We compare two successional models as guides for restoring native riparian understory species along a 160-km stretch of the Sacramento River in California. In 2001 and 2007, we surveyed cover, frequency, and richness of native and exotic understory species in 15 sites planted (1989)(1990)(1991)(1992)(1993)(1994)(1995)(1996) with overstory species to determine whether native understory species colonized naturally (passive relay floristics model). In 2007, we surveyed 20 additional sites (planted 1997-2003) in 14 of which understory species were planted (initial floristics model) to evaluate whether planting accelerated community recovery. We surveyed 10 remnant forests as references for successional trajectories. Mean cover and frequency of natives changed little over time in sites where they were not planted initially; increases in native cover in a few sites were primarily due to a single common species (Galium aparine). Species composition shifted from light-demanding to shade-adapted species, both exotic and native, in response to a doubling of overstory cover. Sites with high intensity understory plantings had greater cover and frequency of native understory species than unplanted sites, but were still low relative to reference forests. Light-demanding natives (e.g., Artemisia douglasiana, Rubus ursinus, and grasses) established in sites where they were planted; however, a shade-adapted species (Carex barbarae) did not survive well. Our research indicates that the passive relay floristics and the initial floristic composition approaches serve to restore a few common native understory species, but that planting species as site conditions become appropriate (active relay floristics model) will be needed to restore entire native understory communities.

Large-scale ecosystem restoration projects seldom undergo comprehensive evaluation to determine project effectiveness. Consequently, there are missed opportunities for learning and strategy refinement. Before our study, monitoring information from California's middle Sacramento River had not been synthesized, despite restoration having been ongoing since 1989. Our assessment was based on the development and application of 36 quantitative ecological indicators. These indicators were used to characterize the status of terrestrial and floodplain resources (e.g., flora and fauna), channel dynamics (e.g., planform, geomorphology), and the flow regime. Indicators were also associated with specific goal statements of the CALFED Ecosystem Restoration Program. A collective weight of evidence approach was used to assess restoration success. Our synthesis demonstrates good progress in the restoration of riparian habitats,

No abstract