Handbook of Ecological Restoration

Volume 2
Restoration in Practice

Edited by
Martin R.Perrow
ECON
University of East Anglia
and
Anthony J. Davy
University of East Anglia

CAMBRIDGE
UNIVERSITY PRESS
PUBLISHED BY THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGE

The Pitt Building, Trumpington Street, Cambridge, United Kingdom

CAMBRIDGE UNIVERSITY PRESS

The Edinburgh Building, Cambridge CB2 2RU, UK
40 West 20th Street, New York, NY 10011-4211, USA
477 Williamstown Road, Port Melbourne, VIC 3207, Australia
Ruiz de Alarcon 13, 28014 Madrid, Spain
Dock House, The Waterfront, Cape Town 8001, South AtTica
http://www.cambridge.org

@ Cambridge University Press 2002

This book is in copyright. Subject to statutory exception

and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without
the written permission of Cambridge University Press.

First published 2002

Printed in the United Kingdom at the University Press, Cambridge

Typeface Swift 9/12 pt. System YITEX

20 [TEl

A catalogue record for this book is available from the British Library

Library of CongressCataloguing in Publication data

Handbook of ecological restoration / [edited by] Martin R. Perrow, Anthony J. Davy.
p. cm.
Includes bibliographical references.
Contents: [1] Principles of restoration - [2]Restoration in practice.
ISBN 0 521 791294
1. Restoration ecology. 1. Perrow, Martin R. (Martin Richard), 1964- II. Davy, A. J.
QH541.15 .R45 H36 2002
333.95'153 - dc21 2001043443

ISBN 0 521 79129 4 hardback

Contents

List of contributors page ix 9 Beaches 197

Foreword xiii CLIVE A. WALMSLEY
Preface xv
10 Coastal dunes 214
Part 1 Restoration policy and infrastructure SIGURDURGREIPSSON

1 The Americas: with special reference 11 Saltmarshes 238

to the United States of America 3 JOY B. ZEDLER AND PAUL ADAM
MOHANK. WALl, NlRANDERM. SAFAYAAND
12 Rivers and streams 267
FATIH EVRENDILEK
PETERW. DOWNS, KEVINS. SKINNER
2 Europe 32 AND G. MATT KONDOLF
F. JANE MADGWICKAND TIM A. JONES 13 Lakes 297
3 Africa 57 ERIK JEPPESEN AND ILKKA
WILLIAM M. ADAMS SAMMALKORPI

4 Asia 78 14 Freshwater wetlands 325

BRYAND. WHEELER,RUSS P. MONEY
4a India 78 AND SUE C. SHAW
B. B. DHAR AND M. K. CHAKRABORTY
15 Polar tundra 355
4b China: progress in the reclamation BRUCEC. FORBESAND JAY
of degraded land 89
D. McKENDRICK
MING H. WONGAND ANTHONY
D. BRADSHAW 16 High-elevation ecosystems 376
KRYSTYNA M. URBANSKA AND JEANNE
5 Oceania 99 C. CHAMBERS
IAN D. HANNAMAND ALISONM. COCHRANE
17 Atlantic heathlands 401
Part 2 The biomes NIGELR. WEBB

6 Marine and coastal ecosystems 121 18 Calcareous grasslands 419

STEPHENJ. HAWKINS,JANETTER. ALLEN, MICHAELJ. HUTCHINGSAND
PAULINEM. ROSSAND MARTINJ. GENNER ALAN J. A. STEWART

7 Seagrasses 149 19 Prairies 443

MARKS. FONSECA,W. JUDSON KENWORTHY, SCOTT D. WILSON
BRIANE. JULIUS, SHARONSHUTLERAND
20 Semi-arid woodlands and
STEPHANIE FLUKE
desert fringes 466
8 Coral reefs 171 JAMESARONSON,EDWARDLE FLOc'H
SUSANCLARK AND CARLOS OVALLE

vii
viii Contents

21 Australian semi-arid lands 24 Tropical dry forest: Area de

and savannas 486 Conservaci6n Guanacaste, northwestern
DAVIDJ. TONGWAYAND JOHN A. LUDWIG Costa Rica 559
DANIEL H. JANZEN
22 Temperate woodlands 503
PETER BUCKLEY, SATOSHI ITO AND Index 585
STEPHANE McLACHLAN Colour plates between pages 6-7 and 558-559
23 Tropical moist forest 539
KAREN D. HOLL
Contributors

Paul Adam University of Nevada,

School of Biological Sciences 920 Valley Road, Reno NV 89512, USA
University of New South Wales, Susan Clark
Australia
Department of Marine Sciences and
William M. Adams Coastal Management
Department of Geography University of Newcastle
University of Cambridge Newcastle upon Tyne NE1 7RU, UK
Cambridge CB2 3EN, UK Alison M. Cochrane
Janette R. Allen 10 Valentine Avenue
Port Erin Marine Laboratory Parramatta
University of Liverpool NSW 2150, Australia
Port Erin David N. Cole
Isle of Man IM9 6JA, UK Aldo Leopold Wilderness Research Institute
James Aronson Rocky Mountain Research Station
USDA Forest Service Missoula MT 59801, USA
Centre d'Ecologie Fonctionelle et Evolutive
CNRS (UPR 9056) David K. Conlin
34293 Montpellier, France Department of Biology
Robin F. Bay Colorado College
Department of Biology Colorado Springs CO 80903, USA
Colorado College B. B. Dhar
Colorado Springs Co 80903, USA Association of Indian Universities
New Delhi 110 002, India
Anthony D. Bradshaw
School of Biological Sciences Peter W. Downs
University of Liverpool PWA Ltd
Liverpool, UK Corte Madera CA 94925, USA
and
Peter Buckley School of Geography
Imperial College at Wye University of Nottingham
University of London
Nottingham NG7 2RD, UK
Wye
Ashford TN25 5AH, UK James J. Ebersole
Department of Biology
M. K. Chakraborty Colorado College
Central Mining Research Institute (CMRI), Colorado Springs CO 80903, USA
Barwa Road,
Fatih Evrendilek
Dhanbad 826001, India
Department of Landscape Architecture
Jeanne C. Chambers Mustafa Kemal University
Rocky Mountain Research Station, 31040 Anatakya
USDAForest Service, Hatay, Turkey

ix
x List of contributors

Stephanie Fluke Michael J. Hutchings

National Oceanic and Atmospheric Department of Biology
Administration University of Sussex
Office of General Counsel for Natural Resources Brighton BN1 9QJ, UK
St Petersburg FL 33702, USA Satoshi Ito
Mark S. Fonseca Forest Science Division
National Oceanic and Atmospheric Administration University of Miyazaki
National Ocean Service Miyazaki 889-2192, Japan
Center for Coastal Fisheries and Habitat Research
Beaufort NC 28516, USA Daniel H. Janzen
Department of Biology
Bruce C. Forbes University of Pennsylvania
Arctic Centre Philadelphia PA 19104, USA
University of Lapland
FIN-96101 Rovianemi, Finland Erik Jeppeson
National Environmental Research Institute
Martin J. Genner DK-8600 Silkeborg, Denmark
Biology and Ecology Division
School of Biological Sciences Tim A. Jones
University of Southampton DJEnvironmental
Southampton SO16 7PX, UK Harper's Mill
Berrynarbor
Sigurdur Greipsson North Devon EX34 9TE, UK
Department of Biological and Environmental
Sciences Brian E. Julius
Troy State University National Oceanic and Atmospheric Administration
Troy 36082, USA National Ocean Service
Damage Assessment Center
Ian D. Hannam
Silver Spring MD 20910, USA
10 Valentine Avenue
Parramatta W. Judson Kenworthy
NSW 2150, Australia National Oceanic and Atmospheric Administration
National Ocean Service
Stephen J. Hawldns
Center for Coastal Fisheries and Habitat Research
Marine Biological Association UK
Citadel Hill Beaufort NC 28516, USA

Plymouth PLl 2PB, UK G. Matt Kondolf

and
Department of Landscape Architecture and
Biology and Ecology Division Environmental Planning
School of Biological Sciences University of California-Berkeley
University of Southampton Berkeley CA 94720, USA
Southampton SO16 7PX, UK
Edouard Le Floc'h
Karen D. Holl
Centre d'Ecologie Fonctionelle et Evolutive
Environmental Studies Department
CNRS (UPR 9056)
University of California-Santa Cruz
34293 Montpellier, France
Santa Cruz CA 95064, USA
 List of contributors xi

John A. Ludwig Ilkka Sammalkorpi

Division of Wildlife and Ecology Watercourse Planning and Restoration
CSIRO Finnish Environment Institute
Winnellie FlN-00251 Helsinki, Finland
Darwin
Sue C. Shaw
NT 0822, Australia
Wetland Research Group
F. Jane Madgwick Department of Animal and Plant Science
European Freshwater Programme University of Sheffield
c/o WWF Denmark Sheffield SlO 2TN, UK
DK-2200 Copenhagen N, Denmark Sharon Shutler
Jay D. McKendrick National Oceanic and Atmospheric
PO Box 902 Administration
Palmer AK 99645, USA Office of General Counsel for Natural
Resources
Stephane McLachlan
Silver Spring MD 20910, USA
Environmental Science Program
University of Manitoba Kevin S. Skinner
Winnipeg R3T 2N2, Canada School of Geography
University of Nottingham
Russ P. Money
Nottingham NG7 2RD, UK
Wetland Research Group
Department of Animal and Plant Alan J. A. Stewart
Science Department of Biology
University of Sheffield University of Sussex
Sheffield S10 2TN, UK Brighton BNl 9QJ, UK
Carlos Ovalle David J. Tongway
CRI Quilamapu Gungahlin Homestead
Instituto de Investigaciones CSIRO Sustainable Ecosystems
Agropecuarias (INIA) GPO Box 284
Casilla 426 Canberra
Chillan, Chile ACT,Australia

Pauline M. Ross Krystyna M. Urbanska

Geobotanisches Institut ETH
Marine Biological Association UK
Citadel Hill CH-8044 Zurich, Switzerland
Plymouth PLl 2PB, UK Mohan K. Wali
and OSU Environmental Science Graduate Program
Faculty of Science and Technology and
University of Western Sydney School of Natural Resources
Richmond The Ohio State University
NSW 2753, Australia Columbus OH 43210, USA

Nirander M. Safaya Clive A. Walmsley

Reclamation Division Countryside Council for Wales
North Dakota Public Service Commission Fford Penrhos
Bismarck ND 58505, USA Bangor LL57 2LQ, UK
xii List of contributors

Nigel R. Webb Ming H. Wong

NERC Centre for Ecology and Hydrology Institute for Natural Resources and
Winfrith Technology Centre Environmental Management
Dorchester DT2 8ZD, UK and
Department of Biology
Bryan D. Wheeler
Hong Kong Baptist University
Wetland Research Group
Hong Kong SAR
Department of Animal and Plant Science
People's Republic of China
University of Sheffield
Sheffield S10 2TN, UK
Joy B. Zedler
Scott D. Wilson Botany Department and Arboretum
Department of Biology University of Wisconsin-Madison
University of Regina Madison WI 53706, USA
Regina
Saskatchewan S4S 0A2, Canada
7 . Seagrasses
MARK S. FONSECA, W. JUDSON KENWORTHY, BRIAN E. JULIUS,
SHARON SHUTLER AND STEPHANIE FLUKE

INTRODUCTION plasticity within individual species. For example,

the canopy height of the smallest species lmown,
Seagrasses are marine flowering plants consisting
Halophila decipiens,usually never exceeds 10 cm
of 12 genera and approximately 60 species growing
while species of Zostera can have canopies exceed-
in all of the world's oceans with the exception of
ing 5-7 m in height. The diversity of clonal growth
the most polar regions (den Hartog, 1970; Phillips
forms and sexual reproductive strategies is accom-
&Menez, 1988). Nearly all seagrasses grow in uncon-
panied by phenotypic variation that allows the lim-
solidated sediments in water depths ranging from
ited number of seagrass species to occupy a wide
the intertidal zone to as deep as 35-50 m. They
range of environmental conditions from wave-swept
are vascular plants anchored to soft sediments by
shorelines to relatively deeper regions of continen-
a functional and complex rhizome and root system,
tal shelves. Only a few other macroscopic plants
with the exception of the genus Phyllospadix which
growing in the ocean are capable of filling the niche
grows on solid substrates along the Pacific coast of
type that seagrasses occupy.
the United States. The seagrass leaf canopy baffles
Even though taxonomic biodiversity is limited
the flow of water, and together with their rhizome
in seagrasses, the diversity of size and morpholog-
and root mat seagrasses stabilise sediments, cleanse
ical forms is accompanied by different growth and
the water column of fine particles, and recycle nu-
survival strategies uniquely adapted to the environ-
trients between the sediments and overlying waters
ments where the plants thrive. The range of growth
(Fonseca, 1996). Numerous species of invertebrates
strategies is also responsible for the patterns of sea-
and large vertebrates consume seagrasses as a por-
grass bed development seen throughout the world.
tion of their diet, and the complex structure and
This is especially evident in multi-species tropi-
physical stability provided by seagrasses form the
cal seagrass communities where distinctive succes-
basis for productive ecosystems consisting of plant
sional processes are evident in the formation of
and animal epiphytes, benthic macroalgae, inverte-
stable climax communities and in their response
brates, mobile vertebrates and numerous other or-
to disturbance (Zieman, 1982). In tropical seagrass
ganisms (Thayer et at, 1984). Many of the animal
communities, colonising and climax species can be
species that utilise seagrasses rely on their struc-
readily distinguished from one another and the
tural complexity to provide shelter and sources of
unique attributes of these species can be utilised to
food for their juvenile stages. This is one of the
enhance their protection and restoration (Fonseca
most important biological functions of the seagrass
et at, 1987).
ecosystem.
The lack of taxonomic biodiversity in seagrasses
is compensated by a wide diversity of size and RATIONALE FOR RESTORATION
morphological growth forms. The size and biomass
of seagrass varies over an order of magnitude Fortunately, in many countries, the battle to recog-
(Kenworthy et at, 2000), resulting partly from nise seagrasses as critical coastal ecosystems wor-
genotypic differences as well as from phenotypic thy of conservation and restoration has been won.

'49
'5° MARKFONSECA ET AL.

This recognition can be credited to the publication restoration then becomes a much more difficult
of thousands of papers from dozens of countries task, because it is nearly impossible to replace the
around the world representing years of research. To attributes seagrasses provide, and a way to correct
the best of our knowledge, research has yet to record the physicochemical properties of the system must
a seagrass bed which is anything but a faunal-rich, be found before reintroduction of the seagrasses can
highly productive ecosystem, that stabilises the sea begin.
floor, limits coastal erosion and filters the water col- We posit that the issues regarding seagrass
umn (Wood et a!., 1969). Thus, the ecological and restoration are not the technology of planting and
sociological value of seagrasses has been broadly raising seagrass beds, but the failure to apply basic
established (Wyllie-Echeverria et a!., 2000). Where ecological principles in implementing restoration
these values are not recognised, it often appears actions. Seagrasses can be readily transplanted and
to be the result of local political and development when sites are appropriately selected (see below
interests overriding conservation values (personal and discussion in Fonseca et a!., 1998a), signifi-
observation). cant restoration successes have emerged. In fact,
Threats to seagrass ecosystems and causes of new technologies are continually being developed
degradation arise from a wide variety of sources. in both the deepwater (Perth, Western Australia:
Eutrophication, coastal construction, motor vessel Fonseca et a!., 1998b; E. Paling, personal commu-
operation, fishing practices and many other activ- nication) and shallow water (Tampa Bay, Florida:
ities have led to both local and regional losses of J. Anderson, personal communication) approaches.
seagrasses (Short & Wyllie-Echeverria, 1996; Fonseca Also, improvements in large-scale seeding tech-
et a!., 1998a). Losses of seagrass also occur through niques are being advanced which have promise with
natural processes such as disease (Muelstein, 1989; some seagrass species (Granger et a!., 2000; Orth
Robblee et al., 1991), tropical cyclones (Preen et a!., 2000). We are only just beginning to recognise
et al., 1995) and overgrazing by invertebrates (Rose the many situations in which opportunities for sub-
et al., 1999). Where the species composition and stantial restoration have either been squandered or
life-history strategies promote recolonisation, sea- serious mistakes in site selection have been made,
grasses can recover naturally from perturbations largely because those involved did not understand
(Preen et al., 1995). However, in many instances ei- the habitat requirements and/or the life history of
ther the severity of the environmental modifica- the plants with which they were working.
tion responsible for the declines or the extremely The ecological value of seagrasses translates into
slow rate of natural recovery leads to long-term enormous commercial and social benefits. For ex-
losses. For example, in climax tropical communi- ample, in the Indian River Lagoon, Florida seagrass
ties dominated by Thalassia testudinum the time to meadows have been described as the marine equiv-
full recovery in severely damaged vessel ground- alent of tropical rainforests providing the ecolog-
ing sites can be more than a decade (Kenworthy ical basis for fisheries worth about US$25000 per
et al., 2000; Whitfield et al., in press). In these in- hectare or a total of approximately 1 billion dol-
stances, loss of seagrasses leads to numerous un- lars a year (Virnstein & Morris, 1996). Seagrass-
desirable and difficult-to-reverse conditions, most dependent fisheries and wildlife communities are
importantly the elimination of habitat structure the economic foundation for commercial and recre-
and the sediment stabilisation properties of the ational fishermen as well as for a variety of indus-
canopy and rhizome mat. A negative feedback on tries and people that utilise the coastal zone for
the ecosystem results; once the seagrass cover is commerce and personal enjoyment. Socially, these
lost and with it the self-sustaining properties of the values are transferred to the health and well-being
system provided by the seagrasses, modification of of the families of these user groups and the re-
the sediments and degradation of the water col- gional economies of nations worldwide. The many
umn may proceed without interruption. Seagrass physical. biological. economic and social attributes
 Seagrasses 151

combine to make seagrasses an essential and ecolog- holes caused by vessel groundings or altering water
ically important habitat in coastal marine ecosys- flow. Off-site selections, in our experience, have had
tems (Wyllie-Echeverria et aI., 2000); consequently higher probabilities of restoration failure because
they are in need of restoration where they have been inexperienced resource managers choose inappro-
anthropogenically injured or lost (Fonseca et aI., priate sites. They are frequently under the impres-
1996; Sheridan, 1999). sion that open habitat areas are prime sites for
restoring seagrass when, in reality, the sites selected
either cannot support seagrass, or currently support
PRINCIPLES OF RESTORATION
only low levels of seagrass.This fallacy has been ad-
We base our assessment of the status of seagrass dressed in detail in several publications (Fredette
restoration on a perspective from within the United et aI., 1985; Fonseca et aI., 1987, 1998a; Fonseca, 1992,
States legal framework. Seagrass beds in United 1994). Suffice it to say, Fredette et aI.'s (1985) con-
States coastal waters are generally viewed as pub- dition 'If seagrass does not grow there now, what
lic trust resources, and such injuries to these re- makes you think it can be established?' best sums
sources are considered losses suffered by the pub- the problem. Recently, Calumpong et aI. (in press)
lic. A number of federal and state laws include listed the criteria for off-site selection that can be
liability provisions which allow the public to be used to avoid off-site selection problems. By giv-
compensated for injuries to seagrasses (for exam- ing attention to these details of site selection, the
ple, the United States's National Marine Sanctuary probability of successful restoration can be greatly
Act of 1972, 16 USC 1431 et seq.). To evaluate this enhanced.
loss in a fair and reasonable manner, we must con-
sider not only the static loss in area and/or de- RESTORATION IN PRACTICE
gree of the injury, but also the loss of resource
services provided by the seagrass bed between the We have dealt previously with what we consider to
time it is injured and the time it recovers to be the status of this aspect of restoration (Fonseca
100% of pre-injury conditions (Fonseca et aI., 2000a). et aI., 1998a). However, there are at least four major
This approach is consistent with the 'no-net-loss' of deficiencies in the process of seagrass restoration.
wetlands policy that has become a benchmark of First, the choice of an appropriate metric for evalu-
restoration strategies in the United States. Our more ating restoration has been elusive. We present here
recent approach substitutes for the 'mitigation' or for the first time findings from a panel of United
'replacement ratio' used to identifY the amount States seagrass experts that considered what are the
of habitat to be generated to offset the amount appropriate metrics for tracking the performance
lost. In the past, use of replacement ratios has fre- of a seagrass restoration project. Second, setting
quently led to undercompensation oflost resources fair, reasonable and consistent ratios for replace-
because lost interim ecological services were not ment of damaged seagrasses has also been at issue.
addressed. We review the methodology used by the National
Effecting no-net-loss and achieving recovery of in- Oceanic and Atmospheric Administration (NOAA)
terim resource services requires that the injured for defining the interim loss of resource services
site be fully rehabilitated (on-site restoration), al- accrued by damage to seagrass beds and the pro-
ternative compensatory restoration sites be found cess of computing compensatory restoration. Third,
(off-site) or some degree of both. To limit the we feel that the weakest part of seagrass restora-
scope of discussion, we are focusing on in-kind tion has been the selection of the restoration site.
restoration (i.e. seagrass service loss replaced by sea- We delve into the pitfalls of site-selection strategy -
grass service gains). On-site restoration can often the point in the process where most plans go
be achieved, but may require engineering interven- awry. For completeness, we briefly review the extant
tions to 'fix' the site, such as filling excavation methodologies for restoring seagrass beds. Fourth
'52 MARK FONSECA ET AL.

and finally, finding a realistic basis for computing removed from the impact area. Thus, any functions
cost of these projects has been a vexing issue for affected by spatial elements of ecosystem linkages
years. Here we provide an evaluation of cost for are lost (i.e. geographic setting). Second, the lost pro-
the planning, implementation and monitoring of duction was removed from a specific point in time.
a seagrass restoration project based on a United Therefore, in some instances it cannot be returned
States federal court case successfully prosecuted by in a way to avoid disruption of ecosystem functions,
NOAA. such as the loss of last year's spawn of herring or set
of bay scallops that might occur as a result of injury
to a seagrass bed. Moreover, if there were a longer
period of time between the injury and full recov-
Definition of injury and evaluation oflost ery from the injury, then one could argue that re-
interim services
planting conducted a long time after an impact has
Defining lost resource services less value than ones conducted sooner. This realisa-
Computation oflost resource services requires three tion is the basis for NOAA'smore recent approach to
assessments: (1) area of habitat lost; (2) the length objectively and quantitatively standardise the prob-
of time needed for the functions associated with lem of computing interim lost services by habitat
that area (and lost to the ecosystem at large dur- equivalency analysis (HEA).This approach provides
ing the period of the injury) to recover to their a basis for setting replacement ratios and arriving
pre-impact levels; and (3) the shape of that recov- at a quantity of persistent area of given quality that
ery function (Fonseca et al., 2000a). Using seagrass has been defined as an appropriate metric of suc-
ecosystems as an example, if 1 hectare of seagrass cess (Fonseca, 1989, 1992, 1994; Fonseca et al., 1998a,
were destroyed today and replanted tomorrow and, 2000a).
for argument's sake, reached standards of equiva- Determination of interim loss and its implemen-
lency (e.g. shoot density, biomass, coverage) in two tation into the restoration process is tightly inte-
years, the interim loss of ecological services over grated with the establishment of a restoration plan.
this two-year period would be relatively low. How- While such a plan must identify the mechanics of
ever, if the restoration of this site were not under- the physical restoration itself, the plan must also
taken immediately and if the site required seven have a clear definition of injury, site selection, moni-
years to reach its pre-impact state, the level of com- toring protocols and success. As mentioned earlier,
pensation due the public for the interim losses from those guidelines have been established (Fonseca,
this same 1-hectare injury would be substantially 1989, 1992, 1994), but have not yet been quantita-
higher. This highlights the weakness of fixed com- tively coupled with the issue of interim loss to de-
pensation ratios. termine replacement ratios.
Actual projects rarely enjoy tight temporal coup- Recently, NOAAdeveloped and implemented HEA
ling between either the injury and on-site repair using basic biological data to quantify interim lost
work, or between the injury and the additional resource services (NOAA, Damage Assessment and
restoration required to compensate for the ecolog- Restoration Program, 1997a). While sharing many
ical services lost from the time of the injury until of the same principles as other methods incorpo-
full recovery. Among other issues, it is very difficult rating interim losses into replacement ratio calcu-
to consistently locate and successfully create new lations for wetlands (Unsworth & Bishop, 1994; King
seagrass habitat that meets ecologically responsible et al., 1993), HEA focuses on the selection of a spe-
site-selection criteria, especially those criteria which cific resource-based metric(s) as a proxy for the af-
preclude simply substituting naturally unvegetated fected services (e.g. seagrass short-shoot density in
bottom for vegetated bottom (Fonseca et a!., 1998a). the example discussed below), rather than basing
Finding large areas of suitable substrate for restora- its calculations on a broad aggregation of injured
tion in close proximity to the impacted area is rare, resources. Determination of this metric was one of
and often results in restoration at sites physically the conclusions from the expert panel as discussed
 Seagrasses 153

in Box 7.5 (biomass, as opposed to shoot density, has by the injured and replacement resources to be cap-
not yet been adopted because of a lack of empirical tured and incorporated into the replacement ratio
data on the recovery rate of belowground biomass, (NOAA, Damage Assessment and Restoration Pro-
whereas recovery rate of shoots is a robust data set; gram, 1997b). Without specification of a quantifi-
this choice is an extremely generous concession to able resource metric, analysis of the recovery of the
the responsible parties). This approach has the ad- resource following injury and/or the success of the
vantage of making HEA applicable not only to a restoration project may be difficult to evaluate pre-
wide range of different habitats, but to injuries to cisely. For example, in the wetlands context, alterna-
individual species as well (see Chapman et aI. [1998] tive metric specifications may lead to significantly
for a discussion of HEA applied to the calculation of different maturity horizons (Broome et aI., 1986)
compensation for historic salmon losses). Addition- as well as the level of functional equivalence ulti-
ally, the selection of a resource-based metric allows mately achieved by the restoration project (Zedler &
for differences in the quality of services provided Langis, 1991). .
154 MARK FONSECA ET AL.

Description of the compensatory restoration At its most basic level, REA determines the appro-
scaling approach priate scale of a compensatory restoration project
Accurate determination of the appropriate target by adjusting the project scale such that the present
scale of compensatory restoration1 projects is neces- value of the compensatory project is equal to the
sary to ensure that the public and the environment present value of interim losses due to the injury
are adequately compensated for the interim service of that action (e.g., freshwater diversion projects in-
losses. For injuries to seagrass resources, NOAA has tended to create wetland acreage).2 This 'balancing'
employed REA as the primary methodology for scal- of gains and losses is accomplished through a four-
ing compensatory restoration projects. The princi- step process (NOAA, 1997a). First (step 1), the ex-
pal concept underlying REA is that the public and tent, severity, and duration of the injury (from the
the environment can be made whole for injuries time of the injury until the resource reaches its
to natural resources through the implementation point of maximum recovery), and functional form
of restoration projects that provide resources and of the recovery curve must be determined, in or-
services of the same type, quality and comparable der to calculate the total interim resource service
value. REA has been applied in cases centered on losses. Next (step 2), the resource services provided
seagrass injuries because those incidents typically by the compensatory project over the full life of the
meet the three criteria defined by NOAA:(1) the pri- project must be estimated to quantifY the benefits
mary category of lost on-site services pertains to the attributable to the restoration. This step is analo-
biological function of an area (as opposed to direct gous to the previous one and requires estimation
human uses, such as recreational services); (2) fea- of both the time required for the compensatory
sible restoration projects are available that provide restoration project to reach its maximum level of
services of the same type and quality and are com- service provision and the functional form of the
parable in value to those lost; and (3) sufficient maturity curve. After these resource service losses
data on the required REA input parameters exist or and gains have been quantified, the scale of the com-
are cost-effective to collect. If these criteria are not pensatory project is adjusted until the projected fu-
met for a particular injury, other valid, reliable ap- ture resource service gains are equal to the interim
proaches and methodologies are available for scal- losses associated with the injury (step 3). This pro-
ing the chosen compensatory restoration projects cess is depicted graphically in Fig. 7.1, where the
(NOAA, 1997b). These criteria for the use of REA scale of the compensatory restoration project is ad-
were upheld by the US District Court (United States justed until the area under the maturity curve (the
of America v. Melvin A. Fisher et. aI. 1997 92-10027-CIV- total resource service gains, represented by area B)
DAVIS).Of equal importance to the Mel Fisher de- is equal to the interim lost resource services (repre-
cision was the decision by the US District Court in sented by area A). Because these services are occur-
UnitedStates of Americav. GreatLakesDredge& Dock ring at different points in time, they must be trans-
Co. 1999 97-2510-CIV-DAVIS to uphold the use of the lated into comparable present value terms through
REA as a proper method by which to scale compen- the use of a discount rate.
satory restoration.

'Compensatory restoration refers to any action taken to gains from the compensatory project exceed the total
compensate for interim losses of natural resources and discounted losses. This situation occurs when the scale of the
services that occur from the point of the injury until preferred project can only be adjusted according to a binary
recovery of those resources/services to baseline. Conversely, or stepwise function rather than a continuous function. or
primary restoration refers to actions that return the injured when the resulting amount of natural resources/services
natural resources and services to baseline. generated by a restoration action cannot be tightly controlled
2In some instances, it may be beneficial to all parties following implementation.
involved to implement a project where the total discounted
 Sea grasses '55

en
CD
Injured Area The importance of site selection
U
">
CD Clearly one of the largest problems with seagrass
(/)
CD
u transplanting is finding an appropriate place to con-
::;
0
en
duct the restoration and install the plantings. It
CD
II: is not advised to plant seagrasses in areas with
Time no history of seagrass growth, or where the afore-
mentioned disturbances have not ceased. Planting
Compensatory Restoration Project should not be done under those circumstances be-
en
CD
U cause of the low probability of success. Planting may
os;
CD be done in open, unvegetated areas among patches
(/)
CD
u
of seagrass, but only for the goal of experimen-
::;
0
en
tal manipulations and/or the evaluation of plant-
CD
II: ing techniques (keeping in mind that these among-
Time patch locations are not a strong test of the efficacy
Fig. 7.1. Graphical depiction of how habitat equivalency of a technique as they are embedded within viable
analysis (HEA) sets the compensatory restoration to equal seagrass territory). Seagrass patches migrate, alter-
interim loss of resource services. This is achieved by nately colonising currently unvegetated sea floor
setting the total services (hectare-years) lost until and dying out where seagrass is located presently
complete recovery back to pre-injury conditions (area A) (Marba et al., 1994; Marba & Duarte, 1995; Fonseca
is equalled by the services rendered under the et at, 1998a, 2000b). Thus, the spaces between the
compensatory project (area B). patches today may be naturally colonised by sea-
grasses in the future.
Discounting is a standard economic procedure
Campbell et al. (2000) provide a decision strat-
that adjusts for the public's preferences for having
egy for assessing the selection planting sites that
resources available in the present period relative to
include measures of light, epiphytisation, nutrient
a specified time in the future. Because of discount-
loading, water motion, depth, proximity of donor
ing, plantings that occur longer after an impact
site and alternative actions (Fig. 7.2). Similarly, Fon-
are worth less in present-value terms than plant-
seca et al. (2000a) and Calumpong et at (in press) give
ings conducted shortly after an impact, and there-
the following criteria for the selection of a restora-
fore more planting must be done as time elapses.
tion site away from the original injury site:
Finally (step 4), appropriate performance standards
associated with the compensatory restoration must
be developed to ensure that the project provides the
.. It is at depths similar to nearby seagrass beds

anticipated level of services. Well-defined and meas-

urable standards are essential to the success of the
. It was anthropogenically disturbed
It exists in areas that are not subject to chronic

project regardless of whether the restoration will be

implemented by the parties responsible for the orig-
. storm damage
It is not undergoing rapid and extensive natural re-

inal resource injury or by the management agency

(trustees) using monetary damages which are recov-
. colonisation by seagrasses
Seagrass restoration
sites
has been successful at similar

..
ered.
There is sufficient area to conduct the project
In Box 7.5, we present the outcome of a national Similar quality habitat would be restored as was lost.
workshop that set the stage for NOAA to provide
reasonable and fair assessments of injuries to sea- These selection criteria have been used success-
grasses and the effort needed to recover the lost re- fully in the US Federal Court as the basis for sea-
sources which must be assumed by the responsible grass restoration projects (United States of America
party. v. MelvinA. Fisheret al. 1997). By considering these
156 MARK FONSECA ET AL.

Light
requirement NO-----.
known
--.- Undertake
Yes
evaluation
From
transplant No
bioregion

Can site No
water quality
be
enhanced?

Fig. 7.2. Decision flow diagram regarding site selection for restoration. From Campbell
et aT.(2000). lsat = saturation irradiance, Ie = compensation irradiance.

criteria, it is apparent that transplantation should potential of being controlled by those conducting
probably not be undertaken for the purposes of en- transplants:
hancing recovery from natural disturbance events
as these events have both an ecological and evolu-
. Similarity of environmental conditions of donor

tionary function in determining the survival and

fitness of the seagrass ecosystem. When possible, re-
. and recipient beds.
Choice of species: preferably same as that lost, but
pioneering species may be substituted to initiate a
habilitation of the primary injury site should be
performed to restore or accelerate the recovery of
baseline service flows, with compensatory restora-
. project.
Presence of grazers or sediment burrowers: these
bioturbating organisms may need to be excluded or
tion used to compensate the public for interim ser-
plantings may have to be conducted in large patches
vice losses that accrue while the site reaches its pre-
injury levels of service provision. . to dissuade them from their activities.
Source of planting material: similar depths and
environmental conditions and from over as broad
Critical factors influencing transplant success a geographic area as possible to ensure genetic
Numerous factors have been determined
transplanting success. Some are of the crop-risk
to affect
. diversity.
Time of year: seagrass should be planted at a time to
type, are extrinsic and cannot be controlled. Oth-
ers involve issues of protocol. In a survey of North
American seagrass planting projects, Fonseca et al.
. ensure the longest period before seasonal stressors.
Cost: many variations of cost have been given but
standardised costs are elusive; based on recent cases
(1998a) listed the following as factors that had the in the US federal court, a contracted project that
 Seagrasses 157

includes site surveys, planting, monitoring and re- plant size remain some of the best general guide-
porting will cost (in 1996 US dollars) ~US$630 000 lines for matching donor and recipient beds.
per hectare. The characteristics of the species, such as fast
growing vs. slow growing; pioneering vs. climax, an-
It is essential to study the substrate-energy (ex- nual vs. perennial growth, etc. must be considered
posure) regime, and optical water quality (clarity before conducting transplantation work. For exam-
or light availability) of the area that will be trans- ple, Halodule spp. and Halophila spp. are fast-growing
planted so that suitable source materials can be pioneering species while Thalassia spp. and Enhalus
identified. Areas exposed at low tides should be care- spp. are slow-growing climax species. Halophila spp.
fully mapped so as to place plantings with mini- rapidly colonise disturbed areas like those with
mal exposure to air, unless the plants are regularly moving sand bars and are under-canopy species,
occurring in the intertidal zone (e.g. in the Pacific requiring low light. Although a climax species
Northwest of the United States). Planting in high may have been disturbed, it is often advisable to
wave energy or tidal current areas will require plant- first install a faster-growing species to stabilise the
ing in larger groups to avoid disruption (Fonseca environment.
et a!., 1998a, b). Planting in larger groups also ap- Another important factor in the selection of sea-
pears to be an effective method of deterring physi- grass for transplanting, besides their intrinsic recov-
cal disruption of the planting by marine organisms. ery rate, is their growth habit (Short & Short, 2000).
However, as suggested by Addy (1947), matching wa- Transplanting can be rendered almost wholly inef-
ter depths, temperature, salinity, water clarity and fective if meristematic regions of these plants are
'58 MARK FONSECA ET AL.

Mono-meristematic leaf-replacing Di-meristematic leaf-replacing

M
/

I "-
M M

Mono-meristematic non-leaf-replacing Di-meristematic non-leaf-replacing

""- ""-
M M

Fig. 7.3. The four basic growth forms of seagrass. In the case of the mono-meristematic,
leaf-replacing form, each terminal shoot on the runner is a viable planting unit -
comparatively low modular integration is present meaning that most often each shoot has
high potential for contributing to spatial colonisation. The other three forms require at
least three to four short shoots be maintained on the runner for a complete planting unit,
as well as an intact rhizome apical meristem. From Short & Short (2000).

damaged or not incorporated in sufficient quantity Florida, indicating that it would be necessary to
in a planting unit to initiate recolonisation. Short use exclosure cages to ensure the survival of trans-
& Short (2000) summarise the morphotypes of sea- planted seagrasses in some areas. Recently, we have
grass (Fig. 7.3). seen that planting in clumps of at least 20-50 em
Seagrass grazers can have disastrous effects on on a side deters many animals from disturbing the
plantings. Seagrass grazers include sea-urchins, plantings (authors' unpublished data).
gastropods and herbivorous fishes. Some migra- Minimisation of disturbance to the source bed is
tory waterfowl such as geese and ducks have paramount in seagrass transplanting so as not to ex-
been observed to decimate seagrass plantings (per- acerbate injury to local populations. With present
sonal observation). Significant grazing of natural techniques focusing on the use of wild, vegetative
Syringodium filiforme beds points out the general stocks, this may be achieved by conducting the
susceptibility of seagrasses to grazing (Rose et aI., transplantation in phases, or dispersing the collec-
1999). Fonseca et a1. (1994 and references therein) tion effort, thus allowing the source bed to recover.
found significant disturbance by rays in Tampa Bay, Harvesting of donor stock should also be done from
 Seagrasses '59

I
I
I
I
I Is rhizome
I growthrate
I high?
I
I Yes
I
Yes
I
I
Yes I
I
I
N0 -=::l
Try larger
I plugs OR
I increase
I Yes density of
sprigs
I
I Consider sprigs
I for transplant
I unit
I
I

Minimize handling

Fig. 7.4. Decision tree for choosing seedlings or whole, mature plants for transplanting.
We caution that the technology for seed establishment is not as developed as for the use
of sprigs or cores. From Campbell et al. (2000).

beds over as broad a geographical area as possible. over, Fonseca et aI. (1998a) suggested that Ruppia,
This may help avoid loss of genetic diversity in the Halophila, HaIodule and Zostera spp. can recover in
planted bed (sensu Williams & Orth, 1998) and may small patches «0.25 m2) within a year with shoot
actually incorporate the full local range of genetic density returning to normal. Furthermore, Fonseca
diversity into the planting. et aI. (1998a) cautioned that patches >~30 m2 in
Fortunately, many pioneering species can be high-current areas may never recover. Campbell et aI.
harvested with minimal disturbance to the beds (2000) also provide a decision tree for selection of
(Fonseca et a!., 1994). However, for climax species, planting stock for both sexual and asexual propag-
harvesting of donor beds may cause long-lasting ules that focuses on intrinsic propagation rates
damage and harvesting from these beds should only (Fig. 7.4).
occur when the beds are under some anthropogenic Choosing the time-frame for planting is an obvi-
source of physiological stress that does not seem ous concern, and as with all crops, the appropriate
likely to abate or if they are in imminent danger of time for seagrass varies with geographical region.
physical removal (e.g. dredging). In general, the best strategy is to plant at a time
The size of the source or donor bed should first be just after the period of highest seasonal stress, when
assessed to determine if recovery will proceed after natural populations are experiencing recovery. For
removal of the sods, cores or sprigs. This is especially example, eelgrass (Zosteramarina) should be planted
true when transplanting vegetative stock, as a large in the autumn in North Carolina, and other mid-
amount of material is needed. Spacing harvesting at Atlantic regions in the United States, because sum-
~0.25 m for small cores or sods «0.15 x 0.15 m) is mer is the period of maximum physiological stress
often sufficient to avoid long-lasting damage. More- at that location (Moore et al., 1997).
160 MARK FONSECA ET AL.

...
et aI. (1991) recommend a minimum of one rhi-
zome apical and at least three shoots per rhizome
segment.

-O.5m Seagrass should be planted either directly into

the bed (sprig) or anchored using a variety of de-
vices such as rods, rings, nails or Rebar. U-shaped

r I
metal staples with attached bare root sprigs (no
sediment) have been widely used as planting units
(Derrenbacker & Lewis, 1982; Fonseca et aI., 1982) or,
when negative buoyancy is not required, bamboo

- O.1m

Fig. 7.5. Diagrammatic representation of the two most

skewers may be substituted (Davis & Short, 1997).
Plants have also been woven into biodegradable
mesh fabric and attached to the sediment surface
as a planting unit (Fonseca et aI., 1998a). Rocky in-
widely used seagrass planting methods; those with tertidal species, such as Phyllospadix spp., have been
sediment such as cores and plugs and those without attached to boulders.
sediment, usually anchored with metal or wood staples.
When using anchoring devices, one must con-
sider using biodegradable or natural materials such
Planting methods as boulder over metal or plastics. As mentioned
above, when using staples, one can choose metal
Fonseca et a1. (1998a) list 14 categories of planting
(US$O.Oleach) or can modify 'shish kebab' bamboo
methods for seagrass in the United States. For these
sticks by bending them into a V (Davis & Short,
methods, source material can be vegetative stock or
1997) which when purchased in bulk could cost
seeds. Transplantation using vegetative stock typic-
only US$0.006 a piece. In tropical areas where bam-
ally requires available wild stock as a source and
boo is plentiful, this could be a more economical
is labour-intensive and invariably expensive. How-
medium to use. Bamboo is also biodegradable. Using
ever, it often gives faster, more reliable coverage
either kind of staple, planting units are made by
than seed methods (but see review by Grth et aI.,
grouping plants and attaching the root-rhizome
2000). Most projects today are carried out using ei-
portion under the bridge of a staple and securing
ther small sods or sprigging of sediment-free units
the plants with a paper-coated metal twist-tie. This
(Fig. 7.5).
can either be prepared beforehand or the planting
unit can be pinned directly to the substrate during
Sediment-free methods planting.
For most sediment-free methods, plants are dug up When using nails, boulders or Popsicle sticks
using shovels, the sediment is shaken from the roots (Merkel, 1988), the technique is more or less sim-
and rhizomes and the plants are placed in flow- ilar and the planting unit is tied to the anchor-
ing seawater tanks or floating pens. For vegetative ing instrument. Frames, such as Short's TERFdevice
stocks, Fonseca et a1. (1998a) recommend a mini- (F. T. Short, in Fonseca et aI., 1998a), have great
mum of one apical shoot per planting unit. The promise for rapid and non-diver-assisted planting at
number of short shoots on a long shoot should depth. A cage deployment system that has shoots
be maximised whenever possible, so as to derive attached to the bottom is lowered onto the sea
benefits from the clonal nature of the plant. Also, floor and retrieved after the shoots have rooted
the plants should be collected and planted on and their paper ties have decomposed. This elim-
the same day, kept in water with the same am- inates the need for divers in deeper water, can
bient temperature and salinity, and kept as moist be used in chemically polluted areas, and provides
as possible when out of the water. When using initial protection of the plantings from biological
vegetative stocks of ThaIassia testudinum, Tomasko disturbance.
 Seagrasses 161

Seagrass with sediment methods The plug method utilises tubes as coring devices
The sod or turf method consists of planting a shovel- to extract the plants with the sediment and rhi-
full of seagrass with sediment and rhizomes intact. zomes intact. The plugs are planted directly into
This is the easiest method, and is most applica- the seagrass bed after creation of a hole to receive
ble for hard, compact substrates and deep-rooted the contents of the tube. The core tubes are usually
and large species such as Enhalus acoroides.The only made of 4-6-cm diameter PVC plastic pipe with caps
equipment needed are shovels and large basins for for both ends to initially create a vacuum and keep
the sods. However, if the donor site is far away, sediments from washing out the bottom. The tube
transporting the sods may present a problem as is inserted into the sediment, capped (which creates
the weight of the material is a physical burden. a vacuum), pulled from the sediment and capped
Some species, such as E. acoroides,Posidonia spp. and at the other end to avoid losing the plug. This can
Thalassia spp. may have very deep root-rhizome sys- only be done with soft but cohesive sediments and
tems requiring removal of a tremendous amount of generally only for small species to avoid excessive
sediment to harvest the belowground plant struc- leaf shearing (unless extreme care is taken to avoid
tures all intact (Fig. 7.6). To our knowledge, this the shearing, which adds measurably to the cost
has only been accomplished in Western Australia ofthe process). When the donor bed is far away from
(by E. Paling, of Murdoch University; see review in the planting site, many tubes are needed which also
Fonseca et al., 1998b). Furthermore, harvesting an adds to the cost.
entire sod may constitute one of the most severe Sod pluggers extract a plug out of the donor
perturbations in a seagrass meadow, inhibiting re- bed which is then extruded into a peat pot; the
covery in the donor bed. method was first used by Robilliard & Porter (1976)

Fig. 7.6. Harvest of Posidonia sod near Perth, Western Australia. Photo courtesy E. Paling.
162 MARK FONSECA ET AL.

and modified by Fonseca et a!.(1994).Because these personal communication). Ruppiamaritima has been
pots are typically only a few em across, they may be successfully transplanted from laboratory culture
inserted into the bottom by liquefYing the sediment stock (Bird et a!., 1994), but all these species are
with a hand tool. Mter the peat pot is planted, its naturally fast growing (i.e. pioneering) and it is un-
side walls must be ripped off or torn down and the clear whether laboratory culture is a cost-effective
pot pushed into the sediment to allow the rhizomes means of restoring naturally prolific species. More-
to spread out. over, questions regarding the ability to maintain
genetic structure of the population have not been
Sowing of seed solved. Given the growing emphasis on mechanised
Seed planting holds promise for large-scale restora- plantings using wild stock, laboratory culture will
tion but is currently more applicable only in low- probably only be cost-effective when techniques are
energy areas where the seeds can settle and germ- developed for slow-growing species, hence avoiding
inate and where there are few seed predators. long-term donor bed impacts.
This method was first introduced by Thorhaug
(1974) with ThaIassia testudinum. A seedling grow-
Monitoring the restoration
out method for T. testudinum has been registered by
Lewis (1987). The availability of seeds must also be Monitoring of the restoration project is necessary
considered. Large areas in the Chesapeake Bay have to provide data required to evaluate the viability
been established by sowing seeds from a small boat of the project based on the performance standards
(R.J. Orth, personal communication). Work contin- (defined below). This permits timely identification of
ues in this highly promising area (Orth et aI., 2000). problems or conditions that may require corrective
Experiments using seeds pelletised to increase their action to ensure the success of the project.
density to facilitate sinking and seeds embedded in
biodegradable mesh are presently being carried out Monitoring schedule and activities
by Granger and his colleagues (Granger et aI., 2000). Field collection of data for performance monitor-
Experiments on planting depth also indicate that at ing should occur for four years after planting. Orig-
least for Zostera marina, seeds should be within the inal plantings should be monitored for three years
top 2 em of the sediment for best germination and and potential remedial plantings in year 2 should
that sowing densities should be 400-1000 seeds per be monitored for three years for a total monitoring
square metre (Granger et aI., 2000). period of four years. Under this schedule the moni-
toring would be conducted as follows:
Laboratory cultured stocks year 1 - day 60, 180, 365
This approach uses plants reared and grown in the year 2 - day 180, 365
laboratory from plant fragments. It may become es- year 3 - day 180, 365
pecially applicable for large-scale plantings where a year 4 - day 180, 365
large amount of planting units is needed. This tech- The precise dates are weather-dependent. In carbon-
nique also holds promise for reducing or eliminat- ate sediments, each surviving planting unit should
ing donor bed damage and this has been shown to receive an additional spike of constant-release phos-
be minimal for pioneering species, such as HaIoduIe phorous fertiliser (0-39-0, nitrogen-phosphorus-
wrightii and Syringodium fiIiforme (Fonseca et aI., potassium) at day 60 of year 1. Alternatively, bird
1994). This approach also has the potential to main- roosting stakes could be installed about every
tain donor stocks for unscheduled plantings and 5-10 m along scars (see Box 7.3).
could theoretically supply genetically variable and
disease-resistant plants. Data collection
Several aspects of this approach remain con- Monitoring should focus on documenting the num-
troversial. So far, three species have been success- bers of apicals at planting time, planting unit sur-
fully propagated in the laboratory, Ruppia maritima, vival, shoot density and areal coverage under the
HaIophiIa decipiensand H. engeImannii(M. Durako, following schedule and definitions. This monitoring
 Seagrasses 163

Box Tr la
for intertidal, bare-rool (e.g., staple teGbnique)and
RESTORING SLOW-G.ROWING SPECIES nine persons for subtidal bart;root plallting. 1'he use

ly plant
another, faster-spreading congener, such as HaloduIe inter- and subtidal with nine persons in one day.
wrightii to achieve 'compressed succession' (see Box 7.1). Fonseca et at (1994) provide a comparison among
s

conditIOn tor the slow-growmg one to

while the faster one stabilises the sediment and pot plugs being the most rapid method and cores
provides a functional seagrass habitat. being the slowest method.
Iti

'compressed sm:cession'<of H. wrfghtii can be enhallced basis, travel, reporting and monitoring. The timed
by fertilising transplantS with bird roosting stak~.The trials also did not accurately measure the effect of
boredom on the speed of the,process.

growth of H. wrightii transplants. In cases where there should be near enough so the shoots can be planted
is a mixed meadow of T. testudinum and H. wrightii the same day. Overnight storage of material,
f an iniured area can be accelerated particularly bare-root material, should be placed in

sufficiently controlled experiment to determine the

STEPS IN A GENERALIZED PROTOCOL
. Study the . storage capability for seagrasses.
Be prepared to manage the workforce with r,

air and waves and currents; (e) substrate

type - avoid clays and high organic sediments;
(d) rate of siltation -
plants often cannot withstand

. animal dis
Determine time-frame and budget by evaluating the
typical staffing requirements. Merkel (1992)
. p
Conduct thorough monitoring (see below) and be
prepared to conduct remedial plantings.

protocol applies to original plantings for three years 2. Survival. Each site should be examined for survival
(years 1-3) and to remedial plantings for three years of all planting units during each survey in year 1
(years 2-4). (days 60, 180 and 365) or until coalescence. Survival
of each species should be expressed as a percent-
1. Apical counts. Prior to planting, one planting unit age of the original number, but the actual whole
out of every 100 collected should be examined for number should also be reported.
the number of rhizome apicals. 3. Shoot density. A separate (from survival) random

L-
164 MARK FONSECA ET AL.

selection of three planting units per 100 planted of all video tapes and other photography should
should be assessed for number of shoots per plant- be turned over to the permitting agency follow-
ing unit at each survey time until coalescence be- ing project completion by the party conducting the
gins. Mter some planting units begin to coalesce, monitoring.
three randomly selected locations per 100 m2 (100
planting units) should be surveyed for shoot den- Remedial plantings and/or project modifications
sity over a 1 m2 area at 0.0625 m2 (25 cmx25 em)
resolution. Shoot density should be monitored for If data from a monitoring report establishes that
three years. the performance standards are not being met or are
4. Areal coverage. The randomly selected planting projected not to be met, remedial plantings of those
units (may be the same as shoot density selection)
should be surveyed for coverage at each survey time
starting at day 180 of year 1. Measurements should
be taken at a 0.0025 m2 (5 em x 5 em) resolution
prior to coalescence and over a 1 m2 area at 0.0625
m2 (25 em x 25 em) resolution after coalescence for
each seagrass species present at each survey time.
Areal coverage should also be monitored for three
years.
5. Video tape transects. Five 100-m transects along
randomly selected portions of the planted area
should be video tape recorded to establish perma-
nent visual documentation of the progression of
areal coverage of seagrass through time. A tape
measure should be laid along the central (long)
axis of the scar and should be included in the video
tape to allow physical reference of locations within
the scar. Video recordings should be taken at each
survey time during the monitoring period of three
years. Observation-based assessment of success may
be substituted if quadrats are used in accordance
with a Braun-Blanquet survey method (Fonseca et
a!., 1998a) or if the data are obtained from the
video tape (making the observational data base
available for cross-checking). The same number of
sample points must be obtained with the same spa-
tial extent (i.e. survey each scar). Similarly, Braun-
Blanquet observations of cover at every metre along
each scar may also be obtained from the video tape
to obtain estimates of planting performance.

Reporting requirements
Monitoring reports should include copies of raw
data gathered in each survey, an analysis of the
data, and a discussion of the analysis. Originals of
all video tapes recorded since the previous report
should be provided with each new report. Originals
 Seagrasses 165

affected seagrass species should occur. If there is Based on past experience in seagrass restoration
a recurring problem with survival of plantings or efforts, it is assumed that 30% of the planted area
replantings in a particular area, remedial planting should require remedial planting in year 2. All orig-
should occur in another suitable area in as close inal plantings should be monitored for three years.
proximity as possible, subject to the approval of Remedial plantings should also be monitored for
permitting agencies. three years.
166 MARK FONSECA ET AL.
 Seagrasses 167

Table 7.1. Top: General distribution (%) of costsby task Costs of restoration
(United States of America v. Salvors Inc.); bottom:
From our experience, there is a general set of fac-
summary of costs by specific actions (Fisher Natural
tors that drive up the cost of seagrass transplanting,
ResourceDamage Assessmentclaim)
particularly inappropriate site selection, inexperi-
ence, and disturbance events (requiring remedial
Percentage
planting). Consistent estimates of planting costs in
Task of total costs
dollars remain elusive, but recent restoration plans
Map and ground-truth 5.5 in the United States that have been litigated in the
Planting 18.5 federal courts have shown the full cost of a restora-
Monitoring 58.7 tion distributed among the various tasks (Table 7.1)
Contractor 8.3 at ~US$590 000 for a 1.55 acre area or ~US$940 000
Government oversight 9.1 per hectare (1996 dollars). Two important points
here are that: (1) the actual costs of collecting and
Type of cost US$ (1996 values) installing planting units is less than 20% of the ac-
tual cost of the entire project; (2) while monitor-
Damage assessment costs ing costs at first glance may appear high relative to
Federal assessment costs
planting costs, it is important to note that monitor-
(up to 26 October 1996) 211130
ing represents a labour-intensive, multi-year effort
Interest on federal to ensure that performance standards are met and
assessment costs at
necessary mid-course corrections are undertaken.
judgment 26 553
The majority of planting costs on the other hand
Subtotal 237683 are incurred at a single point in time. This cost
pattern is not unique to seagrass projects, but is
Restoration costs
commonly observed in natural-resource restoration
Primary restoration costs
projects across different types of habitats. We con-
(vessel-generated holes sider these data to be much more indicative of the
in sea floor - restoration
real costs of executing a restoration project than
deemed not feasible with
0 previously presented (e.g.Fonseca et aI., 1982).
current technology)
Restoration site selection
analysis 5465
CONCLUDING REMARKS
National Environmental Policy Act
compliance/permitting costs 14695 In this chapter we have dealt with what we consider
Preparation of to be some of the critical issues that must be ad-
map/ground-truthing sites 14 314 dressed in the implementation of effective restora-
Collection, preparation and tion projects. These issues include: (1) choice of an
installation of planting units 64846 appropriate metric, representative of the array of
NOAArestoration services provided by a resource, by which to mea-
oversight/supervision costs 17650 sure success; (2) evaluation oflost interim resources;
Subtotal 116970 (3) appropriate selection criteria for off-site restora-
tion projects; and (4) accurate project cost estima-
Monitoring costs tion. A fifth issue presented itself as we edited the
Monitoring of compensatory paper - the role of disturbance. Disturbance is a fun-
prop scar areas 205650 damental ecological process and we noticed that it
Contractor profit on repeatedly worked its way into our discussions, sig-
restoration/monitoring work 29028 nalling its obvious but subtle role in influencing the
Grand total for claim 589331 outcome of restoration projects. Finally we review
methods, but we do not view these to be a weak
168 MARK FONSECA ET AL.

link in the process, per se. The weakness in meth- Calumpong, H. P., Phillips, R. c., Moppets, E. G., Estacion,
ods arises when workers do not study past efforts. J. S., de Leon, R. O. & Alava, M. N. (1992). Performance
Rather, failure of restoration arises in general from of seagrass transplants in Negros Island, central
not considering the broader context of ecological in- Philippines and its implications in mitigating degraded
juries, particularly issue (3).When restoration plans shallow coastal areas. In Proceedings of the 2nd RP-USA
are sent to us for consideration, the first aspect of Phycology Symposium/Workshop, eds. H. P. Calumpong &
the plan that we look at is the choice of a restora- E. G. Menez, pp. 295-313. Laguna, Philippines:
tion site. Almost without fail for those with little Philippine Council for Aquatic and Marine Research
restoration experience, a site is selected that is not and Development.
damaged and does not need repair (e.g. planting in Calumpong, H. P., Fonseca, M. S. & Kenworthy, W. J. (in
spaces among naturally patchy seagrass). press). Seagrass transplantation. In Global Seagrass
In the United States as elsewhere around the ResearchMethods, eds. F. T. Short & R. G. Coles
world, we have largely won the battle to recognise Amsterdam, The Netherlands: Elsevier.
the value of seagrasses as a national resource. How- Campbell, M. 1., Bastyan, G. R. & Walker, D. I. (2000). A
ever, the acceptance by US federal courts of our decision-based framework to increase seagrass
metrics for assessing success, the concept of interim transplantation success. In Proceedings of the 4th
resource service losses and the methods for quanti- International SeagrassBiology Workshop, eds. G. C.
fYing them, and the logic for selecting planting sites Pergent-Martini, M. C. Buia & M. C. Gambi, pp. 332-335.
has given us an unprecedented ability to foster effec- Corsica, France: Societa Italiana di Biologia Marina,
tive restoration of these habitats. More importantly, Biologia Marina Mediterranea.
perhaps, is the signal that this has sent to the de- Chapman, D., Iadanza, N. & Penn, T. (1998). Calculating
velopment community and responsible parties: that resource compensation, an application of the
this resource is of vital national importance and its service-to-service approach to the Blackbird Mine,
destruction cannot be tolerated by the public. While hazardous waste site. NOAA Technical Paper no. 97-1.
transplanting seagrass is not technically complex, Contributions in Marine Science,32, 41-48.
in order to meet the goal of maintaining or increas- Davis, R. & Short, F. T. (1997). An improved method for
ing seagrass area, careful attention to detail must transplanting eelgrass, Zostera marina 1. Aquatic Botany,
be paid to the entire process of planning, planting 59, 1-16.
and monitoring - a process that does not lend itself den Hartog, C. (1970). The Seagrasses of the World.
to oversimplification. As with all terrestrial crops, Amsterdam: North-Holland.

there are inherent risks with seagrass and failures Derrenbacker, J., Jr & Lewis, R. R. (1982). Seagrass Habitat
will inevitably occur. Given that despite collective Restoration, Lake Surprise, Florida Keys.Tampa, FL:
millennia of human experience we trade stock fu- Mangrove Systems Inc.
tures on the probability of successful cultivation of Durako, M. J., Hall, M. 0., Sargent, F. & Peck, S. (1992).
food crops, restoration of seagrass ecosystems will Propeller scars in seagrass beds: an assessment and
suffer from at least this kind of risk. experimental study of recolonisation in Weedon Island
State Preserve, Florida. In Proceedings of the 19th Annual
REFERENCES Conferenceon Wetlands Creation and Restoration, ed. F.J.
Webb, Jr, pp. 42-53. Tampa, FL: Hillsborough
Addy, C. E. (1947). Eel grass planting guide. Maryland Community College.
Conservationist,24, 16-17. Fonseca, M. S. (1989). Regional analysis of the creation
Bird, K. T., Jewett-Smith, J. & Fonseca, M. S. (1994). Use of and restoration of sea-grass systems. In Wetland Creation
in vitro propagated Ruppia maritima for seagrass and Restoration: The Status of the Science,vol. 1, Regional
meadow restoration. Journal of Coastal Research, 10, Reviews, eds. J. A. Kusler & M. E. Kentula, pp. 175-198.
732-737.
Corvallis, OR: Environmental Research Laboratory.
Broome, S. W., Seneca, E. D. & Woodhouse, W. W., Jr Fonseca, M. S. (1992). Restoring seagrass systems in the
(1986). Long-term growth and development of United States. In Restoring the Nation's Marine
transplants of the salt marsh grass Spartina alterniflora. Environment, ed. G. W. Thayer, pp. 79-110. College Park,
Estuaries, 9, 63-74.
MD: Maryland Sea Grant College.
 Seagrasses 169

Fonseca, M. S. (1994). A Guide to Planting Seagrasses in the between the seagrasses Thalassia testudinum and
Gulf of Mexico. Galveston, TX: Texas A&M University. Halodule wrightii in Florida Bay. Oikos, 72, 349-358.
Fonseca, M. S. (1996). The role of seagrasses in nearshore Fredette, T. J., Fonseca, M. S., Kenworthy, J. W. & Thayer,
sedimentary processes: a review. In Estuarine Shores: G. W. (1985). Seagrass transplanting: 10 years of U.S.
Hydrological, Geomorphological and Ecological Interactions, Army Corps of Engineers research. In Proceedings of the
eds. C. Roman & K. Nordstrom, pp. 261-286. New York: 12th Annual Conferenceon Wetland Restoration and
John Wiley. Creation, ed. F.J. Webb, pp. 121-134. Tampa, FL:
Fonseca, M. S., Kenworthy, W. J. & Phillips, R. C. (1982). A Hillsborough Community College.
cost-evaluation technique for restoration of seagrass Granger, S. 1., Traber, M. S. & Nixon, S. W. (2000). The
and other plant communities. Environmental influence of planting depth and density on
Conservation, 9, 237-241. germination and development of Zostera marina seeds.
Fonseca, M. S., Thayer, G. W. & Kenworthy, W. J. (1987). In Proceedingsof the 4th International Seagrass Biology
The use of ecological data in the implementation and Workshop, eds. C. Pergent-Martini, M. C. Buia & M.c.
management of seagrass restorations. Florida Marine Gambi, pp. 55-58. Corsica, France:
Research Publication, 42, 175-187. Kenworthy, W. J., Fonseca, M. S., Whitfield, P. E. &
Fonseca, M. S., Kenworthy, W. J., Courtney, F. X. & Hall, Hammerstrom, K. (2000). Experimental manipulation
M. O. (1994). Seagrass planting in the southeastern and analysis of recovery dynamics in physically
United States: methods for accelerating habitat disturbed tropical seagrass communities of North
development. Restoration Ecology, 2, 198-212. America; implications for restoration and management.
Fonseca, M. S., Kenworthy, W. J. & Courtney, F. X. (1996). In Proceedingsof the 4th International Seagrass Biology
Development of planted seagrass beds in Tampa Bay, Workshop, eds. G. Pergent, C. Pergent-Martini, M. C. Buia
Florida, U.S.A. 1: Plant components. Marine Ecology & M. C. Gambi, pp. 385-388. Corsica, France:
Progress Series, 132, 127-139. King, D. M., Bohlen, C. C. & Adler, K. J. (1993). Watershed
Fonseca, M. S., Kenworthy, W. J. & Thayer, G. W. (1998a). Management and Wetland Mitigation: A Frameworkfor
Guidelines for the Conservation and Restoration of Seagrasses Determining Compensation Ratios. University of Maryland
in the United States and Adjacent Waters. Silver Spring, System Draft Report no. UMCEES-CBL-93-098.
MD: U.S. Department of Commerce, National Oceanic Washington, DC: Office of Policy, Planning and
and Atmospheric Administration. Also retrievable in Evaluation.
.pdf format trom: http://shrimp.bea.nmfs.gov/ Lewis, F. G. III (1987). Crustacean macrofauna of seagrass
library/digital.html/ and http://www.cop.noaa.gov and macro algae in Apalachee Bay, Florida, USA.Marine
Fonseca, M. S., Kenworthy, W. J. & Paling, E. (1998b). Biology, 94, 219-229.
Restoring seagrass ecosystems in high disturbance Marba, N. & Duarte, C. M. (1995). Coupling of seagrass
environments. In Ocean Community Conference, 16-19 (Cymodoceanodosa) patch dynamics to subaqueous dune
November 1998. Washington, DC: Marine Technology migration. Journal of Ecology,83, 381-389.
Society. Marba, N., Cebrian, J., Enriquez, S. & Duarte, C. M. (1994).
Fonseca, M. S., Julius, B. E. & Kenworthy, W. J. (2000a). Migration of large-scale subaqueous bedforms
Integrating biology and economics in seagrass measured using seagrass (Cymodocea nodosum)as tracers.
restoration: how much is enough and why? Ecological Limnologyand Oceanography, 39, 126-133.
Engineer, 15, 227-237. Merkel, K. W. (1988). Eelgrass transplanting in South San
Fonseca, M. S., Kenworthy, W. J. & Whitfield, P. E. (2000b). Diego Bay,California. In Proceedingsof the California
Temporal dynamics of seagrass landscapes: a EelgrassSymposium,eds. K.W. Merkel & R. S. Hoffman,
preliminary comparison of chronic and extreme pp. 28-42. National City, CA: Sweetwater River
disturbance events. In Proceedings of the 4th International Press.
Seagrass Biology Workshop, eds. G. Pergent, G. C. Merkel, K.W. (1992).A Field Manualfor Transplantation
Pergent-Martini, M. C. Buia & M. C. Gambi, pp. 373-376. Techniques for the Restorationof PacificCoastEelgrass
Corsica, France: Meadows.National City, CA:Pacific Southwest
Fourqurean,j. W., Powell, G. V. N., Kenworthy, W. J. & Biological Services Inc.
Zieman, J. C. (1995). The effects of long-term Moore, K. H., Wetzel, R. J. & Orth, R. J. (1997). Seasonal
manipulation of nutrient supply on competition pulses of turbidity and their relations to eelgrass
'70 MARK FONSECA ET AL.

(Zostera marina) survival in an estuary. Journal of Seagrass Biology Workshop, eds. G. Pergent,
Experimental Marine Biology and Ecology, 215, 115-134. C. Pergent-Martini, M. C. Buia & M. C. Gambi,
Muelstein, L. K (1989). Perspectives on the wasting disease pp. 131-134. Corsica, France:
of eelgrass Zostera marina. Diseases of Aquatic Organisms, Short, E T. & Wyllie-Echeverria, S. (1996). Natural and
7, 211-221. human-induced diturbance of seagrasses. Environmental
National Oceanic and Atmospheric Administration Conservation, 23, 17-27.
(NOAA), Damage Assessment and Restoration Program Thayer, G. W., Kenworthy, W. J. & Fonseca, M. S. (1984).
(1997a). Habitat Equivalency Analysis: An Overview. Policy The ecology of eelgrass meadows of the Atlantic coast:
and Technical Paper no. 95-1. Silver Spring, MD: NOAA a community profile. U.S. Fish and Wildlife Service,
Damage Assessment Center. FWSjOBS-84j02. Washington, DC: US Department of the
National Oceanic and Atmospheric Administration Interior.
(NOAA), Damage Assessment and Restoration Program Thorhaug, A (1974). Transplantation of the seagrass
(1997b). Scaling Compensatory Restoration Actions, Guidance Thalassia testudinum Konig. Aquaculture, 4, 177-183.
Document for Natural Resource Damage Assessment Under the Tomasko, D. A, Dawes, C. J. & Hall, M. O. (1991). Effects of
Oil Pollution Act of 1990. Silver Spring, MD: NOAA the number of short shoots and presence of the
Damage Assessment Center. rhizome apical meristem on the survival and growth of
Orth, R. J., Harwell, M. c., Bailey, E. M., Bartholomew, transplanted seagrass Thalassia testudinum. Contributions
A, Jawad, J. T., Lombana, A v., Moore, K A, Rhode, in Marine Science, 32, 41-48.
J. M. & Woods, H. E. (2000). A review of issues in Unsworth, R. E. & Bishop, R. C. (1994). Assessing natural
seagrass seed dormancy and germination: implications resource damages using environmental annuities.
for conservation and restoration. Marine Ecology Progress Ecological Economics, 11, 35-41.
Series, 200, 277-288. Virnstein, R. W. & Morris, L. J. (1996). Seagrass Preservation
Phillips, R. C. & Menez, E. G. (1988). Seagrasses. and Restoration: A Diagnostic Plan for the Indian River
Smithsonian Contribution to the Marine Sciences Lagoon. Technical Memorandum no. 14. Palatka, FL:
no. 34. Washington, DC: Smithsonian Institution Press. St Johns River Water Management District.
Preen, A, Lee-Long, W. J. & Coles, R. G. (1995). Flood and Whitfield, P. E., Kenworthy, W. J., Fonseca, M. S. &
cyclone related loss, and partial recovery of more than Hammerstrom, K (in press). Role of storms in the
1000 km2 of seagrass in Hervey Bay, Queensland expansion and propagation of disturbances initiated by
Australia. Aquatic Botany, 52, 3-17. motor vessels in subtropical seagrass beds. Journal of
Robilliard, G. A & Porter, P. E. (1976). Transplantation of Coastal Research.

eelgrass (Zostera marina) in San Diego Bay. San Diego, CA: Williams, S. L. & Orth, R. J. (1998). Genetic diversity and
Undersea Sciences Department, Naval Undersea Center. structure of natural and transplanted eelgrass
Robblee, M. B., Barber, T. R., Carlson, P. R., Durako, M. J., populations in the Chesapeake and Chincoteague Bays.
Fourqurean, J. W., Muehlstein, L. K, Porter, D., Yarbro, L. Estuaries, 21, 118-128.
A., Zieman, R. T. & Zieman, J. C. (1991). Mass mortality Wood, R. J. E, Odum, W. E. & Zieman, J. C. (1969).
of the tropical seagrass Thalassia testudinum in Florida Influence of seagrasses on the productivity of coastal
Bay (USA). Marine Ecology Progress Series, 71, 297-299. lagoons. In Memorias de Symposium Internacional Lagunas
Rose, C. D., Sharp, W. c., Kenworthy, W. J., Hunt, J. H., Costeras,eds. A Castanares, A Ayala & E B. Phleger,
Lyons, W. G., Prager, E. J., Valentine, J. E, Hall, M. 0., pp. 495-502. Mexico, DF: UNAM-UNESCO.
Whitfield, P. & Fourqurean, J. W. (1999). Sea urchin Wyllie-Echeverria, S., Arzel, P. & Cox, P. A (2000). Seagrass
overgrazing of a large seagrass bed in outer Florida Bay. conservation: lessons from ethnobotany. Pacific
Marine Ecology Progress Series, 190, 211-222. ConservationBiolology, 5, 329-335.
Sheridan, P. (1999). Trajectory for structural equivalence Zedler, J. B. & Langis, R. (1991). Comparisons of
of restored and natural Halodule wrightii beds in Texas. constructed and natural salt marshes of San Diego Bay.
Gulf ResearchReports, 10, 81-82. Restoration and Management Notes, 9, 21-25.
Short, E T. & Short, C. A (2000). IdentifYing seagrass Zieman, J. C. (1982). The ecologyof seagrassesof south Florida:
growth forms for leaf and rhizome marldng a community profile. us Fisheries and Wildlife Service.
applications. In Proceedingsof the 4th International Washington, DC: US Department of the Interior.