Prehispanic Settlement Patterns and Agricultural Production in Tepeaca, Puebla, Mexico, AD 200 - 1519

Open Access
Anderson, J. Heath
Graduate Program:
Doctor of Philosophy
Document Type:
Date of Defense:
September 03, 2009
Committee Members:
  • Kenneth Gale Hirth, Dissertation Advisor
  • Kenneth Gale Hirth, Committee Chair
  • David Lee Webster, Committee Member
  • James William Wood, Committee Member
  • Timothy Michael Murtha Jr., Committee Member
  • Puebla
  • Mexico
  • settlement
  • sustainability
  • EPIC
  • simulation
  • Tepeaca
This dissertation is an investigation of the relationship between settlement patterns and agricultural resources from AD 200 to 1519 in the area surrounding Tepeaca, Puebla, Mexico. In the late 15th and early 16th centuries, Tepeaca was the capital of an Aztec tributary province called Tepeacac that had been conquered by Motecuhzoma Ilhuicamina. Tepeaca enjoyed a strategic location in a natural geographic constriction on an ethnic frontier between the Nahua-speaking populations of the central plateau and the Mixtec-speaking populations to the southeast in Oaxaca. Because of this, it was used as an important military strongpoint and node for interregional trade for the Aztec Empire and later Cortés during his conquest of Mexico. Until the mid-1990s, the only information available about Tepeaca came from ethnohistoric sources. From 1994 to 1997, Penn State archaeologists Ken Hirth and James Sheehy undertook an ambitious program of surface survey, test excavation, and cave exploration within a 560 km2 area around the present-day community of Tepeaca. The survey portion was conducted as a full-coverage reconnaissance. Team members systematically surveyed 100% of the survey area, making surface collections wherever substantial scatters were found and marking the location of each collection on aerial photos. These photos were subsequently digitized and incorporated into a geographic information system (GIS). The sherds from surface collections and excavations were analyzed in a laboratory located in Tepeaca and encoded in a digital database. The result is a census-like record of all prehispanic settlement remains within the survey area. To date, these data are arguably the most extensive, fine-grained survey data available anywhere in Central Mexico. In 2002, Ron Castanzo used the survey and excavation data to build a ceramic chronology and reconstruct settlement for the Formative Period (ca. 950 BC – AD 200). From modest beginnings in the Middle Formative, Castanzo showed that the Terminal Formative was a time of pronounced growth in terms of overall population size and the maximum size of the largest settlements. In this dissertation, I fill in the rest of the prehispanic settlement history by reconstructing settlement patterns for the Classic (AD 200 – 600), Epiclassic (AD 600 – 900), Early Postclassic (AD 900 – 1200), and Late Postclassic (AD 1200 – 1519) periods. To facilitate the present settlement study, I used stratigraphic information from test excavations conducted within the survey area in conjunction with cross-dates from known sequences in adjacent areas to construct the first ceramic chonology for these periods. Following Castanzo, I used the distribution, proximity, and density of ceramic scatters to reconstruct prehispanic settlement. As I demonstrate in Chapter Five, the marked population and settlement growth Castanzo documented for the Terminal Formative did not continue in the Classic Period. In fact, the Classic Period in the Tepeaca area was a time of stagnation or decline in population growth and general settlement dispersal. Although the trend of settlement dispersal continued into the Epiclassic Period, population growth resumes during that period and remains steady throughout the Early and Late Postclassic. The Late Postclassic is the most problematic of the phases, however, because the ceramic chronology for this period is incomplete. Because the only reliable diagnostic ceramics for the Late Postclassic are polychrome service wares, the settlement patterns and population estimates I reconstruct for the centuries leading up to the Spanish Conquest underestimate the true extent and magnitude of prehispanic occupation. Ethnohistoric sources indicate that the population within the PAT survey area was at its greatest during this period. In all time periods, settlement was markedly dispersed. Small, isolated residences and hamlets of 100 inhabitants or less constituted the overwhelming majority of settlement types from AD 200 to 1519. Larger communities were rare, but became more frequent in the Epiclassic and later periods. The configuration of these larger settlements also underscores the dispersed nature of settlement patterns in the Tepeaca area. Even the largest settlements in terms of estimated population size tended to grow in area as they grew in number of inhabitants instead of nucleating and concentrating more people into a small, densely populated area. This stands in marked contrast to the pattern evident in the Basin of Mexico, and constitutes one of the most dispersed, stable settlement patterns evident anywhere in Central Mexico. To investigate some of the basic causal factors driving the settlement patterns, I use a simulation model to investigate maize productivity within the survey area. I employ a model developed by the United States Department of Agriculture called the Erosion Productivity Impact Calculator (EPIC). EPIC uses soils, weather, and management input data to simulate the effect of soil and nutrient loss on agricultural productivity. I develop two measures, initial and sustained productivity, to characterize the landscape in terms that would have been important to prehispanic farmers. Initial productivity refers to the high yields farmers could have expected in the first few years of cultivation, before nutrient and soil loss from repeated cropping reduced soil fertility. Sustained productivity refers to the amount of time a plot of land could be farmed before annual productivity decreased below the threshold necessary to sustain a farming household for one year. As I show in Chapter 7, the Tepeaca landscape is a very risky one from the perspective of a prehispanic subsistence farmer in terms of initial and sustained productivity because of marked interannual rainfall variability. Even cultivating the best lands within the survey area, most households would have had to reduce the amount of maize in the diet in order to harvest enough maize to satisfy their immediate requirements and retain enough surplus for storage and consumption in lean years. As I discuss in Chapter 8, the riskiness of maize agriculture and the benefits of a supplementary source of calories other than maize such as maguey may have been major causal factors in producing the dispersed settlement patterns evident in the Tepeaca area. If cultivated in the familiar infield/outfield configuration documented ethnographically in other parts of Mesoamerica, the inclusion of an infield with each constituent household within a community would have forced larger communities to grow in area as they grew in population size. Although I do not have positive evidence for this practice in the Tepeaca area, this practice has been linked to highly dispersed settlement patterns elsewhere in Central Mexico, notably the Teotihuacan Valley. In sum, the stability of the dispersed settlement patterns in the Tepeaca area in the face of regional political changes like the rise and decline of Teotihuacan, the political fragmentation and reorganization during the Epiclassic Period, and the advent of the Aztec Empire is best explained as the result of a very stable accommodation by prehispanic farming households to a risky agricultural landscape.