A mixed blessing for restoration in arid shrublands: Native perennial grasses reduce erosion and weeds, but also native shrubs

Background/Question/Methods

Shrubs are integral components of many arid ecosystems, as they are long-lived, drought-tolerant, and provide valuable wildlife habitat. In arid regions, native shrub restoration is hindered by both abiotic and biotic challenges. Abiotic challenges include wind erosion and water stress, which often limit revegetation in arid ecosystems. Biotically, shrubs can be difficult to restore because they are weak competitors against invasive annual species. Under these challenging conditions, planting native shrubs into denuded sites is unlikely to be successful unless additional steps are taken to mitigate restoration barriers. We hypothesized that applying short-term irrigation and seeding native perennial grasses would facilitate native shrub establishment by reducing water stress, erosion and weed invasion during the first years of shrub establishment. At two former agricultural fields in Nevada’s Mason Valley, we used a blocked split-plot experimental design to test the separate and combined impacts of initial irrigation and perennial grass seeding on long-term (5-year) restoration outcomes. Plots assigned to the irrigation treatment were watered using deficit irrigation for two years. Seeding treatments included a control (no tilling or seeding), a till-only treatment, and four different drill-seeded perennial grass monocultures.

Results/Conclusions

Irrigation and perennial grasses were successful in mitigating abiotic and biotic barriers that commonly hinder shrub establishment in arid regions. After the first two years, irrigation had increased the density and biomass of seeded perennial grasses by more than ten-fold. The combination of irrigation and seeded grasses was associated with significantly lower weed density, weed biomass and wind erosion. Three years after irrigation ended, perennial grasses remained >3 times more abundant in previously irrigated plots than in non-irrigated plots. Previously irrigated plots also had significantly less bare ground, less annual plant cover and less weed biomass than non-irrigated plots. Large (>2 m) plant canopy gaps were least abundant in plots that had been both irrigated and seeded with perennial grasses. Although they successfully resisted erosion and invasion, native perennial grasses failed to facilitate native shrubs. Native shrub cover and density were highest in plots that had been tilled and irrigated, but lacked perennial grasses. Our results indicate that (1) initial irrigation had clear and persistent restoration benefits, and (2) a tradeoff exists between the benefits (reduced erosion and weed invasion) and costs (reduced shrub establishment) of seeding perennial grasses into arid shrubland sites.