Cook-Austral Islands Research Articles

Magmatic water content in HIMU basalts from the Cook-Austral Islands: constraints on degassing processes and source composition from clinopyroxene phenocrysts

The amount of water recycled during subduction is unclear. Water/Ce estimates in HIMU (high-238U/204Pb) basalts are variable, ranging from < 100 to > 250 in glasses and melt inclusions. Because clinopyroxene (cpx) is a common early liquidus phase and the cpx/melt partitioning of water is well constrained, cpx phenocrysts provide an additional constraint on magmatic water contents. We present water and trace element concentrations in HIMU-basalt-hosted cpx phenocrysts from the Austral Islands. Calculated melt [H2O] (up to 4.3 wt.%) and H2O/Ce ratios (22–825) are higher than in olivine-hosted inclusions, as are estimated equilibration pressures. Correlation between estimated melt [H2O] and crystallization or entrapment pressure suggests significant water loss during magma ascent. Most ocean island basalts (OIBs) span a limited range in H2O (~ 1–1.5 wt.%), and low (< 100) H2O/Ce ratios are primarily observed in melts with unusually high [Ce] (up to 350 ppm). Additionally, [H2O] and H2O/Ce in some suites correlate with entrapment pressure despite having quench pressures high enough to prevent significant water loss from open- or closed-system degassing (< 100 MPa). Polybaric “sparging”, whereby low-P melts re-equilibrate with CO2-rich fluids exsolved at higher pressure, may result in water loss at pressures less than CO2 saturation. This may more accurately describe OIB degassing processes than open or closed system degassing. After correcting for degassing, primary Australs melts likely have H2O/Ce of ~ 300–600. If applicable to OIB sources in general, this limits the total water budget of the mantle, including the mantle transition zone, to < 2.4 ocean masses.

Zinc isotopic evidence for recycled carbonate in the deep mantle

Carbonate, the major carbon reservoir on Earth’s surface, can enter into the mantle by subduction. However, evidence for recycled surficial carbonates in the deep mantle is still scarce. Ocean island basalts from Cook-Austral islands and St. Helena Island, widely called HIMU basalts because of their high μ = 238U/204Pb sources, are thought to be fed by mantle plumes originating in the lower mantle. Here we report exceptionally high δ66Zn values (δ66Zn = 0.38 ± 0.03‰) of these HIMU lavas relative to most published data for oceanic basalts (δ66Zn = 0.31 ± 0.10‰), which requires a source contributed by isotopically heavy recycled surficial carbonates. During subduction of the oceanic lithosphere, melting of mixed surficial carbonates and basaltic crust in the deep mantle generates carbonatite melts, which metasomatizes the nearby mantle and the resultant carbonated mantle ultimately evolves into a high-δ66Zn HIMU source. High-δ66Zn signatures of HIMU basalts, therefore, demonstrate that carbonates can be transported into Earth’s deep mantle.

Sr–Pb–Nd–Hf isotopes and 40Ar/39Ar ages reveal a Hawaii–Emperor-style bend in the Rurutu hotspot

Current Pacific absolute plate motion (APM) models include two major, long-lived hotspot tracks: the ∼80 Ma Hawaii–Emperor and ∼79 Ma Louisville seamount tracks. Prior to ∼50 Ma, these two tracks show significant inter-hotspot drift, most likely due to a strong southern motion of the Hawaiian plume. With the addition of a third long-lived Pacific hotspot track, featuring a well-constrained hotspot location and a Hawaii–Emperor-style bend, we can achieve a more robust evaluation of the relationship between APM models and inter-hotspot drift. Such a candidate exists in the Rurutu hotspot track. The youngest volcanism (∼10 to 0.2 Ma) associated with the Rurutu hotspot is located in the Cook-Austral Islands and is known as the “Atiu” or “Young Rurutu” trend. Here we show with 40Ar/39Ar ages and Sr–Pb–Nd–Hf “isotopic fingerprints” that islands and seamounts ranging between ∼70 to 40 Ma fit the compositional trend and APM model-predicted locations of the Rurutu hotspot.The Rurutu hotspot has a distinctive HIMU- to FOZO-like composition, characterized by high 206Pb/204Pb (≥20.0) and low 87Sr/86Sr (≤0.704) ratios. HIMU is a rare composition, found in only two other hotspots worldwide. Uniquely, elevated Rurutu 208Pb/204Pb ratios relative to 206Pb/204Pb define a steeper regression that more closely reproduces the Northern Hemisphere Reference Line than the shallower regression defined by traditional HIMU. Rurutu also is the only HIMU-type hotspot known to continue into the West Pacific, allowing its unique composition to effectively be used as a fingerprint. The Rurutu signature can be traced through non-Samoan “interloper seamounts” in the Samoan region, to a hotspot bend in the southern Tuvalu Islands, continuing through the ∼74 to 63 Ma Gilbert Ridge and the ∼120 to 78 Ma Wake seamounts. However, precisely locating the Hawaii–Emperor style bend in Rurutu is complicated by the younger Samoan hotspot partially overlapping the Rurutu hotspot track, and Cretaceous volcanic structures predating regional Rurutu volcanism. Therefore, we present a computational method that tests a range of preexisting APM models, all generating Hawaii–Emperor style bend locations for Rurutu within error of each other. Our results confirm the presence of a third major hotspot track in the Pacific as originally proposed by W.J. Morgan (1972), with a ∼50.5–49.0 Ma bend at 178.48°E and 8.56°S.Paleomagnetic data for Hawaii indicate that changes in southern hotspot motion may explain its bend, while negligible latitudinal motion in Louisville and an unchanging distance to Rurutu (Tarduno et al., 2003; Koppers et al., 2012; Konrad et al., 2018) are consistent with subduction-driven plume motion explaining the Hawaii–Emperor bend. The presence of a defined ∼135° bend in the Rurutu hotspot track, coupled with minimal inferred latitudinal motion at that time (Konrad et al., 2018), leaves room for a small change in Pacific plate motion at ∼50 Ma.

Thallium isotope systematics in volcanic rocks from St. Helena – Constraints on the origin of the HIMU reservoir

It is generally accepted that subduction of the oceanic crust plays an important role in creation of mantle heterogeneity. HIMU (stands for high μ=(238U/204Pb)) is a rare and probably not volumetrically large mantle component, which is sampled at St. Helena in the Atlantic and in the Cook-Austral Islands in the South Pacific. Previous studies that used radiogenic isotopes have specifically invoked subducted hydrothermally altered oceanic crust (AOC) as a significant component in the source of HIMU. However, because radiogenic isotopes are affected by both parent-daughter fractionation and time that cannot be independently determined, it is difficult to unambiguously use these tracers to determine the source material of the HIMU reservoir.Thallium isotopes can be used to identify the involvement of sediments rich in Mn oxides or AOC, because these two components exhibit high Tl concentrations and highly fractionated Tl isotope compositons, whereas the upper mantle is depleted in Tl and homogenous with respect to Tl isotopes. Here we report Tl isotope data for a suite of St. Helena lavas previously characterized to investigate if the HIMU mantle reservoir can be linked to AOC. Samples show large Tl isotopic variation, from ε205Tl=−9.8 to ε205Tl=+3.9 and do not correlate with Sr, Nd, Hf or Pb isotopes. Eight samples have ε205Tl lighter than the mantle value (ε205T=−2), with the lightest values similar to those observed for modern AOC. These features are consistent with the presence of different proportions of recycled oceanic crust in the source of St. Helena mantle source. Our results, therefore, support previous studies that inferred AOC as a significant component in the HIMU mantle reservoir and show that the Tl isotopic system can be used as a powerful tracer to detect contributions of AOC in the creation of mantle heterogeneities.

Slab-derived components in the subcontinental lithospheric mantle beneath Chilean Patagonia: Geochemistry and Sr–Nd–Pb isotopes of mantle xenoliths and host basalt

In subduction zones, ultramafic xenoliths hosted in alkaline basalts can yield significant information about the role of potential slab-derived components in the subcontinental lithospheric mantle (SCLM). Chemical and isotopic heterogeneities in such xenoliths are usually interpreted to reflect melt extraction followed by metasomatic re-enrichment. Here we report new whole-rock major, trace element and isotopic (Sr–Nd–Pb) data for a Proterozoic suite of 17 anhydrous spinel-lherzolites and Eocene (new K–Ar data) host alkaline basalt found near Coyhaique (~46°S), Aysén Region, Chile. These Patagonian nodules are located in a current back-arc position, ~100km east of the present day volcanic arc and ~320km from the Chile Trench.The mantle xenoliths consist of coarse- to medium-grained spinel-lherzolites with trace element compositions characteristic of a subduction zone setting, such as pronounced negative Nb, Ta and Ti anomalies coupled with significant enrichment of LILEs (e.g., U) and chalcophile elements (W, Pb and Sn). Most of them are characterized by flat to depleted light-rare earth element (LREE) patterns (Ce/YbN=0.6–1.1) coupled with less radiogenic Sr–Pb (87Sr/86Sr=0.702422–0.703479; 206Pb/204Pb=18.212–18.539) and more radiogenic Nd isotopic compositions (143Nd/144Nd=0.512994–0.513242), similar to the depleted mantle component (DMM or PREMA). In contrast, samples with slight LREE enrichment (Ce/YbN=1.3–1.8) show more radiogenic Sr–Pb (87Sr/86Sr=0.703791–0.704239; 206Pb/204Pb=18.572–18.703) and less radiogenic Nd isotopic compositions (143Nd/144Nd=0.512859–0.512934), similar to the EM-2 reservoir. These new geochemical and isotope data suggest that the Coyhaique spinel-lherzolites are derived from a heterogeneous SCLM resulting from mixing between a depleted mantle component and up to 10% of slab-derived components. The enriched component added to the SCLM represents variable extents of melts of both subducted Chile Trench sediments and modified oceanic crust throughout the initial stages of the Farallón–Aluk ridge collision during Paleocene to Eocene time. However, based on the tectonic evolution of southern South America, we cannot exclude the influence of long-lived subduction events beneath south Patagonia. Although we believe that the studied samples were brought to the surface in this geodynamic context, there is no evidence that ocean island basalt (OIB)–like melts related to the Farallón–Aluk asthenospheric slab window affected the peridotite composition.The host alkaline basalt is a single unit with a HIMU-like OIB signature characterized by marked positive Nb–Ta anomalies coupled with negative anomalies in highly incompatible and fluid-mobile elements (Rb, K, Pb, and Sr). The compositional similarity between the HIMU-like OIB mantle source and the host basalt is also evident from trace element ratios [(Ba–Th–K–La–Zr)/Nb] as well as by the low 87Sr/86Sri (0.703039–0.703058) and relatively high 143Nd/144Ndi (0.512880–0.512874) and 206Pb/204Pb (19.333–19.389) isotopic ratios. The low 206Pb/204Pb ratios compared to end-member HIMU lavas (e.g., Sta. Helena and the Cook-Austral Islands) suggest that this region was modified by processes associated with a prolonged period of subduction related to the Andean orogenesis and the recycling of several oceanic plates beneath the continent, following the Mesozoic breakup of Gondwana or an even older subduction-related event with young recycling ages (<2Ga).

P34 クックーオーストラル諸島火山岩類の K-Ar 年代及び鉛同位体組成の時間空間変化

P34 クックーオーストラル諸島火山岩類の K-Ar 年代及び鉛同位体組成の時間空間変化

Iron isotope tracing of mantle heterogeneity within the source regions of oceanic basalts

Mineralogical variations in the Earth's mantle and the relative proportions of peridotitic versus enriched and potentially crustally-derived pyroxenitic domains within the mantle have important implications for mantle dynamics, magma generation, and the recycling of surface material back into the mantle. Here we present iron (Fe) stable isotope data (δ57Fe, deviation in 57Fe/54Fe from the IRMM-014 standard in parts per thousand) for peridotite and garnet–pyroxenite xenoliths from Oahu, Hawaii and explore Fe isotopes as tracer of both peridotitic and pyroxenitic components in the source regions of oceanic basalts. The pyroxenites have δ57Fe values that are heavy (0.10 to 0.27‰) relative to values for mid-ocean ridge and ocean island basalts (MORB; OIB; δFe57∼0.16‰) and the primitive mantle (PM; δFe57∼0.04‰). Pyroxenite δ57Fe values are positively correlated with bulk pyroxenite titanium and heavy rare earth element (REE) abundances, which can be interpreted in terms of stable isotope fractionation during magmatic differentiation and pyroxene cumulate formation. In contrast, the peridotites have light δ57Fe values (−0.34 to 0.14‰) that correlate negatively with degree of melt depletion and radiogenic hafnium isotopes, with the most depleted samples possessing the most radiogenic Hf isotope compositions and lightest δ57Fe values. While these correlations are broadly consistent with a scenario of Fe isotope fractionation during partial melting, where isotopically heavy Fe is extracted into the melt phase, leaving behind low-δ57Fe peridotite residues, the extent of isotopic variation is far greater than predicted by partial melting models. One possibility is derivation of the samples from a heterogeneous source containing both light-δ57Fe (relative to PM) and heavy-δ57Fe components. While pyroxenite is a viable explanation for the heavy-δ57Fe component, the origin of the depleted light-δ57Fe component is more difficult to explain, as melting models predict that even large (>30%) degrees of melt extraction do not generate strongly fractionated residues. Multiple phases of melt extraction or other processes, such as metasomatism, melt percolation or the assimilation of xenocrystic olivine with light δ57Fe values may need to be invoked to explain these light δ57Fe values; a caveat to this is that these processes must either preserve, or generate correlations between δ57Fe and Hf isotopes. Published variations in δ57Fe in mantle melting products, such as MORB and OIB, are also greater than predicted by melting models assuming derivation from δ57Fe-homogeneous mantle. For example, OIB from the Society and Cook-Austral islands, which have radiogenic Pb and Sr isotope compositions indicative of recycled components such as subduction modified, low-Pb oceanic crust and terrigenous sediments have heavy mean δ57Fe values (∼0.21‰) significantly distinct to those of other OIB and MORB, which could explained by the presence of heavy-δ57Fe pyroxenite cumulate or pyroxenitic melt components, whereas large degree partial melts, such as komatiites and boninites, display light Fe-isotopic compositions which may reflect sampling of refractory, light-δ57Fe mantle components. Iron stable isotopes may therefore provide a powerful new means of fingerprinting mineralogical variations within the Earth's mantle and identifying the mineralogy of depleted and enriched components within the source regions of volcanic rocks.

Phantom Archean crust in Mangaia hotspot lavas and the meaning of heterogeneous mantle

Lavas from Mangaia in the Cook–Austral island chain, Polynesia, define an HIMU (or high μ, where μ=U238/Pb204) global isotopic end-member among ocean island basalts (OIB) with the highest 206,207,208Pb/204Pb. This geochemical signature is interpreted to reflect a recycled oceanic crust component in the mantle source. Mass independently fractionated (MIF) sulfur isotopes indicate that Mangaia lavas sampled recycled Archean material that was once at the Earth's surface, likely hydrothermally-modified oceanic crust. Recent models have proposed that crust that is subducted and then returned to the surface in a mantle plume is expected to transform to pyroxenite/eclogite during transit through the mantle. Here we examine this hypothesis for Mangaia using high-precision electron microprobe analysis on olivine phenocrysts. Contrary to expectations of a crustal component and, hence pyroxenite, results show a mixed peridotite and pyroxenite source, with peridotite dominating. If the isotopic compositions were inherited from subduction of recycled oceanic crust, our work shows that this source has phantom-like properties in that it can have its lithological identity destroyed while its isotope ratios are preserved. This may occur by partial melting of the pyroxenite and injection of its silicic melts into the surrounding mantle peridotite, yielding a refertilized peridotite. Evidence from one sample reveals that not all pyroxenite in the melting region was destroyed. Identification of source lithology using olivine phenocryst chemistry can be further compromised by magma chamber fractional crystallization, recharge, and mixing. We conclude that the commonly used terms mantle “heterogeneities” and “streaks” are ambiguous, and distinction should be made of its lithological and isotopic properties.

Coupled Hf–Nd–Pb isotope co-variations of HIMU oceanic island basalts from Mangaia, Cook-Austral islands, suggest an Archean source component in the mantle transition zone

Although it is widely accepted that oceanic island basalts (OIB) sample geochemically distinct mantle reservoirs including recycled oceanic crust, the composition, age, and locus of these reservoirs remain uncertain. OIB with highly radiogenic Pb isotope signatures are grouped as HIMU (high-μ, with μ=238U/204Pb), and exhibit unique Hf–Nd isotopic characteristics, defined as ΔεHf, deviant from a terrestrial igneous rock array that includes all other OIB types. Here we combine new Hf isotope data with previous Nd–Pb isotope measurements to assess the coupled, time-integrated Hf–Nd–Pb isotope evolution of the most extreme HIMU location (Mangaia, French Polynesia). In comparison with global MORB and other OIB types, Mangaia samples define a unique trend in coupled Hf–Nd–Pb isotope co-variations (expressed in 207Pb/206Pb vs. ΔεHf). In a model employing subducted, dehydrated oceanic crust, mixing between present-day depleted MORB mantle (DMM) and small proportions (∼5%) of a HIMU mantle endmember can re-produce the Hf–Nd–Pb isotope systematics of global HIMU basalts (sensu stricto; i.e., without EM-1/EM-2/FOZO components). An age range of 3.5 to <2Ga is required for HIMU endmember(s) that mix with DMM to account for the observed present-day HIMU isotope compositions, suggesting a range of age distributions rather than a single component in the mantle. Our data suggest that mixing of HIMU mantle endmembers and DMM occurs in the mantle transition zone by entrainment in secondary plumes that rise at the edge of the Pacific Large Low Seismic Velocity Zone (LLSVP). These create either pure HIMU (sensu stricto) or HIMU affected by other enriched mantle endmembers (sensu lato). If correct, this requires isolation of parts of the mantle transition zone for >3 Gyr and implies that OIB chemistry can be used to test geodynamic models.

Constraints on the origin of the HIMU reservoir from He–Ne–Ar isotope systematics

We present a new set of He–Ne–Ar isotopic compositions for HIMU and EM1 lavas from the Cook–Austral Islands in the south Pacific. 3He/ 4He of the HIMU lavas are lower than MORB values, as previously demonstrated, but a coherent variation in 3He/ 4He with Pb isotopes indicates two-component mixing to form the lavas. One component is the HIMU reservoir with 3He/ 4He of 6 Ra or lower, and the other is the local lithosphere, which is commonly involved in EM1 lavas. Relative abundances of radiogenic and nucleogenic 4He, 21Ne and 40Ar show systematic variations. In 4He/ 40Ar*– 4He/ 21Ne* space, the HIMU lavas define a trend that is parallel to, but offset from the trend previously observed for other OIBs. Using 4He/ 21Ne* as a monitor of elemental fractionation of noble gasses, fractionation-corrected 4He/ 40Ar* is higher than the 4He/ 40Ar* production ratio in the mantle, reflecting a HIMU reservoir with a lower K/U (approximately 3000) than canonical mantle value (13000). Radiogenic 3He/ 4He and the low K/U are best explained by a model where the HIMU reservoir was formed by direct accumulation of, or deep mantle metasomatism with, ancient subducted oceanic crust modified by hydrothermal alteration and dehydration during its subduction. If the subducted oceanic crust with fractionated K/U forms a large isolated reservoir in the mantle, previous estimates of K/U and K concentrations for the bulk silicate Earth, that did not take this reservoir into consideration, will be too high. Moreover, the mass balance calculation indicates that the subducted oceanic crust may make a significant contribution to the U mantle budget, requiring there to be less in the primitive mantle. However, a contribution from the primitive mantle is not ruled out entirely, unless the subducted oceanic crust had a relatively high U concentration and low K/U over geologic time and was totally preserved in a convecting mantle.

Geochemical characteristics and origin of the HIMU reservoir: A possible mantle plume source in the lower mantle

Combined Pb-Sr-Nd-Hf-Os isotopes, together with major and trace element compositions, were determined from clinopyroxene and olivine phenocrysts, along with whole rocks, for ocean island basalts with high μ (μ = 238U/204Pb) (HIMU) and enriched mantle isotopic characteristics from Cook-Austral Islands. Clinopyroxene and olivine separates record reliable isotopic information of the sources because of minimized in situ radiogenic ingrowth and their lower susceptibility to crustal contamination. Coherent isotopic systematics in multi-isotope spaces defined by the HIMU samples are best explained by recent mixing of melts derived from the HIMU reservoir and the local shallow mantle. The isotopic compositions of the HIMU reservoir are constrained to be low ɛNd (≤+4), low ɛHf (≤+3), and moderately radiogenic 187Os/188Os (0.14–0.15) in association with radiogenic Pb isotopes (206Pb/204Pb ≥ 21.5). Since ancient oceanic crust would have had exceptionally radiogenic 187Os/188Os, moderately high 187Os/188Os precludes recycled oceanic crust as the only contributor to the HIMU reservoir. Instead, mantle metasomatized with partial melts from subducted oceanic crust is a likely candidate for the HIMU reservoir. Moreover, partial melting of oceanic crust in equilibrium with Mg perovskite would fractionate U/Pb, Sm/Nd, and Lu/Hf, which are in accordance with the time-integrated U/Pb, Sm/Nd, and Lu/Hf deduced from Pb, Nd, and Hf isotopic compositions of the HIMU reservoir, respectively, with a formation age of 2–3 Ga. We thus propose that the HIMU reservoir was formed by hybridization of a subducted oceanic crust-derived melt with the ambient mantle and then stored for several billion years in the lower mantle.

Lithium isotope systematics of lavas from the Cook–Austral Islands: Constraints on the origin of HIMU mantle

Radiogenic isotope variations in lavas from the Cook–Austral volcanic chain have delineated three distinct mantle sources: a HIMU component, a depleted component (DM), and an enriched component (EM). To better constrain the mantle sources for South Pacific hot spot volcanism, we determined lithium isotopic compositions of lavas from Raivavae, Rapa, Mangaia and Tubuai of the volcanic chain. The study includes whole rock and mineral analyses. In general, δ 7Li of most olivines resemble bulk rock composition whereas clinopyroxenes are variably lighter. This implies that clinopyroxene is more susceptible to diffusion-induced fractionation, in agreement with previous studies. Olivine δ 7Li values span a narrower range than whole rock values, and do not extend to the very heavy compositions previously reported in HIMU bulk lavas. This discrepancy likely reflects alteration of bulk lavas, and suggests that Li-isotope analyses of bulk lavas should be interpreted with caution. Olivines from HIMU lavas have heavy δ 7Li values (up to 6.2‰), and extend beyond the range reported for fresh MORB. Correlations between Li-isotopes and radiogenic isotopes suggest that the heavy δ 7Li values in HIMU olivines are a source characteristic and not the result of post-magmatic alteration. Although the Li-isotope composition of recycled, dehydrated oceanic crust is currently under debate, our results suggest that HIMU lavas derive from a source containing recycled dehydrated oceanic crust, and that the “heavy” Li-isotope signature of altered oceanic crust is partially preserved during passage through the subduction factory.

New constraints on the HIMU mantle from neon and helium isotopic compositions of basalts from the Cook–Austral Islands

High 4He/ 3He ratios of 100 000 to 160 000 found at HIMU ocean islands (“high μ,” where μ is the U/Pb ratio) are usually attributed to the presence of recycled oceanic crust in the HIMU mantle source. However, significantly higher 4He/ 3He ratios are expected in recycled crust after residence in the mantle for periods greater than 1 Ga. In order to better understand the helium isotopic signatures in HIMU basalts, we have measured helium and neon isotopic compositions in a suite of geochemically well-characterized basalts from the Cook–Austral Islands. We observe 4He/ 3He ratios ranging from 56 000 to 141 000, suggesting the involvement of mantle reservoirs both more and less radiogenic than the mantle source for mid-ocean ridge basalts (MORBs). In addition, we find that the neon isotopic compositions of HIMU lavas extend from the MORB range to compositions less nucleogenic than MORBs. The Cook-Austral HIMU He–Ne isotopic compositions, along with Sr, Nd, Pb, and Os isotopic compositions, indicate that in addition to recycled crust, a relatively undegassed mantle end-member (e.g., FOZO) is involved in the genesis of these basalts. The association of relatively undegassed mantle material with recycled crust provides an explanation for the close geographical association between HIMU lavas and EM (enriched mantle)-type lavas from this island chain: EM-type signatures represent a higher mixing proportion of relatively undegassed mantle material. Mixing between recycled material and relatively undegassed mantle material may be a natural result of entrainment processes and convective stirring in deep mantle.

One hundred million years of mantle geochemical history suggest the retiring of mantle plumes is premature

Linear chains of intraplate volcanoes and their geochemistry provide a record of mantle melting through geological time. The isotopic compositions of their lavas characterize their mantle sources, and their ages help backtrack these volcanoes to their original, eruptive source regions. Such data may shed light on a much-debated issue in Earth Sciences: the origin of intraplate volcanism and its underlying mantle and lithosphere dynamics. We show here that three major Western Pacific Seamount groups, ∼ 50–100 million years in age, display distinct Sr, Nd, Hf, and Pb isotopic signatures that can be traced back in time, both geographically and geochemically, to three separate, recently-active intraplate volcanoes in the South Pacific Cook–Austral Islands. Their unique 100 million year history, which shows a persistent geochemical fingerprint, suggests formation from large volumes of laterally fixed, long-lived source regions. Such longevity is unlikely to be attained in the relatively dynamic upper mantle. Therefore, these sources are likely anchored deep in the mantle, isolated from homogenization by mantle convection, and imply a primary origin from deep mantle plumes rather than resulting from lithosphere extension.

Lead isotopic compositions in olivine-hosted melt inclusions from HIMU basalts and possible link to sulfide components

Pb isotopic compositions of melt inclusions in olivine phenocrysts of lava samples from Mangaia and Rarotonga, Cook–Austral islands, have been determined by secondary ion mass spectrometry. The Pb isotopic compositions of melt inclusions from Rarotonga are consistent with those of bulk rock. On the other hand, Pb isotopic compositions of sulfide-free melt inclusions from Mangaia Island are widely distributed along the join between HIMU (i.e. highly radiogenic Pb) and less radiogenic components. The variation is much wider than that in bulk-rock samples that are invariably end-member HIMU values. In contrast, Pb isotopic compositions of sulfide and carbonate inclusions are restricted to HIMU end-member values. The variations in Pb isotopic ratios can be explained by mixing between the HIMU component and another component with less radiogenic Pb, and suggests that the HIMU component is enriched in sulfide and carbonate phases as well as Pb compared with the less radiogenic component. It seems paradoxical that a sulfide-rich component is associated with highly radiogenic Pb, because sulfides generally have extremely low U/Pb and Th/Pb ratios, which result in quite unradiogenic Pb. Subducted oceanic crust is not a likely source for such a sulfide-rich component because sulfides in subducted crust may not survive dehydration processes in subduction zones, although it can produce highly radiogenic Pb. The association of highly radiogenic Pb and sulfides implies that sulfides and radiogenic Pb in the HIMU source originate from distinct reservoirs. A possible origin of the sulfides is the Earth’s core, because the core is the largest sulfur budget in the Earth. The highly radiogenic Pb may originate from subducted oceanic crust which resides at the core–mantle boundary. Alternative source for radiogenic Pb is Ca-perovskite in the lower mantle, which is the main host of incompatible elements in the lower mantle and has high U/Pb and Th/Pb ratios. The core-derived sulfides would exchange Pb isotopes with subducted oceanic crust or Ca-perovskite during upwelling in the lower mantle, resulting in production of a sulfur-rich reservoir with highly radiogenic Pb.

Isotope and trace element variations in lavas from Raivavae and Rapa, Cook–Austral islands: constraints on the nature of HIMU- and EM-mantle and the origin of mid-plate volcanism in French Polynesia

We have examined the major and trace element and Sr–Nd–Hf–Pb–Os isotope variations in two suites of lavas from the islands of Raivavae and Rapa from the Cook–Austral chain to better constrain the nature and origin of the source(s) of these lavas. Combined with data from other Cook–Austral islands and seamounts, the Raivavae and Rapa samples reveal the presence of three distinct components involved in the genesis of these lavas, a HIMU component, a depleted (or DM) component, and an “enriched” (EM) component. Northern Cook–Australs lavas (those north of the Austral Fracture Zone (AFZ)) falling along the “Macdonald” pseudo-age progression reflect mixing of the HIMU and DM components, whereas lavas south of the AFZ reflect mixing of the DM and EM components.Strong depletions in potassium and other fluid soluble elements in HIMU lavas suggest that the HIMU source contains ancient recycled (dehydrated) oceanic crust. The EM lavas have systematically less radiogenic Os-isotopes than the HIMU lavas. Therefore, the enriched or EM component cannot be generated through the addition of pelagic or terrigenous sediment to the HIMU source. Instead, the EM component in the southern Cook–Australs may represent recycled subduction-modified sub-arc mantle. High 208Pb/204Pb values in the DM component relative to East Pacific Rise mid-ocean ridge basalts (MORB) preclude derivation of this component from assimilated lithospheric mantle. The Sr–Nd–Pb isotope characteristics of the DM component resemble Pacific alkalic seamount lavas, and may derive from easily fusible components (e.g., pyroxenite veins) dispersed in the Pacific upper mantle.The Raivavae and Rapa lavas define a Hf–Nd isotope correlation that is considerably steeper than the mantle Hf–Nd isotope array. The unusually low εHf values (for a given εNd value) characteristic of the HIMU source are consistent with the presence of recycled oceanic crust. However, the presence of low εHf values in the Rapa EM lavas reveals that anomalously low εHf values are not restricted to the HIMU source. Regardless of the ultimate origin of the Raivavae and Rapa lavas, the comparatively low εHf values in these lavas indicate that multiple low-εHf components are present in the mantle. The addition of these components may help resolve the apparent terrestrial Hf–Nd mass-imbalance with respect to estimates of the bulk earth Nd- and Hf-isotope composition.Islands and seamounts from the Cook–Austral chain do not follow a clear age progression as predicted by the standard plume model for the origin of ocean island basalts (OIB). Previous models for the origin of the Cook–Austral chain have appealed to the existence of multiple small plumes. However, these models are implausible because they do not explain how several deep-seated plumes become aligned with each other in the direction of plate motion. We propose that volcanism along the Cook–Austral chain (and perhaps in much of French Polynesia) is generated through self-perpetuating melting anomalies that preferentially sample enriched components in a marble-cake mantle. These melting anomalies preferentially form or strengthen where lithospheric boundary conditions are favorable, such as at pre-existing transform faults or where the lithosphere has been thinned by a previous period of volcanism.

Noble gas study of HIMU and EM ocean island basalts in the Polynesian region

Noble gas isotopes of HIMU and EM ocean island basalts from the Cook-Austral and Society Islands were investigated to constrain their origins. Separated olivine and clinopyroxene (cpx) phenocrysts were used for noble gas analyses. Since samples are relatively old, obtained from the oceanic area and showing chemical zoning in cpx phenocrysts, several tests on sample preparation and gas extraction methods were performed. First, by comparing heating and crushing methods, it has been confirmed that the crushing method is suitable to obtain inherent magmatic noble gases without radiogenic and cosmogenic components which were yielded after eruption, especially for He and Ne analyses. Second, noble gas compositions in the core and the rim of cpx phenocrysts were measured to evaluate the zoning effect on noble gases. The result has been that noble gas concentrations and He and Ne isotope ratios are different between them. The enrichment of noble gases in the rim compared to the core is probably due to fractional crystallization. Difference of He and Ne isotope ratios is explained by cosmogenic effect, and isotope ratios of the trapped component seem to be similar between the rim and the core. Third, leaching test reveals no systematic differences in noble gas compositions between leached and unleached samples. 3He/ 4He ratios of HIMU samples in the Cook-Austral Islands are uniform irrespective of phenocryst type (olivine and cpx) and age of samples (10–18 Ma), and lower (average 6.8 R A) than those of the Pacific MORB. On the other hand, 3He/ 4He of EM samples in the Cook-Austral Islands are similar to MORB values. EM samples in the Society Islands show rather higher 3He/ 4He than MORB. Ne, Kr and Xe isotope ratios are almost atmospheric within analytical uncertainties. 40Ar/ 36Ar are not so high as those of MORB. Anomalous noble gas abundance pattern such as He and Ne depletion and Kr and Xe enrichment relative to atmospheric abundances was observed. Furthermore, Ne/Ar and Kr/Ar show correlation with some trace elemental ratios like La/Yb. Lower 3He/ 4He of HIMU than MORB values requires relatively high time-integrated (U + Th)/ 3He for the HIMU source, which suggests that the HIMU source was produced from recycled materials which had been once located near the Earth’s surface. Moreover, extreme noble gas abundance pattern and strong correlation of Ne/Ar and Kr/Ar with La/Yb indicate that the HIMU endmember is highly depleted in light noble gases and enriched in heavy noble gases. Such feature is not common to mantle materials and is rather similar to the noble gas abundance patterns of the old oceanic crust and sediment, which supports the model that the HIMU source originates from subducted oceanic crust and/or sediment. If the HIMU source corresponds to the oceanic crust which subducted at 1–2 Ga as suggested by Pb isotope studies, however, the characteristic 3He/ 4He of HIMU (6.8 R A) would be too high because radiogenic 4He produced by U and Th decay should dramatically decrease 3He/ 4He. To overcome this problem, the He open system model is introduced which includes the effects of 4He production and diffusion between the HIMU source material and the surrounding mantle. This model favors that the HIMU source resides in the upper mantle, rather than in the lower mantle. Furthermore, this model predicts the thickness of the HIMU source to be in the order of 1 km. In contrast to low and uniform 3He/ 4He character of HIMU, 3He/ 4He of EM are rather variable. Entrainment of upper mantle material and/or a less-degassed component are required to explain the observed 3He/ 4He of EM in the Polynesian area. Participation of the less-degassed component would be related to the “superplume” below the Polynesian region.

Geomorphology and Geodynamics of the Cook-Austral Island-Seamount Chain in the South Pacific Ocean: Implications for Hotspots and Plumes

Among Pacific hotspot tracks, the Cook-Austral island-seamount chain is distinctly anomalous in geodynamic behavior, exhibiting repetitive episodes of volcanism at multiple sites, uplift of selected islands long after initial immersion by subsidence, and multiple alignments of volcanic edifices. Cook-Austral islands include a variety of disparate geomorphic types: volcanic islands without reefs, with fringing reefs, and with barrier reefs enclosing shallow lagoons, low-lying atolls; and makatea islands composed of volcanic cores surrounded by annular limestone tablelands. Neogene hotspot volcanism along the chain, built across Cretaceous-Paleogene seafloor, has occurred above three different mantle source regions, now located near its southeastern end, center, and northwestern end. Each hotspot has given rise to a separate age trend of volcanism along the same broad hotspot track, as the Pacific plate drifted over them, and each was initiated by eruption through the same region of lithosphere now located near the northwestern end of the compound chain. Age-distance trends of hotspot volcanism along the nearby Society and Marquesas chains depart less from expectation, but both the orientation and age trend of the Marquesas hotspot track also are anomalous. Lithospheric properties have seemingly influenced loci of South Pacific hotspot magmatism more than plume theory implies. Cook-Austral-Society-Marquesas volcanic suites are distinctly more alkalic than typical Hawaiian tholeiites, and display marked isotopic heterogeneities that reflect derivation from multiple mantle sources, even for individual volcanoes. As the volcanic assemblages all were erupted from the SOPITA-Superswell region of anomalously hot mantle including diverse recycled components, hotspots that built the several island chains may reflect bleeding of magma through lithospheric flaws from a buoyant subjacent pillow of atypical mantle, poised to melt readily from varied stimuli. Monotonic island subsidence is expected following cessation of hotspot volcanism, but episodes of post-immersion uplift have produced six makatea islands along the Cook-Austral chain. Makateas are tablelands of Neogene limestone, representing uplifted fringing reefs, and form annular girdles surrounding volcanic island cores. Despite limited erosional degradation, makatea surfaces provide a measure of net post-immersion island uplift. Paleoshoreline indicators of changing relative sea levels on Cook-Austral islands reflect the combined influence of seafloor thermotectonic subsidence, thermal rejuvenation associated with renewed hotspot activity, flexural bulges surrounding the isostatic loads of subsided volcanic edifices built out of sequence, the last-interglacial eustatic highstand in global sea level, and a mid-Holocene hydro-isostatic highstand in regional sea level. Key paleoshoreline indicators include emergent reef flats and microatolls recording former low-tide levels, paleobeachrock recording paleotidal range, and shoreline notches formed by solution and bioerosion at former high-tide levels. The elevations of paleonotches with respect to modern shoreline notches incised into the same seacliffs provide the most precise measures of shoreline emergence. Steady but slow thermotectonic subsidence is undetectable for mid-Holocene paleoshoreline features, but has measurably altered the elevations of last-interglacial features. In the Cook Islands, paleoshoreline features on Rarotonga and Aitutaki are compatible with theoretical expectations for thermotectonic subsidence since the last interglacial, and with a post-mid-Holocene drawdown in hydro-isostatic sea level intermediate between amounts calculated and observed regionally. Four makatea islands, apparently uplifted along the flexural bulge surrounding Rarotonga, display slightly greater post-mid-Holocene emergence, unexpected amounts of post-last-interglacial emergence, and net makatea uplifts similar but not identical to calculated inferences. Excess elevations of last-interglacial terraces on the makatea islands may conceivably reflect continued magmatic inflation of the Rarotonga volcanic edifice long after cessation of surface volcanism. In the Austral Islands, makatea uplift reflects either thermal rejuvenation of lithosphere by hotspot volcanism or concurrent passage of the islands over an asthenospheric bump, or both. As for the Cook makateas, post-last-interglacial emergence of Rurutu implies persistence of uplift beyond the expected time interval, whereas post-mid-Holocene emergence of Tubuai can be ascribed entirely, within inherent uncertainty limits, to hydro-isostasy alone. The distribution of subsided and uplifted islands along the Cook-Austral chain reflects multiple hotspot activity, probably related to multiple mantle diapirs of local character rather than to deepseated plumes. Rapid Pacific plate motion can generate elongate hotspot tracks from transient hotspot activity unrelated to columnar advective plumes.

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

Isotopic variation in ocean island basalts (OIB) is commonly attributed to mixing between four distinctive isotopic components in the mantle. One of these, HIMU, is the only contributor of lead with 206Pb/ 204Pb > 19.2 to the system and is therefore invoked as a mixing component in very many OIB. Extreme HIMU mantle, as represented by OIB with 206Pb/ 204Pb > 20.5, must have been generated by elevated 238U/ 204Pb ratios ( μ values) over time periods of 1.5–2 Ga. If high μ also developed in mantle much more recently than 1.5 Ga, then its isotopic results today could broadly simulate mixtures between extreme HIMU and the other isotopic components. 207Pb/ 204Pb ratios lying systematically below the Northern Hemisphere Reference Line and the MORB data field are quite common in ocean island basalts. These are described as negative-Δ7/4Pb OIB, and have fairly homogeneous compositions for other isotopic systems, and similar trace element characteristics to the extreme HIMU OIB from St. Helena and the Cook-Austral Islands, including low K/Nb and Ba/Nb ratios, but high Ce/Pb and μ values. These compositions can be generated during the Phanerozoic and Proterozoic from mantle with chemical characteristics similar to those of the mantle source of extreme HIMU lavas. A μ value of 20 ± 2 is estimated for the HIMU source and used to model the source of negative-Δ7/4Pb OIB. These OIB are most unlikely to be produced by: 1. (1) Mixing between MORB and HIMU mantle, both because this cannot generate negative Δ7/4Pb, and because they have elevated 87Sr/ 86Sr and Δ8/4Pb relative to MORB-HIMU mixtures, but high Ce/Pb and low Ba/Nb that do not permit admixture of any enriched component. 2. (2) Ageing of the observed high μ values in the oceanic lithosphere, because there is no relationship between the age of the lithosphere, the magmatic μ value and the extent of negative Δ7/4Pb. Further, A is shown to be substantially increased during the melting process. Ce is shown to be more incompatible than Pb during melting in most negative-Δ7/4Pb 0113 suites, and in some MORB and extreme HIMU suites. Higher Ce/Pb ratios in HIMU and negative-Δ7/4Pb OIB than in MORB may just reflect the melting process. There is thus no compelling evidence that the increased μ of HIMU mantle is due to depletion in Pb relative to MORB mantle. Ratios involving K and Nb appear to be frequently fractionated by a potassic phase such as phlogopite in the source. Higher Ce/Pb and Nd/Pb in negative-Δ7/4Pb OIB than in extreme HIMU 0113 cannot be generated by melting, and along with the elevated 87Sr/ 86Sr and Δ8/4Pb, probably reflect secular changes in the process that generates HIMU mantle. Most negative-Δ7/4Pb OIB could be derived from fairly pure young HIMU mantle. Large volumes of continental and oceanic basalts with similar isotopic and chemical signatures, apart from zero or slightly positive Δ7/4Pb, merely require minor incorporation of enriched components (crust or mantle) into melts derived from young HIMU mantle. Recognition of this as a major mantle reservoir would eliminate the need for MORB-source mantle contributions to most OIB.

An analysis of the altimetric geoid in various wavebands in the Central Pacific Ocean: constraints on the origin of intraplate features

SEASAT profiles were filtered in three different wavebands and interpolated to elaborate maps of the very short (25–110 km) and short (80–320 km and 280–820 km) wavelengths of the geoid anomalies for the South Pacific. For the very short wavelengths, we mapped the geoid roughness i.e. the variations of amplitude of the anomalies. Two lines following present-day plate motions and separated by 800–1000 km, were identified. The first is formed by the Tuamotu Archipelago, the Pitcairn-Crough alignment and the Easter-Sala y Gomez chain. The second comprises the Cook-Austral Islands, a seamount cluster mapped close to the East Pacific Rise, south of 30°S and the Chile Rise. Alignments with a N150°E direction and corresponding to traces of old thermal activity, were mapped on lithosphere older than 40 m.y. These features are also spaced of 800–1000 km. The map for the 80–320 km waveband shows a pattern of 200 km wavelength undulations. Several trends are superimposed, but the N95°E direction is dominant in the area. We associate these geoid undulations to lithospheric phenomena, such as fracturations and/or small magmatic ridges. The 280–820 km waveband reveals linear anomalies, roughly following the direction of present-day absolute plate motions. The linear anomalies have a 400–500 km wavelength and are nearly continuous between the Pacific and the Nazca plates. The topography is located on the highs, but the anomalies generally precede the active ends of the volcanic chains. This result favours the hot-line hypothesis previously suggested by several authors. We associate these bands to thermal heterogeneities probably due to a convective regime. Since the main topographic features shall be emplaced over the ascending limbs of two neighbouring cells, the 800–1000 km spacing of the topography imposes a minimum size of 400–500 km for the cells. So, the convective cells cannot be limited to a thin layer beneath the lithosphere, but will affect part of the upper asthenosphere. The existence of hot lines aligned parallel to plate motions, can explain the intermittent magmatic activity observed for the volcanic lineaments of the South Pacific. Volcanic activity can be the result of the interaction between the hot lines and lithospheric zones of weakness and/or of instabilities in the convective cells. The regular spacing observed for topography formed before and after 40 m.y. suggests that the size of the cells is rather constant through time.