Restoration of Fish Habitat in Relocated Streams
DESIGN OF RELOCATED STREAM CHANNELS
Highway Locations In Flood Plains
The base flood plain of a stream or river is that part of its valley that would be inundated by a 1
percent chance flood, that is, a flood having a 1 percent chance of being equaled or exceeded in
any given year.*
*The discharge or flow corresponding to any given frequency can be obtained from a flood frequency curve- Such curves are prepared by hydraulic engineers from rainfall and watershed data and stream discharge records.
Highway construction in flood plains, including the relocation of stream channels, should be
coordinated with local, State and Federal water resource and flood-plain management agencies
during the planning stage. The preliminary location studies should investigate alternatives
avoiding encroachments on flood plains. When a proposed location involves significant
encroachments, the flood risks and the impacts on natural and beneficial flood plain values
should be evaluated and the location should be adopted only after it has been shown to be the
only practicable alternate. (U.S. Water Resource Council's Floodplain Management Executive
Order 11988, 43 FR 6030, February 10, 1978)
Planning Stream Relocations
A relocated stream channel should be designed to carry about the same discharge as the original
natural channel it replaces. Usually, this will be a discharge in the range of the 50 percent chance
flood, (Q2), to the 10 percent chance flood, (Q10). Greater flows than these should be
accommodated by overflow on the flood plain. A relocated channel should also approximate the
hydraulic gradient of the original stream.
The length and slope of the original stream can be measured in the field, or scaled from the large-scale topographic maps made for the highway location studies. The average dimensions of the
old channel can be scaled from the map or determined from field cross sections. With this
information, the designer can sketch a preliminary channel on the topographic map. Where
possible, this channel should have the same average gradient, and approximate the curvature of
the original stream. However, if constraints such as excessive excavation or the need to preserve
valuable land exist, the designer may have to accept a shorter channel with less curvature.
The location and design team should examine this preliminary channel location on the ground.
The inspection group should include a specialist on fish and fish habitat who can estimate the
amount of fish habitat space, the hiding cover, and other features that determine the habitat
quality of the original stream. The soils engineer or geologist can estimate the types of materials
that might be encountered in the new channel, and their probable stability against erosion, and
identify rock outcrops and other conditions that may affect the cost of channel excavation. The
landscape architect may be able to suggest changes in the channel alignment that would minimize
damage to large trees and other vegetation.
The field examination will indicate where the preliminary channel can be improved. After these
changes are made, the new channel should be analyzed for hydraulic capacity, water velocity and
high water elevations. The soils in the new channel may also be tested as possible highway
Unlike canals and ditches, natural streams are inefficient carriers of water. Their channels are
undulating, with steep and flat reaches. Widths are variable, and the channels are filled with
obstructions that impede the flow.
A relocated stream should approximate this natural condition. The size of channel required will
depend upon the assumed design discharge, (usually Q2 or Q10), the slope, hydraulic roughness
and shape (whether wide and shallow or narrow and deep). For design purposes, the hydraulic
engineer can divide the proposed channel into reaches of similar gradient. Knowing the
proposed gradient, channel shape and roughness, he can compute a water surface profile, and
determine the probable water depths and velocities in each reach.
The soil survey should establish the types of materials likely to be encountered in the new
channel, especially the range of particle sizes. Knowing the velocities in each reach, the designer
can estimate the probable erosion that might take place in these materials. If this erosion seems
excessive, the designer may change the shape of the channel, or reduce the gradient, or introduce
roughness to reduce the velocity. This is a cut and try process, and it may require several trials to
arrive at the best solution.
For good habitat, a stream must have adequate depth at low flows. Wide channels with shallow
depths at medium and low stages should be avoided for this reason, and also to reduce
undesirable destruction of bank vegetation.
Hydraulic roughness is an important factor in the hydraulic computations. For a given discharge,
an increase in roughness reduces the velocity and increases the depth of flow. In the Manning
formula for flow in channels, roughness is represented by an empirical coefficient, n, which may
vary from 0.012 for concrete-lined canals to 0.025 or more for unlined earth channels. For the
latter, a good approximation of n can be obtained from Figure 19 if the average particle size of
the materials composing the bed is known.
The roughness coefficient for natural streams may range from 0.03 up to about 0.08, depending
on the coarseness of the bed materials, and the vegetation growing in the bed and banks.
Large rocks placed in a new channel to improve habitat conditions increase channel roughness.
The total effect depends on the number and size of the rocks, the width of channel in which they
are placed, the depth of flow, and the initial roughness of the channel. If the roughness is already
0.045 or greater, adding large rocks at the rate of about 1 rock per 300 sq ft (27 m2) of the stream
bed will not significantly increase n, and the effect of the rocks can be ignored.
Table 2, derived from computer simulation studies, shows the effect of large -rocks in channels
of various widths. As can be seen, isolated rocks spaced far apart in wide channels have little
effect on roughness. An increase in n of less than about 15 percent is not significant and can be
disregarded for hydraulic computations.
One large rock per 300 sq. ft. (27 m2) of streambed is optimum for creating fish habitat.
However, greater concentrations of rocks may be used to reduce velocities in steep channels and
to create stair-step pools and cascades. If the number used is large enough to be effective, these
rocks will materially increase the roughness of an excavated channel. Manning's n for such a
channel would be comparable to that of mountain streams with beds of cobbles and large
boulders, which commonly range from 0.05 to 0.07, or even higher.
|Percentage Increase In Manning's n Caused By Habitat Rocks In A Channel of n = 0.035 Flowing At 5 ft (1.5m) Depth
|Number of 5 ft (1.5 m) diameter rocks placed in a 50 ft (15 m) reach of stream
|20 ft (6 m)
||40 ft (12 m)
||60 ft (18 m)
||100 ft (30 m)
Resistance of Channels to Erosion
The resistance of new channels to erosion depends on the mix of large and small particles in the
bed and banks, and their cohesiveness. Noncohesive fine soils are easily eroded; coarse gravels
and clays are more resistant. The relative erodibility of a proposed channel relocation can be
estimated from soil samples taken at intervals along its length. Analysis of these samples will
give the average particle size, d5O that is, the size of which one half of the sample, by weight, is
finer. Knowing this average particle size, the probable safe channel velocity in that material can
be estimated from Figure 20.
In a stable channel the bottom and sides remain essentially the same from year to year. All new
channels undergo some rearrangement of the bed and bank materials, but will gradually and
naturally stabilize as the finer materials wash away leaving the coarser materials behind.
However, this natural process may take many years, and the end result of such unrestricted bank
erosion may be a wider and shallower stream than is desired. Furthermore, fish habitat
downstream may be damaged by excessive sedimentation. A reasonable degree of stability can
be achieved in vulnerable places such as sharp bends, or where the banks are of new fill or easily
eroded materials by supplying bank protection.
When suitable rock is available, dumped riprap will provide the most economical bank
protection. Protective riprap blankets must be individually designed for the sites they are to
protect, and in most cases a filter will be required between the riprap and the native bank
materials to prevent undermining. The rocks composing the riprap should be large enough that
they will not be moved by the design flood. Riprap on the outside banks of bends should be
composed of larger rocks than for straight reaches. The riprap should extend up the bank as high
as the water surface of the design flood, and the bank above that level should be grassed. It is
good practice to place soil in the crevices of the riprap to encourage the growth of vegetation.
This helps to stabilize the riprap and improves its appearance. A typical design for riprap is
shown in Fig. 21.
Where large rocks are not available for riprap, smaller rocks and cobbles can be utilized by
enclosing them with wire mesh to form protective mattresses.
For long-range stability, it is important to establish vegetation on the banks as soon as possible.
All exposed earth banks should be top soiled and seeded concurrently with the excavation. In
places it may be desirable to plant willows or other quick-growing native woody vegetation,
rather than wait for nature to reestablish the bank vegetation.
Fish Habitat In The Relocated Channel
The preliminary studies for a channel relocation will establish the approximate amount of habitat
space existing in the original stream. About this much habitat space should be provided in the
new channel. The number, type and location of habitat structures and large rocks will affect the
stability and conveyance capacity of the new channel, and so should be considered during the
Where check dams are needed to control the gradient, their locations can be determined from the
profile and plotted on the plans. Additional check dams may be used to provide fish habitat.
Generally, these should be located where the current at medium stages is fairly swift-2 to 3 ft/sec
(0.6 to 0.9 m/sec)-otherwise the downstream plunge pool may fill with sediment.
Large rocks placed in a stream have the effect of increasing bed roughness. This retards the
current velocity, reduces erosion and may provide improved habitat for fishes. This is often the
best way to improve a deep, fast run which otherwise would be uninhabitable by fishes. Whether
to control channel velocities by check dams or large rocks will depend on relative costs and the
availability of materials. It is impractical to specify in advance the locations of habitat rocks but
the designer may specify the number of rocks to be placed in a particular reach of stream, leaving
the exact placement to the field forces.
Plans and Specifications for Stream Relocations
By constructing roads and stream relocations concurrently most of the channel excavation can be
disposed of in the road fills. The road excavation also may be a source of large rocks and
boulders for check dams and habitat rocks. The limits of the stream relocation should be shown
on the contract plans in sufficient detail for the project engineer to mark the channel for the
contractor and measure the excavation involved. The project engineer should have some
flexibility to make minor changes in alignment, depth and channel side slopes to take advantage
of favorable excavation.
The specifications should require the contractor to save and stockpile logs from the clearing and
large rocks from the road and channel excavation for later use in riprap and habitat structures.
Vegetation suitable for transplanting should also be saved.