LITHOSTRATIGRAPHY

Description of sedimentary deposits recovered during Leg 192 generally followed standard ODP methodology. Thus, most procedures used are only briefly outlined, but, where we depart significantly from ODP conventions, we provide a more detailed explanation. Procedures for the description of igneous rocks are described in "Igneous Petrology," "Alteration," and "Structural Geology".

We described each core in detail and entered our observations into AppleCORE (version 8.1f) software. For cores with considerable variation at the meter scale, we first recorded our observations on visual core description (VCD) forms and then condensed and summarized these observations for entry into AppleCORE. AppleCORE generates a simplified, annotated graphical description (barrel sheet) of each core. Barrel sheets are presented alongside corresponding core photographs (see the "Core Descriptions" contents list). AppleCORE also creates a data file of depth-facies information for use in the ODP Janus database.

The lithologies of the sedimentary intervals recovered are represented on barrel sheets by symbols in the column headed "Graphic Lithology" (Fig. F3). Classification follows the scheme of Mazzullo et al. (1988) with minor modifications as outlined below. Bed thickness is characterized as very thick bedded (>100 cm thick), thick bedded (30-100 cm thick), medium bedded (10-30 cm thick), thin bedded (3-10 cm thick), and very thin bedded (1-3 cm thick) (McKee and Weir, 1953). Grain-size divisions (i.e., sand, silt, clay, etc.) are those of Wentworth (1922). Sediments are represented by a single graphic lithology pattern if they are relatively homogeneous across the interval described. In homogenous sediments with two or three dominant components not described by a single standard lithology pattern, strips are plotted showing the relative abundance of each component.

The stratigraphic distribution of geological features such as degree of bioturbation, primary sedimentary structures, soft-sediment deformation, accessory minerals, discrete trace fossils, and diagenetic features are indicated schematically in columns to the right of the graphic log. Specific stratigraphic information concerning important or unusual discrete structures is listed in the description column. In addition, VCD forms containing the detailed observations can be obtained from ODP. Figure F4 contains a key to the full set of symbols used on the graphic columns.

Features related to sampling, analysis, and processing of the cores are indicated on the barrel sheets using symbols defined in Figure F4. Locations of samples taken for shipboard lithostratigraphic analysis (smear slides, thin sections, carbonate analysis, etc.) are indicated in the "Sample" column, with analysis types represented by the codes listed in Figure F4.

The description column summarizes the lithologies observed and the approximate geologic age provided by shipboard paleontological studies. The description generally includes key characteristics of all the major and important minor sediment lithologies (e.g., color, composition, sedimentary structures, and visible trace and body fossils). We estimated general sediment color trends by comparison to Munsell Soil Color Charts (1975). Detailed color data were quantified with an automated color reflectance spectrophotometer (see "Color Reflectance Spectrophotometry").

Lithologic names consist of a principal name based on composition, degree of lithification, and/or texture. For lithologies with a mixture of components, the principal name is preceded by major modifiers (in order of increasing abundance) that refer to those components making up >25% of the sediment. Names of minor components that represent between 10% and 25% of the sediment follow the principal name after a "with," in order of increasing abundance. Thus, an unconsolidated sediment containing 30% nannofossils, 25% clay minerals, 20% foraminifers, 15% quartz silt, and 10% ferromanganese nodules would be described as a clayey nannofossil ooze with ferromanganese nodules, quartz silt, and foraminifers. These naming conventions follow the ODP sediment classification scheme (Mazzullo et al., 1988), with the exception that during Leg 192 we did not distinguish a separate mixed-sediment category (Fig. F5). We did not encounter neritic shallow-water carbonate sediments or chemical sediments except as accessory minerals and do not address these categories below.

Granular sediments are subdivided on the basis of composition and abundance of different grain types estimated from visual examination of the core, smear slides, thin sections, and by shipboard measurements of carbonate content (see "Inorganic Carbon") and shipboard XRD analyses (see "X-Ray Diffraction"). Grain-size divisions are those of Wentworth (1922). For volcaniclastic sediments the term "ash" ("tuff" if lithified) is used in place of sand, and the term "lapilli" is used for granule and cobble size categories. Larger volcanic clasts (breccia) were not encountered. Size-textural qualifiers were not used for pelagic sediment names (e.g., nannofossil clay implies that the dominant component is siliciclastic clay rather than clay-size nannofossils). For thin section analysis of microfacies, we used the textural classification scheme of Dunham (1962). We classified matrix-supported rocks as "mudstone" if they contain <10% and "wackestone" if they contain >10% grains. Grain-supported rocks are classified as "packstone" if clay and silt are present or "grainstone" if silt and clay are absent.

Terms that describe lithification vary depending upon the dominant composition:

1. Sediments derived predominantly from calcareous pelagic organisms (e.g., calcareous nannofossils, foraminifers): the lithification terms ooze, chalk, and limestone reflect whether the sediment can be deformed with a finger (ooze), can be scratched easily by a fingernail (chalk), or cannot be scratched easily (limestone).
2. Sediments derived predominantly from siliceous microfossils (diatoms, radiolarians, and siliceous sponge spicules): the lithification terms ooze, porcellanite, and chert reflect whether the sediment can be deformed with a finger (ooze), cannot be easily deformed manually (porcellanite), or displays a glassy luster (chert). Note that the terms porcellanite and chert do not indicate the mineralogy of the silica.
3. Sediments derived predominantly from siliciclastic material: if the sediment can be deformed easily with a finger, no lithification term is added and the sediment is named for the dominant grain size; for more consolidated material the lithification suffix "-stone" is appended to the dominant size classification (e.g., clay vs. claystone).
4. Sediments composed of sand-size volcaniclastic grains: if the sediment can be deformed easily with a finger, the interval is described as ash; for more consolidated material, the rock is called tuff. The term lapilli is used for clasts that are granule or cobble size, and no distinction is made based on degree of lithification.

The sediment names we applied were based largely on analysis of thin sections and smear slides. Tables summarizing data from thin sections and smear slides are included (see the "Core Descriptions" contents list). These tables include information about the sample location, whether the sample represents a dominant (D) or a minor (M) lithology in the core, and the estimated abundance of different grain sizes and different grain types. Although we tried to ensure consistency among observers and through time, thin section and smear slide analyses on the ship are qualitative; percentages listed are visual estimates of the relative abundances. The mineral identification of finer grained particles can be difficult using only a petrographic microscope, and the abundance of sand-size grains tends to be underestimated in smear slides because they are often incorporated into the smear unevenly. Classification of limestone facies generally requires a thin section. Additional information on composition was obtained by XRD analysis (see "X-Ray Diffraction") and measurement of carbonate content (see "Inorganic Carbon").

For selected samples, the relative abundances of the main silicate and carbonate minerals were determined with a Philips model PW1729 X-ray diffractometer using Ni-filtered CuK radiation. Typical sampling frequency was one per core but was higher in intervals with two or more different lithologies. Each bulk-sediment sample was freeze-dried and crushed. The powder was mounted with a random grain orientation into an aluminum sample holder. Instrument conditions were as follows: 40 kV, 35 mA; goniometer scan from 2° to 70° 2 for bulk samples; step size of 0.01° 2; scan speed at 1.2° 2/min; count time of 0.5 s at each step. After converting peak intensities to values appropriate for a fixed slit width, we used an interactive software package (MacDiff 3.3.0 PPC) on a Macintosh computer to identify the main minerals. Most diffractograms were peak-corrected to match the main calcite peak at 3.035 Å. In the absence of both quartz and calcite, no peak correction was applied. Identifications are based on multiple peak matches, using the mineral database provided with MacDiff.

Peak areas were measured only to estimate calcite/dolomite ratios. In all other cases, minerals were grouped as major or trace components, depending on relative peak heights. Accurate estimation of the proportions of clay minerals, glass, or amorphous opal is impossible with the bulk samples used for analysis, although mixtures of these phases with calcite make up most of the analyzed samples. Relative abundances reported in this volume are useful for general characterization of the sediments, but they are not precise, quantitative data.

Carbonate carbon concentrations were measured using a Coulometrics 5011 CO₂ coulometer. Typical sampling frequency was one per core but was higher in intervals with two or more different lithologies. The coulometer consists of a reaction vessel and a coulometer cell. The coulometer cell is filled with a proprietary solution containing monoethanolamine (ME) and a colorimetric indicator. A platinum cathode and a silver anode are placed in the cell and the cell assembly is positioned between a light source and a photodetector. For samples analyzed during Leg 192, ~10 mg of freeze-dried, ground sediment was acidified with 2-M HCL in the heated reaction vial. The CO₂ evolved is carried on a free air stream through a scrubbing system and into the coulometric cell, where the CO₂ is quantitatively absorbed and reacts with ME to form a titratable acid. Decreased pH causes the colorimetric indicator to fade. The photodetector monitors the change in the solution's color as percent transmittance. An increase in percent transmittance causes a current to be passed through the cell, thereby back-titrating the solution. When the solution returns to its original color, the current stops. The percentage of inorganic carbon (IC) is calculated from the current passed through the cell and the mass of sample analyzed; the weight percentage of carbonate is determined from the IC, with the assumption that all IC is present as calcium carbonate (CaCO₃):

CaCO₃ wt% = IC wt% × 8.332. (1)

In addition to visual estimates of the color, we routinely measured the reflectance of visible light from cores using a Minolta spectrophotometer (model CM-2002) mounted on the archive multisensor track (AMST). The AMST measures the archive half of each core section and provides a high-resolution stratigraphic record of color variations for visible wavelengths (400-700 nm). We covered freshly split cores with clear plastic wrap and placed them on the AMST, making measurements at a 4- or 5-cm spacing. The AMST skips empty intervals and intervals where the core surface is well below the level of the core liner, but the AMST cannot recognize relatively small cracks or disturbed areas of core. Thus, AMST data may contain spurious measurements that should, to the extent possible, be edited out of the data set before use. Each measurement recorded consists of 31 separate determinations of reflectance in 10-nm-wide spectral bands from 400 to 700 nm. Additional detailed information about measurement and interpretation of spectral data with the Minolta spectrophotometer can be found in Blum (1997), Balsam et al. (1997, 1998), and Balsam and Damuth (2000).