ecosystem restoration and management: scientific

599

Ecosystem Restoration andManagement: ScientificPrinciples and Concepts

Wallace Covington, William A. Niering, Ed Starkey, andJoan Walker

+ Ecosystem restoration in ecosystem management

+ Ecosystem restora don defined

+ Naturalness and the evolutionary environment

+ Examples of ecosystem restoration approaches

600 W. Covington et al./Ecosystem Restoration and Management

“Once we restore, we are no longer retreating, tying onlyto slow the wave of destruction. We begin to actually advance,to regain lost ground. Can we really do it, or is the idea onlyhubris, human arrogance rearing its head one more time?. ..The short answer is: yes, we can really do it - to somedegree . At worst we can produce something that mimics thereal thing and that, given enough time, could become the realthing . . ..‘I John P. Wiley, Jr., 1989

This paper summarizes current thinking regardingecological restoration from an ecosystem managementpoint of view. The intended audience is natural resourceprofessionals, natural resource interest groups, andinterested members of the public. We discuss ecologicalrestoration concepts in the context of three ecologicalrestoration efforts with which we have been involvedand which are particularly important to contemporarypublic land management: ponderosa pine ecosystems,forest ecosystems of the Western Hemlock Zone of thePacific Northwest, and tidal wetlands of the Northeast.In discussing these examples we emphasize scientificprinciples and concepts fundamental to ecologicalrestoration. We close our paper with a discussion ofecological restoration and human habitat needs.

Others have presented cogent syntheses of ecologi-cal restoration (MacMahon and Jordan 1994), restora-tion ecology (Jordan et al. 1987), and small group andcommunity-based ecological restoration (Nilsen 1991).Although we draw upon these resources for some ofour discussion, our goal is different - to discussecological restoration in the light of contemporaryecosystem management concepts.

Concern about the degradation of public lands andassociated natural resources has been a driving force infederal land management policy since its inception(Dana and Fairfax 1980). A building consensus suggeststhat unless something is done to reverse the deteri-oration of ecosystem health, current and future genera-tions will continue to incur increasing costs whilesimultaneously enjoying fewer benefits from publiclands. Of particular concern is the cumulative effect ofecosystem simplification such that ecosystems are atrisk of catastrophic losses of biological diversity andhuman habitats (Myers 1984). A cornerstone of thefederal government’s ecosystem management ap-proach to solving these problems is ecosystem restora-tion. The Report of the Interagency Ecosystem Man-agement Task Force (Anon 1995) stated:

“The goal of the ecosystem approach is to restore

cal diversity of ecosystems and the overall qualityof life through a natural resource managementapproach that is fully integrated with social andeconomic goals.”

In a similar vein, the Ecological Society of America’sreport, “The Scientific Basis for Ecosystem Management,”discussed the importance of ecological restoration in thepractice of ecosystem management (Christensen et al.1995). In a previous report by the Ecological Society ofAmerica, Lubchenco et al. (1991) selected ecosystemrestoration as one of twelve featured topics for priorityresearch in an ecological research agenda in support ofsustaining the biosphere. Restoration and maintenanceof ecosystem health is seen as central to ecosystemmanagement by such diverse groups as the Society ofAmerican Foresters (1993), the Southwest ForestAlliance (1996), the Sierra Club (e.g., see Berger 1997),and the American Forest and Paper Association ForestResource Board (1993). Restoration of ecosystem healthis, in fact, an international theme. The United Nations(1992) recognized ecosystem restoration as a centralconcern in the Rio Declaration on Environment andDevelopment in Principle 7 which declares, “States shallcooperate in a spirit of global partnership to conserve,protect and restore the health and integrity of theEarth’s ecosystems.” But what is ecosystem restoration?

Modern principles and concepts of ecologicalrestoration began with the thinking of Aldo Leopold(Flader and Callicott 1991). Soon after the beginning ofLeopold’s professional career as a forester in the South-western United States, he recognized the rapid deteri-oration of forest and range lands because of over-grazing, intensive logging, and predator extirpation(Flader 1974). This spurred Leopold to call for viewingnatural resource management as the practice of landhealth (now termed ecosystem health [Rapport 19951).Leopold recognized that the practice of ecosystemhealth required reference points - healthy, intactecosystems still functioning as they had beforedisruption by intensive industrialization - and thatthose reference points were highly limited in theUnited States. Referring to the need for referencepoints for the practice of land health, Aldo Leopoldsaid, “The first step is to reconstruct a sample of whatwe had to begin with.”

Upon his re-entry into the academic community as aprofessor of game management, he joined forces withothers at the University of Wisconsin to rebuild ap-proximations of the naturally functioning ecosystems

Humans as Agents of Ecological Change 601

American settlement (Jordan et al. 1987). Using CivilianConservation Corps, student, faculty, and communityvolunteers, Leopold and others began the task of ecolo-gical restoration of representative Wisconsin eco-systems at the University of Wisconsin Arboretum in1935. That same year, Leopold began the restoration ofhis beloved sand county farmland just an hour’s drivenorth of the campus. Although rehabilitation ofdegraded land was a widespread goal in the 193Os, theecological restoration work of Leopold and others wasdifferent. The ecological restoration projects in Wiscon-sin had as their goal the restoration of native eco-systems in contrast to others where the goal was thesimple revegetation of derelict lands, stopping acceler-ated erosion, or improving the productive potential ofland.

Over the next 50 years, ecologists working on thearboretum restoration projects learned much about thestructure and function of ecological systems throughtrial and error. In 1981 the University of Wisconsinarboretum began publishing Restoration and Manage-ment Notes as a forum for the interchange of ideas andexperiences among practicing restorationists. By the198Os, the synergy between ecological restoration (thepractice) and restoration ecology (the biological sci-ence) became so apparent that it lead to the recognitionof restoration ecology as a focus for developing andtesting ecological theory (Jordan et al. 1987). In 1987 aprofessional society, the Society for Ecological Restora-tion, was formed; it held its first annual meeting in1988. In 1993 the Society began publishing its scientificjournal, Restoration Ecology.

One of the first tasks of the Society for EcologicalRestoration was defining ecological restoration and itsprinciples and concepts. This task is ongoing and muchdiscussion and debate continue among scientists andpractitioners of ecological restoration (Jackson et al.1995, Aronson and Le Floc’h 1996, Covington andSampson 1996, Higgs 1997, Dobson et al. 1997). None-theless, progress has been made such that basic defi-nitions are generally agreed upon.

The dictionary definition of “restoration” is the actof bringing back to an original or unimpaired condi-tion. Thus, ecological restoration has as its goal therestoration of degraded ecosystems to emulate moreclosely, although not necessarily duplicate, conditionswhich prevailed before disruption of natural structuresand processes, i.e., environmental conditions whichhave influenced native communities over recent evolu-tionary time (see Box 1).

Box 1Society for Ecological Restoratbn MissionStatement and Summary of EnvironmentalPolicies (Society for Ecological Restoration

1993)

The mission of the Society Is to promote ecologicalrestoration as a means of sustaining the diversity of life onEarth and reestablishing an ecologically healthy relatian-ship between nature and culture. Ecological restoration Isthe process of reestablishing to the extent possible thestructure , funct ion, and integr i ty of indigenous eco-systems and the sustaining habitats that they provide. Toadvance its mission, (1) SER serves as a forum fordlscuaion and exchdnge of Ideas; (2) SER raises aware-ness and promotes the expanded use of ecologicalrestoration; (3) SER works to advance the science and artof ecological restoration, SER welcomes the participationof anyone Interested in ecological restoration.

Ecological restoration involves management actionsdesigned to accelerate recovery of degraded ecosys-tems by complementing or reinforcing naturalprocesses. Ecological restoration has been viewed asecosystem medicine where the practitioner is helpingnature heal (Nilsen 1991), that is, building upon thenatural recovery processes inherent in the ecosystem.Restoration ecology, the biological discipline whichundergirds ecological restoration, deals with researchand management experimentation to determine themechanisms that control recovery of degraded eco-systems, and with discovering ways for safely restoringdegraded ecological systems to more nearly naturalconditions.

Ecosystem restoration is founded upon fundament-al ecological and conservation principles and involvesmanagement actions designed to facilitate the recoveryor re-establishment of native ecosystems. A centralpremise of ecological restoration is that restoration ofnatural systems to conditions consistent with theirrecent evolutionary environments will prevent theirfurther degradation while simultaneously conservingtheir native plants and animals (Society for EcologicalRestoration 1993). Practitioners of ecological restora-tion recognize that a failure to include human inter-actions with restored systems is not only unrealistic,but also undesirable for their long-term sustainability.In fact, in cases where novel conditions prevent naturalsystem functions, ongoing management may be re-quired to compensate for the unnatural conditions.Examples of such a circumstance are those in whichrestored sites are too small to support natural pred-ator-prey dynamics or to accommodate naturaldisturbance regimes.


Ecological restoration is related to other practices ofecological healing such as rehabilitation, reclamation,and bioremediation, although the goals of ecosystemrestoration (restoration of natural conditions) are gen-erally more ambitious (MacMahon and Jordan 1994).Although restoration goals are often more ambitiousthat those of rehabilitation, this does not necessarilyimply that restoration is more expensive (see dis-cussion of restoration of Ponderosa Pine Ecosystems,below). The term “reclamation” first came into com-mon usage after the U.S. Surface Mine Control andReclamation Act of 1977 (Jackson et al. 1995). Recla-mation refers to attempts tore-establish elements of thestructure and function of ecosystems, but not completerestoration to any specified prior condition. Rehabi-litation has as its goal making the land useful again,but, as in reclamation, the goal is not restoration topredisruption ecological conditions (National Acad-emy of Science 1974). Rehabilitation might involve, forexample, establishment of agricultural land on a sitepreviously occupied by grassland. Reclamation, re-habilitation, and restoration have as their goals a conti-nuum of outcomes from the least to the most similar tothe predisturbance ecosystem (Jackson et al. 1995).Thus, all share to a greater or lesser extent some of thesame techniques and can be viewed as closely allied.

In many respects ecological restoration might bestbe judged by whether the techniques used are settingthe ecosystem on a trajectory that will eventually leadto the recovery of original ecosystem structure andfunction (Bradshaw 1984, MacMahon and Jordan1994). The underlying assumption of such a view is thatfacilitating partial recovery of ecosystem structure andfunction can lead to re-establishment of natural self-regulatory mechanisms which in turn will eventuallylead to restoration of the original ecosystem dynamics.

Ecological restoration, therefore, consists of a broadvariety of practices designed to restore natural ecosys-tem structure and function. It is related to reclamation,rehabilitation, and other land recovery practices buthas as its goal re-establishment of the original eco-system structure and function. For example, in the caseof southwestern ponderosa forests, ecosystem restora-tion might consist of removing most of the trees thatpostdate Euro-American settlement, raking heavyfuels from the base of the old-growth trees, prescribedburning, removing introduced noxious plants, andsowing with native herbaceous seeds. In the case ofconifer forests of the Pacific Northwest and tidalwetlands of the northeastern United States, differentrestoration activities are required (see below).

However, ecosystem restoration should not be con-strued as a fixed set of procedures, nor as a simplerecipe for land management. Rather, it is a broad intel-

lectual and scientific framework for developing mutu-ally beneficial human:wildland interactions compat-ible with the evolutionary history of native ecologicalsystems. In other words, ecosystem restoration consistsnot only of restoring ecosystems, but also of devel-oping human uses of wildlands which are in harmonywith the natural history of these complex ecologicalsystems.

“Natural” is one of the most controversial concepts inecosystem management (Christensen et al. 1995). Al-though it can be an ambiguous term, it is one that isecologically, aesthetically, spiritually, and politicallyimportant as evidenced by its use in such expressionsas natural area, natural range of variability, naturalhistory, and natural processes. When used to connotethe evolutionary environment, it is fundamental toconservation biology and restoration ecology. In thecontext of restoration ecology ynaturalN implies nativespecies, structures, and processes, in contrast to exoticspecies, structures, and processes. Indigenous ecolo-gical components and processes are natural. Alienecological components and processes are not. At theheart of these distinctions is the evolutionary ecologyprinciple that species which have interacted overevolutionary time will have developed coevolvedregulatory mechanisms and interdependencies thatlead them to function as relatively self-regulatingecological systems.

Naturalness is difficult to quantify. Karr (1981) pro-posed an index of biological integrity for assessing thenaturalness of aquatic ecosystems. His proposed indexranged from 12 in areas without fish to 60 in areas withfish composition equivalent to those in undisturbedareas. Karr’s index integrates attributes such as fishspecies richness, indicator taxa (both tolerant and in-tolerant of pollution), species and trophic guild relativeabundances, and the incidence of hybridization,disease and anomalies such as lesions, tumors or finerosions. Similar approaches could be used in otherecosystems. For example, in forest ecosystems the fo-cus might be on developing an index which integratesplant species richness, indicator animal taxa, and theincidence of hybridization, disease and anomalies inkey plant and animal species.

Anderson (1991) wrestled with the problem andsuggested that naturalness could be assessed by theproportion of native to non-native species, the amount

Humans as Agents of Ecological Change-

603

of human energy needed to maintain current eco-system conditions, and the relative change in theecosystem if human inputs to the system cease. Theconcept of naturalness is sometimes best understood inthe context of defining what is unnatural (Hammondand Holland 1995). In this view, a natural ecosystemwould be constituted of indigenous (native) speciesinteracting in a self-sustaining manner, i.e., species per-sistence by natural recruitment as opposed to managedreproduction, population dynamics regulated internal-ly, disturbance regimes functioning within their pre-disruption range of variability, and trophic dynamicsthat are sustainable over time. An unnatural ecosystemwould have a high proportion of non-native species,wide swings in population dynamics requiringmanagement actions to prevent ecosystem simplifi-cation, and exotic disturbance regimes far outsidethose present before ecosystem degradation.

From a restoration ecology point of view the mostimportant definition of “natural” is related to the con-cept of the evolutionary environment, a key element ofdefining the reference conditions for ecologicalrestoration projects.

The concept of the evolutionary environment is centralto conservation biology and restoration ecology. Theterm evolutionary environment refers to the environ-ment in which a species or groups of species evolved -the environment of speciation (sometimes referred toas the habitat of speciation) (Mayr 1942, Smith 1958,Geist 1978).

Over evolutionary time species not only adapt totheir evolutionary environment, but they may alsocome to depend upon those conditions for their conti-nued survival (Mooney 1981, Wilson 1992). Thus, thegreatest threat to biological diversity is the loss ofevolutionary habitats (Noss 1991), and the greatesthope for reversing the losses is restoration of thesehabitats (MacMahon and Jordan 1994). But on whattime-scale is the evolutionary environment measured?

This question has no simple answer. Evolution is anongoing process and rates of evolution are a function ofgeneration time, population structure, genetic vari-ability, selection pressure, and other factors. Today’sspecies are the product of millions of years of evolu-tion. However, in the context of contemporary com-munities the relevant evolutionary environment isgenerally considered to be that of the past several thou-sand years (for most forest ecosystems this would ap-proximate 50-100 times the average generation time ofthe longest lived ecological dominant trees). Based onevolutionary principles, MacArthur (1972) concluded

that “...the length of time it normally takes for a speciesto split and diverge sufficiently to be regarded as twospecies is a small, uncertain number of thousands ofyears.” A fundamental assumption is that an environ-mental factor can be considered as part of a species’evolutionary environment when that factor has beenof sufficient intensity and duration for the factor toexert selection pressure such that the species hasbecome adapted to it.

For North America, the recent evolutionary envi-ronment is typically taken to include Native Americansas participants in evolutionary processes over the pastten thousand years (Parsons et al. 1986, Kay 1995,Bonnicksen et al. this volume). However, the environ-mental pressures associated with Euro-Americansettlement, especially the introduction of exotic plants,animals, and land use practices, as well as the dis-ruption of natural disturbance regimes (see White etal., this volume), are unprecedented in the recent evo-lutionary environment and thus viewed as disruptingevolutionary trajectories (Covington et al. 1994) andleading to pervasive degradation of ecological systems.

Ecological restoration is now seen as an approachfor reversing ecosystem degradation and setting eco-systems on a trajectory more consistent with their evo-lutionary environment. With this approach in mind,we now present an overview of examples of ecologicalrestoration in three major types with which we havedetailed experience: ponderosa pine forests of theSouthwest, conifer forests of the Western HemlockZone of the Pacific Northwest, and tidal wetlands ofthe Northeast. We use ecological restoration work insouthwestern ponderosa pine to illustrate the use ofdetailed historical and field research based knowledgeto design small-scale (l-10,000 acre) ecologicalrestoration experiments. Conifer forests of the PacificNorthwest are used to illustrate the use of ecologicalrestoration research in the design of a regional eco-system management approach. Our final example,restoration of tidal wetlands in the Northeast, is used toextend our discussion of ecological restorationprinciples beyond forest ecosystems to aquatic andwetland ecosystems.

The Problem

The evolutionary environment of southwesternponderosa pine ecosystems is dominated by natural


disturbance regimes (e.g., fires, predation, defoliation),which have varied in kind, frequency, intensity, andextent (Covington et al. 1994). These disturbance re-gimes served as natural ecological checks and balanceson populations and insured spatial and temporalhabitat diversity (Cooper 1960, Covington and Moore1994b). Natural fire regimes were particularly im-portant in shaping the communities present at the timeof Euro-American settlement.

Previous research has established that southwesternponderosa pine forests were much more open beforeEuro-American settlement (ca. 1870) than they aretoday (Pearson 1950, Cooper 1960, Madany and West1983, Covington and Sackett 1986, Covington andMoore 1994b). Before settlement, the combination offrequent (every 2-5 years), light surface fires, grasscompetition, and a climate unfavorable for pineestablishment had maintained an open and park-likelandscape, dominated by grasses, forbs, and shrubswith scattered groups of ponderosa pine trees. AfterEuro-American settlement, heavy livestock grazing,fire suppression, logging disturbances, and favorableclimatic events favored the invasion of the openpark-like vegetation by dense ponderosa pineregeneration.

Various authors (e.g., Cooper 1960, Weaver 1974,Kilgore 1981, Covington and Moore 1994a, 1994b, Kolbet al. 1994) have described symptoms of ecosystemdegradation of ponderosa pine ecosystems includingincreases in tree density, forest floor depth, and fuelloading and consequent problems such as: (1) de-creases in soil moisture and nutrient availability; (2)decreases in growth and diversity of both herbaceousand woody plants; (3) increases in mortality in theoldest age class of trees; (4) decreases in stream andspring flows; (5) accumulation of fuels; and (6)increases in fire severity and size. These symptoms areconsistent with the general ecosystem health distresssyndrome for terrestrial ecosystems as discussed byRapport (1995) and Rapport and Yazvenko (1996), i.e.,reductions in species diversity, leaching of nutrients,reduction in primary productivity, increasedamplitude of oscillations of component species,increase in diseases, and reduction in size of dominantorganisms.

Reference Condi t ions

Reference conditions in southwestern ponderosa pineecosystems come from three lines of evidence: histo-rical records, retrospective ecological analyses, andanalogous sites in the Sierra Madre Occidental which

Histor ica l Records

Reports from early travelers illustrate the changes inappearance of ponderosa pine forests since settlement.E.F. Beale who travelled through northern Arizona1858 is quoted by C.F. Cooper (1960) as follows:

“We came to a glorious forest of lofty pines,through which we have travelled [sic] ten miles.The country was beautifully undulating, and al-though we usually associate the idea of barren-ness with the pine regions, it was not so in thisinstance; every foot being covered with the finestgrass, and beautiful broad grassy vales extendingin every direction. The forest was perfectly openand unencumbered with brush wood, so that thetravelling [sic] was excellent.”

Cooper (1960) went on to state, “The overwhelmingimpression one gets from the older Indians and whitepioneers of the Arizona pine forest is that the entireforest was once much more open and park-like than itis today.”

Before European settlement of northern Arizona inthe 1860s and 187Os, periodic natural surface fires oc-curred in ponderosa pine forests at frequent intervals,perhaps every 212 years (Weaver 1951, Cooper 1960,Dieterich 1980, Swetnam and Baisan 1996). Severalfactors associated with European settlement caused areduction in fire frequency and size. Roads and trailsbroke up fuel continuity. Domestic livestock grazing,especially overgrazing and trampling by cattle andsheep in the 1880s and 189Os, greatly reduced herba-ceous fuels. Active fire suppression, as early as 1908 inthe Flagstaff area, was a principal duty of early forest-ers in the Southwest. A direct result of interrupting andsuppressing these naturally occurring, periodic fireshas been the development of overstocked forests.

Cooper (1960) cites the writings of early expeditionleaders, Whipple and Beale, both of whom travelledthrough northern Arizona in the 1850s. They reportedthat the condition of the southwestern ponderosa pineforest ” . ..was open and park-like with a dense grasscover.” These early descriptions of the open nature ofpresettlement ponderosa pine forests are in agreementwith results of recent research which found thatcanopy coverage by trees of presettlement origin rangefrom 17 percent to 22 percent of the surface area forunharvested sites near Flagstaff, AZ (White 1985,Covington and Sackett 1986, Covington et al. 1997).

Retrospective Ecological Analyses

Cooper (1960) stated that the structure of the south-


of east-central Arizona is actually that of an all-agedforest composed of even-aged groups. He noted greatvariation in diameter within a single age class. White(1985), in a study conducted on the Pearson NaturalArea near Flagstaff, reported that successful establish-ment of ponderosa pine in presettlement times wasinfrequent (as much as four decades between re-generation events). White also determined that stemswere strongly aggregated, the aggregation rangedfrom 3 to 44 stems within a group, with a groupoccupying an area that ranged from 0.05-0.7 acres.“Ages of stems within a group were also variable withthe most homogeneous group having a range of 33years and the least having a range of 268 years (White1985).” White’s findings of a pattern of uneven-agedgroups near Flagstaff are in contrast to the results ofCooper (1960) for the White Mountains.

Madany and West (1983) discuss the effects thatmany years of heavy grazing and fire suppression havehad on ponderosa pine regeneration in southern ZionNational Park, Utah. They suggested that ponderosapine seedling survival was probably greater in the early1900s than in the presettlement days because ofreduced competition of grasses (through grazing) withpine seedlings, and the reduced thinning effect thatfires once had on seedlings in presettlement times.

The Restoration Process

A fundamental issue is what treatment or combinationof treatments is necessary for rapidly restoring somefacsimile of a healthy ponderosa pine ecosystem. Thetwo leading management plans for ecological resto-ration of ponderosa pine ecosystems are prescribedburning and thinning from below (Williams et al. 1993,Covington and Moore 199413, Arno et al. 1995, Clarkand Sampson 1995).

Previous research has shown that althoughprescribed burning alone (without thinning or manualfuel removal) can reduce surface fuel loads, stimulatenitrogen availability, and increase herbaceous product-

. ivity, it can cause high mortality of the presettlementtrees (40 percent mortality over a 20-year period) andlethal soil temperatures under presettlement tree

. canopies (Covington and Sackett 1984, 1990, 1992,Harrington and Sackett 1990, Sackett et al. 1996).Although some thinning of postsettlement ponderosapine trees was accomplished by prescribed burning(Harrington and Sackett 1990), results were localized,unpredictable, and difficult to control. Furthermore,reburning, even under very conservative prescriptions(low air temperatures and low windspeed), can producedangerous fire behavior because the continuing high

density of postsettlement trees provides a continuousfuel ladder and thus a high crown fire potential. Clearly,existing research shows that prescribed burning alone(without some mechanical fuel treatments) in today’sunnatural ecosystem structure will not restore naturalconditions in ponderosa pine/bunchgrass ecosystems.Thus, some combination of thinning, manual fuelremoval, and prescribed burning will be necessary forrapidly restoring these systems to natural conditions.

Example of Deta i led Ecologica l Restorat ionExper imentat ion

In 1993 a small-scale ecosystem management researchproject was initiated at the Gus Pearson Natural Areanear Flagstaff, Arizona, to test ecological restorationhypotheses (Covington et al. 1997). The research wasguided by the general hypothesis that: (1) bothrestoration of ecosystem structure and reintroductionof fire are necessary for restoring rates of decom-position, nutrient cycling, and net primary production(NPP) to natural (presettlement) levels; and (2) that therates of these processes will be higher in an ecosystemthat is operating within some facsimile of its naturalstructure and disturbance regime. Specifically, theresearch hypothesis was that re-establishing pre-settlement stand structure alone (thinning post-settlement trees) will result in lower rates ofdecomposition, nutrient cycling, and NPP compared tothinning, forest floor fuel manipulation, andprescribed burning in combination, but that both ofthese treatments will result in higher rates of theseprocesses compared to controls (see below). Theyfurther hypothesized that without periodic burning tohold them in check, pine seedling populationirruptions will recur on the thin-only treatment.

Specific questions addressed were:

1 . How has ecosystem structure (by biomass compo-nent) and nutrient storage changed over the pastcentury of fire exclusion in a ponderosa pine/bunchgrass ecosystem?

2 . What are the implications of these changes forNPP, decomposition, nutrient cycling, and otherkey ecosystem characteristics?

3 . Does partial restoration (restoring tree structurealone by thinning postsettlement trees) differfrom complete restoration (the same thinning,plus forest floor removal, loading with herbaceousfuels, and prescribed burning) in its effects on eco-system structure and function?

Using a systematic approach, the authors established

606 W. Covington et aLlEcosystem Restoration and Management

replicated small plot studies to test these hypotheses.Details on the experimental design, variables measured,and analytical techniques are available in Covington etal. (1997). In addition to tree density, herbaceousvegetation cover, and fuel loading, a broad range ofecological attributes related to ecosystem health arebeing monitored.

Dendrochronological analysis revealed that foreststructure had changed substantially in the study areasince fire regime disruption. Particularly striking wasthe population irruption of ponderosa pine from 24.3trees per acre in 1876 to 1,254 trees per acre in 1992. Theirruption of smaller diameter pine trees also created acontinuous tree canopy cover at the expense ofherbaceous vegetation. In 1876, only 19 percent of thesurface area was under pine canopy with the balance(81 percent) representing grassy openings, whereas in1992 pine canopy covered 93% of the area, with only7% left as grassy openings.

Thinning resulted in the removal of a total of 5,500bd.ft./ac (3,700 bd.ft./ac of 9-16 in dbh trees, 1,800 bd.ft./ac of 5-8.9 in dbh trees). Most of the smaller diametertrees (629 trees per acre in the 14.9 in dbh class) wereutilized as latillas for adobe home construction. Amajor problem in utilization was what to do with the 37tons per acre of thinning slash. Because there was nomarket for this material, it was hauled (70-80,18-wheeldump truck loads) to a borrow pit and burned.

In the complete restoration treatment, approximate-ly 21.3 tons per acre of duff were removed by raking,some of which was utilized as garden mulch. Addi-tional treatments on the complete restoration treat-ment (addition of mown grass and prescribed burning)left 4.3 tons per acre. Fire intensities were low with anaverage flame length of six inches. Overall, soil heatingwas negligible except under heavy woody fuels andcambial heating was low.

During the 1995 growing season, soil moisture andtemperature were consistently higher in treated areasthan in the control. These microenvironmental differ-ences between treated and control areas will likelyresult in higher rates of key soil processes, such as fineroot production, litter decomposition, and nitrogenmineralization in treated stands, and these changes inthe soil process rates will, in turn, increase herbaceousproduction and tree growth and resistance to insectattack. In this regard, resin flow of presettlement treeswas higher on the thinned area than on the controlarea, as was foliar toughness, suggesting increasedresistance to bark beetles and foliage feeding insects.No changes in populations of turpentine beetles wasobserved.

Herbaceous production responded markedly to the

treatments, with the greatest response to date in thegrassy substratum. By 1995, the treated areas were pr@ducing almost twice as much herbaceous vegetation asthe controls.

Preliminary results from this ecosystem restorationresearch are encouraging. The combination of thinningand burning changed forest structure such that therestored area has shifted from fire behavior fuel model9 (Anderson 1982), where crown fires are common, tofuel model 2, where surface fires occur but wherecrown fires are impossible.

The reduction in tree competition has improvedon-site moisture availability and has likely increasedinsect resistance of presettlement trees. Grasses, forbs,and shrubs are responding favorably as well, indica-ting a shift away from a net primary productivity domi-nated by pine toward a more diverse balance across abroader variety of plants,

Ecosystem restoration research requires a long-term, interdisciplinary commitment. Ecosystemattributes being measured are likely to be in transitionfor the next lo-20 years before they stabilize aroundsome long-term mean. Therefore, Covington et al.(1997) plan to continue this project for the next 24 years(eight 3-year burning intervals), with subsequent burn-ing coinciding with the natural burning season duringthe spring and summer. Data on other attributes will beused to increase understanding of the restorationtreatments on a range of ecosystem characteristics.

The preliminary results from the small plot studiesare so encouraging that the authors have joined withthe Arizona Strip District of the Bureau of Land Man-agement to test practical ecosystem restoration treat-ments on an operational scale in the Mount TrumbullResource Conservation Area, north of the GrandCanyon (Taylor 1996, Covington 1996). At MountTrumbull, they are working with the BLM using anadaptive ecosystem management approach (Waltersand Holling 1990) to restore over 3,000 acres ofponderosa pine to conditions approximating those thatexisted before Euro-American settlement. They aremonitoring a subset of the basic ecosystem healthattributes measured in the small plot study, but byvirtue of the larger size of the treatment areas, they areable to measure some variables which operate on alarger-scale such as passerine bird populations, com-munity structure of selected insect guilds (e.g.,butterflies), and variables indicative of landscape-scaleecosystem health. The hope is that through such a setof integrated, adaptive ecosystem restoration projectsmany of the symptoms of ecosystem pathology can bealleviated while simultaneously increasing under-standing of ecosystem structure and function.


The Problem

.

Timber harvesting has been an important element ofthe economy of the coastal Pacific Northwest since thearrival of European settlers in the 1800s. Although theextensive forests were initially considered to beobstacles to be cleared to make way for agriculture,their economic value as sources of wood was soonrecognized. In the late 1800s and early 19OOs, rates ofharvest began to increase, and escalated greatlyfollowing World War II. By the late 198Os, most forestson private lands had been harvested at least once, andold-growth forests were generally limited to federallands. More than 80 percent of pre-logging old-growthforests in the region had been removed (Booth 1991).

During the 1970s and 198Os, management of federalforests became increasingly controversial. Much of theconcern focused on the habitat requirements of thenorthern spotted owl (Strix occidentalis cuurina), a resi-dent of old-growth forests. Largely because of conti-nued loss of suitable habitat and the lack of regulationsor policies to protect northern spotted owls, in 1990 theU.S. Fish and Wildlife Service listed the subspecies asthreatened under the Endangered Species Act of 1973.Listing of the northern spotted owl greatly reduced thequantity of timber sold from federal lands in westernOregon and Washington, and northern California.

Although preservation of the spotted owl has been afocal point of the debate, the owl also served as asurrogate for other organisms of old-growth forests,and of the forest itself. The real questions were muchbroader, and included the following: What are appro-priate management objectives for federal forests?, Howmuch old-growth forest should be retained?, Is clearcutlogging an acceptable harvest method?

In an attempt to resolve the complex issue, on April2,1993, President Clinton convened a forest conferencein Portland, Oregon, at which he instructed federalagencies to work together and develop a “scientifically

. sound, ecologically credible and legally responsible”plan to restore, protect, and maintain the long-termhealth of forests, wildlife, and waterways. This plan

. would also provide for human and economic concernsand “produce a predictable and sustainable level oftimber sales and nontimber resources that will notdegrade or destroy the environment.”

Following the conference, an interdisciplinarygroup of scientists was brought together as the ForestEcosystem Management Assessment Team (FEMAT),to “identify management alternatives that attain thegreatest economic and social contribution from the

forests of the region and meet the requirements of theapplicable laws and regulations...” (USDA et al. 1993).Subsequently the President chose an alternative(“Option 9”) which would include approximately 7.4million acres within Late-Successional Reserves, 2.6million acres within Riparian Reserves, and 1.5 millionacres of Adaptive Management Areas within whichapplication and testing of ecosystem managementtechniques are encouraged. New, more restrictive,standards and guidelines were also developed fortimber harvest within the approximately 4 millionacres of federal lands outside of reserves and otherareas withdrawn from timber harvest. Following thepreparation of a Final Supplemental EnvironmentalImpact Statement (USDA/USDI 1994a), on April 13,1994, the Secretaries of Agriculture and Interior signeda joint Record of Decision (USDAKJSDI 1994b) to im-plement this plan.

The Record of Decision applies to federal forestswithin the range of the northern spotted owl. How-ever, in this example we focus on the Western HemlockZone, the most extensive vegetation zone in westernOregon and Washington, and the most important interms of timber production (Franklin and Dyrness1973). This zone is dominated by western hemlock(Tsugu hete~ophyllu), western redcedar (Thuju plicutu),Douglas-fir (Pseudotsugu menziesii), and grand fir (Abiesgmndis).

The Forest Ecosystem Management AssessmentTeam was specifically asked to develop alternatives forlong-term management which met the followingobjectives (USDA et al. 1993):

l Maintenance and/or restoration of habitat condi-tions for the northern spotted owl and the marbledmurrelet (B~uchy~umpus mu~mo~utus) that will pro-vide for viable populations, well distributed withintheir current ranges on federal lands.

l Maintenance and/or restoration of habitat condi-tions to support viable populations, well distributedacross their current range, of species known (or rea-sonably expected) to be associated with old-growthforest .

l Maintenance and/or restoration of spawning andrearing habitat on federal lands to support recoveryand maintenance of viable populations of anadro-mous fish species and stocks and other fish speciesand stocks considered “sensitive” or “at risk.”

l Maintenance and/or creation of a connected or in-teractive old-growth forest ecosystem on federallands.

Accomplishing these objectives will require an un-precedented application of principles of ecological


restoration across 24.3 million acres of federal lands(USDA et. al 1993), at stand-level, watershed, andregional scales.

Reference Condi t ions

Pollen records from the Pacific Northwest suggest thatforests of modern composition were first establishedabout 6,000 years BP following retreat of lowlandglaciers (Brubaker 1991). These forests consist of long-lived conifer species which become massive as theyage. With time, these forests develop characteristicPacific Northwest old-growth attributes: patchy,multi-layered canopies with trees of several age classes,large live trees, and an abundance of snags and fallenlogs. Although rate of development for these structuralcharacteristics is variable, old-growth characteristicscommonly begin to appear in unmanaged forests at175-250 years of age (USDA et al. 1993).

These forests provide habitat for an exceedingly richbird and mammal fauna (Harris 1984), with somespecies occupying unique ecological niches. One suchspecies, the red tree vole (Aborimus longicuudus), is themost specialized vole in the world and the mostarboreal mammal in North America (Maser et al. 1981).Northern flying squirrels (Gluucomys sabrinus) are theonly North American forest mammal that consumeslichens as their primary forage (Harris 1984). Both ofthese mammals are important prey species for thenorthern spotted owl, which is closely associated withold-growth forests. Arthropods are also very diverse,with as many as 7,000 species inhabiting these forests(USDA et al. 1993). Although forests of the PacificNorthwest are relatively young, as measured on geolo-gical time scales, a rich fauna has co-evolved with theseplant communities.

Natural fire-return intervals for these forests arevariable, ranging from less than 100 to several hundredyears (Agee 1991). Intense fires are important elementsof the natural fire regime, but lower severity fires alsooccur (Agee 1991). Unlike southwestern ponderosapine forests, fire suppression has only minimally influ-enced forests of the region. Fire suppression becameeffective after 1910, and with long fire-return intervals,the effects of 85 years of fire exclusion have been rela-tively minor (Agee and Edmonds 1992).

The forest was not an unbroken block of old-growth. Prior to European settlement, the regionallandscape consisted of a shifting mosaic of forestcommunities in varying stages of successional devel-opment following disturbance. During the early lBOOs,prior to extensive fires caused by settlers during the184Os, nearly 40 percent of forests in the Coast Range of

These younger forests, however, were structurallydifferent than those regenerated after timber harvest.The following discussion focuses on management ofyoung stands to accelerate the development of struct-ural attributes of old-growth forests, and provides anexample of restoration challenges faced by forestmanagers in the Pacific Northwest.

Restora t ion Goa ls and Treatments

The heart of the plan is a 10 million-acre federal system(USDA and USDI 1994a) of late-successional andriparian reserves. These reserves were established toprovide an inter-connecting network of late-successional and old-growth forests, but at the presenttime they contain large areas of younger forest. Morethan 50 percent of the area within late-successional andriparian reserves supports younger forests (USDA andUSDI 1994a). Most of these younger stands wereusually established following fire or timber harvest.Regeneration of these stands was designed to producehigh yields of lumber, not to produce old-growth forestcharacteristics.

With enough time, some of these forests and associ-ated processes, communities and species will assumeold-growth characteristics without intervention.Others may not, or will only do so over greatly length-ened time frames.

Acceleration of the development of old-growthcharacteristics in these stands is desirable, and wouldlikely improve the long-term potential for success ofthe Northwest Forest Plan. Northern spotted owls andmarbled murrelets may be important beneficiaries ofsuch management. For both of these species, there isconcern about population viability during a transitionperiod lasting until habitat conditions improve signi-ficantly (USDA et al. 1993).

Most young forests within the reserve system havebeen managed under an even-aged system. Followingclearcutting, regeneration was accomplished with sitepreparation, planting of nursery-grown seedlings, andcontrol of competing vegetation. Standing snags werecommonly removed because of safety concerns. Oftenthese forests were thinned lo-20 years after establish-ment to control density and ensure uniform spacing.Many such stands subsequently go through a stemexclusion or self-thinning stage during which much ofthe understory is lost (Oliver and Larson 1990).

The primary objective of these practices is to create auniform conifer stand which quickly achieves crownclosure and dominates nonconifers (Tappeiner et al.1992). This process of stand development is, however,quite different from that following natural disturbance.


continue for 40-100 years (Agee 1991), rather than as apulse of regeneration extending for a relatively briefperiod. Many old-growth trees were apparently estab-lished in relatively open conditions with little compe-tition for 100 years or more (Franklin et al. 1981). Suchstands may not have gone through the self-thinningstage associated with commercial forests. Understoriespersisted, and younger trees were established andbecame intermediate canopies. The dominant old treesretained deeper crowns, thus providing nest sites andthermal cover for wildlife species such as spotted owlsand marbled murrelets.

The Northwest Forest Plan encourages the use ofsilviculture to “accelerate the development of youngstands into multilayered stands with large trees anddiverse plant species, and structures that may, in turn,maintain or enhance species diversity.” Treatment willfocus on stands that have been regenerated after treeharvest or on stands that have been thinned and willinclude, but not be limited to: (1) thinning or managingthe overstory to produce large trees, release advancedregeneration of conifers, hardwoods, or other plants, orto reduce risk from fire, insects, disease, or other envi-ronmental variables; (2) underplanting and limitedunderstory vegetation control to begin development ofmultistory stands; (3) killing trees to make snags andlogs on the forest floor; (4) reforestation; and (5) use ofprescribed fire (USDA et al. 1993, USDA and USDI1994).

Tappeiner et al. (1992) discussed systems which canbe used to accelerate the development of standstructures which are important for northern spottedowls, and to grow habitat in stands where it is unlikelyto develop naturally. Simulated outcomes of severalsilvicultural prescriptions, based on data from actualstands, were included. The following is an example of aprescription and simulated response for a Douglas-firforest in the Coast Range of Oregon.

.

.

The stand contained Douglas-fir, grand fir, and bigleaf maple, with an initial density of 881 stems/acre. Atage 40 years, 50 percent of the conifers with diametersfrom 10 to 16 inches were removed and one-hundredconifers/acre were planted. At age 60 years, 50 percentof trees with diameters from 10 to 22 inches were con-verted to snags, logs on the forest floor, or removed,and 100 conifers/acre were planted. At age 80 years, 60percent of the conifers 8 to 22 inches in diameter weremade into snags, down logs, or removed, with 224conifers and 70 hardwoods remaining. At age 120 years,the stand was projected to have a multi-story canopywith 53 percent cover, and several trees/acre greaterthan 42 inches in diameter. Untreated, the stand wouldhad a single layered canopy, no trees with diametersgreater than 42 inches, and a sparse understory.

Diameter distributions of the treated stand were similarto those measured in stands providing suitable habitatfor northern spotted owls (Tappeiner et al. 1992).

Although this example was developed for a Douglas-fir forest in the Coast Range of Oregon, similar ap-proaches could be applied in other areas, including themixed conifer forests of southern Oregon and northernCalifornia. Tappeiner et al. (1992) provide the followingsuggestions for silvicultural systems designed toemulate natural disturbance and stand development:

l Favor some large trees with numerous limbs for po-tential nest sites.

. Use hardwoods to help develop a multi-layeredstand.

l Encourage the growth of advanced regeneration ofshade-tolerant conifer and hardwood species.

l Establish new regeneration by planting or seedingin young stands after making small openings or re-ducing overstory density in parts of a stand.

l Vary the distribution of overstory trees when thin-ning. Make openings for new regeneration and torelease advanced regeneration.

l When thinning, leave some trees in the smallercrown size classes to help promote a layered stand.

. In stands with irregularly spaced trees, consider acrown thinning to release individual trees whilemaintaining the irregular spacing.

Although the primary goal of silviculture systems forlate-successional reserves is the restoration of old-growth characteristics, important economic benefitscan be derived from thinning these stands. Educationprograms may be required, however, to illustrate theecological benefits of such activities. Importantly,managers must establish and maintain a high level ofpublic trust when planning silviculture systems forreserves.

Adapt ive Management and Ecosystem Restora t ion

Ecosystems are extremely complex, and often it isdifficult or impossible to predict accurately the impactof management actions on future conditions. Failure toact also has unforeseen consequences. The NorthwestForest Plan resolves this dilemma by requiring adap-tive management (Holling 1978, Walters 1986).

Adaptive management as envisioned in the forestplan is “a continuing process of action-based planning,monitoring, researching, evaluating and adjustingwith the objective of improving the implementationand achieving the goals” of management standards


and guidelines. Simply stated, managers and scientistslearn from experience and use this knowledge toimprove subsequent actions.

Silvicultural activities in late-successional reservesprovide a good example of the application of adaptivemanagement. Initially, existing knowledge of silvicult-ure is reviewed and synthesized as a basis for develop-ment of a prescription. During and after implementa-tion, a scientifically and statistically credible monitoringand evaluation program determines whether develop-ment of old-growth characteristics has been achieved,and, finally, new knowledge and information isincorporated into new prescriptions for similar areas.

Because forest succession occurs slowly from ahuman perspective, long-term monitoring programsare required. Successional data from these programswill provide important information upon which torefine simulation models of stand development, andmore accurately project the outcomes of silviculturesys tems .

Restorat ion Chal lenges

Restoration of old-growth forests and processes will bedifficult and exceedingly complex. We have focused onthe use of silviculture to restore old-growth character-istics at the stand level. We have not addressedimportant issues such as landscape level integration,restoration of fire as a natural process, or restoration ofanadromous fish habitat. Stand-level restoration willprovide key building blocks with which to restore andmaintain landscape and regional scale ecosystems. Theultimate ecological test of the Northwest Forest Planwill be its ability to provide an evolutionary environ-ment which maintains native species and naturalprocesses and allows them to continue to evolve.

The Problem

Over the last half century the salt marshes along theNortheast coastline have been severely impacted be-cause of human activities (Anon 1961, Tiner 1984). Theyhave been dredged, filled, and impounded, which haseither destroyed the marsh vegetation or modified itfloristically. In Connecticut, 40 percent of the originalsalt marshes that fringed Long Island Sound have beendestroyed. Thus, the need for restoration is obvious notonly to compensate for these losses but also to restoredegraded systems. Because of tidal restriction along thevalley marshes which characterize many of the NewEngland type salt marshes, vast acreages of Spautina-

dominated communities have been transformed intomonocultures of Phragmites australis (common reed)(Haslam 1973, Niering and Warren 1980). Those filledbut not developed also exhibit a similar monoculture.There is an urgent need to restore tidal flushing torestricted systems, reclaim filled marshes and encour-age Spartina plantings in those areas where marshescan be recreated and thus reconnect these productivetidal wetlands with the surrounding estuarine waters.

The Refe rence Marsh

The reference salt marsh is an undisturbed estuarine,emergent wetland dominated by Spartim alternifora(saltwater cord grass) and Spartina patens (salt meadowcordgrass) (Miller and Egler 1950, Niering and Warren1980, Nixon 1982, Teal 1986, Bertness 1992). The former,growing l-2 meters in height, characterizes the lowmarsh and it is flooded by every tidal cycle. It isreplaced landward by the high marsh S. patens which isflooded periodically by spring high tides. A third belt ofhums gerardii (black grass) forms a border near theupland and is replaced by an upper-most border ofPanicum virgatum (switch grass) and/or Iva frutescens(marsh elder). Where the disturbance occurs along themarsh/upland interface Phragmites may form the typi-cal upper border vegetation. This marsh system isflushed by estuarine waters with salinity ranging from2030 ppt. It exhibits also a distinctive set of animalpopulations some of which are restricted to the lowmarsh whereas others are more typical of the highmarsh (Olmstead and Fell 1974).

Goals o f Restora t ion

The major goal is to restore salt marsh systems losthistorically and recreate the ecological link betweenthe tidal marsh and contiguous estuary so that the highproductivity of the tidal marsh-estuarine system canbe restored. These systems carry on a diversity of func-tional roles or values in terms of finfish and shellfishproductivity and shoreline stabilization (Niering 1985,Mitsch and Gosselink 1993).

The Restoration Process

A diversity of restoration strategies can be employeddepending upon the nature of site degradation. In thecase of tidal restriction, the aim is to restore tidal flush-ing and attempt to recreate a salinity regime similar tothat which previously existed (Rozsa and Orson 1993).With salinities above 20 ppt, Phragmites will be killed orsufficiently suppressed to favor the re-entry of theSpartim grasses (Capotosto and Spencer 1988, Tiner


1995). In other sites the restoration process may involveremoval of spoil or dredged material from a filled saltmarsh which may now be Phrugmites-dominated. Herespoil is removed to expose the original salt marsh peatsurface (Capotosto 1993, Waters 1995) and then thecreek channels are recreated to restore tidal flushing.Marsh elevations are critical in terms of tidal flooding

b to favor the establishment of Spartina alterniflouu. Agentle slope that falls within the range of mean highand mean low tide is required. This is also the pre-requisite for any planting of S. uZternij7oru where shore-line stabilization is desired or where new marsh isbeing created. High-marsh Spurtinu putens can be re-created at slightly higher elevations either undernatural conditions or by planting, but high marsh is notas easy to recreate as low marsh because it is controlledby a complex of environmental factors. Another para-meter in the restoration process is to avoid flooding ofprivate property during major storms. The use andavailability of self-regulating tidal gates invented byThomas Steinke, Director of the Fairfield ConservationCommission, has greatly aided the marsh restorationprocess in highly developed areas (Steinke 1986,1995a,b).

Several case histories will be briefly described to illus-trate the feasibility of salt marsh restoration in variousecological settings (Rosza and Orson 1993). Such effortsalso serve as invaluable models where one can actuallyobserve firsthand results of ecological restoration.

The Hammock River (Connecticut) valley marsh onLong Island Sound represents a 250-acre system inwhich the upper reaches were restricted by tidal gatesearly in the century, transforming the typical Spurtinu-dominated wetland to a monoculture of Phragmites.To reverse this trend, one of the four tide gates wasopened in the summer of 1985. Now more than a de-cade later Phrugmites has been dramatically suppressed

. and the area is now dominated by Spurtinu and othersalt marsh vegetation.

.Another successful marsh restoration project is at

the Barn Island Wildlife Management Area (Connecti-cut) on the Connecticut/Rhode Island border. Here avalley salt marsh was severely restricted in the late1940s with only an 0.5-m opening connecting the adja-cent estuary (Miller and Egler 1950). The salt marsh wastransformed from a Spartinu-dominated vegetation to aTyphu ungustifloliu (narrow-leaved cattail) marsh withareas of Phrugmites restricted to the marsh borders. In1978 a five-foot culvert was installed, followed by an-

other seven foot shortly thereafter in order to increasetidal flushing. Salinity of the adjacent estuary wasrestored (28-33 ppt) to the impounded area. By the mid1980s most of the cattail was dead and Spurtinu grassesdominated the area (Sinicrope et al. 1990, Barrett andNiering 1993). With the restoration of the plant com-munity, typical salt marsh animal populations havealso become established (Fell et al. 1991). After morethan a decade, functional equivalence is beingrestored, not unlike the productivity and trophicstructure in the nearby reference system.

Marsh restoration by planting has also been docu-mented in the Northeast but especially in Southeast(North Carolina) where extensive salt marsh areashave been created on dredged material (Broome et al.1986, Broome et al. 1988, Broome 1990). In theNortheast, this has been done on a less extensive scalewithin the intertidal low marsh zone especially to favorshoreline stabilization (Garbisch and Garbisch 1994).Recreating the elevation typical of the low marsh sothat the site is flooded by every tidal cycle is, aspreviously mentioned, the major prerequisite. PlantingS. ulternifloru 20 in apart and protecting it from grazerssuch as geese until well established is critical. S.uZternifloru plantings following an oil spill have resultedin successful restoration (Bergen et al. 1995).

The ultimate goal of salt marsh restoration is to create aself perpetuating ecosystem with its productivity,biogeochemical cycling, and food chain supportcomparable to the reference or control marsh system.Much literature now exists on this subject from theSoutheast, where marsh restoration and creation havebeen underway since the 1970s (Craft et al. 1988, Saccoet al. 1994, Thompson et al. 1995). Literature is alsoavailable from the Northeast (Fell et al. 1991, Allen et al.1994, Peck et al. 1994, Spelke et al. 1995). It has beenshown that various aspects of functional equivalencecan be attained within a decade or less dependingupon the region. For some functions several decadesmay be required. Thus, monitoring a minimum of fiveyears upon the completion of the project is basic. Thisallows time for management correction and assessingthe development of functional equivalence.

A further requirement is the establishment of abuffer zone along the upland to accommodate conti-nued sea level rise. Two studies in the Northeast havedocumented the potential effects of sea-level rise onthe marsh vegetation toward a more hydric/less pro-


ductive phase (Warren and Niering 1993, Nydick et al.1995). The current rate of sea level rise is one inch peryear and this rate is predicted to increase in the futurewith climate warming (Warrick 1993). Therefore, anundeveloped buffer of 50-100 ft or more is needed toprovide for the landward movement of the marsh withsea level rise and overall protection of the restoredsystem.

Finally, it should be noted that the potential forrestoration should not substitute for permitting saltmarsh destruction because of development. Allpossible and prudent alternatives should be exploredprior to sacrificing a self-perpetuating functional tidalsalt marsh ecosystem (Oviatt et al. 1977, Kusler andKentula 1989, Moy and Levin 1991).

Far too many ecological restoration projects have beenstarted without clear definition of restoration goals andwith little attempt to evaluate the success quantita-tively. Bradshaw (1993) argued strongly that ecologicalrestoration should follow a scientific approach: (1) beaware of other relevant work, (2) carry out experimentsto test ideas, (3) monitor key indicator parameters, (4)design further experiments and tests based on resultsof monitoring, and (5) publish peer reviewed resultsand conclusions. Kaufmann et al. (1994) discussed theimportance of a systematic approach in ecosystemmanagement including the determination of referenceconditions, determination of current conditions, andusing coarse- and fine-filter analyses to determine ifgoals are met. Oliver et al. (1994) suggested an eight-step systematic approach for achieving forest healthwhich is relevant to ecological restoration. Walters(1987) emphasized the importance of clearly statingassumptions about ecosystem behavior, the building ofexplicit models that synthesize this knowledge, andtesting this knowledge in adaptive learning experi-ments.

Based on these ideas and other sources, we havedeveloped a stepwise systems analytic approach to thedesign of ecosystem restoration experiments (Table 1).Most steps are straightforward and broadly discussedin the adaptive ecosystem management literature. Ofthese steps, step 3, determining reference conditions ismost explicit to ecological restoration.

Adaptive ecosystem restoration and managementinvolves a broad variety of practices for designing andtesting ecological restoration treatments. A systematicdiagnosis of the ecosystem pathology is an essential

Table 1. A systems analytic approach to adaptive ecosystemrestoration.

1.

2 .

3 .

4 .

5 .

6 .

7 .

8 .

Clearly diagnose the symptoms and causes of theecosystem health problem. What are the symptomsthat suggest the ecological system has been degradedand what are the underlying mechanisms (Rapport1995, Covington et al. 1994)?

Determine reference conditions. What was thecondition of the ecosystem before degradation(Kaufmann et al. 1994, Morgan et al. 1994)?

Set measurable ecological restoration goals (NationalResearch Council 1992). How close to referenceconditions do you intend to get? How will you know ifyou are moving in the right direction?

What factors are most limiting to the restorationprocess?

Develop alternative ecosystem restoration hypotheses(Walters and Holling 1990).

Design restoration treatments that will allow you totest the alternative hypotheses (Bradshaw 1993).

Monitor ecosystem conditions and evaluatehypotheses .

Feed the results back into the design andimplementation of ecological restoration treatments -adapting management based on results and changinggoals (Kaufmann et al. 1994).

ecological techniques, retrospective ecological analysis,and dendrochronology, along with other techniques(see Kaufmann et al. 1994, Morgan et al. 1994), are usedto determine the natural structure and function of theecological system to be restored. Goals and perform-ance measures must be defined in measurable terms.Assumptions about ecosystem dysfunction must bestated. A specific set of scientifically-based alternativetreatments for restoring ecosystems to the desiredcondition must be developed. Finally, monitoring andevaluation procedures are used to determine wherethe restoration worked and where it did not. A centralassessment is whether the ecosystem being restoredhas been set on a trajectory such that structural andfunctional equivalency to the reference system will beattained. This information is then fed back into thebody of scientific and managerial knowledge for futureecosystem management decisions.

Philosophically, ecosystem restoration is founded onLeopold’s belief that a workable ecosystem conser-


symbiosis with land, economic, public, and private;” as that can accelerate recovery of the ecosystem. Close“a protest against destructive land use;” as an effort “to attention must be given to restoration of both structurepreserve both utility and beauty;” as “a state of har- and processes, including natural disturbance regimes,mony between men and land;” and finally, as “a if restoration is to be successful. These managementpositive exercise of both skill and insight, not merely a actions can be viewed as working hypotheses to benegative exercise of abstinence and caution.” tested in closely monitored management experiments.

Given the wide variety of human needs and goalsfor federal lands and waters, it seems unlikely that vastareas will be restored to completely natural conditions.In fact, keeping some areas in a somewhat artificialstate may be desirable so long as such action does notimpair their long-term sustainability. Areas dedicatedto wood fiber production, livestock grazing, intensiverecreation, and many other human habitat uses mightfall into this category.

For ecological restoration to serve as a viable ap-proach to implementation of ecosystem management,it must not be viewed as a rigid set of procedures nor asa simple recipe. Rather, it must be viewed as a broadintellectual framework for meeting ecological habitatneeds for all organisms, including those of humans, bymanaging in harmony with the natural ecosystemprocesses and components characteristic of the recentevolutionary environment of the biota.

Setting degraded ecosystems on the path to re-covery of natural structure and function seems broadlywarranted. Ecological restoration, even though partial,can go a long way toward reducing the undesirablesymptoms of dysfunctional ecosystems and benefit notonly future generations but also those involved in therestoration process. In fact, some have suggested thatthe transformation of those involved in ecologicalrestoration is one of the major benefits to be gainedfrom such projects (Dodge 1991, Jordan 1993).

Agee, J.K. 1991. Fire history of Douglas-fir forests in the PacificNorthwest. pp. 25-33. In Wildlife and vegetation of un-managed Douglas-fir forests. USDA Forest Service GeneralTechnical Report PNW-GTR-285. Portland, OR: PacificNorthwest Research Station.

Agee, J.K., and R.L. Edmonds. 1992. Forest protection guide-lines for the northern spotted owl. In: Final draft recoveryplan for the northern spotted owl. Appendix E. U.S. De-partment of Interior, Washington, DC.

Allen, E.A., P.E. Fell, M.A. Peck, J.A. Gieg, CR. Guthke, andM.D. Newkirk. 1994. Gut contents of common mummi-chogs, Fundulus heteroclitus L., in a restored impoundedmarsh and in natural reference marshes. Estuaries 17: 462,471.

Ecological restoration has as its goal the restoration ofdegraded ecosystems to resemble, or emulate moreclosely conditions that prevailed before disruption ofnatural structures and processes. A key concept inrestoration ecology is that of the reference conditionsdefined as the range of ecosystem conditions (inclu-ding structure and function) which have prevailedover recent evolutionary time. Underlying the idea ofreference conditions is the concept of the evolutionaryenvironment -the environment in which species haveevolved. Ecological restoration consists of manage-ment actions designed to accelerate recovery bycomplementing or reinforcing natural processes.

Key ecological restoration concepts and principleshave been discussed using examples from ponderosapine ecosystems of the Southwest, forest ecosystems ofthe Pacific Northwest, and salt marsh ecosystems of theNortheast. A wide variety of methods can be used todetermine reference conditions, ranging from histori-cal documentation to paleoecological reconstructionsand to locating contemporary ecosystems which canserve as analogs for what the degraded ecosystemmight have been like had it not been disturbed.

Once reference conditions have been determined,then sets of management actions must be identified

American Forest and Paper Association. 1993. Sustainable for-estry principles and implementation guidelines. Forest Re-sources Board, American Forest and Paper Association,Washington, DC.

Anderson, H.E. 1982. Aids to determining fuel models for esti-mating fire behavior. USDA Forest Service General Tech-nical Report INT-122. Intermountain Forest and RangeResearch Station, Ogden, UT.

Anderson, J.E. 1991. A conceptual framework for evaluatingand quantifying naturalness. Conservation Biology 5: 347-351.

Anon. 1961. Connecticut’s coastal marshes: a vanishing re-source. The Connecticut Arboretum Bulletin, no. 12. Connect-icut College, New London, CT.

Anon. 1995. Report of the Interagency Task Force on Ecosys-tem Management. U.S. Government Printing Office,Washington, DC.

Amo, SF., J.H. Scott, and M.G. Hartwell. 1995. Age-classstructure of old growth ponderosa pine/Douglas-firstands and its relationship to fire history. USDA ForestService Research Paper INT-RP-481. Intermountain Forestand Range Experiment Station, Ogden, UT.

Aronson, J., and E. Le Floc’h. 1996. Hierarchies and landscapehistory: dialoguing with Hobbs and Norton. RestorationEcology 4: 327-333.


Barrett, N.E., and W.A. Niering. 1993. Tidal marsh restoration:trends in vegetation change using a Geographical Infor-mation System (GIS). Restoration Ecology 1: 18-28.

Bergen, A., M. Levandowsky, T. Gorrell, and C. Alderson.1995. Restoration of a heavily oiled salt marsh usingSpartina ulternifloru seedlings and transplants: effects onpetroleum hydrocarbon levels and soil microflora. Pres-ented to The 20th Annual Conference on Contaminated Soils;Analysis, Site Assessment, Fate, Environmental and HumanRisk Assessment, Remediution, and Regulation. October 1995.University of Massachusetts.

Berger, J.J. 1997. Nine ways to save our forests. Sierra 82(4): 38-39.

Bertness, M.D. 1992. The ecology of a New England saltmarsh. American Scientist 80: X0-268.

Booth, D.E. 1991. Estimating prelogging old-growth in the Pa-cific Northwest. journal of Forestry 89: 25-29.

Bradshaw, A.D. 1993. Restoration ecology as science. Restoru-tion Ecology 1: 71-73.

Bradshaw, A.D. 1984. Ecological principles and land reclama-tion practice. Landscape Planning 11: 3548.

Broome, S.W. 1990. Creation and restoration of tidal wetlandsof the southeastern United States. pp. 37-72. In: J.E. Kuslerand M.E. Kentula (eds.), Wetland Creation and Restoration:the Status of the Science. Island Press, Washington, DC.

Broome, SW., C.B. Craft, and E.D. Seneca. 1988. Creation anddevelopment of brackish-water marsh habitat. pp. 197-205. In: J. Zelazny and J.S. Feierabend (eds.), Increasing OurWetland Resources. Proceedings of the Corporate Conser-vation Council of the National Wildlife Federation, Wash-ington, DC., 4-7 October 1987. National WildlifeFederation, Washington, DC.

Broome, S.W., E.D. Seneca, and W.W. Woodhouse, Jr. 1986.Long-term growth and development of transplants of thesalt-marsh grass Spurtinu ulternifloru. Estuaries 9: 63-74.

Brubaker, L.B. 1991. Climate change and the origin of oldgrowth Douglas-fir forests in the Puget Sound Lowlands.pp.17-24. In: Wildlife and vegetation of unmanagedDouglas-fir forests. USDA General Technical ReportPNW-GTR-285. Portland, OR: Pacific Northwest ResearchStation.

Capotosto, P. M. 1993. Restoration of a dredge disposal site inMumford Cove, Groton, Connecticut. State of Connecticut.Department of Environmental Protection, Field Servicesand Boating Division, Hartford.

Capotosto, P.M., and E. Spencer. 1988. Controlling Phrugmitesby tidegute management in the State of Connecticut. State ofConnecticut, Department of Health Services, Madison,CT.

Christensen, N.L., et al. 1995. The Scientific Basis for EcosystemManagement. Committee on Ecosystem Management, Eco-logical Society of America.

Clark, L.R., and R.N. Sampson. 1995. Forest Ecosystem Health inthe Inland West: A Science and Policy Reader. Forest PolicyCenter, American Forests, Washington, DC.

Cooper, CF. 1960. Changes in vegetation, structure, andgrowth of southwestern pine forests since white settle-ment. Ecology 42: 493499.

Covington, W.W. 1996. Implementing adaptive ecosystemrestoration and management in long-needled pine forests.pp. 4347. In: Conference on adaptive ecosystem restora-tion and management. USDA Forest Service General

Technical Report RM-GTR-278. Rocky Mountain Forestand Range Experiment Station, Fort Collins, CO.

Covington, W.W., R.L. Everett, R.W. Steele, L.I. Irwin, T.A.Daer, and A.N.D. Auclair. 1994. Historical and anticipatedchanges in forest ecosystems of the Inland West of theUnited States. Sustainable Forestry 2: 13-63.

Covington, W.W., and M.M. Moore. 1994a. Southwestern pon-derosa forest structure and resource conditions: changessince Euro-American settlement. Forestry 92(l): 39-47.

Covington, W.W., and M.M. Moore. 1994b. Postsettlementchanges in natural fire regimes: ecological restoration ofold-growth ponderosa pine forests, Sustainable Forestry 2:153-181.

Covington, W.W., and S.S. Sackett. 1984. The effect of a pre-scribed burn in southwestern ponderosa pine on organicmatter and nutrients in woody debris and forest floor. For-est Science 30: 183-192.

Covington, W.W., and S.S. Sackett. 1990. Fire effects on pon-derosa pine soils and their management implications. pp.105-111. In: Effects of fire management of southwesternnatural resources. USDA Forest Service General TechnicalReport RM-191. Rocky Mountain Forest and Range Exper-iment Station, Fort Collins, CO.

Covington, W.W., and S.S. Sackett. 1992. Spatial variation insoil mineral nitrogen following prescribed burning in pon-derosa pine. Forest Ecology and Management 54: 175-191.

Covington, W.W., and R.N. Sampson. 1996. Restoration offorest ecosystem health: western long-needled pines.Transactions of the 6Zst North American Wildlife and NaturalResources Conference. pp. 511-518.

Covington, W.W., P.Z. Fule, M.M. Moore, SC. Hart,T.E. Kolb,J.N. Mast, S.S. Sackett, and M.R. Wagner. 1997. Restoringecosystem health in ponderosa pine forests of the South-west. Journal of Forestry 95(4): 23-29.

Craft, C.B., Broome, S.W., and E.D. Seneca. 1988. Nitrogen,phosphorus and organic carbon pools in natural andtransplanted marsh soils. Estuaries 11: 272-280.

Dana, S.T., and SK. Fairfax. 1980. Forest and Range Policy.McGraw-Hill , New York.

Dieterich, J.H. 1980. Chimney Spring forest fire history. USDAForest Service Research Paper RM-220. Rocky MountainForest and Range Experiment Station, Fort Collins, CO.

Dobson, AI’., A.D. Bradshaw, and A.J.M. Baker. Hopes for thefuture: restoration ecology and conservation biology. Sci-ence 277: 515-522.

Dodge, J. 1991. Life work. pp. 24. In: R.L. N&en (ed.). HelpingNature Heal. Ten Speed Press, Berkeley, CA.

Fell, P.E., K.A. Murphy, Peck, M.A. and M.L. Recchia. 1991.Re-establishment of Melumpus bidentutus (Say) and othermacroinvertebrates on a restored impounded tidal marsh:comparison of populations above and below the im-poundment dike. Journal of Experimental Marine BiologyEcology 152: 33-48.

Flader, S.L. 1974. Thinking like a mountain: Aldo Leopold and theevolution of an ecological attitude toward deer, wolves, and for-ests. University of Missouri Press, Columbia, MO.

Flader, S.L., and J.B. Callicott (eds.). 1991. The River of theMother of God and other essays by Aldo Leopold. University ofWisconsin Press, Madison, WI.

Franklin, J.F., K. Cromack, Jr., W. Denison, A. McKee, C. Ma-ser, J. Sedell, F. Swanson, and G. Juday. 1981. Ecologicalcharacteristics of old-growth Douglas-fir forests. USDA


Forest Service General Technical Report PNW-118. PacificNorthwest Forest and Range Experiment Station, Port-land, OR.

Franklin, J.F., and CT. Dyrness. 1973. Natural vegetation ofOregon and Washington. USDA Forest Service GeneralTechnical Report PNW-8. Pacific Northwest Forest andRange Experiment Station, Portland, OR.

MacMahon, J.A., and W.R. Jordan, III. 1994. Ecological resto-ration. pp. 409438. In: G.K. Meffe and CR. Carroll (eds.),Principles of Conservation Biology. Sinauer Associates,Sunderland, MA.

iGarbisch, E.W., and J.L. Garbisch. 1994. Control of upland

bank erosion through tidal marsh construction on re-stored shores: Application in the Maryland portion ofChesapeake Bay. Environmental Management 18: 677-691.

. Geist, V. 1978. Life Strategies, Human Evolution, and Environ-mental Design: Toward a Biological Theory ofHealth. SpringerVerlag, New York.

Madany, M.H., and N.E. West. 1983. Livestock grazing-fireregime interactions within montane forests of Zion Na-tional Park. Ecology 64: 661-667.

Maser, C., B.R. Mate, J.F. Franklin, and C.T. Dyrness. 1981.Natural History of Oregon Coast Mammals. USDA ForestService General Technical Report PNW-133. Pacific North-west Forest and Range Experiment Station, Portland, OR.

Mayr, E. 1942. Systematics and the Origin of Species. ColumbiaUniversity Press, New Yorki.

Hammond, M., and A. Holland. 1995. Ecosystem health:Some prognostications. Environmental Values 4: 283-286.

Harrington, M.G., and S.S. Sackett. 1990. Using fire as a man-agement tool in southwestern ponderosa pine. pp. 122-133. In: Effects of Fire Management of Southwestern Natu-ral Resources, USDA Forest Service General Technical Re-port RM-191. Rocky Mountain Forest and RangeExperiment Station, Fort Collins, CO.

Harris, L.D. 1984. The Fragmented Forest: Island BiogeographyTheory and the Preservation of Biotic Diversity. University ofChicago Press, Chicago, IL.

Haslam, S.M. 1973. Some aspects of the life history andautecology of Phragmites communis Trin.: A review. Pol.Arch. Hydrobiol. 20: 79-100.

Miller, W.R. and F.E. Egler. 1950. Vegetation of theWequetequock-Pawcatuck tidal marshes, Connecticut.Ecological Monographs 20: 143-172.

Mitsch, W.J., and GosseIink, J.G. 1993. Wetlands. 2nd edition.Van Nostrand Reinhold, New York.

Mooney, H.A. 1981. Adaptations of plants to fire regimes: in-tegrating summary. pp. 322-323. In: Fire Regimes and Eco-system Properties. USDA Forest Service General TechnicalReport WO-26. Washington, DC.

Morgan, I’., G.H. Aplet, J.B. Haufler, H.C. Humphries, M.M.Moore, and W.D. Wilson. 1994. Historical range of vari-ability: A useful tool for evaluating ecological change. Sus-tainable Forestry 2: 87-111.

Higgs, E. 1997. What is good ecological restoration. Conserva-tion Biology 11: 338-348.

Jackson, L.L., N. Lopoukhine, and D. HiIIyard. 1995. Ecologi-cal restoration: A definition and comments. RestorationEcology 3: 71-75.

Moy, L.D., and Levin, L.A. 1991. Are Spartina marshes a re-placeable resource? A functional approach to evaluation ofmarsh creation efforts. Estuaries 14: 1-16.

Myers, N. 1984. Genetic resources in jeopardy. Ambio 13: 171-174.

Jordan, W.R., III. 1993. Rituals of restoration. The Humanist,November/December: 23-26.

Jordan, W.R., III, M.E. Gilpin, and J.D. Aber. 1987. RestorationEcology: A Synthetic Approach to Ecological Restoration. Cam-bridge University Press, New York.

Karr, J.R. 1981. Assessment of biotic integrity using fish com-munities. Fisheries 6(6): 21-27.

National Academy of Science. 1974. Rehabilitation Potential ofWestern Coal Lands. BalIinger, Cambridge, MA.

National Research Council, Committee on Restoration ofAquatic Ecosystems. 1992. Restoration of Aquatic Ecosystems:Science, Technology, and Public Policy. National AcademyPress.

Kaufmann, M.R., et al. 1994. An ecological basis for ecosystemmanagement. USDA Forest Service General Technical Re-port RM-246. Rocky Mountain Forest and Range Experi-ment Station, Fort Collins, CO.

Niering, W.A. 1985. Wetlands. The Audubon Series NatureGuides. Alfred A. Knopf, New York.

Niering, W.A., and Warren, R.S. 1980. Vegetation patternsand processes in New England salt marshes. Bioscience 30:301-307.

Kay, C.E. 1995. Aboriginal overkill and native burning: impli-cations for modern ecosystem management. Western Jour-nal of Applied Forestry 10: 121-126.

KiIgore, B.M. 1981. Fire in ecosystem distribution and struc-* ture: Western forests and scrublands. pp. 58-89. In: Fire Re-

gimes and Ecosystem Properties. USDA Forest ServiceGeneral Technical Report WO-26. Washington, DC.

N&en, R. (ed.). 1991. Helping Nature Heal: An Introduction toEnvironmental Sestoration. Ten Speed Press, Berkeley, CA.

Nixon, S.W. 1982. The ecology of New England high saltmarshes: A community profile. FWS/OBS-81/55. U.S. Fishand Wildlife Service, Office of Biological Services, Wash-ington, DC.

. Kolb, T.E., M.R. Wagner, and W.W. Covington. 1994. Con-cepts of forest health. Forestry 92: 10-15.

Kusler, J.A., and Kentula, M.E. 1989. Wetland creation andrestoration: The status of the science. Vol. 2: Regional re-views. EPA/600/3-89/038. U.S. Environmental ProtectionAgency, Washington, DC.

Noss, R.F. 1991. From endangered species to a biodiversity.pp. 227-245. In: K. Kohm (ed.), Balancing on the Brink of Ex-tinction: The Endangered Species Act and Lessons for the Fu-ture. Island Press, Washington, DC.

Nydick, K.R., A.B. BidweII, E. Thomas, and J.C. Van Kamp.1995. A sea-level rise curve from GuiIford, Connecticut,USA. Marine Geology 124: 13-159.

Oliver, CD., and B.C. Larson. 1990. Forest Stand Dynamics.McGraw-Hill , New York.

Lubchenco, J., et al. 1991. The sustainable biosphere initiative: Oliver, CD., D.E. Ferguson, A.E. Harvey, H.S. Malany, J.M.An ecological research agenda. Ecology 72: 371412. Mandzak, and R.W. Mutch. 1994. Managing ecosystems

MacArthur, R.H. 1972. Geographical Ecology: Patterns in the Dis- for forest health: an approach and effects on uses and val-tribution of Species. Princeton University Press, Princeton, ues. Sustainable Forestry 2: 113-133.NJ. Olmstead, N.C., and P.E. Fell. 1974. Tidal marsh invertebrates of

616 W. Covington et aLlEcosystem Restoration and Management

Connecticut. Connecticut Arboretum Bulletin, no. 20. Con-necticut College, New London, CT.

Oviatt, CA., S.W. Nixon, and J. Garber, 1977. Variation andevaluation of coastal salt marshes. Environmental Manage-ment 1: 201-211.

Parsons, D.J., D.M. Graber, J.K. Agee, and J.W. vanWagtendonk. 1986. Natural fire management in nationalparks. Environmental Management 19(l): 21-24.

Pearson. G.A. 1950. Management of Ponderosa Pine in the South-west. USDA Monograph 6.

Peck, M.A., P.E. Fell, E.A. Allen, J.A. Gieg, C.R. Guthke, andM.D. Newkirk. 1994. Evaluation of tidal marsh restoration:Comparison of selected macroinvertebrate populations ona restored impounded valley marsh and an unimpoundedvalley marsh within the same salt marsh system in Con-necticut, USA. Environmental Management 18: 283-293.

Rapport, D.J. 1995. Ecosystem health: an emerging integrativescience. In: D.J. Rapport, C.L. Gaudet, and I’. Calow (eds.),Evaluating and Monitoring the Health of Large-Scale Ecosys-tems. NATO AS1 Series. Series I: Global environmentalchange, Vol. 28. Springer-Verlag, Berlin.

Rapport, D.J., and S.B. Yazvenko. 1996. Ecosystem distresssyndrome in ponderosa pine forests. pp. 3-9. In: Confer-ence on Adaptive ecosystem restoration and manage-ment: Restoration of cordilleran conifer landscapes ofNorth America. USDA Forest Service General TechnicalReport RM-GTR-278. Rocky Mountain Forest and RangeExperiment Station, Fort Collins, CO.

Ripple, W.J. 1994. Historic spatial patterns of old forests inwestern Oregon. Forestry 4549.

Rosza, R., and R.A. Orson. 1993. Restoration of Degraded SaltMarshes in Connecticut. pp. 196-205. In: Frederick J.Webb, Jr. (ed.), Proceedings of the Twentieth Annual Confer-ence on Wetlands Restoration and Creation, May 2993. Insti-tute of Florida Studies, Hillsborough Community College,Plant City, FL.

Sacco, J.N., E.D. Seneca, and T.R. Wentworth. 1994. Infaunalcommunity development of artificially established saltmarshes in North Carolina. Estuaries 17: 489-500.

Sackett, S.S., S.M. Haase, and M.G. Harrington. 1996. Lessonslearned from fire use for restoring southwestern ponder-osa pine ecosystems. pp. 54-61. In: Conference on adap-tive ecosystem restoration and management. USDA ForestService General Technical Report RM-GTR-278. RockyMountain Forest and Range Experiment Station, Fort Col-lins, co.

Sinicrope, T.L., P.G. Hine, R.S. Warren, and W.A. Niering.1990. Restoration of an impounded salt marsh in NewEngland. Estuaries 13: 25-30.

Smith, Maynard. 1958. Theory of Evolution. Penguin Books,New York.

Society for Ecological Restoration. 1993. Ecological restora-tion, conservation and environmentalism: Mission state-ment and environmental policies of the Society forEcological Restoration. Restoration Ecology 1: 206-207.

Society of American Foresters. 1993. Sustaining Long-Term For-est Health and Productivity. Society of American Foresters,Bethesda, MD.

Southwest Forest Alliance. 1996. Forests forever!A plan to restoreecological andeconomic integrity to the Southwest’s nationalfor-

Spelke, J.A., P.E. Fell, and L.L. Helvenston. 1995. Populationstructure, growth and fecundity of Melampus bidentatus(Say) from two regions of a tidal marsh complex in Con-necticut. The Nautilus 108: 42-47.

Steinke, T.J. 1986. Hydrologic Manipulation and RestoringWetland Values: Pine Creek, Fairfield Connecticut. pp.377-383. In: J.A. Kusler, M. L. Quammen, and G. Brooks(eds.), National Wetland Symposium; Mitigation of impactsand Losses. Association of State Wetland Managers, NewOrleans, LA.

Steinke, T.J. 1995a. The self-regulating tidegate. Atlantic Water-fowl Council, Delaware Department of Natural Resourcesand Environmental Control, Division of Fish and Wildlife,Dover, DE.

Steinke, T.J. 199513. Restoration of degraded salt marshes in PineCreek, Fairfield, Connecticut. Atlantic Waterfowl Council,Delaware Department of Natural Resources and Environ-mental Control, Division of Fish and Wildlife, Dover, DE.

Swetnam, T.W., and C.H. Baisan. 1996. Historical fire regimepatterns in the southwestern United States since AD 1700.In: C.D. Allen (ed.), Proceedings of the Second La Mesa FireSymposium, March 29-30, 1994, Los Alamos, NM. NationalPark Service, Washington, DC.

Tappeiner, J., et al. 1992. Managing stands for northern spot-ted owl habitat. In Final draft recovery plan for the northernspotted owl. Appendix K. U.S. Department of the Interior,Washington, DC.

Taylor, R. 1996. Mt. Trumbull ecosystem restoration project.pp. 74-78. In: Conference on adaptive ecosystem restora-tion and management. USDA Forest Service GeneralTechnical Report RM-GTR-278. Rocky Mountain Forestand Range Experiment Station, Fort Collins, CO.

Teal, J.M. 1986. The ecology of regularly flooded salt marshesof New England: A community profile. Biological Report85(7.4). U.S. Fish and Wildlife Service, Washington, DC.

Thompson, S.P., H.W. Paerl, and M.C. Go. 1995. Seasonalpat-terns of denitrification in a natural and a restored saltmarsh. Estuaries, in press.

Tiner, R. 1995. Phragrnites -Controlling the all-too-commonCommon Reed. Massachusetts Wetlands RestorationTechnical Notes. Wetlands Restoration and Banking Pro-gram, Technical Note Number 1: l-3.

Tiner, R.W. 1984. Wetlands of the United States: Current Statusand Recent Trends. National Wetlands Inventory, Fish andWildlife Service, U.S. Department of the Interior, Wash-ington, DC.

United Nations. 1992. Adoption of agreements on environ-ment and development: The Rio Declaration on Environ-ment and Development. UNEP.A/CONF.151/5/Rev.l. 13June 1992.

U.S. Department of Agriculture and U.S. Department of Inte-rior. 1994a. Final supplemental environmental impactstatement on management of habitat for late-successionaland old-growth forest related species within the range ofthe northern spotted owl. Portland, OR. 2 vol.

U.S. Department of Agriculture and U.S. Department of Inte-rior. 1994b. Record of decision for amendments to ForestService and Bureau of Land Management planning docu-ments within the range of the northern spotted owl. Port-land, OR.


Protection Agency. 1993. Forest Ecosystem Management: AnEcological, Economic, and Social Assessment. Portland, OR.

Walters, C. 1987. Adaptive Management of Renewable Resources.Macmillan, New York.

Walters, C.J., and C.S. Holling. 1990. Large-scale managementexperiments and learning by doing. Ecology 71: 2060-2068.

Warren, R.S., and Niering, W.A. 1993. Vegetation change on aNortheast tidal marsh; interaction of sea-level rise and

I) marsh accretion. Ecology 74: 96-103.Warrick, R.A. 1993. Climate and sea level change: A synthesis.

I pp. 2-24. In: R.A. Warrick, E.M. Barrow, andT.M.L. Wigley(eds.), Climate and Sea Level Changes Observations Projections

4 and Implications. Cambridge University Press.Waters, R.F. 1995. Tidal wetland restoration on a dredge dis-

posal site: Patterns of initial plant establishment, MumfordCove, Groton, Connecticut. M.A. Thesis. Connecticut Col-lege, New London, CT.

Weaver, H. 1951. Fire as an ecological force in the southwest-em pine forests. Journal of Forestry 57: 12-20.

Weaver, H. 1974. Effects of fire on temperate forests: WesternUnited States. pp. 279-319. In: T.T. Kozlowski and C.E.Ahlgren (eds.), Fire and Ecosystems. Academic Press, NewYork.

White, AS. 1985. Presettlement regeneration patterns in asouthwestern ponderosa pine stand. Ecology 66(2): 589-594.

Wiley, J. 1989. Phenomena, comment and notes. A new eco-logical society will devote itself to the toughest mission ofall: the restoration of land left for dead. Smithsonian 19(12):32-33.

Williams, J.T., R.G. Schmidt, R.A. Norum, P.N. Omi, and R.G.Lee. 1993. Fire-related considerations and strategies insupport of ecosystem management. USDA Forest ServiceStaffing Paper, Washington, DC.

Wilson, E.O. 1992. The Diversify of Life. W.W. Norton andCompany, New York. 424 p.

ecosystem restoration and management: scientific

Documents