The oldest oceanic crust is Jurassic seafloor in the western Pacific Ocean Basin (Fig. F1). Sampling deep ocean basin sites at water depths of >6000 m involves drilling the hard Cretaceous chert sequences that lie a few hundred meters above basement. This technical challenge is part of an unfolding drama of three decades of drilling by the Deep Sea Drilling Project (DSDP) and ODP in the western Pacific abyssal plains. The most recent penetration of the ~170-m.y.-old crust by ODP involved the reentry of Hole 801C during Leg 185 as part of the "Subduction Factory experiment."
Much of the following historical summary is taken from the Acknowledgments of the Leg 129 Initial Reports volume (Lancelot, Larson, et al., 1990). DSDP Legs 6 and 7 in 1969 were the first to search the western Pacific for the Earth's oldest oceanic crust and sediments. The search ultimately took 20 yr and 10 cruises of DSDP/ODP (Legs 6, 7, 17, 20, 32, 33, 60, 61, 89, and 129) to achieve the final goal. Many people were involved; the most persistent members of the "Old Pacific Club" included B.C. Heezen, E.L. Winterer, S.O. Schlanger, R. Moberly, I. Premoli Silva, R. Larson, Y. Lancelot, W. Sliter, D. Bukry, R.G. Douglas, and H.P. Foreman. During the early legs, drilling sites were targeted with single-channel seismic records characterized by acoustically opaque chert layers that obscured the underlying volcanic basement. Often, coring was frustrated by these impenetrable cherts, as well as by volcaniclastic sediments and basalts of Cretaceous age. To those who shipped out repeatedly and returned home with more questions than answers, what had started as an oceanographic exercise turned into an ongoing quest for the "old Pacific."
Leg 129 brought the JOIDES Resolution, with improved drilling capabilities, where the Glomar Challenger failed. Also, preparations for Leg 129, led by Yves Lancelot and Roger Larson, included four multichannel seismic expeditions to the area searching for seismic "windows" through the Cretaceous volcaniclastic sediments and solid basalts.
This combination of improved science and technology was finally successful in 1989 at Site 801 in the Pigafetta Basin, where Jurassic sediments of Bajocian–Bathonian age were discovered overlying ~170-Ma oceanic crust (Lancelot, Larson, et al., 1990).
During Leg 185, Hole 801C was deepened by 340 m into basement, providing a total basement section of 470 m, making it the sixth deepest drill site into normal oceanic crust. Recovery of oceanic basement was very good (47%) and included fresh basaltic glass, which has enabled study of the primary basalt chemistry (Fisk and Kelley, 2002). A high-quality set of logs were run to 388 m in basement and, with the extended core penetration, have revealed a remarkable record of magnetic reversals in the Jurassic section (Tivey et al., 2005; Steiner, 2001). The hole is in good condition, and it remains a legacy site into the world's oldest oceanic crust.
The primary motivation for returning to Hole 801C, seaward of the Mariana Trench, was to sample the upper oxidative zone of alteration of this oldest in situ oceanic crust. Previous drilling during Leg 129 only penetrated 63 m into "normal" Jurassic basement. Based on basement rocks from Hole 504B and other basement sites with sufficient penetration, the upper oxidative zone of alteration, which contains the overwhelming majority of some element budgets (e.g., K, B, etc.), lies in the upper 200–300 m of the basaltic crust.
An important objective of Hole 801C in the Pigafetta Basin and Hole 1149D close to the Izu-Bonin subduction zone in the Nadezhda Basin was the establishment of their radiometric and biostratigraphic ages, hence providing constraints on the birth of the Pacific plate. Combined with previous results from Pringle (1992) for Holes 801B and 801C, Koppers et al. (2003) arrived at a multistage history for the site. The oldest part of the Pacific plate was formed at the spreading ridges at 167.4 ± 1.4/3.4 Ma. Alkaline volcanics, which compose the upper unit sampled in basement, were erupted at ~7 Ma after the tholeiites at ~160.1 ± 0.6 Ma. The older age for the tholeiitic basement has been confirmed by radiolarian ages, which range from late Bajocian to middle Bathonian (167–173 Ma) (Bartolini and Larson, 2001).
This difference in age between the lower tholeiitic unit and the upper alkaline unit was recognized based on differences in the structural characteristics of this basement section (Pockalny and Larson, 2003) (Fig. F2). Thin layers comprising hydrothermal deposits separate these sequences, which, in addition to the difference in isotopic age, show distinct major and trace element compositions (Fisk and Kelley, 2002). Using a half-spreading rate of 60–70 km/m.y. and the ages given above, the younger volcanic sequence and hydrothermal activity took place as much as 500 km away from the spreading ridge that formed the basement at Site 801.
As discussed by Koppers et al. (2003), the ages defined at Site 801 provide a calibration point on the geological reversal timescale, as this site is located at the oldest point for Mesozoic magnetic anomalies. Hole 1149D basement potentially gives another important calibration point of 127.0 ± 1.5/3.6 Ma for Anomaly M12, which is younger than current timescale compilations (134.2 ± 2.1 Ma) (Gradstein, 1995). This might suggest that the dated basalt was not formed at a spreading center but was erupted slightly off-axis (Koppers et al., 2003).
Bartolini and Larson (2001) show that the age of Site 801 constrains the time of formation of the Pacific plate to 175–170 Ma, just after the initial separation of the Pangaea supercontinent in the central Atlantic at 190–180 Ma. The authors also identify a time of extensive subduction zone magmatism (175–159 Ma) at the eastern and western edges of Pangaea and suggest that the initial plate separation of Pangaea increased subduction rates along its outer margins and altered the plate boundaries in the Pacific superocean, ultimately leading to formation of the Pacific plate.
Drilling results into basement in the western Pacific basins have provided the ground truth for the geologic timescale, global plate reconstructions, and mapping magnetic reversals in a critical period of Earth's history.
During ODP, few deep holes have been drilled into oceanic crust. Hole 504B, the most renowned of the deep crust drilling sites of ODP, is located on 5.9-Ma crust in the Panama Basin south of the Costa Rica Rift axis (Alt et al., 1996). Drilling this hole occupied seven ODP legs and penetrated >2000 m into oceanic basement, reaching close to the Layer 2–3 transition zone. On the other hand, two other sites (DSDP Site 534 and ODP Site 765) in addition to Site 801 have sampled the oldest crust in the oceans (Table T4). Given the range in crustal age and spreading rate, these sites have proved invaluable in understanding the geochemical balance between altered basaltic crust and seawater involved in global continent–ocean geochemical cycles.
In addition to being the oldest oceanic crust, Hole 801C deeply samples crust formed at fast spreading rates (130–140 km/m.y. whole rate) (Tivey et al., 2005). Thus, Site 801C provides a unique record of crustal structure and accretionary process of oceanic crust formed in a fast-spreading environment.
Pockalny and Larson (2003) used downhole logging data and basement stratigraphy to determine the spreading environment and crustal accretion history of the ocean basement cored in Hole 801C. High-resolution microresistivity data obtained with the Formation MicroScanner (FMS) were used to measure the dip of the extrusive layers and indicate a 10°–30° increase in dip downhole with lava flow contacts dipping back toward the original ridge axis. This structural pattern and the high proportion of massive flows relative to pillow units are consistent with prevailing crustal accretion models proposed for faster spreading ridges (e.g., >60 km/m.y.). Combined with geochronological, geochemical, and structural information (Fig. F2), these authors defined the emplacement history of the lavas erupted at Site 801: the shallowest 100 m of the drilled section (e.g., Sequences I–III) were emplaced just off the ridge (Sequence III) or significantly farther off-axis up to 5–15 m.y. later (Sequences I and II). The remainder of the drilled section (Sequences IV–VIII) has geochemical, lithological, and physical trends that are assumed to be representative of crust created at fast-spreading ridges.
The composition of the basement in terms of proportions of massive vs. pillowed lavas and interflow breccias, hyaloclastites, and sedimentary horizons is one of the major challenges when drilling a "one-dimensional" hole in ocean crust. In addition to the problem of lateral extrapolation of the lithologies, the extent of recovery during coring is highly variable. For Leg 185, determining the proportion of basement lithologies was fundamental to reconstruction of the bulk geochemistry for global flux studies. For example, interflow segments are often poorly sampled by drilling but richly are concentrated in alteration phases (carbonate and alkali-bearing clay minerals). The primary data used to reconstruct the bulk composition of the site include the alteration mineralogy (Talbi and Honnorez, 2003), chemical analyses of core materials (Kelley et al., 2003), and geochemical and geophysical logs. Of particular note are two studies (Barr et al., 2002; Revillon et al., 2002) which integrated downhole logging data, geochemistry, and formation structures using data from the FMS. The reconstructed logging-based lithological sequence consists of thick massive flow units (27.4%), pillow units (33%), breccia units (31%), sediments (1.4%), and hydrothermal deposits (1.3%), with 5.9% unclassified due to unreliable tool response in intervals where hole conditions were poor. The proportion of pillow basalts doubled and the amount of breccia increased sixfold from that reported using core description alone, demonstrating convincingly that core-logging integration is essential to providing an accurate representation of the ocean crust section and the input flux to subduction zones (see also "Site 801: A Geochemical Reference Site for Global Budgets and the Aging of Oceanic Crust" in "Geochemical Sinks and Recycling in Subduction Zones.").
The magnetic properties of the basement section of Hole 801C were determined from borehole measurements down to 837 meters below seafloor (Tivey et al., 2005) as well as paleomagnetic and rock magnetic measurements on 480 individual samples and continuous core segments (Steiner et al., 1999; Steiner, 2001). The results from these different data sets support a remarkable paleomagetic history for Hole 801C.
The rock-magnetic properties of the Jurassic tholeiitic basalts are unusual in their excellent state of preservation (except near the zones of hydrothermal alteration). The iron oxides are ordinary titanomagnetite with iron/titanium ratios identical to those in modern, normal mid-ocean-ridge basalts (N-MORB). The basalt remanent intensities and Curie temperatures are the same as those of very young (e.g., 8 Ma) ocean crust. The logging data observed that the horizontal and vertical field measurements are "in phase," unambiguously demonstrating that the site formed in the Southern Hemisphere at a paleolatitude of 22.7° ± 5°S (Tivey et al., 2005). The paleomagnetic data demonstrate that after crustal formation in the Middle Jurassic, the site first moved north, approximately to the paleoequator by the Oxfordian, reversed direction, and transited south to ~15°–20°S by the Albian–Aptian, and reversed direction again to transit monotonically northward across the paleoequator to its present location at 18.6°N.
These observations at Site 801 demonstrate that the Jurassic magnetic quiet zone (JQZ) does not result from either low-temperature oxidation of the magnetic minerals due to their great age or from originally anomalous magnetic mineralogy. Instead, the JQZ most likely results from anomalous behavior of the geomagnetic field in the Middle Jurassic, from excessively fast reversing of the geomagnetic field, and possibly also accompanied by abnormally low geomagnetic field intensity.
The Jurassic basement of Hole 801C contains evidence for multiple geomagnetic field reversals, with six polarity reversals in the drilled volcanic crust (Tivey et al., 2005; Steiner et al., 1999). An appreciable portion of the section recorded transitions of the magnetic field from one polarity interval to the other. Because of potential hiatuses in the basement section and likely changes in the rate of crustal construction during volcanic episodes, an estimate of reversal frequency or polarity interval duration is necessarily imprecise. However, the vertical extrusive volcanic section probably was constructed in ~50,000 yr at this spreading rate (130–140 km/m.y.), and probably the majority of the extrusive section has been penetrated and sampled. Hence, the logging studies suggest an average reversal period of 10,000–15,000 yr (Tivey et al., 2005), which is consistent with that estimated from the paleomagnetic data (Steiner, 2001) This represents approximately an order of magnitude higher sustained geomagnetic reversal frequency than observed at any other time in the past 300 m.y.
This reversal frequency alone could produce the quiet anomaly signature by cancellation of the magnetic signal by stacked opposite-polarity basalt segments. Tivey et al. (2005) calculated the average magnetic moment of the crustal section to be 1930 Am2, assuming an average magnetization of 5 A/m, but when alternating polarity is taken into account this is reduced to an effective moment of ~650 Am2. This moment is equivalent to a 500-m-thick source layer with a magnetization of only ~1 A/m, which also provides a possible explanation for the appreciable reduction in magnetic anomaly amplitude over the area including Site 801.
Studies of contemporaneous stratigraphic sequences (Steiner et al., 1987) also observed an abnormally high frequency of reversals during the same portion of the Middle Jurassic and led Steiner et al. (1987) to suggest at that time, that rapid reversing of the geomagnetic field might be the explanation of the minimal magnetic anomaly signature within the JQZ. McElhinny and Larson (2003) summarized paleointensity studies and concluded that this is also a time of anomalously low dipole field intensity. However, the common observation of very reduced geomagnetic field intensity during reversals (~10%) suggests that a sustained lower geomagnetic field intensity during the extent of the JQZ is quite probable. Thus, the paucity of a magnetic anomaly signature over the JQZ crust may result from the combination of very high reversal frequency with relatively low dipole field intensity.
Finally, the magnetic logging data demonstrated that the appreciably altered tholeiites adjacent to the upper and lower hydrothermal units correspond to a paleolatitude of ~16°S for normally magnetized rocks. Moreover, sedimentary, radiometric, and paleomagnetic data demonstrated that the alkalic basalts overlying tholeiite basement were emplaced when the site was in an approximately equatorial location; hence, the hydrothermal activity was not contemporaneous with alkalic basalt emplacement. Instead, the site occupied a paleolatitude of ~16°S at two different times, several million years after crustal construction and in the mid-Cretaceous. Because midplate volcanism was widespread throughout the western Pacific Basin during the mid-Cretaceous, the polarity changes suggested that the hydrothermal deposits were precipitated in Site 801 tholeiitic crust in association with that activity, ~60 m.y. after construction of the oceanic crust. The evolution of the volcanic units and the magnetic stratigraphy at Site 801C are illustrated in the cartoon depicted in Figure F3 and taken from Tivey et al. (2005).
More than 50 samples of fresh basaltic glass were recovered from Sites 1149 and 801, providing pristine samples of the igneous liquid that forms Jurassic Pacific crust. These valuable samples record mid-ocean-ridge processes, mantle composition, and mantle temperature at a time preceding the Cretaceous "superplume" event in the Pacific (Fisk and Kelley, 2002). Hole 801C records higher Fe8 (10.77 wt%) and marginally lower Na8 (2.21 wt%) compared to MORB from the modern East Pacific Rise (EPR), suggesting deeper melting and a temperature of initial melting that was 60°C hotter than today. Trace element ratios such as La/Sm and Zr/Y, on the other hand, show remarkable similarities to the modern southern EPR, indicating that Site 801 was not generated on a hotspot-influenced ridge and that mantle of similar composition has fed spreading ridges over the past 170 m.y. Hole 801C basalt chemistry indicates that higher temperatures of mantle melting beneath Pacific ridges may have preceded the initiation of the Cretaceous superplume.