Modeling of Displacement by Woodlands and Cheatgrass

Ecological provinces in the
Intermountain West

We developed two independent, rule-based predictive models of displacement of sagebrush and other native vegetation by encroaching pinyon-juniper woodlands and cheatgrass, employing a variety of topographic and vegetation variables (Suring and others 2005). Simple rule-based models may be effective in displaying broad-scale patterns of risk in susceptible plant communities (Woodbury and Weinstein 2008). The models were deterministic, (i.e., the model assumptions completely determine the outcome of the model, with no element of randomness involved), and were intended to display broad-scale patterns of risk across the ecoregion, summarized at watershed, (i.e., 5^th hydrologic unit code) or larger scales. The models were applied first to vegetation, and then to habitats for individual species of concern (pinyon-juniper and cheatgrass models) and species groups (cheatgrass model only). The models were intended to display risk of displacement of native vegetation in 30 years by the two stressors; thus, areas mapped as high risk may or may not have been converted to cheatgrass or woodland-dominated sites at present.

We identified environmental variables thought to be most important for estimating the risk that sagebrush will be displaced by pinyon-juniper woodlands or cheatgrass by reviewing the literature, including studies of dispersal mechanisms for pine seeds and juniper berries, and consulting with species experts on the ecological relationships and traits of pinyon and juniper species and cheatgrass. The rule sets for both models used ecological provinces (described by Miller and others 1999b and West and others 1998) as a geographic basis for model development (see figure on the right). The characterization of landscape conditions within the provinces provided a useful ecological context for model development.