Background/Question/Methods SageSTEP is a regional experiment designed to evaluate several different methods to maintain sagebrush habitat while reducing fire fuel loads in the Great Basin.  Land management options, including prescribed fire, mechanical thinning of shrubs and trees, and herbicide application are currently, or have been proposed, as fuel treatments that might restore or maintain healthy and diverse native plant communities.  Results of this project will provide resource managers with improved information to make restoration management decisions with reduced risk and uncertainty.  Seven sites were chosen to conduct this experiment within the arid sagebrush steppe biome in the Great Basin.  Each site chosen contained a soil that is associated with an ecological site represented by an Artemisia tridentata Nutt. ssp. wyomingensis overstory and an intact native bunchgrass understory.  Furthermore, Bromus tectorum L. occurred at each site at varying levels but never in dominate amounts.  Due to the high variability in soils and climate across this biome, we determined the similarity among these sites based on pre-treatment floristic composition and abiotic factors.  We determined plant species constancy and indicator species for each site.  Plant-environmental relationships among sites were analyzed using non-metric multidimensional scaling (NMS).
Results/Conclusions Preliminary results from NMS ordination revealed that site differences where driven by abiotic factors, primarily soil sand versus clay content.  Other factors influencing site differences included elevation, precipitation, and frost free days.  The ordination explained 55 % of the variation found among the seven sites.  Constant plant species were composed of native perennial bunchgrasses and Artemisia tridentata wyomingensis.  Indicator species (plant species unique to particular sites) were composed of native annual forbs.  Overall, sites tended to be more similar to each other than different, although we did obtain a good spread of environmental factors among sites. We believe this variation in abiotic conditions among sites with similar vegetation associations provides a unique set of experimental units to allow for a wide range of inference throughout this important arid ecosystem for testing responses of management actions.