IODP Proceedings    Volume contents     Search

doi:10.2204/iodp.proc.346.104.2015

Geochemistry

Site U1423 was the second of three sites drilled in the Japan Basin. Similar to Sites U1422 and U1424, the primary goal behind drilling was to generate high-resolution paleoceanographic records, especially including the history of eolian dust and IRD. However, the water depth (1785 m) and landward proximity of Site U1423 contrasts with those at the other two sites. Furthermore, the location of Site U1423 had not been previously drilled. In short, the sedimentary record and geochemical profiles were expected to be somewhat different from those at Sites U1422 and U1424, but we had limited information before operations.

The main focus of the geochemistry program at Site U1423, other than meeting minimum requirements, was to generate profiles of solids, gases, and interstitial waters that could constrain fluxes of reactants and products resulting from microbial decomposition of organic matter. With this goal, comparisons to profiles at Sites U1422 and U1424 might provide insight to interactions between the “deep biosphere” and overlying waters of the sea. We endeavored to construct a set of detailed records, although this was difficult given the limited time during operations at the site (~2.9 days) and the transit (~6 h) to the next site (U1424). In particular, while drilling at Site U1423 cores arrived on deck every 25–45 min, which left minimal time for analyses.

Sample summary

The geochemistry group collected and analyzed a range of samples. These included the following (Tables T10, T11, T12):

  • 1 mudline (ML) water sample from inside the core liner above sediment of Core 346-U1423A-1H. We collected 50 mL.

  • 50 interstitial water samples from whole-round squeezing (IW-Sq) from Hole U1423A (44 samples) and the lower cores of Hole U1423B (6 samples) at nominal 4.5 m spacing. For most samples, we collected between 35 and 45 mL of interstitial water.

  • 0 interstitial water samples from Rhizons.

  • 41 sediment samples from the interstitial water squeeze cakes of Holes U1423A (35 samples) and U1423B (6 samples). All interstitial water squeeze cake samples from lithologic Unit I were analyzed, whereas alternate interstitial water squeeze cake samples from lithologic Unit II were analyzed.

  • 50 headspace (HS) gas samples. All of these samples were “paired” with the IW-Sq samples.

  • No gas samples on voids (Vacutainer) were taken because no large gas voids occurred during core recovery at Site U1423.

Carbonate and organic carbon

Sediment recovered at Site U1423 had only minor amounts of CaCO3 (~0.4 wt%) (Fig. F22). In general, CaCO3 contents decrease from ~1 wt% at 1.4 m CSF-A to 0.2 wt% at 18 m CSF-A and remain relatively low further downcore. The exceptions are samples centered around 56.3 and 70.3 m CSF-A, which have slightly elevated CaCO3 (Table T10), and one sample at 41.8 m CSF-A with a CaCO3 content of 14.3 wt%. This sample coincides with a zone rich in foraminifers (see “Biostratigraphy”).

Total carbon content of sediment is primarily controlled by the total organic carbon (TOC) content, which has variable and high values in lithologic Subunit IA (Fig. F22). In the uppermost 20 m below seafloor, TOC ranges from 2.1 wt% in the uppermost sample at 1.45 m CSF-A to 0.2 wt% at 8.8 m CSF-A. The variability partly reflects the nature of discrete sampling of shallow sediment in the marginal sea. Sediment sequences in this region, particularly lithologic Subunit IA, have numerous centimeter-scale dark and light layers containing different amounts and compositions of organic matter (Föllmi et al., 1992; Tada et al., 1992). Our samples, which come from intervals squeezed for interstitial water, represent short thicknesses (originally ~5 cm) separated nominally by 4.5 m. In any case, similar TOC variability has been observed in shallow sediment at other drill sites of the Japan Basin (Tamaki, Pisciotto, Allan, et al., 1990). Samples from lithologic Subunit IB, between ~83 and 103 m CSF-A, generally have lower TOC contents than those of lithologic Subunit IA. Interestingly, the highest TOC levels of Subunit IB are found near the base of the unit. Deeper than 103 m CSF-A and within lithologic Unit II, TOC contents are low, averaging 0.54 wt%.

The mean value of total nitrogen (TN) is 0.2 wt%. The maximum TN content (0.3 wt%) is found at ~30 m CSF-A.

Manganese and iron

The dissolved Mn profile (Fig. F23) exhibits a fairly smooth trend with depth. Close to the seafloor, Mn concentration is below detection limit (determined to be 0.5 µM). However, by 1.5 m CSF-A, dissolved Mn concentrations spike to 120 µM (Table T11). Beneath this depth, Mn concentrations decrease. The decrease is concave downward and decreases below detection limit at ~37 m CSF-A. At ~70 m CSF-A, dissolved Mn begins to increase again, reaching ~25 µM at ~175 m CSF-A.

The initial rise in dissolved Mn likely indicates a zone where microbes utilize Mn oxides and hydroxides to consume organic carbon (Froelich et al., 1979; Masuzawa and Kitano, 1983). Although our profile suggests dissolved Mn decreases to zero at the seafloor, the depth of zero dissolved Mn may lie somewhere between the seafloor and the uppermost sample at 1.5 m CSF-A (Masuzawa and Kitano, 1983). Our samples lack sufficient depth resolution to document a possible change in the slope of dissolved Mn concentration at shallow sediment depth. The second and deeper low in dissolved Mn, from ~37 to 70 m CSF-A, suggests a horizon where a mineral removes Mn and then redissolves. This zone corresponds to the alkalinity maximum (Fig. F24). It is possible that dissolved Mn2+ released at shallow depth diffuses downhole to a zone of high HCO3 concentration to form rhodochrosite (MnCO3). In turn, with burial, the rhodochrosite may recrystallize as a different carbonate mineral, such as siderite, releasing dissolved Mn2+ at depth. Such an interpretation is consistent with theory (Middelburg et al., 1987), and previous studies provide evidence for this process in this marginal sea, in particular at ODP Site 799 (Masuzawa and Kitano, 1983; Matsumoto, 1992). However, a thorough study comparing interstitial water and solid phases is required to confirm this interpretation.

The Fe profile at Site U1423 (Fig. F23) exhibits a high degree of variance, which may indicate the complexity of Fe cycling within the sediment, the challenges of measuring Fe, or both. Unlike Mn, Fe can be a major element in multiple authigenic minerals, including sulfides, phosphates, carbonates, and clay (Berner, 1981). Thus, the “bumpy” profile, where variations in dissolved Fe occur over short depth intervals, may reflect intervals where Fe is removed or liberated from different mineral phases. However, Fe also readily precipitates from water when exposed to oxygen. It is possible that when whole-round interstitial water samples are separated and squeezed, exposure to atmospheric oxygen causes precipitation of variable amounts of dissolved Fe. Furthermore, Fe concentrations may be near the detection limit of the instrument, which is quantified to be 0.9 µM for this site.

Smooth interstitial water Fe profiles have been obtained using Rhizon samples at certain locations (e.g., the Arctic Ocean [Backman, Moran, McInroy, Mayer, and the Expedition 302 Scientists, 2006]). Moreover, the Rhizons can be inserted over very short depth increments in the event that dissolved Fe concentrations truly change over a short vertical distance. It would be intriguing to examine Fe concentrations using Rhizons across short depth intervals. No Rhizon sampling was done at this site, however.

Alkalinity, ammonium, and phosphate

The alkalinity, NH4+, and PO43– profiles at Site U1423 (Fig. F24) are similar to those at Site U1422 (see Fig. F25 in the “Site U1422” chapter [Tada et al., 2015c]) with two notable differences. First, the prominent inflection in alkalinity observed at Site U1422 is not found at Site U1423. Second, through the upper 50 m below the seafloor, concentrations of NH4+ and PO43– are slightly greater at Site U1423 than at Site U1422.

Alkalinity increases from 2.4 mM at the seafloor to 28 mM at 51 m CSF-A. Below this depth, alkalinity decreases steadily to 15 mM at 244 m CSF-A. The profile is concave downward to 50 m CSF-A and slightly concave downward deeper than 50 m CSF-A. The lack of a shallow alkalinity inflection at Site U1423 suggests that a prominent sulfate–methane transition (SMT) does not exist at this location, although a change in slope can be observed at 27 m CSF-A.

Ammonium increases from 93 µM near the seafloor to 2352 µM at 56 m CSF-A. Below this depth, NH4+ drops steadily to ~1776 µM at 244 m CSF-A. The profile is concave downward until 50 m CSF-A but fairly linear below 50 m CSF-A.

Phosphate increases from 1.4 µM at the seafloor to 106 µM at 23 m CSF-A. Below this depth, PO43– decreases steadily to <10 µM at 136 m CSF-A. This decrease in PO43– has a concave downward profile.

Somewhat similar to Site U1422, the magnitude and overall shape of the profiles relate to microbial diagenesis of organic matter and authigenic mineral precipitation. Organic matter landing on the seafloor has a nominal C:N:P ratio of 106:16:1 (Froelich et al., 1979). During sediment burial, Bacteria and Archaea consume the organic matter through a suite of microbial reactions, which release HCO3, NH4+, and PO43– to interstitial water. Dissolution of metal oxides also releases dissolved PO43– to shallow sediment (Cha et al., 2005). At great depth, HCO3 and PO43– precipitate into or onto mineral phases (Berner, 1981).

The higher NH4+ and PO43– concentrations in the upper 50 m at Site U1423 compared to those at Site U1422 suggest greater input of organic carbon from at least the last few million years. However, NH4+ and PO43– concentrations throughout deeper depths are higher overall at Site U1422 than at Site U1423, indicating a relatively higher organic carbon input at Site U1422 on a longer timescale. One possibility is that at Site U1423, although the products of organic decomposition occur in higher concentrations in Holocene and Late Pleistocene sediment, present lower concentrations than at Site U1422 when integrated over a longer time interval. These differences may be explained by a time of slow sedimentation (see “Stratigraphic correlation and sedimentation rates”), when organic inputs accumulated more slowly or when the HCO3, NH4+, and PO43– products were lost through diffusion or mineral precipitation.

Interestingly, the NH4+ concentration gradient at Site U1423 (as well as Site U1422 and other locations; see “Geochemistry” in the “Site U1422” chapter [Expedition 356 scientists, 2015c]) implies a modest upward flux of nitrogen from the interstitial water to the bottom water. Given the previous section and discussion about the Mn profile, we wonder whether NH4+ concentration decreases to zero at the seafloor or, alternatively, at some sedimentary horizon below the seafloor.

Volatile hydrocarbons

Methane is the only hydrocarbon gas in all HS samples at Site U1423. No ethane or heavier hydrocarbons were detected.

In general, CH4 concentrations at Site U1423 (Fig. F25) are much lower than those at Site U1422. Concentrations are nearly zero over the upper 30 m CSF-A. Methane values increase downhole, reaching a peak of 2006 ppmv at 70 m CSF-A. The CH4 peak is prominent, but recall that saturation values of 10,000 ppmv were regularly measured at Site U1422 (see “Geochemistry” in the “Site U1422” chapter [Tada et al., 2015c]). Below 70 m CSF-A, CH4 concentrations decrease rapidly to below 10 ppmv at 155 m CSF-A and remain so to the bottom of Hole U1423B.

Degradation of organic matter produces modest amounts of CH4 between 30 and 130 m CSF-A. Production is probably greatest at ~70 m CSF-A, where peak CH4 concentration occurs and dissolved alkalinity and NH4+ concentrations are highest. Because of the relatively low concentrations, no major gas voids were observed at Site U1423.

A downhole comparison of CH4 and SO42– concentrations (Fig. F26) shows that an SMT occurs at Site U1423. However, this SMT is diffuse, occurring over ~20 m between 30 and 50 m CSF-A.

Sulfate and barium

Profiles of dissolved SO42– and dissolved Ba at Site U1423 (Fig. F27) are somewhat similar to those at Site U1422. The major difference is the shape of the SO42– profile.

Dissolved SO42– concentration in the mudline sample is 28.6 mM, which compares favorably with an inferred 28.5 mM in JSPW (see “Geochemistry” in the “Methods” chapter [Tada et al., 2015b]). From the seafloor to ~50 m CSF-A, values steadily decrease. This decrease in SO42– exhibits a concave downward profile, in contrast to the more linear profile at Site U1422 (Fig. F27). Deeper than ~50 m CSF-A, SO42– concentrations are consistently <0.7 mM.

Dissolved Ba concentrations are <30 µM until ~27 m CSF-A when values steadily rise. The highest Ba standard included in the inductively coupled plasma–atomic emission spectroscopy (ICP-AES) calibration curve was 364 µM, and this value is exceeded at 41.75 m CSF-A. Ba concentrations of each sample below this depth exceed this highest standard and are thus unconstrained. Therefore, any Ba values greater than 364 uM are untrustworthy and Ba is thus reported here as “preliminary.”

In contrast to Site U1422, the upward flux of CH4 and anaerobic oxidation of methane (AOM) play a minor role at Site U1423. Instead, the dominant net sink of SO42– below the seafloor is through microbial decomposition of solid organic carbon. This alternative pathway for sulfate reduction can be expressed in simplified form as follows (Berner, 1980):

2CH2O + SO42– → HS + 2HCO3 + H+.

Because this reaction dominates SO42– consumption and occurs over a thick depth interval, the SO42– profile has significant curvature and the alkalinity profile lacks a prominent kink.

Site U1423 is interesting to studies of Ba cycling in deep-sea sediment. Ba is released into pore water from barite dissolution when SO42– concentrations approach zero. As SO42– is depleted at depth, methanogenesis becomes the primary process oxidizing organic material, and thus high barium concentrations commonly occur with high CH4 concentrations. Site U1423 indicates that high dissolved Ba concentrations can occur without high CH4 concentrations at depth, which is unusual. The lack of methane suggests that fluxes of organic material are relatively low.

Calcium, magnesium, and strontium

At the mudline, dissolved Ca concentration is 10.2 mM, which compares closely to the 10.4 mM inferred in JSPW (see “Geochemistry” in the “Methods” chapter [Tada et al., 2015b]). Calcium concentrations smoothly decrease to ~4.6 mM at 32 m CSF-A (Fig. F28). This is close to but shallower than the depth where alkalinity reaches a maximum (~51 m CSF-A). Deeper than 32 m CSF-A, Ca concentrations steadily rise, reaching 13 mM at the base of Hole U1423B (244 m CSF-A).

Magnesium concentration in the mudline sample (51.7 mM) is lower than that of inferred JSPW (53.3 mM) (see “Geochemistry” in the “Methods” chapter [Tada et al., 2015b]). Below the seafloor, dissolved Mg concentrations steadily decrease to 40 mM at ~22 m CSF-A. Deeper in the sediment column, Mg concentrations continue to decrease, reaching 23 mM at the base of Hole U1423B.

The Sr profile at Site U1423 is much smoother than at Site U1422, presumably because of improved shipboard analytical techniques. Dissolved Sr concentration is 87 µM at the seafloor, which compares to 92 µM inferred in JSPW (see “Geochemistry” in the “Methods” chapter [Tada et al., 2015b]). The shallowest interstitial water sample at 1.45 m CSF-A has a Sr concentration of 93 µM. At ~27 m CSF-A, Sr concentrations begin to increase, reaching 172 µM at ~70 m CSF-A (Fig. F28). The initiation of the Sr rise, and by inference the release of Sr into interstitial water, occurs at approximately the same depth as the minimum in dissolved Ca, the inflection in dissolved Mg, and is ~20 m shallower than the maximum in alkalinity. Below 75 m CSF-A, Sr concentrations continue to increase, but the rate of increase is more gradual. Dissolved Sr is 202 µM at 244 m CSF-A.

As at Site U1422 (see “Geochemistry” in the “Site U1422” chapter [Tada et al., 2015b]), two primary processes impact the Ca, Mg, and Sr profiles at Site U1423. Between ~20 and 40 m CSF-A, there appears to be a phase that consumes Ca and Mg while releasing Sr. It is likely that dolomite [Ca,Mg(CO3)2] is precipitating in this interval, perhaps concurrent with dissolution of CaCO3, which would add Sr to the interstitial water. The anomalous peak in CaCO3 concentrations in the sediment occurs at the bottom of this interval at 41 m CSF-A. We note that the decrease in Mg concentrations (14 mM) is much greater than that of Ca concentrations (6 mM) over the upper 30 m (Fig. F28). Further discussion involving flux-based calculations of a dissolved species into and out of the horizon will aid this interpretation. Specifically, as evidenced by the diffusion gradients, dissolved Ca is also entering the “precipitation” horizon from depth, whereas dissolved Mg is leaving the precipitation horizon to depth. If one considers fluxes above and below the probable reaction interval, the consumption of Ca and Mg from interstitial water is much closer to 1:1, which is consistent with dolomite formation.

The other main process affecting the Ca, Mg, and Sr profiles at Site U1423 is basalt alteration, as was seen in many of the ODP Leg 127/128 interstitial water profiles (Murray et al., 1992). This general reaction consumes Mg and produces both Ca and Sr (Gieskes and Lawrence, 1981).

Salinity, chlorinity, sodium, and pH

Water beneath the thermocline in the marginal sea is often called JSPW. It has a practical salinity (S) of 34.06 ± 0.01 (Sudo, 1986). Once calibrated to molarity, measured concentrations of conservative ions in the mudline sample have values similar to JSPW (Table T11). This includes Cl and Na. We measured concentrations of 541 mM or 547 mol/kg of seawater for Cl and 468 mM or 473 mol/kg of seawater for Na, which compare to 551 mol/kg of seawater Cl and 473 mol/kg of seawater Na for seawater with S of 34.06.

The depth profiles of salinity, Cl, and Na each indicate a zone that is less saline than JSPW between the seafloor and ~52 m CSF-A (Fig. F29). Although salinity, Cl, and Na values are not measured precisely with standard shipboard instruments, they collectively suggest water with salinity ~32 in this depth interval. Interstitial water that is less saline than JSPW was also found in shallow sediment at sites drilled during Leg 127. At the time, the shipboard scientific party attributed the low-salinity waters to sulfate reduction (Tamaki, Pisciotto, Allan, et al., 1990). However, this does not seem a reasonable explanation because although sulfate reduction may decrease the total mass of dissolved ions in seawater, it would not decrease Cl or Na concentrations.

There appear to be two general possibilities for the relatively low salinity water. First, and given the proximity of Site U1423 to land, groundwater could be extending from the coast in relatively permeable horizons beneath the seafloor. This occurs along many continental margins, as seen, for example, at ODP drill sites across the New Jersey margin (Malone et al., 2002; van Geldern et al., 2013). Second, this water represents the evolving remnant of interstitial water that was in diffusive exchange with bottom water of the sea that was fresher sometime in the recent past. Most of the world’s deep ocean water was more saline during the Last Glacial Maximum (LGM) relative to present day. As such, interstitial water salinity usually increases over the uppermost 40–50 m below the seafloor (Schrag et al., 1996; Adkins et al., 2002). However, this marginal sea has a complicated late Quaternary history, in part because shallow sills restrict deepwater flow to and from the adjacent Pacific Ocean (Oba et al., 1991). Low-salinity deep water during the LGM may explain the δ18O-depleted benthic foraminifers found in sediments of this age in this marginal sea (Oba et al., 1991).

Irrespective of origin, the presence of lower salinity water in shallow sediment will enhance deviations in concentration profiles.

Potassium

K concentration (Fig. F29) is ~10.1 mM in the mudline sample, which compares to 10.3 mM for inferred JSPW (see “Geochemistry” in the “Methods” chapter [Tada et al., 2015b]). After an increase to ~11.6 mM at 1.45 m CSF-A, concentrations remain high to deeper than 8.75 m CSF-A when K concentrations decrease steadily, reaching 7.4 mM at 244 m CSF-A. An increase in K concentrations immediately below the seafloor was also found at Site U1422 and may reflect exchange during authigenic mineral formation. The decrease in K concentrations 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 and von Breymann, 1992).

It is noteworthy that K concentrations between 10 and 60 m CSF-A deviate from the general downhole trend to lower values. This deviation may result from the aforementioned freshening.

Lithium and boron

Li concentration (Fig. F30) is 22 µM in the mudline sample. Li values decrease to as low as 15 µM at 13 m CSF-A. Below this depth, Li concentrations increase to 110 µM at 244 m CSF-A.

B concentration (Fig. F30) is 412 µM at the mudline, which is approximately that of JSPW. Values increase to >1072 µM at the bottom of Site U1423 (244 m CSF-A). The profile is very similar to that presented for ODP Site 795 (Brumsack and Zuleger, 1992). They suggested multiple possible explanations for the downhole variations in B. We note that below 25 m CSF-A, B concentrations increase more rapidly than at shallower depths, suggesting that deep microbial reactions may somehow affect B concentrations in interstitial water.

Silica

Silica concentrations in interstitial water at Site U1423 (Fig. F30) increase from initial values of 98 μM in the bottom water (0 m CSF-A) to ~1550 μM at 244 m CSF-A. The opal-A/opal-CT boundary was not drilled at Site U1423, so the expected decrease in dissolved H4SiO4 at depth (Tamaki, Pisciotto, Allan, et al., 1990) was not observed.

The profile of Si measured using ICP-AES and the profile of H4SiO4 measured using the spectrophotometer match fairly well.

Preliminary conclusions

The geochemistry at Site U1423 has similarities and differences to that at Sites 794 and 795, as well as Sites U1422 and U1424. At all these sites, modest amounts of organic carbon land on the seafloor. Microbes then utilize this organic material to drive a series of reactions that conform to an established sequence (Froelich et al., 1979). The sum of the reactions releases HCO3, PO43–, and NH4+ to interstitial water, whereas component reactions affect concentrations of select dissolved constituents. For example, metal oxide reduction releases Mn and Fe to interstitial water and sulfate reduction removes SO42– from interstitial water. The changing interstitial water composition also leads to mineral dissolution and precipitation.

Superimposed on this organic diagenesis framework are three other processes that impact all sites in the region. Two of these processes—alteration of basalt and recrystallization of biogenic silica—occur deeper than 300 m CSF-A but still affect water composition throughout the sediment column. The other apparently important process is changing chemistry at the seafloor, specifically the presence of lower salinity deep water in the recent past.

The primary difference between the sites is the degree of organic diagenesis, at least over long time frames. For example, CH4 levels at Site U1422 are at least 16 times higher than those at Site U1423, and this leads to a prominent SMT and significant AOM at the first location. Potentially, different types of organic matter could cause major variations in alkalinity, NH4+, PO43–, SO42–, and CH4 concentrations, as they are all involved in organic matter degradation. More likely though, differential supply of similar organic matter over time causes the variations at the sites.