Orbital, suborbital, and
higher frequency (millennial) variation is clearly exhibited by fluctuations in
mean grain size, sorting, and proportions of the grain-size fractions. An
~10,000-yr cyclicity is most strongly shown throughout the succession by changes
in grain-size sorting. In general, grain size decreases and sorting increases
(standard deviation decreases) during warmer intervals (low 
18O
of planktonic foraminifers) with higher sea level. Clearly, sedimentation at
Site 1017 was strongly influenced by climatic and oceanographic oscillation, yet
the exact mechanisms are not easily resolved. Below, we discuss some of the
oceanographic and bathymetric constraints and lines of evidence important to
consider in the interpretation of the late Quaternary record.
The slope setting (955 m) of Site 1017 is bounded upslope to the north by the Santa Lucia Bank and to the northeast by the continental shelf, and downslope to the south by the Arguello Submarine Canyon and to the east and southeast by a series of upper slope feeder canyons (Fig. 1). Siliciclastic sediments, which dominate this setting (Lyle, Koizumi, Richter, et al., 1997; Tada et al., Chap. 25, this volume), are derived from riverine sources in the Santa Ynez, Santa Maria, and Big Sur regions of the California coast (Irino and Pedersen, Chap. 23, this volume). Neritic sediment transport in the coastal zone and outer shelf is to the south under the influence of littoral drift and the equatorward-flowing California Current, respectively. Frequent, volumetrically minor, graded sand layers at Site 1017 indicate that some turbidity currents flowed directly down the slope or spilled out of the Arguello Submarine Canyon and its feeder system to flow over the southern Santa Lucia Slope. These coarse-grained turbidites are concentrated within two intervals between 320 and 510 cmbsf and 1390 and 1470 cmbsf, which were deposited during the MISs 2 and 4 lowstands. The ultimate source of the coarsest sediments is the nearshore region to the north and northeast (Big Sur and Santa Maria areas; Trask, 1952) where sands of similar composition can be found (including beach lag deposits nearly identical to a remarkable heavy mineral-, zircon-, and spinel-enriched sand layer present at 507 cmbsf in Hole 1017E). The elemental composition of even the finest silt and clay fraction of the sediment deposited during MISs 2 and 4 lowstands has a distinctly mafic provenance, most likely from the Mesozoic Franciscan Complex that outcrops along the central California coast north of Site 1017 (Irino and Pedersen, Chap. 23, this volume). Longshore transport of sand from the north was increasingly diverted into the feeder canyons adjacent to Site 1017 (Fig. 1) during intervals of lower sea level and prevented from continuing on into the Santa Barbara Basin to the east (Behl, 1995) as it does during the present highstand.
In addition to the southward (equatorward) flow and transport due to the California Current (0-500 m depth) and shallow littoral currents, the margin is also influenced by the poleward-flowing California Undercurrent (Hickey, 1979). In modern studies, this current is strongest at 150-300 m water depths with peak speeds ~30-50 cm/s. During winters and other periods of low equatorward wind stress, the countercurrent's flow can extend all the way to the seafloor (~1000 m), albeit at lower velocities (Ramp et al., 1997). A deeper (~300-500 m) and weaker, seasonal equatorward undercurrent also flows over the continental slope as part of the California Current system (Hickey, 1998). At present, the highest velocity flows in both undercurrents are considerably shallower than the seafloor at Site 1017 (955 m), but it is possible that these oceanic currents may have influenced sediment transport and sorting at Site 1017 in the past. This could have occurred either directly, by expansion or deepening of the undercurrents to the seafloor at 1000 m, or indirectly, by resuspending and transporting sediment initially deposited upslope from this site. Both undercurrents can be disrupted and even reversed by large cyclonic and anticyclonic eddies offshore of northern and central California that may reach the seafloor in water as deep as 2000 m (S.R. Ramp, pers. comm., 1998).
Textural variation in the stratigraphic succession at Site 1017 was possibly influenced by either (1) the effects of sea-level fluctuation on offshore transport in the manner postulated and documented by sequence stratigraphers (e.g., Wilgus et al., 1988; Van Wagoner et al., 1990), (2) by fluctuation in the strength of bottom currents passing over the continental slope (Hollister and Heezen, 1972), or (3) by both. If grain size on the Santa Lucia Slope was primarily controlled by current action (i.e., coarser intervals are contourites), then both mean grain size and sediment sorting should be positively correlated (McCave, 1985). As current strength increases, so should grain size and sorting as the finer fraction of the noncohesive "sortable silt" is resuspended and removed, leaving behind a narrower spectrum of coarser silt and/or fine sand (McCave et al., 1995). This is not the case for the upper Quaternary succession at Site 1017, where grain size is negatively correlated with sorting, in which larger mean diameter is almost always associated with a greater standard deviation (poorer sorting). This relationship is explained by inspection of the relative contributions and modes of the two or three distinct components resolved from deconvolution of the grain-size distributions (data set TU). These grain-size spectra show that samples with coarser mean grain sizes are more poorly sorted, chiefly because of the presence of the coarse silt component (Fig. 5).
If sea level was the primary control of textural variation at Site 1017, then increased grain size may reflect enhanced resuspension and transportation off a narrowed and shallower shelf during times of lowered sea level (i.e., MIS 3 stadials and MISs 2, 4, and 6). Increased quantities of sediment could have been transported more or less directly downslope by nepheloid plumes and turbidity currents with some deflection by contour-following undercurrents. An alternate possibility is that off-shelf sediment transport was primarily routed through the feeder canyons to the east and southeast into the Arguello Submarine Canyon (Fig. 1). Hyperpycnal flow (possibly fed by suspended sediments from the Santa Ynez River) or low-density turbidites could easily ride up the right bank of the canyons (Mulder and Syvitski, 1995; Normark et al., 1998) and blanket the southern Santa Lucia Slope with fine-grained deposits. In both sea-level related scenarios, grain size and sorting on the slope would be modified primarily by the addition of the coarser silt fraction from shelfal regions to the finer hemipelagic sediments of the upper slope. Either mechanism could produce the thin intervals of faint silty laminations or sand layers that are intermittently present in this stratigraphic sequence (Tada et al., Chap. 25, this volume). The importance of sea level is also indicated by the observation that the finest mean grain size and best sorting (lowest standard deviation) corresponds to the highest sea level of this interval ~130-120 ka (Neumann and Hearty, 1996). Elemental geochemistry by Irino and Pedersen (Chap. 23, this volume) indicates increased contribution of mafic and ultramafic components to the sediment during the last glacial maximum compared with MIS 1 (Holocene). We, too, find trace amounts of augite and uvarovite (a Cr-rich garnet)—two minerals associated with mafic and metamorphic (serpentinite) components of the Franciscan Formation that outcrops in the central Californian Big Sur region (Trask, 1952; Fig. 1)—consistently present in the coarse and fine silt fractions of samples taken from coarser intervals, but only rarely present in the coarse silt fractions of samples from finer intervals and never in their fine silt fraction. These data support the hypothesis that the coarser sediments represent increased downslope transport from the central Californian coast.
It is possible that individual grain-size components, especially the fine and coarse silt fractions, are not derived from different sources (i.e., shelf turbidites vs. hemipelagic sediment) but may represent episodic differences in the strength of bottom currents. Because each sample spans ~150-yr accumulation, the grain-size spectra may integrate sediments deposited under entirely different flow conditions. In this scenario, the coarse silt component would represent deposition during higher than average current flow rates. Although variability in the location, persistence, and size of mesoscale eddies along the California margin may have produced significant variations in deep geostrophic flow (e.g., the "benthic storms" of Hollister and McCave, 1984), we do not know why contour current flow would be so distinctly bimodal as to produce separate grain-size classes instead of varying continuously over a range of velocities.
