Debris flows as agents of morphological heterogeneity at low-order confluences, Olympic Mountains, Washington Lee Benda† Earth Systems Institute, 310 North Mount Shasta Boulevard, Suite 6, Mount Shasta, California 96067, USA Curt Veldhuisen‡ Skagit System Cooperative, 25944 Community Plaza Way, Sedro Woolley, Washington 98284, USA Jenelle Black§ Hart Crowser Incorporated, 1257 Main Street, Fortuna, California 95540, USA ABSTRACT were limited. Our field data and informa- iment supply that are often accompanied by tion from seven other studies indicate how changes in the size of sediment delivered. Effects of tributary junctions on longi- variation in debris flow volume and com- Morphological changes also occur upstream tudinal patterns of riverine heterogeneity position, stream energy, and valley width at of fans due to valley constrictions and fan- are relevant to both fluvial geomorphology the point of deposition influence the rela- induced reductions in channel gradient that and riverine ecology. We surveyed 10 km tionship between low-order confluences and impede transport of sediment and organic ma- of small- to moderate-sized mountain chan- channel morphology. terial. Morphological effects upstream and nels in the Olympic Peninsula, Washington, downstream of alluvial and debris fans include to investigate how low-order confluences Keywords: slope stability, fluvial geomor- forming steeper and shallower stream gradi- prone to debris flow deposition directly and phology, debris flows, natural hazards. ents, terraces, and wider floodplains (Small, indirectly influenced channel and valley 1973; Grant and Swanson, 1995; Schmidt and morphology. In the Olympic Mountains, INTRODUCTION Rubin, 1995; Benda et al., in press), channel debris flows scour sediment and organic meanders and braids (Benda, 1990; Knighton, material from steep first- and second-order Effects of tributary confluences on mor- 1998), wider and deeper channels (Richards, channels and create deposits (debris fans) phology of streams and rivers have been rec- 1980; Benda et al., in press), boulder deposits at tributary junctions in higher-order ognized in a number of landscapes over the oftenleading to rapids (Wohl and Pearthree, streams. In lower-energy depositional en- past half-century. Tributary junctions can af- 1991; Grimm et al., 1995; Griffiths et al., vironments there were statistically signifi- fect development of floodplains and terraces, 1996), ponds (Everest and Meehan, 1981), cant relationships among debris fans at planform and hydraulic geometry, and sub- mid-channel bars (Best, 1988), and log jams low-order confluences and gravel substrate, strate size in receiving channels, which may (Hogan et al., 1998). Channel effects are fur- wide channels, and numbers of logs and lead to increased physical heterogeneity. This ther differentiated according to whether fans large pools. Effects of debris fans on chan- aspect of fluvial geomorphology is directly form by debris flows (Benda, 1990; Wohl and nel morphology extended upstream and relevant to emerging perspectives in riverine Pearthree, 1991), flash floods—including downstream of fan perimeters, indicating ecology that emphasize the role of physical those generated by accelerated post-fire ero- the importance of indirect (offsite) effects of heterogeneity in maintaining diverse and pro- sion and sedimentation (Schmidt and Rubin, debris flows. Consequently, certain aspects ductive aquatic and riparian habitats. Fans cre- 1995; Benda et al., in press)—and less punc- of channel morphology (e.g., pool density, ated by debris flows, fire-induced erosion and tuated runoff-generated floods (Harvey, 1997). substrate texture, and channel widths) were sedimentation, and large floods also have im- Alluvial and debris flow fans can create nonuniformly distributed, reflecting the plications for the principle of disturbance morphological conditions in channels that dif- role of network topology and disturbance ecology, particularly as it pertains to the dy- fer from reaches located up- and downstream history on the spatial scale of morphologi- namics of habitat formation. The purpose of of confluences. As such, tributary confluences cal heterogeneity. Moreover, heterogeneity this paper is to evaluate the role of debris flow can be viewed as agents of morphological het- of channel morphology increased in prox- fans on channel heterogeneity along small- to erogeneity that affect characteristics of terrac- imity to low-order confluences prone to de- moderate-sized channels in the Olympic es, floodplains, bars, channel width, depth, bris flows. In contrast, confluence effects in Mountains, Washington. A further objective is substrate, and log jams. Heterogeneity arises higher-energy depositional environments to evaluate how stream energy affects debris simply due to the occurrence of different mor- fan impacts on channel morphology. phological conditions at and near junctions relative to areas upstream and downstream of †E-mail: [email protected]. Changes in channel morphology occur ‡E-mail: [email protected]. downstream of alluvial and debris fans be- them and also due to the increased range of §E-mail: [email protected]. cause of abrupt increases in discharge and sed- conditions that can occur at junctions. Chang- GSA Bulletin; September 2003; v. 115; no. 9; p. 1110–1121; 10 figures; 7 tables. For permission to copy, contact [email protected] 1110 ᭧ 2003 Geological Society of America DEBRIS FLOWS AS AGENTS OF MORPHOLOGICAL HETEROGENEITY es in morphological heterogeneity at junctions at low- to high-order confluences. Debris flow energy streams in managed and unmanaged have relevance for fluvial geomorphology, in- deposits can contribute to habitats, including basins on northwestern Olympic Peninsula. cluding sediment transport and storage forming of ponds that become occupied by Study sites included 4 km of third- through (Schmidt and Rubin, 1995; Knighton, 1998), fish and beaver (Everest and Meehan, 1981), fifth-order channels in Finley Creek Basin floodplain development (Grant and Swanson, releasing nutrients due to buried organics (Se- (Quinault River), Matheny Creek Basin (Qui- 1995), and channel morphology (Best, 1988; dell and Dahm, 1984), depositing woody de- nault River), and Sitkum Creek Basin (Sitkum Benda, 1990). Hence, certain aspects of fluvial bris that creates sediment wedges and pools River). In those basins, landslides and their geomorphology are affected by location and (Hogan et al., 1998), depositing boulders that channel effects occurred in areas of primarily size of fans that impinge on channels. trap sediments and create complex habitats unmanaged forests. Study sites also included The perspective of confluences as agents of (Reeves et al., 1995), and forming wider val- 6 km of channel in the Sekiu River Basin cov- morphological heterogeneity in streams and ley floors containing larger floodplains (Grant ered in second-growth forests (Fig. 1, Table rivers is complimentary with emerging per- and Swanson, 1995). Spates of debris flows 1). The majority of landslides inventoried in spectives in riverine ecology. Specifically, may also contribute to varying composition of the Finley, Matheny, and Sitkum River Basins physical heterogeneity is necessary for provid- riparian forests (Nierenberg and Hibbs, 2000) were triggered during a large storm in March ing a range of habitats needed for species and and to increased diversity of channel mor- of 1997 that generated flood flows with recur- population persistence (Reeves et al., 1995; phology (Nakamura et al., 2002). rence intervals of 10–50 yr (U.S. Geological Fausch et al., 2002; Weins, 2002; Poole, Despite the constructive effects of debris Survey, 1997). 2002). Moreover, morphological heterogeneity flows, they also can have negative biological Study sites were selected where channel ge- can be directly linked to biodiversity (Resh et consequences. These include immediate burial ometry encouraged deposition of debris flows al., 1988). In general, however, the role of of existing habitat and direct mortality of at low- to higher-order confluences (Benda branching networks in riverine ecology re- aquatic biota (Everest and Meehan, 1981), in- and Cundy, 1990). Across the study basins, mains an outstanding question (Fisher, 1997), creased fine sediment in gravels onsite and annual precipitation averages 2800–4000 mm although recent fieldwork and modeling have downstream that suffocates fish eggs (Everest yrϪ1, mostly as rain between October and specifically pointed to the importance of trib- et al., 1987; Scrivener and Brownlee, 1989), April. Lithology varies between marine vol- utary confluences in fluvial geomorphology and increased bedload transport and lateral canic rocks to marine sandstones, siltstones, and aquatic ecology (Benda and Dunne, channel movement due to heightened sedi- and mudstones. Steep and mass wasting-prone 1997b; Rice et al., 2001; Gomi et al., 2002). ment supply that scours fish eggs (Tripp and topography characterize the Olympic Moun- The issue of confluence effects in river Poulin, 1988). tains, and erosion is dominated by shallow ecology also relates to the concept of ‘‘distur- The objective of this paper is to contribute landsides and debris flows (Reid, 1981). bance,’’ an ecological principle that recogniz- to the growing body of research on confluence The Sekiu study basins were logged begin- es the important role of dynamic watershed effects in rivers by examining the role of de- ning in the 1960s. At least some of the streams processes (i.e., fires, floods,