The Science and Values of Restoration EcologyMark A. Davis,1,3and Lawrence B. Slobodkin2It has been 22 years since William Jordan III and theUniversity of Wisconsin Arboretum published the firstissue of Restoration and Management Notes, 16 yearssince the founding of the Society for Ecological Restor-ation (SER), and 10 years since SER published the firstissue of its flagship journal, Restoration Ecology. In thisshort time, restoration ecology has become a leader inNorth American conservation efforts. Believing it isimportant that the field has a strong scientific foundation(Bradshaw 1993), restoration ecologists have emphasizedconcepts such as ‘ecosystem health’ and ‘ecosystem integ-rity’ when articulating restoration goals and frequentlyhave invoked ecological principles when describing andjustifying their objectives (SER 2002). Although ecologyplays a central and essential role in the implementation ofrestoration projects, we believe that defining restorationgoals and objectives is fundamentally a value-based, notscientific, activity.Since its inception, SER has taken the lead in develop-ing and articulating paradigms of restoration. SER’s mostrecent major publication, The SER Primer of EcologicalRestoration (SER 2002), is developed around the notionthat communities and ecosystems are ecological entities. Inthe Primer, the goal of restoration is stated to be ‘‘theprocess of assisting the recovery of an ecosystem that hasbeen degraded, damaged, or destroyed.’’ The Primerdescribes a recovered ecosystem as one that ‘‘containssufficient biotic and abiotic resources to continue its devel-opment without further assistance’’ and for which ‘‘poten-tial threats to the health and integrity of the restoredecosystem have been eliminated.’’ An ecosystem is alsoconsidered restored when it ‘‘apparently functions nor-mally for its ecological stage of development, and signsof dysfunction are absent.’’ The idea that communities andecosystems possess traits such as health and integrity, thatthey exhibit an organic development, that their ‘‘health’’can be injured or harmed and then can be restored throughinformed efforts of ecologists is reminiscent of earlierecological claims of communities and ecosystems as inte-grated entities (Clements & Shelford 1939).Attributes such as ‘‘health’’ and ‘‘integrity’’ can be mean-ingfully applied to entities that have been directly shapedby evolution, such as individual organisms. Organismsnormally have clearly defined boundaries and a myriadof homeostatic mechanisms that maintain those bound-aries while the organism develops, matures, and repro-duces. However, communities and ecosystems are notshaped as entities by evolution.Today, communities are no longer believed to be tightlyorganized systems (Slobodkin 2003). They are believed tolack coherence (Gould 2002) and clear boundaries (Stiling1999). A community or ecosystem does not possess distinctboundaries nor does it have mechanisms that have evolvedto regulate particular processes. Communities do notexhibit any kind of evolutionary imperative, such as repro-duction, as do individual organisms. The terms ‘‘com-munity’’ and ‘‘ecosystem’’ are useful in a practical sensefor referring to species and processes occurring in a par-ticular locale (O’Neill 2001), but this does not mean thatthere actually exists some integrated entity out there calledan ecosystem that grows, lives, reproduces and dies, or canbe injured or healed.If ecological communities and ecosystems lack anyintrinsic evolutionary or ecological purpose, one cannotvalidly invoke any ecological (or evolutionary) rationale toestablish particular restoration goals. As noted by Diamond,‘‘this goal [of restoration ecology] is not itself a self-evidentmandate. It is a choice based on values, and it is only one ofmany possible choices’’ (Diamond 1987). Restorationistshave often tried to justify their goals by presenting themas fulfillling various ecological imperatives, e.g., restoringecosystem health and restoring indigenous environments.However, characterizing communities and ecosystems as‘‘healthy’’ or ‘‘damaged’’ is a value-based, not scientific,assessment (Lackey 2001).Architecture uses mathematics, physics, and engineeringin its efforts to achieve a particular result of aesthetic andsocial value. In an analogous fashion, restorationists mustuse ecology, and often geology, soil science, and more toachieve results of social value. Often, their results are alsoof great beauty as well. Perhaps, ‘‘ecological architecture’’might be a more apt characterization of the work ofecological restoration, because the term acknowledgesthe central roles played by both values and science.Ultimately, it is important that restorationists do theirbest to clearly distinguish between their science and theirvalues in their discussions with the public and policy1Department of Biology, Macalester College, Saint Paul, MN 55105,U.S.A.2Department of Ecology and Evolution, Stony Brook University, StonyBrook, NY 11794, U.S.A.3Address correspondence to Mark A. Davis, email [email protected]Ó 2004 Society for Ecological Restoration InternationalMARCH 2004 Restoration Ecology Vol. 12 No. 1, pp. 13 1makers as well as amongst themselves (Kapustka & Landis1998; Lancaster 2000). Restorationists and their supportersmust make their cases in the same socio- politico arena asany other advocacy group and justify the merits of theirpreferences to the various stakeholders in the same way,using social, cultural, economic, health, and ethical argu-ments. Whether those preferences are for a historicalenvironment, a species-rich environment, a particular setof species, or some other type of landscape, restorationistscannot logically or ethically invoke ecology or evolution asa justification for these preferences. Ecology and evolu-tion, and other scientific disciplines, appropriately comeinto play during the actual implementation of the statedsocial goals. Consider the following proposed definition ofecological restoration:‘‘Ecological Restoration is the process ofrestoring one or more valued processes orattributes of a landscape.’’This definition does not invoke questionable ecologicalconcepts such as ecosystem health and ecosystem develop-ment, and it acknowledges the important role values playin the field. It also permits restorationists to define a widerange of