on the origin of em-i end-member - do plumes exist? · on the origin of em–i end-member ... os,...

NJbMinerAbh 179 1 85ndash100 Stuttgart August 2003

On the origin of EMndashI end-member

Michele Lustrino and Luigi Dallai Rome

With 3 figures

Lustrino M amp Dallai L (2003) On the origin of EMndashI end-member ndash N Jb Mi-ner Abh (179) 085 ndash100 Stuttgart

Abstract On the basis of geochemical data available for oceanic basalts distinctmantle reservoirs have been identified during the last decades Among these reservoirsthe EM ndashI end-member (Enriched Mantle type I) shows almost unique geochemicalcharacteristics The definition of the EMndash I end member and the petrogenetic proces-ses related to its origin are still debated the geochemical composition of EMndashI basaltshas been interpreted in terms of crust-mantle interaction during 1) mantle recycling ofsubducted crustal material 2) detachment of lithospheric keels in over-thickened re-gions 3) thermal erosion by anomalously hot mantle Key features of the EM ndashI end-member are low uranogenic Pb ratios (206Pb204Pb lt 17) slightly radiogenic Sr iso-topes (87Sr86Sr ~ 0706) unradiogenic Nd (143Nd144Nd ~ 05121) unradiogenic Hf(176Hf177Hf ~ 02826) slightly radiogenic Os (187Os188Os ~ 0135ndash0145) high207Pb206Pb (gt 085) and 208Pb206Pb (gt208) EMndashI basalts show apparent or inexis-tent Ba Pb Eu or Sr anomalies in primitive mantle-normalized diagrams and have BaNb ratios ranging from 35 to 474 CePb from 12 to 246 NbU from 105 to 718 SrNd from 62 to 364 and EuEu from 083 to 125 On these basis the origin and evo-lution of the EM ndashI end-member cannot be explained by a single petrological modeland even the definition of elemental and isotopic parameters is controversial

Introduction

The description of geochemical heterogeneities in the Earthrsquos mantle and theproposal of models to explain these anomalies have been some of the most im-portant goals of the modern igneous petrology (eg Zindler amp Hart 1986Hofmann 1997) Many studies carried out during the last decades have evi-denced chemical anomalies deviating from the relatively homogeneous upperasthenospheric depleted mantle composition In particular incompatible traceelement investigations coupled with radiogenic and stable isotopic (Sr Nd PbOs O Hf He and Ne) studies of oceanic basalts allowed for the identificationof distinct geochemical reservoirs in the Earthrsquos mantle

DOI 1011270077-775720030179-0085 0077-7757030179-0085 $ 400atilde 2003 E Schweizerbartrsquosche Verlagsbuchhandlung D-70176 Stuttgart

NJBMin_2003_179_p85ESchweizerbartsche

86 M Lustrino and L Dallai

Fig 1 SrndashNdndash Pb isotopic systematics of EMndashI basalts from continental and oceanicsettings Pitcairn islands and seamounts (Pacific Ocean Woodhead amp Devey 1993Eiler et al 1995 Eisele et al 2002) AphanasyndashNikitin Rise (Indian Ocean Maho-ney et al 1996 Borisova et al 2001) Urubici magma-type (Paranagrave Brazil Peate etal 1999) Walvis Ridge (Atlantic Ocean Richardson et al 1984) Kerguelen plateau(Indian Ocean Frey et al 2002) Sardinia (Italy Gasperini et al 2000 Lustrino etal 2000 2002) DMM = Depleted Morb Mantle HIMU = High m EMndashI and EMndashII= Enriched Mantle I and II NHRL = Northern Hemisphere Reference Line Hart1984) Contour of Pb isotopic field of DMM is a broken line because of the extremevariability of Atlantic Pacific and Indian MORBs in terms of 206Pb204Pb 207Pb204Pband 208Pb204Pb

On the origin of EM ndashI end-member 87

Among oceanic island basalts (OIBs) at least three geochemical end-mem-bers have been identified (Zindler amp Hart 1986) HIMU (high m where m is238U204Pb ratio) EMndashI and EMndashII (Enriched Mantle I and II) Mixing ofthese three end-members together with the DMM (Depleted MORB Mantle)component can explain many of the geochemical features of oceanic basalts(eg Carlson 1995) Together with these mantle end-members other termssuch as EMndashIII (Heaman 1989) PREMA (PREvailing MAntle Stein ampHofmann 1994) FOZO (FOcus ZOne Hart et al 1992) PHEM (PrimitiveHe Mantle Farley amp Craig 1992) and C component (Hanan amp Graham1996) have been proposed as other possible reservoirs

Originally thought for basalts from oceanic islands these terms have beentransposed and widely used also to explain the petrogenesis of basalts occur-ring in continental settings (Chung et al 1995 Milner amp Le Roex 1996Peate et al 1999 Lustrino et al 2000 2002) Among these reservoirs thedefinition of the EMndashI end-member is probably the most puzzling Since thepioneering work of Zindler amp Hart (1986) the type locality for EMndashI haschanged from Walvis Ridge (ODP Site 525A) in Atlantic Ocean (Richardsonet al 1984 Zindler amp Hart 1986) to Pitcairn islands and seamounts in Paci-fic Ocean (Woodhead amp Devey 1993 Eisele et al 2002) Tristan and Inac-cessible islands in Atlantic Ocean (Hofmann 1997) Aphanasy Nikitin Rise(Mahoney et al 1996 Borisova et al 2001) and eventually Kerguelen is-lands in the Indian Ocean (Frey et al 2002) Trace element and isotopic com-positions have been interpreted in terms of mixing of various components suchas pelagic andor terrigenous sediments sub-continental lithospheric mantle(SCLM) ancient oceanic plateaux oceanic crust lower continental crustThese components could be recycled either via 1) subduction 2) delaminationand detachment processes 3) thermal erosion by anomalously hot mantle in avariably depleted asthenosphere Metasomatism of mantle peridotite by inter-action with low VPb low RbSr LREE-rich carbonatite melts has been pro-posed as on alternative model to explain the EMndashI end-member Features(Carlson 1995) Up to now it has not been possible to define a commonmechanism (or combination of processes) which can account for the geochem-ical features of oceanic and continental EMndashI basalts from different geologi-cal settings

Radiogenic and stable isotopes constraints

The key features of the EMndashI end-member can be summarized as follow(Fig 1) low uranogenic Pb ratios (eg 206Pb204Pb lt 17) coupled with slightlyradiogenic Sr isotopes (87Sr86Sr ~ 0706) unradiogenic Nd (143Nd144Nd ~05121) unradiogenic Hf (176Hf177Hf ~ 02826) and slightly radiogenic Os


(187Os188Os ~ 0135 ndash0145) Hf and Os isotopes of EMndashI have not beenplotted in Fig 1 because of the paucity of data available in literature Typicalfeatures of EMndashI basalts are also high 207Pb206Pb (gt085) and high 208Pb206Pb (gt208 Lustrino 2000) With regards to the Sr- Nd- and Pb-isotopiccompositions many of the EMndashI basalts define broad compositional fields(Fig 1 andashc) which can be interpreted in terms of mixing processes between apure EMndashI reservoir and the HIMU-DMM component the DMM prevailingover HIMU Still this simple mixing model is not sufficient to explain thegeochemical characteristics of several EMndashI basalts (Fig 1 dndash f) Indeed thestrong variations of the 207Pb204Pb and 208Pb204Pb (and 207Pb206Pb and208Pb206Pb) ratios unlikely reflect a single EMndashI reservoir In particular the207Pb204Pb ratios of the Aphanasy ndashNikitin Rise basalts range from ~154 to156 with a nearly constant 206Pb204Pb (~17) Similarly decoupling of the207Pb204Pb ratio from the 206Pb204Pb ratio is envisaged for Kerguelen and toa minor extent for Sardinian EMndashI basalts The 208Pb204Pb (thorogenic) ra-tios are even more puzzling the Pitcairn basalts show constant and high ratiosfor these isotopes (208Pb204Pb ~ 389) above those of other EMndashI basaltswith comparable 206Pb204Pb ratios (208Pb204Pb ~ 375ndash382 with 206Pb204Pb~ 175ndash180)

Proposed models

The Pitcairn islands and seamounts are the type-locality for EMndashI basalts(Woodhead amp Devey 1993) so that a wealth of data have been produced forthese lavas On the basis of trace element and SrndashNdndashPbndashOsndashHf isotopic ra-tios Eisele et al (2002) inferred that 1) the geochemical characteristics of theEMndashI component are strongly influenced by a small fraction of pelagic andterrigenous sediments (generally lt 5 in fraction) 2) even a consistentamount (up to 60 ) of recycled SCLM does not to contribute significantly tothe Os isotopic budget Because the addiction of Os-rich and Re-poor SCLMresults in low (subchondritic) 187Os188Os ratios (lt0127) the relatively radio-genic Os composition of EMndashI (187Os188Os gt 0135) requires the involvementof some Re-rich component with high time-integrated 187Re188Os ratios

Addiction of ancient SCLM andor sediment recycling have been calledupon for the origin of the EMndashI signature of continental basalts (eg Chunget al 1995 Milner amp Le Roex 1996 Smith 1998) the involvement of lowercontinental crust may also produce the trace element and isotopic composi-tions of EMndashI-like oceanic and continental igneous products (eg Hart ampZindler 1989 Carlson 1995 Lustrino et al 2000 Tatsumi 2000 Bori-sova et al 2001 Frey et al 2002)

On the origin of EMndashI end-member 89

The lower crust-like signature of the EMndashI basalts may result from asource process rather than magma contamination en route to the surface (egLustrino 2000) as consequence of delamination andor detachment of densegranulitic to eclogitic lower crust (density up to 35 gcm3 Wolf amp Wyllie1994) Amphibolite to granulite (and eclogite) metamorphic reactions in thelower crust are associated with partial melting and density increase (with in-creasing garnet content) LILE and LREE are readily mobilized during partialmelting and low RbSr and UPb ratios can be produced (eg Rudnick ampTaylor 1987) The SmNd ratio is virtually unaffected by this process becauseREEs are less mobile under these metamorphic conditions (eg Farmer 1992Dostal et al 1998) Lower crust has lower RbSr (002 ndash014) and lower UPb(005 ndash011) ratios (and lower time-integrated 87Sr86Sr and 206Pb204Pb) thanupper crust (RbSr = 026ndash035 UPb = 008 ndash015 Rudnick amp Fountain1995 Wedepohl 1995 Gao et al 1998 Liu et al 2001) conversely SmNd(0120 for average lower continental crust and 0121 for average upper crust)and HfndashNd-isotopic ratios of these two reservoirs overlap (Vervoort et al2000 and references therein) These characteristics can be used to stress thehigher mobility of 87Rb and 238U isotopes relative to 147Sm by considering therelations of 87Sr86Sr 143Nd144Nd and 206Pb204Pb with Hf isotopic ratios Theupper and lower crustal lithologies have generally eHf and eNd lt 0 whereasmantle rocks have positive values Therefore mantle sources contaminatedwith lower crustal lithologies can evolve toward slightly radiogenic 87Sr86Srunradiogenic 143Nd144Nd and unradiogenic 206Pb204Pb compositions In par-ticular the SrndashNd isotopic field of EMndashI basalts is offset to low Nd frommantle array as evidenced by the low DNd (down to ndash50) (Zindler et al1982 Hart et al 1986)

The Os isotopic ratios of lower crust can vary from subchondritic to sup-rachondritic values in the case of cumuliticrestitic or differentiate composi-tions respectively (eg Esser et al 1993 Reisberg et al 1993) No correla-tion between Os and Pb isotopic ratios has been evidenced (Eisele et al2002) this is likely due to the influence of the recycled material (oceanic crustvs sediments) on the Pb isotopic composition Overall the covariationsamong Sr Nd Pb and Hf isotope compositions can be interpreted as resultingfrom similar factors whereas the Os isotopic composition reflects the overallamount of recycled material and the degree of peridotite melting Similarlythe Hf isotopic composition of the lower crust is variable from subchondriticvalues (eHf lt ndash30 in the case of cumulitic composition andor with metamor-phic garnet) to radiogenic (eHf gt 20 in the case of differentiation processes inthe presence of garnet residual from melting or differentiation events (egVervoort et al 2000)


The study of oxygen isotope composition of EMndashI basalts and their mainphases (eg olivine clinopyroxene) can be considered a potential tool to tracethe addiction of crustal material to their mantle source This because of thestrong fractionation between 18O and 16O isotopes in upper crustal and mantlereservoirs (James 1981 Ito et al 1987 Baldridge et al 1996) High d18O val-ues (up to 25 permil) are typical of terrigenous and pelagic sediments and crustalrocks variably exposed to low temperatures (lt300 C) water-rock interaction(eg Eiler et al 2000 and references therein) On the other hand high tem-perature (gt350 C) hydrothermal alteration shift original d18O toward very lowvalues (down to ndash10 permil Gregory amp Taylor 1981) Peridotitic mantle andMORBs have a rather uniform d18O values (55ndash59 permil Ito et al 1987 Mat-tey et al 1994 Eiler et al 1997) and the measurements that deviate sensiblyfrom this range (down to 46 permil and up to 75 permil Harmon amp Hoefs 1995)have been related to different degrees of alteration of the rocks andor theirMORB glass fraction (Eiler et al 1997) Investigation performed via laserfluorination of MORB olivine (Eiler et al 1997) evidenced uniform d18O val-ues (516 plusmn 009 permil) these values are consistent with the Dbasaltndasholivine = 04 ndash05 permil inferred on the basis of several studies (Anderson et al 1971 Matt-hews et al 1998 Appora et al 2003) Thus olivine d18O values ~52 plusmn 02 permilcan be considered the baseline to check isotopic anomalies of the source

Woodhead et al (1993) interpreted the high d18O values (59ndash74 permil) meas-ured for basaltic glasses of Pitcairn seamounts in terms of source contamina-tion by subducted crustal material with high d18O (gt15 permil) Accordingly the18O-enriched compositions of Pitcarn glasses would result from the incorpora-tion of significant amounts (up to 9 wt ) of crustal rocks A substantial con-tribution of recycled sediments to Pitcairn mantle source was ruled out on thebasis of homogeneous and low d18O values of olivine phenocrysts (521 permil plusmn008) similar to the d18O values of olivine from MORBs (Eiler et al 1995)The high whole rocks d18O values were interpreted as due mostly to sample al-teration and not to anomalous mantle sources (Eiler et al 1995) The 18O16Oratios of olivine from Pitcairn limit the possible amount of crustal material in-volved in EMndashI source to be 1ndash2 wt (Eiler et al 1995) More data areneeded to constrain the role of crustal recycling in EMndashI basalts and to evalu-ate the d18O variations as a function of different degrees of partial melting (Ei-ler et al 2000)

Further constraints on the genesis of the EMndashI mantle source could be ob-tained by noble gas (HendashNe) isotope systematics The 3He4He (RRA) ratiosof Polynesian EMndashIndashOIBs (~95 at Rarotonga Hanyu et al 1999 8ndash13 atPitcairn Eisele et al 1997 and references therein) and MORB values (~8ndash9Kurz et al 1982 Hilton et al 1993 Niederman et al 1997) overlap Thesevalues are lower than those reported for Loihi lavas (~20ndash30 Honda et al


1991 1993) and suggest that EMndashI basalts (and MORB) were derived from asource with high (U+Th)3He ratios These latter ratios may result from the in-teraction of subducted crustal rocks with very low 3He (primitive He) and high4He (derived by radioactive decay of U and Th) contents Thus the 3He4He ra-tio of EMndashI basalt would arise from the recycling of oceanic crust (very low3He4He) in the upper mantle whereas the Loihi lavas would represent melts ofundegassed mantle If this holds true the higher 3He concentration of MORB(between 10ndash11 and 10ndash9 cm3 STPg) relative to Loihi and Kilaueha (between10ndash13 and 10ndash10 cm3 STPg Hiyagon et al 1992 Honda et al 1993) is the op-posite of what expected for a depleted (and therefore degassed) MORB source(Honda et al 1993 Ozima amp Igarashi 2000) Moreover HIMU basalts show3He4He ratios higher (Hawaii and Iceland) and lower (St Helena MangaiaTubuai) than MORB thus requiring different origins for the HIMU source atleast from the He point of view (eg see Hanyu et al 1999 2001)

Neon isotopic data are virtually absent for EMndashI basalts and the availableanalyses are mostly confined to MORB and HIMU basalts These results indi-cate at least two components distinct from that of the present atmosphere a20Ne22Ne-rich solar (primitive) component and a 21Ne22Ne-rich nucleogenic(U- and Th-derived) component (Honda et al 1991 1993 Matsumoto et al1997 Hanyu et al 2001) The possible origins of the radiogenicnucleogeniccomponent (21Ne derived from U and Th) is still uncertain due to similar21Ne4He production rates (~1 10ndash7) in mantle and crust (Ozima amp Podosek2002) Therefore the 21Ne22Ne-rich component cannot be automatically inter-preted in terms of recycling of subducted crust (Honda et al 1993)

Trace elements constraints

The trace element compositions of EMndashI oceanic basalts (eg Ba Nb Pb SrEu Nd) are variable resulting in a wide range of inter-elemental ratios (egBaNb ranging from 35 to 474 CePb from 12 to 246 NbU from 105 to718 SrNd from 62 to 364 and EuEu from 083 to 125 Fig 2) The sameinter-elemental variability has been documented also for EMndashI-like basaltsemplaced on continental settings from Taiwan (Chung et al 1995 Smith1998) Urubici (Serra Geral formation Brazil Peate et al 1999) and Sardinia(Italy Gasperini et al 2000 Lustrino et al 2000 2002) BaNb = 109 ndash350 CePb = 33ndash211 NbU = 134 ndash508 SrNd = 87ndash350 and EuEu =101ndash125 (Fig 2) EMndashI basalts show primitive mantle-normalized patterns(Fig 3) with positive anomalies for barium (Sardinia Aphanasy-Nikitin RiseKerguelen Urubici) lead (Sardinia Aphanasy-Nikitin Rise Kerguelen) andstrontium (Sardinia Kerguelen) together with patterns where these anomaliesare virtually absent (Pitcairn) Similarly throughs at Nb can be envisaged only


Fig 2

in some EMndashI suites (eg Sardinia Aphanasy-Nikitin Rise Urubici Fig 3)The involvement of lower continental crust in the genesis of EMndashI basaltswould change the original trace element ratios because this reservoir has highand variable BaNb (16ndash224) and low and variable CePb (28ndash85) and NbU


Fig 2 Interelemental ratios for EM ndashI basalts from continental and oceanic settingsReferences as in Fig 1 Note the great variability of key trace element ratios and theimpossibility to propose unique values for the EMndashI end member

(8ndash25 Gao et al 1998 Liu et al 2001 and references therein) EuEu andSrNd ratios of lower crust are dependent on the amount of cumulate plagio-clase in the lower crust (see discussion in Frey et al 2002) On the basis ofthe above mentioned trace element data it is possible to distinguish EMndashI


Fig 3 Primitive mantle-normalized diagrams (normalizing factors after Sun ampMcDonough 1989) of the most mafic EM ndashI lavas references as in Fig 1 n = numberof samples Sardinia = hawaiite alkali basalt basanite tholeiitic basalt basaltic ande-site (Mg 071ndash 046) Pitcairn = tholeiitic and transitional basalt hawaiite (Mg 072 ndash051) Aphanasyndash Nikitin Rise = tholeiitic and transitional basalt (Mg 066 ndash047)Urubici = tholeiitic basalt and basaltic andesite (Mg 048 ndash044) Kerguelen = tholeii-tic and transitional basalt (Mg 069 ndash060)

from HIMU and DMM reservoirs (Weaver 1991 Carlson 1995 Dostal etal 1998) In contrast trace element content cannot be effectively used to dis-tinguish EMndashI from from EMndashII basalts because many of their chemicalcharacteristics are shared (eg Hemond et al 1994 Carlson 1995)


Concluding remarks

EMndashI end-member is one of the most puzzling aspects of modern isotopic to-pology The following conclusions arise from the data produced in recentyears

d EMndashI basalts are characterized by relatively high Ba Nb Pb Sr Eu andNd contents high but variable BaNb and SrNd and low and variable CePb and NbU ratios

d EMndashI end-member is likely characterized by relatively low time-integratedRbSr and low to very low SmNd and UPb ratios resulting in time-inte-grated slightly radiogenic Sr- unradiogenic Nd- and unradiogenic Pb-iso-topic compositions These compositions are likely derived from source con-tamination by crustal rocks with low RbSr SmNd and UPb ratios Pos-sible mechanisms for EMndashI origin include origin from ancient lithosphereinteraction with pelagicndash terrigenous sediments andor digestion of sub-ducted plagioclase-rich large igneous provinces or lower crustal lithologies

d The EMndashI basalts occur in oceanic and continental settings Proposed mod-els suggest that the SCLM likely represents the locus where source contam-ination andor metasomatism occur due to the effects of lithospheric keeldelamination within the asthenospheric depleted mantle

d The EMndashI end-member although unique is unlikely to represent a singlecomponent The marked differences in trace element contents and radio-genic isotope compositions of EMndashI basalts cast for a heterogeneous reser-voir resulting from distinct processes (addiction of crustal material re-cycling of oceanic sediments) each of these petrogenetic processes is likelyto prevail over the others according to specific geodynamic settings

Acknowledgements

The authors wish to thank the official review of R Vollmer M L thanks EnricaMascia and Bianca Lustrino for their continue assistance during the writing of themanuscript This work is granted by the Italian Agency CNR (Agenzia 2000 and Pro-getto Coordinato Agenzia 2000 ML) and by the University of Rome La SapienzaldquoProgetto Giovani Ricercatorirdquo (ML)

References

Anderson A T Clayton R N amp Mayeda T K (1971) Oxygen isotope thermo-metry of mafic igneous rocks ndash J Geol 79 714 ndash729

Appora I Eiler J M Matthews A amp Stolper E M (2003) Experimental de-termination of oxygen isotope fractionations between CO2 vapor and soda ndashmeli-lite melt ndash Geochim Cosmochim Acta 67 459 ndash471


Baldridge W S Sharp Z D amp Reid K D (1996) Quartz-bearing basalts Oxy-gen isotopic evidence for crustal contamination of continental mafic rocks ndash Geo-chim Cosmochim Acta 60 4765 ndash4772

Borisova A Yu Belyatsky B V Portnyagin M V amp Sushchevskaya N M(2001) Petrogenesis of olivine ndashphyric basalts from the Aphanasey Nikitin Riseevidence for contamination by cratonic lower continental crust ndash J Petrol 42277ndash319

Carlson R W (1995) Isotopic inferences on the chemical structure of the Mantle ndashJ Geodyn 20 365 ndash 386

Chung S L Jahn B M Chen S J Lee T amp Chen C H (1995) Miocene ba-salts in northwestern Taiwan evidence for EM-type mantle sources in the conti-nental lithosphere ndash Geochim Cosmochim Acta 59 549 ndash555

Dostal J Cousens B amp Dupuy C (1998) The incompatible element characteris-tics of an ancient subducted sedimentary component in ocean island basalts fromFrench Polynesia ndash J Petrol 39 937ndash 952

Eiler J M Farley K A Valley J W Hauri E Craig H Hart S R ampStolper E M (1997) Oxygen isotope variations in ocean island basalt pheno-crysts ndash Geochim Cosmochim Acta 61 2281ndash 2293

Eiler J M Farley K A Valley J W Stolper E M Hauri E H amp CraigH (1995) Oxygen isotope evidence against bulk recycled sediment in the mantlesources of Pitcairn Island lavas ndash Nature 377 138ndash141

Eiler J M Schiano P Kitchen N amp Stolper E M (2000) Oxygen-isotopeevidence for recycled crust in the sources of mid-ocean-ridge basalts ndash Nature403 530 ndash534

Eisele J Sharma M Galer S J G Blichert-Toft J Devey C W amp Hof-mann A W (2002) The role of sediment recycling in EM ndashI inferred from OsPb Hf Nd Sr isotope and trace element systematics of the Pitcairn hotspot ndashEarth Planet Sci Lett 196 197ndash 212

Esser B K amp Turekian K K (1993) The Osmium isotopic composition of thecontinental crust ndash Geochim Cosmochim Acta 57 3093 ndash3104

Farley K A amp Craig H (1992) Mantle plumes and mantle sources ndash Science 258821ndash822

Farmer G L (1992) Magmas as tracers of lower crustal composition an isotopic ap-proach ndash In Fountain D M Arculus R amp Kay R W (eds) Continentallower crust Elsevier 363 ndash390

Frey F A Weis D Borisova Y A amp Xu G (2002) Involvement of continentalcrust in the formation of the Cretaceous Kerguelen plateau new perspectives fromODP Leg 120 Sites ndash J Petrol 43 1207ndash1239

Gao S Luo T C Zhang B R Hang H F Han Y W Zhao Z D amp Hu Y K(1998) Chemical composition of the continental crust as revealed by studies inEast China ndash Geochim Cosmochim Acta 62 1959 ndash1975

Gasperini D Blichert -Toft J Bosch D Del Moro A Macera P TeacuteloukP amp Albaregravede F (2000) Evidence from Sardinian basalt geochemistry for re-cycling of plume heads into the earthrsquos mantle ndash Nature 408 701ndash704


Gregory R T amp Taylor H P (1981) An oxygen isotope profile in a section of Cre-taceous oceanic crust Samail ophiolite Oman evidence for 18O buffering of theoceans by deep (gt5 km) seawater-hydrothermal circulation at mid-ocean ridges ndashJ Geophys Res 86 2737ndash 2755

Hanan B B amp Graham D W (1996) Lead and helium isotope evidence from ocea-nic basalts for a common deep source of mantle plumes ndash Science 272 991ndash 995

Hanyu T Dunai T J Davies G R Kaneoka I Nohda S amp Uto K (2001)Noble gas study of the Reunion hotspot evidence for distinct less-degassed mantlesources ndash Earth Planet Sci Lett 193 83ndash98

Hanyu T Kaneoka I amp Nagao K (1999) Noble gas study of HIMU and EMocean island basalts in the Polynesian region ndash Geochim Cosmochim Acta 631181ndash1201

Harmon R S amp Hoefs J (1995) Oxygen isotope heterogeneity of the mantle de-duced from global 18O systematics of basalts from different geotectonic settings ndashContrib Miner Petrol 120 95ndash114

Hart S R (1984) A large-scale isotope anomaly in the southern hemisphere mantlendash Nature 309 753 ndash757

Hart S amp Zindler A (1989) Constraints on the nature and development of chemi-cal heterogeneities in the mantle ndash In Peltier W R (ed) Mantle convectionplate tectonics and global dynamics ndash Gordon and Breach Sci Publ 261ndash 387

Hart S R Gerlach D C amp White W M (1986) A possible new SrndashNdndashPbmantle array and consequences for mantle mixing ndash Geochim Cosmochim Acta50 1551ndash1557

Hart S R Hauri E H Oschmann L A amp Whitehead J A (1992) Mantle plu-mes and entrainment isotopic evidence ndash Science 256 517ndash 520

Heaman L M (1989) The nature of the subcontinental mantle from SrndashNdndashPb iso-topic studies on kimberlitic perovskite ndash Earth Planet Sci Lett 92 323 ndash 334

Hemond C Devey C W amp Chauvel C (1994) Source compositions and meltingprocesses in the Society and Austral plumes (South Pacific Ocean) element andisotope (Sr Nd Pb Th) geochemistry ndash Chem Geol 115 7ndash 45

Hiyagon H Ozima M Marty B Zashu S amp Sakai H (1992) Noble gases insubmarine glasses from mid oceanic ridges and Loihi seamount constraints onearly history of the Earth ndash Geochim Cosmochim Acta 56 1301ndash1316

Hilton D R Hammerschmidt K Loock G amp Friedrichsen H (1993) Heliumand argon isotope systematics of the central Lau basin and Valu Fa ridge evidenceof crustmantle interactions in a back-arc basin ndash Geochim Cosmochim Acta 572819 ndash2841

Hofmann A W (1997) Mantle geochemistry the message from oceanic volcanism ndashNature 385 219 ndash229

Honda M McDougall I Patterson D B Doulgeris A amp Clague D A(1991) Possible solar noble-gas component in Hawaiian basalts ndash Nature 349149ndash151

ndash ndash ndash ndash ndash (1993) Noble gases in submarine pillow basalt glasses from Loihi andKilauea Hawaii a solar component in the Earth ndash Geochim Cosmochim Acta57 859 ndash 874


Kurz M D Jenkins W J Schilling J G amp Hart S R (1982) Helium isotopicvariations in the mantle beneath the central North Atlantic Ocean ndash Earth PlanetSci Lett 58 1ndash14

Ito E White W M amp Gopel C (1987) The O Sr Nd and Pb isotope geochem-istry of MORB ndash Chem Geol 62 157ndash176

James D E (1981) The combined use of oxygen and radiogenic isotopes as indicatorsof crustal contamination ndash Ann Rev Earth Planet Sci 9 311ndash 344

Liu Y S Gao S Jin S Y Hu S H Sun M Zhao Z B amp Feng J L (2001)Geochemistry of lower crustal xenoliths from Neogene Hannouba basalt northChina craton implication for petrogenesis and lower crustal composition ndash Geo-chim Cosmochim Acta 65 2589 ndash2604

Lustrino M (2000) Volcanic activity during the Neogene to Present evolution of theWestern Mediterranean area a review ndash Ofioliti 25 87ndash101

Lustrino M Melluso L amp Morra V (2000) The role of lower continental crustand lithospheric mantle in the genesis of PliondashPleistocene volcanic rocks fromSardinia (Italy) ndash Earth Planet Sci Lett 180 259 ndash270

ndash ndash ndash (2002) The transition from alkaline to tholeiitic magmas a case study fromOrosei ndashDorgali Pliocene volcanic district (NE Sardinia Italy) ndash Lithos 63 73ndash113

Mahoney J J White W M Upton B G J Neal C R amp Scrutton R A(1996) Beyond EMndash1 lavas from AfanasyndashNikitin Rise and the Crozet Archipel-ago Indian Ocean ndash Geology 24 615 ndash618

Matsumoto T Honda M McDougall I Yatsevich I amp OrsquoReilly S Y(1997) Plume-like neon in a metasomatic apatite from the Australian lithosphericmantle ndash Nature 388 162ndash164

Matthews A Stolper E M Eiler J M amp Epstein S (1998) Oxygen isotopefractionation among melts minerals and rocks ndash Miner Mag 62 1021ndash1091

Mattey D Lowry D amp Macpherson C (1994) Oxygen isotope composition ofmantle peridotite ndash Earth Planet Sci Lett 128 231ndash 241

Milner S C amp Le Roex A P (1996) Isotope characteristics of the Okenyenya ig-neous complex northwestern Namibia constraints on the composition of the earlyTristan plume and the origin of the EM 1 mantle component Earth Planet ndash SciLett 141 277ndash291

Niedermann S Bach W amp Erzinger J (1997) Noble gas evidence for a lowermantle component in MORBs from the Southern East Pacific Rise decoupling ofHelium and Neon isotope systematics ndash Geochim Cosmochim Acta 61 2697ndash2715

Ozima M amp Igarashi G (2000) The primordial noble gases in the Earth a keyconstraint on Earth evolution models ndash Earth Planet Sci Lett 176 219 ndash232

Ozima M amp Podosek F A (2002) Noble gas geochemistry ndash 2nd ed CambridgeUniv Press


Peate D W Hawkesworth C J Mantovani M M S Rogers N W amp Tur-ner S P (1999) Petrogenesis and stratigraphy of the high-TiU Urubici magmatype in the Paranagrave flood basalt Province and implications for the nature of lsquoDup-alrsquo-type mantle in the South Atlantic region ndash J Petrol 40 451ndash 473

Reisberg L Zindler A Marcantonio F White W Wyman D amp Weaver B(1993) Os isotope systematics in ocean island basalts ndash Earth Planet Sci Lett120 149 ndash167

Richardson S H Erlank A J Reid D L amp Duncan A R (1984) Major andtrace elements and Nd and Sr isotope geochemistry of basalts from the Deep SeaDrilling Project Leg 74 Walvis Ridge transect ndash In Moore T C Jr Rabino-wits P D et al (eds) Init Repts DSDP 74 739 ndash754

Rudnick R L amp Fountain D M (1995) Nature and composition of the continentalcrust a lower crustal perspective ndash Rev Geophys 33 267ndash309

Rudnick R L amp Taylor S R (1987) The composition and petrogenesis of thelower crust a xenolith study ndash J Geophys Res 92 13981ndash14005

Smith A D (1998) The geodynamic significance of the Dupal anomaly in Asia ndash InMantle dynamics and plate interactions in east Asia geodynamics ndash AGU Mono-graph 27 89ndash105

Stein M amp Hofmann A W (1994) Mantle plumes and episodic crustal growth ndashNature 372 63ndash68

Sun S S amp McDonough W F (1989) Chemical and isotopic systematics of ocea-nic basalts implications for mantle compositions and processes ndash In SaundersA D amp Norry M J (eds) Magmatism in the ocean basins ndash Geol Soc LongSpec Publ 42 313 ndash345

Tatsumi Y (2000) Continental crust formation by crustal delamination in subductionzones and complementary accumulation of the Enriched Mantle I component inthe mantle ndash Geochem Geophys Geosyst

Vervoort J D Patchett J Albaregravede F Blichert-Toft J Rudnick R ampDownes H (2000) HfndashNd isotopic evolution of the lower crust ndash Earth PlanetSci Lett 181 115 ndash129

Weaver B L (1991) The origin of ocean island basalt end-member compositionstrace element and isotopic constraints ndash Earth Planet Sci Lett 104 381ndash 397

Wedepohl K H (1995) The composition of the continental crust ndash Geochim Cos-mochim Acta 59 1217ndash1232

Wolf M B amp Wyllie P J (1994) Dehydration-melting of amphibolite at 10 kbarthe effects of temperature and time ndash Contrib Miner Petrol 115 369 ndash 383

Woodhead J D amp Devey C W (1993) Geochemistry of the Pitcairn seamounts Isource character and temporal trends ndash Earth Planet Sci Lett 116 81ndash 99

Woodhead J D Greenwood P Harmon R S amp Stoffers P (1993) Oxygenisotope evidence for recycled crust in the source of EM-type ocean island basaltsndash Nature 362 809 ndash 813

Zindler A amp Hart S (1986) Chemical geodynamics ndash Ann Rev Earth Planet Sci14 493 ndash 571


Zindler A Jagoutz E amp Goldstein S (1982) Nd Sr and Pb isotopic systema-tics in a three-component mantle a new perspective ndash Nature 298 519 ndash523

Received August 2 2002 accepted February 17 2003

Authorsrsquo addresses

Michele Lustrino Dipartimento di Scienze della Terra Universitagrave degli Studi diRoma La Sapienza P le A Moro 5 00185 Rome Italy CNR ndash Istituto di GeologiaAmbientale e Geoingegneria Universitagrave degli Studi di Roma La Sapienza P le AMoro 5 00185 Rome Italy Corresponding authore-mail michelelustrinouniroma1it

Luigi Dallai CNR ndash Istituto di Geologia Ambientale e Geoingegneria Universitagravedegli Studi di Roma La Sapienza P le A Moro 5 00185 Rome Italy


Fig 1 SrndashNdndash Pb isotopic systematics of EMndashI basalts from continental and oceanicsettings Pitcairn islands and seamounts (Pacific Ocean Woodhead amp Devey 1993Eiler et al 1995 Eisele et al 2002) AphanasyndashNikitin Rise (Indian Ocean Maho-ney et al 1996 Borisova et al 2001) Urubici magma-type (Paranagrave Brazil Peate etal 1999) Walvis Ridge (Atlantic Ocean Richardson et al 1984) Kerguelen plateau(Indian Ocean Frey et al 2002) Sardinia (Italy Gasperini et al 2000 Lustrino etal 2000 2002) DMM = Depleted Morb Mantle HIMU = High m EMndashI and EMndashII= Enriched Mantle I and II NHRL = Northern Hemisphere Reference Line Hart1984) Contour of Pb isotopic field of DMM is a broken line because of the extremevariability of Atlantic Pacific and Indian MORBs in terms of 206Pb204Pb 207Pb204Pband 208Pb204Pb








Proposed models
















Fig 2









Concluding remarks






Acknowledgements


References


















































































Proposed models
















Fig 2









Concluding remarks






Acknowledgements


References
























































































Fig 2









Concluding remarks






Acknowledgements


References


















































































Concluding remarks






Acknowledgements


References











































































on the origin of em-i end-member - do plumes exist? · on the origin of em–i end-member ... os,...

Documents