19961054 Ver 2_More Info Received_20110712

Maher, Niki
From: Mark Fonseca [[e-mail]]
Sent: Tuesday, July 12, 2011 2:44 PM
To: Maher, Niki
Subject: Responding to your SAV request
Attachments: 2011 FINAL REPRINT fonseca_Addy revisited.pdf; Integrating biology and economics in
seagrass restoration 2000.pdf; Rest Hndbk_Chp7_reduced.pdf; Mark_Fonseca.vcf
Nice talking with you, Niki. Hope you don't bake too much up there in Raleigh. It is 72oF here today, light
breeze, drinks being served on the patio, etc etc.
Anyway - first document is the review paper i mentioned; just got pdf today (2011 Final Reprint..pdf....).
focus on the abstract - you'll get the point about what constitutes 'success' and the rather warped way it is
measured.
the 2nd document, Rest Hndbk_Chp7.pdf with the informative title of "Seagrasses" has the detailed breakout
of costs you were looking for on pg 167 (note USD 1996, values).
Fonseca, M.S., Kenworthy, W.J. Julius, B.J. Fluke, S., Shutler, S. 2002 Seagrasses, p 149-170 In A.J.
Davy and M. Perrow (eds.) Handbook of Ecological Restoration. Vol 2. Chp 7. Cambridge Univ. Press.
third document - see this web link for the national guidelines document:
http://www.col2.noaa.gov/pubs/das/das l 2.pdf
• Fonseca, M.S., W.J. Kenworthy, GV.!tfidyer 1998 Guidelines for mitigation and restoration of
seagrass in the United States and adjacent waters. NOAA COP/Decision Analysis Series. 222p
We have lots of papers on this subject - but these are good entry points - we also have extensive experience with
preparing this kind of information for assisting NOAA General Counsel and the Department of Justice in claims
cases; depositions, expert reports, federal court trial material, etc. if you find yourself going down that road.
Note that our valuation method is called HEA; or habitat equivalency analysis. The short on this is that no one
is going to successfully argue that seagrass meadows are not valuable - bad decision because they would be
overwhelmed with admissible information that would make them look very, very stupid. So - given that point is
universally conceded, we develop costs based on tlaq ?W to restore the injury (balance lost services with
restored services, discounted over time, etc.).
4th optional document with more on how discounting, etc., works: Integrating biology and economics..pdf.
Fonseca, M.S., B.E. Julius, W.J. Kenworthy. 2000. Integrating biology and economics in seagrass
restoration: how much is enough and why? Ecol. Engineer. 15:227-237
Anyway, let me know if you have any questip.,; ?,azn confident I have cured any insomnia issues you may
have.
Best
Mark
NOAA's Beaufort Laboratory
Mark S. Fonseca, Ph.D., Chief
Applied Ecology and Restoration Research Branch
Center for Coastal Fisheries and Habitat Research
National Centers for Coastal Ocean Science, NOS, NOAA
101 Pivers Island Road
Beaufort, North Carolina 28516-9722
USA
[phone] voice
[phone] fax
[phone] cell
Website: http://www.ccfhr.noaa.gov/
----------------------------------------- ><>
2
Addy Revisited: What Has Changed with
Seagrass Restoration in 64 Years?
Mark S. Fonseca
ABSTRACT
A brief appraisal of the present state of seagrass restpration jn the context of the 64-year-old seminal publication by C.E.
Addy reveals that early observations were prescient;and have remained the basis for our collective attempts to conduct
open system seagrass restoration. Our ability to:ensure I restoration success remains limited. A flawed philosophical frame-
work for choosing restoration, frequently exacerbated by management inexperience and failure to apply known standards
for site selection, continues to plague the process. Moreover, seagrass restoration has become an on-demand attempt to
overcome hysteresis and shift a habitat from one stable state (unvegetated) to what is arguably a more complex stable
state (vegetated) by artificial colonization methods. These methods are frequently overwhelmed by natural processes that
ordinarily rely on orders of magnitude more propagules and years of recruitment classes. As a result, the expectations for
successful seagrass restoration, like most wild community restoration projects, are often unrealistic and improperly held
to an even higher standard than agricultural crops. Limited attention to project monitoring, lack of practical recovery
metrics, and, in some cases, limited scientific knowledge of seagrass (e.g., population ecology, genomics, and landscape
dynamics) limit our capability to generate quantitative guidance and realistic expectations. Improving the probability of
successful restoration depends not so much on overcoming technical transplanting issues, but on avoiding injudicious
reliance on restoration to solve higher-level resource mana jement issues. Overall, conservation of seagrass remains a
more reasonable alternative than restoration, and Addy is revealed as a keen observer of the field.
Keywords: history, methods, restoration success, seagrass
I -N Jespite a number of reviews
regarding seagrass restoration
(Phillips 1960, Fonseca 1992, 1994,
Fonseca et al. 1998, Calumpong and
Fonseca 2001, Short et al. 2002a,
Pickerell et al. 2005, 2006, Orth et
al. 2009, Paling et al. 2009), there are
few new, unreported dimensions of
the practice that carry an expectation
of improved outcomes in the years
since Addy's (1947a) seminal paper
on the subject. However, it may be
just as interesting to note what has
and has not changed during this time,
and why, rather than reiterating the
pitfalls of habitat restoration (sensu
Race and Christie 1982, Race and
Fonseca 1996).
In 1947, C.E. Addy, a biolo-
gist with the U.S. Fish & Wildlife
Service, published a two-page paper
in the Maryland Conservationist (a
quarterly) titled "Eel grass planting
guide." Addy's paper (1947a) was
prompted by attempts to facilitate the
recovery of eelgrass (Zostera marina)
after the pr6clpitous decline of the
1930s- `(Cotten 194, Renn 1934,
Addy ,pcf l ylwarc?` 1944), ascribed
to a slime mold infection (Short et al.
1987, 1988). Addy's short paper had
a number of interesting comments
regarding early efforts to transplant
eelgrass using both vegetative shoots
and seeds:
Theridea` of..ttansplanting eelgrass is
noCmew and has;. been tried many
times under the auspices of the old
Biological Survey and in more recent
years by the Fish and Wildlife Ser-
vice. Attempts were even made by
the service to transplant seeds and
runners of the West Coast variety.
In all numerous attempts were made
to re-establish eelgrass in local areas
from Massachusetts to North Caro-
lina by the Government, private
agencies and clubs. No notable
achievements were made in these
plantings and many of them failed
completely. (p. 17)
Ecological Restoration Vol. 29, Nos. 1-2, 2011
ISSN 1522-4740 E-ISSN 1543-4079
©2011 by the Board of Regents of the
University of Wisconsin System.
Unfortunately, detailed information
on these projects appears to be lost in
obscurity (but see limited reference in
Addy and Johnson 1948 and Cottam
and Munroe 1954). Addy (1947a) and
Addy and Aylward (1944), however,
recounted how the natural recovery
process was apparently occurring quite
vigorously by 1943 in many areas that
had been denuded during the 1930s.
From Addy's descriptions, it appeared
that there were persistent (or perhaps
isolated) strains, apparently unafHicted
pockets of eelgrass that were supplying
seed to create small, isolated patches
March/June 2011 ECOLOGICAL RES7'ORATION 29:1-2 4 73
j<
? ? l ii : 9
so
Other da;urrtents
Prier-rnv~d dnc.umnnis r?\Q•.
h? ?`
?i .. , ?
25 __
III 1,1Z
7
(
\ '
\0

W
D 20

M
5
s

i
j
I
195h 14G t9(S 'R?U 1615 19&} '9115 t99U "911 11%#I 2G75-• JG'(?
YEAR N4IDPOINT
Figure 1. Historical trend in seagrass restoration literature for peer-reviewed (white) and all
other (gray) documents. Records from before about 1998 (dotted line) are based on the literature
reviewed by Fonseca and others (1998) that included project reports and obscure printing
venues; more recent records are from a Web of Science search of peer-reviewed literrature using
the term "seagrass restoration."
(pers. obs.; the late Mel Durfee, a
fisherman then living on Nanaqua-
ket Pond in Rhode Island showed me
eelgrass patches in the late 1970s that
he claimed had persisted through the
blight of the 1930s, providing anec-
dotal evidence that some relicts of the
former distribution did indeed remain
to form founder groups). Addy's obser-
vations of these dynamics apparently
led him to a series of experiments and
recommendations of how to accelerate
the eelgrass recovery.
Admittedly, attention to seagrass
restoration has not been consistent
in the intervening decades (Figure 1).
Interest lapsed from the 1940s until
the late 1970s and then enjoyed a
sustained research and application
posture only since the late 1980s (see
reviews by Fonseca et al. 1998; Paling
et al. 2009, van Katwijk et al. 2009).
Here, I review which aspects of sea-
grass restoration were evident early
on and how those rather prescient
observations have been rediscovered
and amplified in recent times.
Addy: Then and Now
Transplanting Stock
Addy recommended limiting harvest
of vegetative stock to extensive beds
(one acre or more. [? 4,000 m2]) a
practice continued today so as to
reduce impacts to isolated populations
(but see Fonseca 1994, where, for
efficiency purposes, removal of entire
small patches in high-energy areas was
recommended). Addy also speculated
that sods may be kept longer than a
day, if kept cool and moist-another
practice unchanged today. In addition,
he advised harvesting as close to the
planting site as possible (thus allowing
type-matching).
The last recommendation is not
consistent with Williams atad•'Orth
(1998), who assert that maintenance
of the natural population's genetic
structure could be better attained with
harvest from larger spatial areas. Con-
versely, it has been argued that use of
local plants could be beneficial owing
74 4 March/June 2011 ECOLOGICAL RESTORAT ION 29:1-2
to the selection of locally adapted gene
complexes (Hammerli and Reusch
2002). Incorporating genetic diversity
considerations in seagrass restoration
remains difficult on a practical basis,
not only due to costs, available exper-
tise, and rapid evolving technology,
but also due to our lack of understand-
ing of the role of genetics in seagrass
bed maintenance and resiliency (e.g.,
van Dijk et al. 2009). However, the
original intent of Addy to select simi-
lar "types" or locally adapted plants
still is supported by current under-
standing of plant genetics (Williams
and Orth 1998, Hughes and Stacho-
wicz 2004, Ruesch 2006, Procaccini et
al. 2007, Hughes et al. 2009). There-
fore, while Addy's guidance was based
almost exclusively on logistics, those
protocols seems to mostly fit well, if
serendipitously, with current thinking
on seagrass population ecology and
genomics.
Areas to Be Planted
Addy suggested choosing locations
that would be useful to waterfowl, but
given the potential devastating effect
of waterfowl on eelgrass plantings
(pers. obs. and review by Valentine
and Duffy 2006) it is likely that such
a practice would be counterproduc-
tive to successful restoration. How-
ever, Addy's recommendations about
bottom type ("similar to that from
which transplants were taken" and "of
mud containing a moderate amount
of sand" p. 16) are still core elements
of site selection (Phillips 1960, Fon-
seca et al. 1998, Short et al. 2002a);
unfortunately, these recommenda-
tions are used in only -9% of studies
(Fonseca et al. 1998, 38).
Addy's suggestions about elevation
are also mirrored in current guidance
to choose depths similar to those
observed in nearby, healthy beds (Fon-
seca et al. 1998, Short et al. 2002a).
Nonetheless quantitative determi-
nation of the appropriate elevation
(or depth) for planting remains an
elusive goal. It is likely that this will
quickly change as global positioning
systems (GPS)-based techniques for

obtaining high-resolution vertical
measures become more affordable and
less complicated and are combined
with precise tidal and vertical datum.
Presently, site selection errors con-
stitute the single greatest challenge in
the restoration process, despite Addy's
intuitive guidance (Fonseca, pers. obs.).
Many of these errors could be avoided
if the plant's growth strategy were
considered. Some species spread faster
than others and can be augmented by
simple facilitation techniques; thus,
uncomplicated calculations of planting
density and recovery projections can
and should be made. When site selec-
tion has been effectively evaluated and
implemented, better success has been
achieved (Paling et al. 2009). This was
not a pressing issue for Addy, as eelgrass
is a seagrass species possessing many
attributes that make it amenable to
transplantation: it has vigorous sexual
and asexual reproduction, possesses
only apical shoots, responds well to
changes in sediment elevation, inhab-
its various habitat types from silt to
cobble sediments, and so on (Fonseca
et al. 1998).
Few generic site evaluation tools
have been developed in the interven-
ing decades. Short and others' (2002a)
Preliminary Transplant Suitability
Index (PTSI) and Transplant Suitabil-
ity Index (TSI) are among the first for
systematic site evaluation since Phil-
lips's work in the early 1980s. Even
those indices, however, do not repre-
sent a dramatic departure from Addy's
guiding principles. The development
of similar indices for the subtropi-
cal species is another prime area for
research.
Since Addy's time, issues surround-
ing seagrass restoration have expanded
to include a variety of other consider-
ations, including the need to quantify
restoration in ways that meet emerg-
ing policies guiding restoration, such
as comparative (among restored and
injured sites) function and extent. For
seagrass in general, adherence to some
simple metrics for site selection can be
utilized (sensu Fonseca et al. 1998).
Sites should have the following:
• Dgp'th that is similar to nearby nat-
ural beds
• Anthropogenic (as opposed to natu-
ral) disturbance
• Sufficient acreage to achieve goals
• Similar characteristics as other suc-
cessfully restored sites
• Similar quality habitat as that which
was lost
Sites should not be the following:
• Subject to chronic storm disturbance
(discussed below)
• Undergoing rapid and extensive nat-
ural recolonization
• Among patches of existing seagrass
Equipment
Little has changed with regard to the
kind of equipment needed to trans-
plantseagrasses:iSmall.boats, manually
operated extraction. tools (i.e., shov-
els),- and wetted burlap (for moist-
ening and cooling of plants out of
the water) are all still commonly used
today (but see mechanization, under
Methods, below).
Time of Planting
Addy's' 'reeoir rhendation of spring
plaritlAgs off vegetative sprigs of eel-
grass also remains applicable (Fonseca
et al. 1998). Even casual observers of
eelgrass reproductive ecology would be
able to deduce this timing. However,
exceptions to planting times are often
made to avoid periods of disturbance,
such as heat or cold stress, biotur-
bation (biologically mediated distur-
bance), or storm seasons. For example,
eelgrass planting south of Chesapeake
Bay should,4,.,performed in the fall,
not spring ?(Fp`hseca,, et al, 1998) in
orderso acl ieve,the,maximum growth
before,the annual stress period, which
is summer in this region. Another
exception is made in Chesapeake Bay
for seed planting methods, where
seeds are typically spread in the fall
of the year. Orth and others (1994,
2009) found that for this Mid-Atlantic
area theyres,F,nee of seeds prior to the
sprlrlgiperminarjon period was impor-
tant for success. Planting times for
other regions and species may also
vary from Addy's norm (Fonseca et
al. 1998).
Methods
Use of shovel-size sods (planting with
sediment attached) remains a mainstay
of vegetative transplant techniques.
Nearly 30% of the eelgrass transplant
projects reported as of 1995 used an
attached sediment method (Fonseca
et al. 1998, 37). Nonetheless, adop-
tion of sediment-free methods (plants
only) that dispense with the burden
of moving large masses of sand and
mud have been widely used with good
results (Fonseca et al. 1998). The most
recent, significant progress with this
approach has been the use of plant-
ing frames. With this method, shoots
are attached to heavy metal frames
and dropped overboard (Short et al.
2002b). Both eelgrass and shoalgrass
(Halodule wrightii) have been used in
this approach, and the plants soon
become rooted and independent of
the frames, allowing recovery and
reuse of the frames. The advent of
mechanized techniques (Fonseca et
al. 1998, Paling et al. 2001, Traber
et al. 2003, Fishman et al. 2004, Bell
et al. 2008, Orth et al. 2009, Uhrin
et al. 2009) is the first major shift in
technology for vegetative plantings
since transplanting began. However,
mechanized techniques remain in
the development stage for most areas
(Uhrin et al. 2009).
Addy recommended harvesting
and sowing eelgrass seed in the fall
and spring, respectively; however, the
effectiveness of this technique "has
not been fully determined" (Addy
1947a, 16). Addy's quiet enthusiasm
for this technique was reinforced by
observations that "natural reproduc-
tion from seed is taking place rap-
idly and on a large scale," as well as
by Addy's other work from that time
(Addy 1947b). Nonetheless, he rec-
ommended use of both sexual and
asexual propagules. These sections by
Addy (1947a) provide information
about sowing seeds that was largely
ignored for decades. In the late 1990s,
March/June 2011 ECOLOGICAL RESTORATION 29:1-2 tO 75

Granger and others (2002) and Orth
and others (2000) began experiment-
ing with the technique (also see Tho-
rhaug 1974, Churchill et al. 1978).
Addy's sequence of harvesting flow-
ering shoots and storing them until
seeds are released, followed by over-
winter storage of the seed, is funda-
mentally the same approach that Orth
and others (2003, 2009), Granger
and others (2002), and Pickerell and
others (2005) have used effectively.
In summary, Addy's (1947a) paper
dealt primarily with a half-dozen
issues that are still considered today:
transplanting stock, site selection,
equipment, planting method, propa-
gation, and planting by seed. Addy
implied that in general, substantial
experimentation in eelgrass restoration
had been underway by 1947. Unfor-
tunately, the detailed transplanting
methodology used by Addy was not
followed by substantive information
on the success of the technique. The
absence of monitoring data raises the
possibility that the protocols Addy so
clearly reported were in part theorized,
rather than derived from field applica-
tion and observations. If so, this would
constitute a comingling of observation
and practice, leading to speculation
that Addy was drawing logical, yet
untested, conclusions. Clarification
of this point may be lost to history.
Restoration Success-
Unrealistic Expectations?
Addy's work was generally optimis-
tic, but new challenges have emerged.
New resource management practitio-
ners who have only recently entered
into their positions often have unre-
alistic expectations of seagrass restora-
tion outcomes. They often lack direct
project experience, and some have
not received appropriate training. As
a result, their philosophical framework
for utilizing habitat restoration is often
deficient, and their expectations for
success are far higher than are justified
by data. For example, Fonseca and
others (1998) conducted a survey of
most of the extant projects at the time
and found that successful establish-
ment of seagrass cover occurred in
<50% of the projects-a classic "glass
half-full, glass half empty" scenario
for characterizing success. Given;these
data, it is unrealistic to expect'letter
results from any gi'ven' seagrass-?'esto-
ration project, and Aadys'cla m that
eelgrass transplanting was an effective
accelerant of the recovery process must
be viewed in this context.
From a natural recovery stand-
point, successful recruitment and
maintenance of seagrass 40% of the
time would be a highly significant
event. In contrast, 40% is considered
a low success rate when attempting
to offset human impacts via restora-
tion projects. Another, perhaps more
useful perspective iiiiy bo gained by
comparing seagrass re'stoi tion results
with marketplace speculation on crop
futures. Despite our collective millen-
nia of terrestrial agricultural practice
on landscapes where we can exert con-
trol over a wide variety of ecological
factors, it is still a profitable exercise
to speculate on crop success iii any
given year. Given our' limited knowl-
edge of seagrasses (s' c'om'pared to
crops) and the lack ofco'ntrols that we
can exert on wild, open systems like
seagrass restoration sites, we should
not be surprised at moderate success
(defined as persistent target acreage)
rates for seagrass planting. From this
perspective, 50% begins to look like
the glass is half full, not half empty.
However, early controversial results of
large-scale restoration attempts (e.g.,
Stein 1984) contributed not only to
lowering the statistical mean success
rate but also discouraging regulatory
personnel from attempting seagrass
restoration (Fonseca, pers. obs.). All
of this begs the question whether cur-
rent permit requirements (which are
often highly variable among projects
worldwide; Fonseca, pers. obs.) con-
stitute an appropriate performance
level. Although beyond the scope of
this limited review, there are indica-
tions that seagrass restoration suc-
cess is on the rise (sensu Paling et al.
2009).
76 4 March/June 2011 ECOLOGICAL RESTORATION 29:1-2
New Issues-Or Old
Issues Redefined
Community-Based Restoration
The recent trend (within the last
decade in particular) of mobilizing
local communities to conduct seagrass
restoration deserves some attention.
Although community-based efforts
can be successful for marshes (Curtin
et al. 2008) and land-based portions of
seagrass planting operations (Short et
al. 2002a), there are both positive and
negative aspects of community-based
restoration (CBR). On the positive
side, CBR educates young people to
become environmental stewards with a
conservation-minded lifestyle. It raises
public awareness of the importance of
seagrasses, which creates the potential
to change zoning to protect seagrasses
and limit development projects that
adversely impact seagrasses.
On the negative side, CBR gen-
erally lacks project monitoring, thus
giving a false sense of success. On the
other hand, perceived or actual limited
success could create the impression
that restoration is a lost cause. Finally,
there is little evidence that it has actu-
ally reduced any negative impacts on
seagrasses. Absence of critical proj-
ect evaluation may indicate more
outcome-based actions are needed.
Disturbance, New Stable
States, and Hysteresis
One of the key issues for present-day
seagrass restoration not mentioned by
Addy is disturbance. Disturbance of
planted seagrass beds particularly by
animals is emerging as a key factor in
limiting the success of seagrass resto-
ration projects as well as the natural
spread of seagrasses into historical dis-
tributions. The role of bioturbation in
limiting natural seagrass recruitment
to these sites as well as establishment
of planting under restoration is well
documented (Orth 1975, Suchanek
1983, Harrison 1987, Merkel 1988,
Valentine and Heck 1991, Fonseca et
al. 1994, Philippart 1994, Valentine
et al. 1994, Townsend and Fonseca
J
Figure 2. Low tide at Crown Beach, Alameda, California, in 2004. The intertidal portion of this site is heavily colonized starting in the late spring
(March, left) by what appears to be a genetically distinct, annual form of eelgrass (Zostera marina, sensu Talbot et al. 2004). Note the excavations
that indicate frequent disturbance of the site by rays that feed on the site at high tide. The frame standing upright (at left) with the survey team
is 1 m2. In August (right), the eelgrass seedlings have extensively colonized and grown to about 15 cm diameter patches since the cessation of ray
disturbance. Photos by C. Addison
1998). Bioturbation sometimes com-
pletely eliminates plantings in very
short periods of time (hours to days):
• Callianassid shrimp (Callianassa
spp.): west Florida shelf (Fonseca et
al. 2008)
• Crabs: New Hampshire (Davis et al.
1998), New York (seed predation)
(Wigand and Churchill 1988)
• European green crab (Carcinus
maenas): Rhode Island (pers, obs.)
• Ducks: North Carolina (pers. obs.)
• Fish herbivory: Florida Keys (pers.
obs.), Setubal, Portugal (pers. obs.)
• Lugworms: Canada (Harrison 1987),
Great Britain (Philippart 1994)
• Rays: Florida (pers. obs.), North Car-
olina (Townsend and Fonseca 1998),
San Francisco Bay (pers. obs.), Tampa
Bay (Fonseca et al. 1994), Virginia
(Orth 1975)
• Sting rays: San Diego Bay (Merkel
1988)
• Urchins: Florida (Camp et al. 1973,
Valentine and Heck 1991, Valentine
et al. 1994)
Dudgeon 2004). If an unvegetated
state is stable, especially in the pres-
ence of bioturbation, a prolonged
effort to introduce and maintain
seagrass may be required in order to
overcome system hysteresis and shift
the system back to a sustainable, stable
state of seagrass habitat. Such an effort
may be beyond the scope of many res-
toration projects. The hysteresis of the
unvegetated,, dition or, state is, ;4
course;i t]lm' itedkw,, sites influenced
by bio,ttrbat ' Even without biotur-
bation many other factors (e.g., storm
waves that cause sediment movement
and resuspension) will, in the absence
of the seagrass bed, give rise to light-
limiting conditions and physical insta-
bility; these factors can make the sea-
grass recolonization process difficult
and pffild'h`g fl e°unvegetated state.
In essence, seagrass restoration is a
calculated attempt, to shift a habitat
from one stable state (unvegetated)
to what is arguably a more complex,
higher-energy stable state (vegetated).
The artificial colonization methods
that are used to achieve the vegetated
state are dwarfed by natural processes
that typically rely on orders of mag-
nitude more propagules and years
of recruitment classes in order to be
"successful." This makes restoration
of these open, uncontrolled systems
especially challenging.
Disturbance might even influence
the selection of local genetic strains of
seagrass. At one site in San Francisco
Bay, a broad intertidal flat is success-
fully colonized only by a genetically
distinct annual form of eelgrass, even
though biennial forms exist only meters
away, below a 50 cm shelf (Talbot et al.
2004). The elevation and tidal regime
of this site is well within the vertical
range of other biennial beds around
the Bay (Fonseca et al., unpubl. data).
It is possible that seasonally extreme
physical disturbance by stingrays dic-
tates that only an annual life history
strategy can survive on this site, as eel-
grass seeds do not germinate until early
spring, after the peak ray disturbance
period (Figure 2). Thus, bioturbation
on this flat may be creating a situa-
tion where only the annual life history
strategy can withstand the periodic
disturbance.
Concerns not addressed (or per-
haps not present) in Addy's time are
now emerging regarding the influ-
ence of bioturbation, both on natural
It seems that bioturbation is a key
concern in restoration efforts because
it can create and sustain a new, unveg-
etated, stable state (sensu Petraitis and
M re / tine 2011 ECOLOGICAL RESTORATION 29:1-2 4 77

seagrasses and restored seagrass beds,
as animal populations adjust to human
pressures. It is possible that bioturba-
tion has significantly increased since
1947. Are the numbers of bioturba-
tors such as stingrays on the rise due
to overfishing of key predators (Myers
et al. 2007), or is there a prolifera-
tion of herbivores, especially in marine
protected areas (Ferrari et al. 2008)?
If so, then restoration practitioners
may need to include temporary bio-
turbation exclusion devices more often
(sensu Fonseca et al. 1994). However,
most of these devices are labor inten-
sive to construct, place, and main-
tain, and they may pose problems
for regulatory agencies (for example,
U.S. Army Corps of Engineers) who
oversee the permit process for place-
ment of structures; this is particularly
true when an agency is trying to avoid
conflict with people using mobile fish-
ing gear and simultaneously prevent
harm to wildlife. A significant amount
of work is needed to develop cost-
effective means for inhibiting grazing
on seagrass transplants and physical
disruption in general.
Introduction of nonindigenous spe-
cies may also pose a new challenge to
recolonization of historically vegetated
areas. One example is the presence of
the European green crab in the north-
western Atlantic and northeastern
Pacific (Davis et al. 1998).
Vessel Groundings
Unlike Addy's time when the number
of motor vessels was comparatively
few, escalating numbers of boats oper-
ating in the coastal zone have led to a
rapid increase in boating-related inju-
ries to seagrass beds (South Florida
Natural Resources Center 2008), as
well as programs to mitigate these
impacts (Kirsch et al. 2005). This has
become a significant regional manage-
ment dilemma, particularly in areas
with slow-recovering species such as
turtlegrass (7balassia testudinum).
Because injuries to slow-spreading sea-
grass like turtlegrass persist for many
years, cumulative impacts are elevated
and become especially evident when
new injuries occur in relatively short
time spans. With short injury return
intervals, even seagrasses such as eel-
grass, which has comparatively high
recovery rates, can be overwhelmed
by vessel injuries (Ortl e't!'4l.16k).
Final Observations
and Comments
In revisiting Addy's paper I was struck
by the confluence of those observa-
tions and today's practices. Another
way of saying this. is that, while many
improvements have been 'made iri sea-
grass restoration?rechiiol'ogy, in many
instances they remain grounded on
observations made over six decades
ago. This is not unlike the conclu-
sions we reached on wetland resto-
ration in general, also on a decadal
review schedule (Race and Fonseca
1996, commenting particularly on
Race and Christie 1982).
Perhaps the most problematic and
challenging issue with seagrass resto-
ration is the lack d'f adequate prepara-
tion and subject knoMedge 'Byj `ttrany
involved in the devef6p''neni"o' f res-
toration projects. Too often, little or
no attention is given to review of the
available literature before and during
project planning. This may be exacer-
bated by the participation of personnel
who are unfamiliar with seagrasses or
their restoration. In addition, many
important findings are not published
at all, or information is not ptoJided in
easily accessible forrnarsAt is extremely
important that more results be pub-
lished in journals, so that they are
critically reviewed and evaluated by
peers and more broadly disseminated
to other practitioners. The combina-
tion of lack of adequate knowledge on
the part of some involved in restoration
project planning and implementation,
along with inaccessibility of documents
that are often rich in practical infor-
mation, has produced.a tremo dous
amount of low-quality.- restoration
projects and some redundant work
(Fonseca et al. 1998).
My personal observation is that
the significant issues of seagrass
78 4 March/June 2011 ECOLOGICAL RESTORA710N 29:1-2
S'
restoration are often not involved
with installing plants; there are many
viable techniques (see review: Fon-
seca et al. 1998). Rather, the selec-
tion of methods and protocols is often
a market-based issue where various
constraints (i.e., regulatory author-
ity, bidding competition, local skills,
environmental setting, availability of
seagrass species and funds) often inter-
act and perhaps compete to produce a
restoration plan that is a compromise
and often unrealistic at best. Perhaps
the most recurring failure in seagrass
restoration plans is the site selection
process, where project plans call for
off site planting locations that fail on
multiple points (see above list under
"Areas to be planted"). Thus, to the
uninitiated, this process frequently
produces the impression that seagrass
restoration is "experimental."
Certainly, the high cost of a well-
planned and appropriately moni-
tored project could give the appear-
ance of a research project, but risk
(50% of seagrass restoration projects
met their goals; Fonseca et al. 1998)
should not be confused with experi-
mentation. While experimentation is
always welcomed to improve success
and reduce costs, that does not mean
that all seagrass restoration is experi-
mental. That is, seagrass restoration
is a proven positive alternative to no
action at all. Experimentation with
adjustments to methodology is typi-
cal for any agricultural practice; and,
given the comparative lack of site con-
trol in open system restoration (as is
the case with seagrass), adaptation to
local conditions and logistics is logical.
The distinction is that experimental
manipulation is unnecessary for effec-
tive seagrass restoration, because there
are proven, standard methods (e.g.,
Montin and Dennis 2003). Rather, as
stated earlier, the expectations for sea-
grass restoration are grossly unrealistic,
because it is frequently being held to
much higher standards compared to
agricultural crops. This leaves us with
the well-known adage that it is far less
expensive to conserve a seagrass bed
than it is to try to restore it.
In summary, there are several issues
that need focused consideration:
• Validation. Useful (informative) mon-
itoring data must be obtained for
making midcourse corrections (i.e.,
adaptive project management).
• Site selection. Existing published guid-
ance must be used to avoid choosing
poor sites. (Fonseca et al. 1998, Short
et al. 2002a; see site criteria under
"Areas to be planted," above),
• Technologies. Appropriate, cost-effec-
tive restoration technologies that cap-
italize on seagrass reproductive ecol-
ogy should be used; seagrass popu-
lation ecology should be fully uti-
lized to enhance project design and
efficiency.
• Knowledge base. The lack of scientific
rigor in investigating past work and
inadequate information dissemina-
tion results in redundancy of effort
and hampers progress.
The intent of this review was to
stress that seagrass restoration is a
viable, but not simplistic, manage-
ment option. Outcomes of restora-
tion projects depend to a great extent
on heeding past lessons, operational
details, and personal experience
(sensu Phillips 1960). However, all
the shortcomings discussed above can
be resolved through rededication to
science-based project management
and a decision process that, in reality,
is easily achievable. It is essentially the
blueprint put forth by Addy over 60
years ago.
Acknowledgments
I would like to thank my colleagues at the
Center for Coastal Fisheries and Habitat
Research, Christine Addison, Jud Kenwor-
thy and Amy Uhrin for their long collabo-
ration and discussions that strongly influ-
ence my thoughts on the subject of seagrass
restoration. I also thank the guest editors
of the special volume, Dr. Kay McGraw
and Dr. Ron Thom, for their patience, and
the four peer reviewers together with the
associate editor Chris Reyes for providing
valuable comments and suggestions for
improving this article and Meg Hannah for
shepherding it to a final product.
References
Addy, C.E. 1947a. Eel grass plant-
ing guide. Maryland Conservationist
24:16-17.
_. 1947b: Germination of eelgrass
seed. Journal of Wildlife Management
11:279.
Addy, C.E., D.A. Aylward. 1944. Status
of eelgrass in Massachusetts during
1944. Journal of Wildlife Management
8:269-275.
Addy, C.E. and R.H. Johnson. 1948.
Status of eelgrass along the Atlan-
tic coast during 1947. Pages 73-78
in Proceedings of the 1948 Northeast
Game Conference.
Bell, S.S., A. Tewfik, M.O. Hall and M.S.
Fonseca. 2008. Evaluation of seagrass
planting and monitoring techniques:
Implications for assessing restora-
tion success and habitat equivalency.
Restoration Ecology 16:1-10.
Calumpong, H.P. and M.S. Fonseca.
2001. Seagrass transplantation. Pages
425-444 in F.T. Short, R.G. Coles
and C.A. Short (eds), Global Sea-
grass Research Methods. Amsterdam:
Elsevier.
Camp, D.K.1,S.P. Cobb and J. Van-
Breedveld.:1-9,7, 3. Qvergrazing of sea-
gra?ses by a regular,urchin Lytechinus
variegatus. Bioscience 23:37-38.
Churchill, A.C., A.E. Cok and M.I.
Riner. 1978. Stabilization of sub-
tidal sediments by the transplanta-
tion of the seagrass Zostera marina L.
New York Sea Grant Report Series.
NYSSGP-RS-78-15.
Cottam, C. 1934, Past periods of eelgrass
scarcity.,Rb,odora 36:261-264.
Cotom,, C. and D.A. Munro. 1954. Eel-
grass status and environmental rela-
tions. Journal of Wildlife Management
18:449-460.
Curtin, C.A., P.C. Delano and L.M.
Valdes-Weaver. 2008. Utilization of a
citizen monitoring protocol to assess
the structure and function of natural
and stabilized fringing salt marshes in
North Carolina. Wetlands Ecology and
Management 16:97-118.
Davis, R.C., F.T. Short and D.M. Burdick.
1998. Quantifying the effects of green
crab damage to eelgrass transplants.
Restd at olt F.?ology ,:297-302.
Ferrari, N. l aventos and S. Planes.
2008. Assessing effects of fishing pro-
hibition on Posidonia oceanica sea-
grass meadows in the Marine Natural
Reserve of Cerbe`re-Banyuls. Aquatic
Botany 88:295-302.
Fishman, J., R.J. Orth, S. Marion and
J. Bieri. 2004. A comparative test of
mechanized and manual transplant-
ing of eelgrass, Zostera marina, in
Chesapeake Bay. Restoration Ecology
12:214-219.
Fonseca, M.S. 1992. Restoring seagrass
systems in the United States. Pages
79-110 in G.W. Thayer (ed), Restor-
ing the Nation: Marine Environment.
College Park: Maryland Sea Grant
College.
_. 1994. A guide to planting sea-
grasses in the Gulf of Mexico. Texas
A&M Sea Grant College Program,
TAMU-SG-94-601.
Fonseca, M.S., W.J. Kenworthy, F.X.
Courtney and M.O. Hall. 1994. Sea-
grass planting in the southeastern
United States: Methods for accelerat-
ing habitat development. Restoration
Ecology 2:198-212.
Fonseca, M.S., W.J. Kenworthy,
E. Griffith, M.O. Hall, M. Finkbeiner
and S.S. Bell. 2008. Factors influ-
encing landscape pattern of the sea-
grass Halophila decipiens in an oceanic
setting. Estuarine Coastal and Shelf
Science 76:163-174.
Fonseca, M.S., W.J. Kenworthy and G.W.
Thayer. 1998. Guidelines for the con-
servation and restoration of seagrasses
in the United States and adjacent
waters. NOAA Coastal Ocean Pro-
gram, Decision Analysis Series No. 12.
Granger, S., M. Traber, S.W. Nixon
and R. Keyes. 2002. Part I. Col-
lection, processing, and storage. In
M. Schwartz (ed), A Practical Guide
for the Use of Seeds in Eelgrass (Zostera
marina L.) Restoration. Narragansett:
Rhode Island Sea Grant College.
Hammerli, A. and T.B.H. Reusch. 2002.
Local adaptation and transplant dom-
inance in genets of the marine clonal
plant Zostera marina. Marine Ecology
Progress Series 242:111-118,
Harrison, P.G. 1987. Natural expansion
and experimental manipulation of sea-
grass (Zostera spp.) abundance and
the response of infaunal invertebrates.
Estuarine Coastal and Shelf Science
24:799-812.
Hughes, A.R. and J.J. Stachowicz. 2004.
Genetic diversity enhances the resis-
tance of a seagrass ecosystem to dis-
turbance. Proceedings of the National
Academy of Sciences 101:8998-9002.
Hughes A.R., J.J. Stachowicz and S.L.
Williams. 2009. Morphological and
physiological variation among seagrass
Marcb/June 2011 ECOLOGICAL RESTORATION 29:1-2 4 79
t;
(Zostera marina) genotypes. Oecologia
159:725-33.
Kirsch, K.D., K.A. Barry, M.S. Fonseca,
P.E. Whitfield, S.R. Meehan, W.J.
Kenworthy and B.E. Julius. 2005.
The Mini-312 Program: An expe-
dited damage assessment and resto-
ration process for seagrasses in the
Florida Keys National Marine Sanc-
tuary. Journal of Coastal Research
21:109-119.
Merkel, K.W. 1988. Eelgrass transplant-
ing in South San Diego Bay, Califor-
nia. Pages 28-442 in K.W. Merkel and
R.S. Hoffman (eds), Proceedings of the
California Eelgrass Symposium, May
27-28, 1988, Chula Vista, California.
National City CA: Sweetwater River
Press.
Montin, G.J. and R.F. Dennis. 2003. A
shallow water technique for the suc-
cessful relocation and/or transplan-
tation of large areas of shoalgrass
(Halodule wrightii). Pages 1-8 in P.J.
Cannizzro (ed), Proceedings of the Thir-
tieth Annual Conference on Ecosystem
Restoration and Creation. Tampa FL:
Hillsborough Community College.
Myers, R.A., J.K. Baum, T.D. Shepherd,
S.P. Powers and C.H. Peterson. 2007.
Cascading effects of the loss of apex
predatory sharks from a coastal ocean.
Science 30:1846-1850.
Orth, R.J. 1975. Destruction of eelgrass
Zostera marina by the cownose ray,
Rhinoptera bonasus, in the Chesapeake
Bay. Chesapeake Science 16:205-208.
Orth, R.J., J. Bieri, J.R. Fishman, M.C.
Harwell, S.R. Marion, K.A. Moore,
J.F. Nowak and J. van Montfrans.
2003. A review of techniques using
adult plants and seeds to transplant
eelgrass (Zostera marina L.) in Ches-
apeake Bay and the Virginia Coastal
Bays. Pages 1-17 in S.F. Treat and
R. R. Lewis (eds), Proceedings of the
Conference on Seagrass Restoration: Suc-
cess, Failure, and the Costs of Both,
Sarasota, Florida. Valrico FL: Lewis
Environmental Services.
Orth, R.J., J.R. Fishman, D.J. Wilcox and
K.A. Moore. 2002. Identification and
management of fishing gear impacts
in a recovering seagrass system in the
coastal bays of the Delmarva Penin-
sula, USA. Journal of Coastal Research
(Special Issue) 37:111-129.
Orth, R.J., M.C. Harwell, E.M. Bailey,
A. Bartholomew, J.T. Jawad, A.V.
Lombana, K.A. Moore, J.M. Rhode
and H.E. Woods. 2000. A review of
issues in seagrass seed dormancy and
germination: Implications for conver-
sation and restoration. Marine Ecology
Progress Series 200:277-288.
Orth, R.J., M.L. Luckenbach and K.A.
Moore, 1994. Seed dispersal in a
marine macrophyte: Implications for
colonization and'restoration. Ecology
75:1927-1939.
Orth, R.J., S.R. Marion,?S-.'Grange'i and
M. Traber. 2009. Evaluation of a
mechanical seed planter for trans-
planting Zostera marina (eelgrass)
seeds. Aquatic Botany 90:204-208.
Paling E.I., M. Fonseca, M. van Katwijk
and M. van Keulen. 2009. Seagrass
restoration. Pages 687-713 in G.M.E.
Perillo, E. Wolanski, D. Cahoon and
M.M. Brinson (eds), Coastal Wetlands:
An Ecosystem Iniegrated`Approaeh:
Amsterdam: Elsevier.
Paling, E.I., M. van Keul'en, K. Wheeler,
J. Phillips and R. Dyhrberg. 2001.
Mechanical seagrass transplantation
on Success Bank, Western Australia.
Ecological Engineering 16:331-339.
Petraitis, P.S. and S.R. Dudgeon. 2004.
Detection of alternative stable states in
marine communities. Journal ofExper-
imental Marine Biology and Ecology
300:343-371.
Philippart, C.J.M. 1994. Interactions
between Arenicola marina and 7_os-
tera noltii on a tidal flat in the Wadden
Sea. Marine Ecology 7?rove' s Series
111:251-257.
Phillips, R.C. 1960. Observations on the
ecology and distribution of the Florida
seagrasses. Florida State Board of Con-
servation Professional Papers Series
No. 2.
Pickerell, C.S. Schott and S. Wyllie-
Echeverria. 2005. Buoy-deployed
seeding: Demonstration of a new
eelgrass (7_osterq.mariw L.)„plant-
ing methods. pcologici Mngineering
25:127-136. `
_. 2006. "Buoy-deployed seeding: A
new low-cost technique for restoration
of submerged aquatic vegetation from
seed." SAV Technical Notes Collection
(ERDC/TN SAV-06-2), U.S. Army
Engineer Research and Development
Center, Vicksburg MS. http://el.erdc
.usace.army.mil/sav/index.htn-d
Procaccini, G., J.L. Olsen and T.B.H.
Reusch. 2007. Contribution of genet-
ics and genomics to seagrass biology
and conservation. Journal of Exper-
imental Marine Biology qnd Ecology
350:234-259. 1 " ? c .
Race, M.S. and D.R. Christie. 1982.
Coastal zone development: mitigation,
marsh creation, and decision-
making. Environmental Management
6:317-328.
Race, M.S. and M.S. Fonseca. 1996.
Fixing compensatory mitiga-
tion: What will it take? Ecological
Applications 6:94-101.
Renn, C.E. 1934. Wasting disease of
Zostera in American waters. Nature
134:416.
Reusch, T.B.H. 2006. Does disturbance
enhance genotypic diversity in clonal
organisms? A field test in the marine
angiosperm Zostera marina. Molecular
Ecology 15:277-286.
Short, F.T., R.C. Davis, B.S. Kopp, C.A.
Short and D.M. Burdick. 2002a.
Site selection model for optimal res-
toration of eelgrass, Zostera marina
L. Marine Ecology Progress Series
227:253-267.
Short, F.T., B.W. Ibelings and C. den
Hartog. 1988. Comparison of a cur-
rent eelgrass disease to the wasting
disease in the 1930s. Aquatic Botany
30:295-304.
Short, F.T., B.S. Kopp, J. Gaeckle and
H. Tamaki. 2002b. Seagrass ecol-
ogy and estuarine mitigation: A low-
cost method for eelgrass restoration.
Fisheries Science 68:1759-1762.
Short, F.T., L.K. Muhlstein and D. Porter.
1987. Eelgrass wasting disease: Cause
and recurrence of a marine epidemic.
Biological Bulletin 173:557-562.
South Florida Natural Resources Center.
2008. Patterns of propeller scarred sea-
grass in Florida Bay: Associations with
physical and visitor use factors and
implications for natural resource man-
agement. Resource Evaluation Report,
SFNRC'Fechnical Series 2008:1.
South Florida Natural Resources
Center, Everglades National Park,
Homestead FL.
Stein, G. 1984. Port's expensive seagrass
program fails to take root. Miami
Herald. Local section, June 10, 1984.
Suchanek, T.H. 1983. Control of sea-
grass communities and sediment dis-
tribution by Callianassa (Crustacea,
Thalassinidea) bioturbation. Journal of
Marine Research 41:281-298
Talbot, S.L., G.K. Sage, J.R. Rearick and
R.M. Salazar. 2004. Genetic structure
of Zostera marina in San Francisco
Bay: Preliminary results of microsatel-
lite analysis. Report to Caltrans. www
.biomitigation.org/reports/files/SF_
Eeigrass_Genetics_0_ 1578.pdf
80 4 March/June 2011 ECOLOGICAL RESTORA770N 29:1-2
Thorhaug, A. 1974. Transplantation of the
seagrass Thalassia testudinum Konig.
Aquaculture 4:177-183.
Traber, M., S. Granger and S. Nixon.
2003. Mechanical seeder provides
alternative method for restoring eel-
grass habitat (Rhode Island). Ecological
Restoration 21: 213-214.
Townsend, E. and M. Fonseca. 1998. The
influence of bioturbation on seagrass
landscape patterns. Marine Ecology
Progress Series 169:123-132.
Uhrin, A.V., M.O. Hall, M.F. Merello and
M.S. Fonseca. 2009. Evaluation of the
success of mechanized transplanting
of two tropical seagrasses. Restoration
Ecology 17:359-368.
Valentine J.F. and J.E. Duffy 2006. The
central role of grazing in seagrass
ecology. Pages 463-401 in A.W.D
Larkum, R.J. Orth and C.M. Duarte
(eds), Seagrass: Biology, Ecology and
Their Conservation. Dordrecht: Kluwer
Academic Publishers.
Valentine, J.F., K.L. Heck, P. Harper
and M. Beck. 1994. Effects of bio-
turbation in controlling turtlegrass
(7halassia testudinum Banks ex Konig)
abundance: Evidence from field enclo-
sures and observations in the north-
ern Gulf of Mexico. Journal of Exper-
imental Marine Biology and Ecology
178:181-192.
Valentine, R. and K.A. Heck. 1991. The
role of sea urchin grazing in regulat-
ing subtropical seagrass meadows: Evi-
dence from field manipulations in
the northern Gulf of Mexico. Jour-
nal;of E erimeni4l Marine Biology and
Ecology T5I "I
4i21(5r230. .
van Dijk; B.I. 'van Tussenbroiek,
K: Jimenez-Duran, G.J. Marques-
Guzman and J. Ouborg. 2009. High
levels of gene flow and low popula-
tion genetic structure related to high
dispersal potential of a marine angio-
sperm. Marine Ecology Progress Series
390:67-77.
van Katwijk> M.M, A.R. Bos, V.N.
de Jonge,, L.S.A.M. Hanssen, D.C.R.
Hermus and D.J. de Jong. 2009.
Guidelines for seagrass restoration:
Importance of habitat selection and
donor population, spreading of risks,

and ecosystem engineering effects.
Marine Pollution Bulletin 58:179-188.
Wigand, C. and C.A. Churchill. 1988.
Laboratory studies on eelgrass seed
and seedling predation. Estuaries
11:180-183.
Williams, S.L. and R.J. Orth. 1998.
Genetic diversity of natural and trans-
planted eelgrass populations in the
Chesapeake and Chincoteague Bays.
Estuaries 21:118-128.
Mark S. Fonseca is a supervisory ecologist
with the National Oceanic and Atmo-
spheric Administration, specializing in
seagrass ecology with special emphasis on
restoration, landscape ecology, and associ-
ated hydrodynamics. He can be contacted
at NOAA, NOS, Center for Coastal Fish-
eries and Habitat Research, Beaufort Lab-
oratory, 101 Pivers Island Road, Beaufort,
NC 28516-9722, mark.. fonseca 0a
noaa.gov.

March/June 2011 ECOLOGICAL RESTORATION 29:1-2
(0 81
t.??.

;A.
ELSEVIER
cological Engineering 15 (2(X)0) 227-237
ECOLOGICAL
ENGINEERINC
www. clscviel'.com,7ocatNeco le n
Integrating biology and economics in seagrass restoration:
How much is 'enough and why?
Mark S. Fonseca a,*, Brian E. Julius b, W. Judson Kenworthy a
c " National Oceanic and Atmospheric Administration (NOAA), National Orean Service.
Center for Coastal Fisheries care/ Habitat Research, 101 Pirers Island Roud, Beaufort, NC 28516, USA
h National Oceanic and Annospherh Adbninisiration (NOAA), National Orean Service, Deanage Asscss'rrtent Center,
SS.UC4 Root 10118, 1705 East-West Higlmeq, Silver Spring, MA 20910, (,SA
Received 30 March 1999; accepted 10 March 2000
Abstract
Although success criteria for seagrass restoration have been in place for some time, there has been little consistency
regarding how much habitat should be restored for every unit area lost (the replacement ratio). Extant success criteria
focus on persistence, area, and habitat quality (shoot density). These metrics, while conservative, remain largely
accepted for the seagrass ecosystem. Computation of the replacement ratio using economic tools has recently been
integrated with seagrass restoration and is based on the intrinsic recovery rate of the injured seagrass beds themselves
as compared with the efficacy of the restoration itself. In this application, field surveys of injured seagrass beds in the
Florida Keys National Marine Sanctuary (FKNMS) were conducted over several years and provide the basis for
computing the intrinsic recovery rate and thus, the replacement ratio. This computation is performed using the
Habitat Equivalency Analysis (HEA) and determines the lost on-site services pertaining to the ecological function of
an area as the result of an injury and sets this agairliJ? ',ht';J f e>ne>acc between intrinsic recovery and recovery afforded
by restoration. Joining empirical field data with :economic theory has produced a reasonable and typically
conservative means of determining the level of restoration and this has been fully supported in Federal Court rulings.
Having clearly defined project goals allows application of the success criteria in a predictable, consistent, reasonable,
and fair manner. Published by Elsevier Science R.V.
Ke,rsrards: Seagrass; Restoration; Success criteria; Goals; Discounting; host ecological services: Recovery: Habitat equivalency;
Population growth
1. Introduction
' Corresponding author. Telfax: Guidelines for site selection, monitoring, and
252-7289784, .: + 1-252-7288729; + I- success criteria for seagrass beds have changed
E-mail addresses: nrrrk.l'onsec,,t(atnoaa.gov (M.S. Fonseca), little in the past decade (Fonseca, 1989. 1992,
brian.juliuswnoaa.gov (B.F. Julius).jud.kenworthy(q)noaa.gov 1994). These criteria focus on achieving an initial
(W1 Kenworthy). level of planting unit survival that could generate
0925-8574 00'S - see front matter Published by Elsevier Science 13N.
P11: 50925-8574(00)00078-1
I:

228 A,. S. Fon.vecu eI u1. l:inluQicul F.ngineerinrg 15 (2000) 227 2.37
the targeted acreage of seagrass in a prescribed
period of time. Shoot density has also been com-
bined with survival and acreage as an indication of
the (asexual) reproductive capacity of the plant-
ings. However, physical setting, which influences
seagrass landscape pattern (Fonseca and Bell.
1998), could also alter the rate of bottom cover-
age. Thus, basic ecological information in the
form of intrinsic population growth and coverage
rates, net population growth rate, and environ-
mental setting have been combined to provide
guidance and set expectations of resource man-
agers faced with restoring injured seagrass ecosys-
tems. However. the manner in which these data
are applied in order to determine the quantity of
habitat restoration that must he performed, has
often been inconsistent. In this paper we explore
how basic ecological data on habitat recovery,
restoration effectiveness, and society's value sys-
tem may be linked to provide fair and consistent
computations on the extent of habitat restoration
that must be performed to compensate for anthro-
pogenic injuries.
The purpose of habitat restoration (the term
`mitigation' is sometimes used) is to 'compensate
for environmental damage or loss of habitat
through replacement of functions, values, andior
acreage' (Race and Fonseca. 1996). Federal wet-
land regulations require the traditional sequence
of injury avoidance, minimization, and, as a last
choice, compensation through active restoration.
Compensatory restoration has been seen as it
means of ameliorating wetland losses. In many
cases, some quantity of wetland must be generated
at some time after the initial injury has occurred.
The amount of wetland to be generated compared
with the amount of wetland injured is generally
referred to as the 'replacement ratio' and is usually
(but not always) greater than unity.. inferring that
the replacement habitat is equal to or larger than
the injured area. That ratio has varied widely
among habitat types, regions, and governmental
agencies, from less than unity to as much as 5 U
of restored habitat for every 1 U lost (pers. obs.).
The actual value of the replacement ratio has, to
all appearances. emerged from value judgments
about the criticality of the injured wetland itself,
e.g. was it endangered species habitat? is it difficult
to replace? and how long will it take to reach
pre-injury functions? High replacement ratios may
also be driven by the generally discouraging track
record of mitigation projects (Nicholas, 1992;
Roberts, 1993). Moreover, it has been suggested
that projects with low replacement ratios must be
then followed by other projects with higher ratios
of replacement, in order to maintain a regional
baseline of wetla.pd, acreage (Race and Fonseca,
1996).
The replacement ratio should be set to recoup
all lost ecosystem services -- in particular, the loss
of resource functions and products that occur
between the time of habitat injury and the time to
full recovery. Because the concepts of success and
functional equivalency are so closely tied, planning
for successful restoration and,,'or mitigation re-
quires early incorporation of interim loss consider-
atiotls;, Hpwever,.as mentioned earlier, the manner
in which! intecirn ecosystem losses computed has
been inconsistent. Often, the ratio appears to be
inversely proportional to the degree of public
interest in the project causing the habitat injury.
Computation of lost resource services requires
three assessments, (1) acreage of habitat lost; (2)
the length of time needed for the functions associ-
ated with that area (and lost to the ecosystem at
large during the period of the injury) to recover to
their pre-impact levels; and (3) the shape of that
recovery function. Using seagrass ecosystems as
an example, if 1 ha of seagrass were destroyed
today and replanted tomorrow and, for argu-
ment's sake, reached standards of equivalency in 2
years, the interim loss of ecological services over
this 2-year period would be relatively low. How-
ever, if the restoration of this site were not under-
taken immediately and if the site required 7 years
to reach its pre-impact state, the level of compen-
sation due the public for the interim losses from
this same 1-acre injury would be substantially
higher, highlighting the weakness of fixed compen-
sation ratios.
Actual projects, rarely enjoy tight temporal cou-
pling either between the injury and on-site repair
work, or between the injury and the additional
restoration (beyond that necessary to return the
injured site to baseline) required to compensate for
the ecological services lost from the time of the
injury until full recovery. Among other issues,
M.S. Fonseca el al. Evobgical Etkgineering 15 (2(00) 217-237
it is very difficult to consistently locate and suc-
cessfully create new seagrass habitat that meets
ecologically responsible site selection criteria
(which precludes simply substituting naturally un-
vegetated bottom for vegetated bottom). Finding
large acreage of suitable substrate for restoration
in close proximity to the impacted area is rare,
and often results in restoration at sites physically
removed from the impact area. Thus, any func-
tions affected by spatial elements of ecosystem
linkages (i.e. geographic setting) are lost. Second,
the lost production was removed from a specific
point in time. Therefore, in some instances it
cannot be returned in a way to avoid disruption
of ecosystem functions, such as the loss of last
year's spawn of herring 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 recovery from the injury, then one
could argue that plantings conducted longer after
an impact have less value than ones conducted
sooner. This realization is the basis for the new
approaches by National Oceanic and Atmo-
spheric Administration (NOAH) to standardize
the problem of computing interim loss services
objectively and quantitatively, which then provide
a basis for setting replacement ratios and arriving
at a quantity of persistent acreage of given qual-
ity, that has been defined as an appropriate metric
of success (Fonseca, 1989, 1992, 1994; Fonseca et
al., 1998).
Determination of interim loss and its imRle-
mentation into the restoration process is tightly
integrated with the establishment of a restoration
plan. While such a plan must identify the mechan-
ics of the physical restoration itself, the plan must
also have a clear definition of injury, site selec-
tion, monitoring protocols, and success. As men-
tioned earlier, those guidelines had been
established (Fonseca, 1989, 1992, 1994), but have
not yet been quantitatively coupled with the iss'ug.,
of interim loss to determine replacement ratios.
Recently, NOAA developed and implemented a
protocol termed 'Habitat Equivalency Analysis'
(HEA) that utilizes basic biological data to quan-
tify these interim lost resource services (NOAA,
1997a). While sharing many of the same principles
as other methods for incorporating interim losses

into replacement ratio calculations for wetlan,
(Unsworth and Bishop, 1994; King et al., 199
HEA focuses on the selection of a specific r
source-based metric(s) as a proxy for the affects
services (e.g. seagrass short-shoot density in tl
example discussed below), rather than basing i
calculations on a broad aggregation of servics
c.,im, -acted. This approach has the advantage c
making HEA applicable not only to a wide rang
of different habitats, but to injuries of individu?
species as well (see Chapman et al., 1998, for .
discussion of HEA applied to the calculation o
compensation for historic salmon losses). Addi
tionally, the selection of a resource-based metri<
allows for differences in the quality of service!
provided by the injured versus replacement re-
sources to be captured and incorporated with tht
replacement ratio (NOAA, 1997b). Without spe-
cification of a quantifiable resource metric, analy-
sis of the recovery of the resource following injury
and' or the success of the restoration project may
be difficult to evaluate precisely. For example, in
the wetlands context, alternative metric specifica-
tions may lead to significantly different maturity
horizons (Broome et al., 1986) as well as the level
of functional equivalence ultimately achieved by
the restoration project (Zedler and Langis, 1991).
In the remainder of this paper, we report on how
this linkage was established by reviewing the theo-
,retical and. biological bases of a restoration plan
that,wa, s developed in response to the destruction
of a subtropical climax seagrass bed (Thaiassia
testudinum), how HEA was utilized in the plan,
and how this procedure influenced project goals
and success criteria.
2. Case study: an example of how the HEA may
be applied
An example of applying HEA to habitat
restoration occurred in a recent Federal court case
(United States of America vs. Melvin A. Fisher et
al., 1997) to provide compensation for the loss of
1.63 acres of seagrasses (turtlegrass, T. tes-
tudinum) within the Florida Keys National
Marine Sanctuary (FKNMS). Extremely energetic
hydrodynamic conditions at the injury site to-

2230 ttt..S. lintserrr er ul F.cntu,t;ir<r! F.ngineeong 15 (20001 227- 2_t7
gether with intense grazing of the seagrass by
nocturnal animals prevented successful establish-
ment of plantings. Therefore, off'-site restoration
was chosen in the form of in-kind (same species)
repair of T. testudinum beds damaged by boat
propeller scars (prop scars). This approach fo-
cused initially on planting a native pioneering
seagrass species, 11ulodule wrightii, to facilitate the
eventual recovery of the slow-growing T. te,s-
tudinum. This sequence, termed 'compressed suc-
cession' (M. Moffler, pers. commun.), promotes
more suitable conditions for T. tesiudinton to
naturally encroach upon the prop scar while stabi-
lizing the site and preventing additional erosion.
Project success was to be quantified by four
parameters, (1) an average of minimum one hori-
zontal H. ivrighid rhizome apical per planting unit
must be installed at planting: (2) survival of plant-
ing units would be not less than 75'%, at the end of
year 1; (3) seagrass shoot density would not be
statistically different from that of nearby natural
beds; and (4) the target acreage of bottom cover-
age would be achieved within a 3-year monitoring
period. Additionally, if monitoring indicated that,
performance standards were not being met or
were not been projected to meet, remedial plant-
ings of those affected areas were designed into the
plan. However, all remedial plantings reset the
monitoring clock for that portion of the project.
The ultimate success criterion was unassisted per-
sistence of target bottom coverage by the seagrass
plantings for 3 years, with photo documentation
providing a common basis of assessment, percep-
tion, and historical reference.
Key factors in NOAA's development of a
restoration plan have been the issues of pre-pro-
ject planning, particularly regarding site suitabil-
itv. Here, sites were reviewed for the suitable use
of the following criteria, (1) they were adjacent to
natural seagrass beds at similar depths. (2) they
were anthropogenically disturbed; (3) they existed
in areas that were not subject to chronic storm
disruption; (4) they were not undergoing rapid
and extensive natural recolonization by sea-
grasses; (5) seagrass restoration had been success-
ful at similar sites: (6) there was sufficient acreage
to conduct the project, and (7) similar quality of
habitat would be restored as was lost. The
restoration of seagrass prop scars created by ves-
sel impacts represented NOAA's preferred ap-
proach to seagrass restoration off-site. In order to
select a planting site that could accommodate the
project's size, the amount of restored acreage was
computed using the HF.A, which is reviewed next.
3. Description of the compensatory restoration'
scaling approach
Accurate determination of the appropriate scale
of compensatory restoration projects is necessary
to ensure that the public and the environment are
adequately compensated for the interim service
losses resulting from the injuries to natural re-
sources. For injuries to seagrass resources, NOAA
has employed HEA 'As the primary methodology
for scaling compensatory restoration projects. The
principal concept ''underlying HEA is that the
public `and the environment can be made whole
for injuries to natural resources through the im-
plementation of restoration projects that provide
resources and services of the same type, quality
and comparable value. HEA has been applied to
cases centered on seagrass injuries because those
incidents typically meet the three criteria defined
by NOAA and upheld by the US District Court
(United States of America vs. Melvin A. Fisher et
al., 1997) for use of HEA, (1) the primary cate-
gory of lost on-site services pertains to the biolog-
ical function of an area (as opposed to direct
human uses, such as recreational services); (2)
feasible restoration projects are available that
provide services of the same type and quality and
are comparable in value to those that were lost;
and (3) sufficient data on the required HEA input
parameters exist and are cost effective to collect.
Note that if these criteria are not met for a
particular incident, other valid. reliable ap-
proaches and methodologies are available for
'C ompensutory. restoration' refers to any action taken to
compensate for interim losses of natural resources and services
that occur from the point of the injury until recovery of those
resourcesjservices to baseline. Conversely. 'primary restora-
tion' refers to actions that return the injured natural resources
and services to baseline.
M.S. Fonseca er al. ' Eralogic•al Engineering 15 (2000) 227-a`17
scaling the chosen compensatory restoration
projects (NOAH, 1997b).
At its most basic level, HEA determines the
appropriate scale of a compensatory restoration
project by adjusting the project scale such that the
present value of the compensatory project is equal
to the present value of interim losses due to, the
injury'. This `balancing' of gains and loss'es' is
accomplished through a four-step process
(NOAA, 1997x). First (step 1), the extent, sever-
ity, duration of the injury (from the time of the
injury until the resource reaches its point of max-
imum recovery), and functional form of the recov-
ery curve must be determined, in order to
calculate the total interim resource service losses.
Injured Resource
0 L
A
d
Time
Compensatory Restoration Project
d
e
10
L
B
Time
Fig. 1. Depiction of how habitat equivalency analysis accounts
for those resource services lost by an injury (A) and those
returned in the compensatory restoration project (t3), The
curves in this figure represent undiscounted, rather than the
discounted, service flows (cumulative provision of services over
time) in order to depict more clearly the biological processes of
recovery and maturity.
In some instances. it may be beneficial to all parties
involved to implement a project where the total discounted
gains from the compensatory project exceed the total dis-
counted losses. This situation occurs when the scale of the
preferred project can only be adjusted according to a binary or
step-wise function rather than a continuous function, or when
the resulting amount of natural resources;:services generated by
a restoration action cannot be tightly controlled following
implementation of that action (e.g. freshwater diversion
projects intended to create wetland acreage.
I,
Next (step 2), the resource services provided b
the compensatory project, over the full life of th
project, must be estimated to quantify the benefit
attributable to the restoration. This step i
analogous to the previous one and requires esti
mation of both the time required for the compen
satory restoration project to reach its maximun
level of service provision and the functional fom
of+the maturity curve. After these resource servict
losses and gains have been quantified, the scale o
the compensatory project is adjusted until the
projected future resource service gains are equa'
to the interim losses associated with the injury
(step 3). This process is depicted graphically in
Fig. 1, where the scale of the compensatory
restoration project is adjusted until the area under
the maturity curve (the total resource service
gains, represented by area B) is equal to the
interim lost resource services (represented by area
A). Because, these services are occurring at dif'fer-
ent,points of time, they must be translated into
comparable present value terms through the use
of a discount rate. Discounting is a standard
economic procedure that adjusts for the public's
preferences for having resources available in the
present period relative to a specified time in the
future. Because of discounting, plantings that oc-
cur longer after an impact are worth less in
present value terms than plantings conducted
shortly. after an impact, and therefore more plant-
ing, must be done as time elapses. Finally (step 4),
appropriate performance standards associated
with the compensatory restoration must be devel-
oped to ensure that the project provides the antic-
ipated level of services. Well-defined and
measurable standards are essential to the success
of the project regardless of whether the restora-
tion will be implemented by the parties responsi-
ble for the original resource injury or whether the
trustees will receive monetary damages to imple-
ment the projects themselves.
As part of the scaling process, it is not feasible
ta,measure and quantify each of the individual
..re.sc urge services provided by seagrass habitats,
such, as fish and benthic production, sediment
stabilization, nutrient cycling, water quality en-
hancement, and primary productivity. Thus, es-
sential to the successful application of HEA is the
212 M.S. H'urfseru et ul i;-eologieul bigineering 15 (2000! 237 237
development of a resource metric that is closely
correlated with the services provided by both the
injured and compensatory habitats. An appropri-
ate metric captures relevant differences in the
quantities and qualities of services provided by
the injured and compensatory habitats. In past
NUAA seagrass cases, short-shoot density has
been used as the resource metric for quantifying
the resource services provided by the injured and
compensatory habitats. Short-shoot density is rel-
atively easy to measure nondestructively and rep-
resents an important metric of plant coverage that
is the basis for the functional role of seagrass
habitat in providing food, shelter, sediment stabi-
lization, and nutrient cycling services. Increases
and/or decreases in shoot density generally indi-
cate the growth status of the entire population
and not just individuals within the population. It
is, however, a conservative metric because it does
not account for the ecological services provided
by the below-ground production and function of
roots and rhizomes, a portion of the T. testudinum
plant community that takes many more years to
develop than shoot density so as to provide nutri-
ent cycling and sediment stabilization equivalent
to that of natural beds. Moreover, in both the
compensatory project selection process and in the
development of an appropriate resource metric, it
is important to consider the landscape context as
well as the biophysical characteristics of the site
(e.g. access by fauna to the site, material flows to
and from adjacent communities, erosion control).
4. Development of model input
In order to conduct the HEA computation, an
empirical assessment of natural seagrass recovery
rates was required. If natural recolonization was
very high, then planting would have little strategic
advantage in accelerating recovery and the differ-
ence in the recovered discounted services as the
result of planting versus natural recovery would
be low, indicating that the project would not
substantively accelerate the return of ecosystem
services. Determination of these recovery rates is
therefore critical for implementing the HEA. Esti-
mates for relative recovery rates of different spe-
cies of tropical and temperate seagrasses are
generally known and several studies reported the
critical abundance and the growth parameters
needed to begin formulation of the recovery
model (den Hartog, 1971; Patriquin, 1973;
'Lieman, 1982; Williams, 1987; Fonseca et at.,
1987; Williams. 1990; Duarte, 1991; Tomasko et
al., 1991; Gallegos et al., 1993; Short et al., 1993;
Gallegos et al., 1994). However, population
growth rates for seagrasses range widely among
geographic' region5 and recovery rates depend on
the severity of the injury. Therefore, we have
recently completed several experiments that
provide the requisite data for a frequently injured
seagrass ecosystem, 5yringodium filifortne, in the
FKNMS (Sargent et al., 1995).
To calculate the required compensation under
HEA we estimated the time it takes for the in-
jured resources to recover to the pre-injury base-
line. For the 'same' reasons that seagrass density
was 'pr'evi'6usly chosen as a metric for planting
performance (Fonseca, 1994), we elected to use
short-shoot density as the metric for assessing
recovery.
We began to develop our model approach using
a recently injured S. ft'liforme meadow in the
FKNMS, a seagrass species that, like H. wrighth,
has a comparatively high population growth rate
compared with the target species, T. testudinumn
(Fonseca et al., 1987). Using these faster-spread-
ing species was critical for us in order to develop
and calibrate our modeling approach within a
short time (3 years). In addition, this injury was
operationally quite similar to the compressed suc-
cession approach taken in Section 2, not only
because S. filiforme also spreads much more
quickly than T. testudinum, but because it per-
forms a facilitation role for T. testudinum recovery
similar to that of H. wrightd.
In the example we are presenting here, between
25 and 501%o of the surface sediment layer was
removed by the injury event, leaving only 10- 20
cm of unconsolidated sediment in the injury area.
This degree of sediment disturbance was thought
to affect seed abundance, and it is known that the
growth and development of some seagrasses is
limited by sediment depth ("Lieman, 1972). Thus,
because of the severity of the injury we found it.
M.S. Fonseca el al. Ecological bygineering 13 (3000) 217 337
necessary to collect in situ data on recovery dy-
namics to supplement literature values and to
calibrate model predictions with actual recovery
rates. To accomplish this, we established perma-
nent stations at three sites along the extent of the
injured area and in the adjacent undisturbed side
population (USP), where we obtained population
data for S. filforme short-shoot density, short-
shoot demography, apical meristem density, hori-
zontal rhizome growth rates, vegetative
reproduction rates, and apical branching rates
(Kenworthy and Schwarzschild, 1998). The three
sites were sampled at least twice annually for 4
years.
Because T. testudinum grows so slowly,% we
derived short-shoot abundance data by sampling
the USP. Population growth data for the model
was obtained from our previous research and
other literature, with the initial assumption that
recovery was based solely on asexual reproduc-
tion. Kates of horizontal rhizome growth, produc-
tion of new apicals, production of new
short-shoots, and natural mortality for T. tes-
tudinum were determined from an exhaustive re-
view of the literature (Patriquin, 1973; Fonseca et
a1., 1987: Duarte and Sand-Jensen, 1990; Duarte,
1991; Tomasko et al., 1991; Gallegos et al., 1,99,3;,
Durako, 1994). The immigration of T. testudi(?unt,,,
rhizome apical meristems into the injury wasi,
modeled in the same manner as S. filiforme; but
slight differences in plant morphology required
modifications to the S. filiforme model structure.
Apical meristem densities in the injury were set
assuming the same proportion of apical meristems
to short-shoot per square meter in the USP. We
also collected data on seedling abundance of both
S. filiforme and T. testudinum in order to refine
the model and determine the relative contribu-
tions of seed and vegetative recruitment to
recovery.
We constructed deterministic population dy-
namics models of both S. filifortne and T. tes-
tudinutn recovery in STELLA 11 software (High
Performance Systems Inc., Hanover, NH) operat-
ing on a Macintosh Personal Computer. Although
T. testudinum was the target species, modeling the
faster-spreading S. filiforine provided an impor-
tant model validation step given the time con-
straints imposed in Section 2. The model wi
constructed to predict the recovery of short-sho<
densities, rhizome apical meristems, and othe
pp.pltlation characteristics on a square meter bas;
within the injury area. In its present configure
tion, the model contains stocks (populations) r.
rhizome apical meristems (primary apicals an.
branch apicals) in the injury and in the adjacen
USP of S. faliforme and T. testudinum. Division o
the rhizome apical meristem is the fundamenta
process that forms new shoots and causes growtl
of horizontal internodes in rhizomatous clona
plants. Thus, rhizome apicals are the major sourct
of vegetative reproduction and horizonta
expansion for most seagrasses (Tomlinson, 1974).
and were identified as the primary means of rc-
covery in the injury. This assumption was confi-
rmed for S. filiforine during initial sampling of the
injury 1 year after the disturbance when we
recorded densities of 1.2-6.6 seedlings m --' ver-
sus rhizome apical densities of 15.6-28.1 m-2.
Less than 25'%o of the seedlings observed had
begun vegetative reproduction, and most were
only single shoots. In contrast, the rhizomes of S.
filifortnc: immigrating from the adjacent USP were
spreading and branching vigorously, in-
dicating that the initial stage of recovery was
" ea lily dependent on recruitment from the USP.
Likewise, we estimated there were 0.26 T. tes-
tudinum seedlings per m', a minor component of
recovery.
4.1. Model predictions - Syringodium filiforme
Model output was generated to predict the
number of short-shoots per square meter in the
injury over a 10-year period after the injury (Fig.
2a), We validated the output by sampling the
three;sites in 2, 3, and 4 years after the injury. For
the S. filiforme model, predicted short-shoot den-
sities in the injury reached the approximate mean
density of the USP after about 3 years. In 3.5
years the predicted density exceeded the measured
mean density (2161 vs. 2045) and then stabilized,
reaching a steady state population of'short-shoots
in approximately 6-7 years. Although the model
accurately predicted that the injury would recover
to ambient shoot density in the USP in approxi-

234
8
r
ru.$. Fullceett e•1 ,d. F.cnlugirui EuKurerriJ,g`iS (,°CHk3} ???-i,??
A. model under the damage scenario, the model was
initially run for a 50-year period assuming an
UPOWC1 initial population in the injury of 0 (100`%, dam-
to*wCt age). The model population was allowed to reach
a stable equilibrium and the relative number of
r primary and branch apicals and the percentage of
r -- modada" short-shoots in each age class were recorded. We
• C" as
f 25cm Z"quads also ran the recovery model assuming 901%, dam-
' 20em 20cm guads
loaa•10em guam age in order to examine its performance as com-
0 0 t 2 7 4 S 6 7 a
Years Since l3 afurbance

S
so 90%damwa /

m
E 100
/
p?
p
t
SO
/ 100 % darn"*

pared with the 10011/, injury scenario. These values
T10 were then entered into the model as initial condi-
tions and 'the model was run to determine the
number of years `f`or the complete recovery. We
normalized the short-shoot density in a given year
as a percentage of the ambient baseline short-
shoot density and used that as a proxy for percent
g, services lost. This formed the curve defining area
_ B (Fig. 1), whereas area A in Fig. 1 was deter-
mined by the review of literature and best profes-
sional judgment.
d'1 s redid that
to 1s 1 20 26
Years Since DlsWrbance
The results of these no a run, p
recovery io' the` mean short-shoot density in the
USP'regia Yes`' pp ieoximately 3 years for the com-
paratively`faster-growing S. filiJorine (Fig. 2a) but
17.5 years for the slower-growing T. testudinum
(Fig. 2b) in the case of 1001)/,, injury, meaning that
3o after 17.5 years the age structure and apical to
short-shoot ratio of T. iesiudinun in the injury
were similar to those observed in the USP. An
Fig. 2. Results of the STELLA 11 model predictions for S.
fdifi;rme (top panel) and 7'. tee+iudinum (bottom panel) recov-
ery. Top panel shows the horizontal lines that represent the
95',`.. confidence intervals around the average density of S.
filiJ rme short-shoots in the undisturbed population. Also
shown short-shcxit densities are in the injury over a 2.5-year
interval based on core and quadrat samples. Bottom panel
shows the predicted recovery horizon in the injury for 7'.
testudinum under two damage scenarios and the mean density
of T. intudintem short-shoots in the USP.
initial injury of 90,1/,, suggested a similar time to
complete recovery.. indicating that a much larger
portion of the original short-shoot population
than 90%, would have to be left intact in order for
mately 3 years, it underestimated the overcompen-
sation response of the short-shoot density ob-
served in the injury population.
4.2. Model predictions - - Thalassia testudinum
To compute the initial values for each stock
component (primary apicals, branch apicals, and
short-shoot age classes) of the T'. testudinum
more rapid recovery to occur.
In Section 2, two factors contribute to a re-
placement ratio less than one. The first is that the
injured area is expected to recover over time. If
this area had been lost in perpetuity, the replace-
ment ratio would have been greater than one,
since the total amount of restoration required
would have been 1.63 acres to replace the services
provided by the permanently injured habitat, plus
additional acreage to account for the interim
losses in the period prior to the implementation of
restoration. and in, the post-restoration period
prior to the restored habitat reaching full maturity
(i.e. maximum service provision). The second fac-
tor relates to the specific estimated recovery and
AI-S. Fonseca 0 ail. %Etvlo,Qii'ul Engineering is (2000) 277, 2,77
maturity horizons for the iniured and restored
habitats, respectively. Because of the long recov-
ery horizon, many of the total service losses occur
far in the future. In the present value calculation,
these losses are weighted less heavily than losses
occurring closer to the present time (Julius, 1997).
Conversely, the compensatory restoration project
is expected to provide a large percentage of its
maximum annual service flow soon after the com-
pletion of the project. These early year benefits
are weighted more heavily in the present value
calculations than benefits occurring far in the
future. Thus, in this example, the combination of
expected recovery of the injured resource, plus the
greater weight given to early term benefits than
late term losses, results in a replacement ratio less
than one. Again, this is predicated upon our
assertion that the present value of those services
provided far in the future is less than the value of
the same level of services provided today, and that
present-day values are what should drive the
restoration process. Therefore, the project goals
were set at 1.55 acres and the previously estab-
lished performance criteria were applied to formu-
late the overall definition of project success.
5. Future directions

Although the wedding of' basic biological infor-
mation (recovery functions) with economic princi-
ples (discounting services) has yielded a
reasonable and predictable means of assessing a
party's level of responsibility, thereby setting fair
and consistent restoration goals, new issues are
emerging regarding the application of the HEA.
For example, a recent study (Fonseca et al„ sub-
mitted), revealed that the placement of a dredged
material island among the patchy seagrass beds
located in a wave-swept portion of'Southern Core.
Sound, North Carolina, towered the wave energy
on the lee side of the island and promoted a shift
in seagrass cover from - 15`4 bottom cover to
over 60% cover. Using simple extent or shoot
density as a measure of scagrass impact, the direct
loss of seagrass by creation of the island may have
been offset to some degree by an increase in cover
at similar densities through the coalescence of
patchy beds in the lee. The question is, does th
offset constitute built-in mitigation? Other issue
such as up-front mitigation trade-offs betwet
animal communities that might use patchy vast
more continuous cover (i.e. functional differenct
among scagrass habitats), should be considerec
However, future studies will need to consider th
degree to which (if any), modification of fray
mented seagrass beds may provide some kind c
inherent mitigative function by increasing loca
percent cover of the seafloor by seagrass as a1
offset to habitat injuries. In general, this exampl,
reveals the importance of incorporating othe
habitat attributes, such as landscape charaeteris
tics, into our evaluation of the true equivalency o';
the selected compensatory project as comparec
with the injured area.
6. Conclusions
Computations of interim loss have often been
divorced from ecological relevance, and conse-
quently replacement ratios have not been suffi-
ciently quantitative to provide predictable
standards for setting performance criteria, compli-
ance, and, ultimately success. New, economically
based models provide a means of standardizing
the interim loss computation. When coupled with
mensurative experimental data, this method has
been shown, through successful litigation, to
provide a reasonable basis for documenting in-
jury, setting goals, and gauging restoration
success.
Acknowledgements
.Thanks are extended to Arthur Schwarzschild
for his assistance in the STELLA modeling and in
the field. We also thank Harold Hudson and
Paula Whitfield for their assistance in the field
and Stephanie Fluke and Erik Zobrist, who as-
sisted in the development of the restoration plan
that was implemented in the Section 2. Finally, we
appreciate many constructive comments of Patti
Marraro, David Meyer, Gordon Thayer, the
anonymous reviewers, and the editor.
236
References
M.S. Font eeu el ul. Eculugirul Engineering iS (2000) 227 2?7
Broome, S.W., Seneca, E.D.. Woodhouse. W.W. Jr.
1986. .Long-term growth and development of transplants of
the salt marsh grass Spartina uhernt fluru. Estuaries 9,
63.
Chapman. D., Ladanza. N., Penn. T., 1998. Calculating re-
source compensation. an application of the service-to-ser-
vice approach to the Blackbird Mine. hazardous waste site,
NOAA technical paper 97-I. 18 pp.
den llartog. C., 1971. The dynamic aspect in the ecology of
seagrass communities. Thalassia Jugosl. 7, 101 112.
Duarte, C'.M., 1991. Allometric scaling of scagrass form and
productivity. Mar. Ecol. Prog. Ser. 77, 289--300.
Duarte. C.M., Sand-Jensen, K., 1990. Scagrass colonization:
patch formation and patch growth in C'+ntodarea nodose.
Mar. Ecol. Prog. Set. 65. 183 191.
Durako. M.J.. 1994. Scagrass die-off in Florida Bay (USA).
changes in shoot demographic characteristics and popula-
tion dynamics in 7'halacsiu tesrudinum. Mar. F,col. Prog.
Ser. 110. 59 66.
Fonseca, M.S.. 1989. Regional analysis of the creation and
restoration of seagrass systems. In: Kusler. J.A.. Kentula.
M.E. (Eds.), Wetland Creation and Restoration: The
Status of the Science. Regional Reviews. vol. L EPA 6003-
89+038a. Environmental Research Laboratory, Corvallis,
OR, pp. 175 -198.
Fonseca, M.S., 1992. Restoring seagrass systems in the United
States. In: Thayer, G.W. (Ed.), Restoring the Nation's
Marine Environment Maryland Sea Grant College, Col-
lege Park, MD, Publication UM-SG-TS-92-06, pp. 79-
110.
Fonseca, M.S., 1994. A guide to planting seagrasses in the
Gulf of Mexico. Texas A & M University Sea Grant
College Program, pp. 26.
Fonseca, M.S., Bell, S.S., 1998- Influence of physical setting on
Scagrass landscapes near Beaufort, North Carolina. USA.
Mar. Ecoi. Prog. Set. 171, 109 121.
Fonseca. M.S., Thayer, G.W., Kenworthy, W.J.. 1987. The
use of ecological data in the implementation and manage-
ment of Scagrass restorations. FL Mar. Res. Pubi. 42,
175-189
Fonseca, M.S.. Kenworthy, W.1., Thayer, G.W., 1998. Guide-
lines for the conservation and restoration of scagrass in the
United States and adjacent waters. NOAA COP'Decision
Analysis Series # 12. pp. 22?
Fonseca. M.S., Whitfield. P.L., Kelly. N. M., Bell, S.S. Statis-
tical modeling of seagrass landscape pattern and associated
ecological attributes in relation to hydrodynamic gradients.
Ecol. Appl. submitted.
Gallegos, M.E.. Merino. M.. Marba. N.. Duarte. C.M.. 1993.
Biomass and dynamics of T hulassiu iemalinum in the Mcx-
ican Caribbean: elucidating rhizome growth. Mar. Ecol.
Prog. Set, 95, 185 192.
Gallegos. M. F . Merino. M.. Rodriguez, A., Marba, N., Du-
arte, C M.. 1994. Growth patterns and demography of
pioneer Caribbeitn seagrasses liuluclule u'rightu and Sr-
ringtiil'ititn' ftlifornte.-I Mar. Ecol. Prog. Ser. 109. 99
104.
Julius. B.F., 1997. Expert report submitted in U.S. vs. Melvin
A. Fisher et al., pp. 25.
Kenworthy, W.J., Schwarzschild, A.C.. 1998. Vertical growth
and short-shoot demography of Svringodium jdl fortne in
outer Florida Bay, USA. Mar. Ecol. Prog. Sec 173, 25
37.
King, D.M., Bohlen. C.C.. Adler. K.1., 1993. Watershed man-
agement and wetland mitigation: a framework for deter-
mining compensation ratios. University of Maryland
System Draft Report # . UMCEES-CBL-93-098.
pp. ".
NOAA, 1997x. National Oceanic and Atmospheric Adminis-
tration. Damage Assessment and Restoration Program,
1997x. Flabiiat eq'ui6lency analysis: an overview. Policy
and Technical Paper Series, No. 95-1, pp. 18.
NOA.A, 1997b. National Oceanic and Atmospheric Adminis-
tration, Damage Assessment and Restoration Program.
1997b. Scaling compensatory restoration actions, guidance
document for natural resource damage assessment under
the Oil Pollution Act of 1990, pp. 143.
Nicholas, S.. 1992. Beyond no net loss, National Wetlands
Newsletter 14. pp. 11
Patriquin, D.G., 1973. Estimation of growth rate, production
and age of the makine angiosperm 7hctlassio t midinutn.
Konig. Carib., J. ScL 13, 111 123.
Race! M.S.; Fonseca, M.S.. 1996. Fixing compensatory mitiga-
tion: what will it take" F.col. Appl, 6, 94-101-
Roberts, L., 1993. Wetlands trading is it loser's game, say
ecologists. Science 260, 1890 -1891
Sargent, F.J., Leary, T.J., Crcwz, D.W., Kreuer, C,R., 1995.
Scarring of Florida's seagrasses, assessment and manage-
ment options. Florida Department of Environmental Pro-
tection, FMRi Technical Report TR-l, pp. 46.
Short. F.T., Montgomery. J.. Zimmerman, C.F.. Short, C.A..
1993. Production and nutrient dynamics of a Srringodium
filiforme Kutz. seagrass bed in Indian River Lagoon, Flor-
ida. Estuaries 16, 323-334.
Tomasko, D.A., Dawes, C.J., Hall. M.O.. 1991. Effects of the
number of short-shoots and presence of the rhizome apical
meriStcm on the survival and growth of transplanted sea-
grass Thalus.siu testudintun, Contrib. Mar. Sci. 32, 41
48.
Tomlinson. P.B_. 1974, Vegetative morphology and meristem
dependence - the foundation of productivity in sea-
grasses. Aquaculture 4. 107 130.
United States of America vs. Melvin A. Fisher et al., 1997.
92-10027-CiV-DAVIS.
Unsworth, R.E.. Bishop. R.C.. 1994- Assessing natural re-
source damages using environmental annuities. Ecol. Leon.
11. 35 41.
Williams. S.L, 1987. 4ompctition between seagrasses 77urias-
sia te.rtudintun and Y.i ringoditan filtforme in a Caribbean
Lagoon. Mar. Lcol. Prog. Ser. 35. 91 98.
{

lr?
M.S. Fonseca et ul.l Ecological Engineering JS (2000) 227 2?7
Williams, S.L., 1990. Lxperimental studies of Caribbean
seagrass bed development. facol. Monogr. 60, 449-
469.
Zedler, J.B., Langis, R.. 1991. Comparisons of constructed
and natural salt marshes of Sari Diego Bay. Restoration
and Management Notes 9, pp. 21-25.

Zieman, J.C.. 1972. Origin of circular beds of Thalassia (Spe
matophyta, Hydrocharitaccae) in south Biscayne Ba
Florida, and their relationship to mangrove hammock
Bull. Mar. Ski. 22, 559- 574
Zieman. J.C., 1982. The ecology of seagrasses of south Flo
ida! a community profile. USFWSOBS-82;25, pp. 158.
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

Pt1t3LIS1111) 13Y [HE PRESS sYNDtCA"IE M- 1fit- UNIVI.RSVIt Ob CAMBRIDG1
l'he Pitt Building.. Trumpington S[r'eet, Cambridge, United Kingdom
CAMBRIDGE USIVE.RSI"rY PRESS
bcr Edinburgh Building, Cambridge CB2 2RU, UK
40 Nest 20th Street. New Y("-k, NY 10011.4211. USA
477 Williamstown Road, Port Melbourne. 171C :3207, Australia
Ruiz de Alarcon 13, 28014 Madrid, Spain
Clock Flouw- Me Waterfront, Cape'1'own 8001. South Africa
http ;Pu?+•w.c•.unbridge.arg
c Cambridge University Press 2002
This book is in copvt'il:ht. Subject to stat.utorv exception
mid to the provisiorts of relevant collective licensing agtccmejrts
no reproduction (A any lxart may take plate without
the written permission of Cambridge Iinivvr.SEEN Tress.
First published 200'.
Printed in the United Kingdom at the University Press, Cambridge
}tvpefcc Swill. `).12 pt. Sen f1fif 2 - 11131
4 catalogue record €inr this book is available fi•arn the 81iti.tia Library
Lilwary of Congress C;amlontaijkg irr Mihiiration data
Handbook L)f ecological restoration I edited byj Martin R. Perrow. Anthony j. Davy.
p. cm,
includes bibliographical reterences.
Contents: III Principles of restoration - 121 Restoration in practice.
ISBN 0 521 79124 4
1. Restoration ecology. 1. Perrow, Martin R. (Martin Richard'}, 1964- 11. flaw, A. ).
Q11541.15 R4i 1116 2002
3:3195 153 - dc-21 -100104'1443
ISBN 0 521 79129 4 hardback
Contents
I.fst qJ, Contributors page ix 9 Beaches 197
Foreword Xiii CLIVE A. WATMSL.F.Y
1'reftice xv 10 Coastal dunes 214
Part i Restoration policy and infrastructure SIGURDUR GREIPSSON
1 The Americas: with special reference 11 Saltnlarshes 238
to the United States of America 3 JOY 13. LEDLER AND PAUL ADAM
MOHAN K. WALL, NIRANDER M. SAFAYA AND 12 Rivers and streams 267
FATIH EVRENDILEK
PETER W. DOWNS, KEVIN S. SKINNER
2 Europe 32 AND G, MATT KONDOLF'
F. JANE MADGW'ICK AND TIM A. JONES 13 Lakes 297
3 Africa 57 ERIK JEPPESEN AND ILKKA
WILLIAM NI. ADAMS SAMMALKORPI
4 Asia 78 14 Freshwater wetlands 325
BRYAN D. 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
BRUCE C. FORBES AND JAY
of degraded land 89
D. ?4LcKEIvDRLCK
MING II. W'ONG AND ANTHONY
D. BRADSHAW 16 lligh-elevation ecosystems 376
KRYS"FYN.A M. URBANSKA AND JEANNE
5 Oceania 99 C'.'( HAMBERS
IAN D. HANNAM AND ALISON M. COCHRANE
1.7 Atlantic heathlands 401
L
Part 2 The biomes
NIGEL R. WEBB
6 Marine and coastal ecosystems 121 18 Calcareous grasslands 419
STEPHEN J. HAWKINS, JANETTE R. ALLEN, MICHALL J. HUTC:HINGS AND
P.AUTANF M. ROSS AND MAR"I"IN J. GF.NNER ALAN J. A. STIANART
7 Seagrasses 149 19 Prairies 443
MARK S. FONSECA, W. JUDSON KENWORTHY, SCOTT D. WILSON
BRIAN E. JULIUS, SHARON SHUTTER AND
20 Semi-arid WOC)t11L1I1dS and
STEPHANIE I-LUKE
desert fringes
466
8 Coral reefs 171 JAMES ARONSON, FDIA%ARD LE F-LOC'H
SUSAN CLARK AND CARLOS OVALLL
Vii
viii Contents
21 Australian semi-arid lands
and savannas
DAVID .1. TONGWAY AND JOHN A. LUDWIG
22 Temperate woodlands
PETER BUCKLEY, SATOSHI ITO AND
STt-FHANE Mc LAC'HLAN
2:3 Tropical moist forest
KAREN D. ROLL
24 "(Topical dry forest: Area de
486 C:onservaci6n Guanacaste, northwestern
(:ost;I Rica 559
503 DANII:I, tt. JANZEN
Irtdex 585
Colour plates behveerf pages 6-7 and 558-551)
539
Contributors
Paul Adana
School of` 13iologlcal ` (iences'
University of Ne w° South Wales,
AiaStralia
William M. Adams
Department of'Geography
University of Cambridge
Cambridge CB2 3P.N, UK
Janette R. Allen
Part Erin Marine laboratory
University t.7i Llvvlpool
Port Erin
Isle of Man 3M9 6JA, UK
James Aronson
Centre d'Ecologie Fonctioneile et Evolutive
C:NRS (1.11'R 9056)
34203 Montpellier, France
Robin F. Bav
Department of Biology
C. _ r..dta College
Cola?r.tdo Springy Co 8090:3. USA
Anthony D. Bradshaw
School of Biological Sciences
University of Liverpool
Liverpool, UK
Peter Buckle y
Imperial College at Wye.
University of London
Wye
Ashf6rd TN25 5AII, UK
M. K. C hakraborty
Central Minim Research Institute (C:M0
Barwva Road.
D?hanbad 826001, Indira
Jranne C. Chambers
F:?,rry Mountain Research Station,
tiDA f orest Service,
University of Nevada,
920 Vallev Road, Reno NV 89532, LISA
Susan Clark
Department of Marine Sciences and
Coastal Management
University of Newcastle
Newcastle upon Tyne NEI 7RU, UK
Alison M. Cochrane
10 Valentine Avenue
Parrarn atta
NSW 2350, Australia
David N. Cole
Aldo Leopold Wilderness Research Institute
Rockv Mountain Research Station
USDA Forest Service Missoula MT 59801. LISA
David K. Conlin
Department of Biology
Colorado College
Colorado Springs CO 8090:3, LISA
B. 13. Dhar
Association of Indian Universities
New Delhi 110 002, Indira
Peter W. Downs
Parh'A Ltd
Corte Madera CA 94925, USA
and
School of Geography
University of Nottingham
Nottingham NG7 2RD, UK
James J. Ebersole
Department of Biology
Colorado College
Colorado Springs CO 80903, USA
Faatih Evrendilek
Department of Landscape Architecture
Mustafa Kemal University
31040 Anatakya
Hatay, Turkey
ix
x List of contributors
Stephanie Fluke
Nationat Oce mic and Atmospheric
Adrninistrstion
Office of General Counsel liar Natural Resources
Si F`ewrsburg FL 33702, USA
Marl: S. Fonseca
National Oceanic and Atmospheric Administration
National Ocean Service
Center for Coastal Fisheries and Habitat Research
Beaufi,rt NC 28516, USA
Bruce C. Forbes
Arctic Centre
University of" Lapland
FIN-96101 Rovianemi, Finland
Martin J. Genner
Biology and Ecolo'?y Division
School of Biolol-ical Sciences
University of S(luthalnl)1011
Southampton 501.6 7PX. 1K
Sigurdur C;reipscin
Department of Biological and Environmental
Sciences
'Troy State University
Troy 6082. USA
lava D. Haunam
10 Valentine Avenue
Parramatta
NSGV 2150, Australia
Steplaeii ,(. Hawkins
Marine Biological Association UK
Citadel Hill
Plymouth PL1 211B, L!K
and
Biology and Ecology Division
School of Biological Sciences
i3niverity of Srnith<vnpton
Southampton. 5016 7PX, L7K
Karen D. Holl
Environmental Studies Department
University of Cahftlrnia-Santa Cruz
Santa Cruz CA 95064, USA
Michael ?. Hutchin<"'s
Department of Biology
Umversity of Sussex
Bril;llton BNI 9Q, UK
Satoshi Ito
Forest Science Division
University of Miyazaki
Miyazaki 889-2192, japan
Daniel H. Jarizen
Department of Biology
University of Pennsylvania
Philadelphia PA 19104, USA
Ltik Jeppeson
National Enviroimmital Research institute
DK-8600 Silkeborg, Denmark
'Pilo A. Jonas
UJEnviromilental
Harper:, Mill
Berrynarbor
North Devon FiX34 9TB. UK
Brian 1. Julius
National Oceanic and Atmospheii( Administration
National Ocean Service
Damages Assessment Cenrer
Silver Spring N41.) 20910, USA
W. Judson Kenworthy
National Oceanic and Atmospheric Adniiiiistr•ation
National Ocean Service
Cc•rtter firr Coastal Fisheries and Habitat Research
l3eaU1C)rt NC 28516, I15A
G, Matt Kmidolf
Deparnnent of Landscape architecture anti
Environmental Planning;
iniversity of'California-Berkeley
Berkelev CA 94720, USA
Edouard Le Floc`h
Centre (IT"cologie Fonctionelle et F.volutive
CNRS (UPR 9056)
34293 Ylompellier, France.
List of contributors
john A. Ludwig llkka Sa1TImalkorhi
Division of Wildlife and Ecology Watercourse Plarining and Restoration
C:SIRO Finnish Environment Institute
Winnellie HN-00251 Helsinki. Finland
Darwin
Ni 0822, Australia Sue C. Shaw
Welland Rescaich GrOLIp
F. Jmle Nlad*vick Department of Animal and Plant Scicnce
I"uropean Freshwater Programme University of Sheffield
c/o WWI` Deiuuark Sheffield 510 2'T'N, UK
171-2200 Copenhagen N, Denmark
Sharon Shutle^r
jay D. Mc?Kendrick National Oceanic and Atmospheric
PC) Box 902 Administration
Palmer` AK 99645, USA Office of"t'cneral Counsel fe r Natural
Stellilane M1t Lachlan
Resource's

Iiily'It{anT1ltlataal `YCIe'alt'e' Program
Silver" Shang MD
20910. USA
Univer,,iiy of Manitoba Kevin S. Skinner
41'innipt- R3T 2N.. Canada Sc'hocil of Gcograplly
Uriivcrsity, of Nottingham
Russ P. N.Joney
V(aTlillgllam NG7 2RI7. UK
Welland Re waI'ch Grollp
Department of Animal aid Plant elan J. A. Stewart
Sciertce Department of ffiohi*'
I,niversity of Sheffield University of Sussex
Sheffield S10 2TN, UK Brighton BNI 9Q.1. LlK
Carlos Ovalle David J. Tongway
CRI Quilairaapit GtIngahlin Homestead
Imtituto de Invtstigaaciones CSIRO Sustainable I Cn;+vsle:ms
Agropecuarias (INIA) GIIO Box 284
C.asilla 426 Canberra
ChilLin. Chile ACT. Australia
Pauline M. Ross Krvstvna. M. Urbanska
Marine Biological Association UK ?,NtabtltanlSt he'S Institut F'"ITI
Citade=l Hill (;11-8044 Zurich, Switzerland
IlIvnlouth 1111 21113, UK Mohan K. Wah
and OSU F110a'ol'lmetltal SCieflCt? (;Fad'UMe PI'ograill
Faculty (A Sa ience° and Tvchnoloa and
University of Wes'lern Sydney School of Natural Resources
Richmond The Ollio State University
NSW 275:3, Australia Columbus 011 43210, USA
Nirander W Saf-aya Clive A. Walm,?,k •y
Recl.imation Division Countryside Council liar Wales
North Dakota Public Service Commission F1hrd I'vnrhos
Bismarck ND 58505. USA Bangor 1-1571 2I,Q. UK
X1
xii List of contributors
Nigel R. Webb
N1,RC: Ccmre t0v Ecology and Hydrolm,y
Winfrith "Technology Centre
Dorchester DT2 8ZD. UK
Bryan t). wheclcr
WOLInd KeseM-Cll Group
Department of Animal and Plant Science
Univerrsity of Sheffield
Sheffield S10 2TN, UK
Scott 1). Wilson
Department of Biology
University of Regina
Regina
Saskatchewan S4S OAZ. Canada
Ming It. VVong
Instilute for Natural KcS0111'teS and
EnVirOnmental Manage111Cnt
and
Department of Biology
Hong King Baptist University
longg Kong SAR
People's Republic of China
Joy 13. Zedlcr
Botany Department and Arboretum
University ot,wisconsin-Madisoil
Madison WI 53706. USA
Seagrasses
MARK S. FONSECA, W. JUDSON KENWORTHY, BRIAN E. JULIUS,
SHARON SHUTLER AND STEPHANIE I 'LUKE
INTRODUCTION
Seagrasses are marine flowering plants consisting
ol'12 genera and approximately 60 species growing
in all of the world's oceans with the exception of
the most polar regions (den Hartog, 1970; Phillips
Meiiez, 1988). Nearly all seagrasses grow in uncon-
solidated Sediments ill water depths ranging froin
the intertidal zone to as deep as :35-50 in. They,
are vascular plants anchored to soft sodinic tits by
a functional and complex rhizcune and root system,
with the exception of the genus Phyl.lospadix which
grows on solid substrates along the Pacific coast of'
the United States. The seagrass leaf canopy baffles
the flow of water, and together with their rhizome
and root mat seagrasses stabilise sediments, cleanse
the water column of fine particles, and recy?-Jv nu-
trients between the sediments and overlying wafters
{Ponwca. 19961. Numerous species of invericbrates
and large vertebrates consume seagrasses as a por-
rion of their diet, and the complex structure and
phvsi(,al stability provided by seagrasses form the
basis for productive ecosystems consisting of plant
;inc] animal epiphytes, benthic macroalgae, inverte-
brates, mobile vertebrates and numerous other or-
ganisms (Thayer et al., 1984). Many of the animal
species that utilise seagrasses rely on their struc-
tural complexity to provide shelter and sources of
f6od for their juvenile stages. This is one of the
most important biological functions of the seagrass
ecosystem,
The lack of taxonomic biodiversity in seagrasses
is compensated by a wide diversity of Size and
morphological growth forms. The size and biomass
of seagrass varies over an order of magnitude
(Kenwortity et al., 2000), resulting partly from
genotypic differences as well as from phenotypic
plasticity within individual species. For example.
the canopy height of the smallest species known,
Hcaiophilca decipiens. usually never exceeds 10 cm
while species of Zostera can have canopies exceed-
ing 5-7 in in height. The diversity of clonal growth
forms and sexual reproductive strategies is accotn-
panied by phenotypic variation that allows the lim-
ited number of seagrass species to occupy a wide
range of environmental conditions from wave-swept
shorelines to relatively deeper regions of continen-
tal shelves. Only a few other macroscopic plants
growing in the ocean are capable of filling the niche
type that seagrasses occupy.
Even though taxonomic biodiversity is limited
in seagrasses, the diversity of size and morpholog-
ical forms is accompanied by different growth and
survival strategies uniquely adapted to the environ-
ments where the plants thrive. The range of growth
strategies is also responsible for the patterns of sea-
grass bed development seen throughout the world.
This is especially evident in multi-species tropi-
cal seagrass communities w=here distinctive succes-
sional processes are evident in the formation of
stable climax communities and in their response
to disturbance (Zieman, 1982). In Tropical seagrass
communities, cotonising and climax species can be
readily distinguished from one another and the
unique attributes oaf these species can be utilised to
enhance their protection and restoration (Fonseca
t't 01., 1987).
RATIONALE FOR RESTORATION
Fortunately, ill many cotuatries, the battle to recog-
nise seagrasses as critical coastal ecosystems wor-
thy of conservation and restoration has been won.
149
150 MARK FONSLCA ET At-,
Vhi?, recognition can he crediti-cl to lhr puhiic'aimm
al, 111m sancta of paper" trnnl do/cns (If cc!untllL"
aIoltnd thc° wtllI(I re I) rCSPniing )'(:'d I's 171 re,e:irah To
the hest ( A olIr kIIo °ledge. rescarc I I I laa yet to re(oI(I
;I seagrass Nett which i, alivthind; hilt a fannkal-rik 11,
highly productive ccusmlcm. that stahihws the v
floor, limits coastal ctosfti) and filters the waiter (ol-
uliin Iti1ood et all., 1969y thus, the ecological and
soctolw4ic l value of seagrclsses has been bro,idly
e tablished i%. vllie-f:chevcrria e! (Ii.. -?000) Where
these values are not recognised it olten appears
to be the result of ]veal pKitical and develop runt
interests overriding Cmunnown values ipersaum
obse nna t i can j.
grass ccosymums and causes c71
Threats to sea!
dcgradatic'an arise from a wide variety r>f' Sources.
lutr0f:7liication, coastal construction, nwwr vessel
cipelation, fishing practices and manly other activ-
hies have led to both local and regional tosses of
sear,iasses (.Short & Wyl-H Lcheverria, 1996; l riseca
ct al.. 1998a(. Losses of magrass alai occur through
natural processes such ,i5 disease f4eluelstein. 19ft9:
Kobblee et all , 1991 i, trccpiial cyo.fimvs fl Wish
"t tat-, 1r}951 and Uyer} 1'alliiS' by 111veriehlaWN (Ro,o
et ui., 1999)- Where the species koml)vsflion and
lilcditstory, stratc,.;rics promote reenlonisali(,n, aca
rassca ran rcrovcr natur:llly front lVilurhatioll-
fPree'll ei (FL. 19951. 1107\A+t'1'l'I', ill lllany ul,tall(cs e'a-
thcr the scvurily of the etlviroll mcntal luodifica-
tfon responsible for the declines or the extremely
slow rate of natural recovery leacl.s to king:; tern)
losses. For cxampie, in climax tropical Couninuni-
ties dvniinatt,d by Jhoiassf<a t?'Snlllirilf?tt the time To
full recvveiy in severer damaged vessel *rouud-
fng sites can be more than a decade ll entvonhy
et it!., 20W Whitfield lit lat.. in pressl. In these In
Stan((",, loss of scag'rasses )cads to nu11101,nia tltl
desirable and difficuii to reveres rcanditions, most
in'ipurtan% the elimination of habitat sea ucturc
and the sediment stabilisation prolrerties of rite
c,m,lpy and rhizome mat. A negarm, Icedbaclk fill
the' CkOSYStelIa I'cttiltS; 011Ce the sedg'rass COW1 i,,
lost and with it the self-Sustaining properties of the
sVstcni provided by the seagiasses, modification of
the sediments and dei?,,radation A the water col..
1111111 may proceed without interruption Seagrass
rcsturatilln then Ivcm(es a much more diftktdt
I'l"k, Iac,=aum it IN rlearll irlaf?ossililc to rcplace The
altrihtlleN ',e•ai;ra NN Cs provide, and a way to ccarroit
IIle hhy1;acacitcilmal properties (ii the sva lcl i must
he fitlluci bc;i>rc reiutn?durtiam r7f the si a#,lass(? Ball
1?c.11n.
VV'(; posit that the lsstu•s legardult; scat rats
mtoratinn are not the teclinolo??,y of planting; and
raising sc t,.;rass beet";, but the failure to apple basic
ecological principles in Implementing restoration
actions. se agrasses can be retaddy transplanted and
when sites are appmphaady belened {see below
mul (hsc-tlsmon in FvnseCa et ui 1.9988), sfgn.ifi-
c;ant resloratioll smC esses have enl(:'rgvd. In fact,
new ter htaullrhics arcs continually heing, developed
in Noll) the decpwarcr )Perth, Western Australia:
Fonseca ct cll., 19981r; E. Paling, personal cormi ts-
nlcatirrni and staallow water (Tampa Hay, florid;,:
1. Ander.`e n personal cammunicationi approaches.
;'alson imprammints in l.q scale seeding tech-
niques are being advanced which have promise with
saatue :ae;agrass sfw(ro lC tatager et at.. 2000; Orth
(, cit., ?11(kli, We are only twat he,giluimn to
recognise
the' many Nil ilallofl^ in which rcfrf?artllnnies firr St1h
SiAntlal re"lorallorl have either h(,vII miuandered or
s'Craacls iilltitakC'1 in iitc scic(.iion have hce'n inad-c,
I il,;;( ly ix'c:aslac tho"k. in~,okcd did viol understand
the habitat reiluirclnents atld"'ol the lifi' historv of.
thc plants evith wjhich Thad well, vvorkiig;
The cxological value of scagrasses tianslates No
enonnmis conin ercial and social benefits, For ex-
taniple, in the hidian )acct Lagoon, Florida m Wass
ineadows page been described as the marine equiv-
alent of tropic;ll raint6rests providing the ecolog-
ical basis for fiNherics worth ;about US.S25000 per
hectare or a lot.al of approximately 'I billion dol-
I,ar a ye.n 1V'irlistein ;i %Iorris, 1996!. 5cragrast;
clef7Cndcn1 hsherirrs and wildlite k0111111tlnities are
the erouolnic foundation for commercial and vecrae-
ational fishCrnte`n as well as fire a variety of iudils-
I! res and people that utilise ill(. coastal zone for
tlietscr
commcrcc and personal cii}oymcut. Socially
values are transferred to the health and well-being
of rile families of these user grOLIpS and the rL
giorial economies of nations v"orldwide. The many
physical. biological, cCUllortllC and Social attributes
Seagrasses 151
combine to ctt:ake scam-asses an essential and ecolog-
ically important habitat in coastal marine ecos.ys-
rems fWvllie-F.cheverria apt al., 2000); consequently
they are in need of restoration where they have been
anthropogenically injured or lost (Folrseca et cal..
1995: Sheridan, 19991.
PRINCIPLES OF RESTORATION
We base our assessment of the status of seagrass
restoration on as perspective front within the United
States legal framework. Scagrass beds in United
States coastal waters are generally viewed as pub-
lic trust resources, and such in to these re-
sources art' considered losses suffered by the pub-
hc. A nluuber of ?:clc r?rl and state laws includo
liability provisions which allow the public to be
componsatecl fitr ia.ltwies to seagrasses (lot, exaam-
pk% the United etates's ?rational Marine Sanctuary
Act cif 1972, 16 USC: 1431 et' se°d.), to evaluate i:his
loss in a faaia' and reasonable !minter, we tmtst con-
sider not only the static loss in area anchor de-
gYree of the injury. but also the loss of resource
services- provided by the se•agrass bed between i1he
time it is injured and the time it recovers to
1170% of pre-injury conditions (Fonseca et al_ 2000ca).
'this approach is consistent with the 'no-net-loss' of
Wetlands policy that has become a benchmark of
restoration strategies in the united states- Out' more
recent approach substitutes for the `mitigation' or
replacement ratio' used to identify the amount
cal' habitat to be generated to offset the amount
lost. lit the past. use (if' replaCelttent r=atios has fre-
quently led to unclerc?ortapensation of lost resources
because lost interim ecological services were tint
.addressed.
Effecting no-net-loss and achieving recovery of in-
terim resource services requires that the ilzjured
site be fully rehabilitated (on-site restoration), al-
ternative compensatory restoration sites be found
(off-site) or sortie degree of both. To limit the
.;eope of discussion. we are focusing on itt-kind
1%,? oratioal (i.e. seagraass service loss replaced by sea-
!,Ii i- service gaills). Oil-site restoration can often
1,,, achieved, but may require en ineering; imeiven-
taoals to Aix' the site, such as filling excavation.
hales caused by vessel groundings or ;-altering water
flaw. ()tfsite selections, in our experience, have had
higher probabilities of restorMiott f1lflUre because
inexperienced resource managers choose inapprcr
priate sites. They, are frequently under the impres-
sion that open habitat areas ire prime sites lot
restoring se.agrass when, in reality, the sites selected
either cannot supl"Wort seagrass, or currently support
only low levels of seagrass. This fallacy has been ad-
dressed in detail in several publications (Fredette
rat al., 1985: Fonseca rl ctl., 1987, 1998x: Fonseca, 1992,
1994). Suffice it to say, Fredette et cry's (1985) con-
ehiiort 'If seagrass does not grow there now. what
makes you think it can be establisliedT hest sums
the problem. Recently, Gatumpong el oil. (in press)
listed the criteria for off=site selection that can be
used to avoid off-site selection problems. By giv-
ing attention to these details of site selection, file
probability of sucCesslarl restoration can be greatly
enhanced.
RESTORATION IN PRACTICE
We have clealt previously with what we consider to
be the statues of this aspect of' restoration (Fonseca
rt ill., 1.998cr). however, there are at 1U,i1t fi;ur major
deficiencies in the process of seagr iss restoration.
First, the choice of an appropriate metric for evalu-
ating restoration has been elusive. 11'e present here
for the first time findings front a panel of United
States seagrass experts that considered what are the
appropriate metrics for tracking; the performance
of a seagrass restoration project, second, setting
laat, reasonable and consistent ratios for replace-
nit tit of damaged seagraasses has also been at issue.
We r(-,riew the ntethoc1010111' y used by the National
0ccuiic and Atmospheric Administration (NOAA)
for defining the interim loss of resource services
accrued by damage to seagrass beds and the pro-
cess ot'computing compensatory restoration. !laird,
we feel that the weakest part of seagrass restora-
tion has been the selection of- the restoration site.
We delve into rite pitfalls of site-selection strategy
the point in the process where most plans go
awry. For- completencss, we briefly review the extant
methodologies for restoring seagrass beds. Fourth
152 MARK FONSECA ET AL.
and finally, finding: a realistic basis ftr computing
cost of these projects has been a vexing; issue for
vears. Here we provide an evaluation of cost for
the ptanning;, implementation and numitoriml of
.r sc lg lass restoration project teased on a l nited
Stales federal court ease stwcessfuily prim ci ted by
NOW
Definition of injury and evaluation of lost
interim services
Defining lost resource services
Computation of lost resource services mcphres three
assessments f 11 area of habitat howl: 121 the length
of time needed for the hinctiunw associated with
that area Land lost to the ecosystem at lat;ge dur-
ing the period of tilt, injurpj to recover to their
pre impact levels and ! ii the shape of that recov
crv function {Fonse(a I Ell. 20000) (:sillg seagrass
ecoss mellis as all example, it I hectare of seagrass
were ckrstmycd today and replanted tomorrow and.
lot argument's salve, reached standards of equiva-
Icncv ie.g. shoot density, batntrass, rr?icragc ftt twO
vears, the interim loss of ecological ser-vircs over
this two-year period woidd he relatrvcly loiv. How-
ever, if the restoration of this site bSerc not under-
taken inunediately and if ill(, site required seveli
Years to reach its preompact state, the level of com-
pensation clue the public for the interim losses from
this same 1-hectare injilry would be substantially
higher-, This highlights the weakness of fixed com-
pensation ratios.
Actual projects rarely enjoy tight temporal coup-
ling hetween either the injury and on-site repair
?roa1:, or between the injury and the additional
restoration required to crmlpensate for tilt' ecolog-
ical services lost frgm tilt, tinge of the injury' until
full recovery. Among other issues, it is very difficult
to consistently locate and successfully create new
suvrass habitat that meets ecologically responsible
site-sc lection criteria. especially those criteria which
preclude simply substituting naturally unyegetated
bottom tart vegetated bottom (Fonseca et 0. 1998n).
Finding large areas of suitable substrate for remora
tion in close proximity n) the impacted area is rare,
and often rescilts in restoration at Sites physically
removed from the impact area. Thus, any 1unctions
affected by spatial elements of e(osystcna linkages
are lost fl,n geographic settings). Second, the lost pro-
ductiota was removed li-orn a specific point ita time.
Therefore. in sore inmances it ca nivit be returned
ida w., to avoid disruption of ecosystem functions,
such as the Ins,, of last year's spawn of herring or set
of hay scallops that might occur as a result of injury
to a magma W. Moreover, if there were a longer
period of time between the injury and full recov-
ery from the Injury, then one coWd argue that re-
planting conducted a long time after an impact has
less value than ones conducted sooner. This realisa-
tion is the basis for NOAH' more recent approach to
objectively and quantitatively standardise the prob-
lem of computing interim lost services by habitat
eWitvidency analysis Ifil..A). This approach provides
a basis Rkr setting replacement ratios crud arriving
at a quantity of persistent area ofgiven quaMy that
ha, 11C°en defined as all appropriate raaetric of suc-
ce:,s H!""eca, 1989. 1992, 1994: Fonseca et ai„ 19980.
2000o I.
Determination of interim loss and its implemen-
tation into the restor<rtictll plcwue s is tightly inte-
grated with the estahlishruent ofa rc:storatiotl plan.
W'hile such a phn must ideniity the mechanics of
the physical restoration itself, the plat, must also
have a clear definition of injury, site selection, Monf-
toring protocols anct success. As mentioned earlier.
those guidelines have been established (Fonseca,
1999, 1992, 1994), but have not yet been quarinta-
iWely coupled with the issue of interim loss to cfe-
terinine replacement ratios.
Recently, NOAH developed and implemented HFA
using basic biological data to gUantAN! interim lost
resource services (N()-kA. Damage ;Assessment and
Restoration Program, 1997((). VV`hile sharing many
of the same principles as other methods incorpo-
rating interim losses into replacement ratio caicu-
lanons for wetlands 0 insworth & Bishop. 1994: king
I A._ 1993). 1IFA focuses on the selection of a spe-
cific resource-based tuetricts) as a proxy fir the at-
tec red services (e.g seagrass short-shoot density in
(he example discussed below). rather than basing
its calculations oil a broad aggregation of ininwd
rewources. Determination of this metric was one elf
the conclusions front the expert 1)anel as discussed
Seagrasses 153
in Box 7.5 (bioinass, as opposed to shoot density, has
not yea: been adopted because of a lack of ernpisicaal
data oil the recovery rate of belovvground Noma",
whereas recovery rate of shoots is a robust data see:
this choice: is all extremely generous Concession Tx-)
the responsible parries[. '11his approach has the ad-
vantage of making 1- EA applicable not only to a
aide range of clifferent habitats, but to injurics to
individual species as well {set, C;hapiraan et cad. 11998
for as discussion of IDA applied to the calculation of'
compensation for historic salmon losses). Addition-
ally, the selection of it resource-based metric allows
fir difftrnlaCCS in the cluality Of sCrViCes providCcl
by' tilt' injured and replacenlent resources to be caEr
turCd Mid illCOrpUrated 111tt3 than rej)lact_tnCllt ratio
(NOAA, Dalliage Asscstment and Kcstciration Pro-
giana, 19971a). Without sp c•ifacotion of as quantifi-
able resource metric, .analysis of i he recovery cof the
resource following injury and?lor tilt: success of the
restoration project may be difficult to evaluate prt-
cisely, for example, in the wetlands context, alterna-
live metric specifications may lead to significantly
different maturity horizons (lirootne of al.. 1986)
as well as the level of functional equivalence ulti-
mately achieved by the restoration proJect (7.edjur N
hangi.s, 1991).
Box 70 Application of the habitat equivalency
analysis (HEA)
An csample of applying; HEA to habitat restor.i;icaai
occtirred in a r:-cent federal court case to prn?i:ic
oiripcr,<,iCiOri f )r the loss of 1.63 acres of ?t -r:a, cs
?t.U r(li'h CA4a, P?al 'c: i,,Srudinutn)witllirl the rIcial'i
keta \ational lv arine Sanc•tuary`((I,itteti Shltar ;far, ; ,:a
,4vin A, Fisher et at-, 1997). Exi cTiwiv energetic
hydredvil'i nip` conditions <?[ Ow mjt ry 3ite, Together
with inrcrst L arazim, cC nc- nocturnal
herhivoreo prevented Yuccr,siiil c<tnblishrueut o,"
seagrass plantings. Theieforv, otfsite restar,o,o,n xvia.i
shorn its the form of ii.-kiii,l (sane sl7ecies) repair of
T. heeds pry i,,)i ty d.iintt?d by bleat propeller
scar, (1)10f` Scat-s). fht> 11ij)l0,1(71 ViLused initially on
planting :; n.oivc pioneeri,ig svagaass species, llakn de
to f,uilitarethe eventualrecoveryof'tlu-
41L 1aCT FU1 Ulf , Fi ?'tClt Ta ld err. This SeC[aicnc. C, Lt'Fl;:f'if
nmaur?sc.l succ:ea :rl?;' (fit. Iv offier, pccsc?nal
rc,nuaaunieatitan) proinotes more suitable cw)LH;:ons lbr
tllc slower•growing T tesittainurri to ratrc?.zt h : aturaily
upon the prop sv.;r whii,, fowpotatily aailr,l ling the,
site and preverlin;l, tae},[icmal crosion with a [pore
rapidly gro w`ug aprcit- Box 7.3;. Project Success
was to be yuanufird by t<,ur parameters; (1.) at
planting, a minimum average of tine horizontal
I', r1iizoinr per planting unit must be
in?t.tlle<l (2) 7 <al of'plantr`ng units, at the end
L) r;tr 1; (. ,hoot density (as compared w
nea:1)y natural beds): and (4) achieve the tar} c:.tctc ai o
of bortom Coverage within .a threo-year anallitoriaig
period. Ndd ition ally, of monitoring indicated that
performance standards were not. being met or Y,,, re
not prti;ec wd to ]w me,, romedial plantings of t1low
ca'ffc Ied wort desii;:wd into the plan. iiov vet r,
all rc-ttwdi,rl pi;anta a?3 inset ttu: monitoring else, !err
tb,it piri iisia of the project. The ultimate success
ctiwi-ion was unassisted persistence cl rar >et bottoara
e'e,t.t.t r by the sew,--i,s pl, ntings f?r O-M gals,
tl5ltln ['llf7tal-tlt)CLn7a.II:.tiic?rl tf) prUVade as ai,Taut'a)rl
17a51S of assessment pe7rCpttvi:, and hislf,rit,d
ref-rance.
`Key factors in the Nitiortal Oceanic and
Atriospheric Adnaini'! I' Lion (NOAA)'s dr,eloprtlent cif'
rei o ation plan, h,n%o bec ii issues of prQ-pisgjeci
phtnniiat;, panic ul:uly rr aerdir:g site strirab!lity. Here,
sill" sere resic.vrd fur Uit ahiiity uimf, the following*
cvluri:t Ii thee; t,'Gre adjacent to natural seagrass bads
,it %imil.ir dcp€hs, (2) they were anthropogenically
4r{ltatl:c'tl, ,.;r they existed in areas that were not
sub?ii,rt to chronic storm disruption (41 they wkre not
ondergoing rapid and extensiva• nio of %i rt i olonisat ion
by Seagxrasses: {5) seagrass restoration had been
successful at similar sites: (6) there was sufficient area
to conduct the pmject; and (7) similar quality habitat
would be restored as was lust. '17xe restoration of
seagrass prop scars created by vessel impacts
represented ?vtWV- prc i? rred approach to seagras,
restoration off--,nip, t11 ordai io select a plantin;, sit,
that could :+rcotaanv_xlarc: rile proiect's size. rile ?rncrutlt
ofresli;rrtl ve:, 4+ra5 awirputed using the tl£:A.
154 MARK I`:0NSECA ET AL.
Description of the compensatory restoration
scaling approach
Accurate determination of the appropriate target
scatle, of comlaetas<rtory restoration` projects is lieces-
saty to ensure that the public and the environment
are adequately compensated for the interim service
losses. For injuries to seagrass resources. NOAA Ills
employed H FA as the primary methodology for scat
ing compensatory restoration projects. The princi-
pal concept underlying HEA is that the public and
the environment can be inade whole for injuries
to natural resources throuf-1i the implementation
of restoration projects that provide resources and
services of the same type. quality and comparable
value. HEA has been applied in cases cantered on
seagiass injuries because those incidents typically
ineet the three criteria defined by NOAA: (1) tile pri
mare c alergoty of last on-site services pertains to the
biological f`tuaction of ,111 area (as oppt,rsed to direct
human uses, such aas re(redtionatl services); i2) fea-
sibic restoration projects Bare aavailaable that provide
services of tlae same type and duality and acre com-
parable in value to those lost; and (3) sufficient
data oil the required HLA input paratnetcTs exist or
are cost-effective to collect. If these criteria ate not
met for a partictrla r injury, other valid, ro,hable ap-
proaches and methodologies are available far .scat-
ing the chosen compensatory restoration projects
(NOAA, 1997b). These criteria for the use of HEA
were upheld by the US I)istrict Court (Utdted States
of Atilcrica tip. Melvin A Fisher et. at. 1997 92-10027-C1V-
UAVIS). Of equal importance to the Mel Fisher de-
cision. was the decision by the US District Court in
United Staters a). Arraerko v. Great lakes [)redge Er Dock
Co. 1999 9 2: 10-CfV-131'VIS to uphold the use of the
[ILA as a proper method by which to scab conapen-
saloi-V r'est'oration.
,,,U[t' Ct'Sti?tallull tolor> to dn1 :MiOll ?,Ik,T to
e lia nitefiill rat 9:;d rlf i"e4« .t'?ul7rC C? Illd
rh,lt ucrtir h-091 Illy point of t11,2 injury 11711 fl
1VC,1TI-V Qf thn?f.r 1.0( MO:Set'tii&l tat bdSC1111c. (,)rn=aTSv1i ,
111.1r:"o"inoll r,tar? tit t1.1ton., "hat rt:turt) rL=C iniurOd
n:a. .. _, ?U:t,°5 and `lit olilc i11 1.17 [.es, it n,ay be hirnc>flo-11 1" a}1 laattiv,
Iit Vt)1VLtt te,1 ll 1t7,Cr:'iLt11 t ]?r??j ('Ct 1R'11 t'rP 1f10 i,"rr:dl etl`;c QllntfCt
At its most basic level. HEA determines the appro-
priate scale of tit compensatory restoration project
bu adjusting the project scale such that the present
value of the compensatory project is equal to tile
present value of interim losses due to the injury
of that action fe,g., freshwater diversion projects in-
tended to create wetland acreage).' '['his 'balancing'
of gains and losses is accomplished through a four-
step process (NOAA, 1997a). First (step 1). the ex-
tent. severity, and duration of the injury! (from the
!late of the injury until the resource reaches its
point of maxiraaum recovery), and functional forts
of the recovery curve must be determined, in or-
(let- to ealcttLIte the total interim resource service
losses. Next (step 2), the resource set-vices provided
by the Compensatory Project over the full life of the
project must be estunatted to quantify the benefits
attributable to the restoration, -1-his step is analo-
wous,to tlae previous one and requires estimation
of both the time reciiiired f"or the compensatory
restoration project to reach its tnaxinnrnt level of
service provision and the functional firm of the
maturity curve. After these resource service, losses
and gains have been quantified, the scale of the com-
pensatory project is adjusted until the projected fu-
WIT rCsource service gains are equal to the interim
losses associated with the injury' (step 1). '17his pro-
cess is depicted graphically in Fig. 7.1, where the
scale of the compensatory restoration project is atd-
jtisted until the area under the maturity curve (the
total resource service gains, repro*sc>nted by area B)
is equal to the interim lost resource services (repre-
sented by area A). Because these somices are occur-
ring at (Iiflerent points in time, they must be trans-
lated into comparable present value terms through
the use of a discount rate-
S.ll, iraat :h Cut11(oll? ,kwly plojett C%Ceed tale tort!
c ue-crtntecl (inst?s. ihr4 tiItuanon uoout whe? the "C the
(1 i}cl^.t':i lrolttet t-,tn onk, ht ,cljuerr+3 ac?ordua to tsiilttu
ur stcr'??'ise function rather than i i,witinumi, hurction, clr
Mien the scant i[1t; atnouelt of 1 ourlt at°??>u rceti t'rv'ie i'S
n, rrr?t d he a reNloiation aa.tia;l c,miwt he tts?,lztlti? conrro;ted
t„clawing uillalrndutli.ttaJU.
Seagrasses 155
i? i?s<<?.t At?a
y
U A /
vt ?
Iune
:n
J
m
m
a?
Compematwy Re tOwhon Prutgt
a
Time
iii.''.(. cmtph"d cltvciko Q hmv haboat t1u vatlrr N
a llalv,is t{iI A, scat the
ini,, i ant i,>sN 0- s, ;nutr? ,t°r%icr,. Ihiw ,? ochre u? <t ha
wits ng the total summons ilaert ai'si lust unfit
ccnlipitAt, rercovery hack to pre-oliury conditions Calm A)
is equalled I)N7 tha wrvicos rendert:d under the
runtpensatcov prujict Cocas Is).
Discounting is a standard ecovoniic procedure
that adjusts fin- We public's pmbrerlms Wr hawing
resounvs available in tiler present period relative to
as specified time ill the- future. Becau e 0f discoant-
ing. plantings (hart occur lonnerr after an itnpat I
are worth last; in present-value terms than pl;tnt-
ings conducted shortly atter :ill impact, and there,-
tore more planting must be done as time elapses.
Finally (step 4). appropriate performance standards
associated with the compensatory restoration must
be developed to ensure that the project l:)°twiclts the
anticipated level of'servicets. Well-defined 4nd ineas-
umble: standards are essential to the success cif' the
project regardless of'whether the restoration will be
i ril,Aetnented by the parties responsible fear the orig-
inal resource injury or by ill(, management agency
ift-ustees) isialg monetary damages which are recov-
ered.
In Box 75, we present the outcome of a national
wowkshop that set the stage fr)r NOAA to pm ide
reasonable and fair assessments of injuries to w&
grasses and the cffi)rt needed to recover the last re-
sources which must be assumed by the responsible
party.
The importance of site selection
clearly one of the wi-gent problems with sea,gmss
transplariling is finding in appropriate place to con-
du,?l the restoration aind install the plantings. It
is not advised to plant seagmsses in areas wall
no Nstog of semgrass cgrowtll, or where the afore-
mentioned disturbances have not ceased. Planting
should toil be clone under these ciroutnstances btr-
cau.se ofthe low probability ofsuc? r,, Pl-ml ig may
be done in open. unvegetated area, among patches
of seaginim but on, A. We goal of experimen-
tal manipulations attd;or the evaluation elf plant-
rig techniques Teepitlg in mind that these aniong-
patch locations are heat it string test ()I, tile efficacV
W a technique as they are embedded within viable
,,,iwgss territory). Seagtass patches migiaw altst--
nately colotlising, c'tarreai(iy anse,,;clated sea ilo)or
and dying out where eagrass is located pres(rntly
lMa rha rt al., 1994; Marba & Duarte, 1995; Fonseca
et id.. 199841, 2000b). `giros, the spaces betwc'en ills'
patches today may be naawrally colonised by sea-
grasses in the future.
Campbell et A (2000) provide a decision stral-
e y for assessing; ill(' selection) planting sites that
intltade moasures of, light. opiphyttaaitiotl, nutrient
loading, water mot0n, depth, proximity of donor
site and alternative actions fFig 7.2). Similarly, Fatl-
seca el aV (2000a) and i-'aliunipong et fit. (ill press) give
the hallo wing (.rtleria for tilt, selection of at rt"4 tr<a-
An tine .1t4°av Amn We etdWnal injury slit'.
• It is at dvpths similar to neasil)V seagra*s bads
it was ant1r0t)0g0TlkA1)V disturbed
It exists in areas that are not st.ibject to chr nits
storm damage
+ It is not cindergaing rapid and extensive: natural re-
colonisation by seagrasses
• Seagrass restoration has been successful it sunilar
sites
• Viere is suflicient area to conduct file project
• SimiMrqualhv habitat mTmId be retitored as wars lost
These selection criteria have been used success-
Wily in the US 11det°ml Court as the basis for sea-
grass restoration projects Wrote l Stia:e,S rtf Arrtrrir i
v. Win A. Wher et A '2997. By considering these
156 MARK F?ONSECA ET AC.
Llf tlt
re,'JU tImerlt t--- No..... ?.
r-
Kncoin -
__ I
YL CfiCertake
f t ,4,jatlon I
F I n r I ---- t
tt;3 4??d f't

7iC cq!co I j I
N?. _-

YC?.
t
_.
I Gin site
Isal S tP { - 'No - -I Vfa P c laniy
be c nh anc e'J'l -- -?

Y',
Reiect site ?.

15. et ,rl ?yt<? ? Can ti;It, `-'_ NU ___
weeter <aUalrty
higt,". bt

L
h era f„ ? cr nti?a?f
Yes h ab*lR

r
7
I R 19? .'B i F r X:.'T-i(y C1
ff?;:)t11 Ot CIU 'Of do!1c,r
i
t read;?sr ?,^adU+1?
Iii,; " ";,)..I Close diay,rmm tcL!;#rdmg s,tk, srlrc14un fw. rc,t ?r;euun, 114Mt (:au1pL<ril
rt ri;, t>o(f(?; l,.r[ i0h II tk - I(1s1'1xcnsatwn irr,rdiancc.
criteria, it is ;apparent that trallspl,antation should
probably not be undertaken lot the purposes of c'r1-
hancing rcrccrverV from inatur.rl disturbance events
as these events have both an ecological and cerolu-
tiorlary function in determining the survival and
fitness of'the seagrass ecosystem. When possible, re'
habilitation of the primary injury site should be
performed to restow or accelerate the recovery of
baseline service flow,,, with compenSatorv restor:l-
tiotl used to cprnpCn,iAtC the public f0r interim ser-
vice losses that ILCIAIC While the site rCachcs its pre-
injury levels of service provision
Critical factors influencing transplant success
NumCrous fector5 havc been dt?1Vl1nirlCd to affect
transpl.antim, success. Some are of the crop-riyl<
type, are extrinsic and cannot be controlled, oth-
ers involve issues of protocol, in a SLIYATy of North
American seagrass planting projects, Fonseca ct al,
(199811) listed the following as factors that had the
potentiai of being controlled by those conducting
transplants.
• Siniilarit. of environmental conditions of donor
and reC pient beds.
• Choice of species. preferably same as that lost. but
pioneering species inq be substituted to initiate a
project.
• Presence of grazers or sediment burrowers: these
bioturbating organisms may need to be excluded or
plantings naayhavc° to he conducted in large patches
to dissuade them from their activities,
• Source of planting material: similar depths find
EInvironnlcntal totlditions grad from over as broad
a geogiaph'ic area as possible to ensure genetic
diversity
l'inic of vear: u>agrass should be planted at it time to
ensure the longest period before seasonal stressors_
• Cost: many v:ari.mons of cost have been given but
standardised costs are elusive; based on recent cases
ill the IIS federal Court, it contracted project that
Seagrasses 157
Box 7.2 The need for physical stabilisation
after boat grounding: United States of America
v. Great Lakes Dredge S Dock Co.
Another example of the difficulties faced in primary
restoration of'seagrass habitat was recently illustrated
in another Federal Court case (United States gj*Arnerka
v. Great I,akes Predge 6 flack Co. and Coastal Marine Towing
lnc. 1999) where a large tugboat grounded on a shallow
wagrass-Pontes coral bank in 1993 and destroyed
7200 m" of habitat, The grounding site was located in
an exposed, high-energy environment where seagrass
transplantation was deemed inappropriate. The expert
case team assembled by NGAA and the LIS
Department of justice recommended that t lie primary
restoration plan should include filling and regrading
the trench made by the tugboat to physically stabilise
the site. It was assumed that once the site was
physically stable the scagrasses would recolonise
naturally: but slowly. Interim losses of seagrass would
be compensated off=site in a plan similar to the plan
for the United States of America v. Melvin Fisher et at,
described previously. Attempts to negotiate a
settlement with the responsible parties proceeded for
five years without a resolution. In September 1998 the
site was impacted by Hurricane Georges which
severely damaged the partially recovered portions of
the injury and effectively set back the recovery clock
(Whitfield et al., in press). Seagrass 'beds adjacent to
the injured area were unaffected. The impart of the
hurricane confirmed initial concerns that the
grounding site was physically unstable and vulnerable
to further injury. Clearly, there are caws in
high-energy environments where physical restoration
is needed to stabilise injured sites to promote the
recovery of seagrasses. In the primary restoration plan
for this case, the amount of sediments excavated by
the grounding; called for substantial in situ
engineering to recreate the bank structure. This is not
surprising, since it took nature between 500 and 1000
years to form the bank that was destroyed. Recent
studies have shown that large vessel groundings are
becoming more common on seagrass banks in south
Florida and elsewhere in. the Caribbean (6trhitfield
et al„ in press). W'hitfic:ld et cll. (in press) have
documented the instability of these injuries and
recommend that physical regrading is necessary to
prevent further damage during severe storms.
includes site sunVeys. planting, monitoring and re-
(aortirag will cost (in 1996 Us dollar,) L1S$630000
per hectare.
It is CSSerrtial to study the substrate-energv ( ' ex-
posurO regime, and optical water quality (clarity
or light availability) of the area that will be trans-
planted so that suitable source materials can be
identified. Areas exposed at low tides should be care-
fully mapped so as to place plantings with mini-
mal exposure to air, unless the plants are regularly
occurring in the intertidal zone (e.g. in the Pacific
Northwest of the United States). Planting in high
wave energy or tidal current areas will require plant-
ing in larger groups to avoid disruption (Yonseca
et (it., 1998a, hi. Planting in larger groups also ap-
pears to be an effective method of deterring physi-
cal disruption of the planting by marine organisms.
However, as suggested by Addy (1947), matching wa-
ter depths, temperature, salinity, water clarity and
pliant We remain sonic of the best general guide-
lines Cor matching donor and recipient beds.
The characteristics of the species, such as fast:
growing vs. slow growing; pioneering vs. climax, an-
nual vs, perennial growth, etc, must be considered
before conducting transplantation work. For exam-
ple, Hulodule spp, and Halopiiiia spp. are fast-growing
pioneering species while Thcalassiaa spp. and lfttlalus
spp. are slow-growing climax species, flalophila spp.
rapidly colonise disturbed areas like those with
moving; sand bars and are under-canopy species,
requiring low light. Although it climax species
may have been disturbed. it is often advisable to
first install a faster-growing species to stabilise the
environment.
:Another important factor in the selection of sea-
grass for transplanting, besides their intrinsic recov-
ery rate, is their growth habit {Short & Short, 2000).
Transplanting can be rendered almost wholly )net-
festive if' tneristematic regions of these plants are
158 MARK FONSECA FT AL.
??inlul-(?ieri•-tcm tti?:? la;ef-reph?.cintt
M
All
lturnr-mcrisurm:lll? ?un?it•t+t-±??p1:11:mg.
z
M
U1 me(isternatk iVA lepiaculg
M
E)i-nrrrltil(•urltii nett-I e:11?rU)lacin"
41,
N1
I ig. Ili( !our ba is ?,rowtli forms of xagra<s. In rh,, Lasc of ttx mone>-mcnstemauic.
k•atr('plaru,,k Klrm. matt! t(•rmill:d shoot ()n true Iminrr is a tiai)l( hlantinh omit -
L(?11 )a1'a11 A'LIV 14A% Ill(A.14114lC lI,m !N Ili 01"W IlUiIIlIII , Ihal I IsI (l(It,I'1 ('u(II sIIoUl Ik;I?
ttl;yt: pllrenr;al tier )III illuatin%4 to spati,II iU1CIT ti5-ti1m s11 L' Utllt•I till'er' h0InIs r(A quirr :1t
IrasI i1ir1'r to I-our short sht")Is hr mainlelin;rd c?±?. tIlk, Runner for a conlpiete piaulting unit.
'IN %k('11 ;1, an *Intact ri).izome° apa al merislon". From Y-Ion R Nhort ;2000j.
damaged or not incorporated in sufficient quantity
in a planting unit to initiate recolonisation. Short
& Shorl (2000) summarise the morphotypea of sea-
gr:ass it ig. 7.3i.
Stiy;r,'155 grazers can have dlSiIStr0115 vflvlcis on
plantings. Scat rass grocer im"lidt, sea Ur(htn,.
gTastro,pods and herhnvorous fishes. Sorue tttigra
t0l?y water-tiJwl such as geese anti ducks haV('
been observed to decimate seagia;s plantings )per-
sonal observation). Significant grazing of natural
Svrulgodiurit fili/ur,ntbeds points out the general
susceplibility of sctgrasses to grazing (Rose et al-
1900). Fonseca et al. 11994 and references therein)
totted significant disntrbance by rays in 'Campy Bay.
Florida, indicating that it would be necessary to
use exclosure cages to ensure the survival of trans-
plantecl scagrasses in some areas. Recently, we have
seen that planting, in clumps of at least 20-50 cin
on 'i side deters many :nlimr is from disturbing the
F1Llntin s (authors' unpublished data).
yfinjtnisatioll 0f'dIsturh,III(e to the source bed is
paramount in scagr•ass transplanting so as not to ex-
accrbate injury to local populations. With present
techniques focusing on the use of wild, vegetative
stocks, this ni, y be achieved by conducting the
transplantation in phases, or dispersing the collec-
tion effort. thus allowing the source bed to recover.
Harvesting of donor stock should also be done from
?' M
Seagrasses 159
1
Sexual Propaguie 1 Vegetative Propagule
?, ( ? Is fiizome
Seed production No 1 I growth rate No
high?
YPS Reject , Enhancement
t irensplantamt i 1 Yes- _ possible
Is seed 1 I No
viability high? No l 1 Yes
( - Reject
Yes planting unit
i
1 derislty of ( No
Is seedling
establishment No - 1 active growing Try larger
high? 1 i shoots high? plugs OR I
Potent+'al with ( increase
Yes habitat 1 Yes density of
enh incement sprigs
Consider Consider sprigs Consider plugs
I Lseedlings for 1 for transplant for tr 4nsolant
ranspkanl unit _unif u.
r „ t
Minimize handling
Fib;. 7.4. Recision irctc Inr choosing seedlings or whole, rnature plant,, for transplanting.
We caution that the tecluiolog)" for seed r-ttablishment is not as developed as for the use
. Front Campbell of al. (2000).
of sprit or (ore,
beds over as broad a geographical area as possible.
This may help avoid loss of genetic diversity in the
planted beer (sense Williams & Orth, 1098) and may
actually incorporate the full local range of frenetic
diversity into the planting.
Fortunately, many pioneering species can be
harvested with minimal disturbance to the beds
(Fonseca et al., 3994). However, for climax species,
harvesting of donor beds may cause long-lasting
damage and harvesting from these beds should only
occur when the beds are under some anthropogenic
source of physiological stress that does not. seem
likely to abate or if they are in imminent danger of
physical removal (e.g. dredging).
The size of the source or donor bed should first be
assessed to determine if recovery will proceed `after
removal of the sods, cores or sprigs. This is especially
true when transplanting vegetative stock, as a large
amount of material is needed. Spacing harvesting at
-0.25 to for small cores or sods (.: 0.15 x 0.15 m) is
often sufficient to avoid long-lasting damage. More-
over, Fonseca et al. (1998x) suggested that Kuppia,
Halopf la, Hulodule and Zostera spp, can recover in
small patches i<-0.25 m") within a year with shoot
density returning to normal. Furthermore, Fonseca
er ill. (1998x) cautioned that patches ::.-,30 m-' in
high-current areas may never recover. Campbell et cal.
(2000) also provide a decision tree for selection of
planting stock for troth sexual and asexual propag-
trles that focuses on intrinsic propagation rates
('Fig. 7.4).
Choosing the time-Frame for planting is an obvi-
ous concern, and as with all crops, the appropriate
time fbr seagrass varies with geographical region.
In general, the best strategy is to plant at a time
just after the period of highest seasonal stress, when
natural populations are experiencing recovery. For
example, celgrass (Zostera marina) should be planted
in the autumn in ]worth Carolina, and other mid-
Atlantic regions in the United States, because sum-
mer is the period of maximum physiological stress
at that location (Moore et al., 1997).
160 MARK FONSECA ET AL.

hg la;,qW,t Mok n.pWWmatwn or ns,> m?.nt
VVIde.; tutIA IC.tt'i', INs III dIIIi IIINtl:?u_Ib: tGI? ,, tti1li
rdimrnt stacta ,t, , WTI Id piIli +s .,Hnd these vuitlat11
diin?uu, t?su,tlly aDLITIVA "Oh Uinta; nr woul staples
Planting methods
Fonseca et al. (1998a'i list 14 categories of planting
methods fc>t seaArns in the United Statey lbr these
methods, source matena_I can be vegetative stock or
seeds. Transplantation using vegetative stock typic-
Ally requires available wild stock as a source and
is labour-intensive and invariably expensive. How-
ever, it often gives faster, more rehab)e (vverage
than >eecl methods abut see review by orth ct al..
2000). Most I)rraiCcis t?xlay are Carried out wing ei-
ther small scads or sprigging of sediment-free units
(Fig. ?.5i.
Sediment-free methods
For most sediment-free methods, plants are dug up
using shovels, the sediment is shaken from the roots
and rhizomes and the plants are placed in [loin
ing seawater tanks or floating pens. For vegetative
stocks. Fonseca et IT 0998a) recommend a mini-
mum of one apical shoot per planting wait. Thee
number of short shoots on a long shoot should
be maximised whenever possible, so as to derive
benefits from the clonal nature of the plant. Also.
the plants should be collected and planted on
the same clay kept in water with the same am-
bient temperature and salinity, and kept as moist
as possible when out of tlae water. When using
vegetative stocks of Thcaltassia testatdinurn. Tomasko
et al. 09911 recommend a minimum of one rhi-
zome apical and at least three shoots per rhizome
wynaent.
Seagrass should he planted either directly into
the bed (sprig) or anchored using a variety of de-
w"n such as rods, rings, nails or Rebar. U-shaped
metal staples 1Vith .attached hare root sprigs {no
sodimeritl have been widely used as planting units
i,Dcrrenbacker & L(,wis, 1982: Fonsma et uf., 1982) or.
when negativv huoyaniy is not required, bamboo
skoivers may be suhstituted iDavis & Short. 1997).
Plants have also been woven into biodOgradablc
mesh fabrit: and attached to the sediment surface
as a planting unit i,honseca V1 A. 1998ct). Rocky in-
W11 1d,11 species. such as T4,a•Ih,llxulix spy, have been
attached to boulders.
When using anchoring devices, one must cola-
Sider using biodegradable or natural Materials such
as boulder over metal or plastics. As mentioned
above, when using staples, one can choose metal
I S50.01 each) or csin intuhl r Wsh kebab' bamboo
sticks by bending, them into a V iDavis h, `;host,
19971 which when purchased in bull: could cost
only USS0,006 a piece. In tropical areas where batn-
bnr is plentiful, this could bc' amore economical
medium t„ use. Ratnbtm is also hiodegratlable. Using
either ),in(] of staple, planing units are made by
grouping plants and attaching the root-rhizome
portion under the bridge of a staple and securing
the plants with a papernwated metal twist-tie. Fhis
tan either he prepared beforehand or Be planting
unit can be pinned cfirecfily to the substrate during
planting.
When using nails. boulders or Popsicle sticks
(NW &L 1988L the teclanigti(' is more or less sim-
ilar and the planting unit is tied to the anchor-
ing irtstrunu°nt. Frames, such as Short's T RY device
if,. 1. Short. in Fonseca et tai., 1998a), have great
promise for rapid and not-diver-assisted planting at
depth. A cage deployment system that has shoots
attached to the bottom is lowered onto the sea
door and retrieved after the shoots have rooted
and their paper As have decompose? This ehm-
inates the need for divers in deeper water, can
be used in chemically polluted :areas, and provides
initial protection of' the plantings from biological
disturbance.
Seagrasses 161
5eagrass with sediment methods
I7ae sad or turf method consists (Aplanting a shovel-
full of seagrass with sediment and rhizomes intact.
This is the easiest method, and is most applica-
ble for hard, compact substrates and sleep-rooted
and large species such as F?nhulus acoroides. The only
equipment needed are shovels and barge basins- for
the sods. However, if' the donor site is far away,
transporting the sods may present a problem as
the weight of the material is a physical burden.
Soane species. such as F. acoroirles, Posidor is spp. and
Thalassia spp, may have very deep root-rhizome sys-
tents requiring removal of a tremendous 4arnount of
sediment to harvest the belowsrotind plant struc-
tures all intact (Fig. 76), 'To our lcnawledge, this
has only been accomplished in We-stern Australia
jby E. Paling, of Murdoch University; see review in
Fonseca eat all„ 1998h). Furthermore, hayA'esting am
entire sad may constitute one of the most severe
perturbations in a seagr<ass meadow, inhibiting re-
coven, in the donor bed.
-p?.w y,?"'" pry
x??.. ..
Ili , lr%T5r uS 1 1: - a'
The plus; method utilises tubes as coring devices
to extract the plants with the sediment and rhi-
zomes intact. Ilse plugs are planted directly into
the seagrass bed after creation of a hole: to r(Iceisrc
the contents of t he tube. The core tubes are usually
made of 4-6-cm diameter PVC; plastic pipe with caps
for both ends to initially create a vacuum and keep
sedinuuni? front washing out the bottom. The tube
i, iv,,t t I d into the sedinent. capped twhich creates
a vacuum), pulled from the sedinient and capped
at the other end to avoid losing the plug. This can
only be done with soft but cohesive sediments and
generally only far small species to aatoid excessive
leaf shearing (unless extreme care is taken to :avoid
the shearing„ which adds measurably to the cast.
ofthe process). When the donor bed is far away from
the planting site, many tubes are needed which also
adds to they cost.
Sod plugger", extract a plug out of' the donor
beet which is then extri.tded into a peat pot; the
method vvas first used by Robilliard & Porter j1976)
2

162 MARK FONSFCA ET AL.
and modified by honscca et al. (1994). Recause these
pots are typically ono a few cm across. they may be
inserted into the bottom by liquefying the sedinicni
with a hand tool. after the peat pot is planted. its
side wails must be ripped off nr torn dawn and the
pert pushed into the sediment to allow tho rhizcnncs
to spread out
Sowing of seed
Seed planting; holds promise by l rr oscale reswra-
tion but is currently mere applicable ranly in low-
energy areas where the seeds can settle and gerrn-
mate and vvhere there are few seed predators.
This method was first introduced by Thorhang,
(1374) with Thalassia testridinurn. A seedling; g;rbw-
out method for T testudinuni has been registered by.
Lewis 119{47). The availability of seeds must alsci be
considered. Large areas in Be Chesapeake Bay have
been established by sowing seeds from a small Not
fR. 1. Orth, personal c0nununicati0y)). VV01'R ce>ntfn-
ues in 09 highly promismg area (Orth et ell., 20001
Experiments using seects prlletised to increase their
density to facilitate sinking and seeds cinhedded its
biodegradable mesh are presently being; curried out
by Granger and his colleagues (Granger of & 2000).
Experiments on planting depth also indicate that at
least toy 9os4,rri rNuirra, seeds should be within the
UT 2 cm of the sediment for best germination and
that sowing densities should be 400-1000 seeds per
sdu are metre (granger et al_ 2000).
Laboratory cultured stocks
This approach uses plants reared and grown in the
laboratory from plant fragments. 1t I= become es-
pecially applicable for large-'scale plantings where a
large amount of plantiry units is needed. This tech-
ruclue also holds promise for reducing or clituinat-
ing donor bed damage and this has been shown to
be rnirmnal by pioneering species, such as HaAdule
twrightii and Sy-ti ngodruni fili16rrue Wonseca el al.,
1994). This approach also has the potential to main-
tain donor stocks for unscheduled plantings and
could theoretically supply genetically variable and
disease-resistant plants.
Several aspects of this approach remain con-
troversiaL So far, three species have been suc(vss-
fully propagated in the laboratory, Ruppia rnaritinui,
iltalogihifce d4QQns and 11 rngelmamin (M. Durako,
personal communication). Ruppia rrdritit?id has been
successfully transplanted from labtirator,y ectlture
stock (Bird er aL 1994), but all these species are
naturally fast growing, ti.e, pioneering) and it is un-
c fear whether Laboratory culture is as cost-offeo ivc
mom% of restoring naturally prtmh& spaAes. More-
ovcr, questions regarding; the ability to maintain
rrenetic tarr.tcttn-r of the: 1101?111ation have nog, been
salved. Given the growing ermplams on me•charuwd
plantings using wild :stack, laboratory culture will
probably only be cost-effective when tccliniques are
developed fiar slow-growing species, hence avoiding;
long-term donor bed impacts.
Monitoring the restoration
Monitoring; of the restoration project is necessary
to provide data regnimd to ea,aluate the viability
of Be project based on then performance standards
(defined below). Fhis pern&% timely identification of problems or conditions that may require corrective
action to ensure the suvrem of the project.
Monitoring schedule and activities
Field adlectiein of data by perfiirrnanev uatinM,
ing; should occur fear tout 'years after planting. Orig-
inal plantings should be monitored for three years
and potential remedial plantings in year 2 Should
be monitored tot- three years for a total monitoring
period of fiiur years. Under this schedule the moni-
toring xvould be conducted as follows
year 1 - day 60, lilt), 365
year 2 - day 180, 365
year 3 - day 180, 365
year 4 - day 1$0, 365
The precise dates are weathendeperndem. W carbon-
ate sedinients, each surviving, planting unit "110111d
receive an additional spits of constant-release phos-
phorous thodiser 10-39--0, nitrogen-phosphorus-
potassium) al day 60 of year 1. Mternatively, bird
rociging stake could be installed about every
5-10 in along scmi {see= Rox 73).
Data Collection
Mordwring should {locus on documenting the num-
bers of apicals at planting time. planting unit sur-
Oval, shoot density and areal coverage under the
Wowing schedule and definitions. This numironng
Seagrasses 16;
Brix 7.3 Transplanting strategies
RESTORING SLOW-GROWING SPECIES
i'1ur,6co J aal, (1992) developt-d a stt iiv y tor resturi;lg
?Io +-s;r?t"iin ? specirs such ?s Htata t, !e?-ta,r,a?t?m b,,
nltrtiic?kin<_ ,,;+u;zf hu<<csvcn i7tcy ?tii[iu4i4? Mart
anatl;c•r?_ - {_t??t?-,htc,?dut -.,?n?c?f?er. ??.?rh ?e i. ..'?;a°
t ?t'ti7 ', ,u.?l,lc't'r ?rOitti?r? 5,i'ci ,,., c :'15{Oal' BSfC itc?s ?.7 !.
''.his tenlpoTarq Suh?titutiun t?f:a i.?titef•prnsving'ppr•cir?
for a sloes-growinc tint l?r ern<rtc5 ,i mon. smr>able -
condition for the + i?'i?'c1'1ng onc 10 C?tat7li5lt ilwil
while the faster that' :rahll,,.; th,l scdiiiwiit aril
provides a functional sea r, s Iml)ttat
tiascd wi a series of ecofo ical Held experiments
if owqure,ut et tai, 1993) and traubp! aittusg, studies
ji t n tlt- < I t+i.., 2000), we have ticlcrm ned that
tonilxcswd sac-e.,ion`of H. WII.s+.L; can be Oils, ncc'ci
U v fartili, i u'atl,l?l;i_tts with bird rt>t7:un , 7..atcc?. ,`ft
rrtxcrltu•ti° t?orki hs lal??rin:; bird r,wstin>; ?L.ke? cn?cr'
the rr.;n}'tartts_ ic?al?irils raostin:q vn thr,tay;e?,
d,-It°ca+te n nric?n€s itu?i tltc? w?atcr, stinwl.rtit.; [;,}iid
?r,?xth r _ r?in,pla!t Is ]n , I,(" ii11,?11, 1!;ere
a r lx _} of and H, ,
»e ri of an area can accelerated by
placc,neaat of stai,e ::lone.
STEPS IN A GENERALIZED PROTOCOL
+ Study the site to be restored and determine at
the followinst parameters; (a) seagrass bed histor-
(species composition, cause of lass); (b) expo?ure3 rn
air and wavcs and currents 1. ; substrate
I pc - rvwd cl.n`s and high w" ,iotic sedr ratios.
;[ti I,itV ut si;tatia?n - }?iantc viten cannot tltthsratad
much more than 25"< huriai and vertical g;r0wLh is
13o'0hle ariiV ftor same pc6ecs and gel presence; of
,17tlttaal di,tu0mlicr.
• Detemono tine i an,r .tnd l?ttci;ct ,3V t siuatin„ the
?hicat, ,?i iat' Ietlttircr,?er,rs, Mclkcl 11992_
protocol applies to original plantings ft if three years ?.
(years 1-31 and io rernedical plantings for three Years
(years 2-41
estinasated a minimum of seven persons were required
for intertidal, bare-root (e.} ., staple techniyuet and
nine persons for suhtidal bare-root planting. The use
of scuba incurs higher costs that need to he factored
in. In the Philippines. Calumpong et cal. (199:2)
at eomplished planting; of 32 0.125 m2 sods in the
ntct int l sul?ti(hl with nine persons in one day.
}'f,.? eI t " rt; pr wide :, , mpa ison among?
'cr.d iairilautl, .isiir", li;7t(d tc w"'f`l.cyfound that
ci)11I?ciion > fihriL,mmi pistun costs ranged
fiom t-3 to 3.5 minutes her hl,ini;;tt; unit with peat
put plugs beitw tint, infk?i :,ipid ttictlcnd ;tn{ iCx°es
being the slowest method.
It is imperative that one recognises what is not
timed in that report, such as mobilisation and
demobilisation, bath daily and on a project-scale
bae4s. travel, rcipor:int= i,d zas??ritoring. The timed
11ial, also =did not .rccuaa;,0ti rnc,?surc the effect of
boor dom oil the slat'e'd Write p_axz s?
+ Lm,,Ite a donor beet ihat ,rnarchc, th, conditions at
the planting site, Pm ?c?;tr,?tiar tncihods, this
should be near enough ?o the ,Lott;: tan iae planted
the same clay. Overntgl nr:t: ,;
particularly bare-root inau-_d. ,,..+aid hr lilaced ill
moving seawater at ambient ternp2t,itulc'i and
salinities.'1o our knowledge, there has not been a
sufficict,tt? controlled L,%rerixnent t.o determine the
+ Ft. pwirtied to nrulagc t: _ r''?,rce with regard
to the Icdit,tn ut teaks. l1; .; borciL>rn, varying
t,t=`,?s tutttn,?* tndi+?idual? t., . r :1 iE,cful strategy.
* t,:,lrefull. dclirtetttE° plot, io t,ttllitatc, I,,uznitoring,
+ Ct,rr,itS?tr .a31 th.- a:urenacntirnecl p;Itrtiti.+l «5ts,
71)e-5e include <lie tl0inc:ati0n, rcp<;ru, tnubili,atiota
and dvIll obilisaIloll, insol,tme. ovclliead, hencht"
napping;, planting operalioll', n,oniuirint!, remedial
planting and a IW profit rnars;irn Icu crantractors,
+ Conduct thorough moltittit i-at; (s+ o below) and be
prepared to conduct pLuitings.
1, typical counts. Prior to planting, one planting unit
out of every 100 collected should lye examined for
the number of rhizome apic,als. 3
Survival, Each site should he examined for survival
of all planting units during each survey in gear I
(days 600 180 and 365) or until coalescence. Survival
of each species should he expressed as a perceni-
ag;e of the cyri „inal number. but the actual W1101e
number should also be reported.
Shoot densitY. A Separate (front survival) random
164 MARK FONSECA ETAL.
selection of three planting units per 100 planted
should be assessed for number of'shoots per plant-
ing unit at each survey time until coalescence be-
gins. After some planting units begin to coalesce,
three randomly selected locations per 100 m' (100
planting units) should be surveyed fir shoot den-
sity over a 1 m' area at 0.0(525 m' 125 ctn-.25 cm)
resolution. Shoot density should be monitored for
three years.
9. Areal coverage. The randomly selected planting;
units (may he the same as shoot density selection)
should he surveyed for coverage at each survey time
starting at day 180 of year 1. Measurements should
be taken at a 0.0025 1n' (5 cm x 5 cm) resolution
prior to coalescence and over a 1 m' area at 0,0625
111' (25 cm .t 25 cm) resolution after coalescence for
each seagrass species present at each survey title.
Areal coverage should also be monitored for three
years.
5. Video tape transects. Five 100-tn transacts along
randomly selected portions of the planted area
should be video tape recorded to establish perma-
rlent visual documentation of the progression of
areal coverage of seagrass through time. ,A tape
treasure 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 takers at each
survey time during the monitoring period of three
years. Obsenfation-based assessment of success may
be substituted if quadrats are trscd in accordance
with a Braun-Blancluet survey method (Fonseca et
al., 1998x) 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. surrey each scar). Similarly, Braun-
Blancluet 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
of ail sdeo tapes and other photography should
be turned over to the permitting agency follow-
ing project completion by the party conducting the
monitoring.
Remedial plantings and/or project modifications
If' data fionl a monitoring report establishes that
the performance standards are not being; met or are
prajcct d not to be met, remedial plantings of those
Box 7.4 Performance standards
Although it is the overall objective to restore the
species that was injured,, performance criteria may
also be based on the success of planting of
pioneering .seagrass species, as found when Flalodule
is installed to expedite the recovery of Tturlassfa
(Fonseca et al.. 1998x).
APIGALS
A mitiiniutn average of one horizontal rhizome
apical per unit should be maintained in all original
planting and remedial planting.
SURVIVAL
77ae survival rate shall be considered successful if a
tniniuaum of 75% of the planting units have
established themselves by the end of year 1. If it is
determined that less than 75% survival has occurred
by the end of year 1. then remedial planting should
occur during the next available planting period to
bring the percentage survival rate to the minimum
standard by the next monitoring survey.
G GROWTH
The third success criterion should be the measured"
growth rate of bottom coverage. The growth rate
should be considered successful if, starting after one
year, the planting is projected to achieve the total
desired acreage of bottom coverage, with 95%
statistical confidence. within the three-year
monitoring period for original plantings. If this
t ritedoo is not met, then remedial planting; should
-`occur during the next available planting period.
Seagrasses 165
affected seagrass species should occur. If there is
a recurring problem with survival of plantings or
replantings in a particular area, remedial 'planting
should occur in another suitable area in as close
proximity as possible. subject to the approval of
permitting agencies.
Rased on past experience in seagrass restoration
efforts, it is assumed that 30% of the planted area
should require remedial planting in year 2. All orig-
inal plantings should be monitored for three years.
Remedial plantings should also be monitored for
three years.
80x7.5 National workshop for defining the
metrics of assessment for seagrass restoration
projects
Nation;d Oceanic and Atmospheric Admmjn ,trauon
iNOAX is &-sig;nated as a trustee for natural resources
under several United States lass' and, lot that
capacity, is authorised to act on behalf of the public to
seek compensation for injuries to its trust resources.
Under each of these statutes, natural resource damage
_claims are composed of three basic components: (t} the
cost of restoring the injured resource to baseline;
(7) compensation for interim lost resource services'
from (he turxt orthe injury until restoration occurs,
aml 131 171(i cost ofrerfo:-ming the damage
Habitat equivalency analysis (HLA) is one of the inore
avqucntly used methodologies available W iaatural
resource trustees for calculating they appropriate scale
of restoration projects necessary to compensate for
interim resource service losses (Fonseca et eaf., 2000<a).
The basic appratob underlying HEA is to determine
the amount of eoniprnsahory habitat to be restored,
enhanced and,'or created, such that the total services
provid<-d by the compensatory project over its
tunction,tl like pan .tie equal to the total services lost
due to the resource injury.,
While [TEA is conceptually and computationally .
straightforward, proper application of this approach
requires a detailed understanding of the biok?glcal -and
ecological processes that affect the recovery and
1'40rahk% tilt Co-prchcusivc L^nviroomeutai Rcsponsc.
U11-ri cI1,11iun, Liability; Act the Clear, GVater An. the
National Mann, latter sties Act, and Ow Oil Pollutioit Act
of 1190.
Se 1 i,'es here rarer to the ILnctiuns .hat one resource
pertorrns liar anc,thel of f<+r hvmaYts.
productivity of injured and restougl habitats. In order
to gain a better understanding of these processes,
NOAA is undertalciri; a systemalic, expert review of
the ecological assumptions made within the HEA
framcworl, for a number of habitats for which HEA is
most fre?luewly applied. Seagrass habitats were
selected as the first habitat to review for two primary
reasons: (7) N0AA expects that due to the frequeno, of
injuries to seagrasses (more than 70000 hectares of
injured seagrass in Florida alone), HEA will be
commonly applied in cases involving injuries to
seagrass habitats. and (21 the relatively small number
ofspecies of .,eagrasses present within areas under
NOAA's trusteeship made this habitat a logical
starling-point Cron) which to develop and refine the
teview process for more diverse, comptex babitats,
WORKSHOP PARTICIPANTS
Under the direction of NOAA's Southeast Fisheries
Science. Center, several academic experts in addition to
NUAA staff were selected to discuss the underly in
assumptions of HEA for seagrass environments. The
woriloroup was assembled to be geographically diverse,
:,s well as to reflect a range ofspeCiaitics within the
field of seagrass biology/ecology. "Die workgroup
participants were: Susan Bell. University of South
Florida; Kenneth Moore, Virginia Institute of Marine
Science; Mary Rut kulshaus, norida Siate [University;
Frederick Short, University of New Hampshire; Charles
Simenstad, University of Washington; Mark Fonseca,
at`or a detailed discussion of HE:A, sec Nnrifzmal oceanic
and Atmospheric Administration, Damage Assewiivnt and
Re>torufoT7 Program (March 1995. revised Ocwbct LQ00),
habitat EgviV31euCy Analssls_ An Crvertiew, unnublished
report.
166 MARK FONSECA ET At.
NMFS Southeast Fisheries Science Center (Modt,r:aTm):
John Cubit, NOAA Darr age Assessment Center; Briaq
Julius. NOAA Damage Assessment Center; Arthur
Schwarzschild, NMFS Southeast. Fisheries Science
Center: Erik Zobrist, NOAA Restoration Center.
WORKSHOP CONCLUSIONS
Prior to the workshop, all participants were provided
with background materials on the theory and
application of HEA. as well as a series of null
hypotheses developed by *IOAA to capture the major
issues relative to implementation of 1-I A for seagrass
environments. Each of the null hypotheses discuss
presented below, accompanied by a suniniary of the
conclusions reached by the workgroup.
Null hypothesis is Recovery of functional attributes
can be forecast based on seagrass biomass and
density alone4
The general conclusion of the workgroup was that
seagrass biomass represents a more comprehensive,
metric of habitat htnction than seagrass shoot density.
In general, biomass was cited as preferable to density
measures because density measures do not capture
the below-ground component of seagrass systems,
which may be important in determining the long-term
persistence of recovering systems tparticularly fiar'7ba
lassiaa tesivdinunt); and seagrass shoot density develops
much more rapidly than function-, and may overshoot:
haseline )hoot density before achieving equilibrium.
Canopy volume (shoot density times the height of
the seagrass canopy) was proposed as a preferable
measure to biomass. While in most cases canopy
volume would be expected to be highly correlated witli
total (above- and belowground) biomass, canopy
volume has the added advantage of being easily
measured in a 110n-destructive manner. Despite the
apparent advantages of the canopy volume measure for
capturing within-patch functions, the workgroup cited
among-patch attributes of landscape structure. scale
and setting that would also be expected to significantly,
influence functional performance and recovery rates.
4Functional attribtttc,, ref r to the range of ecole}gieal
services provided by seagrass habitats including, but suit
c; _ e r aIud connectedness among vegetated patches
were among the measures cited as important in
capturing the among-patch aspects of seagrass habitats.
Null hypothesis z: Forecasting seagrass bed recovery
is independent of scale
I"li general conclusion of the workgroup was that
seagrass bed recovery likely will be influenced by the
spatial scale of the injury. One rationale provided for
rejecting this hypothesis wets that current data suggest
seagrass beds inodify their environment, and thus the
size severity and shape of an injury will alfec:t the
recovery` process, Other participants suggested that as
thc'scale of the injury increases, landscape features
will be increasingly important in determining and
restoring functional attributes of seagrass beds. Stated
differently, participants expected differences in the
recovery of a continuous seagrass bed versus a beef of
the same total density or biornass, but distributed in
discrete patches.
Dull hypothesis 3: Overcompensation responses by
seagrass in injured areas (e.g. generation of shoot
densities higher than un-impacted controls) does not.
constitute enhanced ecological/biological functions
lbe general conclusion of the workgroup was that
overcompensation responses by seagrass in injured
areas do not constitute enhanced functions. Higher
seagrass density does not necessarily indicate higher
production. In addition, external controls, such as
light availability, will serve as limiting factors on the
function of a particular habitat. The workgroup also
concluded that the presence of an observed
overcompensation response may be an artifact of
density-based measures of recovery, while
biomass-based measures for the same area may not
exhibit the same response
The outcome of this workshop set the stage for
NOAA to provide a reasonable and fair technical basis
to assess seagrass injuries and support the recovery of
the lost resource services for which the responsible
party is liable,
liniatetl ia, primary production. faunal use, nutrient
filtration, and sediment stabilization.
Seagrasses 167
'fable 7.1. Topg General distribution Pln) of costs b'y tosk
(United States of America v. Salvors Inc.); bottom:
suannt(lTy of costs by spec-z('ic actions (Fishes Natural
Resource Damage Assessment claini)
Percentage
'i"ask of total costs
Map and ground-truth 5.5
Planting 185
Monitoring 58.7
Contractor 8.3
Government oversight 91
Type of cost
US$ (1996 valued
tlarrias;e assessment costs
Federal assessment costs
(up to 26 October 1996) 211 130
Interest on federal
assessment costs at
judgment 26 553
Subtotal 237 683
Restoration costs
Primary restoration costs
(vessel-generated holes
in sea floor - restoration
deemed not feasible with
current technology) 0
Restoration site selection
analysis 5 465
National Environmental Policy, Act
compliance!pertnitting costs 14695
Preparation of
mapiground•truthing sites 14 314
Collection, preparation and
installation of planting units 64846
NOAA restoration
oversightjsupervision costs 17 650
Subtotal 116970
Monitoring costs
Monitoring of compensatory
prop scar areas 205650
Contractor profit on
restorar[c wn'1onitoring work 29028
Grand total for claim 589331
Costs of restoration
From our experience, there is a general set of fac-
tors that drive up the cost ofseagrass transplanting,
particularly inappropriate site selection, inexperi-
ence, and disturbance events (requiring remedial
planting). Consistent estimates of planting costs in
dollars retrain elusive, but recent restoration plans
in the United States that have been litigated in the
federal courts have shown the full cost of a restora-
tion distributed among the various tasks (Table 71)
at - US$590000 for a 1.55 acre area or -TS$940000
per hectare (1996 dollars). Two in7portant points
here are that: (1) the actual costs of collecting and
installing; planting units is less than 20% of the ac-
tual cost of rile entire project; (2) while monitor-
ing costs at. first glance may appeal- high relative to
planting costs, it is important to note that monitor-
ill, represents a labour-intensive, multi-,year effort
to ensure that performance standards are met and
necessary mid-course corrections are undertaken.
The majority of'' planting costs on the other hand
are incurred at a single point in time. Tl1is cost
pattern is not unique to seagrass projects, but is
commonly observed in natural-resource restoration
projects across different types of' habitats. We con-
sider these data to be much more indicative of the
real costs of executing a restoration project than
previously presented (e.g. Fonseca et al., 1982).
CONCLUDING REMARKS
In this chapter we have dealt with what we consider
to be some of the critical issues that must be ad-
dressed in the implementation of effective restora-
tion projects. These issues include: (1) choice of an
appropriate metric, representative of the array of
services provided by a resource, by which to mea-
sure success, (2) evaluation of lost interim resources;
(3j,appropriate selection criteria for off=site restora-
tion projects; and (4) accurate project cost estima•
Lion. A fifth issue presented itself as we edited the
paper - the role of disturbance. Disturbance is a fun-
damental ecological process and we noticed that it
repeatedly worked its ways into our discussions, sig-
nalling its obvious but subtle role in influencing the
outcome of restoration projects. Finally we review
methods, but we do not view these to be a weak
168 MARK FONSECA ETAL,
link in the prcacess, p(rr se. The vveaakness in meth-
ods arises when workers do not study past efforts.
Rather, failure of restoration arises in general from
not considering the broader context of ecological in-
juries. particularly issue (3i, When restoration plans
are sent to us for consideration, the first aspect of
the plan that we look at is the choice of a restora-
tion site. Almost without fail for those with little
restoration experience. a site is selected that is not
damaged and does not need repair (e.g, planting in
spaces anion; naturally patchy seagrass).
h7 the United States as elsewhere around the
world, we have largely won the battle to recognise
the value of seagrasses as a national resource. How-
ever, the acceptance by US federal courts of our
metrics for assessing success, the concept of interim
resource service losses and the methods tar quanti-
fying them, and the logic for selecting planting sites
has given us an unprecedented ability to luster efiec-
five restoration of these habitats. More importantly,
perhaps, is the signal chat this has sent to the de-
velopment community and responsible parties: that
this resource is of vitro national importance and its
destruction cannot be tolerated by the public. While
transplanting seagrass is not technically complex,
in order to meet the goal of maintaining or increas-
ing seagrass area, careful attention to detail must
he paid to the entire process of planning, planting
and monitoring - a process that does not lend itself'
to oversimplification. As with all terrestrial crops,
there are inherent risks with seagrass and failures
will inevitably occur. Given that despite collective
millennia of human experience we trade stock fu-
tures on the probability of successful cultivation of
food crops, restoration of scag;rass ecosystems will
suffer from at least this kind of risk.
REFERENCES
Addv C. E f 1947j. Eel grass planting guide. Afarylarat
Cartservationist. 24. 16-17
.
Bird. K T, lewctt-Smith. J. & I'onseca, M. S. 11994). Usc of
in vitro propagated Ruppia mantima for seagrass
meadow restonition. jour-no! aJCoustal Re<eurch. 10.
?32-731.
Broonre, S W. Srne(a, t?. 1). & Woodhouse. VV W. Jr
l NMI Long-tcttn growl h and dcvelopntcnt of
uanspiants of the salt marsh glass Slrartind atternffhora.
Ys t u a ries. 9. 6 1--74.
C;ilumpong. H 11.. Phillips. R. C., Moppcts. L G.. irstacion,
j. 5 de f.con, R. cl. & Alav.a, !vi. N. 11992). Performance
of sc:cgrass tranvplaots in Negros island, central
Philippines and it, OnplWatiains In sirltigaung degraded
,IlAloty coastal atrca,. In P1,0,(redoigs of the grad KI'-I:&A
Ph),coloQ, Svmposiu'rfbVorksh la, eds. H. P Caltimpong &
1. C, %lenez, pp. 295-_i13. Laguna. I'll III plij lies.
Philippine Council for Aquatic and Marine Research
and t)rvclopment
C 11milrong. If. f.. 1c>,ts va. hi. S. & Kensorihti. VV. ). (in
(tress;. Svagrass transl,tiam,mon. In (;lohad Seagrass
ke -arch ;Viethods, c(Is. 1, 1 Shorl & R. C. Colts
Am"terdam, The ` ighcrlands: Llwvier
C.Imphcll, M. I-. Bast yan. G. R, & Walker, r). 1. 12000). A
dI, )( mon-bascd ifamework to increase sea;rass
transplantation success. In Proceedings of the 4th
International Seagrass i3iolog VVorkshop, eds. G. C.
Pergent-Martini, M. C. Buia & M. C. Gambi, pp. 332-335.
Corsica. Ctancc: aocieta ltaliana di Btotogia Marina.
tiiofa„ia Marina A9cditerranea.
Chapimin, U., ladmiza, N, ,,,- f'cnn. 'I. 119981. Calculaiing
Vesaurrc tontpcnsaricm, an .app]ie:alion of the
,vise toacrsue aLrproa<h to the Naci,bitd Stine.
lnat.ardou, uasta• sue. 1y0AA Jechnica) Paper no. 97-1.
("Wilrihutions in Marine Science. 32. 41-48.
D)ivis. R- S Short, F I, t1997a. An improved method for
transplanting; eetgrass. Lostero marina 1.. Aquatic Botany,
59, 1-1b.
den Ila rtos; C. f 1970!, t*he Seagrusses of tht, Nhrrhl.
Amsterdam: North-flolland.
Dk i l-cnbac. leer. 1.. ]r '" Lewis, R. R. 11982), seasruss habitat
Reetnruliom Lake Surprise, Honda Kely .'Limpa, FI:
Man-41mi' Systc-nls Int.
I)mako. M 1.. Hall. NI 0. Sargent, F. & r'ee'k. S. 179921.
Propeller scars in wagrass beds an assessment and
crpcrmivnr, l study of recolonisation in Weedon Island
State Preserve, Florida. In Proceedings of the 19th Annual
Conference on Wetlands Creation and Restoration, ed. F. 1.
l-Vebb, jr, pp. 42-53. Tampa. 11: ifillsborough
Community (.o leke.
honscca. Nr. S. 11989). Regional analysis of the creariort
;Ind rCStotation of sea-gras% svstems, fit I'Vellamil Creation
unit Kesforation: The Slants of tYu Science, vol, 1, Regional
Reviews, eds. J. A. Kuslcr & A1. F. KentL141 pp. 17/5-198.
C'rrl?;allis, (7R: Environmental Research Laboratory.
Fonseca, M ti. 11992). Restorim* seagrass systents in the
United States. In Restoring the Nation's Marne
&nvircnrrnenf. ed. G. W. Thayer pp 79-110 College Park,
%it): Rldrviand St ;I Grdnt C:vllcgc.
Seagrasses 16g
Fonwc a. M. S, i,199-11.A C,ufdr to Plowing' Seograsses in the
Gulf of Mexico. Galveston, Ta, Texas A&M University
Fonseca. M. S. (19961, The role or seagrassvs in nearshore
sedimentary processes: a review. In Estuariate Shores:
Hudroingical, Georrtorphological arid Ecological triteractioris,
eds. C_; Roman & K. Nordstrom. pp. 261-286. Vow 7orkc
(ohn Wilev.
Fonseca, M. S., Kenworthy. W 1. & Phillips, R. C• (1982). A
cost-evaluation technique 1ior tv."loration of s< rkrass
and olher plant communities, lsrnirorurirntai
Conser' •ation, 9. 237-241.
Fonseca. M .S., hltavcr, G. W & Kenworihy, NJ. J, (1987).
l-lic ttse,)I ecological data in the implementation acid
manai.cn cnt of scaorais restorations. Florida Mor-itie
Reseordr PuNication, 42. 175-18?.
Fonseca, K4. 5.. Kentworthy, W. (., CourtnEy, F X. tai half.
M. 0. 119941. Seagrass planting in the ,oullwastern
United States: methods for accelerating habitat
cicVelri}?lilent. Restoral on Nolrqy, 2. 198-212,
Fons'i.,, M S.. Kc wo,tit,, Cnu;tnev. 1. X 119961
1)rvt°lapr,i,nr of piartt<•d tii°a?riss heel>, its lamp) Ray,
I lorida, U.S.A. 1: PI,i t compoiwwt ,1,karln, Ferilogy
Pt-ogri cs Series, 132. 127-131).
Hmso(rj, M.?. hemwor lip. W. (. N. Thaver- (;. WS'. i• 1991541;.
Guideline s )or the Gonsenution and Restorotiori oaf Sea•grasses
!ii the IFritted Stores and Aiifacent Wafters. Jilver Spring,
MD: 11.5. Departrnenr of Commerce. National Oceanic
and Atmospheric Administration.:,Mso retrievable in .
pdf format frond: http_•!rstiriinp.bi:a.nmh.kotti-
libr rV4 digital.html^ and hup1;;www.eop.noaa.gov
Fonseca. M. S., Kcnworthy. W. a. & Paling. F (1998b).
Restoring scagiass ccos,>',!i?n!a in high di,;mt- unto
environmews. in (7ceari Comrrimity Cortferenre, 16-ir)
Noe,,nl1wr 1998 Washington, DC: Marine Technology
So( it,Ey.
J-onscca, M. S.. Julius, BF. K Kenwearihy, W. ). 120000).
lurvoramig biology and economics ill seagraslc
restoration: bow much is eilau;;h and why'? Ecological
Engineer, 15, 227-237
Fonseca, M. S-, Kenworthy, A•Ar. •J. & 1Nhitfiel<I, P. f;. i2000hI.
Temporal dynamics of seagrass landscapes: a
pi-dimmarv comparison of chronic and extrcnic
disturbance events. In Proceedings of. the 4th Iriiernatumal
5eag?ass E3ir?lr t tMorkshop, eds. G. Pcr; cm, G. C
i'crgcnt htEirtini, Ri. C. Buia & Kit f:. Gatlihi, hP 37:>-3?f;
Cki?sica• Fr:?nca.
tvoilrt'IUreAn. J. W_ Powell, G. V N., Krnworthv, W I. &
7.icnian, J. C'. 11995), The efti•cis of'lon"'rerm
manipulanon of nutrient supply on competition
between the scagrasses thoWsia t"Studiflum anti
Ifaloiitile bc°ri,glttij in Florida 13av. Oikos, 72, 349-358.
t'redette. L I-. Fcnucca. M. S., Kcnworthy, J. W. &'1"1ayer,
G. W. +.1985'(. 5'ctlgras's transplanting: 10 .yrcrcus of U.S.
Arnjy Corps of CngiAeers rc;canch. In 1°rc}re?rilinks off. tha
12th Annual timje?rera'e fret VVctliand ke<torotion anal
(:caution, rd. F. 1. 4Lcbh, lip. 121-1 34, T,11111)3. Pf :
Jtil!sbc+rcrtigh Ceitti,nunity (.ollr;c.
Glangci. S. I... "i'ralx?r M. S. & Nixon, S. W. 120001 The
influence of planting depth and density on
germination and deVelapment of 2ostc;na marina sceds_
In Proceedings r)) the 4th international Secagrass Bioloq
Workshop, eds. C. Pergent-Martini. M. C. Baia & Nl.C.
Gambi. pp. 55-58. Corsica, [:ranee:
Kenworthy, W. 1., Fonseca. M. S•. Whitt-iitld, R F.
Itamn'lersirom, K. 120001. Experimew,ii manipulation
and 'in.aly>i5 of recovery dy'nainics in pliysically
distuirbed nopit;d seagrass ceinuliurrirles of North
lnieric' c implications for rcmorztion and man<itemcnt.
In fInIl a'c Eirts of uux 4tlt lnt e r notioric' Sc a --INS Bioir3q
Works)wP, rds t. N,i- ew. C:. Porgcm. Mardnt. M. ( . H11w
K, M. C. G:ambi, PPP 3"s5 38S Olstca, Prance
King, I). M- Bohlen. ( ( . K• Adler, K. I. ±1s")93; k1wcrstwd
Marcagemeru and LVcrland M:tigatiott: A iraniewilk lsir
Ue(cminir>,g Compensation Ratios. Univorsiw of Nlarvtand
S?:?tem, Craft Report no. t-1+1t-IT.S-CT31_-93-09.S.
Washimnall, t3C: Office of Policy, Planning and
Evaluation.
Lcrsvis, i. G- ilt i1987?. Crustac ran macrotaulla of''waora'sN
and macroalgat, in Ap.ak4dwc Fay, Florida(, t'SA. Marine
Biology, 94. 219-221.).
Marba, N. K Duarte. C:. M. (119961. Coupling of seagrass
(yniodocea nodosal pats h dynamics to subaqucous dune
migration. Journal of Ecology. 83. 381-389.
Marba, N., Cebrian,,l.. Enric)uez, S. & Duarte, C. M. 1195341.
Migration of lar Caka ?r bac)acous bt!dIm-ms
measured using cc.igras, K;rmodaceo ond)auml ae tracrrx
Limmo4)Kv anal C)ceantfgraph?;, 39. 126-1:3:1.
Merkel. K. W, W)88) I clgiass iransplaiiiing in South San
Dicgo Ray, California- in Prrxredings of the Califirrrti0
f el,grass Symzposium. ock K. W, Merkel &, R. S. Floffman.
pp. 2z-42. National i:rty. CA. Ssweettwarrr River
41c-rkel, K. W, (19921. A Field yfarmal toi Tr arisplantatiiin
Techniques for the Resfmatiort of f'uc!jx Coust Pclg''riss
'lcaLlmvs. National Cit4,, CA, Pacific Sonthwesi
131okrgical Setti'tres tm.
Moore, K. H., AVctzd. R.,{. & thih, R 1. 11997). Sca?rulal
pulsrs of turbidity ?tnd their rrlati+>ns is e+Agra«
170 MARK FONSECA ET AL.
Lt rfl rricrratta; sttr'.n'al Io 'In c"uaIy. jortr!fill ,i(
1 rlr,rr(Irte`rl ul a1[tt'tn[' Nii?ln_tiY time I whit,. 215. 115-1 34.
4lurlstrirl. I . K. r19K9, !'crya?cUuc, on ic.c >,aastinn dIsua;e.
ni r? 3gras5 Losu-ni marina. U m"o rtj':lgttaric O L,anisms.
7, 21 1--771.
\ario ial O cultic and \Irncisphett+ Ad I nwnitlon
?c1Arli. Daniurc .;>?cti,rn, nt and Rcsuoruion Program
i1997a, Halalt,at 7: pivulcr,) ,An+aii'SLs'- An Over-rest. p(fllc ?
,nd lcchnna' paper no 95-1 Stlvc•t-Spring, \10ISO AA
Danla,, ASscisment (enter.
NaiiorIat Olvanic and auncrsphcrtr ACImInlstr.ttion
M )AA), Danrtge Anse?ssnieut and Retitot.?rtic>n Vrnl,r.un
(19'd7Ir1. Sruatfn,g Compertsatarp Resrmr i= rlit wf& Glrl Nmy
!k c ?me:d: jor %ohuztl Rese,urre Damage .ssessracw IR¢der the
0:1 P0111,110v, tt If 799x0. Sav(r `I-Irlflg %1D P?UAA
1').inrage Aavr>?nlcnf C.rrurr
(seta. 1% ?. I1I week, 141 L_ tS;tii;•y. I' Pal. LSarlholomt;c,
\ ;,mead. [ . T.. l outt?,:aaa, . A , Mown., K A R_ltodc11 Wcuods, H l ± (!(}0! A retioictiv cal 1,1.i;a° in
eirrass Seed dorma]lcti .111d kcrnunat11111 uti.il lMIu?IS
h con. avation and wtav;atww %famrr 1`.wio> l'rvurss
Srrrc± 200, 2,'7-2,"
Phi lips, R. C: & liner' 1_ C, ; I9-1 `'Si Scogt'arw.i,'s.
Stttitivoniaan Contribution to the 1larmc `?Cie,trc,
un. 34. Washington. 11C: Srilithsonlan Iusutudon I'rrss.
Provil, ;A.. Lvk'-L?mg, W-1. & C;c:alca. R. G. i 1995I, flood wind
cvclranc related lass. 'ind partial recovery of Move than
1000 kui` of sca?,ra.ws in 11civcv Bay, (,)ueenslanci
Australia Aquatic l,tol:atty. 52, 3-17
Robilhard. G. A. & Porter, V. f. {19761. itarrsphdracinorr uj
c r,Qrnsr{'lontcrs nt:arin,ij Ili Seta Mcgo Liar Sall Diego, CA_
i;nd«_'rsea StiIcu(es DI-palinle , ^Jav,tl !Jndorsca ("W"r.
Robblvc, h4. 13. Barhet, 1 R, Carlson, R V Durako. NI. I.,
footiltnean. J lh'. Atuelrlsutin. 1. I<. porter. 13.. 1'artsra, 1..
A_, 1 Icman. R. 1. & lieulan. J- C. (19991. M<us nun'lalitV
,>i ttit: trope al seag tss Tlurlns.la reshuPiruarn its 1lorida
R;s !11SAi, Mcrine ficuioiCi, J'nQf,xy Sc'ri;>s, 71. L9n2w-
Kosc ?. 0- Sharp, W, C.. Kcmaorthy, AV I_. Hunt. j If-
Icon,' W, (_, Piaklcr, L_ I_. Vatcnline. 1.1c., 1fall, Nil, 0
11hIllit'ld, V & Fourtlutc'o:' 1. AV 110991. Seat urchin
over raurt of ;.t lafg( sr,igr'w' hrd in frtttur hlorlda Bav
Marble Vcol,izv Pniuivss Scnc;, 190, 211-2'22.
Shu.rrdan, P.11999L'PrajcCtcirv (n ynulural mrlivakmcc
01'taste!red and oatcrra9 Haird,ra crrip hL"r Iird. in lrxas
GNlf Rewa'rch Rc'ptrrrs. 10. "(1..,82.
Stan 1- L & short. Q A 1200 v Wcm;Vmg vow"
sZrowth Ibrni, lair leaf and rhiunn>' in.crkilig
application-. In Pracecdb 4s uJ rite 4th }•ttr'mawnul
Suagnlt s ploic>?y 41'c ksla.rp. cads. G, Pa rgent.
( I'crgent M.11IIIn. Nl. (,. iiuia & \1. (, ( amhi.
I,)1?, 1:31-I 4- C:orica, Irall(vr
Shr±t'i, F 'l h kV,Ilic-Ii(1u'%WfI'1a, `; {ITY11 Aatttral <snd
humarumdund dbuil tnrr of ?c'agrassc?s. Enr;rtrrrrrerr?ui
Couservaliw:, 2:3. 1?-27.
Ih:VNr, (;. G1'. Kemacitthy, A'Vr ?. Fonwt? i. M.',. I1984),
I'liv i-ialo? • of eel ra?s rncadoiws of th(r A[iant1 Coa'I
a Conttnunit'v prof Ie_ ti.5 [ ish and \4'ildlitt s l-vice.
1b'ashington, Dt` US Deparlinenr of tilt
Interior'
Thiirhaiit;, 1 ,197,1), Transplantation of the tit grass
i'li"cs.cia ustudmunt KGt;iy;..%#qu aadbum 4. 177083,
icihr,t,ko. D. A , i1awc " t.. 1. & 11;11€. 1%1' (1 1119')U. I flects of
the Inwilvi If short ,how' and pic"eme of till'
rhia:nll7c api(al cocnsleno oil the survival and growlh ofranSphtnud sea, rays Thtllas;ia reviiii1autrt. Cownriuf!01s.
Pt MOI'WIT S iC!l1c, 32, 41-48
.
n;,cnrth. R. ` Y 81t hop' R C H9941
A>5arasui }i.ltural
!'l'tiUUk?"t. Lflnagu C. iVrronnic til'i1 annttlt.1P".
iut;tCal t,cattotrfics, 11. 3a-11.
\l! lon ITI, R. W & %Inrri?. I. 11'+',+61 Sc at,>vvs l r 'St't'iYmwi
ttntl Rrst.wtznwi A 0-i(ignoslir film ;w the luclian Rive?.
lUi;iloli' Ic,. note,+l Mrnu.+randurn sit.>. 1-1. ['a,alkac, I?L:
St Johns River \1'airr KMJn,t}tt.vn'Jn Dislritl.
\lrhirtielcl. P. 1- Kenworlby. 11. 1., Ionw(a, M, S. X
Hantmerstroltl, X. iin prv,,q Role of iorins if,. (tie
expansion and ptolwMmion of diilurhanres initiated by
ntutor vessels in subtropical seaert?' beds. jow-Tiadl (tf
Coastal Research.
V%ilhani,. S. L & t)rih, R_ j_ 199r41, Genetic dM!isiiv amt
stiuctun, of natural wind transplasucd tvlgra.5,
rxipulations in the C:heaapcakc told Chincoteague Bay's.
sniarte's. 21. 118-128.
'Wooal. R. I. I'., odunt, W I w Zicttu r'. i C. I N691.
Itlllu"Tue ofw'agr.asscS on tilt.' prodUCtIVil c,; Coastal
l.tgcniru. iri N10rturilas de .St?niposium trttt'rnacierutl Laguna.,
Cuswrns, ('d;- A- C,?Slamarcy. A. ,A,yala & 1. B. Pnk'ge2r,
pp 495-502. i'rh'x3t'o. Of : I !,'0 s14'hl M 0,
WN,ilic KLhevc-rriat, ti arrcl, 11, & Cox, V. A {20001 lwagrat s
con;en°.tricnt: Ica,c,It,S !'tom <>rhnolaotanv. Yua'ific
Ct?rLSir'ihd tlti}r, kttala+it7;;?. 3, 3329-1,35
Ledler I I & Lanpy R. 1Nok 0mifwisms of
C,utsiructcd end natturai salt mar,hcN oI W, Die°£o Bav.
We 10100'iu uncl hicat aicntc?tN Nnt's. 9 2i-25
Boom t, e 119821, he euA gy (I swt,rasmy IT sa'rurh Fordo:
a cu rimwlily 1 nylle Ill) Fishcric5 and WlldhiC Service,
b1aShin}tton. 1A , I'I Ilcpatrtnicnt o[,ihc;nltil':ta.