Monitoring Stream and Watershed Restoration

Chapter 2: Steps for Designing a Monitoring and Evaluation Program for Aquatic Restoration

Philip Roni, Martin C. Liermann, Chris Jordan, and E. Ashley Steel


Stream enhancement and watershed restoration methods, such as replanting or fencing of riparian areas, road removal and reduction of sediment, placement of instream structures, and reconnecting isolated channels and sloughs, have been used in North American watersheds for more than 70 years (Tarzwell 1934; White 2002). Despite the large financial investment in aquatic restoration in recent decades, monitoring and research to evaluate project effectiveness occurs infrequently and often is inadequate to quantify biological response (Reeves and Roelefs 1982; Reeves et al. 1991; Beschta et al. 1994; Chapman 1996; Roni et al. 2002). While published evaluations of restoration techniques are scattered throughout the scientific literature, their results often are inconclusive, statistically insignificant, and highly variable (NRC 1992; Minns et al. 1996; Roni et al. 2003). Even fundamental minimum requirements of experimental design (e.g., replication, controls) are rarely met in habitat restoration projects (Minns et al. 1996). Objectives for many restoration projects and associated monitoring often are poorly stated or inappropriate. Moreover, because most evaluations are reachscale, site-specific case studies, the results are not broadly applicable.

Few rigorous quantitative evaluations exist, in part, because little information has been developed on appropriate restoration monitoring designs. Concurrent evaluations of multiple restoration activities to assess watershed-scale habitat effects on biota, especially fish populations, are extremely rare. Reeves et al. (1997), Slaney et al. (1994), and Solazzi et al. (2000) represent a few of the studies that have examined both physical and biological responses at a broad watershed scale. Although these types of studies can be expensive, they assess the cumulative effect of restoration on watershed-scale recovery of both habitat and populations and, thus, are essential for restoration planning, evaluating, and developing a predictive understanding of restoration effectiveness.

Similar to assessment of the effects of various other anthropogenic activities on aquatic systems (impact assessment), studies evaluating stream, watershed, and estuarine restoration often are hindered by the lack of spatial replication, inadequate pre- and postproject monitoring (temporal replication), and the presence of confounding effects due to a variety of uncontrollable environmental factors (Hurlbert 1984; Walters et al. 1988; Minns et al. 1996; Downes et al. 2002). Guidance for setting up a monitoring program and for determining the appropriate statistical designs for monitoring physical and biological changes at site, reach, watershed, and ecosystem levels is needed. Effective monitoring requires an understanding of the temporal and the spatial scales, the nature of both the restoration actions and the response, and historic and current conditions (Reeves et al. 1995; Downes et al. 2002). Appropriate monitoring and evaluation programs for stream restoration will differ by project type, as well as by region, geomorphology, scale, and a host of other factors.