RESULTS AND DISCUSSION

In the depth region studied here, the sedimentation rate is ~30 m/m.y. (from paleomagnetic data: Kanazawa, Sager, Escutia, et al., 2001) and the content of organic carbon ranges from 0.10 to 0.50 wt% (Table T1). Stein (1986, 1990) classified the sediments on the basis of the sedimentation rate and the content of organic carbon. According to this classification, the samples studied here are formed under an open-ocean environment. This is also indicated by the concentration profiles of dissolved sulfate in pore water against depth (Kanazawa, Sager, Escutia, et al., 2001). The dissolved sulfate concentration is >24 mM, which indicates that sulfate reduction is not the predominant mechanism for the decomposition of organics in this region (Fig. F2). Dissolved nitrate is observed even at 5 mbsf, and ammonium increases steadily with depth (Kanazawa, Sager, Escutia, et al., 2001). The maximum of dissolved Mn is found at ~6 mbsf (Fig. F2) (Kanazawa, Sager, Escutia, et al., 2001). These results suggest that the sediment at Site 1179 is in a fairly oxic condition.

The major element composition of the sediment samples is shown in Table T1. The silica content of the sediment samples ranges from 56 to 66 wt%, and the Al₂O₃ content is between 11 and 17 wt%, mainly because of the presence of clay minerals. As expected, the MnO content is highest near the seafloor (0.31 mbsf). This is due to the upward diffusion flux of dissolved manganese within the sediments. Evidence for this is the increase in dissolved manganese with depth up to 6 mbsf (Fig. F2). Near the sediment/water interface, reduced manganese (Mn²⁺) is oxidized to Mn(IV), as shown by the direct observation of the oxidation state of Mn.

The oxidation state of manganese (Mn⁴⁺/Mn²⁺ ratio) can be elucidated by the shift of the pre-edge peak in the Mn K-edge XANES spectrum (Schulze et al., 1995). Qualitatively, it is observed that the whole spectrum shifts to higher energy because of the increase in the attractive force between the electrons and the nucleus when the effective positive charge of the nucleus increases (i.e., at a higher oxidation state). To determine the average valence of Mn, XANES spectra of reference materials made by mixing -MnO₂ and MnSO₄ were measured (Fig. F3A). The pre-edge structure was fitted by the combination of a Gaussian function and a second-order polynomial function (Schulze et al., 1995). The calibration line between the average valence of Mn and the shift of the peak defined by the centroid of the Gaussian function is shown in Figure F3B, suggesting that the average valence of Mn can be estimated by the pre-edge structure. The XANES spectra at the pre-edge region for the sediment samples are shown in Figure F3C. It is clear that Mn is more oxidized at 0.60 mbsf compared with Mn at 1.82 mbsf. The depth profile of the average valence of Mn (Fig. F4) shows that Mn in the sediment is mainly divalent below 0.60 mbsf. The reduced state of Mn at depth, as determined directly by XANES, is consistent with the results of the depth profiles of Mn abundances in sediments and pore water.

Sugisaki and his coworkers (Sugisaki et al., 1982; Yamamoto, 1983; Sugisaki, 1984; Sugitani, 1996) suggested that the MnO/TiO₂ ratio can be used to distinguish pelagic and continental sources of siliceous sediments. Sugitani (1996) studied the MnO/TiO₂ ratios of marine sediments at various distances from land. Sugitani (1996) divided these values, plotted in Figure F5, into two groups: sediments <110 km away from land are hemipelagic, and those >600 km are pelagic. We tentatively define the boundary of these two groups at 400 km, which corresponds to an MnO/TiO₂ ratio of ~0.5. This value can be used for classifying sediments into the hemipelagic and pelagic origins. The MnO/TiO₂ ratio of the samples studied here varies from 0.171 to 10.9 (Fig. F6). If we use the value obtained from the sediment samples of Site 1179, the sediments near the seafloor (above 0.6 mbsf) would be classified as pelagic. However, when we use the sediment below ~1 mbsf, the MnO/TiO₂ ratio is more representative of the hemipelagic region. This is caused by the change in Mn concentration, which is a result of the early diagenesis. We suggest that the high mobility of Mn in the sediment column, confirmed by the depth profiles of Mn abundances in the sediments and pore water and the oxidation state of Mn in the sediments by XANES, can obscure the clear characterization of the depositional environment using the MnO/TiO₂ ratio. It is necessary to consider such a diagenetic effect when employing the MnO/TiO₂ ratio to characterize sediments and sedimentary rocks, as has been shown by Murray (1994) and the multiple references therein.

The degree of the Ce anomaly obtained from REE abundances can be employed as another signature of the depositional environment of siliceous sediment. To compare the MnO/TiO₂ ratio and the degree of the Ce anomaly, we also measured the REE abundances in the sediment samples at Site 1179. The results are shown in Table T2. Some of the REE patterns normalized with post-Archean shales from Australia (PAAS) (Taylor and McLennan, 1988) are shown in Figure F7. PAAS was used for the normalization, because the REE pattern of PAAS normalized with the C1 chondrite (Anders and Grevesse, 1989) creates a smooth REE pattern. Although North American Shale Composite (NASC) (Gromet et al., 1984) is generally used to obtain the ratio of normalized abundances of La and Ce, it was not used here for the normalization of the REE patterns of the sediments. This is because the REE pattern of NASC normalized with C1 chondrite did not create a smooth REE pattern, which suggests that the normalization with NASC is not appropriate for the interpretation of the fine structures of REE patterns. It must be noted that the REE patterns of the sediments studied here are similar, which suggests that the chemical processes in this sedimentary setting were rather similar during the past ~1 m.y. (corresponding to a sediment depth of ~35 mbsf). Slight variations in the REE abundances are observed, although the shapes of the REE patterns are very similar. Among the major elements, TiO₂ shows a slight correlation with REE abundances, represented by the Nd content, as shown in Figure F8, where the data from the Hole 1179B cores are plotted. Since the origin of TiO₂ is considered terrigenous, the abundances of REE indicate the influence of terrigenous material within the sediments. The most striking feature recorded in the REE patterns of sediments is the degree of the Ce anomaly, noted as Ce/Ce* (Shimizu and Masuda, 1977):

Ce/Ce* = Ce_N/(La_N Pr_N)^1/2,

where N indicates the normalized abundance with a proper reference material. According to Shimizu and Masuda (1977), the normalization was conducted with chondrite. A negative Ce anomaly implies that the sediment was formed under a pelagic environment, whereas a positive or no Ce anomaly implies formation at the continental margin. Murray (1994) summarized the abundances of REE and other elements and provided a discrimination diagram by a bivariate plot of La_N/Ce_N and Al₂O₃/(Al₂O₃+Fe₂O₃). La_N/Ce_N normalized with NASC was used here instead of Ce/Ce*, where the meaning of the La_N/Ce_N ratio is similar to Ce/Ce*. The continental margin sediments contain relatively high amounts of Ce, leading to a low La_N/Ce_N ratio, and the pelagic sediments, depleted in Ce, show a high La_N/Ce_N ratio. The present data, when plotted on this discrimination diagram (Fig. F9), are concentrated near the boundary of the two fields of the pelagic and the continental margin. This indicates that the sediments at Site 1179 exhibit intermediate characteristics between pelagic and continental margin origin.

It is possible that the Ce oxidation state can change during diagenetic processes, which may lead to the fractionation of Ce from other REEs and to the alteration of the initial REE signature. We measured Ce L_III-edge XANES to elucidate the cerium oxidation state. The selected spectra, with the results of peak deconvolution by the combination of one arctangent and three Lorentzian functions, are shown in Figure F10. The peak at ~5.726 keV can be ascribed to Ce(III), and the peaks at 5.730 and 5.738 keV are assigned to Ce(IV) (Takahashi et al., 2000a, 2002a). The XANES spectra show that Ce(III) is predominant in the sediment samples at Site 1179, although it is not obvious whether Ce(IV) is incorporated in the sediments because of the low quality of the spectra. However, it seems that some Ce(IV) (<10% of total Ce) is incorporated in the sediments at 0.60 mbsf. This indicates that the redox change of Ce did not affect the REE patterns of the sediment samples, due to possible diagenetic effects, unlike the variation of the MnO/TiO₂ ratio (Fig. F7C). This supports the idea that the REEs, including both Ce(III) and Ce(IV), are fairly immobile during diagenetic processes and that the REE signature is useful to deduce depositional environment of sediments.

In the depth profile of the La_N/Ce_N ratio, a slight decrease in the La_N/Ce_N values near the surface of the sediments was observed, even though the La_N/Ce_N ratio was rather stable throughout the sediment column (Fig. F11). It is possible that the slight variation in La_N/Ce_N values is related to the content of Mn, which increases in the subsurface to give larger MnO/TiO₂ ratios, since the MnO₂ phase can accumulate Ce by oxidative sorption of Ce (Takahashi et al., 2000b, and references therein). The enrichment of Ce induces the decrease in the La_N/Ce_N ratio. However, it must be noted that the increase in the MnO/TiO₂ ratio suggests that the deposition took place in a pelagic region, whereas the decrease in the La_N/Ce_N ratio (or enrichment of Ce), presumably caused by the increase of Mn, indicates that the deposition was at the continental margin. This inconsistency between the La_N/Ce_N and MnO/TiO₂ ratios could result from the misuse of geochemical indicators. The large La_N/Ce_N ratio indicates that the deposition was in a pelagic region because the increase in the La_N/Ce_N ratio with deeper seawater was generally observed (e.g., Piepgras and Jacobsen, 1992). If the REEs in the sediments were mainly associated with a mineral phase like silica, the fractionation of Ce by the sediments would have not occurred, which suggests that the Ce anomaly or the La_N/Ce_N ratio of the sediments was similar to that of seawater. In this case, the large La_N/Ce_N ratio suggests that the deposition of the sediments was in a pelagic region. However, if the REE in the sediments was somehow incorporated in the MnO₂ phase, the La_N/Ce_N ratio, smaller than that of seawater, induced by the enrichment of Ce in the sediments may have led to the continental origin. This discussion implies that the La_N/Ce_N ratio should be employed for siliceous sediments for the interpretation of the depositional environment using the ratio.

The REE signature suggests that the sediment samples from Site 1179 exhibit an intermediate character between pelagic and continental margin sediments. This implication is compared with other geochemical studies. Roser and Korsch (1988) reported a discriminant function diagram for the provenance signatures of sandstone-mudstone suites using major element ratios. Two discriminant functions (DFs) are as below:

DF 1 = 30.638 TiO₂/Al₂O₃ – 12.541 Fe₂O_3(total)/Al₂O₃ + 7.329 MgO/Al₂O₃
+ 12.031 Na₂O/Al₂O₃ + 35.402 K₂O/Al₂O₃ – 6.382.

DF 2 = 56.500 TiO₂/Al₂O₃ – 10.879 Fe₂O_3(total)/Al₂O₃ + 30.875 MgO/Al₂O₃
– 5.404 Na₂O/Al₂O₃ + 11.112 K₂O/Al₂O₃ – 3.89.

This method was to be used when the influence of biogenic CaO and SiO₂ should be excluded to estimate the provenance of the sediments, which is valid for the samples studied here. Based on the two discriminant functions, as shown in Figure F12, our data lie mainly at the boundary of the quartzose sedimentary provenance, the intermediate igneous provenance, and the felsic igneous provenance, suggesting that clastic materials from heterogeneous sources were added to the biogenic silica (diatoms and radiolarians) to produce the sediments at Site 1179. Nakai et al. (1993) studied two sediments in the northwestern Pacific on the basis of REE abundances and Sr and Nd isotopes: one (V20-122) was from 46.3°N, 161.4°E, and the other (RC14-105) was from 39.4°N, 157.3°E. Site 1179 (41.4°N, 159.6°E) is located between these two points. The above study suggested that arc volcanic products and weathered surface matter from the Japanese Islands might be the source of the sediment samples. This is consistent with the interpretation based on the geochemical discrimination diagram, which showed the influence of materials from continents and volcanic products. Further study is required for the distinctive identification of the source material incorporated in the sediments at Site 1179, which show geochemical characteristics pointing to a continental margin, despite the fact that the continents are far away.