Downcore variations in total reflectance and r/b ratio show cyclicities regardless of depth or age. However, the underlying lithologic controls for the observed variations in the color data are poorly known and differ between the Congo and Angola Basins and south of the Walvis Ridge.
Stratigraphic variations of the color data from the dark green-colored sediments from Site 1077 show periodic cycles. Moreover, the periodicities of the total reflectance and the r/b ratio appear to be different (Fig. 4A). Because clay dominates at this site, underlying lithologic changes of the other sediment components are only subtle. Spectral analysis from the spliced records for total reflectance and the r/b ratio was performed only for the first 120 meters composite depth (mcd) because some poor-recovery intervals occur between 130 and 200 mcd. Results from the spectral analysis reveal that the main peaks in the variance power of the color data are not related to the lengths of the sections. Sections were, on average, 1.5 m long. This corresponds to a peak at ~0.66 on the frequency scale (Fig. 4B, Fig. 4C). For the r/b ratio, the variance power is concentrated at 8, 4, and 2.3 mcd periodicities and at 5.5, 3.2, and 2 mcd periodicities for total reflectance (Fig. 4B, Fig. 4C). These periodicities can be interpreted in terms of age, assuming a mean sedimentation rate of 100 m/m.y. at Site 1077 (see Giraudeau et al., Chap. 19, this volume). Both the total reflectance and r/b ratio records display cycles that are close to the orbital periods. The most striking feature is the dominance of the near-23-k.y. periodicity in the r/b ratio, which is observed for different depth intervals (Fig. 4B). Spectral analysis of the total reflectance reveals a more complicated picture showing a larger number of competing cycles in accordance with spectral analysis performed for Site 1075 color data (see Berger et al., Chap. 22, this volume).
Although the relationship between the r/b ratio and organic carbon contents is weak, the dominance of the precession cycle in the r/b ratio suggests a strong influence of productivity variation. The close tie between precession and productivity has been evidenced in this area for late Quaternary sediments (Schneider et al., 1997a). The r/b ratio may be used as an indicator of oceanic productivity in the Congo Basin—at least for the last million years.
Cyclicities obtained for total reflectance are not easily interpreted in terms of climate forcing (Berger et al., Chap. 22, this volume). Further comparison with other parameters is required to deconvolute the influence of the different climate components (oceanic or continental). At Site 1077, total reflectance data show a weak positive correlation with magnetic susceptibility measurements (see "Lithostratigraphy" section, "Site 1077" chapter, this volume). In this area, variations in the magnetic susceptibility have been interpreted to result from the dilution with biosiliceous material, which is stronger during glacial periods and would therefore result in lower magnetic susceptibility values (Schneider et al., 1991, 1997a; Thiessen, 1993). This suggests that biogenic opal and the associated organic matter might influence variations in total reflectance. Additional studies will explore opportunities for detecting climatic effects on land and at sea, based on color reflectance.
For sediments with calcium carbonate contents >20 wt%, total reflectance is a suitable proxy for reconstructing changes in carbonate sedimentation. The spliced records for total reflectance were selected from Sites 1082, 1083, 1084, and 1085, located along a north–south transect from the Walvis Bay to the Southern Cape Basin. The color data exhibit high-frequency-scale variability. Total reflectance may be used as an additional stratigraphic tool between biostratigraphic and paleomagnetic tie-points (Fig. 5A). High total reflectance values are observed between the biostratigraphic ages of 0.46 and 0.26 m.y. and at the Brunhes/Matuyama boundary (0.78 m.y.). These periods of high total reflectance values, occurring in the upwelling-influenced areas (Sites 1082 and 1084) and also in more oceanic-influenced areas (Sites 1083 and 1085), indicate either high carbonate productivity or better carbonate preservation. The carbonate proxy records from upwelling-influenced areas (Sites 1082 and 1084) support a switch from siliceous to carbonate production during these periods.
The total reflectance records vs. age from two high sedimentation–rate sites (1082 and 1083) show that the transition to high reflectance values occurs between 0.6 and 0.55 m.y., close to the Milankovitch/Croll boundary, which marks the boundary to the dominance of the 100-k.y. cycles in the climate records. The transition is sharp at Site 1083 and gradual at Site 1082. Similar changes in the carbonate sedimentation have been observed from color measurements on deep-sea gravity cores from the same area (22°S–35°S; Schneider et al., 1997b). Superimposed on this high-reflectance pattern between 0.5 and 0.26 m.y., fluctuations in the carbonate proxy are observed. The magnitude of these changes appears to be small compared with carbonate records from the Southern Ocean for this time interval; for example, at the transition from marine isotope Stages 12 to 11 (Howard and Prell, 1994; Howard, 1997). Ongoing paleontological studies will improve the stratigraphy for these sites and will allow a more detailed assessment of the history of carbonate sedimentation in this area.