Removal of non-native, invasive plants from riparian areas throughout the western United States is widespread and expensive, yet the ecological yield on such investments is extremely variable.  The objectives of non-native species removal efforts range from salvaging water (through reducing evapotranspiration) and minimizing fuel-related fire damage to native riparian forests to restoring native species and recovering degraded riparian ecosystems.  We monitored response of riparian ecosystems to Tamarix and Elaeagnus removal along a 320 km reach of the Middle Rio Grande, NM.  Herpetile, bird, mammal and bat community composition were measured by collaborators in a replicated, controlled, landscape-scale experiment.  Our focus was on the response of plant communities and native riparian Populus forests to non-native species removal.

We quantified vegetation and stand-level responses over five years through measuring plant species composition, riparian forest canopy and stand structure, as well as the effects of fuel reduction efforts on survival, growth and fitness of native forests following wildfire.  We hypothesized that growth rates and fitness of native species would increase due to competitive release following removal of non-natives.  Further, we hypothesized that mortality of native species associated with wildfires would decrease following fuel reduction treatments and that richness and vigor of native vegetation would increase.  Through examining changes in species composition, functional shifts, and incremental growth of individuals and stands, we evaluated the ecological returns on the investment.

Results/Conclusions

We found that the frequency of non-native herbaceous species was higher in areas disturbed during control efforts, resulting in higher proportion of non-native herbaceous species in treated sites compared to controls.  The risk of re-invasion by exotic species following removal efforts can be minimized by evaluating site conditions that might facilitate post-treatment invasion (e.g., presence of local seed sources, vegetative propagule sources, vectors, the creation of ‘open sites’).  Vertical forest structure was reduced in treated areas, but recovery of such habitat structure has been expedited through active transplanting of native shrubs.  Patterns of herbaceous vegetation recovery were strongly related to site-specific factors (light, litter depth, available moisture, and flooding); factors which are influenced by non-native species removal and/or that can be manipulated by managers.  We found little or no response of riparian cottonwood fitness or fire-associated mortality rates due to fuel reduction efforts.  Non-native species control is only one element toward restoration of native riparian vegetation; in certain settings removal of non-natives may have unintended negative ecological effects.