The Archaeological Geography of Small Architectural Sites of the Mogollon Plateau Region of East-Central Arizona

Item Type text; Electronic Dissertation

Authors Mehalic, David Steven

Publisher The University of Arizona.

Rights Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.

Download date 23/09/2021 10:51:00

Link to Item http://hdl.handle.net/10150/265814

THE ARCHAEOLOGICAL GEOGRAPHY OF SMALL ARCHITECTURAL SITES OF THE MOGOLLON PLATEAU REGION OF EAST-CENTRAL ARIZONA

DAVID S. MEHALIC

______

A Dissertation Submitted to the Faculty of the

SCHOOL OF ANTHROPOLOGY

In Partial Fulfillment of the Requirements For the Degree of

DOCTOR OF PHILOSOPHY

In the Graduate College

THE UNIVERSITY OF ARIZONA

2012

THE UNIVERSITY OF ARIZONA GRADUATE COLLEGE

As members of the Dissertation Committee, we certify that we have read the dissertation prepared by David S. Mehalic entitled The Archaeological Geography of Small Architectural Sites of the Mogollon Plateau Region of East-Central Arizona and recommend that it be accepted as fulfilling the dissertation requirement for the

Degree of Doctor of Philosophy

______Date: 11/26/2012 Barbara J. Mills

______Date: 11/26/2012 J. Jefferson Reid

______Date: 11/26/2012 Suzanne K. Fish

______Date: 11/26/2012 Gary Christopherson

Final approval and acceptance of this dissertation is contingent upon the candidate’s submission of the final copies of the dissertation to the Graduate College.

I hereby certify that I have read this dissertation prepared under my direction and recommend that it be accepted as fulfilling the dissertation requirement.

______Date: 11/26/2012 Dissertation Director: Barbara J. Mills

STATEMENT BY AUTHOR

This dissertation has been submitted in partial fulfillment of requirements for an advanced degree at the University of Arizona and is deposited in the University Library to be made available to borrowers under rules of the Library.

Brief quotations from this dissertation are allowable without special permission, provided that accurate acknowledgment of source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the Graduate College when in his or her judgment the proposed use of the material is in the interests of scholarship. In all other instances, however, permission must be obtained from the author.

SIGNED: David S. Mehalic

ACKNOWLEDGMENTS

Barbara Mills has been a direct, insightful, and supportive adviser throughout my time in graduate school. I am extremely grateful for all her guidance, helping me grow from an inquisitive field school student to a more responsible researcher and cultural resources professional. I am also honored to have worked with Jeff Reid, Suzanne Fish, and Gary Christopherson, as well as Jeff Dean and Phil Guertin. Linda Martin, Bruce Donaldson, Heather Provencio, Pete Taylor, Jeremy Haines, and many other individuals from the Apache-Sitgreaves, Coconino, and other National Forests exposed me to the Forest Service and shared their knowledge of the region and its archaeology. I am thankful for the encouragement and opportunities they offered. Time spent in the Mogollon Rim and surrounding country surveying and exploring was a true pleasure. Looking back, it’s hard to remember a bad day in the field. William Gillespie and Mary Farrell, my mentors on the Coronado National Forest, as well as colleagues Chris Schrager and Kathy Makansi, expanded my vision of archaeology and cultural resources management, providing many opportunities for personal growth and as an archaeologist, while also encouraging me to continue pursuing my student research and interests in other areas. Their patience is most appreciated, and their mentorship has been remarkable. It has been an honor to help preserve and study the cultural heritage of the American Southwest. Many fellow students at the University of Arizona ensured an enjoyable and stimulating graduate school experience, helping dismiss many of the stereotypes we learn to fear before entering a competitive program. The consistently collegial atmosphere afforded by the School of Anthropology, the Arizona State Museum, and the other departments I experienced was remarkable. My dad, Charles “Chuck” Mehalic, encouraged me to keep on pursuing archaeology and supported me in many ways through my lengthy student career, despite the recognition of a well-schooled engineer that it may not be the most practical pursuit (although I know he was glad someone is taking care of it). I regret that I didn’t take care of business while you were still around to see it. I have always felt nothing but support from all of my family and feel very fortunate, especially for my grandmothers. Mary Mehalic assured I never forgot that I needed to see my degree through to the end, and Mary Rapp instilled a love of natural history. Like her, I have also grown accustomed to stopping the car for eye-catching stones and suspect some sort of genetic or otherwise cosmic relationship is at play. Finally, my most sincere thanks and appreciation to Daniela Magalhães Klokler, the best thing to ever happen to me. I especially appreciate all of the encouragement during difficult times when things didn’t always seem possible, as well as the valuable help in the field and during the lengthy writing process.

DEDICATION

For my dad…

TABLE OF CONTENTS

LIST OF FIGURES ……………………….……………………………………….. 8 LIST OF TABLES ………………………….…………………………………….... 12 ABSTRACT…………………………………………………………………….…... 13

CHAPTER 1. INTRODUCTION TO THE SMALL SITES “PROBLEM” AND GIS SOLUTIONS………………………………….……….……………………… 14 Research Goals……..………………………….………………………………… 20 Organization of the Dissertation……………….………………………………… 22 CHAPTER 2.ARCHAEOLOGICAL GEOGRAPHY AND GIS FOR LANDSCAPE STUDIES…….…………………………….………………………. 26 Archaeological Geography…………………………….………………………… 27 Landscape Archaeology……………..…………………………………………… 30 Archaeological GIS: Presence, Probability and Prediction….…………………… 40 Reconciling Deductive and Inductive Approaches to Predictive Modeling…...… 46 Operationalizing Archaeological GIS for Small Site Studies……………………. 52 CHAPTER 3. THE ARCHAEOLOGICAL SETTING..………………………….. 56 Early Studies in Culture History………………………………………………. 57 Processual and Behavioral Approaches and the Rise of Cultural Resources Management………………………………………………………………… 62 Ongoing Management, Rodeo-Chediski, Healthy Forests, and Managing Future Forest Uses……………………………………..…………………… 75 Ethnographic Insights and Historical Land Use of the Mogollon Plateau……. 77 Comparing and Deconstructing Site Typologies……………………………… 88 Summary of Expectations………………..………………………………………. 91 Environmental Determinism…………………………………………………... 91 Agricultural Suitability………………………………………………………. 92 Migrant Communities and Seasonal Uses of Common Pool Resources……… 94 Distance Decay Model – Proximity to Large Sites…………………………… 96 Competition and Conflict – A Tragedy of the Commons?...... 97 Summary………………………………………………………………………. 98 CHAPTER 4. IDENTIFYING SMALL SITES WITHIN THE MODERN LANDSCAPE………………….…………………………………………………… 99 Archaeological Site and Survey Data………………..…………………………... 111 Archaeological Site Databases………………………………………………… 112 Comparing Site Databases…………………………………………………….. 123 Archaeological Survey Database……………………………………………… 126

TABLE OF CONTENTS - Continued

CHAPTER 5.SITUATING SMALL SITES WITHIN THE ENVIRONMENT….. 133 DEM and Derived Terrain Characteristics………………………………………. 134 Derived Terrain and Hydrological Surfaces………………………………….. 143 Derived Hydrography………………………………………………………… 146 Terrestrial Ecosystems Survey……………………………...…………………… 151 Site Location Probability Models…………….…………………………………. 160 Models for Assessing Agricultural Suitability……………………..……………. 170 Results………………………………………………………………………… 188 Discussion..……………………………………………………………………... 190 Summary………………………………………….…………………………… 193 CHAPTER 6. SITUATING SMALL SITES WITHIN THE SYSTEMIC LANDSCAPES OF THE MOGOLLON PLATEAU……………….……………… 194 Community Organization and Small Architectural Sites……………………...…. 195 Social Conflict and Culture Change in the Mogollon Plateau Region…………... 201 A Tragedy of the Commons?…...…..……………………………………………. 211 Common Pool Resources, Sustainability, and Culture Change…...………….….. 215 Issues of Scale and Site Function……………………………………………... 221 Discussion and Summary………...……………………….……………………… 233 CHAPTER 7. CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE RESEARCH………………………………………………………………………... 240 Ethnic Identity and Landscape Significance………………….………………….. 241 Implementing Predictive Models for CRM on Public Lands……………………. 246 Management Recommendations…………………………………………………. 251 Research Questions for Future Investigations at Small Sites…………………. 256 REFERENCES …………………………………………………………………….. 261

LIST OF FIGURES

Figure 1.1. An alignment of several cobbles and small boulders of limestone on the Apache-Sitgreaves National Forest, near Red Hill………… 15 Figure 1.2. Some examples of small site types and their relationships with archaeological and systemic context………………………………. 16 Figure 1.3. The study area outlined in yellow and corresponding Ranger District names overlying shaded relief Landsat imagery, with locations of modern towns and major roads (black) and interstates (red). The Mogollon Rim coincides with much of the southern boundary of the study area………………………………………… 19 Figure 2.1. Flow chart of the examination of small sites in the study area 54 Figure 3.1. The transmission line surveyed by the Cholla Project (background) from a site recorded by CARP, with a small valley suggested to be a field location in between, crossed by a Forest Service road…….. 66 Figure 4.1. The Mogollon Rim and the view toward the mountainous transition zone……………………………………………………... 100 Figure 4.2. Elevation throughout the study area derived from USGS National Elevation Dataset…………………………………………………... 102 Figure 4.3. Climatic gradient mapped for the study area by the Terrestrial Ecosystem Survey project (the null data areas are Mormon Lake on the west and Pinetop-Lakeside on the east)……………………. 104 Figure 4.4. Vegetation communities digitized from Brown and Lowe (1980)... 105 Figure 4.5. All surveyed areas in the study region (less-than-complete surveys shown as hatched symbols)………………………………………... 113 Figure 4.6. All recorded archaeological sites in the study area………………... 115 Figure 4.7. Site type frequencies from the Lakeside and Black Mesa Ranger Districts of the Apache-Sitgreaves National Forests………………. 123 Figure 4.8. Site type frequencies form the Mormon Lake and Mogollon Ranger Districts of the Coconino National Forest………………… 124 Figure 4.9. Side-by-side comparison of Native American site type frequencies from the Apache-Sitgreaves and Coconino NF study areas………. 125 Figure 4.10. Density of small structure sites in the study area, revealing the Mogollon Rim cluster in the southeast, and the Chevelon/Chavez cluster along the northern boundary, to the northwest (compare with Figure 4.6)…………………………………………………… 127 Figure 4.11. Density of larger pueblo sites (believed to include five or more rooms) in the study area………………………………………….. 129

LIST OF FIGURES - Continued

Figure 4.12 Locations of the Mogollon Rim and Chevelon/Chavez site clusters within the study area, showing previous survey in the area……….. 131 Figure 5.1. Red Knoll, a minor point of topographic prominence among an area of many small architectural sites, as well as the interface of juniper woodland and grassland…………………………………… 137 Figure 5.2. Histogram of elevation values for 2,000 random points located within the ASNF study area………………………………………. 139 Figure 5.3. Histogram of elevation values for 2,000 random points located within previously surveyed areas in the ASNF study area………… 139 Figure 5.4. Histogram of elevation values for single room and jacal sites in the ASNF portion of the study area…………………………………. 140 Figure 5.5. Histogram of elevation values for field house sites in the ASNF study area………………………………………………………… 140 Figure 5.6. Histogram of elevation values for room block sites in the ASNF study area………………………………………………………… 141 Figure 5.7. Box plot comparing (bottom to top) elevations of random sample points within the AS Study area, random sample points within previously surveyed areas, field houses, single rooms and jacal structures, and larger room block sites……………………………. 142 Figure 5.8. Histograms of classified slope values for ASNF study area and site subsets…………………………………………………………….. 145 Figure 5.9. Terrain aspect in the ASNF study area displayed as a ratio of the total area in 45 degree intervals; the distribution exhibits the area’s north-sloping terrain……………………………………………….. 145 Figure 5.10. Distribution of field house site aspect measurements for the AS study area sites; values less than zero indicate flat locations identified by the aspect algorithm…………………………………. 146 Figure 5.11. Archaeological sites with surface architecture in the Chevelon/Chavez cluster and their association with ephemeral, intermittent, and perennial streams and springs…………………… 149 Figure 5.12. Archaeological sites with surface architecture in the Mogollon Rim cluster and their association with ephemeral, intermittent, and perennial streams and springs……………………………………………………………… 150 Figure 5.13. Comparison of ratios of total sites per TES (blue) and total amount of study area per TES unit (red)…………………………………… 154 Figure 5.14. The seven TES units in the ASNF study area with the most archaeological sites………………………………………………… 158

LIST OF FIGURES - Continued

Figure 5.15. Field houses, single room and jacal structures, and room blocks estimated to have five or more rooms by TES unit in the ASNF portion of the study area…………………………………………… 159 Figure 5.16. Archaeological site density probability model digitized from Plog (1981), resulting from sample surveys completed in the Little Colorado Planning Unit……………………………………………. 161 Figure 5.17. Plog’s (1981) SYMAP model and the Mogollon Rim and Chevelon/Chavez site clusters…………………………………….. 163 Figure 5.18. Site location probability model developed by the ASNF following the Rodeo-Chediski fire…………………………………………… 166 Figure 5.19. Inductive probability model and the Mogollon Rim and the Chevelon/Chavez site clusters…………………………………….. 167 Figure 5.20. An example of elevated plains in piñon-juniper woodland typical of the broad swales regularly identified as locations suitable for agriculture by survey crews working in the Chevelon/Chavez site cluster……………………………………………………………… 171 Figure 5.21. Mean annual rainfall (cm) derived from TES unit ranges in the ASNF study area…………………………………………………... 174 Figure 5.22. Mean annual frost free days derived from TES unit ranges in the ASNF study area…………………………………………………... 175 Figure 5.23. Mean annual snowfall by TES Unit in the ASNF portion of the study area and the Mogollon Rim and the Chevelon/Chavez site clusters…………………………………………………………….. 176 Figure 5.24. Examples of soils derived from sandstone (A) and limestone (B) parent materials. Many of these soils are intermixed within the TES units and cannot be easily separated at the scale used to create the survey, although variation is readily apparent on the ground…. 179 Figure 5.25. Length of the frost-free season expressed by the linear equation of Dean and Kaldahl (1999) derived from historical weather records.. 182 Figure 5.26. Predictive model of annual precipitation, interpolated from (Dean and Kaldahl 1999:17)……………………………………………… 184 Figure 5.27. Results of agricultural risk model combining interpolated surfaces of annual precipitation and frost-free days. Blue zone is susceptible to short growing seasons, green zone appears to be highest potential growing season and precipitation, yellow is sufficient growing season low precipitation, and red is extremely low precipitation (but low risk of frost)………………………….. 187 Figure 6.1. Density of small architectural sites and room blocks in the study area……………………………………………………………….. 199

LIST OF FIGURES – Continued

Figure 6.2. Locations of Great Kiva sites (shown in bright green) in the study area and interpolated density of small architectural sites (number per square km)…………………………………………………… 202 Figure 6.3. Elk pictograph from rock shelter on the CNF, near the western margins of the Chevelon/Chavez site cluster……………………. 221 Figure 6.4. Bird eggs tucked within a pocket of grassland amidst the Ponderosas, one of the many fortuitous finds encountered during a hunting trip, a gathering foray, or an archaeological survey. 222 Figure 6.5. Interpolated density of artifact scatters (sites that include Ceramic period components)………………………………………………. 223 Figure 6.6. Distributions of archaeological sites whose primary feature types include presumed agricultural functions, such as check dams, terraces, water control devices, rock alignments, and agricultural areas……………………………………………………………….. 228 Figure 6.7. The view from the Mogollon Rim on the ASNF southwest toward the mountainous transition zone and the headwaters of the Tonto Basin………………………………………………………………. 236

LIST OF TABLES

Table 3.1. Chronologies for portions of the study are and surrounding region… 69 Table 3.2. Various typologies proposed for small sites in the study area and surrounding regions…………………………………………………. 90 Table 5.1. Descriptive statistics for elevation of random points, random within surveyed areas, field houses, other small sites, and room blocks in the ASNF study area………………………………………………… 138 Table 5.2. Terrestrial Ecosystem Survey units within the ASNF portion of the study area and corresponding site associations and proportion surveyed……………………………………………………………... 152 Table 5.3. Narrative topographic, soil, snow cover, and drainage types for the TES units that include the most archaeological sites (USDA 1989)… 156 Table 5.4. Single rooms, field houses, and room block sites by probability classes in the Sitgreaves inductive probability model……………….. 168 Table 5.5. Single rooms, field houses, and room block sites by probability class within the Mogollon Rim site cluster in the Sitgreaves inductive probability model……………………………………………………. 168 Table 5.6. Average rainfall, snowfall, and frost-free days by TES unit in the ASNF study area…………………………………………………….. 172 Table 6.1. Some common pool resources in the study area and their characteristics………………………………………………………... 219

ABSTRACT

This dissertation explores some of the thousands of smaller Native American archaeological sites with meager architectural elements commonly found along part of the southern edge of the Colorado Plateau in east-central Arizona in an area known as the

Mogollon Plateau. Small surface structures of less than five rooms were typically built of a combination of stone masonry and wattle and daub, and they are generally interpreted as evidence of repeated occupations of limited duration, primarily dating between AD

800 and 1300. Accordingly, these small sites have also served a number of roles in ongoing discussions of settlement systems and land use, and they present challenges for cultural resources management. The fundamental characteristics (or lack thereof) typically used to classify small sites have traditionally relegated them to settlement pattern studies rather than extensive excavation, generating a broad range of hypotheses concerning their significance and drawing heavily upon historical ecology. GIS methods are used to explore several ecologically and socially-driven models and examine the roles of small architectural sites in archaeological and systemic landscapes. Common pool resources offer some explanatory power regarding small sites, but some have suggested competition and conflict led to a “tragedy of the commons” and environmental degradation. Two primary site concentrations are identified, and the evidence supports an interpretation of extensive and sustainable use of the area, much of which seems to have been a frontier. Recommendations for research-driven management and preservation of cultural resources are provided. 14

CHAPTER 1.

INTRODUCTION TO THE SMALL SITES “PROBLEM”

AND GEOGRAPHIC INFORMATION SCIENCE SOLUTIONS

Some of the most common Native American archaeological remains encountered in many areas of the American Southwest consist of little more than modest architectural features ranging from curvilinear rock alignments to more formal constructions of a few rectangular rooms (Figure 1.1). Typically, the associated surface artifacts are sparse and unremarkable assemblages, usually capturing limited attention from survey crews. Along the southern edge of the Colorado Plateau, these sites have been assigned a number of sometimes interchangeable labels, ranging from agrarian inspired field houses and farmsteads, to the more obscure, like “carports” and isolated rooms. Research in other regions has elicited other colorful terms, including summer homes. Many have been less optimistically recorded as artifact scatters, falling victim to the fate of construction materials that do not persist in the archaeological record, or the scant architectural remains being obscured from view by a subtle veneer of vegetation or recently deposited sediment. In other cases, natural outcrops at other sites have been mistaken for architecture, often only becoming clear after dramatic alterations to the environment, like wildfire or excavation (e.g., Haines et al. 2004).

The degree to which actual variation within the documented archaeological record has been obscured or misinterpreted by classification is particularly troublesome 15

Figure 1.1. An alignment of several cobbles and small boulders of limestone on the Apache- Sitgreaves National Forest, near Red Hill. given the many implied functions associated with the various site typologies investigators have employed (Figure 1.2). Archaeological and systemic context are implied in different ways by site typologies, and presumed functions associated with specific site types may derive from questionable or merely speculative inferences.

Although these small architectural sites are some of the most commonly encountered phenomena identified by archaeologists who undertake large-scale cultural resources inventories for various types of projects, researchers have traditionally, and probably justifiably at least to a certain extent, focused upon more

Figure 1.2. Some examples of small site types and their associations with archaeological and systemic context.

spectacular structures, especially the large Ancestral Pueblo sites of the southern

Colorado Plateau occupied during the final few centuries before and during European colonization (e.g., Fewkes 1904; Haury 1962, 1985; Herr 2001; Hough 1903; Mills et al.

1999; Plog 1981a and b; Plog 1984; Upham 1982). The large villages where Ancestral

Pueblo communities coalesced and thrived for generations and the marvels like cliff dwellings and rock shelters are some obvious examples that have attracted the focus of researchers. Many of the investigations of these sites laid important foundations for understanding the archaeology of the region’s Native American traditions, but the smaller and less spectacular remains have not usually received an appropriate proportion of the attention given their abundance.

During the past two decades this legacy of research in the American Southwest has faced the changing character of archaeological research, and more active (and often legally mandated) cultural resources management undertakings have directed more 17 attention to the less impressive categories of archaeological sites (e.g., Ciolek-Torrello et al. 1990, 1994; North et al. 2004; Vierra and Ford 2007). Simple artifact scatters, perhaps with potential for subsurface deposits and architectural remains, are among the most frequently recorded types of sites; previously undocumented, large habitations are less rarely encountered. Within the Mogollon Plateau region, many investigators agree we are unlikely to discover new large pueblos of more than 100 rooms, although the possibility exists. Regardless, there are some indications that large sites may not always be the appropriate location to seek archaeological information to address certain questions. Rice’s (2001:3) review of field houses and farmsteads among the Hohokam led him to conclude that “large Hohokam sites are not the optimum context for documenting the subsistence practices of a particular local. Rather it is field house sites and farmsteads that are the key locations for documenting regional variability in

Hohokam subsistence.”

Ward and others (1978) presented a landmark contribution to the study of “small sites,” defined as “ones whose size and artifactual assemblage suggest a limited temporal occupation by a small group of people, gathered at the locality to carry out a specific, seasonally-oriented set of activities” (Pilles and Wilcox 1978:1). This baseline definition and the accompanying studies proved to be an important contribution to Southwest archaeology and continue to be cited with some regularity, uniting a number of important perspectives that explored variability among small sites “as components in larger settlement-subsistence systems” (Pilles and Wilcox 1978:2). Given the changing nature of archaeological investigations in the Southwest, as well as an increasing amount of 18 available data regarding both archaeological surveys and environmental variability, new synthetic investigations of small sites are warranted.

This dissertation explores some of the thousands of associated smaller sites with less ostentatious architectural elements that are also generally attributable to Ancestral

Pueblo occupations along part of the southern edge of the Colorado Plateau (Figure 1.3).

Those sites that show evidence of small surface structures (five rooms or less), typically built with a combination of stone masonry and wattle and daub, have commonly been interpreted as evidence of repeated occupations of limited duration and primarily date between A.D. 800 and 1350. Accordingly, they have also served a number of roles in ongoing discussions of settlement systems, and researchers have drawn upon some of the most prominent theoretical perspectives in modern archaeological research, especially in the American Southwest (e.g., Barton et al. 2006; Hantman 1983, 1989; Haury 1985;

Herr 2001; Kohler 1992; Lightfoot and Plog 1984; Lightfoot and Most 1989; McAllister and Plog 1978; Netting 1993; North et al. 2003; Peeples et al. 2006; Pilles 1978; Plog

1978; Preucel 1987, 1990; Reid 1982). The fundamental characteristics (or lack thereof) typically used to classify small sites have traditionally relegated them to settlement pattern studies rather than extensive excavation. As these studies have continued to mature, they have generated a broad range of hypotheses concerning the significance of small sites in broader settlement systems, drawing heavily upon historical ecology, land use studies, and landscape archaeology.

Figure 1.3. The study area outlined in yellow and corresponding Ranger District names overlying shaded relief Landsat 19 imagery, with locations of modern towns and major roads (black) and interstates (red). The Mogollon Rim coincides with much of the southern boundary of the study area.

Research Goals

Despite a wide array of approaches, there is a central theme among most of the hypotheses researchers have presented regarding small architectural sites. Virtually all concur that the groups who constructed and occupied small sites in a myriad of ways shared an intimate relationship with their surroundings that encompassed intertwined subsistence practices and ideological beliefs. Interactions with long-lasting effects, including cultivation of domesticates and a wide array of horticultural pursuits, played out on many different levels. This probably accounts for the fact that the most commonly reoccurring hypotheses have strong undertones of environmental determinism. However, we have begun to recognize the social and ideological significance of these trends as well, which are playing active roles in research and cultural resources management alike.

I examine these issues in greater depth through a detailed analysis of the archaeological geography of small sites located in the westernmost portions of the

Apache-Sitgreaves National Forests (ASNF), while also drawing upon some portions of the easternmost portions of the Coconino National Forest (CNF) (Figure 1.3). The study area includes the Black Mesa and Lakeside Districts of the Sitgreaves National Forest

(managed jointlywith the Apache as the ASNF), and portions of the Mormon Lake and

Mogollon Rim Districts of the CNF. Within areas now managed as part of the National

Forest System, several generations of archaeological investigations have given rise to a peculiar nomenclature that often implies function within settlement systems, while simultaneously and necessarily obscuring variation. I demonstrate some of the important

21 roles archaeological geography can play in the integration and assessment of underlying assumptions and hypotheses concerning settlement systems, especially those that incorporate small sites. Following recent and widespread adoption of GIS innovations, I advocate more widespread recognition for Archaeological GIS (AGIS), validating the growing importance it plays in mundane issues of cultural resources management as well as the growing potential for synthesizing, analyzing, and presenting the results of archaeological research. It is particularly well-suited to the archaeological phenomena examined in this dissertation. As Doolittle’s (2000:5) landmark review of Native

American cultivations notes,

Accepting that farmers operate on the earth’s surface, that both natural and human

forces constantly affect farming practices, that cultivation involves numerous

elements, including plants, soil, moisture, and temperature, and that spatial or

regional patterns which can be mapped are created as a result, agriculture,

especially that of native North America, is as much a geographical as it is an

anthropological topic.

Because the archaeological record suggests these small sites played such an important role in settlement systems prior to European colonization, they should contribute accordingly to our interpretations and representations of these societies. By situating my analysis within the realm of archaeological geography, I evaluate inferences about the different roles that these sites played in settlement systems, as well as the existence of

22 less accessible aspects of the archaeological and systemic landscape (Heilen 2005; Heilen et al. 2008).

GIS methods are used to explore several hypotheses regarding the significance of small architectural sites, including some of their presumed associations with agriculture.

Their associations with the environment and other archaeological sites are explored through a review of inductive and deductive approaches. In doing so, I explore the concept of common pool resources, which may be an appropriate way to characterize past use of the area. Is there evidence for long-term, sustainable uses of common pool resources, or does the period when small architectural sites appear to be most abundant indicate a scenario that unfolded into a “tragedy of the commons”?

Organization of the Dissertation

In Chapter 2, I review the concept of archaeological geography and its relationships with recent developments in the field of landscape archaeology. I also explore their relationships with archaeological applications and present the methodology used to operationalize GIS studies of small site. This includes an exploration of deductive as well as inductive methods. The following chapter reviews the archaeological setting and provides context for research of small sites in the Mogollon

Plateau region, paying particular attention to examples from areas that are now managed as parts of the National Forest system. The cool pine forests and incised canyons have attracted generations of Southwestern archaeologists, and the legacy of their attention to its large, late sites provided an important foundation for understanding the past. After a

23 lengthy lull in research, the area became the subject of early advances brought about by the New Archaeology, spawning inventive approaches to applied archaeology and cultural resources management. Re-energized academic interest during the beginning of the era of federal historic preservation combined with these efforts to produce a significant amount of data regarding the area’s archaeological geography. In Chapter 4, I review inventories of small sites within the study area, highlighting the classification schemes and some of their underlying assumptions. The formation of this inventory and the associated sources of error are often complex but also reveal issues applicable in other regions. Some past excavations of small sites are also reviewed, along with an introduction to the subsets of sites included in more detailed analyses that follow.

In Chapter 5, I situate these archaeological sites within the environmental setting of the Mogollon Plateau. After introducing the primary sources of data, I formulate and test simple models based on hypotheses concerning the presumed significance of the distributions of small sites. This provides an opportunity to critically examine the legacy of research regarding some of the most common types of prehispanic archaeological sites and presumed interactions of various environmental and landscape phenomena.

Chapter 6 tackles the difficult issue of situating small sites within the broader social setting. As Fish and Fish pointed out in the landmark 1978 study on small sites,

“The strongest conclusion is an unavoidable necessity to consider the entire social setting of the small site in order to achieve a meaningful interpretation” (1978:57). These characteristics are difficult to assess, but spatial relationships among sites suggest the presence of identifiable communities, and variations in artifacts and architectural styles

24 suggest potential distinctions among community groups. As the analysis shows, many of these issues are difficult to assess from the state of existing databases. However, there are some convincing examples of evidence of social conflict and change through time, as well as cooperation and community integration. I also address potential vectors of anthropogenic influence upon environmental conditions. There are also important insights regarding potential spiritual significance of landscape components, particularly the historically documented shrines and other traditional places throughout the area. I argue that common pool resources played an important role in land tenure and resource acquisition in the region, and speculation about long-term environmental degradation and depletion of common pool resources warrants more critical analysis and attention to formation processes.

The final chapter synthesizes the findings and offer insights for the roles AGIS can play in the management and preservation of archaeological landscapes in the study area and beyond. Recommendations for future research can help direct research-driven management and preservation of these sites. Increasingly large wildfires presumably brought about by a legacy of past overgrazing and fire suppression have spurred large- scale fuel reduction projects, including expansive use of prescribed fire, recent intensive thinning of small diameter trees through the Healthy Forest Act, and future proposed projects through the Four Forest Initiative, a collaborative project among National Forests that manage much of the expansive Ponderosa pine belt that includes much of the study area. Models of archaeological potential and sensitivity are proposed as tools for planning and managing future projects. These applications will be useful for other

25 agencies and researchers. These analyses highlight the roles that AGIS can play and also demonstrates the influences of various vectors of scale and resolution upon, providing important insights that can be applied to surrounding regions, much of which is also public land facing a growing number of pressures, and these formation processes are an important consideration.

CHAPTER 2.

ARCHAEOLOGICAL GEOGRAPHY

AND GIS FOR LANDSCAPE STUDIES

There has been a long history of interest among Southwest archaeologists regarding resumed interactions and relationships among archaeological sites and ecological, environmental, and even ideational landscape characteristics. Enduring archaeological concepts like settlement patterns, land use history, and landscape studies have expanded archaeological focus from excavation contexts to much broader scales, but their intricate relationships with the concept of archaeological geography have not been fully explored. In this chapter I review the concept of archaeological geography and the growing roles GIS is playing in studies, paying particular attention to the diverse intersections among social and spatial landscapes of Southwest archaeology. I introduce archaeological geography as a unifying concept for the roles that GIS can play in archaeological research and cultural resources management. This synthetic discussion offers an important theoretical background for understanding the significance of small sites. I also review the development of predictive modeling applications and provide a framework for exploring distributions of small archaeological sites and their potential contributions toward research-oriented cultural resources management.

Archaeological Geography

The relationship between archaeology and geography traces its deep roots to ancient Greek, Babylonian, Chinese, and Roman societies and the timeless human curiosity for the past. Geography is a remarkably holistic discipline, encompassing countless natural properties and phenomena of the terrestrial environment, largely the domain of physical geography, as well as the implications of myriad influences of human behavior upon the environment, commonly referred to as human geography. Modern physical geography has grown to share interests remarkably akin to scholars in fields like ecology, geology, hydrology, and climatology. In a similar sense, human geography, which includes the study of human interactions with environmental characteristics and the resulting patterns among many other topics (including historical, political, economic, and population geography), is remarkably akin to many of the preoccupations of anthropological archaeologists. Historical geography and environmental history also approach many ecological and environmental studies undertaken by anthropological archaeologists (Williams 1994). These links have been reaffirmed by the ongoing nature of settlement pattern studies and the burgeoning growth of various “landscape” concepts in both fields.

Modern historical geography emerged during the first half of the 20th century, in concert with geography’s rise to the ranks of an independent discipline in many academic departments throughout the world (Butlin 1993). Conceptions of historical geography from this time are clearly associated with the development of settlement pattern studies,

28 including many that speak to issues of environmental determinism and human ecology

(Butlin 1993; Chakrabarti 2001; Trigger 1989). When grappling for a definition of the field of historical geography, many have noted a concern with “reconstruction of past geographies,” including developmental sequences for specific regions and their associated environmental impacts (Darby 1936).

Sauer’s The Morphology of Landscape (1941) encouraged geographers to examine “cultural landscapes,” and regional studies became a common approach to historical geography. His (1941:1) “protest against the neglect of historical geography” reaffirmed the importance of cultural landscapes and articulated the close relationship between archaeology and historical geography, emphasizing the importance of fieldwork and offering examples of “archaeo-geographic study.” Taking historical geography to task, he noted the need to study climatic variability and the reciprocal relationships upon humans, including human groups as geomorphological agents. He alluded to landscape- scale studies, including settlement patterns and land use history, as “the best recognized observational items used in reconstructing changes and continuities” (Sauer 1941:21), which he saw as another critical task for historical geography. In this sense, it seems

Sauer should be credited for coining the notion of archaeological geography, although current studies need not necessarily follow the Berkeley School approach and modern geography has largely abandoned human or cultural geography in favor of approaches that mirror changes in anthropology by including a greater diversity of critical approaches

(Bauder and Engel Di-Mauro 2008).

Doolittle’s notable examination of Cultivated Landscapes of Native North

America (2000) offers an interesting in-road to the modern divide that still seems to exist between geography and anthropology. It is “written from a decidedly geographical, not an anthropological perspective,” and he reaffirms his definition of geography (2005:5) as

“the study of the surface of the earth, with emphasis on the shaping processes and combinations of elements such as soils and human activities that result in distinctive regions. It involves understanding systemic relationships between certain elements rather than synthesizing diverse data.”

More recently, Chakrabarti (2001) invoked archaeological geography to address the historical developments of India’s Ganga Plain, a notably more complex review than I am attempting here. Mosher and Wilkie (2005) applied historical archaeo-geography to explore vestiges of federal legislation at the recent household level, and van der Leeuw

(2008) touted the rise of archaeo-geography in the study of socio-environmental interactions.

I define archaeological geography as the science of documenting, representing, interpolating, and analyzing the characteristics of archaeological landscapes and their associations with other characteristics of natural landscapes in order to understand systemic landscapes. Archaeological geography requires consideration of the formation processes of the archaeological record, ranging from the preservation of material remains to the methods used to collect, store, display, interpret, and visualize data gathered from archaeological landscapes. In addition to material remains and their settings within the modern environment, studies of archaeological geography require consideration of

30 experiences afforded by past environmental conditions along with the communities who inhabited them. This definition relies upon the important distinction between archaeological and systemic landscapes discussed by Heilen and others (Heilen 2005;

Heilen et al. 2008), presented below along with an overview of landscape archaeology, and followed by a discussion of how AGIS may be used to operationalize some of these landscape perspectives. Ecological approaches to systemic landscapes are easily criticized for failing to account for agency and individual action, and for failing to grasp the fundamental aspects of “man-land relationships” (Basso 1996:67). Different historical trajectories and contingencies highlight the importance of examining past land use from different perspectives, since a single process or point of view clearly cannot account for all situations.

Landscape Archaeology

Landscape archaeology has become a prominent theoretical perspective among contemporary archaeological researchers, providing a foundation for explorations of a wide variety of regions and time periods (e.g. Aston 1984; Bender 1992; Bowser and

Zedeño 2009; Bradley 1991; Bradley et al. 1994; Branton 2009; David and Thomas 2008;

Gartner 1999; Hicks 2002; Kelso and Most 1990; Ross 2001; Ruppel et al. 2003; Snead

2008; Tilley 1995, 2001; Van Dyke 2007; Whittlesey 1998a, 1998b; Zedeño 1997, 2000;

Zedeño et al. 1997). Its prominence among Southwest archaeologists led Fowles (2010) to identify a unique “Southwest School” within landscape archaeology. It is an important

31 critique of prior materialist perspectives that emphasized economic issues like resource extraction and energy potential at the expense of the experiential affordances of landscapes and associate processes of their social negotiation (Fowles 2009). Noting the influences of legislative changes, especially NAGPRA, and their influences upon archaeological research, Fowles (2010) highlights the resurgent roles ethnography as played in Southwest archaeology. The development of the Southwest school, in part as a critique to the strongly phenomenological British approaches to landscape studies, emphasizes the fact that multiple ontologies should inform archaeological research; collaboration with descendant communities provides an alternative perspective that offers different modes of perception and interpretations of archaeological information, as well as the natural setting and how it is idealized by different groups.

Landscape archaeology advocates a shift from viewing the environment as either a passive backdrop or a forcible determinant of culture, to a more dynamic perspective that seeks to identify social meaning by viewing landscape as a process that is continually renegotiated and is both an influence upon and influenced by human experience (Pool and Cliggett 2008). Landscape is thus seen as cognized environment, whereby behaviors and beliefs result in the formation of places. The physical and cognitive aspects of the environment are transformed into landscapes of memory, identity, and reflections of social order.

The inherent vagueness of a definition of landscape archaeology has made it attractive to researchers, despite differences of time, social complexity, and environment

(Branton 2009). This has encouraged a wide variety of different invocations of the

32 landscape concept in an attempt to comprehend the past and elucidate meaning from the places that archaeologists study and their surroundings, rather than simply focusing upon artifacts recovered within the confines of site boundaries and archaeological cultures.

Although the inherent vagueness of the landscape perspective has been lauded for its flexibility, it has often been difficult to operationalize (Anschuetz et al. 2001). However, it can encourage archaeologists to remember that sites were not isolated but represent intersections of multiple social landscapes, and the significance of a particular place is strongly determined by its associations with other places situated within these landscapes.

Although working with the historical frontier landscapes, Heilen and Reid (2009:135) identified “strategically cognized landscapes” as a useful way for understanding ways

“complex patterns of frontier settlement can thus be modeled in terms of the interaction of relatively simple, evolving strategies.”

Anschuetz et al. (2001:191) also emphasized the social networks might be inferred from landscapes, noting that they “are worlds of cultural product and represent the record of dynamic processes of human interaction with their environments,” optimistically noting that “material culture encodes information in patterned ways, [and] the use of inductive methods can decode archaeological observations to help make inferences about past meanings that underlie observed regularities and deviations.”

The landscape approach has also offered a realm for addressing sacred landscapes. As McPherson’s (1992) examination of sacred geography among the

Navajo, Welch’s study of Apache place names, and reviews of Zuni place names have indicated, important landscape characteristics often include features and locations that are

33 not readily recognized by standard archaeological survey methods (Basso 1996; Ferguson

1985; Ferguson and Hart 1985; Mills and Ferguson 1998; Senior 2004). Fowles (2009) applies these principals in his villagescape approach.

I advocate for a behavioral approach to landscape archaeology via archaeological geography, building upon concepts introduced by Heilen (2005) and Heilen, Reid, and

Schiffer (2008). Although they may seem misapplied to those more familiar with behavioral geography than behavioral archaeology, their distinction between archaeological and systemic landscapes recognize some of the inherent issues archaeologists must address, including formation process. Building upon Reid’s

(1995:19) updated distinction between archaeological and systemic context, Heilen

(2005:35) notes:

“Archaeological landscapes are the landscapes of archaeological materials and

evidence that archaeologists objectify and operationalize through instances of

scientific observation, inference, and analysis. Systemic landscapes are the

landscapes of past and present behavioral systems in which people–material,

human–environment behaviors and activities are performed. Although

archaeological and systemic landscape contexts overlap in time and space, the

properties of past systemic landscapes are inferred from the properties

archaeological landscapes.”

He goes on to note that “archaeological landscapes are the material correlates of a particular systemic geography on a given landscape at a given time,” while systemic landscapes “can be conceptualized as a spatio-temporal medium of human-material interactions …the sum total of landforms, natural and cultural resources, and human modifications to the natural environment with which a particular behavioral system interacted” (Heilen 2005:36-37). Although these definitions demonstrate notable differences between the two types of landscapes, both are networks that are formed through dynamic, hierarchical processes (Heilen 2005:37-39).

Many have warned that interpretive shortcomings regarding studies of landscape- scale behaviors frequently arise when an inappropriate scale of analysis is adopted or studies fail to account for the full breadth of a given group’s mobility or territory

(Hegmon et al. 1999; Lekson 1996; Nelson 1993a, 1993b; Nelson and Hegmon 2001;

Zedeño 1997). Regions should not necessarily be assumed to represent an entire sequence or occupational continuity, and narrow focus upon individual ecological localities may hinder interpretation when an interregional perspective is needed to identify cultural, rather than ecological differences (Wilshusen and Ortman 1999).

Ethnohistoric accounts and cross-cultural studies provide important context for interpreting archaeological landscapes, potentially indicating the importance of expansive viewsheds, territorial boundaries, and exchange networks that are otherwise difficult to identify. Their results should be useful tools for documenting, preserving, and shedding light on some of the less tangible vestiges of the past, such as travel routes, agricultural

35 and foraging areas, and sacred and ideological components of the landscape (Ferguson

2007; Knapp and Ashmore 1999; Minnis et al. 2006; Phillips 2009; Snead 2002).

An important challenge that remains for Southwest archaeologists is to operationalize the results of landscape studies into ongoing identification, interpretation, management, and preservation of cultural resources. In this sense, it is notable to remember cultural resources include far more than traditionally-defined archaeological sites alone. Landscape-scale aspects of an archaeological culture must be recognized alongside other less detectable affordances such as the imposing view of a sacred mountain, or the entry to a new spring. Ongoing cases in the Southwest like litigation surrounding the Arizona Snow Bowl on the Coconino National Forest and the Mount

Graham International Observatory on the Coronado National Forest demonstrate the significance of these issues. Growing efforts to formally recognize extensive Traditional

Cultural Properties on public lands, such as the Santa Rita and Santa Catalina mountains to the south and north of Tucson, ensure these will be important issues in the future and will need to be integrated with more traditional, site-based archaeological research.

The lasting impacts of prehistoric land use practices upon modern ecological conditions have received increasing attention (Barton et al. 2004; Duff et al. 2010;

Fertelmes and Barton 2007; Hegmon et al. 2006; Kohler 1992a, 1992b; Redman 1999), probably at least in part due to the prevalence of modern debates about anthropogenic climate change. Indeed, studies that correlate land-use practices and their relationships with climate change should be an important avenue for future research (Blinman 2008).

Enduring, long-term changes to soils brought about by past agricultural practices have

36 received growing attention (Holliday 2004; Homburg and Sandor 2010). Sustainability and resource conservation arise in these studies, particularly when addressing issues like migration and abandonment (Spielmann et al. 2011).

In some cases, it appears that settlement duration and land use intensity have been strong enough to sustain lasting effects upon modern ecosystems (Fertelmes and Barton

2007; Homburg et al. 2004; Peeples et al. 2006; Periman 2005; Roos 2008b; Sullivan

2000). The increasing attention that has been paid to potential modern associations between existing ecological conditions and archaeological phenomena highlight the potential impacts of past and modern land use traditions upon one another. In a similar vein, soil studies have also yielded complex and intriguing insights regarding prehistoric agriculture, although conclusions remain tenuous until fields and other related features are accurately identified and investigated (Homburg et al. 2004; Homburg and Sandor

2010; Wills and Dorshow 2012). It is important to temper these insights with careful consideration of formation processes and the impacts of more recent activities like grazing, vegetation management projects like chaining, and timber harvesting (Abruzzi

1999; Butzer 1996, 2002). While the focus of this review has been upon Native

American cultures, historical archaeology has an important role to play in recognizing and understanding the landscape-scale impacts of the more dramatic and obvious post- colonial land-use practices (e.g., Church 2002). Such studies are surprisingly under- represented in the literature. Although we must be careful to recognize the tenuous nature of some of our underlying assumptions, these conditions provide an important role for archaeological investigation of past settlement systems.

Lansing (1991) explored the concept of engineered landscapes in Bali, where impressive and permanent alterations to the physical environment resulted in changes to productive functions and social relations, and the resulting surpluses created by these efforts helped support complex sociopolitical changes. These behaviors also demarcated differential productivity across the landscape that probably also reinforced sociopolitical conditions, perhaps even becoming associated with ceremonial cycles and the community’s social memory. Earle and Doyel (2008) applied these concepts to the case of Hohokam canals, which may provide one of the best examples of this scenario in the

Southwest, although there are clear avenues for pursuing these ideas in other areas, such as the San Juan Basin and Chaco Canyon. Emphasis has also been placed on catastrophic events that can dramatically interrupt the evolutionary model presented by Earle and

Doyel (2008), and these issues have been explored by in more detail in a debate among

Ensor et al. (2003) and Waters and Ravesloot (2001, 2003).

These issues are critical in the examination of small-scale farming societies, common throughout much of the prehistoric Southwest, often requiring recurring settlement shifts to promote sustainability (Cameron 1995; Hard and Merrill 1992; Lipe

1995; Nelson and Anyon 1996; Nelson and Hegmon 2001; Sullivan 1992b). Shifts could occur either within or between regions, raising serious issues concerning occupation duration and seasonality and underscoring the notion that an archaeologist’s notions of abandonment, sedentism, and migration are not always consistent with Native American use and perception of the landscape in the past. Migration has been recognized as a critical component of life among many peoples of the American Southwest, and there

38 have obviously been many intermingled networks that have given rise to overlapping landscapes of memory. They have certainly varied in scale, both in relation to the social groups involved, distances traveled, and frequency, but they share common characteristics that they have brought about transformations, be they social, political, or ideological, and often in terms of contact among different cultures (Mills 2011).

Landscape studies offer a solid, integrative stage upon which archaeologists may communicate their goals and findings with interested communities, ranging from Native

Americans to recent immigrants, schoolchildren to the elderly, as well as commercial interests like agroforestry and range management. Sharing perspectives on landscape and land use can also provide important means for engaging descendant communities, sustaining a growing trend in the American Southwest (Fowles 2010). Consultation with

Native Americans can reveal insights into the myriad of agricultural practices (Doolittle

2000; Doolittle and Neely 2004; Doyel 1993), hunting and gathering territories (Kelly

1995) and other horticultural pursuits (Nelson 1993b). Kuwanwisiwma and Ferguson

(2009:98) report on research conducted in the Grand Canyon, where Hopis were able to identify high potential agricultural areas based upon associations of soil types and modern stands of plants such as greasewood, saltbush, and rabbitbrush. These types of interactions between archaeologists and tribal members can provide opportunities to re- connect communities with ancestral territories and traditions, while also aiding archaeological interpretation. These efforts can also be directly integrated into resource management plans to nurture these types of opportunities, as well as integrate tribal concerns into cultural and natural resource management decisions (Murray et al. 2009).

Although conceptualizations will differ, focus on landscape-scale trends can provide common ground for discussion among resource managers and descendant communities.

Advancements in landscape studies can also benefit modern non-Native populations. Often composed of relatively recent immigrants in the Southwest, modern communities impose land-use patterns that may conflict with traditional values as well as archaeological traces of the past. Land use studies can play an important role in helping direct goals of modern ecological restoration projects undertaken by land management agencies, many of which have good but perhaps misguided intentions to restore precolonial conditions. Land use studies provide an important opportunity to engage an interested public, as well as other social and physical scientists, in heritage issues and communicate advancements in the field. These efforts can also benefit preservation goals. As mentioned previously, historical archaeology’s contributions to land use studies in the Southwest have been somewhat limited. Some of the same issues that have been dealt with in prehistoric studies can be applied to historical contexts.

The landscape approach is particularly relevant for archaeological sites located on public lands, where the archaeological record has been afforded special consideration and protection. Cultural resource managers are responsible for identifying, interpreting, and, in many cases, preserving these sites located within ancestral territories of displaced

Native American groups. At the same time, public lands are being administrated in an attempt to balance multiple land uses with ongoing efforts to also “restore” ecosystems to

“natural” conditions that presumably existed prior to colonization. Often, these projects

40 take a landscape-scale approach, seeking to use mechanical vegetation treatments and prescribed fire over project areas spanning thousands of acres.

Archaeological GIS: Presence, Probability, and Prediction

Archaeological applications of GIS have become increasingly common during the past three decades, and they continue to offer potential for future advancements in research as well as cultural resources management (CRM) (Aldenderfer 1996; Conolly and Lake 2006; Gillings and Wheatley 2005; Kvamme 1989, 1999:155). The inherent spatial nature of most archaeological data encouraged archaeologists to invest in GIS applications, but advances in archaeological theory have also played an important role.

The increasing prevalence of CRM mandates elevated the powerful combination of relational databases and computerized maps to prominence, facilitating data management requirements. The visualization and analytical capabilities of GIS have also provided archaeologists with more accessible means of searching for patterns among complex data sets through new methods of modeling, prospecting, and spatial analysis (Kvamme 1999;

McCoy and Ladefoged 2009). The preponderance of Global Positioning Systems (GPS) mapping capabilities, remotely sensed data, and mobile computing has also facilitated integration of archaeological data with GIS.

Archaeology’s ongoing relationship with geography and growing use of GIS applications encouraged Llobera (2010) to argue Archaeological Information Science

(AISc) warrants consideration, particularly given the importance of scientific

41 visualization of data. He defines AISc as “the generation, representation, and manipulation of archaeological information within the context of information systems”

(Llobera 2010:218). His past work with visual landscapes (Llobera 2007) and the potential for other similar applications in archaeology, including simulation, demonstrate some of the uniquely inherent issues associated with understanding past landscapes, whether archeological or systemic. While his view of AISc is a useful approach to encourage future developments in visualization of archaeological data, I follow a more conservative approach and focus on more widely-accepted archaeological applications of

GIS and how they can be applied to address some lingering issues, such as small site distributions.

Improvements in our ability to compile and share data have encouraged multi- disciplinary studies that incorporate evidence from ever larger geographic areas, revealing important trends that are not always apparent from a narrower regional perspective. The availability of archaeological data has increased dramatically, and our ability to compile, analyze, and display information that includes large-scale geographic areas and diverse sites has supported many important advancements in regional studies

(e.g., Varien et al. 2007; Vivian et al. 2006; Wilcox et al. 2007). GIS has provided accessible platforms for exploring distributions of archaeological materials and their intersections with environmental conditions (McCoy and Ladefoged 2009). These realities have shed new light on some persistent questions in the Southwest, and they offer opportunities to address new issues in the future.

Modeling prehistoric site locations, especially in relation to subsistence resources and settlement patterns, has been an important concern of archaeological method and theory for more than 50 years (Maschner and Stein 1995:61). Developments in ecological anthropology, encouraged in large part by the work of Steward (1937, 1938) and his student Willey (1953), who essentially founded the study settlement patterns in

American archaeology, highlighted the importance of understanding relationships among archaeological site locations and features of the natural environment. Eventually, these perspectives encouraged more detailed studies of site locations, particularly during the

1970s (e.g. Euler and Gumerman 1978; Hodder and Orton 1976; Plog and Hill 1971).

Formal statistical analyses of point patterns were most common (Kohler 1988), but early developments in computing and GIS technology helped archaeologists began to overcome the difficulties of dealing with vast sets of complex archaeological and environmental data, giving rise to some of the earliest attempts to create predictive models. Eventually, GIS provided accessibility to even more complex analyses that were previously impossible due to computational intensity (Kohler 1988; Kvamme 1989:166,

1999, 2006).

More robust models of past landscapes have been produced as new methods of detecting, collecting, and analyzing evidence of past behavior are implemented. Satellite and terrestrial remote sensing offer increasingly available and accessible sources of data that can aid identification of various characteristics of the landscape, including the archaeological record. GIS provides an important means for collecting and analyzing this information (e.g., Lightfoot 2004), potentially providing insights to areas outside of

43 population centers (Wells et al. 2004). For example, Kruse (2007) examined potential for runoff agriculture in Perry Mesa, providing important insights to the relationships with habitation sites. Advancements in land use studies provide opportunities to refine identification and preservation efforts in cultural resource management. Traditional field methods should benefit from new technologies that are becoming increasingly available, and there has been a convergence with landscape ecology (Heilen 2005).

While archaeologists were beginning to understand advancements in early GIS technology and explore possible applications in predictive modeling, federal agencies, especially those working in the western United States, found themselves responsible for managing the cultural resources of vast tracts of land and meeting the responsibilities set forth in the National Historic Preservation Act and other legislation (Kvamme 1999).

Land management agencies sought out means of accurately predicting the locations of unknown archaeological sites based on the locational attributes of known sites (Kohler

1988). Aside from fulfilling their obligations to identify and document “significant” historic properties, they also hoped to facilitate management and planning early in the decision-making process (Berry 1984; Kvamme 1995, 2006; Warren 1990b). Despite warnings that predictive models were neither a panacea nor a substitute for fieldwork, they were lauded for their potential to improve efficiency in cultural resources management (Kohler and Parker 1986:398; Kvamme 2006; Sebastian and Judge 1988).

Although many predictive models have proven useful, most would agree they have fallen short of delivering the idealistic promises many archaeologists hoped for, giving way to a more realistic understanding of the complexity of the issues

44 archaeologists seek to model, as well as data requirements. Nevertheless, the past two decades have witnessed a dramatic increase in the use of predictive models (Kvamme

2006; Mehrer and Wescott 2006). Among the wide array of potential approaches, logistic regression probability modeling has emerged as the most common and successful multivariate technique, being recognized for its supposed potential to examine “specific landscape requirements that went into prehistoric decision-making processes” (Maschner and Stein 1995:72; Warren 1990a, 1990b).

As some argued that predictive modeling research stagnated and could only be reinvigorated by shifting debate from methods to theory (Church et al. 2000), researchers responded during the past decade with a proliferation of new approaches founded upon the realization that specific situations require specific models. Dixon et al. (2005) developed a predictive model to identify locations where archaeological materials might be exposed at the edges of melting glaciers and ice patches. Duke and Steele (2010) developed a weights-of-evidence model to predict the likelihood of geological deposits to produce suitable material for stone tool production and their associations with Middle and

Upper Paleolithic sites in Europe. Graves (2011) highlighted the importance of distinguishing site categories and regions within a study area in her applications of predictive models for Neolithic sites in Scotland. The theoretical basis of predictive models has also received renewed attention and challenges (Kamermans et al. 2009;

Verhagen and Whitley 2011).

While regional land use studies are important advancements in Southwestern archaeology, the underlying data upon which they are constructed retain limitations

45 regarding sample size, observer bias, objectives, intensity, and quality (Kowalewski

2008). Data derived solely from surface survey contexts offer particular challenges, and the abstractions necessary to incorporate them into regional studies should require careful consideration. Site and phase-based structures of place and time may impose unsuitable categories for addressing the nuances of land use systems. These issues have been addressed in part through more holistic approaches to settlement systems, which have emphasized multi-scalar perspectives on both space and time (Kowalewski 2008;

Kulisheck 2003; Minnis 2000; Nelson and Anyon 1996; Snead 2004; Wandsnider 1998;

Wandsnider and Camilli 1996). Failure to account for the full breadth of a group’s territory or their degree of mobility can wreak havoc upon archaeological interpretations.

Occupational duration and seasonality may be important considerations as well, and narrowly focusing upon particular ecological settings or relationships may limit interpretation when an inter-regional perspective is needed to identify cultural rather than environmental influences. Roth (2011) reports on recent excavations from the Mimbres

Mogollon region and warns that the presence of earlier components obscured by surface characteristics have the potential to re-write previously accepted notions of prehistoric land use. The terminal culmination of long-term adaptations may receive undue attention at the expense of the persistence of land use strategies.

Although archaeologists are usually quick to espouse the power and potential of

GIS innovations, the flip side of this fact is the ease of producing inaccurate or misleading cartographic representations of archaeological phenomena. Archaeological site location information is protected under federal and local statutes and should not be

46 disseminated indiscriminately. Data security is a critical issue that has not received enough attention, and the ability to disclose thousands of site locations with the click of a button is troubling.

GIS applications play an increasingly important role in heritage resource management programs and archaeological research, and they will continue to do so, especially in regions now managed as public land (as is the case for most of the study area). This may be particularly true in light of increasing attention to the intersections and interrelationships among ecological and human systems, a long-standing interest of anthropologists and geographers alike. Cartographic representations play an important role in this process, but offer a number of challenges, including the need to maintain confidentiality regarding site locations. Protecting site location information has been a particular concern in my own research. I obscure particular locations of sites through the use of site density surfaces throughout the text. In cases where point symbols are presented, they are at such a scale that it would be difficult to determine the exact location. Individuals interested in obtaining site location information for research or other purposes should contact Heritage program personnel at the appropriate National

Forest. Site and survey data has not been fully integrated into AZSITE, a GIS application intended to consolidate information about cultural resources throughout the state of

Arizona.

Reconciling Inductive and Deductive Approaches

to Predictive Modeling

The potential benefits and shortcomings of predictive modeling have been explored in part through an ongoing debate centered on the distinction between so-called inductive, or correlative, and deductive, or explanatory, models. Reexamination of this debate may help alleviate perceived problems associated with the use of predictive modeling in archaeology and reveal opportunities for future developments. Although

GIS provides a set of tools that are neither inherently deductive nor inductive, the ways they are used can present a perceptible dichotomy, with many distinguishing inductive modeling approaches from deductive ones (Church et al. 2000; Kohler 1988; Kohler and

Parker 1986; Kvamme 1999, 2006; Sebastian and Judge 1988; Verhagen and Whitley

2011; Warren 1990a). Many believe this distinction reflects the maturation process of any scientific discipline, which includes transitions from descriptive phases towards those that focus on explanation (Sebastian and Judge 1988). In archaeology, this has entailed transitions from simple documentation of the archaeological context toward attempts to apply this data to explain changes within the systemic context. Clearly, both approaches are needed, along with attempts to incorporate alternative ontologies in order to achieve greater diversity in approaches and interpretation.

Deductive modeling processes typically begin with theory regarding human behavior in the systemic context, with the goal of identifying law-like principles that may be used to predict archaeologically significant locations. Inductive approaches begin with basic data, typically from archaeological survey projects, and seek to identify statistical relationships with environmental variables to estimate the locational

48 characteristics of the population of archaeological data from which the sample was selected. This distinction focuses on the “procedural logic” of the modeling process

(Kohler 1988:35-37; Sebastian and Judge 1988:4).

The inductive approach has been more frequently and successfully used to create predictive models (Kvamme 1999; 2006). Sometimes referred to as empiric-correlative models, these models rely on statistical inferential procedures intended to “correlate sample site locations with environmental features and forecast unknown sites in areas with similar environmental features” (Kohler and Parker 1986:400), potentially providing robust means for assessing the significance of different environmental variables without relying upon conjecture or subjective judgment to determine and quantify significance

(Kvamme 2006:12). Inductive models are also (relatively) easy to implement compared to the deductive approach, although significant effort is still required to compile, evaluate, manipulate, and interpret the data.

The benefits of inductive models have been highly touted for management and planning purposes, but the research potential of such models has been questioned.

Although they have become common practice for effectively predicting archaeological site locations, critics have repeatedly lamented several perceived shortcomings (Ebert

1988, 2000; Kohler 1988; Kohler and Parker 1986; Sebastian and Judge 1988; Verhagen and Whitley 2011). Many of these arguments reflect Leibenstein’s (1976:13) observation that mere “clairvoyance” is not a scientific pursuit if it is divorced from the power of coherent and communicable explanation (Kohler and Parker 1986:397). As Premo

(2004:855-856) advises, “although identifying a global trend is a necessary first step in

49 exploratory spatial analysis, in many cases, quantifying localized patterns of spatial dependence is more important to formulating detailed, defensible archaeological interpretations.”

Ultimately, inductive models result in correlations between the potential presence of archaeological remains and certain modern, mapped, environmental characteristics.

Critics note that these correlations are associated with contemporary characteristics, both of the natural environment and the archaeological record, rather than aspects of the environment that were significant in the past and were dealt with in the systemic context

(Ebert 2000:130; Kvamme 2006:19; Sebastian and Judge 1988:5). More importantly, correlation is not the same as causation, and these models do not necessarily help us understand why archaeological materials are more likely to occur in certain locations.

Pursuit of explanation requires moving beyond this inductive process and seeking ways of linking these findings with the past, rather than pursuing the inductive approach as an end in itself (Ebert 2000).

Inductive models simply project the characteristics of the observed patterns among a sample to the sample universe (Sebastian and Judge 1988:5-6). Their reliance upon correlation with mappable environmental data may obscure the complex and, sometimes, indirect interactions of groups with the environment. This can become a particularly important consideration when the sample of archaeological data may represent a palimpsest of different cultural traditions that adapted to different, changing environmental characteristics at different times. Also, the ability to obtain environmental data can lead to its uncritical use, perhaps encouraging reliance upon data of questionable

50 accuracy, as well as misinterpretation of the significance of the mapped data (e.g., “blue line” drainages) (Ebert 2000). Environmental data of inappropriate scale (such as broad categories that assume uniform distribution) can also lead to spurious correlations

(Church et al. 2000:142).

Despite much criticism, the inductive approach has been most frequently employed, and the alternative may be even more problematic than the shortcomings of inductive models. Deductive models must be based upon assumptions regarding the systemic nature of human behavior and the structure of the natural environment. This approach may allow archaeologists to address the goals and means of decision-making processes in the past, but it is incredibly difficult to operationalize and validate scientifically (Kvamme 2006:12; Sebastian and Judge 1988:8). Our limited understanding of how settlement systems functioned in the past and how groups conceptualized their interactions with the environment and ideological landscape characteristics, makes it difficult, if not impossible, to derive the types of law-like statements that are fundamental for the formulation of meaningful deductive models

(Altschul 1988; Kvamme 1999; Sebastian and Judge 1988). Also, it is important to consider the roles of human agency in the formation of different trajectories.

Predictive variables are frequently derived from ethnographic sources, but this can be a troublesome endeavor in itself due to the potential pitfalls of ethnographic analogy, as well as our inability to identify inconclusive connections among decision- making processes and environmental variables (Kohler and Parker 1986:432-439;

Kvamme 2006; Maschner and Stein 1995). Also, the statistical methods that are so

51 closely tied to inductive approaches provide a robust method for obtaining weights that can be used to combine variables in an appropriate manner. A significant weakness of deductive approaches is the relative weights assigned to variables are “often based on little more than conjecture” (Kvamme 1999:173, 2006). For these reasons, the explanatory promises of deductive modeling have lagged behind the utility of inductive models.

Kvamme (1999, 2006), who played an important role in the development of the field, has championed this argument and believes “they need not be different but can and should be one in the same (2006:13). He notes that many critics of inductive modeling fail to recognize the fact that variables included in inductive models are often selected through deductive reasoning (Kvamme 1999:173). Still, the debate has proven useful for critically examining different aspects of the modeling process and illustrating the problems and limitations that archaeologists have faced in developing useful, meaningful models. Attempts to combine the two approaches may prove most useful for future applications (e.g., Finke et al. 2008), and this underscores the idea that modeling should be a reflexive, ongoing process.

Particular situations may prove more amenable to one approach or another or a combination, and ongoing analysis of the utility of models and potential combinations of different approaches will likely prove to be most useful. Locally adaptive models have been proposed to adjust to particular landscape characteristics and incorporate variable catchment areas without violating mathematical principles (Carleton et al. 2012). Also, there appears to be ample room to explore the translation of model results into

52 meaningful archaeological interpretations (Graves McEwan 2012), as well as useful applications for fieldwork, consultation, and planning.

Operationalizing Archaeological GIS for Small Site Studies

Having reviewed the enduring relationships between archaeology and geography, the development of landscape archaeology, and the growing roles GIS is playing in landscape-scale studies of archaeological landscapes, I now present an overview of the methods proposed to explore distributions of small architectural sites of the Mogollon

Plateau region. This framework is used to operationalize a GIS-based analysis of the archaeological geography of small sites by comparing inductive and deductive models of the archaeological and systemic landscape relationship they may represent.

Although specific to the study area, the general approach should prove useful for other regions where small sites are widespread and they impose vexing research questions and management issues.

A flowchart (Figure 2.1) depicts the general analytical approach pursued in the following chapters. Beginning with a review of past archaeological survey projects and previously recorded archaeological sites, major trends are identified within the data by assessing the typologies of small site types have been proposed and implemented for

CRM purposes. These typologies are used to reclassify the site data into subsets. Some of these subsets may presume some systemic functional significance (e.g., field houses are small structures located adjacent to agricultural fields and associated with their use),

53 whereas others allude to simple architectural forms (e.g., single room) or surface assemblage composition (e.g., lithic scatter). While the locations of previously recorded sites are critical, locations known not to contain sites are important for evaluating the accuracy of proposed models (i.e., do models accurately predict the absence of sites?).

In the same way that various categorical typologies have been deductively applied to small sites, various researchers working in the region have posited hypotheses regarding the significance of small site locations and their potential associations with environmental characteristics and locations of other site types. These are reviewed in more detail in the following chapter to identify deductive models and appropriate test parameters to assess the significance of these hypotheses. Ethnohistoric accounts are also reviewed to suggest possible influences on socio-ecological relationships and possible ideational landscape components.

Known site locations and randomly selected sets of non-site locations are compared to assess whether site locations are random or not (null hypothesis), or if there are significant associations with environmental parameters derived from Digital

Elevation Model, the Terrestrial Ecosystem Survey, and other parameters. Inductive probability models are generated for sites in the area are compared to deductive models assessing agricultural productivity, site density, proximity to large and late sites, and other indications of community organization. These models are used to assess evidence of cooperative and negotiation of landscape affordances, as well as evidence of competition for access to common pool resources and how these patterns may have changed through time.

Survey data

Site data

Presence Absence Identify deductive models Subsets Null sets

Association Agricultural Field Single 1 2 with large Suitability Houses Rooms late sites? Model 3 4 Room Other Competition Common Blocks Sites Violence? Pool Resources?

Elevation TES units Soil Vegetation Climate

Inductive probability model Evaluate Random or Significant?

Evaluate vs. null sets

Compare results

Tragedy of the Commons or Sustainable Common Pool Resources?

Management recommendations Applications Future research

Figure 2.1. Flow chart of the examination of small architectural sites in the study area.

The results of this comparative modeling process are evaluated to assess the archaeological landscapes of the Mogollon Plateau region. The lengthy history of archaeological research in the area indicates a long period of low population density followed by a dramatic population increase. Past uses of the landscape indicate extensive use of certain regions characteristic of common pool resources. How do these landscape- scale models of past land use and landscape associations correlate with previous archaeological research? The results are used to evaluate whether or not there is evidence for sustainable use of common pool resources, and if a tragedy of the commons scenario may have brought about dramatic settlement changes, including aggregation into increasingly larger sites and limited occupation of some areas in the final few centuries before European contact.

These results are compared with interpretations derived from past inventory and excavation projects, as well as ethnohistorical accounts of land use history and landscape affordances. Cartographic representations of models developed for the study area should prove useful for assisting resource management decisions and consultation with descendant Native American communities, and it is hoped these could prove to be a useful source of conversation and future research. Visualization of the results all provides insight for archaeological survey methods and identification of areas that may have been significant in systemic contexts but have not been recognized in archaeological landscapes. Comparing representations of systemic landscapes and their archaeological signatures with modern landscape features and land uses may reveal productive avenues for preserving sites and managing and interpreting cultural resources more effectively.

CHAPTER 3.

THE ARCHAEOLOGICAL SETTING

The Mogollon Plateau region of the Sitgreaves and Coconino National Forests has attracted the attention of prominent and amateur archaeologists alike for more than a century, an important legacy of research that established chronological control and systematized cultural sequences that remain instructive (Reid and Whittlesey 1997).

Some of the great names of Southwestern archaeology conducted research in the area, and the evidence they found helped define some of the pre-contact “cultures” of the

Southwest. The Mogollon Rim is a prominent transition between physiographic provinces, as well as ecological and geological zones. It seems that the Rim may have been a boundary of sorts for cultural traditions as well, with the concept of the “frontier” commonly appearing when archaeologists discuss the prehistory of the area.

Archaeologists have seen influence from diverse areas throughout the Southwest before the Colonial era, resulting in an intriguing confluence of traditions. The area continues to be actively investigated, thanks in large part to federally-mandated management of archaeological and other cultural resources on National Forest lands. Advances in the practices of archaeological research, and the growing prominence of cultural resources management, re-energized academic interest in the region. As a result of these trends, more attention has been paid to small sites that typically remained outside the focus of most of the early researchers.

The development of this legacy of research has generated important sets of data that can be used to address a number of hypotheses that have been suggested regarding the significance of small sites. In the following sections, I review previous research in the study area, paying particular attention to instructive studies of small sites in the study area and surrounding regions, as well as other cross-cultural settings. Past excavations of small sites are also reviewed, along with an introduction to the subsets of sites included in more detailed analyses that follow. I highlight some of the expectations derived from alternative opinions regarding small sites. This sets the stage for GIS-based analyses to assess these hypotheses in the following chapters.

Early Studies in Culture History

Among the earliest to explore the American Southwest, including portions of the study area, were a number of influential individuals, such as Adolph Bandelier, Jesse

Walter Fewkes, Frederick Webb Hodge, Walter Hough, Alfred Kidder, Cosmos and

Victor Mindeleff, and Leslie Spier. They and many others played fundamental roles in the initial discovery and exploration of a number of ancient settlements. For good reason, they were typically attracted to the region’s largest, and usually more recent, villages. In general, the Zuni and Hopi regions received considerably more attention than other areas throughout the Little Colorado River drainage. “Clan-migration legends” (Kidder

1962:267) strongly influenced archaeological interpretation in these areas, including small sites.

Bandelier (1884) presumably included single-room sites among “small house ruins,” which he differentiated from “communal houses.” He also suggested they generally pre-dated their communal counterparts. Sites with small structures were typically equated with pre-Pueblo occupation, as observed in the neighboring San Juan region (Spier 1917). Although many, including Bandelier, slighted their significance, others were apparently intrigued. Cosmos Mindeleff (1896) emphasized the agricultural functions of single-room sites by naming some of them “farming outlooks.” Exploring an area along Clear Creek near the study area, he interpreted their arrangement as haphazard but in generally close proximity to the drainage and noted minimal investment in construction, leading him to interpret them as “small temporary shelters or farming outlooks, occupied only during the season when the fields about them were cultivated and during the gathering of the harvest, as is the case with analogous structures used in the farming operations among the pueblos of to-day” (Mindeleff 1896:246). He also emphasized their probable association with larger sites nearby and the diverse roles that they played in “horticultural” subsistence networks. Prudden (1903), working primarily in the San Juan region, emphasized the importance of single-room sites as eventual components of “unit pueblos” that were structurally identical to the composition of larger sites, an idea applied later to other areas.

Although large sites were often the focus, a number of sacred and ceremonial sites were documented; in many cases, some of these earliest accounts remain the best evidence of their archaeological remains, which attracted a significant amount of looting in some cases, and were seriously damaged by modern land use practices in others

(Ferguson and Hart 1985; Greenwood 1983). Many shrine sites included features similar to small structures or that could potentially be confused with collapsed structural mounds.

Although many of these spring and peak shrines have been disturbed, their importance to and in many cases continued use by modern Native groups denotes their long-standing significance (Whiteley 2011).

When the first conceptual framework for Southwest archaeology was established in 1927 during a meeting organized at Pecos by A. V. Kidder, precise chronological control was not yet available (although calls for archaeological tree-ring samples were put forward at the time by the innovator of tree-ring dating, A. E. Douglass). The relative, developmental Pecos Classification sequence was primarily based on architectural and ceramic attributes and their stratigraphic relationships, and its formation relied upon issues of typology, classification, and the identification of cultural boundaries. Its geographic significance was acknowledged as being limited to the

“Anasazi” tradition of the northern Southwest (Cordell 1997:167) and it was understood that different regions would conform to varying degrees with the stages (Roberts 1935).

More attention was paid to the study area in 1929 when Emil Haury worked with others to assist A. E. Douglass in establishing a revolutionary degree of chronological control, prospecting for dendrochronological samples at large Ancestral Pueblo sites in the pine forests of the Mogollon Rim. Haury led the team to the Bailey, Show Low, and Pinedale

Ruins (now all at least partially on private land, and are either badly disturbed or destroyed) (Haury 1962; Mills and Herr 1999:269; Reid and Whittlesey 1997:180). This

60 opened the door for several other developmental sequences proposed during the following decades.

Ultimately, most of the sequences from Anasazi areas have been derived from

Kidder’s Pecos Classification. Most reflect a linear or evolutionary sequence of development that is apparent in Kidder’s work, from seemingly simpler, to more complex technology and settlement patterns. Roberts (1935) focused upon material culture in his sequence, noting a general increase of technological sophistication and complexity, followed by a regressive stage. He explained some single-room sites as a stage of development that evolved from pit houses to aggregated pueblos, an approach later used by others (Wilcox and Pilles 1978:3). Longacre (1964) proposed a sequence that emphasized the establishment of farming and larger villages, using ceramics to date the phases. Plog (1974a) later used this revised chronological sequence to examine findings from the Hay Hollow Valley. Among all of the sequences, the most important and influential for the Mogollon Rim region of the Sitgreaves National Forest has been

Haury’s Forestdale Cultural Sequence, based upon investigations in the Forestdale Valley just south of the Rim from 1939 to 1941. He helped establish the Mogollon as a unique cultural tradition and highlighted the contrast between Mogollon and Anasazi archaeology in the Forestdale Valley, an area he called “the front doorstep of the vast

Anasazi domain. The environment alone could not be blamed for the differences one saw. It was clear that some other force was at work, and that had to be an inheritance of cultural values that coursed along a different track than that of the Anasazi” (Haury

1985:xvii). In addition to his earlier work with Douglass prospecting for beams at large

61 sites like Pinedale, Bailey, and Show Low Ruins, Haury (1985) drew upon observations from the surrounding Silver Creek and Grasshopper areas, which he believed shared a similar chronological sequence (Mills and Herr 1999:269).

Woodbury (1961) classified agricultural features in the vicinity of Point of Pines, located southeast of the study area in the more broken and mountainous terrain of the transition zone. He formally defined field houses as “one-room structures associated with prehistoric fields, serving for summer shelter of farmers and temporary storage of the harvest” (Woodbury 1961:xiii). He also argued that social groups probably “farmed a variety of locations simultaneously as insurance against crop losses. In addition to fields marked by terraces or stone borders [the remaining components of his classification], much larger areas of bottomlands and prairie were undoubtedly farmed.” He also found no evidence of irrigation among what he characterized as small and simple systems that effectively managed water distribution and soil erosion for a significant period of time.

Woodbury (1961:14) suggested that field houses were particularly important because they could yield evidence regarding the activities associated with farming, such as keeping careful watch on crops during summer and the approach of the harvest (Cushing

1974[1920]), as well as for temporary storage purposes. He was also confident that field houses could yield chronological information, which was generally lacking among the extensive terraces and borders in the area. He also noted extramural semi-circular walls, interpreted as wind breaks and speculated that the forest had descended significantly and that many of the sites that were now surrounded by trees were once located among more open vegetation, be it woodland or grassland (Woodbury 1961:2).

John P. Wilson (1969), from Harvard University’s Peabody Museum, approached the study area from a different direction, exploring the limits of the Northern Sinagua archaeological culture. He explored the Anderson Mesa area, including Chavez Pass

(Nuvakwewtaqa), and continued east toward Clear Creek and Chevelon Canyons, noting increasingly-Mogollon characteristics in the vicinity of Heber and McDonald Canyon

(Solometo 2004:159; Wilson 1969). Technological differences in plain ware and black- on-white production signaled a fairly clear break in “archaeological cultures” on either side of Clear Creek, and fortified sites on rocky, limestone soils suggested population expansion and a concern for defense.

Following Haury’s pioneering work in the Forestdale Valley, little other excavation occurred in the study area until the advent of the Arizona State Museum’s

Archaeological Salvage Program, conducted in cooperation with the Arizona State

Highway Department. Vivian (1969) reported on the results of several excavations in the vicinity of the communities of Pinedale and Clay Springs, Grebinger and Bradley (1969) reported on excavations at two sites near the Forestdale Valley, noting evidence for multiple occupations. These excavations were important for demonstrating subsurface deposits at presumed small structures did not necessarily meet expectations from surface indications and suggest that even detailed surface documentation of small sites are problematic in regards to occupational sequence and site function.

Processual and Behavioral Approaches and the Rise of Cultural Resources Management

The rise of the “American New Archaeology” during the 1960s began to shift focus from culture history towards issues associated with cultural processes (Trigger

1989:294-295). Important early advances in this approach were completed in the Hay

Hollow Valley area, to the east and north of the study area, by Paul Martin’s field school, which arose from collaboration between the Field Museum and University of Chicago students in the early 1960s and completed a significant amount of excavation as well as survey through 1971 (Mills 2005). These projects inspired a number of important studies that firmly shifted focus upon social processes (e.g., Hill 1970; Longacre 1970; Plog

1974a; Plog and Hill 1971). Plog’s subsequent work with the Chevelon Archaeological

Research Project (CARP) continued to apply this perspective to other areas to the west

(Plog et al. 1976), and research at Grasshopper Pueblo was also influenced (Reid and

Whittlesey 1997, 1999). These new perspectives also emphasized the importance of sampling by exploring some of the underlying assumptions that had driven past interpretations (Trigger 1989:310). Behavioral archaeology rose as an important critique of processual approaches that failed to account for formation processes.

The growth of contract archaeology projects during and after the 1960s extended the influence of Haury’s chronology northward, throughout a larger portion of the

Sitgreaves and surrounding areas (Ciolek-Torrello 1981; Lightfoot 1978; Reid 1982;

Vivian 1969). This development also brought renewed attention to small sites, especially due to the fact that intensive surveys became more common (Gregory 1975; Rohn 1963;

Woodbury 1961). This led to increased attention to the geography of agricultural systems and their relationships with other settlement types. Still, Pilles and Wilcox (1978:3)

64 lamented that small sites “have usually been viewed as functionally specific structures related to storage and farming practices. This concept has not changed significantly since the first mention of small sites in the 1880’s.” Research in the Pinedale area demonstrated that sites located to the north and south of the Mogollon Rim region tend to have nearly identical ceramic assemblages, although proportions of types and wares vary

(Ciolek-Torrello 1981).

Donaldson (1975), working under the supervision of Fred Plog, championed the use of a random sample procedure using traverses oriented along cardinal directions, whose starting longitude or latitude was chosen randomly using a computerized system.

He and Plog undertook additional large-scale, systematic random sample surveys of planning units for the USDA, which led to the publication of a formal Cultural Resources

Overview (Plog 1981a) for much of the ASNF and adjacent BLM land to the north.

These efforts formed a foundation for future survey methods for large timber sale projects. Also, Donaldson’s subsequent role as the ASNF Forest Archaeologist during the following years encouraged implementation of sampling methods, as well as ongoing studies from the Chevelon Archaeological Research Project (CARP), which documented hundreds of sites in the area. Plog (1978) examined relationships among various site types and environmental variables prior to widespread availability of such information and identified some trends, but recognized that more research was needed.

Gregory (1975) completed a two stage nearest-neighbor analysis of sites from the

Hay Hollow Valley dating between A.D. 100-1100. The first stage of his analysis examined all sites and uncovered only random associations. However, the second

65 rendition examined all sites except for single room structures, showing that the sites were clustered. This provides an important perspective for the present study. Hantman (1983), who worked under F. Plog, provided important contributions to understanding the differences in archaeological cultures across the southern portion of the Colorado Plateau.

Reid (1982b) applied a site function approach that emphasized diversity measures of assemblage contents to assess settlement variability. The underlying assumption of this approach being that the longer a particular location is occupied, the total cycle of activities that are conducted at the location will increase. In turn, artifact diversity is assumed to directly correlate with the diversity of activities. Modal results of diversity measures were used to classify sites within the Cholla Project’s study area (Figure 3.1), which includes portions of the Chevelon region, and to assess settlement system variability. This approach implicitly invokes several aspects of formation processes, such as the fact that the formation histories of artifact assemblages at a given site are influenced by highly variable use lives and associated discard rates (Schiffer 1987;

Schlanger 1990; Sullivan 1980). An important factor in successfully and accurately applying this approach is reliance upon surface assemblages, particularly those of uneven accuracy and reliability, as well as the difficulty of comparing such sites with assemblages from excavated contexts.

Figure 3.1. The transmission line surveyed by the Cholla Project (background) photographed from a site recorded by CARP, with a small valley suggested to be a field location in between, crossed by a Forest Service road.

Dosh (1989) reported on data recovery efforts at the Ada Mesa Site, located along

Upper Coyote Creek, a perennial tributary of the Little Colorado River just past the eastern margins of the study area. AR-03-05-01-62 included a three-room masonry structure located along the margins of an earlier pithouse occupation. The research design sought evidence of past subsistence strategies and chronometric samples, along with pollen and flotation samples. Unfortunately, all of the tree-ring samples failed to yield dates, but the radiocarbon and archaeomagnetic dates situated the site’s components comfortably within the regional chronology: the pithouse component spanned the

Georgetown and San Francisco phases (ca. A.D. 600 – 900) and the pueblo component dated between A.D. 1100 – 1200 (Dosh 1989:146-150, 157). The pollen and plant macrofossil remains provided the most interesting results. Maize was abundant and ubiquitous among samples from the earlier pithouse component of the site is not particularly surprising; the complete absence of maize from the site’s later component was unexpected. Situated adjacent to what appears to be prime agricultural land, and apparently confirmed by the earlier component, the preponderance of wild plants, particularly cheno-ams, along with small, circular manos typically assumed to be associated with processing of wild plant foods, suggests the site’s primary function was gathering rather than agriculture (Dosh 1989:152-153). A larger but contemporaneous

40-room pueblo is located about a half mile from the site (Laumbach 1980). A three- room structure could have been needed to oversee active fields, but some evidence of maize would be expected even if the entirety of the resulting crop was harvested and immediately transported to the larger site. This contradicts the notion that maize- intensive agriculture strained available agriculturally-viable sites.

Two university-based research programs have made major contributions to studies of the archaeology of the study area: the Chevelon Archaeological Research

Project (CARP) and the Silver Creek Archaeological Research Project (SCARP). Both trace their origins to research dating backing many years. CARP was originally directed by James Hill and Fred Plog during the early 1970s. SCARP was a more recent manifestation of the University of Arizona Archaeological Field School and, investigated the Mogollon Rim region of the ASNF beginning in the early 1990s, including a later

68 return to the Forestdale Valley on the Fort Apache Indian Reservation. Innovative and integrative studies of the regional prehistory have fine-tuned Haury’s chronology, established chronological control in surrounding regions (Table 3.1), and shed considerable light on the late Ancestral Pueblo archaeology (Mills et al. 1999).

As is common throughout much of the Southwest, the Forestdale sequence demonstrates a shift from semi-subterranean pithouse structures (Haury and Sayles 1985) to above ground masonry architecture, as well as a movement from isolated, hilltop settings to locations with presumed greater agricultural suitability between ca. A.D. 200-900 (Roos

2008b). Test excavations at sites representative of several phases were completed by

SCARP, expanding our understanding of community organization and other issues, especially during the Pueblo periods (ca. A.D. 1100-1400). Excavations at Bailey Ruin

(Duwe 2005; Mills et al. 1999), Pottery Hill (Mills et al. 1999), Bryant Ranch (Cano

2003), Roundy Pueblo (excavated in conjunction with the Forest Service’s Passports in

Time program), and four great kiva sites (Herr 1999, 2001; Herr et al. 1999; Mills and

Duwe 2003; SCARP 2012) were the primary focus of excavations, but work was also completed at an earlier pithouse site called Hall Point (Roos 2005, 2008a), as well as a small architectural site with an associated artifact scatter and agave, called the Agave Site

(Mills and Margolis 2002). Particular attention was also given to community aggregation at large sites during the late Pueblo period and to changes in craft production and specialization (Duwe 2005; Kaldahl et al. 2004; Mills 1998; Scholnick 2003).

Table 3.1. Chronologies for portions of the study area and surrounding regions.

Silver Creek Eastern Sinagua (Wilson 1969) Flagstaff Chevelon Pecos (Mills and Herr 1999) (Kamp and Whittaker 1999) (Tuggle 1982) (Woodbury 1979) Corduroy Group 2 G Pueblo I 800-900 800-900 500-850 700-900 Dry Valley Group 3 Rio de Flag E 900-1080 900-1050 900-1064 850-1050 Pueblo II Early Carrizo Group 4 Angell-Winona D 900-1100 1080-1150 1050-1100 1064-1100 1050-1125 Padre Late Carrizo Group 5 1100-1150 B 1150-1200 1100-1200 Elden 1125-1200 1150-1250 Pueblo III Linden 1100-1300 1200-1275 Group 6 Turkey Hill A Pinedale 1200-1300 1250-1300 1200-1275 1275-1325 Canyon Creek Group 7 Pueblo IV 1325-1385 1300-1400 1300-1600

SCARP also surveyed 2.9 square miles in the Forestdale Valley, providing an opportunity to integrate Haury’s original findings with new survey data and the site database maintained by the White Mountain Apache Tribe (Mills et al. 2008:38; Seidel 2004).

Twenty-eight sites located throughout the Silver Creek, Grasshopper, Forestdale,

Hay Hollow, and Chevelon regions have yielded 1,233 tree-ring dates, forming the basis of our understanding of the chronology of the Mogollon Rim region. Although the

Forestdale sequence inherently suggests a certain sense of continuity through time, several gaps and spikes in the tree-ring record suggest that the region was not occupied continuously from A.D. 200 to 1400 (Mills and Herr 1999:278), and probably with even less frequency during the previous ten millennia. However, tree-ring dates suggest that the region was occupied discontinuously throughout the sequence and that a zone of hybridity developed through prolonged patterns of population movement (Haury 1985;

Mills and Herr 1999). The dramatic increase in site frequency beginning about a millennium ago suggests a level of growth that could only have occurred through immigration (Herr 2001; Newcomb 1999).

Carrizo phase (ca. A.D. 1080-1200) settlements, the most abundant of any phase in the Sitgreaves, usually took the form of small pueblos occupied by one or two households (Mills and Herr 1999). A few slightly larger sites exist, but only a small portion of them had more than 10 rooms (Newcomb 1999). People were dispersed across the landscape in small farmsteads and great kivas served to strengthen ties among groups in the area, where land was abundant, but people and their labor were scarce (Herr 2001,

2002). The types of sites present at this time are representative of others throughout the

71 greater region that were constructed during the 11th and 12th centuries (Mills and Herr

1999), such as the Petrified Forest area (Burton 1993). Although later Linden phase sites are common in the Mogollon Rim region, it seems that population levels declined during the 13th century in the Silver Creek area (Mills 1999:508), and probably in surrounding areas as well. Also, settlement patterns shifted from numerous small farmsteads to fewer aggregated villages, followed by widespread movement out of the area by Pueblo peoples at the end of the 14th century. Roos (2008b) simplified the Forestdale chronology to reflect these trends, identifying the early Pueblo Period from A.D. 1100-1200 and maintaining traditional Pecos Classification periods for the later periods (although with dates modified to include SCARP findings) (Mills et al. 1999).

Two other SCARP investigations are particularly relevant to the study presented here. Kunen (n.d.) reviewed associations of agave parryi and various archaeological sites in the Silver Creek and Chevelon drainages, building upon a relationship previously recognized by Minnis and Plog (1976). She found most agave is associated with limited activity sites or those with few structures and little if any sign of permanent habitation.

Also, agave stands were primarily associated with the Carrizo and Linden phases rather than later periods associated with population aggregation. They were not found in direct association with middens, but rather in rocky areas only marginally-suitable for agriculture. Kunen (n.d.) also recognized that roasting pits had not been sufficiently studied in the area, and associated rock pile and terrace features may be under- recognized.

The Agave Site (AZ P:11:344 ASM), a small-architectural locus, was investigated briefly during the summer of 2002 (Mills and Margolis 2002), but the field season was cut short by the Rodeo-Chediski wildfire, the largest in state history at the time. Located on a low hillside about a mile from a 20-room site believed to be contemporaneous (ca.

A.D. 1000-1150), surface indications included a fairly extensive midden and several shaped masonry blocks on an eroding surface suggesting the presence of a one or two- room structure. An intensive surface collection was undertaken over much of the site and test excavations were completed. Excavations confirmed the presence of an intact architectural unit, revealing several fairly large, faced stones extending into the hillside.

Subsequent closure of the National Forest due to wildfire prevented resumption of the excavations, but it is suspected that at least one or more rooms are present. Also, examination of the surface collections (Alcorta 2002; Mills and Margolis 2002; Richard

2002) revealed evidence for multiple occupational components, suggesting periodic re- use of the site and implying support for previous evidence of a meaningful association between archaeological sites and agave in the study area. There have also been indications of possible anthropogenic soils associated with agave sites in the area, but this has not been examined in detail.

Stephen Plog re-initiated CARP in 1997 after a hiatus of many years, returning to the Chevelon and Clear Creek areas to study expressions of social and community organization. Some of the most significant outcomes of this work were completed by

Solometo (2004, 2006), who examined the case for social conflict as an important factor in culture change in a portion of the study area generally lying between Chavez Pass and

Heber, encompassing the middle reaches of Chevelon, Clear Creek, and Jacks Canyons.

Building upon the work of F. Plog (1978) and Gregory (1990), they characterize a shift from canyon bottoms and rims during the Archaic, to increasing use of smaller seasonal drainages within piñon and juniper woodland. These settlements included more terrace and check dam features with presumed agricultural functions, and a preference toward sandstone deposits and their derived soils. From the 10th through 13th centuries A.D., the most common sites are characterized as loosely clustered groupings of two to three rooms

(16 to 35 square meters in size), although single rooms with well-developed middens are common (Solometo 2006:40). Aggregation was more common after the middle of the

11th century, typically with associated fortifications or largely inaccessible landforms; little use of the area is indicated after the 14th century.

Solometo (2006:41) characterizes rectangular pueblos with enclosed plazas as

“The earliest archaeologically visible defensive strategies in the study region.” Plog et al.

(2001) investigated some of these sites and suggests violent social conflict during the mid-12th century resulted in abandonment of most small sites and the persistence of enclosed pueblos. Some later sites were found to be located upon topographic prominences away from presumed agricultural areas and incorporated presumably defensive architectural elements, such as loopholes (Solometo 2006; White 1976).

Solometo’s (2004, 2006:45) proposed typology for these sites includes the lookout, isolated refuge, proximate refuge (to support larger groups from nearby settlements), and year-round defensive habitations.

Building upon CARP data and other surveys, Peeples et al. (2006:13) suggest:

“the archaeological record of this land use strategy in the Chevelon Crossing

region appears in the form of regularly spaced settlements, created as farmers

repeatedly moved to clear and cultivate a fresh patch of land and avoided

previously farmed patches that had entered an α-phase ecological reorganization

and long period of woodland regrowth that left them agriculturally unproductive

for generations.” [Peeples et al. 2006:13]

In a related study, Fertelmes and Barton (2007:12) suggest that upland farming areas were initially attractive for maize agriculture and offered fertile soils that could be accessed by clearing the land with a minimal amount of labor, transforming piñon-juniper woodlands that dominated these areas with a form of shifting, swidden agriculture that used fire to clear the land (citing Matson et al. 1988 and Kohler and Matthews 1988).

Fertelmes and Barton (2007) suggest that satellite imagery reveals modern vegetation patterns in the middle Chevelon Creek Drainage that have been influenced by prehistoric agricultural practices. Thirty-five sites with “visible above ground room blocks or agricultural features such as check dams” were selected for analysis. They note a significant labor investment was needed to clear land for agriculture and construct habitation and storage structures but suggest the vegetation has remained impoverished despite not being farmed for several centuries. “In this sense, agriculture in this region was initially productive but ultimately unsustainable. Short-term land use practices destroyed the long-term productivity of this region for maize horticulture” (Fertelmes and

Barton 2007:14). Peeples et al. (2006) also assert that later sites are situated in locations

“favoring more defensible agricultural fields,” and “by ca. 1300, this combination of social and natural processes had set off a cascade of socio-ecological changes that may have triggered an ecological reorganization of the piñon-juniper woodland and certainly resulted in the human abandonment of the Chevelon Crossing region for centuries”

(Peeples et al. 2006:14).

Ongoing Management, Rodeo-Chediski, Healthy Forests, and Managing Future Forest

Uses

Following the Rodeo-Chediski Wildfire, extensive surveys were undertaken in advance of proposed timber salvage projects. North et al. (2003:vi) characterized prehispanic settlement in portions of the Rodeo-Chediski Wildfire area as “scattered farmsteads that were widely distributed along the many north-south oriented drainages that traverse the region.” Post-fire surveys offered an opportunity to examine areas with greatly improved surface visibility, which had hampered interpretations in the past. Low site density areas in Ponderosa pine forests with heavy accumulations of pine duff and other debris begged the question of the influences of surface visibility. New sites were identified in many areas, but other suspected features at some sites appeared to be more likely to be natural outcrops in some cases. Removal of vegetation provided a unique opportunity to re-assess the site inventory and surface assemblages and architecture at sites that previously suffered from poor surface visibility. Subsequent erosion in severely

76 burned areas also limited the duration of this opportunity, as many sites were threatened by erosion.

North (2004) highlighted the fact that sites are most frequently encountered in upland settings. This is a prominent feature of the archaeological landscape, and plays prominently in settlement system hypotheses. But it is important to recall that alluvial deposits are generally much deeper within drainages than topographic prominences.

Extensive additional surveys were also completed by the Forest Service. Low site density areas in the southwestern portion of the ASNF study area were generally confirmed, as well as higher site densities in the southern portion of the ASNF east of

Heber toward Show Low. Upon those higher, typically more deflated surfaces, surface finds are more likely. Redeposition of these materials along the slope or near the base of it is also common. Drainage bottoms, on the other hand, are high-energy environments, where depositional patterns fluctuate dramatically with seasonal precipitation patterns.

Interestingly enough, many of the most common classes of sites within these settings date to the historic period.

The USDA Forest Service, especially its Southwest Region, has undergone a dramatic shift since Cultural Resources Management became firmly established in the

1970s. Great strides in the programs came about as the result of subsequent lawsuits that recognized the inability of the agency to adequately comply with federal preservation laws. Personnel increases and formalized strategies were adopted, in large part to facilitate cultural resources compliance for timber harvesting. Projects have shifted from traditional timber sales to “ecosystem restoration projects,” often covering large

77 geographic scales. Catastrophic wildfires, in part caused by historic fire suppression and wood use practices have encouraged thinning of dense Ponderosa pine forest. The

Healthy Forests Restoration Act of 2003 (or Healthy Forest Initiative) instituted large- scale thinning projects on National Forest lands. Although lambasted by some as the “No

Trees Left Behind” Act, the resulting projects were intended to reduce fuel loads in stands of timber that were deemed overgrown and unhealthy due to past timber harvesting and fire suppression activities.

In addition to the burden of completing large-scale surveys and mitigating potential effects to archaeological sites, archaeologists can play other roles in these projects. The lengthy time scales of archaeological analysis may be appropriate for helping to assess the nature of the landscapes that are to be restored. On a more practical note, a better understanding of past land use behaviors may help streamline survey strategies and consultation with the public, other state and federal agencies, and Native tribes.

Ethnographic Insights and Historical Land Use of the Mogollon Plateau

Ethnohistoric accounts of agricultural pursuits in the Mogollon Plateau region offer some important insights to past land use practices and landscape perception.

Evidence of precontact irrigation agriculture in the Mogollon Rim region has been limited to the vicinity of a few perennial streams and speculation regarding the areas surrounding some larger sites (Lightfoot and Plog 1984; Plog 1981; Tuggle et al. 1984),

78 although additional evidence has been found farther to the north from the Rim in the vicinity of Snowflake and the Hay Hollow Valley. This supports the common interpretation that the subsistence economy was primarily based upon dry farming methods augmented by hunting and gathering, or perhaps vice versa. Dry farming practices are distinguished from rain-fed practices by the fact that moisture does not necessarily arrive at the required times as precipitation, but is retained and delivered to targeted crops as a result of soil, topographic, and other associated conditions. Limited precipitation and short growing seasons have made dry farming a particularly risky venture. Although evidence of this lifeway changed with the centuries, agricultural pursuits, including rain-fed agriculture, continued to be practiced in the upland, mountainous areas immediately north of the Mogollon Rim by Mormon settlers during the 19th century, providing important ethnographic insights (Abruzzi 1993, 1999).

Permanent streams in the lowlands farther to the north allowed irrigation agriculture, but

Mormon settlers knew that rainfall was too sparse in those areas to support dry farming techniques. Rain-fed practices were generally less labor intensive than irrigated agriculture, and Lightfoot and Plog (1984) witnessed that these activities were still practiced to some extent in recent times (1984:186). Fields were selected for the presence of deep soils and in favor of locations that capitalized on available runoff from adjacent slopes. Fields were generally situated in valleys between elevations of 6000 and

8000 feet. Planting during the spring resulted in harvests during late summer, provided that rainfall was sufficient and frost damage was avoided. Crop yields were greatly affected by these conditions, as documented by repeated productivity declines during

79 several periods of the 19th century (Abruzzi 1993; Lightfoot and Plog1984). Diversion agriculture was probably used in some locations, although it is more common in other surrounding regions (Damp et al. 2002; Mabry 1996).

“If there is a distinction between dry farming and rainfed agriculture, it is that the former is the latter, but the latter is not necessarily the former” (Doolittle 2000:121), and

“where terrain is not a problem, rainfed agriculture may result in no appreciable traces being left on the present-day landscape” (122). In fact, some estimates suggest as little as one in ten of the world’s rainfed fields result in features that would be identifiable after abandonment. This is probably at least partially attributable to a need to abandon fields after a few seasons to allow the soil fertility to recover during the prehistoric era

(Doolittle 2000:122). In areas where lack of precipitation is a primary concern, methods to divert flow and conserve soil moisture may result in enduring evidence. In areas where length of growing season is a primary concern but there is abundant precipitation, there are few strategies farmers can use to stave off killing freezes and those available would not likely result in detectable archaeological evidence.

This type of correlative identification of presumed rainfed fields in the Southwest typically involve relatively small areas associated with the small structures known as

“field houses” (Doolittle 2000:164). Because most of the areas in the Southwest that receive a suitable precipitation for rain-fed farming are in mountainous regions, areas that are suitable for such cultivation are usually limited in size. Field houses often occur as

“remote and isolated” sites and the agricultural association is typically assumed. For example, Henderson’s report on survey in the vicinity of Chavez Pass noted a

80 proliferation of single rooms located atop the highest points on mesas in areas with direct exposures, and went on to assume that their location is “undoubtedly a function of their association with mesa top field areas” (1979:26). Artifact assemblages indicative of short occupations suggested that few were contemporaneous and that fields had been rotated to circumvent potential problems associated with soil fertility.

Soil moisture is a more critical issue for agriculture than rainfall itself, since plants absorb water through their root systems rather than their leaves. Timing of soil moisture is critical, and studies have shown that significant yields of maize may still be produced so long as sufficient soil moisture is present when plants begin to sprout and pollinate (Doolittle 2000:219). This situation can be achieved by careful selection of field locations, including areas with shallow water tables, such as near seeps, and soils that are more prone to retain water at appropriate depths, especially sandy sediments

(Doolittle 2000:220; Perramond 1994). The importance of retaining snow melt has been documented in many instances in the vicinity of the study area, including Hopi and Zuni

(Beaglehole 1937; Cushing 1974[1920], 1979; Hack 1942), and these accounts provide important ethnographic evidence of rain-fed farming practices, such as the use of sagebrush to trap wind-blown sediment, movement of field boundaries for similar practices, slight modification of planting to take advantage of the slight mounds of previous years’ crops, and the targeting of particular settings associated with stratigraphy and topography.

Espejo’s account from one pueblo along the Rio Grande includes the observation that “in each planted field the worker has a shelter, supported by four pillars, where food

81 is carried to him at noon and he spends the siesta; for usually the workers stay in their fields from morning until night” (1966:220, in Doolittle 2000:145), and Obregón’s

(1584[1928]:325) account of Espejo’s journey notes the presence of “shacks” in each field. Elevated structures in fields were documented among several groups, most commonly citing the need to prevent birds from feeding on crops.

Few studies of Western Apache rainfed fields have been completed, but it seems likely that early groups who occupied areas where higher elevation ensured increased precipitation, small fields were probably cleared or sought out in forested areas, planting multiple crops and tending them with simple tools, such as stone hoes (Pool 1994:93).

This practice likely encouraged the creation of small mounds associated with desirable plants; an archaeological signature of this practice is unlikely (other than, perhaps, rare but recurring implements), and it is suspected that such fields would have comprised relatively small patches whose locations would have changed quite regularly. Doolittle laments that “identifying plain fields archaeologically is, to be sure, tantamount to looking for the proverbial needle in a haystack. Out of necessity, or perhaps frustration, most prehistorians have contented themselves with inferring that agricultural people occupying sites near fertile, well-watered, frost-free lands must have farmed those lands”

(2000:162). He presents a number of cases where locations of rainfed agriculture have been inferred from observed relationships of archaeological sites with arable land throughout North America (2000:163), including Chavez Pass (Henderson 1979), the

Upper Little Colorado area including Chevelon Canyon (Plog 1970; Plog 1978, 1981b), and southwestern Colorado (Van West 1994).

“One of the greatest problems for understanding dry farming prehistorically is the paucity (in fact, in most cases, the lack) of evidence which survives to present day”

(Doolittle 2000:235). A commonly employed strategy is soil modification to encourage the creation of similar micro-environments, such as mulching to retain soil moisture, encourage moisture infiltration, regulate soil temperature and texture, and prevent soil loss. These practices may result in more visible archaeological signatures (e.g., Ellis

1978), or may even be related to natural processes, such as the eruption of Sunset Crater

(Pilles 1978). Clearly, such practices must have varied among and between different groups and in response to varying environmental conditions, as well as perceived need, including the construction of features ranging from the simple to the complex to impede overland flow of water and take advantage of nutrients and detritus transported by water

(Treacy and Denevan 1994:93). Various forms of terraces (assigned a number of often interchangeable names by archaeologists, some of which imply function, while others, such as “rock alignment,” are more general) can function to increase soil depth, create level surfaces, control erosion, regulate water and temperature, and improve soil fertility

(Doolittle 2000:257-264; Rankin 1989). Water harvesting often involves the construction of an earthen diversion dams across a shallow arroyo that bisects a fairly level field.

Complex water harvesting systems are known from surrounding regions, including

Chaco, Mesa Verde, the Verde Valley (Fish and Fish 1984), and below the Mogollon

Rim on the Tonto National Forest (Wood and McAllister 1984). Also, slightly farther to the north of the Mogollon plateau area, larger drainages that supported late Pueblos like

Shumway, Fourmile, and others were probably used to support agricultural pursuits.

In the Mogollon Rim region the most frequently identified agricultural features besides field houses have been check dams and rock alignments. Lightfoot (1981; and

Lightfoot and Plog 1984) reported several terrace systems in the vicinity of the study area, near Snowflake and within the Hay Hollow Valley. One cannot discount the notion that many similar features were constructed of perishable materials, such as fallen trees, or earth, now unrecognizable in the archaeological record (e.g., Cushing 1974[1920]:152-

153; Ferguson 1985:106-107, 110; Ford 1985:36; Winter 1976:85). Juniper is a suitable material for such simple constructions.

Also, it is noteworthy to consider the fact that archaeological features with implied agricultural functions, including field houses, may have served other purposes.

Sullivan (1987) provides an important example of a fortuitous find at a small site that spawned critical examination of the roles that small sites played in presumed settlement- subsistence systems. AZ I:1:17 (ASM) consisted of a single room masonry structure with additional associated features that burned catastrophically, presumably around A.D.

1064. Systematic pollen and flotation sampling revealed a paleobotanical assemblage that ran counter to many earlier assumptions regarding reliance upon agricultural products. Foodstuffs commonly thought of as supplemental, or worse, as starvation foods, were far more prominent in this specific assemblage. Still, it could represent an episode corresponding with poor agricultural production and remains a somewhat unique assemblage in terms of composition, analysis, and fortuity.

Although Sullivan’s (1995) research in the vicinity of the South Rim of the Grand

Canyon is somewhat distant from the current study area, it holds interesting insights for

84 the problems at hand. His presentation of feature and assemblage data for the one-room structures investigated by Statistical Research (Whittlesey 1992) and the Upper Basin

Archaeological Research project provides an important discussion of the archaeological variability of a class of sites traditionally interpreted as “agricultural field houses”

(Sullivan 1986). He suggests that a null model of these sites draws into question the idea that the area’s occupants were primarily agriculturalists. Also, his examination of

“agricultural terraces” from the same area (Sullivan 2000), in concert with a similar study previously completed by Homburg (1992), questions the presumed significance of domesticates to their subsistence system, as well as assumptions regarding depletion of soil mineral content due to cultivation.

Fish and Fish (1984) investigated an intensive field system in the vicinity of

Sacred Mountain, within the Verde Valley on the Coconino National Forest. Although the locality along Beaver Creek does not appear to be an appropriate analog for this dissertation’s study area given the extensive canals and evidence of labor-intensive production, it does provide some interesting information. They noted 24 small structures situated almost exclusively along canals within the field system, and excavated one

“horseshoe structure.” Stacked cobbles marked the remains of walls that presumably supported a brush superstructure, enclosing approximately 8.75 square m (Fish and Fish

1984:155). The recovered artifact assemblage was diverse, but no features were observed, leading to an assumption of seasonal occupation and a “fieldhouse” function.

It is also interesting to note that the soils associated with this field system were particularly rocky, providing abundant raw material for field construction. Ongoing use

85 appears to have required a significant amount of labor for maintenance and reconstruction following micro-environmental changes. This supports ethnographic evidence that similar agricultural features required a great deal more labor beyond the initial investment in construction (Doolittle 2000).

Research at Chavez Pass (Upham 1984, Upham and Bockley 1989) proposed the argument that small sites could represent a record of previously unrecognized mobile groups. Although thought-provoking, research in other areas of the Southwest suggests this is probably a spurious conclusion and the small sites are most likely associated with larger villages in the region. This also raises the notion that the inclusion of small sites in population estimates may result in exaggerated figures (Preucel 1990:181) if they are temporarily occupied by inhabitants of larger nearby sites.

As Preucel (1990:184) points out, “The assumption that all one to three room structures functioned as field houses must be critically evaluated.” Small structures could have served a wide array of purposes, including temporary shelter for wayfarers or traders, shrines or pilgrimage sites, shelters and storage facilities for piñon-gatherers or harvesting other wild plants for food, fuel, or craft production, camps for forays to more distant areas to collect stone or other minerals or clays, hunting camps, or even, in historical times, temporary shelters for shepherds. Unfortunately, the simple fact of the matter is that no one can definitively ascertain the specific function or functions of small architectural sites without excavation, including analysis of extramural areas. Still, even detailed excavations may not reveal the full range of behaviors associated with a structure’s use.

Preucel (1990:42) defined field houses as “isolated dwellings and associated garden plots owned by a single family unit.” Perhaps their most diagnostic feature was their small size and the absence of ceremonial architecture,” and they were typically occupied by no more than a single extended family. Ellis (1978) suggested “vacation homes” as an alternative perspective. During historical times, typical Tewa field houses consisted of one or more single-room structures with stone foundation and adobe walls, averaging about 4 m by 6 m and usually featuring a single doorway and window (an architectural features that has not been documented prehistorically), central hearth for cooking, a roof area typically used for sleeping, and minimal other furnishings. Modern field houses were typically used for storage of crops and garden tools. Preucel (1990:43) points out that “Field houses were not always required in the cultivation of fields” and other structures could be used instead, such as dugouts, lean-tos, windbreaks, and ramadas, or people could simply establish open camps. Although from a different region, this demonstrates that small sites, even of similar function, could include a wide range of architectural variability that would depend upon local customs, individual choices and preferences, as well as availability of raw materials suitable for construction.

Preucel (1990) also noted that the spatial structure of small sites was determined by field location, which was a function of land tenure; Western Pueblo ownership of house and garden plots was governed by clan or lineage membership, while Eastern

Pueblos were typically determined by a farmer’s bilateral descent group. Individual farmers typically worked fields at different locations in order to diversify their holdings, a common risk reduction mechanism with different environmental locations and a common

87 characteristic among Pueblo societies. Their use could vary among late-seasonal, bi- seasonal (spring and fall), and continuous-seasonal strategies (at least during peaceful times) (Preucel 1990:50). He termed this temporal variability as “seasonal circulation” and noted it was “one of the most characteristic features of Pueblo social and economic life during the historic period” (51). He offered a dual residence model (53-55) of the mobility afforded by a network of permanently occupied sites (including villages and smaller hamlets) and seasonal nodes consisting of farming communities and field houses.

Hamlets were distinguished by their lack of large-scale ritual architecture and institutions and their dependent social and ceremonial relationship with a parent village and they originate from farming communities becoming established on a more permanent basis.

Preucel (1990:44) distinguished farming communities from field houses, citing Moenkopi as a Hopi example. “Farming communities were formed of individual field houses clustered together for social, economic, or security reasons” (Preucel 1990:51), a pattern that is also true for seasonally occupied farming villages at Zuni (Ferguson 1996). These villages were occupied on a seasonal basis but included multiple extended families (56).

His research suggested larger pueblos tend to have more farming communities and they are primarily associated fields located five miles or more from a main village, whereas field houses were most common at fields less than five but more than one mile from the main village (Preucel 1990:48-49).

Comparing and Deconstructing Site Typologies

Documentation of archaeological sites in the study area has been influenced by two primary factors: the nature of the archaeological survey projects (e.g., sample surveys for large timber sales, systematic surface surveys, reconnaissance surveys, etc.) and the site typologies used by these researchers. This leads to issues concerning identification of specific classes of archaeological sites in the field as well as the resulting archives. Some of the typologies used by researchers to discuss small sites in the study area and surrounding regions are shown in Table 3.2.

Comparison of these different site types identifies several interesting trends. First and foremost, different researchers have applied many different typologies. Variations in the systemic and archaeological implications of the different site types are apparent.

Whereas some are more likely to identify sites including only a single room as just that, others would apply a functional significance by including them within an overly broad category. These issues are not unique to the study area. Crown (1983) and others have applied a similar approach to small Hohokam sites, distinguishing field houses from farmsteads on the basis of occupation duration and the size of inferred associated social unit (Rice 2001). Field houses are conceptualized as small structures with less than 10 square m of floor area that lack floor features or formal hearths. Occupied temporarily by no more than a single family for a portion of the growing and/or harvest season, they are usually located within fields. Farmsteads are defined as “larger structures located near

89 agricultural fields and occupied by a full household unit for the duration of the growing season” (Rice 2001:1) They are more likely to include cemeteries and multiple structures arranged in a more formal layout, typically associated with deeper midden deposits than field houses, and were presumably the location of more intensive processing of a greater diversity of crops than occurred at field houses. Given the limited occupation duration of both site types, it seems difficult to reliably differentiate these on the basis of artifact assemblages (especially from surface contexts alone). It is also worth emphasizing the fact that field house sites may include the remains of multiple structures, but they are less likely to mimic the formal layout of farmsteads due to more frequent reconstruction.

This review of past research demonstrates that a significant amount of attention has been paid to small architectural sites. However, these diverse perspectives have probably resulted in quite a bit of confusion. New approaches are needed to address the significance of small sites, rather than simply relying on field crews to accurately assign a presumed function based on preliminary reviews of surface assemblages, features, and their own preconceptions. Given the large number of small architectural sites in the study area, more focused attention upon their relationships with ecological conditions and other aspects of the archaeological landscape are needed.

Table 3.2. Various typologies proposed for small sites in the study area and surrounding regions. Plog (1981b) Pilles (2000) Ellis (1978) Artifact scatters Lithic/Sherd/Artifact Scatter Field houses Water control devices Field House (seasonal, low walls, Pinon-pickers shelters and (terraces and grids) few artifacts) storage structures True Pueblo architecture 1-2 Room Habitation (full walls, Herding and (usually 2-4 rooms) diverse artifacts) hunting shelters Petroglyphs/pictographs 3-4, 5-8…Room Pueblos Roasting pits Shrines Field or Agricultural Site Sweathouses Rock shelters Pit House Site Animal traps U-shaped structures Misc. Other Categories Great kivas Wilcox (1978) Compounds North et al. 2003 Fieldhouses Pithouse sites Artifact scatters Farmsteads (year round) Field houses Field houses Hamlets (seasonal) Pithouse sites Seasonally-occupied villages Room block sites Reid (1982a) Rock art sites McAlllister and Plog (1978) Lithic Hunting/gathering camps Sherd-Lithic Moore (1978) Field houses Cobble compound Field houses/seasonally utilized Abandoned/ with pueblo farm structures (SUFS) aborted settlements Pit house Hunter or shepherd lodges Neighborhood settlements Cobble rooms Eagle-hunt houses Cobble pueblo Pinon pickers shelters Dosh (1988) Masonry pueblo Shrines Habitation sites Rockshelter Habitations Small architectural sites Retreats Temporary camps/resource areas Plog (1981a) Travel lodges Lithic/Ceramic/Artifact Migration shelters Scatters Ciolek-Torrello et al. (1994) Pithouse sites Playhouses Farmsteads Pueblo sites Granaries Field houses Great kivas Towns Hamlets Compounds Seasonal villages Villages Water control Resource procurement/processing Devices Petroglyphs/Pictographs Herr 2001 Shrines One-room Solometo (2006) Rock shelters Great Kiva Lookouts Pit house Refuge (isolated) 2 rooms Hartman (1990) Habitation (>5 masonry rooms) Refuge (proximate) 5-8 rooms Food gathering/preliminary Small structure (2-5 masonry Defensive habitation15-25 rooms processing camps rooms, >2 jacal) Farmsteads/small Limited activity (inc. rock art, architectural sites(3-8 rooms) water control) Field houses (1-2 rooms)

Summary of Expectations

The previous sections have clearly shown there is has been a lengthy history of archaeological research in the area. Many sites have been documented, and they have inspired a number of theories and expectations regarding the distributions of the most abundant types of sites, as well as the more extraordinary, such as the large villages of the late Pueblo periods. In the following paragraphs, I expound upon five primary perspectives identified in previous research and their expectations regarding the distribution of small archaeological sites. These form the basis of the GIS-based models tested in the following chapters to assess the archaeological geography of small sites.

Environmental Determinism

Many researchers working in the study area have hypothesized that various types of small sites are associated with specific ecological settings, including elevation, vegetation type, soil type, aspect, slope, distances to drainages (however they may be defined), and other topographic characteristics, including prominences. A fundamental step is establishing whether or not these relationships are random. As we will see in the following chapters, a quick glance at the geographic distribution of small sites suggests they clearly are not arbitrary, but the specific relationships among environmental parameters warrant further attention.

In the following chapter I assess these relationships by comparing known site locations with GIS-based representations of some of those environmental characteristics.

Simple visual examinations and statistical tests are used to assess these relationships and determine which, if any, are strongly recorded with small site locations. Various subsets are examined to attempt to ascertain variations among site types, including subsets of field houses and similar-sized structures.

This is a common approach to traditional inductive probability models in archaeology, and a regression-model is presented and tested against existing archaeological survey data. The resulting probability and sensitivity surfaces derived from this model are examined in more detail and compared against the results of later analyses. Inductive probability, or predictive, models have been criticized for being slanted toward environmental determinism, but they are often effective, as indicated by various model performance statistics. These statistical techniques, most notably logistic regression, provide robust means for assessing the significance of different environmental variables without relying upon conjecture or subjective judgment to determine and quantify significance (Kvamme 2006:12). The results should also serve as a baseline for assessing other models.

Agricultural Suitability

Given the presumed association of many small sites with agricultural pursuits, it seems to be a natural progression to explore a more deductive approach to agricultural

93 suitability in the study area. Models of agricultural productivity have been applied to other areas with surprising results (e.g., Wills and Dorshow 2012). Although similar environmental variables may be used in an inductive approach, a deductive model drawing upon some basic assumptions regarding agricultural pursuits in the area and in other similar settings is proposed.

These accounts suggest a sufficient number of frost-free days is required to raise a crop, and abundant precipitation is needed to support rain-fed agriculture. Schlanger and

Larralde (2008) contend that Ancestral Pueblo occupation often corresponds with distributions of piñon-juniper woodland because “dry-land maize agriculture requires good summer monsoons plus 120 frost free days,” a phenomenon at often found at the same elevation preferred by piñon pine. Locations likely to receive sufficient passive runoff from rainfall events are also favored, and there is a preference for deep soils derived from sandstone parent material. These are all important variables to consider and

I attempt to derive these landscape characteristics from available data and examine the results of a deductive modeling process against the locations of known sites, including subsets of sites with presumed agricultural functions. These results are also compared against the previous inductive model to assess their potential significance and utility.

Interestingly, dense stands of piñon-juniper woodland on Cedar Mesa were previously determined to correspond with areas that received higher amounts of precipitation and were marked by deeper and sandier aeolian soils (Haase 1983). These ecotones were determined to be the most suitable lands for dry farming, whereas “the sparse piñon-juniper is associated with thin, non-arable soils, and this plant community,

94 along with exposed bedrock and the lower elevation shrublands, was not considered to have much agricultural potential” (Matson et al. 1988:248).

The site databases from both the ASNF and CNF identify broad classes with implied agricultural functions. Does this model identify likely locations of agricultural fields that have not been previously recognized during survey projects, or do they simply conform to expectations and lend support to the few field areas that have been previously suggested? Do the results imply that agricultural suitability was or was not a primary factor in locating small sites?

Migrant Communities and Seasonal Uses of Common Pool Resources

Research by SCARP and others suggests much of the study area was sparsely populated throughout the Archaic and early Formative periods. Herr (2001) proposed a frontier model to account for community organization in the area, which was probably occupied seasonally and sporadically, but witnessed a growing population by A.D. 1000 that could not be accounted for by natural population growth alone (Newcomb 1999).

Groups of people were migrating into the area from surrounding regions; communal great kiva sites were apparently constructed to integrate these groups, who were occupying a

“land-rich and labor-poor environment” (Herr 2001:7).

Bayman (2007) and Bayman and Sullivan (2008) applied the concept of common pool resources to studies of land use and land tenure in the Papagueria and Grand Canyon regions, and it may be appropriate for understanding some aspects of the study area.

Common pool resources can be either natural (such as oceans, lakes, forests, etc.) or man- made resources, but they generally comprise systems so large that they are difficult to control and prevent others from using them (Ostrom 1990). Non-agricultural resources may have been primary attractants in the area, including fuel wood and wild foods (Floyd and Kohler 1990; Schlanger and Larralde 2008), which are found in expansive belts of forest. The attraction of these ecosystems for hunter-gatherers is attributed to the importance of piñon nuts as a food source for both humans and fauna, although massive die-off of woodlands may have occurred during extended periods of drought and slow recovery of the woodlands could have made agriculture more reliable than hunting and gathering in these situations, and this may have influenced settlement patterns among various groups adapted to piñon-juniper woodland environments. Just as these non- agricultural resources covered broad zones that probably could not have been easily controlled, suitable agricultural lands may have been a similar type of widely available common pool resource.

Preucel (1990:57) suggests the “relative abundance and spatial distribution of seasonal sites can be used as a rough index of seasonal circulation,” offering a resilient adaptation to reduce risk from environmental and climatic variation. If common pool resources characterized some of the economic importance of the landscape, small sites would be expected to be widely distributed and persistent, but perhaps shifting through time.

Distance Decay Model – Proximity to Large Sites

Haury (1985), Preucel (1990), and others have highlighted the relationships between small architectural sites and larger settlements. This is often characterized as a distance decay model (or gravity model), generally suggesting that seasonal circulation results in a combination of both temporary shelters, such as field houses or other limited activity sites, and permanent village residences. In this manner, seasonal circulation would have facilitated use of more distant fields, presumably supplying direct economic benefits afforded by local environmental factors. Preucel (1990:177) also sees it as a conflict resolution strategy that “may have been linked to the minimization of the inevitable social stress which attends communal village life.”

Preucel’s findings (1990:179), as well as Haury’s suspicion (1985), suggest increased population aggregations generally require traveling greater distances, as suitable and productive areas in the vicinity of population centers become occupied more quickly. In the study area, aggregation increased through the Pueblo periods, apparently culminating in occupation of a limited number of large sites. The distance-decay model would suggest high densities of small sites surrounding these highly populated areas. In a similar vein, Doolittle’s work among smallholder agriculturalists (1988) observed a trend that periods of use tend to become lengthier through time because of the cumulative value of the small investments in cultivating the land. Although this principle was primarily elucidated on the basis of investment in canals and other related technology, it is also applicable to other labor investments that may seem less technologically

97 advanced, but still required a substantial amount of effort in accordance with various environmental or social factors, such as demographic pressure.

Competition and Conflict – A Tragedy of the Commons?

Finally, I explore evidence for competition and conflict in the study area and how this may be reflected in distributions and other characteristics of small sites. Solometo

(2004, 2006) has suggested the region was “organized for war,” and it is evident in the types of small sites encountered in parts of the Chevelon Canyon area. Others have suggested Pueblo III period hilltop sites, such as Pottery Hill, may have been situated for defensive purposes.

Johnson et al. (2005), working with data from the Central Mesa Verde region of southwestern Colorado, used an agent-based model that incorporates soil science, paleoclimatology, and hydrology to emphasize long-term fuel (wood) use patterns. Their findings suggest that various characteristics regarding settlement location were strongly influenced by wood supplies that quickly became limited in the region. Is there evidence for a “tragedy of the commons” scenario (Hardin 1968; Kohler 1992b; Kohler and

Matthews 1988), whereby previously sustainable use of common pool resources eventually exceeded available supplies, or evidence for climatic variability or other event that disrupted previously established patterns of land use? The defensive postures identified by Solometo (2004, 2006) could have been a result of overuse of potentially productive agricultural land (Fertelmes and Barton 2007; Peeples et al. 2006). Peeples et

98 al. (2006:13) note that “the thin and rocky soils that characterize the study area today have not been cultivated since prehistoric farmers abandoned it over 700 years ago.”

Could persistent vegetation characteristics be the result of past agricultural practices or are other land uses during the historical era, including grazing, fuel wood harvesting, and mechanical removal of vegetation to improve conditions for livestock? These trends may be difficult to fully assess throughout the study area, but there should be suggestions based upon the distributions of small sites and assessments of their long-term use.

Summary

Clearly, the perspectives reviewed above are interwoven and not necessarily mutually exclusive. Many touch upon similar aspects of the socio-ecological setting.

The next chapter focuses on situating small sites within their environmental setting, and these relationships are then explored further by addressing the relationships among different categories of site types. Then, in an attempt to glean more meaning from some of the most commonly encountered archaeological sites in the study area, I hope to shed new light on the nature of the archaeological landscape of land now being managed by the National Forests and provide insights regarding the communities who previously negotiated their livelihoods and identifies with the landscape of the Mogollon Plateau.

CHAPTER 4.

IDENTIFYING SMALL SITES WITHIN THE MODERN LANDSCAPE

The Mogollon Rim forms the geologic and geographic “spine” of Arizona.

Depending on where one encounters this feature in east-central Arizona, it may range from a dramatic topographic prominence offering broad vistas of the basin and range country to the south, or to a far more subtle rise that passes through the conifer forests and slowly transitions to the plateaus of the north. Stretching across the central part of

Arizona, it supports an ecological transition zone characterized by a remarkable diversity of flora and fauna, as well as rocks and minerals. The many valleys and plains that emanate from the Rim have experienced a florescence of human cultures for at least

12,000 years. In that time, the high peaks, springs, streams, caves, and a host of other natural features have held significance for many groups. The most recent immigrants and current residents of the area have transformed the landscape in many significant ways and have intersected a variety of past landscape uses.

Drainages that flow south from this part of the Rim feed into Arizona’s Central

Mountains, dissected canyon country that eventually gives way to the Basin and Range physiographic province (Figures 1.3 and 4.1). North-flowing canyons eventually feed into the broad valleys that emanate from its crest, opening up to the southern edge of the

Colorado Plateau and leading toward the Little Colorado River, which parallels much of the Rim after its northern exit from the White Mountains. The study area’s southern

100

Figure 4.1. The Mogollon Rim and the view toward the mountainous transition zone.

boundary generally coincides with the Mogollon Rim and the Little Colorado River is roughly parallel and 40 km north of the study area’s northern boundary (see Figure 1.3).

Colloquially, the area is generally referred to as “the Rim country,” but the relatively flat terrain north of the Rim is sometimes called the Mogollon Plateau. The most prominent modern communities in the area include many that were founded by

Mormon colonists, who followed Brigham Young’s call in 1870s to settle the arable valleys along the Little Colorado and its tributaries, including Silver Creek (Abruzzi

1993). Notably, most of the larger modern communities are located on the margins of the

101 study area (see Figure 1.3). Heber and Overgaard are exceptions however, and there is a series of smaller communities between Heber-Overgaard and Pinetop-Lakeside, including Clay Springs, Pinedale, and Linden, along with a number of other privately owned parcels. Most of the country immediately north of the National Forest boundaries is characterized by “checkerboard” ownership, characterized by alternating square-mile sections of private, State, and BLM property. National Forest property was once included in this ownership pattern but was aggressively consolidated by Forest Service

Lands and Survey personnel, especially in the 1970s and 1980s.

Although best characterized as a transition between different biomes, the Rim region is also notable because of the influence that it plays upon local and regional climate, especially precipitation. As a general rule of thumb within the study area, the orographic effects of the Rim cast a rain-shadow over the terrain that gently slopes toward the Little Colorado River basin (Kaldahl and Dean 1999). Precipitation and elevation generally decrease as one moves farther away from the Rim. In a similar sense, the particular topographic relief of a given site also influences local precipitation patterns.

This results in diverse vegetation communities influenced by the many tributaries that drain the Rim country.

Both of these general trends, plus their influence upon the average length of the frost-free season, would have undoubtedly been discernible and have probably played vital roles in how different communities have made a living from the land. The range in elevation and the orographic rain-shadow effect also influence snow cover, which can be significant and accumulate for lengthy periods at higher elevations (Figure 4.2).

102

Figure 4.2. Elevation throughout the study area derived from USGS National Elevation Dataset. 102

103

This climatic gradient has also been documented by the Forest Service’s

Terrestrial Ecosystem Survey (TES) (Figure 4.3). The TES of the Apache-Sitgreaves NF

(USDA Forest Service 1989) was generated from stereoscopic analysis of 1:24,000 aerial photographs. The TES is a process still being completed for all National Forests in

Arizona. It maps variability and interactions among climate, topography, geological and soil characteristics, and plant communities to identify specific ecosystem units. A similar survey was completed for the CNF in 1995. On the ASNF, the survey methods (USDA

1989) were used to delineate 123 Terrestrial Ecosystem Units (TEU), which were subsequently verified through field inspections. These were combined with CNF units and dissolved according the Climatic Gradient Segment and Step identified by the survey

(see Figure 4.3).

In a similar and related sense, the study area exhibits an ecological gradient, which also corresponds with topography and related geological and soil conditions

(Kaldahl and Dean 1999; Laing et al. 1989; Plog 1981a, b). The highest elevations, typically above 8,200 feet, although subject to local topographic variation, support pine- fir forests with patches of open meadow, giving way to lower elevation ponderosa pine forests (Figure 4.4). Although often exceedingly dense in some places but also widely affected by the large and recent wildfires, these forests supported more park-like settings in the past, with fewer but larger trees spaced apart at greater distances. These settings also support a host of other useful plants that probably thrived when they were more actively associated with horticultural communities in the past (Moore 1979; Rainey and

Adams 2004).

104

Figure 4.3. Climatic gradient mapped for the study area by the Terrestrial Ecosystem Survey project (the null data areas are Mormon Lake on the west and Pinetop-Lakeside on the east) 104

105

Figure 4.4. Vegetation communities digitized from Brown and Lowe (1980). 105

106

Interspersed among the dominant belt of Ponderosa pine are stands of oak trees, as well as increasing densities of Alligator juniper and then piñon pine as one moves farther away from the Rim. These piñon-juniper woodlands between 6,400 and 7,200 are ubiquitous across the northern portion of the study area, giving way to grasslands that are primarily located beyond the Forest boundary. While the belts of Ponderosa pine and piñon -juniper woodland may appear fairly homogenous, the predominant species are interspersed to varying degrees. This diversity has been mapped in more detail by the

TES.

These ecotones are situated upon tablelands of variable relief, generally alternating between north trending ridges and drainages. Originating from reworked sedimentary deposits known as Rim Gravels, the entrenched canyons typically pass through deposits of limestone and sandstone before giving way to areas with deeper soils typically derived from sandstone (Plog 1981a:4-5; USDA 1989). The study area is flanked by the Springerville Volcanic Field on its eastern end, where there is a preponderance of cinder cones and basalt. Volcanic deposits also characterize much of the northwestern margins of the study area as well.

One of the primary potential shortcomings that GIS-based analyses of archaeological and systemic landscapes must face is the problematic tendency of imposing present environmental conditions on the past. Although this approach may be suitable for stable aspects of the environment such as geology, it may be quite unsuitable for environmental characteristics that vary through time. Because of this variability, locations now considered suitable for a given land use practices, such as dry farming,

107 may not have been so at different times in the past. Conversely, locations that were suitable in the past may not appear so in light of current environmental conditions.

These substantial issues may be mitigated in part by factoring in measures of low and high frequency variations in environmental conditions (Dean 1988, 1996, 2010).

Dendroclimatic reconstructions provide useful indicators of high frequency variability, while low frequency fluctuations may be detectable in pollen, geologic, and packrat midden studies. Regarding vegetation, these studies have generally shown the communities do not seem to have shifted dramatically in the past 1500 years, although there have been climatic variations (Kaldahl and Dean 1999) and changes in the fire regime (Roos 2008b).

There has been no extensive effort to systematically assess possible changes in the main vegetation communities during the past two thousand years. Some generalizations are appropriate, but a more detailed, integrative analysis in the vein of Periman’s (2005) study of north-central New Mexico’s Rio Del Oso Valley would be a valuable contribution. Anderson (1993) studied pollen and plant macrofossil remains recovered from Potato Lake, providing insights regarding climatic and biogeographic changes during the past 35,000 years. Results demonstrate that cooler and wetter conditions prevailed throughout the early and late Wisconsin (until ca. 10,500 B.P.) then shifted toward current conditions.

In addition to paleoclimatic variability, the influences of modern land use systems upon environmental conditions need to be considered. In the American Southwest, this has been dramatic. In fact, recent historical land use practices may be more substantial

108 sources of transformation of archaeological, systemic, and “natural” landscapes than paleoecological variability. Forest Reserves were created for the protection and management of timber reserves and watersheds, among other interests. The Black Mesa

Forest Reserve was originally created in 1898, and the CNF and ASNF were established in 1908 (Baker et al. 1988:41). Private ownership through the Homestead Act and other legal routes further concentrated focus on reliable water resources, as well as access to timber and rangeland. Furthermore, private ownership has regularly excluded these areas from archaeological inventory and resource protection under federal law, while authorized developments on federal lands of obliterated, altered, or otherwise obscured sites.

The replacement of indigenous management regimes with Euro-American practices, termed “the shift from a forest for people to a forest for trees” by Huntsinger and McCaffrey (1995:157), has also influenced environmental conditions, dramatically in some cases. The preponderance of spring modifications, stock tanks (regularly constructed during the late 19th and early 20th centuries through present times in locations that are naturally wetter than surrounding locations), and dams have had dramatic effects on current hydrologic regimes. Just as recent ecosystem characteristics are not necessarily representative of the past conditions, the locations of water sources are also not directly applicable to the past. Spring locations change through time, the water table has been influenced by wells and pumping, and dams and earthen tanks have disrupted the natural flow of watercourses.

109

There are a host of modern cultural formation processes that have affected our knowledge of the region’s archaeological geography, including the establishment of federal and private property boundaries. Ongoing uses of the landscape have been influenced by these boundaries, resulting in several modern landscape formation processes. Vehicles for transferring public property into private ownership also dramatically encouraged the growth of transportation systems. Forest Homestead applicants often targeted valuable resources like perennial water sources and arable land, the same settings often associated with large, late Ancestral Pueblo villages in parts of the study area. Historic land tenure systems encouraged control of reliable water sources and productive agricultural land, and this has influenced archaeological research in the region, a pattern commonly repeated throughout much of the American Southwest. The ongoing elaboration of contemporary landholding practices has resulted in vast expanses of public land intersected by diverse patterns of private ownership, which often include many of the areas that attracted Native American occupants. This reality complicates many aspects of archaeological research and cultural resource management. Some specific complications in regard to the archaeological record include destruction of portions of the record attributable to a variety of land uses, balkanization among groups of academic researchers (i.e., Reid 1999) and land management archaeologists, and innumerable irregularities in the interpretation, documentation, and management of the archaeological record.

Springs have been important components of this landscape for Native communities and recent immigrants alike. These quintessential place-makers have been

110 dramatically altered in most cases, whether encased in concrete, excavated, or diverted to a wellhead or pump and re-appropriated for livestock. On National Forests and other public lands, springs have been targeted by many modern land uses, including homesteading, ranching, farming, mining, and lumber milling, among others. However, several springs in the surrounding area have been documented as archaeological sites and as shrines through ethnographic research (Greenwood 1983). Some of these sites, such as Mineral Springs, on the eastern margins of the study area, are referred to as “bead springs” in homage to the numerous beads left as offerings.

The inherent conflicts between private property owners and federal land management have strongly influenced documentation, preservation, and interpretation of historic sites. Within the United States, a barbed wire fence with often questionable accuracy can mean the difference between prison time and six-figure profits, as demonstrated by often misguided looting ventures at large sites on private inholdings surrounded by National Forest, such as Bailey Ruin, Double Circles Pueblo, Pinedale

Ruin, and other larger sites, including Pottery Hill and Roundy Pueblo, that are located on Forest land but in close proximity to private property.

It is critical to recognize the influences of past vegetation management projects such as chaining and other invasive, mechanical methods, but also grazing, fuel wood harvesting, and altered fire regimes. Landis and Bailey (2004) demonstrate some of these issues at nearby Anderson Mesa, along the northwestern margins of the study area. They also caution that “blanket prescriptions” for tree reduction of piñon-juniper woodlands and savannas undermines the heterogeneity of these ecosystems, especially as they

111 existed before European settlement of the area around 1860. They note that this variability is dependent upon soil types, generalized by parent material (sandstone, limestone, and basalt), which can also correspond proportionally with vegetation associations, including tree and grass cover (Thatcher and Hart 1974). Unlike Ponderosa pine forest (Roos 2008b), there is little evidence to support the idea that low-intensity surface fire regimes have dominated in piñon-juniper woodland (Baker and Shinneman

2004). Influences of topography and soil types are also important factors to consider

(Romme et al. 2003).

Native American reservation boundaries and ancestral territories frequently coincide with federal and private property, especially in Arizona, and the study area is a clear example of both. Much of the length of the southern boundary of the study area is shared with the Fort Apache Indian Reservation, and the ancestral territories of several other tribes encompass at least portions of the area, including Zuni, Hopi, Navajo, and other Apache groups (Senior 2004). Although the landscapes of these ancestral territories have been dramatically altered in many cases by modern land uses, the archaeological record of the area offers an opportunity to critically reassess the area’s socio-ecological history.

Archaeological Site and Survey Data

Previously compiled information regarding the locations of archaeological sites, as well as non-site locations documented by previous archaeological surveys, form the

112 backbone of my analysis of small sites. Generations of archaeologists have applied their unique skills and interpretations of the past in a variety of ways while working in the study area. Although there is significant variability in the data and many potential sources of error, they offer important insights to our growing understanding of the archaeological landscape of the Mogollon Plateau region. The site and survey databases are reviewed in the following sections, paying particular attention to the categories that have been employed to organize the data.

Archaeological Site Databases

Obviously, the most important component of the present study is the spatial and descriptive information of the sites themselves. The datasets used for this analysis were received from the ASNF and CNF Heritage Resources Management Programs. The

ASNF includes locations of approximately 6,983 archaeological sites spanning all five districts of the Forest. The ASNF portion of the study area includes 83% (n=5797) of these sites despite the fact that it accounts for only 42% of the entire ASNF, which includes more than 2 million acres. The two CNF districts include 1,535 locations.

Survey coverage has been somewhat variable throughout the area (Figure 4.5), and many of the surveys completed on the ASNF followed sampling procedures that resulted in less-than-complete coverage. These protocols were largely used in areas of presumably low site density, largely based upon elevation ranges of known sites revealed from earlier sample surveys for overview and planning projects. Although several assumptions are

113

Figure 4.5. Survey coverage in the study area (less-than-complete surveys are shown as hatched symbols).

113

114 required to estimate total survey coverage, I estimate that approximately 18% of the study area has been completely surveyed. Assumptions regarding width of linear surveys and areas of point surveys affect the total, as well as sampling intensity in less-than-complete surveys.

When displayed as a density surface interpolated from point locations of sites throughout the study area (Figure 4.6), the density distribution exhibits some notable concentrations. In the western portion, site density is quite variable, and there is a notable increase along the northeastern boundaries of the CNF district. This concentration continues to the east along the northern boundary of the study area.

Towards the eastern third of the study area, there is an indication of slight break in density, which eventually shifts toward the southern boundary closer to the Mogollon

Rim.

The errors associated with the archaeological inventory data are often complex and tedious, but they also reveal issues applicable in other regions. Data reliability can be a significant issue for many data sources, and the site databases used here are no exception. A brief review of their origins is warranted. Although earlier investigators primarily associated with educational and research institutions had collected and published an array of site information, formal and reconnaissance surveys were not systematically documented by federal land management agencies until the early 1970s, following passage of the National Historic Preservation Act of 1966 and subsequent amendments, as well as Executive Order 11593, which opened the first large-scale hiring of Forest Service archaeologists (Gillio 2005:36). Survey areas were initially plotted on

115

Figure 4.6. All recorded archaeological sites in the study area.

116 paper copies of 15’ USGS topographic maps, then the 7.5’ series as they became more widely available. Official survey projects were few at first, but as cultural resources compliance work became more institutionalized and labor intensive, the number of surveys grew rapidly in support of more traditional Forest Service efforts, including engineering, range, timber, and a variety of other projects.

Growing data management needs for cultural resource management programs led to adoption of the Cultural Resources Automated Information System, or CRAIS, in

1977, giving rise to the first computerized database of archaeological and historical site locations and survey projects for the Southwestern Region of the Forest Service (Gillio

2005:50-51). The growing backlog of data entry (originally involving punch cards) became apparent during the following decade, which included a burgeoning amount of timber sales, but an updated site form and recording procedures were established.

Discussion led to the formal definition of archaeological sites, defined since 1987 by

USDA Region 3 Forest Service Handbook 2309.24 as:

“A location of purposeful prehistoric or historic human activity. An activity is

considered to have been purposeful if it resulted in a deposit of cultural materials

beyond the level of one or a few accidentally lost artifacts. Locations of human

activity not classifiable as sites by this definition should be considered isolated

finds.

A cultural resource qualifying as a site under this definition should exhibit at least

one of the following:

117

a. One or more features.

b. One formal tool if associated with other cultural materials, or more

than one formal tool.

c. An occurrence of cultural material (such as pottery sherds, chipped

stone, or historic items) that contains one of the following:

1. Three or more types of artifacts or raw material.

2. Two types of artifacts or material in a density of a least ten items per

100 square meters.

3. A single type of artifact or material in a density of at least 25 items per

100 square meters.

These criteria may be modified, where appropriate, based on a professional

cultural resource specialist's judgment.

The boundary of a cultural resource site shall minimally include:

a. All features, formal tools, and identifiable activity areas.

b. All areas of artifactual debris exhibiting a density of ten or more

cultural items per 100 square meters.”

These site definition criteria purposefully left some room for interpretation because of the great range of variability in the regional archaeology, as well as within

118 even a single National Forest. Professional archaeologists working in certain regions or in areas of high site density versus low density indubitably apply the criteria differently, and their professional judgment in concert with recommendations from Forest Service archaeologists familiar (hopefully, but not always) with a particular region are relied upon to correctly apply the definition. Regardless of the definition, a wide array of sites has been recorded, although there has been significant variation among regions.

By the 1990s, true to the form of government fueling development in geospatial applications, GIS became a hot topic of discussion among Forest Service archaeologists.

GIS layers representing archaeological and historical sites and cultural resource surveys were recognized by regional managers as important corporate data, paving the way to dedicate needed resources to generate the data (Gillio 2005:144-145). Various National

Forests employed different methods of generating this information. Years of backlog on the ASNF were eventually added to the corporate database by tracing boundaries of sites and surveys that were eventually scanned and digitized. Point, line, and polygon layers were originally created for both sites and surveys.

CRAIS was eventually replaced by a new relational database called Heritage

Infra, which began to take hold after 2000. Legacy data from CRAIS was transferred to

Infra, and Infra is now the appropriate venue for updating GI and related attributes for both surveys and sites. Site data from the Coconino National Forest followed a slightly different trajectory, as many sites were originally registered following Museum of

Northern Arizona protocols. Acceptance to use the CRAIS database was slow at the

Forest level, but subsequent Infra and GIS reporting requirements led to their eventual

119 inclusion in the appropriate databases. Regardless, data from both Forests has been compiled by many institutions and many individuals, and the interpretive and classificatory hurdles are immense.

At this time, systematic information regarding sites and surveys on the National

Forests located in Arizona is not included in AZSITE, although many sites and surveys have been registered with the system (and via previous institutions, such as the Arizona

State Museum). Unfortunately, the sporadic nature of this documentation and registration with federal institutions offers much potential for misinterpretation and underreporting of previous research. Federal reporting requirements have directed development of centralized, federally-controlled repositories, but a desire to control access to site location information has also played a significant role. Eventual inclusion in AZSITE seems like a desirable and expected future development, but there seems to be limited movement in that direction at this time. The ability to control information serves important functions, requiring individuals interested the area to consult with specific personnel. Interestingly, a similar model of local control has been adopted for tribally- controlled records.

The archaeological site location databases do not address “off-site” remains, typically recorded as isolated occurrences on the National Forest lands in the study area and not assigned formal site designations (Bintliff and Snodgrass 1988; Heilen 2005:92-

93; Rossignol 1992). The topic of formally inventorying such finds was debated among

Forest Service archaeologists in Arizona and the Regional Office in Albuquerque, eventually resulting in the conclusion that isolated finds would not be considered (Gillio

120

2005). Features without associated artifacts, low-density scatters of artifacts that do not meet site-definition criteria, and single artifacts are not formally tracked, and in those cases where off-site finds have been systematically documented, they have not yet been integrated into GI databases. If this effort is undertaken, it will be equally important to identify surveyed areas where off-site materials have not been documented.

Few of the sites included in this analysis have yielded absolute dates.

Fortunately, primarily due to the success of tree-ring dating in the study area and surrounding regions, along with subsequent ceramic seriation (Hays-Gilpin and van

Hartesveldt 1998; Mills and Herr 1999), it is possible to assign relative dates to many of them. CARP researchers seized upon varying proportions of white ware ceramics to establish chronologically-sensitive combinations of technological characteristics

(Solometo 2004). In those cases where specific phase or period assignments have not been possible, broader categories have been used, primarily determined on the basis of the presence and absence of specific wares. In other cases, a well-established chronology of projectile point forms for the region (Tagg 1994) has proven useful, although not especially for the time period in question here.

Despite the success of ceramic cross-dating in the study area and surrounding areas, it is not entirely possible to reliably re-construct the temporal distribution of small sites throughout the study area. Diagnostic artifacts have not been consistently documented for all sites, and even in those cases where great care has been taken to inventory them, one must acknowledge the fact that surface assemblages do not necessarily reveal the entire length of a site’s occupation, and they have indubitably been

121 subjected to a number of natural and cultural formation processes. Just as measures of artifact diversity are strongly influenced by sample size, so are indices of the presence and absence of various wares.

As technology has evolved, cartographic representations of site locations have shifted. There has been a general movement away from retaining point and line features, since virtually all features of interest can be represented by a polygon. This has been done for a number of practical reasons. Cartographic scale was an important element in the past, especially prior to the widespread availability of 7.5’ USGS topographic maps.

Even in these cases, a simple dot drawn on a map could easily represent a diameter of

100 m or more.

The growing and increasingly widespread use of GPS has alleviated some of these locational issues, but verification of previously recorded site locations has been limited.

The geographic accuracy and precision of these locations are often dubious. As will be shown, many of the most densely occupied areas of the study area occur within piñon- juniper woodland, often of such a density that traditional orienteering methods are difficult to implement and reproduce. Aside from the practicality of seeing landmarks on the horizon to accurately plot oneself on a map, surveyors experience varying degrees of disorientation due to the nature of the landscape and vegetation.

The reliability of archaeological landscape data is a significant issue, and reveals numerous sources of potential error. Older site location data is generally more subject to error than more recent locations (particularly since the GPS era), particularly in terms of locational accuracy. The resolution of the data can also vary greatly. Misrepresentations

122 can be generated through the database development process through classification errors, inappropriate generalization, and exaggeration (or underestimation). Map projections and coordinate systems can also introduce error. Accuracy of associated attributes is particularly problematic, resulting from interpretation errors, topographic or other coding errors, or perhaps even inadequate description. A peculiar problem in the study area has been the shifting boundaries of Ranger Districts through years, since the FS site number incorporates a District number. This has resulted in some sites being assigned more than one FS number, and in many cases they have also been assigned museum numbers (ASM

MNA, etc.) and project-specific numbers (be they FS projects or CARP, for example).

The methods and techniques used to generate GIS datasets can also introduce a variety of inaccuracies, including digitization and editing errors. Ultimately, human error is just difficult to fully account.

In order to attempt to alleviate many of these potential sources of error concerning site characteristics, I select subsets of sites that span various portions of the study area to examine some of the more fine-grained research questions. Many of these sites have been recently inspected. In the case of some characteristics of these groups of sites, there is a certain level of consistency that relieves some of the doubt regarding the site locations and their interpretation. In some cases, these interpretations provide interesting insights. Also, despite the many potential problems associated with the site location data, it is possible to take some solace in the fact that spatial autocorrelation can alleviate some of these concerns. Although many site locations may be inaccurate, the scale of analysis dictates that the surrounding locations are likely similar to the actual location in many

123 regards due to the nature of the geographic information being analyzed. Also, by examining catchment areas surrounding sites through focal or neighborhood analytical functions it is possible to account for characteristics of surrounding areas.

Comparing Site Databases

For the first phase of the analysis, I examine landscape-scale trends regarding the relationships among various types of small architectural sites with their environmental settings. In order to accomplish this, it is necessary to identify subsets of sites for comparative purposes. Some obvious variation in the site typologies is apparent in

Figures 4.7 and 4.8, but there are also many common trends.

2500

2000

1500

1000

500

Figure 4.7. Site type frequencies from the Lakeside and Black Mesa Ranger Districts of the Apache-Sitgreaves National Forests.

124

500

400

300

200

100

Figure 4.8. Site type frequencies form the Mormon Lake and Mogollon Ranger Districts of the Coconino National Forest.

On the eastern side of the study area (ASNF), the most common site class by far consists of artifact scatters (i.e., surface assemblages that include flaked stone, ceramic, and/or ground stone artifacts, but no discernible features were identified). Field houses and small sites with single rooms are the second and fourth most abundant; sites estimated to have five or more masonry rooms are the third most abundant site type.

Lithic scatter and pithouse sites are present, but their relative frequency seems to confirm earlier research suggesting the area was sparsely occupied until the Pueblo periods.

In the western portion of the study area (CNF), field houses are the most abundant type of Native American archaeological site, followed by artifact and lithic scatters.

Small clusters of rooms have been documented in a slightly different manner,

125 distinguishing sites not considered to be field houses with less than four rooms from those with more than five. Another notable category identified in the western portion of the study area is agricultural field areas. These have been noted anecdotally on the eastern side but have not been afforded a separate category.

In order to further facilitate analysis of site typologies across the study area, similar types are merged and compared side-by-side in Figure 4.9. Although the total numbers are far less for the CNF sites, the relative frequencies throughout the two management units is fairly similar once the disparate types have been combined.

Figure 4.9. Side-by-side comparison of Native American site type frequencies from the Apache- Sitgreaves and Coconino NF study areas.

126

Removing some layers from the site typologies employed throughout the study area sheds a considerable amount of light on the distributions of small architectural sites

(Figure 4.10). The vague aggregations seen among the total site distribution earlier are much clearer and seem to form notable clusters. They are identified in Figure 4.10, and referred to as the Mogollon Rim and Chevelon/Chavez site clusters in the remainder of this text. However, it is also important to further consider the nature of the archaeological survey database when interpreting these distributions.

Archaeological Survey Database

Locations of archaeological surveys throughout the study area are presented in

Figure 4.11 along with sites identified as field houses, other small architectural sites

(single rooms, jacal structures, and other structures with less than five rooms), and larger

Pueblo sites (more than 5 rooms). Large portions of the area have been surveyed, especially in the southern portions in the vicinity of the Mogollon Rim. There have been numerous intensive surveys in the northern portion, but they have not been as widespread. Throughout the area, nearly all previous inventory projects consisted of pedestrian surface survey. Survey standards have typically involved examining evenly spaced transects, typically no more than 50 foot-wide intervals. Subsurface “probing” has rarely been completed as a component of these inventories, and certainly not in any systematic manner, such as intensive “shovel-testing” methods that are common practice in other regions.

127

Figure 4.10. Density of small structure sites in the study area, revealing the Mogollon Rim cluster in the southeast, 127 and the Chevelon/Chavez cluster along the northern boundary, to the northwest (compare with Figure 4.6).

128

Records of previous Forest Service-sponsored inventories of historic properties constitute areas among sites, as well as their relationships with a number of geographic and environmental factors. However, in those cases where measurements are derived from relationships with other sites (e.g., Euclidean distance to great kivas, or relative composition of vegetation within a catchment area) failure to recognize those areas that have not been inventoried for historic properties can lead to significant biases in interpretation or false conclusions. In a similar sense, it is important to recognize areas where sites have not been found. Although Kvamme (2006) has implored others to not concern themselves with non-site locations in GIS models because there could potentially be subsurface deposits, we must recognize that the archaeological geography documented in the area is predominantly based upon surface characteristics.

Many of the previously completed inventory projects reflect the legacy of

“sample surveys” completed on the ASNF. In these cases, one must also use caution and recognize that polygons representing some of these inventory projects do not accurately reflect the specific areas on the ground that were surveyed by archaeologists. Many such surveys were completed during the 1980s, when timber sales were a driving force in cultural resources compliance on the National Forests. In areas where previous surveys, expert opinion, and agreements with the State Historic Preservation Officer indicated likelihood of encountering archaeological sites was low, various survey strategies were used (Donaldson 1975). These efforts account for much of the large-scale survey in regions with low densities of small architectural sites and large Pueblo hamlets and villages.

129

129 Figure 4.11. Density of larger pueblo sites (believed to include five or more rooms) in the study area.

130

Closer inspection of the site density distributions identifies two fairly obvious conglomerations of Pueblo sites (Figures 4.12). I refer to these as the Mogollon Rim and

Chevelon/Chavez site clusters. The cluster boundaries I have defined are based upon site locations within the study area boundaries. Additional sites certainly do exist beyond the administrative boundaries, but they have not been included in the density surfaces.

Relationships with survey results in surrounding areas will await a later project. Also, the current shapes of these clusters have no doubt been influenced by survey coverage, and future projects will add new insights.

The Mogollon Rim cluster forms a dense arc that extends from east-southeast of

Heber-Overgaard, through the Pinedale region, and through the Show Low area on the southeast. Much of this area has been fairly intensively surveyed. Although the inventory records indicate fairly low site density to the north of the center of the cluster, survey has been far less intensive. Still, the surveys completed in the area suggest low site density. To the west, there has also been a notable amount of survey and corresponding low site density (see Figure 4.12). Survey coverage has been more extensive within the Mogollon Rim cluster, and a much higher proportion of it has been intensively surveyed than the Chevelon/Chavez area along the northeastern boundary of the study area.

The Chevelon/Chavez cluster (see Figure 4.12) is located north and west of these low-density zones. The distributions of small sites in this cluster exhibit several distinct breaks. Figure 4.12 suggests these could simply be attributable to survey coverage.

While this is true, the sporadic nature of the survey coverage itself is also attributable to

131

Figure 4.12. Locations of the Mogollon Rim and Chevelon/Chavez site clusters within the study area, showing previous survey 131 in the area.

132 the intervening canyons that dissect the terrain. Also, the areas included within this cluster are on the respective ends of the ASNF and CNF administrative units and largely outside of any commercial timber stands. Furthermore, many sites in the area were recorded by CARP and corresponding survey areas have not been thoroughly documented.

A tremendous amount of effort has been expended during the past several decades to document archaeological sites throughout the study area. Intensity of work has been variable, and different groups of researchers have focused on specific areas, as one might expect. Synthesizing the geographical information about small architectural sites from two different areas has identified some interesting trends in the data. Clearly, much work remains to survey new areas and consolidate more information about small sites in a more systematic manner. The associations of these site locations with some environmental and climatic issues are explored in further detail in the following chapter.

133

CHAPTER 5.

SITUATING SMALL SITES WITHIN THE ENVIRONMENT

Identifying small architectural sites and their associated archaeological landscapes highlights important issues regarding site typologies and the inferred significance of different types of sites. Clearly apparent clusters of sites have been identified, as well as areas of extremely low site density even though those areas have been surveyed extensively. This sets the stage for the following chapter and explores the relationships among these small sites in greater detail. Before approaching those issues, however, this chapter situates small architectural sites within the environment of the Mogollon Plateau region by exploring site relationships with representational models of environmental characteristics. Predictive models of various types are also discussed, as well as a deductive approach to modeling agricultural suitability in the study area.

Four fundamental sources of geographic data are used to situate small architectural sites within the environment. The archaeological site data and survey information reviewed in the previous chapter provide sample and reference points associated with the archaeological landscape. A digital elevation model (DEM) and the

Terrestrial Ecosystem Survey (TES) introduced earlier support the exploration of a wide array of derived geodatasets. After introducing and analyzing the specific data sources, including some of their potential and limitations, I explore their relationships with small sites in a comparative approach. These findings may be used to assess some of the hypotheses regarding the significance of small architectural sites discussed in the

134 previous chapters. In some cases, subsets of archaeological sites are examined more carefully for comparative purposes and to assess the relationships among site classes and whether or not perceived differences relate to particular environmental variables. This also provides background for issues explored in more detail in the next chapter.

In the following sections, I present geographic data acquired from USDA Forest

Service sources, as well as other federal, state, and private entities. Visualization and analysis was completed using ArcGIS v10.0, and all data is projected in Universal

Transverse Mercator projections using the North American Datum of 1983. Despite associated sources of error and abstraction, these geographic data provide an important foundation for a number of derived surfaces.

DEM and Derived Terrain Characteristics

The USGS National Elevation Data (NED) provide 1 arc second (approximately

30 m resolution) elevation data for the entire study area (see Figure 4.2). Although accessed on various occasions, the information presented here was downloaded on

August 28, 2012. Data representing 1/3 arc second (approximately 10 m resolution) measurements are also available for the study area. However, given the landscape-scale analysis objectives at hand, the 1 arc second data are acceptable, and previous studies have shown that increased spatial resolution does not necessary improve the significance or meaning of landscape metrics associated with archaeological sites (Zimmerman and

Artz 2006). The NED DEM benefit from

135

“efficient processing methods ... developed to filter production artifacts in the

existing data, convert to the NAD83 datum, edge-match, and fill slivers of

missing data at quadrangle seams. One of the effects of the NED processing steps

is a much-improved base of elevation data for calculating slope and hydrologic

derivatives” (NED metadata 2012).

A DEM is essentially a special type of raster data that divide an area into small parcels (in this case, cells measuring approximately 30 m by 30 m) of the same elevation.

Although subject to issues of generalization, scale, and locational accuracy, the DEM can provide important estimations of elevation, slope, aspect, and other terrain characteristics.

As a general rule of thumb, lower resolution DEM are more likely to obscure variation in these terrain measurements, since the underlying elevation values for a given area become averaged over an increasingly large area. In some cases where specific processes are occurring in very specific settings, this could be problematic. In the case of the issues being explored in this dissertation, the archaeological sites being studied were undoubtedly associated with behaviors and land use practices that must have extended well beyond the confines of a small architectural structure to include much of the surrounding area.

As discussed in the previous chapter, elevation has often been cited as a controlling factor in land use practices, particularly when agricultural functions are assumed. Also, previous researchers have noted the important influence of topography

136 upon the environment, including the orographic rain-shadow effect of the Rim (Kaldahl and Dean 1999) and the ecological gradient (USDA 1989). In this vein, it is important to note that vegetation is strongly correlated within elevation as well, particularly in regard to broad-scale communities, such as the broad belts of Ponderosa pine forest and piñon- juniper woodland that parallel much of the Mogollon Rim in the study area. It is worthwhile to consider the fact that elevation may serve as a proxy measure of vegetation within the study area, and elevation is not susceptible to the ongoing changes in community composition that may have been brought about by environmental change during the past millennia, as well as human influences like grazing, timber harvesting, and large-scale range management projects, such as chaining.

Elevation indubitably played important roles in Native American use of the study area during the Pueblo periods. Given the region’s propensity for snowfall and short growing seasons at higher elevations it is not surprising that it could serve as a limiting factor for agricultural pursuits. Ethnographic studies among modern Native American groups who used the area historically also note a wide variety of significant resources that vary with elevation. Although high elevation zones may not be suitable for agricultural pursuits, they provide access to a host of different plants and animals. Ongoing collection of spruce boughs (Senior 2004) demonstrates influences of elevation that may not conform with modern notions of land use.

Seemingly minor variations in topography were significant beyond the hunting and gathering opportunities they provided. Minor points of topographic prominence have served as shrines and other types of landscape markers. In many cases these peaks are

137 isolated form larger mountain ranges and may not be particularly prominent, as demonstrated by ethnographic accounts and archaeological examples (Figure 5.1).

Although prominent mountain peaks and other comparably rare features attract the most attention, less obvious topographic variations served as important places as well, particularly in the broad, elevated plains of the Mogollon Plateau.

Figure 5.1. Red Knoll, a minor point of topographic prominence among an area of many small architectural sites, as well as the interface of juniper woodland and grassland.

In order to assess variation elevation among small sites in the ASNF portion of the study area, measurements were derived for each the subsets of field houses, single room and jacal structures, and room blocks. Two separate samples of 2,000 random

138 points were also included in the analysis. The first set was selected from the entire study area, while second set was selected only from those areas that had been previously surveyed. The basic descriptive statistics presented in Table 5.1 demonstrate that the mean, maximum, and standard deviation values are much higher than those of the selected archaeological sites.

Table 5.1. Descriptive statistics for elevation of random points, random within surveyed areas, field houses, other small sites, and room blocks in the ASNF study area.

Sample N Minimum Maximum Mean Std. Deviation Random 2000 1758.7 2637.9 2069.698 146.6188 Random Survey 2000 1765.0 2630.0 2142.835 148.6085 Fieldhouse 689 1773.3 2304.9 2002.896 70.9955 Single room 414 1755.5 2296.6 1989.865 63.4632 Room block 517 1748.9 2277.9 2010.116 60.7442

The range of values is much more restricted among the site samples, and the mean values vary by just slightly more than 20 m. In order to explore these distributions in greater detail, histograms of each subset are presented in Figures 5.2 through 5.6. The subset of random points exhibits fairly normal distribution, while the random points selected from previously surveyed areas exhibit a bimodal distribution, with peaks around

2,000 and 2,300 meters. The first peak generally corresponds with the first sample of random points; the high number of cases around 2,300 m in the survey sample reflects the focus of many previous survey projects upon higher elevation areas where timber projects have been completed during the past few decades.

139

Figure 5.2. Histogram of elevation values for 2,000 random points located within the ASNF study area.

Figure 5.3. Histogram of elevation values for 2,000 random points located within previously surveyed areas in the ASNF study area.

140

Figure 5.4. Histogram of elevation values for single room and jacal sites in the ASNF portion of the study area.

Figure 5.5. Histogram of elevation values for field house sites in the ASNF study area.

141

Figure 5.6. Histogram of elevation values for room block sites in the ASNF study area.

The field house and single room/jacal samples exhibit much different distributions than the random samples, and it is certainly not attributable to the smaller sample size alone. The bimodal distribution of single room/jacal elevations shows peaks around

1,940 and 2,010 m. Field houses exhibit a normal distribution, but the vast majority of the sites occur between 1,800 and 2,100 m. There is also a notable tail of high values, with some sites clustering between 2,100 and 2,300 m. The elevations of room block sites are also normally distributed but skew slightly toward lower values.

142

Figure 5.7 Box plot comparing (bottom to top) elevations of random sample points within the AS Study area, random sample points within previously surveyed areas, field houses, single rooms and jacal structures, and larger room block sites.

The relationships among these site subsets and random samples are documented in further detail in Figure 5.7. The differences in the ranges among the site samples and the randomly selected points are quite clear. Also, the similarities among the site sub- samples, particularly the much more restricted ranges of values, are apparent. However, paired sample t-Tests of the distributions indicate no probability that they share the same central tendencies at the 95% confidence interval. The same results were confirmed with paired sample Kolmogorov-Smirnov tests of the elevation distributions among the sample

143 groups. This demonstrates that small architectural sites were clearly not randomly distributed throughout the study area. Also, it appears that single rooms and jacal structures, field house structures, and larger room blocks all occupied a similar range of elevation throughout the landscape, but there were still significant differences. The large number of outliers may be attributable for some of these results, but the fairly large sample sizes are also significant.

However, the room block sites exhibit the tightest distribution but also have many outliers, and they also appear to be situated at slightly higher elevations than single rooms and field houses. These smaller sites may have been situated at more diverse locations to take advantage of different environmental settings and micro-climates influenced by local topography as well the Rim’s rain-shadow. Because of the ecological gradient present within the study area and its strong correlation with elevation, it is reasonable to state that these sites occupied generally similar environmental settings. However, there may still be important local variations that are obscured by grouping sites in this fashion.

Derived Terrain and Hydrological Surfaces

Site elevations derived from the DEM highlight some important trends in the study area. Terrain analysis allows compound topographic and hydrologic characteristics to be derived for the study area. Hydrologic, geomorphic, and biological processes are all affected by the topographic characteristics of the landscape (Moore et al. 1991). The complexities of terrain have a strong influence on the processes that hydrologists and

144 watershed managers seek to study, so extensive effort has been directed toward accurately incorporating three-dimensional terrain data in analysis. Interpreting terrain elements and hydrography from DEMs can be accomplished with a wide variety of methods (Moore et al. 1991; Singh and Fiorentino 1996) to derive an array of primary and secondary attributes. Speight (1974) identified more than 20 attributes that aid landform analysis. The most commonly interpreted primary topographic attributes calculated in GIS include slope, aspect, catchment area, flow path, slope profile curvature, plan (contour) curvature, and elevation (Moore et al. 1991:15). Measures of convexity, concavity, and flatness can be used to characterize landforms across vast regions with digital terrain data (Blaszczynski 1997), and this landform characterization can be used to map and analyze a variety of ecosystem processes at a variety of scales.

This can also be a useful tool for morphostratigraphic landscape analysis (Wells 2001).

Slope, slope curvature, aspect, and catchment area have all been used in varying combinations to estimate soil moisture. At the scale of the hill slope, slope curvature has also been demonstrated as a determinant of processes of erosion and deposition (Moore et al. 1991). Attempts have been made to derive properties like erosion potential, productivity, A and B horizon thickness, and water storage from some of these topographic characteristics.

A first step in accessing many of these derived surfaces is to compute slope values. Slope could be a useful attribute for assessing agricultural suitability. Extremely steep slopes would probably not have been selected unless extensive terraces or other water-control devices were used to modify the terrain. Slope surfaces can be reclassified

145 to convey the preference for relatively flat areas for dry farming, but many researchers have also noted that agricultural areas are often assumed to correspond with small ephemeral and intermittent drainages, in slightly more sloping areas near habitation sites.

Among the small structure sites, 93.8% are located at areas of slope of less than

10 degrees. Values less than 6 account for 81.3% of the total sample, and the mean value is 3.6. The mean value for the entire study area is 5.5, a significantly higher value, and the slope values for small architectural sites are clearly skewed towards lower values.

Extracting point values provide a useful estimate, but the tendency for sites in the study area to be located near areas of higher slope values warrants further consideration.

Figure 5.9. Terrain aspect in the ASNF study area displayed as a ratio of the total area in 45 degree intervals; the distribution exhibits the area’s north-sloping terrain.

146

Comparing aspect values across the landscape and among different site types suggest that it did not play a significant role on a landscape-scale in terms of the decision- making process that contributed to where field house sites were located (Figure 5.8).

Similar results were found for other classes of architectural sites in the study area, and previous researchers have noted no significance in aspect values among archaeological sites in the area (Barnes 2004; EcoPlan 2012; North et al. 2003). Intuitively, aspect seems likely to have been an important influence on site location, but no evidence of such a pattern has been identified to date. It may prove more useful in future studies of more restricted geographic areas, including more localized and site-specific settings. Also, future insights could prove its utility in combination with a function of other landscape characteristics. Although the locations of specific structures may not reveal a strong association with aspect values, nearby agricultural areas may and warrant future consideration.

Derived Hydrography

A unifying characteristic among most of the spatial characteristics of the archaeological landscapes of the American Southwest and much of the rest of the world is the tendency of archaeological sites to be located in close proximity to a source of water. It seems nearly universal that other resource specialists, including those who have participated in the Forest Service’s para-archaeology program, are often quick to point out this association.

147

Watersheds and their sub-basins delineated from DEM are typically the fundamental unit for hydrologic analysis. Flow accumulation surfaces calculated for these watersheds essentially represent the total number of cells in a surface that route water to each target cell in a surface. These surfaces are an approximation of the total area that contributes to the runoff that could theoretically be delivered to each cell. Since successful agriculture would require the selection of locations that receive suitable runoff from adjacent slopes, reclassified flow accumulation surfaces can represent this variability across a surface. Reclassified flow accumulation and derived terrain surfaces such as slope can be combined to highlight relatively flat areas that potentially receive more runoff than other locations.

Most hydrological analysis recognizes the watershed or a subdivision of a watershed as the fundamental unit. Archaeologists must find a way to relate these results to human behavior. In the case of agricultural land, typical sites (habitations and temporary camps) may or may not be situated directly within agricultural land

(agricultural features such as terraces and check dams may occur, but are relatively rare in most of the study area; see Table 3.1). However, ethnographic examples suggest the groups who occupied the study area probably did not live an extreme distance from their agricultural plots. Catchment areas of varying shapes (circular being the most commonsensical) and sizes serve as a useful method for characterizing the relationship between varying indices of agricultural suitability and the presence/absence of archaeological sites and their associated characteristics.

Steps similar to those outlined above were used to model the hydrography of the

148 study area. In the ASNF, streams have been differentiated as perennial streams, intermittent drainages, and ephemeral washes (Figures 5.11 and 5.12). Springs have also been mapped, although their correspondence with spring locations that would have existed a millennium ago are unknown. Drainages and spring locations also offer key distinctions between the two site clusters. Comparing site distributions with other representations of surface water is futile in most of the area, since virtually all of it on this portion of the ASNF consists of fairly recently constructed reservoirs.

The most obvious difference is presented by the association with springs. The

Mogollon Rim cluster is located in the vicinity and immediately downstream from a band of springs located to the south. By contrast, there are extremely few springs documented in the vicinity of the Chevelon/Chavez cluster. Despite the lack of perennial streams and springs in this area, surface water is available sporadically along intermittent streams and the primary drainages. Water is also available sporadically in some scattered small, shaded catchments within these drainages and canyons, where local conditions are favor water’s persistence.

One characteristic shared by the two clusters is a lack of apparent association with perennial streams. Although they are present in the Chevelon area, the vast majority of the sites are located away from these watercourses, and there appears to be a strong association with ephemeral drainages. Among the Mogollon Rim site cluster, there are few perennial streams. Show Low Creek appears to the east of this cluster; although few sites are included site samples, this is largely attributable to the fact that much of the

149

Figure 5.11. Archaeological sites with surface architecture in the Chevelon/Chavez cluster and their association with ephemeral, intermittent, and perennial streams and springs.

150

Figure 5.12. Archaeological sites with surface architecture in the Mogollon Rim cluster and their association with ephemeral, intermittent, and perennial streams and springs.

151 stream is located outside of the ASNF and is privately owned. Sites in the area are present but have not been recorded in such a systematic manner. Regardless, the pattern holds that relatively few of the sites are located along perennial streams. Unlike the

Chevelon/Chavez cluster, those in the Mogollon Rim cluster exhibit a stronger association with intermittent rather than ephemeral drainages. These intermittent drainages receive more water from springs and snowmelt than those to the northwest, a point also documented in the TES and discussed in more detail below.

Terrestrial Ecosystems Survey

Although primarily intended as a planning tool for natural resource management and engineering purposes, the resulting TES information is also pertinent to issues raised in previous chapters regarding the distributions of archaeological sites. The Sitgreaves portion of the study area includes 55 of these units (Table 5.2). Although several of the units are classified as complex (i.e., they can be broken down into further subunits based on terrain or other characteristics), they have not been mapped to that level of detail. I focus on the larger mapping units in the following analysis, but return to the subunits at a later point to further explore their relationships with small sites.

Several TES units exhibit surprisingly high proportions of the total sites in the study area when compared to the total proportion of the study area (Figure 5.13). TES

152 map units 41, 51, 61, 178, 181, 182, and 191 include the vast majority of the sites (Figure

5.14 and 5.15); in many cases they include a much larger proportion of the site sample

Table 5.2. Terrestrial Ecosystem Survey units within the ASNF portion of the study area and corresponding site associations and proportion surveyed.

TES Site Ratio Total Total Acres Ratio Total Total TES Unit Ratio TES Unit Count Sites Study Area Acres Acres Surveyed Unit Surveyed 4 1 0.000 98.56 0.000 85.87 0.871 16 5 0.001 1444.14 0.002 1101.42 0.763 41 166 0.029 11946.37 0.014 1170.08 0.098 43 76 0.013 15826.85 0.018 3830.40 0.242 44 3 0.001 1785.92 0.002 46.51 0.026 51 1430 0.247 80642.01 0.093 25337.92 0.314 52 265 0.046 29561.30 0.034 4140.09 0.140 53 444 0.077 73327.13 0.084 23409.66 0.319 54 156 0.027 25003.82 0.029 6548.55 0.262 55 24 0.004 6770.14 0.008 287.87 0.043 58 33 0.006 5336.34 0.006 1434.40 0.269 61 167 0.029 6477.83 0.007 3155.94 0.487 178 274 0.047 19287.17 0.022 13114.97 0.680 179 45 0.008 28676.46 0.033 9877.24 0.344 181 266 0.046 31027.82 0.036 24976.08 0.805 182 405 0.070 38573.24 0.044 24273.44 0.629 183 190 0.033 36033.06 0.041 15092.89 0.419 186 165 0.029 16518.41 0.019 3387.09 0.205 187 208 0.036 18289.16 0.021 5957.51 0.326 189 19 0.003 18067.33 0.021 6034.48 0.334 191 658 0.114 33099.93 0.038 26926.73 0.813 192 152 0.026 49219.78 0.057 39394.83 0.800 193 79 0.014 30624.54 0.035 27794.77 0.908 196 9 0.002 33255.67 0.038 26567.54 0.799 197 8 0.001 30814.45 0.035 28353.35 0.920 198 64 0.011 6112.44 0.007 2601.11 0.426 199 4 0.001 18200.85 0.021 13191.42 0.725 201 3 0.001 7746.32 0.009 7074.31 0.913 202 8 0.001 12433.80 0.014 11759.77 0.946 203 1 0.000 6765.81 0.008 6554.29 0.969 206 4 0.001 12988.05 0.015 6608.41 0.509

153

207 4 0.001 2179.42 0.003 2155.29 0.989 208 10 0.002 1042.72 0.001 910.88 0.874 502 11 0.002 7147.03 0.008 2846.72 0.398 503 14 0.002 7779.63 0.009 2224.33 0.286 504 62 0.011 18492.62 0.021 4274.80 0.231 505 17 0.003 2004.90 0.002 1558.99 0.778 515 38 0.007 8655.63 0.010 4677.89 0.540 516 1 0.000 89.24 0.000 6.18 0.069 523 72 0.012 11551.96 0.013 7014.48 0.607 531 8 0.001 3357.97 0.004 657.04 0.196 532 69 0.012 31983.00 0.037 30999.89 0.969 534 1 0.000 1259.28 0.001 1249.35 0.992 536 11 0.002 3602.88 0.004 3399.58 0.944 537 9 0.002 11143.66 0.013 11130.53 0.999 538 1 0.000 3036.13 0.003 3004.07 0.989 540 7 0.001 992.20 0.001 306.31 0.309 561 2 0.000 1569.02 0.002 1313.27 0.837 567 0 0.000 3952.08 0.005 2964.07 0.750 575 3 0.001 310.14 0.000 244.91 0.790 580 7 0.001 8877.16 0.010 1775.44 0.200 591 15 0.003 2054.02 0.002 1440.37 0.701 592 26 0.004 2855.59 0.003 957.89 0.335 624 2 0.000 367.43 0.000 367.29 1.000 997 1 0.000 23.11 0.000 17.54 0.759 998 2 0.000 1027.30 0.001 544.22 0.530 999 63 0.011 27718.32 0.032 969.49 0.035

than the study area. Relevant characteristics of these units are listed in Table 5.3. There is a clear distinction among the seven different units, with the first three consisting of soils derived from sandstone and limestone parent materials, ephemeral streams, limited canopy cover, and rare snow cover. The remaining four units primarily consist of deep, gravelly soils derive from alluvium, in areas with closed canopies that are prone to snow cover but retain subsurface moisture from snowmelt.

154

0.300

0.250

0.200

0.150

0.100

0.050

0.000

154 Figure 5.13. Comparison of ratios of total sites per TES (blue) and total amount of study area per TES unit (red).

155

Although probably they have changed to a certain extent during the past several centuries, the dominant vegetation taxa within each unit also exhibit interesting differences (USDA 1989). The three units associated with the Chevelon/Chavez cluster all include one-seed juniper. Ponderosa pine is the dominant taxa in the Mogollon Rim cluster, and alligator juniper is either the second or third most dominant taxa (except for

TES unit 191). Perhaps most interestingly in the Mogollon Rim cluster, all include

Quercus gambelii; although not dominant, oak trees form groves, reproducing from acorns as sprouts from the root structure, even following wildfire. Although not well- represented in archaeobotanical assemblages of the Southwest (Huckell 1999:480), acorns have been an important food source for humans in general and have been found in archaeological contexts (including Grasshopper and Tonto National Monument).

Surprisingly, piñon was also remarkably rarely recovered in the assemblages from

SCARP excavations (Huckell 1999:472-473), although they have regularly been documented in assemblages from other areas. Despite the paucity of archaeological evidence, which very well could be attributable to processing behaviors, acorns and piñon are still valued by many Native American groups in the area, and they are also an important source of forage for a variety of different game animals. The zonal patterning of vegetation in the study, influenced by the orographic effect of the Rim, probably has not changed significantly in the past millennium, although the boundaries may have shifted slightly (Dean 1996) and composition has no doubt been influenced by historical land use practices.

156

Table 5.3. Narrative topographic, soil, snow cover, and drainage descriptions for the TES units that include the most archaeological sites in the ASNF study area (USDA 1989).

TES Topography Vegetation Soils Snow Streams Unit Nearly level to Low canopy cover Derived from limestone Snow cover is rare Ephemeral strongly sloping and sandstone parent 41 elevated plains materials from the Kaibab formation. Intricate pattern of Low re-vegetation Shallow and rocky soils Snow rarely Ephemeral two subunits, potential from limestone and accumulates 51 elevated plains sandstone parent with some slope. materials, some areas with excessive limestone Complex pattern Woodland Shallow and gravelly Snow rarely Dendritic on elevated plains productivity is high sandy loams derived accumulates. ephemeral and plain slopes. (transition from from sandstone parent 61 “woodland to materials, with commercial timber” subsurface clay is nearby). horizons. Single unit on Transition between Moderately deep, very Patchy snow cover Dendritic level to sloping woodland gravelly sandy loams December through ephemeral plains. commercial timber. derived from alluvium March. Heavy and sandstone. subsurface horizons 178 remain wet through early spring from snowmelt and following summer rains.

156

157

Intricate pattern on Ponderosa Pine Mixed deep gravelly Continuous snow cover Ephemeral level to sloping canopy sandy loams, residuum from November to and dendritic plains. derived from deep April, horizons remain intermittent 181 alluvium, subsurface wet through early horizons contain heavy spring from snowmelt clays. and following summer rains. Intricate pattern of Ponderosa Pine Deep and very cobbly Continuous snow cover Dendritic hills and valleys canopy sandy loams, some from November ephemeral and and moderately subsurface horizons through April, soil intermittent steep hills and contain heavy clay. remains wet through 182 slopes with early spring from substantial rock snowmelt but drains cover. more quickly following summer rains. Intricate pattern of Ponderosa Pine Deep sandy loam Continuous snow cover Dendritic moderately canopy derived from old from November ephemeral and sloping elevated alluvium through April, soils intermittent and valley plains. remain wet through early spring from 191 snowmelt and following summer rains and they receive additional moisture from runoff.

157

158

Figure 5.14. The seven TES units in the ASNF study area with the most archaeological sites. 158

159

200 Fieldhouse 180 SingleRoom/Jacal 160 Roomblocks5plusTES 140

120

100

Figure 5.15. Field houses, single room and jacal structures, and room blocks estimated to have five or more rooms by TES unit in the ASNF portion of the study area.

159

160

Although capable of providing a wide array of insights and possible correlations among locational characteristics and ecological conditions, the TES data are clearly susceptible to the problem of “ecological fallacy,” due to the ability to modify areal units and aggregate populations in a manner that may obscure more fine-grained relationships

(Harris 2006). These relationships are not necessarily erroneous, but they require a certain degree of caution.

Site Location Probability Models

F. Plog (1981b) presented a deterministic model of archaeological site probability derived from large-scale sample survey of the eastern portion of the study area, just slightly encroaching upon the eastern margins of the CNF’s Mogollon Rim Ranger

District. Plog’s SYMAP model also extends farther to the east than is shown in Figure

5.16, toward Springerville, but there is a fairly broad area of low site density that corresponds with the Springerville Volcanic Field and continuing east toward the headwaters of the Upper Little Colorado River, another site cluster worthy of mention but not considered in great detail in the present study. The intervening area of low site

161

Figure 5.16. Archaeological site density probability model digitized from Plog (1981b), resulting from sample surveys (green lines) completed in the Little Colorado Planning Unit.

162 density has been surveyed fairly extensively due to past timber production and ongoing vegetation management projects. Recent surveys associated with fuel management projects again confirmed this trend, noting associations with springs, apparent use as hunting grounds documented by isolated projectile point finds, and small artifact scatters apparently representing short-term camps associated with hunting and gathering. Shrines in the surrounding area were probably integrated in patterns of seasonal use, as much of the higher elevation terrain accumulates deep snow cover. Run-off from snowmelt is an important source for the intermittent streams emanating from the mountainous area.

F. Plog’s (1981b) SYMAP model was deterministically created from 439 survey transects used to derive a sample of known archaeological sites on a portion of the ASNF.

The procedure supported visualization of site density (Figure 5.16) and was classified into very high, high, moderate, and low (all areas not in the previous three categories).

Plog (1981b:35) was careful to point out that the boundaries were imprecise and suitable only for planning purposes, as they were based upon the sample survey. Figure 5.17 demonstrates some areas of particularly high site density eluded Plog’s initial sample survey of the study area, as he had warned. Much of the Mogollon Rim site cluster had not yet been identified, although many of the larger sites were known at the time.

Following the Rodeo-Chediski Wildfire during summer 2002, extensive archaeological surveys were completed in advance of several proposed timber salvage, hazard remediation, and other vegetation management projects (Haines et al. 2004). In addition to extensive surveys completed by the Forest Service following Rodeo-Chediski,

163

Figure 5.17. Plog’s (1981) SYMAP model and the Mogollon Rim and Chevelon/Chavez site clusters.

164

SWCA was contracted to survey extensive parts of the Forest (North et al. 2004). The fires provided unparalleled surface visibility, conditions that plagued some past survey projects. Dense accumulations of vegetation, especially pine duff, often led surveyors to caution that survey results were limited by visibility and there could be good potential for near-surface finds. Previously documented areas of varying site density were identified, noting few sites in the southwestern portion of the Sitgreaves NF where a number of surveys were completed previously for timber sales, as well as significantly higher site density in the vicinity of Linden, as documented by previous surveys. That particular wildfire event included extensive portions of the Mogollon Rim site cluster, but did not ravage the extensive piñon-juniper woodland belt as far northwest as the

Chevelon/Chavez region. Although there have been some notable recent fires in that area, surface visibility conditions in those ecotones are typically less limited by vegetation and other debris. Surface visibility is generally not as improved following surface fires as in Ponderosa pine forests.

Survey results from those projects, coupled with prior experiences of Forest personnel, including Bruce Donaldson, led to the creation of a sensitivity surface for archaeological sites to be used in future planning efforts. A particular concern has been to support more accurate contract specifications when hiring outside professionals to complete surveys, as well as to budget and plan more reliably for in-house survey projects and cultural resources compliance needs. Barnes (2004) documented the model, noting the importance of elevation, slope, and distance to piñon-juniper woodland and

165 drainages (Figure 5.18). A step-wise logistic regression equation was used to incorporate the values into an inductive site location probability model, which was reclassified into five categories. That model included all presumed known precolonial Native American sites but did not further stratify the sample.

Comparing the model with known distributions of sites with surface architecture reveals some interesting trends (Figure 5.19). Much of the northwestern portion of the area was characterized as having very high potential, including areas adjacent to known site clusters that have not yet been surveyed extensively. This is attributable to the preponderance of piñon-juniper woodland in the northern part of the Black Mesa district, as well as the preponderance of relatively flat and broad areas. The apparent divide among the two site clusters corresponds with a divide between areas of high and very high site probability. The Mogollon Rim cluster coincides with a band of high probability interspersed by moderate and low probability zones, although there are several patches or islands of high probability.

As seen in Table 5.4, single rooms and sites with jacal structures, field houses, and room blocks are distributed among the probability categories in the predictive model in fairly similar proportions. However, if the two site clusters are examined more closely, it is clear that the Chevelon/Chavez portion of the cluster on the ASNF is more strongly correlated with very high and high probability locations. The Mogollon Rim

166

Figure 5.18. Site location probability model developed by the ASNF following the Rodeo-Chediski fire. 166

167

Figure 5.19. Inductive probability model and the Mogollon Rim and Chevelon/Chavez site clusters.

168

Table 5.4. Single rooms, field houses, and room block sites by probability classes in the Sitgreaves inductive probability model.

Category Single % Single Field % Field Room % Room Room Room house House block Block Very 132 31.9% 204 29.6% 169 32.7% High High 229 55.3% 385 55.9% 287 55.5% Moderate 34 8.2% 60 8.7% 34 6.6% Low 19 4.6% 37 5.4% 25 4.8% Very Low 0 0.0% 3 0.4% 2 0.4%

Table 5.5. Single rooms, field houses, and room block sites by probability class within the Mogollon Rim site cluster in the Sitgreaves inductive probability model.

Category Single % Single Field % Field Room % Room Room Room house House block Block Very High 6 4.3% 39 11.5% 16 7.7% High 103 74.6% 257 75.8% 161 77.0% Moderate 21 15.2% 33 9.7% 25 12.0% Low 8 5.8% 9 2.7% 6 2.9% Very Low 0 0.0% 1 0.3% 1 0.5%

cluster (Table 5.5) includes only 6 of the 132 (4.5%) single room sites in very high probability locations, 39 of 204 field houses (19.1%), and just 16 of the 169 room blocks

(9.5%) located in very high probability locations.

The model can also be assessed by examining the locations of sites recorded after it was created. Within the ASNF portion of the study area, the site database identifies approximately 874 sites have been recorded since the model was created in 2004.

Among these sites, at least 111 only include historical components and do not warrant further consideration given the nature of the model. The distributions of the remaining

763 Native American sites generally support the model’s reliability, but suggest more

169 work is needed. One notable exception has been the extreme eastern end of the

Sitgreaves, along the northern portion of the Forest boundary, where numerous artifact scatters have been recorded as a result of surveys for various projects.

More recently, EcoPlan (2012), a private cultural resources consultant, was contracted to update the ASNF Cultural Resources Overview and Management Plan originally created by Plog (1981a, b). They used chi-square tests to evaluated associations of archaeological sites with a wide array of environmental variables derived from the TES and other sources. Unfortunately, their logistic regression probability model attempts to predict the likelihood of site locations through the ASNF, including the

Alpine, Clifton, and Springerville Ranger Districts to the east and southeast of the study area, immediately west of the Arizona-New Mexico state line. Their model uses five independent variables: positive correlations with elevation ranging between 6000-7000 feet, and gravelly loam soils (included in the previous TES discussion), and negative correlations with the presence of Douglas Fir (essentially mixed-conifer forests at high elevations) and cobbly loam soils (also found at high elevations).

Predictive models in general have suffered from a one-size-fits-all approach to

Native American sites, so probability values were extracted for the site subsets included in this study to assess their relationships. Future modeling efforts would benefit from exploring smaller subsets of sites. Land management agencies could tend to judge the benefits of predictive models by their ability to accurately identify a specific number of archaeological sites in a given survey area in order to support more accurate contract

170 specifications when hiring private consultants, as well as estimating the expense of completing projects on an in-house basis. This motivation probably is not the most productive route for assessing predictive models, and certainly not for understanding the behaviors manifested in the archaeological record.

Models for Assessing Agricultural Suitability

Data derived the TES and DEM were examined in an attempt to explore different aspects of agricultural suitability, which has been proposed as a primary function of many small architectural sites in the study area. The previous chapter reviewed some ecological characteristics that might favor agriculture, including a sufficient number of frost-free days, adequate precipitation, surface flow accumulation and proximity to drainages, and soil suitability (which, as discussed previously, have not been mapped in a manner that represents their intricacies in the vast majority of the study area). Also, ranges in the elevation of site locations and preferences toward areas of gentle slopes

(Figure 5.20) have been identified as common characteristics of many of the sites assumed to have agricultural functions.

171

Figure 5.20. An example of elevated plains in piñon-juniper woodland typical of the “broad swales” regularly identified as locations suitable for agriculture by survey crews working in the Chevelon/Chavez site cluster.

An important component of assessing agricultural suitability is further consideration of data regarding precipitation and length of the growing season.

Fortunately, these factors were estimated as part of the TES (Table 5.6, Figures 5.21 and

5.22). Another aspect of this information is average annual snowfall (Figure 5.23).

Snowfall deserves additional consideration; although mean annual snowfall is not dramatically different in the area of the Mogollon Rim site cluster, snow accumulation

172 and the persistence of snow cover certainly varied. In the more open piñon-juniper woodland and grassland with less or even no canopy cover, snow melts far more quickly than Ponderosa pine communities. Although canopy cover among within the Ponderosa forest in the past was likely far less than the current results of timber harvesting and fire suppression, it still would have provided shade and maintained snow cover for longer intervals. This would have provided additional sources of water, but also would have posed discomforts in some areas that probably precluded winter occupation in some places and made lower elevation and lower precipitation areas more favorable, in some regards.

Table 5.6. Average rainfall, snowfall, and frost-free days by TES unit in the ASNF study area.

TES Unit Rainfall (cm) Avg rainfall cm) Snowfall (cm) Frost-free days 4 62-74 68 150 90 16 50-62 56 120 100 41 30-42 36 40 160 43 30-42 36 40 160 44 30-42 36 40 160 51 34-46 40 80 150 52 34-46 40 80 150 53 34-46 40 80 150 54 34-46 40 80 150 55 28-46 37 80 150 58 34-46 40 80 150 61 40-52 56 90 140 178 44-56 50 100 130 179 50-62 56 120 100 181 50-62 56 120 120 182 50-62 56 120 100 183 50-62 56 120 100

173

186 46-56 51 100 130 187 46-56 51 100 130 189 40-55 47 110 115 191 58-70 64 120 100 192 58-70 64 120 100 193 58-70 64 120 100 196 58-70 64 120 100 197 58-70 64 120 100 198 50-62 56 Null 100 199 58-70 64 120 100 201 69-81 75 150 90 202 69-81 75 150 90 203 69-81 75 150 90 206 58-81 66 140 100 207 69-81 75 140 90 208 50-74 62 150 90 502 34-46 40 80 150 503 34-46 40 80 150 504 34-46 40 80 150 505 44-56 50 100 130 515 44-56 50 100 130 516 34-46 40 80 150 523 44-56 50 100 130 531 34-46 40 80 150 532 50-62 56 120 100 534 50-62 56 120 100 536 50-62 56 120 100 537 50-62 56 120 120 538 50-62 56 120 100 540 34-46 40 80 150 561 58-70 64 140 100 567 58-70 64 140 100 575 42-54 48 20 170 580 34-46 40 80 150 591 50-62 56 120 120 592 44-56 50 100 130 624 44-56 50 60 150

174

Figure 5.21. Mean annual rainfall (cm) derived from TES unit ranges in the ASNF study area.

175

Figure 5.22. Site clusters and mean annual frost free days derived from TES unit ranges in the ASNF study area. 175

176

Figure 5.23. Mean annual snowfall by TES Unit in the ASNF portion of the study area and the Mogollon Rim 176

and Chevelon/Chavez site clusters.

177

The eastern cluster is situated within an area that is clearly pushing the limit for a sufficient growing season length for maize agriculture, as the cluster is primarily within areas of less than 130 frost-free days. The portion of the northwestern area where most of the small architectural sites are located features longer growing seasons. Throughout the ASNF portion of the study area outside of the Mogollon Rim site cluster there is consistently low site-density in areas with less than 120 annual frost-free days. This presents clear evidence of two distinct adaptations represented in this part of the study area. This coincides with changes in other archaeological evidence noted before along the west-east axis just north of the Mogollon Rim.

Doolittle cautions that “The association of environments and sites is a convenient, yet tenuous way of attempting to understand landscapes of cultivation” (2000:165).

Despite these potential shortcomings, I attempt to explore these potential associations more carefully. In the following paragraphs, the previously mentioned characteristics potentially associated with agricultural success are explored in an attempt to assess their potential significance as models of agricultural suitability. Unfortunately, examples of agricultural implements, such as hoes, found in association with presumed agricultural areas have been relatively rare finds (Doolittle 2000:168). This is somewhat problematic, since the association of agricultural implements with arable land is more convincing evidence of agriculture than site association alone. However, their presence is also problematic, as they could be curated or discarded in other locations, or used for other purposes, such as construction of domestic features. Flaked implements could also be

178 mistaken for tools associated with wood procurement, and ground stone implements could serve other practical or even ceremonial functions, as suggested by Rohn (1971).

Unfortunately, review of site records confirms that documentation of surface assemblages of flaked stone artifacts has been inconsistent. Formal tools like projectile points are regularly noted (and often collected), but informal tools are usually mentioned in passing, at best. Agricultural areas have typically been inferred from their proximity to known sites. In some cases, low density scatters of ground stone artifacts have also been noted in association with possible agricultural areas.

Specific soil types have long been suspected to be more favorable than others.

Deep soils derived from sandstone parent material are often speculated as being important for dry land farming, as they are more likely to retain moisture for a longer period and to a greater depth (Figure 5.24). The most common soils associated with small sites in the area are typic eutroboralfs, typically consisting of deep sandy loams with varying but relatively high degrees of gravel and cobbles, and lithic ustochrepts, typically consisting of loams of varying depths and fairly high amounts of gravel of varying sizes (USDA 1989). Although North et al. (2003) found no significant associations among site locations and geology or soil type, their review was limited to a portion of the area burned in the Rodeo-Chediski fire, and subsequent planning documents (EcoPlan 2012) have identified possible associations with large geological units across the ASNF, including some that largely coincide with the large site clusters.

179

Figure 5.24. Examples of soils derived from sandstone (A) and limestone (B) parent materials. Many of these soils are intermixed within the TES units and cannot be easily separated at the scale used to create the survey, although variation is readily apparent on the ground.

180

Several underlying assumptions must form the basis of a model of agricultural suitability in the study area. First, although environmental conditions including temperature and rainfall vary throughout and within the study region in any given year, a static model must assume these variations occurred throughout time in such a manner that decision-making processes that shaped the long-term use of the study area were the primary influencing factors on past settlement and subsistence systems. Mounting evidence suggests the area was inhabited intensively for a period of at least several centuries, indicating recurrent use of the area for hunting, gathering, and agriculture.

Although adjustments to high and low-frequency environmental variation (Dean 1996,

2010) would have influenced behavior and biotic communities, the model requires one to assume a certain level of environmental uniformitarianism. Different species may respond in a variety of ways to different hydrologic periods, so presumed relationships with specific resources may vary considerably through time, but general patterns are evident for the Mogollon Rim region (Kaldahl and Dean 1999). Also, the orographic effects of the Rim are stable components of the environment whose influence upon environmental variability has not changed significantly. The prevailing atmospheric and topographic conditions that influence climate and environment in this area have been in place for substantially longer than archaeological sites with surface architecture (Dean

2010:328). To summarize the challenges presented to agriculturalists in the study area,

“the Rim presents a climatological gradient where the limiting factor of precipitation gives way to the limiting factor of temperature in crop production” (Kaldahl and Dean

181

1999:15), as confirmed by a thorough review of historical weather records. “This climatic setting created a spatially and temporally stable ecological gradient that prehistoric mixed-economy farmers would certainly have appreciated and factored into their migratory and community-building decisions” (Kaldahl and Dean 1999:28). Their review of historical climate records confirmed (16-17):

“The strong inverse relationship between climate station elevation and frost-free

season is expressed by the linear equation:

f = 331.45 – 0.102(e)

where f = frost-free season in days,

and e = site elevation in meters”

The results of this model are depicted in Figure 5.25.

A similar but more complex equation was developed to predict annual precipitation (P), incorporating distance from the Mogollon Rim (d) to account for the orographic effects of the Mogollon Rim (information from weather stations below the

Rim was excluded), and elevation (e) in meters. Step-wise regression was used to evaluate the effects of distance from the Rim and elevation; (d) is expressed on the natural logarithmic scale to account for the non-linear relationship with total precipitation, resulting in the following equation (Kaldahl and Dean 1999:17):

P = 2.038 + 0.01025(e) – 1.893[ln(d)]

182

Figure 5.25. Length of the frost-free season expressed by the linear equation of Dean and Kaldahl (1999) derived from historical weather records.

183

The result is a predictive model of annual precipitation during the historical period

(Figure 5.26), created from the previously discussed DEM and a Euclidean distance surface created from a polyline feature representing the Rim (which generally coincides with most of the southern portion of the study area), using Spatial Analyst ArcGIS tools to complete the necessary math algebra.

184

Figure 5.26. Predictive model of annual precipitation, interpolated from (Dean and Kaldahl 1999:17).

185

This model generally conforms to the findings from the TES classification of precipitation, but there are some notable differences. First, the TES data were estimated as intervals, whereas this model produces a smoother surface of continuous ratio-scale data. Direct comparison of the values requires conversion from inches to cm. Although the TES data may be less precise because of how it was recorded and it does not account for the more continuous nature of precipitation, it does afford insights into the distinction between winter snowfall and summer rains (the dominant components of the region’s bimodal precipitation pattern). Snowfall values have not been modeled with the same precision, and historical weather station data do not account for variable snow accumulation and persistence as influenced by canopy cover.

Dry farming suitability could be modeled with a number of simple approaches that incorporate slope, elevation, average number of frost-free days, average annual precipitation, soil characteristics, and distances to ephemeral or intermittent drainages.

Understanding how suitability is distributed across the landscape can benefit archaeological interpretation of past human behavior and provide insights for future efforts to document and investigate the archaeological record. However, it is clear that soil and geological data are not suitable for such interpretation at this point beyond the scale of noting that high-elevation and volcanic deposits are less likely to include sites.

Soil variation has not been systematically mapped throughout the study area at a level of detail that would allow clear delineation of preferential types, particularly in areas where sandstone and limestone deposits are intricately mingled, such as the Chevelon region.

186

Additional factors that warrant future investigation include incident solar radiation and more complex slope and terrain measurements. These appear to be issues that would be more influential at a local scale rather than the landscape scale approach used here. As such, these localized topographic influences upon cold air drainage, soil moisture retention, and other micro-environmental characteristics warrant more careful attention in future site-specific studies.

The ways these varying factors and weighted and combined will strongly influence the output. There is no clear indication of an indubitable method to accurately integrate these factors. Different strategies were probably used within the main site clusters, and some important differences have been shown in regards to the ecosystem characteristics. The relationships between slope values and drainage patterns require further consideration. Larger canyons and intermittent streams offered different horticultural opportunities, but higher energy environments. Understanding use of these areas is an important avenue of future research, and many sites associated with these areas have indications of extremely lengthy occupations.

One rendition of this model is presented in Figure 5.27. This is one of the possible representations of the challenges presented by climactic factors, reflecting the intersections of the climatic gradients of precipitation and frost-free days, and it was created using the following tasks:

187

Figure 5.27. Results of agricultural risk model combining interpolated surfaces of annual precipitation and frost-free days. Blue zone is susceptible to short growing seasons, green zone appears to be highest potential growing season and precipitation,yellow is sufficient growing season low precipitation, and red is extremely low precipitation (but low risk of 187 frost).

188

Task 1. Reclassify slope values greater than 20 degrees to 0, < 20 degrees

between 0 and 1.

Task 2. Reclassify annual precipitation surface to values between 0 and 1

Task 3. Reclassify annual frost-free day values less than 120 to 0, values >120

between 0 and 1.

Task 4. Combine surfaces

Task 5. Reclassify weighted suitability values to assess variability in agricultural

risk.

Task 6. Examine results of distribution of suitable land across the landscape and

compare with archaeological site locations.

The results were reclassified to reflect four main zones: areas with high annual precipitation but extremely short growing seasons, moderate precipitation with sufficient but potentially short growing seasons, limited precipitation but sufficient growing seasons, and very limited precipitation but long growing seasons. These categories reflect abstractions of the risks that mixed-economy horticulturalists would have perceived and adapted to on the Mogollon Plateau.

Results

189

The Mogollon Rim site cluster is located outside the areas of highest agricultural suitability when length of growing season is incorporated and weighted strongly. The

Chevelon/Chavez site cluster is located in an area where the threat of frost is less significant, but precipitation is a far more limiting factor. The clustering of the sites in the study area has been suggestive of two somewhat distinct archaeological traditions.

This investigation of conditions suitable for horticultural pursuits also suggests the two site cluster areas are associated with two fairly distinct adaptations. In the southeast, groups were afforded more runoff from snowmelt, which accumulated more deeply and lasted until later in the spring due to the canopy cover afforded by the Ponderosa pine forest. These conditions, as well as a relative preponderance of springs, supported more intermittent drainages, around which settlements tended to cluster. To the northwest, precipitation was far more limited. Spring snowmelt would have been far less reliable since snow does not tend to persist due to the more open canopy conditions, although it does persist in certain localized settings. Settlements appear to have targeted more elevated plain settings with ephemeral washes. Horticultural pursuits likely seized upon summer rains and crops would have been less threatened by the earliest frosts of the fall due to increased distance from the Mogollon Rim and its orographic effects. The canyons surrounding these elevated plains may have alleviated cold-air drainage concerns, but such localized phenomena require more detailed site-specific considerations. Experimentation with neighborhood relief values from DEM surfaces could reveal patterns in cold-air drainage, but any relationships at this point would be

190 purely speculative and require precise recording of temperature ranges throughout extended periods of time at particular geographic locations (e.g., Adams 1978; Lange

2001).

Discussion

The spatial break noted among the sites is probably the most obvious insight. It suggests a fairly substantial territorial break among different social groups, or at least groups using different socio-ecological adaptations. Plog’s (1981b) model based on a random sample of survey transects spanning the survey area identified a portion of the site distribution in the ASNF greatly underestimates the amount of sites in the area and the nature of high-density areas, but it used a simple deterministic method and was completed prior to the collection of a large amount of site data collected during survey projects. The inductive probability model includes subsequent findings and identifies high site-density areas in the Forest. It matches quite well with known site locations, and small architectural sites largely fall within the high probability locations. Sites found since the model generally indicate its reliability in the ASNF portion of the study area.

However, it includes all Native American sites and therefore does not account for variability among site types. Also, it was based upon a site database that contains numerous errors regarding site location, probable duplication of sites in some areas, and variable application of site definition standards. The tendency to create a one-size-fits-all

191 model has been rampant in archaeology, and it also often to arise out of necessity.

Breaking these relationships down farther by site type should be an important step in future efforts, and further stratifying study areas by environmental zones and other locational factors. A locally-adaptive inductive probability model (e.g. Carleton et al.

2012) could prove particularly useful given the great variation across the administrative boundaries of the National Forests.

The agricultural suitability model highlights the challenges subsistence-scale agriculturalists would have faced in the region by focusing upon the climatic gradient and on the influences of the Mogollon Rim on precipitation and length of growing season.

Horticulturalists would have faced the ongoing need to assess the risks of balancing the importance of the length of the growing season with the amount of available precipitation. The model highlights the fact that some portions of the Mogollon Rim site cluster were located in areas with marginal growing season lengths, on average, but greater available precipitation. To the northwest, farmers would have faced the challenge of ensuring sufficient precipitation for their crops. Accepting these factors as important influences on agricultural adaptations, stratification of the study area by these zones may aid future studies and strengthen models of agricultural suitability.

Adaptations to two distinct environments may indicate farming communities using genetically different varieties of crops, including corn. This could be an important avenue for future research on archaeobotanical remains recovered from excavation contexts. The Mogollon Plateau region and the easily accessible country to the south of

192 the Mogollon Rim offered a diverse array of common pool resources to diverse groups, each with their own agricultural and non-agricultural strategies. Buffer zones between these groups may have functioned to maintain their distinctiveness.

The potential associations between small site locations and geological units and soil maps available for the area are intriguing, but have been mapped at such a scale that they are difficult to incorporate into agricultural suitability. Studies of more localized conditions could add key components to future GIS reconstructions. Gardens have been an understudied component of Native American traditions of cultivation, and “gardening was an integral subsistence activity, wherever people resided permanently,” although it also varied considerably (Doolittle 2000:117). Unlike fields, they typically contain multiple crops, require ongoing maintenance throughout the year, are more likely locations for experimentation, and they are typically located in greater proximity to one’s residence. Still, the distinction between gardens and fields is subjective, and similar suites of crops may be produced in either. However, it is also worth noting that the general characteristics of gardens make them more likely to produce reliable if not smaller crops every year (Fish and Fish 1984:69-70). Edgar Anderson’s (1952:136-151) dump-heap hypothesis suggests that middens located near permanent or otherwise long- term residences are also highly likely locations for plant domestication and experimentation (in Doolittle 2000:82).

The intricacies of past uses of today’s common plant species remains poorly understood. Cheno-am and maize are generally the most abundant food plants identified

193 in pollen and macro-botanical analyses from the archaeological sites discussed here, but cultivation of and reliance upon other species is elusive. Obvious candidates discussed previously include agave, piñon, juniper, mullein, pine, and perhaps even osha, hops, and herbs, such as sage and oregano, along with a number of grasses and greens. Other plants besides agave considered indicators of human occupation include wolfberry and bear grass. Associations with trees and shrubs have received less attention.

Summary

There is a strong climatic gradient in the study area. Vegetation communities, precipitation, snowfall, and length of growing season are all strongly correlated with elevation. Small architectural sites are clearly concentrated in a particular range of elevation. Site location models have focused on this relationship, and a host of environmental variables, the vast majority of which are closely tied to elevation, can be explored. Exploratory analysis of correlation coefficients can be problematic. However, inductive probability models have proven useful and reliable. Representational models of other factors that may have influenced past behavior are also useful. A deductive approach to agricultural risk was proposed, but it suffers from a lack of consideration of micro-environmental variability. Given the strengths and weaknesses of different types of models, it is clear that multiple approaches are always favorable, and they should be examined for future avenues of research.

194

CHAPTER 6.

SMALL ARCHITECTURAL SITES

AND THE SYSTEMIC LANDSCAPES OF THE MOGOLLON PLATEAU

The previous chapter contributed some new insights into the long legacy of attempting to understand the relationships among environmental characteristics and locations of archaeological sites on the Mogollon Plateau. By focusing upon small architectural sites, some interesting patterns emerge regarding their distributions and associations with modern environmental conditions. These sites represent a palimpsest of archaeological landscapes that have endured in the study area over the course of several centuries, but the systemic relationships among small architectural sites and other types of sites warrant additional consideration. In this chapter, I return to some of the models introduced in the closing section of Chapter 3 and explore the relationships among different classes of sites, drawing upon more detailed evidence from specific areas. The clustering of sites demonstrated earlier suggests the study area could be characterized as a basin of attraction. The issue of site function warrants closer attention, and relationships with common pool resources are also explored. The relationships among Ancestral

Pueblo sites and these types of resources, including potential sources of wood, agricultural land, hunting territory, and foraging grounds for various edible and otherwise usable plants, have encouraged different theories about possible over-exploitation of such resources. Evidence of cooperation and integration has been emphasized at some sites,

195 while competition and conflict have been suggested as well. There has been discussion of a “tragedy of the commons” scenario unfolding in part of the study area. Is there evidence of sustainable use of common pool resources or indications their use led to a

“tragedy of the commons”?

Community Organization and Small Architectural Sites

Herr (2001) expounded upon Plog’s (1983, 1984) distinction between strong and weak patterns in the distribution of settlements throughout the landscape, noting areas with small sites and evidence of unspecialized production, use of local materials, and a high degree of variability in artifact and architectural design as examples of weak patterns. Strong patterns include more centralized settlement structures and increased similarity among architecture and artifacts, including craft specialization. Herr’s perspective on the “land-rich and labor-poor” environment demonstrates how the

Mogollon Rim region was a frontier landscape, marginalized from socio-economic and political influences that dominated surrounding regions, including notably “stronger patterns” among the Chaco, Mimbres, and Hohokam traditions.

Past surveys indicate much of the area was probably occupied seasonally and sporadically during most of the prehistoric period. Lithic scatters and isolated projectile points associated with the Archaic period are encountered periodically, but not in great abundance and most frequently in closer proximity to the major canyons and drainages.

196

Pit house villages comprise a very small portion of the total sites, although this is probably at least partially attributable to difficulties in identification of pit houses in some survey contexts. Excavations at artifact scatters suspected to include pit houses have even found their remains to be elusive (Tuggle 1982), suggesting less formal shelters were often used.

The Silver Creek area witnessed a growing population by A.D. 1000 that could not be accounted for by natural population growth alone, and Newcomb’s (1999) demographic reconstructions suggest dispersed, small architectural sites accounted for a greater amount of the population in the Silver Creek drainage than larger sites, at least until the 12th century, when the population began to aggregate into fewer villages. This trend generally seems to have been followed in adjacent regions as well, including the western portions of the study area. Although populations seem to have been growing rapidly at times, they were still probably fairly low overall, ranging from estimates of less than 1 (early) to no more than 14 people per km2 in the 13th century in specific portions of the Chevelon region (Plog 1974b).

Herr (2001:41) found that small sites tended to be fairly evenly scattered throughout the Silver Creek area and in the vicinity of intermittent drainages. She also determined that they did not tend to spatially cluster around great kiva sites, which appear to have been the primary form of integrative community architecture in the area in the

11th and 12th centuries. The association with drainages was alluded to in the previous chapter by comparing site locations with Euclidean distance calculations from

197 intermittent and ephemeral drainages. A similar distribution of small sites is apparent in the northwestern portion of the study area, but in association with more ephemeral drainages, and has been noted in numerous survey projects.

Within the Chevelon/Chavez cluster, researchers have noted a preponderance of small settlements that include more terrace and check dam features with presumed agricultural functions, often exhibiting a preference toward sandstone deposits and their derived soils (which tend to be intermingled with limestone deposits to varying degrees, as discussed earlier). From the 10th through 13th centuries, the most common sites are characterized as loosely clustered groupings of two to three rooms (16 to 25 square meters in size), although single rooms with well-developed middens are common

(Solometo 2006:40). Aggregation seems to have become more common in this region after the middle of the 12th century, sometimes in association with fortifications or largely inaccessible landforms. Although fortified sites in the Chevelon region seem to have been abandoned around A.D. 1250 and there is little use of the area after A.D. 1300, a brief return to unfortified and dispersed use of the areas between the canyons has been noted.

As discussed earlier, Haury (1985), Preucel (1990), and others have highlighted the relationships between small architectural sites and larger settlements. This is often characterized as a distance decay (or gravity) model and proposes that seasonal circulation results in a combination of both temporary shelters, such as field houses or other limited activity sites, and permanent village residences. In this manner, seasonal

198 circulation would have facilitated use of more distant fields, presumably supplying direct economic benefits afforded by local environmental factors. Preucel (1990:177) also sees it as a conflict resolution strategy that “may have been linked to the minimization of the inevitable social stress which attends communal village life.”

This perspective suggests that as populations become increasingly aggregated into larger settlements, they will be required to travel greater distances to reach suitable and productive areas as the immediate vicinity of a population center becomes occupied. In the study area, aggregation gradually increased through the Pueblo periods before increasing fairly rapidly in the mid-1100s, eventually culminating in occupation of a far more limited number of larger sites. The distance-decay model would suggest high densities of small sites surrounding these highly populated centers. As seen in Figure

6.1, this pattern is presented in the data. However, it is important to remember that the vast majority of small sites in the study area pre-date the largest communities in the area.

Although there is evidence of more intensive occupation during these later periods, survey data have repeatedly demonstrated the earlier periods of occupation were far more extensive in their distribution across the landscape.

Doolittle’s work among smallholder agriculturalists (1988) observed a trend that periods of use of marginal lands tend to become lengthier through time because of the cumulative value of the small investments in cultivating the land. Although this principle was primarily elucidated on the basis of investment in canals and other related technology, it is also applicable to other labor investments that may seem less

199

Figure 6.1. Density of small architectural sites (classified by number per square km) and large room 199 block sites in the study area.

200 technologically advanced, but would have still required a substantial amount of effort in accordance with various environmental or social factors. Moore (1978) argued that rather than representing a function of urbanization, “seasonally utilized farm structures” were instead a result of inconvenient access, and they simply duplicated features at larger village sites. This suggests population estimates could be inflated if not properly accounting for seasonal circulation.

Given the fact that large surface pueblo sites are almost always the latest precolonial sites in the region, it is interesting to observe their distribution in relation to the interpolated density surface of small sites, which primarily pre-date the large late sites

(surveys throughout the study area have repeatedly demonstrated that a very limited proportion of small sites are associated with these later periods, and those that do include later ceramic types often include earlier components). In this respect, small sites in the study area diverge from the pattern seen in many Eastern Pueblo regions, such as Preucel

(1990) discusses, where there are many small sites dating to later periods.

Two primary clusters of large sites have been identified to date. The Mogollon

Rim cluster of large sites generally mirrors the much larger distribution of smaller sites in the surrounding area, and a similar association is seen in the northwestern portion of the study area. The intervening area includes a fairly high number of small sites distributed through a large area, but fewer large sites have been identified than the other areas. As seen in the previous chapter, these areas are characterized by lengthier growing seasons

201 but lower precipitation, and the most common sites are located in the vicinity of ephemeral streams with apparently-productive pockets of sandstone-derived soils.

Interestingly, great kiva sites, which tend to pre-date the largest room blocks, are present within the area that includes fewer large sites (Figure 6.2).

Social Conflict and Culture Change

in the Mogollon Plateau Region

Social conflict manifests itself in many ways, with warfare being among the most extreme expressions. Modern human societies are fraught with instances of social conflict as a driving factor in culture change, but despite numerous modern examples of this phenomenon and the fact that such conflict has been a persistent human phenomenon for millennia, archaeological investigations of social conflict during prehistory have been hampered by lack of a clear definition of the phenomenon and an understanding of its place within the broader realm of social conflict. Operationalizing the concept of warfare in the context of archaeological research has proven problematic. Social conflict has been invoked as an important causal factor in social change in the Mogollon Plateau region, particularly during the 13th and 14th centuries, and it warrants more detailed consideration.

202

Figure 6.2. Locations of Great Kiva sites (shown in bright green) in the study area and interpolated 202 density surface of small architectural sites (number per square km).

203

Warfare and other forms of social conflict have been long-standing points of discussion in Southwestern archaeology, but debate about these phenomena as an explanatory factor of culture change and its archaeological evidence became a point of more frequent debate in the past two decades (Arkush and Allen 2006; Billman et al.

2000; Farmer 1997; Haas 1990, 2001; Haas and Creamer 1993, 1996, 1997; Hegmon

2000; Kohler and Turner 2006; Kuckelman 2000, 2010a and b; LeBlanc 1997, 1999,

2000; Lekson 1996, 1999, 2002; Liebmann et al. 2005; Lowell 2007; Nichols and Crown

2008; Plog 2003; Potter and Chuipka 2010; Rice 1998; Rice and LeBlanc 2001;

Schaafsma 2000; Solometo 2004, 2006; Wilcox and Haas 1994). Much of the resulting frustration has arisen between those who support the invocation of warfare as an important explanatory factor for change in Southwestern prehistory and those who oppose this position, resulting, at least in part, from disagreement and contradictory assumptions regarding the various meanings of concepts associated with social conflict, as well as the measures for identifying them in the archaeological record.

Vencl (1984:124) argues that the many difficulties associated with operationalizing concepts of social conflict in the archaeological record require us to consider that “every cultural change in archaeology is to be investigated a priori for a possible relation to warfare.” Yet, Haas’ suggestion that “the absence of warfare is somewhat more difficult to recognize archaeologically” underscores the potential to overestimate the pervasiveness of traces of behavior that may be related to warfare in its many forms (Haas 2001:331). LeBlanc (1999:8, 55) warns that “the threat of warfare

204 will result in many of the same behaviors as will outright conflict.” These perspectives represent a double-edged sword for Southwest archaeology: we run the risk of uncritically “pacifying the past” by holding claims of warfare to impossibly high standards, while at the same time needing to be careful to not accept isolated evidence alone as an indication of violent, intergroup social conflict (Keeley 1996, 2001).

The propensity of violent, social conflicts to elevate beyond the level of the individual to larger social groups, with the intention of altering the balances of social and political power has become an inherent characteristic of human society. In many cases, the mere threat or fear of such escalated levels of violence manifest themselves in human behaviors in highly visible ways, many of which can produce lasting, physical evidence amenable to archaeological analysis. Recognizing that these behaviors are deeply integrated into various forms of human practice highlights the challenges of validating extreme forms of social conflict as primary explanatory factors for social change

(Pauketat 2009). Warfare indubitably occurred among prehistoric groups in the

American Southwest, but invocation of the concept as a major explanatory factory warrants careful consideration and examination in the context of broader, regional trends.

Skeletal evidence of violent death is probably the most commonly cited direct evidence of warfare, but osteological indicators of trauma do not necessarily imply warfare, as it may result from other forms of social conflict, both within and between social groups. Also, it is often questionable to extrapolate evidence from a limited number of burials to a larger group, let alone argue for conclusive interpretations of

205 osteological evidence in all cases (Milner 1999; Milner et al. 1991; Turner and Turner

1999). Key indicators of intergroup conflict include scalping, decapitation, embedded projectile points, and blunt-force trauma (LeBlanc 1999; Milner 1995; Steadman

2008:53). Other less direct skeletal evidence can also indicate warfare, including mass burials, population profiles that under-represent a certain segment of the society (such as potential male warriors or women and children potentially taken as captives), and evidence that individuals were interred either haphazardly at the time of site abandonment or were not buried shortly after death (e.g., skeletal damage or disarticulation attributed to scavengers).

Another line of evidence that has received attention is iconographic representation of social conflict. Rock art, ceramics, and murals offer the most potential in the

Southwest (Farmer 1997; Haas 2001; Schaafsma 2000; Wilcox and Haas 1994), although there are significant interpretive challenges in comparison with more overt examples from other regions of the world. Depictions of weapons, as well as actual weapons themselves, may also be cited as evidence of warfare. However, in nearly all cases potential weapons can be considered multifunctional, and their context is probably most important for identifying them as evidence of warfare.

LeBlanc (1999) suggests that less direct lines of evidence than skeletal data may be more useful for identifying warfare, and there are several potential sources, although their association with warfare may be ambiguous. Architectural data have traditionally been offered as evidence of warfare in the southwest, and Farmer’s (1957) typology of

206 defensive systems continues to provide a foundation for interpretation (LeBlanc 1999;

Wilcox and Haas 1994). Fortifications, palisades, settlement locations selected for topographic prominence or inaccessibility, and other characteristics determined to be

“defensive” have all been commonly cited as evidence of warfare, although they could also be indicative of less severe forms of social conflict (Farmer 1957; Keeley et al. 2007;

Kenzle 1997; LeBlanc 1999; Liebmann et al. 2005; Powers and Johnson 1987; Saile

1975; Towner 2003; Welch and Bostwick 2000; Wilcox and Haas 1994). Burning of architectural features is also offered as evidence of the severity of social conflict, although its meaning can also be inconclusive. Although these types of archaeological evidence do not necessarily imply the occurrence of lethal conflict between different groups, they may imply a concern for such a threat.

Settlement pattern data have also been cited as evidence of social conflict (Hill and Wileman 2002; LeBlanc 1999; Wilcox and Haas 1994), although the behaviors that produced such evidence are not necessarily related to violent, organized conflict among social groups. Generalized population aggregation is often associated with elevated levels of social conflict, and the inward focus of much late Ancestral Pueblo architecture has repeatedly been interpreted as a defensive response. While aggregation could be viewed as evidence of a failure to resolve conflict and potential or perceived threats of violence from other social groups, this could also be seen as evidence of a community attempting to integrate multiple social groups or resolve intragroup social conflict

(Adams 1989; Bernadini 1996; Ensor 2000; Kenzle 1997; Nelson 2000). Intervisibility

207 among settlements is frequently cited as evidence of potential social conflict (e.g., Haas and Creamer 1993; Welch and Bostwick 2000). Although this is a characteristic that can be quantified and objectively explored, the results require subjective interpretation, as do many other potentially tactical characteristics, and it could be attributable to other behaviors, such as ceremonial or other communal events (Dean 2010). Furthermore, this characteristic also seems to be more suggestive of active cooperation and concern, rather than outright competition.

Examinations of site distributions have often noted the tendencies of sites to cluster, with the resulting formation of “no-man’s lands” or buffer zones between these clusters (LeBlanc 1999, 2006; Wilcox and Haas 1994). These buffer zones, often viewed as an indication of sociopolitical autonomy, are apparent throughout the world at different times and are offered as “one of the most ‘legible’ signatures of warfare” by

LeBlanc (1999:70). However, sociopolitical autonomy does not necessitate the occurrence of lethal, sanctioned, intergroup conflict at any particular scale or intensity.

Furthermore, the economic and ecological considerations of such patterns potentially offer greater interpretive power regarding issues of subsistence and cooperation, not to mention the simple fact that people usually tend to live within proximity to one another.

Possible “no-man’s lands” include the southern portion of the Black Mesa Ranger

District, encompassing a high-elevation portion of the study area dominated by

Ponderosa pine forest interspersed with higher-elevation and cold-air drainage stands of mixed conifer and spruce-fir forest along with stands of aspen, the eastern portion of the

208

Lakeside RD, which includes the western portion of the Springerville Volcanic Field, and the western portion of the CNF part of the study area, which also includes a high elevation area whose western margins give way to the Verde River area and the Southern

Sinagua tradition (Reid and Whittlesey 1997). There has not been enough systematic survey to argue for the presence or absence of sparsely-used areas to the north of the study area, but to the south of the ASNF portion of the study area there is an intervening region of fairly rugged terrain between apparent site density concentrations to the south, such as the Forestdale Valley, and to the southwest in the Grasshopper region.

The challenges of time and space systematics provide further difficulty when interpreting settlement pattern data, particularly when data is limited to surface survey alone. Contemporaneity and seasonality can be difficult issues to address, and assessments of intervisibility and aggregation can be questionable if these issues are not examined. Limited activity sites are difficult to interpret, and their assignment to a particular group at a particular time can be virtually impossible. Territory formation is an important aspect of how different groups create places and pursue their livelihoods, but territories are not necessarily an indication of the occurrence of warfare or other forms of extreme social conflict.

In sum, the various expressions of warfare and other forms of social conflict are potentially manifested in a wide array of human behaviors. In most cases, including osteological evidence of trauma, these lines of evidence do not unambiguously identify the presence of warfare. Many may indicate a perceived threat or socialization for fear of

209 conflict, rather than the actual outcomes of violent, intergroup conflict. Inferences concerning large groups and populations require multiple sources of archaeological information and evaluation of multiple working hypotheses. A priori assumptions that warfare has always been present are not sufficient proof. At a minimum, multiple lines of evidence must indicate involvement of groups rather than individuals, including some direct indication of physical violence, and one or more sources of indirect evidence suggesting the prevalence of social conflict in a given region at the time period in question. While this array of archaeological evidence may indicate the presence of warfare or some other form of social conflict, its role as a primary causal factor in culture change will require more substantial regional investigations.

Solometo (2006:41) characterizes rectangular pueblos with enclosed plazas as

“The earliest archaeologically visible defensive strategies in the study region.” Plog et al.

(2001) investigated some of these sites and suggests violent social conflict during the mid-12th century resulted in abandonment of most small sites and the persistence of enclosed pueblos. Some sites were found to be located upon topographic prominences away from presumed agricultural areas and incorporated presumably defensive architectural elements, such as loopholes (Solometo 2006; White 1976). Solometo’s

(2004, 2006:45) proposed typology for these sites includes the lookout, isolated refuge, proximate refuge (to support larger groups from nearby settlements), and year-round defensive habitations. The evidence that the region was increasingly “organized for war” toward the end of the area’s precolonial occupation hinges on only 13 sites included in

210 the current study area (Solometo 2006:41). Her discussion included three additional sites located north of the study area, in the vicinity of East Sunset Mountain.

Although the sample of defensive sites is rather small, it should not be dismissed.

Their locations along Clear Creek and Chevelon Canyon are noteworthy, as the general area seems to mark a frontier between the clusters of large, primarily late sites (see

Figures 6.1 and 6.2). This general area has also been used to delineate part of the southeastern boundary of the Northern Sinagua archaeological culture area. And, the canyons seem to have likely been used as travel routes and important landscape features.

Interestingly some of these are similar to types of sites identified by Senior (2004) as possible candidates for Apache and Navajo occupations. Most importantly, nearly all of these sites are situated in areas of topographic prominence, especially along Clear

Creek and Chevelon Canyon and their tributaries. These types of topographic prominences (see Solometo 2006:44, 46) are fairly rare in much of the rest of the study area. The sum of the evidence for social conflict in the area suggests “the conduct of war…is suggestive of a socially distant enemy…unlikely to have been caused environmental stress or land shortage” (Solometo 2006:56).

Migration from these areas to Anderson Mesa is suggested around the beginning of the 14th century, but re-occupation of less defensible small architectural sites on the flats near the canyons has been documented during the final 50 years before the apparent end of extensive use of the area. Although archaeological evidence is scant and probably under-recognized, the region continued to be used periodically by various groups through

211 the following centuries. Interactions with new groups arriving in the area warrant further investigation and re-consideration of survey strategies.

A Tragedy of the Commons?

Much of the study area has been characterized by relatively low-productivity from modern standards, and it has been referred to as both a frontier and a marginal area.

Slatter (1979) expounded on Plog’s model of Pueblo abandonment, suggesting a lack of equilibrium between food supplies and population growth in marginal areas was met by agricultural innovation among growing populations. These innovations were signaled by water control devices and redistribution networks with populations occupying more favorable areas where environmental variation was not as significant a problem, serving to reduce subsistence stress among populations in marginal areas. Competition for subsistence resources was exacerbated by migrant populations, disrupting re-distribution networks and causing social conflict and warfare that encouraged movement to less marginal areas, where new organizational strategies were developed to moderate the influences of environmental variable. This model, as presented by Slatter (1979), is remarkably similar to resilience theory, which has been invoked by numerous researchers in recent years, including the Chevelon area (Peeples et al. 2006).

Some have suggested evidence of social conflict in the Chevelon Canyon area could be attributable to a “tragedy of the commons” scenario (cf. Hardin 1968; Kohler

212 and Matthews 1988). As noted earlier, Fertelmes and Barton (2007) suggested a significant labor investment was needed to clear land for agriculture and construct habitation and storage structures in the Chevelon/Chavez area. They also hypothesize piñon-juniper vegetation communities have remained impoverished despite not being farmed for several centuries. “In this sense, agriculture in this region was initially productive but ultimately unsustainable. Short-term land use practices destroyed the long- term productivity of this region for maize horticulture” (Fertelmes and Barton 2007:14).

Peeples et al. also assert that later sites are situated in locations “favoring more defensible agricultural fields,” and “by ca. 1300, this combination of social and natural processes had set off a cascade of socio-ecological changes that may have triggered an ecological reorganization of the piñon-juniper woodland and certainly resulted in the human abandonment of the Chevelon Crossing region for centuries” (Peeples et al. 2006:14).

An important point missing from the discussion of environmental degradation has been the influences of modern formation processes. Forest Service records indicate several surveys have been completed in the area surrounding Chevelon Crossing. Also, there are numerous roads and trails, as well as an electric transmission line. The areas in question are located fairly near the Forest boundary, and there is a small community of houses on the opposite side of the transmission line. Use of the area for grazing has not been considered either. Although pressure for fuel wood harvesting is not remarkably high, there is also recreational use, especially hunting and camping. Prior to proposing centuries-long environmental degradation brought about by mixed-economy

213 horticulturalists who apparently used the area primarily on a seasonal basis, more careful consideration of the impacts of modern land uses should be undertaken.

Kohler and Matthews (1988:539) explicitly state that prior research by Stiger

(1979) overemphasized the presumptive role of slash-and-burn cycles intended to improve soil fertility in the Four Corners region, but they reaffirm the significance of deforestation brought about by agricultural intensification. Also, Clay et al. (1985) found no evidence that dry farming practices on sandy, aeolian soils like those at Mesa Verde would have made the soils infertile. An experiment undertaken in Mesa Verde National

Park beginning in 1918 calls into question some of the assumptions regarding soil fertility and the need to rotate field locations. A small plot cleared of piñon-juniper woodland produced crops in all but two of the following 17 years without the application of water or fertilizers (Franke and Watson 1936). Continuation of the same protocol during an additional four decades resulted in only a single crop failure (Stiger 1979;

Wenger 1980:56). There is some evidence that fire was used to clear land, although intensive slash-and-burn agriculture probably was not used; fire has undoubtedly played an important role within these ecosystems and there is evidence it was used by Apaches to clear land (Roos 2008), and its use in piñon-juniper woodland tends to encourage growth of annual and perennial edible plants, as well as general overall increase in biomass (Sullivan 1996:147-148). However, intense and severe burns would be counter- productive and seem an unlikely strategy.

214

Presumed locations of agricultural fields in most of the Chevelon area appear to typically be around small ephemeral washes in sandy soils. Water control devices are evidence of efforts to capitalize on the limited flows of these ephemeral washes.

Although ephemeral, runoff in these areas would have provided nutrients periodically.

The TES surveys of the area, which included soil scientists, were not focused on past anthropogenic influences on ecosystems, but they noted no landscape scale impairment of soils in these most heavily used areas (USDA 1989). Geoarchaeological studies of agricultural settings in the Southwest have identified some evidence of depletion of soil nutrients, especially nitrogen (e.g., Sandor et al. 2008). However, even in those cases, greenhouse experiments showed these “impaired” soils were capable of growing crops, and simple fertilization alleviated any shortcomings. Although there is little direct archaeological evidence of fertilization of soils in the Southwest, several indirect methods were employed, including the use of simple structures to slow runoff, additions of organic debris, fallowing, and widely spaced plantings. All of these strategies are appropriate to the study area and it seems highly likely that they would have functioned to maintain the productivity of the area. Given the subsequent re-use of the presumably fragile upland environments, it seems unlikely that the settlement shifts seen in the study area are attributable to anthropogenic degradation of soil fertility, as has been demonstrated in other areas (Sullivan 2000).

Native American horticultural pursuits have clearly left a mark on the landscapes of the Mogollon Plateau region. Rather than focusing upon the potential shortcoming of

215 those adaptations or their susceptibility to environmental stress, I believe that the common pool resources of the Mogollon Plateau region offer insights to the affordances the landscape offered to its occupants, who appear to have been attracted from wide- ranging areas. Low population density and archaeological indications of somewhat fluid social identity in some areas provided opportunities for groups to use fairly expansive areas and interact with others, developing social networks that continued to influence the region after land use shifted toward more intensive strategies in the later Pueblo periods.

Common Pool Resources, Sustainability, and Culture Change

Ostrom (1990:30) defines common pool resources as “a natural or man-made resource system that is sufficiently large as to make it costly (but not impossible) to exclude potential beneficiaries from obtaining benefits from its use.” Typical examples include bodies of water such as rivers, lakes, and oceans used as fishing grounds, foraging territories including expansive forests, mountainous areas (a hallmark of the study area), and other ecotones covering large geographic areas, such as hunting territories. In many cases potable sources of water can also be considered common pool resources. Certain landscape elements such as sacred mountains or other land forms that play an important role in a group’s ritual life could also be considered common pool resources, depending upon their characteristics. Two important aspects of common pool resources are excludability and subtractability (Ostrom et al. 1994). Excludability refers

216 to the degree to which it is possible to prevent an individual or groups from extracting resources, and subtractability refers to the degree to which use of a resource excludes others from obtain the same benefit (Ostrom 1990; Ostrom et al. 1999).

Unlike predictions from game theory, field and laboratory experiments have shown that face-to-face interactions often deter over-extraction from common pool resources in some, and costly punishment can further enforce such behaviors (Anderies et al. 2011). In this sense, various socio-ecological conditions provide means of balancing short-term material benefits with more sustainable long-term gains. These mechanisms of control via collective action may not be readily apparent in simplified models offered by game theory and behavioral and experimental economics (Anderies et al. 2011;

Poteete et al. 2010). Common pool resource experiments have shown that cooperation is supported by face-to-face communication and also encourages sanctioning of “free- riders” (Ostrom et al. 1994). Economists have relied upon experiments to assess these relationships in large part due to the difficulty of eliciting such information from ethnographic settings. However, Wiessner’s (2009) study showed that the anonymity afforded by the experimental environment tended to remove institutional norms of reciprocity and sharing, as well as punishment. Regardless, Henrich et al. (2004) completed a cross-cultural study of 15 small-scale farming groups and found that social institutions play an important role in structuring decision-making processes among these groups and there is a strong tendency toward cooperation rather than self-interest. Two important factors influencing the development of social institutions to control common

217 pool resources include the degree to which a community relies upon the resources and the ecological conditions of the resources themselves (Ostrom 2005).

Bayman and Sullivan (2008) explored the significance of common pool resources in their comparative study of the Papaguería region of southwestern Arizona and the

Upper Basin of the Grand Canyon area, in an attempt to understand some of the macroeconomic aspects of the socio-political organization of communities occupying areas that have traditionally been considered hinterlands outside of regions archaeologists have typically considered to be core areas, exhibiting stronger patterns of centralization.

Their findings suggest that long-term macroeconomic adaptations associated with fluctuating common pool resources were supplanted in some cases (but not all) by competition for economic control of peripheral areas at later times. Tainter and Plog

(1994) identified these earlier weak patterns associated with small sites as exhibiting evidence of low population density and mobile adaptations, with little evidence of formalized or centralized sociopolitical control. Subsequent changes in control of joint- use territories and common pool resources signaled shifts in land tenure systems and socio-political systems.

In many cases, it is important to consider the fact that multiple common pool resources can be “nested” within the same ecological setting (Sarker et al. 2008). While most studies have focused upon access to a single common pool resource, it is important to recognize the multiplicity of potential resources, as well as their interconnections and interdependence. This perspective has been applied most often in regard to watershed

218 management, but offers insight to past uses of common pool resources as well. In the study area, common pool resources are influenced strongly by the climatic gradient influenced by the orographic effects of the Mogollon Rim. The variations in the forested belts emanating from the Rim are crossed by numerous drainages of varying depths.

These intersections provided numerous opportunities for overlapping resources and interactions in their use (Table 6.1).

Borrowing from complex adaptive systems, the study area can be conceptualized as offering an array of attractors that can best be characterized as common pool resources. Migrations and subsequent population growth contributed to the burgeoning basin of attraction that fueled transformation of the socio-political structure of settlement organization through the Pueblo periods in the study area. The small sites associated with these developments clearly were not organized randomly, and they cluster within specific settings. There is a clear pattern of more extensive use early, followed by more intensive use in later periods.

In the case of wild plants, the best assurance to “control” a resource may have simply been to assure one’s arrival occurred at the right place at the right time. The subtractability of some wild plant foods could be affected by bear, deer, and other animals. Certain plants present opportunities on varying schedules – even arboreal foods like acorns, juniper berries, and piñon nuts are available at varying intervals and are not always reliable on an annual basis (Floyd and Kohler 1990). Furthermore, these and

219

Table 6.1. Some common pool resources in the study area and their characteristics.

Common Pool Resource Excludability Deductibility Distribution in Study Area Edible wild plants Difficult Variable Widespread Medicinal wild plants Variable but unlikely Variable Variable but widespread Fuel wood Difficult Low Widespread Building materials Difficult Low Widespread Craft materials Variable but unlikely Variable Variable but widespread Hunting territories Difficult Variable Widespread Agricultural land Possible Low Widespread Precipitation Impossible None Highly variable Springs Possible Low Variable Streams Possible None Variable Trails/routes Possible but unlikely None Unknown, likely widespread Shrines Possible but unlikely None Few confirmed, many surrounding Other sacred sites Possible but unlikely None Unknown, likely widespread Other rocks and minerals Variable Low Variable, some widespread

219

220

other trees and shrubs provide many resources for non-subsistence needs, including adhesive pitch, bedding, fuel, medicine, and basketry and construction materials, among other uses. Although not typically regarded as a source of subsistence, inner bark layers also offered undesirable sustenance in starvation scenarios, as documented archaeologically and ethnographically in other regions. Even if famine foods were not regularly used, seasonal circulation in their proximity provided opportunities to share this knowledge across generations in order to provide a buffer against low-frequency shortfalls.

Hunting territories must have been an important component of the systemic landscape, and they indubitably intersected with gathering activities in many instances.

Hunting grounds may have also been associated with other activities, such as pilgrimages to shrines or for collection of specific plants or other materials necessary for the performance of ceremonies. Hunting grounds may occur in low-site density areas and be characterized by isolated projectile points or low-density lithic scatters. The absence of ceramics at such locations does not necessarily imply areas were not used during the

Ceramic period, but that activities occurring in those areas did not necessarily require the use of ceramics. Use of higher elevation portions of the study area for hunting large game (Figure 6.3) is also attested to by modern hunters, who comprise a significant modern use of common pool resources in the area. These forays would have provided opportunities to access other resources (Figure 6.4).

221

Figure 6.3. Elk pictograph from rock shelter on the CNF, near the western margins of the Chevelon/Chavez site cluster.

Issues of Scale and Site Function

Artifact scatters are the most common Native American sites in the study area (Figure

6.5). The name of this site category also does not imply a particular type of function or associated activities, unlike many of the other site typologies presented in the past.

Although artifact scatters do not include evidence of above-ground architecture, many likely included small structures constructed of perishable materials, and many were also probably used for brief periods of time, perhaps repeatedly in association with

222

Figure 6.4. Bird eggs tucked within a pocket of grassland amidst the Ponderosas, one of the many fortuitous finds encountered during a hunting trip, a gathering foray, or an archaeological survey.

Figure 6.5. Interpolated density of artifact scatters (sites that include Ceramic period components).

223

224 seasonal activities that did not require enduring shelters. Ethnohistoric accounts point to a wide array of potential activities (see Table 3.1), including tending agricultural plots.

Reid (1982b) advocated for a settlement function approach, focusing on the behavioral activities represented at sites by grouping sites according to artifact assemblage diversity. Evidence of above-ground architecture does not necessarily imply specific activities were completed at that area. This warrants a return to the issues of site typology raised earlier. It is clear that the strategies that have evolved over the past several decades have infused an unnecessary degree of confusion and misguided implications in how we represent archaeological landscapes, especially in the case of small architectural sites. Assemblage diversity at small sites can address this issue.

However, it has not been undertaken or recorded systematically on a regional scale at a level beyond gross artifact classes.

One of the main hurdles to comparing sites in this manner on a broad scale is the issue of chronological precision. Gregory’s (1990) survey of the Sand Draw area noted that chronological uncertainty even affected how site boundaries were delineated; one site (AR03-01-04-440) included four separate architectural components, all single room structures (one may have had a second room), all with ceramics indicating use during the

Pueblo II period. Had the surface assemblage not appeared to span a continuous period, the features would have been assigned different site numbers. The architectural features

“could have resulted from sequential and non-overlapping construction and occupation or use of the four architectural units as easily as from absolutely contemporaneous occupation or use of them” (Gregory 1990:5). Whether or not excavation would solve 225 this issue is unknown, but he suspected the fairly low-density artifact scatter more likely resulted from non-contemporaneous use; a concentration of diverse lithic raw materials suggested a possible Archaic-period component. Gregory also noted the remarkably diverse ceramic assemblage, including plain, corrugated, white, and red wares, suggesting, “Mogollon, Anasazi, and Sinagua characteristics are apparent in the ceramic assemblages, and no definitive assignment of the remains to only one of these traditions is possible…but should be an important topic of future research” (Gregory 1990:8). This variation has been noted in other studies as well (e.g. Hantman 1983).

Cholla Project excavations in the Chevelon region (Reid 1982a) suggested some sites may have been occupied by groups originating from other areas, or by groups integrated into different exchange networks, as indicated by variations in ceramic assemblages and flaked stone artifacts. Plog’s (1989) review of the Sinagua identified similar expressions in the archaeological record of the Chevelon area. Cholla Project and

CARP excavations suggest mixed-economy horticulturalists took advantage of the gradient of ecological zones and niches afforded by the wide-scale influences of the Rim as well as the opportunities afforded by deeper canyons and broader plains. “Chevelon may have been used by different groups of people simultaneously, and in different ways”

(Whittlesey and Reid 1982:170). However, Cholla Project excavations suggested one apparent commonality in most cases among those who occupied the Chevelon region during its periods of highest population: lack of evidence of year-round occupation.

Returning to Gregory’s (1990) work at Sand Draw, high site density observed in his survey area and surrounding regions fits well with the similar trend observed in the

226

Silver Creek area (Newcomb 1999) of dramatic population growth in the 10th and 11th centuries in areas that had been used sporadically before that period. Although the high site density suggests intensive use of the area, it is probably only intensive in a relative sense. “Assuming that none of the sites was absolutely contemporaneous with any other and that no location was occupied or used more than once, a single, relatively small social group visiting the area only twice in every decade could have easily produced the remains observed” (Gregory 1990:11). Of course, some sites in close proximity to one another were almost certainly occupied contemporaneously and/or reoccupied. More likely, most small sites in the Chevelon/Chavez cluster represent a period of extensive rather than intensive use. Multiple structures in close proximity to one another could also reflect a cultural preference toward building new temporary shelters rather than re- occupying previously built structures, particularly if they had been established by a different social or ethnic group.

Doolittle (1988) found that periods of use tend to become lengthier through time because of the cumulative value of the small investments of smallholders. Although this principle was primarily elucidated on the basis of investment in canals and other related technology, it is also applicable to other labor investments that may seem less technologically advanced, but still required a substantial amount of effort. Short-term sedentism (Nelson and LeBlanc 1986) and “fallow valley occupation” (Nelson and

Anyon 1996) provide reasonable explanations of the extensive land use practices inferred from the distributions of small architectural sites. The site databases indicate the vast majority of archaeological features with presumed agricultural or horticultural functions

227

(terraces, check dams, water control devices, rock alignments, agricultural areas, etc.) occur in the Chevelon/Chavez region (Figure 6.6). Many of the field house and other small pueblo sites include associated check dams, terraces, and other types of water control devices. Assuming agricultural functions for some of these features may be tenuous, and more detailed studies of local conditions should be undertaken other possible functions than agricultural terracing.

Common pool resources account for a significant portion of the available subsistence base in the study area. Although there has been a strong focus upon agricultural suitability and identification of domesticated crops in archaeological investigations (both survey and excavation), the breadth of the foraging territory offered by the Mogollon Plateau region supplied resources of varying degrees of reliability and necessity. As population densities grew communities must have exerted some form of primary claims over particular resources; others, however, must have been difficult to enforce, given the nature of their geographic extent, seasonality, and varying reliability.

If the majority of the field house structures identified by archaeologists were indeed associated with agricultural pursuits, the burgeoning preponderance of above- ground architecture could be seen as an attempt to signal ownership of agricultural land.

The rapid migration and population growth beginning in the Silver Creek region around

A.D. 900 (Newcomb 1999:51) must have been a time when newcomers with tenuous claims to resources needed to exercise a certain degree of innovation, whether in language or other aspects of their social identity to facilitate their integration into

Figure 6.6. Distributions of archaeological sites whose primary feature types include presumed agricultural functions, 228 such as check dams, terraces, water control devices, rock alignments, and agricultural areas.

a developing system of land tenure and exchange networks. Social conflict and warfare would have posed another avenue, and the perceived threat of conflict may be responsible for some of the defensive posturing seen in Pueblo architecture. Regardless of whether small architectural sites were strongly associated with agricultural pursuits, similar strategies could have been employed to signal access to the other types of resources offered in the area as well.

Even after the dramatic population growth and subsequent peak around 1100, several larger communities persisted until the 1380s in the Silver Creek area, before it and the surrounding regions were largely abandoned. These later sites are often located in high-density areas of small sites, most of which pre-date these larger communities.

This suggests that the more intensive land use strategies associated with larger sites were not impeded by past use of the area, suggesting sustainable use of common pool resources and explicitly not supporting evidence of a tragedy of the commons scenario.

Although larger sites were occupied at this time, the overall population was probably not as high as it had been during the 11th and 12th centuries, when there was a major population influx. The proportion of small sites dating to the latest periods varies throughout different areas, but is generally very low. Not surprisingly, they are often in proximity to large sites and are also appear as later components of sites dating to earlier periods.

These large communities also seem to have been places where different identities were integrated into developing social networks. The Coalescent Communities Database

(Hill et al. 2004) noted a trend from dispersed subsistence-settlement systems in the 230 southern Southwest beginning in the late 13th century, characterizing the subsequent period as an interval of social conflict and socio-economic intensification. Subsequent network analysis of later sites has noted remarkably diverse associations of decorated pottery types, suggesting the communities that persisted through these tumultuous times in the region acted as “brokerage sites” (Peeples and Haas, in press).

The land use strategies associated with common pool resources encouraged flexibility and innovation among the various ethnic and social groups that occupied the area. Hill’s (2004) review of the complex relationships among social identity and varying degrees of linguistic innovation and focus among the Tohono O’odham, represented in ideological as well as material realms, contrasts localist and distributed behavioral strategies associated with perceived rights and claims to resources and

“economic privileges.” She found groups associated with more reliable resource bases tend to be more resistant to linguistic innovation and exhibit more “focus” in their speech. On the other hand were those whose land-use patterns exhibited more distributed characteristics: “Through accepting linguistic innovation and diversifying their linguistic repertoires, they hope to enhance their secondary claims on resources. By adding strengthened secondary claims on their limited primary claims on resources, they seek a resource base sufficient to sustain their well-being” (Hill 2004:126). The importance of regional variation in the Tohono O’odham example primarily associates these variations with access to water and its limitations upon social group size (Hill 2004:134), but the principles are applicable to other areas. More stable peer groups, especially among children, typically promoted sociolinguistic focus; contrary situations are more likely to

231 encourage sociolinguistic accommodation. The importance of geographic identity and resource access and its associations with language probably have deep roots in the

Southwest, and the Mogollon Plateau offers an important to area to explore future research on the topic. The structure of resources in the area and past archaeological investigations in the Mogollon Rim area offer many parallels to these issues, and subsequent inhabitants of the area provide even more opportunities to explore these issues.

Archaeologists’ understanding of abandonment, an important consideration in any region of the Southwest as well as a terminological disconnect with Native Americans, has been modified in some circles by injecting the role of agency and recognizing that abandonments are an important component of settlement strategies, as the products of economic, social, and political decisions (Nelson 1999; Nelson and Hegmon 2001;

Nelson and Schachner 2002). Rather than necessarily implying a failure to adapt, abandonment can signal successful outcome of negotiation of the “agents of change and agents of continuity.” These agents tend to be particularly fundamental components among regional, small-scale farming societies.

Innovation in language and other expressions of social identity clearly characterized the area throughout subsequent historical times as well. Senior’s (2004) review of ethnohistorical accounts, historical maps, and interviews with informants from descendent communities highlights the multi-ethnic nature of the region. Apache use of the area is well documented, and their permeable home-ranges and wide-ranging patterns of annual migration and resource use (Diehl and Herr 2011). The area has also been used

232 for a lengthy period by sheep herders, including Navajo and even Basque groups; the

Heber-Reno sheep drive connects the area to the Phoenix Basin. Ethnic co-residence and intermarriage among ethnic groups occurred at some of these large communities, as documented at Grasshopper and other sites in the mountainous transition zone (Dean

1988; Reid 1998; Reid and Whittlesey 1999; Senior 2004). Goodwin (1942) documented many trails throughout the area, and many of the canyons and other smaller drainages likely served as routes that connected different groups to parts of the study area for a wide range of purposes, including Chevelon Canyon and Clear Creek.

Social identity and primary and secondary claims to resources were probably also tested during times of environmental stress, including broad patterns that influenced large portions of the southwest (Dean 1996; Dean and Funkhouser 1995). Farming communities would have been particularly susceptible to these changes. Throughout the

Southwest, there was wide-scale variability among regions, especially between A. D.

1000 to 1150 and 1300 to 1650. Disruption of the division between the bimodal precipitation pattern of the study area and the summer-dominant precipitation pattern to the east resulted in a “chaotic configuration” and “an unprecedented change in conditions to which the human and biotic populations of the region had become accustomed” (Dean and Funkhouser 1995:94). The period of high population growth in the Silver Creek area appears to correspond with a period of generally favorable local environmental conditions (Dean and Funkhouser 1995:99), but biotic communities and their uses by different groups were influenced by climatic fluctuations more severely in other regions.

The Mogollon Rim provided more reliable rainfall than surrounding regions and the

233 region experienced fewer years of severe drought than the Colorado Plateau area

(Kaldahl and Dean 1999).

Upland farming strategies require careful assessment to balance the need to conserve seed but to also to take an appropriate amount of risk to yield a sufficient harvest. Lower risk strategies that avoided upland areas for long periods of time could have compromised a group’s access to suitable areas as populations increased and new land tenure systems began to outweigh the primary hunting and gathering uses that prevailed previously. Economic uncertainty would have encouraged risk-assessment strategies and knowledge of alternative subsistence strategies in times of stress, which must have been passed down through oral histories that incorporated aspects of ceremonial life, place names, and other landscape characteristics spanning broad geographic areas (e.g., Minnis 1996). Mobility, diversification, and expanding social networks could ensure access to different resource bases, including foods that are not generally preferred.

Discussion and Summary

This chapter addressed several important issues associated with small architectural sites and their relationships with some tangible aspects of systemic landscape characteristics, but it is important to consider these interpretations at varying scales. Subsistence stress and economic uncertainty influence adaptations among groups of subsistence-scale agricultural communities, encouraging innovation, borrowing, and

234 behavioral flexibility (Dean 1996). Some of the trends identified in the study area reflect strategically cognized landscapes (sensu Heilen and Reid 2009), although the precise significance of some aspects of the settlement systems in place remain poorly understood.

One of the more difficult issues to clearly access at the landscape scale of analysis, at least at this point in time, is what may have happened to small sites through time, especially in the final few centuries of the precolonial era. Clearly, small architectural sites appeared in great abundance during the 11th and 12th centuries A.D.

Newcomb (1999) demonstrated this in demographic reconstructions of the Silver Creek area, and survey reports have noted a similar trend in other areas, including the

Chevelon/Chavez site cluster. But there is potential danger in assuming the significance of small sites spatially associated with later core areas, as well as in assuming the temporal span of small-site occupations and the functional significance they may have served. Surface contexts and erroneous suppositions may lead to spurious conclusions.

The southeastern portion of the study area, in the vicinity of Show Low and the

Mogollon Rim, arcing to the west through the current communities of Linden, Pinedale,

Clay Springs, and toward Heber and Overgaard, offered access to an important suite of resources, including stands of oak trees. Heavy snow accumulations along the northern margins of the Rim provided ample runoff from snowmelt, saturating soils through the early spring, and providing opportunities for agricultural pursuits following the threat of the last frost. Careful selection of locations with favorable local topographic settings would have increased the likelihood of success. However, the area offered a broad zone of habitats to support a more mixed-economy adaption. Welch (1996) demonstrates the

235

Apache model is an appropriate model of the Mogollon settlement-subsistence system in the Grasshopper area, and it is probably applicable to much of the study area, especially prior to widespread community aggregation. This would equate with extensive land use strategies, which seem to be represented by the widespread and repeated, but probably not always year-round use of a large area.

The archaeological landscape has been homogenized to a certain degree by focusing upon landscape-scale analyses. Situating small architecture sites within systemic landscapes supports a better understanding of how these sites may be associated with modern land uses and landscapes. It is also important to understand the sites of the

Mogollon Plateau in the context of surrounding regions. The population growth seen in the area during between the 10th and 12th centuries seems to occur in surrounding areas as well. Although my focus has been squarely upon those areas immediately north of the

Rim, there are numerous travel corridors to the south through the mountainous transition zone that must have connected populations in different areas (Figure 6.7). To the west on the CNF, several routes approach the Verde River. Toward the central part of the study area, Cherry and Tonto Creek would have provided access to the Tonto Basin, while

Canyon Creek, Salt Creek, and Carrizo Creek connected with the Grasshopper region.

To the east, Limestone Ridge, the Forestdale Valley, and Corduroy Creek all offered ready routes linking the country above and below the Mogollon Rim. Clearly, there were many opportunities to access surrounding regions, providing opportunities to interact with other groups, especially during early periods in the pre-colonial area when mobility was an important aspect of many lifeways.

236

Figure 6.7. The view from the Mogollon Rim on the ASNF southwest toward the mountainous transition zone and the headwaters of the Tonto Basin.

On an even larger geographic scale, periods of population growth in the study area overlapped with significant shifts in the Hohokam region, as well as the depopulation of the Four Corners area. The establishment of large towns in the Middle and Upper Colorado and other regions and large-scale migrations toward Hopi influenced use of the region and the traditional systems that governed use of the region’s common pool resources (Adams 2002). Lange and Adams (2012) have been surveying and excavating north of the study area, finding a similar mixture of Mogollon, Sinagua, and

Ancestral Pueblo as Gregory alluded to in the Sand Draw area. Their preliminary

237 findings suggest the area was probably used by various groups, eventually showing stronger affinity with sites from the Homol’ovi area along the Middle Little Colorado.

Ongoing research continues to affirm the frontier aspects of much of the

Mogollon Plateau region. The frontier characteristics of this area must have been fluid, changing through time as new groups arrived in various pulses. The structure of resources in the area, influenced by the climatic gradient paralleling the Rim to the north and cross-cut by drainages feeding into the Little Colorado, attracted many different groups and offered opportunities for interaction. This influenced a variety of exchange networks of material goods as well as ideas and experiences.

There clearly is evidence of social conflict in the area. It has primarily manifested itself in architectural forms. Burned rooms have been identified at many sites, although their significance is often equivocal and does not necessarily require violence or social conflict. The presence of several fortified sites is in the Chevelon and Clear Creek regions is noteworthy. The topography of these canyons, including associated rock formations, and the high-energy environments of the streams offered unique opportunities. Their persistence as travel routes seems to be suggested by the tendency of older sites to be located more closely to these main drainages. The use of fortified sites for long periods of time also warrants consideration and further investigation, including persistent use and subsequent re-use during following the end of the prehispanic era.

Even during periods of intensification and aggregation, the area was characterized by abundant land and a limited pool of labor. The shift toward more intensive agriculture that is apparent in the locations of the increasingly few 13th and 14th century villages (and

238 a limited number of small architectural sites) was a choice made in particular areas where it was worth doing. Even in these settings, forays into surrounding territory would have been needed. Joint use territories are common in ethnographic studies of Native

American communities in the Southwest, and the common pool resources criss-crossing the country. Salt pilgrimages among the Zuni and Tohono O’odham cover long distances. Forays into hunting grounds provided opportunities to encounter other groups venturing into sparsely occupied areas. Small structures are clearly related with common pool resources, and often in occur in various environmental settings as multiple and nested arrangements nested types. Past excavations have shown that there is a high degree of variability in construction methods, as well as the activities and resources that were used.

The flow chart depicting the approach adopted by this dissertation (see Figure

2.1) outlined a comparison of inductive and deductive approaches to assessing the significance of small architectural sites. They are clearly widespread and represent an extensive land use pattern, which itself is evidence of common pool resources.

Ethnographic examples of hunting territories and perspectives on the area from tribal consultation also support common pool resources as a useful model for understanding past use of the area. Areas with mixed ceramic assemblages and the interconnections among Sinagua, Mogollon, and Ancestral Pueblo groups, and the networks of their interactions (Hantman 1983). In addition to ceramic vessels passing through exchange networks, different aspects of subsistence and cuisine may have played important roles.

Exchange and use of different genetic crops could have occurred in this area.

239

Common pool resources provide a good model for the area in general also because much of the Mogollon Plateau region has been managed in this manner in more recent historical times as well. They can arise in different circumstances and present different opportunities to different groups. Extensive use of common pool resources does not necessitate a tragedy of the commons scenario, although it can arise. Societies have developed a number of mechanisms to maintain the stock of various common pool resources, and identifying those behaviors is an important avenue for future research.

240

CHAPTER 7.

CONCLUSIONS AND

RECOMMENDATIONS FOR FUTURE RESEARCH

The primary goal of this dissertation has been to present a comprehensive overview of the interpretations of small sites in the Mogollon Plateau region, and to reassess these interpretations through a comparative approach spanning a fairly large geographical area. This has been attempted by compiling a comprehensive database of small sites for the area to understand their variability. Two primary clusters of sites have been identified: one in the southeastern portion of the study area near the Mogollon Rim, and the other to northwest, primarily along the Forest boundary from in the vicinity of

Chevelon Canyon, Clear Creek, and Chavez Pass. Their previous characterization as a fairly homogeneous class belies their continued research potential. Many more of these small sites will undoubtedly be located and excavated but the data compiled here represent the largest database of sites so far recorded in the area.

Because my investigation of small sites has relied so heavily upon legacy data it is therefore susceptible to many sources of error inherent in such data. Many of these sites were documented on the behalf of the Forest Service in order to comply with historic preservation policies. This generates a significant amount of work to accurately document locations, describe characteristics of surface assemblages of artifacts and architectural remains, and report this work in a reproducible manner. This has proved challenging, and inspections of previously recorded archaeological sites have revealed

241 many irregularities. Land management agencies should regularly undertake extensive update and quality control reviews of site and survey databases; reliance upon GIS and

GPS data has become increasingly common in planning and fieldwork alike. Assuring accurate and precise data is critical to ensure preservation and efficient planning. In order for Heritage Program planning to be and remain effective on local, regional, and national scales, it must be based upon an accurate synthesis of documented cultural resources. Such studies, typically called cultural resources overviews, have successfully documented a number of important classes of archaeological sites, but others have been under-recognized, as have their interconnections as components of landscape networks.

This information should be used to project the distributions of various types of cultural resources, employing predictive modeling, in part, to identify appropriate strategies for identifying as yet unknown sites.

Ethnic Identity and Landscape Significance

The climatic and ecological gradients and transitional zones along the margins of the Mogollon Rim country developed into areas that drew upon burgeoning basins of attraction during the Pueblo period. This encouraged processes that have blended social groups and ecological conditions, evident prehistorically through ceramic studies and network analyses; historically through accounts of the relationships among Apache, Zuni,

Navajo, Hopi, and other tribes and Mormon colonists; as well as through today’s more recent settlers, retirees, descendant communities, immigrants, and even tourists.

242

My examination of small architectural sites reveals distributional trends that are not readily apparent when examining all of the site data in unison. By removing the most abundant class of Native American archaeological sites (artifact scatters), there appears to be a fairly obvious separation between concentrations of sites with evidence of above- ground architecture: the Chevelon/Chavez cluster in the northwest portion of the study area, and the Mogollon Rim cluster, to the southeast. Comparison with modern environmental data suggests these two clusters represent somewhat distinct adaptations to the environment. In the northwest portion, people were presented with lengthier growing seasons but more limited amounts of precipitation, and their small settlements tended to cluster along ephemeral rather than intermittent and perennial drainages. Closer to the

Mogollon Rim, precipitation is more plentiful, but the growing season is also more abbreviated within much of the site cluster. Snowfall and the period of time it remains on the ground also vary significantly between the two areas, with the Rim communities receiving more runoff from snowmelt. Also, the soils in that area tend to retain moisture for a lengthy period of time.

Archaeologists have tended to emphasize evidence of agricultural intensification and have begun to explore the possibility of more enduring environmental transformations, although these claims seem largely speculative and more evidence is required to explore associations of modern vegetation communities and the archaeological landscape. Descendant indigenous communities living in the vicinity of the study area tend to offer different experiences and perspectives on the significance of small sites than those that have driven archaeological research. San Carlos Apache,

243

Tonto Apache, White Mountain Apache, Navajo, Hopi, and Zuni have traveled through the area for centuries, collecting plants as sources of food, dye, medicine, and tools.

Apache groups collected various seeds, nuts, and other plant foods and tools along seasonal routes from farming villages (Basso 1971; Goodwin 1942). These routes also provided access to sacred sites, including several in the vicinity of the study area, including Chevelon Butte, Sunset Mountains, the White Mountains, and Green’s Peak, along with a host of many other less prominent mountains, springs, canyons, and other natural places, creating places and situating themselves within time and space (Anyon et al. 2005; Ferguson and Hart 1985; Greenwood and White 1970).

Koyiyumptewa (1993) documented a critical distinction in the spiritual values

Hopi ascribe to piñon-juniper woodland, noting its importance in ceremonies as well as general emotional well-being, versus Euro-American perspectives that have been more linear, and profit and management-focused. Clearing the land to improve range conditions for livestock can be seen as a particularly striking intersection of these perspectives, especially through the violent chaining and mastication methods used.

Industrial timber harvesting within the mixed-conifer and Ponderosa pine forests offers a parallel example. Despite this conflict, Koyiyumptewa (1993) expresses fear about disclosing the importance of native plants – since making such knowledge public could result in commercialization of spiritually-significant plants found in the woodland communities. In this vein, it seems apropos that any discussion of Native American destruction of woodlands should first start by documenting the extensive effects of behaviors from the more recent past; coupled with grazing, these activities have almost

244 certainly transformed these communities. For example, mature piñon provide ongoing supplies of fuel wood from their branches, in addition to edible nuts that have been a staple for centuries. Would mixed-economy subsistence farmers not recognize their impacts and simply proceed knowingly with land-use behaviors that could compromise their long-term well-being?

Although often unimpressive by themselves, the collection of small middens, thermal features, and architectural elements at these sites are important contexts for assessing long and short-term land use patterns and environmental change (Adams and

Fish 2007; Adams and Van West 2005). Recognizing the significance of this and similar studies and actively integrating their results into management decisions on public lands offers challenging goals for archaeologists and other anthropologically inspired resource managers. These issues include myriad inter-related goals arriving from cultural resource management, archaeological research, tribal relations, and the implementation of a wide array of modern land uses. The Forest Service and other land management agencies have begun to address climate change, water, and other sustainability issues (Kimball 2007), and the data yielded from this examination of small sites should play an active role in these missions and provide avenues for future research and more effective management of cultural resources.

Recognizing ecological transformations in unison with the structure, function, and time depth of human land use is critical, but these small sites also offer connections to the past for people, and they speak to issues beyond subsistence. They represent a palimpsest of the social networks that developed in within landscapes imbued with deep meaning for

245 many different groups. Small sites are also important legacies for Native people and their histories, and their significance should be recognized in cultural resources management, as well as the multiple use principles of public land management. By placing these sites within various landscape models with the tools of GIS, the limitations of archaeological site definitions can be exposed and landscape affordances can be explored, moving beyond traditional and often narrow definitions of site significance, be it within the setting of resource management or in the discussion of broader socio-ecological and cultural implications.

This is a critical moment in cultural resources management, especially in the

American Southwest. Modern populations have surged, and National Forests and other public lands remain open to many land uses, as well as threats from past mis- management, including overly aggressive fire suppression and grazing policies. The interconnections among cultural and natural resources and their active management can either strain our ability to recognize, document, and preserve many underappreciated aspects of the archaeological record, or offer insights for how people have used and changed the land through time. Political, economic, and social realities provide opportunities for cultural or heritage resource management programs to play growing roles in a variety of socio-ecological management decisions. Implementation of archaeological GIS is prone to data limitations and inappropriate use, but the potential applications they offer can be far more than “dots on maps” and offer insights to past and present values, as well as traditional and modern perspectives and their intersections.

The vast resources expended assessing the potential effects of large-scale fuel reduction

246 projects on National Forests may be directed toward more appropriate management techniques.

Implementing Predictive Models for CRM on Public Lands

The development and diversity of GIS-based predictive models offers great potential for supporting CRM activities on public lands. Despite the potential of these applications, recent examples and debates demonstrate that a single model or type of model cannot be appropriate for all situations. One-size-fits-all models are particularly troubling despite the fact they have been the most common approach. In order to fully reap the benefits of predictive models, it is important to identify the intended purposes of a given model, as well as the fact that multiple predictive models may be needed to fully address the various types of cultural resources likely to be encountered in an area. Rather than focusing on specific methods for creating predictive models, in the following paragraphs I present a general strategy for implementing predictive models as a tool for aiding CRM on public lands.

The first step in the process should be recognizing those circumstances where predictive models (or certain types of predictive models) are an appropriate tool for cultural resource management. All areas may not be suitable for traditional inductive models, such as regions where previously documented archaeological sites and surveys are extremely limited, as seen in Plog’s early SYMAP model. Reliability and accuracy of information about known archaeological sites should also be considered (e.g., Rua 2009),

247 along with additional sources of information about site locations nearby, such as private in-holdings and various types of federal and state ownership (Kohler and Parker

1986:406). Reliance on the concept of the archaeological site as the sole, meaningful analytical unit for archaeological analysis should be questioned; common pool resources like potential horticultural areas, hunting territories, gathering locations, and areas with spiritual or ceremonial significance should be considered wherever possible. The accuracy, precision, and completeness of available environmental data should also be assessed; the tendency to homogenize the environment masks a great deal of variability.

Although data availability is less of a limiting factor for modern environmental data, paleoenvironmental conditions also warrant consideration (Kvamme 2006:19), particularly in regions where environmental changes are perceived to have been significant.

In addition to available data, the suitability of predictive models can be dependent upon the degree of variability among ecological factors, especially in regards to the distribution and concentration of consumable resources, including wild plants, water, arable land, and suitable materials for tool production, as well as terrain characteristics like topographic diversity (Dean 1984). As a general rule of thumb, areas where resources are evenly distributed may be less suitable for predictive models than areas where distributions are patchy. Resource productivity, environmental and climatic stability, and social and ritual factors also warrant investigation. An individual’s current understanding of a region’s archaeology should always draw upon local expertise in the selection of a particular type of model. While traditional inductive models may be useful

248 in many cases, deductive approaches should also be evaluated. For example, a deductive model could predict the likelihood of looting based upon proximity to travel routes, populated areas, and site types, among other factors (Dore and Wandsnider 2006). A combination of different types of models may be most useful in some cases (e.g.,

Robinson et al. 2010; Seramur et al. 2009). Once an area is determined to be suitable for a particular type of predictive model, the next step should be assessing the archaeological and environmental data to determine how the area should be divided for modeling purposes, as well as how different categories of cultural resources should be separated to create different models. Just as various ecological settings may require stratification for sampling purposes, different models may be needed to account for different environmental and behavioral settings. In many locations, separating prehistoric from historic sites is a likely first step. Kvamme (2006:18) makes a similar argument for segregating rock shelters and cliff dwellings from other prehistoric sites, since their locations are largely dependent upon idiosyncratic geological variables rather than human decision-making processes. However, in many cases small rock shelters and caves may have served similar functions as sites with above-ground masonry architecture.

Models should be tested against data that is independent from what was used to create the model. Predictive models should also be tested through sample surveys to assess their accuracy and reliability. When inaccuracies are identified, models should either be rejected or modified to correct the perceived issues, whenever possible. As mentioned previously, the modeling process should be iterative and reflexive, rather than a one-time effort. Use of predictive models should be restricted when confidence in the

249 model is low, but in cases where models have proven to be reliable, complete survey may only be required in areas where sites are predicted to occur. Supplemental survey, perhaps as just a sample, may be recommended to confirm that areas not expected to include archaeological remains are, in fact, devoid of sites. If additional sites are found in these areas, additional survey should be completed in low-probability areas, up to complete survey as deemed appropriate. Models could also be used to assess the reliability of survey results from contract archaeologists.

Although predictive models can be useful tools for prioritizing fieldwork, they should be supplemented by other methods whenever possible, including remote sensing, reconnaissance survey, and interviews with informants who may be familiar with the area in question (e.g., Siart et al. 2008). High-probability areas should be prioritized for survey to fulfill agency requirements under Section 110 of the NHPA. Another important aspect of implementing these models is making the information readily available to land managers. High-probability locations for archaeological sites should be pointed out early in the planning process of various types of projects, such as vegetation management and range improvements. Including predictive models in the planning stages of projects should help reduce unnecessary impacts to sensitive areas and clarify budgetary needs for completing compliance surveys. Predictive models can also play a role in management of emergency situations, such as wildfires, where suppression activities can result in unintended damage or destruction to cultural resources. Finally, the results of the modeling process should be explored as a tool for consulting with Native American tribes, who may be able to provide new insights into the meaning of their results. Their

250 input should help identify new ways of analyzing data through alternative ontologies of natural and cultural resources that are not reflected in current inventories.

GIS modeling of archaeological landscapes should be an ongoing process. A review of Plog’s planning work for the ASNF demonstrates that small sample surveys may identify local patterns in specific areas, but it is problematic to attempt to translate those results to a large geographic scale. Macro- and meso-economic processes can be difficult to assess without large-scale surveys. In a similar sense, the findings presented in this dissertation will be subject to future adjustments. Landscape-scale analysis of site distributions should be an ongoing process that integrates inductive and deductive methods. These studies should also try to incorporate landscape affordances that range beyond the traditional archaeological focus on subsistence and economy, issues that are intimately tied to spiritual and ideological realms among Native communities.

Descendant communities should be given the opportunity to review these findings and offer their own insights to assist in the management of common pool resources in ancestral territories that have been managed in a very different manner since Euro-

American colonization.

Developments in landscape archaeology (e.g., Fowles 2010) have shifted focus from sites to broader scales, and this should aid consultation with Native Americans, provided the significance of these perspectives are communicated within land management agencies. Recognizing palimpsests of archaeological landscapes and relating them to systemic relationships with the environment, although modified by historic uses and climatic variability, is a useful avenue for incorporating indigenous

251 perspectives into planning activities, rather than simply focusing on single sites on a case- by-case basis.

Management Recommendations

Many of the tentative conclusions and hypotheses explored in the preceding chapters provide a useful framework that may guide future interventions at small sites.

This framework can serve as a tool to ensure that recommendations regarding eligibility for nomination to the National Register of Historic Places take into account the full range of possibilities that could contribute to the admittedly amorphous use of “Criterion D,” or in this case, archaeological sites that “possess integrity of location, design, setting, materials, workmanship, feeling, and association and…(d) that have yielded, or may be likely to yield, information important in prehistory or history” (National Register Bulletin

No. 15).

Although potential sources of paleoecological information, such as landforms or paleontological sites, are not usually considered NRHP-eligible in and of themselves, they should be if they have the potential to reveal information regarding the past environment associated with a significant site, especially considering the recent emphasis on climate change. This point underscores another shortcoming of what has become standard practice for most archaeological survey in the American Southwest. Pedestrian survey solely focused upon material culture may overlook important resources, such as packrat middens, alluvial, lacustrine, and other sedimentary deposits, including sources

252 for pollen analysis. Stands of trees should also be considered for their information potential. Old-growth piñon-juniper woodlands can be identified with the aid remote sensing, and dendrochronology can approximate the structure of pre-Colonial settlement conditions. Furthermore, some of these stands may be high-probability locations for identifying Navajo, Apache, and other under-recognized occupations, which are remarkably under-represented in current site inventories.

One may reasonably argue that the same avenues of additional investigation also apply to the use of Criterion A, since they are inherently related to events. Although these events may not conform to singular episodes, they are important trends that have captured the attention of generations of archaeologists, and, perhaps more importantly, they represent a tangible legacy that bridges a disconnect between living populations whose ties to the land were dramatically altered by historical shifts in land tenure systems. Also, the guidelines can be useful for ensuring that testing and mitigation plans for small sites present appropriate research questions and apply appropriate field methods for addressing them. These characteristics are not all intuitively identifiable through standard archaeological survey practices and warrant further consideration.

Beyond the initial documentation of sites, there is an ongoing need to manage them. Eligibility determinations based on surface assemblages alone often dictate whether the location is excluded from project areas, a “strategy” commonly referred to as

“flag and avoid” in a lengthy history of preservation recommendations. When this route is selected in timber sales and other vegetation management projects, it commonly results in the creation of “islands.” Although there also appears to be an “edge effect,” or an

253 association of site locations with project boundaries, that has not yet been fully explored but appears to be associated with the locations of roads and natural landforms. While these modifications to project areas may help to achieve natural resource goals, like the creation of vegetation mosaics intended to mimic perceived “natural” conditions after project implementation, this can also result in dramatic modifications to a site’s surroundings, which could include under-recognized site types such as fields.

The site databases require ongoing attention, and much of the legacy data should be re-evaluated. I have tried to account for errors where possible, but significant and continuing work should be undertaken, and projects with the potential to adversely affect sites should carefully scrutinize decisions based upon legacy data. Perhaps most importantly, the sensitivity of archaeological site location must be acknowledged.

Distribution of site locations and data availability is a critical aspect of operationalizing our knowledge of the archaeological landscape and understanding how it intersects with modern landscape values and uses. Some of the information presented in this dissertation could play a critical role in legal issues, including ancestral land claims, repatriation, historic preservation, and land use histories. These uses highlight the importance of cartographic representations of past archaeological and systemic landscapes, as well as the need to balance effective data management and timely and appropriate distribution of site locations while also protecting sensitive information. I have tried to obscure site location information as much as possible in this dissertation, but recognize the difficulties of emphasizing and communicating the significance of archaeological geographic information. The conundrum of the need to share information to protect and study sites

254 but also balance those issues with necessary confidentiality is a major challenge, particularly as we continue to move through the Information Revolution and our access to geographic information and technology continues to expand.

Land management agencies have encouraged investigation of climate change upon regional landscapes and localized conditions (e.g., Secretary of the Interior Order

No. 3289, 2009). These perspectives have re-emphasized a significant part of land management agencies’ intended purpose to secure the nation’s watersheds, wildlife, land, and important cultural heritage resources. Conservation, education, modified fire suppression practices, and increasing effort toward managing vegetation in ways intended to restore ecological functions have been some of the recent strategies to address climate change. Yet, recent years have witnessed the largest wild fires in recorded history and mounting evidence suggests palpable effects of climate change are manifesting themselves (Nabhan and Coder 2004). “Tribal values are critical to determining what is to be protected, why, and how to protect the interest of their communities” through “use of the best available science, including traditional ecological knowledge” (Secretary of the Interior Order No. 3289, 2009).

Much of the study area has been altered by grazing. Generations have tended cattle and sheep, giving rise to a vast network of fences, modifications to local hydrology, and changes to plant communities, which have been particularly destructive in some cases, especially the widespread land clearing efforts undertaken throughout much of 20th century, known as chaining. Heavy equipment with cutting mechanisms of various types has largely replaced the chain during efforts to restore vast tracks of rangeland invaded

255 by underappreciated plants like juniper and manzanita. Wildfires that would have been suppressed in the past have instead been managed. The carbon footprints of some of these projects are troubling, as are the cost, and landscape-scale projects can impose great strain on under-funded and under-staffed land managers.

A fundamental concept in land management is recognizing the presumption of the natural characteristics of a given ecosystem, as well as the means used to restore it. This dissertation and similar studies offer important considerations for vegetation management programs. “The intuition of archaeologists concerning rates of resource use in prehistory, and the long-term effects of those rates, are probably not very accurate” (Johnson et al.

2005:106). Worldwide forest degradation is an ongoing problem and our efforts to

“restore” it rest upon a number of underlying assumptions. History has shown that past attempts to “restore” or “improve” conditions of ecological systems often result in unintended consequences.

The Mogollon Rim and Chevelon/Chavez site clusters both warrant consideration as a Traditional Cultural Property. Ellis (1974) made the argument that all of Hopitutskwa should be considered a sacred site. Shrines and cinder cones, eagle nest sites, springs, archaeological sites, and other natural landforms have long been recognized by Native

Americans as traditional places, and locations of high site concentrations warrant similar recognition. This should play a role in planning land management activities. Aside from the complications of attempting to mitigate the effects of large-scale timber and other potentially destructive vegetation management projects, this could provide a more sensitive approach to project planning that avoids confrontations and respects Native

256 respect for ancestral remains of both archaeological and ideological significance. These areas should also receive heightened attention when routine road maintenance and improvement projects are completed.

Senior’s consultation and research (2004) after the Rodeo-Chediski fire documented some economic uses of the area by the Hopi, but it is not considered an area to be used for agriculture or other intensive resource extraction. Hopitutskwa “is an area of shrines, sacred natural features, eagle trapping locations and regions where salt is obtained” (Ellis 1974a:8). Use has been made of the region by priests who visit the shrines to perform specific rites, to trap eagles, and to gather herbs, plants, minerals and other materials needed to perform their ceremonies and other rituals (Kuwanwisiwma and

Ferguson 2009; Senior 2004). The southern portion of the study area was also an important territory for hunting, and it was conceived of as an area more suitable for pursuits other than agriculture or grazing (Senior 2004).

Research Questions for Future Investigations at Small Sites

Researchers have recognized that small sites, especially field houses, can help overcome some of the challenges posed by uneven comparative data sets, especially in regards to pollen and macrobotanical studies (Adams and Fish 2006; Adams and Van

West 2005). Given archaeology’s penchant for a long-term perspective, this approach offers interesting avenues to explore the various trajectories of species on both biological and temporal scales. The geographic distributions of field houses and their presumed

257 association with horticulture of domesticates suggest there could be genetic distinctions among varieties of maize and other types of plants.

There are clearly limitations to a landscape-scale study of small sites. New studies focusing on localized conditions and manifestations of specific influences upon horticultural practices warrant further attention. Site records often note potential associations between surface structures and nearby agricultural land, but these areas have rarely been documented, tested, or protected. Although there appear to be correlations with locations near drainages, more detailed studies could reveal patterns from strategies to use these areas for agriculture, including the variability among agricultural features like check dams, terraces, and other water control devices. Flow accumulation surfaces and topographic variability among drainage areas may yet reveal some insights into these matters.

Chronology is an important aspect of studying culture change. Because virtually all of the small sites have been dated through ceramic cross-dating along, and these dates are of highly variable reliability, I have not tried to focus upon fine-grained changes through time, instead relying upon broad-scale changes observed through a long-term accumulation of site data, albeit of highly variable reliability. More detailed studies of ceramic assemblages should provide further avenues for addressing chronologically- sensitive issues in site location and will benefit our understanding of small sites.

There has been remarkably little discussion of the husbandry of, or other anthropogenic influences upon, large and woody plants among the prehistoric societies of the southwestern United States and northwestern Mexico (Doolittle 2000). Instead, focus

258 has been upon fuel wood as another potential tragedy of the commons scenario.

Although I reject impaired soil fertility as a cascading disruption of the ecological systems that attracted people to the area, further research into the compounding influences of fuel wood harvesting could build a stronger argument, particularly if we are predisposed to that conclusion.

Rather than simply comprehending historical pursuits as occurring in opposition to or being determined by environmental factors, landscape archaeology can expose associations and meaning among multiple landscapes and provide insights to the ways that past groups humanized and cognized their surroundings. This also provides opportunities to explore historical landscapes and incorporate insights from archaeology, economics, geography, history, geology, range management, civil engineering, and land surveying, all of which play important roles in our representations of regional archaeological geography.

Ultimately, much of the forest lands seem to have suffered the effects of the tragedies of the commons in more recent times. Nineteenth-century colonizers staked their claims to the prime agricultural land in much of the area. In some regards, archaeological sites themselves became common pool resources. Undocumented excavations at many sites offered recreational opportunities that helped satisfy curiosities about those who came before. In particular instances archaeological sites have become economic opportunities targeted by looters, “digging pots” from the National Forest lands. On the other hand, projects intended to improve habitat conditions and harvest

259 usable products from the forests encounter archaeological sites as “roadblocks” and areas to be avoided, sometimes removing areas from stocks of available timber.

Surveys in advance of fuels projects should recognize the importance of phasing up front, and not only in terms of the scheduling treatments like mechanized thinning and prescribed fires, but also reserving (demanding) funding for subsequent surveys to document previously unrecognized remains and reassess inferences concerning site function. Imperiled archaeological features should receive priority treatments to assess problem-oriented research, including paleoenvironmental reconstruction.

Although I have focused on Native American sites, the results of this research could provide some insights regarding historical archaeology as well. The intersections among prehistoric land use patterns and historical pursuits to extract resources from the mountainous areas of the study area may reveal insights concerning the ways that various social groups depended upon different components of the landscape for their livelihoods, as well as the social meaning ascribed to its attributes. How are they related with modern pursuits to extract similar and different resources from forest environments? How do current values about common pool resources, many of which must have also been important in the past (e.g., hunting, fuel wood, and forage), intersect with different aspects of the archaeological landscape? The breadth of our knowledge of the archaeological geography of the Mogollon Rim region is attributable in large part to archaeological surveys completed in advance of range and timber projects, as well as public infrastructure like roads and utility corridors. This ongoing industrialization and resource extraction and management in the area is the next chapter in how the landscape

260 is transformed and comprehended. Future studies of common pool resources using GIS methods provide opportunities to explore these issues in greater detail to support more harmonious and sustainable use of the landscape.

261

REFERENCES

Abruzzi, William S. 1987 Water and Community Development in the Little Colorado River Basin. Human Ecology 13:241-269. 1993 Ecology and Mormon Settlement in the Little Colorado River Basin. University Press of America, Inc., New York. 1999 Ecology, Resource Distribution, and Mormon Settlement in Northeastern Arizona. American Anthropologist 91:642-655.

Adams, E. Charles 1979 Cold Air Drainage and the Length of Growing Season in the Hopi Mesas Area. Kiva 44:285-296. 1989 Passive Resistance: Hopi Responses to Spanish Contact and Conquest. In Columbian Consequences, Volume I, edited by D. H. Thomas, pp. 77-91. Smithsonian Institution Press, Washington. 2002 Homol’ovi: An Ancient Hopi Settlement Cluster. University of Arizona Press, Tucson.

Alcorta, Marissa 2002 Student paper on the Ceramic Assemblage at the Agave Site. University of Arizona Archaeological Field School. Ms. on file, School of Anthropology, University of Arizona, Tucson.

Aldenderfer, Mark 1996 Introduction. In Anthropology, Space, and Geographic Information Systems, edited by Mark Aldenderfer, and H. Maschner, pp. 3-18. Oxford University Press, New York.

Altschul, Jeffrey H. 1988 Models and the Modeling Process. In Quantifying the Present and Predicting the Past: Theory, Method and Application of Archaeological Predictive Modeling, edited by W. J. Judge and L. Sebastian, pp. 61–96. Government Printing Office, Washington, D.C.

Anderies, John M., Marco A. Janssen, François Bousquet, Juan-Camilo Cardenas, Daniel Castillo, Maria-Claudio Lopez, Robert Tobias, Björn Vollan, Amber Wutich 2011 The Challenge of Understanding Decisions in Experimental Studies of Common Pool Resource Governance. Ecological Economics 70:15711579

Anderson, Edgar 1952 Plants, Man and Life. University of California Press, Berkeley.

Anderson, R. Scott

262

1993 A 35,000 Year Vegetation and Climate History from Potato Lake, Mogollon Rim, Arizona. Quaternary Research 40:351-359.

Anschuetz, Kurt F., Richard Wilshusen, and Cherie Scheick 2001 An Archaeology of Landscapes: Perspectives and Directions. Journal of Archaeological Research 9:157-211.

Anyon, Roger, T.J. Ferguson, and Chip Colwell-Chanthaphonh 2005 Natural Setting as Cultural Landscapes: The Power of Place and Tradition. In Gottfried, Gerald J., Brook Gebow, Lane Eskew, Carleton Edminster (compilers). Connecting Mountain Islands and Desert Seas: Biodiversity and Management of the Madrean Archipelago II. Proc. RMRS-P-36, pp. 273-276. Department of Agriculture, Forest Service, Rocky Mountain Research Station, Fort Collins.

Arkush, Elizabeth N., and Mark W. Allen (editors) 2006 The Archaeology of Warfare: Prehistories of Raiding and Conquest. University of Florida Press, Gainesville.

Aston, Michael A. 1985 Interpreting the Landscape: Landscape Archaeology in Local Studies. Batsford, London

Baker, Robert D., Robert S. Maxwell, Victor H. Treat, and Henry C. Dethloff. 1988 Timeless Heritage: A History of the Forest Service in the Southwest. United States Department of Agriculture. Intaglio, Inc., College Station.

Baker, William L., and Douglas J. Shinneman 2004 Fire and Restoration of Piñon-Juniper Woodlands in the Western United States: A Review. Forest Ecology and Management 189:1-21.

Bandelier, Adolph F. 1884 Report by A. F. Bandelier on His Investigations in New Mexico during the years 1883-1884. Fifth Annual Report of the Executive Committee, pp. 55-98. Archaeological Institute of America.

Barnes, Benjamin 2004 A Predictive Model of Archaeological Site Locations on the Sitgreaves National Forest. Manuscript on file, Apache-Sitgreaves National Forests Supervisor’s Office, Springerville, AZ.

Barton, C. Michael, J. Bernabeu, J.Emily Aura, Oreto Garcia, Steven Schmich, and Lluis Molina 2004 Long-term Socioecology and Contingent Landscapes. Journal of Archaeological Method and Theory 11(3):253-295.

263

Basso, Keith H. 1996 Wisdom Sits in Places: Landscape and Language Among the Western Apache. University of New Mexico Press, Albuquerque.

Bauder, Harald, and Salvatore Engel-Di Mauro (editors) 2008 Critical Geographies: A Collection of Readings. Praxis (e)Press. Electronic document: http://www.praxis-epress.org/CGR/fpages.pdf, accessed October 1, 2012.

Bayman, James M. 2007 Papaguerian Perspectives on Economy and Society in the Sonoran Desert. In Hinterlands and Regional Dynamics in the Ancient Southwest, edited by Alan P. Sullivan III and James M. Bayman, pp. 109-124. The University of Arizona Press, Tucson.

Bayman, James M., and Alan P. Sullivan III 2008 Property, Identity, and Macroeconomy in the Prehispanic Southwest. American Anthropologist 10(1):6-20.

Beaglehole, Earnest 1937 Notes on Hopi Economic Life. Yale University Publications in Anthropology 15. Yale University Press, New Haven.

Bender, Barbara 1992 Theorising Landscapes, and the Prehistoric Landscapes of Stonehenge. Man 27:735-755.

Bernardini, W. 1996 Transitions in Social Organization: A Predictive Model from Southwestern Archaeology. Journal of Anthropological Archaeology 15:372-402.

Berry, Michael S. 1984 Sampling and Predictive Modeling on Federal Lands. American Antiquity 49:842- 853

Billman, B. R., P. M. Lambert, and B. L. Leonard 2000 Cannibalism, Warfare, and Drought in the Mesa Verde Region During the Twelfth Century A.D. American Antiquity 65:145-178.

Bintliff, John and Anthony Snodgrass 1988 Off-Site Pottery Distributions: A Regional and Interregional Perspective. Current Anthropology 29(3):506-513.

Blaszczynski, Jacek S.

264

1997 Landform Characterization with Geographic Information Systems. Photogrammetric Engineering & Remote Sensing 63(2):183-191

Blinman, Eric 2008 2000 Years of Cultural Adaptation to Climate Change in the Southwestern United States. AMBIO: A Journal of the Human Environment 37(sp14):489-497.

Bowser, Brenda J., and Maria N. Zedeño (editors) 2009 The Archaeology of Meaningful Places. University of Utah Press, Salt Lake City.

Bradley, Richard 1991 Rock Art and Perception of Landscape. Cambridge Archaeological Journal 1:77- 101.

Bradley, Richard, Felipe C. Boado, and Ramón F. Valcarce 1994 Rock Art Research as Landscape Archaeology: A Pilot Study in Galicia, North- west Spain. World Archaeology 25:374-390.

Branton, Nicole 2009 Landscape Approaches in Historical Archaeology: The Archaeology of Places. In International Handbook of Historical Archaeology, edited by Teresita Majewski and David Gaimster, pp. 51-65.

Brown, Gregory B. 2001 Archaeological Data Recovery at AZ P:10:118 (ASM), West of Heber, Navajo County, Arizona. Logan Simpson Design, Tempe.

Brown, David E., and Charles H. Lowe 1980 Map of Biotic Communities of the Southwest. General Technical Report RM-78. U.S. Department of Agriculture, Forest Service, Rocky Mountain Forest and Range Experiment Station, Tempe.

Burton, James H. 1993 Provenance determination of obsidian artifacts from Shelltown by back-scattered electron microscopy. In Shelltown and the Hind Site: A Study of Two Hohokam Craftsman Communities in Southwestern Arizona, edited by William S. Marmaduke and Richard J. Martynec, pp. B1-B11. Northland Research, Inc., Flagstaff, v.2.

Butlin, Robin A. 1993 Historical Geography through the Gates of Space and Time. Edward Arnold, New York.

Butzer, Karl W.

265

2002 Archaeology as Human Ecology: Method and Theory for a Contextual Approach. Cambridge University Press, Cambridge.

Cameron, Catherine M. 1995 Migration and the Movement of Southwestern Peoples. Journal of Anthropological Archaeology 14:104-124

Cano, Jenny 2003 Evaluating the Role of Surface Assemblages in Defining Occupational Histories in Bryant Ranch. Unpublished Master's Thesis, Department of Anthropology, The University of Arizona, Tucson.

Carleton, W. Chris, James Conolly, and Gyles Ianonne 2012 A locally-adaptive model of archaeological potential (LAMAP). Journal of Archaeological Science 39:3371-3385.

Chakrabarti, Dilip K. 2001 Archaeological Geography of the Ganga Plain: The Lower and Middle Ganga. Permanent Black, Delhi.

Church, Minette C. 2002 The Grant and the Grid: Homestead Landscapes in the Nineteenth-century Borderlands of Southern Colorado. Journal of Social Archaeology 2:220-244.

Church, Tim, R. Joe Brandon, and Galen R. Burgett 2000 GIS Applications in Archaeology: Method in Search of Theory. In Practical Applications of GIS for Archaeologists: A Predictive Modeling Toolkit, edited by Konnie L. Wescott and R. Joe Brandon, pp. 135-155. Taylor & Francis, Philadelphia.

Ciolek-Torrello, Richard 1981 Archaeological Investigations, Apache-Sitgreaves National Forests: Archaeological Survey of the Stott Timber Sale, Pinedale Ranger District, Navajo County, Arizona. Museum of Northern Arizona, Department of Anthropology, Flagstaff.

Ciolek-Torrello, Richard, Steven D. Shelley, Jeffrey H. Altschul, and John Welch 1990 Roosevelt Rural Sites Study, Volume 1: Research Design. Statistical Research Technical Series No. 28. Statistical Research Inc., Tucson.

Ciolek-Torrello, Richard, Steven D. Shelley, and Su Benaron (editors) 1994 The Roosevelt Rural Sites Study Volume 2: Prehistoric Rural Settlements in the Tonto Basin, Part 1. Statistical Research Technical Series No. 28. Statistical Research, Inc., Tucson.

Clay, Vickie L., Kenneth L. Petersen, and Kenneth W. Decker

266

1985 Sediment and Chemical Analyses of Soil Conservation Service Designated Soils in the Dolores Project Area, Southwest Colorado. Technical Reports DAP-207. Dolores Archaeological Program, Dolores, Colorado. Ms. submitted to U.S. Bureau of Reclamation, Upper Colorado Region, Salt Lake City.

Collins, Timothy W. 2005 The Production of Hazard Vulnerability: The Case of People, Forests, and Fire in Arizona's White Mountains. Unpublished Ph.D. dissertation, Department of Geography, Arizona State University, Tempe.

Conolly, James, and Mark Lake 2006 Geographical Information Systems in Archaeology. Cambridge University Press, New York.

Cordell, Linda S. 1997 Archaeology of the Southwest. Academic Press, San Diego.

Crown, Patricia L. 1983 Introduction: Field Houses and Farmsteads in South-Central Arizona. In Hohokam Archaeology Along the Salt-Gila Aqueduct, Central Arizona Project, Vol. 5: Small Habitations on Queen Creek, edited by Lynn S. Teague and Patricia L. Crown, pp. 3-22. Arizona State Museum Archaeological Series No. 150, Arizona State Museum, Tucson.

Cummings, Ellen 1971 The Roubicek Site: ARIZ. P:11:6. Unpublished Honor’s thesis. Ms. on file, Arizona State Museum, Tucson.

Cushing, Frank H. 1974 [1920] Zuñi Breadstuff. In Indian Notes and Monographs, Vol. 8. Museum of the American Indian, Heye Foundation, New York. 1979 Zuñi: Selected Writings of Frank Hamilton Cushing. Edited by Jesse Green. University of Nebraska Press, Lincoln.

Damp, Jonathan, Stephen Hall, and Susan Smith 2002 Early Irrigation on the Colorado Plateau near Zuni Pueblo, New Mexico. American Antiquity 67: 665-676

Darby, Henry C. 1936 An Historical Geography of England before A.D. 1800. Cambridge University Press, Cambridge

Darby, Henry C. (editor)

267

1976 A New Historical Geography of England before AD 1600. The University of Cambridge Press, Cambridge.

David, Bruno, and Julian Thomas (editors) 2008 Handbook of Landscape Archaeology, Left Coast Press, Walnut Creek.

Dean, Jeffrey S. 1988 A Model of Anasazi Behavioral Adaptation. In The Anasazi in a Changing Environment, edited by G. J. Gumerman, pp. 25-44. Cambridge University Press, Cambridge. 1996 Demography, Environment, and Subsistence Stress. In Evolving Complexity and Environmental Risk in the Prehistoric Southwest, edited by Joseph A. Tainter and Bonnie Bagley Tainter, pp. 25-56. Santa Fe Institue Studies in the Sciences of Complexity Volume XXIV. Addison-Wesley Publishing Company, Reading. 2010 The Environmental, Demographic, and Behavioral Context of the Thirteenth- Century Depopulation of the Northern Southwest. In Leaving Mesa Verde: Peril and Change in the Thirteenth-Century Southwest, edited by Timothy A. Kohler, Mark D. Varien, and Aaron M. Wright, pp. 324-345. The University of Arizona Press, Tucson.

Dean, Jeffrey S., and Gary S. Funkhouser 1995 Dendroclimatic Reconstructions for the Southern Colorado Plateau. In Climate Change in the Four Corners and Adjacent Regions: Implications for Environmental Restoration and Land-Use Planning, pp. 85-104. Proceedings of NTIS Conference 9409325. Grand Junction.

Diehl, Michael W., and Sarah Herr 2011 The Autumn of Foraging in the Little Green Valley. Kiva 76(3):317-342.

Dixon, E. James, Manley, William F., and Lee, Craig M. 2005 The Emerging Archaeology of Glaciers and Ice Patches: Examples from Alaska’s Wrangell-St. Elias National Park and Preserve. American Antiquity 70: 129–143.

Donaldson, Bruce R. 1975 An Archaeological Sample of the White Mountain Planning Unit, Apache- Sitgreaves National Forest, Arizona. Archaeological Report No. 6, USDA Forest Service, Southwestern Region, Albuquerque.

Doolittle, William E. 1988 Intermittent Use and Agricultural Change on Marginal Lands: The Case of Smallholders in Eastern Sonora, Mexico. Geografiska Annaler. Series B, Human Geography, Vol. 70(2):255-266. 2000 Cultivated Landscapes of Native North America. Oxford University Press, New York.

268

Doolittle, William E., and James A. Neely 2004 The Safford Valley Grids: Prehistoric Cultivation in the Southern Arizona Desert. Anthropological Papers of the University of Arizona Number 70. The University of Arizona Press, Tucson.

Dosh, Steven G. 1988 Subsistence and Settlement Along the Mogollon Rim A.D. 1000-1150. MNA Research Paper 39. The Museum of Northern Arizona Press, Flagstaff. 1989 Ada Mesa Site: Archaeology of Upper Coyote Creek, Apache County, Arizona. Prepared for Apache-Sitgreaves National Forests. Project No. R89-23-01. Northland Research, Inc., Flagstaff.

Doyel, David E. 1993 Prehistoric Non-Irrigated Agriculture in Arizona. Estrella Cultural Research Paper No. 3. Arizona State Historic Preservation Office, Phoenix.

Doyel, David E., and Sharon Debowski 1980 Prehistory in the Dead Valley, East-Central Arizona: The TG&E Springerville Report. Arizona State Museum Archaeological Series 144, Tucson.

Duff, Andrew I., Karen R. Adams, and Susan C. Ryan 2010 The Impact of Long-Term Residential Occupation of Community Centers on Local Plant and Animal Resources. In Leaving Mesa Verde: Peril and Change in the Thirteenth-Century Southwest, edited by Timothy A. Kohler, Mark D. Varien, and Aaron M. Wright, pp. 156-179. Amerind Studies in Archaeology, v. 5. University of Arizona Press, Tucson.

Duke, Christopher, and James Steele 2010 Geology and Lithic Procurement in Upper Paleolithic Europe: A Weights-of- Evidence Based GIS Model of Lithic Resource Potential. Journal of Archaeological Science 37:813-824.

Duwe, Samuel Gregg 2005 Communities of Practice and Ancient Apprenticeship in the American Southwest: Pigment Analyses of Pueblo IV Period Ceramics from Bailey Ruin, East Central Arizona. Master's Report, Department of Anthropology, The University of Arizona.

Earle, Timothy, and David E. Doyel 2008 The Engineered Landscapes of Irrigation. In Economies and the Transformation of Landscape, edited by Lisa Cliggett and Christopher A. Pool, pp. 19-46. AltaMira Press, Lanham, MD.

Ebert, James I.

269

1988 Remote Sensing in Archaeological Prospection and Prediction. In Quantifying the Present and Predicting the Past: Theory, Method, and the Application of Archaeological Predictive Modeling, edited by W. James Judge and Lynne Sebastian, pp. 429-492. U.S. Department of the Interior, Bureau of Land Management Service Center, Denver. 2000 The State of the Art in “Inductive” Predictive Modelling: Seven Big Mistakes (and Lots of Smaller Ones). In Practical Applications of GIS for Archaeologists: A Predictive Modeling Toolkit, edited by Konnie L. Wescott and R. Joe Brandon, pp. 129- 134. Taylor & Francis, Philadelphia.

EcoPlan 2012 Cultural Resources Overview and Management Plan. Draft submitted to the Apache-Sitgreaves National Forests. Manuscript on file, ASNF Supervisor’s Office, Springerville.

Ellis, Florence Hawley 1951 Patterns of Aggression and the War Cult in Southwestern Pueblos. Southwestern Journal of Anthropology 7(2):177-201. 1974 The Hopi: Their History and Use of Lands, pp. 25-278, in Hopi Indians. Garland Press. 1978 Small Structures Used by Historic Pueblo Peoples and Their Immediate Ancestors. In Limited Activity and Occupation Sites: A Collection of Conference Papers, edited by Albert W. Ward, pp. 59-68. Contributions to Anthropological Studies No. 1. Center for Anthropological Studies, Albuquerque, New Mexico.

Ensor, Bradley E. 2000 Social Formations, Modo de Vida, and Conflict in Archaeology. American Antiquity 65:15-42.

Ensor, Bradley E., Marisa Ensor, and Gregory W. De Vries 2003 Hohokam Political Ecology and Vulnerability: Comments on Waters and Ravesloot. American Antiquity 68:169-181.

Euler, Robert C. and George J. Gumerman (Editors) 1978 Investigations of the Southwestern Anthropological Research Group: An Experiment in Archaeological Cooperation. Museum of Northern Arizona, Flagstaff.

Farmer, James D. 1997 Iconographic Evidence of Basketmaker Warfare and Human Sacrifice: A Contextual Approach to Early Anasazi Art. Kiva 62:391-420.

Farmer, M. F. 1957 A Suggested Typology of Defensive Systems in the Southwest:. Southwestern Journal of Anthropology 13:249-266.

270

Ferguson, T. J. 1985 Patterns of Land Use and Environmental Change on the Zuni Indian Reservation, 1846-1985: Ethnohistorical and Archaeological Evidence. Expert testimony submitted to the United States Claims Court as evidence in the case of Zuni Indian Tribe v. United States, Docket 327-81L. 1996 Historic Zuni Architecture and Society: An Archaeological Application of Space Syntax. Anthropological Papers of the University of Arizona No. 60. University of Arizona Press, Tucson. 2007 Zuni Traditional History and Cultural Geography. In Zuni Origins: Toward a New Synthesis of Southwestern Archaeology, edited by David A. Gregory and David R. Wilcox, pp. 377-406. University of Arizona Press, Tucson.

Ferguson, T. J., and E. Richard Hart 1985 A Zuni Atlas. University of Oklahoma Press, Norman.

Fertelmes, Craig, and C. Michael Barton 2007 Using Remote Sensing to Assess the Impact of Prehistoric Agriculture on Modern-Day Vegetation Cover in the U.S. Southwest. SAS Bulletin 30(1):11-15.

Fewkes, Jesse W. 1904 Two Summer’s Work in Pueblo Ruins. Twenty-second Annual Reports of the Bureau of American Ethnology. U.S. Government Printing Office, Washington, D.C.

Finke, Peter A., E. Meylemans, and Johan Van de Wauw 2008 Mapping the Possible Occurrence of Archaeological Sites by Bayesian Inference. Journal of Archaeological Science 35:2786-2796.

Fish, Paul R., and Suzanne K. Fish 1978 The Small House in a Social Network: Some Mesoamerican Examples. In Limited Activity and Occupation Sites: A Collection of Conference Papers, edited by Albert E. Ward, pp. 17-23. Contributions to Anthropological Studies No. 1. Center for Anthropological Studies, Albuquerque, New Mexico. 1984 Agricultural Maximization in the Sacred Mountain Basin. In Prehistoric Agricultural Strategies in the Southwest, edited by Suzanne K. Fish and Paul R. Fish. Anthropological Research Papers No. 33. Arizona State University, Tempe.

Floyd, Lisa M. and Timothy A. Kohler 1990 Current Productivity and Prehistoric Use of Piñon (Pinus edulis, Pinaceae) in the Dolores Project Area, Southwestern Colorado. Economic Botany 44:141-156

Ford, Richard I.

271

1985 Zuni Land Use and Damage to Trust Land. Expert testimony submitted to the United States Claims Court as evidence in the case of Zuni Indian Tribe v. United States, Docket 327-81L.

Fowles, Severin M. 2009 The Enshrined Pueblo: Villagescape and Cosmos in the Northern Rio Grande. American Antiquity 74(3):448-466. 2010 The Southwest School of Landscape Archaeology. Annual Review of Anthropology 39:453-468.

Franke, Paul R., and Don Watson 1936 An Experimental Corn Field in Mesa Verde National Park. In Symposium on Prehistoric Agriculture, University of New Mexico Bulletin, Anthropological Series, Vol. 1, No. 5, pp. 35-41. University of New Mexico, Albuquerque.

Gabler, Brandon M. 2009 Panarchy on the Plateau: Modeling Prehistoric Settlement Pattern, Land Use, and Demographic Change on the Pajarito Plateau, New Mexico. Unpublished Ph.D. dissertation, Department of Anthropology, University of Arizona, Tucson.

Gartner, William G. 1999 Late Woodland Landscapes of Wisconsin: Ridged Fields, Effigy Mounds, and Territoriality. Antiquity 73:671-683.

Gillio, David “A” (Compiler) 2005 Flagging the Trail: One Hundred Years of Managing Cultural Resources. USDA Forest Service, Southwestern Region. Heritage Resources Management Report No. 17. Albuquerque.

Gillings, Mark 2012 Landscape Phenomenology, GIS and the Role of Affordance. Journal of Archaeological Method and Theory. 19(4):601-611.

Gillings, Mark, and David Wheatley 2005 Geographic Information Systems. In Handbook of Archaeological Methods, edited by Herbert D. G. Mashner and Christopher Chippindale, pp. 373-422. AltaMira Press, Lanham, MD.

Goodwin, Grenville 1942 The Social Organization of the Western Apache. University of Chicago Press, Chicago.

Graves, Dorothy 2011 The Use of Predictive Modelling to Target Neolithic Settlement and Occupation

272

Activity in Mainland Scotland. Journal of Archaeological Science 38:633-656.

Graves McEwan, Dorothy 2012 Qualitative Landscape Theories and Archaeological Predictive Modelling - A Journey Through No Man’s Land? Journal of Archaeological Method and Theory 19:526-547.

Grebinger, Paul, and Bruce Bradley 1969 Excavations at a Prehistoric Camp Site on the Mogollon Rim, East Central Arizona. Kiva 34:109-123.

Greenwood, Ned H. 1983 High-Altitude and Non-Village Ceremonialism of the Cibola Branch of the Anasazi. Report on file, Apache-Sitgreaves National Forests Supervisors Office, Springerville.

Greenwood, Ned H., and C. W. White 1970 Mogollon Ritual: A Spatial Configuration of a Non-Village Pattern. Archaeology 25(4):298-301.

Gregory, David A. 1975 Defining Variability in Prehistoric Settlement Morphology. In Chapters in the Prehistory of Eastern Arizona, IV. Fieldiana Anthropology 65:40-46. 1990 Sand Draw Area Cultural Resource Survey, Chevelon Ranger District, Apache- Sitgreaves National Forests, Coconino County, Arizona. DAG Enterprises. ASNF Report No 1991-01-012A. On file, Springerville, AZ.

Haas, Jonathan 1990 Warfare and Tribalization in the Prehistoric Southwest. In The Anthropology of War, edited by J. Haas, pp. 171-189. Cambridge University Press, Cambridge. 2001 Warfare and the Evolution of Culture. In Archaeology at the Millennium: A Sourcebook, edited by Gary Feinman and T. Douglas Price, pp. 329-350. Kluwer Academic/Plenum, New York.

Haas, Jonathan, and Winifred Creamer 1993 Stress and Warfare among the Kayenta Anasazi of the Thirteenth Century A.D. Fieldiana Anthropology No. 21. Field Museum of Natural History, Chicago.

Haase, William R. 1983 Pueblo II and Pueblo III Settlement Patterns on Cedar Mesa, Southeastern Utah. Unpublished Master’s thesis, Washington State University, Pullman.

Hack, John T.

273

1942 The Changing Physical Environment of the Hopi Indians of Arizona. Papers of the Peabody Museum of American Archaeology and Ethnology, Vol. 35, No. 1. Harvard University, Cambridge, Massachusetts.

Haines, Jeremy (editor) 2004 Draft Results of Archaeological Survey for the Rodeo-Chediski Environmental Impact Statement. On file, Apache-Sitgreaves National Forests Supervisor’s Office.

Hammack, Laurens C. 1969 Highway Salvage Archaeology in the Forestdale Valley, Arizona. Kiva 34(2- 3):58-89.

Hard, Robert J. and William L. Merrill 1992 Mobile Agriculturalists and the Emergence of Sedentism: Perspectives from Northern Mexico. American Anthropologist 94:601-620.

Hardin, Garrett 1968 The Tragedy of the Commons. Science 162:1243-1248.

Harris, Trevor M. 2006 Scale as Artifact: GIS, Ecological Fallacy, and Archaeological Analysis. In Confronting Scale in Archaeology: Issues of Theory and Practice, edited by Gary Lock and Brian Molyneaux, pp.39-53. Springer, New York

Hantman, Jeffrey L. 1983 Social Networks and Stylistic Distributions in the Prehistoric Plateau Southwest, Unpublished PhD dissertation, Department of Anthropology, Arizona State University, Tempe. 1989 Surplus Production and Complexity in the Upper Little Colorado Province, East- Central Arizona. In The Sociopolitical Structure of Prehistoric Southwestern Societies, edited by Steadman Upham, Kent G. Lightfoot, and Roberta A. Jewett, pp. 363-388. Westview Press, Boulder.

Hartman, Dana (editor) 1990 Small Site Utilization Along the Mogollon Rim. MNA Research Paper 41. The Museum of Northern Arizona Press, Flagstaff.

Haury, Emil W. 1962 HH-39: Recollections of a Dramatic Moment in Southwestern Archaeology. Tree- Ring Bulletin 74:11-14. 1985 Mogollon Culture in the Forestdale Valley, East-Central Arizona. University of Arizona Press, Tucson.

Haury, Emil and E.B. Sayles

274

1985 An Early Pit House Village of the Mogollon Culture, Forestdale Valley, Arizona. In Mogollon Culture in the Forestdale Valley, East-Central Arizona, edited by Emil W. Haury, pp.281-372. The University of Arizona Press, Tucson.

Hays-Gilpin, Kelley, and Eric van Hartesveldt 1998 Prehistoric Ceramics of the Puerco Valley, Arizona: The 1995 Chambers-Sanders Trust Lands Ceramic Conference. Museum of Northern Arizona Ceramic Series No. 7. Museum of Northern Arizona, Flagstaff.

Hegmon Michelle 2000 The Archaeology of Regional Interaction: Religion, Warfare, and Exchange Across the American Southwest. University of Colorado Press, Boulder.

Hegmon, Michelle, Margaret C. Nelson, Roger Anyon, Darell Creel, Stephen A. Leblanc, and Harry J. Shafer 1999 Scale and Time-Space Systematics in the Post-A.D. 1100 Mimbres Region of the North American Southwest. Kiva 65:143-166.

Hegmon, Michelle, Margaret C. Nelson, Karen G. Schollmeyer, Michael Elliott, and Michael Diehl 2006 Agriculture, Mobility, and Human Impact in the Mimbres Region of the United States Southwest. In Managing Archaeological Data and Databases: Essays in Honor of Sylvia W. Gaines, edited by Jeffery L. Hantman and Rachel Most, pp. 107-121. Archaeological Research Paper No.57, Arizona State University, Tempe.

Heilen, Michael P. 2005 An Archaeological Theory of Landscapes. Unpublished Ph.D. dissertation, Department of Anthropology, University of Arizona, Tucson.

Heilen, Michael P., and J. Jefferson Reid 2009 A Landscape of Gamble and Guts: Commodification of Land on the Arizona Frontier. In The Archaeology of Meaningful Places, edited by Brenda Bowser and M.Nieves Zedeño, pp. 132–148. University of Utah Press, Salt Lake City, Utah.

Heilen, Michael P., J. Jefferson Reid, and Michael B. Schiffer 2008 Landscape Formation Processes. In Handbook of Landscape Archaeology, edited by B. David and J. Thomas, pp. 601-608. Left Coast Press, Walnut Creek, California

Henderson, T. Kathleen 1979 Archaeological Survey at Chavez Pass Ruin, Coconino National Forest, Arizona: The 1978 Field Season. Ms. on file, USDA Forest Service, Flagstaff.

Henrich, J., R. Boyd, S. Bowles, C. Camerer, E. Fehr, H. Gintis (editors)

275

2004 Foundations of Human Sociality: Economic Experiments and Ethnographic Evidence from Fifteen Small-Scale Societies. Oxford University Press, Oxford

Herr, Sarah A. 1999 Organization of Migrants on a Pueblo Frontier. Unpublished Ph.D. dissertation, Department of Anthropology, The University of Arizona. 2001 Beyond Chaco: Great Kiva Communities on the Mogollon Rim Frontier. Anthropological Papers of the University of Arizona No. 66. The University of Arizona Press, Tucson. 2002 Frontier Communities of the Mogollon Rim. In Boundaries and Territories: Prehistory of the U.S. Southwest and Northern Mexico, edited by M. Elisa Villalpando, pp. 83-94. Arizona State University Anthropological Research Papers 54. Arizona State University, Tempe.

Herr, Sarah, Elizabeth Perry, and Scott Van Keuren 1999 Excavations at Three Great Kiva Sites. In Living on the Edge of the Rim: Excavations and Analysis of the Silver Creek Archaeological Research Project, 1993- 1998. Arizona State Museum Archaeological Series 192. Arizona State Museum, Tucson.

Hicks, Ronald 2002 Ways of Inhabiting the World: Landscape Archaeology. American Anthropologist 104:315-320.

Hill, Jane H. 2004 Two Styles for Language and Social Identity among the Tohono O’odham. In Identity, Feasting, and the Archaeology of the Greater Southwest, edited by Barbara J. Mills, pp. 124-138. The University Press of Colorado, Boulder.

Hill, J. Brett, Jeffery J. Clark, William H. Doelle, and Patrick D. Lyons 2004 Prehistoric Demography in the Southwest: Migration, Coalescence, and Hohokam Population Decline. American Antiquity 69:689-716.

Hill, Paul, and Julie Wileman 2002 Landscapes of War: The Archaeology of Aggression and Defense. Tempus Publishing Inc., Charleston.

Hodder, Ian, and Clive Orton 1976 Spatial Analysis in Archaeology. Cambridge University Press, Cambridge

Holliday, Vance T. 2004 Soils in Archaeological Research. Oxford University Press, New York.

Homburg, Jeffrey A.

276

1992 Soil Fertility Study. In Archaeological Investigations at Lee Canyon: Kayenta Anasazi Farmsteads in the upper Basin, Coconino County, Arizona, edited by Stephanie M. Whittlesey, pp. 145-161. Technical Series No. 38. SRI Press, Tucson.

Homburg, Jeffrey A., and Jonathan A. Sandor 2010 Anthropogenic Effects on Soil Quality of Ancient Agricultural Systems of the American Southwest. Catena 85:144-154.

Homburg, Jeffrey A., Jonathan A. Sandor, and Dale R. Lightfoot 2004 Soil Investigations. In The Safford Valley Grids: Prehistoric Cultivation in the Southern Arizona Desert, edited by William E. Dolittle and James A. Neely, pp.62-78. Anthropological Papers No. 70. University of Arizona, Tucson.

Hough, Walter W. 1903 Archaeological Field Work in Northeastern Arizona, the Museum-Gates Expedition of 1901. In Annual Report of the U.S. National Museum, 1901, pp. 279-358. Government Printing Office, Washington, D.C.

Huckell, Lisa W. 1999 Paleoethnobotany. In Living on the Edge of the Rim: Excavations and Analysis of the Silver Creek Archaeological Research Project 1993-1998, edited by Barbara J. Mills, Sarah A. Herr, and Scott Van Kuren, pp. 459-504. Arizona State Museum Archaeological Series 192. University of Arizona, Tucson.

Huntsinger, Lynn, and Sarah McCaffrey 1995 A Forest for the Trees: Forest Management and the Yurok Environment, 1850 to 1994. American Indian Culture and Research Journal 19(4):155-192.

Isendahl, Christian 2008 Applied Agro-Archaeological Research in the Bolivian Yungas. The SAA Archaeological Record 8(3):21-27.

Johnson, C. David, Timothy A. Kohler, and Jason Cowan 2005 Modeling Historical Ecology, Thinking about Contemporary Systems. American Anthropologist 107(1):96-107.

Kaldahl, Eric J., and Jeffrey S. Dean 1999 Climate, Vegetation, and Dendrochronology. In Living on the Edge of the Rim: Excavations and Analysis of the Silver Creek Archaeological Research Project 1993- 1998, edited by Barbara J. Mills, Sarah A. Herr, and Scott Van Kuren, pp. 11-29. Arizona State Museum Archaeological Series 192. University of Arizona, Tucson.

Kaldahl, Eric J., Scott Van Keuren, and Barbara J. Mills 2004 Migration, Factionalism, and the Trajectories of Pueblo IV Period Clusters in the

277

Mogollon Rim Region. In The Protohistoric Pueblo World, A.D. 1275-1600, edited by E. Charles Adams and Andrew I. Duff, pp. 85-94. University of Arizona Press, Tucson

Kamermans, Hans, Martijn van Leusen, and Phillip Verhagen 2009 Archaeological Prediction and Risk Management. Leiden University Press, Leiden, Netherlands.

Kamp, Kathryn, and John Whittaker 1999 Surviving Adversity: The Sinagua of Lizard Man Village. University of Utah Anthropological Papers, n°. 120. University of Utah Press, Salt Lake City.

Keeley, Lawrence H. 1996 War Before Civilization. Oxford University Press, New York. 2001 Giving War a Chance. In Deadly Landscapes: Case Studies in Prehistoric Southwestern Warfare, edited by Glen E Rice and Steven A. LeBlanc, pp. 331-342. University of Utah Press, Salt Lake City

Keeley, Lawrence H., Marisa Fontana, and Russel Quick 2007 Baffles and Bastions: The Universal Features of Fortifications. Journal of Archaeological Research 15:55-95.

Kelly, Robert L. 1995 The Foraging Spectrum: Diversity in Hunter–Gatherer Lifeways. Smithsonian Institution Press, Washington, DC.

Kelso, William M., and Rachel Most (editors) 1990 Earth Patterns: Essays in Landscape Archaeology. University Press of Virginia, Charlottesville.

Kenzle, S. C. 1997 Enclosing Walls in the Northern San Juan: Sociophysical Boundaries and Defensive Fortifications in the American Southwest. Journal of Field Archaeology 24:195-210

Kidder, Alfred V. 1962[1924] An Introduction to the Study of Southwestern Archaeology. Yale University Press, New Haven

Kimball, Abigail R. 2007 Memorandum entitled Climate Change; Water; and Kids. Sent to all employees of the USDA Forest Service, October 11th.

Knapp, A. Bernard, and Wendy Ashmore 1999 Archaeological Landscapes: Constructed, Conceptualized, Ideational. In

278

Archaeologies of Landscape: Contemporary Perspectives, edited by Wendy Ashmore and A. Bernard Knapp, pp. 1-30. Blackwell Publishers, Malden, Massachusetts.

Kohler, Timothy A. 1992a Prehistoric Human Impact on the Environment in the Upland North American Southwest. Population and Environment: A Journal of Interdisciplinary Studies 13:255- 268. 1992b Field Houses, Villages, and the Tragedy of the Commons in the Early Northern Anasazi Southwest. American Antiquity 57:617-635.

Kohler, Timothy A. and Meredith Matthews 1988 Long-Term Anasazi Land Use and Forest Reduction – A Case Study from Southwest Colorado. American Antiquity 53(3):537-564.

Kohler, Timothy A., and Sandra C. Parker 1986 Predictive Models for Archaeological Resource Location. Advances in Archaeological Method and Theory 9:397-452.

Kohler, Timothy A., and K. K. Turner 2006 Raiding for Women in the Pre-Hispanic Northern Pueblo Southwest? A Pilot Examination. Current Anthropology 47:1035-1045.

Kolb, Michael J. and James E. Snead 1997 It’s a Small World after All: Comparative Analyses of Community Organization in Archaeology. American Antiquity 62:609-628

Kowalewski, Stephen A. 2008 Regional Settlement Pattern Studies. Journal of Archaeological Research 16:225- 285.

Koyiyumptewa, Bruce K. 1993 Spiritual Values of the Piñon-Juniper Woodland: A Hopi Perspective. In Managing Piñon-Juniper Ecosystems for Sustainability and Social Needs; Proceedings of the symposium 1993 April 26-30; Santa Fe, New Mexico. Gen. Tech. Rep. RM-236, coordinated by Earl Aldon and Douglas Shaw. Fort Collins, CO: USDA, Forest Service, Rocky Mountain Forest and Range Experiment Station.

Kruse, Melissa 2007 The Agricultural Landscape of Perry Mesa: Modeling Residential Site Location in Relation to Arable Land. Kiva 73(1):85-102.

Kuckelman, Kristin A. (editor) 2000 The Archaeology of Castle Rock Pueblo: A Thirteenth Century Village in Southwestern Colorado. Electronic document http://www.crowcanyon.org/castlerock.

279

Accessed online April 5, 2011.

Kuckelman, Kristin A. 2010a The Depopulation of Sand Canyon Pueblo, A Large Ancestral Pueblo Village in Southwestern Colorado. American Antiquity 75(3):497-525. 2010b Catalysts of the Thirteenth-Century Depopulation of Sand Canyon Pueblo and the Central Mesa Verde Region. In Leaving Mesa Verde: Peril and Change in the Thirteenth-Century Southwest, edited by Timothy A. Kohler, Mark D. Varien, and Aaron M. Wright, pp. 180-199. The University of Arizona Press, Tucson.

Kulisheck, Jeremy 2003 Pueblo Population Movements, Abandonments, and Settlement Change in Sixteenth and Seventeenth Century New Mexico. Kiva 69(1):30-54. 2005 The Archaeology of Pueblo Population Change on the Jemez Plateau, A.D. 1200 to 1700: The Effects of Spanish Contact and Conquest. Unpublished Ph.D. dissertation, Southern Methodist University, Dallas.

Kunen, Julie L. n.d. Evidence for Agave Cultivation in the Silver Creek and Chevelon Creek Drainages. Master’s Report, on file, School of Anthropology, University of Arizona, Tucson.

Kuwanwisiwma, Leigh J., and T. J. Ferguson 2009 Hopitutskwa and Ang Kuktota: The Role of Archaeological Sites in Defining Hopi Cultural Landscapes. In The Archaeology of Meaningful Places, edited by Brenda J. Bowser and M. Nieves Zedeño, pp. 90-106. The University of Utah Press, Salt Lake City.

Kvamme, Kenneth L. 1989 Geographic Information Systems in Regional Archaeological Research and Data Management. In Archaeological Method and Theory, Volume 1, edited by Michael B. Schiffer, pp. 139-203. University of Arizona Press, Tucson. 1995 A View From Across the Water: the North American Experience in Archaeological GIS. In Archaeology and Geographical Information Systems, edited by Gary Lock and Zoran Stancic, pp. 1-14. Taylor and Francis, Bristol, PA. 1999 Recent Directions and Developments in Geographical Information Systems. Journal of Archaeological Research 7:153-201. 2006 There and Back Again: Revisiting Archaeological Locational Modeling. In GIS and Archaeological Site Location Modeling, edited by. Mark W. Mehrer and Konnie L. Wescott, pp. 3-38. Taylor & Francis, Boca Raton.

Laing, Larry, Norman Ambos, Tom Subridge, Christine McDonald, Chris Nelson, and Wayne Robbie 1989 Terrestrial Ecosystems Survey of the Apache-Sitgreaves National Forests. United States Department of Agriculture, Forest Service, Southwestern Region, Albuquerque.

280

Landis, Andrew Gascho, and John Duff Bailey 2004 Reconstruction of Age Structure and Spatial Arrangement of Piñon-Juniper Woodlands and Savannas of Anderson Mesa, Northern Arizona. Forest Ecology and Management 204:221-236.

Lange, Richard, and E. Charles Adams 2012 Rock Art Ranch Field School. Electronic document, http://www.statemuseum.arizona.edu/arch/rock_art_ranch_school.shtml, accessed November 21, 2012.

Lanner, Ronald M. 1981 The Piñon Pine: A Natural and Cultural History. University of Nevada Press, Reno.

Laumbach, Karl W. 1980 An Archaeological Inventory of the H Bar V Reservoir and Adjacent Areas near Springerville, Arizona. New Mexico State University, Cultural Resources Management Division, Department of Sociology and Anthropology, Report No. 415, Las Cruces.

LeBlanc, Steven A. 1997 Modeling Warfare in Southwestern Prehistory. North American Archaeologist 18:235-276. 1999 Prehistoric Warfare in the American Southwest. University of Utah Press, Salt Lake City. 2000 Regional Interaction and Warfare in the Late Prehistoric Southwest. In Archaeology of Regional Interaction: Religion, Warfare, and Exchange Across the American Southwest and Beyond 5:41-70.

Leibenstein, Harvey 1976 Beyond Economic Man: A New Foundation for Microeconomics. Harvard Universities Press, Cambridge, MA.

Lekson, Stephen H. 1996 Landscape in Ruins: Archaeological Approaches to Built and Unbuilt Environments. Current Anthropology 37:886-892. 1999 Chaco Death Squads. Archaeology (May/June):67-73. 2002 War in the Southwest, War in the World. American Antiquity 67:607-624.

Liebmann, Matthew, T. J. Ferguson, and Robert W. Preucel 2005 Pueblo Settlement, Architecture, and Social Change in the Pueblo Revolt Era, A. D. 1680-1696. Journal of Field Archaeology 30(1):45-60.

Lightfoot, Dale R.

281

2004 Landscape and Context: GIS Analysis and Mapping. In The Safford Valley Grids: Prehistoric Cultivation in the Southern Arizona Desert, edited by William E. Doolittle and James A. Neely, pp. 38-47. Anthropological Papers of the University of Arizona Number 70. The University of Arizona Press, Tucson.

Lightfoot, Kent G. 1978 An Archaeological Survey of the Nicks Camp Timber Sale, Apache-Sitgreaves National Forests Pinedale Ranger District, Apache County, Arizona. Office of Cultural Resource Management Report No. 36. Department of Anthropology, Arizona State University, Tempe. 1981 Prehistoric Political Development in the Little Colorado Region, East-Central Arizona. Unpublished Ph.D. dissertation, Department of Anthropology, Arizona State University, Tempe.

Lightfoot, Kent G., and Rachel Most 1989 Interpreting Settlement Hierarchies: A Reassessment of Pinedale and Snowflake Settlement Patterns. In The Sociopolitical Structure of Prehistoric Southwestern Societies., edited by Steadman Upham, Kent G. Lightfoot, and Roberta A. Jewett, pp. 389-418. Westwiew Press, Boulder.

Lightfoot, Kent G., and Fred Plog 1984 Intensification along the North Side of the Mogollon Rim. In Prehistoric Agricultural Strategies in the Southwest, edited by Suzanne K. Fish and Paul R. Fish, pp. 179-196. Anthropological Research Papers 33. Arizona State University, Tempe.

Lipe, William D. 1995 The Depopulation of the Northern San Juan: Conditions in the Turbulent 1200s. Journal of Anthropological Archaeology 14:143-169.

Llobera, Marcos 2007 Reconstructing Visual Landscapes. World Archaeology 39:51-69 2011 Archaeological Visualization: Towards an Archaeological Information Science (AISc). Journal of Archaeological Method and Theory 18:193-223.

Longacre, William A. 1964 A Synthesis of the Upper Little Colorado Prehistory. In Chapters in the Prehistory of Eastern Arizona II, by Paul S. Martin, John B. Rinaldo, William A. Longacre, Leslie G. Freeman, Jr., James A. Brown, Richard H. Hevly, M. E. Cooley, Hugh C. Cutler, and Stevens F. F. Seaberg. Fieldiana: Anthropology 55:155-170.

Lowell, Julia C. 2007 Women and Men in Warfare and Migration: Implications of Gender Imbalance in the Grasshopper Region of Arizona. American Antiquity 72(1):95-123.

282

Mabry, Jonathan B. (editor) 1996 Canals and Communities: Small-Scale Irrigation Systems. Arizona Studies in Human Ecology. University of Arizona Press, Tucson.

Maschner, Herbert D. G., and Julie S. Stein 1995 Multivariate Approaches to Site Location on the Northwest Coast of North America. Antiquity 69:61-73.

Matson, R. G., William D. Lipe, and William R. Haase IV 1988 Adaptational Continuities and Occupational Discontinuities: The Cedar Mesa Anasazi. Journal of Field Archaeology 15(3):245-264.

McAllister, Shirley Powell, and Fred Plog 1978 Small Sites in the Chevelon Drainage. In Limited Activity and Occupation Sites: A Collection of Conference Papers, edited by Albert E. Ward, pp. 17-23. Contributions to Anthropological Studies No. 1. Center for Anthropological Studies, Albuquerque, New Mexico.

McCoy, Mark D., and Thegn N. Ladefoged 2009 New Developments in the Use of Spatial Technology in Archaeology. Journal of Archaeological Research 17:263-295.

McPherson, Robert S. 1992 Sacred Land, Sacred View. Brigham Young University, Provo.

Mehrer, Mark W., and Konnie L. Wescott (editors) 2006 GIS and Archaeological Site Location Modeling. Taylor & Francis, New York.

Mills, Barbara J. 1998 Migration and PIV Community Reorganization in the Silver Creek Area, East- Central Arizona. In Migration and Reorganization: The Pueblo IV Period in the American Southwest, edited by Katherine A. Spielmann, pp. 65-80. Arizona State University Anthropological Research Papers 51. Arizona State University, Tempe. 2005 Curricular Matters: The Impact of Field Schools on Southwest Archaeology. In Southwest Archaeology in the Twentieth Century, edited by Linda S. Cordell and Don D. Fowler, pp. 60-80.The University of Utah Press 2011 Themes and Models for Understanding Migration in the Southwest. In Movement, Connectivity, and Landscape Change in the Ancient Southwest, edited by Margaret Nelson and Colleen Strawhacker, pp.345-361. University Press of Colorado, Boulder

Mills, Barbara J., and Samuel Duwe 2003 Annual Report of the University of Arizona Archaeological Field School, submitted to the U.S. Forest Service, Apache-Sitgreaves National Forests, Springerville.

283

Mills, Barbara and T.J. Ferguson 1998 Preservation and Research of Sacred Sites by the Zuni Indian Tribe of New Mexico. Human Organization 57(1):30-42

Mills, Barbara J., and Sarah A. Herr 1999 Chronology of the Mogollon Rim Region. In Living on the Edge of the Rim: Excavations and Analysis of the Silver Creek Archaeological Research Project 1993- 1998, edited by Barbara J. Mills, Sarah A. Herr, and Scott Van Keuren, pp. 269-293. Arizona State Museum Archaeological Series 192. University of Arizona, Tucson.

Mills, Barbara J., Sarah A. Herr, and Scott Van Keuren (editors) 1999 Living on the Edge of the Rim: Excavations and Analysis of the Silver Creek Archaeological Research Project 1993-1998. Arizona State Museum Archaeological Series 192. University of Arizona, Tucson.

Mills, Barbara J., and Michael Margolis 2002 Annual Report of the University of Arizona Archaeological Field School, submitted to the U. S. Forest Service, Apache-Sitgreaves National Forests, Springerville, AZ.

Mills, Barbara, Mark Altaha, John Welch, and T.J. Ferguson 2008 Field Schools without Trowels: Teaching Archaeological Ethics and Heritage Preservation in a Collaborative Context. In Collaborating at the Trowel's Edge: Teaching and Learning in Indigenous Archaeology, edited by Stephen W. Silliman. The University of Arizona Press, Tucson.

Milner, George R. 1995 An Osteological Perspective on Prehistoric Warfare. In Regional Approaches to Mortuary Analysis, edited by Lane A. Beck, pp.221-44. Plenum, New York. 1999 Warfare in Prehistoric and Early Historic Eastern North America. Journal of Archaeological Research 7:105-151.

Milner, George R., Eve Anderson, and Virginia G. Smith 1991 Warfare in Late Prehistoric West-Central Illinois. American Antiquity 56:581-603.

Mindeleff, Cosmos 1896 Aboriginal Remains in Verde Valley, Arizona. Thirteenth Annual Report of the Bureau of American Ethnology, pp. 179-261. Washington, D.C.

Minnis, Paul E. 1996 Notes on Economic Uncertainty and Human Behavior in the Prehistoric Southwest. In Evolving Complexity and Environmental Risk in the Prehistoric Southwest, edited by Joseph A. Tainter and Bonnie Bagley Tainter, pp. 57-78. Proceedings Volume XXIV, Sante Fe Institute Studies in the Sciences of Complexity. Addison-Wesley

284

Publishing Company, Readig, MA. 2000 Prehistoric Agriculture and Anthropogenic Ecology of the North American Southwest. In Archaeology of Drylands: Living on the Margins, edited by Graeme Barker and David Gilbertson, pp. 271-286. Routledge, Florence.

Minnis, Paul E., and Stephen E. Plog 1976 A Study of the Site Specific of Agave Parryi in East-Central Arizona. Kiva 41(3- 4):299-308.

Minnis, Paul E., Michael E. Whalen, and R. Emerson Howell 2006 Fields of Power: Upland Farming in the Prehispanic Casas Grandes Polity, Chihuahua, Mexico. American Antiquity 71(4):707-722.

Moore, Bruce 1978 Are Pueblo Field Houses a Function of Urbanization? In Limited Activity and Occupation Sites, compiled by Albert E. Ward, pp 9-16. Center for Anthropological Studies, Contributions No. 1. Albuquerque, New Mexico. 1979 Pueblo Isolated Small Structure Sites. Unpublished Ph.D. dissertation, Department of Anthropology, Southern Illinois University, Carbondale.

Moore, I.D., R.B. Grayson, and A.R. Ladson 1991 Digital Terrain Modelling: a Review of Hydrological, Geomorphological and biological applications. Hydrological Processes 5:3-10

Mosher, Anne E., and Laurie A. Wilkie 2010 Historical Archaeo-Geographies of Scaled Statehood: American Federalism and Material Practices of National Prohibition in California, 1917–1933. Archaeologies: Journal of the World Archaeological Congress 6(1):82-114.

Murray, Wendi F., Nicholas C. Laluk, Barbara J. Mills, and T. J. Ferguson 2009 Archaeological Collaboration with American Indians: Case Studies from the Western United States. Collaborative Anthropologies 2:65-89.

Nelson, Ben A. 2000 Aggregation, Warfare, and the Spread of the Mesoamerican Tradition. In Archaeology of Regional Interaction: Religion, Warfare, and Exchange Across the American Southwest and Beyond 5:317-337.

Nelson, Ben A., and Roger Anyon 1996 Fallow Valleys: Asynchronous Occupations in Southwestern New Mexico. Kiva 61:275-294.

Nelson, Ben A., and Stephen A. LeBlanc

285

1986 Short-Term Sedentism in the American Southwest. University of New Mexico Press, Albuquerque.

Nelson, Margaret C. 1993a Classic Mimbres Land Use in the Eastern Mimbres Region, Southwestern New Mexico. Kiva 59:27-47. 1993b Changing Occupational Patterns among Prehistoric Horticulturalists in Southwestern New Mexico. Journal of Field Archaeology 20:43-87. 1999 Mimbres during the Twelfth Century: Abandonment, Community and Reorganization. University of Arizona Press, Tucson.

Nelson, Margaret C., and Michelle Hegmon 2001 Abandonment is not as it seems: An approach to the relationship between site and regional abandonment. American Antiquity 66:213-235.

Nelson, Margaret C., and Gregson Schachner 2002 Understanding Abandonments in the North American Southwest. Journal of Archaeological Research 10:167-206.

Netting, Robert McC. 1993 Smallholders, Householders: Farm Families and the Ecology of Intensive Sustainable Agriculture. Stanford University Press, Stanford.

Newcomb, Joanne M. 1999 Silver Creek Settlement Patterns and Paleodemography. In Living on the Edge of the Rim: Excavations Analysis of the Silver Creek Archaeological Research Project 1993-1998, edited by Barbara J. Mills, Sarah A. Herr, and Scott Van Keuren, pp. 31-52. Arizona State Museum Archaeological Series 192, Arizona State Museum, University of Arizona, Tucson.

Nichols, Deborah L., and Patricia L. Crown (editors) 2008 Social Violence in the Prehispanic Southwest. The University of Arizona Press, Tucson.

North, Chris, Michael S. Foster, and Louise Senior 2003 Archaeology at the Edge of the Rim: An Area Heritage Resource Survey of Portions of the Rodeo-Chediski Burn Area, Black Mesa and Lakeside Ranger Districts, Apache-Sitgreaves National Forests, Navajo County, Arizona. SWCA Cultural Resources Report No. 2003-56. SWCA, Inc., Phoenix.

Obregón, Baltasar, George Hammond, and Agapito Rey 1584[1928] Obregón’s History of the 16th Century Explorations in Western America, entitled Chronicle, Commentary, or Relation of the Ancient and Modern Discoveries in New Spain and New Mexico, Mexico. Wetzel Publication Co., Los Angeles

286

Ostrom, Elinor 1990 Governing the Commons: the Evolution of the Institutions for Collective Action. Cambridge University Press, New York. 2005 Understanding Institutional Diversity. Princeton University Press, Princeton.

Ostrom, Elinor, Roy Gardner, and James Walker 1994 Rules, Games, and Common-Pool Resources. University of Michigan Press, Ann Arbor.

Ostrom, Elinor, Joanna Burger, Christopher Field, Richard Norgaard, and David Policansky 1999 Revisting the Commons: Local Lessons, Global Challenges. Science 284(5412):278-282.

Pauketat, Timothy R. 2009 Wars, Rumors of Wars, and the Production of Violence. In Warfare in Cultural Context: Practice, Agency, and the Archaeology of Violence, edited by Axel E. Nielsen and William H. Walker, pp. 244-262. The University of Arizona Press, Tucson.

Peeples, Matthew A., and W. Randall Haas In press Brokerage and Social Capital in the Prehispanic U.S. Southwest. American Anthropologist

Peeples, Matthew A., C. Michael Barton, and Steven Schmich 2006 Resilience lost: intersecting land use and landscape dynamics in the prehistoric southwestern United States. Ecology and Society 11(2): 22. Electronic document http://www.ecologyandsociety.org/vol11/iss2/art22/, accessed November 19

Periman, Richard D. 2005 Modeling Landscapes and Past Vegetation Patterns of New Mexico’s Rio Del Oso Valley. Geoarchaeology 20(2):193-210.

Perramond, Eric P. 1994 An Historical Geography of Zuni Agriculture. Unpublished Master’s thesis, Department of Geography and Anthropology, Louisiana State University.

Phillips, David A. 2009 Adoption and Intensification of Agriculture in the North American Southwest: Notes toward a Quantitative Approach. American Antiquity 74(4):691-707

Pilles, Peter J., Jr.

287

1978 The Field House and Sinagua Demography. In Limited Activity and Occupation Sites: A Collection of Conference Papers, edited by Albert E. Ward, pp. 119-133. Center for Anthropological Studies, Albuquerque, New Mexico. 2000 General Requirements for Conducting Surveys on the Coconino National Forest. Coconino National Forest Heritage Resources Section. Flagstaff.

Pilles, Peter J., Jr., and David R. Wilcox 1978 The Small Sites Conference: An Introduction. In Limited Activity and Occupation Sites: A Collection of Conference Papers, edited by Albert E. Ward, pp. 1-4. Center for Anthropological Studies, Albuquerque, New Mexico.

Plog, Fred 1970 Water Control and the Upper Little Colorado. Unpublished paper presented at the 1970 Pecos Conference symposium on Prehistoric Southwestern Water Control Systems. Pecos National Monument, New Mexico. On file at the Arizona State Museum, University of Arizona, Tucson. 1974a The Study of Prehistoric Change. Academic Press, New York. 1974b Settlement Patterns and Social History. In Frontiers in Anthropology, edited by Murray Leaf, pp. 68-91. D. Van Nostrand Co., New York 1978 An Analysis of Variability in Site Locations in the Chevelon Drainage, Arizona. In Investigations of the Southwestern Anthropological Research Group: An Experiment in Archaeological Cooperation, edited by Robert C. Euler and George J. Gumerman, pp. 53-71. Museum of Northern Arizona, Flagstaff. 1981a Cultural Resources Overview: Little Colorado Area, Arizona. USDA Forest Service and the Bureau of Land Management, Albuquerque. 1981b Managing Archaeology: A Background Document for Cultural Resource Management on the Apache-Sitgreaves National Forests, Arizona. Cultural Resources Management Report No. 1. U.S.D.A. Forest Service, Southwestern Region. 1983 Political and Economic Alliances on the Colorado Plateaus, A.D. 400-1450. In Advances in World Archaeology, vol.2, edited by Fred Wendorf and Angela E. Close, pp. 289-330. Academic Press, New York. 1984 Exchange, Tribes and Alliances: The Northern Southwest. American Archaeology 4(3):217-223. 1989 The Sinagua and Their Relations. In Dynamics of Southwest Prehistory, edited by George J. Gumerman, pp. 263-291. School of American Research, Santa Fe.

Plog, Fred, and James N. Hill 1971 Explaining Variability in the Distribution of Sites. In The Distribution of Prehistoric Population Aggregates, edited by George J. Gumerman, pp. 7-36. Prescott College, Anthropological Reports 1.

Plog, Fred T., James N. Hill, and Dwight W. Read 1976. Chevelon Archaeological Research Project, 1971-1972. Monograph II of the Archaeological Survey, University of California, Los Angeles

288

Plog, Fred, George J. Gumerman, Robert C. Euler, Jeffrey S. Dean, Richard H. Hevly, and Thor N. V. Karlstrom 1988 Anasazi Adaptive Strategies: The Model, Predictions, and Results. In The Anasazi in a Changing Environment, edited by George J. Gumerman, pp. 230-276. Cambridge University Press, Cambridge.

Plog, Stephen 2003 Social Conflict, Social Structure, and Processes of Culture Change. American Antiquity 68:182-186.

Plog, Stephen, Julie P. Solometo, Wes Bernardini, and Carrie C. Heitman 2001 Understanding Organizational Change in the Chevelon Archaeological Drainage, Arizona: A Preliminary Report on the 1997-2000 Research of the Chevelon Archaeological Research Project. Report submitted to the Apache-Sitgreaves National Forest, Springerville, AZ.

Pool, Michael D. 1994 Implications of the Western Apache Settlement System for the Early Mogollon Period. In Mogollon VII: The Collected Papers of the 1992 Mogollon Conference Held in Las Cruces, New Mexico, edited by Patrick Beckett, pp. 89-103. COAS Publishing & Research, Las Cruces.

Pool, Christopher A. and Lisa Cliggett 2008 Introduction:Economies and the Transformation of Landscapes. In Economies and the Transformation of Landscape, edited by Lisa Cliggett and Christopher A. Pool, pp. 1- 16. AltaMira Press, Lanham.

Poteete, Amy R., Marco A. Janssen, and Elinor Ostrom 2010 Working Together: Collective Action, the Commons and Multiple Methods in Practice. Princeton University Press

Potter, J. M., and J. P. Chuipka 2010 Perimortem Mutilation of Human Remains in an Early Village in the American Southwest: A Case for Ethnic Violence. Journal of Anthropological Archaeology 29:507- 523

Powers, M. A., and B. P. Johnson 1987 Defensive Sites of Dinétah. Cultural Resources Series, Vol. 2. USDI New Mexico Bureau of Land Management, Albuquerque.

Premo, Lucas S.

289

2004 Local Spatial Autocorrelation Statistics Quantify Multi-scale Patterns in Distributional Data: An Example from the Maya Lowlands. Journal of Archaeological Science 31:855-866.

Preucel, Robert W. 1987 Settlement Succession on the Pajarito Plateau, New Mexico. Kiva 53:3-33. 1990 Seasonal Circulation and Dual Residence in the Pueblo Southwest: A Prehistoric Example from the Pajarito Plateau, New Mexico. Garland Publishing, New York.

Prudden, T. Mitchell 1903 The Prehistoric Ruins of the San Juan Watershed in Utah, Arizona, Colorado, and New Mexico. American Anthropologist 5(1):224-288.

Rainey, Katharine and Karen R. Adams 2004 Compendium A: Ethnographically Documented Uses of Plants. Crow Canyon Archaeological Center, Cortez, Colorado. Electronic document http://www.crowcanyon.org/ResearchReports/Archaeobotanical/Plant_Uses/compendium _A.asp 1. Accessed October 2012

Rankin, Adrianne G. 1989 Agricultural Features. In Settlement, Subsistence, and Specialization in the Northern Periphery: The Waddell Project, Vol. 1, edited by Margerie Green, pp. 949- 979. Cultural Resources Report 65. Archaeological Consulting Services, Tempe.

Redman, Charles L. 1999 Human Impacts on Ancient Environments. University of Arizona Press, Tucson.

Reid, J. Jefferson (editor) 1982a Cholla Project Archaeology, Volume 2: The Chevelon Region. Archaeological Series No. 161. Cultural Resource Management Division, Arizona State Museum, University of Arizona, Tucson. 1982b Analytic Procedures for Interassemblage-Settlement System Analysis. In Cholla Project Archaeology, Volume 1, edited by J. Jefferson Reid, pp. 193-204. Arizona State Museum Archaeological Series No. 161. University of Arizona Press, Tucson. 1995 Four Strategies after Twenty Years: A Return to Basics. In Expanding Archaeology, edited by James M. Skibo, William H. Walker, and Axel E. Nielsen, pp. 15-21. University of Utah Press, Salt Lake City. 1998 Return to Migration, Population Movement, and Ethnic Identity in the American Southwest: A Peer Reviewer’s Thoughts on Archaeological Inference. In Vanishing River: Landscapes and Lives of the Lower Verde Valley, edited by Stephanie Whittlesey, Richard Ciolek-Torello, and Jeffrey Altschul, pp. 629-638. SRI Press, Tucson. 1999 The Grasshopper-Chavez Pass Debate: Existential Dilemmas and Archaeological Discourse. In Sixty Years of Mogollon Archaeology: Papers from the Ninth Mogollon

290

Conference, Silver City, New Mexico, 1996, edited by Stephanie M. Whittlesey, pp. 13- 22. SRI Press, Tucson.

Reid, J. Jefferson, and Stephanie Whittlesey 1997 The Archaeology of Ancient Arizona. The University of Arizona Press, Tucson. 1999 Grasshopper Pueblo: A Story of Archaeology and Ancient Life. The University of Arizona Press, Tucson.

Reid, J. Jefferson, Michael B. Schiffer, and William J. Rathje 1975 Behavioral Archaeology: Four Strategies. American Anthropologist 77(4):864- 869.

Rice, Glen E. 2001 The Role of Field Houses and Farmsteads in Hohokam Settlement Systems. P- MIP Technical Report No. 2001-10. United States Department of Interior, Bureau of Reclamation, Arizona Projects Office, Phoenix, Arizona.

Rice, Glen E., and S. A. LeBlanc (editors) 2001 Deadly Landscapes: Case Studies in Prehistoric Southwestern Warfare. University of Utah Press, Salt Lake City.

Richard, Adam 2002 Analysis of the Agave Site Lithic Assemblage. Student paper, University of Arizona Archaeological Field School. Ms. on file, School of Anthropology, University of Arizona, Tucson.

Roberts, Frank H. H. 1935 A Survey of Southwestern Archaeology. American Anthropologist 37:1-35.

Rohn, Arthur H. 1963 Prehistoric Soil and Water Conservation on Chapin Mesa, Southwestern Colorado. American Antiquity 28(4):441-455. 1971 Mug House, Mesa Verde National Park, Colorado. Archaeological Research Series, no. 7-D. National Park Service, Washington, D.C.

Romme, William H., Lisa Floyd-Hanna, and David D. Hanna 2003 Ancient piñon-juniper forests of Mesa Verde and the West: a cautionary note for forest restoration programs. In Proceedings of the Conference on Fire, Fuel Treatments, and Ecological Restoration, edited by Philip N. Omi and Linda A. Joyce, pp. 335-350. USDA Forest Service Proceedings RMRS-P-29. Rocky Mountain Research Station, Fort Collins.

Roos, Christopher I. 2005 Mobile Farmers? Placing the Hall Point Site in Local Context. Paper

291 presented at the 70th Annual Meeting of the Society for American Archaeology, Salt Lake City, UT. 2008a Palimpsests and Short-Term Occupations of Early Pithouse Period Sites in East-Central Arizona. Paper presented at the 73rd annual meeting of the Society for American Archaeology, Vancouver, BC. 2008b Fire, Climate, and Social-Ecological Systems in the Ancient Southwest: Alluvial Geoarchaeology and Applied Historical Ecology. Unpublished Ph.D. Dissertation, University of Arizona.

Rossignol, Jacqueline 1992 Concepts, Methods, and Theory Building: A Landscape Approach. In Space, Time, and Archaeological Landscapes, edited by Jacqueline Rossignol and LuAnn Wandsnider, pp. 3-19. Plenum Press, New York.

Roth, Barbara 2011 Site Visibility and the Interpretation of Mimbres Mogollon Demography in the U.S. Southwest. Journal of Field Archaeology 36(3):221-231.

Ruppel, Timothy, Jessica Neuwirth, Mark P. Leone, and Gladys-Marie Fry 2003 Hidden in View: African Spiritual Spaces in North American Landscapes. Antiquity 77:321-335.

Saile, David G. 1975 Pueblo Architecture in the Context of Cultural Conflict. An Open lecture Presented at the University of Newcastle upon Tyne.

Sandor, Jonathan A., Jay B. Norton, Jeffrey A. Homburg, Deborah A. Muenchrath, Carleton S. White, Stephen E. Williams, Celeste I. Havener, and Peter D. Stahl 2007 Biogeochemical Studies of a Native American Runoff Agroecosystem. Geoarchaeology 22:359-386.

Sarker, Ashutosh, Helen Ross, and Krishna K. Shrestha 2008 Interdependence of Common-Pool Resources: Lessons from a Set of Nested Catchments in Australia. Human Ecology 36:821-834.

Sauer, Carl O. 1925 The Morphology of Landscape. University of California Publications in Geography Vol. 2, No. 2. University of California Press, Berkeley. 1941 Forward to Historic Geography. Annals of the Association of American Geographers 31(1):1-24.

SCARP 2012 Website address: http://web.arizona.edu/~scarp/index.htm. Accessed November 12, 2012.

292

Schaafsma, P. 2000 Warrior, Shield, and Star: Imagery and Ideology of Pueblo Warfare. Western Edge Press, Santa Fe.

Schiffer, Michael B. 1987 Formation Processes of the Archaeological Record. University of New Mexico Press, Albuquerque.

Schlanger, Sarah H. 1990 Artifact Assemblage Composition and Site Occupation Duration. In Perspectives on Southwestern Prehistory, edited by Paul E. Minnis and Charles L. Redman, pp. 103- 121. Westview Press, Boulder, Colorado.

Schlanger, Sarah, and Signa Larralde 2008 Why is Cultural Resource Density High in the Piñon-Juniper Woodland? In Ecology, Management, and Restoration of Piñon-Juniper and Ponderosa Pine Ecosystems: Combined Proceedings of the 2005 St. George, Utah and 2006 Albuquerque, New Mexico Workshops, compiled by Gerald J. Gottfried, John D. Shaw, and Paulette L. Ford, pp. 146-147. Proceedings RMRS-P-51, Rocky Mountain Research Station, USDA Forest Service, Fort Collins.

Scholnick, Jonathan B. 2003 Village Formation During the Pueblo III to Pueblo IV Period Transition: Contextualizing Bryant Ranch Pueblo, Arizona . Unpublished Master's Thesis, Department of Anthropology, The University of Arizona, Tucson.

Sebastian, Lynee, and W. J. James Judge 1988 Predicting the past: Correlation, explanation and the use of archaeological models. In Quantifying the Present and Predicting the Past: Theory, Method, and Application of Archaeological Predictive Modeling, edited by W. James Judge and Lynne Sebastian, pp. 1–18. US Department of the Interior, Bureau of Land Management, Denver.

Seidel, Andrew 2004 Archaeological Match-Making: A Survey Concordance. In University of Arizona Archaeological Field School in Archaeology and Heritage Resource Management 2004, edited by B. J. Mills. A report prepared for the White Mountain Apache Tribe, Manuscript on file at the Department of Anthropology, University of Arizona, Tucson.

Senior, Louise 2004 Rim Country Ethnicity: An Ethnographic Resources Inventory of the Rodeo- Chediski Burn Area. SWCA Cultural Resources Report No. 03-164. SWCA Environmental Consultants, Tucson. On file, Apache-Sitgreaves National Forests Supervisor’s Office.

293

Singh, Vijay P., and Mauro Fiorentino 1996 Geographical Information Systems in Hydrology. Kluwer Academic Publishers, Boston.

Slatter, Edwin D. 1979 Drought and Demographic Change in the Prehistoric Southwest United States: A Preliminary Quantitative Assessment. Unpublished Ph.D. dissertation. University of California, Los Angeles.

Snead, James E. 2002 Ancestral Pueblo Trails and the Cultural Landscape of the Pajarito Plateau, New Mexico. Antiquity 76:756-765. 2004 Ancestral Pueblo Settlement Dynamics: Landscape, Scale, and Context in the Burnt Corn Community. Kiva 69(3):242-270. 2008 Ancestral Landscapes of the Pueblo World. University of Arizona Press, Tucson.

Solometo, Julie P. 2001 Tactical Site of the Chevelon and Clear Creek Drainages. In The Archaeology of Ancient Tactical Sites, edited by John Welch and Todd Bostwick, pp.21-36. Arizona Archaeologist No. 32. Arizona Archaeological Council, Tempe. 2004 The Conduct and Consequences of War: Dimensions of Conflict in East-Central Arizona. Unpublished PhD dissertation, The University of Michigan. 2006 The Dimensions of War: Conflict and Culture Change in Central Arizona. In The Archaeology of Warfare: Prehistories of Raiding and Conquest, edited by Elizabeth N. Arkush and Mark W. Allen, pp. 23-65. University Press of Florida, Gainesville.

Speight, J.Garry 1974 A Parametric Approach to Landform Regions. In Progress in Geomorphology, edited by David Linton, Eric Brown and Ronald Waters, pp.213-230. Alden Press, London.

Spielmann, Katherine A., Margaret Nelson, Scott Ingram, and Matthew Peeples 2011 Sustainable Small-Scale Agriculture in Semi-Arid Environments. Ecology and Society 16(1):26. Electronic document: http://www.ecologyandsociety.org/vol16/iss1/art26/, accessed November 10

Spier, Leslie 1917 Notes on Some Little Colorado Ruins. Anthropological Papers of the American Museum of Natural History 18(4)335-361.

Steward, Julian H. 1937 Ecological Aspects of Southwestern Society. Anthropos 32:87-104.

294

1938 Basin-Plateau Aboriginal Sociopolitical Groups. Bulletin 120, Bureau of American Ethnology, Washington, DC.

Stiger, Mark A. 1979 Mesa Verde Subsistence Patterns from Basketmaker to Pueblo III. Kiva 44:133- 144.

Sullivan, Alan P. 1980 Prehistoric Settlement Variability in the Grasshopper Area, East-Central Arizona. Unpublished PhD dissertation, University of Arizona, Tucson. 1983 Seeds of Discontent: Implications of a “Pompeii” Archaeobotanical Assemblage for Grand Canyon Anasazi Subsistence Models. Journal of Ethnobiology 7:137-153. 1986 Prehistory of the Upper Basin, Coconino County, Arizona. Archaeological Series No. 167. Arizona State Museum, Tucson. 1992a Investigating the Archaeological Consequences of Short-Duration Occupations. American Antiquity 57:99-115. 1992b Pinyon Nuts and Other Wild Resources in Wewstern Anasazi Subsistence Economies. In Research in Economic Anthropology, Supplement 6:195-239. 1995 Behavioral Archaeology and the Interpretation of Archaeological Variability. In Expanding Archaeology, edited by J. M. Skibo, William H. Walker, and Axel E. Nielson, pp. 178-186. University of Utah Press, Salt Lake City. 1996 Risk, Anthropogenic Environments, and Western Anasazi Subsistence. In Evolving Complexity and Environmental Risk in the Prehistoric Southwest, edited by Joseph A. Tainter and Bonnie Bagley Tainter, pp. 145-168. Proceedings Volume XXIV, Sante Fe Institute Studies in the Sciences of Complexity. Addison-Wesley Publishing Company, Reading. 2000 Effects of Small-Scale Prehistoric Runoff Agriculture on Soil Fertility: The Developing Picture from Upland Terraces n the American Southwest. Geoarchaeology 15:291-313.

Sulivan, Alan P., III and Anthony S. Tolonen 1998 Evaluating Assemblage Diversity Measures with Surface Archaeological Data. In Surface Archaeology, edited by Alan P. Sullivan III, p. 143-155. University of New Mexico Press, Albuquerque.

Tagg, Martyn D. 1994 Projectile points of east-central Arizona: forms and chronology. In Middle Little Colorado River Archaeology: From the Parks to the People, edited by Anne Trinkle Jones and Martyn D. Tagg, pp. 87–115. The Arizona Archaeologist No. 27. Phoenix: Arizona Archaeological Society.

Tainter, Joseph and Fred Plog

295

1994 Strong and Weak Patterning in Southwestern Prehistory: the Formation of Pueblo Archaeology. In Themes in Southwest Prehistory, edited by George Gummerman, pp.165-181. School of American Research Press, Santa Fe.

Thatcher, A. P., and V. L. Hart 1974 Spy Mesa yields better understanding of piñon-juniper in range ecosystem. Journal of Range Management 27:354-357.

Tilley, Christopher Y. 1995 Rocks as Resources: Landscapes and Power. Cornish Archaeology 34:5-57. 2001 Thinking Places. Cambridge Archaeological Journal 11:130-132.

Towner, R. H. 2003 Defending the Dinétah: Pueblitos in the Ancestral Navajo Heartland. University of Utah Press, Salt Lake City.

Treacy, John M., and William M. Denevan 1994 The Creation of Cultivable Land Through Terracing. In The Archaeology of Garden and Field, edited by Naomi F. Miller and Kathryn L. Gleeson, pp. 91-110. University of Pennsylvania Press, Philadelphia.

Trigger, Bruce G. 1989 A History of Archaeological Thought. Cambridge University Press, Cambridge.

Tucker, Catherine M. 2010 Private Goods and Common Property: Pottery Production in a Honduran Lenca Community. Human Organization 69(1):43-53.

Tuggle, H. David 1982 Cholla and the Chevelon Archaeological Research Project. In Cholla Project Archaeology Volume 2: The Chevelon Region, edited by J. Jefferson Reid, pp. 180-194. Archaeological Series No. 161. Cultural Resources Management Division, Arizona State Museum, University of Arizona, Tucson.

Tuggle, H.David, Jeffrey J. Reid, and R.C. Cole 1984 Fourteenth Century Mogollon Agriculture in the Grasshopper Region of Arizona. In Prehistoric Agricultural Strategies in the Southwest, edited by Suzanne K. Fish and Paul R. Fish. Anthropological Research Papers, no.33. Arizona State University, Tempe.

Turner, Christy G., II, and Jacqueline A. Turner 1999 Man Corn. University of Utah Press, Salt Lake City.

Upham, Steadman

296

1982 Polities and Power: An Economic and Political History of the Western Pueblo. Academic Press, New York 1984 Ecological and Political Perspectives on Labor Intensive Agriculture and Exchange During the 14th Century. In Prehistoric Agricultural Strategies in the Southwest, edited by Suzanne K. Fish and Paul R. Fish, pp. 291-307. Anthropological Research Papers No. 33. Arizona State University, Tempe

Upham, Steadman, Gail M. Bockley 1989 The Chronologies of Nuvakwewtaqa: Implications for Social Processes. In The Sociopolitical Structure of Prehistoric Southwestern Societies, edited by Steadman Upham, Kent G. Lightfoot, and Roberta A. Jewett, pp. 447-490. Westview, Boulder

USDA 1989 Terrestrial Ecosystems Survey of the Apache-Sitgreaves National Forests. United States Department of Agriculture, Forest Service, Southwestern Region.

Van Dyke, Ruth 2007 The Chaco Experience: Landscape and Ideology at the Center Place. School of Advanced Research Press, Santa Fe.

Van West, Carla R. 1994 Modeling Prehistoric Agricultural Productivity in Southwestern Colorado: A GIS Approach. Reports of Investigations 67, Washington State University, Department of Anthropology, Pullman.

Varien, Mark D., Scott G. Ortman, Timothy A. Kohler, Donna M. Glowacki, and C. David Johnson 2007 Historical Ecology in the Mesa Verde Region: Results from the Village Ecodynamics Project. American Antiquity 72(2):273-299.

Vencl, S. 1984 War and Warfare in Archaeology. Journal of Anthropological Archaeology 3:116-132.

Verhagen, Phillip, and Thomas G. Whitley 2011 Integrating Archaeological Theory and Predictive Modeling: A Live Report from the Scene. Journal of Archaeological Method and Theory 18(1).

Vierra, Bradley and Richard Ford 2007 Foragers to Farmers in the Northern Rio Grande, New Mexico. Kiva 73:117-130.

Vivian, R. Gwinn

297

1969 Archaeological Salvage on the Pinedale and Clay Springs Sections, Payson- Showlow Highway, State Route 160, A Preliminary Report. Ms. on file, Arizona State Museum, Tucson.

Vivian, R. Gwinn, Carla R. Van West, Jeffrey S. Dean, Nancy J. Akins, Mollie S. Toll, and Thomas C. Windes 2006 Ecology and Economy. In The Archaeology of Chaco Canyon: An Eleventh- Century Pueblo Regional Center, edited by Stephen H. Lekson, pp. 45-66. School of American Research Press, Santa Fe.

Wandsnider, Luann 1998 Landscape Element Configuration, Lifespan, and Occupational History: Ethnoarchaeological Observations and Archaeological Applications. In Surface Archaeology, edited by Alan P. Sullivan, pp. 21-39. University of New Mexico Press, Albuquerque.

Wandsnider, Luann, and Eileen L. Camilli 1996 Land-Use Histories from the Mesilla Bolson (South-Central New Mexico): Evidence from Spatial Patterning in Surface Artifact Assemblages. In Interpreting Southwestern Diversity: Underlying Principles and Overarching Patterns, edited by Paul R. Fish and J. Jefferson Reid, pp. 211-239. Anthropological Research Papers No. 48. Arizona State University, Tempe.

Ward, Albert E. (compiler and editor) 1978 Limited Activity and Occupation Sites: A Collection of Conference Papers. Contributions to Anthropological Studies No. 1. Center for Anthropological Studies, Albuquerque, New Mexico.

Warren, Robert E. 1990a Predictive Modelling in Archaeology: A Primer. In Interpreting Space: Geographic Information Systems and Archaeology, edited by Kathleen M. S. Allen, Stanton W. Green, and Ezra B. W. Zubrow, pp. 90-111. Taylor & Francis, New York. 1990b Predictive Modelling of Archaeological Site Location: A Case Study in the Midwest. In Interpreting Space: GIS and Archaeology, edited by Kathleen M. S. Allen, Stanton W. Green, and Ezra B. W. Zubrow, pp. 201-215. Taylor & Francis, New York.

Waters, Michael R., and John C. Ravesloot 2001 Landscape Change and the Cultural Evolution of the Hohokam Along the Middle Gila and Other River Valleys in South-Central Arizona. American Antiquity 66:285-299. 2003 Disaster or Catastrophe: Human Adaptation to High- and Low-Frequency Landscape Processes—A Reply to Ensor, Ensor, and De Vries. American Antiquity 68(2):400-405.

Welch, John R.

298

1996 Archaeological Measures and Social Implications of Agricultural Commitment. PhD. dissertation, Department of Anthropology, University of Arizona, Tucson.

Welch, John R., and Todd W. Bostwick (editors) 2000 The Archaeology of Ancient Tactical Sites. Arizona Archaeologist No.32. Arizona Archaeological Council, Tempe.

Wells, Lisa 2001 Replacing Humans on the Landscape: A Geomorphological Approach to Reconstructing Archaeological Settlement Patterns. In Earth Sciences and Archaeology, edited by Paul Goldberg, Vance T. Holliday, and C. Reid Ferring, pp.107-142. Kluwer Academic/Plenum, New York.

Wells, E. Christian, Glen E. Rice, and John C. Ravesloot 2004 Peopling Landscapes Between Villages in the Middle Gila River Valley of Central Arizona. American Antiquity 69:627-652.

Wenger, Gilbert R. 1980 The Study of Mesa Verde National Park. Mesa Verde Museum Association, Mesa Verde National Park, Colorado.

White, Christopher 1976 Prehistoric Warfare in the Chevelon Creek Area: An Ecological Approach. In Chevelon Archaeological Research Project, 1971-1972, edited by Fred Plog, James Hill and Dwight Read, pp.126-145. UCLA Archaeological Survey Monograph no.2

Whiteley, Peter M. 2011 Hopi Place Value: Translating a Landscape. In Born in the Blood: On Native American Translation, edited by Brian Swann, pp. 84-108. University of Nebraska Press, Lincoln.

Whittlesey, Stephanie M. (editor) 1992 Archaeological Investigations at Lee Canyon: Kayenta Anasazi Farmsteads in the Upper Basin, Coconino County, Arizona. Technical Series No. 38. SRI Press, Tucson.

Whittlesey, Stephanie M. 1998a Archaeological Landscapes: A Methodological and Theoretical Discussion. In Vanishing River, Landscapes and Lives of the Lower Verde Valley, edited by Stephanie. Whittlesey, Richard Ciolek-Torrello, and James Altschul, pp. 17-28. SRI Press, Tucson. 1998b Landscapes and Lives Along the Lower Verde River. In Vanishing River, Landscapes and Lives of the Lower Verde Valley, edited by Stephanie Whittlesey, Richard Ciolek-Torrello, and James Altschul, pp. 703-721. SRI Press, Tucson.

Whittlesey, Stephanie M., and J. Jefferson Reid

299

1982 Analysis of Inter-Assemblage Variation and Settlement System Reconstruction. In Cholla Project Archaeology Volume 2: The Chevelon Region, edited by J. Jefferson Reid, pp. 151-179. Archaeological Series No. 161. Cultural Resource Management Division, Arizona State Museum, University of Arizona, Tucson.

Wiessner, Polly 2009 Experimental Games and Games of Life among the Ju/’hoan Bushmen. Current Anthropology 50(1):133-138.

Wilcox, David R. 1978 The Theoretical Significance of Field Houses. In Limited Activity and Occupation Sites, edited by Albert E. Ward, pp. 25-32. Center for Anthropological Studies, Albuquerque.

Wilcox, David R., and Jonathan Haas 1994 The Scream of the Butterfly: Competition and Conflict in the Prehistoric Southwest. In Themes in Southwestern Prehistory, edited by George J. Gumerman, pp. 211-238. School of American Research Press, Santa Fe.

Wilcox, David R., David A. Gregory, and J. Brett Hill 2007 Zuni in the Pueblo and Southwestern Worlds. In Zuni Origins: Toward a New Synthesis of Southwestern Archaeology, edited by David A. Gregory and David R. Wilcox, pp. 165-209. University of Arizona Press, Tucson.

Willey, Gordon 1953 Prehistoric Settlement Patterns in the Virú Valley, Peru. Bureau of American Ethnology Bulletin 155. Washington, D.C.

Williams, Michael 1994 The Relations of Environmental History and Historical Geography. Journal of Historical Geography 20(1):3-21.

Wills, W. H., and Wetherbee Bryan Dorshow 2012 Agriculture and Community in Chaco Canyon: Revisiting Pueblo Alto. Journal of Anthropological Archaeology 31:138-155.

Wilshusen, Richard H., and Scott G. Ortman 1999 Rethinking Pueblo I Period in the San Juan Drainage: Aggregation, Migration, and Cultural Diversity. Kiva 64:369-399.

Wilson, John P. 1969 The Sinagua and Their Neighbors. Unpublished Ph.D. dissertation, Harvard University.

300

Winter, Joseph C. 1976 Hovenweep 1975. Archaeological Report 2. Department of Anthropology, San Jose State University, California.

Wood, J. Scott, and Martin E. McAllister 1984 Second Foundation: Settlement Patterns and Agriculture in the Northeastern Hohokam Periphery, Central Arizona. In Prehistoric Agricultural Strategies in the Southwest, edited by Suzanne K. Fish and Paul R. Fish, pp. 271-290. Anthropological Research Papers 33. Arizona State University, Tempe.

Woodbury, Richard B. 1961 Prehistoric Agriculture at Point of Pines, Arizona. Memoirs of the Society of American Archaeology, 17. Salt Lake City.

Wu, Jianguo, John L. David, and George D. Jenerette 2003 Linking Land Use Change with Ecosystem Processes: A Hierarchical Patch Dynamics Model. In Integrated Land Use and Environmental Models, edited by Subharajit Guhathakurta, pp. 99-119. Springer, Berlin.

Zedeño, M. Nieves 1997 Landscapes, Land Use, and the History of Territory Formation: An Example from the Pueblo Southwest. Journal of Archaeological Method and Theory 4:67-103. 2000 On What People Make of Places: A Behavioral Cartography. In Social Theory in Archaeology, edited by Michael B. Schiffer, pp. 97-111. University of Utah Press, Salt Lake City.

Zedeño, M. Nieves, Diane Austin, and Richard Stoffle 1997 Landmark and Landscape: A Contextual Approach to the Management of American Indian Resources. Culture and Agriculture 19:123-129.

Zimmerman, Larry J., and Joe Alan Artz 2006 Scale, Model Complexity, and Understanding: Simulation of Settlement Processes in the Glenwood Locality of Southwestern Iowa, 1976 and 2000. In Confronting Scale in Archaeology: Issues of Theory and Practice, edited by Gary Lock and Brian Leigh Molyneaux, pp.129-144. Springer, New York.