We summarize here the principal results of the shipboard interstitial water program. At many of the sites, either specific sampling strategies were implemented or unique chemical behavior(s) were discovered (e.g., at Sites 1077 and 1079; see Table 1), and additional details of the chemical distributions within the interstitial waters may be found in the individual site chapters (this volume). Of particular interest are the extremely high alkalinities and large concentrations of NH4+ observed at Site 1084 in the Northern Cape Basin. This unique site is discussed separately below.
Throughout the Benguela upwelling system, the rates of supply and decomposition of biogenically produced organic matter exert a profound influence on the concentration-depth gradients of many dissolved chemical species, including alkalinity, SO42–, NH4+, PO43–, and H4SiO4. With the high concentrations of organic carbon in these sediments (Meyers et al., Chap. 21, this volume), the production of alkalinity in the uppermost ~50 mbsf reflects the progressive degradation of organic matter in the youngest sediments. Inter- and intrabasinal contrasts in the alkalinity profiles (Fig. 2) most likely reflect subtle variations in organic matter reactivity and in sedimentation rate. The observed increases in alkalinity are tied to the total production of CO2 by reduction/consumption of SO42– and other electron acceptors. The asymptotic distribution of alkalinity at depth confirms that the dominant chemical process affecting its distribution is the destruction of organic matter in the shallowest sediments. At some sites, however, carbonate reprecipitation reactions also affect the alkalinity profile, albeit to a lesser degree, as discussed in the individual site chapters (this volume).
The suboxic degradation of organic matter (Froelich et al., 1979) also consumes dissolved SO42– via the reaction (assuming a Redfield stoichiometry)
Information may be gained by examining the relationship between sedimentation rate and the rate of SO42– consumption (Fig. 3). In theory, all other parameters being equal (notably, the concentration and type of organic matter), complete SO42– consumption at drilling sites with slower sedimentation rates should occur at greater depths than at sites where faster sedimentation occurs (Berner, 1980). At sites with slower sedimentation, the downward diffusion from the overlying seawater reservoir will replenish that SO42– consumed during early degradation of organic matter, which, in turn, delays the successive onset and completion of SO42– destruction. Thus, the sedimentation rate and apparent rate of SO42– consumption are often positively correlated. In regimes where such a relationship is not observed, other factors, such as the nature of the organic matter, must be more important than the sedimentation rate control.
Within the Congo Basin, the depth to the complete consumption of SO42– is relatively uniform (Fig. 2; Table 1), and a positive relationship is observed between the sedimentation rate and the SO42– consumption rate, although gradients at the three sites are quite similar (Fig. 3; Sites 1075, 1076, and 1077). As outlined above, this implies that the variation in organic matter among these sites is minimal. Such an argument can also be made, albeit less convincingly, for the relative distributions at the Walvis Ridge/Basin sites (Fig. 3; Sites 1081, 1082, and 1083). In contrast, at Site 1079 within the Angola Basin, the consumption of SO42– is not achieved until 50 mbsf, which is significantly deeper than at Sites 1078 and 1080, also located in the Angola Basin (Fig. 2; Table 1). It is clear that sedimentation rate is not the dominant controlling parameter of organic degradation in the Angola Basin because Site 1079 records the deepest penetration of dissolved SO42– (Fig. 2), yet has the highest sedimentation rate of the three Angola sites (Fig. 3). Finally, the depositional regimes within the Cape Basin display the largest variability of organic matter diagenesis of the entire leg. Because of the extremely high organic matter loadings at Site 1084 (Fig. 2), SO42– is completely consumed within the uppermost 5 mbsf (as described below), whereas at Site 1086, the very high carbonate concentration, the low organic matter concentration, and the low sedimentation rate (Fig. 3) combine to allow the penetration of SO42– to the relatively great depth of ~200 mbsf. Overall, when one considers the entire database of the southwest African margin sites drilled during Leg 175—with the exception of Sites 1080 and 1084 (Fig. 3, inset)—there is a broad positive relationship between sedimentation rate and SO42– reduction, indicating that the general relationship between these two parameters holds true and that the local variations caused by contrasts in export production modify, but do not obscure, this overall pattern.
Variations in organic matter diagenesis suggested by the distributions in alkalinity and dissolved SO42– are also confirmed by the downhole variations in dissolved NH4+ (Fig. 2). Like CO2, NH4+ is a product of organic matter degradation, but it is produced more continuously than alkalinity throughout a sediment column. Ammonium also is involved in clay exchange reactions and can either be released through exchange with K+ or be taken up during authigenic clay formation (e.g., Stevenson and Cheng, 1972). The sites within the Congo Basin record essentially identical downhole dissolved NH4+ patterns, which are more subdued than at the Angola and Walvis sites (Fig. 2). The largest increases in dissolved NH4+ are found at Site 1082 (Wal-vis Basin) and Site 1084 (Cape Basin), which are two of the sites with the highest concentrations of organic matter. Those sites with the deepest SO42– penetration (e.g., Sites 1079, 1081, and 1086) also record the smallest increases in dissolved NH4+. In general, as with the SO42– and alkalinity profiles, these inter- and intrabasinal variations in dissolved NH4+ speak to the relative similarity among the three sites drilled within the Congo Basin, the heterogeneity within the other basins, and the overall strong influence of sedimentary organic carbon on the interstitial water chemistry.
Dissolved PO43– and H4SiO4 distributions within interstitial waters also reflect organic and biogenic inputs, remineralization, and dissolution. At each site, dissolved PO43– increases to maximum values usually within the uppermost ~50 mbsf (Fig. 4), reflecting degradation of organic matter. Below this maximum, dissolved PO43– decreases to minimal values at depth, suggesting authigenic uptake into apatite phases. Postcruise studies focusing on the specific rates of PO43– generation, when interpreted in the context of the site-by-site variations in alkalinity and NH4+ distributions, will provide additional quantitative constraints regarding the remineralization of organic matter.
Dissolved H4SiO4 in the interstitial waters along the southwest African margin is always greater than that in average seawater (Fig. 4), indicating an additive contribution of H4SiO4 from the dissolution of diatoms (see Giraudeau et al, Chap. 19, this volume). As with many other dissolved constituents, the distributions of dissolved H4SiO4 between the Congo Basin sites are comparable, providing further confirmation of the broad geochemical similarity among these sites. Other sites record varying degrees of shallow enrichments, diatom dissolution, and deeper flattenings (or decreases) of the chemical gradient (see individual site chapters, this volume). At these greater depths, decreases in dissolved H4SiO4 most likely are responding to authigenic clay formation. Also, stratigraphically local variations in the sedimentary diatom abundance exert a strong control on the concentration of dissolved H4SiO4 (see Giraudeau et al, Chap. 19, this volume).
The discovery of widespread and pervasive dolomite layers throughout the southwest African margin was one of the principal findings of Leg 175 (Wefer et al. Chap. 16, this volume). The distributions of dissolved Sr2+, Ca2+, and Mg2+ provide important constraints on the diagenetic dissolution of biogenic carbonates as well as the secondary precipitation of authigenic dolomites (Fig. 5).
Increases with depth of dissolved Sr2+ indicate the dissolution of biogenic calcite, which releases Sr2+ to the interstitial waters. Throughout the Benguela Current system, there appear to be two modes of such diagenetic release. First, as typified by Site 1077 in the Congo Basin (Fig. 5, inset), many sites display a rapid increase in dissolved Sr2+ through the uppermost 50 mbsf, followed by gradual increases to total depth. Even at locations with only minimal absolute increases in dissolved Sr2+ over the entire recovered section (such as within the Congo and Angola Basins and at Site 1084 in the Cape Basin), there often is a shallow rapid increase downhole. Second, as typified by Site 1086 in the Cape Basin, several sites display an essentially linear increase with depth of dissolved Sr2+, without the rapid, shallow increase, indicating the continual dissolution of biogenic calcite with depth.
In the absence of significant exchange with either igneous basement rocks or volcanic ash dispersed through the sediment column, distributions of dissolved Ca2+ and Mg2+ predominantly reflect carbonate recrystallization processes. These Ca2+ and Mg2+ changes are best interpreted in the context of the Sr2+ variations discussed above, which indicate dissolution of biogenic calcite throughout the sediment column. Despite this dissolution, which will serve as a source of dissolved Ca2+, at essentially all sites the concentration of Ca2+ decreases through the uppermost ~50 mbsf. This decrease reflects the precipitation of diagenetic carbonate, driven by the marked increase in alkalinity through the same portion of the section at each site (Fig. 2).
Sinks for dissolved Mg2+ in this environment include dolomite precipitation and uptake by clay minerals. In all likelihood, both sinks are operating through the recovered sequences, and the challenge is to resolve the relative contribution of each. Following the logic of Baker and Burns (1985), it appears most likely that dolomite precipitation is occurring according to
As Baker and Burns (1985) note, this stoichiometry is intended to be representative, not unique, of dolomite formation by an intermediate process involving both CaCO3 (calcite or aragonite) dissolution and the supply of dissolved species from solution.
We are able to identify more specifically the depth ranges over which dolomitization is occurring by detailed investigation of the dissolved Mg2+ profiles (Fig. 5; also see volume site chapters). Many sites record a significant negative excursion in the dissolved Mg2+ profile superimposed on a broad linear decrease from seawater values at the surface to minimum values at total depth. At a given site, the location of this negative excursion (Fig. 5, shaded region of Mg2+ profile at Sites 1075 and 1082) coincides with the minimum in dissolved Ca2+, as well as with the alkalinity maximum. Assuming that the linear Mg2+ decrease represents a background level of uptake by clay minerals, the negative excursion appears to delineate the depth range through which dolomite precipitation is occurring. These depth intervals also are indicated by a minimum in the depth profile of the Ca2+/Mg2+ (molar) ratio (Fig. 5). If correct, this interpretation confirms the relationships among organic matter degradation (acting as a bicarbonate source), CaCO3 dissolution (recorded in the Sr2+ profile), and secondary dolomite precipitation (delineated by the negative Mg2+ excursion and the Ca2+ minimum).
Site 1084 possesses several qualities that justify its separate consideration (Fig. 6). Because of the extremely high levels of sedimentary organic matter at this site, the alkalinity and dissolved NH4+ concentrations are the second highest ever observed in the history of Deep Sea Drilling Project (DSDP)/ODP drilling, being exceeded only by those observed at ODP Site 688 along the Peruvian margin (Kastner et al., 1990). Sulfate is completely consumed within the uppermost 5 mbsf (Fig. 2), and dissolved PO43– reaches a maximum value (540 µM) that is far greater than that found at any other Benguela site (Fig. 3, Fig. 6).
The carbonate system at Site 1084 is also affected by these organically mediated constituents. Authigenic carbonate precipitation apparently begins both more intensively (causing a lower Ca2+ minimum) and earlier (at a more shallow depth) than at any other Leg 175 site, as indicated by the extreme drawdown of dissolved Ca2+ within the uppermost 5 mbsf (compare Fig. 6 and Fig. 5). Also, whereas the dissolved Mg2+ at all other sites records linear decreases on which are superimposed negative excursions, the depth profile of dissolved Mg2+ at Site 1084 increases to maximum values at 80 mbsf (Fig. 6). Because of the strong similarity between the alkalinity and Mg2+ profiles, we suggest that this increase may reflect complexation of Mg2+ with the ultra-high levels of bicarbonate (J. Gieskes, pers. comm., 1997). Also, at Site 1084 dissolved Cl– records a decrease with depth, unlike at any other site (Fig. 7).
The behavior of dissolved H4SiO4 (Fig. 3) and dissolved Sr2+ (Fig. 5, Fig. 6) sheds further light on the chemical processes acting at Site 1084. Whereas the dissolved H4SiO4 profile plateaus relatively shallowly at Site 1084, the maximum value is not significantly greater than that at the other Cape Basin sites, nor is it significantly greater than that observed elsewhere along the southwest African margin (Fig. 3). Also, the dissolved Sr2+ profile is actually lower here than at any other Cape Basin site and is approximately equal to that observed elsewhere (Fig. 5, Fig. 6). These two observations, in concert with the relatively modest CaCO3 concentrations in the sediment (Fig. 3), suggest that the elevated sedimentary organic matter concentrations at Site 1084 reflect the primary accumulation of organic matter, rather than organic matter uniquely associated with diatomaceous or carbonate plankton.
At most sites drilled during Leg 175, dissolved Cl– records a pronounced subsurface maximum found in the upper ~50 mbsf (Fig. 7). Although the sharpness of this maximum and its exact depth varies somewhat from site to site, its presence through a variety of depositional and diagenetic regimes suggests that the occurrence is regionally significant and relevant. Because Cl– generally behaves conservatively in the absence of significant clay dewatering reactions or brine influences, we interpret these increases as records of a diffusionally damped signal of the chlorinity of seawater during glacial periods. Because glacial periods last longer than interglacials, the putatitive Cl– signal may reflect a stacked, multiglacial signal. Postcruise studies of δ18O variability at the high-resolution sites (Table 1) will shed further insights into the preservation of glacial chemical signatures.