IODP Proceedings    Volume contents     Search


Lithostratigraphy and petrology

This section outlines the procedures used to document the composition, texture, and sedimentary structure of geologic materials recovered during Expedition 334. These procedures included visual core description, smear slide and petrographic thin section description, digital color imaging, color spectrophotometry, and X-ray diffraction (XRD). Because many of the geologic techniques and observations used to analyze sedimentary cores are similar to those used to analyze igneous cores, the methods are presented together. However, as conditions warranted, different procedures were used to characterize sedimentary and crystalline basement rocks. In instances where the description protocols differ, both are described in detail.

All data were uploaded into the IODP-USIO Laboratory Information Management System (LIMS), and observations were entered using the DESClogik application in Tabular Data Capture mode. Additional details are provided below. A glossary of common geological terms used to describe the basement cores can be found in Table T1.

Core sections available for sedimentary, petrographic, and structural observation and interpretation included both the working and archive halves. Sections dominated by soft sediment were split using a thin wire held in high tension. Pieces of hard rock were split with a diamond-impregnated saw such that important compositional and structural features were preserved in both the archive and working halves. The split surface of the archive half was then assessed for quality (e.g., smearing or surface unevenness) and, if necessary, scraped lightly with a glass slide or spatula. After splitting, the archive half was imaged by the SHIL and then analyzed for color reflectance and magnetic susceptibility using the SHMSL (see “Physical properties”). In rare instances, the archive-half was reimaged when visibility of sedimentary structures or fabrics improved following treatment of the split core surface.

Following imaging, the archive sections of the sediment core were macroscopically described for lithologic, sedimentary, and structural features. Lithostratigraphic units were characterized by visual inspection, and smear slide samples were used to determine microfossil and sedimentary constituents and abundances to aid in lithologic classification. Cores from the basement were also described visually and subsequently with the aid of thin sections for primary and secondary igneous features. All descriptive data were entered into DESClogik using the data capture software. Based on preliminary visual descriptions and physical properties data, thin section samples and samples for XRD were extracted from the working half. All descriptions and sample locations were recorded using curated depths and then recorded on standard graphic report forms (barrel sheets) and documented on visual core description (VCD) graphic reports (Fig. F1).

Visual core descriptions for sediment

Color and composition

Color was determined qualitatively for core intervals using Munsell Color Charts (Munsell Color Company, Inc., 2000). Visual inspections of the archive halves were used to identify compositional elements of the sediment, including concretions, nodules, chert, and tephra.

To emphasize differences in the composition of volcanic sandstones, the rocks were classified using a scheme developed by Fisher and Schmincke (1984). In general, coarser grained sedimentary rocks (63 µm to 2 mm average grain size) are designated as “sand(stone),” where the volcaniclastic components were <25% of the total clast content the sample is termed a volcaniclastic rock. Volcaniclastics can be both (1) reworked and commonly altered heterogeneous volcanic material like lava, tuff, and tephra and (2) fresh, or less altered, compositionally homogeneous loose pyroclastic material resulting directly from explosive eruptions on land or effusive/explosive vents on the seafloor.

If there are ≥25% volcaniclasts but <25% pyroclasts in the sediment/sedimentary rock, it is designated as a “volcaniclastic sand(stone).” If the clast composition is 25%–75% pyroclasts, the sediment/sedimentary rock is classified as “tuffaceous sand(stone),” but if clast composition is ≥75% pyroclasts, the rock/sediment is classified as a “tuff” or “ash.” Depending on grain size and degree of compaction, the nomenclature is adjusted accordingly, as shown in Table T2.

Textures, structures, and sedimentary fabric

When visible at low magnification, sediment grain size was determined using the Wentworth scale (Wentworth, 1922). Grain size, particle shape, and sorting were also noted; however, these textural attributes required inspection at high magnification and were only described on smear slides and thin sections (see below).

Sedimentary structures observed in recovered cores included bedding, soft-sediment deformation, bioturbation, and early diagenetic mineral formation. Bed thickness was defined according to Boggs (2006) and included the following units:

  • Very thick bedded = >100 cm.

  • Thick bedded = 30–100 cm.

  • Medium bedded = 10–30 cm.

  • Thin bedded = 3–10 cm.

  • Very thin bedded = 1–3 cm.

  • Laminae = <1 cm.

Some samples were inspected with a 10× hand lens for micrograded bedding (i.e., graded bedding occurring within laminations) and indications of preferred particle orientation, including lineation and imbrication of elongated detrital and biogenic material.

The abundance of bioturbation is constrained using the semiquantitative ichnofabric index as described by Droser and Bottjer (1986, 1991) and the thickness. The indexes refer to the degree of biogenic disruption of primary fabric, such as lamination, and range from 1 for nonbioturbated sediment to 6 for total homogenization:

  • 1 = no bioturbation recorded; all original sedimentary structures preserved.

  • 2 = discrete, isolated trace fossils; up to 10% of original bedding disturbed.

  • 3 = approximately 10%–40% of original bedding disturbed. Burrows are generally isolated, but locally overlap.

  • 4 = last vestiges of bedding discernible; approximately 40%–60% disturbed. Burrows overlap and are not always well defined.

  • 5 = bedding is completely disturbed, but burrows are still discrete in places and the fabric is not mixed.

  • 6 = bedding is nearly or totally homogenized.

The ichnofabric index in cores was identified with the help of visual comparative charts (Heard and Pickering, 2008). Distinct burrows that could be assigned to specific ichnotaxa were also recorded.

Smear slides and thin sections

Smear slides and thin sections are useful for identifying and reporting basic textural and compositional attributes, even if the results are only semiquantitative. We estimated the texture of the sediments with the help of a visual comparison chart (Rothwell, 1989). However, errors can be large, especially for the fine silt- and clay-size fractions. Smear slide analysis also tends to underestimate the amount of sand-size grains because sand-sized particles are difficult to incorporate evenly onto the slide. Nevertheless, in order to define unit boundaries and subunits, smear slides are a fast and efficient way to evaluate differences in lithology, texture, and composition, as constrained by point counting.

Point counts were conducted in four defined areas at 100×–200× magnification. This method works well for sandstones and sandy siltstones, and the theoretical 2σ error for a total of 200 counted particles is between 3% and 7%, depending on the portion of the total inventory (van der Plas and Tobi, 1965). The various components were then binned into several categories (e.g., feldspar, pyroxene, sedimentary lithics, etc.) to facilitate optimum reproducibility among different scientists.

For fine sediments (silt and silty claystones), rough estimations were made regarding the matrix using the visual comparison chart of Rothwell (1989). These estimates were supplemented with point counting of the coarser fraction. The coarse and fine fractions were subsequently combined and normalized to 100% of the total component inventory. Because most of the point counting is based on estimates, the 2σ error for the finer sediments is much higher and must be considered when interpreting the results; the overall trends, especially for the coarser grains, are considered more reliable. We normalized the data for the finer grained sediments against three principal classes. For example, quartz or feldspar, as well as volcaniclastic or sedimentary lithic content, were normalized to total mineral and total lithic contents, respectively. Additionally, the inventory of tephra layers, pods, and dispersed ash layers was documented separately to account for the presence of glass shard textures and mineral composition. This method cannot be applied to tuffs because many fine glass shards are destroyed during smear slide preparation.

Results are summarized in the smear slides (see “Core descriptions”). The relative abundance of major components was also validated by XRD (see “X-ray diffraction”), and the absolute weight percent of carbonate was verified by chemical analysis (see “Geochemistry and microbiology”).

The sample location of each smear slide was entered into the DESClogik system with a sample code of “SS,” using the Samples application. The position of each specimen is shown on the VCD slide editor column of the VCD application.

X-ray diffraction

We completed routine XRD analyses of bulk powders using a Bruker D-4 Endeavor diffractometer mounted with a Vantec-1 detector using nickel filtered CuKα radiation. Our principal objective was to determine the mineral phases present using identified peaks. Most of the samples were selected from intervals adjacent to whole-round samples, and most are part of sampling clusters with physical properties and carbonate. A few additional samples were collected periodically from such unusual lithologies such as carbonate-cemented claystone and volcanic ash. Samples were freeze-dried, crushed with a ball mill, and mounted as random bulk powders. Standard locked coupled scan conditions were

  • Voltage = 40 kV,

  • Current = 40 mA,

  • Goniometer scan 2θ = 5°–70°,

  • Step size = 0.015°,

  • Scan rate = 0.1 s/step, and

  • Divergence slit = 0.3 mm.

The upper and lower limits of each peak on the diffractogram were adjusted following the guidelines shown in Table T2. Calculations of relative mineral abundance used a matrix of normalization factors derived from integrated peak areas and singular value decomposition (Table T3). As described by Fisher and Underwood (1995), calibration of singular value decomposition factors depends on the analysis of known weight percent mixtures of mineral standards that are appropriate matches for natural sediments (Fig. F2).

In the final assessment, calculated values of a mineral’s weight percent should only be regarded as relative percentages within a four-component system where clay minerals + quartz + plagioclase + calcite = 100%. How close those estimates are to the absolute percentages within the mass of total solids will depend on the abundance of amorphous solids (e.g., biogenic opal and volcanic glass), as well as the total of all other minerals that occur in minor or trace quantities. For most natural samples, absolute errors are probably between 5% and 10% (Fisher and Underwood, 1995). Thus, the primary value of bulk powder XRD data should be to identify spatial and temporal trends in sediment composition and to assist with core-log integration.

Basement description

In order to preserve important features and structures, all cores were visually examined before being split. Large pieces of basement core were marked on the bottom with a red wax pencil to preserve orientation when they were removed from the split core liner. In many cases, pieces were too small to be oriented with certainty. Each piece was numbered sequentially from the top of each core section and labeled on the outside surface. Broken core pieces that could be fit together along fractures were assigned the same number and lettered consecutively from the top down (e.g., 1A, 1B, and 1C). Plastic spacers were placed between pieces with different numbers. The presence of a spacer may represent a substantial interval without recovery. Fitted core pieces were not glued, but wrapped together into plastic foil.

Characterization of the basement is based on visual core description and thin section analyses. Morphology of magmatic contacts, veins, lava, and porphyroclastic flows were initially identified on the whole-round cores and further described and measured after splitting the cores. A hand lens (10×) and binocular microscope helped to distinguish between the different types of sedimentary and igneous rocks. Petrographic texture, grain size, and phenocryst mineralogy as well as modal abundance were determined using thin section analysis. When describing and assigning a name to a rock interval, cores were divided based on changes in mineralogy, texture, grain size, composition, the occurrence of chilled margins, or tectonic contacts.

VCD forms were used to describe each basement core. A key summarizing the symbols used on the hard rock VCDs is given in Figure F3. On the VCDs, the following information is displayed from left to right:

  • Section scale from 0 to 150 cm;

  • Piece number;

  • Photograph of the archive half of the core;

  • Piece orientation;

  • Lithology symbol;

  • Phenocrysts abundance, if igneous;

  • Phenocrysts minerals, if igneous;

  • Alteration intensity;

  • Information about abundance and filling of veins;

  • Description of primary and secondary magmatic structures in the rock; and

  • Any additional comments.

In instances where an individual rock fragment could not have rotated about a horizontal axis during drilling, an arrow pointing toward the top of the section was inserted into the Piece orientation column. The term “vein” describes any filled, crosscutting fractures, which include breccia-filled fractures, epigenetic mineralized veins, shear veins, and vein networks. Vein geometry and mineralogy is given in detail in the Additional comments column. Comments also include accessory mineral occurrence.

Thin sections were studied to complete and refine the hand specimen observations. In the case of igneous rocks, textures were defined at the microscopic scale according to the degree of crystallinity of the groundmass (holohyaline to holocrystalline). A visual estimate of the modal abundance of the phenocrysts was made, average crystal size for each mineral phase was determined, and the mineral shape and mineral habit were also determined using thin sections. The textural terms “euhedral,” “subhedral,” “anhedral,” and “interstitial” are used to describe the habit of crystals. Grain shape was divided into four classes:

  • Equant (aspect ratio = <1:2),

  • Subequant (aspect ratio = 1:2 to 1:3),

  • Tabular (aspect ratio = 1:3 to 1:5), and

  • Elongate (aspect ratio = >1:5).

Basalt and all other rocks exclusively occur within the cored units that have been ascribed to the transition zone between upper plate sediment and basement arcward of the trench. Here, the constituent blocks of this transition zone are clasts composed of basalt, breccia, limestone, radiolarian chert, and argillaceous matrix material. Sediments in the basement are considered to be lithics and are therefore described in the same manner as the sedimentary cover. Outboard of the trench, the predominant basement rocks are basalt, which are distinguished on the basis of matrix appearance; the presence of phenocryst minerals; alteration state; and overprinting by intrusions, faulting and veining.

All classifications together with the basement core descriptions and associated shipboard analyses were archived electronically in DESClogik. DESClogik is a VCD program that stores visual (macroscopic and/or microscopic) description information about core structures at a given section index. Figure F3 displays graphic patterns for all basement lithologies encountered during Expedition 334.