Content uploaded by george 'mathias Kondolf
Author content
All content in this area was uploaded by george 'mathias Kondolf on Jan 29, 2015
Content may be subject to copyright.
PROFILE
Hungry Water: Effects of Dams and Gravel Mining
on River Channels
G. MATHIAS KONDOLF
Department of Landscape Architecture and Environmental
Planning
University of California
Berkeley, California 94720, USA
www.ced.berkeley.edu/,kondolf/
ABSTRACT / Rivers transport sediment from eroding up-
lands to depositional areas near sea level. If the continuity of
sediment transport is interrupted by dams or removal of
sediment from the channel by gravel mining, the flow may
become sediment-starved (hungry water) and prone to
erode the channel bed and banks, producing channel inci-
sion (downcutting), coarsening of bed material, and loss of
spawning gravels for salmon and trout (as smaller gravels
are transported without replacement from upstream). Gravel
is artificially added to the River Rhine to prevent further inci-
sion and to many other rivers in attempts to restore spawning
habitat. It is possible to pass incoming sediment through
some small reservoirs, thereby maintaining the continuity of
sediment transport through the system. Damming and min-
ing have reduced sediment delivery from rivers to many
coastal areas, leading to accelerated beach erosion. Sand
and gravel are mined for construction aggregate from river
channel and floodplains. In-channel mining commonly
causes incision, which may propagate up- and downstream
of the mine, undermining bridges, inducing channel instabil-
ity, and lowering alluvial water tables. Floodplain gravel pits
have the potential to become wildlife habitat upon reclama-
tion, but may be captured by the active channel and thereby
become instream pits. Management of sand and gravel in
rivers must be done on a regional basis, restoring the conti-
nuity of sediment transport where possible and encouraging
alternatives to river-derived aggregate sources.
As waters flow from high elevation to sea level, their
potential energy is converted to other forms as they
sculpt the landscape, developing complex channel
networks and a variety of associated habitats. Rivers
accomplish their geomorphic work using excess energy
above that required to simply move water from one
point on the landscape to another. In natural channels,
the excess energy of rivers is dissipated in many ways: in
turbulence at steps in the river profile, in the frictional
resistance of cobbles and boulders, vegetation along
the bank, in bends, in irregularities of the channel bed
and banks, and in sediment transport (Figure 1).
The transport of sand- and gravel-sized sediment is
particularly important in determining channel form,
and a reduction in the supply of these sediments may
induce channel changes. The supply of sand and gravel
may be the result of many factors, including changes in
land use, vegetation, climate, and tectonic activity. This
paper is concerned specifically with the response of
river channels to a reduction in the supply of these
sediments by dams and gravel mining.
Sediment is transported mostly as suspended load:
clay, silt, and sand held aloft in the water column by
turbulence,incontrasttobedload:sand,gravel,cobbles,
andboulders transported by rolling, sliding, and bounc-
ing along the bed (Leopold and others 1964). Bedload
ranges from a few percent of total load in lowland rivers
to perhaps 15% in mountain rivers (Collins and Dunne
1990), to over 60% in some arid catchments (Schick
and Lekach 1993). Although a relatively small part of
the total sediment load, the arrangement of bedload
sediments constitutes the architecture of sand- and
gravel-bed channels. Moreover, gravel and cobbles have
tremendous ecological importance, as habitat for ben-
thic macroinvertebrates and as spawning habitat for
salmon and trout (Kondolf and Wolman 1993).
The rate of sediment transport typically increases as
a power function of flow; that is, a doubling of flow
typically produces more than a doubling in sediment
transport (Richards 1982), and most sediment trans-
port occurs during floods.
Continuity of Sediment Transport
in River Systems
Viewed over a long term, runoff erodes the land
surface, and the river network carries the erosional
products from each basin. The rates of denudation, or
lowering of the land by erosion, range widely. The
Appalachian Mountains of North America are being
denudedabout0.01mm/yr(Leopoldand others 1964),
the central Sierra Nevada of California about 0.1
KEY WORDS: Dams; Aquatic habitat; Sediment transport; Erosion;
Sedimentation; Gravel mining
Environmental Management Vol. 21, No. 4, pp. 533–551 r1997 Springer-Verlag New York Inc.
mm/yr (Kondolf and Matthews 1993), the Southern
Alps of New Zealand about 11 mm/yr (Griffiths and
McSaveney 1983), and the southern Central Range of
Taiwan over 20 mm/yr (Hwang 1994). The idealized
watershed can be divided into three zones: that of
erosion or sediment production (steep, rapidly eroding
headwaters), transport (through which sediment is
moved more or less without net gain or loss), and
deposition (Schumm 1977) (Figure 2). The river chan-
nel in the transport reach can be viewed as a conveyor
belt, which transports the erosional products down-
stream to the ultimate depositional sites below sea level.
The size of sediment typically changes along the length
of the river system from gravel, cobbles, and boulders in
steep upper reaches to sands and silts in low-gradient
downstream reaches, reflecting diminution in size by
Figure 1. Diagram of energy dissipation in
river channels.
Figure 2. Zones of erosion, transport, and deposition, and the river channel as conveyor belt for sediment. (Reprinted from
Kondolf 1994, with kind permission of Elsevier Science-NL.)
G. M. Kondolf
534
weathering and abrasion, as well as sorting of sizes by
flowing water.
Transport of sediment through the catchment and
along the length of the river system is continuous.
Increased erosion in the upper reaches of the catch-
ment can affect the river environment many miles
downstream (and for years or decades) as the increased
sediment loads propagate downstream through the
river network. On Redwood Creek in Redwood Na-
tional Park, California, the world’s tallest trees are
threatened with bank erosion caused by channel aggra-
dation (building up of sediment in the channel), which
in turn was caused by clear-cutting of timber on steep
slopes in the upper part of the catchment (Madej and
Ozaki 1996, Janda 1978).
Along the river channel conveyor belt, channel
forms (such as gravel bars) may appear stable, but the
grains of which they are composed may be replaced
annually or biannually by new sediment from upstream.
Similarly, the sediments that make up the river flood-
plain (the valley flat adjacent to the channel) are
typically mobile on a time scale of decades or centuries.
The floodplain acts as a storage reservoir for sediments
transported in the channel, alternately storing sedi-
ments by deposition and releasing sediment to the
channel by bank erosion. For example, the Carmel
River, California, is flanked by flat surfaces (terraces)
that step up from the river. The lowest terrace is the
channel of sand and gravel deposited by the 1911 flood,
but the surface now stands about 4 m above the present,
incised channel (Kondolf and Curry 1986). By 1960,
the terrace had been subdivided for low-density hous-
ing, despite the recent origin of the land and the
potential for future shifts in channel position.
A river channel and floodplain are dynamic features
that constitute a single hydrologic and geomorphic unit
characterized by frequent transfers of water and sedi-
ment between the two components. The failure to
appreciate the integral connection between floodplain
and channel underlies many environmental problems
in river management today.
Effects of Dams
Dams and diversions are constructed and operated
for a wide variety of purposes including residential,
commercial,andagriculturalwatersupply;floodand/or
debris control; and hydropower production. Regardless
of their purpose, all dams trap sediment to some degree
and most alter the flood peaks and seasonal distribution
of flows, thereby profoundly changing the character
and functioning of rivers. By changing flow regime and
sediment load, dams can produce adjustments in allu-
vial channels, the nature of which depends upon the
characteristics of the original and altered flow regimes
and sediment loads.
Dams disrupt the longitudinal continuity of the river
system and interrupt the action of the conveyor belt of
sediment transport. Upstream of the dam, all bedload
sedimentandallorpart of the suspended load (depend-
ing upon the reservoir capacity relative to inflow)
(Brune 1953) is deposited in the quiet water of the
reservoir (reducing reser voir capacity) and upstream of
the reservoir in reaches influenced by backwater. Down-
stream, water released from the dam possesses the
energy to move sediment, but has little or no sediment
load. This clear water released from the dam is often
referred to as hungry water,because the excess energyis
typically expended on erosion of the channel bed and
banksforsomeyearsfollowingdam construction, result-
inginincision(downcutting of the bed) and coarsening
of the bed material until equilibrium is reached and the
material cannot be moved by the flows. Reservoirs also
may reduce flood peaks downstream, potentially reduc-
ing the effects of hungry water, inducing channel
shrinking, or allowing fine sediments to accumulate in
the bed.
Channel Incision
Incision below dams is most pronounced in rivers
with fine-grained bed materials and where impacts on
flood peaks are relatively minor (Williams and Wolman
1984). The magnitude of incision depends upon the
reservoir operation, channel characteristics, bed mate-
rial size, and the sequence of flood events following
dam closure. For example, the easily eroded sand bed
channel of the Colorado River below Davis Dam, Ari-
zona, has incised up to 6 m, despite substantial reduc-
tions in peak flows (Williams and Wolman 1984). In
contrast, the Mokelumne River below Camanche Dam
in California has experienced such a dramatic reduc-
tion in flood regime (and consequent reduction in
sediment transport capacity) that no incision has been
documented and gravels are reported to have become
compacted and immobile (FERC 1993).
Reduction in bedload sediment supply can induce a
change in channel pattern, as occurred on Stony Creek,
atributary totheSacramento River 200 km north of San
Francisco. Since the closure of Black Butte Dam in
1963, the formerly braided channel has adopted a
single-thread meandering pattern, incised, and mi-
grated laterally, eroding enough bedload sediment to
compensate for about 20% of the bedload now trapped
by Black Butte Dam on an annual average basis (Kon-
dolf and Swanson 1993).
Effects of Dams and Gravel Mining on Rivers 535
Bed Coarsening and Loss of Spawning Gravels
Channel erosion below dams is frequently accompa-
nied by a change in particle size on the bed, as gravels
and finer materials are winnowed from the bed and
transported downstream, leaving an armor layer, a
coarse lag deposit of large gravel, cobbles, or boulders.
Development of an armor layer is an adjustment by the
river to changed conditions because the larger particles
areless easily mobilized by the hungry water flows below
thedam. The armor layer maycontinue to coarsen until
the material is no longer capable of being moved by the
reservoir releases or spills, thereby limiting the ultimate
depth of incision (Williams and Wolman 1984, Dietrich
and others 1989).
The increase in particle size can threaten the success
of spawning by salmonids (salmon and trout), which
use freshwater gravels to incubate their eggs. The
female uses abrupt upward jerks of her tail to excavate a
small pit in the gravel bed, in which she deposits her
eggs and the male releases his milt. The female then
loosens gravels from the bed upstream to cover the eggs
and fill the pit. The completed nests (redds) constitute
incubation environments with intragravel flow of water
past the eggs and relative protection from predation.
The size of gravel that can be moved to create a redd
depends on the size of the fish, ranging in median
diameter from about 15 mm for small trout to about 50
mm for large salmon (Kondolf and Wolman 1993).
Below dams, the bed may coarsen to such an extent
that the fish can no longer move the gravel. The Upper
Sacramento River, California, was once the site of
extensive spawning by chinook salmon (Oncorhynchus
tshawytscha), but massive extraction of gravel from the
riverbed, combined with trapping of bedload sediment
behind Shasta Dam upstream and release of hungry
water, has resulted in coarsening of the bed such that
spawning habitat has been virtually eliminated in the
reach (Figure 3) (Parfitt and Buer 1980). The availabil-
ity of spawning gravels can also be reduced by incision
below dams when formerly submerged gravel beds are
isolated as terrace or floodplain deposits. Encroaching
vegetation can also stabilize banks and further reduce
gravel recruitment for redds (Hazel and others 1976).
Gravel Replenishment Below Dams
Gravels were being artificially added to enhance
available spawning gravel supply below dams on at least
13 rivers in California as of 1992 (Kondolf and Mat-
thews 1993). The largest of these efforts is on the Upper
Sacramento River, where from 1979 to 2000 over US$22
million will have been spent importing gravel (derived
mostly from gravel mines on tributaries) into the river
channel(Denton 1991) (Figure 4). While these projects
can provide short-term habitat, the amount of gravel
added is but a small fraction of the bedload deficit
below Shasta Dam, and gravels placed in the main river
havewashedoutduring high flows, requiring continued
addition of more imported gravel (California Depart-
ment of Water Resources 1995). On the Merced, Tu-
olumne,andStanislaus rivers in California, a total of ten
sites were excavated and back-filled with smaller gravel
to create spawning habitat for chinook salmon from
1990 to 1994. However, the gravel sizes imported were
mobile at high flows that could be expected to occur
every 1.5–4.0 years, and subsequent channel surveys
have demonstrated that imported gravels have washed
out (Kondolf and others 1996a,b).
On the border between France and Germany, a
series of hydroelectric dams was constructed on the
River Rhine (progressing downstream) after 1950, the
last of which (the Barrage Iffezheim) was completed in
the 1970s. To address the sediment deficit problem
downstream of Iffezheim, an annual average of 170,000
tonnes of gravel (the exact amount depending on the
Figure 3. Keswick Dam and the channel of the Sacramento
River downstream. (Photograph by the author, January 1989.)
G. M. Kondolf
536
magnitude of the year’s runoff) are added to the river
(Figure 5). This approach has proved successful in
preventing further incision of the riverbed downstream
(Kuhl 1992). It is worth noting that the quantity of
gravel added each year is not equivalent to the unregu-
lated sediment load of the Rhine; the river’s capacity to
transport sediment has also been reduced because the
peak discharges have been reduced by reservoir regula-
tion. The amount of sediment added satisfies the
transport capacity of the existing channel, which has
been highly altered for navigation and hydroelectric
generation.
Sediment Sluicing and Pass-Through
from Reservoirs
The downstream consequences of interrupting the
flux of sand and gravel transport would argue for
designing systems to pass sediment through reservoirs
(and thereby reestablish the continuity of sediment
transport). To date, most such efforts have been under-
taken to solve problems with reservoir sedimentation,
particularly deposits of sediment at tunnel intakes and
outlet structures, rather than to solve bedload sediment
supply problems downstream. These efforts have been
mostcommon in regions with high sediment yields such
as Asia (e.g., Sen and Srivastava 1995, Chongshan and
others 1995, Hassanzadeh 1995). Small diversion dams
(such as those used to divert water in run-of-the-river
hydroelectric generating projects) in steep V-shaped
canyons have the greatest potential to pass sediment.
Because of their small size, these reservoirs (or fore-
bays) can easily be drawn down so that the river’s
gradient and velocity are maintained through the dam
Figure 4. Gravel replenishment to
the Sacramento River below Keswick
Dam. (Photograph by the author,
January 1991.)
Figure 5. Bargeartificiallyfeedinggravelinto the River Rhine
downstream of the Barrage Iffezheim. (Photograph by author,
June 1994.)
Effects of Dams and Gravel Mining on Rivers 537
at high flow. Large-capacity, low-level outlets are re-
quired to pass the incoming flow and sediment load.
If low-level outlets are open at high flow and the
reservoir is drawn down, a small reservoir behaves
essentially as a reach of river, passing inflowing sedi-
ment through the dam outlets. In such a sediment
pass-through approach, the sediment is delivered to
downstream reaches in essentially the same concentra-
tion and seasonal flood flows as prevailed in the predam
regime. This approach was employed at the old Aswan
Dam on the River Nile and on the Bhatgurk Reservoir
on the Yeluard River in India (Stevens 1936). Similarly,
on the River Inn in Austria and Germany, floodwaters
with high suspended loads are passed through a series
of hydropower reservoirs in a channel along the reser-
voir bottom confined by training walls (Hack 1986,
Westrich and others 1992). If topographic conditions
are suitable, sediment-laden floodwater may be routed
around a reservoir in a diversion tunnel or permitted to
pass through the length of the reservoir as a density
current vented through a bottom sluice on the dam
(Morris 1993). The Nan-Hwa Reservoir in Taiwan was
designed with a smaller upstream forebay from which
sediment is flushed into a diversion tunnel, allowing
onlyrelatively clear water to pass into the mainreservoir
downstream (Morris 1993).
If sediment is permitted to accumulate in the reser-
voir and subsequently discharged as a pulse (sediment
sluicing), the abrupt increase in sediment load may
alter substrate and aquatic habitat conditions down-
stream of the dam. The most severe effects are likely to
occur when sediment accumulated over the flood sea-
son is discharged during baseflow (by opening the
outlet pipe or sluice gates and permitting the reservoir
to draw down sufficiently to resuspend sediment and
movebedload), when the river’s transporting capacityis
inadequate to move the increased load. On the Kern
River, the Southern California Edison Company (an
electric utility) obtained agency permission to sluice
sand from Democrat Dam in 1986, anticipating that the
sand would be washed from the channel the subsequent
winter. However, several years of drought ensued, and
the sand remained within the channel until high flows
in1992(Figure6)(DanChristenson, California Depart-
mentofFishandGame,Kernville, personal communica-
tion 1992).
On those dams larger than small diversion struc-
tures, the sediment accumulated around the outlet is
usually silt and clay, which can be deleterious to aquatic
habitat and water quality (Bjornn and Reiser 1991).
Opening of the low-level outlet on Los Padres Dam on
the Carmel River, California, released silt and clay,
which resulted in a large fish kill in 1980 (Buel 1980).
The dam operator has since been required to use a
suction dredge to maintain the outlet (D. Dettman,
Monterey Peninsula Water Management District, per-
sonal communication 1990). On the Dan River in
Danville, Virginia, toxicity testing is required during
sluicing of fine sediments from Schoolfield Dam (FERC
1995). Accidental sluices have also occurred during
maintenance or repair work, sometimes resulting in
substantial cleanup operations for the dam operators
(Ramey and Beck 1990, Kondolf 1995).
Less serious effects are likely when the sediment
pulse is released during high flows, which will have
elevated suspended loads, but which can typically dis-
perse the sediment for some distance downstream. The
Jansanpei Reservoir in Taiwan is operated to provide
Figure 6. Sand deposited in the bed of
the Kern River as a result of sluicing from
Democrat Dam in 1986. (Photograph by
the author, December 1990.)
G. M. Kondolf
538
power for the Taiwan Sugar Company, which needs
power for processing only from November to April. The
reservoir is left empty with open low-level outlets for the
first two months of the rainy season (May and June), so
sediments accumulated over the months of July–April
can be flushed by the first high flows of the season
beforestoring water in the latter part of therainyseason
(Hwang 1994).
At present, sediment pass-through is not commonly
donein North America, probably because of thelimited
capacity of many low-level outlets and because of con-
cern that debris may become stuck in the outlets,
making them impossible to close later, and making
diversions impossible during the rest of the wet season
until flows drop sufficiently to fix the outlets. These
concerns can probably be addressed with engineering
solutions, such as trash racks upstream of the outlet and
redundancies in gate structures on the low-level outlet.
Large reservoirs cannot be drawn down sufficiently to
transport sediment through their length to the outlet
works, for such a drawdown would eliminate carryover
storage from year to year, an important benefit from
large reservoirs.
In most reservoirs in the United States, sediment is
simply permitted to accumulate. Active management of
sediment in reservoirs has been rare, largely because
the long-term costs of reservoir storage lost to sedimen-
tationhavenotbeen incorporated into decision-making
and planning for reservoirs. Most good reservoir sites
are already occupied by reservoirs, and where suitable
replacement reservoir sites exist, the current cost of
replacement storage (about US$3/m3in California) is
considerablyhigherthanoriginalstoragecosts.Mechani-
cal removal is prohibitively expensive in all but small
reservoirs, with costs of $15–$50/m3cited for the
Feather River in California (Kondolf 1995).
Channel Narrowing and Fine Sediment
Accumulation Below Dams
While many reservoirs reduce flood peaks, the de-
gree of reduction varies considerably depending upon
reservoir size and operation. The larger the reservoir
capacity relative to river flow and the greater the flood
pool available during a given flood, the greater the
reduction in peak floods. Flood control reservoirs
typically contain larger floods than reservoirs operated
solely for water supply. Downstream of the reservoir,
encroachment of riparian vegetation into parts of the
active channel may occur in response to a reduction in
annual flood scour and sediment deposition (Williams
and Wolman 1984). Channel narrowing has been great-
est below reservoirs that are large enough to contain
the river’s largest floods. In some cases, fine sediment
delivered to the river channel by tributaries accumu-
lates in spawning gravels because the reservoir-reduced
floods are inadequate to flush the riverbed clean.
On the Trinity River, California, construction of
Trinity Dam in 1960 reduced the two-year flow from 450
m3/sec to 9 m3/sec. As a result of this dramatic change
in flood regime, encroachment of vegetation and depo-
sition of sediment has narrowed the channel to 20%–
60% of its predam width (Wilcock and others 1996).
Accumulation of tributary-derived decomposed gra-
nitic sand in the bed of the Trinity River has led to a
decline of invertebrate and salmonid spawning habitat
(Fredericksen, Kamine and Associates 1980). Experi-
mental, controlled releases were made in 1991, 1992,
1993,1995, and 1996 to determinethe flows required to
flush the sand from the gravels (Wilcock and others
1996).
Such flushing flows increasingly have been proposed
for reaches downstream of reservoirs to remove fine
sedimentsaccumulated on the bed and to scourthe bed
frequentlyenough to prevent encroachment of riparian
vegetation and narrowing of the active channel (Reiser
and others 1989). The objectives of flushing flows have
not always been clearly specified, nor have potential
conflicts always been recognized. For example, a dis-
charge that mobilizes the channel bed to flush intersti-
tial fine sediment will often produce comparable trans-
port rates of sand and gravel, eliminating the selective
transport of sand needed to reduce the fine sediment
content in the bed, and resulting in a net loss of gravel
from the reach given its lack of supply from upstream
(Kondolf and Wilcock 1996).
Coastal Erosion
Beaches serve to dissipate wave action and protect
coastal cliffs. Sand may be supplied to beaches from
headland erosion, river transport, and offshore sources.
If sand supply is reduced through a reduction in
sediment delivery from rivers and streams, the beach
may become undernourished, shrink, and cliff erosion
may be accelerated. This process by which beaches are
reduced or maintained can be thought of in terms of a
sediment balance between sources of sediment (rivers
and headland erosion), the rate of longshore transport
along the coast, and sediment sinks (such as loss to
deeper water offshore) (Inman 1976). Along the coast
of southern California, discrete coastal cells can be
identified, each with distinct sediment sources (sedi-
ment delivery from river mouths) and sinks (losses to
submarine canyons). For example, for the Oceanside
littoral cell, the contribution from sediment sources
(Santa Margarita, San Luis Rey, and San Dieguito rivers
and San Mateo and San Juan creeks) was estimated,
Effects of Dams and Gravel Mining on Rivers 539
under natural conditions, at 209,000 m3/yr, roughly
balancing the longshore transport rate of 194,000
m3/yr and the loss into the La Jolla submarine canyon
of 200,000 m 3/yr (Figure 7) (Inman 1985).
Thesupply of sediment to beaches from rivers canbe
reduced by dams because dams trap sediment and
because large dams typically reduce the magnitude of
floods, which transport the majority of sediment (Jen-
kins and others 1988). In southern California rivers,
mostsedimenttransportoccursduringinfrequentfloods
(Brownlie and Taylor 1981), but it is these energetic
events that flood control dams are constructed to
prevent. On the San Luis Rey River, one of the principal
sources of sediment for the Oceanside littoral cell,
Henshaw Dam reduced suspended sediment yield by 6
million tonnes (Figure 8), total sand and gravel yield by
2 million tonnes (Brownlie and Taylor 1981).
Ironically, by trapping sediment and reducing peak
flows, the flood control dams meant to reduce property
damage along rivers contribute to property damage
along the coast by eliminating sediment supply to the
protective beaches. For the rivers contributing sedi-
ment to the Oceanside littoral cell as a whole, sediment
from about 40% of the catchment area is now cut off
by dams. Because the rate of longshore transport (a
function of wave energy striking the coast) is un-
changed, the result has been a sediment deficit, loss of
beach sand, and accelerated coastal erosion (Inman
1985).
The effects of sediment trapping by dams has been
exacerbated in combination with other effects such as
channelization and instream sand and gravel mining
(discussed below). Although sluicing sediment from
reservoirs has been considered in the Los Angeles
Basin, passing sediment through urban flood control
channels could cause a number of problems, including
decreasing channel capacity (Potter 1985). ‘‘Beach
nourishment’’ with imported sediment dredged from
reservoirs and harbors has been implemented along
many beaches in southern California (Inman 1976,
Allayaud 1985, Everts 1985). In some cases, sand is
transported to critical locations on the coast via truck or
slurry pipelines. The high costs of transportation, sort-
ing for the proper size fractions, and cleaning contami-
nated dredged material, as well as the difficulty in
securing a stable supply of material make these options
infeasible in some places (Inman 1976).
To integrate considerations of fluvial sediment sup-
ply in the maintenance of coastal beaches into the
existing legal framework, a system of ‘‘sand rights,’’
analogous to water rights, has been proposed (Stone
and Kaufman 1985).
Gravel Mining in River Systems
Sand and gravel are used as construction aggregate
for roads and highways (base material and asphalt),
pipelines (bedding), septic systems (drain rock in leach
fields), and concrete (aggregate mix) for highways and
buildings. In many areas, aggregate is derived primarily
Figure 7. The Oceanside littoral cell, showing estimated sand
and gravel supply from rivers, longshore transport, and loss to
the La Jolla submarine canyon (in m3/yr). (Adapted from
Inman 1985, used by permission.)
Figure 8. Cumulative reduction in suspended sediment sup-
ply from the catchment of the San Luis Rey River due to
construction of Henshaw Dam. (Adapted from Brownlie and
Taylor 1981.)
G. M. Kondolf
540
from alluvial deposits, either from pits in river flood-
plains and terrances, or by in-channel (instream) min-
ing, removing sand and gravel directly from river beds
with heavy equipment.
Sand and gravel that have been subject to prolonged
transport in water (such as active channel deposits) are
particularly desirable sources of aggregate because
weak materials are eliminated by abrasion and attrition,
leavingdurable,rounded,well-sorted gravels (Barksdale
1991). Instream gravels thus require less processing
than many other sources, and suitable channel deposits
are commonly located near the markets for the product
or on transportation routes, reducing transportation
costs (which are the largest costs in the industry).
Moreover, instream gravels are typically of sufficiently
high quality to be classified as ‘‘PCC-grade’’ aggregate,
suitable for use in production of Portland Cement
concrete (Barksdale 1991).
Effects of Instream Gravel Mining
Instream mining directly alters the channel geom-
etry and bed elevation and may involve extensive
clearing, diversion of flow, stockpiling of sediment, and
excavation of deep pits (Sandecki 1989). Instream
mining may be carried out by excavating trenches or
pits in the gravel bed, or by gravel bar skimming (or
scalping), removing all the material in a gravel bar
above an imaginary line sloping upwards from the
summer water’s edge. In both cases, the preexisting
channel morphology is disrupted and a local sediment
deficit is produced, but trenching also leaves a headcut
onitsupstreamend. In addition to the direct alterations
of the river environment, instream gravel mining may
induce channel incision, bed coarsening, and lateral
channel instability (Kondolf 1994).
Channel Incision and Bed Coarsening
By removing sediment from the channel, instream
gravel mining disrupts the preexisting balance between
sediment supply and transporting capacity, typically
inducing incision upstream and downstream of the
extraction site. Excavation of pits in the active channel
alters the equilibrium profile of the streambed, creating
a locally steeper gradient upon entering the pit (Figure
9). This over-steepened nickpoint (with its increased
stream power) commonly erodes upstream in a process
known as headcutting. Mining-induced incision may
propagate upstream for kilometers on the main river
(Scott1973,Stevensand others 1990) and up tributaries
(Harvey and Schumm 1987). Gravel pits trap much of
the incoming bedload sediment, passing hungry water
downstream, which typically erodes the channel bed
and banks to regain at least part of its sediment load
(Figure 9).
A vivid example of mining-induced nickpoint migra-
tion appears on a detailed topographic map prepared
from analysis of 1992 aerial photographs of Cache
Creek, California. The bed had been actively mined up
to the miner’s property boundary about 1400 m down-
streamofCapay Bridge, with a 4-m high headwall on the
upstream edge of the excavation. After the 1992 winter
flows, a nickpoint over 3 m deep extended 700 m
upstream from the upstream edge of the pit (Figure
10). After the flows of 1993, the nickpoint had migrated
another 260 m upstream of the excavation (not shown),
and in the 50-yr flood of 1995, the nickpoint migrated
under the Capay Bridge, contributing to the near-
failure of the structure (Northwest Hydraulics Consul-
tants 1995).
On the Russian River near Healdsburg, California,
instream pit mining in the 1950s and 1960s caused
channel incision in excess of 3–6 m over an 11-km
length of river (Figure 11). The formerly wide channel
oftheRussianRiver is now incised, straighter,prevented
from migrating across the valley floor by levees, and
thus unable to maintain the diversity of successional
Figure 9. Incision produced by instream gravel mining. a:
The initial, preextraction condition, in which the river’s
sediment load (Qs) and the shear stress (t) available to
transport sediment are continuous through the reach. b: The
excavation creates a nickpoint on its upstream end and traps
sediment, interrupting the transport of sediment through the
reach. Downstream, the river still has the capacity to transport
sediment (t) but no sediment load. c: The nickpoint migrates
upstream, and hungry water erodes the bed downstream,
causing incision upstream and downstream. (Reprinted from
Kondolf 1994, with kind permission of Elsevier Science-NL.)
Effects of Dams and Gravel Mining on Rivers 541
stages of vegetation associated with an actively migrat-
ing river (Florsheim and Goodwin 1993). With contin-
ued extraction, the bed may degrade down to bedrock
or older substrates under the recent alluvium (Figure
12). Just as below dams, gravel-bed rivers may become
armored, limiting further incision (Dietrich and others
1989), but eliminating salmonid spawning habitat.
In many rivers, gravel mining has been conducted
downstream of dams, combining the effects of both
impacts to produce an even larger sediment deficit. On
the San Luis Rey River downstream of Henshaw Dam,
five gravel mining operations within 8 km of the
Highway 395 bridge extract a permitted volume of
approximately 300,000 m3/yr, about 50 times greater
than the estimated postdam bedload sediment yield
(Kondolf and Larson 1995), further exacerbating the
coastal sediment deficit.
Incision of the riverbed typically causes the alluvial
aquifer to drain to a lower level, resulting in a loss of
aquifer storage, as documented along the Russian River
(Sonoma County 1992). The Lake County (California)
Planning Department (Lake County 1992) estimated
that incision from instream mining in small river valleys
could reduce alluvial aquifer storage from 1% to 16%,
depending on local geology and aquifer geometry.
Undermining of Structures
The direct effects of incision include undermining
of bridge piers and other structures, and exposure of
buried pipeline crossings and water-supply facilities.
Headcutting of over7mfromaninstream gravel mine
downstream on the Kaoping River, Taiwan, threatens
the Kaoping Bridge, whose downstream margin is now
protected with gabions, massive coastal concrete jacks,
and lengthened piers (Figure 13).
On the San Luis Rey River, instream gravel mining
has not only reduced the supply of sediment to the
coast, but mining-induced incision has exposed aque-
ducts, gas pipelines, and other utilities buried in the
Figure 10. Nickpoint upstream of 4-m-deep gravel pit in the bed of Cache Creek, California, as appearing on a topographic map
of Cache Creek prepared from fall 1992 aerial photographs. Original map scale 1:2400, contour interval 0.6 m.
Figure 11. Longitudinal profile of the Russian River, near
Healdsburg, California, showing incision from 1940 to 1991.
(Redrawn from Florsheim and Goodwin 1993, used by permis-
sion.)
G. M. Kondolf
542
bedand exposed the footings of a major highwaybridge
(Parsons Brinkeroff Gore & Storrie, Inc. 1994). The
Highway 32 bridge over Stony Creek, California, has
been undermined as a result of intensive gravel mining
directly upstream and downstream of the bridge (Kon-
dolfand Swanson 1993). Municipal water supply intakes
have been damaged or made less effective on the Mad
(Lehre and others 1993) and Russian (Marcus 1992)
rivers in California as the layer of overlying gravel has
decreased due to incision.
Channel Instability
Instream mining can cause channel instability
through disruption of the existing equilibrium channel
form or undercutting of banks caused by incision.
Gravel mining in Blackwood Creek, California, caused
incision and channel instability upstream and down-
stream, increasing the stream’s sediment yield fourfold
(Todd 1989). As a nickpoint migrates upstream, its
incision and bank undercutting release additional sedi-
ment to downstream reaches, where the channel may
aggradeand thereby become unstable (Sear and Archer
1995). Incision in the mainstem Russian River propa-
gated up its tributary Dry Creek, resulting in undercut-
ting of banks, channel widening (from 10 to 400 m in
places), and destabilization, increasing delivery of sand
and gravel to the mainstem Russian River (Harvey and
Schumm 1987).
Figure 12. Tributary to the Sacramento
River near Redding, California, eroded to
bedrock as a result of instream mining.
(Photograph by author, January 1989.)
Figure 13. Undercutting and grade con-
trol efforts along the downstream side of
the Kaoping Bridge over the Kaoping
River, Taiwan, to control incision caused by
massive gravel mining downstream. (Pho-
tograph by the author, October 1995.)
Effects of Dams and Gravel Mining on Rivers 543
A more subtle but potentially significant effect is the
increased mobility of the gravel bed if the pavement
(the active coarse surface layer) (Parker and Klingeman
1982) is disrupted by mining. Similarly, removal of
gravel bars by instream mining can eliminate the
hydraulic control for the reach upstream, inducing
scour of upstream riffles and thus washout of incubat-
ing salmon embryos (Pauley and others 1989).
Secondary Effects of Instream Mining
Among the secondary effects of instream mining are
reduced loading of coarse woody debris in the channel,
which is important as cover for fish (Bisson and others
1987). Extraction (even bar skimming at low extraction
rates) typically results in a wider, shallower streambed,
leading to increased water temperatures, modification
of pool-riffle distribution, alteration of intergravel flow
paths, and thus degradation of salmonid habitat.
Resolving the Effects of Instream Mining
from Other Influences
In many rivers, several factors potentially causing
incision in the channel may be operating simulta-
neously, such as sediment trapping by dams, reduced
channel migration by bank protection, reduced over-
bank flooding from levees, and instream mining. How-
ever, in many rivers the rate of aggregate extraction is an
order of magnitude greater than the rate of sediment
supply from the drainage basin, providing strong evi-
dence for the role of extraction in causing channel
change. On Stony Creek, the incision produced by
Black Butte Reservoir could be clearly distinguished
from the effects of instream mining at the Highway 32
bridge by virtue of the distinct temporal and spatial
patterns of incision. The dam-induced incision was
pronounced downstream of the reservoir soon after its
construction in 1963. By contrast, the instream mining
(at rates exceeding the predam sediment supply by
200%–600%, and exceeding the postdam sediment
supply by 1000%–3000%) produced incision of up to 7
m centered in the mining reach near the Highway 32
bridge, after intensification of gravel mining in the
1970s (Kondolf and Swanson 1993) (Figure 14).
Management of Instream Gravel Mining
Instream mining has long been prohibited in the
United Kingdom, Germany, France, the Netherlands,
and Switzerland, and it is being reduced or prohibited
Figure 14. Sediment budget for Stony Creek, California. (Reprinted from Kondolf and Swanson 1993, used by permission of
Spring-Verlag, New York.)
G. M. Kondolf
544
in many rivers where impacts are apparent in Italy,
Portugal, and New Zealand. In the United States and
Canada, instream mining continues in many rivers,
despite increasing public opposition and recognition of
environmental effects by regulatory agencies. Instream
mines continue to operate illegally in many places, such
as the United States (Los Angeles Times 1992) and
Taiwan.
Strategies used to manage instream mining range
widely, and in many jurisdictions there is no effective
management. One strategy is to define a redline, a
minimum elevation for the thalweg (the deepest point
in a channel cross section) along the river, and to
permit mining so long as the bed does not incise below
this line (as determined by annual surveys of river
topography). The redline approach addresses a prob-
lem common to many permits in California, which have
specified that extraction is permitted ‘‘xfeet below the
channel bed’’ or only down to the thalweg, without
stating these limits in terms of actual elevations above a
permanent datum. Thus the extraction limits have
migrated vertically downward as the channel incises.
Another approach is to estimate the annual bedload
sediment supply from upstream (the replenishment
rate) and to limit annual extraction to that value or
some fraction thereof, considered the ‘‘safe yield.’’ The
replenishment rate approach has the virtue of scaling
extractiontotheriver load in a general way,but bedload
transport can be notoriously variable from year to year.
Thus, this approach is probably better if permitted
extraction rates are based on new deposition that year
rather than on long-term average bedload yields. More
fundamentally,however,thenotionthat one can extract
at the replenishment rate without affecting the channel
ignores the continuity of sediment transport through
the river system. The mined reach is the ‘‘upstream’’
sediment source for downstream reaches, so mining at
the replenishment rate could be expected to produce
hungry water conditions downstream. Habitat manag-
ers in Washington state have sought to limit extraction
to 50% of the transport rate as a first-cut estimate of safe
yield to minimize effects upon salmon spawning habitat
(Bates 1987).
Current approaches to managing instream mining
are based on empirical studies. While a theoretical
approach to predicting the effects of different levels of
gravelminingon rivers would be desirable, the inherent
complexity of sediment transport and channel change
makes firm, specific predictions impossible at present.
Sedimenttransport models can provide an indication of
potential channel incision and aggradation, but all such
models are simplifications of a complex reality, and the
utility of existing models is limited by unreliable formu-
lation of sediment rating curves, variations in hydraulic
roughness, and inadequate understanding of the me-
chanics of bed coarsening and bank erosion (NRC
1983).
In 1995, the US Department of Transportation
issued a notice to state transportation agencies indicat-
ing that federal funds will no longer be available to
repair bridges damaged by gravel mining, a move that
may motivate more vigorous enforcement of regula-
tions governing gravel mining in rivers by states.
Floodplain Pit Mining
Floodplain pit mining transforms riparian woodland
or agricultural land into open pits, which typically
intersect the water table at least seasonally (Figure 15).
Floodplain pit mining has effectively transformed large
areas of floodplain into open-water ponds, whose water
level commonly tracks that of the main river closely, and
which are commonly separated from the active channel
byonly a narrow strip ofunmined land. Because the pits
are in close hydrologic continuity with the alluvial water
table, concerns are often raised that contamination of
the pits may lead to contamination of the alluvial
aquifer. Many existing pits are steep-sided (to maximize
gravel yield per unit area) and offer relatively limited
wetlands habitat, but with improved pit design (e.g.,
gently sloping banks, irregular shorelines), greater
wildlife benefits are possible upon reclamation (An-
drews and Kinsman 1990, Giles 1992).
In many cases, floodplain pits have captured the
channel during floods, in effect converting formerly
off-channel mines to in-channel mines. Pit capture
occurs when the strip of land separating the pit from
the channel is breached by lateral channel erosion or by
overflowing floodwaters. In general, pit capture is most
likely when flowing through the pit offers the river a
shorter course than the currently active channel.
When pit capture occurs, the formerly off-channel
pit is converted into an in-channel pit, and the effects of
instream mining can be expected, notably propagation
of incision up- and downstream of the pit. Channel
capture by an off-channel pit on the alluvial fan of
Tujunga Wash near Los Angeles created a nickpoint
that migrated upstream, undermining highway bridges
(Scott 1973). The Yakima River, Washington, was cap-
tured by two floodplain pits in 1971, and began under-
cutting the highway for whose construction the pits had
been originally excavated (Dunne and Leopold 1978).
High flows on the Clackamas River, Oregon, in 1996
resulted in capture of an off-channel pit and resulted in
2 m of incision documented about 1 km upstream
Effects of Dams and Gravel Mining on Rivers 545
(Figure 16) and caused undermining of a building at
the gravel mine site (Figure 17).
Off-channel gravel pits have been used successfully
as spawning and rearing habitat for salmon and trout in
Idaho (Richards and others 1992) and on the Olympic
Peninsula of Washington (Partee and Samuelson 1993).
In warmer climates, however, these off-channel pits are
likely to heat up in the summer and provide habitat for
warm-water fish that prey on juvenile salmonids. During
floods, these pits may serve as a source of warm-water
fish to the main channel, and juvenile salmon can
become stranded in the pits. The Merced River, Califor-
nia, flows through at least 15 gravel pits, of which seven
were excavated in the active channel, and eight were
excavated on the floodplain and subsequently captured
the channel (Vick 1995). Juvenile salmon migrating
towards the ocean become disoriented in the quiet
water of these pits and suffer high losses to predation by
largemouth and smallmouth bass (Micropterus salmoides
and M. dolomieui). On the nearby Tuolumne River, a
1987 study by the California Department of Fish and
Game estimated that juvenile chinook salmon migrat-
ing oceanward suffered 70% losses to predation (mostly
in gravel pits) in the three days required to traverse an
80-km reach from LaGrange Dam to the San Joaquin
River (EA 1992). To reduce this predation problem,
funding has been allocated to repair breached levees at
one gravel pit on the Merced River at a cost of
Figure 15. Floodplain pit along Cotton-
wood Creek near Redding, California.
(Photograph by author, January 1989.)
Figure 16. Incision of Clackamas River
approximately one mile upstream of
captured gravel pit near Barton, Or-
egon. The three men on the right are
standing on the bed of a side channel
that formerly joined the mainstem at
grade, but is now elevated about 2 m
above the current river bed, after up-
stream migration of a nickpoint from
the gravel pit. View upstream. (Photo-
graph by author, April 1996.)
G. M. Kondolf
546
US$361,000 (Kondolf and others 1996a), and refilling
of two pits on the Tuolumne River has been proposed at
a cost of $5.3 million (McBain and Trush 1996).
Aggregate Supply, Quality, and Uses
Aggregates can be obtained from a wide variety of
sources (besides fluvial deposits), such as dry terrace
mines, quarries (from which rock must be crushed,
washed, and sorted), dredger tailings, reservoir deltas,
andrecyclingconcreterubble. These alternative sources
usually require more processing and often require
longer transportation. Although their production costs
are commonly higher, these alternative sources avoid
many impacts of riverine extraction and may provide
other benefits, such as partially restoring reservoir
capacity lost to sedimentation and providing opportuni-
ties for ecological restoration of sterile dredger tailings.
In California, most aggregate that has been pro-
duced to date has been PCC-grade aggregate from
instream deposits or recent channel deposits in flood-
plains.These deposits were viewed as virtually infinite in
supply, and these high-grade aggregates have been used
in applications (such as road subbase) for which other,
more abundant aggregates (e.g., crushed rock from
upland quarries) would be acceptable. Given that de-
mand for aggregate commonly exceeds the supply of
sand and gravel from the catchment by an order of
magnitude or more, public policy ought to encourage
reservation of the most valuable aggregate resources for
thehighestenduses.PCC-gradeinstreamgravelsshould
be used, to the extent possible, only in applications
requiring such high-quality aggregate. Upland quarry
and terrace pit sources of lower-grade aggregate should
be identified, and alternative sources such as mining
golddredgertailingsorreservoir accumulations, should
beevaluated.Whereverpossible,concreterubble should
berecycledto produce aggregate for many applications.
Reservoir sediments are a largely unexploited source
of building materials in the United States. In general,
reservoir deposits will be attractive sources of aggre-
gates to the extent that they are sorted by size. The
depositional pattern within a reservoir depends on
reservoir size and configuration and the reservoir stage
during floods. Small diversion dams may have a low trap
efficiency for suspended sediments and trap primarily
sand and gravel, while larger reservoirs will have mostly
finer-grainedsand,silt,andclay(depositedfromsuspen-
sion) throughout most of the reservoir, with coarse
sediment typically concentrated in deltas at the up-
stream end of the reservoir. These coarse deposits will
extend farther if the reservoir is drawn down to a low
level when the sediment-laden water enters. In many
reservoirs, sand and gravel occur at the upstream end,
silts and clays at the downstream end, and a mixed zone
ofinterbeddedcoarseand fine sediments in the middle.
Sand and gravel are mined commercially from some
debris basins in the Los Angeles Basin and from Rollins
Reservoir on the Bear River in California. In Taiwan,
most reservoir sediments are fine-grained (owing to the
caliber of the source rocks), but where coarser sedi-
ments are deposited, they are virtually all mined for
construction aggregate (J. S. Hwang, Taiwan Provincial
Water Conservancy Bureau, Taichung City, personal
communication1996).InIsrael,the2.2-km-longShikma
Reservoir is mined in its upper 600 m to produce sand
and gravel for construction aggregate, and in its lower 1
km to produce clay for use in cement, bricks, clay seals
Figure 17. Building undercut by bank
erosion as the Clackamas River flows
through a captured gravel pit near Bar-
ton, Oregon. (Photograph by the author,
April 1996.)
Effects of Dams and Gravel Mining on Rivers 547
for sewage treatment ponds, and pottery (Laronne
1995, Taig 1996). The zone of mixed sediments in the
mid-section of the reservoir is left unexcavated and
vegetated so it permits only fine-grained washload to
passdownstreamintothelowerreservoir, thereby ensur-
ing continued deposition of sand and gravel in the
upstream portion of the reservoir and silt and clay in
the downstream portion. The extraction itself restores
some of the reservoir capacity lost to sedimentation.
Similarly, on Nahal Besor, Israel, the off-channel Lower
Rehovot Reservoir was deliberately created (to provide
needed reservoir storage) by gravel mining. Water is
diverted into the reservoir through a spillway at high
flows, as controlled by a weir across the channel (Cohen
1996).
Extraction of reservoir sediments partially mitigates
losses in reservoir capacity from sedimentation. Be-
cause of the high costs and practical problems with
construction of replacement reservoir storage and/or
mechanical removal of sediment, restoration of reser-
voir capacity may be seen as one of the chief benefits
from mining aggregate and industrial clays from reser-
voirs. If these benefits are recognized, mining reservoir
deposits may become more economically attractive in
the future, especially if the environmental costs of
instream and floodplain mining become better recog-
nized and reflected in the prices of those aggregates. In
the United States, construction of reservoirs was often
justified partially by anticipated recreational benefits,
and thus reservoir margins are commonly designated as
recreation areas, posing a potential conflict with an
industrial use such as gravel mining. Furthermore,
wetlands may form in reservoir delta deposits, posing
potential conflicts with regulations protecting wetlands.
Conclusions
Comprehensive management of gravel and sand in
river systems should be based on a recognition of the
natural flow of sediment through the drainage network
and the nature of impacts (to ecological resources and
to infrastructure) likely to occur when the continuity of
sediment is disrupted. A sediment budget should be
developed for present and historical conditions as a
fundamentalbasis for evaluation of these impacts, many
of which are cumulative in nature.
The cost of sediment-related impacts of existing and
proposed water development projects and aggregate
mines must be realistically assessed and included in
economic evaluations of these projects. The (very real)
costs of impacts such as bridge undermining, loss of
spawning gravels, and loss of beach sand are now
externalized, borne by other sectors of society rather
than the generators of the impacts. The notion of
sediment rights (analogous to water rights) should be
exploredasaframework within which to assess reservoir
operations and aggregate mining for these impacts.
Sediment pass-through should be undertaken in
reservoirs (where feasible) to mimic the natural flux of
sediment through the river system. Pass-through should
be done only during high flows when the sediment is
likely to continue dispersing downstream from the
reservoir. The cost of installing larger low-level outlets
(where necessary) on existing dams will generally be
less than costs of mechanical removal of sediments over
subsequent decades. In larger reservoirs where sedi-
ment cannot be passed through a drawn-down reser-
voir, alternative means of transporting the gravel and
sand fractions around (or through) reservoirs using
tunnels, pipes, or barges should be explored.
Flushing flows should be evaluated not only in light
of potential benefits of flushing fine sediments from
mobilized gravels, but also the potential loss of gravel
from the reach due to downstream transport.
The regional context of aggregate resources, market
demand, and the environmental impacts of various
alternatives must be understood before any site-specific
proposal for aggregate extraction can be sensibly re-
viewed. In general, effects of aggregate mining should
be evaluated on a river basin scale, so that the cumula-
tive effects of extraction on the aquatic and riparian
resources can be recognized. Evaluation of aggregate
supply and demand should be undertaken on the basis
of production–consumption regions, encompassing the
market for aggregate and all potential sources of aggre-
gate within an economical transport distance.
The finite nature of high-quality alluvial gravel re-
sources must recognized, and high-quality PCC-grade
aggregatesshouldbereserved only for the uses demand-
ing this quality material (such as concrete). Alternative
sources should be used in less demanding applications
(such as road subbase). The environmental costs of
instream mining should be incorporated into the price
of the product so that alternative sources that require
more processing but have less environmental impact
become more attractive.
Instream mining should not be permitted in rivers
downstream of dams by virtue of the lack of supply from
upstream or in rivers with important salmon spawning
(unless it can be shown that the extraction will not
degrade habitat).
Acknowledgments
The concepts presented in this paper have drawn
uponresearchover a decade and interesting discussions
G. M. Kondolf
548
with many colleagues, including Ken Bates, Koll Buer,
Brian Collins, Cathy Crossett, Peter Geldner, Peter
Goodwin, Murray Hicks, Jing-San Hwang, Steve Jones,
Pete Klingeman, John Laronne, Han-Bin Liang, Bob
MacArthur, Graham Matthews, Scott McBain, Gregg
Morris, Mike Sandecki, Mitchell Swanson, Jen Vick, Ed
Wallace, Peter Wilcock, and John Williams. This paper
has benefitted from critical comments from Mary Ann
Madej, Graham Matthews, and an anonymous reviewer.
The research upon which this paper is based was
partially supported by the University of CaliforniaWater
Resources Center (UC Davis), as part of Water Re-
sources Center project UCAL-WRC-W-748, adminis-
tered by the Center for Environmental Design Re-
search, and by a grant from the Beatrix Farrand Fund of
the Department of Landscape Architecture, both at the
University of California, Berkeley.
Literature Cited
Allayaud, W. K. 1985. Innovations in non-structural solutions
to preventing coastal damage. Pages 260–290 in J. McGrath
(ed.),California’sbattered coast, proceedings from aconfer-
ence on coastal erosion. California Coastal Commission.
Andrews, J., and D. Kinsman. 1990. Gravel pit restoration for
wildlife: a practical manual. The Royal Society for Protec-
tion of Birds, Sandy, Bedfordshire.
Barksdale, R. D. 1991. The aggregate handbook. National
Stone Association, Washington, DC.
Bates, K. 1987. Fisheries perspectives on gravel removal from
river channels. Pages 292–298 in Realistic approaches to
better floodplain management. Proceedings of the eleventh
annual conference of the Association of State Floodplain
Managers, Seattle, June 1987. Natural Hazards Research
and Applications Information Center, Special Publication
No. 18.
Bisson, P. A. and eight coauthors. 1987. Large woody debris in
forested streams in the Pacific Northwest: Past present, and
future. Pages 143–190 in E. O. Salo and T. Cundy (eds.),
Proceedings of an interdisciplinary symposium on stream-
side management: Forestry and fisher y interactions. Univer-
sity of Washington Press, Seattle.
Bjornn, T. L., and D. W. Reiser. 1991. Habitat requirements of
salmonids in streams. Pages 83–138 in Influences of forest
and rangeland management on salmonid fishes and their
habitats. American Fisheries Society Special Publication 19.
Brownlie, W. R., and B. D. Taylor. 1981. Sediment manage-
ment for southern California mountains, coastal plain, and
shoreline. Part C. Coastal sediment delivery by major rivers
in southern California. Report. 17-C, Environmental Qual-
ity Lab, California Institute of Technology, Pasadena.
Brune, G. M. 1953. The trap efficiency of reservoirs. Transac-
tions of the American Geophysical Union 34:407–418.
Buel, B. 1980. Effects of Los Padres Reservoir silt release.
Unpublished memo. Monterey Peninsula Water Manage-
ment District, Monterey, California.
California Department of Water Resources. 1995. Sacramento
River gravel restoration phase II study: A plan for continued
spawning gravel replenishment between Keswick Dam and
clear Creek. Technical Information Record TIR ND-95-1.
California Department of Water Resources, Northern Dis-
trict, Red Bluff, California.
Chongshan, Z., W. Jianguo, and L. Quigmei. 1995. Experi-
ment study of approach for sediment removed from reser-
voirs. Pages 149–154 in Proceedings of sixth international
symposium on river sedimentation, New Delhi, India.
Cohen, M. 1996. Structures and sills in river channels. Pages
42–44 in J. B. Laronne (ed.), Reservoirs as a source of water
for the Negev conference proceedings. Ben-Gurion Univer-
sity, Be’er Sheva, Israel (in Hebrew).
Collins, B., and T. Dunne. 1990. Fluvial geomorphology and
river gravel mining: A guide for planners, case studies
included. California Division of Mines and Geology Special
Publication 98. Sacramento.
Denton, D. N. 1991. Sacramento River gravel restoration
progress report. Unpublished report. California Depart-
ment of Water Resources, Red Bluff, California, January
1991.
Dietrich, W. E., J. W. Kirchner, H. Ikeda, and F. Iseya. 1989.
Sediment supply and development of coarse surface layer in
gravel bedded rivers. Nature 340:215–217.
Dunne, T., and L. B. Leopold. 1978. Water in environmental
planning. W. H. Freeman & Sons, San Francisco.
EA (EA Engineering, Science, and Technology). 1992. Don
Pedro Project fisheries studies report (FERC Article 39,
Project No. 2299). Report to Turlock Irrigation District and
Merced Irrigation District.
Everts, C. H. 1985. Effects of small protective devices on
beaches. Pages 127–138 in J. McGrath (ed.), California’s
battered coast, proceedings from a conference on coastal
erosion. California Coastal Commission.
FERC (Federal Energy Regulatory Commission). 1993. Final
environmental impact statement, propose modifications to
the Lower Mokelumne River Project, California, FERC
Project No. 2916-004. Washington, DC.
FERC (Federal Energy Regulatory Commission). 1995. Order
amending and approving sediment flushing plan, STS
Hydropower Limited and Dan River Incorporated, FERC
Project No. 2411-012. Washington, DC.
Florsheim, J., and P. Goodwin. 1993. Geomorphic and hydro-
logic conditions in the Russian River, California: Historic
trends and existing conditions. Discussion document, pre-
pared for California State Coastal Conservancy, Oakland.
Fredericksen, Kamine, and Associates. 1980. Proposed Trinity
River Basin fish and wildlife management program. Unpub-
lished report to US Water and Power Resources Service
(now the US Bureau of Reclamation).
Giles, N. 1992. Wildlife after gravel: Twenty years of practical
research by The Game Conservancy and ARC. The Game
Conservancy, Fordingbridge, Hampshire.
Griffiths, G. A., and M. J. McSaveney. 1983. Hydrology of a
basin with extreme rainfalls—Cropp River, New Zealand.
New Zealand Journal of Science 26:293–306.
Hack, H. P. 1986. Design and calculation of reservoirs of run
of river stations incorporating sedimentation. Pages 107–
Effects of Dams and Gravel Mining on Rivers 549
112 in W. Bechteler (ed.), Transport of suspended solids in
open channels, proceedings of Euromech 192. Munich,
Germany. June 11–15, 1985.
Harvey, M. D., and S. A. Schumm. 1987. Response of Dry
Creek, California, to land use change, gravel mining and
dam closure. Pages 451–460 in Erosion and sedimentation
in the Pacific Rim, proceedings of the Corvallis symposium,
August 1987. International Association of Hydrological
Sciences Publication 165.
Hassanzadeh, Y. 1995. The removal of reservoir sediment.
Water International 20:151–154.
Hazel, C., S. Herrera, H. Rectenwald, and J. Ives. 1976.
Assessment of effects of altered stream flow characteristics
on fish and wildlife. Part B: California case studies. Report
by Jones and Stokes, Inc. to US Department of Interior, Fish
and Wildlife Service.
Hwang, J. S. 1994. A study of the sustainable water resources
system in Taiwan considering the problems of reservoir
desilting. Taiwan Provincial Water Conservancy Bureau,
Taichung City, Taiwan.
Inman, D. L. 1976. Man’s impact on the California coastal
zone. Summary report to California Department of Naviga-
tion and Ocean Development, Sacramento.
Inman, D. L. 1985. Budget of sand in southern California;
river discharge vs. cliff erosion. Pages 10–15 in J. McGrath
(ed.),California’sbattered coast, proceedings from aconfer-
ence on coastal erosion. California Coastal Commission.
Janda, R. J. 1978. Summary of watershed conditions in the
vicinity of Redwood National Park. US Geological Survey
Open File Report 78-25, Menlo Park, California.
Jenkins, S. A., D. L. Inman, and D. W. Skelly. 1988. Impact of
dam building on the California coastal zone. California
Waterfront Age September.
Kondolf, G. M. 1994. Geomorphic and environmental effects
of instream gravel mining. Landscape and Urban Planning
28:225–243.
Kondolf, G. M. 1995. Managing bedload sediments in regu-
lated rivers: Examples from California, USA. Geophysical
Monograph 89:165–176.
Kondolf, G. M., and R. R. Curry. 1986. Channel erosion along
theCarmel River,Monterey County,California.Earth Surface
Processes and Landforms 11:307–319.
Kondolf, G. M., and M. Larson. 1995. Historical channel
analysis and its application to riparian and aquatic habitat
restoration. Aquatic Conservation 5:109–126.
Kondolf, G. M., and W. V. G. Matthews. 1993. Management of
coarse sediment in regulated rivers of California. Report
No. 80. University of California Water Resources Center,
Davis, California.
Kondolf, G. M., and M. L. Swanson. 1993. Channel adjust-
mentstoreservoir construction and instream gravel mining,
Stony Creek, California. Environmental Geology and Water
Science 21:256–269.
Kondolf, G. M., and P. R. Wilcock. 1996. The flushing flow
problem: Defining and evaluating objectives. Water Resources
Research 32(8):2589–2599.
Kondolf, G. M., and M. G. Wolman. 1993. The sizes of
salmonid spawning gravels. Water Resources Research 29:2275–
2285.
Kondolf, G. M., J. C. Vick, and T. M. Ramirez. 1996a. Salmon
spawning habitat rehabilitation in the Merced, Tuolumne,
and Stanislaus Rivers, California: An evaluation of project
planning and performance. Report No. 90, University of
California Water Resources Center, Davis, California.
Kondolf, G. M., J. C. Vick, and T. M. Ramirez. 1996b. Salmon
spawninghabitatrehabilitation on the MercedRiver, Califor-
nia: An evaluation of project planning and performance.
Transactions of the American Fisheries Society 125:899–912.
Kuhl, D. 1992. 14 years of artificial grain feeding in the Rhine
downstream the barrage Iffezheim. Pages 1121–1129 in
Proceedings 5th international symposium on river sedimen-
tation, Karlsruhe, Germany.
Lake County. 1992. Lake County aggregate resource manage-
ment plan. Lake County Planning Department, Resource
Management Division, Lakeport, California. Draft.
Laronne, J. B. 1995. Design of quarrying in the Shikma
Reservoir, final report to Mekorot, Israeli Water Supply
Company, Geography Department, Ben-Gurion University,
Be’er Sheva, Israel. July, 15 pp. (in Hebrew).
Lehre, A., R. D. Klein, and W. Trush. 1993. Analysis of the
effects of historic gravel extraction on the geomorphic
character and fisheries habitat of the Lower Mad River,
Humboldt County, California. Appendix F to the draft
program environmental impact report on gravel removal
fromtheLowerMad River.Departmentof Planning, County
of Humboldt, Eureka, California.
Leopold, L. B., M. G. Wolman, and J. P. Miller. 1964. Fluvial
processesingeomorphology. W.H. Freeman, San Francisco,
522 pp.
Los Angeles Times. 1992. Brothers get jail time for river
mining. Article by Jonathan Gaw. 17 June.
Madej, M. A., and V. Ozaki. 1996. Channel response to
sedimentwavepropagation and movement, RedwoodCreek,
California, USA.Earth Surface ProcessesandLandforms21:911–
927.
Marcus,L.1992. Status report: RussianRiverresourceenhance-
mentplan.CaliforniaCoastalConservancy,Oakland, Califor-
nia.
McBain, S. M., and W. Trush. 1996. Tuolumne River channel
restoration project, special run pools 9 and 10. Report
submitted to Tuolumne River Technical Advisory Commit-
tee (Don Pedro Project, FERC License No. 2299) by McBain
and Trush, Arcata, California.
Morris, G. L. 1993. A global perspective of sediment control
measures in reservoirs. In S. Fan and G. L. Morris (eds.),
Notes on sediment management in reservoirs: National and
international perspectives. US Federal Energy Regulatory
Commission, Washington DC.
Northwest Hydraulics Consultants. 1995. Cache Creek stream-
way study. Unpublished report to Yolo County Community
Development Agency, Woodland, California.
NRC (National Research Council). 1983. An evaluation of
flood-level prediction using alluvial-river models. Commit-
tee on Hydrodynamic Computer Models for Flood Insur-
ance Studies, Advisory Board on the Built Environment,
Commission on Engineering and Technical Systems, Na-
tional Research Council. National Academy Press, Washing-
ton DC.
G. M. Kondolf
550
Parfitt, D., and K. Buer. 1980. Upper Sacramento River
spawning gravel study. California Department of Water
Resources, Northern Division, Red Bluff.
Parker, G., and P. C. Klingeman. 1982. On why gravel bed
streams are paved. Water Resources Research 18:1409–1423.
Parsons Brinkerhoff Gore & Storrie, Inc. 1994. River manage-
ment study: permanent protection of the San Luis Rey River
Aqueduct crossings. Report to San Diego County Water
Authority.
Partee, R. R., and Samuelson, D. F. 1993. Weyco-Brisco ponds
habitat enhancement design criteria. Unpublished report,
Grays Harbor College, Aberdeen, Washington.
Pauley, G. B., G. L. Thomas, D. A. Marino, and D. C. Weigand.
1989. Evaluation of the effects of gravel bar scalping on
juvenile salmonids in the Puyallup River drainage. Univer-
sity of Washington Cooperative Fishery Research Unit Re-
port. University of Washington, Seattle.
Potter, D. 1985. Sand sluicing from dams on the San Gabriel
River—is it feasible? Pages 251–260 in J. McGrath (ed.),
California’s battered coast, proceedings from a conference
on coastal erosion. California Coastal Commission.
Ramey, M. P., and S. M. Beck. 1990. Flushing flow evaluation:
The north fork of the Feather River below Poe Dam.
Environment, Health, and Safety Report 009.4-89.9. Pacific
Gas and Electric Company, Department of Research and
Development, San Ramon, California.
Reiser, D. W., M. P. Ramey, and T. A. Wesche. 1989. Flushing
flows. Pages 91–135 in J. A. Gore and G. E. Petts (eds.),
Alternatives in regulated river management. CRC Press,
Boca Raton, Florida.
Richards, C., P. J. Cernera, M. P. Ramey, and D. W. Reiser.
1992. Development of off-channel habitats for use by juve-
nile chinook salmon. North American Journal of Fisheries
Management 12:721–727.
Richards, K. 1982. Rivers: Form and process in alluvial chan-
nels. Methuen, London, 358 pp.
Sandecki, M. 1989. Aggregate mining in river systems. Califor-
nia Geology 42(4):88–94.
Schick, A. P., and J. Lekach. 1993. An evaluation of two
ten-yearsedimentbudgets, Nahal Yael, Israel. PhysicalGeogra-
phy 14(3):225–238.
Schumm, S. A. 1977. The fluvial system. John Wiley & Sons,
New York.
Scott, K. M. 1973. Scour and fill in Tujunga Wash—a fanhead
valley in urban southern California—1969. US Geological
Survey Professional Paper 732-B.
Sear, D. A., and D. R. Archer. 1995. The effects of gravel
extraction on the stability of gravel-bed rivers: A case study
from the Wooler Water, Northumberland, UK. Paper pre-
sented to the 4th workshop on gravel bed rivers. Gold Bar,
Washington.
Sen, S. P., and A. Srivastava. 1995. Flushing of sediment from
small reservoir. Pages 149–154 in Proceedings of sixth
international symposiumonriver sedimentation, New Delhi,
India.
Sonoma County. 1992. Sonoma County aggregate resources
management plan and environmental impact report, draft.
Prepared by EIP Associates for Sonoma County Planning
Department, Santa Rosa, California.
Stevens, J. C. 1936. The silt problem. Paper No. 1927. Transac-
tions American Society of Civil Engineers.
Stevens, M. A., B. Urbonas, and L. S. Tucker. 1990. Public–
private cooperation protects river. APWA Reporter Septem-
ber: 25–27.
Stone, K. E., and B. S. Kaufman. 1985. Sand rights, a legal
system to protect the shores of the beach. Pages 280–297 in
J. McGrath (ed.), California’s battered coast, proceedings
from a conference on coastal erosion. California Coastal
Commission.
Taig, M. 1996. Use of sediment accumulated in flood reser-
voirs. Pages 25–30 in J. B. Laronne (ed.), Reservoirs as a
source of water for the Negev Conference Proceedings.
Ben-Gurion University, Be’er Sheva, Israel (in Hebrew).
Todd, A. H. 1989. The decline and recovery of Blackwood
Canyon, Lake Tahoe, California. In Proceedings, interna-
tional erosion control association conference. Vancouver,
British Columbia.
Vick, J. 1995. Habitat rehabilitation in the Lower Merced
River: A geomorphological perspective. Masters thesis in
Environmental Planning, Department of Landscape Archi-
tecture, and Report No. 03-95, Center for Environmental
Design Research, University of California, Berkeley.
Westrich, B., S. Al-Zoubi, and J. Muller. 1992. Planning and
designing a flushing channel for river reservoir sediment
management. Pages 861–867 in 5th international sympo-
sium on river sedimentation. Karlsruhe, Germany.
Wilcock, P. R., G. M. Kondolf, W. V. Matthews, and A. F. Barta.
1996. Specification of sediment maintenance flows for a
large gravel-bed river. Water Resources Research 32(9):2911–
2921.
Williams, G. P., and M. G. Wolman. 1984. Downstream effects
ofdamsonalluvialrivers. US Geological Survey Professional
Paper 1286.
Effects of Dams and Gravel Mining on Rivers 551
