Previous   |    Next

doi:10.2204/iodp.proc.346.109.2015

Geochemistry

Sites U1428 and U1429 are located in the East China Sea, just west of the southern Japanese island of Kyushu (“Background and objectives”). These sites are located ~7.4 km from each other and have high sedimentation rates with relatively high carbonate contents compared to other sites drilled during Expedition 346. As this location sediments offer a prime opportunity to reconstruct paleoceanographic records from a variety of proxies based on foraminiferal carbonate and organic geochemistry, it is important to establish the baseline geochemical properties of the interstitial water and solid samples to support these efforts.

The terrestrial source of material is an important component at Sites U1428 and U1429, as these sites are possibly under the influence of the Yangtze (Changjiang) and Yellow (Huanghe) Rivers. A piston core (MD98-2195), collected from the IMAGES IV cruise near this site, served as one of the presite survey data sources. According to a broad array of geochemical analyses performed on this core (Kawahata et al., 2006; Ijiri et al., 2005; Kawahata and Oshima, 2004), the site has a high sedimentation rate along with high organic carbon contents (1.5 wt% for the last ~42 k.y.). During Expedition 346, core recovery of the sites was limited to 150 and 186 m CSF-A, respectively, because of the presence of sand layers at deeper depths.

The very rapid sedimentation rates determined for piston Core MD98-2195 are intriguing to interstitial water studies. At such rates, downward advection of water can generally outpace upward diffusion for most dissolved species. Consequently, unusual interstitial water profiles can develop. However, the geochemistry program at Sites U1428 and U1429 began as an effort to construct carbonate and organic matter profiles in line with the paleoceanographic objectives of the expedition.

Sampling

The initial geochemistry plan was a straight path. As there would be three deep holes and multiple logging exercises at a single location (Site U1428), we would take routine sets of samples for solid-phase, interstitial water, and gas analyses in Hole U1428A. The solid-phase samples would be a mix of squeeze cake and discrete samples because the squeeze cake samples represent an amalgamation of sediment over 5 cm or more. The water samples would include one squeezed interstitial water (IW-Sq) sample for Cores 346-U1428A-1H through 3H, one IW-Sq every other core for Cores 4H through 9H, and one IW-Sq every 30 m over the next several hundred meters. The water program would be limited partly because we knew that the high sedimentation rate might lead to complicated profiles of dissolved species, difficult to interpret without numerous measurements. We would, however, supplement the routine water sampling with 15 Rhizon samples at the top of Hole U1428A, primarily to connect shallow and deep interstitial water profiles of certain dissolved constituents (Mn, Fe, and NH₄⁺). The gas samples would be one headspace sample per core, as mandated for safety reasons.

Interstitial water alkalinity is the first measurement completed in the Geochemistry Laboratory. The limited sampling in Hole U1428A provided an intriguing result within hours after drilling commenced: the alkalinity profile (presented below) had an unusual convex upward profile that peaked somewhere between 36 and 55 m CSF-A. We noted this oddity but placed minimal effort toward understanding its cause, given objectives and scheduling for Site U1428 and our preconceived ideas regarding profiles of dissolved constituents in rapidly accumulating sediment. The termination of Hole U1428A in sand horizons deeper than 140 m CSF-A, the detour to Site U1429, and the return to Hole U1428B (“Operations”) modified geochemistry plans at this location considerably, and we put effort into generating potential glacial–interglacial oscillations in sediment composition and constraining the cause of the alkalinity peak.

Upon completing both sites, the geochemistry group collected and analyzed a range of samples. These included the following (Tables T13, T14, T15, T16, T17, T18):

• 102 sediment samples taken from interstitial water squeeze cakes and also additional core sampling (CARB and SRA).

• 2 mudline (ML) samples, representing water from inside the liner of the uppermost cores in Holes U1428A and U1429A.

• 18 interstitial water samples from whole-round squeezing (IW-Sq) of discrete intervals in Holes U1428A and U1429A (9 samples from each hole).

• 59 interstitial water samples from Rhizons (IW-Rh) of discrete points: 15 over the upper 8.5 m in Hole U1428A as planned, 20 in Hole U1429A, and 24 in Hole U1428B to “chase” the origin of the alkalinity peak (below).

• 41 headspace (HS) gas samples.

Interstitial water sampling modifications

Several aspects of interstitial water sampling at Sites U1428 and U1429 warrant further mention. For all IW-Sq samples, a 0.20 µm Whatman Puradisc filter was added “in series” to the standard 0.45 µm filter between the squeezer and the sample syringe. This approach was chosen to address the consistent offset in concentrations of certain dissolved constituents between IW-Sq and IW-Rh samples, as the discrepancy might reflect filter size (discussed in previous Expedition 346 site chapters).

Rhizon sampling in Holes U1429A and U1428B was carried out in an unconventional manner in an effort to locate the precise depth of the alkalinity peak in these holes, and therefore, the sequence and timing of Rhizon sampling proceeded on the basis of initial “low-resolution” alkalinity measurements. For example, once it was established that the alkalinity peak in Hole U1428B lay between Sections 346-U1428B-6H-1 and 7H-4, we pulled the remaining sections from Core 6H-1 while they equilibrated for NGR logging and took second and third batches of Rhizons. The procedure is not ideal for absolutely accurate and precise alkalinity measurements because it adds time between sample recovery and extraction. However, such “real-time” geochemical tracking may be ideal for addressing certain questions and cannot be accomplished practically with IW-Sq samples.

We had difficulty inserting Rhizons into cores deeper than 40 m CSF-A at previous sites (see “Geochemistry” in the “Site U1422” chapter and “Geochemistry” in the “Site U1424” chapter [Tada et al., 2015c, 2015d]). To conduct the above sampling, we used the end of a “swizzle stick” to make a small hole in the sediment before inserting the Rhizon. This is a piece of wood with a diameter slightly smaller than the Rhizon.

Interstitial water observations

While chasing the alkalinity peak at these sites with Rhizons, we made two observations that seem pertinent to understanding the geochemistry. First, the odor of H₂S was detectable in core sections between 42 and 54 m CSF-A. Second, the water flow was extremely fast for several Rhizons. In particular, 10 mL of water was collected within 12 min for Samples 346-U1428A-6H-1, 50 cm, 12H-2, 80 cm, and 15H-1, 55 cm, and Samples 346-U1428B-5H-7, 20 cm, and 6H-2, 30 cm. These are very porous (and permeable) horizons, as also seen in the gamma ray attenuation (GRA) density records (“Physical properties”). Additionally, it is important to note that newer syringes have better suction than preused syringes and reduce extraction time.

Carbonate and organic carbon

A relatively high resolution solid-phase analysis plan was conducted at Sites U1428 and U1429. In addition to squeeze cake samples, two additional discrete sediment samples were collected per core for the measurement of carbonate and organic carbon contents (CARB and SRA samples).

Calcium carbonate contents in Hole U1428A display a cyclic pattern, alternating between low values of ~15 wt% and high values of ~35 wt% (Fig. F32). It appears that the profile illustrates glacial–interglacial changes since marine isotope Stage (MIS) 11. CaCO₃ contents increase during the interglacial period and decrease during glacial times. Deeper than 124 m CSF-A, CaCO₃ contents fluctuate significantly, which is associated with the alternating mud and sand layers near the bottom of the hole. CaCO₃ contents at Site U1429 (Fig. F33) have a similar CaCO₃ profile to Site U1428, but the stratigraphy is different because Site U1429 has a higher sedimentation rate. CaCO₃ analyses were normally performed only in samples from the first hole of each drilled site, but for Site U1428, CaCO₃ determinations were also performed in Hole U1428B. This approach was used in order to better constrain the observed long-term variability as first measured in Hole U1428A.

The organic carbon contents vary from a relative high in the uppermost samples at Site U1428 (total organic carbon [TOC] = ~1.5 wt%) to almost zero in the basal sand layers. This change is a gradual decrease with depth but exhibits a sawtooth pattern with up to 0.20 wt% variance. However, at Site U1429 there is no apparent decrease and TOC contents remain fairly stable, close to 1 wt%, all along the record with variation from 0.7 to 1.6 wt%. An exception to this is seen in samples taken at tephra horizons (80–90 m CSF-A for Hole U1428A and 64–65 m CSF-A for Hole U1429A) and sand layers (starting at 133 m CSF-A for Hole U1428A and 156 m CSF-A for Hole U1429A). The higher sedimentation rate found at Site U1429 appears to enhance the preservation of the sedimentary organic matter, as can be perceived by the relatively high and constant TOC contents at that site.

Total nitrogen (TN) contents are low, as observed at previous sites (Figs. F32, F33). However, they closely follow the TOC pattern of each site, decreasing over depth. Site U1428 shows slightly higher TN concentrations for the uppermost 50 m CSF-A and slightly lower TN concentrations deeper than 115 m CSF-A. The C:N ratio is low and follows the TOC trend but is meaningful because of very low N content, as discussed by Kawahata et al. (2006). Ship conditions (sea state, which interferes with the weighing of sample) and low TOC and N contents at this site caused measured values to be near the error for this method, thus the C:N values are not necessarily reflecting a definitive origin for the organic matter.

Alkalinity and sulfate

The most immediate and perhaps most unusual shipboard geochemical finding at Sites U1428 and U1429 is the alkalinity profile (Fig. F34). Alkalinity is 2.37 mM in both mudline samples. This value is very close to the sum of dissolved inorganic carbon and dissolved B for deep water with a salinity of 35 (Table T10 in the “Methods” chapter [Tada et al., 2015b]). Below the seafloor, alkalinity rises to a maximum between 45 and 48 m CSF-A. The rise is complicated and convex downward at both sites. In Hole U1428B, where a series of Rhizons were placed to find the depth, the maximum is at 47.5 ± 0.8 m CSF-A. Although nearly coincident in terms of depth, maximum alkalinity is greater at Site U1429 (29.6 mM at 46 m CSF-A) than at Site U1428 (23.5 mM). From 48 m CSF-A to the base of the holes at both sites, alkalinity generally decreases, reaching 7.4 at ~162 m CSF-A in Hole U1428A and 12.8 mM at ~177 m CSF-A in Hole U1429A.

The shapes of the alkalinity profiles at Sites U1428 and U1429, at least over the uppermost 45–48 m, are different from most marine sediment sequences, including the sites drilled previously during Expedition 346. They strongly suggest very rapid sedimentation and a “point source” of alkalinity. An extreme sedimentation rate would lead to a convex downward alkalinity profile because dissolved HCO₃^– cannot diffuse through porous media very fast. Such a rate is also consistent with age determinations of piston Core MD98-2195 (Kawahata and Ohshima, 2004). Although multiple sources for the excess alkalinity at 45–48 m CSF-A might be suggested, the dissolved SO₄^2– profiles (Fig. F34) indicate the proximal cause.

The two mudline samples have dissolved SO₄^2– concentrations of 29.1 and 29.4 mM. These values are slightly higher than that for seawater with a salinity of 35 (29.0 mM). Dissolved SO₄^2– concentrations decrease below the seafloor. The decrease is somewhat complicated, with the rate of concentration change varying with depth. In general, however, the rate accelerates toward 0 mM at ~47 m CSF-A. For example, at Site U1428, SO₄^2– decreases by 9 mM over the uppermost 36 m of sediment but 20 mM over the next 11 m of sediment. Deeper than 47 m CSF-A, SO₄^2– concentration is 0 mM. In summary, shallower than 47 m CSF-A, the dissolved SO₄^2– profiles at Sites U1428 and U1429 are near mirror images of the alkalinity profiles.

Anaerobic oxidation of methane (AOM), a recurring theme in geochemistry discussions of Expedition 346 drill sites, provides the basic explanation for the alkalinity and SO₄^2– profiles at Sites U1428 and U1429. Methane produced at depth reacts with SO₄^2– in the interstitial water. This microbial reaction generates HCO₃^– and HS^–, which collectively manifest as an inflection in interstitial water alkalinity profiles. Indeed, core sections around 47 m CSF-A had a very strong H₂S odor.

Volatile hydrocarbons

The headspace gas profiles at Sites U1428 and U1429 (Fig. F35) enhance the above interpretation involving AOM. CH₄ concentrations at Site U1428 are <10 ppmv from the seafloor to ~45 m CSF-A. Deeper than ~45 m CSF-A, CH₄ values increase relatively slowly, at least compared to several previous sites, reaching 7900 ppmv at ~55 m CSF-A. The beginning of this increase corresponds to the horizon where interstitial water SO₄^2– concentrations approach the detection limit. Deeper than ~55 m CSF-A, CH₄ values fluctuate between 8800 and 2200 ppmv until 180 m CSF-A. No ethane or higher hydrocarbons were found in samples from this site.

The downhole CH₄ profile at Site U1429 has a similar trend. CH₄ concentrations are <10 ppmv for the upper ~50 m CSF-A. At this depth, CH₄ values increase slowly, with a maximum value of 13,500 ppmv at ~62 m CSF-A. Deeper than ~62 m CSF-A, CH₄ concentrations fluctuate between 11,200 and 2,500 ppmv until 180 m CSF-A.

In summary, headspace CH₄ concentrations support our inferences from the alkalinity and sulfate profiles. There is a relatively sharp sulfate–methane transition (SMT) somewhere between ~47 and 50 m CSF-A at both Sites U1428 and U1429.

Barium

At locations with a prominent SMT, dissolved Ba concentrations typically rise when sulfate disappears and barite begins dissolving. At Site U1428, dissolved Ba concentrations are very low (<15 µM) until 45.3 m CSF-A and then begin to increase rapidly, reaching a maximum of 540 µM at 74.35 m CSF-A (Table T14; Fig. F36). This coincides with the depth interval over which sulfur disappears, which likely causes barite to dissolve. Site U1429 shows a similar trend at similar depths, but with a lower resolution of samples (Table T17; Fig. F36).

Calcium, magnesium, and strontium

The Ca profile exhibits a concave upward profile beginning at 10.1 mM at the mudline and decreasing to 2.6 mM at 48.97 m CSF-A at Site U1428 (Table T14; Fig. F37). The Site U1429 mudline has a Ca concentration of 10.1 mM and a similar shape to Site U1428 (Table T17; Fig. F37). Concentrations appear to decline slightly quicker than at Site U1428, reaching 2.3 mM by 45.85 m CSF-A, but this may be an artefact of sampling resolution. Ca remains low for the rest of the site (<6 mM).

The Mg profile for Site U1428 is similar to Ca, expressing a concave upward trend in the upper ~50 m CSF-A. Bottom water has a Ca concentration of 52.1 mM and concentrations rapidly decline from 51.7 to 36.9 mM between 32.8 and 64.31 m CSF-A. Deeper than this, concentrations decline more gradually, reaching a minimum of 30.6 mM at 131.35 m CSF-A. Site U1429 shows a similar trend, but with lower sampling resolution, starting at 52.5 mM in the mudline and decreasing to 27.9 mM at 176.54 m CSF-A.

The Sr profile at Site U1428 is also concave upward for the upper ~50 m CSF-A, starting at 91.5 µM at the mudline and decreasing to 77.95 mM at 46.8 m CSF-A. Deeper than 46.8 m CSF-A, concentrations increase to 120.36 µM at 102.85 m CSF-A. Site U1429 has a similar shaped profile with a Sr concentration of 91.7 µM at the mudline and reaching a minimum of 69.54 µM at 44.2 m CSF-A before increasing to ~115 µM at the deepest samples from 124 to 176 m CSF-A.

Bromide

The dissolved Br^– concentration profiles are very similar at Sites U1428 and U1429 (Fig. F38). Br^– concentration in the mudline is 0.87 mM for both sites, followed by a downhole increase in concentration that occurs in a step-wise fashion. Br^– concentrations remain relatively constant from the seafloor to 45 m CSF-A at 0.85–0.88 mM, whereas deeper than 45 m CSF-A there is a step-wise increase in concentrations, which continue to increase downhole to a maximum of 1.04 and 0.97 mM at Sites U1428 and U1429, respectively.

Ammonium and phosphate

The NH₄⁺ and PO₄^3– profiles at Sites U1428 and U1429 display the same general trends, but with overall higher concentrations observed at Site U1429 (Tables T14, T17; Fig. F39). Concentrations of both nutrients are relatively low in the mudline sample with dissolved NH₄⁺ and PO₄^3– concentrations being 26 and 3 µM for Site U1428, and 25 and 1.7 µM for Site U1429 (Fig. F40), respectively. Below the seafloor, NH₄⁺ concentrations increase gradually downhole, reaching higher absolute concentrations at Site U1429, which probably reflects increased organic carbon degradation at that site. PO₄^3– also reaches higher absolute concentrations at Site U1429, further suggesting that this site experiences increased organic carbon degradation. The general trend of the PO₄^3– profiles shows an increase downhole to a maximum at ~50 m CSF-A at both sites, with concentrations of ~58 and ~89.5 µM at Sites U1428 and U1429, respectively. PO₄^3– concentrations then decrease downhole to a minimum of <5 µM by ~160 m CSF-A at Site U1428 and ~177 m CSF-A at Site U1429.

Manganese and iron

Mn concentration at the mudline is 20.8 µM at Site U1428 (Table T14; Fig. F41). Mn concentrations reach a maximum of 68.2 µM in the uppermost Rhizon sample at 0.05 m CSF-A. Mn concentrations decline to 40.8 µM at 0.5 m CSF-A, where there is an inflection point and the rate of decline decreases, reaching 27.2 µM at 2.1 m CSF-A. The decline continues, reaching 6.8 µM at 17.35 m CSF-A and remaining <5 µM for the remainder of the hole, with many samples dropping below the detection limit (0.2 µM). At Site U1429, the overall trend is similar (Table T17; Fig. F41). The Site U1429 mudline sample has a lower concentration of 4.71 µM than Site U1428, and the maximum measured Mn concentration at Site U1429 (22.0 µM) occurs at 7.85 m CSF-A.

At Site U1428, Fe concentration in the mudline is below the detection limit (0.9 µM) and reaches a maximum of 4.1 µM at 0.05 m CSF-A. Concentrations rapidly decline to below the detection limit again at 0.15 m CSF-A and remain at, or near, detection for the remainder of the hole. Fe concentration in one anomalous Rhizon sample is 2.4 µM at 64.31 m CSF-A. Sampling resolution is low at Site U1429 and does not capture any Fe concentrations >3 µM, except for one anomalous value of 7.13 µM at 64.36 m CSF-A.

Chlorinity and sodium

Both Cl^– and Na show little variation with depth (Figs. F42, F43). At Site U1428, chlorinity is 549.3 mM in the bottom water sample and varies between 545 and 562 mM for all the samples (Table T14). The range of chloride concentrations at Site U1429 is between 547 and 567 mM, with 552.3 mM at the mudline and no trend with depth (Table T17).

The mudline has a Na concentration of 468.0 mM at Site U1428 and 471.0 mM at Site U1429. Concentrations vary between 454 and 475 mM for all the samples at both sites, with no downhole trends.

Potassium

K concentration in the mudline is 10.4 mM at Site U1428 and 10.3 mM at Site U1429 (Tables T14, T17). Just below the seafloor, K concentrations at Site U1428 increase to 11.24 mM at 0.05 m CSF-A and continue to increase to 12.0 mM at 7.85 m CSF-A (Fig. F44). K then decreases to a minimum of 9.2 mM in the deepest sample (131.35 m CSF-A). Site U1429 expresses a similar trend at lower resolution sampling.

The rise in K immediately below the seafloor is similar to other sites drilled during Expedition 346 and may reflect exchange during authigenic mineral formation. The decrease in K with depth perhaps results from further reactions with ash and basalt (Murray et al., 1992), although formation of glauconite would also remove K⁺ from interstitial water (Föllmi et al., 1992).

Lithium and boron

At the mudline, Li concentration is 25.35 µM at Site U1428 and 25.62 µM at Site U1429 (Tables T14, T17). Li concentrations remain fairly constant, ranging from 25.35 to 32.28 µM at Site U1428 until 55.35 m CSF-A; deeper than this depth Li concentrations begin to increase (Fig. F45). Site U1429 follows a similar trend. The maximum Li concentration in Site U1428 (55.4 µM) is reached at the deepest sample analyzed (at 102.85 m CSF-A). At Site U1429, the maxima in concentrations of Li (59.77 and 70.11 µM) occur at 100.55 and 124.05 m CSF-A, respectively. The deepest sample at Site U1429, 176.54 m CSF-A, has a lower concentration of 41.74 µM.

At the mudline, B concentration is 414.0 µM at Site U1428 and 413.1 µM at Site U1429 (Fig. F45). Just below the seafloor at Site U1428, B concentrations are enriched to 474.1 µM at 0.05 m CSF-A and reach a maximum of 495.5 µM at 7.85 m CSF-A. B then declines gradually to concentrations similar to bottom water at 409.7 µM at 102.85 m CSF-A. Site U1429 shows that this trend continues and has a minimum concentration of 346.46 µM at the deepest sample, 176.54 m CSF-A.

Silica

H₄SiO₄ concentrations measured by spectrophotometry and Si concentrations measured by inductively coupled plasma–atomic emission spectroscopy (ICP-AES) agree within 2% of the measured value for all of the squeezed samples (Tables T14, T17). Bottom water has Si concentrations of 165.4 µM at Site U1428 and 103.7 µM at Site U1429. The squeezed samples at Site U1429 show a slight overall increase with depth, reaching 905.89 µM at 131.35 m CSF-A before declining to 200.28 µM at 161.67 m CSF-A (Fig. F46).

Rhizon samples from Site U1428, measured only by ICP-AES, add detail to the Si profile and show enrichment of Si in interstitial water up to 676.1 µM just below the seafloor at 0.05 m CSF-A. Si concentrations decline to 580.2 µM at 2.1 m CSF-A, but squeezer samples show increases deeper than 2.1 m CSF-A to 783.8 µM at 36.35 m CSF-A. High-resolution Rhizon samples detail a rapid decline in Si concentrations to 589.14 µM at 40.8 m CSF-A and then an increase to the maximum concentration of 981.17 µM at 55.35 m CSF-A. There is a rapid decrease to 878.2 µM at 64.31 m CSF-A and concentrations vary between ~860 and 900 µM for the remainder of Site U1428. Site U1429 shows a similar trend but at lower resolution.

Rhizon commentary

Our experimentation with Rhizon sampling continued at Sites U1428 and U1429. As at previous sites, the chemistry of water from IW-Sq and IW-Rh samples is similar. Two findings are worth highlighting.

First, water can be extracted from deeper sediment than we have previously managed by using a wooden swizzle stick, which is thinner in diameter than the Rhizon, to make a pathway for the Rhizon to be inserted into the sediment.

Second, the ultimate geochemistry program suggests a powerful future for such sampling. In Holes U1428B and U1430B, Rhizon sample locations were selected based on the multisensor loggers (Whole-Round Multisensor Logger [WRMSL] and/or Special Task Multisensor Logger [STMSL]) for GRA density and magnetic susceptibility records. This meant that the location of Rhizon samples could be selected in real time to target interesting centimeter scale intervals from the physical property record.

Preliminary conclusions

The carbonate content and interstitial water geochemistry stand out as unique features of Sites U1428 and U1429. Carbonate contents display cyclic features, which upon initial assessment are most likely related to glacial cycles and eustatic sea level changes. Upon further shore-based analyses, carbonate contents will likely help to support correlation between the two sites. Interstitial water chemistry indicates that the SMT at both sites is located at very similar depths near 47 m CSF-A, which is an unusual finding, as the stratigraphy does not correlate exactly at this depth between the sites because of differing sedimentation rates. Our preliminary conclusion based on both the geochemistry and sedimentology is that a volcanic tephra at this horizon affects the geochemical profiles at these sites. This tephra appears to have a much higher porosity and permeability than the upper and lower sedimentary mud that bound it. Therefore, the more typical diffusion gradients observed in interstitial water geochemistry at the other sites of Expedition 346 were not observed at Sites U1428 and U1429. Future and further geochemical study, including modeling, of these systems will be required to better constrain how the local (to regional) physical and chemical oceanographic processes affect the unusual aqueous geochemistry at these locations.

Top of page   |    Previous   |    Next