PECOS From Folsom to Fogelson: The Cultural Resources Inventory Survey of Pecos National Historical Park

Pecos Pueblo and the Upper Pecos Valley have played a long and significant role in the prehistory, history, and archeology of the American Southwest (Figure 1.1). The prominence of Pecos is reflected both by the volume of the historical and archeological literature about it and by the stature of the scholars drawn here (Bancroft 1962; Bandelier 1881, 1892; Bolton 1949; Cordell 1998; Hayes 1974; Hewett 1904; Kessell 1979; Kidder 1924, 1932, 1958; Kidder and Amsden 1931; Kidder and Shepard 1936; Schroeder and Matson 1965; Winship 1896).

Figure 1.1. North-central New Mexico with the location of Pecos National Historical Park. (click on image for an enlargement in a new window)

Historians and archeologists alike have focused their efforts to a large degree on Pecos Pueblo and the nearby Spanish Mission church and convento. Kidder conducted a variety of excavations at nearby pueblos, and Cordell (1998) has recently completed a long-term study of Rowe Pueblo, but if anything these exceptions emphasize the scarcity of archeological investigation of the Upper Pecos Valley surrounding Pecos Pueblo.

This volume reports the results of a 100 percent inventory survey of the nearly 2,396 hectares (5,920 acres) of the main unit of Pecos National Historical Park. It provides the first systematic analysis of the prehistoric and historic archeology of a large area surrounding Pecos Pueblo. Our primary goal is to describe the archeological materials found, but we also examine two aspects of prehistoric Puebloan society in the Upper Pecos Valley: (1) the development of the Pecos community and (2) exchange of material goods between Pecos and other Rio Grande Pueblos, and between Pecos and a variety of nomadic Plains groups. The emphasis of this volume is on the prehistoric Puebloan occupation of the park (A. D. 1200-1600). Sites of this affiliation are the most numerous in the project database. The historic portion of this report places the nineteenth and twentieth century Euro-American sites recorded by the survey into chronological perspective, examines economic and social change through census records, and describes and places into context the Santa Fe Trail and other circulation features within the park. This survey did not record the Spanish Mission complex, and very few sites were located that could be assigned to the Spanish Mission/Colonial period. Information for these structures and for the entrada and Colonial periods can be found in the recent NPS Draft Historic Structures Report (Ivey 1996).

The following sections of this chapter are intended to provide the background and context for the study. We begin by describing the need for an archeological inventory of Pecos National Historical Park. This is succeeded by descriptive summaries of the culture history and modern environment of the Upper Pecos Valley. These summaries lay the groundwork for a discussion of the project research objectives. The chapter closes with brief descriptions of the analyses performed in the following chapters.

The Need for an Archeological Inventory

It may seem obvious that the archeological resources of a national historical park as significant as Pecos should be carefully inventoried, but the need for inventory is based not only on logic but on federal law as well. Section 110(a)(2) of the National Historic Preservation Act of 1966 (as amended in 1980) directs federal agencies to "establish a program to locate, inventory, and nominate to the Secretary [of the Interior] all properties under the agency's ownership or control . . . that appear to qualify for inclusion on the National Register." An archeological inventory survey is usually necessary to identify such properties or sites.

This mandate to inventory has been strengthened recently by the National Park Service's Systemwide Archeological Inventory Program (SAIP), the purpose of which is "to conduct systematic, scientific research to locate, evaluate, and document archeological resources under National Park Service stewardship" (National Park Service 1992b:i).

While law and policy dictate that inventories be performed, almost all inventories are conducted for very specific reasons. At Pecos, impetus for an archeological survey was provided by the legislated expansion of the park in 1990, which increased the park's size by nearly 2,527 hectares (6,244 acres). The original monument, or "core area" consisting of 148 hectares (365 acres) surrounding Pecos Pueblo and the Spanish Colonial church and convento, was expanded by 2,248 hectares (5,555 acres) to include the surrounding Forked Lightning Ranch. In addition, two detached units, totaling nearly 279 hectares (689 acres), were added to preserve the Pigeon's Ranch and Cañoncito portions of the Glorieta National Battlefield (Figure 1.2).

Figure 1.2. Boundary history of Pecos National Historical Park. (click on image for an enlargement in a new window)

Although the archeology of the original monument was relatively well known as the result of previous inventory efforts (Nordby 1992), knowledge of the nature and condition of archeological resources in the new areas was poor at best. And, even within the core monument, the results of previous surveys were unreported. Because many areas were unsurveyed, others were unreported, and information often differed in quality and consistency between areas, it became apparent that the most effective and efficient means of establishing an inventory database suitable for both management and research purposes was to inventory the entire park, including those areas previously surveyed.

The Pecos Cultural Resources Inventory Survey, referred to hereafter as the Pecos CRIS, was designed to address both management and research requirements, although the present volume focuses on the research results of the survey. The results reported here are the product of the first systematic and detailed inventory of the full range of the park's cultural resources. Design of the inventory was undertaken in 1994-1995. Fieldwork was carried out during the summers of 1995, 1996, and 1997, resulting in the identification of 629 sites within 2,396 hectares (5,920 acres). At the time of fieldwork, most of the land in the Glorieta Battlefield units was in private ownership so these units were not surveyed. Complete inventory of the Glorieta units remains a goal for the future. Two additional years (1998 and 1999) of data analysis and research produced this volume.

Upper Pecos Valley Culture History

The following discussion of the culture history of Pecos National Historical Park and the Upper Pecos Valley utilizes the Rio Grande chronological sequence developed by Wendorf and Reed (1955). This chronology is based on changes in settlement patterns, architecture, and artifacts. For this reason it differs from the ceramic chronology for the region, which uses only changes in form, composition, and decoration of ceramic vessels to define different periods. Using the Wendorf and Reed chronology helps place the culture history of Pecos in the broader context of the northern Rio Grande region. The ceramic chronology for the prehistoric sites recorded by the project is revisited in Chapter 4 of this volume. The review of the historic period culture history is organized using the periods and themes developed for the NPS Cultural Landscape Overview of the park (Cowley et al. 1998). These two chronological constructions overlap during the early historic era (A.D. 1540-1600).

Preceramic Period (11,500 B.C.-A.D. 600)

Paleoindians were the first human occupants of the Southwest, and evidence of their presence in the Upper Pecos Valley is scant. Paleoindian economy focused on big-game hunting, targeting now-extinct large fauna, such as mastodon and Bison antiquus. Most Paleoindian sites in the Southwest are found around ancient lakebeds where water and game were plentiful (Judge 1973:330). Sites are generally open-air lithic scatters, recognized most easily by the presence of distinctive, well-formed projectile point types. These spear points are remarkable for their consistency in shape and size across a broad geographical area and for the high quality of the knapping techniques used to create them.

The elevation and relatively small size of the Upper Pecos Valley may have made it a marginal habitat for many species of large fauna, and hence for the Paleoindian groups dependent on them. One Clovis-like point (11,000-10,000 B.C.) was previously found in the nearby Sangre de Cristo Mountains (Nordby 1990:33; Wendorf and Miller 1959:48). Evidence for Paleoindian use of Pecos National Historical Park is limited to a single isolated Folsom point midsection (9,000-8,000 B.C.) found in the Tecolote Mountains during the present survey.

It is possible that some of the nondiagnostic lithic scatters in the valley are the result of Paleoindian activities, but until more detailed technological analyses of the lithic debitage are performed, the cultural affiliation of these assemblages will remain uncertain. The lack of Paleoindian sites also may be the product of differential exposure, as the severe erosion that has exposed many such finds in surrounding areas is not widespread in the Upper Pecos Valley (Cordell 1998).

The Archaic culture in the Southwest is defined by a broad subsistence regime including hunting of a variety of medium to small game as well as plant gathering. During the late Archaic period some groups began experimenting with horticulture (Irwin-Williams 1973), although horticultural products probably accounted for only a small portion of the diet. Sites range from open-air lithic scatters with hearths, basin metates and one-hand manos, to south facing rockshelters with deep deposits (Dick 1965; Irwin-Williams 1973). Stemmed and diagonally notched dart points and knives are diagnostic of this adaptation, as are certain patterns of raw material use.

Archaic materials from the Upper Pecos Valley and surrounding mountain areas are known from private collections (Nordby 1990:33). No positive Archaic sites were known within the park until the present survey, which recorded three Archaic site components. Further south in the Upper Pecos Valley Anschuetz (1980) reports four Late Archaic sites which contain hearths and an abundance of lithic debitage in areas surveyed near Rowe Pueblo (Cordell 1998). Lang (1977), in a survey around San Cristobal Pueblo in the Galisteo Basin, suggests the presence of two separate Archaic traditions: the Oshara, associated with the northern Southwest (Irwin-Williams 1973); and the Cochise, an arid-lands adaptation known from southern Arizona, New Mexico, and northern Mexico (Dick 1965). It is not known yet whether a similar distinction can be made among the Archaic materials from the Upper Pecos Valley.

Developmental Period (A.D. 600-1200)

During the Developmental period, sedentary habitation became more common in the northern Rio Grande region. This change in residential permanency was accompanied by increased dependence on agriculture, especially corn (Zea mays). Early in the period, habitations take the form of pithouses; later, small above-ground structures are built. Around A.D. 700, the bow and small, usually corner-notched, arrow points replace the atlatl and heavy dart points of the Archaic tool-kit. Settlements are usually near water sources.

Permanent dwellings first appear in the Upper Pecos Valley during the Developmental period. Five pithouses constructed around A.D. 800-950 are the only early Developmental sites known in the area (Akins 1994; Nordby and Creutz 1993a). Very little is known about settlement in the valley between about A.D. 950 and 1200. Excavations at Rowe (Cordell 1998) and Forked Lightning (Kidder 1958) found evidence of small adobe structures dating to the eleventh and twelfth centuries, but the exact nature of these structures could not be determined due to later structures covering them. The small number of sites between A.D. 950 and 1200 suggests that after abandonment of the pithouses, the valley was not permanently inhabited, although it may have been used intermittently or seasonally by small groups. It is also possible that sites occupied during this period are simply buried deeply in areas with little erosion.

Coalition Period (A.D. 1200-1325)

The Coalition period is generally divided into Early (A.D. 1200-1250) and Late (A.D. 1250-1325) phases. It is marked by intensification of agriculture and by the coalescence of most of the population into larger pueblos. These characteristics seem to appear simultaneously in many portions of the Southwest (see Adler 1996). It is thought that much of the cultural change associated with the Coalition period is influenced by an influx of population to the Rio Grande Valley during the abandonment of the Four Corners region in the mid- to late 1200s.

During the early Coalition, at least two pueblos are present in the Upper Pecos Valley. The first, an adobe structure underlying Rowe Pueblo, is dated between A.D. 1240 and 1250 (Cordell 1998). The second site, Forked Lightning Pueblo, is dated from approximately A.D. 1225 to 1300 based on the presence of early white ware ceramics and the lack of glaze ware (Kidder 1958:42).

During the Late phase of the Coalition period (A.D. 1250-1325), the Upper Pecos Valley appears to experience a population influx, as indicated by the construction and occupation of as many as six pueblos of 50 rooms or more. In addition to continued occupation at Forked Lightning and Rowe, Black-on-white House, the structure underlying Pecos Pueblo (A.D. 1300-1425), was constructed, as were Shin'po or Hill House (PECO 307; LA 267; A.D. 1300-1350), Loma Lothrop (PECO 227; LA 277; A.D. 1315-1450), and possibly Dick's Ruin (PECO 434; LA 276; A.D. 1300-1425) (Cordell 1998; Kidder 1958:42, 47, 55-63). The locations of these pueblos are shown in Figure 1.3. The sources of the populations establishing these sites are unknown, but many may be settlers from the Rio Grande Valley attracted to the relatively cool and moist Upper Pecos Valley.

Figure 1.3. Locations of major prehistoric sites in the Upper Pecos Valley. (click on image for an enlargement in a new window)

It is during the Coalition period that the Southern Plains first supported farmsteads along tributaries of the Canadian River. Most of the few southwestern goods at these sites are exotic items, such as decorated ceramics, obsidian, turquoise, and shell from the Gulf of California (Lintz 1991:93). The presence of these materials on the Southern Plains suggests that reciprocal exchange between Plains and Pueblo groups involving relatively small amounts of raw materials or crafted items began at this time (Habicht-Mauche 1991:51-52; Krieger 1946).

Classic Period (A.D. 1325-1600)

The Rio Grande Classic period is divided into three phases, Early (A.D. 1325-1400), Middle (A.D. 1400-1525), and Late (A.D. 1525-1600). In the northern Rio Grande region, the Classic period is distinguished by the presence of glaze painted ceramics and very large residential structures. Researchers are now beginning to believe that during the Classic period the Rio Grande Valley was organized into a number of shifting alliances, with each community independent, yet linked to others through a complex network of economic, social, and political relations (Kessell 1979:12; Spielmann 1994; Wilcox 1981, 1984, 1991).

Exchange probably became an economic necessity for residents of both Pueblo and Plains during the Classic period, as climatic fluctuations and movement of Athapaskans into the Southern Plains, and population growth and depletion of natural resources in the Rio Grande, caused widespread subsistence stress. Pecos, as an active participant in the commerce between the Rio Grande and Plains, undoubtedly played a significant role in the development of economic networks in the northern Rio Grande Valley.

During the Early phase of the Classic period (A.D. 1325-1400) in the Upper Pecos Valley, masonry structures were built at Rowe Pueblo (LA 108; A.D. 1340-1450) and Arrowhead (PECO 710; LA 251; A.D. 1370-1450), and possibly at Hobson-Dressler (LA 32388; A.D. 1350-1400) (Cordell 1998; Holden 1955:103; National Park Service 1992a:Fig. 3). Both Arrowhead and Hobson-Dressler are located near Glorieta Pass (Figure 1.3). Structures on the Pecos mesilla were expanded, and residence continued at Loma Lothrop, but Forked Lightning Pueblo, Dick's Ruin, and Shin'po were probably no longer used as residences.

The Middle phase of the Classic period (A.D. 1400-1525) was a time of great change in the Upper Pecos Valley. A massive, four-story structure enclosing a central plaza—what we now know as the Pecos Pueblo Quadrangle—was constructed on the mesilla in about A.D. 1425 (Kidder 1958). The Hobson-Dressler site was deserted, while occupation at Rowe, Loma Lothrop, and Arrowhead persisted for a few years. By 1450, all human habitation in the valley was concentrated at Pecos Pueblo. Pecos began to produce its own glaze wares at this time.

During the Late phase of the Classic period (A.D. 1525-1600) Pecos Pueblo developed into a prosperous and influential community. It served as a center of trade and contact between the Southern Plains and Eastern Pueblos. Beginning in the early 1500s Pecos not only produced its own glaze ware, it distributed it to the Plains and other pueblos outside the Upper Pecos Valley as well (Kidder and Shepard 1936; Shepard 1942:154-155; Welker 1997:171).

Pecos Pueblo/Spanish Mission Period (A.D. 1540-1838)

To describe the historic period occupation at Pecos, the structure of this culture history now shifts to the time periods developed by the Cultural Landscape Overview (Cowley et al. 1997). These periods were developed to capture changes in the use and appearance of cultural landscapes in the Pecos area and not to split time into mutually exclusive sequential units. Some of the broader Cultural Landscape Overview periods are here subdivided using shorter periods derived from historical reconstructions for the Pecos area.

In 1540 several Pecos Indians trading at Zuni encountered the Coronado expedition. A year later, in 1541, Coronado launched his ill-fated search for Quivira from Pecos (Kessell 1979; Winship 1896). A Franciscan friar who came to New Mexico with Coronado remained at Pecos after Coronado's departure but died on the Plains continuing the fruitless search for Quiviran gold (Kessell 1979:26-27; Winship 1896:534-535).

Other Spanish expeditions passed or stopped at Pecos: the Rodriguez-Chamuscado expedition in 1581; the expedition of Antonio de Espejo in 1582; and in 1590, that of Gaspar Costaño de Sosa, who, hoping to make his fortune, entered the pueblo world by following the Pecos River north from Mexico. Meeting resistance at Pecos, the first pueblo he encountered, his soldiers stormed the mesilla and after much fighting, subdued the powerful village (Kessell 1979:31-62; Schroeder and Matson 1965). In 1598 Don Juan de Oñate arrived in the northern Rio Grande Valley with a wagon train of settlers and friars and established the first Spanish settlement near San Juan Pueblo. Oñate assigned missionaries to all the Rio Grande pueblos, but the initial missionary at Pecos lasted only three months (Kessell 1979:82). Sometime between 1617 and 1620 a Franciscan friar became a resident at Pecos, and a church was constructed not far from the pueblo (Ivey 1996). The Spanish era in the Upper Pecos Valley had begun.

Spanish Colonial Period (A.D. 1600-1821)

The Spanish Colonial period in New Mexico can be divided into three phases: Pre-Revolt (A.D. 1600-1680), Pueblo Revolt (A.D. 1680-1692), and Post-Revolt (A.D. 1692-1821). These intervals are distinguished by differences in colonial policies regarding land and Indians and by differing magnitudes of Spanish presence. The first and last phases of Spanish occupation are separated by the Pueblo Revolt, a 12-year period during which the Spanish were forced to retreat from New Mexico.

Early Colonial/Pre-Revolt (A.D. 1600-1680)

The first eighty years of Spanish rule in northern New Spain saw the development of deep rifts between the Church and the colonial government, primarily over the use of Puebloan labor and access to its fruits. Using Puebloan labor, numerous churches were constructed, and Spanish haciendas, usually supported by tribute from the pueblos, were established under the encomienda.

The encomienda was a major element of colonial policy in northern New Mexico in the years before the Pueblo Revolt. Under this system, pueblo villages were required to provide tribute to a Spanish settler awarded an encomienda by the governor. In exchange, the encomendero protected the pueblo from raiding and supplied the governor with military support (Kessell 1979:99). An encomendero was usually a high ranking official in the Spanish government in New Spain, or head of an influential Spanish family. This system took valuable food and goods from the pueblos, making it harder for village residents to support their own households. At the same time, Franciscan missions were established at large pueblos to convert and Christianize the souls residing there. Missionaries used pueblo labor to build the churches and associated buildings, and the pueblos provided materials and labor to support the missions.

Evidence of the Spanish colonial occupation at Pecos is largely confined to the mission church and convento. Material manifestations of Spanish influence are the use of form-molded adobes, a variety of metal artifacts, and the appearance of new ceramic forms, such as soup plates and candle sticks.

Over the next 80 years the incremental and wearing effects of disease, drought, and tribute mounted. On August 10, 1680, the normally autonomous pueblos acted jointly and simultaneously to rid themselves of Spanish rule and oppression. During the Pueblo Revolt numerous Spanish families and friars were killed or driven from their homes and churches, and ultimately from northern New Spain. The priest from Pecos attempted to escape to Galisteo Pueblo, but was murdered just outside its walls. The friar and other Spanish citizens remaining at Pecos Pueblo were killed (Kessell 1979:228). Spanish reign over New Mexico ceased for a time (Hackett and Shelby 1970).

Pueblo Revolt (A.D. 1680-1692)

The years between the revolt and the reconquest are poorly known due to the absence of European chroniclers. The lack of major stylistic or technological change suggests that the Spanish presence (and absence) only had little effect on indigenous craft production (see also Snow 1973:56). During the years following the revolt, the Spanish settlement of Santa Fe was occupied by Tano and Tewa speaking Puebloans. El Pope, the primary instigator of the revolt, exhorted every pueblo to obliterate every vestige of Spanish occupation. Churches were burned and dismantled, and church bells, santos, crucifixes, and vestments were destroyed. Even sheep and cattle were killed.

The revolt was a time of great upheaval in Rio Grande Pueblo society and culture. Three generations of Puebloans had grown up under the influence of the Spanish, and every aspect of Puebloan life—religious, economic, social, and political—had been affected. Disease and starvation had taken a heavy toll on pueblo populations (Kessell 1979:489-495; Levine and LaBauve 1994:36). As a result, the socioeconomic network that had previously existed in the region could not be reconstructed during the short time the Spanish were gone. Unsuccessful attempts at reconquest by the displaced colony also had a destabilizing effect on the Puebloans. When de Vargas returned to the province in 1692 he found the pueblos strongly divided (Simmons 1979:186).

At Pecos, archeological evidence of the destruction accompanying the Pueblo Revolt is limited. The massive seventeenth century church was completely razed, reportedly by Tewas, with whom the Pecos had been at odds for years (Kessell 1979:239, 241). Little but the foundations of this church have been identified archeologically. Adolph Bandelier reported seeing portions of what he believed to be the Pecos church bell on Glorieta Mesa (Bandelier 1881:99-101). The kiva in the middle of the convento may represent an act of vengeance against the Spanish religion (Hayes 1974:32), though this interpretation has recently been questioned (Ivey 1997). Otherwise there is little physical evidence to suggest that any conflagration or upheaval occurred at Pecos.

Late Colonial/Post-Revolt (A.D. 1692-1821)

In 1692, Don Diego de Vargas marched into Santa Fe and repossessed the capital. He then toured the Rio Grande, requiring and receiving the submission of each pueblo. Though his initial reception by the pueblos was quiet, the complete recolonization of New Mexico was more problematic, bloody, and time-consuming. Revolts cropped up over the following years, but this time the Spanish response was to send more settlers and soldiers. The encomienda was abolished, and relations between church and state became less contentious. The reconquest was successful; the Spanish foothold in northern New Spain was firm.

Pecos sided with the Spanish during the reconquest despite the presence of an anti-Spanish faction within the Pueblo (Hayes 1974:10; Kessell 1979). A new church was built inside the foundations of the earlier church, and materials from the seventeenth century convento were recycled for use in constructing a new convento on the same site (Ivey 1996; White 1996).

By the turn of the eighteenth century, Spaniards and Puebloans on the frontier faced a common foe, the Comanche. Raiding from the Southern Plains, the Comanche were less interested in trading than the Apache. As early as 1717, Pecos warriors were probably fighting side-by-side with the Spanish against the Comanche (Hayes 1974:13). Severe attacks against Pecos occurred throughout the 1700s. The pueblo's population continued to decline as warriors and farmers were killed in raids, and as smallpox ravaged the Rio Grande.

By 1773, the church at Pecos had been reduced to a visita of Santa Fe, with a priest visiting only occasionally (Adams and Chavez 1956:212; Hayes 1974:16). In 1776, Dominguez estimated the Pecos population at 269 individuals and reported that the "miserable wretches" at Pecos could not use their irrigated fields to the north and east of the pueblo because they were so frequently attacked by the Comanche. As a consequence, they resorted to wandering afar, selling their own possessions to support themselves (Adams and Chavez 1956:213).

Ultimately the reduced number of people living at Pecos Pueblo and the threat of attack produced additional consequences. After 1700, official grants of at least four square leagues of surrounding land were made to each pueblo (Simmons 1979:182). These grants were intended to provide for the safety and subsistence of the pueblos by forbidding encroachment on Indian land under tillage. The founding of the Spanish villages of San Miguel and San Jose del Vado, along the Pecos River south and east of the Pueblo, in 1794 and 1802, respectively (Hall 1984:3-5), proved that these pueblo grants were not secure, particularly when the land was not being used intensively. In 1814, tracts of irrigated land immediately south of the Pecos Pueblo boundary were granted to families from San Miguel, and the stage was set for the alienation of Pecos Pueblo's lands (Hall 1984:17).

Early Settlement, Santa Fe Trail, Civil War Period (A.D. 1812-1880)

Mexican/Santa Fe Trail (A.D. 1821-1846)

Under the Treaty of Cordova, signed on August 24, 1821, Mexico gained independence from Spain, and New Mexico became part of Mexico. Mexican independence brought two major changes to New Mexico—more lenient land grant policies and expansion of trade (Levine et al. 1995). Mexican law and custom changed the status of Indians, making them full Mexican citizens and thereby eliminating the ward status they held under Spanish rule (Hall 1984:32). Puebloans now had the right to own and dispose of land. At Pecos, this change in status officially removed the Pueblo's right to limit use to members of the Pueblo. With a dwindling population, the residents of Pecos had not used much of their land for years. By 1825 the fertile cienega north of the pueblo, was largely controlled by non-Indians (Hall 1984:47).

It was not long before the villages of San Jose del Vado and San Miguel del Vado were joined by additional Spanish settlements both north and south of Pecos Pueblo. Spanish names now frame the landscape along the Pecos River—Los Trigos (Wheat), Las Ruedas (Wheels), El Gusano (The Worm) (Almaraz 1988:88; Cobos 1983). In 1825, only one baptism, one marriage, and one burial were recorded at Pecos compared to 176 baptisms, 40 marriages, and 40 burials for El Vado (Almaraz 1988:88; Kessell 1979:562; Levine and LaBauve 1994). In 1838, the last residents of Pecos left their home to join the pueblo of Jemez.

During the early years of the nineteenth century the economy of the Upper Pecos Valley remained at a subsistence level with fishing and hunting used to augment agriculture and stock raising. Land along the river was used for farming, while uplands provided grazing for goats, sheep, and cattle. Five Spanish homestead sites are located within the boundaries of Pecos National Historical Park. These sites have a period of occupation extending from 1820 to the 1930s. Many of the families may not have been residents of the Upper Pecos Valley year-round, but they tilled and grazed the land in order to retain their claims.

While the settlers, often of mixed Spanish and Indian descent, were colonizing the Pecos area, purchase of the vast Louisiana Territory by the United States brought American explorers and traders westward. In September of 1821, at the time of Mexican independence, William Becknell headed west from Missouri, creditors at his back, with visions of bounteous trade. He and his goods were welcomed warmly at Santa Fe, and the focus of Mexican frontier commerce began to shift from the merchants in Chihuahua to those in Missouri (Almaraz 1988:86; Simmons 1988:94). American wagons travelling the route that became known as the Santa Fe Trail brought cloth, clothes, jewelry, mirrors, window glass, and other luxuries and necessities. Traders headed east with buffalo robes, buckskins, blankets, gold from the Ortiz Mines, and silver Mexican reales (Simmons 1988:97).

United States Territorial (A.D. 1846-1880)

In Santa Fe, on August 15, 1846, Colonel Stephen W. Kearny, commander of the Army of the West, claimed New Mexico for the United States. The American Territorial period in New Mexico is distinguished by a growing interest in commerce and a market economy that demanded a dependable means of transportation. Initially, Conestoga wagons filled with families and goods lumbered toward the west along the Santa Fe Trail. By 1850, long distance stagecoach routes were established to transport travelers as well as the mail.

Pigeon's Ranch, near Glorieta Pass was one stopover along the Santa Fe Trail. Alexander Pigeon (or Encher Pigeon) (Oakes and Swanson 1995:15), a French-American from St. Louis who soon changed his name to Alexander Valle (Hall 1984:149), established the stop in the early 1850s. It served as a forage and supply station for the U.S. Army in addition to providing food, entertainment, and lodging for travelers (Oakes and Swanson 1995:15-18). Today, only three adobe rooms, a mounded ruin, a well, and stone corrals remain.

In 1858 Martin Kozlowski, an immigrant from Poland who had fought Apaches with the U.S. Army in southern Arizona and New Mexico, started a ranch near a spring south of Pecos Pueblo. Travelers along the Santa Fe Trail frequented this spring and Kozlowski's Trading Post became a favorite stop for overland stages and other travelers (Kajencki 1990:47-66). Kozlowski obtained permission from Archbishop Lamy to use materials from the Pecos church to build barns and outbuildings (see Appendix E for recently collected dendrochronological in formation from Kozlowski's Trading Post).

In late 1861, General Sibley, commanding a Confederate Army force of 2,600 men, marched from Texas in an attempt to capture the western territories, and particularly, the gold fields of Colorado and California. Meeting only minor resistance, the Confederate forces moved up the Rio Grande, taking foodstuffs and supplies from villages along the way. Their march was stopped in late March of 1862 when they were engaged and ultimately defeated at Glorieta Pass by a force of Union Army regulars and Colorado volunteers (Swanson 1988).

Kozlowski's Trading Post served as the Union headquarters and hospital during the battle (Haecker 1998). The Union dead were buried "beneath the shadow of the ancient Pecos Church" (Hall 1984:116). The exact location of their original resting place is presently unknown. Pigeon's Ranch was used as the site of the Confederate field hospital and then as a burial ground. A mass grave containing Confederate dead was discovered and excavated in 1987 (Oakes 1990).

Railroad/Tourism (A.D. 1880-1941)

In 1878 the Denver and Rio Grande Railroad (D&RG) began laying track across northern New Mexico and through Glorieta Pass. The railroad had a profound effect on the economy of the territory. Lumber for cross ties was brought out of the Sangre de Cristo Mountains, and supplies were needed for work crews. The mountains surrounding Pecos came alive with lumber camps and sawmills. The Atchison, Topeka and Santa Fe (AT&SF) collected carloads of timber at Baughl's Station (or Bowll's Switch), located near the northern end of Glorieta Mesa, at the opening of Glorieta Pass.

In addition, there were rumors of mineral riches along the Pecos River. An 1881 newspaper article reported that a group of prospectors were at work in the mountains above Pecos, although few mines in the Pecos Canyon achieved steady production. The Evangeline Mine on the Pecos River north of the park produced copper laced with gold and silver, but its ores could not be exploited using the smelting techniques of the era. The mine failed but was revived in the late 1920s when processing technology changed (Glover and McCall 1988:115).

Homesteading, particularly of Hispanic sites now within the park, peaked during this interval. A gristmill was constructed on a bend in the river to serve the local residents (White and Porter 1996). There is also what appears to be an exploratory mine within the park, on the east side of the Pecos River. Created by a local Pecos man, it apparently produced no ore (Dobie 1931:222-238; a more detailed description of this site is given in Chapter 10).

Archeologists first visited and studied Pecos Pueblo during this time. In the early 1880s Adolph Bandelier visited and mapped the pueblo and some of its associated structures (Bandelier 1881). In 1914, A. V. Kidder began stratigraphic excavations at Pecos Pueblo. These were among the first in North America and formed the basis for the Rio Grande pottery sequence as well as the Pecos Classification (Kidder 1927). With research also came preservation. In 1920, the owners of the Gross-Kelly Company, a local mercantile business, deeded 80 acres surrounding the pueblo, including the church and convento, to the Archdiocese of Santa Fe, which turned the deed over to the Museum of New Mexico and the School of American Research in Santa Fe (Kessell 1979:481; Kidder 1958:325). In 1935, Pecos became a New Mexico State Monument, and in 1965, a National Monument (Ivey 1987).

Ranching Period (A.D. 1925-1993)

By 1925, most of the homesteads and properties along the Pecos River and in the former Pecos Pueblo grant were owned by Tex Austin, an international rodeo promoter. His Forked Lightning Ranch was both a working dude ranch and a cattle operation. Guests arrived on the train and spent their time horseback riding, exploring Pecos Pueblo, and relaxing at the elegant ranch house above the confluence of the Pecos River and Glorieta Creek (1920s Forked Lightning Ranch brochure on display at Pecos National Historical Park). Austin enlarged Kozlowski's Trading Post, building corrals, sheds, barns, a tennis court, and a polo field. He went bankrupt, and in 1939, Colonel E. E. "Buddy" Fogelson purchased the Forked Lightning Ranch.

Fogelson/Monument Period (A.D. 1941-1991)

Colonel Fogelson and his wife, actress Greer Garson, raised Santa Gertrudis cattle on the ranch and spent their summers there. Some of the outbuildings constructed by Austin were torn down during their tenure. A well was dug in the Middle Pasture south of the Forked Lightning Ranch house, and additional corrals were constructed. During the 1960s, a program of vegetation clearing was undertaken to increase pastureland. Bulldozers were used to remove pinon and juniper trees in flat areas, mostly on the west side of the river (Gilbert Ortiz, personal communication 1995; see below and Chapter 5 for more details on the effects of this clearing on archeological sites). In the 1960s, the Fogelsons donated land to the newly formed Pecos National Monument. Additional land was donated in the 1980s, along with money for a monument Visitors Center. After Colonel Fogelson's death in 1987, the northern half of the ranch was sold to the R. K. Mellon Foundation, which then donated it to the National Park Service, through the Conservation Fund, in 1991.

The Environment

The Upper Pecos River Valley has been used by humans for nearly 12,000 years. It is likely that the valley has always been attractive because of its position as a natural corridor between the Great Plains and the Rio Grande Valley. Another attraction of the area is the degree of topographic, floral, and faunal diversity offered within relatively short distances.

Although it is difficult to reconstruct most aspects of prehistoric environment without a variety of detailed studies, the following discussion presents information on several aspects of the modern environment of the Upper Pecos Valley, both as a basis for understanding prehistoric and historic human adaptation to the Pecos area and as a basis for making inferences about aspects of earlier environments.

Topography

The Pecos River originates in the Sangre de Cristo Mountains and follows a narrow canyon until it reaches the modern village of Pecos, where it runs through a broad valley approximately 9.7 km (6 mi) wide and 10.5 km (6.5 mi) long. Numerous mountain streams feed the river along its route from the mountains to the Upper Pecos Valley. There are three major topographic divisions in the Pecos National Historical Park—the Pecos River Valley proper, the Tecolote Range, and Glorieta Mesa. Glorieta Mesa is outside the park. The Pecos River runs along the Tecolote Range at the eastern edge of the valley, separating the foothills from the broader expanse of valley floor west of the river. The valley floor ends abruptly at the base of Glorieta Mesa, which runs north-south and towers over the surrounding landscape (Figure 1.4).

Figure 1.4. Upper Pecos Valley topography. View SW from the east side of the Pecos River, overlooking the river, the Forked Lightning Ranch, the open valley floor, and Glorieta Mesa beyond.

The rugged Tecolote Range east of the river is dissected by numerous steep-sided drainages. These canyons flow intermittently to the west or southwest into the Pecos River. The ridges separating the drainages are narrow and flat. Elevation ranges from 2,048 m (6,720 ft) at the river to 2,304 m (7,560 ft) at the highest point in the northeastern part of the park (Figure 1.5).

Figure 1.5. Rugged topography east of the Pecos River in the Tecolote Range, view NE.

Glorieta Mesa, on the western side of the valley, dominates the landscape and separates the valley from the Galisteo Basin and the Rio Grande Valley. The mesa is outside the park, but plants, mammals, stone, minerals, and other resources were undoubtedly used by Pecos Valley residents. The base of the mesa is talus, from which steep sides, interspersed with cliffs, rise to around 2,400 m (close to 7,875 ft). Water running off the mesa's eastern slope enters Glorieta Creek, which drains the valley floor. Historically intermittent, the flow of Glorieta Creek has been continuous since the late 1960s, although the magnitude of the flow fluctuates wildly (Nordby 1990). From its origin at Glorieta Pass, the creek runs through the park from the vicinity of Pigeon's Ranch southward along the west side of the Pecos Pueblo mesilla. It eventually joins the Pecos River immediately below the Forked Lightning Ranch house.

Between the Pecos River and Glorieta Mesa the Pecos Valley is a gently sloping, undulating surface dotted with piñon and juniper and stands of ponderosa pine. Although dissected by small intermittent drainages and dotted with a few higher knolls, there are no major physiographic features. Elevation along the river is around 2,048 m (6,720 ft), and the highest point on the valley floor is a minor hill some 2,168 m (7,112 ft) in height. In addition to Glorieta Creek and the Pecos River there are five known springs within the park: Trading Post Spring (adjacent to Kozlowski's Trading Post), Mud Spring (south of Kozlowski's), Poison Spring (east of Pecos Pueblo), an unnamed spring north of Pecos Pueblo, and an unnamed spring north of Shin'po (Eininger et al. 1995).

Geology

The earliest deposits in the Pecos vicinity are Precambrian amphibolites, granites, quartzes, pegmatites, schists, and quartzites formed in the Sangre de Cristo Mountains and now present within the park as surface cobbles.

During the Mississippian period (ca. 290 million years ago), marine erosion of the earlier pre-Cambrian deposits resulted in the formation of conglomerates and sandstones that constitute the earliest sedimentary rocks in the Pecos area (Nordby 1990:9). Other limestones and sedimentary rocks are represented in the Magdalena group and the Madera formation of Pennsylvanian age. The Sangre de Cristo formation, which covers the Pecos River Valley floor, was created during periods of occasional submergence followed by complete withdrawal of the sea.

The Yeso formation, consisting of red sandstones and siltstones, and gray limestone beds, was deposited on top of the Sangre de Cristo formation during the Permian, a period of dry climatic conditions. These bright red deposits show up in the Pecos area only on the steep slopes of Glorieta Mesa. Two later episodes deposited layers of gray sandstone (Glorieta sandstone) and gray limestone (San Andres formation) on Glorieta Mesa. The Bernal formation, a deposit of maroon sandstone and siltstone beds, is exposed by erosion on portions of Glorieta Mesa. In the drainages in the valley, the Bernal formation is associated with thin gypsum beds.

During the Triassic, major uplifting formed the Tecolote Range and raised the northern scarp of Glorieta Mesa (Nordby 1990:14). These episodes were followed by deposition of Santa Rosa sandstone, which is light tan in color and has a fairly restricted distribution. The Pecos River has completely eroded the valley's Mesozoic rocks, and as a result either of uplifting or downcutting, the river's course, which is still marked by cobbles, has clearly moved eastward.

Many of these deposits contain materials that proved useful to the prehistoric inhabitants of Pecos. Quartzite and other rocks of Precambrian age were utilized for grinding tools and hammer stones, while the later Magdalena deposits provided cherts and chalcedonies for flaked stone tools as well as sandstone slabs for architecture.

In more recent Quaternary times, alternating episodes of erosion and alluviation have created the characteristic riverine terraces, as well as the erosion responsible for formation of the Pecos mesilla. The arkosic sandstone that caps the mesilla is fractured and breaks away in blocks and most likely provided much of the building stone used at masonry sites. The shale eroding on the mesilla probably also provided the adobe for building and possibly clay for making pots (Nordby 1990:15).

Soils

Several soil types are present in the park. According to Spielmann (199la: Map 2), only two of these, a fine sandy loam (Manzano or Mb), and an undulating soil association known as the Vibo-Ribera (or VB) are suitable for cultivation.

The fine sandy loams occur on fans and floodplains and are deep and well thinned with slopes up to 3 percent. These soils have slow permeability, and available water capacity, referring to the soil's ability to hold water for use by plants, is high. Runoff is medium; water erosion hazard is moderate, although hazard of blowing wind is high (Hilley et al. 1981:24).

The Vibo-Ribera association soils occur on fans with slopes from 1 to 9 percent. They are moderately permeable with a high available water capacity, medium runoff, moderate water erosion, and a high hazard of blowing wind. Ribera soils are on the higher slopes, from 5 to 9 percent, and have moderate permeability and water capacity, medium runoff, moderate erosion, and a high wind hazard (Hilley et al. 1981:40). The VB association is appropriate for dry farming (Spielmann 1991a: 102).

According to Spielmann (1991a: 102-104), most of the VB and fine sandy loam soils occur to the north, northwest, and northeast of Pecos Pueblo. There are also areas of Vibo-Ribera soils 2 km (1.2 mi) southeast of the pueblo. The majority of the fine sand loam soils are a minimum of 2 km (1.2 mi) northeast of Pecos Pueblo, in the vicinity of the town of Pecos. Spielmann (1991a) measured approximately 175 hectares (432 acres) of potentially irrigable land and 465 hectares (1,149 acres) of land that could have been dry farmed by the population at Pecos Pueblo within these soil units. In addition, there are more Vibo-Ribera soils farther south that probably could have been farmed that she did not include in her estimates.

Climate

Of the many different aspects of climate, those most important to Pecos agriculturalists were moisture and temperature. The timing of moisture availability affects crop growth and also the productivity of wild plants and animals. Temperature affects germination and maturation of crops, and the length of the frost-free season is especially important to successful corn maturation.

The Pecos area averages between 40 and 50 cm (16 and 20 inches) of precipitation annually. About three-quarters of this moisture comes between late spring and early fall, with about one-third falling during the summer monsoons in July and August. In the winter the Pecos Valley receives anywhere from 25 to 89 cm (10 to 35 inches) of snow. The Pecos River is fed by mountain snow melt, which means that irrigation agriculture is affected by winter snowfall and precipitation that feeds the river during the spring and summer.

The growing season in the Pecos area averages 127 days, which is marginal for corn agriculture. Spielmann (1991a: 116) estimates that it falls below 120 days, the minimum needed for corn, one out of every three years. In addition, frosts are more common in late spring than early fall, and they are usually more harmful than frosts late in the growing season. As Snow (1991:73) states, "If prehistoric growing season length at elevations above 6,000 feet was highly variable, as it is today in northern New Mexico and adjacent areas, relying on a mature maize crop from one year to the next would have been an undependable strategy."

Even with a shortened growing season, the inhabitants of Pecos most likely harvested green corn, which would not provide seed for the following year, but would supply short-term calories and food for storage (Snow 1991:73). In Snow's (1991:75) view, the probable inability of peripheral populations, especially those above 6,000 feet, to produce seed corn would have limited their "agricultural independence."

Certain aspects of prehistoric climate are critical to understanding how the agricultural subsistence system might have functioned and the problems that farmers may have faced. The Palmer Drought Severity Index (PDSI) is one way to estimate climatic stress on agriculture (Rose et al. 1982). PDSI is an index of meteorological drought and it is calculated using estimates of potential evapotranspiration. Retrodiction of PDSI backward to time intervals without climatic records is based on tree-ring indices, and is, as best as possible, a measure of the moisture available to plants during the growing season. The June PDSI for the Santa Fe area tree-ring index, which includes samples from Glorieta Mesa, covers the years A.D. 883 to 1988 and is the most appropriate record for the Pecos area. Analyses of these values reveal quite variable conditions over time and are discussed in Chapter 6.

Another important aspect of prehistoric climate is the timing of rainfall during the growing season. In her research on corn growth, Spielmann (1991a: 109) found "that the most severe damage is experienced if stress occurs during the tasselling/silking stage. Moisture stress at this time can reduce grain yield by as much as 50%." Typically, the Santa Fe-Pecos area gets summer monsoons during the same time that corn is in the tasselling/silking stage. Unfortunately, the June PDSI does not provide information on the monsoon season, which begins in July.

Based on tree-ring reconstructions of prehistoric climate, Dean (1996:40) has very recently identified two patterns of precipitation, one characteristic of the "northwestern two thirds of the Southwest" the other of the "southeastern third of the region," that are strikingly similar to the present day. The southeastern pattern, which includes the Santa Fe-Pecos area, is dominated by a unimodal, summer-dominant regime (Dean 1996:Figure 4), while the northwestern pattern is a bimodal, summer-winter regime. The patterns are separated by a sinuous transition zone that winds through Arizona and New Mexico.

This geographic division was stable for over 500 years, then, between A.D. 1250 and 1450, "the long term pattern broke down into chaotic distributions of three or four principal components that exhibit no logical geographic patterning" (Dean 1996:43). The southeastern group, which includes Santa Fe-Pecos, remained stable. The northwestern group, however gave way to a variable mixture of rainfall regimes" (Dean 1996:44).

Within the northwestern regime area, the combination of high population densities, falling water tables, floodplain erosion, and severely depressed annual rainfall between A.D. 1275 and 1300 (the "Great Drought") resulted in the virtual abandonment of the San Juan drainage as a place of residence for agricultural societies. Population shifted to the south and east into the Little Colorado Valley, the Mogollon Highlands, and the Rio Grande drainage.

Because the Rio Grande Valley maintained its summer-dominant pattern during the 200 years of precipitation chaos in the northwestern regime area, it appears to have become highly attractive for late thirteenth and early fourteenth century settlement.

Flora and Fauna

The Upper Pecos Valley falls within the Upper Sonoran Life Zone (Bailey 1913), but other zones are accessible within short distances of the park. Rodriguez-Bejarno's (1993) vegetation map shows that most of the park is dominated by piñon-juniper. Along the Pecos River, riparian vegetation, such as cottonwood, willow, and reeds, is predominant. At higher elevations, oak, pinon, and juniper woodlands give way to ponderosa pine, Douglas fir, and aspen forests. Above the treeline, in the Sangre de Cristo Mountains north of the valley, grassland meadows predominate (Cordell 1998).

Within the park, the present vegetation is a reflection of many years of use for ranching and grazing. During the 1960s piñon and juniper were bulldozed in order to increase the amount of grass available for grazing (see below and Chapter 5). Today's landscape probably does not reflect prehistoric vegetation, which undoubtedly changed through time as trees were cut for building and firewood, and fields were cleared to grow corn.

Archeological evidence for the presence and potential use of certain plant species is available from excavations at Rowe Ruin (Cordell 1998:61). Most of the edible wild plant remains identified at Rowe are common disturbance species, such as pigweed (Amaranthus sp.), goosefoot (Chenopodium sp.), and purslane (Portulaca sp.). The Rowe excavations also revealed remains of tansy mustard, several cacti, Indian ricegrass, dropseed, juniper seeds, piñon nuts, and American plum. Wild tobacco (Nicotiana rustica) and wood from juniper, piñon, oak, and willow are nonfood items found at Rowe.

There is, and probably was, a diversity of fauna in the area. Parmenter and Lightfoot (1994) identify 24 different mammals, seven reptiles, and two amphibians within the park. They also found 189 arthropods. Historically, larger game animals, such as bighorn sheep, elk, mule deer, and pronghorn were found in the area.

A separate inventory of birds (Mukai 1989) has identified 109 species, of which 24 are year-round residents. There are fish in the river, most notably rainbow and brown trout, but both are introduced species. Fishes native to the river probably included "white sucker, long-nosed dace, and possibly the fathead chub, fathead minnow, Rio Grande chub, and Rio Grande cutthroat" (Eininger 1995:7).

Faunal remains from Rowe Ruin include the common species found at most prehistoric pueblos in the Southwest, but the percentage of large game is much higher at Rowe than other contemporaneous pueblos outside the valley (e.g., 39.3 percent vs. 18.5 percent at Tijeras Pueblo), and it appears that meat was a greater part of the Rowe diet than we might expect for Pueblo agriculturalists (Cordell 1998:62-63). Turkey bone also was found at Rowe, but Cordell did not find any dung or eggshells that might indicate they were raised there (Cordell 1998:64). Turkeys were bred and raised at Arroyo Hondo and Tijeras Pueblo, and probably at other prehistoric pueblos in the northern Rio Grande.

Recent Impacts to the Pecos Environment

Modern vegetation cover rarely reflects prehistoric conditions, and this is especially true in the Upper Pecos Valley. During the 1950s and 1960s, large bulldozers were used to clear piñon and juniper from portions of the valley. The fact of this clearing was known when the Pecos survey was begun, though its extent and effect were not well understood. Steps were taken during site recording to acknowledge and track vegetation removal, and in Chapter 5 this information is used to give a more detailed analysis of the effects of clearing on surface archaeological features. To summarize the results of that analysis: artifact scatters on affected sites are larger than those on unaffected sites, reflecting the horizontal movement of materials by bulldozer blades and treads; structures are less likely to occur on affected sites; and all nonportable features are more likely to be described with ambiguous terms on affected sites. The implications of these findings for interpretation are difficult to assess. They serve as a caveat emptor, and the reader must take vegetation removal impacts into consideration when appraising the project results.

Research Objectives

Based on the project research design (Head 1997a), data recovered by the survey are used to address research questions pertinent to both the prehistoric and historic periods in the Upper Pecos Valley. For the prehistoric and early historic period (ending at 1700) two problems related to the development of large sedentary villages are examined. The first issue concentrates on the development of the Pecos community, while the second examines the nature of exchange relationships between Pecos, other Rio Grande pueblos, and nomadic Plains groups.

Community Development

The prehistoric Pecos community evolved over a period of seven centuries from a small cluster of pithouses to a single large aggregated pueblo. In order to understand the changes that accompanied this transformation, our research focuses on aspects of community development, specifically: population growth and aggregation, agricultural intensification, and settlement.

Aggregation occurred in many portions of the northern Southwest during the thirteenth and fourteenth centuries (Adams 1991; Adler 1996; Kintigh 1985; Powers and Orcutt 1999; Reid 1989). A variety of theories have been proposed to explain the formation of large villages throughout the Southwest. Kidder (1924) put forth the need for defense as one of the earliest theories of aggregation, a theme that has been echoed recently (Haas and Creamer 1993, 1996). Other current explanatory models have involved minimization of conflict over resources (Hunter-Anderson 1979), population growth and environmental degradation (Kohler 1989; Orcutt et al. 1990), adaptation to climatic variability, particularly rainfall (Dean et al. 1985; Plog 1983; Plog et al. 1988), technological specialization, particularly agricultural intensification (Hill and Trierweiler 1986), and organization and legitimization of access to good quality arable land (Adler 1990, 1992). All of these discussions of population aggregation address aspects of population density, the nature of the resources being utilized, the technology used to exploit them, and methods of social integration (Cordell et al. 1994).

We begin our examination of the development of the Pecos community by investigating three aspects of the process of aggregation: (1) changes in the location, number, size, and structure of Puebloan sites; (2) the measurement of population aggregation within each time period; and (3) based on these data, estimations of population over time to determine whether the observed rate of growth is indicative of local population growth or migration from surrounding areas.

Changes in subsistence were clearly an important part of the development of the Pecos community. Agricultural intensification is frequently seen as a correlate of aggregation (Kohler 1989), and we examine whether it occurred with the formation of Pecos Pueblo. Four measures of agricultural intensification are evaluated: (1) increased investment in agricultural features; (2) change in ceramic vessel size through time as an indicator of greater reliance on agricultural products; (3) site location in relation to plant productivity; and (4) grinding intensity as measured by change in the shape, size, and degree of wear shown by manos.

Settlement patterns are closely related to subsistence. We expect that the distribution of prehistoric sites across the Pecos landscape changed over time as a consequence of aggregation and increasing dependence on agriculture. Three aspects of these changes are examined: (1) the relationship of sites to key natural resources and landscape features; (2) the relationships of sites to other sites of the same type as a means of inferring the degree of cooperation or competition among sites; (3) the relationship among different site types as a means of inferring change in subsistence patterns.

Exchange

It has long been recognized that Pecos Pueblo played a pivotal role in the exchange of a variety of subsistence and craft items. Kidder and Shepard (1936) identified the presence of nonlocal trade ware ceramics at Pecos. These nonlocal ceramics were manufactured by Puebloan groups in the Rio Grande Valley and eastern Arizona, then traded to Pecos. Likewise, obsidian from the Jemez Mountains occurs in substantial quantities in Pecos Pueblo deposits, again apparently as the result of exchange with other Rio Grande pueblos.

Complementing trade with the Rio Grande Valley and other Puebloans, exchange also occurred between Pecos and a variety of nomadic groups from the southern Plains. This trade is thought to begin in the late prehistoric period (ca. A.D. 1450) and likely involved the exchange of meat and hides from the Plains for corn, pottery, and obsidian from Pecos (Kessell 1979; Kidder 1932; Spielmann 1991a). Spielmann (1983, 1991a) has argued that the relationship between Pecos and other pueblos on the eastern edge of the Rio Grande Valley was mutualistic in that it provided resources each needed on a regular basis.

Two topics on Pecos exchange are investigated using the survey data: (1) information on the organization of production of utility and white ware ceramics; and, (2) identification of the quantity, geographic distribution, temporal duration, and sources of nonlocal lithic materials present at Pecos.

Historic Period Investigations

By the close of the eighteenth century, the Pecos Pueblo social and political organization, subsistence economy, settlement, and exchange systems were irrevocably changed as a result of Spanish colonization. Although Pecos Pueblo was not abandoned until 1838, archeological remains of the nineteenth and twentieth centuries are almost exclusively of Spanish or American (referred to collectively as Euro-American) cultural affiliation and reflect a new and different pattern of human use of the Upper Pecos Valley. The occupation of most of these sites occurred after the emergence of Mexico and the United States as independent nation-states. Although the Pecos Valley is initially on the far frontier of each of these political and economic systems, interaction with them is increasingly significant and crucial to understanding local settlement and subsistence.

The discussion of the Euro-American materials recorded by the Pecos survey has three objectives: (1) to provide accurate and replicable dates for the nineteenth and twentieth century sites through examination of the artifact data collected in the field; (2) to examine how the settlement and subsistence economy changed over time through consideration of census records pertaining to the area; and (3) to describe remains of the Santa Fe Trail within the park and to consider the effect it had on the subsistence economy of the Upper Pecos Valley.

Organization of the Volume

The chapters that follow furnish the setting and background for the project, then present the data gathered by the survey and apply that data to the research topics presented above.

In Chapter 2, Susan Eininger summarizes the history of archeological investigation at Pecos National Historical Park and in the Upper Pecos Valley. Archeological excavations, surveys, ancillary studies, and historical research conducted over the last 100 years are discussed.

Chapter 3, also by Susan Eininger, describes the inventory survey methodology and organization. Survey logistics, procedural definitions, data collection objectives, recording techniques, and artifact sampling procedures are presented and discussed.

In Chapter 4, Melissa Powell and Karl Benedict use ceramic types with known production spans to date the Pecos survey sites. Ceramic mean dating and ceramic probability data are the primary methods used to construct a seven-period chronology. The classification of sites into these temporal divisions provides a chronological framework for many of the analyses performed in subsequent chapters of the report.

In Chapter 5, Genevieve N. Head groups the sites recorded by the survey into three types using site architectural characteristics. Habitation, seasonal, and special use categories are created based on the presence or absence of structures and the number of feature types present on sites. In this same chapter, Head provides descriptions of the habitation sites in the park and examines the effects of aggregation on subsistence and society through changes in agricultural features and kivas.

In Chapter 6, Karl Benedict and Janet D. Orcutt describe the locations of different site types utilizing several environmental variables and nearest-neighbor analyses. Expectations for site locations are derived in part from changes in climate that affected agricultural productivity and are measured using the PDSI.

In Chapter 7 Janet D. Orcutt uses the survey data to reconstruct the number of people in the Pecos area during the Puebloan occupation. The survey data do not inform well on the development of Pecos Pueblo so other data are used to reconstruct its population. The population fluctuations in the Pecos area are compared to those in the larger Rio Grande area.

In Chapter 8, Melissa Powell assesses how increasing aggregation and changes in regional interaction at Pecos affect ceramic technology through time. Her chapter focuses on the cultural, temporal, and spatial variability in the survey ceramic assemblages. She investigates the functional content of ceramic assemblages as evidence for the intensification of agriculture. Finally, using petrographic and compositional data for utility and white ware ceramics, she evaluates evidence for the organization of pottery production at Pecos.

In Chapter 9, David Kilby and Joseph V. Cunningham contribute the first systematic analysis of lithic materials in the Upper Pecos Valley. Using chipped and ground stone data, they evaluate expectations for activities at different site types. An X-ray fluorescence study identifies obsidian sources for the large pueblos at Pecos. The authors also examine other chipped stone raw material sources to better understand interactions with Plains groups. Analysis of the survey manos is used to examine postulated increases in grinding activity.

In Chapter 10, Jeffrey L. Boyer, James L. Moore, Natasha Williamson, and Genevieve N. Head place the historic sites recorded by the survey into chronological perspective and provide details on census records and the Santa Fe Trail. In the first portion of the chapter, Boyer and Williamson use mean dates derived from field recorded artifact information to place the sites into the themes and landscapes presented earlier and derived from the Cultural Landscape Overview of the park (Cowley et al. 1997). In the second portion of the chapter, Williamson provides insight into changes in the life of the Upper Pecos Valley as revealed through a review of census records from the late nineteenth and early twentieth centuries. In the final portion of the chapter, Moore describes the ruts of the Santa Fe Trail that are still visible within the park, as well as several other circulation features discovered during survey. He places each in context through comparison with historic maps and documents.

Finally, in Chapter 11, Janet D. Orcutt and Genevieve N. Head review the conclusions of the previous chapters and integrate the results to provide a summary of the survey findings.