The relationship between plant community composition and invasibility has been studied extensively but seldom in the context of ecosystem restoration. Experimental riparian restoration plots differing in species composition and density were established and evaluated for susceptibility to invasion by giant reed, a common riparian invader in California, and natural recruitment by riparian species over time. Plots were planted in 2002 with cuttings of common threesquare (a sedge), seepwillow (a shrub), and Goodding's willow (a tree) at two densities in monoculture and all possible mixture combinations. Giant reed rhizomes were introduced into half of the plots in the spring of 2003, while the remaining plots were allowed to undergo natural recruitment for an additional year. In late winter 2004, giant reed rhizomes were planted in the remaining plots. Both planting groups were followed for one growing season to evaluate giant reed establishment, survival, and growth. Community composition affected giant reed performance, particularly in 2003 before natural recruitment occurred. In that year, plots containing seepwillow + willow had the lowest giant reed shoot production, growth, and survival. All plots containing seepwillow were resistant to colonization by natural recruitment in 2004, but none of the planting treatments affected giant reed success in that year. Giant reed was more successful overall in 2004 despite deeper shade and drier soils. This pattern could be attributed to larger initial rhizome size in 2004, which allowed giant reed to overcome environmental stress during establishment. Planting density did not impact giant reed or natural recruitment independently, but may affect environmental parameters and warrants further study as a potential contributor to restoration success. Our results indicate that choice of species composition in restoration might impact giant reed invasion success initially, but community resistance might not be sustainable and maintenance-free over time.
