Beschreibung: The Nifonea submarine volcano rises 1000 m above the seafloor of the Vate Trough back-arc basin behind the New Hebrides island arc. This large volcanic edifice has a caldera of ∼8 km diameter and is connected to two ∼20 km long volcanic rift zones in the back-arc basin. We present new chemical and isotope data for volcanic glasses and whole-rocks from both the volcano and its rift zones. Lavas from Nifonea volcano show an evolution towards more incompatible element enrichment, with the most enriched lavas being the youngest eruption products on the caldera floor. These are products of significant fractional crystallization, show minor contamination by hydrothermal fluids (〈0·3%) and reflect mixing of melts derived from depleted upper mantle and melts from an enriched source similar to those occurring in the North Fiji Basin. The enrichment in Nb of these lavas is comparable with that of some lavas from the New Hebrides island arc (e.g. Mota Lava island), where these coexist with typical island arc basalts. The lavas erupted along the rift zones in the Vate Trough back-arc basin are relatively depleted in incompatible elements, indicating melting of depleted upper mantle with a minor addition of a sediment-derived fluid. Our observations suggest that the mantle beneath Vate Trough is heterogeneous on a small scale (〈20 km) and that the occurrence of these enriched and fertile mantle portions has a stronger control on melting processes than the influx from the subducting slab, as all samples were recovered at a similar distance from the trench.

Beschreibung: The Louisville Seamount Trail is a 4300 km long volcanic chain that has been built in the past 80 m.y. as the Pacific plate moved over a persistent mantle melting anomaly or hotspot. Because of its linear morphology and its long-lived age-progressive volcanism, Louisville is the South Pacific counterpart of the much better studied Hawaiian-Emperor Seamount Trail. Together, Louisville and Hawaii are textbook examples of two primary hotspots that have been keystones in deciphering the motion of the Pacific plate relative to a set of "fixed" deep-mantle plumes. However, drilling during Ocean Drilling Program (ODP) Leg 197 in the Emperor Seamounts documented a large ~15° southward motion of the Hawaiian hotspot prior to 50 Ma. Is it possible that the Hawaiian and Louisville hotspots moved in concert and thus constitute a moving reference frame for modeling plate motion in the Pacific? Alternatively, could they have moved independently, as predicted by mantle flow models that reproduce the observed latitudinal motion for Hawaii but that predict a largely longitudinal shift for the Louisville hotspot? These two end-member geodynamic models were tested during Integrated Ocean Drilling Program (IODP) Expedition 330 to the Louisville Seamount Trail. In addition, existing data from dredged lavas suggest that the mantle plume source of the Louisville hotspot has been remarkably homogeneous for as long as 80 m.y. These lavas are predominantly alkali basalts and likely represent a mostly alkalic shield-building stage, which is in sharp contrast to the massive tholeiitic shield-building stage of Hawaiian volcanoes. Geochemical and isotopic data for the recovered lavas during Expedition 330 will provide insights into the magmatic evolution and melting processes of individual Louisville volcanoes, their progression from shield-building to postshield and (maybe) posterosional stages, the temperature and depth of partial melting of their mantle plume source, and the enigmatic long-lived and apparent geochemical homogeneity of the Louisville mantle source. Collectively, this will enable us to characterize the Louisville Seamount Trail as a product of one of the few global primary hotspots, to better constrain its plume-lithosphere interactions, and to further test the hypothesis that the Ontong Java Plateau formed from the plume head of the Louisville mantle plume around 120 Ma. During Expedition 330 we replicated the drilling strategy of Leg 197, the first expedition to provide compelling evidence for the motion of the Hawaiian mantle plume between 80 and 50 Ma. For that reason we targeted Louisville seamounts that have ages similar to Detroit, Suiko, Nintoku, and Koko Seamounts in the Emperor Seamount Trail. In total, five seamounts were drilled in the Louisville Seamount Trail: Canopus, Rigil, Burton, Achernar, and Hadar Guyots (old to young). By analyzing a large number of time-independent in situ lava flows (and other volcanic eruptive products) from these seamounts using modern paleomagnetic, 40Ar/39Ar geochronological, and geochemical techniques, we will be able to directly compare the paleolatitude estimates and geochemical signatures between the two longest-lived hotspot systems in the Pacific Ocean. We drilled into the summits of the five Louisville guyots and reached volcanic basement at four of these drilling targets. In two cases we targeted larger seamount structures and drilled near the flanks of these ancient volcanoes, and in the other three cases we selected smaller edifices that we drilled closer to their centers. Drilling and logging plans for each of these sites were similar, with coring reaching 522.0 meters below seafloor (mbsf) for Site U1374 and 232.9, 65.7, 11.5, 182.8, and 53.3 mbsf for Sites U1372, U1373, U1375, U1376, and U1377, respectively. Some Expedition 330 drill sites were capped with only a thin layer of pelagic ooze between 6.6 and 13.5 m thick, and, if present, these were cored by using a low-rotation gravity-push technique with the rotary core barrel to maximize recovery. However, at Sites U1373 and U1376 no pelagic ooze was present, and the holes needed to be started directly into cobble-rich hardgrounds. In all cases, the bulk of the seamount sediment cover comprised sequences of volcanic sandstones and various kinds of basalt breccia or basalt conglomerate, which often were interspersed with basaltic lava flows, the spatter/tephra products of submarine eruptions, or other volcanic products, including auto-brecciated flows or peperites. Also several intervals of carbonate were cored, with the special occurrence of a ~15 m thick algal limestone reef at Site U1376 on Burton Guyot. In addition, some condensed pelagic limestone units were recovered on three of the other seamounts, but these did not exceed 30 cm in thickness. Despite their limited presence in the drilled sediment, these limestones provide valuable insights for the paleoclimate record at high ~50° southern latitudes since Mesozoic times. Several Louisville sites progressed from subaerial conditions in the top of volcanic basement into submarine eruptive environments, or drilling of the igneous basement immediately started in submarine volcanic sequences, as was the case for Sites U1376 and U1377 on Burton and Hadar Guyots. At three sites we cored 〉100 m into the igneous basement: 187.3 m at Site U1372, 505.3 m at Site U1374, and 140.9 m at Site U1376. At the other sites we did not core into basement (Site U1375) or we cored only 38.2 m (Site U1377) because of unstable hole conditions. Even so, drilling during Expedition 330 resulted in a large number of in situ lava flows, pillow basalts, or other types of volcanic products such as auto-brecciated lava flows, intrusive sheets or dikes, and peperites. In particular, the three holes on Canopus and Rigil Guyots (the two oldest seamounts drilled in the Louisville Seamount Trail), resulted in adequate numbers of in situ lava flows to average out paleosecular variation, with probable eruption ages estimated at ~78 and 73 Ma, respectively. Remarkably, at all drill sites large quantities of hyaloclastites, volcanic sandstones, and basaltic breccias were also recovered, which in many cases show consistent paleomagnetic inclinations compared to the lava flows bracketing these units. For Site U1374 on Rigil Guyot we also observed a magnetic polarity reversal in the cored sequence. Overall, this is very promising for determining a reliable paleolatitude record for the Louisville Seamounts following detailed postcruise examinations. The deeper penetrations of several hundred meters required bit changes and reentries using free-fall funnels. Basement penetration rates were 1.8–2.5 m/h depending on drill depth. In total, 1114 m of sediment and igneous basement at five seamounts was drilled, and 806 m was recovered (average recovery = 72.4%). At Site U1374 on Rigil Guyot, a total of 522 m was drilled, with a record-breaking 87.8% recovery. Most outstandingly, nearly all Expedition 330 core material is characterized by low degrees of alteration, providing us with a large quantity of samples of mostly well-preserved basalt, containing, for example, pristine olivine crystals with melt inclusions, fresh volcanic glass, unaltered plagioclase, carbonate, zeolite and celadonite alteration minerals, various micro- and macrofossils, and, in one case, mantle xenoliths and xenocrysts. The large quantity and excellent quality of the recovered sample material allow us to address all the scientific objectives of this expedition and beyond.

Beschreibung: [1] The Terceira Rift formed relatively recently (∼1 Ma ago) by rifting of the old oceanic lithosphere of the Azores Plateau and is currently spreading at a rate of 2–4mm/a. Together with the Mid-Atlantic Ridge, the Terceira Rift forms a triple junction that separates the Eurasian, African, and American Plates. Four volcanic systems (São Miguel, João de Castro, Terceira, Graciosa), three of which are islands, are distinguished along the axis and are separated by deep avolcanic basins similar to other ultraslow spreading centers. The major element, trace element and Sr-Nd-Pb isotope geochemistry of submarine and subaerial lavas display large along-axis variations. Major and trace element modeling suggests melting in the garnet stability field at smaller degrees of partial melting at the easternmost volcanic system (São Miguel) compared to the central and western volcanoes, which appear to be characterized by slightly higher melting degrees in the spinel/garnet transition zone. The degrees of partial melting at the Terceira Rift are slightly lower than at other ultraslow mid-ocean ridge spreading axes (Southwest Indian Ridge, Gakkel Ridge) and occur at greater depths as a result of the melting anomaly beneath the Azores. The combined interaction of a high obliquity, very slow spreading rates, and a thick preexisting lithosphere along the axis probably prevents the formation and eruption of larger amounts of melt along the Terceira Rift. However, the presence of ocean islands requires a relatively stable melting anomaly over relatively long periods of time. The trace element and Sr-Nd-Pb isotopes display individual binary mixing arrays for each volcanic system and thus provide additional evidence for focused magmatism with no (or very limited) melt or source interaction between the volcanic systems. The westernmost mantle sources beneath Graciosa and the most radiogenic lavas from the neighboring Mid-Atlantic Ridge suggest a mantle flow from Graciosa toward the Mid-Atlantic Ridge and hence a flux of mantle material from one spreading axis into the other. The Terceira Rift represents a unique oceanic rift system situated within the thickened, relatively old oceanic lithosphere and thus exhibits both oceanic and continental features.

Beschreibung: Abundant volcanic activity occurs in the back-arc region of the northern Tofua island arc where the Northeast Lau Spreading Centre (NELSC) propagates southwards into older crust causing the formation of numerous seamounts at the propagating rift tip. An off-axis volcanic diagonal ridge (DR) occurs at the eastern flank of the NELSC, linking the large rear-arc volcano Niuatahi with the NELSC. New geochemical data from the NELSC, the southern propagator seamounts, and DR reveal that the NELSC lavas are tholeiitic basalts whereas the rear-arc volcanoes typically erupt lavas with boninitic composition. The sharp geochemical boundary probably reflects the viscosity contrast between off-axis hydrous harzburgitic mantle and dry fertile mantle beneath the NELSC. The new data do not indicate an inflow of Samoa plume mantle into the NELSC, confirming previously published He isotope data. The NELSC magmas form by mixing of an enriched and a depleted Indian Ocean-type upper mantle end-member implying a highly heterogeneous upper mantle composition in this area. Most NELSC lavas are little affected by a slab component implying that melting is adiabatic beneath the spreading center. The DR lavas show the influence of a component from the subducted Louisville Seamount Chain, which was previously thought to be restricted to the nearby arc volcanoes Niuatoputapu and Tafahi. This signature is rarely detected along the NELSC implying little mixing of melts from the low-viscosity hydrous portion of the mantle wedge beneath the rear-arc volcanoes into the melting region of the dry mantle beneath the NELSC.