The Amount of Recycled Crust in Sources of Mantle-Derived ......The Amount of Recycled Crust in...

www.sciencemag.org SCIENCE VOL 316 20 APRIL 2007

Published 29 March 2007 on Science Express DOI: 10.1126/science.1138113

The Amount of Recycled Crust in Sources of Mantle-Derived Melts

Alexander V. Sobolev,* Albrecht W. Hofmann, Dmitry V. Kuzmin, Gregory M. Yaxley, Nicholas T. Arndt, Sun-Lin Chung, Leonid V. Danyushevsky, Tim Elliott, Frederick A. Frey, Michael O. Garcia, Andrey A. Gurenko, Vadim S. Kamenetsky, Andrew C. Kerr, Nadezhda A. Krivolutskaya, Vladimir V. Matvienkov, Igor K. Nikogosian, Alexander

Rocholl, Ingvar A. Sigurdsson, Nadezhda M. Sushchevskaya, Mengist Teklay

*To whom correspondence should be addressed. E-mail: [email protected]

RESEARCHARTICLES

The Amount of Recycled Crust inSources of Mantle-Derived MeltsAlexander V. Sobolev,1,2* Albrecht W. Hofmann,1 Dmitry V. Kuzmin,1,3 Gregory M. Yaxley,4Nicholas T. Arndt,5 Sun-Lin Chung,6 Leonid V. Danyushevsky,7 Tim Elliott,8 Frederick A. Frey,9Michael O. Garcia,10 Andrey A. Gurenko,1 Vadim S. Kamenetsky,7 Andrew C. Kerr,11Nadezhda A. Krivolutskaya,2 Vladimir V. Matvienkov,12 Igor K. Nikogosian,13,14Alexander Rocholl,15 Ingvar A. Sigurdsson,16 Nadezhda M. Sushchevskaya,2 Mengist Teklay17

Plate tectonic processes introduce basaltic crust (as eclogite) into the peridotitic mantle. Theproportions of these two sources in mantle melts are poorly understood. Silica-rich melts formedfrom eclogite react with peridotite, converting it to olivine-free pyroxenite. Partial melts of thishybrid pyroxenite are higher in nickel and silicon but poorer in manganese, calcium, andmagnesium than melts of peridotite. Olivine phenocrysts’ compositions record these differencesand were used to quantify the contributions of pyroxenite-derived melts in mid-ocean ridge basalts(10 to 30%), ocean island and continental basalts (many >60%), and komatiites (20 to 30%).These results imply involvement of 2 to 20% (up to 28%) of recycled crust in mantle melting.

It is widely accepted that the heterogeneity ofthe convecting mantle observed in thecomposition of mantle-derived magmas is

largely due to subduction and recycling ofoceanic crust into the deep mantle (1, 2). Tounderstand the role of crustal material in creatingcompositional heterogeneities in the mantle andto evaluate the geodynamical consequences ofthis contribution, one must quantify the crustal

input to the mantle sources of common, mantle-derived magmas in mid-oceanic ridges basalts(MORBs), ocean islands (OIBs), and largeigneous provinces (LIPs). It is not possible to useincompatible element abundances in basalts toconstrain the proportion of recycled component inthe magma source because concentrations of theseelements are also sensitive to the extent of melt-ing. Similarly, the use of isotope ratios for mak-ing such quantitative estimates is compromisedby the isotopic variability of subducted materialsinvolved in the recycling process (2). We used analternative approach based on a combination ofmajor elements and compatible trace elements inparental melts, because these are more uniform inthe mantle and are strongly controlled by theresidual phases in equilibrium with partial melts(3–5).

Our method has its basis in the experimentaland theoretical prediction that high-pressure (P >3.0 GPa) melting of typical recycled oceaniccrust (in the form of eclogite with a separate SiO2

phase) and reaction of this melt with peridotiteproduces olivine-free pyroxenite (5). We showthat further melting of this hybrid lithology in theabsence of residual olivine is more voluminousthan the melting of peridotite (at a given pressureand temperature) and that pyroxenite-derivedmelts are characteristically enriched in Si andNi but depleted in Mg, Ca, and Mn comparedwith their peridotite-derived counterparts. Thisdifference arises because olivine principallycontrols the composition of melt produced inperidotite, whereas pyroxene mainly controls thecomposition of melt from olivine-free hybridpyroxenite (5–8). Experimental data predict (9)that, as such pyroxenite-derivedmelts rise towardthe surface, the decrease in pressure causes theirsaturation in olivine. This olivine is unusually Nirich and Mn and Ca poor. With use of a new,large data set of high-precision analyses of

olivine phenocrysts from OIBs, LIPs, MORBs,and komatiites, we show that hybrid pyroxeniteis a common lithology in upwelling mantle and amajor contributor to tholeiitic (silica-saturated)and transitional (moderately silica-undersaturated)magmas of OIBs and LIPs emplaced on thickoceanic or continental lithosphere.

Olivine data set. We use olivine phenocrystsas probes of parental melt composition, becauseolivine is the first phase to precipitate at lowpressures in almost all mantle-derived magmasand because its forsterite content is an excellentmeasure of the degree of fractional crystallizationallowing reconstruction of the parental meltcomposition.

Olivine phenocrysts were analyzed by elec-tron microprobe using high probe currents andlong counting times (10). This procedure rou-tinely yields detection limits of around 6 to 15parts per million (ppm) and errors (2 standarderrors) of 15 to 30 ppm for trace elements (Ni,Ca, Mn, Cr, Co, and Al) and 0.01 mole percent(mol %) for forsterite content [defined as Fo =Mg/(Mg + Fe)], checked by repeated analysis ofSan Carlos olivine standard (11). In the followingdiagrams we use only high-precision data.

We have analyzed nearly 17,000 grains ofolivine phenocrysts from 229 samples of tholei-itic to transitional compositions coveringMORBs(40 samples) from Mid-Atlantic Ridge, EastPacific Rise, South-East Indian Ridge, andKnipovich Ridge; OIBs (138 samples) fromHawaiian Islands and Emperor Seamounts,Canary Islands, Reunion, Azores, and Iceland;LIPs (36 samples) from Ontong Java Plateau,Siberia, Emeishan, Karoo, Afar, and NorthAtlantic Province; komatiites and associatedpicrites (15 samples) from the Archean Abitibigreenstone belt in Canada and the Belingwe beltin Zimbabwe and South Africa; Proterozoickomatiitic basalts from Gilmour Island, Canada;and komatiites and picrites from Gorgona Island,Colombia. Most samples are picrites or olivinebasalts containing large amounts of fresh, high-magnesium olivine phenocrysts. The samples aresubdivided into four groups: (i) MORB; (ii)within plate magmas (WPM, magmas eruptedfar from plate boundaries) forming OIBemplaced over thin lithosphere (<70 km thick),WPM-THIN; (iii) WPM (OIB and LIP)emplaced over thick lithosphere (>70 km thick),WPM-THICK; and (iv) komatiites and asso-ciated magmas, KOMATIITES. Details of sam-ple locations, references for sample descriptions,and their group correspondence are presented intable S2a.

The most-magnesian olivine compositions(defined by olivines phenocrysts with Fo within1 mol % from amaximum Fo) for each specimenwere averaged (table S2a) for the plots shown inFig. 1. Individual olivine analyses are presentedon fig. S4 and tables S2, c to f.

In addition to Mn and Ni concentrations,which strongly correlate with Fo (Fig. 1, A and

RESEARCHARTICLES

1Max Planck Institute (MPI) for Chemistry, Post Office Box3060, 55020 Mainz, Germany. 2Vernadsky Institute ofGeochemistry and Analytical Chemistry, Russian Academyof Sciences, Kosygin Street 19, 119991 Moscow, Russia.3Institute of Geology and Mineralogy, Siberian Branch ofRussian Academy of Sciences, Koptuyga prospekt 3,630090 Novosibirsk, Russia. 4Research School of EarthSciences, Australian National University, Canberra, ACT0200 Australia. 5Laboratoire de Géodynamique desChaînes Alpines, Université de Grenoble, 38401 Grenoblecedex, France. 6Department of Geosciences, National TaiwanUniversity, Post Office Box 13-318, Taipei 106, Taiwan.7Australian Research Council, Centre of Excellence in OreDeposits and School of Earth Sciences, University ofTasmania, Private Bag 79, Hobart, Tasmania, 7001, Aus-tralia. 8Department of Earth Sciences, Queen’s Road, WillsMemorial Building, University of Bristol, Bristol BS8 1RJ,UK. 9Department of Earth, Atmospheric, and PlanetaryScience, Massachusetts Institute of Technology (MIT), 77Massachusetts Avenue, Cambridge, MA 02139, USA.10Department of Geology and Geophysics, University ofHawaii, 1680 East-West Road, Honolulu, HI 96822, USA.11School of Earth, Ocean and Planetary Sciences, CardiffUniversity, Main Building, Park Place, Cardiff CF10 3YE,UK. 12P. P. Shirshov Institute of Oceanology of RussianAcademy of Sciences, Nakhimovsky prospekt 36, 117997Moscow, Russia. 13Department of Petrology, Faculty ofGeosciences Utrecht University, Budapestlaan 4, Utrecht,Netherlands. 14Department of Petrology, Faculty of Earthand Life Sciences, Vrije Universiteit, De Boelelaan 1085,1081 HV Amsterdam, Netherlands. 15Department of Earthand Environmental Sciences, University Munich, 80333Munich, Germany. 16South Iceland Nature Centre, Strandvegur50, Vestmannaeyjar, IS 900, Iceland. 17Department of EarthScience, University of Asmara, Asmara, Eritrea.

*To whom correspondence should be addressed. E-mail:[email protected]

20 APRIL 2007 VOL 316 SCIENCE www.sciencemag.org412

C), we also plot Mn/Fe and Ni versus Mg/Feratios (Fig. 1, B and D). These ratios do not varysignificantly with olivine fractionation (seemodel curves Frac 1 and Frac 2) but neverthelessrange considerably (Fig. 1, B and D). Mostolivine phenocrysts from MORBs and manyfrom komatiites have Mn and Ni contents similarto those of peridotite-derived melts. In contrast,most olivines from the WPM-THICK group aresignificantly depleted in Mn and enriched in Ni.Their concentrations are not compatible with themelting products of common peridotites. The ol-ivines from the WPM-THIN group have inter-mediate Mn and Ni contents.

Concentrations of Ca also provide somediscrimination in spite of the greater overlap.Most olivines from the WPM-THICK group aretoo low in Ca to have precipitated from peridotite-derived melts (shown as experimental-basedmodel compositions and fractionation trajecto-ries, Fig. 1E).

Chromium is strongly controlled by garnetand spinel in peridotites and thus might be usefulto decipher products of high-degree melting of

peridotite, which leave residuals (restites) free ofCr-rich phases (12). Olivines from Archeankomatiites have the highest Cr values and matchcompositions of olivines from a spinel- andgarnet-free refractory restite (Fig. 1F). Theycould, therefore, be derived directly from high-degree melting of peridotite. In the other groupsof olivines, Cr is markedly lower than expectedin equilibrium with peridotite at high pressures(see experimental data on lherzolite melting, Fig.1F). The lowest Cr contents are found in MORBolivines, indicative of residual Cr spinel.

Cobalt (Fig. 2A) shows nearly uniform cor-relation with Fo for all rock groups, with possiblyonly minor (around 5%) relative enrichment inWPM-THICK and WPM-THIN over MORB(estimated from group average Co/Fe of tableS2a). Decoupling of Co andNi yieldsNi/Co ratiosof many WPM-THICK olivines that are un-usually high for the Earth mantle (Fig. 2B).

Mn/Fe is the parameter least dependent onolivine fractionation (as shown by the modelfractionation curves in Fig. 3). Thus, it is diag-nostic of parental magma compositional differ-

ences. There is a significant negative correlationof Ni/Mg versus Mn/Fe (linear correlation coef-ficient r is 0.66 for 238 samples) in spite of strongdependence of Ni/Mg on the degree of olivinefractionation (see fractionation trajectories inFig. 3A). This correlation improves (r = 0.88 for103 samples) for the subset of olivines with anarrower range of Fo contents (Fo89 to Fo91).MORB olivines are the lowest in Ni and highestinMn, whereas olivines from theWPM-THICKgroup are the highest in Ni and lowest in Mn,with olivines from theWPM-THIN group beingintermediate.

To minimize the effects of olivine fractiona-tion, we show parameters Ni/(Fe/Mg) and Ca/Fein Fig. 3, B and C. This procedure also reducesthe scatter in the ordinate significantly, thushighlighting the differences between geodynamicsettings.

Fate of recycled oceanic crust. In subductionat P > 2.5 GPa, the basaltic and gabbroic portionsof the oceanic crust are transformed completelyto eclogite (clinopyroxene and garnet) with a freeSiO2 phase (13–15). Unless silica has been

Fig. 1. (A to F) Average compositionsof the most highly magnesian olivinephenocrysts in each sample. Concen-trations and their ratios are given inppm versus forsterite content of olivinein mol %. Olivine group names are asdefined in text. PM, Herzberg indicatescompositions of olivine in equilibriumat 0.1 MPa with melt originally gen-erated at 3.0 to 5.0 GPa from fertileperidotite (12), calculated by Petrologsoftware (41) for oxygen fugacity corre-sponding to quartz-fayalite-magnetite(QFM) buffer using the Herzberg model(4), PM, Kinzler, olivine compositionssimilar to PM, Herzberg but with Nicalculated by using Ni partitioning be-tween olivine and melt from Kinzleret al. (28). Frac 1 is the trend of olivinecomposition during fractional crystalli-zation from a melt derived from fertileperidotite at 3 GPa and 1515°C (12).Fractionation of olivine modeled up to20% for oxygen fugacity correspond-ing to QFM buffer using the Herzbergmodel (4). Frac 2 is similar to Frac 1 butcalculated for 35% crystallization ofmelt derived at 4.0 GPa and 1630°C(12). Green ellipse indicates field ofolivine compositions compatible withperidotitic source. In (F), HRZ, Herzbergstands for calculated compositions ofolivine from spinel- and garnet-freeharzburgite restite using (4); LHRZ,Walter and HRZ, Walter indicate exper-imental olivines from lherzolite- andgarnet-free harzburgite residual assem-blages, respectively, produced by high-pressuremelting of fertile peridotite (12).Black ellipse marked “ol + low Ca-Px”indicates field of olivine compositions from refractory garnet- and spinel-free assemblage of olivine and low-Ca pyroxene.

800

1000

1200

1400

1600

86 87 88 89 90 91 92 93 94

Mn

MORB

WPM-THIN

WPM-THICK

KOMATIITES

Frac-1

Frac-2

PM, Herzberg

A

1.1

1.2

1.3

1.4

1.5

1.6

1.7

1.8

1.9

86 87 88 89 90 91 92 93 94

100M

n/F

e

B

1000

1500

2000

2500

3000

3500

4000

4500

5000

86 87 88 89 90 91 92 93 94

Ni

PM, Kinzler

C

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

86 87 88 89 90 91 92 93 94

Ni/(

Mg

/Fe)

/100

0

D

0

500

1000

1500

2000

86 87 88 89 90 91 92 93 94

Fo mol%

Cr

Walter HRZ

Walter LRZ

Herzberg HRZ

OL+Low-Ca Px

F

1000

1500

2000

2500

3000

86 87 88 89 90 91 92 93 94

Fo mol%

Ca

E

www.sciencemag.org SCIENCE VOL 316 20 APRIL 2007 413

RESEARCH ARTICLES

removed during subduction (16), this combina-tion will also be the relevant assemblage duringrecycling to the upper mantle (17).

In the ascending mantle (e.g., a mantle plumeor upstream flow of convecting mantle), thesilica-oversaturated eclogite starts melting athigher pressures than the peridotite and produceshigh silica melt (18, 19). This melt reacts witholivine from peridotite, producing pyroxenes andgarnet (5, 8, 19). Previous studies have envi-sioned that this reaction creates a refertilizedperidotite enriched in pyroxene (19, 20). Thisconclusion would predict variable mixingproportions of individual ingredients (eclogite-derived high-Si melt and peridotite) that are

drastically different in composition. Melting suchvariable source compositionswould create highlynonlinear correlations of 187Os/188Os and 87Sr/86Srisotope ratios in the melts, and this is contradictedby the strongly linear correlations observed inHawaiian basalts (21, 22), which are thought tohave a significant eclogite component (3, 5).

However, it has been shown experimentally(23) and proposed on the basis of Korzinskii’stheory (24) that, under conditions of local equilib-rium, the reaction between high-Si eclogite-derivedmelt and peridotite produces an olivine-free lithol-ogy enriched in pyroxene (5). This fundamentallydiffers from a partial reaction (19, 20) because itleads to a stable pyroxenite lithology (hybrid py-

roxenite) generated by roughly fixed proportionsof high-Si melt and peridotite [constrained byreaction stoichiometry between 40 and 60weight % (wt %) of melt (5)] irrespective of theinitial proportions of the reaction ingredients.Consequently, the hybrid pyroxenite has nearlyuniform chemical and isotopic composition,thus constituting a single mixing endmember.Binary mixing of melts derived from peridotiteand this pyroxenite leads to near-linear 87Sr/86Srversus 87Os/188Os trends (5).

Other predicted geochemical consequencesof replacement of olivine by pyroxene are a sig-nificant decrease of the bulk distribution co-efficient between crystals and melt (Kd) for Ni

Fig. 2. (A and B) Cobaltand nickel to cobalt ratioversus Fo of average Mg-rich olivine phenocrysts.Pink band at Ni/Co 20 ±1 represents estimatedvalues for bulk silicateEarth (BSE), core, andchondrites (39). Arrowsindicate trend of olivinecompositions due to themantle melting (melting)and magma crystalliza-tion (cryst). All other sym-bols are as on Fig. 1.

100

110

120

130

140

150

160

170

180

190

200

86 87 88 89 90 91 92 93

Co

pp

m

A

10

15

20

25

30

35

86 87 88 89 90 91 92 93

Fo mol%Fo mol%

Ni/C

o

BSE, Core, Chondrites

B

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

2.2

2.4

2.6

1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9

Ni/(

Mg

/Fe)

/100

0

PYR, HERZBERG

PYR, KINZLER

MIX, HERZBERG

MIX, KINZLER

PYROXENITE

PERIDOTITE

B

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

2.2

2.4

2.6

1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9

100N

i/Mg PYROXENITE

PERIDOTITE

A

0

1

2

3

4

5

1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9

100Mn/Fe100Mn/Fe

100Mn/Fe

100C

a/F

e PYROXENITE

PERIDOTITE

C

C

Fig. 3. (A to C) Composition of average Mg-rich olivine phenocrysts andequilibrium olivines from peridotite- and pyroxenite-derived melts.Elements and their ratios are shown in ppm. PYR, Herzberg and PYR,Kinzler stand for compositions of olivine equilibrium at 0.1 MPa withmelt generated at 3.5 GPa by melting of pyroxenite (table S3) calculatedfor oxygen fugacity corresponding to QFM buffer using model (4) for Fe-Mgpartitions and models (4, 28) for Ni partition, respectively. MIX Herzberg andMIX Kinzler correspond to equilibrium olivine compositions calculated for themixture of pyroxenite-derived and peridotite-derived melts using model (4)for Fe-Mg partitions and models (4, 28) for Ni partition, respectively. Blackellipse indicates olivine compositions corresponding to pyroxenitic source. Allother symbols are as on Fig. 1.


RESEARCH ARTICLES

(5, 6) and a decrease in the ratio of the bulkcoefficients of Mn and Fe (7). These featuresoccur because olivine is the major silicate phasein peridotite concentrating Ni and the only sili-cate phase in peridotite having Kd for Fe greaterthan Kd for Mn (7). The bulk Kd changes willincreaseNi and lower theMn/Fe ratio of pyroxenite-derived melt compared with peridotite-derivedmelts. In addition, melting of pyroxenite yieldslower Ca compared with peridotite (8). Addi-tional predicted differences are higher meltfractions for hybrid pyroxenite than peridotiteand higher Si and lower Mg in pyroxenite-derived melt (5, 25, 26).

These predictions were tested by experimen-tal melting of a model hybrid pyroxenite (5).Experiments were run at P = 3.5 GPa andtemperatures between 1400° and 1570°C in aconventional 1.27-cm piston-cylinder apparatusat the Australian National University (10, 19).These results, together with published experi-mental data for melting of peridotite (12),confirm the Ni and Mn relationships as well asmelting rates predicted above (table S3 and fig.S5). From these data, we calculate equilibriumolivine compositions at low pressure (4, 27, 28)in order to compare them with the naturalphenocryst data (Fig. 3). We used these resultsto estimate mixing proportions of melts derivedfrom the two end-member sources for the olivinedata sets representing different geodynamicsettings. The end-member melt compositionswere calculated from averaging experimentaldata on melting of pyroxenite and peridotite (10).

Quantitative estimates. We assumed mix-tures (in 10% intervals) of the end-member melts(10) and calculated the composition of equilibri-um olivines. The calculated mixing trajectoriesfor the two different models for Ni partitioning

are consistent with natural olivine data (Fig. 3B).The relation between Mn/Fe of modeled olivinesand mixing proportions (10) was used to com-pute the amount of pyroxenite-derived compo-nent for individual samples (Fig. 4). Olivinesfrom the WPM-THICK group of basalts yield anaverage of 61 ± 16% (standard deviation)pyroxenite-derived component, similar to resultsderived from Ni contents in Hawaiian meltinclusions and olivines only (5). The olivinesfrom some continental LIPs (specific suits fromSiberia and Karoo) indicate almost pure pyrox-enitic sources. Corresponding results for theother groups are for WPM-THIN, 30 ± 13%;for Archean komatiites, 21 ± 10%; and forMORB, excluding one unusual sample from theSouthern Atlantic (see below), 17 ± 12% [similarto predictions of (25)]. Because of the uncertain-ties involved in estimating the end-membercompositions, the differences between groupsare better constrained than the absolute numbers.Although MORBs contain the lowest proportionof pyroxenite-derived melt, the spread of MORBdata is significant, and many samples do containsubstantial amounts of pyroxenite-derived com-ponent [the extreme is the enriched in silicaMORB sample from the Southern Atlantic (29)with 100% pyroxenite-derived component]. Thecalculations show that the Archean komatiitescontain a significant amount of pyroxeniticcomponent (maximum of 30% for samples fromCanada and Belingwe), although the largestamount is in Proterozoic komatiitic basalts fromGilmour Island, Canada (up to almost 40%).From these calculations, an estimate of theamount of recycled oceanic crust (10) yields4% for MORB, 11% for WPM-THIN group,16% for WPM-THICK group, and around 13%for Archean komatiites. The highest estimate of

the amount of recycled oceanic crust (10) yieldOntong Java high-Mg lavas: 13% to 28%.

Silica-undersaturated basalts. Most of themagmas analyzed in this study are silica saturated(tholeiites or transitional). Only a few samples aremoderately silica-undersaturated alkali magmas(e.g., Azores and Afar). Our database is repre-sentative of the normal oceanic crust, several ofthe world’s major suites of flood basalts (LIPs),several of the major modern mantle plumes (30),and some komatiites. The strongly silica-undersaturated associations not covered hereinclude continental rift basalts, many smallerocean islands consisting mostly of alkali lavas,and also some larger-flux plumes (30) such asPitcairn, Tahiti, and Cape Verde islands. Why aresuch basalts that are highly enriched in in-compatible elements, and therefore presumablygenerated by very low degrees of melting, nearlyalways undersaturated in silica? This observationappears to contradict our model; one wouldexpect that silica-saturated melts generated fromhybrid pyroxenites should be prevalent, especial-ly at very low melt fractions. There are severalpossible explanations. (i) Avolatile (mostly CO2-)triggered melting of peridotite may be thedominant mechanism forming strongly silica-undersaturated alkaline magmas at temperatureslower than hybrid pyroxenite melts (31). (ii)Low-degree melts of silica-saturated eclogitemay be retained in the source because of theirhigh viscosity, thus preventing production of thehybrid pyroxenite (5). (iii) Melting of hybridpyroxenite at the contact with peridotite mayproduce low-degree, silica-undersaturated meltsat lower temperatures than melting of hybridpyroxenite itself (8). (iv) Melting of bimineraliceclogites (no free silica phase) formed fromsilica-undersaturated recycled crust producesundersaturated alkaline magmas (16).

What controls the amount of pyroxenite-derived melt? By following the method outlinedabove, we estimated the proportions of meltderived from pyroxenite and peridotite for eachparental magma. These proportions depend onseveral interrelated parameters, namely the thick-ness of lithosphere, the potential mantle temper-ature (Tp) (32), and the amount of recycled crustin the upwelling mantle (Fig. 5). Because at thesame Tp pyroxenite melts at higher pressure thanperidotite (26), a thick lithosphere (which im-poses a high lower limit on the depth of melting)will suppress low-depth peridotite melting andtherefore favor a high proportion of pyroxenite-derived melts (33, 34). The extreme case is foundin some continental flood basalts (specific suitesof Siberia and Karoo at table S2a) where theamount of pyroxenite-derived melt is nearly100%. In such a case, the amount of recycledmaterial cannot be estimated because the perido-titic component contributes no melt. In contrast, athin lithosphere (MORB, Iceland, Azores, andDetroit seamount) favors a higher proportion ofperidotite-derived melt because of the increasingdegree ofmelting of peridotite at shallower depths.

KOMATIITES

0

1

2

3

4

5

Fraction of pyroxenite derived meltFraction of pyroxenite derived melt

Fraction of pyroxenite derived melt Fraction of pyroxenite derived melt

Nu

mb

er o

f sa

mp

les

WPM-THICK

0

5

10

15

20

25

30

35

Nu

mb

er o

f sa

mp

les

WPM-THIN

0

1

2

3

4

5

6

7

8

Nu

mb

er o

f sa

mp

les

MORB

0

2

4

6

8

10

12

0 0.1 0.2 0.3 0.6 0.7 0.8 0.9 10.2 0.3 0.4 0.5 0.4 0.5

0 0.1 0.2 0.3 0.6 0.7 0.8 0.9 10.4 0.50 0.1 0.2 0.3 0.6 0.7 0.8 0.9 10.4 0.5

0.6 0.7 0.8 0.9 1

Nu

mb

er o

f sa

mp

les

0.10

Fig. 4. Estimated amounts of pyroxenite-derived component in the parental melt for 229 samplesof four different groups.


RESEARCH ARTICLES

A high Tp is an important condition to main-tain sufficient buoyancy of mantle plumes or anyother upstream mantle flow, and this buoyancylimits the amount of dense eclogite that they cancarry (35). Also, a high Tp affects mostly theproportion of peridotite-derived melt becausefractional melting imposes a rather stringent up-per limit to further melting at high melt fractions(36). High melt fractions are restricted to theeclogite and hybrid pyroxenite assemblages (26)(fig. S5). The peridotite assemblage produceslower melt fractions than pyroxenite (fig. S5) oreclogite (19, 25) at any given temperature andpressure, and its actual extent ofmelting thereforedepends strongly on the specific Tp.

Lastly, why are the proportions of recycledcomponent lower beneath mid-ocean ridges thanin thick-lithosphere settings? We suggest severalexplanations: (i) Relatively low amount of denserecycled component inMORBs is limited by theirTp, which is too low to carry more. (ii) For sta-tistical reasons, plumes are more likely to en-counter more-common thick lithosphere thanless-common thin lithosphere and few plumesimpinge directly on ridges, so we are forced todeal with very-small-number statistics. (iii) De-troit seamount represents one case where a(Hawaiian) plume has encountered thin litho-sphere and where our results do indicate a highfraction of recycled crust, similar to those foundon the island of Hawaii on the thick lithosphere.

This amount is significantly higher than for Ice-land, probably reflecting the effect of a higher Tpof the Hawaiian plume (4). (iv) The surfaceexpression of a plume emplaced under thicklithosphere requires high Tp, which is necessaryfor carrying a significant amount of recycledcrust (35), allowing melts to form at higher pres-sures than for ordinary peridotite (5, 25, 26),and melting a peridotite at higher pressures [e.g.,komatiites (37)].

Heterogeneous versus homogeneousmantle.The model presented here assumes that therecycled crustal component was not fully mixedwith peridotite during subduction and mantleconvection and thus that the formation of theolivine-free hybrid lithology may take place. Onthe other hand, homogenization of crustal ma-terial within the peridotite mantle should create arange of ultramafic lithologies with variableamounts of olivine, similar to a model byKelemen et al. (20). Under these circumstances,the major-element contents of partial melts willcorrespond to the eutectic-like composition, buf-fered by the peridotite assemblage, whereas thecompatible trace elements (Ni and Mn) will becontrolled by the bulk partition coefficients ofthis assemblage and thus by the amount of ol-ivine and pyroxene present in the system. There-fore, the amount of recycled crust can still beestimated on the basis of these trace elements, buttheir abundances will no longer correlate with the

buffered major elements (Si, Ca, and Al). ForHawaiian basalts, such correlations with Ni arepresent (5), which requires a strongly heteroge-neous source.

Input from Earth’s core? The Ni excess inmantle olivines from Siberian LIP (38) and theelevated Fe/Mn ratios in Hawaiian lavas (7) havebeen explained by input from Earth’s core to thesources of mantle plumes. This suggestion isconsistent with 186Os/188Os ratios for someHawaiian and Gorgona lavas (39) but is contra-dicted by the fact that concentrations of highlysiderofile (Pt) and chalcophile (Cu) elementsreported for Hawaiian basalts are not affected bythis process (5). Our olivine data provide strongarguments against any notable core contribution toNi or Fe excess in the sources of mantle-derivedmagmas. Cobalt does not show significant excessin olivines (Fig. 2 and fig. S3) and is effectivelydecoupled fromNi. As a result, the Ni/Co ratio inmost Ni-rich mantle plume olivines exceeds 30 atthe typical mantle Fo range of 89 to 91 (Fig. 2).This is not expected from a core contribution,because Ni/Co ratios for both mantle and core arealmost equal and close to the chondritic value ofabout 20 (40). In addition, Ca is significantlydepleted in many high-Ni and low-Mn olivinesfrom the WPM-THICK group, which cannot beexplained by core contribution. Lastly, theolivines from Gorgona komatiites, which doshow significant excess in 186Os (39), do notindicate large anomalies in Ni and Mn/Fe,whereas Koolau lavas with the highest Niexcess and lowest Mn/Fe ratio in olivines donot show significant elevations in 186Os/188Osratios (39). This suggests complete decouplingof these potentially strong indicators of core-mantle exchange.

References and Notes1. A. W. Hofmann, W. M. White, Earth Planet. Sci. Lett. 57,

421 (1982).2. A. W. Hofmann, in The Mantle and Core, vol. 2 of Treatise

on Geochemistry, H. D. Holland, K. K. Turekian, Eds.(Elsevier, Amsterdam, 2003), p. 61.

3. E. H. Hauri, Nature 382, 415 (1996).4. C. Herzberg, M. J. O'Hara, J. Petrol. 43, 1857 (2002).5. A. V. Sobolev, A. W. Hofmann, S. V. Sobolev,

I. K. Nikogosian, Nature 434, 590 (2005).6. Ni was first proposed as a monitor for replacement of

olivine by pyroxene in (20).7. M. Humayun, L. P. Qin, M. D. Norman, Science 306, 91

(2004).8. C. Herzberg, Nature 444, 605 (2006).9. S. M. Eggins, Contrib. Mineral. Petrol. 110, 387 (1992).

10. Materials and methods are available on Science Online.11. E. J. Jarosevich, J. A. Nelen, J. A. Norberg, Geostand

Newsl. 4, 43 (1980).12. M. J. Walter, J. Petrol. 39, 29 (1998).13. V. S. Sobolev, A. V. Sobolev, Dokl. Akad. Nauk SSSR 237,

437 (1977).14. T. Kogiso, M. M. Hirschmann, D. J. Frost, Earth Planet.

Sci. Lett. 216, 603 (2003).15. T. Kogiso, M. M. Hirschmann, M. Pertermann, J. Petrol.

45, 2407 (2004).16. T. Kogiso, M. Hirschmann, Earth Planet. Sci. Lett. 249,

188 (2006).17. A. Yasuda, T. Fujii, K. Kurita, J. Geophys. Res. 99, 9401

(1994).18. A. E. Ringwood, D. H. Green, Tectonophysics 3, 383

(1966).

Xpx ~ 10-15%Xrc ~ 5 %

Tp~1350oC

Siberian LIP Hawaii Detroit Smt MORB

per

ido

tite

+ e

clo

git

ep

yro

xen

ite

mel

ts

per

ido

tite

mel

ts

Xpx ~ 100%Xrc ~ 20 %

Tp~1550oC

Xpx ~ 60%Xrc ~ 20 %

Tp~1550oC

Xpx ~ 40%X rc~ 20 %Tp~1550oC

Iceland

LITHOSPHERE

Xpx ~ 20%X rc~ 10 %

Tp~1450oC

LITHOSPHERELITHOSPHERE

LITHOSPHERE

eclo

git

e m

elts

Fig. 5. Schematic model illustrating interplay between amount of recycled crust, thickness oflithosphere, and Tp. Blue, upwelling peridotitic mantle; red, recycled oceanic crust (eclogite withfree SiO2 phase); black dots, melting; yellow, reaction zones forming hybrid pyroxenite; pink,refractory restite after eclogite melting; and green, lithosphere. Xpx, amount of pyroxenite derivedmelt in the mixture with peridotite-derived melt, and Xrc, amount of recycled oceanic crust in theperidotitic mantle (42).


RESEARCH ARTICLES

19. G. M. Yaxley, D. H. Green, Schweiz. Mineral. Petrogr.Mitt. 78, 243 (1998).

20. P. B. Kelemen, S. R. Hart, S. Bernstein, Earth Planet.Sci. Lett. 164, 387 (1998).

21. E. H. Hauri, M. D. Kurz, Earth Planet. Sci. Lett. 153, 21(1997).

22. J. C. Lassiter, E. H. Hauri, Earth Planet. Sci. Lett. 164,483 (1998).

23. R. P. Rapp, N. Shimizu, M. D. Norman, G. S. Applegate,Chem. Geol. 160, 335 (1999).

24. D. S. Korzhinskii, Theory of Metasomatic Zoning(Clarendon, Oxford, 1970).

25. M. M. Hirschmann, E. M. Stolper, Contrib. Mineral. Petrol.124, 185 (1996).

26. G. M. Yaxley, Contrib. Mineral. Petrol. 139, 326 (2000).27. P. Beattie, Contrib. Mineral. Petrol. 115, 103 (1993).28. R. J. Kinzler, T. L. Grove, S. I. Recca, Geochim.

Cosmochim. Acta 54, 1255 (1990).29. V. S. Kamenetsky et al., Geology 29, 243 (2001).30. N. H. Sleep, Annu. Rev. Earth Planet. Sci. 20, 19 (1992).31. G. H. Gudfinnsson, D. C. Presnall, J. Petrol. 46, 1645 (2005).32. D. McKenzie, M. J. Bickle, J. Phys. Earth 38, 511 (1990).33. J. P. Morgan, Geochem. Geophys. Geosystems 2,

2000GC000049 (2001).34. G. Ito, J. J. Mahoney, Earth Planet. Sci. Lett. 230, 29 (2005).

35. M. Pertermann, M. M. Hirschmann, J. Petrol. 44, 2173(2003).

36. I. Kushiro, Annu. Rev. Earth Planet. Sci. 29, 71(2001).

37. N. Arndt, J. Geophys. Res. 108, XX (2003).38. I. D. Ryabchikov, Dokl. Earth Sci. 389, 437 (2003).39. A. D. Brandon, R. J. Walker, Earth Planet. Sci. Lett. 232,

211 (2005).40. W. F. McDonough, S. S. Sun, Chem. Geol. 120, 223

(1995).41. L. V. Danyushevsky, J. Volcanol. Geotherm. Res. 110, 265

(2001).42. D. McKenzie, M. J. Bickle, J. Petrol. 29, 625 (1988).43. We thank B. Schulz-Dobrick for supervising purchase and

establishing electron microprobe facility in MPI; HawaiianScientific Drilling Project team, Koolau Scientific DrillingProject team, Ocean Drilling Program team, A. T. Anderson,E. A. Mathez, and N. Gitahi for providing samples;E. J. Jarosevich for supplying microprobe standards;A. Yasevich for sample preparations and N. Groschopffor maintaining the electron microprobe. The paperbenefited from discussions with M. Hirschmann,P. Kelemen, C. Herzberg, B. McDonough, I. Ryabchikov,L. Kogarko, A. Kadik, E. Galimov, V. Batanova. Constructivereviews of C. Herzberg, and P. Kelemen improved the

clarity of the manuscript. The study was supported byWolfgang Paul Award of the Alexander von HumboldtFoundation, Germany, to A.V.S. Partial support from thefollowing sources is also acknowledged: RussianFoundation for Basic Research (RFBR) grants 06-05-65234and НШ-4264.2006.5 and Russian Academy of Sciencesand Deutsche Forschungsgemeinschaft grant HO1026/16-1 to A.V.S., RFBR grant 06-05-64651 to N.M.S.,NSF grants EAR03-36874 to M.O.G. and EAR-0105557to F.A.F., and Netherlands Research Center for IntegratedSolid Earth Science grant 6.2.12 to I.K.N. This is Schoolof Ocean and Earth Science and Technology, Universityof Hawaii, contribution no. 7104.

Supporting Online Materialwww.sciencemag.org/cgi/content/full/1138113/DC1SOM TextFigs. S1 to S5Tables S1 to S4Data set

29 November 2006; accepted 16 March 2007Published online 29 March 2007;10.1126/science.1138113Include this information when citing this paper.

Genes Required for Mitotic SpindleAssembly in Drosophila S2 CellsGohta Goshima,1,3* Roy Wollman,2,3 Sarah S. Goodwin,1 Nan Zhang,1Jonathan M. Scholey,2 Ronald D. Vale,1,3† Nico Stuurman1,3

The formation of a metaphase spindle, a bipolar microtubule array with centrally alignedchromosomes, is a prerequisite for the faithful segregation of a cell’s genetic material. Using afull-genome RNA interference screen of Drosophila S2 cells, we identified about 200 genes thatcontribute to spindle assembly, more than half of which were unexpected. The screen, incombination with a variety of secondary assays, led to new insights into how spindle microtubulesare generated; how centrosomes are positioned; and how centrioles, centrosomes, andkinetochores are assembled.

The diamond-shaped mitotic spindle hasbecome one of the most widely recog-nized images in biology, emblematic of

life’s propagation through cell division. In highereukaryotes, the process of spindle formationbegins after nuclear envelope breakdown (NEB)when microtubules (MTs), generated both fromcentrosomes and from the vicinity of chromatin,are organized into a bipolar array (1–5). Sisterchromatids bind toMTs emanating from oppositepoles, are aligned in the middle of the bipolarMTnetwork, and then ultimately separate and moveapart during anaphase. Failures early in mitosisresult in the formation of an abnormal metaphase

spindle, which can lead to mitotic delay and,potentially, chromosome missegregation duringthe ensuing anaphase.

To understand the mechanism of metaphasespindle assembly, it is critical to identify the pro-teins required for this process and then dissecthow they function. Many mitotic proteins havebeen identified through genetic and RNAi screens(6–10), but the inventory is likely incomplete.Here, we present a genome-wide screen for mi-totic spindle morphology in Drosophila S2 cellsand the functional analysis of unexpected genesdiscovered through the screen.

Identification of genes involved in meta-phase spindle formation by high-throughputmicroscopy. Because the percentage of S2 cellsin mitosis is low (~1%), we conducted ourRNAi screen in the presence of dsRNA (double-stranded RNA) to Cdc27 (a subunit of theanaphase-promoting complex) to delay ana-phase and thereby increase the percentage ofmetaphase cells (~10% of the population). Thus,our screen was designed to investigate the as-sembly of the metaphase spindle, but not ana-phase or cytokinesis. We also rescreened the

final hits without Cdc27 RNAi–induced mitot-ic arrest. The majority (88%) showed identicalphenotypes, although a few genes only manifestclear phenotypes under conditions of mitoticarrest (table S1).

Using our custom, full-genome (14,425 genes)Drosophila RNAi library (11), we treated S2cells with dsRNA for 4 days, conditions thatgenerally reduce protein levels by >80% (12, 13).After dsRNA treatment, cells were fixed andstained for DNA, g-tubulin, MT, and phospho-histone H3 (pH3) in 96 well plates, and about 40sites per well were imaged by automated micros-copy with a high numerical aperture air ob-jective to obtain relatively high-resolution imagesof mitotic spindles (Fig. 1A). To reduce the com-plexity of this large amount of image data, a cus-tom computer code was used to identify, crop,and arrange mitotic spindles into galleries, whichwere then blindly scored by an observer (Fig. 1Band fig. S1). In addition, computer algorithmsmeasured eight parameters of spindle shape, aswell as the intensity of g-tubulin, cell number, andmitotic index (Fig. 1C) (11). More than 4,000,000spindles were analyzed in this screen.

Before beginning this screen, we annotated49 genes that produce mitotic defects in S2 cells(table S2). Of these 49 genes, 45 were identifiedas hits in the primary screen, indicating a highsuccess rate of identifying mitotic phenotypes.However, our final list of genes should not beconsidered as a complete or universal inventory,because genes can be missed (particularly thosewith subtle phenotypes), and some phenotypes(or lack thereof) may be specific to S2 cells.False positives by off-target effects of dsRNAcan occur in RNAi screens (14, 15), so pre-cautions were taken to minimize gene overlap inthe dsRNA design, and all unexpected hits wereconfirmed with another dsRNA that had nooverlap with the first dsRNA (11). To learnmore about the functions of interesting genes,

1Howard Hughes Medical Institute and the Department ofCellular and Molecular Pharmacology, University of Cal-ifornia, San Francisco, 600 16th Street, San Francisco, CA94158, USA. 2Department of Molecular and Cellular Biol-ogy, University of California, Davis, One Shields Avenue,3203 Life Sciences, Davis, CA 95616, USA. 3PhysiologyCourse, Marine Biological Laboratory, 7 MBL Street, WoodsHole, MA 02543, USA.

*Present address: Institute for Advanced Research, NagoyaUniversity, Furo-cho, Chikusa-ku, Nagoya, 464-8601, Japan.†To whom correspondence should be addressed. E-mail:[email protected]


RESEARCH ARTICLES

Corrected 19 April 2007: Minor typographic erros have been corrected.

www.sciencemag.org/cgi/content/full/1138113/DC1

Supporting Online Material for


Alexander V. Sobolev, Albrecht W. Hofmann, Dmitry V. Kuzmin, Gregory M. Yaxley, Nicholas T. Arndt, Sun-Lin Chung, Leonid V. Danyushevsky, Tim Elliott, Frederick A. Frey, Michael O. Garcia, Andrey A.Gurenko, Vadim S. Kamenetsky, Andrew C. Kerr, Nadezhda A. Krivolutskaya, Vladimir V. Matvienkov, Igor K. Nikogosian, Alexander

Rocholl, Ingvar A. Sigurdsson, Nadezhda M. Sushchevskaya, Mengist Teklay

Published 29 March 2007 on Science Express

DOI: 10.1126/science.1138113

This PDF file includes:

Materials and Methods Figs. S1 to S5 Tables S1, S2a, S2b, and S3 References

Other Supporting Online Material for this manuscript includes the following: available at www.sciencemag.org/cgi/content/full/1138113/DC1

Table S2 (complete) as zipped file

Sobolev et al, Supporting Online Material www.sciencemag.org/cgi/content/full/1138113/DC1

20 APRIL 2007 VOL 316 SCIENCE www.sciencemag.org

S1

Electron probe microanalysis. Olivine phenocrysts were analyzed for Si, Fe, Mg, Ca, Ni, Mn, Cr, Al and Co with a Jeol Jxa 8200 electron probe at Max-Planck Institute of Chemistry. Each olivine grain was analyzed in the geometrical center at 20 kV accelerating voltage and probe current of 300 nA. A small olivine subset was analyzed at 30 kV and 200nA. Details of analytical conditions are presented in Table S1. These conditions have been found optimal to obtain best detection limits (see Fig S1). The intensity of the Co Kα line was additionally corrected for overlap with the shoulder of FeKβ second order line using the following linear equation established by analyzing Fe bearing standards free of Co: CoOc=CoOm-0.0011×FeOm-0.013. [S1] Where CoOc and CoOm, FeOm-are corrected and measured values correspondingly.

The constant in equation [S1] was obtained from analyses of San Carlos olivine standard with known Co content (Table S1). Note that the linear approximation of equation [S1] may not be valid for highly Mg (Fo>93) and low Mg (Fo<70) olivines.

In addition San Carlos olivine standard USNM 111312/444 (S1) was run (3 points per each 30-50 measurement points) as an

unknown to monitor drifts in calibration and estimate accuracy of analysis. All measurements of Si, Fe, Mn, Ni, Ca and Al were corrected for deviation of this standard from the reference values (Table S1).

For trace elements, the above measurement conditions routinely yield detection limits of around 6-15 ppm based on 3 sigma criteria by Jeol standard software, and errors of 15-30 ppm for trace elements and 0.01 mol% for Fo content (2 standard errors) established by counting statistics and reproducibility of olivine standard (see Fig S1-S2). Precision and

El St Cryst Line Peak BG(+) BG(-) S.C.O.

Si 1 TAP Kα 90 90 - 190740 *

Al 2 TAP Kα 240 120 120 170 #

Fe 1 LIF Kα 90 90 - 74230 *

Mn 3 LIFH Kα 120 60 60 1084 *

Mg 1 TAP Kα 90 45 45 298040 *

Ca 4 PETJ Kα 120 60 60 665 #

Ni 5 LIF Kα 150 70 70 2907 *

Co 6 LIFH Kα 120 120 - 140 **

Cr 7 PETJ Kα 120 60 60 95 #

Table S1. Typical analytical conditions of electron probe microanalysis of olivine. Probe current 300 nA, acceleration voltage 20kV.


A. V. Sobolev, A. W. Hofmann, D. V. Kuzmin, G. M. Yaxley, N. T. Arndt, et al.

Supporting Online Material Methods and Samples

1.41

1.43

1.45

1.47

1.49

1.51

1.53

1.55

90 90.1 90.2 90.3 90.4 90.5

100M

n/Fe

100

110

120

130

140

150

160

170

90 90.1 90.2 90.3 90.4 90.5

Fo mol%

Co

ppm

600

620

640

660

680

700

720

740

760

90 90.1 90.2 90.3 90.4 90.5

Ca

ppm

2800

2820

2840

2860

2880

2900

2920

2940

2960

2980

3000

90 90.1 90.2 90.3 90.4 90.5N

i ppm

measuredcorrected

40

50

60

70

80

90

100

110

120

130

90 90.1 90.2 90.3 90.4 90.5

Fo mol%

Cr p

pm

1040

1050

1060

1070

1080

1090

1100

1110

1120

1130

1140

90 90.1 90.2 90.3 90.4 90.5

Mn

ppm

Fig. S2. Reproducibility of San Carlos olivine standard analyzed three times every 30 to 50-th measurement points. Measured- represent uncorrected measurements, corrected – stands for measurements corrected for the calibration drift.

Notes for Table S1: Standards (St) - 1- San-Carlos olivine USNM 111312/444 (S.C.O.); 2-pure Al2O3; 3-Rhodonite; 4-Wollastonite; 5-pure NiO; 6- pure Co-metal; 7-pure Cr2O3. Peak and background (BG) counting time in seconds. *-Accepted values for (S.C.O.) from ref (S1) in ppm, **-value measured by LA-ICP MS.

Fig. S1. Detection limit for Mn as function of probe current and peak counting time (in seconds, sec). Detection limit was determined for San Carlos olivine standard with Jeol software using background statistics and 3 sigma criterion. Typical measurement conditions used in literature are marked as “routine analysis”. Analysis at higher beam current and counting times published by (S3) is marked as “precision analysis”. The analytical conditions used in this study are indicated as “high precision analysis”.

0

10

20

30

40

50

60

70

0 50 100 150 200 250 300 350

Probe current, nA

Det

ectio

n lim

it, p

pm

30 sec60 sec120 sec180 sec

high precision analysis

precision analysis

routine analysis

Mn

800

900

1000

1100

1200

1300

1400

1500

800 900 1000 1100 1200 1300 1400 1500

Mn by EPMA, ppm

Mn

by L

A-I

CPM

S, p

pm

100

150

200

250

300

100 150 200 250 300

Co EPMA, ppm

Co

LA

-IC

PMS,

ppm

Fig. S3. Comparison between Co and Mn concentrations in olivine measured by EPMA and LA-ICPMS. Error bar corresponds to ± 2 standard errors for both EPMA and LA-ICP MS



S2

accuracy of Co and Mn measurements were independently checked by LA-ICP MS analysis using Thermo Finnigan Element 2 mass spectrometer and New Wave Up213 YAG laser (213 nm wavelength) setup. KL2-G (S4) and NIST 612 glasses were used as a external standards and Ca in olivine as reference value. Figure S3 shows that both Co

and Mn concentrations measured by EPMA correspond to those analyzed by LA-ICP MS within 20 ppm (2 standard errors).

Olivine database. The averaged most highly magnesian olivines for each sample are presented in Table S2a. Individual olivine analyses are presented on Fig S4 and in Tables S2,c,d,e,f separately for each group: MORB in Table S2c, WPM-THIN in Table S2d, WPM-THICK in Table S2e and KOMATIITES in

Table S2f. These tables are included as separate spreadsheets together with Table S2a in the Excel file called Table S2. Table S2a is also placed in the end of this file. Tables S2c,d,e,f include concentrations of oxides and their relative standard deviations in % (signal counting statistics) for individual olivine grains. Each analytical point in addition to

sample name has a unique number. Table S2a includes group title, sample name, reference for sample description (if published), information on locality, number of averaged high-Mg grains, forsterite content, element concentrations in ppm, standard errors of the mean in ppm, characteristic ratios of elements and calculated amount of pyroxenitic component (X px, see below for explanation).

Melting of pyroxenites. The model hybrid pyroxenite composition from Sobolev et al. (S3) (their Table S2, column 50%) was chosen for the high pressure experimental investigation. A synthetic starting material with this composition (Px-1) was prepared for the experiments by blending high purity oxides and carbonates under analytical grade

acetone for several hours, until the material was homogenous and very fine-grained. The resultant dried powder was then pelletised and fired for 12 hours at 1100°C to decarbonate and partially fuse the components. FeO was added after firing in the form of synthetic fayalite (Fe2SiO4), and again blended under acetone. The final mixture was then dried overnight at 200°C and subsequently stored at 110°C. The actual composition of the Px-1

A

500

1000

1500

2000

2500

3000

3500

70 75 80 85 90 95

Fo

Mn

ppm

WPM-THICKKOMATIITESWPM-THINMORB

C

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

70 75 80 85 90 95

Fo

Ni p

pm

E

0

500

1000

1500

2000

2500

3000

3500

70 75 80 85 90 95

Fo

Ca

ppm

B

1.1

1.2

1.3

1.4

1.5

1.6

1.7

1.8

1.9

70 75 80 85 90 95

Fo10

0Mn/

Fe

D

0

0.5

1

1.5

2

2.5

70 75 80 85 90 95

Fo

Ni/(

Mg/

Fe)/1

000

F

0

500

1000

1500

2000

2500

70 75 80 85 90 95

Fo

Cr p

pm

Fig. S4. Composition of olivine phenocrysts: all data from database. Group of MORB olivines shown on top of all groups.



S3

starting material was checked by electron probe microanalysis of quenched glass prepared by fusing the starting mix at 1300°C in an Ar atmosphere in a box furnace and quenching in water (Table S3).

High pressure experiments were run at a pressure of 3.5 GPa and temperatures between 1400-1570°C in a conventional 1.27cm piston-cylinder apparatus at the Australian National University. The Px-1 starting material (≈1 mg) was loaded into graphite capsules, which were sealed by arc-welding in Pt outer capsules. The

capsules were placed centrally into NaCl-pyrex sleeves with MgO inserts, an assembly which requires no friction correction. Type B thermocouples (Pt6Rh94/Pt30Rh70) were employed, with the thermocouple join placed within a fraction of a millimeter of the Pt capsule. Runs were brought to run temperature and pressure simultaneously. Temperature was controlled throughout by Eurotherm controllers attached to the thermocouple and is accurate to ±10°C. Pressure is accurate to ±0.1 GPa. Experiments were maintained at the

desired PT condition long enough to allow a close approach to equilibrium, and then quenched by terminating power to the furnace.

After recovery from the post-run assembly, the Pt capsules were mounted in 1 inch diameter epoxy buttons, sectioned longitudinally and polished, in preparation for analysis by scanning electron microscopy at the Australian National University and electron probe microanalysis at the Max Planck Institute for Chemistry in Mainz.

Pyroxenite-derived component in melts. In order to estimate endmember compositions we have averaged compositions of melts of pyroxenite and peridotite (see Fig.S5). For pyroxenite we averaged all melt compositions except numbers C-2298 and C-2327 (Table S3) produced at low temperature and low percentage of melting. For peridotites we averaged 7 compositions of melts from runs 30.12, 30.07, 30.14, 30.1 (3.GPa), 40.06, 40.07, 40.05 (4.GPa) of Walter, 1998 (S2). The endmember compositions are shown in Table S3. We assumed that the mixtures consist of endmembers with similar temperatures, and we therefore excluded (from the averaging procedure) experimental melts with very low or very high temperatures (see Fig.S5).

The calculated endmember melts were then mixed (in 10% intervals) producing mixed melt compositions. For these melts the composition of equilibrium olivines were

calculated using Herzberg’s model (S5). The calculated Mn/Fe ratios of the olivines and the amounts of pyroxenitic endmember (Xpx) yield a straight line: Xpx = 3.48–2.071×(100Mn/Fe) [S2]

We used this equation to calculate the amount of pyroxenitic endmember for each average olivine composition in Table S2a.

The calculated endmember compositions are derived from experiments at high pressures and temperatures. Therefore, they are directly relevant primarily to thick-lithosphere settings, but we have applied them for all settings. Equivalent estimates for lower pressures and temperatures (1.0 GPa and 1300-1350oC) have been calculated using pMELTS (S6). This yielded similar degrees of melting and residual assemblages (50 to 60% of melting and low-Ca pyroxene dominated residue) as those derived from high pressure and temperature runs (Table S3). This gives us some confidence that the same Mn/Fe ratios can be applied to lower pressures, but experimental confirmation will be needed. Additional confidence in the results is derived from the observation that the extreme compositions observed in both MORB and WPM-thick settings match the calculated endmember compositions reasonably well.

Run DT T Phases proportions, wt% Melt compositions in oxide wt%, Ni in ppm

N h o C melt Opx Cpx Ga SiO2 TiO2 Al2O3 FeO MnO MgO CaO Na2O K2O Cr2O3 Ni

C-2298 45 1400 8 0. 79 14 60.70 2.76 14.40 6.80 0.068 4.44 7.71 2.34 0.70 0.07 453

C-2317 48 1450 16 0 71 12 55.76 2.33 14.20 7.63 0.081 7.45 7.95 2.16 0.44 0.06 626

C-2301 5 1500 33 0 60 6 53.30 1.58 14.21 8.25 0.106 10.67 8.61 1.75 0.13 0.09 810

C-2330 5 1520 42 0 55 3 52.54 1.29 14.22 8.65 0.112 11.66 8.65 1.98 0.14 0.11 850

C-2332 5 1540 72 24 4 0 52.00 0.86 13.55 8.19 0.125 14.08 9.00 1.80 0.08 0.20 796

C-2318 6 1550 76 18 7 0 52.38 0.83 13.39 7.95 0.117 14.67 8.77 1.74 0.09 0.20 843

C-2319 5 1575 79 20 0 0 52.20 0.79 13.07 8.11 0.126 15.44 8.52 1.67 0.07 0.22 841

Px-1, Bulk composition 52.67 0.64 11.26 7.55 0.119 18.48 7.05 1.52 0.06 0.25 1000 Peridotite-derived endmember 46.79 0.85 11.50 9.68 0.185 19.07 10.00 0.84 0.47 0.39 642

Pyroxenite -derived endmember 52.56 1.07 13.71 8.24 0.117 13.32 8.72 1.79 0.10 0.16 830

Table S3. Proportions of phases produced and melt compositions from melting of pyroxenite at 3.5 GPa and 1400-1575oC.

Note for Table S3: Melt, Opx, Cpx, Ga, - proportions of melt, orthopyroxene, clinopyroxene and garnet in experimental runs in mass fractions, calculated using list square method from bulk composition and compositions of phases. DT- run duration in haurs, T- temperature in oC. The amount of Ni has been calculated using the method described in Sobolev et al, 2005 (S3). This method uses known phase proportions and the distribution of Ni between phases, to calculate Ni in the phases on the basis of mass balance and known bulk Ni content. The assumed bulk Ni concentrations are: 1000 ppm for pyroxenite (Px-1) and 1900 ppm for peridotite (S3).

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1350 1400 1450 1500 1550 1600 1650 1700

Mel

t fra

ctio

n

Pyroxenite 3.5 GPaPeridotite 3.0 GPaPeridotite 4.0 GPa

0.800

1.000

1.200

1.400

1.600

1.800

2.000

2.200

1350 1400 1450 1500 1550 1600 1650 1700

100M

n/Fe

mel

t

0.000

0.200

0.400

0.600

0.800

1.000

1.200

1.400

1.600

1.800

1350 1400 1450 1500 1550 1600 1650 1700

Temperature C

100N

i/Mg

mel

t

Fig. S5. Melt fraction and composition of experimental melts of pyroxenite and fertile peridotite. Temperatures for experimental runs of fertile peridotite (S2) are extrapolated to 3.5 GPa by using slope 100C/GPa. Data for pyroxenite from Table S3. Crystalline phase coexisting with melt are: Ol - olivine; Opx - orthopyroxene; Cpx -clinopyroxene; Ga - garnet. Blue and red ellipses indicate analyses averaged to calculate peridotite-derived and pyroxenite-derived endmembers respectively.



S4

Oceanic crust in magma sources. A final step is to estimate the actual amount of recycled oceanic crust (Xcrc) from the determined proportion of pyroxenite-derived melt. This quantity is linked to the proportion of hybrid pyroxenite-derived melt (Xpx) by the degree of melting of eclogite (Fe), the amount of eclogite-derived melt needed to produce hybrid pyroxenite from peridotite (Xe), and the degrees of melting of peridotite (Fpe) and pyroxenite (Fpx) (S3):

[S3]

Following assumption of Sobolev et al,

(2005) (S3), we propose that the extent of melting of eclogites reaches a maximum of 50%. This limit is unlikely to be exceeded significantly during fractional melting, because this process removes Na (S7) and thereby renders the residue highly refractory. Therefore it is also not strongly dependent on the potential temperature of the rising mantle material. The amount of pyroxenite produced by reaction of the primary (eclogite-derived) melt with peridotite is prescribed by the stoichiometry of the reaction, which also yields proportions of the reactants of close to 50:50 (S3). For these, proportions, the amount of reaction pyroxenite equals the amount of initial eclogite. The melt fraction produced by the pyroxenite can be estimated at low pressure (1-2 GPa by modeling using pMELTS software (S6)), yielding a maximum of about 60 % for batch melting at 1 GPa and 1320 to 1350°C. At higher pressures (3.5 GPa) we must rely on the experimental data (Table S3, Fig. S5), which also yield a maximum melt fraction of about 50-60% for batch melting for 1540-1550°C.

These values will be significantly lower for fractional melting (S8), for the same reason as discussed for eclogite melting (i.e. early Na removal). Adopting an amount of 50 % pyroxenite melting will therefore result in a minimum estimate in the amount of recycled crust present. Finally, we estimate the melt fraction of peridotite using published models (S5) for MORB (10%), Iceland, as a proxy for the WPM-THIN group (20%), Hawaii, as proxy for the WPM-THICK group (10%) and Archaean komatiites -40% (S9). This yields the following average estimates for the amounts of recycled oceanic crust in the mantle sources: 4% for MORB, 11% for WPM-THIN group, 16% for WPM-THICK group, and around 13% for Archean komatiites. According to (S10) the melt fraction of peridotite for Ontong Java high-Mg magmas corresponds to 30%, which yield the amount of recycled oceanic crust of 13-28% in the mantle source of these magmas.

Supporting Online References

S1. E. J. Jarosevich, J. A. Nelen, J. A. Norberg, Geostand. Newsl. 4, 43 (1980).

S2. M. J. Walter, Journal of Petrology 39, 29 (J1998).

S3. A. V. Sobolev, A. W. Hofmann, S. V. Sobolev, I. K. Nikogosian, Nature

434, 590 (2005).

S4. K. P. Jochum et al., Geostandards Newsletter-the Journal of Geostandards and

Geoanalysis 24, 87 (2000).

S5. C. Herzberg, M. J. O'Hara, Journal of Petrology 43, 1857 (2002).

S6. M. S. Ghiorso, M. M. Hirschmann, P. W. Reiners, V. C. Kress,

Geochemistry Geophysics Geosystems 3 (2002).

S7. T. Kogiso, M. Hirschmann, Earth and Planetary Science Letters 249, 188

(2006).

S8. I. Kushiro, Annual Review of Earth and Planetary Sciences 29, 71 (2001).

S9. N. Arndt, Journal of Geophysical Research-Solid Earth 108 (2003).

S10. J. G. Fitton, M. Godard, in Origin and Evolution of the Ontong Java PlateauJ. G.

Fitton, J. J. Mahoney, P. J. Wallace, A. D. Saunders, Eds. (Geological

Society, London, Special Publication, 229 London, 2004) pp. 151-178.

S11. B. Sylvander, D. M. Christie, L. V. Danyushevsky, in Eos. (1996), vol. 77,

pp. F691.

S12. M. R. Perfit et al., Earth and Planetary Science Letters 141, 91 (1996).

S13. A. V. Sobolev, Petrology 4, 209 (1996).

S14. V. S. Kamenetsky et al., Earth and Planetary Science Letters 160, 115 (1998).

S15. V. Kamenetsky, Earth and Planetary Science Letters 142, 479 (1996).

S16. A. P. Leroex, H. J. B. Dick, L. Gulen, A. M. Reid, A. J. Erlank, Geochimica Et

Cosmochimica Acta 51, 541 (1987).

S17. V. S. Kamenetsky et al., Geology 29, 243 (2001).

S18. N. M. Sushchevskaya et al., Geochemistry International 43, 222 (2005).

S19. in Table S2b.

S20. T. Elliott, J. Blichert-Toft, A. Heumann, G. Koetsier, V. Forjaz, Geochimica

Et Cosmochimica Acta 71, 219 ( 2007).

S21. S. Huang, M. Regelous, T. Thordarson, F. A. Frey, Geochemistry Geophysics

Geosystems 6 (2005).

S22. A. C. Kerr, R. W. Kent, B. A. Thomson, J. K. Seedhouse, C. H. Donaldson,

Journal of Petrology 40, 873 (1999).

S23. G. M. Yaxley, V. S. Kamenetsky, M. Kamenetsky, M. D. Norman, D.

Francis, Contributions to Mineralogy and Petrology 148, 426 (2004).

S24. L. M. Larsen, A. K. Pedersen, Journal of Petrology 41, 1071 (2000).

S25. K. G. Cox, B. G. Jamieson, Journal of Petrology 15, 269 (1974).

S26. Y. G. Xu, S. L. Chung, B. M. Jahn, G. Y. Wu, Lithos 58, 145 (2001).

S27. S. L. Chung, B. M. Jahn, Geology 23, 889 (1995).

S28. M. Teklay, A. W. Hofmann, G. E. Brugmann, J. C. Lassiter, Geochimica Et

Cosmochimica Acta 66, A767 (2002).

S29. A. V. Sobolev, I. K. Nikogosian, Petrology 2, 111 (1994).

S30. J. M. Rhodes, M. J. Vollinger, Geochemistry Geophysics Geosystems 5, Q03G13

(2004).

S31. F. A. Frey et al., Journal of Geophysical Research-Solid Earth and Planets 96,

14347 (1991).

S32. F. A. Frey, M. O. Garcia, M. F. Roden, Geochimica Et Cosmochimica Acta 58,

1441 (1994).

S33. E. H. Haskins, M. O. Garcia, Contributions to Mineralogy and Petrology 147,

162 (2004).

S34. M. D. Norman, M. O. Garcia, Earth and Planetary Science Letters 168, 27

(1999).

S35. M. O. Garcia, D. J. P. Foss, H. B. West, J. J. Mahoney, Journal of Petrology

36, 1647 (1995).

S36. V. V. Matvienkov, O. G. Sorokhtin, Okeanologiya 38, 742 (1998).

S37. A. A. Gurenko et al., Chemical Geology 233, 75 (2006).

S38. A. A. Gurenko, T. H. Hansteen, H.-U. Schmincke, in Proceedings of the

Ocean Drilling Program, Scientific Results P. P. E. Weaver, H.-U. Schmincke, J.

V. Firth, W. Duffield, Eds. (1998), vol. 157, pp. 375-401.

S39. N. T. Arndt, A. J. Naldrett, D. R. Pyke, Journal of Petrology 18, 319 (1977).

S40. N. T. Arndt, Journal of Petrology 27, 279 (1986).

S41. M. J. Bickle et al., in The geology of the Belingwe Greenstone Belt, Zimbabwe M. J.

Bickle, E. G. Nisbet, Eds. (Balkema, Rotterdam, 1993) pp. 175-214.

S42. N. T. Arndt, G. E. Brügmann, K. Lehnert, C. Chauvel, B. W. Chappell, in

Geochemistry and Mineralization of Proterozoic Volcanic Suites T. C. Pharaoh, R.

D. Beckinsale, D. Rickard, Eds. (Geol. Soc. Spec. Publ., 1987) pp. 133-145.

S43. A. C. Kerr, Lithos 84, 77 (2005).

Table S2b. Locations (GPS coordinates) and description for Icelandic samples

Sample Locality Description latitude longitude

01-7 Reykjanes, Haleyjabunga Glassy olivine phyric picrite N63o 48.954' W22o 38.646' 03-102 Reykjanes, Lagafell Glassy olivine phyric picrite N63o 52.906' W22o 32.611'

03-106 Reykjanes, Haleyjabunga Glassy olivine phyric picrite N63o 48.873' W22o 38.724'

01-15 Reykjanes, Sulur Glassy olivine –phyric basalt N63o 54.094' W22o 32.558'

01-12 Reykjanes, Stapafell Glassy olivine-plagioclase phyric picrite N63o 54.601' W22o 31.370'

01-8 Hengill, Midfell Glassy olivine phyric picrite with rare plagioclase and clinopyroxene xenocrysts N64o 10.140' W21o 04.257'

01-10 Hengill, Midfell Glassy olivine phyric picrite with rare plagioclase and clinopyroxene xenocrysts N64o 10.103' W21o 04.118'

01-19 Hengill, Maelifell Glassy olivine phyric basalt with rare plagioclase and clinopyroxene xenocrysts N64o 06.252' W21o 10.744'

03-131 Kistufell Glassy olivine phyric basalt N64o 47.806' W17o 13.745'




01-55 Theistareykir, Laufrandarhraun Glassy olivine phyric picrite N65o 56.285' W17o 05.047'

01-57-4 Theistareykir, Laufrandarhraun Glassy olivine phyric picrite N65o 55.791' W17o 04.463'


01-41 Theistareykir, Theistareykjahraun Glassy olivine phyric basalt N65o 57.547' W17o 04.120'




01-51 Theistareykir, Langavitishraun Glassy olivine phyric basalt with rare clinopyroxene phenocrysts N65o 56.058' W16o 52.282'

03-224 Snaefellsness, Enni Glassy olivine phyric basalt N64o 54.146' W23o 45.776'

03-226 Snaefellsness, Sydri-Raudamelur Olivine phyric basalt N64o 52.296' W22o 17.368'

03-220 Snaefellsness, Ytri-Raudamelur Olivine phyric basalt N64o 52.717' W22o 20.860'

)111

( +−

+−

=

ee

e

pe

px

px

pxe

ecrc

XF

FFF

XX

F

XX

Sobo

lev

et a

l, Su

ppor

ting

Onl

ine

Mat

eria

l ww

w.sc

ienc

emag

.org

/cgi

/con

tent

/full/

1138

113/

DC

1

20 A

PRIL

200

7 VO

L 31

6 SC

IEN

CE

ww

w.s

cien

cem

ag.o

rg

S5

Tab

le S

2a. A

vera

ge c

ompo

sitio

ns o

f mos

t Mg-

rich

oliv

ine

phen

ocry

sts (

see

text

of p

aper

for d

efin

ition

). GR

OUP

Geod

ynam

ic se

tting

Refer

ence

s Sa

mple

Geog

raph

ic

Loca

lity

N Fo

Al

ppm

Fe pp

m Mn

ppm

Mg pp

m Ca

ppm

Ni pp

m Co

ppm

Cr pp

m 10

0*Mn

/Fe

100*

Ni/M

g Ni

/(Mg/F

e)/10

00

100C

a/Fe

X px

Mn/F

e ST

E Fo

wt%

Al p

pm

Fe pp

m Mn

ppm

Mg pp

m Ca

ppm

Ni pp

m Co

ppm

Cr pp

m

MORB

MO

RB

(S11

) 11

3-16

Ind

ian O

cean

10

5.22 o

E SE

IR

24

89.49

27

0 79

979

1293

29

6327

19

95

2109

389

1.617

0.7

12

0.569

2.4

94

0.135

0.07

7 50

2 10

26

8 13

38

5 MO

RB

MORB

(S

11)

145-

3 Ind

ian O

cean

11

6.72 o

E SE

IR

9 89

.90

298

7703

2 12

24

2982

38

1761

25

00

36

9 1.5

90

0.838

0.6

46

2.285

0.1

91

0.0

5 4

422

8 23

4 9

27

9

MORB

MO

RB

unpb

30

-29

Indian

Oce

an

124.2

8 oE

SEIR

10

89

.60

236

7918

5 12

80

2968

74

1843

21

15

39

0 1.6

17

0.712

0.5

64

2.328

0.1

35

0.1

0 10

72

4 14

49

9 20

71

8 MO

RB

MORB

(S

11)

144-

13

Indian

Oce

an

115.2

1 oE

SEIR

29

89

.16

269

8231

5 13

16

2944

71

1921

19

93

36

3 1.5

99

0.677

0.5

57

2.334

0.1

71

0.0

5 6

377

8 26

5 20

39

5 MO

RB

MORB

un

pb

18-1

0 Ind

ian O

cean

12

6.74 o

E SE

IR

24

89.60

27

2 78

878

1239

29

5854

19

21

2303

372

1.571

0.7

78

0.614

2.4

35

0.229

0.06

4 43

8 9

280

12

38

7

MORB

MO

RB

(S11

) 69

-1

Indian

Oce

an

88.92

oE S

EIR

11

89.89

31

9 76

966

1205

29

7733

20

36

1972

299

1.566

0.6

62

0.510

2.6

46

0.239

0.05

10

417

8 28

7 10

29

4 MO

RB

MORB

(S

11)

88-2

2 Ind

ian O

cean

96

.83 oE

SEI

R 10

88

.18

213

8862

7 14

53

2878

37

2153

15

72

38

0 1.6

39

0.546

0.4

84

2.429

0.0

88

0.1

1 8

755

14

526

10

18

7

MORB

MO

RB

unpb

5-

2 Ind

ian O

cean

12

6.52 o

E SE

IR

33

89.63

25

7 78

288

1238

29

4600

18

75

2383

364

1.582

0.8

09

0.633

2.3

95

0.207

0.04

3 30

7 6

235

3 17

2 MO

RB

MORB

un

pb

16-1

9 Ind

ian O

cean

12

7.58 o

E SE

IR

18

88.40

23

5 87

179

1357

28

9113

19

20

1953

348

1.557

0.6

76

0.589

2.2

02

0.258

0.15

5 11

12

18

675

21

48

6

MORB

MO

RB

unpb

J-P

E-1

Pacif

ic Oc

ean

Garre

tt F.Z

. 32

90

.82

395

6962

7 10

84

2997

51

2071

28

17

48

0 1.5

57

0.940

0.6

54

2.975

0.2

59

0.0

2 10

17

1 3

169

3 38

11

MORB

MO

RB

(S12

) SQ

-1

Pacif

ic Oc

ean

Siqu

eiros

F.Z

. 89

90

.38

377

7245

4 11

31

2962

61

1985

26

61

137

413

1.561

0.8

98

0.651

2.7

39

0.250

0.03

4 19

1 3

201

2 15

1

5 MO

RB

MORB

(S

13)

FZ 64

9-5(

i) At

lantic

Oce

an

9 N M

AR

36

89.75

30

1 76

919

1213

29

3090

19

53

2444

374

1.577

0.8

34

0.642

2.5

39

0.216

0.04

5 30

9 4

200

5 15

4 MO

RB

MORB

(S

13)

FZ 64

9-5(

ii) At

lantic

Oce

an

9 N M

AR

28

88.76

30

2 84

161

1235

28

9205

18

32

2410

371

1.467

0.8

33

0.701

2.1

77

0.444

0.04

7 27

6 6

173

14

30

6

MORB

MO

RB

(S13

) 64

9(i)

Atlan

tic O

cean

9N

MAR

12

89

.63

314

7804

4 12

22

2933

79

1929

25

01

144

360

1.566

0.8

53

0.665

2.4

72

0.239

0.05

7 41

0 6

334

18

20

2 5

MORB

MO

RB

(S13

) 64

9(ii)

Atlan

tic O

cean

9N

MAR

10

88

.87

310

8345

2 12

40

2898

34

1875

24

34

155

390

1.486

0.8

40

0.701

2.2

47

0.406

0.08

9 63

9 10

58

5 20

55

5

8 MO

RB

MORB

un

pb

AII-6

0, St

n9, D

9, 9-

20

N. A

tlanti

c 43

N MA

R 88

89

.81

332

7641

6 11

84

2931

34

1896

25

43

43

0 1.5

49

0.868

0.6

63

2.481

0.2

74

0.0

2 2

152

2 10

3 2

11

3

MORB

MO

RB

unpb

AI

I-32 1

1-92

N.

Atla

ntic

43N

MAR

92

90.70

24

2 70

116

1151

29

7618

20

32

2461

502

1.642

0.8

27

0.580

2.8

98

0.082

0.01

5 10

9 2

97

9 7

5

MORB

MO

RB

unpb

AI

I-32 1

2-6(

i) N.

Atla

ntic

43N

MAR

74

90.83

24

0 69

316

1124

29

8835

23

40

2221

428

1.622

0.7

43

0.515

3.3

76

0.124

0.03

3 20

0 3

134

14

19

10

MO

RB

MORB

un

pb

AII-3

2 12-

6(ii)

N. A

tlanti

c 43

N MA

R 13

90

.88

252

6893

4 11

75

2989

50

2174

22

91

51

0 1.7

05

0.766

0.5

28

3.153

-0

.048

0.0

1 3

57

2 16

6 33

10

3 MO

RB

MORB

un

pb

AII-3

2 12-

2 N.

Atla

ntic

43N

MAR

85

91.13

25

4 66

863

1089

29

8972

22

21

2339

531

1.629

0.7

82

0.523

3.3

21

0.110

0.03

4 20

1 4

132

27

18

10

MO

RB

MORB

(S

14)

AII-3

2 12-

7 N.

Atla

ntic

43N

MAR

55

91.03

21

3 67

956

1115

30

0251

23

93

2252

453

1.641

0.7

50

0.510

3.5

22

0.084

0.03

4 21

1 4

152

22

11

10

MO

RB

MORB

(S

15)

AII-3

2 11-

178

N. A

tlanti

c 43

N MA

R 68

90

.88

260

6915

5 11

40

2998

97

2072

24

79

52

1 1.6

48

0.827

0.5

72

2.996

0.0

70

0.0

3 5

202

3 12

6 12

15

6 MO

RB

MORB

(S

15)

CH-3

1 Dr0

8 695

N.

Atla

ntic

Famo

us

17

91.21

30

6 66

586

1039

30

0597

21

00

2721

592

1.561

0.9

05

0.603

3.1

54

0.251

0.09

9 67

0 12

37

1 8

81

8

MORB

MO

RB

(S15

) CH

-31 D

r08 1

36-0

2 N.

Atla

ntic

Famo

us

19

91.38

29

8 64

581

1012

29

7930

20

19

2620

595

1.567

0.8

79

0.568

3.1

27

0.237

0.07

10

523

9 26

5 7

89

8

MORB

MO

RB

(S15

) AR

P-73

10.03

N.

Atla

ntic

Famo

us

52

91.30

29

5 65

425

1039

29

8895

20

46

2535

582

1.589

0.8

48

0.555

3.1

26

0.193

0.05

9 34

5 5

180

5 31

8 MO

RB

MORB

(S

16)

AG32

4-68

S.

Atla

ntic

Bouv

et 44

90

.72

319

6999

6 10

88

2978

91

2024

26

53

45

0 1.5

54

0.891

0.6

23

2.891

0.2

64

0.0

4 7

270

5 18

5 7

41

10

MO

RB

MORB

un

pb

AG-3

2 3-3

6 S.

Atla

ntic

Bouv

et 98

89

.95

282

7547

4 12

17

2940

27

1869

22

39

42

7 1.6

12

0.761

0.5

75

2.476

0.1

44

0.0

4 4

264

4 15

6 4

40

6

MORB

MO

RB

unpb

AG

-32 3

-41

S. A

tlant

ic Bo

uvet

73

90.22

30

1 74

107

1199

29

7481

19

02

2432

474

1.618

0.8

18

0.606

2.5

66

0.132

0.02

5 14

9 3

122

10

35

7

MORB

MO

RB

(S17

) 51

8-60

/1 S.

Atla

ntic

Bouv

et 5

86.70

22

5 99

549

1199

28

2320

11

15

3478

536

1.204

1.2

32

1.226

1.1

20

0.99

0.0

5 5

308

6 34

5 11

24

22

MORB

MO

RB

unpb

13

-12/4

9a

Atlan

tic O

cean

Ro

manc

he F

.Z.

60

89.35

19

8 80

862

1332

29

5314

23

92

1999

13

9 27

0 1.6

48

0.677

0.5

47

2.958

0.0

70

0.0

3 3

230

4 20

6 6

13

1 6

MORB

MO

RB

unpb

13

-12/4

9c

Atlan

tic O

cean

Ro

manc

he F

.Z.

58

88.98

19

8 82

768

1357

29

0684

24

55

1953

14

5 26

4 1.6

40

0.672

0.5

56

2.966

0.0

87

0.0

4 5

268

5 17

0 13

13

1

5 MO

RB

MORB

un

pb

13-1

2/49d

At

lantic

Oce

an

Roma

nche

F.Z

. 19

88

.90

185

8379

5 13

88

2919

45

2415

18

72

143

250

1.656

0.6

41

0.537

2.8

81

0.053

0.06

7 41

5 7

229

31

18

2 10

MO

RB

MORB

un

pb

13-1

2/49e

At

lantic

Oce

an

Roma

nche

F.Z

. 36

87

.69

185

9253

9 15

14

2867

53

1715

20

76

154

322

1.636

0.7

24

0.670

1.8

54

0.095

0.02

4 10

2 3

160

4 10

2

5 MO

RB

MORB

un

pb

13-1

2/49f

Atlan

tic O

cean

Ro

manc

he F

.Z.

17

89.39

19

5 80

132

1331

29

3645

24

43

1985

14

1 26

0 1.6

61

0.676

0.5

42

3.048

0.0

44

0.0

6 6

478

9 33

9 11

12

2

8 MO

RB

MORB

un

pb

13-1

2/49g

At

lantic

Oce

an

Roma

nche

F.Z

. 21

89

.23

201

8129

9 13

38

2929

63

2402

19

96

145

262

1.646

0.6

81

0.554

2.9

54

0.074

0.06

3 43

8 8

358

11

15

2 6

MORB

MO

RB

unpb

13

-11/1

At

lantic

Oce

an

Roma

nche

F.Z

. 21

88

.11

240

8921

9 13

58

2876

67

1705

19

96

154

307

1.522

0.6

94

0.619

1.9

11

0.330

0.01

3 75

2

130

2 9

2 4

MORB

MO

RB

unpb

13

-11/2

At

lantic

Oce

an

Roma

nche

F.Z

. 36

87

.44

205

9409

0 13

36

2851

63

1854

15

68

164

341

1.420

0.5

50

0.517

1.9

70

0.542

0.05

3 34

1 6

219

6 11

1

4 MO

RB

MORB

un

pb

13-1

2/66

Atlan

tic O

cean

Ro

manc

he F

.Z.

2 90

.56

283

7090

2 11

93

2960

41

2433

22

79

130

356

1.682

0.7

70

0.546

3.4

32

-0.00

1

0.01

13

253

0 15

71

18

8 4

7 MO

RB

MORB

un

pb

13-1

2/78

Atlan

tic O

cean

Ro

manc

he F

.Z.

13

89.34

21

1 79

948

1314

29

1469

23

95

2059

14

7 28

3 1.6

43

0.707

0.5

65

2.996

0.0

80

0.0

8 5

562

10

433

18

23

3 9

MORB

MO

RB

unpb

13

-13/5

At

lantic

Oce

an

Roma

nche

F.Z

. 23

89

.16

230

8145

1 13

09

2914

85

1722

17

67

161

357

1.607

0.6

06

0.494

2.1

14

0.154

0.03

2 23

8 3

195

3 3

2 2

MORB

MO

RB

unpb

13

-13/1

4 At

lantic

Oce

an

Roma

nche

F.Z

. 2

88.27

16

4 87

742

1448

28

7388

18

83

1501

14

2 28

4 1.6

51

0.522

0.4

58

2.146

0.0

64

0.0

8 0

653

0 39

11

8

0 17

MO

RB

MORB

(S

18)

22-4

0 Ar

ctic o

cean

Kn

ipovic

h ridg

e 31

87

.56

213

9354

7 15

12

2865

42

1813

17

75

155

288

1.616

0.6

19

0.579

1.9

38

0.136

0.04

6 21

6 4

350

4 35

8 MO

RB

MORB

(S

18)

22-2

3 Ar

ctic o

cean

Kn

ipovic

h ridg

e 17

87

.42

219

9421

1 15

24

2849

00

1824

17

47

157

291

1.618

0.6

13

0.578

1.9

36

0.132

0.08

4 40

6 7

602

6 34

7 W

PM-T

HIN

OIB

(S19

) 01

-7

Icelan

d Re

ykjan

es, H

aleyja

bung

a 52

90

.22

344

7415

6 11

76

2976

83

2335

25

18

48

7 1.5

86

0.846

0.6

27

3.148

0.1

98

0.0

3 10

25

2 4

217

11

12

9

WPM

-THI

N OI

B (S

19)

03-1

02

Icelan

d Re

ykjan

es, L

agafe

ll 31

90

.66

355

7078

0 11

42

2989

21

2440

25

20

54

9 1.6

14

0.843

0.5

97

3.447

0.1

41

0.0

5 14

36

2 7

264

17

23

16

W

PM-T

HIN

OIB

(S19

) 03

-106

Ice

land

Reyk

janes

, Hale

yjabu

nga

65

90.17

31

4 74

317

1175

29

6602

23

08

2498

462

1.581

0.8

42

0.626

3.1

05

0.208

0.03

7 19

7 4

210

5 10

7 W

PM-T

HIN

OIB

(S19

) 01

-15

Icelan

d Re

ykjan

es, S

ulur

8 88

.39

237

8768

7 13

72

2905

73

1989

23

07

154

434

1.564

0.7

94

0.696

2.2

68

0.243

0.12

8 87

3 15

68

1 31

43

2

16

WPM

-THI

N OI

B (S

19)

01-1

2 Ice

land

Reyk

janes

, Stap

afell

7 87

.61

249

9289

2 14

57

2857

89

1987

21

23

36

8 1.5

69

0.743

0.6

90

2.139

0.2

34

0.1

3 5

952

15

695

35

73

18

W

PM-T

HIN

OIB

(S19

) 01

-8

Icelan

d He

ngill,

Midf

ell

11

90.25

47

4 73

667

1150

29

6824

24

60

2464

671

1.561

0.8

30

0.612

3.3

39

0.251

0.10

18

722

9 39

7 34

39

26

WPM

-THI

N OI

B (S

19)

01-1

0 Ice

land

Heng

ill, M

idfell

25

89

.80

373

7695

3 12

24

2947

65

2672

23

03

54

3 1.5

91

0.781

0.6

01

3.472

0.1

88

0.0

5 12

39

1 5

347

19

30

14

W

PM-T

HIN

OIB

(S19

) 01

-19

Icelan

d He

ngill,

Mae

lifell

5 90

.24

426

7353

3 11

55

2958

22

2444

27

31

149

643

1.571

0.9

23

0.679

3.3

24

0.228

0.19

9 13

81

23

719

17

80

6 23

W

PM-T

HIN

OIB

(S19

) 03

-131

Ice

land

Kistu

fell

67

88.43

29

5 86

728

1335

28

8535

21

70

2305

16

9 30

7 1.5

40

0.799

0.6

93

2.502

0.2

94

0.0

2 9

178

4 19

6 7

13

1 6

WPM

-THI

N OI

B (S

19)

03-1

64

Icelan

d Ki

stufel

l 25

88

.84

331

8404

6 12

87

2911

32

2174

24

11

167

356

1.531

0.8

28

0.696

2.5

87

0.311

0.05

10

320

8 42

9 18

18

2

8

Sobo

lev

et a

l, Su

ppor

ting

Onl

ine

Mat

eria

l ww

w.sc

ienc

emag

.org

/cgi

/con

tent

/full/

1138

113/

DC

1

20 A

PRIL

200

7 VO

L 31

6 SC

IEN

CE

ww

w.s

cien

cem

ag.o

rg

S6

GROU

P Ge

odyn

amic

settin

g Re

feren

ces

Samp

le Ge

ogra

phic

Lo

cality

N

Fo

Al pp

m Fe

ppm

Mn pp

m Mg

ppm

Ca pp

m Ni

ppm

Co pp

m Cr

ppm

100*

Mn/F

e 10

0*Ni

/Mg

Ni/(M

g/Fe)

/1000

10

0Ca/F

e X

px M

n/Fe

STE

Fo w

t% A

l ppm

Fe

ppm

Mn pp

m Mg

ppm

Ca pp

m Ni

ppm

Co pp

m Cr

ppm

WPM

-THI

N OI

B (S

19)

03-1

40

Icelan

d Ki

stufel

l 45

88

.56

305

8584

3 13

11

2892

25

2161

23

59

167

321

1.527

0.8

16

0.700

2.5

17

0.320

0.03

9 22

3 4

186

13

17

1 6

WPM

-THI

N OI

B (S

19)

03-1

61

Icelan

d Ki

stufel

l 50

88

.48

286

8640

8 13

26

2886

57

2157

23

41

169

317

1.535

0.8

11

0.701

2.4

97

0.303

0.04

10

263

6 21

5 8

17

1 6

WPM

-THI

N OI

B (S

19)

01-5

5 Ice

land

Theis

tarey

kir,P

icrite

s 24

90

.30

356

7331

3 11

68

2969

01

2439

24

89

144

584

1.594

0.8

38

0.615

3.3

27

0.182

0.06

19

430

9 35

6 31

33

2

21

WPM

-THI

N OI

B (S

19)

01-5

7-4

Icelan

d Th

eistar

eykir

,Picr

ites

32

90.06

37

9 74

874

1187

29

5369

23

71

2467

589

1.586

0.8

35

0.625

3.1

67

0.199

0.04

15

268

6 34

3 22

26

15

WPM

-THI

N OI

B (S

19)

01-5

6-1

Icelan

d Th

eistar

eykir

,Picr

ites

48

90.21

29

9 73

806

1189

29

5885

24

35

2423

552

1.611

0.8

19

0.604

3.2

99

0.147

0.04

10

284

6 22

7 15

18

13

WPM

-THI

N OI

B (S

19)

01-4

1 Ice

land

Theis

tarey

kir

15

89.76

43

7 77

047

1178

29

4013

21

77

2800

14

7 63

1 1.5

29

0.952

0.7

34

2.826

0.3

16

0.0

8 17

58

4 9

406

8 20

2

15

WPM

-THI

N OI

B (S

19)

01-4

4 Ice

land

Theis

tarey

kir

8 89

.16

415

8159

2 12

67

2920

04

2169

25

70

155

580

1.553

0.8

80

0.718

2.6

58

0.266

0.11

48

803

17

608

76

73

4 53

W

PM-T

HIN

OIB

(S19

) 01

-56-

2 Ice

land

Theis

tarey

kir,P

icrite

s 9

90.49

35

0 71

973

1149

29

7838

24

25

2491

15

3 58

6 1.5

96

0.836

0.6

02

3.370

0.1

77

0.1

0 30

71

6 18

41

4 43

67

3

44

WPM

-THI

N OI

B (S

19)

01-5

4 Ice

land

Theis

tarey

kir,P

icrite

s 45

87

.95

293

9033

5 14

07

2870

24

2183

22

97

172

399

1.557

0.8

00

0.723

2.4

17

0.258

0.04

9 29

8 5

226

13

14

1 8

WPM

-THI

N OI

B (S

19)

01-5

1 Ice

land

Theis

tarey

kir

6 91

.39

367

6565

0 10

35

3032

39

2433

22

87

138

885

1.577

0.7

54

0.495

3.7

07

0.217

0.15

38

1176

25

35

9 10

8 10

7 4

71

WPM

-THI

N OI

B (S

19)

03-2

24

Icelan

d Sn

aefel

lsnes

s 4

89.14

29

2 82

132

1272

29

3273

19

94

2363

15

0 51

3 1.5

49

0.806

0.6

62

2.428

0.2

75

0.2

5 12

18

23

29

1054

17

86

3

27

WPM

-THI

N OI

B (S

19)

03-2

26

Icelan

d Sn

aefel

lsnes

s 15

88

.28

232

8838

3 14

09

2896

92

2079

19

84

161

364

1.594

0.6

85

0.605

2.3

52

0.181

0.08

9 58

0 11

73

6 69

33

3

16

WPM

-THI

N OI

B (S

19)

03-2

20

Icelan

d Sn

aefel

lsnes

s 5

88.93

24

5 83

462

1344

29

1820

20

73

2067

15

6 45

8 1.6

11

0.708

0.5

91

2.483

0.1

47

0.2

1 23

14

84

17

1033

93

55

6

32

WPM

-THI

N OI

B (S

20)

SM-6

Az

ores

Sa

o Mige

l 26

87

.43

166

9461

0 14

69

2862

86

2062

17

06

161

308

1.552

0.5

96

0.564

2.1

80

0.268

0.05

3 34

1 5

243

16

19

2 5

WPM

-THI

N OI

B (S

20)

SM-1

0 Az

ores

Sa

o Mige

l 11

7 88

.58

184

8525

2 11

81

2877

85

1919

23

87

151

553

1.385

0.8

29

0.707

2.2

51

0.615

0.01

2 78

2

79

7 6

1 3

WPM

-THI

N OI

B (S

20)

SM-1

6 Az

ores

Sa

o Mige

l 12

2 89

.18

233

8108

2 10

94

2907

82

1812

23

12

149

672

1.349

0.7

95

0.645

2.2

34

0.689

0.02

2 14

8 3

100

10

26

1 3

WPM

-THI

N OI

B (S

20)

SM-3

6a

Azor

es

Sao M

igel

13

89.88

22

2 75

771

1140

29

2770

19

22

2605

13

9 52

5 1.5

04

0.890

0.6

74

2.537

0.3

68

0.1

0 9

761

12

481

26

40

3 13

W

PM-T

HIN

OIB

(S20

) SM

-36

Azor

es

Sao M

igel

75

87.91

17

0 89

944

1397

28

4715

23

57

1774

15

4 35

5 1.5

53

0.623

0.5

60

2.620

0.2

67

0.0

3 2

205

4 12

6 13

8

1 3

WPM

-THI

N OI

B (S

20)

SM-4

Az

ores

Sa

o Mige

l 4

88.20

17

7 87

865

1297

28

5844

18

87

1925

14

5 41

1 1.4

76

0.674

0.5

92

2.147

0.4

25

0.1

1 3

796

5 46

5 5

23

5 16

W

PM-T

HIN

OIB

(S20

) SM

-33

Azor

es

Sao M

igel

59

88.68

16

3 84

172

1269

28

7065

23

94

2159

14

9 41

4 1.5

08

0.752

0.6

33

2.844

0.3

59

0.0

2 2

175

3 12

1 17

14

1

3 W

PM-T

HIN

OIB

unpb

T-

18

Azor

es

Terce

ria

3 88

.45

199

8605

0 12

03

2867

61

1894

24

49

152

447

1.398

0.8

54

0.735

2.2

01

0.587

0.12

9 83

8 32

65

3 98

21

9 6

44

WPM

-THI

N OI

B un

pb

T-6

Azor

es

Terce

ria

43

88.73

24

3 84

248

1275

28

8597

18

97

2537

14

4 45

0 1.5

13

0.879

0.7

40

2.251

0.3

49

0.0

4 5

277

4 16

3 14

16

1

7 W

PM-T

HIN

OIB

(S21

) 12

03a -

038R

01W

Pa

cific

Ocea

n De

troit s

mt

57

89.53

34

4 78

587

1160

29

2482

18

37

2987

15

3 60

9 1.4

76

1.021

0.8

03

2.338

0.4

26

0.0

2 7

173

3 18

2 8

16

1 7

WPM

-THI

N OI

B (S

21)

1203

a -03

2R

Pacif

ic Oc

ean

Detro

it smt

13

89

.73

369

7692

4 11

20

2924

12

1813

30

25

151

628

1.456

1.0

34

0.796

2.3

57

0.467

0.07

14

520

9 49

5 10

32

2

16

WPM

-THI

CK

LIP

(S22

) BR

-2

Atlan

tic O

cean

Mu

ll isla

nd

17

87.25

34

4 96

138

1384

28

6280

20

00

2239

445

1.440

0.7

82

0.752

2.0

80

0.500

0.07

19

524

9 32

8 5

35

17

W

PM-T

HICK

LIP

(S

22)

BR-5

At

lantic

Oce

an

Mull i

sland

17

87

.99

421

9066

5 12

66

2890

61

1921

29

30

43

1 1.3

96

1.014

0.9

19

2.119

0.5

92

0.0

8 11

55

8 7

423

15

17

12

W

PM-T

HICK

LIP

(S

22)

BR-6

At

lantic

Oce

an

Mull i

sland

11

87

.83

435

9178

2 13

22

2881

14

1982

24

59

55

2 1.4

41

0.854

0.7

83

2.159

0.4

99

0.1

2 17

90

4 14

37

6 8

28

16

W

PM-T

HICK

LIP

(S

22)

BHL-

15

Atlan

tic O

cean

Mu

ll isla

nd

33

88.24

40

4 89

264

1333

29

1557

21

19

2386

510

1.493

0.8

18

0.730

2.3

74

0.391

0.04

7 32

6 5

230

3 17

7 W

PM-T

HICK

LIP

(S

22)

BHL-

19

Atlan

tic O

cean

Mu

ll isla

nd

16

87.83

46

3 92

074

1353

28

9015

21

51

2562

567

1.469

0.8

87

0.816

2.3

36

0.440

0.07

21

557

11

305

31

72

22

W

PM-T

HICK

LIP

(S

22)

BHL-

34

Atlan

tic O

cean

Mu

ll isla

nd

24

87.77

41

0 92

639

1335

28

9280

20

43

2655

476

1.441

0.9

18

0.850

2.2

05

0.499

0.05

10

396

6 25

9 5

22

13

W

PM-T

HICK

LIP

(S

22)

BCH-

14

Atlan

tic O

cean

Mu

ll isla

nd

10

86.50

46

1 10

1517

13

75

2829

42

1904

27

01

41

1 1.3

55

0.955

0.9

69

1.875

0.6

77

0.1

2 19

87

5 12

59

4 9

32

16

W

PM-T

HICK

LIP

(S

22)

BCH-

24

Atlan

tic O

cean

Mu

ll isla

nd

14

89.35

43

7 80

685

1195

29

4511

20

43

2912

654

1.482

0.9

89

0.798

2.5

33

0.414

0.09

6 68

8 11

49

5 16

28

14

WPM

-THI

CK

LIP

(S22

) BC

H-27

At

lantic

Oce

an

Mull i

sland

24

88

.90

426

8408

3 12

33

2929

30

2042

28

49

62

6 1.4

66

0.973

0.8

18

2.428

0.4

46

0.0

6 8

423

6 26

7 8

8

7 W

PM-T

HICK

LIP

(S

22)

BCH-

33

Atlan

tic O

cean

Mu

ll isla

nd

20

88.38

41

9 87

922

1262

29

0959

19

58

2782

550

1.435

0.9

56

0.841

2.2

27

0.510

0.06

10

434

7 29

5 7

16

9

WPM

-THI

CK

LIP

(S22

) AM

-7a

Atlan

tic O

cean

Mu

ll isla

nd

18

86.21

40

7 10

2948

14

38

2801

59

1956

23

96

32

8 1.3

97

0.855

0.8

80

1.900

0.5

90

0.0

6 13

43

0 7

303

20

18

11

W

PM-T

HICK

LIP

(S

22)

BB-2

2 At

lantic

Oce

an

Mull i

sland

12

86

.68

319

9973

1 13

45

2823

19

1621

25

49

38

5 1.3

49

0.903

0.9

00

1.626

0.6

90

0.1

2 53

78

5 19

65

4 12

1 63

32

WPM

-THI

CK

LIP

(S22

) B-

26

Atlan

tic O

cean

Mu

ll isla

nd

3 86

.66

450

9999

0 14

17

2826

39

2101

26

06

48

1 1.4

17

0.922

0.9

22

2.101

0.5

47

0.3

1 52

21

77

39

1420

33

97

75

WPM

-THI

CK

LIP

(S22

) B-

29

Atlan

tic O

cean

Mu

ll isla

nd

12

87.29

39

3 95

617

1355

28

5840

19

55

2696

448

1.417

0.9

43

0.902

2.0

44

0.547

0.09

21

613

10

532

13

23

15

W

PM-T

HICK

LIP

(S

23)

BB

Atlan

tic O

cean

Ba

ffin B

ay

1 92

.65

386

5570

9 89

8 30

5505

20

51

3450

11

8 82

8 1.6

13

1.129

0.6

29

3.682

0.1

43

0.0

0

W

PM-T

HICK

LIP

(S

24)

9597

6 At

lantic

Oce

an

Gree

nland

16

92

.05

487

6047

3 94

4 30

4500

18

94

3749

12

5 88

6 1.5

62

1.231

0.7

45

3.132

0.2

49

0.0

8 14

56

4 6

391

12

9 2

14

WPM

-THI

CK

LIP

(S24

) 27

141

Atlan

tic O

cean

Gr

eenla

nd

32

90.93

46

9 68

080

1032

29

6916

19

44

3499

13

6 84

8 1.5

15

1.178

0.8

02

2.856

0.3

44

0.0

5 14

39

0 4

286

11

14

1 16

W

PM-T

HICK

LIP

(S

24)

G-27

142

Atlan

tic O

cean

Gr

eenla

nd

21

91.06

37

9 67

782

1062

30

0460

18

72

3128

14

3 72

5 1.5

67

1.041

0.7

06

2.762

0.2

36

0.0

7 8

500

7 27

7 16

24

2

12

WPM

-THI

CK

LIP

unpb

Di

s-1

Atlan

tic O

cean

Di

sko

75

88.50

31

0 86

270

1329

28

8974

20

28

3006

468

1.541

1.0

40

0.897

2.3

50

0.292

0.03

6 21

7 3

142

8 6

5

WPM

-THI

CK*

LIP

(S10

) 11

85A

9R 2

92-9

6 Pa

cific

Ocea

n On

tong J

ava P

latea

u 80

87

.36

250

9424

2 14

63

2835

14

2121

21

56

167

377

1.553

0.7

60

0.717

2.2

50

0.267

0.02

3 15

7 3

124

6 7

1 3

WPM

-THI

CK*

LIP

(S10

) 11

87A

10R

7 46-

49

Pacif

ic Oc

ean

Onton

g Jav

a Plat

eau

15

88.76

26

9 83

907

1115

28

8325

15

84

3313

15

2 74

3 1.3

28

1.149

0.9

64

1.887

0.7

31

0.0

8 6

604

8 36

2 15

48

2

30

WPM

-THI

CK*

LIP

(S10

) 11

87A

6R 6

116-

119

Pacif

ic Oc

ean

Onton

g Jav

a Plat

eau

4 89

.33

292

7992

8 12

02

2913

33

1910

24

48

157

506

1.504

0.8

40

0.672

2.3

90

0.367

0.07

25

377

7 70

4 32

70

3

41

WPM

-THI

CK

LIP

(S25

) P2

3-9

Afric

a Ka

roo

21

88.28

33

3 87

816

1245

28

7906

16

23

2681

15

5 83

6 1.4

17

0.931

0.8

18

1.848

0.5

27

0.0

4 8

491

7 33

9 11

44

2

31

WPM

-THI

CK

LIP

(S25

) P2

3-32

Af

rica

Karo

o 25

89

.49

361

7908

0 11

38

2930

22

1632

28

22

149

1106

1.4

39

0.963

0.7

61

2.063

0.4

84

0.0

5 7

303

4 24

0 9

11

2 24

W

PM-T

HICK

LIP

(S

25)

N356

Af

rica

Karo

o 29

90

.42

124

7208

2 92

2 29

6170

10

84

4392

486

1.279

1.4

83

1.069

1.5

04

0.835

0.05

4 39

0 5

276

9 9

11

W

PM-T

HICK

LIP

(S

26)

EM-4

3 Ch

ina

Emeis

han

26

90.47

25

6 72

181

1286

29

8304

26

12

2566

15

5 47

0 1.7

81

0.860

0.6

21

3.619

-0

.207

0.0

4 14

28

4 5

301

39

18

1 19

W

PM-T

HICK

LIP

(S

27)

EM-5

5 Ch

ina

Emeis

han

5 90

.59

213

7131

0 10

36

2988

30

2084

34

18

55

1 1.4

53

1.144

0.8

16

2.922

0.4

73

0.2

6 5

1869

29

12

61

163

80

24

W

PM-T

HICK

LIP

(S

27)

EM-5

7 Ch

ina

Emeis

han

4 90

.69

198

7101

3 10

20

3010

66

1728

33

99

48

7 1.4

37

1.129

0.8

02

2.433

0.5

07

0.4

0 27

30

14

41

1492

38

19

9

51

WPM

-THI

CK

LIP

(S27

) EM

-58

China

Em

eisha

n 4

90.76

20

8 69

952

1001

29

8979

17

96

3548

523

1.431

1.1

87

0.830

2.5

67

0.519

0.35

22

2640

36

11

04

186

143

60

W

PM-T

HICK

LIP

un

pb

4270

Si

beria

, Nor

ilsk

Gudc

hikhin

skay

a unit

15

79

.90

170

1466

01

1770

25

3513

19

14

2942

299

1.207

1.1

61

1.702

1.3

06

0.982

0.09

3 60

1 9

371

8 28

7 W

PM-T

HICK

LIP

un

pb

XC51

-130

Si

beria

, Nor

ilsk

Gudc

hikhin

skay

a unit

7

82.41

16

0 12

8631

16

19

2622

06

1698

30

16

37

9 1.2

58

1.150

1.4

80

1.320

0.8

77

0.1

6 6

1071

18

66

0 25

24

12

WPM

-THI

CK

LIP

unpb

XC

51-1

30(n

) Si

beria

, Nor

ilsk

Gudc

hikhin

skay

a unit

9

82.74

13

0 12

6790

15

82

2645

03

1805

31

02

200

383

1.247

1.1

73

1.487

1.4

24

0.899

0.10

5 74

7 11

46

6 12

18

2

19

Sobo

lev

et a

l, Su

ppor

ting

Onl

ine

Mat

eria

l ww

w.sc

ienc

emag

.org

/cgi

/con

tent

/full/

1138

113/

DC

1

20 A

PRIL

200

7 VO

L 31

6 SC

IEN

CE

ww

w.s

cien

cem

ag.o

rg

S7

GROU

P Ge

odyn

amic

settin

g Re

feren

ces

Samp

le Ge

ogra

phic

Lo

cality

N

Fo

Al pp

m Fe

ppm

Mn pp

m Mg

ppm

Ca pp

m Ni

ppm

Co pp

m Cr

ppm

100*

Mn/F

e 10

0*Ni

/Mg

Ni/(M

g/Fe)

/1000

10

0Ca/F

e X

px M

n/Fe

STE

Fo w

t% A

l ppm

Fe

ppm

Mn pp

m Mg

ppm

Ca pp

m Ni

ppm

Co pp

m Cr

ppm

WPM

-THI

CK

LIP

unpb

SU

-50

Sibe

ria, N

orils

k Gu

dchik

hinsk

aya u

nit

22

80.80

16

3 14

0093

16

91

2565

05

1930

30

17

32

7 1.2

07

1.176

1.6

48

1.378

0.9

83

0.0

7 3

501

6 34

1 7

19

6

WPM

-THI

CK

LIP

(S28

) AK

-31

Afric

a, Af

ar

Zawl

-Kelk

elti, E

mba T

eker

a, Er

itrea

5

87.10

37

4 96

459

1382

28

3503

19

33

2112

16

9 37

1 1.4

32

0.745

0.7

19

2.003

0.5

16

0.1

6 31

11

49

17

711

16

58

2 22

W

PM-T

HICK

LIP

(S

28)

AK-3

3 Af

rica,

Afar

Za

wl-K

elkelt

i, Emb

a Tek

era,

Eritr

ea

9 89

.34

475

7997

2 11

86

2917

58

1891

27

12

158

480

1.483

0.9

29

0.743

2.3

64

0.412

0.10

17

711

9 42

1 8

26

3 12

W

PM-T

HICK

LIP

(S

28)

AG-1

0 Af

rica,

Afar

Ad

i Ghe

bray

, Emb

a Tek

era,

Eritre

a 1

88.30

51

3 87

011

1286

28

5738

19

08

2562

15

4 49

3 1.4

77

0.897

0.7

80

2.193

0.4

23

0.0

0

W

PM-T

HICK

LIP

(S

28)

AG-1

5 Af

rica,

Afar

Ad

i Ghe

bray

, Emb

a Tek

era,

Eritre

a 17

84

.66

358

1127

74

1530

27

0733

17

43

2154

18

2 29

5 1.3

56

0.795

0.8

97

1.546

0.6

73

0.0

7 8

483

8 25

9 8

35

3 15

W

PM-T

HICK

OI

B (S

29)

RY-7

Ind

ian O

cean

Re

union

, Pito

n de

Neige

73

85

.86

224

1054

45

1567

27

8606

19

97

2377

329

1.486

0.8

53

0.900

1.8

94

0.405

0.02

3 14

9 3

104

11

5

3 W

PM-T

HICK

OI

B (S

29)

RY-1

1 Ind

ian O

cean

Re

union

, Pito

n de

Neige

13

89

.82

262

7702

8 11

25

2957

86

1675

29

70

53

4 1.4

60

1.004

0.7

74

2.174

0.4

59

0.0

8 9

545

11

444

28

38

12

W

PM-T

HICK

OI

B (S

29)

OPF

Indian

Oce

an

Reun

ion, P

iton d

e la F

ourn

aise

82

83.98

18

5 11

8585

16

96

2704

79

1988

20

16

24

3 1.4

30

0.745

0.8

84

1.677

0.5

21

0.0

3 2

186

3 10

4 5

3

2 W

PM-T

HICK

OI

B un

pb

IKI-2

2 Ha

waii

Kilau

ea Ik

i 8

88.28

25

5 87

957

1205

28

8325

17

80

3059

16

1 60

0 1.3

70

1.061

0.9

33

2.023

0.6

45

0.1

3 16

95

7 12

44

1 10

49

3

17

WPM

-THI

CK

OIB

unpb

IK

I-44

Hawa

ii Ki

lauea

Iki

20

88.54

28

8 85

808

1176

28

8577

17

42

3210

15

8 61

4 1.3

71

1.112

0.9

54

2.030

0.6

43

0.0

5 4

368

5 31

9 12

33

2

6 W

PM-T

HICK

OI

B un

pb

K97-

12

Hawa

ii Ki

lauea

Iki

34

87.81

23

6 91

085

1259

28

5538

18

68

2915

15

9 54

4 1.3

82

1.021

0.9

30

2.050

0.6

21

0.0

5 6

365

6 20

7 20

17

1

7 W

PM-T

HICK

OI

B un

pb

K97-

14b

Hawa

ii Mo

una L

oa, P

uu W

ahi

67

88.41

25

5 86

630

1195

28

7459

13

76

3023

16

1 60

7 1.3

79

1.051

0.9

11

1.589

0.6

27

0.0

3 15

4 4

220

7 14

1

7 24

7 W

PM-T

HICK

OI

B (S

30)

SR-1

7-9.0

Ha

waii

Maun

a Loa

, HSD

P-2

31

90.22

26

4 73

484

998

2949

88

1352

33

00

165

663

1.357

1.1

19

0.822

1.8

40

0.671

0.02

2 13

4 2

131

2 26

1

15

WPM

-THI

CK

OIB

(S30

) SR

-18-

0 Ha

waii

Maun

a Loa

, HSD

P-2

19

90.53

25

7 71

919

994

2993

11

1359

39

18

69

2 1.3

82

1.309

0.9

41

1.889

0.6

20

0.0

8 11

57

8 8

279

6 37

27

WPM

-THI

CK

OIB

(S30

) SR

-31-

1.0-2

.0 Ha

waii

Maun

a Loa

, HSD

P-2

7 89

.32

196

8135

6 11

33

2959

86

1791

28

83

58

4 1.3

93

0.974

0.7

92

2.201

0.5

99

0.1

3 18

93

2 20

67

5 11

8 14

3

14

WPM

-THI

CK

OIB

(S30

) SR

-98-

2.5-3

.0 Ha

waii

Maun

a Loa

, HSD

P-2

31

89.94

27

8 76

622

1031

29

7952

14

28

3928

710

1.346

1.3

18

1.010

1.8

63

0.695

0.05

7 39

4 5

275

6 49

18

WPM

-THI

CK

OIB

(S30

) SR

-98-

2.5-3

.0(n)

Ha

waii

Muan

a Loa

, HSD

P2

60

89.75

29

9 76

967

1045

29

3323

13

98

3830

15

1 73

6 1.3

58

1.306

1.0

05

1.817

0.6

70

0.0

2 4

172

3 16

5 7

26

1 11

W

PM-T

HICK

OI

B (S

30)

SR-1

01-7

.0 Ha

waii

Maun

a Loa

, HSD

P-2

50

90.00

31

8 74

911

1035

29

3507

13

85

3727

15

0 78

4 1.3

82

1.270

0.9

51

1.849

0.6

21

0.0

3 4

188

3 12

5 3

24

1 6

WPM

-THI

CK

OIB

(S30

) SR

-104

-5.5

Hawa

ii Ma

una L

oa, H

SDP-

2 33

89

.89

276

7700

6 10

54

2979

65

1405

37

00

74

5 1.3

68

1.242

0.9

56

1.824

0.6

49

0.0

5 4

387

6 21

7 3

21

9

WPM

-THI

CK

OIB

(S30

) SR

-104

-5.5(

n)

Hawa

ii Mu

ana L

oa, H

SDP2

70

89

.77

287

7719

7 10

57

2947

44

1413

37

29

144

735

1.370

1.2

65

0.977

1.8

30

0.646

0.03

2 20

3 3

149

3 13

1

8 W

PM-T

HICK

OI

B (S

30)

SR-1

13-6

.5 Ha

waii

Maun

a Loa

, HSD

P-2

63

89.58

27

3 77

943

1057

29

1517

14

09

3702

568

1.357

1.2

70

0.990

1.8

08

0.673

0.03

3 21

0 2

153

2 10

5 W

PM-T

HICK

OI

B (S

30)

SR11

7-4.0

Ha

waii

Maun

a Loa

, HSD

P-2

13

89.67

30

5 76

422

1046

28

8664

13

97

3680

836

1.368

1.2

75

0.974

1.8

28

0.649

0.11

10

782

8 37

3 6

39

15

W

PM-T

HICK

OI

B (S

30)

SR-1

20-2

.0-2.5

Ha

waii

Maun

a Loa

, HSD

P-2

35

89.65

25

8 78

973

1075

29

7548

14

29

3632

732

1.361

1.2

21

0.964

1.8

09

0.664

0.04

6 29

6 5

203

4 22

8 W

PM-T

HICK

OI

B (S

30)

SR-1

20-2

.0-2.5

(n)

Hawa

ii Mu

ana L

oa, H

SDP2

40

89

.30

275

8046

2 10

85

2923

28

1453

35

65

150

1.3

49

1.220

0.9

81

1.806

0.6

89

0.0

4 5

305

5 19

7 4

19

1 0

WPM

-THI

CK

OIB

(S30

) SR

-133

-10.0

Ha

waii

Maun

a Kea

, HSD

P2

7 87

.33

211

9519

3 13

10

2854

97

2011

28

15

42

8 1.3

76

0.986

0.9

39

2.113

0.6

33

0.1

3 9

942

11

649

23

70

12

W

PM-T

HICK

OI

B (S

30)

SR-1

34-2

.6 Ha

waii

Maun

a Kea

, HSD

P2

25

86.71

21

2 98

879

1354

28

0742

20

66

2725

17

2 40

6 1.3

70

0.971

0.9

60

2.089

0.6

46

0.0

6 6

397

6 27

0 19

13

2

9 W

PM-T

HICK

OI

B (S

30)

SR-1

36-6

.1 Ha

waii

Maun

a Kea

, HSD

P2

16

89.20

21

0 82

042

1101

29

4758

15

37

3140

630

1.342

1.0

65

0.874

1.8

73

0.703

0.08

6 55

0 8

426

21

63

12

W

PM-T

HICK

OI

B (S

30)

SR-1

38-6

.5 Ha

waii

Maun

a Kea

, HSD

P2

31

89.28

26

3 80

686

1096

29

2513

15

31

3197

15

7 66

7 1.3

59

1.093

0.8

82

1.898

0.6

68

0.0

5 5

347

6 20

6 9

33

1 8

WPM

-THI

CK

OIB

(S30

) SR

-139

-9.5

Hawa

ii Ma

una K

ea, H

SDP2

39

89

.05

243

8233

8 11

25

2912

74

1666

30

77

69

8 1.3

66

1.056

0.8

70

2.024

0.6

53

0.0

5 5

349

5 25

3 10

22

6 W

PM-T

HICK

OI

B (S

30)

SR-1

44-2

.0 Ha

waii

Maun

a Kea

, HSD

P2

23

88.58

16

8 86

738

1181

29

2769

17

19

2961

464

1.361

1.0

12

0.877

1.9

82

0.664

0.07

5 47

0 8

323

20

20

14

W

PM-T

HICK

OI

B (S

30)

SR-1

45-2

.5-2.6

Ha

waii

Maun

a Kea

, HSD

P2

39

88.55

25

3 86

149

1170

29

0005

17

28

2962

16

4 65

6 1.3

59

1.021

0.8

80

2.005

0.6

69

0.0

5 4

340

5 18

9 9

11

1 7

WPM

-THI

CK

OIB

(S30

) SR

-148

-8.5

Hawa

ii Ma

una K

ea, H

SDP2

19

88

.00

201

9007

4 12

41

2874

69

1877

26

79

55

2 1.3

77

0.932

0.8

39

2.083

0.6

30

0.0

8 7

561

9 38

5 29

21

13

WPM

-THI

CK

OIB

(S30

) SR

-149

-1.0

Hawa

ii Ma

una K

ea, H

SDP2

31

87

.92

190

9145

4 12

58

2895

28

1908

26

55

54

8 1.3

75

0.917

0.8

39

2.086

0.6

34

0.0

6 8

414

7 26

7 14

19

7 W

PM-T

HICK

OI

B (S

30)

SR-1

57-6

.1 Ha

waii

Maun

a Kea

, HSD

P2

24

89.11

19

6 82

064

1126

29

2116

16

40

3079

596

1.372

1.0

54

0.865

1.9

99

0.641

0.06

4 42

3 7

272

21

23

13

W

PM-T

HICK

OI

B (S

30)

SR-1

59-2

.6 Ha

waii

Maun

a Kea

, HSD

P2

42

88.75

24

4 84

807

1154

29

1029

16

48

3036

15

9 64

3 1.3

60

1.043

0.8

85

1.943

0.6

65

0.0

4 4

288

3 21

1 9

14

1 8

WPM

-THI

CK

OIB

(S30

) SR

-170

-1.2

Hawa

ii Ma

una K

ea, H

SDP2

19

88

.30

213

8827

1 11

93

2897

85

1648

29

40

62

7 1.3

51

1.014

0.8

95

1.867

0.6

85

0.0

7 8

530

8 34

7 20

40

16

WPM

-THI

CK

OIB

(S30

) SR

-170

-3.1

Hawa

ii Ma

una K

ea, H

SDP2

46

87

.98

218

8982

2 12

17

2861

90

1668

29

33

62

9 1.3

55

1.025

0.9

21

1.857

0.6

77

0.0

4 5

282

4 18

5 11

16

7 W

PM-T

HICK

OI

B (S

30)

SR-1

75-5

.2 Ha

waii

Maun

a Kea

, HSD

P2

36

88.95

24

1 83

748

1123

29

3387

15

46

3210

696

1.340

1.0

94

0.916

1.8

45

0.707

0.05

4 35

2 6

214

16

26

11

W

PM-T

HICK

OI

B (S

30)

SR-1

84-3

.0 Ha

waii

Maun

a Kea

, HSD

P2

4 89

.35

201

8022

2 11

15

2930

05

1746

30

96

165

513

1.390

1.0

57

0.848

2.1

76

0.604

0.29

10

2119

9

1187

81

8

6 9

WPM

-THI

CK

OIB

(S30

) SR

-188

-6.5

Hawa

ii Ma

una K

ea, H

SDP2

6

89.55

24

0 78

725

1053

29

3534

15

41

3340

15

2 55

6 1.3

38

1.138

0.8

96

1.958

0.7

12

0.1

5 10

11

09

25

654

48

213

6 30

W

PM-T

HICK

OI

B (S

30)

SR-2

60-8

.5 Ha

waii

Maun

a Kea

, HSD

P2

5 89

.73

267

7739

6 10

04

2941

14

1434

38

47

134

590

1.297

1.3

08

1.012

1.8

52

0.797

0.13

9 98

0 11

56

7 14

53

6

14

WPM

-THI

CK

OIB

(S30

) SR

-277

-7.5-

8.0

Hawa

ii Ma

una K

ea, H

SDP2

2

90.86

19

8 69

347

906

2998

34

1361

36

93

146

380

1.307

1.2

32

0.854

1.9

63

0.777

0.14

13

999

0 70

9 11

55

1

38

WPM

-THI

CK

OIB

(S30

) SR

-291

-3.0-

3.8

Hawa

ii Ma

una K

ea, H

SDP2

38

89

.24

201

8111

4 11

15

2926

28

1645

33

27

149

603

1.374

1.1

37

0.922

2.0

28

0.637

0.04

7 31

6 4

210

12

18

1 12

W

PM-T

HICK

OI

B (S

30)

SR-2

98-6

.0 Ha

waii

Maun

a Kea

, HSD

P2

27

89.03

24

7 82

446

1102

29

1201

15

75

3328

14

9 66

4 1.3

37

1.143

0.9

42

1.910

0.7

15

0.0

6 6

424

8 23

6 10

49

2

11

WPM

-THI

CK

OIB

(S30

) SR

-308

-0-0

.6 Ha

waii

Maun

a Kea

, HSD

P2

12

89.10

27

7 81

993

1099

29

1738

15

57

3478

15

1 63

9 1.3

40

1.192

0.9

77

1.899

0.7

07

0.1

3 6

952

17

611

26

108

3 8

WPM

-THI

CK

OIB

(S30

) SR

-310

-2.4-

2.8

Hawa

ii Ma

una K

ea, H

SDP2

70

89

.51

265

7943

7 10

82

2947

97

1548

33

05

150

742

1.362

1.1

21

0.891

1.9

49

0.662

0.03

3 19

2 3

130

2 14

1

5 W

PM-T

HICK

OI

B (S

30)

SR-3

19-7

.5 Ha

waii

Maun

a Kea

, HSD

P2

42

89.02

25

7 82

543

1097

29

1302

15

51

3384

15

3 68

0 1.3

29

1.162

0.9

59

1.879

0.7

30

0.0

4 5

316

5 20

2 5

26

1 10

W

PM-T

HICK

OI

B (S

30)

SR-3

32-9

.8-10

.0 Ha

waii

Maun

a Kea

, HSD

P2

7 90

.31

222

7333

6 98

1 29

7371

14

76

4121

13

7 48

7 1.3

38

1.386

1.0

16

2.013

0.7

11

0.1

4 9

1015

15

62

3 15

11

2 2

21

WPM

-THI

CK

OIB

(S30

) SR

-335

-3.2-

3.4

Hawa

ii Ma

una K

ea, H

SDP2

55

89

.89

286

7596

6 10

37

2937

64

1500

33

45

149

785

1.365

1.1

39

0.865

1.9

75

0.656

0.04

4 26

8 4

201

4 25

1

9 W

PM-T

HICK

OI

B (S

30)

SR-3

37-3

.4-3.7

Ha

waii

Maun

a Kea

, HSD

P2

58

89.78

26

5 76

896

1041

29

4088

15

04

3344

15

0 74

0 1.3

54

1.137

0.8

74

1.956

0.6

78

0.0

4 5

318

4 18

7 6

25

1 10

W

PM-T

HICK

OI

B (S

30)

SR-3

53-7

.2-7.4

Ha

waii

Maun

a Kea

, HSD

P2

62

88.01

22

9 89

908

1210

28

7179

16

60

3108

16

5 55

3 1.3

45

1.082

0.9

73

1.846

0.6

97

0.0

3 3

236

3 17

7 9

22

1 6

WPM

-THI

CK

OIB

(S30

) SR

-361

-2.5-

2.8

Hawa

ii Ma

una K

ea, H

SDP2

26

89

.66

231

7798

5 10

50

2942

11

1498

35

12

148

602

1.346

1.1

94

0.931

1.9

21

0.695

0.04

5 28

6 4

173

5 12

1

11

WPM

-THI

CK

OIB

(S30

) SR

-363

-7.0-

8.0

Hawa

ii Ma

una K

ea, H

SDP2

57

89

.03

228

8275

1 11

07

2923

65

1584

33

46

151

603

1.337

1.1

44

0.947

1.9

14

0.713

0.03

6 22

6 4

155

4 14

1

13

WPM

-THI

CK

OIB

(S30

) SR

-375

-0.5-

1.0

Hawa

ii Ma

una K

ea, H

SDP2

20

89

.65

243

7802

7 10

48

2942

22

1534

34

68

146

738

1.343

1.1

79

0.920

1.9

67

0.702

0.06

8 44

4 6

244

6 32

1

11

WPM

-THI

CK

OIB

(S30

) SR

-377

-6.7-

6.9

Hawa

ii Ma

una K

ea, H

SDP2

38

89

.36

276

7998

9 10

60

2923

96

1534

36

17

146

731

1.325

1.2

37

0.990

1.9

18

0.739

0.04

5 25

7 3

210

11

8 1

11

WPM

-THI

CK

OIB

(S30

) SR

-382

-4.0-

4.4

Hawa

ii Ma

una K

ea, H

SDP2

19

88

.26

238

8764

8 11

66

2868

14

1619

32

56

156

660

1.330

1.1

35

0.995

1.8

47

0.728

0.08

11

561

8 32

0 9

28

2 24

Sobo

lev

et a

l, Su

ppor

ting

Onl

ine

Mat

eria

l ww

w.sc

ienc

emag

.org

/cgi

/con

tent

/full/

1138

113/

DC

1

20 A

PRIL

200

7 VO

L 31

6 SC

IEN

CE

ww

w.s

cien

cem

ag.o

rg

S8

GROU

P Ge

odyn

amic

settin

g Re

feren

ces

Samp

le Ge

ogra

phic

Lo

cality

N

Fo

Al pp

m Fe

ppm

Mn pp

m Mg

ppm

Ca pp

m Ni

ppm

Co pp

m Cr

ppm

100*

Mn/F

e 10

0*Ni

/Mg

Ni/(M

g/Fe)

/1000

10

0Ca/F

e X

px M

n/Fe

STE

Fo w

t% A

l ppm

Fe

ppm

Mn pp

m Mg

ppm

Ca pp

m Ni

ppm

Co pp

m Cr

ppm

WPM

-THI

CK

OIB

(S30

) SR

-389

-1.8-

2.0

Hawa

ii Ma

una K

ea, H

SDP2

35

89

.76

303

7720

7 10

41

2944

10

1505

34

25

147

795

1.349

1.1

63

0.898

1.9

49

0.690

0.04

2 29

7 4

232

3 42

2

11

WPM

-THI

CK

OIB

(S30

) SR

-565

-0.7

Hawa

ii Ma

una K

ea, H

SDP2

4

90.28

28

7 73

350

962

2963

80

1261

41

17

149

547

1.312

1.3

89

1.019

1.7

20

0.766

0.01

8 20

5 5

610

12

12

6 3

WPM

-THI

CK

OIB

(S30

) SR

-682

-9.8-

10

Hawa

ii Ma

una K

ea, H

SDP2

50

88

.70

289

8469

0 11

38

2893

80

1576

32

12

158

723

1.343

1.1

10

0.940

1.8

61

0.701

0.03

4 24

2 4

163

6 17

1

5 W

PM-T

HICK

OI

B (S

30)

SR-6

84-5

.0-5.5

Ha

waii

Maun

a Kea

, HSD

P2

28

88.91

28

4 83

330

1111

29

0795

15

50

3209

16

1 73

1 1.3

33

1.103

0.9

20

1.860

0.7

23

0.0

4 4

274

4 18

4 6

31

2 8

WPM

-THI

CK

OIB

(S30

) SR

-690

-4.0

Hawa

ii Ma

una K

ea, H

SDP2

16

89

.67

307

7786

9 10

32

2941

48

1587

36

51

150

723

1.325

1.2

41

0.967

2.0

38

0.738

0.08

4 62

2 5

293

10

18

3 10

W

PM-T

HICK

OI

B (S

30)

SR-6

94-3

.0-3.2

Ha

waii

Maun

a Kea

, HSD

P2

2 89

.01

278

8282

2 11

04

2920

28

1601

35

44

157

770

1.332

1.2

14

1.005

1.9

33

0.723

0.07

13

428

4 52

5 7

63

6 65

W

PM-T

HICK

OI

B (S

30)

SR-9

30-9

.1 Ha

waii

Maun

a Kea

, HSD

P2

10

89.28

27

6 80

768

1122

29

2698

15

45

3395

15

5 69

5 1.3

89

1.160

0.9

37

1.913

0.6

05

0.1

0 9

746

10

575

27

85

3 27

W

PM-T

HICK

OI

B (S

30)

SR-9

42-5

.7 Ha

waii

Maun

a Kea

, HSD

P2

40

88.14

22

9 88

920

1206

28

7592

14

68

3102

16

4 60

1 1.3

56

1.079

0.9

59

1.651

0.6

74

0.0

4 5

267

4 19

0 7

10

2 8

WPM

-THI

CK

OIB

(S30

) SR

-954

-10.7

Ha

waii

Maun

a Kea

, HSD

P2

48

89.52

28

9 78

227

1059

29

0821

14

87

3631

14

7 77

9 1.3

53

1.249

0.9

77

1.900

0.6

80

0.0

3 4

212

3 15

3 2

13

1 5

WPM

-THI

CK

OIB

(S30

) SR

-961

-14.0

Ha

waii

Maun

a Kea

, HSD

P2

44

89.44

26

1 79

445

1082

29

2711

15

06

3537

765

1.362

1.2

08

0.960

1.8

95

0.662

0.04

4 31

1 4

195

3 29

7 W

PM-T

HICK

OI

B (S

30)

SR-9

62-1

8.1

Hawa

ii Ma

una K

ea, H

SDP2

66

89

.51

284

7902

9 10

75

2933

75

1488

35

80

152

773

1.361

1.2

20

0.964

1.8

82

0.665

0.03

4 20

8 3

147

3 15

1

5 W

PM-T

HICK

OI

B (S

31)

KI-1

1 Ha

waii

Maun

a Kea

post

shiel

d 18

86

.98

132

9816

6 14

23

2854

90

2424

22

33

36

6 1.4

50

0.782

0.7

68

2.470

0.4

80

0.0

7 6

472

8 30

9 20

14

9 W

PM-T

HICK

OI

B (S

31)

KI-8

Ha

waii

Maun

a Kea

post

shiel

d 9

88.26

18

1 89

202

1229

29

1934

18

32

2448

536

1.378

0.8

39

0.748

2.0

54

0.630

0.12

8 81

8 14

60

4 30

60

10

WPM

-THI

CK

OIB

(S31

) KI

-4

Hawa

ii Ma

una K

ea po

st sh

ield

11

87.74

13

9 92

943

1295

28

9526

20

43

2588

446

1.393

0.8

94

0.831

2.1

98

0.598

0.11

9 79

1 11

49

9 68

30

18

WPM

-THI

CK

OIB

(S31

) KI

-3

Hawa

ii Ma

una K

ea po

st sh

ield

8 88

.79

251

8497

9 11

50

2927

85

1590

29

16

65

4 1.3

53

0.996

0.8

46

1.871

0.6

80

0.1

2 11

90

1 17

59

5 29

10

0

13

WPM

-THI

CK

OIB

(S31

) KI

-3(n

) Ha

waii

Maun

a Kea

, pos

t shie

ld 5

88.82

23

4 84

025

1145

29

0374

15

62

2951

15

9 64

2 1.3

62

1.016

0.8

54

1.859

0.6

61

0.2

1 11

15

10

28

1002

68

13

4 3

20

WPM

-THI

CK

OIB

(S32

) Ko

o-10

Ha

waii

Koola

u, Ma

kapu

u 7

87.13

17

9 96

799

1263

28

5148

12

74

2967

479

1.305

1.0

40

1.007

1.3

16

0.780

0.14

7 10

30

10

495

8 24

8 W

PM-T

HICK

OI

B (S

32)

Koo-

17a

Hawa

ii Ko

olau,

Maka

puu

19

89.11

24

1 82

175

1085

29

2709

11

99

3910

613

1.321

1.3

36

1.098

1.4

59

0.747

0.08

6 54

1 11

41

7 13

98

12

WPM

-THI

CK

OIB

(S32

) Ko

o-19

a Ha

waii

Koola

u, Ma

kapu

u 43

88

.82

248

8427

6 11

11

2912

60

1257

37

07

61

0 1.3

18

1.273

1.0

73

1.492

0.7

53

0.0

4 4

275

5 17

0 8

36

6

WPM

-THI

CK

OIB

(S32

) Ko

o-30

a Ha

waii

Koola

u, Ma

kapu

u 2

86.74

18

0 99

604

1348

28

3591

13

54

3505

404

1.353

1.2

36

1.231

1.3

60

0.681

0.04

11

459

8 36

2 32

47

21

WPM

-THI

CK

OIB

(S32

) Ko

o-49

Ha

waii

Koola

u, Ma

kapu

u 34

89

.00

235

8300

5 10

96

2921

94

1272

38

03

57

9 1.3

20

1.302

1.0

80

1.532

0.7

48

0.0

3 5

227

6 21

2 7

37

10

W

PM-T

HICK

OI

B (S

32)

Koo-

55

Hawa

ii Ko

olau,

Maka

puu

41

88.52

21

5 87

012

1156

29

2015

12

94

3584

557

1.329

1.2

27

1.068

1.4

88

0.731

0.03

5 21

3 4

165

6 18

7 W

PM-T

HICK

OI

B un

pb

S10(

i) Ha

waii

Koola

u, Ma

kapu

u 23

86

.78

194

9874

5 12

89

2819

59

1239

31

29

171

475

1.306

1.1

10

1.096

1.2

54

0.751

0.04

3 30

7 5

213

6 35

1

6 W

PM-T

HICK

OI

B un

pb

S10(

ii) Ha

waii

Koola

u, Ma

kapu

u 2

87.86

23

0 90

851

1142

28

6085

11

94

4503

16

9 48

2 1.2

57

1.574

1.4

30

1.314

0.8

79

0.3

5 3

2433

12

16

67

21

149

5 24

W

PM-T

HICK

OI

B (S

33)

R8-1

.1-11

97.1

Hawa

ii Ko

olau,

KSDP

23

88

.67

248

8515

7 11

36

2899

45

1366

35

16

152

620

1.334

1.2

13

1.033

1.6

04

0.720

0.04

6 29

9 5

323

14

49

2 7

WPM

-THI

CK

OIB

(S34

) LO

-02-

04

Hawa

ii Lo

ihi sm

t 28

87

.70

236

9199

3 12

83

2853

14

1984

27

53

52

8 1.3

94

0.965

0.8

88

2.156

0.5

95

0.0

5 7

395

7 24

8 22

29

10

WPM

-THI

CK

OIB

(S34

) LO

-02-

02

Hawa

ii Lo

ihi sm

t 19

88

.29

219

8815

2 12

33

2890

83

2036

28

11

54

2 1.3

99

0.972

0.8

57

2.309

0.5

85

0.0

7 5

515

8 28

2 66

35

17

WPM

-THI

CK

OIB

(S35

) 15

8-9

Hawa

ii Lo

ihi sm

t 44

89

.15

239

8197

7 11

66

2929

71

1781

31

00

65

9 1.4

23

1.058

0.8

67

2.172

0.5

36

0.0

3 5

193

3 13

1 15

12

6 W

PM-T

HICK

OI

B (S

35)

186-

5 Ha

waii

Loihi

smt

3 89

.43

309

7963

7 11

31

2933

26

2015

28

79

71

2 1.4

20

0.981

0.7

82

2.531

0.5

42

0.3

4 14

24

15

45

1594

84

12

3

51

WPM

-THI

CK

OIB

(S35

) 18

7-1

Hawa

ii Lo

ihi sm

t 7

89.67

30

3 78

190

1095

29

5195

17

74

3242

679

1.401

1.0

98

0.859

2.2

69

0.582

0.12

16

898

14

488

36

121

29

W

PM-T

HICK

OI

B (S

36)

M234

3-7

Hawa

ii Lo

ihi sm

t 9

88.14

21

5 89

237

1319

28

8692

21

94

2742

483

1.478

0.9

50

0.848

2.4

59

0.421

0.13

17

973

14

586

26

11

12

W

PM-T

HICK

OI

B (S

36)

M233

5-12

Ha

waii

Loihi

smt

13

89.44

25

8 79

807

1143

29

4008

18

87

3091

683

1.432

1.0

51

0.839

2.3

65

0.516

0.08

11

619

11

378

29

28

23

W

PM-T

HICK

OI

B (S

36)

M234

0-3

Hawa

ii Lo

ihi sm

t 3

87.56

22

2 93

624

1291

28

6763

19

11

3023

511

1.379

1.0

54

0.987

2.0

41

0.628

0.32

26

2393

37

12

03

88

156

39

W

PM-T

HICK

OI

B (S

36)

M234

3-10

Ha

waii

Loihi

smt

37

87.16

21

0 95

978

1382

28

3538

22

72

2781

350

1.440

0.9

81

0.941

2.3

67

0.500

0.03

5 18

6 4

157

11

8

8 W

PM-T

HICK

OI

B (S

36)

M234

3-8

Hawa

ii Lo

ihi sm

t 5

88.16

21

1 88

961

1295

28

8238

21

80

2804

434

1.456

0.9

73

0.865

2.4

50

0.468

0.13

37

914

11

569

51

9

36

WPM

-THI

CK

OIB

(S36

) M2

340-

1 Ha

waii

Loihi

smt

3 88

.21

279

8834

6 12

21

2876

78

1899

29

62

62

0 1.3

82

1.030

0.9

10

2.149

0.6

20

0.1

7 40

13

40

14

269

43

99

28

W

PM-T

HICK

OI

B (S

36)

MK23

37-4

Ha

waii

Loihi

smt

19

87.11

22

4 96

131

1351

28

2756

18

26

2823

468

1.405

0.9

98

0.960

1.9

00

0.573

0.07

10

524

10

364

43

18

11

W

PM-T

HICK

OI

B (S

36)

M234

3-9

Hawa

ii Lo

ihi sm

t 17

87

.09

228

9672

6 13

58

2839

61

2223

25

77

44

5 1.4

04

0.908

0.8

78

2.298

0.5

76

0.0

8 6

585

12

312

64

40

10

W

PM-T

HICK

OI

B (S

36)

M234

3-5

Hawa

ii Lo

ihi sm

t 3

88.87

21

7 83

863

1242

29

1338

22

11

2792

504

1.481

0.9

58

0.804

2.6

36

0.416

0.30

19

2098

23

15

55

74

36

30

W

PM-T

HICK

OI

B (S

36)

M233

6-5

Hawa

ii Lo

ihi sm

t 10

86

.86

220

9850

9 13

85

2832

69

2150

26

95

38

5 1.4

06

0.951

0.9

37

2.183

0.5

70

0.1

0 9

734

13

499

44

56

14

W

PM-T

HICK

OI

B (S

36)

M233

6-4

Hawa

ii Lo

ihi sm

t 21

87

.57

235

9349

6 12

98

2865

00

1996

27

14

53

4 1.3

88

0.947

0.8

86

2.135

0.6

08

0.0

6 5

425

9 30

4 27

20

14

WPM

-THI

CK

OIB

(S36

) M2

338-

5 Ha

waii

Loihi

smt

5 88

.54

231

8629

0 12

21

2902

33

2311

31

49

42

3 1.4

14

1.085

0.9

36

2.679

0.5

53

0.1

9 12

13

56

27

891

22

67

4

WPM

-THI

CK

OIB

(S36

) M2

336-

7 Ha

waii

Loihi

smt

6 87

.20

200

9550

0 13

68

2830

74

2372

25

00

35

4 1.4

33

0.883

0.8

43

2.483

0.5

16

0.1

8 20

13

41

16

790

66

96

29

W

PM-T

HICK

OI

B (S

36)

M233

5-R

Hawa

ii Lo

ihi sm

t 22

89

.42

232

7964

9 11

43

2929

60

1962

30

15

64

1 1.4

35

1.029

0.8

20

2.463

0.5

11

0.0

6 7

391

9 37

9 35

32

15

WPM

-THI

CK

OIB

(S36

) M2

335-

2 Ha

waii

Loihi

smt

5 87

.41

196

9404

8 13

34

2840

70

2341

28

70

32

6 1.4

18

1.010

0.9

50

2.489

0.5

46

0.1

9 4

1438

27

65

9 49

72

19

WPM

-THI

CK

OIB

(S36

) M2

335-

8 Ha

waii

Loihi

smt

23

86.21

19

2 10

3052

14

62

2803

46

2559

24

38

29

8 1.4

19

0.870

0.8

96

2.483

0.5

45

0.0

6 6

420

8 28

0 39

61

11

WPM

-THI

CK

OIB

(S36

) M2

343-

4 Ha

waii

Loihi

smt

13

87.93

20

7 90

549

1334

28

6982

20

45

2715

428

1.473

0.9

46

0.857

2.2

58

0.432

0.09

17

674

12

429

123

42

25

W

PM-T

HICK

OI

B un

pb

5328

Ha

waii

Kaua

i-Hae

na

1 88

.92

259

8303

9 10

46

2899

96

1401

38

74

55

4 1.2

59

1.336

1.1

09

1.687

0.8

75

0.0

0 0

0 0

0 0

0

0 W

PM-T

HICK

OI

B un

pb

0433

c-033

R Pa

cific

Ocea

n Su

iko sm

t 11

88

.46

326

8637

1 11

56

2879

78

1559

33

70

162

655

1.338

1.1

70

1.011

1.8

05

0.711

0.08

9 61

0 9

530

21

19

2 13

W

PM-T

HICK

OI

B un

pb

0433

c-024

R Pa

cific

Ocea

n Su

iko sm

t 21

86

.77

280

9805

4 12

94

2798

60

1627

29

43

174

552

1.319

1.0

51

1.031

1.6

59

0.750

0.06

7 46

0 7

317

12

37

2 12

W

PM-T

HICK

OI

B (S

37)

LG-9

9-58

At

lantic

Oce

an

La G

omer

a 5

89.13

25

2 81

918

1209

29

2195

19

05

2839

15

5 47

7 1.4

76

0.972

0.7

96

2.326

0.4

25

0.1

0 11

71

2 6

557

46

160

3 12

W

PM-T

HICK

OI

B (S

37)

GC-9

8-59

At

lantic

Oce

an

Gran

Can

aria

5 89

.46

221

7978

0 11

82

2946

70

1711

30

46

153

464

1.481

1.0

34

0.825

2.1

44

0.415

0.14

5 87

1 7

1514

37

96

3

6 W

PM-T

HICK

OI

B (S

37)

GC-9

8-35

At

lantic

Oce

an

Gran

Can

aria

5 89

.21

180

8115

2 11

96

2918

39

1753

29

97

153

434

1.474

1.0

27

0.833

2.1

61

0.430

0.16

8 10

22

23

1094

60

20

9 4

15

WPM

-THI

CK

OIB

(S37

) GC

-126

5 At

lantic

Oce

an

Gran

Can

aria

3 89

.56

171

7862

9 12

19

2936

71

2085

27

05

152

367

1.551

0.9

21

0.724

2.6

52

0.271

0.23

19

1481

21

17

05

81

45

1 37

W

PM-T

HICK

OI

B (S

38)

953C

-101

R-5 1

16-1

23

Atlan

tic O

cean

Gr

an C

anar

ia 5

88.23

19

7 87

674

1202

28

6031

13

52

3547

15

4 48

6 1.3

71

1.240

1.0

87

1.542

0.6

44

0.1

7 8

1214

18

77

5 44

71

4

13

WPM

-THI

CK

OIB

(S38

) 95

3C-9

3R-0

5 13-

21

Atlan

tic O

cean

Gr

an C

anar

ia 14

88

.67

222

8479

5 11

61

2886

63

1447

34

71

152

518

1.369

1.2

03

1.020

1.7

06

0.648

0.08

3 55

5 8

335

6 39

3

10

Sobo

lev

et a

l, Su

ppor

ting

Onl

ine

Mat

eria

l ww

w.sc

ienc

emag

.org

/cgi

/con

tent

/full/

1138

113/

DC

1

20 A

PRIL

200

7 VO

L 31

6 SC

IEN

CE

ww

w.s

cien

cem

ag.o

rg

S9

GROU

P Ge

odyn

amic

settin

g Re

feren

ces

Samp

le Ge

ogra

phic

Lo

cality

N

Fo

Al pp

m Fe

ppm

Mn pp

m Mg

ppm

Ca pp

m Ni

ppm

Co pp

m Cr

ppm

100*

Mn/F

e 10

0*Ni

/Mg

Ni/(M

g/Fe)

/1000

10

0Ca/F

e X

px M

n/Fe

STE

Fo w

t% A

l ppm

Fe

ppm

Mn pp

m Mg

ppm

Ca pp

m Ni

ppm

Co pp

m Cr

ppm

WPM

-THI

CK

OIB

(S38

) 95

3C-9

3R-0

6 45-

55

Atlan

tic O

cean

Gr

an C

anar

ia 4

82.31

16

7 12

8922

17

48

2610

52

1930

24

85

187

219

1.356

0.9

52

1.227

1.4

97

0.674

0.20

12

1377

18

90

8 61

74

5

6 W

PM-T

HICK

OI

B (S

38)

953C

-93R

-4 18

-24

Atlan

tic O

cean

Gr

an C

anar

ia 5

87.75

18

1 91

591

1300

28

5495

16

02

3239

16

8 41

6 1.4

19

1.135

1.0

39

1.749

0.5

44

0.1

8 13

12

84

16

702

53

180

3 26

KO

MATI

ITES

KO

MATI

ITES

(S

39)

MUN-

24

Cana

da

Munr

o 2.7

Ga

65

92.34

39

2 58

523

925

3070

10

1562

33

16

14

41

1.580

1.0

80

0.632

2.6

69

0.210

0.03

4 22

3 4

182

4 4

6

KOMA

TIIT

ES

KOMA

TIIT

ES

(S39

) MU

N-41

4 Ca

nada

Mu

nro 2

.7 Ga

63

89

.42

254

8021

8 12

38

2948

87

1625

29

42

73

2 1.5

43

0.998

0.8

00

2.026

0.2

88

0.0

3 3

205

4 16

6 5

5

8 KO

MATI

ITES

KO

MATI

ITES

(S

39)

Ch-1

1a

Cana

da

Munr

o 2.7

Ga

32

88.82

29

9 84

803

1339

29

3033

16

45

3034

791

1.579

1.0

35

0.878

1.9

40

0.212

0.04

5 32

7 7

214

8 6

16

KO

MATI

ITES

KO

MATI

ITES

(S

40)

M-62

6 Ca

nada

Al

exo 2

.7 Ga

11

92

.81

334

5513

2 86

2 30

9510

14

90

3461

1299

1.5

64

1.118

0.6

17

2.703

0.2

43

0.1

0 7

772

13

474

7 10

20

KOMA

TIIT

ES

KOMA

TIIT

ES

(S40

) M-

767

Cana

da

Alex

o 2.7

Ga

8 93

.53

390

4982

9 80

1 31

3247

14

84

3481

1308

1.6

07

1.111

0.5

54

2.978

0.1

55

0.1

4 14

10

82

13

575

10

41

30

KO

MATI

ITES

KO

MATI

ITES

(S

40)

M-66

6 Ca

nada

Al

exo 2

.7 Ga

33

94

.30

421

4388

1 71

9 31

5979

14

84

3353

1263

1.6

38

1.061

0.4

66

3.381

0.0

91

0.0

5 9

404

5 27

4 5

6

20

KOMA

TIIT

ES

KOMA

TIIT

ES

(S41

) MZ

-4

Zimba

bwe

Belin

gwe 2

.7 Ga

9

92.60

36

1 56

373

885

3071

89

1494

36

30

128

1327

1.5

71

1.182

0.6

66

2.651

0.2

30

0.1

1 10

79

6 13

47

7 6

8 2

32

KOMA

TIIT

ES

KOMA

TIIT

ES

(S42

) G-

21

Cana

da

Gilm

our 1

.9 Ga

15

92

.03

359

6084

1 94

8 30

5902

17

62

3351

1058

1.5

58

1.095

0.6

67

2.897

0.2

56

0.0

8 8

626

9 32

7 11

18

25

KOMA

TIIT

ES

KOMA

TIIT

ES

(S42

) G-

18

Cana

da

Gilm

our 1

.9 Ga

2

91.83

35

2 62

561

968

3058

04

1787

33

04

10

71

1.547

1.0

80

0.676

2.8

56

0.278

0.36

3 26

62

46

1468

14

35

38

KOMA

TIIT

ES

KOMA

TIIT

ES

(S42

) G-

15

Cana

da

Gilm

our 1

.9 Ga

63

88

.74

222

8527

7 12

81

2924

37

1796

21

90

65

2 1.5

03

0.749

0.6

39

2.106

0.3

71

0.0

4 3

263

4 15

6 2

3

5 KO

MATI

ITES

KO

MATI

ITES

(S

42)

G-8a

Ca

nada

Gi

lmou

r 1.9

Ga

70

88.67

24

4 85

637

1284

29

1770

17

93

2617

675

1.500

0.8

97

0.768

2.0

94

0.377

0.03

6 20

9 3

124

3 5

9

KOMA

TIIT

ES

KOMA

TIIT

ES

(S43

) GO

R-94

-19

Gorg

ona

Gorg

ona 0

.9 Ga

71

91

.44

549

6460

7 10

51

3004

52

2140

35

93

92

0 1.6

26

1.196

0.7

73

3.313

0.1

15

0.0

2 4

126

2 11

7 2

5

5 KO

MATI

ITES

KO

MATI

ITES

(S

43)

GOR9

4-35

Go

rgon

a Go

rgon

a 0.9

Ga

13

92.58

53

3 56

770

930

3083

31

2120

34

47

11

67

1.638

1.1

18

0.635

3.7

34

0.090

0.08

10

613

11

396

10

16

41

KO

MATI

ITES

KO

MATI

ITES

(S

43)

GOR9

4-34

Go

rgon

a Go

rgon

a 0.9

Ga

27

91.50

47

3 64

764

1042

30

3202

22

74

3247

1003

1.6

09

1.071

0.6

94

3.511

0.1

50

0.0

5 12

40

2 7

256

12

12

15

KO

MATI

ITES

KO

MATI

ITES

(S

43)

GOR9

4-32

Go

rgon

a Go

rgon

a 0.9

Ga

18

92.79

54

4 54

632

907

3059

20

2121

34

82

11

95

1.661

1.1

38

0.622

3.8

82

0.043

0.07

18

527

9 26

6 12

10

29

Note

s for

Tab

le S2

a. Co

ncen

tratio

ns an

d stan

dard

erro

rs of

mean

(STE

) are

show

n for

elem

ents

in pp

m an

d for

Fo i

n mol%

. N-

numb

er o

f ave

rage

d hig

h-Mg

oliv

ines (

see

text fo

r defi

nition

). Sa

mples

mar

ked

(i) a

nd (i

i) re

pres

ent d

iffere

nt co

mpos

itiona

l gro

ups o

f oliv

ine w

ithin

the sa

me sa

mple.

Sam

ples m

arke

d (n

) rep

rese

nt an

add

itiona

l inde

pend

ent p

opula

tion

of oli

vines

analy

zed f

or th

e sam

e sam

ple. R

efere

nces

inclu

de sa

mples

desc

riptio

n, un

pb- u

npub

lishe

d. *-

Onton

g Jav

a Plat

eau s

ample

s wer

e arb

itrary

class

ified a

s WPM

-THI

CK, b

ut ac

tually

may

belon

g to W

PM-T

HIN

at the

time o

f form

ation

.

The Amount of Recycled Crust in Sources of Mantle-Derived ......The Amount of Recycled Crust in...

Documents

Transcript of The Amount of Recycled Crust in Sources of Mantle-Derived ......The Amount of Recycled Crust in...