As determined by XRD, sediment on the Bermuda Rise is comprised of three basic components (Table T2): calcite (1%-43%), aluminosilicates (52%-92%), and amorphous silica (0%-23%). Aluminosilicate minerals are dominated by quartz, illite-type clays, and Na- and K-feldspars. Minor phases (<10%) include Ca-feldspar, dolomite, chlorite, kaolinite, muscovite, and halite. The small amount of halite probably precipitated from pore water during drying. Shipboard smear slide analyses show that the amorphous silica is of biogenic origin. Interestingly, some of this silica may be transported from shallow depths to the Bermuda Rise rather than produced in the overlying water column (D. Winter, pers. comm., 1999).
Physical properties changes over the studied sediment interval correspond to changes in mineralogy (Fig. F5). Intervals of high bulk density are marked by high concentrations of calcite (Fig. F5A). Conversely, intervals of low bulk density correspond to high concentrations of aluminosilicate minerals (Fig. F5C). Intervals of high P-wave velocity and low bulk density relate to high concentration of biogenic silica (Figs. F5C, F5D, F5E).
Carbonate and biogenic silica typically serve as a dilutant in marine sediment. Thus, the various aluminosilicate phases were examined by recalculating their abundance on a calcite- and biogenic silica-free basis. In general, the relative proportions of quartz, feldspars, and clay are fairly constant downcore. The exceptions are muscovite, which is generally more abundant over intervals of low bulk density, and illite-type clays, which are generally more abundant during periods of high bulk density (Fig. F5G).
Bulk sediment from the Bermuda Rise shows a wide range in major element oxide compositions (Table T3), although this range is less than that determined for core-top samples across the entire northwest Atlantic (Ericson et al., 1961) (Fig. F6). Much of the chemical variation is caused by differences in the abundance of calcite (CaCO3). As evident by the strong correlation between CaO and calcite (Fig. F7), calcium is hosted almost entirely in calcite. When this carbonate dilution effect is removed, SiO2, TiO2, Al2O3, and K2O concentrations vary within the relatively narrow ranges of 42%-56%, 0.57%-0.73%, 12.0%-15.4%, and 1.1%-3.4%, respectively. However, in spite of carbonate dilution, several downcore trends in major elements are evident. Of particular interest is that the Al2O3/TiO2 and K2O/Al2O3 ratios show excellent correlations with downcore variations in bulk density (Figs. F5E, F5F). The SiO2/Al2O3 ratio also shows correlations with P-wave velocity and the abundance of biogenic silica determined by XRD (Fig. F5C).
Major element oxide concentrations can be used to assess mineralogical compositions determined by XRD. The expected total SiO2, Al2O3, and CaO concentrations for each sample were calculated by multiplying the abundance of each mineral phase by the weight percent of Si, Al, and Ca contained in each phase (Table T4). For example, albite (NaAlSi3O8) constitutes 20% of Sample 172-1063D-4H-3, 108-109 cm, at 25.38 mbsf. Thus, albite contributes 13.5%, 3.8%, and 0% of the total SiO2, Al2O3, and CaO concentrations, respectively, in the bulk sample. The sum of calculated SiO2, Al2O3, and CaO concentrations for all minerals renders expected total oxide concentrations. The mean difference for SiO2 determined by XRD and XRF is 6.6% with a standard deviation of 3.1%; similarly, the difference for Al2O3 is 0.5% with a standard deviation of 1.0%, and for CaO the mean difference is 2.2% with a standard deviation of 1.2% (Table T5). This comparison suggests that determination of the major phases is relatively accurate; however, the consistent overestimation of calculated relative to measured SiO2 possibly reflects problems in determining the abundance of biogenic silica by XRD.
There is an increase in the percentage of silt and fine sand-sized particles (10-100 µm) across zones of high P-wave velocity (Fig. F8). The volume of sediment in this moderately coarse size range compares closely with the abundance of biogenic silica (Fig. F9). Indeed, examination with a scanning electron microscope shows that bulk sediment in Sample 172-1063D-4H-4, 72-74 cm, at 26.52 mbsf, is composed of abundant and fragile sand-sized biogenic silica grains (Fig. F10). The changing particle size during sonication likely reflects destruction of biogenic silica grains. The very coarse mode at 200-400 µm evident in some samples is possibly a combination of sand-sized foraminifers and authigenic particles.