Channel-reach morphology in mountain drainage basins
David R. Montgomery* ] John M. Buffington* J
Department of Geological Sciences, University of Washington, Seattle, Washington 98195
ABSTRACT
A classification of channel-reach morphology in mountain drainage basins synthesizes stream morphologies into seven distinct reach types: colluvial, bedrock, and five alluvial channel types (cascade, step pool, plane bed, pool riffle, and dune ripple). Coupling reach-level channel processes with the spatial arrangement of reach morphologies, their links to hillslope processes, and external forcing by confinement, riparian vegetation, and woody debris defines a process-based framework within which to assess channel condition and response potential in mountain drainage basins. Field investigations demonstrate characteristic slope, grain size, shear stress, and roughness ranges for different reach types, observations consistent with our hypothesis that alluvial channel morphologies reflect specific roughness configurations adjusted to the relative magnitudes of sediment supply and transport capacity. Steep alluvial channels (cascade and step pool) have high ratios of transport capacity to sediment supply and are resilient to changes in discharge and sediment supply, whereas low-gradient alluvial channels (pool riffle and dune ripple) have lower transport capacity to supply ratios and thus exhibit significant and prolonged response to changes in sediment supply and discharge. General differences in the ratio of transport capacity to supply between channel types allow aggregation of reaches into source, transport, and response segments, the spatial distribution of which provides a watershed-level conceptual model linking reach morphology and channel processes. These two scales of channel network classification define a framework within which to investigate spatial and temporal patterns of channel response in mountain drainage basins.
INTRODUCTION
Geologists and engineers have long recognized fundamental differences between mountain channels and their lowland counterparts (e.g., Surell, 1841; Dana, 1850; Shaler, 1891). In contrast to self-formed flood-plain channels, the gradient and morphology of mountain channels are tremendously variable and prone to forcing by external influences. Although mountain channels provide important aquatic habitat (e.g., Nehlsen et al., 1991;Frissell, 1993), supply sediment to estuaries and the oceans (e.g., Mil-liman and Syvitski, 1992), and transmit land use disturbances from headwater areas down through drainage networks (e.g., Reid, 1993), they have received relatively little study compared to lowland rivers.
Improved ability to relate morphology and processes in mountain channels would facilitate understanding and predicting their response to both human and natural disturbance. Classification schemes can organize such understanding into conceptual models that provide further insight into channel
*E-mail: dave@bigdirt.geology.washington.edu ^E-mail: jbuff@u. washington.edu
processes (e.g., Schumm, 1977). With few exceptions (e.g., Paustian et al., 1992; Whiting and Bradley, 1993), classifications of mountain channels are not process based, which compromises their use for assessing channel condition, response potential, and relations to ecological processes.
In order to provide a useful general classification of mountain channels, a typology should be applicable on more than a regional basis, yet adaptable to regional variability; otherwise proliferation of regional channel classifications could impede rather than enhance communication and understanding. Moreover, a classification should rely on aspects of channel form that reflect channel processes. Furthermore, it should encompass the whole channel network, rather than consider only channels inhabited by desirable organisms or indicator species. A process-based understanding of spatial linkages within a watershed is essential for assessment of channel condition, prediction of channel response to disturbance, and interpretation of the causes of historical channel changes.
Herein we systematize a channel classification that expands on Schumm's (1977) general delineation of erosion, transport, and deposition reaches and provides a framework for examining channel processes in mountain drainage basins. We also report a field test of the classification using data from drainage basins in Oregon and Washington and propose a genetic explanation for the distinct channel morphologies that we recognize. The tie to channel processes and morphogenesis provides a defensible theoretical and conceptual framework within which to classify channel morphology, assess channel condition, and interpret response potential. In particular, coupling of process-based channel classification with landscape-specific spatial linkages can provide insight into how disturbances propagate through drainage basins. Our classification arose from field work in mountain drainage basins where we repeatedly observed the same general sequence of channel morphologies down through the channel network. Here we draw on previous work and our own field observations to discuss these morphologies and propose a theory for the origin of distinct alluvial channel types. Although developed based on literature review and field observations in the Pacific Northwest (Montgomery and Buffington, 1993), subsequent field work confirms the relevance of the classification in other mountainous regions.
Channel-reach Morphology
A voluminous literature on channel classification attests to the wide variety of morphologies exhibited by stream channels. No single classification can satisfy all possible purposes, or encompass all possible channel types; each of the channel classifications in common use have advantages and disadvantages for use in geological, engineering, and ecological applications (see discussion in Kondolf, 1995). Although stream channels possess a continuum of characteristics identifiable at spatial scales that range from individual channel units to entire drainage basins (Frissell et al., 1986), channel reaches of at least 10 to 20 channel widths in length define a useful scale over which to relate stream morphology to channel processes, response potential, and habitat characteristics.
GSA Bulletin; May 1997; v. 109; no. 5; p. 596-611; 12 figures; 3 tables.
596
CHANNEL-REACH MORPHOLOGY IN MOUNTAIN BASINS
TABLE 1. DIAGNOSTIC FEATURES OF EACH CHANNEL TYPE
	Dune ripple	Pool riffle	Plane bed	Step pool	Cascade	Bedrock	Colluvial
Typical bed material	Sand	Gravel	Gravel-cobble	Cobble-boulder	Boulder	Rock	Variable
Bedform pattern	Multilayered	Laterally oscillatory	Featureless	Vertically oscillatory	Random	Irregular	Variable
Dominant roughness elements	Sinuosity, bedforms (dunes, ripples, bars) grains, banks	Bedforms (bars, pools), grains, sinuosity, banks	Grains, banks	Bedforms (steps, pools), grains, banks	Grains, banks	Boundaries (bed and banks)	Grains
Dominant sediment sources	Fluvial, bank failure	Fluvial, bank failure	Fluvial, bank failure, debris flows	Fluvial, hillslope, debris flows	Fluvial, hillslope, debris flows	Fluvial, hillslope, debris flows	Hillslope, debris flows
Sediment storage elements	Overbank, bedforms	Overbank, bedforms Overbank		Bedforms	Lee and stoss sides of flow obstructions	Pockets	Bed
Typical confinement	Unconfined	Unconfined	Variable	Confined	Confined	Confined	Confined
Typical pool spacing (channel widths)	5 to 7	5 to 7	None	1 to 4	<1	Variable	Unknown
We recognize three primary channel-reach substrates: bedrock, alluvium, and colluvium. Bedrock reaches lack a contiguous alluvial bed and reflect high transport capacities relative to sediment supply; they are typically confined by valley walls and have steep slopes. In contrast, alluvial channels exhibit a wide variety of morphologies and roughness configurations that vary with slope and position within the channel network, and may be either confined, with little to no associated flood plain, or unconfined, with a well-established flood plain. We recognize five distinct alluvial reach morphologies: cascade, step pool, plane bed, pool riffle, and dune ripple. Colluvial channels form an additional reach type that we recognize separately from alluvial channels, despite the common presence of a thin alluvial substrate. Colluvial channels typically are small headwater streams that flow over a colluvial valley fill and exhibit weak or ephemeral fluvial transport. Each of these channel types is distinguished by a distinctive channel-bed morphology, allowing rapid visual classification. Diagnostic features of each channel type are summarized in Table 1 and discussed below.
Cascade Channels
The term "cascade" connotes tumbling flow, although its specific morphologic definition varies and often is applied to both channel units and reaches (e.g., Bisson et al., 1982; Grant et al., 1990). Our delineation of cascade channels focuses on streams in which energy dissipation is dominated by continuous tumbling and jet-and-wake flow over and around individual large clasts (e.g., Peterson and Mohanty, 1960) (Fig. 1A). Cascade channels generally occur on steep slopes, are narrowly confined by valley walls, and are characterized by longitudinally and laterally disorganized bed material typically consisting of cobbles and boulders (Fig. 2A). Small, partially channel-spanning pools spaced less than a channel width apart are common in cascade channels. Tumbling flow over individual grain steps and turbulence associated with jet-and-wake flow around grains dissipates much of the mechanical energy of the flow (Fig. 3A).
Large particle size relative to flow depth makes the largest bed-forming material of cascade reaches effectively immobile during typical flows. Studies of steep-gradient channels report that large bed-forming grains typically become mobile only during infrequent (i.e., 50-100 yr) hydrologie events (Grant etal., 1990; Kondolf etal., 1991; Whittaker, 1987b). Mobilization of these larger clasts is accompanied by high sediment transport rates due to the release of finer sediment trapped under and around large grains (Sawada et al., 1983; Warburton, 1992). During lesser floods, gravel stored in low energy sites is mobilized and travels as bedload over larger bed-forming clasts (Griffiths, 1980; Schmidt and Ergenzinger, 1992). Gravel and finer material
are locally stored on stoss and lee sides of flow obstructions (i.e., large grains and large woody debris) due to physical impoundment and generation of velocity shadows. One tracer study (Kondolf et al., 1991) showed that material in such depositional sites was completely mobilized during a seven-year recurrence-interval event, whereas no tracer movement was observed during flows of less than the annual recurrence interval.
These observations suggest that there are two thresholds for sediment transport in cascade channels. During moderate recurrence-interval flows, bedload material is rapidly and efficiently transported over the more stable bed-forming clasts, which have a higher mobility threshold corresponding to more infrequent events. The lack of significant in-channel storage (Kondolf et al., 1991) and the rapid scour of depositional sites during moderately frequent high flows suggest that sediment transport is effectively supply limited in cascade channels. Bedload transport studies demonstrate that steep channels in mountain drainage basins are typically supply limited, receiving seasonal or stochastic sediment inputs (Nanson, 1974; Griffiths, 1980; Ashida et al., 1981; Whittaker, 1987). Because of this high transport capacity relative to sediment supply, cascade channels function primarily as sediment transport zones that rapidly deliver sediment to lower-gradient channels.
Step-Pool Channels
Step-pool channels are characterized by longitudinal steps formed by large clasts organized into discrete channel-spanning accumulations that separate pools containing finer material (Figs. IB and 2B) (Ashida et al., 1976, 1981; Griffiths, 1980; Whittaker and Jaeggi, 1982; Whittaker and Davies, 1982; Whittaker, 1987a, 1987b; Chin, 1989; Grant etal., 1990). Primary flow and channel bed oscillations in step-pool reaches are vertical, rather than lateral, as in pool-riffle channels (Fig. 3B). The stepped morphology of the bed results in alternating critical to supercritical flow over steps and subcritical flow in pools (Bowman, 1977; Chin, 1989). Step-pool channels exhibit a pool spacing of roughly one to four channel widths (Bowman, 1977; Whittaker, 1987b; Chin, 1989; Grant et al., 1990), significantly less than the five to seven channel widths that typify self-formed pool-riffle channels (Leopold etal., 1964; Keller and Melhorn, 1978). Steps provide much of the elevation drop and roughness in step-pool channels (Ashida etal., 1976; Whittaker and Jaeggi, 1982; Whittaker, 1987a, 1987b; Chin, 1989). Step-pool morphology generally is associated with steep gradients, small width to depth ratios, and pronounced confinement by valley walls. Although step-forming clast sizes typically are comparable to annual high flow depths, a stepped longitudinal profile also may develop in steep sand-bedded channels (G. E. Grant, 1996, personal commun.).
Geological Society of America Bulletin, May 1997
597
MONTGOMERY AND BUFFINGTON
Step-forming material may be viewed as either a kinematic wave (Lang- (Church and Jones, 1982), or as macroscale antidunes (McDonald and bein and Leopold, 1968), a congested zone of large grains that causes in- Banerjee, 1971; Shaw and Kellerhals, 1977; Grant and Mizuyama, 1991). creased local flow resistance and further accumulation of large particles    Step-pool sequences form through armoring processes under high dis-
598
Geological Society of America Bulletin, May 1997
CHANNEL-REACH MORPHOLOGY IN MOUNTAIN BASINS
MS: $m
Figure 1. (Continued—caption on facing page).
charges and low sediment supply (Ashidaetal., 1981; Whittaker and Jaeggi, 1982). Grant et al. (1990) suggested that low sediment supply and infrequent discharges capable of moving the coarsest sediment are required for development of stepped-bed morphology, and Grant and Mizuyama (1991) suggested that step-pool formation requires a heterogeneous bed mixture and near-critical flow. Furthermore, step spacing corresponds to maximum flow resistance, providing stability for a bed that would otherwise be mobile (Whittaker and Jaeggi, 1982; Abrahams etal., 1995).
Step-pool channels have several sediment transport thresholds. Large bed-
forming material generally is mobile only during relatively infrequent hydro-logic events (Whittaker, 1987a, 1987b; Grant et al., 1990), although Warbur-ton (1992) showed that step-forming clasts in steep proglacial channels may be mobile annually. Significant movement of all grain sizes occurs during extreme floods, and step-pool morphology is reestablished during the falling limb of the hydrograph (Sawada et al., 1983; Whittaker, 1987b; Warbuton, 1992). During more frequent discharges, finer material stored in pools travels as bedload over stable bed-forming clasts (Ashida et al., 1981; Whittaker, 1987a, 1987b; Ergenzinger and Schmidt, 1990; Grant et al., 1990; Schmidt and Ergenzinger, 1992). In a series of tracer tests in a step-pool channel, Schmidt and Ergenzinger (1992) found that all of the tagged particles placed in pools mobilized during frequent, moderate discharges and were preferentially redeposited into pools. Transport of all the pool-filling material indicates that sediment transport of non-step-forming grains is supply limited. Bedload studies in step-pool channels demonstrate complex relations between discharge and sediment transport; transport rates are dependent on seasonal and stochastic sediment inputs, flow magnitude and duration, and antecedent events (Nanson, 1974; Griffiths, 1980; Ashida et al., 1981; Sawada et al., 1983; Whittaker, 1987a, 1987b; Warburton, 1992). Ashida et al. (1981), for example, observed a 10 hr lag between the hydrograph peak and onset of bed-load transport for step-pool channels scoured of all pool-filling sediment during previous storms. Hydrograph peaks and bedload transport were, however, directly correlated during a subsequent storm due to the availability of sediment deposited in pools. Warburton (1992) suggested three phases of sediment transport in step-pool channels: a low-flow flushing of fines; frequent high-flow mobilization of pool-filling gravel (also noted by Sawada et al., 1983); and less-frequent higher-discharge mobilization of step-forming grains.
Although step-pool and cascade channel morphologies both reflect supply-limited transport, they are distinguished by differences in the spatial
Geological Society of America Bulletin, May 1997
599
MONTGOMERY AND BUFFINGTON
density and organization of large clasts. Step-pool channels are defined by discrete channel-spanning steps less than a channel width in length that separate pools spaced every one to four channel widths. Cascade channels are defined by ubiquitous tumbling and jet-and-wake flow over a series of individual large clasts that together exceed a channel width in length, with small, irregularly placed pools spaced less than a channel width apart. The regular sequence of pools and steps in step-pool channels probably represents the emergence of a fluvially organized morphology in alluvial channels. In contrast, the disorganized large clasts of cascade channels may include lag deposits forced by nonfluvial processes (e.g., debris flows, glaciers, and rock falls).
Plane-Bed Channels
The term "plane bed" has been applied to both planar bed phases observed to form in sand-bed channels (Simons et al., 1965) and planar gravel and cobble-bed channels (Florsheim, 1985) like the coarse-grained, threshold canals described by Lane and Carlson (1953). Our use of the term refers to the latter and encompasses glide (run), riffle, and rapid morphologies described in the fisheries literature (e.g., Bisson et al., 1982). Plane-bed channels lack discrete bars, a condition that is associated with low width to depth ratios (Sukegawa, 1973; Ikeda, 1975, 1977) and large values of relative
Figure 2. Schematic planform illustration of alluvial channel morphologies at low flow: (A) cascade channel showing nearly continuous, highly turbulent flow around large grains; (B) step-pool channel showing sequential highly turbulent flow over steps and more tranquil flow through intervening pools; (C) plane-bed channel showing single boulder protruding through otherwise uniform flow; (D) pool-riffle channel showing exposed bars, highly turbulent flow through riffles, and more tranquil flow through pools; and (E) dune-ripple channel showing dune and ripple forms as viewed through the flow.
roughness (ratio of 90th percentile grain size to bankfull flow depth). Church and Jones (1982) considered bar formation unlikely at relative roughnesses of 0.3 to 1.0. Plane-bed reaches occur at moderate to high slopes in relatively straight channels that may be either unconfined or confined by valley walls. They typically are composed of sand to small boulder grain sizes, but are dominantly gravel to cobble bedded.
Plane-bed channels differ morphologically from both step-pool and pool-riffle channels in that they lack rhythmic bedforms and are characterized by long stretches of relatively featureless bed (Figs. 1C and 2C). The absence of tumbling flow and smaller relative roughness distinguish plane-bed reaches from cascade and step-pool channels (Fig. 3C). Plane-bed channels lack sufficient lateral flow convergence to develop pool-riffle morphology due to lower width to depth ratios and greater relative roughness, which may decompose lateral flow into smaller circulation cells. However, introduction of flow obstructions may force local pool and bar formation.
Plane-bed channels typically exhibit armored bed surfaces calculated to have a near-bankfull threshold for mobility, although elevated sediment loading can cause textural fining and a lower calculated mobility threshold (Buffington, 1995). Plane-bed channels with armored bed surfaces indicate a transport capacity greater than sediment supply (i.e., supply-limited conditions), whereas unarmored surfaces indicate a balance between transport capacity and sediment supply (Dietrich et al., 1989). Nevertheless, beyond
600
Geological Society of America Bulletin, May 1997
CHANNEL-REACH MORPHOLOGY IN MOUNTAIN BASINS
Figure 3. Schematic longitudinal profiles of alluvial channel morphologies at low flow: (A) cascade; (B) step pool; (C) plane bed; (D) pool riffle; and (E) dune ripple.
the threshold for significant bed-surface mobility, many armored gravel-bedded channels exhibit a general correspondence between bedload transport rate and discharge (e.g., Milhous, 1973; Jackson and Beschta, 1982; Sidle, 1988), implying transport-limited conditions. The above observations suggest that plane-bed channels are transitional between supply- and transport-limited morphologies.
Pool-Riffle Channels
Pool-riffle channels have an undulating bed that defines a sequence of bars, pools, and riffles (Leopold et al., 1964) (Fig. ID). This lateral bedform oscillation distinguishes pool-riffle channels from the other channel types discussed above (Fig. 2D). Pools are topographic depressions within the channel and bars are corresponding high points (Fig. 3D); these bedforms
are thus defined relative to each other (O'Neill and Abrahams, 1984). Pools are rhythmically spaced about every five to seven channel widths in self-formed, pool-riffle channels (Leopold et al., 1964; Keller and Mellhörn, 1978), but channels with a high loading of large woody debris exhibit smaller pool spacing (Montgomery et al., 1995). Pool-riffle channels occur at moderate to low gradients and are generally unconfined, and have well-established flood plains. Substrate size in pool-riffle streams varies from sand to cobble, but typically is gravel sized.
Bar and pool topography generated by local flow convergence and divergence may be either freely formed by cross-stream flow and sediment transport, or forced by channel bends and obstructions (e.g., Lisle, 1986). Free-formed pool-riffle sequences initially result from internal flow perturbation that causes flow convergence and scour on alternating banks of the channel; concordant downstream flow divergence results in local sediment accumulation in discrete bars. Topographically driven convective accelerations reinforce convergent and divergent flow patterns, and thus pool-riffle morphogenesis (Dietrich and Smith, 1983; Dietrich and Whiting, 1989; Nelson and Smith, 1989). Alluvial bar development requires a sufficiently large width to depth ratio and small grain sizes that are easily mobilized and stacked by the flow (Church and Jones, 1982). Bar formation in natural channels appears to be limited to gradients <0.02 (Ikeda, 1977; Florsheim, 1985), although flume studies indicate that alternate bars may form at steeper gradients (Bathurst et al., 1983; Lisle et al., 1991). Bedform and grain roughness provide the primary flow resistance in free-formed pool-riffle channels.
Pool-riffle channels have heterogeneous beds that exhibit a variety of sorting and packing, commonly with a coarse surface layer and a finer subsurface (Leopold et al., 1964; Milhous, 1973). Armored gravel-bed channels typically exhibit a near-bankfull threshold for general and significant bed-surface mobility (e.g., Parker et al., 1982; Jackson and Beschta, 1982; Andrews, 1984; Carling, 1988; Buffington, 1995). Movement of surface grains releases fine sediment trapped by larger grains and exposes finer subsurface sediment to the flow, contributing to a steep rise in bedload transport with increasing shear stress (Milhous, 1973; Jackson and Beschta, 1982; Em-mett, 1984). Bed movement is sporadic and discontinuous, depending on grain protrusion (Fenton and Abbott, 1977; Kirchner et al., 1990), friction angle (Kirchner et al., 1990; Buffington et al., 1992), imbrication (Komar and Li, 1986), degree of burial (Hammond et al., 1984; Buffington et al., 1992), and turbulent high-velocity sweeps of the channel bed. Very rarely is the whole bed in motion, and material eroded from one riffle commonly is deposited on a proximal downstream riffle.
Pool-riffle channels, like plane-bed channels, exhibit a mixture of supply-and transport-limited characteristics depending on the degree of bed-surface armoring and consequent mobility thresholds. Unarmored pool-riffle channels indicate a balance between transport capacity and sediment supply, while armored surfaces represent supply-limited conditions (e.g., Dietrich et al., 1989). Nevertheless, during armor-breaching events, bedload transport rates are generally correlated with discharge, demonstrating that sediment transport is not limited by supply once the bed is mobilized. Considerable fluctuations in observed transport rates, however, reflect a stochastic component of grain mobility caused by grain interactions, turbulent sweeps, and transient grain entrapment by bedforms (Jackson and Beschta, 1982; Sidle, 1988). Magnitudes of bedload transport also may vary for similar discharge events, depending on the chronology of antecedent transport events (Milhous, 1973; Reid et al., 1985; Sidle, 1988). Although both pool-riffle and plane-bed channels display a mix of supply- and transport-limited characteristics, the presence of depositional barforms in pool-riffle channels suggests that they are generally more transport limited than plane-bed channels. The transport-limited character of both of these morphologies, however, contrasts with the more supply-limited character of step-pool and cascade channels.
Geological Society of America Bulletin, May 1997
601
MONTGOMERY AND BUFFINGTON
Dune-Ripple Channels
Dune-ripple morphology is most commonly associated with low-gradient, sand-bed channels (Figs. IE, 2E, and 3E). A flow regime-dependent succession of mobile bedforms provides the primary hydraulic resistance in dune-ripple channels (e.g., Kennedy, 1975). However, even gravel-bed channels can exhibit a succession of multiple-scale bedforms during extreme discharges (e.g., Griffiths, 1989; Dinehart, 1992; Pitlick, 1992). The bedform configuration of dune-ripple channels depends on flow depth, velocity, bed-surface grain size, and sediment transport rate (e.g., Gilbert, 1914; Middle-ton and Southard, 1984), but generally follows a well-known morphologic sequence with increasing flow depth and velocity: lower-regime plane bed, ripples, sand waves, dunes, upper-regime plane bed, and antidunes (Gilbert, 1914; Simons et al., 1965; Harms et al., 1975). In channels transporting moderately to poorly sorted sediment, migrating bedload sheets composed of thin accumulations of sediment also may develop (Whiting et al., 1988). Several scales of bedforms may coexist in a dune-ripple channel; ripples, bedload sheets, and small dunes may climb over larger mobile dunes. A complete theoretical explanation for the development of such multiple-scale bedforms does not yet exist, but they are typically associated with low relative roughness. Dune-ripple channels also exhibit point bars or other bedforms forced by channel geometry. In contrast to the threshold sediment transport of plane-bed and pool-riffle streams, dune-ripple channels exhibit "live bed" transport (e.g., Henderson, 1963), in which significant sediment transport occurs at most stages. Hence, dune-ripple channels are effectively transport limited. The frequency of bed mobility and the presence of ripples and/or dunes distinguish dune-ripple channels from pool-riffle channels.
Colluvial Channels
Colluvial channels are small headwater streams at the tips of a channel network that flow over a colluvial valley fill and exhibit weak or ephemeral fluvial transport (Fig. IF). Little research has focused on colluvial channels, even though first-order channels compose approximately half of the total length of a channel network (Montgomery, 1991). Dietrich et al. (1982) recognized that shallow flows in headwater channels have little opportunity for scour, and therefore sediment delivered from neighboring hillslopes generally accumulates to form colluvial valley fills. Benda and Dunne (1987) examined sediment in steep headwater valleys in the Oregon Coast Range and concluded that beneath a water-worked coarse surface layer, the valley fill consists of relatively unsorted colluvium delivered from surrounding hill-slopes. Shallow and ephemeral flow in colluvial channels appears insufficient to mobilize all of the colluvial sediment introduced to the channel, resulting in significant storage of this material (Dietrich and Dunne, 1978; Dietrich et al., 1982; Benda, 1990). Large clasts, woody debris, bedrock steps, and in-channel vegetation further reduce the energy available for sediment transport in colluvial channels. Intermittent flow may rework some portion of the surface of the accumulated material, but it does not govern deposition, sorting, or transport of the valley fill.
Episodic transport by debris flows may account for most of the sediment transport in steep headwater channels. A sediment budget for a small basin in northern California indicated that debris flows account for more than half of the long-term sediment yield (Lehre, 1982). Swanson et al. (1982) estimated that only 20% of the total sediment yield from a first-order channel in the Cascade Range is accommodated by fluvial transport. Hence, the long-term sediment flux from low-order channels in steep terrain appears to be dominated by debris-flow processes. Differences in channel profiles support the hypothesis that different processes dominate the erosion of steep headwater channels and lower-gradient alluvial channels in the Oregon Coast Range (Seidl and Dietrich, 1992).
Dietrich and Dunne (1978) recognized that the residence time of sediment in headwater debris-flow-prone channels was on the order of hundreds of years. Kelsey (1980) also estimated that the sediment stored infirst-and second-order channels is scoured by debris flows every 300 to 500 yr. Benda (1990) proposed a conceptual model for the evolution of channel morphology in steep headwater channels that involves cyclical alteration of bed morphology from gravel to boulder to bedrock in response to episodic sediment inputs. The accumulation of colluvial valley fills during periods between catastrophic scouring events indicates that transport capacity, rather than sediment supply, limits fluvial transport in colluvial channels.
Bedrock Channels
Bedrock channels lack a continuous alluvial bed. Although some alluvial material may be temporarily stored in scour holes, or behind flow obstructions, there is little, if any, valley fill. Hence, bedrock channels generally are confined by valley walls. Evidence from both anthropogenic badlands and mountain drainage basins indicates that bedrock channels are steeper than alluvial channels having similar drainage areas (Howard and Kerby, 1983; Montgomery et al., 1996). It is reasonable to adopt Gilbert's (1914) hypothesis that bedrock channels lack an alluvial bed due to high transport capacity associated with steep channel gradients and/or deep flow. Although bedrock channels in low-gradient portions of a watershed reflect a high transport capacity relative to sediment supply, those in steep portions of a watershed may also reflect recent catastrophic scouring.
Forced Morphologies
Flow obstructions can force a reach morphology that differs from the free-formed morphology for a similar sediment supply and transport capacity. In forested mountain drainage basins, for example, large woody debris may force local scour, flow divergence, and sediment impoundment that respectively form pools, bars, and steps (Fig. 1G). In an extreme example, Montgomery et al. (1996) found that log jams forced alluvial streambeds in otherwise bedrock reaches of a mountain channel network in western Washington.
Forced pool-riffle and step-pool channels are the most common obstruction-controlled morphologies in forested mountain drainage basins. A forced pool-riffle morphology is one in which most pools and bars are forced by obstructions such as large woody debris, and a forced step-pool channel is one in which large woody debris forms most of the channel-spanning steps that define the bed morphology. Forced morphologies can extend beyond the range of conditions characteristic of analogous free-formed morphologies (i.e., to steeper gradients and/or lower sediment supply). We recognize forced morphologies as distinct channel types because interpretation of whether such obstructions govern bed morphology is important for understanding channel response.
Intermediate and Other Morphologies
The channel types described above represent identifiable members along a continuum that includes several intermediate morphologies: riffle bar (pool riffle-plane bed); riffle step (plane bed-step pool); and cascade pool (step pool-cascade). Mixed alluvial and bedrock reaches exhibit subreach scale variations in alluvial cover. In our experience, however, it is simple to replicate identification of the seven basic reach types, even though they lie within a continuum of channel morphologies. Whether intermediate channel types are useful for classification purposes depends on the context of the application. Although our proposed classification does not cover all reach types in all environments (e.g., estuarine, cohesive-bed, or vegetated reaches), we have found it to be applicable in a variety of mountain environments.
602
Geological Society of America Bulletin, May 1997
CHANNEL-REACH MORPHOLOGY IN MOUNTAIN BASINS
TABLE 2. STUDY AREA CHARACTERISTICS
Study area	Geology	Drainage area	Relief	Land use
		(km2)	(m)	
Finney Creek, Washington	Phyllite, greenschist, glacial sediments	128	1476	U.S. Forest Service, state forestry
Boulder River, Washington	Phyllite, glacial sediments	63	1985	U.S. Forest Service wilderness area
South Fork Hoh River,	Sandstone, glacial sediments	129	>882	State forestry, national park
Washington				
□eton Creek, Oregon	Sandstone	8	327	Private forestry
FIELD TEST
Process differences associated with reach morphology should result in distinct physical characteristics for each reach type. Data compiled from field studies in the Pacific Northwest reveal systematic association of channel types with slope, drainage area, relative roughness, and bed-surface grain size. Furthermore, these data suggest an explanation for the origin of distinct channel types.
Study Areas and Methods
Field surveys were conducted in four drainage basins in western Washington and coastal Oregon: Finney Creek, Boulder River, South Fork Hoh River, and Deton Creek (Table 2). In each study area, channel reaches 10-20 channel widths in length were surveyed throughout the drainage basin. Each reach was classified into one of the above-defined channel types. Reach slopes were surveyed using either an engineering level or a hand level and stadia rod. Topographic surveys and channel-spanning pebble counts of 100 grains (Wolman, 1954) were conducted at representative cross sections. Reach locations were mapped onto U.S. Geological Survey 1:24,000 scale topographic maps from which drainage areas were measured using a digital planimeter. Reach slopes were determined from topographic maps for some additional reaches where morphologies were mapped, but slope and grain-size measurements were not collected. We also included in our analysis data collected using similar field methods in related studies in
initiation Debris
Figure 4. Idealized long profile from hillslopes and unchanneled hollows downslope through the channel network showing the general distribution of alluvial channel types and controls on channel processes in mountain drainage basins.
western Washington and southeast Alaska (Montgomery et al., 1995; Buff-ington, 1995).
Results
In each study area, there is a general downstream progression of reach types that proceeds as colluvial, cascade, step pool, plane bed or forced pool riffle, and pool riffle (Fig. 4); we encountered no dune-ripple reaches in the study basins, although we observed them in neighboring areas. Bedrock reaches occur at locally steep locations throughout the channel networks, and not all of these channel types are present in each watershed. Furthermore, the specific downstream sequence of reach types observed in each drainage basin reflects local factors controlling channel slope, discharge, sediment supply, bedrock lithology, and disturbance history.
Data from alluvial, colluvial, and bedrock reaches within each study basin define distinct fields on a plot of drainage area versus reach slope (Fig. 5). These data provide further evidence that, for a given drainage area, bedrock reaches have greater slopes, and hence greater basal shear stress and stream power, than either alluvial or colluvial reaches (Howard and Kerby, 1983; Montgomery et al., 1996). Alluvial reaches occur on slopes less than about 0.2 to 0.3, and different alluvial channel types generally segregate within an inversely slope-dependent band within which pool-riffle and plane-bed channels occur at the lowest slopes, and step-pool and cascade channels occur on steeper slopes. Colluvial reaches occur at lower drainage areas and extend to steeper slopes. Data from colluvial reaches define a relation between drainage area and slope that contrasts with that of lower-gradient alluvial reaches. This general pattern holds for each of the study basins, implying consistent differences among colluvial, alluvial, and bedrock reaches in mountain drainage basins.
The different drainage area-slope relation for colluvial and alluvial channel reaches implies fundamental differences in sediment transport processes. For equilibrium channel profiles, channel slope (S) and drainage area (A) are related by
S = KA-™'" (1)
where K, m, and n are empirical variables that incorporate basin geology, climate, and erosional processes (e.g., Howard et al., 1994). A log-linear regression of reach slope and drainage area data from alluvial and colluvial channels in Finney Creek yields m/n values of 0.72 ± 0.08 (R2 = 0.72) and 0.26 ± 0.05 (R2 = 0.58), respectively, which implies long-term differences in sediment transport processes between these channel types. This correspondence between the inflection in the drainage area-slope relation and the transition from colluvial to alluvial channels is consistent with the interpretation that scour by debris flows is the dominant incisional process in colluvial channels (Benda, 1990; Seidl and Dietrich, 1992; Montgomery and Foufoula-Georgiou, 1993).
Although slope ranges of free-form alluvial channel types overlap, they have distinct medians and quartile ranges (Fig. 6). Examination of the composite slope distributions indicates that reaches with slopes of less than 0.015 are likely to have a pool-riffle morphology; reaches with slopes of
Geological Society of America Bulletin, May 1997
603
MONTGOMERY AND BUFFINGTON
10v
E
©
Figure 5. Drainage area versus §■ reach slope for channels in the Fin- "to ney Creek watershed, Washington.
10"' -
10-2r
10"
I 111 III _—111 III
colluvial
A	bedrock
A	colluvial
•	pool-riffle
o	plane-bed
■	step-pool
□	cascade
10'
10'
II I I.....I    'I   II.....I      III 111 III"
alluvial
10° 10° 10'
drainage area (m2)
10<
10!
0.015 to 0.03 typically have a plane-bed morphology; reaches with slopes of 0.03 to 0.065 are likely to have a step-pool morphology; and alluvial reaches with slopes greater than 0.065 typically have a cascade morphology. These core slope ranges define zones over which each channel type is the
0.20
E
Q.
o
0.15 -
0.10 -
0.05 ■
0.00
CD
0) = o ■
eg
CL
0)
it
o o
Q.
"D
<D .Q
I
CD C 03
Q.
O
0 Q.
1
CL
2 «5
a>
CO Ü V) CO Ü
channel type
Figure 6. Composite slope distributions for channel reaches surveyed in this and related studies (Buffington, 1995; Montgomery et al., 1995); boxes represent inner and outer quartiles; vertical lines represent inner and outer tenths.
most likely to occur; however, the distributions overlap and channel type is not uniquely related to reach slope. Furthermore, forced pool-riffle reaches span the slope ranges for pool-riffle and plane-bed reaches, indicating that introduction of large woody debris can extend a forced morphology to slopes where such a morphology would not be expected under low woody debris loading (Montgomery et al., 1995). Nonetheless, the general segregation of reach type by slope allows prediction of likely channel morphology from topographic maps or digital elevation models.
Relative roughness (the ratio of the ninetieth percentile grain size to the bankfull flow depth [cL^/D]) and reach slope together differentiate alluvial reach types (Fig. 7): pool-riffle channels have relative roughness less than about 0.3 and occur on slopes <0.03; plane-bed channels exhibit relative roughness of roughly 0.2 to 0.8 on slopes of 0.01 to 0.04; step-pool reaches occur on steeper slopes and have relative roughness of 0.3 to 0.8; and the size of the largest clasts on the bed of steeper cascade reaches can approach those of bankfull flow depth. Relative roughness and reach slope together provide a reasonable stratification of channel morphology. In pool-riffle and plane-bed channels relative roughness increases rapidly with increasing slope, whereas there is little relation between relative roughness and slope for steeper step-pool and cascade reaches.
Composite bed-surface grain-size distributions for pebble counts from different channel types exhibit systematic coarsening from pool-riffle through cascade channels. For reaches in the Finney Creek watershed (Fig. 8), the median grain size increases from 17 mm for pool-riffle channels to 80 mm for cascade morphologies, and dg4 increases from 57 mm to 250 mm. These systematic changes in bed-surface grain-size distributions indicate that progressive fining of the bed material accompanies the formation of different channel types downstream through a channel network.
The data reported above demonstrate that qualitatively defined channel types exhibit quantitatively distinguishable characteristics. Our data further indicate that channel morphology is related to reach-average bankfull shear stress (Fig. 9). Bedrock channels occur in reaches with the greatest shear stress; cascade and step-pool reaches plot at lower values, which in turn are greater than those for plane-bed and pool-riffle channels. Hence, it appears that, in part, local flow hydraulics influence the general distribution of channel types in a watershed.
604
Geological Society of America Bulletin, May 1997
CHANNEL-REACH MORPHOLOGY IN MOUNTAIN BASINS
1.0
E
E
O
cn -o
0.5
-B^qľ
□
□
□
y
□
□
100
9 Pool-Riffle
O Plane-Bed
■ Step-Pool
D Cascade
0.00
0.05
0.10
0.15
0.20
slope (m/m)
Figure 7. Composite plot of relative roughness (d90/D) versus field surveyed reach slope for data from alluvial reaches in our study areas.
ORIGIN OF REACH-LEVEL MORPHOLOGIES
The typical downstream sequence of channel morphologies (Fig. 4) is accompanied by a progressive decrease in valley-wall confinement, which in stream-formed valleys may reflect opposing downstream trends of sediment supply (<2S) and transport capacity (Qc). Transport capacity is defined here as a function of the total boundary shear stress and is distinguished from the effective transport capacity (Qc'), which is a function of the effective shear stress available for sediment transport after correction for shear stress dissipation caused by hydraulic roughness elements. Transport capacity generally decreases downstream due to the slope decreasing faster than the depth increases, whereas total sediment supply generally increases with drainage area, even though sediment yield per unit area often decreases (Fig. 10). This combination may result in long-term patterns of downstream
CD C
c
CD O i_
CD CL
CD >
4-»
ra
E
o
,oo«P **
i o*"
I pool-riffle]
iOB -  • '
100
200
300
grain size (mm)
Figure 8. Aggregated cumulative grain-size distributions for alluvial channels of reaches with different bed morphologies in the Finney Creek watershed.
deposition and development of wide flood plains and unconfined valleys. Insignificant sediment storage in a valley segment indicates that virtually all of the material delivered to the channel is transported downstream. In contrast, thick alluvial valley-fill deposits imply either a long-term excess of sediment supply over transport capacity, or an inherited valley fill.
These general patterns and our field observations discussed above lead us to propose that distinctive channel morphologies reflect the relative magnitude of transport capacity to sediment supply, which may be expressed as the ratio qx = Q/Q . Colluvial channels are transport limited (gr << 1), as in-
Figure 9. Plot of drainage area versus reach. Average shear stress for bedrock (triangles), cascade and step-pool (circles), and plane-bed and pool-riffle (squares) channel morphologies are from the South Fork Hoh River study area.
Drainage Area (rrr)
Geological Society of America Bulletin, May 1997
605
MONTGOMERY AND BUFFINGTON
a
O cd
&
o
tí O
a*
CA
C/3 ft CL
B 3
"H-•T
O
drainage area
supply limited
transport limited
Figure 10. Schematic illustration of generalized relative trends in sediment supply (gs) and transport capacity (Qc) in mountain drainage basins.
dicated by the accumulation of colluvium within valley bottoms. In contrast, the lack of an alluvial bed indicates that bedrock channels are supply limited (qľ» 1). For a given drainage area (and thus Qs), bedrock reaches have greater slopes and shear stresses (Figs. 5 and 9), implying that they have higher transport capacities and thus greater qľ values than other channel types. Alluvial channels, however, probably represent a broad range of qľ: steep alluvial channels (cascade and step-pool) have higher shear stresses (Fig. 9) and thus higher Qc and qľ values for a given drainage area and sediment supply; the lower-gradient plane-bed and pool-riffle channels are transitional between qľ >1 and qľ ~\, depending on the degree of armoring (e.g., Dietrich et al., 1989) and the frequency of bed-surface mobility; and the live-bed mobility of dune-ripple channels indicates that qľ < 1. The variety of alluvial channel morphologies probably reflects a broad spectrum of qľ expressed through fining and organization of the bedload (Fig. 11), which leads to formation of distinct alluvial bed morphologies that represent the stable bed form for the imposed qť This hypothesized relation between qľ and stable channel morphologies in mountain drainage basins provides a genetic framework for explaining reach-level morphologies that elaborates on Lindley's (1919) regime concept. An alluvial channel with qľ > 1 will become stable when the bed morphology and consequent hydraulic roughness
produce an effective transport capacity that matches the sediment supply
Different channel types are stabilized by different roughness configurations that provide resistance to flow. In steep channels energy is dissipated primarily by hydraulic jumps and jet-and-wake turbulence. This style of energy dissipation is pervasive in cascade channels and periodic in step-pool channels. Skin friction and local turbulence associated with moderate particle sizes are sufficient to stabilize the bed for lower shear stresses characteristic of plane-bed channels. In pool-riffle channels, skin friction and bed-form drag dominate energy dissipation. Particle roughness in dune-ripple channels is small due to the low relative roughness, and bedforms govern hydraulic resistance. The importance of bank roughness varies with channel type, depending on the width to depth ratio and vegetative influences, but in steep channels bank resistance is less important compared to energy dissipation caused by tumbling flow. These different roughness configurations represent a range in qx values that varies from high in cascade reaches to low in dune-ripple channels.
Our hypothesis that different channel types represent stable roughness configurations for different q values implies that there should be an association of channel type and roughness. Even though the general correlation of morphology and slope (Fig. 6) implies discrete roughness characteristics among channel types, different channel morphologies occurring on the same slope should exhibit distinct roughness. Photographs and descriptions of channel morphology from previous studies in which roughness was determined from measured velocities (Barnes, 1967; Marcus et al., 1992) allow direct assessment of the roughness associated with different channel types. For similar slopes, plane-bed channels exhibit greater roughness than pool-riffle channels, and step-pool channels, in turn, appear to have greater roughness than plane-bed channels with comparable gradients (Fig. 12). Moreover, intermediate morphology reaches plot between their defining channel types. These systematic trends in roughness for a given slope strongly support the hypothesis that reach-level channel morphology reflects a dynamic adjustment of the bed surface to the imposed shear stress and sediment supply (i.e., the specific qx value).
CHANNEL DISTURBANCE AND RESPONSE POTENTIAL
Natural and anthropogenic disturbances that change hydrology, sediment supply, riparian vegetation, or large woody debris loading can alter channel processes and morphology. The effect that watershed disturbance has on a particular channel reach depends on hillslope and channel coupling, the sequence of upstream channel types, and site-specific channel morphology. In particular, the variety and magnitude of possible morphologic responses to
valley segment
channel reach
colluvial
alluvial
bedrock
colluvial	dune-ripple	pool-riffle	plane-bed	step-pool	cascade	bedrock
Qc « Qs (transport limited)
Qc » Qs (supply limited)
Figure 11. Schematic illustration of the transport capacities relative to sediment supply for reach-level channel types.
606
Geological Society of America Bulletin, May 1997
CHANNEL-REACH MORPHOLOGY IN MOUNTAIN BASINS
c
A Step-Pool/Cascade
I Plane-Bed/Step-Pool
□ Plane-Bed
# Pool-Riffle/Plane-Bed
O Pool-Riffle
A
CO
Gradient (m/m)
Figure 12. Plot of reach roughness coefficient (Manning's n) versus reach slope for channels classified according to our system using data and photographs in Barnes (1967) and Marcus et al. (1992). Note that channel types interpreted to reflect greater relative transport capacity have higher roughness over similar slopes.
a given disturbance depend on channel type, external influences (e.g., confinement, riparian vegetation, large woody debris), and disturbance history. Together these considerations provide an integrative approach for examining spatial and temporal patterns of channel disturbance and response in mountain watersheds.
Spatial Distribution of Channel Types
The spatial distribution of channel types and their coupling to both hill-slopes and one another can strongly influence the potential for a channel to be affected by a disturbance. In general, the degree of hillslope-channel coupling changes downstream through mountain channel networks, resulting in changes in both the characteristics and delivery mechanisms of sediment supplied to a channel (e.g., Rice, 1994). Furthermore, the general downstream progression of channel morphologies in mountain drainage basins (Fig. 4) causes an association of hillslope coupling and channel type. Headwater colluvial channels are strongly coupled to adjacent hillslopes, and net sediment transport from these weakly fluvial reaches is affected by the frequency of upslope debris flows and mass movements. Valley-wall confinement allows direct sediment input by hillslope processes to cascade and step-pool channels, which makes them prone to periodic disturbance from hillslope failures. Debris flows can dominate the disturbance frequency in headwater portions of the basin, scouring high-gradient channels and aggrading the first downstream reach with a gradient low enough to cause deposition of the entrained material (e.g., Benda and Dunne, 1987). Consequently, the effects of debris-flow processes on channel morphology can be divided into those related to scour, transport, and deposition. Farther downstream, the coupling between hillslopes and lower-gradient channels (i.e., plane-bed, pool-riffle, and dune-ripple) is buffered by wider valleys and dep-ositional flood plains, making these reaches less susceptible to direct disturbance from hillslope processes. Sediment characteristics, delivery, and trans-
port are generally dominated by fluvial processes in these lower-gradient channels, although forcing by large woody debris and impingement of channels on valley walls can have a significant influence on the local transport capacity and sediment supply (e.g., Rice, 1994).
The downstream sequence in which channel types are arranged also affects the potential for a disturbance to impact a particular reach. Position within the network and differences between q values allow general aggregation of channel reaches into source, transport, and response segments. In steep landscapes, source segments are transport-limited, sediment-storage sites subject to intermittent debris-flow scour (i.e., colluvial channels). Transport segments are morphologically resilient channels with a high qx (i.e., bedrock, cascade, and step-pool channels) that rapidly convey increased sediment loads. Response segments are channels with a low qx (i.e., plane-bed, pool-riffle, and dune-ripple) in which significant morphologic adjustment occurs in response to increased sediment supply. These distinctions build upon Schumm's (1977) concept of erosion, transport, and deposition zones within a watershed to provide a conceptual model that allows identification of reach-specific response potential throughout a channel network.
The spatial distribution of source, transport, and response segments governs the distribution of potential impacts and recovery times within a watershed. Downstream transitions from transport to response reaches define locations where impacts from increased sediment supply may be both pronounced and persistent. Transport segments rapidly deliver increased sediment loads to the first downstream reach with insufficient transport capacity to accommodate the additional load. Consequently, the "cumulative" effects of upstream increases in sediment supply may be concentrated in response segments where longer time and/or significant morphological change is required to transport the additional sediment. In this regard, reach-level classification identifies areas most sensitive to increases in upstream sediment inputs. Hence, downstream transitions from transport to response segments can provide ideal locations to monitor network response and should serve as critical components of watershed monitoring studies. Most important, the relation between channel type and response potential provides a direct link between upstream sediment inputs and downstream response. Identification of source, transport, and response segments thereby provides a context for examining connections between watershed modifications, impacts on channel morphology, and biological response.
Influence of Channel Type
Differences in confinement, transport capacity relative to sediment supply, and channel morphology influence channel response to perturbations in sediment supply and discharge. Thus, it is important to assess channel response potential in the context of reach type and location within a watershed. An understanding of reach morphologies, processes, and environments allows reach-specific prediction of the likely degree and style of response to a particular perturbation. Small to moderate changes in discharge or sediment supply can alter channel attributes (e.g., grain size, slope, and channel geometry); large changes can transform reach-level channel types. On the basis of typical reach characteristics and locations within mountainous watersheds, we assessed the relative likelihood of specific morphologic responses to moderate perturbations in discharge and sediment supply for each channel type (Table 3).
Channels with different bed morphology and confinement may have different potential responses to similar changes in discharge or sediment supply. Changes in sediment storage dominate the response of colluvial channels to altered sediment supply because of transport-limited conditions and low fluvial transport capacities (Table 3); depending on the degree of valley fill, increased discharge can significantly change channel geometry. In contrast, bedrock, cascade, and step-pool channels are resilient to most discharge or
Geological Society of America Bulletin, May 1997
607
MONTGOMERY AND BUFFINGTON
TABLE 3. INTERPRETED REACH-LEVEL CHANNEL RESPONSE POTENTIAL TO MODERATE CHANGES IN SEDIMENT SUPPLY AND DISCHARGE
Width   Depth Roughness Scour depth  Grain size  Slope  Sediment storage
Dune ripple	+	+	+	+	0	+	+
Pool riffle	+	+	+	+	+	+	+
Plane bed	P	+	P	+	+	+	p
Step pool	0	P	P	P	p	p	p
Cascade	0	0	P	0	p	0	0
Bedrock	0	0	0	0	0	0	0
Colluvial	p	p	0	p	p	0	+
Notes: +—	-likely, o-	—unlikely, p-	—possible.				
sediment-supply perturbations because of high transport capacities and generally supply-limited conditions. Many bedrock channels are insensitive to all but catastrophic changes in discharge and sediment load. Lateral confinement and large, relatively immobile, bed-forming clasts make channel incision or bank cutting unlikely responses to changes in sediment supply or discharge in most cascade and step-pool channels. Other potential responses in step-pool channels include changes in bedform frequency and geometry, grain size, and pool scour depths, whereas only limited textural response is likely in cascade channels. Lower gradient plane-bed, pool-riffle, and dune-ripple channels become progressively more responsive to altered discharge and sediment supply with decreasing qx, smaller grain sizes, and less channel confinement. Because plane-bed channels occur in both confined and unconfined valleys, they may or may not be susceptible to channel widening or changes in valley-bottom sediment storage. Smaller, more mobile grain sizes in plane-bed and pool-riffle channels allow potentially greater response of bed-surface textures, scour depth, and slope compared to cascade and step-pool morphologies. Unconfined pool-riffle and dune-ripple channels generally have significant potential for channel geometry responses to perturbations in sediment supply and discharge. Changes in both channel and valley storage are also likely responses, as well as changes in channel roughness due to alteration of channel sinuosity and bedforms. There is less potential for textural response in dune-ripple than in pool-riffle and plane-bed channels simply because of smaller and more uniform grain sizes. At very high sediment supply, any of the above channel types may acquire a braided morphology (e.g., Mollard, 1973; Church, 1992). The general progression of alluvial channel types downstream through a channel network (Fig. 4) suggests that there is a systematic downstream increase in response potential to altered sediment supply or discharge.
The above predictions of response potential are largely conceptual, based on typical reach processes, characteristics, and locations within a drainage basin. Nevertheless, our approach provides a rational, process-based alternative to channel assessments based solely on descriptive typologic classification. For example, a channel-reach classification developed by Rosgen (1994) recognizes 7 major and 42 minor channel types primarily on the basis of bed material and slope; there is also the option of more detailed classification using entrenchment, sinuosity, width to depth ratio, and geomor-phic environments. However, the classification lacks a basis in channel processes. The lack of an explanation of the rationale underlying Rosgen's (1994) assessment of response potential for each minor channel type emphasizes this shortcoming. Furthermore, Rosgen's (1994) classification combines reach morphologies that may have very different response potentials: Rosgen's (1994) C channels may include reaches with dune-ripple, pool-riffle, plane-bed, or forced pool-riffle morphologies; his B channels may include plane-bed, forced-pool riffle, and step-pool morphologies; and his A channels may include colluvial, cascade, and step-pool reaches. Although bed material and slope provide a convenient classification for many channels, the lack of a process-based methodology compromises such an approach to structuring channel assessments, predicting channel response, and investigating relations to ecological processes.
External Influences
Channel response potential also reflects external influences on channel morphology, the most prominent of which are confinement, riparian vegetation, and large woody debris loading. Valley-wall confinement limits changes in both channel width and flood-plain storage and maximizes channel response to increased discharge by limiting overbank flow. Although there is a general downstream correspondence between channel type and valley-wall confinement in many mountain watersheds, structural controls and geomorphic history can force confinement in any portion of the channel network.
Riparian vegetation influences channel morphology and response potential by providing root strength that contributes to bank stability (e.g., Shaler, 1891; Gilbert, 1914), especially in relatively noncohesive alluvial deposits. The effect of root strength on channel bank stability is greatest in low-gradient, unconfined reaches, where loss of bank reinforcement may result in dramatic channel widening (Smith, 1976). Riparian vegetation is also an important roughness source (e.g., Arcement and Schneider, 1989) that can mitigate the erosive action of high discharges.
Large woody debris provides significant control on the formation and physical characteristics of pools, bars, and steps (Heede, 1985; Lisle, 1986; Montgomery et al., 1995; Wood-Smith and Buffington, 1996), thereby influencing channel type and the potential for change in sediment storage and bedform roughness in response to altered sediment supply, discharge, or large woody debris loading. Woody debris may decrease the potential for channel widening by armoring stream banks; alternatively, it may aid bank erosion by directing flow and scour toward channel margins. Furthermore, bed-surface textures and their response potential are strongly controlled by hydraulic roughness resulting from in-channel wood and debris-forced bed-forms (Buffington, 1995). Although large woody debris can force morphologic changes ranging from the scale of channel units to reaches, its impact depends on the amount, size, orientation, and position of debris, as well as channel size (Bilby and Ward, 1989; Montgomery et al., 1995) and rates of debris recruitment, transport, and decay (Bryant, 1980; Murphy and Koski, 1989). In general, individual pieces of wood can dominate the morphology of small channels, whereas debris jams are required to significantly influence channel morphology in larger rivers where individual pieces are mobile (Abbe and Montgomery, 1996). Thus, the relative importance of large woody debris in controlling channel morphology and response potential varies through a channel network.
Temporal Changes in Channel Morphology
The spatial pattern of channel types within a watershed provides a snapshot in time of a channel network, but history also influences the response potential of mountain channels, because past disturbance can condition channel response. Temporal variations in macroscopic channel morphology reflect (1) changes in large woody debris loading (e.g., Beschta, 1979;
608
Geological Society of America Bulletin, May 1997
CHANNEL-REACH MORPHOLOGY IN MOUNTAIN BASINS
Heede, 1985); (2) changes in discharge and sediment input (e.g., Hammer, 1972; Graf, 1975; Megahan et al., 1980; Coats et al., 1985); and (3) routing of sediment waves through the channel network (e.g., Gilbert, 1917; Kelsey, 1980; Church and Jones, 1982; Madej, 1982; Reid, 1982; Beschta, 1983).
Channels in which large woody debris forces pool formation and sediment storage are particularly sensitive to altered wood loading. For example, removal of large woody debris from forced pool-riffle channels may lead to either a pool-riffle or plane-bed morphology (Montgomery et al., 1995). Similarly, loss of large woody debris may transform a forced step-pool channel into a step-pool, cascade, or bedrock channel, depending on channel slope, discharge, and availability of coarse sediment.
Changes in reach-level channel type resulting from increased sediment supply typically represent a transient response to a pulsed input, although a longer-term response may result from sustained inputs. A landslide-related pulse of sediment may result in a transient change to a morphology with a lower qľ that subsequently relaxes toward the original morphology as the perturbation subsides. Pool-riffle reaches, for example, can develop a braided morphology while transmitting a pulse of sediment and subsequently revert to a single-thread pool-riffle morphology. Channel reaches with high qľ should recover quickly from increased sediment loading, because they are able to rapidly transport the load downslope. Reaches with a low qľ should exhibit more persistent morphologic response to a comparable increase in sediment supply. Transient morphologic change can also result from debris-flow scour of steep-gradient channels. For example, collu-vial and cascade channels that are scoured to bedrock by a debris flow may slowly revert to their predisturbance morphologies.
The spatial pattern of channel types provides a template against which to assess channel response potential, but the disturbance history of a channel network also is important for understanding both current conditions and response potential. Reach-level channel morphology provides a general indication of differences in response potential, but specific responses depend on the nature, magnitude, and persistence of disturbance, as well as on local conditions, including riparian vegetation, in-channel large woody debris, bank materials, and the history of catastrophic events. Furthermore, concurrent multiple perturbations can cause opposing or constructive response, depending on both channel type and the direction and magnitude of change. Hence, assessment of either present channel conditions or the potential for future impacts in mountain drainage basins should consider both disturbance history and the influences of channel morphology, position in the network, and local external constraints.
CONCLUSIONS
Systematic variations in bed morphology in mountain drainage basins provide the basis for a classification of channel-reach morphology that reflects channel-forming processes, serves to illustrate process linkages within the channel network, and allows prediction of general channel response potential. The underlying hypothesis that alluvial bed morphology reflects a stable roughness configuration for the imposed sediment supply and transport capacity implies a fundamental link between channel processes and form. The association of reach types and ratios of transport capacity to sediment supply combined with identification of external influences and the spatial coupling of reaches with hillslopes and other channel types provides a conceptual framework within which to investigate channel processes, assess channel conditions, and examine spatially distributed responses to watershed disturbance. Integration of this approach into region-specific landform and valley segment classifications would provide a common language to studies of fluvial processes and response to disturbance. This classification, however, is not ideal for all purposes; characterization of river planforms, for example, is useful for classifying flood-plain rivers. The development of specific
restoration designs requires further information on reach-specific characteristics. Our classification simply characterizes aspects of reach-level channel morphology useful for assessing channel condition and potential response to natural and anthropogenic disturbance in mountain drainage basins.
ACKNOWLEDGMENTS
This research was supported by the Sediment, Hydrology, and Mass Wasting Committee of the Washington State Timber-Fish-Wildlife agreement through grant FY95-156 and by the U.S. Forest Service through cooperative research agreements PNW 93-0441 and 94-0617. Tamara Mas-song, Carolyn Trayler, and Matt Coglon provided assistance in the field. We thank Jim Knox, Gordon Grant, and Andrew Marcus for insightful reviews of the manuscript, and Mike Church for thorough critiques that sharpened the discussion.
REFERENCES CITED
Abbe, T. B., and Montgomery, D. R., 1996, Large woody debris jams, channel hydraulics and habitat formation in large rivers: Regulated rivers: Research and Management, v. 12, p. 201-221.
Abrahams, A. D., Li, G., and Atkinson, J. E, 1995, Step-pool streams: Adjustment to maximum flow resistance: Water Resources Research, v. 31, p. 2593-2602.
Andrews, E. D., 1984, Bed material entrainment and hydraulic geometry of gravel-bed rivers in Colorado: Geological Society of America Bulletin, v. 95, p. 371-378.
Arcement, G J., and Schneider, V. R., 1989, Guide for selecting Manning's roughness coefficients for natural channels and flood plains: U.S. Geological Survey Water-Supply Paper 2339, 38 p.
Ashida, K.,Takahashi, T., andSawada, T., 1976, Sediment yield and transport on a mountainous small watershed: Bulletin of the Disaster Prevention Research Institute, v. 26, p. 119-144.
Ashida, K., Takahashi, T., and Sawada, T., 1981, Processes of sediment transport in mountain stream channels, in Erosion and sediment transport in pacific rim steeplands: International Association of Hydrological Sciences Publication 132, p. 166-178.
Barnes, H. H., 1967, Roughness characteristics of natural channels: U.S. Geological Survey Water-Supply Paper 1849, 213 p.
Bathurst, J. C, Graf, W H., and Cao, H. H., 1983, Bedforms and flow resistance in steep gravel-bed channels, in Mutlu Sumer, M., and Muller, A., eds., Mechanics of sediment transport: Rotterdam, Netherlands, A A. Balkema, p. 215-221.
Benda, L., 1990, The influence of debris flows on channels and valley floors in the Oregon coast range, USA: Earth Surface Processes and Landforms, v. 15, p. 457^166.
Benda, L., and Dunne, T., 1987, Sediment routing by debris flows, in Beschta, R. L., Blinn, R., Grant, G E., Ice, G., and Swanson, F. I., eds., Erosion and sedimentation in the Pacific rim: International Association of Hydrological Sciences Publication 165, p. 213-223.
Beschta, R. L., 1979, Debris removal and its effects on sedimentation in an Oregon Coast Range stream: Northwest Science, v. 53, p. 71-77.
Beschta, R. L., 1983, Channel changes following storm-induced hillslope erosion in the upper Kowai basin, Torlesse Range, New Zealand: New Zealand Journal of Hydrology, v. 22, p. 93-111.
Bilby, R E., and Ward, J. W, 1989, Changes in characteristics and function of woody debris with increasing size of streams in western Washington: Transactions of the American Fisheries Society, v. 118, p. 368-378.
Bisson, P. A., Nielsen, J. L.,Palmason, R. A., and Grove, L E., 1982, A system of naming habitat types in small streams, with examples of habitat utilization by salmonids during low streamflow, in Armantrout, N. B., ed., Proceedings of a Symposium on Acquisition and Utilization of Aquatic Habitat Inventory Information: Portland, Oregon, Western Division of the American Fisheries Society, p. 62-73.
Bowman, D., 1977, Stepped-bed morphology in arid gravelly channels: Geological Society of America Bulletin, v. 88, p. 291-298.
Bryant, M. D., 1980, Evolution of large, organic debris after timber harvest: Maybeso Creek, 1949 to 1978: Portland, Oregon, Pacific Northwest Forest and Range Experiment Station, U.S Department of Agriculture, Forest Service General Technical Report PNW-101, 30 p.
Buffington, J. M., 1995, Effects of hydraulic roughness and sediment supply on surface textures of gravel-bedded rivers [master's thesis]: Seattle, University of Washington, 184 p.
Buffington, J. M., Dietrich, W. E., and Kirchner, J. W, 1992, Friction angle measurements on a naturally formed gravel streambed: Implications for critical boundary shear stress: Water Resources Research, v. 28, p. 411^125.
Carling, P., 1988, The concept of dominant discharge applied to two gravel-bed streams in relation to channel stability thresholds: Earth Surface Processes and Landforms, v. 13, p. 355-367.
Chin, A., 1989, Step pools in stream channels: Progress in Physical Geography, v. 13, p. 391-407.
Church, M., 1992, Channel morphology and typology, in Carlow, P., and Petts, G E., eds., The rivers handbook: Oxford, United Kingdom, Blackwell Scientific Publications, p. 126-143.
Church, M., and Jones, D., 1982, Channel bars in gravel-bed rivers, in Hey, R. D., Bathurst, J. D., and Thorne, C R., eds., Gravel-bed rivers: Fluvial processes, engineering and management: Chichester, United Kingdom, John Wiley and Sons, p. 291-338.
Geological Society of America Bulletin, May 1997
609
MONTGOMERY AND BUFFINGTON
Coats, R., Collins, L., Florsheim, J., and Kaufman, D., 1985, Channel change, sediment transport, and fish habitat in a coastal stream: Effects of an extreme event: Environmental Management, v. 9, p. 35—48.
Dana, J. D., 1850, On denudation in the Pacific: American Journal of Science, ser. 2, v. 9, p. 48-62.
Dietrich, W. E., and Dunne, T., 1978, Sediment budget for a small catchment in mountainous terrain: Zeitschrift für Geomorphologie, Supplementband 29, p. 191-206.
Dietrich, W. E., and Smith, J. D., 1983, Influence of the point bar onflow through curved channels: Water Resources Research, v. 19,p. 1173-1192.
Dietrich, W. E., and Whiting, R, 1989, Boundary shear stress and sediment transport in river meanders of sand and gravel, in Ikeda, S., and Parker, G., eds., River meandering: American Geophysical Union Water Resources Monograph 12, p. 1-50.
Dietrich, W. E., Dunne, T., Humphrey, N., andReid, L., 1982, Construction of sediment budgets for drainage basins, in Swanson, F. J., Janda, R. J., Dunne, T., and Swanston, D. N., eds., Sediment budgets and routing in forested drainage basins: Portland, Oregon, Pacific Northwest Forest and Range Experiment Station, U.S. Department of Agriculture, Forest Service, General Technical Report PNW-141, p. 2-23.
Dietrich, W. E., Kirchner, J. W, Ikeda, H., and Iseya, F, 1989, Sediment supply and the development of the coarse surface layer in gravel-bedded rivers: Nature, v. 340, p. 215-217.
Dinehart, R. L., 1992, Evolution of coarse-gravel bedforms: Field measurements at flood stage: Water Resources Research, v. 28, p. 2667-2689.
Emmett, W. W, 1984, Measurement of bedload in rivers, in Hadley, R. F, and Walling, D. E., eds., Erosion and sediment yield: Some methods of measurement and modeling: Norwich, United Kingdom, GeoBooks, p. 91-109.
Ergenzinger, P., and Schmidt, K.-H., 1990, Stochastic elements of bed load transport in a step-pool mountain river, in Sinniger, R. O., and Monbaron, M., eds., Hydrology in mountainous regions. II—Artificial reservoirs, water and slopes: International Association of Hy-drological Sciences Publication 194, p. 39—46.
Fenton, J. D., and Abbott, J. E., 1977, Initial movement of grains in a stream bed: The effects of relative protrusion: Proceedings of the Royal Society of London, v. 352A, p. 532-537.
Florsheim, J. L., 1985, Fluvial requirements for gravel bar formation in northwestern California [master's thesis]: Areata, California, Humboldt State University, 105 p.
Frisseil, C. A., 1993, Topology of extinction and endangerment of native fishes in the Pacific Northwest and California (U.S.A.): Conservation Biology, v. 7, p. 342-354.
Frisseil, C. A., Liss, W. J., Warren, C. E., and Hurley, M. D., 1986, A hierarchical framework for stream habitat classification: Viewing streams in a watershed context: Environmental Management, v. 10, p. 199-214.
Gilbert, G. K, 1914, The transportation of debris by running water: U.S. Geological Survey Professional Paper 86, 263 p.
Gilbert, G. K, 1917, Hydraulic-mining debris in the Sierra Nevada: U.S. Geological Survey Professional Paper 105,154 p.
Graf, W. L., 1975, The impact of suburbanization on fluvial geomorphology: Water Resources Research, v. 11, p. 690-692.
Grant, G. E., and Mizuyama, T, 1991, Origin of step-pool sequences in high gradient streams: A flume experiment, in Proceedings of the Japan-U.S. workshop on snow avalanche: Landslide, debris flow prediction and control: Tskuba, Japan, Organizing Committee of the Japan-U.S. Workshop on Snow Avalanche, Landslide, Debris Flow Prediction and Control, p. 523-532.
Grant, G. E., Swanson, F. J., and Wolman, M. G., 1990, Pattern and origin of stepped-bed morphology in high-gradient streams, Western Cascades, Oregon: Geological Society of America Bulletin, v. 102, p. 340-352.
Griffiths, G. A., 1980, Stochastic estimation of bed load yield in pool-and-riffle mountain streams: Water Resources Research, v. 16, p. 931-937.
Griffiths, G. A., 1989, Form resistance in gravel channels with mobile beds: Journal of Hydraulic Engineering, v. 115, p. 340-355.
Hammer, T. R., 1972, Stream channel enlargement due to urbanization: Water Resources Research, v. 8, p. 1530-1540.
Hammond, F. D. C, Heathershaw, A. D., and Langhorne, D. N., 1984, A comparison between Shields' threshold criterion and the movement of loosely packed gravel in a tidal channel: Sedimentology, v. 31, p. 51-62.
Harms, J. C, Southard, J. B., Spearing, D. R., and Walker, R. G., 1975, Depositional environments as interpreted from primary sedimentary structures and stratification sequences: Society for Economic Paleontologists and Mineralogists Short Course 2,161 p.
Heede, B. H., 1985, Channel adjustments to the removal of log steps: An experiment in a mountain stream: Environmental Management, v. 9, p. 427—432.
Henderson, F. M., 1963, Stability of alluvial channels: Transactions of the American Society of Civil Engineers, v. 128, p. 657-686.
Howard, A. D., andKerby, G., 1983, Channel changes in badlands: Geological Society of America Bulletin, v. 94, p. 739-752.
Howard, A. D., Dietrich, W. E., and Seidl, M. A., 1994, Modeling fluvial erosion on regional to continental scales: Journal of Geophysical Research, v. 99, p. 13971-13986.
Ikeda, H., 1975, On the bed configuration in alluvial channels: Their types and condition of formation with reference to bars: Geographical Review of Japan, v. 48, p. 712-730.
Ikeda, H., 1977, On the origin of bars in the meandering channels: Bulletin of the Environmental Research Center, University of Tsukuba, v. 1, p. 17-31.
Jackson, W. L., andBeschta, R. L., 1982, A model of two-phase bedload transport in an Oregon coastrange stream: Earth Surface Processes and Landforms, v. 7, p. 517-527.
Keller, E. A., and Melhorn, W. N., 1978, Rhythmic spacing and origin of pools and riffles: Geological Society of America Bulletin, v. 89, p. 723-730.
Kelsey, H. M., 1980, A sediment budget and an analysis of geomorphic process in the Van Duzen River basin, north coastal California, 1941-1975: Geological Society of America Bulletin, v. 91, p. 1119-1216.
Kennedy, J. F, 1975, Hydraulic relations for alluvial streams, in Vanoni, V, ed., Sedimentation
engineering: American Society of Civil Engineers Manual 54, p. 114—154. Kirchner, J., Dietrich, W. E., Iseya, F, and Ikeda, H., 1990, The variability of critical boundary shear stress, friction angle, and grain protrusion in water-worked sediments: Sedimentology, v. 37, p. 647-672.
Komar, P. D., and Li., Z., 1986, Pivoting analyses of the selective entrainment of sediments by size and shape with application to gravel threshold: Sedimentology, v. 33, p. 425-436.
Kondolf, G. M., 1995, Geomorphological stream channel classification in aquatic habitat restoration: Uses and limitations: Aquatic conservation: Marine and Freshwater Ecosystems, v. 5, p. 127-141.
Kondolf, G. M., Cada, G. F, Sale, M. J., and Felando, T, 1991, Distribution and stability of potential salmonid spawning gravels in steep boulder-bed streams of the eastern Sierra Nevada: Transactions of the American Fisheries Society, v. 120, p. 177-186.
Lane, E. W, and Carlson, E. J., 1953, Some factors affecting the stability of canals constructed in coarse granular materials: Proceedings of the Minnesota International Hydraulics Convention, International Association for Hydraulic Research and American Society of Civil Engineers, p. 37—48.
Langbein, W. B., and Leopold, L. B., 1968, River channel bars and dunes—Theory of kinematic waves: U.S. Geological Survey Professional Paper 422-L, 20 p.
Lehre, A. K, 1982, Sediment budget of a small Coast Range drainage basin in north-central California, in Swanson, F. J., Janda, R J., Dunne, T, and Swanston, D. N., eds., Sediment budgets and routing in forested drainage basins: Portland, Oregon, Pacific Northwest Forest and Range Experiment Station, U.S. Department of Agriculture, Forest Service General Technical Report PNW-141, p. 67-77.
Leopold, L. B., Wolman, M. G., and Miller, J. P., 1964, Fluvial processes in geomorphology: San Francisco, California, W. H. Freeman, 522 p.
Lindley, E. S., 1919, Regime channels: Proceedings of the Punjab Engineering Congress, v. 7, p. 63-74.
Lisle, T. E., 1986, Stabilization of a gravel channel by large streamside obstructions and bedrock bends, Jacoby Creek, northwestern California: Geological Society of America Bulletin, v. 97, p. 999-1011.
Lisle, T. E., Ikeda, H., and Iseya, F, 1991, Formation of stationary alternate bars in a steep channel with mixed-size sediment: A flume experiment: Earth Surface Processes and Land-forms, v. 16, p. 463^169.
Madej, M. A., 1982, Sediment transport and channel changes in an aggrading stream in the Puget Lowland, Washington, in Swanson, F. J., Janda, R. J., Dunne, T, and Swanston, D. N., eds., Sediment budgets and routing in forested drainage basins: Portland, Oregon, Pacific Northwest Forest and Range Experiment Station, U.S. Department of Agriculture, Forest Service General Technical Report PNW-141, p. 97-108.
Marcus, W. A., Roberts, K, Harvey, L., and Tackman, G., 1992, An evaluation of methods for estimating Manning's n in small mountain streams: Mountain Research and Development, v. 12, p. 227-239.
McDonald, B. C, and Banerjee, I., 1971, Sediments and bed forms on a braided outwash plain:
Canadian Journal of Earth Sciences, v. 8, p. 1282-1301. Megahan, W, Platts, W. S., andKulesza,B., 1980, Riverbed improves through time: South Fork
Salmon River, in Symposium on watershed management: New York, American Society of
Civil Engineers, p. 380-395. Middleton, G. V, and Southard, J. B., 1984, Mechanics of sediment movement: Society of Economic Paleontologists and Mineralogists Short Course 3,401 p. Milhous, R. T, 1973, Sediment transport in a gravel-bottom stream [Ph.D. dissert.]: Corvallis,
Oregon State University, 232 p. Milliman, J. D., and Syvitski, J. P. M., 1992, Geomorphic/tectonic control of sediment discharge
to the ocean: The importance of small mountainous rivers: Journal of Geology, v. 100,
p. 525-544.
Mollard, J. D., 1973, Air photo interpretation of fluvial features: Edmonton, Canada, Proceedings of the 9th Canadian Hydrology Symposium, p. 341-380.
Montgomery, D. R., 1991, Channel initiation and landscape evolution [Ph.D. dissert.]: Berkeley, University of California, 421 p.
Montgomery, D. R., and Buffington, J. M., 1993, Channel classification, prediction of channel response, and assessment of channel condition: Olympia, Washington State Department of Natural Resources Report TFW-SH10-93-002, 84 p.
Montgomery, D. R., and Foufoula-Georgiou, E., 1993, Channel network source representation using digital elevation models: Water Resources Research, v. 29, p. 3925-3934.
Montgomery, D. R., Buffington, J. M., Smith, R. D., Schmidt, K M., and Pess, G., 1995, Pool spacing in forest channels: Water Resources Research, v. 31, p. 1097-1105.
Montgomery, D. R., Abbe, T. B., Buffington, J. M., Peterson, N. P., Schmidt, K M., and Stock, J. D., 1996, Distribution of bedrock and alluvial channels in forested mountain drainage basins: Nature, v. 381, p. 587-589.
Murphy, M. L, and Koski, K V, 1989, Input and depletion of woody debris in Alaska streams and implications for streamside management: North American Journal of Fisheries Management, v. 9, p. 427^136.
Nanson, G. C, 1974, Bedload and suspended-load transport in a small, steep, mountain stream: American Journal of Science, v. 274, p. 471—486.
Nehlsen, W, Williams, J. E., and Lichatowich, J. A., 1991, Pacific salmon at the crossroads: Stocks at risk from California, Oregon, Idaho, and Washington: Fisheries, v. 16, p. 4-21.
Nelson, J. M., and Smith, J. D., 1989, Evolution and stability of erodible channel beds, in Ikeda, S., and Parker, G., eds., River meandering: American Geophysical Union Water Resources Monograph 12, p. 321-377.
O'Neill, M. P., and Abrahams, A. D., 1984, Objective identification of pools and riffles: Water Resources Research, v. 20, p. 921-926.
Parker, G., Klingeman, P. C, and McLean, D. G., 1982, Bedload size and distribution in paved gravel-bed streams: Journal of the Hydraulics Division, American Society of Civil Engineers, v. 108, p. 544-571.
610
Geological Society of America Bulletin, May 1997
CHANNEL-REACH MORPHOLOGY IN MOUNTAIN BASINS
Paustian, S. J., and 13 others, 1992, A channel type users guide for the Tongass National Forest, Southeast Alaska: U.S. Department of Agriculture Forest Service, Alaska Region RIO Technical Paper 26,179 p.
Peterson, D. F, and Mohanty, P. K., 1960, Rume studies of flow in steep, rough channels: Journal of the Hydraulics Division, American Society of Civil Engineers, v. 86, p. 55-76.
Pitlick, J., 1992, Row resistance under conditions of intense gravel transport: Water Resources Research, v. 28, p. 891-903.
Reid, L, Frostick, L. E., and Layman, J. T, 1985, The incidence and nature of bedload transport during flood flows in coarse-grained alluvial channels: Earth Surface Processes and Land-forms, v. 10, p. 33^14.
Reid, L., 1982, Evaluating and mapping sources and temporary storage areas of sediment, in Sediment budgets and routing in forested drainage basins: Portland, Oregon, Pacific Northwest Forest and Range Experiment Station, U.S. Department of Agriculture, Forest Service General Technical Report PNW-141, p. 138-142.
Reid, L. M., 1993, Research and cumulative watershed effects, Berkeley, California, Pacific Southwest Research Station: U.S. Department of Agriculture, Forest Service General Technical Report PSW-GTR-141,118 p.
Rice, S., 1994, Towards a model of changes in bed material texture at the drainage basin scale, in Kirkby, M. J., ed., Process models and theoretical geomorphology: Chichester, United Kingdom, Wiley and Sons, p. 158-172.
Rosgen, D. L., 1994, A classification of natural rivers: Catena, v. 22, p. 169-199.
Sawada, T, Ashida, K., and Takahashi, T, 1983, Relationship between channel pattern and sediment transport in a steep gravel bed river: Zeitschrift fur Geomorphologie, Supplement-band 46, p. 55-66.
Schmidt, K.-H., andErgenzinger,P, 1992, Bedload entrainment, travel lengths, step lengths, rest
periods—Studied with passive (iron, magnetic) and active (radio) tracer techniques: Earth
Surface Processes and Landforms, v . 17, p. 147-165. Schumm, S. A., 1977, The fluvial system: New York, John Wiley and Sons, 338 p. Seidl, M., and Dietrich, W. E., 1992, The problem of channel incision into bedrock, in Schmidt,
K.-H., and de Ploey, J., eds., Functional geomorphology, Catena Supplement 23: Crem-
lingen, Germany, Catena Verlag, p. 101-124. Shaler, N. S., 1891, The origin and nature of soils: U.S. Geological Survey 12th Annual Report,
p. 213-345.
Shaw, J., and Kellerhals, R., 1977, Paleohydraulic interpretations of antidune bedforms with applications to antidunes in gravel: Journal of Sedimentary Petrology, v. 47, p. 257-266.
Sidle, R. C, 1988, Bed load transport regime of a small forest stream: Water Resources Research, v. 24, p. 207-218.
Simons, D. B., Richardson, E. V., and Nordin, C. F, 1965, Sedimentary structures generated by flow in alluvial channels, in Middleton, G. V, ed., Primary sedimentary structures and their
hydrodynamic interpretation: Tulsa, Oklahoma, Society of Economic Paleontologists and Mineralogists, p. 34-52.
Smith, D. G., 1976, Effect of vegetation on lateral migration of anastomosed channels of a glacier meltwater river: Geological Society of America Bulletin, v. 87, p. 857-860.
Sukegawa, N., 1973, Condition for the formation of alternate bars in straight alluvial channels, in Proceedings of the international symposium on river mechanics: Bangkok, Thailand, International Association for Hydraulic Research, A58-1-A58-11.
Surell, A., 1841, Etude sur les torrents des Hautes-Alpes: Paris, France.
Swanson,F. J.,Fredriksen,R L.,andMcCorison,F. M., 1982, Material transfer in a western Oregon forested watershed, in Edmonds, R. L., ed., Analysis of coniferous forest ecosystems in the western United States: Stroudsburg, Pennsylvania, Hutchison Ross Publishing, p. 233-266.
Warburton, J., 1992, Observations of bedload transport and channel bed changes in a proglacial mountain stream: Arctic and Alpine Research, v. 24, p. 195-203.
Whiting, P. J., and Bradley, J. B., 1993, A process-based classification for headwater streams: Earth Surface Processes and Landforms, v. 18, p. 603-612.
Whiting, P. J., Dietrich, W. E., Leopold, L. B., Drake, T. G., and Shreve, R. L., 1988, Bedload sheets in heterogeneous sediment: Geology, v. 16, p. 105-108.
Whittaker, J. G., 1987a, Modeling bed-load transport in steep mountain streams, in Beschta, R L., Blinn, T, Grant, G. E., Ice, G. G., and Swanson, F. J., eds., Erosion and sedimentation in the Pacific rim: International Association of Hydrological Sciences Publication 165, p. 319-332.
Whittaker, J. G., 1987b, Sediment transport in step-pool streams, in Thorne, C. R., Bathurst, J. C, and Hey, R. D., eds., Sediment transport in gravel-bed rivers: Chichester, United Kingdom, John Wiley and Sons, p. 545-579.
Whittaker, J. G., and Davies, T. R. H., 1982, Erosion and sediment transport processes in step-pool torrents in Walling, D. E., ed., Recent developments in the explanation and prediction of erosion and sediment yield: International Association of Hydrological Sciences Publication 137, p. 99-104.
Whittaker, J. G., and Jaeggi, M. N. R., 1982, Origin of step-pool systems in mountain streams: Journal of the Hydraulics Division: Proceedings of the American Society of Civil Engineers, v. 108, p. 99-104.
Wolman, M. G., 1954, A method of sampling coarse bed material: Transactions, American Geophysical Union, v. 35, p. 951-956.
Wood-Smith, R. D., and Buffington, J. M., 1996, Multivariate geomorphic analysis of forest streams: Implications for assessment of land use impact on channel condition: Earth Surface Processes and Landforms, v. 21, p. 377-393.
Manuscript Received by the Society July 20,1995 Revised Manuscript Received August 12,1996 Manuscript Accepted October 3,1996
Printed in u.s.a.
Geological Society of America Bulletin, May 1997 611