New Geospatial Approaches to the Anthropological Sciences
edited by Robert L. Anemone and Glen C. Conroy
New Geospatial Approaches
                to the Anthropological Sciences

Chapter 8. Understanding the Ecological Decision-Making of Tiwanaku Pastoralists through Geospatial Agent-Based Models

B Vining and S Burns

Additional Chapter details here


Fig. 1. The location of the Suches study region in the central Andes.

Fig. 2. Recorded archaeological sites in Suches. Data from Vining 2016; Stanish et a., 2010.

Fig. 3. Comparison of Middle-Late Formative period settlement, characterized by nucleated villages (top) with Middle Horizon settlement characterized by scattered individual hamlets.

Fig. 4. Alpaca herd growth curves based on three different estimated rates.

Fig. 5. The distribution of ecological niches for each cultural phase predicted by maximum-entropy modeling and recorded archaeological sites.

Fig. 6. A comparison of MaxEnt (top) and WLC (bottom) niche-distribution models. Graphic representations (left) show the spatial distribution of predicted niches. Histograms (right) show the loading of known archaeological locations into model scores, based on frequency.

Fig. 7. A comparison of three ABM model runs simulating changing herd ecology at the Formative-Middle Horizon transition: (a) village-based management, with minimal reproduction rates and optimal herd sizes in which no "move" is triggered; (b) village-based management, with increased reproductive rates and minimal optimal herd sizes in which "move" decisions are triggered, causing new locations to be colonized; (c) initial village-based management, with increased reproduction rates and optimal herd sizes, triggering frequent "move" decisions and the colonization of new territories.

Fig. 8. The relationship between observed Middle Horizon sites and the averaged results of one hundred runs of the PED-ABM. Neighborhoods represent significant clusters of predicted settlements at a 90-99% confidence interval, given the herding strategies shown in figure 8.7c.