This dissertation reports on an intense effort to model the natural ecology of multiple resources in a large study area of the southwest United States. The dynamic simulation of standing crops of fuel woods and populations of wild herbivores is designed to provide natural resources in an agent-based model of long-term human settlement strategies in a bounded landscape under changing natural and social environmental conditions. Model agents are simulated households that settle particular areas to survive on a spatially and temporally variable landscape providing different resources at different locations. These households are required to obtain basic necessities in the forms of water, food, and fuels as supplied by the model world. The resources modeled here include woody species available as fuels, the primary productivity of other native species that provide the food to support wild herbivores, and three animal species commonly recovered from archaeological contexts. The three animal species are mule deer, black-tailed jackrabbits, and cottontail rabbits. Populations of these animals are simulated based on the annual productivity of their preferred foods as supplied by the native vegetation communities associated with study area soils. Requiring model households to satisfy basic natural resource needs is intended to improve the fit of simulated household settlement patterns with the long- term settlement patterns observed from the archaeological record of agrarian peoples in the Mesa, Verde region from A.D. 600 to 1300. Result indicate that inclusion of these critical natural resources on the model landscape as requirements of sustained life for model households does improve the goodness-of-fit between simulated and observed settlement patterns. |