j ourna l homepage: www.e lsev ie r .com/ locate / l i thos
A radiogenic isotopic (He-Sr-Nd-Pb-Os) study of lavas from the Pitcairnhotspot: Implications for the origin of EM-1 (enriched mantle 1)
G. Garapić a,⁎, M.G. Jackson a, E.H. Hauri b, S.R. Hart c, K.A. Farley d, J.S. Blusztajn c, J.D. Woodhead e
a Department of Earth Science, University of California, Santa Barbara, USAb Department of Terrestrial Magnetism, Carnegie Institution of Washington, Washington, DC 20015, USAc Department of Marine Geology and Geophysics, Woods Hole Oceanographic Institution, 266 Woods Hole Road, Woods Hole, MA 02543, USAd Geological and Planetary Sciences Division, California Institute of Technology, Pasadena, CA 91125, USAe School of Earth Sciences, The University of Melbourne, Victoria 3010, Australia
We present new He-Sr-Nd-Pb-Os isotopic compositions and major and trace-element concentrations for tensubaerially-erupted lavas and one seamount lava associated with the Pitcairn hotspot. The most geochemically-enriched lavas at the Pitcairn hotspot have signatures that are consistent with recycled sediments derived fromupper continental crust. Pitcairn lavas have elevated Ti, which also supports the presence of a mafic protolith inthe Pitcairn mantle. A subset of Pitcairn seamount samples, including the seamount sample presented here, aretholeiitic. Tholeiitic lavas are uncommon at ocean hotspots located far from mid-ocean ridges. Like tholeiites thaterupted in Hawaii, the presence of tholeiites in the Pitcairn magmatic suite can be explained by melting a silica-saturated recycled mafic component in the Pitcairn mantle source. We also present the highest 3He/4He ratio(12.6 Ra, ratio to atmosphere) from the Pitcairn hotspot. This sample anchors the high 206Pb/204Pb portion of thePitcairn array and provides evidence for a plume component in the Pitcairn mantle. In contrast, Pitcairn lavasthat have the lowest 206Pb/204Pb are the most geochemically enriched, and have the highest 87Sr/86Sr and lowest143Nd/144Nd in the Pitcairn suite; these EM-1 end-member lavas have MORB-like 3He/4He (~ 8 Ra, ratio toatmosphere). Recycled oceanic crust and sediment suggested to be in the Pitcairn EM-1 mantle are expected tohave low 3He/4He (b 0.1 Ra). Therefore, the higher, MORB-like 3He/4He in Pitcairn EM-1 lavas is paradoxical, butmight be explained by diffusive exchange of helium, but not the heavy radiogenic isotopes, with the ambientmantle over billion-year timescales.
Geochemical analyses of ocean island basalts (OIB) erupted athotspots reveal that the Earth's mantle is compositionally heteroge-neous (e.g., Gast et al., 1964; Hofmann, 1997, 2003; Stracke, 2012;White, 2010; Zindler and Hart, 1986). Compilations of radiogenicisotopic measurements made on OIB show that several distinct isotopiccompositions emerge from the global dataset (Zindler and Hart, 1986),including EM-1 (enriched mantle 1), which is characterized by theunradiogenic 206Pb/204Pb ratios with moderately high 208Pb/204Pb and87Sr/86Sr, EM-2 (enriched mantle 2) which is characterized by interme-diate 206Pb/204Pb ratios and highly radiogenic 87Sr/86Sr, and HIMU (highμ, or high 238U/204Pb), which is characterized by the most radiogenic Pbisotopic compositions in the oceanic mantle. Geochemically-depletedcompositions are also sampled by hotspots, and the origin of thedepleted component in plumes is not well understood and may host
New Paltz, New Paltz, NY, USA.
elevated (primordial) 3He/4He ratios (Farley et al., 1992; Hanan andGraham, 1996; Hart et al., 1992). Mixing of the EM-1, EM-2 and HIMUend-members with the depleted composition is thought to generatemuch of the diversity in OIB.
While the field of mantle geochemistry has advanced to a state ofcareful description and classification of the various mantle species, andtheir possible mixing relationships, the origin of these species stillremains poorly understood. Models explaining the sources of themantle end-members often call for subduction injection of surface mate-rials, including continental crust, oceanic crust, and sediments into themantle over geologic time (Hofmann andWhite, 1982). Followingmixingwith ambient mantle, the subducted protoliths are sampled by regions ofupwelling mantle, called plumes, and melted beneath hotspots. Alterna-tive hypotheses for the origin of the various end-members include delam-ination of subcontinental lithosphere (Mahoney et al., 1991; McKenzieand O'Nions, 1983), CO2-flux induced melting of pristine mantle domainin the lowermantle (Collerson et al., 2010), andmore exoticmetasomaticprocesses, wherefluids in themantle impart enriched geochemical signa-tures on specific domains of the mantle (e.g., Geldmacher et al., 2008;Menzies, 1983; Menzies and Murthy, 1980; Niu and O'Hara, 2003; Pilet
et al., 2011; Salters and Sachi-Kocher, 2010;Workman et al., 2004). How-ever, a consensus is emerging for the origin of EM-2 as recycled marinesediment (Workman et al., 2008) with a terrigenous origin (Jacksonet al., 2007). Similarly, the long-standing hypothesis that HIMU repre-sents recycled oceanic crust (Eiler et al., 1997; Hofmann and White,1982; Zindler et al., 1982) is gaining strength (Cabral et al., 2013; Hanyuet al., 2011, 2014; Kawabata et al., 2011).
Among all mantle end-members, no consensus has emerged for theorigin of EM-1 (e.g., Hofmann, 1997; Hart, 2011) and yet numerousmechanisms have been proposed (e.g., Blichert-Toft et al., 1999;Brandenburg et al., 2008; Chauvel et al., 1992; Collerson et al., 2010;Eiler et al., 1995; Eisele et al., 2002; Escrig et al., 2004; Gasperini et al.,2000; Geldmacher et al., 2008; Hanan et al., 2004; Honda andWoodhead, 2005; Lassiter and Hauri, 1998; Mahoney et al., 1991;Rehkämper and Hofmann, 1997; Salters and Sachi-Kocher, 2010;Tatsumi, 2000; Weaver, 1991; Willbold and Stracke, 2010; Woodheadand Devey, 1993; Woodhead and McCulloch, 1989). Here, armed witha new geochemical dataset from the Pitcairn hotspot, we argue forthe presence of recycled sediment in the Pitcairn EM-1 mantle(e.g., Eisele et al., 2002; Honda and Woodhead, 2005), and we suggestthat Pitcairn lavas also sample recycled oceanic crust. We also reportnew helium data on Pitcairn lavas, including the highest 3He/4He ratiofrom Pitcairn, and the new data provide insights into the evolutionof the poorly-understood high 3He/4He mantle and EM-1 mantlereservoirs.
2. Methods
The ten subaerial lavas from Pitcairn island reported here werecollected by Ken Farley during the 1989 Helios Expedition. We do notreport ages for the samples in this study, but subaerial lavas fromPitcairn have ages between 0.4 and 1.0 Ma (Duncan et al., 1974). Fig. 1shows locations for the subaerial samples presented in this study.Based on comparison with the geological formations as mapped byCarter (1967), the samples in this studywere collected from the Tedside(Pit-1, Pit-3, Pit-4, Pit-7, Pit-8) and Pulawana volcanic formations (Pit-6,Pit-16, Pit-11, Pit-12, Pit-13). Samples Pit-6, 12, 13 and 16 werecollected in place from outcrops. Samples Pit-1, 3, 4a, 7, 8 and 11 werecollected as cobbles on thewave cut terrace; it is very likely that cobbleson the terrace derive from the overlying cliff as there are no rivers totransport cobbles and longshore transport is not plausible.
Fig. 1.Mapof Pitcairn islandwith sample locations from this study. The subaerial samples, and th65DS-4 was obtained from volcano 1 located ~90 km ESE of Pitcairn.
We also report new geochemical data on a single submarine sample,65DS-4, dredged between 1503 and 1276 m water depth on AdamsSeamount (25°21′26″, 125°18′59″), located ~90 km ESE of Pitcairn.This sample was recovered aboard the FS Sonne on December 1989 onthe 65th cruise of this vessel (Stoffers et al., 1990; Woodhead andDevey, 1993). The seafloor extending ~90 km ESE of Pitcairn Island ispopulated by a submarine volcanic field hosting approximately 90volcanic cones and seamounts, including Adams Seamount (Hekinianet al., 2003), the largest seamount in this volcanic field (Hekinianet al., 2003; Woodhead and Devey, 1993). Several of these volcanicfeatures exhibit evidence of being active, and submarine dredgesreturned samples with fresh volcanic rocks and glass; there was littleevidence for sediment cover (Stoffers et al., 1990). Seafloor backscatterimages revealed that several of the volcanic edifices are highlyreflective, further supporting a young age for these seamounts, andsubstantiating the hypothesis that that these seamounts (includingAdams Seamount) mark the active portion of the Pitcairn hotspots(Hekinian et al., 2003; Stoffers et al., 1990; Woodhead and McCulloch,1989). K/Ar age determinations of lavas recovered from AdamsSeamount range from 2 to 7 kyr (Guillou et al., 1997) and volcanicrocks recovered from this seamount exhibit evidence of only slightalteration (Hekinian et al., 2003). Additionally, geochemical dataobtained on seven dredges of Adams Seamount, including new datapresented here on sample 65DS-4, indicate that this seamount has aPitcairn hotspot pedigree (Woodhead and Devey, 1993). Given this sig-nificant body of evidence, we adopt the finding that Adams Seamountmarks the active end of the Pitcairn hotspot.
Cut rock slabs were crushed in plastic bags to avoid contaminationwith metal from the hammer, and the freshest chips were separatedand powdered in agate. With the exception of sample 65DS-4, majorelements and a subset of trace elements were measured on powders byXRF in the laboratory of Michael Rhodes at the University of Massachu-setts Amherst, and the methods and associated errors are reported inRhodes (1988). The remaining trace elements were measured by ICP-MS at the Geoanalytical Lab at Washington State University (WSU)(Knaack et al., 1994). For sample 65DS-4, major and trace elementswere measured at WSU by XRF and ICP-MS (analytical methods anderrors are reported in Johnson et al. (1999); Knaack et al. (1994)). AtWSU, precision for the analyses of SiO2, Al2O3, TiO2 and P2O5 in basaltsis 0.11 - 0.33% (1σ) of the amount present, and 0.38 - 0.71% (1σ) of theamount present for FeO, MgO, CaO, Na2O, MnO and K2O (Johnson et al.,
e single submarine sample,were collected prior toGPS availability. The submarine sample
Table 1Whole rock major, trace and isotope analyses for Pitcairn samples from this study, and Os-isotopic analyses on olivines.
Data in italics are previously published in Eiler et al. (1995; 1997) where NA indicates “present but not analyzed”. * 3He/4He on sample PIT-16, was measured twice, on coarse(11.8 Ra; 4.3*10−9 cc STP/g) and fine (13.3 Ra; 5.5*10'1°) olivine fractions, to verify the elevated 3He/4He ratio.
3G. Garapić et al. / Lithos 228–229 (2015) 1–11
1999). Trace element analyses by ICP-MS at WSU have a precision of0.77 - 3.2% (1σ) for all elements except for Th (9.5%) and U (9.3%)(Knaack et al., 1994). Sample 65DS-4 is an olivine cumulate, andwe pres-ent XRF major element data on the whole rock and on (olivine-free)matrix separated from the sample.
For the subaerial lavas, chemical separation protocols and Sr and Ndisotopic measurements were made at the Woods Hole OceanographicInstitution and follow the same protocols outlined in Hauri and Hart(1993); Sr and Nd isotopic data have never been reported for these sam-ples. Sr-isotopic compositions were fractionation corrected relative to
XF0 = (n(MgO)/(n(MgO) + n(FeO)))*100, where n = number of moles;Multiple spots were measured on two olivine grains. The grains were designated as oil(n = 8 spots) and ol2 (n = 4 spots).
4 G. Garapić et al. / Lithos 228–229 (2015) 1–11
86Sr/88Sr=0.1194, and 87Sr/86Sr ratioswere adjusted to an SRM987valueof 0.710240. The external precision for the 87Sr/86Sr ratio is estimated tobe 45 ppm, and is based on repeat analyses of the SRM987 Sr standard.Nd-isotopic compositions were fractionation corrected relative to a146Nd/150Nd value of 0.7219. The 143Nd/144Nd ratios were adjusted to avalue of 0.511847 for the La Jolla Nd standard. The external precision onthe 143Nd/144Nd is estimated to be 40 ppm, and is based on repeat mea-surements of the La Jolla Nd standard. Pb isotopic data for the subaerialsamples were previously reported for all samples in Eiler et al. (1995),except for sample Pit-6, which is reported here; analytical proceduresand fractionation correction are the same as outlined in Hauri and Hart(1993), and Pb-isotopic ratios for sample Pit-6 were adjusted toSRM981 values from (Todt et al., 1996) to be consistent with the datareported in Eiler et al. (1995). Sr, Nd and Pb chemical separation protocolsand isotopic analyses for the submarine lava, 65DS-4, follow thoseoutlined in Hart and Blusztajn (2006); Sr, Nd and Pb isotopic measure-ments were made on the Neptune multi-collector ICP-MS at WHOI.After adjusting to 0.710240 and 0.511847 for SRM987 and La Jolla Ndstandards, respectively, the external precision is estimated to be between15 and 25 ppm. Pb-isotopicmeasurementswere adjusted to the SRM981values reported in (Todt et al., 1996), and external reproducibility is esti-mated to be better than 120 ppm for the 206Pb/204Pb, 207Pb/204Pb and208Pb/204Pb ratios.
Osmium isotopic analyses were performed on 315 to 820 mg ofolivine separated from six Pitcairn samples. Samples were dissolvedby HF-HCl dissolution, subjected to a gentle dry down at b100 °C, thenloaded in Carius Tubes as described in Shirey and Walker (1995);chemical separation of Os is as described in Shirey and Walker (1995).The isotopic analysesweremade at DTMby negative thermal ionizationmass spectrometry (N-TIMS) (Hauri and Kurz, 1997). In-run precisionwas better than 0.1% (2σ standard error of the mean) on the187Os/188Os ratio for all runs. Total procedural blanks were b5 pg Osand were within analytical errors. Analyses of the DTM osmiumstandard yielded an average 187Os/188Os of 0.11378 ± 10 (2σ, n = 5).Helium isotopic measurements were made on 0.9 to 3.8 g of olivinephenocrysts by crushing in vacuo and follow the methods outlined inFarley et al. (1993). Uncertainties on the 3He/4He ratios are estimatedto be ±05 Ra; the estimated uncertainties account for both blankcorrection and uncertainty in standards of similar size to samples. Thesample with the highest 3He/4He, Pit-16, was measured twice on morethan 4 g of olivine, on coarse (11.8 Ra) and fine (13.3 Ra) olivine fractions(the average value of the two measurements, 12.6 Ra, is used hereafter),to verify the elevated 3He/4He ratio; the raw ratioswere 10.5 and 11.5 Ra,so evenwithout blank correction the sample is unusual. This sample wascollected on a sea-cliff, where erosion enhanced by continuous waveaction precludes long-term exposure at the surface, which enables us toexclude a cosmogenic helium origin for the elevated isotopic ratio inthis sample.
The new isotopic data are presented in Table 1, together withpreviously published data on these samples. Additionally, in Table 2 wereport major element abundances for olivine phenocrysts from subma-rine sample 65DS-4; these data complement previously-published olivinecompositional data on subaerially-erupted lavas from the hotspot (Eileret al., 1995). Microprobe analyses were made on Cameca SX-100 at UCSanta Barbarawith the following operating conditions: 15 kV acceleratingvoltage, 15 nA beam current and 1 μm beam diameter.
Fig. 2. TAS diagram shows that Pitcairn lavas are prevailingly alkalic, but some seamountsamples are slightly tholeiitic and cross the alkali-tholeiite divide. The new data forPitcairn seamount sample 65DS-4 show that it is a tholeiite (whole rock and matrix dataare shown for this sample). The previously published data shown in this figure (and inall subsequent figures) were obtained from the GEOROC database (http://georoc.mpch-mainz.gwdg.de/georoc) on October 1, 2013 and verified with the original references.
3. Data
All of the samples in this study are porphyritic in texture. According toa TAS classification (Fig. 2), subaerial Pitcairn lavas are predominantlyalkalic and span the whole suite of compositions from alkaline basalts totrachy-andesites. Most of the submarine lavas are also alkalic, but a limit-ed number of seamount samples cross the alkali-tholeiite division ofMacdonald and Katsura (1964) and are tholeiitic, as was noted by
Hekinian et al. (2003). The new seamount sample, 65DS-4, is clearly tho-leiitic, based on its whole-rock and matrix major element composition.
Fig. 3 shows variation diagrams of MgO versus major elementoxides. Olivine fractionation dominates the major element variationsfor bulk rock compositions N8–10 wt.% MgO. At MgO b 8–10 wt.%,clinopyroxene fractionation is apparent from reduced CaO/Al2O3 ratios.Fractionation of Fe- and Ti-rich phases is evident in lavas withMgO b 5 wt.%.
The panels in Fig. 4 display isotopic compositions for the new Pitcairnhotspot dataset compared to previously published Pitcairn data and otherOIB. Separate fields are designated for Samoa, which represents theextremeEM-2 end-member, and forHawaii. A separate distinctionwithinthe Hawaiian field is made for Koolau lavas because these lavas host anenriched end-member that has an EM-1 pedigree. However, Pitcairnand Hawaiian Koolau lavas have distinct EM-1 flavors. In both,208Pb/204Pb versus 206Pb/204Pb and 207Pb/204Pb versus 206Pb/204Pbdiagrams, Pitcairn and Hawaii form two separate fields, and highlightthe difference in the EM-1 composition between the two hotspots.However, in 143Nd/144Nd versus 87Sr/86Sr isotopic space, Pitcairn overlapswith Hawaiian Koolau lavas, highlighting the complementary EM-1 sig-natures present at both hotspots, in spite of their different Pb-isotopiccompositions. The plot of 208Pb/204Pb versus 206Pb/204Pb reveals the pres-ence of a third component in Pitcairn lavas (located at the lowest vertex of
Fig. 3. Variation diagrams of major element oxides for all Pitcairn lavas. The data are normalized to 100 wt.% on a dry basis. Olivine dominates the major element variations for bulk rockcomposition where MgO N 8–10 wt.%. At MgO b 8–10 wt.%, clinopyroxene fractionation is apparent from reduced CaO/Al2O3 ratios. Fractionation of Fe- and Ti-rich phases is evident inlavas with MgO b 5 wt.%. For some of the seamount samples, both whole rock and glass data are plotted when it was available on the same sample. For 65DS-4 whole rock (65DS-4 wr) and matrix (65DS-4 m) are plotted separately.
5G. Garapić et al. / Lithos 228–229 (2015) 1–11
the triangle outlining the Pitcairn dataset in Pb-isotope space in Fig. 4),and this component is characterized by lower 208Pb/204Pb at a given206Pb/204Pb. Curiously, this third component is only visible in208Pb/204Pb versus 206Pb/204Pb isotope space, and is not evident in otherisotope projections. Sample 65DS-4 defines the end-member composi-tion for this new component, and it plots closer to Hawaiian Koolaulavas in Pb-isotopic space than any previously examined Pitcairn lava.We emphasize that a subset of prior analyses of lavas from the Pitcairnhotspot have Sr, Nd and Pb isotopic compositions that are similar to65DS-4, but this sample slightly expands the field of the Pitcairn hotspotin 208Pb/204Pb versus 206Pb/204Pb isotopic space.
Fig. 5 shows 3He/4He ratios in olivine in subaerial samples (reportedhere for the first time) and seamount glasses [reported in] (Honda andWoodhead, 2005), plotted against 206Pb/204Pb. At a given 206Pb/204Pb
ratio, the 3He/4He of the olivines is bracketed by the 3He/4He of theglass dataset; only the Pit-16 olivine 3He/4He measurement is higherthan any of the 3He/4He measurements on glasses. The 3He/4He of thePitcairn olivines presented here show no evidence for reduced3He/4He by radiogenic ingrowth of 4He. Fig. 5 also shows a possiblerelationship between 206Pb/204Pb and 3He/4He, where the highest3He/4He ratio is found in the sample with the highest 206Pb/204Pb. Incomparison, the two lavas with the lowest 206Pb/204Pb have the lowest3He/4He ratios, 7.7 Ra (measured in a glass) and 8.0 Ra (measured inolivine), indicating that EM-1 lavas at Pitcairn haveMORB-like 3He/4He.
Fig. 6 shows 187Os/188Os ratios and Os concentrations measured onolivines from samples from this study compared to the whole rockmeasurements of Pitcairn lavas from previous studies (Eisele et al.,2002; Reisberg et al., 1993). 187Os/188Os ratios in olivines from this
Fig. 4. Isotopic data for Pitcairn lavas reveal a typical EM-1 signature: low 143Nd/144Nd, moderately high 87Sr/86Sr and highly unradiogenic 206Pb/204Pb. The plot of 208Pb/204Pbagainst 206Pb/204Pb reveals the presence of a third component in Pitcairn lavas, which is the second component that is highlighted in the dashed triangle. The third component(called low-κ EM-1) is anchored by geochemically-more depleted rejuvenated lavas. Data fields for Samoa and Hawaii (including Koolau volcano) are shownwith the global OIBdatabase, and these data are taken from Georoc database and verified with the original references. The Pitcairn isotopic data are from the following sources: (Eisele et al., 2002;Honda and Woodhead, 2005; Woodhead and Devey, 1993; Woodhead and McCulloch, 1989; Woodhead et al., 1993).
6 G. Garapić et al. / Lithos 228–229 (2015) 1–11
study range from 0.1342 to 0.1432, which falls within the range previ-ously identified in whole-rock measurements of the most Os-richPitcairn samples (0.131 to 0.148 for samples with Os N 50 pg/g) (see
Fig. 5. 3He/4He ratios andHe concentrations in olivine (in subaerial samples, reportedherefor the first time) and seamount glasses (reported in Honda and Woodhead, 2005). Onlythe data for glasses with 4He N 10−7 cc STP/g are plotted in the right panel. The line inthe left panel connects two replicate measurements on coarse and fine olivine fractions(sample PIT-16). The diagram to the right shows that the sample with the highest3He/4He ratio (which is shown as the average of two measurements of this sample,11.8 Ra and 13.3 Ra, measured on coarse and fine olivine fractions, respectively) also hasthe weakest EM-1 signature and instead trends toward the depleted mantle component(FOZO?) sampled by Pitcairn lavas. The two lavas with the lowest 206Pb/204Pb have thelowest 3He/4He ratios, 7.7 Ra (measured in a glass) and 8.0 Ra (measured in olivine), indi-cating MORB-like 3He/4He signature in the EM-1 mantle.
Fig. 2b in Eisele et al. (2002)). As observed by Eisele et al. (2002),there are no clear relationships between Os isotopic ratios and theother radiogenic isotopic compositions presented in this study. Rheni-um was not measured on the olivines, so an age-correction cannot beperformed. However, recent studies on magmatic olivines from Samoa(Jackson and Shirey, 2011), Iceland (Debaille et al., 2009) and theCook Islands (Hanyu et al., 2011) indicate that Re/Os ratios in magmaticolivines are low, and the subaerial lavas from Pitcairn are young, so anyage-correction to the 187Os/188Os ratio of the olivine will be small.
Fig. 6. The new olivine 187Os/188Os measurements in this study fall in the same range asidentified in the most Os-rich whole rocks (N50 pg/g; Eisele et al., 2002) and supportthe previous observations that the EM-1 end-member has radiogenic Os, which isconsistent with a recycled origin for EM-1 lavas in Pitcairn (Eisele et al., 2002).
Fig. 7. 206Pb/204Pb plotted against the major element compositions of Pitcairn lavas, theHawaiian dataset is from Jackson et al. (2012) and the global OIB dataset from JacksonandDasgupta (2008).With the exception of sample 65-DS-4matrix, the displayed Pitcairndata shows only data with MgO N 9%, and all lavas are olivine fractionation corrected bythe addition or subtraction of equilibrium olivine so that they were in equilibrium withmantle olivine (forsterite 90). The data shows that, together with Hawaii, Pitcairnseamount lavas anchor the high SiO2-rich portion of the Pitcairn array.
7G. Garapić et al. / Lithos 228–229 (2015) 1–11
Fig. 7 displays SiO2 data plotted against 206Pb/204Pb for Pitcairn,together with the global OIB dataset from Jackson and Dasgupta(2008) and a Hawaiian dataset from Jackson et al. (2012). With theexception of thematrix analysis from sample 65DS-4, only Pitcairn sam-ples with MgO N 9% are plotted, and all lavas were corrected for olivinefractionation. The data show that, with Hawaii, Pitcairn lavas anchor thelow 206Pb/204Pb, SiO2-rich portion of the global OIB array.
4. Discussion
4.1. Recycled sediment in the source of Pitcairn EM-1 lavas
Previous work on the EM-1 signature hosted in Pitcairn lavas (Eiseleet al., 2002; Honda andWoodhead, 2005; Woodhead and Devey, 1993;Woodhead and McCulloch, 1989) argued that the data are consistentwith sediment addition to the mantle source: EM-1 Pitcairn lavas havehigh 87Sr/86Sr and 187Os/188Os, low 143Nd/144Nd, and Nb-depletion(Nb/Nb ∗ b1) (e.g. Eisele et al., 2002; Honda and Woodhead, 2005).Additionally, EM-1 lavas from Pitcairn tend to have low Ce/Pb ratios(e.g. Honda and Woodhead, 2005), which is a characteristic of marinesediments (e.g. Plank and Langmuir, 1998).
More recently, Willbold and Stracke (2010) argued for recycling oflower continental crust to generate EM-1 signatures in oceanic lavas,which follows prior work suggesting a similar origin for the EM-1 man-tle (Escrig et al., 2004; Hanan et al., 2004; Tatsumi, 2000; Willbold andStracke, 2010). Willbold and Stracke (2010) relied on Eu/Eu* to distin-guish between upper and lower continental crust in the mantle sourceof Pitcairn lavas: Upper continental crust has negative Eu/Eu*, andlower continental crust has positive Eu/Eu*. Intriguingly, the Pitcairnlavas in Willbold and Stracke (2010) (refer to Fig. 4 of said reference)show no clear Eu-anomaly, and most of the Pitcairn lavas in Willboldand Stracke (2010) actually have Eu/Eu* values slightly less than one,which is consistent with the addition of recycled upper continentalcrust.While Eu/Eu*may be useful for distinguishing a lower continentalcrust component in several EM-1 hotspots globally, it may not support alower continental crust component at the Pitcairn hotspot, and permitsa model whereby upper continental crust and derivative sedimentsexist in the Pitcairn mantle.
Sediments derived from upper continental crust have the necessarygeochemical ingredients for generating an EM-1mantle source sampledby Pitcairn lavas. However, upper continental crust and derivative sed-iments are highly heterogeneous (Plank and Langmuir, 1998; Rudnickand Gao, 2003). To better characterize the type of the crustal materialthat contributes to the EM-1 source, we point out two distinctive
geochemical feature of EM-1 as sampled by Pitcairn lavas: 1. Pitcairnlavas exhibit a range of highly unradiogenic 206Pb/204Pb ratios associat-ed with relatively high 208Pb/204Pb, and 2. The most geochemically-enriched Pitcairn lava with Hf-isotopic data departs from the globalarray formed by 143Nd/144Nd and 176Hf/177Hf and trends toward a com-ponentwith high 176Hf/177Hf at a given 143Nd/144Nd (Eisele et al., 2002).The former geochemical characteristic indicates that the protolith con-tributing to the enriched mantle source beneath Pitcairn must alsohave high Th/U (Eisele et al., 2002). Sediments can have high Th/U,and pelagic clays have among the highest Th/U, as they are theweathering products of mature upper continental crust (where U hasbeen preferentially mobilized relative to Th over time) (Plank andLangmuir, 1998). Eisele et al. (2002) demonstrated that the Th/U ratioin pelagic sediments will generate Pb-isotopic systematics appropriatefor Pitcairn hotspot lavas over time. Pelagic clays also satisfy the176Hf/177Hf systematics in Pitcairn lavas since they do not contain detri-tal zircon, a Hf-rich phase. The presence of detrital zircon in a sedimentwould yield lower Lu/Hf ratios at a given Sm/Nd ratio, and the departureof Pitcairn lavas from themantle array suggests the presence of a zircon-poor pelagic sediment. Therefore, among possible subducted protolithsin the EM-1mantle, pelagic sediments have the combination of high Th/U and high Lu/Hf required to generate the Nd, Hf and Pb isotopic com-positions in Pitcairn EM-1 lavas. Therefore, based on the geochemicalcharacteristics inferred for the Pitcairn EM-1 mantle source, incorpora-tion of a pelagic sedimentary protolith into the mantle sourced byPitcairn EM-1 lavas remains a plausiblemodel for the origin of thisman-tle end-member (Eisele et al., 2002; Honda and Woodhead, 2005;Woodhead and Devey, 1993; Woodhead and McCulloch, 1989).
4.2. The presence of subducted oceanic crust in the EM-1 Pitcairn mantle
Eisele et al. (2002) modeled the mantle source of the EM-1 Pitcairnmantle and found that, in addition to supporting a role for recycled pe-lagic sediment in the EM-1 mantle, the model results permit the pres-ence of a mafic component in the EM-1 Pitcairn mantle. Highconcentrations of TiO2 in OIB lavas, including lavas from the Pitcairnsuite, have also been used to argue for a recycled mafic component intheir mantle sources (Prytulak and Elliot, 2007). This is becausemeltingof a primitive or depleted mantle peridotite cannot generate the elevat-ed TiO2 observed in primitive OIB lavas from many hotspots, even atvery low degrees of melting. We emphasize that elevated Ti concentra-tions do not translate to particularly striking Ti, Ta and Nb (TITAN)anomalies (Jackson et al., 2008) in the new dataset; this is because theconcentrations of these elements are generally complemented byhigher concentrations of elements of similar incompatibility. Whilethe presence of a mafic component in the Pitcairn mantle is indicatedfrom elevated TiO2 in Pitcairn lavas, it is important to evaluate the na-ture of this mafic component.
With the exception of the tholeiitic seamount samples from thePitcairn hotspot (including sample 65DS-4), most Pitcairn lavas arealkalic and silica undersaturated, which suggests that the mafic compo-nent in the mantle source is silica-undersaturated (i.e., a silica-deficientpyroxenite; Herzberg, 2011). This silica-undersaturated mafic compo-nent may be responsible for elevating the TiO2 in alkalic Pitcairn lavas.
The less common tholeiitic samples from the Pitcairn hotspot offerimportant clues to silica-saturated recycled mafic compositions in thePitcairn mantle. Together with Koolau lavas from Hawaii (e.g., Hauri,1996), lavas from the Pitcairn hotspot, in particular the submarinePitcairn seamount lavas (e.g., Eisele et al., 2002; Hekinian et al., 2003;Woodhead and Devey, 1993), anchor the SiO2-rich portion of the globalOIB array (Fig. 7, Jackson and Dasgupta, 2008). The similarity in SiO2
contents in both the Pitcairn seamount lavas (including sample 65-DS4) andKoolau lavasmay suggest thatmantle sources at both EM-1 lo-calities may generate SiO2-rich lavas. The observation of tholeiitic sam-ples at the Pitcairn hotspot is important and can give important clues tothe nature of the EM-1mantle, as oceanic hotspot lavas erupted far from
8 G. Garapić et al. / Lithos 228–229 (2015) 1–11
mid-ocean ridges (i.e., excluding near-ridge hotspots like Iceland andGalapagos) are predominantly alkalic. Among hotspots that eruptedfar frommid-ocean ridges, excludingHawaii, tholeiitic lavas are uncom-mon (b10%). Of 14 such hotspots tabulated by (Jackson and Dasgupta,2008), all plot, on average, in the alkali basalt field of Macdonald andKatsura (1964).
The new tholeiitic Pitcairn seamount sample, 65DS-4, not only hasthe highest olivine-corrected SiO2 abundance in the Pitcairn dataset,but it also extends the Pitcairn field closer to the Koolau EM-1 field inPb-isotopic space than any previously examined Pitcairn lava (Fig. 4).The geochemical affinities suggest that there may be broad similarities,including geochemically-enriched Sr and Nd isotopic compositions andelevated SiO2, in the mantle sources sampled by the low 208Pb/204Pbcomponent in Pitcairn (which is exhibited in purest form by 65DS-4;Fig. 4) and the EM-1 component in Koolau lavas. The origin of theSiO2-rich nature of Koolau lavas from Hawaii is argued to be the resultof eclogite melting in the Hawaiian mantle plume (Hauri, 1996;Herzberg, 2011; Sobolev et al., 2005), where the eclogite is suggestedto be a recycled oceanic crust that preserved low U/Pb ratios throughthe subduction zone and generated relatively low time-integrated206Pb/204Pb (Jackson and Dasgupta, 2008; Jackson et al., 2012). Hauri(1996) showed that existing peridotite melting experiments at pres-sures N2 GPa (i.e., pressures that are slightly lower than the pressureat the base of mature oceanic lithosphere beneath Pitcairn and Hawaii)cannot generate the major element compositions observed in Koolaulavas, and more recent compilations of experimental data support thisobservation (see discussion in Jackson et al. (2012)). Hauri (1996)Hauri (1996) also showed that elevated FeO at a given SiO2 in Koolaulavas is a result of eclogite melting. Given existing experimentconstraints, the high FeO and SiO2-rich tholeiites from Hawaii are bestexplained by melting a silica-saturated, mafic protolith in the mantlesource (Hauri, 1996; Jackson et al., 2012). Sample 65DS-4 and a subsetof Pitcairn seamount samples plot within or close to the HawaiianKoolau field in a plot of SiO2 vs. FeO (Fig. 8); sample 65DS-4 plots inthe field for Mauna Loa lavas, which also host elevated FeO at a givenSiO2 abundance, and like Koolau lavas, are argued to be melts of aneclogite-bearing mantle source (Hauri, 1996), (Lassiter and Hauri,1998), (Blichert-Toft et al., 1999), (Huang and Frey, 2005), (Sobolevet al., 2005), (Herzberg, 2006), (Jackson et al., 2012). Therefore, we sug-gest that a subset of EM-1 Pitcairn lavas, which have SiO2 abundancesthat anchor the high-SiO2 portion of the data arrays in Figs. 7 and 8,
Fig. 8. FeOT plotted against SiO2 of Pitcairn lavas and Mauna Loa, Koolau and Loihi fromHawaiian dataset (obtained from GEOROC database on March 13, 2014). All lavas areolivine fractionation corrected by the addition or subtraction of equilibrium olivine sothat they were in equilibrium with mantle olivine (forsterite 90). With the exception ofsample 65-DS-4 matrix, the displayed Pitcairn data have MgO N 9%, and Hawaiian datahave MgO N 6.5%.
sample amantle source that is lithologically similar to themantle sourceof high-SiO2 Hawaiian lavas (Mauna Loa and Koolau).
4.3. Helium isotopic constraints on the mantle source beneath Pitcairn
4.3.1. Origin of the depleted high 3He/4He component hosted in Pitcairnlavas: entrained depleted upper mantle or a depleted plume (FOZO?)component?
Thenewdata provide awindow into the high 3He/4He component inthe Pitcairn suite. The Pitcairn lava (Pit-16) with the highest 3He/4He(12.6 Ra, ratio to atmosphere) is also the sample with the highest206Pb/204Pb that has been examined for helium isotopes (Fig. 5). Addi-tionally, with increasing 206Pb/204Pb, Pitcairn lavas exhibit increasinglydepleted geochemical signatures, including higher 143Nd/144Nd andlower 87Sr/86Sr (Fig. 4). While the high 3He/4He Pitcairn sample, Pit-16, is not in itself geochemically-depleted (143Nd/144Nd is 0.512590,87Sr/86Sr is 0.704687), it plots in a region of isotopic space that is closerto the most depleted Pitcairn component (but farther from the EM-1domain) than other lavas from Pitcairn that have been characterizedfor helium isotopes (Fig. 4). Therefore, going in the direction of thedepleted Pitcairn component in isotopic space, 3He/4He appears toincrease. Since MORB lavas tend to have 3He/4He ratios that clusternear 8±1Ra (e.g., Grahamet al., 1992), the new 3He/4Hemeasurementsuggests that the depleted component in the Pitcairn mantle is unlikethe typical uppermantlematerial sampled atmid-ocean ridges. Instead,thismixing componentmay be thehigh 3He/4He component that is pro-posed to reside in the lower mantle (Hart et al., 1992).
Based on radiogenic Sr-Nd-Pb isotopic compositions, four distinctmantle components EM-1, EM-2, HIMU and the depletedMORBmantleare suggested to encompass the global OIB dataset, and these end-members have been plotted as the vertices of a three-dimensional tetra-hedron in Sr-Nd-Pb isotopic space (e.g., Hart et al., 1992). Each hotspotforms arrays within the tetrahedron that appear to converge on acommon component with geochemically depleted Sr and Nd isotopiccompositions, and Hart et al. (1992) argued that this component, calledFOZO (Focus Zone), is distinct from (i.e., less geochemically-depletedthan) the MORB mantle and has high 3He/4He. Indeed, the Pitcairnlava with the highest 3He/4He is far from being as depleted as theMORB mantle.
Like other hotspot suites, geochemically-depleted lavas from thePitcairn hotspot appear to converge on the common component ofHart et al. (1992), but the range of 3He/4He data in the subset of deplet-ed Pitcairn lavas examined by Honda and Woodhead (2005) wereinconsistent with a FOZO component in the Pitcairn hotspot. Instead,Honda and Woodhead (2005) proposed that the MORB-mantle is thedepleted component sampled by geochemically-depleted Pitcairnlavas. The new 3He/4He isotopic data presented here suggest that thedepleted component in the Pitcairn mantle is not like the mantlesampled at mid-ocean ridges. Instead, the terrestrial mantle reservoirhosting high 3He/4He, like that sampled by sample Pit-16, is proposedto reside in the lower mantle (Hart et al., 1992). If this is the case,then the depleted Pitcairn component was incorporated into thePitcairn plume in the lower mantle, and is not a component entrainedin the upper mantle. While the 3He/4He ratios in the geochemically-depleted Pitcairn lavas are not extraordinarily elevated, they do providehints of a non-MORB component in the geochemically depleteddomains of the Pitcairn plume. Honda and Woodhead (2005) showedfrom neon isotope data that Pitcairn samples of clear EM-1 signaturehad a major component of solar neon, but the data did not fall nearthe N-MORB correlation line.
4.3.2. MORB-like 3He/4He in the EM-1 mantleA notable feature of the new dataset is the observation that the
Pitcairn lavaswith themost geochemically-enriched isotopic signatureshave 3He/4He similar to MORB. The two Pitcairn lavas that havethe most extreme EM-1 signatures (i.e., lowest 143Nd/144Nd and
9G. Garapić et al. / Lithos 228–229 (2015) 1–11
206Pb/204Pb and highest 87Sr/86Sr), and that have also been character-ized for helium isotopic ratios, are found to have MORB-like 3He/4He(Fig. 5). However, the other mantle end-members associated withrecycled crustal materials, EM-2 and HIMU, have 3He/4He ratios lowerthan MORB, as do lavas from several other hotspot localities(e.g., Tristan, Gough, Comores, Canaries) that do not sample end-member heavy radiogenic isotope compositions (Barfod et al., 1999;Class et al., 2005; Day and Hilton, 2011; Graham et al., 1992; Hanyuand Kaneoka, 1997; Jackson et al., 2007; Kurz et al., 1982; Moreiraet al., 2011; Parai et al., 2009). 3He/4He ratios range down to 4.3 Ra inSamoan EM-2 lavas with 87Sr/86Sr of 0.7186 (Jackson et al., 2007),(Jackson et al., 2009), and the low 3He/4He is interpreted to be the resultof the incorporation of terrigenous sediment with low 3He/4He (downto 0.01 Ra; (Podosek et al., 1980), (Ballentine and Burnard, 2002))into peridotitic mantle (N8 Ra); such sediment is U and Th-rich andwill generate 4He, and the mixture of sediment and mantle peridotitewill evolve relatively low 3He/4He in the EM-2 reservoir over time(Jackson et al., 2009). Similarly, lavas sampling the HIMU mantlegenerally have 3He/4He b 7 Ra (e.g., St. Helena, Mangaia, Tubuai, olderseries at Rurutu; (Graham et al., 1992), (Hanyu and Kaneoka, 1997),(Parai et al., 2009), (Hanyu et al., 2011)) or b8 in OIB with less extremeHIMU signatures (e.g., Canaries; (Day and Hilton, 2011)). The sub-MORB 3He/4He ratios may relate to the presence of recycled U and Th-rich oceanic crust into the HIMU mantle (e.g., (Hofmann and White,1982)). Note that this U–Th enrichment is opposite the low U/Pbassumed above for the eclogitic component of Koolau EM-1 lavas. Ifthe high HIMU mantle is host to recycled oceanic crust, which isdegassed and will have highly radiogenic 3He/4He (b0.1 Ra; (Moreiraand Kurz, 2001)), an important question is how lavas sampling thismantle reservoir have 3He/4He that is significantly higher than arecycled oceanic crust. A number of solutions have been presentedthat attempt to explain the observation of sub-MORB 3He/4He ratios inHIMU lavas that are not as low as degassed oceanic crust (e.g., Classet al., 2005; Day and Hilton, 2011; Hanyu and Kaneoka, 1997; Hanyuet al., 1999, 2011; Hart et al., 2008; Hilton et al., 2000; Moreira andKurz, 2001; Moreira et al., 2011; Parai et al., 2009).
Like the HIMU and EM-2 reservoirs, Pitcairn EM-1 lavas are alsosuggested to host recycled crustal components, including recycledsediment and oceanic crust. However, Pitcairn EM-1 lavas haveMORB-like 3He/4He, which is higher than end-member HIMU and EM-2 lavas and is therefore notable. Recycled crustal materials, like thosesuggested to exist in the EM-1mantle beneath Pitcairn, should generate4He by U and Th decay over time, and this should generate mantlereservoirs with low 3He/4He (b0.1 Ra). A clue to the apparent paradoxof MORB-like 3He/4He in the Pitcairn may be provided by the higherdiffusivity of He relative to the other radiogenic isotopic tracers (Sr,Nd and Pb). Owing to the high diffusivity of helium, Hart et al. (2008)suggested that mantle heterogeneities with length-scales smaller than5 to 10 km will not preserve their helium isotopic signatures overtime scales of 1.5 Ga, but their Sr, Nd and Pb isotopic signature willremain relatively unchanged. If the recycled sediment heterogeneitiesin the EM-1 mantle are distributed over relatively short length scales(5 to 10 km or less), and if the EM-1 mantle is sufficiently old(N1.5 Ga), then the helium isotopic composition of the EM-1 mantlemay become overwhelmed by the 3He/4He signature of the mantlereservoir hosting the recycled materials, here inferred to be ambientupper mantle, which is 8 Ra. Of course, this requires that the 4Heproduction of the recycled component is low compared to the ambientmantle helium concentration (Hart et al., 2008). If the time-integratedhelium concentrations of the upper mantle greatly exceed the concen-trations of the recycled oceanic crust and sediment (Day and Hilton,2011), and if the length scales of the recycled materials are small(≪5–10 km; (Hart et al., 2008)), the 3He/4He of the resulting mantlereservoir is more likely to have a 3He/4He similar to the ambientupper mantle. A diffusive decoupling mechanism like that proposedby Hart et al. (2008) may explain the presence of MORB-like 3He/4He
in a mantle reservoir that hosts recycled oceanic crust and sediments.Because neon diffuses many orders of magnitude slower than helium(in olivine, Cherniak et al. (2014)), the solar neon signature of Hondaand Woodhead (2005) could be retained in the recycled materialwhile aMORB-like helium is imprinted by diffusive exchange. Of course,a more sinister explanation would simply postulate a helium (but notSr, Nd or Pb) isotope exchange in the uppermost mantle betweenmigrating EM-1 melts (or magma chambers) and the ambient MORBmantle country rocks. It is of course possible that EM-1 has the samehelium value as MORB by coincidence.
5. Conclusions
The extreme enriched mantle EM-1 signature in Pitcairn lavas likelyhosts an enriched protolith derived from upper continental crust, andthis protolith exhibits many geochemical similarities with pelagicsediment. Elevated primary-melt Ti in the EM-1 lavas suggests thepresence of recycled pyroxenite in the mantle source. We identify atholeiitic sample (65-DS4) from Adams Seamount at the leading edge ofthe Pitcairn hotspot, and the tholeiitic composition of this sample comple-ments tholeiites previously described from the Pitcairn hotspot. The newtholeiitic Pitcairn seamount sample not only has the highest olivine-corrected SiO2 abundance in the Pitcairn dataset, but it also extends thePitcairn field closer to the Hawaiian Koolau field in Pb-isotopic spacethan any previously examined Pitcairn lava. The geochemical affinitiessuggest that there may be broad similarities including geochemically-enriched Sr andNd isotopic compositions and elevated SiO2 in themantlesources sampled by tholeiitic Pitcairn lavas and the EM-1 component inKoolau lavas. The origin of the SiO2-rich nature of Koolau lavas fromHawaii is argued to be the result of eclogite melting in the Hawaiianmantle plume. The isotopic and major element affinities between thePitcairn EM-1 tholeiitic sample 65DS-4 and theEM-1 component sampledby tholeiitic Koolau lavas fromHawaii suggests a common origin, andweargue that a silica-saturated mafic protolith is in the source of the EM-1mantle at both hotspots. We also present the highest 3He/4He (12.6 Ra)reported for Pitcairn lavas, and it is associated with a third componentwith the highest 206Pb/204Pb and the most geochemically-depleted87Sr/86Sr and 143Nd/144Nd ratios in the Pitcairn suite. We also show thatEM-1 end-member lavas fromPitcairn have 3He/4He that is indistinguish-able from the MORB mantle. If the recycled sedimentary and maficcomponents in the Pitcairn EM-1 occur at relatively short length scales,distributed in a peridotitematrixwith 3He/4He of 8 Ra, diffusive exchangeof helium over timescales N1.5 Ga will permit the 3He/4He of the PitcairnEM-1 mantle source to evolve toward values that typify the ambientupper mantle sampled by MORB. Alternatively, exchange of heliumbetween EM-1 melts and melt conduits in the upper mantle host rockmay result in EM-1 helium isotopic ratios evolving toward valuestypical of the upper mantle.
Acknowledgments
We acknowledge the seagoing efforts of Peter Stoffers, the chiefscientist during the 1989 cruise of the FS Sonne, and H. Craig, the chiefscientist of the Helious Expedition that visited Pitcairn Island in 1989.The authors acknowledge the constructive comments from Al Hofmannon a prior version of this manuscript. MGJ acknowledges support fromNSF grants OCE-1153894, EAR-1348082, EAR-1347377 and EAR-1145202.
References
Ballentine, C., Burnard, P., 2002. Production, release and transport of noble gases in thecontinental crust. Reviews in Mineralogy and Geochemistry 47, 481–538.
Barfod, D., Ballentine, C., Halliday, A., Fitton, J., 1999. Noble gases in the Cameroon line andthe He, Ne, and Ar isotopic compositions of high μ (HIMU) mantle. Journal ofGeophysical Research, Solid Earth 104, 29509–29527.
Brandenburg, J., Hauri, E.H., van Keken, P.E., Ballentine, S., 2008. A multiple-system studyof the geochemical evolution of the mantle with force-balanced plates and thermo-chemical effects. Earth and Planetary Science Letters. http://dx.doi.org/10.1016/j.epsl.2008.08.027.
Carter, R.M., 1967. The geology of Pitcairn Island, South Pacific Ocean. Vol. 231. BishopMuseum Press.
Chauvel, C., Hofmann, A.W., Vidal, P., 1992. HIMU-EM: The French Polynesian Connection.Earth and Planetary Science Letters 110, 99–119.
Cherniak, D., Thomas, J., Watson, E., 2014. Neon diffusion in olivine and quartz. ChemicalGeology 371, 68–82.
Class, C., Goldstein, S., Stute, M., Kurz, M., Schlosser, P., Grand, P., 2005. Comore island: awell-constrained “low 3He/4He” mantle plume. Earth and Planetary Science Letters233, 391–409.
Collerson, K.D., Williams, Q., Ewart, A.E., Murphy, D.T., 2010. Origin of HIMU and EM-1domains sampled by ocean island basalts, kimberlites and carbonatites: The role ofCO2-fluxed lower mantle melting in thermochemical upwellings. Physics of theEarth and Planetary Interiors 181, 112–131.
Day, J., Hilton, D., 2011. Origin of 3He/4He ratios in HIMU-type basalts constrained fromCanary Island lavas. Earth and Planetary Science Letters 305, 226–234.
Debaille, V., Tronnes, R., Brandon, A., Waight, T., Graham, D., Lee, C.T.A., 2009. Primitiveoff-rift basalts from Iceland and Jan Mayen: Os isotopic evidence for a mantle sourcecontaining enriched subcontinental lithosphere. Geochimica et Cosmochimica Acta73, 3423–3449.
Duncan, R., McDougall, I., Carter, R., Coombs, D., 1974. Pitcairn island— another Pacific hotspot? Nature 251, 619–682.
Eiler, J.M., Farley, K.A., Valley, J.W., Stolper, E.M., Hauri, E.H., Craig, H., 1995. Oxygenisotope evidence against bulk recycled sediment in the mantle sources of Pitcairn is-land lavas. Nature 377, 138–141.
Eiler, J., Farley, K., Valley, J., Hauri, E.H., Craig, H., Hart, S., Stolper, E., 1997. Oxygen isotopevariations in ocean island basalt phenocrysts. Geochimica et Cosmochimica Acta 61,2281–2293.
Eisele, J., Sharma, M., Galer, S.J.G., Blichert-Toft, S., Devey, C.W., Hofmann, S., 2002. The role ofsediment recycling in EM-1 inferred from Os, Pb, Hf, Nd, Sr isotope and trace elementsystematics of the Pitcairn hotspot. Earth and Planetary Science Letters 196, 197–212.
Escrig, S., Capmas, F., Dupre, B., Allegre, C.J., 2004. Osmium isotopic constraints onthe nature of the DUPAL anomaly from Indian mid-ocean-ridge basalts. Nature.http://dx.doi.org/10.1038/nature02904.
Farley, K.A., Natland, J.H., Craig, H., 1992. Binary mixing of enriched and undegassed(primitive?) mantle components (He, Sr, Nd, Pb) in Samoan lavas. Earth andPlanetary Science Letters 111, 183–199.
Farley, K., Basu, A., Craig, H., 1993. He, Sr and Nd isotopic variations in lavas from the JuanFernandez Archipelago, SE Pacific. Contributions to Mineralogy and Petrology 115,75–87.
Gasperini, D., Blichert-Toft, S., Bosch, D., DelMoro, A., Macera, D., Télouk, P., Albarède, F.,2000. Evidence from Sardinian basalt geochemistry for recycling of plume headsinto the Earth's mantle. Nature 408, 701–704.
Gast, P.W., Tilton, G.R., Hedge, C., 1964. Isotopic composition of lead and strontium fromAscension and Gough Islands. Science. http://dx.doi.org/10.1126/science.145.3637.1181.
Geldmacher, J., Hoernle, K., Klügel, A., van den Bogaard, P., Bindeman, I., 2008.Geochemistry of a new enriched mantle type locality in the northern hemisphere:Implications for the origin of the EM-I source. Earth and Planetary Science Letters265, 167–182.
Graham, D.W., Humphris, S.E., Jenkins, W.J., Kurz, M.D., 1992. Helium isotopegeochemistry of some volcanic rocks from Saint Helena. Earth and Planetary ScienceLetters 110, 121–131.
Guillou, H., Garcia, M., Turpin, L., 1997. Unspiked K–Ar dating of young volcanic rocksfrom Loihi and Pitcairn hot spot seamounts. Journal of Volcanology and GeothermalResearch 78, 239–249.
Hanan, B.B., Graham, D.W., 1996. Lead and Helium Isotope Evidence from Oceanic Basaltsfor a Common Deep Source of Mantle Plumes. Science 991–995.
Hanan, B.B., Blichert-Toft, J., Pyle, D.G., Christie, D.M., 2004. Contrasting origins of theupper mantle revealed by hafnium and lead isotopes from the Southeast IndianRidge. Nature 432, 91–94.
Hanyu, T., Kaneoka, I., 1997. The uniform and low 3He/4He ratios of HIMUbasalts as evidencefor their origin as recycled materials. Nature. http://dx.doi.org/10.1038/36835.
Hanyu, T., Kaneoka, I., Nagao, K., 1999. Noble gas study of HIMU and EM ocean islandbasalts in the Polynesian region. Geochimica et Cosmochimica Acta 63, 1181–1201.
Hanyu, T., Tatsumi, Y., Kimura, J.I., 2011. Constraints on the origin of the HIMU reservoirfrom He–Ne–Ar isotope systematics. Earth and Planetary Science Letters 307,377–386.
Hanyu, T., Kawabata, H., Tatsumi, Y., Kimura, J.I., Hyodo, H., Sato, K., Miyazaki, T., Chang, Q.,Hirahara, Y., Takahashi, T., Senda, R., Nakai, S., 2014. Isotope evolution in the HIMUreservoir beneath St. Helena: Implications for the mantle recycling of U and Th.Geochimica et Cosmochimica Acta 143, 232–252.
Hart, S.R., 2011. The Mantle Zoo: New Species, Endangered Species, Extinct Species.Mineralogical Magazine 3, 983.
Hart, S.R., Blusztajn, J.S., 2006. Age and geochemistry of the mafic sills, ODP site 1276,Newfoundland margin. Chemical Geology 235 (3-4). http://dx.doi.org/10.1016/j.chemgeo.2006.07.001.
Hart, S.R., Kurz, M., Wang, Z., 2008. Scale length of mantle heterogeneities: Constraintsfrom helium diffusion. Earth and Planetary Science Letters. http://dx.doi.org/10.1016/j.epsl.2008.03.010.
Hauri, E.H., 1996. Major-element variability in the Hawaiian mantle plume. Nature 382,415–419.
Hauri, E.H., Hart, S.R., 1993. Re–Os isotope systematics of HIMU and EMII oceanic islandbasalts from the south Pacific Ocean. Earth and Planetary Science Letters 114,353–371.
Hauri, E.H., Kurz, M., 1997. Melt migration and mantle chromatography, 2: a time seriesOs isotope study of Mauna Loa volcano, Hawaii. Earth and Planetary Science Letters153, 21–36.
Hekinian, R., Cheminee, J.L., Dubois, J., Stoffers, P., Scott, S., Guivel, C., Garbe-Schonberg, D.,Devey, S., Bourdon, B., Lackschewitz, K., McMurty, S., Drezen, E.L., 2003. The Pitcairnhotspot in the South Pacific: distribution and composition of submarine volcanicsequences. Journal of Volcanology and Geothermal Research 121, 219–245.
Herzberg, C., 2006. Petrology and thermal structure of the Hawaiian plume from MaunaKea volcano. Nature 444. http://dx.doi.org/10.1038/natur e05254.
Herzberg, C., 2011. Identification of Source Lithology in the Hawaiian and Canary Islands:Implications for Origins. Journal of Petrology 113–146.
Hilton, D., Macpherson, C., Elliott, T., 2000. Helium isotope ratios inmafic phenocrysts andgeothermal fluids from La Palma, the Canary Islands (Spain): implications for HIMUmantle sources. Geochimica et Cosmochimica Acta 64, 2119–2132.
Hofmann, A.W., 2003. Sampling mantle heterogeneity through oceanic basalts: Isotopesand trace elements. In: Carlson, R.W. (Ed.), The Mantle and Core. Treatise onGeochemistry Vol. 2. Elsevier, New York, pp. 1–44.
Hofmann, A.W., White, W.M., 1982. Mantle plumes from ancient oceanic crust. Earth andPlanetary Science Letters 57, 421–436.
Honda, M., Woodhead, J.D., 2005. A primordial solar-neon enriched component in thesource of EM-I-type ocean island basalts from the Pitcairn Seamounts, Polynesia.Earth and Planetary Science Letters 597–612.
Huang, S., Frey, F.A., 2005. Recycled oceanic crust in the Hawaiian Plume: evidence fromtemporal geochemical variations within the Koolau Shield. Contributions toMineralogy and Petrology 149, 556–575.
Jackson, M.G., Dasgupta, R., 2008. Compositions of HIMU, EM1, and EM2 from globaltrends between radiogenic isotopes and major elements in ocean island basalts.Earth and Planetary Science Letters 276, 175–186.
Jackson, M.G., Shirey, S.B., 2011. Re–Os systematics in Samoan shield lavas and the use ofOs-isotopes in olivine phenocrysts to determine primary magmatic compositions.Earth and Planetary Science Letters 312, 91–101.
Jackson, M.G., Hart, S.R., Koppers, A.A.P., Staudigel, S., Konter, J., Blusztajn, S., Kurz, M.,Russell, S., 2007. The return of subducted continental crust in Samoan lavas. Nature448. http://dx.doi.org/10.1038/natu re06048.
Jackson, M.G., Hart, S.R., Saal, A.E., Shimizu, N., Skovgaard, A.C., 2008. Globally elevatedtitanium, tantalum, and niobium (TITAN) in ocean island basalts with high3He/4He. Geochemistry, Geophysics, Geosystems Q04027. http://dx.doi.org/10.1029/20 07GC001876.
Jackson, M.G., Kurz, M.D., Hart, S.R., 2009. Helium and neon isotopes in phenocrysts fromSamoan lavas: Evidence for heterogeneity in the terrestrial high 3He/4He mantle.Earth and Planetary Science Letters 287, 519–528.
Jackson, M.G., Weis, D., Huang, S., 2012. Major element variations in Hawaiian shieldlavas: Source features and perspectives from global ocean island basalt (OIB)systematics. Geochemistry, Geophysics, Geosystems 13 (9). http://dx.doi.org/10.1029/2012GC004268.
Johnson, D., Hooper, P., Conrey, R., 1999. XRF analysis of rocks and minerals for major andtrace elements on a single low dilution Li-tetraborate fused bead. Advances in X-RayAnalysis 41, 843–867.
Kawabata, H., Hanyu, T., Chang, Q., Kimura, J.I., Nichols, A., Tatsumi, Y., 2011. The Petrologyand geochemistry of St. Helena alkali basalts: Evaluation of the oceanic crust-recycling model for HIMU OIB. Journal of Petrology 52. http://dx.doi.org/10.1093/petrol ogy/egr003.
Knaack, C., Cornelius, S., Hooper, P., 1994. Trace element analyses of rocks andminerals byICP-MS. open-file report. Washington State University.
Kurz, M., Jenkins, W., Hart, S., 1982. Helium isotopic systematics of oceanic islands andmantle heterogeneity. Nature 297, 43–47.
Lassiter, J.C., Hauri, E.H., 1998. Osmium-isotope variations in Hawaiian lavas: evidence forrecycled oceanic lithosphere in the Hawaiian plume. Earth and Planetary ScienceLetters 164, 483–496.
Macdonald, G.A., Katsura, T., 1964. Chemical Composition of Hawaiian Lavas. Journal ofPetrology 5. http://dx.doi.org/10.1093/petrology/5.1.82.
McKenzie, D., O'Nions, R.K., 1983. Mantle reservoirs and ocean island basalts. Nature 301,229–231.
Menzies, M., 1983. Mantle Ultramafic Xenoliths in Alkaline Magmas: Evidence for MantleHeterogeneity Modified by Magmatic Activity. Continental Basalts and MantleXenoliths. Shiva Publishing, pp. 92–110.
Menzies, M., Murthy, V.R., 1980. Nd and Sr isotope geochemistry of hydrousmantle nodules and their host alkali basalts: Implications for local heterogene-ities in metasomatically veined mantle. Earth and Planetary Science Letters 46,323–334.
Moreira, M., Kurz, M., 2001. Subducted oceanic lithosphere and the origin of the ‘high μ’basalt helium isotopic signature. Earth and Planetary Science Letters 189, 49–57.
Moreira, M., Kanzari, A., Madureira, P., 2011. Helium and neon isotopes in São Miguelisland basalts, Azores Archipelago: New constraints on the “low 3He” hotspot origin.Chemical Geology 322–323, 91–98.
Niu, Y., O'Hara, M.J., 2003. Origin of ocean island basalts: A new perspective frompetrology, geochemistry, and mineral physics considerations. Journal of GeophysicalResearch. http://dx.doi.org/10.1029/2002JB002048.
Parai, R., Mukhopadhyay, S., Lassiter, J.C., 2009. New constraints on the HIMUmantle fromneon and helium isotopic compositions of basalts from the Cook–Austral Islands.Earth and Planetary Science Letters 277, 253–261.
Pilet, S., Baker, M.B., Müntener, O., Stolper, S., 2011. Monte Carlo Simulations ofMetasomatic Enrichment in the Lithosphere and Implications for the Source ofAlkaline Basalts. Journal of Petrology. http://dx.doi.org/10.1093/petrology/egr007.
Plank, T., Langmuir, C.H., 1998. The chemical composition of subducting sediment and itsconsequences for the crust and mantle. Chemical Geology 145, 325–394.
Podosek, F., Honda, M., Ozima, M., 1980. Sedimentary noble gases. Geochimica etCosmochimica Acta 44, 1875–1884.
Prytulak, J., Elliot, T., 2007. TiO2 enrichment in ocean island basalts. Earth and PlanetaryScience Letters 263, 388–403.
Rehkämper, M., Hofmann, S., 1997. Recycled ocean crust and sediment in Indian OceanMORB. Earth and Planetary Science Letters 147, 93–106.
Reisberg, L., Zindler, A., Marcantonio, F., White, W., Wyman, D., Weaver, B., 1993. Osisotope systematics in ocean island basalts. Earth and Planetary Science Letters 120,149–167.
Rhodes, J., 1988. Geochemistry of the 1984 Mauna Loa eruption: Implications for magmastorage and supply. Journal of Geophysical Research 93, 4,453–454,466.
Rudnick, R.L., Gao, S., 2003. Composition of the continental crust. In: Rudnick, R.L. (Ed.),Treatise on Geochemistry. Vol. 3. Elsevier, New York, pp. 1–64.
Salters, V.J., Sachi-Kocher, S., 2010. An ancient metasomatic source for the Walvis Ridgebasalts. Chemical Geology. http://dx.doi.org/10.1016/j.chemgeo.2010.02.010.
Shirey, S., Walker, R., 1995. Carius tube digestion for low-blank rhenium–osmiumanalysis. Analytical Chemistry 67 (13), 2136–2141.
Sobolev, A.V., Hofmann, A.W., Sobolev, S.V., Nikogosian, S., 2005. An olivine-free mantlesource of Hawaiian shield basalts. Nature 590–597.
Stoffers, P., Hekinian, R., Ackermand, D., Ackermand, D., Binard, N., Botz, R., Devey, C.,Hansen, D., Hodkinson, R., Jeschke, G., Lange, J., v Perre, E., Scholten, J., Schmitt, M.,Sedwick, P., Woodhead, J.D., 1990. Active Pitcairn hotspot found. Marine Geology95, 51–55.
Stracke, A., 2012. Earth's heterogeneous mantle: A product of convection-driveninteraction between crust and mantle. Chemical Geology 10, 330–331. http://dx.doi.org/10.1016/j.chemgeo.2012.08.007.
Tatsumi, Y., 2000. Continental crust formation by crustal delamination in subductionzones and complementary accumulation of the enriched mantle I component in themantle. Geochemistry, Geophysics, Geosystems 1. http://dx.doi.org/10.1029/2000GC000094.
Todt, W., Cliff, R.A., Hanser, A., Hofmann, A.W., 1996. Evaluation of a 202Pb–205Pbdouble spike for high precision lead isotope analysis. In: Basu, A., Hart, S.(Eds.), Earth processes: reading the isotopic code. AGU Geophysical MonographVol. 95. http://dx.doi.org/10.1029/GM095P0429.
Weaver, B.L., 1991. The origin of ocean island basalt end-member compositions: traceelement and isotopic constraints. Earth and Planetary Science Letters 104, 381–397.
White, W.M., 2010. Oceanic Island Basalts and Mantle Plumes: The GeochemicalPerspective. Annual Review of Earth and Planetary Sciences. Annual Reviews.http://dx.doi.org/10.1146/annurev-earth-040809-152450.
Willbold, M., Stracke, A., 2010. Formation of enriched mantle components by recycling ofupper and lower continental crust. Chemical Geology 276, 188–197.
Woodhead, J.D., Devey, C.W., 1993. Geochemistry of the Pitcairn seamounts, I: sourcecharacter and temporal trends. Earth and Planetary Science Letters 81–99.
Woodhead, ï.D., McCulloch, S., 1989. Ancient seafloor signals in Pitcairn Island lavas andevidence for large amplitude, small length-scale mantle heterogeneities. Earth andPlanetary Science Letters 94, 257–273.
Woodhead, J.D., GreenWood, P., Harmon, R.S., Stoffers, P., 1993. Oxygen isotope evidencefor recycled crust in the source of EM-type ocean island basalts. Nature 362. http://dx.doi.org/10.1038/36280 9a0.
Workman, R.K., Hart, S.R., Jackson, M.G., Regelous, M., Farley, K.A., Blusztajn, J., Kurz, J.,Staudigel, H., 2004. Recycled metasomatized lithosphere as the origin of the EnrichedMantle II (EM2) end-member: Evidence from the Samoan Volcanic Chain.Geochemistry, Geophysics, Geosystems. http://dx.doi.org/10.1029/2003GC000623.
Workman, R.K., Eiler, J.M., Hart, S.R., Jackson, M.G., 2008. Oxygen isotopes in Samoanlavas: Confirmation of continent recycling. Geology. http://dx.doi.org/10.1130/G24558A.1.
Zindler, A., Hart, S., 1986. Chemical Geodynamics. Annual Review of Earth and PlanetarySciences 14, 493–571.
Zindler, A., Jagoutz, E., Goldstein, S., 1982. Nd, Sr and Pb isotopic systematics in a three-component mantle: A new perspective. Nature 1036 (298), 519–523.
