Natural and human impacts on centennial sediment accumulation patterns on the Umpqua River margin, Oregon

Natural and human impacts on centennial sediment accumulation patterns on the Umpqua River margin, Oregon

Marine Geology 339 (2013) 44–56 Contents lists available at SciVerse ScienceDirect Marine Geology journal homepage: www.elsevier.com/locate/margeo ...

2MB Sizes 6 Downloads 78 Views

Marine Geology 339 (2013) 44–56

Contents lists available at SciVerse ScienceDirect

Marine Geology journal homepage: www.elsevier.com/locate/margeo

Natural and human impacts on centennial sediment accumulation patterns on the Umpqua River margin, Oregon Robert A. Wheatcroft a,⁎, Miguel A. Goñi a, Kristin N. Richardson a, Jeffry C. Borgeld b a b

College of Earth, Ocean, and Atmospheric Sciences, Oregon State University, Corvallis, OR, USA Department of Oceanography, Humboldt State University, Arcata, CA, USA

a r t i c l e

i n f o

Article history: Received 29 November 2012 Received in revised form 16 April 2013 Accepted 17 April 2013 Available online 3 May 2013 Communicated by J.T. Wells Keywords: Oregon sedimentation radioisotopes continental shelf fluvial dispersal systems

a b s t r a c t Quantifying patterns of sediment accumulation rates (SARs) over the past ~ 125 years on continental margins provides insight into diverse processes spanning the land–ocean boundary. In particular, temporal changes in the export of fluvial sediment can lead to changes in ocean margin SARs, whereas spatial patterns of accumulation reflect the net effect of wave-driven resuspension and current transport averaged over many years. To explore these issues we quantified SARs using 210Pb geochronology at 73 stations on the shelf and upper slope off the Umpqua River, Oregon. Three types of 210Pb profiles were observed: a well-defined roughly 10-cm-thick surface mixing layer (SML) underlain by a zone of logarithmic decrease on the distal mid to outer shelf and slope (type 1, n = 45); a deep (20–30-cm thick) SML underlain by a zone of logarithmic decrease in the sandy inner shelf (type 2, n = 8); and a composite profile with two distinct zones of logarithmic decrease below an ~ 10-cm-thick SML in a mid-shelf depocenter adjacent to the river mouth (type 3, n = 21). Type 3 profiles imply a 2–4-fold increase in the SAR that occurred, on average, in 1967 ± 13 y. Such an increase in SARs is consistent with the history of industrial logging in the Umpqua basin, which peaked in the two decades after World War II and coincided with a wet phase of the Pacific Decadal Oscillation (1944–1978) when average and peak river flows were elevated. Comparison of the Umpqua River shelf depocenter with the well-studied Columbia and Eel River systems reveals some important similarities and differences between these ‘marine-dispersal dominated’ systems, whereby forcing (e.g., river–ocean coherence) on the Umpqua is comparable to that of the Eel, whereas the spatial pattern is more similar to that of the Columbia. These seemingly paradoxical results can be reconciled by considering the relative significant wave height during periods of elevated sediment delivery. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Spatial and temporal patterns of sediment accumulation on continental margins provide important information on processes occurring across the land–ocean boundary and are key to understanding whole-margin geology and biogeochemistry. Variations in sediment supply from uplands due to both natural (e.g., seismic, volcanic; Major et al., 2000; Hovius et al., 2011) and anthropogenic (e.g., land use, fire; Pasternack et al., 2001; Warrick et al., 2012) perturbations can be recorded in offshore sedimentary sinks (e.g., Gomez et al., 2007; Sommerfield and Wheatcroft, 2007). Because fluvial sediment supply is a fundamentally integrative process, deciphering supply fluctuations yields valuable information on the connectivity of sediment routing systems (e.g., Allen, 2008; Jerolmack and Paola, 2010) and provides context for the delivery of diverse bioactive constituents (e.g., Wheatcroft et al., 2010; Hatten et al., 2012; Goñi et al., 2013). Once delivered to the coastal ocean, sediment undergoes a second ⁎ Corresponding author. Tel.: +1 541 737 3891. E-mail address: [email protected] (R.A. Wheatcroft). 0025-3227/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.margeo.2013.04.015

set of processes that determines its initial pattern of deposition and eventual accumulation. Thus, patterns in wave energy, along- and across-shelf currents, and shelf physiography can have important effects on sediment dispersal and accumulation (e.g., Wright and Nittrouer, 1995; Geyer et al., 2004; Walsh and Nittrouer, 2009). Due to the complexity of these processes a general, predictive theory of margin sedimentation has yet to be formulated. All of the forcing processes mentioned above vary on a range of timescales and, because sediment accumulation rate also varies as a function of observation time (Sadler, 1981; Schumer and Jerolmack, 2009), it is important to stipulate the time scale of interest. Herein, our concern is sediment accumulation on decadal to centennial timescales over the past ~125 y. There are three reasons for this focus. First, this period of time is when relatively complete historical records of important processes (e.g., rainfall, river discharge, logging intensity) exist, thus there is the potential of establishing rigorous cause-and-effect relationships that can be used to decipher pre-historic sediment archives. Second, in the Pacific Northwest, the location of our research efforts, the period of time from the late 19th century to the present has been when human effects on the landscape have been most pronounced. The brevity of

R.A. Wheatcroft et al. / Marine Geology 339 (2013) 44–56

large-scale human impacts in our study area contrasts with the situation in many other parts of the globe where anthropogenic effects have existed for centuries to millennia (e.g., Pasternack et al., 2001; Oldfield et al., 2003; Gomez et al., 2007). Lastly, studying sediment accumulation over decadal to centennial timescales is facilitated by the existence of 210 Pb, a particle-reactive, naturally occurring radionuclide that has a half-life of 22.2 y, and therefore records processes over the past ~125 y (Robbins, 1978; Nittrouer et al., 1979). Our focus is to document and understand patterns of sediment accumulation over decadal to centennial time scales on an active continental margin, in particular the upper slope and shelf off the Umpqua River, Oregon. Studies of the Umpqua, a small, mountainous river (SMR) system, provide an opportunity to add to the growing body of knowledge on this type of river dispersal system (Nittrouer, 1999; Wheatcroft, 2000; Carter et al., 2010). Doing so is important, because SMRs supply roughly 50% of the sediment to the global ocean (Milliman and Farnsworth, 2011) and a similar fraction of bioactive constituents, in particular particulate organic carbon (e.g., Lyons et al., 2002; Hatten et al., 2012; Goñi et al., 2013). Although much is known about SMRs and how they differ from large river systems, we still lack an appreciation of how variations in key parameters combine to determine the location and character of margin depocenters (e.g., Walsh and Nittrouer, 2009). This lack of understanding arises for two reasons. First, the realization that SMRs are globally significant is a relatively recent one, thus the number of detailed case studies is small compared to the potential diversity of system behaviors (i.e., we are data starved). Second, as noted above, there is a general lack of theory by which to categorize SMRs. A recent compilation of river dispersal systems by Walsh and Nittrouer (2009) is an important advance toward a theoretical understanding of margin sedimentation. In this study, river sediment load, mean significant wave height, tidal range, and shelf width were used to identify five types of river dispersal systems. Of the five, the most common were ‘proximal accumulation dominated’ (PAD) and ‘marine dispersal dominated’ (MDD) systems, with the former defined on the basis of low wave and tidal energies. In contrast, MDD systems are characterized by high wave and/or tidal current energies that preclude deposition proximal to the river mouth and cause efficient along- and across-shelf sediment dispersal. Because of the energetic wave conditions along the US West Coast (Allan and Komar, 2006), all river dispersal systems in the region were classified as MDDs (Walsh and Nittrouer, 2009). The objectives of this paper are threefold. First, we seek to quantify sediment accumulation rates over the past 125 y on the shelf and upper slope (b200 m) off the Umpqua River. Because spatial gradients in SARs can be large on continental shelves (e.g., Sommerfield and Nittrouer, 1999), a high-density array of stations was studied. Second, we interpret the patterns of sediment accumulation, both in space and time, in light of potential natural and anthropogenic forcings. Specifically, we are interested in determining whether large-scale timber harvesting in the Umpqua River basin has had a measurable impact on the accumulation of sediment on the continental margin. Third, we compare our results to other marine dispersal dominated systems on the U.S. West Coast (e.g., Columbia and Eel Rivers) with the aim of further refining this important class of sediment dispersal system. 2. Methods 2.1. Study area The Umpqua River drains a 12,103-km 2 mountainous basin that heads in the High Cascades (maximal elevation: Mount Thielsen, 2799 m) and then flows through the Western Cascades, Klamath Mountains, and Oregon Coast Range (OCR) before discharging into the Pacific Ocean around 43.7° N. The geology, geomorphology, and sediment production of the Umpqua basin are varied (Fig. 1a, b),

45

with the permeable, low-relief volcanic rocks of the High Cascades resulting in little runoff and sediment production (Jefferson et al., 2010), whereas the more deeply dissected mountains of the Western Cascades support higher rates of runoff, with mass wasting the dominant mechanism of hillslope sediment erosion (Grant and Wolff, 1991). Further west, the South Fork of the Umpqua River passes through the rugged Klamath Mountains, which comprise a complex assemblage of meta-sedimentary, volcanic and intrusive igneous rocks with variable sediment production rates (Wallick et al., 2011). Lastly, the mainstem Umpqua River flows through the OCR, which in the Umpqua basin is composed of the Paleocene to Eocene soft marine sediments of the Elkton and Tyee Formations (Molenaar, 1985); both of which are susceptible to landsliding (e.g., Roering et al., 2005). Except for its central portion, the Umpqua River basin is densely forested with Sitka spruce and western hemlock dominating near the coast and Douglas-fir elsewhere (Franklin and Dyrness, 1973). Since colonization by Euro-Americans in the mid 19th century a range of land use activities, including placer mining, in-stream gravel mining, farming, ranching, and timber harvesting, have occurred in the basin that may have altered sediment production and transport (Wallick et al., 2011). Of these land uses, timber harvesting has likely been the most important. For example, Douglas County, whose boundaries follow closely that of the Umpqua basin, ranked second in the nation in timber production from 1949 to 1970 (Wall, 1972). A recent analysis of Landsat imagery by Cohen et al. (2002) indicates that approximately 18% of the forested portion of the Umpqua basin was logged between 1972 and 1995 (Fig. 1c). Because this was a period of declining timber harvest in the region (discussed below), we can expect that the cumulative watershed effects of logging, which included extensive road building, log ‘drives’ and splash damming (Brown and Krygier, 1971; Reid and Dunne, 1984; Miller, 2010), to have been potentially greater in earlier parts of the 20th century. The Umpqua River basin's climate is highly seasonal, with wet, stormy winters (NDJFM) and relatively dry summers (AMJJASO). Average annual precipitation (1971–2000) ranges from ~ 800 mm/y in the interior lowland portion of the basin to > 2700 mm/y in the OCR (Fig. 1d). Peak flows measured at the U.S. Geological Survey's gauging station near Elkton (station number 143210), which captures runoff from ~ 79% of the basin, are due to winter frontal systems (i.e., ‘atmospheric rivers,’ Ralph et al., 2006; Dettinger, 2011), with the largest flows resulting from rain-on-snow events. During the period of record (1905–present), the largest measured flow was during the December 1964 flood, when discharge reached 7505 m 3/s. This flood was likely the highest in the basin since the rain-on-snow event of 1861 (Wallick et al., 2011). Since the 1950's several hydroelectric and flood-control dams were constructed mainly in the North Fork of the Umpqua, but also in the Upper Cow Creek basin. Because these dams are in the upper part of their respective basins they likely have little impact on the mainstem discharge measured at Elkton (Wallick et al., 2011). As mentioned above, sediment production within the basin varies as a function of geologic province. To estimate the relative contribution of different parts of the basin to total sediment load, we used data compiled by Curtiss (1975) from five gauging stations in the basin (Fig. 1b). The estimated load at the mainstem station (143210, ‘Umpqua River near Elkton’) is 3.2 Mt/y, with contributions of 1.5 Mt/y and 0.7 Mt/y from the South and North Forks of the Umpqua, respectively (stations 143120 and 143195). Subtracting these amounts from the Elkton load provides an estimate for the 1733-km2 portion of the basin between the South and North Fork stations (‘between stations’; Fig. 1b) and the Elkton station of 0.9 Mt/y (yield = 525 t/km2/y). To estimate the load of the 2564 km2 area downstream of Elkton, which includes the Smith River, Elk Creek and numerous other tributaries draining the OCR, we have averaged the yield obtained from stations

46

R.A. Wheatcroft et al. / Marine Geology 339 (2013) 44–56

Fig. 1. Maps of the Umpqua River basin showing (a) elevation, the main tributaries (Smith, North and South Forks), and Mt. Thielsen (MT), (b) the main geological/geomorphic regions, the locations of the USGS gauging stations used for estimating sediment load (red dots), and subdivisions of the basin mentioned in the text (dashed lines), (c) the spatial distribution of harvested forest parcels between 1972 and 1995, as well as the boundary of Douglas County, Oregon (red line), and (d) average annual precipitation (1971–2000). Sources: stream channel and elevation data were obtained from websites maintained by the U. S. Geological Survey's, National Hydrography Dataset and Earth Resources Observation and Science Center, respectively; geomorphic regions, and the location of the USGS gauging stations are after Curtiss (1975) and Wallick et al. (2011); timber harvest data are from Cohen et al. (2002); and annual precipitation data are from Wieczorek and LaMotte (2010).

143220, 143207 and the ungauged mainstem (Fig. 1b). The result is 364 t/km2/y or 0.9 Mt/y. When summed these estimates imply a total load of 4.1 Mt/y for the Umpqua River of which roughly 18% is from the North Fork, roughly 38% from the South Fork, and the remainder (45%) is from the (mainly) Oregon Coast Range province above and below the Elkton gauging station. These load estimates result in a basin-averaged sediment yield of 340 t/km2/y, which for comparison is approximately five times less than that of the Eel River (Sommerfield and Nittrouer, 1999) and roughly 100-times less than some SMRs on the North Island of New Zealand and Taiwan (Milliman and Farnsworth, 2011). The marine portion of the study area is the continental shelf and uppermost slope adjacent to the Umpqua River; a region along the central Oregon margin demarcated by Heceta Bank to the northwest and Cape Arago to the south (Fig. 2b). Within this region, the shelf is approximately 30-km wide with a slope of ~ 0.3°, a relatively smooth bathymetry, and coast-parallel isobaths that diverge offshore to the north. The major exception to this along-margin uniformity is a narrow topographic high (‘Siltcoos Bank’) that extends to the northwest from the mid-shelf. Unlike many other areas of the U.S. West Coast, a submarine canyon does not dissect the shelf even though the Umpqua has a watershed size comparable to other rivers with offshore canyons (e.g., Rogue, Klamath/Trinity, Eel, Salinas). Prior sedimentologic studies of the Umpqua margin based on grab sampling and box coring have provided a broad description of sediment grain size patterns and cross-shelf facies variations (Kulm et al., 1975) that provided guidance during our coring efforts.

Along the Oregon coast, poleward winds and downwelling-favorable conditions are prevalent during winter, leading to offshore displacement of the equatorward-flowing California Current and the development of the poleward Inshore Countercurrent (also known as the Davidson Current) (e.g., Hickey, 1998; Strub and James, 2000). Given the circulation over the shelf during winter, northward transport of both surface and near-bed sediment plumes is expected to dominate as it does on the Columbia and Eel margins (e.g., Sternberg, 1986; Geyer et al., 2004). Lastly, recent work by Kniskern et al. (2011) has demonstrated that the Umpqua River discharge peaks are coherent with energetic waves and strong, southerly winds. 2.2. Field sampling The majority of the sampling was conducted during cruise W0906C on the R/V Wecoma in June–July 2009. Seabed samples were collected along a series of shore-perpendicular transects separated along-margin by 8 km and extending from 70 to 200-m water depth (Fig. 2a). Transects are identified by letter from A (south) to H (north); thus, for example, core B90 is taken along the B transect at a nominal water depth of 90 m. A 20 by 30-cm box corer was used to collect cores at sites with a muddy substrate (typically deeper than 80 m), whereas a hydraulically dampened gravity corer (see Stevens et al., 2007) was used to recover cores from sandy bottoms at sites shallower than 80 m. Core lengths ranged from 35 to 70 cm. From both corers a 10.8-cm diameter tube core was sectioned at 1-cm intervals. Samples for grain-size and radionuclide analyses

R.A. Wheatcroft et al. / Marine Geology 339 (2013) 44–56

47

Fig. 2. Map of study area showing (a) the location of the coring stations on shore-perpendicular transects (A–H), and (b) the regional setting of the Umpqua River margin. In panel (a) slow core locations are indicated by gray circles, whereas box core locations are indicated by white squares. Contours are water depth in meters. Numbers in parentheses adjacent to rivers in panel (b) are the average sediment load in Mt/y estimated at each gauging station (Wheatcroft and Sommerfield, 2005). Major geographical and bathymetric features of the region are also identified (e.g., Heceta Bank, Cape Arago, Siltcoos Bank).

were stored in refrigerators until laboratory processing. In addition, box cores were subsampled with two orthogonal Plexiglas trays that were X-radiographed onboard ship using a digital radiography system (Wheatcroft et al., 2006). Additional cores were collected within the study area between 1998 and 2011 during cruises aboard R/V Wecoma (e.g., Wheatcroft and Sommerfield, 2005; Reimers et al., 2012). Similar sample collection and processing protocols were followed on these cruises. 2.3. Laboratory analyses Samples for grain size determination were homogenized and analyzed using standard sieve and pipette techniques (Galehouse, 1971; Ingram, 1971). Briefly, samples were disaggregated sonically and organic matter oxidized using 30% hydrogen peroxide for 48 h. Particles of coarse silt size and larger were separated by wet-sieving samples through a 32-μm sieve. The portion that did not pass was dried and shaken for 20 min on stacked sieves at 0.25-ϕ intervals (ϕ = − log2 (d) and d is in mm; Ingram, 1971). The grain size distribution of material b32-μm was determined using a settling column and Stokes' Law, with a solution of sodium hexametaphosphate (1 g/L) added to inhibit flocculation. Fine silts were separated at 0.5-ϕ intervals, whereas clays were separated at 1-ϕ intervals. The graphical technique of Inman (1952) was used to determine sediment size statistics. In this study, we report the weight percent of mud (i.e., clay + silt). Samples for γ-ray spectroscopy were weighed wet, dried for 2–4 days at 60 °C, and then reweighed, thereby providing water content estimates. Samples were then ground to a consistent texture using a mortar and pestle and placed in pre-weighed polystyrene counting jars. Roughly 30 g of sediment was counted for ≥ 24 h on one of two identical Canberra GL2020RS LEGe planar γ-ray detectors with 2000-mm 2 windows, and the 46.5 and 351.9-keV photopeaks used to quantify 210Pb and 214Pb activities, respectively. The latter was used to calculate supported 210Pb activities. Detector efficiency and self-absorption corrections were made using NIST standards following standard techniques (e.g., Gilmore and Hemingway, 1995) and activities were decay corrected to the time of core collection. [Note that these detectors are not optimized for measuring higher

energy γ rays, therefore Cs-137 (photopeak of 661.6 keV) is rarely above the minimum detection limits.] Sediment accumulation rates (SARs) were estimated from the excess 210Pb profiles by assuming that the system was at steady state and there was negligible biodiffusive mixing below ~ 10-cm depth in the seabed (discussed below). Thus, there is a balance between advection (burial) and radioactive decay (e.g., Nittrouer et al., 1979; Anderson et al., 1988): S

∂A ¼ −λA ∂z

1

where S is the SAR (mm/y), A is excess 210Pb activity (dpm/g; 1 dpm/g = 16.67 Bq/kg), z is depth within the seabed, and λ is the 210Pb decay constant (0.0311/y). With the boundary conditions that Az = A0 at z = 0 and Az = 0 at z = ∞, the linearized solution to the advection-decay balance is: λ lnAz ¼ lnA0 − z: S

2

An unweighted, least squares line fit to ln Az versus z data yields a slope (i.e., λ/S) that is used to calculate the SAR. Using the water content data, profiles of excess 210Pb activity versus cumulative mass were obtained, and these profiles were used to estimate mass accumulation rates (g/cm 2/y). Standard parametric statistical techniques were used to test the significance of the regression lines (i.e., testing the null hypothesis that the slope was equal to 0) and estimate 95% confidence limits for the SAR (Sokal and Rohlf, 1981). 3. Results 3.1. Grain size and X-radiography Surficial grain-size analyses reveal the Umpqua margin to be a classic example of a mid-shelf mud deposit (McCave, 1972), whereby grain size at inshore stations (b80 m) is a muddy sand (typically >90% sand), stations in the mid-shelf region (90 to 120 m) are

48

R.A. Wheatcroft et al. / Marine Geology 339 (2013) 44–56

sandy muds, and stations on the outer shelf (120 to 150 m) are again sands, but with a higher mud fraction (25–50% mud) than in the inshore stations (Fig. 3a). The grain size of the upper slope (150 to 200 m) is uniformly mud. This pattern is two-dimensional on the shelf, however, in that in the northern sections of the study area (transects G & H), there is no mud on the mid shelf, and sands extend from the inner shelf to the shelf break. The width of the mud deposit also varies along margin, such that in plan view, the mud patch has a mitten-like shape (Fig. 3b), with the maximal cross-shelf width on the D transect. The likely cause of this shape is the presence of Siltcoos Bank, which may have a significant, albeit unstudied, effect on bottom boundary layer flow (cf. Hastings et al., 2012). X-radiographs from the box cores contain little evidence for undisturbed layers on the Umpqua margin. Instead the upper 40 cm of the seabed is characterized by abundant biogenic sedimentary structures, including diverse feeding and locomotion structures, as well as lined and unlined burrows and tubes. In a few instances, cores contain vague contacts, but these were visible only after applying horizontal edge detection and contrast enhancement algorithms (e.g., Wheatcroft et al., 2006). 3.2.

210

Pb geochronology

Within the 73 cores studied herein, there are essentially three types of excess 210Pb profile. The most common profile type (‘type 1’, n = 44),

characteristic of the distal mid to outer shelf and upper slope, exhibits a well-defined surface-mixing layer (SML) that is 8 to 12 cm thick with excess 210Pb activities in the range of 6 to 25 dpm/g (Fig. 4a–c). Underlying the SML is a zone where the excess 210Pb activity decreases logarithmically, thereby permitting a single estimate of the sediment accumulation rate (SAR). Type-2 profiles (n = 8) are restricted to sandy, inshore (≤80 m) sites and are characterized by a thick SML (up to 30 cm) with low excess 210Pb activities of 0.5 to 3 dpm/g (Fig. 4d–f). Below this zone excess 210Pb activities also decrease logarithmically, again permitting estimates of a single SAR. The third profile type (type 3, n = 21), which occurs mainly within the muddy portions of the proximal mid-shelf, is characterized by a roughly 10-cm thick SML underlain by two zones of logarithmically decreasing excess 210Pb activity that have distinct slopes (Fig. 5a–f). In all cases, the upper (i.e., younger) zone above the ‘inflection point’ (the point where the slope changes) is characterized by a gentler excess 210Pb gradient; this composite profile allows two estimates of the SAR. By assuming that physical and biological sediment mixing is absent below the SML, estimates of apparent SARs for type-1 and type-2 profiles were readily obtained using Eq. 2 (Figs. 4 & 5; Table 1, Supplementary information). The treatment of type-3 profiles was slightly more complex. In this case, an analysis of covariance (ANCOVA) was used to test whether the slopes of the regression lines above and below the inflection points were statistically different (Sokal and Rohlf, 1981). This proved to be the case in all but one of the twenty-one type-3 profiles (Table 1), thus we obtained two SAR estimates, which subsequently are referred to as the ‘upper’ and ‘lower’ SARs, at these twenty sites. Overall results, which include the upper SARs, indicate that they range over an order of magnitude from 0.4 mm/y to 5.6 mm/y. To a first approximation, the SAR pattern is similar to the grain size pattern (Fig. 3b), with maximal values in the landward edge of the mid-shelf mud deposit adjacent to the Umpqua River mouth (Fig. 6a). There, in water depths of 85–100 m accumulation rates are 3 to 5.6 mm/y, whereas inshore the SARs decrease by a factor of two or more. Offshore of the muddy depocenter, in water depths >115 m there is a broad region of the shelf and upper slope where SARs are 0.5 to 2 mm/y, which is essentially the relative rate of sea-level rise in this portion of the Oregon coast (Komar et al., 2011). If the lower SARs from the type-3 composite 210Pb profiles are used instead, then the maximal sediment accumulation rate within the entire study area is only 2.3 mm/y and the depocenter adjacent to the river mouth is nonexistent (Fig. 6b). 4. Discussion 4.1. Nonsteady sediment accumulation rates

Fig. 3. Sediment grain size patterns on the Umpqua margin. (a) Histogram of mud (i.e., b63 μm) content (wt.%) as a function of water depth. (b) Contour map of mud content (wt.%) showing a well-defined mid-shelf mud deposit directly offshore and north of the river mouth. Depth contours in meters.

Before discussing the implications of the spatial pattern of sediment accumulation (e.g., Fig. 6), it is necessary to explore in depth the evidence for, and causes of, the apparent increase in recent sediment accumulation rates. The main evidence for a change in SARs on the shelf is the type-3 excess 210Pb profiles that exhibit a slope break at depth (Fig. 5). An alternative explanation for this type of profile is that deep biodiffusive mixing extends to the inflection point depth (e.g., Nittrouer et al., 1983/1984; Anderson et al., 1988), and that by not accounting for this diffusive bioturbation we have overestimated the upper SARs. Although particle displacement below 210Pb-derived surface-mixing layers is certainly possible (e.g., Wheatcroft and Drake, 2003), for several reasons we believe this explanation to be less compelling. Unfortunately, we cannot provide direct evidence for a lack of deep biodiffusive mixing, therefore the discussions below should be assessed in light of this important caveat. The first piece of indirect evidence in support of a lack of deep biodiffusive mixing is that the inflection-point depth varies by a factor of two (14 to 30 cm) and shows no systematic variation with water depth or position relative to the river mouth. This pattern is

R.A. Wheatcroft et al. / Marine Geology 339 (2013) 44–56

49

Excess Pb-210 (dpm/g) 0.1 0

1

10

30

0.1

10

0.1

30

B150 0.7 mm/y -10

-10

-20

-20

-20

a

b

-30

0.1

1

10

0.1

30

1

10

0.1

30

0

0

-10

-10

-10

-20

-20

-20

-30

-30

-30

-40

C70a 3.0 mm/y -40

D70 -40 1.4 mm/y

d

30

10

30

e

-50

c

-30

0

-50

10

D125 1.0 mm/y

-10

-30

1

0

A200 0.5 mm/y

Depth (cm)

1

0

-50

1

H80 1.1 mm/y

f

Fig. 4. Representative type-1 and type-2 profiles of excess 210Pb activity (dpm/g) as a function of depth in the sediment (cm). Error bars in depth represent the 1-cm-thick interval of the samples, whereas error bars in activity are the 1-σ counting errors. Panels (a) to (c), which are from offshore sites, illustrate canonical 210Pb profiles (type 1) with a well-defined surface mixing layer (SML) of order 10-cm thick underlain by a zone of logarithmic decline (cf. Nittrouer et al., 1979). Panels (d) to (f) depict type-2 profiles measured in shallow portions of the shelf that have lower activities and thicker SMLs. In all panels, the solid line is a best-fit regression that yields the listed sediment accumulation rate in mm/y.

difficult to reconcile with the likely distribution of megabenthic animals capable of mixing to great depths. For example, time-series sampling on southern California's Palos Verdes shelf (Stull et al., 1986) documented the recruitment and multi-year persistence of a

large echiuran worm (Listriolobus pelodes) that materially impacted shelf sediments. The distribution of the echiuran, however, was spatially aggregated and its depth within the sediment column was consistent throughout its range. There is no evidence for such spatial

Excess Pb-210 (dpm/g) 0.1

1

-10

1

A90 1.9 & 0.5 mm/y 1950

0.1

10

0

B80 3.2 & 1.5 mm/y 1988

-10

-20

-30

-30

-30

0.1 0

-10

b

-40

1

D80 5.0 & 1.9 mm/y 1975

10

0.1

-40 1

10

0.1

D115 1.9 & 0.5 mm/y 1961

-10

-20

-20

-20

-30

-30

-30

-40

d

-40

c 1

10

0

0

-10

10

C80 5.5 & 1.8 mm/y 1984

-10

-20

a

1

0

-20

-40

Depth (cm)

0.1

10 20

0

e

-40

F150 1.9 & 0.5 mm/y 1948

f

Fig. 5. Representative type-3 profiles of excess 210Pb activity (dpm/g) as a function of depth in the sediment (cm) that have an inflection point in the data below the SML. The paired best-fit lines were tested against the null hypothesis of equivalent slopes and rejected at the 0.05-level (Table 1). The date listed in each panel is the approximate year of the inflection point assuming a constant initial concentration of excess 210Pb (Appleby and Oldfield, 1978). See text for details.

50

R.A. Wheatcroft et al. / Marine Geology 339 (2013) 44–56

Table 1 Summary statistics for the analysis of covariance (ANCOVA) with the sample size (n), coefficient of determination (r2) and slope of the upper and lower regression lines and the overall p-value of the ANCOVA (Sokal and Rohlf, 1981). Core

A80 A90 B80 B90 B110 C80 C90 CD110 UA41 UA43 D80 D80a D90 D100 D110 D115 DE90 UM2 F110 F150 G110

Upper 2

n

r

5 5 6 6 5 5 4 5 5 4 6 5 5 5 4 4 4 5 4 5 4

0.98 0.95 0.97 0.95 0.63 0.88 0.98 0.90 0.93 0.98 0.98 0.91 0.96 0.99 0.98 0.98 0.83 0.94 0.95 0.96 0.94

Lower slope

n

r2

slope

p-Value

0.075 0.165 0.097 0.106 0.133 0.056 0.093 0.075 0.088 0.079 0.062 0.131 0.102 0.157 0.161 0.163 0.067 0.108 0.190 0.165 0.203

5 3 4 3 4 6 4 4 4 3 7 4 6 3 4 3 4 5 3 3 3

0.99 0.99 0.99 0.98 0.96 0.97 0.86 0.78 0.95 0.99 0.90 0.99 0.89 0.99 0.99 0.94 0.98 0.90 0.99 0.99 0.99

0.242 0.605 0.212 0.299 0.372 0.173 0.184 0.376 0.176 0.703 0.163 0.423 0.263 0.475 0.515 0.673 0.249 0.247 0.520 0.683 0.682

0.0001 0.0005 0.0008 0.0048 0.0308 0.0016 0.1553 (NS) 0.0322 0.0299 0.0001 0.0088 0.0005 0.0262 0.0001 0.0002 0.0219 0.0041 0.0304 0.0074 0.0006 0.0066

aggregation in the Umpqua data. Second, it is unlikely that deepdwelling animals mix sediment in a biodiffusive manner, and it is only biodiffusive mixing that can confound estimates of sediment accumulation (Nittrouer et al., 1983/1984). Rather a great number of studies (e.g., Rhoads, 1974; Smith et al., 1986; François et al., 2002; Hughes et al., 2005) have shown that large, deep-dwelling animals mix sediment non-locally and that the main effect of this mixing is to introduce significant vertical variation in tracer concentrations. In fact there is some evidence for such mixing in both type-1 and type-3 profiles (e.g., Fig. 5c, d), but the majority of profiles are relatively ‘smooth’ implying lack of deep, non-local mixing.

The third piece of evidence that argues against deep, biodiffusive mixing as the cause of type-3 excess 210Pb profiles is provided by grain size data across the inflection-point depth. In six randomly chosen cores with type-3 profiles, grain size measured at three depth intervals above (post) and three depth intervals below (pre) the inflection-point depth showed that in all cores the upper depth horizons (i.e., post inflection point) were finer. In two of the cores, the fining (expressed as % mud) was not statistically significant, but in the others (i.e., A80, B80, D80 and DE90) the grain size change was significant (t-test, Sokal and Rohlf, 1981) and consisted of a 2% to 13% increase in mud content (Table 2). It is difficult to envision how deep mixing would lead to the observed grain size variation. Instead such vertical variation in a conservative property suggests allogenic changes involving either the rate of supply of sediment to the shelf depocenter or a fining of sediment supplied by the Umpqua River (discussed below). If one accepts that the apparent increase in sediment delivery to the shelf is real, then several questions arise, including: (1) when did it occur, (2) by how much did it increase, and (3) why did it occur. To estimate when the change in sediment accumulation rate occurred we have used two methods. The first employs the constant initial concentration model (CIC), which assumes that changes in sediment accumulation rate are possible and will result in a proportionate change in the 210Pb flux to the seafloor (Appleby and Oldfield, 1978; Robbins, 1978; Blais et al., 1998). Thus, the time of deposition of the inflection point (Tip), is defined as: −1

T ip ¼ k

ln

Ao Aip

3

where Aip is the excess 210Pb activity at the inflection point depth and all other terms are as defined previously (Section 2.3). Because the result is sensitive to both activities, and surface activities can be particularly variable, we averaged the upper three sample points to estimate Ao, and calculated inflection-point age for each core. This approach yielded an average inflection-point age of 1967 ± 13 y, with a total range between the years 1948 and 1988. The second method assumes a zero age for sediment at the base of the surface-mixing

Fig. 6. Maps showing the spatial distribution of sediment accumulation rates (mm/y) estimated from excess 210Pb profiles. Panel (a) represents results from type-1 and type-2 profiles, as well as the upper (post inflection point) type-3 best-fit lines. Panel (b) is from type-1 and type-2 profiles and the lower (pre inflection point) type-3 best-fit lines. See text for details.

R.A. Wheatcroft et al. / Marine Geology 339 (2013) 44–56

Core

Pre

±s.d.

Post

±s.d.

p-Level

A80 B80 C80 D80 D115 DE90

50.2 39.8 37.9 50.1 44.3 45.0

7.9 0.2 3.2 3.8 3.6 2.0

61.0 41.5 39.3 58.7 46.0 58.1

1.6 0.7 2.7 1.8 1.5 4.8

0.05 0.03 0.24 (NS) 0.04 0.30 (NS) 0.02

400

Cumul. Res. Q (m3/s)

Table 2 Summary of mud content (wt.%) for samples below (pre) and above (post) the inflection point. p-levels for a two-tailed t-test are given in the column on the right (Sokal and Rohlf, 1981). s.d. = standard deviation.

51

a

200 0

-200 -400 1900 8000

1920

1940

1960

1980

2000

1920

1940

1960

1980

2000

1920

1940

1960

1980

2000

6

4X

Post-SAR (mm/y)

2X 4

2

0 0

1

2

3

Pre-SAR (mm/y) Fig. 7. Bivariate plot of sediment accumulation rate (SAR) in mm/y above (post) and below (pre) the inflection points observed in type-3 excess 210Pb profiles. The dashed lines represent two and fourfold increases in the SAR.

6000 4000 2000 0 1900 2

c

Board Feet Harvestedx 109

layer (SML), and then divides the upper SAR into the linear distance between the base of the SML and the inflection-point depth. Applied to each type-3 profile, this approach yielded a similar result for the mean age of the inflection point: 1966 ± 14 y. The question of by how much did the rate of sediment accumulation increase can also be approached in two ways. First, in terms of linear sediment accumulation rates at sites with type-3 profiles, a simple linear regression of the lower (pre) and upper (post) SARs indicates an average increase of 2.7 (r 2 = 0.91), with nearly all cores falling in the range of a two to fourfold increase (Fig. 7). Second, to estimate the shelf-wide increase in mass accumulation rate (MAR) associated with the timber harvesting we calculated area-weighted average MARs based on the pre- and post-inflection point data and obtained values of 0.11 g/cm 2/y and 0.14 g/cm 2/y, respectively. Differencing these values results in a post inflection point (i.e., post 1967) increase of 0.03 g/cm2/y, which spread over the ~1750 km2 of the gridding domain translates to an approximately 0.5 Mt/y increase in sediment accumulation on the shelf. Depending on the long-term annual load estimate used for the Umpqua River (e.g., Curtiss, 1975; Karlin, 1980; Wheatcroft and Sommerfield, 2005), this enhanced accumulation implies a load increase from 12 to 36%. Lastly, there is the question of causality. Absent a major volcanic or seismic perturbation to the river basin (e.g., Major et al., 2000; Hovius et al., 2011), which is the case for the past 300 y, there are at least two additional factors that may explain the increase in SAR during the second half of the 20th century: hydroclimatic changes and land use changes. In the case of the former, stream flow data indicate that both the average discharge and peak discharge for the Umpqua River were anomalously high during the middle portion of the 20th century (Fig. 8a, b). In particular, the cumulative residual of annual discharge (e.g., Inman and Jenkins, 1999; Wheatcroft and Sommerfield, 2005), which is a measure of temporal variation in overall runoff in the basin, exhibits a prolonged period of increase from approximately 1945 to

Qpeak (m3/s)

b

1.5 1 0.5 0 1900

Year Fig. 8. Important hydroclimatic and land use data for the Umpqua River basin during the 20th century. Panel (a) is the cumulative residual of mean annual discharge (m3/s) measured at the Elkton gauging station. Upward trending portions of the line indicate anomalously wet periods, whereas downward trends indicate dry periods (cf. Wheatcroft and Sommerfield, 2005). Panel (b) is the peak annual discharge (m3/s) measured at the Elkton gauging station. Panel (c) is a record of board feet (×109) harvested in Douglas County, Oregon, which coincides closely with the boundary of the Umpqua River basin (Fig. 1c). The source for the discharge data is the U.S. Geological Survey, whereas the timber harvest data are from Andrews and Kutara (2005).

1975 (Fig. 8a). In addition, the top seven peak annual discharges, including the December 1964 flood, measured on the Umpqua River during the period of record occurred during that same 30-y period (Fig. 8b). This temporal pattern in average discharge and peak discharge is consistent with the historical record of the Pacific Decadal Oscillation (PDO) (Mantua et al., 1997), which is known to influence the climate of the Pacific Northwest (Dettinger et al., 1998; Brown and Comrie, 2004; Praskievicz and Chang, 2009). Thus, the cool phase of the PDO that existed from 1944 to 1977 coincides almost exactly with the period of elevated mean and peak flows of the Umpqua River; whereas the preceding ~25-y period (1920–1944) was in the warm phase of the PDO and average discharge and peak discharge were lower than the long-term average. In addition to these hydroclimatic changes, there is a long history of land use – primarily industrial logging – in the river basins of the Pacific Northwest that likely has increased the flux of sediment from the uplands. Beginning in the late 19th century the vast coniferous forests of the Pacific Northwest have provided a huge volume of timber that fueled the region's economy, but with myriad cumulative effects on the ecology, biogeochemistry and geomorphology of the area (e.g., Reid, 1993). In particular, forest clear cutting and road building, by destabilizing hill slopes and increasing the frequency of mass wasting events (Brown and Krygier, 1971; Reid and Dunne, 1984; Montgomery et al., 2000), have been shown to increase sediment flux from small (i.e., 10s ha) catchments by over an order of

52

R.A. Wheatcroft et al. / Marine Geology 339 (2013) 44–56

magnitude (e.g., Swantson and Swanson, 1976; Grant and Wolff, 1991; Lewis, 1998). Considering the potential widespread effects of logging in the Pacific Northwest, direct data on pre-satellite era (~ 1970) amounts and spatial patterns of road building and timber harvesting are surprisingly rare. Tax records compiled by Andrews and Kutara (2005) were therefore used to reconstruct the quantity of board feet harvested from Douglas County, which has a 96% overlap with the Umpqua River basin (Fig. 1c). Because county-specific records extend to only 1925, whereas statewide records are continuous to 1905, we regressed Douglas County timber harvest numbers to those for the entire state during 1925–1944. The resultant power–law regression equation (r 2 = 0.87) was then used to estimate the Douglas County timber harvest for 1905–1925. The composite curve (Fig. 8c) can be broken into four periods: (1) a pre-1940 period of slowly increasing timber production punctuated by a slight drop during the Great Depression; (2) a 15-y period of rapidly increasing timber harvesting that culminated in the peak harvest year of 1955; (3) a 35-y period of fluctuating, but elevated harvests; and (4) a rapid decline after about 1990 that has stabilized at roughly 25% of the peak harvest. In summary, the temporal coincidence of intensive timber harvesting with wet hydroclimatic conditions in the years following World War II may have resulted in a multi-decade period of increased sediment yield from the uplands. To assess the extent to which temporal changes in hydroclimate could increase sediment delivery, we estimated annual sediment loads using a time-constant rating curve (Wheatcroft and Sommerfield, 2005) for the period 1905 to 2009. The resultant average annual loads before and after the inflection-point date (i.e., 1967) are 1.4 Mt/y and 1.2 Mt/y, respectively. However, because there is a large amount of inter-annual variation (e.g., coefficient of variations are 146% and 122%, respectively) these average loads are not statistically significant (t-test, p = 0.26). This result suggests that hydroclimatic changes alone cannot explain the increase in SARs measured on the Umpqua shelf and that changes in sediment yield must have occurred (i.e., a time-varying rating curve is needed; Warrick et al., 2013). The most likely cause of a change in sediment production is widespread logging in the uplands (Fig. 8c). That there should be a measurable logging signal on the shelf is by no means a given, however. The reason for this statement is that although paired catchment studies in the Pacific Northwest and elsewhere have demonstrated that logging leads to elevated sediment fluxes from disturbed uplands (e.g., Swantson and Swanson, 1976; Reid and Dunne, 1984; Grant and Wolff, 1991; Lewis, 1998), it is uncertain whether these effects will scale up to whole river basins that are 1000 s km 2 in area. There are several reasons for this uncertainty. First, averaged over an entire river basin the relative area disturbed by logging in any one year is a small fraction of the basin. In the case of the Umpqua, the Landsat data indicate that b 1% of the basin was logged annually during the peak years in the mid 1980's (Cohen et al., 2002). If this area is increased in proportion to the board feet harvested in the mid 1980's compared to the peak harvest year of 1955, then the maximal area logged in a year increases to only 1.1% of the basin, which is still far less than typical harvest intensity in paired watershed studies (>80%; e.g., Lewis, 1998). The second factor to consider is the increased storage capacity of large basins; whereby sediment liberated from the hill slope is deposited before reaching the channel network or is sequestered in lower floodplains or estuaries (e.g., Walling, 2006; Lancaster et al., 2010). Lastly, there are a variety of autogenic factors involving thresholds (e.g., landslides, river bank failures) that may conspire to “shred” environmental signals as they propagate through sediment routing systems (e.g., Jerolmack and Paola, 2010). All of these factors were used by Ambers (2001) to explain the lack of a logging signal in a flood control reservoir in the Western Cascades. In the present case, however, it appears that a logging signal has indeed been recorded in the Umpqua shelf sediments in terms of both a fining of grain size and increased accumulation rate. This

finding implies that the cumulative effects of logging (e.g., Reid, 1993) are sufficient to overwhelm any sources of autogenic noise. Further, there is a relatively short time lag between the period of maximal upland disturbance (i.e., 1945–1955; Fig. 8c) and the estimated age of the SAR increase on the continental shelf (~ 1967). This close temporal coupling between a perturbation and a signal in this small, mountainous river system is likely due to the limited storage capacity for fine-grained sediment in the basin and the apparent fine-grained nature of the signal recorded in the marine sink. That is, it is unlikely that coarse material fluxes would have had such a short response time (e.g., Pearce and Watson, 1986). 4.2. Comparison to other marine dispersal dominated systems Two rivers – the Columbia and Eel – provide well-studied dispersal systems with which to compare results obtained herein. All three systems discharge their sediment into an energetic coastal ocean with large waves, strong, mainly along-shelf, currents, and a relatively narrow shelf. As such these dispersal systems should be classified similarly within the Walsh and Nittrouer (2009) hierarchy, and in fact the Eel and Columbia are both listed as examples of marine dispersal dominated (MDD) systems. An important attribute of MDD systems is that energetic waves preclude deposition of fine-grained sediment in shallow water depths proximal to the river mouth, and in this regard the Umpqua matches predictions. A closer look at these three river systems, however, suggests that there are some important differences in both the pattern of centennial sediment accumulation on the shelf and the processes by which those sediment accumulation patterns were likely formed. In terms of pattern, both the Columbia and the Umpqua dispersal systems are characterized by well-defined mid-shelf mud deposits, with regions of the outer shelf that have significantly coarser sediment (Fig. 3; Nittrouer et al., 1979). In addition, maximal sediment accumulation rates are found directly off the river mouth, albeit at intermediate (70–90 m) water depths because of the energetic wave conditions characteristic of the Pacific Northwest (Allan and Komar, 2006). There are also clear proximal–distal indicators within the Columbia and Umpqua depocenters, although the nature of these indicators differs between the two systems. On the Columbia shelf, there is a distinct distal fining along-shelf to the northwest (Nittrouer et al., 1979), whereas on the Umpqua shelf the fraction of fine-grained sediment decreases to the northwest (Fig. 3). In contrast, the Eel River's dispersal system is characterized by fine-grained sediment across the entire mid and outer shelf, maximal sediment accumulation rates are displaced 10–20 km along-shelf, and there are no clear proximal–distal indicators (Sommerfield and Nittrouer, 1999; Wheatcroft and Borgeld, 2000). These differences can be explained by considering two key features of the river systems. The first involves the relative amount of sediment delivered by the river compared to the dispersal capacity of the adjacent coastal ocean, which is a combination of wave resuspension and current speed/direction. Because all of these terms, especially the latter, are difficult to quantify precisely, we discuss them in relative terms only. At ~18 Mt/y (Wheatcroft and Sommerfield, 2005), the Eel River's annual sediment load is roughly twice that of the Columbia and 5–10 times greater than the Umpqua. Wave energy on all three margins is extreme compared to global averages, but appears to be greatest on the Umpqua, with the Eel and Columbia slightly lower (Allan and Komar, 2006). Given the lack of long-term data on sediment flux (Ogston et al., 2004 on the Eel is the exception), it is reasonable to assume that sediment dispersal capacity on all three margins to be roughly the same. Thus, the greater fluvial sediment load, along with the fact that it discharges onto a narrower shelf, is the likely reason that fine-grained sediment on the Eel margin extends across the shelf onto the upper slope, whereas there is a coarse outer-shelf on both the Columbia and Umpqua. Moreover, the relatively low sediment input by the Umpqua results in

R.A. Wheatcroft et al. / Marine Geology 339 (2013) 44–56

the along-shelf coarsening (Fig. 3), whereby fine-grained sediment is simply winnowed from the seabed at a rate in excess of delivery. The second factor to consider is the level of river–ocean coherence for all three systems. Set mainly by the source basin's area, river– ocean coherence potentially exists for small rivers like the Umpqua and Eel (Kniskern et al., 2011), but does not for large systems like the Columbia. The latter rivers are characterized by a multi-month lag between the period of time that sediment is delivered to the coast and the period of time that maximal waves, and hence sediment dispersal occurs. Thus, on the Columbia fluvial sediment flux is greatest during the late-Spring driven by snowmelt in the Rockies and Cascades, whereas sediment resuspension and along-shelf sediment fluxes are highest in the late Fall and Winter when storms occur in the Gulf of Alaska. Therefore, the Columbia River delivers most of its sediment load during relatively quiescent wave conditions that allow deposition immediately off the river mouth, but that sediment is reworked northwestward during the subsequent winter storm period (Nittrouer et al., 1979; Sternberg, 1986). This pattern of delivery and reworking causes maximal accumulation rates to be near the river mouth and the clear proximal–distal fining that is observed on the Washington shelf. For the Umpqua and Eel, maximal sediment delivery occurs during Fall and Winter storms when strong along-shelf winds and energetic wave conditions exist (Kniskern et al., 2011). On the Eel margin, this river–ocean coherence is thought to facilitate the development of wave-supported gravity flows that deliver large amounts of sediment to the mid-shelf with few proximal–distal indicators (Wheatcroft et al., 1997; Traykovski et al., 2000; Wheatcroft and Borgeld, 2000; Harris et al., 2005). Subsequent wave and current reworking does little to change the pattern of sediment accumulation, which is essentially set at the event time scale (Sommerfield and Nittrouer, 1999). But why then does the Umpqua, which as a small river has river– ocean coherence (Kniskern et al., 2011), share some attributes with the Columbia River dispersal system? To address this question consider that on all shelves there is a right-skewed probability distribution function (pdf) of significant wave height (Hs) that governs the frequency of wave resuspension (Fig. 9). [Note that although other wave parameters influence wave orbital speeds and thus sediment resuspension (Wiberg and Sherwood, 2008), to a first approximation Hs is the dominant factor.] Within a given shelf's Hs spectrum there is a value, here termed Hs′, which represents wave conditions when sediment is delivered and this parameter determines the degree to which a system is event vs. post-event dispersal dominated. On the Columbia, the Spring–Summer delivery of sediment means that Hs′ is relatively small (Fig. 9), thus post depositional reworking is large

Fig. 9. Schematic showing a generic probability distribution function of significant wave height. The relative location along the x-axis that a river delivers most of its sediment determines the importance of post-depositional reworking. Systems such as the Eel deliver their sediment during extreme wave conditions, thus there is little post-depositional reworking of terrestrial sediment. In contrast, the Columbia delivers the majority of its sediment during relatively low wave conditions, thus there is a great deal of post-depositional reworking. The Umpqua is intermediate and therefore its depocenter shares attributes of the Eel and Columbia.

53

and a depocenter is formed with clear proximal–distal indicators. On the flood-dominated Eel shelf, Hs′ is large during fluvial sediment delivery, thus post-depositional reworking is minimal. Analysis of NDBC buoy records by Kniskern et al. (2011) indicates that Hs′ on the Umpqua shelf is, in a relative sense, intermediate (Fig. 9). That is, even though wave energy is high during Umpqua River floods, a larger fraction of high-wave events occurs during non-flood periods (so called ‘dry storms’; Ogston et al., 2004), thus post-depositional wave-induced sediment reworking is greater than on the Eel, but less than on the Columbia. Therefore, the Umpqua River depocenter shares elements of both the Columbia and Eel. The preceding suggests that an additional criterion to consider in the Walsh and Nittrouer (2009) dispersal system hierarchy is the relative wave energy during periods of peak sediment delivery (i.e., Hs′). 4.3. Broader implications The presence of a mid-shelf depocenter where fine-grained sediment is accumulating at rates 2–4 times greater than ambient shelf settings has important implications for the biogeochemistry, ecology and sedimentology of the continental shelf. In particular, recent results reported by Hastings et al. (2012) indicate that particulate organic carbon (POC) contents in the depocenter are elevated relative to ambient sediments and have a more terrestrial signature. Therefore, we expect terrestrial burial fluxes of POC to be significantly greater within the depocenter. Based on the relatively young Δ 14C ages of fluvial POC (~ 200 to 400 years), it appears a large fraction of the terrestrial organic matter carried by the Umpqua River is modern across a wide range of discharges (Goñi et al., 2013). Thus, the elevated burial fluxes within the depocenter likely involves an appreciable fraction of biogenic POC, as opposed to petrogenic POC (Hilton et al., 2008; Blair et al., 2010), suggesting that a possible side effect of the industrial logging within the Umpqua basin is enhanced sequestration of photosynthetically derived organic matter from the uplands. Whether this enhanced sequestration is greater than the opposing effect of greater terrestrial POC degradation due to surface disturbances (e.g., Stallard, 1998; Dymond, 2010) is an important unknown. Although the correlation between static seafloor properties (e.g., grain size, carbon content) and the species composition, biomass, and the abundance of macrobenthos have been recognized for some time, it has been demonstrated (see discussion in Snelgrove and Butman, 1994) that processes are in fact more likely to determine pattern. Despite this realization, recent efforts, motivated by marine spatial planning (e.g., Gaines et al., 2010), have focused on mapping habitat variability on margin-wide spatial scales. Termed the ‘benthoscape’ in analogy with terrestrial landscape ecology (Zajac, 2008), acoustic remote sensing techniques are most often used in mapping efforts. Because these techniques are by definition indirect, the resultant classifications of the benthoscape are fairly crude (e.g., rock vs. mud). The growing appreciation that terrestrial depocenters, where sediment and organic carbon are accumulating at elevated rates, exist on continental margins offshore small mountainous river systems (cf. Walsh and Nittrouer, 2009) suggests that a more sophisticated consideration of the ‘sediment-covered seafloor’ of active margins would be fruitful. Lastly, this study demonstrates that the cumulative effects of timber harvesting at the basin scale are expressed as an increase in sediment accumulation rates and a shift in sediment grain size toward finer particles in the marine sink. Because a similar pattern has been observed on the Eel margin (e.g., Sommerfield et al., 2002; Leithold et al., 2005; Sommerfield and Wheatcroft, 2007), it is tempting to imply that timber harvesting results in the delivery of more fine grained sediment to river channels and that this material is simply propagated through the sediment routing system (i.e., the grain size signal on the shelf sink is set by the source). Unfortunately, the complexities of sediment delivery to the seabed and its subsequent reworking make this simple conclusion premature. Rather it may

54

R.A. Wheatcroft et al. / Marine Geology 339 (2013) 44–56

simply be that timber harvesting, by increasing the frequency of mass wasting events on hillslopes, leads to an increase in sediment export, but no change in grain size. Instead the observed fining may arise from the inability of post depositional reworking to winnow fines (e.g., Law et al., 2008) under increased deposition rates. 5. Summary To better understand spatial and temporal patterns of centennial sediment accumulation on continental margins a dense array of cores was collected on the shelf and upper slope off the Umpqua River, Oregon. Data obtained from the cores included X-radiographs, surficial grain size, and profiles of excess 210Pb, the latter of which were used to estimate linear sediment accumulation rates (SARs) and mass accumulation rates (MARs). The following conclusions can be made from these data: • There is a mid-shelf mud deposit directly offshore and to the north of the Umpqua River mouth that extends from approximately 90-m water depth to 120 m. The shape of the mud deposit in plan view appears to be influenced by a topographic high, Siltcoos Bank. • There is little evidence of preserved bedding in the X-radiographs, suggesting that either event beds are not formed on the Umpqua margin or that they are destroyed by post-depositional bioturbation. • Three types of excess 210Pb profiles were measured. Type-1 and type-2 profiles had 10-cm or >20-cm thick surface-mixing layers (SMLs), respectively that were underlain by a region of logarithmically decreasing activity. These profile types yielded a single estimate of the SAR. Type-3 profiles had ~10-cm-thick SMLs underlain by a composite profile with a slope break that permitted an upper and lower SAR to be estimated. • If the upper, more recent SARs from the type-3 profiles are included, then accumulation rates range from 0.4 to 5.6 mm/y and there is a clearly defined depocenter at ~90-m water depth where SARs are 2 to 3 times greater than ambient shelf sediments (Fig. 6a). If the lower (earlier) SARs from the type-3 profiles are used, then the depocenter essentially disappears and the rates vary from 0.4 to 2.3 mm/y. • The likely explanation for the composite, type-3 profiles is the temporal coincidence of wet hydroclimatic conditions with widespread industrial logging in the Umpqua basin in the years following World War II, which increased the flux of sediment from the uplands. • The existence of a logging signal on the shelf and the short lag between the period of maximal timber harvesting and the change in sediment accumulation rate suggest a close coupling between source and sink in this small, mountainous river system that likely stems from the limited storage capacity in the basin and the cumulative effects of the land use disturbance. Supplementary data to this article can be found online at http:// dx.doi.org/10.1016/j.margeo.2013.04.015. Acknowledgments Help in field collection and laboratory analysis of samples was provided by the captain and crew of the R/V Wecoma (Ret.), Daryl Swenson, Chris Moser, Paul Walczak, Yvan Alleau, Gabe Acosta, Danielle Asson, Jessica Bechler, Kaitlyn Butz, Andy Gray, Roxanne Hastings, Jeff Hatten, Nguyen Minh-Chau, Erik Mulrooney, Jason Padgett, Isaac Shepherd, Lauren Smith, and Maryann Tekverk. Jon Warrick (USGS) provided excellent comments on an earlier draft of the manuscript. Comments by an anonymous reviewer and Basil Gomez (U Hawaii) were helpful. Funding for this work was provided by the National Science Foundation's Carbon and Water Program, including grant EAR 0628487 to M. A. Goñi and R. A. Wheatcroft and grant EAR 0628490 to J. C. Borgeld.

References Allan, J.C., Komar, P.D., 2006. Climate controls on US West Coast erosion processes. Journal of Coastal Research 22, 511–529. Allen, P.A., 2008. From landscapes into geological history. Nature 451, 274–276. Ambers, R.K.R., 2001. Using the sediment record in a western Oregon flood-control reservoir to assess the influence of storm intensity and logging on sediment yield. Journal of Hydrology 244, 181–200. Anderson, R.F., Bopp, R.F., Buessler, K.O., Biscaye, P.E., 1988. Mixing of particles and organic constituents in sediments from the continental shelf and slope off Cape Cod: SEEP-I results. Continental Shelf Research 8, 925–946. Andrews, A., Kutara, K., 2005. Oregon's Timber Harvests: 1849–2004. Oregon Department of Forestry, Salem, OR (160 pp.). Appleby, P.G., Oldfield, F., 1978. The calculation of 210Pb dates assuming a constant rate of supply of unsupported 210Pb to the sediment. Catena 5, 1–8. Blair, N.E., Leithold, E.L., Brackley, H., Trustum, N., Page, M., Childress, L., 2010. Terrestrial sources and export of particulate organic carbon in the Waipaoa sedimentary system: problems, progress and processes. Marine Geology 270, 108–118. Blais, J.M., France, R.L., Kimpe, L.E., Cornett, R.J., 1998. Climatic changes in northwestern Ontario have had a greater effect on erosion and sediment accumulation than logging and fire: evidence from 210Pb chronology in lake sediments. Biogeochemistry 43, 235–252. Brown, D.P., Comrie, A.C., 2004. A winter precipitation ‘dipole’ in the western United States associated with multidecadal ENSO variability. Geophysical Research Letters 31, L09302. Brown, G.W., Krygier, J.T., 1971. Clear-cut logging and sediment production in the Oregon Coast Range. Water Resources Research 7, 1189–1198. Carter, L., Orpin, A.R., Kuehl, S.A., 2010. From mountain source to ocean sink: the passage of sediment across an active margin, Waipaoa Sedimentary System, New Zealand. Marine Geology 270, 1–10. Cohen, W.B., Spies, T.A., Alig, R.J., Oetter, D.R., Maiersperger, T.K., Fiorella, M., 2002. Characterizing 23 years (1972–95) of stand replacement disturbance in western Oregon forests with Landsat imagery. Ecosystems 5, 122–137. Curtiss, D.A., 1975. Sediment yields of streams in the Umpqua River basin, Oregon. U. S. Geological Survey Open-File Report. (1 plate with text). Dettinger, M.D., 2011. Climate change, atmospheric rivers, and floods in California—a multimodel analysis of storm frequency and magnitude changes. Journal of the American Water Resources Association 47, 514–523. Dettinger, M.D., Cayan, D.R., Diaz, H.F., Meko, D.M., 1998. North–south precipitation patterns in western North America on interannual-to-interdecadal timescales. Journal of Climate 11, 3095–3111. Dymond, J.R., 2010. Soil erosion in New Zealand is a net sink of CO2. Earth Surface Processes and Landforms 35, 1763–1772. François, F., Gerino, M., Stora, G., Durbec, J.P., Poggiale, J.C., 2002. Functional approach to sediment reworking by gallery-forming macrobenthic organisms: modeling and application with the polychaete Nereis diversicolor. Marine Ecology Progress Series 229, 127–136. Franklin, J.F., Dyrness, C.T., 1973. Natural vegetation of Oregon and Washington. U.S. Forest Service General Technical Report PNW-8, Portland, OR (417 pp.). Gaines, S.D., White, C., Carr, M.H., Palumbi, S.R., 2010. Designing marine reserve networks for both conservation and fisheries management. Proceedings of the National Academy of Sciences of the United States of America 107, 18286–18293. Galehouse, J.S., 1971. Sedimentation analysis. In: Carver, R.E. (Ed.), Procedures in Sedimentary Petrology. Wiley, New York, NY, pp. 69–94. Geyer, W.R., Hill, P.S., Kineke, G.C., 2004. The transport, transformation and dispersal of sediment by buoyant coastal flows. Continental Shelf Research 24, 927–949. Gilmore, G., Hemingway, J.D., 1995. Practical Gamma-Ray Spectrometry. John Wiley and Sons, Chichester, UK (314 pp.). Gomez, B., Carter, L., Trustum, N.A., 2007. A 2400 yr record of natural events and anthropogenic impacts in intercorrelated terrestrial and marine sediment cores: Waipaoa sedimentary systems, New Zealand. Geological Society of America Bulletin 119, 1415–1432. Goñi, M.A., Hatten, J.A., Wheatcroft, R.A., Borgeld, J.C., 2013. Particulate organic matter export by two contrasting small mountainous rivers from the Pacific Northwest, U.S.A. Journal of Geophysical Research—Biogeosciences 118, 1–23. http://dx.doi.org/ 10.1002/jgrg.20024. Grant, G.E., Wolff, A.L., 1991. Long-term patterns of sediment transport after timber harvest, Western Cascade Mountains, Oregon, USA. International Association of Hydrological Sciences Publication, No. 203, pp. 31–40. Harris, C.K., Traykovski, P.A., Geyer, W.R., 2005. Flood dispersal and deposition by nearbed gravitational sediment flows and oceanographic transport: a numerical modeling study of the Eel River shelf, northern California. Journal of Geophysical Research 110 (C9), C09025. Hastings, R.H., Goñi, M.A., Wheatcroft, R.A., Borgeld, J.C., 2012. A terrestrial organic matter depocenter on a high-energy margin: the Umpqua River system, Oregon. Continental Shelf Research 39–40, 78–91. Hatten, J.A., Goñi, M.A., Wheatcroft, R.A., 2012. Chemical characteristics of particulate organic matter from a small, mountainous river system in the Oregon Coast Range, USA. Biogeochemistry 107, 43–66. Hickey, B., 1998. Coastal oceanography of western North America from the tip of Baja California to Vancouver Island. In: Robinson, A.R., Brink, K.H. (Eds.), The Sea. The Global Coastal Ocean, 11. John Wiley & Sons, New York, NY, pp. 345–393. Hilton, R.G., Galy, A., Hovius, N., Chen, M.-C., Horng, M.-J., Chen, H., 2008. Tropicalcyclone-driven erosion of the terrestrial biosphere from mountains. Nature Geoscience 1, 759–762. http://dx.doi.org/10.1038/ngeo333.

R.A. Wheatcroft et al. / Marine Geology 339 (2013) 44–56 Hovius, N., Meunier, P., Ching-Weei, L., Hongey, C., Yue-Gau, C., Dadson, S., Ming-Jame, H., Lines, M., 2011. Prolonged seismically induced erosion and the mass balance of a large earthquake. Earth and Planetary Science Letters 304, 347–355. Hughes, D.J., Brown, L., Cook, G.T., Cowie, G., Gage, J.D., Good, E., Kennedy, H., MacKenzie, A.B., Papadimitriou, S., Shimmield, G.B., Thomson, J., Williams, M., 2005. The effects of megafaunal burrows on radiotracer profiles and organic composition in deep-sea sediments: preliminary results from two sites in the bathyal north-east Atlantic. Deep-Sea Research Part I 52, 1–13. Ingram, R.L., 1971. Sieve analysis. In: Carver, R.E. (Ed.), Procedures in Sedimentary Petrology. Wiley, New York, NY, pp. 49–67. Inman, D.L., 1952. Measures for describing the size distributions of sediments. Journal of Sedimentary Petrology 22, 125–145. Inman, D.L., Jenkins, S.A., 1999. Climate change and the episodicity of sediment flux of small California rivers. Journal of Geology 107, 251–270. Jefferson, A., Grant, G.E., Lewis, S.L., Lancaster, S.T., 2010. Coevolution of hydrology and topography on a basalt landscape in the Oregon Cascade Range, USA. Earth Surface Processes and Landforms 35, 803–816. Jerolmack, D.J., Paola, C., 2010. Shredding of environmental signals by sediment transport. Geophysical Research Letters 37, L19401. Karlin, R., 1980. Sediment sources and clay mineral distributions off the Oregon coast. Journal of Sedimentary Petrology 50, 543–560. Kniskern, T.A., Warrick, J.A., Farnsworth, K.L., Wheatcroft, R.A., Goñi, M.A., 2011. Coherence of river and ocean conditions along the US West Coast during storms. Continental Shelf Research 31, 789–805. Komar, P.D., Allan, J.C., Ruggiero, P., 2011. Sea level variations along the U.S. Pacific Northwest coast: tectonic and climatic controls. Journal of Coastal Research 27, 808–823. Kulm, L.D., Roush, R.C., Harlett, J.C., Neudeck, R.H., Chambers, D.M., Runge, E.J., 1975. Oregon continental shelf sedimentation: interrelationships of facies distribution and sedimentary processes. Journal of Geology 83, 145–175. Lancaster, S.T., Underwood, E.F., Frueh, W.T., 2010. Sediment reservoirs at mountain stream confluences: dynamics and effects of tributaries dominated by debrisflow and fluvial processes. Geological Society of America Bulletin 122, 1775–1786. Law, B.A., Hill, P.S., Milligan, T.G., Curran, K.J., Wiberg, P.L., Wheatcroft, R.A., 2008. Size sorting of fine-grained sediments during erosion: results from the western Gulf of Lions. Continental Shelf Research 28, 1935–1946. Leithold, E.L., Perkey, D.W., Blair, N.E., Creamer, T.N., 2005. Sedimentation and carbon burial on the northern California continental shelf: the signatures of land-use change. Continental Shelf Research 25, 349–371. Lewis, J., 1998. Evaluating the impacts of logging activities on erosion and suspended sediment transport in Caspar Creek watersheds. U.S. Forest Service General Technical Report PSW-GTR-168, pp. 55–69. Lyons, W.B., Nezat, C.A., Carey, A.E., Hicks, D.M., 2002. Organic carbon fluxes to the ocean from high-standing islands. Geology 30, 443–446. Major, J.J., Pierson, T.C., Dinehart, R.L., Costa, J.E., 2000. Sediment yield following severe volcanic disturbance—a two-decade perspective from Mount St. Helens. Geology 28, 819–822. Mantua, N.J., Hare, S.R., Zhang, Y., Wallace, J.M., Francis, R.C., 1997. A Pacific interdecadal climate oscillation with impacts on salmon production. Bulletin of the American Meteorological Society 78, 1069–1079. McCave, I.N., 1972. Transport and escape of fine-grained sediment from shelf areas. In: Swift, D.J.P., Duane, D.B., Pilkey, O.H. (Eds.), Shelf Sediment Transport: Process and Pattern. Dowden, Hutchinson and Ross, Stroudsburg, PA, pp. 225–248. Miller, R.R., 2010. Is the Past Present? Historical Splash-Dam Mapping and Stream Disturbance Detection in the Oregon Coastal Province. (Unpublished M.S. thesis) Oregon State University (96 pp.). Milliman, J.D., Farnsworth, K.L., 2011. River Discharge to the Coastal Ocean: A Global Synthesis. Cambridge University Press, Cambridge, UK (319 pp.). Molenaar, C.M., 1985. Depositional relations of Umpqua and Tyee Formations (Eocene), southwestern Oregon. American Association of Petroleum Geologists Bulletin 69, 1217–1229. Montgomery, D.R., Schmidt, K.M., Greenberg, H.M., Dietrich, W.E., 2000. Forest clearing and regional landsliding. Geology 28, 311–314. Nittrouer, C.A., 1999. STRATAFORM: overview of its design and synthesis of its results. Marine Geology 154, 3–12. Nittrouer, C.A., Sternberg, R.W., Carpenter, R., Bennett, J.T., 1979. The use of Pb-210 geochronology as a sedimentological tool: application to the Washington continental shelf. Marine Geology 31, 297–316. Nittrouer, C,A., DeMaster, D.J., McKee, B.A., Cutshall, N.H., Larsen, I.L., 1983/1984. The effects of mixing on Pb-210 accumulation rates for the Washington continental shelf. Marine Geology 54, 201–221. Ogston, A.S., Guerra, J.V., Sternberg, R.W., 2004. Interannual variability of nearbed sediment flux on the Eel shelf, northern California. Continental Shelf Research 24, 117–136. Oldfield, F., Wake, R., Boyle, J., Jones, R., Nolan, S., Gibbs, Z., Appleby, P., Fisher, E., Wolff, G., 2003. The late-Holocene history of Gromire Lake (NE England) and its catchment: a multiproxy reconstruction of past human impact. The Holocene 13, 677–690. Pasternack, G.B., Brush, G.S., Hilgartner, W.B., 2001. Impact of historic land-use change on sediment delivery to a Chesapeake Bay subestuarine delta. Earth Surface Processes and Landforms 26, 409–427. Pearce, A.J., Watson, A.J., 1986. Effects of earthquake-induced landslides on sediment budget and transport over a 50-yr period. Geology 14, 52–55. Praskievicz, S., Chang, H., 2009. Winter precipitation intensity and ENSO/PDO variability in the Willamette Valley, Oregon. International Journal of Climatology 29, 2033–2039.

55

Ralph, F.M., Neiman, P.J., Wick, G.A., Gutman, S.I., Dettinger, M.D., Cayan, D.R., White, A.B., 2006. Flooding on California's Russian River: the role of atmospheric rivers. Geophysical Research Letters 33, L13801. Reid, L.M., 1993. Research and cumulative watershed effects. U. S. Forest Service General Technical Report PSW-GTR-141 (118 pp.). Reid, L.M., Dunne, T., 1984. Sediment production from forest road surfaces. Water Resources Research 20, 1753–1761. Reimers, C.E., Özkan-Haller, H.T., Berg, P., Devol, A., McCann-Grosvenor, K., Sanders, R.D., 2012. Benthic oxygen consumption rates during hypoxic conditions on the Oregon continental shelf: evaluation of the eddy correlation method. Journal of Geophysical Research 117, C02021. Rhoads, D.C., 1974. Organism-sediment relations on the muddy sea floor. Oceanography and Marine Biology: An Annual Review 12, 263–300. Robbins, J.A., 1978. Geochemical and geophysical applications of radioactive lead. In: Nriagu, J.O. (Ed.), Biogeochemistry of Lead in the Environment. Elsevier Scientific, Amsterdam, pp. 285–393. Roering, J.J., Kirchner, J.W., Dietrich, W.E., 2005. Characterizing structural and lithologic controls on deep-seated landsliding: implications for topographic relief and landscape evolution in the Oregon Coast Range, USA. Geological Society of America Bulletin 117, 654–668. Sadler, P.M., 1981. Sediment accumulation rates and the completeness of stratigraphic sections. Journal of Geology 89, 569–584. Schumer, R., Jerolmack, D.J., 2009. Real and apparent changes in sediment deposition rates through time. Journal of Geophysical Research 114, F00A06. Smith, J.N., Boudreau, B.P., Noshkin, V., 1986. Plutonium and 210Pb distribution in northeast Atlantic sediments: subsurface anomalies caused by non-local mixing. Earth and Planetary Science Letters 81, 15–28. Snelgrove, P.V.R., Butman, C.A., 1994. Animal sediment relationships revisited: cause versus effect. Oceanography and Marine Biology: An Annual Review 32, 111–177. Sokal, R.R., Rohlf, F.J., 1981. Biometry, 2nd edition. W.H. Freeman and Company, New York, NY (859 pp.). Sommerfield, C.K., Nittrouer, C.A., 1999. Modern accumulation rates and a sediment budget for the Eel shelf: a flood-dominated depositional environment. Marine Geology 154, 227–241. Sommerfield, C.K., Wheatcroft, R.A., 2007. Late Holocene sediment accumulation on the northern California shelf: oceanic, fluvial, and anthropogenic influences. Geological Society of America Bulletin 119, 1120–1134. Sommerfield, C.K., Drake, D.E., Wheatcroft, R.A., 2002. Shelf record of climatic changes in flood magnitude and frequency, north-coastal California. Geology 30, 395–398. Stallard, R.F., 1998. Terrestrial sedimentation and the carbon cycle: coupling weathering and erosion to carbon burial. Global Biogeochemical Cycles 12, 231–257. Sternberg, R.W., 1986. Transport and accumulation of sediment on the Washington continental shelf, USA. Journal of the Geological Society of London 143, 945–956. Stevens, A.W., Wheatcroft, R.A., Wiberg, P.L., 2007. Seabed properties and sediment erodibility along the western Adriatic margin, Italy. Continental Shelf Research 27, 400–416. Strub, P.T., James, C., 2000. Altimeter-derived variability of surface velocities in the California current system: 2. Seasonal circulation and eddy statistics. Deep-Sea Research Part II 47, 831–870. Stull, J.K., Haydock, C.I., Montagne, D.E., 1986. Effects of Listriolobus pelodes (Echiura) on coastal shelf benthic communities and sediments modified by a major California wastewater discharge. Estuarine, Coastal and Shelf Science 22, 1–17. Swantson, D.N., Swanson, F.J., 1976. Timber harvesting, mass erosion, and steepland forest geomorphology in the Pacific Northwest. In: Coates, D.R. (Ed.), Geomorphology and Engineering. Dowden, Hutchinson and Ross, Stroudsbury, PA, pp. 221–236. Traykovski, P., Geyer, W.R., Irish, J.D., Lynch, J.F., 2000. The role of wave-induced density-driven fluid mud flows for cross-shelf transport on the Eel River continental shelf. Continental Shelf Research 20, 2113–2140. Wall, B.R., 1972. Log production in Washington and Oregon: an historical perspective. U.S. Forest Service Resource Bulletin PNW-42 (89 pp.). Wallick, J.R., O'Connor, J.E., Anderson, S., Keith, M., Cannon, C., Risley, J.C., 2011. Channel change and bed-material transport in the Umpqua River basin, Oregon. U. S. Geological Survey Scientific Investigations Report 2011–5041 (112 pp.). Walling, D.E., 2006. Human impact on land–ocean sediment transfer by the world's rivers. Geomorphology 79, 192–216. Walsh, J.P., Nittrouer, C.A., 2009. Understanding fine-grained river-sediment dispersal on continental margins. Marine Geology 263, 34–45. Warrick, J.A., Hatten, J.A., Pasternack, G.B., Gray, A.B., Goñi, M.A., Wheatcroft, R.A., 2012. The effects of wildfire on the sediment yield of a coastal California watershed. Geological Society of America Bulletin 124, 1130–1146. Warrick, J.A., Madej, M.A., Goñi, M.A., Wheatcroft, R.A., 2013. Trends in the suspendedsediment yields of coastal rivers of northern California, 1955–2010. Journal of Hydrology. http://dx.doi.org/10.1016/j.jhydrol.2013.02.041. Wheatcroft, R.A., 2000. Oceanic flood sedimentation: a new perspective. Continental Shelf Research 20, 2059–2066. Wheatcroft, R.A., Borgeld, J.C., 2000. Oceanic flood deposits on the northern California shelf: large-scale distribution and small-scale physical properties. Continental Shelf Research 20, 2163–2190. Wheatcroft, R.A., Drake, D.E., 2003. Post-depositional alteration and preservation of sedimentary event layers on continental margins, I. The role of episodic sedimentation. Marine Geology 199, 123–137. Wheatcroft, R.A., Sommerfield, C.K., 2005. River sediment flux and shelf sediment accumulation rates on the Pacific Northwest margin. Continental Shelf Research 25, 311–332. Wheatcroft, R.A., Sommerfield, C.K., Drake, D.E., Borgeld, J.C., Nittrouer, C.A., 1997. Rapid and widespread dispersal of flood sediment on the northern California margin. Geology 25, 163–166.

56

R.A. Wheatcroft et al. / Marine Geology 339 (2013) 44–56

Wheatcroft, R.A., Stevens, A.W., Hunt, L.M., Milligan, T.G., 2006. The large-scale distribution and internal geometry of the fall 2000 Po River flood deposit: evidence from digital X-radiography. Continental Shelf Research 26, 499–516. Wheatcroft, R.A., Goñi, M.A., Hatten, J.A., Pasternack, G.B., Warrick, J.A., 2010. The role of effective discharge in the ocean delivery of particulate organic carbon by small, mountainous river systems. Limnology and Oceanography 55, 161–171. Wiberg, P.L., Sherwood, C.R., 2008. Calculating wave-generated bottom orbital velocities from surface-wave parameters. Computers and Geosciences 34, 1243–1262.

Wieczorek, M.E., LaMotte, A.E., 2010. Attributes for MRB-E2RF1 catchments by major river basins in the conterminous United States: 30-year average annual precipitation, 1971–2000. Digital Data Series, DS-491-21.U. S. Geological Survey, Reston, VA. Wright, L.D., Nittrouer, C.A., 1995. Dispersal of river sediments in coastal seas: six contrasting cases. Estuaries 18, 494–508. Zajac, R.N., 2008. Challenges in marine, soft-sediment benthoscape ecology. Landscape Ecology 23, 7–18.