Abstract:
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 in the Emperor Seamounts during Deep Sea Drilling Project Leg 55 and Ocean Drilling Program Legs 145 and 197 documented a large approximately 15 degrees southward motion of the Hawaiian hotspot prior to approximately 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 Expedition 330 to the Louisville Seamount Trail. Existing data from dredged lava suggest that the mantle plume source of the Louisville hotspot has been remarkably homogeneous for as long as 80 m.y. This lava is composed of predominantly alkalic basalt and likely represents a mostly alkalic shield-building stage, which contrasts sharply with the massive tholeiitic shield-building stage of Hawaiian volcanoes. Geochemical and isotopic data for core samples recovered 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 perhaps 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. The drilling strategy of Expedition 330 replicated that of Leg 197, the first expedition to provide compelling evidence for the motion of the Hawaiian mantle plume between approximately 80 and 50 Ma. For that reason drilling targeted seamounts in the Louisville Seamount Trail equivalent in age to Detroit, Suiko, Nintoku, and Koko Seamounts in the Hawaiian-Emperor Seamount Trail. In total, six sites were drilled on five seamounts: Canopus, Rigil, Burton, Achernar, and Hadar Guyots (from oldest to youngest). Analysis of 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 will allow direct comparison of the paleolatitude estimates and geochemical signatures of the two longest-lived hotspot systems in the Pacific Ocean. In all cases, the summits of the flat-topped seamounts (i.e., guyots) were drilled, and volcanic basement was reached at four of these drilling targets. In two cases larger seamount structures were targeted and drilled near their flanks; in the other three cases smaller edifices were selected and 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 a thin layer of pelagic ooze between 6.6 and 13.5 m thick, which was cored using a gravity-push approach with little or no rotation of the rotary core barrel assembly 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 sedimentary cover comprised sequences of volcanic sandstone and various kinds of basalt breccia or basalt conglomerate that often were interspersed with basaltic lava flows, the spatter/tephra products of submarine eruptions, or other volcanic products, including autobrecciated lava flows or peperites. Also, several intervals of carbonate were cored, with the particularly interesting occurrence of a approximately 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 limited presence in the drilled sediments, these limestone occurrences provide valuable insights for the paleoclimate record at high approximately 50 degrees southern latitudes since the Cretaceous. Several Louisville sites progressed from submarine to subaerial eruptive environments at the top of the volcanic basement. However, at Sites U1376 and U1377 on Burton and Hadar Guyots, igneous basement immediately began with submarine volcanic sequences. More than 100 m of igneous basement was cored at three sites: 187.3 m at Site U1372, 505.3 m at Site U1374, and 140.9 m at Site U1376. At the other sites basement was not cored (Site U1375) or recovery was limited to 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 autobrecciated lava flows, intrusive sheets or dikes, and peperites. In particular, the three holes on Canopus and Rigil Guyots, with probable eruption ages estimated at approximately 75-77 and 73 Ma, respectively, likely have adequate numbers of in situ lava flows to average out paleosecular variations of the geomagnetic field. Remarkably, at all drill sites large quantities of hyaloclastites, volcanic sandstone, and basaltic breccia were also recovered, and in many cases these show consistent paleomagnetic inclinations compared to the intercalated lava flows. In the cored sequence for Site U1374 on Rigil Guyot both normal and reversed polarity was documented. Overall, the cored basement sequences are very promising for determining a reliable paleolatitude record for the Louisville Seamounts following detailed postexpedition examinations. 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 was drilled at five seamounts, with an average recovery of 72.4%. At Site U1374 on Rigil Guyot, a total of 522 m was drilled, with a record-breaking 87.8% recovery.