Sediment Considerationsand Dam Breach on the Lower Snake River
Idaho Fish & Game Report to the Director 5/1/98
Will removal of dams create sediment that will harm fish?
The following information is provided to give a sense of the types of considerations and possibilities when fish, flowing water, and sediment interact. Because potential impacts are dependent upon a myriad of factors (e.g., sediment type, fish species, fish size, water temperature, and others), extrapolation from one situation to another is problematic at best. Clearly, sediment can effect fish, given the variables and unknowns; however, probably the most productive considerations relate to how to manage sediment and its potential impacts. Lastly, from an ecosystem perspective the sediment that will travel downstream following dam breach on the lower Snake River should be considered episodic in nature; a large-scale event whose detrimental effects will be transient. Restoration of Idaho's salmon and steelhead resources is long-term in nature, and like other anadromous populations in the Northwest have survived past episodic perturbations.
Definitions and Units Suspended sediment (SS) is typically measured in milligrams per liter (mg/L). Turbidity is an optical quality and can be measured in a number of different units, but generally these units relate the amount of light that can be passed or refracted from a suspension or the distance through which something can be seen. The relationship between SS and turbidity measured at one concentration of SS in one stream could be achieved at a lower or higher concentration of SS in another stream, depending upon the type of sediment in suspension. Jackson turbidity units (JTUs) are a measure of turbidity and relate the distance a light can be seen through a suspension.
Direct Impacts of Suspended Sediment on Fish
Direct impacts of fish include lethal and sublethal responses. These effects are related to a number of factors and among these factors some generalities can be made.
Dose -- dose is defined as the concentration of SS multiplied by the duration of exposure. SS concentration alone without duration factored in does not correlate with effects on fish nearly as well as dose does. As a result, when considering potential impacts of SS on fish, duration should be considered. The higher the dose, the greater the impact.Lethal Impacts, Examples from the Laboratory
Particle size and shape -- The larger the particle (up to a point) and the more angular, the greater the impact.
Water temperature -- For salmonids, the higher the temperature, the greater the impact.
Fish size and lifestage -- The smaller the fish, the greater the impact. This is believed to be related to the fish's ability to physically expel particles that enter the mouth.
Presence of other stressors -- The greater the stress the fish is under prior to exposure, the greater the impact.
Species -- Some species are affected by SS differently than others. This may be due to size and behavioral differences. From laboratory experiments, it appears that chinook and steelhead may be more resistant to SS than coho or sockeye.
Recalling that sediment type affects impacts on fish, none of the studies below necessarily occurred with sediment similar to that behind the lower Snake River dams. Nonetheless, differential impacts under different conditions does provide some insight into the principles governing potential impacts to fish from SS.
At a concentration of Fraser River sediment in suspension at 2,100 mg/L for 96 hours, sockeye salmon experienced no mortality. As the concentration increased to 3,148 mg/L, trauma to gill tissues was evident. In one trial, 50% mortality occurred at 17,650 mg/L while in another trial 90% mortality occurred at 13,000 mg/L. Coho salmon were used in a subsequent experiment by the same researchers using the same Fraser River sediment, and the fish never experienced more than 50% mortality even at 22,700 mg/L for 96 hours. Unfortunately, chinook and steelhead were not tested in either of these experiments.
In other experiments with a different sediment type, using coho and steelhead of similar size and under similar conditions (5,471 mg/L, 96 hours, 18.7C), coho experienced 10% mortality, but no mortality occurred in steelhead. In another experiment, it took almost twice the concentration of SS to elicit the same mortality rate in chinook as in sockeye.
In addition to different responses between species, differences also occur between different size fish of the same species. In one experiment with coho, 50% of the test fish died at concentration of 8,200 mg/L for 96 hours. After these fish grew an additional 6mm and weight increased 40%, the concentration required to kill 50% of them over 96 hours increased to 22,700 mg/L.
Sub-lethal responses can occur at doses of SS cited above and less. The sub-lethal responses of fish observed in the laboratory include: elevated plasma cortisol and hematocrit levels (indicating stress), changed feeding patterns and reactive distances, decreased growth rates, decreased disease resistance, and use of turbidity for predator avoidance (a beneficial impact). Some of these responses were measured under conditions more favorable than those observed in the wild where wild populations not only persist, but thrive.
Suspended Sediment Impacts at the Population Level
From the examples above, it is clear that sediment can affect fish on an individual basis when they cannot seek refuge. On a population level those effects and the ramifications are unclear; at least partly because suspended sediment and turbidity are factors fish have evolved with. Indeed, some of the largest producers of salmon on the west coast of North America are quite turbid and regularly carry a significant SS load. For instance, the Taku River (mean turbidity range 240-400 JTUs) and the Fraser River (>100 JTU's and up to 1,050 mg/L) in Southeastern Alaska. In the case of the Fraser River, the high level of SS also corresponds with the peak of smolt emigration.
Chinook have been observed rearing for extended periods in rivers with SS concentrations up to 61mg/L. Estuaries which are generally turbid, typicaly have SS concentrations of >15mg/L. A survey published in 1959 reported that peak values of SS in northwest streams reported that SS seldom exceeded 500 mg/L but at times could go as high as 7000 mg/L.
Sediment flushing from the Spencer Dam on the Niobrara River in Nebraska resulted in a SS concentration up to almost 22,000 mg/L. Over a period of five years these flushing activities resulted in mortality of at least 22,741 fish. However, the fish impacted were primarily resident, bottom-dwelling species, not migrating salmon and steelhead. When the Harpster Dam on the South Fork Clearwater River and the Lewiston Dam on the mainstem Clearwater River were breached, no dead fish were observed by IDFG personnel.
Generally, the main concern with sediment effects on salmonids in the Northwest is the detrimental effects on spawning and rearing areas and in impacting egg-to-parr survival. From an ecosystem perspective, in the case of the lower Snake River, rearing habitat could only be improved by breach. Spawning activity in the pools of the dams is negligible relative to what could potentially occur in a free-flowing Snake River and tributaries following breach. In 1993, an estimated 10 redds were constructed in the pools formed by the lower four Snake River dams, in 1994, 9 redds, and an abbreviated survey in 1995 found 0 redds.
Breach Options Being Considered for the Lower Snake River Dams
Options for breach under consideration by the U.S. Army Corps of Engineers, Walla Walla (the Corps), include different combinations of the number of dams breached at once. The two options being looked at most seriously at this time are breach of all four dams in one year and breaching two dams each year over two successive years. Due to administrative policies, the latter option would likely result in work beginning one year earlier than the former option. Removal of all four dams at once would return the river to a natural state sooner, decrease construction duration, and more work could be done out of the water under low flow conditions. Work would begin around August 1 and finish in December or January. The earthen portion of all four dams could be removed in six months at a cost of about $207 million dollars. Adult steelhead and chinook are in the river during the late summer and fall and fish ladders could remain operational. It's expected that under this scenario, significant amounts of sediment wouldn't begin moving downstream until the November - December time frame.
Sediment Behind the Lower Snake River Dams
An estimated 100-150 million cubic yards of sediment reside in the pools behind the lower Snake River dams. In Lower Granite pool, this sediment is believed to be primarily silt with a smaller component of sand. From 47% to 95% of the particle sizes moving as bedload during the 1992 drawdown were less than 0.5mm in size, depending upon location and date, relative to peak drawdown. All of this material was less than 16mm in diameter.
Removal of sediment from the bankfull channel of the Snake River through erosion is expected to take no longer than 5 years. At least 50% of this sediment is expected to be moved downstream within 2 years, depending upon conditions. This time frame is similar to that estimated on the Elwha River following removal off the Glines Canyon and Elwha dams and the Condit Dam on the White Salmon River. About 95% of the fine silt is expected to be removed in the first year following removal of the Condit Dam. During the 1992 drawdown, suspended sediment increased from a background level of 9.5 mg/L to a high of 1,928 mg/L. Most measurements were substantially less than 510 mg/L, the next highest reading. The 1992 drawdown occurred during March and peak drawdown lasted 2 weeks. Inflow into Lower Granite pool during the drawdown averaged 30,100 cfs, which was 54% of the 18-year average. During the duration of the 1992 drawdown, the Snake River and tributaries eroded down to their original channels.
Based on observations during the 1992 drawdown, the Corps believes that there will be little opportunity to stabilize accumulated sediment and use it in the riparian reclamation process. Rather, due to the incised nature of the landscape and instability of the sediment, most of the sediment will be eroded and moved downstream. The ultimate fate of this sediment is uncertain, but a substantial portion will likely settle out in McNary pool.
Snake River Channel Stabilization and Fall Chinook Use
The amount of time for the river channel and stream banks to stabilize are not known at this time, but the Corps is in the process of developing a model to answer these questions. The Elwha River is expected to remain unstable for 6-10 years following removal of the Glines Canyon and Elwha dams. Much of the 15 million cubic yards of sediment behind these dams is fine-grained silt and clay; potentially similar in size to the sediment behind the lower Snake river dams. As the gravel and sandbars of the Elwha River are restored following dam removal, it is estimated that the streambed will elevate 1-5 feet.
The amount of spawning habitat that would become available to fall chinook following breach in the lower Snake River is being addressed by the USFWS and USGS cooperatively, but that work will not be completed until June. However, based on maps from the mid 1930s, the Snake River, under what is now Lower Granite pool contained 13 rapids, 46 pools, and 23 riffles. The average size of the riffles was 1,600,000 square feet, ranging in size from 94,000 square feet to 4,900,000 square feet.
Considering the short amount of time it took in for the Snake River to erode down to its original channel in 1992, it's possible that recolonization of the lower Snake could conceivably begin in a relatively short time frame, although time will be required to achieve streambed stability and a clearing of the gravels.
Managing Sediment Impacts
Sediment can affect fish individually and, under the right circumstances, on a population basis. Resolution of exactly what the sediment impacts following dam breach will be, will remain elusive; however, due to the absence of projects similar in size and scope. Perhaps the best question when considering breach of the lower four Snake River dams, and the sediment behind them is how best to manage the sediment and minimize impact.
Timing is probably the best method available for minimizing impacts. Within any given year, breach should occur consistent with timing that considers both the number of native species affected and species particularly susceptible to extirpation. Between year timing will also be critical. Questions of whether to protect strong brood years or weak ones will have to be addressed as well as which species warrant the highest level of protection. An additional management action for consideration is augmentation of spring flows with storage water following breach, to decrease the concentration of SS and thereby potentially reduce any impacts.
Appendix 3.1 IDAHO's ANADROMOUS FISH STOCKS:
Their Status and Recovery Options
Report to the Director Idaho Fish & Game 5/1/98
Issue Paper: Sediment Considerations and Dam Breach on the Lower Snake River
learn more on topics covered in the film
see the video
read the script
learn the songs