By: Brett Hill
The use of upland floodwater farming techniques by prehistoric farmers is frequently cited as an ingenious adaptation to agriculturally marginal, semi-arid environments. These techniques divert and retain water and soil in small hill-slope plots that overcome rainfall limitations by capturing runoff from a larger watershed in a concentrated area. Another, less frequently discussed, consequence of floodwater farming is its potential for exacerbating erosion and rendering agricultural plots less fertile. Indeed, many prehistoric check dams and other features were probably attempts to counteract the loss of soil in these already tenuous agricultural efforts. Floodwater techniques were effective in producing crops in otherwise non-arable locations, yet it was necessary to maintain a precarious balance between the benefits of floodwater farming and the potential for soil degradation. The frequent abandonment and relocation of prehistoric communities utilizing such techniques may, in fact, reflect a failure to maintain such a balance.
This project focuses on the prehistoric Salado communities of the Tonto Basin in central Arizona, where settled farming populations lived for over a thousand years until regional abandonment sometime in the 15th century. An abundance of upland archaeological remains indicate that floodwater farming on hill-slopes surrounding the basin was an important part of the Salado subsistence strategy. We are using the ArcView Geographic Information System (GIS) with data curated by the Archaeological Research Institute of Arizona State University to model floodwater farming and it’s potential contribution to erosion in the Tonto Basin. Using data on topography, soils, hydrology, precipitation, vegetation, and settlement locations we have created three basic models that can be interrelated for analysis.
First, we created a model of likely floodwater farming locations based on terrain. Ethnohistoric and archaeological records indicate that floodwater farming was typically practiced on slopes of from 3% - 10% and in watersheds ranging from 1 – 10 hectares. Thus, we have identified areas satisfying these topographic constraints as having been the most likely ancient floodwater field locations.
Second, we created a model of use intensity based on the locations of the largest known sites in the region. Assuming that landscape use intensity would be greatest near large sites and decrease with distance, we used a terrain factored cost surface, radiating out from sites with over 20 rooms, to a distance of 8 k, to assign a use intensity value ranging from 1 to 0 to surrounding catchments. Assuming furthermore that degradation would be cumulative, we additively overlayed use intensity values for catchments that were utilized through three periods, such that use intensity values range from 3 to 0.
Third, we created a model of potential erosion using the Universal Soil Loss Equation (USLE). Using known rainfall factors, soils and terrain data, and variable vegetation parameters we calculated the tons per acre, per year, of soil lost under various conditions. The GIS allows us to easily explore erosional consequences of changing vegetation and rainfall patterns, as they would affect upland farmers.
Research is continuing on this project along two fronts. One advantage to the use of GIS in the evaluation of ancient environmental degradation is that we can combine the data we have developed to create maps pinpointing locations for field analysis. We can easily identify locations that display a high potential for both ancient use and susceptibility to erosion for closer examination. In this way we improve the efficiency and precision of field research, and develop comparative data for areas displaying different potential in our models.
A second avenue of continuing research will be to further explore the integration of potential erosion and landscape use patterns. For example, we can evaluate how changing USLE parameters based on location specific changes in land use might affect soil loss, and hence agricultural productivity. Likewise, we can use hypothesized settlement location strategies in iterative simulations to develop a better understanding of the role of upland farming resources and their degradation in those strategies.