---

Compilation of Focus Teams and Hydrogeomorphic Advisory Team Recommendations:
Integrating Abiotic Factors in the Goals Project: Tenets of the Hydrogeomorphic Advisory Team

---

Contents

Plant Focus Team Recommendations

Fish and Aquatic Invertebrate Focus Team Recommendations

Mammals, Amphibians, Reptiles and Terrestrial Invertebrates Focus Team Recommendations

Shorebirds and Waterfowl Focus Team Recommendations

Other Birds Focus Team Recommendations

Integrating Abiotic Factors in the Goals Project: Tenets of the Hydrogeomorphic Advisory Team

---

Notice

The following is a compilation of the recommendations produced by each of the focus teams and the HAT. You will note that some of the terminology used in these reports differs from that used in the Goals Report. For example, some teams discuss regions and subregions that are defined differently, and some use different names for habitats. This is because the terminology used within the Project has evolved over time, and some has been developed specifically to facilitate presentation of the Goals.

In order to preserve the original intent of the focus team authors, no changes, other than minor formatting changes, have been made in the teams' reports. We believe the intent of their recommendations is made clear by the information contained in the reports.

Hydro-Geomorphic Advisory Team

Integrating Abiotic Factors in the Goals Project: Tenets of the Hydrogeomorphic Advisory Team

This paper presents information regarding some of the physical considerations associated with restoring baylands habitats. It includes the HAT's organizing principles and other summary points. It also includes questions posed by the focus teams and our brief answers to them.

The information we are providing here is very limited, and we recognize that any large-scale effort to restore the baylands will require substantial regional and site-specific investigation.

I. Organizing Principles

1. The preferred approach to implementation of the goals should be the restoration of natural, self-sustaining systems that can adjust to changes in physical processes, with minimum ongoing human intervention.

2. In those cases where the restoration of natural, self-sustaining systems is not possible or is not preferred, a phased eventual transition to such systems should be considered. For example, some salt ponds may be maintained to support species currently using this type of habitat. However, it would be preferable to eventually replace these with natural, self-sustaining habitat types, such as mudflats, tidal marshes, salt pannes, etc., if this can be accomplished.

3. Restoration planning and design should be based on expected regimes and variability of physical processes, including hydrology, sediment, salinity, water quality, and biogeochemistry.

4. Restoration planning and design should account for both natural causes of variability, such as drought, and anthropogenic causes of variability, such as alteration of freshwater flows.

5. Restoration planning and design should recognize the range of temporal scales. This applies both to the rate of evolution of a restored site in providing wetland functions, such as transition from an intertidal mudflat to a vegetated tidal wetland, and potential changes in the controlling physical factors, such as expected changes in the bay sediment budget.

6. Restoration planning and design should recognize the spatial scale (size and location) of restoration sites. Key examples include uniform sedimentation rates across small sites versus higher sedimentation rates near a levee breach at larger sites, and greater sedimentation rates in the South Bay than in Central Bay.

7. Restoration planning and design should consider its effects on regional physical processes, such as sediment transport.

II. Links to External Influences

1. Restoration planning should be conducted in recognition of the links to major and local influences that are external to the Bay itself.. These primarily include the oceanic influence, and watershed input from the Sacramento, San Joaquin River System, and the local river and creek contributions.

2. The restoration of the Bay Ecosystem must be tied to the restoration and maintenance of these external links.

II. Planning and Design Guidance

1. Site design should be focused on the creation of an appropriate "template" which will evolve towards a dynamic equilibrium within the shortest time to provide the range of ecosystem function and complexity characteristic of appropriate reference sites.

2. Preference should be given to the restoration of large sites, capable of providing the complexity of habitat, highest channel order, and ecosystem resilience.

3. Preference for habitat type should recognize those ecosystems which are limited in their potential extent by controlling abiotic factors. For example, intertidal wetlands should be given preference where conditions are suitable because of the extremely limited opportunities for creation.

4. It is recognized that at some sites it will be infeasible to develop a self-sustaining system, and that artificial control structures may be required. Based on experience with these systems, there is a greater uncertainty in site evolution and a greater need for oversight. Therefore, an adequately funded maintenance and management organization is needed to ensure that restoration goals are met in perpetuity.

IV. Advancing Restoration Science

1. The goal of advancing restoration science is to improve the ecological effectiveness of restoration projects.

2. Advancement of restoration science should come from a comprehensive program linking academic research, numerical and physical modeling, pilot projects, monitoring, restoration success evaluation, and information exchange between researchers, practitioners, regulators, and interest groups.

3. This comprehensive program should incorporate natural reference systems, previous restoration projects, and new restoration projects.

4. Principles of adaptive management should be incorporated into all restoration projects. Adaptive management would include both incorporating project-specific monitoring results into ongoing site management as well as exchanging information for other restoration efforts.

V. Evaluating Restoration Success

1. Adaptive monitoring and maintenance of the systems should be based on conformance to the expected evolutionary trend and to reference conditions rather than specific conditions at any given time.

RMG/Focus Team Questions and HAT Responses

The questions shown below in bold type were submitted to the HAT by the RMG on 3 November 1997, and the HAT members worked together to develop the following responses:

How long would it take for marsh "restoration" to take place in areas of subsidence?

As for most of the questions, the response to this depends on the definition of "restoration," as well as a consideration of temporal and spatial scales. The question reflects the understanding that restoration requires raising the subsided marsh plain back to an appropriate elevation in the intertidal zone and a restoration of the range of functions a tidal marsh provides. A brief review of these will provide some insights regarding the process.

Process

In subsided San Francisco Bay tidal wetlands, restoration will proceed primarily by deposition of suspended sediment (as opposed to the accumulation of organic matter), since most of the marsh plain rise will occur at elevations below that suitable for vegetation. Our observations indicate that deposition will continue until the marsh plain reaches a steady elevation relative to the Mean Higher High Water (MHHW), supporting primarily a pickleweed vegetation cover under saline conditions, or a more diverse cover of pickleweed .and other plant species under brackish-saline conditions.

It is important to remember that the processes vary through time and space in ways that preclude any exact equilibrium or steady-state.

Depth of Subsidence

The depth of subsidence in the diked wetlands varies dramatically around the Bay, depending on the various subsidence properties. In the South Bay, maximum subsidence has been about 15 feet (New Chicago Marsh in Alviso), due to both groundwater withdrawal, soil compaction, and oxidation of the organic faction by microbial actions in part. The nearby Salt Ponds (adjacent to Alviso Slough) appear to have subsided about 8 feet. Another Salt Pond (100 acre site at Cooley Landing in E. Palo Alto adjacent to the Dunbarton Bridge) has only subsided 1-2 feet. In the North Bay, subsidence of 3- to 6-feet appears common in sites studied.

Rate of Deposition

The rate of sediment deposition is affected by numerous parameters; major factors include suspended sediment concentration, depth of water column, local wave climate, salinity regime, presence of vegetation and others. Some simple models are available incorporating the first two factors for prediction of sediment accumulation rates. Reasonably good field information is available on the rates of deposition at a number of locations around the Bay to predict deposition rates for small to moderate size (up to about 200 acres) subsided sites. These include the Alviso Marina site, Warm Springs Marsh Restoration, Baumberg Tract in the South Bay, some Petaluma River marshes, and other sites in the North Bay. These indicate fairly rapid rates of accretion under present conditions for most of the Bay.

For example, the Alviso Marina (about a 5 acre site) was last excavated (for boat use) to a depth of about -15 ft NGVD (vertical datum that corresponds approximately with mean sea level) in 1976. The marina accumulated silt rapidly, and was only marginally functional by the early 80's. By 1990, it had accumulated about 16 feet of sediment, and vegetation began encroaching. By 1995, it was mostly covered with (brackish marsh) vegetation. This corresponds with monitoring in the Warm Springs Marsh (a 200 acre site) in which initial deposition rates have been extremely rapid (up to 5 feet per year), and an overall rapid pace of deposition.

Initial rates of over 2 feet of accretion per year are common in deeply subsided sites. These rates decrease as the marsh plain rises (smaller water column and associated sediment above). Using local data for calibration, it has been predicted that about 10- to 15-years would be required for sediment deposition in a subsided South Bay salt pond (marsh plain elevation currently about -3 ft NGVD) to raise the marsh plain to an elevation where native vegetation would become established. While the amount of sediment available for deposition decreases as the marsh plain rises, the establishment of vegetation accelerates the rate of rise towards steady elevation relative to the tides by reducing turbulence and adding organic matter. This estimate of 10- to 15 years is probably applicable to similar small to moderate sites in the South Bay, which has the highest rates of deposition. In the North Bay, there are reports of initial cumulative deposition rates of about 0.5 m/yr at the Petaluma River Marsh Restoration Site. Based on a series of site comparisons, there are estimates that it would require about 35 years for the Sonoma Baylands site to reach a steady elevation relative to the tides.

Temporal Considerations

The above observations are based on the historical and existing suspended sediment concentrations and rates of sea level rise. While these are not likely to change quickly, it is important to recognize the long-term future sediment supply to, and sediment loss from the Bay system may change and that the rate of sea level rise may increase. These topics are described more fully in the response on pages 6 and 7.

Spatial Considerations

The restoration sites monitored to date have been small (generally less than a couple hundred acres). Concurrent opening of large numbers of subsided sites will require consideration of the regional sediment supply. As an example, at its maximum depth, the 200-acre Warm Springs site aggraded at a rate of almost 5 ft/year, corresponding to an annual accumulation of perhaps 1.0 to 1.7 million cubic yards. This represents a significant fraction of the total net Bay sediment available of about 5 million yards/year.

Restoration Process

It is probably clear to all the participants that the term "restoration" is a controversial topic, which covers a wide range functions/values. As applied to this question, we recognize that the subsided site will evolve through the states of:

• subtidal, unvegetated

• low intertidal (mostly mudflats, unvegetated)

• mid intertidal (vegetated by lower marsh vegetation)

• high intertidal (mature marsh plain vegetation)

From a process perspective, we can state with some assurance that the evolution will proceed through the above states at a predictable time frame. However, these represent only the broadest categories, and do not reflect the complexity that we see in an ancient marsh compared with a recently restored site. For example, it is unlikely the slough channel system will achieve the multiple channel orders and sinuosity in a recently restored site compared with an unaltered reference site. Likewise, the amount of organic matter and nutrients in the marsh sediments will be less in a site which has undergone extremely rapid rates of mineral soil deposition (such as a subsided marsh reopened to tidal circulation), and the organic matter would be mainly in the uppermost soil layer. Whereas, in the case of a marsh plain that has evolved gradually over thousands of years the organic matter would be distributed throughout the vertical soil profile, down to the contact with ancient mudflat sediments. In view of this, we should not expect recently restored marshes to include this level of complexity for decades or perhaps even centuries. The goal of the restoration plan should be to the create the optimum "template" such that the site will evolve towards a condition of maturity and complexity in the shortest time frame, recognizing that some functions can be restored more quickly than others. The monitoring process should be focused on whether the side is evolving along the desired path rather than the specific state at any one time. This approach is emphasized in the HAT guiding principles and recommended research.

Can we create and maintain large slough channels in restoration (which provide mudflat foraging habitat for shorebirds)?

Marsh slough channels evolve as nature's most "efficient" way of exchanging water and dissipating energy within the intertidal landscape. At any location within the marsh plain system, the slough channel cross-section dimensions and shape reflect a balance between erosion (scour) forces exerted by the tidal flow which tend to expand the channel bed/banks, and the tendency for deposition of suspended sediment to decrease the channel dimensions. At the most basic level, the maximum channel "order" within a marsh complex, and size of the slough channels at a particular location are determined by the size (areal extent) of the intertidal zone. Quite simply, to support large slough channels or complex networks of channels of varying order, we need large marshes. Slough channels hundreds of feet wide, with maximum depths of 25 to 30 feet and broad expanses of unvegetated mudflats were common features of the historical Bay marshes which covered thousands of acres. This image is apparent in the historical view of the EcoAtlas.

The areal extent of restored tidal marsh required to support a given width/depth dimension of a slough channel can be approximated by using the "hydraulic geometry" approach developed in large part Bay Area wetlands scientists. There is also extensive data collected on how rapidly channels respond to changes in tidal area/tidal prism: In response to decreases in marshplain area, the channels decrease rapidly in depth to a new equilibrium level, and more slowly in width. The rate of enlargement of channels with increased tidal exchange depends primarily on the erodibility of the underlying sediment (highly consolidated clay material is relatively resistant to erosion and may require excavation).

Does marsh restoration decrease mud flat habitat? Will restoring tidal marsh areas reduce bayside tidal mud flats used by shorebird and waterfowl species?

The seat-of-the-pants consensus of the HAT seems to be 'not much, if any.'

Can you estimate the decrease in mud flat area with the increase in tidal marsh at a specific site? For a region?

To check this, some simple calculations were made assuming the following:

1. All sediment to fill a pond comes from mud flats from MHHW to -6 ft MLLW, a depth at which bottom sediments can readily be resuspended by wind waves.

2. The longshore distance that contributes sediment to a salt pond is equal to the frontage of the pond on the Bay.

3. The pond bottom must be raised 3 feet (gross estimate from Napa River ponds).

4. No sediment comes from the local watershed.

5. The slope of the resulting mudflat will be the same.

6. The mudflat can not migrate landward.

These assumptions are probably the worst case for mudflats. In reality, some if not most of the sediment that would deposit in the ponds would come from further away in the Bay (or more distant sediment would replenish sediment that moved from the mudflat to the pond) or from the local watershed. These assumptions can be used to calculate the distance the -6 ft contour will migrate landward. This is the cross shore distance of mudflat lost.

Volume of fill = AH = L dh dx (1 - dx/x)

where A=area of pond, H=depth of pond, L=Bay frontage of pond, dh=distance from -6 ft MLLW to MHHW, dx=distance -6 ft MLLW contour moves toward shore, and x=cross shore distance from -6 ft MLLW to MHHW.

This approach was tried for 2 randomly selected salt ponds along the shore of South Bay: the pond east of Mountain View Slough and the pond south of Coyote Hills Slough. For each pond, about 10-15% of the habitat from MLLW-6 ft to MHHW (dx/x~0.1 to 0.15) would be lost for this worst case scenario.

How should salt ponds be restored through phasing of pilot projects, i.e., which ponds should be restored first and how long should one salt pond or pond complex take to develop before another is attempted?

Some of the decisions that would have to be made and some of the factors that would affect these decisions can be listed. If a large area of salt ponds were to be restored, a study would be required to answer this questions specifically and in appropriate detail for making resource management decisions. So far, as best we know, restoration projects have been small enough where this question has not arisen.

Spatial decisions (what order?):

Napa or South Bay first (depends on ecological benefit on baywide scale)?
Restore ponds adjacent to the bay/river shoreline or furthest away from shoreline first?
For South Bay, restore closer to SF or San Jose first?

Phasing decisions (when)?

What criteria should be used to decide when to open up additional ponds to tidal action (seasonal, deposition in previous pond, time)?

Hydrologic factors influencing decisions:

Sediment supply: is seasonal and varies from year to year.
Salinity: in and around the ponds should be maintained at an ecologically safe level. This will also be dependent on freshwater flow, which also varies seasonally and annually.
Tidal currents: May be altered near a restored pond.
Pond plumbing: The ponds have extensive plumbing to transfer water and produce salt.
Restoring tidal action to one pond could affect the flow of water between ponds and thus water quality in the remaining ponds.

Other factors:

Opportunity: What ponds are available at any given time for restoration.
Levees: If a levee breaks during a flood the pond is restored to tidal action.

Can you maintain the variation in salinity in salt ponds without continuing to operate the ponds as a evaporative system?

No. In order to produce hypersaline water from seawater, water (H2O) must be removed (evaporated). An alternative to solar evaporation would be a desalination (reverse osmosis) plant that would produce drinking water and hypersaline water.

Comment on the implications of sea level rise in relation to long-term management of both tidal and diked wetlands.

Atwater and others have described the history of the San Francisco Bay on a geologic time scale. Sea level rose rapidly prior to 8,000 years ago and progressively invaded the valleys, creating the San Francisco Bay system. The rate of rise slowed between 8,000 and 6,000 years ago to approximately the present rate. At this slower rate soil eroded from the land and transported to the bays accumulated along the shores and supported the proliferation of marsh plants. The plants accelerated the rate of deposition of suspended sediment in their midst, as they do today, and continuing accumulation of sediment and plant material raised the surface. As sea level continued its rise, sediment was added to the surface and the rising marsh invaded the land and created the extensive tidal marshes found by the fortyniners.

Sediments enter the bays suspended in the waters of winter freshets. For the bay system as a whole, the Central Valley drainage provided in excess of 80 percent of the sediment entering the bays, with the remainder contributed by local streams. The importance of local sediments supplies probably increased closer to local sources. For example, it is possible that the relative contribution of sediment from the Napa Creek watershed increased upstream from Mare Island. The material from the Central Valley drainage deposits initially where it enters the broad bays. Onshore breezes generate waves during spring and summer days that suspend the newly deposited material and tidal currents circulate it throughout the bays. During a year, most of the material either deposits in locations where it is not further suspended by currents or waves, including deposition on marshes, or exits the Golden Gate.

Human activity wrought large changes in the Bay system. Sediment and water inflows have been altered drastically, and most of the marshlands have been diked and drained. Present evidence indicates that prior to 1849, the limited supply of suspended sediment brought into the bays was not quite sufficient to maintain water depths, and the bays slowly deepened. Hydraulic mining contributed 1.4 billion cubic yards of mud deposit in the bays and on the marshes during the period from the early 1850's to the late 1870's.

Vallejo Bay and Northhampton Bay became mud flats with the Carquinez Strait channel through them. Large deposits filled the upper bays and added large amount of material to San Pablo Bay. Most of the marsh south of Highway 37 and marsh along the western shore evolved on these deposits. Agriculture in the valleys and mountain slopes added to the sediment supply then and since.

The suspended sediment input continued to be higher than natural pre-1850 levels until the water projects began to divert sediment-laden river waters for irrigation and municipal supplies. Total annual input to the system averaged 10.5 Mcy during the period 1923 to 1950, and averaged 7.9 Mcy during the period 1955 to 1990. The CalFed activities suggest that there will be no further reduction. Water diversion is subject to political and legal forces and to the pressures of population growth, however, and the long-term prognosis is uncertain.

The upper bays and San Pablo Bay are now so shallow that suspension by waves and tidal currents move all of the annual winter deposit, except that needed to compensate sea level rise, further down into the system. It circulates and deposits where hydraulic conditions permit. North San Francisco Bay is now slowly filling with accumulating sediment, and there is a plentiful supply to South Bay. About 40 percent of the annual supply exits the Golden Gate.

The central roles of sea level rise and sediment supply in maintaining the elevation of mature marshes is apparent from this description. As long as the sediment supply is sufficient to maintain the elevation relative to MHW as sea level rises, the marshes will endure. It appears that the present supply of suspended sediment is sufficient for a modest rate increase. Excessive increase of the rate of sea level rise or decreases in sediment supply, however, will lower marsh elevations relative to the tides or submerge them.

Restoration requires higher suspended sediment concentrations than does marsh maintenance. Suspended sediment concentrations are highest where there is wave action on mud flats, less elsewhere. Planning restoration of diked former marshlands requires attention to the local supply and to the impacts of nearby large restoration projects on depletion of suspended sediments. Evaluation of such impacts can be made using numerical hydraulic and sediment transport models.

In order to sort out immediate and long term effects of a restoration projects it will be useful to complete the bay-wide evolution of the bathymetric history as this integrates variations over time scales of interest. Then we could develop whole bay sediment transport model(s) with resolution on the order of 50-100m in conjunction with the bathymetric change surfaces at the same resolution. Higher resolution models of individual restoration projects will be useful in predicting an immediate (1-5yr) response, but for long term stability analysis, a full bay transport model will be required.

Once a model is functional and verified with bathymetric change, for it to be useful in predictive scenarios we will need accurate estimates of sediment delivered to the bay including major local streams, and elevation maps for potential restoration sites.

In the southernmost South Bay, inflows from San Jose may be creating a brackish system. Would large scale restoration in this area work for tidal marsh species? If not, could a marsh system be used to keep the freshwater farther from the Bay?

The answer is yes, but it will be expensive and will require maintenance.

There will always be brackish water where treatment plant effluent having low concentrations of dissolved salts mixes with the more saline South Bay water. The location and configuration of this mixing zone can be changed, and from the question it appears that it would be desirable to locate the mixing zone far from the Bay and have tidal marshes colonized with salt-tolerant plants on the margins of the Bay. It might be noted that the historical condition included some amount of brackish marshland associated with the inflows from local creeks. It might also be noted that there are some historical data to suggest that the historical mean daily flow from South Bay creeks combined was about equal to the allowable mean daily sewage effluent, although the historical natural flow was seasonally much more variable. The ongoing occurrence of some amount of brackish tidal marsh in the far South Bay, in conjunction with salt marsh restoration, would reflect historical conditions.

A portion of the most bayward salt ponds, leveed to protect them against the highest tides and storm waves, can be used for mixing the effluent with saltier tidal water, before release into the bay environment. In essence, what is required is a forebay. The pond or forebay should be sufficiently deep to prevent the establishment of brackish water plants and should be connected directly to the Bay with a channel dimensioned to maintain itself by flows created by the tidal prism of the mixing pond. Maintenance of the water depths in the mixing pond will require periodic dredging, because the high concentrations of suspended sediment in South Bay waters, combined with the tranquillity of the mixing pond waters, will cause rapid rates of sedimentation. For some years the dredged material can be used to accelerate the restoration of neighboring marshes.

Dimensions and configuration of the mixing pond and the connecting channel can be determined with the aid of hydraulic and salinity models and specification of acceptable salinities at the discharge. Patterns and rates of sedimentation in the pond can be determined with the use of a sediment transport model.

The large tidal range in the South Bay may be sufficient to provide the necessary mixing in the Bay itself. A pipeline from the plant and a diffuser. Possible located north of Dunbarton Bridge, would avoid local low salinities. Evaluating this solution and determination of the location of the diffuser could easily be completed with conventional models.

The HAT would like to take this opportunity to emphasize the need for the simulation of hydraulics and sediment transport in the design of restoration projects. Even the simple breaching of levees requires that their locations and dimensions be optimized to achieve desired deposition patterns and water circulation. Every project has unique conditions including shape and elevations of the site and suspended sediment and salt concentrations in the flooding water. Model studies are very inexpensive, compared with construction costs or the costs of an unsuccessful project. The HAT is considering how it might help bridge the gap between modelers (the scientists that develop and test models but may know too little about their operational application in natural resource management) and managers (the people in government who make decisions based upon model outputs but who may know too little about their assumptions and uncertainties).

Given that shorebirds and waterfowl need certain elements that are contained in artificial salt ponds and managed marshes (e.g., open water, roosting sites within a kilometer of feeding areas, etc.), are there particular sites with restoration potential, or particular design features that could be incorporated into tidal restoration projects, that will provide the elements required by shorebirds and waterfowl?

At this time the HAT will defer discussion of particular sites. We can discuss some of the assumptions about restoration projects.

Some Assumptions

Large scale tidal marsh restoration on diked historic baylands will occur either through the "natural sedimentation" model or the "dredged sediment/backfill placement" model. This assumption guides the approaches available during and following restoration construction to achieve desired elements.

The desired landscape elements sought within tidal marsh restoration projects are open water areas within the tidal marsh of both shallow (for shorebirds) and deeper (for waterfowl) depths.

Any restoration project proceeds through an evolutionary process from the initial unvegetated (or in some cases submerged vegetation) intertidal or subtidal landscape to early vegetation colonization and ultimately to a vegetated marsh plain dissected with a tidal slough system. The time over which this process occurs can vary widely from site to site and in general cannot be predicted with a high degree of accuracy.

Restored wetlands will be subject to regional sea level rise conditions that will influence the inundation regime of tidal wetlands.

Tidal Marsh Pannes as Open Water Areas

At least two types of ponds ("pannes" in the Goals Project typology) existed historically in tidal wetlands in the San Francisco Estuary: drainage divide pannes located within the vegetated marsh plain between tidal drainage networks, and transitional pannes located at the upland boundary of the tidal wetland. Few examples of drainage divide pannes remain. Petaluma Marsh is probably the best location to find numerous extant pannes of this kind. Hoffman Marsh in Richmond adjacent to Highway 580 had such pannes but they were drained as part of an enhancement project in the mid 1980s. Virtually no historical transitional pannes remain, as these areas have been overtaken by land use conversion. Only where tidal wetlands still have a natural upland edge that is not too steep are these pannes found. Rush Ranch in the Suisun Marsh is one such example, though mosquito control ditches have taken their toll. The pannes along the uplands edge of marshland at the Emeryville Crescent may be analogous to the historical transitional pannes.

Drainage divide and transitional pannes are characterized as small depressions in the landscape that have some type of topographic containment that defines the top elevation of the water surface. Containment features can be small berms in which case the pond could be partially or wholly perched atop the marsh plain, with the pond bottom below or at the height of the adjacent marsh plain, respectively. Containment features can also be the marsh plain itself, in which case the pond bottom is below the marsh plain (i.e., a simple depression).

There may be three water sources for drainage divide pannes. Most prevalent are tidal inputs, the magnitude and frequency of which are related to the height of the pond containment feature relative to the tides. Typically, it is the higher spring tides that reach these ponds. Consequently, the seasonal variability of ponding relates directly to the seasonal variability of the higher spring tides each year (see discussion of tides elsewhere), with the June-July and December-January spring tide series being of particular significance. Direct rainfall can also supply water to these ponds. Finally, emergent groundwater can contribute to surface ponding. Surface water is lost by surface drainage out of pond, groundwater infiltration, and evaporation. Likely the most important characteristic affecting surface water loss is substrate type; the more impervious the substrate (e.g., more clay), the longer duration of ponding (vernal pools are a good analogy).

Water sources for transitional pannes can include all those described for drainage divide pannes plus runoff from adjacent upland areas. Consequently, these ponds can have a greater freshwater influence relative to drainage divide pannes and, depending on annual climatic variability, they may support greater duration of ponding. Surface water is lost in the same manner as for interior ponds.

The HAT presumes that both types of ponds or pannes offer habitat for benthic and aquatic invertebrates as food sources for shorebirds and waterfowl.

Tidal Marsh Channels as Open Water Areas

Channels within tidal marshes are open water areas. Water depth varies with the daily tidal flows and with channel bottom elevation. Minor channels drain at lower tide levels offering exposed channel bottoms through a portion of the tidal cycle depending on their bottom elevation. Major channels either remain submerged throughout all tides (for the largest channels) or may be drained at some of the lowest tides (for the moderately large channels). The HAT presumes that channels can provide a variety of foraging opportunities for shorebirds and waterfowl, including habitats for benthic and aquatic invertebrates and fish.

High Marsh as Roosting Sites

The HAT presumes that roosting sites are needed for passerines and raptors as well as shorebirds and waterfowl. Tidal marshes support roosting sites for passerines and raptors on the tall emergent vegetation along channels and especially along natural levees. Roosting sites are also available along the upland perimeter of tidal wetlands though such availability is strongly affected and defined by adjacent land use. For moderately high tides that do come out of the channels, the tidal marsh plain covered with low vegetation may serve as a roosting site for shorebirds and waterfowl. The shallow pannes of high tidal marsh might also serve as roosting sites for shorebirds and waterfowl.

Creating Shorebird and Waterfowl Habitat in Tidal Marsh Resptoration Projects

No built projects we know of have included drainage divide pannes in their design and construction. Two proposed projects have included such features (Montezuma Wetlands Project and Redwood Landfill Wetland Restoration) and one planned project may include them (Hamilton Army Airfield). The only built projects we know of that have included transitional pannes are Arrowhead Marsh currently under construction by the Port of Oakland and Oro Loma Marsh under construction by the East Bay Regional Park District. Both projects include a variation of the transitional panne idea that does not quite replicate the historic condition but seeks to provide shorebird and waterfowl foraging habitat. The basic issue with pond creation is how to generate the appropriate elevations, perimeter containment features, and substrate, and how to exclude unwanted vegetation colonization.

Under the natural sediment restoration model, ponds may form naturally but as yet we do not have sufficient understanding about how they form or under what time scale formation may occur. Pond formation probably involves some influence of stagnant water (tidal water entrained in the peats or isolate on the marsh surface) on plant survival. Drainage divide ponds could be created within restoration sites after the appropriate elevations have been reached (i.e., return to the site some number of years after construction and do some followup construction work). Though restoration strategies have yet to be developed, what they might entail could be determined through experiments in any existing tidal marsh with appropriate elevations.

Under the dredged sediment/backfill placement restoration model, ponds could be built at the outset through creating the required elevations, containment features and substrate with the dredged sediment or backfill. Strategies to achieve ponds in this manner have been proposed but not yet field tested.

Restoration projects can incorporate some flexibility with respect to channel density and size within some constraints. These constraints include proving adequate amount of tidal circulation throughout a restoration site and the natural processes of sediment transport that form and maintain tidal marsh channels through erosion and deposition. Natural marshes exhibit a wide range of combinations of channel density (defined hydrologically as the total length of channel per unit of marsh surface area [though an ecologist might be interested in the total surface area of channel per marsh surface area or the amount of channel edge]) and channel size. For example, Rush Ranch in Suisun Marsh comprises relatively few channels (i.e., low channel density) but these channels are generally fairly large, whereas Petaluma Marsh comprises numerous channels (i.e., high channel density) but these channels are generally not as large. These differences may have to do with several factors that are specific to regions with the Estuary, such as tidal range, degree of riverine influence, and salinity.

What is the relationship between natural maintenance of tidal channels large and small, including mudflats along the edges of large channels, and the tidal prism provided by tidal marsh restoration.

As pointed out elsewhere, the form of tidal marsh channels in plan view, profile, or cross section is a result of interactions among the erosional and depositional actions of the flowing tides. In a very general way, channel cross-sectional area increases with tidal prism. For example, channels get larger downstream, toward their tidal source. For smaller channels, say 1st-order to 3rd-order, the increase in channel size is due more to gains in depth than width. For larger channels, the increase in channel size is due more to width than depth. It is therefore also commonly observed that smaller tidal marsh channels tend to be u-shaped in cross-section, whereas the larger channels are more v-shaped. In other words, the banks of the larger channels are less steep. The large channels therefore tend to have more area of mudflats, despite the fact that the smaller channels are more likely to completely de-water at low tide.

The relationship between channel form and the tidal prism of the channel has been described for some channels in some areas of the Estuary. The relationship is better described for channels of small to moderate size in saline marshlands. The relationship is not well described for very large channels in any area, or for any size channels in freshwater areas. Historical soundings in tidal marsh channels could be assembled to help describe the relationship between channel size and tidal prism for very large channels, but original field work would be required to explore the relationship for smaller channels in non-saline areas.

A rather crude prediction of the relationship between size of tidal marsh restoration project and amount of channel-associated mudflat could be developed based upon an assumed height of the project plain relative to a local tidal datum, the estimated area of the plain, the expected equilibrium form of the channel in cross-section (i.e., the slope of the banks and channel depth relative to the tides) as evidenced by existing data, and the expected plan form (i.e., sinuousity and length) or density (area of channel per unit area of marsh plain) as evidenced by existing data. Another approach would be to quantify the mudflat associated with different size natural marshes as historically mapped by the US Coast Survey.

What are the local physical controls, including soil characteristics, for seasonal ponding on diked baylands, including farm lands and ranch lands?

The primary control is the distribution and abundance of surface water, as affected by rainfall, levee weep or leakage, ground water discharge, and on-site water management. For diked baylands, it is generally true that surface water exists only until it infiltrates the soil, or while the groundwater rises above the soil surface. There are variations within and among sites due to the interactions among weather (timing, intensity, and duration of rainfall, evapotranspiration), soil conditions (depth, field capacity and related parameters), depth to groundwater, distance to tidal influence and/or uplands, and water management practices (i.e., types and conditions of water control structures and their methods and timing of use). However, a few basic patterns are self-evident. These are listed below.

• For any given set of weather conditions, the amount of seasonal wetlands tends to be inversely related to the amount of drainage.

• The lack of drainage, or the amount of management to provide drainage, tends to be related to the amount of ground subsidence. Simply stated, lower land is more difficult to drain. Therefore the potential for seasonal wetlands increases with subsidence.

• Subsidence tends to be greater for organic soils than mineral soils, and for deep soils than shallow soils. The most subsidence is observed for deep peaty soils.

• Within the diked historical marshlands, the mosaic of organic and mineral soils reflects the historical distribution of tidal marsh channels, with the organic faction increasing with distance from historical channel banks.

• Within a site, in the absence of land management practices that level the land surface, the mosaic of organic and mineral soils produces differential rates of subsidence, which in turn produce topography, and this topography helps control the distribution of surface water, and seasonal wetlands.

• In diked baylands, seasonal wetlands are mainly due to local infiltration of irrigation water or rain that causes the near-surface groundwater level to rise above the ground surface. The influence of the tides and groundwater from hillsides tends to be restricted to areas very near the bay, perimeter levees, or adjacent to steep upland terrain.