Latitude: 27°11.15'S

Longitude: 1°34.62'E

Water depth: 4755 m

Maximum depth of penetration: 213 meters below seafloor (mbsf)

Oldest formation: late Maastrichtian

Time on site: 4.69 days (1300 hr on 24 March–0530 hr on 29 March)

Site 1262 (proposed Site WALV-12A) is located at the northwestern end of the Walvis Ridge drilling transect and represents the deep end-member of the depth transect (Figs. F12, F13). At 4.75 km depth, Site 1262 is close to the level of the present-day lysocline and CCD, which in this sector of the eastern Atlantic Ocean are below 4.8 and 5.0 km, respectively. Close proximity to the CCD appears to be maintained for much of the Cenozoic as the rate of local subsidence has more or less kept pace with the long-term deepening of the CCD. As a result, Site 1262 is uniquely situated to record major changes in regional and/or global ocean carbon chemistry and/or circulation. As the deep end-member of the Walvis Ridge transect, Site 1262 was drilled with the objective of recovering sections suitable for detailing changes in bottom water chemistry and circulation at abyssal depths during several of the key paleoceanographic events of the Paleogene including the Eocene–Oligocene transition, the PETM, and the K/P boundary extinction. Initial results indicate that this objective was achieved.

Three holes, offset ~20 m from each other, were cored at Site 1262 using the APC coring system. A 213-m-thick section of upper Maastrichtian to Pleistocene nannofossil ooze and clay was recovered. Hole 1262A was started 6 m below the mudline and prematurely terminated at 150 mbsf because of a severed core barrel. Hole 1262B was cored from the mudline to 210 mbsf, and Hole 1262C was cored from 90 to 213 mbsf. Nominal recovery exceeded 100% in all three holes. Using magnetic susceptibility (MS) data, cores from the three holes were correlated by depth shifting and representative intervals were spliced together to create a single stratigraphic section with a total length of 236 meters composite depth (mcd).

Three lithologic units and five subunits were recognized (Fig. F14). Unit I (0–46 mcd) consists of upper Miocene to Pleistocene nannofossil ooze and foraminifer-bearing nannofossil ooze with sedimentation rates up to 10 m/m.y. Unit II (46–90 mcd) is divided into three distinct subunits based on the relative abundance of clay. Subunits IIA (46–68 mcd) and IIC (79–90 mcd) are upper Oligocene to upper Miocene and middle to upper Eocene clay units separated by a lower Oligocene nannofossil clay interval (68–79 mcd). Unit III (90–213 mcd) consists of upper Maastrichtian to upper Eocene clayey nannofossil ooze. Biostratigraphic results show the section to be stratigraphically complete in the Pleistocene and in the upper Paleocene–lower Eocene intervals with sedimentation rates up to 12 m/m.y. The middle to upper Eocene is highly condensed. Calcareous microfossils are generally well preserved at this site, particularly in the lower Eocene and upper Paleocene.

Sharp transitions between carbonate- and clay-rich facies at Site 1262 are an expression of the deepening of the CCD and related changes in ocean carbon chemistry and/or circulation. The carbonate-rich facies include the Pleistocene, Pliocene, lower Oligocene, Paleocene and lower–middle Eocene, and Maastrichtian intervals. The clay-rich facies include the Miocene and middle to upper Eocene sections as well as several discrete layers at the P/E and K/P boundaries. Preliminary age assignments indicate that each of the major facies changes at Site 1262 corresponds to a previously documented shift in the level of the CCD. The Pliocene and Pleistocene sedimentation rates of up to 10 m/m.y. are consistent with moderate rates of carbonate dissolution in this part of the Atlantic Ocean (Fig. F15). Dissolution occurred primarily during the Pleistocene glacial maxima, when the lower boundary of NADW shoaled and allowed more corrosive AABW to cross over into the Angola Basin through mid-ocean-ridge fracture zones. The facies transition between lithologic Units II and I reflects on a regional deepening of the CCD during the late Miocene and earliest Pliocene as has also been recorded in other cores from the abyssal Atlantic Ocean. Regional deepening of the CCD indicates that the Angola Basin would have been filled primarily with more corrosive AABW prior to the late Miocene, and the CCD would have shoaled. Similarly, the carbonate-rich lower Oligocene interval (lithologic Subunit IIB) implies a deep CCD, whereas the underlying upper Eocene clay (lithologic Subunit IIC) implies a shallow CCD. The contact between these two units is sharp, indicating that the CCD descent occurred rapidly, possibly in the span of two obliquity cycles as suggested by observations of Pacific cores recovered during Leg 199. The transition back into clay-rich facies in the mid-Oligocene does not imply a shoaling CCD but, rather, the local deepening of the seafloor via subsidence.

The most prominent clay layer at Site 1262 is a 60-cm-thick unit at the P/E boundary (base at 140.04 mcd) that is embedded within a thick and uniform sequence of upper Paleocene and lower Eocene foraminifer nannofossil ooze. The benthic foraminiferal extinction event occurs just below the base of this layer at 140.18 mcd, as does a major shift in nannofossil abundances from Fasciculithus to Zygrhablithus. The basal color contact is relatively sharp, although magnetic susceptibility data indicate a more gradual, steplike increase in clay content over the lower 20 cm, with at least two brief reversals. The upper contact, although gradational, is relatively sharp compared to P/E boundary sections recovered at shallower water depths. Still, the overall pattern is consistent with other pelagic records and is inferred to result from seafloor carbonate dissolution because of the input of methane-derived CO₂. Overlying the clay layer is a sequence of nannofossil ooze, which is noticeably richer in carbonate than the unit immediately underlying the clay layer. This is an important feature of this boundary sequence as it confirms another prediction of the hydrate dissociation model (e.g., an overcompensation in global carbonate deposition driven by weathering feedbacks). In theory, such a feedback would be a natural response to rapid input of a large mass (2000 Gt) of carbon dioxide.

Another anomalous clay layer is present at the K/P boundary at 216.58 mcd. The basal contact of this layer is sharp, both in color and in MS. This clay layer gradually grades upward into clay nannofossil ooze over several meters. The lowest Danian biozones are well represented, if not expanded, in this section. The P and P1a zones are 0.4 and 1.0 m thick, respectively. Preservation of foraminifers and calcareous nannofossils is excellent, particularly in the clay-rich layers above the boundary. Many of the "dwarfed" foraminifer specimens exhibit "glassy" texture and should be particularly useful for geochemical and textural studies. The post-extinction flora is dominated by Thoracosphaera spp. Key marker species Cruciplacolithus primus and Cruciplacolithus tenuis first appear at 214.5 and 213.8 mcd, respectively. The boundary is present in the upper third of a reversed zone, Chron C29R, although postcruise analysis of discrete samples is required to confirm the chron boundaries. The combination of orbitally paced bedding cycles, stratigraphic continuity, and excellent fossil preservation will permit further refinement of key biostratigraphic datums, as well as testing of models concerning the rate of ecosystem recovery following mass extinction.

Pervasive bedding cycles are expressed in the MS, color reflectance, and other high-resolution core logging data at Site 1262. The lower Eocene and upper Paleocene cores, in particular, are characterized by pronounced decimeter- to meter-scale bedding cycles. The variance is concentrated in three frequency bands. The shorter cycles have a frequency close to that of the orbital precession, whereas the longer oscillations have frequencies similar to the 100- and 400-k.y. eccentricity cycles. Assuming pacing by precession, the total number of high frequency cycles in the upper Paleocene and lower Eocene would suggest that the sequence is stratigraphically continuous. The bedding cycles are even more pronounced in the Maastrichtian with power, again, mostly concentrated in the precession and eccentricity bands. Above the K/P boundary, the power shifts into the 100-k.y. eccentricity band. This phenomenon results from a 75% reduction in accumulation rates, primarily in the carbonate component. As previously recognized in most pelagic K/P boundary sequences, carbonate accumulation rates do not recover until much later in the Cenozoic. The presence of these orbitally paced cycles in stratigraphically complete sections provides a unique opportunity to astronomically calibrate the duration of Paleocene and lower Eocene chrons.