Primitive Mantle Composition Research Articles

Large igneous province magma petrogenesis from source to surface: platinum-group element evidence from Ontong Java Plateau basalts recovered during ODP Legs 130 and 192

Abstract A total of 16 Ontong Java Plateau (OJP) basalt samples from Ocean Drilling Program Legs 192 and 130 were analysed for major, trace and platinum-group elements (PGEs; Ir, Ru, Rh, Pt and Pd). Major- and trace-element compositions determined by our study confirm Leg 192 shipboard analyses that indicated a new group of more primitive or ‘Kroenke-type’ basalts, with higher MgO, Ni and Cr, and lower incompatible-element, abundances than the more common Kwaimbaita-type basalts. The PGE abundances quantified here extend the range of the continuum of compositions found in previously analysed OJP basalts and are similar to those present in some komatiites. The PGEs, therefore, cannot be used to differentiate definitively between OJP basalts groups. The two samples analysed from Leg 130 (one from Site 803 and one from Site 807) are akin to the Kwaimbaita-type basalts. Low-temperature alteration has not affected Pd abundances in the Leg 192 basalts as it has in the Solomon Island and the Leg 130 samples. Elemental abundances and ratios along with petrography reveal that the OJP basalts have not experienced sulphide saturation. Positive correlations of Ir and Ru with Cr and Ni attest to the lithophile behaviour of the PGEs and lend more credence to studies suggesting compatibility of these elements in oxide and silicate phases, such as Cr-spinel and olivine. Estimates of sulphur abundance in the mantle, degree of partial melting and pressure of melt initiation were used in conjunction with the model of Mavrogenes & O’Neill to calculate a minimum initial excess temperature of +185–+235°C (1465–1515°C at 3.5–4.0 GPa) above ambient mantle for the OJP source. This is in broad agreement with a fossil geotherm preserved in megacrysts and peridotite xenoliths found in pipe-like intrusives of alnöite that outcrop on the island of Malaita, Solomon Islands. Using the PGEs as a guide, the OJP basalts were modelled using a three-source component melt mix: a 10% garnet peridotite melt of primitive mantle composition, which then passed through the garnet-spinel transition and melted a further 20%, a 30% partial melt of fertile upper mantle and 0–1% of outer core material. The core component was included only in the plume source, and the ratio of plume source to upper mantle source was 19: 1. It is evident from this study that the PGE contents of at least some of the OJP basalts are too high to be generated by primitive mantle sources alone. A PGE-enriched component is required and we suggest that this is outer core material. While a sulphide-rich mantle component could also increase the PGE abundances (assuming that the sulphide is exhausted during partial melting), the sulphur-undersaturated nature of these basalts argues against this. Variations in the amount of outer core in the source (from 0 to 1 wt%) and degree of fractional crystallization can account for the entire range in PGE abundances of OJP basalts.

Large‐ion lithophile element fractionation during the early differentiation of Mars and the composition of the martian primitive mantle

Abstract— Basaltic shergottites display a systematic decrease in K/Th, K/U, and K/La ratios with increasing K content. These trends are interpreted as mixing lines between relatively young martian magmas derived from highly depleted mantle sources and an ancient large‐ion lithophile (LIL) element‐enriched crustal component. One implication of this is that a substantial fractionation of these ratios occurs during the early crustal differentiation on Mars. Isotopic evidence from SNC meteorites and compositional data from Pathfinder and orbital gamma ray spectroscopy suggest that in excess of 50% of the LIL element complement of Mars resides in the crustal reservoir. If so, the primitive mantle of Mars is significantly more volatile‐depleted (i.e., lower K/Th, K/U, K/La) than previously thought but probably (though not necessarily) still less volatile‐depleted than the primitive mantle of the Earth. The La/Th ratios of virtually all SNC meteorites are subchondritic, including those with the most severe LREE‐depletion. Extrapolation of the basaltic shergottite trend suggests that both the depleted mantle end member and the enriched crustal end member have subchondritic La/Th ratios. This is in contrast with the Earth where basalts from LIL element‐depleted sources such as MORB have superchondritic La/Th ratios, complementary to the subchondritic ratios of the continental crust. Accordingly, assuming that the refractory elements are in chondritic proportions for the Mars primitive mantle, an additional major geochemical reservoir must exist on Mars that may not yet have been sampled.

Stability of various hydrous phases in CMAS pyrolite-H 2 O system up to 25 GPa

We carried out a series of melting experiments with hydrous primitive mantle compositions to determine the stability of dense hydrous phases under high pressures. Phase relations in the CaO–MgO–Al2O3–SiO2 pyrolite with ˜2 wt% of water have been determined in the pressure range of 10–25 GPa and in the temperature range between 800 and 1400 °C. We have found that phase E coexisting with olivine is stable at 10–12 GPa and below 1050 °C. Phase E coexisting with wadsleyite is stable at 14–16 GPa and below 900 °C. A superhydrous phase B is stable in pyrolite below 1100 °C at 18.5 GPa and below 1300 °C at 25 GPa. No hydrous phases other than wadsleyite are stable in pyrolite at 14–17 GPa and 900–1100 °C, suggesting a gap in the stability of dense hydrous magnesium silicates (DHMS). We detected an expansion in the stability field of wadsleyite to lower pressures (12 GPa and 1000 °C). The H2O content of wadsleyite was found to decrease not only with increasing temperature but also with increasing pressure. The DHMS phases could exist in a pyrolitic composition only under the conditions present in the subducting slabs descending into the lower mantle. Under the normal mantle and hot plume conditions, wadsleyite and ringwoodite are the major H2O-bearing phases. The top of the transition zone could be enriched in H2O in accordance with the observed increase in water solubility in wadsleyite with decreasing pressure. As a consequence of the thermal equilibration between the subducting slabs and the ambient mantle, the uppermost lower mantle could be an important zone of dehydration, providing fluid for the rising plumes.

Mantle Source Characteristics of Alkali Basalts and Basanites in an Extensional Intracontinental Plate Setting, Western Anatolia, Turkey: Implications for Multi-stage Melting

The alkaline volcanic province of Western Anatolia is characterized by intra-continental plate alkali olivine basalts and basanites extruded along localized extensional basins during Late Miocene (<11 Ma) to Quaternary (>0.13 Ma) time. The rocks are characterized by low 87Sr/86Sr (0.70311-0.70325) and high 143Nd/144Nd (0.51293-0.51298) ratios; they have OIB-like trace-element patterns characterized by enrichment in LILE, HFSE, and L-MREE, and a slight depletion in HREE relative to N-MORB. Trace-element modeling indicates that the mafic magmas formed by variable degrees (˜2% -10%) of partial melting of an isotopically homogenous garnet-bearing mantle source. The degree of melting decreased progressively from early-formed alkali olivine basalts to later basanites. Major- and trace-element systematics reveal two distinct depth ranges of melt segregation from the source mantle: (1) garnet and spinel + garnet stability zone for the Late Miocene rocks; and (2) spinel and spinel + garnet stability zone for the Quaternary rocks. Source concentrations, calculated using an inverse numerical method, show that the mantle from which the alkaline magmas were generated was enriched in all incompatible elements (e.g., LILE, HFSE, and LMREE) relative to depleted MORB mantle (DMM) and primitive mantle (PM) compositions. The isotopically depleted nature of the alkaline rocks relative to bulk silicate earth (BSE) further indicates that this enrichment is a recent event related to small-degree, multi-stage melting processes that involve local metasomatism of the mantle.

Experimentally determined partition coefficients for minor and trace elements in peridotite minerals and carbonatitic melt, and their relevance to natural carbonatites

A laser ablation microprobe coupled to an ICP-MS was used to analyse concentrations of K, Rb, Cs, Sr, Ba, Ti, Zr, Hf, Nb, La, Ce, Sm, Ho, Yb and Lu in experimentally produced garnet, clinopyroxene, amphibole and magnesiocarbonatitic melt. The experiment was conducted at 1050 °C and 2.5 GPa on a synthetic composition designed to simulate natural peridotite melting. Crystal/melt partition coefficients determined from the experiment were used to calculate minor and trace element concentrations in a hypothetical carbonatitic melt equilibrated with a garnet-bearing amphibole lherzolite of primitive mantle composition. The melt is strongly enriched in alkaline earths and rare earths relative to high-field-strength elements and alkalis, and plots within the compositional range of natural carbonatites. However, the fractionation of alkaline earths and REE from alkalis and HFSE is usually larger in natural carbonatites than in the calculated composition. The absolute concentrations of many incompatible elements are also much larger in some natural carbonatites. Because of these features, processes in addition to peridotite melting (such as crystal fractionation, wall rock reactions, and variations in source chemistry) are needed to fully account for the minor and trace element characteristics of carbonatites.

Melting relations of the hydrous primitive mantle in the CMAS–H 2O system at high pressures and temperatures, and implications for generation of komatiites

The hydrous phase relations of primitive mantle compositions in the CaO–MgO–Al 2O 3–SiO 2–H 2O system with 1, 2, and 5 wt.% H 2O have been investigated at 4, 6.5, and 8 GPa. The stability field of orthopyroxene in the residue expands with increasing H 2O content at 4–8 GPa. Dissolution temperature of garnet decreases more rapidly than that of the other phases at 4 and 6.5 GPa. The dissolution curves of garnet and pyroxenes cross at 6 GPa and 1700°C in the 2 wt.% H 2O-bearing system. Garnet is stable until high degree of melting at 8 GPa. Compositions of aluminum depleted komatiites are consistent with the liquid formed by more than 30% melting of the primitive mantle with H 2O up to 5 wt.% at 8 GPa. Aluminum undepleted komatiite can be formed by more than 50 wt.% melting of dry and wet primitive peridotite (with water up to 2 wt.% H 2O) at around 4–6 GPa. Variation in chemistry of komatiites might be explained by partial melting of primitive mantle with various water contents.

Trace-Element Modeling and Source Constraints for Tholeiitic and Cale-alkaline Basalts from a Depleted Asthenospheric Mantle Source, Mt. Erciyes Stratovolcano, Turkey

The Mt. Erciyes stratovolcano was built up in an intraplate tectonic environment as a consequence of Eurasian and Afro-Arabian continental collision. However, the volcanic products generally exhibit a calc-alkaline character; minor amounts of tholeiitic basalts are also present. Tholeiitic basalts show high Fe2O3, MgO, CaO, low K2O, and depleted Ba, Nb, and especially Rb (2.3-5.97 ppm) contents, low 87Sr/86Sr (0.703344-0.703964), and high 143Nd/144Nd (0.512920-0.512780) isotopic ratios. These compositional features show that they were derived from a depleted asthenospheric mantle source, possibly a MORB-like source component. In contrast, calc-alkaline basaltic rocks exhibit relatively high large-ion-lithophile and high-field-strength elements, high 87Sr/86Sr (0.704591-0.70507) and low 143Nd/144Nd (0.51272-0.512394) isotopic ratios. The bulk-rock chemistry of the tholeiitic basalts reflects the chemical composition of the extracted source component. Furthermore, trace-element concentrations may be calculated from an accepted mantle source component (starting composition) for different degrees of partial melting. These calculations also provide a sensitive approach to the origin of tholeiitic basalts. Modeled trace-element compositions of tholeiitic basalts are calculated from a primitive mantle composition. Calculated trace-element compositions imply that tholeiitic basalts are derived by minor fractional melting (1-1.5 %), in the absence of assimilation or deep-crustal melting. The calc-alkaline basalts were subsequently produced from initially tholeiitic basalts by the way of an AFC (assimilation-fractional crystallization) process, with a crustal assimilation of 10-15 %. The geochemical data, partial melting, and AFC modeling all indicate that basaltic products have a complex evolutionary history involving partial melting from a MORB-like mantle source. The assimilation and fractional crystallization processes are considered as providing an example for the chemical evolution of basaltic products, from tholeiitic to calc-alkaline, in an intraplate environment.

The density structure of subcontinental lithosphere through time

This study uses information on composition, thermal state and petrological thickness to calculate the densities of different types of subcontinental lithospheric mantle (SCLM). Data from mantle-derived peridotite xenoliths and garnet–xenocryst suites document a secular evolution in the composition of SCLM: the mean composition of newly formed SCLM has become progressively less depleted, in terms of Al, Ca, mg# and Fe/Al, from Archean, through Proterozoic to Phanerozoic time. Thermobarometric analyses of xenolith and xenocryst suites worldwide show that the mean lithospheric palaeogeotherms rise from low values (corresponding to surface heat flows of 35–40 mW/m2) beneath Archean terranes, to higher values (>50 mW/m2) beneath regions with Phanerozoic crust. The typical thickness of the lithosphere (defined as a chemical boundary layer), ranges from about 250 to 180 km, 180–150 km and 140–60 km for Archean, Proterozoic and Phanerozoic terranes respectively. The depth of this lithosphere–asthenosphere boundary corresponds to a temperature of 1250–1300°C. Using the estimated compositions, average mineral compositions and experimental data on the densities of mineral end-members (tables 1 and 2), we calculate mean densities at 20°C for Primitive Mantle (3.39 Mg m−3) and for SCLM of Archean (3.31±.016 Mg m−3), Proterozoic (3.35±0.02 Mg m−3) and Phanerozoic (3.36±0.02 Mg m−3) age. Curves of density and cumulative density versus depth, which take into account variations in geotherm with tectonothermal age, have been constructed for each age type of lithospheric section to assess the buoyancy of these columns relative to the asthenosphere, modelled as a Primitive Mantle composition. The density curves show that Archean SCLM is significantly buoyant relative to the asthenosphere at depths greater than about 60 km. Proterozoic sections deeper than about 100 km thick also are significantly buoyant. The buoyancy of Archean and Proterozoic SCLM sections, combined with their refractory composition, leads to high viscosities and explains the longevity and stability of old SCLM. Replacement of Archean lithosphere, as beneath the present-day eastern Sino–Korean craton, probably involves mechanical dispersal by rifting, accompanied by the rise of hot, fertile asthenospheric material. Fertile Phanerozoic lithosphere is buoyant when the geotherm is sufficiently high, as in many Cenozoic volcanic provinces. However, as the geothermal gradient relaxes toward a stable conductive profile, Phanerozoic SCLM sections thinner than about 100 km become denser than the asthenosphere, and hence gravitationally unstable. This could help to induce delamination of the SCLM and upwelling of asthenospheric material, beginning a new cycle. The tectonic consequences of such lithosphere replacement would include uplift and magmatism, and basin formation during subsequent thermal relaxation.

Preservation of near-solar neon isotopic ratios in Icelandic basalts

Neon isotopic ratios measured in olivine and basaltic glass from Iceland are the most primitive observed so far in terrestrial mantle-derived samples. Ratios were measured in gas released from olivine and basaltic glass from a total of 10 samples from the Reykjanes Peninsula, Iceland, and one sample from central Iceland. The neon isotopic ratios include solar-like, mid-ocean ridge basalt (MORB)-like and atmospheric compositions. Neon isotopic ratios near the air–solar mixing line were obtained from the total gas released from glass separates from five samples. MORB-like neon isotopic compositions were measured in the total gas released from olivine and glass separates from four samples. Although there is clear evidence for a solar neon component in some of the Icelandic samples, there is no corresponding evidence for a solar helium ratio (320 R a> 3He/ 4He>100 R a). Instead, 3He/ 4He ratios are mainly between 12±2( R a) and 29±3( R a), similar to the range observed in ocean island basalts, indicating that the He–Ne isotopic systematics are decoupled. The mantle source of Icelandic basalts is interpreted to be highly heterogeneous on a local scale to explain the range in observed helium and neon isotopic ratios. The identification of solar-like neon isotopic ratios in some Icelandic samples implies that solar neon trapped within the Earth has remained virtually unchanged over the past ∼4.5 Ga. Such preservation requires a source with a high [Ne solar]/[U+Th] ratio so that the concentration of solar neon overwhelms the nucleogenic 21Ne* produced from the decay of U and Th in the mantle over time. High [Ne solar]/[U+Th] ratios are unlikely to be preserved in the mantle if it has experienced substantial melting. An essentially undegassed primitive mantle component is postulated to be the host of the solar neon in the Icelandic plume source. Relatively small amounts of this primitive mantle component are likely to mix with more depleted and degassed mantle such that the primitive mantle composition is not evident in other isotopic systems (e.g. strontium and neodymium). The lower mantle plume source is inferred to be relatively heterogeneous owing to being more viscous and less well stirred than the upper mantle. This discovery of near-solar neon isotopic ratios suggests that relatively primitive mantle may be preserved in the Icelandic plume source.

Partitioning of major elements between garnet-structured minerals and silicate melt at pressure of 3–15 GPa

High-pressure melting experiments on primitive mantle composition material (PM1) and thoreiitic basalt (JB2) were performed from 3 to 15 GPa, and garnet (majorite)/silicate melt partition coefficients, D, were measured for major elements. In the case of PM1, garnet was the liquidus phase only at around 15 GPa, where the partitioning experiments were performed. Although experimental time of PM1 was changed from 10 to 120 min, no obvious change was found in the observed D values of major elements. The liquidus phase of JB2 was garnet under the present experimental conditions. With increase in pressure, an obvious increase in D Na and decrease in D Ti were found in JB2, but D values of the other elements showed no significant pressure dependence. The observed D values of PM1 and JB2 were compared in partition coefficient (PC)-ionic radius (IR) diagram. Although the absolute D values were different, D profiles of PM1 and JB2 at 15 GPa resembled each other.

Whole rock compositional variations in an upper mantle peridotite (Horoman, Hokkaido, Japan): are they consistent with a partial melting process?

Whole rock major and trace element abundances of the Horoman peridotites were used to understand processes forming lithological and compositional variations in the upper mantle. Similar to other orogenic peridotites, Horoman peridotites range from fertile lherzolites (3 to 4% Al 2O 3 and CaO) to depleted harzburgites (∼0.5% Al 2O 3 and CaO). Abundances of major oxides and compatible to moderately incompatible elements vary systematically with variations in MgO content. Such trends are commonly interpreted as indicating that the peridotites formed as residues from varying degrees of partial melting. The fertile end of these trends coincides with estimates of primitive mantle composition. Because of a mismatch between experimental melting trends for spinel peridotite, especially the Na 2O-MgO trend, the compositional variations of Horoman peridotites are not consistent with formation as residues from partial melting of spinel peridotite. Non-linear trends in minor and trace element versus major element abundance diagrams also preclude a two-component mixing model. Recent melting experiments on garnet peridotite demonstrate that at 3 GPa the near-solidus peridotite has a large amount of subcalcic clinopyroxene (ca. 27%) coexisting with small amount of garnet (ca. 2%). Residues from polybaric melting of such garnet peridotite are consistent with the abundance variations of major and moderately incompatible elements, such as Na and heavy rare-earth elements, in the Horoman peridotites. A similar conclusion is applicable to other orogenic peridotites such as the Ronda peridotite because their major element compositional variations are similar to the Horoman peridotite.

Geochemistry and palaeotectonic setting of amphibolites from the Western Tatra Mountains, southern Poland

Amphibolites from the crystalline basement of the Western Tatra Mountains, which are found as small lenses within migmatitic gneisses and mica schists, were formed during pre- or early Variscan amphibolite-facies metamorphic events, and subsequently intruded by the post-metamorphic Variscan Tatra Granite. The amphibolites occur in both the upper and lower metamorphic complexes, which are separated by a major subhorizontal shear zone in the Western Tatra Mountains. The amphibolites can be divided into three types: massive, striped and garnetiferous. The striped and massive amphibolites, concordant with a dominant S1 foliation, and the garnet amphibolites, which cross-cut the S1 banding in the gneisses, were all originally intrusive dolerites. The striped amphibolites (consisting primarily of hornblende, andesine and quartz), and later, cross-cutting garnet-hornblende-andesine-quartz-bearing amphibolites, predominate in the lower part of the dominantly migmatitic upper complex, and are exposed mainly on the ridges. The massive amphibolites, which contain a similar mineral assemblage, mainly occur in the usual unmigmatized lower structural unit. Chemical studies indicate that three amphibolite suites are present, which probably originated as a series of enriched tholeiites, similar to more recent plume-influenced magmas, which were derived by partial melting of a spinel lherzolite with primitive mantle composition and compositionally slightly modified by crustal contamination. The amphibolites were intruded as dolerites into clastic sediments which had accumulated in an extensional basin floored by attenuated continental crust, a situation similar to that of amphibolites found in metamorphic complexes within the Variscan belt, e.g. in the Orlica–Snieznik area of the Sudetes, where amphibolites chemically similar to those in the Western Tatra also occur. Copyright © 2000 John Wiley & Sons, Ltd.

Origin of planetary cores: evidence from highly siderophile elements in martian meteorites

We present new bulk compositional data for 6 martian meteorites, including highly siderophile elements Ni, Re, Os, Ir and Au. These and literature data are utilized for comparison versus the siderophile systematics of igneous rocks from Earth, the Moon, and the HED asteroid. The siderophile composition of ALH84001 is clearly anomalous. Whether this reflects a more reducing environment on primordial Mars when this ancient rock first crystallized, or secondary alteration, is unclear. QUE94201 shows remarkable similarity with EET79001-B for siderophile as well as lithophile elements; both are extraordinarily depleted in the “noblest” siderophiles (Os and Ir), to roughly 0.00001 × CI chondrites. As in terrestrial igneous rocks, among martian rocks Ni, Os and Ir show strong correlations vs. MgO. In the case of MgO vs. Ni, the martian trend is displaced toward lower Ni by a large factor (5), but the Os and Ir trends are not significantly displaced from their terrestrial counterparts. For Mars, Re shows a rough correlation with MgO, indicating compatible behavior, in contrast to its mildly incompatible behavior on Earth. Among martian MgO-rich rocks, Au shows a weak anticorrelation vs. MgO, resembling the terrestrial distribution except for a displacement toward 2–3 times lower Au. The same elements (Ni, Re, Os, Ir and Au) show similar correlations with Cr substituted for MgO. Data for lunar and HED rocks generally show less clear-cut trends (relatively few MgO-rich samples are available). These trends are exploited to infer the compositions of the primitive Earth, Mars, Moon and HED mantles, by assuming that the trend intercepts the bulk MgO or Cr content of the primitive mantle at the approximate primitive mantle concentration of the siderophile element. Results for Earth show good agreement with earlier estimates. For Mars, the implied primitive mantle composition is remarkably similar to the Earth’s, except for 5 times lower Ni. The best constrained of the extremely siderophile elements, Os and Ir, are present in the martian mantle at 0.005 times CI, in comparison to 0.007 times CI in Earth’s mantle. This similarity constitutes a key constraint on the style of core-mantle differentiation in both Mars and Earth. Successful models should predict similarly high concentrations of noble siderophile elements in both the martian and terrestrial mantles (“high” compared to the lunar and HED mantles, and to models of simple partitioning at typical low-pressure magmatic temperatures), but only predict high Ni for the Earth’s mantle. Models that engender the noble siderophile excess in Earth’s mantle through a uniquely terrestrial process, such as a Moon-forming giant impact, have difficulty explaining the similarity of outcome (except for Ni) on Mars. The high Ni content of the terrestrial mantle is probably an effect traceable to Earth’s size. For the more highly siderophile elements like Os and Ir, the simplest model consistent with available constraints is the veneer hypothesis. Core-mantle differentiation was notably inefficient on the largest terrestrial planets, because during the final ∼ 1% of accretion these bodies acquired sufficient H 2O to oxidize most of the later-accreting Fe-metal, thus eliminating the carrier phase for segregation of siderophile elements into the core.

Re-Os systematics of mantle xenoliths from the East African Rift: age, structure, and history of the Tanzanian craton

In order to understand the effects of contractional and extensional tectonics on thick, mantle roots, we have undertaken a systematic study of mantle xenoliths from the Labait volcano, which lies within the East African Rift on the eastern boundary of the Archean Tanzanian craton. The Re-Os systematics of the Labait xenoliths show that ancient, refractory lithosphere is present to depths of ∼140 km. Above this depth, the mantle section consists of harzburgitic xenoliths with whole rock 187Os/ 188Os between 0.1081 and 0.1140, corresponding to Re depletion (T RD) ages of 2.8 to 2.0 Ga. Chromites from these samples are generally less radiogenic than their corresponding whole rocks and have T RD ages between 2.5 to 2.9 Ga, yielding the best estimate for the age of this portion of the lithosphere. Coupled petrographic and isotopic data for some of these samples indicate they have been variably overprinted by recent addition of Re and/or radiogenic Os. Between 4.4 to 4.7 GPa (∼140 to 150 km depth), peridotites are more fertile and yield younger T RD ages (1.0 Ga to future ages). The highest temperature sample has radiogenic 187Os/ 188Os (0.133), overlapping the range measured for metasomatic xenoliths and the Labait host melilitite (0.13 to 0.14). This range is taken to represent asthenospheric mantle beneath the Tanzanian craton, which has plume-like isotopic characteristics. The suite shows a good correlation on 187Os/ 188Os vs. temperature (hence depth) and 187Os/ 188Os vs. wt.% Al 2O 3 or CaO plots. These trends, which pass above primitive mantle compositions, may reflect mixing of recent plume-derived Os with ancient lithospheric Os, formation of the lowest portion of the lithosphere during successive melting events or a combination of both processes. Our data show that complete delamination of the lithospheric mantle has not occurred beneath the Tanzanian craton during its long tectonic history. However, if the Archean lithosphere was originally thicker than the ∼140 km currently beneath Labait, then the lithosphere has been thinned, either by thermal erosion associated with the rift or by partial delamination during Proterozoic collision. Finally, we see no evidence for extensive lithospheric thinning associated with development of the East African Rift, although overprinting of the lithosphere by rift-related magmas has occurred.

THE COMPOSITION OF MESOZOIC PERIDOTITE TECTONITES OF THE EASTERN ALPS

Numerous small ultramafic bodies are exposed in Mesozoic cover units of the central Eastern Alps. The major occurrences are restricted to the tectonic windows of the Penninic zone and their surroundings. In the Lower Engadin window, three different mantle peridotite groups have been investigated. The Idalp ophiolite, situated at the northern rim of the Engadin window, is believed to be of south Penninic origin (Trumpy, 1972), whereas the Ramosch ophiolite at the southwestern margin is assigned to the north Penninic area (Vuichard, 1984). The tectonic position of the ultramafics in the vicinity of Nauders, at the southern margin, is unknown. Ophiolite remnants of south Penninic affinity are also present in the Glockner nappe of the Tauern Window. At the eastern end of the Alpine orogen similar peridotites appear in the small Rechnitz window. Numerous, usually small and highly serpentinized bodies of ultramafic rocks can be traced in the Matrei zone along the southern margin of the Tauern window; the Matrei Zone comprises both, Penninic and Lower Austroalpine elements. The Reckner ophiolite complex, near the NW corner of the Tauern window, is part of the tectonically higher units of the Lower Austroalpine nappe. Geochemical and petrological investigations reveal considerable differences between these mantle slices. The different ultramafic bodies are influenced by a pervasive regional metamorphism locally reaching amphibolite facies. Primary mineral assemblages of the peridotites have been mostly replaced by metamorphic parageneses. The products of metamorphism are dominantly serpentine minerals accompanied by various combinations of diopside, tremolite, chlorite, hydrogarnet and magnetite. In contrast to all other occurences, the samples from Nauders are characterized by a well preserved primary assemblage. Relic clinopyroxene and spinel can be found in peridotites of the Reckner and in the Matrei zone. Olivine and orthopyroxene pseudomorphed by serpentine minerals are present in less deformed areas only. Preserved clinopyroxenes of these mantle peridotites have up to 7 wt% Al2O3 and 1.5-2.0 wt% Na2O, which corresponds to a jadeite component of more than 10 %. The Cr2O3 content is generally high (up to 1 wt%). Clinopyroxene is frequently zoned with progressive depletion of Al, Na, Ti and Cr towards pure diopside. In samples from Nauders the Fo-content of olivine is about 90% and the NiO concentration varies between 0.4-0.5 wt%. The orthopyroxenes range in XMg from 91-92 and Al2O3 reaches up to 5 wt%. Titanian pargasite (XMg 0.88, about 3.5 wt% TiO2) is a secondary phase in Nauders peridotites only. Spinels from Nauders are different in composition to those of the Reckner complex. They are typically poor in Cr2O3 (about 7 wt%), high in Al2O3 (about 60 wt%) and develop rims which usually have lower Cr contents. Relics of spinel with significantly higher Cr values (Cr# 40-50, Cr2O3 35-40 wt%) are preserved in samples of the Reckner complex. The dominant rock types of all Mesozoic ultramafic bodies are residual lherzolites and harzburgites (XMg 87-94). The serpentinized ultramafics of the southern Penninic realm (Idalpe, Tauern and Rechnitz windows) are generally harzburgites with low Al2O3 (<2 wt%) and CaO contents (Fig. 1). The HREE concentrations (0.3-0.5 times chondrite) display patterns of moderately depleted restitic mantle (Fig. 2). Lherzolites and cumulate rocks (XMg = 0.80) are restricted to a few occurrences within the main mass of harzburgites. In contrast, more fertile lherzolites and harzburgites form ultramafic bodies of the Matrei zone, the Reckner complex, the Ramosch ophiolite and Nauders.The high abundances of Al2O3 (up to 4.6 wt%) and TiO2 (up to 0.22 wt%) are consistent with the estimated composition of a primitive upper mantle. All samples display higher HREE concentrations (up to 2.5 times chondrite) and a significant LREE depletion trend (Figs. 1 and 2). The Cr-Yb projection (Pearce and Parkinson, 1993) has been used to investigate the degree of mantle depletion. This diagram provides further evidence that the mantle material of the south Penninic region represents a more depleted source (about 15% melt depletion) than those of the Matrei zone and the Lower Austroalpine region (up to 5% melt depletion). These data support the model of a slow spreading oceanic environment (Hock and Koller, 1992) for the ophiolites of the Idalpe, Tauern and Rechnitz window, whereas the ultramafics of the Reckner, Matrei Zone and Nauders were probably generated in a pre-oceanic stage.

The Deep Cratonic Mantle Roots Beneath the Canadian Craton: Mantle Xenolith Evidence from Somerset Island Kimberlites

The nature and origin of the mantle roots beneath the Canadian craton are currently poorly constrained. The recently discovered Cretaceous Nikos kimberlite (Pell, 1993) on Somerset Island, in the Canadian Arctic, consists of three individual pipes and hosts an unusually well preserved suite of large (10-20 cm) mantle xenoliths dominated by garnet-peridotite (lherzolite, harzburgite, dunite) showing coarse and porphyroclastic textures, and minor occurrences of garnet-pyroxenite. The bulk compositions and mineral data for fifty-four Nikos xenoliths indicate their highly refractory nature with high olivine abundances (avg. 80 wt.%) and olivine forsterite contents (avg. Fo = 92.3). These compositions are characteristic of peridotite xenolith suites derived from the mantle underlying Archaean cratons such as the Kaapvaal and Siberian cratons (avg. olivine Fo = 92.5; Boyd et al., 1997), but are clearly distinct from the trend defined by oceanic peridotites (avg. olivine Fo = 91.0; Boyd, 1989) and spinel-peridotites from the Canadian Cordillera (Shi et al., in press) that represent the mantle beneath accreted post-Archaean continental terranes along the western margin of the Canadian shield. The Nikos xenoliths yield pressures and temperalures of last equilibration between 20 to 55 kb and 650 to 1300~ and a number of the peridotite nodules appear to have equilibrated in the diamond stability field. The pressure and temperature data define a conductive palaeogeotherm corresponding to a surface heat flow of 44 mW/m 2. Peridotite xenoliths from the central Slave province of the Canadian craton require a lower surface heat flow (-40 mW/m2; Boyd and Canil, 1997) indicating a cooler geothermal regime than that beneath the Canadian Arctic. Many of the kimberlite-hosted peridotites from the Kaapvaal craton in South Africa and parts of the Siberian craton are characterized by high orthopyroxene contents (avg. Kaapvaal 32 wt.%, Siberia 20 wt.%; Boyd et al., 1997). The Nikos peridotites show modal mineral assemblages with moderate to low orthopyroxene contents (avg. 12 wt.%), indicating that the orthopyroxene-rich mineralogy characteristic of the Kaapvaal and Siberian cratons are not present in the subcratonic mantle beneath Somerset Island, and suggesting that the Nikos xenoliths formed by extensive partial melting of primitive mantle compositions. The isotope results obtained to date suggest that the cratonic mantle beneath Somerset Island is heterogeneous and that the pyroxenites, lherzolites, and harzburgites may represent isotopically distinct mantle reservoirs. The Pb isotope data for the Nikos peridotites show that the harzburgites have more radiogenic Pb isotopic compositions than the lherzolites. One pyroxenite sample, in particular, gives a very unradiogenic Pb isotope composition (2~176 = 17.015, 2~176 = 15.349, 2~176 = 37.285) indicative of involvement of an old, most probably Archaean mantle component. Nd and Sr isotope results for the Nikos xenoliths (143Nd/144Nd = 0.51257-0.51285; S7Sr/S6Sr = 0.70412-0.70969) are similar to data previously reported for gamet-peridotites world wide, with the lherzolites plotting at the enriched end of the data array. Nd model age calculations for the Nikos xenoliths indicate ages between 420 and 800 Ma. A Lu-Hf isochron using garnet and clinopyroxene separates from one pyroxenite sample yields an age of 680 Ma, similar to the Nd model ages. The Hf and Nd isotope results may reflect formation ages, but more likely represent mixing between an old Archaean mantle reservoir and a younger metasomatic event possibly associated with kimberlite generation.

Depleted spinel harzburgite xenoliths in Tertiary dykes from East Greenland: Restites from high degree melting

A new collection of mantle xenoliths in Tertiary dykes from the Wiedemann Fjord area in Southeast Greenland shows that this part of the central Greenland craton is underlain by highly depleted peridotites. The samples are mostly spinel harzburgites with highly forsteritic olivines (Fo 87–94, average Fo 92.7). This, together with unusually high modal olivine contents (70–>95%), places the Wiedemann harzburgites in a unique compositional field. Relative to depleted Kaapvaal harzburgites with comparable Fo in olivine, the Wiedemann samples have considerably lower bulk SiO 2 (average 42.6 wt% versus 44–49 wt%). Spinel compositions are similar to those in other sub-cratonic harzburgites. Pyroxene equilibrium temperatures average 850°C, which is above an Archaean cratonic geotherm at an inferred pressure of 1–2 GPa, but low enough so that it is unlikely that the xenoliths represent residual peridotites created during Tertiary magmatism. Among mantle samples, the Wiedemann harzburgites are, in terms of their bulk composition, most similar to harzburgites from the ophiolites of Papua New Guinea (PNG) and New Caledonia (NC). One hypothesis is that the Wiedemann harzburgites, along with PNG and NC harzburgites, formed via dissolution of pyroxene from previously depleted peridotites, possibly beneath a volcanic arc. If so, higher spinel Cr/Al in Wiedemann samples may reflect a deeper origin compared to PNG and NC peridotites. Alternatively, using proposed primitive mantle compositions as a protolith, the Wiedemann harzburgites can be modeled as the residue after extraction of some 40% melt. The composition of this calculated hypothetical melt in terms of CaO, Al 2O 3, FeO, MgO and SiO 2 is similar to published experimental data on high degree melts of peridotite at 2–3 GPa. Munro-type komatiites lie close to these calculated and experimental melts but are slightly displaced towards low degree experimental melts at higher pressure (e.g., 6 GPa). We conclude that the Wiedemann harzburgites formed as a residue after about 40% melting, and that they may represent shallow, refractory residues after polybaric melting initiated at pressures ≥7 GPa and continuing to relatively low pressures (2–3 GPa or less). Extraction and aggregation of polybaric melts would produce liquids similar to Munro-type komatiites.

Potassic and Sodic Igneous Rocks from Eastern Paraguay: their Origin from the Lithospheric Mantle and Genetic Relationships with the Associated Parana Flood Tholeiites

Eastern Paraguay represents the westernmost fringe of Early Cretaceous Parana flood tholeiites (Serra Geral Formation SGF). Besides the SGF, eastern Paraguay has been the site of alkaline magmatism since Mesozoic times: (1) Late Permian-Early Triassic sodic intrusions and lavas: (2) Early Cretaceous potassic igneous rocks and very scarce sodic lavas (3) Late Cretaceous–Oligocene sodic lavas. Two distinct magmatic events are dominant in the Asuncion–Sapucai graben (ASU) eastern Paraguay: (1) wide-spread potassic magmatism and SGF tholeiites (Early Cretaceous) (2) Asuncion sodic magmatism (Late Cretaceous–Oligocene). The potassic rocks form a compositional continuum from moderately to strongly potassic. Two potassic suites are proposed, i.e. basanite to phonolite and alkali basalt to trachyte and their intrusive analogues. The sodic rocks include ankaratites, nephelinites and phonolites. Two similar but distinct parental magmas have been inferred for the potassic suites, both characterized by strongly fractionated REE and negative ‘Ta–Nb–Ti anomalies’. Slight positive Ta and Nb anomalies distinguish the sodic rocks. Sr–Nd isotope data confirm the distinction of the potassic rocks, enriched in radiogenic Sr and low in radiogenic Nd, from the sodic rocks, close to bulk Earth. Crustal contamination does not appear to have been significant in the generation of the investigated rocks, supported by δ 18O data. The source of potassic rocks is constrained by high LILE, LREE, Th, U and K, relative to a primitive mantle composition, and a garnet peridotite is favoured as a possible mantle source for the investigated rocks. The close association of potassic and sodic rock suites in the ASU demands that their parental magmas derived from a heterogeneous subcontinental mantle, variously enriched in incompatible elements. Significant H2O, CO2 and F are expected in the mantle source(s) for the occurrence of coeval carbonatites. Any genetic hypothesis based on a ‘mantle plume’ system is constrained by strong lithospheric mantle characteristics. This does not preclude that thermal perturbations from the asthenosphere may have trggered magmatic activity in the lithospheric mantle. Model ages indicate that two distinct metasomatic events may have occurred during Late and Early–Middle Proterozoic as precursor to the genesis of tholeiitic and alkaline magmatism in the Parana basin. These metasomatic processes were chemically distinct, indicated by the strong differences in Ti, LILE and HFSE concentrations found in both alkaline provinces (e.g. potassic rocks from ASU vs Alto Paranaiba Igneous Province) and Parana tholeiites (low vs high Ti). Ingeneral, the relationships between the alkaline rocksfrom southeastern Brazil, i.e. AltoParanaiba, Ponta Grossa Arch, Serra do Mar, Lages and the flood basalts of the Parana Basin, support a common origin in the lithospheric mantle. Sr–Nd isotope and other geochemical data indicate that a significant role was played by a mantle component depleted in incompatible elements and with high Sm/Nd ratio. This component (N-MORB type) would be represented by the depleted portions of a mantle which was variously metasomatized during Proterozoic times. The isotopic and geochemical features of the modernTristan da Cunha plume are distinctly different from the component depleted in incompatible elements, and its contribution is not apparent in the compositions of the Parana tholeiites.

北海道日高帯，幌満マントルダイアピル内での全岩主化学組成と微量成分組成の挙動

Petrochemical studies based mainly on major and trace element whole-rock XRF analyses have been made on the Horoman and the Nikanbetsu peridotite complexes, Hokkaido, Japan. Incompatible trace element abundances in peridotite samples were obtained by carefully examining the XRF lower limit of detection. The Horoman complex is divided into Lower and Upper Zones on the basis of geological and petrological features. There is a systematic variation between Mg# [100×Mg/(Mg+Fe)] and whole-rock chemical compositions in both zones. There are, however, several differences between the plagioclase lherzolite samples from the Upper Zone and those of the Lower Zone. First, the range of Mg# of the plagioclase lherzolite from the Upper Zone is more variable than that of the Lower Zone. Some samples from the Upper Zone have higher Mg# than the maximum values for plagioclase lherzolite from the Lower Zone. Such samples are significantly poor in Al2O3, CaO, TiO2, Sr and Y. Second, there is a gap in TiO2 content in the compositional trend for the plagioclase lherzolite from the Upper Zone. This observation enables the division of the plagioclase lherzolite into TiO2-rich and-poor types. These results indicate that the compositional variation of the plagioclase lherzolite from the Upper Zone was caused by pervasive partial melting and variability in the extent of segregation of the partial melt. By contrast, the plagioclase lherzolite from the Lower Zone is inferred to have kept its original composition, which is consistent with petrographic evidence for no or very minor partial melting. Spinel lherzolite and harzburgite from the Horoman complex are rich in Sr and Ti relative to Zr and Y. Plagioclase lherzolite from the Upper Zone occurring close to the segregation vein has a character more enriched in melt component and was probably derived by addition of a melt component to the primitive mantle composition. The Nikanbetsu complex may have the same origin as the Horoman complex because of their similarity in major and trace compositions.

The chemical composition of the Earth

The bulk composition of the Earth and the composition of the mantle and core are calculated using the ratios of major and trace elements. The ratios of elements which do not enter the core (lithophile) are the same in the bulk Earth as in the mantle. Bulk earth ratios involving an element that does enter the core (siderophile) are therefore determined from meteorite correlation diagrams of siderophile-lithophile ratios vs. lithophile-lithophile ratios and from primitive mantle composition in elements which do not enter the core (e.g., Al). The composition of the core is determined by difference, without resorting to assumptions about core formation processes. It is found that the core contains about 7.3 wt% silicon and 2.3 wt% sulphur. To account for the seismologically determined density deficit of the core, about 4 wt% oxygen must be added. The present results are compatible with the idea that the core material equilibrated at low pressure, in reducing conditions. Furthermore, we propose that the Earth is closer to CM rather than to C1 for non-volatile element ratios.