Much of the hemipelagic sediments and most of the pelagic sediments recovered during Leg 175 consist of calcareous nannofossils with various admixtures of foraminifers. The varying carbonate content at any one site reflects production in overlying waters (and benthic production), dilution with noncarbonate particles, and dissolution. In most cases, dissolution is the process most responsible for the variations seen. This is true even in rather shallow water depths, as at Site 1081 (805 m). At this site, abundances of calcareous fossils, although tracking one another, have no correlation with those of siliceous fossils (Fig. 9).
Two factors dominate preservation (or dissolution) of calcareous fossils: the saturation of waters in contact with the seafloor, which provides the boundary conditions, and the diagenetic reactions within the uppermost sediment, which determine the chemistry of pore waters in contact with the calcareous particles. In low-productivity situations, deep-water properties are important, and their changes are recorded in terms of fluctuations in preservation. In environments characterized by high production, diagenetic processes will dominate, so that deep-water properties will imprint less prominently on the sediment.
What we expect to see regarding the deep-water environment is an overall trend of increased carbonate preservation since the Miocene carbonate crisis ~12 m.y. ago (Fig. 10). This crisis, which represents an excursion of the pelagic carbonate compensation depth (CCD) to elevations close to the crest of the Mid-Atlantic Ridge, was first identified in the South Atlantic based on results from DSDP Leg 3 (Fig. 10, "Bg72"; Berger, 1972). Subsequent expeditions allowed considerable refinement (Fig. 10, "H&W85"; Hsü and Wright, 1985). The drop of the CCD since the beginning of the late Miocene largely represents the increased intensity of NADW production. Seafloor bathed by NADW preserves carbonate, whereas seafloor bathed by Antarctic Bottom Water does not. The turning up of NADW some 10 m.y. ago is likewise recorded in the overall switch of opal deposition out of the North Atlantic into the North Pacific (Keller and Barron, 1983; Woodruff and Savin, 1989).
The available data do not show the late Neogene trend of increased deep-water carbonate preservation (Fig. 11, Fig. 12, Fig. 13). There is one exception: at Site 1085 (1713 m), carbonate values are distinctly low, (between 9 and 8 Ma), and they rise to a high level shortly after. Site 1086 (at 794 m) is too shallow to trace deep-water properties; carbonate data for Site 1087 (at 1386 m) are not yet available.
Visual comparison of carbonate stratigraphies at the sites on and south of the Walvis Ridge indicate a number of low-value intervals, which can be interpreted as periods of enhanced dissolution (Fig. 11). The first of these, going back in time, is between 0.6 and 0.7 Ma and is perhaps associated with glacial Stage 16. The second is centered between 1.7 and 2 Ma, marking the onset of the Quaternary. Lack of detailed sampling prevents identification of other, older events. There is a long-term trend toward lower values, from 6 to 2 Ma, seen at Sites 1081 and 1082. Presumably, this trend parallels increased productivity (and carbon dioxide production within sediments). Also, there is an overall increase in carbonate values throughout the Quaternary period, suggesting a general decrease in productivity or an increase in intermediate water and upper deep-water saturation.
Nannofossil abundances show striking variability at some sites (the Congo transect, Site 1081). Site 1081 has a nearly barren zone within the lower Pliocene section; at Site 1082, a similar zone is centered on 3 Ma. This suggests high organic productivity at the boundary between the early and late Pliocene, just before the MOM. A shorter barren zone is seen at Site 1081, near 1.5 to 1.6 Ma. Also, at Site 1075, there is such a zone at the base of the section recovered between 2 and 1.5 Ma. Abundance estimates are not sensitive enough, nor sampling dense enough, to provide more detail on dissolution zones or trends.
Benthic foraminiferal abundances show some of the same patterns as carbonate percentages and nannofossils. Again, there are low values in the early Quaternary (Sites 1075, 1077, 1081, and 1083) and around 3 Ma (Sites 1081 and 1084). Observations on the preservation of planktonic foraminifers at Site 1084 support the general impression that calcareous fossils are dissolved in those sections where productivity is high. The lack of sensitivity of the abundance index and the wide spacing of samples make it difficult to extract a more detailed dissolution stratigraphy.
In summary, there is some indication that there were two extended dissolution pulses, one centered near 3 Ma, the other in the early Quaternary. It is likely that these were connected to times of high productivity. If so, this would suggest that productivity decreased somewhat throughout the late Quaternary (as also suggested by the diatom record; see Fig. 14). Organic matter abundance would not necessarily show this reduction in productivity (Fig. 6) because of diagenetic processes that result in an overall decreasing preservation of organic matter with depth below seafloor, which yields a false pattern of productivity increase within the Quaternary.