Previous   |    Next

doi:10.2204/iodp.proc.310.109.2007

Geochemistry

Hole M0008A was drilled more proximal to the island of Tahiti than any other Expedition 310 site and recovered volcaniclastic sediments. This makes Site M0008 unique, and its interstitial water (IW) geochemistry will be discussed separately. In all of the IW profiles, a barrier to diffusion is evident at ~18 mbsf (80.65 mbsl), which corresponds to the position of a large basalt boulder recovered from Sections 310-M0008A-7R-CC through 8R-1. The lack of a chilled margin at the top and bottom of this basalt suggests that it is not a continuous layer of basalt providing an impervious layer. However, the seismic profile used for site selection indicates a strong continuous reflector at approximately the same depth, suggesting the basalt may represent a continuous layer. Ultimately, with no recovery of Section 310-M0008A-7R-1 above the basalt, the nature of the diffusion barrier remains uncertain.

All geochemical pore water data are presented in tables allsitesIW.xls and IWsaturation.xls in “Supplementary Material.”

pH, alkalinity, ammonia, chloride, and sulfate

Above ~18 mbsf (80.65 mbsl) in Subunit A (see “Sedimentology and biological assemblages”), the pH of IW samples is like ambient seawater, but below that depth, pH decreases sharply downhole until pore waters are slightly acidic (Fig. F17A). The IW alkalinity profile essentially traces that of pH (Fig. F17B). This increase in free H⁺ results in an undersaturation of aragonite and calcite as calculated using PHREEQC software (free from USGS; wwwbrr.cr.usgs.gov/​projects/​GWC_coupled/​phreeqc/) and shown in Figure F17N. Just below ~18 mbsf, significant amounts of ammonia are also detected (Fig. F17C), indicating microbial activity. Ammonia appears to diffuse downward from this source just below the diffusion barrier at ~18 mbsf. Chloride in the pore waters is essentially like seawater with a slight depletion at the bottom of the section (Fig. F17D). This precludes the influence of significant amounts of freshwater in these sediments. Sulfate concentrations do not significantly deviate from seawater values at any depth.

Mg, K, Ca, and Sr

Mg concentrations of IW samples are similar to that of seawater (Fig. F17E), whereas K concentrations become depleted with depth below the diffusion barrier at ~18 mbsf. Both Ca and Sr concentrations become highly elevated with depth below the diffusion barrier (Fig. F17G, F17H), indicating dissolution of carbonate debris and/or weathering of the silicate material. The calculated undersaturation of calcite and aragonite in these pore waters (Fig. F17N) suggests that carbonate dissolution must be contributing to these enrichments.

Li, P, Mn, Fe, and Ba

Li is depleted from its seawater value (~174 µg/L) at all depths in Hole M0008A (Fig. F17I), suggesting Li uptake by clays in the siliclastic sediments (e.g., Zhang et al., 1998). P displays little variability with depth (Fig. F17J). Enrichment of Mn in pore waters is observed in all samples from Hole M0008A, but there is an important source at ~20 mbsf (82.65 mbsl) below the diffusion barrier (Fig. F17K). Fe is greatly elevated in IW samples from above the diffusion barrier at ~18 mbsf but is below detection in samples from below the diffusion barrier (Fig. F17L). This pattern is unexpected because conventional wisdom suggests that Mn oxide reduction occurs above the zone of Fe oxide reduction in marine sediments (e.g., Berner, 1980). One possible explanation for this pattern is that most Fe has already been lost from the older sediments below the diffusion barrier. However, iron oxide–rich sediments containing small root fragments indicative of a laterite soil were recovered in this interval of Fe-free pore waters. It is very interesting that Mn can be mobile in these sediments below the diffusion barrier and Fe is not. Ba is highly enriched in the pore waters below the diffusion barrier at ~18 mbsf (Fig. F17M), leading to a calculated oversaturation with respect to of barite (Fig. F17N).

X-ray fluorescence

Five samples of volcanic sand/​silt intervals and nine individual basalt samples were selected for bulk-rock analysis by energy dispersive polarization X-ray fluorescence analyzer (EDP-XRF) (see “Volcaniclastic sediments” in the “Overview of sites” chapter). All analyzed samples were taken from Hole M0008A. Volcanic sand/​silt samples fall into two compositional groups on the basis of their position relative to the gray to orange color transition at Section 310-M0008A-8R-1, 100 cm. The two “upper” samples (Samples 310-M0008A-4R-1, 22–28 cm, and 5R-1,15–20 cm) are from a continuous interval composed of gray sand and silt, whereas the three “lower” samples were taken in two orange-brown sand/​silt/​clay horizons (intervals 310-M0008A-14R-1, 24–28 cm, and 16R-1, 55–60 cm, from the same horizon, and interval 16R-1, 115–120 cm, from another). Whole-rock samples were taken from the boulder (intervals 310-M0008A-7R-CC, 4–13 cm, and 8R-1, 42–51 cm) and cobble (interval 8R-1, 61–80 cm) in the rubble unit above the gray to orange color transition and from pebbles (intervals 9R-CC, 0–5 cm, 10R-1, 27–30 cm, 14R-1, 0–10 cm, 17R-1, 0–5 cm, 17R-1, 10–15 cm, and 17R-1, 18–23 cm) in the rubble units deeper in the hole (Subunit B; see “Sedimentology and biological assemblages”).

Analytical results are shown in Table T1. Low SiO₂ but fairly high K₂O, P₂O₅, and TiO₂ contents (Na₂O was not analyzed) suggest that the volcanic rock samples belong to the alkalic basalt family (i.e., alkalic basalt, basanite, tephrite, and nephelinite). The two texturally similar samples that apparently come from a single boulder (interval 310-M0008A-17R-1, 10–15 cm) show many compositional similarities, but they also have differences, especially with respect to MgO contents. The differences may be due to flow differentiation, which is known to create compositional variation within a lava flow, or to instrumental error. Two samples (Samples 310-M0008A-8R-1, 61–80 cm and 10R-1, 27–30 cm) are fairly primitive (>14.0 wt% MgO and >350 ppm Ni), but one of them (Sample 8R-1, 61–80 cm) is highly olivine-pyroxene phyric, and therefore its composition may have been changed by crystal accumulation of olivine and pyroxene. Despite their small number, rock samples apparently show downhole compositional variation (Fig. F18A–F18C). Samples from the upper part of the hole have slightly higher incompatible element content (e.g., K and Rb) but lower SiO₂ than those from the lower part of the hole. This indicates that the shallow basalt samples are compositionally more alkalic than the deeper samples.

Volcaniclastic sand/​silt samples are compositionally different from the whole-rock basalt samples in that they have lower SiO₂ contents that translate to lower total weight percent of all major oxides, which is obviously due to higher volatile contents, mainly seawater, as evidenced by their high Cl contents. Sulfur is also high in the sand/​silt samples, particularly in the upper samples. The only rock sample that is high in both S and Cl contents is a scoriaceous pebble (Sample 310-M0008A-9R-CC, 0–5 cm) that probably contains a lot of pore water. Nevertheless, the upper sand/​silt samples show many compositional similarities with the basalts but the lower samples do not. The upper and lower sand/​silt samples show apparent compositional differences (Fig. F19A–F19C). Because downhole compositional variations are observed in pore water chemistry, compositional differences between the upper and lower sand and silt horizons above and below the gray–orange color transition are probably real. Unlike the downhole compositional variation of the basalts, the sand/​silt compositional differences are most probably related to changes in depositional settings.

Both Tahiti-Nui and Tahiti-Iti are predominantly composed of lava flows of moderately alkalic to strongly alkalic basalts; plutonic rocks are subordinate in amount and only exposed near their eroded centers (e.g., McBirney and Aoki, 1968; Cheng et al., 1993; Duncan et al., 1994). Differentiated rocks were erupted mostly in the waning stages of volcanic activity. Thus, the lithological makeup of the volcaniclastic sediments drilled during Expedition 310, particularly the preponderance of lithic basalt clasts, is a direct reflection of the geology of the island. Compositionally, the rock samples and the two upper sand/​silt samples analyzed plot with the rest of the igneous samples from Tahiti (Fig. F20A–F20C). The two primitive basalts plot with the relatively fewer high-MgO basalts from Tahiti, whereas the other basalt samples plot with the more abundant, lower MgO basalts. Many of the lower MgO basalts in Tahiti can be modeled as crystal fractionation daughters of the higher MgO parents (e.g., Cheng et al., 1993; Duncan et al., 1994). The two upper volcaniclastic sand/​silt samples plot between the two basalt groups, either because their lithic basalt components are less fractionated than the lower MgO basalts or because they are similar to the latter but contain additional pyroxene mineral components. The three lower sand/​silt samples generally plot outside the Tahiti lava field, suggesting that their compositions have been modified or altered.

Several detailed geochemical and geochronological investigations have shown temporal evolution in the geochemistry of Tahiti lavas. For example, three volcanic series have been identified in Tahiti-Nui (Duncan et al., 1987). The oldest, or Series A, lavas were erupted before 1.3 Ma and include rocks that now crop out in the center of the volcano. These are predominantly moderately alkalic to transitional basalts that are parental to the few observed tholeiitic basaltic andesites on the island (McBirney and Aoki, 1968). Series B lavas represent the main exposed shield phase of volcanism and were erupted between 1.3 and 0.6 Ma. They consist mainly of moderately alkalic basalts and some of their differentiates. Series C lavas, the youngest, represent a late-stage valley-filling eruption phase that occurred between 0.6 and 0.3 Ma. They consist of highly alkalic lava series such as basanite and tephrite and their differentiates. With decreasing age, therefore, there is an increase in the alkalinity and fractionation of the volcanic lavas erupted in Tahiti-Nui. This is most probably the cause of the increasing alkalic character of the lithic basalt components with decreasing depth (Fig. F17A–F17C). In other words, the older (deeper) basaltic clasts mainly came from the younger, more alkalic Series B or C lavas capping most of the island, whereas the younger (shallower) clasts originated from the older, less alkalic Series A or B lavas. This hypothesis can be tested with detailed petrographic and chemical analyses of these samples and basalt clasts from other sites. Alternatively, the downhole variation is incidental and may be due to a change in the provenance of the basalt clasts.

As mentioned earlier, the downhole compositional change of the volcaniclastic sand and silt is most probably real because pore water chemistries show downhole compositional variation. More importantly, this change is consistent with the visually observed color change from gray to orange and an increase in the amounts of iron oxides/​oxyhydroxides and clay (mud) with depth.

Top of page   |    Previous   |    Next