Conserving and restoring grasslands and shrub-steppe has important ecological and socioeconomic implications in arid rangelands of the world. Among the greatest contemporary threats to the structure, function and biological diversity of desert grassland and shrub savanna ecosystems of the southwestern United States is the displacement of mesophytic grasses by xerophytic woody plants. The needs of land managers in addressing this issue have surpassed the information provided by inventories and traditional monitoring programs. Land managers must understand the mechanisms behind shrub expansion in order to identify grasslands that are at risk and shrublands that may be viable candidates for grassland restoration. As yet, there are no objective criteria for making these key distinctions. Through a combination of field sampling and spatial modeling we sought to create a predictive understanding of the biophysical conditions under which grasslands are converted to shrublands. We asked: What are the physiographic settings of persistent grass patches and what environmental factors contribute to their persistence? In communities that are a mosaic of grass and shrub patches, which of these are stable and which are in the process of transitioning to shrub dominance?  What do spatial patterns of past shrub expansion suggest about future landscape change? As a first step in addressing these questions, we characterized landscape-scale perennial plant cover across grassland-to-shrubland gradients within the Jornada Basin near Las Cruces, NM. Soil was characterized to 150 cm or to restrictive calcium carbonate horizon and the relationships between perennial plant assemblages and soil characteristics were explored using multivariate statistics. Changes in shrub and perennial grass patch structure and distribution were quantified using time-series aerial photography (1940s - 2000s). Field survey and aerial photography data were then used to develop a model of patch structure dynamics focused on characterizing patterns of vegetation structure and cover that portend dramatic shifts in shrub-grass abundance.

Results/Conclusions

Remnant grass patches were associated with restrictive soil horizons. Expansion of shrub patches was dependent on local soil characteristics and the context of the surrounding landscape (e.g., landform and adjacent vegetation structure). Although historical events are doubtless important triggers for shrubland expansion into grasslands, there is a strong signature of soils and landscape position in the pattern and rate of shrubland expansion. This signature can be used to distinguish at-risk areas from areas with high, inherent grassland or savanna resilience.