Funny & Jokes

Abrupt climate variability since the last deglaciation based on a high-resolution, multi-proxy peat record from NW Iran: The hand that rocked the Cradle of Civilization?

Description
We present a high-resolution (sub-decadal to centennial), multi-proxy reconstruction of aeolian input and changes in palaeohydrological conditions based on a 13000 Yr record fromNeor Lake's peripheral peat in NW Iran.Variations inrelative
Categories
Published
of 16
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.
Related Documents
Share
Transcript
  Abrupt climate variability since the last deglaciation based on ahigh-resolution, multi-proxy peat record from NW Iran: The hand thatrocked the Cradle of Civilization? Arash Shari fi  a ,  b ,  * , Ali Pourmand  a ,  b , Elizabeth A. Canuel  c , Erin Ferer-Tyler  c ,Larry C. Peterson  b , Bernhard Aichner  d , Sarah J. Feakins  d , Touraj Daryaee  e ,Morteza Djamali  f  , Abdolmajid Naderi Beni  g , Hamid A.K. Lahijani  g , Peter K. Swart  b a Neptune Isotope Laboratory (NIL), Department of Marine Geosciences, Rosenstiel School of Marine and Atmospheric Science, University of Miami, 4600Rickenbacker Causeway, Miami, FL 33149-1098, USA b Department of Marine Geosciences, Rosenstiel School of Marine and Atmospheric Science (RSMAS), University of Miami, 4600 Rickenbacker Causeway,Miami, FL 33149-1098, USA c Virginia Institute of Marine Science, College of William  &  Mary, P.O. Box 1346, Gloucester Point, VA, USA d University of Southern California, Department of Earth Sciences, 3651 Trousdale Pkwy, Los Angeles, CA 90089-0740, USA e Samuel M. Jordan Center for Persian Studies and Culture at the University of California, Irvine, CA 92697, USA f  Institut M   editerran  een de biodiversit   e et d ’  Ecologie (IMBE   e  UMR CNRS 7263 / IRD 237), Europ ^ ole M   editerran  een de l ’   Arbois, BP 80,13545 Aix-en-Provence cedex 04, France g Iranian National Institute for Oceanography and Atmospheric Science (INIOAS), Marine Geology Division, P.O. Box 14155-4781 Tehran, Iran a r t i c l e i n f o  Article history: Received 21 January 2015Received in revised form22 June 2015Accepted 2 July 2015Available online xxx Keywords: Holocene climateCompound-speci fi c biomarkerCradle of CivilizationAtmospheric dustOmbrotrophic peatYounger DryasIran a b s t r a c t We present a high-resolution (sub-decadal to centennial), multi-proxy reconstruction of aeolian input andchangesinpalaeohydrologicalconditionsbasedona13000YrrecordfromNeorLake'speripheralpeatinNWIran.Variationsinrelativeabundancesofrefractory(Al,Zr,Ti,andSi),redoxsensitive(Fe)andmobile(KandRb)elements, total organic carbon (TOC),  d 13 C TOC , compound-speci fi c leaf wax hydrogen isotopes ( d D), carbonaccumulation rates and dust fl uxes presented here  fi ll a large gap in the existing terrestrial paleoclimate re-cords from the interior of West Asia. Our results suggest that a transition occurred from dry and dusty con-ditionsduringtheYoungerDryas(YD)toarelativelywetterperiodwithhighercarbonaccumulationratesandlowaeolianinputduringtheearlyHolocene(9000 e 6000YrBP).ThisperiodwasfollowedbyrelativelydrieranddustierconditionsduringmiddletolateHolocene,whichisconsistentwithorbitalchangesininsolationthat affected much of the northern hemisphere. Numerous episodes of high aeolian input spanning a fewdecades to millennia are prevalent during the middle to late Holocene. Wavelet analysis of variations in Tiabundances as a proxy for aeolian input revealed notable periodicities at 230, 320, and 470 years with sig-ni fi cant periodicities centered around 820,1550, and 3110 years over the last 13000 years. ComparisonwithpalaeoclimatearchivesfromWestAsia,theNorthAtlanticandAfricanlakespointtoateleconnectionbetweenNorthAtlanticclimateandtheinteriorofWestAsiaduringthelastglacialterminationandtheHoloceneepoch.We further assess the potential role of abrupt climate change on early human societies by comparingour record of palaeoclimate variability with historical, geological and archaeological archives from thisregion. The terrestrial record from this study con fi rms previous evidence from marine sediments of theArabian Sea that suggested climate change in fl uenced the termination of the Akkadian empire. Inaddition, nearly all observed episodes of enhanced dust deposition during the middle to late Holocenecoincided with times of drought, famine, and power transitions across the Iranian Plateau, Mesopotamiaand the eastern Mediterranean region. These  fi ndings indicate that while socio-economic factors aretraditionally considered to shape ancient human societies in this region, the in fl uence of abrupt climatechange should not be underestimated. ©  2015 Elsevier Ltd. All rights reserved. *  Corresponding author. Neptune Isotope Laboratory (NIL), Department of Marine Geosciences, Rosenstiel School of Marine and Atmospheric Science, University of Miami,4600 Rickenbacker Causeway, Miami, FL 33149-1098, USA. E-mail address:  oshari fi @rsmas.miami.edu (A. Shari fi ). Contents lists available at ScienceDirect Quaternary Science Reviews journal homepage: www.elsevier.com/locate/quascirev http://dx.doi.org/10.1016/j.quascirev.2015.07.0060277-3791/ ©  2015 Elsevier Ltd. All rights reserved. Quaternary Science Reviews 123 (2015) 215 e 230  1. Introduction Since the beginning of the Neolithic Era, the area in West Asiathat extends from southwestern Iran and the Arabian Peninsula totheeasternandsoutheasternMediterraneanSea,alsoreferredtoasthe “ CradleofCivilization ” andthe “ FertileCrescent ” ,haswitnessedthe birth of agriculture and development of some of the earliesthuman societies (Leick, 2010; Mellaart, 1975; Riehl et al., 2013).Evidence is mounting that abrupt climate change during the Ho-locene epoch (beginning 11700 before present, BP) may haveplayed a transformative role in the growth and deterioration of human civilizations (Brooks, 2006; Cullen et al., 2000; deMenocal,2001; Riehl, 2009). Although the amplitude of climate variabilitywas reduced during the Holocene relative to the last glacial period(Grootes et al.,1993), episodes of abrupt climatic change have beendocumented in marine and terrestrial records throughout the Ho-locene in both hemispheres (see review by Mayewski et al., 2004).On a regional scale, the climate of West Asia is governed bycomplex interactions between the mid-latitude Westerlies, the Si-berianAnticycloneandtheIndianOceanSummerMonsoon(Fig.1).While a number of paleoclimate studies have drawnpotential linksbetween abrupt climate change and the rise and fall of civilizationsacross the Fertile Crescent (Cullen et al., 2000; Staubwasser andWeiss, 2006), high-resolution (sub-decadal to centennial) terres-trial archives of climate variability with well-constrained agemodels are scarce from this region (Nicoll and Küçükuysal, 2013).Such records are needed in order to address the uncertaintyregarding the timing and regional signi fi cance of climatic transi-tions and their potential in fl uence on early human societies, andthe extent to which anthropogenic activities may have in fl uencedthe climate of the last deglacial period.Interpreting the available proxy reconstructions of Holoceneclimate variability in the interior of West Asia is not straightfor-ward. For example, Stevens et al. (2001) found disagreement be-tween relatively lower  d 18 O values from Lake Zeribar during theearlyHolocene,whicharegenerallyinterpretedtorepresentwetterconditions (Roberts et al., 2008), and pollen and macrofossil datafor this period from Lake Zeribar, which were interpreted to indi-cate drier conditions during this period. Theyconcluded that lower d 18 O values may have been due to a shift in the timing of precipi-tation (i.e., protracted summers). Other explanations that havebeen suggested for this discrepancy include underestimating hu-man impact on vegetation during the early Holocene, and thedelayed reaction of biomes to postglacial climate change (Djamaliet al., 2010; Roberts, 2002). Stevens et al. (2006) further exam- ined geochemical and biological evidences from Lake Mirabad inwestern Iran during this period and concluded that the early Ho-locene was dry. Based on microfossil assemblages and pollen data,Wasylikowa et al. (2006) concluded that Zeribar lake levels werevariable during the early Holocene. Pollen and ostracod assem-blages from lakes in Turkey, Iran and Georgia suggest that conti-nental (dry and variable) climate dominated over the interior of West Asia during the early to middle Holocene (Connor andKvavadze, 2008; El-Moslimany, 1982; Grif  fi ths et al., 2001; Wasy-likowa, 2005). Available records of palaeo-vegetation changes,however, fall short of disentangling human  versus  climate impact(Djamali et al., 2009a), indicating the need for high-resolutionpalaeoclimate reconstructions that are independent of vegetationtypes that may have been in fl uenced by agriculture as well asclimate (Roberts et al., 2011).In this contribution, we present inorganic and organic proxyreconstruction of aeolian input and palaeohydrological changesover the last 13000 Yr from an ombrotrophic (rain fed) peat mirelocated at the periphery of Neor Lake in NW Iran. We examine thepossibility of an atmospheric teleconnection during the last glacialtermination and the Holocene by comparing our results with re-cords from the North Atlantic, African lakes and eastern Mediter-ranean. We further investigate the potential in fl uence of abruptclimate change on major early human societies from West Asia bycomparing our  fi ndings with historical and archaeological recordsfrom this region. 2. Study area Neor Lake (37  57 0 37 00 N, 48  33 0 19 00 E) is a seasonally rechargedbody of water formed over a tectonic depression on the leeward fl ank of the Talesh (Alborz) Mountains in NW Iran (Fig. 1 and Fig 2A). The tectonic depression, which formed within an andesiticbedrock as a result of displacements during the Eocene (Madadiet al., 2005) does not receive water from any permanent riversand has fostered a peripheral peat mire in primarily southernsection of the lake for at least the last 13000 years. Precipitation inthis high-altitude peat mire (~2500 m above sea level, m.a.s.l.)consistsofrainandsnow.Waterleavesthelakethroughanincisionin the north that has been arti fi cially dammed for the last fewdecades. The lake surface area is reduced by more than 50% duringperiods of low precipitation (Fig. 2B) (Madadi et al., 2005). Mass accumulation in the peat mire is primarily driven by accretion of decomposing biomass and wet and dry aeolian deposition.The mean annual precipitation (30-years average) recorded atthe nearest meteorological station in Ardebil, located at1332 m.a.s.l. and 50 km NE of the lake, exceeds 300 mm. Precipi-tation is highest in Mayand November (Fig. 2B) and the dry seasonlasts from July to September. Mean annual temperature at thestation is 15.4   C and the mean maximum and minimum temper-atures of the warmest and coldest months of the year are 25   C(July) and   7.9   C (January), respectively. As Neor peat mire is1200 m higher than the meteorological station, it is not unrea-sonable to expect higher annual precipitation and lower tempera-ture at Neor relative to the weather station. Vegetation in the lakebasin is composed of Irano-Turanian mountain steppe, dominatedby thorny-cushions plants. Local nomado-pastoral communitiesexploit both the steppe and the Neor peat mire vegetation. 3. Materials and methods  3.1. Peat cores In the summer of 2010, we recovered a 7.5-m core from thesouthwest part of Neor peat mire (Fig.1), as well as two additionalcoresofsimilarlengthfromwithin2mofthemaincore.Half-barrelcores were collected in 1-m increments of 7-cm diameter using aRussian split corer until the basal bedrock was reached. The  fi nalsegmentofthecoreincludeddense,laminatedgyttjaandamixtureof andesitic gravel from the bedrock. The cores were photographedon site and transferred into PVC core liners, sealed in non-reactiveplastic sheets and stored at 4   C and constant humidity at the corerepository of the Rosenstiel School of Marine and AtmosphericScience (RSMAS), the University of Miami. The physical propertiesof the core segments were logged, and changes in core dimensionswere closely monitored throughout the course of the study as peatwater content tended to vary with time. Each core segment wastransferred into an especially-designed polymethyl methacrylatecore holder and PVC liner with scale bars for reference. The coreswere subsequently imaged using a Geotek Multi-Sensor CoreLogger (GeotekMSCL) at the Department of Marine Geosciences(MGS), RSMAS and the images were calibrated for cross-core res-olution, light intensity and white balance. The position of eachsample taken for discrete organic and inorganic geochemical ana-lyses was determined based on comparing down-core XRF  A. Shari  fi  et al. / Quaternary Science Reviews 123 (2015) 215 e  230 216  elemental intensities and core images. Samples were taken fromareas of high terrigenous input, which correspond to high XRF in-tensities, areas of high organic content with low XRF intensities aswell as midpoints in order to capture the entire range of variabilityand minimize the effect of aliasing. Sub-samples were taken fororganic and inorganic geochemical analyses using a pre-cleanedceramic knife and the samples were dried at 45   C over 5 days.The water content was measured for each sample gravimetricallybased on sample masses before and after evaporation. The sam-pling resolution of each proxy record is discussed in sections3.3 e 3.9.  3.2. Age model Based on variations in the intensity of lithogenic elementsmeasured by XRF (see section 3.3. below), 19 sub-samples weresubmittedfor 14 CdatingtotheNational OceanSciencesAcceleratorMass Spectrometry Facility (NOSAMS), Woods Hole OceanographicInstitution. At NOSAMS, samples were pretreated using acid-onlyor acid-base-acid protocols prior to graphitization based on theirorganic contents. Ages were based on  14 C analysis and associateduncertainties were calibrated using CALIB 6.0 program (Stuiveret al., 1998) by utilizing IntCal13 calibration dataset (Reimer et al.,2013). Ages are reported as calibrated year before present (cal. YrBP).  3.3. Major and trace element analysis by XRF scanning  High-resolution measurements of the relative abundances of trace and major elements were performed on an AVAATECH XRF-core scanner in the Palaeoclimatology Lab at MGS-RSMAS. Allcore sections were scanned twice (10 kV,1000  m A, no  fi lter; 30 kV,1000  m A, Pd  fi lter) to acquire the range of elements reported here.For all measurements, the surface area was irradiated for 20 s of integration time at 2-mm intervals (average resolution of 3.5 yr)using a window 2-mm high by 12-mm wide. The raw XRF spectrawere then processed using the Canberra WinAxil software withstandard software settings and spectrum- fi t models, and Fig. 1.  Schematic position of major synoptic systems over West Asia and the location of Neor peat mire (star). Shaded area marks the region known as the  “ Fertile Crescent ” . Theapproximate current location of the Intertropical Convergence Zone (ITCZ) is also shown. IOSM refers to Indian Ocean Summer Monsoon. The Neor Lake's catchment basin andperipheral peat mire are shown in the bottom panel. Open triangle denotes the location of the peat core for this study.  A. Shari  fi  et al. / Quaternary Science Reviews 123 (2015) 215 e  230  217  abundances are reported as the intensity of each element in countsper second (cps). The JR-1 standard (Geological Survey of Japan)was used for calibration, and calibrated data were compared withprevious measurements to check for discrepancies. We interpretthe XRFelemental abundances as a qualitativeproxy forchanges inatmospheric dust input throughout the core. In addition, wecalculated aeolian  fl uxes based on discrete analysis of Ti concen-trationsfromasub-setof60samplesthatcorrespondedtointervalsof high, intermediate and low Ti intensities from XRF measure-mentsandthebulkdrydensityof thepeatcoreas describedbelow.  3.4. Bulk density Bulk density of peat material is required for quantifying atmo-spheric deposition. However, accurate measurement of bulk den-sity in peat and poorly consolidated sediment is challenging. As aresult, bulk densityof sediments is oftenassumed to be constant inthe sedimentary record or is measured on discrete samples usingdestructive gravimetric and volumetric techniques (Boelter, 1966;Chambers et al., 2011; Janssens, 1983). In addition to sample loss,this approach is not suitable for organic-rich sediments with highporosity.We utilized a novel, alternative approach using Gamma RayAttenuation Porosity Evaluator (GRAPE) data (Evans, 1965). Thebulk density based on gamma ray attenuation measurements aspart of the GEOTEK-MSCL is calculated from the followingequation: r  ¼  1 m dlnI  0 I   (1) where  “ r ”  is sediment bulk density,  “ m ”  is the Compton attenuationcoef  fi cient,  “ d ”  is the sediment thickness and  “ I  0 ”  and  “ I  ”  are thegamma intensity of the source and measured through the sample,respectively. To generate high quality density data, correctionsmust be implemented for experimental factors such as watercontent, beam spreading and attenuation through the PVC coreliner. We took an empirical approach to calibrating the gammadensitymeasurementsbyusinghalf-cylinderaluminumcalibrationtubes of varying thicknesses that resembled the split peat cores(Fig. SI-1). Gamma counts were measured for 100 s along thecalibration tubes that were encased in PVC core liners and  fi lledwith water. The average densities at different thicknesses werethen calculated using the following equation: r a v  ¼  d i D  * r  Al  þ ð D   d i Þ D  * r water   (2) where  “  r av ”  is the average density,  “ d i ”  is the thickness of thealuminum calibration tube,  “ D ”  is total thickness and  “ r  Al ”  and “ r water  ”  are the density of aluminum and water, respectively. Thenatural log of gamma intensities was then plotted against theaverage density of aluminum and water for various tube thick-nesses ( r av    d  vs.  ln (cps) ). The best- fi t line through the measuredvalues deviated from the theoretical line (Fig. SI-2) as a function of thevariablesmentionedabove. Themulti-sensorwascalibratedformeasuring wet bulk density using the following empirically-derived equation and the Gamma ray attenuation counts from theNeor peat core at increments of 1 cm and 10 s of integration time:  y  ¼  0 : 07  x  þ 8 : 8991 (3) where  y  ¼  ln ð cps Þ  and  x  ¼  r    d . Parameter  d  is the core thickness.Measurements of wet bulk density were performed on an adjacentcore and transferred to the main core using the near-identicalpro fi les of lithogenic elements. The dry bulk densities were calcu-lated using the equation below and the water content of sub-samples extracted from the main core: r dry  ¼  1   Water   %100    r wet   (4) where  “ r dry ”  and  “ r wet  ”  are the dry and wet bulk densitiesrespectively.  3.5. Total organic carbon (TOC) and stable isotope composition of organic carbon ( d 13 C  TOC  ) Basedonelementalvariationsidenti fi edusingtheXRFdata,400sub-samples (average resolution of 30 Yr) were collected for TOCand  d 13 C analyses. Between 0.002 and 0.015 g were sampled from0.5 to 1.5 g of dried and homogenized peat material. Samples weretransferred into pre-cleaned tin cups and treated with 10% HCl toremove inorganic carbon according to the procedure described inHedgesandStern(1984).TotalorganiccarbonwasmeasuredwithaThermo Scienti fi c ® Flash 2000 CHN analyzer at the Virginia Insti-tute of Marine Science (VIMS). Quality control and data reproduc-ibility procedures included analyzing duplicate samples andmultiple acetanilide standards and use of an acetanilide standard Fig. 2.  A) Location of Neor peat mire with respect toTalysh (Alborz) Mountains. B) 30-year average of minimum and maximum monthly mean air temperature (dark red andyellow lines) and mean monthly precipitation (blue line) as recorded in Ardebilmeteorological station.  A. Shari  fi  et al. / Quaternary Science Reviews 123 (2015) 215 e  230 218  curve with a correlation coef  fi cient ( R  2 )  >  0.99. The  d 13 C weredetermined at the RSMAS stable isotope laboratory (SIL) using aCostech elemental combustion system interfaced with a ThermoScienti fi c ® Delta V Advantage continuous  fl ow isotope ratio massspectrometer. Reproducibility ( < 0.01 ‰ ) was assessed fromrepeated analysis of acetanilide and glycine standards.  3.6. Measurement of carbon accumulation rate Changes in the Long Term Rate of Carbon Accumulation(LORCA), which is a function of carbon production and preserva-tion, were calculated based on 71 samples (average resolution of ~118 Yr) following the method described in Tolonen and Turunen(1996) and using the following equation (Page et al., 2004): LORCA   g C m  2  y  1   ¼  r    10000   r dry    c   (5) where  “ r  ”  is the average accumulation rate (mm y  1 ),  “ r dry ”  is drybulk dry density (g cm  3 ) and  “ c  ”  is carbon content (g C g  1 dryweight).  3.7. Ti concentration and aeolian  fl ux In order to quantify the  fl ux of dust to Neor peat mire over thelast 13000 years, 60 samples (average resolution of ~220 Yr) werecollected to be analyzed for Ti concentration. Approximately 0.2 gof peat material was transferred to capped, high-purity quartzcrucibles and ashed in a muf  fl e furnace at 750   C for 1 h. Approx-imately 0.01to0.02 gof theashed residuewas transferredto a pre-cleaned 6-ml PFA Savillex vial and dissolved in 5 mL of concen-trated HNO 3 e HCl e HF mixture (2:2:1, volumetric ratios). The vialswereplacedinanultrasonicbathfor2hat80  Candthenheatedat220   C on a hotplate overnight. After complete digestion was ach-ieved, about 0.1gof thesolutionwas dilutedin 40 gof 0.45 molL   1 HNO 3  and Ti concentration was measured against a mono-elemental standard solution (Spex CertiPrep) of known concen-tration( ± 5 ‰ )bysample-standard-samplebracketingtechniqueonthe multi-collector inductively coupled plasma mass spectrometer(MC e ICPMS) at the Neptune Isotope Laboratory (NIL), RSMAS.Concentrationscalculatedbasedon 48 Ti, 49 Tiand 50 Tiisotopeswereidentical within uncertainties. The accuracy and precision of Timeasurements were evaluated by comparing three replicate ana-lyses of USGS certi fi ed reference materials BCR-2 and BHVO-2 with16 and 26 literature compilations (GEOREM, http://georem.mpch-mainz.gwdg.de/), respectively. Results for both reference mate-rials agreed well with literature values within analytical un-certainties (Fig. SI-4). Contributions from the procedural blankwere negligible (smaller than 0.1%). Nevertheless, corrections wereapplied to all measurements prior to calculating Ti concentrations.Aeolian  fl uxes were calculated based on Ti concentrations inpeat ash residues (Ti ash ) and assuming an average Ti concentrationof 0.40% (4010  m g g  1 ) in continental crust (Wedepohl, 1995)through the following equations (Bao et al., 2012): Ti ash  m  g g   1   ¼  Ti  m  g g   1     Ash ð % Þ 100  = ð 0 : 4 Þ  (6)  Aeolian Flux  m  g cm  2  y  1   ¼  Ti ash  m  g g   1    r dry   g cm  3    r   cm y  1  (7) where  “ r  ” : is the average accumulation rate (mm y  1 ) and  “ r dry ”  isthe dry bulk density (g cm  3 ) measured as described above.  3.8. Lipid extraction and n-alkane separation Total lipids were extracted from approximately 1 g of 97 driedand homogenized peat samples (average resolution of ~140 Yr)with a solution of 2:1 dichloromethane:methanol (DCM/MeOH)using an Accelerated Solvent Extractor (Dionex ASE-200). The lipidextracts werepartitioned intothe organic phase followingadditionof methanol and NaCl-saturated water. Samples were concentratedto 1 ml using a Turbo-vap system under N 2  gas and half of thesample was used to separate  n -alkanes from the remaining lipidsthroughcolumnchemistry.Sampleswereblownto “  justdry ” underN 2  and re-dissolved in hexane. Columns were packed with 5%water-deactivated silica gel (100 e 200 mesh) using hexane. Sam-ples were eluted by three solvents,10 mL of hexane, 10 mL of 25%toluene in hexane and 10 mL of MeOH. The  fi rst fraction wasreduced to 0.5 mL under N 2  using the Turbo-vap. Concentrations of all  n -alkanes present in the sample were identi fi ed using gaschromatography mass spectrometry (GC e MS) and individual  n -alkanes were quanti fi ed relative to an internal standard (5 a -cho-lestane). The abundance of hydrophilic  fl oating and submergedplants (medium chain  n -alkanes: C 23 , C 25 ) relative to terrestrialplants (characterized by long chain  n -alkane: C 29 , C 31 ) was quan-ti fi ed according to  “ P aqueous ”  (P aq ) ratio employing the equationbelow (Ficken et al., 2000): P  aq  ¼ ð C  23  þ  C  25 Þ = ð C  23  þ  C  25  þ  C  29  þ  C  31 Þ  (8)  3.9. Compound-speci  fi c hydrogen isotope analysis Alkanoic acids were separated from the total lipid extract of 39 samples (average resolution of ~340 Yr) using column chro-matography (5 cm    40 mm Pasteur pipette, NH 2  sepra bulkpacking, 60 Å). Separation was achieved by eluting with 2:1DCM:isopropanol, followed by 4% formic acid in diethylether,yielding neutral and acid fractions, respectively. The acid fractionwas esteri fi ed with 5% HCl and 95% methanol of known isotopiccomposition at 70   C for 12 h to yield corresponding fatty acidmethyl esters (FAMEs). FAMEs were obtained by liquid-liquid-extraction using hexane as a non-polar solvent, and traces of water picked up during sample processing were removed bypassing the sample through a column of anhydrous Na 2 SO 4 .FAMEs were further puri fi ed using column chromatography(5 cm    40 mm Pasteur pipette, 5% water-deactivated silica gel,100 e 200 mesh), eluting with hexane, followed by FAMEs elutedwith DCM.Compound speci fi c hydrogen isotopic values were obtainedusing gas chromatography isotope ratio mass spectrometry(GC e IRMS). We used a Thermo Scienti fi c ® Trace gas chromato-graph equipped with a Rxi-5ms column (30 m    0.25 mm,  fi lmthickness1 m m)andaprogrammabletemperaturevaporizing(PTV)injector operated in solvent split mode with an evaporation tem-perature of 60   C. The GC was connected via a GC Isolink with apyrolysis furnace (at 1400   C) via a Con fl o IV interface to a Del-taVPlus isotope ratio mass spectrometer. The H 3 þ -factor wasdetermined daily to test linearity and accounted for 6.1 ppm mV   1 on average. Reference peaks of H 2  were co-injected between  n -alkanoic acid peaks during the course of a GC e IRMS run; two of these peaks were used for standardization of the isotopic analysis,while the remainders were treated as unknowns to assess preci-sion. Except for the case of co-elution, precision of these replicateswas better than 0.6 ‰ . The data were normalized to the VSMOW/SLAP hydrogen isotopic scale by comparing with an external stan-dardcontaining15 n -alkanecompounds(C 16 toC 30 )withknown d Dvalues obtained from A. Schimmelmann, Indiana University,  A. Shari  fi  et al. / Quaternary Science Reviews 123 (2015) 215 e  230  219
Search
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