IODP Proceedings    Volume contents     Search

doi:10.2204/iodp.proc.337.103.2013

Lithostratigraphy

In general, drilling in Hole C0020A was divided into two phases of drilling operations. The first phase covered the depth interval from 636.5 to 1256.5 m mud depth below seafloor (MSF), from which only cuttings samples were retrieved, and the second drilling phase covered the depth interval from 1256.5 to 2466 m MSF, in which both core and cuttings samples were collected.

Cuttings observations

Cuttings samples from 636.5 to 926.5 m MSF were composed mainly of dark olive to olive-gray silty claystone with abundant diatoms, common volcanic glasses, common quartz, few feldspars, and few lithic fragments (Fig. F1). The first cuttings sample was highly contaminated by cement. Few palagonite and opaque minerals (e.g., pyrite or other iron sulfide) were identified in smear slides along with common sponge spicules and few shell fragments. Rare semiconsolidated fine sandstone fragments were visible in >4 mm cuttings samples. Semiconsolidated silt grains in smear slides were poorly sorted with moderate to subangular roundness.

Cuttings recovered from 926.5 to 1116.5 m MSF were dominated by diatom-bearing clayey siltstone with abundant lithic fragments, quartz, and feldspar. In some samples, volcanic glass and sponge spicules were common, whereas pyrite or other iron sulfides and shell fragments were present in only minor amounts. Semiconsolidated siltstone and fine sandstone in smear slides were subangular to rounded with moderate to poor sorting. Fragments of sandstone were identified in all cuttings samples. In addition, plant remains were rarely visible in cuttings. The sandstone content in the cuttings ranged between 5% and 15%. Colors observed in cuttings from this depth interval corresponded frequently to lithology. For example, well-cemented siltstone fragments were light in color (5Y 5/2) and clayey siltstone and clayey siltstone with fine sandstone (5Y 4/2) were dark olive.

Cuttings samples from 1116.5 to 1236.5 m MSF consisted mainly of medium silt to sand with very common wood fragments found most frequently between 1206.5 and 1226.5 m MSF (as much as 15% of the total material). Sand and silt grains in smear slides were rounded to moderate and were poor to moderately sorted. Siliceous microfossil remains were also common, often as broken fragments. Volcanic glass and iron sulfides were also present.

Only one cuttings sample (337-C0020A-92-SMW) was recovered from 1116.5 to 1186.5 m MSF. The material from this interval consisted of loose, fine to coarse sand grains and was highly contaminated with drilling mud. A rare amount of silty fragments were observed. Loose, fine to medium sand was present in samples from 1186.5 to 1236.5 m MSF. Moreover, lithic fragments, pumice, sulfides, and clayey siltstones were also observed. The amount of clayey silt increased slightly and became more consolidated with depth to a more semiconsolidated consistency (cuttings Samples 92-SMW through 95-SMW).

Cuttings samples from 1236.5 to 1256.5 m MSF consisted mainly of semiconsolidated clayey siltstone with fine sand. In clay-rich layers, diatom fossils were well preserved; however, in coarser fractions, fractured diatoms and shells along with volcanic glass were more frequently observed (Fig. F2A). Wood fragments were also observed, although less commonly than from 1186.5 to 1236.5 m MSF. Silt grains in smear slides were rounded to subrounded in poorly sorted material.

In cuttings samples between 1256.5 and 1506.5 m MSF, the clay fraction generally increased from 10%–30% to 50%, whereas the sand fraction varied between 0% and 30% over the same interval (Fig. F1). Higher sand fractions in the cuttings corresponded to the increased prevalence of sandstones at the same core depths. Between 1506.5 and 1696.5 m MSF, the clay fraction increased to 60%–75%, whereas the sand fraction varied between 5% and 35%. This interval demonstrated a transition from a silt-dominated grain size distribution to one that was bimodally dominated by clay and silt. Between 1696.5 and 1826.5 m MSF, clay generally decreased to 30%, silt increased to 50%, and sand varied between 5% and 30%. Wood fragments were observed in most samples between 1186.5 and 1826.5 m MSF in abundances classified as rare or few (Fig. F1). In some sand-rich intervals, common wood and lignite fragments were observed. In general, smear slides were dominated by clay minerals, quartz, lithic fragments, and volcanic glass. Feldspar, mica, ferromagnesian silicates, and organic matter were also frequently observed in this interval. Smear slide observations indicated that the amount of biosiliceous material decreased with depth over this interval, with diatoms and sponge spicules present only in rare to few amounts below 1366.5 m MSF, with no diatoms observed below 1786.5 m MSF (see “Cuttings contamination”). In contrast, glauconite increased with depth from rare to absent above 1536.5 m MSF to common to dominant between 1886.5 and 1916.5 m MSF (Fig. F1).

Cuttings samples between 1826.5 and 2046.5 m MSF were dominated by fragments of coal. Silt content was relatively high and ranged between 40% and 70%, except for one sample with only 10% silt. Sand content is <25%. Glauconite was visible in almost all cuttings samples in this interval.

Olive-black silty shale is the most common lithology in cuttings samples between 2046.5 and 2396.5 m MSF. Fine sandstone laminations, few volcanic ash components, pumice, rare to common coal fragments, and some shell fragments are also present within the samples. Clay content is low at 2046.5 m MSF (~25%) and increases downhole (70%–80%). Clay content is nearly constant to 2246.5 m MSF and decreases downhole to 45% at 2396.5 m MSF. Glauconite is visible in varying amounts in smear slides. Quartz, lithic fragments, clay minerals, and volcanic glass are also observed in smear slides.

In cuttings samples between 2396.5 and 2466.5 m MSF, silty mudstone with fine sandstone containing authigenic carbonates was common to abundant. These authigenic carbonate fragments reacted slowly with hydrochloric acid and produced a yellow fluid during reaction, suggesting possible siderite. Smear slides from these samples also contained fine-grained siderite. Abundant lithic fragments, common quartz, few volcanic glass, rare glauconite, and few organic matter were observed in smear slides. Between 2426.5 and 2466 m MSF, clay and silt are the dominant size fraction in cuttings samples with dominant coal fragments.

Core observations

Cores 337-C0020A-1R, 2R, 5R, and 6R largely comprised fine to medium sandstone (Fig. F3A) and less common siltstone, whereas Cores 3R and 4R consist mainly of siltstone (Fig. F3). Plant debris and shell fragments were commonly observed. Planar laminations were common throughout these cores. A coarsening-upward interval and smaller fining-upward intervals were observed in Cores 5R and 6R. Consolidated siltstones showed higher X-ray CT values than semiconsolidated sandstones, corresponding to higher densities of consolidated siltstones. Pebbles display higher CT values than the surrounding sandstone. Clay, quartz, and lithic fragments are common to abundant in nearly every smear slide. Minor amounts of feldspar, volcanic glass, pyrite, mica, and glauconite were visible in some samples. Glauconite observed in these cores indicates reworking of original shelf sediment. Dolomitic cements and a few dolomitic concretions are sometimes also observed. Diatom-bearing mudstones are present in a very rare amount, possibly as contaminations from drilling mud (Fig. F2B).

Cores 337-C0020A-7R and 8L contained shale, conglomerate, siltstone, and sandstone (Fig. F3C). The conglomerate intervals contained rounded volcanic rocks, which were identified as gabbro and altered tuffite by examination of thin sections. Shells and plant fragments were commonly observed in these cores, along with planar and wavy laminations. Slight to heavy bioturbation was commonly observed in Core 8L. Section 8L-2 contained two carbonate nodules. Smear slides taken from these nodules contained dominantly dolomite. Both bioturbation and carbonate nodules are visible in the X-ray CT scan image. Quartz, clay, and volcanic glass are exhibited in nearly every sample of these cores. Lithic fragments, mica, and a few iron sulfides are observable in the smear slides.

Cores 337-C0020A-9R through 14R contained sandstone, siltstone, and shale. Planar, wavy, and cross-laminations were common throughout these cores, more frequently than in Cores 1R through 8L. Bioturbation was also more common in Cores 9R through 14R (e.g., Fig. F3E). X-ray CT scan images showed the morphology and orientation of the burrows that were horizontal and vertical. Mud clasts and upward-fining and upward-coarsening intervals were common. Glauconite was observed in sediments as faint green bands in cores and as sharp thin laminations such as in Core 14R. Sediments were generally rich in quartz. They showed a certain amount of clay minerals, lithic fragments, volcanic glass, and organic matter (Fig. F2). Olivine was present in some smear slides. A thin coal layer appeared in Core 14R, which may indicate the presence of more terrigenous organic material in the sequence. Lithogenic grains were subangular to subrounded and moderate to poorly sorted. In some depth intervals, siltstones were also aragonitic. Shell fragments and wood fragments were the most common fossils. These fragments were observed in intermittent layers throughout the sections. Parallel laminations were the main sedimentary structure observed in Cores 11R through 14R.

Cores 337-C0020A-15R through 25R were dominated by several coal horizons with intervening coaly shales, siltstones, and sandstones. The description of the coal was also based on X-ray CT images because different lithotypes were difficult to differentiate using the split surface of sections. Almost all coal horizons consisted of detritic to xylo-detritic coal with some layers of xylitic coal. Shaly coal was also observed. Water content, color, and vitrinite reflectance measurements of the coal suggested that the coal showed low maturity (Ro = 0.36%; see “Physical properties”). Amber and pyrite were visible in the coal layers. Pyrite occasionally developed as veins but also appeared more frequently as a series of disseminated crystals scattered within coal and coaly shale layers (Fig. F3G). The boundary between the underlying sediments and the coal above was represented most frequently by a gradational transition between organic-rich shale or less frequently by organic-rich sandstone and coal. Contacts between coalbeds and sediment above the coal were usually sharp and dominated by sand. These sandstones were fine to medium grained and often well sorted. Sediments between the coal horizons in Cores 15R through 19R consisted primarily of fine sandstone. Cores 20R and 21R were dominated by shale and siltstone, followed by a succession of fine sandstone, siltstone, and shale with coal horizons. In general, clay minerals, quartz, and lithic fragments were common in microscopic observations. Glauconite was rare in the entire interval (1959.5–1967.79 m core depth below seafloor, Method B [CSF-B]). Grains in smear slides were mostly moderate to subrounded and were moderately sorted, with a few well-sorted samples. Authigenic carbonate occurred as cement within Cores 15R through 25R (Figs. F2, F3). Bands of carbonate cementation and carbonate nodules that included siderite were visible. Glauconite was present in Cores 15R through 20R but was not observed in Cores 21R through 23R. The sandstones in Cores 21R through 24R have common olivine content.

Major sedimentary structures were observed visually as well as in the X-ray CT scans. Common structures in Cores 337-C0020A-15R through 25R were fining- and coarsening-upward sequences often associated with parallel and/or wavy laminations. Flaser and lenticular bedding and cross-lamination were only common in shale or silty intervals such as Core 21R.

Shells or marine fossiliferous material were observed only in Core 337-C0020A-21R. However, plant remains, present as wood fragments and organic-rich layers, were abundant in Cores 15R through 25R (Fig. F2E, F2F). Burrows, indications of bioturbation, were present in this interval. The shape and orientation of burrows were visible in the X-ray CT scans. In Cores 15R through 17R and 21R, horizontal burrows were present in sandstones. Horizontal burrows were also observed in shales and siltstones.

Cores 337-C0020A-27R and 28R were dominated by silty shale intercalated with thin siltstone layers in the cores. In Core 27R, shale was present as massive, homogeneous, and consolidated material. Grains were subangular to rounded and moderate to well sorted. Core 28R also contained fine to medium dolomitic sandstone intercalated with siltstone layers. Grains were moderate to poorly sorted. Abundant carbonate banding and nodules and glauconitic cement were observed in Core 28R. Carbonate bands and nodules were visible in both X-ray CT scan images (as high CT values) and split core sections (Fig. F3I). Figure F2G shows a photomicrograph from a smear slide of a carbonate band, which consisted of dominant siderite. Carbonate-cemented layers and nodules appear as shades of very light gray in the X-ray CT scan images. Plant remains and shell fragments were the most common fossils found. Gastropods were commonly observed in Core 27R, whereas only a few bivalves and sponge spicules were visible. Sediments of these cores were rich in quartz, clay minerals, and lithic fragments. Mica was visible in few amounts. Parallel laminations were the major sedimentary structure observed in both Cores 27R and 28R. Wavy laminations were found in Core 28R along with coarsening- and fining-upward successions and many carbonate bands. Bioturbation (i.e., vertical and horizontal burrows) was observed in Core 28R.

Cores 337-C0020A-29R through 32R consisted of fine to medium sandstone with minor shale layers associated with coalbeds and coaly material. The upper cores contained shale intercalated with fine sandstone and siltstone. Section 30R-2 contained one coal horizon with an organic-rich shaly parting (shaly or sandy layer within a coal horizon). Sandstone was often intercalated with thin carbonaceous siltstone and shale layers, and organic and coal material were common in these cores. Grains were subangular to well rounded and moderate to well rounded. Sandstone layers were often dolomite and glauconite rich (Fig. F2I). Lithic fragments were also a dominant component in the smear slides. Few bivalves were observed. Lenticular and wavy bedding along with parallel laminations and coarsening- and fining-upward successions were common.

Scanning electron microscopy

Observations of selected coal and carbonate layers from Cores 337-C0020A-18R, 29R, and 30R were performed on board with a scanning electron microscope with energy dispersive spectrometry (Fig. F4). Pyrites, which were mostly framboidal, siderite, dolomite, and naturally formed barite particles were observed in Core 18R. The presence of barite particles, however, needs further verification because of a possibility of barite contamination from the riser drilling mud. The structure of FeS2 in the coal of Core 18R showed marcasite.

Mineralogical and geochemical analyses

Semiquantitative X-ray diffraction

Cuttings

Semiquantitative X-ray diffraction (XRD) analysis provided a measure of downcore changes in the relative abundance of four major mineralogical components (i.e., total clay, quartz, plagioclase, and calcite) and was not an absolute percentage of the total sediment, as it did not include some major components (e.g., volcanic glass and biogenic silica) and minor components that generally comprise <10% of marine sediments (e.g., sulfides, oxides, and ferromagnesian silicates). However, semiquantitative XRD measurements provided a good measure of relative lithologic changes, such as the relative changes in the amount of sand and clay. Analysis of cuttings material between 636.5 and 1116.5 m MSF showed a relative dominance of total clay minerals (62%–78%) with lower, but significant, fractions of plagioclase and quartz (15%–20% and 13%–20%, respectively; Fig. F5). Between 950 and 1070 m MSF, relative abundance of quartz and plagioclase increased, which is in contrast to a decline in total clay mineral content at the same depths. The increase in quartz and plagioclase agreed well with the increase in sand content observed below 920 m MSF. Among this four-component semiquantitative analysis, clays, plagioclase, and quartz constitute almost all of the material, whereas calcite is below detection by XRD in each sample above 1116.5 m MSF. The low abundance of calcite over this depth interval as measured by XRD agreed well with the general lack of biogenic calcareous material observed in smear slides.

XRD measurements of cuttings samples below 1296.5 m MSF, which were measured at an approximate resolution of 50 m, support the overall downcore trends determined from macroscopic and microscopic observations of cuttings. The relative abundance of total clay decreased between 1331.5 and 1451.5 m MSF and then increased again to 1701.5 m MSF (Fig. F5) in a pattern that agreed well with the visual estimates of clay. Over the same interval, quartz and plagioclase followed a pattern inverse to total clay, which correlated well with the silt and sand content in the visual estimates. The relative abundance of calcite was below detection in all cuttings samples below 1446.5 m MSF, which was <5% of the four-component mixture in cuttings Sample 337-C0020A-136-SMW (1446.5–1456.5 m MSF).

Cuttings XRD measurements between 1696.5 and 2466.5 m MSF show 54%–73% total clay with 25%–30% quartz and 10%–15% plagioclase. These measurements demonstrate an appreciable decrease in plagioclase relative to cuttings samples above 1696.5 m MSF. XRD measurements in cuttings between 2046.5 and 2466 m MSF show a consistent pattern of 54%–73% total clay with 8%–37% quartz and 10%–31% plagioclase. Calcite was below detection in all cuttings samples across this depth interval.

Core

XRD measurements from core samples showed more pronounced downcore variation between samples than in cuttings because of the higher resolution and discrete sampling that captured individual lithologies (Fig. F5). XRD measurements provided semiquantification that confirmed the visual and smear slide descriptions of lithologies. Total clay content increased systematically from sandstone to siltstone to silty shale to shale. Likewise, plagioclase content systematically decreased between sandstone, siltstone, silty shale, and shale. Quartz was generally higher in sandstone than in shale; however, there was no clear difference in relative quartz abundance between siltstone or silty shale and any other lithology. Calcite was generally below detection in siltstone and shale, present in amounts <5% in sandstone, or up to 47% in calcite-cemented sandstone.

Between 1276 and 1371 m CSF-B, quartz and plagioclase combined were ~60% of the four-component system and decreased between 1371 and 1604 m CSF-B, associated with an increase in total clay over the same depth interval from 20%–30% to >80%. This pattern generally matched that observed in the cuttings with a trend of increasing shale relative to sandstone with depth. Calcite in this interval was either below detection or 1% and 7% in two samples.

XRD measurements from Cores 337-C0020A-15R through 25R (1919–2002.345 m CSF-B) capture the variation between sandstone (17%–27% total clay and 26%–56% quartz) and shale (70%–82% total clay and 18%–31% quartz) (Fig. F5). As in the cuttings samples, plagioclase was lower when compared to Cores 1R through 14R. Two samples, from very hard carbonate-cemented sandstone, contained calcite relative abundance of 23% and 47%.

XRD samples between 2002 and 2462 m CSF-B in Cores 337-C0020A-25R through 29R were composed of 60%–80% total clay with 5%–25% each of plagioclase and quartz, in agreement with analysis of the cuttings samples. Increases in quartz and plagioclase to 25%–75% were observed in Cores 30R through 32R and associated with an increased prevalence in sandstone. As observed in the cuttings samples, plagioclase is lower relative to quartz compared to the shallower cores. In Cores 25R through 32R, calcite was below detection in all samples except for one sample from carbonate-cemented sandstone, which contained 13% calcite.

Qualitative X-ray diffraction

Cuttings

Qualitative analysis of °2θ peaks in diffraction patterns showed a broadly consistent mineral assemblage, which supports the selection of the four components used in semiquantitative analysis. Quartz, plagioclase, pyrite, and illite/muscovite were present in all or the vast majority of the cuttings samples. No specific clay mineralogical analyses were conducted; therefore, clay mineralogy was only presented in a general discussion, as bulk analysis presents significant errors in analysis of specific clay minerals. Illite was the most common clay mineral present in the cuttings, and has an overlapping peak with muscovite in bulk analysis. In addition, peaks associated with chlorite and smectite group minerals were commonly observed. This agrees well with the regional assessment that illite, chlorite, and smectite clays are dominant along the northern Japan margins (Rateev et al., 1969) and further confirmed by Deep Sea Drilling Project (DSDP) and Ocean Drilling Program (ODP) drilling near the Shimokita region (Kurnosov et al., 1980; Mann and Müller, 1980; Shipboard Scientific Party, 2000). The dominant presence of these minerals suggests that the illite-smectite-chlorite clay, quartz, and plagioclase contained in the standards used for semiquantification in this analysis that were originally selected for the Nankai Trough (e.g., Underwood et al., 2003) are appropriate for semiquantitative analysis of Shimokita sediments.

Peaks associated with pyrite were observed in nearly every sample above 1336.5 m MSF. Above 1056.5 m MSF, clay minerals and micas (usually illite and muscovite) appeared regularly, but clays and micas were observed less frequently, especially in the deeper sediments. Plagioclase was the dominant feldspar, with less frequent occurrence of alkali feldspars; however, alkali feldspar peaks were frequently observed between 691.5 and 741.5 m MSF and between 1801.5 and 2351 m MSF. Cuttings below 920 m MSF were also marked by the appearance of peaks associated with dolomite, likely authigenic carbonate. This is consistent with observations in several smear slides (e.g., 337-C0020A-129-SMW through 131-SMW). Peaks associated with cristobalite were commonly observed, possibly present with volcanic glass or from conversion of amorphous biosiliceous material during drying.

Core

XRD analysis from cores was largely similar to the cuttings, with quartz, plagioclase, illite, and muscovite observed in almost every core sample. Alkali feldspar peaks were frequently observed, correlating with the increased presence of alkali feldspars observed in cuttings samples from the same intervals. Chlorite peaks were only observed in Cores 337-C0020A-15R through 20R and 23R through 24R. Peaks of siderite, dolomite, calcite, or aragonite were observed frequently in samples from Cores 1R, 16R, 19R through 22R, 24R, 26R through 29R, and 32R in intervals near observed carbonate nodules, bands, or cement. Pyrite was commonly observed, especially in Core 15R and shallower. Pyroxene minerals were frequently observed in samples from Cores 13R through 32R. A minority of XRD samples contained peaks associated with amphiboles, olivine, or heavy minerals. Table T1 shows the summary of qualitative XRD data from cores.

X-ray fluorescence

Cuttings

Many samples contained >60% SiO2 and, together with Al2O3, decrease continuously between 824 and 1050 m MSF (Fig. F6). This decrease was associated with the simultaneous decrease of clay and diatoms over this interval, as observed in smear slides and XRD (Fig. F5). High SiO2 in Hole C0020A was influenced by the high silica content of diatoms and quartz, with additional silica from plagioclase, and clay minerals. High values of Si/Al (a proxy for sands because sands are rich in silica and clays are rich in aluminum) were consistent with high values of quartz from the XRD data. This was in good agreement with our visual description of cuttings that also showed sand between 636.6 and 1236.6 m MSF. The high biogenic silica content can increase Si/Al, adding uncertainty to the use of Si/Al as a sand proxy; however, the incorporation of X-ray fluorescence (XRF) Si/Al with observations such as sand/silt content from visual observations or XRD data (Figs. F1, F4) suggests this proxy is effective in this environment.

Calcium values were consistently low above 1106.5 m MSF, apart from one value at ~1050 m MSF that indicated possible authigenic carbonate. Dolomite is thought to have first-order control on Mg/Al values (Limmer et al., 2012). Given that dolomite is present in multiple cuttings samples (e.g., Sample 337-C0020A-74-SMW [1056.5–1066.5 m MSF]), the high Mg/Al value could be linked to the authigenic precipitation of dolomite in cuttings samples. Mg/Al also showed a strong positive correlation with both CaO and Ti/Ca. Authigenic carbonate nodules observed in cores near the Shimokita region during DSDP Leg 56 and ODP Leg 186 were primarily dolomitic (Okada, 1980; Ijiri et al., 2003). Therefore, increases in Mg/Al in Hole C0020A can indicate, and potentially identify, intervals in which authigenic dolomite was present in cuttings. However, one particularly high Mg/Al ratio was noted at 1050 m MSF corresponding to the appearance of the Mg-rich olivine (i.e., forsterite) (Fig. F6) in the qualitative XRD data (Table T2). The chemical index of alteration (CIA) (Nesbitt and Young, 1982), using the correction of Singh et al. (2005), in which CaO* is replaced by CaO/Na2O, was also measured. The CIA value showed a significant correlation with Mg/Al and CaO, suggesting a further proxy for carbonate weathering. The relationships between CaO, CIA, and Mg/Al have been confirmed by the cross-plots (Fig. F7).

Below 1106.5 m MSF, Si/Al is generally lower than above this depth, most likely as a result of the general decrease in diatoms and sponge spicules; however, variation within this unit serves as a sand proxy. In cuttings samples, Si/Al increases in intervals with higher sand content, and core samples collected from sandstone have increased Si/Al. There is a good correlation between XRF Si/Al and quartz content between 1296.5 and 1946.5 m MSF, although overall quartz content and Si/Al values are not statistically significant. K/Al values are lower in the shale-rich layers at the bottom of this depth interval, which is consistent with increased chemical weathering in the finer, more clay-rich sediments.

Between 1256.5 and 1826.5 m MSF, the CIA shows little fluctuation, with an increase in K/Al and Si/Al at ~1400 m MSF, where the CIA declines. This drop in Al2O3 matches an increase in the proportion of lithic fragments observed in the cuttings. Also, an increase in the CaO ratio is found above 1400 m MSF, where an increase in carbonate is observed in the cuttings samples. In cuttings, a consistent inverse relationship between CaO and the CIA is seen (Fig. F7).

Only four cuttings samples were analyzed between 1826.5 and 2046.5 m MSF, although they cover three different lithologies. The uppermost sample (337-C0020A-226-SMW) is a silty shale, and as a result has low K/Al, Si/Al, Ti/Al ratios. In contrast, Sample 242-SMW, taken from a coal, has the highest K/Al and Ti/Al values from the hole. Sample 254-SMW, taken from a coal-rich sand, has very low SiO2 and Al2O3 values.

Cuttings samples below 2046.5 m MSF show a sharp increase in K/Al and Si/Al between the hole bottom and ~2200 m MSF as the lithology becomes sandier. The remaining element ratios appear relatively stable, although sodium fluctuates, which also corresponds to aluminum above 2200 m MSF.

Core

In comparison to cuttings, the geochemistry record for cores is slightly more diverse between 1256.5 and 1826.5 m MSF. Many elements and ratios fluctuate in value, especially between 1600 and 1800 m CSF-B. This region is an area where faulting is observed and perhaps more importantly marks the first appearance of carbonate nodules within the sediments. The declines in K/Al and Si/Al are consistent with an increase in the relative proportion of shale within the sediment.

Core data from 1826.5 to 2046.5 m MSF show a relative increase in Ti/Ca and substantially lower K/Al values than the cuttings samples. Analysis of element values suggests this is due to fluctuations in aluminum, probably linking to the change between sand, silt, and coal. Also, one very high Mg/Al value was observed that is linked to a sharp increase in Fe/Al and Ca/Al. This value is associated with a siderite nodule observed in Core 337-C0020A-24R. Fe/Al and Mg/Al have a consistent positive statistical relationship because iron and magnesium both increase when aluminum declines. Mg/Al values increase consistently wherever dolomite is observed.

Both element values and ratios in cores closely follow the pattern of the cuttings. The differences occur at the bottom of the hole, where values for potassium, sodium, and silica are lower, resulting in a decrease in the K/Al ratio. Small increases in Ti/Ca, Ca/Al, and Mg/Al at 2400 m CSF-B are probably linked to more siderite production. The CIA value slightly declines from 77% to 70% during this interval.

Some elemental geochemical ratios have been used as proxies for chemical weathering based on the relative mobility of some chemical elements relative to others (Duzgoren-Aydin et al., 2002; Wei et al., 2006). For example, the K/Al ratio has been widely used as a proxy for chemical weathering, based on the principle that potassium is highly mobile in water and will therefore be lost under intense chemical weathering (e.g., Nesbitt et al., 1980; Derry and France-Lanord, 1996; Limmer et al., 2012). Therefore, a decreasing K/Al ratio in cores between 1256.5 and 2046.5 m CSF-B indicates more chemical weathering. A general increase of Ti/Ca is visible in cores between 1256.5 and 2046.5 m CSF-B. The Ti/Ca ratio has been applied as a proxy for the relative amount of terrigenous input, assuming that Ti is sourced from coarse-grained sediments. Overall, the only statistically significant correlations occur in proxies used for either carbonate weathering or carbonate precipitation. This means that the composition of sediments at Site C0020 is strongly influenced by these processes. Chemical weathering and terrigenous input show a minor influence on the proxies (Fig. F7).

Interpretation of units

Based on macroscopic and microscopic descriptions of cuttings and core samples, XRD, and XRF data, we defined four different lithologic units at Site C0020. To obtain a more precise interpretation, some of the units were divided into subunits. The units are summarized in Table T3.

Unit I

  • Interval: cuttings Samples 337-C0020A-25-SMW through 98-SMW

  • Depth: 647–1256.5 m MSF

Unit I consists primarily of diatom-bearing silty clay. The interval between 1116.5 and 1236.5 m MSF is an exception. The upper part (1116.5–1216.5 m MSF) of this interval consists of loose sand, and the lower part (1216.5–1236.5 m MSF) consists of semiconsolidated sandstone. The lithology of Unit I was observed only from the cuttings samples available from the first phase of riser drilling during Expedition 337. Cuttings samples were taken at 10 m depth intervals from 636.5 to 1256.5 m MSF with a gap of 70 m between 1116.5 and 1186.5 m MSF. A total of 56 cuttings were available and subjected to macroscopic (visual and binocular) and microscopic (smear slides) observations, as well as XRD and XRF analyses. However, because of high contamination from the drilling mud ingredients (e.g., walnut shells, mica, and metal shards) associated with the mud-loss countermeasure, XRD and XRF analyses for the cuttings samples between 1116.5 and 1256.5 m MSF were not conducted.

To obtain a more precise interpretation, we divided this unit into four subunits. Subunit Ia is characterized by semiconsolidated and consolidated diatom-bearing clayey siltstone to silty shale with volcanic glass (647–926.5 m MSF; cuttings Samples 25-SMW through 58-SMW). The Subunit Ia/Ib boundary is defined by the top of the first occurrence of fine sand in cuttings samples (between Samples 58-SMW and 59-SMW). Subunit Ib consists of semiconsolidated diatom-bearing clayey siltstone with common fine sandstone (926.5–1116.5 m MSF; Samples 59-SMW through 88-SMW). At the top of Sample 92-SMW (1116.5 m MSF), cuttings consist mainly of unconsolidated to semiconsolidated sandstone and silty sandstone with rare clayey siltstone. This lithologic unit (Subunit Ic; 1116.5–1236.5 m MSF) is visible to 1236.5 m MSF (Sample 95-SMW). The Subunit Ic/Id boundary was characterized by a change in lithology to semiconsolidated clayey siltstone (1236.5–1256.5 m MSF; Samples 97-SMW through 98-SMW).

Unit II

  • Intervals: cuttings Samples 337-C0020A-105-SMW through 216-SMW; Cores 337-C0020A-1R through 14R

  • Depth: 1256.5–1826.5 m MSF

The Unit I/II boundary is defined by lithologic changes found in cuttings samples (between Samples 337-C0020A-98-SMW and 105-SMW). A rapid decrease of sand content is visible at this boundary. The boundary is also characterized by the first occurrence of rare glauconite in cuttings samples (1266.5–1276.5 m MSF; Sample 111-SMW). In addition, the amount of biosiliceous material decreases, whereas the amount of plant remains increases. Unit II consists mostly of silty shale, with some intervals of sandstone and siltstone. The interval between 1256.5 and 1826.5 m MSF was characterized at the top by a relatively high silt content (70%–90%) (Fig. F1). The silt content decreases to the middle part of the unit and shows the lowest content at 1556.5 m MSF (15%). Silt content then increases again to the bottom of this unit (50%).

To obtain a more precise interpretation, this unit was divided into two subunits. Subunit IIa includes mainly shale, siltstone, and sandstone associated with marine fossiliferous material (1256.5–1506.5 m MSF; Samples 105-SMW through 153-SMW and Cores 337-C0020A-1R through 6R). Subunit IIb consists of organic-rich shales and sandstone with plant remains (1506.5–1826.5 m; Samples 154-SMW through 216-SMW and Cores 7R through 14R).

The principal difference between Subunits IIa and IIb is the presence of abundant organic material mostly in the form of wood/lignite fragments and glauconite in cuttings samples in Subunit IIb. The sand fraction generally increases to the bottom of Subunit IIb, whereas the amount of biosiliceous material decreases. Organic-rich silty shale with abundant volcanic glass comprises the upper part of this subunit, with grains that were mostly moderate to subangular and moderate to well sorted. The lower part of this subunit was dominated by fine to medium sandstone with few silt and shale intercalations. Grains were moderate to subangular and moderately to well sorted. Parallel and wavy laminations are commonly observed throughout Subunit IIb, whereas cross- and ripple cross-lamination as well as coarsening- and fining-upward intervals were only observed in some parts of Subunit IIb. Joint, normal fault, and mineral-filled fractures were found in the middle part of this subunit. Bioturbation ranged from slight to moderate and was common throughout this subunit. Unit II demonstrated a transition from a siltstone-dominated lithology in Subunit IIa to one that is bimodally dominated by clay and sandstone in Subunit IIb.

Unit III

  • Intervals: cuttings Samples 337-C0020A-217-SMW through 260-SMW; Cores 337-C0020A-15R through 25R

  • Depth: 1826.5–2046.5 m MSF

The Unit II/III boundary was selected at the point where plant remains and lignite were the dominant fraction in the cuttings samples. However, the first occurrence of coal in cores appears in the lower part of Core 337-C0020A-14R (1824.6 m CSF-B). To be consistent with defining unit boundaries on the basis of cuttings samples, it was decided to set the boundary below cuttings Sample 216-SMW. The lithology of Unit III was determined from macroscopic and microscopic observations on 11 cores (337-C0020A-15R through 25R) and 22 cuttings samples (337-C0020A-217-SMW through 261-SMW) available from the second phase of drilling operations. This unit is dominated by several coal horizons that are divided by coaly shales, siltstones, and sandstones.

Unit IV

  • Intervals: cuttings Samples 337-C0020A-261-SMW through 391-SMW; Cores 337-C0020A-26R through 32R

  • Depth: 2046.5–2466 m MSF

The Unit III/IV boundary was selected at the base of the last dominant occurrence of plant remains/lignite in cuttings samples (337-C0020A-260-SMW). The absence of thick coal layers differentiates this unit from Unit III. Unit IV covered the deepest part of Hole C0020A. The identification of the lithology in this unit was supported by the recovery of seven cores (337-C0020A-26R through 32R) and 42 cuttings samples (261-SMW through 391-SMW). Unit IV was dominated by silty shales in the upper part, sandstone intercalated with siltstone in the middle part, and shales associated with sandstone, siltstone, and a thin coal layer in the lower part.

Unit IV is divided into two subunits. The boundary between these two subunits is drawn by the first common occurrence of plant remains/lignite in a cuttings sample (384-SMW). Shale and sandstone associated with carbonate and glauconitic material characterize Subunit IVa (2046.5–2426.5 m MSF; cuttings Samples 261-SMW through 371-SMW and Cores 26R through 29R). Subunit IVb shows clay and silt as the dominant size fraction in cuttings samples (2426.5–2466 m MSF; cuttings Samples 384-SMW through 391-SMW and Cores 30R through 32R). Coal fragments are also dominant. Clay and silt are the dominant size fraction in cuttings samples. Coal fragments are also dominant in the cuttings samples. In cores of Unit IV, fine to medium sandstone with minor shale layers with one coal layer is visible.

Depositional environment

In general, sand-sized lithic fragments and biotite occur throughout the hole. Furthermore, sediments are often poorly to moderately sorted and grains are often moderate to subangular rounded (except the deeper parts in the hole). These details suggest short and/or sometimes fast transport distances of the sediments.

The interpretation of Unit I is limited to only cuttings samples because no cores were collected until 1276.5 m CSF-B. Diatom-bearing clay- and siltstones are interpreted as hemipelagic sediments representative of a slope to abyssal plain setting. Toward the top of Unit II, the occurrence of sponge spicules and diatoms decreases, whereas rare glauconite was observed. Glauconite is a hydrated illite-group mica mineral that typically forms in water depths of 5–500 m in the presence of high biological activity (Cloud, 1955; McRae, 1972). Thus, glauconite is an indicator for shallow-marine environments, typical of continental shelf settings. Glauconite formation is limited by high sedimentation rates. In the uppermost part of Unit II, sediments are likely still deposited in a deep-marine setting but with a position nearer to the shelf margin. Occurrence of abundant and dominant glauconite is probably caused by transport from the shelf to the slope by mass flows (cf. Core 337-C0020A-2R).

The first occurrence of bioturbation appears in Core 337-C0020A-6R (1495–1499.135 m CSF-B). In addition, cross-laminae that are convex-upward are visible in this core. These laminae could be interpreted as hummocky cross-stratification, which is typical for storm sediments. In Core 8L and deeper, bioturbation and glauconite are very common. Burrows in this area are normally horizontal and often occur in shales and siltstones. Detailed ichnofacies studies were not conducted, but with the observation of glauconite in the sediments, we suggest that the burrows belong to the Cruziana ichnofacies, which is typical for continental shelf environments. Symmetric cross-stratifications, which are visible in this part of the hole, are typical for low-energy and shallow-marine environments. Therefore, we conclude that sediments from the end of Subunit IIa to the middle part of Subunit IIb (1495–1634.905 m CSF-B; Cores 6R through 10R) were deposited on the shelf between the fair-weather wave base and the storm wave base (Fig. F8).

Horizontal and vertical burrows occur in the lower part of Subunit IIb (1737.5–1828.095 m CSF-B; Cores 11R through 14R). Bioturbation is observed in clayey, silty, and sandy sediments. Parallel and cross-lamination as well as wavy bedding are visible in Cores 11R through 14R. Several shell fragment–rich layers are observable in Core 12R, and some shell fragments, likely of the genus Ostrea, appear in Core 13R. Greater diversity of different burrows in different types of sediments and the sedimentary structures suggest deposition in an intertidal and lagoonal environment.

In the middle of Core 14R, the depositional environment changes slightly. Organic-rich shales enclose a thin coal horizon. Asymmetric cross-stratification above the coal layer suggests a high-energy nearshore environment. Several coal horizons and organic-rich shales are present in Unit III until Core 25R (1995–2002.345 m CSF-B). High gelification of the coal, no inertinite (inert component in the coal), and areas with high clay and sand content in coal horizons suggest very wet conditions in the swamp. Glauconite, mostly horizontal burrows, and different types of sedimentary structures are visible in Unit III. Cores 15R through 19R (1919–1960.295 m CSF-B) are sand dominated and show coarsening- and fining-upward cycles and parallel laminations. Cores 20R and 21R (1959.5–1972.71 m CSF-B) are rich in shale and siltstone with flaser and lenticular bedding, wavy bedding, and coarsening- and fining-upward cycles. Core 22R (1973–1979.19 m CSF-B) contains planar bedding and cross-bedding within sandstones, whereas Core 23R (1981.5–1991.25 m CSF-B) is again shaly with authigenic carbonate bands possibly containing siderite. Authigenic carbonate bands are also visible in Cores 24R (sandstone rich; 1991–1994.965 m CSF-B) and 25R (mostly fine grained; 1995–2002.345 m CSF-B). Lenticular bedding is typical for tidal flats. The absence of carbonate bands in the upper part of Unit III suggests a tidal flat environment with cycles of changing environments from intratidal to supratidal (marsh and swamp). Sandstones immediately above the coal seams could be related to deltaic environments (channels of an estuary or of a fluvial-dominated delta). Siderite bands and nodules are typical for euxinic environments (brackish) and can be found in back-barrier marine environments with lagoons, tidal flats, tidal channels, and flood tidal deltas (Boggs, 2006). Initial XRD, XRF, and smear slide results suggest that siderites are a common constituent in observed carbonate bands. No glauconite was observed near siderite bands. Cycles of frequently changing environments from back-barrier to wetland are also interpreted for this interval. Burrows and glauconite in Unit III show the presence of marine conditions. It is not known if these frequent changes in depositional environment are caused by sea level fluctuations. Sediments of Unit IV look very homogeneous over wide parts of the hole (cuttings and cores). Carbonate bands, horizontal burrows, fining- and coarsening-upward cycles, cross-bedding, and lenticular and flaser bedding are visible throughout Unit IV. Therefore, the depositional environment of this unit is similar to that of Unit III. Wetland conditions are only observable in the lowermost part of this unit (coal and organic-rich shale). Siderite layers are again an indicator for a back-barrier environment with tidal flats (lenticular bedding) and tidal channels. Sandy distributary channels from deltas are also possible environments preserved in Unit IV. The coal horizon in the lowermost part of this succession is similar to the coal layers in Unit III, suggesting a similar depositional environment.

Units III and IV are ~700 m thick in total. These parts of the hole were deposited in almost the same depositional environment, showing fluctuations between intertidal and wetland (marine/brackish influence) conditions. This also suggests that the sediment accumulation is in balance with tectonic subsidence, which seems to be high in this area.

Drilling disturbances

Recovery in Hole C0020A was excellent overall, even at such deep penetration depth. Although Cores 337-C0020A-1R through 7R had <50% recovery, core recovery in deeper parts of the hole, especially around the coal formation interval, was high (80%–100%).

Drilling disturbance (Fig. F9) varied depending on core and lithology. Unit II experienced slight disturbance or fracturing, except for Cores 337-C0020A-5R and 6R, where parts of the cores were moderately to heavily fractured. In Unit III, drilling disturbance was also generally low, especially in the shale and silt regions. Semiconsolidated sandstones were more likely to be affected by drilling disturbance. The cores of Unit IV were also mostly found in good condition with minor disturbance throughout. However, moderate to heavy disturbance was found in some sections of Cores 28R and 31R.

Injections of drilling mud and fluid used in riser drilling during Expedition 337 caused complications to visual observations of the cores. Semiconsolidated materials were commonly observed in Hole C0020A, and drilling mud often easily penetrated the rock, causing possible false lamination structure in the cores, which might be misinterpreted as natural sedimentary structure preserved in the cores (Fig. F9). Drilling mud may also contribute additional contamination to smear slide analysis. In this case, it was possible that the mud layers caused by mud injections were observed on smear slides under the microscope, which could also lead to misinterpretation of fossils or other materials present on the slides.

Cuttings contamination

During Expedition 337, drilling mud circulation also caused high contamination of cuttings samples. Cuttings samples were washed to eliminate this contamination. In addition to drilling mud components, other material such as walnut shells and mica from lost circulation material (LCM) and metal shards from drilling equipment were observed in cuttings samples. During the washing process, semiconsolidated material was easily broken, and fractions finer than 250 µm were lost through the mesh. This is one possible reason why cuttings and cores containing a large amount of sandstone were not in good agreement in the lithologic description (e.g., cuttings Sample 337-C0020A-114-SMW and Core 337-C0020A-1R).

Drilling mud may contaminate samples for XRD and XRF analyses as well. For this reason, XRD and XRF analyses were not conducted on the cuttings samples between 1116.5 and 1256.5 m MSF because it was obvious that the samples were highly contaminated with LCM. Although attempts were made to avoid samples with high contamination, in some cases it is evident that the mud penetrated the cutting chips easily and it is extremely difficult to differentiate this type of mud from clay originally present in the cuttings samples. In this case, the contamination can be seen only after measuring XRD and XRF of samples. For example, potassium in cuttings samples was often higher than that observed in the cores, likely due to introduction by drilling fluid. However, overall trends are nearly the same. Therefore, it is suggested that the influence of contamination is only low.

Other than from drilling mud, some of the cuttings samples probably represent a mix from different depths. Poor agreement between XRD and XRF results of cuttings and cores (e.g., 1256.5–1506.5 m MSF) could be due to this mixture of material from a wide range of depth intervals (e.g., recirculated material or cave-in material). On the other hand, washing-out of semiconsolidated and unconsolidated sand in cuttings as well as different lithologies taken from cores can also cause differences in XRD and XRF results. Variability in elements and mineral abundances measured from XRF and XRD are much lower in cuttings than in cores, suggesting the influence of a dominant lithology can overprint the downhole elemental variation in cuttings. In this case, rather than having cuttings or core profiles only, it is quite beneficial to have both sets of analysis to help identify contamination.

Conclusion

Based on macroscopic and microscopic descriptions of cuttings and core samples during Expedition 337, which was supported by XRD and XRF data as well as all available data (e.g., logging, micropaleontology, and physical properties data), we defined four different lithologic units at Site C0020. The succession of lithofacies in Hole C0020A also provides insight into the evolution of depositional environments in the site region.

  • Unit I (647–1256.5 m MSF) consists primarily of diatom-bearing silty clay. This unit clearly represents a marine offshore environment.

  • Unit II (1256.5–1826.5 m MSF) consists mostly of silty shale with some intervals of sandstone and siltstone. The principal differences between this unit and the upper unit are the tendency for decreasing amounts of biosiliceous material and increasing plant and glauconite material toward the deeper part of the unit. Based on those phenomena, Unit II was divided into two different subunits: sandstone and siltstone associated with marine fossiliferous material and organic-rich shale and sandstone associated with plant remains. This unit reveals more offshore transition environment, which then gradually changes into a shallow-marine setting. The lower part of this unit is situated in the intertidal zone.

  • Unit III (1826.5–2046.5 m MSF) is dominated by several coal horizons that are divided by coaly shales, siltstones, and sandstones. Almost all coal horizons consist of detritic to xylo-detritic coal with some layers of xylitic coal. Sometimes shaly coal was observed. Water content, color, and vitrinite reflectance measurements of the coal suggest that the coal has low maturity (Ro = 0.36%). Bioturbation and sedimentary features like flaser bedding, lenticular bedding, or cross-bedding suggest a nearshore depositional environment with tidal flats and tidal channels. The presence of coal, coaly shale, and siderite bands at the bottom of this unit suggest a back-barrier marine environment in combination with wetlands (e.g., salt marsh or swamp). Small terrestrial influence might occur within sand bodies that overlie coal horizons. This could be due to channels from fluvial-dominated or tidal-dominated deltaic estuarine environments.

  • Unit IV (2046.5–2466 m MSF) is dominated by silty shales in the upper part, sandstone intercalated with siltstone in the middle part, and shale associated with sand, silt, and a thin coal layer in the lower part. The absence of thick coal layers differentiates this unit from Unit III. Although one coal layer appears again at the bottom of the unit, it is only a thin layer, which may indicate the presence of another coal-rich unit below this interval. Unit IV suggests deposition in the same environment as Unit III. Units III and IV represent high-frequency changes of depositional environment. Within a few meters, sediments indicate tidal flats and tidal channels that are overlain by organic-rich marsh sediments and terminate in the formation of a peat. Often, coal layers are bounded at the top by sandstones that can be interpreted as channel sandstones of a fluviodeltaic system.