The succession is divisible into at least three unconformity-bounded sequences of late Miocene to Pleistocene age (Fig. 3). The sequence boundaries are characterized by well-developed seismic offlap and are informally designated pp3(s), pp4(s), and m0.5(s). Core and log data from Site 1072 confirm that seismic reflection pp5(s), tentatively interpreted as a sequence boundary at the beginning of the leg, is related to fining-upward sediments and an increase in paleowater depth. Ages are consistent with those obtained at Site 1071: pp3(s), late Pleistocene (younger than 0.78 Ma); pp4(s), early Pleistocene-possibly latest Miocene (1.1-7.4 Ma), most likely of Pliocene-Pleistocene age; and m0.5(s), late Miocene. Surface pp5(s) is dated as late Miocene (5.9-7.4 Ma). Coring and logging at Site 1072 suggest that the sandy lower intervals above sequence boundaries m0.5(s) and pp4(s) thin in a seaward direction. Each of these units is a composite, including seismically imbricated intervals 10-25 m thick that shoal upwards. Upward coarsening and possible shoaling (from benthic foraminifers) beneath surface pp4(s) are consistent with a larger scale transition from transgressive to highstand sedimentation.
The sedimentologic column at Site 1072 is divided into two units on the basis of accessory components: glauconite, carbonate, and pyrite nodules (Fig. 7). Unit I extends from 0 to 152.13 mbsf, and Unit II extends from 152.13 to 274.38 mbsf. These units are considered close genetic equivalents to Units I and II at Site 1071; the contact between them corresponds to surface pp4(s). Thick, unrecovered intervals in both units suggest that sands are present in these sections. Unit I is characterized by intervals of dark gray to dark greenish gray silty clays, clayey silts, and clays interbedded with olive gray sandy mud, sandy silt, clayey sand, and/or muddy sand. Slumping/microfaulting is common at the top and bottom of the unit. Bioturbation is associated with intervals where clays are interbedded with coarser sediments. The lower boundary of the unit is rich in glauconite, shells, granules, and pebbles. Glauconite is generally present toward the top of Unit II in olive gray sandy silts with scattered granules, shells, and wood fragments. Carbonate and pyrite nodules, wood fragments, and discrete burrows are associated with dark gray to olive gray silty clay from the middle to the base of Unit II. Cemented intervals of poorly sorted, glauconitic, pebbly, medium- to coarse-grained quartz sandstone are present at the base of this unit.
As at Site 1071, biostratigraphic resolution at Site 1072 is limited for calcareous microfossils
because of strong carbonate dissolution and shallow-water depths that were unfavorable to these
planktonic organisms. Nannofossils provide relatively useful zonations for the Pleistocene and
early Pliocene to late Miocene. Planktonic foraminifers have limited use for biostratigraphic
zonation at Site 1072, although it was possible to identify late Miocene and late Pliocene to
Pleistocene ages based on the presence/absence of rare taxa. Pleistocene benthic foraminiferal
faunas vary from assemblages dominated almost exclusively by Elphidium excavatum to more
diverse assemblages, probably reflecting changing paleodepths or substrates related to
glacial/interglacial cycles and/or depositional systems. Benthic foraminiferal species abundances
indicate that paleodepths were slightly deeper at Site 1072 than at Site 1071 during the
Pleistocene, ranging from inner neritic (0-50 m) to upper middle neritic (~50-65 m) at Site
1072. Miocene biofacies are characterized by Buliminella gracilis and Uvigerina juncea,
indicating middle neritic paleodepths (50-100 m). Organic microfossils are common to
abundant in most samples studied, particularly in the pre-Pleistocene sediments. Dinocysts are
relatively more abundant in pre-Pleistocene sediments as well, whereas terrestrial
palynomorphs (pollen and spores) dominate Pleistocene sediments. Reworking is clearly
evident in many samples. In addition, it is possible that some well-preserved organic
microfossils are indistinguishable from in situ fossils. Surface pp4(s) can be recognized in a
condensed interval of Pliocene-early Pleistocene age (Fig. 8).
Several magnetic polarity boundaries are recorded in sediments recovered from Hole 1072A. The
Brunhes/Matuyama boundary was found at 62.3 mbsf within a clayey silt (Subunit IC), which
was confirmed by nannofossil biostratigraphy. This boundary coincides with a marked increase
both in magnetization intensity and susceptibility downhole. This evidence suggests that the
sedimentary environment, which supplied magnetic minerals, changed across the boundary, or
that the boundary is a diagenetic front during the Brunhes Chron associated with the
sedimentation above. For Subunit IC, below the assumed Brunhes/Matuyama boundary,
magnetic polarity is reversed down to pp4(s) (131-145 mbsf), including normal zones at 127.5
135 mbsf and 141.8-144 mbsf, which are associated with slump/sand layers and intervals of
low recovery. Below pp4(s), within the upper silty clay layer of Subunit IIA, magnetic polarity
is dominantly normal with thin reverse zones at 177-178, 181-182, 206-208, and 215-218
mbsf. For the lower muddy sands of Subunit IIA, near surface pp5(s), magnetic polarity is
normal throughout; magnetization intensity and susceptibility in this interval are lower than in
the overlying upper silty clay layer.
Downhole profiles of interstitial water at Site 1072 are complex, but somewhat similar to those at Site 1071. Pore waters are significantly fresher than seawater (Cl- to 469 mM, a 16% decrease) at shallow subsurface depths (< 150 mbsf; Fig. 7). The salinity minimum is a nonequilibrium feature caused by large-amplitude oscillations in the salinity of overlying water, perhaps a proxy for the rise and fall of sea level. Freshwater has access to the sediment column during presumed glacial stages, whereas seawater covers the sediment column during interglacial stages. The salinity minimum at Site 1072 is less pronounced and ~30 m deeper than the minimum at Site 1071. This difference may reflect the elevation offset between the two locations during the last transgression. Decreases in sulfate, with corresponding increases in alkalinity, ammonia, and phosphate, occur at two distinct intervals in the sediment column: at 30 mbsf in the shallow salinity minimum, and below 150 mbsf. Upper and lower zones of significant organic-matter diagenesis also were observed at similar depths at Site 1071. However, changes in pore-water concentrations are more pronounced at Site 1072, suggesting higher overall rates of organic-matter diagenesis at Site 1072. In particular, sulfate drops to 1 mM in the upper zone at Site 1072, but only to 6 mM at Site 1071.
Hydrocarbon gases were monitored by headspace sampling for every core recovered at Site 1072. As at Site 1071, hydrocarbon gases generally are at or near detection limit (C1< 5 ppmv). Exceptions are three samples near 30 mbsf, where C1 and C2 rise to 1056 ppmv and 4 ppmv, respectively. The presence of C1 at a depth where interstitial-water sulfate approaches zero indicates a thin zone of bacterial methanogenesis at Site 1072, unlike at Site 1071.
A comprehensive set of physical properties measurements was acquired on cores from Hole 1072A, with the exception of the PWL component of the MST, which was not employed because of the presence of incompletely filled core liners. The natural gamma-ray (NGR) component of the MST reaches a local maximum at 57.1 mbsf; this maximum is located immediately above the boundary between Subunits IA/IB and IC and is at the level of sequence boundary pp3(s). The overall maximum NGR value is at 151.6 mbsf; this depth is immediately above the boundary between Subunits IC and IIA and is at or slightly below the inferred level of sequence boundary pp4(s). In recovered intervals, the overall shape of the NGR curve obtained from physical properties measurements is in good agreement with that derived from logging. Wet bulk density values for Hole 1072A average 2.00-2.10 g/cm3, with generally little variation. The maximum values in wet bulk density are found just above the boundary between Subunits IC and IIA [~157.8 mbsf; near pp4(s)] and from the well cemented sandstone recovered at 268.79 mbsf. The shape of the physical properties wet bulk density curve is in good agreement with that obtained from logging at Hole 1072D, although logging-derived values are consistently somewhat higher. Discrete P-wave velocities are generally between 1600 and 1800 m/s, with exceptions noted below. Velocities in excess of 5000 m/s are associated with well-cemented sandstones recovered at 165.35 and 268.79 mbsf. Other high-velocity intervals (>2000 m/s) are located at ~36.6 mbsf and between ~147 and 151 mbsf; the latter interval corresponds to the Subunit IC/IIA lithic boundary and is near the inferred depth of pp4(s). In addition, this higher velocity interval appears to be of sufficient thickness and density contrast to be seismically resolvable. Conversely, the highest velocity intervals, associated with indurated sandstones, may be too thin to be seismically resolvable. Resistivity measurements were taken at least once per recovered section; trends in physical properties and logging resistivity data are consistent overall, although physical properties resistivity values are on average ~20%-25% lower. The highest resistivity value measured in recovered cores was 3.23 ‡m at 147.54 mbsf; this is just above the Subunit IC/IIA [pp4(s)] lithic/sequence boundary. Other measurements include porosity, shear strength, thermal conductivity, and magnetic susceptibility. Physical properties data appear to be consistent with logging data and will prove useful in refining velocity models for seismic data in the vicinity of Sites 1071 and 1072.
Hole 1072A was logged to total depth (TD; 300 mbsf) with the triple combo logging string (dual induction resistivity, neutron porosity, and density tools), plus the spectral gamma ray and the Lamont temperature tools. Log data are of good quality, except for washed-out intervals near the bottom of the hole. A repeat run from 110 mbsf to the bottom of the drill pipe (61 mbsf) confirmed the log responses of the main run. The pipe became stuck while rigging down; this forced the severing of the pipe and moving to Hole 1072B to continue logging.
Hole 1072B was washed to 307 mbsf and logged in four wireline runs. The first run was an induction-sonic string (measuring resistivity and sound velocity, with a spectral gamma-ray tool for correlation to other log runs) that failed to pass a bridge at ~90 mbsf. The tool string was pulled out, the hole was reamed, and a repeat run was made with the long-spaced sonic tool (LSS) in place of the sonic-digital tool (SDT). Despite difficulties in passing several bridges on the way down, a successful run was logged from TD up to the pipe at 43 mbsf. Good velocity data were collected, with cycle-skipping observed in only two thin intervals of marked velocity variation corresponding to indurated sandstone. Overall, the LSS tool, despite its lack of a receiver array, performed better than the SDT in the variable diameter, sandy conditions encountered in Hole 1072B. The third wireline run, from 307 to 50 mbsf, utilized the formation microscanner (FMS) tool. Roughly 20% of the hole was washed beyond the maximum opening of the caliper, but images of the remaining 80% provided good detail of bedding features in intervals of poor core recovery. The final logging operation was a vertical seismic profile (VSP) using the Schlumberger well seismic tool (WST) tool. These data provided interval velocities that compared well with the shipboard P-wave measurements on discrete samples from Hole 1072A, and yielded a time-to-depth conversion that should give seismic core-log correlations a high degree of precision.
Logging while drilling (LWD) was conducted at Holes 1072C and 1072D to 100 and 356 mbsf, respectively. Measurements included resistivity, spectral gamma ray, porosity, density, and photoelectric effect. In addition, borehole diameter was statistically derived from the density measurements.
Preliminary shipboard integration of logs and core data has been useful, both in assessing the
character of unrecovered intervals and calibrating the log measurements. Log data quality are
generally good to excellent, except where sand-rich washed-out intervals are encountered. These
unconsolidated, sand-rich layers correlate to zones of low resistivity, velocity, and gamma-ray
values. Intermediate resistivity, velocity, and gamma-ray data correspond to silty intervals; clay
rich intervals show high resistivity, velocity, and gamma-ray values. The spectral gamma ray,
as well as the photoelectric factor, identify high glauconite concentrations in the upper 154 m
(Pliocene-Pleistocene). Of particular note is an incompletely recovered glauconite-rich sand that
logs show has variable thickness in each of the three holes it was penetrated (149.5-154 mbsf,
Hole 1072A; 149.5-151.5 mbsf, Hole 1072B; 149-154.5 mbsf, Hole 1072D). These holes
were a total of 40 m apart; the variable thickness of the sand is interpreted to reflect erosional
relief at presumed sequence boundary pp4(s). Two intervals of well-cemented glauconitic
quartz sandstone can be detected by especially high resistivity, density, and velocity: one is a
few meters above pp4(s), near 150 mbsf; the other is at 275-277 mbsf, ~20 m above surface
m0.5(s). Throughout most of Hole 1072B, FMS images delineate bed boundaries and internal
structures such as slumping. High as well as low resistive spots in these images may indicate
various kinds of nodules, clasts, and burrows.