The record of productivity of the Angola-Benguela upwelling system resides in a number of indicators, most prominently in organic matter and opal accumulation, and in the various chemical changes driven by bacterial activities and other diagenetic reactions. On the basis of these indicators, the sites can be ranked according to the overall productivity of overlying waters, and trends in productivity can be discerned.
Starting in the late Miocene, the strength of the Benguela Current has continuously increased. This is apparent in the history of eastern boundary upwelling in the South Atlantic (Siesser, 1980; Meyers et al., 1983). The record of organic matter deposition in the Angola Basin and on the Walvis Ridge suggests that rates of upwelling accelerated near 6 Ma and near 3 Ma, but held steady (or decreased somewhat) within the Quaternary (Fig. 5).
A general increase in deposition of organic carbon is evident at all sites on and south of the Walvis Ridge (Fig. 6; north of the ridge, penetration is insufficient to say). There is some indication for the steps seen earlier at 6 Ma (Site 1082) and 3 Ma (Sites 1081, 1082, and 1084). However, substantial variability exists, and the widely spaced sampling intervals make it difficult to recognize any steps. Within the Quaternary record, no clear trends are apparent. Maximum values in sediments of late Quaternary age in the sites north of the Walvis Ridge may be sampling artifacts (extremes are more likely where sampling is dense, other factors being equal). The early Quaternary maximum suggested in the DSDP data of Meyers et al. (1983; Fig. 5) is seen at Site 1081 and, perhaps, Site 1083, but not at Sites 1082 and 1084. Site 1085 also has a maximum value in the early Quaternary.
The overall trend of increasing organic carbon deposition is difficult to interpret in any detail. Decay of organic matter continues deep within the sediment, so that a trend toward lower values with depth below seafloor is expected from diagenesis alone. The most obvious sign of decay of organic matter at depth was the high gas content of hemipelagic sediments. Biogenic methane and carbon dioxide were roughly equally important at most sites, based on headspace measurements (Fig. 7). The methane production only sets in after the available sulfate in interstitial waters has been used up for oxidation of organic matter. Carbon dioxide is generated throughout the zone of bacterial activity. The presence of a methane maximum (100 to 150 m at Site 1081) suggests that the rate of destruction of organic matter decreases considerably below 125 m.
The high gas content in most of the cores recovered during Leg 175 made handling and measuring physical properties difficult. However, it also kept sediments from compacting as quickly as they might have otherwise, thus allowing much faster drilling than anticipated. The site with the highest TOC values is Site 1084, off Lüderitz, at the edge of the most active Namibian upwelling cell. Its sediments showed extremely high diagenetic activity (with a mixture of malodorous gases emanating from the cores, in addition to methane and carbon dioxide).
The rapid reduction of sulfate in the uppermost part of the sediment column attests to the vigorous activity of bacteria within the sediment (Fig. 8), as do the extremely high ammonia values in the interstitial waters, especially at Site 1084. High alkalinity of interstitial water results from these processes, which leads to the precipitation of carbonates including dolomites. Intense degradation of organic matter is also indicated by the rapid increase of phosphate values below the seafloor. Formation of apatite and other phosphatic minerals presumably accounts for the decrease of phosphate values at depth. Apparently, dissolved silicate values are similarly constrained, but at a rather well-defined range of values (near 1000 µM), suggesting precipitation of Opal CT or uptake by clay minerals. This indicates that diatoms also are subject to long-term destruction.
Given the pore-water measurements concerning sulfate, ammonia, phosphate, and silicate, we can rank of the Leg 175 sites in terms of the intensity of organic-driven diagenesis. We use reported values at 10 mbsf, as well as the depth at which the sulfate ion drops below 1 mM. Sulfate, ammonia, and phosphate values correlate well among the sites (r is between 0.7 and 0.8). However, the correlation between the silicate values and the other variables is rather poor. Thus, the intensity of organic-driven reactions will not predict silicate values well (also illustrated in Fig. 8).
The ranking of the sites in terms of intensity of diagenetic reactions (driven by organic matter supply) is as follows: (1) Site 1084; (2) Site 1078; (3) Site 1076; (4) Site 1077; (5) Site 1083; (6) Site 1082; (7) Site 1075; (8) Site 1079; (9) Site 1085; (10) Site 1081; (11) Site 1087; and (12) Site 1086. Site 1080 is not ranked. The first six sites in this list have intense diagenesis and may be considered "high-productivity" sites.
The ranking in terms of dissolved silicate at 10 mbsf is as follows: (1) Site 1075; (2) Site 1083; (3) Site 1082; (4) Site 1084; (5) Site 1077; (6) Site 1076; (7) Site 1081; (8) Site 1085; (9) Site 1078; (10) Site 1087; (11) Site 1086; and (12) Site 1079. Site 1080 is not ranked. The first six sites in this list show elevated levels of silicate and may be considered "high-opal" sites.
Sites that appear in the first six ranks of both lists are (1) Site 1084; (2) Site 1083; (3) Site 1076; (4) Site 1077; and (5) Site 1082.
Sites that appear in the last six ranks of both lists are (1) Site 1079; (2) Site 1085; (3) Site 1081; (4) Site 1087; and (5) Site 1086.
Thus, on the whole, high-productivity sites also tend to be high-opal sites, although this is by no means a strict correspondence. The low-productivity low-opal sites compose the set of pelagic carbonates, except for Site 1079, off Lobito, with its high rate of terrigenous sediment accumulation.
The ranking given above, being derived from indices dependent on reactions within the upper ~50 m of the sediment column, refers to Quaternary conditions. Concerning the abundance of diatoms within Quaternary sediments, it is seen that the high-opal sites (as identified by dissolved silicate in pore waters) are the high-diatom sites as well. The exact rank order depends on which criteria are used (average diatom abundance, last 2 m.y., or maximum abundance, for example). In any event, Sites 1075, 1082, 1083, and 1084 would be at the top. As mentioned, the Congo Fan (Sites 1075, 1076, and 1077) was identified previously as a region of high opal deposition (as mentioned above). River supply and peri-estuarine pumping of silicate-rich subsurface waters are the processes responsible for high opal deposition. Sites 1082, 1083, and 1084 reflect the effects of coastal upwelling and associated eddies and filaments penetrating the Benguela Current.
All sites on and south of the Walvis Ridge tend to show a maximum abundance of diatoms near and just before 2 Ma. This is the MOM, first discovered on the Walvis Ridge by Leg 75 scientists (Shipboard Scientific Party, 1984). We confirm it here as a phenomenon common to the entire Benguela upwelling system and explore it in some detail.