Paleomagnetism and the Hotspot Test

The main goal of Leg 197 was to obtain sufficient basement penetration at several Emperor Seamount sites to test the hypothesis that the Hawaiian hotspot migrated southward during Late Cretaceous to early Tertiary time. This hypothesis further challenges long-standing notions about the tectonic implications of the bend in the Hawaiian-Emperor Seamount chain. The bend is usually considered to represent a large change in Pacific plate motion at 43 Ma; this interpretation can be found in nearly all textbook descriptions of hotspots and plate motions.

Because of the record-setting basement penetration achieved during the leg, we were able to meet all of our goals. A key part of our study plan involved assessing whether secular variation had been averaged by lava flows recovered at a given site through a large number of shipboard paleomagnetic and rock magnetic measurements. These data have provided a firm basis both for an initial assessment of the results of the Leg 197 paleolatitude experiment and for guiding the shore-based work that must be completed to finalize the hotspot motion test.

Prior to the leg, only two time-averaged paleomagnetic data sets were available to address fixity of the Hawaiian hotspot during formation of the Emperor Seamounts. Paleolatitudes derived from analyses of basement cores recovered from Site 433 on Suiko Seamount (Kono, 1980) and Site 844 on Detroit Seamount (Tarduno and Cottrell, 1997) clearly differ from the latitude of Hawaii (~19) (Fig. F4). Although each paleolatitude represents the summary of a large data set composed of many paleomagnetic measurements, rates of motion ultimately depend on paleolatitudes for many separate times; therefore, additional data were needed both to confirm the motion suggested and to better constrain potential hotspot drift rates. Our shipboard data allow us to draw preliminary conclusions on both of these aspects.

The paleolatitudes suggested from our preliminary paleomagnetic analysis of the basement cores recovered at Sites 1203, 1204, 1205, and 1206 also clearly differ from the latitude of Hawaii. The values are consistent with prior data from Suiko and Detroit Seamounts and the hypothesis that the Hawaiian hotspot moved southward from 81 to 43 Ma at rates of 30-50 mm/yr. These values, which are within the range of velocities typical of some lithospheric plates, force us to reconsider the cause of the Hawaiian-Emperor bend, rates of mantle convection, and Pacific plate reconstructions based on the fixed hotspot assumption. The latter issue is of particular interest for ODP studies, as Leg 198 and Leg 199 scientists will require accurate paleolatitude control in their studies of Paleogene paleoceanography in the Pacific Basin.

Whereas the implications of the Leg 197 shipboard paleomagnetic data are exciting and multifaceted, the results are nevertheless preliminary and must be supported by shore-based paleomagnetic, rock magnetic, and radiometric measurements. In particular, analyses employing detailed stepwise thermal demagnetization are needed to better resolve our estimates, which are currently based solely on alternating-field demagnetization data. Thermal demagnetization is necessary because secondary magnetic minerals with intermediate to high coercivities could carry important magnetizations that might contaminate the characteristic directions we have isolated in our shipboard analyses.

Evidence for such magnetic mineral carriers has indeed been found in cores from some sites (especially Site 1204), as recorded in observations under reflected-light microscopy and in the results of rock magnetic measurements. Additional shore-based rock magnetic analyses including magnetic hysteresis, Curie temperature, and low-temperature measurements are needed to further characterize the magnetic carriers in the basement rocks recovered at each site.

The shore-based thermal demagnetization and rock magnetic data hold the promise of achieving much more than confirming our preliminary paleolatitude values. Because thermal demagnetization is a more efficient means of magnetic cleaning than the application of alternating magnetic fields (given the presence of high-coercivity magnetic minerals), the data may allow us to better constrain the uncertainty limits of each paleolatitude data set. Several units with limited recovery were reserved for shore-based thermal demagnetization study; inclusion of data from these units will increase the precision of the mean paleolatitudes. In addition, for Sites 1205 and 1206 thermal demagnetization analysis of recovered soils and deeply weathered lava flow tops (Fig. F51) will provide paleolatitude constraints based on a natural recording medium that averages significantly more time than a given lava flow. Finally, the application of thermal demagnetization analyses together with a host of rock magnetic measurements will allow us to isolate and identify magnetic overprints. If properly understood, such overprints can be used to reorient cores and obtain paleodeclination information (Cottrell and Tarduno, in press). Complementary studies of zeolite assemblages in the recovered cores (Fig. F52) may allow us to constrain the temperatures responsible for magnetic overprints and use them with greater confidence for tectonic studies.

The principal motivation for obtaining paleodeclination data is to address predictions of relative plate motion studies (e.g., Cande et al., 1995) that call for a rotational component of Hawaiian hotspot motion during Late Cretaceous to early Tertiary time. In addition to the use of magnetic overprints, we will use FMS, general purpose inclinometer tool (GPIT), and Deutsche Montan Technologie (DMT) color scanner data to reorient basement cores. Veins and fractures have been imaged in the recovered cores with the DMT system (Fig. F53). Similar features can be seen in the high-quality FMS images (which are automatically oriented with respect to north with the GPIT data) obtained from logging at Detroit Seamount Site 1203 (Fig. F54). We plan to use these data to orient declinations derived from core pieces having veins and fractures and to use the information to constrain a Late Cretaceous paleomagnetic pole for the Pacific plate.

Additional constraints on paleodeclination may become available through shore-based analyses of downhole data collected with the GBM. This was the first deployment of a magnetometer at an ODP site with a sensor to record tool rotation. Preliminary analyses of the data indicate the rotation history of the tool was successfully recorded (Fig. F55).

In addition to obtaining cores and data needed for the test of the hotspot hypothesis, during Leg 197 we recovered an outstanding collection of basement rocks that will be used to investigate inclination anomalies of geomagnetic origin. These anomalies, although much smaller than those associated with the debate over hotspot drift rates, are nevertheless important for our understanding of the geodynamo. In fact, because few data are available of Late Cretaceous to early Tertiary age from the Pacific region, the paleomagnetic data resulting from thermal demagnetization studies of the Leg 197 sites will carry considerable weight in efforts to study the geometry of the past geomagnetic field. Similarly, the recovery of basaltic glass and whole rocks with favorable rock magnetic characteristics bodes well for paleointensity investigations that will be part of postcruise science studies.

Source and Melting History of the Hawaiian Hotspot

One of the significant initial achievements of Leg 197 has been the setting of a new record for total basement penetration (1220 m; 52% recovery). This augurs well for a variety of planned postcruise investigations, including the long-term petrochemical variability of source and melting, volcano-stratigraphic and environmental setting of eruptions, the timing and duration of volcanism, and the cooling and alteration conditions of lava flows produced by the Hawaiian hotspot.

Observations of lava flow thickness, vesicularity, crystallinity, and morphology, together with analysis of volcaniclastic sediment, have provided a picture of eruptions in subaerial to shallow-water conditions at Detroit and Koko Seamounts and waning subaerial activity at Nintoku Seamount (Fig. F56). Further study of core material from all sites and integration with the downhole logging data, particularly FMS images from Site 1203 (Fig. F17), promises to reveal additional details about eruption rate, volume of flows, and distance from source.

From a limited number of shipboard geochemical measurements, we believe we have captured the transition from Hawaiian tholeiitic shield stage to alkalic postshield stage at each of the volcanic complexes. Between Sites 1203 and 1204 and previously studied Sites 883 and 884, we have a range of compositions at Detroit Seamount that covers most of the variability seen in the volcanoes of the island of Hawaii (Fig. F57). Site 1205 (Nintoku Seamount) basalt is dominantly alkalic but includes tholeiitic compositions, whereas Site 1206 (Koko Seamount) basalt is dominantly tholeiitic but includes alkalic basalt compositions. We did not sample any of the posterosional stage of evolved compositions that occur at the end of Hawaiian island volcanic activity, except as cobbles in a conglomerate above basement at Site 1205.

We have hints that we have sampled different source compositions through the variability of trace element ratios, such as Ti/Zr (Fig. F58). It will remain for shore-based studies of additional trace elements and isotopic compositions (Sr, Nd, Pb, Hf, and He) to evaluate and define these suspected source heterogeneities. We are assured, however, that appropriate material has been recovered in the cores for such a comprehensive geochemical program. Unaltered olivine (Fig. F59), which will be separated for He isotopic studies, was observed at all sites. We will probe melt inclusions in olivine and feldspar (Fig. F60) to discover parental melt compositions. Fresh glass is common and will be the source of information about melt volatile content and magma evolution. The latter subject will also be addressed through studies of zoned feldspar phenocrysts (Fig. F10). Opaque minerals provide a rich source of information about cooling rates of lava flows and subsequent alteration history (Fig. F61). Low-temperature alteration will be studied through the composition of secondary minerals in multiple generations of vesicle and vein fillings (Fig. F62).

Time is an important aspect of the volcanic history recorded at each site. Excellent material was recovered for both mineral (feldspar) and whole-rock age determinations to estimate the timing and duration of volcanism through 40Ar-39Ar incremental heating radiometric dating. In particular, the evidence for prolonged volcanic activity (soil horizons at Site 1205 and alternating shallow-water volcaniclastic sediment with subaerial lava flows at Site 1206) will be investigated, as will the apparent underestimate of current ages from Nintoku and Koko Seamounts relative to the long-term age progression within the Emperor Seamounts (Clague and Dalrymple, 1987).