Evidence for Large-scale Agave Cultivation
in the Marana Community

Suzanne K. Fish, Paul R. Fish, and John H. Madsen

Linkage of agave cultivation with a farming technology represented by widespread remains in the Marana Community has been one of the significant consequences of Northern Tucson Basin Survey research (S. Fish, R Fish, Miksicek, and Madsen 1985; S. Fish, R Fish, and Madsen 1990a), illuminating a new dimension of prehistoric agriculture in the Sonoran Desert. Fields marked by rockpiles and low stone alignments cover many hundreds of hectares. Interdisciplinary study of these prehistoric agricultural complexes has detailed the nature and extent of agave cultivation during the later portions of the Hohokam sequence.


Agave or century plant species have been a source of food and fiber for most aboriginal groups of North America living within the distributional range of these drought-adapted perennial succulents. Stiff, spiny leaves pointing upward in a rosette contain fibers used for cordage, sandals, brushes, textiles, and other items. During its lifetime of from 5 to 20 years, the agave stores carbohydrates in its tissues to fuel a single and final flowering episode in which a tall, blossom-bearing stalk emerges (Fig. 7. 1). When agave is harvested prior to energy expenditure in flowering, stored carbohydrates are converted by roasting to a sugary, high calorie, and nutritious food (Fig. 7.2). Large Mesoamerican species were utilized in a somewhat different manner by extraction, just prior to flowering, of copious, sugary sap used primarily in historic times as a fermented base for pulque, mescal, and tequila. Unlike annual crops such as corn, beans, and squash, cultivated agaves exhibit no clear morphological markers that can be used to distinguish archaeological remains of crops from wild, gathered plants.

The importance of agave as a gathered resource among Southwestern Indians is well attested in the ethnographic literature (Castetter and others 1938; Pennington 1963; Felger and Moser 1985). Processed products were circulated widely among groups living at a distance from natural stands. Spanish documentary references to agave "plantations" have been interpreted as evidence of cultivation in Baja California at the time of first Jesuit contact (Castetter and others 1938: 50; Nabhan 1985) and in northern Sinaloa (Perez de Ribas 1968). However, cultivation that has been historically documented in the southwestern United States and adjacent northwestern Mexico is relatively recent, limited to small-scale plantings, and has not established a role for agave as a major aboriginal cultigen.

The apparent absence of cultivated agave among northerly indigenous peoples of the Sonoran Desert and its limited role in the agriculture of southern groups contrasts with the pre-Columbian and historic ubiquity of this crop farther south. Agave is a major cultigen throughout the rest of highland Mexico, including portions of Durango and Zacatecas often considered within the greater Southwestern cultural sphere. The potentially active role of Southwestern Indians in spreading indigenous species beyond their natural ranges was included by Richard Ford (1981: 21) as an issue deserving attention in the definition of an agricultural complex distinctive to the American Southwest.

Agave species of the Sonoran Desert grow mainly on rocky slopes of hills and mountains and are lacking in valleys (Gentry 1972: 1). Many records for the Pimans, or O'odham describe acquisition through the harvesting of wild stands in such topographic situations and often at considerable distances from home bases (Castetter and others 1938: 50; Curtin 1981: 48). Upland gathering has been commonly assumed to account for charred agave long known to occur in Hohokam roasting pits at lower elevation habitation sites away from natural populations (Haury 1945: 39; Fewkes 1912; Hayden 1957: 103).

Botanical attributes and distributional associations of Agave parryi Englem, Agave murpheyi E. Gibson, and other species recovered from archaeological sites have

Figure 7. 1. Agave murpheyi with flowering stalk. (Photograph by Wendy Hodgson.)

been cited as suggesting potential aboriginal transport or manipulation of indigenous species (Minnis and Plog 1976; Crosswhite 1981: 58-59; Ford 1981; Gentry 1972, 1982). More recently, extensive flotation analyses for charred plant materials at Hohokam residential sites in the Tucson Basin (Miksicek 1988) and the Phoenix Basin (Miksicek 1984; Bohrer 1987; Gasser 1988a, 1988b; Mitchell 1989) have yielded impressive amounts of agave. Locations of these sites coincide poorly with natural distributions in southern Arizona (Fig. 7.3), yet in each case, charred agave occurs in substantial amounts and is often among the more common kinds of botanical material. Prehistoric cultivation rather than acquisition of gathered plant products from uplands has been proposed on the basis of overall quantity and variety of plant parts (Miksicek 1984; Gasser and Miksicek 1985). Ongoing studies in the Marana Community have demonstrated the cultivation of agave by identifying replicated associations of agricultural features, specialized artifacts, field-side

Figure 7.2. Roasted hearts or heads of agave, prepared by
removal of leaves and extended baking in a pit.
(Photograph by Charles H. Miksicek.)

processing facilities, and charred agave (P. Fish and others 1984; S. Fish, P. Fish, Miksicek, and Madsen 1985; S. Fish, P. Fish, and Madsen 1985, 1990a). Complexes of these agricultural remains document the productive capacity for supplying this abundantly utilized resource.


In the northern Tucson Basin, as in most other regions, all facets of prehistoric agricultural activity have not left equally tangible or easily retrieved records. Santa Cruz floodplain canals are buried by active alluvial and colluvial deposition. Surface features are also seldom preserved on arable bottomlands of larger tributaries. Some Piman cultivation methods with probable Hohokam antecedents involve brush diversions, low earthen ridges, and short ditches. Such ephemeral structures

Figure 7.3. A comparison of the modern range of agave species with
evidence for prehistoric production and use in southern Arizona.

Figure 7.4. Typical rockpile feature in the Marana survey area.

often are obliterated in single planting seasons and are unlikely to persist archaeologically. Farming that employed ephemeral constructions can only be inferred by reference to ethnographic practices in similar situations of settlement location and topographic opportunity (Chapter 4).

With the exception of floodplains and portions of some alluvial fans, however, stone surface features related to agriculture are relatively well preserved throughout the Marana area. Most are small devices, informally constructed from unshaped local rock. These kinds of remains are readily obliterated by later prehistoric or modern surface disturbance and would be spottily preserved in areas experiencing pervasive historic development. Distributional characteristics hinder systematic recording even where preservation is optimal. The features are areally diffuse, but may be so widespread in the aggregate that boundary definition for large complexes presents problems. Associated artifacts are usually sparse.

The ubiquitous unit of stone features and other associated remains has been termed a "rockpile field," after its most frequent feature. Rockpiles or rounded heaps were constructed from cobbles scattered on the bajadas (Fig. 7.4). Excavated cross sections of undeflated piles reveal that cobbles often cap smaller mounds of soil (Fig. 7.5). Rockpile size is variable but rarely exceeds 1.5 m in diameter and 75 cm in height. Short, linear features of one to several cobble courses that served as contour terraces (Fig. 7.6) and checkdams are often interspersed in small fields and are always present in large complexes.

Most of these rockpile fields are located in Zones 2 and 4 of the Marana Community (Fig. 3.9). Zone 2, the focus of this chapter, occupies middle elevations of the eastern bajada and was rarely chosen for residence. More than 485 hectares (more than 2 square miles) of large, 10-ha to 50-ha fields without habitation have been located here. Segments of three typical fields are mapped in Figures 7.7, 7.8, and 7.9. Rockpiles, terraces, and checkdams occur in substantial numbers in Zone 4 in

Figure 7.5. Cross section of an excavated rockpile.

Figure 7.6. Typical contour terrace in the Marana survey area.

Figure 7.7. Distribution of features in a portion of site AZ AA:12:108,
a large rockpile field in Zone 2 of the Marana Community.

conjunction with large sites, small sites, and isolated structures, but never independent of habitation as in Zone 2 and never of an areal extent comparable to the larger fields below (Fig. 3.9).

Rockpiles and Planting

The presence of rockpile complexes has not been recorded routinely by a majority of archaeologists; the fact that instances have been reported throughout Hohokam territory (Fig. 7.3; S. Fish, P Fish, Miksicek, and Madsen 1985) is a reflection of their ubiquity. However, these features received little directed study until recently. Multiple lines of evidence now indicate that the rockpiles as well as the terraces and checkdams, are facilities of agricultural production. Doyel (1984: 43) has suggested that heaped cobbles in the New River area north of Phoenix originated as residuals of surface clearing, but this seems an improbable explanation for a majority of rockpile occurrences. Rockpiles are sometimes located in the midst of dense concentrations of surface rock and occasionally are found in alluvial situations necessitating

Figure 7.8. Distribution of features in a portion of site AZ AA: 12:470,
a large rockpile field in Zone 2 of the Marana Community.

importation of rock for construction. A function of clearing the surface to increase runoff onto more arable land (Evenari and others 1971; Irwin-Williams 1986) also seems unlikely for the Tucson Basin and most other Hohokam locations. Tucson rockpiles on bajadas are topographically situated so as to receive optimal runoff from ridge catchments; diversion devices for directing water to more suitable plots are not present in any field.

Rockpiles enhance the planting environment, as do the more obviously agricultural terraces and checkdams with which they occur. The uneven, porous surface of the piles allows penetration of rainfall and runoff in contrast to surrounding hard-packed and impermeable ground surfaces. The rocks then act as a mulch, slowing evaporation of soil moisture by blocking capillary action and preserving higher moisture levels beneath. This mulching effect of rocks in desert soils has been measured experimentally in Israel (Evenari and others 1971) and by the authors in gauged soil-moisture experiments now underway.

The response of plants today to the microhabitat of rockpiles is demonstrated by the relatively dense seasonal concentrations of annuals, the distribution of perennials, and the presence of lichens and mosses that require

Figure 7.9. Distribution of features at site AZ AA. 12:469, one of the many small groups
of rockpiles interspersed among the large fields in Zone 2 of the Marana Community.

substantial moisture. Modern plant response has been quantified by comparing root biomass in soil directly beneath rockpiles, and in adjacent control areas. Root weight in rockpile soil is an average 80 percent higher than in the controls (S. Fish, R Fish, Miksicek, and Madsen 1985: 109, Fig. 4). These observations further strengthen an interpretation of rockpiles as moisture-enhancing facilities for crop plants.

An additional benefit of rockpiles became apparent from experimental plantings. Rockpiles provide significant protection from rodent predation that damages leaves but is concentrated on stem bases and roots. Animals cannot as easily gain access to these parts by digging beneath the plants in rockpiles as is possible with agaves growing directly in soil. The practices of Mayan traditional farmers in Yucatan (Caballero 1986) provide parallels in the use of piled rocks for protecting young sabal palms. Gary Nabhan, of the Phoenix Botanical Garden and Native Seed Search, has observed similar practices by the Lower Pima of Sonora to protect cucurbits.

Figure 7.10. Three basin cross sections from north (top) to south (bottom) in the
Marana Commuinity, showing elevational limits of rockpile fields in Zone 2.

Zone 2 Distributions

Although earlier dates for small rockpile loci have been indicated by ceramic associations and radiocarbon determinations (Chapter 8), fields in the large category (10 ha to 50 ha) belong exclusively to the early Classic period. These mid-bajada rockpile fields are homogenous in physical and locational attributes. All are situated on ridges of dissected Pleistocene alluvial fans. Caliche or calcium carbonate layers occur consistently between 20 cm and 40 cm below the surface and form an impermeable barrier that traps moisture in the upper soil within reach of shallow-growing agave roots. Covering a total of about 5 square kilometers (2 square miles), community fields occur within a band less than 2.5 km (1.5 miles) wide and 20 km (12 miles) long. All are positioned between elevations of 625 m and 670 m (2030 and 2200 feet) above sea level (Figs. 3.9, 7.10).

Most of these agricultural complexes derive a linear outline from their locations along ridge tops paralleling trans-bajada drainages. Cobble features are infrequent on bottomlands of secondary drainages between ridges. At least some of these intervening drainage bottoms were farmed with the aid of earthen checkdams, as attested by occasionally preserved berms. Corn pollen was recovered from the upstream planting surface behind one such checkdam during berm excavation at site AZ AA: 12:470, a large rockpile field to the east and upslope of the Marana Mound Site.

No indications of field house structures or concentrated domestic refuse occur within Zone 2 fields. Roasting pits filled with ash and fire-cracked rock are found in many fields, and multiple pits usually occur in the largest complexes. Stone artifact assemblages are marked by steep-edged core tools and thin, knifelike implements made from tabular stone with flaked and ground edges. Field size, proportions of agave-related implements in stone tool assemblages, and sizes of roasting pits indicate a production emphasis in mid-bajada fields that is not duplicated in other zones of the Marana Community.

Figure 7.11. Typical large roasting pit located in a large
rockpile field in Zone 2 of the Marana Community.

Roasting Pits

Rockpile agricultural complexes are most directly linked to agave by roasting facilities. Pit-roasting is the common method of preparing agave among historic Southwestern Indians (Castetter and others 1938). In some cases, pits were dug relatively deeply into the ground and in other cases, agave was roasted in shallow troughs with earth heaped above. Extended cooking up to 48 hours in sealed pits prepares the nutritious heart, inflorescence, and leaf bases for eating. Briefer, controlled roasting of leaves is also a process that may be used to aid in the removal of leaf fibers for use in cordage and crafts. Rocks lining pits or mixed in the fill retain heat and facilitate the roasting process. Indians of historic times often reused both pits and rocks.

Roasting pits are consistently found in Marana Zone 2 fields, often in loose, easily excavated sediments of shallow and ephemeral drainages. They range from several meters up to 50 m in the longest dimension and average 1.5 m in depth (Fig. 7.11). Pit size tends to correlate with field size. Discrete pit shapes from individual roasting episodes cannot be easily distinguished in the large features among the intrusions and accretions of seasonal reuse over many years (Fig. 7. 11). Pits are filled with ash, charcoal, fire-cracked rock, and occasional artifacts. Large sherds (discussed further in Chapter 8) appear to have been used to protect food items from scorching during roasting and as scoops in the excavation of pit fill. Whole and partial examples of such formal and informally shaped sherd artifacts with worn edges have been recovered from large mid-bajada pits.

Botanical Evidence

Flotation of more than 600 liters of fill from one or more excavated roasting pits in 16 fields (S. Fish, P. Fish, and Madsen 1990a; Miksicek 1988) has produced charred

Figure 7.12. Agave knife from the Marana Community.

specimens of agave from each. Fragmentary fibers are most abundant, but recovered plant parts include marginal teeth, terminal spines, leaf bases, and heart (compound stem) fragments. Plentiful wood charcoal, dominated by mesquite and ironwood, has been identified; fuel sources reflect an environment similar to the present and trees readily available in adjoining drainages.

The identity and number of cultivated agave species is currently unresolved. Variation in remains from Marana roasting pits suggests more than one cultivar (S. Fish, P. Fish, Miksicek, and Madsen 1985). Floral and fruit parts for taxonomic determination are unavailable in roasting pits and other site contexts. Arizona species may have been transplanted, particularly Agave murpheyi (following Gentry 1972; Crosswhite 1981; S. Fish and Nabhan 1991; Hodgson and others 1989). However, as with most other Southwestern aboriginal crops, cultivars may have included varieties of ultimate Mesoamerican origin.

Inflorescenses, or flower stalks were gathered from natural agave stands by Southwestern groups. However, by the time the inflorescence had emerged, a majority of stored energy and moisture had been consumed by its growth, leaving a much reduced potential for edible use of the heart. Groups who acquired agave solely by gathering often prevented stalk development in maturing plants intended for future harvest by inhibiting or destroying the precursor tissues (Castetter and others 1938). Cultivators are typically even more rigorous than gatherers in this kind of plant management (Parsons and Parsons 1990). Occasionally the Hohokam may have inadvertently or intentionally allowed flower stalk growth to proceed. Charred stalk remains are rare, but have been recovered in an excavated Marana site in Zone 1 on the lower bajada where they appear to have been used as roofing materials (Miksicek 1987).

Agave pollen is seldom recovered from Hohokam sites, in keeping with the presumed rarity of flowering. It has not been identified in sediments from Marana roasting pits or in agricultural features. As with charred stalk fragments, the infrequent occurrences of pollen in residential sites (Spaulding 1974; S. Fish 1987b, 1989) may represent accidental or intentional instances of inflorescence development for special purposes, or cases in which wild agave products were obtained.

Field Artifact Assemblages

Artifacts recovered in systematic surface collections from the rockpile fields strengthen the inference of agave cultivation. Broad, flat stone tools made on raw materials with naturally tabular fracture (Fig. 7.12) are prominent in field assemblages. The common term in the ethnographic and archaeological literature for this tool type is agave or mescal knife. Varying in outline from rectangular to rounded, such specialized implements were used historically by Southwestern groups to sever agave leaves from the hearts (Castetter and others 1938). Supporting this analogy, calcium oxalate crystals like those present in agave tissues have been observed in microscopic examinations of Hohokam knife surfaces (Bernard-Shaw 1990b). In collections from Marana fields, 9 percent of more than 400 stone artifacts, including flaking debris, and 19.2 percent of retouched tools are agave knives and fragments. Knives are not concentrated near the roasting pits or in any other sector, but are widely scattered throughout the fields.

Figure 7.13. Pulping planes from the Marana Community.

A second distinctive and numerous artifact type is a steep-edged core tool often designated a scraping or pulping plane (Fig. 7.13). Such tools have been associated experimentally, and in ethnographic reports, with fiber removal from agave leaves (Salls 1985; Rogers 1939: 51-53; Osborne 1965: 47-49; Kowta 1969: 52-69; Hester and Heizer 1972: 109-110; Bernard-Shaw 1990b). Experiments conducted in the course of this study have

Figure 7.14. Ground stone object of uncertain
function from a rockpile field in Zone 2 of the
Marana Community. (Height is 23 cm.)

shown pulping planes to be efficient in stripping marginal spines from leaves. Precautionary removal of spines from agave leaves and hearts before transport to processing facilities has been observed among modern Otomi cultivators (Parsons and Parsons 1990). As with agave knives, pulping planes are scattered throughout the fields. Together these two artifact types represent 53.9 percent of retouched tools, an assemblage emphasis unique among sites in the Marana survey area.

One complete and two fragmentary ground stone objects of uncertain function were collected from Marana rockpile complexes. These are formally shaped artifacts resembling a "T" in outline. The complete specimen (Fig. 7.14) is approximately 64 cm by 42 cm in maximum dimensions, 11 cm thick, and weighs 12.8 kg (28 pounds). Additional examples have been found in rockpile fields elsewhere (for example, Debowski and others 1976). In a discussion of the distribution and significance of such objects, Alan Ferg (1986) proposed a possible symbolic or ceremonial function. An alternative possibility based on field contexts might be a functional association with agave processing, although detectable use-wear is lacking on the Marana specimens.

Cultivation in Rockpile Fields

Stands of wild agave are absent on northern Tucson Basin bajadas. Within a 100-km (62-mile) radius of the Marana Community, all species but one occur well above community boundaries at elevations greater than 925 m (3035 feet; Gentry 1982). The exception, Agave murpheyi, has been collected at lower elevations only under Tohono O'odham cultivation. It is unlikely that naturally occurring prehistoric stands at lower elevations have been extirpated historically. Agave persists widely in other habitats under postcontact conditions such as cattle grazing; natural populations depleted by Hohokam overuse should have had ample opportunity to recover in the intervening 500 to 600 years. Furthermore, according to Robert McDaniel, of the Department of Plant Sciences at the University of Arizona, experiments in commercial agave production at the nearby University of Arizona Marana Farms have shown that reproduction by seed for the most likely species is inhibited by damage to seedlings from the intense summer heat and sun at lower bajada elevations. Successful reproduction appears possible only through cloning by previously established plants or by human transplantation of clones or offsets beyond the seedling state of development. This latter method is the universal means of planting by agave cultivators in historic and modern times, even in environments where heat presents no danger to seedlings.

Mexican agave cultivation of the magnitude suggested by Tucson Zone 2 fields typically involves similar basin slopes (K. Johnson 1977; Parsons and Parsons 1990; West 1948, 1968). Agave is considered a crop primarily for planting on land of secondary agricultural value that cannot be irrigated. On sloping terrain agaves are used to stabilize rock features such as terraces and to trap soil and surface runoff. Adapted to and conditions, agaves survive droughts and spotty rainfall more successfully than annual crops, producing reliable long-term yields. Annual crops requiring more predictable moisture may be interspersed in the better watered portions of fields or in years when rainfall is judged to be particularly promising.

Intercropping in Marana fields on a minor scale seems indicated by a few instances of corn pollen from soil samples in and around several rockpiles. This pollen also might have been generated by nearby plantings in earthen checkdams across adjoining small drainages. Corn and cotton pollen have been identified at Hohokam rockpile loci elsewhere, such as along the Gila River in situations where floodwater diversion onto the fields may have been possible (S. Fish 1984b).

The combined evidence from Marana fields presents a strong case for large-scale agave cultivation. Roasting pits that are separated from habitations are localized in or near field areas. All excavated pits have contained agave. Both the pits and the associated distinctive stone tools fit Southwestern ethnographic correlates for harvesting and processing agave. Knives and pulping planes are distributed throughout fields rather than clustered about roasting pits or in other circumscribed locations. Finally, the predominant topographic situation of rockpile fields closely resembles dry basin slopes extensively planted with agave in highland Mexico.


Construction Labor

The fact that agricultural features in Marana fields persist intact and apparently serviceable until today suggests simultaneous cultivation of the great majority of complexes during at least the latter part of community occupation in the early Classic period. Only a few small fields appear to date to Preclassic times, as indicated by surface ceramics and sherds in excavated roasting pits. Numbers of rockpiles and meters of cobble alignments were tabulated in multiple 50-m squares within three large fields in order to characterize feature densities. Averaged figures were then used to calculate feature totals for the area covered by all large mid-bajada fields. Estimates of 42,000 rockpiles and 120,000 m of alignments were projected. Experimental rockpile construction gave an approximate range of 20 minutes to one hour, depending on size and availability of cobbles. A 40-minute mean per rockpile results in an initial construction effort of 28,000 person-hours or 14 person-years with 5-day work weeks. Experimental building of comparable cobble terrace alignments produced a figure of 1.65 m per hour (P. Fish and S. Fish 1984: 156-157). An initial investment of 72,500 person-hours or slightly more than 36 person-years would have been required to construct the terraces and checkdams in the Zone 2 fields.

A total of more than 100,000 person-hours or 50 person-years are estimated to have been expended by farmers in constructing cobble features prior to the demands of planting, tending, harvesting, and processing crops. Any perishable components of the system and maintenance of field facilities would have been additional. Intriguing but unanswered questions concerning labor investment include whether the large fields were completed as unitary construction events or by accretion over a period of continued use and to what degree field layout may have been a centrally coordinated effort.

Agave Yields

In Mexico, agaves often line terrace walls for soil stabilization and runoff control, and a similar planting pattern seems likely for rockpiles. Quantities of these features are therefore used to estimate potential agave yield. As a minimum estimate, a single plant is posited per rockpile, and one for every 2 m of alignment. According to this formula, 102,000 plants could have been growing at one time in large mid-bajada fields of the Marana Community. Harvest comes at plant maturity, for which an average 10-year figure is used. Thus only one-tenth of the plantings, or 10,200 agaves, would have been available in a given year.

Yield in food and fiber is based on relatively small Southwestern species best suited to lower elevations. Because species identification from charred macrofossils is not yet possible, per plant estimates would be distinctly low should it be determined in the future that large agaves of Mesoamerican origin were also utilized by the Hohokam. Yields for two end products are calculated independently; use of the edible heart should not have precluded fiber extraction from the leaves of the same plant.

Agaves harvested in a single year could have supplied approximately 365 grams of fiber per plant (S. Fish, P. Fish, Miksicek, and Madsen 1985) or a total for the Marana fields of 3.72 metric tons. At an average of 4 kg, agave hearts could have produced 40.8 metric tons of edible product. Hearts furnish 347 calories and 4.5 g of protein per 100 g (Ross 1944). According to FAO recommendations (FAO/WHO 1973), agaves in Zone 2 fields would have supplied the equivalent of annual caloric requirements for 155 individuals and protein requirements for 110. In an alternative projection of nutritional role more in keeping with probable prehistoric diet, 20 percent of caloric needs for 775 persons and protein for 550 could have been met by agave. Although specific variables in estimates are subject to refinement or change, it is clear that these fields could have supplied a significant increment in Classic period Hohokam diet and economy.

The ratios of plants to features used in these calculations are likely to produce a low estimate. Particularly for moderate-sized species, individual rockpiles and 2-m intervals of linear features might have supported several plants at once; agaves produce offsets in natural habitats and tend to form clusters. Superfluous offsets tapping mother plants may have been removed to concentrate nutrients and water for rapid maturity and harvest. However, if two or three plants could mature in tandem or rapid succession, the above yield estimates would be superseded by several magnitudes. Results of experimental plantings suggest that multiple plants per feature would have appeared naturally by cloning and that retaining more than one would have offered a strategy for achieving higher long-term production rather than the most rapid returns.


The regional distribution of rockpile fields has implications for the organization of labor and land tenure. Large complexes may have necessitated communal labor and consensual recognition of intrafield boundaries by many cultivators. In one large field, a small boulder with petroglyphs resembles field boundary markers of the Hopi and Zuni (Forde 1931: 233, 235; Cushing 1920: 153). As among many Southwestern native peoples, kin-based or other corporate groups may have controlled arable land, with use-rights for individuals or households assigned to particular plots. The patterning of fields with respect to habitation also can be compared by time and setting. Small fields on the upper bajada occur in conjunction with habitation. Only fields on the middle bajada are situated at a distance from residential sites, indicating a system of agricultural tenure not based on immediate proximity.

Large fields undoubtedly drew upon the labor of many cultivators, whether solely through individual incentive or more integrated planning. Indicative of communally scheduled efforts in the complexes are the huge roasting facilities. Although features of large size show repetitive use from year to year, excavation produced little evidence of discrete small firings. It appears that numbers of cultivators roasted their harvests together. Communal events would have been an efficient use of woody fuel in a desert environment, but would have entailed coordination of fuel acquisition, harvest, and preroasting preparation of the plants. By contrast, farmers using the modest pits of smaller fields could process their harvest according to individual convenience.

In an environment where aridity circumscribes agricultural activity, opportunities to expand production are limited. Agave cultivation on marginal bajada slopes would have offered an optimal solution to restricted sources of supplemental water. These plants are adapted to low and unreliable moisture to a greater degree than annual crops. More poorly watered and previously uncultivated land could therefore be used to satisfy expanding needs for foodstuffs. In addition, fibers could support craft manufacture and furnish highly portable raw materials and finished products for external trade.

Cultivation of mid-bajada areas with sparse prior use represents an and land version of intensification(Boserup 1965), through expansion of extensive land use practices rather than more frequent cropping of land already under cultivation. Intensification is also indicated by an increased labor cost in travel time per unit of production (G. Johnson 1977: 490). Farmers of the mid-bajada complexes had to invest more time going to and from their fields than did farmers whose fields and habitations were adjacent. It is not surprising, then, that large-scale cultivation was concentrated in those community sectors where denser populations coincided with poorer access to irrigable land or floodwater opportunities.

Potential investment in rockpile fields on mid-bajadas was virtually unlimited. Combinations of topographic and hydrologic variables similar to those of the largest fields are duplicated widely in the Classic period Marana Community. The location and extent of rockpile fields can best be understood as the outcome of investment decisions that took into account other productive needs and alternatives.

The largest fields and the greatest total acreage occur above the Zone 1 lower bajada sites with the highest population densities and poorest potential for floodwater farming (Figs. 3.2, 3.9). These fields are located at a distance from habitations and in a slightly drier portion of the bajada than areas to the south. Even though higher labor input per unit return was likely entailed than in fields near residences, expansion of bajada cultivation must have been economically advantageous under local circumstances. To the south in Zone 1, less dense populations with more favorable floodwater situations constructed fewer and mostly small rockpile fields. Farther south still, fields are absent on bajadas opposite the numerous settlements along the river at the Tucson Mountains. The agricultural labor of these river dwellers evidently could be expended more effectively in irrigated farming.

Recognition of the mid-bajada field configuration as a temporal innovation rests on access to comprehensive regional data sets. Such information confirms that the technology and the crop were not new. Only the locational concentration, the remarkable size of individual complexes, the expansion of total acreage in rockpile fields, and a concomitant emphasis on their yield are unique to the early part of the Classic period. Productive strategies at that time seem best understood in light of the demographic conditions described in Chapter 3. Agave production in mid-bajada fields represents an economic reformulation of agricultural effort by particular segments of the community and a form of intensification in response to increasing population and demand.

Next: Chapter 8 | Previous: Chapter 6

UAP Online Books  |   UAP Home Page

The University of Arizona Press, 2/10/01 5:03PM