Site 1253 is located ~200 m seaward of the deformation front, in the deepest part of the Middle America Trench (Figs. F3, F13) Operationally, one primary goal for this site was to recore the sediments immediately above the sill encountered during Leg 170, drill and core for the first time through the sediments below the sill, and core >100 m into the oceanic sections. The other major task was to install a CORK-II observatory in the deep igneous section, where coring and logging were used to identify depths at which to set the packer and osmotic fluid and gas samplers.

There are two primary science objectives for this site. The first goal is to determine the bulk composition and distribution within the incoming plate of key element and isotopic tracers, to provide a baseline in physical, mineralogical, and chemical characteristics against which changes during shallow subduction processes may be measured or inferred. They also provide a starting point from which to examine the recycling from subduction trench to volcanic arc (or deeper mantle) of important components such as CO₂. The second major objective at this site is to investigate the hydrology and thermal structure of the igneous section entering the trench. This objective will be addressed using temperature and pressure data and fluid and gas samples recovered from the observatory 1- to 2-years and 5- to 6-years postcruise. Interstitial water chemistry analyzed shipboard during Leg 205, together with that from Leg 170, provides evidence for contemporary flow of seawater at depth at both reference sites. Although not a primary objective for Leg 205, the ash recovery at the base of the sediment section also tells an interesting story.

The seismic record (Fig. F13) in the vicinity of Hole 1253A (CMP number 3210) images the sedimentary sequence of middle Miocene age with a clear change in resolution at ~6.1 s TWT. This represents the top of the calcareous Unit U3 (Kimura, Silver, Blum, et al., 1997) at ~180 mbsf. Beneath the sedimentary sequence, the strong reflector at 6.34 s TWT images the top of a gabbro sill as revealed by drilling results from Leg 170 at Site 1039. The coherent reflection pattern below the top of the sill is difficult, if not impossible, to interpret below the drilled depth of 600 mbsf.

One hole was drilled at Site 1253, which was partially cored and into which we installed a long-term hydrologic borehole observatory. After setting a reentry cone and 16¹/₂-in casing into the seafloor, we reentered this hole with the rotary core barrel (RCB) and drilled without coring to ~370 mbsf. RCB coring below 370 mbsf penetrated 30 m of calcareous and locally clay-rich sediments with intermittent ash layers (average recovery = 75%) before encountering a gabbro sill between 400 and 431 mbsf (average recovery = 74%). Below the sill were ~30 m of partially lithified calcareous sediments with intermittent ash layers (average recovery = 20%). This interval was followed by coring ~140 m into a second igneous unit (average recovery = 75%), with local zones of 55%–50% recovery.

After coring, operations focused on preparing the hole for logging and CORK-II installation. The hole was opened to 14³/₄ in; 10³/₄-in casing was installed to a depth of ~413 mbsf and cemented in place to inhibit communication between the borehole and the formation. After drilling out the cement shoe and drilling a rat hole with an RCB bit, we logged the hole. Because we conducted operations at or very near Leg 170 drill sites where LWD was conducted, our logging focused on the igneous section at Site 1253. Here, we ran the triple combination (triple combo) and Formation MicroScanner (FMS)-sonic strings to determine physical properties, fracture distribution, and structure of the basement rocks. After an initial logging run encountered an impassable bridge in the shallow sediment section, casing was run into the uppermost part of the sill to stabilize the hole for subsequent CORKing operations. The subsequent logging run encountered a bridge at 530 mbsf that limited the triple combo and first FMS-sonic run to the interval between 530 and 413 mbsf; on a second pass, the FMS-sonic tool passed below the bridge and the hole was logged upward from 566 mbsf. A miniaturized temperature logger was run along with the Lamont-Doherty Earth Observatory Temperature/Acceleration/Pressure tool (TAP).

After logging, we assembled the CORK-II components, including a 4¹/₂-in casing screen, casing packer, and casing made up to the instrument hanger. The entire assembly was then lowered into the hole and latched in to seal the borehole outside of the 4¹/₂-in casing. The OsmoSampler with integral temperature sensors was lowered through the center of, and latched into the bottom of, the 4¹/₂-in casing. The final operation was to inflate the packers and shift spool valves that would connect the CORK-II pressure monitoring system to the formation, which would completely seal the zone to be monitored. Problems with the "go-devil" used for this step made it difficult to determine whether the packer had inflated or the valves had turned for pressure monitoring. Alvin dives since then have confirmed that the installation is fully operational. Three absolute pressure gauges including a data logger are installed within the instrument hanger head. One sensor monitors pressure within the sealed-off fluid sampling zone at the bottom of the hole; one monitors pressure variations present within the borehole above the sealed-off section; and the third sensor provides seafloor reference pressures. One additional sampling line extends from the CORK-II head all the way down to the screened interval below the packer and is available for future pressure/fluid sampling purposes. The specifics of the CORK-II installation, relative to the structure and petrology of the igneous sections, are discussed in more detail below.

Lithostratigraphy and Sediment Geochemistry

Sediment coring began at Site 1253 at 370 mbsf, where nannofossil chalks with minor clay interlayers were recovered, closely similar to those at Leg 170 Site 1039. Other significant grains identified are siliceous sponge spicules, diatoms, and zeolites derived from the degradation of volcanic glass shards. Volcanic detritus (glass, altered glass, and mineral fragments) is ubiquitous, varying between ~3% and 10% of the total. Tephra layers (<1% of total stratigraphic thickness) are typically thin (<5 cm), with mafic layers accounting for >70% of the layers identified. A thick (8 cm) siliceous white tephra was recovered at 398.8 mbsf (Fig. F14). Diagenesis has resulted in moderate lithification in the section, except immediately above the gabbro sill, where the sediment is much more clay rich, laminated, and lithified. This section (Core 205-1253-4R) is less calcareous (<2% CaCO₃), with clays and zeolites forming increasingly large volumes of the sediment in the last 3 m above the gabbro sill. Thin chert layers are also seen at 395.4 mbsf (interval 205-1253A-4R-1, 53–61 cm). Below the first gabbro sill, in Core 205-1253A-10R, less lithified nannofossil chalks were recovered. These are identified as the same lithologic unit as above the sill, but they are dominated by a clastic granular limestone, defined as packstone with clay. Minor amounts of baked sediments, usually inferred to be out of place, were recovered within and below the gabbro sill. Bulk sediment chemistry, by inductively coupled plasma–atomic emission spectroscopy (ICP-AES), largely map the minor variations in lithology, with SiO₂ increasing and CaO and Sr decreasing in the more clay-rich interval immediately above the sill. The TiO₂ and Al₂O₃ in the sediments are largely controlled by the ash contribution; relatively constant Ti/Al ratios through the calcareous and clay sediment sections suggest relatively homogenous amounts of volcanic detritus throughout the section. Baked sediments have chemistry similar to the dominant lithology. No appreciable increases in Fe, Mn, or transition metal concentrations were noted above the sill, in contrast to the increases in Cu, Ni, Zn, and V observed at ~80 m above the sill at Site 1039.

Biostratigraphy

Because of the small amount of new sediment recovery expected during Leg 205, the shipboard science party did not include a micropaleontologist. Samples were taken for a shore-based participant. Results are expected to help constrain the ages of the gabbro sill and the lower igneous unit.

Igneous and Metamorphic Petrology

Coring at this site penetrated two separate igneous subunits (Fig. F15). The upper subunit is a gabbro sill (Fig. F16) and is similar to that encountered at Leg 170 Holes 1039B, 1039C, and 1040C. The sill (Subunit 4A) has been subdivided tentatively into 2 subunits, based on the distribution of voids, veins, grain size variation with depth, and the proportions of plagioclase to pyroxene. The lower igneous subunit (Subunit 4B) has been tentatively subdivided into seven subunits using the same criteria. Both the upper and lower igneous sections contain plagioclase and clinopyroxene phenocrysts, with rare olivine, orthopyroxene, and ilmenite and magnetite (Fig. F17). Subunit 4B, particularly below 513 mbsf, is more glass rich and more altered. Phenocrysts are set in a groundmass that typically varies between microcrystalline and fine grained, with occasional medium-grained material. A 1.3-m-thick interval of cryptocrystalline material is present at 513 mbsf (Fig. F18), where larger amounts of glass and a greater degree of alteration are observed. The petrologic data suggest that Subunit 4B is either a sill complex with multiple intrusions or a series of thick and slowly cooled lava flows. Its possible that changes in petrology and physical properties at ~513 mbsf mark the change from a sill complex to basement; postcruise dating and detailed analysis will be necessary to evaluate the two possible origins of Subunit 4B.

Discrete alteration is highest at the tops of the subunits, is generally higher in the lowermost cores, but is generally low (1%–5%) overall. Veins sampled below 485 mbsf in Subunit 4B contain up to ~11 wt% carbonate, although quantification is difficult because vein material is mixed with various amounts of igneous rock. Diffuse alteration of the bulk rock, in the form of zeolite formation and clay replacement of minerals and glass, ranges from ~10%–50%, with higher levels of alteration seen below 513 mbsf. Chemically, all rocks from both subunits are of basaltic composition (46–49 wt% SiO₂ and 6–9 wt% MgO), with compositional variation in part due to olivine, clinopyroxene, and plagioclase fractionation. Variations in elements such as Ti, V, Ba, and Zr indicate that Subunits 4A and 4B are not comagmatic and possibly could have been derived from different mantle sources. Chemical and isotopic analyses beyond those available shipboard will be necessary to determine whether these mantle sources are associated with the Galapagos hot spot, the EPR, or both.

Structural Geology

The most evident feature of the sediments above the magmatic intrusion (in cores firm enough to preserve original structures) is the tilted bedding with 30° average dip; paleomagnetically reoriented dip azimuths show a westward orientation with a cluster toward 221°, similar to the geometry observed at Sites 1039 and 1040. Small normal faults perpendicular to bedding, with millimeter-scale offsets are common throughout, as are pressure solution structures 3–10 cm long. A conjugate system of reverse faults is observed in interval 205-1253A-2R-3,128–135 cm, in which the principal stress orientation reconstruction gives a north–south oriented (183°), horizontal one. This stress field is oriented at ~30° to the convergence direction (De Mets et al., 1990). The close association of reworked pelagic sediment in the lower part of Subunit U3C, westward-tilted bedding, and the magmatic intrusion suggest that the lowermost part of the sedimentary section was deformed during the gabbro emplacement. At Site 1253, the sediments do not show a clear signal that the subhorizontal shortening observed in the cores is related to the incipient stage of subduction, despite the presence of the deformation front only several hundred meters to the east.

The magmatic intrusions were carefully analyzed with respect to the CORK-II experiment and to provide data for comparison to the FMS data (Fig. F19). A sediment/gabbro contact was recovered in interval 205-1253A-27R-1, 1–6 cm, and it dips 72°, although this piece sits at the top of the core above possible dropstones. The intrusions are commonly cut by magmatic veins. Dilational joints are also frequent, usually filled with a film of green minerals (clay and zeolite?), and rarely are present as open fractures (Fig. F20). The paleomagnetic reorientation of the fractures to the real geographical coordinates has been done with particular care; most joints share preferred orientations with magmatic veins, but two populations do not. Overall fracture density increases with depth (Fig. F19). Common in the lower part of the deeper magnetic intrusion (Core 205-1253A-36R and below) are brittle shear zones, represented by enechelon Riedel shears, usually showing a reverse sense of movement.

Physical Properties

Variations in physical properties correlate with major lithologic changes between sediments and igneous units (Fig. F15). A limited number of measurements indicate decreased porosity and increased grain density and P-wave velocity within sediments immediately above and between the igneous units; these differences may reflect alteration (recrystallization) and porosity reduction caused by emplacement of the igneous units.

Clear trends in the physical property data are: (1) the small but systematic increase in velocity, bulk density, and grain density and decrease in porosity within the lower igneous unit and (2) the higher natural gamma ray (NGR) in the upper igneous unit and the clear shift in NGR emissions at 512 mbsf within the lower igneous unit. The cause of these trends in porosity, density, grain density, and velocity with depth in the lower igneous unit is unclear. The differences in NGR emissions suggest chemical differences between and within the igneous units, which may reflect primary compositional differences or varying degrees of alteration within igneous units that were initially chemically similar. The fact that the trends in porosity, density, and velocity are not correlated with the NGR trend suggests that the processes that control porosity, density, and P-wave velocity are separate from the chemical or lithologic processes that affect the NGR.

Paleomagnetism

Shipboard magnetic studies on the archive-half sections and discrete samples (Fig. F15) established a reliable set of magnetic polarity reversals and investigated rock magnetism, especially the domain state of magnetic minerals in the sediments and igneous rocks. The small amount of coring above the sill yielded a reversal stratigraphy consistent with that seen in Holes 1039B, 1039C, and 1040C during Leg 170. Sediments below the sill generally showed negative polarity, but low recovery and high drilling disturbance preclude identification of a magnetic chron or subchron. Good recovery in the upper gabbro sill allows identification of several intervals of normal and reversed magnetic polarity. In the lower igneous unit, the upper part (between 450 and 513 mbsf) is primarily within an interval of reversed polarity. Two brief intervals of possibly normal polarity are identified, but discrepancies between archive-half and discrete sample results preclude firm identification. Below this depth, multiple intervals of normal and reversed polarity are observed. Postcruise age dating will be necessary to provide an absolute framework for this chronostratigraphy. Saturation remanent magnetization and Lowrie's test of the sediments show three separate unblocking temperatures, interpreted to reflect the presence of goethite, pyrrhotite (or griegite), and magnetite. In the igneous section, magnetization is often unstable and appears to reflect largely multidomain (>100 m) magnetic minerals, presumed to be magnetite. Intervals of more stable magnetization and high magnetic intensity are observed at 400 mbsf in the upper unit and at 462–474, 513–523, and 572–593 mbsf in the lower unit.

Inorganic Geochemistry

Interstitial water chemistry was used to investigate in situ diagenetic reactions and the possibility of fluid flow in basement (Fig. F21). Several features in the pore water chemistry suggest a role for enhanced ash alteration and associated authigenic mineral formation just above and below the sill. Higher Na and much lower K and Si are observed just above the sill, and the Ca and Sr gradients stop decreasing at this depth. Cl concentrations are very slightly freshened (1.5%) relative to seawater, which may reflect opal-A or clay dehydration reactions immediately above the sill. The implied liberation of Na, Ca, and Sr to the fluids suggests ash alteration. The sharp decrease in K and Si are consistent with the uptake of these liberated elements via the authigenic formation of zeolites and quartz, also observed lithologically. Just below the sill, the Mg concentration in the fluid is quite low, consistent with the authigenic formation of more Mg-rich clays associated with ash alteration. Clear overall gradients with depth are noted for calcium, strontium, sulfate, silica, and lithium. The gradients parallel those measured during and after Leg 170 (Kimura, Silver, Blum, et al., 1997) but are shifted deeper by ~40 m, thus maintaining the same depth relationship to the top of the sill. The gradients trend toward values typical of modern seawater in the intervals just above and below the sill.

Organic Geochemistry

Organic geochemistry at this site reflects the low heat flow of the incoming plate, with all hydrocarbon concentrations measured for shipboard safety requirements being below the detection limit of the gas chromatographs. Calcium carbonate concentrations in the sediments above and below the sill range from 32 to 65 wt% and overlap those of Site 1039, except in the laminated, clay-rich sediments just above the sill, where values drop to <2 wt%. CaCO₃ in the igneous rock is low, (<0.4 wt%, except in veins) even in the top and bottom of Subunit 4A and the top of Subunit 4B, immediately adjacent to the sediment section. Veins below 485 mbsf typically contain carbonate (<11 wt%), where some of the differences may be due to variable dilution with igneous material. Total organic carbon is low, and frequently below detection limit, throughout. Sulfur concentrations are <2 wt% in the sediment sections and near zero in the igneous section, except for one vein sampled at 546.1 mbsf.

Microbiology

Sediment whole rounds (5 cm) were taken for contamination testing (microspheres and PFTs) and postcruise microbiological measurements (adenosine 5'-triphosphate [ATP] assay, cell counts, and DNA extractions and analysis). As expected with RCB coring in partially lithified sediments, contamination was significant and variable. In the igneous section, veined intervals were taken as whole rounds (up to 40 cm) and split under sterile conditions. Aliquots will be used for DNA extraction and analysis, culturing experiments and cell counts, fluorescent in situ hybridization studies, and for studies of mineral alteration and chemical change associated with microbial activity. Contamination tests, although difficult to use quantitatively, indicate that the tracers were delivered to all but one cored interval. Interior tracer concentrations are variable, but microsphere concentrations are lower to very low in the interiors.

Downhole Measurements

At Site 1253, the Davis-Villinger Temperature-Pressure Probe (DVTPP) was deployed twice in an attempt to determine the in situ temperature and pressure of the formation. The first measurement was performed directly beneath the casing of the reentry cone at a depth of 60 mbsf; the second was at a depth of 150 mbsf. Prior to these measurements, the bottom water temperature was determined using a high-resolution and calibrated miniaturized temperature data logger (MTL) (Pfender and Villinger, 2002) attached to the video system during reentry, giving a bottom water temperature of 1.989°C at Site 1253. At Site 1039, the bottom water temperature was 1.81°C, as measured by two different tools (water-sampling temperature probe and Adara). The cause of this difference is not clear. The MTL was also affixed to the triple combo logging tool near the TAP and run during logging. Unfortunately, thermal changes attributed to the curing of the cement used to tag the casing to the formation created a large signal visible in the temperature record in the upper logged interval, and temperature differences recorded between the two runs indicate that equilibrium formation temperatures had not been attained. The two DVTPP runs encountered difficulties with electronic noise and excessive tool motion, precluding their use to provide high-quality temperature or pressure measurements at this site.

Logging

The hole was logged upward with one pass of the triple combo and one pass of the FMS from 530 mbsf to the bottom of the casing shoe at 413 mbsf (Figs. F15, F19). On the second pass, the FMS slid past the obstruction at 530 mbsf and the hole was logged upward from 564 mbsf, where a second bridge was encountered. Measured inclination of the hole was very small (0.5°–1.6°). The caliper data indicate that the hole diameter in the logged portions of the upper and lower igneous units (423–431 and 461–561 mbsf) was relatively uniform, ranging mostly between 10 and 12 in. Thin intervals of increased hole diameter are present at 482, 485, 487–489, and 502–504 mbsf. The caliper reached maximum extension between 435 and 461 mbsf, corresponding to the sedimentary section between the igneous units.

The logs can be clearly separated into three intervals on the basis of obvious changes in hole diameter, velocity, resistivity, bulk density, and porosity, corresponding to the upper igneous subunit (Subunit 4A: gabbro sill), the sediments below, and the lower igneous subunit (Subunit 4B) (Fig. F15). In the logged part of sill (413–431 mbsf, with a 14³/₄-in hole above 423 mbsf and a 9³/₈-in hole below), porosities are low and densities, resistivities, and P-wave velocities are high. In the sedimentary section of enlarged borehole (431–461 mbsf) high porosities and low bulk densities, resistivities, and P-wave velocities identify sediments. A return to high bulk densities, resistivities, and P-wave velocities at 461 mbsf indicates the top of the lower igneous subunit. The logs better identify the exact depth of the lower igneous subunit than do core depths, because of partial recovery and the standard curatorial practices of moving any recovered material to the top of the core. The NGR intensity is distinctly higher in Subunit 4A than in the sediments and Subunit 4B. Natural gamma logs are not available below 513 mbsf, where NGR measured on cores using the multisensor track (MST) suggest a small increase in K, U, or Th concentrations in the lower part of the lower igneous subunit.

The logging data identify a change in the character of the resistivity and P-wave velocity logs in Subunit 4B at ~491–493 mbsf. Above this depth in the lower subunit, values are relatively homogenous; below, the logs have similar average values but a more spiky character. FMS images indicate a change in character at a depth of ~508 mbsf. Above that depth, conductive features are generally discontinuous. Below, more closely spaced, thin, near-horizontal to slightly dipping conductive features are present in several intervals that are separated by intervals of poor images and irregular borehole size that could be fractured material or sediment interlayers. Intervals of decreased bulk density but no corresponding velocity decrease may indicate a fractured interval rather than sediment interlayers, which should cause a velocity decrease. Based on the bulk density and sonic logs, potential fractured intervals are inferred at 466–468, 484–486, 490–493, and 506–508 mbsf. Sediment interlayers thinner than the vertical resolution of the sonic tool (107 cm), would not be clearly distinguishable in this log, but the general high density and low porosity in areas of smaller borehole diameter preclude the presence of any significant sediment layers.

FMS images can be used to characterize structure and fabric in the igneous units. The gabbro sill (Subunit 4A) between 419 and 426 mbsf exhibits a blocky texture with an ~0.5-m size to the blocks. Between 426 and 432 mbsf, the formation appears more massive with thin conductive features at a 0.5- to 2-m spacing, although it is difficult to trace the conductive features across the four FMS pads. At the very top of Subunit 4B (463–467 mbsf), curved conductive features (fractures or irregularities in the borehole wall) are common. Between 467 and 493 mbsf, the formation appears more massive to blocky, with 0.5- to 1-m spacing between thin conductive features. These conductive features can be clearly traced across the four pads only between ~472 and 478 mbsf. Between 487 and 493 mbsf, irregular to curved vertical conductive features are present, representing possible fractures or irregularities in the borehole wall. From 493 to 498 mbsf, conductive features are rare, becoming more common again between 498 and 508 mbsf. At 508 mbsf, the character of the FMS image changes to more closely spaced conductive features (<0.5-m spacing). In rare cases, such as at 513–514 mbsf, these conductive features can be traced across the four pads and suggest a low dip angle. Image quality between 514.5 and 518 mbsf is poor because of an enlarged borehole. Relatively low (3800–4000 m/s) P-wave velocities and low (5–15 m) spherically focused resistivities occur at similar depths. Below 518 mbsf, the layered character returns, but the absolute value of resistivity increases. Imaging is poor from 525–527, 534–539, and 542–555 mbsf. From 539 to 541 and 555 to 563 mbsf, the image is characterized by more closely spaced (<0.5 m) thin, nearly horizontal conductive features. These conductive features appear to dip to the southwest. The static FMS images indicate that both intervals have high resistivity. Therefore, it appears unlikely that these are sediment layers.

Synthesis Topics

Central American Ash Eruption

Leg 205 coring recovered an exceptionally thick Central American ash layer (Fig. F14) deep in the section, deposited in sediments of ~17 Ma, when the plate was more than ~1650 km distant from the Middle America Trench. A thick (~8 cm), white siliceous (>62 wt% SiO₂) ash was recovered at 398.8 mbsf at Site 1253, in the lithified and laminated clay-rich sediments just above the gabbro sill. This tephra contrasts with the majority of the ash layers recovered at Site 1253, which are thin (<2 cm), dark gray in color, and typically have 40–50 wt% SiO₂. These tephras appear to be the product of primary air fall events rather than forming through redeposition as volcaniclastic turbidites or hyaloclastites. The white ash is altered, containing fine sand– and silt-sized particles of cloudy, partially palagonitized glass. Shipboard chemical analysis allowed ash data to be evaluated in terms of Nb/Y and Zr/Y ratios. These tracers are regarded as being little affected by alteration because they are immobile in aqueous solutions. They also distinguish effectively between lavas generated from a hot spot source such as the Galapagos and volcanic arc sources such as Central America. The white tephra has a distinct chemical signature from those in the dark ash layers; the white ash is interpreted as being from a volcanic arc source, presumably Central America, whereas the dark ashes appear to derive from a Galapagos source.

Using the thickness of the tephra (8 cm) and the maximum grain size of the glass shards (200 together with the inferred distance from the site of eruption in Central America (presumed to be Costa Rica), it is possible to compare this deposit to those from other major eruptions. The white tephra is similar, in both thickness and grain size as a function of inferred distance, to the late Pleistocene Toba eruption of Sumatra (Ninkovich et al., 1978). It is far greater in size than the large, well-documented Quaternary events in the Campanian Province of Italy, or on Santorini in the Aegean Sea. The presence of a second white ash at interval 205-1253A-4R-3, 44–46 cm, raises the possibility that the 8-cm-thick white tephra is simply the largest in a series of powerful events. Both occurred during a period of strong explosive volcanic activity in the eastern Circum-Pacific region (Kennett et al., 1977), which has been linked to rapid spreading rates and plate reorganization in the Pacific. Both siliceous white tephras lie within an atypical 4-m-thick section of laminated claystone, almost carbonate free (<2 wt% CaCO₃), which is over and underlain by bioturbated nannofossil chalks. This interval may represent a period when the biological productivity of the eastern Pacific was in a state of temporary collapse precipitated by a period of powerful explosive volcanism.

Fluid Flow in the Incoming Plate

As at Site 1039, 1.4 km to the west, interstitial water chemistry determined at Site 1253 is also indicative of fluid flow in or below Subunit 4B, where the chemical composition of the fluid is inferred to approach values typical of modern seawater. Figure F21 shows depth profiles for major, minor, and biogeochemical components determined shipboard in the sediment interval above and below the sill. In the limited sediment interval cored, Site 1253 profiles for Ca and Sr mimic those at Site 1039 (Fig. F21). Highest Ca values (~18.5 mM) seen at ~300 mbsf at Site 1039 likely reflect the effects of mafic ash alteration, which liberates Ca. Mg-calcite and dolomite production are also suggested by Mg and Mg/Ca profiles at Site 1039 (Kimura, Silver, Blum, et al., 1997). In the pore waters from the deeper sediments at Site 1253, Ca and Sr decrease by ~20%–30%, toward, but not to, values typical of seawater. A similar magnitude change is seen in the Si content of the pore fluids above and below the sill, excluding the exceptionally low values seen at the immediate boundary with the sill (where quartz precipitation was noted). Li contents increase by ~60% over the lowermost 30 m of the section above the sill. Sulfate concentrations in the pore waters from Site 1253 are relatively uniform (27.2–28.6 mM, with no clear depth variation) and nearly of seawater composition. This contrasts with values of 12–20 mM measured higher in the section at Site 1039. These gradients are in directions opposite to those expected for most biogeochemical and fluid/rock reactions in deep siliceous and calcareous sediments at low and elevated temperatures, which would be expected to reduce sulfate and to release Si, Ca, and Sr while consuming Mg and Li. The gradients observed at Site 1253, like those at Site 1039, suggest communication with a fluid of nearly seawater composition at depths below those from which interstitial waters have been recovered. At Site 1039, residence time calculations based on Sr and Li isotopes and concentrations (Silver et al., 2000) indicate that the gradients toward seawater are supported by flow within the last 15–20 ka. The gradients at Site 1253 are closely similar to those at Site 1039, supporting an argument for recent flow here also, which may have extracted heat from the plate to produce the unusually low heat flow in this region. The nature of this large regional-scale flow system, presumed responsible for the large heat flow anomalies as well as the chemical gradients, remains enigmatic; the CORK-II was installed at this site in hopes of providing necessary new information for better understanding the flow system.

Igneous Stratigraphy

The petrology of the igneous units can be combined with paleomagnetic and rock magnetism studies and logging results to better understand the nature of the two units and the internal structure of Subunit 4B (Fig. F15). Paleomagnetic results show that the sill (Subunit 4A) spans several polarity reversals, implying multiple pulses of magma intrusion, although the elapsed time cannot be evaluated until age dating is completed postcruise. Magnetic intensity is highest at the top of the sill, indicating more stable magnetization, probably because of the presence there of finer-grained magnetite than at deeper levels in the sill. The petrologic boundary between the two subunits approximately corresponds with a polarity reversal boundary. Subunit 4A-2 is composed entirely of microcrystalline gabbro. Logging results show a large hole diameter at the top of Subunit 4A-2, which corresponds to the 14³/₄-in hole drilled to provide a rathole for the casing installation. Seismic velocity and shallow resistivity (considered more reliable in the igneous units; see "Downhole Logging") are relatively high and uniform, and the cores recovered are massive in appearance, breaking into large pieces. At the base of the sill, recovery drops, the hole size increases, and velocity and resistivity decrease in general and exhibit a more spiky character, suggesting that fractured rock is present or possibly thin (<1 m) sediment interlayers.

The lower igneous subunit (Subunit 4B) begins at a depth of 450 mbsf in the core reference frame, which was used for petrologic and paleomagnetic work, and at ~460 mbsf in the logging data. A depth of 460 mbsf for the top of Subunit 4B is considered more reliable, given the very low recovery at the top of the subunit and the standard curatorial practice of moving all recovered material to the top of the core. Subunit 4B was subdivided into seven subunits, using the same criteria used for Subunit 4A. Within each subunit, multiple alternations between microcrystalline and fine-grained material may indicate the presence of multiple cooling units. Subunits 4B-1 through -3 all formed during what may be a single reversed polarity interval, although dating is required. Magnetic intensity is again high at the top of Subunit 4B, and decreases with depth. The logging data show that Subunits 4B-1 through 4B-3 are characterized by high, and relatively uniform, resistivity and P-wave velocity. There is a marked increase in resistivity at the top of Subunit 4B-4, which corresponds to a short massive interval that was drilled very slowly (0.75 m/hr) with high recovery. In this interval, conductive features are rare in the FMS data (see "Downhole Logging"). There is a hint of increased P-wave velocity at and below this interval, seen in the logging data and as measured in the cores. From ~490 mbsf to the base of the logged section, the borehole character become more heterogeneous, with intermittent highs and lows in resistivity and seismic velocity. At 508 mbsf, FMS images change to more closely spaced conductive features, which are continuous across all four pads at rare intervals, such as 513–514 mbsf. This is an interesting depth, as it corresponds to a thin layer of rock with true basaltic texture and a return to high magnetic susceptibility, similar to that seen at the top of the sill and the top of Subunit 4B, and interpreted as indicating single domain (<100 multiple periods of magmatic activity. The MST natural gamma measurements on the core suggest increased K, U, or Th concentrations in this lower part of Subunit 4B. Glass is more abundant below this depth, discrete and diffuse alteration is more extensive, and carbonate-bearing veins are present. Despite these differences, the generally microcystalline and fine-grained material below this depth share many textural, mineralogical, and chemical similarities to the overlying sections.

CORK-II Installation

Details of the CORK-II installation in Hole 1253A are shown in Figure F22 with the petrological and structural character of key depths as shown in Figures F15 and F19. The center of the packer was set at ~473 mbsf, with the inflatable element being between 471.5 and 475.5 mbsf. The cores indicate that this is an interval of high recovery of massive rock with relatively few fractures. The logging results (see "Downhole Logging") show this to be in an area of relatively uniform physical properties (high resistivity, bulk density, and P-wave velocity). Interpretation of FMS images indicates a massive-blocky formation, with 0.5- to 1-m spacing between thin conductive features, which can be traced across the four pads. The upper OsmoSampler, located inside a 7.35-m-long screen in the 4504 mbsf (Fig. F19). A 2-m pressure screen is located within the casing screen, and a fluid sampling line runs from this screen to the CORK-II wellhead. Figure F19 shows this to be an interval of modest recovery of moderately fractured rock composed of alternating microcrystalline and fine-grained material (see "Igneous Petrology"). Logging data in Figure F15 show this to be an interval of generally uniform hole diameter, with minor variations in bulk density and Pwave velocity. FMS images show closely spaced conductive features. The lower sampler is dangled in the open hole between 512.1 and 519.5 mbsf. This is again a zone of moderate recovery and fracture density in a cryptocrystalline (basaltic) to microcrystalline part of the section, with relatively high concentration of voids and 10% to locally 50% secondary mineral formation. The logging data (Fig. F15) show this to be an interval of decreased resistivity and sonic velocity and variable hole diameter. In the upper part of this interval, FMS images show closely spaced (<0.5 m) shallowly dipping conductive features that are continuous and can be traced across the four FMS pads. The intervals for the osmotic samplers were chosen using a combination of scientific and operational constraints. Originally, the intervals between 513–521 (now OsmoSampler 2) and 560–568 mbsf were targeted, where the latter is a zone of high fracture density and maximum alteration in largely microcrystalline rock. However, the bridge encountered by the logging tools at 530 mbsf restricted the OsmoSampler deployment to shallower levels. The upper pressure screen, located above the packer, was set into the sediments between the two igneous subunits, where sediments collapsing around the screen are expected to make an effective seal. The final installed configuration for this modified CORK-II geochemical and hydrologic borehole observatory is shown in Figure F22.