The inorganic and organic carbon results are compiled from Leg 175 shipboard analyses for Sites 1081, 1082, 1084, and 1085 (Wefer, Berger, Richter, et al., 1998), from shore-based analyses for Site 1087 (Meyers and Robinson, Chap. 2, this volume), and from shipboard (Shipboard Scientific Party, 1984) and shore-based analyses (Meyers et al., 1984; Rullkötter et al., 1984) for Site 532. Subbottom depths have been converted to approximate sediment ages using the shipboard biostratigraphic summaries (Shipboard Scientific Party, 1984; Giraudeau et al., 1998). As more refined postcruise chronostratigraphies become available for the Leg 175 results, the shipboard age assignments for the sites of this drilling leg may undergo minor adjustments. Moreover, as determinations of mass accumulation rates become available, more sophisticated interpretations of the paleoproductivity records than attempted in this summary will become possible.
Three important patterns in the magnitude and variability of CaCO3 concentrations in sediment deposited over the past 10 m.y. emerge from comparison of the results from the six sites. First, concentrations are notably higher in sediments at Sites 1087 and 1085 than at the more northerly locations (Fig. F2). Second, strong minima in concentrations occur in the upper Pliocene and lower Quaternary sections of sediments from Sites 1084, 1082, 532, and 1081 but are weakly expressed or absent in time-equivalent sections of Sites 1087 and 1085. These minima correspond to the early Matuyama Diatom Maximum described by Lange et al. (1999) and indicate an excursion from coccolith-dominated production to one dominated by diatoms. Third, well-developed alternations between high and low CaCO3 concentrations are present at all sites and especially at Sites 1084, 1082, and 532. These variations in carbonate content correspond to light-dark color cycles that are described at these locations (Dean et al., 1984; Diester-Haass et al., 1986; Vidal et al., 1998; Wefer, Berger, Richter, et al., 1998), with the lighter portions of the cycles having higher CaCO3 concentrations.
The different patterns indicate that different combinations of changes in delivery of calcareous material, dilution by noncalcareous components, and carbonate dissolution fueled by oxidation of organic matter or corrosiveness of subsurface water masses occurred at the six locations.
The pattern of increasing TOC concentration since the Miocene that was documented at DSDP Sites 362 and 532 on Walvis Ridge (Siesser, 1980; Meyers et al., 1983) also appears at the sites in Walvis and Cape Basins (Fig. F3). However, this trend is not as well developed at Sites 1087 and 1085 as at the more northerly locations, where TOC concentrations range between 5 and 10 wt% in Pliocene and Pleistocene sediments (Fig. F3). Such concentrations are notably higher than in most deep-sea sediments from the South Atlantic (~0.3 wt%) (Premuzic et al., 1982; Keswani et al., 1984). The elevated TOC concentrations under the Benguela Current reflect the combination of elevated paleoproductivities, which deliver abundant organic matter to the sediments, and the high accumulation rates of sediments, which favors preservation of the organic matter.
Well-developed alternations between higher and lower TOC concentrations exist at Sites 1084, 1082, 532, and 1081 (Fig. F3). Like the similar alternations in CaCO3 concentrations, these accompany light-dark color cycles in the sediments. However, changes in sediment color are not as closely linked to TOC concentrations as they are to CaCO3 concentrations (Dean et al., 1984; Diester-Haass et al., 1986; Vidal et al., 1998). Nonetheless, good correspondence exists between higher TOC concentrations and lower CaCO3 concentrations in sediment from Sites 1084, 1082, 532, and 1081 (Fig. F4). Some of this correspondence is probably related to the Matuyama Diatom Maximum (Lange et al., 1999) and associated elevated production of organic matter, yet some of it may indicate dissolution of CaCO3 as a result of oxidation of organic matter (e.g., Diester-Haass et al., 1986, 1992). An additional factor that is probably involved is downslope transport of organic carbon-rich sediments from the shelf edge that is related to glacial-interglacial changes in sea level (Diester-Haass et al., 1986, 1990, 1992; Summerhayes, et al., 1995).
Organic C/N values calculated for sediment samples using TOC and total nitrogen concentrations vary from 1 to 22 (Fig. F5). Most of these atomic ratios are intermediate between unaltered algal organic matter (5-8) and fresh land-plant material (25-35) (e.g., Emerson and Hedges, 1988; Meyers, 1994). The low C/N values occur in sediment that is poor in organic carbon; these values may be biased by the tendency of clay minerals to absorb ammonium ions generated during the degradation of organic matter (Müller, 1977). Because of their setting offshore of a coastal desert, it is likely that delivery of land-derived organic matter to these sediments has been minimal since the late Miocene initiation of the Namib Desert, which has been linked to the onset of coastal upwelling by pollen contents of Site 532 sediment (van Zinderen Bakker, 1984).
C/N values are higher in sediments richer in TOC than in those lean in TOC (Fig. F5). Many of the values are higher than in fresh algal organic matter. C/N values that are elevated above algal values are common in organic carbon-rich marine sediments (Suess and Müller, 1980; Meyers, 1997). They evidently result from the selective loss of nitrogen as organic matter settles from the photic zone, because nitrogen-bearing proteins are more labile than other organic matter components such as carbohydrates and lipids (Verardo and McIntyre, 1994). This type of preferential nitrogen depletion and consequent carbon enrichment is recognized in other organic carbon-rich sediments, such as those present in the Mediterranean Sea (Meyers and Doose, 1999; Nijenhuis and de Lange, 2000), the equatorial Atlantic (Verardo and McIntyre, 1994), and the northwest Mexican slope (Ganeshram et al., 1999). The C/N elevations are most pronounced when TOC concentrations are highest, suggesting that a higher rate of organic matter delivery leads to diminished organic matter degradation. The existence of the diagenetically elevated C/N values is further evidence that organic matter preservation was enhanced during accumulation of the Pliocene-Pleistocene sediments under the Benguela Current, presumably because of the presence of a strongly developed oxygen minimum zone along the continental margin of Cape and Walvis Basins.
Rock-Eval pyrolysis was originally developed to characterize the organic matter present in oil source rocks, which typically is more thermally mature and present at higher concentrations than commonly found in nonlithified sediments obtained by scientific ocean drilling. Rock-Eval analyses have nonetheless proved valuable in helping to determine organic matter sources in DSDP and Ocean Drilling Program (ODP) samples. The hydrogen index (HI) and the oxygen index (OI) relate to the origin of the total organic matter and are commonly plotted against each other in a Van Krevelen-type diagram in which a comparison of elemental H/C and O/C ratios is approximated. In the HI-OI plot, three main types of organic matter and their thermal alteration pathways are defined. Type I organic matter is especially rich in aliphatic hydrocarbons and hydrocarbon-like components and is derived from algae and microbial biomass. Type II organic matter is moderately rich in these aliphatic components and originates from the waxy coatings of land plants and partially degraded algae. Type III organic matter is poor in hydrocarbon-like materials but rich in cellulose and lignin. Land-plant organic matter is usually rich in these woody components and consequently has lower HI and higher OI than found in lipid-rich and cellulose-poor algal organic matter. Type III organic matter, therefore, usually typifies woody land-plant matter, but it may also represent poorly preserved algal organic matter.
The source distinction between continental and marine organic matter that can be made from the results of Rock-Eval pyrolysis becomes blurred by diagenesis. Oxidation of organic matter affects both HI and OI values. As hydrocarbon-rich organic matter (Type I or II) is oxidized, its hydrogen content decreases and its oxygen content increases and it takes on the HI-OI characteristics of Type III vascular plant organic matter. A further constraint on the use of Rock-Eval pyrolysis for determination of organic matter source is that samples should contain at least 0.5 wt% TOC to yield meaningful results (Katz, 1983; Peters, 1986).
The results of Rock-Eval analyses of Leg 175 organic carbon-rich sediments show that their organic matter content appears to be dominated by varying mixtures of Type II algal material and Type III land-plant material (Fig. F6). The indication of large proportions of land-derived organic matter in many of the sapropels conflicts, however, with C/N ratios in these sediments that are too low (<20) for land-plant organic matter (Wefer, Berger, Richter, et al., 1998). The contradiction between the Rock-Eval source characterization and the elemental source characterization is evidence that the marine organic matter has been moderately oxidized, because well-preserved Type II organic matter has high HI values (Espitalié et al., 1977; Peters, 1986).
Comparison of TOC concentration with Rock-Eval HI values shows that sediment samples with higher TOC concentrations also have higher HI values (Fig. F7), which is a pattern also found in sapropels in the Mediterranean (Bouloubassi et al., 1999; Meyers and Doose, 1999). Higher TOC concentrations also correspond to higher C/N ratios (Fig. F5). As a consequence, sapropels having higher HI values also have higher C/N ratios, which is an indication of partial but not extensive alteration of marine organic matter during sinking and incorporation into bottom sediments. The correspondence between higher Rock-Eval HI values, higher elemental C/N ratios, and higher TOC concentrations suggests that the rate of export production and the degree of preservation of marine organic matter in bottom sediments are related and have increased as algal productivity has increased since the Miocene.
Rock-Eval Tmax values are generally <425°C in the Miocene-Pleistocene sediments from the locations sampled during Leg 175 (Wefer, Berger, Richter, et al., 1998; A. Rosell-Melé, unpubl. data). These values indicate that the organic matter in sediments from Cape and Walvis Basins and Walvis Ridge is thermally immature with respect to petroleum generation (Espitalié et al, 1977; Peters, 1986). Because organic matter is sensitive to temperatures that are only slightly elevated (>60°C), the thermal immaturity of the Miocene-Pleistocene organic matter is evidence of low heat flows along the margin of the southeastern South Atlantic Ocean since the times these sediments were deposited. Moreover, the thermal immaturity also is evidence of minimal contributions of recycled detrital organic matter derived from erosion of ancient sedimentary rocks present on the African landmass.
TOC concentrations in sediments younger than the latest Miocene are larger and more variable than those in older sediments (Fig. F3). Maximum TOC concentrations appear within the last 2 m.y. at most sites but not in the most recent sediment, suggesting that upwelling once induced greater biological productivity than at present. Moreover, Pliocene-Pleistocene sediments at Sites 1087 and 1085 are not nearly as rich in organic carbon as those closer to Walvis Ridge (Wefer, Berger, Richter, et al., 1998), indicating that upwelling in most of Cape Basin has never created similarly elevated levels of productivity as at the more northerly locations. The greater concentrations of CaCO3 present in sediments at Sites 1087 and 1085 (Fig. F2), and by implication less opal, is further evidence of the absence of similarly elevated productivity at these locations over the past 10 m.y.
The well-developed variations in TOC concentrations that are present in Pliocene-Pleistocene sediments at Sites 1084, 1082, 532, and 1081 have been postulated to represent glacial-interglacial changes in the intensity and locus of the Benguela Current (Diester-Haass et al., 1990, 1992). This interpretation was necessarily constrained to essentially a single location (Site 362/532) by availability of cored sequences. The expanded geographic coverage provided by the multisite comparison in this summary shows that upwelling has not migrated up the coast as the Benguela Current has evolved but has instead remained in the region of Walvis Basin and Walvis Ridge, where the trade winds are able to transport surface water offshore. The TOC variations consequently are probably the result of glacial-interglacial variations in trade wind intensity and in sea level rather than reorientation of the Benguela Current.