Re-Establishing A Mobile Stream Bed Lower Clear Creek Watershed,
Shasta County, CA, USA
V.L. Finney
Water Resources Geologist, USDA Natural Resources Conservation
Service, 2121 C, 2nd St., Davis, CA.,95616
M.A. Cocke
Civil Engineer, USDA Natural Resources Conservation Service, 2121
C, 2nd St., Davis, CA.,95616
T.J. Viel
Fisheries Biologist, USDA Natural Resources Conservation Service,
No. 3 Horseshoe Lane, Weaverville, CA., 96093
ABSTRACT: Located in northern California, Lower Clear Creek's natural in-stream flow and sediment transport were disrupted initially in 1902 with the construction of a low head dam (4.6 m), McCormick-Saeltzer Dam (MCSD), and more drastically in 1963 with the construction of Whiskeytown Reservoir (86 m), an impoundment reservoir for waters being diverted from the Trinity River and Clear Creek watersheds to the Sacramento River for power generation.
The greatly reduced Clear Creek flows below McCormick-Saeltzer
Dam have created reaches of armored streambed. Under the existing
water management, flows are inadequate to mobilize the bed of
the stream and to recruit gravel from flood plain sources. One
effort to improve spawning habitat has been the placing of gravel
downstream of MCSD into Clear Creek. The purpose of this paper
is to assess the potential for regulating stream flows that optimize
spawning opportunities for chinook salmon with existing channel
and flood plain sediments.
Lower Clear Creek in northern California is a 12,700 ha sub-watershed of the Clear Creek watershed, a west bank tributary to the Sacramento River, Figure 1. Located in the northwestern portion of the upper Sacramento River Basin, Clear Creek is unique among west-side tributaries to the Sacramento River because it is a largely undeveloped perennial stream near a growing metropolitan area.
Whiskeytown Reservoir, part of the Trinity Division of the
Central Valley Project, impounds Clear Creek and also stores water
diverted from the Trinity River through the Clear Creek tunnel.
Closure of Whiskeytown Dam occurred on May 2nd, 1963. All of
the Trinity River diversion and 87 percent of the natural flow
of Clear Creek are diverted from the reservoir through the Spring
Creek Tunnel into the Sacramento River above Keswick Dam. The
remaining 13 percent of the water flow now comprises lower Clear
Creek. (Modified from Western Shasta Resource Conservation District,
1996)
2. GEOMORPHOLOGY
Lower Clear Creek, below Whiskeytown dam, can be grouped into two reaches. The upper canyon-bound reach of Clear Creek has stream slopes in the range 0.6 to 2.0 percent as measured from USGS, 1:24,000 scale, topographic quadrangles. Tributaries to the canyon bound reach typically have stream slopes greater than 4 percent. The lower reach has an average stream gradient of 0.3%. (Castro, 1996)
The lower reach has lost its natural meander pattern. In places, the stream runs in straight highly entrenched channels dug to facilitate gravel mining. Steep bluffs, composed of The Riverbank and Red Bluff formations, occur where Clear Creek has cut into these formations and where hydraulic placer mining historically occurred. For a distance of several hundred meters immediately below McCormick-Saeltzer Dam (MCSD), Clear Creek has cut a narrow canyon into the exposed metamorphic rocks of the Eastern Klamath Terrane. The flood plain, flood plain terraces, and the adjacent valley walls are potential sources of gravel. The lower reach has multiple flow channels. The major flow channel contains sections of riffle-pool sequences with alternating point bars. Multiple flow channels have been created by past dredge-mining for gold and present gravel mining activities. A few beaver dams exist in areas of multiple channels. (Modified from Western Shasta Resource Conservation District, 1996)
Up to 10,000 spawning chinook salmon frequent Clear Creek
(Brown, 1996a). The digging of redds by fish and reduced stream
flows contribute to streambed armoring. Streambed samples taken
in 1996 using a McNeil sediment sampler show a decrease in armoring
downstream to the Sacramento River (Castro, 1996).
3. STUDY REACH 8
Clear Creek below MCSD, flows for 1.3 km through a narrow canyon.
From this canyon to the Sacramento River, the average stream
gradient is 0.3%. Using the Rosgen Level II stream classification
system (Rosgen, 1994), the stream types are C4 and C5. This
lower 9.2 km of Clear Creek has a present sinuosity of 1.1 using
the definition of sinuosity (P) equal to channel length (Lc) divided
by valley length (Lv) (Chorley et al., 1985). Study Reach 8,
at stream km 9 and discussed in this paper, is a 140 m riffle/run/riffle.
A 3-dimensional sketch of Reach 8 for an average stream flow
of 34 m3/s on Sept. 4, 1996 is shown in
Figure 2.
4. BED MATERIAL CHARACTERISTICS
Particle-size distribution was evaluated in July and September of 1996 at established reaches below MCSD (Castro, 1996). The USFWS (Kohler, 1996) conducted pebble counts using the Wolman procedure (Wolman, 1954) and collected McNeil core samples (McNeil et Ahnell, 1964) in order to develop a particle-size distribution for the surface and subsurface bed material. The percent fines (<0.85mm) for all samples taken below MCSD were below 10%, which is typical for Pacific Northwest rivers (Stober et al., 1983: Chapman and McLeod, 1987). Lisle and Eads (1991) report that thresholds of concern fall most commonly around 20%. However, this standard or criteria is only applicable to spawning sites, and not to randomized sampling across or along the stream channel. Thus, some caution must be used when comparing the Clear Creek sediment samples to standards established specifically for spawning gravel.
Nonetheless, it is meaningful to plot the McNeil and Wolman
Pebble Count samples against an envelope for approximate limits
of suitable spawning gravel. Figure 3, shows samples re-plotted
for particle size ranges used in model runs and spawning gravel
limits modified from California Fish and Game criteria (CDWR 1986).
The d50 (mean particle size) of all pebble counts
ranged from 2.4 to 12.7 cm (Brown, 1996b). The d50
particle range of the McNeil samples ranged from 1.7 to 2.5 cm.
McNeil streambed samples taken prior to spawning activity plot
near or outside the fine-grained limit for spawning gravel. Two
of the samples are plotted in Figure 3. Although the McNeil samples
may indicate a potential for fine sediment to accrue in redds
(spawning areas), plots of the McNeil samples (Figure 3) also
suggest that digging fish can make functional redds. It's also
worth noting that the plotted armored material falls within the
CDWR criteria. A discussion of the mobility of streambed sediments
is included in the section on Stream Hydraulics and Particle Transport.
5. FLOW REGIME
Hydroelectric water use associated with Whiskeytown Dam and the
Spring Creek Power Plant has reduced streamflow in the creek.
Flows (Q,s) are measured at the Igo gage, located about 11.3
km downstream of the dam. From 1940 to 1963, the discharge at
the gage averaged 368,000,000 m3/y. From 1963 through
1996, it has averaged 143,000,000 m3/y, a reduction
of 61 percent. Current minimum releases from the reservoir are
2.8 m3/s during November and December, and 1.4 m3/s
the rest of the year, giving a minimum annual release of 51,800,000
m3. Figure 4 displays the pre- and post-dam mean monthly
Q's at the Igo gage.
Pre- and post-dam annual peak Q's by exceedance probabilities
and recurrence intervals are shown in Figure 5.
6. Stream Hydraulics and Particle Transport
The HEC-RAS, Version 1.0, (1995) model was run for five selected
Q's in the normal flow channel and the two high flow channels
in Reach 8. These Q's were aggregated by water surface elevation
to develop a stage discharge relationship for Reach 8. Channel
Q's range from 0.1 m3/s to 457 m3/s, Figure
6.. Five Q's were examined in detail: 4.25 m3/s, 19
m3/s, 42 m3/s, 85 m3/s, and 258
m3/s. These Q's were selected because they are either
a currently significant Q or are proposed as increased Q's for
improved management. The first Q is the presently selected management
release from Whiskeytown Dam. The second and fifth Q's are the
1.25 yr.- and 25 yr.- peak flow events (Figure 5). The third
and fourth Q's have been proposed as increased flow targets for
habitat improvement. HEC-RAS was used to calculate the hydraulic
parameters used in the sediment transport equations. Parameters
were calculated for the unvegetated portions, channels, of the
cross-section.
The Meyer-Peter Müller formula (1948) was used to examine the potential mobilization of the streambed in Reach 8 as characterized by McNeil samples. Calculations show no bedload transport of sediments at discharges of 4.25 and 19 m3/s. The equation calculates sediment transport (Qs)of these sediments at 10.69 t/d at a discharge (Q) of 42.48 m3/s, 46.40 t/d at a Q of 84.95 m3/sec., and 77.10 t/d at a Q of 258 m3/s.
The Toffaletti sediment transport equation (1969) calculates
the transport of sediment as bedload (Qb)and suspended
load (Qss). The Toffaletti equation was run off of
a floppy disk of sediment transport equations compiled by West
Consultants, Inc. (Teal, 1996) to assess the relationship of water
discharge (Q) and sediment transport (Qs) of the decomposed
granitic (DG) sediment through Reach 8. A bed sample of the DG
material was analyzed for percent weight by particle size distribution,
Figure 7. The five selected Q's described above were used in
Hec-Ras runs to calculate input values to the Toffaletti equation.
Transport of the four particle size fractions accounting for
ninety-five percent of the weight of the DG bed sample for each
Hec-Ras run are plotted in Figure 8. The size fraction 0.50 to
1.0 mm accounting for fifty-four percent of the DG material incurs
significant transport at a Q of 19 m3/s.
| Percent by Weight | Particle Size Range in mm | Geometric mean in mm | Bedload Sediment Transport in T/d | Suspended Sediment Transport in T/d | Total Sediment T/d |
| 0.3 | .062 - .125 | 0.09 | 0.0 | 66.7 | 66.7 |
| 2.7 | .125 - .250 | 0.18 | 0.1 | 187.5 | 187.6 |
| 21 | .250 - .500 | 0.35 | 0.6 | 286.9 | 287.5 |
| 54 | .500 - 1.000 | 0.71 | 1.9 | 125.0 | 126.9 |
| 16 | 1.000 - 2.000 | 1.41 | 0.6 | 6.8 | 7.4 |
| 4 | 2.000 - 4.000 | 2.83 | 0.1 | 0.4 | 0.5 |
| 1 | 4.000 - 8.000 | 5.66 | 0.0 | 0.0 | 0.0 |
| 1 | 8.000 - 16.000 | 11.34 | 0.0 | 0.0 | 0.0 |
| 0 | 16.000 - 32.000 | 22.63 | 0.0 | 0.0 | 0.0 |
| 0 | 32.000 - 64.000 | 45.26 | 0.0 | 0.0 | 0.0 |
7. FISH USAGE
Lower Clear Creek from MCSD to the Sacramento River, approximately 6 miles (10 km.), is currently utilized by fall-run Chinook salmon, Oncorhynchus tshawyscha, for spawning (CDFG, 1993). Late-fall chinook also spawn in Clear Creek as perhaps do steelhead (Brown, 1996b). All salmonids including winter-run chinook use the creek for rearing habitat (ibid).
The California Department of Water Resources (CDWR, 1986) stated that the quantity and quality of fishery habitat in Clear Creek have declined significantly during the preceding 20 years, due to low sustained flow releases (Q's) below Whiskeytown Reservoir, reduced incidence and intensity of flushing flows, mining of spawning gravel sources, increased amounts of sand-size sediment, and encroachment of riparian vegetation. The 1986 CDWR report goes on to say analyses of gravel samples taken during the study consistently showed excessive sand.
Study reach 8, stream mile 5.6 (km. 9), discussed in this
paper is a 460 foot (140-meter) riffle/run/riffle habitat sequence
bordered on its upper and lower limits by pool habitat. This
reach is the upper limit of available spawning habitat for fall-run
Chinook before the migration barrier at MCSD. Spawning habitat
at the upstream pool riffle interface appears good with slight
embeddedness of gravel (<20%) and appropriate gravel size
distributions (>50% 2.54 cm. through 10 cm. gravel).
8. CONCLUSIONS
This study has looked at the relationship between stream flows (Q's) and the existing streambed in Reach 8. Armoring does not appear to be limiting the spawning activities of fall-run Chinook salmon. Salmon were observed reworking up to one foot (one-third meter) depth of the streambed in construction of their redds (spawning areas).
A total and discharge related sediment budget needs to be developed to model the quantities of sediment that can be transported through the stream system, preventing accruement of fine grained sediment in redds. This will require systematic sampling of Qs, measurements of Q and the development of a sediment rating curve.
The absolute values of calculations of sediment transport is not at issue. At issue is the conclusion that lower Q's will not transport spawning gravel and that a threshold Q must be exceeded before movement of spawning gravel will occur.
Sediment transport of the finer sand fractions at low Q's
has management implications. Namely, suitable stream bed habitat
can be maintained by scheduling flow releases to insure stream
transport of finer fractions through reaches used by salmonids
to construct redds. To determine management Q's, the methodology
used in this paper needs to be extended to the entire lower reaches
of Clear Creek and calculations verified with field measurements.
ACKNOWLEDGMENTS
The authors wish to thank Janine Castro, NRCS geomorphologist,
Portland, Oregon and the US Fish and Wildlife Service in Red Bluff,
California for sharing data and Thomas Share, NRCS civil engineering
technician, Davis, California for assistance in developing figures.
The authors also wish to thank the Shasta County RCD, sponsors
of the Lower Clear Creek Watershed, for the opportunity to participate
in a comprehensive planning effort and the NRCS staff at the Redding
field office for their assistance. The authors also wish to express
special thanks to the numerous volunteer reviewers of this paper
and especially to reviewer Walt Sykes, NRCS Water Resource Planner,
Davis, California.
REFERENCES
Brown, M.R. 1996a. Benefits of Increased Minimum Instream Flows on Chinook Salmon and Steelhead in Clear Creek, Shasta County, California, U.S. Fish and Wildlife Service, Northern Central Valley Fishery Resource Office, P.O. Box 667, Red Bluff, California 96080, 17 p.
Brown, M.R. 1996b. Personal Communication, U.S. Fish and Wildlife Service, Northern Central Valley Fishery Resource Office, P.O. Box 667, Red Bluff, California 96080.
CDFG 1993. Restoring Central Valley Streams: A Plan for Action, California Department of Fish and Game, Sacramento, CA.
CDWR 1986. Clear Creek Fishery Study, The Resources Agency, Department of Water Resources, Northern District, State of California, 70 p.
Chorley, R.S., Schumm, S.A., Sugden, D.E. 1995. Geomorphology, Methuen, London and New York, 605 p.
Castro, J. 1996. Lower Clear Creek Monitoring Project, Shasta County, California, Natural Resources Conservation Service, 101 SW Main Street, Suite 1300, Portland, Oregon, 97204, unpublished.
Chapman, D.W., and K.P. McLeod 1987. Development of Criteria for Fine Sediment in the Northern Rockies EcoRegion, Battelle Columbus Laboratories, EPA Contract No. 68-01-6986.
HEC-RAS 1995. HEC-RAS River Analysis System, Hydraulic Research Manual, Version 1.0, U.S. Army Corps of Engineers, Hydrologic Engineering Center, Davis, California, Two Manuals.
Kohler, E. 1996. Personal Communication, U.S. Fish and Wildlife Service, Northern Central Valley Fishery Resource Office, P.O. Box 667, Red Bluff, California 96080.
Lisle, T.E., and R.E. Eads 1991. Methods to Measure Sedimentation of Spawning Gravels, USDA Forest Service Research Note PSW-411, 7pp..
McNeil, W.J., and W.H. Ahnell 1964. Success Of Pink Salmon Spawning Relative To Size Of Spawning Bed Materials, U.S. Fish and Wildlife Service Special Scientific Report-Fisheries No. 469, 15pp..
Meyer-Peter, E., and R. Müller 1948. Formulas for bed load transport. Int. Assoc. Hydraul. Res., 2nd Meeting, Stockholm, pp. 39-64.
Rosgen, D.L. 1994. A Classification of natural rivers. CATENA, ISSN 0341-8162, Vol. 22, No. 3, pp. 169-199.
Stober, Q.J., C.R. Steward, and F. Winchell 1983. Tolt river Fisheries and Instream Flow Analysis, Final Report, Department of Water, Seattle, Washington.
Teal, Marty 1996. Personal Communication. West Consultants, Inc. 2111Palomar Airpot Rd., Suite 180, Carlsbad, Ca. 92009-1419.
Toffaleti, F.B. 1969. "Definitive Computation of Sand Discharge in Rivers", ASCE Journal of the Hydraulics Division, V. 95, No. HY1: pp. 225-248.
Western Shasta Resource Conservation District 1996. Lower Clear Creek Watershed Analysis, Shasta County, California, seven chapters and six appendices.
Wolman, M.G. 1954. A method of sampling coarse river-bed material.
Transactions of the American Geophysical Union 35(6): pp. 951-956.