RESULTS

Terrigenous mineral input dominates the sediments found along the northern California margin, comprising an average of 70-80 wt% of the bulk sediment composition (Fig. 2). As such, there are only slight downcore variations in the dry bulk density, and linear sediment rates generally reflect changes in the amount of terrigenous mineral contribution. The overall terrigenous accumulation rate is significantly greater at Site 1018 compared to Site 1020, a difference that becomes more pronounced during glacial periods. Terrigenous MAR at Site 1018 averages 15.06 g(cm² · k.y.)^-1 (N = 174; = 9.10) but varies from 5 to 20 g(cm² · k.y.)^-1 throughout the late Pleistocene. At Site 1020, which is farther offshore and deeper, terrigenous mass accumulation and variability are significantly reduced, averaging about 8.81 g(cm² · k.y.)^-1 (N = 336; = 3.29). Both sites display a general trend toward increased terrigenous flux during glacial periods, although the details of this comparison vary slightly at each location. Spectral analysis confirms a strong relationship between terrigenous flux and glacial-interglacial periodicity with clear periodicity concentrated at the orbital frequencies of eccentricity, tilt, and precession (Fig. 3A, B). Cross-spectral analysis shows that the terrigenous flux records are coherent and in-phase (i.e., greater terrigenous flux during glacial ¹⁸O values) with SPECMAP at 100, 41, and 19 k.y. (Fig. 4).

Grain-size data also display important differences between Sites 1018 and 1020. Average grain-size distributions for terrigenous sediment from Site 1018 are coarser and more poorly sorted than those deposited at Site 1020 (Fig. 5). Terrigenous minerals at Site 1018 are characterized by a moderately sorted, broad size distribution centered with a mode near 4 µm that peaks at ~1 wt% (Fig. 5A). Downcore there is little variation, with mean grain-size values averaging 3.87 µm (N = 164; = 0.53) and sorting values (shown as standard deviation of the grain-size distribution) of 7.26 µm ( = 0.82) during the past 300 k.y. At Site 1020, size distributions show a size mode near 3 µm that peaks at 2 wt%, are better sorted, contain a significant amount of smaller material, and are concave up in the coarse tail (Fig. 5B). They are remarkably uniform throughout the late Pleistocene, with average mean grain-size values of 2.89 µm (N = 244; = 0.17) and sorting values of 5.18 µm ( = 0.49). Grain-size data vary at a higher frequency than terrigenous mass accumulation rate but show no obvious relationship to glacial cycles or orbital periodicity (Fig. 3C, D).

A subset of samples was selected for clay mineral analysis. Ideally, we wished to compare mineralogical variation at both sites; however, constraints imposed by time and instrumentation permitted only analysis of Site 1020 at the time of publication. The normalized clay mineral content varies considerably downcore (Fig. 6). In general, illite is most abundant, comprising as much as 50% of the clay fraction, but its percentage varies throughout the interval. Chlorite and smectite percentages are less variable but also less abundant, averaging ~15% and 7%, respectively. Kaolinite contributes little to the clay mineral assemblage in these sediments. The low-resolution nature of clay mineralogy data makes it difficult to identify temporal patterns; however, the data show a general increase in the input of chlorite minerals during intervals of increased terrigenous flux. Likewise, smectite shows a general inverse pattern and is reduced during these intervals. This is more clearly evident in the upper and lower portion of the record and less so during isotopic Stages 6 through 8.