certain desired restorative actions (e.g., eradication
<br />of exotic species, reintroduction of extinct native
<br />species), and philosophical differences among stake-
<br />holders and disagreements over who should bear the
<br />social and economic costs of restoration. Resolving
<br />resource- management issues across entire river basins
<br />and resolving conflicting interests among stakehold-
<br />ers requires degrees of coordination and cooperation
<br />rarely achieved in human society [Naiman, 1992].
<br />However, as the public increasingly recognizes the
<br />link between ecological integrity and ecosystem goods
<br />or services such as clean water or productive fisheries,
<br />shifts in values may induce people to rethink assump-
<br />tions about what is sociopolitically acceptable in res-
<br />toration scenarios. For example, should reduced flood
<br />flows downstream from a dam constrain restoration
<br />efforts, or should restoration include greater flood -
<br />flow releases from the dam? Many factors assumed to
<br />be constraints twenty years ago are being re- examined
<br />as opportunities to restore rivers today. Rather than a
<br />dichotomy between pro - development and pro- environ-
<br />ment, many scientists and practitioners are realizing
<br />that there is a middle ground in which some functions
<br />can be restored without great cost to water users.
<br />River restoration can also be advanced by treating
<br />restoration projects as experiments that can teach us
<br />about ecosystem operation. Most restoration projects
<br />have been implemented without the study design,
<br />baseline data, and post - project appraisal needed to
<br />learn from them [Downs and Kondolf, 2002; Bern-
<br />hardt et al., 2005]. Much of the published literature,
<br />which forms the basis of our ecological understand-
<br />ing, describes research conducted at space -time scales
<br />much smaller than those appropriate for restoration.
<br />Furthermore, many restorative actions are applied at
<br />scales too small to produce the intended effects on bi-
<br />otic populations and assemblages [Pretty et al., 2003].
<br />A major limitation in advancing scientific knowledge
<br />to guide predictive restoration is the lack of opportuni-
<br />ties to conduct large -scale experiments, where whole
<br />system responses can be evaluated at scales that match
<br />management actions. For example, restoration of flow
<br />regimes below existing water control structures pres-
<br />ents tremendous opportunities to learn about system -
<br />specific responses that can guide future restoration ac-
<br />tions [Poll et al., 2003]. Experimental flood releases
<br />such as those on the Colorado River in Grand Canyon
<br />[Collier et al., 19971 provide opportunities to pose
<br />and test hypotheses regarding the ecosystem effects
<br />of these floods. Despite the lack of standard experi-
<br />mental features such as randomization of controls and
<br />treatments, or replication, the flood releases create
<br />quasi- experiments that provide important knowledge
<br />about river response to restoration efforts [Block et
<br />al., 2001].
<br />Viewing restoration projects as experiments affords
<br />a framework for engaging scientific involvement
<br />early in the process and strengthens the rationale
<br />for monitoring the results of the restoration action.
<br />Adaptive management coupled with effective moni-
<br />toring facilitates learning from experience [Walters,
<br />1997; Rogers, 2003], and has been repeatedly identi-
<br />fied as a critical and missing component of existing
<br />river management programs such as that on the Platte
<br />River [National Research Council, 2005]. We cur-
<br />rently have far too few experiments at appropriate
<br />scales that are conducted adaptively and thus we have
<br />not yet developed scientific guidelines for how best
<br />to restore adaptively or over what timescale adaptive
<br />management should be applied.
<br />In summary, recent overviews of the state of river
<br />restoration in the U.S. have highlighted existing prob-
<br />lems and suggested directions for improvement. We
<br />suggest that river restoration can be most effectively
<br />advanced with increasing emphasis on (i) implement-
<br />ing restoration within a clearly articulated scientific
<br />conceptual framework and a watershed context, (ii)
<br />restoring process rather than form, and (iii) monitor-
<br />ing and learning from ongoing restoration efforts. It
<br />is not unreasonable for society to expect a return on
<br />their investment in river restoration.
<br />References Cited
<br />Angermeier, P.L (1997), Conceptual roles ofbiologi-
<br />cal integrity and diversity, in Watershed restora-
<br />tion: principles and practices, edited by J.E. Wil-
<br />liams, C.A. Wood and M.P. Dombeck, pp. 49 -65,
<br />American Fisheries Society, Bethesda, MD.
<br />Baron, J.S., N.L. Poff, P.L. Angermeier, C.N. Dahm,
<br />P.H. Gleick, N.G. Hairston, R.B. Jackson, C.A.
<br />Johnston, B.G. Richter, and A.D. Steinman
<br />(2002), Meeting ecological and societal needs for
<br />freshwater, Ecological Applications, 12, 1247-
<br />1260.
<br />Bernhardt, E.S., M.A. Palmer, J.D. Allan, and the
<br />National River Restoration Science Synthesis
<br />Working Group (2005), Restoration of U.S. riv-
<br />ers: a national synthesis, Science, 308.
<br />Block, W.M., A.B. Franklin, J.P. Ward, J.L. Ganey,
<br />
|