Investor Relations

Compilation and time-series analysis of a marine carbonate [delta] 18O,[delta] 13C, 87Sr/86Sr and [delta] 34S database through Earth history

Compilation and time-series analysis of a marine carbonate [delta] 18O,[delta] 13C, 87Sr/86Sr and [delta] 34S database through Earth history
of 21
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Similar Documents
  Compilation and time-series analysis of a marine carbonate  δ 18 O,  δ 13 C, 87 Sr/  86 Sr and  δ 34 S database through Earth history A. Prokoph  a, ⁎ , G.A. Shields  b,d,1,2 , J. Veizer   c,3 a  SPEEDSTAT, 19 Langstrom Crescent, Ottawa, ON, Canada K1G 5J5  b Geologisch-Paläontologisches Institut, Westfälische Wilhelms-Universität Münster, Corrensstrasse 24, D-48149, Münster, Germany c  Department of Earth Sciences and Ottawa-Carleton Geoscience Centre, University of Ottawa, Ottawa, ON, Canada K1N 6N5 d  Department of Earth Sciences University College London, Gower Street, London WC1E 6BT, UK  Received 21 February 2007; accepted 20 December 2007Available online 18 January 2008 Abstract The Sr, S, O and C isotope database of marine carbonates contains over 55,000 published isotope values of low-Mg calcite from diageneticallylittle altered Phanerozoic fossil shells as well as samples of whole rocks and calcite cements of Ordovician to Archean age. Carbon and oxygenisotope data for the shell material are divided into habitat subsets (high-, mid-, low-latitude and deep sea), and whole rock data are separated bymineralogy into calcite/dolomite subsets. Trend, correlation, wavelet, and spectral analyses on Gaussian-filtered isotope records were applied todetect and quantify similarities and patterns in temporal records with the following results:(1) Oxygen isotope trends from the  “ high-latitude ”  and  “ deep-sea ”  habitats are almost indistinguishable through the last 115 Ma, consistent withthe existence of the  “ oceanic conveyor belt  ”  throughout this interval;(2) All oxygen isotope habitat records show a strong, coherent 30 – 45 Ma ( ∼ 38 Ma) cyclicity throughout the Cretaceous and the Cenozoic(3) Up to 70% of the multi-million year variability in the  δ 18 O record of the last 115 Ma can be simulated by the following equation: d 18 O  x ð Þ¼ 0 : 64sin2 p t  = 120 Ma þ 0 : 9 ð Þþ  X  sin2 p t  = 38 : 3 Ma þ 1 : 1 ð Þ with  X   ranging from 0.4 – 0.6 ‰  for the  “ low- “ ,  “ high-latitude ”  and  “ deep-sea ”  habitats, to 0.8 ‰  for the  “ mid-latitude ”  realm.(1) A  ∼ 120±20 Ma cycle occurs in the Paleozoic and Neoproterozoic  δ 18 O record, consistent with paleoclimate variability as interpreted fromsedimentological and faunal records.(2) The offset of   δ 13 C values between  “ deep water  ”  and  “ high-latitude ”  vs. surficial habitats at lower latitudes is consistent with the operation of a biological pump in the oceans since at least the Cretaceous.(3) Sr and S isotope records exhibit a  ∼ 60 – 70 Ma cyclicity throughout the Phanerozoic.© 2008 Elsevier B.V. All rights reserved.  Keywords:  isotopes; databases; oxygen; carbon; strontium; sulfur  1. Introduction Over recent decades some hundred thousand isotope analysesofC,O,S,andSrhavebeencarriedoutforgeologicstudies.Their  purposes varied, as did the rock and fossil material used, theaccuracy and resolution of the stratigraphy, and the care takenduring sample selection. Commencing with Keith and Weber (1964) and Veizer and Compston (1974) and Veizer and Hoefs (1976), several such isotope databases have been compiled inan attempt to reconstruct global multi-million-year records of   Available online at Earth-Science Reviews 87 (2008) 113 – ⁎  Corresponding author. Tel.: +1 613 247 1072; fax: +1 613 520 5613.  E-mail addresses: (A. Prokoph),,  (G.A. Shields), (J. Veizer). 1 Tel.: +49 251 83 33937; fax: +49 251 83 38312. 2 Tel.: +44 20 7679 2363; fax: +44 20 7679 2433. 3 Tel.: +1 613 562 5800x6461; fax: +1 613 562 5192.0012-8252/$ - see front matter © 2008 Elsevier B.V. All rights reserved.doi:10.1016/j.earscirev.2007.12.003  seawater isotope variability. At that time, the primary challengewas to detect and overcome diagenetic alteration of srcinalisotope signals as well as advancing the limits of analytical precision.Subsequently,inorder toavoid asmuchaspossiblethe problems of diagenesis, effort has been mainly concentrated onunalteredlow-Mgcalciteshellmaterial,preferablyfromarticulate brachiopods,belemnitesandforaminifera(e.g.Veizeretal.,1999;Lisiecki and Raymo, 2005). More recently, the oxygen isotoperatios of phosphatic shells of conodonts or fish teeth enamel, andof aragonitic shells (e.g. ammonites), have also been used as paleoenvironmental tracers (e.g., Picard et al., 1998).The Phanerozoic marine  87 Sr/  86 Sr trend has remained without significant change for about three decades because this ratio isrelatively insensitive to habitat variability or to the fossil materialused (Veizer et al., 1999). In contrast, oxygen and carbon isotoperatiosarehighlysensitivetoenvironmentalandbiologicalfactors.For example, the application of oxygen isotopes to thereconstruction of seawater temperature has to concern itself with water depth, salinity, pH and the mode of shell growth.Moreover,thespecies-specificphysiological( “ vital ” )factorsmayreflect additional variables, such as coral photosymbiosis withalgae that, in turn, can modify the isotope fractionation of carbon(Abramovitch et al., 2003). As a consequence, with an improvedunderstanding of isotopic proxies for reconstruction of past environments, there has emerged a need to address not only thetemporal variability of the signal but also the variability due tohabitat and organism-specific isotope fractionation parameters(e.g., Spero and DeNiro, 1987).In this review, we constructed separate Phanerozoic oxygenand carbon isotope records for   “ deep water  ”  and for   “ high- ” , “ mid- ” and “ low-latitude ” surfacewaterinanattempttoimprovethe reconstruction of paleoenvironmental conditions. Wecalibrated these datasets to an articulate brachiopod standard,in order to address the differences in habitat and physiology of the organisms. Brachiopods were chosen as a standard becausethey have been extant since the Cambrian and, except for theCenozoic, the bulk of Phanerozoic O and C isotope measure-ments were carried out on their shells. The Precambrian recordsare based on calcitic or dolomitic components of whole rock carbonate samples.The aims of this review are:(1) toprovidethescientificcommunitywithanupdated(uptoFebruary 2006) compilation of marine Sr, O, C, and Sisotope data from published literature for the entire Earthhistory. The database is fully annotated, grouped intodifferent habitats where deemed necessary, and consis-tently referenced to thenew geologic time-scaleGTS2004(Gradstein et al., 2004);The storage of the database in a secure databank where it can be available to the community for their own retrievaland manipulation is the primary goal of this publication, because with Veizer's retirement, its continuation andeven survival cannot be guaranteed. The supplementarygoals listed below are only tentative suggestions for future advances in development and interpolation of thedatabase.(2) to carry out data processing that transforms the raw datainto habitat-specific geologic records;(3) to determine long-term trends, periodic patterns andinterrelationships within and between the isotope recordsusing time-series analytical techniques;(4) to interpret the results of the time-series analysis in termsof paleoenvironmental reconstructions.We particularly avoided any reconstruction of recordsthrough time intervals where no data are available. Differencesin temporal data density are clearly highlighted. In this way, thereader can easily recognize for which time interval and at whichresolution the suggested reconstructions and interpretations arerobust or less so. All raw data used for record compilationsassociated with this article can be found at  doi:10.1016/j.earscirev.2007.12.003. 2. Phanerozoic fossil data 2.1. Data selection This review is an update and restructuring of the Ottawa – Bochum O, C, S and Sr isotope database for marine carbonates(Veizer et al., 1999; Shields and Veizer, 2002; Kampschulte andStrauss, 2004). The updated database includes oxygen andcarbon isotope data for fossil shells from 146 references, threereferences for sulfur isotope values from marine sulfates, and 43references for strontium isotope ratios of fossil shells andmicrite. In order to extend the fossil carbonate database to cover the entire Phanerozoic, it has been supplemented by Cambrianwhole rock carbonate data (11 references for   87 Sr/  86 Sr and 16for   δ 18 O and  δ 13 C).The database includes over 39,000  δ 18 O and  δ 13 C values for the low-Mg calcite or aragonite of diagenetically little alteredshells of ammonites, belemnites, brachiopods, inoceramids,trilobites, benthic and planktonic foraminifera, as well as micriteandconodontdatafor   87 Sr/  86 Sr(Table1).The380 δ 34 S sulfate  dataare from marine barites and structurally substituted sulfate. A set of 2282 whole rock samples, mostly deriving from references upto 2002, has been included to fill the large stratigraphic gaps betweenshellsamplesintheCambrian.Forconsistency,thedeep-sea carbon and oxygen isotope data of the last 67 Ma were takenentirely from the benthic foraminifera dataset of  Zachos et al.(2001). Isotope data for shell material with Mn concentrations N 350 ppm, some duplicate data entries, and Holocene foramini-fera data from the Ottawa – Bochum database (Veizer et al., 1999)havenotbeenincorporatedintothenewdatabase.Inaddition,theabundant Quaternary isotope data from the literature have not  been utilized in this database. The only exceptions are thearticulate brachiopod data because of their importance as astandard for the Phanerozoic isotope records.The individual samples of the database represent discretemeasurements that are subject to stratigraphic uncertainty as wellas errors in both analytical and palaeoenvironmental in-terpretations.Allsampleagesweretransformedfromtheiroriginaltime scale into the GTS2004 time scale (Gradstein et. al., 2004)and include also estimates of the ±1 σ  stratigraphic uncertainty. 114  A. Prokoph et al. / Earth-Science Reviews 87 (2008) 113  –  133  Samples with a cumulative stratigraphic uncertainty  N 5% of the mean sample age were excluded from the database. All theseuncertainties were assessed very conservatively for cases whereno detailed stratigraphy or age has been provided. A minimum1 σ  uncertainty of 0.5% of the mean age has been applied to allsamples, even if the srcinal source stated a narrower ageuncertainty. Nevertheless, the assignment of accurate strati-graphic mean ages and uncertainties is often problematic andwill certainly change with updated time scales and radioactivedecay constants. 2.2. Oxygen and carbon isotope data The Phanerozoicdatabaseincludesinformation on timescale,stratigraphic uncertainty, sample age, fossil groups, habitat,literature source, sample location and other comments. Thedatabase is organized as an MS EXCEL spreadsheet with sheetsfor different habitats and a miscellaneous category.The updated carbon and oxygen database has been structuredaccording to the following habitat designations:(1) surface waters between 58° – 90° paleolatitudes ( “ high-latitude ” ),(2) surface waters between 32° – 58° paleolatitudes ( “ mid-latitude ” ),(3) surface water between 32°N and 32°S paleolatitudes( “ low-latitude ” ),(4)  “ deep sea ”  below 300 m water depth.We recognize that this division is somewhat arbitrary because the extent of paleogeographic zones will vary withclimates of greenhouse/icehouse types. Nevertheless, with the present-day state of the art, any further specification would haveto be speculative.The surface water habitats (1 – 3) do not exactly follow the polar circles and tropics for the following reasons:(1) Paleogeographic reconstructions of the Paleozoic are not known well enough for delineation of exact climaterealms. In this study, the widening of the tropics to 32°latitude allows us to accommodate almost all pre-Jurassicfossil data into the  “ low-latitude ”  category, an assignment more reflective of their habitat. The Cretaceous plankticforaminifera from latitudes  ∼ 25 – 32° are also morerepresentative of a tropical than of a temperate climate(Huber et al., 1995). Table 1Seawater isotope databaseMaterial/fossil All All Low-latitude Mid-latitude High-latitude Deep sea Miscellaneous All Veizer et. al.(1999)Sr 34S C O C O C O C O C O C/O C/OBelemnites 435 150 164 962 975 69 73 628Plankticforaminifera1666 602 766 489 493 109 163 136 115 76Brachiopods 1235 3794 3889 99 101 9 11 4 3733Benthicforaminifera(deep sea)9763 11175 9 9 680 # Benthicforaminifera(neritic)17 17Trilobites 4 4Inoceramids 7 61 61Bivalves 50 50 1Oysters 63 102 125 119Ammonites 58 58Bivalve larvas 5 5Echinoderms 3 3Gastropods 8 8Corals 11 3 3 2Conodonts 623Micriticcarbonate186Total: fossils 4226 0 4550 4823 1550 1569 187 247 9763 11175 452 458 16503/ 182774559Structurallysubstitutedsulfate225Barite 155Whole rock 1355 68 7872 7222Total 5581 448 12422 12045 1550 1569 187 247 9763 11175 452 458 16503/ 182774559 # Portion of  Veizer et al. (1999) database not included in this database.115  A. Prokoph et al. / Earth-Science Reviews 87 (2008) 113  –  133  (2) Samples from latitudes 58 – 60°, as a group, havesignificantly heavier   δ 18 O values than other   “ mid-latitude ” samples.Most of the belemnite and planktic foraminifera data of theOttawa – Bochum database (Veizer et al., 1999) are now part of the  “ mid-latitude ”  realm subset, and most of their brachiopodsamples part of the  “ low-latitude ”  realm subset, respectively. Inorder to avoid local geographic bias on the global perspective of this study, data from sections that were sampled at very highstratigraphic resolution were down-sampled by including only ∼ 20 kA averages.The databasealso includesafifth category( “  Miscellaneous ” )for data that do not fulfill our requirements for compiling aconsistent isotope record. Samples included into this group are:(1) Fossils with aragonite shells, for example ammonites.(2) Fossils that cannot be assigned reliably to surface water above the thermocline or to the deep sea, such as thethermocline/subthermocline planktic foraminifera (e.g.,Rotaliporae, Subbotinea) or the shallow water benthicforaminifera.(3) Samples from epicontinental seas or marine basins withrestricted water exchange with the open ocean, such as theMediterranean Sea during the late Miocene (i.e. theMessinian Crisis) or the Western Interior Seaway of  North America. Most data from these environments arenot retained in the database.(4) Samples exhibiting probable isotopic disequilibrium between seawater and shell (i.e.  “ vital effects ” ). For example, the planktic foraminifera with high  δ 13 C isotopevalues that likely lived in photosymbiosis with algae.(5) Fossil groups of inoceramids, bivalves, corals, as well asunaltered surface dwelling tropical planktic foraminifera(e.g., Pearson et al., 2001) because, at this stage, theycould not be calibrated to the brachiopod standard.(6) Paleozoic samples with  δ 18 O of less than − 14 ‰ , becausethey are almost certainly diagenetically altered. Thesesamples were also excluded from data analyses andinterpretations by Veizer et al. (1999) and subsequent  publications.Data from the  “  Miscellaneous ”  category are not included inany plots and data analysis, but have not been discarded asthey may become useful once more information and relatedmetadata have been collected. Many of these fossil groupshave already been shown to serve as useful paleoenviron-mental proxies for high-resolution and regional studies (e.g.,Picard et al., 1998; Dromart et al., 2003; Zakharov et al.,2001).Insufficient stratigraphic range and density of data withineach fossil group generates some problems with habitat assignment. Among the sampled fossils, brachiopods have thelongest stratigraphic range, and they have been studied for vitaland diagenetic effects. In addition, they represent the largest shallow water dataset for the  “ low-latitude ”  habitat. In thisstudy, they therefore provide the baseline to which all other fossil groups are calibrated. The details of the calibration areoutlined in Section 4.2.In summary, only brachiopods, belemnites, trilobites, and planktic foraminifera are included in categories (1) to (4)above. Whole rock data have been added to cover the other-wise stratigraphically incomplete Cambrian fossil dataset for the  “ low-latitude ”  realm (Fig. 1). The  “ mid-latitude ” ,  “ high- Fig. 1. Raw  δ 18 O (A) and  δ 13 C (B) data from the  “ low-latitude ”  realm for the last 543 Ma. Time-scale GTS2004 (Gradstein et al., 2004).116  A. Prokoph et al. / Earth-Science Reviews 87 (2008) 113  –  133  latitude ”  and  “ deep-sea ”  oxygen isotope subsets have nodata entries older than 200 Ma, 156 Ma and 126 Ma, re-spectively (Fig. 2). The carbon isotope data for the  “ mid-latitude ” ,  “ high-latitude ”  and  “ deep-sea ”  habitats are gen-erally from the same sample populations as the oxygen isotopedata.  “ High-latitude ”  carbon isotope data are, however, sparse(Fig. 3). 2.3. Strontium isotope data The entire Phanerozoic  87 Sr/  86 Sr database of marinecarbonates is composed of 4226 samples from diageneticallyunaltered or demonstrably little altered low-Mg calcite shellmaterial of brachiopods, belemnites, corals, inoceramids, planktic foraminifera, oysters, micritic carbonate (mostlynanoplankton), as well as from apparently well preservedconodontapatite(Table1).Thesedataarecomplementedby400whole rock samples that fill stratigraphic gaps in the fossil data(Fig. 4).The  87 Sr/  86 Sr database is organized into three subsets( “ fossil ” ,  “ omitted ” , and  “ whole rock  ” ). Omitted data haveinsufficient quality (e.g. diagenetic alteration) or missingessential information on the sample (e.g., precise location andage). The  “ fossil ”  and  “ omitted ”  subbase includes data on age,stratigraphic uncertainty, sample age, data source, fossil group,and location. All measured  87 Sr/  86 Sr values are normalized to NBS 987 of 0.710250. The whole rock subset is organized for consistency with the PMCID 1.1. database (Shields and Veizer,2002), which is the only source of these data.It is not necessary to classify samples into  “ habitat  ”  subsetsas the strontium isotope ratio of marine carbonate is habitat and physiology independent; moreover the world's oceans are wellmixed with respect to  87 Sr/  86 Sr (Peterman et al., 1971; Veizer and Compston, 1974; McArthur et al., 2000). 2.4. Sulfur isotope data of marine sulfate The  δ 34 S database of seawater sulfate is composed of 380 barite or structurally substituted sulfate samples from threeliterature references (Table 1, Fig. 5), and is organized into three subsets ( “ fossil ” , “ omitted ” , and  “ whole rock  ” ) with informationon sample age (in GTS2004), stratigraphic uncertainty,  δ 34 Ssample, and literature reference.There is no known habitat related bias to the  δ 34 S ratio of seawater sulfate in unrestricted marine environments, and nosubsets were therefore generated. 3. Precambrian whole rock data The majority of data utilized in this database was compiled previously in the Precambrian marine carbonate isotope database:Version1.1of ShieldsandVeizer(2002).Wehaveaddedliteraturedata from 2001 – 2006, with particular emphasis on the Neoproter-ozoic (Jaffrés, 2005; Jaffrés et al., 2007) and the Archean. Fig. 2. Raw  δ 18 O data for the  “ mid-latitudes ”  (A),  “ high-latitude ”  (B) and  “ deep-sea ”  (C) categories for the last 200 Ma. Time-scale GTS2004 (Gradstein et al., 2004).117  A. Prokoph et al. / Earth-Science Reviews 87 (2008) 113  –  133
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!