Background/Question/Methods Invasive species, such as Tamarix spp. (tamarisk), often alter ecosystem properties and reduce native species diversity. However, native and exotic species diversity are often positively correlated within plots sampled across landscapes. Thus, a goal of restoration that removes dominant invasive species should include changing the relationship between native and exotic species by reducing richness and abundance of exotic species while increasing native species. Tamarisk has invaded thousands of hectares of riparian zones in the southwestern US, including ecosystems in Grand Canyon National Park. The National Park Service and partner organizations began a tamarisk management project in 2000, which included a restoration study investigating the effects of tamarisk removal on native plant diversity and composition. Plant species composition and diversity were sampled along transects in tamarisk-dominated stands prior to tamarisk removal. Transects were then resampled one to three years following tamarisk removal. Paired reference stands containing little or no tamarisk also were sampled to determine community diversity and structure in uninvaded stands. Species diversity and compositional changes were assessed using data collected before and after tamarisk removal. We also investigated changes in relationships between native and exotic diversity and abundance to determine whether restoration goals were achieved. 
Results/Conclusions After removing tamarisk, species composition tended to shift toward reference conditions as predicted. These patterns were less clear when tamarisk was removed from the species × transect data matrix reflecting the fact that reduction in tamarisk abundance drove much of the patterns in compositional shifts. Total plant cover and abundance of functional groups tended to decrease following tamarisk removal. Before tamarisk removal, native and exotic species diversity and abundance were positively correlated. After tamarisk removal, the native-exotic relationships had decreased or became non-significant, reflecting a dramatic decline in exotic richness and abundance during the treatment applications. However, native diversity varied little between pre- and post-treatment sampling times. These results suggest that investigating changes in native-exotic species relationships - and not solely the response of native species - may be an informative tool to determine the success of invasive species removal restoration programs and to help determine thresholds for management actions.