METHODS

Textural analyses were performed using settling-tube techniques for the coarse-grained fraction (>50 µm) and Sedigraph 5000 D techniques for the silt and clay fractions (<50 µm) (Giró and Maldonado, 1985). Grain-size distributions were plotted in a cumulative curve and frequency histogram for each sample. Textural statistical parameters (mean, standard deviation, skewness, and kurtosis) were calculated using moment measurements on sample populations containing quarter-phi interval classes in all fractions (Friedman and Sanders, 1978; Forrest and Clark, 1989). The following ranges of standard deviation were adopted for the classification of sediment into sorting classes (Friedman and Sanders, 1978; Alonso and Maldonado, 1990; Ercilla et al., 1994):

moderately well sorted sediments: 0.50 to 0.80,

moderately sorted sediments: 0.80 to 1.40,

poorly sorted sediments: 1.40 to 2.00,

very poorly sorted sediments: 2.00 to 2.60.

The fine-grained fraction (pelite fraction, <50 µm) was plotted in cumulative curves to identify sediment types according to hydrodynamic depositional processes (Gonthier et al., 1981). Three different cumulative curve shapes were identified by the slope gradient according to Riviere (1977): (1) hyperbolic (0°-15°), (2) logarithmic (15°-45°), and (3) parabolic (>45°). These curve shapes correspond the granulometric facies and sedimentary processes described by Riviere (1977) and Gonthier et al. (1981). The hyperbolic curves are observed in pelagic/hemipelagic deposits. Logarithmic curves tending toward hyperbolic represent turbiditic/contouritic silty clays, whereas those tending toward parabolic correspond to turbiditic/contouritic silts and sandy silts. The parabolic curves occur in turbiditic/contouritic coarse-grained deposits (sand and silt).

The nature of the pelite fraction (<50 µm) was examined with a Hitachi S-750 scanning electron microscope (SEM). These data were included within the sedimentological data set to define the sedimentary facies.

Sand fraction (>63 µm) composition was studied using a binocular microscope. Abundances of different components were estimated by counting about 300 grains per sample. The following components were identified and counted: light minerals (quartz, mica, feldspar, and others), heavy minerals, rock fragments, neoformation minerals (pyrite and pyritized tests and burrows, glauconite), bioclasts (pelagic foraminifers, benthic foraminifers, siliceous biogenic materials, pteropods), and others that include unidentifiable remains. For sand fraction components, the ranges used to describe the nature of the sediments are as follows:

biogenous sediments: >60% biogenous components,

mixed sediments: from 40% to 60% biogenous and terrigenous

components, terrigenous sediments: >60% terrigenous components.

For the quantitative mineral analysis of marine sediments, which are multiphase mixtures, pure standards are not available to obtain calibration constants. Thus, standardless methods of quantitative X-ray analysis must be used. Many authors have developed complicated methods, from a practical point of view. To obviate the major problems of these methods, Rius et al. (1989) developed a simple standardless method that directly determines calibration constants using only diffracted intensities and calculated phase-absorption coefficients through a least-squares procedure. The relative abundances of illite, kaolinite, chlorite, quartz, calcite, dolomite, feldspar, and plagioclase were estimated following a method according to Rius et al. (1989), which has been used to analyze marine sediments (Palanques et al., 1990).

For the analysis of mineral phases, sediment samples were ground for 20 min in an agate mill and packed into sample holders. X-ray diffractograms were taken on a Siemens d-500 X-ray diffractometer. Cu K radiation, generated at 40 kV and 20 mA, a graphite monochromator, and a scintillation counter were used. Peaks were scanned from a 2 of 4° to 50°, with a step-scan range of 1.2–2, an angular increment of 0.05/2, and a counting time of 3 s/step. Diffractograms were visually interpreted with the help of a computerized search-and-match routine using Joint Committee on Powder Diffraction Standards files.

The carbonate content of the samples was determined using a Bernard calcimeter, according to the method described by Vatan (1967). This method measures the pressure of the CO₂ released when the sample is attacked with dilute HCl at atmospheric pressure. The bulk sample was previously pulverized to achieve total homogenization. Carbonate contents are expressed as weight percent CaCO₃, assuming that all carbonate was present as pure calcite (Table 1).