Late Quaternary History of Colluvial Deposition and Erosion in Hollows, Central California Coast Ranges

Late Quaternary history of colluvial deposition and erosion in hollows, central California Coast Ranges

STEVEN L. RENEAU Earth and Environmental Sciences Division, M.S. D462, Los Alamos National Laboratory, Los Alamos, New Mexico 87545 WILLIAM E. DIETRICH Department of Geology and Geophysics, University of California, Berkeley, California 94720 DOUGLAS J. DONAHUE | National Science Foundation Regional Facility for Radioisotope Analysis, University of Arizona, Tucson, A. J. TIMOTHY JULL J Arizona 85721 MEYER RUBIN U.S. Geological Survey, National Center 971, Reston, Virginia 22092

ABSTRACT side slopes and noses where contours are straight workers have proposed that the deposition of and convex, respectively. The concave topog- coarse permeable debris in hollows leads to sta- Colluvial deposits in hollows are the pre- raphy forces colluvial debris to converge toward bility, and subsequent erosion of weathered bed- dominant source of debris flows in many the axes of hollows, resulting in long-term depo- rock on the adjoining side slopes and noses mountainous areas, and the depositional his- sition (for example, Dietrich and others, 1986; results in topographic inversion. This model has tory of hollows can provide insight into land- Reneau and others, 1989). Landslides in the col- been proposed for colluvial deposits in the Ap- slide frequency and long-term hillslope proc- luvial deposits are the primary source of debris palachian Mountains and in Texas (Bryan, esses. Detailed study of colluvial deposits in flows in many mountainous areas, and the initial 1940; Mills, 1981). 20 hollows in the northern San Francisco Bay landslides and subsequent erosion by debris These conceptual models are not mutually area, California, reveals diverse histories of flows supply much sediment to stream channels exclusive, and each may be partially correct in erosion and deposition. Basal radiocarbon (for example, Pierson, 1977; Dietrich and explaining the evolution of hollows in a region. ages range from 1 to 29 ka and document Dunne, 1978; Lehre, 1981,1982; Okunishi and Detailed knowledge of the depositional history deposition of colluvium throughout this pe- lida, 1981; Dietrich and others, 1982; Tsuka- of hollows and the processes active at these sites riod at different sites. Ages from multiple moto and others, 1982; Reneau and Dietrich, is essential to resolve the relative importance of stratigraphic levels confirm that the deposits 1987a, 1987b). The long-term history of deposi- the above models. are cumulative and have thickened through tion and erosion in hollows is thus needed to In this paper, previous research on the deposi- the Holocene. Radiocarbon dates and strati- evaluate debris-flow hazards and the flux of sed- tional history of hollows in the western United graphic observations document unconformi- iment to streams. States is reviewed, followed by the presentation ties in hollows, reflecting incomplete evacua- Conceptual models of the evolution of collu- of detailed stratigraphic observations and chron- tion of colluvium during both Pleistocene and vial deposits in hollows are varied and have led ologic data from hollows in part of the central Holocene events and providing evidence for a to differing interpretations of the geomorphic California Coast Ranges. The research presented cycle of alternating accumulation and evacua- significance of the deposits. Several workers herein expands the earlier work of Reneau and tion of colluvium. This cycle has apparently have proposed that colluvial deposits in hollows others (1986), incorporating a greatly enlarged been affected by major climatic changes, and owe their existence to climatic changes, with data set. In particular, radiocarbon dates from the common occurrence of basal ages at ca. 9 stream incision under one climatic regime re- multiple stratigraphic levels and from different to 14 ka may record widespread slope insta- placed by colluvial deposition following a major longitudinal positions in several colluvial depos- bility during the Pleistocene-Holocene transi- climate change. Cotton and TePunga (1955) its help document the depositional history of in- tion. By analogy with modern landsliding, the postulated that a change from warmer intergla- dividual hollows in greater detail than previous- increased erosion may have been caused by cial to colder periglacial conditions led to collu- ly possible. The dating constrains the timing of an increased frequency of high-intensity vial deposition in New Zealand hollows, and past erosional events and improves the under- storms during extended periods of meridional Schlocker (1974) and Shlemon and others standing of the late Quaternary history of hol- flow in the upper atmosphere. The acceler- (1987) proposed that a change from a wetter lows in central California and the effects of ated discharge of colluvium from hillslopes glacial to a drier interglacial climate led to depo- regional climatic changes on hillslope erosional may have contributed to stream aggradation sition in California hollows. An alternative view processes. in diverse parts of California. is that continued accumulation of colluvium in hollows inherently leads to instability, and hol- PREVIOUS RESEARCH ON INTRODUCTION lows are thus characterized by a cycle of alter- DEPOSITIONAL HISTORY nating storage and discharge of colluvium. A OF HOLLOWS Hollows are an important part of many cyclic model, independent of climatic changes, landscapes, playing a critical role in the flux of has been proposed for sites in coastal California, Knowledge of the depositional history of hol- debris from hillslopes to streams. As defined by Oregon, Washington (for example, Pierson, lows in western North America has developed Hack and Goodlett (1960) and Hack (1965), 1977; Dietrich and Dunne, 1978; Lehre, 1981, through studies in California, Oregon, and hollows are parts of hillslopes where contours 1982; Dietrich and others, 1982), and Japan Washington during the past 15 yr. Dietrich and are concave-out from the slope, contrasting with (Okunishi and lida, 1981). In other areas, some Dunne (1978) initially described colluvial de-

Geological Society of America Bulletin, v. 102, p. 969-982, 13 figs., 2 tables, July 1990.

969

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/102/7/969/3381052/i0016-7606-102-7-969.pdf by guest on 24 September 2021 970 RENEAU AND OTHERS

posits in Oregon Coast Range hollows and residence time radiocarbon ages from buried that colluvial deposits in hollows are typically proposed that they were sites of episodic land- soils. Shlemon and others (1987) also proposed cumulative, thickening over time, and that the sliding, with the landslide scars filled with that the climatic change at the onset of the Hol- period of accumulation can exceed 10,000 yr. colluvium derived from the adjacent slopes. Ob- ocene led to initial accumulation of colluvium in This overlaps with the time scale of major cli- served textural variations included an increase in former Pleistocene stream channels. A similar matic change, and clusterings of basal radiocar- gravel content toward the base of the deposits, model had previously been advanced by bon dates suggest that climatic changes have and the basal gravels were interpreted as record- Schlocker (1974) for deposits on the Marin Pe- influenced the evacuation of colluvium from ing the winnowing of fine sediment by overland ninsula, immediately north of San Francisco. hollows in some regions. Relative dating criteria flow when the deposits were thin. Subsequently, Additional radiocarbon age determinations suggest that a wide range in age of colluvium Dietrich and others (1982) reported similar have been made from hollows in coastal Oregon may be present and that unconformities record- deposits on the Olympic Peninsula in Washing- and Washington. Benda and Dunne (1987) re- ing partial evacuation are present in some ton and estimated that colluvial accumulation ported basal radiocarbon dates from recent deposits. for 1,000 to 10,000 yr was required to replace landslide scars in three Oregon Coast Range hol- the sediment evacuated in a landslide. On the lows of 1.6, 6.4, and 9.4 ka and inferred that the STUDY AREA basis of detailed observations and measurements period between successive failures in this area is of landslide scars in central California grass- thus in the range of thousands of years. Basal The study sites are within the central Califor- lands, Lehre (1981, 1982) proposed essentially radiocarbon dates from nine additional hollows nia Coast Ranges of Marin and Alameda Coun- the same model for colluvial deposits north of in the Oregon Coast Range extend from 4 to ties, in the northern San Francisco Bay area San Francisco, emphasizing the role of revegeta- >40 ka, with seven sites clustering between 4 (Fig. 1). The climate is Mediterranean, with wet tion in stabilizing debris within landslide scars and 7.5 ka (Reneau, 1988). Basal radiocarbon winters and dry summers, moderated by coastal and the role of gullying in evacuating some dates from eight hollows on the Olympic Penin- fog. Average annual rainfall ranges from about hollows. sula of Washington range from 7-13 ka and 600 to 1,000 mm/yr and is greatest at sites near- A more diverse depositional history was sug- show a clustering between 7 and 10 ka (Reneau est the coast (Rantz, 1971). Native vegetation is gested by the study of colluvial deposits in the and others, 1989). The early Holocene was a varied and includes grassland; a northern coastal Redwood Creek basin of northwest California period of warmer, drier climate in Washington, scrub community dominated by coyote brush by Marron (1982, 1985). Abrupt changes in tex- and the basal ages suggest a period of wide- (Baccharis pilularis) and poison oak (Rhus di- ture and soil development that record probable spread landsliding during this drier period versiloba); a mixed hardwood forest dominated unconformities were observed within some de- (Reneau and others, 1989). In addition, dates by California laurel (Umbellularia californica), posits, and Marron proposed that partial remov- from multiple stratigraphic levels in 11 Oregon coast live oak (Quercus agrifolia), and Pacific al of colluvium by landsliding is common. On and Washington hollows document that the madrone (Arbutus menziesii); and a mixed the basis of relative weathering and soil devel- deposits there are cumulative and have progres- bishop pine (Pinus muricata) hardwood forest. sively thickened over time (Reneau, 1988; Ren- opment criteria, she also suggested that the col- Available paleoecological data in the San eau and others, 1989). luvium varies widely in age; radiocarbon dates Francisco Bay area record an increase in certain of 7-10 ka were obtained from one deposit, and In summary, the research to date indicates coniferous species in the latest Pleistocene that most deposits studied were more weathered than that one. Some deposits in Marron's study area have no topographic expression, and this, in combination with strongly weathered colluvium, suggests long-term stability of some sites. In central California, Dietrich and Dorn (1984) used pollen analysis to locate an 11-13.5 ka stratigraphic level within two colluvial deposits near Clear Lake. This level occurs near the base of one deposit and in the upper part of the second, recording differing periods of accumulation. Reneau and others (1986) subsequently reported basal radiocarbon dates from 11 Cali- fornia hollows. Most of these dates cluster between 9 and 15 ka, and Reneau and others (1986) proposed that evacuation of colluvium was more thorough and possibly more frequent under the latest Pleistocene vegetation and climate. Shlemon and others (1987) used soil stratigraphy to interpret the history of one hollow in Pacifica, south of San Francisco. A 6-m section in the upper basin was interpreted as a cumula- Figure 1. Location map of dated hollows in northern San Francisco Bay area. CC, Clare- tive deposit, with uninterrupted deposition for mont Canyon; GPB, Grizzly Peak Boulevard; LTC, Lone Tree Creek; MPC, Madrone Park 8,000-10,000 yr, whereas a recurrence interval Circle; NSP, North San Pedro Road; PCG, Pike County Gulch; SB, Skyline Boulevard; SPR, of about 1,000-4,000 yr for debris flows from San Pedro Ridge; TV, Third Valley. Late Pleistocene pollen sites of Rypins and others (1989) the lower basin was proposed based on mean at Point Reyes are shown by triangles.

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/102/7/969/3381052/i0016-7606-102-7-969.pdf by guest on 24 September 2021 QUATERNARY COLLUVIAL DEPOSITION AND EROSION, CALIFORNIA COAST RANGES 971

indicate a cooler, moister climate, as discussed Group (Radbruch, 1969). Colluvial soils devel- in hollows, leaving the basal colluvium in place by Johnson (1977), Adam and others (1981), oped on these units typically have high friction (Fig. 2B). Recent landslide scars on San Pedro and Rypins and others (1989). The nearest rec- angles and high porosities, and landslides in the Ridge have been relatively stable since initial ord spanning the Pleistocene-Holocene bound- colluvium are generally mobilized into rapid de- failure (Reneau, 1988), although significant, ary is from the Point Reyes Peninsula in Marin bris flows rather than slower earthflows or progressive erosion of colluvium can in some County (Fig. 1), where the major change from a slumps (for example, Reneau and Dietrich, cases occur over a period of years. For example, closed coniferous forest of Douglas fir, grand fir 1987b). a history of successive failures over an 18-yr (?), and bishop pine to coastal scrub and grass- Several studies have documented that hollows period is well documented in a relatively gentle land occurred between ca. 10.3 and 9.4 ka are the predominant source of shallow, debris- (17°) grassland hollow in the Lone Tree Creek (Rypins and others, 1989). Conifers were prob- flow-producing landslides in this region (Lehre, basin first studied by Lehre (1982) (Fig. 3), as ably also common at the San Pedro Ridge study 1981, 1982; Smith and Hart, 1982; Reneau and summarized by Wilson and others (1989). sites (Fig. 1) in the latest Pleistocene, contrasting Dietrich, 1987a, 1987b; Shlemon and others, with the modern mixed hardwood forest; an ex- 1987; Ellen and others, 1988; Howard and oth- STRATIGRAPHY AND amination of charcoal fragments from three San ers, 1988; Smith, 1988). Landslide characteris- CHRONOLOGY Pedro Ridge hollows, dated to 10-21 ka, re- tics for forested San Pedro Ridge (Fig. 1), where vealed a mixture of conifers and hardwoods, al- most of the radiocarbon dates in this study were In order to constrain the depositional history though species-level identification was not pos- obtained, are presented in Reneau and Dietrich of hollows, 35 radiocarbon dates have been ob- sible (O. K. Davis, 1986, written commun.). (1987a, 1987b) and Reneau (1988); similar tained from 20 hollows in the northern San Records from the last interstadial, ca. 30-29 ka, characteristics were seen in many other parts of Francisco Bay area. The sites consist of 10 hol- are available from Tomales Bay in Marin the region with similar soils and topography. On lows on San Pedro Ridge in Marin County, 7 County (Mason, 1934; Berger and Libby, 1966) San Pedro Ridge, recent landslide scars have a additional hollows in Marin County, and 3 hol- and from coastal Santa Cruz County (Adam and characteristic size that is much smaller than the lows in the Berkeley Hills of Alameda County others, 1981). These also suggest a cooler cli- size of each hollow (Reneau and Dietrich, (Fig. 1; Table 1). Study sites were chosen based mate than today, although the relative precipita- 1987b). The landslides typically evacuated col- on the availability of charcoal in the lower part tion is uncertain. luvium from only a part of a hollow, and debris of a deposit, so that the approximate initiation of Bedrock at the study sites is diverse but com- flows initiated in the upper part of a hollow deposition could be determined. Large charcoal prises units generally resistant to deep-seated caused varying amounts of erosion downslope fragments or collections of many small frag- landsliding. Most Marin County sites are under- (Fig. 2). In many cases, little erosion occurred in ments were dated using conventional beta lain by sandstone and shale of the Mesozoic the lower hollow or was limited to the uprooting counting methods where possible, and accelera- Franciscan assemblage, although sites in Third of trees and stripping of the uppermost soil layer tor mass spectrometry dating of small fragments Valley near Inverness are underlain by deeply (Fig. 2A), although the triggering of secondary was used when charcoal was less abundant. The weathered Mesozoic granitic rocks (Blake and landslides downslope by debris flows in some dates include samples from multiple stratigraph- others, 1974). Study sites in the Berkeley Hills of cases resulted in major evacuation (Fig. 2B; see ic levels in five deposits and different longitudi- Alameda County are underlain by Miocene sed- also Ellen, 1988). Failure planes developed nal positions in six deposits. Because of their imentary rocks, primarily chert of the Monterey within the colluvium in about half of the scars critical importance in understanding the history of hollows, site descriptions, including stratigraphic observations, are presented in some detail below. A / tributary /' / , M/ i hollow / .' / ' iii' '

runoff and colluvial transport path

temporary discontinous channel

bedrock-colluvium boundary

Figure 2. Sketches of some typical characteristics of recent landslides in San Pedro Ridge hollows. (A) Debris flow mobilized from landslide in upper hollow, causing minor erosion of colluvium in lower hollow. Old degraded landslide scar present in tributary hollow. (B) Debris flow originating in upper hollow, triggering secondary landslide downslope.

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/102/7/969/3381052/i0016-7606-102-7-969.pdf by guest on 24 September 2021 972 RENEAU AND OTHERS

The study sites vary in slope gradient, in upslope drainage area, and in the magnitude of topographic convergence (Table 2), and these variables may strongly influence the history of specific sites. For example, the analyses of Ren- eau and Dietrich (1987b) predict that min- imum failure volume will decrease with increas- ing slope gradient, so that failures could occur more frequently on steeper slopes. In addition, sites with large drainage areas, which typically branch and include tributary hollows (Reneau and Dietrich, 1987a), should be more influenced by landslides triggered upslope than are sites with small drainage areas. Local topographic convergence is important because it affects the flux of colluvium and runoff into hollows, with faster depositional rates and greater pore- pressure development generally expected in hollows with stronger convergence. The local topography can be quantified by a convergence angle, which is the angle between the orientation Figure 3. Photograph of 1974 landslide scar at Lone Tree Creek 1 site, taken in 1987. A of the hollow axis and the orientation of the history of successive failures is documented here; a fresh scar in center of photograph occurred adjacent side slopes (Fig. 4); sites with conver- in 1987, and, to left of person, a smaller failure occurred in 1982. Person at upper cross section gence angles of 30°-50° will be informally re- of Figure 12. ferred to as "distinct hollows" in this paper, and sites with convergence angles of 10°- 30° as "subtle hollows."

TABLE 1. RADIOCARBON DATES FROM SAN FRANCISCO BAY AREA HOLLOWS San Pedro Ridge

Site name Depth Radiocarbon age Laboratory Most of the age determinations for this study (m) (yr B.P.) number have been obtained from San Pedro Ridge San Pedro Ridge above San Rafael (Figs. 1 and 5), and this area Arbutus landslide, axis 2.6 14.280 ± 450* W-5456 provides the most detailed depositional records. Arbutus landslide, side 1.8 21,000 ± 500* W-5448 Calyptroderma landslide 2.5 15,550 + 160 AA-1616 Reneau and others (1986) reported 9 basal dates Cascade landslide 1.0 2360 ± 70 AA-2548 3.1 10,080 i 80 AA-2547 from 6 sites on this ridge, and a total of 22 dates Chockstone landslide, headscarp 3.2 9600 ± 110 AA-1612 from 10 hollows are now available (Table 1). 4.2 9830 ± 120 AA-1611 Chockstone landslide, lower scar 2.3 7650 ± 290 Beta-27254 These include radiocarbon dates from multiple Hole, uppermost scar 1.7 13,720 ± 300* W-5641 Hole, upper scar 1.8 11,200 ± 670* UCLA-2443 stratigraphic levels in four hollows and, in addi- Hole, head of gully 1.7 12,370 ± 550 GX-11083 tion, dates from different longitudinal positions Hole, mid-gully 3.2 13,000 i 300 W-5638 4.8 13,650 ± 300 W-5634 in four hollows. Bedrock at all sites is sandstone Hungry Vulture swale 1.5 1230 ± 250* W-5639 Island landslide, headscarp 1.5 21,160 ± 400 W-5667 and shale of the Franciscan assemblage, and 2.5 25,850 ± 330 AA-1613 vegetation is a mixed hardwood forest. Island landslide, island 0.55 1735 ± 95 AA-1614 2.5 9370 ± 80 AA-2546 The Hole. The best exposed and most thor- Lindenwood landslide, lower scar 1.5 18,995 ± 1140* UCLA-2435 Lindenwood landslide, mid-scar 1.5 13.350 ± 600* W-5635 oughly dated site on San Pedro Ridge is re- Mosquito landslide 1.5 11,670 + 1680 - 1390' GX-11163 ferred to as "the Hole," the largest landslide scar Rodine landslide 2.0 9740 ± 660* GX-11162 in the study area of Reneau and Dietrich Other Marin County sites (1987b) (Fig. 6). The site grades from a subtle Lone Tree Creek 1 2.5 12,290 ± 140 AA-1618 4.2 12,915 ± 760* UCLA-2441 hollow in the upper reaches to a distinct hollow Lone Tree Creek 2 2.65 9960 ± 100 AA-1617 along the lower scar; a pre-1982 channel head Madrone Park Circle 2.0 10,220 ± 100 AA-1619 North San Pedro Road 3.5 29,150 ± 450 AA-1610 was probably present at the downslope end of Pike County Gulch 3.5 6260 ± 250 Beta-27249 Third Valley 1, upper scar 2.5 10,560 ± 100 Beta-26812 the scar. The scar is about 100 m long, with a Third Valley 1, lower gully 3.2 9130 ± 400* (none) maximum depth of 7.0 m (Figs. 6C and 7A), Third Valley 2 2.2 10,090 ± 90 Beta-26502 3.3 10,580 ± 130 Beta-26503 and provides extensive exposures of the bedrock Berkeley Hills surface and the colluvial stratigraphy. The lower Claremont Canyon ? 11,875 ± 720* UCLA-2442 part of the scar has a steep-sided main axis with Grizzly Peak Blvd. 6 24,330 ± 350 AA-2545 Skyline Blvd. 5 24,030 ± 310 AA-2544 a distinct concave bedrock profile (Fig. 7A), and the landslide scarp extends up into and high- •Presenled in Reneau and others (1986). lights several subtle tributary hollows (Figs. 6B

TABLE 2. SITE DATA FROM DATED HOLLOWS IN THE SAN FRANCISCO BAY AREA

Site name Deposit Hollow Convergence Distance Upslope thickness gradient angle below drainage (m) (degrees) (degrees) ridge area

San Pedro Ridge Arbutus landslide 2.6 31 26 55 1,200 Calypuoderma landslide 2.1 35 16 135 1,300 Cascade landslide 3.6 29 44 95 6,300 Chockstone landslide, headscarp 5.5 30 41 60 2,000 Chockstone landslide, lower scar 2.3 27 47 95 5,500 convergence Hole, headscarp 2.3 38 II 45 400 angle Hole, inner scar 7.0 35 30 90 2,800 Hungry Vulture swale 1.5 31 47 155 7,900 Island landslide, headscarp 3.0 23 48 50 2,400 Island landslide, island 2.5 22 50 90 5,900 Figure 4. Schematic topographic map of Lindenwood landslide, lower scar 1.8 29 26 90 1,500 Lindenwood landslide, mid-scar 1.5 29 26 75 1,100 hillslope, illustrating convergence angle. Mosquito landslide 1.5 28 35 60 2,600 Rodine landslide 2.0 30 19 40 800

Other Marin County sites lows, suggesting more prolonged periods of flu- Lone Tree Cr. 1, upper scar 5.0 17 45 50 2,200 Lone Tree Cr. 1, lower scat 4.2 17 45 70 3,200 vial erosion here. The other sites in distinct Lone Tree Cr. 2 4.6 17 45 85 4,400 Madrone Park Circle 3.5 30 7 95 1,000 hollows yielded basal dates between 7 and 26 North San Pedro Road 3.5 14 16 65 700 ka, and four sites have yielded 9-12 ka dates. Pike County Gulch 3.5 25 46 155 7,300 Third Valley 1, upper scar 2.5 28 28 65 3,200 Eight radiocarbon dates have been obtained Third Valley 1. lower gully 3.2 14 48 190 21,300 Third Valley 2 5.5 21 35 55 1,900 from landslide scars in distinct hollows in three

Berkeley Hills adjacent basins near the Hole (Fig. 8), and each Claremont Canyon ? 15 ? 160 5,600 of these sites displays a unique stratigraphy. The GriKly Peak Blvd. 7+ 35 31 55 ? Skyline Blvd. 6+ 24 45 70 4,900 Island and Chockstone scars are the second and third largest scars in the study area of Reneau and Dietrich (1987b), with volumes of about 1,250 and 590 m3, whereas the 150-m3 size of and 6C). The deepest part of the deposit con- dates have been obtained from six distinct hol- the Mosquito scar is more typical. sists of about 4.0 m of coarse, angular, strat- lows on San Pedro Ridge, including the lower The Island landslide (Fig. 8A) is unusual in ified gravels overlain by finer-textured, matrix- part of the Hole. The youngest, a 1230 yr B.P. that a small part of the original surface was left supported colluvium. The lower gravels suggest date from Hungry Vulture swale (Table 1), was surrounded by the landslide scar, allowing two a talus apron produced by raveling off a bedrock obtained at the downslope end of a long hollow. separate cross sections to be examined. It is also slope; analogous modern deposits are present The deposit overlies water-polished bedrock; the lowest-gradient landslide studied on San below bedrock outcrops in some San Pedro similar bedrock was not seen beneath other hol- Pedro Ridge, with a 22°-23° slope. At the Ridge hollows. Radiocarbon dates obtained from the deepest part of the deposit at the Hole document rapid deposition of the lower gravels in the latest Pleis- tocene, and ages of 13,650 and 13,000 yr B.P. were obtained 4.8 and 3.2 m deep, respectively (Fig. 7A; Table 1). At the head of the inner gully, an additional date of 12,370 + 550 yr B.P. from the overlying matrix-supported colluvium, at a depth of 2.0 m (Fig. 7A), shows that coarse gravel deposition had ceased by this time. No evidence of unconformities has been found in exposures along the 100 m of scar, and a date of 13,720 yr B.P. from the uppermost scar, where maximum thickness is only about 2.5 m, reveals that deposition here began contemporaneously with deposition in the deeper part downslope (Fig. 7A). In the headscarp exposure, continu- ous deposition is suggested by a progressive decrease in gravel content higher in the section Figure 5. Topographic map showing dated hollows on San Pedro Ridge and locations of (Fig. 7B); this apparently records gradual areas shown in Figures 6B and 8A. A, Arbutus landslide; C, Calyptroderma landslide; Ca, changes in soil texture in the source area over Cascade landslide; Ch, Chockstone landslide; H, the Hole; HV, Hungry Vulture swale; I, Island time, possibly due to the progressive burial of landslide; L, Lindenwood landslide; M, Mosquito landslide; R, Rodine landslide. Roads shown bedrock exposures. by dashed lines. LC, Lindenwood Court in Glenwood subdivision of San Rafael. Contour Other Distinct Hollows. Basal radiocarbon interval is 50 ft; topography from San Quentin and San Rafael 7.5' quadrangles.

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/102/7/969/3381052/i0016-7606-102-7-969.pdf by guest on 24 September 2021 974 RENEAU AND OTHERS

headscarp, a 3-m-thick section displays a distinct unconformity at a depth of 1.35 m (Fig. 8B). Clasts in the lower colluvium are very friable, suggesting a greater age than most deposits, and radiocarbon dates of 21,160 and 25,850 yr B.P. from depths of 1.5 and 2.5 m confirm the greater apparent age (Table 1). The unconformity is marked by an abrupt change in color and texture, and lenses of well-sorted sand in the overlying deposit record overland flow; similar recent deposits of sand are seen at the downslope end of many landslide scars in this area. The sandy horizon grades up into the root-permeated A horizon, and the upper part of this A horizon contains sand apparently derived from a small landslide scar present immediately upslope (Fig. 8A). Two dates have been obtained from the lower cross section at the "island," yielding an age of 9370 yr B.P. for the base of the 2.5-m-thick section and an age of 1735 yr B.P. for a gravelly layer at a depth of 0.55 m (Fig. 8B, Table 1); the latter date documents that deposition continued into the late Holocene here. At the island, two discontinuous stone lines are present at depths of 1.0 and 1.55 m, and they may record uncon- Figure 6. (A) Photograph of the Hole, a 100-m-long, 2,600-m3 landslide scar that occurred formities or distinct depositional events in the in January 1982. Failure plane in the upper scar and on left side of the lower scar is at an lower hollow; the lower stone line seems to be approximately 1.5-m-deep soil horizon, below main rooting zone. Inner scar is a maximum of traceable to the unconformity in the headscarp, 7 m deep, and failure of openwork gravels here may have been due to rapid undrained loading bracketing the post-21 ka landslide to between caused by a debris flow generated upslope. (B) Topographic map of the Hole; surveyed in 1.7 and 9.4 ka. In sum, the dating and stratig- 1983 and 1984. (C) Isopach map of pre-failure thickness of colluvium at the Hole; depths to raphy at the Island landslide document that bedrock obtained from reconstructed topography, exposures in scar, and soil pits. Hachured different parts of this hollow have different his- line is 1982 scarp, which extends up into several subtle tributary hollows. Isopach contour tories and that incomplete evacuation of collu- interval = 1.0 m. vium has occurred in multiple Holocene landslides. Dates from different longitudinal positions tion (Figs. 8B and 9). Because most of the lower sites discussed above, and a date of 2360 yr B.P. were also obtained from the Chockstone land- section was eroded in 1982, the presence or ab- from a depth of 1.0 m (Table 1) documents slide scar (Figs. 8 and 9), recording a history sence of an unconformity cannot be determined, continued deposition into the late Holocene, as different from that of the Island basin. The col- although the existence of four recognized land- seen in the Island basin. No unconformity was luvium at the headscarp of the scar reached 5.5 slide scars upslope (Fig. 8A) suggests that such recognized in this section. m in thickness, consisting of a 3-m-thick layer of incomplete evacuation is possible. Subtle Hollows. Radiocarbon dates have coarse, angular, openwork gravel overlain by In the Mosquito landslide scar (Fig. 8), a been obtained from five landslide scars in subtle, finer-textured colluvium with matrix-supported basal age of 11,670 yr B.P. was obtained from weakly convergent hollows on San Pedro Ridge, gravel (Figs. 8B and 10); the lower gravel layer the lower axis (Table 1). A complex series of including the upper part of the Hole. All have pinched out downslope as the deposit thinned to overlapping scars is present in the upper basin yielded basal dates between 9 and 16 ka (Table about 3 m. Radiocarbon dates of 9830 and 9600 (Fig. 8A), and soil stratigraphy in the headscarp 1), overlapping with dates from the distinct hol- yr B.P. obtained from depths of 4.2 and 3.2 m of the 1982 scar indicates that partial evacuation lows, although the latter tend to be younger. (Table 1; Figs. 8 and 9) document that the of the lower hollow occurred during at least one Unconformities are also present in the subtle gravel layer was deposited very rapidly, as seen older landslide event; a 13-cm-thick lens of col- hollows, indicated by multiple dates from two at the Hole, although deposition of the gravels in luvium derived from the B horizon upslope sites. At the Arbutus landslide (Fig. 5), two dif- the Chockstone basin occurred 3,000-4,000 yr (Munsell color 7.5 YR 5/6) unconformably ferent dates were obtained from a single cross later. No unconformity is apparent in this sec- overlies a truncated BA horizon (10 YR section (Fig. 11). A date of 21,000 ± 500 yr B.P. tion, and the gravel content progressively in- 3/2-3/3) (Fig. 8B). From the soil stratigraphy, from large charcoal fragments within coarse creases with depth, as seen at the Hole. Two an estimated 1.2 m of the upper soil was eroded basal gravel on the side of the hollow contrasts recent landslides upslope from the headscarp during this pre-1982 landslide, followed by 0.3 with a date of 14,280 + 450 yr B.P. from (Fig. 8A) thus left no obvious stratigraphic m of deposition prior to 1982. smaller, rounded charcoal fragments within record. Two dates have been obtained from a 3.6-m stratified, finer-textured basal gravels in the axis. A basal date of 7650 yr B.P. from the lower section in an additional distinct hollow on San These dates record erosional events separated by Chockstone scar documents deposition begin- Pedro Ridge. A basal date of 10,080 yr B.P. about 5,700-7,700 yr. At the Lindenwood ning significantly later than at the headscarp sec- from Cascade landslide is similar to those of the landslide (Fig. 5), basal dates from two separate

contour interval 1 meter Figure 6. (Continued). arbitrary datum

cross sections were obtained. A date of 18,995 ± 1140 yr B.P. from basal gravels at the downslope end of the scar contrasts with a 13,350 ± 600 yr B.P. date from finer-textured, matrix- supported basal colluvium 15 m upslope at a mid-scar location (Fig. 11), suggesting an additional late Pleistocene unconformity. No evidence of unconformities younger than the latest Pleistocene dates was seen at the Arbutus or Lindenwood sites, or at the other dated sites in subtle hollows. Older colluvium is also present in some subtle hollows on San Pedro Ridge, as indicated by the degree of soil profile development. Colluvial deposits in the 1 -26 ka age range have brownish colors (Munsell hue 10 YR or 7.5 YR), whereas greater ages in other deposits are shown by their redder hues (5 YR) and higher clay contents. A basal date of 29,150 yr B.P. (Table 1) was obtained from one deposit with 5 YR hues in a subtle, low-gradient hollow along North San Pedro Road (Fig. 1), about 2.5 km northeast of the San Pedro Ridge sites. In some cases, the older deposits occur unconformably beneath Figure 7. (A) Longitudinal profile of the Hole, showing pre-failure ground surface, bedrock younger colluvium, as was observed in a small surface, and locations of dated charcoal samples. (B) Gravel content in headscarp of the Hole.

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/102/7/969/3381052/i0016-7606-102-7-969.pdf by guest on 24 September 2021 ridgecrest A nose hollow axis channel flow lines dirt road radiocarbon dates 4 1982 landslide scar OLDER LANDSLIDE SCARS

distinct, vegetated

î A degraded t ^

^ I vague

dark (10 YR 3/2-2/3) Chockstone Landslide Island Landslide Mosquito Landslide upper soil, with matrix- ß headscarp lower scar headscarp "island" headscarp supported clasts [A, AB, BA horizons] unconformity gravelly layer 1.7 ka - brown (7.5 YR 4/4-5/6) - N lower colluvium, with matrix-supported clasts 1 - 1 - unconformity 1 - 1 - ; Jna-'/I 11.7 ka //TlUt [B horizon] CD O 21.2 ka &&&1 — • CO ' \ > , predominantly openwork, •e \ \ " ' \ N N , clast-supported gravels / . > X S G O • <1 1 co 2 • 2- 2 H I / '-" " ' with yellowish brown TD / « ' C 7.7 ka (10 YR 4/4-5/6) matrix 3 Trhrfi 9.4 ka 0 25.6 ka rrmfn 5 sand lens 3 1 O . ^o-q. . c= scarps are reduced to rough slope breaks. (B) Simplified sections through colluvial deposits, showing 5- CV •.

Chockstone Figure 9. Longitudinal profile of Chockstone basin, showing % Landslide 1982 landslide scar and locations of dated charcoal samples. Thickness of colluvium determined from exposures in scar, soil pits, and seismic lines. Cross sections at sample locations are Bedrock surface also shown and illustrate the variable shape of the deposi- 20 m tional zone. The upper cross section is in a steep-sided sandstone trough, with the colluvium consisting of angular openwork gravel; relatively high deposi- 30 m i •7650 +/- 290 BP tional rates in this trough may have produced the convex slope landslide scar in a subtle hollow 20 m west of seen in the longitudinal profile. Chockstone landslide (Fig. 8A), recording incomplete evacuation in a previous landslide. In other cases, the older colluvium occupies buried one such deposit. Similar deposits with no topo- bedrock hollows that have no topographic ex- graphic expression are also present in the Red- pression; a small 1982 landslide scar on a side wood Creek basin of northwest California County. The occurrence of these old colluvial slope 10 m east of the Hole (Fig. 6B) exposes (Marron, 1982) and on Bolinas Ridge in Marin deposits suggests relatively great stability of some sites.

Other Marin County Sites

Two additional dated sites have similar bedrock and vegetation to the San Pedro Ridge

Arbutus Landslide

21,000 +/- 500

14,280 +/- 450 BP

13,350 +/- 600 BP

Lindenwood Landslide

Franciscan 18,995 +/- 1140 BP sandstone

Figure 11. Cross sections across Arbutus and Lindenwood landslide scars, showing locations of dated charcoal samples, bedrock- Figure 10. Photograph of coarse, angular, openwork gravel at colluvium boundary, and approximate pre-1982 ground surface headscarp of Chockstone landslide, dated at ca. 9.6-9.8 ka. (dashed lines).

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/102/7/969/3381052/i0016-7606-102-7-969.pdf by guest on 24 September 2021 978 RENEAU AND OTHERS

sites, located in the Pike County Gulch basin on vium prior to failure; an isopach map of the Third Valley 1, the steep upper hollow contains Bolinas Ridge and near Madrone Park Circle in deposit is presented in Wilson and others a large 1982 landslide scar, and extensive gully- Mill Valley (Fig. 1). The Pike County Gulch site (1989). A distinctive horizon that contains ing occurred in 1982 in the lower-gradient swale is a large (750 m3) 1982 landslide scar and has a small, scattered fragments of charcoal and downslope. A basal date of 10,560 yr B.P. was relatively large drainage area of about 7,300 m2 common reddish clasts occurs within the de- obtained from the upper scar, and a basal date of (Table 2). The colluvium is very loose and grav- posit. The horizon also generally contains 9130 yr B.P. was obtained from the gully 125 m elly, containing a high percentage of angular coarser gravel than the lower, more yellowish downslope (Table 1). The gully exposure is lo- sandstone clasts, and has yielded a relatively colluvium. The charcoal-bearing layer overlies cated farther downslope than the other dated young basal date of 6260 + 250 yr B.P. (Table an unconformity in the deposit and postdates a hollows and has a much larger drainage area of 1). Additional exposures downslope in the major erosional event in this hollow. Whereas in about 21,000 m2 (Table 2). No stratigraphic debris-flow path reveal a basal unit containing the upper scar it occurs 2-3 m above bedrock, in breaks suggestive of unconformities were seen at subrounded clasts overlain by colluvium with the lower scar the horizon rests on bedrock in either location. At Third Valley 2, another large angular gravel, recording a transition from flu- the axis and drapes over the lower colluvium on 1982 landslide scar, two dates have been ob- vial erosion to colluvial deposition as the deposit the side (Fig. 12); gravel layers in the lower tained from a single stratigraphic section. Ages extended downslope. colluvium are truncated by the upper unit. Two of 10,580 yr B.P. for the lower colluvium and The Madrone Park Circle site is a 1982 land- radiocarbon dates obtained from the basal part 10,090 yr B.P. 1.1m higher in the section (Table slide scar in a subtle forested hollow. Present of the upper horizon in different parts of the scar 1) document rapid initial deposition of collu- topographic convergence is very low (Table 2), are essentially the same at ca. 12.3 ka (Table 1), vium. The nearly identical basal ages from Third and the topographic expression of this hollow approximating the age of the earlier erosional Valley 1 and 2 suggest similar histories for these has been nearly erased; a topographic map of the event. two hollows. hillslope is presented in Reneau and Dietrich As a comparison with Lone Tree Creek 1, a (1987a). A date of 10,220 yr B.P. was obtained single date was obtained from a pre-1943 land- Berkeley Hills from a depth of 2.0 m in a deposit reaching at slide scar in the next hollow to the south (Lone least 3.5 m in total thickness, indicating that Tree Creek 2). Maximum thickness here was Radiocarbon dates have been obtained from most of the deposition here has occurred during about 4-5 m before failure, and a date of 9960 ± three hollows in the Berkeley Hills. Reneau and the Holocene. 100 yr B.P. was obtained from a depth of 2.65 others (1986) reported a date of 11,875 yr B.P. Two adjacent grassland hollows underlain by m (Table 1). Although no basal age could be for a low-gradient site in Claremont Canyon Franciscan sandstone have been studied at the obtained here, the available date suggests a pe- (Table 1; Fig. 1). The dated exposure is at the head of Lone Tree Creek on the south slope of riod of accumulation similar to that at Lone lower end of a hollow where it merges with a Mount Tamalpais (Fig. 1). Lone Tree Creek 1 is Tree Creek 1. broad colluvial apron above the main valley a large 1974 landslide scar with extensive Two additional dated sites in Marin County floor and is atypical of the dated deposits. Depo- exposures of colluvium (Figs. 3 and 12). The are in Third Valley on Inverness Ridge (Fig. 1), sition here may have been associated with either deposit reached a maximum thickness of about within a mixed bishop pine-hardwood forest the hollow upslope or the valley fill downslope. 3 5.5 m and contained about 4,000 m of collu- developed on weathered granitic bedrock. At Additional dates have been obtained from

Lone Tree Creek #1

base of 12,300 yr B.P. upper scar horizon

Figure 12. Cross sections across Lone Tree Franciscan Creek 1 landslide scar, sandstone showing approximate pre-1974 ground surface (dashed lines), bedrock- colluvium boundary, and base of location of dated horizon. 12,300 yr B.P. lower scar horizon 4 m _i i Franciscan sandstone

two large deposits near the crest of the Berkeley Ridge suggest late Pleistocene landslides in each tral California hollows record the burial of Pleis- Hills to the southeast (Fig. 1). Sites below basin separated by roughly 4,000-8,000 yr. An tocene stream channels by colluvium following Grizzly Peak Boulevard and along Skyline additional late Pleistocene landslide is probably the change to a drier climate. The postulated Boulevard are both on chert of the Monterey recorded by the unconformity at Lone Tree lack of colluvial deposition in the wetter periods Group; present vegetation includes introduced Creek 1, and Holocene unconformities are pres- of the Pleistocene presumably reflected much eucalyptus and pine, whereas the native vegeta- ent in the Mosquito and Island basins on San greater runoff that was capable of transporting tion probably consisted of grassland or north Pedro Ridge. the colluvial debris shed from adjacent hill- coastal scrub. Both deposits are very gravelly, slopes. This hypothesis implies a synchroneity of with the lower sections dominated by stratified, GENERALIZED MODELS OF basal ages that would correspond to the period angular, openwork gravels. Dates from the DEPOSITION AND EROSION of changing climate. Pollen analyses from Clear lower parts of both deposits are about 24 ka IN HOLLOWS Lake, 110 km north, show that a change toward (Table 1), suggesting similar histories for many species indicating drier conditions began ca. 13 deposits on this ridge crest. The history of hollows in the central Califor- ka and was generally complete by ca. 10 ka nia Coast Ranges appears to be complex, and (Adam, 1988). Although the latest period of Summary none of the general conceptual models discussed deposition in many hollows did begin during in the Introduction is adequate to characterize and immediately after the period 13-10 ka Study of the 20 hollows discussed above re- fully the history of these sites. Evidence pertain- (Fig. 13), deposition was not restricted to this veals significant variations in stratigraphy and ing to each of these models is discussed below. climatic transition and instead occurred under a chronology, documenting diverse depositional range of climates. Older basal ages range from histories. Basal ages range from 1 to 29 ka, Long-Term Stability ca. 29 to 14 ka, extending from the last intersta- although about half are between 9 and 14 ka dial until after the peak of the late Wisconsinan (Fig. 13). Rapid initial deposition of coarse Long-term stability of some hollows is re- glaciation, and younger ages of 1-8 ka docu- openwork gravels occurred at the Hole and at corded by the presence of old colluvial deposits ment deposition beginning well after the Chockstone landslide, and of finer-textured col- mantling bedrock hollows with no topographic Pleistocene-Holocene transition (Fig. 13). luvium at Third Valley 2. Late Holocene deposi- expression. Subtle topographic noses formed on Although a model of climatically induced tion is recorded by dates from upper levels at the colluvial deposits have been observed at two deposition of colluvium is probably inadequate Island and Cascade sites. Colluvium in down- locations in central California, on Bolinas Ridge for most dated sites, it may apply to the downslope parts of hollows is generally younger than in Marin County and in the Berkeley Hills, al- slope parts of some hollows reflecting relatively in the upslope parts, shown by paired dates at though such inversion of topography is atypical small changes in the location of channel heads. the Chockstone, Island, and Third Valley 1 of hollows in this region. No evidence was seen Reneau and others (1986) and Montgomery and landslide scars, and by the stratigraphy at Lone for the relief inversion mechanisms proposed by Dietrich (1988) have suggested that the location Tree Creek 1, although at the Hole, deposition Bryan (1940) and Mills (1981), involving depo- of channel heads and the length of hollows in in different parts began synchronously, and at sition of bouldery debris on less resistant part reflect the local climate and are thus sensi- the Lindenwood site, the younger of two dates bedrock. tive to regional climate changes. For example, was obtained upslope. Incomplete evacuation of for slope gradients of 20°-35°, channel heads colluvium in earlier erosional events is indicated Climatic Control on Colluvial Deposition are typically about 10-20 m farther upslope in by probable unconformities in five of the 20 the wetter climate of coastal Oregon, where an- dated hollows. Radiocarbon dates from the Ar- Schlocker (1974) and Shlemon and others nual rainfall is about 1,500 mm/yr, than in cen- butus and Lindenwood basins on San Pedro (1987) proposed that colluvial deposits in central coastal California, with about 800 mm/yr of rain (D. R. Montgomery, unpub. data). There is, however, much variation in the location of channel heads in each area, associated with recent landsliding, and there is also much overlap between the two populations (Montgomery and Dietrich, 1988). Changes in channel-head location caused by recent landslides in many cases exceed 20 m and are probably more important in most steep hollows than those induced by regional climatic changes.

Basal Radiocarbon Age (ka) Cyclic Instability Figure 13. Histogram of basal radiocarbon dates from hollows in San Francisco Bay area, with each date recording the approximate initiation of deposition at different locations. Be- Evidence for a recurring cycle of accumula- cause most hollows with dates from two longitudinal positions yield variable ages, such paired tion and evacuation of colluvium in hollows is dates are included as separate sites. Several dates are below recognized unconformities, and provided by the presence of probable uncon- unconformities may also be present at other sites. These dates, therefore, do not necessarily formities in five of the 20 deposits reported in record the last evacuation event at each site. this study, recording partial erosion of colluvium

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/102/7/969/3381052/i0016-7606-102-7-969.pdf by guest on 24 September 2021 980 RENEAU AND OTHERS

in both Pleistocene and Holocene landslides. ning the period of déglaciation. Late Wiscon- runoff to drain, such as through coarse basal The 1-8 ka basal ages at Hungry Vulture swale, sinan glaciation in California began ca. 26 ka gravel layers or into highly fractured bedrock. Pike County Gulch, and Chockstone landslide (Atwater and others, 1986), and deposition be- The stable colluvial deposits require exceptional may also record Holocene landslides. As men- ginning during the early stages of this glaciation conditions to fail, such as unusually high- tioned earlier, Marron (1982, 1985) similarly is documented by ca. 24 ka dates from two de- intensity rainfall that results in greatly elevated reported the presence of unconformities in col- posits in the Berkeley Hills and a ca. 26 ka date pore pressures, repeated large storms that result luvial deposits in northwest California, and from San Pedro Ridge. One dated deposit in in progressive upslope failures from landslide Shlemon and others (1987) reported evidence Marin County documents accumulation that scarps or channel heads, or secondary failures for multiple Holocene failures in a hollow at began ca. 29 ka, during the previous interstadial. triggered from debris flows originating upslope. Pacifica, California. The stratigraphic and chron- A simple climatically controlled model of filling For example, the four largest landslide scars in ologic evidence for recurring landslides in both of Pleistocene stream channels or gullies with the study area of Reneau and Dietrich (1987b) the late Pleistocene and the Holocene, combined colluvium following the change to Holocene on San Pedro Ridge are spatially clustered, sug- with the well-documented occurrence of recent climate, as proposed by Schlocker (1974) and gesting failure of relatively stable deposits asso- landslides in hollows, supports a model of cyclic Shlemon and others (1987) for sites in the San ciated with a local rainfall cell of exceptional accumulation and evacuation of colluvium from Francisco Bay area, is largely inconsistent with intensity. Periods with increased storm intensity hollows. A model of climatically controlled these data. Such a model, however, may apply and frequency over broad regions could cause deposition, however, as discussed above, cannot to the downslope parts of some hollows, reflect- widespread evacuation of the more stable depos- be ruled out for the downslope parts of some ing climatically induced changes in channel- its, and such a period at ca. 14-9 ka is suggested hollows. In addition, the available basal ages head location. by the clustering of basal ages from hollows. from hollows suggest that landsliding has not A more important climatic influence may be Extended periods of widespread landsliding, by been uniformly distributed through time, but in- present in the timing of landslides. The common discharging exceptional volumes of colluvium stead has been influenced by climatic changes. occurrence of basal ages at ca. 14-9 ka, at sites from hillsides, may also contribute to aggrada- The clustering of ages at ca. 14-9 ka (Fig. 13) with varying topography and vegetation, sug- tion along streams. for sites with varying topography, including gests more widespread landsliding in the latest Support for a connection between widespread both relatively steep and gentle slope gradients Pleistocene than in the Holocene as proposed landsliding in hollows and aggradation along and large and small drainage areas, varying earlier by Reneau and others (1986). It is un- streams is present in the stratigraphy of two vegetation, and at recognized unconformities, likely that such a clustering reflects an inherent drainages on the Point Reyes Peninsula (Rypins argues for widespread hillslope instability during time scale of failure, owing to widely varying and others, 1989). Aggradation in these valleys the Pleistocene-Holocene transition. Landsliding site conditions. Specifically, stability analyses began ca. 12 ka, associated with thick debris- in hollows, however, was clearly not restricted suggest that root strength along the margins of a flow deposits, and slowed considerably by ca. 10 to this period, and all hollows were not "cleaned potential failure imposes constraints on landslide ka. Periods of stream aggradation in this region out" at this time. size and that each colluvial deposit must reach a have also occurred during the Holocene (Pape, critical size before failure can occur (Reneau and 1978; Haible, 1980), although the relative DISCUSSION Dietrich, 1987b). This critical size should vary timing in different basins and the magnitude as a function of vegetation, slope gradient, soil of these aggradation events has not been The chronologic data and stratigraphic obser- texture, and local hydrology. In addition, depo- determined. vations presented in this study are generally con- sitional rates of colluvium vary greatly between Evidence for major stream aggradation dur- sistent with a cycle of alternating accumulation sites (Reneau, 1988), and such a critical deposit ing the Pleistocene-Holocene transition has also and evacuation of colluvium in California Coast size should thus be reached at very different been cited from other nonglaciated parts of Cali- Range hollows, as inferred from modern proc- times in different hollows. The clustering of fornia (Marchand, 1977; Swan and others, esses. Unconformities in the deposits reflect basal radiocarbon ages thus seems to record cli- 1977; Ponti and others, 1980; Weldon, 1983, partial erosion of colluvium in both Pleistocene matic fluctuations that are superimposed on a 1986; Lettis, 1985; Ponti, 1985; Wells and oth- and Holocene events, as is often caused by mod- cycle of alternating accumulation and evacua- ers, 1987, 1989). The timing of stream aggrada- ern landslides (Reneau and Dietrich, 1987b). tion of colluvium, producing regional fluctua- tion is well constrained by Weldon (1986) for Longitudinal variations in age within a hollow tions in the discharge of debris from hillslopes. Cajon Creek in the Transverse Ranges, where are also consistent with landslides that involve We propose that failures on steep slopes at aggradation extended from ca. 17 to 6 ka and only part of a hollow in each event. The obser- sites with typical loamy, matrix-supported col- culminated at ca. 14-12 ka. This fill event at vation that colluvium is generally younger luvium are generally scattered through time, Cajon Creek was the largest in the past 50,000 downslope may be partially explained by an in- with the timing of failure controlled in part by yr (Weldon, 1986). Farther east in the eastern creasing frequency of debris flows, and conse- the depositional rate of colluvium and the Mojave Desert, Wells and others (1987, 1989) quently more frequent erosional events, as the strength provided by local vegetation (for exam- used dated shorelines of pluvial Lake Mojave to upslope drainage area increases in a hollow ple, Reneau and Dietrich, 1987b). In addition to constrain fan aggradation between ca. 12.5 and (Reneau and Dietrich, 1987a). these typical steep sites, considerable additional 9.5 ka. The apparent similarity in timing of hill- Basal radiocarbon dates from hollows show a colluvium is stored in relatively stable deposits. slope erosion and stream aggradation over di- broad range in age. The latest period of deposi- Deposits that are comparatively stable include verse parts of California suggests that there may tion at most dated sites began between 16 and 9 sites with low slope gradients and sites where have been a common meteorologic control. This ka, particularly from 14-9 ka (Fig. 13), span- local hydrologie conditions allow subsurface is at a time when atmospheric general circula-

tion models indicate that major changes in the from the Great Basin, Benson and Thompson man is gratefully appreciated. The manuscript location and strength of the jet stream occurred (1987) suggested greater climatic variability dur- in its various stages has benefited from reviews over western North America (for example, ing the period 13-10 ka, with major high stands by G. Curtis, R. Dorn, C. Harrington, H. Kelsey, Kutzbach, 1987). in several lakes followed by rapid retreat. In D. Montgomery, and T. Oberlander. Support Modern landsliding in colluvium occurs dur- addition, Spaulding and Graumlich (1986) pro- for this project was provided by National Sci- ing periods of relatively long-duration high- posed that vegetation changes in the southwest- ence Foundation Grants EAR84-16775 and intensity rainfall (Caine, 1980; Cannon and ern deserts reflect pronounced meridional flow EAR84-51175, a grant from the Pacific Gas Ellen, 1985, 1988; Neary and Swift, 1987; from 12-8 ka that contrasted with zonal flow and Electric Company, and a Department of Wieczorek, 1987; Wieczorek and Sarmiento, during full-glacial time. The correlation between Energy postdoctoral fellowship. 1988). In coastal California, these storms result the basal radiocarbon ages from hollows and from unusual meteorological conditions involv- these diverse records of climatic change supports ing the formation of a high-latitude high- the inference that hillslope erosion in central REFERENCES CITED pressure region, or "block" near the Gulf of California hollows was affected by climatic Adam, D. P., 1988, Pollen zonation and proposed informal climatic units for Clear Lake, California, cores CI.-73-4 and CL-73-7, in Sims, J. D., ed.. Alaska, a weakening and southward shift of the changes at the end of the Pleistocene. Late Quaternary climate, tectonism, and sedimentation in Clear Lake, northern California Coast Ranges: Geological Society of America Spe- semi-permanent North Pacific high, and the cial Paper 214, p. 63 80. Adam, D. P., Byrne, R., and Luther, E., 1981, A late Pleistocene and Holocene formation of a low-pressure region along the CONCLUSIONS pollen record from Laguna de las Trancas, northern coastal Santa Cruz California coast. These events allow the conver- County, California: Madrono, v. 28, p. 255 272. Atwater, B. F., Adam, D. P., Bradbury, J. P., Forester, R. M„ Mark, R. K.. gence of cold polar air masses and warm, moist The detailed study of colluvial deposits pre- Lettis, W. R„ Fisher, G. R., Gobalet, K. W„ and Robinson, S. W„ 1986, A fan dam for Tulare Lake, California, and implications for the subtropical air masses that have been responsi- sented herein documents a complex history of Wisconsin glacial history of the Sierra Nevada: Geological Society of America Bulletin, v. 97, p. 97-109. ble for the periods of heaviest rain (Weaver, erosion and deposition in California Coast Benda, L., and Dunne, T., 1987, Sediment routing by debris flow, in Beschla, 1962; Cayan and Namias, 1982; Monteverdi, R. L., Blinn, T„ Grant, G. E., Swanson, F. J., and Ice, G. G., eds., Range hollows. A cycle of alternating accumula- Erosion and sedimentation in the Pacific Rim: International Association 1982; Namias, 1982; Hirschboeck, 1987). Ex- tion and evacuation of colluvium probably of Hydrological Sciences Publication 165, p. 213 223. Benson, L., and Thompson. R. S„ 1987, The physical record of lakes in the amples include the storms of 21-24 December characterizes most sites, with landslides occur- Great Basin, in Ruddiman, W. F.. and Wright, H. E„ Jr.. eds., North 1955, 21-23 December 1964, 18-27 January America and adjacent oceans during the last deglaciation: Boulder. ring during the late Pleistocene as well as the Colorado, Geological Society of America, The Geology of North Amer- 1969, and 3-5 January 1982, all periods of ex- ica, y. K-3, p. 241 260. Holocene. This cycle of erosion and deposition Berger, R., and Libby, W., 1966, UCLA radiocarbon dates V: Radiocarbon, ceptional landslide activity and major flooding has apparently been influenced by regional cli- v. 8, p. 467 497.

in parts of California. Similar blocking events matic fluctuations, with more widespread land- Blake, M. C., Jr.. Barlow, J. A., Frizzell, V. A, Jr., Schiocker, J., Sorg, D„ have also been linked to historic and late Holo- Wentworth, C. M., and Wright, R. H., 1974, Preliminary geologic map sliding occurring at ca. 14-9 ka. The period of of Marin and San Francisco Counties and parts of Alameda, Contra cene lakes in the Mojave Desert of southeastern increased landsliding may have been caused by Costa, and Sonoma Counties, California: U.S. Geological Survey Mis- cellaneous Field Studies Map MF-574. California associated with heavy rainfall in the extended periods of meridional flow in the Bryan, K., 1940, Gully gravure A method of slope retreat: Journal of Gco- San Bernardino Mountains (Enzel and others, morphology, v. 3, p. 331 - 344. upper atmosphere during the terminal Wiscon- Caine, N„ 1980, The rainfall intensity-duration control of shallow landslides 1989; Wells and others, 1989). In addition, sinan deglaciation that led to frequent blocking and debris flows: Geograliska Annaler, v. 62A, p. 23 27. Cannon, S. H., and Ellen, S. D.. 1985. Rainfall conditions for abundant debris Wells and others (1989) proposed that Lake events over the North Pacific Ocean and more avalanches in the San Francisco Bay region, California: California Mojave high stands at ca. 13.7-11.7 ka record Geology, v. 38, p. 267 272. frequent intense storms over California. The ac- 1988, Rainfall that resulted in abundant debris-ilow activity during the more frequent flood events in the San Bernar- celerated discharge of colluvium from hillslopes storm, in Ellen. S. D.. and Wieczorek, G. F., eds., Landslides, floods, and marine effects of the storm of January 3 5, 1982, in the San dino Mountains, suggesting a higher frequency may in turn have contributed to stream aggrada- Francisco Bay region. California: U.S. Geological Survey Professional Paper 1434, p. 27-33. of such blocking events in the latest Pleistocene. tion in diverse parts of California. Hollows, Cayan, D. R., and Namias, J., 1982, Anomalous climatic conditions leading to flood producing rains in the San Francisco Bay area: Paper presented at The blocking events occur as a result of ex- which are the primary storage site for colluvium conference on debris flows, landslides, and floods in the San Francisco on many hillslopes, may provide a critical link Bay region, January 1982, August 23 26, 1982, Stanford University, treme meridional flow in the upper atmosphere California. (for example, Lamb, 1972; Hirschboeck, 1987), between climatic fluctuations, sediment produc- Cotton. C. A., and TePunga. M. T.. 1955, Fossil gullies in the Wellington tion, and stream aggradation in many areas. landscape: New Zealand Geographer, v. 2, p. 72-75. and by analogy with modern météorologie con- Dietrich, W. E„ and Dorn. R„ 1984, Significance of thick deposits of colluvium on hillslopes: A case study involving the use of pollen analysis in the ditions, the accelerated hillslope erosion pro- coastal mountains of northern California: Journal of Geology, v. 92. posed for ca. 14-9 ka may reflect extended ACKNOWLEDGMENTS p. 147 158. Dietrich. W. E., and Dunne, T., 1978, Sediment budget for a small catchment in periods of meridional flow. Historic periods mountainous terrain: Zeitschrift fur Geomorphologie, Supplementband characterized by frequent blocking display 29. p. 191-206. The authors thank the California Department Dietrich, W. E., Dunne, T.. Humphrey, N. F., and Reid. L. M.. 1982. Construc- greater climatic variability than do periods of of Parks and Recreation for access to China tion of sediment budgets for drainage basins, in Swanson, F. J., Janda, R. J„ Dunne, T., and Swanston. D. N., eds., Sediment budgets predominantly zonal atmospheric flow (for ex- Camp and Mt. Tamalpais State Parks, the City and routing in forested drainage basins: U.S. Department of Agriculture Forest Service General Technical Report PNW-141, p. 5 23. ample, Lamb, 1977), and the existence of more of San Rafael for access to Barbier Memorial Dietrich. W. E., Wilson, C. J., and Reneau, S. L„ 1986, Hollows, landslides, pronounced meridional flow during the terminal and colluvium in soil-mantled landscapes, in Abrahams, A. D.. ed.. Park, and Oscar Brostrom and Jack and Patty Hillslope processes: Boston, Massachusetts, Allen & Unwin, p. 361 388. Wisconsinan déglaciation is supported by di- Wright for access to sites near Inverness. O.K. Ellen, S. D., 1988, Description and mechanics of soil slips/debris flows in the storm, in Ellen, S. D„ and Wieczorek, G. F., eds., Landslides, floods, verse paleoenvironmental records that indicate Davis generously provided an examination of and marine effects of the storm of January 3 5, 1982, in the San Francisco Bay region, California: U.S. Geological Survey Professional major climatic variability at this time. For ex- charcoal fragments from San Pedro Ridge. Field Paper 1434, p. 63 112. ample, sea surface-temperature reconstructions Ellen, S. D.. Cannon. S. H., and Reneau, S. L„ 1988. Distribution of debris assistance by D. Chambers, L. Collins, L. Den- flows in Marin County, in Ellen, S. D., and Wieczorek, G. F., eds.. in the Pacific Ocean from cores off the Oregon gler, E. Hughes, P. Jordan, A. Lagios, J. Mani- Landslides, floods, and marine effects of the storm of January 3- 5, 1982, in the San Francisco Bay region, California: U.S. Geological coast suggest much greater variability during the takos, M. Mauldon, D. Montgomery, G. Muller, Survey Professional Paper 1434, p. 113-131. period ca. 15-10.5 ka than during the Holocene Enzel, Y., Cayan, D. R„ Anderson, R. Y„ and Wells, S. G„ 1989. Atmospheric A. Niederhov, R. Olshansky, S. Raugust, circulation during Holocene lake stands in the Mojave Desert: Evidence (Moore, 1973). On the basis of lake-level data K. Rich, J. Sturman, C. Wilson and M. Young- of regional climate change: Nature, v. 341, p. 44-47.

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/102/7/969/3381052/i0016-7606-102-7-969.pdf by guest on 24 September 2021 982 RENEAU AND OTHERS

Hack, J. T., 1965. Geomorphology of the Shenandoah Valley, Virginia and Namias, J., 1982, Meteorologic and oceanographic conditions for the en- storm of January 3-5, 1982, in the San Francisco Bay region, Califor- West Virginia, and origin of the residual ore deposits: U.S. Geological hancement or suppression of winter rains over California, in Storms, nia, U.S. Geological Survey Professional Paper 1434, p. 185-203. Survey Professional Paper 484, 84 p. floods, and debris flows in southern California and Arizona, 1978 and Smith, T. C„ and Hart, E. W., 1982, Landslides and related storm damage, Hack, J. T.. and Goodlett, J. C., I960, Geomorphology and forest ecology of a 1980, Proceedings of Symposium, September 1980: Washington, D C, January 1982, San Francisco Bay region: California Geology, v. 35, mountain region in the central Appalachians: U.S. Geological Survey National Academy Press, p. 25-41. p. 139-152. Professional Paper 347,66 p. Neary, D. G., and Swift, L. W„ Jr.. 1987, Rainfall thresholds for triggering a Spaulding, W. G„ and Graumlich, L. J., 1986, The last pluvial climatic episode Haible, W. W., 1980, Holocene profile changes along a California coastal debris avalanching event in the southern Appalachian Mountains, in of southwestern North America: Nature, v. 320, p. 441-444. stream: Earth Surface Processes, v. 5, p. 249-264. Costa, J. E., and Wieczorek, G. F., eds., Debris flows/avalanches: Swan, F. H.. Ill, Hanson, K. L., and Page, W. D., 1977, Landscape evolution Hirschboeck, K. K., 1987, Catastrophic flooding and atmospheric circulation Process, recognition, and mitigation: Geological Society of America and soil formation in the western Sierra Nevada foothills, California, in anomalies, in Mayer, L, and Nash. D., eds„ Catastrophic flooding: Reviews in Engineering Geology, v. 7, p. 81-92. Singer, M. J., ed., Soil development, geomorphology, and Cenozoic Boston, Massachusetts, Allen & Unwin, p. 23-56. Okunishi, K., and lida, T., 1981, Evolution of hillslopes including landslides: history of the northeastern San Joaquin Valley and adjacent areas, Howard. T. R., Baldwin. J. E.. III. and Donley, H. F„ 1988. Landslides in Japanese Geomorphological Union Transactions, v. 2, p. 291-300. California (American Society for Agronomy, Soil Science Society of Pacifica, California, caused by the storm, in Ellen, S. D., and Wieczorek, Pape, D. A., 1978, Terraced alluvial fills in Contra Costa County, California America, and Geological Society of America, joint field session guide- G. F„ eds.. Landslides, floods, and marine effects of the storm of January [M.S. thesis]: Berkeley, California, University of California, 47 p. book): Davis, California, University of California Press, p. 300-311. 3-5, 1982, in the San Francisco Bay region, California: U.S. Geological Pierson, T. C., 1977, Factors controlling debris-flow initiation on forested Tsukamoto, Y., Ohta, T„ and Noguchi, H., 1982, Hydrological and geo- Survey Professional Paper 1434, p. 163-183. hillslopes in the Oregon Coast Range [Ph.D. thesis]: Seattle, Washing- morphological studies of debris slides on forested hillslopes in Japan, in Johnson, D. L., 1977, The late Quaternary climate of coastal California: Evi- ton, University of Washington, 166 p. Walling, D. E., ed., Recent developments in the explanation and predic- dence for an Ice Age refugium: Quaternary Research, v. 8. p. 154-179. Ponti, D. J., 1985, The Quaternary alluvial sequence of the Antelope Valley, tion of erosion and sediment yield: International Association of Hydro- Kutzbach, J. E., 1987, Model simulations of the climatic patterns during the California, in Weide, D. L., ed.. Soils and Quaternary geology of the logical Sciences Publication 137, p. 89-98. déglaciation of North America, in Ruddiman, W. F., and Wright, southwestern United States: Geological Society of America Special Weaver, R. L„ 1962, Meteorology of hydro!ogically critical storms in Califor- H. E., Jr.. eds.. North America and adjacent oceans during the last dé- Paper 203, p. 79-%. nia: U.S. Weather Bureau Hydrometeorological Report 37, 207 p. glaciation: Boulder, Colorado, Geological Society of America, The Ponti, D. J., Burke, D. B„ Marchand. D. E., Atwater, B. F„ and Helley, E. J., Weldon, R. J., 1983, Climatic control for the formation of terraces in Cajon Geology of North America, v. K-3. p. 447-462. 1980, Evidence for correlation and climatic control of sequences of Creek, southern California: Geological Society of America Abstracts Lamb. H. H.. 1972, Climate: Past, present, and future: Volume 1. Fundamentals late Quaternary alluvium in California: Geological Society of America with Programs, v. 15. p. 429. and climate now: London. United Kingdom, Methuen, 613 p. Abstracts with Programs, v. 12, p. 501. 1986, The late Cenozoic geology of the Cajon Pass: Implications for 1977, Climate: Past, present, and future: Volume 2. Climatic history and Radbruch, D. R., 1969, Area! and engineering geology of the Oakland East tectonics and sedimentation along the San Andreas fault [Ph.D. thesis]: the future: London. United Kingdom. Methuen, 835 p. quadrangle, California: U.S. Geological Survey Map GQ-769, 15 p. Pasadena, California, California Institute of Technology, 400 p. Lehre, A. K„ 1981. Sediment budget of a small California Coast Range drain- Rantz, S. E., 1971, Mean annual precipitation and precipitation depth- Wells. S. G„ McFadden, L. D„ and Dohrenwend, J. C., 1987, Influence of late age basin near San Francisco, in Davies, T.R.H., and Pearce. A. J., eds., duration-frequency data for the San Francisco Bay region, California: Quaternary climatic changes on geomorphic and pedogenic processes Erosion and sediment transport in Pacific Rim steeplands: international U.S. Geological Survey-HUD San Francisco Bay Region Environment on a desert piedmont, eastern Mojave Desert, California: Quaternary Association of Hydrological Sciences Publication 132. p. 123-139. and Resources Planning Study Basic Data Contribution 32, 23 p. Research, v. 27, p. 130-146. 1982. Sediment mobilization and production from a small mountain Reneau, S. L., 1988, Depositional and erosional history of hollows: Application Wells, S. G., Anderson, R. Y„ McFadden, L. D„ Brown, W. J., Enzel, Y.. and catchment. Lone Tree Creek, Marin County, California [Ph.D. thesis]: to landslide location and frequency, long-term erosion rates, and the Miossec, J.-L., 1989, Late Quaternary paleohydrology of the eastern Berkeley, California, University of California, 375 p. effects of climatic change [Ph.D. thesis]: Berkeley, California, University Mojave River drainage, southern California: Quantitative assessment of Lettis. W. R.. 1985, Late Cenozoic stratigraphy and structure of the west of California, 328 p. the late Quaternary hydrologic cycle in large arid watersheds: Las margin of the central San Joaquin Valley, California, in Weide, D. L., Reneau, S. L., and Dietrich, W. E., 1987a, The importance of hollows in debris Cruces, New Mexico Water Resources Research Institute Report 242, ed.. Soils and Quaternary geology of the southwestern United States: flow studies: Examples from Marin County, California, in Costa, J. E„ 253 p. Geological Society of America Special Paper 203, p. 97-114. and Wieczorek, G. F., eds., Debris flows/avalanches: Process, recogni- Wieczorek. G. F„ 1987, Effect of rainfall intensity and duration on debris flows Marchand. D. E., 1977, The Cenozoic history of the San Joaquin Valley and tion, and mitigation: Geological Society of America Reviews in Engi- in central Santa Cruz Mountains, California, in Costa. J. E„ and adjacent Sierra Nevada as inferred from the geology and soils of the neering Geology, v. 7, p. 165-180. Wieczorek, G. F., eds.. Debris flows/avalanches: Process, recognition, eastern San Joaquin Valley, in Singer, M. J., ed„ Soil development, 1987b, Size and location of colluvial landslides in a steep forested and mitigation: Geological Society of America Reviews in Engineering geomorphology, and Cenozoic history of the northeastern San Joaquin landscape, in Beschta, R. L., Blinn, T., Grant, G. E., Swanson, F. J., and Geology, v. 7, p. 93-104. Valley and adjacent areas. California (American Society for Agronomy, Ice, G. G., eds.. Erosion and sedimentation in the Pacific Rim: Interna- Wieczorek, G. F., and Sarmiento, J., 1988, Rainfall, piezometric levels, and Soil Science Society of America, and Geological Society of America, tional Association of Hydrological Sciences Publication 165, p. 39-48. debris flows near La Honda, California, in storms between 1975 and joint field session guidebook): Davis, California. University of Califor- Reneau. S. L„ Dietrich, W. E„ Dorn, R. I., Berger, C. R., and Rubin, M., 1986, 1983, in Ellen, S. D., and Wieczorek, G. F„ eds., Landslides, floods, and nia Press, p. 39-50. Geomorphic and paleoclimatic implications of latest Pleistocene radio- marine effects of the storm of January 3-5, 1982, in the San Francisco Marron, D. C.. 1982, Hillslope evolution and the genesis of coltuvium in carbon dates from colluvium-mantled hollows, California: Geology, Bay region, California: U.S. Geological Survey Professional Paper 1434, Redwood National Park, northwestern California: The use of soil de- v. 14, p. 655-658. p. 43-62. velopment in their analysis [Ph.D. thesis]: Berkeley, California, Univer- Reneau. S. L.. Dietrich, W. E., Rubin, M., Donahue, D. J., and Jull, A.J.T., Wilson, C. J„ Reneau, S. L., and Dietrich, W. E., 1989, Hydrologic and sity of California, 187 p. 1989, Analysis of hillslope erosion rates using dated colluvial deposits: erosional processes in hollows, Lone Tree Creek, Marin County. Cali- 1985. Coll uvi urn in bedrock hollows on steep slopes. Redwood Creek Journal of Geology, v. 97, p. 45-63. fornia, in Brown, W. M., Ill, ed., Landslides in central California (28th drainage basin, northwestern California, in Jungerius, P. D., eds., Soils Rypins, S„ Reneau, S. L., Byrne, R., and Montgomery, D. R., 1989, Palyno- International Geological Congress field trip guidebook T381): Wash- and geomorphology: Catena Supplement 6, p. 59-68. logic and geomorphic evidence for environmental change during the ington, D.C., American Geophysical Union, p. 75-89. Mason, H. L„ 1934, Pleistocene flora of the Tomales Formation: Carnegie Pleistocene-Holocene transition at Point Reyes Peninsula, central Institution of Washington Publications, v. 415, p. 81-179. coastal California: Quaternary Research, v. 32, p. 72-87. Mills, H. H,. 1981. Boulder deposits and the retreat of mountain slopes, or Schlocker, J.. 1974, Geology of the San Francisco North quadrangle, "gully gravure" revisited: Journal of Geology, v. 89, p. 649-660. California: U.S. Geological Survey Professional Paper 782, 109 p. Monteverdi, J. P., 1982, Comparisons between three record rain-producing Shlemon, R. J., Wright, R. H., and Montgomery, D. R„ 1987, Anatomy of a storms in north-central California: Paper presented at conference on debris flow, Pacifica, California, in Costa, J. E., and Wieczorek, G. F., debris flows, landslides, and floods in the San Francisco Bay region, eds.. Debris flows/avalanches: Process, recognition, and mitigation: January 1982. August 23-26. 1982, Stanford University. California. Geological Society of America Reviews in Engineering Geology, v. 7. Montgomery, D. R., and Dietrich, W. E.. 1988. Where do channels begin?: p. 181-199. Nature, v. 336, p. 232-234. Smith, T. C„ 1988, A method for mapping relative susceptibility to debris MANUSCRIPT RECEIVED BY THE SOCIETY JULY 12, 1989 Moore. T. C., Jr., 1973, Late Pleistocene-Holocene oceanographic changes in flows, with an example from San Mateo County, in Ellen, S. D., and REVISED MANUSCRIPT RECEIVED OCTOBER 23, 1989 the northeastern Pacific: Quaternary Research, v. 3, p. 99-109. Wieczorek, G. F„ eds.. Landslides, floods, and marine effects of the MANUSCRIPT ACCEPTED OCTOBER 27, 1989

Printed in U.S.A.

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/102/7/969/3381052/i0016-7606-102-7-969.pdf by guest on 24 September 2021