Elsevier

Lithos

Volumes 170–171, June 2013, Pages 224-238
Lithos

A composite, isotopically-depleted peridotite and enriched pyroxenite source for Madeira magmas: Insights from olivine

https://doi.org/10.1016/j.lithos.2013.03.002Get rights and content

Highlights

  • Nickel, Mn and Ca in olivine represent powerful tracers in unraveling deep mantle processes.

  • Isotopically depleted peridotite and enriched pyroxenite compose the Madeira magma source.

  • The amount of the recycled crust in the Madeira magma source varies but does not exceed ~ 10%.

  • The recycled crust endmember appears to be common to both the Madeira and Canary hotspots.

Abstract

The Madeira and Canary island/seamount chains represent two adjacent hotspot tracks in the eastern North Atlantic, which were derived from mixed peridotite–pyroxenite mantle sources. They possess systematically different Sr–Nd–Pb, Os and Hf isotope signatures, implying mixing of DMM-, HIMU- and EM-type components from different mantle lithologies. The lithological nature of these postulated mantle endmembers (e.g., if formed by peridotite, pyroxenite and/or eclogite) is still a subject of debate. We studied the chemical composition of olivine phenocrysts (focusing on their Ni, Mn and Ca concentrations) from the lavas covering the entire volcanic history of the Madeira Archipelago (~ 0–5 Ma). We demonstrate that Ni × FeO/MgO and Mn/FeO ratios and Ca-concentrations of olivine correlate with Sr, Nd and Pb isotopic compositions of their host lavas. The amounts of peridotite- and pyroxenite-derived melt fractions in the parental Madeira magmas were inferred from olivine compositions and independently confirmed by modeling of two-component source melting using trace elements. Our calculations demonstrate that the amount of recycled crust (= eclogite) in the Madeira magma source varies but does not exceed ~ 10%. Strong linear relationships between chemical composition of olivine and radiogenic isotopes of the host lavas allow us to determine the isotopic composition of peridotite and pyroxenite endmembers of the Madeira hotspot. The peridotite endmember has a highly depleted isotopic composition 87Sr/86Sr = 0.7022–0.7026, 143Nd/144Nd = 0.5132–0.5135, 206Pb/204Pb = 17.4–18.6, 207Pb/204Pb = 15.4–15.49, and 208Pb/204Pb = 36.8–38.2, whereas the pyroxenite endmember has an enriched composition 87Sr/86Sr = 0.7031–0.7035, 143Nd/144Nd = 0.5127–0.5130, 206Pb/204Pb = 19.9–21.3, 207Pb/204Pb = 15.59–15.69, and 208Pb/204Pb = 39.6–41.2. Our new data confirm the existence of recycled (pyroxenitic) lithology in the Madeira magma source that was previously interpreted to represent relatively young (< 1 Ga) recycled oceanic crust. The isotopically depleted, DMM-like (peridotitic) component sampled by the younger post-erosional stage magmas is thought to reflect the ultramafic portion of the recycled oceanic lithosphere, whereas the HIMU-type pyroxenite of similar age to the peridotite is derived from recycled oceanic crust. Combining our new results with the data from the Canary hotspot, we conclude that the < 1 Ga recycled crustal component is common to both plume sources. The Canary volcanism also reflects the presence of older (> 1 Ga) recycled crust and a contribution from African subcontinental lithosphere.

Introduction

Identification of source components and determination of their relative contribution to the origin of mafic, mantle-derived magmas represent one of the main petrologic problems in understanding the evolution of the Earth. The existing systematic difference in radiogenic isotope composition between mid-ocean ridge basalts (MORB) and ocean island basalts (OIB) is traditionally explained by the presence of isotopically different mantle components. These include high time-integrated U/Pb mantle (HIMU), enriched mantle one (EM1) and two (EM2) (Hart, 1988, White, 1985, Zindler and Hart, 1986, Zindler et al., 1982), whose origin is ascribed to the recycling of oceanic crust, marine sediments and continental lithosphere through the deep mantle back to the surface (Allègre and Turcotte, 1986, Hofmann and White, 1982, White and Hofmann, 1982), in addition to the depleted N-MORB source (DMM) and ubiquitous (FOZO or C) mantle components (Galer and O'Nions, 1985, Hanan and Graham, 1996, Hart et al., 1992). An open question concerns the lithologies (e.g., pyroxenite, eclogite and/or peridotite) that host these endmembers in the Earth's mantle.

A number of studies (e.g., Dasgupta et al., 2007, Hirschmann and Stolper, 1996, Ito and Mahoney, 2005a, Ito and Mahoney, 2005b, Kawabata et al., 2011, Mallik and Dasgupta, 2012, Phipps Morgan and Morgan, 1999, Sims and DePaolo, 1997, Sobolev et al., 2005, Sobolev et al., 2007, Sobolev et al., 2011, Stracke and Bourdon, 2009, among others) have investigated the effects of two- (and more) component melting of the ascending mantle, dealing with mineralogically, chemically and isotopically distinct lithologies. Recently, Sobolev et al., 2005, Sobolev et al., 2007 proposed a method for the quantitative assessment of the presumed recycled material in the source of intraplate magmas. The basic assumption of this approach is that quartz-normative melts, which are generated by partial melting of eclogite (the high-P and -T manifestation of subducted ocean crust in the mantle), react with ambient mantle peridotite and convert it to olivine-free “reaction pyroxenite”. If the newly formed reaction pyroxenite subsequently melts along with ambient peridotite due to further upwelling and adiabatic decompression, the pyroxenite-derived melts should be enriched in Ni and depleted in Mn, as compared to melts generated by peridotite alone (because Ni is less and Mn is more compatible in pyroxenite than in peridotite). Mixing and pooling of both (i.e., peridotite- and pyroxenite-derived) melts lead to a “hybrid” parental magma, from which olivine will eventually crystallize at shallower depth. As demonstrated by Sobolev et al., 2005, Sobolev et al., 2007, the Ni and Mn contents of this olivine can be used to determine the original proportions of the contributing parental lithologies and, thus, to infer the amount of recycled material in the magma source.

Most of the lavas analyzed by Sobolev et al., 2005, Sobolev et al., 2007 were tholeiitic or transitional basalts from large oceanic islands and continental lava flows. Therefore, the proposed model best explains the origin of silica-saturated mantle-derived magmas. The question as to whether a similar mechanism can produce silica-undersaturated alkali basaltic lavas, which are common in the late shield and rejuvenated volcanic stages of almost all oceanic island volcanoes (e.g., Hawaii, Azores, Madeira, Cape Verde and Canary Islands) was shortly reviewed by Sobolev et al. (2007) and then in the later studies by Pilet et al. (2008) and Aulinas et al. (2010) among others. In particular, Sobolev et al. (2007) proposed several possible scenarios for generating silica-undersaturated melts within the framework of the “reaction pyroxenite” model. These are: (1) a CO2-trigerred melting of peridotite at temperatures below the solidus of hybrid pyroxenite melts (e.g., Gudfinnsson and Presnall, 2005); (2) melting of hybrid pyroxenite in contact with peridotite at temperatures lower than the solidus of pyroxenite alone (e.g., Herzberg, 2006); (3) melting of silica-deficient mafic garnet pyroxenite or bimineralic eclogite (no free silica phase is present) possibly formed from silica-undersaturated recycled crust (e.g., Hirschmann et al., 2003, Kogiso and Hirschmann, 2006), and (4) possible retention of low-degree, highly viscous, silica-rich eclogite-derived melts not in contact with peridotite, preventing the formation of reaction pyroxenite (Sobolev et al., 2005). Alternatively, Pilet et al. (2008) suggested a model where the recycled components in the sources of ocean island alkali basalts can be derived from veined/metasomatized oceanic or continental lithosphere, which is in fact very similar to the model previously proposed for Canary Islands by Lundstrom et al. (2003).

Since the behavior of Ni and Mn is source-indicative in the “hybrid” pyroxenite model and the Earth's peridotitic mantle and subducted/recycled crustal material have substantially different radiogenic isotope signatures (e.g., Hart, 1988, Zindler and Hart, 1986), coherent relationships are expected between the Ni and Mn concentrations of olivine phenocrysts and the isotopic composition of their host lavas, if the “reaction pyroxenite” model is correct. Indeed, significant correlations between Ni, Mn (and also Ca) contents of olivine phenocrysts and Sr, Nd, Pb and Os isotope ratios in their host lavas have been described for Iceland, the Hawaiian and the Canary Islands, allowing the chemical (isotopic) compositions of their recycled components to be constrained (Gurenko et al., 2009, Gurenko et al., 2010, Sobolev et al., 2008).

Here, we report a new dataset for the elemental composition of olivine phenocrysts and demonstrate their links to the Sr–Nd–Pb isotopic compositions of the host lavas from the Madeira Archipelago. Madeira is well-suited to further test the relationship between olivine composition and bulk radiogenic isotope ratios, because the geochemistry of its lavas is interpreted to reflect progressive melting of a heterogeneous mantle plume source containing both fertile (recycled pyroxenite/eclogite) and depleted (peridotite) lithologies derived from recycled oceanic lithosphere (Geldmacher and Hoernle, 2000, Geldmacher et al., 2006), but without a significant contribution of DMM from the upper mantle. Furthermore, the Madeira Archipelago is located only ~ 500 km north of the Canary Islands, also a hotspot system, but the lavas from both archipelagos have distinct Sr–Nd–Pb, Os and Hf isotope signatures (e.g., Geldmacher et al., 2001, Geldmacher et al., 2005, Geldmacher et al., 2011, Hoernle and Tilton, 1991, Hoernle et al., 1991, Hoernle et al., 1995, Widom et al., 1999). The purpose of this study was to investigate (1) if similar coherent relationships between olivine composition and radiogenic isotopes are present in Madeira lavas, as were found in the neighboring Canary Island lavas (Gurenko et al., 2009, Gurenko et al., 2010); (2) if independent modeling of two-component source melting using trace elements in the host lavas confirms the results obtained from the olivine studies; and (3) if the composition of recycled components in the Madeira and Canary magma sources matches that proposed by Geldmacher and Hoernle (2000). Finally, this work aims to place additional constraints on the origin of the isotopically depleted peridotitic component in the Madeira plume and to estimate the possible amount of recycled crust in the magma source.

Section snippets

Geological setting

The Madeira Archipelago in the eastern North Atlantic is located on 140 Ma old oceanic crust at the end of a slightly curved NE–SW oriented age-progressive chain of islands and seamounts interpreted as a > 70 Ma-old hotspot track (Fig. 1). It consists of five islands: (1) the island of Madeira (with the age of volcanic rocks ranging from 0 to > 5.2 Ma), (2) Desertas Islands (Ilhéu Chão, Deserta Grande and Bugio; 1.9–5.1 Ma), which are situated on a submarine ridge extending more than 60 km SSE from

Mineral chemistry

Olivine phenocrysts were analyzed for major (Si, Fe, Mg) and minor (Ti, Al, Cr, Mn, Ca, Ni, Co) elements, using the JEOL Superprobe JXA-8200 electron microprobe (Max Planck Institute for Chemistry, Mainz, Germany) and the technique described by Sobolev et al. (2007) and Gurenko et al. (2009). Forsterite (Fo) contents, Ni, Mn and Ca concentrations and Ni × FeO/MgO and Mn/FeO ratios of olivine (Ol) phenocrysts together with bulk-rock radiogenic isotope data are listed in Table 1, Table 2. The

Amount of the recycled crustal component in the source of Madeira magmas

The original weight fractions of reaction pyroxenite, peridotite and eclogite restite in the mantle source, and thus the amount of recycled material in the mantle source can be computed using mass balance, if the following parameters are defined (Sobolev et al., 2005):

  • 1.

    The proportion of pyroxenite-derived melt in the parental magma (Xpxm), which can be directly acquired from Ni × FeO/MgO and Mn/FeO ratios in olivine using the parameterization of Sobolev et al. (2008) (Table 2).

  • 2.

    The respective

Summary and conclusions

As previously shown for Iceland (Sobolev et al., 2008), the Canary Islands (Gurenko et al., 2009, Gurenko et al., 2010) and presently demonstrated for Madeira, the concentrations of Ni, Mn and Ca in early-crystallized olivine phenocrysts represent powerful tracers in unraveling deep mantle processes. We observe strong linear relationships of Ni and Mn concentrations normalized to FeO and MgO contents and absolute Ca concentrations in olivine with the Sr–Nd–Pb radiogenic isotope ratios of their

Acknowledgments

We thank the Museum of Natural History (Washington, DC, USA) for providing us with the standards for electron microprobe analysis. We thank Meritxell Aulinas and one anonymous referee for their thorough reviews and constructive comments that helped us to improve the manuscript. Editorial handling by Andrew Kerr is gratefully acknowledged. We acknowledge financial support from the Max Planck Society, the DFG grants HO1833/5 and HO1833/8 to KAH and the Chair of Excellence grant ANR-09-CEXC-003-01

References (76)

  • A.A. Gurenko et al.

    Major, trace element and Nd-Sr-Pb-O-He-Ar isotope signatures of shield stage lavas from the central and western Canary Islands: Insights into mantle and crustal processes

    Chemical Geology

    (2006)
  • A.A. Gurenko et al.

    Enriched, HIMU-type peridotite and depleted recycled pyroxenite in the Canary plume: a mixed-up mantle

    Earth and Planetary Science Letters

    (2009)
  • A.N. Halliday et al.

    Incompatible trace elements in OIB and MORB and source enrichment in the sub-oceanic mantle

    Earth and Planetary Science Letters

    (1995)
  • S.R. Hart

    Heterogeneous mantle domains: signatures, genesis and mixing chronologies

    Earth and Planetary Science Letters

    (1988)
  • K. Hoernle et al.

    Sr–Nd–Pb isotopic evolution of Gran Canaria: evidence for shallow enriched mantle beneath the Canary Islands

    Earth and Planetary Science Letters

    (1991)
  • A.W. Hofmann et al.

    Mantle plumes from ancient oceanic crust

    Earth and Planetary Science Letters

    (1982)
  • G. Ito et al.

    Flow and melting of a heterogeneous mantle: 1. Method and importance to the geochemistry of ocean island and mid-ocean ridge basalts

    Earth and Planetary Science Letters

    (2005)
  • G. Ito et al.

    Flow and melting of a heterogeneous mantle: 2. Implications for a chemically nonlayered mantle

    Earth and Planetary Science Letters

    (2005)
  • T. Kogiso et al.

    Partial melting experiments of bimineralic eclogite and the role of recycled mafic oceanic crust in the genesis of ocean island basalts

    Earth and Planetary Science Letters

    (2006)
  • C.C. Lundstrom et al.

    U-series disequilibria in volcanic rocks from the Canary Islands: plume versus lithospheric melting

    Geochimica et Cosmochimica Acta

    (2003)
  • A. Mallik et al.

    Reaction between MORB-eclogite derived melts and fertile peridotite and generation of ocean island basalts

    Earth and Planetary Science Letters

    (2012)
  • B. Peucker-Ehrenbrink et al.

    Hydrothermal lead transfer from mantle to continental crust: the role of metalliferous sediments

    Earth and Planetary Science Letters

    (1994)
  • J. Phipps Morgan et al.

    Two-stage melting and the geochemical evolution of the mantle: a recipe for mantle plum-pudding

    Earth and Planetary Science Letters

    (1999)
  • S. Pilet et al.

    The metasomatic alternative for ocean island basalt chemical heterogeneity

    Earth and Planetary Science Letters

    (2005)
  • S. Schwarz et al.

    Internal structure and evolution of a volcanic rift system in the eastern North Atlantic: the Desertas rift zone, Madeira archipelago

    Journal of Volcanology and Geothermal Research

    (2005)
  • K.W.W. Sims et al.

    Inferences about magma sources from incompatible element concentration ratios in oceanic basalts

    Geochimica et Cosmochimica Acta

    (1997)
  • A. Stracke et al.

    The importance of melt extraction for tracing mantle heterogeneity

    Geochimica et Cosmochimica Acta

    (2009)
  • M.F. Thirlwall

    Pb isotopic and elemental evidence for OIB derivation from young HIMU mantle

    Chemical Geology

    (1997)
  • R.K. Workman et al.

    Major and trace element composition of the depleted MORB mantle (DMM)

    Earth and Planetary Science Letters

    (2005)
  • C.J. Allègre et al.

    Implications of a two-component marble-cake mantle

    Nature

    (1986)
  • R. Dasgupta et al.

    Partial melting experiments of peridotite + CO2 at 3 GPa and genesis of alkalic ocean island basalts

    Journal of Petrology

    (2007)
  • J.M.D. Day et al.

    Pyroxenite-rich mantle formed by recycled oceanic lithosphere: oxygen-osmium isotope evidence from Canary Island lavas

    Geology

    (2009)
  • S.J.G. Galer et al.

    Residence time of thorium, uranium and lead in the mantle with implications for mantle convection

    Nature

    (1985)
  • J. Geldmacher et al.

    The 40Ar/39Ar age dating of the Madeira Archipelago and hotspot track (eastern North Atlantic)

    Geochemistry, Geophysics, Geosystems

    (2000)
  • J. Geldmacher et al.

    Hafnium isotopic variations in East Atlantic intraplate volcanism

    Contributions to Mineralogy and Petrology

    (2011)
  • G.H. Gudfinnsson et al.

    Continuous gradations among primary carbonatitic, kimberlitic, melilititic, basaltic, picritic, and komatiitic melts in equilibrium with garnet lherzolite at 3–8 GPa

    Journal of Petrology

    (2005)
  • A.A. Gurenko et al.

    Crust-primitive magma interaction beneath neovolcanic rift zone of Iceland recorded in gabbro xenoliths from Midfell, SW Iceland

    Contributions to Mineralogy and Petrology

    (2006)
  • A.A. Gurenko et al.

    Source components of the Gran Canaria (Canary Islands) shield stage magmas: evidence from olivine composition and Sr–Nd–Pb isotopes

    Contributions to Mineralogy and Petrology

    (2010)
  • Cited by (19)

    • Petrogenesis of shield volcanism from the Juan Fernández Ridge, Southeast Pacific: Melting of a low-temperature pyroxenite-bearing mantle plume

      2019, Geochimica et Cosmochimica Acta
      Citation Excerpt :

      Considering that the volume of pyroxenites can make up to 10% of the mantle source (Pearson and Nixon, 1996; Stracke et al., 2003), in addition to the above-mentioned evidence, the involvement of pyroxenite in the JFR source is plausible. The importance of pyroxenite in the source has been reported for many intraplate oceanic islands such as the Canary Islands (e.g., Gurenko et al., 2006, 2009, 2010; Day et al., 2009; Klügel et al., 2017), Hawaii (e.g., Herzberg, 2006; Garcia et al., 2015), Madeira (Gurenko et al., 2013), Mauritius (Paul et al., 2005), and the Galápagos (Trela et al., 2015). Considering the proposed presence of pyroxenite in the source, we attempted to quantify its abundance in the mantle plume beneath the JFR and determine the thermodynamic parameters for the peridotite/pyroxenite source.

    • Mineralogical and geochemical evidence for polybaric fractional crystallization of continental flood basalts and implications for identification of peridotite and pyroxenite source lithologies

      2018, Earth-Science Reviews
      Citation Excerpt :

      Another effect of crystallization of peridotite-derived magmas at the Ol + Cpx + L cotectic is to drive liquids to progressively more SiO2-undersaturation with decreasing Mg#. Consequently, crystallization of a peridotite-derived melt at ~ 1.0 GPa can result in a CaO depleted, Si-undersaturated liquid composition that is indistinguishable from a magma derived from the melting of low SiO2 pyroxenite such as those found at the Canary Islands and Madeira (Herzberg, 2011; Gurenko et al., 2013). Furthermore, recycled pyroxenite is also known to be far more heterogeneous than mantle peridotite, both chemically and mineralogically, and produces a diverse range of derivative magmas that may be very similar to peridotite-derived magmas (Keshav et al., 2004; Herzberg, 2011; Lambart et al., 2012, 2013, 2016).

    View all citing articles on Scopus
    View full text