Recycled Crustal Material Research Articles

Melt depletion and subsequent metasomatism in the shallow mantle beneath Koolau volcano, Oahu (Hawaii)

Xenoliths from Pali (Oahu, Hawaii) include samples of the mantle lithosphere underlying Koolau shield volcano. Most such xenoliths are spinel peridotites, the remainder being plagioclase-spinel peridotites, and garnet-free pyroxenites. Clinopyroxene separates from Pali peridotite xenoliths have relatively depleted 87Sr/86Sr (0.70309–0.70346) and 143Nd/144Nd (0.512967–0.513206). The mantle lithosphere beneath the Koolau volcano has a range of present-day epsilonNd = 6.9–11. On a 143Nd/144Nd versus 147Sm/144Nd diagram, they define a 61 ± 20 Ma errorchron, within error of the estimated age of ~80–85 Ma of the Pacific lithosphere at Oahu, and an intercept of epsilonNd = 6.8. We interpret the Pali spinel peridotites as samples of the Pacific lithosphere residual to melt extraction of the Pacific crust. These rocks were not metasomatized by melts having isotopic composition similar to the shield-building Koolau volcano, but they could have been metasomatized by melts related to the Honolulu Volcanic Series. Plagioclase mineral separates from two peridotites are in 87Sr/86Sr isotopic disequilibrium with coexisting clinopyroxenes. The plagioclases are light rare earth element depleted and have “enriched” 87Sr/86Sr ratios (0.70436 and 0.70443). The delta18O of olivine (5.09–5.12‰) and clinopyroxene (5.32–5.33‰) from spinel peridotites are typical for oceanic upper mantle rocks. In contrast, delta18O of olivine, pyroxene, and plagioclase in plagioclase-bearing peridotites are enriched by ~0.5‰ relative to the spinel peridotites (e.g., delta18O ratios of olivines from plagioclase peridotites are 5.4–5.5‰). It is here interpreted that the plagioclases represent phases that formed by reaction with or precipitation from a melt that intruded the Pacific lithospheric mantle. The Pacific lithospheric isotopic composition is reflected in the clinopyroxene separates. The high delta18O and 87Sr/86Sr of the plagioclase resemble Koolau lavas and suggest that the melt passing through the lithosphere was plume-related. The elevated delta18O of the plagioclase indicates that the melt from which it fractionated had a component of recycled crustal material not acquired from local crustal contamination. It is also estimated here that the duration of the porous melt flow of the Koolau-like melt though the mantle lithosphere was short-lived, ~10^2–10^4 years.

Hydrogen and oxygen isotope geochemistry of A-type granites in the continental margins of eastern China

Hydrogen and oxygen isotope studies of both whole-rock and mineral separates were carried out for five representative A-type granites in eastern China. From North to South, the whole-rock δD and δ 18O values vary from −145 to −82‰ and from −1.2 to 5.0‰ for Nianzishan pluton, from −135 to −110‰ and from 1.6 to 6.7‰ for Shanhaiguan pluton, and from −145 to −111‰ and from 1.5 to 9.1‰ for Laoshan massif, respectively. The highest whole-rock δD and δ 18O values are observed in Suzhou pluton and range from −81 to −59‰ and from 3.5 to 9.2‰, respectively. Homogeneous δD and δ 18O values are found in Kuiqi massif, ranging from −113 to −99‰ and from 5.6 to 8.5‰, respectively. The results show that the large variabilities in the hydrogen and oxygen isotope ratios are mainly related to the diverse processes of magma evolution, such as magma degassing, fractional crystallization as well as post-magmatic alteration by local meteoric water under subsolidus conditions. The initial δD and δ 18O values of the A-type granites in eastern China are constrained to be about −50±5‰ and 7.5±1.0‰, respectively. This is consistent with mantle-like protoliths for granitoid rocks and thus precludes the possibility that the A-type granites is derived from the anatexis of the 18O-rich upper crust. However, recycling of crustal materials by earlier subduction cannot be excluded in the processes of generating the A-type granitic magmas. The variation in the D-depletion of the granites with latitude implies that the paleogeographic localities of the A-type granites have not changed significantly since the Mesozoic.

Hydrogen isotope composition of mantle-derived mica megacryst from ion micro probe analysis

The hydrogen isotope composition of a mantle-derived mica megacryst from Cenozoic basanite from Nushan, Anhui Province has been determined by ion micro probe. The results demonstrate that δ D and water content of the megacryst were heterogeneous on the micro scale, which resulted from reaction with meteoric water after being brought to the surface. The primary δD of mica megacrysts was about - 23‰, suggesting the recycled crustal materials in its source. By combining these values with those of other researchers, it is believed that the hydrogen isotope composition of the mantle is heterogeneous at least on the large scale.

Re–Os isotope geochemistry of Tertiary picritic and basaltic magmatism of East Greenland: constraints on plume–lithosphere interactions and the genesis of the Platinova reef, Skaergaard intrusion

Re–Os abundance and isotopic studies on a small number of samples from the lowermost part of the East Greenland lava pile and the nearby Skaergaard intrusion show that picrite and ankaramite lavas tend to have high Os abundances and low Re/Os ratios, with most rocks having near-chondritic values for initial 187 Os/ 188 Os . One olivine basaltic sample with very high 187 Os/ 188 Os has most likely been affected by upper crustal contamination. There is no clear evidence for recycled crustal material in the ancestral Iceland plume source or for a significant degree of interaction of the picrites with ancient subcontinental lithospheric mantle of the North Atlantic craton. The Platinova Reef, the precious metal enriched zone in the Skaergaard intrusion, has very high Au and Pd contents but very low Pt and Os contents compared to platiniferous horizons in other layered intrusions. These characteristics are almost certainly due to the very late stage in the evolution of the Skaergaard magma chamber at which the Platinova Reef formed. The Platinova Reef yields a Re–Os T CHUR model age that is younger than the age of the intrusion, which was most likely produced by some form of post depositional disturbance to the Re–Os isotopic system. The high PGE contents of the East Greenland volcanic rocks indicate that they were formed from magmas that had the potential to form major accumulations of Ni–Cu–PGE sulphide deposits. Had similar magmas undergone significant interaction with continental crust elsewhere in East Greenland, they may have formed massive magmatic sulphide deposits. Thus, the potential for the discovery of large magmatic sulphide deposits of the Noril'sk-Talnakh-type in East Greenland must be considered excellent.

Sources of ocean island basalts: A review of the osmium isotope evidence

The Os isotope system provides insight into the origin of mantle source heterogeneity which is complementary to that provided by the incompatible element isotope systems of Sr, Nd and Pb. The Os isotope system is both an extremely sensitive tracer of crustal contamination and the only isotope system which can clearly distinguish between contamination of magmas in the crust vs. contamination in the lithospheric mantle. Os isotopes therefore provide important constraints regarding whether basalts are recording plume or lithospheric signatures. Basalts inferred to be recording mantle plume signatures indicate that the plume sources contain an enriched component which is more radiogenic in Os than primitive upper mantle. While this might be attributed to recycled crustal material, the expected correlations between Os and Pb isotopic signatures are largely absent. An alternative possibility is that mantle plumes contain a component of lower mantle which is radiogenic in Os. The Os isotopic compositions of plumes may be further enriched in some cases by the addition of recycled crustal components which also produces distinctive signatures in the incompatible element isotope systems. HIMU basalts, which have both extremely radiogenic Os and Pb isotopic signatures, can be produced by the addition of 15–25% recycled oceanic crust to a plume source already slightly enriched in Os due to a radiogenic lower mantle component. The origins of EMI and EMII mantle sources are currently less well constrained by Os isotopic signatures, but might be attributed to recycling of oceanic crust plus pelagic sediment, and metasomatized subcontinental lithospheric mantle, respectively.

Oxygen isotope composition of mantle-derived materials from Anhui-Jiangsu basalt province, East China

The oxygen isotopic composition of mantle-derived materials from Anhui-Jiangsu basalt province indicates the mantle heterogeneity in δ18O and the coherence between δ18O and Sr-Nd isotopic composition for the mantle beneath the province. The oxygen isotopic compositions of the basalts from the Luhe-Yizheng area, the clino- and orthopyroxenes in spinel lherzolite xenoliths from the province and mineral megacrystals from Nushan and Panshishan are as same as those of normal mantle in a first approximation. The basalts from the Jiashan-Lai’an area have high δ18O, and Nd-Sr isotopic compositions of the basalts are the same as those of EM-type enriched mantle. The residual of recycled crustal materials supply the relevant enriched component, and mantle mixture produces the enriched source from which the basalts were derived.

Stable Isotope Evidence for Earth's Raw Materials

Variations in abundances of stable, nonradiogenic isotopes in natural materials may be used to study provenance of those materials or processes involved in their formation. In dealing with Earths mantle, the variations of interest include those which occur within the mantle, those which occur between mantle and other terrestrial reservoirs, and those which occur between the Earth and other solar system bodies. The last of these may shed light on chemical compositions of inaccessible regions of the mantle.Two examples are discussed here: (a) oxygen isotopes, which serve as tracers of recycled crustal material and provide stringent tests of twoor multi-component models of terrestrial accretion, and (b) potassium isotopes, which severely constrain processes for depletion of moderately volatile elements in the whole Earth.

An enriched mantle source for potassic basanites: evidence from Karisimbi volcano, Virunga volcanic province, Rwanda

Lavas from Karisimbi, the largest volcano in the Virunga province in the Western Branch of the African rift on the Zaire-Rwandan border, constitute a suite of mafic potassic basanites and more evolved potassic derivatives. All of the lavas are potassic with K2O/Na2O≥1, and enriched in incompatible elements, with chondrite normalised (La/Yb)n>18 and Nb/Zr>0.25. The 87Sr/86Sr and 143Nd/144Nd ratios reflect these enriched compositions, varying from 0.7052 and 0.51258 respectively in the K-basanites to 0.7132 and 0.51226 in the most evolved K-trachyte, although at MgO abundances >4% there is no systematic variation of isotope ratios with fractionation. At >4% MgO, lava compositions were controlled by assimilation and fractional crystallization in a sub-volcanic magma chamber. Trace-element and isotope variations in the more mafic lavas appear to reflect mixing between a “primitive” K-basanite (PKB) magma and a Sr-rich end-member, similar to melilite nephelinites from the neighbouring volcano, Nyiragongo. Both endmembers are mantle-derived and isotopically distinct, with the PKB being characterised by 87Sr/86Sr up to 0.707 and 143Nd/144Nd as low as 0.51236. Alternatively, isotope variations may be the time-integrated response to trace-element fractionations in a variably enriched mantle source. The Pb isotope variations within Karisimbi are complex. In the more evolved lavas all three ratios increase coherently with fractionation, whereas in the mafic varieties 206Pb/204Pb remains roughly constant at ∼19.2 while 207Pb/204Pb and 208Pb/204Pb vary from 15.67 to 15.78 and 39.49 to 40.80 respectively, defining sub-vertical trends, consistent with PKB-nephelinite magma mixing. The Nd and Sr isotopes indicate trace-element fractionation in the PKB source at ∼1 Ga, similar to ages derived from the overlying crust and suggesting a lithospheric origin. Elevated 208Pb/204Pb and 208Pb*/206Pb* values of the PKB are also consistent with Th/U fractionation at a similar time. However, this 1Ga age contrasts with that derived from the elevated 207Pb/204Pb ratios which indicate U/Pb fractionation during the Archaean. Crustal contamination can be excluded as the major control of Pb isotope variation in the PKB because their high Ce/Pb ratios (∼27) are similar to those typical of oceanic basalts. Parent/daughter trace-element fractionation and the high Ti, Nb and Ta abundances of the PKB lavas are all consistent with enrichment of a lithospheric source region by small-degree silicate melts at ∼1Ga. Comparison between measured and time-integrated trace-element ratios suggests that the degree of melting associated with recent magmatism was ≥5%. These data show that significant Th/U and Rb/Sr fractionation can be produced by intra-mantle melting processes and that high 208Pb/204Pb and 208Pb*/206Pb* values can evolve within the upper mantle and do not necessarily require the recycling of crustal material. Comparable isotope features in continental flood basalts and DUPAL ocean island basalts may be explained in a similar way.

Rhenium-Osmium and Samarium-Neodymium Isotopic Systematics of the Stillwater Complex

Isotopic data for the Stillwater Complex, Montana, which formed about 2700 Ma (million years ago), were obtained to evaluate the role of magma mixing in the formation of strategic platinum-group element (PGE) ore deposits. Neodymium and osmium isotopic data indicate that the intrusion formed from at least two geochemically distinct magmas. Ultramafic affinity (U-type) magmas had initial epsilon(Nd) of -0.8 to -3.2 and a chondritic initial (187)Os/(186)Os ratio of approximately 0.88, whereas anorthositic affinity (A-type) magmas had epsilon(Nd) of -0.7 to +1.7 and an initial (187)Os/(186)Os ratio of approximately -1.13. These data suggest that U-type magmas were derived from a lithospheric mantle source containing recycled crustal materials whereas A-type magmas originated either by crustal contamination of basaltic magmas or by partial melting of basalt in the lower crust. The Nd and Os isotopic data also suggest that Os, and probably the other PGEs in ore horizons such as the J-M Reef, was derived from A-type magmas. The Nd and Os isotopic heterogeneity observed in rocks below the J-M Reef also suggests that A-type magmas were injected into the Stillwater U-type magma chamber at several stages during the development of the Ultramafic series.

Hygromagmaphile element distributions in oceanic basalts as fingerprints of partial melting and mantle heterogeneities: a specific approach and proposal of an identification and modelling method

Summary On the basis of the properties of hygromagmaphile (HYG) elements we propose a specific method of identifying and modelling the evolution of mantle geochemical properties using simple mass balance equations and taking into account a large volume of data obtained for basaltic series belonging to characteristic oceanic domains. This method uses both binary and ternary HYG element ratio diagrams to distinguish between the effects of partial melting and mantle heterogeneity. It is assumed that batch partial melting is the simpler and more suitable process of basaltic magma generation, which produces linear correlations on binary HYG element ratio diagrams. Melting of a homogeneous source is projected on ternary HYG element ratio diagrams as a single point approximating the composition of the mantle source. We show that basalt HYG element ratios are closely dependent on four main factors, namely the nature of melting (batch partial melting), the degree of partial melting, the composition of the mantle and the mineralogy of the mantle. Partial-melting trends and complex mantle heterogeneities are clearly indicated by binary diagrams, and a simple and regular suboceanic mantle HYG element array is obtained in ternary diagrams. Taking into account the close relation with isotopic trends, the HYG element array is considered to be the fingerprint of ancient mantle differentiation. It can be reproduced by a simple fractional crystallization process, and mostly excludes a large-scale and intensive chemical mixing process. These results are in good agreement with a marble cake mantle model, limiting convection effects to a mechanical stirring of the upper mantle, but they do not necessitate any recycling of crustal material into the suboceanic mantle.

A strontium, neodymium and oxygen isotope study of hydrothermal metamorphism and crustal anatexis in the Trois Seigneurs Massif, Pyrenees, France

Nd, Sr, and O isotope analyses have been made on metamorphic and igneous rocks and minerals from a 310–340 Ma Hercynian-age metamorphic terrane in the Pyrenees, France. Lower Paleozoic shales and phyllites have 87Sr/86Sr values of 0.707–0.717 at 310 Ma, but model values at 310 Ma of 0.709–0.736 (based on assumed depositional age of 450 Ma and an initial 87Sr/86Sr=0.707). On a regional scale, 87Sr/86Sr was homogenized to about 0.713 to 0.717 in the higher-grade pelitic schists during metamorphism. Much of this 87Sr/86Sr exchange occurred at very low grades (below the biotite isograd), but significant changes also accompanied the δ18O lowering of the phyllites (+13 to +16) during their transformation to andalusite- and sillimanite-grade schists (δ18O=+11 to +12); all of these effects are attributed to pervasive interactions with hydrothermal fluids (Wickham and Taylor 1985). The data also show that a syn-metamorphic plutonic complex, dominated by a biotite granite body, was derived by mixing of a relatively mafic magmatic end-member (87Sr/86Sr∼ 0.7025–0.7050 and δ18O∼ +7.5 to +8.0) with two metasedimentary sources, both having 87Sr/86Sr∼0.715 and δ18O∼ +10.0 to +12.0, but with one being more homogeneous than the other. The more homogeneous component and the (mantle-derived?) magmatic end-member dominate at low structural levels within the complex. The less homogeneous end-member that dominates at high levels is clearly derived from the local Paleozoic pelitic schists. A Rb-Sr age of 330±20 Ma was obtained on hornblende from a deep level within the complex, which fixes this age for the regional metamorphism, as well. Although a post-metamorphic granodiorite magma body at Trois Seigneurs also displays heterogeneities in δ18O and 87Sr/86Sr (and thus does not give a clear-cut Rb-Sr isochron), the data are consistent with an emplacement age between 260 and 310 Ma, similar to ages of other late granodiorites in the Pyrenees. 143Nd/ 144Nd is very uniform within the Hercynian crust, both at Trois Seigneurs (ɛNd=−3 to −7) and elsewhere in the Pyrenees; almost all igneous lithologies have depleted-mantle, mid-Proterozoic model ages, consistent with efficient recycling of crustal material following original crustal accretion in this area at about 1600 Ma or earlier. Rb-Sr mineral ages exhibit a complex cooling history reflecting late Hercynian and Mesozoic thermal events. Our results show that profound homogenization of the 87Sr/86Sr and 18O/16O ratios of large volumes of the crust can occur during regional metamorphism and crustal anatexis, particularly in regions undergoing extensional tectonics. Such processes can significantly modify the isotopic compositions of the protoliths of granitic magmas; this may explain why many peraluminous Hercynian granitoids of Western Europe have anomalously low (87Sr/86Sr) initial values compared to their probable sedimentary parent rocks.

Recycling of the continental crust

In order to understand the evolution of the crust-mantle system, it is important to recognize the role played by the recycling of continental crust. Crustal recycling can be considered as two fundamentally distinct processes: 1) intracrustal recycling and 2) crust-mantle recycling. Intracrustal recycling is the turnover of crustal material by processes taking place wholly within the crust and includes most sedimentary recycling, isotopic resetting (metamorphism), intracrustal melting and assimilation. Crust-mantle recycling is the transfer of crustal material to the mantle with possible subsequent return to the crust. Intracrustal recycling is important in interpreting secular changes in sediment composition through time. It also explains differences found in crustal area-age patterns measured by different isotopic systems and may also play a role in modeling crustal growth curves based on Nd-model ages. Crustal-mantle recycling, for the most part, is a subduction process and may be considered on three levels. The first is recycling with only short periods of time in the mantle (<10 m.y.). This may be important in explaining the origin of island-arc and related igneous rocks; there is growing agreement that 1–3% recycled sediment is involved in their origin. Components of recycled crustal material, with long-term storage (up to 2.5 b.y.) in the mantle as distinct entities, has been suggested for the origin of ocean island and ultrapotassic volcanics but there is considerably less agreement on this interpretation. A third proposal calls for the return of crustal material to the mantle with efficient remixing in order to swamp the geochemical and isotopic signature of the recycled component by the mantle. This type of recycling is required for steady-state models of crustal evolution where the mass of the continents remains constant over geological time. It is unlikely if crust-mantle recycling has exceeded 0.75 km3/yr over the past 1–2 Ga.

Sulphur isotope heterogeneity in the mantle from ion microprobe measurements of sulphide inclusions in diamonds

The sulphur isotope composition of the mantle has been generally considered to be 0±2‰1. There is, however, considerable uncertainty in the mean δ34S value of the mantle and in its range because there have been very few analyses of unaltered sulphide phases of direct mantle origin1–4 and δ34S values of sulphides from large mafic intrusions range from –9 to +17‰5. We report here ion microprobe δ34S determinations from two mantle derived sulphide inclusions (≈100 µm) preserved in a diamond (δ13C = –4.25‰) of peridotitic(?) paragenesis from Premier (South Africa) and in four diamonds of eclogitic paragenesis from Orapa (Botswana). Three octahedral sulphides have lower δ34S values (+2.3±1.4‰) than three platelet-like sulphides (+8.2±0.9‰). This evidence for mantle heterogeneity for sulphur is interpreted in terms of recycling of crustal material into the eclogitic region of diamond growth in the mantle.

Diamonds and carbon isotopes

Diamond crystals may contain useful geochemical evidence of the deep portions of the upper mantle in their carbon isotopes. Two recent studies of types I and II diamonds showed that variations in δ13C may be related to carbon reservoir sources in the mantle (Nature, 303, 791–792, and 793–795, 1983). In the case of type II diamonds, it was noted that there is a strong—although not total— correlation with eclogitic suites. Type II diamonds are low in nitrogen content, which would be consistent with aluminum in their eclogitic inclusions having acted as a nitrogen getter. The range in δ13C values was −0.5 to −31.9‰, which actually is broad enough to include the variations of all known diamonds. With regard to individual crystals, however, it was confirmed that, on average, type II diamonds are isotopically lighter than type I. It is suggested that type II diamonds are a sampling of an open carbon isotope reservoir, with minimal evidence of fractionation. The variability may reflect the influences of recycled crustal material that reacted in the mantle.

Introductory remarks

In the past decade or so several major advances have significantly increased our understanding of continental crustal evolution. First, there is widespread recognition that the Wilson Cycle of plate opening and closure was responsible for the formation of Mesozoic-Cainozoic orogenic belts. Indeed the application of this plate tectonic scenario has been so successful that it is now possible to interpret the deformation of most Palaeozoic and several Proterozoic orogenic belts by accretionary or collisional processes. Secondly, there has been a surge of multidisciplinary research activity on Precambrian rocks, and especially on those formed before about 2.5 Ga B .P . during the Archaean. In the light of these advances, we are now beginning to understand the evolution of the continental crust in a continuous, long-term perspective with a global tectonic model that seems to be increasingly applicable to a wide range of young and old environments. This meeting report highlights one of the principal current controversies regarding continental crustal evolution: how much crust was created during, and by the end of, the Archaean, and how much subsequently? This question is directly related to the controversial role of secular recycling of crustal materials. According to one view (Armstrong), based mainly on isotopic and continental freeboard arguments, the total mass of continental crust was differentiated early in the Earth’s history, perhaps before 3.7 Ga B.P ., with negligible crustal growth since that time. This ‘steady-state’ model clearly implies efficient recycling of the bulk of eroded continental materials through the mantle throughout geological time

A Discussion on the evolution of the Precambrian crust - Problems of the evolution of the continental crust

In introducing this symposium I shall draw attention to some of the main problems concerning the early stages in the evolution of the continental crust. First, we wish to know how early the continental crust began to form, how fast it accumulated, when and by what processes the granitic and granodioritic rocks which form so much of it separated from the mantle, and whether these granitic materials were added continuously to the crust or in sharply defined phases. A few years ago it was thought that by studying 87Sr/86Sr initial ratios, it would be possible to discriminate between granitic rocks derived directly from the mantle and those derived from pre-existing crust. This method has shown that increments of granitic material are still being added to the crust from the mantle, as in Iceland (Moorbath &amp; Walker 1965). But recycling of crustal material through the mantle by way of subduction zones, depletion of the deeper parts of the crust in rubidium and the smaller difference between mantle and crustal strontium isotope ratios early in the Earth’s history make this method less decisive when it is applied to very old granites such as those which surround the greenstone belts of the Rhodesian craton and the Barberton mountain land. It has been proposed that the greenstones in these regions accumulated in an island arc environment on a thin crust (Anhaeusser, Mason, Viljoen &amp; Viljoen 1969) and that most of the enveloping granites represent subsequent additions to the crust (White, Jakes &amp; Christie 1971). Yet before 2550 Ma ago when the Great Dyke of Rhodesia was intruded, the granites which constitute much of the Rhodesian craton had already been emplaced and the crust there must have reached almost its present thickness. Structural and stratigraphic evidence suggest that much of the granitic material represents reactivated continental crust older than the rocks of the greenstone belts (Shackleton 1970). Isotopic studies of lead from the Tanzanian, Rhodesian and Canadian cratons indicate that a protocrust existed at least 4000 Ma ago in these regions (Robertson 1970).

Isotopic and chemical constraints on models of magma genesis in volcanic arcs

Lead isotopic compositions of New Zealand and Lesser Antilles calc-alkaline volcanic arc magmas and of oceanic sediments are very similar, and so distinct from the isotopic composition of abyssal tholeiites that a simple genetic connection between abyssal tholeiites and volcanic arc magmas is ruled out. Recycling of crustal material through the mantle seems the best explanation for the Pb isotopic similarityof volcanic arc magmas and oceanic sediments. Massbalance calculations demonstrate that the flux of K, Rb, Cs, Pb, Ba and Th in oceanic basaltic crust created continuously along mid-ocean rises is insufficient by factors of 2 to 10 to supply the amounts required for generation of andesites and continental rocks at observed rates. Sediment contamination linked with the Green and Ringwood model for magma genesis can explain both the isotopic and chemical data.