Mafic Plinian volcanism and ignimbrite emplacement at Tofua volcano, Tonga

Tofua Island is the largest emergent mafic volcano within the Tofua arc, Tonga, southwest Pacific. The volcano is dominated by a distinctive caldera averaging 4 km in diameter, containing a freshwater lake in the south and east. The latest paroxysmal (VEI 5-6) explosive volcanism includes two phases of activity, each emplacing a high-grade ignimbrite. The products are basaltic andesites with between 52 wt.% and 57 wt.% SiO(2). The first and largest eruption caused the inward collapse of a stratovolcano and produced the 'Tofua' ignimbrite and a sub-circular caldera located slightly northwest of the island's centre. This ignimbrite was deposited in a radial fashion over the entire island, with associated Plinian fall deposits up to 0.5 m thick on islands > 40 km away. Common sub-rounded and frequently cauliform scoria bombs throughout the ignimbrite attest to a small degree of marginal magma-water interaction. The common intense welding of the coarse-grained eruptive products, however, suggests that the majority of the erupted magma was hot, water-undersaturated and supplied at high rates with moderately low fragmentation efficiency and low levels of interaction with external water. We propose that the development of a water-saturated dacite body at shallow (<6 km) depth resulted in failure of the chamber roof to cause sudden evacuation of material, producing a Plinian eruption column. Following a brief period of quiescence, largescale faulting in the southeast of the island produced a second explosive phase believed to result from recharge of a chemically distinct magma depleted in incompatible elements. This similar, but smaller eruption, emplaced the 'Hokula' Ignimbrite sheet in the northeast of the island. A maximum total volume of 8 km(3) of juvenile material was erupted by these events. The main eruption column is estimated to have reached a height of similar to 12 km, and to have produced a major atmospheric injection of gas, and tephra recorded in the widespread series of fall deposits found on coral islands 40-80 km to the east (in the direction of regional upper-tropospheric winds). Radiocarbon dating of charcoal below the Tofua ignimbrite and organic material below the related fall units imply this eruption sequence occurred post 1,000 years BP. We estimate an eruption magnitude of 2.24x10(13) kg, sulphur release of 12 Tg and tentatively assign this eruption to the AD 1030 volcanic sulphate spike recorded in Antarctic ice sheet records.

Magma Evolution in the Primitive, Intra-oceanic Tonga Arc: Petrogenesis of Basaltic Andesites at Tofua Volcano

Tofua volcano is situated midway along the Tonga oceanic arc and has undergone two phases of ignimbrite-forming activity. The eruptive products are almost entirely basaltic andesites (52 center dot 5-57 wt % SiO2) with the exception of a volumetrically minor pre-caldera dacite. The suite displays a strong tholeiitic trend with K2O <1 wt %. Phenocryst assemblages typically comprise plagioclase + clinopyroxene +/- orthopyroxene with microlites of Ti-magnetite. Olivine (Fo(83-88)) is rare and believed to be dominantly antecrystic. An increase in the extent and frequency of reverse zoning in phenocrysts, sieve-textured plagioclase and the occurrence of antecrystic phases in post-caldera lavas record a shift to dynamic conditions, allowing the interaction of magma batches that were previously distinct. Pyroxene thermobarometry suggests crystallization at 950-1200 degrees C and 0 center dot 8-1 center dot 8 kbar. Volatile measurements of glassy melt inclusions indicate a maximum H2O content of 4 center dot 16 wt % H2O, and CO2-H2O saturation curves indicate that crystallization occurred at two levels, at depths of 4-5 center dot 5 km and 1 center dot 5-2 center dot 5 km. Major and trace element models suggest that the compositions of the majority of the samples represent a differentiation trend whereby the dacite was produced by 65% fractional crystallization of the most primitive basaltic andesite. Trace element models suggest that the sub-arc mantle source is the residuum of depleted Indian mid-ocean ridge basalt mantle (IDMM-1% melt), whereas radiogenic isotope data imply addition of 0 center dot 2% average Tongan sediment melt and a fluid component derived from the subducted altered Pacific oceanic crust. A horizontal array on the U-Th equiline diagram and Ra excesses of up to 500% suggest fluid addition to the mantle wedge within the last few thousand years. Time-integrated (Ra-226/Th-230) vs Sr/Th and Ba/Th fractionation models imply differentiation timescales of up to 4500 years for the dacitic magma compositions at Tofua.

Pb-, Sr-, and Nd-isotopic composition of basaltic rocks from the Lau Basin

New Pb, Sr, and Nd isotope data are presented for 64 samples from the six backarc sites drilled during Leg 135. Systematic changes in Pb and Sr compositions illustrate significant isotopic variations between and within sites as well as provide two key pieces of information. First, a recent influx of asthenosphere with Indian Ocean mantle affinities has occurred and has successfully displaced older "Pacific" asthenosphere from the mantle underlying the backarc region. Second, clear evidence exists for mixing between these two asthenospheric end-members and at least one "arc-like" component. The latter was not the same as most material currently erupting in the Tofua Arc, but it must have had a more radiogenic Pb-isotope signature, perhaps similar to rocks analyzed from the islands of Tafahi, and Niuatoputapu. A comparison between the isotopic variations and the tectonic setting of the drill sites reveals consistent and important information regarding the mantle dynamics beneath the evolving backarc basin. We propose a model in which the source of upwelling magmas changes from Pacific to Indian Ocean asthenosphere with the propagation of seafloor spreading, a model with important implications for the rate of mantle influx into this region. Although the chemistries of backarc magmas have been profoundly influenced by this process, an additional consequence is the advection of Indian Ocean asthenosphere into the sub-arc mantle source. The isotopic compositions of arc rocks from the vicinity have been reevaluated on the basis of the proposed mantle advection model. We suggest that the slab-derived flux of trace elements into the arc wedge has remained relatively uniform with time (i.e., ~40 Ma), so that the change in arc chemistry results from mantle source substitution, rather than from differences in the composition of the downgoing plate.

Geochemistry of ODP Hole 135-839B lavas

Four petrographic lava types occur, ranging from aphyric to moderately phyric clinopyroxene-olivine tholeiitic basalts (Unit 1); olivine-clinopyroxene picritic basalts, sparsely to strongly olivine-phyric (Unit 3-type); olivine-clinopyroxene basalts (clinopyroxene dominant) (Unit 4); and moderately to strongly phyric two-pyroxene-plagioclase basaltic andesites (Unit 9-type). The olivine phyric lavas contain forsteritic olivines (extending to Fo92), and very magnesian Cr-rich spinels similar to those occurring in boninitic lavas. The basaltic andesites are mineralogically and petrographically indistinguishable from the modern Tofua Arc basaltic andesites, one notable feature being the highly calcic cores in plagioclase phenocrysts (up to An95). The forsteritic olivines, the Cr-spinels, and the calcic plagioclases are unlikely to have been precipitated in the lava compositions in which they occur, and are thought to have been incorporated from highly primitive melts by way of mixing processes (as advocated by Allan, this volume). Notwithstanding the evidence for mixing, the major element chemistries of the Unit 1- and Unit 9-type lavas are shown to be consistent with the derivation of the Unit 9-type basaltic andesites by means of fractional crystallization, through magmas of similar chemistry to Unit 1. Some trace element discrepancies in the modeling, and the relative volcanic stratigraphy of Site 839, however, preclude a direct liquid line of descent between the actual recovered units. Trace element data as well as TiO2 and Na2O data clearly illustrate the arc-like affinities of the magmas, with strong highfield-strength element depletion and large-ion-lithophile element enrichment. The abundance patterns are very close to those of the Tofua and Kermadec arc magmas, and also Valu Fa. Pb-, Sr-, and Nd-isotopic compositions indicate closest affinities with a "Pacific" MORB source, apparently characteristic of the western, older part of the Lau Basin. A subduction-related isotopic contribution is, however, inferred. The sources of the Site 839 magmas are thus inferred to be similar to, but less depleted geochemically, than those of the modern Tofua Arc magmas. The Site 839 sequence is interpreted as an older remnant of a volcanic construct of the "proto-Tofua arc", originally developed adjacent to the Tonga Ridge. Opening of the eastern Lau Basin, because of southward migrating propagators, has split and isolated the sequence, leaving it stranded within the modern Lau Basin.

Geochemistry of ODP Site 135-834 basalts

New major, trace element, and isotope data (Pb, Sr, and Nd) reveal an impressive compositional variation in the basalts recovered from Site 834. Major element compositions span almost the entire range observed in basalts from the modern axial systems of the Lau Basin, and variations are consistent with low-pressure fractionation of a mid-ocean-ridge-basalt (MORB)-like parent, in which plagioclase crystallization has been somewhat suppressed. Trace element compositions deviate from MORB in all but one unit (Unit 7) and show enrichments in large-ion-lithophile elements (LILEs) relative to high-field-strength elements (HFSEs) more typically associated with island-arc magmas. The Pb-isotope ratios define linear trends that extend from the field of Pacific MORB to highly radiogenic values similar to those observed in rocks from the northernmost islands of the Tofua Arc. The Sr-isotope compositions also show significant variation, and these too project from radiogenic values back into the field for Pacific MORB. The variations in key trace element and isotopic features are consistent with magma mixing between two relatively mafic melts: one represented by Pacific MORB, and the other by a magma similar to those erupted on 'Eua when it was part of the original Tongan arc, or perhaps members of the Lau Volcanic Group (LVG). Based on our model, the most radiogenic compositions (Units 2 and 8) represent approximately 50:50 mixtures of these MORB and arc end-members. Magma mixing requires that both components are simultaneously available, and implies that melts have not shown a compositional progression from arc-like to MORB-like with extension at this locality. Rather, it is apparent that essentially pristine MORB can erupt as one of the earliest products of backarc initiation. Indeed, repetition of isotopic and trace element signatures with depth suggests that eruptions have been triggered by periodic injections of fresh MORB melts into the source regions of these magmas. The slow and almost amagmatic extension of the original arc complex envisaged to explain the observed chemistry is also consistent with the horst-and-graben topography of the western side of the Lau Basin. Given the similarities between basalts erupted at the modern Lau Basin spreading centers and MORB from the Indian Ocean, the overwhelming evidence for involvement of mantle similar to Pacific MORB in the petrogenesis of basalts from Site 834 is a new and important observation. It indicates that the original arc was underlain by asthenospheric material derived from the Pacific mantle convection cell, and that this has somehow been replaced by Indian Ocean MORB during the last ~5.5 Ma.

Isotopic composition of rocks from the Lau/Tonga back-arc basin, data of DSDP Leg 21 and ODP Legs 91 an 135

New trace element, Sr-, Nd-, Pb- and Hf isotope data provide insights into the evolution of the Tonga-Lau Basin subduction system. The involvement of two separate mantle domains, namely Pacific MORB mantle in the pre-rift and early stages of back-arc basin formation, and Indian MORB mantle in the later stages, is confirmed by these results. Contrary to models proposed in recent studies on the basis of Pb isotope and other compositional data, this change in mantle wedge character best explains the shift in the isotopic composition, particularly 143Nd/144Nd ratios, of modern Tofua Arc magmas relative to all other arc products from this region. Nevertheless, significant changes in the slab-derived flux during the evolution of the arc system are also required to explain second order variations in magma chemistry. In this region, the slab-derived flux is dominated by fluid; however, these fluids carry Pb with sediment-influenced isotopic signatures, indicating that their source is not restricted to the subducting altered mafic oceanic crust. This has been the case from the earliest magmatic activity in the arc (Eocene) until the present time, with the exception of two periods of magmatic activity recorded in samples from the Lau Islands. Both the Lau Volcanic Group, and Korobasaga Volcanic Group lavas preserve trace element and isotope evidence for a contribution from subducted sediment that was not transported as a fluid, but possibly in the form of a melt. This component shares similarities with that influencing the chemistry of the northern Tofua Arc magmas, suggesting some caution may be required in the adoption of constraints for the latter dependent upon the involvement of sediments from the Louisville Ridge. A key outcome of this study is to demonstrate that the models proposed to explain subduction zone magmatism cannot afford to ignore the small but important contributions made by the mantle wedge to the incompatible trace element inventory of arc magmas.

Sedimentary and subsidence history of ODP Sites 135-840 and 135-841

Sedimentary sections recovered from the Tonga platform and forearc during Ocean Drilling Program Leg 135 provide a record of the sedimentary evolution of the active margin of the Indo-Australian Plate from late Eocene time to the Present. Facies analyses of the sediments, coupled with interpretations of downhole Formation MicroScanner logs, allow the complete sedimentary and subsidence history of each site to be reconstructed. After taking into account the water depths in which the sediments were deposited and their subsequent compaction, the forearc region of the Tofua Arc (Site 841) can be seen to have experienced an initial period of tectonic subsidence dating from 35.5 Ma. Subsidence has probably been gradual since that time, with possible phases of accelerated subsidence, starting at 16.2 and 10.0 Ma. The Tonga Platform (Site 840) records only the last 7.0 Ma of arc evolution. However, the increased accuracy of paleowater depth determinations possible with shallow-water platform sediments allows the resolution of a distinct increase in subsidence rates at 5.30 Ma. Thus, sedimentology and subsidence analyses show the existence of at least two, and possibly four, separate subsidence events in the forearc region. Subsidence dating from 35.5 Ma is linked to rifting of the South Fiji Basin. Any subsidence dating from 16.2 Ma at Site 841 does not correlate with another known tectonic event and is perhaps linked to localized extensional faulting related to slab roll back during steady-state subduction. Subsidence from 10.0 Ma coincides with the breakup of the early Tertiary Vitiaz Arc because of the subduction polarity reversal in the New Hebrides and the subsequent readjustment of the plate boundary geometry. More recently, rapid subsidence and deposition of a upward-fining cycle from 5.30 Ma to the Present at Site 840 is thought to relate to rifting of the Lau Basin. Sedimentation is principally controlled by tectonic activity, with variations in eustatic sea level playing a significant, but subordinate role. Subduction of the Louisville Seamount Chain seems to have disrupted the forearc region locally, although it had only a modest effect on the subsidence history and sedimentation of the Tonga Platform as a whole.

Grain size composition and geochemistry of volcaniclastic sediments in the Lau Basin

Early Pliocene to Pleistocene volcaniclastic sediments recovered during Ocean Drilling Program Leg 135 from Sites 834 to 839 in the Lau Basin show a wide range of chemical and mineralogical compositions extending the spectrum previously known from the Lau Basin, Lau Ridge and Tofua Arc. The following major types of volcaniclastics have been distinguished: (1) primary fallout ashes originating from eruptions on land, (2) epiclastic deposits that resulted from subaerial and submarine eruptions, (3) subaqueous fallout and pyroclastic flow deposits resulting from explosive submarine eruptions, and (4) hyaloclastites resulting from mechanical fragmentation and spalling of chilled margins of submarine pillow tubes and sheet-lava flows. Vitric shards are mostly basaltic andesitic to rhyolitic and broadly follow two major trends in terms of K2O enrichment: a low-K series (LKS) with about 1 wt% K2O at 70 wt% SiO2, and a very low-K series (VLKS) with only about 0.5 wt% K2O at 70 wt% SiO2. Sites 834 and 835 on "old" backarc basin crust, >4.2 and 3.4 m.y. old, comprise LKS rhyolites >3.3 m.y. old. Calc-alkaline basaltic turbidites originating from the Lau Ridge flowed in at 3.3 Ma. In the period from 3.3 to 2.4 Ma basaltic andesitic to rhyolitic, fine-grained LKS and VLKS volcaniclastics were deposited by turbidity currents and subaerial fallout. Three thin, discrete fallout layers (2.4-3.2 m.y. old) with high-K calc-alkaline compositions probably erupted in New Zealand. Volcaniclastics from Site 836, all <0.6 m.y. old, make up 24% of the sediments and comprise local basaltic andesitic to andesitic hyaloclastites with low Ba/Zr ratios of 0.9 to 1.4 and polymict andesitic sediments with Ba/Zr ratios of up to 5.5, containing clasts altered to lower greenschist facies. In Sites 837-839, drilled on young crust (1.8-2.1 m.y. old), volcaniclastics make up 45%-64% of the total sediment. Glass compositions are often bimodal with a mafic and a rhyolitic population. Large-volume rhyolitic, silt- to lapilli-sized volcaniclastics are interpreted as pyroclastic flows from explosive eruptions on a seamount 25-50 km away from the sites. Ba/Zr ratios are 2 to 4, partially overlapping with some Lau Basin basement lavas that show an "arc" signature, and they can reach values >5 in thin volcaniclastic layers <0.6 m.y. old.

(Table 1) Sedimentary thicknesses, lithologies, ages, and interpreted paleowater depths for ODP Sites 135-840 and 135-841

The sedimentary succession drilled at Sites 840 and 841 on the Tonga forearc allows the sedimentary evolution of the active margin to be reconstructed since shortly after the initiation of subduction during the mid Eocene. Sedimentation has been dominated by submarine fan deposits, principally volcaniclastic turbidites and mass-flows derived from the volcanic arc. Volcaniclastic sedimentation occurred against a background of pelagic nannofossil sedimentation. A number of upward-fining cycles are recognized and are correlated to regional tectonic events, such as the rifting of the Lau Basin at 5.6 Ma. Episodes of sedimentation dating from 16.0 and 10.0 Ma also correlate well with major falls in eustatic sea level and may be at least partially caused by the resulting enhanced erosion of the arc edifice. The early stages of rifting of the Lau Basin are marked by the formation of a brief hiatus at Site 840 (Horizon A), probably a result of the uplift of the Tonga Platform. Controversy exists as to the degree and timing of the uplift of Site 840 before Lau Basin rifting, with estimates ranging from 2500 to 300 m. Structural information favors a lower value. Breakup of the Tonga Arc during rifting resulted in deposition of dacite-dominated, volcaniclastic mass flows, probably reflecting a maximum in arc volcanism at this time. A pelagic interval at Site 840 suggests that no volcanic arc was present adjacent to the Tonga Platform from 5.0 to 3.0 Ma. This represents the time between separation of the Lau Ridge from the Tonga Platform and the start of activity on the Tofua Arc at 3.0 Ma. The sedimentary successions at both sites provide a record of the arc volcanism despite the reworked nature of the deposits. Probe analyses of volcanic glass grains from Site 840 indicate a consistent low-K tholeiite chemistry from 7.0 Ma to the present, possibly reflecting sediment sourcing from a single volcanic center over long periods of time. Trace and rare-earth-element (REE) analyses of basaltic glass grains indicate that thinning of the arc lithosphere had begun by 7.0 Ma and was the principle cause of a progressive depletion of the high-field-strength (HFSE), REE, and large-ion-lithophile (LILE) elements within the arc magmas before rifting. Magmatic underplating of the Tofua Arc has reversed this trend since that time. Increasing fluid flux from the subducting slab since basin rifting has caused a progressive enrichment in LILEs. Subduction erosion of the underside of the forearc lithosphere has caused continuous subsidence and tilting toward the trench since 37.0 Ma. Enhanced subsidence occurred during rifting of the South Fiji and Lau basins. Collision of the Louisville Ridge with the trench has caused no change in the nature of the sedimentation, but it may have been responsible for up to 300 m of uplift at Site 840.

Pumice rafting and faunal dispersion during 2001-2002 in the Southwest Pacific: record of a dacitic submarine explosive eruption from Tonga

A new influx of sea-rafted pumice reached the eastern coast of Australia in October 2002, approximately 1 year after a felsic, shallow-marine explosive eruption at a previously unknown volcano (0403-091) along the Tofua volcanic arc (Tonga). The eruption produced floating pumice rafts that first became stranded in Fiji in November 2001, approximately I month after the eruption. Strandings of sea-rafted pumice along shorelines have been the only record of products from this submarine explosive eruption at the remote, submerged volcano. Computed drift trajectories of the sea-rafted pumice using numerical models of southwest Pacific surface wind fields and ocean currents indicate two cyclonic systems disturbed the drift of pumice to eastern Australia, as well as the importance of the combined wave and direct wind effect on pumice trajectory. Pumice became stranded along at least two-thirds (>2000 km) of the coastline of eastern Australia, being deposited on beaches during a sustained period of fresh onshore winds. Typical amounts of pumice initially stranded on beaches were 500-4000 individual clasts per in, and a minimum volume estimate of pumice that arrived to eastern Australia is 1.25 x 10(5) m(3). Pumice was beached below maximum tidal/storm surge levels and was quickly reworked back into the ocean, such that the concentration of beached pumice rapidly dissipated within weeks of the initial stranding, and little record of this stranding event now exists. Most stranded pumice clasts ranged in size from 2 to 5 cm in diameter; the largest measured clasts were 10 cm in Australia and 20 cm in Fiji. The pumice has a low phenocryst content (3500 km) and period of pumice floatation (greater than or equal to1 year), confirm the importance of sea-rafted pumice as a long-distance dispersal mechanism for marine organisms including marine pests and harmful invasive species. Billions of individual rafting pumice clasts can be generated in a single small-volume eruption, such as observed here, and the geological implications for the transport of sessile taxa over large distances are significant. An avenue for future research is to examine whether speciation events and volcanicity are linked; the periodic development of globalism for some taxa (e.g., corals, gastropods, bryozoa) may correlate in time and/or space with voluminous silicic igneous events capable of producing >10(6) km(3) of silicic pumice-rich pyroclastic material and emplaced into ocean basins. (C) 2004 Elsevier B.V. All rights reserved.

Geochemistry of ODP Leg 135 igneous rock samples

Ocean Drilling Program Leg 135 provided igneous rock cores from six sites drilled on a transect across the Lau Basin between the Lau Ridge remnant arc and the modem spreading ridges of the Central and Eastern Lau Spreading Centers. The drill cores sampled crust from the earliest stage of backarc extension (latest Miocene time, about 6 Ma), and younger crust (late Pliocene, about 3.8-2 Ma, and middle Pleistocene, about 0.64-0.8 Ma). Nearly all of the igneous samples are from tholeiitic basalt flows; many of them are interbedded with arc-composition volcaniclastic sediments. Rock compositions range from olivine-plagioclase-clinopyroxene basalt, with up to 8% MgO, to oceanic andesites with less than 3.2% MgO and silica contents as high as 56%. The oldest rocks recovered are close in composition to rocks formed at the modern Central and Eastern Lau Spreading Centers and have MORB-like characteristics. Generation of the oldest units was coeval with arc-tholeiitic volcanism on the Lau Ridge less than 100 km to the west. The arc and backarc melts came from different mantle sources. At three sites near the center of the basin, the crust is arc-tholeiitic basalt, two-pyroxene basaltic-andesite, and two-pyroxene andesite. These rocks have many similarities to modem Tofua Arc lavas yet they were drilled within 70 km of the MORB-like Eastern Lau Spreading Center. Estimates of the minimum age for these arc-like rocks indicate that they are late Pliocene (about 2 Ma). These ages overlap the age of the nearby Eastern Lau Spreading Center. The heterogeneous crust of the Lau Basin carries many of the signatures of supra-subduction zone (SSZ) melts but also has a distinct MORB-like component. Mixing between SSZ and MORB mantle sources may explain the variations and the spatial distribution of magma types.