Plant community response to riparian invasive tree removal by biocontrol vs. active means for 40 sites over five years

Background/Question/Methods

Restoration of riparian plant communities in the Southwestern U.S. frequently involves removal of undesirable species, with a particular focus on Tamarix spp (tamarisk, salt cedar).  Tamarix is controlled using chemical, mechanical and biological means, the latter of which employing a rapidly spreading beetle (Diorhabda spp.) first released in 2003.  Each of these approaches alone or in combination has potentially negative consequences on the remaining plant community, which is expected to recover as a consequence of weed removal.  We monitored vegetative response to these three categories of Tamarix removal three times over five years in 40 sites that were established for the Dolores River Restoration Program (DRRP) along the Dolores River in Colorado. Each site was sampled using a point intercept method along five 50m transects, with percent cover measured by species, which were categorized by nativity and functional group. A BACI (Before-After-Control-Impact) design enabled comparisons both over time and between treatments.

Results/Conclusions

Native species cover was negatively correlated with initial Tamarix cover, but on average native cover increased over time and most dramatically with the two active means of Tamarix removal (chemical and mechanical).  Increases in native cover, including the differences in removal, were primarily due to increases in perennial species, including forbs, grasses and shrubs.  Annual grasses, which were almost exclusively exotic, had a net decrease in cover over the study period.  Although reduction in Tamarix cover and native species recovery was slower when Tamarix was only subject to biological control, recovery was also more consistently positive; sites subjected to chemical and/or mechanical removal had greater increases of understory exotic species.   While there has been concern about loss of bird habitat as a consequence of defoliation of Tamarix by the bio-control beetle, our sites with bio-control still contained 10% live Tamarix cover even after nearly a decade since the beetle was released there. Given the concurrent recovery of native species in these sites and the cost and negative impacts of active means of Tamarix removal, these results suggest that biological control may be a preferable management approach in these systems.