Restoring Freshwater Tidal Wetland Systems

Pool spacing, channel morphology, and the restoration of tidal forested wetlands of the Columbia River, USA.  Heida Diefenderfer and David Montgomery.

Large wood structures have been recognized as having an impact in both aquatic systems (rivers) and terrestrial ecosystems (old growth forests)

The nature, structure and function of large wood in tidal channels has been forgotten.

Wood in rivers can obstruct flow and sediment transport

May create pools

Anastomosing channel patterns may be forced

Forested tidal channels receive effects from both oceanic tides and terrestrial watershed processes.

High percentage of reproduction occurs in high marshes occurs on fallen trees

However, they may also float, dredging up living trees

Large wood as animal habitat

Shelters juveniles

Tree depressions, when trees float off, are used by fish

Enhances macroinvertebrate richness

Tidal channels in vicinity of Gray’s River, with a 7.7 ft. tidal range, were found to have a “forced step-pool” morphology:

An observed sequence of log-jams with pools between them

And an average number of channel-widths per pool

Pool structure is reversed from that of a mountain stream

Pools are upstream of log jams

Tails are upstream of pool

This suggests that the dominant energy is associated with the flooding tide.

Implications for restoration

Density of logs in natural forest system is high

Legacy logs under diked pastures have emerged due to erosion after tidal flow has been re-connected.

More wood is needed (imported LWD)

Systems, when mature should provide wood.

Beaver positively effect functioning of such systems

Tanner, Curtis, Jeff Cordell, Jane Rubey and Lucinda Tear. 2002. Restoration of freshwater intertidal habitat functions at Spencer Island, Everett, Washington.  Restoration Ecology 10(3):564-576

Spencer Island is located between Steamboat and Union Sloughs in the Snohomish estuary.

Diked in early 1900’s; used for grazing

Flooding excluded by dikes, tidegates, ditches

Southern tip is open to the Snohomish River at river kilometer 6.1

Spencer Island is 412 acres.

Restoration is limited to south 59 acres

Phalaris arundinacea (reed canarygrass) covered two thirds of this area, with some cattail

Elevation is 0.2 to 3.0 m above MLLW in the reed canarygrass areas.

Higher areas support blackberry, alder, willow, poplar and Sitka spruce.

Restoration goals

Re-establish tidal conditions

Displace reed canarygrass

Provide habitat for juvenile salmon

Flooding not consistent with waterfowl management to north

760 m  cross levee was constructed to isolate restoration area

Two 15 m and one 60 m dike breach were made in 1994 in historical south dike

This resulted in restoration of a 4 m tidal range

Some salinity was expected; none has appeared.


Pre-breach, forest covered eastern third; it subsequently began to die off.

P. arundinacea began to die off and retreat to higher elevations; deeper areas became mudflats.

Shrubs and emergents replaced forest

Permanent open water areas became drained

Results, continued

Juvenile fish used site: chum, coho and chinook salmon, plus other non-game fish (peamouth chub, stickleback)

Invertebrates were less diverse than in reference site, but rich in salmon prey.

Reference site 3.2 km downstream

Naturally breached between 1965-70.

Lythrum salicaria colonized parts of site, so Gallerucella beetles were introduced as a bio-control.

Hood, Greg.  2002.  Application of landscape allometry to restoration of tidal channels.  Restoration Ecology 10(2):213-222.

Allometry: the relationship of the parts of an organism as it grows.

In natural tidal channels, parts with predictably related measurements are:

Outlet width and depth

Channel length, perimeter and surface area

In an artificially dredged slough on the lower Chehalis River, natural slough allometry was not maintained

As a result, sediment was retained and channels filled.

Intertidal sedge habitat  abundance was related to slough size for smaller sloughs, but not for larger.

Author suggests that sedge habitat in large sloughs was destroyed by log storage and transport.

Abundance of salmonid prey of terrestrial origin (aphids and adult flies) in slough surface waters was related to slough perimeter and sedge habitat.

Allometric perspective has several implications:

1. Rather than requiring similar size reference sites, allometric relationships may be used to design channels.

2. Correct initial allometry can increase ecological predictability.

Site:  an excavated slough and a reference slough off of the Chehalis River across the stream from Cosmopolis.

Floodplain is a Picea sitchensis-dominated wetland with freshwater species

Marsh habitat is 0.7 to 1.5 m lower and is a monoculture of Carex lyngbyei.

Reach is tidal and oligohaline


In natural sloughs, width, depth and area of natural sloughs were positively correlated with slough length; p<0.0001.

Slough perimeter and area were correlated; p<0.0001.

Excavated slough dimensions fell outside of the 95% confidence bands of natural slough dimensions.

Excavated slough was wider, deeper and had greater surface area than natural sloughs of its length.

It is therefore unlikely to be morphologically stable.

Allometry is different in different river systems.

Willapa, South Fork Willapa, Chehalis and North River were compared.

Freshwater discharge, gradients, soils likely result in locally differing allometry.

Sedge bench size also is related to slough dimensions, except for large channels

Historical photos show heavy use of channel edges and banks for log transport and storage.

Sedge benches on larger sloughs may have been destroyed by log storage.

Aphid export was correlated with sedge bench width because they are dependent on the habitat and less mobile

Dipteran export was not correlated with sedge bench width.

Diptera rear in mudflats, sedge benches and floodplain, so are not dependent on the sedges.