when is restoration not? incorporating landscape-scale ... · at the landscape scale or in some...

Ecological Engineering 26 (2006) 27–39

When is restoration not?Incorporating landscape-scale processes

to restore self-sustaining ecosystemsin coastal wetland restoration

Charles Simenstada,∗, Denise Reedb, Mark Fordc

a School of Aquatic and Fishery Sciences, University of Washington, Seattle, WA 98195-5020, USAb Department of Geology and Geophysics, University of New Orleans, New Orleans, LA 70148, USA

c Coalition to Restore Coastal Louisiana, 746 Main Street #B101, Baton Rouge, LA 70802, USA

Received 24 January 2005; received in revised form 24 August 2005; accepted 26 September 2005

Abstract

With increasing restoration initiatives for coastal wetlands, the question of ‘What are we restoring to?’ becomes more pressing.The goal of this paper is to explore restoration concepts, examples, and challenges from the Pacific and Gulf coasts. One of thefundamental concepts explored is change over time – either in the controlling processes or the restoration structure – and howsuch changes can be meshed with the goals of various restoration efforts. We subsequently review the concepts of ecosystemt he contexto ems beingr to plan fort ange overt©

K

1

ah

caleiscotificm-thea-of

0

rajectories, alternative restoration approaches, and the ideal attributes of functional self-sustaining restoration in tf realities of restoration planning, design, and implementation. These realities include the dynamics of the ecosystestored, very real constraints that are imposed by the contemporary physical and human landscape, and the needhe long term development of restoration sites recognizing that both project performance and expectations may chime.

2005 Elsevier B.V. All rights reserved.

eywords: Restoration; Landscape-scale process; Coastal wetland; Self-sustaining ecosystem

. Introduction

Coastal wetland restoration is rapidly approachingscale of planning, design, and implementation that

as surpassed its origins in individual mitigation and

∗ Corresponding author.E-mail address: [email protected] (C. Simenstad).

restoration actions. Large, complex, landscape-sprograms occurring in the Everglades, San FrancBay, and coastal Louisiana require additional scienunderstanding of entirely different spatial and teporal perspectives. While it is argued that evenperformance of individual regulatory wetland mitigtion actions suffer from a lack of considerationwatershed setting and landscape function (NRC, 2001),

925-8574/$ – see front matter © 2005 Elsevier B.V. All rights reserved.doi:10.1016/j.ecoleng.2005.09.007

28 C. Simenstad et al. / Ecological Engineering 26 (2006) 27–39

restoration designed to address increasing degrada-tion of entire coastal zones will likely not approachor meet their goals without promoting self-sustaininglandscapes and incorporating large-scale disturbancedynamics.

At the beginning of the 21st century, consider-able environmental science is focused on fixing theproblems wrought by generations of environmentalexploitation combined with either ignorance of or dis-regard for the consequences. Current Federal initiativescall for the restoration of tens of thousands of kilo-meters of stream corridor and hundreds of thousandsof hectares of wetlands (EPA, 2000). The cumulativeeffects of mining waste, dams, land use change, andurbanization have left few watersheds intact and riverrestoration is in such demand that curricula and pro-fessional training courses are now commonplace. Inthe lower reaches of rivers, in estuaries, and at thecoast, fundamental changes in riverine inputs com-bined with local landscape alterations mean that fewtraces of historical ecosystem function remain. Thisis most commonly the case in terms of natural distur-bance effects—management of river flows and ‘pro-tective’ measures for local communities have removedregular hydrologic pulsing as a normative agent ofecosystem change and only the largest and mostcatastrophic events, for the most part uncontrollable,remain.

With increasing restoration initiatives, the questionof ‘What are we restoring to?’ becomes more pressing.F tifico crit-i ri d toh rredp hichi ra-t ationb vedo raisei ghta , buta dis-t tionm of ad nter-a mayb But

floods, fires, droughts, and storms disrupt everydaylife of local communities and frequently lead to callsto return again to more ‘management’ or ‘prevention’measures.

As restoration plans proceed for almost all largerivers and estuaries around the U.S., the disconnectbetween societal goals and ecosystem functions is per-haps most obviously shown in the lack of a clearunderstanding of the term “restoration”.NRC (1992)defined the term to mean returning an ecosystem to“a close approximation of its condition prior to distur-bance” which requires “reestablishment of predistur-bance aquatic functions and related physical, chemicaland biological characteristics.” Thus, the NRC recog-nized both ecosystem structure and function must beaddressed in conjunction with the natural ecosystemdynamics of processes: “the term restoration should beapplied only to those activities directed to rebuildingan entire ecosystem.” Among restoration practitioners,the importance of the ‘indigenous, historic ecosystem’as a template for restoration was noted by the Societyfor Ecological Restoration (Aronson et al., 1993) andclearly distinguished from rehabilitation.Middleton(1999)summarizes an evolving view of restoration asseeking to establish ‘a site that is self-regulating andintegrated within its landscape, rather than to reestab-lish an aboriginal condition that can be impossible todefine and/or restore within the context of current landuse or global climate change’.

The goal of this paper is to explore restoration con-c andG oredi sseso ngesc tione sys-t , andt ora-t ng,d onc aree sys-t

nala dinga froms listicg ing

or society exposed to the rationale, whether scienr economic, that more natural ecosystems provide

cal goods and services (Daily et al., 1997), the answes usually easy. People desire that which they useave that is now gone, or, if the degradation occurior to the current generation’s experience, that w

s more ‘natural’. Society also wishes that ‘restoion’ be made permanent—that ecosystem degrade reversed and conditions in the future be improver their current state. For science, such goals

mportant questions about not only how we michieve desirable recovery of natural ecosystemslso whether they can be sustainable. In terms of

urbance regimes, the societal context for restoraakes things even more challenging. The valueynamic range of ecosystem processes, including innual variations and occasional extreme events,e well established from a scientific perspective.

epts, examples, and challenges from the Pacificulf coasts. One of the fundamental concepts expl

s change over time – either in the controlling procer the restoration structure – and how such chaan be meshed with the goals of various restorafforts. We subsequently review the concepts of eco

em trajectories, alternative restoration approacheshe ideal attributes of functional self-sustaining restion in the context of realities of restoration planniesign, and implementation. Although our focus isoastal wetlands, many of the principles exploredqually applicable to other seriously threatened eco

ems.We herein argue that, to be ecologically functio

nd self-sustaining, restoration requires understannd reinstating fundamental ecosystem processesite to landscape scales. Our emphasis is on: (1) reaoal setting; (2) ecological functions; (3) the emerg

C. Simenstad et al. / Ecological Engineering 26 (2006) 27–39 29

paradigms of a relatively young science; (4) the need torecognize high uncertainty in restoration technologiesand responses; (5) the rationale behind taking a precau-tionary, adaptive approach.

2. Trajectories of change

A fundamental premise of restoration ecology is thatrelease or diminution of stressors will reinitialize phys-ical, geochemical, ecological, and other ecosystem pro-cesses in a direction toward a more natural, unstressedstate. This progression of ecosystem recovery over timehas been characterized as a pathway or trajectory ofecosystem redevelopment toward a less compromisedstate, or even the attainment of a fully functioning sys-tem comparable to “target” reference sites. In graphicalform, these trajectories have been referred to as “per-formance curves” (Kentula et al., 1992), “restorationtrajectories” (Hobbs and Norton, 1996) or “functionalequivalency trajectories” (Simenstad and Thom, 1996).The validity of such trajectories, or at least the realityand predictability of the time required to reach equiv-alency, has often been called into question (Zedler andCallaway, 1999). Despite notable cases where wet-land mitigation or creation sites have not followedparticularly rapid or definable trajectories, there arewell-documented cases where restoration of naturalprocesses has resulted in obvious trajectories and oftenover relatively rapid time periods (Morgan and Short,2 rrene rnsa ofa s ofs in thel

ea overy( hepr temp ouss bil-ig d bymc ewe osed

to the previous definition for restoration, rehabilita-tion (as well asenhancement) is confined to the iso-lated manipulation of individual ecosystem elementsto a less degraded state, rather than full restoration ofthe structural and functional attributes and processesof the predisturbance state (NRC, 1992; Middleton,1999). It might be argued that many rehabilitationactivities are knowingly conducted under the guise ofrestoration, but perhaps many more activities resultunavoidably and unexpectedly in reallocation evenwhen the explicit objective is restoration. At the riskof corroborating more jargon in the restoration liter-ature, theAronson and Le Floc’h (1996a,b)conceptof reallocation is a useful concept because both inten-tional creation and unintentional shifts in ecosystemstate are typically the consequence of deviating fromrestoring the full complement of natural ecosystemprocesses.

Some constraints to achieving restoration in thestrict sense of the term also result from pervasivechanges to the ecosystem processes that promotedthe pre-existing condition that even vastly improvedresource management cannot resolve. Often, theresulting steady state ecosystem is in the short termcomparable to or undetectable from the pre-existingecosystem. However, the lack of fully vetted ecosystemprocesses may be ultimately expressed in wideningdifferences between the restoring and pre-existingstates, or even the eventual shift to rehabilitation orreallocation.

thes m-i cturea evenb ion.T ver-itmpstac s one cog-n onu gicalc

002; Tanner et al., 2002; Thom et al., 2002; Wat al., 2002). Such wide variation in response pattend rates is hardly surprising given the variabilitypproaches to “restoration,” the types and leveltressors, antecedent conditions, and changesandscape setting.

Aronson and Le Floc’h (1996a,b) describe threlternative ecosystem phases of ecosystem recFig. 1) that differ in their ability to actually reverse trocesses that led to degradation: (1)restoration, whichequires reactivating hydrological and other ecosysrocesses and allowing reintroduction of indigenpecies to the point that “thresholds of irreversity” are circumvented; (2)rehabilitation, where oneroup of species or ecosystem service is favoreodified management over the short term; (3)reallo-

ation, where entirely new trajectories promote ncosystems and uses over the long term. As opp

While the desirable functions may result fromtructure of ecosystems, it is typically the dynacs of ecosystem processes that sustain that strut the landscape scale or in some cases maye the underlying mechanism behind the functhus, sustainable restoration is contingent on reco

ng both ecosystem structure and processes: “Restora-ion means returning an ecosystem to a close approxi-ation of its condition prior to disturbance. Accom-lishing restoration means ensuring that ecosystemtructure and function are recreated or repaired, andhat natural dynamic ecosystem processes are oper-ting effectively again.” (NRC, 1992). However, therux of achieving sustainable restoration still hingexpectations of a predictable endpoint, and the reition that ‘desirable’ trajectories are often basedncertain and unpredictable responses of ecoloommunities.


Fig. 1. Alternative ecosystem trajectories over three phases, illustrating the notions of restoration, reallocation and rejuvenation as well as thatof “thresholds of irreversibility” (Aronson and Le Floc’h, 1996a).

3. Approaches to ecosystem recovery

There are three basic approaches to ecosystemrecovery that purposefully address ecosystem struc-ture but which encompass the reintegration of dynamicprocesses to varying degrees:passive, active, andcre-ation. The alternative pathways of ecosystem recoveryoften vary as a function of these approaches, as well asthe restoration, rehabilitation or reallocation end-points(Fig. 2; Kauffman et al., 1995; Middleton, 1999).

In passive approaches, the accidental or incidentalremoval of barriers to degraded ecosystem processeslead to their reinstatement either in whole or to a largepart. No further actions are, or need to be, taken to

Fig. 2. Interactions among wetland structure and function withecosystem processes.


facilitate restorative ecosystem change. In most casesof passive restoration, the re-establishment of naturalhydrological cycles increases disturbance to the site,promoting natural dynamic, rather than static, ecosys-tem processes. For example, many studies evaluatingecosystem responses to removal or gapping of lev-ees surrounding a former wetland often describe howflooding events or natural pulses influence both thenear-term development of the site (e.g., sedimentationrate) and its response to long-term factors such as sealevel rise (Simenstad and Warren, 2002; Orr et al.,2003). Cessation of practices that lead to the degrada-tion of wetlands, such as cattle grazing, will also assistin passive restoration of an ecosystem by removing adetrimental disturbance (Esselink et al., 2000; Bos etal., 2002).

Active approaches to restoration are accomplishedthrough more “engineered” actions that intentionallyand specifically re-create wetland structure and pro-cesses. This occurs in areas where these processesonce existed or where they still exist, but in a muchdegraded form. This may involve removal of processbarriers (passive restoration if conducted in isolation)and active management or enhancement of processesbeyond those which passively reoccur. For example,reestablishing tidal hydrology to a drained and lev-eled estuarine wetland might be combined with diggingpilot channels to encourage tidal channel developmentand vegetative plantings to promote growth of nativemarsh vegetation and prevent colonization by inva-s ks toa ent’a ntin-u ls ina

eren andd sys-t nds,s butm reates2

cals ry, but thev andi hes.

Ultimately, the long-term performance of a restoredwetland as a functioning ecosystem, regardless of theapproach, will depend on reintroducing some form ofnatural dynamics and disturbances into the wetland sys-tem (Middleton, 1999; Orr et al., 2003). Such restoredwetlands, if they persist and become self-sustaining, aremost likely to resemble natural wetlands in the region ifhydrological and topographical variability, subsurfaceprocesses, and the hydrogeomorphic and ecologicallandscape and climate are considered (NRC, 2001).

Socioeconomic, public safety and other legitimateconstraints often do not allow ecosystem restoration.Anthropogenic disruption of naturally dynamic ecolog-ical processes has often allowed, as designed, humaninfrastructure to occupy coastal wetlands; completerestoration would in turn threaten this more contempo-rary human “footprint” in marshes. In some cases, “par-tial restoration” (Fig. 3), essentially rehabilitation, hasallowed muted recovery of some fundamental ecosys-tem processes. In many coastal areas, marsh rehabilita-tion has been promoted by manipulated reintroductionof tidal influences through various water control struc-tures (e.g., slot gates, self-regulating tide gates, etc.).The rationale in acknowledging these constraints is thatsome recovery of natural ecosystem processes, albeitsignificantly truncated from natural dynamics, rehabil-itates natural functions to some degree. This is certainlytrue in many cases but there are trade-offs and negativeeffects are common. At the minimum, some functionsare enhanced while others may either be unaffectedo resi veg-ea yp na ,1 h“ otr ftenl wf sys-t

ioni ela-t thea aled eara n

ive species. In most cases such restoration seechieve specific goals and an ‘adaptive managempproach is sometimes employed to guide the coal modification of the system to achieve the goapurposefully modified ecosystem.Creation is the establishment of wetlands wh

one previously existed. In this case the conceptesign may be based on wetlands elsewhere in the

em. The process regime dominating these wetlauch as hydrology, will depend on local conditionsost such restoration efforts focus on attempts to c

tructure rather than natural process or function (NRC,001).

However, our point is not to focus on techniemantics about approaches to ecosystem recoveo convey to restoration managers and the publicery real differences in expectations, sustainabilitynvestment involved with each of these approac

t

r even negatively affected. Water control structun coastal Louisiana saline marshes do sustaintation but preclude much nekton exchange (Rozasnd Minello, 1999) in what is typically a dynamicallulsed system (Rozas, 1995), and limit sedimentationd other geochemical processes (Reed et al., 1997999; Kuhn et al., 1999). What is important in sucpartial restoration” is to acknowledge that it is nestoration at all, but exceedingly targeted (and oegitimately so) rehabilitation that will seldom alloull ecosystem recovery and may even impact ecoems outside the project area (Hood, 2004).

A fourth constraint on tidal wetland restorats scale. Although as yet poorly quantified, the rionship between tidal ecosystem restoration andttributed functions of a natural marsh is highly scependent. Many relationships are, in fact, non-linnd dictated by thresholds. A 0.1 km2 marsh restoratio


Fig. 3. Pathways of ecosystem recovery and the reintegration of disturbance processes (Kauffman et al., 1995; Middleton, 1999).

should not be expected to necessarily function as a0.1 km2 segment of a 10 km2 natural marsh. Thisis illustrated most definitively in the tidal geome-try relationships between marsh size and tidal chan-nel metrics (Williams and Orr, 2002; Williams et al.,2002). Restoration expectation should acknowledgethese constraints and be equally scale dependent.

4. Prerequisites of functional, self-sustainingsystems—restoration realities

4.1. Restore processes, not structure

The ability of specific actions and approaches toachieve functional, self-sustaining restoration is con-tingent on the goals the project. Not every restorationproject is necessarily dependent upon a “walk-away”

assurance that the restoration will ultimately result in anaturally functioning system with little to no humanintervention or management. Conversely, stakehold-ers and decision-makers do not always understand thatsome of the more engineered or uncertain restorationprojects will require intensive, and potentially long-term, investment in public resources to maintain theirexpected level of performance. In Louisiana, where lossof coastal wetlands has exceeded 52 km2 per year fordecades (Barras et al., 2003) manipulation of marshhydrology is still used as an arguable approach torestoration (LCWCRTF, 2003). Structural marsh man-agement in coastal Louisiana is usually designed toimpact both channel flow and marsh water levels.Hydrology is altered in order to achieve the statedgoals of the management, which normally includerestoration, conservation or enhancement of emergentmarsh or specific vegetation types. Historically this


was a relatively successful manipulation for the spe-cific purpose of enhancing waterfowl habitat (Baldwin,1967) but more recent plans seek to achieve restora-tion goals through the control of salinity and/or waterlevels using systems of control structures and levees(Cowan et al., 1988). These plans can be used to controlmarsh hydrology passively or actively. Several studies(Boumans and Day, 1994; Cahoon, 1994; Reed et al.,1997) indicate that this restoration approach severelyimpacts the natural flow of material, especially sedi-ments, into the marshes during both tidal and extremeevents. Even though this impairment of natural processregimes has been noted as a fundamental feature of suchmarsh management (EPA SAB, 1998) there are stillproposals to employ such techniques in new restora-tion initiatives in San Francisco Bay (Harrison et al.,2001).

4.2. Recover natural ecosystem dynamics

Natural ecosystem dynamics are the source ofmany functions and services attributed to coastal wet-lands and other hydrologically structured ecosystems(Middleton, 1999). However, recovering natural sys-tem dynamics must include the fluctuations and dis-turbances that in most cases account for the long-termstructure and function of coastal wetlands. Some ofthese disturbances cannot be controlled or manipu-lated in initiating restoration, but must be considerednonetheless when designing restoration projects. Fore ct ona ls andp eta

d-i ssest stemp u-l nte l comp h-o es tow ich im-m1 ta

the Columbia River estuary (Simenstad et al., 1992),to name just a few systems currently under intensiverestoration activity or consideration.

Tropical storms and hurricanes can affect coastalwetlands in a variety of ways. Low intensity hurricanesand tropical storms can deliver excess rains, which willboth flood a system, and also help flush saltwater fromthe system. This is especially true for restoration effortsusing dedicated dredged materials, which often carryhigh salinities. Higher intensity hurricanes often haveassociated large tidal surges, which contain high sedi-ment loads but can also flood coastal wetlands with saltwater for periods of time beyond normal tidal cycles.Such tides can assist in the die-back of less flood- andsalt-tolerant plant species and promote new plant ger-mination. Many plants, for example, are also adapted towater borne seed dispersal, water borne propagule dis-persal, or vegetative reproduction. Although existingplant assemblages and wetland geomorphology may betolerant to disturbance conditions, sediment restructur-ing may readjust the local mosaic of plant assemblages.For instance, floods routinely cause large-scale intro-duction of large logs into coastal estuaries of the PacificNorthwest. These can disturb estuarine marsh plains asthe wood settles and refloats at higher tides (Simenstadet al., 2003).

Neither hurricanes nor floods can be controlled, butboth positive and negative effects of such large-scaledisturbances need to be incorporated into restorationplanning. Restoration designs can be adaptable to, ore istur-b tageso ento ific,t tly,r ten-s andi ent.F ra-t theo ofs phicea or-t e ofa

nce,c stem

xample, drought can have a serious negative effen ecosystem (e.g.,Visser et al., 2002), while tropicatorms and hurricanes can result in both negativeositive impacts (e.g.,Dingler et al., 1995; Cahoonl., 1995).

Alteration of the natural dynamics of fluvial floong is one of the more important landscape procehat have led to degradation of many coastal ecosyrocesses (Middleton, 1999). Seasonal, and partic

arly ‘pulse’, flooding is one of the more importacosystem processes that has shaped the naturaosition, variability and diversity of wetlands througut coastal zones of the world. The consequencetland integrity of altering the historically dynamydrologic regime are amply evident for the Kissee River in the greater Everglades ecosystem (Kobel,995; Toth et al., 1998), coastal Louisiana (Boesch el., 1994; Roberts, 1997; DeLaune et al., 2003) and

-

ven embrace, the ultimate consequences of dance. Conversely, the cost and other disadvanf investing in a restoration project that is contingn resisting disturbance in order to maintain a spec

argeted function may ultimately be futile. Importananges of impacts in terms of spatial scale and inity of specific disturbances must be consideredncorporated into restoration planning and assessmrequency, intensity (high, medium, low), and du

ion of such events can be important controls onutcome of restoration efforts. A high level floodhort duration may not have the same geomorffect as a low level flood of long duration (Wolmannd Miller, 1960), but both scenarios may be imp

ant and beneficial aspects of the disturbance regimn ecosystem.

The nature of the relationships among disturbaomplexity, resistance and resilience of an ecosy


is often hard to determine, but must be attempted ifadequate disturbance regimes are to be reintroduced byrestoration of coastal ecosystems. Life history require-ments of faunal and floral species should also be takeninto account in restoration planning, to ensure targetspecies can persist in restored systems. Land or hydro-logic restoration without the associated plant estab-lishment and animal usage should not be consideredrestoration. Just one example of the role of disturbanceregimes can be seen in flood pulsing which is welldocumented as providing several ecological benefitsto wetlands such as flushing, sediment and nutrientinput, and the limitation of non-flood tolerant uplandspecies (Middleton, 2002). From a hydrological stand-point, reestablishing natural hydrologic regimes maybe difficult, and in some cases impossible. Crevassecuts (breaches) and cut tidal channels in restorationprojects can help replace some of this action and arepotentially effective mechanisms to allow influence ofextreme flood and drainage events.

Fire, whether from lightening strikes or prescribedburns, and often viewed as a negative disturbance inany ecosystem, can reset the plant community allowingless competitive, “pioneering” species (Grime, 1977) achance to establish, as well as sending a pulse of nutri-ents and minerals to the soils. Mangroves are thought tobe, at least in some cases, fire dependent (Middleton,1999). Similarly, allowing native herbivores to grazereintroduces natural disturbance into the restorationprocess, which can improve diversity and acceleraten te-r ts”fn eda anyoG antc hani

4

evem ringt rac-t tting( s as mely

scale-dependent. For instance, the role of coastal Gulfof Mexico wetlands to protect more inland ecosystemsfrom major physical disturbances, such as hurricanesand related storm surges, is associated with the hun-dreds of square kilometers of the coastal wetland fringe(Suhayda, 1997). Conversely, strategically located andoften narrow wetlands such as fringing mangroves mayaccount for significant filtration of nutrients fluxingfrom landward watersheds, thus protecting seawardseagrass and reef ecosystems. The essential landscapeecology concept is that important processes and theirinteracting functions can be very explicit spatially.Restoration must consider two issues in this landscapecontext: (1) the role of landscape processes on the func-tion of the restoration project, and (2) the potentialoutcome and sustainability of a restoring wetland ina landscape with extensively modified processes (seebelow).

The performance of restoration with a goal of recov-ering habitat of motile species can depend very muchon the landscape context. Anadromous fishes, suchas juvenile Pacific salmon, can benefit considerablyfrom strategically positioned restoration sites that offerunique functions (e.g., refuge from predation, highlynutritional food resources) that are disproportional tosimilarly designed restoration sites in other locationsalong the estuarine gradient (Gray et al., 2002). Evenif restoring marshes are still early in their developmenttoward equivalency with natural reference systems, andmay actually provide less than optimum conditions forf forfi rineg icala rd int

an-n ce,i oryr ncea eend gan-i m’sp ola-t ;H set hocs y tob licit

utrient cycling through the deposition of fecal maial. In large-scale herbivory events, called “eat-ouor wildlife such as muskrat (Ondatra zibethica) andutria (Myocastor coypus), the soil surface is exposnd the root zone is often disturbed allowing mpportunistic plant species a chance to grow (Ford andrace, 1998). In both of these cases, growth of the plommunity will take a successional path different tf left undisturbed.

.3. Incorporate landscape context

Coastal wetland restoration is unlikely to achiaximum desired performance without conside

he approach, location, size and other “design” chaeristics in the context of the greater landscape seNRC, 2001). The function of coastal wetlands hatrong landscape context, although it may be extre

eeding or refuge from predation, the opportunitysh to occupy a particular position along the estuaradient may be more important for their physiologdaptation than other saline habitats further seawa

heir migration.Ignoring landscape setting in restoration pl

ing can affect the overall cumulative performanncluding the spatial distribution, time and trajectequired to achieve the desired level of performand dynamic equilibrium. Spatial lags have bescribed between simulated restoration of an or

sm’s habitat and the recovery of the organisopulations, and this has been attributed to a “perc

ion” effect (O’Neill et al., 1992; Tillman et al., 1997uxel and Hastings, 1999). The cumulative respon

o restoration that is based on opportunistic, adelection of restoration sites and designs is likele additive at best; only strategic, spatially exp


restoration planning incorporating landscape scaleprocesses is likely to create a cumulative response thatis synergistic and complementary.

4.4. Recognize and adapt to system-scaleconstraints

Many coastal systems are now so highly alteredthat restoration efforts are subject to fundamental con-straints. In some cases these constraints reflect cur-rent landscape or ecosystem characteristics, such aslimited sand resources for barrier island rebuildingin Louisiana (Van Heerden and DeRouen, 1997) andinvasions by non-native species in San Francisco Bay(Cohen, 1998; Nichols et al., 1986). More commonly,these constraints represent societal preferences for allo-cation of physical or financial resources. Altered fresh-water flow regimes may be one of the most pervasiveconstraints on coastal ecosystem restoration (Dynesiusand Nilsson, 1994), both in terms of restoring historicstructure and in meeting assumptions and expectationsof process-limiting factors such as sediment accretionrates. The altered flow regimes of almost all U.S. riversconstrain restoration of riparian and estuarine habitatsthat rely on annual flood cycles but in many cases thedams causing the alteration are considered permanentlandscape features—their water supply, hydropower,navigation and recreational purposes considered bysociety of more import than natural variations in riverflow. However, even dams should not be consideredi ovaltK hewa ater-s ningf

theS d asi for-m usedfa 99a ered theS s ared t-l nd

restoration sites, this amounts to 2.0 g/cm2/yr which issimilar to field estimates of sediment accumulation byReed (2002)of 3.6 g/cm2/yr. While the current rate ofsediment input appears adequate to maintain the ele-vation of the remaining marshes in the Delta, furtherrestoration may well be sediment limited. The Deltahas been converted into agricultural land and the result-ing drainage has caused subsidence rates of 3–5 cm/yr(Deverel and Rojstaczer, 1996). Thus, as years pass theamount of sediment needed to restore these lands totheir former intertidal elevation becomes greater andsediment availability emerges as a real constraint onsuccess.

Antecedent conditions and the extent of restorationopportunities may be most limited in urban and indus-trialized estuarine and coastal settings. This is aptlyrepresented by the attempts to restore critical segmentsof the Duwamish River estuary in Puget Sound, Wash-ington State, where at best only rehabilitation is feasible(Simenstad et al., 2005). The Duwamish River estuaryis a system that has been heavily assailed by toxic con-tamination, intense anthropogenic disturbance, exten-sive modification of the watershed and hydrogeomor-phic driving forces, resulting in an urban/industriallandscape that leaves only scattered, small patchesoffering opportunities for rehabilitation. Less than 3%of the historic estuarine wetlands remain, more than65% of the historic watershed area and 70–75% of thefreshwater inflow have been diverted from the estuary,and a legacy of contamination by metals (chromium,c oly-c car-b ists.T lan-n egicp ctedl g ($1 ms izedb , thec uar-i s (int nif-i d ac oalsa inge strials

mmutable constraints, as exemplified by dam remhat has occurred as part of restoration (Hill et al., 1994;anehl, 1997) or is increasingly planned around torld (Wunderlich et al., 1994; Shuman, 1995). Yet, thessociated constrained floodplains lower in the wheds are often not included in restoration planollowing dam removal.

The management of water resources inacramento-San Joaquin river system is viewe

mposing severe constraints of the restoration ofer tidal marshes which have been drained and

or agricultural land for almost a century.Wrightnd Shoellhammer (2004)estimate that between 19nd 2002, 4.4 million metric tonnes of sediment weposited in the Delta, largely derived fromacramento River. Assuming that these sedimenteposited in the∼75 km2 of tidally connected we

and remaining in the Delta, many of them wetla

admium, copper, lead, zinc), pentachlorphenol, phlorinated biphenyls and other halogenated hydroons, and polycyclic aromatic hydrocarbons pershis landscape limits any restorative efforts to ping around the few opportunities rather than stratlanning for optimum allocation across a less impa

andscape. The cost of rehabilitation can be numbin.0–8.0 million ha−1 in the Duwamish) and long-terustainability of rehabilitation sites may be jeopardy the lack of contaminant source control. Howeverommitment to compensate for loss of historic estne functionality and to recover endangered speciehe case of the Duwamish, two evolutionarily sigcant units (ESU) of Pacific Salmon) has prompteoordinated, rehabilitative effort that has realistic gnd is rebuilding a public investment by recoverlements of natural landscapes in an urban/induetting.


Even if some systems are not irreversibly altered,many natural processes may require considerablylonger to implement or facilitate restoration than mightbe expected under more natural conditions. Subsidenceof leveed wetlands can result in considerable elevation“debts” that will require decades to recover under nat-ural sediment accretion regimes, and perhaps centuriesif available sediment sources have been significantlyreduced (Deverel and Rojstaczer, 1996). The prospectof long-term tidal lakes does not necessarily fit moststakeholders’ definition of coastal wetland restoration!Other examples of system constraints include: reducedrecruitment of native flora and fauna due to restrictedsources proximal to a restoration site; conversely,extensive recruitment of non-indigenous species; influ-ence of unnaturally high exposure to ecosystem engi-neers and other disturbance factors such as grazing bydomesticated geese (Simenstad et al., 2005).

4.5. Avoid conflicting goals

Compromising wetland restoration may in somecases be counterproductive to the intent of both sidesof the compromise. Restoration projects are often sub-jected in the planning stage to sharply divergent pur-poses among stakeholders. The easiest solution is toincorporate a compromise into the restoration design.This may involve modification of the restoration goaland its manifestation in the overall design, or evendivision of the available restoration site into differentb ro-m entb sir-a d byt cificN tuar-i rouso ora-t uaryw efta s theS ishR allyc rol-o ,2 fec-t esst oal

because function is frequently dependent on restorationsite size (e.g., functions associated with tidal channelgeomorphology;Hood, 2002) or may even result indepressed function (e.g., predator attraction from onesite to the adjacent site, or persistent external impacts;Hood, 2004).

4.6. Plan for the long-term landscape

If it is scientifically prudent to incorporate land-scape context in designing and implementing restora-tion projects, it is imperative also to plan strategi-cally and deploy restoration at the landscape scale.The demand for instant gratification often results ina “gardening” approach to restoration that circumventslife-history, natural variability and meso- or long-termcycles, disturbance, succession and other long-termfacilitating processes that dynamically shape land-scapes. Ecosystem processes that dictate a long-termapproach to restoration include soil development andstochastic event-driven disturbances that “reset” land-scape structure (Middleton, 1999). Similarly, sea levelrise, tide regime changes and other emerging changes inregional forcing factors cannot be ignored, especially inregions where they are already a contingency (Orr et al.,2003).

4.7. Learn and adapt, by monitoringprocess-based performance measures

s onc bove,i oree ilet eedi eedst tificr eea lea bjec-t tificu eses,m ratest osti t bet bleslt e the

ut adjoining restoration projects. However, compising goals do not necessarily result in a “differ

ut equal” situation because the likelihood of deble performance of one or both may be threatene

he compromise situation. For example, in the Paorthwest region, competing pressures for both es

ne fish habitat and waterfowl habitat has on numeccasions resulted in the partitioning of large rest

ion sites into two: one parcel reconnected to the estith full tidal dynamics and the remaining parcel ls a freshwater impoundment. In cases such apencer Island restoration project in the Skokomiver estuary, the two restoration projects are actuontained within one relict levee system and hydgy of one is linked through the other (Tanner et al.002). While dual objectives may sometimes be ef

ively met by this approach, it may also lead to lhan the maximum potential benefit for either g

Given the inherent uncertainties and constraintoastal ecosystem restoration we have described at is incumbent that restoration programs become mxplicit learning and experimental platforms. Whhere is often a token tribute to this recognized nn many restoration plans, adaptive management no be applied as it was designed in all its scienigor (Walters, 1986; Halbert, 1993; Lee, 1993; Lnd Lawrence, 1986). One of the more applicabspects of adaptive management is that it is an o

ive process or framework that characterizes scienncertainties, develops strategies to test hypotheasures the response to the test and incorpo

he results into future decisions. Perhaps the mmportant point is that (restoration) strategies musreated as experiments in a framework that enaearning from the results (Lee, 1993). Particularly inhe case of coastal ecosystem restoration, wher


link between processes, structure and function are typ-ically poorly understood, assessment of restoration“experiments” cannot typically be limited to monitor-ing structural attributes. The ability to learn and corrector realign your experimental restoration strategy, orto apply an improved strategy to subsequent restora-tion, depends upon understanding the ecosystem pro-cesses that resulted in the response (structure). Mon-itoring ecosystem processes is not typically includedin restoration monitoring plans, but process-based per-formance measures can provide a much more directindication of what “needs to be fixed” to avoid contin-ued or future restoration dysfunction.

Understanding system responses to natural varia-tions and extremes of ecosystem processes can bemisleading if based solely on restoring ecosystems.Reference or “benchmark” sites are fundamental inexplaining how ecosystem processes affect structuraland functional attributes and scales of the natural vari-ation in these processes and responses (NRC, 1992,1995). Tracking variation in both structural attributesand ecosystem processes at naturally dynamic sites pro-vides critical context in interpreting comparable infor-mation from restoration sites in the same landscape,exposed to the same landscape-scale forcing. Opti-mally, multiple reference sites at different positions inthe landscape should be monitored to capture the mostlikely range of functional equivalency trajectories.

5

an“ tionp g thep ctinge withss gest lessa dless.T whatr inew ora-t ande bili-t hiev-a oeco-

nomic constraints. The actual challenge is not whetheror not restoration is acceptable, but whether we havethe technical and scientific abilities to accomplish it.At the present state of knowledge, we will never knowwhat we have accomplished with restoration actionsif we do not address them as structured experimentswithin a broader application of implementing adaptivemanagement. We particularly need to start learningfrom dedicated case or “demonstration” studies that areintended to develop tools for comprehensive planningof integrated watershed-coastal wetland restorationat landscape scale. To understand the broader contextof how these natural and restoration settings fit into adynamic landscape requires commitment to long-termmonitoring of dedicated reference sites, preferably a“benchmark system.” Only by such a scientific deploy-ment of restoration experiments might we reach astage where we can apply adaptive models (conceptualto simulation) for mechanistic understanding, pre-dictability and management of restoration processes,and what can and cannot be realistically achieved.

Choi (2004)effectively synthesized the need forsuch a ‘futuristic’ approach to restoration, which is to:(i) set realistic and dynamic (instead of static) goalsfor future, instead of past, environment; (ii) assumemultiple trajectories acknowledging the unpredictablenature of ecological communities and ecosystems; (iii)take an ecosystem or landscape approach, instead ofad hoc gardening, for both function and structure; (iv)evaluate the restoration progress with explicit criteria,b ermm

R

A sing

A story:.

A tora-semi-

and

B nticdings–20,

B ohn-eed,

. Summary and conclusions

Restoring aquatic ecosystems is far fromoff-the-shelf” science or technology, and restoraractitioners have a poor track record of addressinrerequisites described above, much less conduffective pre- and post-assessment monitoringtandardized methods (Bernhardt et al., 2005). Under-tandably, the number and difficulty of challenhat must be overcome to actually address, muchccomplish, coastal ecosystem restoration are enhe greatest difficulty is to assess expectations ofestoration can realistically accomplish, and determhether it can even approach the definition of rest

ion at all. Conversely, acceptance of rehabilitationnhancement is not tantamount to failure; a reha

ated or enhanced ecosystem may be the only acble goal given the antecedent, landscape and soci

ased on quantitative inference; (v) maintain long-tonitoring of restoration outcomes.

eferences

ronson, J., Le Floc’h, E., 1996a. Vital landscape attributes: mistools for restoration ecology. Restor. Ecol. 4, 377–387.

ronson, J., Le Floc’h, E., 1996b. Hierarchies and landscape hidialoging with Hobbs and Norton. Restor. Ecol. 4, 327–333

ronson, J.C., Le Floc’h, E., Ovalle, C., Pontanier, R., 1993. Restion and rehabilitation of degraded ecosystems in arid andarid lands. II. Case studies in southern Tunisia, central Chilenorthern Cameroon. Restor. Ecol. 3, 168–187.

aldwin, W.P., 1967. Impoundments for waterfowl on South Atlaand Gulf Coastal marshes. In: Newsom, J.D. (Ed.), Proceeof the Marsh and Estuary Management Symposium. July 19pp. 127–133.

arras, J., Beville, S., Britsch, D., Hartley, S., Hawes, S., Jston, J., Kemp, P., Kinler, Q., Maertucci, A., Porthouse, J., R


D.J., Roy, K., Sapkota, S.K., Suhayda, J.H., 2003. Historic andPredicted Coastal Louisiana Land Changes: 1978–2050. USGSOpen File Report 03-334, pp. 1-27.

Bernhardt, E.S., Palmer, M.A., Allan, J.D., Alexander, G., Barnas,K., Brooks, S., Carr, J., Clayton, S., Cahm, D., Follstad-Shah,J., Galat, D., Gloss, S., Goodwin, P., Hart, D., Hassett, B., Jenk-inson, R., Katz, S., Kondolf, G.M., Lake, P.S., Lave, R., Meyer,J.L., O’Donnell, T.K., Pagano, L., Powell, B., Sudduth, E., 2005.Synthesizing U.S. river restoration efforts. Science 308, 636–637.

Boesch, D.F., Josselyn, M.N., Mehta, A.J., Morris, J.T., Nuttle, W.K.,Simenstad, C.A., Swift, D.J.P., 1994. Scientific assessment ofcoastal wetland loss, restoration, and management in Louisiana.J. Coast. Res. (SI 20).

Bos, D., Bakker, J.P., de Vries, Y., van Lieshout, S., 2002. Long-termvegetation changes in experimentally grazed and ungrazed backbarrier marshes in the Wadden Sea. In: Bos, D. (Ed.), Grazing inCoastal Grasslands, pp. 111–130.

Boumans, R.M., Day Jr., J.W., 1994. Effects of two Louisiana marshmanagement plans on water and materials flux and short-termsedimentation. Wetlands 14 (4), 247–261.

Cahoon, D.R., 1994. Recent accretion in two managed marshimpoundments in coastal Louisiana. Ecol. Appl. 4 (1), 166–176.

Cahoon, D.R., Reed, D.J., Day, Jr., J.W., Steyer, G.D., Boumans,R.M., Lynch, J.C., McNally, D., Latif, N., 1995. The influence ofHurricane Andrew on sediment distribution in Louisiana coastalmarshes. J. Coast. Res., 280–294 (SI 18).

Choi, Y.D., 2004. Theories for ecological restoration in changingenvironment: toward ‘futuristic’ restoration. Ecol. Restor. 19,75–81.

Cohen, A., 1998. Accelerating invasion rate in a highly invaded estu-ary. Science 279, 555–558.

Cowan, J.H.J., Turner, R.E., Cahoon, D.R., 1988. Marsh manage-ment plans in practice: do they work in coastal Louisiana, USA?Environ. Manage. 12 (1), 37–53.

Daily, G.C., Alexander, S., Ehrlich, P.R., Goulder, L., Lubchenco, J.,an,

plied16.

D 003.anc-stuar-

D landseous–2367.

D mea-oast.

D ation266,

E aticitedp. 4.

E andment:.

Esselink, P., Zijlstra, W., Dijkema, K.S., van Diggelen, R., 2000. Theeffects of decreased management on plant-species distributionpatterns in a salt marsh nature reserve in the Wadden Sea. Biol.Conserv. 93, 61–76.

Ford, M.A., Grace, J.B., 1998. The interactive effects of fire andherbivory on a coastal marsh in Louisiana. Wetlands 18, 1–8.

Gray, A., Simenstad, C.A., Bottom, D.L., Cornwell, T.J., 2002. Con-trasting functional performance of juvenile salmon in recoveringwetlands of the Salmon River estuary, Oregon, USA. Restor.Ecol. 10, 514–526.

Grime, J.P., 1977. Evidence for the existence of three primary strate-gies in plants and its relevance to ecological and evolutionarytheory. Am. Nat. 111, 1169–1184.

Halbert, C.L., 1993. How adaptive is adaptive management? Imple-menting adaptive management in Washington State and BritishColumbia. Rev. Fish. Sci. 1, 261–283.

Harrison, C., Enright, C., Schmutte, C., Baye, P., October 2001.Demonstration of a microtidal marsh and lagoon system: Endan-gered estuarine species recovery and waterfowl management inSuisun Marsh. In: Abstract Fifth Biennial State of the EstuaryConference, San Francisco, CA.

Hill, M.J., Long, E.A., Hardin, S., 1994. Effects of dam removal onDead Lake, Chipola River, FL. In: Proceedings of the AnnualConference of Southeast Association of Fish and Wildlfie Agen-cies, pp. 512–523.

Hobbs, R.J., Norton, D.A., 1996. Towards a conceptual frameworkfor restoration ecology. Restor. Ecol. 4, 93–110.

Hood, W.G., 2002. Application of landscape allometry to restorationof tidal channels. Restor. Ecol. 10, 213–222.

Hood, W.G., 2004. Indirect environmental effects of dikes on estu-arine tidal channels: thinking outside of the dike for habitatrestoration and monitoring. Estuaries 27, 273–282.

Huxel, G.R., Hastings, A., 1999. Habitat loss, fragmentation andrestoration. Restor. Ecol. 7, 309–315.

Kanehl, P.D., 1997. Changes in the habitat and fish community of thelen

K .L.,heast

K an,achre-

ntalC.

K mee

K logy626.

L e and.

L fromon.

L Taskthening,

Matson, P.A., Mooney, H.A., Postel, S., Schneider, S.H., TilmD., Woodwell, G.M., 1997. Ecosystem services: benefits supto human societies by natural ecosystems. Issues Ecol. 2,

eLaune, R.D., Jugsujinda, A., Peterson, G., Patrick, W., 2Impact of Mississippi River freshwater reintroduction on enhing marsh accretionary processes in a Louisiana estuary. Eine Coast. Shelf Sci. 58, 653–662.

everel, S.J., Rojstaczer, S., 1996. Subsidence of agriculturalin the Sacramento-San Joaquin Delta, California: role of aquand gasous carbon fluxes. Water Resources Res. 32, 2359

ingler, J.R., Hsu, S.A., Foote, A.L., 1995. Wind shear stresssurements in a coastal marsh during Hurricane Andrew. J. CRes. SI 21, 295–305.

ynesius, M., Nilsson, C., 1994. Fragmentation and flow regulof river systems in the northern third of the World. Science753–762.

PA, 2000. Principles for the Ecological Restoration of AquResources. EPA841-F-00-003. Office of Water (4501F), UnStates Environmental Protection Agency, Washington, DC,

PA Science Advisory Board (SAR), 1998. Ecological impactsevaluation criteria for the use of structures in marsh manageEnvironmental Protection Agency, EPA-SAB-EPEC-98-003

Milwaukee River, Wisconsin, following removal of the WooMills Dam. N. Am. J. Fish. Manage. 17, 387–400.

auffman, J.B., Case, R.L., Lytjen, D., Otting, N., Cummings, D1995. Ecological approaches to riparian restoration in nortOregon. Restor. Manage. Notes 12, 12–15.

entula, M.W., Brooks, R.P., Gwin, S.E., Holland, C.C., ShermA.D., Sifneos, J.C., 1992. In: Hariston, A.J. (Ed.), An Approto Improving Decision Making in Wetland Restoration and Cation. U.S. Environmental Protection Agency, EnvironmeResearch. Lab./Island Press, Corvallis, OR/Washington, D

obel Jr., J.W., 1995. An historical perspective on the KissimRiver Restoration Project. Restor. Ecol. 3, 149–159.

uhn, N.L., Mendelssohn, I.A., Reed, D.J., 1999. Altered hydroeffects on Louisiana salt marsh function. Wetlands 19, 617–

ee, K.N., 1993. Compass and Gyroscope: Integrating SciencPolitics for the Environment. Island Press, Washington, DC

ee, K.N., Lawrence, J., 1986. Adaptive management: learningthe Columbia River basin fish and wildlife program. EnvirLaw 16, 431–460.

ouisiana Coastal Wetlands Conservation and RestorationForce (LCWCRTF), 2003. The 2003 Evaluation Report toU.S. Congress on the Effectiveness of Coastal Wetland Plan


Protection and Restoration Act Projects. Department of NaturalResources, Baton Rouge, LA, 27 pp.

Middleton, B.A., 1999. Wetland restoration, Flood Pulsing and Dis-turbance Dynamics. John Wiley and Sons Inc., New York, NY,388 pp.

Middleton, B.A., 2002. Flood Pulsing in Wetlands—Restoring theNatural Hydrological Balance. John Wiley and Sons Inc., NewYork, NY, 320 pp.

Morgan, P.A., Short, F.T., 2002. Using functional trajectories to trackconstructed salt marsh development in the Great Bay estuary,Maine/New Hampshire, USA. Restor. Ecol. 10, 461–473.

National Research Council (NRC), 1992. Restoration of AquaticEcosystems—Science, Technology and Public Policy. NationalAcademy Press, Washington, DC, 576 pp.

National Research Council (NRC), 1995. Wetlands: Characteristicsand Boundaries. National Academy Press, Washington, DC, 308pp.

National Research Council (NRC), 2001. Compensating for WetlandLosses under the Clean Water Act. National Academy Press,Washington, DC, 322 pp.

Nichols, F.H., Cloern, J.E., Luoma, S.N., Peterson, D.H., 1986. Themodification of an estuary. Science 231, 567–573.

O’Neill, R.V., Gardner, R.H., Turner, M.G., 1992. A hierarchicalneutral model for landscape analysis. Landsc. Ecol. 7, 55–61.

Orr, M., Crooks, S., Williams, P.B., 2003. Will restored tidal marshesbe sustainable? San Francisco Estuary Watershed Science 1 1 (5),http://repositories.cdlib.org/jmie/sfews/vol1/iss1/art5.

Reed, D.J., De Luca, N., Foote, A.L., 1997. Effect of hydrologicmanagement on marsh surface sediment deposition in coastalLouisiana. Estuaries 20, 301–311.

Reed, D.J., Spencer, T., Murray, A.L., French, J.R., Leonard, L.,1999. Marsh surface sediment deposition and the role of tidalcreeks: implications for created and managed coastal marshes. J.Coast. Conserv. 5, 81–90.

Reed, D.J., 2002. Understanding tidal marsh sedimentation inRes.,

R sippi.

R se of0.

R an-ring a

S g damRes.

S ater-mbiaater-ental

S tra-col.

Simenstad, C.A., Warren, R.S., 2002. Introduction to the special issueon dike/levee breach restoration of coastal marshes. Restor. Ecol.10, i.

Simenstad, C.A., Wick, A., Van de Wetering, S., Bottom, D.L.,2003. Dynamics and ecological functions of wood in estuar-ine and coastal marine ecosystems. In: Gregory, S.V., Boyer, K.,Gurnell, A. (Eds.), The Ecology and Management of Wood inWorld Rivers, American Fisheries Society Symposium, vol. 37.Bethesda, MD, pp. 265–277.

Simenstad, C.A., Tanner, C., Cordell, J., Crandell, C., White, J., 2005.Challenges of habitat restoration in a heavily urbanized estuary:evaluating the investment. J. Coast. Res. 40, 6–23.

Suhayda, J., 1997. Modeling impacts of Louisiana’s barrier islandson wetland hydrology. J. Coast. Res. 13, 686–693.

Tanner, C.D., Cordell, J.R., Rubey, J., Tear, L., 2002. Restoration offreshwater intertidal habitat functions at Spencer Island, Everett,Washington. Restor. Ecol. 10, 564–576.

Thom, R.M., Ziegler, R., Borde, A.M., 2002. Floristic developmentpatterns in a restored Elk River estuarine marsh, Grays Harbor,Washington. Restor. Ecol. 10, 487–496.

Tillman, D., Lehman, C.L., Kareiva, P., 1997. Population dynam-ics in spatial habitats. In: Tilman, D., Kareiva, P. (Eds.), SpatialEcology. Princeton University Press, Princeton, NJ, pp. 3–20.

Toth, L.A., Melvin, S.L., Arrington, D.A., Chaberlain, J., 1998.Hydrologic manipulations of the channelized Kissimmee River.BioScience 48, 757–764.

Van Heerden, I., DeRouen, K., 1997. Implementing a barrierisland and barrier shoreline restoration program—the State ofLouisiana’s perspective. J. Coast. Res. 13, 679–685.

Visser, J., Sasser, C.E., Chabreck, R.H., Linscombe, R.G., 2002. Theimpact of severe drought on the vegetation of a sub-tropical estu-ary. Estuaries 25, 1184–1195.

Walters, C., 1986. Adaptive Management of Renewable Resources.MacMillan Publishing Company, New York, NY.

Warren, R.S., Fell, P.E., Rozsa, R., Brawley, A.H., Orsted, A.C.,ora-estor.

W redstuary.

W try:n in

W rces

W trans-aquinFall

W n of

Z tion:l. 7,

the Sacramento-San Joaquin Delta, California. J. Coast.605–611 (SI 36).

oberts, H.H., 1997. Dynamic changes of the holocene Missisriver delta plain: the delta cycle. J. Coast. Res. 13, 605–627

ozas, L.P., 1995. Hydroperiod and its influence on nekton uthe salt marsh: a pulsing ecosystem. Estuaries 18, 579–59

ozas, L.P., Minello, T.J., 1999. Effects of structural marsh magement on fishery species and other nekton before and duspring drawdown. Wetl. Ecol. Manage. 7, 121–139.

human, J.R., 1995. Environmental considerations for assessinremoval alternatives for river restoration. Regul. RiversManage. 11, 249–261.

imenstad, C.A., Jay, D.A., Sherwood, C.R., 1992. Impacts of wshed management on land-margin ecosystems: the ColuRiver estuary as a case study. In: Naiman, R.J. (Ed.), Wshed Management: Balancing Sustainability and EnvironmChange. Springer-Verlag, p. 543.

imenstad, C.A., Thom, R.M., 1996. Functional equivalencyjectories of the restored Gog-Le-Hi-Te estuarine wetland. EAppl. 6, 38–56.

Olson, E.T., Swamy, V., Niering, W.A., 2002. Salt marsh resttion in Connecticut: 20 years of science and management. REcol. 10, 497–513.

illiams, P.B., Orr, M.K., 2002. Physical evolution of restobreached levee salt marshes in the San Francisco Bay eRestor. Ecol. 10, 527–542.

illiams, P.B., Orr, M.K., Garrity, N.J., 2002. Hydraulic geomea geomorphic design tool for tidal marsh channel evolutiowetland restoration projects. Restor. Ecol. 10, 577–590.

olman, M.G., Miller, J.P., 1960. Magnitude and frequency of foin geomorphic processes. J. Geol. 68, 54–74.

right, S.A., Shoellhammer, D.H., 2004. Suspended-sedimentport where rivers become an estuary: Sacramento-San JoRiver Delta, Water Year 1999–2002. Eos Trans. AGU 85,Meet. Suppl. Abstract H53C-1262.

underlich, R.C., Winter, B.D., Meyer, J.H., 1994. Restoratiothe Elwha River ecosystem. Fisheries 19, 11–20.

edler, J.B., Callaway, J.C., 1999. Tracking wetland restorado mitigation sites follow desired trajectories? Restor. Eco69–73.

when is restoration not? incorporating landscape-scale ... · at the landscape scale or in some...

Documents