Aim: We tested the hypothesis that the diversity and abundance of aquatic macrophytes are negatively related with Eichhornia crassipes abundance in its introduced range, but not in its native range. 

Location: Upper Parana River Floodplain, Brazil and Southeast China

Methods: We sampled aquatic macrophytes patches in Brazil (native range) and China (introduced range) along a biomass gradient of E. crassipes. For each patch, we obtained values of species richness and aquatic macrophytes percentage cover, as response variables in regression models. We also used species accumulation curves to quantify the total plot diversity in dominated and non-dominated plots for both countries. Finally, we compared the influence of E. crassipes dominance on community composition and beta diversity with Permanova and Permdisp, respectively.

Results: The regression analyses revealed a negative correlation between macrophyte richness and cover and E. crassipes biomass only in the introduced range. The cumulative number of species decreased at a higher extent in plots dominated by E. crassipes in China, compared to Brazil. Also, species composition changed and beta diversity decreased in the dominated plots in China, but not in Brazil.

Main conclusions: The reduction of all diversity attributes related to E. crassipes probably results from its engineer species role, which decreases littoral region habitat heterogeneity and affects rare species in the introduced range. Differences between countries may be associated with impacts of water hyacinth on native macrophytes since this plant grows very fast and is highly competitive. Although less probable, biotic resistance at the establishment phase of water hyacinth in sites with higher number of native species is also a possibility. Regardless of the main mechanism explaining our patterns, it is suggested that invasion by water hyacinth is a cause for concern for its higher impacts in the introduced ranges than the native ranges.