This research determines and demonstrates the ability to simulate dynamic multispecies animal habitat suitability with forest succession models. A literature review of dynamic animal habitat models is presented. The structure of an existing forest simulation model (MASS10) was modified from a basal area-based model to a volume-based model (DYNAM10). The forest model was calibrated using data from permanent-plot growth and vegetation samples collected by USDA Forest Service Forest Survey procedures. The theoretical growth parameters used to simulate stand development were validated. Predictions of DBH and height growth, as well as stand-level behavior, were verified. A subroutine, VEGDYN, was added to DYNAMlO to simulate 34 structural vegetation parameters required by animal Habitat Suitability Index (HSI) models. Predictions of the structural parameters were verified. Ten animal-species HSI models were linked to DYNAMlO via the program HSI.FOR, and predicted dynamic HSI values were verified by hand-calculation. Typical patterns of dynamic HSI predictions are presented and discussed.

The interaction of land-use and climate can cause non-linear “state” changes in ecosystems, characterized by persistent differences in structure and function. Changes in land-use and climate on the Colorado Plateau may be driving many ecosystems toward undesired states where energy-intensive measures are required to return to previous states. Landscape classification systems based on “ecological potential” offer a robust framework to evaluate ecological conditions. Ecological sites are a popular landscape classification system based on long-term ecological potential and are widely used throughout the western US. Ecological sites have been described extensively for rangelands and woodlands on DOI Bureau of Land Management lands; however, they have yet to be described on USDA Forest Service (USFS) lands. In this thesis, I describe a statistical approach to ecological site delineation and the development of state-and-transition models, diagrams that illustrate ecosystem dynamics and responses to disturbances. In Chapter 2, I used a large inventory dataset and multivariate statistical procedures to classify plots based on life zone, soils, and potential vegetation, effectively delineating statistical ecological site-like groups. Most of the statistical ecological sites matched ecological sites already described by the USDA Natural Resources Conservation Service (NRCS). Additionally, I described one new ecological site that has not been described by the NRCS in the Colorado Plateau region. In Chapter 3, I examined empirical evidence for alternative states in mountain ponderosa pine (Pinus ponderosa Lawson & C. Lawson) and upland piñon-juniper ecosystems. Using multivariate statistical procedures, I found that plots cluster into groups consistent with generalized alternative states identified in a priori conceptual models. Additionally, I showed that ponderosa pine clusters were true alternative states and piñon-juniper clusters were not true alternative states because they were confounded by similarities in climate. Ponderosa pine clusters were differentiated by overstory ponderosa pine density and corresponded to three states: current potential, high fuel load, and reduced overstory. These results illustrate the range of ecosystem variability that is present throughout the study area and present evidence for alternatives states caused by historical land-use. This project is the first to propose ecological sites and state-and-transition models on USFS lands in this region. These techniques could be applied to areas that do not have formally described ecological sites and state-and-transition models and could help identify ecological sites that may have been overlooked using other means of delineation. Additionally, these methods can be used to evaluate the range of ecological variability throughout an area of interest and to improved understanding of ecosystem dynamics.