Clark R. Alexander, Tim Elfers & Amy Simoneau

Abstract
Sedimentary processes on accretionary continental slopes are poorly characterized because these settings are relatively uncommon during the present high-stand of sea level. A suite of 139 cores has been collected to examine the processes of sediment accumulation and sediment delivery to the Eel River continental slope in waters <800 m deep. Spatial surveys of ²¹⁰Pb accumulation rate (0.2-7.5 g/cm ²y) and surficial sediment grain size (4-8.5 phi) demonstrate that sediment is accumulating throughout the Eel River margin, reflecting a rapid and widespread redistribution of the fluvial discharge. Downcore grain-size profiles of heterogeneous sediments suggest that hemipelagic and other episodic sediment-delivery processes may be important for the slope in areas proximal to the river. In areas more distal to the river, downcore profiles reveal relatively homogeneous sediments, delivered dominantly by hemipelagic processes. A budget for fine-grained sediment shows that this upper portion of the slope contains, at a maximum, 15% of the river's annual discharge; a combined budget for the shelf and upper slope suggests that at least 65% of the annual sediment load is not accounted for in these areas. The Eel River continental slope, because of its large sediment input, the narrow shelf adjacent, and its energetic environment is a good analogue for continental margins during transitional stands of sea level and for modern, tectonically active settings because processes typically associated with these geological settings ( i.e. , supply of sediment to the slope and downslope mass movement) are actively occurring.

Introduction
Sedimentary processes on continental slopes have received comparatively little attention and thus are poorly known relative to processes on shelves and in the deep sea. In tectonically active environments, sedimentary processes are complex and typically display temporal as well as regional-scale spatial variability. On the Eel margin, slope sedimentary processes are potentially influenced by a variety of factors that can dramatically influence sediment accumulation and redistribution patterns: earthquakes, rapid deposition from major floods, shoaling internal waves, storms and high-velocity currents, and subsurface gas. During transitional periods and lowstands of sea-level, most slopes directly received sediment from fluvial discharge. Since the Holocene Transgression, sediment input has been effectively trapped nearshore in estuaries, by coastal fronts, or on shelves. The Eel River continental slope is useful as an analogue for slope processes active during transitional periods of sea level because Holocene sediment is escaping the narrow shelf and accumulating on the slope. Observations from this study serves as a tool for better interpreting the stratigraphy of deposits formed during transitional periods of sea-level, and for understanding the interactions between sediment supply and depositional sequences.

Methods
Field
Cores were taken at 139 stations from 1995-1998 using single-spade box corers (60 cm long), kasten corers (3-m long), and piston and gravity corers (3-10 m long) deployed from research vessels, and using hydraulic push cores collected with the Monterey Bay Aquarium Research Institute ROV Ventana. Subcores for radiochemical and sedimentological analyses were taken from box cores and Kasten cores at 1-2 cm intervals. Transect lines are alphabetically lettered from south to north, starting with the "A" transect line just north of the Eel Canyon (Fig. 1). Stations are labeled with transect line and numbered with the nominal water depth.

Laboratory
Piston and gravity cores were split and photographed using a digital camera. Samples of approximately ~3 cm vertical thickness were removed from piston and gravity cores to provide foraminiferal material for ¹⁴C analyses. Sediment samples for ²¹⁰Pb and ¹³⁷Cs analyses were ground to a powder, sealed and equilibrated for 20 days. Activities of radionuclides other than ¹⁴C were determined using gamma spectroscopy. Detector efficiency as a function of material density for each radionuclide was determined from a series of standards covering the range of densities observed in the study area. Total ²¹⁰Pb activity was directly determined by measuring the 46.5-KeV gamma peak. Supported levels of ²¹⁰Pb were determined by measuring the gamma activity of ²¹⁴Pb (295 and 352 KeV) and ²¹⁴Bi (609 KeV). Self-absorption corrections for ²¹⁰Pb were made on each sample following the technique of Cutshall et al. (1983). ¹³⁷Cs activities were determined by measurement of its 662-KeV gamma peak. ¹³⁷Cs is an impulse tracer (produced from atmospheric nuclear tests), which was first introduced into the environment in significant amounts in about 1954 and had peak input in 1963. Given a calculated ²¹⁰Pb accumulation rate, one can predict how deeply ¹³⁷Cs should be observed to penetrate since its introduction ((accumulation rate * elapsed time between 1954 and core collection) + thickness of the surface mixed layer). ¹⁴C ages of selected intervals were assessed using Accelerator Mass Spectrometry (AMS) techniques. The target preparation and ¹⁴C analyses were performed at the Lawrence Livermore AMS Facility. Grain size was determined at 1/4 phi intervals using sieves (for the sand fraction) and using a Sedigraph 5100 (for the silt and clay fraction). The Sedigraph, by quantifying x-ray beam attenuation in a settling cell, measures Stoke's settling velocities and provides data comparable to the pipette method. Sediment statistics were calculated using the method of moments.

Figure 1: Location of the study site indicating all stations sampled in this project. Note the a breached anticline at about 40°55' N, which effectively divides the slope into two distinct depositional environments. South of the anticline, the slope morphology is dominated by a large, arcuate, failure feature (the Humboldt Slide) at the base of which large, slope-parallel ridge and trough morphology is observed. North of the anticline, the slope is relatively smooth, incised only by shallow slope gullies (Field et al. , 1999). These gullies become common in water depths below 250-300 m, and extend out onto the mid-slope plateau and into the head of Trinity Canyon. The major physiographic provinces of the Eel margin include: the shelfbreak at 150 m water depth, an upper slope at 150-450 m water depth; a broad, gently dipping mid-slope plateau in water depths of 450-800 m; and a steep, lower slope that is present in water depths from 800 m to 2700 m.

Figure 2: Distribution of ²¹⁰Pb-derived mass accumulation rates. Accumulation rates on the Eel River slope range from <0.02 g/cm²y in areas of overconsolidated muds, exposed in the headwall of the Humboldt Slide, to approximately 7.5 g/cm² y. Representative profiles of ²¹⁰Pb and ¹³⁷Cs are shown in Figs. 3A&B. Rates are generally highest (0.2-0.6 g/cm²y) in shallow (150 m) water depths and shallower portions of the mid-slope plateau (450-600 m). Note that rates decrease seaward across the mid-slope plateau, indicating a decrease in sediment delivery to the lower slope. Rates exhibit a minimum in intermediate water depths (on the upper slope; see fig. 4). High- resolution geophysical profiles (Field et al. , 1999) show erosional truncation of seabed reflectors in this depth range, indicating that lower rates have existed in this region in the past. Sediment is bypassing this region, either by non-deposition on this relatively steep portion of the slope profile, or through some resuspension process. Conduits for bypassing may exist in the slope gullies, which have their heads in this depth range, although preliminary radiochemical data suggest that these channels are not active on 100-y timescales. Resuspension would most likely be caused by internal waves, which impinge on the Eel slope and expend their energy in this region (D. Cacchione, pers. comm.). Anomalously high accumulation rates (>0.6 g/cm²y) are present within the ridge and trough morphology at the bottom of the Humboldt Slide failure feature. Because the ²¹⁰Pb profile from O450, the central station in this anomalous region did not exhibit a steady-state decay profile, it was not possible to calculate a ²¹⁰Pb accumulation rate (Fig. 5A). However, a rate was estimated based on the steadily increasing activity of ¹³⁷Cs from the top to the bottom of the 55-cm-long box core (Fig. 5B).

Figure 3: Representative radiochemical profiles from the Eel slope. A) Profile of excess ²¹⁰Pb showing a 3-cm-thick surface mixed layer and exponential decay of excess activity below. B) ¹³⁷Cs profile for the same core. Good agreement between ¹³⁷Cs penetration depths observed and those predicted based on accumulation rates indicate that ²¹⁰Pb accumulation rates are good estimates of actual accumulation rates in most cases in the study area. ¹⁴C ages from station Y450 (2680 y B.P. at 192 cm, 3510 y B.P. at 327 cm, and 3910 y B.P. at 457 cm) suggest that rates have been relatively constant over at least the past ~4000 y.

Figure 4: The depth distribution of ²¹⁰Pb mass accumulation rates both documents a minimum in rates on the upper slope and demonstrates that rates on the upper slope significantly differ by approximately a factor of two between the south and north sides of the anticline (significant at 99% level, t=3.0, 26 df). The upper-slope minimum in rates is less pronounced north of the anticline. Accumulation rates in >450 m water depth are not significantly different north and south of the anticline (at 99% level, t=1.4, 18 df). The open circles represent stations from north of Trinidad Head (Fig. 1), where slope sedimentation patterns appear to differ from those observed farther south. Note the anomalously high rates in the Humboldt Slide and Eel Canyon.

Figure 5: A) Excess ²¹⁰Pb profile from the region of anomalously high accumulation rates in the Humboldt Slide. Variability in the profile is not a result of grain size differences and may represent discrete downslope mass flow events. B) The ¹³⁷Cs profile exhibits increasing activity toward the bottom of the core, suggesting that the base of the 55-cm long core was deposited after the 1963 peak in atmospheric weapons testing. Assuming that the base of the core represents the 1963 peak in ¹³⁷Cs input, the accumulation rate would be 1.7 g/cm²y. High rates are also measured in the upper reaches of the Eel Canyon, suggesting that the upper canyon is actively trapping sediment derived from the Eel River.

Figure 6: Distribution of surficial sediment grain size. Sediments on the shelfbreak and upper slope are significantly coarser south of the anticline (4-8 phi) than north of it (6-8 phi; significant at 99% level, t=2.2, 36 df). On the inner mid-slope plateau, sediments do not differ significantly (at 99% level, t=2.0, 8 df). However, sediments coarsen in the seaward portions of the outer mid- slope plateau, suggesting a decrease in supply of fine-grained material.

Figure 7: A depth transect of all size data for the northern area demonstrates that sediments display a coarsening in 250-350 m water depths, a depth range similar to that where we observe a decrease in accumulation rates. Grain-size distributions exhibit coarser mean grain size, better sorting and a decrease in the amount of clay-sized material in the upper-slope region, suggesting that shoaling internal waves are acting to resuspend bottom sediments and remove finer-grained material or are inhibiting accumulation in this region of the slope.

Figure 8: A) Two kasten cores (2-m long) were collected to compare the stratigraphies of the northern portion (Y450) and the southern portion (O450) of the study area. In core Y450, sediments are texturally homogeneous below a finer-grained surface layer. The deep sediments have a narrow range of sizes between 6.9-7.4 phi and are characterized as moderately sorted clayey silts. In contrast, core O450 contains a diverse sedimentology with textures spanning a broad range of sizes (4.9-7.8 phi) and sorting, from well-sorted fine sands to poorly sorted silty clays. Texturally distinct beds in O450 range in thickness from 10-50 cm. Sand-silt-clay percentages downcore demonstrate that sand is not a significant component in the northern slope (C), whereas sand is a variable and important component in the southern slope sediments (B). Inset shows detailed analysis of one distinct bed in O450.

The variable texture in the southern part of the study area in comparison to the relatively homogeneous sediments documented in the north suggests that several mechanisms deliver sediment to the southern slope, whereas sediment delivery to the northern slope is more uniform and may be dominated by a single process. Data presented in Walsh and Nittrouer (1999) for station Y450 show that the input of sediment through the water column approximately equals the flux of sediment to the seabed on ²¹⁰Pb timescales. This hemipelagic input is also present in the south, but is either present in smaller quantities because of the character of local circulation patterns in relation to the timing of major sedimentation events, or is accumulating in areas not yet investigated adequately ( i.e. , Eel and Trinity Canyons). Because hemipelagic processes are not sufficient to transport the coarse sediments observed on the slope in some areas, another process is necessary to provide the coarser-grained material to the southern part of the study area. Potential mechanisms include hyperpycnal flows produced during flood events and episodic downslope mass flows produced by remobilization of sediment that has accumulated seaward of the river at the shelf break. The narrow shelf, high discharge of the river during floods, rapid accumulation rates along the shelf break and several initiation mechanisms for downslope flow ( e.g. , earthquakes, large wave events) suggest that these failures could occur in the study area.

The distribution of accumulation rates allows us to calculate a sediment budget for fine-grained ( i.e. , silt and clay) sediment delivery to the Eel River continental slope (Table 1). The contribution of sand to the budget has been removed using grain size data for each core. This budget, which includes the shelfbreak, upper slope and mid-slope plateau, uses a sediment input value of 14x10⁶ t/y to the Eel margin, representing the combined input from rivers discharging into the study area - the Eel, Mad and Van Duzen Rivers (Sommerfield and Nittrouer, 1999). The five slope regions in the budget, representing water depths from 150-800 m in the study area, document that about 15% of the annual discharge of the Eel River is accumulating in the shelfbreak, upper slope and mid-slope plateau. The budget for the shelf (Sommerfield and Nittrouer, 1999) documents 20% retention of the annual discharge and therefore the greatest quantity of continental shelf and slope sediment that can be accounted for is approximately 35%. Thus, a repository for at least 65% of the Eel's discharge remains to be determined. Analyses are now ongoing to evaluate the importance of the Eel and Trinity Canyons, and the deep-sea fans associated with these canyons, as repositories for sediment. Preliminary ²¹⁰Pb data from the head of the Eel Canyon indicates that sediment may be accumulating rapidly in this area on 100-y timescales. The seaward decrease in rates on the outer mid-slope plateau suggests that the lower slope (>800 m) is not a significant repository for Eel-derived sediment.