].Notethatthelargest[87Os788Os ratiosare associated with the lightestcarbonisotoperatiosat bothsitesandthat the amplitudeof Os isotopeexcursionis similarat bothsites.Depthis reportedin metersbelowseafloor(mbsf).The lightcarbonisotoperatiosthatcorrespond to the CIE providethe basisfor correlatingthe two Os isotoperecords.Age assignments are baseduponthe carbonisotopestratigraphy andan ageof 55 Ma for the initiationof theCIE [NorrisandRShl, 1999]andan estimateddurationfor theCIE of 220 kyr. [Rfihlet al., 2000]. These ageassignments are intendedto facilitatecorrelationof the two recordsbasedon our interpretation of the carbonisotopedata.Note thatthe 55 Ma agefor the LPTM conflictswith currentlyassignedagesfor nearbybiostratigraphic andmagnetostratigraphic datums [Berggrenet al., 1995].This inconsistency reflectsthe needfor an improvedand internallyconsistent geochronology for the upper Paleoceneto the lowermostEocene[Attb•3, et al., 1998]. systematic offsetin [•?Os/•SOs ratiobetweenthetwositesof-0.03 •?Os/•Os units,with the North Atlantic beingmoreradiogenic than the Indian Ocean. The highestOs concentrationsat both sites occurcloseto butnotexactlycoincident withthehighest •7Os/•8Os ratios. Bulk sediment Os concentrations and •?Os/•S•Os ratios do not show the hyperbolicrelation requiredby two-componentphysical mixing models.This indicatestwo-componentphysicalmixing of isotopicallydistinctend-membersis not controllingthe observed Os isotopevariations. 5. Discussion 5.1. A Seawater Os Isotope Record of the LPTM from DSDP Sites 213 and 549 We interpretthe •7Os/•Os variationsat Sites213 and549 asa recordof the Os isotopiccompositionof seawateracrosstheLPTM. Severallinesof argumentsupportth•s•nterpretation. First,by analogy to other marine sedimentsused to reconstructthe seawater Os isotoperecord [Ravizza, 1993], the Os concentrations at both sitesare sufficientlyhighthathydrogenous (i.e., seawaterderived) Os shoulddominatethe sedimentOs budget.Second,the absolute •?Os/•S•Osratios are similar to one anotherand to existingdata fromPacificsedimentsof similarage(seeTable2 anddiscussion of LL44-GPC3 in section).It is unlikely that sucha similarity would resultas a consequence of physicalmixing of isotopicallydistinct components. Third, the Os isotopeexcursionat bothsitesmustbe nearly synchronousbecauseit coincidesso with the CIE. Fourth, theamplitudeof the •?Os/•SOsexcursion, asdefinedby themaximum 187Os/188Os and the post-CIEplateau,at thesewidely separatedsitesis essentiallythe same. As notedabove,there is a systematicoffset betweenthe North AtlanticandIndianOceansitesof-0.03 187Os/•8Os units.It is pos- RAVIZZA ET AL.: LPTM Os ISOTOPE EXCURSION Table 2. Summaryof •S7Os/•SSOs Variationsin Cores From the Pacific, Indian, and Atlantic Oceans Location/Site North Pacific/LL44-GPC3 Indian Ocean/DSDP 213 North Atlantic/DSDP 549 Approximate Duration" -- 3 million years -- 220,00 years = 220,00 years 159 5.2.1. Changes in total Os flux to the ocean associatedwith enhanced continental input during the LPTM. The simplest interpretationof the LPTM Os isotope shift is that it reflects an 187Os?SSOsepisode of increased chemical weathering on the continents, which resultsin an increasein the total flux of radiogenicOs to 0.32-0.41 0.34-0.41 0.36-0.44 the oceans. A crude estimate of the relative increase in Os flux can beobtainedby considering the •87Os/•8SOs ratioof seawater to be the result of mixing betweenunradiogenichydrothermalOs and "Duration estimates are based on sedimentation rate estimates for GPC3 radiogeniccontinentalOs with fixed •7Os/•Os for both end[Janecekand Rea, 1983] and the estimatedduration of LPTM carbon members. At steadystatethemolefractionF of 188Os in seawater isotopeevent [RiJhleta!., 2000]. derived fi'om continentalweatheringis sible that this difference is the result of a subtle offset between the F = (R,- R•,)/(R,- R•,), (1) •87Os/1S8Os of thebulksediment andthatof contemporaneous seawhereR denotesthe iS7Os/188Os ratioandthesubscripts s, h, andr denoteseawater,hydrothermal,and rivers, respectively.To comOceans.Analysesof the surfacesof manganesecrustsindicatea paretherelativeincreasein continentalOs duringtheLPTM peak with thepost-LPTMplateau,we assumeeachcanbe approximated similar contrast may exist in the modern ocean [Burton et al., steadystatemixtureandcalculatetheratioof themix1999]. In spiteof the uncertaintyas to the causeof this systematic asa separate offset betweenSites 213 and 549, synchronousOs isotopeexcur- ing fractionsas sionsof similar amplitudeat thesetwo widely separatedsitesproFl.l,r•/Fr•,,t_l.l,l• = [R,(LPTM)- R•,]/[R•(post-LPTM) - R•,]. (2) vide strongevidencefor a transientwhole-oceanshift to more water. It is equally likely that this differencereflectsa true contrast in the 187Os/188Osratio between the North Atlantic and Indian continent-like •87Os/•8SOs ratiosduringtheLPTM. Using•S?Os/•SSOs ratioof thehydrothermal end-member (0.12)and Theyoungest samplefromtheSite213 recordhasa •87Os/188Os rationearlyas high assamplesfrom theLPTM intervalof thiscore. It is difficult to judge the significanceof this samplebecauseit is an isolatedoccurrence.If abruptexcursionsof thissortarecommon thelS7Os/•SSOs ratiosfortheLPTM andpost-LPTM fromSite549 yieldsa fractionratioof 1.23.Thisratioindicates a 23% largercontributionof continental Os to seawaterat theheightof theLPTM comparedto the period immediatelythereafter.Using data from in the marine Os record, the likelihood that the correlatedOs and Site213 to makethe samecalculation yieldsa 26% largerconticarbonisotopevariationsdescribedaboveis fortuitousincreases. nentalOs contributionduringtheLPTM. It is alsopossiblethatthissamplerecordspartof a minorwarming 5.2.2. Changesin the •S?Os/•SSOs of rivefineinput duringthe eventfollowingthe LPTM because it coincides with a localmini- LPTM. The magnitudeof thechangein riverine•S?Os/•SSOs ratio mumin bulkcarbonate 8•sO(Figure1) andis notdiscordant in the that would have been requiredto producethe observedseawater •S?Os/•SSOs-(5 •sOcorrelation(Figure3a). This studyof theLPTM Os recordcanbe estimated by assuming otherrelevantparameters anda studyof recentglacialepisodes [Oxburgh,1998]aretheonly (F andRn) remainedconstant.As above,we assumethat the LPTM high-resolution studiesof seawaterOs isotopevariationsmadeto andtheperiodof timeimmediatelythereaftercanberepresented as date.Consequently, thereis little basisfor evaluatinghow com- separatesteady state mixtures of Os derived from continentaland monlyabruptexcursions in themarineOsisotoperecordoccur.The hydrothermalsources.In thiscase,the mixingequationis recastin relativelygoodagreement betweenNeogenerecordsbasedon dif- the form ferentsedimentcoressampledat low temporalresolution[Pegram R,,•.= R,F + Rh(I-F), (3) etal., 1992;PegrainandTurekian,1999;Peucker-Ehrenbrink et al., 1995; Ravizza, 1993; Reuschet al., 1998] suggestthat suchlarge- amplitude, abruptexcursions arenotubiquitous. Nevertheless, additionalhightemporalresolution Os isotopedatasetsareneededto better address this concern. wherethe notationusedis identicalto (1). Writing separateequations for LPTM andpost-LPTM conditionsandsubtracting themyields R,w(LPTM)- R,,(post-LPTM) = F [R,(LPTM) - R,(post-LPTM)]. (4) For a widerangeof assumed •8?Os/•a8Os for post-LPTMinput, 5.2. Cause and Significanceof the LPTM •S7Os/•SSOsExcursion Our preferredinterpretationof the seawaterOs isotoperecordis thattheexcursion to higherseawater •87Os/•SSOs ratiosresultsfrom rangingfrom0.5 to 1.5,theincreasein •a?Os/•aaOs in LPTM riverine input requiredto producethe Os isotopeexcursionobserved in seawater corresponds to a •8?Os/•SSOs ratioroughly20% larger thanpost-LPTMvalues.Giventhelargerangeof •a?Os/•a8Os ratio measured in modern rivers (0.4- 2.5) [Levasseur et al., 1999; an increasein the total flux of Os suppliedto the oceanby continental weathering. The LPTM Os isotope excursion could also Sharmaet al., 1999],we cannotruleoutthepossibilitythatchanges sive. Throughoutthis discussionwe make the implicit assumption that fluxes of unradiogenicOs to the ocean, related to seafloor hydrothermalactivity and/orcosmicdustdeposition,remainedconstantduring the LPTM becauseat present,there is no evidenceto suggesta transientreductionof theseinputs. sibleto statewith any degreeof certaintythatincreasedcontinental Os flux to the oceanduringthe LPTM to the oceancausedthe ex- in the •a?Os/•aaOs of riverineinputto the oceansis an important resultfroman increasein theaverage•87Os/•88Os ratioof soluble contributorto the LPTM Os isotopeexcursion. Os from the continents.Althoughwe considerthesetwo scenarios 5.2.3. Discerningbetweenchangesin the flux and isotopiccomseparatelybelow, these two possibilitiesare not mutually exclu- positionof rivefine Os during the LPTM. Althoughit is notpos- cursion to higher•8?Os/•88Os ratios,thereareseverallinesof argument that are supportiveof this interpretation.These include (1) independent evidenceof anaccelerated hydrologiccycleduringthe 160 (a) RAVIZZA ET AL.' LPTM EXCURSION Gibsonet al., 1993;Kaiho et al., 1996;RobertandKennett,1994]. However,the timescaleof kaolinite formationis long compared to the durationof the LPTM [Thiry, 2000]; thereforeclay mineral- 0.44 0.42 ogy cannot provide time-resolved information about continental weathering intensity. Climate simulations [Sloan and Thomas, ß 1998] indicatethat increasedprecipitationover land surfaceswas likely duringthe LPTM. Changesin nannofossilmorphologyassociatedwith LPTM suggestan adaptationtowardmore buoyant •(•• 0.40[] i forms, consistent with the inference of decreased surface water • 0.38 .... ß[] 0.36 0.34 Os ISOTOPE -2.5 -2 [][]•l ß -1.5 -1 -0.5 density[Aubryeta!., 2000]. In the contextof theseindependent lines of evidence,it is reasonableto interpretthe shift in the •?Os/•88Osratio of seawater at the LPTM as the result of increased ß [] 0 DSDP DSDP 0.5 549 213 1 (per (b)150 -t Silicate Weathering I Activation Energy for Os flux from the continents.The Os isotopedata make a significant contributionto this line of investigationbecausethey provide the only compellingevidencethat enhancedweatheringandrunoff was globally rather then locally significant. Laboratory measurementsof the activationenergy of silicate weathering[Kumpet al., 2000; Whiteet al., 1999]predicta specific relationshipbetweensilicateweatheringratesandtemperature.At both Sites 213 and 549 the 45•O of bulk carbonate and •a?Os/laaOs are correlatedwith one another(Figure 3a). Consideringoxygen isotopesasa proxyfor localtemperatureandOs isotopesasa proxy for weatheringflux, thisplot qualitativelyindicatesweatheringflux = 70 kj/mol increases with increasingtemperature. To examinethisrelationship c 100 ---• - 30 kj/mol in a more rigorousway would requireknowledgeof an average temperatureincreaserepresentative of weatheringcontinentalenvironments.We contendthat a reasonableguessat sucha temperature .•_ increaseis in the rangeof 4ø-7øC. This is basedon seasurfacetemperatureestimates[Kennettand Stott, 1991; Thomasand Shackleton, 1996; Thomas et al., 1999] showing increasesof 40-6 ø and r..) 1%4ø at high and low latitudes,respectively.This temperature rangeand the 20-30% increasein weatheringflux estimatedabove are consistentwith an effective activationenergyfor weatheringof -30 kJ/mol (Figure 3b), a value at the low end of the rangesmea0 1 2 3 4 5 6 7 suredin silicate weatheringrate experiments[White et al., 1999]. Delta T (øC) Althoughthiscomparisonis a grossextrapolation,it showsthatinterpretingthe Os isotopeexcursionas a proxyfor weatheringflux Figure3. (a) The plotof ]•7Os/•aOsversus$•O of bulkcarbonate, which yieldsan estimatedweatheringrate increasethat is consistentwith suggests thatseawater •87Os/188Os ratiosincrease withincreasing temperaavailable data constrainingboth the LPTM temperatureincrease tures.Note that bulk carbonateoxygenisotopedata are only a proxy for and the temperaturedependenceof silicateweatheringrates. local temperaturevariations.The samplefi'omSite 549 that plotsaway from the correlationis from the $•*C minimumandhasexperienced the most The abruptness of the onsetof the LPTM alsosuggests that the extensivecarbonatedissolution;percentcarbonatein this sampleis only associated Os isotopeexcursionis morelikely relatedto changing i x 50 -'t- 3% versus40-50% aboveand below the CIE. Thus we suspectthe larger ratherthanchanging•7Os/•Os $•aOof thissamplereflectsthe influenceof dissolution-resistant benthic totalOs flux fi'omthe continents of river input. The short and transientnatureof the Os isotope excursionassociated with theLPTM makesit unlikelythattectonic ciated with the LPTM, which are used to define the dark field labeled processescould producethe observedOs isotopicexcursionby "LPTM." The curvesrepresentthe calculatedtemperaturedependenceof varying the natureof the exposedlithologiesduring the course silicateweatheringratesbasedon activationenergiesmeasuredin the laboof the LPTM. This is a fundamentallyimportantcontrastbetween ratory.The overlapof the LPTM field with the 30 kJ/toolactivationenergy curveindicatesthat Os isotope-based weatheringflux estimatesare consis- interpretingshort-termvariationsin the Os isotopiccomposition tent with available data constrainingboth the LPTM temperatureincrease of seawater,as we do here, and interpretinglong-termCenozoic andthe temperaturedependenceof silicateweatheringrates. trends. There is one particular scenario of variable rivefine foraminifer.(b) The percentincreasein weatheringflux, inferredfi'om Os isotopedata, and the estimatedaveragesurfacetemperatureincreaseasso- •a?Os/•88Os throughtheLPTM thatwarrants specificconsideration: LPTM, (2) the magnitudeof theincreasein temperature andweatheringflux estimatedfor theLPTM, and(3) the abruptonsetof the LPTM. Each of these is discussed in turn below. the possibilitythat the Os isotopeexcursionmight reflect a brief episode of erosion of sedimentary organic matter with large •8?Os/•aaOs ratios.The clearcorrelationof high •a?Os/•aaOs with lightiS•*Cvalues(Figure2) motivates usto address thisscenario ex- Severalindependent linesof evidencesuggestan accelerated plicitly becausea similarcorrelationis apparentin theNeogeneOs hydrologiccycleoperatedduringtheLPTM. Many workershave isotoperecord [Ravizza, 1993], albeit on a much longertimescale. arguedthat increasedabundanceof kaolinite and/or increased In addition,exactlythis mechanismis invokedto explainthe longkaolinite/illite ratiosassociatedwith someLPTM sectionsprovide direct evidence of warm, humid climates [Cramer et al., 1999; term shift in bulk carbonisotopesto lighter valuesas a resultof the initial stagesof the collision of India and Asia from the Paleocene RAVIZZA ET AL.' LPTM Os ISOTOPE EXCURSION 161 Median Grain Size ( phi - 50) (Janecek and Rea 1 983) to the Eocene and to suggestthat the unusualwarmth associated with thisperiodmay have resultedfrom a relatedincreasein atmospheric CO2 levels [Beck et al., 1995, 1998]. If the CIE is attributed to a net decreasein the sizeof the sedimentaryorganiccarbonreservoir rather than gas hydrate release,then the excursionto higher 8.2 8.4 8.6 8.8 9.0 9.2 9.4 LL4 4- GPC3 •*7Os/•'8Os caneasilybeattributed toerosionof radiogenic ancient sedimentaryorganic matter. Other workers have arguedthat the CIE is too large and abruptto be explainedby burial and erosionof sedimentaryorganicmatter,and favor attributingthe CIE to gashydratedecomposition [Dickenset al., 1997;Thomasand Shackleton, 1996].Recentworkdemonstrating a veryabrupt(<104years)onset for the CIE [Bains et al., 1999; Norris and RShl, 1999; ROhl et al., 2000] stronglysupportsthislatterview andargueagainstinterpreting the Os isotopeexcursionas the resultof the erosionof ancient sedimentaryorganic matter. There are other scenariosthat might give rise to largeand rapid Grain Size 10 187Os/ 188Os changesin the •7Os/•88Osof rivefineinput that cannotbe discounted.For example,shiftsin continentalclimate,suchaschanges in precipitationpatternsor/andmeanannualtemperatures, couldbe calleduponto producethis effect eitherby increasingweatheringin one region at the expenseof anotheror by changingthe extent to whichan isotopicallyheterogeneous terrainis weathered.Therefore our interpretationof the LPTM Os isotopeexcursionas evidence for increasedcontinentalweatheringratesduringthe LPTM must be regardedas tentative.Better constrainingthe processes responsiblefor thisOs isotopeexcursionis importantbecausethe implied increasein chemical weatheringduring this episodeof unusual warmthsupportsthe operationof a feedbackin whichatmospheric CO2 levelsare regulatedby the temperaturedependence of silicate weatheringrates. 5.3. Comparison of the LPTM to Recent Glacial-Interglacial Variations in the Seawater Os Isotope Record 15 Pegionof P-E Epoch boundary 2O 25 0.2 0.4 0.6 During the last two majorglacialperiods(isotopestages2 and6), brief excursions to lower •*?Os/•88Osratios have been found in Pa- cific metalliferoussedimentsand interpretedas an indication of diminishedchemicalweatheringon the continentsduringcold, arid glacial periods [Oxburgh, 1998]. Both climatically and in terms of the directionof the Os isotopeexcursion,the LPTM is the oppositeof the muchmorerecentglacialepisodes.Justasglobalaridity andcoldertemperatureswereinvokedby Oxburghto explainshifts to lower•87Os/•Osduringrecentglacialperiods,we attributethe higher•87Os/•Osat theLPTM to generallywarmer,morehumid conditionsasdescribedabove.The coherentresponseof the marine Os isotoperecord to these contrastingclimate regimes supports the notionthat high-resolutionOs isotopestudiesmay be broadly applicableto studyingthepossiblefeedbackbetweenglobalweathering ratesandclimatechange. Two additionalaspectsof the comparisonbetweenrecentglacial episodesand the LPTM deservemention.First is the significant difference of the duration of the excursion. While the •'7Os/•Os minima associatedwith recentglacial episodesspanat mosta few thousandyears, the LPTM event spannedroughly 200,000 years [Bainset al., 1999]. Thus the slopeof the seawaterOs curveduring the LPTM recovery is modest comparedto that associatedwith glacial transitions.Second,interpretationof the LPTM Os excursion is lesscomplicatedthan thoseassociatedwith recentglacials. The late Paleoceneand early Eoceneare consideredto be ice-free periods[Sloanand Thomas,1998;Zachoset al., 1993]. Therefore other potential explanationsof glacial Os isotopeexcursions,re- 0.8 1 1.2 187Os/188Os (Pegram and Turekian 1 999) Figure 4. A reinterpretation of thetimingof Os isotopevariationsin LL44GPC3 [Pegramand Turekian,1999]basedon grainsizedata[Janecekand Rea, 1983].The abruptshiftto smallermediangrainsizeof eolianmaterial (largerphi values)marksthe Paleocene-Eocene (P-E) epochboundaryin severalPacificcores[Rea, 1998].This correlation indicates thatthelong- term•S7Os/•aaOs of seawater is characterized by a broadminimumin the vicinity of the P-E boundary. latedto weatheringof glaciatedterrains[Peucker-Ehrenbrink and Blum, 1998], are not directly relevant. 5.4. Paleocene-Eocene Transition as Recorded in LL44-GPC3 and Its Relationship to the LPTM The mostcompletesingleOs isotoperecordis fromLL44-GPC3, a North Pacific pelagic clay sequence[Pegram and Turekian, 1999]. A dramatic decreasein eolian grain size occursin several Pacific corescloseto the Paleocene-Eocene(P-E) epochboundary [Rea, 1998] includingGPC3. This changein grain size in GPC3 [Janecekand Rea, 1983] providesa stratigraphic markerwhichallows identificationof the portionof the GPC3 Os isotoperecord that includesthe LPTM and P-E boundaryintervals(Figure4). Detailed correlationof Os isotopedata betweenPacific GPC3 and the North Atlantic and Indian Oceanis not possiblebecauseof the 162 RAVIZZA ET AL.: LPTM OsISOTOPEEXCURSION relativelylow temporalresolutionof theGPC3 record.We discuss thisminimum, such asophiolite weathering andrecycling ofextratheGPC3 Os isotoperecordherefor two reasons. First,thelower- terrestrial Osderived fromtheK-Tboundary impact [Pegram and resolution GPC3 recordprovidesa broadertemporalcontextin whichlong-termwarmingfrom the Paleocene intotheEocenecan Turekian, 1999],remainreasonable hypotheses. beexamined. Second, andmoreimportantly, thesimilarity in abso- 6. Conclusions lute •87Os/•8OsratiosbetweenGPC3 and the North Atlanticand a large(10%) increase in the•8?Os/•aaOs IndianOceansites(Table2) provides strongempiricalevidence We havedocumented sitesthatis coincident withthelate thatOsisotopic variations atallthreesitesarecontrolled mainlyby ratioat two widelyseparated Paleocene thermalmaximum.On the basisof the coherence of changes in the•87Os/•a8Os ratioof seawater through time. The GPC3recordis characterized by a minimum(•a?Os/•a•Os -- these twonewrecords withoneanother andwithexisting data we haveconcluded thattherewasa tran0.32) duringthe periodof mostrapidgrainsizedecrease and fromPacificsediments ratios higher•a?Os/•aOs ratios(= 0.41)beforeandafterthegrainsize sientwhole-oceanshift to higherseawater•87Os/•88Os transition (Figure3). This featureprobablyspansseveralmillion during the LPTM. The transientnatureof the excursionand indeyearson the basisof sedimentation rate estimates[Janecekand Rea, 1983;Kyteet al., 1993].The higher187Os/188Os ratiosafter theP-E epochboundary are substantiated by a spotanalysis of Pacificmetalliferous sediments of earlyEoceneage(•aVOs/laaOs -0.48[Peucker-Ehrenbrink etal., 1995].A singleanalysis of leachableOsin a pelagic•clay of latePaleocene age(187Os/188Os -• 0.45) [Peucker-Ehrenbrink et al., 1995]is distinctlyhigherthenthe GPC3databut similarto thehighestvaluesmeasured at Site549. pendent supporting dataindicative of an accelerated hydrologic cyclesuggest thatthisOsisotope excursion isevidence of a globallysignificant increase in chemical weathering. Thisinterpretation supportsthe operationof a feedbackbetweenclimate and chemical weathering ratesandsuggests thatCO2drawdown associatedwithsucha feedback mayhaveplayedanimportant role in arresting LPTM warming. Thevalidityof thisinterpretation dependscriticallyuponour assumption that the •87Os/•88Os of The rangeof •a?Os/ta•Os ratiosmeasuredin Sites 213 and 549 riverinefluxdidnotchange significantly duringtheLPTM.Comarevery similarto thosereportedfromGPC3 (Table2). However, parisonof theLPTM Os excursion reportedhereto thoseassocithe shapeof the high- and low-resolutionrecordsis different. atedwithrecentglaciation reveala coherent picturein whichbrief The higher-resolution datasetsspanning theLPTM arecharacter- periods of extreme warmthandhumidity arecharacterized byexizedbya localmaximum coincident withtheCIE (Figure2), while cursions to high18708/18808 ratios,whiletheopposite is trueof the GPC 3 recordis characterized by a broadminimumas de- briefepisodes of extremecoldandaridity.Thusavailable Osisothata weathering-based mechanism doesindeed scribed above.In total,thedatasuggest thatthelong-term transi- topedatasuggest tionfromthelatePaleocene intotheearlyEoceneis characterized actto ameliorate Earth'sclimate[Walkeretal., 1981]duringcliby generallylow •?Os/•aOs.ratioswithat leastonebriefexcursion maticextremeson timescales of 104-10s years.We stressthat to higher•VOs/•8•Os withinthebroadminimum defined by the thesehigh temporalresolutionstudiesreveal little aboutthe factorscontrolling thelong-term evolution of theOsisotopic comIt is noteworthy thattheagemodelfor GPC3[Kyteet al., 1993] position of seawater. In thecontextof establishing thecoherence datafromLL44GPC3, usedby PegrainandTurekian[1999]yieldsa mid-Paleocene age of thesenewLPTM Osdatawithexisting GPC3 record. for the localminimum in the seawatert87Os/188Os recordratherthan we havewe reinterpreted age-depth relationsin thiscoreandinfer epochboundary falls withina broad the Paleocene-Eocene boundary agewe suggest here.In using thatthe Paleocene-Eocene record,indicativeof a relatheeoliangrainsizedatato reinterpret thetimingof thisminimum, minimumin the seawater187Os/188Os we open the possibilitythat the generallylow •a?Os/•a•Os ratio tiveincrease in theamount of seawater Osderived fromyoung of seawaterduringthe Paleocene-Eocene transitionresultedfrom increased supply of unradiogenic Osfromrecent magmatic activity. mantle-derived rocks. This is consistentwith evidence of widespread volcanism at this time. Severalpotentialsourceshavebeenidentified:(1) increased seafloorhydrothermal activityrelatedto platereorganization [Owen andRea, 1985],(2) extensive NorthAtlanticvolcanism [Thomas Acknowledgments.We thankLary Ball and the W.H.O.I. ICP Facility for operationandsupportof theFinniganMAT sectorfieldinductivelycouand Shackleton,1996], and (3) Caribbeanvolcanism[Bralower pledplasmamassspectrometer (SF-ICPMS) in makingthesemeasurements. et al., 1997].The prospectof drawinga causativelink between GR thanksB. Peucker-Ehrenbrink and J. Howard for helpfulcomments thelow •VOs/•aaOs ratioof seawater andanyof thesemagmatic on earlydraftsof this manuscript.T Bralower,R. Oxburgh,andan anony- episodes is important because eachof thesehasbeensuggested as mousreviewel'offeredmanyconstructivecomments.Curatorsof theOcean DrillingProgramprovidedsamplematerialessential forthisstudy.Thiswork a potentialtriggerfor methane hydratedecomposition. At present, wassupportedby NSF A wardsOCE-99190500 (RDN), OCE-9819296 (GR) thislink is onlyspeculative andotherpossibilities for explaining and EAR-9804864 (GR). References Aubry,M.-P., Stratigraphic (dis)continuity and temporal resolutionof geological events in the upperPaleocene-lower Eocenedeepsea record,in Late Paleocene-EarlyEoceneClimatic and Biotic Events in the Marine and Cuba: hnplicationsfor the LPTM and for re- able moment in the evolution of the calcare- gionaltectonichistory,Micropaleontology, 45, 5-18, 1999. Aubry,M.-P., S. G. Lucas,andW. A. Berggren (Eds.), Late Paleocene-EarlyEocene Cli- et al., pp.37-66,ColumbiaUniv.Press, New matic and Biotic' Events in the Marine and York, 1998. TerrestrialRecor&, 513 pp.,ColumbiaUniv. Aubry,M.-P.,andA. Sanfilippo. LatePaleocene- Press, New York, 1998. earlyEocenesedimentary historyin western Aubry, M.-P., C. Requirand,and J. Cook. The Terrestrial Records, editedby M.-P. Aubry Rho•nboaster-Tribrachiatus linage:A remark- ous nannoplankton,GFF, 122, 15-18, 2000. Bains, S., R. M. Corfield, and R. D. Norris. Mechanisms of climatewarmingattheendof the Paleocene,Science,285, 724-727, 1999. Beck, R. A., D. W. Burband,W. J. Sercombe, T. L. Olson,andA. M. Khan.Organiccarbon exhumationand globalwarmingduringthe earlyHimilayancollision,Geology,23, 387390, 1995. RAVIZZA ET AL.: LPTM Os ISOTOPEEXCURSION Beck, R. A., A. Sinha, D. W. Burbank, W. J. Sercombe, and A.M. Khan. Climatic, oceano- graphic, and isotopic consequences of the Paleocene India-Asia collision, in Late Paleocene-EarlyEoceneClimaticand BioticEvents in the Marine and Terrestrial Records, edited warming, palaeoceanographic changes and benthicextinctionsat theendof thePalaeocene, ratios inferred fi'om metalliferous carbonates, Earth Pktnet. Sci. Lett., 160, 163-178, 1998. Koch,P. L., J. C. Zachos,and P. D. Gingerich. Robert, C., and J. P. Kennett. Antarctic sub- Correlationbetweenisotoperecordsin marine and continental carbon reservoirs near the Palaeocene/Eoceneboundary, Nature, 358, Univ. Press,New York, 1998. 319-322, 1992. Berggren,W. A., andR. D. Norris.Biostratigra- Kump, L. R., S. L. BrantIcy,and M. A. Arthur. Chemicalweathering,atmospheric CO2, and phy,phylogeny,andsystematics of Paleocene climate, Anntt. Rev. Earth Planet. Sci., 28, 611trochospiral plankticforaminifera, Micropaleon- Publ. SEPM Soc. Sediment. Geol., 54, 129212, 1995. Bemer, R. A. GEOCARB ll: A revised model for atmospheric CO2overPhanerozoic time,Am.J. Sci., 294, 56-91, 1994. Bickle, M. J., K. Caldeira, and R. A. Berner. The J. D. Wright. Mioceneseawater•?Os/•Os Nature, 353, 225-229, 199 I. by M.-P. Aubryet al., pp. 103-117,Columbia tology,43, p. 116, 1997. Berggren,W. A., D. V. Kent, C. C. Swisher,and M.-P. Aubrey. A revised Cenozoic geochronology and chronostratigraphy,Spec. 163 667, 2000. Kyte, F. T., M. Leinen, G. R. Heath, and L. Zhou. Cenozoicsedimentationhistoryof the central North Pacific: Inferences fi'om the elemental geochemistryof core LL44-GPC3, Geochim. Cosmochim.Acta, 57, 1719-1740, 1993. Levasseur,S., J. L. Birck, and C. J. Allegre.The tropical humid episode at the PaleoceneEocene boundary: Clay-mineral evidence, Geology,22, 211-214, 1994. R6hl, U., T. J. Bralower, R. D. Norris, and G. Wefer. New chronologyfor the late Paleocene thermal maximum and its environmen- tal implications,Geology,28, 927-930, 2000. Schroeder,T. A palynologicalzonationfor the Paleocene of the North Sea Basin, J. Micro- palaeontol., 11, 113-126, 1992. Shar•na,M., G. J. Wasserburg,A. W. Hofinann, andG. J. Chakrapani.Hi•nalayanuplift andosmium isotopesin oceansandrivers,Geochim. flux and the oceanic mass Cosmochim. Acta, 63, 4005-4012, 1999. Sloan, L. C., and E. Thomas. Global climate of the balance of osmium, Earth Planet. Sci. Lett., 174, 7-23, 1999. stances associated with a climatic "event," osmimn riverine late Paleoceneepoch: Modeling the circum- Norris, R. D., and U. R6hl. Carboncyclingand in lxtte Paleocene-EarlyEoceneClimatic and Biotic Events in the Marine and Terrestrial chronologyof climate warming during the geochemicalcarbon cycle: Discussionand Palaeocene/Eocene transition, Nature, 401, Records,editedby M.-P. Aubryet al., pp. 138reply, Geology,26, 477-478, 1998. 775-778, 1999. 157, New York, Colu•nbia Univ. Press, New Bralower, T. J., D. J. Thomas, J. C. Zachos, Owen, R. M., and D. K. Rea. Sea-floor hydroYork, 1998. M. M. Hirschmann,U. Roehi,H. Sigurdsson, thermalactivitylinks climateto tectonics:The Stott, L. D., A. Sinha, M. Thiry, M.-P. Aubry, E. Thomas, and D. L. Whitney. HighEocene carbon dioxide greenhouse,Science, andW. A. Berggren. Globaldelta•aC changes resolution records of the late Paleocene 227, 166-169, 1985. acrossthe Paleocene-Eocene boundary:Critethermal maximum and circum-Caribbean Oxburgh,R. Variationsin the osmiumisotope ria for terrestrial-marine correlations in Corvolcanism:Is there a causallink? Geology, compositionof seawaterover the past200,000 relation of the Early Paleogenein Northwest 25, 963-966, 1997. years, Earth Planet. Sci. Letr, 159, 183-191, Europe,Geol. Soc.Spec.Publ., 101, 381-399, Broecker,W. S., and A. Sanyal. Does atmosneed for mass balance and feedback in the 1998. pheric CO2 police the rate of chemical weathering?Global Biogeochem.Cycles,12, 403-408, 1998. Burton, K. W., B. Boutdon, J.-L. Birck, C. J. Allegre, and J. R. Hein. Os•nium isotope variationsin the oceansrecordedby Fe-Mn crusts,Earth Planet. Sci. Lett., 171, 185-197, 1999. Cramer, B. S., M.-P. Aubry, K. G. Miller, R. K. Olsson,J. D. Wright, and D. V. Kent. An exceptionalchronologic,isotopic,andclay mineralogicrecordof thelatestPaleocenethermal •naximum, Bass River, N J, ODP 174AX, 1996. Palmer, M. R., and J. M. Edmond. Controls over the strontiumisotope compositionof river water, Geochim. Cosmochim. Acta, 56, 20992111, 1992. Pegram,W. J., and K. K. Turekian. The osmium Thiry, M. Paleoclimatic interpretation of clay •nineralsin •narinedeposits:An outlookfrom the continental origin, Earth Sci. Rev., 49, 201-221,2000. Thomas, D. J., T. J. Bralower, and J. C. Zachos. isotopiccompositionchangeof Cenozoicsea waterasinfe•xedfi'oma deep-sea corecorrected the late Paleocene thermal maximum: for meteodticcontributions,Geochim.Cosmochim. Acta, 63, 4053-4058, 1999. isotopesfi'om Deep Sea Drilling ProjectSite 527, Walvis Ridge, Paleoceanography,14, Pegram, W. J., S. Krishnaswami, G. Ravizza,and 561-570, 1999. Thomas, E., and N.J. Shackleton. The Paleocene- K. K. Turekian. The record of seawater 187Os/186Osvariationthroughthe Cenozoic, New evidencefor subtropicalwarmingduring Eocene benthic formniniferal Stable extinction and Earth Planet. Sci. Letr, 113, 569-576, 1992. stableisotopeano•nalies,in Con'elationof the Peucker-Ehrenbrink,B., and J. D. Blum. The Early Paleogenein NorthwestEurope,Geol. effectsof global glaciationon the osmium Soc.Spec.Publ., 101, 401-441, 1996. Walker. A blastof gasin the latestPaleocene: isotopic composition of continental runoffand Walker, J. C. G., P. B. Hays,andJ. F. Kasting.A Simulatingfirst-ordereffectsof massivedisseawater, Geochim. Cosmochim.Acta, 62, negativefeedbackmechanismfor the longsociationof oceanic methanehydrate, Geol3193-3203, 1998. term stabilizationof Earth'ssurfacetemperaogy, 25, 259-262, 1997. ture, d. Geophys.Res., 86, 9776-9782, 1981. Edmond,J. M., Y. and Huh. Chemicalweathering Peucker-Ehrenbrink,B., and G. Ravizza. Case studyof continental run-offof Os:TheBaltic White, A. F., A. E. Blmn, T. D. Bullen, D. V. yieldsfi'o•nbasementand orogenicterrainsin Sea,Geology,24, 327-330, 1996. Vivit, M. Schulz,andJ. Fitzpatrick.The effect hot and cold climates,in TectonicUplift and of temperatureon experimental and natural ClimateChange,editedby W. F. Ruddiman, Peucker-Ehrenbrink,B., G. Ravizza, and A. W. Hofmann. The marine 187Os/186Os record of chemicalweatheringratesof grantitoidrocks, pp. 329-351, Plenum,New York, 1997. Geochim. Cosmochim. Acta, 63, 3277-3291, the past 80 million years, Earth Planet. Sci. Gibson, T. G., L. M. Bybell, and J. P. Owens. 1999. Lett., 130, 155-167, 1995. LatestPaleocenelithologicand biotic events ratio Zachos, J. C., K. C. Lohmann, J. C. G. Walker, in nefftic depositsof southwesternNew Jer- Ravizza,G. Variationsof the 187Os/186Os and S. W. Wise. Abrupt climate changes of seawaterover the past28 million yearsas sey,Paleoceanography, 8, 495-514, 1993. and transientclimatesduringthe Paleogene: inferred from metalliferous carbonates, Earth Hassler, D. R., B. Peucker-Ehrenbrink, and A marineperspective,J. Geol., 101, 191-213, Planet. Sci. Lett., 118, 335-348, 1993. G. Ravizza. Rapid determinationof Os isoBull. Soc. Geol. Ft., 170, 883-897, 1999. Dickens, G. R., M. M. Castillo, and J. C. G. topic compositionby spargingOsO4 into a magneticsectorICP-MS, Chem.Geol., 166, Nature, 359, 117-122, 1992. 1- 14, 2000. Janecek,T. R., and D. K. Rea. Eolian deposition in the northeast Pacific Ocean: Cenozoic historyof atmospheric circulation,Geol.Soc. Am. Bull., 94, 730-738, 1983. Kaiho, K., et al. Latest Paleocene benthic foraminiferal Raytoo,M. E., and W. F. Ruddiman.1992, Tectonicforcing of late Cenozoicclimate, extinction.. and environmental changesat Tawanui, New Zealand, Paleoceanography, 11, 447-465, 1996. Kennett,J. P, and L. D. Stott. Abrupt deep-sea Rea, D. K. Changesin atmospheric circulation during the latest Paleoceneand earliest Eoceneepochsandsomeimplications for the global climate regime, in Late PaleoceneEarly EoceneClimaticand BioticEventsin the Marine and Terrestrial Records,editedby M.-P. Aubyr et al., pp. 118-123,Columbia Univ. Press,New York, 1998. Reusch,D. N., G. Ravizza, K. A. Maasch,and 1993. M.-P. Aubry, lnstitut des Sciencesde l'Evolution, Universit• Montpellier II, 34095 Montpellier Cedex 05, France. J. Blusztajn, R. N. Noms, and G. Ravizza, Departmentof GeologyandGeophysics, Woods Hole OceanographicInstitute,WoodsHole, MA 02543. ([email protected]) (Received May19,2000;revised December 14, 2000; acceptedDece•nber27, 2000.)
© Copyright 2026 Paperzz