An inadequate basis for precisely predicting the outcome of lotic ecosystem recovery, whether due to unaided natural processes or management techniques or both, exists because: (1) the field of ecology has not yet matured as a rigorous predictive science; (2) the precise sequence of events, including climatic occurrences, affecting the recovery process may be unique events and thus rarely or never repeated; and (3) even when attempts are made to control the recolonization process through introduction of species, etc., the interaction of these species may not follow deterministic models. Although this symposium focuses on lotic ecosystems, such systems are influenced strongly by exports from the surrounding land mass and, under certain circumstances, this may be the overriding influence on the recovery process; therefore, unless the boundary conditions are determined realistically, the recovery process may not follow desirable pathways. Despite the lack of a robust theoretical support base for lotic ecosystem recovery, some remarkable and rapid recoveries have occurred to either a close approximation of the original condition or to a condition ecologically superior to the damaged condition. In some cases, the recovery was due entirely to natural processes and, in others, often followed relatively straightforward management practices. There is evidence indicating that lotic ecosystem restoration is both cost effective and likely to produce satisfying results relatively rapidly. It is both fortunate that this is the case, since society is likely to support such efforts when the results have been extraordinarily successful, and unfortunate since restoration ecology needs a predictive capability. © 1990 Springer-Verlag New York Inc.
