JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 103,NO. D15, PAGES 18,885-18,894, AUGUST 20, 1998
Is a periglacial biota responsiblefor enhanceddielectric responsein basal ice from the Greenland Ice Core Project ice core? Jean-Louis Tison,1 RolandSouchez, 1EricW. Wolff,2 JohnC. Moore,3 MichelR. Legrand, 4 andMarfinedeAngelis 4 in the 6 rn of basaldielectric silty ice sequence from the GreenlandIce Core Pr ect (GRIP) Abstract. Athe detailed profiling (DEP) conductivity profile (oO75% effi-
ciency) to explain the value of 1 observedfor the DEP peaksof the YoungerDryas.This is certainlyalsotruefor the results from the basal ice described here and therefore
suggeststhat part of the ammoniuminitially presentin the gaseousform (NH3), as shownin this section,was
nextlogicalstepwasto measure NH4+ andCI' profilesin
transformed in NH4+ priorto incorporation intothe ice
the basalice sequence.Theseare shownin Figure 4a and are comparedwith the c•ooprofile. The chloride profile showsa contrastingbehaviorwith very low values(about 1 gM CI') throughoutall the silty ice and showshigher valuesin the clear ice layers of units 1 and 2 (7-11 gM CI'). This fits the idea of contrastedorigins for the two end-membersin the mixing model [Souchezet al., 1995a, b]. The clearice of the baseof the growingice sheet'sendmember,thoughtto have formed closerto the coastin the easternmountainous range,carrieslower/5 values(colder ice) and higherC1øvaluesrelativeto the silty ice of local origin, formed "in- situ,"far away from seaspraysources, and in the absenceof the present-dayice sheet(warmer ice).
lattice, to sustaina sufficientproductionof 2-D defects. This is quite plausible, even in the presenceof very limited amounts of liquid water, given the very high solubility of NH3. Two main conclusionscan be drawn from the developments above: 1. The assumption of negligibleacidityin the basalice layerappearsto be correct. 2. For the threepoints(locatedat 3022.275, 3022.328, and 3023.298 m depth,respectively)wherechloridealso contributessignificantlyto the calculatedconductivity,it is clear that (2) overestimatesc•oo.In fact, no ice (except dopedice) hasbeenmeasured beforein whichbothcomponentshavebeenpresent.Equation(2) may notbe appropriate in this case because the L and D defects can combine,
Onthecontrary, theNH4+ profilefaithfullymimicsthe and their concentrationsare the resultof an equilibrium.A revisedequationshouldaccountfor thisinteraction,which will reducec•oobelow that expectedfrom (2). Estimatesin which we make a crudeattemptto allow for this alsogive a goodfit to the data. However,for this basalice in general, the assumptionthat ammoniumdominatesthe conbasalice,the[H+] is negligible. It should benotedin this ductivityseemscorrect. regardthat sincethe ionic balance([H-^c¾])is negativefor c•oo profile, andthe correlationbetweenthe two variablesis strong(Figure 4b), suggestinga causalitylink. It is therefore temptingto apply (1) to the basalice of GRIP to test its validity. Since no ECM measurementis available,we will consideras a first approximationthat as in the Dye-3
all thesamples measured (downto -50 geql-1),thisassertion is probablyvalid. Equation(1) is thereforereducedto
{Joo(gSm -1) = 9 + [NH4 +] + 0.55[C1']
(2)
and calculatedc•ooare plottedin Figure 4c, wherethey are comparedwith the observedvalues.The relative error on c•oo(expressedas (C•calc.- C•obs.)/C•obs. in percent)was alsocomputedandis shownin Figure4f. Major departures from the observedvalues clearly occur in the clear ice wherethe chloridecontentis higher(up to 60% relativeerror). Therefore(2) wasfurtherreducedto pureice andNH 4 conductivitydependency only:
{Joo(gSm 'l) = 9 + [NH4 +]
(3)
and the results were plotted in Figures 4d and 4g. The agreementis now quitegood,with a meanrelativeerrorof 1.61% and maximumdepartures from the observedvalues of about20%. This is exactly the range of precisionreckoned by Moore et al. [1994b] for the ammoniumcalibration of c•oo,higher up in the GRIP core, given the relatively coarsechemicalsamplingprocedure.For comparison,Figures4e and 4h plot the calculatedc•ooandrelative errorsusinga bestfit linearregressionacrossthe data.The result is not significantlydifferent, apart from a slightly better mean value on relative errors (-0.06 instead of
1.61%).Theproportionality factorbetween c•oo and[NH4+] in the bestfit linear regressionis alsoa bit higherthan 1 (1.12). Moore et al. [1994a] gave an estimateof this proportionalityfactor from theoreticalconsiderations on the relationshipbetweenc•ooand the concentrations, charges, and the mobility of the Bjermm defects.They suggestthat oneD defectshouldproducea maximumvalue of 0.66 for the proportionalityconstant,and they further deducethat
6.
Origin of the Ammonium Signal Ammonium(NH4+) is originallyemittedasammonia
(NH3), predominantlyfrom continentalsourcesof biogenic origin, mainly bacterial decomposition in soils and biomassburning.Legrand et al. [1992] and Fuhrer et al.
[ 1993]discussed theoriginof theNH4+ signalin thedetailed profiles measuredalong the GRIP core above the basalsilty ice. The backgroundlevel is generallyhigherin the present interglacial (0.4-0.8 gM) than in the last glacialperiod(around0.2 gM). Theseauthorssuggestthat soil emissionsare the principal sourcefor this low-signal record,with a possibleslightcontributionfrom the seaice zone.Ammoniumpeaksoccurin bothclimaticperiodsbut reachmuch higher levels in the presentinterglacial(up to
600ngg-1= 33gM) thanduring thelastglacial (onlyup to a few gM). Legrandet al. comparethe strongand narrow peaksthey observebetween100 and 600 m depthin the core (330-2500 yearsB.P.) to the measurements of organic acidsat the samelevels (formate(HCOO'), acetate (CH3COO-), and oxalate(C204'-)). Formateis dominant,
showing clearsynchronous peakswithNH4+ andrelative abundances closetothoseexpected fromtheNH4+?rlCOO ' molarratio.NH4+ thusprimarilyoccursas ammonium formatein the peaks,and this is usedby Legrandet al. to demonstratethat biomassburningfrom high-latitudeforest
firesis themostlikelysource forthehighNH4+ levelsin the studied sections of the core.
The picturelooks,however,quitedifferentin the caseof the basalice from GRIP, as shownin Figure5. Oxalateis now the dominantorganicacid species(about2 ordersof magnitudehigherthanin all of the othersectionsmeasured in the GRIP core) and of all three organicacidsmeasured
18,892
TISON ET AL.: PERIGLACIAL BIOTA AND DIELECTRIC SIGNAL IN THE GRIP CORE
Formate (HCOO'- gM ) 0.00
400
2.23
4.45
6.68
Acetate (CH3CO0' - gM )
Oxalate (C204 = - gM )
8.90 0.00 1.70 3.40 5.10 6.80 8.50 0.00
5.68
11.35
17.03
22.70 22.20
'/'
350 ß
300
ß
ß
19.43
16.65
250
13.88
200
11.10
150
-8.33
100
5.55 41,
5O
2.78
0
.................. 100
• .................. 200 300
0.00
400
Formate (HCOO' - ng.g-1)
0
100
200
300
400
500
0
Acetate (CH3CO0' - ng.g4)
500
1000
1500
2000
Oxalate (C204 = - ng.g-1)
Figure 5. Correlationsare shownbetweenammoniumand(a) formate,(b) acetate,and(c) oxalatefor the samplesin the basalsilty ice sequence from GRIP.
in this case, is the one that showsthe strongestpositive
medium[Stummand Morgan, 1996].We haveseenfrom
correlation (r2 = 0.81) with ammonium. Oxalateis known thecomparison betweencalculated andobserved DEP that by to be presentin plants(lichens,for example)and alsoto [H+] mustbe negligible,andthisis furthersupported be oneof theby-products of thebacterialdegradation of organicmatter.In particular,productionof ammoniumoxalateduringuricacidbreakdownis a well-described process in the literature(see, for example,Hutchinson[1950] and Legrandet al. [1998]). In this regard,the closeassociation of oxalate and ammoniumhas beenrecentlydemonstrated as being one of the fingerprints of ornithogenic soils (contaminatedby animal excreta, in that case,penguin's droppings)in the aerosolsand snowdepositsof the Terre Adelie area [Legrandet al., 1998]. The datafrom Figure5 thereforesuggestthe existenceof considerablydeglaciated areas,with significantplant coverand efficientbiological activity, in the closevicinity of the formationsite of the local end-member involved in the basal ice sequenceat GRIP. Conversely, this is a further argumentto demonstratethat part of the ice in this basalsequenceoriginated in a periglacial environment in the central part of Greenland,before the settlingof the present-dayice sheet. In situ growth of the ice sheetor "invasion"by a growing ice sheet born in the Eastern Greenland mountain range seem equally plausible mechanismsin this context.
the observationthat the ion.icbalance([H-^c¾])is negative
for all the samples.The full chemicalpictureof thebasal ice layer is, however,far more complexsincemajor
cations like Ca++andMg++alsoshowa goodlinearrelationshipto C204-- and since someof the ammonium could also exist in the formate form (see Figure 5a), but thisis beyondthe scopeof thepresentwork. 7.
Conclusion
DEP conductivities of the basalsiltyice sequence appear to be essentiallycontrolledby intracrystallineprocesses dueto pureice andto the effectof additionalBjerrumD de-
fectscaused by theinclusion of bothNH3 andNH4+ into the ice lattice. In the past therehasbeen somediscussion [Moore et al., 1994b] of whetherthe dielectricresponseof ammoniumpeaksin shallowerpartsof the GRIP corewas entirely due to ammonium or could also indicate the formation
of L defects from formate.
The fact that the
pointswith the highestammoniumcontent(in Figure 5) are clearly not associatedwith formatebut probablywith However, the contrast in chlorine content between the oxalate,yet they fit on the samecalibrationcurve as the local and the "foreign"end-members(seeabove),between earlier results, strongly suggeststhat only ammonium the stableisotopesand the gas contentsof the upperunit (producingD defects)is involved.The electricalbehavior of the basalice [Souchez,1997],andbetweentheisotopic of the basalsilty ice differs from that observedhigherup compositionof the embeddeddebris[Weiset al., 1997] all in thecore,wherehighooo(observedduringwarmclimatic favor the latterhypothesis. episodes)alsocorrespond to highdc conductivities (ECM), As underlinedin section4, the behaviorof oooin Figure indicating a considerable contribution from[H+] eventually 3f showsthat ammoniumis tightly associatedwith the occurringat the boundariesof the crystals.This is undergaseousphase,which suggeststhat it initially partly oc- standablesince ice formed (or transported)close to the curred as NH 3. This implies alkaline conditionsin the bedrockand loaded with debrisinclusionsis likely to be
TISON ET AL.: PERIGLACIAL
BIOTA AND DIELECTRIC
alkaline. Another unique feature of the basal sequenceis the sourcefor the NH3 emissions.Whilst long-distance transportof the productsfrom biomassburning events seemsto control the bulk of the peaksin the higher parts of the GRIP core, the dominance of oxalate and its clear
correlationwith ammoniumin the basal sequencesupport local biogenic production in a periglacial biotope as a likely source.Consideringthat thereis a net excessof oxalate as compared to what would result from uric acid degradationalone, this local biogenic productioncovers both plants and animals. This is of crucial importance sinceit obviouslyimplies the absenceof the ice- sheetto allow for significantinputs from this low level and lowintensitysource.It thereforestronglycorroborates the findingsdeducedfrom othervariablesanddiscussed in previous studies [Souchez et al., 1995a, b, 1994; Tison et al., 1994]. Acknowledgments. This work is a contribution to the GreenlandIce Core Project (GRIP) organizedby the European ScienceFoundation.We thank the GRIP participantsand supporters for their cooperative effort. We also thank the National Science Foundationsin Belgium, Denmark, France, Germany, Iceland, Italy, Switzerland, and the United
SIGNAL IN THE GRIP CORE
18,893
Greenland) ice cores by ion chromatography, J. Chromatogr., 640, 250-258, 1993. Legrand, M., F. Ducroz, D. Wagenbach,R. Mulvaney, and J. Hall, Ammonium in coastal antarctic aerosol and snow:
Role of polar ocean and penguin emissions,J. Geophys. Res., 103,
11,043-11,056,
1998.
Moore, J.C., and J.G. Paren, A new technique for dielectric logging of Antarctic ice cores, J. Phys., Paris, 48(C1), 155-160,
1987.
Moore, J.C., R. Mulvaney, and J.G. Paren, Dielectric stratigraphy of ice: A new techniquefor determiningtotal ionic concentrations in polar ice cores, Geophys. Res. Lett., 16(10), 1177-1180, 1989. Moore, J.C., J.G. Paren, and H. Oerter, Sea salt dependent electrical conduction in polar ice, J. Geophys. Res., 97(B13), 19,803-19,812, 1992a. Moore, J.C., E.W. Wolff, H.B. Clausen, and C.U. Hammer,
The chemical basis for the electrical stratigraphyof ice, J. Geophys. Res., 97(B2), 1887-1896, 1992b. Moore, J.C., A.P. Reid, and J. Kipfstuhl, Microstructure and electrical properties of marine ice and its relationship to meteoric ice and sea ice, J. Geophys.Res., 99(C3), 51715180, 1994a. Moore, J.C., E.W. Wolff, H.B. Clausen, C.U. Hammer, M.R.
Legrand,and K. Fuhrer, Electrical responseof the SummitGreenland ice core to ammonium, sulphuric acid, and hydrochloric acid, Geophys. Res. Lett., 21(7), 565-568, 1994b.
Kingdom, aswellastheXII Directorate of theE.C.This'paper Mulvaney, R., E.W. Wolff, and K. Oates, Sulphuric acid at is a contributionto the Belgian Global Change Program (Science Policy Office). J.-L. Tison is ResearchAssociate at the Belgian F.N.R.S.
References
grain boundariesin Antarctic ice, Nature, 331, 247-249, 1988.
Paren, J.G., and J.W. Glen, Electrical behaviour of finely divided ice, J. Glaciol., 21(85), 173-191, 1978. Petrenko,V.F., Electrical Propertiesof ice, Spec.Rep. 93-20, 69 pp., Cold Reg. Res. and Eng. Lab., Hanover, N.H., 1993.
Alder, B., J. Geiss,N. Groegler,and A. Renaud,Gas composi- Souchez, R., The build up of the ice sheet in Central tion in ice samples collected by E.G.I.G. in Greenland, Greenland, J. Geophys. Res., 102(C12), 26,317-26,323, Medd. Groenl., 177(2), 101-107, 1969. Camplin, G.C., J.W. Glen, and J.G. Paren, Theoretical models for interpretingthe dielectricbehaviourof HF-dopedice, J. Glaciol., 21(85), 123-142, 1978. Caroll, D., Ion exchangein clays and other minerals, Geol. Soc. Am. Bull., 70, 749-780,
1959.
Davidson, D.W., Clathrate hydrates, in Water: A Comprehensive Treatise, vol.2, edited by F. Franks, pp. 115-234, Plenum, New York, 1973.
1997.
Souchez, R., J.-L.
Tison, R. Lorrain,
L. Janssens, M.
Stievenard, J. Jouzel, A. Sveinbj6rnsdottir, and S.J. Johnsen,Stable isotopesin the basal silty ice preservedin the GreenlandIce Sheetat Summit:Environmentalimplications, Geophys. Res. Lett., 21(8), 693-696, 1994. Souchez,R., L. Janssens,M. Lemmens,and B. Stauffer, Very low oxygen concentration in basal ice from Summit, Central Greenland, Geophys. Res. Lett., 22(15), 2001-
Fuhrer, K., A. Neftel, M. Anklin, and V. Maggi, Continuous 2004, 1995a. measurements of hydrogen peroxide, formaldehyde, Souchez, R., M. Lemmens, and J. Chappellaz, Flow-induced calcium and ammoniumconcentrations along the new GRIP ice core from Summit, Central Greenland, Atmos. Environ., 27A(12), 1873-1880, 1993.
Gough, S.R., and D.W. Davidson, Dielectric properties of clathrateices, in Physicsand Chemistryof Ice, editedby.E. Whalley, S.J. Jones,and L.W. Gold, pp. 51-55, R. Soc. of Can., Ottawa, Ont., Canada, 1973.
Hammer, C.U., Acidity of polar ice cores in relation to absolute dating, past volcanism and radio echoes, J. Glaciol., 25(93), 359-372,
1980.
Hammer, C.U., Initial direct current in the build-up of space chargesand the acidity of ice cores, J. Phys. Chem., 87, 4099-4103,
1983.
Hellferich, F., Ion Exchange, 624 pp., McGraw-Hill, New York, 1962.
Hobbs, P.V., Ice Physics, 837 pp., Clarendon, Oxford, England, 1974. Hutchinson,G.E., Survey of existing knowledgeof biogeochemistry, 3, The biogeochemistryof vertebrate excreta, Bull. Am. Mus. Nat. Hist., 96, 71-94, 1950.
Legrand, M., M. De Angelis, T. Staffelbach,A. Neftel, and B. Stauffer, Large perturbationsof ammonium and organic acids
content
in
the
Summit-Greenland
ice
core:
Fingerprint from forest fires?, Geophys.Res. Lett., 19(5), 473-475,
1992.
Legrand,M., M. De Angelis, and F. Maupetit,Field investigation of major and minor ions along Summit (Central
mixing intheGRIPbasal icededuced fromtheCO2andCH4
records,Geophys.Res. Lett., 22(1), 41-44, 1995b. Stumm, W, and J.J. Morgan, Aquatic Chemistry: Chemical Equilibria and Rates in Natural Waters,3rd ed., 1022 pp., Wiley-Interscience, New York, 1996. Tison, J.-L., Diamond -wire saw cutting techniquefor investigating textures and fabrics of debris-ladenice and brittle ice, J. Glaciol., 40(135), 410-414, 1994. Tison, J.-L., T. Thorsteinsson,R. Lorrain, and J. Kipfstuhl, Origin and developmentof texturesand fabricsin basalice at Summit, Central Greenland, Earth Planet. Sci. Lett., 125(1-4), 421-437, 1994. Weis, D., D. Demaiffe, R. Souchez, A.J. Gow, and D.A.
Meese, Ice sheet development in Central Greenland: Implicationsfrom the Nd, Sr and Pb isotopiccompositions of basal matedhal, Earth Planet. Sci. Lett., 150, 161-169, 1997.
Wolff,
E.W., J.C. Moore, H.B. Clausen, C.U. Hammer, J.
Kipfstuhl, and K. Fuhrer, Long-term changesin the acid and salt concentrationsof the Greenland Ice Core Project ice core from electrical stratigraphy, J. Geophys. Res., 100(D8), 16,249-16,263, 1995. Wolff, E.W., W.D. Miners, J.C. Moore, and J. Paren, Factors
controlling the electrical conductivityof ice from the polar regions: A summary, J. Phys. Chem., BlO1, 6090-6094, 1997a.
18,894
TISON ET AL.: PERIGLACIALBIOTA AND DIELECTRICSIGNAL IN THE GRIP CORE
Wolff, E.W., J.C. Moore, H.B. Clausen, and C.U. Hammer,
R. Souchez and J.-L. Tison, Universit6 Libre de Bruxelles,
Climatic implications of background acidity and other chemistryderived from electrical studiesof the Greenland Ice Core Project ice core, J. Geophys. Res., 102(C12),
D6partementdes Sciencesde la Terre et de l'Environnement,
26,325-26,332,
E.W. Wolff, British Antarctic Survey, National Environment Research Council, Cambridge, England, U.K. (e-mail:
[email protected]. ac.uk)
1997b.
M. de Angelis and M.R. Legrand, Laboratoire de Glaciologie et G6ophysiquede l'Environnement,CNRS, Grenoble, France. (e-mail:
[email protected]) J.C. Moore, Arctic Center, University of Lapland, Rovianemi, Finland. (e-mail:
[email protected])
CP 160-03, Avenue F.D. Roosevelt 50, Bruxelles B-1050,
Belgium. (e-mail:
[email protected])
(ReceivedNovember 13, 1997; revisedMarch 26, 1998; acceptedMarch 31, 1998.)
