Ecological correlates of invasion by Arundo donax.

摘要

Arundo donax L. (Poaceae) invades riparian areas throughout California and other warm, coastal states in the US. Once established, it can spread throughout a watershed via flood-dispersal of its rhizomes. In a series of single-factor experiments, the direct effects of rhizome size, light levels, and temperature on A. donax establishment from rhizome pieces were determined. Rhizomes of any size can sprout, but large rhizomes (> 176 g) produce taller shoots than small rhizomes. Production of new root and rhizome biomass was greatest in light shade compared with all other light treatments. There was no effect of temperature on the number of roots or shoots produced. A greenhouse experiment was conducted to test the effects of competition, nitrogen, and their combination on rhizome expansion from previously established A. donax plants. Only nitrogen enrichment had a significant positive effect on rhizome length, suggesting that with adequate nitrogen supplies, giant reed can spread belowground regardless of the competitive environment. In a two-year experimental riparian restoration experiment that manipulated diversity in the restored community, native species richness did not affect A. donax performance in either year. In the first year, A. donax plants established more slowly in Baccharis salicifolia + Salix gooddingii plots. A. donax was more successful overall in the second year of the experiment and did not depend on the original community type or colonizer cover. Colonization by native species was lower in plots that were originally planted with B. salicifolia. In a three-year study of A. donax establishment in three southern California field sites, A. donax performance was explained by positive relationships with pre-planting rhizome weight, soil moisture, and percent bare ground, while inverse relationships were found with species richness and community diversity. However, when each site was considered singly, A. donax performance was influenced by different variables in each location, suggesting a high degree of phenotypic plasticity. Based on the research presented here, land managers should remove all rhizome material from the field, include flood-deposited rhizomes in early detection efforts, prioritize watersheds receiving nitrogenous effluent, and consider incorporating invasion theory in designing restoration projects.