Content uploaded by Albrecht W. Hofmann
Author content
All content in this area was uploaded by Albrecht W. Hofmann on Sep 02, 2014
Content may be subject to copyright.
Mantle geochemistry: The message from oceanic volcanism
Abstract
Basaltic volcanism 'samples' the Earth's mantle to great depths, because
solid-state convection transports deep material into the (shallow)
melting region. The isotopic and trace-element chemistry of these
basalts shows that the mantle contains several isotopically and
chemically distinct components, which reflect its global evolution. This
evolution is characterized by upper-mantle depletion of many trace
elements, possible replenishment from the deeper, less depleted mantle,
and the recycling of oceanic crust and lithosphere, but of only small
amounts of continental material.
The Mesozoic-Cenozoic geological evolution of the East Asian continental margin and a series of Cenozoic back-arc basins (South China Sea, West Philippine Sea, Sea of Japan and Shikoku Basin) developing in the West Pacific region have been predominately resulted from the westward subduction of Pacific plate, accompanying with the effects of some intraplate processes. However, how the intraplate processes work remains unclear. Here we present new geochemical and Sr-Nd-Pb-Hf isotope data for the lava samples collected from Deep Sea Drilling Project Expeditions 6, 58, and 59 in the Shikoku Basin (SKB) and its adjacent region. High Nb/La ratio and enriched Sr-Nd-Pb-Hf isotopic compositions indicate that the spreading-related lavas in the SKB were affected by the enriched mantle type I (EMI) component, which can be solely produced by mixing enriched mantle components and ambient mantle. We suggested that EMI components in the asthenosphere mantle beneath the region from northeast China to Sea of Japan may be involved in the back-arc spreading processes by means of asthenospheric toroidal flow around the subducting Philippine Sea slab, or the EMI components have preexisted in the asthenos-pheric mantle beneath the SKB prior to the northward migration of the Philippine Sea Plate since Equator.
Mantle plumes—upwellings of buoyant rock in Earth's mantle—feed hotspot volcanoes such as Hawai‘i. The size of volcanoes along the Hawai‘i–Emperor chain, and thus the magma flux of the Hawaiian plume, has varied over the past 85 million years. Fifteen and two million years ago, rapid bursts in magmatic production led to the emergence of large islands such as Pūhāhonu, Maui Nui and Hawai‘i, but the underlying mechanisms remain enigmatic. Here, we use new radiogenic Ce–Sr–Nd–Hf isotope data of Hawaiian shield lavas to quantify the composition and proportion of the different constituents of the Hawaiian plume over time. We find that most of the Hawaiian mantle source is peridotite that has experienced variable degrees of melt depletion before being incorporated into the plume. We show that the most isotopically enriched LOA‐type compositions arise from the aggregation of melts from more depleted, trace element‐starved peridotite, causing the over‐visibility of melts from recycled crust in the mixture. Our results also show that upwelling of chemically more depleted, and thus less dense, more buoyant mantle peridotite occurred synchronously to an observed burst of magma production. Buoyancy variations induced by variably depleted peridotite may not only control the temporal patterns of volcanic productivity in Hawai‘i, but also those of other plumes world‐wide. The excess buoyancy of depleted peridotite may therefore be an underrated driving force for convective mantle flow, trigger and sustain active upwelling of relatively cool plumes, and control the geometry of mantle upwellings from variable depths.
The Islamic Peninsula (Saray Volcano) is located east of Lake Urmia and west of Sahand Volcano. It is composed of alternating lava flows and pyroclastic deposits, subsequently intruded by dikes and dome structures. The dominant trachytic and leucitite lava flows contain xenoliths of varying sizes and morphologies. The primary minerals in the trachytic host rock include sanidine, biotite, and opaque minerals. In contrast, the leucitite host rocks are mainly composed of leucite and clinopyroxene, set within a fine-grained groundmass of leucite, clinopyroxene, olivine, and opaque
minerals. The boundary between the host rock and xenoliths is sharp, lacking any reaction rims. The xenoliths primarily consist of clinopyroxene, biotite, and apatite. Some clinopyroxene and biotite crystals within the leucitite xenoliths display elongated and oriented textures. The clinopyroxene compositions in both the xenoliths and the trachytic and leucitite host rocks range from diopside to fassaite. Micas in the trachytic host rock and associated xenoliths are biotite and phlogopite, whereas those in the leucitite host rock and its xenoliths are exclusively phlogopite. The micas within the leucitite host rock and xenoliths are primary biotites, whereas those in the trachytic host rock exhibit both primary and recrystallized biotite phases. The estimated
crystallization temperatures for clinopyroxenes in the trachytic and leucitite xenoliths and host rocks range between 1200–1300°C. Specifically, clinopyroxenes in the trachytic host rock formed at temperatures of 1027–1111°C, while those in the leucitite host rock crystallized between 996–1143°C and 1044–1140°C. The calculated pressure for clinopyroxenes in both rock types is approximately 6–10 kbar, corresponding to an estimated depth of 20–35 km. The negative Ti, Nb, and Ta anomalies in the leucitite host rock suggest varying degrees of crustal contamination and a possible association with subduction-related active continental margin settings. The close similarity in the Mg-number of clinopyroxene and biotite in both xenoliths and host rocks indicates a common parental magma. However, subsequent heteromorphic differentiation under distinct physico-chemical conditions resulted in the formation of compositionally similar but mineralogically different rock types. The mantle source of the Saray volcanic activity is identified as a phlogopite-bearing garnet lherzolite.Structurally, two major fault zones, the ENE-WSW trending Saray-Aq Gonbad fault and the SSE-NNW trending Teymurlu-Gamichi fault, intersect in the central part of the peninsula, creating a fractured and weakened zone that facilitated magma ascent. The ascending magma underwent crystallization in magma chambers, forming cumulate phases that were later transported to the surface by magmatic eruptions.
Vanadium (V) isotopes were used to trace magmatic processes during planetary differentiation, however, whether and how they could be fractionated in the terrestrial mantle remains unclear. Here, we report V isotopic compositions for samples from the Balmuccia (BM) and Baldissero (BD) massifs in the Italian Alps, and Zhimafang in Dabie-Sulu orogen of China, as well as for xenoliths in basalts from Tariat (Mongolia), Tok (Siberia), and Panshishan and Tashan in eastern China. These samples are grouped, according to their origin, as follows: Group 1 combines 20 fertile to moderately melt-depleted, unmetasomatised lherzolites from massifs and xenoliths; Group 2a combines five strongly melt-depleted peridotites (harzburgites, dunites) and 12 associated vein pyroxenites from the Italian Alps; Group 2b samples are three composite xenoliths (pyroxenite veins in peridotites) from Tariat. Group 3 comprises strongly metasomatized peridotites (two garnet lherzolites and a wehrlite) from Tok and Zhimafang. The δ51V values in Group 1 peridotites range from −1.14‰ to −0.82‰ and show broad covariations with melt-extraction indices (e.g., Al2O3). Melting models indicate that V isotopes may fractionate during partial melting with △51Vresidue-melt of −0.15‰ to −0.10‰. The δ51V values in Group 2a peridotites range from −1.19‰ to −0.98‰, possibly involving minor redistribution of V between residual peridotites and parental melts of pyroxenite veins that show δ51V from −1.09‰ to −0.92‰. Group 2b composite xenoliths show contrasted δ51V values between peridotite hosts (−1.05‰ to −0.85‰) and pyroxenite veins (−1.32‰ to −0.99‰). Host-vein δ51V differences in the xenoliths range from 0.05 to 0.27‰, likely reflecting diffusion-driven kinetic V isotope fractionation during melt intrusion and the formation of the pyroxenites. The δ51V values in Group 3 peridotites (−1.04‰ to −0.92‰) are close to the BSE estimate (−0.91 ± 0.09‰). Fe-Ca-V-rich melt metasomatism may slightly lower δ51V, while fluid metasomatism preserves earlier magmatic values.
We report first δ51V data for mineral separates from mantle rocks (Groups 1 and 3). They show narrow, BSE-like ranges in spinel (−0.89‰ to −0.85‰), garnet (−0.85‰ to −0.80‰), orthopyroxene (−0.99‰ to −0.95‰) and phlogopite (−1.02‰ to −0.96‰), but a greater δ51V range (−1.13‰ to −0.77‰) in clinopyroxene. Clinopyroxene from Group 1 displays a spongy texture, while clinopyroxene from Group 3 exhibits textural and trace element evidence for metasomatism. The lower δ51V values in clinopyroxene may reflect isotope fractionation during solidification of evolved interstitial liquids. However, mass balance calculations indicate that mineral-scale variations have a negligible impact on whole-rock V isotopic compositions. Overall, V isotopes can be fractionated in the mantle with δ51V from −1.32‰ to −0.82‰ by a combination of partial melting and melt-solid interactions.
A defense is conducted of the proposition that continental crust is recycled into the mantle and that the earth is in a near-steady state, with essentially constant volumes of ocean and crust through geological time. A contrasting view, that the continental crust has grown with time, has been repeatedly expressed by Moorbath (1977, 1978). It is pointed out that his arguments against recycling are not persuasive, and that the evidence which he presents is easily accommodated in a no-continental-growth model. Attention is given to evidence of continent and ocean volume, early planetary differentiation, continental accretion and destruction, evidence for the subduction of sediment, a quantitative simulation of the no-growth hypothesis, the Pb isotope evolution, the apparent single-stage evolution of Pb and Sr isotopes in ancient rocks, the inert gases He-3 and Ar-36, stable isotopes in sediments and ocean water, and steady-state chemical sedimentology.
Mantle and crust evolution is discussed in terms of two simple transport models. In model I, continents (j = 3) are derived by melt extraction over the history of the earth from undepleted mantle (j = 1), and the residue forms a depleted mantle (j = 2), which today is the source of mid-ocean ridge basalts. In model II _ additions to continents are derived from a mantle reservoir 2, which becomes more de-pleted through time by repeated extraction of melts. Transport equations were solved for stable s, radio-active r, and daughter d isotopes for arbitrary mass growth curves MAT). For both models the isotopic composition and --,trations of trace elements are shown to mince to simple mathematical ex-pressions which readily permit calculations of basic evolutionary parameters from the data. For long-lived isotopes (X -I >> 4.5 aeons) for model I the deviations in parts in 10 4 of the ratio of a daughter isotope to a stable reference isotope of the same element in reservoirs j = 2, 3 from that of i is given by ed,* = Qd* tmX is. Here t hi j is the mean age of the mass of j, f7' is the enrichment factor of the ratio of a radio-active isotope to a stable isotope relative to that in 1, and Qa * = const. Thus for long-lived isotopes such as I 'Srn and 87Rb the only time information that can be obtained from model I from measurement 6f the relative chemical enrichment factors and isotopic ratios at a single time is the mean age of the mass of the continental crust and the complementary depleted mantle reservoir. This mean age is independent of the long-lived parent-daughter system. An analogous result is obtained for model II, where Ed.2 * = Qd * (f i
It is noted that different physicists and geologists have in recent years espoused not less than four groups of theories of the physical behavior of the Earth's interior. Recent observations of submarine geology, heat, and rock magnetism have tended to support some form of continental drift rather than the older concept of a rigid earth.The Hawaiian Islands are one of seven, parallel, linear chains of islands and seamounts in the Pacific Ocean of Tertiary to Recent age. Their nature had previously been explained in terms of a series of volcanoes along parallel faults. Horizontal shear motion along these faults was supposed to be extending them southeasterly.The inadequacies of this explanation are pointed out. If there are convection currents in the Pacific region and if the upper parts of these cells move faster than the central parts, sources of lava within the slower moving cores could give rise to linear chains of progressively older volcanic piles such as the Hawaiian Islands. This view is shown to be compatible with seismic observations and age determinations.
It has been 23 years since the oceanic mantle was first shown to be isotopically heterogeneous (Gast, Tilton & Hedge, 1964). The richness and diversity of this heterogeneous character has been expanded over the intervening years with the application of new isotopic tracer systems (Nd, Hf, He, Xe) to volcanic rocks from most of the oceanic islands. Currently, this isotopic heterogeneity can be embraced as mixtures between four or five fairly distinct isotopic components (DMM: depleted MORB mantle; HIMU: a high U/Pb component; and two enriched components, EMI and EMII; Zindler and Hart, 1986). The DMM component comprises the upper mantle, and has been depleted over geologic time by processes associated with extraction of continental crust. This is the only mantle reservoir for which a location and evolutionary process are well established. Chase (1981) and Hofmann & White (1982) proposed that subducted oceanic crust ± sediment would acquire significant isotopic heterogeneity while aging in the mantle, and that recycling of this material could provide some of the isotopic diversity observed in oceanic basalts.
The island of Tristan da Cunha, located at 37°S, 12°W in the South Atlantic, is the largest of a group of three islands, the
others being Nightingale and Inaccessible. Tristan da Cunha comprises a continuous series of alkaline lavas ranging in composition
from ankaramitic basanite through phonotephnte and tephnphonolite to phonolite. Moderately porphyritic basanite is the dominant
rock type (˜on the island.
Major and trace element variations in the lavas describe well-defined trends with increasing differentiation which are generally
consistent with control by fractional crystallization of phenocryst phases. None of the lavas can be considered to be primary
in composition, mg-numbers range from 40 to 62, and covariation of certain minor and trace elements (e.g., Sr, Ba, P) suggests the presence
of at least two distinct fractionation trends. Sr, Nd, and Pb isotopic analyses of a subset of the lavas confirm previously
published data for the island, but show a slightly greater range: 87Sr/86Sr = 0˙70495–0˙70517; 143Nd/144Nd = 0˙51259–0˙51247; 206Pb/204Pb = 18˙47–18-˙74.
Quantitative modelling of the compositional variations suggests that the ankaramitic basanites are partial olivine + clinopyroxene
+ titanomagnetite (±minor plagioclase) cumulates, with ˜40% crystal accumulation being required to account for the most porphyritic
varieties. The range in composition from basanite to phonotephrite can be accounted for by up to 50% fractional crystallization
of clinopyroxene, olivine, titanomagnetite, and plagioclase, with minor apatite and, in some models, amphibole. Average proportions
of these phases in the fractionate are Cpx 40, TiMgt 20, Plag 30, Oliv 10. As much as 20% amphibole fractionation is required
in models involving the phonotephrites. The compositions of the evolved tephriphonolites and phonolites are consistent with
extensive (up to 80%) fractional crystallization of an initial basanitic magma with clinopyroxenc (1–4%), amphibole (17–23%),
plagioclase (6–20%) alkali feldspar (0–13%), and titanomagnetite (4–6%)±minor apatite and sphene being the dominant fractionating
phases.
Inferred trace element and isotopic characteristics of the source regions of the Tristan lavas are distinct from those giving
rise to the Walvis Ridge, Gough Island, or Discovery Seamount basalts. Normalized trace element abundances of the Tristan
lavas are more similar to those of Marion Island, whereas Sr-, Nd-, dnd Pb-isotopic ratios are most similar to nearby Inaccessible
Island lavas. If Tristan da Cunha is the present-day surface expression of the upwelling mantle plume that previously gave
rise to the Walvis Ridge, then the source material tapped by the lavas is distinctly heterogeneous (or has changed with time)
within the limits imposed by maintaining the general characteristics of DUPAL-type mantle.
Simple arguments show that ascending thermal plumes will entrain their surroundings as the result of coupling between conduction of heat and laminar stirring driven by the plume motion. In the initial stages of ascent of a plume fed by a continuous buoyancy flux (a starting plume) the plume consists of a large buoyant head followed by a narrow vertical conduit. Laboratory experiments reported here show that the spherical head entrains ambient material as it rises, while the axial conduit carries hot source material to the stagnation point at the cap of the plume, from where it spreads laterally into thin laminae. We develop an analysis of the effects of entrainment on the structure and dynamics of starting plumes. The analysis predicts that under conditions appropriate to the earth's mantle large volumes of cooler lower mantle will be stirred into the head of a plume by the time it reaches the top of the mantle if it originates at the core-mantle boundary. The result is a major cooling and enlargement of the head. Source material ascending in the trailing conduit will undergo little contamination or cooling until the conduit is deflected from the vertical by large scale shear associated with plate motion. This plume structure explains the close association of high-temperature melts (komatiites or picrites) with more voluminous, lower temperature basalts in Archaean greenstones and modern continental flood basalt provinces: the picrites can be produced by melting in the hot axial conduit and the basalts from the cooler bulk of the head. More generally, we put forward stirring in plumes as one plausible mechanism contributing to compositional heterogeneity in hotspot melts.The predicted diameter of plume heads originating at the core-mantle boundary is ∼ 1000 km and this is expected to enlarge to ∼ 2000 km when the plume collapses beneath the lithosphere. This result is in excellent agreement with the observed extent of volcanism and uplift associated with continental flood volcanism. It also provides support for the hypothesis that at least some plumes originate at the core-mantle boundary.
A wide range of geophysical and geochemical observations pertaining to convection in the earth's mantle and the dynamics of the tectonic plates is discussed. It is inferred that the dominant model of mantle convection is a plate-scale flow and that the plates are an integral part of this flow. Upwelling buoyant plumes, that cause volcanic hotspots, are inferred to comprise a secondary model of convection arising from a relatively weak thermal boundary layer at the base of the mantle. A significant viscosity increase is inferred by perhaps two to three orders of magnitude, through the depth of the mantle, with a large part of this increase occurring through the transition zone. With ridges sampling the top of the mantle and plumes sampling the bottom, these features offer explanations for the main geochemical characteristics of, and differences between, mid-ocean ridge basalts and oceanic island (hotspot) basalts. -from Authors
Flood basalt volcanism, defined by huge accumulations of magma erupted over relatively short time intervals, is a primary means of extracting melts from the mantle to form new crust. Even though the duration of volcanism is short within any given province, the volumetric sum of oceanic and continental flood basalt accumulations in the Phanerozoic rivals the amount of crust created in the worldwide system of convergent margins. The larger flood basalt provinces consist of millions of cubic kilometres of tholeiitic basalt erupted over time intervals as short as a million years or less. Magma production rates required to supply flood basalt volcanism are at least an order of magnitude greater than is typical of intraplate or convergent margin volcanism. The very high melting rates associated with flood basalt volcanism are difficult to explain by simple adiabatic decompression of mantle rising beneath an extensional continental terrain and may require a contribution from anomalously hot mantle. In most cases, flood basalt eruptions can be correlated in time and space with the initial stages of activity of volcanic hot spots leading to the conclusion that the anomalously hot mantle may be supplied by a plume from the deep mantle. In their chemical and isotopic characteristics, some flood basalts mimic the composition of the basalts associated with the companion intraoceanic hot spot thereby supporting the flood basalt-hot spot link. Others trend toward compositions similar to oceanic lavas derived from depleted mantle. Many flood basalts, however, have chemical characteristics typical of shallow (i.e. lithospheric) fractionation and isotopic signatures more typical of continental crust than of the convecting mantle. Some of these characteristics may be imposed upon magmas by crustal or lithospheric-mantle contamination, others may be signatures of magma sources in the subcontinental lithospheric mantle.