Allocating scarce resources to restore large areas of degraded habitat requires understanding factors limiting ecosystem recovery over a range of temporal and spatial scales. We took advantage of large-scale riparian restoration efforts along the Sacramento River, CA (>2500 ha restored to date) that have focused on planting overstory tree and shrub species to compare the relative importance of within-site factors (e.g. overstory cover, elevation on the floodplain) vs. the surrounding landscape (e.g. percent remnant forest cover) on the cover and species richness of naturally-establishing native, understory species. We surveyed understory species composition in 15 and 35 restored forest sites along a 150-km of the river in 2001 and 2007, respectively. We also planted and seeded seven native riparian species into a factorial experiment to test the effects of overstory cover (with and without), grass control (herbicide or control) and distance from forest edge (<50 m and >250 m) on seedling establishment and survival.
Results/Conclusions
Results of our field survey show that native understory species cover did not increase significantly over time in sites where none was planted (2001: 20.7 ± 4.0, 2007: 32.3 ± 6.4) and was dominated by a single widespread species (Galium aparine). Both field survey and experimental results indicate that native understory cover is higher in sites with more overstory cover which shades out non-native understory grasses and forbs. The effect of landscape factors varies by dispersal guild with wind-dispersed species more strongly influenced by surrounding forest cover, whereas cover of gravity-dispersed species is correlated with distance to the river. Planting experiments showed that several understory species had higher survival when planted under an existing tree canopy. The grass herbicide treatment had a minimal effect on total understory cover as a reduction in non-native grass cover resulted in increased non-native forb cover. Our results support the following management recommendations: 1. native understory species colonize sufficiently slowly that they will need to be introduced to restoration sites, particularly at locations distant from seed sources; 2. most forest species should be introduced later during the successional process once site conditions are more favorable; and 3. establishing an overstory cover is the most effective method to control non-native understory species.