Downcore variations in the mass input of terrigenous material provide an important signal of changes in source area climate. In modern systems, large variations occur in the amount of sediment delivered to the seafloor of the northeast Pacific that are directly linked to large flood events along coastal areas and more efficient transport of sediment at times of peak river discharge (Griggs and Hein, 1980; Griggs, 1987). On longer time scales, however, sediment yield is more closely related to sediment supply in the source area, which can vary with a combination of factors: basin relief and area, changes in sea level, and climatic factors such as precipitation or vegetative cover (Milliman and Syvitski, 1992; Rea, 1992). Within the time scale of the late Pleistocene, tectonic factors (basin relief and area) do not change significantly. The deposition of terrigenous material on the seafloor mainly reflects some combination of sediment supply from shallow shelf regions exposed during glacioeustatic sea-level changes and variations in precipitation/erosion in the source area drainage basin associated with shifts in climate.
Both Sites 1018 and 1020 show a significant increase in terrigenous accumulation rate during times of glaciation. Glacial terrigenous flux values are more than twice the values observed during interglacial episodes. Spectral and cross-spectral analysis performed on the 0-300 ka portion of these records confirms this pattern and shows that increased terrigenous flux values are coherent and in-phase with oxygen isotopic records at each of the major Milankovitch periodicities. Implicit in this pattern is that there was an enhanced supply of terrigenous mineral grains to the northern California margin coincident with periods of glacial advance in the Northern Hemisphere. To distinguish whether this increase is related to sea-level changes or to climatic factors requires some knowledge of the provenance, transport, and depositional processes associated with these deposits.
Terrigenous mineralogy provides a reasonably good tracer of provenance in this area of the Pacific (Karlin, 1980; Krissek, 1982). Modern transport of clay minerals suggests that terrigenous material deposited at Site 1018 monitors the supply of sediment transported from sources draining Northern and Central California, whereas sediment at Site 1020 receives a mixture of material supplied from northern sources draining the Columbia River Basin and southern sources draining the Coastal Range of California and southern Oregon (Karlin, 1980). The downcore record of clay mineral abundance at Site 1020, however, shows increased illite and chlorite content during intervals of enhanced terrigenous input, suggesting that much of this material may be transported from southern source regions. Depositional records at Site 1020, therefore, probably represent a distal record of sediment supplied to Site 1018. This is not surprising given the similar temporal variability observed in terrigenous input at each location. However, to interpret whether terrigenous mass flux data record changing sea level or climatic factors requires more information about transport pathways and depositional processes.
Recent studies involving terrigenous mineral size distributions suggest that they may permit reasonably good delineation of the dominant mode of sediment transport to the seafloor (Ellwood and Ledbetter, 1977, 1979; Rea and Hovan, 1995; Joseph et al., 1998). These studies show that distinct differences exist between the size distributions associated with sediment found in hemipelagic, drift, and turbidite depositional regimes. Although a few thin turbidite layers were identified throughout the upper sections of sediments at these sites (Lyle, Koizumi, Richter, et al., 1997), grain-size distribution data suggest we were successful in avoiding them during sampling for this study. Grain-size distributions for Site 1018 are characterized by a coarser mean size broadly distributed over 1-30 µm and are more similar to sediments found in higher energy drift deposits. Size distributions for Site 1020 are finer grained with better sorting, reflecting sedimentation under lower energy hemipelagic conditions. When plotted together, the grain-size data fall along a mixing line between end-members represented by samples from each site, which indicates that sediments from Site 1020 may also represent a distal counterpart to Site 1018 with respect to transport mode and deposition process (Fig. 7).
This has important implications in our interpretation of the terrigenous flux records. If lower sea level during glacial episodes caused increased terrigenous mineral flux at both sites during these times, then grain-size data should also show similar temporal patterns. But records of terrigenous grain size contain much higher frequency variability than terrigenous accumulation rate data and display no obvious Milankovitch periodicity (Fig. 2, Fig. 3). Thus, it seems likely that increased terrigenous input during glacial times reflects enhanced supply from the continental source region. These findings are consistent last glacial maximum climate reconstructions made from vegetational history and lake-level records in the southwestern United States (Van Devender et al., 1987; Mifflin and Wheat, 1979) but contrast with those from the Northwest (Barnosky et al., 1987).
Perhaps just as significant is the remarkable similarity observed between grain-size distributions throughout glacial cycles of the late Pleistocene (Fig. 5). Glacial size distributions are nearly identical to interglacial size distribution within sediments examined at each site and give no indication of changes in the energy of transport or type of depositional processes associated with sea level or other glacial-interglacial variability.