washington geology, september 1998, vol. 26, no. 2/[email protected]...

NA

TU

RA

L

RECLAMATION ISSUE

� Flood plains, salmon habitat, and sand and gravel mining, p. 3

� Reclamation of flood-plain sand and gravel pits as off-channelsalmon habitat, p. 21

� Innovations and trends in reclamation of metal-mine tailings, p. 29

� Surface Mine Reclamation Awards, p. 43

� Early Miocene trace fossils from southwest Washington, p. 48

� Radiocarbon ages of probable coseismic features from theOlympic Peninsula and Lake Sammamish, Washington, p. 59

� Notes on the new Washington State fossil, Mammuthus columbi, p. 68

RE

SO

UR

CE

S

WASHINGTONGEOLOGY

valleywall

gravel pit lakes

point ofavulsion

1995

upperpit

1995avulsion

lowergravel pit

lakes

1996 avulsionand present path

abandoned

mainstem

channel

valleywall

point ofavulsion

1996

East Fork Lewis River

avulsion rejoinsmain stem

RIVER AVULSION INTO FLOOD-PLAIN GRAVEL PIT MINES 1995/96

VOL. 26, NO. 2/3SEPTEMBER 1998

2 Washington Geology, vol. 26, no. 2/3, September 1998

WASHINGTONGEOLOGY

Vol. 26, No. 2/3September 1998

Washington Geology (ISSN 1058-2134) is published four times eachyear by the Washington State Department of Natural Resources,Division of Geology and Earth Resources. This publication is free uponrequest. The Division also publishes bulletins, information circulars,reports of investigations, geologic maps, and open-file reports. A list ofthese publications will be sent upon request.

DIVISION OF GEOLOGY AND EARTH RESOURCES

Raymond Lasmanis, State GeologistJ. Eric Schuster, Assistant State GeologistWilliam S. Lingley, Jr., Assistant State Geologist

Geologists (Olympia)Joe D. DragovichWendy J. GerstelRobert L. (Josh) LoganDavid K. NormanStephen P. PalmerPatrick T. PringleKatherine M. ReedHenry W. (Hank) SchasseTimothy J. WalshWeldon W. Rau (volunteer)

Geologist (Spokane)Robert E. Derkey

Geologists (Regions)Garth Anderson (Northwest)Charles W. (Chuck) Gulick

(Northeast)Rex J. Hapala (Southwest)Lorraine Powell (Southeast)Stephanie Zurenko (Central)

Senior LibrarianConnie J. Manson

Library InformationSpecialistLee Walkling

EditorKatherine M. Reed

Senior Cartographer/GIS SpecialistCarl F. T. Harris

CartographersKeith G. IkerdAnne Heinitz

Production Editor/DesignerJaretta M. (Jari) Roloff

Data CommunicationsTechnicianJ. Renee Christensen

Administrative AssistantJanis G. Allen

Regulatory ProgramsAssistantMary Ann Shawver

Clerical StaffPhilip H. DobsonCathrine Kenner

MAIN OFFICEDepartment of Natural ResourcesDivision of Geology

and Earth ResourcesPO Box 47007Olympia, WA 98504-7007

Phone: (360) 902-1450Fax: (360) 902-1785

(See map on inside back coverfor main office location.)

Internet Connections:Library inquiries:

[email protected]@wadnr.gov

Subscriptions/address changes:[email protected]

FIELD OFFICEDepartment of Natural ResourcesDivision of Geology

and Earth Resources904 W. Riverside, Room 209Spokane, WA 99201-1011

Phone: (509) 456-3255Fax: (509) 456-6115E-mail: [email protected]

Publications available from theOlympia address only.

Copying is encouraged, but pleaseacknowledge us as the source.

Printed on recycled paper.Printed in the U.S.A.

URL: http://www.wa.gov/dnr/htdocs/ger/ger.html

Cover Photo: The East Fork Lewis River, Clark County, southwest Wash-ington, in June 1994. Ponds created by gravel mining here cover approxi-mately 200 acres on both sides of the river and are about 30 ft deep. The val-ley is approximately 1 mile wide in this area. During 1995 and 1996, the riveravulsed through a gravel pond (shown by solid line) and abandoned about1,700 ft of channel. Later avulsion forced the river into the lower gravel pitsand abandoned an additional 3,200 ft of channel. Small arrows indicate the1994 path of the river (today’s abandoned channel). See related article, p. 3.

Association of American StateGeologists Present Results ofSTATEMAP Program

Raymond Lasmanis, State Geologist

Washington Department of Natural Resources

Division of Geology and Earth Resources

PO Box 47007, Olympia, WA 98504-7007

On March 18, 1998, the Association of American StateGeologists hosted a reception in the foyer of the RayburnHouse Office Building in Washington, D.C. Members of Con-gress and their staff, committee staff, and representatives ofagencies that use geologic maps were invited guests. Forty-five state geologists participated, and each state presented aposter displaying a STATEMAP product and derivative infor-mation that addresses societal needs.

I presented the poster for Washington State. It consisted oftwo derivative maps derived from a digital version of the1:100,000-scale Seattle quadrangle. Developed using Depart-

ment of Natural Resources GIS capabilities, one of the postermaps showed locations of rock, sand, and gravel resources.The second map showed earthquake liquefaction hazard areasranked by the potential severity of the hazard. The narrativewith the poster listed various applications of digital geologicmap data and Division clients, such as:

� Forest resource management – WeyerhaeuserCorporation, Simpson Timber, Boise Cascade

� Fisheries management – Skagit System Cooperative,Washington Association of Federated Tribes

� Mineral resource exploration – BHP Minerals, WesternMining Corporation

� Federal and State agencies – U.S. Forest Service,Washington Department of Ecology

� Local governments – Clark County, Lewis County,Tacoma-Pierce County Health Department

� Research – Central Washington and Evergreen Stateuniversities, Pacific Northwest Laboratory

State Geologist Ray Lasmanis with Washington STATEMAP poster.

Continued on page 75

Washington Geology, vol. 26, no. 2/3, September 1998 3

Flood Plains, Salmon Habitat, andSand and Gravel Mining

David K. Norman

Department of Natural Resources


PO Box 47007, Olympia, WA 98504-7007

C. Jeff Cederholm


Resource Planning and Asset Management Division

PO Box 47014, Olympia, WA 98504-7014

William S. Lingley, Jr.



PO Box 47007, Olympia, WA 98504-7007

INTRODUCTION

People who live or work along rivers and their flood plainsknow these are hydrologically and biologically dynamic envi-ronments. Rivers flood and shift their courses from time totime, resulting in natural cycles of erosion and deposition ofsand and gravel. The river and its banks are also home to manyfish and wildlife species. In this age of rapid land development,however, people have turned to rivers (dredging and gravel barmining) and flood plains (gravel pit lakes) as sources of sandand gravel for construction aggregates.

Gravel mining practices such as channel dredging are typi-cally done with either a dragline or suction dredge (Fig. 1) andare conducted mostly in the Columbia River, where a shippingchannel is maintained, and the Cowlitz River, where large vol-umes of sediment from the 1980 Mount St. Helens eruption arestill being deposited. Gravel bar mining (scalping) (Fig. 2 ) isperformed in many Washington rivers for aggregate and in anattempt at flood control (Collins, 1996). Miners no longer diginto the banks of rivers (Fig. 3). Also, the Army Corps of Engi-neers, state agencies, and some local governments have recog-nized the environmental effects of in-stream mining. TheGrays Harbor County Planning Department, for example, hasmandated reduced rates of gravel removal on the Satsop, Wy-noochee, and Humptulips Rivers (Collins and Dunne, 1990).(See Fig. 6.)

Flood-plain mining, digging gravelpit lakes adjacent to rivers, is a commonactivity in Washington. “Flood plain”as defined in Washington’s ShorelineManagement Act (Revised Code ofWashington [RCW] 90.58) means the100-year flood plain, the area suscepti-ble to flooding by a stream for whichthere is a one percent chance of inunda-tion in any given year (Washington De-partment of Ecology, 1994). The 100-year flood plain definition must be usedwith caution because the small flood-frequency data sets result in large po-tential for error in identifying the 100-year event and determining the area af-fected (Mount, 1995). Also, as water-sheds become more developed androofs and roads present large areas ofimpermeable surfaces, high-intensitypeak runoff occurs more frequently andwith greater volume (Booth, 1991).The definition of the 100-year floodplain is important in Washington be-cause most regulation of developmentand mining depends on where this“line” is drawn. Furthermore, the geo-

morphic flood plain, or the area where fluvial erosion has cre-ated a flat valley, is generally much larger than the “calculated”100-yr flood plain and is where mining occurs.

Flood plains support diverse plant and animal populations,as well as much of our agricultural system (Teskey and Hink-ley, 1977). Vegetation along river banks provides habitat formost wildlife. Significantly, the riverine environment is hometo the many species of Pacific salmon (Oncorhynchus sp.) in-digenous to Washington (Palmisano and others, 1993; Wash-ington Department of Fish and Wildlife, 1997; Sedell andLuchessa, 1982). The riparian area is where aquatic and terres-trial ecosystems interact and is essential to both fish and wild-life. Streamside plants shade the water, help moderate watertemperature, and promote stream-bank stability, as well as pro-viding the organic nutrient load in the aquatic ecosystem.These plants are the source of large in-stream woody debristhat provides refuge and food sources (Washington Depart-ment of Natural Resources, 1996). Salmon use gravel channelsfor spawning and clay-bottom channels, with their richsubaqueous flora, for overwintering, feeding, and refuge.Side-channels (yazoo tributaries) and oxbow lakes are types ofwallbase channels (Fig. 4A) or off-channel sites that are pri-mary overwintering habitats for juvenile coho salmon and cut-throat trout (Peterson, 1980; Cederholm and Scarlett, 1981;

plume

Figure 1. A suction gravel dredge and barges at work in the Chehalis River in the mid-1960s.

This type of mining now occurs chiefly in the Columbia and Cowlitz Rivers to remove sediment de-

posited after the Mount St. Helens eruption. Note the suspended sediment plume moving down-

stream. (Photo by Lloyd Phinny, Washington Department of Fisheries.)


Peterson and Reid, 1984; Cederholmand Scarlett, 1991). Wallbase channelscommonly look like swamps, ponds, orsmall tributaries and are connected tothe main stem of the river by a smallstream called an ingress/egress channel.

Rivers in Washington generallyhave steep gradients and V-shapedcross sections in their upper reaches.Headward erosion occurs where streamgradients rapidly steepen, as in theOlympic and Cascade mountains. Asthe rivers approach the ocean or findtheir way across the Columbia Basin,gradients decrease, valleys broaden,flow velocity decreases, and gravelsare deposited. Deposition is greatest atthe gradient changes. The river accom-modates this aggradation by lateral(horizontal) migration. Traces of chan-nel shifts and stream meanders arecommon flood-plain features; oxbowlakes and side channels are testament tothe history of this activity (Fig. 4A,B).

Flood-plain gravel mining gener-ally occurs where the gravel bedloadhas been deposited, for example, at gra-dient changes or above or below topo-graphic constrictions through which ariver flows (such as Union or Selahgaps on the Yakima River). In Wash-ington, many gravel pits near rivershave been excavated for fill materialfor major highway projects, a processthat further complicates flood-plainfunctions. Examples are I-90 along theYakima River, I-5 where it crosses theSkookumchuck River at Centralia, I-82near metropolitan Yakima on the Ya-kima and Naches Rivers, and in severalplaces along the East Fork Lewis Rivernear Vancouver. Gravel sources lo-cated near the point of use significantlyreduce the cost of the aggregate.

Seeking the lowest cost material,gravel miners commonly choose to ex-cavate large, deep ponds adjacent toactive river channels (Fig. 4A). Thesepits have the potential to significantlychange the physical and ecologicalfunction of flood plains (Collins, 1996,1997). Wherever a channel shifts into agravel pit or multiple pits that are largerelative to the scale of the flood plainand the river’s sediment transport re-gime, natural recovery of original floodplain environment and similar channelmorphology could take millennia (Collins, 1997). The time forrecovery is highly dependent on the availability of sediment,particle size, gradient, and the size of excavations to be filled.

Regardless of the best planning and intentions, impacts offlood-plain mining may simply be delayed until the river iscaptured by the gravel pit. While capture may not occur in thenext 100-year flood event, it is likely to occur in the future as

development and consequent flood magnitude increase. In thelong term, stream capture by gravel pits is a near certainty. Be-cause the gravel pits have a lower base elevation, there is riskof rapid channel change into the pits during high flows, a pro-cess termed avulsion. The flooded pits “capture” the stream.The effects of avulsion are similar to those of in-stream miningdiscussed in Evoy and Holland (1989), Collins and Dunne

gullies, ridges,and scrape marks

point barmining

Figure 2. Scalping a point bar in the Wynoochee River in about 1965. Current rules require that

the post-mining ridges and gullies on the gravel bar to be graded to 2 percent toward the river and

that no areas where fish could be trapped remain. Gravel bar mining has been restricted in several

rivers, such as the Wynoochee where over-mining of gravels was proven by gravel bedload trans-

port studies conducted for Grays Harbor County (Collins and Dunne, 1990). (Photo by Lloyd

Phinny, Washington Department of Fisheries.)

sedimentpond

sediment beingreleased from pond

sedimentplume

Wynoochee River

gravel stockpilesand processing area

Figure 3. This plume of turbid water (lighter gray area at top) from a gravel mine on the Wy-

noochee River was created by bar scalping and shoreline mining, no longer practiced in Washing-

ton. The small pond at the left center of the photo was used as a settling pond. Note turbid water

discharge from the sediment pond to the river at the center of photo. (Photo taken about 1965 by

Lloyd Phinny, Washington Department of Fisheries.)


(1990), Netsch and others (1981), Kondolf and Graham Mat-thews (1993), Kondolf (1993, 1994), and Williamson and oth-ers (1995a,b). They may include:

� lowering the river bed upstream and downstream ofmining operations, causing river bed erosion and (or)channel incision and bank erosion and collapse,

� eroding of footings for bridges or utility rights-of-way,

� changing aquatic habitat,

� unnaturally simplifying the complex natural streamsystem,

� increasing suspended sediment, and

� abandoning reaches of spawning gravels or damagingthese gravels by channel erosion or deposition of siltsin spawning and rearing reaches.

Flood-plain mine lakes, the loss of natural flood-plain habi-tat, and the isolation of the flood plain from its river by armor-ing and dikes that protect against avulsion are semipermanentconsequences of flood-plain mining. Careful reclamation mayrestore some of the previous function of a flood plain; how-ever, it may be impossible to replace lost habitat in the nearterm if a substantial amount of a flood plain has been convertedto pit lakes (Collins, 1996, 1997).

Careful siting, planning, limiting mining, a thoroughhydrogeological analysis, use of alternative resources, and in-novative reclamation can mitigate and reduce some mining im-pacts. This article describes river processes and mining opera-tions and discusses some examples of flood-plain mines. Rec-lamation of these mines is the subject of the article by Norman(this issue, p. 21).

t

operatingsand & gravel

pit/lake

draglinebucket

pit/lake 25–100 deep¢

alluvial gravels(hyporheic zone)

not to scale

water table

bedrock

wetlands

geomorphic flood plain

wallbasechannel(pond)

abandonedchannel

sidechannel

draglinetower

& stockpiles

abandonedgravel pit

oxbowlake

ingress/egresschannel

riparianzone

erosionpoin

tbar

dike

main stemriver channel10–15 deep¢

water table

wallbasechannel(pond)

tailho

ld

berm

dike

/lev

ee/b

erm

wallbasechannelcutoff

pointbar

leve

e/

dike

/

wallbasechannel

t

Figure 4. A. Diagram showing a flood plain

with an active, meandering river channel and

dragline digging a gravel pit. Note the rela-

tion of the pit to the water table and main

stem river channel with dikes built to protect

the pit and channelize the river. Also shown

are point bars, wallbase channels (including

side channels, ponds, and oxbow lakes) and

associated wetlands, all of which are impor-

tant for salmon habitat. B. Oxbow lakes with

narrow riparian zones on the broad and flat

Chehalis River flood plain near the Chehalis

airport. Oxbow lakes such as these that are

hydraulically connected to the main stem of a

river can provide off-channel habitat for

salmon. Here, much of the riparian zone has

been removed for agriculture.

river

oxbow lake

oxbow

river

oxbow

oxbow

B

A


FLOOD-PLAIN HYDRAULIC PROCESSES

Many Washington river valleys reflect millions of years of flu-vial processes and in some places modification by glaciers.Most lowland river valleys are filled with alluvial sand, gravel,silt, and clay, which have been spread out across the adjacentvegetated flood plains wherever flood flows have left the mainchannel. Small irregularities in the channel cause local varia-tions in flow velocity, which in turn contribute to variations inerosion and sediment deposition. From these variations, theriver begins slow lateral displacement, depositing on onestream bank while eroding, scraping, and undercutting theother bank as it gradually migrates from side to side across itsvalley floor. The valley widens over time by erosion, particu-larly on the outside of river bends and where the flow impingesagainst the valley walls. The river’s course appears snakelikewhen viewed from above (Leopold and others, 1964).

The water strikes with greatest force against the outer (con-cave), downstream bank of the meander bend, causing erosion.Flow is slower on the inside (convex) bank of a meander wheredeposition occurs. These deposits are called point bars, gravelbars, or lateral accretion surfaces. Over time, meanders can en-large sideways and shift downstream by a combination of ero-sion and sedimentation processes. A meandering river gener-ally retains the same approximate channel width, but a givenriver channel may eventually occupy most parts of a flood plainthroughout its geologic history (Fig. 4A,B).

Meander bends do not enlarge indefinitely because theybecome sizable detours for the flowing water. Eventually, theriver will cut off a meander and abandon the longer, curvedchannel for the shorter, more direct route. The cutoff meandersections are termed oxbows. Where they are filled with water,they are called oxbow lakes or ponds (Fig. 4A,B). Meandercutoffs most commonly occur during floods, when large vol-umes of water have enough additional momentum to carve theshortcut and bypass the meander bend.

As river profiles or individual reaches become stable over aperiod of years, they achieve a balance between erosion anddeposition. At equilibrium, the river ora reach is a graded stream, one in whichthe slope, velocity, and discharge com-bine to transport its sediment load withneither erosion nor sedimentation(Mackin, 1948). That balance is gov-erned by the elevation of its base level(for example, an adjacent ocean or lakelevel) and the many physical and bio-logical factors that control equilibrium.Base levels and the stream’s equilib-rium profile can be altered by humanintervention, such as dam building, ac-tivities that cause river avulsion, or re-moval of gravel bars. Gradually thestream’s profile will reach a new equi-librium, but this may change the distri-bution of erosion and deposition sitesand also the shape of the channel.

A flood plain is a flat or gently slop-ing region along a stream channel, typi-cally extending laterally to the adjacentslopes, onto which the stream expandsduring floods (Fig. 4A,B). With time,lateral erosion associated with mean-dering, channel sediment deposition,and additional overbank sedimentation

during floods combine to create a flood plain. The width of theflood plain is a function of many factors, including the size ofthe stream, relative rates of meander migration and downcut-ting, and the ease with which the valley walls are eroded. Dur-ing periods between floods, the meandering stream occupiesonly a portion of the breadth of the flood plain. Slow lateralchannel migrations over long periods of time tend to maintain asingle flood plain with a fairly level surface. In aggradingreaches, where cobbles and boulders are the dominant bedloadand sediment transport rates are high, the river may take on abraided morphology. Braided rivers are characterized by achannel dividing and reuniting in an intricate interlaced fash-ion. Early maps of the state show that many rivers flowing fromthe Cascades had braided channels.

If conditions are suddenly and significantly changed, rela-tive rates of meandering and downcutting may accelerate, andthe old flood plain may be cut up by development of new, nar-rower channels. Incision can result from an increase in the riv-er’s discharge, reduction in the size of its bedload, dredging,avulsion into a gravel pit, or upstream levee construction.

As a valley widens, the stream’s flood plain also widens.The flood plain width can be limited, as it is along the en-trenched Yakima River canyon between Ellensburg and Rozadam (Figs. 5, 6), or very broad, as it is along the Chehalis Riverdownstream of Oakville, where it is more than 2 mi wide(Waitt, 1979; U.S. Army Corps of Engineers, 1970). A floodplain’s width can also be out of proportion to its river channel.For example, the Chehalis River appears to have an undersizedchannel in the wide valley along its lower reaches. During thePleistocene epoch, the Chehalis River valley was the outlet forthe enormous amounts of glacial outwash and meltwater fromthe Puget ice lobe (Bretz, 1913; Eddy, 1966). The modern Che-halis River drains a much lower and smaller watershed area.The upper reaches of rivers such as the Stillaguamish, Cowlitz,and east fork of the Lewis (Mundorff, 1984) flow in more re-stricted glacier-carved valleys; only in the lower parts of thesevalleys do they appear underfit in proportion to their overallflood-plain size.

Figure 5. Yakima Canyon near Umtanum gaging station; view to the north, toward Ellensburg.

Note the entrenched river with meander loops cut into the basalt flows. The resistant basalt con-

strains the channel, and no flood plain can develop here, and no off-channel habitat can be cre-

ated as on a flatter flood plain.


BIOLOGICAL PROCESSESIN FLOOD PLAINS

Off-Channel Habitat for Salmon

Rivers are not isolated ribbons of water flowing through a val-ley; they are connected to their valleys and flood plains by bothsurface water and shallow ground-water sources (Collins,1996). Wallbase channels (Fig. 4A,B) form on flood plainswhere runoff reoccupies abandoned channels created by themigration of main stem rivers (Peterson and Reid, 1984). Wall-base channels develop along meander scars and oxbows or be-hind gravel bars; in most places they appear as swamps, ponds,or small tributaries and are connected to the main stem (Fig. 7).

During spring and fall, juvenile salmon migrate out of mainrivers into tributaries, including wallbase channels, where theyseek refuge from high flows and turbidity for varying periodsof time (Cederholm and Scarlett, 1981; Peterson and Reid,1984; Peterson, 1982; Scarlett and Cederholm, 1984; Brownand Hartman, 1988). In fall, these movements are initiated dur-ing the first freshet (Peterson, 1980). Increased discharge andturbidity in the main river relative to clear water flowing fromwallbase channels attracts large numbers of juvenile coho andcutthroat. While in these channels, salmon take advantage of arich food supply of aquatic insects and relatively stable hydro-logical conditions. Juvenile salmon typically feed in the shal-lows and seek cover from predators in deeper water or inwoody debris complexes and emergent vegetation. The growththat the juvenile salmon immigrants (Fig. 8) are able to acquirein these habitats improves their overall size and survival rates(Peterson, 1980; 1982).

Poorly planned mining exposes these salmon habitats to po-tential avulsion. Many gravel pit mines have been establishedin active wallbase channels, sloughs, and (or) marshes (Col-lins, 1996, 1997), resulting in destruction of salmon habitat. Inrare instances, flood-plain mining has created effective off-channel habitat (see Norman, this issue, p. 21).

Benthic Insects and the Hyporheic Zone

Benthic macroinvertebrates are an important food source forjuvenile salmon (Mundie 1969). Many benthic invertebrates,such as caddis flies, mayflies, and stone flies, spend all or partof their lives in a clean gravelly substrate that has abundantoxygenated water (Pennak 1978; Merritt and Cummins, 1984).Healthy streams are characterized by a high species diversityof benthic macroinvertebrates and a moderate number of indi-viduals of any given species group (Brookes, 1989). Unhealthystreams, on the other hand, are characterized by low species di-versity and large numbers of individuals in any species group.

The hyporheic zone (Fig. 4A), the shallow unconfined aq-uifer under the flood plain that is in hydraulic continuity withthe river, may extend miles horizontally across the width of theflood plain and many yards beneath the surface (Stanford andWard, 1988). This zone provides extensive interstitial (inter-granular) habitat for benthic macroinvertebrates. The hypo-rheic zone serves as a refuge for benthic insects duringdroughts and high-flow events, and it is capable of re-supplying the population of an area at and immediately belowthe stream bed once conditions in the stream improve. Thiszone plays a major ecological role in streams, especially those

Clea

rwate

r R.

Satsop

Rive

rPA

CIF

ICO

CE

AN

Yakima

River

Elu

m

Skookumchuck

River

Newa

ukum

R.

Nor

th

Fork

South

For

k

Naches

River

Wes

tF

ork

East

Fork

Cle

R.

Salmon Cr.

Oakville

Rive

r

Cowli

tz

Hum

ptul

ips

Riv

erW

ynooch

eeR

iver

Chehalis

River

Columbia

River

East Fork

Lewis RiverCo

lumbia

River

Columbia R

iver

90

90

405

5

5

205

101

101

12

1212

97

82

82

2

97

2

101

14

22

281

28

14

5044

MosesLake

Kennewick

PascoRichland

Olympia

Kelso

Longview

Vancouver

CleElum

Centralia

YakimaMorton

Chehalis

White Salmon Goldendale

Raymond

OREGON

Forks

Bremerton

Everett

NorthBend

Leavenworth

Wenatchee

SelahGap pits

Parkerpits

La Center

Ridgefieldpits

Tacoma

Seattle

Aberdeen

ToledoMayfield

dam

Union Gap

Rozadam

Ellensburg

Toledopits

IFA Nursery

120o

119o

46o

121o122o

123o124o

47o

48o

0

0

20 mi

30 km

Figure 6. Rivers, towns, and flood-plain gravel mines in Washington referred to in this article.


that have coarse substrates where mostaquatic insects spend important parts oftheir life cycle (Ward, 1992). Overlylarge and deep flood-plain mines maysignificantly interrupt local hyporheicprocesses through removal of gravelsand (or) by changing water-table eleva-tions.

Effects of Avulsion

Salmon depend on a river system thatfunctions well. Salmon lay their eggs ingravels, where cool, oxygenated waterflows freely over the eggs, assuringtheir normal development. Generally,the most productive river substrates are those that are stableand have a wide range of particle sizes. They also produce themost diverse invertebrate populations (Brookes, 1989). If thegravel substrate is disturbed or is filled and covered by silt andclay, as can happen during floods and avulsion episodes,salmon egg nests (redds) can be eroded away or smothered.Additionally, the benthic invertebrate community is likely tochange and become less productive when the stream bed is al-tered. Where velocities are persistently high, many kinds ofbenthic macroinvertebrates may be absent.

When a river breaches a pit, the river biota can be cata-strophically changed. Water temperatures may rise duringsummer and early fall because the relatively slack water in thepits is exposed to sunlight for long periods. Water temperaturesabove 75oF may be fatal to some salmon species, and tempera-tures between 60o and 75oF can increase their metabolic ratesand stress levels. While moderate increases in water tempera-ture can increase growth rates, large increases can cause dis-

ease outbreaks and may kill significant numbers of adult andjuvenile fish (Bjornn and Reiser, 1991).

Pits that are warmer than the adjacent river may be idealhabitat for warm-water fish, such as large mouth bass or yellowperch, which are predators of juvenile salmon (Kondolf andothers, 1996). This may be particularly applicable in the lowerreaches of the Yakima River, where low flows and warmingriver water during summers is a chronic problem. Additionally,smolting juvenile salmon may become disoriented in the quietwaters of a pit that has been captured by a river (Ken Bates,Washington Department of Fish and Wildlife [WADFW], writ-ten commun., 1998).

RIVERS AND FLOOD PLAINS AFFECTEDBY MINING IN WASHINGTON

Flood-plain mining has occurred on many rivers in Washing-ton State and is highly concentrated along the Yakima River

Figure 8. Coho smolt being measured for growth after spending the winter in the pond shown in

Figure 7. Scale is in centimeters. Smolts that spend the winter in wallbase channels and ponds are

25 percent longer and weigh more than smolt that migrate immediately to the ocean (Cederholm

and Scarlett, 1991). Coho that overwinter in ponds such as this one grow larger and have a better

chance of survival once they migrate to the ocean.

location ofof insetingress/

egresschannel

wallbasechannel

pondflow

Figure 7. A wallbase channel pond on the Clearwater River that serves as off-channel habitat for salmon. This pond covers only 3 acres and has a

maximum depth of 12 ft. This remnant of a channel cutoff is partially covered with lily pads and is surrounded with a robust riparian zone. The inset

shows a close-up of the pond area.

ingress/egress

channel

wallbasechannel

pond


and its tributaries (Naches and Cle Elum Rivers), the ChehalisRiver and its tributaries (Wynoochee, Satsop, Newaukum, andSkookumchuck Rivers), and the Cowlitz River and East ForkLewis Rivers (Fig. 6)(Collins, 1996, 1997). Collins estimatesthat about one-sixth of Washington’s gravel production was re-moved from riverine sources between 1970 and 1991, and mostof this mining was located on flood plains and active gravelbars. However, this paper does not attempt to address all theflood-plain gravel pits in the state.

Many currently permitted mines cover several hundredacres of flood plain that have been excavated or are scheduledfor excavation. The depth and size of many of these mines wasrestricted by economics and the types of machinery used.

Gravel mining has created pit lakes ranging from a fewacres to several hundred acres and with depths averaging about30 ft. Several mined lakes are as deep as 90 ft, and some are asmuch as five times as deep as the adjacent river. Some of thesedeep lakes are within 50 ft of the active river channels. Thishead differential makes them highly vulnerable to river avul-sion during high flows. Rivers have avulsed at several mines inWashington (see Table 1) in the last 14 years, most of them dur-ing storms in 1995 and 1996.

Some Effects of River Diking and Channelization

Mining in depositional reaches commonly occurs on both sidesof a river (for example, mines on the north and south banks ofthe East Fork Lewis River). The typical method of attemptingto prevent a river from flooding a pit is armoring the activechannel bank with riprap or building dikes and levees. Wherebank armoring and dikes tend to minimize lateral erosion, theymay reduce the supply of sand and gravel to the active channeland prevent maintenance or creation of habitat diversity in boththe main stem river and side channels.

Channelization using bank armoring and dikes attempts toconfine the river to one main channel and prevent it from mi-grating across the flood plain. Channelization can smooth and(or) straighten a channel, a process that reduces energy dissipa-tion and moves channel and bank instability and floodingproblems (as well as increased sediment load) to downstreamreaches. When the channel or thalweg (the deepest part of thechannel) is moved or deflected by levees or dikes, the realign-ment may temporarily affect the downstream channel andbanks as the thalweg scours through salmon spawning habitatsor impinges on channel banks (WADFW, 1998).

Channelization can change flow velocity and the substrate,which may change benthic macroinvertebrate populations onwhich fish depend for food (Brookes, 1989). Additionally,where armoring interferes with natural meander patterns, anydeposition will be confined to the active channel. Sedimentmay have to be removed on a regular basis just to maintainchannel depth and flow capacity (Brookes, 1989). Gravelbuildup in channelized rivers can result in the river beingperched above adjacent flood plains—and the adjacent gravelpits—and making it prone to avulsion.

Some levees (also called flood berms) that are used to pro-tect gravel pits from flooding and avulsion are composed ofmaterials originally carried by the river to the pit site. The ma-terials are, therefore, of a size that the river can easily transportagain. These dikes can be washed away during a flood orgradually made porous as fines are washed out. Through thiswashing process a levee can become a windrow of unconsoli-dated gravel that is highly vulnerable to collapse. Channel mi-gration and single-event bank failures reduce the width of thedike and weaken it.

Additionally, dikes and levees can be damaged by dewater-ing an adjacent pit by pumping or flow though an outlet chan-nel. During floods, when the river’s level is abnormally high, alarge static head exists between the higher river water surfaceand that of the pit. Here, the potential for dike collapse andavulsion is great. The dikes are weakened by the constant headdifference. Floods that overtop dikes can allow a river to rap-idly cut through these structures as the flow drops from theriver surface into the pit, causing headward erosion. In effect,dikes and levees are only short term solutions to long termnatural processes such as channel migration and flooding.

Consequences of River Avulsion

Typically, avulsion occurs during floods. Avulsion is charac-terized by a sudden change in the course of a river that causes itto break through a low point such as a meander neck (to form anoxbow lake) or to rush into a gravel pit. Avulsion events occurin gravel pit lakes because the pit surface is lower than theriver. The old river channel can be partially or completelyabandoned in the avulsion process.

Pits are typically wider and deeper than the river (Fig. 4A)and out of proportion to the overall scale of the river (Collins,1997; Woodward-Clyde Consultants, 1980a,b). The severityof the consequences of breaching pits depends on the relativescale of the mining operation and the river. Significantamounts of material must be moved from the river bed andbanks to fill a deep pit.

Once avulsion and breaching begin, the knickpoint, wherethe river enters the pit (Fig. 9), immediately moves upstream,and the riverbed starts to be scoured (Kondolf, 1993; Kondolfand others, 1996; Collins and Dunne, 1990; Collins, 1996,1997). This scouring cuts downward and lowers the river ele-vation, a condition called incision. Avulsion into gravel pits bythe Clackamas River at Clackamas, Oregon, in February of1996 resulted in incision of more than 2 yards over a distanceof one-third of a mile upstream of the pit (Kondolf and others,1996). Hazardous consequences of this kind of erosion can beundercutting of levees, bridge supports, pipelines, and utilitytowers and other structures. For example, when a gravel pit onthe alluvial fan of Tujunga Wash near Los Angeles, California,captured the river, the knickpoint migrated upstream to under-mine nearby highway bridges (Scott, 1973).

The negative effects caused by breaching a small (5 acres orless) or shallow (10 ft) flood-plain pit may be negligible andmay add channel complexity to a channelized river. If the pit issmall and shallow relative to the river, it may quickly fill withsediment. However, even small gravel pit lakes in locationsthat are prone to avulsion or near structures can be problematic.For example, in Clark County near Vancouver where SalmonCreek crosses I-5, an avulsion in 1996 into small (


consequently will erode if they are not replenished with coarsebed material. The river, in essence, mines its own gravel bars,becomes less stable, and is inhospitable to salmon.

The time required for the river to re-establish equilibriumdepends on the size and depth of the pit, the river’s ability totransport sediment, and the availability of sediment (Wood-ward-Clyde Consultants, 1980a,b). It can also depend on theinstabilities caused by the headcut upstream or bedload deple-tions downstream. If the river has already eroded down to bed-rock in its upper reaches and sediment is not available, equilib-rium may not be re-established for a very long time.

EXAMPLES OF FLOOD-PLAIN SAND ANDGRAVEL MINING AND STREAM CAPTURE

Recent avulsion and stream capture events at gravel mines inWashington are listed in Table 1. The following paragraphsdiscuss some of these events.

Cowlitz River Upstreamof Toledo

In November 1995 after heavy rains, the Kirkendoll revetmentat river mile 38 on the Cowlitz River failed (Fig. 10). About aquarter of the Cowlitz flow rushed through the opening andacross the IFA Nursery, then found its way to a gravel pit nearthe river 3,200 ft downstream between river miles 36 and 37.The mine pit partially filled with sediment and debris in a mat-ter of days. Part of the river severed access to about 13 homesuntil residents were able to construct a temporary dike of loosesand and gravel across the breached area as waters receded.During the February 1996 (Fig. 11) event, the Cowlitz Riverbreached this temporary dike and re-established itself in the1995 channel. During the November 1995 event, the CowlitzRiver flow was greater than 68,400 ft3/sec as measured belowMayfield dam, approximately 18 mi upstream. (This value isthe maximum flow that can be recorded on that instrument, andno estimate was made as to amount of flow above that value.)During February 1996, the flow at the Mayfield dam was not ashigh, but tributaries downstream of the dam were contributingmore water than they had in November 1995 (StephanieZurenko, Washington Department of Natural Resources[WADNR], oral commun., 1997).

Because the pits behind the Kirkendoll revetment wereonly 20 acres in area and less than 20 ft deep, no obviouschanges to the grade of the river occurred. During flooding andfilling of the pit, scouring deepened the new channel and aknickpoint migrated upstream, as evidenced by a moving waveimmediately upstream of the pit and at the revetment (Stephanie

Zurenko, WADNR, oral commun., 1997). However, no agen-cies did a survey of the elevation of the river bed before and af-ter the avulsion to determine if the event had caused incision.

When the February 1996 flood receded, water was divertedback into the pre-November 1995 channel by repairing thedike. The scouring and downcutting between the IFA Nurseryand the gravel pit were then evident. According to the U.S.Army Corps of Engineers, complete abandonment of the 1995channel would likely have occurred if the revetment had notbeen replaced and reinforced with riprap in the summer of 1996because the existing Cowlitz River bed was higher than thearea protected by the revetment (Stephanie Zurenko, WADNR,oral commun., 1997). Most of the material deposited in the pitconsisted of fine sediment. The large amount of sand, silt, and

soil present was probably derived fromthe nursery. Stumps and logs were alsodeposited in the pit. Because of the de-bris, miners now use an excavatorrather than the former dragline to exca-vate gravels. The large volume of sand,soil, and debris in the pit likely makesthis gravel mine less valuable.

Cowlitz River at Toledo

Extensive gravel mining occurs alongthe Cowlitz River at Toledo (Fig. 12).Before the mid-1930s, the course of theCowlitz River near the gravel pits be-tween river miles 34 and 35 consistedof two or more channels (Collins,

Location or operation River Year Location Date mined Acres

Upstream Ridgefield pits East Fork Lewis 1995 sec. 19, T4N, R2E 1960s 6

Ridgefield pits East Fork Lewis 1996 secs. 13, 24, T4N, R2E 1980s-90 70

Salmon Creek Park ponds Salmon Creek 1996 sec. 35, T3N, R1E early 1970s 5

Pits upstream of Toledo Cowlitz 1995, 1996 sec. 10, T11N, R1W ongoing 20

Gravel pits at Toledo Cowlitz 1995, 1996 secs. 8, 17, T11N, R1W ongoing 108

Mouth of Wynoochee River Wynoochee 1984 sec. 18, T17N, R7W 1960s 20

Walker pit Yakima 1996 sec. 36, T11N, R20E ongoing 12

Parker pit Yakima 1996 sec. 20, T12N, R19E 1980s 35

Selah Gap pits Yakima 1996 sec. 31, T14N, R19E ongoing 250

Gladmar Park Yakima 1996 sec. 13, T18N, R17E 1960s 30

I-90 pits Yakima 1996 sec. 29, T18N, R18E 1960s 20

Table 1. Recent avulsions or stream captures

knickpoint

excavation

knickpointmigration

deposition within piterosion

A

B

C

Figure 9. A. A profile of a streambed before mining. B. The same

profile showing a depression excavated in the stream bed when stream

flow is not strong enough to move the bedload. The knickpoint is where

the stream profile abruptly changes. C. During high flows, or once avul-

sion and breaching begin, the knickpoint immediately retreats up-

stream as the sediment at that location is removed by erosion and de-

posited in the pit or downstream and the river tries to re-establish its

grade and equilibrium. The high flow also attacks the downstream side

of the depression where the slope also changes. (Modified from Kon-

dolf, 1993.)


oo

oo

oo

oo

oo

oo

o

oo

o

o

o

o

oo

o

o

o

oo o o o o

oo

oo

oo

++

++

+

+

++

+++

+

+

+

+

+

++

+

+

+

+

+

+

++

++

++

+

+

++

++

+

++

++

++

+

++

+

++

++

++

++

+

+

+

+

+

+

+ + + + + +

+

++

+

+

+

++

+

bre

ach

edd

ike

TO

LE

DO

gra

vel

pit

IFA

Nu

rser

y

Cow

litz

Riv

er

Kir

ken

doll

rev

etm

ent

scou

red

path

gra

vel

pit

lak

es

18

54

19

37

19

74

19

84

av

uls

ion

pa

th1

99

5/6

oo

o ++

+is

ola

ted

ho

mes

Fig

ure

10

.V

ert

ica

la

irp

ho

to(J

uly

19

96

)o

fth

eC

ow

litz

Riv

er

fro

mT

ole

do

toa

pp

roxim

ate

ly3

miu

pri

ve

r.F

low

isto

the

left

.A

rro

ws

sh

ow

the

sco

ure

dp

ath

so

fth

eN

ove

mb

er

19

95

an

dF

eb

rua

ry1

99

6fl

oo

ds.T

he

Kir

ke

nd

oll

reve

tme

ntw

as

de

sig

ne

db

yth

eU

.S.A

rmy

Co

rps

ofE

ng

ine

ers

toco

nta

ina

flo

wo

f4

8,0

00

ft3/s

ec.H

om

es

lab

ele

da

reth

ose

iso

late

db

yth

e1

99

5/1

99

6fl

oo

ds.T

he

rive

rb

roke

thro

ug

hth

ere

ve

tme

ntin

toth

e1

85

4ch

an

ne

lb

eca

use

itw

as

hig

he

rth

an

the

ad

jace

ntfl

oo

dp

lain

do

wn

str

ea

mo

fth

ere

ve

tme

nt.

Th

eg

rave

lp

ith

ad

fille

dw

ith

se

dim

en

td

uri

ng

the

No

ve

mb

er

19

95

flo

od

,so

the

rew

as

no

sig

nif

ica

ntd

ep

osit

ion

the

refr

om

the

Fe

bru

ary

19

96

flo

od

.P

ea

kfl

ow

du

rin

gth

eN

ove

mb

er

19

95

flo

od

ma

yh

ave

exce

ed

ed

the

68

,40

0ft

3/s

ec

me

asu

red

atth

eM

ayfi

eld

Da

msta

tio

n1

3m

iup

str

ea

mo

fT

ole

do

,b

utth

eF

eb

rua

ry1

99

6p

ea

kfl

ow

wa

sle

ss

tha

n4

4,9

00

ft3/s

ec

atth

ed

am

(Lu

isF

uste

,U

SG

S,o

ralc

om

mu

n.,

19

98

),w

he

rea

ve

rag

ew

inte

rfl

ow

sa

rea

pp

roxi-

ma

tely

20

,00

0ft

3/s

ec.

Th

eU

SG

Sca

lcu

late

sth

e1

00

-ye

ar

eve

nt

flo

wto

be

91

,44

8ft

3/s

ec

(Willia

ms,

19

85

).T

he

pit

sa

tT

ole

do

are

pre

do

min

an

tly

inth

e1

93

7a

nd

18

54

ch

an

ne

ls.

No

teth

at

pri

or

to1

93

7,

the

Co

wlitz

Riv

er

ha

dm

ore

tha

no

ne

ch

an

ne

la

tth

eT

ole

do

gra

ve

lp

its

(Co

llin

s,

19

96

).A

fte

r1

93

7,

the

Co

wlitz

wa

sch

an

ne

lize

d.

Arr

ow

sin

dic

ate

the

loca

tio

ns

of

the

No

ve

mb

er

To

led

od

ike

failu

rea

nd

pa

rtia

lca

ptu

reo

fth

eC

ow

litz

.D

ike

failu

reco

nti

nu

ed

un

tilw

ate

rre

ce

de

da

fte

rth

eF

eb

rua

ry1

99

6e

ve

nt.

Th

ere

isn

olo

ng

er

ag

ag

ing

sta

tio

na

tT

ole

do

,so

the

reis

no

info

rma

tio

na

bo

ut

flo

od

flo

wa

tth

ep

it.

Th

ed

ike

at

the

up

pe

re

nd

of

the

gra

ve

lp

itla

ke

wa

sre

bu

ilt

by

the

min

ing

co

mp

an

y.

Th

ep

itis

sti

llco

nn

ecte

dto

the

rive

ra

tth

elo

we

re

nd

toa

llo

wfi

sh

acce

ss.


1996). The Cowlitz River, as mapped in1854 and even in 1941, had a muchmore complicated course and floodplain than it does today (Fig. 10) (Col-lins, 1996). The river is now channel-ized and pinned between the gravel pitlakes and the bluffs on which the townof Toledo has been built. Mining occursin the 1937 and 1854 channels, the sidechannels, sloughs, marshes, and ox-bows along the river.

During the November 1995 andFebruary 1996 floods, the gravel pits,including stockpiles, and processingareas were inundated (Fig. 13). Thedike (constructed of sand and gravel)protecting the gravel pits from the riverfailed and allowed some of the water toflow into one of the ponds close to theriver. The river breached a riparianbuffer area as it exited the lower pond.After the flood, sand and gravel wasbulldozed by the mine operator to forma new dike in the area of the upperbreach, forcing the river into its formerchannel. The lower breach was leftopen to allow fish access to the ponds,which are no longer mined. The re-mainder of the site is being mined. As atthe pits upstream of Toledo, the effectsof the avulsion at Toledo on the bed ofthe Cowlitz River were not measured,and the total volume of material movedwas not estimated.

Yakima River at Parker

During the February 1996 flood, part ofthe Yakima River avulsed through thegravel pits near Parker between rivermiles 105 and 106 on the south side ofUnion Gap (Fig. 14A,B). Ice jams mayhave played a role in stream diversionsin this instance. The river entered sev-eral gravel pit ponds totaling 35 acresand averaging 10 ft deep. Miningstopped at this depth because a 100-ftthick clay layer underlies the gravel inthis reach (Len Sali, Columbia Redi-Mix, oral commun., 1998). Because thesurfaces of the gravel pit lakes werelower than the river level and the low-flow course of the river was around the

¬ Cowlitz River

Kirkendollrevetment breach

IFANursery

gravelstockpiles

floodedgravel

pits

draglinetower

1854 channel(avulsion path)isolated

houses

Cowlitz

River

screens

left open forfish passage

TOLEDO

area ofdike failure

avulsionpath

Cowli

tz R.

gravel pit

lakes

Figure 12. The gravel pits near Toledo on the Cowlitz River in 1994, prior to partial avulsion into

the pits closest to the river. Large arrows indicate the position of the channel during the flooding;

small arrows, where the dike failed and the lower end of the pit lake left open for fish passage.

Figure 13. The flooded Toledo pit, February 1996. Both the 1995 and 1996 floods inundated the

sand and gravel pits, stockpiles, and processing areas across the river from Toledo. Dams up-

stream were forced to release water at high rates to prevent the impounded water from overtop-

ping, so “flood control” efficiency was lower.

Figure 11. The area of the Kirkendoll re-vetment failure on the Cowlitz River and thenew channel established during the Febru-ary 1996 flood. The dragline tower (to the leftof the mine equipment at the center of thephoto) is about 60 ft high. The housesisolated by flood waters are in the upper left.During maximum flood flows, the fields in thelower right were inundated. (Photo by AlexNagygyor, DNR Central Region, 1996).


gravel pits, it was inevitable that theriver would avulse at flood stage.

The flood breached the dikes in twoplaces at the north end of the ponds.The miner attempted to fill the breacheswith rip rap; however, the WashingtonDepartment of Fish and Wildlifestopped the operation before fillingwas complete because the miner had nopermits. One breach was partiallyfilled, and the other remains as it wasafter it was breached during the flood.The dike openings where the channelflows into the ponds are approximately100 ft in width. On their upstream ends,the pits have begun filling with graveland are now about 4 ft deep. At thelower end of the mine, the river exitsthrough two outlets that are about 50 ftwide (Len Sali, Columbia Redi-Mix,oral commun., 1998).

The river currently has a braided,meandering course through the ponds.In this instance, the channel has be-come more complicated, and becausethe ponds were shallow, no negative ef-fects have been recorded by regulatoryagencies. However, there is no moni-toring of the site. No further extractionwill occur at the site, but crushing andasphalt batching of material brought tothe site is ongoing.

Yakima River at Selah Gap

During the flood of February 1996 atSelah Gap, ice jams, high water flow,and gravel mining combined to causeavulsion into Washington’s largest(areally) flood-plain mining area. Themined area covered approximately 250acres, and the maximum depth of exca-vation was about 25 ft (Fig. 15). Whenthe dike upstream of the pit wasbreached, the entire Yakima River flowentered the gravel pits between rivermiles 118 and 120, exiting at the down-stream end of the site (Fig. 16).

During this flood, the peak flow of the Yakima River was27,200 ft3/sec as measured at the Umtanum gaging station18 mi upstream and above Roza dam. The calculated 100-yearevent at Umtanum station is 27,459 ft3/sec (Williams and Pear-son, 1985b). Large ice jams played an undefined but likely sig-nificant role in the avulsion. Approximately 8,000 ft of channelwas abandoned when the flood waters receded.

Results of the avulsion are difficult to quantify. About 6 to8 ft of incision occurred immediately upstream. There was lo-cal knickpoint migration as evidenced by a migrating standingwave and increased bank erosion as the river began to re-estab-lish its grade (Lorraine Powell, WADNR, oral commun.,1996). We estimate that at least 300,000 yd3 of gravel wasscoured from the river bed and deposited as a layer a minimumof 6 ft thick in the excavated pits over a 33-acre area. We alsoestimate that more than 100,000 yd3 of gravel was moved from

the river bed during the flood and deposited on gravel bars andprivate lands just upstream of the pits.

Dike building (Fig. 17) forced the Yakima River back intoits pre-February 1996 channel by September 1996. As a result,river bed disruption and incision were halted or slowed. How-ever, because of concerns about such floods in the future, themining company decided to rebuild the dikes that tightly con-strain the river. These dikes increase the erosive power of theriver further downstream. The mine operator also installed anengineered armored spillway at a low point in the dike that al-lows the river to overtop it there and reduce its flow. This sillkeeps the bedload in the main channel and reduces the potentialfor incision in this area.

East Fork Lewis River near La Center

On the East Fork Lewis River, in Clark County, intensive min-ing has occurred between river miles 8 and 9. The river has not

I-82

breacheddikes

Yakima River

avul

sion

path

s

re-enters existing channel

breached dikeYakima River

I-82

Figure 14. A. The Yakima River and gravel pits near Parker in 1994. A portion of the river

avulsed through these gravel pits between river miles 105 and 106 on the south side of Union Gap.

Flow is to the right, and the airplane from which this photo was taken is positioned above Union

Gap. The point of dike breaching and the new path of river and abandoned channel are indicated

by the lines. B. In this 1998 photo, a portion of the Yakima River is flowing through the Parker

gravel pits. Photo is taken from above Interstate Highway 82. Points of breaching are indicated by

arrows. (Photo by Mary Ann Shawver, WADNR.)

A

B


been dammed and has one of westernWashington’s few remaining wild (na-tive) steelhead runs. This fish has beenlisted as threatened on this river.Gravel mine ponds on both sides of theriver cover approximately 200 acres,average about 30 ft deep, and are sitedin abandoned channels in this formerlybraided river system (Fig. 18) wherethe river issued from a V-shaped valleyonto a broad complex flood plain. The

1866

floodedpits

Yakim

a Rive

r

I-8

2

1979

xx

xx

x

xx

xx

x

x

x

x

x

xx

x

x

x

xx

x

x

x

xx

x

x

x

x

x

x

x

x

x

x

xx

x

x

xo

o

o

o

o

o

o

o

o

o

oo

oo

oo

o

o

o

o

o

o

o

oo

o

o

oo

o

o

o

oo

oo

oo

oo

oo

o

x

x

x

xx

x

x

x

x

x

x

x

oo

o

x

1936

1866

1936

old meanderscars

1972xxx

1866

1936

1991o o o

1979

1958

1996

avuls

ion

pat

h

Figure 16. Selah Gap pits between Interstate Highway 82 and the Yakima River. Most of the gravel pit lakes are located in the 1866 and 1936

channels. Traces of numerous meander scars and oxbow cutoffs from earlier, natural lateral channel migrations remain, west of the river. Note the

path of the February 1996 avulsion. In this September 1996 photo, dikes have returned the Yakima River to its February 1996 channel.

Selah Gap

YakimaRiver

pit

lakes

Yakim

a Rive

r

path

of

avu

lsio

n

I-82

chan

nel a

band

oned

after

avul

sion

Figure 15. The Yakima River and gravelpits near Selah Gap in 1994. During the floodof February 1996, breaching of the dikecaused the entire flow to enter the gravel pitsbetween river miles 118 and 120 and eventu-ally exit at the downstream end. After theflood, diking and other river training devicesbuilt by the mining company diverted theriver back into its original channel.


current pits occupy approximately two-thirds of the mile-widevalley and most of the 100-year flood plain in this reach. Sev-eral dike failures, avulsion events, and diversions have oc-curred along the river.

During November 1995, the river avulsed through a gravelpit pond at mile 9 (sec. 19) and abandoned about 1,700 ft ofchannel (Fig. 19) and spawning gravels. During February1996, the east fork flooded again. The Heissen gaging station

1854

1937

1984

avulsion path 1995/6

xxx

1954

1963o o o

x

x

x

x

x

x

x

x

x

x

x

x

x

xx

x

x

x

x

x

x

x

x xx

x

xx

x x

x

x

x

x

x

xx

x

x

oo

o

oo

oo

oo

ooo

oo

oo

o

o

oo

ooo

oo

o

oo

ooooo

ooo

oo

oo

o

o

o

o

o

o

o

oo

oo o o

oo

o

o

o

o

ooo

o o

o

o

o

o

o

oo o

oo

o

o

o

oo

oo

oo

o

o

o

o

oooo

o

o

Figure 18. Vertical air photo (1996) showing the paths of abandoned river channels from the 1995/1996 avulsions and older channel scars. The

braided channel shown in 1854 and 1858 surveys is now occupied by gravel pit lakes. At that time, the river course was much more complicated

(Collins, 1996) and exhibited a braided morphology.

dike

avulsion path

Figure 17. In the spring of 1996, the mining company rebuilt the dike along the Yakima River by dumping oversize rock in the area near the old

gravel pits; this dike eventually rechannelized the river.


located approximately 10 mi upstreamof the gravel pits was washed out in thisflood, but the peak flow was indirectlyestimated at 28,600 ft3/sec by a U.S.Geological Survey (USGS) hydrolo-gist. Earlier calculations (Williams andPearson, 1985a) indicated the 100-yearpeak flow event at Heissen to be 20,046ft3/sec. Since the 1996 flood event, the100-year event has been recalculatedby the USGS at 22,200 ft3/sec, usingthe 28,600 ft3/sec observation as themaximum peak flow (Sumioka and oth-ers, 1998). No avulsion of the east forkinto the lower gravel pits occurred dur-ing this event. However, significantbank erosion set the stage for eventualstream capture.

In November of 1996, the riveravulsed through six closely spacedgravel pit ponds on the south side of theriver and the river eventually aban-doned about 3,200 ft of channel andspawning gravels. Avulsion began nearan outside bend of the river near thehaul road from the pits (Fig. 20). Lat-eral channel migration and the series offloods severely eroded the road, whichhad acted as the dike to keep the riverout of the pits. The main stem was cap-tured and now flows through the south bank gravel pits(Fig. 21).

Some of the results of the avulsions are:

� about 10 ft of channel downcutting as the knickpointmigrated upstream,

� increased erosion along the south bank,

� abandonment of about 4,900 ft of channel wheresalmon and steelhead had spawned, and

� sluggish flow through the gravel pits.

The depth of downcutting was estimated as the height dif-ference between the bed of the abandoned channel and bed ofthe current channel. The severity of the effects may be relatedto the fact that these ponds were much deeper and wider thanthe normal river channel. We estimate that it will require morethan 2 million yd3 of sand and gravel, which must be derivedfrom the channel and banks, to refill the 70-acre pits throughwhich the river flows.

STATE MINING REGULATIONS

The principle law regulating activity on the shorelines of thestate is the Shoreline Management Act (RCW 90.58), which isadministered by cities and counties. Many local jurisdictionsalso regulate flood-plain mining under provisions of condi-tional use permits or other land-use ordinances and the GrowthManagement Act (RCW 36.70A). Several counties also havemining ordinances. The Shoreline Management Act generallyapplies to mining activities on the 100-year flood plain and as-sociated wetlands. Regulation may vary according to the “mas-ter plan” of each local jurisdiction. For instance, in LewisCounty, the Shoreline Management Act applies only to areaswithin 200 ft of the floodway of flooding that occurs with rea-sonable regularity, although not necessarily annually, or the

ordinary high water mark if the floodway is not mapped. Whenmines are proposed, the local jurisdiction becomes the LeadAgency under provisions of the State Environmental PolicyAct (SEPA). The Washington Department of Ecology (Ecol-ogy) must also give its approval to terms of a Shoreline or Con-ditional Use Permit issued by a local jurisdiction. Ecology’s1994 Shoreline Management Guidebook recommends that lo-cal governments encourage miners “to locate activities outsidethe shoreline jurisdiction”— in other words, generally 200 ftfrom the floodway, or off the 100-year flood plain.

Several other agencies of secondary importance have re-sponsibility for regulating mining on flood plains and rivers.Any work, including mining, that uses, diverts, obstructs, orchanges natural flow or the bed of any waters of the state re-quires a Hydraulic Project Approval (RCW 75.20-100) fromthe Washington Department of Fish and Wildlife (WADFW).However, because WADFW jurisdiction is restricted to watersof the state, the agency may not have jurisdiction over flood-plain mining if mining does not involve the active channel. Af-ter an avulsion has occurred WADFW would have jurisdictionover any work occurring in the pits/river. The Department ofNatural Resources administers the Surface Mine ReclamationAct (RCW 78.44), which generally requires mines to be re-claimed immediately after each segment is mined. The 1993 re-vision of this law requires that most mines in flood-plain envi-ronments be reclaimed as beneficial wetlands. As part of thereclamation plan for a mine, the act also requires that “wheremining on flood plains or in river or stream channels is contem-plated, a thoroughly documented hydrologic evaluation thatwill outline measures that would protect against or would miti-gate avulsion and erosion as determined by the department” beincluded.

The U.S. Army Corps of Engineers regulates excavation orfilling of wetlands through section 404 of the Clean Water Act

pit

lake

spit

lakes

river

abandonedchannel

point ofavulsion

upperpit

eventual dike failureand avulsion point

seen in Fig. 18

path

of

avu

lsio

n

dike

failu

re

haulroad

Figure 19. The East Fork Lewis River at flood stage in February 1996. The river had already

avulsed through an upper gravel pit at the top of photo and later avulsed into the pits at the center

and bottom right of photo. Avulsion into the downstream pits began near an outside bend of the

river located near the severely eroded haul road out of the pits. (Photo by Dan Miller, Friends of the

East Fork.)


(Title 33). Most flood-plain mineswould occur in wetlands or portions ofwetlands and would be required to ap-ply for a section 404 permit.

The Growth Management Act(RCW 36.70A) directs counties to des-ignate “mineral resource areas” and“critical areas” such as wetlands or“fish and wildlife conservation areas”that include water of the state (Lingleyand Jazdzewski, 1994). However, theGrowth Management Act has not beenfully implemented and does not applyto all counties.

As wild salmon populations dwin-dle and the fish are listed as threatenedand endangered species, the Endan-gered Species Act may play a role in fu-ture siting of riverine mining. It is notyet clear what that role will be.

A complete discussion of regula-tions for mining is given in Norman(1994).

Recommendations forPlanning and Siting

Mine site selection or plans for mineexpansion should not be made on thebasis of a perception that, becausegravel mining has occurred historicallyin an area, it is appropriate to continuemining or that river rock is the onlysource of construction aggregate in thatmarket. Potential consequences of min-ing can be severe, and the function ofthe flood plain can be lost, along withfish and wildlife habitat. Whereverpossible, large gravel mines should belocated in uplands away from the river

East Fork Lewis River

parked sparpole andportable screen

haul roadpoint ofavulsion

pitlake

A

gravel stockpiles and crusher/screens

abandonedmain stem

channel

haul road

point ofavulsion

A

drift boatgoing into pit

oldhaul road/

dike

thalweg ofabandoned channel

gravel delta depositedin old gravel pit

East F

orkLew

is Rive

r ®

Figure 20. A. Before avulsion of the EastFork Lewis River. In the center of this 1990photo are the bank and haul road that werebreached in 1996. The river is at the topright. A gravel pit lake is at the lower left. Apole and portable screen are at the left.B. After avulsion, the haul road/dike hasbeen breached, and the river flows intogravel pits. The remains of the haul road canbe seen on the left across the river. “A” indi-cates same spot on each photo. The aban-doned channel of the river is shown in thecenter right of photo. Approximately 10 ft ofincision has occurred immediately upstreamof where the river entered the pits. TakenFebruary 1998 from approximately the sameposition as A. C. This February 1998 viewshows the present position of the East ForkLewis River; the drift boat is entering theformer pit area. At the top right is the rem-nant of the old haul road. The gravels andcobbles in the foreground are the old riverbed after avulsion. About 10 ft of incision hasoccurred here, on the basis of the differencebetween the height of the abandoned chan-nel thalweg and the present-day channel.

A

B

C


valley floors. A poor second choice is to locate mining on ter-races and the inactive flood plain, that is, above the 100-yearflood plain. In Washington, upland deposits offer ample rocksupplies. Mining these deposits eliminates potential for streamcapture or river avulsion. Furthermore, pits in these locationshave a good potential for successful long-term reclamation.

In parts of Europe and the eastern United States, gravel isslowly being replaced by crushed quarry rock for use as con-struction aggregate. In these areas, the transition is occurringmainly because limited gravel resources are nearly depleted.However, substitution of crushed quarry rock for gravel hasdistinct environmental advantages in that considerably morerock is produced from quarries for the surface disturbance andquarries can easily be located away from flood plains and aqui-fers. The relatively small disturbance occurs because mostquarries contain 100 percent usable rock, whereas gravel de-posits have high porosity (meaning less material per volume)and fines that are not economic.

If mine plans call for sites on flood plains, then wide, topo-graphically higher, and thickly vegetated buffers should beconsidered as a means of reducing the probability of river avul-

sion in the near term. In some places, buffers may be effectivebecause the river may not utilize the entire flood plain. How-ever, in most instances, buffers only delay the inevitable. De-termining an adequate distance between the flood-plain mine-pit lake and the river will depend on understanding the rate ofriver meandering and the risk of avulsion.

If flood-plain mining is approved, the main goals of plan-ning, siting, and reclamation are:

� The mining should not increase the potential for river avul-sion.

� Fish and wildlife habitat should be protected.

� Riparian areas should be protected, both to provide habitatand to improve flood-plain stability.

� Reclamation should ultimately enhance salmon habitat(Norman, this issue, p. 21).

� If there is potential for migration of the river into a gravelpit, the site must be reclaimed in a way that is hydrologi-cally compatible with the adjacent river.

flood-plaingravel pits

valleywall

1996 avulsion path

1995 avulsion path

valley wall

AB

C

D

E

F

abandonedriver

channel

abandoned river channel

river

chan

nel

pits

main stem

Figure 21. Vertical air photo of the East Fork Lewis River showing the course of the river in November 1997. At A, the river changed its course into

a gravel pit in 1995. The river re-entered its channel at B. The channel between A and B is now abandoned. C indicates where the haul road was

breached and the river avulsed into the gravel pits in 1996. D marks the start of the channel abandoned in this event. E is along the avulsion path as

the river finds its way through the series of ponds. The river re-enters its old channel at F. The valley walls are approximately 1 mi apart near the

gravel pits.


Environmental Analysis

Before any new mining or expansion is allowed on a floodplain, miners must make a rigorous environmental analysis ofthe mining plan. This should include, at minimum, a geohydro-logical analysis of the affected reaches of the river system. Athorough plan will also include:

� a topographic map of the existing conditions andsurrounding lands at a 2-ft contour interval and at anappropriate scale, as well as flood profiles;

� maps and cross sections that show depths and locationsof all bodies of water, the stream profile, and theelevation of the river bed measured from a permanentstation, such as a nearby bridge, near the gravel mine;

� a geomorphic analysis that identifies historic channelsand channel migration trends on the basis ofexamination of all available data, such as historic airphotos and maps, and that considers geological andartificial controls of the channel, such as armoredbanks, dikes, bridges, dams, and other mine sites;

� a detailed chronology and description of historicalprecipitation, flooding, discharge, sediment transport,including description of sediment sizes in and adjacentto the proposed mine site;

� maps of vegetation and analysis of its role in flood anderosion control, as well as a description of the relationbetween the sediment distribution and the biota,especially as it applies to bank erosion and avulsion;

� an analysis of avulsion or stream capture potential,including the consequences of stream capture, channelincision, and scouring;

� an analysis of potential damage to neighboringproperties, fish and wildlife habitat, bridges, andrights-of-way and the effects of existing or proposeddikes and levees and their long-term maintenance;

� an analysis of channel stability, magnitude andfrequency of the 5-, 10-, 25-, and 100-year floods,channel and flood-plain hydraulics near the proposedmine site, and any previous stream capture events; flowpaths for the 5-, 10-, 25-, and 100-year floods beforeand after the mining project;

� a carefully documented study of potential impacts tosalmon species listed under the Endangered SpeciesAct.

In summary, because flood-plain gravel pits are in thedynamic riverine environment, the environmental analysisshould be undertaken with great care, taking into account long-term stability of the site, and include proposals for enhancingor restoring the site’s fish and wildlife habitat.

ACKNOWLEDGMENTS

Funding for this project was provided by the U.S. Environmen-tal Protection Agency (USEPA) through the Tri-State Agree-ment of Oregon, Idaho, and Washington to share and developtechnical information regarding mining. We thank Nick Cetoand Bill Riley (USEPA) for helpful discussions, Brian Collinsfor discussions regarding riverine mining, and the followingfor discussions, reviews, and comments: Peter Wampler andFrank Schnitzer, Oregon Department of Geology and MineralIndustries; Ray Lasmanis, Matt Brunengo, Lorraine Powell,

Stephanie Zurenko, Rex Hapala, and Larry Dominguez,WADNR; Don Samuelson, Grays Harbor College; Ken Bates,WADFW; Al Wald, Washington Department of Ecology; Les-lie Lingley, U.S. Forest Service; and Beth Norman, Pierce Col-lege. We thank Jari Roloff for help with editing and graphicsand Keith Ikerd for drafting Figure 6.

REFERENCES CITED

Bjornn, T. C.; Reiser, D. W., 1991 Habitat requirements of salmonidsin streams. In Meehan, W. R., editor, Influences of forest andrangeland management on salmonid fishes and their habitats:American Fisheries Society Special Publication 19, p. 83-138.

Booth, D. B., 1991, Urbanization and the natural drainage system—Impacts, solutions, and prognoses: Northwest EnvironmentalJournal, v. 7, no. 1, p. 93-118.

Bretz, J H., 1913, Glaciation of the Puget Sound region: WashingtonGeological Survey Bulletin 8, 244 p., 3 plates.

Brookes, Andrew, 1989, Channelized rivers—Perspectives for envi-ronmental management: Wiley, 326 p.

Brown, T. G.; Hartman, G. F., 1988, Contribution of seasonallyflooded lands and minor tributaries to the production of cohosalmon in Carnation Creek, British Columbia: American Fisher-ies Society Transactions, v. 117, no. 6, p. 546-551.

Cederholm, C. J.; Scarlett, W. J., 1981, Seasonal immigrations of ju-venile salmonids into four small tributaries of the ClearwaterRiver, Washington, 1977–1981. In Brannon, E. L.; Salo, E. O.,editors, Proceedings of the salmon and trout migratory behaviorsymposium: University of Washington School of Fisheries, p. 98-110.

Cederholm, C. J.; Scarlett, W. J., 1991, The beaded channel—A low-cost technique for enhancing winter habitat of coho salmon. InColt, John; White, R. J., editors, Fisheries bioengineering sympo-sium: American Fisheries Society Symposium 10, p. 104-108.

Collins, B. D., 1996, Geomorphology and riverine gravel removal inWashington State. In Booth, D. B., Geology and geomorphologyof stream channels—Course manual: University of WashingtonCenter for Urban Water Resources Management, p. IX-1–IX-45.

Collins, B. D., 1997, Application of geomorphology to planning andassessment of riverine gravel removal in Washington. In Booth,D. B., Geology and geomorphology of stream channels—Coursemanual: University of Washington, p. IX-1–IX-46.

Collins, B. D.; Dunne, Thomas, 1990, Fluvial geomorphology and

river-gravel mining—A guide for planners, case studies included:

California Division of Mines and Geology Special Publication 98,

29 p.

Eddy, P. A., 1966, Preliminary investigation of the geology and

ground-water resources of the lower Chehalis River valley, and

adjacent areas, Grays Harbor County, Washington: Washington

Division of Water Resources Water-Supply Bulletin 30, 70 p.,

2 plates.

Evoy, Barbara; Holland, Mel, 1989, Surface and groundwater man-agement in surface mined-land reclamation: California Divisionof Mines and Geology Special Report 163, 39 p.

Kondolf, G. M., 1993, The reclamation concept in regulation ofgravel mining in California: Journal of Environmental Planningand Management, v. 36, no. 3, p. 395-406

Kondolf, G. M., 1994, Environmental planning in regulation and man-agement of instream gravel mining in California: Landscape andUrban Planning, v. 29, no. 2-3, p. 185-199.

Kondolf, G. M.; Graham Matthews, W. V., 1993, Management ofcoarse sediment on regulated rivers: University of California Wa-ter Resources Center Report 80, 128 p.


Kondolf, G. M.; Vick, J. C.; Ramirez, T. M., 1996, Salmon spawning

habitat rehabilitations in the Merced, Tuolumne, and Stanislaus

Rivers, California—An evaluation of project planning and per-

formance: University of California Water Resources Center Re-

port 90, 147 p.

Leopold, L. B.; Wolman, M. G.; Miller, J. P., 1964, Fluvial processesin geomorphology: W.H. Freeman and Company, 522 p.

Lingley, W. S., Jr.; Jazdzewski, S. P., 1994, Aspects of growth man-agement planning for mineral resource lands: Washington Geol-ogy, v. 22, no. 2, p. 36-45.

Mackin, J. H., 1948, Concept of the graded river: Geological Societyof America Bulletin, v. 59, no. 5, p. 463-512.

Merritt, R. W.; Cummins, K. W., editors, 1984, An introduction to theaquatic insects of North America; 2nd ed.: Kendall/Hunt Publish-ing Co., 722 p.

Mount, J. F., 1995, California rivers and streams—The conflict be-tween fluvial process and land use: University of California Press,359 p.

Mundie, J. H., 1969, Ecological implications of the diet of juvenilecoho in streams. In Northcote, T. G., editor, Symposium onsalmon and trout in streams: University of British Columbia Insti-tute of Fisheries H. R. MacMillan Lectures in Fisheries, p. 135-152.

Mundorff, M. J., 1984, Glaciation in the lower Lewis River basin,southwestern Cascade Range, Washington: Northwest Science,v. 58, no. 4, p. 269-281.

Netsch, Norval; Rundquist, L. A.; Moulton, L. L. 1981, Effects offloodplain gravel mining in Alaska on physical factors importantto salmon spawning. In Washington Water Research Center, Pro-ceedings from the conference, Salmon-Spawning Gravel—A Re-newable Resource in the Pacific Northwest?: Washington WaterResearch Center Report 39, p. 154-167.

Norman, D. K., 1994, Surface mining in Washington—Regulatory re-sponsibilities of federal, state, and local government agencies,1994: Washington Division of Geology and Earth ResourcesOpen File Report 94-4, 26 p.

Norman, D. K., 1998, Reclamation of flood-plain sand and gravel pitsas off-channel salmon habitat: Washington Geology, v. 26, no. 2/3, p. 21-28.

Palmisano, J. F.; Ellis, R. H.; Kaczynski, V. W., 1993, The impact ofenvironmental and management factors on Washington’s wildanadromous salmon and trout; Summary and recommendations:Washington Forest Protection Association; Washington Depart-ment of Natural Resources, 371 p.

Pennak, R. W., 1978, Fresh-water invertebrates of the United States;2nd ed.: John Wiley & Sons, Inc., 803 p.

Peterson, N. P., 1980, The role of spring ponds in the winter ecologyand natural production of coho salmon (Oncorhynchus kisutch) onthe Olympic Peninsula Washington: University of WashingtonMaster of Science thesis, 96 p.

Peterson, N. P., 1982, Population characteristics of juvenile cohosalmon (Oncorhynchus kisutch) overwintering in riverine ponds:Canadian Journal of Fisheries and Aquatic Sciences, v. 39, no. 9,p. 1303-1307.

Peterson, N. P.; Reid, L. M., 1984, Wall-base channels—Their evolu-tion, distribution, and use by juvenile coho salmon in the Clear-water River, Washington. In Walton, J. M.; Houston, D. B., edi-tors, Proceedings of the Olympic wild fish conference: PeninsulaCollege Fisheries Technology Program, p. 215-225.

Scarlett, W. J.; Cederholm, C. J., 1984, Juvenile coho salmon fall-winter utilization of two small tributaries of the Clearwater River,Jefferson County, Washington. In Walton, J. M.; Houston, D. B.,editors, Proceedings of the Olympic wild fish conference: Penin-sula College Fisheries Technology Program, p. 227-242.

Scott, K. M., 1973, Scour and fill in Tujunga Wash—A fanhead valleyin urban southern California—1969: U.S. Geological Survey Pro-fessional Paper 732-B, 29 p.

Sedell, J. R.; Luchessa, K. J., 1982, Using the historical record as anaid to salmonid habitat enhancement. In Armantrout, N. B., edi-tor, Acquisition and utilization of aquatic habitat inventory infor-mation—Proceedings of a symposium held 28–30 October, 1981,Portland, Oregon: American Fisheries Society, p. 210-223.

Stanford, J. A.; Ward, J. V., 1988, The hyporheic habitat of river eco-systems: Nature, v. 335, no. 6185, p. 64-66.

Sumioka, S. S.; Kresch, D. L.; Kasnick, K. D., 1998, Magnitude andfrequency of floods in Washington: U.S. Geological SurveyWater-Resources Investigations Report 97-4277, 91 p., 1 plate.

Teskey, R. O.; Hinkley, T. M., 1977, Impact of water level changes onwoody riparian and wetland communities; Volume 1, Plant andsoil responses: U.S. Fish and Wildlife Service Biological ServicesProgram FWS/OBS-77/58, 30 p.

U.S. Army Corps of Engineers, 1970, Flood plain information—Ya-kima and Naches Rivers Yakima–Union Gap, Washington: U.S.Army Corps of Engineers, 57 p.

Waitt, R. B., 1979, Late Cenozoic deposits, landforms, stratigraphy,and tectonism in Kittitas Valley, Washington: U.S. GeologicalSurvey Professional Paper 1127, 18 p.

Ward, J. V., 1992, Aquatic insect ecology; Volume 1, Biology andhabitat: John Wiley & Sons, Inc., 438 p.

Washington Department of Ecology, 1994, Shoreline managementguidebook; 2nd ed.: Washington Department of Ecology Shore-land and Coastal Management Program, 93-104A, 626 p.

Washington Department of Fish and Wildlife, 1997, Wild salmonidpolicy—Draft environmental impact statement, recommended al-ternative justification statement: Washington Department of Fishand Wildlife, 1 v.

Washington Department of Fish and Wildlife, 1998, Integrated bankprotection guidelines—Draft: Washington Department of Fishand Wildlife, p. 1-1–1-6.

Washington Department of Natural Resources, 1996, Draft habitatconservation plan, March 1996: Washington Department of Natu-ral Resources, 1 v.

Williams, J. R.; Pearson, H. E., 1985a, Streamflow statistics anddrainage-basin characteristics for the southwestern and easternregions, Washington; Volume I, Southwestern Washington: U.S.Geological Survey Open-File Report 84-145A, 424 p., 1 plate.

Williams, J. R.; Pearson, H. E., 1985b, Streamflow statistics anddrainage-basin characteristics for the southwestern and easternregions, Washington; Volume II, Eastern Washington: U.S. Geo-logical Survey Open-File Report 84-145B, 662 p., 1 plate.

Williamson, K. J.; Bella, D. A.; Beschta, R. L.; Grant, Gordon;Klingeman, P. C.; Li, H. W.; Nelson, P. O., 1995a, Gravel distur-bance impacts on salmon habitat and stream health; Volume I—Summary Report: Oregon Water Resources Research Institute[under contract to] the Oregon Division of State Lands, 49 p.

Williamson, K. J.; Bella, D. A.; Beschta, R. L.; Grant, Gordon;Klingeman, P. C.; Li, H. W.; Nelson, P. O., 1995b, Gravel distur-bance impacts on salmon habitat and stream health; Volume II—Technical background report: Oregon Water Resources ResearchInstitute [under contract to] the Oregon Division of State Lands,225 p.

Woodward-Clyde Consultants, 1980a, Gravel removal studies in Arc-tic and subarctic floodplains in Alaska; Technical report: U.S.Fis

washington geology, september 1998, vol. 26, no. 2/[email protected]...

Documents