United States
Environmental Protection
Agency
Office 01 Ecosystems & Communities
Region 10
1200 Sixth Avenue
Seattle WA 981 01
Aquatic Resources Unit
Alaska
Idaho
Oregon
Washington
December 1997
EPA910-R-97-007
Wetland and Riparian
Restoration:
Taking a Broader View
Proceedings of a Conference
.WJ
A ~
4
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
E:....O.
-: . .
..
Northwest Chapter
Society for Ecological Restoration
Pacific Northwest Chapter
Society of Wetland Scientists
J
-------
Wetland and
Riparian
Restoration:
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
1997
-------
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,
Washington.
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
wording.
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.
Availability
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
O
Printed on Recycled and
Recyclable Paper
-------
Contents
Page
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
Implications
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
Tolerance
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
Arizona
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
Regeneration
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
VI
-------
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
sea980350031.doc
Vll
-------
Sponsorship
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
via
-------
Acknowledgments
We especially appreciate and gratefully acknowledge
the enthusiasm, excellent contributions, and patience
of all the many authors who made this Proceedings
possible.
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.
IX
-------
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
locations.
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
restorationists.
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.
Introduction
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.
XI
-------
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.
Xli
-------
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.
xm
-------
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).
Restoration
An increasingly important research theme and the
one driving these Proceedings, concerns wetland
and riparian habitat restoration, enhancement, and
XIV
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
xv
-------
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
1993).
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
xvi
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.
xvn
-------
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).
Integration
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
efforts.
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
1996).
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
XVlll
(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
restoration?
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
1996).
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").
xix
-------
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.
1995a).
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
xx
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
projects.
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-
sions.
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.
Conclusions
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
xxi
-------
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
cit).
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.
XXll
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.
16pp.
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.
159pp.
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.
xxm
-------
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
pp.
Hamilton, P. and K. B. Macdonald. 1979. Estuarine
and wetland processes, with emphasis on model-
ing. Marine Science 11, Plenum Press, NY. 653
pp.
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
cit).
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
XXIV
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,
Canada.
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
pp.
Mann, C. and M. Plummer. 1997. Qualified thumbs
up for habitat plan science. Science 278:2052-
2053.
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.
xxv
-------
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
275:499-500.
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
XXVI
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-
1663.
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-
92.
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-
727.
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.
30pp.
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.
174pp.
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,
XXVH
-------
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-
1188.
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.
XXVlII
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
pp.
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.
sea/972650006.doc
-------
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
Seattle-1997
-------
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-
sessment.
Introduction
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.
Methods
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
environment.
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
wetlands.
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"
manner.
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
complex.
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
Improvement
Export of Primary
Productivity
Fish Habitat
Bird Habitat
Stabilization of Shoreline and
Channels
Access to Transportation
Corridors
Recreation
Priority Species Habitat
Aesthetic Value
Sediment Stabilization
Sediment Toxicant Retention
Nutrient Retention and
Transformation
Invertebrates
(includes rearing, feeding,
migration, and shallow water
refuge functions)
non-anadromous
anadromous
migratory
overwintering
nesting
Other Species Habitat Herptiles
Mammals
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
area.
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.
Results
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
colonization.
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"
categories.
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
Attributes
Maiysville
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
0'
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
Island.
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
areas.
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
influence.
Discussion - Management Plan
Development
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
functions.
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
-------
SEWIP
Development
Footprint & Tidal
Restoration/
Enhancement
Sites
//Maalsby
> // 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
wetlands.
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.
Conclusions
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"
perspective.
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
11
-------
Snohomish Estuary
Table 2. Summary of development footprint impacts and restoration credits for Poortinga site in Snohomish Estuary
Site
Impacts to Water
Quality Functions
in Acre Points
Restoration Credits
for Water Quality
Functions in Acre
Points
Impacts to Wildlife
Functions in Acre
Points
Restoration Credits
for Water Quality
Functions in Acre
Points
Development Footprint for
Vegetated Wetlands
Priority #1 Restoration
Site, Poortinga Property
(350 acres)
10,761
None
None
16,350
4,500
None
18,326
Table 3. Calculated compensation ratios for Snohomish Estuary - vegetated wetlands
Restoration
Priority and
Location
Priority 1 -
Poortinga
Priority 1 -
Poortinga
Priority 2 - N.
Spencer (Beringer)
Priority 2 - N.
Spencer (Beringer )
Priority 3 - E. Smith
Island
Priority 3 - E. Smith
Island
Priority 4 & 4A -
East Mainland
Priority 4 & 4A -
East Mainland
Wetland
Functions
Wildlife
Water
Quality
Wildlife
Water
Quality
Wildlife
Water
Quality
Wildlife
Water
Quality
Average
Performance
of Functions
for
Development
Footprint
(per acre)
13
31
13
31
13
31
13
31
Average
Increase in
Performance
for
Restoration-
Enhancement
Area, (per
acre)
57
51
54.5
41.8
51.2
32.8
53.4
26.6
Compensation
Ratio to offset
function loss
& temporal
loss (i.e. 1.25
x functional
loss ratio)
0.3:1
0.8:1
0.3:1
0.9:1
0.3:1
1.2:1
0.3:1
1.5:1
Compensation
Ratio Based on
Priority &
Acreage
1x
1.3x
1.5x
2x
Reasons for
Selected
Compensation
Ratio
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
12
-------
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
permanently.
Acknowledgments
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
Everett.
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.
13
-------
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
Raton.
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,
WA.
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/
OBS-83/05.
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.
sea971830001.doc
14
-------
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
conditions.
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.
Introduction
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.
15
-------
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
(SEWIP).
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
form:
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
wetland)
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
16
-------
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
Recreation
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
Recreation
Channel Stabilization
Access to Deep Water
(transportation)
Anadromous Fish
Resident Fish
Migratory Bird
Overwintering Bird
Breeding Bird
Invertebrate
Reptile and Amphibian
Mammal
SET RESTORATION GOAL
i
IDENTIFY WETLANDS FOR RESTORATION
SET RESTORATION OBJECTIVE FOR EACH WETLAND
SET RESTORATION TACTICS FOR EACH WETLAND
SCORE RESTORATION POTENTIAL
ESTIMATE REGIONAL RESTORATION VALUES
Figure 1. Planning steps in developing a restoration plan linked to a regional wetland management plan.
17
-------
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
Restoration
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
restoration.
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
Slough.
Wetland 5KKK: Raise dissolved oxygen to
acceptable levels.
18
-------
_
Wetland 5L: Increase water supply in spring
and early summer to encourage nesting by
waterfowl.
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
"mosaic").
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.
Discussion
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
19
-------
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
Score
Before
SITE Hectares Restoration
146
147
339
340
341
342
343
360
361
362
363
— •• • — .
9.2
33.9
7.8
6.8
23.8
5.0
7.0
1.8
9.9
14.9
8.8
— — ,
15
14
12
10
13
18
15
11
15
12
9
• •
Net
Increase
After Pts/ha
72
75
67
67
66
66
68
70
74
68
64
58
61
55
57
53
4.8
54
59
59
57
56
Water Quality Improvement Group of Functions
Total Score
Increase Before
ha/Pts Restoration
531
2,064
423
387
1,260
239
373
105
583
848
489
10
14
38
47
41
26
41
17
10
9
9
Net
Increase
After Pts/ha
79
82
79
79
79
50
79
66
79
50
50
69
68
41
33
38
23
38
50
68
40
40
Total
Increase
ha/Pts
637
2,315
317
223
907
116
265
89
683
605
356
-------
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
Species
Water Quality Improvement
AVERAGE SCORE/ACRE IN:
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
gain/acre
AVERAGE compensation ratio by 0.5
acre of impact (including 1.25
multiplier)
56
64
98
34
1.7
2.1
65
75
105
30
2.2
2.8
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.
SEA/971900001.doc
21
-------
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.
Introduction
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
Study
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.
~n
-------
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
23
-------
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
gain.
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-
lands).
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
method.
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.
Advantages
1. Compensation projects occur in advance of un-
avoidable, adverse impacts to wetlands allowing
regulators to measure their success over a span of
years.
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-
quirements.
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.
24
-------
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.
Disadvantages
1. Mitigation planning is weakened when ad-
vanced compensatory mitigation projects create in-
centives to substitute "compensation" for
"avoidance" and "minimization" under the mitigation
sequence.
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
below.
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
environments.
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
25
-------
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-
ues.
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.
Advantages
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
property.
5. Cooperation and coordination between programs
with overlapping and/or complementary goals and
objectives.
6. Planning includes local needs, conditions, and per-
spectives.
Disadvantages
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
unit.
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-
lands.
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.
26
-------
Wetland Compensation Banking
Integrating Wetland Compensation
Banking Into Local Wetland
Planning
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
Activities
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-
tion.
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-
tion.
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.
27
-------
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
independently.
Advantages
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.
Disadvantages
1. Lack of regulatory guidance and experience for
implementing wetland compensation banking and
local wetland planning.
28
-------
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
projects.
Discussion of Advantages and
Disadvantages
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
success.
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
29
-------
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,
WA.
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,
NY.
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
WA.
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
30
-------
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.
sea971900006.doc
31
-------
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.
Introduction
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.
32
-------
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
(SEPA).
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
33
-------
Watershed Analysis
Table 1. Summary of Watershed Processes and Resources Addressed by the Washington
State Watershed Analysis Modules
Watershed Analysis
Module
Mass Wasting
Surface Erosion
Hydrology
Riparian Function
Channel Condition
Fish Habitat
Water Supply / Public
Works
. - I .- * ,-*"" »
Watershed Processes and Resources Addressed
.* — :>- ' "' .: - """' '>:':^l N. - '
• Debris Torrents
• Landslides
• Earthflows
• Hillslope Surface Erosion
Gullying
- Dry Ravel
Sheetwash
• 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
Resource
Vulnerability
Low
Medium
High
Low
Standard Rules
Standard Rules
Standard Rules
Medium
Standard Rules
Minimize
Prevent
High
Prevent
Prevent
Prevent
34
-------
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
resources.
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
developed.
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).
Sediment
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.
35
-------
Watershed Analysis
Hydrology
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-
tions.
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.
Temperature
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
36
-------
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
Habitat
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
37
-------
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-
agement.
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.
Monitoring
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
watershed.
Quartz Mountain Watershed
Analysis
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
38
-------
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,
1994).
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
fisheries.
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
40
-------
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
analysis.
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
41
-------
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
terraces.
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
slopes.
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
Creek.
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.
42
-------
Watershed Analysis
Prescriptions
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
period.
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
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
watershed.
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
activities.
Watershed Analysis and
Restoration
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
43
-------
Watershed Analysis
Resource of
Concern
Threats to
Resource
Confidence in
Linkage
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
explored
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.
44
-------
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.
Conclusions
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,
WA.
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
28:371-383.
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-
436.
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.
45
-------
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-
801.
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,
WA.
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.
sea972120014.doc
46
-------
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
whole.
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.
47
-------
Prioritization of Restoration Projects
Table 1. Prioritization Questions and Scoring for Tributary/Habitat Problems
Question
Score
1. If nothing is done, will the problem get
worse?
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
concerns)?
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
app.
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.
117pp.
sea972040003.doc
48
-------
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
Zone)
+
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.
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.
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.
49
-------
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.
Introduction
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.
50
-------
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
channel
• Vertical distribution of foliage
• Many sizes and species of trees in the adjacent
forests.
Longitudinal Changes in Ecological
Processes
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
1989).
51
-------
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
1990).
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
Northwest
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
systems.
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
variability.
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
52
-------
Table 1. Stand type classes identified in the assessment of riparian forest stands
al.1994).
Stand Type
Open
Brush
Description
Recently disturbed (landslide, harvest, gravel bar);
unvegetated or herbs and grasses
Shade
potential
LOW
in the Deer Creek basin, (after Collins et
LWD
recruitment
potential
LOW
Bank
reinforcement
potential
LOW
Vegetated - small diameter (<6") trees and shrubs, LOW
herbs
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
LOW
MOD
MOD
LOW
MOD"
MOD3
MOD
LOW
MOD
MOD
MOD
MOD8
MOD"
MODa
Mature
Conifer - C
Mixed-M
Hardwood -
>18" average DBH
HIGH
HIGH
HIGHb
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.
53
-------
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
habitats).
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-
ceed.
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/
54
-------
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
55
-------
in
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
channel
Open: Ugh bar
Open: floodplaln or
upland
Brush : conifer
(dense, small dimen-
sion, low growth
potential)
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
cedar).
- 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"
DBH).
cedar - 90 yr.
D.fir-70yr.
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.
D.fir-90yr.
100 yr.
130 yr.
hemlock, fir - 100 yr.
cedar -150 years
Cost
Labor
S80/ac
•release-
$40/ac
-plant-
$100/ac
S150/ac
•release-
$150/ac
j
-plant-
$100/ac
-release-
$150/ac;
-plant-
SlOO.ae
Cost
Planting Stock
$40/aca
S50/acb for 200
tree/ac
-plant-
$50/ac
-ptant-
S25/ac
Log
Value
Net
PNV
OPTION 1
-$120/ac
OPTION 1
-S190/BC
OPTION 1
-$ISO/ac
OPTION 1
-$300/ac
OPTION 1
-S275/ac
-------
Sapling: conifer
Sapling: deciduous
Sapling:
mixed
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
trees/ac.
- 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-
growth").
- 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
"old-growth")
OPTION 2:30 yr.
OPTION 3 :30yr.
80 years
130 yr.
80 yr.
OPTION I:70yr.
(0 yrs in silver fir zone
"old-growth")
OPTION 2: 50 yr.
OPTION 3:50 yr.
-release -
SISO/ac;
-comm.
thin-
S400/ac
(
-------
oo
Pole: deciduous
Pole: mixed
- objective: large conifer (LWD recruitment and shade).
- OPTION 1 : no action (best option in silver fir zone "old-
growth").
- OPTION 2: thin, no extraction, plant Douglas fir and cedar
to200trccs/ac.
- OPTION 3: commercial thin , plant Douglas fir and cedar to
200trees/ac.
- objective: large conifer (LWD recruitment and shade).
- OPTION 1 : no action (best option in silver fir zone "old-
growth").
- OPTION 2: thin, no extraction, plant Douglas fir and cedar
to 200 trees ac.
- OPTION 3 : commercial thin , plant Douglas fir and cedar to
200trees/ac.
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.
-thin-
$200acre
-comm. thin-
SlOOO/ac
(ifS200'MBF
logging and
haul; 5 MBF/
acre)
•plant-
SlOO/ac
-thin-
$200/ac
-comm. thin-
SlOOO/ac
CgS200/
MBF logging
and haul1, 5
MBF/ ae)
-plant-
SlOO/a
•plant-
$50/ac
-plant-
S25/acre
•comm. thin -
SlOOO/ac
(conifer (S
S200/MBF;
5 MBF/ac)
-comm. thin-
SlOOO/ac
(conifer @
S200/MBF; 5
MBF/ac)
OPTION 2
-$350/ac
OPTION 3
-SlSO/ac
OPTION 2
-$325/ac
OPTION 3
-$125/ac
-------
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).
Conclusions
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-
ington.
Berg, D. R. 1995. Riparian silvicultural system de-
sign and assessment. In the Pacific Northwest Cas-
cade Mountains. Ecological Applications. 5,1:87-
96.
Bilby, R. E. 1981. Role of organic debris dams in
regulating the export of dissolved and particulate
matter from a forested watershed. Ecology
62:1234-1243
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.
118:368-378
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.
59
-------
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.
216pp.
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
102:320-352.
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.
15:133-302.
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,
AZ.
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-
329.
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,
60
-------
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-
349.
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-
137.
Washington Forest Practices Board (WFPB). 1993.
Standard Methodology for Conducting Watershed
Analysis. Washington State Department of Natural
Resources. Olympia, WA.
sea972300013.doc
61
-------
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
(0.843(0085x),r2=0.71).
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-
cession.
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.
62
-------
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.
SEA7972270003.doc
63
-------
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
system.
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-
tions.
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.
64
-------
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,
OR.
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.
sea971830003.doc
65
-------
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.
Introduction
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.
66
-------
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
environment.
Methods
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
States.
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
birders.
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
Mapping
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
wetland.
Habitat Availability and Suitability
Analysis
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.
67
-------
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
analysis.
Results
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
1995).
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
1982).
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
are.
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
68
-------
Base map reproduced with permission of
Satellite Images Inc., Anacortes, Washington.
t
N
Okm
I
Skm
I
IOkm
I
Legend ~ Wetlands identified on NW I maps
0-
-D
-------
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
Analysis
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.
70
Discussion
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,
OR.
SEA972120013.doc
71
-------
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.
Introduction
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
1986).
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
1996).
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
1997).
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
climates.
Methods
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
species.
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.
Results
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
73
-------
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
plantations.
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
74
-------
Table 1. Home Range and Dispersal Distances of Selected Amphibians.
Species
SALAMANDERS
NORTHWESTERN SALAMANDER
Ambystoma gracile
JEFFERSON SALAMANDER
Ambystoma jeffersonii
SPOTTED SALAMANDER
Ambystoma maculatum
MARBLED SALAMANDER
Ambystoma opacum
MOLE SALAMANDER
Ambystoma talpoideum
SMALL-MOUTHED SALAMANDER
Ambystoma texanum
FROGS
PACIFIC TREEFROG
Hyla regilla
NORTHERN LEOPARD FROG
Rana pipiens
POOL FROG
Rana lessonae
WOOD FROG
Rana sylvatica
TOADS
AMERICAN TOAD
Bufo americanus
COMMON TOAD
Bufo bufo
NATTERJACK TOAD
Bufo calamita
Home Range (m2)"
Mean
(<200)
[1-181]
[1-39]
{12-14}
[13-226]
[<1-23]
[1-5]
[68-503]
{60}
[50-150]
[50-150]
Dispersal
Distances (m)b
{1,000}
(250)
{625} (252)
{1,610}
{152}
[0-125] (64)
[157-250]
[6-220] (150)
[0-450] (194)
[81-261]
[0-1 25] (52)
[500-914]
[34-300]
{15,000}
{2,530} (1,1 26)
400
3,000
[70-1,600]
1,400 [100-200]
Location
Washington
Kentucky
Indiana
New York
Michigan
Indiana
Michigan
Kentucky
Indiana
South Carolina
Indiana
California
Michigan
Austria/Hungary
Minnesota
Virginia
Minnesota
Minnesota
Switzerland
Germany
England
References
Stringer pers. comm.
Douglas & Monroe 1981
Williams 1970
Bishop 1941 in Williams
1970
Wacasey1961 in Williams
1970
Williams 1970
Kleeberger & Werner 1983
Douglas & Monroe 1981
Williams 1970
Semlitsch 1981
Williams 1970
Brattstrom & Warren 1955
Dole 1965
Tunner 1991 in Dingle
1996
Bellis 1962
Berven & Grudzien 1990
Ewert 1969 in Christein &
Taylor 1978
Bellis 1959
Heusser1968
Sinsch 1988a
Beebee 1983
Minimum polygon: [Range], (mean), {linear distance}.
[Range], (mean), {maximum}.
75
-------
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
76
-------
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
corridors.
77
-------
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-
morphosis.
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
species.
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
rare.
Breeding amphibians may select medium-depth
water because of optimum temperatures that
78
-------
Lentic-Breeding Amphibians
Table 2. Oviposition Water Depths of Selected Pacific Northwest Amphibians
Species
SALAMANDERS
Water Depth
(cm)a
Depth Below
Surface (cm)" Location
References
NORTHWESTERN SALAMANDER
Ambystoma gracile
LONG-TOED SALAMANDER
Ambystoma macrodactylum
FROGS
RED-LEGGED FROG
Rana aurora
CASCADES FROG
Rana cascadae
BULLFROG (INTRODUCED)
Rana catesbeiana
SOUTHERN LEOPARD FROG
Rana sphenocephala
PACIFIC TREEFROG
Hyla regilla
[15-61]
[0-122]
200-300"
[7-22]
[3-39] (11)
[30-152] [61-122]
[26-47] (28)
[5-18]
[42-54]
(32) & (40)
[5-45] (20)
(7)
200-300
Washington
Washington
Washington
12-65 California
tracks Washington
substrate0
Slater 1936
Richter &
Roughgarden
unpublished
manuscript
Crisafulli personal
communication
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
touching
surface
surface to Washington
subsurface
Washington
Sype 1974
Richter personal
observation
(4)&(15) South Carolina Caldwell 1986
Richter personal
observation
TOADS
WESTERN TOAD
Bufo boreas
[5-100]
Oregon
Olson 1988
a [Range], (mean).
b Personal observation.
°Eggs typically a fixed distance above the bottom.
79
-------
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
metamorphosis.
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-
tion).
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
80
-------
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
carbon.
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.
81
-------
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
82
-------
Lentic-Breeding Amphibians
Table 3. Puget Sound Lowland Shallow Marsh Hydrophytes Used as Amphibian Oviposition Sites.
Amphibian Species
Additional Material
Used
Scientific Name
Common Name
LONG-TOED SALAMANDER
Ambystoma macrodactylum
PACIFIC TREEFROG
Hyla regilla
OREGON SPOTTED FROG
Rana pretiosa
1 to 2 mm Diameter, Thin-stemmed Emergents
Roothairs, rootlets & thin
twigs.
Some roothairs, rootlets &
thin twigs.
Some variable-diameter
emergents.
POACAE
Alopecurus aequalis
A. geniculatus
Agrostis aequivalvis
Calamagrostis
canadensis
Glyceria grandis
G. elata
G. occidentalis
CYPERACEAE
Carex obnupta
Eleocharis palustris
E. ovata
E. acicularis
JUNCACEAE
Juncus acuminatus
J. bufonius
GRASS FAMILY
Shortawn foxtail
Water foxtail
Alaska bentgrass
Bluejoint reedgrass
American mannagrass
Tall mannagrass
Western mannagrass
SEDGEFAMILY
Slough sedge
Common spikerush
Ovate spikerush
Needle spikerush
RUSH FAMILY
Tapered rush
Toad rush
NORTHWESTERN
SALAMANDER
Ambystoma macrodactylum
RED-LEGGED FROG
Rana aurora
3 to 6 mm Diameter, Medium-stemmed Emergents
CYPERACEAE.
Carex athrostachya
Some variable-diameter
emergents.
C. utriculata
CYPERACEAE
Carex obnupta
MENYANTHACEAE
Menyanthes trifoliata
APIACEAE
Oenanthe sarmentosa
POLYGONACEAE
Polygonum amphibium
P. hydropiperoides
P. punctatum
POTAMOGETONACEAE
Potamogeton natans
P. emersum
P. gramineus
SPARGANIACEAE
Sparganium emersum
S. eurycarpum
SEDGE FAMILY
Slender-beaked sedge
Beaked sedge
RUSH FAMILY
Slough sedge
BUCK-BEAN FAMILY
Buckbean
PARSLEY FAMILY
Water-parsley
BUCKWHEAT FAMILY
Water smartweed
Waterpepper
Dotted smartweed
PONDWEED FAMILY
Floating-leaf pond
weed
Emersed pondweed
Grass-leaved pond
weed
BUR-REED FAMILY
Simple:stem Bur-reed
Broadffuit Bur-reed
ROUGH-SKINNED NEWT
Taricha granulosa
WESTERN TOAD
Bufo boreas
Variable Diameter Emergents
Leafy submerged plants. CYPERACEAE
Scirpus cyperinus
S. microcarpus
Grasses & leafy grass-like
plants.
SEDGE FAMILY
Woolgrass
Small-fruited Bullrush
83
-------
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-
84
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-
able.
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.
Discussion
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-
tion.
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
85
-------
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-
tions.
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.
Acknowledgments
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-
GTR-285.
86
-------
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
26:353-360.
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:
173-174.
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
44:2047-2056.
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
1987:386-395.
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.
87
-------
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.
23:878-882.
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-
425.
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,
ML
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-
38.
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.
FWS/OBS-79/31.
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
5:1579-1595.
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,
England.
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-
166.
-------
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-
478.
Douglas, M. E. and B. L. Monroe Jr. 1981. A com-
parative study of topographical orientation in Am-
bystoma (Amphibia: Caudata). Copeia 1981:460-
463.
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
1966:60-74.
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-
653.
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:
66:1762-1768.
Forman, R. T. T. 1983 Corridors in landscapes:
Their ecological structure and function. Ecology
(CSSR)2:375-387.
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-
0921.
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-
166.
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,
Australia.
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-
316.
Jameson, D. L. 1957. Population structure and
homing responses in the Pacific tree frog. Copeia
1957:221-228.
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
17:176-177.
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-
GTR-285.
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
52:116-124.
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-
285.
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.
90
-------
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.
44:63-100.
Olson, D. H. 1988. The ecological and behavioral
dynamics of breeding in three sympatric anuran
amphibians. Ph.D. Dissertation. Oregon State
University.
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,
Canada.
Payne, N. F. 1992. Techniques for Wildlife Habitat
Management of Wetlands. Biological Resource
Management Series, McGraw-Hill, New York,
NY.
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
91
-------
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
26:413-422.
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-
111.
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
7:131-136.
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.
11:178-184.
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.
92
-------
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-
GTR-229.
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
50:104-108.
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-
174.
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.
93
-------
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.
SEA/972370004
94
-------
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.
Introduction
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.
95
-------
Eelgrass Impacts
Methods
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
distribution.
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.
Irradiance
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
terminal.
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
96
-------
z
r-.
<0
,....
a>
,...:
"
Ii)
,....
a>
,...:
"
,...:
"
,...:
"$~-
122.3490° W
97
-------
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
day.
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.
Results
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
distribution.
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.
Irradiance
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
habitats.
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
98
-------
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).
99
-------
u
z:
<
»—i
Q
1200-
1100-
1000-
900-
800-
700-
600-
500-
400-
300-
200-
100-
0
DOCKSTRUCTURE-
Photosynthesis
Saturated
-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.
_J
J
s
-1-
-3-
-4-
-6
o
o
co
°°
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.
100
-------
Table 1. Taxa of fish, macroinvertebrates and macrophytes observed at the Clinton Terminal in late
summer 1994.
COMMON NAME SCIENTIFIC NAME
IN OR NEAR
EELGRASS
UNDER
DOCK
Invertebrates
barnacle
bay mussel
brooding anemone
coon-striped shrimp
Dungeness crab
heart cockle
helmet crab
horse clam
kelp crab
leather star
nudibranch
nudibranch
plumose anemone
purple star
red rock crab
snail (chink shell)
sunflower star
Benthic Plants
diatoms
eelgrass
gracilaria
sea lettuce
sugar wrack
Vertebrates
copper rockfish
crescent gunnel
flatfish
kelp greenling
penpoint gunnel
pile perch
saddleback gunnel
sanddab
sculpins
shiner perch
striped perch
tubesnout
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
Hexagrammos
decagrammus
Apodichthys flavidus
Rhacochilus vacca
Pholis ornata
Citharichthys spp.
various unidentified species
Cymatogaster aggregata
Embiotocca lateralis
Aulorhynchus flavidus
101
-------
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).
Discussion
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
102
-------
&
£
I
1
180-
160-
140-
120-
100-
80-
60-
Af\
20-
n-
departing
V
8
f
o
o
0
o
o
o
fl
o
o
approaching
O _*r^
\°.
O
\^ 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
departing
0 °5
O ^Siy
1020 1030 1040 1050
110011101120113011401150
120012
TIME (HRS)
Figure 5. Current speeds in propeller wash plume. Current meter was positioned in
Area F (see Figure 2).
250'
Ed
U
200H
150-j
50
900
950
—I—
1000
1050
1100 1150
1200
1250
TIME (HRS)
Figure 6. Bottom irradiance at Area F during ferry docking and departures on
October 3,1994 (see Figure 2).
103
-------
Reflective Foil
Incident
Irradiance (PAR)
400uM
Figure 7. Average PAR Reflected Off Aluminum Foil and Wood Surface of Dock.
104
-------
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
constructed.
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
105
-------
UNDISTURBED
MEADOW
CONSTRUCTION
MAINTENANCE
& OPERATION
DISTURBANCE DURING
CONSTRUCTION
DOCK IN PLACE
MAINTENANCE
DISTURBANCE
FERRY PROP
DISTURBANCE
IMPACTS
EELGRASS
LOSS&
RETARDED
RECRUITMENT
LOWER PROD.,
EELGRASS LOSS &
RETARDED
RECRUITMENT
MITIGATION
MODIFY
PROCEDURES;
REPLANT
MOVE DOCK
OFFSHORE
LOWER
LIGHT
INCREASED
SESSILE
PREY
LOWER PROD.,
EELGRASS LOSS &
RETARDED
RECRUITMENT
EELGRASS LOSS &
RETARDED
RECRUITMENT
EEGLRASS
LOSS
RED. DOCK WIDTH;
GLASS BLOCKS;
GRATE;
FEWER PILINGS;
REFLECTIVE
MATERIAL;
PLANT ADJACENT
AREAS
FEWER PILINGS;
REPLANT
REPLANT
(OVERPLANT)
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.
Acknowledgments
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
Washington.
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-
390.
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.
sea7972260009.doc
107
-------
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.
Introduction
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.
108
-------
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
alternative.
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.
109
-------
- .;;:::;Corte
Ecological
Muzzi Mars
PACIFIC
OCEAN
:i;::::OYv-:v;':^$iin Mated
SOUTH SAN .
FRANCISCO
BAY
£$Sa n •:Lor"efrzo'WillOUK*; •1!lllft:-pte:;:v''llfe":
\ .••r;:^> jjt;|f (I^I'vll^'^^^lMMSI
Ha'y^itfM^ililv.-^^SfS^-islii
- - --:•'-:::'•>.•:•:-• '• ••••'•••••• •:'•:'•".- i/.-y-v.'.-'.-.•'.•'.•'•'.•'.:•::• -,•:•:• x•:•••:-••:-: ••••:• ,vX;t•;,,.•;.;.;
iiin6::i;MaisHlilP;|i^^;:;j|l?;:^
raberxTract
Palo Alto "^-^£^ Coyote Cr'eek y
Figure 1. Location of tidal marshes surveyed in this study.
110
-------
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.
Methods
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.
Sinuosity
Slough channel sinuosity is the ratio of slough
channel length (Lc) to the straight line length (Ld) of
the slough channel drainage system:
S = ^
Ld
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
channels).
Ill
-------
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
centimeter.
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.
Results
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).
112
-------
Selected
Slough System
Selected Slough System
Figure 2. Tidal Slough Channels at Corte Madera Ecological Reserve
(top) and Muzzi Marsh (bottom).
113
-------
West
Corte Madera Ecological Reserve
Longitudinal Profile, 1993
East
£ "•
o ">
IBM #4
Corla Madera
c &
o
.Pay \
> »
i-
0-
200 400 660 800
Distance in feet
1000
1200
1400
Muzzi Marsh, Longitudinal Profile
West
East
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).
114
-------
500 FEET
Selected'N
Slough System
Transect D.te of Photography. 1992
Figure 4. Tidal Slough Channels at Laumeister Tract and Faber Tract.
115
-------
West
-200
Laumeister Tract
Longitudinal Profile, 1993
East
200 400 600 800 1000 1200 1400 1600 1800 2000
Distance in feet
South
O
.3
c
o
'5
3
Faber Tract
Longitudinal Profile, 1993
North
1*
4-
>-
1-
o-
7-
San
Francis
Creek
\
k
\
julto Top of Levee Road
\
Ranae Pole \
Start ot Transect Rebar
\
/ TBM #1 \
/ \ \
/ \
\
' \
"^ •- ^1 ^
Y 1 V-~J.. — 1
1 1 "
1 " \
1 \
- - - . \
: \
\
'
V
-200 6 200 400 600 800 1000 1200 1400 1600 1800
(^
20
Distance in feet
Figure 5. Topographic Profile at Laumeister Tract (top) and Faber Tract
(bottom).
116
-------
Table 1. Morphometric Characteristics of Selected Slough System.
CORTE MADERA ECOLOGICAL RESERVE
Slough Channel Order
1
2
3
4
Number of Sloughs in
Order
19
4
2
1
Length of Sloughs in
Order (mi.)
0.26
0.19
0.06
0.05
Drainage Density of Selected Slough System = 56 mi./mi/
Sinuosity = 1.11
MUZZI MARSH
Slough Channel Order
1
2
3
4
Number of Sloughs in
Order
10
3
1
0
Length of Sloughs in
Order (mi.)
0.22
0.05
0.12
Lower Marsh
• Drainage Density of Selected Slough System = 39 mi./mi.^
• Sinuosity = 1.51
Upper Marsh
• Drainage Density of Entire Marsh = 9 mi./mi.^
LAUMEISTER TRACT
Slough Channel Order
1
2
3
4
Number of Sloughs in
Order
13
5
1
0
Length of Sloughs in
Order (mi.)
0.42
0.28
0.22
o
Drainage Density of Selected Slough !
Sinuosity = 1.21
FABER TRACT
Slough Channel Order
1
2
3
4
Number of Sloughs in
Order
33
9
2
1
Length of Sloughs in
Order (mi.)
0.56
0.34
0.28
0.04
Drainage Density of Selected Slough System = 41 mi./mi/
Sinuosity =1.10
117
-------
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
118
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.
virginica.
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.
Discussion
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).
MARSH
CMER
No. of quadrats sampled
Mean
Range
Muzzi Marsh
No. of quadrats sampled
Mean
Range
Laumeister Tract
No. of quadrats sampled
Mean
Range
Fa her Tract
No. of quadrats sampled
Mean
Range
SPFO
39
0
0
41
20.55
0-87.5
38
22.17
0-67.5
42
9.29
0-67.5
SAVI
39
47.2
0-87.5
41
60.37
0-87.5
38
37.83
0-87.5
42
67.14
2.5-87.5
Table 3. Stem Length, Total Stem Length and Density for Spartina foliosa.
Parameter
Stem Length (cm)
Number of stems
Mean
Range
Total Stem Length (cm)
Number of quadrats
Mean
Range
Density (/m^)
Number of quadrats
Mean
Range
Muzzi Marsh
123
68.9
8-118
8
1059.2
319-1626
8
153.8
50-250
Laumeister Tract
324
67.9
14-98
21
1047.4
271-1986
21
154.3
40-280
Faber Tract
181
61.7
5-102
9
1240.6
141-2936
9
201.1
50-460
119
-------
100
80
g 60
3
cr
I 40
3
E
3
u
20
muzzi marsh
fabcr tract
laumcistcr
60 90
height interval (cm)
120
Figure 6. Cumulative Frequency for Sparlina foliosa Stem Length at
Muzzi Marsh, Faber Tract and Laumeister Tract.
120
-------
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
121
-------
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
vegetation.
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.
Conclusion
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.
Acknowledgements
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.
122
-------
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:
109-126.
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
District.
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
District.
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-
83/23.
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-
231.
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.
123
-------
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-
38-56.
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-
125.
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.
sea971820026.doc
124
-------
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.
Introduction
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.
125
-------
N>
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).
Sampling
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.
Results
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
127
-------
San Francisco Bay Salt Marshes
Table 1. Biogeographic variables calculated for the 83 study sites.
Biogeographic Variable Description
Shape Factor
Cover Type
Tidal Regime
Area
Perimeter
Natural Perimeter
Unnatural Perimeter
Aquatic Perimeter
Channelization
Isolation
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)
Mean/Mode
0.42
3.0
3.0
0.34
2.96
4.06
62.49
32.87
17.3
0.35
SE
0.22
0.09
0.1
0.67
0.28
1.33
3.55
3.54
2.04
0.05
r Value
-0.441
0.729
0.86
0.3
0.383
-0.108
-0.616
0.664
0.702
-0.254
p Value
<0.0001
<0.0001
<0.0001
0.0059
<0.0001
0.3308
<0.0001
<0.0001
<0.0001
0.0206
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
128
-------
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.
Discussion
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.
129
-------
REGIONAL RESTORATION PLANNING FOR
SAN FRANCISCO BAY SALT MARSHES
POTENTIAL SPECIES DIVERSITY INDICES
Redwood City
East
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
marshes.
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.
131
-------
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.
Conclusions
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
132
-------
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
133
-------
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-
703.
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
2:75-92.
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
134
-------
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-
1030.
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
8:1087-1097.
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
35:68-76.
SEA/972500001
135
-------
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
projects.
Overview
"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
community.
Using a readily available resource, reclaimed
wastewater, the town integrated these benefits into
construction and operation of groundwater recharge
ponds.
Jones & Stokes Associates, 2600 V Street, Suite 100, Sacramento, California 95818-1914.
136
-------
Desert Riparian Habitat
Groundwater Conservation Efforts and
Benefits
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
exhibits.
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).
137
-------
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
Methods
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
facilities.
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.
138
-------
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
Participation
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
139
-------
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.
Conclusions
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,
CA.
sea971820017.doc
140
-------
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
Ornithologist
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.
Introduction
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-
naire.
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.
141
-------
Design and Costs
• Enhancement of fish and wildlife habitat
• Estuarine wetland restoration
• Estuarine wetland restoration and wildlife
enhancement
• 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
hectare.
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
U.S.
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
these?
6. Who paid for this program?
One of the biggest problems that we encountered
was finding the right person with whom to discuss
142
-------
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
follows:
• 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
reported.
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
Project
Component Quantity
Convert oil 3
canals to marsh
Fertilizing
Monitoring
Engineering
design
Total
Unit
ea
LS
LS
LS
Equipment,
Materials,
Supplies, etc.
$76,059
$3,370
$4,017
$42,449
$125,895
Labor
$44,670
$3,564
$2,074
$50,308
Total Costs
($1994)
$126,236
$7,250
$6,369
$44,385
$184,240
Total Costs
($1992)
$120,729
$6934
$6,091
$42,449
$176,204
143
-------
Design and Costs
Table 2. Cost by Component for Palo Alto Harbor marsh construction.
Project Component
Mobilization
Demolition (includes
launch ramp)
Clearing and grubbing
Earthwork
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
handrail
Retaining wall
Bollards
Cable
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
grass)
Marsh plants
(gumweed)
Disking
Hydroseeding
Siltfence
Maintenance
Design
Total
Quantity
49,422
49,422
14,540
1,320
130
140
124
98
1,520
50
50
4
1
3
7
6
7
102
370
903
903
1,590
392
17,327
0.23
1,990
Unit
LS
LS
LS
cy
cy
sqft
sqft
If
If
If
ea
If
If
If
ea
ea
ea
ea
ea
ea
LS
cy
cy
ea
ea
ea
ea
sqyd
acre
If
LS
LS
Equipment,
Materials,
Supplies, etc.
$11,400
$15,000
$10,000
$197,688
$45,962
$12,504
$6,600
$390
$32,739
$19,518
$5,880
$3,040
$50
$50
$50
$2,000
$2,723
$3,086
$480
$6,523
$5,500
$3,500
$12,765
$1,806
$3,928
$14,310
$1,568
$2,946
$0
$5,970
$1,500
$165,000
$594,476
Total
Costs
Labor ($1994)
N/A $11,920
$15,684
$10,456
$206,706
$48,059
$13,074
$6,901
$408
$34,232
$20,408
$6,148
$3,179
$52
$52
$52
$2,091
$2,847
$3,227
$502
$6,820
$5,751
$3,660
$13,347
$1,888
$4,107
$14,963
$1,640
$3,080
$0
$6,242
$1,568
$172,527
$621,591
Total Costs
($1992)
$11,400
$15,000
$10,000
$197,688
$45,962
$12,504
$6,600
$390
$32,739
$19,518
$5,880
$3,040
$50
$50
$50
$2,000
$2,723
$3,086
$480
$6,523
$5,500
$3,500
$12,765
$1,806
$3,928
$14,310
$1,568
$2,946
$0
$5,970
$1,500
$165,000
$594,476
144
-------
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.
Conclusions
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.
Acknowledgments
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
16:3-8.
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.
sea/972260012.doc
145
-------
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.
Introduction
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
146
-------
~~
.
EVERETT
SEATTLE
. .
BREMERTON
.
ISSAQUAH
. SITE
0-
.TACOMA
v
.OLYMPIA
Jp~~
\ "..;;:<-
''A'Ui,-"ake
L/\r!
'0 Youngs
~,
"',
/
/.
:1"'.0
'--...-
'"
,
~
,
~~
~" '
~T~GOM",A
.~
V
\
~~
'-. "'.
(5 '1 ~"
. '-!
BLAGK.0
DIAMOND
S"1
/1::
o I ~
~.\ ;
'v' \
/ .
. I
'-.,
FEDERAL
WAY
"'--.
\
)
\
)
}
. /' --,'
~
~,'" ""'" <".'~"."""";?'C+/, ,
~ f'h"r <)VI'
"\~',, ';"/~'
\"'~~~~~,""',,',,.',,'. r".'. FS'i
~A!-Ta!5e.J N
''& ,apps
0-\~C<.,
\ \ -,
~
s
*
0 7,500 15,000
Scale in Feet
Figure 1. Site Location Map.
~~IT SHAPIRO
- &. ASSOCIATES. INC
---
147
-------
WETLAND GRASSES.
RUSHES. FORBS
UPLAND GRASSES, FORBS
DMA RIGHT-OF-WAY
WETLAND GRASSES.
RUSHES. FORBS
UPLAND GRASSES, FORBS
LOU BARDY POPLARS
HORSE CHESTNUTS
UPLAND GRASSES,
FORBS, FRUtT TREES
UNVEGETATED
FILL
BLACKBERRY,
RED-OSIER DOGWOOD
_WETLAND GRASSES,
RUSHES, FORBS
UNFILLED SETBACK
SEWER EASEMENT
UPLAND GRASS
Figure 2. Existing Vegetation Racetrack Site.
Em«u«ntWefands
Soub-Shrub/Open Water Wetlands
ScrutKShnjbWedands
HI
-------
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
Basin.
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
149
-------
I/I
o
RACING, INC.
AUBURN RACE TRACK
CONCEPTUAL HABITAT MITIGATION PLAN
Figure 3. First Mitigation Proposal (1992).
—RA.
RAEDEKE ASSOCIATES. INC.
§3"OIT
-------
AUBURN THOROUGHBRED RACETRACK
OFFStreWeTLAHOUirtQATIOHPLAN ' ''-"\
Figure 4. Second Mitigation Proposal.
ID ronuno OCIUMO njamnt -
n«.
-------
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
banks.
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
152
-------
Figure 6. Fourth Mitigation Proposal.
Figure 7. Final Mitigation Proposal.
153
-------
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
channel.
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
Summary
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,
Communication
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.
Consistency
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.
154
-------
Theory to Reality
Collaboration
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.
Consensus
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
emphasis.
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
appendices.
sea972270005.doc
155
-------
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.
Introduction
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.
156
-------
Jetty Island, Washington
MARYSV1LLE
0 1800 3800 7600
••g
SCAIE ( ft )
POSSESSION SOUND
Figure 1. Location and Vicinity Map
157
-------
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.
Methods
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
158
-------
Jetty Island, Washington
FEDERAL NAVIGATION CHANNEL
FISH/INVERT
REFERENCE
AREA
POINT
JETTY
ISLAND,^
PRESTON
POINT
MAULSBY
SWAMP
DONOR/REFERENCE
MARSH
PROJECT AREA
1SH/INVER'
REFERENCE
LEGEHS
2^2DSALTMARSH
^JVVVVl MUD FLAT
BRACKISH MARSH
Figure 2. Project Area
159
-------
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
noted:
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
quadrat.
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.
Results
Physical
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.
Biological
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
160
-------
"!'j
~.
=
..,
1'0
Vol
1.0
1.0
VI
1:0'
(1)
..,
~
'0
::r
o
o
(Jq
..,
1:0'
'0
::r
o
-.,
'1\
!,"
L: :
o A .~
" "') ~.,
""<~.
0'<
-------
Table 1. Mean key prey biomass and the prey weighing factor (PFW) from epibenthic sampling in
Port Gardner Bay, 1995.
Biomass
(grams/square meter)
Month
April
May
Elevation
(ftMLLW)
+9
+6 or +7
+3
+9
+6 or -1-7
+3
Reference
area
0.0006
0.0003
0.0050
0.0011
0.0000
0.0139
Project area
Inside
0.1 161'
0.0000
—
0.0350*
0.0875
—
Exposed
0.0000
0.0006
0.0014
0.0000
0.0000
0.0003
Prey Weighing Factor
Reference
area (= before)
1.00
1.00
1.00
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)
category.
162
-------
Jetty Island, Washingtc
area. Key prey taxa were also not taken from +6 ft
MLLW in the exposed project area or reference
areas.
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
conditions.
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
success.
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.
Vegetation
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
163
-------
Table 2. Pre- and post project habitat area (acres) and quality (habitat units).*
Mud
Sand
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
0.0
0.0
0.0
0.0
0.0
(loss)
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
Change
2.6
4.0
6.5
-26.3
-13.2
(acres)
Habitat Units
Before
1.5
2.4
12.1
23.8
39.9
-16.6%
After Change
140 138
305 302
1528 1516
2 -22
1975
1935
4853%
*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.
Assumptions
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.
3
>>
111
Reference Berm
Inner
Taxon
Jaumea carnosa
Salicornia virginica
Distichlis spicata***
Atriplex patula
Elymus mollis
Total cover**
Mean
21.45
8.18
12.38
5.45
17.73
68.20
SD
18.91
10.42
8.77
6.90
18.24
25.27
Mean
25.19
9.69
32.75
3.47
23.50
97.41
SD
31.59
15.19
23.57
4.23
31.59
35.86
P
0.6699
0.7380
0.0003
0.3607
0.414
0.0078
Point
Mean
0.12
18.20
1.46
0.56
—
31.84
SD
0.44
22.18
3.80
1.96
...
21.81
Unplanted Area
03
fin*
o Jaumea carnosa*
= Salicornia virginica
Distichlis spicata**
Q) P
5 3 Total cover
0 «
*p < 0.05
**p < 0.01
*"p < 0.001
69.35
24.15
0.10
11.60
109.40
18.06
15.61
0.31
6.09
11.39
83.75
13.38
0.63
5.31
87.25
16.75
24.69
0.96
3.14
33.23
0.0170
0.1170
0.0269
0.0011
0.4101
Mean
11.13
26.25
5.25
6.38
49.88
SD
17.03
22.70
8.51
9.40
29.42
165
-------
120
LJ Jmimea carnosa
Distichlisspicala •••
Elymus moll a
Salicornia virginlca
[U Atriplex patula
Spergularia marina
10
Reference-Upper Benn-Upper
Area
Figure 4. Key species present in the Distichlis/Elymus zone, 1995.
Pokit
Upper Lower
Reference
Upper Lower
Planted Area
Point Cove
Figure 5. Mean total vegetative cover (%) in late summer 1993 and 1995.
166
-------
Jaumea carnosa •
Sallcornla vlrglnlca
Diiilchlis spicata ••
Spergularla marina *
S
8
Reference-Lower
Bonn-Lower
Area
Unpbnted Beim-Lower
Figure 6. Key species present in the Jaumea/Salicornia zone, 1995.
4T
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.
167
-------
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.
Discussion
Physical
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.
Biological
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
168
-------
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
berm.
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.
Summary
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
MLLW.
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.
169
-------
Jeiiy Island, Washinglon
Acknowledgments
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.
sea/972260004.doc
170
-------
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.
171
-------
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
172
-------
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
Oregon.
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.
sea971820018.doc
173
-------
,".
..
"
..., .,.
'~
... .
......
'"
,
l',~'
"f
"'"
]74
""-.
~,
"
~
I~."
~
-------
Figure 3.
175
-------
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
Introduction
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.
176
-------
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.
177
-------
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.
178
-------
Percent Cover
uo-
120-
100-
80-
60-
40
20
78
] AQAL
3 DISP
79
80
Q POPA
HH HOLA
82
Year
84
CALY
TRRE
88
93
^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
CALY
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.
179
-------
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
surface.
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.
Discussion
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
dominate.
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
180
-------
Elevation (m NGVD)
1.6 4
Mitchell Marsh
0.51
Control
Restored
Control
Restored
Figure 4. Elevation distribution 1988 of the restoration site and controls in Mitchell Marsh and
Y-Marsh in the Salmon River estuary.
1.20
LOOt
0.80 h
0.60 h
0.40 h
0.20 t-
0.00
Elevation (m)
Creek 1
Cross-section 4
234
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.
181
-------
Salmon River Salt Marsh
species have begun to be replaced by common re-
gional salt marsh species.
Conclusions
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
established.
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.
Acknowledgements
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-
135.
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.
182
-------
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.
sea971820020.doc
183
-------
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.
Introduction
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.
184
-------
"$: BATIQUITOS LAGOON
~.=::I~ ENHANCEMENT PROJECT
..
Poinsenia lone
Interpretive
Center
t. '.
Qc .
OSI.
rh Q 4"el1.
'1T lie
Figure 1. Diagrammatic illustration of the Batiquitos Lagoon Enhancement Project.
LEGEND
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
-
DO
VI
-------
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
wildlife.
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-
ment.
• 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-
goon.
• 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
Construction
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
186
-------
Table 1. Approximate acreage of habitats to the enhanced by the project.
Habitat type
Open nontidal water/seasonal mudflats
Subtidal(-1.6ftMLLW)
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
Existing
415
0
0
111
0
67
593
After
construction
0
148
144
174
38
94
593
100
California least terns
Adult Pairs Fledglings
i—i—r~r—i—i—r—i—r
1978 1979 1960 1981 1982 1983 1984 1985 1986 1987 1988
Years
i—i—\—r
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.
187
-------
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.
Discussion
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
project.
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,
188
-------
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
CA.
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,
CA.
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,
CA.
sea972260018.doc
189
-------
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
initiation.
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.
190
-------
0
8
NATURAL MARSH
PLANTED CONTROL
UNPLANTED CONTROL
ELAPSED TIME
(months)
—£— STRAW
....>... ALFALFA
PEAT
INORGANIC N ADDITIONS
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
unknown.
Elevation
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.
Vegetation
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
192
-------
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
densities.
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:
489-500.
972020008.doc
194
-------
The Sum Exceedance Value as a Measure
of Wetland Vegetation Hydrologic
Tolerance
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.
Introduction
Background
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.
195
-------
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.
1992).
Purpose
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
196
-------
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
site.
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).
Methods
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
197
-------
Ground Elevation
Ground Elevation - 30 cm
I I I I I I I I I I
JFMAMJJASONDJ
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
99.00
1 I
s s
s £ S i 5 S
date
.Deep Marsh Well (gmd. elev. = 98.366 m)
99.00
Grass Lake Marsh Water Elevations 1990-1994
97.00
I I
date
. Shallow Marsh Plot (grnd etev. = 97.540 m)
Figure 2. West Chicago Prairie and Grass Lake Marsh hydrographs from water level
data.
198
-------
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
199
-------
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
(1985).
Table 1. Cover classes, cover class ranges, and range
midpoints for vegetation plots.
Cover Class
1
2
3
4
5
6
7
Cover Class
Range
0-1%
1-5%
5-25%
25-50%
50-75%
75-95%
95-100%
Range Midpoint
0.5%
3.0%
15%
37.5%
62.5%
85%
97.5%
Surveying
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
Ranges
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
200
-------
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
range.
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.
Hydropatterns
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.
201
-------
Average Water Depths and SEVio 's for Different Communities at
Wadsworth Prairie -1995
elevation
date
-WPSEV=14
• SM SEV=36
- SMA SEV=56
- DMA SEV*94
0.40
Average Water Depths and SEV M 's for Different Communities at West
Chicago Prairie-1994
-0.80
data
-WPSEV=28
- SM SEV= 45
-SMASEV=77
- DMA SEV= 88
Average Water Depths and SEVto 's for Different
Communities at West Chicago Prairie -1995
0.60
date
E
-WPSEV=.19
• SM SEV= 57
- SMA SEV= 83
•DMASEV=106
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
1995.
202
-------
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
gradient.
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
203
-------
15
95% Confidence Intervals
20
40
60
80
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
SEV(30)
too
Preferred • Tolerated O Avoided
Figure 5. Sum exceedance value ranges for selected species at Grass Lake Marsh 1994.
204
-------
Helianthus grossoserratus Martens.
Veronicastrumvtrginicum(L.) Farw
Solidago canadensis L
Viola pratincola Greene
Pycnanthemum
vfrginianum (L) Dur & Jacks
Euthamia grammifolia (L) Sahsb.
Lysimachia quadrifloraStms
Carex stricta Lam.
Calamagrostis
eanadens/s (Michx.) Beauv
Apocynum sibmcum Jacq.
. PofygonumamphibiumL.
. LycopusamencanusMuh\.
LycopusvirgmicusL
: Galium obtusum Bigel.
Asfe/s/mp/exWilld
Stachys palustns L
^ Scutellaria galericulata L.
Typhalatifolial.
Ludwigia palustns L
, ScirpusacutusMuM
SEV(30)
Preferred
Tolerated
Avoided
Figure 6. Sum exceedance value ranges for selected species at West Chicago Prairie
1994.
205
-------
Agrostis alba L.
Melilotus alba Medic.
Oxypolis rigidior (L.) Coulter and Rose
Achillea millefolium L
Poa pratensis L
Pycnanthemum
virgintenum (L) Our & Jades
Veron/castrvm virgin/cum (L) Fatw. ,
Viola pratincola Greene.:'i . • '..
Hellanthus grosseserratus Martens. :
SolUagocanadensisL.
Poa compressa L ::;•/
Lysimactiia quadrifotta Sims: ,:
ScutellartagalericulataL. :^
Lycopus amaricanus Muhl.
Lycopusvirginicusl. : ;; •
Calamagmstis canadensis (Michx.) Beai
Polygonumamptiibium L
TyphalaafoliaL
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
1995.
206
-------
Figure 8. Sum exceedance value ranges for selected species at Wadsworth Prairie 1995
207
-------
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.
Conclusions
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.
Acknowledgments
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.
208
-------
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,
England.
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,
England.
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
4(2):99-116.
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
109(2):367-375.
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
9(2):9-10.
209
-------
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-
1255.
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-
114.
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-
1424.
Spence, D. H. N. 1982. The zonation of plants in
freshwater lakes. Advances in Ecological Research.
12:37-125.
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.
sea971820015.doc
210
-------
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.
Introduction
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.
211
-------
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
calculations.
Table 1. Cover classes used in herbaceous
vegetation sampling (Daubenmire 1959).
Cover Class
6
5
4
3
2
1
+
Range of
Coverage (%)
95-100
75-95
50-75
25-50
5-25
1-5
< 1
Mid-point of
Range (%)
97.5
85.0
62.5
37.5
15.0
3.0
0.5
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
Establishment
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
212
-------
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
Prevalence
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
213
-------
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
1991
8.82
4.73
23.91
1.34
22.57
7.48
4.41
22.51
18.06
2.15
3.27
0.16
0.91
10.75
0.42
0.25
% of Species % of Species Native to
Hydrophytic1 central Illinois'
72
65
93
82
64
plant species
1992
0.00
22.46
29.79
6.59
23.20
4.32
2.15
19.56
17.83
0.96
0.00
0.12
9.67
2.09
0.00
0.12
at the restored
1993
6.33
0.00
12.67
3.28
8.91
6.26
0.57
41.99
41.34
1.04
11.73
1.24
0.88
0.00
2.69
4.50
73
65
83
82
70
wetland,
1994
12.07
0.67
21.49
6.03
15.33
2.97
1.58
27.88
25.08
1.66
17.58
0.44
1.76
0.77
4.76
3.27
1991-1995.
1995
0.60
5.56
24.84
13.73
11.11
5.06
1.03
25.03
22.47
3.35
7.93
2.41
8.42
0.82
0.35
4.32
214
-------
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
observed.
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.
215
-------
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
wetland.
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.
81/33.
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
216
-------
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
66:259-265.
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,
IL.
Sea/971890015
217
-------
Illinois Keltic Marsh Restoration
oo
Appendix A. Top importance values for vegetation at the restored wetland, 1991-1995.
1991
1992
1993
1994
1995
Bidens tripartite
(22.57)
Eleocharis palustris
(18.06)
Rorippa islandica
(10.75)
Alisma plantago-aquatica
(8.82)
Ambrosia artemisiifolia
(4.73)
Echinochloa crus-galli
(4.41)
Carex frankii
(3.55)
Lemna minor
(3.27)
Carex tribuloides
(2.70)
Eleocharis obtusa
(2.54)
Bidens tripartita
(23.20)
Ambrosia artemisiifolia
(22.46)
Eleocharis palustris
(17.83)
Bidens aristosa
(6.59)
Polygonum punctatum
(5.86)
Polygonum pensylvanicum
(3.58)
Carex tribuloides
(2.24)
Echinochloa crus-galli
(2.15)
Rorippa islandica
(2.09)
Carex frankii
(1.54)
Eleocharis palustris
(41.34)
Lemna minor
(11.73)
Bidens tripartita
(8.91)
Alisma plantago-aquatica
(6.33)
Typha latifolia
(4.50)
Carex frankii
(4.25)
Bidens aristosa
(3.28)
Sagittaria latifolia
(2.69)
Bromus inermis
(1.55)
Juncus acuminatus
(1.55)
Eleocharis palustris
(25.08)
Lemna minor
(17.58)
Bidens tripartita
(15.33)
Alisma plantago-aquatica
(12.07)
Bidens aristosa
(6.03)
Sagittaria latifolia
(4.76)
Typha latifolia
(2.63)
Amaranthus tuberculatus
(2.21)
Carex frankii
(2.18)
Eleocharis acicularis
(1.96)
Eleocharis palustris
(22.47)
Bidens aristosa
(13.73)
Bidens tripartita
(11.11)
Lemna minor
(7.93)
Ambrosia artemisiifolia
(5.56)
Leersia oryzoides
(3.35)
Typha latifolia
(3.21)
Polygonum punctatum
(2.71)
Eleocharis acicularis
(2.55)
Phalaris arundinacea
(2.41)
-------
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
velvetleaf
three-seeded mercury
three-seeded mercury
silver maple
quackgrass
water plantain
tamarisk waterhemp
pigweed
waterhemp
common ragweed
swamp milkweed
hairy aster
panicled aster
swamp marigold
beggar-ticks
beggar-ticks
beggar-ticks
awnless brome grass
sedge
sedge
sedge
sedge
fox sedge
scaly-bark hickory
milk spurge
spurge
lamb's quarters
Canada thistle
thistle
bull thistle
horseweed
dogwood
red haw
Hydrophytic
No
-
No
Yes
No
Yes
Yes
~
Yes
No
Yes
No
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
-
Yes
Yes
No
No
--
No
No
-
No
No
Yes
Native to central Illinois2
No
Yes
Yes
Yes
No
Yes
Yes
-
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
—
Yes
Yes
Yes
Yes
—
No
No
"
No
Yes
Yes
Yes
219
-------
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
Queen-Anne's-lace
panic grass
barnyard grass
yerba de tajo
needle spikerush
spikerush
spikerush
spikerush
fireweed
annual fleabane
fleabane
tall boneset
wild strawberry
Peruvian daisy
white avens
rough avens
ground ivy
honey locust
jewelweed
rush
path rush
lettuce
wild lettuce
rice cutgrass
duckweed
false pimpernel
common water horehound
moneywort
yellow wood sorrel
ditch stonecrop
reed canary grass
fog-fruit
buckhorn
Hydrophytic
No
--
Yes
Yes
Yes
Yes
Yes
Yes
No
No
--
No
No
No
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
--
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
Native to central Illinois
No
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
No
Yes
Yes
Yes
Yes
No
—
Yes
Yes
Yes
Yes
No
Yes
Yes
No
Yes
No
220
-------
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
hydropiperoides
Polygonum lapathifolium
Polygonum pensylvanicum
Polygonum persicaria
Polygonum punctatum
Polygonum ramossissimum
Polvqonum scandens
Common Name
red-stalked plantain
Kentucky bluegrass
water smartweed
smartweed
wild water pepper
pale smartweed
giant smartweed
spotted lady's thumb
dotted smartweed
bushy knotweed
climbing buckwheat
Hydrophytic'
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Native to central Illinois2
Yes
No
Yes
No
Yes
Yes
Yes
No
Yes
Yes
Yes
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
smartweed
purslane
American pondweed
drooping coneflower
marsh yellow cress
pale dock
curly dock
grass-leaved arrowhead
arrowhead
sandbar willow
hardstem bulrush
softstem bulrush
figwort
giant foxtail
yellow foxtail
prickly sida
horse nettle
poison ivy
red clover
white clover
narrowleaf cattail
No
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
No
No
Yes
No
No
Yes
No
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
Yes
Yes
No
No
No
221
-------
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
cattail
blue vervain
ironweed
speedwell
common blue violet
violet
grape
Hydrophytic
Yes
Yes
--
--
Yes
--
Yes
Native to central Illinois
Yes
Yes
Yes
-
Yes
--
Yes
Reed (1988).
Vaftetal. (1993).
222
-------
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.
Introduction
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
1992).
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.
223
-------
South Carolina Coastal Plain Wetlands
Materials and Methods
Sites
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).
Measurements
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
individual.
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
224
-------
Table 1. Species Codes and Relative Flood Tolerance of the Twelve Study Species (McKnight et al.
1980).
Aperies
Species Code
Blackgum
Cherrybark Oak
Dogwood
Hackberry
Mockernut Hickory
Overcup Oak
Persimmon
Swamp Chestnut Oak
Southern Red Oak
Swamp Tupelo
Water Hickory
White Oak
Table 2. Non-Sheltered
Classes.
BG
CO
D
H
MH
OC
P
SC
SR
ST
WH
WO
Percent Survival
Terrace WD
Moderately Tolerant
Weakly Tolerant to Intolerant
Intolerant
Moderately Tolerant to Intolerant
Intolerant
Moderately Tolerant
Moderately Tolerant
Weakly Tolerant
Intolerant
Tolerant
Moderately Tolerant
Moderately Tolerant to Intolerant
Species Performance Across All Terraces and
MWD SPD
Drainage
PD
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
Classes.
Terrace
WD
MWD
SPD
PD
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
225
-------
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
Carolina.
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.
Results
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.
226
-------
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
227
-------
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
class.
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
classes.
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-
races.
Discussion
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.
SPD
Terrace
WD
MWD
SRS
WWC
SEF
Overall
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,
wo
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
Terrace
WD MWD
bg,co,h,oc,p,sc,sr, bg,co,h,oc,p,sc,sr.st,
st,wo wo
co,h,oc,p,sc,sr,
st,wo
co,h,oc,p,sc,sr,
st,wo
co,h,oc,p,sc,sr,st,
wo
All Terraces and
SPD
bg,co,mh,oc,p,sc,w
o
bg,co,oc,p,st
bg,co,oc,p,sc,st, wo
Drainage Classes.
PD
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,
sc,sr,st,wh,wo
Overall bg,co,d,h,oc,p,sc, bg,co,d,h,mh,oc,p,
sr,st,wh,wo sc,sr,st,wh,wo
bg,co,h,oc,p,sc,sr,
st,wh,wo
bg,co,d,h,mh,oc,p,
sc,sr,wh,wo
bg,co,d,mh,oc,p,
sc,sr,st,wh
bg,co,d,h,mh,oc,p,
sc,sr,st,wh
bg,co,h,oc,p,sc,sr,st,
wh,wo
bg,co,d,h,mh,oc,p,sc,
sr,st,wh,wo
bg,co,mh,oc,p,sc,sr,
st,wo
bg,co,oc,p,sc,sr,st,
wh,wo
Table 6. Best Species Performance in Each Drainage Class
Class
Non-Sheltered
Survival
Sheltered %
Survival
Non-Sheltered
Total Height
WD
% co,oc,p,sc,sr,wo
bg,co,d,mh,oc,
p,sc,sr,wh,wo
bg,co,d,h,oc,p,
sc,sr,wo
Sheltered Total bg,co,d,h,oc,p,
Height
Table 7. Suite
South Carolina
Class
of Species Exhibiting the
L Coastal Plain.
WD
MWD
co,oc,sc,wo
bg,co,oc,p,sc,sr,st,
wo
bg,co,oc,p,sc,sr,st,
wh,wo
bg,co,d,h,mh,oc,
p,sc,sr,st,wh,wo
Best Performance
MWD
from the Four Classification Categories.
SPD
bg,co,oc,sc
bg,co,oc,p,sc,st
wo
co,h,oc,p,sc,sr,
st,wo
bg,co,d,h,mh,
oc,p,sc,sr,st,wh
Across All Study
SPD
PD
oc
bg,co,oc,p,sc,st
bg,co,oc,p,sc,st,wo
bg,co,oc,p,sc,
sr,st,wh,wo
Drainage Classes in the
PD
229
-------
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.
230
-------
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,
GA.
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,
FRG.
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.
sea/972260019.doc
231
-------
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.
Introduction
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,
1992).
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).
Methods
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
233
-------
Figure 1. Buffalo River drainage basin.
234
-------
Figure 2. The lower Buffalo River.
235
-------
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
analysis.
Analyses
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
organisms
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):
H'=-I(PiInPi)
i= 1
K =
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
1981
relative percentage
of species i from
1982
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
236
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
-------
°f
* , acc,
York (modified from Karr et al. 1986).
iC 'ntegrity metrics for the Buff*l° River. New
Category
Metric
Scoring criteria
5 3 1
Species richness and
composition
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
A.
<5
5- 20
>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
CPUE*
11. Percent hybrids
Expectations for metric 10 vary with
stream order. See Appendix A.
>0- 1
>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).
237
-------
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
Lepisosteidae
Leoisosteus osseus
Clupeidae
Alosa oseudoharenous
Dorosoma ceoedianum
Cyprinidae
Carassius auratus
Cyprinus carpjo
(Hybrid)
Notemioonus crvsoleucas
Nocomis micropooon
Cyprinella spiloptera
Notropis hudsonius
Notropis atherinoides
Pimephales notatus
Notroois cornutus
Semotilus atromaculatus
Catostomidae
Carpiodes cvprinus
Catostomus commersoni
Hypentelium nigricans
Moxostoma sp.
Ictaluridae
Ameiurus nebulosus
Ameiurus melas
Ictalurus punctatus
Noturus flavus
Esocidae
Esox lucius
Esox masouinonqy
Osmeridae
Osmerus mordax
Salmonidae
Oncorhvnchus mvkiss
Q. kisutch
0. tshawvtscha
Salmo trutta
Common Name
longnose gar
alewife
gizzard shad
goldfish
carp
carp x goldfish
golden shiner
river chub
spotfin shiner
spottail shiner
emerald shiner
bluntnose minnow
common shirfer
creek chub
quillback
white sucker
northern hog sucker
redhorse spp.
brown bullhead
black bullhead
channel catfish
stonecat
northern pike
muskellunge
rainbow smelt
rainbow trout
coho salmon
Chinook salmon
brown trout
Taxa
-
*
-
Y*
Y*
Y*
Y
1
Y
1
Y
I
Y
1
Y
Y
A
A
A
A
_
.
.
-
.
-
*
*
^
Jt
Trophic
Guild
F
P
P
0
0
0
P
M
M
0
P
O
0
0
M
M
0
M
0
0
0
M
F
F
P
M
P
1
F
F
Intolerant
Species
-
-
-
-
-
-
1
1
-
1
_
.
1
.
_
1
.
.
1
1
.
I
i
I
i
I
i
1
Percopsidae
Percopsis omiscomavcus
trout-perch
M
238
-------
Percichthyidae
Morone americana
Morone chrvsops
Centrarchidae
Ambloplites ruoestris
Leoomis gibbosus
Leoomis macrochirus
Micropterus dolomieu
Micropterus salmoides
Pomoxis annularis
Pomoxis nioromaculatus
Leoomis oulosus
Percidae
Etheostoma niqrum
Perca flavescens
Percina caprodes
Stizostedion vitreum
Sciaenidae
Aolodinotus orunniens
white perch
white bass
rock bass
pumpkinseed
bluegill
smallmouth bass
largemouth bass
white crappie
black crappie
warmouth
johnny darter
yellow perch
logperch
walleye
freshwater drum
E
E
E
E
E
E
D
D
M
M
M
M
M
F
F
P
P
M
M
M
M
F
M
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
Attributes
47-55
Excellent
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
29-37
20-28
11 -19
Good
Fair
Poor
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.
239
-------
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.
Results
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.
Discussion
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
240
-------
2.50 T
0.00
D1981
H1992
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 -
37.8
: :•:::;:
III
1
62.
:
jj
2
4
Reach
34.1
i| j!;
3
38.:
:
4
3
Figure 4. Calculated Percent Similarity Index (PSC) at each reach between 1981 and 1992.
37
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.)
241
-------
123456789 10 11
123456789 10 11
123458789 10 11
Figure 6. Individual metric scores at each reach for 1981 and 1992.
242
-------
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
(62.4).
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
score.
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
River.
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.
243
-------
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
244
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.
26pp.
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.
26pp.
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
pp.
Sweeney, R. A. 1972. River on the mend. Limnos
5(2).
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-
694.
Werner, R. G. 1980. Freshwater fishes of New York
State: a field guide. Syracuse University Press.
Syracuse, NY. 186pp.
sea971810011.doc
245
-------
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
-------
Q.
CO
TS
•H
x:
o
Cfl
c
CD
o
c_
CD
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).
CO
•H
O
-------
Figure A5. Total number of intolerant species vs.
stream order in the Buffalo River Watershed (Metric 5).
CO
14
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).
10
c
•H
CA
•H
U_
C_
-------
APPENDIX B
Water chemistry monitoring data, 1970 to 1993.
-------
1.
E
Dissolved O2 for May
12 -r
10-
8
6 -
V
4-
2-
* 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
Year
— i — i
93 95
Dissolved O2 for June
10 -J
8 -
6 -
E 4 -
2 -
n
^ A
t
1
" *
65 67 69 71 73 75 77 79 81 83 85 87 89 91 93 95
Year
Dissolved O2 for August
5 j
4 -
3 -
* 2-
1 -
0 •
I *
1
1 *
*
65 67 69 71 73 75 77 79 81 83 85 87 89
Year
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).
-------
M
~Z
«
Temperature for May
25 -,
20-
15-
10-
5 -
0 •
*
•« i
* A.
•
65 67 69 71 73 75 77 79 81 83 85 87 89 91
Year
93 95
Temperature for June
25 j
20 -
M .c
3 15 -
VI
"5 in .
0 1U
5 -
« * t
i
1
i
1 1 T 1 1 1 1 T -f 111.-
65 67 69 71 73 75 77 79 81 83 85 87 89 91
Year
— i i
93 95
Temperature for August
30 ]
25 ]
« 20
1 M
o>
O 10-
5-
65 67 69 71 73 75 77 79 81 83 85 87 89 91 93 95
Year
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
-------
E
«
B
E
3
Conductivity in May
500 -,
400-
300-
200-
100-
|
*
^ -
67 69 71 73 75 77 79 81 83 85 87 89 91 93
Year
E
u
o
£
3
Conductivity in June
400 j
300-
200-
100-
0 -
* •
f
^r
70 72 74 76 78 80 82 84 86 88 90 92
Year
94
E
u
0
E
Conductivity in August
450 T
400-
350-
300-
250-
200-
150-
100-
50 -
0 -
• • ^
f
w • i - i 1 1 — — j 1
69 72 75 78 81 84
Year
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
Selected
Abstracts
-------
"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-
jectives.
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.
247
-------
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
248
-------
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,
WA.
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.
sea972030014.DOC
' 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.
249
-------
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
self-maintenance.
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.
sea972030016.doc
1 Battelle Marine Sciences Laboratory. 1529 West Sequim Bay Road, Sequim, Washington 98382.
250
-------
» 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.
sea972040004.doc
1 Jones & Stokes Associates, 2600 V Street, Suite 100, Sacramento, California 95818-1914.
251
-------
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
system.
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,
OR.
sea/972020010.doc
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.
sea972040007.doc
'Department of Environmental Science, Policy, and Management, Univerity of CaHfonua, Berkeley, Odifbnri. 94720.
253
-------
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.
sea/972030006.doc
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.
254
-------
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
82:227-237.
Taylor, S. 1995. Spatial and temporal patterns in the
bottomland hardwood forest of Big Oak Tree State
Park. Master's Thesis. Southeast Missouri State
University.
sea972040006.doc
Department of Biology, Southeast Missouri State University, Cape Girardeau, Missouri 63701.
255
-------
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-
vidually.
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
28:467-490.
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.
sea972030011.doc
'Center for Urban Horticulture, College of Forest Resources, University of Washington. Contact address: 4724 University
View Place NE, Seattle, Washington 98105.
256
-------
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.
257
-------
Selected Abstracts
Factors in plant invasions
Anthropogenic disturbances:
Roads
Reservoirs
Pollution
Agriculture
Grazing
Forestry
Changes in intensity/frequency
of natural disturbances:
• Suppression of fire/flooding
• Dams
Changes in faunal populations:
• Introduction of exotic fauna
• Reintroduction of extirpated
species
• 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
14:285-294.
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).
sea971900012.doc
258
-------
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-
387.
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
42:134-140.
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-
486.
sea972030012.doc
1 Department of Natural Resources, Fernow Hall, Cornell University, Ithaca, New York 14853.
259
-------
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-
978.
sea972030017.doc
1 Monsanto Agricultural Co., 17004 N.E. 37th Circle, Vancouver, Washington 98682-8616.
260
-------
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
14:79-80.
sea972030018.doc
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.
261
-------
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.
262
-------
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-
don.
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
31(3):505-514.
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
14786.
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-
64.
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
Press.
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.
sea972040002.doc
263
-------
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.
sea972030013.doc
1 Department of Geology, University of Iowa, Iowa City, Iowa 52242.
264
-------
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.
Introduction
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.
265
-------
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.
Results
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.
Discussion
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
transplantation.
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
plants.
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".
266
-------
170
210
250
OKIS SINCE PLANTING
330
170
Figure 1. Increases in Area of Marsh Zone Occupied by Tule Clumps and Number of Stems at La Franchi
Demonstration Wetland.
267
-------
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.
Acknowledgements
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.
sea972270001.doc
268
-------
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,
California.
Introduction
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
process.
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
work.
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
monitoring.
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
construction.
• 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
goals.
1 Ogden Environmental and Energy Services Company, 5510Morehouse Drive, San Diego, California 92121.
269
-------
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
construction.
• Intensive field spotting of plant materials and
hands-on coordination with contractors during
construction.
• 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-
uling.
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
context:
• Appropriate agency review for type of
impacts/mitigation required
• Evaluate permitting/processing timeline
• Sequence of review with regulatory
agencies.
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
270
-------
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
incurred.
L. Coordinate with general contractor and land-
scape contractor for installation:
• Explain environmental constraints
• Help develop implementation strategy/
schedule
• 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.
sea972030009.doc
271
-------
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
98104.
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
272
-------
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
sources.
Hydrologic and Ecological
Modeling and Geographic
Information Systems for
Predicting Floodplain Forest
Restoration Following Dam
Removal
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-
sionmakers.
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
98102.
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
273
-------
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
274
-------
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
94132.
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
Project
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,
275
-------
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.
Gregoire2
'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
Wetlands
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
98195-7980.
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
276
-------
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
Review
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
277
-------
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
Watershed
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
River
James M. Helfield1, Miriam L. Diamond
Department of Geography, University of Toronto, 100
Street, George Street, |