Analytical and Modeling Approaches

Scientifically defensible approaches have been developed over the last 2 decades for assessing vegetation, herbivore, and ecosystem dynamics in spatially heterogeneous environments from landscape through regional and even to global scales. A wide variety of models have been developed that are capable of simulating vegetation, biogeochemistry and hydrology dynamics in response to soils, changing climate, and, to a lesser extent, herbivory A body of science in this field does exist focusing on vegetation and ecosystem responses to herbivory. Some models are capable of simulating interactive responses to herbivory, climate, and soils. Although computer modeling has been adopted by BLM to predict horse population responses to management and to assist in the setting of the lower bounds of AMLs (see Chapter 6), modeling has not been used to set the upper bounds of AMLs or to inform AML decisions. That would necessitate models of vegetation and ecosystem dynamics and the ability of such models to represent ecosystem dynamics in spatially heterogeneous environments and mobile herbivore populations. Assessments of AMLs could be made more robust and more informative by using the powerful analytical and modeling approaches.

A first step would be to use geographic information systems (GIS) and remote sensing to a greater extent in setting and evaluating AMLs. BLM uses GIS to some extent to quantify vegetation and forage production potentials in different range sites, as delineated by NRCS or older Soil Conservation Service soil surveys. Forage production estimates for each range site have been combined or scaled up by using GIS to derive forage production. However, there is a potential to do more with spatial data and to derive additional data from remote sensing, for example,

  • Overlay spatial data on equid distributions, which are temporally variable, onto forage production estimates to predict percentage utilization across the landscape. Even if the equid distributions are coarse or estimated, they represent what is known.
  • Use spatial modeling of equid habitat selection on a seasonal basis to provide estimates of equid distributions. Equid habitat-selection patterns will be influenced by distance to water, topography, forage quantity and quality, shrub and tree cover, barriers to movement, conflicting land uses, forage offtake by livestock, and other factors. All those can be represented as GIS data layers.
  • Use remote-sensing data to cross-check and augment estimates of forage production.
  • Use spatially explicit precipitation maps that account for patchy rainfall and topographic gradients to refine estimates of forage production.
  • Estimate snowpack distributions by using remote-sensing products, SNOTEL data, snowpack modeling, and spatial interpolation to estimate areas that are available to horses in winter. The snowpack in turn affects the forage supply for the horses in winter.

As just noted, various vegetation and biophysical-ecosystem models have been developed over the last 3 decades. All have capabilities of simulating realistic scenarios of plant production and vegetation dynamics in response to soils and climate. However, few have focused on simulating vegetation or ecosystem responses to herbivory. Few have explicitly represented herbivores or their dynamic distributions on the landscape. However, one example that does is the application of the SAVANNA ecosystem model to the Pryor



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement