Climate-change versus landslide origin of fill terraces Climate-change versus landslide origin of fill terraces in a rapidly eroding bedrock landscape: San Gabriel River, California Dirk Scherler1,2,3,†, Michael P. Lamb1, Edward J. Rhodes4,5, and Jean-Philippe Avouac1 1Division of Geological and Planetary Sciences, California Institute of Technology, 1200 East California Boulevard, Pasadena, California 91125, USA 2GFZ German Research Centre for Geosciences, 14473 Potsdam, Germany 3Institute of Geological Sciences, Freie Universität Berlin, Malteserstrasse 74-100, 12249 Berlin, Germany 4Department of Geography, University of Sheffield, Winter Street, Sheffield, S10 2TN, UK 5Department of Earth, Planetary, and Space Sciences, University of California–Los Angeles, 595 Charles Young Drive East, Los Angeles, California 90095, USA ABSTRACT lar deposits in other parts of the San Gabriel Because rates of sediment transport can be Mountains argues against a regional climatic quite variable, observational records are often Fill terraces along rivers represent the signal. Our study highlights the potential for too short to yield reliable trends. This is par- legacy of aggradation periods that are most valley aggradation by debris flows in arid ticularly true in semiarid to arid environments, commonly attributed to climate change. In bedrock landscapes downstream of landslides which are characterized by high rainfall, runoff, the North Fork of the San Gabriel River, an that occupy headwater areas. and discharge variability (e.g., Molnar et al., arid bedrock landscape in the San Gabriel 2006). In contrast, aggradational and degrada- Mountains, California, a series of prominent INTRODUCTION tional landforms provide comparatively lon- fill terraces was previously related to climate- ger records of landscape change and usually change–induced pulses of hillslope sediment The way in which landscapes respond to integrate over short-term variability. Amongst supply that temporarily and repeatedly over- climate and other environmental changes is an the most often used landforms for reconstruct- whelmed river transport capacity during the important issue in light of global climate change ing landscape history are river terraces, which Quaternary. Based on field observations, digi- and anthropogenic changes in land use (National are found all across the world. River terraces tal topographic analysis, and dating of Qua- Research Council, 2010). Most geomorphic record the geometry of previous valley floors, ternary deposits, we suggest instead that val- research in this direction has so far focused which can be used to inform about sea-level ley aggradation was spatially confined to the on soil-mantled landscapes and how changes variations (e.g., Merritts et al., 1994), or tec- North Fork San Gabriel Canyon and was a in rainfall, vegetation cover, and runoff lead to tonic rates (e.g., Lavé and Avouac, 2001; Paz- consequence of the sudden supply of uncon- changes in sediment transport on hillslopes and zaglia and Brandon, 2001). Their formation is solidated material to upstream reaches by by rivers (e.g., Knox, 1983; Blum and Törn- most often attributed to landscape-scale effects one of the largest known landslides in the San qvist, 2000). Far less inquiry has addressed of climate change (e.g., Leopold et al., 1964; Gabriel Mountains. New 10Be-derived sur- how rapidly eroding, steep bedrock landscapes Brakenridge, 1980; Knox, 1983; Weldon, 1986; face exposure ages from the landslide depos- respond to environmental changes (e.g., Riebe Bull, 1991; Porter et al., 1992; Poisson and its, previously assumed to be early to middle et al., 2001; DiBiase and Lamb, 2013; Scher- Avouac, 2004; Pazzaglia, 2013; Scherler et al., Pleistocene in age, indicate at least three ler et al., 2015). This is a shortcoming, because 2015). However, other studies have shown that Holo cene events at ca. 8–9 ka, ca. 4–5 ka, and steep hillslopes and higher mass fluxes have the terraces can form even during periods of steady ca. 0.5–1 ka. The oldest and presumably most potential for a greater and more rapid impact downcutting, from intrinsic variations in river extensive landslide predates the valley aggra- on ecosystems and societies. The San Gabriel lateral migration and meander cutoff (Davis, dation period, which is constrained by exist- Mountains in Southern California, for example, 1909; Merritts et al., 1994; Finnegan and Diet- ing 14C ages and new luminescence ages to ca. rate amongst the most rapidly uplifting moun- rich, 2011). 7–8 ka. The spatial distribution, morphology, tains in the United States. Their steep slopes and Traditionally, river terraces are subdivided and sedimentology of the river terraces are sensitivity to wildfires, flash floods, landslides, into strath and fill terraces. Whereas strath ter- consistent with deposition from far-travel- and debris flows account for imminent hazards races are cut into bedrock and typically capped ing debris flows that originated within, and to nearby urban areas (e.g., Eaton, 1935; Lavé with a thin veneer of sediment, fill terraces mined, the landslide deposits. Valley aggrada- and Burbank, 2004; Lamb et al., 2011; Kean represent remnants of valley fills that are often tion in the North Fork San Gabriel Canyon et al., 2013). An assessment of the potential several tens to hundreds of meters thick (e.g., therefore resulted from locally enhanced sedi- risks that are associated with climatic and other Merritts et al., 1994; Pazzaglia, 2013). Most ment supply that temporarily overwhelmed environmental changes requires understanding generally, valley fills form when rivers switch river transport capacity, but the lack of simi- the controls on sediment production and trans- from incision to aggradation. In uplifting land- port in these landscapes and the ways in which scapes that are far away from sea-level changes, †scherler@ gfz -potsdam .de they change with time. aggradation that is due to damming of a river by GSA Bulletin; Month/Month 2016; v. 128; no. X/X; p. 1–21; doi: 10.1130/B31356.1; 19 figures; 3 tables; Data Repository item 2016046.; published online XX Month 2016. For permission to copy, contact [email protected] Geological Society of America Bulletin, v. 1XX, no. XX/XX 1 © 2016 Geological Society of America Scherler et al. a landslide (e.g., Korup et al., 2006), or glacier aggrade and steepen its bed until shear stresses STUDY AREA AND PREVIOUS WORK (e.g., Montgomery et al., 2004; Scherler et al., at the bed become high enough to transport all 2014), can usually be identified by a distinct of its material. The San Gabriel Mountains constitute a spatial association of the fill and the dam. Valley Based on detailed field work and analyti- transpressional basement block adjacent to the fills that are more regionally distributed are in cal tools available at that time, Bull (1991) Mojave segment of the San Andreas fault (Fig. most cases interpreted as resulting from flu- suggested that fill terraces in parts of the San 1A) that consists of mostly Precambrian igne- vial aggradation in response to climate change Gabriel Mountains in Southern California ous and metamorphic rocks as well as Meso- (Bull, 1991). (Fig. 1) were formed by cycles of river aggra- zoic granites. It is bounded in the south by Whether a river incises, aggrades, or main- dation and incision that occurred in response the predominantly reverse-slip Sierra Madre– tains a stable bed, depends on the balance to temporal variations in hillslope sediment Cucamonga fault zone, along which Holocene between the river’s transport capacity and the supply and river transport capacity induced uplift rates of ~0.5–0.9 mm yr–1 have been supply of sediments from hillslopes (Lane, by climatic changes. Moreover, Bull (1991) inferred, based on offset landforms (Petersen 1955; Bull, 1979, 1991). Temporal variations considered the San Gabriel Mountains as a and Wesnousky, 1994; Lindvall and Rubin, in transport capacity are mostly due to changes type locality for the response of an unglaci- 2008). Uplift and exhumation of the San Gabriel in discharge and the slope of the bed, whereas ated, semiarid to subhumid mountain range Mountains are thought to have initiated at ca. both the amount and the caliber of sediment that to past climatic changes. Because the San 12 Ma, and accelerated at ca. 5–7 Ma, when the is supplied from hillslopes may vary. In static Gabriel Mountains are mostly a steep bedrock San Andreas fault replaced the San Gabriel fault equilibrium, rivers neither incise nor aggrade, landscape (Lamb et al., 2011), rather than a as the principal strike-slip fault in this region and their transport capacity is just in balance soil-mantled landscape, with slope angles (Matti and Morton, 1993; Blythe et al., 2002). with the supply of hillslope sediments. Any greater than the angle of repose (DiBiase et al., Spatial differences in the onset and rate of uplift change of the system’s variables may lead to 2012), they are limited in their ability to have between different fault-bounded blocks within departure from equilibrium. For example, if dis- a much thicker soil mantle even under differ- the San Gabriel Mountains are thought to be charge increases, with everything else staying ent Pleistocene climatic conditions. Because of responsible for a northwest-southeast increase constant, the river increases its transport capac- their apparent significance for understanding in landscape-scale relief that coincides with ity and incises. Incision should result in shal- how steep bedrock landscapes respond to cli- younger mineral cooling ages (Blythe et al., lowing of the bed gradient until the river reaches mate change, we have chosen the San Gabriel 2000) and higher 10Be-derived erosion rates a new equilibrium. A decrease in discharge, or Mountains for a detailed analysis using mod- (DiBiase et al., 2010), as well as enhanced an increase in the size or amount of sediment ern analytical tools and the aim to identify the fluvial erosion and more frequent landsliding supply, on the other hand, would force a river to processes responsible for valley aggradation.