Flinn-diagrams (L vs. F, Fig. 1) for all measured samples generally show distributions along the abscissa (F axis), indicating oblate ellipsoid dominance, although samples close to the point of intersection show shifting toward the ordinate (L axis). These later fabrics represent a more spherical and nonpreferential alignment of grains. To observe the shape of the fabric in detail, downhole profiles in each hole are used (Fig. 2). They cover lithostratigraphic Units I and II, from Pleistocene to late Paleocene time.
In Holes 897C and 897D, the values of parameter L range from 1.00 to 1.01 and are fairly uniform (Fig. 2). No linear downcore increase is observed in F value. F values at intervals within Unit I and the lower part of Subunit IIC range from 1.00 to 1.07. Conversely, F values of Subunit IIB and the upper portion of Subunit IIC are fairly low. The inclination of Kmin shows a reversed trend against F value. The inclination of the Kmin axis ranges from 60° to 90°, but an interval of Subunit IIB (300-360 mbsf) shows shallower and varied inclinations. Undisturbed fine sediment should show foliation and Kmin inclination perpendicular to the bedding plane. The fabric of low F value and shallow Kmin inclination suggests that the fabric was disturbed after deposition. A distinct offset is observed at the boundary between Units I and II (300 mbsf). At the same horizon, the porosity trend also shows an offset (Sawyer, Whitmarsh, Klaus, et al., 1994). Although the porosity generally decreases downhole because of compaction, the F value shows no downcore change. This suggests that the magnetic foliation is not developed by compaction. The correspondence of offset horizons may be an important indicator of sediment properties.
Unit I from Site 898 shows fairly low F values, ranging from 1.00 to 1.02. However, F values in Unit II are scattered from 1.00 to 1.06. The Kmin inclination is negative, coincident with the trend of the F value. The low F values are interpreted to indicate postdepositional disturbance as discussed above. Data from Hole 899 are inadequate and do not allow vertical fabric variations to be interpreted.
At Site 900, low F values are observed in the intervals between 120 and 230 mbsf and between 440 and 560 mbsf. F values are scattered in the intervals between 250 and 400 mbsf and above 100 mbsf. A distinct offset in F values at 250 mbsf corresponds to the lithostratigraphic boundary between Subunits IIA and IIB (234.3 mbsf), and at this horizon, the porosity trend shows the reverse offset to that of the F value. The interval between 440 and 560 mbsf shows an interesting change in F value trend. The trend of F value seems to be offset, whereas the slope of the porosity trend changes slightly. The change in sediment property may be related to an unconformity, inferred by the absence of a lower Oligocene microfossil zone (Sawyer, Whitmarsh, Klaus, et al., 1994) at about 470 mbsf.
Two characteristic F value intervals are observed at each site: a fairly low F interval (1.00-1.02), and a scattered and higher F value interval (1-1.07). Except at Site 900, the low F value intervals are characterized by shallow Kmin inclination, which suggests that the fabric was disturbed after deposition. Bioturbation may have destroyed the sedimentary fabric. Subunit IIA of Hole 897C, which shows disturbed fabric, also shows intensive color mottling of Zoo-phycos, Chondrites, and Planolites. However, the bioturbation in Units I through II is commonly observed to a varying degree, which is not strong evidence that the low F value is the result of bioturbation. Drilling disturbance may have occurred at Site 898. The upper portion of Unit I at Site 898 is clearly disturbed, especially the first 17 cores (0-158.1 mbsf). Most structures are smeared out. The scattering of the porosity data above 50 mbsf probably indicates drilling disturbance.
Tectonic disturbance can also affect the observations. In Units I and II, some cases of structural deformation were observed. Cores 149-900A-26R and 27R (234.1-240.0 mbsf) contain faulted and folded beds that were formed before sediment burial (Sawyer, Whitmarsh, Klaus, et al., 1994). At this interval, only two shallow Kmin inclination data are observed and these may show tectonic contributions. However, the low F values in other intervals are not of tectonic origin.
Intervals of low F value and high Kmin inclination at Sites 897 and 898 are characterized by disturbed fabric. The evidence of disturbance from low F value intervals at Site 900 is not supported by Kmin inclination observations. The Kmin fabric inclination is a statistical measurement, so it is possible for fabric subject to weak disturbance to show normal Kmin inclination, but low foliation.
The variations of F value show clear trends and offsets. Some of the changes correspond to lithologic unit boundaries.
The fabric direction observed throughout Units I and II is of various origins, which include current, gravity, and disturbed orientation. As mentioned in the previous section, the lineation factor is not well developed, but may provide some information about paleocurrent. It is important to select the most suitable measure of fabric for paleocurrent analysis. The fabric signature resulting from gravity during deposition does not produce a preferred Kmax direction, but this random component will blur the current analysis. The q parameter was chosen because it eliminates the gravity origin fabric. The q value of gravity induced orientation is typically less than 0.2 (Taira, 1989).
The inclination of Kmin is used to reject samples affected by postdepositional disturbance. To remove the disturbed fabric, I chose the samples with Kmin inclination greater than 60°. After selection, only 45 samples were obtained for paleocurrent analysis. AMS directions oriented by remanent magnetization of Unit I from Holes 897C, 898A, and 900A reveal clustering of Kmax. The magnetic lineation of each hole indicates a general eastward dip, suggesting imbrication of grains (Fig. 3). The AMS directions show a westward trend, but are diverse, probably reflecting the variability of local settings. Unit I in Hole 900A shows a northeast downslope direction, probably caused by turbidity currents from the continental slope (Fig. 4). The submarine canyons around Site 898 are elongated in a west-east direction. The elastics of lower to upper Pleistocene turbidites were probably supplied through this canyon system. Site 898 sediments also show a direction concordant with the direction of the canyons. However, samples from Hole 897C seems to show a different lineation. A possible interpretation is that turbidity currents came from the rise of Vasco da Gama Seamount, situated northwest of Hole 897C.
The current directions for Unit II sediments from Holes 897C, 897D, 899A, and 899B seem to be scattered, and it is difficult to infer paleocurrent directions from them. The lithology of Unit II is characterized by turbidite and contourite facies (Sawyer, Whitmarsh, Klaus, et al., 1994). This scatter may be caused by mixed flow directions, reflecting both contourite and turbidite depositional processes.