United States
Environmental Protection

Office 01 Ecosystems & Communities
Region 10
1200 Sixth Avenue
Seattle WA 981 01

Aquatic Resources Unit

December 1997
Wetland and Riparian
Taking a Broader View
Proceedings of a Conference
A ~
Publication Co-Sponsors
I ~'\t:O ST-'1)'.s
-:5 <5'
~ ;po ft ~
'\ ;
«'1-;: ,-,0
-'11. P,RO\'i:-V
u.s. Environmental Protection Agency
-: . .
Northwest Chapter
Society for Ecological Restoration
Pacific Northwest Chapter
Society of Wetland Scientists

                    Wetland and
                         Taking A
                   Broader View
     Contributed Papers and Selected Abstracts
            Society for Ecological Restoration
International Conference, September 14-16,1995,
                      Seattle, Washington
                Editors: Keith B. Macdonald
                         Fred Weinmann

                           Published by
         U.S. Environmental Protection Agency
                             Region 10
                      Seattle, Washington

Cooperating  Parties

Northwest Chapter, Society for Ecological Restoration
Pacific Northwest Chapter, Society of Wetland Scientists
U.S. Environmental Protection Agency, Region 10
CH2M HILL, North Pacific Region
Publication of this Proceedings volume is the result
of a collaborative effort between the U.S. Environ-
mental Protection  Agency, Region 10, and CH2M
HILL, North Pacific Region, both located in Seattle,
The opinions  expressed herein  are  those of  the
authors and do not necessarily reflect those of the
cooperating or sponsoring agencies.  These proceed-
ings reflect the diverse  points of view and writing
styles of the contributing authors. The  aim of the
editors was to present the papers in an effective for-
mat, not to change  the author's intent  or overall
First published in  1997  by the U.S. Environmental
Protection Agency and CH2M HILL.
Printed in the United States of America.
Copyright® 1997 by the  U.S. Environmental Protec-
tion Agency.
All rights reserved. No part of this publication may
be  reproduced or transmitted in any  form or by  any
means-electronic or mechanical, including  photo-
copying, recording, or any information storage or re-
trieval system-without prior permission  in writing
from the U.S. Environmental Protection Agency.
                                      The United States Government is authorized to pro-
                                      duce  and distribute reprints for governmental pur-
                                      poses notwithstanding  any copyright notation that
                                      may appear hereon.

                                      Preferred Citation
                                      Macdonald,  K.B.,  and F. Weinmann  (eds.) 1997.
                                      Wetland and Riparian Restoration: Taking a Broader
                                      View. Contributed Papers and Selected Abstracts,
                                      Society for Ecological  Restoration, 1995  Interna-
                                      tional Conference,  September 14-16, 1995, Univer-
                                      sity of  Washington, Seattle, Washington,  USA.
                                      Publication EPA 910-R-97-007, USEPA, Region 10,
                                      Seattle, Washington.

                                      Additional copies of this  publication are  available
                                      from  the Environmental Protection Agency, Public
                                      Information Center,  EXA-124,  1200 6th  Avenue,
                                      Seattle, Washington 98101.
                                      Telephone: (206) 553-1200
Printed on Recycled and
Recyclable Paper

Sponsorship [[[ vm
Acknowledgments [[[ ix
About These Proceedings ............................ [[[ x
   New Directions in Wetland and Riparian Restoration
      Macdapald, Keith B., Weinmann, Fred [[[ xi

Taking a Broader View

   An Integrated Wetland Management Program for the Snohomish Estuary
      Stanley, Stephen; Roberts, Paul [[[ 1

   Planning for Wetland Restoration and Enhancement in the Context of Regional
   Wetland Management Plans
      Hruby, Thomas [[[ 15

   Integrating Wetland Compensation Banking with Local Wetland Planning
      Green, William J [[[ 22

   Watershed Analysis as a Tool in Watershed Restoration
      Toth, Steven [[[ 32

   Prioritization of Salmonid Habitat Problems and Habitat Restoration Projects in an
   Urbanizing Watershed
      Neal, Kathryn; Harenda, Mary [[[ 47

   Active Management of Riparian Habitats
      Berg, Dean Rae [[[ 50

   The Relationship of Large Woody Debris, Riparian Vegetation, and Landform in
   Small, Low Gradient Streams

Contents, continued
Creating Tidal Marshes on Dredged Materials: Design Features and Biological
Winfield, Ted; Florsheim, Joan; Williams, Philip [[[108

Regional Restoration Planning for San Francisco Bay Salt Marshes: A
Biogeographic Approach

MacKay, Kevin.......... """"""""""""'"'''''''''' ................................................ ........ ........ 125
Habitat Restoration in an Urban Setting: Integrating Desert Riparian Habitat,
Groundwater Recharge, and Community Participation
Brown, S. c.; Donaldson, J. J.; Kroska N.; Cotter, D. [[[136
Taking a Closer Look
Aquatic Habitat Restoration in the United States: A Review of Design and Costs
Thorn, Ronald M.; Wellman, Katherine; Shreffler, David K.; Scott, Michael J. ...........141

Wetland Mitigation Design: From Theory to Reality

Boule, Marc................................... [[[ 146
Ecological Functions of a SaltmarshlMudnat Complex Created Using Clean
Dredged Material, Jetty Island, Washington
Houghton, Jonathan: Gilmour, Robert H. [[[156
Restoration and Management of the Salmon River Estuary

McDonald, Chris ...... .................. .................. ............................................... .......... .......... 171
Salmon River Salt Marsh Restoration in Oregon: 1978-1995

Frenkel. Robert E. [[[176
Batiquitos Lagoon Enhancement Project: Concept to Construction
Josselyn, Michael; Appy, Ralph; Whelchel, Adam [[[184
Restoration of Spartina Marsh Function: An Infaunal Perspective
Levin, L. A.; Talley, D.; Talley, T.: Larson, A.; Jones, A.; Thayer, G.; Currin, c.;

Lu nd, C[[[ 190
The Sum Exceedance Value as a Measure of Wetland Vegetation Hydrologic
Simon, Scott D.; Cardona, Martha E.; Wilm, Brian W.; Miner, James A.:

Contents,  continued
Works in Progress-Selected Abstracts

   Complexities of Species Conservation Under the Endangered Species Act:
   Ecosystem Management in the Cedar River Watershed
       Erckmann, W. James	247

   Restoration of Biodiversity from the Genetic to Landscape Levels: The Elwha
   River Restoration Project
       Winter, Brian; Bransom, Sarah; Bohman, Jeff	249

   Estuarine Restoration: A Landscape Ecology Perspective
       Shreffler,  David K.; Thorn, Ronald M	250

   Suitability Analysis for Enhancing Wildlife Habitat in the Yolo Basin
       Donaldson, Joseph J.; Rawlings, Marcus; Chainey, Steve;
       Stallman,  Jay	251

   Historical Analysis: A First Step in Stream Restoration
       Grant, Katharine R.; Huckins, Eddie	252

   Multi-Scale Planning for Riparian Woodland Restoration on Streams of the
   Western United States
       Harris, Richard R.; Olsen, Craig; Johnson, Rebecca	253

   Plant Species Diversity in a Semi-Arid Riparian Ecosystem in Southeastern
       Snyder, Keith A.; Guertin, D. Phillip; Ffolliott, Peter F.; Jemison, Roy L	254

   Manipulating the Disturbance Regime in a Southeast Missouri Bottomland
   Hardwood Forest Fragment to Promote Overstory Species Recruitment and
       Taylor, Sylvia H.; Hermann, James W.; Larson, Brad; Williams, Michael;
       Journet, Alan, R. P	255

   Invasive Plant Species Distribution and Abundance in Freshwater Wetlands of
   Puget Sound  Lowlands, King County, Washington
       Houck, Catherine A	256

   Considerations for Restoring Structure, Function, and Diversity to an Ecosystem
   Colonized by Invasive Plants:  A Phragmites Case Study
       Aheam-Meyerson, Laura; Vogt, Kristina A	257

   Biological Control of Purple Loosestrife
       Blossey, Bernd	259

   Rodeo® Herbicide Use to Control Smooth Cordgrass (Spartina alterniflora L.) in
   Pacific Northwest Estuaries
       Crockett, Ron P	260

   Integrated Pest Management Strategies for Reed Canarygrass (Phalaris
   arundinacea L.) in Seasonal Wetlands
       Paveglio, Fred L.; Kilbride, Kevin M.; Crockett, Ron P.; Wiseman, R. Bruce	261

Contents, continued
    Bioengineering in Stream and River Restoration: Success of Enhancing Specific
    Functions Under Multiple Constraints
       Larson, Mark; Hopkins, Betsy	262

    Planting Deep Willow Cuttings via Hydraulic Jetting, for Riparian Erosion Control
    and Habitat Enhancement
       Drake, Lon;Langel, Rick	264

    Revegetation at La Franchi Demonstration Wetland, Santa Rosa, California
       Waaland, Marco	265

    Lessons Learned in Habitat Restoration, Comparison of Two Wetland Mitigation
    Projects in San Diego, California
       Minchin, John L	269

    Wetland Functional Relationships; A Landscape Ecology Perspective
       Fuerstenberg, Robert R	272

    Assessing Wetland Condition and Restoration Potential Within Watersheds Using
    a Performance Criteria Matrix
       Brooks, Robert P.; Cole, Charles Andrew; Wardrop, Denice H.; Bishel, Laurie;
       Prosser, Diann J	272

    Wetland Restoration Success Following Pipeline Construction
       Clancy, Margaret; Antineau, Clayton; Burns, Dennis	272

    Hydrologic and Ecological Modeling and Geographic Information Systems for
    Predicting Floodplain Forest Restoration Following Dam Removal
       Shuman, John; Warr, Karen R.; Greenberg, Gary M.; Bryan, Judith C	273

    Restoring Hydrologic, Geomorphic, and Ecologic Functions in the Big Quilcene
    River Delta
       Fishbain, Larry B.; Williams,  Philip B.; Collins, Brian	273

    Geographic Information Systems Applied to Selection of Forested Wetland
    Restoration Sites
       Schulman, S. Andrew	274

    An Evaluation of the Hydrogeomorphic Approach for Assessing Restored Forested
    Wetland Functions in the Puget Sound Lowlands
       Rains, Mark C.; Lee, Lyndon  C.; Braatne, Jeffrey H.; Mason, Jeffrey A	274

    Use of the Hydrogeomorphic Approach in Design of Riparian Wetland Restoration
    Along the Central California Coast
       Ellis, Linda R.; Lee, Lyndon C.; Fiedler, Peggy L.; Rains, Mark C	275

    Middle Waterway Pilot Restoration Project275
       Clark Jr., Robert C.; McEntee, Dave; Lantor, Judy	275

    Nearshore Industrialized Areas  of Port Gardner, Washington, and Their Effects on
    Early Marine Life History of Anadromous Fishes
       Houghton, Jonathan P.; Kyte,  M. A.; Gregoire, D. L	276

Contents, continued
   Importance of Tidal Channel Geomorphology to Restoring Ecological Functions of
   Coastal Wetlands
       Simenstad, Charles, A.; Rozas, Lawrence P.; Minello, Thomas J.; Reed, Denise J.;
       Coats, Robert N.; Zedler, Joy	276

   Creating Landscape Restoration Opportunities: A Case Study at Rattlesnake Lake
       Rottle, Nancy	277

   An Assessment of the Mycorrhizal Fungal Status of Wetland Prairie Plant Species:
   Implications for Ecological Restoration
       Turner, Stephen D.; Friese, Carl F	277

   Watershed Restoration in Deer Creek, Washington—A Ten-Year Review
       Doyle, James; Fisher, Michelle; Movassaghi, Greta; Nichols, Roger	277

   Stream Restoration in Alaskan Subalpine Placer-Mined Watershed
       Densmore, Roseann V.; Karle, Kenneth	278

   Restoration of Ashbridge Marsh: Potential Consequences for Water Quality
   and the Aquatic Community of the Lower Don River
       Helfield, James M.; Diamond, Miriam L	278

   Native Plant Salvaging: Plant Recycling for Watershed Restoration
       Goeldner, Jo; Maia, Eric; Young, Cindy	279

   Using Citizen Monitoring for Watershed-Level Assessment of Urban
   Restoration Goals
       Preuss, Charles; Fischer, Chris	279

   Characterization of Columbia  Basin Watersheds: Providing Regional Context for
   Stream Restoration Goals
       Goodman, Iris; Jensen, Mark	280

   Restoring Native Communities to the Sandy River Delta
       O'Hara, Kevin R.; Kelly, Virginia	280

   Fundamentals in Restoring Biological Integrity to Regulated, Large Floodplain
   River—A Case Study
       Davis, Mike; Johnson, Scot	280

   Author Index	283


The Northwest Chapter of the Society for Ecological Restoration and the Pacific Northwest
Chapter of the Society of Wetland Scientists are pleased to sponsor the publication of this
series of papers and abstracts on wetland and riparian restoration. The decade of the 90s has
been a period of emergence for the expanding field of aquatic ecosystem restoration and
during this time a confluence has occurred between the critical need for action and in-
creased technical understanding of "how to do it." At the same time, the credibility of
aquatic restoration as a scientific discipline was increased by two published compendiums:
Restoration of Aquatic Ecosystems by the National Research Council of the National Academy
of Science (National Academy Press, 1992) and Restoring the Nation's Marine Environment
CGordon W. Thayer, ed., Maryland Sea Grant College, 1992).

The list of authors represented in this Proceedings is a virtual who's who in the field of
habitat restoration, particularly within the Pacific Northwest. The papers reflect the high
quality work of many authors. We believe that these papers—in their diversity—represent a
collection of thoughtful approaches to restoration issues, a source of reference information
for other practitioners, and a measure of the state-of-the-art at the time of their preparation.
As such, we believe they will assist restoration science through the planning and imple-
mentation of sound habitat restoration projects on the ground.
                          David Chapin, President
                          Northwest Chapter
                          Society for Ecological Restoration
                          Curtis Tann&r, President
                          Pacific Northwest Chapter
                          Society of Wetland Scientists

We especially appreciate and gratefully acknowledge
the enthusiasm, excellent contributions, and patience
of all the many authors who made this Proceedings

Manuscript  preparation  and  publication  of  the
Proceedings reflects a creative and successful public-
private partnership among four co-sponsoring agen-
cies:  the U.S.  Environmental Protection  Agency
(Region 10),  the  environmental  services  company
CH2M HILL, the Northwest Chapter of the Society
for Ecological  Restoration (SER), and  the Pacific
Northwest Chapter  of the  Society  of  Wetland
Scientists (SWS).
The initial invitation that the senior editor organize a
conference session  devoted to wetland restoration
came from  1995 SER Conference Co-Chairs Sono
Hashisaki (Springwood Associates Inc.)  and Tim
White (CH2M HILL). During  an early  planning
session  with Sono and Tim,  we came up with the
conference theme, Taking A Broader View—that we
subsequently invited wetland contributors to consider
in their  talks and papers. With Mary Kentula's (U.S.
EPA, Environmental Research Laboratory, Corvallis)
enthusiastic support the topic rapidly expanded to in-
clude riparian restoration, and two  full days of talks
were  organized.  During  the  SER  Conference,
significant interest was expressed in soliciting papers
from  the speakers and  publishing  a proceedings
focused on broadening perspectives in wetland and
riparian restoration. Our special thanks go to all those
who by  making the 1995  SER  conference  such  a
success, indirectly made our task much easier.
Special thanks also go to Chapter Presidents David
Chapin (SER),  Janet  Morlan and  Curtis   Tanner
(SWS), and their respective Chapter boards, for their
joint support and endorsement of these proceedings.
Mike  Kennedy, Russ  Stepp,  Paul  Lanspery, and
Kathy  Lombardo,  all  of CH2M HILL, approved
financial support for manuscript preparation.  CH2M
HILL   document  processing   staff  Jani   Hatch,
LaDonna Stewart, Julie  Moore, Sue Seefeld, and
Charla Peterson, editor-proofreader Ermina Swager,
and  graphics  designer Lynette  White, all  helped
prepare the camera-ready manuscript.
Karrie Simic (University of Washington, Engineering
Professional Programs) made our task easier by pro-
viding,  with  the  Conference  Chairs'  approval,
electronic copies of selected program abstracts.
The senior editor particularly thanks Fred Weinmann
(U.S. EPA, Region 10) who enthusiastically endorsed
the concept of publishing  these  proceedings, obtained
EPA funding to do so, and then shepherded the final
manuscript through the printers.

About  These Proceedings
 The 1995 International Conference of the Society
for Ecological Restoration,  organized  around  the
theme  Taking A Broader View, was held on  the
University  of  Washington,  Seattle  Campus   in
September 1995. During the conference's  wetlands
sessions,  significant  interest  was  expressed   in
publishing  a  wetland/riparian-based  Proceedings
volume. The papers and abstracts included here were
solicited from conference  participants  throughout
1996.  A  summary  paper  highlighting  changing
perspectives on wetland and riparian restoration was
then added to complete the Proceedings.
  The  papers   and  abstracts  included  in  this
Proceedings volume have been divided among three
groups: Taking  a Broader View,  Taking  a  Closer
Look, and Works in Progress. Where possible, papers
and  abstracts sharing  related themes  have been
grouped together.
  Taking a Broader View—These  14  contributions
are  generally broad  in scope and context.  They
consider the role of wetland and riparian habitats on a
landscape or watershed scale, emphasizing regional
approaches to wetlands restoration  planning. Each
contribution demonstrates the value to be gained from
taking a broader view. In apparent contrast, Klaus
Richter's  paper  focuses on detailed habitat require-
ments  of lentic  breeding amphibians—yet  his con-
clusions stress critical  "landscape linkages" between
wetland  and  upland  habitats, without  which  the
amphibians  would  not survive. Ron  Thorn et  al.
provide a unique perspective on  restoring  shallow-
water eelgrass habitat in Puget Sound, Washington.
 The 11  contributions in Taking  a Closer Look,
cover a broad range of topics but generally focus in
greater  depth  on individual restoration  sites: Jetty
Island,  the  Salmon River,  Batiquitos  Lagoon,  an
Illinois Kettle marsh, and South Carolina bottomland
hardwoods.  Thorn  et  al. explore  the  surprisingly
difficult  task  of  accurately   determining  habitat
restoration project costs, while Marc Boule addresses
the sensitive issue of balancing restoration priorities
among different agencies during a complex  wetland
mitigation negotiation. Papers  by Lisa Levin et al.,
Scott Simon et al., and Poole and Trometer each
recommend methods for evaluating individual resto-
ration projects that could be broadly applied  in other
 Works in Progress, include 38 extended and short
abstracts.  The  first   17  of  these  are extended
abstracts—modified and expanded by  the  authors
following  the SER 95 Conference. In many cases the
addition of  literature  citations  will help the reader
find what  they need more efficiently. The remaining
21 short abstracts are reprinted just as they appeared
in   the  SER  95  Conference  Program.  They  are
included here  to illustrate the tremendous breadth of
studies being pursued by wetland and riparian habitat
 The real purposes of  a Proceedings volume such as
this are to stimulate interest in the topics covered and
encourage  active  networking  among  practitioners
across the country. Current author's addresses are
included for all contributions and readers are  urged to
contact workers  involved in  projects of particular
interest to them.

                                              New Directions in  Wetland and
                                                               Riparian Restoration
                                                                       Keith B. Macdonald1
                                                                            Fred Weinmann2
  Abstract:  Before passage of federal  environ-
 mental legislation in the late 60s and early 70s,
 wetland and riparian studies exhibited a wide
 variety of  disparate research interests. Due to
 their rapid decline and the suggestion that detri-
 tal  export  subsidized valuable coastal  fish and
 shellfish resources, coastal wetlands captured the
 greatest attention.
  More focused research themes emerged after
 passage of the Federal Water Pollution Control
 Act Amendments of 1972 (Section 404). Regula-
 tory agencies and the development community
 focused on defining the shared  fundamental
 characteristics of all wetland types—hydrology,
 hydric soils, and hydrophytic vegetation—as well
 as how they were to be identified and delineated
 in the field. Once adequately defined and classi-
 fied, then inventories  of  wetland types  and
 trends  in  habitat  abundance were   pursued.
 Clearer definition of wetland fundamentals also
 resulted in a greater focus  on unique regional
 differences among wetland types, wetland func-
 tions, and the  perceived "values" of  wetlands
 across  the  nation.  Two additional   research
 themes emerged during this period: the devel-
 opment of treatment wetlands, engineered to en-
 hance water quality; and the rapid-expansion—
 driven  largely  by mitigation requirements—of
 habitat restoration projects.  Subsequent studies
 examined  the relative success of mitigation wet-
 lands—particularly the  degree to  which they
 duplicated the functions and  processes of natural
 wetlands. In the 70s and 80s, wetland and ripar-
 ian habitat restoration  projects were planned
 and installed  project-by-project,  site-by-site.
 Typically,  we stood at the edge of the wetland
 looking in—focusing on  what we saw, how  it
 functioned,  and how it could  be duplicated.
Important new trends have emerged  since the
early 90s,  changes  captured  by  the  SER
Conference theme of Taking A  Broader View.
These new trends are being driven by the need
for cost-effective regulatory streamlining, as well
as increasing recognition that critical issues such
as water quality enhancement, sensitive species
protection,  and habitat restoration need to  be
addressed in a broader landscape, watershed, or
ecosystem management context.  Wetland and
riparian restoration projects are now seen as in-
tegral parts of the broader landscape and there
is  a critical need to identify landscape linkages
and focus on interactive processes that "drive"
the structure  and  function of these  habitats.
Wetland  and  riparian habitat  restoration  is
rapidly moving beyond regulatory-driven  miti-
gation projects into incentive-based, larger scale,
watershed enhancement and  habitat conserva-
tion planning programs. These larger programs
require more careful analysis of project  goals
and  objectives, more effective decisionmaking,
and increasing public involvement.

 One could say that wetlands, as we think of them
today, were officially created after passage of the
Federal Water Pollution Control Act Amendments
of 1972 (FWPCA, PL 92-500; amended again as the
Clean Water Act in 1977). Prior to that time  "wet-
lands" was a colloquial term used mostly by farmers
and sportsmen  to  describe  riverbottom  lands and
seasonally flooded meadows and forests. A contem-
porary dictionary definition was, "land containing
much soil moisture" (Webster 1965). While wetland
research had been expanding in the 50s  and 60s it
stressed   specific   habitats—salt  marsh, prairie
potholes, Cypress swamps, peatlands—and had not
1 CH2M HILL, P.O. Box 91500, Bellevue, Washington 98009-2050.
2 USEPA Region 10, 1200 Sixth Avenue, Seattle, Washington 98101.

New Directions
yet focused on their underlying shared fundamentals
of hydrology, hydric soils, and hydrophytic
vegetation. Conversion of wetlands for agriculture
or construction was still the norm and while the
U.S. Fish and Wildlife Service supported wetland
prote~tion (Shaw and Fredine, 1956), some other
federal agencies promoted drainage and conversion.
Along with the scientists, conservation and sporting
groups were increasingly drawing attention to the
broader environmental consequences of continued
wetland and riparian habitat losses.

Section 404 of the FWPCA, the Dredge-and-Fill
Permit Program, provided the first effective tool for
wetlands protection. In 1977, President Carter's
Executive Orders on the Protection of Wetlands and
Floodplain Management (11990 and 11988, respec-
tively) initiated federal policy reviews that ex-
panded wetlands protection. Subsequently, in 1988,
presidential candidate George Bush's support for
the concept of "No Net Loss" (Conservation Foun-
dation 1988) provided the foundation for present
policies that seek to increase wetland and riparian
acreages across the United States.
Much of the growth in wetland and riparian re-
search, as well as today's expanding restoration pro-
grams, ultimately sprang from these changing
federal mahdates and their incorporation into state
and local statues.
50 Where Have We Been?
What then were some of the older themes of wet-
land and riparian research? And how has the direc-
tion of research changed to meet the needs of new
wetland management regulations?
Early Wetland Research3

The earliest systematic studies of United States
wetlands followed passage of the Swamp Land Act
(1849, as amended). This Act transferred "swamp
and overflow lands" from the federal to state gov-
ernments and actively encouraged wetland drainage
and "reclamation". Various reports document the
reclamation potential for both interior riverbottom
(Mississippi Basin) and coastal wetlands (Wright
1907, Nesbit 1885). U.S. Coast Survey charts pro-
duced at about this same time often include detailed
mapping of coastal wetlands.
Wetland and riparian publications were scarce be-
fore 1960 and generally focused on specific disci-
plinary concerns. Early interests included the land-
building role of coastal marshes (Shaler 1885) and
mangroves (Davis 1940), and their relationship to
changing sea levels (Chapman. 1938). Coastal wet-
land natural history observations (Yapp 1917, Dex-
ter 1947) and floristics (Purer 1942, Mason 1957)
progressed into studies of primary production
(Penfound 1956, Pomeroy 1958) and energy flow
(Teal 1958, Smalley 1959).

Vegetation dynamics and energy flow within cool
temperate peatlands-bogs (muskegs) and fens-also
received early attention (Transeau 1903, Lindeman
1942). Study and protection of prairie pothole
marshes was promoted by duck hunting organiza-
tions as early as 1916 to support production of
migratory waterfowl (Linduska and Nelson 1964).
Early riparian studies focused on vegetation
changes in response to river channel dynamics and
flooding (Fitzpatrick and Fitzpatrick 1902, Turner
1936, Thompson 1961).
Ecosystem Studies

The close of the 50s saw two important bench-
marks for wetland research. In March 1958, the
National Science Foundation sponsored the first
ever interdisciplinary conference on coastal salt
marshes, at the University of Georgia Marine Insti-
tute, Sapelo Island (Ragotzkie 1959). Then, 2 years
later, publication of Salt Marshes and Salt Deserts
of the World (Chapman 1960) provided the first
monographic treatment of wetland research. Both
events greatly increased awareness and visibility of
research on coastal wetlands. Soon after the Sapelo
Island conference, John Teal published his pivotal
salt marsh trophic synthesis paper suggesting that
45 percent of marsh production was exported "to
support an abundance of animals," in adjacent estu-
aries (Teal 1962). Coastal oceanography was a
beneficiary of sharply increased federal research
funding in the early 60s. Teal's suggestion of mas-
sive detrital export captured attention and wetland
research rapidly expanded, especially along the
Atlantic and Gulf Coasts.

U.S. Fish and Wildlife Service initiated the
National Estuary Study (USFWS 1970), and major
symposia were published on estuaries (Lauff 1967)
3 Representative studies only; see Mitsch and Gosselink (1993) for others. Crowcroft (1993) and Wilson (1994) provide
valuable insights into changing ecological perspeCtives since the 30s.

                                                                                        New Directions
and coastal lagoons (Castanares and Phleger, 1969).
Much of the steadily expanding research of the 60s
and 70s  focused on the role of marsh production
(Kibby et al. 1980), nutrient flux, and detrital export
in estuarine  ecosystems (Nixon 1980).  Research
soon suggested a linkage between marsh production
and offshore commercial fisheries (Turner 1977).
Along with concerns over habitat loss from duck
hunters and sports fisherman, it was the potential
linkage between destruction of estuarine wetlands
and declining finfish and shellfish  resources that
provided the strongest political support for regula-
tory protection of wetlands (Clark 1967,  Chabreck
1973, Nixon 1980).
  The federal regulatory landscape was also chan-
ging markedly during this  period.  The  Fish and
Wildlife Coordination Act (USFWS) was enacted in
1967,  the  National Environmental Policy  Act
(NEPA)  in 1969, and both the Endangered  Species
Act and Coastal Zone  Management Act in  1972.
The FWPCA of 1972 was also further amended as
the Clean Water Act in  1977. In many cases, pas-
sage of these federal regulations led to  establish-
ment of closely similar state mandates. The overall
result was closer scrutiny of the potential impacts of
construction projects, a variety of tools to protect
wetland habitats, and the potential use of mitigation
to offset wetland losses (ASWM 1986).

Regional Differences
  The emergence of NEPA and the Clean Water Act
as tools to protect  wetland  and riparian  habitats
against the "onslaught of development" was eagerly
embraced in the  environmentally  active 70s and
80s. There was increasing concern,  however, that
the functions attributed to Atlantic Coast wetlands,
especially  Georgia's intensively  studied  Sapello
Island   marshes,  were  neither  broad   enough
(Gosselink et al. 1974), nor universally characteris-
tic (Onuf et al. 1979).  A national symposium  or-
ganized  in  1978  by  the  National   Wetlands
Technical Council and the Conservation Foundation
broadly addressed the issue of wetlands  functions
and values and resulted  in a landmark publication
(Gresson et al.  1979). Research on  wetland func-
tions led in two different directions, one focusing
on  the  classification of different wetland  types
(Cowardin  et  al.   1979),   while   the  second
documented characteristics of wetlands in different
regions of the United States.
 Research  on wetland functions and classification
led to the development of methodologies for wet-
land functional  assessment  (Adamus 1983), that
ultimately resulted in  development and application
of the presently evolving  hydrogeomorphic  wet-
lands classification system  (Brinson  1993,  1995,
1996; Smith et al. 1995).
 Descriptive literature on  regional  wetlands ex-
panded rapidly in the 70s. The USFWS and  Cali-
fornia Department of Fish and Game published the
Coastal Wetland Series, describing priority coastal
wetlands in  California  (e.g., Macdonald  1976).
Other states quickly followed with similar publica-
tions  (e.g., USACOE 1976;  Shapiro &  Assoc.
1978). The immediate success of these reports, both
as a foundation for future research and for manage-
ment purposes, resulted in  an expanded  series of
USFWS publications, Community Profiles  and  later,
Estuarine Profiles.  While the initial impetus was
coastal wetlands (Zedler 1982, Nixon  1982), studies
of freshwater wetlands and riparian habitats (Faber
et al. 1989) soon followed.  In addition, numerous
studies  of  individual  wetlands, nationwide  (e.g.,
McCormick et al. 1970),  and several  national  sym-
posia focused  on  wetland  types and  processes
(Good et al. 1978, Hamilton and Macdonald 1979,
Brinson et al. 1981, Warner and Hendrix 1984).

Wetland  Delineation
 As a direct practical response  to calls for broad
regulatory protection of wetlands, both the  regula-
tory agencies  and the  development  community
needed to know exactly what was being protected,
how it was to be defined, and how a legally defen-
sible boundary line could be  drawn around it on the
ground. This, of course, led  to  extensive debate
concerning a  legal definition of wetlands and/or
"jurisdictional wetlands" and on standardized meth-
odologies that could be employed for wetland de-
lineation (USACOE  1979,   1987;  Federal   Inter-
agency Committee  1989, EDF/WWF  1992). Publi-
cation  of  the  National   Research  Council's,
Wetlands:  Characteristics and Boundaries  (NRC
1995), has helped  achieve  federal  consensus on
these issues, but the debate continues on some state
and local levels.

New Directions
Wetland Inventories

Along with wetland delineation issues there was
substantial interest in knowing how the nation's
wetland and riparian habitats were distributed, how
much 'acreage of different wetland types remained,
and how much had been lost.
Early inventories are described by Shaw and
Fredine (1956), who focused their own studies on
wetlands for waterfowl support. Following devel-
opment of the Cowardin et al. (1979) wetland clas-
sification, the USFWS initiated the National
Wetlands Inventory (NWI) based on color-infrared
aerial photography. Since field reconnaissance to
confirm general wetland boundaries remains a slow
and costly process, NWI status varies by location.
Some areas of the country still remain to be mapped
while in others NWI is already well into its second
generation of map products. Earlier studies, along
with the NWI data are summarized by Dahl (1990)
and Dahl and Johnson (1991). These data offer our
first quantitative overview of the nation's wetland
resources and provide a baseline against which to .
measure future changes.
. Treatment Wetlands
Increased understanding of wetland and riparian
nutrient fluxes, along with monitoring of wastewa-
ter discharges into natural wetlands, identified the
critical role these systems can play in maintaining
water quality (Gosselink et al. 1974). In the 70s, this
knowledge was applied to a new research theme-
the development of treatment wetlands, expressly
engineered and built for water quality treatment.
This activity has steadily gained importance, con-
tributing significantly to better understanding of
natural wetland functions and practical field meth-
ods for wetland enhancement and creation (Kadlec
and Knight 1996). Under appropriate circum-
stances, treatment wetlands have been shown to
offer subsidiary benefits for wetland-related species
(USEPA 1993). The possible accumulation of
contaminants (trace metals, organic chemicals, etc.)
can, however, remain a concern (Wren et al. 1997).
An increasingly important research theme and the
one driving these Proceedings, concerns wetland
and riparian habitat restoration, enhancement, and
creation. Although occasional pioneering wetland
restoration projects have been implemented for
many years, the field expanded rapidly in the late
70s (Seneca et al. 1975, Garbish 1977, Montanari
and Kusler 1978, Josselyn 1982). This reflected the
passage first of NEP A (1969), which provided the
framework for mitigation programs to balance pro-
ject impacts, and then Section 404 of the Clean
Water Act (1977), which required compensation for
wetland losses.
In subsequent years, analysis of 404 permit deci-
sions showed that despite mitigation requirements
. wetlands continued to be lost (Marcus 1982, Gwin
and Kentula 1990). Further, monitoring of early
wetland restoration programs showed they often feU
short of expectations, particularly in terms of re-
placing some of the functions and processes typical
of comparable natural wetlands (Race 1985, Stick-
land 1986, Zedler 1988, PERL 1990). Clearly the
time required for a restored or created wetland to
replace natural wetland functions is much longer
than first anticipated (> 10 years) and some func-
tions may not be fully replaced (Simenstad and
Thom 1996, Zedler 1996b). These insights ulti-
mately led to political support for the "No Net
Loss" wetlands protection policy (Conservation
Foundation 1988).

The late-80s and 90s have seen publication of in-
creasing numbers of guidebooks recommending
detailed procedures for planning, installing, and
monitoring wetland and riparian restoration and en-
hancement programs (Kusler and Kentula 1990,
Kentula et al. 1992, Zedler 1984, 1996a; Marble
1992, Vanbianchi et al. 1994, Erwin 1996).
Publications focusing on wetland support for
waterfowl (Scott 1982, Smith et al. 1989) and wild-
life (Payne 1992), as well as riparian/stream
benefits for fisheries (Hunter 1991, Spence et al.
1996) have also become available; The credibility
of aquatic restoration as a scientific discipline also
gained substantially with publication of two major,
national compendiums: Restoration of Aquatic Eco-
systems by the National Research Council of the,.
National Academy of. Science (National Academy
Press 1992) and Restoring the Nation's Marine En-
vironment (Thayer 1992).

The newer research themes noted above never
eliminated the older ones-rather they have steadily
expanded the breadth of wetland and riparian

studies  now in progress.  There are now several
excellent  textbooks  covering different aspects  of
wetlands or wetland types (Daiber 1982, Mitsch and
Gosselink   1986,  1993;  Adam  1990,  Malanson
1993).  Wetland-oriented   professional   societies
(Society of Wetland Scientists, Association of State
Wetland Managers)  and journals  (Wetlands,  Wet-
lands  Ecology  and  Management)  have  become
popular, and both peer-reviewed and grey literature
on wetlands has  exploded. The  first edition  of
Mitsch  and Gosselink's text Wetlands (1986) in-
cluded  nearly 700 literature citations; just 8 years
later in their second edition (1993) this number had
risen to nearly 1,500 citations!

The Broader View  Trend
  The accelerating worldwide decline in  biodiver-
sity—the  total  variety of life forms  living on the
planet—has become the overarching environmental
concern of the  90s. Human modification and de-
struction of habitat is clearly recognized as the prin-
cipal  cause of this decline, reflecting in turn the
rapidly expanding  human  population. In Pogo's
now famous admonition  (Earth Day 1971),  "We
have met the enemy and he is us." Public opinion
polls  suggest that protecting biodiversity, like wil-
derness before  it, has now become a significant
quality  of life issue.
  The 90s may have seemed like a tough decade for
the  environment  with  a  considerable  backlash
against the regulatory protections introduced in the
70s. Private property rights are certainly a conten-
tious issue, but frustration  over the complexity and
cost of fragmented,  overlapping, and sometimes
contradictory regulatory mechanisms is a  much
more widespread concern (US General Accounting
Office 1994, McCormick et al. 1995).
  In the 70s and 80s most environmental concerns
were dealt with on a case-by-case basis: site by site,
outfall by outfall, tract by  tract (whether forests or
houses). Regulations were  often interpreted with a
narrow  focus and public involvement was generally
limited. Greater awareness of both  the  need for
regulatory streamlining and the accelerating impact
of habitat  degradation has resulted in significant
changes in how environmental issues are now  being
addressed.  The  90s have witnessed  some striking
changes captured by, "Taking a Broader View," the
guiding   theme   for  the   1995   International
                                  New Directions

Symposium of the Society of Ecological Restora-
tion. Examples illustrating these newer approaches
to several issues germane to wetland and riparian
restoration are outlined below.
Multidisciplinary Training and
Hypothesis Testing
  Given the prominence of wetland issues today, it
is easy to forget that as little as  a decade ago few
students were formally trained as wetland scientists
or restorationists. Rather, individuals "got hooked"
while studying wetland and riparian habitats  from
their  own particular  disciplinary   perspective—
geomorphology, groundwater, botany,  migratory
birds, nutrient cycling, etc. What most of us found
was  that to better interpret our own disciplinary
concerns, we needed to better understand the system
as a whole. Recognition of the benefits of a  truly
multidisciplinary/interdisciplinary approach to wet-
land and riparian science and restoration is  now
accepted as central to our further progress.
  Another important lesson is that it is costly and
ineffective  to  continually  "rediscover" restoration
fundamentals.  We must learn from and build  upon
the experience of earlier workers. This requires that
restoration projects always be designed as  experi-
ments  to  test  clearly  defined hypotheses—about
hydrologic  modifications perhaps, or soil amend-
ments, planting methods, or control  of introduced
species. Effective monitoring is also a critical  com-
ponent if we are to learn anything from experience.

Water Quality Enhancement
  Successive water quality initiatives, focusing first
on point source controls,  adding nonpoint source
controls, and  now  evolving into the watershed-
based "total maximum daily load" (TMDL) process,
clearly follow the trend of taking a broader view.
  For decades,  the primary  water quality  concern
was  treatment of municipal sewage  and industrial
waste discharged into  waterways through ditches,
pipes,  and outfalls—so-called point source pollu-
tion.  The Federal  Water  Pollution Control Act
Amendments   of 1972 were  broadly  crafted to,
"restore and maintain  the chemical, physical, and
biological integrity of  the nation's waters." Imple-
mentation of  the Act initially  focused on  point
source discharges, subjecting them to water-quality

New Directions
and technology-based controls imposed by the
National Pollution Discharge Elimination System
(NPDES) permit process. These controls are gener-
ally imposed outfall-by-outfall, one permit at a time..
Overall, NPDES controls have been very success-
ful, producing dramatic water quality improvements
that allowed marine and freshwater fish, and other
organisl1ls, to return to previously degraded habi-
tats. Yet significant pollution remained.

Since the 70s, EPA-sponsored studies of nonpoint
source pollution have demonstrated that stormwater
flows draining from diffuse locations spread out
over the landscape-from street runoff, faulty septic
systems, land clearing, agriculture, forestry, and
construction practices-also cause significant pollu-
tion of wetlands, streams, and other aquatic habitats
(A WRA 1988, Jackson 1994, Hopkinson and Val-
lino 1995, Pitt 1995). Under the FWPCA, these
no~point sources have been subject to increasingly
stnngent state regulation through "best management
practices" (BMPs) tied to specific industry types
and land uses (Washington Department of Ecology
1992). Joint EPA-NOAA guidance for new state
coastal nonpoint pollution control programs was
implemented in 1993, following the Coastal Zone
Act Reauthorization Amendments of 1990 (EP A
A potential shortcoming of prior cleanup actions
has been that neither point source NPDES permits,
nor state-imposed non point BMPs, consider the
cumulative load of pollutants entering a particular
waterbody, nor its ability to satisfactorily assimilate
and neutralize the total waste load. This need was
foreseen in the FWPCA amendments but has only
very recently begun to be implemented. Sec-
ti~n ~~5(b) of the Act requires states to identify and
pnontIze s~ream segments or waters not meeting
water qualIty standards. Section 303(d) then im-
poses the TMDL requirement that mandates reduc-
tion of pollution to meet water quality standards
designed to protect beneficial uses of each stream
segment or waterbody. A TMDL provides a mecha-
nism for quantitative assessment of water quality
problems and contributing pollutant sources; allo-
cates pollution control responsibilities between
sources in a watershed; and specifies the amount of
pollutant or other stressor to be reduced to meet
water quality standards. EP A recognizes habitat
degradation impacts, and restoration of specific
habitat endpoints (e.g., fish spawning habitat) may
be allowed as part of the TMDL (EP A 1994, Dilks
et al. 1994, WFCA 1997).
Endangered Species Recovery

Changing approaches to implementing the Endim-
gered Species Act (Rohlf 1989, Pulliam and Babbit
1997) also illustrate the trend towards taking a
broader view. Following passage of the Act in 1973
and the first federal listings of threatened and
endangered species, the initial goal was to summa-
rize ecological life history data for each species and
develop an appropriate Recovery Plan (USFWS
1980, 1985). The Act relied on prohibition of a
b~oadly defined "take" to protect endangered spe-
CIes. It soon became apparent, however, that despite
each species' unique ecological requirements, the
common theme shared by all was loss of critical
habitat. The Act permitted "incidental take" of a
listed species following preparation and approval of
a habitat conservation plan (HCP). Early HCPs
focused on the protection of single species, often by
transferring development rights from critical habitat
and clustering higher-density construction on
nearby, less sensitive areas (San Bruno Mountain
HCP, Beatley 1995).

.Focusing habitat protection to favor single spe-
CIes, even endangered species, revealed several con-
cerns: (a) complex, poorly understood life histories
often make it difficult to decide exactly which
habitat features-or other ecological interactions-
favor survival; (b) enhancing habitat characteristics
for one species may place other species occupying
the same habitat at greater risk; and (c) a single en-
dangered species often signals critical losses of a
particular habitat, indicating that several other co-
occurring species are soon likely to be at risk.
Southern California coastal wetlands, for example,
provide critical habitat for several different endan-
gered species--each with different habitat require-
ments (Zedler 1996a). Similarly, local riparian wil-
low thickets provide critical nesting habitat for an
endangered neotropical migrant songbird, while
several other migrants also appe.ar to be in decline.
(Recon 1989). These concerns"-along with new in- .
sights on habitat size and landscape distribution
required to support sustainable species populations
(derived from island biogeography, MacArthur and
Wilson 1967, Soule et al. 1988)-have ultimately
led to the development of large-scale, regional,

                                                                                          New Directions
 multiple species HCPs (City/County of San Diego
 1996). This suggests that HCPs are evolving from a
 process adopted primarily to address single species
 and (construction) development, to a broad-based,
 landscape  level  planning tool utilized to  achieve
 long-term biological and regulatory goals (USFWS/
 NMFS 1996). While the success of such plans is
 still in debate (Kaiser 1997, Mann and Plummer
 1997, Shilling 1997, Luoma  1998), they certainly
 reflect taking a broader view.

 Wetland Economic Benefits
  Society's view of the benefits (values) provided
 by wetland and riparian resources has changed sub-
 stantially as understanding of wetland processes and
 functions has grown (Greeson et al.  1979). Aborigi-
 nal uses  of wetland resources focused on limited
 harvest of plants for food and fiber, fish and  shell-
 fish, and trapping of fur-bearers. Early trappers and
 pioneers  showed  less  restraint  and quickly  ex-
 ploited,  then exhausted, many of  these  same re-
 sources. The "wetlands are worthless swamps" ethic
 persisted well into the 70s, until their role as nurs-
 ery grounds  for coastal fisheries was understood
 (Clark 1967). In addition to their biological  produc-
 tivity and  food chain  support,  Gosselink et al.
 (1974) drew attention  to  wetlands' role in  water
 quality enhancement and  nutrient cycling. Others
 noted their role attenuating flood flows and  medi-
 ating groundwater flux.
  The critical role of riparian forests in creating and
 maintaining stream habitat  for  trout and  salmon
 (Spence et al. 1996), as well as habitat for neotropi-
 cal migrant songbirds (Terborgh 1989), and of prai-
 rie  potholes  and  coastal marshes  for migratory
 shorebirds and waterfowl, are all now more widely
 appreciated. Today wetland  and riparian  habitats
 are also endowed with recreational and aesthetic
 values,  as  well  as  a  broad  array  of  critical
 "ecosystem services"  (ESA 1997). The long-term
 trend from harvestable goods to ecosystem services,
 and less tangible aesthetic values, again illustrates
 the trend of taking a broader view.

 Wetland Restoration Incentives
  Prior to the Clean Water Act 404 Permit Program,
 limited wetland and riparian habitat restoration was
 accomplished by  USFWS,  some  state  resources
 agencies, and private conservation  groups  such as
 the  Native  Conservancy, Ducks  Unlimited, and
Trout Unlimited. Since implementation of the 404
program, mitigation responsibilities have accounted
for the majority of wetland and  riparian  habitat
restoration. Since mitigation is necessarily driven
by development activities, its proponents are not
always the most knowledgeable about habitat resto-
ration, nor committed to more than regulatory suc-
cess.  The present  regulatory  framework—with
mitigation negotiated on a  site-by-site basis—has
often  resulted in small, inappropriately sited wet-
lands  with limited functional value and at a very
high cost-per-acre  restored  (Weinmann and  Kunz
1993). Proactive mitigation banking in which  small
mitigation projects are combined into larger units
can provide several advantages, including a vehicle
for basin-level  planning (Reppert 1992, ELI  1993,
Green 1995a4).
  Some private sector consultants  are  reporting a
decline in development-related wetland mitigation
projects. Perhaps the regulations are working and
more  projects are successfully avoiding impacts to
wetlands, as well as the regulatory expense and de-
lays they are likely to cause. Agencies with major
projects that cannot avoid  wetland and riparian im-
pacts  (port expansions, highway construction), are
actively promoting mitigation banking and river ba-
sin planning (Reppert 1992, USDOT 1997). A 1992
Seattle workshop,  Partnerships and  Opportunities
in Wetland Restoration (USEPA 1993) provided a
call to arms, encouraging innovative restoration op-
portunities   beyond   mitigation   requirements
(LaFountain 1996).
  Historically,  the greatest  wetland and  riparian
losses have  resulted from agriculture,  silviculture,
and grazing, yet these activities were exempt from
404 Permit provisions and until 1985, the  Depart-
ment  of Agriculture continued to subsidize wetland
drainage projects  (Mitsch  and Gosselink  1993).
This  changed in 1985 with passage of the Food
Security Act "Swampbuster" provisions which re-
placed the drainage  subsidies with incentives  to
farmers  to set  aside farmed wetlands  for habitat
protection  (Conservation  Reserve  Program and
Wetlands Reserve Program; Robinson 1996). The
USFWS estimated that 90,000 acres of wetlands
were restored under this program between 1987 and
1990  (Josephson 1992). In 1996 implementation of
a  CRP State  Enhancement  Program allowed the
Department   of  Agriculture  to  approve  state-
submitted cost  sharing proposals that use  CRP to
4 This designation refers to another paper included in this Proceedings volume.

New Directions
address regional enhancement of a particular estu-
ary, river, or ecosystem. Plans are presently under
review to protect and restore environmentally sen-
sitive lands surrounding Chesapeake Bay, Maryland
and in tije Minnesota River Watershed, Minnesota
(Minnesota 1996). The potential to apply the pro-
gram to further salmon recovery in the Pacific
Northwest is also being explored (EDF 1997,
Searchinger 1997).

Within the last few years several of the trends
outlined above have been coming together-partly
in response to federal and state initiatives that en-
courage both ecosystem management and water-
shed-based management. In many cases, these
changes have equally important proactive regional!
watershed planning and regulatory streamlining
components. Some new players are leading these
With diverse habitats, "from the desert to the sea,"
and a high concentration of sensitive endemic spe-
cies (Dobson et al. 1997), San Diego County has
embarked on a comprehensive MSCP and growth
management program. The plan protects a network
of core habitat preserves (mostly local, state, and
federal public lands) connected together by riparian
corridors. Key parcels threatened by development
can be purchased to complete the preserve network,
while regulatory and financial incentives will focus
development away from scarce critical habitats into
less sensitive areas (City/County of San Diego
Further north, the Washington State Department
of Transportation (WSDOT) is also on the cutting
edge, proactively planning for long-term transpor-
tation growth on a watershed basis. Using GIS tech-
nology, WSDOT can pull up stream corridors,
wetlands, fish passage problems, stormwater cul-
verts, water quality data, sensitive species distribu-
tions--even selected road-kill data to determine
wildlife migratory corridors-within a chosen
radius of a proposed project. Working with the
regulatory agencies and local communities,
WSDOT can then determine the best combination
of project design features and mitigation measures
to maximize the overall environmental benefits. By
achieving concensus, limited funding can be lever-
aged into more creative, cost-effective mitigation
(USDOT 1997; Shari Schaftlein, WSDOT personal
communication 1997).

These examples demonstrate that incentives for
restoration and enhancement of wetland and ripar-
ian habitats are now moving well beyond the regu-
latory focus of mitigation requirements. The need
for habitat restoration is clearly increasing, but now
it is in the broader landscape context of watershed
management (Wear et al. 1996, Kentula 1997).
Reestablishing natural wetland and riparian habitat
functions and processes is increasingly seen as an
appropriate response to reduce flood damage, en-
hance water quality, and sustain populations of sen-
sitive species.
What Does This Mean for Wetland
and Riparian Restoration?

The 90s are proving to be a tremendously exciting
and challenging time for wetland and riparian re-
search and restoration. Largely through develop-
ment of landscape ecology (Forman and Godron
1986) and increasing regulatory interest in water-
shed management, our perspective has been funda-
mentally shifted from developing individual
restoration projects to seeing such projects as inte-
gral parts of the larger landscape. Now more than
ever we are challenged to seek understanding and
solutions "outside the box." What then, are some of
the ways that the trend of "taking a broader view"
might alter our practice of wetland and riparian
Wetlands Occur in Landscapes

Landscape ecology, integrating ecological con-
cepts with human activities, has introduced new
perspectives to wetland and riparian studies. How
are landscape elements and ecosystems distributed;
how do energy, materials, or biota flow among
these elements; and what changes might occur in
the landscape mosaic over time (Godron and
Forman 1986)?

One of the most significant. changes is that we
must turn our focus from'inside the wetland (or ri-
parian corridor) to outside the wetland and to its
surrounding landscape (Wear et al. 1996, Kentula
1997). Application of the hydrogeomorphic classifi-
cation of wetlands (Brinson 1993, 1995, 1996)
helps in this regard. How does a particular wetland

                                                                                         New Directions
site fit into the surface and groundwater hydrologic
patterns  and other processes of  the  surrounding
landscape? How do other landscape functions and
processes impact the  wetland, and visa versa? We
need to better understand what is going on in the
surrounding landscape outside the wetland and then
apply this knowledge to better understand how it
impacts what  is happening inside the  wetland. As
Klaus Richter recently put it so elegantly, "You can
engineer these (wetland)  systems  in the best way
possible—but if you put them in the wrong place they
are not going to work! "

Wetlands Function in Watersheds
  It is now recognized that even modest watershed
development—whether urbanization or construction
of forestry roads—accelerates the  initiation of run-
off, as well as increasing its quantity and velocity.
Significant changes begin to occur when as little as
10 percent of the watershed is converted to imper-
vious surfaces. These changes increase the magni-
tude and frequency of flooding  events in  stream
systems  and  wetlands.  This typically results  in
channel downcutting  (which  in turn lowers the ad-
jacent water table), increased bank erosion, and thus
increased sediment flux—as well as the  potential
for increased  contamination  via the runoff (Booth
and Reinelt  1993, Pitt et al. 1995,  May,  1996).
Wetland and  riparian restoration  design must ac-
commodate these potential changes (REIC  1995,
King County 1996).
  Failure  to  account  for these  urban  runoff and
stormwater impacts is responsible  for the failure of
many wetland and riparian enhancement programs
(Shaffer and Kentula  1995a). Horner (1992) has in-
dicated  however,  that  widely  used  engineering
models systematically underestimate  runoff rates,
and stormwater specifications thus need to be sig-
nificantly improved. Unless we clearly understand
the significance and  implications of these  funda-
mental modifications  to watershed hydrology—and
thus to related wetland and stream processes (Poff
et  al.  1997)—our  habitat   restoration  and  en-
hancement activities are always likely to fall short
of their  goals (Frissel and  Nawa 1992, Chapman
Focus on Processes
 Collection of descriptive physical, chemical, and
biological  data from reference sites—typically,
pristine  or  minimally  disturbed  habitats—will
always play a key role in restoration planning, as
well as in  monitoring project success. Identifying
and understanding the natural processes that under-
lie  and "drive" these site  characteristics is  even
more important, however (Frenkel 1995a, Simon et
al. 1995a, Winfield et al. 1995a).
 Fuerstenberg,  Nelson,  and  Blomquist  (1997)
documented structural changes in stream character-
istics,  wetland and  riparian habitat,  and salmon
populations of Puget  Sound's Green River, as the
Green  River Valley was  increasingly modified by
140 years of settlement, agriculture, and urbaniza-
tion. Increasing human modification and control of
river processes resulted in reduced natural variabil-
ity  within the river system, less habitat diversity,
and declines  in macroinvertebrate  and  salmonid
richness. Fuerstenberg et al. concluded that the spe-
cific physical,  chemical,  and biological  structures
and functions observed in the field reflected the un-
derlying, variable processes (e.g., flood pulses, sedi-
ment flux,  woody  debris   recruitment,  nutrient
transformations) operating  in aquatic ecosystems.
Comparison of historic structure and function with
present day attributes helps  define how the under-
lying processes operate. This, in turn, provides an
ecological  basis and organizing framework for re-
gional  habitat  restoration and management  plan-
ning. To quote Bob  Fuerstenberg  (King County,
Seattle, WA), "You've got to ask yourself, what are
the field data telling you about the system?"
 Commenting  on watershed management plans in
general, Dave Caraher (U.S. Forest  Service,  Port-
land, OR) recently echoed this theme, "We do these
detailed science studies but don't take the time to
understand what they really mean. You need  to go
sit  on a rock—out in the  watershed—and  think
about what you 've learned."

Look for Landscape Linkages
 In addition to physical (e.g., hydraulic connec-
tions, groundwater and sediment flux) and chemical
(e.g., nutrient cycling) processes that link wetland
and riparian habitats to  their  surrounding  land-
scapes, potentially important biological interactions
must  also  be  identified   (Fuerstenberg  1995").

New Directions
Wetland and riparian vegetation stabilizes the banks
of rivers, streams, and tidal channels, reduces
erosion and sediment flux, and can reduce nutrient
and pollutant inputs moving between uplands and
adjacent aquatic habitats. The height, density, and
species composition of riparian vegetation can
modify shade, water temperatures, litter fall, and the
contribution of large woody debris (L WD) to
adjacent streams, as well as influencing the
availability of insect prey. The presence of L WD
adds critical habitat complexity, trapping sediment
and enhancing the development of pools and riffles
for fish. The size and species of L WD-western
cedar versus alder, for example-influences its
persistence in the system (Berg 1995a, Rot et al.
While wetland and riparian habitats support their
own unique flora and fauna, they also provide sup-
port during critical life stages of large numbers of
species that live much of their life cycle in nonwet-
land habitats (Mitsch and Gosselink 1993). Resto-
ration programs need to consider the needs of both
resident and transient species. Richter's (1995a)
lentic breeding amphibians include examples of
both permanent wetland residents and other species
thai must move- freely between wetland ponds and
adjacent upland forest habitat to prosper. Anadro-
mous fish, such as salmonids (Spence et al. 1996)-
as well as many species of marine fish and crusta-
ceans-depend on shallow wetland habitats during
larval development, but move out to sea as adults. A
wide variety of migrant shorebirds, waterfowl and
songbirds depend on wetland and riparian areas
during their seasonal migrations. Knowledge of the
biological requirements and interactions of these
species can significantly influence restoration loca-
tion and design.

Species associated with upland habitats can also
provide critical support for wetland dependent spe-
cies. Zedler (PERL 1990) provides the example of
upland pollinator species-bees characteristic of
adjacent coastal-scrub dominated upland---critical
to successful seed production in the endangered an-
nual salt marsh bird's beak (Cordylanthus mariti-
mus ssp. maritimus). The Forgotten Pollinators
(Buchmann and Nabhan 1996) documents similar
more broadly based concerns.
Accessibility of restored wetland and riparian
habitats to native predators-or their protection
from them (as well as from introduced pest species
and domestic animals)-also needs to be given
careful consideration (Soule, et al. 1988, Recon
1989, Jurek 1994).
As many as possible of these various "landscape
linkages" need to be identified and appropriate de-
sign features or mechanisms for their continued
successful operation be incorporated into both re-
gional wetlands plans and individual restoration
Regional Planning for Restoration

Like the old way of doing business for water
quality enhancement or endangered species recov-
ery, wetland and riparian restoration projects are
typically planned and implemented site-by-site, one
project at a time. But what if they all follow the
same planting plan, or all emphasize achievement of
the same hydrologic goals? Habitat diversity begets
species diversity-and community resilience to en-
vironmental purtabations---clearly this is an impor-
tant function to protect.

A promising approach to address this issue is the
development of regional plans that specify the vari-
ety, relative abundance, and most appropriate loca-
tions for pursuing wetland and riparian restoration
projects within a particular watershed, geographic
region, or management area (Harris et al. 1995a;
MacKay 1995a; Stanley and Roberts, 1995a). Zedler
(1996), concerned about dredging of shallow-water
habitats to favor fish over birds, proposed this ap-
proach for scarce southern California salt marsh
habitats. The California Coastal Conservancy is
presently exploring implementation of such a plan.
Tanner (1990) favored a similar approach for re-
storing wetland sites along the heavily industrial-
ized Duwamish Channel, that flows into Puget
Sound; while Shreffler and Thorn (1995a) suggest
that a landscape ecology perspective is appropriate
for developing such a plan.
More Effective Decisionmaking

Planning for larger resto~ation programs (Scodari
et al. 1995) requires more careful analysis of project
goals and objectives: What underlying problems
generated the need for wetland or riparian restora-
tion? What are the greatest restoration needs? How
might they be most cost-effectively accomplished?

                                                                                         New Directions
 Computerized  geographic  information  systems
(GIS) have already emerged as the preferred method
for merging and displaying all of the datasets (GIS
layers)  necessary for  comprehensive  restoration
planning (MacKay 1995a, Schulman 1995", Shuman
et  al. 1995a, Toth 1995a, GIS World 1996, Kyle et
al. 1996).  The quality  of the decisionmaking and
implementation process, however, is as important as
the quality of the science  that goes into the deci-

 Making good decisions can be difficult, especially
when complex habitat restoration and sensitive spe-
cies  problems typically involve multiple stakehold-
ers,  have conflicting objectives, require technical
expertise, deal with  large amounts of money, have
multiple objectives,  and are surrounded by public
interests. Most people do not realize that when such
decisions fail it  is rarely due to  inadequate science
but commonly because  of organizational issues that
could be managed. Most decisions fail because of
organizational problems that reflect problems with
leadership, a lack of a problemsolving process or
training, poor  teamwork,  and  poor  commitment.
These problems can, and often do, lead to conflict
(Ury and Fischer 1981, Clemens  1996).
  Providing leadership in a decision process can ad-
dress these problems. Sound leadership requires that
a systematic problem solving process be used; that a
common vision be clearly understood; that the right
people  are involved;  and  that there  be a well-
defined problem, schedule, and commitment from
participants.  Remember, that implementation  of
habitat  restoration programs begins with the deci-
sion process (CH2M HILL 1997).

Increasing Public Involvement
  Where restoration programs  encompass  large
areas—in watershed management  or  multiple spe-
cies   conservation   plans,  for  example—public
involvement (CH2M HILL 1995) becomes very im-
portant. In  an era of strong competition for limited
financial resources,  the public  increasingly helps
prioritize restoration decisions. They  can also pro-
vide dedicated,  cost-effective help with field sur-
veys,   restoration    plantings,    and   long-term
monitoring programs (Goeldner et al.  1995", Pruess
and Fischer 1995"). These roles  for public involve-
ment are likely to increase as wetland and riparian
habitat restoration moves beyond regulatory-driven
mitigation projects into more incentive-based, larger
scale, watershed enhancement and habitat conser-
vation planning programs.

 Broad recognition of the critical need for cost-
effective regulatory streamlining and the need to ad-
dress the accelerating impact of habitat degradation
has resulted in significant changes in how environ-
mental issues are being addressed in the 90s.  Water
quality enhancement initiatives, having moved from
point source to include nonpoint source controls,
are now evolving into the watershed-based TMDL
process. Endangered  species  recovery  initiatives
have evolved from early HCPs focused on the pro-
tection  of single  species,  into large-scale MSCPs
that provide a broad-based, landscape level  plan-
ning tool  more suitable to achieve long-term bio-
logical  and regulatory goals.  At the same  time,
incentives for wetland and riparian restoration are
moving beyond site-by-site mitigation to encourage
wetland mitigation banking and basin-level restora-
tion planning.
 What  do these changes mean for wetland and ri-
parian restoration? Largely through development of
landscape ecology and increasing regulatory interest
in watershed management, our  perspective has been
fundamentally changed from developing individual
wetland restoration projects to seeing such projects
as integral elements of the broader landscape. We
must turn our focus from inside  the wetland to the
surrounding landscape outside  the wetland. How do
landscape functions  and processes impact  a par-
ticular  wetland site  or riparian corridor, or visa
versa? How have landscape modifications altered
natural  processes  and impacted wetlands? And how
can these trends be reversed?
 New directions for restoration in the 90s demand a
clearer  understanding of the  landscape  setting  of
wetland and riparian habitats;  better understanding
of their functions in watersheds;  a sharper focus on
the natural processes that drive these systems; and
identification of the critical linkages between wet-
lands and their surrounding landscapes. There is
also a critical need to both think about wetlands and
plan  for  their protection and  restoration,  on  a
broader scale.
  Given the strong competition for limited financial
resources that has characterized the 90s, planning

New Directions
for larger restoration programs will require more
effective decisionmaking and increased public
involvement. Wetland and riparian habitat
restoration is rapidly moving beyond regulatory- .
driven rrutigation projects into more incentive-based
larger scale, watershed enhancement and habitat
conservation planning programs.
Literature Cited

Adam, P. 1990. Saltmarsh ecology. Cambridge
Studies in Ecology. Cambridge University Press,
NY. 461 pp.
Adamus, P. R. 1983. A method for wetland func-
tional assessment. 2 Vols. U.S. Dept. of Transpor-
tation, Federal Highway Administration Rpts.
FHW A-IP-82-23/24. Washington, D. C.

American Water Resources Association (A WRA).
1988. Proceedings of the Symposium on Non-
point Pollution, 1988 - Policy, Economy, Man-
agement and Appropriate Technology. Vladimir
Novotny (ed). A WRA, Minneapolis, MN.

Assoc. State Wetland Managers (ASWM). 1986.
. National wetlands symposium: mitigation of im-
pacts and -losses. New Orleans, LA. ASWM,
Berne, NY.
Beatley, T. 1995. Preserving biodiversity through
the use of habitat conservation plans. pp 35-74, In
Porter, D. R. and D. A. Salvesen, Collaborative
planning for wetlands and wildlife. Island Press,
Washington, D. C.

Booth, D. B. and L. Reinelt. 1993. Consequences of
urbanization on aquatic systems-measured ef-
fects, degradation thresholds, and corrective
strategies. In USEPA (1994). Watershed '93 (op

Brinson, M. M. et al. 1981. Riparian ecosystems:
their ecology and status. Biological Services Pro-
gram, USFWS, Washington, D. C. FWS/OBS-
81/17.151 pp.
Brinson, M. M. 1993. A hydrogeomorphic classifi-
cation of wetlands. Wetlands Research Program
Technical Report WRP-DE-4. U.S. Anny Corps
of Engineers, Washington, D. C. 79 pp.

Brinson, M. M. 1995. The HGM approach ex-
plained. National Wetlands Newsletter, Nov-
Dec. '95:7-13.
Brinson, M. M. 1996. Assessing wetland functions
using HGM. National Wetlands Newsletter. Jan-
Feb. '96:10-16.

Buchmann, S. L. and G. P. Nabhan. 1996. The for-
gotten pollinators. Island Press, Washington D. C.
312 pp.
Castafiares, A. A. and F. B. Phleger (eds.) 1969.
Coastal lagoons, a symposium (UNAM-
UNESCO), Universidad Nacional Autonoma de
Mexico, Mexico City.

CH2M HILL. 1995. Public involvement strategies:
. a manager's handbook. Sponsored by, American
Water Works Association Research Foundation,
Denver, CO. .
CH2M HILL, 1997. Decision facilitation handbook.
CH2M HILL, Denver, CO. 173 pp.

Chabreck, R. H. (ed.) 1973. Proceedings: Coastal
marsh and estuary management symposium. Lou-
isiana State University, Baton Rouge, LA. 316 pp.

Chapman, D. W. 1996. Efficacy of structural
manipulations of instream habitat in the Columbia
River Basin. Rivers 5(4):279-293.
Chapman, V. 1938. Coastal movement and the de-
velopment of some New England saltmarshes.
Proc. Geological Assoc. of England. 49:373-384.

Chapman, V. 1960. Salt marshes and salt deserts of
the world. Plant Science Monographs Interscience
Publishers, NY. 392 pp. Second Supplemented
Reprint Edition (1974) Verlag Von J. Cramer,
Gennany. 494 pp.

City/County of San Diego. 1996. Multiple species
conservation program: MSCP Plan. (County
Subarea Plan 1997) San Diego Association of
Governments, San Diego, CA.

Clark, J., 1967. Fish and man: conflict in the Atlan-
tic estuaries. American Littoral Society. Special
Publication No.5. 78 pp.

Clemens, R. T. 1996. Making hard decisions, an
introduction 1.0 decision analysis. (2nd edition). .
Duxberg Press, Belmc>"nt, MA.
Conservation Foundation. 1988. Protecting Amer-
ica's wetlands: an action agenda. The final report
of the National Wetlands Policy Forum. Wash-
ington, D. C. 69 pp.

                                                                                       New Directions
Cowardin, L. M. et al. 1979. Classification of wet-
  lands and  deep  water  habitats of the  United
  States. USFWS  Special  Pub.  FWS/OBS-79/31.
  Washington, D. C. 103 pp.
Crowcroft, P., 1991. Elton's ecologists. A history of
  the Bureau of Animal Population.  University of
  Chicago Press, Chicago, IL. 177 pp.

Dahl, T. E.  1990.  Wetland losses in  the  United
  States, 1780s to 1980s. USFWS, Washington, D.
  C. 21 pp.

Dahl, T. E.  and C. E. Johnson.  1991. Wetlands
  status  and trends in the conterminous  United
  States mid-1970s to mid-1980s. USFWS, Wash-
  ington, D. C. 28 pp.
Daiber, F. G. 1982.  Animals of the tidal marsh. Van
  Nostrand Reinhold Co., NY. 422 pp.
Davis, J. H. 1940. The ecology and geologic role of
  mangroves in Florida. Carnegie Institute, Wash-
  ington, Pub. No. 517:303-412.
Dexter, R. W. 1947. The marine communities of a
  tidal inlet at Cape Ann, Massachusetts: a study in
  bio-ecology. Ecological Monographs 20:261-294.
Dilks, D. W. et al. 1994. Development of the  San
  Luis Obispo Creek demonstration TMDL. pp 367-
  372, In USEPA (1994). Watershed '93 (op cit).
Dobson, A.  P., et al. 1997.  Geographic distribution
  of endangered species in the United States. Sci-
  ence 275:550-553.
Ecological Society  of America (ESA).  1997. Eco-
  system services: benefits supplied  to human so-
  cieties by natural ecosystems. Issues in Ecology 2.
EDF/WWF. 1992.  How wet is a wetland? The im-
  pacts of the proposed revisions to the Federal
  Wetlands   Delineation Manual.  Environmental
  Defense Fund, NY, and World Wildlife Fund,
  Washington, D. C. 170 pp.
Environmental Defense Fund (EOF). 1997. About
  the Conservation Reserve Enhancement Program:
  a Federal/State partnership; and, Use of CRP to
  further salmon recovery in the Northwest. Infor-
  mation Pamphlets, EOF, Washington, D. C.
Environmental Law Institute (ELI). 1993. Wetlands
  mitigation  banking.  ELI,  Washington,  D. C.
Erwin, K. L. 1996. A bibliography of wetland crea-
  tion and restoration literature. Assoc. State Wet-
  land Managers, Berne, NY. 99 pp.
Faber, P. M. et al. 1989.  The ecology of Riparian
  habitats of the Southern California coastal region:
  a community profile. USFWS, Washington, D. C.
  153 pp.
Federal  Interagency  Committee  for   Wetlands
  Delineation.  Federal manual for identifying  and
  delineating jurisdictional  wetlands.  USACOE,
  USEPA, USFWS, and USDA Soil Conservation
  Service, Washington, D.C.  Cooperative  Tech.
  Pub. 76 pp. (1989 manual).
Fitzpatrick, T.  J. and M. F. L. Fitzpatrick. 1902. A
  study of the  island flora of the Mississippi River
  near Sabula, Iowa. Plant World 5:198-201.
Forman, R. and M. Godron. 1986. Landscape Ecol-
  ogy. John Wiley & Sons, NY. 620 pp.
Frissell, C. A. and R. K. Nawa. 1992. Incidence and
  causes  of physical failure  of artificial  habitat
  structures  in streams of Western  Oregon  and
  Washington. North American Journal of Fisheries
  Management 12:182-197.
Fuerstenberg, R. R., K. Nelson, and Rob Blomquist.
  1997. Ecological  conditions and limitations to
  salmonid diversity  in the Green River, Washing-
  ton, USA: Structure, function and process in river
  ecology. King County Dept. Natural  Resources,
  Seattle, WA (in preparation).
Garbisch,  E. W., Jr. 1977. Recent  and planned
  marsh establishment  work  throughout the con-
  tiguous United States-a survey and basic guide-
  lines. Contr.  Rep.  D-77-3.  U.S.  Army Eng.
  Waterways Exp. Sta., Vicksburg, MS.
GIS  World, Inc. 1996.  GIS applications in natural
  resources 2.  (Papers from GIS '92-'95 Symposia,
  Vancouver, British Columbia.) GIS World Books,
  Fort Collins, CO.
Good, R. E., D. F.  Whigham, and R. L. Simpson
  (eds.)  1978.  Freshwater  wetlands:  ecological
  processes  and management potential.  Academic
  Press, NY. 378 pp.

New Directions
Gosselink, J. G., E. P. Odum, and R. M. Pope. 1974.
The value of the tidal marsh. Center for Wetland
Resources, Louisiana State University, Baton
Rouge, LA. LSU-SG-74-03. 30 pp.

Greeson, P: E., J. R. Clark, and J. E. Clark. (eds.)
1978. Wetland functions and values, the state of
our understanding. American Water Resources
Association, Technical Publication Series. Min-
neapolis, MN. 674 pp.

Gwin, S. E. and M. E. Kentula. 1990, Evaluating
design and verifying compliance of wetlands cre-
ated under Section 404 of the Clean Water Act in
Oregon. EPN600/3-90/061. U.S. EPA, Environ-
mental Research Laboratory, Corvallis, OR. 122

Hamilton, P. and K. B. Macdonald. 1979. Estuarine
and wetland processes, with emphasis on model-
ing. Marine Science 11, Plenum Press, NY. 653
Hopkinson, C. S. and J. J. Vallino. 1995. The rela-
tionships among man's activities in watersheds
and estuaries: a model of runoff effects on por-
tions of estuarine community metabolism. Estu-
aries 18(4):598-621.

Homer, R. 1992. Constructed wetlands for storm
runoff water quality control, course materials.
Center for Urban Water Resources Management,
Univ. Washington, Seattle, W A.
Hunter, G. J. 1991. Better trout habitat: a guide to
stream restoration and management. Island Press,
Washington, D. C. 320 pp.

Jackson, G. W. 1994. Farmstead assessments-a
voluntary approach for identifying and imple-
menting farmstead practices to prevent pollution.
pp 697-703, In USEPA (1994). Watershed '93 (op
Josephson, J. 1992. Status of wetlands. Environ-
mental Science and Technology 26:422.

Josselyn, M. (ed.). 1982. Wetland restoration and
enhancement in California. California Sea Grant
College Program. Report No. T-CSGCP-007.
University of California, La Jolla, CA. 110 pp.

Jurek, R. M. 1994. A bibliography of feral, stray,
and free-roaming domestic cats in relation to
wildlife conservation. California Dept. Fish and
Game, Nongame Bird and Mammal Program
Report No. 94-5. April 1994. 24 pp.

Kadlec, R. H. and R. L. Knight. 1996. Treatment
Wetlands. Lewis Publishers, New York. 893 pp.

Kaiser, J. 1997. When a habitat is not a home. Sci-
ence 276: 1636-1638.
Kentula, M. E. et al. 1992. An approach to improv-
ing decision making in wetland restoration and
creation. USEP A, Environmental Research Labo-
ratory, Corvallis, OR.

Kentula, M. E. 1997. A step toward a landscape
approach in riparian restoration. (Guest editor)
Special Supplement, Riparian Restoration. Resto-
ration Ecology 5(45): 1-121

Kibby, H. V., J. L. Gallagher, and W. D. Sanville.
1980. Field guide to evaluate net primary produc-
tion of wetlands. Environmental Research Lab.,
USEP A, Corvallis, OR. EP A-600/8-80-037.
55 pp.

King County. 1996. Wetlands and urbanization: im-
plications for the future. Puget Sound Wetlands
and Stormwater Management Research Program,
King County, Seattle, W A.

Kusler, J. and M. Kentula. (eds.) 1990. Wetland
creation and restoration: the status of the science.
Island Press, Washington, D. C. 594 pp.

Kyle, H. M., M. G. Schmidt, and K. P. Lertzman.
1996. A spatially dynamic riparian buffer model:
conservation on the edge. pp. 341-346, In GIS
'35- Ninth Annual Symposium on Geographic In-
formation Systems, Vancouver, British Columbia,
LaFountain, C. 1996. The most for the money.
National Wetlands Newsletter, Environmental
Law Institute, Washington, D. C. 18(3):26-27.

Lauff, G. H. (ed.) 1967. Estuaries. Amer. Assoc.
Advancement of Science, Washington, D. C. Pub.
83. 757 pp.

Lindeman, R. L. 1942. The trophic-dynamic aspect
of ecology. Ecology 23:399-418;

Linduska, J. P. and A. L. Nelson (eds.). 1964.
Waterfowl Tomorrow. U.S. Fish and Wildlife
Service, Government Printing Office, Washing-
ton, D. C. 770 pp.

                                                                                       New Directions
Luoma, J. R.  1998.  Habitat  conservation plans:
  compromise or capitulation? Audubon January-
  February 1998, 36-43.
MacArthur, R. H.,  and E. O. Wilson. 1967. The
  theory of island biogeography. Princeton Univer-
  sity Press, Princeton.

Macdonald, K. B. 1976. The  natural  resources of
  Mugu Lagoon.  Coastal  Wetland Series  No. 17.
  Calif. Dept. Fish and Game/USFWS. Sacramento,
  CA. 119pp.
Malanson, G. P. 1993. Riparian landscapes. Cam-
  bridge University Press, Cambridge, England. 296
Mann, C.  and M. Plummer. 1997. Qualified thumbs
  up  for  habitat plan science. Science 278:2052-
Marble, A. D. 1992. A guide to wetland functional
  design. Lewis Publishers, Boca Raton, FL.
Marcus, M. L. 1982.  Wetlands policy assessment:
  California  case  study.  Prepared  for Office  of
  Technology Assessment, Washington, D. C.  by
  ESA/Madrone Consultants. Novato, CA.
Mason, H. L. 1957. A flora of the marshes of Cali-
  fornia. University of California Press, Berkeley,
  CA. 878 pp.
May, C. W. 1996. Assessment of cumulative effects
  of urbanization on  small streams in the Puget
  Sound Lowland ecoregion: implications for sal-
  monid resource management. Ph.D. Dissertation,
  Univ. Washington, Seattle, WA. 383 pp.
McCormick, J., R. R. Grant, and R. Patrick (eds.)
  1970. Two studies  of Tinicum Marsh, Delaware
  and Philadelphia Counties, PA. The Conservation
  Foundation. Washington, D. C. 180 pp.
McCormick, Taylor & Assoc.  Inc.  1995. Cost sav-
  ings of expedited NEPA/404 approvals: an analy-
  sis of three project in Pennsylvania. Pennsylvania
  Dept. Transportation and  Federal Highway Ad-
  ministration, Washington, D. C. 13 pp.
Minnesota, State of. 1996. Minnesota River water-
  shed conservation reserve  enhancement program.
  Submitted to U.S. Dept. of Agriculture.  State of
  Minnesota (10/25/96). 48 pp.
Mitsch, W. J. and J. G. Gosselink. 1993. Wetlands.
  (Second Edition)  Van  Nostrand Reinhold Co.,
  NY. 722 pp.
Montanari, J. H. and J. H. Kusler (eds.) 1977. Pro-
  ceedings  of  the  National  Wetland  Protection
  Symposium, Reston, VA, June 1977. U.S. Forest
  Service,  Washington,  D.C.   FWS/DBS-78/97.
  255 pp.
National Research Council (NRC). 1995. Wetlands:
  characteristics and  boundaries.  Committee  on
  Characterization of Wetlands, National Academy
  of Sciences, Washington, D. C.
Nesbit, D.  M.  1885. Tide marshes of the United
  States.  USDA  Misc.  Special  Report,  No.  7,
  Washington, D. C. 259 pp.
Nixon, S. W. 1980. Between coastal  marshes and
  coastal  waters-A  review  of twenty  years  of
  speculation  and  research  on  the role of salt-
  marshes  in  estuarine  productivity  and water
  chemistry, pp. 437-525,  In Hamilton, P. and K. B.
  Macdonald, Estuarine  and wetland  processes,
  with emphasis on modeling. Marine Science, Vol.
  11. Plenum Press, NY.
Nixon, S. 1982. The ecology of New England high
  salt  marshes:  a community profile.  Biological
  Services  Program, USFWS, Washington, D. C.
  FWS/OBS-81/55. 71 pp.
Onuf, C.  P. et al. 1979. An analysis of the values of
  Central and Southern California coastal wetlands.
  pp 186-199, In Greeson P.E., et al (eds.) Wetland
  functions and values: the state of our understand-
  ing (op cit).
Payne, N. F. 1992. Techniques for wildlife habitat
  management of wetlands. McGraw-Hill, NY.
Penfound, W. T. 1956. Primary production of vas-
  cular aquatic plants.  Limnology and Oceanogra-
  phy 1:92-101.
PERL (Pacific  Estuarine  Research  Laboratory).
  1990. A manual for assessing restored and natural
  coastal  wetlands  with  examples from southern
  California. Calif. Sea Grant Rep. T-CSGCP-021.
  San Diego State University, CA.
Pitt, R., et al. 1995. Urban stormwater toxic pollut-
  ants: assessment, services, and treatability. Water
  Environment Research 67(3): 260-275.

New Directions
Poff, N. L. et al. 1997. The natural flow regime: a
paradigm for river conservation and restoration.
BioScience 47(11):769-784.

Pomeroy; L. R. 1958. Productivity of algae in salt
marshes.' pp. 88-95 In Ragotzkie, R. A. (ed) 1959,
op cit.
Pulliam, H. R. and B. Babbit. 1997. Science and the
protection of endangered species. Science
Purer, E. A. 1942. Plant ecology of the coastal salt
marshlands of San Diego County, California.
Ecological Monographs 12:82-111.

Race, M. S. 1985. Critique of present wetlands
mitigation policies in the United States based on
an analysis of past restoration projects in San
Francisco Bay. Environ. Management 9:71-82.

Ragotzkie, R. A. (ed). 1959. Salt marsh conference
proceedings. The Marine Institute, University of
Georgia, Sapelo Island, Georgia, March 25-28,
1958. Athens, GA (April 1959). 133 pp.
Recon (Regional Environmental Consultants). 1989.
Comprehensive species management plan for the
least Bell's vireo. Prepared for San Diego Asso-
ciation of Governments by Recon, San Diego,
CA. 226 pp.

Regional Interagency Executive Committee (RIEC).
1995. Ecosystem analysis at the watershed scale:
federal guide for watershed analysis. Revised
August 1995, Version 2.2. Regional Ecosystem
Office, Portland, OR. 26 pp.
Reppert, R. 1992. Fundamentals of wetlands miti-
gation banking. Concept Paper. U.S. Army Corps
of Engineers, Institute for Water Resources. Re-
port 92-WMB-l.

Robinson, A. Y. 1996. A wetland-friendly Farm
Bill? National Wetlands Newsletter, Environ-
mental Law Institute, Washington, D.C.
18(3): 1 ,28-30.

Rohlf, D. J. 1989. The endangered species act: a
guide to its protections and implementation. Stan-
ford Environmental Law Society, Stanford Law
School, Stanford, CA. 207 pp.
Scodari, P.F., et a1. 1995. Prototype information tree
for environmental restoration plan formulation
and cost estimation. USACOE, Inst. for Water
Resources, Alexandria, V A. IWR Report 95-R-3.
58 pp.

Scott, D. A. 1982. Managing wetlands and their
birds: a manual of wetland and waterfowl man-
agement. International Waterfowl Research
Bureau, Slimbridge, Glos., England. 368 pp.
Searchinger, T. D. 1997. Senior Attorney, Environ-
mental Defense Fund, Washington D. C. Personal
communication, October 1997.

Seneca, E. D., W. W. Woodhouse, Jr., and S. W.
Broome. 1975. Saltwater marsh creation, pp 427-
437. In L. E. Cronin (ed.), Estuarine Research
V 01. II. Geology and Engineering, Academic
Press, NY.
Shaler, N. S. 1885. Preliminary report on sea-coast
swamps of the eastern United States. U.S. Geo-
logical Survey, 6th Annual Report. pp. 359-368.

Shapiro and Assoc., Inc. 1978. Snohomish Estuary
wetlands study. Vols. I-IV. USACOE, Seattle
District, Seattle, W A.

Shaw, S. P. and C. G. Fredine. 1956. Wetlands of
the United States, their extent, and their value for
waterfowl and other wildlife. USFWS, Circular
39, Washington, D. C. 67 pp.

Shilling, F. 1997. Do habitat conservation plans
protect endangered species? Science 276: 1662-
Simenstad, C. A., and R. M. Thorn. 1996. Func-
tional equivalency trajectories of the restored
Gog-Le-Hi- Te estuarine wetland. Ecological
Applications 6:38-56.

Smalley, A. E. 1959. The role of two invertebrate
populations Littorina irrorata and Orchelimum
fidicinium in the energy flow of a salt marsh eco-
system. Ph.D. Dissertation, Univ. of Georgia,
. Athens, GA. 126 pp.

Smith, L. M., R. L. Pederson and R. M. Kaminski.
1989. Habitat management for migrating and
wintering waterfowl in North America. Texas
Tech University Press, Lubbock, TX. 560 pp.

Smith, R. D. et a1. 1995. An approach for assessing
wetland functions using hydrogeomorphic classi-
fication, reference wetlands, and functional indi-
ces. Wetlands Research Program Tech. Rpt.
WRP-DE-9. USACOE, Waterways Experiment
Station, Vicksburg, MS.

                                                                                       New Directions
Soule, M. E., et al. 1988. Reconstructed dynamics
  of rapid extinctions of chaparral-requiring birds in
  urban habitat  islands. Conservation Biology 2:75-

Spence, B. C. et al. 1996. An ecosystem approach to
  salmonid conservation. TR-4501-96-6057. Man-
  Tech Environmental Research Services  Corp.,
  Corvallis,   Oregon.  (Available  from  National
  Marine Fisheries  Services, Portland,  Oregon.)
  356 pp.

Strickland, R., (ed.). 1986. Wetland functions, reha-
  bilitation and creation in the Pacific Northwest:
  the  state of  our understanding.  Proceedings  of
  Conference, April 30-May 2,  1986. Port Town-
  send, WA. Pub.  No. 86-14.  Washington State
  Department of Ecology, Olympia, WA.
Tanner, C. D. 1990. Mitigation planning on an estu-
  ary-wide scale:  an  alternative to  case-by-case
  wetland mitigation policy. Masters thesis, Univer-
  sity of Washington. Seattle, WA. 89 pp.
Teal,  J. M. 1958. Energy flow in the salt marsh eco-
  system, pp.  101-107, In Ragotzkie,  R.  A. (ed)
  1959, opcit.
Teal,  J. M. 1962. Energy flow in the salt marsh eco-
  system of Georgia. Ecology 43:614-624.
Terborgh, J. 1989. Where have all the birds gone?
  Princeton University Press, Princeton, NJ. 207 pp.
Thayer,  G. W.  (ed.) 1992. Restoring the nation's
  marine environment. Maryland SeaGrant Book,
  College Park, MD. 716 pp.
Thompson, K.  1961. Riparian forests of the Sacra-
  mento Valley, California. Annals of Assoc. Amer.
  Geographers 51:294-315.
Transeau, E.  N. 1903.  On the geographic distribu-
  tion and ecological relations of the  bog plant so-
  cieties of  northern  North  America.  Botanical
  Gazette 36:401-420.
Turner, L. M. 1936. Ecological studies in the Lower
  Illinois River Valley. Botanical Gazette 97:689-
Turner, R. E. 1977. Intertidal vegetation and com-
  mercial yields of Penaeid  Shrimp.  Trans. Amer.
  Fisheries Society 106:411-416.
U.S. Fish and Wildlife Service. 1970. National es-
  tuary study. Washington, D. C. 3 vols.
U.S. Fish and Wildlife  Service. 1980. California
  least term. FWS/OBS-80/01.20. Biological Serv-
  ices Program, USFWS, Washington, D. C.
U.S. Fish and Wildlife  Service. 1985. Coachella
  Valley fringe-toed lizard Recovery Plan. USFWS,
  Portland, OR.
U.S. Fish and Wildlife Service and National Marine
  Fisheries Services  (NMFS).  1996.  Endangered
  species habitat conservation planning handbook.
  USFWS/NMFS, Washington, D. C.
U.S. Department of Transportation (USDOT). 1997.
  Integrating watershed management and transpor-
  tation planning. National Teleconference Series,
  Program No. TC-12. Center for  Transportation
  and the Environment, North Carolina State Uni-
  versity, Raleigh, NC (11/7/97).
U.S. General Accounting Office.  1994.  Highway
  planning: agencies are attempting to  expedite en-
  vironmental  reviews,  but   barriers   remain.
  USGAO, Washington, D. C. GAO/RCED-94-211.
Ury, W. and R. Fischer. 1981. Getting to yes: nego-
  tiating an agreement without giving in. Houghton
  Mifflin Co., NY. (Penguin Books 1983). 161 pp.
USACOE. 1976. Alsea wetlands review. USACOE,
  Portland District, Portland, OR. 345 pp.
USACOE.  1979.  Wetland Values: concepts  and
  methods for wetlands evaluation. USACOE, Inst.
  For Water Resources,  Research  Rpt. 79-RI, Ft.
  Belvoir, VA. 109 pp.
USACOE. 1987. USACOE Wetlands  Delineation
  Manual. Environmental Laboratory, USACOE,
  Waterways Experiment Station, Vicksburg, MS.
  Tech. Rpt. 4-87-1. (1987 Manual).
USEPA. 1993.  Constructed wetlands for wastewater
  treatment and wildlife habitat: USEPA, Office of
  Wastewater Management, Municipal Technology
  Branch, Washington, D. C.  EPA 832-R-93-005.
USEPA.  1993. Guidance specifying  management
  measures for sources  of  nonpoint  pollution in
  coastal  waters.  Report 840-B-92-002.  USEPA,
  Office of Water, Washington, D. C.
USEPA.  1993. Partnerships  and opportunities in
  wetland  restoration.  Workshop   Proceedings,

New Directions
Seattle, WA. April 1992. EPA 9IO/R-94-003.
USEP A, Seattle, W A.

USEPA. 1994. Watershed '93: a national confer-
ence or watershed management. EPA 840/R-94-
002 USEPA, Washington, D. C.
Vanbianchi, R. et al. 1994. A citizen's guide to
wetland restoration. Approaches to restoring
vegetation communities and wildlife habitat
structure in freshwater wetland systems. Publica-
tion 91O/R-94-006. USEPA, Region 10, Seattle,
WA. 71 pp.

Warner, R. E. and K. M. Hendrix. 1984. California
riparian systems: ecology, conservation and pro-
ductive management. University of California
Press, Berkeley, CA. 1035 pp.

Washington State Dept. of Ecology. 1992. Storm-
water management manual for the Puget Sound
Basin. Pub. #91-75. Washington State Dept. of
Ecology, Olympia, W A.

Wear, D. N., M. G. Turner, and R. O. Flamm. 1996.
Ecosystem management with multiple owners:
landscape dynamics in a southern Appalachian
watershed. Ecological Applications 6(4): 1173-
Weinmann, F. and K. Kunz. 1993. Wetlands: infu-
sion of restoration into the management formula.
pp 1-17, In USEP A, Partnership and opportunities
in wetland restoration. Workshop Proceedings,
Seattle, WA. April 1992. EPA 310/R-94-003.
USEP A, Seattle, W A.
Western Forestry and Conservation Association
(WFCA). 1997. Watershed analysis and the Total
Maximum Daily Load allocation process. WFCA,
Portland, OR.

Wilson, E. 0., 1994. Naturalist. Island Press,
Shearwater Books, Washington, D. C. (Warner
Books edition 1995). 380 pp.
Wren, C. D. et al. 1997. Wildlife and contaminants
in constructed wetlands and stormwater ponds.
Environment Canada. Tech. Rpt. Series No. 269.
Canadian Wildlife Service, Ontario Region. 109

Wright, J. O. 1907. Swamp and overflow lands in
the United States. U.S. Dept. of Agriculture Cir-
cular No. 76. Washington, D.C.

Yapp, R. H. et al. 1917. The salt marshes of the
Dovey Estuary. Journal of Ecology 5:65-103.

Zedler, J. B. 1982. The ecology of Southern Cali-
fornia coastal salt marshes: a community profile.
Biological Services Program, USFWS, Washing-
ton, D. C. FWS/OBS-81154. III pp.
Zedler, J. B: 1984. Salt marsh restoration: A guide-
book for Southern California. California Sea
Grant College Progt:am. Report No. 7-CSGCP-
009. La Jolla, California.

Zedler, J. B. 1988. Salt marsh restoration: lessons
from California, pp 123-238. In J. Cairns (ed.),
Management for rehabilitation and enhancement
of ecosystems. CRC Press, Boca Raton, FL.

Zedler, J. B. 1996. Coastal mitigation in southern
California: the need for a regional restoration
strategy. Ecological Applications 6:84-93.
Zedler, J. B. 1996a. Tidal wetland restoration: a sci-
entific perspective and Southern California focus.
Calif. Sea Grant College System. Univ. of Cali-
fornia, La Jolla, CA. Report No. T-038. 129 pp.

Zedler, J. B. (ed.) 1996b. Forum: Wetland Mitiga-
tion. Ecological Applications 6(1):33-101.

Taking A Broader

"You can engineer these
(wetland) systems in the best
way possible-but if you put them
in the wrong place they're not
going to work/'
                     Klaus Richter,
                     Wetland Ecologist

                                    An Integrated Wetland Management
                                   Program for the Snohomish Estuary
                                                                       Stephen Stanley1
                                                                           Paul Roberts1
 Abstract:  The  Snohomish  Estuary  Wetland
Integration  Plan (SEWIP)  is an  ecologically
based wetlands management plan, which pro-
vides the basis  for  a coordinated regulatory
program for the protection, restoration, and en-
hancement  of  the Snohomish  Estuary.  The
12,000-acre Snohomish Estuary is located on the
Puget Sound 25  miles north of Seattle in Sno-
homish County (Washington). The program is
an adaptive management process involving tech-
nical experts from the Washington Departments
of Ecology and Fish and Wildlife, in addition to
the Corps of Engineers, Environmental Protec-
tion Agency, and  local jurisdictions including
the City of Everett, the Tulalip Tribe and Sno-
homish County. The 12,000 acres of the estuary
were inventoried and assessed using the Indica-
tor Value Assessment (IVA) protocol2. The IVA
was designed to measure the performance of  16
wetland functions  (e.g., anadromous fish, over-
wintering birds, mammals, sediment retention,
etc.) based on a  series of "indicator questions"
developed  by a  team of  wetland scientists
(ecologists, wildlife biologists, fisheries biolog-
ists) and hydrologists familiar with the estuary.
For planning purposes, it was assumed that the
entire study area was wetland (not jurisdiction-
ally delineated), and it  was mapped as 367
individual wetland complexes on the  basis  of
common  hydrological  boundaries.  Wetland
complexes were categorized as Group 1 (highest
ranked), Group 2 and Group 3 (lowest ranked)
based on assessment  scores  and their  relative
rankings. Analysis of the assessment data indi-
cated that 86 percent of the  wetland complexes
within the estuary have a relatively high level of
performance of the  Wildlife, Fisheries  and
Water Quality Improvement Functions. It was
concluded that the estuary is functioning as  an
integral ecosystem and has a high potential for
successful restoration. Restoration  sites have
been identified with a priority given to tidal res-
toration. The conceptual  management plan
ensures the continuation of agriculture, primar-
ily in one of the  seven Ecological Management
Units (1.0), with  restoration efforts focusing on
non-regulatory incentive measures within this
unit,  including leasing of edges of  fields  for
"habitat crops" in order to provide wildlife cor-
ridors  between  larger  habitat   patches.
Ecological Management  Unit  2.1, an area of
wetlands with low  performance  of assessed
wetland functions, is being considered for poten-
tial future development. Phase II of the plan will
include preparation of an alternatives analysis,
compensation manual, and environmental  as-

 The  "Snohomish  Estuary Wetland  Integration
Plan," (SEWIP) was prepared for 11,623  acres of
the Snohomish Estuary located on the Puget Sound
in the west-central portion of Snohomish County of
Washington State (Figure 1). The plan was com-
pleted  in April  1995 and was funded  by the
Department of Ecology and the Environmental
Protection Agency (EPA) as part of the State Wet-
land Integration Strategy (SWIS).
 The plan consists of  a wetland management  pro-
gram which provides for the protection and restora-
tion  of the most important wetland  ecosystems
within the estuary  and a  wetland compensation
process designed to streamline the local, state and
federal wetland permitting process. When an alter-
natives analysis for the plan is completed in 1996,
the City and County will apply to the  U.S. Army
Corps  of Engineers (Corps of  Engineers) for a
'City of Everett Planning and Community Development, 2930 Weetmore Ave. Suite 100, Everett, Washington 98201-4044.
2Editor's Note-The IVA protocol is further explained in Thomas Hruby's paper this volume.

SEWIP Study Area

                                                                                       Snohomish Estuary
"blanket permit" through either a Letter of Permis-
sion (LOP) or Programmatic Permit. This will
allow both local governments to issue permits for
some types of wetland alteration (including en-
hancement and restoration) within wetland com-
plexes with a low ranking. Alterations would only
be permitted if conducted within the conditions of
the overall permit issued by the Corps of Engineers.
This process should provide for a substantial sav-
ings in time and money relative to the current
permitting process.

  In order to develop a regional wetland manage-
ment plan that is scientifically based it is necessary
to: 1) conduct an inventory of the number and type
of wetland resources; and 2) conduct an assessment
of the ecological functions of those wetlands. The
methods  for conducting  wetland inventories  are
relatively well established because they are based
on plant species, soil types and the hydrological re-
gime present. Numerous  comprehensive resources
are available which list and  identify wetland plants
and soils, and assist in the identification of different
hydrological  regimes.  These resources  include
national and  regional  plant  lists  which  identify
plants by species and  their probability of occur-
rence in wetlands; soil surveys with maps from the
Natural Resource Conservation  Commission (for-
merly Soil  Conservation  Service) identifying soil
makeup and whether it is hydric; and the 1987 and
1989 Federal Interagency Manuals for Identifying
and Delineating Jurisdictional Wetlands.
  Assessing the performance of wetland functions is
a much more difficult  undertaking. Functions are
defined as  the  physical,  chemical and  biological
processes or attributes  that  contribute to the  self-
maintenance of wetland ecosystems.  To assign  a
"quantifiable" numeric  value to wetland functions
for a specific wetland  requires an intricate knowl-
edge  of the interaction of the physical, chemical
and biological components within that wetland's
ecosystem. Typically this requires years of study to
identify the habitat requirements and life forms of
species within  the ecosystem and to identify rela-
tionships    between   the   species   and  their
 Ideally, with this  type of in-depth knowledge a
wetland ecologist would have to only identify the
type of habitat (including basic vegetation struc-
ture, hydrology, landscape features and  connec-
tions, and buffer type) in order to know how  well
wetland functions are being  performed. Yet, this
type of detailed knowledge  of the "quantitative"
performance of wetland functions  is more than a
decade away (Hruby et al.  1994). Further, the ex-
isting  habitat assessment  models,  such  as  the
Wetland Evaluation Technique (WET), the  U.S.
Fish and Wildlife Service's Habitat Evaluation Pro-
cedures (HEP) and the Reppert method (Reppert et
al. 1979) do not specifically measure the perform-
ance of functions, or allow for the  comparison of
wetland performance and value between individual
 Nevertheless,  there  is sufficient  knowledge re-
garding performance  of a  wetland function  at a
general level when certain basic indicators  are pre-
sent. For example, a wetland with a restricted outlet
performs the "water quality improvement" function
better than a wetland with an  unrestricted outlet; a
tidal wetland with dendritic channels performs the
"fishery" function better than  a tidal wetland with-
out such channels.  The SEWIP study selected a
habitat  evaluation method which   measures  per-
formance  at this more general  "qualitative" level
and  ranks  wetlands  in  a  "semi-quantitative"
 In January of 1994, a Technical Advisory Com-
mittee was formed to assist in the development of
the Inventory and Assessment methodology for the
SEWIP effort. The committee consisted of wetland
and  resource scientists from the Corps of Engi-
neers,  EPA, Department  of Ecology, Washington
Department of Fisheries and  Wildlife, Snohomish
County, City of Everett, and local wetland consult-
ing firms.

Inventory of Wetlands
 The initial task of the Technical  Advisory Com-
mittee was to determine how the estuary wetlands
would be mapped and inventoried.
  1. Mapping Units. The committee first dealt with
whether the wetlands should  be mapped according
to the delineation methodology of the 1987 Federal

Snohomish Estuary
 Interagency Manual on Wetland Delineation. It was
 determined that it would be too time consuming to
 field delineate and map wetlands in this manner.
 Though large portions of the estuary are diked and
 drained for farming,  the majority of the soils are
 still hydric, and are saturated during the winter and
 early spring months. Because  the majority of the
 diked portions of the  estuary cannot be drained due
 to high river flows (e.g., in adjacent sloughs) prior
 to March, hydric soils are more than likely satu-
 rated  to  the  surface  for  the  requisite  one week
 period during the growing season.
  In light of this  information, the Committee deter-
 mined that all areas within the estuary should  be
 mapped (unless all three indicators were not pres-
 ent-plants, soils and hydrology and the areas was
 clearly upland) on the basis of common hydrology.
 Each hydrological unit would be called a "wetland
 complex" and no attempt  would be made to deter-
 mine  actual  wetland  boundaries  within  the
  2. Wetland Complex Inventories. Plant  invento-
 ries for each complex were conducted with the esti-
 mated percent of dominant plant species for each
 class (emergent,  scrub-shrub, and forested class)
 recorded, in addition to buffer condition, and wild-
 life observations.
  3. GIS  Mapping. Wetland complexes were base-
 mapped on computer generated copies of Corps of
 Engineers  color-infrared  aerial photos  (8/18/93,
 scale 1:2000) according to common hydrological
 boundaries. Black and white aerial photos (2/24/91,
 scale  1:600) were also used to assist in the mapping
 of units.  The COE  color  infrared  aerials were
 scanned directly into the  City's Geographic Infor-
 mation System and subsequently printed out to use
 for mapping in the field. This allowed for digitiza-
 tion of wetland complex boundaries directly off of
 the field maps.

  4.  Habitat Classification System for Field Map-
 ping. Field teams used a modified Cowardin  Sys-
 tem (Cowardin et al.  1979) to  describe  wetland
 habitat within the estuary. The estuarine intertidal
 and palustrine systems were primarily used with the
 classes being  limited to  unconsolidated  shore,
 aquatic bed, emergent, scrub-shrub, and forested.
Dominant  classes  were   based  on  a  minimum
30 percent areal coverage of the wetland complex
being inventoried and assessed. Secondary classes
were based on a minimum 10 percent areal cover-
age of the wetland complex.

Habitat Assessment Model
  In developing the habitat assessment   model it
was necessary to first identify those functions that
were ecologically critical to the functioning of the
estuary  and  socially important to humans.  The
Technical Advisory Committee selected  17 func-
tions in the estuary (Table 1).

Table 1. Functions for vegetated wetlands
(+6 MLLW)1
Social Significance
Water Quality
Export of Primary

Fish Habitat
Bird Habitat
Stabilization of Shoreline and
Access to Transportation
Priority Species Habitat
Aesthetic Value

Sediment Stabilization
Sediment Toxicant Retention
Nutrient Retention and
(includes rearing, feeding,
migration, and shallow water
refuge functions)

Other Species Habitat  Herptiles
1The functions selected for Sand/Mudflat and Subtidal
 Wetlands (-15 to +6 MLLW) where similar to those for
 vegetated wetlands except that herptiles, mammals,
 and primary productivity were dropped and epibenthic
 and infaunal invertebrates were added.
  In  reviewing several  habitat assessment models,
including the Oregon Freshwater Assessment and

                                                                                        Snohomish Estuary
Wetland  Evaluation Technique  (WET), it was
determined  that  the Indicator  Value Assessment
(IVA) protocol would be used. The primary advan-
tage of this model relative to others, is that it can be
designed specifically for the wetlands being evalu-
ated  and  it  allows comparison  of  functional
performance of wetlands within a defined regional

  The IVA model is based on the assumption that a
wetland with  a greater number of indicators for a
particular function  is performing that function bet-
ter than another study area wetland with fewer indi-
cators present. Numeric scores are assigned to  all
wetlands  within the study area on the basis  of the
number of indicators present for each wetland, and
the  wetlands  are ranked according to their score
from 1 to 100.
  To determine the number and type of indicators
present within an evaluated wetland, the committee
devised a series of "indicator questions."   These
questions, when  answered  in  the field, "indicate"
the  presence  of biological, chemical and physical
conditions that are  important to the functioning of a
wetland ecosystem. The model also includes ques-
tions pertaining  to "social  functions" that are im-
portant to  humans  such  as  recreation  (includes
shoreline access), navigable waterways for coastal-
dependent industry, hunting, fishing, boating, and
wildlife observation.
  Based on their scientific knowledge  of  coastal
wetlands  within Washington and of the estuary, the
Technical Advisory Committee assigned individual
numeric values to  each function of every indicator
question.  The assigned values were assigned de-
pending on whether the indicator question was:
very strongly associated (3);  strongly  associated
(2); or weakly associated (1) with the function be-
ing  assessed.  A multiplier  or fraction was applied
when an  indicator resulted in  significantly higher
values (e.g., the presence of dendritic channels  in-
dicating greatly improved fish habitat), or negative
values (e.g., presence of cattle grazing).
  The habitat  questions  are separated into catego-
ries of wetland "function" such as Fish, Birds and
Other Species,  Primary Productivity,  and  Water
Quality Improvement.  The model  also contains
questions which  subtract from the overall habitat
value if negative features are present on the wet-
land complex being evaluated,  such as farming,
drainage ditches, and filling activities. This evalua-
tion model allows each parcel (as defined by a com-
mon hydrological boundary, such as ditches, dikes
or sloughs)  to be assigned  a  numeric value and,
therefore,  to be compared to all other wetland com-
plexes  within the  estuary.  This  will  allow  the
regulatory agencies to determine where the highest
and lowest performance wetlands  are located  and
provide a basis for offsite mitigation planning.
  Two teams of biologists and resource specialists
(wetland ecologist, fishery biologist, wildlife biolo-
gist, water quality scientist) spent five months from
May 1994 to September 1994 mapping and assess-
ing 367 wetland complexes within the study area. A
List of Assumptions for applying the model was re-
viewed by a subcommittee of the technical advisory
committee and any necessary changes made. A rig-
orous review of all data for consistency with the
List of Assumptions  was then  conducted  by the
team members over a two-month period.

  Analysis of the  assessment  data  indicate  that
86 percent of the wetland complexes within the
estuary have a relatively high level of performance
of  the  Wildlife  and Water Quality  Improvement
Functions. As such, the estuary is functioning as an
integral ecosystem and has high potential for  suc-
cessful  restoration  given  the  presence of hydric
soils, hydrology, and ample adjacent "pristine" sites
which could serve as  sources for plant and animal
  Composite scores for the Water Quality Improve-
ment and Wildlife Attributes  for 367 wetlands
within  the  Snohomish Estuary are presented in
Figure 2.  Based  on  the  individual  frequency
distributions of data for each  attribute, wetland
complexes were separated into "Group  1, 2, or 3"
  The  individual  functions for the Social Signifi-
cance functions  were not  combined for  a  final
overall categorization because  of the  dissimilar
nature  of "transportation  access"  and "recreation/
aesthetics" functions.
  Definitions for the  three wetland groups are as
follows: Group 1 - Wetland complex is very

SEWIP Wetland
Functional Performanc
for Water Quality
and Wildlife
SewerPonds  Marysyille
                 wr Lbtr f E~ RANKT NQ-
                   Mu d ( I at
                         I :  40-100
                         Z:  24-38
                         3:  0-2]

                         1 :  27- 1 00
                         2:  8-28
                         3:  0-7
                                                       WATE H~ QDAt F T?-RAKlltrKlo'
                                                               1 :  36-100
                                                               2:  26-35
                                                               3:  0- 25
 Wetland Groupings

 Veg. Units       Mudflat Units
                                                             571174~Ac r «e
                                                             2849.1 ADI ft
                                                             1490.9 Ac I ft
                                                             1 3797 VAcre«
                                                               73.2 Ac r•«
                                                              1 2 8 . S Ac r « B
                                Figure 2.

                                                                                     Snohomish Estuary
strongly associated with the performance of estuary
functions; Group 2 - Wetland complex is  strongly
associated with the performance of estuary func-
tions; and  Group 3 - Wetland complex is weakly
associated with performance of  estuary functions
but may have a high potential for restoration.

  Wetland rankings should not be compared be-
tween mudflat (non-vegetated) wetland complexes
and vegetated wetland complexes due to the differ-
ent indicator assessment models  used for  each
wetland type (Table 1).

  A total of  11,632  acres  of non-vegetated and
vegetated wetland were classified. A breakdown of
the wetland categorization by  acreage  and  the
ranking scheme is included in Figure 2. In  general,
vegetated wetland complexes with ranking scores
greater than 26 for the Wildlife function and 34 for
the Water  Quality Improvement  functions were
placed in Group 1;  those with scores from 8 to  26
for Wildlife  and 22 to 34  for  Water  Quality
Improvement, were placed in  Group 2; and those
with scores less than 8 for Wildlife and less than  22
for Water  Quality  Improvement, were placed  in
Group 3.
  Over  60 percent  (7090 acres) of  the estuary's
wetland complexes  were  classified within Group 1
for both the Water Quality Improvement and Wild-
life Attributes.  This  area  represents  an  almost
continuous  band  of Group  1  wetland complexes
stretching from the western mudflats along Ebey
and Steamboat Sloughs,  through  South  Spencer
Island and into the north portion (south of Highway
2) of South Ebey Island and across to Diking Dis-
trict 6 (Figure 2).
  Edge units along the north portion of South Ebey
Island appear to play  an important role  in connect-
ing Group  1 wetland  complexes to the north with
similar areas south of Highway  2  on South Ebey
  Group 2 Wetland Complexes (2922 acres) com-
prise  over 25 percent  of estuary study area and are
primarily in the central portion of the estuary adja-
cent  to  Group   1  wetlands.  Group  3  Wetland
Complexes represent  only 14 percent (1619 acres)
of the estuary study area and are entirely limited to
heavily cultivated,  grazed  or  mowed agricultural
 Both the presence of a significant area of Group 1
wetland complexes throughout the estuary and con-
centration of  Group 2 wetland complexes adjacent
to  Group 1 wetland complexes, indicates that  the
estuary is functioning at a relatively high level of
performance for the functions assessed and is still
an intact estuary ecosystem that has a high potential
for enhancement and restoration. Group 3 wetlands
have a high potential for restoration because  the
existing agricultural practices have not permanently
degraded the former estuarine habitat and  do  not
prevent the return of these lands to full tidal

Discussion - Management Plan
 In developing a  management plan the technical
committee addressed the following seven tasks in
sequence:  1) identify and rank the  restoration  and
enhancement goals; 2)  identify restoration  and
assessment sites; 3) identify specific restoration
actions for each site, based on site visits by techni-
cal committee; 4) rank restoration and enhancement
sites based on overall estuary goals, technical,  and
social    feasibility;    5)   calculate   potential
"compensation  credits"  for  restoration and  en-
hancement  sites by re-scoring these sites with IVA
model based on identified restoration  and enhance-
ment  actions (i.e., #3);  6)  identify  development
footprint based on location of lowest performance
wetlands for habitat functions and highest perform-
ance for transportation and water access functions,
and determine if "impact debits" generated can be
offset by "compensation credits"; and 7) develop-
ment  compensation mechanism for "development
footprint" impacts, including identification of com-
pensation ratios.

Identification of Restoration and
Enhancement Goals (Step 1)
 The primary impact to the approximately 12,000
acres  of wetlands within the Snohomish Estuary
has been due to diking and draining for conversion
to agricultural production. This has resulted in an
estimated 74.4 percent (Pentec 1992)  to 85  percent
(Shapiro and Associates 1979) loss of the original
estuarine wetland area.  Estuarine wetlands are an

 Snohomish Estuary
 extremely  important component in the marine and
 terrestrial food chain, providing critical habitat for
 fish, birds, and  other  wildlife including  species
 which are commercially important. Studies in the
 Pacific Northwest have demonstrated that estuarine
 marshes are more productive than other plant com-
 munities. Frenkel measured  the productivity of a
 diked pasture at 1200 grams/meter relative to 2300
 grams/meter2 for the same pasture 10 years after it
 was restored as  estuarine habitat (Frenkel  1990).
 The loss of  estuarine  habitat statewide has  also
 been significant.
   Based on this information the technical committee
 established that  one of the principal goals of the
 management plan is to restore diked palustrine wet-
 lands to tidal influence in order to provide fish
 habitat for the full range of estuary fish  species,
 with emphasis placed  on overwintering and fresh-
 water/saltwater   transition  habitat   for  juvenile
 salmonids. Included within this goal was the resto-
 ration of shoreline migration corridors for juvenile
 salmonids and other fish species.
   The second principal restoration and enhancement
 goal was enhancing  agricultural lands for bird,
 herptile, and mammal habitat, and as wildlife corri-
 dors. This goal was based on the input of farmers
 and other users of the estuary who wished to main-
 tain and protect agriculture  and existing forested
 and scrub-shrub palustrine habitat on the 4000-acre
 South Ebey Island.

   On a regional basis the Snohomish  Estuary is an
 important  habitat for overwintering and migrating
 waterfowl and shorebirds. It provides an alternative
 refuge site for waterbirds during the  winter to es-
 cape severe weather present at estuaries in Northern
 Puget Sound and the Fraser Estuary in Canada. For
 example, ducks can move 15  to 110 km south from
 the Fraser River  Delta to numerous bays and estu-
 aries in Puget Sound (Shreffler & Thorn 1993).

  Agricultural fields are very important to a large
 number of waterbirds. Brennan reports that shore-
 birds and waterfowl feeding  intertidally typically
 roost in adjacent fields at high tide or during storms
 (Brennan et al. 1985). Shreffler and Thorn state that
 large intertidal areas  without adjacent farmland
 usually  support  fewer ducks  and do not provide
complete  wintering habitat  (Shreffler &  Thorn
 1993). Surface feeding ducks  switch to feeding in
freshwater wetlands and agricultural fields in the
winter because eelgrass (Zostera spp.) is not avail-
able  in   intertidal  areas.  Therefore,  this  plan
specifies  the  enhancement  of agricultural  fields
through a combination of flooding and planting off-
season crops in heavily cultivated fields, for forage.

Identification of Restoration and
Enhancement Sites and Actions
and Ranking of Sites (Steps 2, 3,
and 4)
 The central criteria for selecting appropriate res-
toration and enhancement sites and actions was that
they must result in a stable, persistent, resistant and
resilient  ecosystem.  A stable  estuarine  ecosystem
does not readily transform over the long term from
one system, such as an estuarine mudflat, to another
such as a salt marsh, except when it has been sig-
nificantly impacted by human or naturally induced
perturbations  such as  subsidence or excess  sedi-
mentation. A persistent ecosystem is one such as a
forested  tidal swamp,  which maintains  its habitat
structure over a long period of time. An emergent
Carex marsh would be considered to be  resistant
and resilient  if it could readily recover from distur-
bances such as flooding and  erosion which removed
large areas of vegetation.
 The structure of the restored/enhanced estuary  is
critical in obtaining these characteristics of stabil-
ity, persistence,  resistance,  and  resilience.  Struc-
ture consists   of "patches"  of habitat, their shape
and location  in the estuary, and the corridors  con-
necting these patches. Without  all of these ele-
ments, restoration and enhancement efforts have
little chance of success (Shreffler & Thorn 1993).
 In  general,  tidal marshes  of at least 250 to 300
acres appear to be stable and functional ecosystems
(Collins et al. 1987). Patches of larger size tend  to
support more species than small patches. (Shreffler
&  Thorn  1993).  Patches should be connected in a
manner that promotes export of primary productiv-
ity, movement of fish and animals, and a  high
diversity  of   "non-opportunistic" native species.
Shreffler  and Thorn concluded that  "unconnected
habitats are not preferable  and that restoration of
systems supporting large and small habitats, with

                                                                                      Snohomish Estuary
maximum connectance is a more appropriate resto-
ration concept."

  Shreffler  and Thorn (1993) also  state that  "the
chances of successful restoration are higher if the
landscape is relatively intact even  though the de-
gree of disturbance of the restoration site is high."
The habitat assessment demonstrates that the core
of the estuary is intact, based on high functional
performance scores. The potential restoration sites
adjoining these areas  have  only  been temporarily
disrupted by agriculture.

  To  implement  these landscape  principles, the
Management plan focused on restoring large areas
of habitat that are adjacent to the "core" of Group 1
wetlands that are  relatively undisturbed. These tidal
restoration areas, selected and ranked by the tech-
nical committee, are depicted in  Figure 3 and
include the Poortinga Property on  the east main-
land, North Spencer Island,  West Smith Island and
Diking District 6. Enhancement areas within diked
palustrine (agricultural areas) wetlands were con-
centrated primarily on  South  Ebey Island and
Marshland area (not shown on Figure 3).

  In addition to the broader  landscape principals in
selecting restoration and enhancement  sites, the
technical committee used specific criteria based on
the TV A functional performance scores, elevational
data and the presence of any fill or  toxic materials.
For example, any area that scored in Group I for the
Overwintering Birds  or Herptiles  function  was
ranked lower for tidal restoration.

  The technical committee also discussed the effect
that subsidence in the estuary would have upon the
success of  tidal  restoration. Considerable  subsi-
dence appears to have occurred in the estuary as the
result of diking, draining and farming  activities.
Elevation is a primary control on most tidal marsh
processes because it determines the frequency and
period  of   daily  and  seasonal  tidal  inundation
(Frenkel 1990). SEWIP elevation data and vegeta-
tion surveys indicate  that mudflats are generally
found below +6  feet  MLLW; emergent wetlands
generally above +7 feet MLLW; and scrub-shrub,
forested  marshes above +12 feet MLLW. There-
fore, the lower limit for vegetated tidal marshes in
the estuary  appears to approximately be +7 to +8
MLLW.  After considerable debate, the technical
committee determined that the presence of substan-
tial areas of diked palustrine wetland at +8 MLLW,
should not decrease its feasibility for restoration to
tidal habitat  because:  1)  the management  plan is
designed to  ensure the long term protection of a
tidal restoration site; and 2) physical, chemical and
biological estuarine processes will eventually lead
to the establishment of an intertidal marsh and for-
ested  swamp  with  high   performance  of all

  Tidal restoration actions included: the removal of
the maximum practicable length of dike for tidal
restoration; re-connect old sloughs and tidal creeks
with the main tidal channel by overexcavating the
old sloughs  and tidal creeks where they cross the
former  dike location. Frenkel and Morlan found
that  sediment accretion was highest in those areas
that  had  the  greatest  degree of  tidal  exchange
(Frenkel 1990). This typically occurred around tidal
creeks.  The  least amount of  sediment increase
occurred in  areas  which still had dikes in place.
Frenkel and  Morlan also found that tidal creeks
were a  major determinant in  marsh  hydrology
(Frenkel 1990). Removal of large portions of the
dike allows for the accumulation of large woody
debris which will increase wetland elevation in se-
verely subsided areas and allow for the formation
of a more diverse wetland plant community. Other
restoration actions identified by  the committee in-
cluded creating island habitat out of existing dikes.
Enhancement actions for diked palustrine wetlands
included increasing flooding in  agricultural fields
in order to improve bird and amphibian habitat and
creating scrub-shrub corridors along ditches at the
edge of agricultural fields, particularly between the
forested habitat on South  Ebey and  along  the
"water-edge" units.

Identification of Development
Footprint and  Development of a
Compensation Mechanism
(Steps 5, 6, and 7)
  One of the purposes of the SEWIP document is to
determine if there  is an adequate level of compen-
sation credits to offset the impacts from the filling
of wetlands with low functional performance. The
technical and user committees identified a 352-acre
development footprint (Figure 3) within an  existing
area of industrial development with Group III

Footprint & Tidal
                        > // Swamp
       Mudflat Restoration
       Remove Log Rafts
       Mudflat Units
       Development Footprint
       Vegetated Units
       Development Footprint

                                                                                       Snohomish Estuary
wetlands (low functional performance). This area
scored high for the Access to Transportation Corri-
dors  function  given the  immediate presence  of
dredged tidal waterways, rail lines and two major
freeways. The wetland  impacts  for this 352-acre
area were 10,761-acre points (IVA ranking score x
the acreage of the wetland complex) for the Water
Quality function  and  4,500-acre  points  for the
Wildlife function (see Table 2).

  Based on the "restoration/enhancement actions"
identified for the estuary  restoration/enhancement
sites, the IVA model was run again for the affected
restoration  sites and the new "increased" scores
tabulated. From these new  scores or "compensation
credits" it could then be determined if they  were
adequate   to   offset  the  impacts   from   the
"development footprint" located within the area of
Group  El wetlands. The compensation  scores for
the Priority  1 restoration site (Poortinga Property)
are displayed in Table 2. Table 2 demonstrates that
the  Priority #1  Restoration  site has  more  than
enough compensation credits to offset the total im-
pacts from the development footprint for vegetated
  With  adequate compensation credits  established
for the development footprint, the technical  com-
mittee then considered the appropriate compensa-
tion  ratio to apply. By  examining the  average
increase for the Water  Quality  Improvement and
Wildlife Functions for  each restoration and en-
hancement site and comparing that increase to the
impact debits for the corresponding functions in the
Development Footprint, the overall compensation
ratio was determined.
  Table 3 shows that the net increase in function for
the first four restoration sites usually exceeds the
existing level of function  performance. However,
the Water Quality Improvement Function does pro-
vide  less  increase in function per  acre than the
Wildlife Function, so the Compensation Ratios pre-
sented in Table 3 are based on the average loss or
gain in the Water  Quality Improvement Function.
Table 3 shows that the average wetland impacts in
the development footprint relative to  the average
wetland restoration gains will result in  a less than
1:1 acreage  replacement ratio for priority  site  1.
However,  because  there  are  policies  among
resource  agencies  requiring  that  compensation
ratios should be at  least a  1:1 ratio on  a acreage
basis, a 1:1 compensation ratio was selected for the
priority 1 restoration site.
 Priority sites 2, 3,  and 4 demonstrated an progres-
sive increase in the compensation ratio to replace
loss of Water Quality Improvement functions. This
is  due to the decreasing wetland  functional  per-
formance at these restoration sites based on IVA
model results. The compensation  ratio  based on
acreage (column 6) was increased by approximately
25 percent   above  the  functional   loss   ratio
(column 5) to account for the lower priority of the
site and to encourage  restoration of the higher
ranked site.

 Though not yet adopted by local governments or
state  and  federal agencies,  the SEWIP  document
has already proved valuable to local planning agen-
cies, the Port District, and local industries. Prior to
the management plan, compensation within the es-
tuary  was   often  improperly   sited  and   not
appropriately designed.  Purchase of compensation
sites took place without any overview to determine
what the benefit of  the compensation site would be
to  the  overall estuary.  This resulted in  isolated
compensation sites that were much more expensive
to engineer and construct than if a larger, better lo-
cated,  compensation   area  was  purchased   and
restored. The  assessment of the estuary's wetland
resources on a landscape basis has encouraged re-
source  agencies to  coordinate efforts to purchase
higher ranked restoration sites and undertake the
restoration actions identified in the plan which are
designed to provide the greatest ecological benefit
to the estuary.
 More  importantly,  the   SEWEP  process  has
brought together local  governments  and resource
agencies that  normally  have conflicting mandates
for the estuary and forced them to plan  for the fu-
ture of the estuary from a ecological perspective,
rather than  from a fragmented  "market demand"
 The FVA method  proved  to be a relatively accu-
rate and rapid assessment tool for establishing the
relative value of wetlands within the estuary  for
the  purposes  of developing a detailed  wetland

 Snohomish Estuary
   Table 2. Summary of development footprint impacts and restoration credits for Poortinga site in Snohomish Estuary
Impacts to Water
Quality Functions
  in Acre Points
                                                 Restoration Credits
                                                  for Water Quality
                                                  Functions in Acre
Impacts to Wildlife
Functions in Acre
Restoration Credits
 for Water Quality
 Functions in Acre
Development Footprint for
Vegetated Wetlands
Priority #1 Restoration
Site, Poortinga Property
(350 acres)
   Table 3. Calculated compensation ratios for Snohomish Estuary - vegetated wetlands
Priority and
Priority 1 -
Priority 1 -
Priority 2 - N.
Spencer (Beringer)
Priority 2 - N.
Spencer (Beringer )
Priority 3 - E. Smith
Priority 3 - E. Smith
Priority 4 & 4A -
East Mainland
Priority 4 & 4A -
East Mainland
of Functions
(per acre)
Increase in
Area, (per
Ratio to offset
function loss
& temporal
loss (i.e. 1.25
x functional
loss ratio)
Ratio Based on
Priority &


Reasons for
Using less than
1:1 would not be
consistent with
National Policies
for no net loss.

The higher ratio
creates an
incentive for
restoration of
Priority site 1 first

The higher ratio
creates an
incentive for
restoration of
Priority site 1 and
2 first

The higher ratio
creates an
incentive for
restoration of
Priority site 1,2
and 3 first


management  plan.  More rigorous  "quantitative"
models based on years of qualitative observations
would not have allowed this regional  planning
effort to  proceed  in  a timely fashion,  thereby
allowing  development  to  further  fragment and
possibly    degrade    the    overall   ecosystem

  The development  of the SEWIP document would
not have been possible without an unusual level of
support and cooperation between resource agencies
and local governments. Due to the foresight  of in-
dividuals, this  process became a  partnership  in
which jurisdictional politics were set aside so that
land use planning based on ecosystems rather than
political boundaries could take place. We are  grate-
ful to the Environmental Protection Agency (EPA),
Department of Ecology  (DOE)  and Puget Sound
Water Quality Authority for funding this planning
effort; to the members of the SEWIP technical and
user committees; to Lynn Beaton for her assistance
during the grant and development of the plan; and
to Dr. Tom Hruby for his assistance throughout the
process on the development and application  of the
IVA model.

Literature Cited
Adams, L. W., and  L. E. Dove. 1989. Wildlife Re-
  serves  and Corridors in the Urban Environment:
  A Guide to Ecological Landscape Planning and
  Resource  Conservation.  National Institute  for
  Urban Wildlife, Columbia, MD.
Aronson,  J., C. Floret, E. LeFloc'h,  C. Ovalle, and
  R. Pontanier.  1993. Restoration and Rehabilita-
  tion of Degraded  Ecosystems in Arid and  Semi-
  Arid Lands. I. A  view  from the South.  Restora-
  tion Ecology 1(1): 8-17.
Bortleson, G. C., M. J.  Chrzastowski, and  A.  K.
  Helgerson. 1980.  Historical changes of shoreline
  and wetland at eleven  major deltas in the  Puget
  Sound  region,  Washington. Hydrologic Investi-
  gations Atlas HA-617, USGS.
Boule, M. E., G. B.  Shea, and A. L.  Driscoll. 1978.
  Snohomish estuary wetlands study: delineation of
  wetland  boundaries.   Volume  4.   Report   of
  Northwest Environmental Consultants, Inc., to
  United States Army Corps of Engineers, Seattle
  District, WA.

Burrell,  G.  1978. Snohomish estuary wetlands
  study:  classification  and mapping. Volume 3.
  Report of Washington State Department of Game
  to  United States Army  Corps of  Engineers,
  Seattle District, WA.

Collins, L. M., J. N.  Collins,  and L.B. Leopold.
  1987.  Geomorphic  Processes  of an  Estuarine
  Tidal Marsh: Preliminary Results and Hypothe-
  ses, pp. 1049 - 1072 In V. Garner (ed). Inter-
  national Geomorphology 1986 Part L. John Wiley
  and Sons, New York, NY.

	. 1991. Preliminary draft of Jetty Island
  management  plan.  Prepared for the  Port of

Cowardin, L. M., V. Carter, and F. C. Golet. 1979.
  Classification of wetlands and deepwater habitats
  of the United  States. U.S. Fish and Wildlife
  Service, Washington, D.C. FWS/OBS-79/31.

Driscoll, A. L. 1979. Snohomish Estuary Wetlands
  Study (4 volumes). Shapiro and Associates,  Inc.,
  Seattle, WA.
Forman, R. T. T. and M. Godron.  1986. Landscape
  Ecology. John Wiley & Sons, New York, NY.

Frenkel, R. E.  and J. C. Morlan. 1990. Restoration
  of the Salmon  River Salt Marshes:  Retrospect
  and Prospect. Department of Geosciences Oregon
  State University, Corvallis, Oregon. Final Report
  to the U.S.  Environmental  Protection Agency,
  Seattle, WA.
Hruby,  T.,  W. E. Cesanek, and  K. E. Miller. In
  preparation.  Establishing relative wetland values
  on a regional scale:  a step beyond WET.

Josselyn, M.,  J.  Zedler, and  T.  Griswold.  1989.
  Wetland Mitigation Along the  Pacific Coast of
  the United States, pp. 3-36 In  J. A.  Kusler and
  M. E. Kentula (eds). 1990. Wetland Creation and
  Restoration:   The Status of the Science.. Island
  Press, Covelo, CA.

Kusler,  J. A.  and  M. E.  Kentula (eds).  1990.
  Wetland Creation and Restoration: The Status of
  the Science.  Island Press, Covelo, CA.

 Snohomish Estuary
 Leibowitz, S. G . B. Abbruzzeze, P. R. Adamus, L.
   E.  Hughes,  and  G. T. Irish.  1992.  A Synoptic
   Approach  to Cumulative  Impact Assessment.
   EPA/600/R-92/167 October 1992.  EPA Environ-
   mental Research Laboratory, Corvallis, OR.
 Lewis. Roy R. III. 1982. Creation and Restoration
   of Coastal Plant Communities. CRC Press, Boca
 Lovvorn, J.  1995. Panel  discussion (January  13,
   1995) at Puget  Sound  Research  95, Bellevue,
   WA. University of Wyoming, Zoology  Depart-
   ment, Laramie, WY.
 McGuiness.  M.,  et  al.  1988. Snohomish  County
   Watershed Ranking Project, Final  Report. Public
   Works, Snohomish County, Everett, WA.
 National  Research Council (NRC).  1992. Restora-
   tion of Aquatic  Ecosystems.  National Academy
   Press. Washington, D.C.
 Pentec Environmental, Inc.  1992. Port of Everett
   Landscape Analysis for Port Gardner and  the
   Snohomish River Estuary. Pentec Environmental,
   Inc., Edmonds, WA.
 Pickett, S. T. A. and P. S. White (eds). 1985. The
   Ecology  of Natural   Disturbances  and  Patch
   Dynamics. Academic Press, Inc. Orlando, FL.
 Pritchard, D. W.  1967. What is an estuary:  physi-
   cal viewpoint, pp. 3-5 In Lauff, G. H. (ed.),
   Estuaries. Pub No. 83, American Association for
   the Advancement of Science, Washington, D.C.
 Schaefer, G. M., and M. T.  Brown. 1992.  Design-
   ing and protecting river corridors for wildlife.
   Rivers 3(1): 14-26.

 Shapiro and Associates, Inc.  1979. Snohomish Es-
   tuary Wetlands  Study. Volume  1:   Summary
   Report. Prepared for U.S. Army Corps of Engi-
   neers, Seattle District, WA.

 Shapiro & Associates, Inc. 1979. Snohomish Estu-
   ary  Wetlands   Study.  Volume   2:  Baseline
   information  and evaluation.  Prepared for US
   Army Corps of Engineers, Seattle District, WA.
Shapiro & Associates, Inc. 1985. Snohomish River
  wetland units preservation  management  plan.
  Prepared  for  Snohomish County Department of
  Planning and  Community Development, Everett,

Shapiro & Associates, Inc. 1989. Snohomish River
  wetlands management plan. Final report to Sno-
  homish  County  Department  of Planning  and
  Community Development, Everett, WA.

Shreffler, D. K., and R. M.Thom. 1993. Restoration
  of Urban  Estuaries:  New Approaches  for  Site
  Location  and Design. Battelle/Marine Sciences
  Laboratory. Prepared for Washington State De-
  partment  of Natural Resources Aquatic  Lands
  Division, Olympia, WA.

Simenstad,  C. A.  1983. The Ecology of Estuarine
  Channels of the Pacific Northwest Coast: A Com-
  munity Profile. U.S. Fish  and Wildlife Service,
  Division of Biological Services, WA, D. D. FWS/
Simenstad, C. A., C. D. Tanner, R. M. Thorn, and
  L. Conquest.  1991. Estuarine Habitat Assessment
  Protocol. EPA 910/9-91-037. Report to U.S. En-
  vironmental Protection Agency, Region 10, Fish.
  Res.  Inst., University  of Washington, Seattle,
  WA. 201 pp.

Stevens, M. L., and  R. Vanbianchi.  Restoring
  Wetlands  in  Washington.  1993. Department of
  Ecology Publication #93-17, Olympia, WA.
Williams, J. R., H. F. Pearson, and J. D. Wilson.
  1985. Stream flow statistics and drainage-basin
  characteristics for  the Puget Sound region, WA.
  USGS: 84-144 B.

Zedler,  J. B. and M. W. Weller. 1990. Overview
  and  Future Directions,  pp.  405-413 In J, A.
  Kusler  and  M.   E.  Kentula  (eds).  Wetland
  Creation  and  Restoration:  The  Status  of the
  Science. Island Press, Covelo, CA.

                                  Planning for Wetland  Restoration  and
                             Enhancement in the  Context  of  Regional
                                                 Wetland Management Plans
 Abstract:  As development pressures  on wet-
lands increase  in  urbanizing  areas, many are
turning to  regional wetland management plans
as a solution for the problem of balancing social
and environmental  needs.  As  these  regional
plans are being developed they provide signifi-
cant opportunities for the restoration  and en-
hancement of wetlands in the planning areas.
One of the issues, however, in using restoration
as  a mitigation tool is to quantify the loss of
wetland functions through development and de-
termine  whether any  planned restoration and
enhancement  will  actually mitigate  for lost
functions. This paper describes how the Indica-
tor Value  Assessment method (IVA)  is being
used assess the restoration potential of wetlands,
and meet mitigation needs for the Mill Creek
Special Area Management Plan.
  The IVA method provides a relative estimate of
the performance of wetland functions and their
value based on specific indicators. Numeric mod-
els use  the  indicators as  variables  that are
summed or multiplied. Value  is then estimated
based on: 1) the performance score, 2)  the rela-
tive value assigned the function by a local over-
sight/citizens committee, and 3) the area of the
wetland. The restoration potential of a wetland
is assessed  by estimating the value of functions
after restoration based on changes to indicators.
The mitigation potential is then set as the dif-
ference in  value between  existing and  future
  Of the 128 wetlands assessed  in the Mill Creek
SAMP, 42  have been identified as suitable for
enhancement. Initial calculations indicate that
the value of the floodflow alteration functions in
these wetlands  might be increased by about
20 percent; the value of fish  habitat functions
might  be increased  by about 50 percent; the
value of habitat  for other wildlife species might
be increased by about 50 percent, and the value
of water quality improvement functions might
                                                                          Thomas Hruby1
be increased by about 30 percent. The potential
increase  in  the value of  wetland  functions
through restoration/ enhancement also sets the
upper limit of the impacts to wetland values that
might be  allowed through development. Even
when wetlands which have lower relative values
for  the  different  functions  are  chosen for
development, the potential impacts have to be
limited to less than 400 acres of wetlands out of
the 2500 acres present in the basin. To balance
the value of functions lost through filling 400
acres  of   wetlands   will   require   restor-
ing/enhancing all 1465 acres of wetlands identi-
fied in the restoration plan.

 Development pressures on wetlands are increasing
as buildable  land  becomes  scarce in  urbanizing
areas. One solution for accommodating both eco-
nomic and environmental needs is to develop re-
gional management plans that identify, in advance,
wetlands  suitable for development and those for
preservation or restoration. Moreover, such regional
plans provide  significant  opportunities  for  the
restoration and enhancement  of wetlands. The
restoration or enhancement of  wetland functions
can  be used as compensation  for functions lost
through the development when the process is linked
in a regional plan. One issue,  however,  in using
restoration as a tool for compensation is to find a
method that  will adequately quantify the loss of
wetland   functions and  determine  whether  any
planned  restoration and enhancement will  actually
compensate for wetland functions  lost through de-
velopment. (NOTE: In  this paper I use the term
"restoration"  to mean both true restoration and en-
hancement: where restoration is the re-establish-
ment of wetland functions in areas that were once
wetlands, and enhancement is an increase in the
performance of some functions in areas that are cur-
rently wetlands.)
1 Washington State Department of Ecology, P.O. Box 47600, Olympia, Washington 98504-7600.

 Regional Wetland Management Plans
  This paper describes the process being used to in-
 tegrate restoration planning with requirements for
 compensatory mitigation in three regional  wetland
 management plans. The three  wetland management
 plans  under development are  the  Hackensack
 Meadowlands  Special Area Management  Plan  in
 New Jersey (Meadowlands SAMP); the Mill Creek
 Special Area Management Plan in Washington State
 (Mill Creek SAMP), and the Snohomish  Estuary
 Wetland   Integration  Plan  also  in  Washington
   Unavoidable impacts to wetland functions in all
 three plans are to be compensated by undertaking
 specific restoration  actions in the watershed. The
 wetland management plans are the  basis for U.S.
 Army  Corps of Engineers Regional General Per-
 mits under the Clean Water Act, Section 404. Com-
 pensatory mitigation for  impacts  outlined in  a
 development plan is constrained to specific activi-
 ties outlined in a restoration plan. This means that
 anyone proposing to impact wetlands will have to
 limit their compensation to activities outlined in the
 plan. The advantage of such an approach for a de-
 veloper is a more predictable and faster per-mitting
 process.  The advantages from a restoration per-
 spective  is  that  the compensation requirements
 provide some of the capital  necessary  to begin
 restoration efforts on a watershed scale.
   The basic premise in the three wetland manage-
 ment plans is that unavoidable impacts are compen-
 sated by  enhancing or restoring other wetlands in
 the planning region. The goal  is to achieve "no net
 loss of functions and values."  Wetland values lost
 through development are to be compensated by in-
 creasing values elsewhere in the watershed. In order
 for such a process to work, a method is needed that
 will quantify both the impacts to wetland functions
 and those that may be gained through restoration
 activities. No numeric method existed for assessing
 wetland functions that met planning needs when we
 started the process over five years ago in the Mead-
 owlands. As a result a new method was developed
 called  the  Indicator  Value Assessment  Method
 (IVA) (Hruby, et al.  1995).

  The IVA is used to track changes in the perform-
 ance of functions and their values resulting from un-
 avoidable impacts to wetlands  and those chosen for
 restoration.  Losses  in wetland values can  be  as-
 sessed directly from the IVA value scores  of wet-
lands that will be  impacted.   The  amount of
compensation that can be achieved is estimated by
recalculating the  performance and value scores of
functions  for areas that are restored  or enhanced.
Indicators that  would  change or be  added in the
"restored" condition are identified and added to the
assessment models. The compensation potential is
then set as the difference in value between  existing
and potential future conditions.

  Briefly:  The IVA method provides a relative esti-
mate of the performance of wetland functions and
their value based on specific indicators. Numeric
models use the  indicators as variables  that are
summed or  multiplied. Models take the following
           Performance = f(indicators)

  The coefficients reflecting  the importance of an
indicator in the expression of function are assigned
and developed based on local conditions. The wet-
land with the highest  performance in the basin is
scored at 100 and the rest are scored relative to this.
Value  is then estimated based on: 1) the perform-
ance score, 2) the relative value assigned the func-
tion by a local oversight/citizens committee, and 3)
the area of the wetland.
  Value =  f(performance, value of function, area of

  The wetland functions assessed in these planning
processes are chosen locally and reflect local val-
ues. Ten functions were assessed in  the Meadow-
lands SAMP; 12 in the Mill Creek Samp, and 15 in
the Snohomish  estuary plan (Table 1). To simplify
the valuation process, the performance scores for
the individual functions were combined into scores
for groups of functions such as Wildlife  Habitat,
Fish Habitat,  and Water  Quality   Improvement.
These  scores represent the relative value per acre
and  not  performance, because  the  performance
scores  for the individual functions were combined
based  on  the relative  values of the functions (see
Hruby  et al. 1995 for a more detailed discussion of
the calculations).

Developing  A Restoration Plan
  The following section describes  the process used
to plan restoration and enhancement  activities, and
this is  graphically represented in Figure 1. Planning
restoration and enhancement activities in  a water-
shed is initially independent  of the development of
the wetland management  plan and the assessment

                                                              Regional Wetland Management Plans
 Table 1. Wetland functions assessed in three wetland management plans
                                                          Lower Snohomish SEWIP
Meadowlands SAMP
Nutrient Uptake
Retention of Toxics
Export of Production
Floodflow Alteration
Conservation Potential
 Aquatic species
 General Fish
 General Waterfowl
 General Wildlife
Mill Creek SAMP
Nutrient Uptake
Retention of Toxics
Export of Production
Floodflow Alteration
Sediment Stabilization
Groundwater Discharge

Habitat Functions
Aquatic species
Anadromous Fish
Resident Fish
Migratory Bird
Resident Bird
All Other Species
Nutrient Uptake
Retention of Toxics
Export of Production
Sediment Stabilization
Channel Stabilization
Access to Deep Water

Anadromous Fish
Resident Fish
Migratory Bird
Overwintering Bird
Breeding Bird
Reptile and Amphibian
                              SET RESTORATION GOAL
                         SCORE RESTORATION POTENTIAL
Figure 1. Planning steps in developing a restoration plan linked to a regional wetland management plan.

 Regional Wetland Management Plans
 of wetland  functions.  The  linkage  between  the
 restoration plan and the wetland management plan
 occurs in  the  last two  steps.  The same planning
 process has been used in the two SAMP's and in the
 SEWIP. For the  sake of brevity, I will be using
 examples only from the Mill Creek SAMP and the
 SEWIP to illustrate the process.

 Step 1: Setting a Restoration/
 Enhancement Goal
   Most wetland  management plans  are developed
 for a small watershed or basin that is under heavy
 development pressures (e.g., City of Eugene  1992).
 In such cases any remaining wetlands are often de-
 graded by agricultural activities, stormwater runoff,
 or other human impacts. Many of the original func-
 tions performed by the wetlands have been lost, and
 the ecosystems in the watershed could benefit from
 some type of wetland restoration or enhancement.

   The first step in developing a restoration plan  is
 the crafting of  a restoration  goal suited for the
 watershed. In many cases in  may not be possible  to
 "move back the clock" and restore  the wetland
 ecosystem to a pre-industrial state. Furthermore, the
 remaining wetlands may have had their functions
 changed and now perform important new functions.
 For example,  in the Mill Creek watershed the ag-
 ricultural wetlands have become important habitat
 for migratory  waterfowl and raptors, when  in the
 past they were forested and  not accessible to these
 species. Restoration of the wetlands to their original
 forested state would eliminate this important current
 function.  The restoration   goal  for  Mill  Creek
 therefore is  "to establish an interconnected system
 of wetlands  and uplands along Mill Creek that will
 effectively  function  as a natural  ecosystem  that
 supports fish and other wildlife."  Some forested
 wetlands will  be restored but  the goal  is not  to
 restore the full extent of forested wetlands found  in
 the valley before settlement by Europeans.

  Watershed planning is effective only when coop-
 eration and consensus can be achieved among the
 different  owners and  "users"  in the  watershed
 (Euphrat and Warkentin 1994). Thus, it is important
 that the restoration goal for a watershed be set by a
 group with representatives from resource agencies,
 local landowners, and other interested parties.
Step 2:  Identifying Wetlands for
 Once a goal has been established the next step is
to  identify the wetlands best suited for restoration.
There are several ways in which this task may be
approached including 1) developing a map by con-
sensus    of    the    "restoration    committee,"
2) developing  specific  restoration  criteria  and
identifying areas that meet them; or 3) using a geo-
graphic information  system (GIS) (an approach to
restoration planning currently being developed by
the Washington State Department of Ecology for
the Puget Sound Water Quality Authority).
 Wetlands in the Mill Creek SAMP were ranked as
having a High, Medium, or Low potential for resto-
ration based on  the following criteria:  1) proximity
to the major streams in the watershed, 2) landscape
position, 3)  low relative performance  of  existing
functions,  4)  adequate water regime,  5)  location
relative to the 100-year floodplain, and 6) technical
feasibility. Of the  128 wetlands in  the  planning
area,  42  had  a High or  Medium  potential for

Step 3:  Developing Restoration
Objectives for Each Wetland
  Once  suitable  wetlands  have been identified, one
or more restoration  objectives  are developed for
each wetland. Objectives are chosen that  best ac-
complish the overall goal within the limits  of feasi-
bility at the  site. At this point  it is important to
identify where specific functions are to be restored
in the watershed so these can be targeted to the in-
dividual wetlands. Not all functions can be restored
to all  wetlands. Decisions should be  made  by  a
group of wetland experts who know and understand
the local ecosystems.  Restoration objectives were
set by such a group in all three plans developed.
The process  involved both  general discussions in
meetings and at least 2 to 3 days of field visits to
the wetlands that met the original criteria.
 Examples of restoration objectives  for three wet-
lands in the Mill Creek basin are listed below:
      Wetland 2E: Enhance wetland for waterfowl
          habitat and recreate a forested riparian
          buffer along the tributaries of Mullen
      Wetland 5KKK: Raise dissolved oxygen to
          acceptable levels.

      Wetland 5L:  Increase water supply in spring
          and early summer to encourage nesting by

Step 4:  Setting Restoration Tactics for
Each Wetland
  Restoration  objectives are implemented by cany-
ing out specific actions in a wetland. These actions
can be considered as "restoration tactics," and also
need to be  specified in  the plan. Although several
wetlands  may share similar objectives the tactics to
be used depend on conditions in that specific wet-
land. Tactics  need to be determined by a group of
local experts  and  verified by site  visits.  Because
tactics are very specific to the sites in question, they
will vary  between wetlands, basins and watersheds.
Furthermore a single wetland may  be targeted for
more than one tactic.

  The 19  restoration  tactics chosen  for wetlands in
the Mill  Creek  basin and the 15  chosen  for  the
lower Snohomish River estuary are listed Table 2.

Step 5:  Scoring  Restoration Potential
  Many indicators  of functions  in  a wetland will
change if restoration tactics are implemented as pro-
posed in  a  plan. It is the change in indicators that
forms the  basis of estimating the  increases  or
decreases in  the performance and values of func-
tions using  the IVA. New IVA scores are calculated
for each  wetland identified in the restoration plan
by postulating the indicators that would be present
after restoration. Thus, if the tactic  is to add inter-
spersion between land  and water the indicator for
interspersion  is changed to reflect  this (e.g., from
"no  interspersion" to "moderate interspersion"  or
  IVA performance  models are re-calculated  for
each  function based  on the changes to the indica-
tors. If a wetland already has the indicator the score
is not changed. This means that wetlands that  are
performing  functions well would  not  have their
scores changed  as  much as wetlands where func-
tions  are  poorly performed.  Furthermore,  the  in-
crease in  potential performance  is dictated by  the
tactics identified for that wetland. As a result each
wetland ends  up with a different potential increase
in performance.
  The performance scores for each function have
been combined  in the IVA models  developed into
                Regional Wetland Management Plans
scores for the per hectare value of major groups of
functions in each wetland, as described in Hruby, et
al. (1995). Table 3 shows the potential increase in
the per acre value  scores of wildlife habitat func-
tions and water quality improvement functions for
11 wetlands in the lower Snohomish River estuary.
Most of these 11 wetlands currently have low value
scores for relative performance  and thus provide
opportunities for significant increases in value. The
scores  suggest that the  performance of functions
may be increased two or three times over existing
levels by appropriate enhancements.

 Such was not the case in the Mill Creek, however,
where the wetlands identified for restoration only
averaged a 50 percent increase in the per acre value
of the habitat functions, and only a 30 percent in-
crease in the performance of water quality improve-
ment functions (Table 4). The existing wetlands had
fairly high levels of performance and thus could not
be enhanced much further.
Step 6:  Calculating Increase In Value
  The potential for compensation at each wetland is
estimated by multiplying the potential increase in
the performance score by the acreage of the wet-
land, as  shown in Table  3. The potential is  repre-
sented as a dimensionless number we have  called
"hectare-points" (or "acre-points" if English  meas-
ures  are  used).  These   numbers  represent  the
"values" in functions that can be accrued through
restoration and enhancement.

  Impacts are estimated in terms of "hectare-points"
that will be  lost by  multiplying the values  for
groups of functions in the wetlands to be impacted
by the size of the impact. Because the value  scores
generated in the restoration plan are comparable, it
is easy to establish  policies  by which impacts to
some wetlands are to be compensated for by resto-
ration or enhancement of other wetlands.  State  and
federal  policies state that the compensation  for
losses to wetland functions should be at least on a
one to one basis. Thus every hectare-point of impact
needs to be compensated by the  increase of  one
hectare-point through restoration  or enhancement.
In practice, however, the policies being developed
for the regional management plans call for replace-
ment ratios  greater than one to one, to compensate

 Regional Wetland Management Plans
 Table 2. Tactics used to enhance wetland functions.
                    Mill Creek SAMP
                     Everett SEWIP
 Riparian planting of trees
 Plant native emergent species
 Plant native scrub/shrub species
 Restore native forest
 Restore aquatic bed habitat elements
 Excavate ponds
 Add interspersion between vegetation classes
 Add interspersion between water and land
 Restore natural stream characteristics
 Restrict access
 Construct check dams to hold back water
 Reconfigure the stream channel
 Enhance access of floodwaters to wetland
 Add side channels off main stream
 Add structures to control water flows
 Control or remove livestock from wetland
 Control exotic or invasive plant species
 Restore upland buffers
 Modify existing flapgates
 Add specific habitat structures
Remove the maximum possible area of dike
Grade dike to provide natural transitions
Create islands
Reconnect remnant tidal streams
Excavate channels to "pre-dike" depths
Fill interior drainage ditches
Restore upland buffers
Restore a scrub/shrub  edge to ditches and streams
Restore emergent tidal marsh species
Add native emergent wetland plants
Restore full tidal regime
Increase seasonal flooding to 25 percent of wetland
Table 3. Changes in the IVA scores resulting from proposed restoration actions in the lower Snohomish estuary (SEWIP).
Wildlife Habitat Group of Functions

SITE Hectares Restoration
— •• 	 • 	 — .
— — ,
• 	 • 	 	
After Pts/ha
Water Quality Improvement Group of Functions
Total Score
Increase Before
ha/Pts Restoration
After Pts/ha

                                                                         Regional Wetland Management Plans
Table 4. Average scores for impacts and restoration potential used in developing policy for the Mill Creek SAMP.

                                                        GROUP OF FUNCTIONS
                                       Fish Habitat
 Habitat For Other
                                                                        Water Quality Improvement

       Development "footprint"                 10

       Wetlands BEFORE restoration            42

       Wetlands AFTER restoration              65

       NET gain through restoration            23

       RATIO of net loss/acre to net             0.4

       AVERAGE compensation ratio by        0.5
       acre of impact (including 1.25










for the loss of functions between the time a wetland
is impacted and the time that the restoration actions
have achieved their planned level of performance.
  The estimates of values lost and gained for each
wetland generated by the IVA can be summarized in
different ways to help guide policy decisions in the
plan development. Results from  the  Mill  Creek
SAMP are presented in Table 4  as an example. The
policies that are being developed from such data are
as follows:
1. The water  quality improvement group of func-
    tions are the limiting  ones  in terms of estab-
    lishing the amount of compensation required.
    On the average,  an acre of impacts to wetlands
    will require the enhancement  of 2.8 acres  of
    wetlands identified in the restoration plan.

2. Given that the average increases in values differ
   among the three  groups of  functions, it is  the
   limiting one (i.e., water  quality  improvement)
   that will be used to determine the size of com-
  pensation  necessary.  Any extra  values  to  be
  gained in the non-limiting function  groups  rep-
  resent a "benefit" for the natural resources in the
  area, and  are not tradeable for  impacts from
  other projects.
    3. Even when wetlands, which have lower relative
      values for the different functions are chosen for
      development, the potential impacts have to be
      limited to less than 400 acres of wetlands out of
      the 2500 acres present  in the basin. Compensa-
      tion for the value of functions lost through filling
      400 acres   of  wetlands   will  require restor-
      ing/enhancing all 1,465 acres of wetlands identi-
      fied in the restoration plan.
     In conclusion, I would like to note that the full
    process described has now been completed for the
    Mill Creek and the Meadowlands wetland manage-
    ment plans, and we are just beginning the process
    for  the  Lower Snohomish River  Plan. Final deci-
    sions on the how the exchange of credits is to take
    place are still to be made.

    Literature Cited
    City of Eugene. 1992. West Eugene Wetlands Plan.
      City of Eugene, OR.
    Euphrat, F. D., and B. P. Warkentin. 1994. A water-
      shed assessment primer.  EPA 910/B-94-005.
    Hruby, T., W. E. Cesanek, and K.E. Miller. 1995. Esti-
    mating relative wetland values for regional planning.
    Wetlands 15:93-107.


                                      Integrating Wetland Compensation
                                   Banking With LocalWetland Planning
                                                                          William J. Green1
  Abstract: Two recent alternatives to traditional
 wetland compensatory mitigation under the sec-
 tion 404  program  of the Clean  Water Act  are
 being  experimented  with  around  the  nation:
 wetland mitigation  banking and local wetland
 planning programs. Wetland mitigation banking
 is a process that coordinates multiple compensa-
 tory mitigation efforts into one mitigation plan or
 project. Local wetland planning is a process that
 coordinates the mitigation sequence steps for mul-
 tiple development  proposals within a  specific
 landscape management unit, with the goal of bal-
 ancing wetland protection and economic develop-
 ment. There are proponents and opponents of
 both processes. One major advantage often cited
 for these processes is that they allow compensa-
 tory mitigation to occur within a landscape per-
 spective or,  in  keeping  with the theme  of  this
 meeting, they encourage "taking a broader view."
 This paper outlines the opportunities as well as the
 difficulties  associated  with  "taking a broader
 view" in wetland  compensatory  mitigation  and
 focuses attention on future research needs.

  Throughout the United States, communities are
 experimenting with  wetland compensation banking
 and local wetland planning as alternative approaches
 to   satisfy  the   "compensatory  mitigation"  and
 "practicable alternatives  test" requirements of the
 Section 404 program of the Clean Water Act. There
 are  both  advantages and disadvantages  to these
 approaches. One major advantage is  that they  encour-
 age implementation  of  the  Section 404 program
 within a landscape perspective, or,  in  keeping with
 the theme of this meeting, they encourage "taking a
 broader view." This  paper explores the potential of
 integrating wetland  compensation  banking  efforts
 with local wetland planning. First,  I present the re-
 sults of a recent research project that indicates inter-
 est   among   local  government   institutions  for
integrating wetland banking efforts with local wet-
land planning. Second, I review wetland compensa-
tion banking and local wetland planning and consider
their link to the Section 404 program of the Clean
Water Act, the range of activities involved in each
approach, and the advantages and disadvantages as-
sociated with each approach. Third, I explore the idea
of integrating the two approaches by looking at how
their activities either overlap or complement one an-
other, and the advantages and disadvantages of inte-
grating the approaches. Finally, I conclude by provi-
ding a set of  research recommendations  that will
provide information to improve the integration of
wetland compensation banking with local wetland
planning and allow the Section 404 program to move
towards "taking a broader view."

Wetland Compensation Banking
  On  September  15,  1994, the Washington State
Department of Transportation (WSDOT), five  federal
agencies, and two state agencies reached a Memoran-
dum of Agreement (the Agreement) establishing the
WSDOT Wetland Compensation Banking  Program.
The purpose of the Agreement is  to "set  forth the
principles and  procedures that all signatories to the
Agreement will  adhere to in establishing,  imple-
menting, and  maintaining the Washington State
Department of Transportation Wetland Compensation
Bank Program" (WSDOT,  1994). The state and fed-
eral agencies did not include local and tribal govern-
ments in the negotiation process for the Agreement
due to the sheer number of county governments (39),
tribal governments (26), and many more city  govern-
ments in the State of Washington. However, local and
tribal governments  may participate  in  establishing
wetland compensation banks under the Agreement.
Many of these governments have their own  wetland
protection regulations and programs.
'School of Marine Affairs, University of Washington, Seattle, Washington 98195.

                                                                           Wetland Compensation Banking
 In February of 1994, WSDOT contracted with Pro-
fessor Marc Hershman, Director, School of Marine
Affairs, University of Washington, to investigate the
implementation of the Agreement within the legal
and institutional framework of county and tribal gov-
ernments. The goal of the study was to identify im-
portant  issues  WSDOT will  confront  when  it
implements the Agreement within county  and tribal
government  jurisdictions. The  research team  pub-
lished the results of the study in a final report titled
The Washington State Department of Transportation
Memorandum of Agreement for Wetland Compensa-
tion  Banking:  County  and Tribal   Participation
(Hershman and Green, 1995).
  Four key findings of the study relate directly to the
issue  of integrating  wetland  compensation banking
with local wetland  planning. First, each of the county
governments  chose their planning department as the
lead agency to respond to the Agreement. They see
wetland compensation  banking  as directly linked to
planning issues. Second, nearly all the county  and
tribal  governments desired a primary role in locating
wetland compensation  bank  projects. They believe
that local knowledge of land  use and  environmental
protection priorities should determine the location of
wetland bank sites. Third, the county and tribal gov-
ernments exhibit more flexibility in locating wetland
compensatory mitigation projects than currently al-
lowed by state and federal regulations. They believe
that limiting compensatory  mitigation  to "on-site"
and/or "in-kind" projects is too rigid. They encourage
"off-site" and "out-of-kind" compensatory mitigation
projects especially in urbanized areas. Fourth, most
of the county and tribal governments participate in
local planning efforts that include wetland protection.
  These findings indicate that county and tribal gov-
ernments  have an interest  in  integrating wetland
compensation banking and local wetland planning.

Wetlands Compensation Banking
 The  U.S. Fish and Wildlife Service explored the
idea of "banking" in response  to a number of requests
to "bank" fish and  wildlife habitat "credits" to offset
future adverse impacts to those resources from devel-
opment activity (Short, 1988). Soileau et al. (1985),
offer a good description of banking:
  "Mitigation banking is similar to maintaining
 a bank account. A developer undertakes meas-
  ures to create, restore, or preserve fish and
 wildlife habitat in advance  of an anticipated
  need for mitigation for project construction
  impacts.  The benefits  attributable to  these
  measures are quantified,  and  the developer
  receives mitigation credits from the appropri-
  ate regulatory  and/or planning agencies.
  These credits are placed in a mitigation bank
  account from which withdrawals can be made.
  When the developer proposes  a project in-
  volving unavoidable losses offish and wildlife
  resources, the  losses (debits) are quantified
  using the same  method that was used  to de-
  termine  credits, and a withdrawal equal to
  that amount is  deducted  (debited) from the
  bank.  This can be  repeated as long as mitiga-
  tion  credits remain available  in  the  bank. .
  (Soileau et. ai, 1985)"

Links to the Section 404 Program
  Commentators now commonly refer to the original
idea of "habitat banks" as "wetland mitigation banks"
because they nearly  always associate "banking" with
wetland mitigation requirements. Federal mitigation
authority is in the  National Environmental Policy Act
(NEPA) and the Fish and Wildlife Coordination Act.
The Council on Environmental Quality (CEQ) issued
regulations implementing NEPA and clarifying miti-
gation in 1978. The Council regulations  defined the
various  mitigation alternatives "to include: avoiding
impacts, minimizing impacts, rectifying impacts, re-
ducing impacts over time, and compensating for im-
pacts"  due to a  major federal action including the
issuance of a federal permit (Strand,  1993). The U.S.
Army  Corps  of  Engineers (COE)  had  authority
through these statutes and regulations, and as the lead
federal agency implementing the  Section 404 permit
program of the Clean Water Act, to require mitigation
for adverse impacts  to wetlands. However, no law or
regulation required the COE  to include mitigation as
a permit condition.  The CEQ mitigation  alternatives
were the foundation for the current "wetland mitiga-
tion sequence" established by the Section 404 (b)(l)
Guidelines issued by the Environmental Protection
Agency (EPA) in 1989. In  1990 the COE signed a
Memorandum of  Agreement (MOA) with the EPA
whereby the  COE  adopted  the  Section  404  (b)(l)
Guidelines approach to mitigation and agreed to at-
tach mitigation requirements when issuing a permit to
a project that adversely impacts wetlands (EPA/COE,
1990). The Section 404 (b)(l) Guidelines "mitigation
sequence" involves "avoiding" and "minimizing" ad-
verse  impacts  to  wetlands  by   modifying  or

 Wetland Compensation Banking
 redesigning  the  development  project,  and  then
 -compensating" for unmitigated, or unavoidable,  ad-
 verse impacts to wetlands as a condition of issuing a
 Section 404 permit. Wetland compensatory mitiga-
 tion is the replacement of adversely impacted wetland
 functions  and values by restoring, enhancing, and/or
 preserving existing wetlands or by creating new wet-
 lands. Compensatory mitigation is a tool to  off-set
 any loss of wetland functions and values due to  de-
 velopment projects and achieve the overall national
 goal of "no net loss" of wetland functions and values
 (Conservation Foundation, 1988).  The  1990 MOA
 also  states  that  "mitigation banking may  be  an
 acceptable form of compensatory  mitigation under
 specific criteria designed  to ensure an environmen-
 tallv successful bank." (I use the term "wetland com-
 pensation  banking" in this  paper rather than "wetland
 mitigation banking" because banking only involves
 the   last   step  in  the   mitigation   sequence—
 "compensatory mitigation.")

 Major Activities
   The following is a list of the major activities associ-
 ated with wetland compensation banking.
 1. A landscape  approach  to locating compensatory
    mitigation projects  including locations that  are
    "off-site" and "out-of-kind"  from the anticipated
    future unavoidable, adversely impacted wetlands.
 2. The restoration, enhancement, and/or preservation
    of existing wetlands and/or  the  creation of new
    wetlands from uplands for the purpose of compen-
    sating  in advance for anticipated future unavoid-
    able,   adverse   impacts  to  wetlands  due   to
    development activity. In most cases, wetland com-
    pensation  banking  activity  occurs  within  the
    boundaries of the same  landscape unit (watershed,
    sub-watershed,  estuary, etc.) in  which  adverse
    impacts occur.

 3. The assessment and evaluation of wetland func-
    tions and values before and after the occurrence of
    unavoidable, adverse impacts to wetlands to meas-
    ure net  loss.

 4.  The assessment and evaluation of wetland func-
    tions and values before and after the occurrence of
    compensatory mitigation projects to measure  net

 5.  The establishment of a  currency system to allow
   exchanges between the  functions and values  of
   adversely impacted wetlands (debit wetlands) and
   compensatory mitigation  wetlands  (credit  wet-
6. The management, maintenance, and monitoring of
   compensatory mitigation projects to ensure com-
   pliance with the permit conditions.
7. Legal protection "in perpetuity" for compensatory
   mitigation projects  using  in fee purchases, ease-
   ment  restrictions, or  some other legally  binding

Advantages and Disadvantages
  Wetland compensation  banking attracts both propo-
nents and opponents. A healthy debate has ensued re-
garding the approach since its conception in the early
1980s. The following is a list of the major advantages
and disadvantages associated with wetland compen-
sation banking  versus the traditional  case-by-case
approach to compensatory mitigation.

1. Compensation projects occur in  advance of un-
   avoidable, adverse  impacts to  wetlands allowing
   regulators to measure their success over a span of
2. Small impacts, for which compensation  is often
   "not practicable," can be consolidated into a larger
   compensatory mitigation project.
3. Large projects promote efficient  "economies  of
   scale" in  financing, planning,  implementing, as-
   sessing, evaluating,  monitoring, managing, and
   maintaining compensatory mitigation projects.
4. Assessing  and  evaluating   adversely  impacted
   wetlands and compensatory mitigation wetlands is
   standardized resulting in  more consistent imple-
   mentation  of the  compensatory  mitigation  re-

5. Compensatory mitigation is linked to the functions
   and values of a landscape rather than limited to the
   specific functions and values of an adversely im-
   pacted wetland.

6. Permit certainty is increased and conflict reduced
   by removing parts of the compensation  process
   from the permit process.

7. Public  awareness  of  compensatory  mitigation
   projects is increased  because wetland  compensa-
   tion banks tend to be large, high profile projects.

                                                                            Wetland Compensation Banking
8.  Advanced compensatory mitigation "credits" off-
   set the constitutional "takings" issue by creating
   value for wetlands as wetlands.

9.  Advanced compensatory mitigation "credits" pro-
   vide  incentive for credit  producers to purchase
   land appropriate for wetland restoration, enhance-
   ment, preservation, and/or  creation when  these
   sites are available rather than limiting land acqui-
   sition to the time period when compensatory miti-
   gation is required.

  1.    Mitigation planning  is weakened when ad-
vanced  compensatory mitigation projects create in-
centives    to   substitute    "compensation"    for
"avoidance" and "minimization" under the mitigation
  2.    Wetland function and value assessment and
evaluation methods are not adequate to establish an
accurate currency system for the complex transac-
tions between credit wetlands and debit wetlands.
  3.     Large, consolidated compensatory mitigation
projects effect surrounding land use, land value, and
the local tax base.
  4.     It is  difficult to  finance compensatory miti-
gation projects in advance unless they are linked di-
rectly to the budget of a development project.
  5.     Planning for a bank is expensive.
  6.     Potential loss of small wetlands and wetlands
that are  difficult to restore,  enhance,  or create for
compensatory mitigation.
  7.     Off-site and out-of-kind compensatory miti-
gation fragments the environment.
  8.    Legacy of past compensatory mitigation fail-
ures. Wetland restoration, enhancement, and creation
technology and science are not adequate to  achieve
success in compensatory mitigation projects.

Local  Wetland Planning  Programs
 The term  "local wetland  planning"  can  describe
many types of processes. This paper considers a spe-
cific type of local wetland planning process described
 Local  wetland planning is  a process  that balances
wetland protection and development within a  specific
landscape area. Urban and suburban  areas that con-
tain  limited  upland  development  sites and  very
intense  development  pressure are  the  most likely
areas for wetland planning initiatives. The approach
is to allow future development in low value wetlands
when an upland site is not available, while preserving
high value wetlands  and providing "compensatory
mitigation" for  any adverse  impacts to low value
wetlands.  The  intention is to create  a "win-win"
situation where development occurs in the least valu-
able environments, and wetland  preservation  and
"compensatory mitigation" occur in the most valuable

Links to the  Section 404 Program
 Local  wetland planning links to  the Section 404
program in two ways. First, the process is similar to
the "practicable alternatives  test"  required by  the
Section  404 (b)(l) Guidelines issued by the EPA in
1989. The Guidelines assume that if a project is not
"water dependent," meaning it does not require ac-
cess to,  or location within, a wetland, then there will
be  practicable,  upland alternatives, unless clearly
demonstrated otherwise (Strand, 1993). The Guide-
lines also assume that a practicable, upland  alterna-
tive will  have  less adverse  impact to the  aquatic
ecosystem unless, again, clearly demonstrated other-
wise. The "practicable alternatives test" that occurs in
a wetland planning process is different from the tra-
ditional test under an individual permit review. The
analysis involves numerous, future  development pro-
posals rather than a single, immediate development
proposal. And the analysis does not consider alterna-
tive development sites outside the  boundaries of the
landscape unit.
 The second link  between local  wetland planning
and the Section 404 program is the process of ranking
wetland functions and  values within the landscape
and identifying wetland sites suitable for future
development projects,  wetland sites not suitable for
future development, and wetland sites appropriate for
future   compensatory   mitigation  projects.   The
"Special Area  Management Plan" (SAMP), author-
ized and funded under the federal Coastal Zone Man-
agement Act, can include legally binding agreements
on  wetland sites suitable and not suitable for future
development that  can be used to either support or
deny a Section  404 permit application. The EPA pro-
vides funding for wetland  surveys, assessments, and
evaluations  through  its  "Advanced  Identification"
program. However, the COE and EPA do not have to
abide by these determinations when reviewing a Sec-
tion 404 permit application unless the agencies sign a

 Wetland Compensation Banking
 legal agreement  such as a SAMP or a MOA. The
 purpose of the determinations is to facilitate future
 Section 404 permit application reviews.

 Major Activities
   Federal, state,  county, city,  and/or tribal  govern-
 ments  as well as non-governmental  groups  sponsor
 and participate in local wetland plans. Local  wetland
 planning usually involves relatively small landscape
 scales  such as a single watershed, a sub-watershed, a
 river corridor, a  greenway, an estuary, or a complex
 of wetlands. The following is a list of the major ac-
 tivities associated with local wetland planning.
 1. Determine the boundaries of the  landscape plan-
    ning unit.
 2. Identify wetlands in the landscape unit.
 3. Assess and evaluate  the  functions and values of
    the  wetlands  before  the  occurrence  of potential
    development or compensatory mitigation projects.
 4. Rank  wetlands based on their functions and val-
 5. Prioritize  wetlands for preservation,  restoration,
    and enhancement based on their  ranks (some of
    these  sites may be identified as  appropriate for
    compensatory mitigation projects).
 6. Prioritize wetlands suitable for future development
    based on  their ranks  (include consideration  of
    "practicable  alternatives" to  locating a develop-
    ment in a wetland).
   mem in a wetiana;.

7.  Analysis  of present land ownership and zoning
   patterns  for potential land  swaps, transactions,
   and/or re-zoning that supports the wetland plan-
 8. Integrate other land use and resource management
    programs  into the  wetland  planning  process
    (transportation, surface water management, flood
    management, fish and wildlife management, etc.).
 9. In  some cases,  establish a  streamlined  permit
    process for regulated activities that are in accor-
    dance with the  wetland plan.

 Advantages and Disadvantages
  Local wetland planning also attracts proponents and
 opponents. The following  is a list of the major ad-
 vantages  and disadvantages  associated with local
 wetland planning  versus the traditional case-by-case
approach  to  reviewing proposed  development  proj-
ects that adversely impact regulated wetland area and
determining compensatory mitigation requirements.

1.  Future development, compensatory mitigation, and
   wetland protection decisions are based on analysis
   that considers landscape scale processes.
2.  Increased certainty and predictability for both de-
   velopment and environmental protection.
3.  Increased consistency and efficiency in the permit
   review process.
4.  Creative opportunities to create  "win-win"  solu-
   tions by swapping, purchasing, and/or re-zoning
5.  Cooperation  and coordination  between programs
   with overlapping and/or complementary goals and
6.  Planning includes local needs, conditions, and per-

 1. The boundaries  of the  landscape unit  limit the
   "practicable alternatives test." Local wetland plan-
   ning usually does not consider an alternative de-
   velopment site  outside  the landscape planning

2. The "practicable alternatives test"  is more specu-
   lative when applied in a planning context. It is dif-
   ficult to predict future development demands and
   needs, especially over the 20 year planning period
   often used in planning processes.

3. Identifying  wetlands  that rank  as  "low"  value
   wetlands  suitable for future development reduces
   incentives  for   developers   to   "avoid"  and
   "minimize" adverse impacts to "low" value wet-

4. The public can  only review an anticipated future
   development project during the  planning process
   rather  than the actual development project during
   the permit review process.

5. Legacy of past  compensatory  mitigation failures.
   Wetland  restoration,  enhancement,  and creation
   technology and  science is not  adequate to achieve
   success in compensatory mitigation projects.

                                                                           Wetland Compensation Banking
Integrating Wetland Compensation
Banking Into Local Wetland
 It is a recurring theme in the wetland compensation
banking literature that the future of banking will link
with local wetland planning processes (Short,  1988;
Castelle, 1992; Kusler and Lassonde, 1992; Reppert,
1992;  World  Wildlife Fund,  1992; Dennison and
Berry, 1993; McKenzie  and Rylko, 1993; National
Association of Industrial and  Office Parks,  1993;
ELI,  1994;  Brumbaugh and Reppert, 1994; Silver-
stein, 1994; Association of State Wetland Managers,
1995;  COE et. al.,  1996).  Successful compensatory
mitigation is more likely when projects recognize the
system in which they are located (National Research
Council, 1992).  Planning is  the best tool wetland
regulators and developers have to locate compensa-
tory mitigation projects in a manner that is sensitive
to the surrounding ecological systems, land-use prac-
tices, and long-term social goals. "It is sometimes
said about  real estate that value depends upon loca-
tion, location, location. Similarly, the importance of
wetland functions and values also depend upon loca-
tion, location, location" (Kusler, 1992). Coordinating
the activities of wetland  compensation banking with
local wetland planning provides opportunities to re-
duce the costs of wetland protection, increase regu-
latory consistency,  improve wetland functions and
values at the landscape scale,  and facilitate local
community goals, conditions, and perspectives.

Overlapping and Complementary
  Many of  the  activities  associated with wetland
compensation banking and local wetland planning,
directly overlap  or complement one another.  An
overlapping activity is one that occurs  in both ap-
proaches; a complementary activity is one that occurs
in only one of the  approaches, but it facilitates the
goals of the other approach. Integrating the two ap-
proaches would reduce the replication of the overlap-
ping  activities  and  facilitate the  complementary
activities. The following  are lists of overlapping and
complementary activities between wetland compen-
sation banking and local wetland planning.

Overlapping Activities
1.  Landscape  scale  analysis to identify of wetland
   sites for  restoration, enhancement, and preserva-
2. Assessment and evaluation of the functions  and
   values of wetland sites before the occurrence of
   development impacts.

3. Assessment and evaluation of the functions  and
   values of wetland sites before the occurrence of
   restoration,  enhancement,  and/or  preservation
   projects for compensatory mitigation.

4. Establishment of a currency system to allow com-
   plex exchanges between the functions and values
   of adversely impacted wetlands and compensatory
   mitigation wetlands.

5. Planning and analysis confined to a landscape unit
   (watershed, sub-watershed, estuary, etc.).

Complementary Activities

1.  Wetland banking's actual restoration,  enhance-
    ment, and/or preservation projects  for compen-
    satory mitigation.

2.  Wetland banking's assessment and evaluations of
    wetland functions and values after the occurrence
    of development impacts.

3.  Wetland banking's assessment and evaluation of
    the functions and values of wetland sites after the
    occurrence  of restoration, enhancement, and/or
    preservation projects for compensatory mitiga-
4.  Wetland  banking's   long-term  management,
    maintenance,  and monitoring of compensatory
    mitigation projects.
5.  Local  wetland planning's determination of the
    boundaries of the landscape planning unit.

6.  Local  wetland planning's identification of wet-
    lands in the landscape planning unit.

7.  Local  wetland planning's ranking of wetlands
    based on their functions and values.

8.  Local wetland planning's  analysis of present land
    ownership and zoning patterns for potential land
    swaps, transaction, and re-zoning.

9.  Local  wetland planning's integration of other
    land use and resource management programs.

10. Local  wetland  planning's establishment  of  a
    streamlined permit process for  regulated  activi-
    ties that are in accordance with the wetland plan.

 Wetland Compensation Banking
 Discussion of Overlapping and Comple-
 mentary Activities
   Obviously, it would  be a waste of effort for a wet-
 land banking project  and a local wetland planning
 process to  produce 2 landscape analyses, 2 assess-
 ments  and evaluations of wetland  functions and val-
 ues, and 2 currency systems for the same wetlands in
 the same landscape unit. Not only would that result in
 a waste of time  and money, but it would lead to in-
 consistent and perhaps conflicting results.
   Imagine a situation  in which a wetland compensa-
 tion bank moves, independent of local consideration,
 to purchase a degraded wetland area along a river
 corridor that is  deemed  appropriate for restoration
 and enhancement by  its  own  analysis. Imagine  the
 same wetland area is  considered ideal for future  de-
 velopment  by a local wetland planning process. In
 fact, the area is zoned for development and is consid-
 ered an excellent location for shipping and water de-
 pendent type activities. The local community needs
 this type of economic investment to provide jobs. The
 local government is also aware that the amount of
 land zoned for development is shrinking and they  can
 not afford to loose the tax revenues. This scenario is
 not  so imaginary.  Several county  governments in
 Washington State  mentioned  this type of concern
 when they were  asked to review and comment on the
 Washington  State  Department   of  Transportation
 Wetland Banking Agreement (Hershman and Green,
 1995).  Combining  overlapping activities  between
 wetland banking and planning is a very effective way
 to reduce costs, improve regulatory consistency, and,
 perhaps most importantly, provide for coordination
 and  cooperation  between potentially conflicting  ac-
 tivities at the local level.

   Complementary activities present a similar oppor-
 tunity. There are  many activities  associated with
 wetland compensation banking that do not occur in
 local wetland planning and vice versa. Coordinating
 these activities can also reduce conflicts and improve
 the success of both approaches. For example, local
 wetland planning processes identify all wetlands in a
 landscape  unit and ranks each according to their
 functions and values relative to other wetlands in a
 landscape unit. This is more thorough than the analy-
 sis that is likely  to occur in a wetland banking proj-
 ect. Conflicts between the two approaches would be
 reduced if  a wetland banking  project participated
 and/or adhered to the decisions of a local wetland
 plan. Complementary activities also represent an  ex-
 tra tool for both approaches. For instance, the actual
wetland restoration, enhancement, creation, and pres-
ervation efforts that occur as compensatory  mitiga-
tion  in a banking framework directly  accomplish
objectives of a local wetland plan to protect and im-
prove the wetland  resources in the landscape unit.
The  streamlined  permit  process  for development
projects that are  consistent with the local wetland
plan directly improves the objectives of a banking
project to increase permit certainty and predictability
for developers.
  It is possible for complementary activities to over-
come a disadvantage associated  with a  single  ap-
proach  implemented independently  of  the other
approach. A good example of this is when  a local
wetland planning process reduces the potential for a
wetland banking  project to  negatively impact sur-
rounding land use, land value, and the local tax base;
to fragment the environment by  using "off-site" or
"out-of-kind"  compensatory mitigation; or to accu-
mulate large planning expenses. These are included
in the list  of disadvantages  when  using a wetland
compensation banking approach. Complementary ac-
tivities are an opportunity to improve the success of
both approaches.

Advantages and Disadvantages
  Integrating wetland compensation banking with lo-
cal wetland planning presents a number of advantages
and a few disadvantages. The following is a list of the
most important advantages and disadvantages of inte-
grating the two approaches versus implementing them


1. Combining overlapping  activities  reduces costs
   and increases regulatory consistency.

2. Complementary   activities  improve  both  ap-
   proaches in some cases overcoming  a disadvan-
   tage associated  with a single  approach imple-
   mented independent of the other.

3. Compensatory  mitigation projects occur in  the
   most appropriate sites and target priority functions
   and values for the broader ecosystem.


1. Lack  of regulatory guidance  and experience for
   implementing wetland compensation banking and
   local wetland planning.

                                                                           Wetland Compensation Banking
2.  Legacy of past compensatory mitigation  failures.
   Wetland  restoration, enhancement, and  creation
   technology  and  science  are  not adequate  to
   achieve   success  in  compensatory   mitigation

Discussion of Advantages and
  Wetland compensation banking and local wetland
planning, implemented independent of one another.
have advantages and  disadvantages  that are listed
earlier in this paper. It is a different issue to consider
the advantages and disadvantages of integrating these
approaches  versus implementing them separately.  I
already discussed the three  advantages listed above,
the  disadvantages listed above require some clarifi-
cation. Currently, there is very little regulatory guid-
ance  or   experience   for   integrating  wetland
compensation banking  with local  wetland planning.
An  interagency work group just recently  issued for-
mal guidance for establishing wetland compensation
 banks (COE et. al., 1995). The northwest regional of-
 fices of the COE and EPA  are working on  guidance
 for  local wetland planning (personal communication
 with COE staff in the Seattle District Office). Guid-
 ance documents clarify how these approaches should
 be structured so they comply with the Section 404
 program requirements. It will  be difficult to integrate
 wetland banking and planning until guidance exist for
 both. Even  with  guidance, integrating the two  ap-
 proaches will require  an adaptive  learning  period
 before  enough  experience accumulates to avoid the
 hidden pitfalls in innovative approaches.
   Finally, the legacy of past compensatory mitigation
 failures is  inescapable. Wetland  restoration,  en-
 hancement, and creation  technology  and  science are
 not adequate to produce the kind of results necessary
 to achieve "no net loss" of wetland functions and val-
 ues within the compensatory mitigation framework.  A
 recent  series of papers presented as a "Forum"  in
 Ecological Applications is an excellent treatment of
 the subject (the most  applicable were: Zedler, 1996;
 Race and Fonseca, 1996; Mitsch and Wilson, 1996).

 Research Needs
   A number of areas of research will help improve the
 successful   integration   of   wetland compensation
 banking with  local wetland planning.  First,  both
 approaches  rely on landscape analysis  methods  to
 prioritize wetland areas  appropriate for  restoration,
enhancement, or preservation within a specified land-
scape unit such as a watershed. A greater under-
standing  of  how  landscape  ecology  applies  to
wetlands will be very useful. Issues such as the most
useful scale of analysis for wetlands should be deter-
mined (watershed, migratory pathways, material flow
pathways, etc.). Second, both approaches rely on as-
sessment and evaluation  methods to measure wetland
functions and values. Methods need to be established
that are both flexible enough to apply in different re-
gions and rigid enough to provide consistency within
regions.  These methods must  also  be lay person
"friendly" at some level to provide for public partici-
pation.  They  should  also  be  transferable  between
compensatory mitigation activities  at the local wet-
land planning stage and  during the  actual implemen-
tation  stage  within a wetland  banking framework.
Third,  regulators  and developers need  to know ex-
actly how to comply with the regulatory requirements
of the Section 404 program while experimenting with
innovative approaches  like  wetland banking and
planning. Finally,  further understanding of wetland
restoration, enhancement,  and  creation  technology
and  science will greatly improve the odds for com-
pensatory  mitigation  planning  and  implementation
  Research will not answer these needs overnight.
Ongoing wetland compensation banking projects and
local wetland planning processes, and efforts to inte-
grate the two approaches, should be designed as ex-
periments to produce learning  opportunities.  Future
efforts can adapt  and improve based on the  results
from  the earlier  experiments  (Lee,   1993).  This
"adaptive management" approach is the best way to
move  forward with wetland compensation banking
and local wetland planning, conducting important re-
 search to improve the two approaches, and "taking a
broader view" while implementing the Section 404
program requirements.

 Literature Cited
 Association of State Wetland Managers. 1994. Effec-
   tive Mitigation:  Mitigation Banks and Joint Proj-
   ects In the Context of Wetland  Plans. Association
   of State Wetland Managers, Berne, NY.
 Association  of  State  Wetland  Managers.   1995.
   Watershed Management and Wetland Ecosystems:
   Implementing   Fair,   Flexible,  And  Effective
   Approaches,  Background Report.  Wetlands  '95

 Wetland Compensation Banking
   National  Symposium  in  Tampa  Bay,  Florida.
   ASWM, Berne, NY.
 Brumbaugh, Robert and Richard Reppert. 1994. First
   Phase Report. U.S.  Army Corps of Engineers, In-
   stitute for Water Resources, IWR Report 94-WMB-
   4, Alexandria, VA.
 Castelle, Andrew et.  al. 1992. Wetlands  Mitigation
   Banking, Washington State Department of Ecology,
   Publication #92-12,  Olympia, WA.
 Conservation Foundation. 1988. Protecting America's
   Wetlands: An  Action Agenda. National Wetlands
   Policy Forum. Washington, DC.
 Dennison, Mark S. and James F. Berry. 1993. Wet-
   lands:  Guide to Science, Law,  and Technology.
   Noyes Publications. Park Ridge, NJ.
 Environmental Law Institute. 1994. Wetland Mitiga-
   tion  Banking. U.S.  Army Corps  of Engineers. In-
   stitute for  Water  Resources. IWR Report 94-
   WMB-6. Alexandria, VA.
 Hershman,  Marc and  William J. Green.  1995. The
   Washington  State  Department  of Transportation
   Memorandum of Agreement for Wetland Compen-
   sation  Banking:  County and Tribal Participation.
   Washington  State Transportation Center. Seattle,
 Kusler, Jon A.  1992. Mitigation Banks and the Re-
   placement of Wetland Functions and Values.  In
   Effective Mitigation: Mitigation  Banks and Joint
   Projects In  the Context of Wetland Management
   Plans.  Association  of State Wetland  Managers,
   Berne, NY.

 Kusler, Jon A. and Cindy Lassonde. 1992. Effective
   Mitigation: Mitigation Banks and Joint Projects In
   the Context  of Wetland  Management Plans. The
   Association of State Wetland Managers, Berne,

 Lee,  Kai N. 1993. Compass and  Gyroscope: Inte-
   grating Science and Politics for the Environment.
   Island Press. Washington, D.C.

 McKenzie, Tracy P. and Micheal Rylko. 1993. Part-
   nerships in Restoration  Mitigation  Banking.  In
   Partnerships and  Opportunities in Wetland Resto-
   ration. Fred  Weinman and et. al. (eds) U.S. Envi-
   ronmental Protection Agency, Region 10, Seattle

Mitsch, W. J. and Renee F. Wilson. 1996. Improving
  the  Success  of Wetland Creation and Restoration
  with Know-How, Time, and Self-Design. Ecologi-
  cal Applications 6(l):77-83.
National Association of Industrial and Office Parks.
  1993.   Wetland   Incentives:   Non-Regulatory
  Approaches to Protecting Wetlands. Puget Sound
  Water Quality Authority: Olympia, WA.

National  Research  Council.  1992. Restoration of
  Aquatic   Ecosystem.  National  Academy  Press.
  Washington, DC.

Race, M. and M. S. Fonseca. 1996. Fixing Compen-
  satory Mitigation: What Will  It Take? Ecological
  Applications 6(1):94-101.

Reppert, Richard.  1992. Wetland Mitigation Banking
  Concepts. U.S. Army Corps of Engineers, Institute
  for  Water  Resources,  IWR  Report 92-WMB-l,
  Alexandria, VA.
Short, Cathleen. 1988. Mitigation Banking. U.S. De-
  partment of Interior, Fish and Wildlife Service,
  Washington, DC.
Silverstein, Jonathon. 1994.  Taking Wetlands to the
  Bank: The Role of Wetland Mitigation Banking in
  a Comprehensive Approach to Wetlands Protection.
  Environmental Affairs 22:129-161.

Soileau, et. al. 1985. Mitigation Banking: A Mecha-
  nism  for Compensating  Unavoidable  Fish  and
  Wildlife  Habitat Losses.  In Transcripts from the
  North American Wildlife and  Natural Resource
  Conference, pp. 465-474.
Strand, Margaret N. 1993. Federal Wetlands Law. In
  The Environmental Law Reporter: Wetlands Desk-
  book. Environmental Law  Institute. Washington,
  DC. pp.  1-107.

U.S. Army Corps of Engineers,  et. al. 1995. Federal
  Guidance for the Establishment, Use and Operation
  of  Mitigation Banks;  Notice.  Federal  Register
  60(228):58605-58614, November 28,1995.

U.S. Environmental Protection Agency and  the U.S.
  Army Corps of Engineers. 1990. Memorandum of
  Agreement Between the Environmental Protection
  Agency  and the Department of the Army Concern-
  ing the   Determination of  Mitigation  Under the
  Clean Water Act Section  404(b)(l) Guidelines.
  pp. 331-336, In The Environmental Law Reporter:
  Wetlands Deskbook.  Environmental Law  Institute.
  Washington, DC. pp.  331-336.

Washington  State  Department  of   Transportation
  1994.    Washington   State    Department    of

                                                                        Wetland Compensation Banking
Transportation   Wetland   Compensation  Bank    Zedler, J. B. 1996. Ecological Issues in Wetland Miti-
Program Memorandum  of Agreement.  Olympia,      gation: An Introduction to the Forum. In Ecological
WA. September 15, 1994.                             Applications 6 (1): 33-37.

                                               Watershed Analysis as a Tool
                                                      in Watershed Restoration
                                                                                Steven Toth1
  Abstract:  Watershed analysis  in Washington
 State was developed by constituents of the Timber
 Fish  and  Wildlife  (TFW)  agreement (Native
 American tribes, state  agencies, environmental
 groups, and the timber industry). The methodol-
 ogy  addresses  the  cumulative  effects  of  forest
 practices on fish and water resources in the con-
 text  of natural ecological processes. While  other
 watershed  analysis  methodologies  have  been
 developed recently (e.g., Federal Agencies Guide,
 Idaho), TFW watershed analysis is unique  in its
 scientific rigor and ability to affect land manage-
 ment/restoration  decisions.  Close  interaction
 between scientists  and policy  makers allowed
 development of a method that is repeatable, defen-
 sible, and accountable to constituents.
  The watershed analysis process consists of three
 components:  resource assessment, prescriptions
 and  monitoring. An interdisciplinary team of sci-
 entists with expertise in geology, hydrology, fish
 biology, soils, and forest  ecology assesses the
 physical and biological components of the water-
 shed. Natural disturbance processes (e.g., floods,
 fire) and human-related disturbance (e.g., timber
 harvest, grazing) are evaluated  in  the context of
 present land management activities. The evalua-
 tion relies primarily on aerial photographs, moni-
 toring  data,  landform maps  and  geographic
 information systems, with some additional field
 work. The team delineates  specific areas on the
 landscape that are sensitive to management  prac-
 tices  and  have  the  potential  to impact public
 resources. The resource assessment and synthesis
 of  information  for watersheds  ranging in size
 from 100  to  200 square  kilometers  and  takes
 approximately 2 to 3 months.

  In the prescriptions phase, land managers  work
 with scientists to develop options for operating in
 sensitive areas. Potential restoration  actions are
 also  identified  and  prioritized  to improve  or
restore  aquatic habitat conditions.  The entire
watershed  is  addressed regardless  of adminis-
trative boundaries. The direct connection between
the scientists  and land managers ensures  that
information generated by the assessment team is
at a scale  appropriate  for guiding management
decisions in the field.
  Finally, a monitoring  program is developed to
evaluate the implementation of the prescriptions,
test assumptions about  watershed process, and
assess recovery of aquatic resources. Monitoring
priorities are developed jointly with  stakeholders
in the watershed. Preservation and restoration of
ecological processes are evaluated and prioritized
using the information generated from the water-
shed analysis. An example of a watershed analysis
conducted in the central Cascade range of Wash-
ington state is used to illustrate the assessment of a
watershed,  development of prescriptions,  and
monitoring of watershed conditions.

  Strategies  for protecting aquatic resources while
allowing land management activities must be suffi-
ciently flexible to accommodate landscape variability
in the Pacific Northwest (FEMAT 1993). Methodolo-
gies for conducting watershed  analysis in forested
watersheds  have been developed by Federal agencies
(U.S. Department of Agriculture 1994), the state of
Washington (Washington  Forest  Practices  Board
1995),  and  the  state of Idaho (Idaho Department of
Lands 1995) to link land management with scientific
assessments. These procedures are also being used on
private lands in the states of California, Oregon and
Montana and in British Columbia, Canada.
  Although many similarities exist between these pro-
cedures,  varying  objectives  have led  to different
methodologies  for assessing watersheds. The federal
watershed  analysis  process characterizes human,
1 Plum Creek Timber Company, L.P., 999 Third Avenue, Suite 2300, Seattle, Washington 98104.


                                                                                      Watershed Analysis
aquatic and terrestrial conditions and is more appro-
priately described as ecosystem analysis at the water-
shed scale. The federal analysis is driven primarily by
issues identified prior to the start of the analysis.  A
list of approaches is provided for the interdisciplinary
team of resource specialists, but no specific method-
ology has to be followed. The analysis is not a deci-
sion-making process, but establishes the context for
subsequent planning (Regional Interagency Executive
Committee 1995).
  The Washington State and Idaho processes focus
primarily on aquatic resources and the physical proc-
esses that  influence aquatic habitat, but the Wash-
ington state  process has more rigorous  scientific
analysis. Both  processes have tightly choreographed
methodologies  that  the  analysts must  follow.  The
Washington state  process, however, requires scien-
tific assessment of watershed processes  by  an inter-
disciplinary team of certified specialists. The Idaho
process is designed to  be implemented  by a single
trained resource manager. The Idaho analysis focuses
on inherent hazards within a watershed and current
stream conditions, rather  than  characterizing  proc-
esses.  Both procedures  directly influence decision-
making   through   development   of  management
prescriptions based on the watershed assessment.
  Many people  have looked to watershed analysis as a
means for guiding and implementing watershed resto-
ration, despite  the fact that most of these  processes
were not explicitly developed to meet this objective.
For watershed analysis to be useful in development of
restoration plans, data must be gathered at a scale that
can be utilized at the project level. In particular, any
in-stream or riparian restoration work requires an un-
derstanding of  watershed processes such as flooding
history, sediment supply, and geomorphic context at a
local scale. Of the three procedures described previ-
ously,  the  Washington state process is  probably the
best suited for prioritizing and implementing  land
management,  monitoring  and  restoration   projects
because of its more rigorous use of scientific analysis
and ability to  affect  management/restoration  deci-
sions. This paper outlines the  general  methods for
conducting Washington  State  watershed  analysis,
describes restoration approaches, and examines how
information from watershed analysis can be used  to
prioritize and implement restoration projects.
Washington State Procedure for
Watershed Analysis
  Watershed analysis is a regulatory process admin-
istered  by  the Washington  State  Department  of
Natural Resources (DNR) on  state  and private land
ownership in Washington. The analysis is designed to
address the cumulative effects of forest practices on
the public  resources  of fish,  water,  and capital
improvements  (e.g.,  bridges and county  roads).  A
watershed analysis can be initiated by either the DNR
or voluntarily by a private landowner who owns more
than 10 percent of a watershed. Watersheds range in
size from approximately 100 to 200 km2.
  The Washington state watershed analysis  procedure
consists of four distinct components:
1.  Resource  Assessment:   Scientists  identify  hill-
    slope hazards by assessing mass wasting, surface
    erosion, hydrology, and  riparian condition. They
    also identify vulnerable resources by assessing
    fish habitat, stream channels, water quality, and
    capital improvements (Table  1). Sensitive areas
    of the watershed are delineated where hillslope
    hazards can affect a vulnerable resource (e.g., an
    area prone to landslides  that delivers sediment to
    a fish-bearing stream).
2.  Prescriptions:  Land  managers  and  scientists
    design prescriptions for  each  sensitive area. Pre-
    scriptions  are methods for operating in sensitive
    areas to reduce or eliminate potential problems.
    Standard forest practice rules are applied on the
    remainder of the watershed.
3.  Public Review: The public is given the oppor-
    tunity to  review and comment on the findings
    through the  State Environmental Policy Act
4.  Monitoring: A monitoring plan is developed to
    track  changes in  watershed conditions and test
    the effectiveness of prescriptions.  Monitoring is
    voluntary,  although most landowners  have initi-
    ated some monitoring following completion of
    the analysis.
  Watershed analysis was developed by scientists and
managers from state agencies, Indian  tribes and the
timber industry working cooperatively under the state
Timber/Fish/Wildlife  agreement.  Watershed analysis
creates additional forest practice rules tailored to spe-
cific  watersheds using a science-based  assessment
procedure. The assessment  is typically conducted in
two to three months by  an  interdisciplinary team of

Watershed Analysis
 Table 1. Summary of Watershed Processes and Resources Addressed by the Washington
 State Watershed Analysis Modules
Watershed Analysis
Mass Wasting
Surface Erosion
Riparian Function
Channel Condition
Fish Habitat
Water Supply / Public
. - I .- * ,-*"" »
Watershed Processes and Resources Addressed
.* — :>- ' "' .: - """' '>:':^l N. - '
• Debris Torrents
• Landslides
• Earthflows
• Hillslope Surface Erosion
- Dry Ravel
• Road Erosion
• Peak Streamflows
• Summer Low Flows
• Large Woody Debris Recruitment
• Shade/Water Temperature
• Bank Stability
• Historic Channel Disturbance
• Current Channel Condition
• Spatial Distribution of Channel Response Types
• Dominant Habitat Forming/Geomorphic Processes
• Distribution and Relative Abundance of Salmonid Fish
• Existing Habitat Condition
• Fish Habitat Utilization and Preferences
• Location and Sensitivity of Water Supplies/Public Works
Public State Roads and Bridges
Reservoir, Irrigation Structures
- Municipal, Domestic, Hatchery Water Supplies
 Table 2. Management Response for Areas of Resource Sensitivity
Likelihood of Adverse Change and Deliverability



Standard Rules
Standard Rules

Standard Rules
Standard Rules



                                                                                      Watershed Analysis
natural resource specialists that are certified by the
DNR.  The team  may  include  geomorphologists,
hydrologists, soil scientists, biologists, and other spe-
cialists   as  needed.   The   assessment   provides
information on  physical and  biological  processes
within a watershed such  as the  spatial distribution of
processes,  recent  changes in  the  condition of the
watershed, and how these changes influence aquatic
  The resource assessment information is passed onto
a prescriptions team that is typically composed of
land managers and some of the specialists  that par-
ticipated in the resource assessment phase.  The pre-
scription team develops various options for operating
in sensitive areas to provide as much flexibility for
landowners as possible while  still protecting public
resources.  The assessment report  and prescriptions
must be approved by the DNR  and go through public
review. The entire watershed analysis process  typi-
cally takes a full year to complete.
  Monitoring is  a  critical  component of watershed
analysis because there are often limitations in the sci-
entific assessment and many of the prescriptions are
experimental in  nature. Guidelines  for  developing
monitoring plans  are  provided  in  the watershed
analysis manual (Washington Forest Practices Board
 1995); however, monitoring is  currently not required
of landowners. A review of the watershed  analyses,
though, is  required every five years and should pro-
vide information on the effectiveness  of prescriptions
in  protecting  resources.  Analyses  may  also  be
reviewed  sooner if fish habitat  degrades,  prescrip-
tions are not working or new scientific information is
  Washington State watershed analysis is designed to
create options  for conducting timber  management
activities in a manner that maintains  natural rates of
sediment input from hillslopes, minimizes  potential
changes in streamflow and provides adequate riparian
corridors for maintaining temperature, large woody
debris and nutrient levels. The following sections
describe in more detail the procedures used to evalu-
ate  these  watershed  processes  and  relating  the
processes to regulatory prescriptions. For a complete
listing of procedures, refer to  the Washington State
Watershed  Analysis Manual  (Washington  Forest
Practices Board 1995).
  Mass wasting is the dominant source of sediment in
many Pacific Northwest watersheds (Swanston and
Dyrness  1973,  Sidle  et  al. 1985, Megahan  1983).
While landslides often initiate far above fish-bearing
waters,  debris  flows, dam-break floods and  fluvial
processes can route sediment downstream and affect
fish habitat far below the landslide (Swanston et al.
1987). Both the volume and rate of sediment influx is
considered in light of its effects on fish habitat.
  Delivery of coarse sediment to streams is addressed
primarily by  the  Mass  Wasting module. The mass
wasting  analyst maps all historical  mass wasting
within the watershed as defined by the aerial  photo-
graph record and evaluates  the sediment  delivery
potential  to   streams.  A landslide   inventory  is
produced and correlations are made with a number of
variables including geology, slope gradient,  slope
form, and forest practices. The analyst uses these data
to produce a mass wasting hazard map that identifies
areas with the potential for mass wasting.
  Another significant concern is  fine sediment (less
than 2 mm) from mass wasting, hillslope erosion, and
road erosion that  can reduce the viability of eggs in
spawning gravels, reduce rearing habitat by filling the
interstitial spaces  of cobbles and  gravel, and increase
turbidity and nutrient levels in  streams (Lisle and
Hilton  1992, Lisle 1989, Chapman 1988). Fine sedi-
ment input into streams is addressed by the Surface
Erosion and Mass Wasting modules. The surface ero-
sion analyst evaluates recent harvest units and road
construction in the field to examine the potential for
gullying and sheetwash erosion. The road network is
also evaluated  using an  empirically-based model to
estimate sediment production from roads. Predicted
sediment yields are based primarily on road charac-
teristics  including traffic  levels  and road surfacing.
Management-derived  sediment inputs  are compared
to estimates of natural sediment input rates as part of
a crude sediment budget to evaluate hazards. In this
context,  estimates of sediment input  are useful for
estimating  relative contributions  from the  various
sources of sediment  in the watershed.  Other sources
of fine sediment from land use  activities  such as
grazing and agriculture as well as episodic  natural
events such as fire are  considered in the context of
inputs from timber management activities.

 Watershed Analysis
   Removal  of forest cover can  increase snow ac-
 cumulation, allow  for  greater  wind  speeds,  and
 increase  solar radiation,  thereby  increasing  the
 amount of snowmelt, especially during rain-on-snow
 events (Coffin and  Harr 1992).  Increasing  the  size
 and frequency of flood flows from timber harvest is a
 concern because larger and more frequent flood flows
 can change channel morphology through increased
 sediment  and large woody  debris transport  and
 greater bank erosion. Larger and more frequent flood
 flows can also increase the depth  of gravel bed scour,
 potentially scouring  fish redds.
   The hydrology module primarily addresses changes
 in peak flows during rain-on-snow events. The analy-
 sis procedure  for peak flows consists of evaluating
 canopy coverage in  relation to elevation. A modified
 U.S.  Army Corps  of  Engineers (1956) empirical
 snowmelt model is  used to estimate the increase in
 water available for  runoff during various magnitude
 storm events. The water available for runoff is then
 related to  streamflow using either  gage data, field
 estimates of flood flows, or regional regression equa-

 Riparian Corridors

 Large Woody Debris
   Large woody debris (LWD) is vital for maintaining
 fish habitat in most Cascade streams. LWD dissipates
 stream energy, influences sediment storage and trans-
 port,  and provides  habitat and cover for fish both
 directly and through changes in channel morphology
 (i.e., pool formation)(Bilby 1984, Bisson et al. 1987)
   The riparian module addresses  large woody debris
 recruitment to streams  primarily by  assessing the
 condition of riparian areas within 20 to 30 meters on
 either  side  of streams. This  distance equates to
 approximately one-half the site potential tree height
 and generally  accounts for 85 to 100 percent of the
 natural woody debris input to streams (McDade et al.
 1990, Robison and Beschta 1990, Murphy and Koski
 1989). Forest stand  type (coniferous,  mixed  or
 deciduous), relative age,  and density are measured for
 each stream  reach  from  aerial photographs.  The
 assessment  concentrates  on fish-bearing streams, but
 non-fish-bearing streams less than 20 percent gradient
 are also considered.

  The potential for future LWD  recruitment is con-
 sidered  together  with   the   amount   of   present
in-channel LWD to address concerns about present
and future timber management. In-channel LWD data
is collected in the field jointly with the channel and
fish  module  team members.  The  riparian analyst
interacts with the channel analyst to assess channel
recruitment   mechanisms   such   as   meandering
channels, debris flows and bank erosion. The  width
of riparian recruitment can, thus, be extended beyond
20 to  30 meters  depending on  site potential tree
height and channel processes.

 Salmonids require relatively cool  stream tempera-
tures for all  life  history  stages.  During the summer
when stream  temperatures  increase, cool  tempera-
tures are especially important for rearing  juveniles
and spawning adults. Empirical evidence in Wash-
ington   shows that there  is a  strong relationship
between  canopy  cover,   elevation  and  stream
temperature (Sullivan  et al. 1990). This has led to the
development  of a temperature  screen  that specifies
the amount of canopy  coverage needed at a given ele-
vation to meet or exceed state water quality  standards
for stream temperature.
 The riparian module analyst uses aerial photographs
to assess canopy cover and topographic  maps  to
establish elevations. All fish-bearing streams as well
as non-fish-bearing streams that contribute at least 20
percent of the flow to a fish-bearing stream are con-
sidered (Caldwell et  al. 1991). At  given  elevation
zones,  targets for canopy coverage have been  estab-
lished through the state forest practice rules.
 If canopy cover meets specified target levels, the
stream is assumed to be below state maximum water
temperature criteria at that location.  If canopy cover
does not meet specified target levels,  it is assumed
that  maximum   water temperature  standards  are
exceeded. Available temperature data can be used to
identify or verify problem areas.

Channel Condition
 The Channel module assesses past changes in chan-
nel morphology and processes, current channel con-
ditions, and  the potential sensitivity of channels to
changes in  inputs of sediment, wood, and  water.
Channel morphology reflects and integrates processes
operating in  a  watershed because  material  eroded
from hillslopes ultimately is delivered to and routed
through  the   channel network.  Channel  and  fish

                                                                                       Watershed Analysis
habitat condition, therefore, reflect the relative input
of sediment, wood and water relative to the ability of
the channel to either transport or store these inputs.
  The stream channel network is stratified into geo-
morphic units.  Geomorphic  units  are reaches  of
streams that .respond in a similar fashion based on
comparable channel-forming processes. Alluvial fans
or steep-gradient canyon tributaries would be exam-
ples of geomorphic  units. These channel-forming
processes are derived from the general geology and
climate which dictates  stream gradient and confine-
ment and hillslope topography and vegetation. Geo-
morphic units allow assessment of channel conditions
on a watershed basis and provide a context for evalu-
ating the influence of land management activities.
These units become the basis for linking hillslope and
channel processes  during the synthesis  portion  of
watershed analysis.

Fish Habitat
  It is difficult to assess all the factors that affect sal-
monid production because of their wide-ranging life
histories. Anadromous fish populations can vary sub-
stantially simply based on factors outside of their
freshwater life history phase. For this reason, the fish
habitat module  assumes that  evaluation of physical
stream  habitat  characteristics  will provide an ade-
quate  measurement of salmonid production  during
their freshwater life history phase. It is assumed that
degradation of physical habitat features will result in
reductions in salmonid production.
  Two basic  premises of the  fish habitat evaluation
are  that: (1)  physical habitat  characteristics  are
strongly influenced by geomorphic  setting,  and (2)
old-growth  conditions  most  closely  represent the
conditions to which  multiple  species have  adapted
over the past several thousand years. This approach
does not imply that preferred fish habitat only occurs
in old-growth forests, but that knowledge of habitat in
old-growth forests can form the basis for identifying
changes in habitat conditions  (Peterson et al. 1992).
Indices of habitat  conditions  are based  on habitat
utilization  and  on  stream characteristics that  have
supported a  multitude  of species prior  to  human-
induced habitat  changes.  Physical  habitat features
which  are  evaluated include  depth   and  velocity
ranges (grouped as channel units such as pools and
riffles), pool frequency, pool size, cover, spawning
gravels,  and  temperature ranges. Other elements of
fisheries evaluated by the module include historic and
current salmonid fish distribution, relative abundance
of salmonids, and an assessment of factors limiting
fish production.

Synthesis of Channel Condition and Fish
 The channel and fish module analysts jointly deter-
mine  the vulnerability of fish habitat to changes in
physical processes based on local channel conditions
and fish biological requirements. The geomorphic
units  form  the  basis  for assessing  vulnerability
because channel forming processes are assumed to be
similar within  a  geomorphic unit.  Vulnerability is
assessed  for processes  such  as:  (1) debris torrents
(debris flows and dam-break floods); (2) increases in
coarse sediment;  (3) increases in fine  sediment; (4)
changes  in hydrology  (primarily peak flows);  (5)
decreases in large woody debris and shade; and (6)
removal of near-bank riparian vegetation.
 The vulnerability of fish habitat  is  rated  "high,"
"medium," or "low" based on the channel sensitivity
to a  change in these  processes and  the  potential
impact on fish production. A  channel geomorphic
unit may not be particularly sensitive to a given input
process, but if there is concentrated fish use (e.g., a
chum spawning reach) or limited habitat availability
within that unit (e.g., a single area that accounts for
most  of the coho winter rearing habitat), the vulner-
ability rating is changed to reflect the unit's impor-
tance for fish production.

Synthesis, Causal Mechanism Reports,
and  Prescriptions
 An integral part of conducting watershed analysis is
linking hillslope  and  channel  processes.  This may
involve routing sediment derived from management-
induced  landslides through the channel network or
discussing the role of wildfire in producing surface
erosion and its consequent effect on fish production.
 During the synthesis  process, the resource assess-
ment  team identifies  areas  of resource  sensitivity
based on the likelihood of adverse change and deliv-
erability to vulnerable resources. Some hazard areas
identified by the resource assessment  team may not
be considered resource sensitive areas if impacts can-
not be delivered  to  the  resource of  concern  (e.g.,
unstable  slopes  that  do  not  deliver sediment to
streams). The resource sensitive areas  are designated

 Watershed Analysis
 relative to the hazard area, rather than to the stream
 segments with the affected resource.
  The resource assessment team rates both the likeli-
 hood of adverse change and the resource vulnerability
 as  "low," "medium,"  or  "high." These  ratings  are
 placed in a management response matrix (Table 2) to
 provide direction on which situations need prescrip-
 tions and the standard to which prescriptions will be
 written.  There  are   three  potential management
 responses (rule calls): (1) standard  forest  practice
 rules; (2) a  prescription  that minimizes  potential
 impacts to the resource of concern;  and (3) a pre-
 scription  that prevents impacts that would damage or
 prevent the recovery of public resources. While the
 ratings end up simplifying complex technical prob-
 lems, it provides an important link between scientific
 assessment and regulatory policy.

  A primary tool for synthesizing the results of the
 resource  assessment is the causal mechanism report.
 A causal mechanism report is produced for all areas
 of resource sensitivity. These reports  provide a brief
 and focused summary of problem areas  that can be
 used easily  by the  prescriptions team.  The causal
 mechanism  report summary has a  brief  problem
 statement, identifies linkages between land manage-
 ment and resources of concern,  and provides addi-
 tional comments to help guide the prescription team
 in  developing appropriate options  for timber man-

  The  prescription   team   develops  management
 options consistent with the  standard of protection
 required  by  the rule  call in  each causal mechanism
 report.  The team must provide justifications for each
 prescription  to ensure that the standards for protec-
 tion are met. The DNR also provides technical review
 of the  documents prior to initiation of the SEPA
 process to confirm that the watershed analysis meets
 regulatory standards.

  Generally, three types of monitoring are employed
 upon completion of watershed analysis; validation,
 trend and effectiveness monitoring (MacDonald et al.
 1991). Validation monitoring assesses the  scientific
 validity underlying elements  of the resource assess-
 ment.  Examples  of  validation  monitoring might
 include testing of empirical models with local data or
 e%aluating assumptions about woody debris input.
 Trend monitoring  assesses resource conditions over
 time.  Examples of trend monitoring could include
collecting large woody  debris  data  over time or
measuring   streamflow.  Effectiveness   monitoring
evaluates whether prescriptions produced the desired
result.  Examples  of effectiveness monitoring could
include evaluation of prescriptions for road construc-
tion near unstable slopes or silvicultural prescriptions
in riparian areas.
  The watershed analysis manual contains a monitor-
ing  module  to assist  in  the  development of a
monitoring program. The module is designed to assist
in prioritizing  monitoring  projects  and  to develop
specific monitoring objectives and  time frames  for
evaluating the effectiveness of the watershed analysis
process. The monitoring program is typically devel-
oped  in  conjunction with all stakeholders in  the

Quartz Mountain  Watershed
  In order to better understand how watershed analy-
sis is implemented, the following example outlines
the results of  a  watershed analysis  in  the Quartz
Mountain watershed administrative unit (WAU). The
Quartz Mountain watershed analysis was initiated by
Plum Creek  Timber Company in  September 1993.
The watershed analysis team was comprised of repre-
sentatives from Plum Creek, Yakama Indian Nation,
Department  of Ecology,  Department   of  Natural
Resources, U.S. Forest Service, and the Washington
Environmental  Council. The entire watershed analy-
sis  process  was  open  to public  participation. A
detailed description of the analysis is contained in the
Quartz Mountain Watershed Analysis Report (1994).

Overview of the Watershed
  The  Quartz Mountain  WAU consists  of the  area
drained by the North  and South Forks of Taneum
Creek to the point of their confluence (Figure 1).  The
Taneum  Creek watershed is located in  mountainous
forestland east of the Cascades crest, approximately 2
km (1.24 miles) southeast of the town of Cle Emm, in
Kittitas  County,  Washington. The upper  Taneum
Creek watershed  area is approximately  29,400 acres
(46 mi2). About 40 percent  of the watershed is owned
by Plum Creek with the remainder administered by
the U.S. Forest Service.

  Elevations  in the  watershed range from approxi-
mately 853  m  (2,800 feet) at the  confluence of the
North  and South Forks of Taneum Creek to 1,917 m

Type 2-5 Stream
Public Land Survey Line
PCTC Ownership
USFS Ownership
Other Ownership
         FIGURE  1
  Location and Ownership
within Quartz Mountain WAU

 Watershed Analysis
 (6.290  feet)  at Quartz Mountain. The watershed  is
 characterized by a relatively wide low-gradient
 floodplain along both forks of  Taneum  Creek with
 steep sideslopes ranging from 10 to greater than 50
 degrees. The geology of the basin is characterized by
 a complex mix of volcanic,  metamorphic and sedi-
 mentary rocks with a well-mixed soil mantle of vol-
 canic ash.
   Local climate within the Quartz Mountain WAU
 generally consists of warm, dry summers and cold,
 snowy  winters. Based on records at Cle Elum, tem-
 peratures average -3  °C (26°F)  in January and 19°C
 (66°F)  in July. The basin receives most of  its pre-
 cipitation,  principally  as  snow,  during  October
 through March. Annual  precipitation increases with
 elevation from approximately 1,378 mm (35  inches)
 at lower elevations to 2,756 mm (70 inches) or more
 in the upper elevations.
   Fish  are present in most of North and South Fork
 Taneum Creek, as well as in the lower portions of the
 major  tributary streams. Resident fish species include
 rainbow trout (Oncorhynchus   roykiss), westslope
 cutthroat (O.  clarki), brook trout (Salvelinus font-
 mails), and  sculpin  (Cottus   spp.).  Historically,
 Taneum Creek supported anadromous runs of spring
 chinook (O. tshawytscha), coho (0. kisutch), and
 steelhead (O. mykiss), but  an  irrigation  diversion
 ditch approximately 17.7 km (11 miles)  downstream
 of the  WAU boundary has served as a migration bar-
 rier since around 1910. Bull trout (Salvelinus conflu-
 entus)  were not found  in  1993  presence/absence
 surveys, but likely had historical access to the water-
 shed from the Yakima River.
   The watershed area is dominated by ponderosa pine
 forests at lower elevations, mixed-conifer forests  at
 middle elevations, and by true  fir forests  at higher
 elevations.   Before   the  development   of   modern
 methods of fire suppression, wildfire played a major
 role  in shaping the forests of this region  (Agee,

   The Quartz Mountain WAU area was probably first
 used by settlers for grazing cattle and sheep between
 the 1880s and 1930s. Selective removal of large trees
 for railroad ties occurred in  the lower watershed  in
 the 1920s, with roughly 800 acres harvested in four
 locations. Road construction began in the mid-1950s,
 followed by  extensive partial cutting  in the east end
 of the Lower North Fork sub-basin and  upper Frost
 Creek.  The upper half of the WAU was not  entered
 for harvest until 1984. Plum Creek harvests approxi-
mately  150-300 acres  per  year.  The  U.S. Forest
Service is ending timber harvest on their ownership
with current plans to place the area in Late-Succes-
sional Reserve.
Results of Resource Assessment
  Stream channels are shaped by a number of impor-
tant  variables that interact to create characteristics
unique to each stream. Some variables such as the
gradient, valley confinement  and drainage area of a
stream are relatively unchanged by human activities.
Other  variables, however, such  as  the amount  of
coarse and fine sediment, the amount of large wood
in the stream channel,  and the volume and timing of
floods can  be influenced by  timber  management
activities. These variables influence the channel mor-
phology  and dictate the quality of habitat  available
for fish.  Studying  the channel morphology,  water
quality, and riparian area, thus provides a  surrogate
assessment of the health of the stream system for
  The most pervasive problem identified  in the Quartz
Mountain Watershed Administrative Unit (WAU) is
the excessive amount of fine sediment in both forks
of Taneum Creek. Based on McNeil sampling, fine
sediment  less than 1 mm in diameter is abundant in
the active channel, ranging from 17 to 20 percent of
the total volume of sediment in  1993. Fine sediment
and small gravel (up to about 16 mm), fills a signifi-
cant portion of pool volume. Fine sediment accumu-
lation  in riffles and pools fills interstitial  spaces in
gravels,  reducing  the  suitability  of spawning  and
rearing habitat for resident fish. Winter  rearing habi-
tat may be the limiting factor  for resident  fish  sur-
vival in Taneum Creek.
  In  general,  the  Quartz  Mountain WAU stream
channels may be classified as either steep tributaries
or low-gradient mainstem reaches. The low-gradient
mainstem reaches have historically stored and trans-
ported large quantities  of fine  sediment.  Bankfull
flows are sufficiently powerful to mobilize  and trans-
port fine sediment entrained in the channel bed; how-
ever, because  of  abundant  large wood and  other
channel  roughness  elements,  extensive  secondary
channels, and  frequent opportunities  for  overbank
flow,  much  fine  sediment  that is  transportable is
deposited in temporary storage sites.

  The  storage function in the low-gradient mainstem
segments is strongly  influenced  by  large  woody
debris. Both local storage associated with debris jams

                                                                                      Watershed Analysis
and individual pieces of wood, and formation of
secondary channels caused by debris jams are critical
sediment storage functions of large  wood. Moreover,
abundant wood tends to reduce stream energy avail-
able for sediment transport and encourages overbank
flow  locally.  These  factors  all contribute to  the
potential for fine  sediment storage in  low-gradient
mainstem channels.
  A major sediment routing function of steep tributary
streams is the transport of fine  sediment. Neverthe-
less, significant storage of coarse and fine  sediment
occurs in tributary channels. Large woody  debris is
the primary  sediment storage element  in  tributary
channels. The analysis recognized the importance of
long-term recruitment of woody debris  in Type 4
streams to provide storage for fine sediment.
  Residence times for fine sediment in the mainstem
channel are likely to be at least 8 years,  and could be
as much as much as 50 years or more. With respect to
fine sediment,  this residence time estimate suggests
that a reduction of fine sediment inputs  would not be
reflected in stream channels for  a  period of several
years to decades.
  There are a number of sources for fine  sediment
input to the stream system. Soil creep, stream bank
erosion,  landslides,  hillslope surface  erosion,  and
road  erosion all contribute fine sediment in varying
proportions to streams. Soil creep is the rate of down-
slope movement  of the soil column.  The  transfer
process of sediment from hillslopes to streams occurs
primarily through bank erosion and landsliding.
  To identify the relative contributions of each input
variable and to better assess the role of timber man-
agement activities in sediment production, a partial
sediment budget was constructed. A partial sediment
budget identifies sediment sources and quantifies the
rate of sediment production, but does not completely
address  transport  of sediment  through  the  stream
network.  Sediment source estimates are within the
order-of-magnitude accuracy assigned to this type of
  Measurements of bank erosion and  landslides were
collected in sampled segments and extrapolated for
the stream network. A total  of  105 landslides were
inventoried from  field work and aerial  photographs.
Eighty-three  percent  of historic  landslides  have
occurred under natural conditions. The remaining 17
percent  are  associated  with timber  management
activities,  specifically  small, shallow   stream-side
failures in recent harvest units, failures in road cuts,
and one landing sidecast failure.
 Hillslope surface erosion in the watershed is limited
due to the high natural permeability of native soils
and duff cover which limits overland flow. In the vast
majority of harvested areas, recent standard practices
have  prevented  measurable   sediment  delivery  to
streams from surface  erosion.  It appears that  the
highest erosion potential  is  associated with soil-
exposing activities (ground-based harvest or slash
burning) which occur adjacent to streams, rather than
on any certain soil types or slope gradients. Surface
erosion of hillslopes was of insufficient quantity to
include in the sediment budget.
 Sediment delivery to streams from road erosion was
particularly high in relation to other sources for this
watershed. Fine sediment input to streams was high
primarily due to roads with unvegetated cutslopes and
the high proportion of native surface roads with direct
entry to streams. In the lower portion of both forks of
Taneum Creek,  another  problem was  unsurfaced
roads constructed on highly erodible phyllite-derived
soil that ruts readily. Abandoned roads and motorbike
trails contributed relatively small quantities of sedi-
ment compared to active secondary and low-use spur
roads. The erosion from existing roads were rated for
each sub-watershed. Most sub-watersheds (i.e., 7 of
9) received a high hazard rating because the estimate
of fine sediment input  from roads exceeded the total
fine sediment input from all natural sources.
 Based on a  15-year period, fine sediment from roads
in North Fork Taneum Creek is approximately 1.5
times greater than from other sources including land-
slides, surface  erosion of exposed slopes,  and bank
cutting from both hillslopes and remobilized stream
alluvium. The results  of the  sediment budget, how-
ever, indicate that the potential errors in measure-
ments and estimates are of the same magnitude as the
comparative  increases in road sediment over back-
ground. Although the road erosion rates may be over-
estimated, it  is reasonable to  assume that delivery of
road sediment to streams is of the same magnitude as
natural sources.
 The  input of coarse sediment from timber manage-
ment activity was not  perceived as a problem in the
basin at this  time. Coarse sediment is abundant in the
mainstem channels and relatively mobile. The coarse
sediment is stored on lateral  bars (point bars) and in
medial bars.  Sources of coarse sediment are primarily
the terraces adjacent to mainstem channels and, in the

 Watershed Analysis
 lower portion of  the  watershed,  streamside-land-
 sliding. Channel migration and erosion of secondary
 channels are thought to be important fluvial processes
 generating coarse sediment inputs. It is  likely  that
 only the  most extreme events (e.g.,  debris flow or
 massive,   multiple  streamside  landslides)   would
 significantly affect channel behavior as influenced by
 coarse sediment. Coarse sediment residence times are
 on the order of centuries in mainstem channels and
   The primary process through which land manage-
 ment could  accelerate  coarse  sediment input  is
 landsliding. Slope stability is dependent on both slope
 and  rock competence affected by fracture, bedding
 attitude and composition. Slopes  in the upper water-
 shed  underlain primarily with competent metamor-
 phic rock exhibit  areas of small-scale rockfalls with
 talus  aprons, snow avalanches, and  debris flow
 morphology with  attendant shallow soil failures. In
 contrast,  the layered sedimentary and volcanic rocks
 are less competent and have  planes of weakness in
 the form of shale,  siltstone, or  coal or tuffaceous
 interbeds. Failures occur along the weak interbeds.
   Ancient, large-scale mass  wasting features have
 occurred in the basin in response to past climatic
 regime  and  seismic activity. Indicators of recent
 movement  of these features, however, are lacking,
 and we can assume stable conditions with the present
 climate and angle of repose.  Local failures in these
 soils  resulting from lateral support removal  in road
 cuts indicate a sensitivity to disturbances on steeper
   Eight  mass  wasting  map units  were  created  to
 describe the variety of slope processes, deliverability
 to streams, relative  hazard  ratings, and timber man-
 agement  implications related to slope stability in the
 basin. The only mass wasting unit with a high hazard
 rating represented approximately 40 percent of the
 mass  wasting in the watershed.  The mass  wasting
 occurred  on  streamside slopes adjacent to middle and
 lower reaches of both forks of Taneum Creek, unre-
 lated to management activities.

   An important factor controlling  dynamics of coarse
 sediment  is the amount of large wood in the stream. It
 is important in retention of coarse sediment (gravel)
 in steep  tributaries.  In  the low-gradient mainstem
 channels, large wood determines the location of many
 gravel bars, and also tends to reduce gravel transport
 distance.  Bankfull flows  appear  to be sufficient  to
 entrain gravel where the  flow  is  unimpeded  by
obstructions.  Where  large wood is abundant, eddies
develop and  gravel  bars tend to form in  sheltered
portions of the bed.
 The majority of riparian areas in the WAU are made
up  of dense mature conifers  that provide natural
amounts  of shade and large woody debris input to
stream  channels.  The  major  exceptions  include
approximately 350 m (1,100 feet)  of North Fork
Taneum  Creek above its confluence with Lookout
Creek, and substantial portions of Lookout  Creek,
Peaches  Creek,  and upper  Butte Creek, where  the
riparian canopy has been removed by timber harvest
in the last ten years. Short and long-term large woody
debris recruitment potential is greatly reduced. These
reaches are also below minimum  shade levels and,
according to  documentation, have exceeded tempera-
ture standards. Historic selective timber harvest  and
encroachment of roads have also reduced the amount
of shade  and  potential for large woody debris recruit-
ment in the lower portions of both forks of Taneum
Creek. Some  areas in the middle reaches of both forks
have less shade  and woody  debris  recruitment
potential than expected due to natural factors such as
stream meandering on the  North Fork  of Taneum
Creek and a wet  meadow  on South Fork Taneum
  One final variable that can be influenced by timber
management  is  the  duration and  volume of peak
streamflows.  Removing vegetation can increase the
amount of water available  for  runoff and increase
peak flows that  shape the channel morphology.  The
relative  importance  of rain-on-snow  versus spring
snowmelt generated peak flows in this watershed was
difficult  to  assess because of the limited data on
streamflow. Peak streamflows generally occur in the
spring with the melting of the snowpack. Large rain-
on-snow floods  occur infrequently, probably on the
order of once every  10 to 15 years. The Butte Creek
sub-watershed was  the  only area where observed
changes  in the stream channel  could  potentially be
linked to increased peak flows. Since the majority of
runoff in east-side Cascade streams occurs during the
spring, studying changes in seasonal runoff(i.e., the
timing and  volume  of snow-melt generated  flows)
due to timber harvesting is probably  more relevant
than assessing potential changes in rain-on-snow run-
off. The data and/or methods for  such an analysis,
however, are unavailable at present.

                                                                                      Watershed Analysis
 A  number of  prescriptions were  developed  to
improve the condition of the watershed and to avoid
potential problems  in the future.  The prescriptions
not only address forest practices, but identify moni-
toring and restoration opportunities as well.
 A 5-year road improvement and maintenance plan is
being  developed  to  reduce  the  amount  of  fine
sediment entering streams to 50 percent of natural
background inputs. The improvement  work includes
the placement of additional culverts, revegetation of
cutslopes, and abandonment of roads. New roads can
be built provided that sediment production from all
roads in the watershed is reduced to specified annual
target levels.
  Riparian  areas without enough trees  will have 80-
foot no-entry buffers. Certain segments of North and
South  Fork Taneum Creek will have  wider buffers
likely  ranging from 80 to 200 feet to accommodate
natural stream meander.  In most cases, these wider
buffers will be no-entry zones, but limited partial-cut
harvest will be allowed in some areas.
  Timber harvest within  the Butte Creek watershed
will be deferred for three years. For the following two
years,  further harvest will be limited  to partial cuts
less than  120 acres or clearcut harvest of less  than
50 acres.  Monitoring of stream bank erosion  and
recovery of the riparian area will help determine the
amount of allowable harvest after the initial five-year
  Since mass  wasting tends to be a very localized
problem,  timber harvest  or road  construction  in
delineated  areas will require additional review  by a
qualified geologist.

  Monitoring in the Taneum Creek watershed cur-
rently  consists of  trend and validation  monitoring
with other stakeholders through the Yakima Resource
Management  Cooperative.  Annual  McNeil stream
sediment sampling  is used to assess the amount of
fine sediment in spawning gravels. High levels of fine
sediment in spawning gravels reduces the viability of
eggs in fish redds. A road sediment monitoring study
has been initiated to determine the  amount of  sedi-
ment produced from roads  under east-side Cascade
conditions. Sediment production from roads with dif-
ferent soil types, road surfacing, and traffic levels are
assessed at eight different sites within  the watershed
area. A streamflow monitoring study was also initi-
ated with two  stream gages installed in  the South
Fork Taneum Creek to assess potential  changes in
streamflow  as timber  harvest occurs  within  the

  Channel conditions are being assessed over time in
Butte Creek  using photography, permanent channel
cross-section measurements, and annual stream sur-
veys.  Also,  the amount of  in-channel large woody
debris   in  North  Fork Taneum  Creek  is being
monitored to identify  potential watershed restoration

Watershed Analysis and
  Restoration activities in forested watersheds such as
road abandonment, riparian  plantings, and in-stream
structure placement have generally occurred on a site-
specific  basis  without explicitly  considering  the
watershed context. A given site is typically consi-
dered deficient in certain elements and the objective
of the restoration project is to provide or increase the
productivity of those elements. Unfortunately, few of
these projects end up as effective long-term solutions
because they fail to consider natural landscape distur-
bance processes. Many of the projects either fail to
survive during floods  or are not implemented in
locations that provide maximum benefits.
  Watershed  analysis  provides a unique opportunity
to identify "natural"  disturbance regimes for  given
elements in a landscape and  provide guidance on pri-
oritizing areas of the landscape where restoration will
provide the greatest benefit to resources of concern.
  Washington state watershed analysis is the  only
process at present that requires a rigorous methodol-
ogy for identifying local physical process that affect
channel morphology.  Local  information on channel-
forming processes is  critical for prioritizing restora-
tion projects as well  as identifying the potential  for
long-term success of the  project.  Data collected at
broader  scales as  is  done in  the  federal watershed
analysis process may be useful for identifying general
areas  where restoration is desirable, but typically do
not provide the necessary information to make deci-
sions on the long-term success of restoration projects.
For example, if pool  frequency in a stream reach is
considered deficient  for  good quality fish habitat,
data is needed  on the geomorphic context of the area
(i.e.,  the  natural  geologic  setting),  current  and
potential future sediment supply to the area, flooding

 Watershed Analysis
        Resource of
        Threats to
         Confidence in
         Time Frame
         for Recovery
         Action Plan
         Objective with
         Time Frame
Spring Chinook spawning habitat in Segments 12 and 13 of the Yakima River
is degraded by fine sediment, lack of holding pools, and potential scour during
high flows.              	
1.  Landslides from roads built prior to 1980 directly into the Yakima River
    and from debris flows in Cabin Creek.
2.  Surface erosion from extensive road network in Cabin Creek.
3.  Reduction in large woody debris from power lines in riparian area.
4.  Large proportion of watershed is hydrologically immature.
1.  High
2.  High
3.  Moderate
4.  Low
1.   Years for sediment to route through system.
2.   Indefinite, continual input of fines as roads are used.
3.   Indefinite for large woody debris input unless power lines relocated.
4.   Years for hydrologic maturity
 1. Road 2520 and 2550 will be abandoned with culverts removed and sidecast
    material pulled back.
 2. Road 2500 will have any sidecast on slopes >60 percent pulled back.
 3. Drainage problems and sediment delivery to streams will be evaluated and
    addressed for the entire road network.
 4. Alternative silvicultural methods for maintaining vegetation under power
    lines will be investigated with the power company.
 5. Placement of large woody debris to increase holding pool capacity will be
The action plan will reduce sediment delivery to Segment 12 and 13, but
measurable improvement in percentages of fines in gravels is not expected for
10 years or more. Short- and long-term (-50 years) large woody debris input
will increase with silvicultural treatments and placement of wood in-channel.
       Figure 2. Example Restoration Action Plan

 regime, and riparian vegetation conditions before any
 decisions  are  made on  restoration  projects.  Such
 information is not collected or put in the context of
 watershed processes for  either the federal  or Idaho
 watershed analysis processes.

  Watershed analysis can  be a useful tool for guiding
 restoration activities, but many obvious site-specific
 problems exist that can be addressed without expen-
 sive, time-consuming landscape analysis. Data gener-
 ated  from  watershed analysis  is  most  useful  for
 developing  effective,   long-term   watershed-scale
 restoration  plans  and for guiding  site-specific in-
 stream  projects.  Depending  on  the  long-term
 objectives  of the landowner  or agency that is initi-
 ating watershed  analysis, data collection  can   and
                            should be modified to be the most practical and use-
                            ful for guiding decision-making.

                              Five important elements are necessary to prioritize
                            and help implement restoration activities:

                            1.  Identification  of resources  of greatest  concern
                                and their conditions.

                            2.  Identification  of threats to resources and to the
                                potential projects.

                            3.  Evaluation of the  linkages between the threats
                                and the resources

                            4.  A clear statement of the objectives of the project.

                            5.  Identification  of the  time  frame for  expected
                                recovery of the resources.

                                                                                    Watershed Analysis
 These elements provide a systematic approach  to
restoration that is critical for evaluating the costs and
benefits of particular projects. Figure 2  contains an
example restoration action plan.

 Providing  for long-term  restoration  of  aquatic
resources requires an analysis of landscape elements
at a local scale.  Watershed analysis that is scientifi-
cally rigorous and has direct  links to management
decision making can provide useful data for planning
and implementing long-term restoration projects.
  Due  to  the long-term nature of  most recovery
efforts, a systematic  approach to restoration is also
critical. Watershed  analysis that provides data on
watershed process at a local scale can ensure that
restoration efforts will be successful in the long-term.
Active restoration management will be necessary  to
speed the recovery of many aquatic systems, but will
need to rely on good local data collected in the con-
text of landscape processes. Finally,  monitoring  of
restoration projects and the recovery of resources will
provide a  critical feedback mechanism for future
restoration efforts.

Literature Cited
Agee, J. 1994. Fire and weather disturbances in ter-
  restrial ecosystems of the eastern Cascades. Gen-
  eral  Technical Report PNW-GTR-320. U.S.D.A.
  Forest Service, Pacific Northwest Research Station,
  Portland, OR. 52 pp.
Bilby,  R.  E. 1984.  Removal of woody debris may
  affect stream channel stability. Journal of Forestry
   10: 609-613.
Bisson, P. A.,  R. E. Bilby, M. D. Bryant, C. A.
  Dolloff, G. B. Grette, R. A. House, M. L. Murphy,
  K. V. Koski,  and J. R. Sedell. 1987. Large woody
  debris in forested streams in the Pacific Northwest:
  past,  present, and future,  pp. 143-180 In  Pro-
  ceedings,   Streamside   Management:   Forestry-
  Fisheries Interactions.  Edited by  E.G.  Salo and
  T.W.  Cundy.  University of Washington,  Seattle,
Caldwell, J. E., K. Doughty, and K.  Sullivan.  1991.
  Evaluation  of downstream temperature  effects  of
  Type  4/5  waters.  Timber/Fish/Wildlife  Report
  TFW-WQ5-91-004. Washington  Department  of
  Natural Resources. Olympia, WA. 71 pp.
Chapman, D. W.  1988. Critical review of variables
  used to define effects of fines  in redds of large
  salmonids. Trans. Amer. Fish. Soc. 117: 1-21.

Coffin, B. A. and R. D. Harr.  1992. Effects of forest
  cover on volume of water delivery to soil during
  rain-on-snow.   TFW-SH1-92-001.    Washington
  Department of Natural Resources, Forest Practices
  Division, Olympia, WA. 118 pp.

Idaho Department of Lands. 1995. Forest Practices
  Cumulative Watershed Effects Process  for Idaho.
  Boise, ID.

Lisle, T. E. 1989. Sediment transport and resulting
  deposition in spawning gravels, North Coastal Cali-
  fornia. Water Resources Research 25:1303-1319.
Lisle, T. E. and S. Hilton.  1992. The yolume of fine
  sediment in pools: an index of sediment supply in
  gravel-bed  streams.  Water  Resources  Bulletin
MacDonald, L. H., A. W. Smart, R. C. Wissmar.
  1991. Monitoring guidelines to evaluate effects of
  forestry  activities  on  streams  in  the  Pacific
  Northwest and  Alaska.  EPA 910/9-91-001.  U.S.
  EPA. Seattle, WA. 166 pp.
McDade, M. H., F. J. Swanson, W. A. McKee, J. F.
  Franklin, and J. VanSickle. 1990. Source distances
  for coarse woody debris entering small streams in
  western Oregon and Washington. Canadian Journal
  of Forest Research 20:326-330.
Megahan, W. F. 1983. Hydrologic effects of clearcut-
  ting and wildlife on steep granitic slopes in Idaho.
  Water Resources Research 19:811-819.
Murphy, M. L. and K. V. Koski. 1989. Input and de-
  pletion  of  woody debris in  Alaska  streams  and
  implications for stream side management. North
  American Journal of Fisheries Management 9: 427-
Peterson,  N. P., A. Hendry, and T. P. Quinn. 1992.
  Assessment  of cumulative  effects on salmonid
  habitat: some suggested parameters  and target con-
  ditions. Timber/Fish/Wildlife Report TFW-F3-92-
  001. Washington Department of Natural Resources.
  Olympia, WA.
Regional  Interagency Executive  Committee. 1995.
  Ecosystem Analysis at the Watershed Scale, Fed-
  eral Guide for Watershed Analysis,  Version 2.2.
  Portland, OR. 26 pp.

 Watershed Analysis
  Robison, E. G. and R. L. Beschta. 1990. Identifying
   trees  in  riparian areas that  can provide  coarse
   woody debris to streams. Forest Science 36: 790-
  Sidle, R. C.,  A. J. Pearce, and  C. L.  OLoughlin.
   1985.  Hillslope  stability and  land  use.  Water
   Resources Monograph Vol. 11.  AGU,  Washington
   D.C. 140 pp.
  Sullivan, K., J. Tooley, K. Doughty, J. E. Caldwell,
   and P.  Knudsen. 1990.  Evaluation of  prediction
   models and characterization of stream temperature
   regimes   in  Washington.  Timber/Fish/Wildlife
   Report  TFW-WQ3-90-006.  Washington  Depart-
   ment of Natural Resources. Olympia, WA. 224 pp.
  Swanson, F. J., L. Benda, S. Duncan, G.  Grant,  W.
   Megahan, L. Reid, and R. Ziemer. 1987. Mass fail-
   ures and other processes of sediment production in
   Pacific   Northwest  landscapes,   pp. 9-38   In
  Proceedings,  Streamside  Management: Forestry-
  Fisheries Interactions.  Edited by  E.O. Salo and
  T.W.  Cundy. University of  Washington,  Seattle,
Swanston, D. N. and C. T. Dymess. 1973. Stability of
  steep land. Journal of Forestry 71: 264-269.
U.S. Army Corps of Engineers. 1956. Snow hydrol-
  ogy - summary report of the snow investigations.
  North Pacific Division. Portland, OR. 437 pp.
U.S. Department of Agriculture.  1994.  A  Federal
  Guide for Pilot Watershed Analysis, Version 1.2.
  Portland, OR. 202 pp.
Washington  State  Forest  Practices  Board. 1995.
  Standard Methodology  for Conducting Watershed
  Analysis, Version 3.0. Washington Department of
  Natural Resources. Olympia, WA.

                          Prioritization of Salmonid  Habitat  Problems
                                            and  Habitat Restoration Projects
                                                    in an  Urbanizing  Watershed
                                                                                 Kathryn Meal1
                                                                                Mary Harenda1
  An interdisciplinary design team identified and pri-
oritized salmonid  habitat deficiencies  in approxi-
mately 14 miles of stream and riparian corridor in the
Bear/Evans Creek watershed, King County, Wash-
ington. This watershed supports regionally significant
runs of sockeye (Oncorhynchus nerkd), chinook (O.
ishawyisha), and coho (O. kisutch) salmon, and steel
head (O.  mykiss) and cutthroat (O.  clarki) trout in a
rapidly urbanizing area northeast of Seattle (KCSWM
1990). A two-step process was designed to first iden-
tify, locate, and rank the habitat  problems in the
stream system; and second, to apply financial, logisti-
cal, permitting, construction, and landowner-imposed
constraints  to define restoration projects for design
and  construction.  Habitat  restoration  projects  are
undertaken only with the full cooperation of the land-
owner and are entirely at public expense.
  The problem ranking  process involved  walking
streams and documenting physical habitat conditions;
analysis of aerial photos and map overlays; and col-
lecting land ownership information  on all streamside
parcels. Spreadsheet software was  used to organize
the information and perform simple sorts and  calcu-
lations.  Principal  categories  of  habitat problems
included:   fish    passage   barriers,    streambank
degradation, poor instream habitat diversity, and lack
of riparian  vegetation (Johnson and Caldwell  1994).
The team developed prioritization criteria to reflect
pertinent habitat issues and the differing characteris-
tics of the four streams of interest  in the watershed
(Entrance et al. 1994). Prioritization criteria,  in the
form of questions, were applied to each habitat prob-
lem and stream (Table 1). Question scores for each
habitat problem were summed yielding a "tributary
score" representing the relative severity  of a particu-
lar habitat problem in each stream. Tributary  scores
were assigned to individual reaches within a stream
when  distinctions  in habitat problem characteristics
were observed among reaches. Prioritization  scores

1 King County Department of Natural Resources, Surface Water Management Division, 700 Fifth Avenue, Suite 2200,
Seattle, Washington 98104.
for each land parcel were derived by summing the
products of the  tributary  scores and stream reach
lengths of habitat problems observed on each parcel
(Table 2). The resulting "land parcel habitat scores,"
reflecting the relative severity of habitat problems on
each parcel, were ranked to determine priority parcels
for stream restoration in the  watershed.  Raw scores
(incorporating tributary scores but not weighted for
stream length) were referenced as indicators of the
number  of problems on  a parcel. Before  ranking
potential project sites, it was necessary to group par-
cels into clusters that defined a feasible  project
extent. Clusters were kept as small as possible  and
were formed by grouping: 1) contiguous parcels with
the same owner  and land use; 2) parcels  directly
across the  stream from each other;  and 3)  parcels
where problems   could not  be addressed  without
involvement of adjacent parcels. Land parcel habitat
scores were summed to yield cluster scores, which
were adjusted to  ensure that instream diversity prob-
lems were  not counted twice for parcels across the
stream from each other.  Cluster scores were ranked to
determine priority sites  for stream habitat restoration
on each of the four streams and in the watershed  as a
  A secondary set of criteria, which considered finan-
cial, logistical, landowner, permitting and other con-
straints,  was then applied to determine restoration
projects for design and implementation.
  Implementation of habitat restoration projects began
in 1995 and will continue through 1997. Addressing
barriers to fish passage is the highest priority. Other
components of restoration projects include:  enhanc-
ing instream habitat, biostabilizing eroding  stream-
banks, fencing, and revegetating riparian corridors.

 Prioritization of Restoration Projects
    Table 1. Prioritization Questions and Scoring for Tributary/Habitat Problems
     1.      If nothing is  done,  will  the  problem  get
     2.      How much of the stream in the subbasin is
            affected by the problem? (For example, if a
            potential fish barrier is located downstream,
            the potential  for  blocking a majority  of
            upstream areas exists.)
     3.      How many species of salmon or trout using
            the steam are affected by the problem?
            Of the following, how many problems are
            present:  streambank  degradation, lack  of
            instream   diversity,   degraded   riparian
            corridor, fish passage barrier, miscellaneous
            (e.g.,  trash   or  debris,   water  quality
            Has the  tributary  or stream  reach  been
            designated either  a  regionally or locally
            significant resource area by King County?
  Yes = 10 points; No = 0 points

  If >67 percent of stream, score high = 10 points;
  if 33  to 66 percent, score medium =  5 points;
  if <33 percent of stream score low = 0 points.
  If the number of species using the affected reach is
  all, score high =10 points; if 50 percent of species,
  score medium = 5 points; if only  1  species, score
  low = 0 points.
  If 5 of 5 problems exist, score 10 points; if 4 of 5,
  score 8 points; if 3 of 5, score 6 points; if 2 of 5,
  score 4 points; if 1 of 5, score 2 points.
  Regionally  significant  score  10 points;  locally
  significant score 5 points; 0 points for others.
   We found that a clear definition of the goals and
 methodology of the prioritization process is impera-
 tive to gather useful, well-organized data and enable
 cost-effective allocation of limited public funds to
 address the  most  pressing habitat problems. Most
 project sites are privately-owned and landowner par-
 ticipation in restoration efforts is voluntary. Some of
 the highest-ranking projects have not been undertaken
 because  of  unwilling landowners.  When  imple-
 menting  restoration  projects,  we  found  that close
 coordination  with landowners is essential to ensure
 that a project  meets  intended goals. Clearly com-
 municating the intent of the project and providing the
 landowner with a vision of the final product, through
 graphic  design illustrations, pictures,  or visits  to
 similar projects is crucial to ensuring good working
 relationships  with landowners  and successful imple-
 mentation of projects.
Literature Cited
Entranco, King County Surface Water Management
  Division, J. F.  Osborn, Caldwell and Associates,
  and Lee and Associates. 1994. Bear, Evans, Cottage
  Lake, and Mackey Creeks: Habitat Problems, Pri-
  oritization, and  Solution  Development.  40 pp. 4
Johnson,  A.  W.  and  J.  E. Caldwell, 1994.  Bear,
  Evans, Cottage Lake, and Mackey Creeks Habitat
  Assessment.  Prepared  for  Entranco  and  King
  County  Surface Water  Management  Division.
  17 pp. + app.
KCSWM. 1990. Bear Creek Basin Plan, King County
  Surface Water Management Division. Seattle, WA.

                                                                        Prioritization of Restoration Projects
Table 2. Formula for Land Parcel Habitat Scores
                                     Land Parcel Habitat Score =
    Tributary Riparian Score x (Riparian Disturbance Zone Weight x Reach Length of Riparian Disturbance
                      Tributary Bank Erosion Score x Reach Length of Bank Erosion
                                Riparian Shade Score x Total Reach Length
                        Tributary Fish Barrier Score x Reach Length of Fish Barrier
                    Tributary Instream Diversity Score x Reach Length of Poor Diversity
                                           all divided by 100
  The formula components are defined as:
  Tributary  Riparian Score = a number between  0 and 50, which weights  riparian problems for each
  Tributary  Bank Erosion Score = a number between 0 and 50, which weights bank erosion problems for
  each tributary.
  Tributary  Fish Barrier Score = a number between  0 and 50,  which weights fish passage problems for each
  Tributary  Instream Diversity Score = a number between  0  and  50, which weights instream diversity
  problems for each tributary.
  Riparian Shade Score^ = either the number 2, which represents <50 percent shade for at least 200' of
  stream; or the number 0, which represents >50 percent shade for at least 200' of stream.
  Riparian Disturbance Zone Weight^) = a percentage assigned to disturbance zones 1 through 5.
                  Disturbance Zone              Riparian Condition Weight
                  1  - high disturbance                      1.00
                  2  - high to moderate                      0.66
                  3  - moderate                             0.33
                  4  - moderate to low                        0
                  5  - low disturbance                        0
     The score was derived from streamwalk data.
(2)  The weight is based on aerial photo analysis. In the case where more than one disturbance zone occurs on a land parcel,
    each zone length is multiplied by the appropriate weight.

                               Active Management of Riparian  Habitats
                                                                               Dean Rae Berg1
   Abstract: Current approaches to protecting bi-
 otic  diversity  are  passive—let  nature take  its
 course. This approach leaves degraded ecosys-
 tems. If forest and stream processes are disrupted,
 these ecosystems may  not recover for decades.
 Short term project life, as is the case with many
 restoration projects, limit the ability to implement
 long term solutions, such as the recovery of native
 riparian forest structure and in stream popula-
 tions. Only a long term solution, actively managed
 forests and streams, can restore natural processes.
 Forest managers can use  silviculture to create
 benefits for both aquatic  and terrestrial habitats.
 Release of growing space for selected trees allows
 development of larger trees for LWD recruitment
 and increases light to the forest floor and stream
 bottom.  Currently, no systematic design to ac-
 complish these goals exists. An active management
 approach, such as described here, is prescriptive
 and  adaptive  while  providing returns  to  both
 landowners and the riparian ecosystem.

   Forests supply logs to streams, critical elements of
 the typical step  pool morphology (Grant et al 1990).
 Logs, or large woody debris (LWD), set up pools and
 help  retain sediment in  low order streams (Heede
 1985, Keller and Swanson  1979); LWD provides the
 matrix for the capture of fine sediment and small de-
 bris (Megahan 1982 Bilby 1981). Li addition to sedi-
 ment retention, the logs play an important role in both
 aquatic (Bisson  et al. 1987, Bisson et al. 1982) and
 terrestrial  (Franklin et al. 1996, 1986) habitat struc-
 ture. In large rivers, riparian forests are vulnerable to
 the risk of large flood flows; here islands are often
 only vegetated for decades  (Swanson and Lienkaem-
 per 1982). But  in the  upper reaches  where salmon
 spawn and rear,  wood is a primary building block for
 the complex habitat structure  (Sedell et al.  1989,
 Lienkaemper and Swanson 1987, Harmon et al.  1986,
 Grette 1985, Bryant 1983, Triska et al.  1982).
 A mix of both deciduous and conifer forest is typi-
cal on the fresh geomorphic surfaces left from floods
(Agee  1988,  Oliver  and Hinkley  1987). In many
places where early serai hardwood species have colo-
nized, such as off-channel pools, habitat is very pro-
ductive  for over-wintering  salmon  (Heede 1985,
Cederholm and Scarlet  1982). Riparian  forests are
species diverse (Nilsson et al. 1989)  and should be
protected and cultivated for the functions they pro-
vide. Hardwood dominated forests do  provide neces-
sary functions and have always been present along
streams  and rivers in the Pacific Northwest (Naiman
et al. 1992).
 Rivers change in  many ways as they move down-
stream from the headwaters to the sea  (Vannote et al.
1980). Our streams originate in the alpine zone and
flow through the upper reaches of  the watershed
where landform influences  control  the channel, and
forests supply wood and leaf litter. The valley floor
begins to widen and the river begins to develop ter-
races. At higher elevations the valley walls are steep,
and cold runoff has great force as it falls through the
upper watershed (Naiman et al.  1992, Vannote et al.
1980). Headwaters  are biological storehouses for the
food webs that develop downstream. Forests adjacent
to  rivers and streams transpire and process runoff.
Intricate webs of tree roots form  along the banks,
anchoring topsoil.
 In-channel LWD attrition (e.g., decay, mobilization,
extraction) increases the difficulty of reestablishing
conifers in degraded riparian forest stands. Because
large jams are not present to protect bars from flood
flows, the bars cannot revegetate.  Also, hardwoods
colonize many areas and compete with the near-term
conifer  regeneration  and subsequent recruitment of
conifer logs to the  stream channel.  The size of func-
tional wood increases with channel gradient and with
drainage area (Bilby and  Ward 1990).  In  steeper
streams, much of the functional wood was recruited
prior to  logging (Potts  and Anderson  1990, Grette
1985). For these reasons, off-channel habitat in the
low gradient mainstem may be slow to recover.
 Silvicultural Engineer, 15806 60th Avenue West, Edmonds, Washington 98026.

                                                                              Riparian Habitat Management
Structure in Streamside Forests
 Elements of the streamside environment include the
active channel, the associated  flood  plain, and the
zone of influence (after Cummins 1974). The active
channel and floodplain  are  easily recognized  and
have clear biological and physical limits to their ex-
tent. The  zone of influence is broader and encom-
passes the shading, litter fall,  and sources of large
woody debris (McDade et al. 1990). Riparian forest
structural attributes are listed below.

  •  Live, vigorous trees and large  snags as the
     source of LWD

  •  Large  down logs  for  habitat formation  and
     sediment retention

  •  Suspended and embedded wood  in the stream
  •  Vertical distribution of foliage
  •  Many sizes and species of trees in the adjacent
 Longitudinal Changes in Ecological
  In the  upper watershed the nearshore riparian  is
 dominated by brushy early serai trees and shrubs, be-
 cause the force of the water reworks the narrow val-
 ley floor annually.  Trees establish very rarely. The
 narrow stream  channels are choked  with woody
 debris that holds what little sediment is stored. The
 valley wall confines some stream reaches and limits
 development of forests near the active channel.
  As the river widens, terraces begin to build soils
 and the forest produces large logs. The  adjacent for-
 ests generate woody debris that is large enough  to re-
 sist the stream flow (Bilby and Ward 1990); but  in
 large rivers riparian forests are vulnerable to the risk
 of floods. Rivers move large amounts of material and
 rework valley floors. The power of these rivers  is ca-
 pable of drastic modification of valley bottom ripar-
 ian forests.
  Streamside forests protect streambanks  and chan-
 nels because they add roughness to the channel, re-
 duce the  flow  of flood  waters,  and stabilize  soils
 (Heede 1982). Riparian soils vary widely in both dis-
 tribution  and productivity. Scour and deposition are
responsible for some of the spatial variability in site
quality. Site capability for tree  growth is degraded
when fine, nutrient-rich sediments are leached out  of
coarse-grained gravel bars. Elevated drainage capac-
ity  of gravely  soils further reduces  their  nutrient
holding capacity. Sediment deposition  in the form of
silts, and, and gravel from upslope, create sites that
are well drained and productive. However, some an-
aerobic soils are present along streamsides and can
result in reduced productivity.

  Functions of lateral connections to streams include
the input of litter and debris (to drive food chains and
provide channel structure) and the hyporheic zone,
which functions as  both  habitat to stream  inverte-
brates and as a fine filtering system for ground water
recharge (Stanford and Ward 1988). Forest structure
influences the quality of these lateral functions. For-
est stands have  some effect in altering  the  channel
morphology  but they  do  influence   the  hyporheic
zone, a deep, wide, underground flow. The space be-
neath the forest processes  nutrients through the  fil-
tering (e.g., through roots, gravels and sediments) and
biological activity of the flora and fauna  of the inter-
stitial spaces (Stanford and Ward 1988).

Passive and Active Management
  Where stream systems are degraded, the inherent
resilience in  forests  will  eventually  succeed in
recolonizing  the area with large dimension conifer
forests. Acting  passively may not recover the forest
functions in time to be of use to declining fish popu-
lations, however (CSE  1995). An active  approach to
riparian management asks, "Can we grow forests that
sufficiently  supply the proper  size and amount of
wood?" The  alternative is to let the  stands  proceed
through  successional  pathways  toward  functional
stand structure,  perhaps at a  much slower rate. A
managed forest could feasibly develop beneficial for-
est structure in  a much shorter period  than leaving it
to the chance circumstances of natural succession.
Active management  may  accelerate development of
full canopy cover by planting a mix of desirable spe-
cies (e.g., red cedar, spruce, Douglas-fir,  black cot-
ton wood).
  Relationships  between  forest  structure and  in-
stream  channel  morphology need to be  set  up as
working hypotheses  to be tested. If we  are going to
manage this apparently unmanageable ecosystem then
we need to establish simple empirical  models for the
different forest and stream types (e.g., McDade et al.
1990, Bilby and Ward  1990, Mitsch  and  Jorgenson

 Riparian Habitat Management
  An important element of active management is eco-
 nomics. From the forest managers' standpoint, what
 does riparian recovery  cost? We can do better than
 break-even while creating a residual forest capable of
 sustaining its own LWD supply, perhaps even begin-
 ning to restore some of the more subtle functions of
 the riparian forest? The Pacific Northwest is blessed
 with highly productive sites and  high value timber
 species. But  even so our regional forest economics
 are subject to many of the same worldwide  influ-
 ences,  interest  rates, amount  of early  investment,
 growth and recovery rates, and risk of disturbance.
 The economic feasibility  of various silvicultural sys-
 tems can  be evaluated relative to these  parameters
 (Berg 1995).
  When viewed at the stand level, it is almost  over-
 whelming to think that restoration might be recom-
 mended on thousands of  miles  of steams in  the
 Pacific Northwest: Hundreds of thousands of acres of
 degraded  streamside forests. Riparian management
 can improve the function of these streamside forests
 and  restore many of  the habitats  associated with
 streams. Large diameter trees, snags, and down logs
 are the structure most  limited in supply in  terms of
 the LWD  component  of stand  structure. This is
 something we know intuitively, yet it still requires an
 extensive commitment of research effort.

 Basin Scale Silviculture
  Basin scale silviculture is  the next frontier as we
 wrestle with definitions, objectives,  and the increas-
 ingly demanding goals of society (Oliver et al. 1992).
 Watershed planning  and analysis  identify the condi-
 tion of streams and the associated/dependent  re-
 sources such as salmon populations.  Short  term
 projects, as is often the case with restoration projects,
 limit the ability to  implement  long term  solutions
 over large areas, such as the recovery of native  ripar-
 ian forest  structure and survival of instream popula-
 tions.  Basin  level designs  should  include project
 implementation and  monitor follow through, which
 will require longer time frames.

  Streams and the associated riparian zone are  struc-
 tural elements in the larger landscape that is con-
 trolled  more by the shape of the land (Trotter  1990,
 Swanson et al. 1982). Patch size and stand distribu-
 tion over the landscape interact at a  watershed  level,
 and the arrangement  of various stand structures may
 affect   landscape  functions, such as  groundwater
 (Stanford and Ward  1988) and surface water  flow,
stream temperatures (Sullivan et al. 1987), or wildlife
corridors (Raedeke 1988). Vegetation indicates the
condition  of a stream or stand (Schoonmaker and
Mckee 1988); the species and age indicate time or se-
verity  of  the last disturbance (Oliver and Larson
 Collins et  al.  (1994) describe how riparian forests
change through time in a watershed, focusing on cur-
rent  stand conditions  and their relationships to fish
habitat. Washington's  standard methodology  for con-
ducting watershed analysis (WFPB 1993) identifies
three broad  classifications of stand types that can be
related to the stream shading and debris recruitment
functions of the riparian zone. Collins et al. (1994)
added a third role of the riparian forest, resisting bank
erosion, and categorized forest stand types to  a higher
resolution than State methodology (WFPB 1993) so
restoration recommendations  could be more  directly
interpreted from the data (stand types  and ratings are
listed in Table 1).

Riparian Management in the Pacific
 Loss of riparian forests has exacerbated  channel
widening  during floods, increased  stream tempera-
tures in some reaches, and diminished opportunities
for woody debris recruitment. Priority areas are refu-
gia and low  gradient mainstems, which potentially are
the most productive habitats  for steelhead and coho
salmon. Not all  streams  need rehabilitation.  Func-
tional riparian zones should be recognized and main-
tained and yet, in light of all the empirical evidence
presented above, a relatively small amount of our
collective wisdom is used in the design of recovery
 The  extensive  lower elevation, floodplain forests
have been gone from most Pacific  Northwest rivers
for decades  (Sedell and Froggatt 1984). Stream pro-
tection efforts have  largely focused  on  salmonid
habitat; enhancing parameters such as water quality,
shade, LWD, and litter.  Many  basic biological re-
quirements  of other instream species are unknown.
Little consideration is given to  habitat composition,
location, larger scale  connectivity, persistence, and
 Alder dominated riparian  zones  of the Pacific
Northwest are prime examples of second-growth for-
ests  lacking in structural and biological diversity. Past
practices  have left many streams in  dire  need  of
structural enhancement and riparian forest

Table 1. Stand type classes identified in the assessment of riparian forest stands
Stand Type
Recently disturbed (landslide, harvest, gravel bar);
unvegetated or herbs and grasses
in the Deer Creek basin, (after Collins et
               Vegetated - small diameter (<6") trees and shrubs,       LOW

Sapling         6"-12" average DBH (tree diam. @ 4.5 ft. above
   Conifer - C   ground)
   Mixed - M         > 70% conifer                                MOD
   Hardwood -        mixed conifer/deciduous                        MOD
                    > 70% deciduous                              MOD

Pole            12"-18" average DBH
   Conifer - C        > 70% conifer                                HIGH
   Mixed - M         mixed conifer/deciduous                        HIGH
   Hardwood -        > 70% deciduous                              HIGH
  Conifer - C
  Hardwood -
>18" average DBH
a. Moderate in western hemlock zone (below 2000 feet elevation), high in silver fir zone (above 2000 feet elevation)
where average DBH is 12"-18" in mature and old-growth stands.

b. This function is dependent on the size of the stream. In smaller reaches (especially moderate gradient tributary
reaches) mature conifer trees help resist bank erosion by floods. In larger reaches and the low gradient mainstems, large
floods may still significantly erode banks even with mature conifer stands established. However, in the low gradient
mainstems, banks without large conifer have been more severely eroded than banks with large conifer trees.

 Riparian Habitat Management
 restoration. Before the establishment of streamside
 buffers, conifers were systematically removed from
 riparian areas and streams were cleaned  of  woody
 debris. Frequently, stream channels were used to skid
 logs, if not splash damned  and used to drive logs
 down.  The resultant plant community  is generally
 dominated  by  red  alder  and  shrubs,  such  as
 salmonberry,  which in some areas may persist as a
 plant community until a disturbance or intervention.
   In hardwood dominated and debris  poor forests,
 stream channels are  often also limited  in structural
 diversity; lacking a mixture of wood-formed pools,
 known to be important fish habitat. Engineered in-
 stream structures  (e.g., gabions, logs  placed  with
 heavy  equipment ) improve fish  habitat only for a
 short time.  A more sustainable, long-lasting, and less
 expensive method is to establish  or release  riparian
 zone conifers. This approach allows the forest to pro-
 vide recruitment of decay resistant woody debris.

 Recovery of Riparian Forests
   The  economic  motivation  for restoration  is ele-
 gantly  simple; Northwest  salmon streams  once sup-
 ported a thriving industry. Now stocks are declining
 and the source of fishing revenues is  questionable
 (CSE  1995).  Riparian  silviculture  restores native
 watershed  functions  responsible  for  forming and
 maintaining stream habitats  for a variety  of  aquatic
 species, but in a sequence that favors  long-term re-
 covery of the seasonal salmon runs.
   A second focus of restoration improves  habitat re-
 lated to riparian forests and recruitment of wood from
 those  forests. By planting  fast-growing trees  with
 dense  root  mats (e.g.,  cottonwood), silviculture
 speeds recovery of riparian shading and bank stabili-
 zation, especially on low terraces. Recovery of decay
 resistant wood recruitment by various riparian silvi-
 cultural treatments are longer term.

   The goal is to recover the mature conifer forests that
 generally dominated  riparian forests in  the past and
 provided shade, woody debris to streams, and root
 strength to erodable banks. Silvicultural treatment can
 restore the functions of a mature riparian forest using
 an active approach. Wood  delivered to channels aug-
 ments the effects of riparian vegetation on  stabilizing
 stream banks and bars. Placing wood in channels can
 improve habitat in key areas (e.g., pools for  rearing
 Riparian silviculture requires an inventory of stand
and basin riparian conditions. The key to manipulat-
ing structural diversity in riparian areas is the estab-
lishment of conifers over a  long period  of time. If
understory conifers are present, then they may simply
be released. Under planting of conifers may enhance
the opportunity to establish multiple cohorts of trees.
Thinning to wide spacings also allows regeneration of
many species beneath the overstory, even moderately
shade intolerant Douglas-fir.  Stand dynamics will be
particularly  important to monitor in riparian areas
where aggressive, early  serai species like red alder
and salmonberry dominate and are likely to out com-
pete riparian conifer plantations.

 Target species  composition for planting  to meet
long-term LWD targets is  mixed Douglas-fir  and
western red cedar, although  higher elevation stream
reaches are usually dominated by the silver fir zone
(Franklin  and Dyrness 1973) with hemlock and  yel-
low cedar.
 Where stream temperatures are  elevated and  root
strength loss has  been significant, primarily  in the
low-gradient mainstem reaches, short-term objectives
include establishment of  fast-growing  cottonwood
stands to  rapidly recover a dense root mass and to
shade the  stream channel. As these stands mature and
riparian functions are established, gradual conversion
through active management to conifer stands can  pro-

 Sapling  stands of  hardwood,  or mixed hardwood
stands with a conifer understory (e.g.,  hemlock, ce-
dar), may offer opportunities to recover some of the
costs associated with planting conifer species through
commercial thinning. Sapling conifer  stands can be
thinned to increase  growth  rates  and recover large
woody debris recruitment.

 Restoration of riparian  forests  should encourage
planting of species that  both improve bank stability
and provide shade. Given a  historical  channel width
of 40 metres and a tree height of 50 metres, the target
canopy cover  should be 30 percent (Collins et al.
1994).  A short term target may be recover 30 to
40 percent  canopy   cover,  accomplished  through
plantations of fast growing hardwoods (e.g.,  cotton-
wood, alder) and decay resistant conifers  (e.g., red
cedar, Douglas-fir).

 There are  stream  reaches where  the  unvegetated
floodplain extends 200-300 feet from the existing
streambank. Planting sites are not always available at
the  streams edge because  of  the  coarse  boulder/

                                                                              Riparian Habitat Management
cobble substrate. Here a staged revegetation plan can
be developed that attempts to advance the edge of the
existing riparian forest incrementally by decade (e.g.,
plant  the first 25-50 feet from the edge of existing
forest in the first  decade with cottonwood or red ce-
dar; then  advance the next  increment each decade
until the floodplain vegetation is recovered). As the
forest recovers, an active silvicultural program (e.g.,
thinning and planting) may accelerate development of
large  conifer trees. These reaches are a high priority
for recovery.
  High priority stands include those in open and brush
areas,  as well as  newly-formed  and largely unvege-
tated  terraces in low-gradient  mainstem channels.
Restoration options for open  and brushy  areas  typi-
cally include clearing growing space and establishing
mixed Douglas fir and western red cedar, at an esti-
mated cost of $120  to $300 per acre (Table 2). On
unvegetated terraces, initiation of cottonwood is the
preferred option.  Site selection should consider suit-
ability of the selected species to the site, susceptibil-
ity to flooding,  and proximity  to  side channel  or
valley tributary habitats.
  Sapling  stands  of  hardwood,  or  mixed  hardwood
and understory conifer, are the second priority. These
stands may offer opportunities to recover some of the
costs of restoration through commercial thinning be-
fore planting conifer species. Costs range from $275
to $300 per acre  for thinning with no extraction and
planting, and from $125 to $150 per acre for com-
mercial thinning and for active management planting.
Sapling conifer stands are a lower  priority, but they
can be thinned to increase growth rates for large
woody debris recruitment. Thinning alone costs $150
per acre,  but commercial thinning can result  in a
profit of about $600 per acre (Table  2).
  Pole hardwood  or mixed stands can be thinned to
encourage  conifer growth with understory reestab-
lishment through planting of conifer. Costs without
extraction range from $125 to $150 per  acre, costs
with commercial  thinning range from $325 to $350
per acre. Pole conifer stands in the western hemlock
zone (below 2000 feet elevation) can be thinned from
the lower two thirds of the size class (<12" DBH) but
this option is not encouraged because the benefits to
riparian functions are likely to be  small.  Piling and
pole size conifer  stands in the silver fir zone require
no action because they are at or near maximum size
and are dominantly conifer.
Other Stream and Riparian Restora-
tion Measures
  Silvicultural treatment is one method for accelerat-
ing development of a  mature riparian forest  with
many   of   the   functions  previously  mentioned
(Rainville et al. 1985). There are also some opportu-
nities for manipulating wood in channels to augment
the effects of riparian vegetation on stabilizing stream
banks and  bars. These  opportunities are focused in
the low gradient mainstems where  large jams  can
protect newly planted vegetation from floods. Ripar-
ian planting should take advantage of existing jams at
the upstream ends of bars and newly-formed terraces
as a first  priority, but jams may be constructed in
some areas by cabling together existing woody debris
that is scattered across bar tops. Higher terraces with
large jams  at the upstream end, to protect them from
flood flows, are likely to persist in the long-term and
should be addressed first. Additionally,  the value of
off-channel habitats to coho and steelhead indicates
that  sites  providing riparian functions  to  both the
mainstem and off-channels should receive priority.
  Habitat restoration with woody debris is most ef-
fective in small low-gradient channels (Potts and An-
derson  1990). Wood placement is typically used to
create pools for rearing habitat, or,to protect stream
banks from eroding. Placement of woody  debris to
create pools is typically expensive, and usually im-
proves habitat conditions only within a few meters of
the structure. Placement of logs at critical points in-
channel can accelerate  development of pools  and
sediment retention. One must acknowledge the dy-
namic nature of log/pool  location  and  allow for
movement  of logs through the system (Reeves et al
1995). These approaches are temporary  solutions for
maintenance of channel structure. Log placement can
be accomplished a number of ways, including the use
of skyline yarding systems set upon adjacent  cut-
blocks.  This technique, called "back-yarding" (Loren
Kellogg, Oregon State University; personal commu-
nication) has a key advantage of being able to access
the channel without heavy equipment and associated
site degradation. Yarding corridors can be established
at designated points based on critical in-channel con-
ditions in conjunction with standard forest operations.
  Beavers play a significant role in the  riparian eco-
system, by  creating  and  maintaining  off-channel
aquatic habitat.  The primary roles of beaver in sal-
monid ecology are (1) physically creating overwin-
tering habitat for coho salmon by dam  construction,
and (2) helping maintain conifer forests along streams

   Table 2. Silvicultural options for the riparian conditions identified in the riparian assessment, with the target of large conifer in the
   riparian zone for LWD recruitment and shade. All DBH measures are for average of stand, not largest trees.  Costs of stock are based
   on (a) $200/M for cottonwood seedlings and (b) $250/M for cedar and Douglas fir seedlings. Time to objective is based on hemlock,
   Douglas fir or cedar in the Western hemlock zone (< 2000 feet elevation). Where it is relevant, management options and time to
   objective are noted for Silver Fir Zone stands.
Stand type

Open : low bar In
Open: Ugh bar

Open: floodplaln or

Brush : conifer
(dense, small dimen-
sion, low growth

Brush: deciduous

Brush: mixed

Objective and silviculture] activities

-Active diannel; no activity.

-Short term objective: stabilize low terrace to provide shade
and resist bank erosion.

-OPTION 1 : plant in spring with alder or cottonwood depend-
ing on site conditions; 100-200 trees/ac. High risk of failure
due to floods.
-objective: large conifer (LWD recruitment and shade).
OPTION 1 : release planting space; plant cottonwood, cedar.
D. fir at 400 trees/ac; thin at 20 yr.

- OPTION 2: release planting space; plant cedar and Douglas
fir at 200 trees 'ac; no thin.
- objective: large conifer (LWD recruitment and shade).

- OPTION 1 : juvenile spacing (PCT), target of 200 free to
grow ac.

- objective: large conifer (LWD recruitment and shade).

- OPTION 1 : thin deciduous (release growing space) and plant
conifer at 200 trees / ac (100 TPA each of Douglas fir and
- objective: large conifer (LWD recruitment and shade).

- OPTION 1 : release existing conifer by thinning deciduous
(target of 200 conifer/ac); plant to target level if necessary.

24" DBH

30 years to establish
cottonwood (24"

cedar - 90 yr.

70 yr.

100 yr.

hemlock, fir - 70 yr.
cedar - 90 yr.

Time to objective
36 " DBH

SO years to establish
alder (16" DBH).

cedar- 150yr.

100 yr.

130 yr.

hemlock, fir - 100 yr.
cedar -150 years







Planting Stock


S50/acb for 200










Sapling: conifer

Sapling: deciduous


Pole: conifer

- objective: large conifer (LWD recruitment and shade).

-OPTION 1: thin to 1 50-200 tree /ac (no extraction).

- OPTION 2: thin to 1 50-200 tree /ac (commercial thin).

- objective: large conifer (LWD recruitment and shade).

- OPTION 1 : release - no extraction; plant cedar and fir at 200

- OPTION 2: commercial thin, plant cedar and Douglas fir at
200 trees/ac.

- objective: large conifer (LWD recruitment and shade).

- OPTION 1 : release - no extraction, plant cedar and/or
Douglas fir to 200 trees per acre

- OPTION 2: commercial thin deciduous, plant cedar and or
Douglas fir to 200 trees per acre

- objective: large conifer (LWD recruitment and shade).

- OPTION 1 : no action (best option in silver fir zone "old-

- OPTION 2: thin from lower 2/3 of size class (<12" DBH).
no extraction.

- OPTION 3: commercial thin from lower 2/3 of size class
(< 12" DBH).

50 yr.

100 yr.

50 yr.

OPTION 1: 50 yr.(0
yrs in silver fir zone

OPTION 2:30 yr.

OPTION 3 :30yr.

80 years

130 yr.

80 yr.

OPTION I:70yr.
(0 yrs in silver fir zone

OPTION 2: 50 yr.

OPTION 3:50 yr.

-release -
Pole: deciduous

Pole: mixed

- objective: large conifer (LWD recruitment and shade).

- OPTION 1 : no action (best option in silver fir zone "old-

- OPTION 2: thin, no extraction, plant Douglas fir and cedar

- OPTION 3: commercial thin , plant Douglas fir and cedar to

- objective: large conifer (LWD recruitment and shade).

- OPTION 1 : no action (best option in silver fir zone "old-

- OPTION 2: thin, no extraction, plant Douglas fir and cedar
to 200 trees ac.

- OPTION 3 : commercial thin , plant Douglas fir and cedar to

OPTION l:>200yr.

OPTION 2:1 00 yr.

OPTION 3:1 00 yr.

OPTION 1:50 yr.

OPTION 2:30 yr.

OPTION 3 :30yr.

OPTION l:>200yr.

OPTION 2: 130 yr.

OPTION 3:130 yr.

OPTION 1:70 yr.

OPTION 2:50 yr.

OPTION 3:50 yr.

-comm. thin-
logging and
haul; 5 MBF/



-comm. thin-
MBF logging
and haul1, 5
MBF/ ae)




•comm. thin -
(conifer (S
5 MBF/ac)

-comm. thin-
(conifer @
S200/MBF; 5





                                                                             Riparian Habitat Management
by cutting  down  deciduous trees. Managing for a
healthy beaver population may accelerate recovery of
off-channel habitats in low gradient mainstems,  and
may  reduce costs  associated with silvicultural  ma-
nipulations designed to promote large conifer forests.
The risks are that the beaver will either devour plan-
tation stock  or create excessively large impound-
ments. Prescriptions for active management should be
sensitive to beavers and their ecosystem function.
  A riparian reference  stand (RSS) system based on
the riparian conditions (species, age, stocking, valley
form, channel type,  process domain  (Floodplain,
steep cascade), and EPA hydrogeomorphic unit (bog,
hyporheic) could  be implemented as a part  of a re-
gional monitoring program (Bisson and  Raphael
1996). The  Washington  State  watershed  analysis
monitoring module (WFPB 1993) could accelerate
the development of a reference network through  vol-
untary participation  of landowners. The reference
system compliments experimentally  designed treat-
ments for the variety of active management prescrip-
tions (e.g., thinning,  planting, pruning). Controlled
experiments with working hypotheses feed the adap-
tive  management process, advancing the knowledge
of reliable recovery methods (after Rollings 1978).

  Native people have used riparian forests for millen-
nia and yet only  in the last century have we devel-
oped the  technology  to  remove  in  excess  of
80 percent of native forest  cover. With the loss of
these forests comes the loss of function they  had  pro-
vided. Protection  must extend beyond the lower wa-
tershed   where   buffer   strips  have   often  left
dysfunctional hardwood stands, dominated by short-
lived species while the conifers have been harvested,
high-graded for the economic values from the riparian
forest. Often the legacy of the previous forest is all
that remains, as large dimension and rapidly decaying
logs. We are in that twilight where we are even losing
these legacies, large stumps are often the only struc-
ture  that remains  of the native forest. The next cen-
tury  will yield the results of our active management.
We cannot afford to wait passively for the forest and
stream ecosystems to recover while instream species
that depend on riparian forests dwindle to extinction.

Literature Cited
Agee, J.  K.  1988. Successional dynamics in forest
  riparian zones.  In  Streamside Management: For-
  estry and Fishery Interactions. K.J. Raedeke (ed.).
  Institute of Forest Resources, University of Wash-

Berg, D. R. 1995. Riparian silvicultural system de-
  sign and assessment. In the Pacific Northwest Cas-
  cade Mountains.  Ecological Applications. 5,1:87-

Bilby, R. E.  1981. Role of organic debris dams in
  regulating the export  of dissolved and particulate
  matter  from   a  forested  watershed.   Ecology

Bilby, R. E. and J. W. Ward. 1990. Changes in char-
  acteristics and function of woody debris with  in-
  creasing size of streams  in  Western Washington.
  Transactions  of the American Fisheries Society.

Bisson,  P. L. and M. Raphael. 1996.  Riparian eco-
  system research:  Evaluation of different manage-
  ment   approaches  to  protecting  and   restoring
  ecological  processes  in   riparian  zones. Pacific
  Northwest Research Station. Olympia, WA.
Bisson,  P.  A.,  J. L. Nielsen,  R. A.  Palmason,  and
  L. E.  Grove.  1982. A system of naming habitat
  types  in  small  streams, with examples  of habitat
  utilization by  salmonids  during low stream flow.
  pp 62-73 In N. B. Armantrout (ed). Acquisition and
  utilization of aquatic habitat inventory information.
  Proceedings of a symposium held 28-30 October,
  1981, Portland,  OR. Western Division,  American
  Fisheries Society.
Bisson,  P. A., R. E. Bilby, M. D. Bryant, C. A. Dol-
  loff, G. B. Grette, R. A. House, M. L. Murphy, K.
  V. Koski, and J. R. Sedell 1987. Large woody  de-
  bris in forested streams of the Pacific Northwest:
  past and present. Streamside Management: Forestry
  and Fishery Interactions.  Salo, E.  O., and T.  W.
  Cundy.(eds.) Institute  of Forest Resources, Univer-
  sity of Washington, Seattle, WA.
Bryant,  M. D. 1983. Role and  management of woody
  debris in west coast salmonid nursery streams. No.
  Am. Jour. Fish. Mgmt. 3:322-330
Cederholm, C.  J. and W. J. Scarlett. 1982. Seasonal
  immigrations of juvenile  salmonids into  four small
  tributaries  of the Clearwater River, Washington
  1977-1981.  pp  98-110  In E.  L.  Brannon  and
  E. O.  Salo  (eds.) Proceedings of the Salmon and
  Trout Migratory Behavior Symposium.  School of
  Fisheries, University of Washington, Seattle, WA.

 Riparian Habitat Management
 Centre for the Study of the Environment (CSE). 1995.
   Status and Future of Salmon in Western Oregon
   and northern California. University of California,
   Santa Barbara, CA.
 Collins, B. D., T. J. Beechie, L. E. Benda, P. M. Ken-
   nard, C. N. Veldhuisen, V.  S. Anderson, and D. R.
   Berg.  1994.  Salmonid habitat and watershed as-
   sessment  and  restoration  plan  for Deer  Creek,
   North Cascades  of Washington.  Report by  the
   10,000 years Institute to Stilliguamish Indian Tribe
   and Washington Department of Ecology. 232 pp.
 Cummins, K.  W.  1974. Structure  and function of
   stream ecosystems. Bioscience 24(11):631-641.
 Franklin, J.  F. and C.  T. Dyrness.  1969. Natural
   Vegetation of Oregon and Washington. USDA For-
   est Service Res.  Pap. PNW 80 Pacific Northwest
   Forest and Range Experiment Station. Portland OR.
 Franklin, J. F., D. R. Berg, D. Thornburg, and J. Tap-
   penier.  1996 Alternative silvicultural approaches to
   timber harvest: Variable retention  harvest system.
   In Forestry for the 21st Century.  Kathryn Kohm
   (ed.) Island Press. Washington D.C.
 Franklin, J. F., T. Spies, D. Perry, M. Harmon, and A.
   Mckee. 1986. Modifying Douglas fir management
   regimes for nontimber objectives.  In Douglas fir:
   Stand Management for the Future. C. D. Oliver, D.
   P. Hanley, J. A.  Johnson (eds.) Institute of Forest
   Resources  Contribution No. 55.  University  of
   Washington, Seattle, WA.
 Grant, G.  E.,  F. J. Swanson,  and M. G.  Wolman.
   1990. Pattern and origin of stepped bed  morphol-
   ogy in  high gradient streams,  western Cascades,
   OR.  Geological   Society  of  America  Bulletin

 Gregory, S. V., F. J. Swanson, W. A.  Mckee, and K.
   W. Cummins.  1991. An ecosystem perspective of
   riparian zones. Bioscience. 41:540-551.

 Grette, G. 1985. The role of large woody debris in ju-
   venile salmonid rearing habitat in  small streams.
   M.S. thesis,  University  of Washington,  Seattle,
   WA. 105 pp.

 Harmon M. et al. 1986 The ecology of coarse woody
   debris  in temperate  ecosystems. Adv.  Ecol. Res.

 Heede, B. H.  1985. Interactions between streamside
  vegetation and  stream dynamics pp 54-58 In Pro-
  ceedings of Riparian Ecosystems and Their Man-
  agement: Reconciling Conflicting  Uses.  Tucson,

Hollings, C. S.  1978. Adaptive  Environmental As-
  sessment and Management.  London. John Wiley
  and Sons.

Keller, E; H. and  F. J. Swanson.  1979.  Effects  of
  Large organic material on channel form and fluvial
  process. Earth Surface Processes. 4:361-380.

Lienkaemper G. W. and F. J. Swanson.  1987. Dy-
  namics of large woody debris in old growth Doug-
  las fir forests. Can. J. For. Res. 17:150-156.
Mcdade, M. H., F.  J. Swanson, W. A. McKee, and J.
  F. Frank. 1990. Source distances for course woody
  debris entering small streams in Western Oregon
  and Washington.  Can. Jour, of For. Res.  20(3):326-
Megahan, W. F. 1982. Channel sediment storage be-
  hind  obstructions in the forested drainage basins
  draining the granitic bedrock of the Idaho Batholith
  pp 114-121 In F. J. Swanson, R. J. Janda, T. Dunne,
  and D. N. Swanston  (eds.)  Sediment budgets and
  routing  in forested  drainage basins. USDA For.
  Serv. Gen. Tech. Rep. PNW-141 Pac.  Northwest
  Forest and Range Exp. Stn. Portland, OR.
Mitsch, W. J.,  and  S. E. Jorgensen. 1989.  Ecological
  Engineering. Wiley and Sons, Inc. NY. 472 pp.
Naiman, R. J.,  T. J. Beechie, L. E. Benda, D. R. Berg
  (and  others). 1992. Fundamental elements of ecol-
  ogically healthy  watersheds  in the Pacific North-
  west coastal ecoregion. pp  127-188 In  Watershed
  Management: Balancing Sustainability  and Envi-
  ronmental Change. (R. J. Naiman ed.). Springer-
  Verlag, NY.

Nilsson,  C.,  G. Grelsson, M Johansson, and  U.
  Sperens.  1989. Patterns of  species richness along
  riverbanks. Ecology 70(1).

Oliver, C. D. and B. C. Larson. 1990. Forest Stand
  Dynamics. McGraw-Hill, Inc. NY. 467 pp.

Oliver, C. D. and T.  M.  Hinkley.  1987. Species,
  Stand  Structure  and  Silvicultural Manipulation
  Patterns for the  Streamside zones. In  Streamside
  Management:  Forestry  and Fishery Interactions.
  Salo, E. O.,  and  T. W. Cundy (eds.).  Institute of
  Forest   Resources,  University  of  Washington,
  Seattle, WA.

Oliver, C. D., D. R. Berg, D. R. Larsen, and K. L.
  OHara.  1992.  Integrating  management   tools,

                                                                             Riparian Habitat Management
 ecological knowledge, and silviculture, pp 361-382
 In  Watershed Management: Balancing Sustain-
 ability  and environmental  change.  (R.J. Naiman
 ed.). Springer-Verlag, NY.
Potts, D. F. and B. K. M. Anderson.  1990. Organic
 debris and the management of small stream chan-
 nels. W. Jour. Appl. For. 5(l):25-28.
Raedeke, K. J.  1988. Ecology of large mammals in
 riparian systems of the Pacific northwest forests.
 pp. 113-132 In Streamside Management: Forestry
 and Fishery Interactions. K.J. Raedeke (ed.). Insti-
 tute of Forest Resources, University of Washington,
  Seattle, WA.
Rainville R. P., S. C. Rainville, and E. L. Lider. 1985.
  Riparian silvicultural strategies for fish habitat em-
  phasis. In Proceedings of 1985 Society of American
  Foresters National Convention. Ft. Collins, CO.
Reeves, G. H., L. E. Benda, K. M. Burnett, P. A. Bis-
  son, and J. R. Sedell. 1995. A disturbance-based
  ecosystem approach  to  maintaining and restoring
  freshwater habitats of evolutionary significant units
  of anadromous salmonids in the Pacific Northwest.
  American Fisheries  Society  Symposium  17:334-
Salo, E. O. and T. W. Cundy. 1987.  Streamside Man-
  agement: Forestry and Fishery  Interactions. Insti-
  tute of Forest Resources, University of Washington,
  Seattle, WA.
Sedell, J. R. and J. L. Froggatt. 1984. Importance of
  Streamside forests  to large rivers: Isolation of the
  Willamette River, OR, USA from its floodplain by
  snagging and Streamside forest removal.  Interna-
  tionale Vereinigung fur theoretische und Limnolol-
  gie, Verhandlungen 22:1828-1834.
Sedell, J. R., P. A. Bisson, F. J. Swanson, and S. V.
  Gregory. 1989. What we know  about large  trees
  that fall into  streams and rivers.  153 pp. In From
  the forest to the sea:  a  story  of fallen  trees.
  C.Maser, R. F. Tarrant, J. M. Trappe and J. F.
  Franklin.  (Eds.).  GTR. PNW-GTR-229  Pacific
  Northwest Forest and Range Experiment  Station,
  Portland, OR.
Schoonmaker,  P.  and W.  A.  McKee. 1988.  Species
 composition   and   diversity  during   secondary
  succession  of coniferous  forests  in  the  western
  Cascade Mountains. For. Sci. 34(4):960-979.
Stanford, J. A., and J. V. Ward. 1988. The hyporheic
  habitat of river ecosystems. Nature 335:64-66.
Sullivan, K., T. E. Lisle, C. A. Dolloff, G. E. Grant,
  and L. M. Reid. 1987. Stream Channels: The link
  Between Forests and Fishes. In Salo, E.  O., and
  T. W. Cundy. 1987. Streamside Management: For-
  estry and Fishery Interactions. Institute of Forest
  Resources, University of Washington, Seattle, WA.
Swanson,  F. J. and G. W.  Lienkaemper.  1982. Inter-
  actions among fluvial processes, forest vegetation
  and aquatic ecosystems, South Fork,  Hoh River,
  Olympic National Park.  In Ecological research in
  the National Parks of the Pacific Northwest, E. E.
  Starkey, J.  F. Franklin and J. W. Matthews (eds.)
  Oregon State University. Corvallis, OR.
Swanson, F. J. and D. R. Berg. 1991. Ecological roots
  of new approaches to forestry. Forest Perspectives.
Swanson, F. L., S. V. Gregory, J. R. Sedell, and A. G.
  Campbell. 1982. Land and water interactions: The
  riparian zone. In R.L. Edmonds  (ed.) Analysis of
  coniferous forest ecosystems in the western United
  States.   Hutchinson   Ross   Publishing   Co.,
  Stroudsburg, PA.
Triska, F. J., J. R. Sedell, and S. V. Gregory. 1982.
  Coniferous forest streams.  In R.L. Edmonds (ed)
  Analysis of coniferous  forest ecosystems in the
  western United States. Hutchinson Ross Publishing
  Co., Stroudsburg, PA.
Trotter, E. H. 1990. Woody debris, forest-stream suc-
  cession, and catchment geomorphology. Jour. No.
  Am. Benthol. Soc. 9(2): 141-156.
Vannote, R. L., G. W. Minshall, K. W. Cummins, J.
  R. Sedell, and C. E. Gushing. 1980. The river con-
  tinuum concept. Can.  J. Fish. Aquat. Sci. 37:130-
Washington Forest Practices Board (WFPB). 1993.
  Standard Methodology for Conducting Watershed
  Analysis. Washington State Department of Natural
  Resources. Olympia, WA.

                             The Relationship of  Large Woody  Debris,
                                   Riparian Vegetation, and Landform in
                                                 Small, Low Gradient Streams
                                                                                   Byron Rot13
                                                                                R. J. Naiman1
                                                                                    R. E. Bilby2
  Riparian  and  stream  restoration  too  often  is
 attempted without an  appreciation  for how large
 woody debris (LWD) dynamics and riparian compo-
 sition are affected by landform and fluvial processes.
 A  total  of  twenty  one  sites were studied within
 mature and old forests of southern Washington state.
 Stream channels were delineated by bed morphology
 into three classes: bedrock, plane-bed, and forced
 pool-riffle. Geomorphic landform was used to ap-
 proximate the  relative fluvial influence  upon  the
 riparian plant community. Landform was divided into
 three classes based upon height  above  the  active
 channel (floodplain 0 to 1 m above the channel, low
 terrace to 1 to 3 m, high terrace > 3 m), with slope as
 the fourth class. Tree density by species  (trees per
 hectare)  characterized the riparian  forest  while the
 understory was  delineated with Detrended Corre-
 spondence Analysis. Valley constraint was defined as
 the ratio of  valley width  to active channel width, at
 3 m above the active channel.

  Large woody debris volume was significantly influ-
 enced by valley constraint within forced pool-riffle
 channels with volume increasing as a power function
 of  decreasing  valley  constraint  (y=15.854c°1559x,
 r2=0.58). No pattern was observed for bedrock or
 plane-bed channels.  The presence  of off-channel
 habitat for aquatic organisms increased exponentially
 with decreasing valley constraint for all channel types

  Fluvial disturbance had a decreasing impact on ri-
 parian  succession and community composition with
 increasing elevation above the channel. In the over-
 story, deciduous species dominated floodplain land-
 form with red alder (Alnus rubra) at 62 percent of
 stems. Conifers were more common on higher land-
 forms.   Upland  disturbances  controlled  riparian
succession on  low and high terraces, while fluvial
disturbances  controlled floodplain  riparian  suc-
 The diameter of LWD in the channel was related to
the diameter at breast height (dbh) of the  riparian
forest.  In old-growth  stands  (>300 years), LWD
diameter was  significantly greater (o=0.05) than the
average forest diameter for  all sites. This follows
common successional theories for old-growth as the
oldest age cohort begins to die out with some of those
stems falling into the channel. In younger stands (100
to 300 years),  a mixed relationship between LWD
and riparian forest diameter reflected LWD sources
from previous stands, smaller suppressed stems  from
the existing stand, and a wide range of diameters
contributed through bank erosion.
 When planning riparian restoration, the effect of
fluvial disturbances and channel migration  must be
factored into  the plan.  The floodplain landform (in
many cases) can be used to define the channel migra-
tion zone, or the area bounded by historical channel
movement. The  rate  of channel migration  must be
predetermined prior to  restoration. If conifer regen-
eration  is desired  on floodplain landform  (a  risky
prospect), then planting should be restricted to safe
sites: topographic highs, on  nurse logs, and behind
stable  LWD jams. Finally,  all restoration projects
include an element of luck. Luck (or a period of mild
annual floods)  may have  accounted for successful
natural conifer regeneration,  and may be needed for
the success of your restoration plan.

Literature Cited
Fetherston, K. I., R. J. Naiman, and R. E. Bilby. 1995.
  Large  woody  debris, physical processes, and
| Center for Streamside Studies, University of Washington, Seattle, Washington 98195.
J Weyerhaeuser Company, Federal Way, Washington 98477.
3 Point No Point Treaty Council, Kingston, Washington 98346.

                                                                                  Low Gradient Streams
  riparian  forest  development  in  montane  river
  networks  of  the  Pacific  Northwest.  Geomor-
  phology. 13:133-144.
Hupp, C. R. 1988. Plant ecological aspects of flood
  geomorphology  and paleoflood history, pp.  335-
  356 In V. R. Baker, R. C. Kochel, and P. C. Patton
  (eds). Flood geomorphology. John Wiley and Sons,
  New York, NY.
Montgomery, D.  R. and  J. M. Buffington.  1993.
  Channel classification,  prediction of channel  re-
  sponse,  and  assessment  of  channel condition.
  Washington State Timber-Fish-Wildlife agreement.
  Report TFW-SH10-93-002. 84 pp.

Rot, B. W. 1995. The interaction of valley constraint,
  riparian landform,  and riparian plant  community
  composition size and age upon channel configura-
  tion of small, low gradient streams of the western
  Cascade  Mountains, Washington. M.S.  Thesis,
  University of Washington, Seattle, WA. 67 pp.

                                         Location and  Design Influence
                                     Hydrologic Dynamics of Wetland
                                                             Mitigation  Projects

                                                                        Paul W. Shaffer1
                                                                        Mary  E. Kentula2
  Hydrologic conditions are critical to the estab-
 lishment, function,  and persistence of wetlands,
 and are  perhaps  the  most  important,  element
 influencing the successful development  of miti-
 gation  projects.  Despite  the  recognized   im-
 portance of hydrology,  it remains one of the least
 studied attributes of wetlands (Kusler and Kentula
 1990).   To  achieve   wetland  management
 objectives that incorporate restoration techniques,
 it is necessary to develop an understanding of
 how hydrology  influences other structural  and
 functional attributes of wetlands,  and to  un-
 derstand how design and  siting  of projects will
 influence their hydrologic regimes.
  A goal of the U.S. EPA's Wetlands Research
 Program (Leibowitz et al. 1992) is to assess the
 extent to which wetland mitigation projects are
 replacing the structure  and function of naturally-
 occurring wetlands, and to provide technical data
 to  support development  of design  and  siting
 guidelines for wetland mitigation projects. As one
 element of an integrated assessment of naturally-
 occurring   wetlands   and   wetland   mitigation
 projects within the Wetlands Research Program
 (Magee et al. 1993), a 3-year project was  initiated
 in  the fall  of 1993 to monitor  water levels in
 approximately 50 freshwater, emergent wetlands
 (naturally-occurring   wetlands   and   wetland
 mitigation projects) in  the  Portland,  Oregon,
 metropolitan area. The  study wetlands are located
 in diverse land use and  hydrogeomorphic  settings,
 ranging  from small slope wetlands,  to  riparian
 corridors of urban  streams, to wetlands on the
floodplain  of a  large,  highly-regulated  river
 Data from the first year of monitoring demon-
strate wide  variability  in water  level dynamics
within and  among  study wetlands, suggesting
important differences in the control  of short and
long-term variability in water levels (i.e., indi-
vidual storm events and interannual variability).
During individual storm events, land use setting
appears to be the most  important factor affecting
changes in wetland  water levels. Both naturally-
occurring wetlands  and wetland mitigation proj-
ects in urbanized settings experience large (up to
1 m) changes in water levels,  with rapid rise and
recession. Stage increases are much smaller in
wetlands  in rural  areas,  and  recessions  are
considerably slower than in urban systems. The
differences are presumably attributable to exten-
sive  landscape modifications  in the urban areas
(extensive impermeable areas,  channelized runoff,
etc.)  that route water  quickly to wetlands and
stream channels. The  association between land
use   and  hydrology  suggests   that   location
constrains levels of hydrologically-related func-
tions  (e.g.,  flood storage),  and further  suggests
that wetlands should be designed so that functions
are  compatible with   extant  site  conditions.
Further,  where  specific functions  are desired,
projects  should  be placed in settings likely to
maximize potential levels of the  desired func-

  On  an  annual basis,  there are  pronounced dif-
ferences  in hydrologic  regimes  of  naturally-
occurring   wetlands   and  wetland  mitigation
1 ManTech Environmental Research Services Corp., U.S. EPA, NHREEL - Western Ecology Division, Corvallis,
Oregon 97333. Current affiliation Dynamac Corporation, U.S. EPA, NHREEL - Western Ecology Division, Corvallis,
Oregon 97333.
2 U.S. Environmental Protection Agency, NHREEL - Western Ecology Division, Corvallis, Oregon 97333.

                                                                              Hydrologic Dynamics
projects that are closely tied to project design and
wetland morphology. Differences are related most
notably to the presence and type of water control
structure (excavated pond, impoundment) in the
wetland.  During  1994,   wetland   mitigation
projects  had  more   extensive  and persistent
standing water than naturally-occurring wetlands,
observations  consistent with previous research in
other geographic areas (e.g., Confer and  Niering
1992). Results reflect the fact that most naturally-
occurring wetlands are marshes or meadows in
which water  regimes are unregulated, while most
wetland  mitigation  projects  are  ponds  with
structural modifications  designed  to maintain
standing water and regulate  water  levels.  The
annual  range in water levels  (excluding storm
events) in 1994  was  much larger in naturally-
occurring wetlands  than in wetland mitigation
projects  (1.09 vs. 0.56 m) and  in some  wetland
mitigation  projects water  levels varied  by less
than 10 cnxduring the year. Average water depths
were much higher in wetland mitigation  projects
than  in  naturally-occurring wetlands (0.75  vs.
0.30  m), and the average extent  of flooding in
wetland  mitigation projects was about twice that
in naturally-occurring wetlands (52 vs. 26 percent
of  wetland  area).  Standing water  was  present
throughout the year on at least part of the  wetland
in 80 percent of wetland mitigation projects, but
only 20 percent of naturally-occurring wetlands.
  Results show that design of wetland mitigation
projects is resulting in hydrologic regimes that are
fundamentally  different  from  those   of the
naturally-occurring wetlands they  were  built to
replace,  and  that design of wetland mitigation
projects is  leading to changes in  the  relative
abundance  of  wetland  types  for  freshwater
emergent wetlands in the Portland area. Given the
influence of hydrology  on  other   wetland  at-
tributes, changes in  wetland  type  are almost
certainly resulting in  wetlands  with  soils,  vege-
tation, and  habitat functions that are  different
from those of naturally-occurring wetlands in the
region. Results suggest that,  as advocated by
Kentula  et al. (1992), wetland managers  need to
recognize  the  intimate  relationships  between
wetland  structure and function; if functional re-
placement of wetlands is the desired endpoint of
management   activities,  modification  of current
project designs is necessary.
Literature Cited
Confer, S. R., and W. A.  Niering. 1992. Com-
  parison   of  created and  natural freshwater
  emergent  wetlands  in  Connecticut.  (USA)
  Wetlands Ecology and Management 2:143-156.
Kentula, M. E., R. P Brooks, S. E. Gwin, C. C.
  Holland, A.D.  Sherman,  and J. C.  Sifneos.
  1992. An  Approach  to Improving  Decision
  Making in Wetland Restoration  and  Creation.
  Edited by  A.J.  Hairston. EPA/600/R-92/150.
  U.S. Environmental Protection Agency, Envi-
  ronmental Research Laboratory, Corvallis, OR.
Kusler, J.  A. and M. E. Kentula. 1990. Executive
  summary, p. xvii-xxv hi J. A.  Kusler and M. E.
  Kentula  (eds.)   Wetland  Creation  and  Res-
  toration: The  Status  of the Science.  Island
  Press. Washington, D.C.
Leibowitz, S. G., E. M. Preston, L. Y. Arnaut, N.
  E. Detenbeck, C. A. Hagley, M. E. Kentula, R.
  K. Olson, W.  D. Sanville, and R. R. Sumner.
  1992. Wetlands Research Plan FY 1992-96: An
  Integrated Risk-Based  Approach. EPA/600/R-
  92/060.  U.S. Environmental Protection Agency,
  Environmental Research Laboratory,  Corvallis,
Magee, T. K., S. E. Gwin, R.  G.  Gibson, C. C.
  Holland, J. Honea, P. W. Shaffer, J. C. Sifneos,
  and  M. E. Kentula. 1993.  Research  Plan and
  Methods  Manual  for  the Oregon  Wetlands
  Study. EPA/600/R-93/072. U.S. Environmental
  Protection  Agency, Environmental  Research
  Laboratory, Corvallis, OR.

                                   A Portland, Oregon,  Landscape Scale
                                   Prescription  for Rare Wetland Wildlife
                                                                               P. Lynn Sharp1
   Abstract:  Basic  landscape  ecology  principles
 were used to develop a site-specific strategy for
 maximizing the chances for long-term survival of
 several  wetland-associated wildlife species. This
 application  used  existing  data  on  the  area
 requirements of eight locally rare wildlife species
 with strong associations with wetlands. The eight
 species were classified into habitat and area size
 requirement groups. National Wetlands Inventory
 maps were used to identify potential areas of wet-
 land habitat on a  satellite photo. The  primary
 areas of potentially suitable habitat identified for
 the eight  species were the lowlands and islands
 along the Columbia River which included areas of
 pasture, former pasture, forest, agricultural land,
 remnant sloughs and lakes, residential, commer-
 cial  and  industrial  areas.  Patches of potential
 habitat ranged in size from over 3.5 square kilo-
 meters in  size to those smaller than 50 hectares. It
 was concluded that there was an abundance of po-
 tential  habitat for these species in this area and
 that interconnections between patches could be
 restored  or maintained. This general  approach
 identifies  specific sites that should be considered
 for acquisition by entities such as the local  re-
 gional government, METRO, (for its Greenspaces
 program), park districts, and state agencies.  It
 also reveals where interconnections between sites
 should be  considered in land use and development
 decisions by regulatory agencies. These sites could
 also be used as regional wetland mitigation banks
 if wetland restoration and enhancement activities
 would be  appropriate to provide or substantially
 improve habitat quality.

  A number  of wildlife species strongly associated
 with wetlands were identified as rare in the Portland,
 Oregon  metropolitan area by the 1990-1992 Metro
 Greenspaces  Inventory (Poracsky  et  al.  1992), and
from other sources such as the Oregon Natural Heri-
tage Program (1993) and the  Portland  Audubon
Society's Checklist of Portland  Birds. One of the
goals  of the Greenspaces Master Plan  (METRO
1992) is to ensure the long-term  survival of as many
indigenous species of plants and wildlife as possible
in the Portland urban area. The Master Plan identified
three general landscape-scale restoration  principles
that  could   be  applied  throughout   the  area:
(1) consolidate habitat areas;  (2) restore contiguous
riparian vegetation; and  (3) restore connections be-
tween  watersheds at headwaters.  The  Plan  also
included an  overall description  of each watershed
within the study area in terms of  its degree of urbani-
zation, types  and  extent  of natural areas remaining,
and  development pressures.  Recommendations  on
natural area conservation within each watershed were
also made, but often were not very site specific.
  Without some kind of prioritization and attention to
species-specific habitat needs, application of these
three principles is unlikely to meet the needs  of at
least some rare species of concern. The logical next
step, therefore,  would be to identify the habitat and
area requirements of a small number of rare wetland
species thought to represent the  habitat requirements
of most. This habitat information could  be used to
develop a landscape-scale prescription that could en-
hance the likelihood  of  long-term survival of these
species  in the urban  area  through a combination of
restoration, preservation, enhancement, and creation
of the habitats they  require.  Additional population
declines  might  then be  avoided and  eventually,
population increase and even successful reintroduc-
tion of extirpated species might be possible.

  There are  a number of planning and acquisition
level projects by various levels of governments in the
Portland area which, in combination, could greatly
enhance the chances of long-term survival of these
species. This analysis was designed to identify high
priority  areas for wetland conservation efforts to
 Woodward-Clyde Consultants, Suite 900, 111 SW Columbia, Portland, Oregon 97201.

                                                                                    Rare Wetland Wildlife
benefit these species, to  identify the likelihood of
successfully applying the  prescriptions, and  to sug-
gest research on solving some of the difficult prob-
lems  these  species  face  in  a  dynamic  urban


Study Area
  This analysis covers the Portland, Oregon urban and
urbanizing area, located along the south shore of the
Columbia River and including islands and the mouths
of the Willamette and Sandy Rivers (Figure  1). The
only connection between the Cascades Range on the
east and the Coast Range on the west lies adjacent to
the Columbia  River  and in the  agricultural lands
either many miles to the south of Portland or to the
north of Vancouver. Habitats and land uses along the
south side of the Columbia River, from east to west,
include: (1) the Sandy River Delta,  a large delta of
former pasture and  bottomland  hardwoods  on the
west side and which comprises the western end of the
Columbia  Gorge  National Scenic Area; (2)  an  area
rapidly changing from row crops  and pasture to resi-
dential, commercial and industrial area still traversed
by stream corridors, sloughs, remnant lakes and wet-
lands, with the heaviest industrial development west
of the Portland International Airport and extending to
the mouth of the Willamette River; (3) predominately
agricultural lands and remnant wetlands  on Sauvie
Island, a large island separated from the shore by the
Multnomah Channel of the Willamette River; and
(4) on the mainland the  forested peninsula of the
Tualatin Mountains, which extend deeply into urban
Portland from the Coast Range, including the 2,000-
ha Forest  Park, the largest urban park in  the United
  The density of development in the central portion of
Portland is high and most streams have been put in
underground channels. The riparian corridor along
the Willamette River is absent in the most  intensively
developed  areas but is present as at least a narrow
strip of vegetation in the  remainder.  Several natural
parks are located along the Willamette and on its is-
lands within the Portland area.  Riparian corridors
along other major rivers and smaller streams are vari-
able as well.
Wildlife Species
 Eight rare wetland-dependent wildlife species were
selected for this analysis. The main criteria used to
select species were that they were rarely encountered
during  the  Metro  Greenspaces  field  surveys
(Poracsky et al. 1992), were considered to be rare or
sensitive according to the Oregon Natural Heritage
Program (1993), were listed as rare on the Portland
Audubon Society's Portland Bird Checklist, or were
considered  to be rare by local wildlife ecologists and

 The wetland habitats  commonly used by the se-
lected  species were identified  on the  basis of the
Geenspaces data, published literature,  and observa-
tions by local ecologists and birders. References such
as Brown (1985) were reviewed, and information on
home range/territory  size and  habitat  requirements
was summarized. On the basis of habitat use, the
eight species were placed into  a smaller number of
basic  habitat-related  groups representing  combina-
tions of wetland and upland habitats.

Potential Habitat Identification and
 A color poster, derived from 1992 satellite data and
published by Satellite Images, Inc. (reproduced by
permission of Satellite Images, Inc.), was used as the
base map  for this  analysis  because it provided an
overview of the entire urban area and its vegetation.
The National Wetlands Inventory maps (U.S. Fish
and Wildlife Service, various dates) covering the ur-
ban area were examined and  the largest areas of
mapped wetlands, 50 ha or more in size, were marked
on an  overlay  of the satellite image.  The satellite
image  base provided  information on the basic  vege-
tation  type(s) for each  wetland area as well  as the
character  of  the  upland  area  surrounding  each

Habitat Availability and Suitability
 Based on size, vegetation, location, and surrounding
land use, the general suitability of each potential area
of habitat was assessed  for each of the eight species.
This was compared to the known range of occurrence
of each  species in the Portland area. The  general
status  of each species and the  challenges it faces in
the urban environment was described.

 Rare Wetland Wildlife
 Future Research
  Suggestions of research projects that should precede
 reintroduction or population enhancement efforts for
 some species were made, based on the results of the


 Wildlife Species
  The  species  selected were  the northern harrier
 (Circus cyaneus L.),  Virginia rail (Rallus limicola
 Vieillot),  sora  (Porzana  Carolina  L.),  American
 bittern  (Botaurus  lentiginosus  Rackett);  black-
 crowned  night  heron  (Nycticorax  nycticorax  L.),
 muskrat (Ondatra zibethica L.), western pond turtle
 (Clemmys marmorata Baird and Gerard), and western
 painted turtle (Chtysemys picta Schneider). All of
 these species have been adversely affected by habitat
 degradation, including water pollution problems such
 as algae  blooms and oil  spills, water level fluctua-
 tions  due to stormwater  runoff, reed  canarygrass
 invasion  of emergent  areas, habitat fragmentation
 (especially  for  less mobile species such  as the two
 turtle species); predation of turtles and nests by intro-
 duced bullfrogs (Rana catesbeiana Shaw) and other
 predators; and  in  the  case of the muskrat, potential
 competition with  exotic  nutria  (Myocastor coypus
 E. Geoffrey St.-Hila'lre).

 Habitat Relationships
  The basic habitats utilized  by the eight species fit
 into four general  groupings as follows: wet and dry
 meadow  and emergent-northern  harrier; emergent
 and open water-Virginia rail, sora, American bittern,
 muskrat;  emergent, open water  and forest-black-
 crowned night heron; open water, emergent and adja-
 cent   uplands-western pond  and  painted  turtles
 (Godfrey 1966,  Perry 1982, Brown 1985, Boutin and
 Berkenholtz 1987, Holland 1994, Storm and Leonard

  According  to Brown  (1985),  northern harrier
 breeding home ranges vary from 100 to 890 ha, and
 winter home  ranges vary from 12 to 260  ha; pub-
 lished home range sizes for the  bittern  and  black-
 crowned night heron were not found in the literature
 but probably  intermediate between  harrier and  the
 smaller rails; Virginia rail and sora home  ranges
during nesting are  in the <  1 ha range, with sora den-
sity reported as up  to 30 birds per square kilometer.
 Muskrat home ranges are reported as usually <1 ha
but up to 2.4 ha, with a minimum habitat per popula-
tion of 1.6 km of stream (Brown 1985). Boutin and
Birkenholz (1987) report  home range size of indi-
vidual muskrats in the < 1  ha to 1 ha range and point
out that home range size appears to vary with habitat
quality. Densities of muskrats were in the 10-80 ani-
mals per ha range in large marshes  and were much
lower along water courses, ranging from 1 pair per
46 m  of shoreline to 3 or  4 pairs per 1609 m (Perry
 The home ranges of individual western pond turtles
are reported as less than 1 ha by Bury (1972). Data on
home range size of the western  painted turtle could
not be found.  For purposes of this analysis, painted
turtle home ranges were  assumed to be similar to
pond turtles.
 On the  basis of this review, the eight wildlife spe-
cies were grouped into three home range size catego-
ries differing by an order of magnitude: individual
pairs of harriers-over 100 ha; bittern, black-crowned
night heron, and muskrat colonies-10 to 100 ha; and
individual Virginia rails and soras, and colonies of
western pond turtle and western painted turtle-< 1 to
10 ha.
Potential Habitat Identification
  The larger wetland areas identified in the National
Wetlands Inventory maps  are outlined in Figure 1.
These are the areas where opportunities might exist to
enhance  wetland and  upland habitats in  terms of
quantity  and quality for the eight species of concern.
The largest areas are  concentrated in the lowlands
along the Columbia River, including around three
remaining large  lakes:  Sturgeon,  Vancouver,  and
Smith-Bybee;  on islands and along stream corridors
and Sandy River Delta. Forested areas appear dark in
this Figure while some croplands appear nearly white
and are also indicated  by relatively large block size.
The urbanized area appears a  mottled gray  color,
although the rural lands north of Vancouver, Wash-
ington appear to be more urbanized than they actually

  The larger areas of wetlands around the lakes  con-
sist of a combination of emergent, scrub-shrub, and
forested  patches adjacent to these large areas of open
water. In the case of Sturgeon Lake on Sauvie Island,
these wetlands are surrounded  by agricultural  land
which is expected to remain agricultural over the long
term as a result of current zoning and Oregon's land

Base map reproduced with permission of
Satellite Images Inc., Anacortes, Washington.

Legend ~ Wetlands identified on NW I maps

 Rare Wetland Wildlife
 use law. The areas adjacent to Smith-Bybee are inten-
 sively urban and mostly industrial, and development
 has occurred or  is rapidly occurring in the area
 between Smith-Bybee  Lakes and  the  Sandy River
 Delta to the east.  The Columbia Slough, however,
 and wetland mitigation activities in the area between
 the airport  and  Sandy River Delta could  maintain a
 connection   between  the  Delta  and  Smith-Bybee
 Lakes, focused on the Columbia Slough.

 Habitat Availability and Suitability
  Harriers   are  regularly  observed  year-round   at
 Sauvie Island  and around  Vancouver Lake, where
 there is  an  abundance of wet and dry meadows. The
 area along the Columbia River east of Sauvie Island
 did provide habitat but rapid development is expected
 to  eliminate this habitat. The  only other remaining
 areas  of suitable  habitat  large enough  to  support
 breeding pairs are Powell Butte (a City Park), and the
 Sandy River Delta, now under U.S. Forest Service
 ownership.  The Management   Plan  for  the Sandy
 River Delta (USDA Forest Service  1995)  calls  for
 rehabilitation and  creation  of  upland and  wetland
 meadows, including several strategies for control of
 reed canary grass, in the 100-ha or greater size range,
 which should provide nesting habitat for harriers over
 the long term.
  The bittern, black-crowned night heron, rails, and
 muskrat all utilize open water/emergent complexes.
 The  night  heron  also  requires forested  areas  for
 roosting. These areas are also  prevalent  on Sauvie
 Island,  in  the  Smith-Bybee Lakes  area,  in  some
 patches  near the Columbia and  along the Willamette
 River, and  in the Sandy River  Delta. Emergent wet-
 lands in at least some proportion of all of these areas
 are heavily  invaded by reed canarygrass which does
 not appear to be as suitable as habitat for these spe-
 cies as other emergents such as cattail (Typha latifo-
 lia L.) and bulrush (Scirpus).

  The western painted turtle is fairly widely distrib-
 uted in the lakes, sloughs, and islands  along the south
 side of the Columbia. The western pond turtle, how-
 ever, is rarely observed. The relatively large popula-
 tion of introduced bullfrogs combined with habitat
 degradation  is suspected to be a factor in the decline
 of the pond turtle,  along with a loss  of suitable  un-
 disturbed upland nesting habitat adjacent  to remain-
 ing wetland  areas.
 Maintenance of wet and dry meadow habitat blocks
in the 100-ha size range which are not disturbed dur-
ing the  nesting season by mowing, plowing, and at
least somewhat protected from human and dog access
would be an appropriate goal for harrier habitat man-
agement in all of these areas. A secondary but no less
important goal should be to control reed canary grass
(Phalaris arundinacea L.) as much as  possible  be-
cause prey availability  and thus habitat  suitability is
low owing to the height and density of reed canary
grass, according to a USFWS (undated)  draft Habitat
Evaluation Procedures model for the northern harrier.
Late summer mowing could still occur to prevent in-
vasion by trees and shrubs. Upland meadows need to
be included, as nesting  occurs  on the ground  in
uplands as well as wetlands.
 Introduced nutria may be acting as a keystone spe-
cies in creating and maintaining monocultures of reed
canary grass in many urban wetlands in the Portland
area, particularly in the floodplains of local streams.
This suggestion  is based on personal observations of
major changes in  emergent species composition at
created  and natural wetlands in the Portland area con-
current  with rapid nutria population growth. Reed
canary  grass control  measures   might  be  more
successful if nutria populations were also controlled.
A  valuable experiment would be  to  test identical
methods of physical/chemical control  of reed canary
grass with and without nutria population control.
 Habitat for the two rail species, bittern, night heron,
muskrat, and two turtle species overlaps considerably
and usually occurs in complexes that include all of
the elements needed-open water, emergent  vegeta-
tion, scrub-shrub, forested wetland, and adjacent up-
lands.  Preservation  of the  large  blocks  of habitat
needed  by  an individual harrier pair could also bene-
fit these other species  if the harrier habitat includes
marshes, lakes, and channels. In addition, many of the
areas smaller than 100 ha shown on  Figure  1 could
provide suitable habitat  for  populations of these
smaller, less wide-ranging  wildlife   species.  The
habitat  components present and/or which could be
restored at individual sites would determine which of
the above species would find suitable habitat.
 Connections between habitat patches are probably
not so important for harriers and other  birds because
they are highly mobile. Aquatic connections should
be present between smaller patches to provide ave-
nues of movement by species  such as  muskrats and
the two turtle species.

                                                                                 Rare Wetland Wildlife
Literature Cited
Anonymous.  Undated.  Draft  Habitat  Suitability
  Model for the Northern Harrier. Unpublished draft
  manuscript  provided  by  U.S.  Fish  and Wildlife
  Service. 1 p.
Boutin, S. and D. E. Birkenholz. 1987. Muskrat and
  Round-tailed Muskrat. pp. 314-325 In M.  Novak,
  J.A. Baker, M. E. Obbard, and B. Malloch (eds).
  Wild Furbearer Management and  Conservation in
  North  America.  Ontario  Ministry  of   Natural
  Resources, Toronto. 1150 pp.
Brown, E. R.  (ed.) 1985. Management of wildlife and
  fish habitats  in  forests of western  Oregon  and
  Washington.  Pacific  Northwest  Region,  Forest
  Service, USDA,  Publication No.  R6-F&WL-192-
  1985. 2 Vols.
Bury, R.  B.  1972. Habits and home  range of the
  Pacific  pond  turtle  (Clemmys  marmorata)  in  a
  stream community. Ph.D. Dissertation, Univ. Calif.
  Berkeley (not  seen, cited in Holland 1994).
Godfrey, W. E.  1966. The Birds of Canada. Nat. Mus.
  Can. Bull. No. 203, Biol. Series 73. 428 pp.
Holland, D. C. 1994. The western pond turtle: habitat
  and history -  Final report. Prepared for U.S. De-
  partment of  Energy,  Bonneville  Power  Admin.,
  Portland, OR. 279 p. plus appendices.
Metropolitan  Service   District   (METRO).  1992.
  Metropolitan  Greenspaces Program  Master  Plan.
  METRO, Portland, OR.
Oregon Natural  Heritage  Program.  1993.  Rare,
  threatened and endangered plants and animals of
  Oregon. Oregon  Natural Heritage  Program, Port-
  land, OR. 79 pp.

Perry,  H.  R.,  Jr.  1982. Muskrats. pp. 282-325 In
  Chapman, J. A. and G. A. Feldhamer (eds). Wild
  Mammals of North  America,  Biology, Manage-
  ment, and Economics. John Hopkins Univ. Press,
  Baltimore. 1147 pp.

Poracsky, J., P. L.  Sharp, and E. Lev.  1992. Metro-
  politan Greenspaces Program Data Analysis: Field
  Based Biological Data Final  Report. Prepared for
  METRO, Portland, OR. 142 pp.
Portland  Audubon Society.  Checklist of Portland
  Birds. Portland, OR.
Satellite Images, Inc. 1992. Portland-Mt. St. Helens
  from Space  (Poster). Satellite  Images, Inc.,  P.O.
  Box  1257, Anacortes, WA 98221.
Storm, R. M. and W. P. Leonard. 1995. Reptiles of
  Washington and Oregon. Seattle Audubon Society,
  Seattle, WA. 176 pp.
U.S.  Fish  and Wildlife  Service,  various  dates.
  National  Wetland  Inventory  U.S.G.S.  7.5-minute
  Quadrangle Maps:  Portland,  Mt. Tabor, Linnton,
  Sauvie Island, Sandy, Damascus,  Beaverton,  and
  Lake Oswego, Oregon;  Vancouver, Camas,  and
  Washougal, WA.
USDA Forest Service.  1995.  FEIS for the Sandy
  River Delta Plan. USDA Forest Service, Columbia
  River Gorge  National Scenic Area, Hood River,

                              Criteria for the Restoration and  Creation
                                 of Wetland Habitats of Lentic-Breeding
                                    Amphibians of the  Pacific  Northwest

                                                                           Klaus O.  Richter1
  Abstract:  I describe watershed features and
 wetland design guidelines to encourage coloniza-
 tion  and self-sustaining populations of lentic-
 breeding amphibians at wetland restoration and
 creation sites. I describe breeding, feeding and
 refuge habitat. I also characterize migration cor-
 ridors between these habitat components and de-
 scribe  requirements  for  dispersal habitat.  I
 provide criteria based on a review of conserva-
 tion biology and landscape ecology, amphibian
 natural history literature, and on my own studies
 of lentic-breeding amphibians  in the  Puget
 Sound Basin. Specifically, I suggest area, lengths,
 widths and  quality  of  refuge  and movement
 habitats. I also describe a hierarchy of explicit
 wetland attributes,  including current velocity,
 minimum water depths and water level fluctua-
 tion, and open  water and vegetation require-
 ments to encourage amphibian breeding. Finally,
 I  present criteria for wetland size, orientation,
 configuration and buffer  condition to further
 maximize spawning success.

  Success of wetland mitigation depends on replac-
 ing the range of physical and biologic functions that
 have been lost. Most restoration to date has been di-
 rected  towards  creating  appropriate  hydrology
 (Kusler and Brooks  1988)  and  native vegetation
 (Galatowitsch and  van der  Valk  1994), the most
 fundamental  and  crucial wetland  characteristics
 (Mitsch and Gosselink 1993). Methods for improv-
 ing waterfowl production have also been addressed,
 but primarily for wetlands in the prairie pothole and
 midwest region of the United States (Weller 1990).
 Techniques for restoring habitat for a greater diver-
 sity of wildlife, and in  other geographic  locations,
 have only recently been reported (Hammer 1992,
 Payne 1992, Kent 1994a,  Kent 1994b, Wheeler et
 al. 1995); yet a bias remains for developing avifau-
nal habitat (Galatowitsch  and van der Valk 1994,
Ward et al. 1995). Minimal attention  is focused on
amphibians or other non-avian aquatic wildlife, and
on landscape requirements for such taxa.

 Paradoxically, with reported advances in provid-
ing wildlife functions at restoration wetlands, con-
cern about amphibian extinctions and declines has
grown (Corn 1994, Stebbins and Cohen 1995, Olson
and Leonard 1997, among others). The dominant
amphibian assemblages of concern in many of these
writings are  those using wetlands and other lentic
habitats for breeding. Scientists believe there is no
single reason for amphibian losses, but rather that
there are a multitude of factors potentially involved.
In the Pacific Northwest, factors include pathogenic
fungi (Blaustein et al. 1994a),  Ultraviolet B radia-
tion (Blaustein  et al. 1994b,  1995b), agricultural
runoff (Boyer 1993, Boyer and Grue 1995) and the
introduction of exotic animals (Hayes and Jennings
 Habitat loss and deterioration, however, is consid-
ered by many to be one of the foremost reasons for
wetland-breeding amphibian declines (Stebbins and
Cohen  1995). In watersheds undergoing develop-
ment (e.g., urbanization) and other disturbance (e.g.,
logging, grazing), upland habitats essential for land-
dwelling  stages may be  severely altered (Minton
1968, Richter and Azous 1995, Olson and Leonard
1997, Delis et. al. 1996). Specifically, feeding and
refuge  (e.g., aestivation  and hibernation) patches
may be destroyed or fragmented.  Consequently,
movement between these patches and breeding sites
(e.g.,   wetlands)  may  be  hindered,   separating
individuals from these environments and from each
other, and thereby increasing the risk of extinction
(Blaustein et al.  1994c, Sjogren-Gulve and Ray

 Wetlands themselves may be filled, drained and
otherwise  changed by  land-use  conversions that
modify their hydrology, water quality and vegeta-
tion (Booth  1991, Booth and Reinelt 1993, Taylor
1993, Booth and Jackson  1994,  Holland  et.  al.
1995),  destroying amphibian  breeding sites. For
example, there is evidence that in watersheds with
'King County Natural Resources Division, 506 Second Avenue, Suite 720, Seattle, Washington 98104-2311.

                                                                               Lentic-Breeding Amphibians
proportionately  greater  development  amphibians
occur in fewer wetlands and the quality of breeding
habitat remaining may be poorer than in wetlands
less affected by runoff (Minton 1968, Richter and
Azous 1995, Azous and Richter 1995). Collectively,
these activities are hypothesized to account for de-
creasing ranges and dwindling numbers of the Ore-
gon  spotted frog (Rana pretiosa),  (Orchard 1992,
Richter and Azous 1995, Azous and  Richter 1995,
McAllister and Leonard 1990,  1997), western toad
(Bufo boreas)  (McAllister personal  communication,
Leonard,  personal  communication)  and possibly
other lentic-breeding species (Olson and Leonard
  Amphibian declines attributable to habitat losses,
however, can  be stabilized and reversed by water-
shed  protection and wetland restoration if factors
responsible for successful  population survival and
breeding biology are identified and provided. I am
unaware, however, of any scientific  literature that
specifically targets amphibians  for  wetland restora-
tion at a collective watershed-wetland scale. Nor am
I familiar  with  papers that systematically  suggest
and quantify habitat characteristics that may be used
as wetland restoration guidelines. Consequently, the
objectives  of this paper are to review the literature
and identify the diverse biogeographic and wetland
habitat  requirements of amphibians, and  also  to
quantify wetland site-selection and design  criteria
targeting  Puget Sound Basin  amphibians.  Hope-
fully, these restoration criteria may apply to behav-
iorally  similar  species  elsewhere  in temperate

  I develop  watershed  habitat  recommendations
from a review of conservation  biology (e.g., Saun-
ders et al. 1987, Saunders and Hobbs 1991, Forman
1995), population ecology (e.g., Gill  1978a, 1978b;
Sjogren 1994, Sjogren-Gulve and Ray 1996), natu-
ral history  (e.g., Nussbaum et al. 1983,  Blaustein et
al. 1995a), and herptile biology (e.g., Duellman and
Trueb 1986, Stebbins and Cohen 1995) literature of
relevance to lentic-breeding amphibians. I  specifi-
cally searched for empirical data on habitat use and
microclimate associations,  to determine conditions
that maximize wetland colonization and continued
use by amphibians.

  I also review  spawning habitat  use by wetland-
breeding amphibians (e.g.,  Nussbaum  et al. 1983,
Blaustein et al. 1995a, among others), to determine
requirements  for  successful  egg  development.
Moreover, I suggest design criteria based on 1988
through 1996 searches of amphibian breeding habi-
tat at 23 wetlands in the Puget Sound Basin  (Azous
and Richter  1995, Richter and Azous 1995,  Richter
and Leonard unpublished data, Richter and  Rough-
garden unpublished manuscript), and  on pitfall trap
data at 19 Puget  Sound wetlands censused from
1988 through  1995  (Richter  and Azous   1995).
These surveys sampled up to seven species at a site
including  the  Pacific  Northwestern salamander,
Ambystoma gracile Baird; long-toed salamander, A.
macrodactylum Baird; rough-skinned newt, Taricha
granulosa Skilton; red-legged  frog,  Rana  aurora
Baird & Girard; bullfrog, Rana catesbeiana Shaw,
western toad, Bufo boreas Baird & Girard;  and
Pacific treefrog, Hyla regilla Baird & Girard.
  Design criteria also are based on empirical  evi-
dence  from field  experiments  of habitat and  egg
manipulations in A. gracile and R. aurora (Richter
and  Roughgarden  unpublished  manuscript,  Richter
unpublished data), and  on oviposition literature of
Pacific Northwestern and other temperate-latitude
North  American  and European  wetland-breeding
  Finally,  I  assume that recurrently identified  am-
phibian breeding habitat features are beneficial to
spawning and  egg survival (i.e., reproductive fit-
ness), and that characteristics that substantially dif-
fer  from  identified  features  are  detrimental to
colonization and  self-sustaining populations at re-
stored and created  wetlands.

  The  historic literature  has generally  described
wetland-breeding  amphibian  habitat associations
during either aquatic or land-dwelling stages.  The
recent literature augments this knowledge  by em-
phasizing that amphibians respond  to habitats  at
several integrated landscape scales, ranging from
specific wetland  features favored for breeding to
broader watershed attributes. The former  include
oviposition  sites and the latter  include feeding and
refuge patches in addition to the connecting habitat
between them (Brown  1997,  Sjogren-Gulve  and
Ray 1996).  Also,  recent population biology and
conservation theories of amphibian distributions are
incorporating metapopulation and source-sink mod-
els to explain the  spatial and temporal distribution
of viable, self-sustaining populations. Fundamental
to these concepts  is the notion that animals use
more or less  discrete,  suitable feeding and refuge
patches within their home range, and further require

 Lentic-Breeding Amphibians
 favorable  migration  habitat between  home range
 patches, as  well  as unhindered  dispersal  habitat
 leading from natal wetlands to new breeding sites.

 Site-Selection Criteria

 Feeding and Refuge Habitats
   Feeding  and  refuge  habitats  are  critical to the
 land-dwelling  life  stages of wetland-breeding am-
 phibians. In general, many Pacific Northwest spe-
 cies spawn in wetlands during only a few weeks or
 months and spend  most of the  year  in terrestrial
 (upland) feeding and refuge habitats (See Blaustein
 et. al. 1995a, Olson and Leonard 1997 for an  over-
 view of breeding chronology and aquatic  and ter-
 restrial habitats used by amphibians associated with
 lentic habitats). Some low-elevation species  or indi-
 viduals of normally migrating species (R.  aurora,
 Taricha granulosa), as well as most  high-altitude
 species (R. cascadae, A. macrodactylum ), however,
 may remain  at wetlands in which they breed.
   The size and characteristics of feeding and refuge
 habitats away from wetlands (i.e., home range) are
 functions  of climatic condition and habitat struc-
 ture. Amphibians,  in general, prefer  cool, humid
 conditions (Duellman and Trueb  1986, Kleeberger
 and Werner  1983) with T. granulosa and A. gracile,
 often reaching their highest abundance in relatively
 cool forests  that are not extremely wet (Aubry and
 Hall 1991).  I have  trapped the greatest number of
 amphibians  within  stands and conditions  of  cool,
 stable climate,  well-drained soils and high ground
 humidity (Richter personal observation). Such cir-
 cumstances maintain moisture in woody debris, leaf
 litter and duff,  and  are favored as feeding and ref-
 uge sites because they facilitate amphibian  respira-
 tion and provide more  invertebrate food than drier
 sites without logs and other organic debris. Never-
 theless, some wetland-breeding species including T.
 granulosa, some Oregon spotted frogs (R. pretiosd),
 B. boreas, and Cascades frog (Rana  cascaded) may
 be found  in warm,  dry, exposed locations,  espe-
 cially early in spring or at high altitudes.

  There is a strong correlation of terrestrial-breeding
 amphibian distribution  (Plethodontidae) with  large
 woody  debris, dead and decaying wood and  organic
 matter, and  other habitat conditions  favorable to
 thermoregulation, foraging,  and resting (Welsh and
 Lind 1988, Walls et  al.  1992, Dupuis  et al. 1995).
 Because of similar physiology and food preference,
 it may be inferred that land-dwelling  stages of many
 lentic-breeding species prefer similar conditions. In
fact, Corn and Bury (1991) found higher densities
of T. granulosa, and Aubry and Hall (1991) larger
numbers of A. gracile in old-growth stands gener-
ally having greater volumes of downed wood than
mature and younger stands (Spies and Cline 1988).
In contrast, Lehmkuhl et al. (1991) found no rela-
tionship between stand age and amphibian richness,
but did find amphibians least abundant in clearcuts.
Clearly, mature and older forests have more numer-
ous crevices below logs, extensive root zones and
unoccupied rodent and other  small mammal bur-
rows that may be used as temporary refuge during
migration and dispersal (Loredo et al. 1996), feed-
ing (Williams 1970) and for aestivation and hiber-
nation  than younger  immature forests and timber
  Little empirical data is available for home range
sizes (e.g.  post-breeding  through winter hiberna-
tion) of Pacific Northwest species with the excep-
tion of A. gracile, individually  marked and followed
by Stringer (Stringer personal communication). She
identified many  recaptured individuals within 200
m of her surveyed wetland. Home ranges for species
in other regions indicate that home ranges are small
for most ambystoma  salamanders and  somewhat
larger for ranid frogs and toads (Table 1). Some in-
dividual  common  toads of Europe, Bufo bufo  L.,
may even  move  considerable distances to new
feeding and refuge  sites  several times a summer
(Beebee  1983).

  Without additional empirical  studies presumably it
is best to select wetland restoration sites adjacent to,
or as close to, forested habitats as possible. Larger
forested  areas with bigger diameter logs and other
coarse woody debris adjacent to some clearings will
provide habitats to  species with  smaller or more
diverse home range requirements.

Habitat Contiguity

  Dispersal Habitat. The passive colonization  and
self-sustainable occupation of restored  or created
wetlands depends on the ability of amphibians to
disperse from  favorable  source sites.  Dispersal
routes  from  source  populations  are also critical
when populations are eliminated by stochastic proc-
esses such as drought (Pounds and Crump 1994),
disease (Bradford  1991), and pollution (Richter  per-
sonal observation),  or when  populations produce
insufficient offspring to occupy a site permanently
(Gill 1978a,b; Sinsch 1992). Finally, amphibians at
restoration wetlands may benefit as members  of a
metapopulation. A  metapopulation encompassing

Table 1. Home Range and Dispersal Distances of Selected Amphibians.
Ambystoma gracile
Ambystoma jeffersonii
Ambystoma maculatum
Ambystoma opacum
Ambystoma talpoideum
Ambystoma texanum
Hyla regilla
Rana pipiens
Rana lessonae
Rana sylvatica
Bufo americanus
Bufo bufo
Bufo calamita
Home Range (m2)"



Distances (m)b

{625} (252)
[0-125] (64)
[6-220] (150)
[0-450] (194)
[0-1 25] (52)

{2,530} (1,1 26)

1,400 [100-200]

New York
South Carolina



Stringer pers. comm.
Douglas & Monroe 1981
Williams 1970
Bishop 1941 in Williams
Wacasey1961 in Williams
Williams 1970
Kleeberger & Werner 1983
Douglas & Monroe 1981
Williams 1970
Semlitsch 1981
Williams 1970

Brattstrom & Warren 1955
Dole 1965
Tunner 1991 in Dingle
Bellis 1962
Berven & Grudzien 1990

Ewert 1969 in Christein &
Taylor 1978
Bellis 1959
Sinsch 1988a
Beebee 1983
Minimum polygon: [Range], (mean), {linear distance}.
[Range], (mean), {maximum}.

 Lentic-Breeding Amphibians
 several wetlands may maintain healthy populations
 that otherwise could go extinct from inbreeding de-
 pression (Sjogren 1991, 1994, Pechmann and Wil-
 bur 1994). In turn, the  success of colonization and
 the movement between  breeding, feeding and hiber-
 nation  sites, is  considered to be dependent on the
 vagility of individual species, the distance between
 habitat patches, and the environmental characteris-
 tics of the intervening landscape.
  Given their small  size and  physiological con-
 straints amphibians are less mobile than  other ter-
 restrial vertebrates. Moreover, amphibian dispersal
 is taxa and age specific. Urodeles regularly disperse
 up to 1,000  m and anurans sometimes  greater than
 1,000 m from breeding  sites (Table 1). Juveniles are
 generally more vagile than adults (Williams  1970,
 Gill  1978b)  with metamorphs and young moving to
 new wetlands. Most adults are philopatric, faithfully
 returning to breeding  wetlands  during successive
 years (Twitty  1958, 1959, 1966;  Oldham  1966;
 Shoop,  1968; Gill 1978a;  Breden  1987; Sinsch
 1988a, 1992; Berven & Grudzien 1990; Denton and
 Beebee 1993; Sinsch and Seidel 1995).  Some adults
 even return to breeding  sites several years  after their
 obliteration (Heusser 1960, Shoop 1968).
  All habitat traits being equal, dispersal to and
 colonization of a restoration wetland decreases as a
 negative  exponential   of distance  (Wolfenbarger
 1949, Breden 1987, Berven and Grudzien  1990)
 from source populations. Colonization  is also a
 function of the probability that breeding amphibians
 actually find the mitigation site which, in turn, is the
 product of the combined probability that  the site is
 reachable and that dispersers travel towards the site
 from source  wetlands. This collective probability is
 described by the function P = E [1- exp (- n  pd2)],
 where E is the percent  of an amphibian population
 emigrating from source wetlands, p is the  density of
 randomly distributed wetlands within a given area,
 and d is a species' maximum dispersal distance dur-
 ing a single generation (Travis 1994).

  Dispersal   distances  from  amphibian recaptures
 and sightings are presented in Table 1. Theoretical
 probabilities for successful dispersal  under ideal
 conditions can be calculated  from the equation
 above using  Table 1 data and other required land-
 scape and species data.  Clearly,  the  greater and
 closer the number of source wetlands within the
 normal  dispersal distance of a targeted species, the
 greater  the  potential for successful colonization.
 However, until  dispersal distances  for all Pacific
 Northwest wetland-breeding species are  known, a
 1,000 m  maximum distance between  source and
mitigation wetlands is suggested in undisturbed for-
ested landscapes. This recommendation is based on
dispersal distances and gene  flow calculations  in
other species (Gittins  et al.  1980,  Berven and
Grudzien  1990,  Gibbs  1993,  Beebee  1996). Note
however,  that successful  colonization is signifi-
cantly reduced below optimum  dispersal  distances
by roads, fences, housing, farmlands, clearcuts and
other barriers.

  Corridors. Corridors theoretically channel dis-
persing  individuals to newly  created and restored
wetlands, provide migration routes between wetland
breeding, feeding and hibernation patches, and may
increase suitable habitat size for amphibians. Routes
used by amphibians may  vary widely  with animal
thermal tolerances, often depending on acclimation
to specific  latitudes, altitudes and other regional or
localized climatic conditions (Hutchison and Dupre'
1992 and  others referenced therein,  Duellman and
Trueb 1986). Corridors should provide cool, moist
microclimates because  amphibians'  highly perme-
able integuments require that most species have ac-
cess to  moist environments to maintain  or restore
lost body water and to favor adequate cuticular res-
piration (Shoemaker et al.1992).
  Amphibians frequently  migrate  within distinct
habitats between wetlands and uplands during suc-
cessive  years. Semlitsch (1981) found that the Mole
Salamander A. talpoideum, selected  wooded  rather
than open grassy  areas.  Kleeberger  and Werner
(1983)  found  that the  Spotted Salamander  A.
maculatum and  Gittins  et al.  (1980) that the Mar-
bled Salamander A. opacum  Gravenhorst, and B.
bufo prefer moist  soils. Shoop (1968) found that
although A. opacum limited migration to a narrow
10-30 m-wide corridor  near wetlands, it neverthe-
less  crossed open  fields between wetlands and for-
ests. Beneski et al. (1986) found no preference in A.
macrodactylum for habitat quality during migration.

  Most amphibians migrate and disperse during wet
conditions and don't seem to distinguish between
areas with  or without cover during these times. Un-
der drier conditions, however,  amphibian physiol-
ogy, and these studies suggest that animals will be
less  stressed, and hence prefer  to move  within
cooler,  moister  forested and vegetated areas. Nev-
ertheless, they  will traverse  roads and meadows
and, when  necessary, tolerate intervening unsuitable
habitats for limited lengths of time.

  Ambystoma spp. frequently exhibit  identical  mi-
gration  routes between years (A. opacum Douglas
and  Monroe 1981, A.  maculatum, A. opacum

Stenhouse 1985; A. maculatum Shoop 1965, 1968,
Shoop and Doty 1972; A maculatum Kleeberger and
Werner 1983;  A. macrodactylum Beneski et  al.
1986). Some species, including the eastern narrow-
mouthed toad (Gastrophryne  carolinensis  Hoi-
brook),  and   the  striped  newt  (Notopthalmus
perstriatus  Rafinesque) enter wetlands  to breed
from all directions, but generally disperse to upland
sites along select routes (Dodd and Charest 1988).
Conversely, others (B. bufo) migrate to wetlands in
a straight line,  but move away toward upland sites
along more undefined paths (Sinsch 1988a). Conse-
quently, identifying and protecting routes between
wetlands, and between wetlands and upland feeding
and hibernation sites,  is important.

  Roads and increasing traffic volumes are signifi-
cant contributors to amphibian population declines
(Langton 1989, Fahrig et al. 1995). Enabling unhin-
dered movement of animals with distinct migration
routes across highways and roads  through specially
constructed tunnels and underpasses has  decreased
mortality and facilitated dispersal by many species
(Langton  1989).  Unhindered migration  in  frag-
mented habitats also reduces homozygosity result-
ing   from   reproductive  isolation.   Populations
surrounded by roads exhibit higher levels  of genetic
distancing from reduced gene flow than those con-
nected by woods and  other favorable corridors (Reh
1989). Metal culverts, tunnels and fences, however,
may confuse amphibians  using magnetoperception
(e.g., B. bufo) for orientation  and should be avoided
(Sinsch 1989).

  For species that do  not exhibit clear migration or
dispersal routes, undeveloped flood zones,  riparian
strips and other shrublands and forests adjacent to
rivers, streams  and creeks are ideal corridors. These
provide cool, moist microclimate conducive to am-
phibian transit (Hurlbert 1969,  Kleeberger  and
Werner 1983,  Sinsch 1988a) and may already re-
ceive regulatory protection. Moreover, floodplains
and riparian habitats may deter  upland  predators
normally not making regular use of wet  habitats
(Forman 1983). From  riparian  corridors, forested
draws and  other  vegetated depressions may direct
amphibians upslope to feeding and refuge  habitat.

  Although dispersal  distances in amphibians may
exceed 1,000 m (Table  1); individuals of most taxa
remain within  1,000  m of wetlands. Hence, corri-
dors between wetlands, or between suitable home-
range habitats, do not need to exceed this distance.

  Corridor widths should provide a core  area  of
large diameter  logs and other coarse wood to pro-
                        Lentic-Breeding Amphibians
vide refuge habitat - particularly cover to  protect
from desiccation during summer droughts (Dupuis,
et al. 1995), cover to minimize predation, and cover
to limit potential agonistic behavior between spe-
cies, sexes and age  classes. Unfortunately, refuge
requirements for minimizing predation, competition
and agonistic behavior are not known.

  A 30-m core corridor width is suggested to take
competition, predation and other potential density
dependent  regulating mechanisms  into considera-
tion. To maintain soil moisture, retain cool micro-
climates, and provide woody debris for amphibians
establishing temporary refuges during migration and
dispersal, buffers should flank the corridor. Moreo-
ver, to protect vegetation in the core from  the ef-
fects of wind and evapotranspiration, buffer widths
of 2-3  times the height of  adjacent vegetation are
suggested as a rule-of-thumb in high-density stands
(Fritschen  and  Edmonds 1976) and 5 times the
vegetation height in low-density stands (Reifsnyder
1955).  Buffer widths of 60-120 m to maintain soil
moisture and ground temperature, 120-180 m to sta-
bilize air temperature and humidity, and 240  m to
negate  vegetation responses have been identified to
insulate old-growth Douglas-fir  (Pseudotsuga  men-
ziesii) forests from the climatic  effects  of clear-cut
edges (Chen et al.  1990, Chen et al. 1992). Conse-
quently, to protect amphibians in corridors  from
changes in soil moisture and air temperatures, total
corridor widths of 150 to 270 and 270 to 390 m re-
spectively are  required. To retain  intact  and self-
sustaining vegetation in the core a total corridor
width of 510 m would be required.
  These widths  minimize structural  and functional
changes to a core but may not  be essential to mi-
grating and dispersing amphibians.  These animals
will  move through  sub-optimum  vegetation  and
climatic conditions of narrower corridors not pro-
tected  by wide  buffers. Moreover,  depth-of-edge
microclimatic  influences are  also dependent  on
vegetation beyond the buffer's edge. Large shrubs,
and trees adjacent to the buffer's outer edge can re-
duce buffer-width requirements (Wales 1972, Ran-
ney et al.  1981, Forman 1995) without disturbing
climatic conditions within  the core. As a compro-
mise, corridor widths of 150m are suggested with
reduced widths proportional to  the insulative prop-
erties of vegetation  adjacent to core's buffer. Not
surprisingly, meadow  and grass communities  are
poor habitats, exposing  amphibians to greater cli-
matic extremes  and predators  and should not be
used as corridors or count as protective buffers to

 Lentic-Breeding Amphibians
 Watershed Conditions to Protect and Maintain
 Restoration Wetlands
  To provide the full range of biological functions
 of consequence to amphibians the restoration wet-
 land should be located within a watershed  basin in
 which imperviousness (the sum of roofs, sidewalks,
 parking lots  and  roads) does not  exceed  10 to
 15 percent (Schueler   1994,  Richter  and Azous
 1996), unless specific methods are used to minimize
 hydrological and water-quality impacts on wetlands.
 In depressional  wetlands, greater  imperviousness
 leads to elevated discharge and increases  wetland
 water level fluctuations. In  slope wetlands, more
 runoff from impervious surfaces increases current
 velocities.  Both  diminish   suitable  amphibian
 breeding, feeding and rearing habitat (see hydrology
 section).   Moreover,  wetland  invertebrates  and
 plants also decrease in richness and abundance with
 greater water-level fluctuations and concomitant
 pollution loads (Schueler 1994, Ludwa 1994, Hicks
 1995), further reducing amphibian habitat quality.

 Wetland Design Criteria

 Breeding Habitat
  Hydrodynamics and Hydrology.  Hydrodynamic
 and hydrologic modifiers of importance  to amphibi-
 ans include current velocity and the duration, depth
 and frequency of flooding. Collectively, these  fac-
 tors directly influence spawning and egg survivor-
 ship,  and indirectly  affect  breeding  success by
 determining vegetation zonation and the outcome of
 competition and predator-prey relationships.

  Current Velocity.  Lentic  breeding  amphibians
 spawn only in vernal ponds, depressional wetlands,
 or in slow-moving or quiescent water  of riverine
 backwaters and  slope  wetlands  (Savage  1961,
 Nussbaum et al. 1983, Blaustein et al. 1995a, Olson
 and Leonard 1997). Increased discharge to riverine
 and slope wetlands can  increase current  velocity
 preventing breeding, reducing the success of fertili-
 zation, dislodging  eggs  from oviposition  sites, or
 physically damaging  eggs  with suspended  silt,
 sediment and large floating debris (Lind et al. 1996,
 Richter personal observation). Velocities exceeding
 5 cm/second preclude breeding by both R. aurora
 and A. gracile (Richter and  Roughgarden unpub-
 lished manuscript), suggesting that currents through
 mitigation wetlands should not exceed  this thresh-
 old.  Flow  regimes  with current  velocities  ap-
 proaching  2  cm/second may  be optimum in  that
 slight  currents are beneficial in flushing  silt  and
suspended  solids, and  in delivering  oxygenated
water  to  eggs.  Modest currents resist freezing,
thereby providing  open  water  for early  spring
spawning. However currents  can also reduce tem-
peratures, prolonging time to hatching and meta-

 Duration of Water.  All native  lentic-breeding
Pacific  Northwest  amphibians  use  permanently
flooded wetlands. Successive years  of inundation
are essential only to A. gracile, some A. macrodac-
tylum populations that  have  more than a  1-year
larval period, neotenic T.  granulosa, and high-ele-
vation species that overwinter in wetlands. Semi-
permanence is beneficial to many species because it
precludes the establishment of fish, (e.g., Salmoni-
dae),  amphibian (e.g., R.  catesbeiana, A. gracile)
and invertebrate (e.g., crayfish, dragonflies) preda-
tors.  Consequently,  semi-permanent marshes  and
swamps make ideal mitigation wetlands for most
Puget  Sound  Basin  species, whereas  permanent
wetlands may  be targeted  for A. gracile. Rana
catesbeiana and A. gracile may be a threat to other
amphibians in small and structurally simple perma-
nent wetlands. Competitive exclusion between these
predators and other amphibians has yet to be estab-
lished in  large,  structurally  complex wetlands
(Richter and  Azous  1995, Richter personal obser-
vation). Several closely situated wetlands with dif-
fering hydroperiods, or one two-celled wetland  in
which one cell is vernal and the other remains per-
manently  flooded,  would provide  breeding  and
spawning habitat for the full compliment of native

  Water Depth. Pacific Northwest amphibians ex-
hibit distinct minimum and maximum water depths
beyond which oviposition normally does not occur
(Table 2).  From these  and other surveys (Cooke
1975, Scale 1982, Waldman 1982) it can be inferred
that oviposition of most temperate amphibian spe-
cies is at depths between 10 and 100 cm. Under ex-
ceptional circumstances, spawn of most species may
be found in very shallow (e.g., Licht  1969) water
such as flooded ruts and  roadside ditches  (Richter
personal observation) or  very deep in lakes (e.g.,
200-300  cm below the  surface [Crisafulli personal
communication]),  presumably because oviposition
sites  at preferred water depths are unavailable. In
these unique situations however, successful egg and
larval development through metamorphosis may be

  Breeding amphibians may  select medium-depth
water because  of  optimum  temperatures that

                                                                               Lentic-Breeding Amphibians
Table 2. Oviposition Water Depths of Selected Pacific Northwest Amphibians
  Water Depth
 Depth Below
 Surface (cm)"   Location
 Ambystoma gracile
  Ambystoma macrodactylum

  Rana aurora
  Rana cascadae
  Rana catesbeiana
  Rana sphenocephala

  Hyla regilla

[3-39] (11)
[30-152] [61-122]
[26-47] (28)


(32) & (40)

[5-45] (20)

12-65          California
tracks          Washington
Slater 1936
Richter &
Crisafulli personal

Anderson 1967
Richter & Leonard
personal observation.
max 152        British Columbia   Licht1969
tracks          Washington       Richter &
substrate'       -—             Roughgarden
—            Oregon           unpublished
[15-38]                         manuscript
                               Storm 1960
substrate but    Oregon
surface to       Washington
 Sype 1974
 Richter personal
(4)&(15)       South Carolina    Caldwell 1986
 Richter personal

  Bufo boreas
                                                 Olson 1988
  a [Range], (mean).
  b Personal observation.
  °Eggs typically a fixed distance above the bottom.

 Lcntic-Breeding Amphibians
 accelerate  egg development (Duellman  and Trueb
 1986),  and  possibly  because  such  depths  are
 unavailable to some predators. Medium-depth water
 may  also  indicate that  wetlands  are only  semi-
 permanently flooded, thereby eliminating the threat
 of predation, yet retaining sufficient water through
  To  encourage spawning  by Puget  Sound  Basin
 amphibians, mitigation wetlands with depths of 10-
 50 cm should be available from December through
 May. Terraced shores of 10-20,  30-40 and 50-60 cm
 depths, or alternately,  a gradual sideslope gradient
 of 10-horizontal to 1-vertical or shallower slope, to
 a total  depth  of 50 cm, should provide spawning
 habitat for all species. Commonly used 3-vertical to
 1-horizontal or steeper sideslopes provide little op-
 timum preferred depth and should not be graded at
 proposed breeding sites.
   Water Level Fluctuation. Water-level fluctuation,
 especially  maxima and minima, are vitally  impor-
 tant to  wetland biota (Weller 1990). Extremes may
 directly harm amphibians or decrease habitat value
 through simplified floral communities (Azous and
 Richter 1995), reduced food abundance and  variety
 (Munn  and Brusven 1991,  Ludwa 1994), and al-
 tered spatial relationships among physical elements
 (Reinelt and Horner 1995).
   Many temperate  latitude amphibians  may mini-
 mize exposure of  eggs  to fluctuating depths and
 temperatures  by  spawning not only  in mid-depth
 water but also by initially or permanently submerg-
 ing eggs below the surface (Table 2).
   Controlled experiments with several caudates re-
 veal  distinct  and narrow  oviposition preferences
 within the water column. In aquarium studies using
 simulated  vegetation, Miaud (1995), for  example,
 found that three species of Triturus spawn within 10
 cm  of the surface.  My field experiment  using
 wooden dowels as  surrogate vegetation, (Richter
 and Roughgarden  unpublished  manuscript) show
 that every A. gracile breeding female during a given
 night attaches its eggs at  almost  identical  depths
 (i.e., 7.06 cm ±1.27 cm, N=19) below the surface.

  Amphibian egg development  depends on perma-
 nent or partial submergence. Therefore, water levels
 should be stabilized from spawning through hatch-
 ing, which for most Puget Sound Basin  species is
 from  mid-December through  mid-May.  Although
 mean water level fluctuations exceeding 20 cm have
 been correlated to decreased amphibian richness in
 wetlands (Richter  and  Azous  1995, Azous and
Richter  1995), our  experiments  suggest that  ex-
tended drops of more than 7 cm from oviposition
through hatching may kill A, gracile eggs. More-
over,  eggs of A.  macrodactylum and H.  regilla
spawned in shallow water are harmed by  stranding
and desiccation on shore (Richter personal observa-
 Water level declines may also leave eggs spawned
within the water  column  closer  to  the surface,
thereby increasing exposure to  freezing and  to
greater  temperature fluctuations.  Finally, short-
duration water fluctuations may not kill  eggs out-
right but may leave eggs above the surface so that
their weight pulls  them from attachment  sites and
they sink to  the substrate when water levels rise
(e.g.,  A. gracile,  early development stage of  R.
aurora) or float to the surface  (e.g., late develop-
ment stage of R. aurora) (Richter and Roughgarden
unpublished manuscript). At the surface, currents
and wind may route free-floating eggs to shallow
water and other unfavorable  environments, whereas
on the substrate they may be prone to sedimentation
and colder water where eggs develop slower or fail
to develop at all (Richter and Roughgarden unpub-
lished manuscript).

 Water Quality. Amphibians are found in water of
widely varying chemical composition (Cooke and
Frazer  1976).  Ildos  and  Ancona (1994) found
chemical water quality in 42 wetlands well within
the range of  amphibians' tolerability. Richter
(unpublished data) also found water chemistry not
to be directly limiting amphibian  distribution and
spawning when  comparing pH, conductivity, alka-
linity,   TSS,  chlorophyll   a,   ammonia-nitrogen,
nitrate+nitrite-nitrogen,  soluble  reactive  phospho-
rus, total phosphorus  P, fecal  coliforms,  entero-
cocci, microtox, and metals (cadmium Cd, copper
Cu, lead Pb,  zinc  Zn) in 19 wetlands (Reinelt and
Horner 1990) against concentrations  known to be
harmful  (Power et al.  1989).  Within these same
wetlands, however, Azous (1991), found a signifi-
cant negative correlation between amphibian rich-
ness  and water  column conductivity.  Moreover,
Platin (1994) and Platin and Richter (1995) found
R. aurora embryo mortality positively correlated to
a principal  water  quality  component comprising
conductivity, Ca, Mg, and pH,  and negatively cor-
related  to a second principal component including
total P, total  suspended solids, Pb, Zn, Al, total or-
ganic content and dissolved oxygen. Interestingly,
A. gracile egg mortality under similar  conditions
showed  no  correlation  with  these two principal
components but instead correlated with total petro-
leum  hydrocarbons and fecal coliforms.  These  and

other studies (Strijbosch 1979, Beattie and Tyler-
Jones 1992, Rowe et al. 1992, Sadinski and Dunson
1992, Rowe and Dunson  1993,  1995) suggest that
some  species'  distribution and  breeding  success
may be predicted locally by water quality, most no-
tably  conductivity,  pH,  Al, total  cations,  NO2,
chemical  oxygen  demand  and  dissolved  organic

  Because heavy metals and other pollutants adhere
to  suspended solids,  removing  sediments  before
they enter mitigation  wetlands may protect  water
quality. The  settling rate of suspended solids is  a
function of vegetation and  current velocity, which
in turn is determined by slope. Consequently, mod-
erately vegetated or totally  submerged hydrophytes
at emergent wetland gradients of roughly 0.0012 to
0.0006 will provide slope conditions to remove de-
positional sediments in 15-61 cm depth waters  of
2.4- to 6.1-m-wide channels (Adamus et al. 1987).
  Methods to reduce suspended  solids and pollut-
ants at mitigation wetlands can also include captur-
ing runoff in detention facilities and wet ponds,  or
by channeling  water  through biofiltration  swales
prior to wetland discharge (Horner  1988). Swales
should be less than 10 cm deep or have current ve-
locities exceeding  5 cm/s  to deter breeding am-
phibians from placing their eggs in  danger. Short,
50-cm-high   fences  are   also  suggested  around
stormwater  ponds  with widely  fluctuating  water
depths  to deter access by  native  amphibians and
around  facilities with  stable water to prevent colo-
nization by introduced R. catesbeiana (Richter per-
sonal observation).
  Other than outright  death from  toxic spills and
sediment deposition, direct relationships  between
water quality, amphibian distribution, and egg sur-
vivorship remain complex, and may be a reason for
the absence of water-quality criteria for amphibians
(Boyer and Grue 1995). Consequently, until causal-
ity between  water-quality  factors  and amphibian
success  are  assessed  under controlled  conditions
(both in isolation and in  synergy with each  other,
and under diverse  hydrological conditions) it may
be prudent  to meet or exceed  water-quality stan-
dards similar to those suggested to protect human
health by federal, state and local jurisdictions.

  Vegetation. Amphibian  presence at  wetlands is
correlated  with vegetation  that provides shelter,
calling locations, spawning sites and feeding possi-
bilities   (Ildos   and  Ancona 1994).  Vegetation
classes, cover, structure and species may all signifi-
cantly  affect breeding and larval  success. Pacific
                        Lentic-Breeding Amphibians
Northwest  caudates  attach their eggs directly to
vegetation  within the water column (Slater 1936,
Anderson 1967, Richter and  Roughgarden unpub-
lished  manuscript,  Richter and Leonard,  unpub-
lished  data),  whereas  anurans  anchor  eggs  to
vegetation either below or near the surface (e.g., R.
aurora, B.  boreas), or  occasionally  spawn  free-
floating eggs (R. pretiosa; Licht 1969).

  Vegetation Classes.  Aquatic-bed and  emergent
vegetation classes are significant in accounting for
amphibian   richness and breeding  at  wetlands,
whereas the total number of habitat  classes per
Cowardin et al. (1979) is unimportant in the Puget
Sound Basin (Richter and Azous 1995). Vegetation
type is one of the most important characteristics
determining the distribution of seven amphibians in
Italy (Ildos and Ancona  1994). Strijbosch (1979)
found  egg masses of all nine species surveyed oc-
curred disproportionately within particular hydro-
phyte communities in the Netherlands. Interestingly,
Strijbosch found a clear correlation between certain
hydrophyte  communities and the chemical/physical
water-quality data. This confounded direct vegeta-
tional  relationships but highlighted the  complex in-
teractions  of  site  criteria that guide amphibian
habitat use.
  Cover and Interspersion. Although not well quan-
tified,  vegetation cover may be one of the most im-
portant factors  determining  amphibian  breeding
distributions. Most amphibians generally avoid both
open water and densely vegetated sites, instead se-
lecting habitats  with an interspersion  of both fea-
tures  (Strijbosch 1979, Ildos  and Ancona  1994,
Richter  personal  observation).  Comparisons  of
vegetation cover within V4 -m2 circular plots sur-
rounding A. gracile, Rana aurora and R. catesbe-
iana eggs to cover within 1A -m2 randomly chosen
sites, suggests that both dense (95-100 percent) and
light (0-5  percent) cover is  avoided (Richter per-
sonal observation). This may be to reduce predation
in exposed or  highly vegetated areas. In contrast,
some  species  may  exhibit different preferences  in
different parts of their range. For example, B. bufo
was  the  only  species among nine  amphibians
spawning   in  open  water   in  the  Netherlands
(Strijbosch  1979), but was found spawning among
vegetation in England (Beebee 1983).
  Nevertheless,  these  findings  suggest  that most
species may prefer interspersed  open  water and
vegetation  for  oviposition.   Therefore,  a 50:50,
25:75, or 75:25 ratio of open water to  vegetation  is
suggested to encourage spawning.

 Lentic-Breeding Amphibians
  Plant Species.  Amphibians oviposit on a variety
 of plants, although within particular wetlands only
 certain species may be used. Among the caudates,
 Miaud (1995) found the warty newt (Triturus cris-
 tatus) attaching its eggs entirely on one plant spe-
 cies (e.g., Nasturtium) whereas eggs of the palmate
 newt (T. helveticus helveticus Razoumowski) were
 found on four species. Strijbosch (1979) found T.
 cristatus, the alpine newt  (T. alpestris Laurenti),
 and the smooth  newt (T. vulgaris L.), ovipositing
 primarily on reeds (Phragmites spp.). Richter and
 Roughgarden (unpublished manuscript), however,
 found eggs of A. gracile attached to a minimum of
 28 species of herbs, shrubs and trees in 11 wetlands.
 Richter and Leonard  (unpublished data) found A.
 macrodactylum  spawning  on  minimum of  eight
 plant  species  in just one wetland  and  Richter
 (unpublished data) found its eggs on at least four
 different plant species at a second wetland.
   Anura also use a diversity of species for oviposi-
 tion. Licht (1969) found R.  aurora eggs attached to
 submerged  stalks of Typha, Carex and  Potamoge-
 ton,  but Richter and Roughgarden  (unpublished
 manuscript) rarely found R.  aurora  using Typha
 spp., despite its  abundance. Rather, they observed
 wide use of less common Oenanthe, Polygonum,
 Juncus, Scirpus, and Carex. From these surveys and
 studies it can be  inferred that plant species per se is
 unimportant for  oviposition. This frees  mitigation
 ecologists to select any plant species that grows un-
 der the specific  hydrologic conditions required by
 breeding amphibians.

  Species Structure.  Experimental  evidence sug-
 gests  that vegetation structure,  particularly  plant
 shape and stem diameter, are the oviposition criteria
 most important to caudates. The role of vegetation
 structure in anuran  oviposition remains undeter-
 mined. In laboratory tests a linear shape  is selected
 over an arborescent shape by ovipositing T.  helveti-
 cus,  T.  cristatus and T.  alpestris (Miaud  1995).
 Moreover, the thinnest PVC tubing (i.e., simulated
 vegetation)  of 25 u was  preferred by the smallest
 species, T. helveticus, whereas medium tubing of 50
 |j was preferred over thicker tubing of 100  u by T.
 alpestris and T. cristatus.

  Wetland  surveys and controlled field  studies of
 several Pacific Northwest salamanders also confirm
 that distinct  stem widths are preferred by oviposit-
 ing caudates. Field surveys of A. gracile egg masses
 indicated  that the stem diameters of the 28 plant
 species used ranged  between 1 and 8  mm,  and for
 all  species averaged 3 to 4 mm. Moreover, under
controlled  field  studies using  3,  6, and  9 mm
wooden dowels as surrogate hydrophytes, an over-
whelming preference for 3-mm dowels over 6-mm
diameter dowels was found, with 9 mm dowels al-
most totally avoided (Richter and Roughgarden un-
published). Interestingly, the smallest caudate in the
Puget Sound Basin, A. macrodactylum, oviposits
primarily  on very thin (i.e.,  1 to 2 mm diameter
vegetation [Richter unpublished data]) supporting a
direct relationship  between  salamander size and
oviposition substrate size.

  It is clear that Pacific Northwest amphibians use
the most commonly available thinstem hydrophytes
and small twigs and branches available in benign
hydrodynamic and hydrologic  spawning conditions.
Stems need to be grasped readily by females,  hold
eggs firmly, and allow eggs to float within the water
column to fluctuating depths and currents. A partial
list of recommended  shallow-marsh hydrophytes
used by Pacific Northwest amphibians is provided
in Table 3.
  Hardback, (Spiraea douglasii), willow (Salix spp.)
and other shrubs and trees with rigid woody stems
are not  prescribed. Finally, mitigation should not
use aggressive smartweeds (Polygonum spp.) and
cattail (Typha latifolia). Reed canarygrass (Phalaris
arundinaced), yellow flag (Iris pseudacorus), loos-
estrife (Lythrum salicaria, Lysimachia vulgaris) and
other exotics. If they are  found at mitigation  wet-
lands, they  should be excluded  through  selective
vegetation management.

  Interim oviposition sites may be provided at newly
restored and created wetlands by adding thin,  dead
branches and twigs from hardback, red alder (Alnus
rubrd),  western redcedar  (Thuja plicata)  western
hemlock (Tsuga heterophylld) and other species at
spawning locations.

Wetland Layout

  Size.  Given that all  other  habitat features are
equal, wetland size is unrelated to amphibian rich-
ness (Sjogren 1991, Ildos and Ancona 1994, Richter
and Azous 1995). Hence, there is no minimum area
required by breeding amphibians (Cooke and Frazer
1976, Andren and  Nilson  1985,  Denton  1991,
Richter and Azous 1995). Smaller wetlands may be
used more than larger ones by some species (Cooke
and Frazer  1976,  Moler and  Franz  1987) because
larger, and consequently often permanent wetlands
are suitable for predators such as fish, reptiles and
amphibians requiring permanent water. Ambystoma
macrodactylum may be especially abundant in small
wetlands  (e.g., 201  different individuals  captured
over 25  nights in a  130-m2 pond; Leonard and

                                                                             Lentic-Breeding Amphibians
Table 3. Puget Sound Lowland Shallow Marsh Hydrophytes Used as Amphibian Oviposition Sites.
      Amphibian Species
    Additional Material
    Scientific Name
                                                   Common Name
 Ambystoma macrodactylum

 Hyla regilla

 Rana pretiosa
                             1 to 2 mm Diameter, Thin-stemmed Emergents
 Roothairs, rootlets & thin

 Some roothairs, rootlets &
 thin twigs.

 Some variable-diameter
 Alopecurus aequalis
 A. geniculatus
 Agrostis aequivalvis
 Glyceria grandis
 G. elata
 G. occidentalis

 Carex obnupta
 Eleocharis palustris
 E. ovata
 E. acicularis

 Juncus acuminatus
 J. bufonius
  Shortawn foxtail
  Water foxtail
  Alaska bentgrass
  Bluejoint reedgrass

  American mannagrass
  Tall mannagrass
  Western mannagrass

  Slough sedge
  Common spikerush
  Ovate spikerush
  Needle spikerush

  Tapered rush
  Toad rush
  Ambystoma macrodactylum

  Rana aurora
3 to 6 mm Diameter, Medium-stemmed Emergents
                           Carex athrostachya
  Some variable-diameter
  C. utriculata

  Carex obnupta

  Menyanthes trifoliata

  Oenanthe sarmentosa

  Polygonum amphibium
  P. hydropiperoides
  P. punctatum

  Potamogeton natans

  P. emersum
  P. gramineus
                                                       Sparganium emersum
                                                       S. eurycarpum
  Slender-beaked sedge
  Beaked sedge

  Slough sedge



  Water smartweed
  Dotted smartweed

  Floating-leaf pond
  Emersed pondweed
  Grass-leaved pond

  Simple:stem Bur-reed
  Broadffuit Bur-reed
  Taricha granulosa

  Bufo boreas
          Variable Diameter Emergents
  Leafy submerged plants.    CYPERACEAE
                            Scirpus cyperinus
                            S. microcarpus
  Grasses & leafy grass-like
                         SEDGE FAMILY
                          Small-fruited Bullrush

 Lentic-Breeding Amphibians
 Richter unpublished data). Nevertheless, there may
 be instances where some species avoid very small
 pools  (e.g., Triturus <100  m2 [Cooke and Frazer
 1976]). Other species with strong territorial males
 (e.g., introduced R. catesbeiana, and green frog, R.
 damitans  Latreille),  may  be  less  numerous  in
 smaller than in larger wetlands.
  Because  water depth  and seasonal availability,
 interspersion  of open  water and vegetation, and
 specific vegetation  structure  are  all  important
 breeding criteria, wetlands should be big enough for
 these attributes to coexist.
  Configuration. Amphibians favor shorelines for
 oviposition because they spawn in specific vegeta-
 tion at particular depths. Moreover, food and shelter
 for larvae  is more frequently found along vegetated
 shorelines. Consequently, configuration (e.g., length
 of shoreline  in relation to  area) is an important
 characteristic  that   should  maximize  spawning,
 feeding and  cover  habitats.  Additional  shoreline
 habitat can be  created by establishing long lobes
 and scalloped edges.
  Orientation. Egg  development is  a function  of
 water  temperature   (Herreid and  Kinney  1967,
 Brown 1975, 1977).  Clutch numbers increase with
 temperature (Albers and Prouty 1987, Sjogren et. al.
 1988, Sjogren  and  Larsson  1988),  with  warmer
 northern shores exhibiting the highest numbers of
 eggs  among spring-breeding  species (Stenhouse
 1985, Sinsch 1988b).
  Given that all other habitat features are  similar,
 Pacific Northwest amphibians (as well as  those in
 other  higher-latitude regions) show a clear spawn-
 ing preference for northern quadrants of wetlands,
 although the importance of orientation decreases
 with wetland size  (Richter  and Roughgarden un-
 published  manuscript). In controlled  field experi-
 ments where A. gracile could  choose to  oviposit
 eggs along both north and  south shores  and fur-
 thermore select between northwestern and north-
 eastern shorelines, 98 percent of eggs were spawned
 along  the north  shore, and 68 percent of these were
 attached to sites  along the northwestern shore
 (Richter  and  Roughgarden  unpublished  manu-
 script). Presumably the breeding preference for the
 northwestern wetland locations is a consequence of
 higher water temperatures from solar radiation.

  Consequently, mitigation wetlands should provide
 a gradual shallow slope along the northern shore.
 Oval wetlands with  a long axis from  southwest to
 northeast will optimize  preferred  breeding habitat
 along the northwestern side of wetlands. Podloucky
(1989) suggests  500 to 1,000 m2 ponds with, east-

west to north-south ratios of 2:1 to maximize sunny
areas along the northern shore.
  Optimizing water depth through  wetland  bathy-
metry (i.e., gradual side-slope with optimum water
depth and vegetation plantings etc.) along the north-
ern shore and increasing solar radiation through
vegetation management can further raise springtime
water temperatures.

  Buffers. Buffers are essential components of wet-
lands (Wilard and Killer  1990), providing separa-
tion between wetlands and developed environments,
moderating  water level fluctuations,  and trapping
sediments and insoluble pollutants (Castelle et al.
1992). In highly urbanized  areas, wetland buffers
also may be the only  terrestrial habitat  remaining
for amphibians (Richter personal observation); ad-
jacent upland feeding, aestivation and hibernation
sites  may be inaccessible   or  destroyed. In the
Pacific Northwest, R. pretiosa and  R. cascadae
(Dumas 1966) remain within several meters of wa-
ter throughout the year, whereas in other regions,
the  Columbia spotted frog  R.  luetraventris, R,
catesbeiana, R.  damitans (Richter personal obser-
vations) and other Ranids, (Beebee 1996) remain at
the wetland buffer throughout most of their lives.
  Buffers  may be especially important in providing
cover  to  females and to metamorphs.  Female R.
aurora, A.  gradle (Richter personal observation)
and A. macrodactylum (Beneski et. al. 1986, Leon-
ard and Richter unpublished data)  generally wait in
buffers near wetlands until  environmental and bio-
logical conditions favor spawning. They then enter
wetlands during one or a few nights to spawn, and
quickly retreat to the cover of adjacent  buffers or
more distant vegetated upslope habitats.

  Metamorphs of many  species  may  also benefit
from wetland  buffers. They are important to the ti-
ger salamander, Ambystoma tigrinum Green, seek-
ing shelter in rodent burrows during the first days
after emigrating from natal ponds (Loredo  et al.
1996). Surveys of wetland buffers and upland habi-
tats during  summer droughts, suggests  that meta-
morphs of H. regilla, B. boreas, R. aurora, and T.
granulosa may spend several weeks in buffers and
disperse to upland sites only after rains or when the
soil and vegetation are moist (Richter personal ob-
servation).  Vulnerable metamorphs  and juveniles
may have greater cover, moisture, and food within
wetland buffers than in upland sites.

  Guidance  on  buffer  width for  overall wetland
protection is available (Brown and Schaefer 1987,
Kusler and  Kentula  1990, Castelle  et al.  1992,
Castelle  at  al.1994),  although  not  based  on

                                                                               Lentic-Breeding Amphibij
empirical  studies of the unique microclimatic and
other biological requirements of amphibians. Extra-
polations from Rudolph and Dickson (1990)  sug-
gest that 30-95 m buffers retain the full complement
of amphibians adjacent to streams. The Washington
Department of Wildlife (1992) suggests 50 m buff-
ers for R. pretiosa in western Washington and 30 m
for R.  luteiventris in eastern Washington, although
empirical  justification for  either width is unavail-

  In the absence of experimental data it seems pru-
dent to maintain a treed,  riparian zone. A buffer
equal to two to three tree heights circumscribing the
mitigation wetland  may be optimum for litter and
downed wood provision and for sustaining a moder-
ate microclimate  for migrating breeding adults and
newly metamorphosed juveniles, prior to terrestrial
dispersal. The 60-240  m  protective zone recom-
mended for migration corridors (see previous sec-
tion) can be applied to wetlands and may be wider
than the widths based on  tree  heights. Whether a
zone of this width  is necessary to provide staging
habitat for breeding and food and cover for meta-
morphs cannot be  known  until individual species
requirements are established.

  Wetland-breeding  amphibians exhibit  complex
life cycles that include ontogenetic changes in mor-
phology, physiology and behaviors. Typically  each
stage includes carnivorous terrestrial adults, seden-
tary  aquatic  spawn,  motile  aquatic  larvae  and
transitional metamorphs-all constrained by the dis-
tribution  of suitable habitat.  Wetland  restoration
and creation goals must consider habitats  to provide
for all life-stages in order to attract and sustain am-
phibians at wetlands.  Restoration must focus first
on optimizing a number of  concurrent  landscape
elements  that include the  spatial arrangement be-
tween wetland breeding, upland feeding and refuge
patches, and their interconnecting corridors through
the landscape matrix. Restoration should then  in-
clude  wetland design features  that optimize, ovi-
position and egg and larval survival.

  Theoretical  models  for  landscape  structure  of
fragmented populations are available  (Harris and
Scheck  1991, Soule  and  Gilpin  1991,  Forman
1995). Explicit empirical landscape elements, how-
ever, are  seldom considered  in mitigation (Fahrig
and Merriam  1985), although a few have been de-
veloped for  birds (Saunders  and Hobbs  1991),
mammals (Saunders and Hobbs 1991) and amphibi-
ans (Sjogren-Gulve and Ray  1996). Whereas  gen-
eral qualitative habitat needs have been described
for many amphibians,  we  know  relatively  little
about the population biology of amphibians and
how they use the landscape. Few if any empirically
based quantitative data exist on landscape attrib-
utes,  including  the effective  distances  between
wetlands, corridor dimensions or the required area
and characteristics of intervening feeding and ref-
uge  habitats.  Moreover,  conservation  biologists
continue to debate the merits of small multiple
habitat patches as  opposed to single larger patches
and the benefits of corridors (Mann and Plummer
1995). In the  interim, I assume that creation and
restoration connected to or  near natural  wetlands
will increase success. Future studies must address
such critical aspects of successful wetland restora-

  General habitat-use  patterns are species specific,
and may vary between geographic regions and years
(Aubry  and Brookes  1991 and references therein,
Brown  1997). Foremost life history adaptations of
Pacific  Northwest lentic-breeding species  indicate
late winter and early spring migrations, explosive
and  synchronous  breeding,  strategic  selection  of
oviposition sites, rapid egg and  larval development,
and an initiation of metamorphosis prior to wetland
drying  (Wilbur and  Collins 1973;  Brown  1975,
1977; Nussbaum et al. 1983; Blaustein et al. 1995a,
Olson and Leonard 1997, Richter and Roughgarden
unpublished manuscript). The literature review and
field  studies reported here  also suggest  that ovi-
position sites  are  the result of selective pressures
enabling females to spawn successfully, and eggs to
develop rapidly and safely under conditions of rela-
tively distinct  weather, wetland hydrology, water
quality, vegetation and predation (see Collins 1975,
Hairston 1987  and references therein). These find-
ings have significant management implications  for
the restoration and creation of wetlands that may be
applied to encourage amphibian reproduction.
  I am confident the data  and conservative sugges-
tions within this paper can be applied to many wet-
land-breeding  species because,  as  this   paper
suggests, specific landscape and breeding habitat
needs of Pacific Northwest species (as well as other
temperate  latitude,  lentic-breeding  species) have
more common features than differences. By pro-
viding habitat for the species exhibiting the widest
habitat  needs,  species  with more restricted  needs
can be accommodated. For example, all amphibians
in the Pacific Northwest  can successfully breed in
permanent wetlands, but not all breed in shallow
temporary wetlands. Moreover,  by providing  the
hydrologic  requirements  for A.  gracile,  all other


 Lentic-Breeding Amphibians
 species' hydrologic needs are also likely to be met.
 Similarly,  by providing the upland needs  for  B.
 boreas and R. aurora  the refuge habitat of most
 other  amphibians  may  be met. Clearly, additional
 species-specific research will find amphibians under
 a wider range of habitat conditions  (i.e., exhibiting
 wide  tolerances) because of their "plasticity" and
 adaptability. Nevertheless, as an increasing number
 of controlled studies show, amphibians exhibit clear
 and sometimes narrow habitat preferences under
 which survival excels. It is for these conditions that
 restoration wetlands should be built. Studies that
 focus on the entire range of observations, such  as
 amphibian homing rather  than  post-metamorphic
 dispersal,  may greatly overestimate the potential for
 wetland colonization and population stability  of a
 species. Hence,  for  wetland  restoration  I recom-
 mend species-specific  optimum  conditions when-
 ever  they  have  been  identified.   Suggested
 recommendations should  further be adjusted to re-
 flect local climatic conditions  and the breeding and
 spawning  habitat preferences of endemic popula-
   The historical absence of  carefully designed em-
 pirical methods for identifying movement corridors,
 refuge habitat and spawning requirements makes it
 difficult to differentiate  discrete optimum condi-
 tions from those used merely  because of  availabil-
 ity.  Clearly, the range of circumstances over which
 species  are found and breed in nature may not re-
 flect optimum environmental  factors (see above).
 They  may, in fact, differ  from that  observed under
 controlled experiments  and represent conditions un-
 der  which species may  not maintain themselves. To
 determine optimum conditions, site selection and
 designs criteria can be  selected with explicit quan-
 titative characteristics in  mind. These can then  be
 monitored at the restoration wetland for success and
 thereby facilitate  comparisons with natural condi-
 tions to improve amphibian restoration.

   Suggestions that benefit amphibians favor other
 wildlife and vegetation.  For example, breeding birds
 (Brown, and Dinsmorel986)  and mammals (Kent
 1994b) are more abundant  when distance to adja-
 cent wetlands  is less and density of wetlands in-
 creases. Native plant species richness also  increases
 with proximity to  the nearest seed source  (Reinartz
 and  Warne 1993), with wetlands of significantly
 more species per unit area in locations where other
 wetlands are close (M011er and R0rdam 1985).

  Finally, siting mitigation wetlands to benefit am-
phibians by the suggested criteria will preserve up-
land habitat, contribute  undeveloped areas of open
space,  and provide  places for passive  recreation.
Maintaining urbanizing areas in forest cover also
increases   stormwater  infiltration,   promoting
groundwater recharge and flood control. These im-
portant landscape features,  which are disappearing
from  watersheds  under increasing  development
pressures, should be protected and maintained.

 I thank Deanna H. Olson  for providing construc-
tive suggestions on an earlier draft of this manu-
script. A special thanks  to Richard  Robohm who
dropped everything to help me with final edits. I am
also grateful to Fred Weinmann, USEPA Region 10,
and Rich Sumner,  USEPA  ERL-Corvallis and Bill
Sanville, USEPA  ERL-Duluth,  for  providing re-
search funds to carry out the studies underlying the
suggestions provided in this paper.

Literature Cited
Adamus, P. R., E.  J. Clairain, Jr., R. D. Smith, and
  R.  E. Young. 1987. Wetland Evaluation Tech-
  nique (WET), Volume  H:  Methodology.  U.S.
  Army Corps of Engineers Waterways Experiment
  Station, Operational Draft Technical Report Y-87,
  Vicksburg, MS.
Albers, P. H. and  R. M. Prouty. 1987. Survival of
  spotted salamander eggs  in temporary  woodland
  ponds of coastal  Maryland. Environmental Pollu-
  tion 46:45-61.
Anderson, J. D. 1967. A comparison of the life his-
  tories of coastal  and montane populations of Am-
  bystoma   macrodactylum  in  California.  The
  American Midland Naturalist 77:323-355.
Andren,  C. and G. Nilson.  1985. Breeding pool
  characteristics and  reproduction  in   an island
  population of natterjack toads,  Bufo calamita
  Laur., at  the Swedish west coast. Amphibia-Rep-
  tilia 6:137-142.
Aubry  K. B. and M. H. Brookes (contributing edi-
  tors).  1991.  Wildlife and Vegetation of Unman-
  aged  Douglas-fir  forests. U.S.  Forest Service,
  Pacific  Northwest Research  Station, Portland,
  OR. General Technical Report PNW-GTR-285.
Aubry, K. B, and  P. A. Hall. 1991. Terrestrial am-
  phibian communities in the southern Washington
  Coast Range, pp. 327-338  In K.B. Aubry and
  M.H. Brookes (contr. eds.) Wildlife and Vegeta-
  tion of Unmanaged Douglas-fir forests. U.S. For-
  est Service, Pacific Northwest Research Station,
  Portland, OR. General Technical Report PNW-

                                                                              Lentic-Breeding Amphibia
Azous , A. 1991. An analysis of urbanization effects
  on wetland biological communities. M.S. Thesis,
  University of Washington, Seattle, WA.

Azous, A. and K. O. Richter. 1995. Amphibian and
  plant community responses to changing hydrology
  in urbanizing  wetlands, pp. 156-162 In E. Ro-
  bichaud (ed.)  Puget  Sound Research '95 Pro-
  ceedings, Volume 1. Puget Sound Water Quality
  Authority, Olympia, WA.

Beattie, R. C. and R. Tyler-Jones. 1992. The effects
  of low pH and aluminum on breeding success in
  the frog Rana temporaria. Journal of Herpetology

Beebee, T. C. J.  1983. The Natterjack Toad. Oxford
  University Press, Oxford, England.

Beebee, T. C. J.  1996. Ecology and Conservation of
  Amphibians. Chapman & Hall, London, England.
Bellis, E. D.  1959. A  study  of the movement of
  American toads in a Minnesota bog. Copeia 1959:
Bellis, E. D. 1962. Cover value and escape habitats
  of the wood frog in a Minnesota bog. Herpe-
  tologica 17: 228-231.
Beneski, J. T. Jr., E.J. Zalisko and J. H. Larsen Jr.
  1986. Demography and migratory patterns of the
  eastern long-toed salamander, Ambystoma macro-
  dactylum columbianum. Copeia 1986:398-408.
Berven, K. A. and T.  A. Grudzien. 1990. Dispersal
  in the  wood frog (Rana sylvcatica):  Implications
  for  genetic   population   structure.   Evolution
Bishop, S. C. 1941. The salamanders of New York.
  NY State Mus. Bull. No. 324. NY.
Blaustein, A. R., B. Edmond, J. M.  Kiesecker, J. J.
  Beatty, and D. G.  Hokit. 1995b.  Ambient ultra-
  violet  radiation  causes  mortality in  salamander
  eggs. Ecological Applications 5:740-743.

Blaustein, A. R., P. D. Hoffman, D.  G. Hokit, J. M.
  Kiesecker, S. C. Walls, and J.B. Hays. 1994b.  UV
  repair and resistance to solar UV-B in amphibian
  eggs: A link to population declines? Proceedings
  National Academy of Sciences 9:1791-1795.

Blaustein, A. R., D. G. Hokit, R. K.  O'Hara,  and
  R.A. Holt.  1994a Pathogenic  fungus contributes
  to amphibian losses in the Pacific Northwest. Bio-
  logical Conservation 67:251-254.

Blaustein, A. R., D. W. Wake,  and W. P. Sousa.
  1994c.  Amphibian  declines:  Judging  stability,
  persistence,  and susceptibility of populations to
  local and global extinctions. Conservation Biol-
  ogy 8:60-71.

Blaustein, A.  R., J. J. Beatty, D. H. Olson, and R.
  M. Storm. 1995a. The Biology of Amphibians and
  Reptiles in  old-growth  forests  in the  Pacific
  Northwest.  U.S. Forest Service, Corvallis, OR.
  General Technical Report PNW-GTR-337.

Booth, D. B.  1991. Urbanization  and the natural
  drainage system - impacts, solutions, and progno-
  ses. Northwest Environmental Journal 7:93-118.
Booth, D. B. and C. R. Jackson. 1994. Urbanization
  of aquatic systems -  degradation thresholds  and
  limits of mitigation, pp. 425-434. In R.A. Martson
  and V. R.  Hasfurther (eds.) Effects of Human-
  Induced Changes on Hydrologic Systems, Pro-
  ceedings; Annual Summer Symposium of the
  American Water Resources Association, Jackson
  Hole, WY.

Booth, D. B. and L. E. Reinelt. 1993. Consequences
  of urbanization on aquatic systems - measured ef-
  fects,  degradation thresholds,  and  corrective
  strategies, pp. 545-550 In Watershed '93, A  Na-
  tional  Conference  on Watershed Management,
  Alexandria, VA.
Boyer, R. 1993. Evaluation of water quality in rela-
  tion to frogs at Klamath Basin National Wildlife
  Refuges. M.S. Thesis. University of Washington,
  Seattle, WA.
Boyer, R. and C. E. Grue. 1995. The need for water
  quality criteria for frogs. Environmental Health
  Perspectives 103:352-357.
Bradford, D. F. 1991. Mass mortality and extinction
  in a high-elevation population of Rana muscosa.
  Journal of Herpetology 25:174-177.
Brattstrom, B. H. and J. W. Warren. 1955. Observa-
  tions of the ecology and behavior of the Pacific
  Treefrog, Hyla regilla. Copeia 1955:181-191.
Breden,  F. 1987. The  effect of post-metamorphic
  dispersal on the population genetic  structure of
  Fowler's toad, Bufo woodhousei fowleri. Copeia
Brown, C. 1997. Habitat structure and occupancy
  patterns of the montane frog. Rana cascadae, in
  the Cascade Range,  Oregon, at multiple scales:
  Implications  for population dynamics in patchy
  environments. M.S. Thesis, Oregon State Univer-
  sity, Corvallis, OR.
Brown, H. A. 1975. Embryonic temperature adapta-
  tions of the Pacific treefrog, Hyla regilla. Comp.
  Biochem. and Physiol. 51A:863-873.

Lentic-Breeding Amphibians
Brown, H. A. 1977. Temperature and embryonic de-
  velopment of the northwestern salamander, Am-
  bystoma gracile. pp. 209-217 In R.D.  Andrews
  in, R.L. Carr, F. Gibson, B.Z. Long, R.A. Soltero
  and K.C. Svedberg. Proceedings of the Sympo-
  sium on Terrestrial and Aquatic Ecological Stud-
  ies of  the Northwest. Eastern Washington State
  College Press, Cheney, WA.
Brown,  M.  T. and J.  M. Schaefer.  1987.  Buffer
  Zones  for Water, Wetlands, and Wildlife: A Final
  Report on the Evaluation of the Applicability of
  Upland Buffers  for the Wetlands of the Wekiva
  Basin. Center for Wetlands, University of Florida,
  Gainsville, FL.
Brown, M. and J. J. Dinsmore  1986. Implication of
  marsh  size and isolation for marsh management.
  J. Wildl. Manage. 50:392-397.
Caldwell, J. P.  1986. Selection of egg deposition
  sites: A seasonal shift in  the Southern Leopard
  Frog. Copeia 1986: 249-253.
Castelle, A. J., C. Conolly, M. Emers, E. D. Metz,
  S. Meyer, M. Witter, S. Mauermann, T. Erickson,
  and S. S. Cooke. 1992. Wetland Buffers: Use and
  Effectiveness. Adolfson Associates, Inc.,  Shore-
  lands  and Coastal Zone Management  Program,
  Washington Department of Ecology.  Olympia,
  WA. Pub. No. 92-10.
Castelle, A. J.,  A. W.  Johnson, and C. Conolly.
  1994.  Wetland and  stream  buffer size require-
  ments-a review. Journ. Environmental Quality.
Chen, J., J. F. Franklin,  and  T.  A. Spies. 1990.
  Microclimatic pattern and biological response at
  the  clear-cut  edges  of old-growth Douglas-fir
  stands. Northwest Environmental Journal 6:424-

Chen, J., J. F. Franklin,  and  T.  A. Spies. 1992.
  Vegetation responses to edge environments in
  old-growth Douglas-fir forests. Ecological Appli-
  cations 2:387-396.

Christein, D. and D. H. Taylor  1978. Population
  dynamics in breeding aggregations of the  Ameri-
  can  Toad, Bufo americanus (Amphibia,  Anura,
  Bufonidae) Journal of Herpetology 12:17-24.

Collins, J. P.  1975. A comparative study of  the life
  history strategies in a community of frogs. Ph. D.
  Dissertation, University of Michigan, Ann Arbor,

Cooke, A. S. 1975. Spawn site selection and colony
  size of the frog (Rana  temporaria) and the toad
  (Bufo bufo). Journal of Zoology (London) 175:29-

Cooke, A. S. and J. F. D. Frazer. 1976. Characteris-
  tics of newt breeding sites. Journal of Zoology
  (London) 178:223-236.

Corn, P.  S. 1994. What we know and don't know
  about amphibian declines in the west. pp. 59-67.
  In  W.  W. Covington and L. F. DeBano (tech.
  coords.) Sustainable Ecological Systems: Imple-
  menting  an  Ecological  Approach  to  Land
  Management, U.S. Forest Service, Rocky Moun-
  tain Forest  and Range Experiment Station, Fort
  Collins, CO. General Technical Report RM-247.
Corn, P.  S. and R. B. Bury. 1991. Terrestrial am-
  phibian communities  in the Oregon Coast Range
  pp.  305-317 In  K. B. Aubry and M.H. Brookes
  (contr.  eds.) Wildlife and Vegetation of Unman-
  aged Douglas-fir forests.  U.S.  Forest Service,
  Pacific  Northwest  Research  Station,  Portland,
  OR. General Technical Report PNW-GTR-285.
Cowardin, L.  M., V. Carter,  F. C. Golet, and E. T.
  LaRoe.  1979.  Classification  of  Wetlands  and
  Deepwater  Habitats of the United  States. U.S.
  Fish and Wildlife   Service,  Washington  DC.
Delis, P. R., H. R. Mushinsky,  and E. D. McCoy.
  1996.  Decline  of  some west-central Florida
  anuran  populations  in   response   to habitat
  degradation.  Biodiversity  and  Conservation

Denton, J. S.  1991. The distribution and breeding
  sites of newts in Cumbria, England.  Herpetologi-
  cal Journal 1:549-554.

Denton, J. S. and T. J. C. Beebee. 1993. Density-
  related features of natterjack toad (Bufo calamita)
  populations  in  Britain.  Journal  of Zoology
  (London) 229:105-119.
Dingle, H., 1996. Migration: The  Biology of Life
  On the Move. Oxford University Press, Oxford,

Dodd, C. K. Jr. and B. G. Charest. 1988. The herpe-
  tofaunal community of temporary ponds in north
  Florida sandhills: Species composition, temporal
  use, and management implications,  pp. 11-22 In
  R. C.  Szaro, K. E. Severson, and  D.  R. Patton
  (tech. coords.) Management of Amphibians, Rep-
  tiles, and Small Mammals  in North America. Pro-
  ceedings of a  Symposium Flagstaff,  AZ. U. S.
  Forest Service General Technical  Report RM-

Dole, J. W. 1965. Spatial relations in natural popu-
  lations of the Leopard Frog, Rana pipiens Schre-
  ber, in northern Michigan. Am. Mid Nat.  74:464-
Douglas, M. E. and B. L. Monroe Jr. 1981.  A com-
  parative study of topographical orientation in Am-
  bystoma (Amphibia: Caudata). Copeia 1981:460-

Duellman, W.E. and L.  Trueb. 1986. Biology  of
  Amphibians. McGraw-Hill, Inc., New York, NY.

Dumas, P. C.  1966. Studies of the Rana  species
  complex  in   the  Pacific  Northwest.  Copeia

Dupuis, L.  A.,  J. N. M. Smith, and F. Bunnell.
  1995. Relation of terrestrial-breeding amphibian
  abundance to tree-stand age. Cons. Biol. 9:645-
Ewert,  M. A.  1969. Seasonal movements of  the
  toads Bufo americanus  and Bufo cognatus  in
  Northwestern  Minnesota.  Ph.D.  Dissertation.
  University of Minnesota, Minneapolis, MN.
Fahrig, L., J. H.  Pedlar,  S.  E. Pope, P. D. Taylor,
  and S. F. Wegner.  1995. Effect of road traffic on
  amphibian density. Biol. Cons. 73:177-182.

Fahrig, L. and G. Merriam. 1985.  Habitat  patch
  connectivity  and  population survival. Ecology:
Forman, R.  T. T.  1983 Corridors in landscapes:
  Their ecological structure and function.  Ecology
Forman, R. T. T. 1995. Land Mosaics: The  Ecology
  of Landscapes  and Regions. Cambridge  Univer-
  sity Press, New York, NY.
Fritschen, L. J. and  R. Edmonds. 1976. Dispersion
  of florescent particles into and within a Douglas-
  fir forest,  pp 280-301 In  R. J. Engelmann  and
  G. A.  Sehmel  (eds.)  Atmosphere-Surface  ex-
  change  of  particulate  and  gaseous pollutants.
  ERDA  Symposium Series-38.  NTIS CONF-74-

Galatowitsch, S. M. and A. van der Valk. 1994. Re-
  storing  Prairie  Wetlands:  An  Ecological Ap-
  proach.  Iowa State University Press, Ames, IA.

Gibbs, J. P. 1993. Importance of small wetlands  for
  the persistence of local populations of wetland-
  associated animals. Wetlands 13:25-31.

Gill, D. E. 1978a. Effective population size and in-
  terdemic migration rates in a metapopulation of
                      J~eniic-Breeding Amphibians
  the  red-spotted  newt, Notopthalmus viridescens
  (Rafinesque). Evolution 32:839-849.

Gill, D. E. 1978b. The metapopulation ecology of
  the  red-spotted  newt, Notopthalmus viridescens
  (Rafinesque). Ecological Monographs  48:145-

Gittins, S. P., A. G. Parker, and F. M. Slater 1980.
  Population  characteristics of the common toad
  (Bufo bufo) visiting a breeding site in Mid-Wales.
  Journal of Animal Ecology 49:161-173.

Hammer,  D.  A. 1992. Creating  Freshwater Wet-
  lands. Lewis Publishers, Chelsea. MN.
Hairston,  N.  G.  1987. Community Ecology  and
  Salamander Guilds. Cambridge University Press,
  Cambridge, MA.

Harris, L. D. and J. Scheck.  1991. From implica-
  tions to applications: The dispersal corridor prin-
  ciple applied to the conservation of biological
  diversity, pp. 189-220 In  D. A. Saunders and R. J.
  Hobbs  (eds.) Nature Conservation 2: The Role of
  Corridors. Surrey Beatty & Sons  Pty  Limited,
Hayes, M. P. and  M. R. Jennings.  1986. Decline of
  ranid frog species in western north America: Are
  bullfrogs (Rana catesbeiana) responsible? Journal
  of Herpetology 20:490-509.
Herreid,  C. F. and S. Kinney. 1967. Temperature
  and devilment of the wood  frog, Rana  sylvatica,
  in Alaska. Ecology 48:579-590.
Heusser,  H.   1960.    Uber die  beziehungen der
  Erdkrote (Bufo  bufo L.) zu ihrem Laichplatz II.
  Behavior 16:93-109.
Heusser, H. 1968. Die Lebensweise der Erdkrote,
  Bufo  bufo  L.:  Wanderungen  und  Sommer-
  quartiere. Rev. Suis. Zool. 76:444-517.
Hicks, A. 1995. Impervious surface area and  ben-
  thic macroinvertebrate response as an index of
  impact from urbanization on freshwater wetlands.
  M.S. Thesis, University of Massachusetts,  Am-
  herst, MA.
Holland,  C. C., J. Honea, S. E. Gwin, and M. E.
  Kentula. 1995. Wetland degradation and  loss in
  the rapidly urbanizing  area of  Portland,  OR.
  Wetlands 15:336-345.
Horner, R. R. 1988. Biofiltration Systems for Storm
  Runoff  Water Quality Control.  Prepared for
  Washington State Department  of Ecology, Mu-
  nicipality of Metropolitan  Seattle, King County,

 Lentic-Breeding Amphibians
  and Cities of Bellevue, Mountlake Terrace  and
  Redmond, WA.
 Hurlbert, S. H. 1969. The breeding migrations and
  interhabitat wandering of the vermilion-spotted
  newt  Notopthalmus  viridescens  (Rafinesque).
  Ecological Monographs 39:465-488.
 Hutchison,  V. H. and R. K. Dupre'. 1992. Ther-
  moregulation. pp. 206-249 In M.  E.  Feder  and
  W. W. Burggren (eds.) Environmental Physiology
  of Amphibians. The University of Chicago Press,
  Chicago, IL.
 Ildos, A. S. and N. Ancona. 1994. Analysis of am-
  phibian habitat preferences in a farmland area (Po
  Plain, northern Italy. Amphibia-Reptilia 15:307-
 Jameson, D. L.  1957.  Population  structure  and
  homing responses in the Pacific tree frog. Copeia
 Kent, D. M. 1994a. Designing wetlands for wildlife.
  pp. 283-306  In  Donald M  Kent (eds.) Applied
  Wetlands Science and Technology. CRC Press,
  Inc., Boca Raton, FL.
 Kent, D. M. 1994b. Managing wetlands for wildlife.
  pp. 307-329 In Donald M  Kent (eds.) Applied
  Wetlands Science and Technology. CRC Press,
  Inc., Boca Raton, FL.
 Kleeberger,  S. R.  and J. K. Werner.  1983. Post-
  breeding  migration  and summer  movement of
  Ambystoma  opacum.  Journal   of Herpetology
 Kusler, J. A. and G. Brooks (eds.). 1988. Proceed-
  ings of the National Wetland Symposium: Wet-
  land  Hydrology. Association of State  Wetland
  Managers, Inc., Beme, NY.

 Kusler, J. A. and M. E. Kentula (eds.)  1990. Wet-
  land Creation and Restoration: The Status of the
  Science. Island Press, Washington, D.C.

 Langton, T. S. E., ed. 1989. Amphibians and Roads.
  AGO Polymer Products Ltd. Shefford, Bedford-
  shire, England.

 Lehmkuhl, J. F., L. F. Ruggiero  and P.  A. Hall.
  1991.  Landscape-scale  patterns of forest frag-
  mentation and wildlife richness and abundance in
  the southern Washington Cascade range, pp. 425-
  442  In  K.  B.  Aubry  and M.   H.  Brookes
  (contributing editors) Wildlife and  Vegetation of
  Unmanaged  Douglas-fir  forests.  U. S.  Forest
  Service,  Pacific  Northwest  Research  Station,
  Portland, OR. General Technical Report  PNW-
Licht, L. E. 1969. Comparative breeding behavior
  of the red-legged frog (Rana aurora aurora) and
  the western spotted frog (Rana pretiosa pretiosa)
  in southwestern  British  Columbia.   Canadian
  Journal of Zoology 47:1287-1299.

Licht, L. E. 1971. Breeding habits and embryonic
  thermal  requirements of the frog, Rana pretiosa
  pretiosa,  in  the  pacific  northwest.  Ecology

Lind,  A. J., H. H.  Welsh Jr., and R. A. Wilson.
  1996. The effect of a dam on breeding habitat and
  egg survival of the foothill yellow-legged frog
  (Rana boylii) in northwestern California. Herpe-
  tological Review 27:62-67.

Loredo, L, D.  VanVuren, and  M.  L.  Morrison.
  1996. Habitat use and migration behavior of the
  California tiger salamander. Jour. Herp. 30:282-
Ludwa, K. A. 1994. Urbanization effects  on palus-
  trine wetlands: empirical  water quality models
  and development of a macroinvertebrate commu-
  nity-based biological index. M. S. Thesis, Univer-
  sity of Washington, Seattle, WA.
Mann,  C. C. and M. L. Plummer. 1995. Are wildlife
  corridors the right path? Science 270:1428-1430.
McAllister, K.  R. and B. P. Leonard. 1990. Past
  Distribution and  Current  Status of the Spotted
  Frog  (Rana pretiosa)  in  Western  Washington.
  1989 Progress Report. Washington State Depart-
  ment of Wildlife, Olympia, WA.

McAllister, K. R. and B. P. Leonard. 1997. Wash-
  ington State Status Report for the Oregon Spotted
  Frog. Washington Department of Fish and Wild-
  life, Olympia, WA.

Miaud, C. 1995. Oviposition site selection in three
  species  of European newts (Salamandridae) genus
  Triturus. Amphibia-Reptilia 16:265-272.
Minton, S. A., Jr. 1968. The fate of amphibians and
  reptiles  in a suburban area. Journal of Herpetol-
  ogy 2:113-116.

Mitsch, W. J. and J. G. Gosselink. 1993.  Wetlands.
  2nd  Edition.  Van Nostrand Reinhold  Company,
  Inc., New York, NY.

Moler, P.  E. and R. Franz. 1987. Wildlife values of
  small,  isolated wetlands  in  the southeastern
  coastal  plain, pp. 234-241 In R. R. Odom, K. A.
  Riddleberger, and J. C. Ozier (eds.) Georgia De-
  partment of Natural Resources, Atlanta, GA.

                                                                             Lentic-Breeding Amphibiai
M011er, T. R. and C. P. R0rdam 1985. Species num-
  bers of vascular plants in relation to area, isola-
  tion and age of ponds in Denmark. Oikos 45:8-16.

Munn, M. D.  and M. A. Brusven. 1991. Benthic
  macroinvertebrate communities in nonregulated
  and regulated waters of the Clearwater River, ID.
  Regulated   Rivers:   Research   &   Manage-
  ment 6:1-11.

Nussbaum, R. A., D. B.  Edmond Jr., and  R. M.
  Storm. 1983. Amphibians and Reptiles of the Pa-
  cific Northwest. University of Idaho Press, Mos-
  cow, ID.

Oldham, R. S.  1966. Spring  movements  in  the
  American Toad, Bufo americanus. Can. J. Zool.

Olson, D. H. 1988. The  ecological and behavioral
  dynamics  of breeding in three sympatric anuran
  amphibians.  Ph.D.  Dissertation.  Oregon  State
Olson, D. H. and W. P. Leonard. 1997. Amphibian
  inventory  and  monitoring:  A standardized  ap-
  proach  for the Pacific Northwest, pp. 1-21  In
  D. H. Olson, W. P. Leonard and R. B. Bury (eds.).
  Sampling  Amphibians in Lentic Habitats.  North-
  west Fauna 4, Olympia, WA.
Orchard,  S. A.  1992. Amphibian  population de-
  clines in  British  Columbia, pp.  10-13  In C. A.
  Bishop and  K. E. Pettit (eds.) Declines in Cana-
  dian  Amphibian  Populations: Designing  a Na-
  tional Monitoring Strategy. Occasional Paper 76.
  Canadian  Fish and Wildlife  Service. Ottawa,
Payne, N. F. 1992. Techniques for Wildlife Habitat
  Management of Wetlands. Biological Resource
  Management Series, McGraw-Hill,  New York,

Pechmann, J. H. K. and  H. M. Wilbur. 1994. Put-
  ting declining amphibian populations in perspec-
  tive:  Natural  fluctuations  and human  impacts.
  Herpetologica 50:65-84.
Platin, T. J. 1994. Wetland amphibians Ambystoma
  gracile and Rana aurora as bioindicators of stress
  associated with  watershed urbanization.  M.S.
  Thesis, University of Washington, Seattle, WA.

Platin, T. J. and K. O. Richter. 1995. Amphibians as
  bioindicators of stress associated with watershed
  development, pp. 163-170 In E. Robichaud  (ed.)
  Puget Sound Research '95  Proceedings, Volume
  1, Puget Sound Water Quality Authority, Olym-
  pia, WA.
Podloucky, R. 1989. Protection of amphibians on
  roads-  examples and  experiences from  Lower
  Saxony, pp. 15-27 In T. S. E. Langton (ed.) Am-
  phibians and Roads. AGO Polymer Products Ltd.
  Shefford, Bedfordshire, England.

Power, T., K. L. Clark, A. Harfenist, and D. B.
  Peakall. 1989.  A Review and Evaluation of the
  Amphibian Toxicological Literature.  Technical
  Report No. 61, Canadian Wildlife Service, Head-
  quarters, Ottawa, Canada.

Pounds, J. A. and M. L. Crump.  1994. Amphibian
  declines and climate disturbance: the case of the
  golden toad and the harlequin frog. Conservation
  Biology 8:72-85.

Ranney, J.  W., M. C. Bruner, and J. B. Levenson.
  1981. The importance of edge in the structure and
  dynamics of forest islands, pp. 67-95 In R.L. Bur-
  gess and D. M. Sharpe (eds.) Forest Island Dy-
  namics in Man-Dominated Landscapes. Springer-
  Verlag, New York, NY.

Reh, W.  1989. Investigations into the influences of
  roads on the genetic structure of populations of
  the common frog  Rana temporaria. pp.  101-103
  In T.E.S.  Langton  (ed.) Amphibians and Roads.
  AGO Polymer Products Ltd., Shefford, Bedford-
  shire, England.
Reifsnyder, W. E. 1955. Wind profiles in a small
  isolated forest stand. Forest Science: 1:289-297.
Reinarts, J. A. and E. L. Warne. 1993. Development
  of vegetation in small created wetlands in south-
  eastern Wisconsin.  Wetlands 13:153-164.
Reinelt, L. E. and R. R. Homer. 1990. Characteris-
  tics of the hydrology and water quality of palus-
  trine wetlands affected by urban stormwater. King
  County Resource Planning, Seattle, WA.
Reinelt, L. E. and R. R. Horner. 1995. Pollutant re-
  moval from stormwater runoff by palustrine wet-
  lands based  on  comprehensive budgets.  Eco-
  logical Engineering 4:77-79.
Richter, K.  O. and A. L. Azous. 1995. Amphibian
  occurrence and wetland characteristics  in the
  Puget Sound Basin. Wetlands 15:305-312.

Richter, K. O. and A. L. Azous. 1996.  Amphibian
  distribution, abundance and habitat use. pp. 5.1-
  5.11 In Wetlands  and Urbanization, Implications
  for  the  Future.   Puget  Sound  Wetlands  and
  Stormwater Management Research Program, King
  County, Seattle, WA.
Rowe, C. L. and W. A. Dunson. 1993. Relationships
  among abiotic parameters and breeding effort by

 Lentic-Breeding Amphibians
   three amphibians in temporary wetlands of central
   Pennsylvania. Wetlands 13:237-246.
 Rowe,  C. L. and W.  A.  Dunson.  1995. Impacts of
   hydroperiod on growth and survival of larval am-
   phibians in temporary  ponds of Central Pennsyl-
   vania. Oecologia 102:397-403.
 Rowe,  C. L., W. J. Sadinski, and W. A. Dunson.
   1992. Effects of acute and chronic acidification
   on three larval amphibians that breed  in tempo-
   rary  ponds. Archives of Environmental Contami-
   nation and Toxicology  23: 339-350.
 Rudolph, D. C. and J. G.  Dickson. 1990. Streamside
   zone width and amphibian and reptile abundance.
   The Southwestern Naturalist 35:472-476.
 Sadinski, W. J and W. A.  Dunson.  1992. A multi-
   level study of the effects of low pH on amphibi-
   ans of temporary ponds. Journal  of Herpetology
 Saunders, D. A., G.  W. Arnold, A. A.  Burbridge
   and A. J. M.  Hopkins. 1987. Nature Conservation:
   The Role of Remnants of Native Vegetation. Sur-
   rey Beatty & Sons Pty Ltd. NSW, Australia.
 Saunders, D. A. and  R. J.  Hobbs (eds.).  1991. Na-
   ture  Conservation 2: The Role of Corridors. Sur-
   rey Beatty & Sons Pty Limited, NSW, Australia.
 Savage, R. M. 1961. The Ecology and Life History
   of the Common Frog. Pitman, London, England.
 Schueler, T. 1994. The importance of impervious-
   ness.  Watershed Protection Techniques  1:100-

 Scale,  D.  1982.  Physical  factors influencing ovi-
   position by  the  woodfrog, Rana sylvatica, in
   Pennsylvania. Copeia 1982:627-635.
 Semlitsch, R. D. 1981. Terrestrial activity and sum-
   mer  home  range  of  the  mole  salamander
   (Ambystoma  talpoideum).  Canadian Journal of
   Zoology 59:315-322.

 Shoemaker,  V, H., S. S. Hilman, S. D. Hillyard,
   D. C. Jackson, L. L. McClanahan, P. P. Withers,
   and M. L. Wygoda. 1992. Exchange  of  water,
   ions,  and  respiratory gases in terrestrial amphibi-
   ans,  pp.  125-150 In  M. E. Feder and   W.W.
   Burggren   (eds.)  Environmental  Physiology of
   Amphibians.  The  University of Chicago  Press,
   Chicago, IL.

Shoop,  C.  R.  1965.  Orientation of Ambystoma
   maculatum: Movements  to and  from breeding
  ponds. Science 149:558-559.
Shoop, C. R. 1968. Migratory orientation of Amby-
  stoma  maculatum:  Movements  near  breeding
  ponds and displacements of migrating individuals.
  Biological Bulletin 135:230-238.

Shoop, C. R. and T. L. Doty.  1972. Migratory orien-
  tation  by  marbled  salamanders   (Ambystoma
  opacum) near a breeding area. Behavioral Biology

Sinsch U. 1988a. Seasonal changes in the migratory
  behavior  of the toad Bufo  bufo: direction and
  magnitude of movements. Oecologia 76:390-398.
Sinsch U. 1988b. Temporal spacing of breeding ac-
  tivity in  the natterjack  toad Bufo  calamita.
  Oecologia 76:399-407.

Sinsch, U.  1989. Migratory behaviour of the com-
  mon toad Bufo bufo and the  natterjack toad Bufo
  calamita. pp. 113-124 In Amphibians and Roads
  T.E.S. Langton (ed.) AGO  Polymer Products Ltd.
  Shefford, Bedfordshire, England.

Sinsch, U. 1992. Structure and dynamic of a natter-
  jack  toad  metapopulation   (Bufo   calamita).
  Oecologia 90:489-499.
Sinsch U. and D. Seidle.  1995. Dynamics of local
  and temporal breeding assemblages in a Bufo ca-
  lamita metapopulation. Australian Journal of
  Ecology 20:351-361.
Sjogren, G. P. 1991. Extinction and isolation gradi-
  ents in metapopulations: The case of the pool frog
  (Rana lessonae).  Biological  Journal of the Lin-
  nean Society (London) 42:135-147.
Sjogren, G. P.  1994.  Distribution and  extinction
  patterns within a northern  metapopulation of the
  pool frog Rana lessonae. Ecology 75:1357-1367.

Sjogren, P. and K. Larsson. 1988. Thermal require-
  ments for breeding  in Rana lessonae, R.  arvalis
  and R. temporaria at 60° N. Ph.D. Thesis Uppsala
  University, Sweden.

Sjogren, G. P.,  J. Elmberg, and S. A.  Berglind.
  1988. Thermal preference  in the pool frog Rana
  lessonae: impact on the reproductive behavior of
  a northern fringe population. Holarctic Ecology.

Sjogren-Gulve, P. and C. Ray 1996.  Using logistic
  regression  to model  metapopulation  dynamics:
  Large-scale forestry extirpates the Pool Frog. pp.
  111-137  In Metapopulations and Wildlife Con-
  servation  D. R. McCullough  (ed.)  Island Press,
  Washington, D.C.

Slater,  J. R.  1936. Notes  on Ambystoma gracile
  Baird and  Ambystoma  macrodactylum  Baird.
  Copeia 1936:234-236.

Soule, M. E. and M. E. Gilpin. 1991. The theory of
  wildlife corridor capability, pp. 3-8 In D. A. Saun-
  ders and R. J. Hobbs (eds.) Nature Conservation
  2: The Role of Corridors. Surrey Beatty & Sons
  Pty Limited, Australia.

Spies, T. A. and  S. P. Cline. 1988. Coarse  woody
  debris in forests and plantations of coastal Oregon
  pp. 5-24 In C. Maser, R. F. Tarrant, J. M. Trappe,
  and J. F. Franklin (tech. eds.) From The Forest to
  the Sea:  A Story of Fallen Trees.  U.S.  Forest
  Service, Portland, OR.  Gen. Tech.  Rep.  PNW-

Stebbins, R. C. and N. W. Cohen.  1995. A Natural
  History of  Amphibians. Princeton  University
  Press, Princeton, NJ.

Stenhouse, S. J.  1985. Migratory orientation and
  homing in Ambystoma  maculatum  and  Amby-
  stoma opacum.  Copeia 1985:631-637.

Storm, R. M. 1960. Notes on the breeding biology
  of the Red-Legged Frog (Rana  aurora aurora).
  Herpetologica 16:251-259.

Strijbosch, H. 1979. Habitat selection of amphibians
  during their aquatic phase. Oikos 33:363-372.

Sype,  W. E. 1974. Breeding habitats, embryonic
  thermal requirements and embryonic and larval
  development  of the Cascade frog, Rana  cascadae
  Slater. Ph.D.  Thesis.  Oregon State University,
  Corvallis, OR.

Taylor, B.  L. 1993. The influence of wetland and
  watershed morphological  characteristics on wet-
  land  hydrology  and  relationships  to wetland
  vegetation communities. M.S. Thesis University
  of Washington, Seattle, WA.

Travis, J.  1994. Calibrating  our  expectations  in
  studying amphibian populations.  Herpetologica

Tunner, H. G.  1991. Migratory behaviour of Rana
  lessonae and its hybridogenetic  associate Rana
  esculenta of  Lake Neusiedl (Austria, Hungary).
  Proc. 6th Ord. Gen. Meeting Eur. Soc. Herp. Bu-
  dapest, p. 91.

Twitty, V. C. 1958. Experiments on homing behav-
  ior and speciation  in  Taricha.  pp.  415-458  In
  Vertebrate Speciation. ed. W.F. Blair. University
  of Texas Press,  Austin, TX.
                       Lentic-Breeding Amphibians
Twitty,  V. C. 1959. Migration and speciation in
  newts. Science 130:1735-1743.

Twitty, V. C. 1966. Of Scientists and Salamanders.
  W.H. Freeman and Co., San Francisco, CA.

Wacasey,  J. W.  1961. An ecological study of two
  sympatric species  of  salamanders,  Ambystoma
  maculatum and  Ambystoma jeffersonianum  in
  southern Michigan. Ph.D.  Dissertation, Michigan
  State University, East Lansing, MI.

Waldman, B. 1982. Adaptive  significance of com-
  munal oviposition in wood frogs (Rana sylvaticd).
  Behavioral Ecology and  Sociobiology  10:169-

Wales, B. A. 1972. Vegetation analysis of northern
  and southern edges in a mature oak-hickory for-
  est. Ecological Monographs 42:451-471.
Walls, S. C., A. R. Blaustein, and J. J. Beatty. 1992.
  Amphibian biodiversity of the Pacific Northwest
  with  special   reference  to  old-growth  stands.
  Northwest Environmental Journal 8:53-69.
Ward, D. E., G. J.  Hirons, and M. J. Self. 1995.
  Planning for the  restoration of peat wetlands for
  birds, pp. 207-222 In B. D Wheeler,  S.  C. Shaw,
  W. J. Fojt, and R. Al. Robertson (eds.)  Restora-
  tion of Temperate Wetlands. John Wiley & Sons,
  New York, NY.
Washington State  Department of Wildlife. 1992.
  Buffer Needs  of Wetland Wildlife:  Final Draft.
  State  of  Washington  Department  of  Wildlife,
  Habitat Management Division, Olympia, WA.
Weller, M. W.  1990. Waterfowl management tech-
  niques for wetland enhancement,  restoration  and
  creation useful in mitigation procedures, pp. 517-
  528 In J.A. Kusler and M. E. Kentula (eds.) Wet-
  land Creation and Restoration: The Status of the
  Science. Island Press, Washington, DC.
Welsh, H. H. Jr. and A. J. Lind. 1988. Old growth
  forests and the distribution  of the terrestrial  her-
  petofauna. pp. 439-458  In R. C.  Szaro R. C.
  Szaro, K. E.  Severson, and D. R.  Patton (tech.
  coords.) Management of Amphibians,  Reptiles,
  and Small Mammals in North America. Proceed-
  ings of a Symposium Flagstaff, AZ. U.S. Forest
  Service Gen Tech Rep RM-166.
Wheeler,  B. D., W. J. Fojt, and R.  R.  Allan, (eds.)
  1995. Restoration  of Temperate Wetlands. John
  Wiley & Sons, New York, NY.
Wilard, D. E. and  A. K. Killer. 1990.  Wetland dy-
  namics:  Considerations for restored and created
  wetlands, pp.  459-466 In J.A. Kusler and M. E.

  Lentic-Breeding Amphibians
    Kentula (eds.) Wetland Creation and Restoration:       Williams, P.  K. 1970. Seasonal  movements and
    The Status of the Science. Island  Press, Wash-         population  dynamics of four  sympatric  mole
    ington, DC.                                            salamanders, Genus Ambystoma. Ph.D.  Disserta-
  11/-IL    IT  »*   j i  n f~> 11-     ir>-i-> T-   i   -  i         tion, Indiana University, Bloomington, IN.
  Wilbur, H.  M. and J.  P. Collins.  1973. Ecological                             '          °
    aspects of amphibian  metamorphosis.  Science       Wolfenbarger, D. O. 1949. Dispersion of  small or-
    182:1305-1314.                                        ganisms. American Midland Naturalist 35:1-152.


                                 Mitigating Impacts from Ferry Terminals
                                                on Eelgrass (Zostera marina L.)
                                                                             Ronald M. Thorn1
                                                                            David K. Shreffler1
                                                                       Charles A. Simenstad2
                                                                            Annette M. Olson3
                                                                   Sandy Wyllie Echeverria3
                                                                          Jefferey R. Cordell2
                                                                               James Schafer4
   Abstract: The purposes of this study were to
  understand effects of ferry terminals  and ferry
  operations on eelgrass (Zostera marina L.) habitat
  in Puget Sound and to design  appropriate meas-
  ures to avoid, minimize and compensate for asso-
  ciated impacts. Dramatic increases in population
  and ferry traffic in  western  Washington have
  resulted in the need to expand existing terminals.
  Our studies showed that eelgrass habitats near
  ferry terminals were affected by light reduction,
  and other initial and long-term disturbances asso-
  ciated  with  terminal construction  and  mainte-
  nance,  propeller wash, and  bioturbation   by
  macroinvertebrates  (i.e., seastars and Dungeness
  crab). Measures to  avoid and minimize impacts
  are being evaluated, including concrete piles, in-
  corporating transparent plastic materials into the
  dock structure, reflective material under the dock,
  grating over  offloading ramps,  and  reorienting
  slips to minimize propeller wash. Restoration is
  also proposed for portions  of habitats  previously
  destroyed by ferry operations.


  Background and  Purpose
   Eelgrass (Zostera marina L.) habitats occur along
  an  estimated 50 percent  of  the shoreline of Puget
  Sound (Thom and Hallum 1991). This ecosystem is
  rich in fish and invertebrate species, is important to
  migratory  waterfowl  (Baldwin and Lovvorn  1994)
  and functions  as  habitats for reproduction, rearing
  and feeding  of  commercially  and  recreationally
important fisheries species (Phillips 1984). Although
difficult to quantify, losses of eelgrass have occurred
in Puget Sound because of physical alterations and
development of shorelines (Thom and Hallum 1991).
Docks, which  necessarily cross  over  the eelgrass
habitats, pose  a threat to eelgrass through reduction
of light and a  variety of other factors. Dock shading
has been  shown to affect seagrasses in Washington
State (Pentilla and Doty, 1990) and New England
(Burdick  and  Short 1995).  Because commerce and
public  transportation  requires  docks, and because
trade  and use of waterways for transportation  is
growing, the demand to expand existing docks as well
as construct new  docks will continue to impact eel-
grass as well as other coastal habitats. For sustainable
development  to  occur  sound,  scientifically-based
methods to mitigate impacts  of docks on shallow
coastal habitats need to be developed.
  The purpose of our study was to understand the im-
pacts of ferry  terminals on eelgrass habitats in Puget
Sound, and develop measures to mitigate the impacts.
Ferries provide a major mode of transportation for the
public and commerce. In this paper we present our
approach  to addressing impacts of ferry terminals on
eelgrass, and a conceptual model that identifies key
impacts of ferry terminals on eelgrass in Puget Sound
and describes  how these impacts can be avoided,
minimized and compensated. This model is based on
impacts to eelgrass and fisheries resources observed
at one ferry  terminal (Clinton)  that is  slated  for
expansion, and on directed studies of light, currents
and propeller wash.  Studies described herein have
also been  completed at the  Port Townsend and
Edmonds terminals in Puget Sound.
1 Battelle Marine Sciences Laboratory, 1529 W. Sequim Bay Rd., Sequim, Washington 98382.
2 Wetland Ecosystem Team, School of Fisheries, WH-10, University of Washington, Seattle, Washington 98195.
3 School of Marine Affairs, HF-05, University of Washington, Seattle, Washington 98195.
4 Environmental Affairs Office, Washington State Department of Transportation, P.O. Box 47331, Olympia, Washington 98504.

Eelgrass Impacts
  The  investigations  consisted of  eelgrass habitat
 mapping, sampling of standing stock, in situ irradi-
 ance, fish and invertebrate observations, quantifica-
 tion of propeller wash impacts, and testing of light
 enhancement technologies.

 The Clinton Ferry Terminal
  The Clinton Ferry Terminal is located on the south-
 east end of Whidbey Island, Washington (Figure 1).
 The present creosote wood terminal is approximately
 30 m wide and extends over the depth distribution of
 eelgrass. Vessels in excess of 70 m in length serve the
 terminal. During the day, dockings and sailings occur
 on the order of every 15 to 30 minutes.

 Video Surveys
  Eelgrass within approximately 500 m of either side
 of the terminal was mapped in July  1994 using an
 video camera towed from a boat. Video recordings of
 the bottom were made while the boat traveled along
 26 transects that crisscrossed the study area over the
 maximum depth range eelgrass would be expected to
 occur. A Global Positioning System (GPS) mounted
 on board the boat automatically transferred exact sta-
 tion  position to the videotape.  These tapes were
 viewed to score eelgrass cover at  approximately 2-m
 intervals along these transects according to the  fol-
 lowing cover classes: absent (0 percent), low-moder-
 ate (1-50 percent),  and moderate  to  high (50-100
 percent).  These  data  were input to a Geographic
 Information System (GIS)  database using GPS  data
 and mapped. In addition, a vertical  color aerial pho-
 tograph, taken showing  the  general outline of eel-
 grass, was  used  to  delineate the  broad  eelgrass

 Eelgrass Standing Stock
  Based on the video-generated maps, we established
 three transects through the eelgrass habitat.  The tran-
 sects ran parallel to the shoreline and were positioned
 at the inner, mid and outer portions  of the habitat.
 The continuous, 180-m long transects spanned from
 50 m north of the terminal to 50 m south of the termi-
 nal, and included the 30 m section  under the terminal.
 On 29-30 August 1994, using SCUBA,  we surveyed
 points  at 5-m intervals along the each  transect; re-
 cording  depth,  time, eelgrass  shoot  density  (in
0.25 m2 quadrants) and substrata type.
 To document the effect of the dock on light during
summer, mid-day conditions, we recorded photosyn-
thetically active radiation (PAR) in  air using  a 4n
quantum sensor (LICOR) at  1-m intervals along  a
transect which ran from 30 m north to 30 m south of
the terminal.  We also recorded PAR at each of the
points  where eelgrass was sampled  along the tran-
sects during midday (1200-1400 hrs). In addition, we
recorded PAR in the middle of the habitat and in air
using sensors mounted in the habitat and on the dock,
respectively. These measurements were taken at the
same time as the eelgrass sampling.

Fish and Invertebrate Observations
 Fish and macroinvertebrates present within each of
the eelgrass quadrats were noted along with those in
the general vicinity of the transect  lines  and the
Propeller Wash
  Our observations indicated that propeller wash af-
fected at least the outer portion of the eelgrass habi-
tat. To  examine  this further, we recorded  current
speeds within 30cm of the  bottom during several
ferry  dockings and  departures  20 September and
3 October  1994, and on  8 June  1995.  We  used  a
pressure-driven meter during the 1994  visits and a
propeller-driven  meter in 1995. The current  meters
were positioned immediately next to the terminal, on
the south  side,  in a location  where eelgrass was
apparently absent due to sediment disturbances from
propeller wash. We  also made observations of cur-
rents  at other positions further from the area obvi-
ously impacted  by  propeller wash.  During the  3
October  trip, we  recorded irradiance  concurrently
with current speeds  using the  LICOR  sensor posi-
tioned next to the current meter. Finally,  we made ob-
servations and marked on maps of the study area the
maximum extent of the visible surface plumes during
several ferry departures and dockings.

Light Enhancement Technologies
  Because  light would be reduced by the expanded
terminal,  we investigated simple  light-enhancement
technologies  including  artificial  (halogen  lamp)
lighting, concrete blocks with transparent glass cen-
ters (approximately  200 cm2), and placement of  re-
flective material. To evaluate whether quartz halogen

122.3490° W

 Eelgrass Impacts
 lamps could  support photosynthesis, using oxygen
 flux  methods we measured  photosynthetic  rates of
 three  10-cm long leaf sections  placed in glass  jars
 (1L) containing ambient seawater. Ten replicate jars
 were placed in tanks with flowing seawater  at ambi-
 ent sea temperature, and the tank was covered to
 block out ambient light. Two quartz  halogen  lamps
 (500 watts each) were positioned over the jars in the
 tank about 50 cm above the jars. The jars were  sub-
 merged in approximately 40 cm of water. A separate
 set of 10 jars were incubated in ambient sunlight in a
 flowing seawater tank. Two blank jars (i.e., contain-
 ing water but no seagrass) were incubated along with
 each treatment. Incubations were 3 hrs long. The ex-
 periment was run on 29 March 1995 and repeated on
 30 March 1995.
   To evaluate the amount of PAR transmitted through
 the concrete blocks containing glass centers,  the PAR
 sensor was placed immediately below the lower edge
 of the block. Black plastic sheeting was used  to ex-
 clude all light except that passing through the glass
 insert. Ambient PAR  was  concurrently  recorded
 within 1 m of the  block frequently  throughout the
   We evaluated the effect of reflective material on in-
 creasing  light  under   docks   by   attaching  a
 122 x 244 cm sheet of plywood  covered with alumi-
 num foil to the underside of the wooden boat dock at
 Battelle's Marine Science Laboratory. The sheet was
 approximately in the center of the  12-m wide dock.
 Ambient PAR was measured  on  the south side of the
 dock, and under the dock at a point approximately 1
 m below the foil and below  an  adjacent area  of the
 dock. Light  reaching  each  of these points  was
 primarily light reflected off of the water surface.


 Habitat Characteristics
  The ferry terminal will be expanded to the south of
 the existing terminal (Figure 2). The habitats that oc-
 cur in  this area include moderately  sloping coarse
 sand  from above extreme higher high water (EHHW)
 down to about  mean lower low  water  (MLLW).
 Thereafter, the slope decreases and the substrata con-
 sists  primarily of medium to fine sands (Figure 1).
 The terminal is situated along a portion of the beach
 that is somewhat steeper than areas within at least
 1 km north and south of the terminal.
Vegetation (eelgrass and seaweeds)
 Aquatic  vegetation in  the terminal expansion area
consisted  primarily of eelgrass, benthic macroalgae
(Ulva  spp., Laminaria  sp.)  and benthic  diatoms
(Table 1). Ulva was intermingled with the eelgrass,
and Laminaria occurred primarily at depths greater
than -3 m. Both  seaweeds were very patchy in their
 The  distribution  of  eelgrass  is  illustrated in
Figures 1  and  2. Eelgrass was primarily distributed
from MLLW down to -3.6 m MLLW. Within this
depth range, areas with little or no eelgrass occurred
in a band approximately 5m wide on either side of the
terminal.  Shoot  density ranged  from  20 to over
800 m"2 within the main distribution of the habitat.
The apparent zone affected by propeller wash at the
outer edge of the habitat  was also devoid of eelgrass.

 Although light decreased dramatically under the
terminal, it remained relatively high for a distance of
5 m under the south side of the terminal, and was also
relatively high at 2 m under the north side of the ter-
minal (Figure 3). Bottom light measurements made at
midday during a sunny period in August show the
dramatic decrease in PAR with depth, and that eel-
grass occurred in greatest densities where PAR was
greatest (Figure 4). Areas under the dock are distin-
guishable  in Figure 4 as points with very low light
relative to depth.

Fish and Invertebrate Observations
 Twelve fish and 17 macroinvertebrate taxa were ob-
served during the diving studies (Table 1). Although
the eelgrass  habitat held  more species from  each
group  (14 species  of macroin vertebrates,  12 species
of fish), the under-dock habitat did contain  substan-
tial numbers of  species. Eelgrass-associated species
such as tubesnout and piling-associated species such
as bay mussel  were  restricted  to  their  preferred
 We  observed  high  seastar (Pycnopodia  heli-
anthodes) densities (i.e., up to  15 m"2) in some areas
under  or immediately adjacent to the dock. Seastars
commonly feed  on barnacles and mussels  growing
attached to  pilings. Extensive  piles  of mussel shell
and barnacle tests on the bottom under the dock were
evidence that this was occurring at Clinton. Although

                                                r>    /<:
                                                VeHcle capacity

                                              Relocated Put!l\FlsHng Pier
      Passenger Loading
     Future Overhead
                      PREFERRED ALTERNATIVE
                   CLINTON  FERRY TERMINAL
                     ENVIRONMENTAL  STUDY
Figure 2. Preferred Construction Plan Showing Existing Eelgrass and Transplanting Areas (A-H).

                -40 -35 -30 -25 -20 -15 -10 -5  0  5  10  15  20  25 30  35  40
                   NORTH                                     SOUTH

                                    DISTANCE (m)
        Figure 3.  PAR in Air Under the Dock and Next to the Dock.
                   50   100  150  200  250
                IRRADIANCE («M m-2 s-1)
                                                50    100   150   200   250
                                        ZOSTERA DENSITY (No. 0.25m-2)
       Figure 4.  Data from Clinton Terminal showing variation with Depth of (A) Light and
       (B) Eelgrass Density.

Table 1.  Taxa of fish, macroinvertebrates and macrophytes observed at the Clinton Terminal in  late
summer 1994.
                                IN OR NEAR
  bay mussel
  brooding anemone
  coon-striped shrimp
  Dungeness crab
  heart cockle
  helmet crab
  horse clam
  kelp crab
  leather star
  plumose anemone
  purple star
  red rock crab
   snail (chink shell)
   sunflower star

  Benthic Plants
   sea lettuce
   sugar wrack

  copper rockfish
  crescent gunnel
  kelp greenling

  penpoint gunnel
  pile perch
   saddleback gunnel
  shiner perch
  striped perch
Balanus spp.
Mytilus spp.
Epiactus prolifera
Pandalus danae
Cancer magister
Clinocardium nuttallii
Telmessus cheiragonus
Tresus capax
Pugettia producta
Dermasterias imbricata
Melibe leonina
Dirona aurantia
Metridium senile
Pisaster ochraceus
Cancer productus
Lacuna spp.
Pycnopodia helianthodes
unidentified species
Zostera marina
Gracilaria pacifica
Ulva spp.
Laminaria saccarina
Sebastes caurinus
Pholis laeta
various unidentified species
Apodichthys flavidus
Rhacochilus vacca
Pholis ornata
Citharichthys spp.
various unidentified species
Cymatogaster aggregata
Embiotocca lateralis
Aulorhynchus flavidus

Eel grass Impacts
 unquantified, Dungeness crab were at least as dense
 as seastars in many areas near the terminal.

 Propeller Wash
  Measurements at Clinton showed that  the plume
 created during ferry arrival and departure extended up
 to  80m from the dockside end of the vessel. There
 was  considerable variability  in  plume extent  and
 angle depending on  currents, wave action, wind and
 other factors. However, the angle of the main axis of
 the plume typically extended slightly shoreward of
 the main axis  of  the  docking vessel.  Plume events
 occurred  on average every 15-20 minutes during the
 day, and  lasted for 5-10 minutes.  Hence, there are
 only relatively short periods of time when the plume
 is not present.
  Materials  in the  observable plume consisted of fine
 sediments, organic debris and air bubbles. Sediment
 disturbance appeared to be greatest during periods of
 low tides, especially during spring tide series in the
 lunar cycle. At this time the propeller of the vessel is
 closest to the  channel bottom of the slip.  Bubbles
 were always present during plume events. Propeller
 wash accelerated  current speeds from near  zero to
 over 1 m  s"1  (Figure 5).  Maximum  rates recorded
 during a  docking were 3.5 m  s"1. Accelerations were
 accompanied by reductions in light (Figure 6).

 Light Enhancement Experiments
  The quartz halogen lamps supported photosynthesis
 of the eelgrass  leaf sections. Mean productivity under
 the  lights   and  ambient  light  were  2.08  and
 0.74 mgO2/g dry wt/hr, respectively. Calculations of
 the light  attenuation caused by  water  and  distance
 between lights attached to the underside of docks (at
 about 6 m  MLLW) and the depth where eelgrass
 normally  grows (0- to  -3 m MLLW) at this site indi-
 cated that a prohibitively large number of lights (and
 expenditure of energy) would be needed to support
 eelgrass photosynthesis under the docks.
  Approximately 60 percent of the ambient PAR was
 transmitted  through  the concrete block containing a
 glass center. The  foil attached under the dock re-
 flected substantially  more PAR than did the wooden
 dock (Figure 7).

Interpretation of Field Studies
 Shading by the dock is undoubtedly a major factor
causing the loss of eelgrass near the Clinton terminal,
but we believe that propeller wash, bioturbation and
other physical disturbances may be contributing to
the loss. The irradiance measurements clearly showed
that light reaches very low levels under the terminal.
However, lack of eelgrass in a 5-m wide band around
the terminal suggests that other factors are active.
Terminals of this  vintage  (mid-1950's) were con-
structed by hydraulically inserting wood piles into the
sediment. This process,  which involves water jet liq-
uefaction  the  sediment, completely eliminates  eel-
grass  and  likely   drastically  modifies  sediment
conditions such as organic content and redox profile.
Eelgrass, which primarily spreads by rhizome growth
in the region, may take  decades to recover from this
types of disturbance. Annual maintenance of wood
terminals  is  required and  these activities (e.g., barge
grounding and anchoring, propeller scars from tugs
and work boats) may also disturb eelgrass.
 During eelgrass surveys, the  divers noted seastars
foraging extensively for bivalves at the edge  of the
eelgrass  habitat  adjacent to  the  dock. Enhanced
populations  of both seastars and crab may have rami-
fications on  eelgrass in the vicinity of the dock. This
behavior has been documented by us in other regions
of Puget Sound. We have  observed that their foraging
activity disrupts eelgrass, and could retard recruit-
ment of eelgrass. In addition, Dungeness crab bury in
sediments as  a  predator defense mechanism. This
burrowing activity may  also disrupt newly recruiting
eelgrass seedlings. Where population density is great
as at Clinton, burrowing may be a significant factor
inhibiting recruitment of eelgrass.  Dungeness  crab
larvae are known to settle  in  shell piles. The shell
offers  shelter from predation  as well as  enhanced
food resources for  the young crab (Dumbauld et al.
1993). Enhanced crab abundances may be due to the
availability of prime habitat for crab larvae.

 The net effect of the propeller wash is to scour and
redistribute  sediments and  associated  biota  and  to
lower irradiance. Redistribution of sediments is evi-
dent as a characteristic disturbed "ring" adjacent to
the slip channel. This ring is barren of eelgrass. The
reduced  irradiance reaching the  bottom may  have
effects on the growth rate and survival of eelgrass. In
general, eelgrass exists  as deep as light requirements







O _*r^

\^ Tidal Current with
luxuriant growth

^ departing departing
0 °« \ $ approaching / . .
o \ ° / approaching
— \ o i y
O O May A
	 £.. Background 	 „-- °

«oj ^

	 ° ««\ 0°
•jCPoo CT^ Q. "*% o

0 °5
O ^Siy

      1020 1030 1040 1050
                                       TIME (HRS)

 Figure 5.  Current speeds in propeller wash plume. Current meter was positioned in

 Area F (see Figure 2).
                           1100      1150
                                   TIME (HRS)

 Figure 6.  Bottom irradiance at Area F during ferry docking and departures on

 October 3,1994 (see Figure 2).

                                             Reflective   Foil
                                                                       Irradiance  (PAR)
       Figure 7.  Average PAR Reflected Off Aluminum Foil and Wood Surface of Dock.

                                                                                          Eelgrass Impacts
and  suitable  substrata  will  allow.  At  the  outer
(deepest)  edge  of the habitat,  eelgrass  growth  is
probably at a threshold and is highly dependent on
some critical  level of light reaching the plants.  In
areas where light is reduced, such as in the propeller
wash plume, eelgrass photosynthesis may be inhib-
ited  enough to reduce  growth  below  this  critical
threshold for survival. At present, we believe that the
frequently-reduced irradiance associated  with the
plume has caused a reduction in the seaward extent of
the eelgrass habitat near the Clinton terminal. Obser-
vations made  by  us at  Clinton,   other  terminals
(Edmonds, Port Townsend), and examination of aer-
ial photographs of several other terminals, indicate
that  propeller wash is  likely  a significant  factor
affecting eelgrass distribution near the terminals.

Conceptual  Model  of the  Dock  System
Impacts and Mitigation Measures
  The  Clinton Ferry Terminal Construction  Project
will involve replacement of the  existing timber pile
structure with a concrete  pile structure, and expan-
sion of the holding area to accommodate increased
ferry traffic (Figure 2). The dock will be widened by
approximately 19 m and lengthened by approximately
44 m. A  new south slip,  steel wing walls, floating
dolphins,   towers  and  headframe   will  also   be
  The  present condition  of the eelgrass habitat  at
Clinton is due to  historical (not presently active)  as
well as current processes  (Figure 8). Disturbance to
the habitat is partitioned into two phases: construc-
tion, and maintenance and operation.
  Terminal  expansion  will have  short-term  direct
effects and longer term impacts on eelgrass. Initial
construction activities are predicted to have  some
limited effect  on  eelgrass, although these activities
will be largely conducted away from existing beds.
The new terminal deck will cover 320 m2 of eelgrass
presently  located  on the south side  of the terminal.
The proposed mitigation  measures  are directed  at
eliminating any longer-term effects.
  With  potential   impacts  to  eelgrass   identified
through a series of meetings with state and federal re-
source agencies, the Washington State Department of
Transportation undertook a program to identify nec-
essary actions to avoid, minimize and compensate for
these impacts. Impacts expected under the original
design plan for the terminal have been either avoided
or minimized. For example, ferry propeller wash im-
pacts have been avoided by moving the slips further
offshore. Light impacts have been minimized through
incorporation   of   light   transmitting   structures
(concrete blocks with glass centers) in the terminal
deck,  and lengthening  the  terminal.  In  addition,
highly reflective paint (i.e., the type used for painting
white lines on roads) will be used under the terminal
to enhance the albedo.  Lengthening the terminal
reduced the width of the terminal at the point where it
crosses the eelgrass habitat. Maintenance  activities
have been reduced dramatically through the use of
concrete piles and  decking as opposed  to  timber
(which was proposed in  the original plan).  Use of
concrete pilings will result in placement of 1/3 fewer
pilings than  presently  exist.  This  will  reduce the
amount of space for piling communities to develop
and may support fewer  seastars and dungeness crab.
This may result in less bioturbation effects  on eel-
grass.  Fewer  pilings will  also allow more light to
penetrate under the terminal.
  We believe that, with appropriate modifications in
the terminal  expansion, eelgrass  can  be restored in
many of the areas where eelgrass has been eliminated
by past  or on-going disturbances.  These areas are
identified by letters A-H  in Figure 2. Probability of
success varies among the areas. For example,  areas
A, B,  C, D, E, F,  and G are considered areas where
disturbances  can be essentially eliminated and eel-
grass transplants have a moderate to high probability
of being successfully established. Other areas  are less
likely  to succeed due to their experimental nature
(e.g., area H under glass blocks).
  The  overall mitigation goal for the  Clinton Ferry
terminal Expansion project is no net loss of eelgrass
on the project site. This goal will be achieved by:
1.  Reduction in impacts as much as  possible in the
    design  and  construction  of the project (as
    described above);
2.  Avoidance of future  impacts associated with
    operations at this terminal by
3.  Locating all anchors  or pilings for any  floating
    dolphins associated with the  terminal outside  of
    eelgrass areas, b. avoid moorage  of barges over
    eelgrass areas; and,
4.  Reduce local habitat fragmentation through resto-
    ration of previously  disturbed eelgrass  patches
    through  transplantation  of eelgrass  into  these
    areas. Approximately 320 m2 of eelgrass will  be
    directly impacted by the terminal expansion. To-
    tal  restoration  area  (A-H   in   Figure 2)  is

                                DOCK IN PLACE
Figure 8.  Model of Impacts and Mitigation Measures at Clinton Terminal.

                                                                                       Eelgrass Impacts
approximately 3,100 m2, which  would result in an
approximately 10:1  replacement  of the area directly
impacted by the dock expansion. The additional area
is  required  to offset poor transplant  survival and
growth in some areas (Fonseca 1990).
 The National Research Council (NRC 1992) has
recommended that restoration and mitigation projects
should incorporate experiments to help test new tech-
nologies and evaluate methods that may help in future
projects. Thus, an additional goal of the project is to
evaluate  some new concepts in mitigating shading
effects of overwater structures. The use  of glass
blocks in the passenger walkways and reflective paint
under the terminal deck represent ideas that, if proven
feasible  and  successful through  post-construction
monitoring,  could  then be incorporated  into future
terminal expansions elsewhere.

  This research  was supported  by  the Washington
State Department  of Transportation.  We gratefully
acknowledge the  excellent  logistical support and
cooperation  provided  by  WSDOT  staff,  and  the
Clinton terminal managers  and workers. J.  Norris
conducted the video surveys and B. Feist  produced
the GIS-based map of the eelgrass habitat. The map
in Figure  1 was  produced  by the  Transportation
Northwest  Regional  Center at the  University  of

Literature Cited
Baldwin, J.  R. and J. R. Lovvorn. 1994. Habitats and
  tidal accessibility of the marine foods of dabbling
  ducks and brant  in Boundary Bay, British Colum-
  bia. Marine Biology 120:627-638.
Burdick, D. M. and F. T. Short. 1995. The effects of
  boat docks on eelgrass  beds  in  Massachusetts
  coastal waters. Submitted to Waquoit Bay National
  Estuarine  Research Reserve  and  Massachusetts
  Coastal  Zone  Management,  Jackson  Estuarine
  Laboratory, University  of New Hampshire, Dur-
  ham, NH.

Dumbauld, B. R., D. A. Armstrong and T. L. McDon-
  ald. 1993.  Use of oyster shell to enhance intertidal
  habitat  and  mitigate  loss of Dungeness crab
  (Cancer magister)  caused by  dredging. Canadian
  Journal of Fisheries and Aquatic Sciences 50:381-

Fonseca, M. 1990. Regional analysis  of the creation
  and restoration of seagrass systems,  pp. 171-189 In
  Jon A. Kusler and  Mary E. Kentula, eds. Wetland
  creation and restoration the status of the science.
  Island Press.
National Research Council (NRC). 1992. Restoration
  of aquatic  ecosystems. National Academy Press,
  Washington, D.C.
Pentilla, D. and D. Doty. 1990. Results of 1989 eel-
  grass shading studies in Puget Sound. Unpublished
  progress report. Washington State  Department of
  Fisheries, Olympia, WA.
Phillips, R. C. 1984. The ecology of  eelgrass mead-
  ows in the Pacific  Northwest: a community profile.
  U.S. Fish and Wildlife Service FWS/OBS-84/24.
Thorn, R. M. and L. Hallum. 1991. Historical changes
  in the distribution of tidal marshes,  eelgrass mead-
  ows and kelp forests in Puget Sound, pp. 302-313
  In Puget Sound Research 91 Proceedings. Puget
  Sound Water Quality Authority, Olympia, WA.

                                      Creating Tidal Marshes on Dredged
                                             Materials: Design Features and
                                                          Biological Implications
                                                                           Ted P. Winfield1
                                                                           Joan Florsheim2
                                                                           Philip Williams2
  Abstract: The use of dredged materials to create
 tidal marshes is not new, but there is a lack of dis-
 cussion on design parameters necessary for con-
 structing  tidal   marshes   that   are  similar
 structurally  and  functionally  to  natural  tidal
 marshes. Our study was designed  to: (1) investi-
 gate the similarity  and differences  in selected
 physical  and biological components  of natural
 tidal marshes compared to tidal marshes created
 on dredged materials; and (2) identify important
 design  features  to consider  in designing  future
 tidal marsh restoration projects  using dredge
 materials to increase the likelihood of achieving
 functional equivalence. Four tidal  marshes  were
 selected  for study   in  San  Francisco   Bay,
 California, two  that had been constructed using
 dredged  materials  (Muzzi  Marsh   and  Faber
 Tract)  and  two  natural tidal  marshes  (Corte
 Madera  Ecological  Reserve  and   Laumeister
 Tract). There were significant differences in the
 percent cover by tidal marsh vegetation (primarily
 Spartina foliosa  and Salicornia virginicd) between
 the  natural tidal  marshes  and constructed  tidal
 marshes. These differences appeared to be related
 to the poor development of a tidal slough channel
 network at the constructed tidal marshes. At the
 two constructed  tidal  marshes,  plant heights of
 Spartina and S. virginica were greater where there
 was  a   well-developed slough  channel system,
 compared to areas where marsh surface elevation
 has  apparently prevented the formation of slough
 channels. The major conclusion of the study, is
 that the initial elevation of the dredged materials
 when tidal activity is restored, is a major determi-
 nate of  slough channel density and morphology in
 constructed marshes.  The presence  of a  well-
 developed  tidal  slough  channel   system  is
important if the constructed marsh is to achieve
functional equivalence with natural tidal marshes.

 San Francisco Bay (Bay) is an important natural re-
source, containing a variety of habitats ranging from
deep water to shallow intertidal  mudflats and tidal
marshes. Because of the geomorphological structure
of the Bay, maritime trade has become an important
component of the economy of the communities sur-
rounding the Bay. However, historical land use prac-
tices have  greatly reduced some  of the valuable
natural features of the Bay,  especially the acreage of
tidal marshes along the margins of the Bay.
 The historical loss of tidal marshes around the Bay
margins has been well documented  (e.g., Josselyn
1983). Since the middle of the last century, one-third
of the Bay has been lost due to filling and 90 percent
of the Bay's wetlands (tidal marshes and other types
of wetlands) converted to other  uses (National  Re-
search  Council  1992). Prior to  the  gold rush, the
acreage of tidal marshes  (2,200 km2) was twice the
area of the Bay's open water (National Research
Council 1992). Approximately 35 percent of the tidal
marshes were located in  the Bay and the remaining
65 percent  in  the Sacramento  River-San  Joaquin
River  delta (Delta)  (National  Research  Council
1992).  As a result of diking, filling, and anthropo-
genic sedimentation, only about 5 percent of the his-
torical tidal marsh acreage remains (Atwater 1979).
 Sedimentation of navigational channels and harbors
has established a need to dredge to maintain and ex-
pand the maritime trade in the Bay. The disposal of
dredged materials has been controversial in recent
years due to the perceived impacts to the environment
resulting from the deposition of dredged materials,
' ENTRIX, Inc., 590 Ygnacio Valley Road, Suite 200, Walnut Creek, California 94596.
2 Philip Williams & Associates, Ltd., Pier 35, The Embarcadero, San Francisco, California 94133.


                                                                              Tidal Marsh Design Feature
especially in the Bay. The use of dredged materials to
create habitat, including tidal marshes, is not new and
was the subject of the U.S. Army Corps of Engineers
(Corps)  Dredged   Material   Research   Program
(Environmental Laboratory  1978). This program con-
cluded that the use of dredged material as a substrate
for  tidal  marsh   development   was  a  feasible

  Several tidal marsh restoration projects established
on dredged materials have been completed in the Bay
and Delta region, and several projects are in the plan-
ning stage. Hayward Salt Pond No. 3 in the  south
Bay,  Muzzi Marsh  in the north  Bay, and  Donlan
Island and Venice Cut in the San Joaquin River were
all established using dredged materials and have been
subject to long-term monitoring programs (National
Research Council 1994). Faber Tract is another tidal
marsh restoration on dredged materials, but there has
been  only limited  monitoring of  this project, which
occurred early in its project life.
  Determination  of success of these and other resto-
ration projects is a matter of some controversy. The
four projects constructed in the Bay with long-term
monitoring (Salt Pond No. 3, Muzzi Marsh, Donlan
Island, and Venice Cut) are considered to be success-
ful based on perceived attainment of project  goals
and objectives (England and  Nakaji 1990, Landin et
al. 1989, Landin  1985, 1990).  Landin (1985) re-
viewed the results  of multi-year  monitoring of a
number of wetlands constructed with dredged  mate-
rials throughout the United States, including Califor-
nia, and her conclusions regarding the success  of
these projects were generally favorable.
  Determination  of  success  of tidal  marshes  con-
structed  on dredged materials has been hampered by
the lack  of well defined goals and objectives and in-
adequate monitoring. For example, Salt Pond No. 3 is
dominated primarily by Salicomia virginica and no
slough channels have formed naturally. Some may
conclude that  the  presence  of  relatively  dense
S. virginica means that the project was  a success.
However, because Spartina  foliosa  was  originally
planted,  the implied purpose of the project (establish
a Spartina marsh)  was not attained and the resulting
constructed tidal marsh does not approach functional
equivalency of natural Spartina marshes.
  Minimal information has been developed on design-
ing tidal  marsh systems to maximize the likelihood of
achieving functional equivalence  in constructed tidal
marshes. Questions remain about the time frame for
attaining functional equivalency in constructed tidal
marshes, and even if functional equivalency can be
attained. Numerous  studies  have documented the
development of  various  structural and  functional
attributes of constructed tidal marshes but  no clear
pattern  has emerged  from  these   studies  (e.g.,
Cammen et  al.  1974, Seneca et al. 1976, Craft et al!
1988, Zedler 1988, LaSalle et al. 1991, Havens et al.
1995, Simenstad and Thorn 1996). In fact, Havens et
al. (1995) concluded that "[TJhe  question of if and
how long does it take for a  constructed  marsh to
achieve the  same level of function as similar natural
marshes is unresolved."
  The lack of specific guidance relative to designing
and  constructing tidal  marsh  systems to  achieve
functional  equivalency,  or to approach  functional
equivalency in the shortest time possible, is troubling.
We felt that the tidal marshes  constructed on dredged
materials in the Bay offered an opportunity to iden-
tify key components of the tidal  marsh system that
should  be considered in the  design  of future tidal
marshes constructed on dredged materials. Therefore,
this study was designed to: (1) investigate the similar-
ity and differences  in  selected components  of the
physical and  biological structure of natural tidal
marshes compared  to   tidal  marshes created  on
dredged materials and;  (2) discuss important design
features to consider in designing tidal  marsh restora-
tion projects using dredged materials to increase the
likelihood of success.

Description of Study Sites
  Two pairs of sites were studied: (1) Muzzi Marsh
and the Corte Madera Ecological Reserve (CMER) in
the north Bay  (MHHW - 4.34 ft. NGVD); and (2)
Faber Tract and Laumeister Marsh in the south Bay
(MHHW-3.13 ft. NGVD) (
   1). Muzzi Marsh and CMER were  once  part of  a
historical tidal marsh system  that  extended along the
edge  of Corte Madera Bay. While CMER  remained
undiked, Muzzi Marsh was diked in the 1950's for use
as a  future industrial site. Approximately  750,000
cubic yards of dredged material were placed within
the diked portion of the property  (Faber 1983, 1988,
1990),  with the  landward portion of the  site at  a
higher elevation (about  1.3 feet higher than MHHW)
than the bay ward portion (about 1.1 to 2.1 feet below
MHHW). A training dike separated  the upper and
lower  portions.  Tidal  activity  was restored by
breaching the bay ward dike in 1976.


      - .;;:::;Corte
          Muzzi  Mars

                               :i;::::OYv-:v;':^$iin Mated
                                           SOUTH  SAN   .
£$Sa n •:Lor"efrzo'WillOUK*; •1!lllft:-pte:;:v''llfe":
 \    .••r;:^> jjt;|f (I^I'vll^'^^^lMMSI
                -    - --:•'-:::'•>.•:•:-• '• ••••'•••••• •:'•:'•".- i/.-y-v.'.-'.-.•'.•'.•'•'.•'.:•::• -,•:•:• x•:•••:-••:-: ••••:• ,vX;t•;,,.•;.;.;

                                                               Palo Alto "^-^£^ Coyote  Cr'eek  y
    Figure 1. Location of tidal marshes surveyed in this study.

                                                                              Tidal Marsh Design Features
 Faber Tract and Laumeister Tract were once part of
an extensive tidal marsh along the west side of south
San Francisco Bay. The Laumeister Tract remained
undiked,  while the historical  marsh at Faber Tract
was diked and drained in 1945 for cattle grazing. In
1968, 200,000 cubic yards of sediment was deposited
at Faber Tract for the restoration of a tidal marsh and
opened to tidal  circulation  in 1971 (Harvey 1970,
Harvey et al. 1982). Our estimate of the average ele-
vation of Faber Tract following  discharge of  the
dredged material in 1969 ranges from 2.9 to 3.5 feet
NGVD (Gahagan & Bryant Associates et  al.  1994).
The upper portion of Faber Tract ranged in elevation
from 4.2 feet NGVD (-0.1 ft. below MHHW) to an
average elevation of 3.5 feet NGVD near the middle
of the marsh, while the lower  portion of Faber Tract
sloped down to an elevation of 2.6 feet NGVD.


Physical Measurements
  Topographic profiles  were surveyed along a transect
from the  high end to the low end of the marsh plain in
each of the four tidal marsh systems. The preliminary
location  of  each transect was selected using recent
aerial photographs  of each of the tidal marsh systems
and finalized during the field survey. The topographic
profiles provided data  on the  elevation and gradient
of the existing marsh plain and the location and depth
of slough channels not visible on aerial photographs.
In addition, the aerial  photographs were  used to de-
termine the following morphometric characteristics of
the tidal  marsh slough  channel system: stream order,
drainage density, and sinuosity. These measures pro-
vide a quantitative basis for determining  the similar-
ity  of   tidal  slough  channel  systems  between
constructed tidal marshes and natural tidal marshes.
Stream Order
  We used  Horton's system of stream  ordering to
characterize the  tidal slough channels on the marsh
plain (see Straler 1964). We used a consistent scale
of 1 inch equal to 500 feet on all of the morphometric
measurements to address the problem  of map or
photo scale (Leopold  and Miller 1956)  so that  the
marshes could be compared.
Drainage Density

  Drainage density  (D) provides  a measure of the
abundance of slough channels. Drainage density  is
defined as the length of all channels in a marsh (EL)
divided by the marsh area (A):
                    D =
  In tidal marshes, a high drainage density is associ-
ated with frequency and sinuosity of slough channels,
both of which are characteristic of good tidal circula-
tion. The frequency of slough channels refers to the
number present within a specific area. The sinuosity
refers to the length of the channel along its thalweg
(or deepest part) relative to the straight line distance
between its  head in the upper part of the marsh and
its mouth at the Bay.

  Slough channel  sinuosity  is the ratio of slough
channel length (Lc) to the straight line length (Ld) of
the slough channel drainage system:

                    S  = ^
  Sinuosity is a characteristic of slough channel de-
velopment on a marsh plain. Poorly developed slough
channels  tend  to  be relatively straight while well
developed channels tend to be relatively sinuous.

Biological Measurements
  Vegetation structure was investigated along a single
transect established in each marsh. This transect ex-
tended from the upland/tidal marsh boundary of each
marsh to  the low  marsh/mudflat boundary (edge of
the vegetation). The  selection of transects at CMER,
Faber Tract, and Laumeister Tract was based on the
physical characteristics of the marsh and on review of
recent aerial photographs.  The transect surveyed  at
Muzzi Marsh was the same as that used  as part of
continuing studies at Muzzi Marsh (Faber 1990). We
limited our sampling  points  to  the  topographic
transect  because  we  wanted to investigate the
relationship between measured vegetation parameters
and the physical data (e.g., elevation, extent of slough

Tidal Marsh Design Features
  The location of plots along each transect  was de-
termined randomly. At each sampling plot,  the fol-
lowing  data  were collected from a  1-meter-square
quadrat: percent live cover by  cover class (class 1  =
0-1 percent,  class  2  =   2-5 percent,  class   3  =
6-25 percent, class 4  =  26-50 percent,  class  5  =
51-75 percent, class 6 = 76-100 percent), percent bare
ground, and  height of each species.  In plots with  a
cover class for  Spartina foliosa greater  than 3, the
length of each shoot of Spartina in a 0.1-meter-square
quadrat randomly located in one of the four corners
of  the  larger quadrat was measured  to  the nearest

 Data Analysis
  Differences in percent cover of Spartina foliosa and
 Salicornia virginica between pairs of stations (CMER
 vs. Muzzi Marsh and Laumeister Tract vs.  Faber
Tract) were investigated using the Mann-Whitney test
 at  a  significance  level of 0.05.  Analysis  of  other
 marsh vegetation species was not attempted  as  these
 species were not common at all the stations. Differ-
ences  in plant height of the tallest occurrence  of S.
virginica  in  each plot  was  investigated using the
paired /-test.  Analysis of variance was used to inves-
tigate  differences in the  average shoot length of
Spartina between Muzzi  Marsh, Laumeister Tract,
and Faber Tract. Spartina was not present in  the sam-
ple plots at CMER.


Topographic Profiles
  The marsh  plain at CMER was relatively  flat with
an  average elevation of  3.4  feet NGVD  (slightly
higher than MHHW). Slough channel bottoms along
the transect range in elevation between 3.0 to 1.1 feet
NGVD. The marsh  plain  at  Muzzi  Marsh can be
divided into two portions, the upper, higher elevation
marsh from the western levee to about 950 feet at the
location of the training dike, and the lower elevation
marsh from 950 feet to the Bay (Figure 2). The upper
marsh  had an average elevation of 3.3 feet NGVD
(slightly higher than MHHW). The lower marsh had
an  average elevation of  2.3  feet NGVD  between
MLW and  MTL. Slough channel bottoms along the
transect in the upper part of the marsh ranged in
elevation   between  -0.1   and  2.0   feet  NGVD
(Figure 3).
 The marsh plain at Faber Tract sloped toward the
north away from  the original  dredge material dis-
charge locations with a gradient of about 0.0005 from
a distance of 400 feet to 1,800 feet along the transect
(Figure 4). The upper part of the marsh had an aver-
age elevation of about 4.0 feet NGVD and a higher
gradient  of about  0.001. There were no  developed
slough channels in the  upper portion of the  Faber
Tract from a distance of 0 to 400 feet along the tran-
sect (Figure 5). In the lower portion of the marsh, the
average  marsh plain  elevation was  about 3.3 feet
NGVD and slough channel bottoms along the transect
range in  elevation from 3.2 to  0.9 feet NGVD. The
marsh plain at Laumeister Tract was relatively flat
and  had  an average elevation of 4.2  feet NGVD
(between MHW and MHHW)(Figure 4). The bay-
ward portion of the marsh was slightly higher (0.1 to
0.2 feet)  on the average than the landward portion of
the marsh, which indicates greater sedimentation on
the bayward portion of the marsh plain. Slough chan-
nel bottoms along the transect ranged in elevation
between  3.5 to 0.0 feet NGVD (Figure 5).

Tidal Slough Channel Morphology
 Slough  channel  morphology  on the  natural tidal
marshes  was found to differ from that on the con-
structed  tidal  marshes (Table  1). The  natural tidal
marshes  and   lower portions  of  the  constructed
marshes  where the dredged material  was placed at a
relatively low elevation, dense networks  of sinuous
slough channels have developed (drainage densities
range from about  30 to 40 mi/mi2).  In contrast, on
portions  of the constructed marshes where the initial
elevation of the  dredged  materials  was too high
(upper portion of Muzzi Marsh and the upper portion
of Faber Tract) slough  channels were poorly devel-
oped after at least two decades (see Figure 2 [Muzzi
Marsh] and 4 [Faber Tract]).
 The difference  in tidal slough  development is re-
flected in the morphometric  measures made at each
tidal  marsh (Table 1).  At Muzzi Marsh,  the  low
marsh had substantially greater drainage density than
the higher marsh. Although Faber Tract had a higher
slough channel density  than Laumeister Tract, the
sloughs  at Laumeister Tract were more sinuous. At
Faber Tract, there were few tidal slough channels in
the higher southern portion of the marsh. In addition,
the fourth order channel may  be an  artifact of con-
struction of the marsh as it slopes to the north rather
than bayward (Figure 4).

                                                        Slough System
                           Selected Slough System
Figure 2.      Tidal Slough Channels at Corte Madera Ecological  Reserve
              (top) and Muzzi Marsh (bottom).

Corte Madera Ecological Reserve

    Longitudinal Profile, 1993
             £  "•
             o  ">
                    IBM #4
                                                                    Corla Madera
              c  &
                                                                       .Pay \
              > »


                         200       400       660       800

                                         Distance in feet
                               Muzzi Marsh,  Longitudinal Profile
                                         1000  1200   1400  1600   1800   2000  2200   2400  2600
                                          Distance in feet
             Figure 3.     Topographic Profile at Corte Madera Ecological Reserve (top)

                          and Muzzi Marsh (bottom).

                                                                 500  FEET
 Slough  System
                                                 Transect    D.te of Photography. 1992
Figure 4.       Tidal Slough Channels at Laumeister Tract and Faber Tract.

                                        Laumeister Tract

                                   Longitudinal Profile, 1993
                            200    400   600    800    1000   1200    1400   1600   1800   2000

                                          Distance in feet

        Faber Tract

Longitudinal Profile, 1993








julto Top of Levee Road

Ranae Pole \
Start ot Transect Rebar
/ TBM #1 \
/ \ \
/ \
' \
"^ 	 •- ^1 ^
Y 1 V-~J.. — 	 	 1
1 1 "
1 " \
1 \
- - - . \
: \
-200 6 200 400 600 800 1000 1200 1400 1600 1800

                                         Distance in feet

            Figure 5.     Topographic Profile at Laumeister Tract (top) and Faber Tract


Table 1.      Morphometric Characteristics of Selected Slough System.
Slough Channel Order
Number of Sloughs in
Length of Sloughs in
Order (mi.)
              Drainage Density of Selected Slough System = 56 mi./mi/
              Sinuosity = 1.11
Slough Channel Order
Number of Sloughs in
Length of Sloughs in
Order (mi.)

Lower Marsh
       •      Drainage Density of Selected Slough System = 39 mi./mi.^
       •      Sinuosity = 1.51
Upper Marsh
       •      Drainage Density of Entire Marsh = 9 mi./mi.^
Slough Channel Order
Number of Sloughs in
Length of Sloughs in
Order (mi.)
              Drainage Density of Selected Slough !
              Sinuosity = 1.21
Slough Channel Order
Number of Sloughs in
Length of Sloughs in
Order (mi.)
              Drainage Density of Selected Slough System = 41 mi./mi/
              Sinuosity =1.10

 Tidal Marsh Design Features
  The relation of slough channel development to ini-
 tial elevation after placement of dredged material at
 Muzzi Marsh and Faber Tract suggests that the initial
 elevation of the dredged material was the most impor-
 tant determinant of slough channel development  in
 marshes  constructed using dredged  materials.  The
 upper portion of Muzzi Marsh was filled to 4.4 feet
 NGVD, an elevation higher than MHHW. In contrast,
 the lower portion of Muzzi Marsh ranged between 1.0
 to 2.0 feet NGVD (between MTL and MHW) similar
 to the elevation prior to discharge  of dredged mate-
 rial. Only one small slough  channel system  has de-
 veloped  naturally  in  the  upper portion  of  Muzzi
 Marsh. A dense network of sinuous slough channels
 exist on the mudflat and  marsh plain  in  the lower
 portion  of Muzzi Marsh,  indicating  that  the initial
 elevation of the dredged materials has allowed for the
 successful evolution of the slough channel system
 and simultaneous development of the marsh plain.

 Tidal Marsh Vegetation
 Vegetation Cover
   Salicomia  virginica was the most  abundant and
 frequently occurring species at each of the marshes
 surveyed. At CMER, 5. virginica occurred in 38  of
 the  39  plots  sampled while  at  Muzzi  Marsh,  S.
 virginica occurred in 34 of the 41 plots surveyed. The
 greatest number of species (7) was encountered in the
 sample  plots at Laumeister Tract. In addition  to
 Salicornia  virginica  and  Spartina  foliosa,  other
 species included Distichlis spicata, Jaumea camosa,
 Suaeda   sp.,  Atriplex  patula,   and  Frankenia
 grandifolia. S. virginica  occurred  in  35  of the  38
 plots sampled and 5. foliosa occurred in 33 of the  38
 plots. Only S. virginica and S. foliosa were encoun-
 tered at Faber Tract, with S. virginica occurring in  all
 the plots.

  Mean percent cover for S. virginica varied from ap-
 proximately 38 percent at Laumeister  Tract to ap-
 proximately 67 percent at  Faber Tract  (Table 2).  At
 the south Bay and north Bay sites,  the mean  percent
 cover of 5. virginica at constructed marshes (Muzzi
 Marsh and Faber Tract) was significantly greater than
 at the natural marshes (CMER and Laumeister Tract).
  Because of the influence  of elevation on the occur-
 rence of plants in tidal marshes, possible differences
 in percent cover of S, virginica  and  S. foliosa  at
 similar elevation classes between the natural marsh
and constructed marsh were investigated. Since most
of the sample  plots occurred between  MTL  and
MHW, percent cover in this elevation class was com-
pared between the four tidal marshes. The difference
in percent cover of 5. virginica between Muzzi Marsh
and CMER in the north Bay was not significant in
this elevation range, but the percent cover at Faber
Tract was significantly higher than at Laumeister
Tract. Above MHW, the percent cover of S. virginica
was significantly higher at Muzzi Marsh than CMER.
There were not enough plots above MHW at the other
two sites to test for difference in percent cover of S.
Plant Height
  The height of 5.  virginica was significantly greater
at Laumeister Tract  was  than  at Faber Tract and
Muzzi Marsh. The difference between the  other
stations  were not  significant.   Although the mean
height of S. virginica in the  elevation class MTL to
MHW at Muzzi Marsh was greater than  the mean
height at CMER at the same elevation, the difference
was  not  significant.  However, at  the south Bay
stations,  the height of 5. virginica was significantly
greater  at Laumeister Tract than  at  Faber Tract
between MTL and MHW.
Spartina Stem Length and Density
  The mean stem length (Table 3) was significantly
greater  at Muzzi Marsh  than at  Faber Tract or
Laumeister Tract, but Spartina density and total stem
length per  plot  (Table 3)  were not  significantly
different from one another.  Total stem length and
distribution pattern of individual  stem length were
similar to that observed by Zedler (1993)  in San
Diego  County tidal  marshes.  Most of the stems
occurred in the 61 to 90 cm  range, as shown by the
cumulative  percent frequency  curves for the three
sites (Figure 6). Although  the highest stem densities
occurred at Faber Tract, a constructed marsh in south
Bay, the  difference in mean density between marshes
was not significant.

  If we are to construct tidal marshes that are func-
tionally equivalent to natural  tidal  marshes, imple-
menting an approach that takes advantage of the self-
designing capacity of nature  (Mitsch  and  Wilson
1996) is the preferred option.  Attempting to define
the final planform for a constructed tidal marsh pre-
sents several challenges, each of which could extend
the time  it takes the  developing constructed tidal
marsh from approaching functional equivalency with
natural tidal marshes and possibly  even  could prevent

      Table 2.     Percent Cover for Spartina foliosa (SPFO) and Salicornia
      virginica (SAVI).
No. of quadrats sampled
Muzzi Marsh
No. of quadrats sampled
Laumeister Tract
No. of quadrats sampled
Fa her Tract
No. of quadrats sampled








Table 3.      Stem Length, Total Stem Length and Density for Spartina foliosa.
Stem Length (cm)
Number of stems
Total Stem Length (cm)
Number of quadrats
Density (/m^)
Number of quadrats
Muzzi Marsh



Laumeister Tract



Faber Tract




      g  60
     I  40
             muzzi marsh

             fabcr tract

                                60            90

                               height interval (cm)
     Figure 6.     Cumulative  Frequency  for Sparlina foliosa  Stem Length  at

                  Muzzi Marsh, Faber Tract and Laumeister Tract.

                                                                              JTidal Marsh Design Featui
the attainment of desired functions. The lack of de-
velopment of a tidal slough channel system at the up-
per Muzzi Marsh, upper areas of Faber Tract and Salt
Pond 3 appears to be the result of overfilling. Over-
filling is more likely to occur if attempting to achieve
the desired  marsh  planform  during  construction
instead of allowing the system to evolve from lower
elevations as natural sedimentation occurs.
  The differences  in the structure of the vegetation
communities  between the  natural   marshes   and
constructed marshes  appear  to be  related  to  the
development of  the  tidal slough  channel  system,
which in turn is the product of the initial elevation of
the constructed site when tidal activity was restored.
Observations at Muzzi Marsh and Faber Tract indi-
cates that in portions of the marsh where  the initial
surface  elevations were too high, slough  channels
have not developed. At Faber Tract, slough channels
were absent in the upper part  of the marsh, occurred
intermittently in the middle portion of the marsh, and
were abundant in the lower marsh. Cover of S. foliosa
was significantly lower  in  the upper  marsh  area
compared to the area of the marsh where there were
intermittent and abundant  slough channels. Spartina
growing in the  area of  the  marsh  where  slough
channels were abundant was significantly taller than
plants occurring where there were intermittent or no
slough channels. These results suggest that good tidal
flushing is necessary  to  get  maximal  growth of
Spartina and other tidal marsh plant species.
  The cover of  Salicornia virginica between MTL
and MHW at Laumeister Tract was significantly
greater than at comparable elevations at Faber Tract
and the plants were significantly taller. The  shorter
plants at Faber Tract, especially at the higher eleva-
tions, could reduce the value of this area as habitat
for species dependent on the upper tidal marsh areas,
such as the federally endangered salt marsh harvest
mouse (Reithrodotomys raviventris).
  The importance of  tidal  flushing and the develop-
ment of a  full array of tidal channels, including the
small first-order  channels (tidal rivulets  of some
authors), to the  biological functioning  of the  tidal
marsh  system is  well documented.  Simenstad  and
Thorn (1996) surmise that development of the tidal
slough channel system facilitates access to the inter-
tidal marsh surface by fish and motile macroinverte-
brates. Although they did not test whether  fish usage
of the marsh surface was a function of dendritic tidal
slough channel development, they did suggest  that
development  of  a dendritic   tidal  slough  channel
system may be an  indicator of marsh development
(Simensted and Thorn 1996).

  Others have also addressed the importance of tidal
flushing and the development of tidal rivulets on the
marsh surface. Mclvor and Odum (1988) suggest that
the sinuosity of tidal creeks, channel depth and bank
stability may affect the utilization of the tidal creeks
and  adjacent marsh  surface by  fish.  Minello and
Zimmerman (1983) and Rozas  et al. (1988) discuss
the importance of tidal rivulets and suggest that the
tidal rivulets provide access to the marsh surface for
juvenile fish and macroinvertebrates and refuge dur-
ing low tide. Havens et al. (1995) emphasize the im-
portance of tidal rivulets and recommend that tidal
rivulets, as a component of marsh microtopography,
be considered in  the design  of constructed wetlands.
Tidal rivulets also increase the amount of edge in the
marsh system and this "edge" habitat has been found
to be important for fish and  macroinvertebrates (e.g.,
Thomas et al. 1990, Rozas  1992,  Baltz et  al.  1993,
Peterson and Turner 1994).
  In north Bay, the presence of the slough  channels
(tidal rivulets of other authors), especially in the up-
per marsh plain, may have an affect on the presence
of the federally  endangered California clapper rail
(Rallus longirostris obsoletus). Recent work on clap-
per rails in the north Bay suggests that the primary
channels (rivulets) found in the upper marsh may be
important to the existence of this species in a particu-
lar tidal marsh  system (].  Collins, SFEI,  personal
communications). The clapper rail appears to prefer-
entially nest in the upper marsh adjacent to the pri-
mary channels near the upper part of the tidal  zone
and use the primary  channel as travel routes to the
lower marsh where it forages. If this is the case, the
lack of adequate  development of a tidal slough chan-
nel  system, especially  the  primary and secondary
slough channels, could affect the ability of the tidal
marsh to support  clapper rails.
  Construction  of the "final" marsh plain  elevation
can also lead to  a  delay in attainment of functional
equivalency. After  approximately  20 years, the por-
tion of Muzzi Marsh and Faber Tract that were above
MHW when tidal  flushing was restored, were still
structurally different from  nearby natural  marshes.
The delayed development of the tidal slough channel
system at Muzzi Marsh and upper Faber Tract, result-
ing from overfilling the area with dredged materials,
has resulted in greater cover by pickleweed but the
plants are shorter, less robust. The  lack of tidal
flushing related  to the  poor development of a tidal

Tidal Marsh Design Features
 slough channel system in the constructed marshes is
 probably the major factor that has contributed to the
 observed  differences in  the  structure  of the  marsh
  In the Gog-Le-Hi-Te estuarine  wetland restoration
 project (Washington), the main tidal channels were
 dug during construction Simenstad and Thorn (1996).
 Recent monitoring of the constructed  tidal channels
 indicates that they are filling with sediments and that
 a dendritic tidal channel pattern is developing in the
 adjacent intertidal areas, including the marsh surface.
 Construction  of the tidal channels may have slowed
 the attainment of a functional tidal slough channel
 system, thus  delaying  development of the desired
 functional attributes. As  Simensted and Thorn (1996)
 caution, the use of "ecotechnological fixes," such as
 the construction  of tidal channels to  facilitate tidal
 flushing,  may not produce  the  same  endpoints  as
 natural, longer-term processes, such as allowing the
 final marsh surface to develop through natural sedi-
 mentary processes.
  Results  of this  study and analysis of available in-
 formation provides a basis for developing design cri-
 teria for future projects involving the use of dredged
 materials to create or restore tidal marsh habitat. Cre-
 ating the proper physical conditions is key to devel-
 oping the desired  tidal marsh  ecosystem.  If  the
 surface elevation of the created tidal marsh system is
 too high, development of slough channels will be
 slow and, in some cases,  the slough channels may not
 develop. The  resulting tidal marsh habitat  would be
 restricted to high marsh dominated by Salicornia vir-
 ginica with  few other species being present.  The
 resulting marsh system would be incomplete and have
 limited functional capacity relative  to natural tidal
 marsh systems.

  The major conclusion of our study is  that  the initial
 elevation of the dredged materials, when tidal activity
 is restored, is an extremely important determinant of
 slough  channel density  and morphology  in con-
 structed marshes. The presence of a well-developed
 tidal  slough channel system  is important for  devel-
 opment of desired biological structure  and related
 function at the constructed tidal marsh. This conclu-
sion is based on the following findings of this study:

  •   Slough channels appear to form initially  at ele-
     vations  between  MIL  and MHW (-1.5  feet
     below MHHW) the reported approximate lower
     elevation range for Spartinafoliosa.
 •   Slough channels either do not form naturally, or
     form very  slowly, at elevations about 1.5  feet
     below MHHW.

 •   Slough  channel  morphology  at natural  tidal
     marshes differs from constructed marshes in the
     orientation of  the  slough channels, sinuosity
     and drainage density.
 •   There are significant differences in the structure
     of the tidal marsh  vegetation (primarily Spar-
     tina  and S. virginicd) between the natural tidal
     marshes  and constructed tidal  marshes. These
     differences appear to be related to differences in
     the formation of a tidal slough channel network
     at the constructed tidal marshes.
 •   Plant height of Spartina and 5. virginica are
     greater  where  the  drainage density of slough
     channels is higher.  At the two constructed tidal
     marshes, the plant heights are greater where
     there is a well-developed slough channel system
     compared to areas  where marsh surface eleva-
     tion  has apparently prevented the formation of
     slough channels.

 This study was performed as part of a contract let to
Gahagan  & Bryant  Associates by  the  U.S.  Army
Corps of Engineers.  This study is part of a series of
studies designed to provide information for the Long
Term Management Strategy for dredged  materials in
San Francisco Bay. The authors wish to  thank Rick
Olejniczak, project manager for Gahagan & Bryant
Associates for his support. Eric Larson, Bob Batha,
and  Steve Goldbeck from the San  Francisco  Bay
Conservation  and  Development Commission  pro-
vided helpful direction in the design  of the study and
comments on the project report. Eric Larson has been
particularly helpful  in providing comments on the
project report from which this paper was taken. The
authors also wish to thank those who assisted us  in
the field and in the preparation of the main project
report from which this paper was taken.

Literature Cited
Atwater, B. F. 1979. History, landforms,  and vegeta-
  tion of the estuary's tidal marshes.  In T.J. Comonos
  (ed.) San Francisco Bay, The Urbanized Estuary.

                                                                             Tidal Marsh Design Feature
  American Association for the Advancement of Sci-
  ence, Pacific Division, San Francisco, CA.
Baltz, D. M., C. Rakocinski, and J. W. Flegger.  1993.
  Microhabitat use by marsh edge fishes in a Louisi-
  ana estuary. Environmental Biology of Fishes  36:
Carnmen, L. M., E. D. Seneca, and B. J. Copeland.
  1974. Animal colonization of salt marshes  artifi-
  cially established on dredge  spoil. University  of
  North Carolina, Chapel  Hill,  NC.  Sea Grant Pro-
  gram Publication UNC-SG-74-15.
Craft, C. B., S. W. Broome, and E. D. Seneca.  1988.
  Soil nitrogen, phosphorus,  and organic carbon in
  transplanted estuarine  marshes, pp.  351-358  In
  D. D. Hook (ed.) The Ecology and  Management of
  Wetlands. Timber Press.
England, A.  S.  and  F.  Nakaji.  1990. Reclaiming
  flooded  river  islands  as  wetlands  and riparian
  habitat using dredged material. In  Proceedings of
  Beneficial Uses of Dredged Material in the Western
  United States.  U.S. Army Engineers  Waterways
  Experiment Station, Vicksburg, MS.
Environmental  Laboratory.  1978. Wetland habitat
  development with  dredged material:  engineering
  and plant propagation.  Technical Report No. DS-
  78-16, U.S. Army Engineer Waterways Experiment
  Station, Vicksburg, MS.
Faber,  P.  1983. Marsh  restoration  with  natural
  revegetation:  A case study in San Francisco Bay.
  pp. 729-734 In O.  T. Magoon (ed.) Coastal Zone
  '83. American Society of Civil Engineers, NY.
Faber, P. 1988. Current Status of the Muzzi Marsh:
  Report No. 5, Final Addendum. Prepared for the
  Golden Gate Bridge, Highway, and Transportation
Faber, P. M. 1990. Long-Term Observations of the
  Muzzi Marsh, Corte Madera, CA.
Gahagan  &  Bryant  Associates,   Bechtel   Corp,
  ENTRIX Inc., and Philip Williams & Associates.
  1994. A Review of the Physical and Biological Per-
  formance  of  Tidal  Marshes  Constructed  with
  Dredged Materials  in San Francisco Bay, CA. Pre-
  pared for  Corps  of  Engineers,  San  Francisco
Harvey, H.  T. 1970. Letter to Mr. Forest Warthman,
  BCDC Research Assistant, BCDC File.
Harvey, H. T., P. B. Williams, Madrone Associates,
  and San Francisco Bay Conservation and Develop-
  ment Commission.  1982. Guidelines for enhance-
  ment and restoration of diked historic baylands. A
  technical report prepared for San Francisco BCDC.
Havens, K. J., L. M. Varnell,  and J. G. Bradshaw.
  1995. An assessment of ecological conditions in a
  constructed tidal marsh and two natural reference
  tidal marshes in coastal Virginia. Ecological Engi-
  neering 4: 117-141.

Josselyn, M. 1983. The ecology of San Francisco Bay
  tidal marshes:  A community  profile. U.S. Fish and
  Wildlife  Service, Biol. Serv.  Proj. FWS/OBS-

Landin, M. C. 1990. Wetland restoration, protection,
  and establishment by beneficially  using dredged
  material. Proceedings of American Society of Civil
  Engineers Water Forum '92, NY: 114-118.
Landin, M. C. 1985. Long-term monitoring of habitat
  development  sites,  1974-1984. Environmental  Ef-
  fects of Dredging. Information Exchange Bulletin
  D-85-3: 1-7.
Landin, M. C., J. Webb, and P.  Knutson. 1989. Long-
  term  monitoring of  eleven  Corps of Engineers
  habitat development  field sites  built of dredged
  material, 1974-1987. Dredging Operations Techni-
  cal Support Program Technical Report D-89-1. U.S.
  Army Corps of Engineers  Waterways Experimen-
  tation Station.
LaSalle, M. W., M. C. Landin,  and J. G. Sims. 1991.
  Evaluation of the flora and fauna of a Spartina al-
  terniflora marsh established on dredged material in
  Winyah Bay, SC. Wetlands 11:101-108.
Leopold, L. B.  and J. P. Miller.   1956. Ephemeral
  streams-hydraulic factors and their relation to the
  drainage net. USGS Professional  Paper 282-A.
Mclvor,  C.  C. and  W. E.  Odum. 1988.  Food,
  predation  risk, and  microhabitat selection in a
  marsh fish assemblage. Ecology 69:1341-1351.
Minello, T. J. and R. J. Zimmerman. 1983.  Fish pre-
  dation on juvenile brown shrimp, Panaeus aztecus
  (Ives): the effect of simulated Spartina structure on
  predation rates. J. Exp. Mar. Biol. Ecol. 72:211-
Mitsch, W. J. and R. F. Wilson. 1996. Improving the
  success of wetland creation and restoration with
  know-how, time and self-design. Ecological Appli-
  cations 6:77-83.

 Tidal Marsh Design Features
 National Research  Council.  1992.  Restoration of
   Aquatic  Ecosystems.  National  Academy  Press,
   Washington, D.C.
 National Research  Council.  1994.  Restoring  and
   Protecting  Marine  Habitat.  National  Academy
   Press, Washington, D.C.
 Peterson, G. W. and R. E. Turner. 1994. The value of
   salt marsh edge vs. interior as a habitat for fish and
   decapod  crustaceans in  a Louisiana tidal marsh.
   Estuaries 17: 235-262.
 Rozas, L. P. 1992. Comparison of nekton habitats as-
   sociated with pipeline canals and natural canals in
   Louisiana marshes. Wetlands 12: 136-146.
 Rozas, L. P., C. C. Mclvor, and W. E. Odum. 1988.
   Intertidal rivulets and creek banks: corridors be-
   tween tidal creeks and marshes.  Mar. Ecol. Prog.
   Ser. 47: 303-307.
 Seneca, E.  D., S. W.  Broome, W. W. Woodhouse,
   L. M. Cammen, and  J. T. Lyon. 1976. Establishing
   Spartina  alterniflora  marsh in  North  Carolina.
   Environmental Conservation 3: 185-188.
Simenstad, C. A. and R. M. Thorn. 1996. Functional
  equivalency trajectories of the restored Gog-Le-Hi-
  Te  estuarine wetland. Ecological  Applications 6-

Straler, A. N. 1964. Quantitative geomorphology of
  drainage basins and  channel networks,  pp.  4-39
  through 4-76  In  V.T.  Chow (ed.) Handbook of
  Applied Hydrology. McGraw Hill, NY.

Thomas, M. A., R. J. Zimmerman, and T. J. Minello.
  1990. Abundance patterns of juvenile blue crabs
  (Callinectes sapidus) in nursery habitats of two
  Texas bays. Bulletin  of Marine Science 46:  115-

Zedler, J. B.  1988. Salt  marsh restoration: lessons
  from California, pp. 123-138 In J. Cairns (ed.) Re-
  habilitating  Damaged  Ecosystems. CRC  Press,
  Boca Raton, FL.

Zedler, J. B. 1993. Canopy architecture of natural and
  planted cordgrass marshes: selecting habitat evalua-
  tion criteria. Ecological Applications 3: 123-138.


                                        Regional Restoration Planning for
                                          San  Francisco Bay Salt Marshes:
                                                    A Biogeographic Approach
                                                                              Kevin MacKay1
 Abstract:  Eighty-three  wetland  sites,  located
along the southwestern shoreline of San Francisco
Bay, were studied to identify possible relationships
between the diversity of resident wildlife species
and the biogeographic characteristics of the sites.
The distribution of five special status species: the
California  clapper rail,  California black  rail,
Alameda song sparrow, salt marsh harvest mouse,
and the salt marsh wandering shrew, were used as
indicators of species diversity. National Wetland
Inventory maps and aerial photographs were used
to identify and quantify ten biogeographic charac-
teristics for each site: shape, cover type, tidal re-
gime, area, total perimeter, percentage of natural
perimeter,  percentage  of unnatural  perimeter,
percentage  of  aquatic  perimeter,  degree  of
channelization, and isolation. Six of the ten char-
acteristics, cover type  (r = 0.729), tidal regime (r =
0.86), area (r = 0.3),  total perimeter  (r = 0.383),
percentage of aquatic perimeter (r = 0.664), and
degree of  channelization (r = 0.702)  were posi-
tively correlated with  occurrences of the indicator
species. The remaining four characteristics, shape
(r =  -0.441), percentage  of  natural perimeter
(r = .108), percentage  of unnatural perimeter (r =
-0.616), and isolation  (r = -0.254) were negatively
correlated  with  indicator  species  occurrences.
Overall, the relationship between the wetland sites
and  species  diversity  was  best  explained  by
a combination of  characteristics  (r2= 0.631) in-
cluding cover type, tidal regime, total perimeter,
and isolation. The final results of the  study were
used to develop a predictive model to guide the re-
storation of wetland habitat and the conservation
of associated resident wildlife species  in the San
Francisco Bay Estuary.
 Human development of the San  Francisco Bay
Estuary has significantly reduced the area of historic
wetlands and led to  the endangered or threatened
status of certain  wetland-dependent  species.  Before
1850, 1400 km2 of freshwater marsh surrounded the
confluence of the Sacramento and San Joaquin rivers,
and another 800 km2  of saltwater marsh fringed the
Bays' shores. As population increased, tidal marshes
were  diked to create  farmland,  salt  evaporation
ponds, and later residential and  industrial land.
Reclamation of freshwater marshes in the Delta was
essentially complete by the 1920's, but the filling and
conversion  of the Bay's saltwater marshes continued
until  the early the 1970's. Only  125 km2  or five
percent  of the original  2200 km2  of  tidal  marsh
remains today (Nichols, et al. 1986).
 While mitigation to compensate  for these  historic
losses has taken place under federal and state permit
policies for over 25 years, it has been implemented on
a project-by-project basis, often ignoring the regional
perspective. The legacy of this historic development
and site-specific mitigation strategies is the fragmen-
tation of once contiguous bands of tidal marsh habitat
into patches scattered across the regional landscape.
These patches  or remnants support different  vegeta-
tion types and tidal regimes, and vary greatly in size,
shape, cover type,  and proximity  to  other sites.
Within the Estuary the conservation of regional bio-
diversity depends entirely  upon the retention  and
maintenance of these  remnants. Regulatory and plan-
ning agencies are therefore faced with the dual issues
of determining whether the remnants have any practi-
cal conservation values, and if they do,  of how they
should be managed to restore or enhance these values
(Saunders et al. 1991).
' Department of Geography, San Francisco State University, 1800 Holloway Avenue, San Francisco, California 94132.

                              REGIONAL RESTORATION PLANNING FOR
                                 SAN FRANCISCO BAY SALT MARSHES
                                               STUDY SITES
                             Redwood City
                                              East  (43
                                              Palo Alto
          i)  Site Identification Number
                      South San Francisco
                      Map Series 87\02-D
             Figure 1. Map of the southwestern shoreline of San Francisco Bay showing the distribution of the 83 study sites.

                                                                             San Francisco Bay Salt Marshes
 The concept of remnants of natural vegetation as
habitat islands, and the changes that result from the
isolation of these  islands have been the subject of
considerable   debate   in  the  scientific   literature
(Saunders et al. 1991). This debate has centered pri-
marily on the theory of island biogeography and its
applicability to conservation planning (Preston 1962;
MacArthur and Wilson  1967; Soule and Wilcox
1980; Wilcove 1987). In particular, the importance of
biogeographic characteristics such as size, shape and
proximity,  as well  as the  spatial  arrangement of
fragments  within  the  overall  landscape  matrix
(Temple and Wilcox 1991). Although many  studies
have addressed the relationship between these factors
and species diversity in forests, grasslands, chaparral
and coastal sage scrub habitats, limited knowledge
exists relative to their applicability to wetlands.
  The primary objective of this study  is to examine
the influence of the biogeographic characteristics of
individual salt marsh sites on the diversity of resident
wildlife species. These characteristics include cover
type, tidal regime, degree of channelization, site area,
site perimeter, site shape, and isolation. The  second
objective is to use the data produced from this analy-
sis to develop a predictive model that can be used to
guide the regional restoration planning process for
the San Francisco Bay Estuary.

Materials and Methods

Study Area
  The   general  study  area   is  composed   of  a
25-kilometer long band located along the southwest-
ern shoreline of San Francisco Bay, from Belmont
Slough, San  Mateo County in  the north, to  Alviso
Slough, Santa Clara County in the south (Figure 1).
This area was selected because it encompassed a mo-
saic of wetland sites and land use types representative
of the San Francisco  Bay Estuary. In addition, the
area is large enough to provide a regional context for
assessing  species  diversity, and small enough  that
there are no  pronounced  differences in the wildlife
and vegetation associated with the individual sites.
Eighty-three smaller study sites, classified as tidal es-
tuarine  wetlands,  or diked estuarine/seasonal  wet-
lands, were selected within the general study area to
assess  the  potential relationships between the  bio-
geographic characteristics of the individual sites and
the diversity of resident wildlife species.
Indicator Species
  Since a survey of all resident wildlife species that
occur in the study sites was outside the scope of this
study, five species were selected as indicators of spe-
cies diversity: the California clapper rail, California
black rail, Alameda song sparrow,  salt marsh  wan-
dering shrew, and salt marsh harvest mouse. These
species were chosen because of their special  legal
status, as well as the general availability of informa-
tion relevant to their ecological requirements. Species
distribution data was obtained from the U.S. Fish and
Wildlife Service (USFWS 1993, 1994), the California
Department  of Fish and Game's Natural Diversity
Data Base, and other wildlife studies  (Walton 1975,
Posternak 1986, WESCO 1986, WRA 1994).

  Ten  biogeographic variables were  identified  and
quantified for each of the 83 study sites  (Table  1).
These  variables are geographic,  biotic, and hydro-
logic  characteristics that  potentially influence the
biological diversity, or otherwise diminish the long-
term viability of a wetland site.  Jandel Scan/Image
software was used in conjunction with the appropri-
ate National Wetland Inventory maps to calculate the
geographic characteristics of each site.  Dominant
vegetation types and hydrologic regimes were deter-
mined   from  recent   aerial  photographs   (scale
1:12,000), and assigned relative point values based
upon  the ecological requirements of wetland-associ-
ated species.

Data Analysis
  Statistical  analyses were performed  with Jandel
Sigma Stat computer software. Spearman  rank order
correlations were used to determine possible relation-
ships between the number of indicator species and the
biogeographic  variables of the 83 study sites, and
multiple regression analysis to develop a predictive
model for species diversity.


Biogeographic Variables
  The study sites varied greatly in size, shape, and pe-
rimeter (Table 2). The area of the largest and smallest
sites were separated by more than two orders of mag-
nitude. Only  eight percent  (7 of  83) of the sites
exceeded 1 km2 with the majority falling between the


 San Francisco Bay Salt Marshes
   Table 1. Biogeographic variables calculated for the 83 study sites.
   Biogeographic Variable    Description
   Shape Factor

   Cover Type

   Tidal Regime



   Natural Perimeter

   Unnatural Perimeter

   Aquatic Perimeter


Measure of site shape relative to a circle of equal area
Dominant vegetation type
Degree of tidal inundation
Site area in square kilometers
Site perimeter in kilometers
% of site perimeter surrounded by tidal salt marsh vegetation
% of site perimeter site perimeter bordering urban development,  salt ponds,
roads, riprap, or dikes/levees
% of site perimeter slough channels or the Bay
% of site infiltrated by tidal channels
Distance in kilometers to the nearest  "source" site that contains one or more of
the indicator species
  Table 2. Descriptive statistics and results of the Spearman rank order correlation analysis.
Biogeographic Variable
Shape Factor
Cover Type
Tidal Regime
Area (km2)
Perimeter (km)
% Natural Perimeter
% Unnatural Perimeter
% Aquatic Perimeter
Channelization (%)
Isolation (km)
r Value
p Value
 mean  value of  0.34 km2 and  the low  value  of
 0.01 km2.  Site   perimeters  ranged  from  0.38  to
 13.33 km, with a mean value of 2.96 km. Of the three
 perimeter types  measured,  aquatic, natural and un-
 natural,  the  later occupied a greater percentage  of
 overall perimeter (mean  62.5), with the other two
 types registering  means of 32.87 and 4.06 percent re-
 spectively. Only  18 percent (15 of 83) of the sites had
 some amount of natural perimeter,  and thus were
 contiguous with  adjacent sites. Site shape was pre-
 dominately linear with the mean shape  factor equal-
 ing 0.42  (oval/linear).  Only  4 percent  of the sites
 (3 of  83)  registered a  shape  factor  above  0.75,
 roughly  the  equivalent of  a  round/irregular  shape.
 The degree of isolation of the study sites ranged from
0.00 to  1.90 km, with the mean  distance equaling
0.346 km.
                           Cover type and tidal regime  were predominately
                         coastal salt marsh with full tidal inundation. Almost
                         two-thirds (60  percent, 50 of 83) of the sites were
                         coastal salt marsh, 27 percent (23 of 83) upland, and
                         the remaining 13 percent (10 of 83) a mixture of the
                         two habitat  types. Approximately 72 percent (60 of
                         83) of the sites are at least partially inundated by the
                         daily  tidal  cycles.  The  degree  of channelization
                         within these sites varied from 5 to 50 percent, with a
                         mean of 27.9 percent. The remaining 23 sites (28 per-
                         cent)  are completely diked off  and only fill during
                         periods of heavy precipitation.

                         Data Analysis
                           Spearman rank order correlation was used for a pre-
                         liminary  analysis  to  determine  the  relationships

                                                                              San Francisco Bay Salt Marsh
between the individual biogeographic variables and
species diversity. Six of the ten variables were posi-
tively correlated with  increasing  species  diversity
(Table 2). Four variables, shape, natural  perimeter,
unnatural  perimeter and isolation  were negatively
correlated with species diversity. In all but one case,
natural perimeter, these correlations were statistically
significant (p<0.05). The r values for the individual
correlation equations ranged from -0.616 to 0.860.
  Multiple Regression  Analysis. The relatively low r
values derived from the Spearman rank order corre-
lation analysis indicate that high species diversity is
not entirely explained by any single regression model.
As such, a best subsets regression analysis was used
to select the  combination of independent variables
that best contributed to increasing species diversity.
Adjusted r2 values (r2 value adjusted for the number
of variables)  ranged from 0.514 to  0.643, with the
highest value  associated with the combination of in-
dependent variables that included cover type, tidal re-
gime, total perimeter, percentage of natural perimeter,
percentage  of unnatural perimeter,  percentage  of
aquatic perimeter, and isolation.
  However,  as this analysis indicated redundancy or
correlation between  some of the biogeographic char-
acteristics, variables with Variance Inflation Factors
(VIF) higher  than 4.0 were removed from the best-
subset analysis. This final step removed the percent-
age of aquatic perimeter,  percentage of  unnatural
perimeter, and percentage of natural perimeter vari-
ables, resulting in a model with an r2 value of 0.631.
As such, it appears that high species diversity in the
study sites is  correlated with a  combination of inde-
pendent variables including isolation, perimeter, tidal
regime, and cover type.
  Based  upon this  correlation, the four  remaining
variables were used to formulate a model to predict
species diversity in salt marsh sites. This model uses
multiple linear regression  analysis to develop  an
equation  that predicts potential  species diversity
based upon a  site's degree of isolation, total perime-
ter,  tidal regime,  and cover  type. The equation
derived from  the data obtained during this study is:
Potential Species Diversity = -0.734 - (0.258 x isola-
tion) + (0.153  x  total perimeter)  + (0.896  x  tidal
regime) - (0.214 x cover type).
  This equation was  used to formulate Potential Spe-
cies Diversity Indices (PSDI) for the 83 sites exam-
ined in  this study. PSDI scores  ranged from -0.70 to
3.33, with a mean value of 0.975.  The total  PSDI
scores were used to place the 83 study sites into four
categories  representing very high, high, moderate,
and low potential species  diversity (Figure 2). Five
sites,   or  6 percent  of the  total,  fell  within  one
standard deviation (SD) of the high value and were
assigned a very  high  potential  value. Twenty-three
sites,  or 28 percent of the total, fell between the first
and second SD  and were assigned  high potential
value. Twenty sites, or 24 percent, fell between the
second and third SD and were assigned to the moder-
ate value category. The remaining  35 sites  (42 per-
cent)  fell below the third SD and were relegated to
the low potential value category.


Influence of Biogeographic Variables
  The results from this study suggest that the ten bio-
geographic variables have varying levels of influence
over the species diversity of the 83 study sites. Fac-
tors such as area and shape, prominent  in  previous
studies of terrestrial systems,  appear to have only
slight  positive  correlation  with species  diversity in
wetland sites. While others,  such as  perimeter,  that
had negative effects in previous studies  have a sig-
nificant positive  correlation with greater species di-
versity in this study.
  Site area had a slight positive correlation with high
species diversity (r  =  0.30). This figure was some-
what lower than the correlation between area and spe-
cies diversity  in studies of chaparral requiring bird
species in California (r = 0.55), and obligate wetland
birds in Maine (r = 0.61) (Soule et al. 1988, Gibbs et
al. 1991), but was still  higher than the correlation (r =
0.22)  with  facultative  wetland birds in New York
(Grover and Baldassare 1995). This difference can be
explained by two factors.  First, the  loss of historic
coastal marsh habitat in the South Bay has forced the
indicator species to colonize even  the smallest  site.
Secondly, previous  studies  (Power  1972,  Johnson
1975, Western and Ssemakula 1981) concluded that
larger sites support a greater diversity of habitat and
topographic relief, and that these features  are more
significant than area in determining species diversity.
  Thus, because of the relatively homogenous nature
of salt marsh vegetation and the lack of site relief,
larger study sites would  not  necessarily  support  a
higher diversity of species.

                           REGIONAL RESTORATION PLANNING FOR
                              SAN FRANCISCO BAY SALT MARSHES
                                  POTENTIAL SPECIES DIVERSITY INDICES
                         Redwood City
                                              Palo Alto
          Highest Species Diversity

          High Species Diversity

          Moderate Species Diversity

          Low Species Diversity
                                                                      Mountain View
      Source: USGS, Topographic Map South San Francisco Bay, 1975,
      USFWS 1990, Dedrick 1989, and USFWS/NWRC Map Series 87V02-D
Figure 2. Map of the southwestern shoreline of San Francisco Bay showing the potential species diversity indices for the study sites.

                                                                              San Francisco Bay Salt Marshes
 The correlation between  cover type (r =  0.729),
tidal regime  (r  = 0.860),  and channelization (r =
0.702) were consistent with the known habitat prefer-
ences of  the  indicator  species.  California clapper
rails, California  black rails, Alameda song sparrows,
and salt marsh wandering shrews preferred channel-
ized sites with coastal salt marsh vegetation, full tidal
inundation, and a moderate to high level of channeli-
zation. Whereas, occurrences of  salt marsh  harvest
mice  extended to diked marshes  with mixtures  of
upland and coastal salt marsh vegetative types. Geis-
sel et al. (1988) found that salt marsh harvest mice act
as a refugial species as they are forced to move into
areas of lesser cover and higher salinity  by  superior
competitors such as western harvest mice, house mice
and California  meadow voles.  Salt  marsh  harvest
mice are able to live in these areas because of their
ability to tolerate high levels  of salt in their diet
(Shellhammer et al. 1988). For example, Zetterquist
(1977) found  that salt marsh harvest mice were the
last species to be excluded from highly saline diked
  The negative correlation  between the shape factor
(r = -0.44) and high species diversity suggests that the
optimal site is linear in shape. This contradicts the
guidelines provided by much of the scientific litera-
ture (Diamond  1976, Wilson and Willis  1975) that
declares  that  a  round shape is  far superior to any
other shape. This difference can be explained by the
fact that  these  guidelines  were  reportably derived
from  island biogeographic theory rather than from
empirical studies  of species diversity.  In  fact, no
research  has  demonstrated that round reserves  hold
more species than long thin ones (Shafer 1990). The
results of this study suggest that linear sites may be
more visible  to  immigrants,  thus promoting  the
movement of species into the site from adjacent sites,
slough channels, or bays.
  This conclusion is borne out by the relationship
between the perimeter variables and species diversity.
Total perimeter  (r = 0.383) and percent of site pe-
rimeter surrounded by water (r = 0.664) were  both
positively  correlated  with high  species diversity.
Whereas,  the percent of  site perimeter surrounded by
natural vegetation  (r = -0.108) and unnatural vegeta-
tion (r = -0.616) had negative correlations with  high
species  diversity.  This  latter  relationship is much
more  significant and is probably  a result of detri-
mental "edge" effects. This finding is consistent  with
previous  studies (Simberloff 1982, Noss  1983) that
indicate that reserves with  longer  perimeters  appear
to be more susceptible to detrimental edge effects
such as invasion by exotic plants, species predation
and nest parasitism. A long unnatural perimeter made
up of agricultural or urban land would increase the
ability of opportunistic mesopredators such as Nor-
way rats, red foxes,  or feral  cats to move into the
study sites and  feed  on  the eggs and young of the
indicator species.

  However, the positive correlations between total pe-
rimeter and percent surrounded  by aquatic perimeter
and high species diversity is quite dissimilar from the
results of these  earlier studies. This is probably be-
cause of a number of factors specific to wetland sites.
First, physiographic factors such as high salinity and
regular tidal inundation, that encourage colonization
of native salt marsh vegetation, discourage the estab-
lishment of invasive plant species. Secondly,  while
exotic predators are able to move into sites from adja-
cent agricultural or urban land,  a high percentage of
aquatic perimeter would act as a barrier to these spe-
cies. In addition, a long aquatic perimeter would pro-
mote the entry of the indicator species into the site by
swimming or rafting. California clapper  rails,  salt
marsh harvest mice, and salt marsh wandering shrews
are all known to use these methods of dispersal.
  Degree of isolation had a slight negative correlation
(r = -0.254) with the diversity  of indicator species.
This relationship differs from that discovered by Gro-
ver and Baldassarre (1995) in New York wetlands,
who found that wetland isolation was not a factor af-
fecting  species  richness  of wetland  obligate  birds.
However,  this  correlation  is  consistent  with  the
results of Soule et al. (1988), which reflected a  slight
benefit  of patch proximity for small mammals such as
rodents, rabbits and hares. This is further substanti-
ated by the studies of Walton (1975), Shellhammer et
al.  (1982), and  Bias  (1994)  which revealed  the
limited dispersal ability of the indicator species. For
example,  the  mean  dispersal   distance  of  young
Alameda  song  sparrows  is  estimated  to be  185
meters, and salt marsh harvest  mice have not been
known to move  distances greater than 30 meters on a
regular basis (Bias  1994).  As such, the average
distance between sites of 350 meters (0.35 km), may
be too  great to allow easy dispersal  between sites.
This is especially true when the  sites are separated by
non-native  habitat which may  act  as  a  complete
barrier to dispersal.

San Francisco Bay Salt Marshes
Predictive Models
  Conservation planners and biologists have often re-
lied upon multiple regression analysis to obtain equa-
tions that  can be used for predictive purposes. For
example, Johnson and Simberloff (1974) used multi-
ple  regression to determine that area,  latitude and
distance were significant predictors of plant species
number in the British Isles. More recently, Soule et
al. (1988) developed a  predictive  model to anticipate
the  fate of chaparral-requiring birds  in fragmented
canyon  habitats. The  model developed from  the
multiple regression  results  of this study can be used
to predict the relative species diversity of individual
salt marsh habitat sites  based upon a series of habitat
characteristics. The  data derived from this model can
be  extremely  valuable as regulatory  agencies  are
often forced to rely upon inadequate  or preliminary
species  census  data;  a   problem that  has  been
identified as one of  the major pitfalls of conservation
planning (Baughman and Murphy  1990).
   The predictive model can also  be used to predict
potential increases in species diversity that would re-
sult from restoration or enhancement, and  to link
these activities to a landscape-based planning proce-
dure. Sites that might be suitable for restoration can
be  re-assessed  for species  diversity  based  upon
changes to characteristics such as size, shape and ac-
cessibility. For example, the restoration of full tidal
inundation to study  site 52 would result in a 57 per-
cent increase in the PSDI for that site, elevating it one
level to the  very high  species diversity category. In
addition, PSDI scores could be used to identify sites
with lower relative values (Figure 2) that  might  be
suitable for development.  The loss  of these sites, in
turn, would be mitigated through a regional approach
that would enlarge or connect existing sites currently
identified as supporting high species diversity. These
connections  would  facilitate  species  movement  be-
tween the populated sites, encouraging recolonization
of sites following local extinctions, and allowing ge-
netic enrichment of existing populations of wetland-
associated species.

  Although predictive models  are somewhat coarse,
they do provide a statistical tool to assess the impacts
of habitat fragmentation  and to evaluate potential
conservation and restoration strategies developed to
ameliorate these effects. Moreover, the advent of sta-
tistical  software packages  make  these models rela-
tively quick and easy to use,  and the necessary data
can  often  be obtained from existing  aerial  photo-
graphs or National  Wetland Inventory  maps.  How-
ever,  some caution must be emphasized when using
these  models as they only provide a relative measure
of species diversity, not a comprehensive evaluation
of wetland performance and function.

  In recent years, state and federal regulatory agencies
have  rejected isolated,  site-by-site mitigation strate-
gies in favor of regional management plans that strive
to protect critical  habitat functions and preserve the
overall biodiversity of wetland ecosystems. However,
the development and implementation  of these plans
has been a difficult process due to a lack  of agree-
ment  on consistent criteria for assessing relative wet-
land  values,   and  identifying  sites  suitable  for
restoration and enhancement (Rumrill and  Cornu
1995). As a result,  many of  the recent plans  have
been  based upon proven  strategies  that emphasize
functional assessments of internal wetland character-
istics as indicators of conservation value. This failure
to recognize the importance of landscape level proc-
esses upon individual sites can result in the failure of
the regional planning  process, and the eventual ex-
tinction of threatened and endangered wetland-associ-
ated species. Future planning strategies must place a
greater emphasis on site location and on habitat char-
acteristics such as size, shape, and proximity in order
to ensure that  restored sites  are  successfully inte-
grated into the  regional wetland mosaic. While sci-
ence  is  still  decades  away from discovering  and
understanding all  of the major components that gov-
ern landscape  level process in wetland ecosystems,
the results of this  study provide a basis for a series of
guidelines for restoring wildlife and habitat values to
the San Francisco Bay Estuary:
  1. Both large and small sites should be preserved.
  Although the results of this  study demonstrate that
size of a site is not strongly correlated with high spe-
cies diversity, large sites should be preserved, as they
generally support  a greater number of individuals of a
particular species. These larger populations  enable
the species to survive random environmental events
(environmental stochasticity), random variations  in
birth  and death rates (demographic stochasticity), and
random changes in genetic composition (genetic sto-
chasticity).  Small sites should also be preserved as
they  seem to play a  greater  role in population dy-
namics than indicated by their size. These sites may
function  to  increase  the  number  of sources  of
potential  colonists  for  sites  that   have  recently

                                                                               San Francisco Bay Salt
undergone local extinction. Also, small wetlands may
act  as   stepping   stones  that  decrease   inter-site
distance, and facilitate the dispersal of individuals
within the  regional  wetland mosaic.  This  would
increase  the probability of site colonization and lead
to a greater proportion of sites having extant popu-
lations at any one time. This, in turn, would increase
the  likelihood of  metapopulation  persistence  over
time (Gibbs 1993).
  2. Long linear sites are better than round ones.
  Long  linear sites were positively correlated  with
high species diversity, possibly because they are more
visible to immigrants, thus promoting the  coloniza-
tion of the site by species from adjacent areas. This is
particularly important in sites that have long aquatic
perimeters, as this characteristic would enable species
such as  California clapper rails, salt marsh harvest
mice, and salt marsh wandering to enter the site by
swimming or  rafting.  In  addition,  sites with  long
aquatic perimeters would  support a greater percen-
tage of low marsh vegetation which  would increase
the amount of potential habitat available for Califor-
nia clapper rails.
  3.  Sites should be surrounded by buffer zones.
  The results  of the study found that the percent of
site  perimeter surrounded by  unnatural vegetation
types, including agricultural and urban lands,  was
negatively correlated with high species diversity. This
is probably  a direct result  of  detrimental  "edge"
effects related to predation by opportunistic mesopre-
dators. While not specifically addressed in this study
the  provision  of buffer  zones  around sites would
ameliorate this problem by filtering out exotic species
such Norway rats, red foxes, and feral cats. In addi-
tion, buffer zones enlarge the effective size of the site
and  provide critical refuge  areas during periods of
extreme high tide.
  4.  Sites should be located close together.
  The results of this study show a slight negative cor-
relation between the isolation of a site and species di-
versity.  As the distance  between  remnant patches
increases, the probability of  successful dispersal de-
creases,  resulting  in a decreased  species diversity in
the  more  isolated sites.  In addition,  endangered
populations are more likely  to be "rescued," by dis-
persing individuals from other sites, if they are in
close proximity to each other (Brown and Kodric-
Brown 1977). This movement between sites can help
prevent  local  extirpation caused  by  demographic
stochasticity,  and  the loss of  genetic  variabilitv
(Wilcove 1987).

  5. Sites should be connected by corridors.

  Interpopulation dispersal, as mentioned above, is
important for regional species persistence. In cases
where the intersite distance is greater than the disper-
sal abilities  of resident  species,  habitat corridors
should be developed between sites. In addition to pro-
viding habitat for plant and animal species, these cor-
ridors can serve as conduits for intersite movement.
  6. Sites should support a full range of salt marsh
vegetation and be subject to full tidal inundation.
  Both cover  type  and tidal regime were correlated
with increased species diversity. All of the indicator
species occurrences, except  the salt harvest  mouse,
were  recorded  in  sites that  supported  coastal  salt
marsh vegetation and were subject to full tidal inun-
dation. In addition, while diked marshes with mixed
vegetation do support populations of salt marsh har-
vest mice,  the species is at a competitive disadvan-
tage  in  these areas  and  is often relegated to  the
poorest habitat patches.
  7. Sites should be channelized.
  The correlation between  channelization and species
diversity was consistent with the known habitat pref-
erences of the  indicator species.  California clapper
rails, California Black Rails, Alameda song sparrows,
and salt marsh wandering shrews preferred sites with
a  moderate to high level  of channelization.  This is
probably because  their primary food sources consist
of aquatic invertebrate species such as ribbed mussels
and yellow shore  crabs. As  salt marsh harvest mice
feed  primarily on  seeds and terrestrial  insects,  this
characteristic seems  to be less critical to their  sur-
vival. However, they have been known to use chan-
nels as dispersal corridors between sites.
  8.  Proposed projects should  be planned  and re-
viewed from a regional perspective.
  The review of projects  on an individual basis ig-
nores the regional distribution of resident species. For
example, the relative  isolation of plants and animals
in San Francisco Bay  salt marsh sites makes  it highly
unlikely that these sites will be recolonized following
local extinctions. Therefore, the replacement of a de-
veloped  marsh with a newly created or enhanced one
in a different location would be of little value to a lo-
calized species that is unable to reach the new  site.
The  recolonization of  sites could be  encouraged,
rather than  inhibited,  if compensatory  replacement

 San Francisco Bay Salt Marshes
 mitigation policies were changed to emphasize re-
 gional rather than site specific goals. For example,
 wetland  development could be mitigated through an
 off-site banking approach that enlarges or connects
 existing  sites that currently support high species di-
 versity.  These  connections  would facilitate species
 movement between the populated sites and increase
 the overall viability of the  regional populations. The
 creation  or restoration of upland  buffer zones could
 also be used to mitigate project development.

 Literature  Cited
 Baughman, J. R, and D. D. Murphy. 1993. Beware of
   snapshots at the bottleneck—temporal  considera-
   tions in conservation planning. Endangered Species
   Update, 7, 8, and 9.
 Bias, M. A. 1994. Ecology of the salt marsh harvest
   mouse in San Pablo Bay.  Doctoral Dissertation,
   University of California Berkeley, Berkeley, CA.
 Brown, J. H., and A.  Kodric-Brown. 1977. Turnover
   rates in insular biogeography: effect of immigration
   on extinction. Ecology 58:445-449.
 Diamond, J. M., and R. M. May. 1976. Island bio-
   geography and  the design  of  nature reserves. In
   R.M.  May (ed.), Theoretical Ecology: Principles
   and Applications,  pp.  163-186. W.B.  Saunders,
   Philadelphia, PA.
 Geissel,  W., H. Shellhammer, and  H.  T. Harvey.
   1988.  Ecology of the salt marsh harvest mouse in a
   diked  salt marsh. Journal of Mammalogy 69:696-
 Gibbs, J. P., J. R. Longcore, D. G. McAuley,  and
   J. K. Ringelman. 1991. Use  of wetland habitats by
   selected nongame waterbirds in Maine. U.S. Fish
   and Wildlife Service, Fish and Wildlife Research 9.
 Gibbs, J. P. 1993. Importance of small wetlands for
   the persistence of local populations of wetland-as-
   sociated animals. Wetlands 13:25-31.
 Grover, A. M.,  and G. A. Baldassare. 1995. Bird spe-
   cies  richness within beaver  ponds  in south-central
   New York. Wetlands 15:108-118.

 Johnson, M. P., and D. S. Simberloff.  1974. Environ-
   mental determinants of island species numbers in
  the British Isles. J. Biogeography 1:149-154.

Johnson, N. K.  1975. Controls  of number of bird spe-
  cies on montane islands in the Great Basin. Evolu-
  tion 29:545-567.
MacArthur, R. H., and E. O. Wilson. 1967. The the-
  ory of island biogeography. Princeton University
  Press, Princeton, NJ.
Nichols, F. H., J.  E. Cloern, S. N. Luoma, and D. H.
  Peterson. 1986. The modification of an estuary.
  Science 231:567-573.
Noss, R. F.  1983. A regional landscape approach to
  maintain biodiversity. Bioscience 33:700-706.
Postemak, L. D. 1986. A general study of birds using
  three wetland areas of shoreline at Mountain View.
  City of Mountain View, Mountain View, CA.
Power, D. M. 1972. Number of bird species  on the
  California islands. Evolution 26:451-463.
Preston, F. W. 1962. The  canonical distribution of
  commonness and rarity. Ecology 6:88-97.
Rumrill, S. S., and C. E. Cornu.  1995. South Slough
  coastal  watershed  restoration.  Restoration  and
  Management Notes 13:53-57.
Saunders, D. A.,  R. J. Hobbs, and C.  R. Margules.
  1991. Biological consequences of ecosystem frag-
  mentation: a review. Conservation Biology 5:18-28.
Shafer, C. L.  1990. Nature reserves: island theory and
  conservation  practice.   Smithsonian  Institution
  Press, Washington, DC.
Shellhammer, H.,  R. Jackson, W. Davilla, A. M. Gil-
  roy, H. T.  Harvey, and L. Simons. 1982. Habitat
  preferences of  salt marsh harvest mice. Wasmann
  Journal of Biology 40:102-114.
Shellhammer, H., R. Duke, H.  T. Harvey, V.  Jen-
  nings, V. Johnson, and M. Newcomer. 1988. Salt
  marsh harvest  mice  in the diked salt marshes of
  southern San Francisco Bay. Wasmann Journal of
  Biology 46:89-103.
Simberloff, D. S., and L. G. Abele. 1982. Refuge de-
  sign and island biogeographic  theory: effects of
  fragmentation. American  Naturalist 120: 41-50.
Soule, M. E., D. T. Bolger, A. C. Roberts, R. Sauva-
  jot, J. Wright, M. Sorice, and S.  Hill. 1988.  Recon-
  structed dynamics of chaparral-requiring birds in
  urban  habitat   islands.   Conservation  Biology
Soule, M. E., and B. A. Wilcox. 1980. Conservation
  biology:  an  ecological-evolutionary  perspective.
  Sunderland Associates, Sunderland, MA.
Temple, Stanley  A., and Bruce A. Wilcox. 1991. In-
  troduction: predicting effects of habitat patchiness

                                                                           San Francisco Bay Salt Marshes
  and fragmentation, pp. 261-262 In D. A. Saunders,
  R. J. Hobbs, and P. R. Ehrlich (Eds.), Nature Con-
  servation 2: The Role of Corridors. Surrey, Beatty
  & Sons, NSW, Australia.
Terborgh, J. 1976. Island biogeography and conser-
  vation: strategy and limitations. Science 193:1029-
U.S. Fish and Wildlife Service (USFWS). 1993. Un-
  published Sensitive Species distribution data.
U.S. Fish and Wildlife Service (USFWS). 1994. Un-
  published Sensitive Species distribution data.
Vickery, P.  D., M. L. Hunter, and S. M.  Melvin.
  1994.  Effects of habitat area on the distribution of
  grassland  birds in Maine.  Conservation  Biology
Walton, B. J. 1975. The status of the salt marsh song
  sparrows of the San Francisco Bay system,  1974-
  1975. M.S. Thesis, San Jose State University, San
  Jose, CA.
Western Ecological Services  Company (WESCO).
  1986. A Review of the Population Status of the Salt
  Marsh Wandering Shrew. U.S. Fish and Wildlife
  Service, Sacramento, CA.
Western, D. and J. Semakula. 1981. The future of the
  savannah ecosystems: ecological islands of faunal
  enclaves? Aft. J. Ecology 19:7-19.

Wetlands Research Associates (WRA). 1994. Draft
  Environmental Assessment:  Cargil  Salt  Mainte-
  nance  Activities,  Permit Application  No.  4-93,
  SCH# 94023030. San Francisco Bay Conservation
  and Development Commission, San Francisco, CA.
Wilcove, D. S. 1987. From fragmentation to extinc-
  tion. Natural Areas Journal 7:23-29.
Wilson, E. O., and E. O. Willis.  1975. Applied bio-
  geography,  pp. 522-534 In M. L. Cody and J. M.
  Diamond (eds.), Ecology and Evolution of Commu-
  nities. Belknap Press of Harvard University, Cam-
  bridge, MA.
Zetterquist, D. K. 1977. The salt marsh harvest mouse
  in marginal habitats. Wasmann Journal of Biology

                             Habitat  Restoration in an  Urban Setting:
                                     Integrating Desert Riparian Habitat,
                                                Groundwater  Recharge,  and
                                                     Community  Participation
                                                                             S. C. Brown1
                                                                         J. J. Donaldson
                                                                                N. Kroska
                                                                                  D. Cotter
  Abstract: Desert riparian habitat has been scarce
and of great importance to wildlife in the south-
west United States. Because the amount and qual-
ity of this habitat type has been greatly reduced
from historic levels,  restoring desert  riparian
habitat is a high priority in Arizona. This presen-
tation describes a plan that is being implemented
by the  Town of Gilbert, Arizona,  to  establish
riparian habitat and  provide other  wildlife and
public use benefits at  the town's 75-acre ground-
water  recharge pond complex.  The facility is
located  in an urban area and uses treated  water
from the town's wastewater reclamation plant to
help recharge the local groundwater. The project
is funded primarily through two Arizona Depart-
ment of Game and Fish Heritage Grants and has
already won  two state environmental and plan-
ning awards. The project has  been largely imple-
mented by volunteers  from the local community.
Design features  of the project include establish-
ment of riparian vegetation along pond margins,
marsh  vegetation  in  a  permanent pond, and
upland vegetation on pond slopes; wildlife nesting
structures; interpretive exhibits; and blinds for
viewing wildlife. Jones & Stokes Associates, with
assistance from  Wild Seed of Tempe,  Arizona,
assisted  the  town with  preparing the  grant
applications, developing the plan and details for
implementation, collecting and installing cuttings
and seeds, and  implementing the project  using
community volunteers. The project successfully
integrates enhancement of wildlife and public use
values  with  the operation  of a groundwater
recharge facility and could serve as a model for
other similar urban wildlife  habitat restoration
  "We need nature as much in  the city as in the
country side....It is not a choice of either the city or
the countryside:  both are essential, but today it is
nature, beleaguered in the country,  too scarce in the
city which has become precious. (McHarg 1969)"
  Riparian habitat in Arizona is a scarce and precious
resource that is extremely important to wildlife, as
well as to people. Particularly in urban areas, riparian
habitat is scarce. What does exist often is  of low
value for wildlife. The Town of Gilbert, Arizona, has
realized  a way to enhance or create riparian habitat
within its rapidly developing urban  boundaries while
simultaneously  improving the community  and the
environmental education opportunities for people.
  The town developed its wastewater recharge ponds
to provide the following benefits for wildlife, people,
and the community:

  •   Open water and riparian vegetation that pro-
     vides resting and foraging habitat for wildlife,

  •   Opportunities for people to closely view diverse
     wildlife in an urban setting, and

  •   Groundwater recharge and water planning  to
     ensure available water resources for a growing
  Using  a  readily available resource,  reclaimed
wastewater, the  town integrated these benefits into
construction and operation of groundwater recharge
 Jones & Stokes Associates, 2600 V Street, Suite 100, Sacramento, California 95818-1914.


                                                                                  Desert Riparian Habitat
Groundwater Conservation Efforts and
 The town's  growth  and location  in  the northern
Sonora Desert  had previously challenged decision
makers to develop plans for efficient use of ground-
water and renewable water resources. The answer to
this challenge was to  make better use of reclaimed
water, and the town's  leaders developed a policy of
100 percent reuse of water. In addition, under  state
legislation  (1980  Groundwater  Management   Act
[ARS 45-564.A.2]), the town  measures and reports
the volume of  water  recharged  and recovered  and
receives credits  for reclaimed water in the area where
the effects of recharge can be detected and measured.
To ensure the groundwater credits, the town has con-
structed 12 recharge ponds  on 75 acres and a  deep
well  that recovers water for potable use. Water
reclamation has helped the  town achieve  the state's
groundwater  management  goals  for   individual
municipalities and provided a  safe resource for cur-
rent and future water demand (Frost et al. 1993).
  An added benefit of water use planning has been the
enhancement of urban habitat at the recharge ponds.
After the original six  ponds were constructed on 38
acres in  1989,  wildlife began  using the open water
habitat for resting and  foraging. A benefit to the town
was the resulting wildlife viewing opportunities at the
recharge ponds. The Sun Circle Trail along the north
edge  of the pond  complex  provided a viewing site
used  by  local  Audubon  Society members.   The
viewing area  soon attracted many other birdwatchers.
The   absence  of  vegetation   around  the  ponds,
however, limited their habitat value and the diversity
of species that  could  use the pond area, which  was
expanded to six more ponds on 37 acres in 1993.

Plans for Site Improvements
  In 1993, the town took an important step to improve
the resource  values of the pond area  by seeking a
Heritage Fund grant from the Arizona Game and Fish
Department.  Jones  &  Stokes   Associates (JSA)
assisted  the  town  in  identifying opportunities to
enhance the habitat value of the ponds and provide
interpretive   facilities  for  the  community.  JSA
prepared the  grant application, which proposed that
the town and JSA work as a team to design  and
construct numerous wildlife habitat and public use
improvements. JSA would design the  urban wildlife
enhancement  plan and interpretive features and assist
the   town    in   supervising    volunteers   during
construction. The town would coordinate volunteers
for facilities construction and supervise overall plan
implementation.  After award of the initial grant in
1993, JSA, with assistance from Wild Seed, Inc., a
local native seed and revegetation consultant, began
refining  the  design concepts  for wildlife habitat
enhancement and opportunities for public use  and
environmental education at the new ponds.

Design Concepts
  The  town  was awarded two grants  for habitat
improvements. The first,  awarded in  1993 for the
west half of the pond complex, included  establishing
riparian, upland, and marsh vegetation, and designing
and installing bird nest and bat roost boxes, creating
viewing blinds,  designing interpretive exhibits,  and
constructing a viewing ramada. The second, awarded
in 1994,  provided funding for expanding the habitat
and public use improvements to the six original ponds
on  the east half of  the pond complex by planting a
greater diversity of riparian  and upland  species,
planting  an interpretive garden around the  viewing
ramada,  and developing two additional  interpretive
Maximizing Resource Opportunities
  As part of the design concept  planning, existing
resources were assessed and utilized. The  town had
three valuable  resources:  reclaimed water, existing
wildlife, and a central location.
  The  primary  resource  was  the availability  of
reclaimed water, which  provided an excellent oppor-
tunity  to establish riparian  vegetation  around the
recharge ponds. Although the water source is  man-
aged, it  is a valuable resource in the Desert South-
west. Riparian vegetation occupies less than 1 percent
of the  Arizona landscape (Brown  1982); however, it
supports the densest and most diverse wildlife  com-
munities in the state. Its diverse plant species,  vege-
tative  growth  forms, and microclimate conditions
(e.g., increased humidity and lower temperature) pro-
vide a variety of important foods  and habitat condi-
tions for wildlife (Warner 1979). Riparian vegetation
provides  nesting and foraging habitat  for resident
wildlife  and important refuge for migratory species.
Approximately  75 percent  of Arizona's  wildlife,
including  fish,  amphibians, insects,  birds,   and
mammals, depends on a riparian area for some phase
of its life cycle  (Tellman 1992).

Desert Riparian Habitat
  The project site is on the Mohall-Contine soil asso-
ciation, which consists of well-drained, deep soils on
old  alluvian fans. The  area  lies between the flood-
plains of the Salt River and  the Queen Creek Wash.
Because  the permeability of the  soil is  slow, the
ponds are periodically disked to improve permeabil-
ity.  Historic vegetation was probably  creosotebush,
saltbush, mesquite, cactus, and annual grasses. The
historic  land use  was predominantly  irrigated  agri-
culture using an expansive system of canals originally
developed  by  the Hohokam people  native to the
region. Crops planted more  recently include alfalfa,
cotton, vegetables, and citrus.
  Initially the ponds were relatively barren of vegeta-
tion except for a few weedy grass  species. Nonethe-
less, the  ponds provide  water,  a  critical habitat
component that attracts numerous wildlife  species.
Maricopa Audubon  Society  members  have  observed
 140 bird species  at  the  ponds  and on agricultural
lands surrounding the project site. The ponds provide
resting opportunities and  limited foraging opportuni-
ties for various  species; because of a lack of vegeta-
tion, however, the site does not have the cover and
shelter habitat that wildlife need to  occupy the site on
a regular basis.
  The  recharge pond complex  is completely  sur-
rounded  by a chain-link  fence, and public access
within the fenced area is prohibited. The public does
have visual/pedestrian access to the ponds along the
north  boundary via Maricopa  County's  Sun Circle
Trail,  a  more than 200-mile, regional multipurpose
trail. The  pond complex is centrally located  and
accessible to town residents and Gilbert schools.
 Design Elements and Construction
  In planning the new ponds, the Town of  Gilbert
 recognized an opportunity to incorporate features that
 would improve their value for both wildlife and peo-
 ple. The goals of the project were to enhance wildlife
 habitat and  provide environmental education and
 wildlife  viewing  opportunities for town residents.
 These goals could be achieved by establishing diverse
 vegetation types on the project site to ensure the fol-
 lowing results:

  •   Encourage and maximize wildlife use,

  •   Encourage wildlife  occupancy of the  site  by
      providing structural habitat, and
  •   Accommodate  public use  and enjoyment by
     providing additional viewing  and interpretive

  As a result  of the Heritage grant, more extensive
design concepts were developed to enhance the habi-
tat value of the ponds and manage them for wildlife,
as well as for recharging groundwater. Features of the
new ponds (west  half)  that  address these  goals
include curvilinear  embankments  to  improve  aes-
thetics and create more  edge  habitat,  a  permanent
pond, and an elevated viewing area for the public.

West Pond Habitat Enhancements
  Three vegetation types were established at the west
ponds,  including  cottonwood   and willow  thicket
habitat, emergent  marsh  habitat, and desert herba-
ceous and scrub habitat. Establishing this vegetation
took advantage of the existing resources on the site
(water, soil, wildlife) and  met the goals of the project
(increased habitat, improved education opportunities,
and  community  involvement).  Establishing  this
vegetation also met secondary  considerations about
the availability of material, cost, and appropriateness
for wildlife species already using the site or expected
to use the site.
  The cottonwood and  willow  thicket habitat estab-
lished  at  the  west ponds included  Fremont cotton-
wood, Goodding's willow,  seep willow, and coyote
willow. Cuttings of these species had been collected
from the area by volunteers. The cuttings were placed
in wet cold storage  for several weeks until the next
volunteer activity planting.

  The cuttings were  placed at  the  toe of the pond
embankment  slopes and  arranged  in clusters  to
maximize  shade, density, canopy  structure, habitat
value,  edge effect, and diversity. The planting clus-
ters provided 25 percent coverage of the total area of
the inside  slopes of  the ponds; this percent coverage
was intended to provide essential habitat components
and open corridors for wildlife viewing. The clusters
were  located  to minimize interference with mainte-
nance  activities requiring vehicular access, and the
gaps between  planting clusters  were designed to meet
the needs of ducks such as pintails, northern shovel-
ers, green-winged teals,  and  various  diving  ducks
(ruddy  ducks) that  use  a long, low  trajectory  for
landing and takeoffs. Plantings were increased along
the north boundary  to  minimize  disturbance  of
wildlife by partially screening from view the people
on the Sun Circle Trail.

                                                                                    Desert Riparian Habitat
 The emergent marsh habitat was planted in a small,
permanent pond in one comer of a larger pond. Marsh
species, including sedge, spike rush, rush, tule, and
arrow weed, were collected and planted as tubers by
volunteers. The  emergent marsh  habitat provides
nesting  and foraging  habitat for waterfowl,  grebes,
rails, and wetland songbirds.
 Desert  upland  herbaceous  and scrub habitat was
established along the  top of  the pond embankments
where  the only  available  water source  is rainfall.
Volunteers broadcast  seed, with direction from Wild
Seed, Inc., along the  top third of the embankments
and gently raked the seed into the top 1/4-inch of the
soil. Wood chip mulch was donated by a local tree
service, and "a light coverage was  distributed  along
the seeded banks by volunteers. The species  compo-
sition included a diverse mixture of native grasses,
forbs, and shrubs to provide wildlife cover and food
and  to  control  erosion.  Unfortunately  the  pond
embankments  were compacted to 95 percent during
pond construction and only the most tenacious desert
seeds germinated, predominantly saltbush and  a few
desert sunflower.

East Pond Habitat Enhancements
  The same planting  concept was used for the east
ponds  during  the  second Heritage grant with  the
addition of two  desert  riparian plant associations:
Interior Southwest  Riparian  Deciduous Forest and
Woodland and  Sonora  Riparian  and Oasis  Forest
Mesquite Series. These associations are identified as
sensitive by the  Arizona Department of Game and
Fish and were included to create more diverse wild-
life habitat and  vegetative species composition and
expand the educational opportunities of the site. Both
vegetation groups were planted midslope around the
ponds from 1-gallon containers and are provided with
drip irrigation until they become established. Volun-
teers planted  the  vegetation  in holes  previously
drilled by the town's staff. The Interior Southwest
Riparian Deciduous Forest and Woodland is  domi-
nated by Arizona sycamore, but also includes desert
hackberry, wolfberry,  mesquite, Mexican elderberry,
graythorn, desert willow, and hackberry. The Sonora
Riparian and Oasis Forest  is  mesquite-dominant and
includes desert hackberry,  wolfberry, and graythorn.
(Brown 1982).
Wildlife and Public Use Structures

 Food, cover,  and shelter  are  important  habitat
requirements for wildlife. The diversity of riparian
and upland  vegetation  will help meet these require-
ments by providing fruit and seed for forage, protec-
tion from predators, protection from adverse weather
conditions, and  habitat for breeding. The establish-
ment of vegetation is one of the habitat enhancement
goals for the project; however, it will be years before
the vegetation is mature enough to provide tree cavi-
ties and high branches for adequate cover and shelter
for certain species of wildlife. In the  interim, human-
made structures  were designed to provide additional
wildlife cover and shelter habitat.  Nest  and roost
boxes for American kestrels, wood ducks, black-bel-
lied whistling ducks, tree swallows, black phoebes,
and bats were designed and installed to supplement
the riparian vegetation.  The  structures were  con-
structed by the Maricopa Audubon Society volunteers
and placed  on poles around the ponds  by  the town
staff. It is possible that even the  boxes may not be
utilized until surrounding vegetation provides suitable
environmental conditions to attract wildlife.
 A number of public  use improvements  were also
constructed by volunteers. Boy Scouts constructed
benches at the viewing blinds along the Sun Circle
Trail, and  a local architect donated  designs for the
ramada.  The  town also constructed   an  elevated
observation area and access ramp up to the ramada on
the southwest corner of the project site. The viewing
ramada is the main public use area; other facilities
include public  parking,  benches,  and a  screened
viewing area.  Seven interpretive exhibits created by
ISA  explain  the  wildlife,  riparian  habitat,  and
groundwater recharge system at the ponds.

Town Commitment and Citizen
 The Town of Gilbert was responsible for publiciz-
ing the planting days  and soliciting volunteers and
organizations  to participate.  For the  first  planting
effort, more than 50 volunteers donated their time on
a Saturday to plant, dig, or spread seed. They came
from the Audubon Society, the Rotary Club, the Boy
Scouts, the local high school ecology  club, and the
general  public.  The   consultants  organized  and
instructed the volunteers about the  day's  activities.
For the subsequent activities, the Audubon Society
constructed roost  and  nest boxes  and  Eagle Scouts

 Desert Riparian Habitat
 constructed benches, dug planting holes, and planted
 seeds for the interpretive gardens.
  The  future monitoring of the ponds  will be con-
 ducted by  a local junior college biology department
 in cooperation with the high school and elementary
 schools. The monitoring activities will be used to
 teach  students of all levels about soil, water, vegeta-
 tion, and wildlife.

  As the vegetation around  the ponds matures, its
 value  to wildlife will increase and expand the variety
 of wildlife species attracted to the ponds. It will also
 offer  residents and students visiting the ponds on a
 regular basis a unique opportunity to observe changes
 in plant growth and habitat value through time.
  The  ponds provide valuable  wildlife habitat in a
 desert environment, which will  continue to increase
 in value as the vegetation matures and as long as the
 ponds are maintained. As of fall 1995,  11 species of
 waterfowl  were routinely identified at the ponds,
 including  black-bellied  whistling duck, great blue
 heron, northern shoveler, ruddy duck, and  cinnamon
 teal, and a marsh  wren  has moved into the emergent
 marsh  habitat.  The ponds continue to provide the
 Town of Gilbert with economic value in the form of
 water  credits for groundwater  recharge.  They  also
 provide the town with a unique civic and community
 resource, providing passive recreation  opportunities
 for  the community; educational opportunities for
 local  students  and  adults;  and, perhaps  most im-
 portantly, a civic  project people of all ages can be
 actively involved  in  through   volunteer  plantings,
 monitoring, and maintenance.

  The  project received  public  recognition on  two
 occasions  during  1994. It received the Governor's
 Pride in Arizona Award in the Environmental Lead-
 ership category  and the  Best Project Award from the
Arizona Planning Association, recognizing the ponds
as a  site-specific project using  original  ideas  to
address local, regional, or national concerns.

  The project successfully integrates enhancement of
wildlife and public use values with the operation of a
groundwater recharge facility in an urban setting and
could serve  as a model for other,  similar urban wild-
life habitat  projects. Uniting economic values with
wildlife habitat and community values ensures human
and natural benefits.

Literature Cited
Brown, D. E. (ed.). 1982. Biotic communities of the
  American  southwest -  United  States and Mexico.
  (Desert Plants, Volume 4, No. 1-4.) The University
  of Arizona.  Prepared  for  the Boyce Thompson
  Southwestern Arboretum, Superior, AZ.

Frost, L. K., N. Mailman, R. F. Buss, R. D. Johnson,
  K. D. Schmidt, and D. Cotter. 1993.  The Town of
  Gilbert's operational experience with recharge and
  recovery of reclaimed water.  Proceedings  of  the
  29th  Annual  Conference   on  Innovations   in
  Groundwater  Management  and Symposium  on
  Effluent  Management,  August 29-September  2,
  1993, American Water  Resources  Association.
  Tucson, AZ.

McHarg, I. L. 1969. Design with nature. Doubleday
  & Company, Inc. NY.
Tellman,  B.  1992.  Arizona's  effluent dominated
  riparian areas:  issues  and opportunities.  Water
  Resources Research Center. University of Arizona.
  Tuscon, AZ.

Warner, R.  E.  1979. California  riparian study pro-
  gram: background information and proposed study
  design. Department of Fish and Game. Sacramento,

Taking A Closer

"(Wetland) biology is not simple
and you cannot make it simple.
Every time you try to make it
simple you're probably doing
something very wrong and
seriously misleading yourself/'
                     Elizabeth Copper
                     San Diego-1990

                             Aquatic  Habitat Restoration in the United
                                  States:  A Review of Design and Costs
                                                                          Ronald M. Thorn1
                                                                       Katherine Wellman2
                                                                         David K. Shreffler1
                                                                           Michael J. Scott3
  Abstract: The objective of this study was to com-
pile  and  summarize  non-Corps of  Engineers
aquatic habitat  restoration projects  throughout
the United States for the purpose of assisting the
Corps  plan and develop accurate cost estimates
for future projects. A subset of 91 projects were
evaluated from an initial list of well over 200 proj-
ects  identified from a literature search. Project
descriptions and costing were fully developed for
39 of the 91  projects. The projects ranged from
coastal ecosystems to palustrine systems, and were
classified into 16 different types (e.g., wetland con-
struction, bottomland hardwood restoration). We
found that most were largely done on an individ-
ual basis with little impetus to follow set methods
or guidelines for establishing goals, performance
criteria, monitoring, or documentation. Documen-
tation of all aspects of restoration projects is, in
particular given  low  priority.  Costing,  was not
documented at all in some cases or was not acces-
sible even after considerable effort. We recom-
mend that all aspects of a restoration project
should  be  completely  and consistently  docu-
mented. Planning of future projects requires this
type of information to help minimize costs and
maximize the probability of success.

  The U.S. Corps of Engineers is developing guidance
documents to  assist Corps planners in the develop-
ment and implementation  of environmental restora-
tion projects. In the coming decades, the Corps will
be involved with or leading  major  environmental
restoration efforts. The objective of our study was to
compile and compare management measures, engi-
neering features, monitoring techniques, and detailed
costs for a  representative sample of non-Corps of
Engineers environmental restoration projects. Analy-
sis and conclusions reported here are from a project
report entitled: National Review of Non-Corps Envi-
ronmental Restoration Projects  for the Institute for
Water Resources (Shreffler et al. in press). The pres-
ent paper deals primarily with the issue of costing and
cost reporting.

 In order to identify appropriate environmental resto-
ration projects for our review, we conducted a litera-
ture  search of recent  conference proceedings and
contacted colleagues within the restoration  commu-
nity. Over 200 non-Corps projects were identified. Of
these a total of 91 were pursued. We conducted a set
of interviews  over  the phone with experts  or  indi-
viduals  with  specific  knowledge  about particular
restoration projects, followed by a written question-

 Information  from the questionnaire  and supple-
mental project documentation were used to develop
2- to 3-page summaries for each project. Of the 91
projects pursued, 39 were complete enough to be in-
cluded in final analysis.

Results and Discussion

Project Descriptions
 The projects  identified in this study were catego-
rized into 16 types, which reflect the primary features
of these projects. The classification generally follows
Mitsch and Gosselink (1995) for system types and the
National Research Council  (NRC, 1992) for defini-
tions of restoration, enhancement, and creation. The
16 types of projects included:

  •   Bottomland hardwood forest restoration
1 Battelle Marine Sciences Laboratory, 1529 W. Sequim Bay Rd., Sequim, Washington 98382.
2 Battelle Seattle Research Center, 4000 ME 41st St., Seattle, Washington 98105.
3 Battelle Pacific Northwest National Laboratory, P.O. Box 999, K803, Richland, Washington 99352.

Design and Costs
  •   Enhancement of fish and wildlife habitat

  •   Estuarine wetland restoration

  •   Estuarine  wetland  restoration  and  wildlife

  •   Mitigation banks

  •   Stream enhancement

  •   Stream restoration

  •   Water quality remediation

  •   Wetland creation
  •   Wetland restoration and enhancement
  The primary  goals of the various projects included
 reestablishing  historical vegetation, restoring or en-
 hancing habitat for wildlife and fish species,  stabi-
 lizing  shorelines,  mosquito  control, treatment  of
 wastewater, and restoration of hydrology. Many proj-
 ects were conducted as mitigation designed to  offset
 impacts from another project.
  The engineering features cover a wide  range  of ac-
 tivities. For example, the Natural Resources Conser-
 vation  Services (NRCS)  Wetland Reserve Program
 has targeted agricultural  land that can be relatively
 easily returned to wetlands. These projects typically
 involved  little physical  work but include  activities
 such as removal of impediments to natural hydrologi-
 cal processes. Other projects such as the North Fraser
 Harbor Habitat Compensation Bank involved a wide
 variety of actions such as excavation, rip rap shore-
 line stabilization, and placement of sod.
  Monitoring techniques ranged widely from simple
 observations and photographs taken annually to very
 complex and integrated sampling and modeling stud-
 ies.  We found no project that exclusively relied  on
 established procedures such as the HEP or the WET,
 although  some projects employed these techniques
 during part of the assessment or planning phase. This
 result indicated that although standardized techniques
 exist, they have not been widely applied.

  All projects with monitoring information covering
 more than one year indicated some level of success
 relative to goals and criteria for the projects. Because
 monitoring varied  so widely among projects, the suc-
 cess  measures varied widely as  well.  Hydrology,
plant growth and cover, and bird use were most often
cited as being clear indicators of the performance of
the system.
Cost Analysis
 A number of previous attempts have been made to
analyze restoration costs. In these  studies cost esti-
mates are assessed on a total per acre basis and re-
ported in constant cost dollars. King and Bohlen
(1994) reported estimates of average wetland restora-
tion costs (excluding land costs) derived from pri-
mary data on  90 wetland restoration projects. Costs
gathered for their study  ranged from $5 per acre to
$1.5 million per acre. Guinon (1989) presents results
of a cost survey of 25 wetland restoration projects
throughout California and illustrates a wide variation
in costs ($1,626 to $240,000 per acre). NOAA (1992)
summarized some costs  of typical  wetland creation
projects. The  costs of restoration projects presented
in this study range from $485  to over $70,000 per
 U.S.  Department of Interior (DOI 1991) presented
the results of a comprehensive review of the available
literature  on wetland  creation  and restoration, and
wetland mitigation plans. The range of overall resto-
ration  costs, (updated to 1989 dollars), is approxi-
mately $2,000 to $50,000 per acre although one study
showed costs  of $220,000 per acre. The DOI report
suggests that  a  portion  of this variation may  be
explained by  the fact that there is no "typical"  or
standard restoration project.
 Our study attempted to overcome  some  of the past
limitations of restoration cost estimation efforts. We
attempted to derive estimates of the cost of equip-
ment, labor, materials, and supplies for each compo-
nent of a  project from a representative sample of
39 wetland and habitat restoration projects across the
 Individual project contacts were asked to respond to
the following questions (within the context of the
overall survey):
1.  What was the price of the engineering design?
2.  What was the cost of monitoring the program?
3.  What physical  structures  were  built?  How  many
    of each? What was the cost? Year costed?
4.  What were the labor costs associated with each
    major aspect of the project?
5.  Are there maintenance requirements for physical
    conditions or structures? What is the cost  of
6.  Who paid for this program?

  One  of the  biggest problems that we encountered
was finding the right person with  whom to discuss

project costs. In many cases the institutional knowl-
edge about a project no longer existed or  files  in-
cluding cost information  had been  misplaced.  In
many cases  project costs were reported only as the
initial project budget with  no detailed accounting of
actual  expenditures  including, for example, change
orders, delays,  salary and wage  increases  and  etc.
(Table 1). In other cases "reported" cost figures often
did not include costs for land acquisition, permitting,
planning,  design,  maintenance and monitoring.  We
attempted, through follow-up phone calls, to  elicit in-
formation on the true costs of selected projects, those
costs mentioned above as well as the opportunity cost
of individual's time and donated or contributed labor
and equipment.
  While  many  project  contacts   eagerly  reported
"costs" it appears that they were  unaware of where
these figures came from, how they were assessed, or
the implications of inaccurate and unreported hidden
restoration costs  such as project management, over-
head and volunteer labor.  As illustrated in  Table 2,
most costs were reported as  "installed" or lump sum
(Table 2). As a result separate labor, equipment and
material  costs  for a project component were  not
available. In other cases, the  lack of unit descriptions
made estimation  of  costs per unit impossible. As a
result, our final total costs  estimates are probably, on
balance, underestimates of the true costs of  the vari-
ous projects analyzed.
  Because the elements associated with the restora-
tion projects analyzed in this study vary across proj-
ects, and costs  are allocated  in different ways across
our entire sample, it  is impossible  to make any statis-
tically significant comparisons of the costs of specific
                                   Design and Costs
components across projects. However, several com-
ponents appeared more than once in our sample for
which total  and per  unit costs  were reported  as

  •   Gravel removal activity costs range from $3.27
      to $3,239 per ton.

  •   Rip rap  installation costs range from $5.00 to
      $19.00 per ton.

  •   Culvert installation costs range from $150 (for
      48" diameter culvert) to $1,103.85 per ft.

  •   Channel  cleaning costs  range from  $4.00 to
      $8.00 per cu.m.

  •   Erosion control costs range from $1.40 to $4.00
      per sq. ft.

  •   Dike removal costs range from $1.92 to $2.67
      per ft.

  •   Dike/dam/levees  construction costs range from
      $5.00 to $20.00 per linear ft.

  From our work we suggest that there are two pri-
mary variables determining restoration costs: (1) the
specific  project components required to restore the
ecosystem, from conceptual design to monitoring;
and  (2) how the restoration costs are  allocated and
  Costs  of  implementing restoration  projects  are
unique to each project  and  are  significantly influ-
enced by: site access;  grading; site preparation re-
quirements;    difficulty   of   plant    community
establishment;  schedule delays; and, complexity of
 Table 1. Cost by Component for the Barataria-Terrebonne National Estuary

Component Quantity
Convert oil 3
canals to marsh



Supplies, etc.






Total Costs



Total Costs



Design and Costs
Table 2. Cost by Component for Palo Alto Harbor marsh construction.

Project Component
Demolition (includes
launch ramp)
Clearing and grubbing
Stock pile embankment
2-inch decomposed
granite on 4-foot
aggregate base
2-inch AC on 6-foot
aggregate base
Asphalt concrete berm
Retaining wall and
Retaining wall
Detail striping
12-inch solid line
Pavement marking
Directional sign
Park bench (double)
Park bench (single)
Bike rack bollard
Trash receptacle
Irrigation system
Topsoil, plain
Topsoil, mix
Weed control rings
Buffer plants
Marsh plants (cord
Marsh plants
















Supplies, etc.






Labor ($1994)
N/A $11,920






Total Costs








                                                                                       Design and Costs
habitat management programs.  However, other fac-
tors affecting final restoration costs include:  econo-
mies of scale; type of restoration; restoration design;
restoration site quality; adjacent site quality; appro-
priate technology; simultaneous construction multiple
use; and, project management.
  When reporting, analyzing, and comparing restora-
tion costs, it is essential to consider the project scope
and identify all project elements included or excluded
from reported figures. The scope of this study and in-
adequate nature of contact responses did not allow us,
however, to adequately inquire about all significant
components of each project on a per unit basis, assess
costs in context, or specify all those project elements
not reported by the project contact.

  We found that most non-Corps restoration  projects
are largely done on an individual basis with little im-
petus to follow set methods or guidelines for estab-
lishing goals, performance  criteria,  monitoring,  or
documentation. Our review suggests that documenta-
tion  of all aspects of restoration projects is,  in par-
ticular given  low  priority.  Costing,  as indicated
earlier was not documented at all is some cases  or
was  not  accessible even after considerable effort.
This hinders the ability to develop the technologies
for restoration, which in turn limits the predictability
of actions undertaken during restoration projects.
  As part  of any  restoration project  we recommend
that  all aspects of a restoration project should  be
completely and consistently documented especially if
environmental restoration becomes a national focus
for federal agencies. Accurate and consistent record
keeping is useful for documenting the effects of deci-
sions and to show progress towards goals. In addition,
planning  of future projects require this  type of infor-
mation to help minimize costs  and maximize the
probability for success.

  This study was conducted as part of the Evaluation
of Environmental Investments Research  Program
sponsored by the Institute for Water Resources (IWR)
of the U.S. Army Corps of Engineers. We gratefully
acknowledge the support  of Joy  Muncy the  study
manager at IWR. The cost and project information for
Barataria-Terrabonne and  Palo Alto projects were
kindly provided by Mamie Winter and Jim Harring-
ton, respectively.

Literature Cited
Guinon. 1989. Project elements determining compre-
  hensive  restoration costs and repercussions of hid-
  den  and  inaccurate   costs.  Paper  presented  to
  Society  for  Ecological  Restoration and  Manage-
  ment, 1989 Annual Meeting, Oakland, CA.

King, D. M. and Bohlen.  1994. Estimating costs of
  wetland restoration. National Wetlands Newsletter

Mitsch, W. J.  And J. G. Gosselink.  1995. Wetlands.
  Second Edition. Van Nostrand Reinhold, NY.

National Oceanographic and  Atmospheric Admini-
  stration (NOAA). 1992. Restoration guidance docu-
  ment for natural resource injury as a result  of oil
  spills. National Oceanographic  and Atmospheric
  Administration, Washington, D.C.

National Research Council (NRC). 1992. Restoration
  of Aquatic Ecosystems. National  Academy  Press,
  Washington, D.C.
Shreffler, D. K., R. M. Thorn, M. J. Scott, K. F. Well-
  man, M.A. Walters, and M  Curran. In press. Na-
  tional review of non-Corps  environmental restora-
  tion projects. Institute for Water  Resources, U.S.
  Army Corps of Engineers. IWR Report 95-R- 12.
  Alexandria, VA.
U.S. Department of the Interior (DOI).  1991. Esti-
  mating the environmental costs of  OCS oil and gas
  development and  marine oil spills: a general pur-
  pose model.  Report prepared by A. T. Kearney, Inc.
  for the U.S.  Department of Interior, Minerals Man-
  agement Service, Washington, D.C.

                                                      Wetland  Mitigation  Design:
                                                             From  Theory to Reality
                                                                                    Marc Boule1
  Abstract:  Watershed planning,  and other non-
 project  habitat restoration techniques, offer  ap-
 proaches   for   providing   suitable   wetland
 mitigation for proposed project impacts. Experi-
 ence suggests, however, that  these basic ideas do
 not (and often cannot) always consider market
 forces and  other issues facing the project propo-
 nent. Examination of the  selection and design of
 the wetland  mitigation for the Emerald Downs
 Racetrack offers an opportunity to  compare  the
 ideal approach with the reality of political, finan-
 cial and scheduling constraints. These constraints
 may limit the opportunities  for  developing pre-
 ferred or recommended mitigation approaches.

  In the past,  wetland mitigation has been described
 as  more  art  than science. While there has  been some
 truth to  that,  today it  seems that what has become
 equally important is the art of communication. This is
 especially true with large  or  controversial projects
 (they seem to go  hand-in-hand),  where numerous
 agencies and  individual citizens or citizen's groups
 are involved. The controversy, or at least confusion,
 is often complicated by the conflicting mandates be-
 tween, and  sometimes within, the  various resource
 agencies. Consider, for example, the Washington  De-
 partment of Fish and Wildlife, where some personnel
 are responsible for protecting heron rookeries while
 others find herons to be an egregious predator on ju-
 venile salmon  and steelhead.

  This paper describes  a controversial project  that
 went through several years  of planning and environ-
 mental review. The purpose here is not to discuss the
 environmental process; rather,  the goal is to empha-
 size the changes in  project  and mitigation design in
 response  to agency review, and to compare and con-
 trast what was proposed, opposed, and ultimately ap-
proved, and is presently under construction.
Project Description

Pre-Project History
  In  1992 the Longacres  Thoroughbred Racetrack,
just south of Seattle, Washington, was purchased by
The  Boeing Company with the intent of converting
the site into a corporate campus. As a result of that
purchase, the horseracing industry found it necessary
to develop a new track facility. After almost three
years of searching, a site was identified in the City of
Auburn in the Green River Valley, about 20 miles
south of Seattle (Figure 1).
  Initially,  the proposed project  site encompassed
198 acres and required the fill of 53 acres of wetland.
About  36 acres of this was wet agricultural land that
had been abandoned for many years.  Approximately
156 acres  of the site was either actively grazed pas-
ture (including 17 acres of wet pasture) or previously
filled pasture,  at the time it was identified as the po-
tential  racetrack site (Figure 2). Once the project pro-
ponents became  aware of wetland regulatory issues,
they began to search for a mitigation site. Because the
project was located in the watershed of Mill Creek, a
tributary of the Green River, the search was concen-
trated in the Mill Creek Valley.
  Over the past  100 years, and especially the past
15 to 20 years, development in the Green River Val-
ley has resulted  in the loss of wetlands, floodplain,
wildlife habitat,  and other resources. As a result of
these impacts and continued development pressure in
the valley, federal, state, and local natural resource
agencies have combined efforts to develop a Special
Area Management Plan  (SAMP) for the Lower Mill
Creek  Valley. Although the SAMP is not yet com-
pleted, some preferences for  certain  types of devel-
opment or habitat protection in the valley have been
identified. These include identification of areas where
development might be considered more appropriate
and where wetland preservation or restoration might
be more appropriate. The basic goals of the SAMP
process were  used to identify a potential mitigation
1 Shapiro and Associates, Inc., 101 Yesler Way, Suite 400, Seattle, Washington 98104


. .
\ "..;;:<-
'0 Youngs



~" '
'-. "'.
(5 '1 ~"
. '-!
o I ~

~.\ ;
'v' \
/ .
. I

. /' --,'
~,'" ""'" <".'~"."""";?'C+/, ,

~ f'h"r <)VI'
"\~',, ';"/~'
\"'~~~~~,""',,',,.',,'. r".'. FS'i
~A!-Ta!5e.J N
''& ,apps
\ \ -,
0 7,500 15,000
 Scale in Feet 
Figure 1. Site Location Map.

                                                                 UPLAND GRASSES, FORBS
                            DMA RIGHT-OF-WAY
         RUSHES. FORBS
                                                                 UPLAND GRASSES, FORBS
                                                                 LOU BARDY POPLARS
                                                                 HORSE CHESTNUTS
                                                                 UPLAND GRASSES,
                                                                 FORBS, FRUtT TREES
                                                                  RED-OSIER DOGWOOD
                                                                _WETLAND GRASSES,
                                                                  RUSHES, FORBS
                                    UNFILLED SETBACK
Figure 2.  Existing Vegetation Racetrack Site.

                        Soub-Shrub/Open Water Wetlands



                                                                                        Theory to Reality
site for the proposed racetrack facility and to develop
a wetland mitigation plan. One of the SAMP goals is
restoration of wetland habitats within the Mill Creek
  Not much is known about the character of the Green
River  and  Mill  Greek  Valleys  prior  to Anglo-
American settlement. Soils suggest much of the area
was wetland, with a variety of hydric soil types. His-
toric records, such as newspaper stories, refer to im-
penetrable swamps in  the area. Many of the early
roads were made of logs "floated" on top of wet soils.
In the  early 20th century, state forestry maps show
much of the area converted to agriculture, but uncon-
verted areas are designated as wet forest cover.

The Mitigation Planning Process

The First Proposal
  The first mitigation proposal by  the project  propo-
nent was  to use a series of abandoned sewage treat-
ment plant lagoons  immediately adjacent  to  the
project site(Figure 3). The lagoons are used by water-
fowl at present, but the riparian habitat adjacent to
the open water is limited because of the steep slopes
on the berms enclosing the lagoons and because most
of the upland is dominated by non-native species such
as blackberries, Scot's broom, and  weedy grasses and
forbs. In addition, sludge in  the lagoons has  been a
problem for which the City of Auburn is still trying to
find a solution. The proposal, developed by Raedeke
and Associates (1992), was to excavate and  appropri-
ately remediate the sludge material and then create a
diversity of waterfowl and riparian habitat on the site.
  At the U.S. Army Corps of Engineers pre-Applica-
tion presentation to the resource agencies for the pro-
posed  racetrack,  several objections  were  raised
regarding the proposed mitigation. First, much of the
area was identified as wetland and waterfowl habitat,
so there would be  little creation of new wetland to
offset losses of wetland area associated with the proj-
ect. Furthermore, the proposed mitigation site is im-
mediately adjacent  to Auburn Municipal  Airport, so
enhancement  of  waterfowl  habitat  is  not recom-
mended nor highly  regarded by the Federal Aviation
Administration. Finally, the SAMP  process pointed
toward restoration of wetlands in the  vicinity of Mill
Creek, and the abandoned sewage lagoons  are even
farther from Mill Creek than  the project site. In gen-
eral, this proposal was not well received.
The Second Proposal
  With the concerns expressed at the pre-Application
meeting  in  mind, the project team  met with  the
SAMP  committee to  identify potential mitigation
areas. As a result of those discussions, areas of Lower
Mill Creek where restoration might be implemented
were identified and the project proponents began the
search for a site to purchase. In the lower valley, Mill
Creek has been severely  ditched along most of its
length and is generally characterized as a vertical-
walled channel up to 5 or more feet  in depth with lit-
tle  or no overstory riparian canopy. Water  quality,
especially high turbidity, high temperatures, and low
dissolved oxygen (DO), are the major factors  limiting
its value as salmonid habitat.

  Ultimately the proponents were able to identify and
acquire a 56.5-acre parcel about 1 mile south and up-
stream of the project  site. Although it offered  many
opportunities, it also  was bisected  by SR  167, the
major freeway through the area. The  26.5-acre por-
tion of the site west of the freeway has no roadway
access,  straddles 1/4  mile of Mill  Creek, and  con-
sisted of about 19 acres of wetland dominated by reed
canarygrass and other non-native species, 4.6 acres of
native  emergent  and  scrub-shrub  wetland,  and
2.9 acres of upland, primarily abandoned pasture. The
approximately 30-acre site  on the  east side of the
freeway, consisted of 19 acres of grazed wet pasture,
9.2 acres  of  upland pasture, and about  1.5 acres of
upland and filled wetland that supported a house and
barns adjacent to the pasture.
  The project team evaluated this new site in light of
its existing wetland functional values and developed a
restoration plan to initiate plant communities similar
to those thought to have existed in  the area prior to
European settlement. The plan involved reconfiguring
the creek channel away from a ditch  cross-section and
more toward a natural stream channel cross section,
and vigorous mechanical control of reed canarygrass
in combination  with  heavy planting of native  trees
and shrubs (Figure 4). The intent was to create, over
time, an overstory community  that resembled the pre-
settlement community, which would  shade out the
reed canarygrass and would improve water quality in
the creek through shading.
  Some concerns were expressed by resource agen-
cies about  the specifics of  the plan,  particularly
whether shading could effectively eliminate  reed ca-
narygrass and whether reconfiguration  of the creek
channel would result  in significant water quality im-
pacts  during construction.  The  major   concern




Figure 3.  First Mitigation Proposal (1992).


 OFFStreWeTLAHOUirtQATIOHPLAN                  ' ''-"\

Figure 4. Second Mitigation Proposal.
                                            ID ronuno OCIUMO njamnt -
Theory to Realitv
expressed,  however,  was  that  the  approximately
60-acre mitigation area, with only  10 acres of new
wetland creation,  would not adequately  mitigate the
58 acres of proposed wetland impact.

The Third Proposal
  The project proponents then acquired an option on
an additional 20 acres of land immediately upstream
of the initial mitigation site, about half of which  was
upland and included another 1/4 mile  of Mill Creek.
A  third restoration proposal was now prepared  that
 involved  conversion  of another  10 acres  of upland
and fill to wetland, and restoration of an  additional
 10 acres of wetland through control of pasture grasses
and planting of native trees and shrubs (Figure 5).
Additionally, this  third proposal offered  the opportu-
nity to reintroduce Mill Creek  into a portion of its
original channel. At some time in the  past, the creek
had been diverted into a ditch along the back lot of
this property, but remnants of  the original channel
remained. Creation of a short segment of new channel
would  allow the  creek to  be reintroduced into its
original channel, while allowing the ditch to function
as  a backwater channel, for fish refuge during periods
of high flow.
  Resource agency concerns about this  proposal in-
cluded a  perception that wetlands  on this  site were
not significantly degraded (despite the fact they  had
been converted to agriculture  and grazing land  for
almost a century and abandoned only in the last 10 to
20 years), and therefore could  not be  substantially
improved. But more importantly, a concern was ex-
pressed that reintroduction  of the  creek into its old
channel would affect water quality in the  creek be-
cause of potential  erosion from introducing the creek
into an unestablished channel. Furthermore, the relo-
cation would result in the loss of  emergent wetland
communities that  had become established  in  the ex-
isting abandoned creek channel.

The Fourth Proposal
  At this point the  project proponent decided it would
be  expeditious to further reduce project impacts to
expedite the regulatory process. Several elements of
the project were reduced or eliminated, and the over-
all  extent  of the project was reduced. Most impor-
tantly, the abandoned agricultural  land  at the south
end of  the  site, with  its  39 acres  of wetland,
(Figure 2) was eliminated from the  project  area. As a
result, wetland  impacts were  reduced  to  a fill of
17 acres of grazed wet pasture. With this reduction in
the extent of wetland impact, the project proponent
found it  uneconomical  to  include  in the mitigation
plan the 20-acre parcel added in the third restoration
proposal described above.
 At the same time, two long meetings were held with
representatives from the project proponent, the U.S.
Army Corps of Engineers (lead permitting agency),
and all other federal, state, and local agencies, and the
tribes with jurisdiction  or concern over the project.
The purpose of these meetings was to develop a con-
sensus of opinion on what would be appropriate miti-
gation for the proposed  project. The result of these
meetings was a series of guidelines-but not quite per-
formance  standards-describing what the mitigation
needed to accomplish. The overriding focus of these
guidelines was improved waterfowl habitat, because
that was identified by the resource agencies  as  the
greatest habitat value on the proposed  project site.
Fish habitat in Mill Creek was suggested as the sec-
ond most important element in any mitigation plan.
(Interestingly, while  fish habitat is recognized as a
limiting condition in  Mill Creek, impacts of the pro-
posed project on fishery resources was expected to be
limited to stormwater runoff impacts, that, based on
computer modeling and other assessment techniques,
were anticipated to be minimal.)
 Based on  the input from these meetings, a fourth
restoration proposal was prepared for the 56.5-acre
site previously identified for mitigation. The empha-
sis of this  fourth plan was  on waterfowl breeding
habitat, because the project site had been identified as
waterfowl  overwintering habitat.  Interestingly,  be-
cause green-winged teal had once been observed in
the abandoned agricultural area originally  proposed
as the south part of the  project site, this species be-
came a key indicator for state and federal resource
agency personnel, and its habitat requirements were
the major criteria for developing the mitigation plan.
The result was a mosaic of open  water ponds scat-
tered over the site  and  separated by zones of scrub-
shrub  wetland  community  (Figure 6).  A forested
wetland  community  surrounded the outside. Little
was proposed for Mill Creek itself, except to provide
a canopy of forested and scrub-shrub species along its

 Although the goal had been to reflect agency priori-
ties, response to the fourth restoration proposal was
mixed,  and several revisions were  suggested by
agency personnel.  Most importantly,  it was deter-
mined that  there  was  too  great  an   emphasis on

Figure 6.  Fourth Mitigation Proposal.
Figure 7.  Final Mitigation Proposal.

 Theory to Reality
 waterfowl habitat, especially on the west side of the
 parcel. In addition, a suggestion was made to develop
 a new meander to the creek channel and to provide
 backwater  channels  as  salmonid  refuge habitat.
 Finally,  a desire was  expressed  to  create  some
 shallow sedge meadow communities.

 The Fifth Proposal
  Back to the drawing board again,  the project team
 developed a fifth proposal that reflected the most re-
 cent  revisions requested by  the  resource agencies
 (Figure 7). A new creek channel was included, as
 were backwater channels. Open water on the west
 side was limited to two small pond areas adjacent to
 the freeway that do not drain directly to Mill Creek.
 This was the proposal finally approved, with  some
 minor modifications, as part of the  Section 404 per-
 mit for the proposed project (USACOE 1995).
potential  impacts  of  restoring creek  to  original
Proposal: Reduce Project Impact to
17 Acres of Wetland Fill
 Step 4 Mitigation: Reduce area of mitigation to
50 acres, emphasize waterfowl habitat.
 Concerns: Too much  open water, potential long-
term impacts on water quality in creek, need to create
new stream channel and sedge habitat.
 Step 5 Mitigation:  Reduce area of open water, cre-
ate new creek  channel, backwater  channels, and
sedge habitat.
 Concerns: Still a 7-acre net loss in total  wetland
habitat within the Green River Valley.

Lessons Learned
  Controversial projects requiring especially large ar-
 eas of wetland fill can require extensive coordination
 with resource agencies  to determine appropriate miti-
 gation. In this example, development of an acceptable
 wetland  mitigation plan required  extensive negotia-
 tion and  numerous revisions to respond to the diverse,
 and sometimes changing,  concerns of the agencies.
 The following is a brief summary of the mitigation
 design and negotiation process.

 Proposal: 52 Acres of Wetland Fill
  Step 1 Mitigation: Enhance waterfowl habitat  ad-
 jacent to the project site.
  Concerns: Too close  to airport, not enough wetland
 creation, too much emphasis on waterfowl and not on
 wetland restoration.

  Step 2  Mitigation: Restore 40 acres of wet pasture
 and 10 acres of upland to forested swamp, and  en-
 hancement of 1/4 mile of Mill Creek.
  Concerns: Not enough wetland  creation,  potential
 construction impacts of reconfiguring stream channel.

  Step 3 Mitigation:  An additional 10 acres  of
 mostly abandoned wet  pasture for enhancement, an
 additional 1/4 mile of Mill Creek to restore, including
 relocating creek into original channel.
  Concerns: Still not enough wetland creation or gain
 of functional value,  need more waterfowl  habitat,
 Even with numerous meetings  and telephone con-
versations, opportunities for miscommunication  or
missed  communications abound.  To  reduce  this,
every meeting should be documented. Had we  sum-
marized our  meetings with the  SAMP committee,
noting our understanding of what was implied by the
concept of "restoration," we possibly would  have
avoided the effort and delays involved in creating a
plan that did  not meet other agencies' concerns. Had
the resource agencies been clearer early in the proc-
ess, about the importance waterfowl, the emphasis of
the mitigation plan could have been redirected. There
cannot be too much communication with regulatory
agency personnel when developing large and contro-
versial projects.

 The mandates,  and thus the goals, of the myriad of
resource agencies involved in the regulatory process
may vary  significantly. There is  often a marked lack
of consistency in these mandates and goals.  Optimi-
zation for waterfowl habitat is not always the best for
fisheries resources,  and vice versa. Agency represen-
tatives must establish their parameters and priorities-
preferably early  in  the process. Otherwise,  fulfilling
one agency's requirements only  leads to  direct con-
flict with another.

                                                                                      Theory to Reality
 Agency staff should be included in the design proc-
ess. It is  a delicate, politically sensitive process, for
agency  staff often do not  want  to be  seen  as
"encouraging" a project. It also leads to design by
committee, but this might be preferable to, and more
cost- and time-effective than, redesign by committee.

  Without consensus on the part of resource agencies,
a proposed project is destined to flounder, losing time
and  costing money. The project proponent and/or
their consultants are not always in the position to en-
sure consensus,  but  that  should  always  be  their
 The Emerald Downs wetland mitigation experience
was certainly not a model of efficiency, and the proj-
ect  suffered  more  from  political,   non-wetland
concerns  that  have  not been  discussed  here.
Nonetheless,  this  experience  offers  some important
lessons  in  how  to  reduce  the controversy  and
streamline the process.

Literature Cited
Raedeke Associates. 1992. Conceptual Wildlife
  Habitat Mitigation Plan for the Garrett/Lone/
  Auburn Downs Race Track Property, Auburn, WA.
  Prepared for Garrett Enumclaw Company.
U.S. Army Corps of Engineers. 1995. Auburn Thor-
  oughbred Horse Racing Facility, Final Federal
  Environmental Impact Statement (NEPA),
  USACOE, Seattle District, Seattle, WA. 2 vols and

                       Ecological Functions of a Saltmarsh/Mudflat
                                Complex  Created Using  Clean Dredged
                                        Material, Jetty  Island, Washington
                                                                 Jonathan P. Houghton1
                                                                       Robert H. Gilmour1
  Abstract:  The  Port  of Everett and the U.S.
 Army Corps of Engineers created a sand berm
 on an existing island in the winter of 1989 and
 1990.  Four objectives  were established:  (1) to
 balance erosion losses on the exposed west side of
 the island, (2)  to create additional dune grass
 habitat, (3) to create a protected embayment that
 would be colonized by marine invertebrates, and
 (4)  to demonstrate a  beneficial use of clean
 dredged material. Development of the embay-
 ment was expected to allow increased productiv-
 ity and invertebrate biomass, thereby improving
 habitat for juvenile salmonids, other fish, shore-
 birds, and waterfowl. Progress toward meeting
 these objectives has been monitored over 5 years.
 A  5.3-ha (13-ac)  berm  was  constructed with
 clean dredged material, forming a protected soft
 mudflat that covers 7.7 ha (19 ac). Experimental
 planting of six saltmarsh species conducted in
 the spring  of  1990 demonstrated  that  upper
 portions of the embayment (above +9 ft MLLW)
 would support such species. A second planting of
 Jaumea, Salicornia, and Distichlis in this area in
 1991  has   flourished  and   achieved   over
 100 percent  cover in  some  areas.  Natural
 colonization  of  salt-tolerant  species  also  has
 occurred above +10 ft  MLLW. Field sampling
 has demonstrated that juvenile salmon and other
 fish use the embayment  during high tides and
 that the lagoon supports a rich population of
 epibenthic zooplankton  prey. The protected
 mudflat  is  widely used  by migrating dunlin,
 dowitchers, plovers, other sandpipers, and bald
 eagles. The project fully met its objectives by the
 1995 date set for final project evaluation.

  The  Jetty  Island  berm project  was  planned,
designed, and coordinated by the Port of Everett
(Port)  and  the  U.S. Army Corps of Engineers
(Corps). Seattle District, to demonstrate beneficial
use of clean dredged material for habitat develop-
ment. By creating a sand berm on an existing island,
the Port and the Corps expected four objectives to
be met within 5 years: (1) to balance erosion losses
on the exposed west side of the island, (2) to create
additional  American  dunegrass  (Elymus  mollis)
habitat,  (3) to create  a protected embayment that
would be  naturally colonized by marine inverte-
brates, and (4) to demonstrate a beneficial and eco-
nomical disposal option for clean dredged material.
The protected  intertidal  mudflat within the em-
bayment was expected to contribute to increased
productivity and biomass of fish prey  and, there-
fore, improved  habitat for juvenile salmonids out-
migrating to northern Puget Sound.
 Construction of the federal navigation channel in
Everett Harbor from 1894 to 1903 resulted in large
volumes of sediment requiring disposal.  Creation of
Jetty Island began in  1903 with construction of a
rock jetty behind  which these  dredged materials
could be placed. Corps maintenance dredging of the
channel  and settling basins (Figure 1) continued to
build the island until 1969.  Today Jetty Island is
approximately 3 km long and covers approximately
40  ha (100 ac) above mean  higher  high water
(MHHW). Upland vegetation is dominated by Scot's
broom (Cytisus scoparius), blackberry (Rubus spp.),
and many other shrubs  and herbaceous species.
Extensive mudflats and Possession Sound lie to the
west of the island; the Snohomish River estuary, the
federal navigation channel, and the City of Everett
lie to the east (Figure 1).
 In the mid-1980s the Corps identified an oppor-
tunity to enhance and increase existing Jetty Island
habitat with material from maintenance dredging of
the channel and lower settling  basin.  In 1987 an
Agency Technical Advisory Committee (ATAC)
was formed to develop specific project details. An
environmental assessment of the area was com-
pleted prior to placement of dredged material (U.S.
Army Corps of Engineers 1989a). Abundance and
1 Pentec Environmental, Inc., Edmonds, Washington.


                                                                Jetty Island, Washington
    0   1800  3800      7600
          SCAIE ( ft )
Figure 1. Location and Vicinity Map

 Jetty Island, Washington
 distribution  of benthic  infauna  and  epibenthic
 Crustacea  as well  as use of the  project area by
 resource species (salmon, Dungeness crab, birds)
 were  investigated.  Environmental  design  and
 monitoring programs were finalized  in the  Imple-
 mentation  Plan (U.S. Army  Corps  of  Engineers
 1989b),  which stated project  objectives, responsi-
 bilities of participants, performance standards, and
 methods for evaluating and monitoring success.
  Specific  monitoring was required over 5 years. A
 quantitative assessment method was  developed to
 evaluate  pre- and post-project juvenile  salmon
 habitat losses and gains. For the project to be con-
 sidered a success, surveys were  required to demon-
 strate that  (1) juvenile salmonids use the project
 area, and (2) juvenile salmonid habitat is increased
 by at least 10 percent over pre-project levels. Other
 studies evaluated the physical stability of the berm
 and the  success of experimental plantings of salt-
 marsh vegetation within the lagoon formed by the
 berm. The criterion for success of saltmarsh habitat
 development was  that  the vegetative community
 cover at least 40 percent of the  planted area inside
 the berm and not differ statistically from that of the
 saltmarsh reference site located  on the eastern side
 of the island (Figure 2).
  Hydraulic  maintenance  dredging was completed
 on January 6, 1990. A 550-m  (1,800-ft) long berm,
 occupying a little  over 5.3 ha (13  ac), was created
 by placement of 323,000 cy of clean sands and silts
 on the west side of Jetty Island. Placement of the
 berm created a 7.7-ha (19-ac) mudflat within a pro-
 tected embayment on what had been the sandy west
 shore of the island.


 Physical Monitoring
  The Corps developed physical  performance crite-
 ria for the  project  and  monitors physical features
 (i.e.,  coastal  geomorphology,  topography).  Ap-
 proximately 65,000 cy of the berm  was predicted to
 erode over 8 to 10  years. As long as the berm ero-
 sion rate  does not exceed this predicted  rate by
 more than 25 percent, the frequency of nourishment
 required  to  preserve it  should be  about every
 10 years (U.S. Army Corps of Engineers 1989b).
Biological Monitoring
  Epibenthic  zooplankton  was  sampled in April
and May of 1992 and  1995 using a 0.018-m2, bat-
tery  powered epibenthic  sampler.  Three replicate
samples  were analyzed from each of 18  stations
representing  three depth strata (+9, +6,  and +3 ft
mean lower low water  [MLLW]) inside and outside
the berm and at the north and south reference areas
(Figure 2).  Samples were sieved on a  0.25-mm2
screen,  and zooplankters  recognized as  key prey
taxa  for juvenile pink  and  chum salmon  were
enumerated and wet weighed.
  To evaluate the success of the project in terms of
juvenile  salmonid habitat,  an equation was devel-
oped  and adopted by  the ATAC to describe the
quality of the habitats  in  terms of biomass of key
prey species  and other relevant factors. This equa-
tion (U.S. Army Corps of Engineers 1989a; Pentec
Environmental, Inc. 1994) is as  follows (see Table 2
for definitions of factors):
  Habitat Units (HU) = ac X PWF X (Cover factor
+ Slope factor) X RIF
  Fish use of the area  was evaluated using a 40-m
(130-ft) beach seine in the spring of 1992 and 1994
under high tide conditions at locations inside and
outside the berm and at the north and south refer-
ence areas.
  Because the artificial establishment of saltmarsh
vegetation  in urban estuaries of  Puget Sound was
not well  understood,  experimental planting  was
conducted in 1990, immediately after placement of
the dredged  material,  to  determine at which ele-
vations within the embayment saltmarsh vegetation
would survive. Experimental plugs of Deschampsia
caespitosa,  Carex lyngbyei, Triglochin  maritima,
and Scirpus  maritimus were obtained from Wave
Beach Grass Nursery in Florence, Oregon. Plugs of
native saltmarsh  vegetation,  predominantly  Dis-
tichlis spicata, were separately dug from a reference
marsh on  the east side of  the  island (Figure 2).
Dunegrass also was planted over approximately 1.5
ha (3.6 ac) along the western side of the berm.
  By  September 1990, it was apparent that vegeta-
tion would survive only within a limited area of the
embayment;  thus, plantings  in the spring  of  1991
focused  on areas  between +9  and +12 ft  MLLW.
Distichlis-dommated plugs collected from the refer-
ence  marsh  were planted along the inside of the
berm primarily between +11 and +12 ft MLLW to
augment  Distichlis   surviving   from   the   1990

                                                       Jetty Island, Washington

                   PROJECT AREA

Figure 2. Project Area

Jetty Island, Washington
 plantings. Salicornia and Jaumea plugs from Wave
 Beach  Grass  Nursery  were   planted  primarily
 between +9.5 and +11 ft MLLW along the inside
 base  of the berm  (southern tip of the protected
 lagoon) approximately 106 m to the north along the
 inside of the berm and  approximately 40 m  north
 along the island shoreline (east side of the lagoon).
  Field monitoring was conducted in early Septem-
 ber 1993 and in late  August 1995 (Years 3 and 5)
 to measure the success of the saltmarsh plantings as
 well  as natural colonization. In the project planting
 area, three relatively distinct vegetative zones were
  1. A lower zone between +9.5 and +11 ft MLLW
 that was dominated by Jaumea carnosa with a sig-
 nificant  component  of  Salicornia  virginica  (the
 "lower zone" or "Jaumea/Salicornia zone").
  2. An upper zone  beginning at  about +11 and
 + 12 ft MLLW and continuing up into the supralitto-
 ral in some areas that was dominated by Distichlis
 spicata, E.  mollis,  and Spergularia  (the "upper
 zone" or "Distichlis/Elymus zone").
  3.  A supralittoral  zone  generally above  +12 ft
 MLLW that was dominated by  E. mollis from natu-
 ral colonization.
  The  Jaumea/Salicornia   and  Distichlis/Elymus
 zones were sampled as distinct strata;  the supralit-
 toral zone was not sampled. North  of the planted
 area, the pattern of  three distinct zones was  not
 evident, and a single zone dominated by Spergu-
 laria was sampled. The  reference marsh was sam-
 pled  to  gauge success  of the marsh  developing
 within   the  protected   embayment.  A   small
 sub-embayment inside the first small spit that  has
 formed at the end of the berm was also sampled to
 document  natural  colonization that is  occurring
 there. Percent cover was determined by species in
 25,  0.25-m2 quadrats randomly placed over  the
 upper and lower zones within the inner berm and at
 the reference site. Similar  data also were recorded
 for the reference area and the "Point" area at the tip
 of  the berm (Figure 2). Individual  species cover
 values were summed to derive total cover in  the

  Randomized analysis of variance (ANOVA) and
 t-tests (Edgington 1987)  were used to  test for sig-
nificant elevation or  site  differences  in  species
cover and total plant cover.

 A portion  of the western  side of the berm first
eroded during winter storms in 1990/1991 and has
continued to erode (Nelson 1996). Natural beaching
of logs, supplemented  by placement of old boom-
sticks by Port staff, has added some stability to the
berm. Three small sand spits have formed at the end
of the berm (Figures 2 and  3), creating small pro-
tected flats at about  +9 to +10 ft MLLW. By early
1991  some areas along the eastern side of the em-
bayment became eroded  or covered with sand that
had been transported by storms. Elevations changed
from  0.5  to 2.0ft within the embayment. Woody
debris from the dredging formed thick mats (to 20
cm) over  some areas inside  the embayment. These
mats have gradually dissipated to the point that their
coverage is insignificant to epibiota.
 The Corps has completed yearly surveys to assess
the  stability  of  the  berm  and  developed   a
post-project bathymetric  map in  1992 and  1995.
Recent analysis (Nelson 1996) has indicated that the
west shore of the berm has lost up to approximately
150ft to erosion  and  that  the  berm has lost
approximately 37,000 cy of sediments. Renourish-
ment  is  now  being considered on  a 7-  or
8-year cycle to maintain  the integrity  of the berm
with minimal risk of  overtopping or breaching.

Epibenthic Zooplankton  Productivity
 The biomass of epibenthic  fauna from both the
Year 3 and 5 samples was dominated by Crustacea
other than key prey  taxa at all locations except for
the +9 ft stations inside the berm during  April.
Biomass of key prey taxa by month and elevation  in
1995  is presented in Table 1.
 In April the highest biomass of key prey taxa was
found at +9 ft MLLW in  the inner berm of the proj-
ect area.  Biomass  was  significantly  higher (p  <
0.05) at these stations than at the +9 ft stations  at
the reference areas. Also, in April no key prey taxa
were  taken  from +9 ft  MLLW  in   the exposed
project area, nor at +7 ft MLLW inside the berm.
 In May the highest biomass of key prey taxa was
found at +7 ft MLLW inside the berm but the dif-
ference was not significant. No key prey taxa were
taken from +9 ft  MLLW in the exposed  project

L: :
o A .~
" "') ~.,

            Table 1.  Mean key prey biomass and the prey weighing factor (PFW) from epibenthic sampling in
            Port Gardner Bay, 1995.
(grams/square meter)



+6 or +7
+6 or -1-7
Project area
0.1 161'
Prey Weighing Factor
area (= before)

Project area
Inside Exposed
90.70 0.00
314.75 2.00
— 0.09

            'Includes both inside berm + 9 stations. The + 7-ft transect is included in the + 6-ft (MLLW)

                                                                                   Jetty Island, Washingtc
area. Key prey taxa were also not taken from +6 ft
MLLW in  the  exposed project  area  or reference
Salmonid Habitat Evaluation
 Project area prey weighting  factors (PWFs) from
the Year 5  data were greater than 1.0 at the inside
berm +9-ft and +7-ft MLLW elevations, and at the
exposed +6 ft MLLW elevation; but less than 1.0 at
the  exposed +9-ft  and  +3-ft  MLLW  elevation
(Table 1). The high PWFs in the inside berm project
area occur because of the greatly  enhanced abun-
dance of key prey taxa within the protected embay-
ment.  In Year 2  it  was expected that epibenthic
zooplankton  productivity  at  +3-ft MLLW on  the
outside of the  berm would increase  as  the  berm
continued to stabilize. This appears to be the case;
the PWF at +3 ft MLLW, on the outside of the
berm, in Year 2 was 0.01,  and in Year 5 the PWF at
this same elevation was 0.09.
  The  following  analysis  recalculates the habitat
units using Year 5 PWFs,  1995 percent cover data
from project planting area (inner  berm), and habitat
areas (acreages) derived from the Corps' May 1993
reconnaissance flight. The 1993  acreages within
each tidal elevation and substrate type are included
in Table 2 along with acreages that would be extant
(based on pre-project bathymetry) had the berm not
been built. The project area is defined as the area
between  two perpendicular lines  running westward
from the shoreline of Jetty Island  (+12 ft MLLW)
out to the 1993 +2 ft MLLW elevation. A portion of
this area was below +2 ft before project construc-
tion. Changes of less than 1 ft elevation in this area
between  1990  and  1993  probably resulted  from
gradual erosion of material off the berm and were
not considered  significant  in this analysis; acreage
below  +2 ft in 1989 was included in the +2 to +5 ft
elevation category on Table 2.
  The habitat loss of approximately 13.2 ac between
+2 and +12 ft  (the  area   of  berm above +12 ft;
Table 2)  eliminated  production of prey  items for
juvenile  salmon;  to avoid a net  loss  of salmon
feeding opportunities, one  success  criterion for the
project was that the  habitat units calculated by the
above  formula  be  at  least 10 percent  above  the
pre-project  level  at  Year 5  (1995).  The  high
epibenthic zooplankton productivity in the mudflats
within  the  protected embayment generated a high
PWF that  in turn led to a calculation of a high
number of  habitat  units   and a  high percentage
(4,853 percent)   of  increase   over   pre-project

  As in 1992, there is a high degree of variability in
the  YearS epibenthic  zooplankton data  which
leaves  some uncertainty regarding  this  analysis.
However, these data are consistent with those from
the Year 2 epibenthic sampling which indicated a
4,021 percent increase in habitat.  Thus, it can be
concluded that the project has greatly exceeded the
salmonid habitat criterion.
Fish Use of the Project Area
  Beach seining in 1992 and  1994 confirmed that
juvenile salmon were present within the Jetty Island
lagoon  and at two reference areas  during their out-
migration from the Snohomish River system. Mean
catch in the project area in 1994  (1.2 salmon/set)
was somewhat lower than that from the West Jetty
Island reference areas (8.3 salmon/set) and substan-
tially lower than  that from the north end of the
island (71.0 salmon/set). Mean catch/set of juvenile
pink salmon at the exposed sites on the berm was
greater than catches  in  the  reference areas, which
are also  exposed. However,  catches of juvenile
chum salmon were substantially higher at the refer-
ence  areas than  in the  project area, exposed  or
sheltered. Regardless, the presence of juvenile sal-
monids at the six project sites sampled indicates use
of the project area as required by the criterion for
  In addition to chum and pink salmon, several other
fish species were captured in 1994. Numbers of ju-
venile surf smelt (Hypomesus pretiosus) were much
higher  in the  project   area  (49.7/set;  averaging
47 mm) than in  the reference areas. Catch of stag-
horn sculpin  (Leptocottus armatus) was also much
higher in the project area.
  Mean percent  cover and total percent cover for
species sampled in the upper zone at the inner berm
and reference site in  1995 are plotted on Figures 4
and 5.  Similar 1995  data for the  lower vegetative
zone at the inner berm and reference site are plotted
in Figures 5  and  6  (the lower zone  is not repre-
sented at the Point).  Mean  percent cover for  each
species  (Jaumea  carnosa,  Salicornia  virginica,
Spergularia  marina, Distichlis  spicata,  Atriplex
patula, and Elymus mollis)  and mean total percent
cover for upper and lower zones in the inner berm
were compared with  upper  and lower zones at the
reference site (Table 3). In the upper zone, there

 Table 2. Pre- and post project habitat area (acres) and quality (habitat units).*
Elevation (MLLW
+ 12 to + 10ft
+ 10 to +8 ft
+ 8 to +5 ft
+5 to +2 ft
Total +12 to +2 ft

Area gain (loss)
Habitat units gain
Percent change
Without Berm With Berrn ( 1 995)
) Area


HU Area Slope Cover PWF RIF HU
0.0 3.1 0.5 0.0 90.7 1.00 139.8
0.0 6.5 0.5 0.0 90.7 1.01 304.7
0.0 9.6 0.5 0.0 315.0 1.00 1507.3
0.0 0.0 0.5 0.0 0.0 1.00 0.0
0.0 19.2 1951.7

Without Berm With Berm (1995) Total Acres
Area Slope Cover PWF RIF HU Area Slope Cover PWF RIF HU Before After
3.1 0.5 0.0 1.0 1.0 1.5 2.7 0.5 0.0 0.0 1.0 0.0 3.1 5.7
4.9 0.5 0.0 1.0 1.0 2.4 2.4 0.5 0.0 0.0 1.0 0.0 4.9 8.9
24.1 0.5 0.0 1.0 1.0 12.1 21.0 0.5 0.0 2.0 1.0 21.0 24.1 30.6
47.7 0.5 0.0 1.0 1.0 23.8 21.3 0.5 0.0 0.2 1.0 2.0 47.7 21.3
79.7 39.9 47.4 23.0 79.7 66.5



Habitat Units

After Change
140 138
305 302
1528 1516
2 -22

*The project area is defined vertically by the 1989 +12-ft contour and the 1993 + 2-ft contour. A portion of the project area that is currently above +2-ft MLLW was below that elevation before the berm was constructed in 1989.
The difference of 13.2 acres represents area filled to above +12-ft to form the berm.
HU (Habitat Units) = Acres x PWP x (Cover factor + Slope factor) x RIF.
PWF = Prey Weighting Factor, local mean key prey species biomass/reference area key prey species biomass at same elevation. Data from April and May 1992 sampling (see Table 6 in Pentec 1993).
Cover and slope factors are defined in Pentec 1993.
RIF = Relative Regional Importance Factor, a recognition that some habitats are more valuable because of past losses. Assigned a value of 1.3 for saltmarsh areas.

RIF of 1.3 and cover factor of 0.2 assigned to 0.04 acres between +10 and 12 ft with average vegetative cover of 25 percent.
RIF of 1.3 and cover factor of 0.4 assigned to 0.16 acres between +8 and 10 ft with average vegetative cover of 59 percent.

Table 3. Mean percent cover for individual plant species August, 1995.

Reference Berm
Jaumea carnosa
Salicornia virginica
Distichlis spicata***
Atriplex patula
Elymus mollis
Total cover**
Unplanted Area
o Jaumea carnosa*
= Salicornia virginica
 Distichlis spicata**
Q) P
5 3 Total cover
0 «
*p < 0.05
**p < 0.01
*"p < 0.001









                 LJ  Jmimea carnosa

                     Distichlisspicala •••

                     Elymus moll a

                     Salicornia virginlca

                 [U  Atriplex patula

                     Spergularia marina
                       Reference-Upper                   Benn-Upper


     Figure 4.  Key species present in the Distichlis/Elymus zone, 1995.
                     Upper                Lower
Upper               Lower
        Planted Area
Point Cove
    Figure 5.  Mean total vegetative cover (%) in late summer 1993 and 1995.

                                                                                  Jaumea carnosa •

                                                                                  Sallcornla vlrglnlca

                                                                                  Diiilchlis spicata ••

                                                                                  Spergularla marina *
                                                                                Unpbnted Beim-Lower
 Figure 6.  Key species present in the Jaumea/Salicornia zone, 1995.
                                      1995 Mean=97 percent cover
                                                                                           • 1993 DisiicMislElymus
                                                                                           • 1993 Jaumea/Salicornia
                                                                                           •1995 DisiicMislElymus
                                                                                           • 1995 Jaumea/Saticorniu
                  1993 Mean=25 percent cover
                                                           1993 Mean=59 percent cover
                      1995 Mean=50 percent cover
   -10 J-
                                              Distance along transect (m)
Figure 7.  Width of upper (Distichlis/Elymus) and lower (Jaumea/Salicornia) vegetative zones in the
inner berm.

Jetty Island, Washington
 was no  significant  difference in percent cover of
 Jaumea   camosa,   Salicornia  virginica,  Atriplex
 patula, or Elymus mollis between the inner berm and
 the reference site. In the lower zone no significant
 difference was found in percent cover of Salicornia
 virginica between the inner berm and the reference
 site in 1995. However, several other taxa differed
 significantly in the lower zone. Total percent cover of
 all taxa combined differed significantly in the upper
 zone,  but not the lower zone. Changes in total plant
 cover between 1993  and 1995 are shown in Figure 5.
  The 1993 and  1995 width of the upper and lower
 vegetative  zones,  Distichlis/Elymus and  Jaumea/
 Salicornia, respectively, are plotted on Figure 7. Over
 a distance of 99 m along the upper transect, a total of
 0.05 ac was represented  by the Distichlis/ Elymus
 zone in 1995; mean  percent cover over this area was
 97 percent. Along the lower transect (148 m),  a total
 area  of 0.1 Sac was  represented by Jaumea  and
 Salicornia; mean percent cover over this area was
 49.9 percent.


  Before  the  berm  was created,  no mud substrate
 occurred  on  the west side of Jetty Island. A total of
 76.7 ac of sandy habitat was present between the ele-
 vations of +2 ft and +12 ft MLLW. A portion (3 ac)
 of the present-day  project  area  was  below +2 ft
 MLLW before the berm was constructed. Creation of
 the berm placed the post-construction  2-ft contour
 approximately 394 ft west of the  pre-project 2-ft
 contour in some places.
  When the berm was originally created, it occupied
 about  15 ac  above  +12 ft MLLW; currently, only
 about  13.2 ac remain above this elevation due to ero-
 sion of the berm. Thus,  the project has resulted in the
 loss of 13.2 ac of intertidal sandflat.

  As  expected,  depositional  conditions within  the
 19.2-ac mudflat inside the berm have greatly favored
 the production of invertebrates including epibenthic
 prey  of juvenile salmonids. Based on the  1995
 epibenthic zooplankton  abundances  and  the  1993
 bathymetry, the project  has greatly exceeded the cri-
 terion  of  providing  >  110 percent  of pre-project
habitat units  for juvenile salmonids.  Presence of
juvenile salmon within the lagoon in  1992 and 1994
has met the criterion for demonstration of salmonid
use of the area.

  Although  no  infaunal  monitoring  was  required
under the Implementation Plan, qualitative observa-
tions  confirmed a rapid colonization of the projected
embayment formed by the berm. Early in the  spring
of  1990,  ghost shrimp (Callianassa califomiensis)
substantially colonized the mudflat below about +8 ft
MLLW.  Since 1990, qualitative excavations within
the lagoon have shown a continual increase  in the
apparent density and diversity of the  infaunal  com-
munity with polychaetes, crustaceans, and molluscs
all  contributing to  this assemblage. By summer of
1992, small softshell clams  (Mya  arenaria)  were
established  along the firmer muddy sand  along the
inside of the  berm  between about  +7 and  +8 ft
MLLW.  Although elevations  within the lagoon are
relatively high (i.e.,  >  6 ft MLLW)  for Dungeness
crab,  some  use is indicated by an observation of a
small (approximately 3 cm)  crab moulting in the
moist mud of the lagoon.
  The abundance of shorebirds, smelt, and flatfish as
well as the juvenile salmon feeding on benthic and
epibenthic prey within the lagoon demonstrate that
the project area is functioning as an important part of
the Port Gardner ecosystem.
  Natural colonization of vegetation on the  top of the
berm has been remarkable. Seedlings of some  12 to
15  species became established in the spring of 1990,
and the  vegetative  community has  continued to
develop rapidly with species  such as  bighead sedge
and beach peavine that are also dominants on drier
areas of the main island. During an August 1994 sur-
vey, we collected over 40 species of vascular plants
on  the center of the berm (e.g., above +12 ft MLLW).
The berm now supports several shrub and tree species
(e.g., Scot's broom and black cottonwood [Populus
balsamifera trichocarpa]) that have grown to more
than 3 m in height.
  Dunegrass plantings in  1990 were very  successful
and continue  to expand  beyond and between the
planted rows  along the west side of  the top of the
berm. A majority of this area planted with dunegrass
has eroded away, and it appears that the rhizome mat
of  the grass is only  marginally helpful at retarding
this erosion. A similar erosion of the dunegrass com-
munity was  occurring  prior  to  berm  construction
along much of the northwest shore of the island. The
dunegrass plantings and natural colonization by other

                                                                                   Jetty Island, Washingtc
species have significantly retarded wind erosion on
the top of the berm, however. Natural colonization of
dunegrass from waterborne seeds and rhizomes from
eroding areas of the island to the north has developed
into a similarly dense growth along the lower eleva-
tions (+12 to +14 ft MLLW) on the east side of the
  The nearly complete lack of success of the 1990
experimental  plantings  of  saltmarsh plants  in  the
lagoon contrasts sharply with the success of the 1991
plantings  and  subsequent  natural   colonization.
Quantitative sampling in  late summer  of  1993 and
1995  demonstrated considerable success of species
planted in 1991 and in some areas  of natural coloni-
zation. The lack of a statistically significant differ-
ence  in  cover in  some  of the  dominant  species
between  the area inside the berm and  the reference
marsh (Jaumea, Distichlis, Elymus in the upper zone;
Jaumea,  Salicornia in the  lower zone)  is  highly
encouraging. Over much of the planted areas of the
lower  zone,  the combined  cover of  Jaumea and
Salicornia exceeded 100 percent, and Jaumea was in
bloom in early September 1993. Differences in other
species and in total cover remain, however, and it is
unlikely  that the vegetative communities in the two
areas (berm inside and reference marsh) will ever be
identical because of subtle environmental differences,
most  notably  in  salinity regime.  However, it  is
expected that  the  healthy and expanding  saltmarsh
fringe on the  inside  of the berm will continue to
develop and function as a significant and  beneficial
habitat feature that would  otherwise be lacking on the
west side of Jetty Island. A disturbing finding was the
discovery in 1994 of  a large number of plants of the
invasive  exotic Spartina  ?anglica; the Port has initi-
ated an eradication program.
  Sampling of  fish and invertebrate use  of the lagoon
at high tide demonstrates that several functions of a
natural saltmarsh (fish use,  feeding, refuge;  inverte-
brate  production)  are also  provided by this fringe
marsh. Qualitative observations  also  suggest that
important marsh functions of use by shorebirds and
waterfowl as well  as  organic detrital production are
also occurring.
  Beyond its documented use by immense numbers of
juvenile surf smelt and lesser numbers of salmon and
other  fish,  it  seems  likely  that  the  greatest single
ecological benefit of the berm and lagoon has been its
suitability for  a variety  of ecological  functions  of
birds. Virtually all  of  the bird species reported in the
baseline studies have  been observed to  use the berm
and/or lagoon, often  in greater numbers  than seen
elsewhere on the island (Pentec Environmental, Inc.
1996). Construction of the berm added three impor-
tant bird habitats not previously found on the west
side of Jetty Island: unvegetated sandy uplands, salt-
marsh, and mudflats.  Each of these habitats is used
differently by birds, but collectively they constitute a
habitat complex similar to some of the better habitats
found in more natural estuaries. Although the unvege-
tated  sandy uplands have been reduced in area by
vegetative colonization since  construction, the  two
lower spits that have formed on the north end of the
berm continue to offer this type of habitat.

  Monitoring of the Jetty  Island berm  project and
evaluation of performance standards through Year 5
indicate that the  project has  met the criteria estab-
lished in the  Implementation  Plan. Monitoring  has
shown that  the area is being used by juvenile sal-
monids and has met  the criterion for demonstrated
use in at least two years.
  The  productivity  of the  protected  embayment
exceeded  project goals  for  salmonid habitat  by
4,000 percent. As anticipated, the lower energy re-
gime  and mud substrate of the protected embayment
is providing a more productive habitat  for juvenile
salmon prey  resources  than  exists  on  the  more
exposed sandy beaches outside the berm. Monitoring
has shown that the area  is being used by juvenile
salmonids and has met the criterion for demonstrated
use in at least two years. The abundance of juvenile
salmon prey resources has enhanced the project area
habitat value for juvenile salmonid feeding and for
use by juvenile surf smelt, an important forage  fish.
This gain in productivity  within the embayment has
more  than offset productivity lost by filling 13.2 ac of
previously  intertidal  habitat  to greater than +12 ft
  Year 5 monitoring showed that the percent cover of
all plants and  of selected plant species in the planted
portion of the protected  embayment did  not differ
significantly  from  those in  a  reference  marsh
although other taxa did differ significantly.
  The ATAC  is considering a proposal by the Port
and  the  Corps  to  renourish  the  berm in 1997 to
maintain the increased habitat values created.

 Jeiiy Island, Washinglon
   The Port of Everett and the U.S. Army Corps of
 Engineers, Seattle District, have provided logistical,
 technical, and field support for this project. Thorough
 review of project plans and the habitat evaluation
 model used  by the ATAC greatly  improved the
 technical  quality of the  monitoring  program  and
 results.  We would  like to acknowledge Wilber E.
 Ternyik of Wave Beach Grass Nursery in Florence,
 Oregon, who provided  the nursery plants; the many
 state and  federal agency representatives  who volun-
 teered to  assist with the transplanting; and  Dr. Ron
 Thorn (Battelle Northwest) for the original design of
 the vegetation experiment and for technical advice on
 habitat development in urban  estuaries.

 Literature Cited
 Edgington, E.  S.  1987.  Randomization  tests,  2nd
   edition. Marcel Dekker.
Nelson, E. 1996. Jetty Island protective berm erosion
  analysis and renourishment plan. Draft manuscript
  by  the  U.S.  Army  Corps of  Engineers, Seattle
  District. April 5, 1996.

Pentec Environmental, Inc.  1994. Beneficial use of
  dredged material, Jetty Island habitat development
  demonstration  project.  Years 3 and 4 Monitoring
  Report. Report to the Port of Everett, WA.

Pentec Environmental, Inc.  1996. Beneficial use of
  dredged materials, Jetty Island habitat development
  demonstration  project. Year 5  Monitoring Report.
  Report to the Port of Everett, WA.

U.S. Army Corps of Engineers. 1989a. Final  envi-
  ronmental assessment,  Jetty Island habitat devel-
  opment. Seattle District, Seattle, WA.

U.S. Army Corps of Engineers.  1989b. Implemen-
  tation plan for habitat  development project using
  dredged  material at Jetty Island,  Everett,  WA.
  Seattle District, Seattle, WA.

                                      Restoration  and  Management  of the
                                                             Salmon River  Estuary
                                                                            Chris McDonald1
 Abstract:   Located   along  the  north-central
Oregon coast, the Salmon River estuary and sur-
rounding lands were set aside by Congress as the
Cascade Head Scenic Research Area in 1974. This
is the first Scenic Research Area designated in the
United States. The goals of the research area are
to protect scenic values, provide research oppor-
tunities and within the 1085 acre estuary return
lands to a more natural condition.
 The U. S. Forest Service bought 80 acres in 1974
to help achieve restoration  goals within the estu-
ary. Much of the original estuary had been diked
and converted to pasture land or recreation use
changing hydrologic and ecological processes. Res-
toration began in 1978 when dikes were breached
within the 80 acre  pasture to re-establish a salt-
marsh. Studies of the restored saltmarsh began
before  dike  removal  and  continue today. The
Salmon River estuary has  the longest continuous
record of restoration research of any estuary on
the Oregon coast.
 Additional land has been acquired and restored
throughout the years. The Forest Service owns
over  80 percent of the estuarine lands and has
restored 230 acres to salt marsh through dike re-
moval.  The value of  restoring this  estuary has
increased its importance as other bays and estu-
aries along the Oregon  coast have decreased in
natural productivity  and diversity. The estuary
provides other opportunities for research such as
studying estuarine dependence of anadromous fish
stocks. Future restoration plans include additional
removal of dikes, which will allow 120 acres to re-
turn to a more natural condition.
 Western  Oregon and  Washington are known for
their scenic beauty,  productive forests and riparian
ecosystems. Estuarine wetlands are an important but
small part of the Northwest's lands and are increas-
ingly  being recognized for their  contribution  to
healthy ecosystems. Since European settlement, most
of these estuarine lands have decreased in size and
productivity as lands have been diked, drained and
converted to other uses. The  Salmon River estuary is
one of the  least  disturbed along the Oregon coast.
Since 1974, the management goal under the U.S. For-
est Service Administration has been to restore the
estuary to a more natural condition. This 1,085 acre
estuary is one of the state's smallest; however, it in-
cludes one of the largest wetland restoration projects
in Oregon.

  The Salmon  River estuary is located along the
north-central Oregon coast  approximately  7 miles
north  of Lincoln  City (Figure 1). The estuary is part
of a 75-square-mile watershed surrounded by mostly
coniferous forest. Major land uses are forestry, agri-
culture and recreation. The  estuary and associated
wetlands are entirely within  the 9,670-acre Cascade
Head  Scenic Research Area (CHSRA), part of the
Siuslaw National Forest in the Hebo Ranger District.
  People have been part of the Salmon River estuary's
landscape for  centuries. Perhaps  as  long  ago  as
500 years, the  Salmon River became home  to the
Nechesne Indians (Beckham, 1975). Numerous cul-
tural sites can be found along the banks of the river as
the Nechesne culture was oriented toward the ocean.
Legends of these  Indians indicate that they used Cas-
cade Head,  a promontory rising above the sea to the
north  of the Salmon River's mouth, as a vigil site for
individuals seeking power and visions.
  The estuary  and   surrounding  area  began to  be
populated by European settlers in the mid-1800s.
Early settlers cleared the land to raise cattle and grow
vegetables. Major changes came to the estuary in the
1920s when the Salmon River Highway and Oregon
Coast Highway were constructed. With transportation
available, farmers started dairies, and residents of the
Willamette Valley and Portland area had access to the
coast  to  fish and  enjoy  the  scenic beauty. Since
World War II, the area has grown in popularity.
  Historic photos of the estuary show major  changes
began in the early 1960s when dikes were constructed
(Figures 2 and 3). The dikes kept tidal waters from
flooding the salt  marshes so that they could  be used
for pastures. In 1961, U.S. Highway 101 was rerouted
from a narrow winding road through the forest to a
 Hebo Ranger District, Siuslaw National Forest, P.O. Box 324, Hebo, Oregon 97122.

Salmon River Estuary
      Figure 1.  Location of the Salmon River estuary within the Cascade Head Scenic Research Area.
 route nearer the coast. The highway crosses the upper
 portion of the estuary and functions much like a dike.
 During  highway  construction,  Salmon  Creek,  a
 tributary to the Salmon River, was diverted from its
 natural channel into a ditch. In  the early 1970s, an
 entrepreneur built the Pixieland  amusement Park on
 55 acres of diked marshland covered with fill. A  mo-
 bile home park and boat launch area followed else-
 where  in  the estuary. The Salmon  River  area
 appeared to be destined in the late 1970s for devel-
 opment (Figure 3).

  Even though much of the estuary was  being devel-
 oped, much of the area remained in a nearly natural
 state. A  small group of residents in the surrounding
 area became interested in preserving the estuary  and
 adjacent lands  from further development. In  1965,
 they  purchased  Cascade  Head  and   donated  the
headland  to  The  Nature Conservancy. This same
group  of concerned land  owners  then  requested
Representative  Wendell Wyatt  and Senator Bob
Packwood  to consider  formal  protection  of the
estuary  and surrounding lands. In 1974, President
Gerald Ford signed Public Law 93-535 making the
Cascade Head and surrounding lands the first (and
only) scenic research area in the United States.
  The U.S. Forest Service  administers this specially
designated parcel. As stated in the legislation, the
general  management  objectives of the  CHSRA are
"...to provide present and future generations with the
use  and  enjoyment  of certain ocean headlands,
rivers,  streams,  estuaries,  and forested areas,  to
ensure  the protection and encourage  the study  of
significant areas for research and scientific purposes
and to promote a more sensitive relationship between
man and his environment..  ."
  A  Management Plan and EIS were  approved in
1976. The writers of the management plan recognized

                                                                                     Salmon River Estuary
that the value of this estuary  would increase in  im-
portance as other bays and estuaries along the Oregon
Coast decreased in size and productivity. Both long-
and short-term goals were  agreed upon at  that time.
The long-term goal for the estuary is "revitalization
and restoration  of the Salmon River estuary and its
associated wetlands to a functioning estuarine system
free from the influence of man." The interim man-
agement goal is to study the  issue of dike removal
while allowing existing agricultural use to continue.
  The  Act states  that changes  in land ownership
would be necessary to meet legislative  goals and it
specified how lands could be acquired. In 1974, there
were   no  public  lands in the  estuary  and  only
35 percent of the estuary remained in a  semi-natural
state. As landowners became  aware of management
goals they sold their lands to the U.S. Forest Service.
Most lands were acquired from willing sellers, how-
ever,  some land owners were reluctant to sell,  and
those lands were condemned. Money for the land ac-
quisitions came from the Land and Water  Conserva-
tion Fund, a  federal fund that dedicates windfall oil
profits to federal property acquisition.
  The U.S. Forest Service bought 80 acres in 1974 to
assist restoration goals within the estuary. Restoration
began in  1978  when  two-thirds  of the outer dikes
were  breached within a 52-acre pasture. A research
project conducted by a Ph.D. student focused on  col-
lection of baseline vegetation and soils data before
dike breaching,  methods of restoring hydrologic con-
nectivity, and a monitoring strategy to follow the ef-
fects   of  breaching   the  dikes  (Mitchell,  1981).
Vegetation changes  occurred  almost immediately as
pasture grasses  were replaced by  salt marsh species.
No planting  was done. Diking caused the  pasture to
subside in elevation as soils became  compacted  and
dewatered. The restored surface land  is gradually
regaining its  elevation. Seventeen years  of data have
been  collected on the changes in elevation,  vegeta-
tion,  accretion and elevation,  and soil salinity.  The
Salmon River estuary has the longest continuous  rec-
ord of restoration  research of any estuary  on the
Oregon coast.
  Additional  estuarine  land has been acquired and re-
stored throughout the years. The U.S. Forest Service
now owns over  80 percent of the estuarine lands and
has restored 230 acres to low salt marsh through dike
removal. The value of restoring this  estuary  has in-
creased its importance as other bays and estuaries
along  the Oregon coast have been  altered. Future
restoration plans include additional removal of dikes
and tidegates. Several sites west of Highway 101 are
being considered for dike removal in 1996. They in-
clude a two acre pasture near the mouth of the river,
an eight acre pasture with  partially breached dikes
along the north  shore and a 55-acre pasture immedi-
ately west of Highway 101. Other long-term restora-
tion goals are to acquire additional lands, reconnect
Salmon Creek under Highway  101 and restore hy-
drologic conditions east of Highway 101 where wet-
lands have been drained and  streams diverted  from
original channels.

  Almost  22 years after the CHSRA legislation and
19 years after the Forest Service management plan
was approved, it is recognized that the stringent goal
of restoring  the Salmon river estuary to a state "...
free from the influences of man..." may not be en-
tirely feasible. In many cases previous conditions are
unknown. Also, it is not realistic to assume that all
manmade features will be removed. However, resto-
ration  within the salt  marsh has been successful.
Monitoring  started by Diane Mitchell and continued
by Dr. Robert Frenkel (Dept. of Geosciences, Oregon
State University) and his graduate students has shown
that tidal channels are deepening, salt marsh vegeta-
tion has been re-established, and sediment accretion
is occurring.
  Because of its management goals and history, the
Salmon River offers opportunities for research.  With
57 percent of the Salmon River functioning as a near
natural system,  it  provides the best and longest data
set for salt marsh restoration on the Pacific Coast and
a prime area for studying the  fresh and saltwater
interface for anadromous fish. For more information
on possible  involvement in research in the Salmon
River estuary, contact the U.S. Forest Service Pacific
Northwest  Forest Sciences  Laboratory,  Corvallis

Literature Cited
Mitchell, Diane L. 1981. Salt  Marsh  Re-establish-
  ment Following Dike  Breaching in  the Salmon
  River Estuary, Oregon. Corvallis, OR: Oregon State
  University. 171 pp., Ph.D. thesis.
Beckham, Stephen D. 1975. Cascade  Head and the
  Salmon River Estuary: A History of Indians and
  White Settlement. Report prepared  for the U. S.
  Forest Service.

...,  .,.
... . 

Figure 3.

                                Salmon  River Salt Marsh Restoration in
                                                                  Oregon: 1978-1995
  Abstract: The  U.S. Forest Service initiated the
 Salmon River estuary salt marsh  restoration in
 1978 by removing dikes from a 22 ha pasture. In
 1987,  a nearby  58 ha pasture was likewise re-
 turned to tidal  circulation. Vegetation composi-
 tion,  biomass,   elevation,  sediment   accretion,
 salinity, and  substrate texture have been moni-
 tored  for 15  years since  1978. Adjacent natural
 salt marshes  serve as research controls. Species
 composition changed rapidly after dike breaching.
 Pasture species, killed by saltwater,  were  dis-
 placed initially  by   annual   marsh   colonizers
 (Cotula coronopifolia  and  Spergularia  marina),
 After  4 years, regionally common  native species
 (Carex lyngbyei, Salicornia virginica and Distichlis
 spicata) began to replace the annuals and  today
 dominate the  restored marshes. Reestablished low
 marsh differs in species composition from the high
 marsh controls.  Diked pasture elevations were ca.
 35 cm lower than adjacent control  marsh eleva-
 tions,  caused by subsidence.  Subsided restored
 surfaces increased  in elevation at a greater rate
 than control elevations. Sediment-clogged pasture
 creeks deepened 20-60 cm  upon tidal  reconnec-
 tion. Substrate in the restored marsh accreted ca.
 4 cm  per decade.  Estimated above ground  pri-
 mary productivity based on biomass measurement
 of the pasture was comparable to that of natural
 salt marsh controls (ca.  1,200 g/m2/yr) and to
 other  Pacific  Northwest  high  salt  marshes.
 Restored  marsh productivity  was  much greater
 (2,300 g/m2/yr)  than controls. Restoration trends
 differ between the  two restored units. The  58 ha
 unit lacked an intact drainage system characteris-
 tic of the 22 ha unit and surface ponding retarded
 restoration. The U.S.  Forest  Service restoration
 goal  was  to  return  diked pasture to its  pre-
 settlement condition. Restoration was successful in
 reestablishing apparently functioning salt marshes
 that are integrated into the estuary; however, long
 term prospects for complete restoration to pristine
 conditions are more problematic.
                       Robert E. Frenkel1

 Popular, government and academic interest in wet-
land and riparian  ecosystem restoration has bur-
geoned  in  the  past  decade as we become  more
cognizant of the enormous losses and degradation of
these critically valuable biotic systems (NRC 1992,
Mitsch and  Gosselink 1993, National  Academy of
Sciences 1995).  Small in proportion  to the earth's
other  impaired  and impoverished  landscapes, wet-
lands and riparian systems are recognized as having
over arching importance in regulating climate, main-
taining water quality, stabilizing hydrologic regimes,
and providing  crucial wildlife  habitat (Mitsch and
Gosselink 1993).
 Despite the fervor and concern over wetland resto-
ration, there have been few large scale wetland resto-
ration projects conducted over an extended period of
time. As the preferred compensatory mitigation strat-
egy, wetland restoration projects have typically been
on privately owned land, small  in extent and poorly
planned, executed and monitored (Kusler and Ken-
tula 1990).
 When Congress designated the Cascade Head Sce-
nic Research Area (CHSRA) in  1974 (P.L. 93-535), a
unique opportunity arose to restore and monitor a
substantial salt marsh area on public land along the
north central Oregon  coast. The U.S. Forest Service,
which administers the specially designated CHSRA,
was directed by Congress to restore ". . . the Salmon
River estuary and its associated wetlands to a func-
tioning  estuarine system free from the influences of
man . ... to its condition prior to the existing diking
and agricultural use" (U.S. Forest Service 1976).
About 75 percent of  the Salmon River salt marshes
had been diked to create pasture, mostly in the early
1960s. Conversion of coastal marshes has been com-
mon in Oregon where of the  state's original salt marsh
area, about 46 percent has been diked or filled since
the early 19th  century; most conversion has been to
pasturage (Boule and Bierly 1987). Besides accom-
plishing the Forest Service's goal of returning pasture
to near-natural estuarine conditions, restoration of the
Salmon River salt  marshes by dike breaching serves
 Department of Geosciences, Oregon State University, Corvallis, Oregon 97331-5506.


                                                                                  Salmon River Salt Marsh
as a long-term demonstration project that should be
helpful to wetland managers in planning other coastal
restoration programs and projects.

Study Site and Objectives
  The Salmon River project consists of two units re-
stored by dike removal, (1) a 21 ha pasture on the
north shore was returned to tidal circulation in  1978,
and (2) an  additional 63 ha of pasture on the  south
shore were restored in  1987. For comparison,  both
restored pastures  were selected so that they  are
flanked by  relatively intact natural  salt marshes that
served as "controls" (Figure 1).  Restoration was de-
liberate but passive in that dikes were breached al-
lowing  reestablishment of tidal circulation without
planting, seeding or surface  grading. Together with
graduate students,  I monitored these restoration sites
for 17 and 8 years respectively. Throughout the re-
mainder of the paper I refer to the north shore site as
"Mitchell Marsh" and the south shore as "Y-Marsh."
  I address four objectives: (1) will salt marsh vege-
tation reestablish in the absence of planting or seed-
ing; (2) is  it possible to  restore a diked pasture to
original high marsh conditions; (3) what is the rate of
restoration; and (4) is the  pattern of restoration simi-
lar in Mitchell Marsh and Y-Marsh?
  Diane Mitchell initiated the north  shore study in
 1978 and  established 115 one-meter square plots
along 20 transects  in the  restored and control  units.
At each plot she sampled vegetation cover, biomass
and elevation before and after dike removal in  1978
and 1979 respectively (Mitchell 1981). Together with
Mitchell and  Janet Morlan,  I  continued sampling
vegetation  cover in 1980, 1981, 1982, 1984,  1988,
and 1993.  Additionally,  I sampled  interstitial  soil
salinity, soil texture, soil  organic matter, and accre-
tion above  a sand layer placed on the pasture surface
in  1978.  Mitchell also  sampled  elevations  across
47 cross-sections along 8 marsh creeks. With Morlan,
I completely resampled the restored  marsh in  1988,
10 years after dike breaching (Morlan 1991, Frenkel
and Morlan 1991).
  On  the south shore in the  Y-Marsh, I  established
57-meter-square plots along 6 transects and sampled
vegetation cover in 1987 prior to dike breaching and
again in 1988, 1990,  1991, 1992,  1993, and 1995. I
surveyed plot elevations in 1988.
Salt Marsh Reestablishment

Composition and productivity
 Upon  dike removal, saline  (30 to  10 ppt) tides
flooded the  pasture, stressing upland  plants, and
within a year, killing them. Detached litter and dead
plants prevailed for  several  years after dike breach-
ing. Change in selected species composition is shown
in Figure 2 for Mitchell Marsh from pasture in 1978
to extensive low salt marsh  in 1993. We considered
three species groups: upland species (e.g., Holcus la-
natus and Ranunculus repens) that were killed imme-
diately;  residual  species  (e.g.,  Agrostis  alba and
Potentilla pacified)  that diminished in cover; and
both  ephemeral and persistent colonizing species
(e.g., Carex lyngbyei, Cotula coronopifolia, Distichlis
spicata  and Salicornia virginica) that invaded and in-
creased in  cover. Bare ground increased to a  maxi-
mum the second year after dike breaching after litter
was  washed away, and diminished rapidly thereafter.
Six  years after restoration,  two  low  marsh assem-
blages  dominated the Mitchell  site,  an extensive
brackish Carex lyngbyei community and a more lim-
ited  Salicornia virginica-Distichlis  spicata  commu-
nity  in more saline areas.
  A similar pattern of reestablishment occurred  on the
more recently  restored Y-Marsh, except that bare
ground  and litter dominated  and persisted for several
years after dike removal. Colonization by salt  marsh
species  was slow and initially characterized by annual
ephemeral  species (Cotula coronopifolia, Puccinelia
pumila  and Spergularia marina}. By  1993, two as-
semblages were prominent, a Deschampsia cespitosa
community  and a Salicornia virginica-Distichlis spi-
cata community.
  During the same interval, species composition in the
relatively intact control  sites  on both the  Mitchell
Marsh and Y-Marsh  changed very little (Figure 3).
  I estimated above  ground  net primary productivity
from sorted biomass in the Mitchell Marsh following
the method of Kibby et al. (1980). Pasture productiv-
ity was similar to that of the  control high marsh in
1978, about 1000 g/m2-yr. After dike breaching, bio-
mass at first diminished and then gradually increased
to 2300 g/m2-yr in 1988, more than double the value
of either the intact high marsh or the pasture.

    Figure 1. Location of restoration and control area in the Salmon River estuary. Mitchell Marsh:
    Restoration area 51 and 52, Control area 50 and 53; Y-Marsh: Restoration area 57, Control area 54.

        Percent  Cover







          ] AQAL
          3 DISP
     Q POPA
     HH HOLA

                                  ^a  SAVI
                                  CZH  BARE
   Figure 2.  Change in percent cover of principal plant species in the Mitchell Marsh restoration site,
   1978-1993 in the Salmon River estuary.
   AGAL = Agrostis alba, POPA = Potentilla paciflca, CALY = Carex fyngbyei,
   SAVI = Salicomia virginica, DISP = Distichlis spicata, HOLA = Holcus lanatus,
   TRRE = Trifolium repens, and BARE = surface not vegetated by vascular plants.
          Percent  Cover
 Figure 3. Change in percent cover of principal plant species in the Mitchell Marsh control site,
 1978-1988 in the Salmon River estuary.
 SCMI = Scirpus microcarpus, POPA = Poteniilla pacifica, AGAL = Agrostis alba,
 JUBA = Juncus balticus, DECE = Deschampsia cespitosa, and CALY = Carex lyngbyel.

Salmon River Salt Marsh
 Environmental Change
  By  environmental change  I  refer to variation  in
 controls that affect marsh  restoration including, sur-
 face elevation that determines tidal inundation, creek
 development, sediment accretion, salinity,  soil tex-
 ture, and soil aeration.
  During the 17 years that  Mitchell Marsh was diked
 (1961-1978),  the original marsh  surface  subsided
 30-35 cm  due to loss  of  buoyancy, oxidation and
 trampling, while the flanking control marsh surface
 did not subside (Figure 4).  A similar pattern of subsi-
 dence occurred on the Y-Dike side that was diked for
 27 years prior to dike removal; however the Y-Marsh
 site  was  generally  about 20cm  lower than  the
 Mitchell Marsh (Figure 4). Upon reestablishment of
 tidal exchange, the subsided pasture surface increased
 in  elevation at a greater rate than the control  marsh
  With dike removal, tides again  began to flush inac-
 tive,  clogged tidal creeks.  Creeks deepened and nar-
 rowed  due  to  erosion,  but  also  because   of
 sedimentation along  marsh levees (Figure 5). Marsh
 creeks show  greatest deepening low in the  marsh
 where tidal flushing was greatest, and least deepening
 at  creek heads.  In  the  Y-Marsh, the original tidal
 creek pattern had  been severely altered during  the
 diking  period.  In  places creeks  were  obliterated
 leading to ponding and anaerobiosis. Elsewhere the
 basic creek pattern remained and creeks deepened.
  Creek reestablishment was critical to marsh resto-
 ration. It is by creeks that sediments and nutrients are
 exchanged with the estuary. It is by creeks that saline
 water enters and drives compositional change, and, it
 is  by  creeks that anaerobic toxics are removed  and
 aeration enhanced.

  Salt marsh vegetation was fully restored to Mitchell
 Marsh and the  Y-Marsh  within about eight years
 without planting, seeding,  or grading. The only nec-
 essary manipulation was dike removal which reestab-
 lished tidal exchange.

  Composition of restored marsh vegetation, however,
 is very  different now from that prevailing  prior to
 diking and pasture creation in the early 1960s. Based
 on  relatively unaltered natural salt marshes  flanking
 the restored sites, I assumed that pre-dike composi-
 tion was typical of high marsh  with  dominance of
Juncus  balticus,   Deschampsia   cespitosa,   and
Potentilla  pacifica. Because  the pasture  surface
subsided 30-40 cm during  the  diked period,  upon
return  of estuarine circulation, low  marsh  charac-
terizes the restored sites and strongly reflects salinity
and  soil texture  conditions. Today,  an expanse of
almost pure Carex lyngbyei marks the more brackish
and  finer textured substrates. Where conditions are
more marine,  with higher salinities  and  sandier
substrates, Salicomia virginica and Distichlis spicata
  Above ground net primary productivity in the re-
stored  Mitchell  Marsh is  almost twice that of the
pasture and typical high marsh. Enhanced productiv-
ity probably reflects released resources and increased
tidally bom nutrients.  Since reestablishment of salt
marsh is akin to primary succession, it is not surpris-
ing to observe high net productivity (Odum 1969).
  To restore high marsh,  surface elevations would
need to be increased at least 30 cm. Accretion rates
are greater in the restored sites (ca. 5 cm per decade)
than in adjacent controls (ca. 1-2 cm per decade). Ac-
cretion  rates, however, are  more diminished in high
marsh than in lower areas near the river; therefore, I
estimate that it  would take at least 80 to 100 years to
change from a low to high salt marsh.
  Drainage and soil aeration also control salt marsh
restoration. When marshes are diked, natural drainage
patterns disintegrate and are altered by reduced flow,
sedimentation,  livestock trampling of creek banks,
and grading. For high marsh reestablishment, an effi-
cient integrated  creek system  must  redevelop. In
Mitchell Marsh, where the creek system was subdued
but not destroyed, restoration proceeded fairly rap-
idly. In the Y-Marsh,  surface grading had destroyed
the  original creek system  and drainage was  poor.
Marsh vegetation restoration was retarded.
  Poor  drainage  and  low  surface  elevations in the
Y-Marsh led to persistent ponding which created ex-
treme anaerobic  conditions. These were accentuated
by the  accumulation  of dead pasture litter that was
not  removed by tidal water. After  dike breaching,
extensive  patches  of algal mat  dominated the Y-
Marsh for about four to  five years.  Vascular  plants
tolerating   these   conditions  included   ephemeral
Cotula coronopifolia, Puccinelia pumila, Spergularia
marina, and Triglochin coccinea. More typical salt
marsh  species  were  found along creek margins  or
where soil aeration was greater. Gradually, soil aera-
tion increased, algal mats  broke up, and ephemeral

        Elevation (m NGVD)
1.6 4
                Mitchell  Marsh
Figure 4. Elevation distribution 1988 of the restoration site and controls in Mitchell Marsh and
Y-Marsh in the Salmon River estuary.
0.80 h
 0.60 h
 0.40 h
 0.20 t-
      Elevation (m)
                                                                Creek 1
                                                                Cross-section  4
                                      Distance (m)
Figure 5. Elevation across a typical marsh creek cross-section in the Mitchell Marsh restoration site
in 1978 and 1988 showing creek deepening associated with reestablished tidal circulation.

 Salmon River Salt Marsh
 species have begun to be replaced by common  re-
 gional salt marsh species.

  I addressed four objectives. First,  after dike  re-
 moval, typical salt marsh  vegetation reestablished in
 the  course  of about  eight years in the absence of
 planting and seeding. Although  I did  not measure
 marsh  function directly,  species composition,  en-
 hanced productivity, presence of salmonids in marsh
 creeks  (C. Simenstad  pers.  comm., June 30,  1993),
 and abundant waterfowl provide evidence of a func-
 tioning marsh.
  Second,  because  the  diked  pasture  subsided
 30-40 cm relative to the original high marsh, low salt
 marsh vegetation prevailed after dike breaching. It is
 unlikely that original  salt marsh conditions  will be
 restored within 100 years,  if then.
  Third, rates of restoration as measured by  vegeta-
 tion composition vary greatly. Under favorable con-
 ditions, Mitchell Marsh returned to low marsh within
 about six years after which composition changed very
 slowly. In the Y-Marsh, where surface drainage had
 been severely altered causing ponding and anaerobio-
 sis,  marsh vegetation took about eight years to rees-
 tablish and even then was changing annually.
  Fourth, the pattern of restoration was quite different
 in Mitchell Marsh than in the Y-Marsh. Although sur-
 face  subsidence was the  same in the two  restored
 areas, initial elevation, drainage and  anaerobiosis
 were strikingly different. Under anaerobic conditions
 prevailing  in the  Y-Marsh,  restoration of vascular
 plants was retarded, algal mats covered large areas,
 annual ephemeral  plants dominated for several years
 before  more  characteristic  marsh  plants  became

  Despite these differences in restoration pattern and
 rate, the Salmon River salt marsh restoration is a suc-
 cess. The project  demonstrates that one of the best
 opportunities to stem the national and regional loss of
 coastal  wetlands  is to return agriculturally diked
 coastal  wetlands to natural  salt marsh through dike
 removal. This  opportunity  is particularly advanta-
 geous on large tracts of public land. It is under these
 circumstances  that restoration  can  be  carefully
 planned and monitored over decades.
  The Salmon River estuary research project could
not have taken place without the careful meticulous
field and analytical work of Diane Mitchell and Janet
Morlan, for this I am especially grateful. I also wish
to acknowledge funding  assistance from  the U.S.
Forest Service and U.S.  Environmental Protection
Agency, Region 10.

Literature Cited
Boule, M. E.  and K. F. Bierly. 1987. History of estu-
  arine wetland development and  alteration:  What
  have we wrought?  Northwest Environmental Jour-
  nal 3:41-61.
Frenkel, R. E. and J. C. Morlan. 1991. Can we restore
  our salt marshes? Lessons from the Salmon River,
  Oregon. Northwest Environmental Journal 7:119-
Kibby, H. V., J. L. Gallagher, and W.  D.  Sanville.
  1980. Field guide to evaluate net primary produc-
  tion  of wetlands. U.S.  Environmental Protection
  Agency, Environmental  Research  Laboratory, Cor-
  vallis, OR, USA. (EPA-600/8-80-037).
Kusler, J. A. and M. E. Kentula (eds.). 1990. Wetland
  Creation  and Restoration: The Status  of the Sci-
  ence. Island Press, Washington, DC.
Mitchell, D. L. 1981. Salt marsh reestablishment fol-
  lowing dike breaching in the Salmon River estuary,
  Oregon. Ph.D. Dissertation, Oregon State Univer-
  sity, Corvallis, OR.
Mitsch, W. J. and J. G. Gosselink. 1993. Wetlands,
  2nd ed. Van Nostrand Reinhold, New York, NY.
Morlan, J.  C. 1991. Ecological status and dynamics
  of a salt marsh restoration in the Salmon  River es-
  tuary, Oregon. M.S.  Thesis, Oregon State Univer-
  sity, Corvallis, OR
NRC  (National Research  Council Committee on
  Restoration of Aquatic  Ecosystems).  1992. Resto-
  ration  of Aquatic  Ecosystems: Science,  Technol-
  ogy, and Public Policy. National Academy Press,
  Washington, DC.
National  Academy  of  Sciences.  1995.  Wetlands
  Characteristics and Boundaries. National Academy
  Press, Washington, DC.
Odum, E. P.  1969. The strategy of ecosystem devel-
  opment. Science 164:262-270.

                                                                                Salmon River Salt Marsh
U.S.  Forest Service.  1976. Final  environmental
  statement for the management plan: Cascade Head
  Scenic Research Area. U.S. Department of Agricul-
  ture,   Forest   Service    document   USDA-FS-
  FES(Adm)-76-06 on file at the Siuslaw National
  Forest, Corvallis, OR.

                             Batiquitos Lagoon  Enhancement Project:
                                                        Concept to Construction
                                                                           Michael Josselyn1
                                                                                  Ralph Appy2
                                                                            Adam Whelchel1
  Abstract:  The 593 acre Batiquitos Lagoon En-
 hancement Project is a joint restoration project of
 the City of Carlsbad and the Port of Los Angeles.
 The $55  million project is being funded  by the
 Port of Los Angeles as mitigation for the loss of
 deep water  marine habitat in the outer Los Ange-
 les harbor area. The enhancement project involves
 the  re-establishment of a tidal  entrance  to the
 blocked  lagoon  mouth and  dredging  approxi-
 mately 3.1  million  cubic  yards of accumulated
 sediments from the  lagoon basin to increase the
 tidal prism and restore coastal  lagoon habitats.
 The project has  been  under  construction since
 1994 using electric dredges.  A  portion  of the
 dredged material  is  being  used for beach  replen-
 ishment  and  a portion for the  construction  of
 nesting areas for the  California least tern and
 western snowy plover. Use of the nesting sites by
 these birds  has exceeded historic levels and con-
 tributes significantly to the coastal populations of
 these federally listed species. The project  will be
 completed in 1997.

  The Batiquitos Lagoon Enhancement Project (Fig-
 ure  1)  is the  largest  coastal wetland  restoration
 project in California and provides an example of the
 effort and planning required to implement large scale
 projects.  It is one of a series of six lagoons on the
 northern coast of San Diego County, several of which
 are also proposed for restoration. Most of the  other
 proposed restoration projects involve restoration of
 tidal action as was done at Batiquitos.

  The need for restoration at Batiquitos Lagoon was
 clearly recognized in  1976 when the Department of
 Fish  and Game issued its report on the current condi-
 tions in the lagoon (Mudie et al.  1976). The lagoon
 had  been  tidally influenced during historic  times;
 however,  the  construction  of road  and railroad
 bridges and causeways restricted tidal flows. With in-
 creasing agricultural and urbanization pressures on
 the lagoon's watershed, sediments filled the lagoon,
 decreasing its tidal prism. In addition, portions of the
 lagoon  were filled for development, diked for solar
 salt production, and freshwater flows damned and di-
 verted.  These  influences  resulted  in a lagoon envi-
 ronment that was  frequently  closed and sometimes
 completely fresh. In summer months, the closed sys-
 tem dried out and  hypersaline mudflats were preva-
 lent.  In less than  150 years, the lagoon  had been
 converted from a tidally influenced, marine commu-
 nity to  one that altered annually between freshwater
 and hypersaline conditions.
  The lagoon continued to support wildlife, particu-
 larly  migratory birds  (California Coastal Conser-
. vancy  1987).  In   addition,  the  federally  listed,
 endangered  California least  tern (Sterna antillarum
 browni)  and   threatened  western snowy  plover
 (Charadrius alexandrirrus  nivosus) were  observed
 and sometimes nested on the mudflats of the lagoon.
 The concern expressed by most observers was that
 despite the  existing values, the frequent long-term
 closure of the lagoon  coupled  with  the decreasing
 tidal prism due to siltation would eventually eliminate
 the existing wetland functions and values.
  In 1987, the  Port of Los Angeles, the City of Carls-
 bad,  the  California State  Lands  Commission, the
 California Department of Fish and Game, the US Fish
 and Wildlife Service, and the National Marine Fisher-
 ies Service  entered into a formal Memorandum of
 Agreement  to enhance Batiquitos Lagoon for the
 mitigation of Port  fills in San Pedro Bay. The state
 and federal resource agencies determined that the en-
 hancement of Batiquitos Lagoon would be an appro-
 priate mitigation to offset the adverse environmental
 impacts that would result from filling activities in the
 Wetlands Research Associates, Inc., 2169 E Francisco Blvd Ste G. San Rafael, California 94901.
 Port of Los Angeles, Environmental Management Division, PO Box 151, San Pedro, California 90733-0151.

Poinsenia lone
t. '.
Qc .
rh Q 4"el1.
'1T lie
Figure 1. Diagrammatic illustration of the Batiquitos Lagoon Enhancement Project.
F'::'.::j Least Tern & Snowy Plover Nesting Site
Wiltl Low Marsh (+1.34 + 4.0 1"1 NGVDI
- Intertidal (Mudflat) (~.1610 +1.34 lee' NGVOI
- Subtidal (below ~.16 1", NGVDI
o Existing Wetlands

 Batiquitos Lagoon Enhancement
 shallow  waters in Los Angeles Harbor. The mitiga-
 tion  was  both  off-site  and  out-of-kind, but  was
 deemed  valuable because "shallow, estuarine coastal
 embayment habitat  in southern California, with its
 relatively high value to marine fishes and migratory
 birds, has been reduced in area at a greater rate  than
 that of deep water habitat".
  A joint EIR/EIS document was prepared for the en-
 hancement  of the lagoon  (City of Carlsbad  1990).
 The EIR/EIS  evaluated a number of alternatives for
 the enhancement project. These alternatives ranged
 from no action to large scale dredging of the lagoon.
 The  City  and the  California  Coastal  Commission
 eventually determined that a project referred to as
 "Mitigated Alternative B" had the best opportunity to
 restore tidal action to the lagoon and to provide for
 improved habitat for marine and  estuarine fish and
  Engineering studies and specifications  were  pre-
 pared by the City and the Port and construction on the
 project commenced in April of 1994. The purpose of
 this paper is to summarize the construction procedure
 and to document the use of newly constructed nesting
 areas by least terns and western snowy plovers.

 Construction Summary
  The restoration of Batiquitos Lagoon consists of
 several major construction activities (Figure 1).

  •   Dredging the lagoon's central basin  and use of
      the sandy material  for beach nourishment and
      construction of shorebird nesting sites.

  •   Dredging the lagoon's east basin and the use of
      the silty  material as fill within the central basin.

  •   Dredging of the west basin and use of the sandy
      material  for nesting  sites and beach nourish-

  •   Construction of five California least tern and
      western  snowy plover  nesting  sites totaling
      38  acres.

  •   Construction of two 300-foot rock  jetties  ex-
      tending into the ocean from the mouth of the la-

  •   Replacement and/or rock revetment of bridges
     crossing the lagoon

  •   Revegetation of salt marsh vegetation disturbed
     during construction  of the project.
  The enhancement project has a unique approach to
disposal  of bottom materials that are unsuitable for
beach nourishment. Within the west and central ba-
sin, sandy materials are prevalent. However, in the
east basin, the material is largely fine silt and unsuit-
able  for placement on area beaches. As  a result, the
construction  involves an  over-excavation  of sand
from the central basin and the placement of the fine
silts  into the pit created. Approximately 1.6 million
cubic yards of sand will be placed on area beaches at
the conclusion of the project. The sand was also used
to create the nesting areas for shorebirds. These areas
range in size from 1 acre to 19 acres and were created
to be above tidal and flood water influence. Once the
fine materials from the  east basin are  placed  within
the central basin, it will  be capped with  sandy mate-
rial from the dredging in the western basin. At the
end of  the project, approximately 3  million cubic
yards of material will have been moved. The change
in habitat acreage as a result of the project is pro-
vided in Table 1.
  The project is also being constructed in a manner to
avoid impacts to existing natural resources. Dredging
is being conducted using electric dredges to  assure
quiet and reduce air pollution. All dredging is halted
during the least tern/western snowy plover nesting
period (April  1 to September  15). Construction ac-
tivities must also avoid the state listed Belding's sa-
vannah   sparrow  (Passerculus  sandwichensis bel-
dingi) habitat during the summer nesting period. All
salt marsh vegetation disturbed by the project must be
transplanted or replaced  on a one to one ratio. These
measures and others are being supervised by an onsite
environmental monitor throughout the construction.

Utilization of Nesting Sites During
  One of the  conditions established in the permits
issued for the project was that suitable  nesting sites
for California least tern and western snowy plover be
constructed during the project and the utilization of
these sites monitored  (Figure 2). During the first year
of  construction,   one   2  acre  site   (W-l)  was
constructed,  managed,  and  monitored. During  the
second  year of  construction, two additional sites
(W-2-5.2  acres  and E-l-19.1  acres)  were  made
available. Each of the sites was constructed with sand
either trucked  in  or  dredged from  the  lagoon. The
sand contained shell materials. Two of the sites (W-l
and  W-2) were  surrounded  on landward  facing

    Table 1.  Approximate acreage of habitats to the enhanced by the project.
Habitat type
Open nontidal water/seasonal mudflats
Intertidal flats (-1.6 to 3.9 ft MLLW)
Salt marsh (3.9 to 7.0 ft MLLW)
Nontidal nesting flats (+7 ft MLLW)
Other transitional habitats
Total acreage
             California least terns

                 Adult Pairs   	Fledglings
   1978  1979   1960  1981   1982  1983  1984  1985  1986  1987  1988

1990  1991   1992  1993  1994   199S  1996
    Figure 2. Utilitization of Batiquitos Lagoon Breeding Colony given in numbers of adult pairs and

    fledglings from 1978 through 1995.

Batiquitos Lagoon Enhancement
 portions with chain-link fencing up to 7 ft in height.
 On the waterfacing sides, they were  bounded with
 3 foot high polyethylene mesh fencing to retain least
 tem chicks inside  the  nesting area and  to provide
 some predator control. Site E-l was not fenced except
 at the access road.
  To  facilitate monitoring, a grid  system of various
 dimensions was established on  each of  the  three
 nesting sites. On W-l and W-2, a 20x20 meter grid
 system was utilized and on E-l a 60x60 meter grid
 system was used. The grid system assisted in locating
 and monitoring nest sites  and chicks and was supple-
 mented with marking each nest location with wooden
 slats  (tongue depressors). Monitoring  of the nesting
 areas  was conducted daily  either  by observations
 outside the nesting area or by walk-throughs. Counts
 were kept on number of eggs laid, hatched, aban-
 doned or depredated, and chicks fledged.
  During the 1994 breeding season, approximately 72
 adult pairs of the California least tern were observed
 nesting  at the Batiquitos Lagoon breeding colony
 (Whelchel and  Keane  1994). Nesting at Batiquitos
 Lagoon in 1994 represented approximately 5 percent
 of the total nesting pairs in San Diego County and 3
 percent of the State total. Nesting was  initiated on
 May  13th and ended on June 15th. The 72 adult pairs
 initiated a total of 80 nests. A total of  147 eggs were
 laid;  of  these, 129 hatched  (88 percent  hatching
 success). Mean  clutch  size  was   1.84.  Based  on
 fledgling counts, an estimated 68 chicks fledged from
 Batiquitos Lagoon. Fledgling dates ranged from June
 22 to July 25.
  During the  1995 nesting season, 82  adult pairs of
 the California least tern were  observed at Batiquitos
 Lagoon (Whelchel and Keane  1995). These birds ini-
 tiated a total of 93 nests. Nest initiation ranged from
 May 22 to June 20. The distribution of nest initiations
 between the three newly created nesting sites was 60
 (W-I), 30 (E-I), and 3 (W-2). A total of 165 eggs were
 laid; of these 146 hatched (88 percent hatching suc-
 cess).  The mean  clutch  size  was  1.77.  Based on
 fledgling  counts   and   recaptures,   an   estimated
 71 chicks fledged (0.87 fledgling per pair).

  In addition  to least tern, nesting  activity for  the
 western snowy plover was also recorded. This species
did not utilize the constructed nest site in 1994, but
did utilize  E-l  during  the  1995  nesting season
(Whelchel and Keane 1995). Fifteen  nesting attempts
were documented between April 3 and July 25. A to-
tal of  40 eggs were laid and of these  32 hatched
(80 percent hatching success rate). The modal clutch
size was 3 eggs with a mean clutch size of 2.7. Based
on fledgling surveys, an estimated 24 chicks fledged
between June 28 and September 19.
 Predation at the nesting sites was reduced through
the efforts of an active trapping and monitoring pro-
gram  conducted by  Animal Damage Control. Most
predation attacks on least terns were from American
kestrels, great horned owls, and a feral cat. Most pre-
dation attacks on western  snowy plovers were due to
common ravens.

 The construction  of the  Batiquitos Lagoon En-
hancement Project  represents the culmination of a
20-year planning effort. Restoration of degraded sys-
tems is  made difficult by  the conflicts that arise due
to  surrounding land  uses, recreational demands  on
beaches where inlets are proposed, infrastructure and
transportation corridor constraints, protection of fed-
erally and  state listed species,  and implementation
costs. The cooperation of federal and state agencies
coupled with the economic resources of the Port of
Los Angeles and the desire of the City of Carlsbad to
restore its natural resources was the combination that
made this project possible. The project benefits, es-
pecially to beach nourishment and habitat enhance-
ment, provided  the  public incentives to  accept the
 The use  of the  nesting areas by federally listed
shorebirds  has far exceeded expectations. This may
be due in part to the protection afforded these areas
by both passive and active predator control methods.
However, the wide-spread use of the site during con-
struction suggests that the design and implementation
of the project has  been successful. Only long-term
monitoring to be conducted  after the completion  of
the construction will tell whether the restoration itself
functions as designed.

Literature Cited
California State Coastal Conservancy. 1987. Batiqui-
  tos  Lagoon  Enhancement Plan.  Revised Draft.
  Oakland, CA.

City of Carlsbad.  1990. Batiquitos Lagoon Enhance-
  ment Project. Final EIR/EIS.

Memorandum   of   Agreement.   1987.   Agreement
  among the City  of Los Angeles, the City of Carls-
  bad, the California Department of Fish and Game,

                                                                          Batiquitos Lagoon Enhancement
 the  California  State  Lands  Commission,  the
 National Marine Fisheries Service, and the United
 States  Fish  and Wildlife Service to Establish  a
 Project for Compensation of Marine Habitat Losses
 Incurred by Port Development Landfills within the
 Harbor District  of the  City of Los Angeles by
 Marine Habitat Enhancement at Batiquitos Lagoon.
Mudie, P. J., B. M. Browning and J.W. Speth. 1976.
 The Natural  Resources  of  San  Dieguito  and
 Batiquitos Lagoon. Coastal Wetland Series #12.
 Department of Fish and Game, State of California.
Whelchel, A. W. and K.  Keane.  1994.  California
 Least  Tern  Breeding Survey, Batiquitos  Lagoon,
  San  Diego  County,   1994  Season.  Batiquitos
  Lagoon Environmental Monitor, City of Carlsbad

Whelchel,  A.  W. and  K. Keane.  1995.  Western
  Snowy Plover Breeding Survey, Batiquitos Lagoon,
  San  Diego  County,   1995  Season.  Batiquitos
  Lagoon Environmental Monitor, City of Carlsbad,

Whelchel,  A.  W.  and K. Keane.  1995. California
  Least Tern Breeding Survey, Batiquitos Lagoon.
  San  Diego  County,   1995  Season.  Batiquitos
  Lagoon Environmental Monitor, City of Carlsbad,

                               Restoration of  Spartina Marsh  Function:
                                                           An Infaunal Perspective
                                                                                     L. A. Levin1
                                                                                        D. Talley1
                                                                                        T. Talley1
                                                                                      A. Larson1
                                                                                       A. Jones1
                                                                                      G. Thayer2
                                                                                       C. Currin2
                                                                                         C. Lund2
  Although  rarely a focus of wetlands  restoration
 monitoring, sediment-dwelling fauna (infauna) play a
 number of  important roles in salt  marsh function.
 They cycle organic matter by  breaking down detritus
 and mixing sediment; provide  food for fish, birds, and
 other predators;  link primary producers, microbes,
 and fungi with higher order consumers; and represent
 a significant portion of the  faunal  diversity in salt
 marshes. Here we will summarize some key results
 from studies of macrofauna  (animals  retained on a
 300 micron sieve) in restored salt marsh systems on
 both  the Pacific  and Atlantic coasts of the United
 States.  We consider infaunal succession, measures
 and rates of recovery relative to natural marsh refer-
 ence  sites, factors that influence recovery, and possi-
 ble  causes of compositional differences  between
 created and natural marshes.

 North Carolina Sites
  Data presented  here are from studies  of the Port
 Marsh, a 2.2-acre created Spartina alterniflora marsh
 located on the Newport River  in Morehead City, N.C.
 An adjacent natural  marsh was used as a reference
 site.  Experiments were  initiated  in June  1990 to
 (l)test  the effects of 10 different  soil treatments on
 ecosystem recovery, and (2) evaluate  successional
 stages, mechanisms and  recovery potential of flora
 and fauna in created Spartina alterniflora habitats
 relative to an adjacent natural marsh. The treatments
 applied  included  a series  of  organic matter amend-
 ments (straw,  alfalfa, peat) with  and without inor-
 ganic N additions, as well as  planted and unplanted,
 rototilled  and  unrototilled controls, each replicated
 three times (one 2 x 7 m plot in each of three blocks).
Samples were taken from upper and lower elevations
of each plot in the created marsh and the natural ref-
erence marsh  at  1  wk,  1 mo, 3 mo, 6 mo,  12 mo,
15 mo, 18 mo,  22 mo, 27 mo, 47 mo, and 52 mo after

Unamended Treatments
  Recovery of the unamended sediments (planted and
unplanted but  not rototilled) during the first  4 years
after marsh establishment is described in Levin et al.
(in press). Throughout much of the study macrofaunal
densities were reduced in the planted (unamended)
and  unplanted (bare) plots  relative to  the  natural
marsh, especially at upper tidal elevations (Figure 1).
A similar pattern was observed for species richness.
Three taxa, the polychaetes Streblospio benedicti and
Capitella sp.,  and turbellarians, accounted for over
75 percent  of  the created marsh macrofauna during
the first several years after marsh establishment. Even
when macrofaunal  densities  and species richness
were comparable in the created and natural systems,
species   composition  differed  greatly.  Just  over
2 years after marsh creation, the artificial marsh had
92 percent   of the  natural  marsh   species and
88 percent of the individuals present, yet the compo-
sition of the two systems was markedly different. The
natural   marsh  was   dominated  by   oligochaetes
(61 percent),   which  are   burrowing,  subsurface-
deposit feeders, while in the created marsh the oligo-
chaetes were virtually absent, having been replaced,
proportionally, by  Streblospio benedicti,  a  tube-
dwelling, surface-deposit feeder.  5.  benedicti and
other surface-deposit  feeders may be more available
to fish and birds as food, but are less likely to aerate
' Marine Life Research Group, Scripps Institution of Oceanography, La Jolla, California 92093.
2 Southeast Fisheries Science Center, Beaufort Laboratory, 101 Fivers Island Road, Beaufort, North Carolina 28516-9722.

                NATURAL MARSH

                PLANTED CONTROL
                UNPLANTED CONTROL
            ELAPSED  TIME

            —£—  STRAW
            ....>...  ALFALFA

Figure 1. Mean number of macrofauna (>300 microns) collected per 18 cm2 core in the Port Marsh and an adjacent natural
marsh. Data are given for selected organic amendment treatments (straw, alfalfa, and peat incorporated into the soil at 3000 g/m2
with no additional nitrogen), planted and unplanted controls, and an adjacent natural marsh. Data are also presented for inorganic
nitrogen additions including unamended and straw-, alfalfa-, and peat-amended plots. Time indicates months since establishment
of the man-made marsh. Error bars are one standard error, and are included only on natural marsh and straw treatments, for
clarity of presentation.

  Batiquitos Lagoon Enhancement
 sediments or contribute to mechanical breakdown of
 plant detritus. A primary lesson from these results is
 that total abundance data and species richness meas-
 ures alone cannot reflect infaunal functional recovery
 without  the  additional  consideration  of  species
 composition and trophic roles.

 Organic Amendments
  We initially expected that addition of organic matter
 and nutrients to marsh soils  should expedite marsh
 recovery by enhancing growth of Spartina. Addition
 of straw, alfalfa, and peat initially reduced the rates
 of macrofaunal colonization  (Figure 1)  and species
 richness  relative  to  the  unamended  control plots.
 However, after 6 months all organic  additions except
 straw   had infaunal  densities statistically  indistin-
 guishable from control treatments. Negative effects of
 straw on macrofaunal abundance and species richness
 persisted for at least 22 months (Figure 1). The straw
 treatment  exhibited the  lowest redox levels of  any
 treatment during the experiment, and the lack of oxy-
 gen may have contributed to poor infaunal establish-
 ment.  Additions of inorganic N, tracked during the
 first  6 months  only,  appeared  to   further  reduce
 macrofaunal  colonization  in plots  with  organic
 amendments (Figure 1). The source  of this effect is

  The lower tidal heights came to resemble the natu-
 ral marsh more quickly than the higher elevations did
 in terms of both abundance and species richness. The
 greater  inundation  time  at  low  elevations  may
 enhance recruitment.

  Specific associations between fauna and above or
 belowground Spartina biomass have  been reported
 for a variety of infaunal taxa. We therefore note with
 interest that the planted and unplanted control plots
 exhibited similar patterns  of macrofaunal coloniza-
 tion in  most instances (Figure 1), although we did not
 sample  Spartina culms directly. Also, proximity to
 Spartina culms  (measured to  the nearest 10 cm) did
 not appear to affect densities of total macrofauna or
 dominant species in the  created marsh.  Because the
 unplanted  plots and  adjacent walkways were   sur-
 rounded by planted Spartina on all sides but one  (the
creek edge),  we  cannot  infer  that  the observed
successional  patterns  would  apply to  unvegetated
areas larger than 6 x 7 m.

Life Histories
 The role of life history in marsh succession also was
examined by categorizing each species  into one of
four development or dispersal modes, and then evalu-
ating the composition  of the system over  time with
respect to these  categories.  Planktotrophic larval
development was rare among macrofauna in the natu-
ral  (reference)  marsh,  while the created system was
dominated  by planktotrophic forms or direct develop-
ers  with swimming adult stages, especially during the
first several years. Conversely, in the natural system a
substantially larger proportion of its fauna exhibited
direct development with no dispersal stage,  relative to
the  created marsh, although representation of direct
developers increased with time in the man-made sys-
tem. Thus, dispersal potential may partially explain
the  early differences in faunal composition between
the  created and natural marshes.

Southern  California Sites
 Preliminary studies  have examined macrofauna of
three restored  wetland areas: a newly created system
in  the Northern  Wildlife Preserve  of Mission Bay
(unvegetated during the first 3 months);  a  5-year-old
restored Salicornia virginica marsh within Anaheim
Bay; and in 10-year-old restored Spartina foliosa and
S. virginica habitats within the Connector Marsh of
San Diego Bay.  In the 5- and  10-year-old systems,
total macrofaunal densities and species richness were
either  similar  or  were higher in the restored than
natural transects  (Table 1).  In all three of these sys-
tems oligochaetes were the most abundant taxon. The
oligochaete  Paranais litoralis  was   a   dominant
colonizer of the restored  marshes and in some cases
was significantly more  abundant  or  proportionally
better  represented in the  restored than  reference
marshes. At high densities,  adults of  this  species
swim in the water column, providing an effective dis-
persal  mechanism. In  sharp contrast, poorly dispers-
ing tubificid and enchytraeid oligochaetes dominated
natural sediments in most of the Pacific and Atlantic
coast  marshes studied.  5.  benedicti  and Capitella
spp., both  dominant early colonizers in North Caro-
lina, were  present but relatively rare in both created
and natural habitats studied in southern California. In
southern California, the  young Mission  Bay restored
marsh   was   initially   colonized   by   peracarid

                                                                                                                                      An Infaunal Perspective
Table 1. Macrofauna (>300 microns) and oligochaete representation in restored (Rest.) and natural (Nat.) salt marshes of southern California. Restored systems in Paradise
Creek (San Diego Bay), Anaheim Bay, and Mission Bay were 10 yr., 5 yr., and 3 mo. of age, respectively, at the time of sampling. Natural marsh data from Mission Bay
March 1996  have not  yet been processed. P values indicate t tests comparing restored and  natural values within a system. Percent data were  arcsin-transformed for
statistical analyses, then back-transformed for presentation.

Paradise Creek
San Diego Bay (Feb. 1994)
Spartina foliosa transect
Salicornia virginica transect
n = 10 cores
Anaheim Bay (Feb. 1994)
Salicornia virginica transect
n = 10 cores
Mission Bay (Mr. 1996)
Creekbed (unvegetated)
n = 2 cores
Macrofaunal Density
no. per 18.05 cm2 core
Rest. Nat. P
231.2 310.0 NS
159.5 47.0 0.049
211.1 59.9 NS
147 No Data
P. litoralis Density
no. per 18.05 cm2 core
Rest. Nat. P
119.0 99.3 NS
89.6 16.5 0.026
142.9 23.6 0.087
79.5 No Data
% Oligochaetes
Rest. Nat. P
95.4 96.5 NS
93.1 79.4 NS
33.1 86.7 0.068
54. 1 No Data
% P. litoralis
Rest. Nat. P
60.8 28.2 0.002
46.4 42.1 NS
31.9 28.6 NS
54. 1 No Data

 Batiquitos Lagoon Enhancement
 crustaceans   (amphipods  and  tanaids-all  direct
 developers) and annelids that rafted into the system
 on drift plant material.

 Conclusions and Applications
 1.   Organic amendments such as peat or alfalfa can
     retard macrofaunal colonization  initially, but
     should be used if they enhance Spartina growth,
     since  no  adverse  long-term  effects  were
     observed. This  may  not  be   true of  straw.
     Leaching of organic amendments prior to their
     application may ameliorate  the  early inhibition
     of infaunal recruitment.
 2.   Our results  demonstrate that  although  faunal
     densities and species  richness  measures in a
     created marsh can  resemble those of a natural
     system fairly  quickly,  these  similarities  can
     belie compositional differences  that may be  of
     consequence.  In  North   Carolina,  restored
     marshes  exhibited  a  notable  absence   of
     oligochaetes,  while in  southern  California,
     restored marshes exhibited an  unusually high
     representation of a naidid  oligochaete species
     with  high  adult  dispersal  ability. Although
     oligochaetes  comprise  over   3/4  of  the
     macrofauna in most salt marshes,  too little  is
     known  about their functional  roles to  assess
     effects  of  reduced or elevated  oligochaete

 3.   Results   from  both   North   Carolina   and
     California  suggest  that   species'  dispersal
     abilities    affect    macrofaunal   succession.
     However,  planktotrophic  recruits  dominated
     newly  created  marshes  of North Carolina,
     whereas  adult  dispersers appear  to dominate
     those  of California.  In  southern  California,
     where   wetlands   habitats   are  isolated   by
    stretches  of   high-energy  coastline,   adult
    dispersal may be  more effective  than  that of
    long-lived planktonic  larvae. On both coasts
    species lacking effective dispersal mechanisms
    were slowest to recover (Moy and Levin, 1991;
    Levin et al., in prep.).  These  taxa  could  be
    stimulated by 'seeding' restoration sites  with
    adults. Construction of restored  systems that
    directly  border  natural  marshes also  may
    facilitate recruitment by these species.
4.  From the studies conducted, we do not know
    how  long  full  recovery  of  macrofaunal
    communities will take  in restored salt marshes.
    Recovery seems to be both marsh- and taxon-
    specific, with time alone being no guarantee of
    functional equivalence (Sacco et al., 1994). We
    suspect  that the time frame will match that
    required for  restored marsh sediment properties
    (organic  matter,  water content,  and particle
    size) to approach natural levels.

Literature Cited
Levin, L. A., Talley, D. M., Thayer, G. (in press).
  Succession of macrobenthos  in  a  created salt
  marsh. Marine Ecology Progress Series.
Moy, L.  D.,  Levin,  L. A.  1991.  Are  Spartina
  marshes  a replaceable  resource?  A functional
  approach to evaluation of marsh creation  efforts.
  Estuaries 14: 1-16.

Sacco, J., Seneca, E., Wentworth T.  1994. Infaunal
  community development of artificially established
  salt marshes   in North  Carolina.  Estuaries 17:


                           The Sum  Exceedance Value as a Measure
                                        of Wetland Vegetation Hydrologic
                                                                         Scott D. Simon12
                                                                      Martha E. Cardona3
                                                                            Brian W. Wilm1
                                                                          James A. Miner4
                                                                        Douglas T. Shaw5
 Abstract: Although qualitative information exists
on wetland community  hydropatterns  and wet-
land species hydrologic  ranges, it has not been
adequately quantified in  the Midwest. The lack of
quantitative baseline information hampers cur-
rent efforts to restore, create, and manage wet-
lands in the Midwest. To quantify wetland species
occurrence along a hydrologic gradient, we used a
method called the sum  exceedance value (SEV).
The SEV is an integration of the wetland hydro-
pattern that incorporates the magnitude, timing,
and duration of surface  flooding and soil satura-
tion within the growing  season. We installed and
monitored staff gauges and groundwater monitor-
ing well transects within wetland communities of
several high quality Chicago-area natural areas
(Grass Lake  Marsh, Lake County;  Wadsworth
Prairie, Lake County; and West Chicago Prairie,
DuPage County)  to obtain  wetland  community
hydropatterns. We  obtained  plant species loca-
tions, corresponding cover classes, and plot eleva-
tions during yearly July vegetation sampling and
surveying.  Hydropatterns   and  corresponding
SEVs were determined  for  all vegetation plots
based on  bi-weekly water levels  and vegetation
plot elevations. A range of SEVs was computed for
each species  based on  elevational  abundance
within  each study area.  Monitoring results indi-
cate that wetland species distribution is strongly
correlated with elevation and the sum exceedance
value. From these results, we present  mean SEVs
for wet prairie, sedge meadow, shallow marsh and
deep marsh  communities. We also present pre-
liminary preferred, tolerated, and avoided SEV
ranges for all species analyzed for each wetland
community type at these three sites. This research
has many applications  in  wetland management,
restoration, and creation. The methods and data
can be  used for  selected  community or species
management, differentiating community and hy-
drologic disturbance from  natural dynamics, and
identifying target communities, target hydropat-
terns, design criteria, planting locations, and per-
formance  standards  for   wetland  restoration/
creation projects.


 Hydrology is a critical component of all wetland
ecosystems (Mitch  and Gosselink  1993).  Although
many environmental factors (light, soils, temperature,
etc.) influence vegetation distribution, one  dimen-
sional environmental gradients are  often seen as an
oversimplification of many environmental effects on
species distribution (Curtis and Mclntosh 1951). Hy-
drology, though, is an extremely strong and possibly
overriding selection pressure influencing the distri-
bution of vegetation at the wet end of the soil mois-
ture continuum (Mitch and Gosselink 1993;  Spence
1982). Because of the strong influence inundation ex-
erts on the plant community, plant  species are often
stratified along a  hydrologic gradient based upon
species' adaptations to particular  hydropatterns  or
1 Illinois Natural History Survey, 607 E. Peabody Drive, Champaign, Illinois 61820.
2 Correspondence Address: The Nature Conservancy, 601 North University Avenue, Little Rock, Arkansas 72205.
3 University of Illinois Department of Civil Engineering, 205 N. Mathews, Urbana, Illinois 61801.
 Illinois State Geological Survey, 615 E. Peabody Drive, Champaign, Illinois 61820.
5 South Florida Water Management District, 3301 Gun Club Road, West Palm Beach, Florida 33416.

 Sum Exccedance Value
 inundation  regimes (Johnson  1987; Shipley  1991;
 Zimmerman 1987). Temporal and spatial variation in
 hydrology is key to the spatial  and seasonal variabil-
 ity in the vegetation distribution of our wetland com-
 munities (Gosselink and Turner 1978; Niering 1987a;
 Niering 1987b; Weller 1987).
  Several terms are used  to  describe wetland hydro-
 logic signatures. For clarification purposes in this pa-
 per we will use the term hydropattern to describe the
 yearly pattern of water levels within a wetland, simi-
 lar to Mitch and Gosselink's (1993) definition of the
 hydroperiod. We will define the  hydroperiod  instead
 as the  period of inundation  within a particular wet-
 land. The hydropattern is  the "hydrologic signature of
 each wetland type" and strongly influences the type
 of wetland community, its diversity, productivity, and
 distribution of associated plants  and animals (Mitch
 and Gosselink  1993).  Typical marsh hydropatterns
 differ from typical deep  water swamp hydropatterns
 in shape and flooding depth and duration. The hydro-
 pattern within one part of a wetland can differ from
 the hydropattern  in another portion of the  wetland
 based on topography, variation in hydrologic inputs
 and outputs, etc (Zimmerman 1987). A wetland's hy-
 dropattern can also vary  temporally, based on yearly
 differences in precipitation  patterns (Gosselink and
 Turner 1978; van der Valk 1989;  Doss 1993).
   The  hydropattern variation  between and  within
 communities presents difficulties in managing, restor-
 ing, and creating wetland communities. For the wet-
 land manager, differentiating between natural rainfall
 driven  vegetation  dynamics and vegetation  change
 caused by an unnatural hydrologic disturbance is dif-
 ficult without  knowledge of the sites'  hydropattern,
 precipitation dynamics, and wetland vegetation hy-
 drologic preferences. For the  wetland restorationist,
 identifying planting locations is difficult  without
 knowledge of the target community's hydropatterns,
 seasonal  water elevations,  basin  morphology,  and
 wetland vegetation hydrologic preferences. Identify-
 ing the target community's hydropattern and wetland
 vegetation occurrence along the  hydrologic gradient
 is critical for successful  vegetation establishment in
 restoration and creation projects.  Because the success
 of these projects is often dependent on the vegetation,
 successful wetland vegetation  establishment is criti-
 cal.  Wetland vegetation  is typically  planted  with
 minimal information on species'  preferred hydropat-
 terns and water depths. The wetland project manager
 has to make management, planting, construction, or
design decisions without adequate information on the
hydropatterns and vegetation hydrologic tolerance of
natural wetland communities. Consequently, most of
these projects are prone to failure (Kentula  et al.

  Although experienced wetland managers and natu-
ralists are familiar with species' qualitative hydropat-
tern   and  seasonal   water  depth  tolerances, this
information has not been adequately quantified, ham-
pering wetland protection, restoration, and manage-
ment efforts (Maurizi and Poillon 1992; Galatowitch
and van der Valk 1994). Species tolerance to selected
features of the hydropattern have been included in the
literature (Nelson and Anderson 1983; Gunderson et
al.  1986; Picket and Bazzaz  1978;  Robel  1962;
Squires and van  der Valk  1992;  van der Valk et al.
1994; Wentz et  al. 1974, and Whitlow  1979; Thun-
horst 1993). Examples of these features include peak
water levels, mean water  levels,  flooding  duration,
the rate of drawdown,  low  water levels,  drainage
class etc. The information provided on species'  hy-
drologic tolerances  includes anecdotal water level
and flooding evidence, greenhouse studies with pos-
sible applications to  natural systems, and  detailed
statistical gradient analysis not  readily usable for
wetland  managers  and  restorationists. Typically,
these studies describe vegetation  response  to  depth,
duration, or timing of inundation or saturation inde-
pendently and not their combination. The information
does  not facilitate easy application or comparison
between  sites and   studies. Additionally,  reliable
measurements of various  wetland hydropattern fea-
tures (water depth, duration of flooding, rate of draw-
down, timing of drawdown,  etc.) for high-quality
communities have not been  accurately quantified. In
instances where  some knowledge  is available, these
disparate measurements  of hydropattern features do
not adequately represent the hydropattern itself and
do not allow comparisons  of its influence on species
distribution within or between sites and communities.

  We have been conducting a baseline study of the
hydrology,  vegetation, and vegetation response to
natural yearly hydropattern changes in several Illinois
wetland communities. We are attempting to quantify
the hydropattern and wetland vegetation distribution
in response to the hydropattern. In our early attempts
to quantify the hydropattern, we  used several meas-
urements including peak water level, low water level,
depth of flooding,  duration of flooding,  timing of
drawdown, and  rate of drawdown. Unfortunately, we

                                                                                   Sum Exceedance Vain
felt that these measurements did not adequately char-
acterize the hydrologic influence on the plant com-
munity  given  the  variation  in  the  hydrologic
components (depth, duration, timing, etc.) from one
year to the next. We needed an increased level of pre-
cision that included a comprehensive measure of the
hydropattern.  Following work conducted by Gowing
and Spoor (In Press) and Gowing, Spoor, and Mount-
ford (In Press) in the United Kingdom, we  are  ana-
lyzing  an additional  measure  of the  wetland hy-
dropattern, the sum exceedance value  (SEV), in an
attempt to better characterize vegetation response to
hydrologic change. The SEV  is an integrated feature
of the  hydropattern that incorporates the magnitude,
timing, and duration of surface flooding or soil satu-
ration (Spoor 1993).

The Sum Exceedance Value
  The SEV (Figure 1)  is  a simple, quantifiable pa-
rameter that integrates water table depth and duration
of flooding/saturation over a growing season, similar
in concept to the "degree-day" measure  used  to assess
temperature stress:
 'n' is the number of days present in the growing sea-
 son. 'SEV* is the Sum Exceedance Value Index for
 an entire growing season, 'd  ' is a daily value of the
 elevation of the water above a datum located 30 cm
 below the soil surface and is defined to be zero when
 the water table is  below  the datum. We selected
 30 cm as an assumed average wetland plant rooting
 depth (National Research Council 1995). The SEV is
 a valuable index for quantifying the hydropattern be-
 cause the SEV describes how much and for how  long
 water depths are present in a given soil column. For
 example, using  a 30 cm datum (0.30 m), if the water
 were 5 cm (0.05 m) below the soil surface every day
 for a  200 day growing  season,  SEV  would be
 50 m.day, or (0.30 - 0.05 m) * 200 days. In comparing
 SEV values for  two different wetlands (or  two differ-
 ent locations in the  same wetland) over growing sea-
 sons of identical length, high values indicate greater
 wetness. SEV values can be easily calculated at any
 ground  surface elevation  within the wetland from
 water table data using a spreadsheet if the elevation
 and appropriate wells and staff gauges are surveyed.
 From species occurrence and elevation data, an  SEV
range can be calculated for a particular species at a

 Field studies with agricultural crops,  and more re-
cently with wetland plants, have documented strong
relationships  between the soil moisture indices like
the SEV and plant productivity and abundance. Soil
moisture  indices have been  used successfully in
agronomy  to assess   water  stress and  associated
changes in productivity in corn, soybeans, grain sor-
ghum, and other crops (Shaw 1974;  Sudar et. al.
1979; Howell et al. 1976; Hardjoamidjojo et al. 1990;
Evans et  al.  1990). Recent  work by Spoor  (1993),
Gowing and  Spoor (In Press), and Gowing, Spoor,
and Mountford (In Press) demonstrated strong rela-
tionships between the SEV and wetland plant distri-
bution in United Kingdom wet grasslands.
 Although the SEV reduces a site's growing season
hydropattern  to one number, this number incorporates
something many of the other measurements  do not:
the   depth,  duration  and  timing  of   inunda-
tion/saturation combined. For a particular species, re-
searchers  and managers  can compare  species SEV
distributions  between  years,  compare  species SEV
distributions  between  sites,  predict vegetation  re-
sponses to hydrologic change, and determine the ef-
fects  of  modeled  hydropatterns  on  vegetation
distribution for wetland mitigation projects (Konhya
et al.  1995).

 We conducted field work and data analysis  to iden-
tify and  quantify  hydropatterns  of  Illinois' native
wetland  communities  and plant species. Hydropat-
terns  were quantified using the SEV from site eleva-
tion and water levels. Methods include  site selection,
hydrologic monitoring, surveying,  vegetation sam-
pling, SEV calculations, and identifying species' hy-
drologic preferences.

Site Selection
 We randomly selected possible study  sites from the
state  list of natural areas as identified  in the Illinois
Natural Area Inventory (White 1978). Such sites pro-
vide the best information on natural wetland  commu-
nities in Illinois because the plant communities are
relatively intact and  human  disturbance  has  been
relatively light. Over fifty high quality wetland sites
of various types were initially identified from the Illi-
nois Natural  Areas Inventory (White, 1978).  This list

                                                                       Ground Elevation
                                                                     Ground Elevation - 30 cm
                             I   I    I   I    I    I   I    I   I    I

       Figure 1.  Representation of sum exceedance value derived from hydropattern, growing
                  season, and 30 cm below ground elevation.
                                     West Chicago Prairie Water Elevations 1990 -1995
                  1   I
                  s   s
s   £   S   i   5   S
                                            .Deep Marsh Well (gmd. elev. = 98.366 m)
                                    Grass Lake Marsh Water Elevations 1990-1994
                        I    I
                                          . Shallow Marsh Plot (grnd etev. = 97.540 m)
       Figure 2.  West Chicago Prairie and Grass Lake Marsh hydrographs from water level

                                                                                   Sum Exceedance Vak
was screened for sites with minimal hydrologic dis-
turbance and, preferably, existing hydrologic data or
hydrologic monitoring capability (e.g., existing wells
or staff gauges). Ten sites on the short list were ran-
domly selected and visited to determine feasibility for
field sampling. Field surveys were made of each site
to obtain baseline information on soils, plant commu-
nity  composition  and diversity, general hydrology,
and other important features.  Additional information
on plant communities, soils, ownership, disturbance
history, and management activities was obtained from
soil surveys,  topographic  maps, reports,  the Illinois
Natural Areas Inventory, and  other sources.  Of the
sites included within the study,  three will be  dis-
cussed in this paper.

Study Sites
  All three study sites described in this paper (Grass
Lake Marsh Natural Area, Wadsworth Prairie Nature
Preserve, and West Chicago Prairie Natural Area) are
located within Illinois' Northeastern Morainal Divi-
sion (Schwegmann 1973). Grass Lake Marsh Natural
Area (Lake County) consists of shallow and  deep
water marsh  communities as  described  by White
(1978) and Ebinger et al. (1977). The shallow marsh
community is dominated by Sparganium eurycarpum
(bur-reed),  Scirpus fluviatilis  (river  bulrush),  and
Car ex lacustris (lake sedge). The deep marsh com-
munity is dominated by Typha  spp. (cattail)  and
Sparganium  eurycarpum. The site  is adjacent to
Grass Lake of the Chain O' Lakes, with water levels
influenced by Grass Lake and the adjacent Fox River.
  Wadsworth  Prairie  Nature Preserve  (Lake County)
is located near the Des Plaines River and consists of a
mosaic of  marsh,  sedge meadow,  and wet  prairie
communities  as  described  by  White  (1978)  and
Nuzzo (1994). Previous  hydropattem  and  vegetation
information was  documented by  Bremholm (1993).
The  marsh  community  is  dominated by Acorus
americanus (sweet  flag), Sparganium eurycarpum,
Scirpus fluviatilis,  Car ex  lacustris, and Typha  spp.
The sedge community is dominated by Carex stricta
(tussock sedge). The  wet prairie community is domi-
nated by Carex stricta and Calamagrostis canadensis
(bluejoint grass).

  West Chicago Prairie (DuPage County) is located at
the headwaters of Salt Creek and also consists  of a
mosaic of marsh,  sedge meadow,  and wet  prairie
wetland communities as described by White  (1978)
and Lampa (1993). The marsh community is domi-
nated by  Scirpus acutus (hard-stem bulrush)  and
Ludwigia  palustris  (marsh  purslane).  The  sedge
community is dominated by Carex stricta. The wet
prairie community is dominated by Carex stricta and
Calamagrostis  canadensis.  Water table, soils  and
plant community relationships  have  been described
by Kelsey and Hootman (1992).

Hydrologic Monitoring
  Hydrologic data for  Grass Lake Marsh was pro-
vided from  gauges  at  the Fox  River and Chain  O'
Lakes maintained by the United States  Geological
Survey  and  Illinois Department of Transportation
(IDOT 1970-1984; USGS 1995). Specifically, water
level data for the vegetation transects was generated
by  correlating current  water level  data from the
nearby station at Lake Villa (U.S. Geological Survey,
1994) with data available from the Fox River, located
at the edge of the marsh. The correlation has a root
mean square of 0.94. Hydrologic data for West Chi-
cago Prairie was provided  from monitoring wells
maintained by Pat Kelsey at The Morton Arboretum
(Lampa 1992, Kelsey 1995, and Kelsey and Hootman
1992). Water level data for the vegetation transects at
this site was taken from existing well transects.
  At Wadsworth  Prairie, we installed groundwater
monitoring wells and  staff  gauges  along randomly
selected transects parallel to the water level gradient.
Each transect consisted of 5  to 7 shallow wells, staff
gauges in open water  areas, and two deep wells at
either end of the transect to provide information on
site  stratigraphy and underlying aquifers. Each tran-
sect begins  topographically  higher and  ends  lower,
crossing several  plant communities. One  shallow
groundwater monitoring  well was placed within the
center of  each community.  Wells were constructed
from one-inch diameter PVC pipe with a 30 cm long
screen at  the bottom, inserted in  a hand-augured
borehole lined with sand and capped with bentonite.
Wells were permanently capped at the bottom and
temporarily  capped at the top to prevent  rainfall en-
try. All  of the shallow groundwater monitoring wells
were made of the same  materials and placed at the
same depth. Wells were constructed  uniformly to en-
sure that differences in water level readings between
any  two wells were actually differences  in the water
table, and  not caused by differences in well construc-
tion. Because all of the study sites were State Nature
Preserves or Natural Areas, wells were installed dur-
ing the  non-growing season (March) and soil cores
were discarded off site to minimize siltation within

 Sum Exceedance Value
 the wetlands. Wells were surveyed and purged 2 to
 3 weeks after installation. Water levels at staff gauges
 and wells were recorded at least bi-weekly.

 Vegetation Sampling
  Herbaceous vegetation sampling was conducted in
 July  of each year along permanent,  randomly estab-
 lished transects parallel to the well transects. Vegeta-
 tion  was sampled  according to the  quadrat-transect
 method  (Mueller-Dombois  and  Ellenberg   1974).
 Plant species were identified and corresponding cover
 was  collected from  l/4m2  plots spaced  every  two
 meters along the transects. Species  cover was then
 converted to midpoints based upon corresponding
 cover  classes  (Table 1) from (Daubenmire  1959).
 Plant  species nomenclature  follows Mohlenbrock
 Table 1. Cover classes, cover class ranges, and range
 midpoints for vegetation plots.
Cover Class
Cover Class
Range Midpoint
  Elevations of vegetation  plots  and bench  marks
 were determined using a Sokkia C75 level, surveying
 tripod, and range poles. Elevations were taken from
 the center of the plot and read to the thousandth of a
 meter. In plots with several distinct elevations (logs
 or sedge tussocks for example), readings were taken
 at each elevation with  vegetation. All species were
 then assigned a cover class for each elevation.

 Calculating the SEV
  The SEV is a numerical representation of the hy-
dropattern  of any given  elevation within the wetland.
Hydropatterns were generated from shallow ground-
water monitoring  wells  and staff gauges for each
vegetation  community or  zone.  Hydropatterns for
each vegetation plot were modeled from  the hydro-
patterns of the nearest wells or staff gauges and the
elevation of that particular vegetation plot. Therefore,
a hydropattern was  assigned to  the  species within
each plot. By  integrating the hydropattern within the
growing season to 30 cm below the soil surface, an
SEV was calculated for each plot, and, therefore, for
each species in each plot. The  growing  season for
these sites was selected as the time between the last
28° F freeze (50 percent probability) in spring and
first 28°F freeze (50 percent probability) in the fall.
Of the three sites, West Chicago Prairie has the long-
est growing season, starting on April 12th and ending
on October 23rd (USDA 1979). Grass Lake Marsh and
Wadsworth  Prairie have shorter growing  seasons,
starting on April  23rd and  ending on October 28th
(USDA 1970).

Identifying  Species' Preferred Hydrologic
  Although a species could occur over a wide range
of SEVs at a particular site, we attempted to identify
the SEVs where a species was particularly abundant
(high coverage) or rare (low coverage) within a sub-
set  of its entire SEV range.  We anthropomorphically
assigned a species SEV range with high coverage its
"preferred" range, medium coverage its  "tolerated"
range,  and  low  or absent  coverage its "avoided
range." Identifying these three ranges for each species
based on coverage required some data manipulation.
First, cover data was plotted versus SEV classes for
each species  at  each  site.  The observed  coverages
were divided into small, 2 m.day SEV intervals and a
frequency histogram of species coverage within each
interval was calculated. Second, to identify a species'
preferred, tolerated, and  avoided  ranges at a particu-
lar  site, we standardized the observed coverage for
each SEV range by calculating a ratio of the observed
coverage over a hypothetical uniform coverage (as if
the species was uniformly distributed throughout the
site irrespective of SEVs). A species uniform cover-
age was calculated as its total coverage divided by the
number of SEV classes. Third, confidence intervals
(95 percent) for the coverage ratio mean were identi-
fied for each species. Fourth, SEVs were categorized
as "preferred, "tolerated," or "avoided" based on the
location of the coverage  ratio in relation to the confi-
dence intervals. SEVs for coverages above the upper

                                                                                  Sum Exceedance Valu«
confidence limit were designated as the "preferred"
SEV range. SEVs for coverages within the 95 percent
confidence limits were designated as the "tolerated"
SEV range. SEVs for coverages below the lower con-
fidence limit were designated as the "avoided" SEV
 To determine whether the observed coverage distri-
bution along the SEV gradient was significantly dif-
ferent from a hypothetical uniform distribution,  the
null hypothesis HQ: "Species coverage follows a uni-
form distribution along the SEV gradient" was tested
against the hypothesis H,: "Species coverage distri-
bution along the SEV gradient is different from a uni-
form distribution."  The Kolmogorov-Smirnov non-
parametric test of distribution was used to compare
the observed versus the uniform distributions at the
0.01 level. The test statistic is the maximum deviation
of the observed coverage distribution versus the uni-
form  coverage  distribution. Significant  deviation
from the uniform coverage distribution are indicative
of the hydrologic preferences explained above.

Results and Discussion
  The results of our monitoring and associated discus-
sion include preliminary community hydropatterns,
preliminary  community SEVs, preliminary species
SEVs and some notes on SEV comparisons between
sites and years.

  Long-term water level elevations for  marsh com-
munities at  West Chicago Prairie and  Grass Lake
Marsh are provided in Figure 2. The years 1990  and
1993  appear to  have higher growing season  water
levels at both sites.  At Grass Lake Marsh, 1993 was
the wettest  year, with an SEV of  117 m.day.  Other
years between 1990 and 1994 at Grass  Lake Marsh
showed very little variation in SEV, ranging from 87
to 96 m.day. At West Chicago Prairie, 1993 was also
the wettest year, with a SEV of 164 m.day. Both 1991
and 1992 were the driest years during this period at
West Chicago  Prairie,  with  SEVs   of  79  and
53 m.days, respectively. West Chicago Prairie hydro-
patterns include a marked drawdown during the  late
summer of most  years. This drawdown is most pro-
nounced during drier years, as expected.  Such  late
summer drawdowns do not appear to occur at  Grass
Lake Marsh. Lower water levels  occur during the
winter in most years. This seemingly unnatural  hy-
dropattern is possibly caused  by maintenance  of
artificially high water levels on the Fox River by the
Corps of Engineers (presumably for summer naviga-
tion). The water levels are drawn down during the
winter, presumably  in preparation for spring  rains.
Although this hydropattern is different from the other
sites and probably modified, the vegetation at the site
must still adapt to the timing, depth, and duration of
flooding. Because the Grass Lake Marsh community
is of natural-area quality, we  decided to include the
site, even with its potentially  modified hydropattern.
Inclusion of this site  increases project hydropattern
variety, and  allows  species  SEV comparisons be-
tween different hydropatterns.
 Hydropatterns  and  estimated SEVs  for wetland
communities  at Wadsworth Prairie 1995, West Chi-
cago Prairie 1994, and West Chicago Prairie 1995 are
provided in Figure 3. The water levels are relative to
the soil surface elevation and 30 cm below the soil
surface. Differences in the hydropatterns between the
communities  are discernible. As expected, the marsh
communities have the highest water levels throughout
the year, followed by the sedge meadow and wet prai-
rie communities. Water level differences between the
wetland communities  are relatively small. For West
Chicago  Prairie specifically, water level  elevation
differences ranged from  12 to 18cm between wet
prairie (WP)  and sedge meadow (SM), 13 to  17cm
between  sedge  meadow and  shallow marsh (SMA),
and 6 to 23  cm between shallow marsh and  deep
marsh (DMA) throughout the year.

Community Sum Exceedance Values
  The relatively small differences between commu-
nity  hydropatterns  translate  into  correspondingly
larger differences in community SEVs because  these
water level differences are cumulative over the entire
growing season. Although there were site differences
in hydropatterns for each community, the community
SEV  ranges  appear well differentiated. Among the
sites, the mean  SEV30 for wet prairie was 20.30 ±
9.01 m.day,  for sedge meadow was 45.94 +  10.64
m.day, for  shallow marsh was 71.86  + 10.91 m.day,
and for deep marsh was 95.90 ± 8.88 m.day.
  Quantifying a community  hydropattern using the
SEV allows for comparison to other communities and
studies. For example, the mean SEV30for wet prairie
of 20.30 ± 9.01  m.day is comparable to  an SEV^ of
28m.day as  documented by Spoor  (1993)  from a
United Kingdom lowland wet grassland.

                                Average Water Depths and SEVio 's for Different Communities at
                                            Wadsworth Prairie -1995
                                            • SM SEV=36
                                                            - SMA SEV=56
                                                                             - DMA SEV*94
                            Average Water Depths and SEV M 's for Different Communities at West
                                              Chicago Prairie-1994
                                          - SM SEV= 45
                                                                       - DMA SEV= 88
                                 Average Water Depths and SEVto 's for Different
                                  Communities at West Chicago Prairie -1995
                  • SM SEV= 57
                                   - SMA SEV= 83
             Figure 3.  Water depths and sum exceedance values for wet prairie (WP), sedge meadow
                        (SM), shallow  marsh (SMA), and deep marsh (DMA) communities at
                        Wadsworth Prairie, West Chicago Prairie  1994, and West Chicago Prairie

                                                                                    Sum Exceedance Val
Individual Species' SEV Ranges
 SEV ranges were also computed for individual spe-
cies at the three sites. Using Helianthus grossoserra-
tus  (saw-tooth sunflower)  at West Chicago Prairie
1994 as an example (Figure 4), a ratio of observed
over expected coverage is plotted versus the SEV and
preferred, tolerated, and  avoided ranges were com-
piled in relation  to the  confidence interval as de-
scribed in the methods section. Based on the coverage
distribution we recorded for Helianthus grosseserra-
tus at West Chicago Prairie in 1994, we designated its
"preferred" SEV range as 1 to 28 m.day, "tolerated"
range as 28 to 38 m.day, and "avoided" range greater
than 38 m.day.  It is important to note that the  SEV
ranges we have designated as preferred, tolerated, and
avoided, are based on  this limited data and are pre-
liminary. These ranges are a starting point. As addi-
tional data from wetter and drier conditions is added
from additional sites and years, SEV ranges for  vari-
ous species are expected to change.
  Analyzing individual species' SEV ranges allows us
to make preliminary comparisons between many spe-
cies occurring along a theoretical or real hydrologic
gradient. Preliminary species'  SEV ranges based on
our field data are provided for species  at Grass Lake
Marsh  in 1994 (Figure 5), West Chicago Prairie in
1994  (Figure  6), West  Chicago  Prairie  in   1995
(Figure 7), and Wadsworth Prairie in 1995 (Figure 8).
The abrupt delineation between species ranges is not
meant to be indicative  of natural conditions, but was
used for illustrative purposes only. The line graph of
Helianthus  grosseserratus  in Figure 4 is condensed
into the bar graph for the same species in Figure 6.
  Evaluating each site individually, species  occur-
rence along the hydrologic gradients appear to make
biological sense.  As would be expected looking at
species from West Chicago Prairie 1994 (Figure 6),
Helianthus grosseserratus (a common species of wet
prairie communities) is located at the drier end of the
SEV gradient (< 38m.day) than species like Scirpus
acutus  (hard-stem bulrush),  a common species of
deep marsh communities, that occurred at  higher
SEVs (> 49 m.day).
 Within a single community, species' preferred  SEV
ranges also follows this pattern. For the marsh com-
munity, the order of species abundance  from the  drier
end to the wetter end of the SEV gradient is relatively
consistent between sites.  Deep water marsh plants
like Scirpus acutus are consistently located at higher
SEV ranges with average preferred SEV ranges of
66-120 m.day. On the other hand, species like Carex
lacustris (lake  sedge)  are located at  lower average
preferred  SEV  ranges with average preferred SEV
ranges of 27 to 58 m.day, among all three sites. Ad-
ditionally,  the  sequence of  species between these
SEV  endpoints  like Acorus  americanus,  Sagittaria
latifolia  (common  arrowhead),  Eleocharis  smallii
(marsh  spikerush),  and  Sparganium eurycarpum,
seems to  make biological sense, and  support previ-
ously  published community descriptions  (Weller
1987; Squires and van der Valk 1993).

  We used the  Kolmogorov-Smirnov non-parametric
test of distribution  (« = 0.01)  to test species'  ob-
served versus expected values to determine if a spe-
cies'  observed  distribution   is  different  than  an
expected uniform distribution. Based  on these tests,
almost all of the species were not uniformly  distrib-
uted throughout the site. Almost all  of the  species
were significantly more abundant at some SEVs than
others. These results suggest that the SEV method
can quantify a species location along a hydrologic
  Three species, though, failed to reject the test of
uniform distribution, Polygonum amphibium (water
smartweed), Scirpus fluviatilis, and Acorus america-
nus. Polygonum amphibium failed to reject the test at
Wadsworth Prairie  1995 and West Chicago Prairie
1994, while  it  did  reject the test at  West Chicago
Prairie 1995. This  disturbance-tolerant species rap-
idly germinates on moist soils. As water levels drop
over the growing season, this species  is able to ger-
minate, growing quickly at  a  range of  elevations
within a site, and therefore at a broad range of  soil
exceedance values.  Why, then, was this species not
found uniformly distributed at West Chicago Prairie
in 1995?  A  possible explanation includes variation
within the site's hydropattern. Water levels during
1995 (Figure 2) did not drop as much as the previous
year,  providing less habitat for this species to colo-
nize. In this  case, SEV does not adequately  explain
this species occurrence along a  hydrologic gradient.
The rate, depth, and timing of drawdown might better
explain its coverage distribution.
  Scirpus  fluviatilis and Acorus americanus  both
failed to reject  the null hypothesis of uniform distri-
bution at  Grass Lake  Marsh. Although Scirpus  flu-
viatilis  is somewhat  tolerant  to  disturbance  and
fluctuating water levels, Acorus americanus is not as
tolerant. A partial explanation to these species distri-
bution may exist in the  site's topography. Whereas

                                                         95% Confidence Intervals
                                                                               100         120
                      Mean = 1.364   Standard Deviation = 2.695  95% Confidence Interval = [0.659,2.070]
        Figure 4.  Standardized coverage versus sum exceedance values for Helianthus
                   grosseserratus at West Chicago Prairie 1994.
         Equisetum arvense L

         Polygonum amphibium L

         Carex lacustris Willd.

         Sagittaria latifolia Willd.

         Scirpus fluviatilis (Torr.) Gray.

         Acorus americanus (Raf.) Raf.

         Typha latifolia l.

         Lemna minor L

         Sparganium eurycarpum  L
                                                       40        60
                                                 Preferred • Tolerated  O  Avoided
        Figure 5.   Sum exceedance value ranges for selected species at Grass Lake Marsh 1994.

  Helianthus grossoserratus Martens.
  Veronicastrumvtrginicum(L.) Farw
  Solidago canadensis L
  Viola pratincola Greene
   vfrginianum (L) Dur & Jacks
  Euthamia grammifolia (L) Sahsb.
  Lysimachia quadrifloraStms
  Carex stricta Lam.
    eanadens/s (Michx.) Beauv
  Apocynum sibmcum Jacq.
 . PofygonumamphibiumL.
 . LycopusamencanusMuh\.
 : Galium obtusum Bigel.
  Stachys palustns L
 ^ Scutellaria galericulata L.
  Ludwigia palustns L
 , ScirpusacutusMuM
Figure 6.   Sum exceedance value ranges for selected species at West Chicago Prairie

             Agrostis alba L.
             Melilotus alba Medic.
             Oxypolis rigidior (L.) Coulter and Rose
             Achillea millefolium L
             Poa pratensis L
               virgintenum (L) Our & Jades
             Veron/castrvm virgin/cum (L) Fatw.  ,
             Viola pratincola Greene.:'i       . •  '..
             Hellanthus grosseserratus Martens.  :
             Poa compressa L   ::;•/
             Lysimactiia quadrifotta Sims:   ,:
             ScutellartagalericulataL.     :^
             Lycopus amaricanus Muhl.
             Lycopusvirginicusl.   :      ;; •
             Calamagmstis canadensis (Michx.) Beai
             Polygonumamptiibium L
             Carax lanuginosa Uichx.
             Sclrpus acutus Muhl.   L
             Beochaits smallliBritL  :'
             Slum suaveWaK.
             Polygonum hydropiper L
             Ludwigia palustris L
           Figure 7. Sum exceedance value ranges for selected species at West Chicago Prairie

Figure 8. Sum exceedance value ranges for selected species at Wadsworth Prairie 1995

 Sum Exceedance Value
 the marsh communities SEV ranges at the other sites
 were  approximately  40 to  110 m.day,  the  marsh
 community SEV range at Grass Lake Marsh was ap-
 proximately 20 to  85. The  site is exceptionally flat
 (0.25  meters of elevation change over  a 200 meter
 transect). Additionally, limitations in the range of our
 surveying equipment inhibited us from  sampling the
 full range of SEV values within the marsh communi-
 ties,  unlike  our  representative  sampling at West
 Chicago Prairie and Wadsworth Prairie.

   There are several issues we need to address as we
 continue  data analysis and field work. Obviously,
 these  results are preliminary, representing only a se-
 lect number of communities  and sites, over only a
 few years. To increase the  applicability of the data,
 we need to  increase community, geographic,  and
 temporal  coverage. Recording  seedlings (as we did),
 in addition to mature individuals during vegetation
 sampling, may  not provide  the clearest picture of a
 species' "preferred" SEV range. Including seedlings
 (that may not survive the entire growing season) in-
 creases a species' observed SEV range. Additionally
 for some species, the SEV for the entire growing sea-
 son may  not explain species distributions as  well as
 an SEV of the first half of the growing season, the
 previous years SEVs, etc.
   Based on our preliminary  results, the  SEV appears
 to be a usable  method to quantify the hydropattern
 and species and community occurrence along a hy-
 drologic gradient. Sum exceedance values for the wet
 community  averaged  20.90 m.day,  sedge meadow
 community  averaged  45.94 m.day,  shallow  marsh
 community averaged 71.86 m.day, and deep marsh
 community averaged 95.90 m.day.  Sum exceedance
 values for  species  along  the hydrologic gradients
 measured appear to  reflect their intuitive  location
 along a hydrologic gradient.

  Regardless, we feel  that in  addition to the impor-
 tance  of documenting the vegetation and hydrology
 for our remaining high-quality wetland communities,
 this work has  immediate  practical  applications in
 wetland management, restoration, and  creation. For
 wetland restoration and creation  projects, a baseline
 wetland vegetation data set is valuable for identifying
 target communities, target hydropatterns, and  species
 planting locations (combined  with  site  hydropattern
 information) with respect to  water levels, at any time
 for both wet and dry years. The data can also be used
when establishing project  design criteria  and per-
formance standards. Additionally, quantitative infor-
mation on  species' hydrologic  preferences provides
wetland managers an additional tool in managing for
target species and communities, and  discerning be-
tween  natural  wetland  community  dynamics and
long-term, disturbance-caused, hydrologic change.

 We would like to thank the University of Illinois
Water  Resources  Center,  Illinois Department  of
Transportation, and University of Illinois Agricultural
Experiment Station for providing financial support
for this project. We would also like to thank Jim
Anderson  and Ken Klick (Lake County  Forest Pre-
serve District) for site access and vegetation sampling
assistance, Wayne Lampa (Forest Preserve District of
DuPage County) for site access and vegetation sam-
pling assistance, Pat Kelsey (Morton Arboretum) for
providing   West Chicago  Prairie hydrologic data,
Gary Balding (United States  Geological Survey) for
providing Grass Lake Marsh hydrologic data, Randy
Heidorn (Illinois Nature Preserves Commission) for
permission to conduct  research in the Nature Pre-
serves, and Richard Warner  (University of Illinois)
for support and suggestions throughout the project.
We would  also like to thank volunteers and field as-
sistants including  Christina Bolas,  Michelle  Foster,
Yvonne Marlin, Stacy  O'Leary, Cory Rubin, Amy
Seiler, Jeff Stillman, and Jen Verbancic.

Literature Cited
Bremholm, T. 1993. Evaluation of techniques for es-
  tablishment of  sedge meadow vegetation.  Master
  Thesis. Iowa State University. 93 pp.
Curtis, J.  T. and R. P. Mclntosh. 1951. An upland
  forest continuum in the prairie-forest border region
  of Wisconsin. Ecology 32:479-496.
Doss, P. K., 1993. The nature  of a  dynamic water
  table in  a  system of non-tidal, freshwater coastal
  wetlands. J. Hydrology 141:107-126.

Daubenmire, R. 1959. A canopy coverage method of
  vegetational analysis.  Northwest Science 33:43-64.

Ebinger, J. E., R. D. Andrews, H. C. Nilsen.  1977.
  Papers on  the ecology of the Chain O'  Lakes re-
  gion, Illinois. Eastern Illinois University. Charles-
  ton, IL.

Evans, R. O., R. W. Skaggs and R. E. Sneed.  1990.
  Normalized crop susceptibility factors for com and
  soybean to excess water stress. Transactions of the
  American   Society   of  Agricultural   Engineers
  33(4): 1153-1161.
Galatowitch, S.  M. and A. G.  van der Valk.  1994.
  Restoring prairie wetlands: an ecological approach.
  Iowa State University Press. 246 pp.
Gosselink, J. G. and R. E. Turner. 1978. The role of
  hydrology  in freshwater  wetland   ecosystems.
  pp. 63-78, In R. E. Good, D. F. Whigham and R. L.
  Simpson  (ed). Freshwater Wetlands:  Ecological
  Processes  and Management  Potential.  Academic
  Press, Rutgers University, New Brunswick, NJ.
Gowing, D. J. G., Spoor, G. and Mountford, J. O. In
  press. The influence of minor variations in  hydro-
  logical regime on plant communities - implications
  for water management. In C. Joyce and M. Wade
  (eds). European Lowland  Wet  Grasslands:  Biodi-
  versity, Management and Restoration. International
  Center for Landscape Ecology.  Loughborough,
Gowing, D. J. G. and Spoor, G. In press. The effect of
  water table depth on  the distribution  of plant spe-
  cies on lowland wet grassland.  In B.  Sherwood P.
  Jose, and R. Bailey (eds.) United Kingdom  Flood-
  plains. 1995.  The   Linnaean  Society.  London,
Gunderson, L. H., J. R. Stenberg and A. K. Hemdon.
  1986. Tolerance  of  five  hardwood  species  to
  flooding  regimes, pp.  119-133, In R.  Herrmann
  (ed). Interdisciplinary  Approaches to  Freshwater
  Wetlands. Michigan State University Press.
Hardjoamidjojo, S., R. W. Skaggs and G. O. Schwab.
  1990. Corn yield response to excessive soil water
  conditions. Transactions of the American  Society
  of Agricultural Engineers Vol. 33.
Howell, T. A., E. A. Hiler, O. Zokezzi and R. J. Rav-
  ello. 1976. Grain sorghum response to inundation at
  three growth  stages.  Trans. American  Society of
  Agricultural Engineers, 19(5):876-880.
Illinois  Department of Transportation.  1970-1984.
  Division  of Water  Resources.  Water  Resources
  Data for Illinois: water years 1970- 1984.
Johnson, W. C., T. L. Sharik, R. A. Mayes, and E. P.
  Smith.  1987.  Nature and cause of zonation dis-
  creteness  around glacial prairie marshes. Canadian
  Journal of Botany 65: 1622-1632.
Kelsey, P. D.  and R. G. Hootman.  1992. Relation-
  ships between water tables, plant communities, and
  hydric soils: West Chicago Prairie, West Chicago,
  IL. Soil Survey Horizons 33(3):53-58.

Kelsey, P. D.  1995. Unpublished data on West Chi-
  cago Prairie water tables.

Kentula, M. E., R. P. Brooks, S. E. Gwin, C. C. Hol-
  land, A. D.  Sherman and S. J. C.  1992. An  Ap-
  proach to Improving Decision Making in Wetland
  Restoration and Creation. U.S. Environmental Pro-
  tection Agency. EPA/600/R-92/150.

Koyha, K. D., D. T.  Shaw and K. W. Weiler. 1995.
  Hydrologic  design  of  a  wetland:  advantages of
  continuous   modeling.  Ecological   Engineering
Kusler, J. A.,  and G. Brooks (eds.). 1987. Proceed-
  ings of National Wetlands Symposium: Wetland
  Hydrology. Omnipress, Chicago, IL.
Lampa, W.  1992. Results of a three year study of
  wetlands at the West Chicago Prairie. Unpublished
  report. 16 PP + appendices.
Maurizi, S., and F.  Poillon.  1992. Restoration of
  aquatic  ecosystems.  National  Academy  Press,
  Washington, DC. 732 pp.
Mitsch, W. and  J. Gosselink.  1993. Wetlands.  Van
  Nostrand Reinhold. NY. 722 pp.
Mohlenbrock,  R. H.  1986.  Guide to the Vascular
  Flora of Illinois. Southern Illinois University Press.
  Carbondale.  507 pp.
Mueller-Dombois, D. and H. Ellenberg. 1974. Aims
  and methods of vegetation ecology. John Wiley and
  Sons. NY.
National Research Council.  1995. Wetlands charac-
  teristics and boundaries. National  Academy Press,
  Washington, DC. 307 pp.
Nelson, D. C.  and R. C.  Anderson. 1983. Factors re-
  lated to the distribution of prairie plants along  a
  moisture gradient.  American Midland  Naturalist
Niering, W. A.  1987a. Hydrology, disturbance, and
  vegetation Change,  pp. 50-54, In J. A. Kusler and
  G. Brooks (eds.). National Wetlands  Symposium-
  Wetland Hydrology.
Niering, W.   A.  1987b. Wetlands hydrology and
  vegetation dynamics. National Wetlands  Newsletter

 Sum Exceedance Value
 Nuzzo, V.  1994. Management plan for Wadsworth
   Prairie  Nature Preserve and  Wadsworth Savanna.
   Lake County, Illinois. Unpublished report to the
   Lake County Forest Preserve District.
 Pickett, S. T. A. and F. A.  Bazzaz.  1978. Organiza-
   tion of an assemblage of early successional species
   on a soil moisture gradient. Ecology  59(6): 1248-
 Rebel, R. J. 1962. Changes in submerged vegetation
   following a change in water level. Journal of Wild-
   life Management 26(2): 221-224.
 Schwegmann, J. E.  (ed.)  1973. Comprehensive plan
   for the Illinois  Nature  Preserves  System, Part 2.
   The  Natural Divisions  of Illinois, Illinois Nature
   Preserves Commission, Rockford. 32 pp.
 Shaw, W.H. 1974. Weighted moisture stress index for
   com in Iowa. Iowa State J. Research 49(2):  101-
 Shipley, B., P. A. Keddy, and L. P. Lefkovitch. 1991.
   Mechanisms producing plant zonation along a wa-
   ter depth gradient: a comparison with the exposure
   gradient.  Canadian Journal of Botany 69:  1420-
 Spence, D. H. N.  1982.  The zonation of plants in
   freshwater lakes. Advances in Ecological Research.
 Spoor, G. In Preparation. Water regime requirements
   of wetland plant species. Report in progress, Silsoe
   College, Cranfield Institute of Technology, Silsoe,
   England. MK45 4DT.
 Squires, L. and A.  G. van  der Valk. 1992. Water-
   depth tolerances of the dominant emergent macro-
   phytes  of the Delta  Marsh, Manitoba. Canadian
   Journal of Botany 70:1860-1867.
 Sudar, R.A., et. al. 1979. A predictive model of water
   stress in corn and soybeans.  American  Society of
   Agricultural Engineers Paper No. 79-2004.
 Thunhorst, G. A.  1993. Wetland planting guide for
   the northeastern United States. Environmental Con-
   cern, Inc. St. Michaels, MD. 179 pp.
United States Department of Agriculture, Soil Con-
  servation Service. 1979. Soil Survey of DuPage and
  part  of Cook counties,  Illinois. Agricultural  Ex-
  periment Station Report No. 108 217 pp. + maps.
United States Department of Agriculture, Soil Con-
  servation  Service.  1970. Soil  Survey  of Lake
  County.  Illinois Agricultural Experiment Station
  Soil Report No. 88. 82 pp + tables and maps.
United States  Geological  Survey.  1995.  Water  Re-
  sources Data for Illinois Open File.
van der Valk, A, (ed.)  1989. Northern  prairie wet-
  lands.  Iowa  State  University Press,  Ames,  IA.
  400 pp.
van der Valk,  A. G., L. Squires and C. H. Welling.
  1994. Assessing the impacts of an increase in water
  level on wetland vegetation. Ecological Applica-
  tions 4(3):525-534.
Weller, M. W. 1987. Freshwater Marshes. University
  of Minnesota Press. Minneapolis. 147 pp.
Wentz, W. A., R. L. Smith and J.  A. Kadlec. 1974.
  State-of-the-art survey of marsh plant establishment
  techniques: induced and natural; volume n, a se-
  lected annotated bibliography on aquatic and marsh
  plants and their management. US Army Engineers
  Waterways Experiment Station, Vicksburg, MS.
White, J.  1978. Illinois natural areas inventory tech-
  nical report. Vol. 1: Survey methods and results. Il-
  linois Natural Areas Inventory, Urbana. 426 pp.
Whitlow, Thomas H., and Richard W. Harris. 1979.
  Flood tolerance in plants: a state-of-the-art review.
  US Army Corps of Engineers. NTIS, E-79-2.
Zimmerman,  J.H. 1987.  A  multi-purpose wetland
  characterization procedure, featuring the hydrope-
  riod. pp. 31-49 In Kusler, J.  A.,  and G. Brooks
  (eds.).  Proceedings. National  Symposium Wetland
  Hydrology, Association  of State Wetland Manag-
  ers, Chicago, IL.

                               A  Successful East-Central  Illinois  Kettle
                                   Marsh Restoration  and  Its  Developing
                                                      Wetland Plant Community
 Abstract:  The restoration of a kettle marsh in
east-central Illinois is detailed and the develop-
ment of its plant  community  is  documented.
Underground drainage tiles were dismantled in
1990, restoring wetland hydrology to the study
area. Wetland vegetation rapidly colonized the
site, as indicated by five years of vegetation sam-
pling data  beginning in 1991. Ninety-five  plant
species have been recorded  in  the wetland, of
which 68 percent are hydrophytic and 78 percent
are native to central Illinois. General trends in
plant community development are discussed, in-
cluding the effects of extremes in rainfall on spe-
cies composition.  Exotic  and  aggressive  plant
species have generally not been a major concern,
although some efforts aimed at cattail (Typha
spp.) control have  been attempted. A controlled
burn conducted in 1995  appeared to  be  more
detrimental than beneficial.

 Wetlands once totaled about lOtmillion acres in
Illinois or about 23 percent of the land surface (Bell
1981). However, it is estimated that over 90 percent
of this wetland area has already been lost (Suloway
and Hubbell 1994), with some estimates putting the
loss as high  as 99.5 percent (Bell 1981). Most of the
remaining wetland habitat is severely degraded, less
than 6,000 acres remain in their natural, undisturbed
state (IDENR 1994). Most of this loss and degrada-
tion has resulted from Illinois agriculture. The con-
version of Illinois wetlands to prime farmland has
been accomplished principally through  drainage.
Historically  encouraged by  legislation, over 5 mil-
lion acres of wetlands had been drained by Illinois
drainage districts as of 1960 (Bell 1981).

 In response to this wetland crisis, numerous wet-
land creation and restoration projects have been un-
dertaken across Illinois. Many of these attempts
involve the conversion of agricultural land back to
                          Brian W. Wilm1
                       Marilyn J. Morris1
                         Scott D. Simon1

wetland. Included in this collection are scattered,
small-scale projects, as well as larger undertakings
along Illinois rivers, where thousands of acres of
farmland are to be converted or reverted  back to
wetland. All of these restoration or creation proj-
ects, no matter the scale, are of importance in trying
to restore or replace the functions and values lost by
the destruction and degradation of Illinois wetlands.
  The restoration project described here is a small-
scale attempt to replace lost wetland habitat through
an approach low in cost, labor, and maintenance.
Also, by systematically monitoring the restoration
and subsequent community development processes,
important  information  can be gained  for  use in
similar restoration efforts and for the promotion of
restoration science in general.

Site History and Methods
  Illinois Natural History Survey researchers have
been studying a restored marsh community in east-
central Illinois. This restored wetland is the result of
a cooperative effort between the Illinois  Natural
History Survey,  the Champaign County Forest Pre-
serve District, and regional volunteer organizations,
such as the Nature Conservancy  and the Grand
Prairie Friends. Additional funding was provided by
the Illinois Department of Transportation.
  Prior to the restoration process, the study area was
principally fallow, abandoned pasture, having not
been grazed for 10 years  or more. Underground
drainage tiles, in place for at least several decades,
directed  water  to an excavated ditch  running
through the center of the area. Some wetland plant
species, including cattail (Typha spp.)  and willow
(Salix  spp.), persisted along this  ditch, although
much of the vegetation likely  consisted  of non-
wetland species. In 1990, the drainage tile system
was dismantled, restoring "natural" hydrology to the
site.  Some excavation was required in this process,
creating a shallow depression at the eastern edge of
'Illinois Natural History Survey, Center for Wildlife Ecology, 607 East Peabody Drive, Champaign, Illinois 61820.

Illinois Kettle Marsh Restoration
the site. Water now collects in this low-lying basin
of  approximately 7  acres, persisting on  the  site
throughout the year.  Only in late summer does the
entire wetland occasionally dry out.
  Hydrophytic vegetation began to return to the site
almost  immediately  after the dismantling of the
drainage tile system.  For the purposes of this study,
hydrophytic plant  species are those classified,  in
Illinois, as facultative or wetter by Reed  (1988).
Some native plant species were also introduced into
the wetland via seed, rootstock, and/or plantings,
but with virtually no success. These efforts are not
thought to have significantly affected the vegetation
composition of the restored wetland.
  Vegetation  sampling in the wetland  restoration
was initiated in 1991 and has continued on a yearly
basis through 1995. Seven vegetation transects were
established at 30tm intervals, crossing the length of
the wetland. Herbaceous vegetation was  sampled
within  one-quarter meter plots at 5 m  intervals
along each transect. Within each plot, plant species
were identified  and  the estimated percent aerial
coverage  was  determined  (Daubenmire  1959).
These percent aerial  coverages were then converted
to  standard cover  classes (Table 1). Mid-points  of
the ranges of aerial  coverage  were used in further

Table 1. Cover classes used in  herbaceous
vegetation sampling (Daubenmire 1959).
Cover Class
Range of
Coverage (%)
< 1
Mid-point of
Range (%)
  Dominance for an individual plant species is cal-
culated as the total aerial coverage for all quadrats
divided by the area sampled. Relative dominance
for a species is the dominance for that species di-
vided by the total  dominance for all species, times
one hundred. Frequency for a given species is cal-
culated as the number of quadrats in which the spe-
cies occurs divided by the total number of quadrats
sampled. Relative frequency for a species is the fre-
quency for  that species divided by the total  fre-
quency values  for all  species,  times one hundred.
The importance value for any species is defined as
its  relative dominance plus its  relative frequency,
divided by two. All formulas and calculations  fol-
low those defined in Cox (1985).
  A controlled  burn conducted on the surrounding
prairie and savanna restorations was allowed to ex-
tend into the restored  wetland in November 1994.
Vegetation sampling results from 1995 show any ef-
fects of this fire. Also, a one-time attempt to physi-
cally   remove  cattail  from the  restoration was
conducted in 1992.

Results  and Conclusions
  Ideally, restorationists and site managers would
like to predict the success of wetland creations  and
restorations and the development of their associated
plant communities.  Unfortunately, this is not an ex-
act  science.  Studies  have  documented  wetland
community  development on  these types  of sites
(Reinartz and Warne 1993, Wilhelm and Wetstein
1993), but generalizations are  difficult. In a broad
sense, newly developing wetlands are often charac-
terized by opportunistic, invading plant  species  that
may give way to species characteristic of more sta-
ble wetlands as time passes (Mitsch and Gosselink
1986,  Odum  1988).  Regional,  local, and site-
specific influences, however,  all affect  wetland
community  development (Willard and Hiller 1990).
Weather, species interactions (e.g., competition, al-
lelopathy), immigration, and emigration all contrib-
ute to this constant change to which plants are
continually trying  to  adapt (van  der  Valk 1981).
Nonetheless, some  patterns and trends in vegetation
establishment, change, and community development
are evident, particularly in individual cases.

Colonization and  Vegetation
  In this restoration,  development of  the wetland
plant community proceeded at a rapid pace from the
very beginning. In  1991, one year after the restora-
tion of "natural" hydrology, 43 different plant spe-
cies were present (Table 2). Species composition in
this  first-year community  was  also  favorable.
72 percent  of the  plant  species were  classified as
hydrophytic and 73 percent were native  to central
Illinois.  As  indicated  by   importance  values
 (Appendix  A),  the most  prevalent plant  species
 sampled in that  first year are generally considered
 to be valuable or desirable wetland species, with the

possible exception of common ragweed (Ambrosia
artemisiifolia).  Other prevalent plant species  in-
cluded beggar-ticks (Bidens tripartita), spikerushes
(Eleocharis palustrls and E. obtusa), marsh yellow
cress (Rorippa islandica), water plantain (Alisma
plantago-aquatica), millet (Echinochloa crus-galli),
sedges (Carex frankii  and C.  tribuloides),  and
duckweed (Lemna minor). The quick return of wet-
land vegetation to the restoration  is thought to be
attributable to the spread of seed from the previ-
ously existing wet  ditch, although waterfowl and
other water birds may have  also been the  source of
some seed. The drained area had likely been dry  too
long for the survival of any  viable wetland seeds in
the below-ground seed bank, although the seeds of
many wetland species may remain viable beneath
the soil surface  for many years (Galatowitsch and
van derValk 1994).

Trends in Vegetation Composition and
  Changes in the developing plant community, some
of which  appear to be  directly related to rainfall,
have been noted from year to year.  Although  de-
tailed hydrologic data was not collected  until  re-
cently, general, anecdotal  information is known. Of
particular note are the years  of  1992 and  1993.
Rainfall during 1992 was very slight, leading to  dry
conditions and very low water levels in the restored
wetland. As a result of  this drought, three obligate
wetland plant species,  water  plantain, arrowhead
(Sagittaria latifolid), and duckweed, disappeared
from sampling results (Table 3). Other plants, such
as ragweed, swamp marigold (Bidens aristosa), and
smartweed (Polygonum  spp.), increased  greatly
during this time, apparently responding to the drier
conditions  and  the disappearance  of competing
plant species. Plant species diversity  dropped in
1992, as did the percent of hydrophytic species and
the percent  of  species native to central  Illinois
(Table 2), all likely attributable to a lack of water.

  Heavy rains returned  in 1993, resulting  in very
wet conditions at the  wetland restoration  site. Spe-
cies  diversity  again declined, but the percent of
hydrophytic  species  and the  percent  of  species
native to central Illinois both increased (Table  2).
Excessively  wet conditions may  have  caused  the
extirpation of some  plant  species favoring drier
conditions. The three obligate wetland plant species
that disappeared in 1992 all  returned in 1993, while
common ragweed, swamp marigold, and smartweed
all dropped substantially. Common ragweed, in fact,
disappeared completely from sampling results.
  Some general  trends  in  vegetation change and
plant community development  not thought to be
directly related to fluctuation in rainfall have also
been noted. Marsh yellow cress, a common moist-
soil herb, has shown a large decline throughout the
five-year sampling period. Exhibiting an importance
value of 10.75 in the first year of sampling, it has
decreased rapidly, to an importance value under one
in the last  three years of sampling. Marsh yellow
cress was  successful in colonizing and exploiting
the newly  restored wetland, but is apparently not
able to thrive in the developing wetland plant com-
munity. It may not compete well with dense growth
of taller plant species such as beggar-ticks, cattail,
and swamp marigold. The opposite trend has  been
exhibited by swamp marigold, a tall, moist-soil an-
nual.  Initially exhibiting an  importance value  of
only 1.34 in 1991,  it has grown steadily in preva-
lence throughout. Conditions  at the restoration site
are apparently favorable for survival and growth of
this  species,  enabling  its  continued population
growth through reseeding.

Invasive  and Exotic Vegetation
  Potentially troublesome and undesirable species
have generally not been a major problem thus far.
Cattail has  been the only species to cause moderate
concern. Although native and an  important part of
many  wetland communities,  cattails  can spread
rapidly, excluding other plant species and decreas-
ing diversity such that  a  virtual monoculture is
sometimes  the result (Baldassarre and Bolen 1994,
Galatowitsch and van der Valk  1994). This may be
of particular concern in the developing plant  com-
munity  of  a restored wetland where the establish-
ment of other species may be prevented by cattail
(Galatowitsh and  van der Valk 1994). Cattail,  how-
ever,  is  an important component of the restored
marsh and  is the only common, large emergent in
the wetland. It is a valuable food source for muskrat
(Ondatra zibethica) and important habitat for  birds
such  as bitterns,  rails,  grebes,  and  waterfowl
(Baldassarre and Bolen 1994, Herkert 1992, Martin
et al 1951). Reed canary grass (Phalaris arundina-
cea) is another species found  within the wetland
restoration  that has the potential to be problematic.
  Capable of tolerating extensive periods of inunda-
tion, it can sometimes dominate shallow marshes
(Galatowitsch and van der Valk  1994).
  Cattail and reed canary grass  were initially of no
concern in  the restored wetland.  Cattail, however,
began  to show  a  steady  increase. The physical

Table 2. Number of plant species, percent of species hydrophytic, and percent of species
native to central Illinois in the restored wetland, 1991-1995.
Number of Species
1991 43
1992 37
1993 29
1994 34
1995 53
Reed (1988).
Vaftetal. (1993).
Table 3. Importance values

Alisma plantago-aquatica
Ambrosia artemisifolia
Bidens spp. (total)
Bidens aristosa
Bidens tripartite
Carex spp. (total)
Echinochloa crus-galli
Eleocharis spp. (total)
Eleocharis palustris
Leersia oryzoides
Lemna minor
Phalaris arundinacea
Polygonum spp. (total)
Rorippa islandica
Sagittaria latifolia
Typha spp. (total)

for selected
% of Species % of Species Native to
Hydrophytic1 central Illinois'

plant species

at the restored



                                                                             Illinois Kettle Marsh Re
removal of cattail in 1992  was effective, but was
deemed too labor intensive for long-term  mainte-
nance. Although  cattail  had  not really  reached
problematic levels, the controlled burn conducted
on the adjacent prairie and savanna restorations was
allowed to burn into the wetland restoration to help
control cattail and, secondarily, reed canary grass.

Fire Effects
  Almost no positive effects were evident from the
controlled burn. Cattail and reed canary grass both
showed increased importance  values after  the fire
(Table 3). Cattail can apparently  be controlled by
fire only under limited circumstances. For cattail to
be killed, the fire must be hot enough, or the marsh
must be  dry  enough, to  bum into the  sediment or
peat layer of the wetland (Beule  1979, Smith and
Kadlec 1985).

  Reed canary  grass can be controlled  by repeated
exposures to  fire (Hutchison 1990, Galatowitsh and
van  der Valk 1994). This one-time burn, however,
appears to have had the  opposite  effect or, at best,
no effect. Although still a relatively minor compo-
nent of the wetland plant community,  reed canary
grass did rise  to  an importance  value of  2.41  in
 1995, almost six times the previous year's level. Ob-
servations indicated that  this was attributable  to the
spread of existing colonies. No new  colonies were
  Two important  wetland plant species, water plan-
tain  and arrowhead, showed a definite decline after
the  burn. Water plantain,  a common, moist-soil,
emergent, had  been prevalent in the wetland  resto-
ration throughout, with the exception of the  drought
year (1992). However, 1995's sampling results indi-
cated a tremendous drop, from an  importance value
of 12.07  in 1994 to only 0.60. Similarly, the  emer-
gent arrowhead showed  a large decrease after  the
burn. Arrowhead had an importance value of 4.76 in
1994, but dropped dramatically to 0.35 in 1995.

  Populations of several  plants, besides cattail and
reed canary grass,  increased after the  fire. Smart-
weeds, common  ragweed,  swamp marigold, rice
cutgrass  (Leersia oryzoides),  and sedges  (Carex
spp.) all showed greater  importance values  in 1995
sampling results. Whether  or  not all of these in-
creases are directly attributable to the fire is highly
debatable. Sedges appeared to fluctuate throughout
the five-year sampling period and 1995's increase
may  have just been an upward correction from the
unexplained  decline  in   1994.   As   previously
discussed,  smartweeds,   common  ragweed, and
swamp marigold all also  increased in the drought
year of 1992. These plants appear to be disturbance-
tolerant, capable of tolerating and exploiting distur-
bances such as drought and fire, where other species
may not be able to survive or at least compete ef-
fectively.   Smartweeds,  common   ragweed,  and
swamp marigold are all annuals which produce  an
abundance of seed, thereby allowing rapid coloni-
zation of newly exposed ground after disturbance.
  Overall,  plant species diversity rose greatly fol-
lowing the burn, from  34  species in 1994 to 53 in
1995. Species diversity often  increases following a
disturbance such as fire (Triner and Klimstra 1965,
Ricklefs 1983). The increase in species diversity,
unfortunately, was not due to an influx of native,
hydrophytic species, as both the percent of species
hydrophytic and the percent  of species native  to
central Illinois  decreased. Invasion by exotic, up-
land species  accounted for these trends. Although
species diversity increased after the burn, few, if
any, other positive effects were noted. Two highly
desirable species (water plantain and arrowhead)
declined greatly, while  cattail and reed canary grass
both increased.  Decreases  in the percent of species
hydrophytic and the percent  of species native  to
central Illinois were other negatives possibly attrib-
utable to the fire. This one-time burn appears  to
have been more of a detriment to the wetland plant
community than a benefit. Fire may not be an ef-
fective  management tool  in  this  marsh wetland
community and no additional burns are planned at
this time.

Vegetation Sampling Summary
  Five years of vegetation sampling has, thus far,
revealed a total of 95 plant species, representing 67
different genera (Appendix B). Hydrophytic species
accounted for 68 percent of the total. Seventy-eight
percent of species were native  to  central Illinois.
Only  a few of  the sampled plants  (Vemonia spp.,
Asclepias  incarnata, and Carex spp.)  may have
originated, at least partially, from limited plantings
and  seedings in the wetland. One other sampled
species, drooping  coneflower  (Ratibida pinnata),
was originally  introduced into  the  adjacent prairie
and  savanna restorations. Artificially established
vegetation, however, is not thought to have signifi-
cantly contributed to the composition of the wetland
community. Community composition is attributable
to natural  processes of vegetation establishment,
competition, and succession.

 Illinois Kettle Marsh Restoration
 Summary and Conclusions
   Although this wetland restoration is termed a suc-
 cess by  its  researchers, the evaluation of wetland
 restoration and creation success is often quite diffi-
 cult. What exactly does determine success? Ideally,
 specific goals directly related to the proposed func-
 tions of the restored or created wetland should be
 formulated  before  the  process  actually  begins
 (Army Corps 1993). This restoration  project,  un-
 fortunately, was not designed with scientific  ex-
 perimentation and evaluation in mind. Evaluation of
 success is largely subjective in this case. Wetland
 success  evaluation often  includes  monitoring se-
 lected aspects  of vegetation, hydrology,  and/or
 wildlife utilization (Erwin 1990). Monitoring of this
 restoration has  mainly focused on  the previously
 discussed vegetation  sampling. This sampling re-
 vealed  a rapidly-developing,  diverse  flora domi-
 nated by native, hydrophytic vegetation, similar to
 what might be found in a natural wetland commu-
 nity. Assuming no major hydrologic changes and no
 dramatic  increases in potentially problematic vege-
 tation,  this functioning wetland community  will
 likely persist  into the future. This newly restored
 wetland  is already being used by a wide variety of
 mammals, reptiles, amphibians, and birds, including
 two Illinois-endangered species, the northern harrier
 (Circus   cyaneus)  and   the  pied-billed   grebe
 (Podilymbus podiceps) (Herkert  1992). In 1994, a
 pied-billed grebe even nested successfully in the
   Continued sampling and monitoring over the next
 few years will  clarify and  more accurately deter-
 mine any patterns of vegetational  change in the
 wetland. Included in this determination will be the
 true, long-term effects of fire, if any, on vegetation
 composition. For example, will water  plantain  and
 arrowhead both  recover after being  severely af-
 fected by the  fire? Recently installed  staff gauges
 and ground  water monitoring wells will permit ac-
 curate monitoring and documentation of hydrology,
 hopefully allowing the relationship between hydrol-
 ogy and  vegetation composition to  be  successfully
 delineated. Continued vegetation  sampling will be
 supplemented  by detailed data analysis of vegeta-
 tion composition, community change, and hydro-
 logic/vegetational  relationships. Further  sampling
 and analysis of  this  restored wetland  will provide
 restorationists  and site managers  with background
 information that can  hopefully be of use in  similar
 types of restoration efforts. Long-term monitoring
 of this developing wetland community  will provide
 data not commonly available.
 This  restoration effort has successfully created a
functioning kettle marsh wetland in an area of Illi-
nois where wetlands have readily been drained and
degraded. This quality restoration  has realistically
been achieved with relatively minor effort and cost.
Proper site selection was key. This restoration site
was a landscape depression with artificially drained,
hydric  soils, meaning that it was likely wetland in
the past. Natural hydrology was easily restored by
disrupting the underground drainage tiles through
minimal excavation, unlike many other restorations
or creations where extensive excavation is required.
Also unlike many other projects, natural vegetation
establishment processes were the primary source of
the wetland plant  community.  Other  restorations
often invest much time and money in the planting
and seeding of desired vegetation, often with less
than optimal results. Relying principally on natural
revegetation, this restored wetland  has developed a
diverse wetland community  dominated  by native,
hydrophytic species in a very short time. As demon-
strated, proper site selection can  allow for a low
cost approach to restoration, requiring only minimal
labor and low maintenance. The broad-scale attrac-
tiveness of this type of restoration is obvious, espe-
cially when compared to more intensive approaches.
Although small in scale,  the contribution that this
restored wetland makes to the natural resources of
Illinois is nonetheless valuable. Similar restoration
efforts can be repeated on a statewide or regional
basis to significantly help offset the loss and degra-
dation  of wetland habitat.

Literature Cited
Baldassarre, G. A., and E. G. Bolen.  1994. Water-
  fowl ecology and management. John Wiley &
  Sons, New York, NY.

Bell, H. E., m. 1981. Illinois wetlands:  their  value
  and  management. State of Illinois, Institute of
  Natural Resources, Chicago,  IL. Document No.

Beule, J. D. 1979. Control and management of cat-
  tails  in southeastern Wisconsin  wetlands. Wise.
  Dept. Nat. Resour. Tech. Bull. 112, Madison, WI.

Cox, G. W-  1985.  Laboratory manual of general
  ecology.  5th  ed. Wm.  C.  Brown  Publishers,
  Dubuque, IA.

Daubenmire, R. F. 1959. Canopy coverage method
  of vegetation analysis. Northw. Sci. 33:43-64.

Erwin, K. L. 1990. Wetland evaluation for restora-
  tion  and creation, pp. 429-449, In J. A. Kusler and

                                                                           Illinois Kettle Marsh Re
 M. E. Kentula (eds) Wetland Creation and Resto-
 ration - the Status of the  Science. Island Press,
 Washington, DC.

Galatowitsch, S. M., and A. G. van der Valk. 1994.
 Restoring prairie  wetlands:   an  ecological  ap-
 proach. Iowa State Univ. Press, Ames, IA.

Herkert, J. R. (ed). 1992. Endangered and threat-
 ened species of Illinois:  status and  distribution,
 Volume 2 - animals.  Illinois Endangered Species
 Protection Board, Springfield, IL.

Hutchison,  M.   1990.  Vegetation  management
 guideline - reed canary grass (Phalaris arundina-
 cea). Vol.  1, No. 19.  Illinois Nature Preserves
 Commission, Springfield, IL.

Illinois Department of Energy and  Natural  Re-
  sources and Nature of Illinois Foundation. 1994.
 The changing Illinois environment: critical trends.
  Summary report of the critical trends assessment
  project.  Illinois  Department  of  Energy   and
  Natural   Resources   and  Nature   of  Illinois
  Foundation, Springfield, IL.
Linde, A. F. 1969. Techniques for wetland man-
  agement. Research Report 45,  Dept. of Natural
  Resources, Madison,  WI.
Martin, A. C., H. S. Zim, and A. L. Nelson. 1951.
  American wildlife and plants - a guide to wildlife
  food habitats. McGraw-Hill Book Company, New
  York, NY.
Mitsch, W. J. and J. G. Gosselink. 1986. Wetlands.
  Van Nostrand Reinhold, NY.
Odum, W. E. 1988. Predicting ecosystem develop-
  ment following creation and restoration of wet-
  lands, pp.  67-70, In  J. Zelazny and J. S. Feiera-
  bend (eds) Increasing Our  Wetland Resources,
  National  Wildlife Federation  Conference  Pro-
  ceedings. Washington, DC.
Reed, P. B., Jr. 1988.  National list of plant species
  that occur  in wetlands:  Illinois. U.S.  Fish  and
  Wildlife Service, National  Wetlands Inventory,
  Washington, DC. NERC-88/18.13.
Reinartz, J. A. and E. L. Wame. 1993. Development
  of vegetation  in small created wetlands in south-
  eastern Wisconsin. Wetlands 13(3): 153-164.

Ricklefs, R. E.  1983. The economy of nature. 2nd
  ed. Chiron Press, New York, NY.

Smith, L. M. and J. A.  Kadlec. 1985. Fire and her-
  bivory in  a  Great Salt Lake  marsh.  Ecology

Suloway, L., and M. Hubbell.  1994. Wetland re-
  sources of Illinois: an analysis and atlas. Illinois
  Natural History  Survey,  Champaign,  Illinois.
  Special Publication 15.

Taft, J., G. S. Wilhelm, D. Ladd, and L. A. Masters.
  Floristic quality assessment for Illinois. Erigenia.
  In preparation.

Triner, E. J. and W. D.  Klimstra. 1965. Effects of
  burning and  fallowing on vegetation.  Tran. 111.
  Acad. Sci. 58(2):102-114.

van der Valk, A. G. 1981. Succession in wetlands:
  a Gleasonian approach. Ecology 62(3):688-696.

Wilhelm, G.  and L. Wetstein. 1993. Vegetational
  monitoring of a wetland restoration in northern
  Illinois, pp. 755-765, In M. C. Landin  (ed) Wet-
  lands - Proceedings of the 13th  Annual Confer-
  ence, Society of Wetland  Scientists. June  1992,
  New Orleans, LA.
Willard, D. E. and A. K. Hiller. 1990. Wetland dy-
  namics:  considerations for restored and created
  wetlands, pp. 459-466 In J. A. Kusler and M. E.
  Kentula (eds) Wetland Creation and Restoration -
  the Status of the Science. Island  Press, Washing-
  ton, DC.
U.S.  Army  Corps of Engineers. 1993. Guidelines
  for  developing mitigation  proposals. Chicago
  District, U.S. Army Corps of Engineers, Chicago,

       Illinois Keltic Marsh Restoration
Appendix A. Top importance values for vegetation at the restored wetland, 1991-1995.
         Bidens tripartite
         Eleocharis palustris
         Rorippa islandica
         Alisma plantago-aquatica
         Ambrosia artemisiifolia
         Echinochloa crus-galli
         Carex frankii
        Lemna minor
        Carex tribuloides
        Eleocharis obtusa
                                   Bidens tripartita
                                   Ambrosia artemisiifolia
                                   Eleocharis palustris
                                   Bidens aristosa
                                   Polygonum punctatum
                                   Polygonum pensylvanicum
                                   Carex tribuloides
                                   Echinochloa crus-galli
                                   Rorippa islandica
                                  Carex frankii
Eleocharis palustris
Lemna minor
Bidens tripartita
Alisma plantago-aquatica
Typha latifolia
Carex  frankii
Bidens aristosa
Sagittaria latifolia
Bromus inermis
Juncus acuminatus
Eleocharis palustris
Lemna minor
Bidens tripartita
Alisma plantago-aquatica
Bidens aristosa
Sagittaria latifolia
Typha latifolia
Amaranthus tuberculatus
Carex frankii
Eleocharis acicularis
Eleocharis palustris
Bidens aristosa
Bidens tripartita
Lemna minor
Ambrosia artemisiifolia
Leersia oryzoides
Typha latifolia
Polygonum punctatum
Eleocharis acicularis
Phalaris arundinacea

Appendix B. Complete plant species list for the restored wetland, 1991-1995.
Scientific Name
Abutilon theophrastii
Acalypha spp.
Acalypha virginica
Acer saccharinum
Agropyron repens
Alisma plantago-aquatica
Amaranthus rudis
Amaranthus spp.
Amaranthus tuberculatus
Ambrosia artemisiifolia
Asclepias incarnata
Aster pilosus
Aster simplex
Bidens aristosa
Bidens frondosa
Bidens spp.
Bidens tripartite
Bromus inermis
Carex annectens
Carex frankii
Carex spp.
Carex tribuloides
Carex vulpinoidea
Carya ovata
Chamaesyce humistrata
Chamaesyce spp.
Chenopodium album
Cirsium arvense
Cirsium spp.
Cirsium vulgare
Conyza canadensis

Cornus spp.
Crataegus mollis
Common Name
three-seeded mercury
three-seeded mercury
silver maple
water plantain
tamarisk waterhemp
common ragweed
swamp milkweed
hairy aster
panicled aster
swamp marigold
awnless brome grass
fox sedge
scaly-bark hickory
milk spurge
lamb's quarters
Canada thistle
bull thistle

red haw

Native to central Illinois2


 Illinois Kettle Marsh Restoration
  Appendix B. Complete plant species list for the restored wetland, 1991-1995.
Scientific Name
Daucus carota
Dichanthelium spp.
Echinochloa crus-galli
Eclipta prostrata
Eleocharis acicularis
Eleocharis erythropoda
Eleocharis obtusa
Eleocharis palustris
Erechtites hieracifolia
Erigeron annuus
Erigeron spp.
Eupatorium altissimum
Fragaria virginiana
Galinsoga quadriradiata
Geum canadense
Geum laciniatum
Glechoma hederacea
Gleditsia triacanthos
Impatiens capensis
Juncus acuminatus
Juncus tenuis
Lactuca serriola
Lactuca spp.
Leersia oryzoides
Lemna minor
Lindernia dubia
Lycopus americanus
Lysimachia nummularia
Oxalis stricta
Penthorum sedoides
Phalaris amndinacea
Phyla lanceolata
Plantago lanceolata
Common Name
panic grass
barnyard grass
yerba de tajo
needle spikerush
annual fleabane
tall boneset
wild strawberry
Peruvian daisy
white avens
rough avens
ground ivy
honey locust
path rush
wild lettuce
rice cutgrass
false pimpernel
common water horehound
yellow wood sorrel
ditch stonecrop
reed canary grass
Native to central Illinois

                                                                         Illinois Kettle Marsh Restoration
Appendix B. Complete plant species list for the restored wetland, 1991-1995.
Scientific Name
Plantago rugelii
Poa pratensis
Polygonum amphibium
Polygonum hydropiper
Polygonum lapathifolium
Polygonum pensylvanicum
Polygonum persicaria
Polygonum punctatum
Polygonum ramossissimum
Polvqonum scandens
Common Name
red-stalked plantain
Kentucky bluegrass
water smartweed
wild water pepper
pale smartweed
giant smartweed
spotted lady's thumb
dotted smartweed
bushy knotweed
climbing buckwheat
Native to central Illinois2
 Polygonum spp.
 Portulaca oleracea
 Potamogeton nodosus
 Ratibida pinnata
 Rorippa islandica
 Rumex altissimus
 Rumex crispus
 Sagittaria graminea
 Sagittaria latifolia
 Salix exigua
 Scirpus acutus
 Scirpus tabernaemontanii
 Scrophularia spp.
 Setaria faberi
 Setaria glauca
 Sida spinosa
 Solanum carolinense
 Toxicodendron radicans
 Trifolium pratense
 Trifolium repens
 Typha angustifolia
American pondweed
drooping coneflower
marsh yellow cress
pale dock
curly dock
grass-leaved arrowhead
sandbar willow
hardstem bulrush
softstem bulrush
giant foxtail
yellow foxtail
prickly sida
horse nettle
poison ivy
red clover
white clover
narrowleaf cattail


  Illinois Kettle Marsh Restoration
   Appendix B. Complete plant species list for the restored wetland, 1991-1995.
Scientific Name
Typha latifolia
Verbena hastata
Vernonia spp.
Veronica spp.
Viola pratincola
Viola spp.
Vitis riparia
Common Name
blue vervain
common blue violet
Native to central Illinois
   Reed (1988).
  Vaftetal. (1993).

                                    Restoration of Twelve Heavy-Seeded
                                    Hardwood Species in South Carolina
                                                           Coastal Plain Wetlands
                                                                             W. H. Brantley1
                                                                            V. B. Shelburne
                                                                                   D. D. Hook
  Abstract:  The restoration  of altered forested
 wetland ecosystems involves the determination of
 which species are best able to  survive under dif-
 ferent flood regimes, soils, and competition from
 forbs and grasses. In 1991 a study was established
 to measure the survival  and  growth of twelve,
 heavy-seeded  hardwood tree  species on  poorly
 drained, somewhat  poorly drained, moderately
 well drained,  and well  drained drainage classes,
 on  three different geologic terraces across  the
 South  Carolina  coastal  plain.  Species  were
 selected for their wildlife and forest product val-
 ues. On each terrace twelve blocks of species were
 planted in 7x7  rows and were  replicated twice
 more in each drainage class. The interior 5x5 rows
 were  measured, and the second and fourth rows
 contained tree shelters. Survival and growth were
 significantly  affected by tree  shelters and soil
 drainage class. The first three years of  data indi-
 cate that tree shelters  significantly increased  the
 survival of all twelve species.  Growth  responses
 among seedlings of the same  species  were also
 significant,  but  less  dramatic than  survival.
 Drainage class was significant  in species survival
 and growth by the third year, with generally more
 survival and growth found in the well drained and
 moderately well drained drainage classes. Addi-
 tionally, tree  shelters  provided protection from
 wildlife browsing and protection from competing
 forbs  and grasses. Finally, a suite of species  for
 each drainage class was determined  using meth-
 odology which incorporated best species perform-
 ance by total height and percent survival.
  It was estimated in the mid-1980s that 103.3 million
acres of wetlands existed in the continental United
States,  which actually  represents  more  than  a
50 percent decrease in wetland acreage since settle-
ment began (Dahl 1990). Wetlands are important to
society because of the various functions and values
which  they  provide (Walbridge 1993).  Restoring
altered  forested wetland ecosystems requires select-
ing those species  which appear to be best adapted to
survive specific flood regimes, soils, and competition
from forbs and grasses.
  Although a considerable amount is known about the
flood tolerance of many tree species (Teskey and
Hinckley  1977, Hook 1984a, Kozlowski 1984, Tiner
1991), specific information on species-site suitability
is lacking when they  are planted on  wetland sites
where hydrologic regimes and soils have been altered
(Hook 1984b). Thus, there is a real need to develop
better guidelines for matching species and site traits
to improve restoration efforts.
  Information  on the effects of tree shelters  on sur-
vival and growth of tree species is scant and virtually
nothing is known  about their effects on wetland sites
(Manchester et al. 1988, Potter 1989, Tuley 1983 and
1985, Minter et al. 1992, Bardon and Countryman
1992, Lantagne et al. 1990, Ponder 1992, McConnel
  The effects of soil drainage class and tree  shelters
on the  survival and growth of twelve heavy-seeded
hardwood species in the South Carolina coastal plain
was the focus of this study.  It was hypothesized that
species would not differ, with soil drainage class or
tree shelters, in percent survival and height growth.
' Department of Forest Resources, College of Agriculture, Forestry and Life Sciences, Clemson University, 261 Lenotsky
Hall, Box 341003, Clemson, South Carolina 29634-7003.

 South Carolina Coastal Plain Wetlands
 Materials and Methods

   This project was carried out on three separate geo-
 logic terraces across the South Carolina coastal plain.
 The  study  locations  were the Savannah  River  Site
 (SRS; elevation  >  30m mean sea level  [MSL])  in
 New Ellenton, SC;  the Santee Experimental Forest
 (SEF; < 15 m MSL) in Cordesville, SC; and the
 Webb  Wildlife  Center  (WWC;  between 15-30  m
 MSL) located  in Gamett, SC. Two locations (SRS
 and WWC) contained four different drainage classes
 (well   drained   (WD),  moderately  well  drained
 (MWD), somewhat poorly drained (SPD),  and poorly
 drained  (PD) soil while SEF contained only three soil
 drainage classes  (there was no well drained drainage
 class). As  implied,  drainage classes  were  chosen
 based on their relative wetness. Differences in these
 classes  were  based on  several  factors  including
 hydrology  and  soil  characteristics,  including  soil
 aeration and depth to mottling. Depth to mottling on
 the well drained class was  greater than or equal  to
 30 inches;  16  to 29 inches  on the moderately well
 drained  class; 9 to 15 inches on the somewhat poorly
 drained  class; and 3 to 6 inches on the poorly drained
 site.  The soil  series for  each terrace and drainage
 class are listed below.
   Santee Experimental Forest - Flatwoods Coastal
   Plain, Upper Terrace
      Moderately Well Drained - Duplin
      Somewhat Poorly Drained - Wahee
      Poorly Drained - Bethera
   Savannah River Site - Hilly Coastal Plain
      Well Drained - Lucy and Troup
      Moderately Well Drained - Hornsville
      Somewhat Poorly Drained - Smithboro
      Poorly Drained - Rembert

  Webb  Wildlife Center - Flatwoods Coastal Plain,
   Lower Terrace

      Well Drained - Emporia

      Moderately Well Drained - Emporia

     Somewhat Poorly Drained - Elonia
     Poorly Drained - Elonia
 Experimental Design
  Each drainage class contained three replications of
 twelve species planted in 7x7 blocks. Only the inner
 5x5 rows were  measured with  the  second and the
 fourth rows  containing tree shelters; for a total of
 25 trees per block, with 10 trees containing tree shel-
 ters. The study was set up as a split-split plot design.
 For analysis purposes only seedlings 1-20 (rows 1-4)
 were  used to balance the number of sheltered versus
 non-sheltered seedlings. The species involved in the
 study were: water hickory (Carya aquatica (Michx.
 f.) Nuft.), swamp tupelo (Nyssa sylvatica var. biflora
 (Walt.)  Sarg.),  swamp  chestnut   oak  (Quercus
 michauxii Nuft.), mockernut hickory (Carya tomen-
 tosa  (Poir.)  Nuft.), white  oak (Quercus  alba L.),
 southern red oak (Quercus falcata Michx.), overcup
 oak (Quercus lyrata Walt.),  cherrybark oak (Quercus
pagoda Ell.),  blackgum  (Nyssa sylvatica  Marsh.),
 persimmon  (Diospyros  virginiana  L.),  sugarberry
 (Celtis  laevigata Willd.),  and  dogwood  (Comus
florida L.). Species were chosen based on their rela-
 tive abundance within these  areas, their wildlife value
 and their forest products  value. For ease in reading
 the graphs, Table 1  lists the  species with their species
 code  along   with   their relative  flood  tolerance
 (McKnight et al. 1980).

  All  trees  were measured for  survival  and total
 height. Total  height was  obtained by using a meter
 stick  placed beside  the individual. Visual observation
 determined whether the tree was alive or dead. Third
 year height growth  and diameter were also measured,
 but the results are not presented herein. Additionally,
 wildlife browsing and dieback were noted for each

 Species Suites
  The methodology behind  the determination of the
 suite  of species best suited for a particular drainage
 class was  based on four  categories: non-sheltered
 percent  survival, sheltered percent survival,  non-
 sheltered total height and sheltered total height. The
 cutoff point for a species falling into the percent sur-
 vival category was  at least 45 percent survival. The
 cutoff point for a species falling into the total height
 category was at least 40 cm in height. These cutoffs
 were chosen  based on relative performance of indi-
 vidual species within each class. For every terrace
 and drainage class, in each of the  four  categories, a

Table 1. Species Codes and Relative Flood Tolerance of the Twelve Study Species (McKnight et al.
      Species Code
Cherrybark Oak
Mockernut Hickory
Overcup Oak
Swamp Chestnut Oak
Southern Red Oak
Swamp Tupelo
Water Hickory
White Oak
Table 2. Non-Sheltered
Percent Survival

Terrace WD
Moderately Tolerant
Weakly Tolerant to Intolerant
Moderately Tolerant to Intolerant
Moderately Tolerant
Moderately Tolerant
Weakly Tolerant
Moderately Tolerant
Moderately Tolerant to Intolerant
Species Performance Across All Terraces and



  SRS        bg,co,oc,p,sc,sr,wo
  WWC      co,d,h,oc,p,sc,sr,wo
  SEF        NA
  Overall     co,oc,p,sc,sr,wo
                   co,oc,sc,wo              bg,co,oc,sc,sr,wo        oc,st
                   oc                       co,oc,sc                 co
                   bg,co,oc,p,sc,sr,wo       bg,co,oc,p,sc,st,wh       oc
                   co,oc,sc,wo    	bg,co,oc,sc	oc
Table  3.  Sheltered  Percent Survival  Species Performance Across  All  Terraces and Drainage
  SRS      bg,co,d,mh,oc,p,sc,   bg,co,oc,p,sc,sr,       bg,co,h,oc,p,sc,sr,st,    bg,co,oc,p,sc,st
           sr,wh,wo             st,wo                 wo
  WWC    bg,co,d,h,mh,oc,p,    bg,co,oc,p,sc,sr,wo    bg,co,oc,p,sc,srwo      bg,co,oc,p,sc,
           sc,sr,wh,wo                                                       wo
  SEF      NA                  bg,co,oc,p,sc,sr,       bg,co,mh,oc,p,sc,st,    bg,co,oc,p,sc,
                                 st,wo                 wh,wo                stwh
  Overall   bg,co,d,mh,oc,p,sc,   bg,co,oc,p,sc,sr,       bg>co,oc,p,sc,st,wo     bg,co,oc,p,sc,st
           sr,wh,wo             stwo

 South Carolina Coastal Plain Wetlands
 suite of species was chosen based on their ability to
 meet the minimum cutoff requirements. Once a suite
 was determined for every drainage class in each  ter-
 race, the results were compared and any overlapping
 species  were put into an overall group which con-
 sisted of one  particular drainage class on all three
 sites. This process was repeated  for each of the four
 drainage classes. The  final  step  was to compare the
 overall  groups and check for frequency of species
 overlap among the four drainage classes. If a species
 was in either  three or four of the groups it  was in-
 cluded in the final suite of species.  In some cases, a
 species was included if it fell into two of the groups if
 the groups were not in  the same category (i.e.,  if a
 species turned up in both sheltered and non-sheltered
 percent  survival  but not in  any of the total height
 classes it would not be a valid  choice for the final
 suite determination). The result  was a performance
 index which gauged the species most appropriate for
 a drainage class  within the coastal plain of  South
 Statistical Methodology
   Analysis of Variance (ANOVA) was used to com-
 pare the means of the variables being considered. All
 differences in sample means were judged statistically
 significant (or not) by comparing them to the varia-
 tion within samples.  Additionally, Duncan's Multiple
 Range Test (a = .05) was used to rank and compare
 the sample means.


 Savannah River Site (SRS)
   Percent survival of non-sheltered trees at SRS var-
 ied greatly  by drainage  class. Overcup  oak and
 swamp tupelo had 50 percent or higher survival rates
 in the poorly drained class. White oak, southern red
 oak, swamp chestnut oak,  overcup oak, cherrybark
 oak, and blackgum had 50 percent or higher survival
 rates in the  somewhat poorly drained class. White
 oak. swamp chestnut oak, cherrybark oak, and over-
 cup oak had 50 percent or higher survival rates on the
 moderately  well  drained  class while white  oak,
 southern red oak, swamp chestnut oak, persimmon,
 overcup oak, cherrybark oak, and blackgum displayed
 the  highest survival rates in the well drained class.
 The sheltered trees at SRS generally averaged much
 higher survival  rates. Swamp  tupelo, overcup  oak,
 swamp  chestnut oak  and blackgum  all had  survival
rates of at least 60 percent in the poorly drained class
with swamp tupelo displaying  100 percent  rate of
survival. In the somewhat poorly drained class, white
oak, swamp tupelo, southern red oak, swamp chestnut
oak, persimmon, overcup  oak, cherrybark  oak and
blackgum all had survival rates of at least 60 percent.
White  oak,  southern  red oak, swamp chestnut oak,
overcup  oak,   cherrybark  oak  and  blackgum  had
survival rates of at least 45 percent in the moderately
well drained class while every species except swamp
tupelo  and hackberry  had at least 70 percent survival
rates in the well drained class.  The  overall  percent
survival  means in  the well drained  and somewhat
poorly drained classes were not significantly different
from each other but were significantly different from
the moderately well  drained and  poorly  drained
classes. The sheltered trees in every drainage  class
had significantly greater survival rates than  did the
non-sheltered trees.
 Total  height  for the trees without  shelters varied
greatly by drainage class. White oak, swamp chestnut
oak, overcup oak and blackgum all attained heights of
90 cm or better in the  poorly  drained  class while
southern red oak, overcup oak  and  cherrybark oak
were at  least  60 cm  tall at the end of three years
growth in the  somewhat poorly drained class. In the
moderately  well drained class, white  oak,  swamp
tupelo  and overcup oak were at least 70 cm in height
while swamp chestnut oak, hackberry, dogwood and
blackgum were at least 45 cm in height in the well
drained class.  Generally, greater heights were found
in the poorly drained  class  among all twelve  species.
For the sheltered  trees in  the poorly drained class,
swamp tupelo, overcup oak and blackgum grew to
heights of  at  least  140 cm.  Only  blackgum and
cherrybark oak attained heights  of at least 95 cm in
the somewhat poorly  drained  class while white oak,
overcup oak, cherrybark oak and blackgum grew to at
least 110 cm  in the  moderately well drained class.
Only persimmon reached a height greater than 90 cm
in the  well drained class. Again, the greatest total
height  was  generally found  in the  poorly drained
class. The poorly drained class produced significantly
taller trees than the other drainage classes. There was
no  significant  difference  between the  total height
means on the moderately well drained and somewhat
poorly  drained classes,  but  the  moderately  well
drained class produced significantly  taller trees than
the well drained class. In  all  drainage classes, shel-
tered  trees  were  significantly taller than  the  non-
sheltered trees.

Webb Wildlife Center (WWC)
  At WWC only one species, cherrybark  oak,  dis-
played a survival rate greater than 40 percent in the
poorly drained class for the non-sheltered species.
Swamp chestnut oak, overcup oak and cherrybark oak
all  had survival rates  greater than 50 percent in the
somewhat poorly drained class,  while only overcup
oak had a survival rate greater than 50 percent in the
moderately well drained class.  In the well drained
class  every species  except water hickory,  swamp
tupelo, mockernut  hickory and blackgum  had  a
greater than 50 percent  survival rate. For  the shel-
tered trees, white oak, swamp chestnut oak, persim-
mon, overcup  oak,  cherrybark oak and blackgum all
had survival  rates  greater  than 50 percent  in the
poorly drained class. In the  somewhat poorly drained
class white oak, southern red oak, swamp chestnut
oak, persimmon, overcup oak and cherrybark oak all
had greater than 50 percent survival rates while white
oak, southern  red oak, swamp chestnut oak, persim-
mon, overcup  oak, cherrybark oak and blackgum dis-
played at  least  60 percent  survival  rates  in  the
moderately well drained class. Every species with the
exception of swamp tupelo  had 60 percent or greater
survival  rates  in the  well  drained class.  The well
drained class percent survival mean was significantly
larger than the means in  any of the other three drain-
age classes. In  every drainage  class,  the  sheltered
trees  displayed significantly greater percent survival
than the non-sheltered trees.
  The non-sheltered trees displayed much variation in
total height. In the poorly drained class, only overcup
oak displayed extreme total height (over  130 cm).
Swamp chestnut oak and overcup oak were the only
species which  were taller than 70 cm in the somewhat
poorly drained class while swamp chestnut oak, over-
cup oak  and cherrybark oak were the tallest in the
moderately well drained class.   White oak,  swamp
chestnut  oak,  overcup oak and cherrybark  oak each
were  at least  80 cm in height  in the well drained
class. For the sheltered trees the  total height was, on
the average, much  greater.  Using a baseline  of 100
cm in height,  swamp tupelo, swamp chestnut oak,
overcup oak, dogwood, cherrybark oak and blackgum
attained heights of  at least 100 cm in  the  poorly
drained class.  Cherrybark oak, overcup oak,  swamp
chestnut oak, swamp tupelo and white oak were at
least  110 cm  in  height in  the somewhat  poorly
drained class while white oak, swamp chestnut oak,
overcup oak, cherrybark oak and blackgum were at
least 100 cm in the moderately  well drained class.
                jouth Carolina Coastal Plain Wetlands
The greatest height was measured in the well drained
class with white oak, southern red oak, swamp chest-
nut oak, overcup oak,  dogwood and cherrybark oak
attaining heights of at least 150 cm. Overcup oak and
swamp chestnut oak each averaged over 200 cm in
total height in this same class. The well drained class
produced significantly taller trees than the  other
drainage classes. The sheltered trees in  all  drainage
classes were significantly  taller than the  non-shel-
tered trees.
Santee Experimental Forest (SEF)
  At SEF,  non-sheltered trees had varying survival
rates among drainage classes. Only overcup oak had a
greater  than 50 percent survival rate  in  the poorly
drained class. Water hickory, swamp tupelo, swamp
chestnut oak, persimmon,  overcup oak,  cherrybark
oak and blackgum had at least 50 percent survival in
the somewhat poorly drained class while white oak,
swamp  chestnut oak, overcup oak, cherrybark  oak
and blackgum had at least 50 percent survival in the
moderately well drained class. For the sheltered trees,
water hickory, swamp tupelo, swamp chestnut oak,
persimmon, overcup oak, cherrybark oak and black-
gum had 50 percent or greater survival in the poorly
drained class while every species except southern red
oak, mockernut hickory, hackberry,  and dogwood had
greater  than 50 percent  survival in the somewhat
poorly drained class. With the exception of dogwood,
hackberry,  mockernut hickory and water hickory, all
species  in the moderately well drained class had at
least 50 percent survival rates. Generally, blackgum,
cherrybark  oak, overcup oak, persimmon, and swamp
chestnut oak did extremely well in survival across all
drainage classes.  There were  no significant differ-
ences among the overall percent survival rates in any
of the  three drainage classes. However, in every
drainage class,  sheltered trees exhibited significantly
greater survival rates than did the non-sheltered trees.
  The greatest total height generally occurred in the
driest drainage class,  and in general  the  heights
attained at  SEF were lower as compared to the other
two  terraces. Swamp  tupelo, persimmon, overcup
oak,  cherrybark  oak,  and blackgum all  attained
heights  of at least 40 cm in the poorly drained class.
Swamp tupelo, southern red  oak, swamp chestnut
oak, persimmon,  overcup oak, and cherrybark oak
were at least 60 cm in total height  in the moderately
well drained class. The sheltered trees in the poorly
drained class were generally shorter  than the shel-
tered  trees  in  the other two  drainage classes with

South Carolina Coastal Plain Wetlands
blackgum. overcup oak and swamp tupelo attaining
heights of at  least 70 cm. In the somewhat poorly
drained  class, swamp  tupelo, cherrybark oak and
blackgum were at least 80 cm in height while  white
oak, swamp chestnut oak, overcup oak,  cherrybark
oak and blackgum grew to at least 120 cm in height.
The moderately  well  drained class  produced total
heights which were significantly greater than the total
heights found on the somewhat poorly drained and
poorly drained classes. Again, in all drainage classes
sheltered trees were significantly taller than the non-
sheltered trees.
 Species Suites
  The results of the performance index used to segre-
 gate  the  species  most appropriate  for a  particular
 drainage class within the South Carolina coastal plain
 are seen  in Tables 2-6. These tables show the pro-
 gression  for the  final suite  determination in each
 drainage class. Table 2 shows the non-sheltered trees
 which met the 45 percent survival rate cutoff point in
 each  terrace for every drainage class. Table 3 is the
 same as Table 2 except that the trees are sheltered.
 Table 4 shows the nonsheltered trees which met the
 40 cm cutoff point for minimum height for every ter-
 race  and every drainage class. Table 5  shows  the
 same data for sheltered trees.  Table 6 shows the spe-
 cies which performed  best in  each of the four classi-
 fications  (nonsheltered  percent  survival, sheltered
 percent  survival,  non-sheltered total height,   and
 sheltered total height)  in each drainage class for each
 terrace.  Table  7 shows the species  best suited for
 each  particular drainage class  within the coastal plain
 of South Carolina.
  The above performance index indicates that cherry-
 bark  oak, overcup oak, persimmon,  swamp chestnut
 oak,  southern red  oak and white oak will perform
 well  in the well drained class. Cherrybark oak, over-
 cup oak, swamp chestnut oak, white oak, persimmon,
 southern red oak, and swamp tupelo are the desired
 species for survival and  growth in the moderately
 well  drained class while blackgum, cherrybark oak,
 overcup  oak, swamp chestnut oak, persimmon,  and
 swamp tupelo  will  perform well in the somewhat
 poorly drained class. Finally, cherrybark oak, overcup
 oak, blackgum, persimmon,  swamp chestnut oak,  and
 swamp tupelo are the  most likely species to exhibit
the greatest survival and growth in the poorly drained
  Cherrybark  oak,  overcup  oak,  persimmon,  and
swamp chestnut oak are all present in each drainage
class  suite and swamp tupelo  is  present  in every
drainage class except  for the  well drained class.
Swamp tupelo is the only species in the study which
is truly tolerant to flooding; thus, it is not unusual that
it performed  as  well as it did in the wetter drainage
classes. The unusual result is that it ended up in the
final suite of species in the moderately well drained
class. The fact that it was planted  and not  naturally
regenerated partially explains this since flooded con-
ditions are necessary for successful  germination of
this species.  Both overcup oak and  persimmon are
moderately tolerant to flooding and both were in the
final suite of species for the somewhat poorly drained
and poorly drained drainage classes. Likewise, they
were  also in the moderately well  drained and well
drained drainage class suites. In contrast, cherrybark
oak is weakly tolerant to intolerant of flooding, but it
did well not only on the well drained and moderately
well drained soils but also on the  somewhat poorly
drained and  poorly drained  soils.  Williams et al.
(1992) reported in a study  on water oak (Quercus
nigra L.), nuttall  oak (Quercus  nuttallii  L.)  and
cherrybark oak  that first year survival and growth
performance  of cherrybark  oak  were  poor on the
hydric soils.  This  contradicts the findings in  this
study. Likewise, swamp chestnut oak is weakly toler-
ant to flooding but it showed excellent survival and
growth on the poorly drained soils. It is not unusual
to find swamp chestnut oak on the fringes of swamps
or in bottoms, but for it  to behave as it  did  in  a
flooded situation is unusual.  Blackgum responded as
expected. It is a moderately flood tolerant species and
it fell into the final suite of species in both the some-
what  poorly  drained  and poorly  drained  drainage
  Several  species  including dogwood,  mockernut
hickory, and water hickory did poorly  on the wetter
classes on all terraces. Both dogwood and mockernut
hickory are intolerant to  flooding and consequently
did not fare well in the wetter drainage  classes. In the
drier  soils these two species did fairly well in terms
of survival and  total height only when a shelter was
present. The best survival  and growth of water  hick-
ory was found in the well drained class across all ter-
  Although  the  three terraces  involved in the study
cannot  be  statistically compared, some  differences

                                                                        South Carolina Coastal Plain Wetlands
Table 4. Non-Sheltered Total Height Species Performance Across All Terraces and Drains Classes.
bg,co,d,h,p,sc,sr, bg,co,oc,p,sc,sr,st,
wo wo
bg,co,d,h,oc,p,sc, bg,co,mh,oc,p,sc,sr,
sr,st,wh,so wh,wo
NA bg,co,oc,sc,sr,st,wh,
bg,co,d,h,oc,p,sc, bg,co,oc,p,sc,sr,st,
sr,wo wh,wo
Table 5. Sheltered Total Height Species Performance Across
bg,co,h,oc,p,sc,sr, bg,co,h,oc,p,sc,sr.st,
st,wo wo
All Terraces and
bg,co,oc,p,sc,st, wo
Drainage Classes.
  SRS      bg,co,d,h,oc,p,sc,    bg,co,d,oc,p,sc,sr,st,
            sr,st,wh,wo          wh,wo
  WWC     bg,co,d,h,mh,oc,p,   bg,co,d,h,mh,oc,p,
            sc,sr,st,wh,wo       sc,sr,wh,wo
  SEF      NA                 bg,co,d,h,mh,oc,p,
  Overall    bg,co,d,h,oc,p,sc,    bg,co,d,h,mh,oc,p,
            sr,st,wh,wo          sc,sr,st,wh,wo
Table 6. Best Species Performance in Each Drainage Class
Sheltered %
Total Height
% co,oc,p,sc,sr,wo

Sheltered Total bg,co,d,h,oc,p,
Table 7. Suite
South Carolina

of Species Exhibiting the
L Coastal Plain.

Best Performance

from the Four Classification Categories.

Across All Study




Drainage Classes in the


 South Carolina Coastal Plain Wetlands
 among  them  are  apparent.  These  differences  are
 important in  explaining the varying  performances
 among  the species within  each site. The seedlings
 experienced rainfall extremes and constant tempera-
 tures  during the three  growing seasons. The first
 growing season averaged 71 degrees Fahrenheit dur-
 ing the eight month period and had 39 inches of rain-
 fall. The second growing season averaged 73 degrees
 Fahrenheit and had 35 inches of rainfall. And finally
 the third growing  season averaged  73 degrees Fahr-
 enheit and had an abundance of rainfall (52 inches).
   At WWC and SEF, the general trend was that the
 trees attaining the  greatest heights were found in the
 better drained classes. At SRS, the general trend was
 the opposite, with the poorly drained class displaying
 the greatest  total  height  and  incremental  height
 growth  while  the well drained  class had the lowest
 total  height and incremental height  growth. These
 results  are  consistent  with  what  is known  about
 flooding and plant growth. There is extensive docu-
 mentation to support the idea that flooding is gener-
 ally injurious to a  plant's ability to grow. Kozlowski
 (1984) also indicated that flooding adversely impacts
 shoot growth  by inhibiting intemode elongation of
 numerous species  including Alnus rugosa, Platanus
 occidentalis, Betula nigra,  Ulmus americana, Ulmus
 alata and Acer rubrum. Further he found that 39 days
 of  flooding  decreased the  height growth of Ulmus
 americana seedlings, with unflooded plants showing
 a rate of growth three to five times greater than the
 growth rate of flooded plants.
   The hydrology of each terrace in this study is an
 important factor to consider. Dramatic differences in
 hydrologic regimes can be seen among these terraces.
 At  WWC and  SEF, the flood  waters tended to be
 stagnant. They lacked topography and slope; hence,
 drainage was limited. At SRS, however, water had a
 constant opportunity  to flow  through  the  poorly
 drained  site.  There was a definite  slope from  the
 mostly well-drained to poorly-drained site. Also, the
 poorly-drained site was drained by an actively flow-
 ing stream.  Kozlowski  (1984)  found that standing
 water  is much more injurious to  trees than moving
 water. Hunt  (1951) found that both height and root
 growth were reduced more  in stagnant water than in
 moving water. In a  study on swamp and water tupelo,
 Hook et al. (1970) found that the height  growth in
 moving  water was about twice the  height growth
 found  in stagnant water treatments. WWC and SEF
 both had poorly drained soil classes in which  the
 water was stagnant while SRS had moving water in
its  poorly drained  class.  The  height differences
(greater average heights at SRS than at the other two
terraces) of species such as  blackgum  (moderately
flood tolerant) and  swamp  tupelo  (flood  tolerant)
among these poorly drained classes was apparent and
the hydrology seems to be one viable explanation.
 There was also a difference in history of site use.
Sites at SEF and WWC had been cleared and treated
for several years prior to planting the seedlings. In
contrast, the SRS was logged in the fall and planted
the following spring.  Also overcup oak (moderately
flood tolerant) and swamp chestnut oak (weakly flood
tolerant) attained much greater heights at SRS than at
SEF, but attained similar heights at WWC. Each of
the three terraces contain drainage classes with vary-
ing degrees of slopes which effect water movement in
the poor drainage classes among the terraces. At SEF
and WWC the terrain on all drainage  classes was
generally flat, void of any  significant slopes. SRS,
however, has three of its drainage classes located on a
fairly significant slope, with the poorly drained class
located at the bottom of the slope. This slope affected
both the surface and subsurface water movement on
this terrace, by increasing the movement of both.

Summary and Conclusions
1.   Survival of all species  at SEF  and WWC was
    highest on the well drained and  moderately well
    drained  drainage  classes. At SRS,  the highest
    survival rates were found on the well drained and
    somewhat poorly drained classes.
2.   On all terraces, the greatest total heights of most
    species were found in the well drained and mod-
    erately well drained drainage classes.
3.   Among the poorly drained classes across  all ter-
    races, SRS displayed greater total heights of most
    species. This is probably due to the effect  of run-
    ning water vs stagnant water at the site.

4.   Drainage class had a significant effect on tree
    diameters, with generally larger diameters found
    in the wetter drainage classes.

5.   Survival  rate of all  species  on  all terraces
    increased significantly with tree shelters.

6.   Total height and  height growth increment of all
    species was significantly larger with tree shelters.

 Although there was overlap in species performance
(both survival  and height)  across the soil drainage
classes, patterns of species suitability were evident.

Species know to be upland species did better on the
WD and MWD sites. Also,  the  most flood-tolerant
species (OC and ST) did best on the PD sites. Except
for extreme site conditions,  it appears that  it takes
longer than 3 years for  site traits to truly weed out
those species that are least suited to the site. Use of
tree shelters appeared to  increase  the ecological toler-
ance of most species on  dry and  wet sites. The long-
term  implications of  these  responses need further
observation and documentation.

 Literature  Cited
Bardon, R. E.  and D. W. Countryman.  1992.  First
  year survival and growth of underplanted northern
  red oak seedlings in southcentral Iowa. Fifth Work-
  shop on Seedling Physiology and Growth Problems
  in Oak Plantings (Abstracts). USDA Forest  Service,
  General Technical Report NC-158.
Dahl, T. E. 1990. Wetland losses in the United States,
  1780s to 1980s. U.S.  Dept.  of Interior, Fish and
  Wildlife Service, Washington, DC. 21 pp.
Hook, D. D., O. G. Langdon, J. Stubbs, and C.  L.
  Brown. 1970. Effect of water  regimes on  the sur-
  vival, growth, and morphology of tupelo seedings.
  Forest Science 16:304-311.
Hook, D.  D.  1984a.  Adaptations  to  flooding  with
  fresh water,  pp.  265-294  In  Flooding and  Plant
  Growth. Ed. T. T. Kozlowski. Academic Press, NY.
Hook, D. D. 1984b. Waterlogging tolerance  of  low-
  land tree species of the South.  Southern Journal  of
  Applied Forestry 8(3): 136-149.
Hunt, F. M. 1951. Effects of flooded soil on growth
  of pine seedlings. Plant Physiology 25:363-368.
Kozlowski, T.  T. 1984.  Flooding and plant  growth.
  Press, NY. 356 pp.
Lantagne, D. O., C. W. Ramm, and D. I. Dickmann.
  1990. Tree shelters increase heights of planted  oaks
  in a Michigan  clearcut. Northern Journal of Ap-
  plied Forestry 7:24-26.
Manchester, E. H., F. G. Roland,  and D. H. Sims.
  1988. Tree shelters show promise for oak regenera-
  tion. USDA Forest Service, Southern Region, Man-
  agement Bulletin R8-MB25.
McConnel, J. L. 1992. The use of tree shelters and
  underplantings for oak regeneration. Oak Regen-
  eration  Symposium   Proceedings,   Sept.  1992.
  Southern Region, USDA Forest Service, Atlanta,

McKnight, J. S., D. D. Hook, O. G. Langdon, and R.
  L. Johnson. 1980. Flood tolerance and related char-
  acteristics of trees of the bottomland forests of the
  southern United States. Reprinted from Wetlands of
  Bottomland  Hardwood Forests. Proceedings of a
  Workshop on  Bottomland Hardwood Forest Wet-
  lands of the Southeast. USDA Forest Service, Pro-
  ceedings Reprint, pp. 29-69.

Minter, W. F., R.  K.  Myers, and B. Fischer. 1992.
  Effects  of tree shelters on northern red oak seed-
  lings planted in harvested forest openings. Northern
  Journal  of Applied Forestry 9:58-63.
Ponder, F., Jr.  1992. Microclimate and growth of oak
  seedlings grown  in treeshelters inside and outside
  of a  shadehouse.  Fifth  Workshop on  Seedling
  Physiology and Growth Problems in Oak Plantings
  (Abstracts). USDA Forest Service, General Techni-
  cal Report NC-158.
Potter, M. J.  1989. Treeshelters: Their effects on
  microclimate  and tree establishment. Paper pre-
  sented at International Conference on Fast Growing
  and N-Fixing Trees. Phillips University, Marburg,
Teskey, R. D.  and  T. M. Hinckley. 1977.  Impact  of
  water level changes on woody riparian and wetland
  communities. Vol. II. Southern Forest Region. U.S.
  Fish  and Wildl.  Serv., Biological  Serv. Program,
  Washington, DC. FWS/OBS-77/59.46 p.
Tiner, R. G. 1991.  The concept of a  hydrophyte for
  wetland identification: Individual  plants  adapt  to
  wet environments. BioScience 41(4):236-247.
Tuley,  G.  1983. Shelters improve  the growth  of
  young  trees in the  forest. Quarterly Journal  of
  Forestry 77:77-87.
Tuley,  G. 1985. The growth  of young  oak trees  in
  shelters. Journal of Forestry 58(2): 181-195.
Walbridge, M. R. 1993. Functions and values of for-
  ested wetlands in the southern United States. Jour-
  nal of Forestry 92:15-19.

Williams, H. M., B. A. Kleiss, M. N.  Humphrey, and
  C. V. Klimas. 1992.  First-year field performance of
  oak species with varying flood tolerance planted on
  hydric and non-hydric soils. Paper presented at the
  Seventh Biennial Southern Silvicultural  Research
  Conference, Mobile, AL. pp. 409-414.

                                           Fish Community Changes in the
                                          Buffalo River,  Buffalo, New York,
                                Following Water  Quality Improvements
                                                                              E. Ann Poole12
                                                                       Elizabeth Trometer1
  Abstract: Changes in the fish community of the
Buffalo River, Buffalo, New York, were assessed
using three indices; the Shannon-Weiner diversity
index (H1), percentage similarity index (PSC), and
an  index of biotic integrity (IBI). Published fish
catch data from four reaches in the lower 9.7 km
of the river for the years 1981 and 1992 were used.
An IBI was developed based on  indicator fishes
particular to  the geographic area  investigated.
Index-specific analyses among survey sites show
that diversity (H') generally increased, as did bi-
otic integrity (IBI scores) indicating a moderate
degree of recovery in the fish community over the
period evaluated.

  Prior to the early 1800's, the lower Buffalo River
provided plentiful habitat for reproduction, feeding
and growth of a myriad of resident  and  migratory
fish, birds, and wildlife (Poole et al. 1994). Since the
mid-1800's, engineering (e.g., dredging, filling, locks,
canals) and compensating works (e.g., bulkheads and
berms) have been constructed to improve navigation
for upstream industries. Cumulatively, these activities
have resulted  in changes of channel substrate, flow
regime, water quality, and riparian soils and vegeta-
tion, thereby degrading aquatic habitat. By the  early
1900's, municipal and industrial waste discharges to
the Buffalo River degraded water quality such that
the lower river became essentially devoid of any fish
life(Greeley 1929).

  In the late 1960's, water quality began to improve as
a result of increased sewage treatment, pollution
abatement programs, and flow augmentation by in-
dustrial cooling  water pumped from Lake Erie and
discharged into the river  (Sweeney  and  Merckel
1972. Sauer 1979). By 1972, fish started to re-inhabit
the lower river (Sweeney 1972). Plant closures and
process shutdowns associated with economic changes
eliminated still more pollution  sources through the
1980's; however, municipal and industrial waste as-
similation continues to  be an important use  of the
lower river (Bode et al. 1993).
 Though water quality  has improved over the past
twenty years and fish again inhabit the lower river,
this section of the river  still does not meet its desig-
nated best use (Class C fishing and fish propagation);
it is considered by the New York State Department of
Environmental Conservation (DEC) to be "impaired"
(NYSDEC 1993a). Impairment is largely due to con-
taminated sediments, groundwater leaching from in-
active  hazardous  waste  sites  (NYSDEC  1993a),
combined  sewer overflows,  and nonpoint sources
(Irvine and Pettibone 1996). Use impairments directly
related  to  fish are: restrictions on consumption;
tainting; degradation of populations; tumors and other
deformities;  and, loss of habitat (NYSDEC 1989).
The fish community, therefore, should be useful in
terms of assessing  progress toward restoring the
aquatic ecosystem of the Buffalo River.
 Although  water quality improvements  have been
documented in the Buffalo River, there have been no
assessments of improvements in the fish community.
Six fishery surveys  of  the river conducted between
1928 and 1975 were limited to annotated species lists
(NYSDEC 1993b). Although these studies did a great
deal to describe patterns of distribution and relative
abundances of fishes throughout the drainage basin,
detailed analyses of fish community structure were
beyond the scope of the  research.
 Structurally and functionally diverse fish communi-
ties can provide evidence of water quality  in that they
are particularly responsive to  environmental  stress
(Hocutt 1981,  Fausch   et al.  1984).  Response to
1 U.S. Department of the Interior, Fish and Wildlife Service, Lower Great Lakes Fishery Resources Office, 405 North French
 Road, Amherst, New York 14228.
: Correspondence Address: 263 Beattie Ave., #4, Lockport, New York  14094-5648.

chemical stress often involves rapid changes in spe-
cies composition that  can translate into changes in
various  aspects of community structure. Direct  and
indirect responses can include:  increased  frequency
of disease; disrupted reproductive processes; reduced
diversity; reduced size (biomass);  retrogressed spe-
cies composition (i.e., opportunistic/tolerant species);
or, debased food prey items (Ford 1989, Gray 1989).
  The objectives of our study were to:

  •  Document long-term changes in Buffalo River
     fish communities;

  •  Develop and apply an index of biotic integrity
     (IBI) for the river; and,

  •  Evaluate  the  degree to  which  the  Shannon-
     Weiner diversity index  (H1), percentage simi-
     larity (PSC), and an index  of biotic integrity
     (IBI) correspond relative to one another.
  We used  H', PSC,  and an IBI because, when used
together, they may permit more reliable assessment of
natural  groupings than any one method  alone.  We
believe this  approach could be especially advanta-
geous in the analysis of disturbed ecosystems and ob-
served changes in biological structure in response to
stress or rehabilitation.

Study Area
  The  Buffalo River  drainage  basin is  located in
western New York and extends nearly 65 km along a
north-northwest axis from the Allegheny foothills to
eastern Lake Erie (Figure 1). The 1,155 km2 drainage
basin is roughly crescent-shaped and lies in two  ma-
jor physiographic provinces,  the Allegheny Plateau
and Erie-Ontario Lake Plain.  The  drainage basin is
divided into three tributary sub-basins (Cayuga, Buf-
falo, and Cazenovia creeks) and has  a dendritic or
branching pattern. The Buffalo River is formed by the
junction of the Cayuga and Buffalo creeks,  13 km up-
stream of Lake Erie;  Cazenovia Creek enters the river
9.7 km upstream of the lake (Figure 2). Mean annual
discharge  (total)  is  about 17  mVsec (Irvine  et al,
  Land use  within the drainage basin varies. Much of
the upper portion of the drainage  basin is  character-
ized by woods and farmland, but prior to joining the
river,  the tributary creeks also pass through several
small  communities and  receive both  industrial  and
municipal discharges. The lower portion of the river
(9 km) is densely urbanized and industrialized. Active
                        Buffalo River Fish Community
industries (e.g., cereal and flour mills, dye plants, sas
compressors,  chemical refineries)  and  abandoned
grain elevators dominate the shoreline.

  The seasonal flow of the  river varies widely and
corresponds to seasonal differences in precipitation.
In general, the region experiences cold and snowy
winters and moderately warm summers, the moder-
ating effect becoming more pronounced nearer the
lake. Current velocities vary considerably in the river;
Lake Erie and the Buffalo River are subject to wind-
driven waves and seiches, which can result in rapid
and large changes in water levels (more than 2 m over
a 9 hr period, Irvine et al.  1992) that are  greater than
the average annual variation  (1.7 m, International
Joint Commission 1981). Locally,  the  passage  of
large vessels  can  also   cause  similar  short-term
changes in water level as well as current velocities
(McNabbetal. 1986).
  The average  annual depth of the river  varies from
about 2 m at the confluence of Cayuga and Buffalo
creeks to over  8.5 m at the mouth. To improve navi-
gation and develop industry, the lower 9.4 km of the
river have been dredged to a depth of at least 6.7 m
below low water datum, with widths ranging from
46 m to 245 m to allow passage of commercial deep-
draft vessels (Poole et al. 1994).

  Published fish catch data from four reaches in the
lower 9.7 km of the river for 1981 (Makarawicz et al.
1982) and 1992 (Kozuchowski et al.  1993) were used
for this study. Reach 1 extends from the mouth at
Lake Erie upstream for approximately 1.5 km; reach
2 extends from reach 1 upstream for approximately
2.5 km; reach 3 extends from reach 2 upstream for
approximately 4.0 km; and,  reach  4 extends  from
reach  3 upstream for approximately  1.7  km  to
Cazenovia Creek (Figure 2).
  In both Markarwicz et al. 1982 and Kozuchowski et
al. 1993, sampling for adult fishes was conducted at
each reach with an electrofishing boat  following a
path parallel to the shoreline in depths less than 3 m.
In 1981, electrofishing was conducted for approxi-
mately four minutes at each reach on one or two days
per   month   from   April   through   December
(Markarawicz et al.  1982). In 1992, electrofishing
was conducted for approximately 30 minutes at each
reach  once   a week  from   April  through  July
(Kozuchowski  et al. 1993).  At the  end  of the elec-
trofishing period, the  fish were identified to spec.es

      Figure 1.  Buffalo River drainage basin.

Figure 2.  The lower Buffalo River.

Buffalo River Fish Community
(Eddy 1969, Werner 1980), measured, and released in
the same reach (Markarwicz et al. 1982, Kozuchow-
skietal. 1993).
  Only data for the months of April through July were
used  in  this  analysis. The data  were  converted
to a standard catch-per-unit-effort measurement  for

  Diversity,  percent similarity and an index of biotic
integrity were  calculated for each reach, for 1981 and
1992. Diversity is a measure of the number of species
(i.e.,  richness) in a community  and their relative
abundances  (i.e.,  evenness). Low diversity refers to
few species  or unequal abundances;  high diversity to
refers to many species or equal abundances. Diversity
was calculated for each sampling date at each reach
using the Shannon-Wiener  diversity  index (Wash-
ington 1984):
                      Pi   =  n/N
        S             S   =  no. of species
                      n,   =  no.  of organisms in
                             the i* species
                      N  =  total no. of
  The resultant values  were then  averaged by reach
and by  year for 1981 and 1992. Statistical compari-
sons of the  fish community at each  reach were made
between years  using  the  student's  t-test (Excel,
Microsoft Corporation).
  Percent similarity (PSC) is an index used to meas-
ure community similarity (or dissimilarity). The PSC
was used to compare the similarity  of the fish com-
munity  at each reach between  1981 and  1992. The
PSC was calculated using the Whittaker and Fair-
banks' index (Washington 1984):
      i= 1
                      K  =
PSC= 100-0.51 la-b| a   =
                      b   =
                              the total number of
                              species combined
                              from 1981 and 1992
                              relative percentage
                              of species i from
                              relative percentage
                              of species i from
  The PSC comparing 1981 and 1992 was calculated
for each  reach. Percent  similarity  ranges  from  0,
when two samples contain no species in common, to
100, when  the  two samples  are identical  in both
species and individual abundances. The PSC can fail
if the relative proportions of taxa remain the same but
the overall abundance changes; this fault may not be
disqualifying in fisheries studies of polluted rivers as
environmental changes tend to alter dominance rela-
tionships. If the balance of taxa is nearly identical
then the  communities may be functionally the same
(Washington 1984).
  The Index of Biotic Integrity (IBI) is a composite
index  that integrates characteristics of the commu-
nity, population,  and individual organisms. IBI pa-
rameters  or  metrics  include  species  richness,
indicator  taxa, trophic  guilds, fish abundance,  hy-
bridization, and diseases and  anomalies. An IBI for
the Buffalo River  was developed on the basis of
methods  described in Karr (1981) and Lyons (1992).
Eleven metrics similar to those developed by Karr et
al. (1986) were used (Table 1). The twelfth metric de-
fined by  Karr, proportion of individuals with disease,
tumors,  fin damage, and skeletal anomalies, was
dropped  because  this information  was not available
for the years evaluated. The sixth metric was rede-
fined from proportion of individuals as green sunfish
to proportion of individuals as carp, goldfish, or carp-
goldfish  hybrids,  because green sunfish do not occur
in the Buffalo River watershed.
  Each metric consists of fishes that belong to a single
structural or functional group. Individual  species may
belong to more than one group, and hence, contribute
to  more  than  one  metric. Once  the  metrics  were
defined,  all the  fish species found in  the Buffalo
River from both sampling periods were characterized
according to taxonomic groups, feeding guilds, and
tolerance to environmental degradation (Table 2).
  Physical characteristics also influence community
structure  and function.  An important physical factor
that influences community structure and function  is
stream size (Lyons  1992). In  general,  the total num-
ber of fish and the  number of species increase with
stream size. To account for this change, metrics 1, 2,
3, 4, 5, and 10 were modified based on the stream-
order of each reach.
  Stream-order is a classification system used to indi-
cate progressive increases in stream size (Strahler
 1957,  as cited in  Karr et al.  1986). The smallest
streams,  highest in the  watershed,  are first-order.
When two first-order streams join, they form a stream
of the second-order; when two second-order streams
join, they form a stream of the third-order, etc.
Applying this classification  system to the Buffalo

                         *         ,    acc,
York (modified from Karr et al.  1986).
                                                      iC 'ntegrity metrics for the Buff*l° River. New
           Scoring criteria
      5         3          1
  Species richness and
1. Total number of fish species

2. Number of darter species

3. Number of centrarchid species
  (excluding Micropterus spp.)

4. Number of sucker species

5. Number of intolerant species

6. Percent carp, goldfish, or carp x goldfish
                                                                          Expectations for metrics 1 through 5
                                                                          vary with stream order.  See Appendix
                                                                                         5- 20
  Trophic function         7. Percent omnivores1

                         8. Percent macroinvertivorous cyprinids2

                         9. Percent piscivores3
                                                    <21       22-45        >45

                                                    >45       45-20        <20

                                                     >5         5-2         <2
  Fish abundance
10. Number of individuals collected, expressed as

11. Percent hybrids
Expectations for metric 10 vary with
stream order.  See Appendix A.
                                                                                         >0- 1
 '  Omnivores - species with diets composed of at least 25 percent plant material and >25 percent animal material.
 2  Macroinvertivorous cyprinids - Cyprinidae spp. with diets composed of >75 percent macroinvertebrates (insects,
 crustaceans, annelids, etc.).
 3  Piscivores - species with diets composed of  >75 percent fish.
 4  CPUE =  catch-per-unit-effort  (total number of individuals / total  electrofishing effort, in minutes).

      Table 2.  Taxonomic group, trophic guild, and identification of intolerant species of adult
      fishes collected in the Buffalo River watershed since 1928.  Taxa: A = Catostomid; D  =
      Darter; E = Centrarchids, excluding Micropterus SDP.: Y = Cyprinid.  Trophic guild:  0 =
      omnivore; P = planktivore;  F = piscivore; M = macroinvertivore.
Scientific Name
Leoisosteus osseus
Alosa oseudoharenous
Dorosoma ceoedianum
Carassius auratus
Cyprinus carpjo
Notemioonus crvsoleucas
Nocomis micropooon
Cyprinella spiloptera
Notropis hudsonius
Notropis atherinoides
Pimephales notatus
Notroois cornutus
Semotilus atromaculatus
Carpiodes cvprinus
Catostomus commersoni
Hypentelium nigricans
Moxostoma sp.
Ameiurus nebulosus
Ameiurus melas
Ictalurus punctatus
Noturus flavus
Esox lucius
Esox masouinonqy
Osmerus mordax
Oncorhvnchus mvkiss
Q. kisutch
0. tshawvtscha
Salmo trutta
Common Name

longnose gar

gizzard shad

carp x goldfish
golden shiner
river chub
spotfin shiner
spottail shiner
emerald shiner
bluntnose minnow
common shirfer
creek chub

white sucker
northern hog sucker
redhorse spp.

brown bullhead
black bullhead
channel catfish

northern pike

rainbow smelt

rainbow trout
coho salmon
Chinook salmon
brown trout



























     Percopsis omiscomavcus

 Morone americana
 Morone chrvsops

 Ambloplites ruoestris
 Leoomis gibbosus
 Leoomis macrochirus
 Micropterus dolomieu
 Micropterus salmoides
 Pomoxis annularis
 Pomoxis nioromaculatus
 Leoomis oulosus

 Etheostoma niqrum
 Perca flavescens
 Percina caprodes
 Stizostedion vitreum

 Aolodinotus orunniens
        white perch
        white bass
        rock bass
        smallmouth bass
        largemouth bass
        white  crappie
        black crappie
        johnny darter
        yellow perch

        freshwater drum



                  Denotes nonindigenous  species
 Table 3.  Total index  of biotic integrity scores, integrity classes, and the attributes of those
 classes (modified from Karr et al.  1986).
  Total IBI score
  (sum of the 11 metric ratings)
Integrity class
                 Comparable to the best situations without human-disturbance; all
                 regionally expected indigenous species for the habitat and stream
                 order, including the most intolerant forms, are present with a full
                 array of age (size) classes; balanced trophic structure.
            38 -46
            11 -19
Very poor
Species richness somewhat less than best situations, especially due
to the loss of the most intolerant forms; some species are present
with less than optimal abundances or size distributions; trophic
structure shows some signs of stress.

Signs  of additional deterioration include loss of intolerant forms,
fewer species, highly skewed trophic structure (e.g., increasing
frequency of omnivores and carp, goldfish, or carp x goldfish); older
age classes of top piscivores may be rare.

Dominated by omnivores, tolerant forms, and habitat generalists;
few piscivores; growth rates and condition factors commonly
depressed; hybrids and diseased fish often present.

Few fish present, mostly non-indigenous or tolerant  forms; carp,
goldfish, or carp x goldfish common; disease, parasites, findamage,
and other anomalies regular.
                                                 Repeated sampling finds no fish.

 Buffalo River Fish Community
 River,  reaches  1  and 2  were classified a stream-
 order 6 and reaches 3 and 4 were classified a stream-
 order 5.
  Fisheries  data for  the whole basin collected  since
 1928 were  used to  develop plots of species versus
 stream-order, and  the  scoring  criteria  for   those
 6 metrics was estimated  from those plots  (Appendix
 A). The scores  from each metric for each year and
 reach were  summed to give the overall IBI score for
 each reach  and year. The  maximum possible IBI
 score is 55  and  the lowest, 11. The IBI scoring  range
 was divided into  5  sub-ranges, or  integrity classes,
 qualified  as "excellent" (high IBI score) to  "very
 poor"  (low  IBI  score) (Table 3). The IBI scores for
 each reach were compared between years.

  In comparing 1981 and 1992, there was virtually no
 change in diversity (H)  at reach  1  (river mouth);
 however, there was an increase in H' at reaches 2,  3,
 and 4  (Figure  3). The  increase was  significant  at
 reach 3 (P<0.0001) which went from the lowest H' in
 1981 to the highest H' in  1992. In 1981,  there was a
 trend  of decreasing diversity  from  reach  1  (the
 mouth) going upstream to reach  3. In 1992, H' was
 fairly similar for all sites.
  Percent similarity (PSC) indicated that, when com-
 paring  1981  and 1992, the fish community of reach 2
 remained the most similar  (Figure  4). The PSC for
 reaches 1, 3, and 4 were less similar.
  The IBI improved from 1981 to 1992 at all reaches
 (Figure 5). EBI scores reveal significant improvement
 at reaches 2, 3, and 4, although the scores still reflect
 some degree of stress. The integrity class remained
 "fair"  for reach 1, with only  a slight (6%) increase
 from 1981  to  1992. The IBI  score increased most
 dramatically (82%) at reach 2 with a change in integ-
 rity class  rating  from "very poor"  to  "fair." The
 increase was largely due to the decline in percent hy-
 brids (metric 11): a property not addressed by H'. The
 IBI score for reach  3 increased by 10 points  (37%)
 and for reach 4 increased by 14 points (61%), with
 integrity class changes from "poor" to "fair." In  1981,
 there was a trend from the mouth going upstream  of
 decreasing biotic integrity;  in  1992, biotic integrity
 was similar  for all  reaches. Reach 1  displayed the
 least difference in IBI score between 1981 and 1992.

 Closer inspection reveals that the improvements  in
IBI scores at reaches 2, 3, and 4 were due primarily to
increases in the number of: total species (metric 1);
centrarchid species (metric 3); sucker species (metric
4)(reaches  3 and 4);  and,  intolerant  species (metric
5)(reaches 2 and 3), and decreases in the percentage
of omnivores (metric  7)(reaches 2 and 4)(Figure 6).
Overall,  IBI  metric  3   (centrarchids,  excluding
Micropterus) showed the  greatest  net increase in
score from  1981 to 1992.

  Indices of diversity  (H1)  and biotic integrity (IBI)
reveal  discernable improvement in  the fish commu-
nity  of the  lower Buffalo  River between  1981 and
1992.  Diversity  generally  increased (significant at
reach 3), as did biotic integrity (significant at reaches
2, 3, and 4) during the period evaluated. Total num-
ber of species  (richness)  increased  and abundance
(CPUE) remained the same, except at reach 1 where
both declined slightly.
  Overall, PSC's were not consistently compatible
with observed increases in H' and the IBI, especially
at reaches  1 and 2. The degree of similarity in fish
community  composition (PSC) between  1981 and
1992 was very low at  reaches  1,3, and 4, and moder-
ate at reach 2,  indicating a high degree of heteroge-
neity  over  time.  Intuitively,  when  there  is  a
significant  change in  IBI  score between years, one
would expect a low PSC; when there is no significant
change in IBI score, a high PSC might be expected.
This  premise  is  not  substantiated  by results  for
reaches 1 and 2.
  At reach  1, in 1981, 61  percent of the total catch
was  composed  of four species, ranging from 10 to
29 percent each. In 1992, 62 percent of the total catch
was  composed  of the  same four species;  however,
one migratory lake species overwhelmingly predomi-
nated (emerald shiner, 50 percent of total catch). The
prevalence of this species in 1992, weighed heavily in
the PSC calculation in which  the relative proportions
of taxa changed significantly. The  PSC,  therefore,
was relatively low (37.8).

  In contrast, values for H' and IBI were relatively un-
changed at  reach 1; furthermore, the species compo-
sition  of the community at reach 1 was  similar to
nearshore-lake communities. This may indicate niche
overlap  and transition in community  composition
between the river and Lake Erie, potentially affecting
IBI scores  within the  individual metrics.  If so, the
succession  in  habitat structure  from riverine to

                   2.50 T

 Figure 3.  Calculated Shannon-Wiener diversity index (H') of reaches 1 through 4 for 1981 and
 1992 (+/- standard deviation).
70.0 H
:£ 60.0 -
^ 50.0-
1 40.0-

•£ 30.0-
E 20.0-
S. 10.0-
00 -

: :•:::;:





i| j!;




 Figure 4.  Calculated Percent Similarity Index (PSC) at each reach between 1981 and 1992.
Figure 5.  Calculated IBI scores at reaches 1 through 4 for 1981 and 1992. (The number at the
top of each bar is the actual IBI score.)

                            123456789   10  11
                            123456789  10  11
                            123458789  10  11
            Figure 6.  Individual metric scores at each reach for 1981 and 1992.

                                                                              Buffalo River Fish Communm
estuarine would require modification of the IBI met-
rics used.

  At reach 2, in  1981, 76 percent of the total catch
was  composed of four species; two-thirds of which
was  due to two  migratory lake  species (33  percent
emerald shiner and 11 percent gizzard shad). In 1992,
the same four species accounted for 80 percent of the
total catch, of which emerald shiner and gizzard shad
accounted  for  37  and  23  percent,   respectively.
Twelve species were found in 1992  (accounting for
about  19 percent of  the  total catch) that were not
found  in  1981.  The presence  of these additional,
relatively  uncommon species in  1992, however, did
not weigh heavily in the PSC calculation because the
relative proportions of the predominant taxa remained
the same.  The PSC, therefore,  was relatively high
  In contrast, since species were grouped by structure
or function in the IBI, the effect of the uncommon
species on IBI scores was pronounced. For example,
at reach 2, centrarchids accounted for about 3  percent
of the total catch in 1981 and 10 percent of the total
catch in 1992. The number of species, however, in-
creased from 1 in 1981 to 6 in 1992  (including 2 in-
tolerant species)  resulting in a 4 point increase in IBI
  In the initial stages of environmental improvement,
one can expect increasing diversification of commu-
nity  structure and progressive changes in community
function.  A  secondary response  in  the community
may follow, caused by the re-colonization of an intol-
erant species or altered  interactions (i.e., competi-
tion).  In  the  case  of  the  lower  Buffalo  River,
however, quantifiable relationships between recovery
of the fish community and water quality improvement
cannot be made.
  Water chemistry monitoring data  (i.e., dissolved
oxygen, temperature, and conductivity)  indicate dis-
solved oxygen increased between 1970 and 1980. In
the early  1970's, migratory Lake Erie fishes  were
found in upstream reaches of the river where they had
not been seen for at least thirty years (Sweeney and
Merkel  1972). Since 1981, there has been no apparent
improvement  in  water quality (Appendix B); there-
fore, recovery of the fish community appears to lag
behind water quality improvements by 10 to 15 years.
Improvements in the fish community at reaches 2, 3,
and 4 may have occurred  as a result of re-population
by riverine and migrating nearshore-lake fishes  (e.g.,
crappies and  emerald   shiner,   respectively)   and
 increased reproductive success. Long-term collection
 and analysis of larval, as well as adult, fish data will
 be necessary to pursue this line of inquiry.

  In this study, we used a new approach to assess fish
 community changes in the lower Buffalo River based
 on three common-used indices. Our results show that
 the combined  use of H', PSC, and  an IBI minimized
 individual deficiencies within each index and avoided
 potential  misinterpretations of the  data. In order to
 enhance similar studies of the Buffalo River in the
 future, we offer the following suggestions:
  Analytical methods should be developed prior to
 conducting sampling in order to gather  relevant
 information.  Our study was  not designed  with the
 intent of applying an IBI to the results, therefore  sev-
 eral fish  community  attributes  were not measured.
 Data on the incidence of external anomalies (i.e., de-
 formities,  fin erosion, lesions, tumors) would be es-
 pecially useful given existing use impairments. Water
 chemistry data (e.g., dissolved oxygen, temperature,
 biochemical oxygen  demand, suspended sediments)
 would also be useful in the absence of intensive water
 quality monitoring.
  The IBI metrics  used for  this study should be
modified  before applying them  to  other large,
lake-influenced  rivers.  Five  metrics that we used
 were species-poor relative to the overall fish commu-
nity. We believe that this may have caused  the met-
rics to be  more insensitive to environmental changes
than when applied to smaller river systems. Also, a
metric for larval fish (expressed as CPUE) should be
added.  Larval  fish data would be useful to  evaluate
relative reproductive success.
  Metric 2. Number of Percidae species. Four species
of percids, including one species of darter, exist in
the lower Buffalo River.
  Metric 3. Number of centrarchid species. Eight  spe-
cies of centrarchids  exist in the lower Buffalo River
 when two  Micropterus species are included.
  Metric 4. Number of benthic species. Four species
 of Ictaluridae,  in  addition to the four species of Ca-
 tostomidae  (suckers), exist  in the  lower Buffalo
  Metric 8. Percent  macroinvertivores. Eighteen  spe-
 cies of macroinvertivores, including two  cyprinid
 species, exist in the lower Buffalo River.
  Metric  11. Percent non-indigenous  species.  Nine
 non-indigenous species, including hybrid carp x gold-
 fish, exist in the lower Buffalo River.

 Buffalo River Fish Community
 Literature Cited
 Bode, R. W.. M. A.  Novak, and L.E.  Abele. 1993.
   Twenty-year trends in water quality  of rivers and
   streams in New York State based on macroinverte-
   brate data 1972-1992. New York State Department
   of Environmental Conservation. Division of Water.
   Stream Biomonitoring Unit. Albany, NY. 196 pp.
 Eddy, S. 1969. How  to know the freshwater fishes.
   Wm. C. Brown Company. Dubuque, IA. 286 pp.
 Fausch, K.  D.. J. R. Karr, and P.  R. Yant. 1984. Re-
   gional application of an index of biotic integrity
   based on  stream fish communities. Transactions of
   the American Fisheries Society 113:39-55.
 Ford, J.  1989.  The  effects  of  chemical stress on
   aquatic species composition and community struc-
   ture, pp. 99-144,  (eds.) In  S. A. Levin, M.  A. Har-
   well, J. R. Kelly and K. D. Kimball, Ecotoxicology:
   problems  and approaches. Springer-Verlag. New
   York, NY.
 Greeley, J. R. 1929. Fishes of the Erie-Niagara water-
   shed, pp.  150-180,  In A biological survey of the
   Erie-Niagara System. Suppl. 18th Annual Report
   (1928),  New  York  Conservation  Department.
   Albany, NY. 244  pp.
 Gray. J. S.  1989. Effects  of environmental stress on
   species rich assemblages. Biological Journal of the
   Linnean Society 37:19-32.
 Hocutt, C. H. 1981. Fish as  indicators of biological
   integrity. Fisheries 6(6):28-31.
 International Joint Commission. 1981. Lake Erie wa-
   ter level study. Appendix A, Vol. 1,  Lake  Regula-
   tion. International  Joint Commission.  Lake  Erie
   Regulation Study  Board. Winsor, Ontario.
 Irvine. K. N., G. P.  Stein,  and J. K. Singer. 1992. An
   environmental guidebook to the Buffalo River. In
   Proceedings of the International  Symposium on
   Environmental Dredging, September 30 to  October
   2. 1992, Buffalo, NY. 30pp.

 Irvine, K. N. and  G. W.  Pettibone. 1996. Planning
   level evaluation of densities and sources of indica-
   tor bacteria in a mixed  land use watershed. Envi-
   ronmental Technology 17:1-12.

 Karr, J. R.  1981. Assessment of biotic integrity using
   fish communities. Fisheries 6(6):21-27.

 Karr. J. R., K. D. Kurt, P. L. Angermeier,  P. R. Yant,
  and I. J. Schlosser.  1986. Assessing biological in-
  tegrity in  running  waters,  a method  and  its
  rationale. Illinois  Natural History Survey Special
  Publication 5. Champaign, IL. 28 pp.

Kozuchowski,  E. S., E. A. Poole, and C. E. Lowie.
  1993. Buffalo River fisheries assessment: report on
  the results of the 1992 larval and adult fish survey.
  U.S.  Department of the Interior, Fish and Wildlife
  Service, Lower  Great  Lakes  Fishery  Resources
  Office, Amherst, NY. Administrative Report 93-03.
Lee, C. R., D. L.  Brandon,  J.  W. Simmers,  H.E.
  Tatem, and  J. G. Skogerboe.  1991.  Information
  summary,  area  of concern: Buffalo River,  New
  York. Misc.  paper EL-91-9,  Department of the
  Army, Corps of Engineers, Vicksburg, MS. 29 pp.
Lyons, J.  1992. Using the index  of biotic integrity
  (IBI) to measure  environmental quality  in warm-
  water streams of  Wisconsin.  General Technical
  Report  NC-149.  United  States  Department of
  Agriculture.  Forest Service, North Central Forest
  Experiment Station, St. Paul, MN. 51 pp.
Makarewicz, J.C., R.C. Dilcher, J.M. Haynes, and K.
  Shump. 1982. Biological survey: Buffalo River and
  outer harbor of Buffalo, NY.  Final  Report,  Vol-
  umes 1 and  2. State University College  at Brock-
  port, Biology Department, Brockport, NY.
McNabb, C. D., T. Batterson, J. Craig, P. Roettger,
  and M. Siami.  1986. Effects  of commercial ship
  passage on emergent wetlands in Lake Nicolet, St.
  Marys River, 1984 (Part II)  In Environmental  base-
  line  studies  during  1984  of Lake Nicolet and
  Munuscong Bay, St. Marys River, MI, in relation to
  proposed extension of the navigation season. C.R.
  Liston  and C.D. McNabb,  principal  investigators.
  U.S. Department of the Interior, Fish and Wildlife
  Service, Twin Cities, MN and U.S. Army Corps of
  Engineers, Detroit District  Office, MI. Biological
  Report 86(3). 78 pp.

Mikol, G., K.  Roblee, and M. Wilkinson. 1993. Fish
  and wildlife habitat inventory and assessment  of the
  lower Buffalo River Watershed. New York State
  Department   of   Environmental    Conservation,
  Albany, NY. 86 pp.

New York State Conservation Department (NYSCD).
  1929. A biological survey  of the Erie-Niagara sys-
  tem.  Supplemental  to the   Eighteenth  Annual
  Report, 1928.

New York State Department  of Environmental Con-
  servation (NYSDEC). 1989. Buffalo River remedial
  action  plan.  New  York  State  Department  of

                                                                            Buffalo River Fish Community
  Environmental Conservation, Division of  Water,
  Buffalo, NY.

New York State Department of Environmental Con-
  servation (NYSDEC). 1990.  Rotating intensive ba-
  sin studies, Appendix A, water column. New York
  State Department of Environmental Conservation,
  Division of Water, Albany, NY.  114 pp.

New York State Department of Environmental Con-
  servation (NYSDEC). 1993a. Priority water prob-
  lem list - Region 9. New York State Department of
  Environmental  Conservation, Division  of  Water,
  Buffalo, NY.
New York State Department of Environmental Con-
  servation (NYSDEC). 1993b. Reprint of the NYS
  rivers  inventory  resource report for the Buffalo
  River Watershed. New  York State Department of
  Environmental  Conservation, Division  of  Water,
  Raybrook, NY.
Poole,  E. A., E. S. Kozuchowski, and C. E.  Lowie.
  1994. Fish and wildlife habitat restoration on the
  Buffalo River, Buffalo, New York.  Prepared  by
  U.S. Department of the Interior, Fish and Wildlife
  Service, Lower Great Lakes Fishery Resources Of-
  fice, Amherst, New York, for Erie County  Depart-
  ment of Environment and Planning, Buffalo, NY.
 Sauer,  D. E. 1979. An environmental history of the
  Buffalo River. Text  of  presentation, 50th meeting,
  New York Water Pollution Control Association,
  New York City, January 1979. Buffalo Color Cor-
  poration, Buffalo, NY. 88 pp.

Sweeney, R. A. 1971. Survey of benthic macroinver-
  tebrates and analysis of water and sediments from
  the Buffalo River 1970. State University College at
  Buffalo, Great Lakes Laboratory, Buffalo, NY. 24

Sweeney, R.  A. 1972. River on the  mend. Limnos

Sweeney, R. A. and C. Merkel. 1972.  Survey of ben-
  thic macroinvertebrates and analysis of water and
  sediments from the Buffalo River, 1972. State Uni-
  versity College at Buffalo,  Great Lakes Laboratory.
  Buffalo, NY. 26 pp.
Ward,  C. A .B. 1980. A survey of  the crustacean
  zooplankton in the Buffalo River,  1979, Buffalo
  River Study X. State University College at Buffalo,
  Great Lakes Laboratory, Buffalo, NY. 103 pp.
Washington, H. G. 1984. Diversity, biotic, and simi-
  larity indices, a review with  special relevance to
  aquatic ecosystems. Water Resources 18(6):653-
Werner, R. G. 1980. Freshwater fishes of New York
  State: a  field guide. Syracuse University  Press.
  Syracuse, NY. 186pp.


              APPENDIX A

Scoring criteria for metrics 1, 2, 3, 4, 5, and 10
           based on stream order.

Figure A1.   Total number of fish species vs. stream
  order in  the  Buffalo  River Watershed (Metric  1).
                         3      4
                      Stream Order
Figure A2.   Total number, of darter species  vs.  stream
  order in the Buffalo  River Watershed (Metric 2).
                         3      4
                      Stream  Order


      Figure A3.   Total number of centrarchid  species
      (excluding  Micropterus) vs. stream  order in  the
            Buffalo River Watershed  (Metric  3).
                      2.3      4      5

                           Stream Order
Figure A4.   Total number of sucker species  versus  stream
    order  in the Buffalo River watershed  (metric 4).
    Figure A5.   Total number of intolerant species vs.
 stream order in the Buffalo River Watershed (Metric 5).
                            3      4
                          Stream Order
 Figure A6.   Number  of  individuals in sample  (total fish
  collected  for year /  total electrofishing minutes for
  year) vs.  stream order  in the Buffalo River Watershed
                      (Metric  10).

             APPENDIX B

Water chemistry monitoring data, 1970 to 1993.



Dissolved O2 for May
12 -r
6 -
* 1
* * t


	 1 I W 1 	 1 	 1 	 1 	 1 	 1 r- i .1
65 67 69 71 73 75 77 79 81 83 85 87 89 91

— i — i
93 95

Dissolved O2 for June
10 -J

8 -
6 -
E 4 -
2 -
^ A

" *

65 67 69 71 73 75 77 79 81 83 85 87 89 91 93 95
Dissolved O2 for August
5 j

4 -
3 -
* 2-
1 -

0 •
I *
1 *


65 67 69 71 73 75 77 79 81 83 85 87 89

91 93 95

Figure B1.  Dissolved oxygen (mg/l) in the Buffalo River measured near the Ohio Street
Bridge (reach 2) during the months of May, June and August (Lee et al. 1991, Mikol et al.
1993, NYSDEC 1990, Sweeney 1970, Sweeney and Merkel 1972, Ward 1980).



Temperature for May
25 -,

5 -
0 •

•« i
* A.

65 67 69 71 73 75 77 79 81 83 85 87 89 91

93 95

Temperature for June
25 j

20 -
M .c
3 15 -
"5 in .
0 1U
5 -
« * t
1 1 T 	 1 	 1 	 1 	 1 	 T 	 -f 111.-
65 67 69 71 73 75 77 79 81 83 85 87 89 91

— i i
93 95

                             Temperature for August
                     30 ]
                     25 ]

                  «  20

                  1  M
                  O  10-

                       65 67 69 71 73 75 77 79 81  83 85 87 89 91 93 95

Figure B2. Temperature in the Buffaio River
2) during the months of May, June and August (Lee et al. 1
NYSDEC 1990, Sweeney 1970, Sweeney and Merkel 1972, wara



Conductivity in May
500 -,



^ -

67 69 71 73 75 77 79 81 83 85 87 89 91 93


Conductivity in June
400 j



0 -
* •


70 72 74 76 78 80 82 84 86 88 90 92




Conductivity in August
450 T
50 -
0 -
• • ^

w • i - i 	 1 	 1 — — j 	 1
69 72 75 78 81 84
Figure B3.  Conductivity in the Buffalo River measured near the Ohio Street Bridge (reach
2) during the months of May, June and August (Lee et al. 1991, Mikol et al. 1993,
NYSDEC 1990, Sweeney 1970, Sweeney and Merkel 1972, Ward 1980).

Works in

"Seek out diverse friends. By just
hanging out with them, cross-
pollination of ideas occurs."
                      Mary Ann Norris
                      Mattel Strategic Planner
                      El Segundo-1997

                                   Complexities  of Species Conservation
                                       Under the Endangered Species Act:
                                    Ecosystem Management in  the Cedar
                                                                       River Watershed
                                                                       W. James Erckmann1
  Ecological restoration and ecosystem management
 are essential parts of a Habitat Conservation Plan
 for the Cedar  River Municipal Watershed, near
 Seattle, Washington, which  is being prepared  by
 Seattle Water under Section  10 of the  Endangered
 Species Act. This paper will  discuss the major ele-
 ments of the plan (which may change), the proce-
 dural challenges to gaining  approval,  and lessons
 learned from the experience.

  The HCP is intended to provide greater certainty
 for the region's water supply in the face of potential
 listing of a number  of fishes affected by  Seattle
 Water operations.  The plan  attempts  at once  to
 sustain  ecological  values,  including protection  of
 many at-risk species and their habitats, and human
 uses of the environment,  including the drinking
 water for nearly one-fourth of the state population.
  Like  many other regional planning efforts, how-
 ever, this plan  requires approval of local elected
 officials, and federal government must be able to
 give assurances that implementation  of the plan will
 provide a reasonable amount of certainty in water
 supply  operations  for  Seattle  Water, in  turn
 requiring  that  the plan meet its  ecological  ob-
  While the commitments inherent in this HCP have
 substantial  environmental  benefits, the changing,
 unpredictable, and often conflicting political and
 legal  contexts  under which  the  plan  is  being
 developed  make  these commitments  difficult  to
 achieve. Uncertainties in state water resource plan-
 ning,  ambiguities  inherent  in  the  ESA,  legal
 challenges  to   ESA  regulations,  and  potential
 Congressional changes to the act combine to work
 against  effective  ecosystem  management for  the
 many at-risk species addressed by the plan.
 Species of particular concern in the HCP include
several anadromous salmonids, resident bull trout,
northern  spotted  owl, and marbled murrelet. Res-
toration efforts focus primarily on protecting and
restoring natural  processes that create or maintain
aquatic, riparian,  and old-growth forest habitats in
the 90,000-acre montane watershed, and recovery of
depressed populations of anadromous salmonids af-
fected by diversion of water from the Cedar River.
 An   ecological   reserve   comprising   nearly
65 percent of the watershed, and all existing late-
successional  forest, will provide habitat protection
and connectivity.  Silvicultural  and  timber harvest
guidelines outside the reserve, as  well as limited
experimental restoration of second growth inside
the reserve,  should result in long-term improve-
ments in forest structure and habitat complexity in
second-growth forests over time, and restoration of
important riparian functions.
 Watershed  analysis forms the basis for  planning
forest road improvements and abandonment to re-
duce sediment loads to streams, and replacement of
stream crossing structures that block fish passage.
Forest  practices are  also  constrained to minimize
the likelihood of landslides  and erosion  resulting
from cumulative impacts in specific basins.
 The HCP  will  also  include commitment to  a
specific instream flow regime to protect fish habitat,
and other measures to protect and  recover anadro-
mous salmonids.  Because  good  information  on
many of these and other key  species is sorely lack-
ing, the plan will  also include elements of adaptive
management, which presents additional challenges
in the current regulatory arena.
 Lessons in this process are still being learned, but
it  is already clear that the  level  of coordination
among resource agencies needed for such  a plan to
be successful is not typical. The need to accommo-
date varying interests  and  overcome institutional
barriers highlights  the value  of regular,  open
1 Seattle Water Department, 18015 SE Lake Youngs Road, Renton, Washington 98058.

  Selected Abstracts
  communications with all organizations, at both staff       a problem-solving, partnership mode can be adopted
  and management levels. This experience suggests       in the process.
  that ecosystem planning will be most effective when         sEA/9732400iodoc

                                     Restoration of Biodiversity from the
                                        Genetic to Landscape Levels: The
                                            Elwha River Restoration Project
                                                                              Brian Winter1
                                                                          Sarah Bransom1
                                                                              Jeff Bohman2
  The Elwha  River Restoration Project  offers  an
opportunity for the restoration of biodiversity at all
levels of the biological hierarchy -  from the genetic
level  to the landscape  level (Winter and  Hughes
1995). Most (83 percent) of the Elwha River Basin
lies within Olympic National Park and is essentially
pristine. Since the  1920's, two hydroelectric dams
have prevented anadromous salmon and trout from
reaching over 90 percent of their historic spawning
and rearing habitat; the result is diminished  and ex-
tinct fish stocks and a dearth of nutrients  once pro-
vided by decomposing salmon carcasses to the largest
drainage within Olympic National park (NPS 1995).
In addition, the dams have inundated important river-
ine and terrestrial habitat and trapped sediment within
the reservoirs, resulting  in the armoring of the river
bed below the dams, reduction in the size and com-
plexity of the estuary, and steepening of nearshore
marine areas. Studies for the Elwha Report (DOI et
al. 1994) and Environmental Impact Statements for
dam  removal  indicate dam removal and sediment
management are feasible and would provide  good to
excellent chances for full  restoration of the  anadro-
mous fishery and the riverine ecosystem (NPS 1996).
Restoration on smaller  scales  may  not prevent the
fragmentation and ultimate loss of native ecosystems;
by addressing the entire system, the Elwha project of-
fers  a comprehensive  restoration  opportunity. In
addition, most rivers suffer continuing habitat pertur-
bations, whereas the one-time effort of removing the
Elwha and Glines Canyon dams and management of
the accumulated sediments will allow full restoration
to occur.
Literature Cited
Department of the Interior (DOI), Department of
  Commerce, and Lower Elwha Klallam Tribe. 1994.
  The Elwha Report, Restoration of the Elwha River
  Ecosystem and Native Anadromous Fisheries.
National Park Service (NPS). 1995. Final Environ-
  mental Impact Statement, Elwha River Ecosystem
  Restoration. Olympic National Park, Port Angeles,
National Park Service (NPS). 1996. Draft Environ-
  mental Impact Statement, Elwha River Ecosystem
  Restoration  Implementation.  Olympic  National
  Park, Port Angeles, WA.
Winter, B. D., and R. M. Hughes. 1995. American
  Fisheries  Society  draft position   statement  on
  biodiversity. Fisheries 20(4):20-26.
' National Park Service, Olympic National Park, 600 East Park Avenue, Port Angeles, Washington 98362.
2 Lower Elwha S' Klallam Tribe, 2851 Lower Elwha Road, Port Angeles, Washington 98363.

                                                             Estuarine Restoration:
                                         A Landscape Ecology Perspective
                                                                           David K. Shreffler1
                                                                             Ronald M. Thorn1
   Our central thesis is that landscape ecology princi-
 ples provide  the  critical link  between restoration
 ecology theory and effective, practical restoration in
 estuaries. We contend that  recent efforts  in Puget
 Sound to identify damaged natural resources through
 the Natural Resource Damage Assessment (NRDA)
 process could provide a unique opportunity to  use
 landscape ecology  principles  to  restore  lost  or
 degraded wetlands, as well as other habitats, in ecol-
 ogically  optimal locations within  the estuary and
 watershed.  In particular, we will demonstrate  the
 need for greater emphasis in site selection and design
 on habitat size, shape, structure, buffers, accessibility,
 connectance, and self-maintenance to help ensure that
 restored sites are functionally and structurally inte-
 grated into the landscape. Using our landscape-based
 restoration planning procedure, the highest priority
 sites:  (I) have  potential for restoration to historic
conditions, (2) are larger than 2 acres, (3) have ade-
quate buffers, (4) are easily accessed by target spe-
cies, (5) are connected to existing viable  natural
habitats, and (6) have a high probability of long-term
Literature Cited
Shreffler, D. K. and R. M. Thorn. 1993. Restoration
  of urban estuaries; new approaches for site location
  and design. Battelle Pacific Northwest Laborato-
  ries, Richland, WA.
Shreffler, D.  K. and R.  M. Thorn.  1995. Estuarine
  Restoration: A Landscape Perspective, pp. 702-712
  In Puget Sound Research. 1995 Proceedings. Puget
  Sound Water Quality Authority. Olympic, WA.
1 Battelle Marine Sciences Laboratory. 1529 West Sequim Bay Road, Sequim, Washington 98382.


                                            »              Ana|ysis for Enhancing
                                            Wildlife Habitat in the Yolo Basin
  Historically,  California's  Central  Valley was a
mosaic  of wetlands,  woodlands,  grasslands,  and
riparian corridors that provided diverse and important
habitat for wildlife, especially migratory birds. Dur-
ing the past 150 years, much of this valuable habitat
has been lost or degraded by activities that converted
the land to agricultural, residential, and other uses.
Fueled by an intent to reverse a long-term  decline in
populations of water birds that winter in or migrate
through the  Central Valley,  recent efforts  have
focused on restoring large portions of the formerly
vast marshlands and other habitats in  the southern
Sacramento Valley. This presentation will describe
these large-scale restoration  efforts and focus on  the
development of a study to analyze the suitability  for
enhancing wildlife habitat  in  the  170-square-mile
Yolo Basin near Sacramento,  California. The study
was requested in 1991 by a  broad coalition of local
landowners,  conservation  groups, state  and federal
agencies, and local elected officials. Its purpose is to
support the creation of a diverse complex of histori-
cally important habitats in the Yolo Basin. Inherent in
the study's purpose is the necessity for preserving
and balancing  agriculture and other land uses in  the
basin with the  needs of wildlife  and identifying
implementation  and  incentive  mechanisms  (e.g.,
easement and cost-share programs) for private land-
owners. This presentation will describe the study's
context and illustrate the approach and methods used
                   Joseph J. Donaldson1
                         Marcus Rawlings
                            Steve Chainey
                              Jay Stallman

to develop the suitability analysis. The study investi-
gated historical  and existing conditions,  including
soils, landforms, hydrology, vegetation, and agricul-
tural activities and employed a geographic informa-
tion  system in the analysis of restoration suitability.
The  study may serve as a model for restoring and
managing ecosystems in the California Central Valley
and elsewhere.

Literature Cited
Central Valley Habitat Joint Venture.  1990. Central
  Valley  Habitat Joint Venture Implementation Plan.
  Sacramento, CA. February, 1990.
Jones &  Stokes Associates, Inc. 1990. Inventory of
  wetland and  riparian  habitats  of  Yolo County,
  California. Final Report. Prepared for Yolo County
  Community Development Agency, Woodland, CA.
Jones & Stokes Associates, Inc. 1994. Suitability
  analysis for enhancing wildlife habitat in the Yolo
  Basin, Sacramento, CA. Prepared for Central Val-
  ley Habitat Joint Venture, Sacramento, CA.
Yolo Basin Working Group. 1990. Yolo Basin wild-
  life area concept plan.  Prepared  for  California
  Department of Fish and Game, Region II, Rancho
  Cordova, CA.
1 Jones & Stokes Associates, 2600 V Street, Suite 100, Sacramento, California 95818-1914.

                                                                 Historical Analysis:
                                         A First Step in Stream  Restoration
                                                                         Katharine R. Grant1
                                                                               Eddie Huckins
  In  1990,  the Nature  Conservancy created  the
 1.200-acre Middle Fork John Day Preserve (MFJD)
 in eastern Oregon  "to restore the cold water aquatic
 and riverine ecosystem to its natural conditions and
 functions." Federal and  state  restoration activities
 have highlighted the John Day River Basin because it
 lacks dams and hatcheries, making it one of the most
 important remaining wild salmon resources in the
 Columbia Basin. Historically,  the MFJD provided
 some of the best spawning beds in the watershed, and
 today still supports small populations of spawning
 wild chinook, despite extensive channelization  and
 habitat degradation. This project addressed the ques-
 tion, how has this  4-mile stretch of the MFJD river,
 and the river as a whole, changed since the turn of the
 century, and  why?  Analysis  addressed four data
 sources: aerial photos, flow records, climate records,
 and oral history. While the complexity of watersheds
 and the inherent patchiness  of historic data makes it
 difficult to draw firm conclusions about the past, this
 historic analysis provides information important in
 designing the long-term  restoration of the riverine
Literature Cited
Larkin, Geri.  1990. Plant Community Survey at the
  Middle Fork of the John Day River Preserve. Pre-
  pared  for  The Nature  Conservancy. Available
  through Oregon Field Office, Portland, OR.
Oregon Water Resources Department. 1991.  Stream
  Restoration  Program for the Middle Fork Subbasin
  of the John Day River. Prepared in  cooperation
  with U. S. Bureau of Reclamation. Available from
  ORWRD, Canyon City, OR.
Ottersberg,  R. J. 1992.  Detailed Soil Survey of the
  Valley Floor of Dunstan Ranch, Middle Fork John
  Day Preserve.  Prepared  for The Nature Conser-
  vancy.  Available  through  Oregon Field  Office,
  Portland, OR.
Welcher, Karin  E., 1993. Channel Restoration Plan
  and Geomorphology of the Middle Fork John Day
  Preserve.  Prepared  for The Nature Conservancy.
  Available through Oregon Field Offices, Portland,
1 The Nature Conservancy. 821 SE 14th Avenue, Portland, Oregon 97214.

                                         Multi-Scale Planning for Riparian
                                                     Woodland Restoration on
                                                        Streams of the Western
                                                                        United States
                                                                        Richard R. Harris1
                                                                               Craig Olsen
                                                                        Rebecca Johnson
 Prioritization and  planning of riparian woodland
restoration requires  data collection and analysis at
three levels: the watershed as  a whole, the stream
reach, and the site. Landscape measurements  ob-
tained from topographic and land cover maps can be
used to classify stream reaches within a watershed
that have the greatest potential  for restoration. Field
sampling, within the priority reaches focuses on as-
sessment  of geomorphic and vegetation  conditions
relevant to restoration. Standardized cluster analysis
is  applied to  the field sample data to distinguish
groups of samples with similar  structure and compo-
sition occurring  on floodplain  landforms. These
groups are then further evaluated in relation to de-
sired riparian ecological functions to establish targets
for restoration management.
Literature Cited
Harris, R. R. and C. Olson. In press. Two-stage in-
  ventory and analysis for prioritizing riparian resto-
  ration at the stream reach  and community scales.
  Restoration Ecology.
Harris, R. R.,  P. Hopkinson, S. McCaffrey and
  L. Huntsinger. In press. Use of geographical infor-
  mation systems  versus manual techniques for map
  analysis in riparian restoration projects: a compari-
  son. Journal of Soil and Water Conservation.
Olson, C. and R. R. Harris. In press. Applying two-
  stage inventory  and analysis to prioritize riparian
  restoration  at the San  Luis Rey River, San  Diego
  County, California. Restoration Ecology.

'Department of Environmental Science, Policy, and Management, Univerity of CaHfonua, Berkeley, Odifbnri. 94720.

                                                    Plant Species Diversity in a
                                               Semi-Arid Riparian Ecosystem
                                                         in  Southeastern Arizona
                                                                           Keirith A. Snyder1
                                                                           D. Phillip Guertin1
                                                                            Peter F. Ffolliott1
                                                                             Roy L. Jemison2
  Knowledge of community structure and associated
 hydrologic characteristics is fundamental to the suc-
 cessful restoration and conservation of riparian eco-
 systems. High plant species diversity is often cited as
 one of the foremost attributes of riparian areas. Base-
 line data were collected from a small mid-elevation
 watershed in southeastern Arizona to evaluate differ-
 ences in community structure and diversity between
 riparian areas and adjacent uplands, and to determine
 relevant environmental factors that  influence plant
 community  characteristics. Herbaceous  and woody
 plant distribution and abundance was sampled in 192
 quadrats located on transects that radiated away from
 a mixed perennial and intermittent stream. Principal
 Components Analysis (PCA) was  used  to display
 community  patterns of  species composition across
 environmental gradients. PCA indicated that patterns
 of woody plant composition varied with height above
 the  primary stream channel and  hydrologic regime,
 but  herbaceous plant composition appeared unaf-
 fected by proximity to a riparian area. Woody plants
 comprised three distinct communities associated with
 differences in height above the primary stream chan-
 nel: an upland, shrub-dominated Mimosa biuncifera
 community: a mesquite (Prosopis velutind) woodland
 community found primarily on secondary floodplain
 terraces; and a low-lying riparian community domi-
 nated by Platanus wrightii. A fixed effects analysis
 of variance  model  was  used  to  determine whether
 calculated diversity indices varied between riparian
 areas and upland  areas. Overall species  richness,
 based on the combination  of herbaceous and woody
 species sampled, was not consistently higher in low-
 lying riparian areas compared to upland areas. Alpha
 diversity indices  (species richness  (S),  Shannon's
Index (H),  and Simpson's Index (D)) calculated for
woody plants  demonstrated  no consistent pattern;
however, vertical  structural diversity, measured  by
the relative density of woody species within various
height classes, was highest near the stream. Similarly,
herbaceous  plant   diversity  was  not consistently
higher in  the riparian  areas.  Herbaceous  plants
appeared to be more sensitive to overall topographi-
cal differences than  to  fluvial system linkages.
Diversity was not higher in riparian areas  than in the
surrounding uplands for this semi-arid system. How-
ever, PCA demonstrated that the composition of these
low-lying areas was distinct from that of the uplands,
and vertical structural diversity  was highest near the
stream. Therefore, riparian areas make a unique and
important contribution to overall landscape diversity.

Literature Cited
Gregory, S. V., Swanson, F.  J., Mckee, W. A., and
  Cummins, K. W. 1991. An ecosystem perspective
  of riparian zones. Biosciences 41:540-551.
Jemison, R. L. and Ffolliott, P. F. (in press). Condi-
  tions that define a riparian zone  in southeastern
  Arizona. Water Resources Bulletin.
Snyder, K.  A. 1995. Patterns of plant species diver-
  sity and composition in a  semi-arid riparian eco-
  system.  Unpublished M.S.  thesis, University  of
  Arizona, Tucson, AZ.
Snyder, K. A., Guertin,  D. P., Jemison, R. L., and
  Ffolliott,  P. F. (in review). Plant species  composi-
  tion in a semi-arid riparian ecosystem. Vegetatio.
1 School of Renewable Natural Resources, University of Arizona, Tucson, Arizona 85721.
: US DA Forest Sen 'ice, Rocky Mountain Forest and Range Expt. Sta., Flagstaff, Arizona 86001.

                             Manipulating the Disturbance Regime in
                                        a  Southeast Missouri Bottomland
                                             Hardwood Forest Fragment to
                                                 Promote Overstory Species
                                            Recruitment and Regeneration
                                                                         Sylvia H. Taylor
                                                                   James W. Hermann1
                                                                             Brad Larson
                                                                       Michael Williams
                                                                      Alan R. P. Journet
   Big Oak Tree  State  Park  occupies  less  than
  450 hectares in Mississippi County, Southeast Mis-
  souri. Following extensive ditching, draining, and ag-
  ricultural development, Southeast Missouri, once a
  vast  expanse  of bottomland hardwood forest  con-
  taining a few scattered prairies, now contains virtu-
  ally  no  representatives of these original  ecological
  systems.  Close to  8 percent of the park  is an  old
  growth forest fragment with a canopy dominated by
  oaks, hickories,  pecan, and sweetgum. An adjacent
  area, subject to more extensive flooding,  supports a
  canopy dominated  by cypress. In addition, the  park
  contains  over a dozen state or national  champion
  specimens of bottomland woody species.  Studies of
  the structure and composition of these forest frag-
  ments consistently  reveal  a pattern through overstory
  and understory that indicates transition  to a box-
  elder/sugarberry  forest. It is hypothesized  that main-
  tenance   of  the   current  canopy  composition,
  particularly the cypress and oak components,  may
  require a management regime incorporating modified
  managed hydrology and disturbance.
   As a result, a study is currently being conducted, in
  collaboration with  Missouri's Department  of Natural
  Resources, to explore the role of understory removal,
  and understory removal in combination with fire, in
  promoting recruitment of oak seedlings, and to de-
  termine  the growth  rates of nursery grown trans-
  planted seedlings derived from local seed sources.
   Initial results indicate that understory removal, in
 conjunction with fire, selectively removed "late suc-
 cessional" herbaceous and  woody  species, signifi-
 cantly increasing the amount of  photosynthetically
active radiation available to the forest floor. The tie
required to measure regenerative rate in response to
disturbance  mandates  further  monitoring.  Initial
winter  survival of two species of nursery  grown
transplanted seedlings in the park indicate no signifi-
cant seedling predation after six months.
 This presentation described the forest structure and
composition, predictions regarding future unmanaged
transitions, and the consequences of the initial phase
of imposed fire disturbance.

Literature Cited
Abrams, M. D. 1992. Fire and the development of
  oak forests. Bioscience 42:346-353.
Hermann, J. W. 1995. Fire and restoration in the old-
  growth  bottomland hardwood forest fragment of
  Big Oak Tree State Park. Master's Thesis. South-
  east Missouri State University.
Lorimer, C. G. 1994. Tall understory vegetation as a
  factor in the poor development of oak seedlings be-
  neath   mature  stands.  Journal  of Ecology
Taylor, S. 1995. Spatial and temporal patterns in the
  bottomland hardwood forest of Big Oak Tree State
  Park. Master's Thesis. Southeast Missouri State
Department of Biology, Southeast Missouri State University, Cape Girardeau, Missouri 63701.

                                Invasive Plant  Species  Distribution and
                                      Abundance in Freshwater Wetlands
                                                     of Puget Sound Lowlands,
                                                      King County, Washington
                                                                       Catherine A. Houck1
  Invasive plant species often hinder restoration ef-
 forts in freshwater wetlands. Selecting wetlands with
 a low potential for invasion and creating environ-
 mental conditions and community structures that dis-
 courage invasive species are critical  components of
 successful wetland restoration projects.  This study
 characterizes wetland environmental  conditions and
 community structures that are dominated by invasive
 plant species. The distribution and abundance of in-
 vasive plant species were  studied in 19 freshwater
 wetlands  within  the  lowlands  of  King County,
 Washington. Patterns of invasive plant species distri-
 bution,  dominance,  and  abundance  were compared
 among  and  within wetlands and community types.
 Relationships between the degree  of invasion and
 watershed condition and seasonal water levels were
 explored for invasive species as a group and indi-
  Invasive plant species occur within all wetlands and
 community types in King  County,  Washington, but
 vary considerably  in the  number,  frequency,  and
 abundance among and within these habitats. Some of
 this variability was related to watershed condition
 and seasonal water levels. The frequency of invasives
 (as a group) was positively associated with the pro-
 portion of impervious area within a watershed; inva-
 sives were more common  in urbanized watersheds.
 This relationship does not  hold true for all invasive
 species,  however.  Some invasives   (e.g.,  Phalaris
 arundinaceae, Rubus discolor, and Solanum dulca-
 mara) were most abundant  within wetlands in urban-
 ized watersheds,  while  others (e.g.,  Typha latifolia
 and  Juncus effusus) were generally in greater abun-
 dance within wetlands in less urbanized watersheds.
Wetland hydrology, specifically the length of flood-
ing, also was associated with the abundance of some
invasives.  Typha latifolia and Juncus effusus were
more   abundant   under   permanently  flooded
conditions, while Rubus discolor  tended to inhabit
sites with  a limited to nonexistent flooding period.
Phalaris arundinacea was a generalist, thriving in a
range of hydrologic conditions.
 These results suggest that invasion is influenced by
life history attributes of the invader as well as spe-
cific environmental conditions  of a particular site.
Predicting or preventing invasion in wetland habitats
should consider attributes of each  potential invasive
plant species as well as site characteristics.

Literature Cited
Ehrenfeld, J. G. and Schneider, J.  P. 1991.  Chamae-
  cyparisthyoides  wetlands  and  suburbanization:
  effects on hydrology, water quality, and plant com-
  munity composition. Journal  of Applied Ecology

Houck,  Catherine A.  1996.  The distribution and
  abundance of invasive plant species in freshwater
  wetlands of  the  Puget Sound lowlands, King
  County,  Washington. M. S. Thesis, University  of
  Washington, Seattle, WA.

van der Valk  A. G. 1981. Succession in wetlands: a
  Gleasonian approach. Ecology 62:688-696.

van der Valk, A. G., L. Squires, and C.  H. Welling.
  1994. Assessing the impacts of an increase in water
  level on wetland vegetation. Ecological Applica-
  tions 4:525-534.
'Center for Urban Horticulture, College of Forest Resources, University of Washington. Contact address: 4724 University
View Place NE, Seattle, Washington 98105.

                                Considerations for Restoring Structure
                               Function, and  Diversity to an Ecosystem'
                                              Colonized by Invasive Plants: A
                                                           Phragmites Case Study
                                                                  Laura Ahearn-Meyerson1
                                                                            Kristina A. Vogt1
  The objective of our research is to identify ecosys-
 tem processes altered by the colonization of Phrag-
 mites australis  in a disturbed fresh water marsh.
 Phragmites is an invasive rhizomatous reed grass
 that frequently forms a monoculture and inhibits the
 establishment of other plant species. Degraded eco-
 systems  often  result  from multiple  disturbance
 factors occurring at different temporal scales. When
 an invasive plant alters  ecosystem-level processes
 such  as  nutrient cycling, then the  structure  and
 function  of that system  is impacted by that plant
 (Vitousek and Walker 1989).  These factors  may
 create legacy effects that hinder rehabilitation (Vogt
 etal. 1995, see table).
  Cutting and burning can encourage vigorous re-
 growth of phragmites  due to increased light avail-
 ability and thus  are not always effective restoration
 methods  (Marks et al. 1994). Herbicide treatments
 can require repeated applications and affect non-tar-
 get plant species as well (Ahearn et al. 1996).  We
 hypothesize  that   the  clonal   stage  of  stand
 development needs consideration when attempting
 to restore a site dominated by P. australis.  For
 example,  a  recently  established stand  can  be
 eradicated by treatment with herbicide followed by
 cutting. In contrast, increased stand age and detrital
 accumulation may modify nutrient cycles reducing
 the availability  of limiting nutrients to other plant
 species in the area. Phragmites invasion may leave
 legacy effects  that  make  simple eradication  an
 inadequate restoration  technique. These  legacy
 effects necessitate  restoration  of  pre-colonization
 ecosystem  cycles   by   accelerating  rates   of
 decomposition and regenerating desired vegetation.
 We  are  investigating natural and anthropogenic
factors that control ecosystem function and increase
the competitive  advantage of phragmites, including
land use history,  light  and temperature  regimes.
changes in nutrient cycles, hydrology, alteration of
the substrate, and legacy effects of invasive species.
Our hypotheses are 1) it is insufficient to simply
eradicate P.  australis. Legacy effects created by
colonization will hinder the reestablishment of de-
sired species and the return to the pre-disturbance
successional  pathway, 2) Phragmites australis is
modifying the site and causing it to become nutrient
poor. While P. australis may tolerate lower nutrient
availability, nutrients held in the phragmites detritus
are unavailable to other plants due to slow rates of
decomposition, 3) the effects of colonization by P.
australis on structure and function change over time
resulting  in increased degradation as  the site is
occupied by phragmites.
 We have established experimental plots and are
applying  treatments  to  identify  the  competitive
mechanisms of P. australis, the effects of stand age,
and the impacts of P. australis on nutrient pools and
cycles. By using information collected at  the eco-
system level  to modify  plant competition  and
restore ecosystem structure and function, we plan to
develop  methods  to   combat   the  general
phenomenon of invasive clonal species.

Factors Contributing to  Plant
Invasions and Potential  Direct and
Indirect Ecosystem Effects
  This  table  summarizes some of  the  factors that
make ecosystems vulnerable to plant invasions. The
direct  and  indirect effects outlined  in  the  table
below  can  all  contribute to habitat  degradation.
However, some factors  will have stronger impacts
on habitat  quality than others. In  a phragmites
dominated ecosystem, the importance of each effect
1 Yale School of Forestry and Environmental Studies, 205 Prospect T., New Haven, Connecticut 06511.

 Selected Abstracts
           Factors in plant invasions
           Anthropogenic disturbances:
           Changes in intensity/frequency
           of natural disturbances:
           •   Suppression of fire/flooding
           •   Dams

           Changes in faunal populations:

           •   Introduction of exotic fauna
           •   Reintroduction of extirpated
           •   Predator loss
Direct ecosystem effects
Colonization by invasive plants

Changes in community structure
and faunal populations

Decreased plant diversity

Alteration of light and tempera-
ture regimes

Detrital accumulation

Modification of nutrient cycles

Retention of heavy metals

Modification of hydrology

Alteration of soil structure

Decreased regeneration of native
plants/reduced food resources
Indirect ecosystem effects
Modified nutrient availability

Carbon fixation, nutrient avail-
ability and nutrient pulses

Changes in the composition and
diversity of micro/macro flora,
fauna, decomposers and plants

Disrupt symbiotic relationships

Loss of pollination/seed
dispersal vectors

Soil water relations

Catastrophic fire/flooding
 should  vary depending on phragmites  stand  age.
 Identifying how  the impacts  of  these variables
 change with  time is central to  our research and
 needs  to  be considered in  any attempt to restore
 sites dominated by P. australis.

 Literature Cited
 Ahearn, L., Triol, J., Berlyn, G.  1996.  The persis-
   tence of the herbicide glyphosate in soil. Unpub-
   lished paper presented at the annual conference
   for the Society of Ecological Restoration, Rutgers
   University, New Brunswick, NJ.
 Marks.  M., B. Lapin, and J. Randall. 1994. Phrag-
   mites  australis (P.  communis):  Threats, Manage-
   ment, and Monitoring.   Natural  Areas Journal
                 Vitousek, P. M. and L. R. Walker. 1989. Biological
                   Invasion by Myrica faya in Hawaii: Plant Demog-
                   raphy, Nitrogen Fixation, Ecosystem Effects. Eco-
                   logical Monographs 59(3):247-265.
                 Vogt, K. A., D. J. Vogt, P. Boon, A. Covich, F. N.
                   Scatena, H. Asbjornsen, J. L. O'Hara,  J. Perez,
                   T. G. Siccama, J.  Bloomfield, J.  F.  Ranciato.
                   1996. Litter dynamics along stream, riparian and
                   upslope  areas following Hurricane Hugo,  Lu-
                   quillo  Experimental Forest,  Puerto Rico.  Bio-
                   tropica (In press).

                                                                   Biological  Control of
                                                                      Purple  Loosestrife
                                                                                 Bernd Blossey1
 Purple loosestrife (Lythrum salicaria) is a wetland
perennial introduced to North America from Europe
in  the  early  19th century now occurring in large,
monotypic stands throughout much of the temperate
regions.  Because  conventional control techniques
have proven  ineffective, the development of a  bio-
logical weed control program began with surveys and
host specificity screening tests in  Europe in 1986.
Five-insect species demonstrated the host specificity
required  for  introduction to  North America. Four
species are now successfully established. These are a
root-mining  weevil,   Hylobius  transversovittatus,
attacking   the  main   storage  tissue  of   purple
loosestrife;  two  leaf-beetles,  Galerucella  calma-
riensis  and  G.   pusilla  capable  of  completely
defoliating individual plants and entire L.  salicaria
populations;  and  the flower feeding weevil  Nano-
phyes marmoratus. Since 1992, about 400,000 leaf-
beetles and 80,000 Hylobius eggs were shipped to 28
different states and Canada to collaborators in  a wide
range of organizations - many have started their own
mass rearing programs. Increased  attention  is being
given to follow-up studies to monitor target plant and
control agent populations. A standardized monitoring
guideline has been developed. This should allow the
comparison of results across the entire distribution of
L. salicaria.

Literature Cited
Blossey,  B.  1993. Herbivory below ground  and bio-
  logical weed control: life history of a root-boring
  weevil on  purple loosestrife. Oecologia 94:380-
Blossey, B., Schroeder, D, Might, S. D. &. Malecki,
  R. A.  1994a. Host specificity and environmental
  impact of the weevil Hylobius transversovittatus, a
  biological  control  agent  of  purple  loosestrife
  (Lythrum salicaria). Weed Science 42: 128-133.

Blossey B., Schroeder, D., Might, S. D. & Malecki,
  R. A. 1994b. Host  specificity  and environmental
  impact of two leaf beetles (Galerucella calmarien-
  sis and G. pusilla) for the biological control of pur-
  ple loosestrife (Lythrum salicaria). Weed Science

Blossey, B. 1995.  Coexistence of two competitors in
  the same fundamental  niche.  Distribution,  adult
  phenology and oviposition. Oikos 74:225-234.
Blossey, B. & Schroeder, D.  1995. Host specificity of
  three  potential  biological weed  control  agents
  attacking flowers  and seeds of Lythrum  salicaria
  (purple loosestrife). Biological Control 5:47-53.
Might,  S. D, Blossey, B., Laing, J.,  & DeClerck-
  Floate, R. 1995 Establishment  of insect biological
  control agents from Europe against Lythrum  sali-
  caria in  North  America. Environmental  Entomol-
  ogy 24:967-977.
Malecki, R. A., Blossey, B., Hight, S. D., Schroeder,
  D., Kok, L. T.  & Coulson, J. R. 1993. Biological
  control  of  purple  loosestrife.  Bioscience 43:480-
1 Department of Natural Resources, Fernow Hall, Cornell University, Ithaca, New York 14853.

                                         Rodeo®  Herbicide Use to  Control
                              Smooth Cordgrass (Spartina alterniflora
                                         L.)  in  Pacific  Northwest Estuaries
                                                                           Ron P. Crockett1
  Smooth cordgrass (cordgrass; Spartina alterniflora
 L.) is an introduced perennial weed in the estuarine
 ecosystems of Washington State. Cordgrass has in-
 vaded bare  mudflats and competes with desirable
 intertidal  species.  The  presence  of  cordgrass in
 Willapa Bay and other coastal bays impacts recrea-
 tion,  commercial oyster  culture, juvenile  salmonid,
 and migratory waterfowl habitats.  It also  alters  the
 hydrology and affects the accretion  of fine sediment,
 raising the elevation of the mudflats. Spartina control
 efforts include mechanical treatments and the use of
 Rodeo8 herbicide. Mowing and covering plants with
 plastic  represent the  mechanical options evaluated.
 Rodeo"  herbicide  applications  using  a  variety of
 application techniques have been shown to be effec-
 tive  in  controlling both the cordgrass leaves  and
 dense  fibrous  rhizomes.  Application   rates  of
 3.75 quarts/acre broadcast as 2 to 5 percent v/v spray-
 to-wet,  and 33 percent v/v  solution through wiper
 equipment have all been shown to be effective in
 controlling cordgrass. Complete spray  coverage and
 maximizing rain-free  or tidal inundation  intervals
 following treatment are important factors in obtaining
 consistent  control.  The  environmental  and  toxi-
 cological  characteristics  of glyphosate  the  active
 ingredient in Rodeo*,  are well  suited  to its  use in
 these sensitive sites.
 Rodeo® herbicide is a registered trademark of the
Monsanto Company, St. Louis, MO 63167.

Literature Cited
Kilbride, K. ML, F. L. Paveglio, and C. E. Grue. 1995.
 Control of smooth  cordgrass  with Rodeo®  in a
 southwestern Washington estuary. Wildlife Society
 Bulletin 23:520-524.
Paveglio, F. L., K. M. Kilbridge, C. E. Grue, C. A.
 Simenstad, and  K. L. Fresh. 1996. Use of Rodeo®
 and X-77® Spreader to control  smooth  cordgrass
 (Spartina  alterniflora) in  a  southwestern Wash-
 ington  estuary: I.  Environmental Fate.  Environ-
 mental Toxicology and Chemistry 15:961-968.
Simenstad, C. A., J. R. Cordell, L. Tear, L. A. Weit-
 kamp, F. L. Paveglio, K. M. Kilbride, K. L. Fresh,
 and C. E. Grue. 1996. Use of Rodeo® and X-77®
 Spreader  to control smooth cordgrass  (Spartina
 alterniflora) in a southwestern Washington estuary:
 II. Effects on benthic microflora and invertebrates.
 Environmental Toxicology and Chemistry 15:969-
1 Monsanto Agricultural Co., 17004 N.E. 37th Circle, Vancouver, Washington 98682-8616.


                                Integrated Pest Management Strategies
                                for Reed Canarygrass (Phalaris arundi-
                                             nacea L.) in Seasonal Wetlands
                                                                         Fred L. Paveglio1
                                                                        Kevin M. Kilbride1
                                                                          Ron P. Crockett2
                                                                       R. Bruce Wiseman3
  Reed canarygrass (canarygrass; Phalaris arundina-
cea L.) is responsible for the reduction of biodiversity
in wetlands nationwide.  Dense  monocultures  of
canarygrass displace open mudflats and out-complete
native wetland  plant species, reducing habitat value
for aquatic  migratory  birds  and  other  wetland-
dependent  wildlife. Although methods for control of
canarygrass have been studied in the mid-West, little
information is available regarding the effectiveness of
intergraded pest management  techniques  in  the
Pacific Northwest. The objective of this study was to
determine  the  effectiveness of Rodeo® herbicide
(2.25  quarts/acre), disking, mowing, and herbicide-
disking for the control of canarygrass at Ridgefield
National Wildlife Refuge in southwestern Washing-
ton State.  Preliminary findings from the  1994 treat-
ments indicate that  spring  herbicide  application
combined  with fall disking resulted in  the  greatest
control of  canarygrass. Other treatments,  in order of
declining effectiveness, were spring and fall appli-
cation of Rodeo®; fall disking; fall Rodeo® applica-
tion, and mowing. Vegetation response during the
1995 growing season will be discussed. Vegetation
response to treatments will also be determined during
the 1996 growing season. Evaluation of Rodeo®
application rates, timing, and method of application,
as well as surfactant type, for replicated small plot
studies at the Ridgefield National Wildlife Refuge
will also be presented.
 Rodeo® herbicide is  a registered trademark of the
Monsanto Company, St. Louis, MO 63167.

Literature Cited
Paveglio, F. L., K. M. Kilbride, R. B. Wiseman, and
  R. P. Crockett. 1996. Integrated management tech-
  niques show  promise for control of reed canary-
  grass (Phalaris arundinacea) in seasonal wetlands
  (Washington). Restoration and Management Notes
1 U.S. Fish and Wildlife Service, 9317 NE Highway 99, Suite D, Vancouver Washmgton 98665.
2 Monsanto Agricultural Company, 17004 NE 37th Circle, Vancouver, Washingtor198WU.            Ridgef.eld,
3 U.S. Fish and Wildlife Service, Ridgefield National Wildlife Refuge, P.O. Box 457,301 N. Tlnrd Street, K,ag
Washington 98642.

                                                Bioengineering in Stream and
                                                River  Restoration: Success of
                                                Enhancing Specific Functions
                                                     Under Multiple Constraints
                                                                                 Marit Larson1
                                                                              Betsy Hopkins2
  Bioengineering in stream and river restoration is the
 use of living plants in combination with organic and
 inorganic material for re-constructing, stabilizing, and
 establishing new morphological and  vegetative fea-
 tures.  The  proliferation  of modem  bioengineering
 techniques began about two decades ago in Western
 Europe, with the increasing concern over the loss of
 the ecological and hydrological functions of riparian
 systems. Initially, bioengineering was embraced as a
 technically  acceptable alternative to concrete and
 riprap. Bioengineering also augmented natural ripar-
 ian functions, and reestablished the appeal  of a more
 naturalistic landscape. In Europe, in particular, chan-
 nel revegetation techniques began to  offer a con-
 trolled  approach  to  enhancing elements  of urban
 stream systems, which have been entirely disrupted.
  Today, as our understanding of the full  scope of
 challenges in actually restoring and sustaining the
 viability of stream ecosystem functions increases, the
 role of bioengineering bears more careful scrutiny.
 Stream enhancement  projects  using  bioengineering
 should be examined for their contribution to restora-
 tion efforts. Bioengineering projects are often insuffi-
 ciently monitored and lack baseline  data, however,
 making their influence on habitat and the aquatic and
 wildlife  community  difficult  to  quantify. Conse-
 quently, the importance of bioengineering in meeting
 restoration goals often has to be evaluated using indi-
 cators that are available. The bioengineering projects
 we are  discussing at  this conference have been in
 place for various lengths of time in various locations.
 Their success and limitations are evaluated according
 to the goals identified for each project.

  The projects or case studies  explored here have
 three common themes. First, most of these streams
 are  in urbanized watersheds, have undergone extreme
modification, and have problems that would normally
be addressed  with conventional engineering struc-
tures. Second, these projects are implemented on a
limited section of the river system and are  not ac-
companied by comprehensive watershed restoration.
Third, these projects are constrained by space, time,
and funding. For these reasons, restoration goals are
confined  to specific features of the  stream  system.
The  projects discussed  include: Wedel Brook near
Hamburg and the Enz River near Pforzheim  in Ger-
many, and the North River in Massachusetts. Each
project had different goals and succeeded in  achiev-
ing these goals to varying degrees.
  Several  conclusions can be drawn from all three of
these projects, however. First,  the projects  demon-
strated that erosion control through vegetation and
vegetable materials can be successful.  Second, each
project underscores the need to clearly identify the
problems, and to  incorporate an  understanding of
channel geomorphology into the applications of bio-
engineering principles.  Third, the  projects success-
fully focused attention on rivers or sections of rivers
that  are  frequently marginalized  as  resources.  By
elevating the importance of waterways in even ur-
banized or constrained settings, we have a  greater
chance of modifying our actions  that impact these
resources. Since an increasing number of streams and
rivers are subject to the constraints  facing the streams
in these projects, we must continue to evaluate the
effectiveness of bioengineering approaches to stream
and river restoration.

Literature Cited
Allen, H, H.  and C. V.  Klimas. 1986. Reservoir
  Shoreline   Guidelines.  Waterways  Experiment
1 The Bioengineering Group, Inc., 53 Mason Street, Salem, Massachusetts 01970.
2 Bestmann Green Systems. Inc., Salem, Massachusetts 01970.

  Station, U.S. Army Corps of Engineers. Technical
  Report E-86-13, Vicksburg, MI.

Bestmann Green Systems,  Inc.  1993. North  River
  Streambank Stabilization  Bioengineering  Demon-
  stration Project, First Year Monitoring Report. U.
  S. Army Corps of Engineers. Waterways Experi-
  ment station, Vicksberg, MS.

Carey, M. 1996.  Effective low cost streambank bio-
  engineering techniques for the landowner. Interna-
  tional  Erosion  Control Association  Conference
  Proceeding, Seattle, WA. pp. 169-175.

Coppin, N. J. and I. G. Richards, (eds.) 1990. Use of
  vegetation in civil engineering. Butterworths, Lon-
Gore, J. A. and F. D. Shields. 1995. Can Large Rivers
  be Restored? Bioscience 45(3): 142-152.
Gray, D. H. and A. T. Leiser.  1989. Biotechnical
  Slope Protection and Erosion Control. Robert E.
  Krieger Publishing Co. Malabar, FL.
Groeneveld, D. P. and R. H. French. 1995. Hydrody-
  namic control of an emergent aquatic plant (Scirpus
  acutus) in open channels. Water Resources Bulletin
Humborg,  G.  1995. Typologische und morpholo-
  gische  Untersuchungen   an   Bergbaechen   im
  Buntsandstein-Odenwald.    (Classification   and
  morphological research of mountain  streams in the
  Sandstone-Oden Forest,) Institute for Hydraulic and
  Civil   Engineering,  University  of   Karlsruhe,
  Germany. Vol.  192.
Larson, M.  G.  1995. Developments  in River  and
  Stream Restoration in Germany.  Restoration Man-
  agement Notes. Journal  of the Society  for Ecologi-
  cal Restoration  13(l):77-84.
Miller, D. E. 1996. Design  guidelines  for bioengi-
  neering river bank  stabilization. International Ero-
  sion Control Association Conference Proceedings,
  Seattle, WA. pp. 157-168.
                                 Selected Abstracts
 Nadolny,  I.  1994.  Morphologic  und  Hydrologic
  naturnahe Flachlandbaeche unter gewaessertypolo-
  gischeen   Gesichtspunkten.   (Morphology   and
  hydrology in the classification of natural lowland
  streams,)  Institute  for  Hydraulic and  Civil Engi-
  neering,   University   of   Karlsruhe.   Germany
  Vol. 189.

 Peirce,  G. J.  1994.  Adaptive  Modes  in Wetland
  Plants, A preliminary Review. Southern Tier Con-
  sulting, Inc., P.O.  Box 30, West Clarksville   NY

 Rotar, M. A. and J. T. Windell. 1996. Innovative bio-
  engineering  techniques  used  to restore  Boulder
  Creek, CO. International Erosion Control Associa-
  tion Conference  Proceedings, Seattle, WA. pp. 55-

 Schiechlt, H. M. and R. Stem. 1994. Handbuch fuer
  naturmahen  Wasserbau. Oesterreichischer  Agrar-
  verlag. Vienne, Austria.

Schiechtl, H. M. 1980. Bioengineering for land recla-
  mation and conservation. The University of Alberta
Thomas, M. R. and R. E. Smiley.  1995. Stream chan-
  nel stabilization using soil  bioengineering - a case
  history. International Erosion Control Association.
  Conference Proceedings. Atlanta, GA. pp. 63-73.
Thomas, M. R. and R. E. Smiley.  1996. Stream chan-
  nel stabilization using soil  bioengineering. Interna-
  tional  Erosion   Control Association  Conference
  Proceedings, Seattle, WA. pp. 65-71.
U. S. Army Corps of Engineers (ACOE). 1987. Water
  Resources Development  in Massachusetts. New
  England Division.
 Wraith, J. G. 1996. Tannery Creek - Bioengineering
  stream realignment. International Erosion Control
  Association Conference Proceedings, Seattle, WA.

                                              Planting Deep Willow Cuttings
                                                        via Hydraulic  Jetting, for
                                                        Riparian Erosion Control
                                                      and Habitat Enhancement
                                                                                  Lon Drake1
                                                                                 Rick Langel1
  Planting  willow  cuttings for streambank erosion
 control is  an ancient technology, revived  in recent
 decades as part of "bioengineering." Problems in-
 clude difficulties with emplacement, while  slumping
 can easily  undermine the short  cuttings  normally
 used. We have  implemented a new hydraulic jetting
 technique,  which can rapidly (20 to 30 seconds) open
 planting holes  in  streambank alluvium, which  are
 6 feet  deep by  1 to 2 inches diameter.  Slender dor-
 mant willow cuttings pushed into  these  holes extend
 to damp soil and can sprout roots  their entire below-
 ground length. The technology consists of a specially
 machined jetting tip welded to a 6-foot  length  of
 ordinary '/2-inch steel pipe. Two of these jetting pipes
 are  connected  to  a portable high-pressure pump,
 drawing water from the stream and delivering it to the
 jets via sturdy garden hoses. A planting team of 4 is
 most  efficient, with  2 persons jetting holes  and
 2 planting willows. At favorable locations,  a team of
 4 can harvest 1,000 to 1,500 slender willow cuttings
 per  day and plant  1,000 cuttings per  day. All  the
 equipment  is portable and can be carried on muddy
 stream banks. The jetting  equipment plus 1,000 bun-
 dled willow cuttings fit into a pickup truck.
  The  hydraulic jetting works well in  most poorly
 sorted alluvial sands, silts, and fine gravels,  including
 those  saturated with  groundwater.  Even under
 quicksand conditions, the  added water pumps up the
 hydraulic head  so that the jetted holes  remain open
 long enough to  push a cutting down 6 feet. We have
also jetted through rotten logs and can flush stones 1
inch  in diameter up out of the hole. The technique
does not work well in dense clayey glacial till, stoney
gravel or coarse clean sands.
 On the sandy eroding cutbanks  of  Clear  Creek,
Iowa, in 1994, we planted approximately 4,000 sand-
bar willow cuttings. At the end of the first growing
season, we had  a 98 percent survival rate for those
planted below the terrace top. However,  the banks
were left in a vertical position and during the floods
of 1995 from this flashy 88-square-mile watershed,
the cutbanks continued to collapse and damaged por-
tions of this planting. In  spring 1995, we planted
2,000 willows along a broad curving stretch of Hare
Creek, Iowa, in silty colluvium. The  Hare Creek
watershed is very flashy but only 4 square miles in
area. We excavated the  vertical banks back to a 1:1
slope before planting and broadcast seeded them with
oats  and  timothy grass. At least  7 floods  passed
through within  2 months after planting but  willow
survival was  about 96 percent  for  the season  and
erosion  was hardly  measurable. Test plantings in
1996  include red osier dogwood  cuttings for en-
hanced habitat  value. Caveat: before they become
established, willow cuttings can be drowned by pro-
longed flooding. They do not root well  below the
permanent water table and  should not be planted in
the flowing water of streams at baseflow level.
1 Department of Geology, University of Iowa, Iowa City, Iowa 52242.


                                                     Revegetation at La Franchi
                                                          Demonstration Wetland,
                                                            Santa  Rosa,  California
                                                                            Marco Waaland1
  Abstract: Tule (Scirpus californicus) plants at La
Franchi demonstration wetland have been moni-
tored for growth parameters  (number of stems
per tule clump, area of marsh zone occupied by
tule clump) after transplantation into three sepa-
rate marsh zones in winter of 1990. Twenty-nine
of the 30 plants in the sample  survived. Initial
stem numbers and clump area were not measured.
However, it is unlikely stem number exceeded
seven and clump area exceeded 0.5  ft2. By fall of
1991, mean stem density was 121 stems/plant  and
the mean area occupied by individual tule clumps
was 7.6 ft2. As expected, a significant correlation
exists between the number of tule stems and the
area occupied per tule clump. By projecting the
relationship between increase in tule clump area
over time, it was possible to estimate that  the
vegetation  at the demonstration  wetland would
occupy the entire marsh zone approximately 1.5
years after transplantation. At that  time, the  full
animal  waste treatment ability  of the  wetland
would be attained because the surface area of the
tule stems and oxygen transfer to the soil via the
root system would be maximized.

  The  dairy industry is the largest user of the Santa
Rosa Subregional Water Reclamation System's re-
claimed water. However, its viability may be threat-
ened by  the costs associated  with  animal waste
control water quality compliance. Animal waste  con-
trol is an important element of the existing reclama-
tion system and was selected by the  City of Santa
Rosa for a long-term reclamation project.
  The potential loss of any dairies in the Laguna has a
negative effect on the ability of the industry to sur-
vive, and has potentially negative implications for the
Subregional System's irrigation reuse program. Con-
structed  treatment wetlands (i.e., freshwater marsh
systems) have been shown to successfully treat ani-
mal waste to acceptable levels (Maddox and King-
sley,  1989). Wetland wildlife  also  benefit  from
increased marsh  habitat acreage.  The Subregional
System  has  undertaken this  animal waste control
demonstration project to provide a local example of
management practices that can be used on dairies to
meet water quality objectives. This paper presents
preliminary  findings  on  the development of the
planted vegetation of the wetland.

Physical Layout
 The wetland consists of three vegetated zones, sepa-
rated by deeper open water sections. Operating water
depth for the marsh zones was designed to be 1.5 feet.
Topsoil (Huichica loam) was stockpiled during con-
struction and subsequently spread in a layer one-foot
thick in  the marsh zones to provide a planting sub-
strate. The dimensions of each planting zone  were
approximately 45 feet by 50 feet, for a total planting
area of 6,750 ft2.

Revegetation Scheme
 A total of 420 California tule (Scirpus californicus)
plugs were transplanted into the three marsh zones in
rows.  The tule source area  was a freshwater marsh
along Roseland Creek, bordering Sonoma County
Water Agency  mitigation area at  Alpha Farm. Tule
plugs were dug from the soil during the fall and win-
ter of 1990. Plugs consisted of rootwads and several
live stems which were kept intact  and transported by
truck to La Franchi's and transplanted the same day.
Individual rootwads were planted on staggered four
foot centers along each row. Tules were transplanted
into wet or flooded soils at the demonstration wetland
on the  following dates: Zone 1, 12/6/90; Zone 2,
11/18/90; and Zone 3,11/12/90.
1 Golden Bear Biostudies, 257 27th Avenue, San Francisco, California 94121.

 Selected Abstracts
 Data Analysis and Monitoring Methods
   A sampling scheme involving 30 randomly selected
 tule plant individuals was developed to reflect per-
 formance  of the La  Franchi tule  population.  Water
 depth,  tallest stem, total number  of live stems per
 clump  and  individual  plant (clump) circumference
 were measured.  Individual  tule plant circumference
 was measured at the water surface  and converted to a
 cross-sectional2 area. This parameter is referred to as
 the "area of marsh zone occupied by tule clumps." In
 the course of the study, one of the selected sampling
 plants  died and was removed from regression  analy-
 sis. The statistical package "Statgraphics" was util-
 ized to  perform multifactor ANOVA  and simple
 regressions. The regressions lumped together  all re-
 peated observations of stem number and clump area
 to determine the degree of correlation.
   The data were transformed by natural log for the re-
 gression analysis.  The ANOVA sorted the data by
 season to estimate stand growth trend, and by zone to
 determine sources of any variation that  may be ex-
 plained by spatial differences. Monitoring dates were:
 6/6, 8/7, and 11/11, 1991.

   The  growth response of  the 30 sample plants  is
 shown in Figure 1. The three discrete distributions
 represent  the  spring,  summer and fall  monitoring
 events. The number of stems per tule clump  is ar-
 rayed along the Y-axis and the area occupied per tule
 clump  is arrayed along the Z-axis. The X-axis repre-
 sents  the  number  of days  elapsed since the  plants
 were  initially transplanted. The  increase  in  stem
 number and clump area are discussed further below.

   Of the 30  plants  subjected to monitoring, only one
 died; a 97 percent survival  rate. This result demon-
 strates  the resiliency  of the California tule to uproot-
 ing and transplantation shock.

   There was a significant increase in number of stems
 per tule clump over time. The original number of live
 stems is not known, but it is unlikely that any  clump
 exceeded seven live stems,  based  on the clump size
 that could be handled by one person with a shovel.
 Figure  1 shows that mean number of stems increased
 from 14.7  in spring,  to 62.1 in summer and 121 by
fall. By fall, there were significantly fewer live stems
in Cell 3 than either Cells 1 or 2, despite the fact that
Cell 3 was planted first and had three more weeks of
growth than the other cells.
  The mean  area of marsh zone occupied by individ-
ual  tule clumps increased over time from  0.9 ft2 in
spring, to 4.9 ft2 in summer, and 7.6 ft2 by fall. Unlike
stem  growth, there was no significant  difference in
area occupied by tule clumps between the three sepa-
rate marsh zones.


Tule Growth Dynamics
  The 97 percent survival rate exhibited by the Cali-
fornia tule transplanted at the La Franchi Demonstra-
tion Wetland demonstrates a high degree of success.
Corollary studies underway at the Kelly Farm Dem-
onstration Wetland also report high tule survival rates
of 78 percent and 94 percent (CH2M HILL et al, in
prep). It is apparent that the California tule is a hardy
species  able to withstand the  injury and  shock of
  The transplanted tules spread across the revegetated
marsh zones at a rapid rate. The constructed planting
area set  aside for tule marsh totaled approximately
6,750 ft2. In fall  1991, the area occupied by all the
tule clumps was estimated  at 3,990  ft2  (7.6ft2
x 420 plants). Therefore,  by fall, over  50 percent of
the  area of all three marsh zones was occupied by tule

  Another means of assessing  the  development of
marsh at the La Franchi Demonstration Wetland is to
compare the overall  stem density within marshes at
different stages of development. Overall density of
tule stems within a defined area of marsh can vary
from 0.5 stems/ft2 in sparsely vegetated, recently cre-
ated marsh, such as  the Subregional System's Kelly
Farm Demonstration Wetland (CH2M HILL et al, in
prep.), or exceed  15  stems/ft2,  as  found in mature
natural marsh  growing in the Central  Valley where
tules occupy the entire marsh area (Fox 1989). At the
time of the most recent La Franchi wetland monitor-
ing (11/11/91), the average density of tule stems in
This  estimate  assumes  stem  growth  for each plant expands in a geometrically  round  form,  fitting the  simple
 relationship: area = Tcr".

                                OKIS SINCE PLANTING
Figure 1. Increases in Area of Marsh Zone Occupied by Tule Clumps and Number of Stems at La Franchi
Demonstration Wetland.

 Selected Abstracts
 the marsh zone was: 7.5 stems/ft2 [(121 stems/plant
 x 420 plants) -*• 6,750 ft2].  If tule stem density in
 mature marsh can  reach up to 15 stems/ft2, then the
 most recent  estimate of 7.5 stems/ft2 is 50 percent
 that of mature marsh.
   Both the  area of  marsh  zone occupied  by tule
 clumps (50 percent)  and percent stem density com-
 pared to mature marsh (50 percent) indicate a marsh
 in a youthful stage of development.  However, it is
 also a marsh in a phase of rapid vegetation growth.
 As would be expected, the number of stems per tule
 clump and the area of marsh zone occupied  by tule
 clumps  are  highly correlated.  This  log-linear
 relationship shows  a  rapid  rate  of tule  clump
 expansion as the  transplanted stems take root and
 begin sending up  new sprouts. As more stems are
 added the rate of expansion  increases dramatically.
   For the  existing tule  clumps to expand  to the
 extent that they fully occupied the entire area of the
 three  marsh zones, they would have to increase in
 mean area  occupied, from  the current 7.5 ft2 to a
 mean of 16 ft2. A regression of the data for clump
 area against time elapsed since planting indicates a
 significant correlation. By using this limited data set
 it is  possible  to  extrapolate  the  regression  to
 determine approximately when the mean clump area
 would reach  16 ft2.  The regression model predicts
 that the mean clump area would reach the maximum
 of 16 ft2 after approximately 560 days, or 1.5 years
 after   planting.   Within   95 percent  confidence
 intervals, the actual  time could range from 1.0 to 3
 years. At this time,  the full  animal waste treatment
 ability of the wetland will  be attained because the
 surface area of the tule stems and oxygen transfer to
 the soil via the root system will be maximized.
   It  must be stressed that this is only an estimate
 based  on  a  very limited  data set. This  model
 assumes that the growth response will continue in a
 linear fashion. It is  possible that the initial  growth
 rate is exponential but tapers off before the marsh
 zones  were entirely  occupied. This  decline  in
 growth rate would be a natural result of competitive
 interactions between the individual tule clumps for
 increasingly limited  resources such as space, light,
 and nutrients.

  Future studies should look at comparative rates of
 spread amongst differently spaced  planting centers.
Also, studies of individual tule clumps would allow
observation  of whether there  is  a self limiting
maximum to the numbers  of new shoots individual
plants can develop.

Related  Benefits
  Monitoring at a second created wetland showed a
rapid development of wildlife habitat. Seventy-eight
species of birds have been  observed. The total num-
ber of birds  increased by 38 percent within one year
after revegetation, along with an increase in nesting
species. A large aquatic prey base  (crayfish, frogs,
and mosquitofish) has resulted in  colonization by
river otters  and mink. The development of dense
marsh vegetation has enhanced the  nutrient removal
capability of the  marsh. Orthophosphate reduction
averages  33 percent,   despite large  numbers  of
waterfowl. Total inorganic nitrogen was essentially
removed (94-99 percent), and TKN was reduced by
80 percent.

  Carolyn Dixon,  natural resources manager for the
Subregional System, has been closely involved with
the La Franchi project. She assisted in the planting
and conducts the routine marsh monitoring. Kristina
Sloop also deserves credit for her  assistance in all
phases of this project.

Literature Cited
CH2M HILL,  D. W. Smith, R. Williamson, and M.
  E. Waaland. In Preparation. Kelly Farm Demon-
  stration Wetland: Management  Plan and Moni-
  toring  Results. Prepared  for  Santa Rosa Sub-
  regional Water Reclamation System.

Fox, J. P. 1989 Evapotranspiration of macrophytes
  in  California's  wetlands.   Unpublished  report
  prepared for San Francisco Bay/ Delta Hearings.

Maddox, J. J. and J. B. Kingsley. 1989. Waste treat-
  ment for confined swine with an integrated artifi-
  cial wetland and aquaculture system.  In: D.A.
  Hammer (Ed.)  Constructed Wetlands for  Waste-
  water Treatment.  Lewis Publishers, Inc. Chelsea,
  MI. 831 pp.

                               Lessons Learned in Habitat Restoration,
                                  Comparison of Two Wetland Mitigation
                                             Projects in San  Diego, California
                                                                             John L Minchin1
  Abstract:  This paper addresses design and in-
stallation  issues  associated  with  large  scale
riparian mitigation projects in San Diego County,

  Over the past 6 years Ogden's habitat restoration
team  has been involved in well  over 100 wetland
mitigation projects within San Diego County. Two of
the major projects accomplished using this period in-
clude, the Twin Oaks Valley Ranch project, a 308-
acre master  planned recreational community in San
Marcos, in northern San Diego County; and the All-
red-Collins Business Park project, a 140-acre planned
industrial development in San  Diego.  Both projects
involved complex  hydrological constraints, diverse
mitigation requirements, and multidisciplinary coor-
dination with consultant team members to bring about
ultimate  success.  Both projects are similar in that
Ogden  staff were  involved  from  start  of  finish
including initial biological evaluation,  design devel-
opment   and  construction   document  preparation,
installation  monitoring,  and  long-term  biological
monitoring. As a result, much has been learned in the
  The purpose  of this presentation is to compare the
two projects and outline some of the lessons learned
in  implementing  their  mitigation  programs.  This
should help  point out key  issues to consider when
planning similar habitat restoration projects, for the
benefit of other professionals involved in this type of
Twin Oaks Valley Ranch
 Twin Oaks Valley Ranch project included biologi-
cal analysis of existing vegetation; wetland delinea-
tions  and  impact  evaluation; conceptual  wetland
mitigation plan preparation; permit processing  with
regulatory agencies; coordination with civil engineers
for alternative flood channel design; assistance  to
landscape architects and golf course  architects for
construction  document  preparation;  construction
implementation monitoring; and long-term mitigation

The Major Project Challenges at Twin Oaks Included:

•  Interfacing with project  team consultants  to
   develop an integrated  mitigation scheme which
  . would to work around the  development  con-
   straints of the 644  home residential community
   and 18-hole golf course.

•  Interfacing with recreational amenities including
   edge  transitions from the  native habitats to the
   golf course, and incorporating equestrian and pe-
   destrian trail access.

•  Analysis of development impacts and developing
   compensatory mitigation measures acceptable to
   the regulatory agencies.

•  Evaluation of flood control and hydrologic con-
   cerns, developing flood channel treatment alter-
   natives, and developing strategies for check dam
•  Coordination of wetland mitigation requirements
    with  permitting and reviewing agencies, and con-
    vincing them to accept mitigation strategies with
    the incorporation of recreational amenities.

•   Developing  appropriate  vegetation  associations
    for the revegetation effort to  meet the mitigation
1 Ogden Environmental and Energy Services Company, 5510Morehouse Drive, San Diego, California 92121.

 Selected Abstracts
 Allred-Collins Business Park
  Allred-Coliins  Business Park  project included bio-
 logical surveys of existing vegetation; evaluation of
 canyon hydrology, both before and after construction,
 to  verify suitable conditions for plant material estab-
 lishment: conceptual wetland mitigation  plan  prepa-
 ration;  regulatory permit  processing; coordination
 with  project engineers  and contractors  on canyon
 grading and check dam construction; assistance to the
 landscape architects  on  conceptual plan preparation;
 construction implementation monitoring; and  long-
 term mitigation monitoring.

 The Major Project Challenges at Allred-Collins Included:

 •   Evaluation of flood control/hydrologic concerns
     to verify suitability for vegetation.

 •   Incorporating strategies  for check dam construc-
     tion  to  slow down  water and alter  the canyon
     bottom  gradient to accommodate an  expanded
     riparian corridor.

 •   Coordination of wetland mitigation requirements
     with permitting/reviewing agencies.

 •   Developing  appropriate  wetland vegetation asso-
     ciations for the revegetation effort.

 •   Incorporating recreational  park   facilities  and
     pedestrian trail systems.

 •   Working around the industrial park development
     constraints and evaluating drainage outfall.

 •   Addressing  long-term brush management  issues
     along perimeter of the open  space preserve.

 •   Coordination  of  grading  and  check   dam

 •   Intensive field spotting of plant  materials and
     hands-on coordination with  contractors  during

 •  Coordination with maintenance contractors  dur-
    ing long-term maintenance/monitoring period.

 Key Issues Summary and Conclusions:
  Based  on experience gained during these  two large
 mitigation projects, we believe the issues listed below
 should be given key consideration  in wetland habitat
 restoration design.

  A.  Analyze your project site thoroughly including:
      soils, hydrology (before and after construction),
    climatic considerations,  sun exposure, slopes,
    and gradients.
B.  Inventory all biological resources including :

      •   plant communities
      •   sensitive plants and wildlife
      •   endangered species.
C.  Evaluate project development plans and meet
    with other project team members to thoroughly
    understand development  strategy  and sched-
D.  Evaluate impacts including:

      •   Direct impacts
      •   Indirect impacts
      •   Additional construction access issues and
          staging areas
      •   Buffer zones
      •   Future  fire/brush  management zones and
          limits of vegetation clearing
      •   Possible preservation issues
      •   Designation of open space parcels
      •   Future utilities easements.
E.  Establish resource agency and local jurisdiction

      •   Appropriate  agency review for type of
          impacts/mitigation required
      •   Evaluate permitting/processing timeline
      •   Sequence  of review  with  regulatory

F.  Evaluate mitigation  requirements,  strategies,
    and  options.  Anticipate acceptable  mitigation
    ratios for agency concurrence and provide justi-
    fications for ratios being proposed.

G.  Evaluate site potentials for type of mitigation
    required. (Are adequate factors  available  to
    support intended mitigation?):

      •   Soils
      •   Hydrology and depth to water table
      •   Sufficient acreage available
      •   Adjacency of native habitat  and potential
          wildlife/habitat linkages.

H.  Explore off-site mitigation alternatives.

I.   Develop conceptual mitigation plan program.

J.   Develop detailed working drawing/construction
    documents for bidding; make  sure  specifica-
    tions are adequate (i.e., present the  installation

                                                                                       Selected Abstracts
    goals, delineate maintenance requirements, and
    establish monitoring responsibilities).
K.  Order plants well ahead of anticipated installa-
    tion schedule,  at least 9 months to 1 year in
    advance. Anticipate  special  collections,  local
    sources, and realistic implementation schedules.
    Build  in   contingencies   should  delays  be
L.  Coordinate  with general contractor and  land-
    scape contractor for installation:

      •   Explain environmental constraints
      •   Help  develop implementation  strategy/
      •   Set up site observation/monitoring sched-
          ule at key project milestones
      •   Monitory actual installations.
M. Coordinate   final   construction  acceptance/
    maintenance turnover.

N.  Coordinate  with maintenance contractor  for
    long-term maintenance/monitoring.
O.  Conduct actual monitoring:

      •   Horticultural
      •   Botanical.
P.  Coordinate final  turnover  at  end of mainte-
    nance/monitoring period:

      •   Client
      •   Local jurisdictions
      •   Resource agencies.
Q.  Follow up with evaluation of success and les-
    sons learned for future projects.

 Selected Abstracts
 Wetland  Functional Relationships:
 A Landscape Ecology Perspective

 Robert R.  Fuerstenberg
 Sr. Ecologist. King County Surface Water Management
 Division, 700 5th Avenue, Ste. 2200, Seattle, Washington

   The functional  attributes of wetlands  are  not
 solely the result of internal wetland characteristics.
 Many of the  functional values we assign to wetlands
 are the outcome of processes originating some dis-
 tance from the  wetland edge. Failure  to consider
 these outside influences has often resulted in inef-
 fective protection and restoration programs. In order
 to understand wetland functions more clearly, it is
 necessary to  examine the relationship of wetlands to
 the surrounding landscape. Landscape ecology fo-
 cuses on  three characteristics of landscapes: struc-
 ture-the  spatial  relationship  among  elements;
 function-interactions among  the spatial elements;
 and process-the alteration of structure and function
 over time. It therefore provides a useful framework
 for examining wetland functions as  they are linked
 to the greater landscape. Employing the principles
 of landscape ecology,  we may ask how wetland
 functions are coupled to landscapes and influenced
 by the structure, function and processes that occur
 outside and often distant from wetland boundaries.

 Assessing  Wetland Condition and
 Restoration  Potential Within
 Watersheds  Using a Performance
 Criteria Matrix

 Robert P.  Brooks1, Charles Andrew
 Cole, Denice H. Wardrop, Laurie Bishel,
 and Diann J.  Prosser
 'Penn State Cooperative Wetlands Center, Forest Re-
 sources Lab, Penn State University, University Park,
 Pennsylvania  16802.

  Assessing the condition and the restoration poten-
 tial for wetlands and their associated stream reaches
 within watersheds  is an appealing concept, both
 ecologically  and politically.  Yet,  issues of scale
 continue to plague attempts at measuring how indi-
 vidual wetlands functionally  contribute to water-
 shed health,  how the wetlands  themselves  are
 impacted  cumulatively, and  where to effectively
 target locations for restoration and creation projects.
 We are recommending use  of a tailored HGM and
 disturbance classification  system coupled  with  a
 performance  matrix to: 1)  describe the variability
across wetland types, 2) improve site selection pro-
cedures  for created  and restored wetlands,  and
3) predict  the  successional  trajectories  of natural
reference  wetlands, and  creation and  restoration
projects. We recommend that the sources of infor-
mation used  to  develop  regional  classification
schemes and performance matrix  be obtained from
detailed studies of reference wetlands within a re-
gion. Intensive studies of the site characteristics of
reference wetlands and created projects, conducted
by the Penn State Cooperative  Wetlands Center
(CWC), have greatly increased our understanding of
wetlands in their respective landscape positions. We
have shown that natural wetlands classified into five
categories (receptor,  conveyor, donor, fringe,  and
beaver-impounded) differ in their respective  soil
characteristics, plant  communities, hydrologic  and
water quality signatures,  sedimentation rates,  and
habitat potential. In addition, comparisons between
reference wetlands and creation projects show dis-
tinct differences in these attributes. These findings
can be used to define expectations regarding how
natural wetlands will  respond to impacts and to de-
termine whether creation projects are performing
their expected functions. We are extrapolating these
site characteristics to a  larger scale to establish
synoptic criteria to screen portions of  watersheds
regarding   their  condition  and  the  need  for
restoration. We propose that these same criteria can
be used to build a performance matrix, arranged by
HGM and disturbances classes; based on the best
information available. The strength of this approach
is that the reliability and predictability of assess-
ments of wetland condition are expected to increase
as we  move from best  professional judgment to
testable hypotheses.

Wetland Restoration Success
Following Pipeline  Construction

Margaret Clancy1, Clayton Antieau2 and
Dennis  Burns3

'EDAW, Inc., 1505 Western Avenue,  Seattle, WA 98101
2Washington State University Cooperative Extension, Jef-
ferson County, 201 W. Patison, Port Hadlock, Washing-
ton  98339.
3Foster Wheeler Environmental Corporation, 10900 ME
8th Street, Bellevue,  Washington 98004.

  Construction of natural gas pipelines has affected
hundreds of acres of wetlands throughout the Pa-
cific Northwest. Federal regulations require pipeline
companies to restore  wetland habitats to their origi-
nal contours  and vegetated condition following
construction.  However,  criteria  for   evaluating

restoration success  are  usually poorly defined or
focused on a single parameter such as plant cover-
age.  In a study that examined conditions in  120
wetlands one year following pipeline construction,
investigators assessed the degree of restoration suc-
cess  relative to three main criteria: total vegetative
cover,  species diversity, and  average species indi-
cator status (according to Reed 1994, 1988). Data
concerning the percent of native and introduced
species and the survival of planted  woody  stems
were also compiled. Minimum success  thresholds
were established for each criterion and a data in-
dexing system combined data for each criterion into
a single value for  each wetland. Pre- and post-
construction species lists were compared and the ef-
fects of reseeding  were examined. Cover  in  the
majority  of   the  wetlands   studied   exceeded
80 percent and most wetlands  supported a domi-
nance of native hydrophytes. Species diversity was
typically higher following construction. Reseeding
efforts appeared to  have been unsuccessful as  the
species used  in  restoration  seed mixtures  were
poorly represented in post-construction plant com-
munities. Establishing vegetation appeared to have
derived from soil seed banks and/or adjacent seed

Hydrologic and  Ecological
Modeling and Geographic
Information  Systems for
Predicting Floodplain Forest
Restoration Following Dam

John R. Shuman1, Karen R. Warr, Gary
M. Greenberg, and Judith C. Bryan
'St. Johns River Water Management District, P.O. Box
1429, Palatka, Florida 32178-1429.

  Significant  areas  of  floodplain forests are  fre-
quently lost when rivers are  impounded  by dams.
Rodman Dam, a  2,073m long  and 6.7m high
earthen-fill dam on the Ocklawaha River in Florida,
is proposed for removal. This dam inundated or se-
verely  stressed approximately  2,924 ha  of flood-
plain forest on this  canopied, blackwater river. An
important component of the environmental assess-
ment research designed to assist in the reservoir re-
tention or river restoration decision-making process
was  the prediction of floodplain forest restoration if
the dam were removed. Floodplain forest restora-
tion following dam removal was predicted using the
U.S.  Fish and Wildlife Service's FORFLO  forest
succession simulation model, designed to predict
 the impacts of altered hydrologic regimes on flood-
 plain forests. FORFLO is a stochastic,  individual-
 based  computer  model  simulating  the  growth.
 reproduction and death of each tree  in a test plot.
 The primary variables  in FORFLO affecting forest
 succession are soil type, growing degree days, and
 nver flow regime. Water stage statistics in the river
 and  floodplain  with the dam  removed  were pre-
 dicted using HEC-1 and HEC-2 simulation models.
 Bathymetry of the river and floodplain  were also
 measured, digitized, and entered into CIS. FORFLO
 was  used  to predict  the  community  dominants
 within elevational plots across five floodplain tran-
 sects along the 25.7 km longitudinal axis  of the res-
 ervoir. The FORFLO predictions were entered into
 GIS  as another data layer  over bathymetry, with
 maps of predicted floodplain habitats produced at
 intervals through 200 yr of forest succession. Mod-
 eling results demonstrated that the floodplain forest
 canopy would return in about 40 yr after dam re-
 moval, occupying about 3,007 ha of the floodplain,
 only slightly greater than the 2,924 ha of floodplain
forest occupying the study area prior to  dam con-
struction. The use of GIS maps was  an effective
method to  simplify and present  these floodplain
forest succession predictions to non-technical deci-

 Restoring Hydrologic,
Geomorphic, and Ecologic
 Functions in the Big Quilcene
 River Delta

 Larry B. Fishbain1,  Philip B. Williams2,
 and Brian Collins3
 ll2Philip Williams & Associates, Ltd., Pier 35,  The Em-
 barcadero, San Francisco, California 94133.
 332201 Fairview Ave. E., Slip 2, Seattle, Washington
  The Big Quilcene River  is a steep, gravel bed
 stream, which provides spawning and  rearing habi-
 tat for a Washington State listed species of early run
 chum salmon. This habitat has been  degraded by
 alterations in the hydrologic functions of the fluvial
 system that result from disturbances to its physical
 and  ecological structures, including logging in the
 upper watershed, water diversions,  urbanization,
 levee construction, and periodic removal of gravel
 from the channel bed.  Impacts to spawning habitat
 include burial of redds  by excess sediment, washout
 of reeds at high flows,  dewatering reeds  by channe
 migration, and removal of gravel from the channel
 in  an  effort to  reduce flood risks. This paper

 Selected Abstracts
 describes  a project  designed  to enhance  chum
 habitat on the Big Quilcene River delta by restoring
 the structures of the fluvial system that create and
 sustain  anadromous  fish  habitat.  The  project
 includes  removal of levees in the lower 0.5 river
 mile  to  reconnect  the  river  channel with  the
 adjacent  floodplain and marshes. Improvements in
 hydrologic functions include reduced flow depths,
 and more appropriate distribution of conveyance of
 water and  sediment between the channel  and the
 floodplain. Benefits to fish habitat include  lower
 shear stresses and reduced gravel bed mobility, less
 sediment deposition in the channel and more  on the
 floodplain, and increased in-channel structure and
 habitat   diversity.   Additional  benefits  include
 reduced   flood  hazards   and   rejuvenation  of
 floodplain wetland channels.

 Geographic Information Systems
 Applied to Selection of Forested
 Wetland Restoration Sites

 S. Andrew Schulman
 Springwood Associates, Inc., 3644 Albion Place North,
 Seattle, Washington 98103.

   Restoration practitioners have typically depended
 upon pre-existing wetland inventories, such the Na-
 tional Wetlands  Inventory (NWI), to locate oppor-
 tunities for off-site restoration. However, mapping
 techniques employed to develop this inventory are
 limited in their ability to differentiate forested wet-
 lands from other forest communities. As a  result,
 identifying sites for forested wetland restoration
 often  involves extensive ground exploration  or re-
 newed inventory efforts,  either  of which  can be-
 come costly and  time-consuming undertakings. The
 rapidly growing volume of available digital data and
 increasing  access to Geographic Information Sys-
 tem (GIS) software offer  opportunities for stream-
 lining forested wetland restoration site selection. It
 is now not only possible, but cost-effective, to focus
 field exploration  by generating digital maps to indi-
 cate probable areas of degraded forested wetland.

  This paper describes a pilot study carried  out to
 identify suitable sites for restoration to compensate
 wetland impacts from a planned mining operation in
 the  Okanogan Highlands  of eastern Washington.
 Analysis  of the  100-square-mile study area  was
 based entirely upon existing, publicly available da-
 tasets. including vegetation maps, soil maps, surface
 hydrology maps, and Digital Elevation Models.
 Mapped  environmental variables were  analyzed
 using raster GIS software to  identify  statistically
significant spatial correlation with the wetland im-
pact sites and nearby known wetlands.  The result
was a set  of environmental  profiles for  various
wetland types. These environmental profiles were
then applied to the digital database as sorting crite-
ria  to generate a map indicating sites with  the po-
tential to support forested wetlands. This map was
in turn overlaid with an existing forest  cover data
layer to pinpoint likely locations for degraded for-
ested wetland systems.
  In addition to describing the  methods employed,
this paper addresses  advantages and limitations of
this approach to site selection, including issues of
scale and accuracy, costs, and data reliability.

An Evaluation  of the
Hydrogeomorphic Approach for
Assessing Restored Forested
Wetland Functions in the Puget
Sound Lowlands

Mark C. Rains1, Lyndon C. Lee, Jeffrey
H.  Braatne, and Jeffrey A. Mason
!L.C. Lee & Associates, Inc., 221 First Avenue West,
Suite 415, Seattle, Washington  98119.

  The hydrogeomorphic approach (HGM)  to  as-
sessing  wetland  functions was utilized  in  permit-
ting, design, and implementation of restoration and
monitoring  protocols for a forested wetland eco-
system on March Point in Padilla Bay, Washington.
HGM was used to assess depression and slope wet-
land functions in four categories: hydrology,  bio-
geochemistry,  plant  community maintenance, and
habitat/faunal support. Fifteen reference sites, clas-
sified on the basis of hydrologic and geomorphic
features, were used  to develop reference standards
for   impact  assessment,  design,  and  monitoring
protocols. When used in these contexts, HGM pro-
vided a mechanism  whereby (1) impacts could be
assessed and explicit restoration goals could be ne-
gotiated, (2) reference standards could  be  used as
restoration design templates, and (3) the restoration
of wetland  function could be monitored and com-
pared to regional reference  sites, and, if necessary,
appropriate contingency measures could be identi-
fied and implemented. After one growing season,
monitoring results indicate  that all classes of wet-
land functions increased relative to pre-project con-
ditions, yet most functions remained  lower  than
reference   standards.  All  hydrologic  functions
increased  relative to pre-project conditions,  and
surface and subsurface water storage functions were

comparable to reference standards. Biogeochemical
functions also increased relative to pre-project con-
ditions, but water quality conditions did not differ
significantly from  pre-project conditions.  Plant
community maintenance functions increased due to
the introduction of native wetland trees, shrubs, and
herbs.  However,  percent cover, density, and basal
area were  lower  than reference standards.  Habi-
tat/faunal  support functions such  as  habitat struc-
ture, interspersion, and connectivity also increased,
yet they remained lower than  reference standards.
The status of the wetland ecosystem restoration was
assessed after the first growing season using esti-
mates  of change  relative to specific wetland func-
tions. Study results show that HGM can be used to-
assess impacts and  to design  wetland restoration.
Furthermore, HGM can be used to assess the resto-
ration  of  wetland ecosystems  by measuring  rela-
tively  short-term  changes in wetland functions and
by comparing results to pre-project conditions and
to reference standards.

Use  of the Hydrogeomorphic
Approach  in Design of Riparian
Wetland Restoration along the
Central California Coast

Linda R. Ellis1, Lyndon C. Lee1, Peggy L.
Fiedler2,  and Mark C. Rains1
'L.C., Lee & Associates, Inc., 221 1st Avenue W, Seattle,
Washington 98119.
Department of Biology, San Francisco State University,
1600 Holloway Avenue, San Francisco, California

  The hydrogeomorphic approach (HGM) classifies
wetlands based upon their geomorphic setting, wa-
ter source and transport,  and  hydrodynamics. In
each  hydrogeomorphic wetland class,  HGM as-
sesses functions in four categories: hydrology, bio-
geochemistry, plant community maintenance,  and
habitat/faunal support.  The procedure provides  a
mechanism by  which (1) impacts can be  assessed
and  explicit restoration  goals  can be negotiated,
(2) reference standards can be established and used
as restoration design templates, and (3) the restora-
tion  of wetland  functions can be monitored and
compared to regional reference  sites,  and,  if neces-
sary,  appropriate contingency measures  can be
identified and implemented. In this study,  32 river-
ine wetlands and  19 estuarine  wetlands were sam-
pled along the Central California Coast. These data
have been  used to develop reference standards for
impact assessment and design criteria fox a restora-
                                Selected Abstracts
tion of the near-coast reach of Calera Creek in Paci-
ftca,  California. With  reference  standards  clearly
articulated, it has been possible to ascertain which
physical and biological factors need to be addressed
m the restoration design procedure. The reference
standards  show an ability to store surface water in
pits,  sloughs, channels, and ponds adjacent to a
wide, shallow channel located in a broad floodplain.
Energy reduction during high flows is enhanced by
high scrub-shrub and tree components.  Pre-project
conditions are characterized by little surface storage
capacity. Pits, sloughs, channels, and ponds are ab-
sent,  the channel  is narrow, deep, and highly  in-
cised,  and  there  is  virtually   no   floodplain.
Furthermore, velocity reduction is low due to low
scrub-shrub  and tree components. These results
suggest that the design of the ecosystem restoration
should focus on moving the creek out  of its cur-
rently incised channel and re-connecting it to its re-
stored floodplain.  To  complete  the   analysis,
measured and estimated levels of ecosystem func-
tioning were compared for reference standards, pre-
project, and  post-project  conditions.   Projected
effectiveness of the ecosystem restoration is shown
by the changes in the measured and estimated func-
tions. Functional assessment results provided a ba-
sis for negotiating permit conditions.

Middle  Waterway Pilot Restoration

Robert C. Clark, Jr.1, Dave McEntee*,
and Judy Lantor3
'NOAA Restoration Center Northwest, 7600 Sand Point
Way NE, Seattle, Washington 98115-0070.
2Simpson Tacoma Kraft Co., P.O. Box 2133, Tacoma,
Washington 98401.
3US Fish and Wildlife Service (F&WS), 3704 Griffin
Lane SE, Olympia, Washington 98501.

  Restoration of over  3  acres of intertidal, salt
marsh, and riparian habitat in the midst of Tacoma's
(Washington)  heavily  industrialized  Commence-
ment Bay has recently been  accomplished.  The
project, planned over  the past several  years and
constructed in June of 1995, has three basic compo-
nents: the excavation of tidal  channels similar to
those existing in  the adjacent relic mudflats; con-
struction of a vegetated  intertidal bench; and the
removal of upland fill, isolation of a small quantity
of metal debris, and resloping  the head of the wa-
terway The twin goals of this pilot project are to
create productive and diverse estuarine habitats for
fish  and wildlife and  to provide  an outdoor
laboratory for studying various planting strategies,

 Selected Abstracts
 monitoring  protocols,  and  volunteer/stewardship
 models  for  future  natural  resources  restoration
 efforts in the region. This successful joint industry-
 government-public  partnership  between Simpson
 Tacoma Kraft  Company, Champion International,
 and the natural resource trustees [U.S. Depts. of the
 Interior  (F&WS)  and Commerce (NOAA),  WA
 Depts. of Ecology, Fish and Wildlife, and  Natural
 Resources, the Puyallup Tribe of Indians,  and the
 Muckleshoot   Indian  Tribe]  resulted   in   the
 conversion  of  1.5 acres  of  industrial upland to
 estuarine high and  low salt  marsh and intertidal
 mud/sand fiat  habitat. The natural  shoreline edge
 along the +9 to +13  foot  contour was increased
 from  840  to   960 feet.  A  riparian  buffer  and
 transition zone was built to screen, protect, and
 support  the  diversity of the  remaining  Middle
 Waterway mud flats.

 Nearshore Industrialized Areas of
 Port Gardner, Washington, and
 their Effects on Early Marine Life
 History of Anadromous Fishes

 J.P. Houghton1, M.A. Kyte1, and  D.L.
 'Pentec Environmental, Inc., Edmonds, Washington.
 2Port of Everett, Washington.

  Port Gardner, a  natural,  deep-water harbor adja-
 cent to the Snohomish River estuary, is critical both
 to the economy of Everett and to  local fish and in-
 vertebrate resources. Changes in juvenile salmonid
 habitat quality and function (e.g., feeding and ref-
 uge) were evaluated relative to historic development
 of the waterfront.  For decades, commercial and in-
 dustrial  uses  progressively  degraded  nearshore
 habitats, which  function as migration corridors for
 juvenile salmonids, into a maze of piling walls, ver-
 tical bulkheads, finger piers,  and over-water saw-
 mills.  Log  storage  on the  mudflats  created  ob-
 structions to migratory fish  and  reduced  benthic
 productivity by shading or compressing the  sedi-
 ments. Chemical and  perhaps thermal discharges
 from a variety of industries also negatively affected
 water and sediment quality. By the late 1950s, many
 of the historic sawmills were closed although sev-
 eral  pulp mills were active for many more years.

  Large-scale filling along the shoreline  which  con-
 tinued into  the 1970s eliminated large areas of
 shallow water mud fiats, contributing to the overall
 reduced  biological productivity of nearshore  Port
 Gardner. At the same time, however, these  fills of-
 ten covered in-water pile accumulations of sawdust
and debris and simplified the shoreline, eliminating
many of the pile walls and floating and over-water
structures. Fills were either riprapped or bulkheaded
for structural stability.
 More recent and improved water quality regula-
tions and log handling practices, as well as regula-
tion of nearshore dredging and filling, are changing
conditions found along this corridor.  Redevelop-
ment actions, such as replacing vertical bulkheads
with sloped riprap, also may improve productivity
for salmon food organisms and enhance  refuge from
predators. Source controls and redevelopment of the
shorelines in Port Gardner and restoration of inter-
tidal marsh habitats in the estuary are reversing, to
some  degree, historically negative environmental
effects of urbanization on nearshore  habitats  en-
countered by outmigrating juvenile salmon.

Importance of Tidal Channel
Geomorphology to  Restoring
Ecological Functions of Coastal

Charles A. Simenstad1, Lawrence P.
Rozas2, Thomas  J. Minello2,  Denise J.
Reed3, Robert N. Coats4, and Joy Zedler5
'Wetland Ecosystem Team, School of Fisheries, Univer-
sity of Washington, Box 357980, Seattle, Washington
2NOAA-NMFS, 4700 Avenue U, Galveston, Texas
77551,3LUMCON, 8124 Highway 56, Chauvin,
Louisiana 70344-2124
"Philip Williams & Associates, Ltd., Pier 35, The Embar-
cadero, San Francisco, California 94133.
5Pacific Estuarine Research Laboratory, San Diego State
University, San Diego,  California 92182-4614.

 Divergent ecological functions in constructed ver-
sus  natural coastal marshes  may result from inher-
ent differences in marsh geomorphology, especially
morphology of tidal  channel systems. We hypothe-
size that nutrient cycling, species assemblages, pri-
mary  and  secondary  production,  and  species
interactions are strongly influenced by such land-
scape-scale features. Dendritic  channel networks
form through lengthy processes of sediment erosion
and accretion, channel migration and extension, and
plant  and animal community development, proc-
esses that ultimately define complex,  ecologically-
important features such as fish and wildlife habitat.
Although available  geotechnical tools can  explain
influences of tidal prism and range, channel flow
and sediment transport on tidal channel systems, we
still  lack empirical  tests of relationships between

tidal channel geomorphology and ecological func-
tion. Using examples from Gulf of Mexico, Califor-
nia and Pacific Northwest coastal marshes, we infer
how tidal channel planform (e.g., channel density,
bifurcation, and sinuosity), cross section, and slope
influence fish, macroinvertebrate and avifauna ac-
cess, refuge and foraging,  and describe how devel-
opment of tidal  channel  systems and  ecological
functions both differ among these regions. In re-
storing and constructing coastal marshes,  we are
faced  with decisions of how to promote  natural
channel systems, ranging  from no intervention to
extensive engineering manipulations. A better un-
derstanding of  basic  and  empirical geomorphic
principles that regulate natural tidal channel forma-
tion  along  pathways toward predictable ecological
functions will allow us to make more educated deci-
sions about marsh restoration.

Creating Landscape  Restoration
Opportunities: A Case Study at
Rattlesnake Lake

Nancy Bottle1
1 Jones & Jones Architects and Landscape Architects, 105
South Main, Seattle, Washington 98104.

  Changing patterns in visitor recreational use can
serve as catalysts for redesign, creating opportuni-
ties for landscape and habitat restoration. In master
planning for Rattlesnake Lake in the Cedar River
Watershed, Jones & Jones recognized possibilities
for improving habitat and water quality while ad-
dressing  visitor  use issues and expanding  recrea-
tional  trails  and  facilities. To achieve multiple
objectives, we  re-routed  a  section  of road away
from the lake edge and re-sited parking areas to
disturbed land on higher ground. This plan  created
possibilities  for  restoring  riparian  habitat  and
planting new upland forests and meadows, while
providing nearly a mile of recreational trails.  In
developing the design, we proposed the use of bio-
engineering techniques and native riparian plantings
along the lake edge, and plan to employ native for-
est succession to restore upland areas. As a result of
the road re-routing, the client will also daylight and
restore over 100 feet of culverted creek. In  solving
multiple  issues,  support from the client and user
constituency has been maintained. Interpretive  signs
will  highlight natural and cultural history,  explain
restoration   processes,   and   promote  personal
stewardship of the Rattlesnake Lake area.
 An Assessment of the Mycorrhizal
 Fungal Status of Wetland Prairie
 Plant Species: Implications for
 Ecological Restoration

 Stephen D. Turner1, Carl F. Friese
 Department of Biology, University of Dayton Dayton
 Ohio 45469-2320.

  Wetland ecosystems are important because of
 their ecological significance in such areas as natural
 waste and nutrient recycling. Much has been ac-
 complished  in   determining  the  floristics  and
 edaphic characteristics of these areas but  little is
 known about the function of the below ground sys-
 tem. This study  was designed to collect descriptive
 data on the mutualistic mycorrhizal fungal commu-
 nity of a wetland  prairie (Zimmerman  Prairie,
 Greene County, Ohio) associated with plant species
 diversity and soil characteristics. Preliminary analy-
 sis of the soil sample from the wetland prairie indi-
cate a positive correlation between mycorrhizal root
infection and soil moisture gradient for a 7-month
time period.  The overall analysis of the wetland
prairie  will determine the  significance of the my-
 corrhizal fungi and will assist in the restoration of
 other wetlands.

 Watershed  Restoration in Deer
 Creek, Washington-A Ten-Year

 James Doyle1, Michelle Fisher, Greta
 Movassaghi, Roger Nichols
 'United States Forest Service, Mt. Baker Snoqualmie Na-
 tional Forest, 21905 64th Avenue West, Mountlake Ter-
 race, Washington  98043.

  Deer Creek  is  a tributary to  the North Fork
 Stillaguamish River in northwest Washington State.
 Approximately two thirds of the watershed is in fed-
 eral ownership. Deer Creek once supported one of
 the largest populations of native, wild summer run
 steelhead in the Pacific Northwest, and also pro-
 vided  habitat  for  coho salmon  and native char
 populations.  Concentrated   timber  management
 activities on landslide-prone  areas throughout the
 watershed and the occurrence of a number of major
 storm events over the past 30 to 40 years have been
 the major reason  for the loss or degradation of
 aquatic habitat and a subsequent decline  and near
 extinction of the native fish populations. Since the
 early 1980's major landslides and flood events have

 Selected Abstracts
 dropped these anadromous fish populations to criti-
 cal  levels of viability. Beginning in 1984, Deer
 Creek has been the focus for comprehensive water-
 shed  planning  and  restoration.  All landowners
 began  to coordinate  short  and long  term timber
 harvesting operations within the watershed; at this
 time the U.S. Forest Service deferred further timber
 harvesting in Deer Creek due to the critical decline
 in fish populations and the  loss/decline in fish
 habitat conditions. Watershed restoration was initi-
 ated on federal  land  in 1984 with the goal  of re-
 storing  the  watershed's physical  and  biological
 processes that formed and maintained the favorable
 aquatic habitat conditions of the past. Initial resto-
 ration  project emphasis was to reduce the occur-
 rence of mass soil movement (landslides and debris
 torrents)  from logging roads, old harvest units, and
 naturally unstable areas. Another phase of restora-
 tion work occurring at the same time was upgrading
 and stormproofing the existing road system. Second
 stage of work was restoring structure to the stream
 channel network with the  goal of controlling  the
 bedload transport and deposition as well as stre-
 ambank  stability. In  1994 more  than 460 adult
 steelhead returned to Deer Creek, a significant in-
 crease from the past 6 to 7 years when less than 100
 returned  to spawn. This documented  increase has
 been due mainly to the increase in the freshwater
 survival of steelhead juveniles. By protecting and
 restoring aquatic habitat conditions in Deer Creek
 anadromous  fish populations are rebounding in this
 watershed. This is a review of watershed manage-
 ment efforts  as we pass the 10-year mark and look
 toward the future.

 Stream Restoration in an Alaskan
 Subalpine Placer-Mined

 Roseann V. Densmore1, Kenneth Karle2

 National Biological Service, 1011 East Tudor Road,
 Anchorage, Alaska 99503.
 2National Park Service, P.O. Box 9, Denali National Park
 and Preserve, Alaska 99755.

  Techniques to promote riparian ecosystem recov-
 ery  are being tested  in a  subalpine  placer-mined
 watershed in Alaska's Denali National  Park and
 Preserve.  Stream and floodplain geometry were de-
 signed to  allow the stream  to develop floodplains,
 sinuosity,  and pools and riffles similar to premining
 conditions. Bioengineering  techniques, which util-
 ized subalpine plant materials, were tested for sta-
 bilization. Earthwork in 1991 and 1992 emphasized
recontouring  artificially  raised floodplains  to  a
shallow slope to  restore natural floodplain proc-
esses, and constructing a segment of new channel
away from the valley  wall. Anchored alder  and
willow brush bars were  installed to stabilize the
floodplain and collect sediment. A moderate flood
occurred at the end of the 1992 construction.  The
brush bars stabilized the flood-plain surface  and
trapped sediment, but unconsolidated bank material
and  a lack  of bed  armor for the  new  channel
segment led to  some bank erosion, slope changes,
and an increase in sinuosity in  several reaches. To
address these problems,  we tested new techniques
in 1994. These included construction of pools, rock
weirs, and meanders with deeper channels. Bioen-
gineering techniques focused on stabilization  and
revegetation  of the newly-created  meander  cut
banks and point bars. On cut banks, root wads and
brush bars (redesigned to protect both flood-plain
and  bank)  were tested.  On point  bars,  riparian
vegetation and soil from sites away from the present
channel were bulldozed over the new point bars, or
a brush layer was buried with  the ends projecting
into  the channel. All bioengineering methods used
buried willow branches for revegetation.

Restoration of Ashbridge Marsh:
Potential Consequences for Water
Quality and the Aquatic
Community of the Lower  Don

James M.  Helfield1, Miriam L. Diamond

 Department of Geography, University of Toronto, 100
Street, George Street,