Suzanne K. Fish, Paul R. Fish, and John H. Madsen
The prehistoric residents of the Tucson Basin were participants in a Hohokam cultural tradition spanning 116,000 square kilometers (45,000 square miles) in the desert basins of south central Arizona (Fig. 1.1). The name Hohokam is usually reserved for pottery-making inhabitants of this territory, although Late Archaic occupants had adopted a farming lifestyle centuries prior to the earliest manufacture of ceramics about 2000 years ago. One of the major subdivisions of archaeological cultures in the southwestern United States, the Hohokam tradition is distinguished by designs in red paint on buff, brown, or gray pottery during all but the initial and final segments of the ceramic sequence.
The initial interval of ceramic manufacture is known from limited exposures, but has been documented in widespread locations during the first few centuries of the Christian Era. Compared to the precise, tree-ring-dated archaeological chronologies of the northern Southwest, Hohokam sequences are not internally refined or well synchronized among the regions of the tradition. An exhaustive review of chronological data has recently been completed by Jeffrey Dean (1991) and is the source for schema in Figure 1.2. The final years of the tradition remain a subject of debate, marked by a paucity of evidence between the early fifteenth century and reliable Spanish observation in the 1680s. Primarily Piman-speaking Indians of the historic period represent geographic successors in limited sectors of the former Hohokam domain and exhibit only a generalized level of cultural continuity.
The Hohokam are distinguished from other prehistoric peoples of the southwestern United States by a strong cultural orientation toward the Mesoamerican fringe to the south (P. Fish 1989: 21). Both in stylistic elements and in settlement pattern the Hohokam resemble the historic, river-oriented, rancheria cultures of adjacent Sonora and Sinaloa more than the Puebloans of uplands to the north and east. Hohokam ceramic traits with Mexican associations in a variety of periods include incising, grooving, repetitious small design elements, and vessel forms such as molcajetes (chili graters), comals (griddles), tripods, and censers. In the earlier part of the sequence, figurines, censers, and palettes form a suite of ritual items that are rare in the Southwest beyond the northern extent of Hohokam occupation. Similarly, mounds and ballcourts are classes of public architecture with distributional continuity to the south, in contrast to the kivas of Puebloan groups. More than other spheres of material culture, these ceremonial, ideological, and, by inference, organizational expressions have been the operative criteria in delimiting the boundaries of the Hohokam tradition.
The Hohokam lived in pit houses with wattle and daub or brush superstructures throughout the sequence. Adobe rooms were added about A.D. 1100 near the beginning of the Classic period and were often grouped within the walls of a compound. Earthen-banked ballcourts were the common form of public architecture before the Classic period, superseded thereafter by platform mounds supporting adobe structures. Public architecture was constructed at the large sites in differentiated settlement patterns. Hohokam agriculture was distinctive in its scale, intensity, and engineering feats that created hundreds of kilometers of canals along perennial rivers. A variety of alternate farming technologies were employed in other environmental settings.
A watershed in Hohokam cultural development is indicated by the transition to the Classic period about A.D. 1150, although more continuities with Preclassic times are now evident than previously. Canal systems reached their greatest extent, and acreage cultivated by alternate techniques was expanded in many areas. Maximum population densities at the largest settlements and highest overall densities for most regions were achieved after this transition. Greater investment in integrative organization can be seen in the construction of the most massive examples of public architecture, with which socially differentiated personnel were now associated.
Hohokam in all reaches of the tradition shared the challenges of a relatively low, hot, desert environment. Their geographic extent to the north and east coincided closely with the vegetational attributes demarcating the Sonoran Desert. More tropical facies of this desert extend southward from Hohokam territory, and drier facies continue to the west and southwest. Even in the vicinity of the larger mountain ranges in the basin-and-range country of southern Arizona, Hohokam settlements were concentrated in basin interiors, seldom exceeding an elevation of 1100 m (3500 feet).
Bimodal rainfall in the Hohokam area is associated with a greater diversity in growth forms among plant communities and a more arborescent character of vegetation than under winter-dominant precipitation to the west and summer-dominant to the east (Turner and Brown 1982: 182). Associations of shrubby plants near basin floors (Fig. 1.3) predominantly contain mixes of creosote bush, bursage, saltbush, and similar species, but larger plant forms are common in associations on basin slopes (Fig. 1.4). Distinctively large Sonoran Desert perennials include columnar saguaro cactus and the leguminous trees, mesquite, palo verde, and ironwood. Numbers and size of economically important trees and a comparative variety and abundance of edible cacti set off the plentiful plant resources of the Sonoran Desert basins from those of the adjacent Chihuahuan and Mohave deserts.
Parallel series of desert basins are bounded and separated by discontinuous mountain ranges of generally limited mass. Widths of well-defined basins range from 10 km to 32 km (6 to 20 miles), with the broadest rarely wider than 48 km (30 miles; S. Fish and Nabhan 1991: 31). Rock pediments along mountain flanks are overlain by relatively thin soils, but detrital sediments fill the basin interiors to great depth. Water in drainages from bordering uplands deposits suspended soil as it leaves the mountains, creating alluvial fans. These outwash fans coalesce on the lower slopes in zones of active deposition. Slopes descending from bordering ranges are the dominant landforms of inner basins and are called "bajadas,” a term used throughout this monograph. The floodplains of axial drainages are positioned along the basin floors where the bajadas from opposite mountain borders meet.
Yearly precipitation in the area inhabited by the Hohokam rarely surpasses 400 mm (15 inches), but local averages vary by a factor of two. Totals between 175 mm and 300 mm (7 to 12 inches) are widespread, with a few locales receiving less than 150 mm (6 inches). Annual precipitation below 225 mm (9 inches) corresponds with greater variation about the mean and higher summer temperatures in the middle of the Hohokam domain near the modern city of Phoenix and in a swath to the southwest. Phoenix temperatures reach 40° C (100° F) about 90 days per year. To the north, east, and south, higher elevations and somewhat ameliorated conditions occur in conjunction with increasing proportions of rainfall during the summer months.
Climatic extremes in the Phoenix Basin at the center of the Hohokam world (see Fig. 1.1) were countered by hydrological advantages for agriculturalists. The confluence of the Salt and Gila rivers lies just southwest of Phoenix. The two conjoined valley segments upstream from the confluence form the Phoenix Basin, which contains the broadest expanse of flat, irrigable floor in the entire Hohokam tradition. The Salt and Gila transport water from vast upland watersheds to the north and east outside the low desert. Peak flows in the spring carry mountain snowmelt. With insignificant seasonal frost, two cropping seasons were possible for Hohokam irrigators along canal networks of the largest scale in prehistoric North America. More than 500 km (300 miles) of main trunk lines have been mapped (Masse 1981; Nicholas and Feinman 1989).
The densest regional population and largest Hohokam settlements occurred in the Phoenix Basin. Estimates for maximum population range from a low of 30,000 persons to more than 100,000 persons (Haury 1976: 356; Schroeder 1960; Doyel 1991; P. Fish and S. Fish 1991) in an area of 2000 square kilometers (750 square miles). Impressive remains attracted the great majority of early field research by archaeologists, and urban construction has continued to prompt intensive investigations in recent years. In contrast with the archaeology of other Southwestern areas, Hohokam studies have emphasized comparison against the Phoenix sequence rather than the definition of regional variants. This combination of factors has resulted in a strong core-periphery model, in which outlying regions have been considered imitative and less developed.
The core-periphery dichotomy has also been cast as a distinction between riverine and nonriverine or outlying desert regions, based on the implications of differential opportunities for large-scale irrigation (Haury 1950: 546–548). In spite of higher rainfall north, east, and south of the Phoenix core, rivers lack sizable upland watersheds and are intermittent rather than perennial. Basin and floodplain morphologies restrict the width of irrigable land. Peak flows in summer rather than spring and greater frost hazards limit early crops. Settlement was more dispersed in these areas, and it has been generally assumed that populations lacked the productive base for an elaborated cultural development equal to that in the Phoenix Basin. Spotty and sporadic research in the vast geographic remainder of the Hohokam tradition outside the Phoenix Basin did little to modify these perceptions before the last decade. Accelerated investigations elsewhere have now revealed unanticipated levels of population and an equivalent array of material culture, including an increasing number of recorded sites with public architecture.
In the Phoenix Basin, irrigation networks and large sites were mapped in the early part of this century before extensive modern land use, providing a means for recognizing clusters of interrelated sites along a shared canal line. A unit of settlement surrounding a central site with public architecture was termed a "community" on the basis of common interests in the acquisition and distribution of water (Doyel 1974, 1980). Integrative functions are ascribed to the central site, as embodied in communal construction and observances at ballcourts and mounds. Multiple community units along canals as long as 30 km (18 miles) could be inferred from the spacing of central sites at fairly regular intervals (Wilcox and Sternberg 1983: 195; Crown 1987: 154; Gregory and Nials 1985).
At the start of this study, sites with public architecture in outlying regions implied similar functions of multisite integration and some form of community organization outside the Phoenix Basin. These instances did not coincide with large-scale irrigation, however. Configurations of related settlement surrounding such sites, and their productive bases were unknown. In the Tucson Basin, ballcourts and mounds were known to occur both near the Santa Cruz River and in other basin settings.
Defined by the drainage basin of the Santa Cruz, a major desert river, the Tucson Basin (Figs. 1.1, 1.5) is a typical focus of regional research in Southwestern archaeology. Mountains rimming the basin and dividing it from other drainage systems form physical barriers promoting a degree of both natural and cultural closure (Figs. 1.6, 1.7). Annual precipitation is between 225 mm and 300 mm (9 to 12 inches). Rainfall must be concentrated for successful cropping under conditions of rapid runoff and evaporation. Within an average horizontal distance of 24 km (15 miles) between mountains on the east and west, no internal barriers are present to inhibit travel, communication, and exchange.
Well-defined local sequences of decorated pottery, paralleling the ceramic sequence of the Phoenix core, have been established for few other Hohokam regions. The Tucson Basin is the major exception (Kelly 1978). The earliest Hohokam decorated pottery is rare and appears to be imported from the core among assemblages of Tucson plain and red-slipped wares. By the beginning of the eighth century A.D., however, a coherent regional style is evidenced by a series of local red-on-brown designs exhibiting broadly similar trends with red-on-buff decoration of the Phoenix Basin. In the early 1980s, recorded distributions of ballcourt and mound sites were sufficient to suggest substantial levels of prehistoric population and community organization of settlement throughout the Tucson Basin.
Over its long history, the city of Tucson has erased Hohokam remains in a nucleus of most desirable land along the Santa Cruz River and successive outer zones, with few records of even the largest former sites. As in the prehistoric era, however, population and agriculture are of lesser magnitude than in the Salt and Gila valleys. North of Tucson, historic settlement has been late (mainly after 1915) and clustered along a few roads. The dominant economic activity has been cattle raising. Modern agriculture is confined to localized and fragmented strips along the Santa Cruz River and one tributary. These conditions are in marked contrast to the Phoenix Basin where greater urban sprawl and highly mechanized irrigation in leveled fields have obliterated major portions of regional settlement pattern. The northern Tucson Basin promised preservation of a full range of surface remains in environmental settings representative of the basin as a whole.
Optimal conditions for recovering comprehensive settlement patterns and defining territorial organization are found in a broad trans-bajada area near the town of Marana, Arizona (Figs. 1.1, 1.8). A localized high water table on the Santa Cruz floodplain offers the greatest potential for riverine irrigation in the northern Tucson Basin. Elevational diversity is repeated between the river and the low volcanic Tucson Mountains on the west and the more massive Tortolita Mountains on the east.
Diachronic trends in agricultural production and territorial organization throughout the Hohokam sequence can be monitored in this Marana study area. Agricultural occupations beginning with Late Archaic cultivators are continuous until the late Classic period. Population, as measured by settlement, peaks during the early Classic period in tandem with the most expansive and intensive land use. The intersection of societal organization and agricultural production can be examined for technological combinations typifying Hohokam regions outside the core: irrigation on a smaller scale, diversions of short-term flow in ephemeral drainages or floodwater farming, and techniques for capturing overland runoff.
The trappings of community organization are reflected in a platform mound at a preeminent site within this study area. Enclosing this center is a virtually intact framework of concentrated contemporary settlement, surrounded by markedly lower site densities and areas with scarce remains (see Fig. 3.2). Community units farther north, also surveyed in this study, offer comparative territorial configurations in somewhat different settings (Fig. 1.8). The mound center in the Robles Community is 19 km (12 miles) northwest and the center of the McClellan Community is 27 km (17 miles) north.
Because late Classic settlement is largely absent in the Marana study area, distributional evidence for agricultural production and settlement organization in the early Classic period community is not obscured by subsequent prehistoric occupations. The abundant decorated pottery of the early Classic period Tanque Verde phase provides a reliable diagnostic for identifying even small community habitations and activity loci. Unlike the majority of large Hohokam sites with public architecture, the Marana Mound Site has been spared the ravages of serious pothunting. For understanding synchronic relationships in a noncore community, the early Classic configuration in the Marana study area comes as close to an ideal slice of time as can be achieved with Hohokam chronology and settlement.