Abstract
As environmental stress increases positive (facilitative) plant interactions often predominate. Plant-plant associations (or lack thereof) can indicate whether certain plant species favor particular types of microsites (e.g., shrub canopies or plant-free interspaces) and can provide valuable insights into whether "nurse plants" will contribute to seeding or planting success during ecological restoration. It can be difficult, however, to anticipate how relationships between nurse plants and plants used for restoration may change over large-ranging, regional stress gradients. We investigated associations between the shrub, Wyoming big sagebrush (Artemisia tridentata ssp. wyomingensis), and three common native grasses (Poa secunda, Elymus elymoides, and Pseudoroegneria spicata), representing short-, medium-, and deep-rooted growth forms, respectively, across an annual rainfall gradient (220-350 mm) in the Great Basin, USA. We hypothesized that positive shrub-grass relationships would become more frequent at lower rainfall levels, as indicated by greater cover of grasses in shrub canopies than vegetation-free interspaces. We sampled aerial cover, density, height, basal width, grazing status, and reproductive status of perennial grasses in canopies and interspaces of 25-33 sagebrush individuals at 32 sites along a rainfall gradient. We found that aerial cover of the shallow rooted grass, P. secunda, was higher in sagebrush canopy than interspace microsites at lower levels of rainfall. Cover and density of the medium-rooted grass, E. elymoides were higher in sagebrush canopies than interspaces at all but the highest rainfall levels. Neither annual rainfall nor sagebrush canopy microsite significantly affected P. spicata cover. E. elymoides and P. spicata plants were taller, narrower, and less likely to be grazed in shrub canopy microsites than interspaces. Our results suggest that exploring sagebrush canopy microsites for restoration of native perennial grasses might improve plant establishment, growth, or survival (or some combination thereof), particularly in drier areas. We suggest that land managers consider the nurse plant approach as a way to increase perennial grass abundance in the Great Basin. Controlled experimentation will provide further insights into the life stage-specific effectiveness and practicality of a nurse plant approach for ecological restoration in this region.