MIDDLE TO LATE HOLOCENE STREAM DYNAMICS OF THE SANTA CRUZ RIVER, TUCSON, ARIZONA: IMPLICATIONS FOR HUMAN SETTLEMENT, THE TRANSITION TO AGRICULTURE AND ARCHAEOLOGICAL SITE PRESERVATION by Andrea Kelly Lee Freeman Copyright © Andrea Kelly Lee Freeman 1997 A Dissertation Submitted to the Faculty of the DEPARTMENT OF ANTHROPOLOGY In Partial Fulfillment of the Requirements For the Degree of DOCTOR OF PHILOSOPHY In the Graduate College THE UNIVERSITY OF ARIZONA 1997 2 THE UNIVERSITY OF ARIZONA GRADUATE COLLEGE As members of the Final Examination Committee, we certify that we have read the dissertation prepared byAndrea Kelly Lee Freeman entitled Middle to Late Holocene Stream Dynamics of the Santa Cruz River, Tucson, Arizona: Implications for Human Settlement, the Transition to Agriculture, and Archaeological site Preservation. and recommend that it be accepted as fulfilling the dissertation requirement for the Degree of Doctor of Philosophy AlArr C. V Hayn z) Date / ,/ /9 1D ate Date Date Date Final approval and acceptance of this dissertation is contingent upon the candidate's submission of the final copy 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. Dissertation Director C. Vance Hay es, Jr. / 27/ Date 99/ 3 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 copyright holder. SIGNED: <!!7,2---'-' 4 ACKNOWLEDGEMENTS Research for this dissertation was funded by the U. of A. Graduate College, ADoT, the City of Tucson, the Arizona Archaeological and Historical Society, and the Center for Desert Archaeology. Radiocarbon assays were provided by Beta Analytic, Inc. and the University of Colorado Institute for Arctic and Alpine Research. My sincere gratitude goes to Elizabeth Black and Dena McDuffie for formatting and editing and to Elizabeth Gray, Rob Ciaccio, and Catherine Gilman for drafting. Additional mapping and drafting was provided by GEO-MAP, Inc. with particular recognition to Jim Holmlund for long hours and friendly advice when needed. Stratigraphic pits were excavated by Paul Giacomino (Desert Diggers) and Dan Arnit (Innovative Excavating). Josh Edwards, Todd Schmitz, Todd Surovell, and Vance Haynes' "Late Quaternary Geology" class aided me in geologic mapping. Many friends and colleagues have stimulated my intellectual development and supported me with friendship and love. I would like to thank the staff of Desert Archaeology, Inc. (DAI), the founding members of the Quaternary Alliance (Qal), and the EGS for their support. Helga Wi5cherl, Chaz Tompkins, Louise Senior, and Dunbar Birnie allowed me share in the joy of raising their daughters and watching them grow. The unconditional love and cherished phrases from my friends, April Birnie and Anwyn Tompkins, rescued me many times with a sense of calm and simplicity that no one else could muster. Six four-footed friends provided occasional companionship: Zack, Bones, Amos, Lola, Georgia, and Ripley. Numerous bipeds deserve a special note of appreciation for devoted encouragement: S. Bierwirth, K. Harry, B. Miksa, K. Nicoll, M. Slaughter, K. Tankersley, R. VanDyke, L. Young, and T. Young. For those left out, please know that you are in my heart and mind, but not the regulations of the Graduate College. Field discussions with E. Hajic, B. Huckell, G. Huckleberry, A. Meglioli, P. Pearthree, B. Roth, K. Vincent, M. Waters, and colleagues at DAI helped me to refine my thinking on the relationship between streams and archaeology. My committee, Vance Haynes, John Olsen, and Steve Kuhn, greatly improved the content and format of this thesis. All flaws or omissions in this dissertation are expressly my own. Additional members of my examination committee, David Meltzer, William Longacre, Vic Baker, and Owen Davis helped me through many difficult and unpleasant lessons. My advisor, Vance and "ldr," David provided me with the rare (deserved) scowl and unswerving belief in my abilities. Two other professors provided appropriate words and inspiration during especially trying times: C. Kramer and M. Schiffer. My strength to continue in this pursuit came from training and support I received early in my life. For that unique gift, I thank my former swim coaches and my parents, Hilda and Ralph Freeman, who taught me to break down walls. Finally, this research would not have been initiated without the support of my employer, Bill Doe lle. Bill cajoled me into writing this dissertation and provided the support for me to do so. I can never thank him enough. If I can ever be worthy of the faith that those mentioned above have had in me, my life will be fulfilled. 5 DEDICATION To my grandparents... David Carlson Grace Carlson Marvel Carlson Samuel Freeman E. Joan Freeman ...from whom I inherited my life's geography. 6 TABLE OF CONTENTS LIST OF ILLUSTRATIONS 10 LIST OF TABLES 12 ABSTRACT 13 CHAPTER 1: INTRODUCTION 14 STATEMENT OF THE PROBLEM 15 DEFINITION OF THE STUDY AND FOCUS AREA 17 ALLUVIAL GEOLOGY AND ARCHAEOLOGY IN SOUTHERN ARIZONA 22 Arroyo Cutting and Filling 24 Preceramic Archaeology in the Tucson Basin 31 CHAPTER 2: MIDDLE AND LATE HOLOCENE ARCHAEOLOGY IN SOUTHERN ARIZONA INTRODUCTION THE SOUTHWESTERN "ARCHAIC" San Dieguito and Pinto-Gypsum Amargosa The Cochise Culture The Desert Culture PICOSA, Oshara, and the Southwestern Archaic THE MIDDLE ARCHAIC Why So Few Middle Archaic Sites? Characteristics of the Southwestern Middle Archaic Middle Archaic Sites in Southern Arizona Status of the Middle Archaic in Southern Arizona New Hallmarks of the Southern Arizona Middle Archaic Middle Archaic Site Types and Site Distributions THE LATE ARCHAIC/EARLY AGRICULTURAL PERIOD Investigations of Late Archaic/Early Agricultural Sites in Southern Arizona GEOCLIMATIC MODELS FOR ARCHAIC PERIODIZATION Antevs' Altithermal Correlate Human Responses Altithermal Abandonments? MIDDLE TO LATE ARCHAIC TRANSITION 40 40 43 44 45 46 48 48 49 50 52 56 69 70 72 73 74 80 81 82 84 88 TABLE OF CONTENTS (continued) CHAPTER 3: HISTORIC STREAM BEHAVIOR AND HYDROLOGIC PROCESSES OF THE SANTA CRUZ RIVER INTRODUCTION PALEOCOMPETANCE RECONSTRUCTIONS Desert Stream Processes Sediment Characteristics Summary HISTORICAL RECONSTRUCTIONS Historic Descriptions of the Santa Cruz River REGIME-BASED RECONSTRUCTIONS San Xavier Reach Tucson Reach Cortaro Reach MODELLING PREHISTORIC CHANNEL BEHAVIOR 7 91 91 93 93 95 97 97 99 104 105 106 107 108 CHAPTER 4: GEOARCHAEOLOGICAL BACKGROUND OF THE SANTA CRUZ RIVER WITHIN THE TUCSON BASIN 110 OVERVIEW OF THE SANTA CRUZ RIVER TERRACES ........ . 111 QT5 and QT4 (The Univerisity and Cemetary terraces) 112 Qt3 (the Jaynes terrace) 113 Qt2 (the Holoecene terrace) 113 Qt 1 (the Historic terrace) 115 SAN XAVIER 116 Unit I 117 Unit II 117 Unit III 118 Unit IV 118 Unit V 118 Unit VI 119 Unit VII 119 RIO NUEVO SOUTH/A-MOUNTAIN 119 Clearwater Site (AZ BB:13:6) 120 Project Area Background 124 Methods 126 Stratigraphy 127 Correlation 138 Summary 140 ALAMEDA STREET 141 Stratigraphy 142 RILLITO FAN 146 Rillito Fan Site (AZ AA:12:788) 147 Methods 148 TABLE OF CONTENTS (continued) Results of Geomorphic Study Summary INA ROAD AZ AA:12:111/688 AZ AA:12:130 AZ AA:12:503 Geologic Background Previous Geochronological Work in the Project Area Recent Excavation at AA: 12:503 (ASM) by Statistical Research Recent Excavation at AA:12:130 (ASM) by SWCA Results of Testing Along Ina Road Correlation Summary SILVERBELL ROAD SUMMARY 8 149 157 159 160 160 161 162 164 165 165 167 168 169 170 171 CHAPTER 5: ALLUVIAL STRATIGRAPHY, GEOCHRONOLOGY, AND GEOMORPHOLOGY OF THE SANTA CRUZ BEND AND JUHAN PARK SITES INTRODUCTION Santa Cruz Bend (AZ AA:12:746) Stone Pipe (AZ BB:13:425) Square Hearth (AZ AA:12:745) Juhan Park (AZ AA:12:44) Implications of Excavations at the Santa Cruz Bend, Stone Pipe, and Square Hearth Sites GEOLOGICAL INVESTIGATIONS Description of the Area The Santa Cruz Bend Profile Sediments Revealed in Juhan Park Archaeological Testing The Juhan Park Profile SUMMARY CHAPTER 6: ALLUVIAL STRATIGRAPHY, GEOCHRONOLOGY, AND GEOMORPHOLOGY OF THE LOS POZOS SITE INTRODUCTION THE LOS POZOS SITE (AZ AA:12:91, ASM) The Middle Archaic Component The Early Agricultural Period Component METHODS ALLUVIAL STRATIGRAPHY West Side Stratigraphic Trench 173 173 175 176 177 177 178 179 180 181 191 194 200 202 202 203 205 208 214 214 216 TABLE OF CONTENTS (continued) East-Side Excavation GEOCHRONOLOGY AND CORRELATION Intrasite Geochronology and Correlation Intersite Geochronology and Correlation GEOMORPHOLOGY Environmental Implications of Channel Changes Implications for Human Settlement and Site Preservation SUMMARY CHAPTER 7: THE IMPACT OF MIDDLE HOLOCENE STREAM DYNAMICS ON HUMAN SETTLEMENT, THE TRANSITION TO AGRICULTURE, AND ARCHAEOLOGICAL SITE PRESERVATION SUMMARY OF MIDDLE HOLOCENE STREAM DYNAMICS Long-Term Climatic Changes and the Geologic Record of the Santa Cruz River Short-Term Climatic Changes or Small Scale Threshold Breaches HUMAN SETTLEMENT THE TRANSITION TO AGRICULTURE Models of the Transition to Agriculture New Chronology for Maize The Relationship between Stream Changes and the Adoption of Agriculture ARCHAEOLOGICAL SITE PRESERVATION LEVELS OF ARCHAEOLOGICAL AND ENVIRONMENTAL INTERPRETATION 9 223 229 229 234 236 240 242 243 245 245 247 249 251 252 252 254 259 260 261 APPENDIX A: KEY TO GEOLOGIC MAP UNITS IN FIGURE 4.1 263 APPENDIX B: SEDIMENT DESCRIPTIONS FOR FIGURE 6.3 266 APPENDIX C: SEDIMENT DESCRIPTIONS FOR FIGURE 6.4 270 REFERENCES 280 10 LIST OF ILLUSTRATIONS FIGURE 1.1 Location of study and focus areas 18 FIGURE 1.2a Location of sites and relevant geologic cross-sections within the study area 19 FIGURE 1.2b Location of sites and relevant geologic cross-sections with the study area 20 FIGURE 2.1 Distribution of archaeological sites in the study area and their location on the Holocene (Qt2) terrace 77 FIGURE 4.1 Location of Clearwater site showing surficial geologic deposits and relevant Rio Nuevo and Alameda Street cross-sections, A-A', B-13 1. C-C 1 125 FIGURE 4.2 Location of allostratigraphic units identified at the Clearwater site and relevant trench locations referred to in text 128 FIGURE 4.3 Stratigraphic column south wall of Trench 112 129 FIGURE 4.4 Stratigraphic profile of Trench 12, showing the relationship between allostratigraphic unit 1 and possible allostratigraphic unit 3 130 FIGURE 4.5 Profile of trenched area showing a north-to-south cross-section of the alluvial soils (from Elson and Doe lle 1987) 133 FIGURE 4.6 Profile of south wall of Clearwater site stratigraphic pit 135 FIGURE 4.7 Generalized stratigraphic profile of the Clearwater site 137 FIGURE 4.8 Generalized stratigraphic diagram of Alameda Street 143 FIGURE 4.9 Surficial geology of the Rillito Fan project area (after McKittrick 1988) showing location of the Rillito Fan site and trenches described in text 147 FIGURE 4.10 Historic channel locations along Rillito Creek (after Graf 1984) . 154 11 LIST OF ILLUSTRATIONS (continued) FIGURE 4.11 Surficial geology in the Ina Road area (after McKittrick) showing relevant archaeological sites and geologic cross-sections 163 FIGURE 4.12 Generalized diagram of alluvial stratigraphy along Ina Road (from Haynes and Huckell 1986) 168 FIGURE 5.1 Surficial geology of the Juhan Park and Santa Cruz Bend site areas showing relevant geologic cross-sections and stratigraphic profiles 174 FIGURE 5.2 Generalized stratigraphic columns at the Santa Cruz Bend and Square Hearth sites, showing inferred correlation of stratigraphic units by Huckell (1996) 186 FIGURE 5.3 Generalized stratigraphic columns at the Santa Cruz Bend and Square Hearth sites, showing alternate interpretation of correlation between stratigraphic units 190 FIGURE 5.4 Location of excavated trenches at the Juhan Park site . . . . . . . . . 193 FIGURE 5.6 Stratigraphic column for Juhan Park profile 1 195 FIGURE 5.7 Juhan Park profile 6 196 FIGURE 6.1 Surficial geology of the Los Pozos and Wetlands site areas (after McKittrick 1988) showing relevant trenches and excavation areas 204 FIGURE 6.2 Map of Wetlands site archaeological excavations 211 FIGURE 6.3 Profile of east side excavation 215 FIGURE 6.4 Profile of west side stratigraphic trench in pocket FIGURE 6.5 Generalized stratigraphic diagram of the west-side stratigraphic pit at Los Pozos site 217 FIGURE 7.1 Calibrated radiocarbon age ranges on early maize in the American Southwest 257 12 LIST OF TABLES TABLE 1.1 TABLE 2.1 Periodization and chronology of Santa Cruz Valley-Tucson Basin prehistory (adapted from Mabry 1996) 34 Presence of carbonized maize remains at the Middle Santa Cruz River sites 78 TABLE 4.1 Correlation between geologic units defined by different researchers 112 TABLE 4.2 AMS radiocarbon dates from A-Mountain mitigation features . . . 123 TABLE 4.3 Stratigraphic descriptions from the south wall of Trench 10, AZ AA:12:788 150 Stratigraphic descriptions from the north wall of Trench 14, AZ AA:12:10 151 TABLE 4.4 TABLE 5.1 Radiocarbon dates on alluvial stratigraphic units at the Santa Cruz Bend and Square Hearth sites (as reported by Huckell 1996) . . . 183 TABLE 5.2 Alternate division of alluvial stratigraphic units at the Santa Cruz Bend and Square Hearth sites TABLE 6.1 TABLE 6.2 TABLE 6.3 191 Radiocarbon dates from the Middle Archaic component at Los Pozos 206 Radiocarbon dates from the Early Agricultural component at Los Pozos 209 Radiocarbon dates from the Wetlands site 213 TABLE 6.4 Additional radiocarbon dates from the Los Pozos site 218 TABLE 6.5 Estimated age of stratigraphie units at Los Pozos site 233 TABLE 7.1 Earliest radiocarbon dates on cultigens in the Southwest 255 13 ABSTRACT Historic records of arroyo formation have long been used as inferential tools for reconstructing paleoclimate in the American Southwest. These paleoclimatic reconstructions have attempted to demonstrate that synchronous incision of river valleys across the American Southwest was the result of large-scale (regional, global) climatic change. Projected to the past, the inferred chronological boundaries of certain climatic periods have been used by archaeologists as convenient boundaries for demarcating long-term changes in human settlement and subsistence. The rapid accumulation of new data on middle to late Holocene subsistence and settlement along the Santa Cruz River, and the application of new theoretical constructs in hunter-gatherer research require the use of higher resolution data in geoarchaeology. During the past ten years, advances have been made in our understanding of the hydroclimatological processes which cause channel changes on the Santa Cruz River and geologists are now better able to predict the circumstances under which desert streams become arroyos. Together with high-resolution geologic documentation of channel exposures, the prehistoric setting of human occupation along the Santa Cruz River can be addressed at a scale that is more relevant to the archaeological issues of today. The detail derived addresses specific geomorphic and paleoenvironmental variables that operate at the site or regional level and that have the most direct effect on human decision-making. 14 CHAPTER 1 INTRODUCTION Though we often think of geologic resources as fixed, they are not. Geologic resources and features change over time under the force of natural geologic processes. These geologic processes are initiated by extrinsic and intrinsic forces (or agents), which include climatic, tectonic, and eustatic forces, biological and chemical agents, wind, and water. The geologic features created by these processes form the context for all past and living human groups and influence the lithologic, hydrologic, and biological resources that are available to these groups. Of all geologically-formed features, rivers may have the most significant impact on human decision-making, particularly in prehistory. Rivers provide at least one critical resource to those groups — water. Rivers form boundaries to human movement and they provide reliable resources for migratory human groups. They provide stable, reliable soils on which to grow crops, but they also alter the landscape on which human groups attempt to live. For archaeologists, rivers provide both the context for archaeological sites and a force that sometimes removes those sites. To some extent, geoarchaeology is an opportunistic discipline, because it exploits the geologic context in which archaeological resources are found' to address archaeological research issues. Fluvial processes are some of the most active geologic agents and as a result preserve short-term records of human activity that can span hundreds or thousands of years. These archaeological records provide chronological 15 markers by which to measure the activities of rivers and the preserved effects of fluvial processes. Thus, geoarchaeology can be used as a geoscientific, as well as an archaeological, research tool. Deposits left by rivers and the archaeological resources associated with those deposits are interesting to geologists, geoarchaeologists, and archaeologists, all of whom have had a lengthy relationship with the investigation of archaeological sites in river settings. STATEMENT OF THE PROBLEM This dissertation explores the relationship between geologic processes and human occupation of the middle Santa Cruz River, Tucson, Arizona, during the middle to late Holocene. The geologic period encompassed by this study overlaps the archaeological periods known as the Middle and Late Archaic. Inferred climatic changes and the geologic indicators of those changes during the transition from the middle to late Holocene have been used by archaeologists as convenient boundaries for demarcating long-term changes in human settlement and subsistence patterns, including the transition to agriculture. The rapid accumulation of new data used to address such issues from Middle and Late Archaic sites in southern Arizona and the application of new theoretical constructs in hunter-gatherer research require the use of higher resolution data in geoarchaeology. During the past ten years, advances also have been made in our understanding of the hydroclimatological processes which cause channel changes on the Santa Cruz River (Parker 1995). Using data from historic records and recent flooding events in 16 conjunction with climatic modeling, geologists have been better able to predict the circumstances under which desert streams become arroyos (Betancourt and Turner 1990; Parker 1995). These data can be used to understand the prehistoric setting of human occupation in the Tucson Basin at a scale that is more relevant to the archaeological issues being addressed today. A more complete understanding of prehistoric stream behavior can also be used to predict the potential for additional, yet undiscovered, archaeological sites within the prehistoric floodplain of the Santa Cruz River. Utilizing prehistoric records of channel change and archaeological evidence for Middle and Late Archaic sites, this dissertation seeks to evaluate the impact that current and future discoveries of Middle Archaic sites have on our understanding of middle to late Holocene prehistory in the Tucson Basin. The dynamic record left by the Santa Cruz River is used to reconstruct environmental variables important to understanding prehistoric use of the riverine setting, and to predict the potential for discovery of additional Middle and Late Archaic sites within the floodplain of the Santa Cruz River. The dissertation will address the following general issues: What was the nature of middle to late Holocene floodplain development along the Santa Cruz River? How are those stream processes related to the preservation of middle to late Holocene archaeological sites? What was the nature of human settlement and subsistence along the Santa Cruz 17 during the middle to late Holocene? How did channel change during this period influence human decisions regarding settlement and subsistence? • Was the Santa Cruz floodplain environment during the middle Holocene conducive to the introduction of cultigens and is there support for the presence of cultigens during this period? DEFINITION OF THE STUDY AND FOCUS AREAS In order to address the above general issues, the middle Santa Cruz River surrounding Tucson, Arizona, has been subdivided into two areas: the study area and the focus area (Figure 1.1). The study area refers to the stretch of the Santa Cruz River from A-Mountain (Sentinel Peak) to Ina Road. The study area represents the portion of the Santa Cruz that has been investigated by the author. This area has been sampled (i.e., prehistoric alluvium has been documented) rather thinly, providing low-resolution data for portions of the study area. With the exception of the Clearwater site, archaeological sites within the study area but outside the focus area have also been thinly sampled (i.e., tested rather than mitigated). The results of this geologic and archaeological sampling effort are presented in Chapter 4. Sites and relevant geologic sections are presented in Figure 1.2. A portion of the Santa Cruz outside the study area from Pima Mine Road to A-Mountain has been investigated previously by other authors (Haynes and Huckell 1986; Stafford 1986; Waters 1987, 1988) and the results of their work are also reviewed in Chapter 4. 18 Figure 1.1. Location of study and focus areas. 19 J (i" ow....... s,,,0.< .• t . ,2 I " e t <-1 Cf5•t7"----.‘• 11, f „.... .7...,...,) 1 ,,. i /2/7„.--:-....„,. i .: „.., 1 ( fsififtgoN% \,i0 i ',';''',..:,.. ,<> ! \ TUMAMOC HILL ,;.)» ) v V 1 % ‘Ï‘.̀Z,; ., : ff (.•' 1 „ '' ; --,\\ , : ":.4 1 1 —6P ) • , a ,----.• . •‘<. : > -; . , \--------",J L ' \...... — 2 ss, ' ,, •, ,-, 1 , • ..,/,, A MOUNTAIN . EXPLANATION contours ..."'"•• river bank profile location wash road SITE REFERENCES 1 AZ AA:12:745 — square hearth stratigraphic pit 2 AZ AA:12:746 — Santa Cruz Bend PIMA COUNTY - TUCSON, ARIZONA LOCATION OF STRATIGRAPHIC SECTIONS IN THE SANTA CRUZ RIVER FLOODPLAIN BETWEEN A-MOUNTAIN AND PRINCE ROAD contour interval — 100 feet contour interval below 2400 feet — 20 feet 3 AZ AA: 12:746 — Santa Cruz Bend, river bank profile 4 AZ AA:12:44 — Juhon Park profile 6 5 AZ AA:12: 44 — Juhan Park profile 1 B AZ BB:13:16 — Clearwoter site 0 1 .0 2.0 km EMI=n_ 0 0.5 1.0 mi Digital cartography by CEO—MAP, Inc. 1997 Figure 1.2a. Location of sites and relevant geologic cross-sections within the study area. 20 EXPLANATION contours wash road SITE REFERENCES 1 AZ AA:12:130 2 AZ AA:12: 503 3 AZ AA:12:111/688 4 AZ AA:12:10 — Trench 14 Sunset Mesa PIMA COUNTY TUCSON, ARIZONA contour interval — 100 feet contour interval below 2400 feet — 20 feet 5 AZ AA:12:788 — Trench 10 KR() Fan 6 AZ AA:12: 91 — west side stratigraphic pit — L9CATION OF STRATIdRAPHIC SECTIONS IN THE SANTA CRUZ RIVER FLOODPLAIN BETWEEN PRINCE ROAD AND INA ROAD 0 1.0 2.0 km 7 AZ AA:12: 91 — Los Pozos excavations 8 AZ AA: 12:91 — Los Pozos east side excavations 9 AZ AA:12:90 — Wetlands site 0 0.5 .1.0 mi Digital cartography by CEO—MAP, Inc. 1997 Figure 1.2b. Location of sites and relevant geologic cross-sections within the study area. 21 The focus area, which is inset within the study area, refers to the stretch of the Santa Cruz River from Grant Road to Ruthrauff Road. The focus area has been intensively documented, geologically and archaeologically (through site mitigation), providing high-resolution data for a specific portion of the river. The results of research within the focus area are presented in Chapters 5 and 6. Although both the study area and the focus area have been selected, in part, on the basis of the needs of cultural resource management, the areas are quite appropriate for the issues presented in this study. In an attempt to provide a geomorphic and hydrologic explanation for channel change on the Santa Cruz, Parker (1995) has identified certain landforms within the middle Santa Cruz that provide geologic and topographic controls on channel changes. His study reaches are defined on the basis of those landforms. The three reaches that he defines within the Tucson Basin are either partly or fully covered by the research in the study reach and by the efforts of previous investigators (Haynes and Huckell 1986; Stafford 1986; Waters 1987, 1988) outside the study reach. The focus reach samples a limited portion of what Parker (1995) has identified as the Tucson reach of the middle Santa Cruz. A discussion of Parker's (1995) defined reaches is provided in Chapter 3, along with a discussion of the data used to infer the prehistoric character of the Santa Cruz from documented prehistoric alluvium. Research within the study and focus areas provides two different scales of data from which to interpret the relationship between stream processes and prehistoric human decision-making. General trends can be inferred from low-resolution data 22 within and outside the study reach, while specific, higher frequency processes are documented within the focus reach. The scale at which archaeologists and geologists study the past is related to progress made in data acquisition and data interpretation (Stein 1993). During the past ten years, both geologic and archaeological research have made considerable progress, allowing geoarchaeological research to better address issues in both sciences at an appropriate scale. The remainder of this chapter briefly charts the history of progress made in the interpretation of alluvial stratigraphy, the relationship between fluvial processes and climatic change, and the acquisition of additional data on the Middle and Late Archaic periods in southern Arizona. Additional background information related to the Middle to Late Archaic transition in southern Arizona is addressed in Chapter 2. Chapter 3 covers new data on the relationship between channel changes and hydroclimatological processes on the Santa Cruz River. ALLUVIAL GEOLOGY AND ARCHAEOLOGY IN SOUTHERN ARIZONA The investigation of archaeological sites in the American Southwest has established a long and close relationship with the geosciences. The two most influential persons in the early study of Southwestern geoarchaeology were Ernst Antevs and Kirk Bryan. Both scientists were interested in the relationships between climatic change and geologic processes and utilized archaeological resources as a means of measuring these changes. Both scientists also found a place for this research in the American Southwest. Kirk Bryan's influence on Quaternary geology and 23 archaeology in the Southwest was passed on from generation to generation by his students (Haynes 1990), while Antevs began a lifelong collaboration with Ted Sayles that would define the Cochise culture (Haynes 1990) and set preceramic geoarchaeology in southern Arizona on its current path. As time passed, more geologists became interested in the records left by fluvial systems. During the 1960s, Quaternary scientists, many of them Bryan's students (Haynes 1990), struggled to understand the processes that cause changes in river valleys. Detailed observations of these processes in field and laboratory settings enabled them to better understand the mechanics of fluvial geomorphology (Leopold et al. 1964). Interest in field documentation of processes subsided with the development of computers and their use in simulation modeling. The application of computer-generated models to the geoscientific study of hazard reduction led scientists interested in fluvial geomorphology away from studies of the causes of catastrophic events in river systems to mathematically-based management and prediction of future flood events (Baker 1988a, b). Questions of scale became even more important when geologists realized that predictive modeling based on the historical record was poor at predicting global changes in fluvial systems (Baker and Twidale 1991). Research in planetary geomorphology, a product of the 1980s and 1990s, has attempted to address issues of scale. In response to concerns about the nature of global climatic changes and unusually high frequency of extreme flooding events, geologists have attempted to extend their predictive models back into the past, a process that involves the field 24 application of new knowledge. The one issue that pervades attempts to extend predictive models into the past has come full-circle back to the research of Antevs and Bryan — the cause of arroyo cutting and filling. Arroyo Cutting and Filling One of the predominant themes throughout Bryan's and Antevs' research was the cause of arroyo cutting and filling and the relationship of these events to climatic change (Antevs 1936, 1952; Bryan 1925, 1927, 1928, 1940). Interest in these processes was a result of the fact that most large river valleys across the Southwest became entrenched during the period 1865-1915. The relative synchroneity of this event across the region caused scientists to speculate about the cause of arroyo cutting. Discussion of the "arroyo problem" was highly charged and the interests of those involved were usually intertwined with their conclusions (Cooke and Reeves 1976; Haynes 1968, 1990). ...there is a certain correlation between the professional interests of investigators and the conclusions they reach on the causes of arroyo cutting. Agriculturalists, foresters, and conservationists commonly indict man for his excesses. In contrast, some geologists, paleontologists, and archaeologists have sought and found "natural" explanations (Cooke and Reeves 1976:6). Because the events preceding historic entrenchment included a variety of human and natural agencies, the causes of this event were the subject of intense debate. A comprehensive discussion of this debate is found in Cooke and Reeves (1976). Recognizing this implicit warning against ignoring other causes 2, the following 25 overview of the literature briefly touches on the arroyo cutting and filling debate and covers only the parameters encompassed by geomorphic, hydrologic, and climatic processes. Arroyo cutting and filling has been the focus of studies of prehistoric alluvium and is critical to understanding the history of research preceding this study. Over the past 75 years, significant progress has been made in understanding the geomorphic and hydrologic processes involved in arroyo cutting and filling and the significance of these events. Because much of this literature has been covered in previous reviews (Betancourt 1990; Betancourt and Turner 1990; Cooke and Reeves 1976; Graf 1983; Webb 1985), only a cursory examination is presented here. Climatic Changes and Arroyo Incision During the early part of this century, most geologists agreed that arroyo cutting was a synchronous event across the American Southwest, and geologists studying past arroyo cutting events believed that entrenchment was caused predominantly by climatic change'. The question boiled down to whether the climatic change that caused arroyo cutting represented a shift to a wet or a dry period. Original arguments supporting the wet-period incision hypothesis were drawn out of the flawed concept that stream gradient was a function of "balance between erosion and deposition" of sediments (Davis 1902:86). Huntington (1914) applied this concept to arroyo forming processes by arguing that aridity would increase gradients causing transportation of sediments and aggradation, while humidity would decrease the gradient of a stream resulting in entrenchment. He contended further that the loss 26 of vegetative cover during dry periods would overload streams with alluvium, while improved vegetative cover would reduce sediment loads, causing clearer, more erosive flows. Huntington's concept would soon take a subordinate position to Bryan's dry-period incision hypothesis. In his study of the Rio Puerco, Bryan (1928) advanced the argument that dry periods would be accompanied by a reduction of vegetative cover resulting in increased and more powerful runoff which would initiate gullying in critical reaches. Bryan's concept derived from an understanding that stream gradients were not necessarily in equilibrium. He further contended that discontinuous arroyos were integrated by headward migration of cutting during subsequent floods, forming continuously entrenched arroyos. Bryan's concepts were soon adopted by other alluvial stratigraphers (Antevs 1952; Euler et al. 1979; Hack 1939; Haynes 1968; Leopold and Miller 1954) to support the idea that two periods of synchronous arroyo cutting in Holocene prehistory represented dry periods (the Altithermal and the "Great Drought"). The application of radiometric dating to alluvial sequences supported the timing of these arroyo cutting events (Haynes 1968). Some of these studies of prehistoric alluvium have suggested that lowered groundwater during drought periods would increase the erodability of unconsolidated sediments near streams (Euler et al. 1979; Haynes 1968). But understanding the relationship between climate and erodability of sediments was still considered much more complex. In addition to the difficulty in identifying specific cause-and-effect relationships, the scale and complexity of 27 relationships between precipitation, temperature, soils, water table, plant growth, and other biological activity was still poorly understood. The simple dichotomy created by earlier studies was considered too restrictive. The concept that united earlier studies (synchroneity) would be put to the test in future research. Synchroneity vs. Asynchroneity The apparent synchroneity of arroyo incision across the Southwest during the historic period allowed scientists to utilize the concept of synchronous incision to reconstruct past river behavior. It is important to note the scale at which climatological changes were supposed to have driven changes in hydrologic processes. Most of the studies cited above, and a later study relating Holocene ice advances to valley degradation (Brackenridge 1980), deal with processes occurring on the order of several millennia. That synchronous incision caused by climatic changes has occurred during certain periods is undeniable (Knox 1995); however, not all channel cutting events are caused by climatic changes. Evaluation of both the scale and timing of these events is critically important to interpreting whether or not they are synchronous. In his well-documented case for synchroneity, Haynes (1968) demonstrated that the periods of entrenchment were short compared to the periods of aggradation. The synchroneity of cutting events is also more difficult to evaluate than the synchroneity of filling events. This is partly due to conditions of preservation and the presence of material evidence for the timing of cutting events. Soil formation can continue for long periods on an unincised surface. Further, Haynes (1968:613) suggests that the 28 perturbations caused by climatic changes resulted only in trends favoring one process over another: This is not to say that either erosion or deposition began everywhere at the same time. It does suggest, however, that there were periods when, throughout the Southwest, processes of aggradation were dominant and other periods when degradation predominated. However, probabilistic treatments of flood hydrologic processes argue against synchronous arroyo development (Hirschboeck 1987; Knox 1983). When combined with the errors inherent in radiometric dating of these events', the localization of storm events, and the effect of response time of the fluvial system to climatic perturbations (Brackenridge 1981), the concept that synchronous processes, caused by a single climatic anomaly, could encompass an entire region becomes less compelling. Another study has demonstrated that the number, character, and chronological position of degradation and aggradation phases over the past 15,000 years are out of sequence across several valleys in southern Arizona (Waters 1985; Waters and Kuehn 1996). Rainfall, Climatic Change, and Prehistoric Incision In order to address the relationship between climatic change and arroyo incision, several scientists conducted studies specifically geared to determine the effects of rainfall on arroyo development (Leopold 1951; Leopold et al. 1966; Leopold and Miller 1954; Thorthwaite et al. 1942). The intensity of rainfall appears to be a critical factor in arroyo development. Leopold and Miller (1954) determined that monsoonal moisture is the source of most heavy rainfall in the American Southwest. 29 Martin (1963) used this concept in his interpretation of palynological data from the Southwest in order to argue that increases in summer rainfall caused arroyo incision. His work was subsequently scrutinized and overturned (Antevs 1962; Mehringer 1967). Many of these studies probably focused too heavily on vegetation, but the lasting conclusion was that rainfall intensity caused large floods and entrainment as well as entrenchment of river sediments. Floods and Hydrology Betancourt and Turner (1990) provide convincing evidence that regardless of the hydrologic and geomorphic constraints posed by human and natural activity, entrenchment in large valleys across Arizona could not have occurred without significant flood flows. These flood flows are attributed to three types of storms: frontal systems, which occur in the winter; dissipating tropical cyclones, which tend to occur in early fall; and intense, localized convective storms, which occur during the summer "monsoon" season (Hirschboeck 1985; Webb and Betancourt 1992). The entrenchment of historic arroyos throughout the American Southwest seems to have been caused by extraordinary flood events (Betancourt and Turner 1990; Gregory 1917; Hereford 1985; Huntington 1914; Love 1983; Thornthwaite et al. 1942; Tuan 1966; Webb 1985). These floods succeed in entraining channel materials because discharges exceed the threshold for erosion (Bull 1979), causing channel entrenchment. Relationships between large floods and atmospheric circulation patterns are well documented (Hirschboeck 1985, 1987; Maddox et al. 1980). Although flooding 30 events in the Southwest do cluster temporally (Ely 1992), large-scale atmospheric circulation patterns cause spatial-clustering in large flood events (Hirschboeck 1987; Knox 1983), arguing against their synchroneity. Though the time-spans in which these major floods are clustered are short, they have a significant impact on the character of desert streams and tend to cross the threshold conditions that cause perturbations in stream systems (Baker 1988; Bull 1988; Kochel 1988; Kochel et al. 1982). Parker (1995) has further elucidated this relationship, demonstrating that monsoonal moisture due to its flashy, but spatially restricted precipitation pattern does not produce discharges large enough to sustain energy through the stream system, while tropical and frontal storms are able to sustain such energy due to their widespread distribution and sustained periods of moisture. Despite the well-documented relationship between large-scale atmospheric circulation patterns and the response of fluvial systems today, the prehistoric relationship between atmospheric circulation and channel change is poorly studied. Global climatic modeling cannot accurately predict the intricacies of time-clustered flood events. So, how can prehistoric alluvial records be used to determine the effects of climatic change on stream systems? Modeling the Relationship between Hydroclimatic Processes and Channel Change Recently, Parker (1995) has attempted to utilize the physical characteristics of the river, the relationship between meteorological events and streamflow, and historic records of meteorological events, flooding, and channel change to develop a predictive 31 model for channel changes along the Santa Cruz River. He also attempts to explain Holocene alluvial history, documented near San Xavier and Ina Road, by extending this model into the past. The controls that Parker (1995) describes and the model he creates are addressed in greater detail in Chapter 3. He uses high-resolution events to develop this predictive model, but is unable to test the model because of the relatively low resolution of the Holocene stratigraphic record. Recent archaeological research in the Tucson Basin has exposed additional sections of the Holocene alluvial record allowing for higher-resolution testing of Parker's (1995) model. The transitions in human settlement and subsistence covered by this archaeological record can also be used to determine whether the results of modeling channel change can be used to address the scale of archaeological research questions being posed by middle to late Holocene archaeology. Preceramic Archaeology in the Tucson Basin Ten years ago the preceramic (Archaic) period in the Tucson Basin was poorly known and even more poorly understood. Several large surveys (Doelle 1985; Fish et al. 1985; Huckell 1984a; Simpson and Wells 1983; Tagg and Huckell 1984) had produced numerous preceramic sites in the upper bajada and surrounding piedmont. A few Archaic sites were known to exist in the basin itself (Betancourt 1978; Haynes and Huckell 1986), but fewer yet were excavated (Elson and Doelle 1987). Even less well known was the extent of Archaic archaeological resources buried under the floodplain of the Santa Cruz River. Indication that archaeological sites of preceramic 32 age existed below the floodplain consisted of occasional clusters of artifacts, fire-cracked rock, human bone, or the rare hearth, discovered in the most unfortunate of circumstances — fortuitous exposure, in landfill pits, by construction equipment, or exposure after large-scale flooding events (Betancourt 1978; Haynes and Huckell 1986). During the past six years, numerous new discoveries of preceramic remains have been made along the Santa Cruz River as part of projects to improve the interstate and other parcels of land near the river (Diehl 1996a, 1996b; Ezzo and Deaver 1996; Freeman 1995, 1996, 1997; Gregory 1995, 1997a, 1997b; Huckell 1990, 1995; Huckell et al. 1994; Mabry 1995, 1996a; Mabry and Clark 1994). Contract archaeologists have discovered large Late Archaic (also referred to as Early Agricultural Period') settlements along the Santa Cruz River dating between 3500 and 2000 B.P. These earliest farmers utilized a river that was far more reliable than the dry arroyo that exists today. In addition to the late preceramic evidence recovered from floodplain sites, the past two years have witnessed additional evidence for an earlier (Middle Archaic) occupation along the Santa Cruz River (Gregory 1995, 1997a). Though hints of this earlier occupation have been recovered through fortuitous discoveries and geologic investigations of the exposed banks of today's dry arroyo (Haynes and Huckell 1986; Huckell 1996a), recent excavation of a Middle Archaic component suggests that a Middle Holocene occupation of the Santa Cruz River did exist and may have been the precursor to an Early Agricultural period occupation of the river valley. In light of 33 this new evidence, it has become necessary to re-evaluate the nature of the Middle to Late Archaic period transition in the Tucson Basin. Archaic Periodization Discoveries of these early farming villages have caused archaeologists to re-evaluate the current cultural-historical scheme and to propose a new sequence for the Archaic period in southern Arizona (Table 1.1). Under this scheme, the Early Archaic period, which is poorly known from in southern Arizona, encompasses the period of time from 10,500 to 8000 B.P. (Mabry 1996a; but see, Haynes 1967:278; Huckell 1984:137; Huckell 1996b; Huckell and Haynes 1995; Martin 1963:57; Waters 1986b), equivalent to Sayles and Antevs, (1941) Sulphur Spring stage of the Cochise culture. The Early Archaic is defined predominantly on radiocarbon dates and artifacts recovered from Ventana Cave and Whitewater Draw, and is characterized by simple ground stone milling equipment and flaked stone implements. Early studies placed these milling tools in association with late Pleistocene megafauna (Sayles and Antevs 1941; Haury 1960); however, further study of the stratigraphy at Double Adobe suggests that the two occupations are not contemporaneous (Waters 1986a). Nevertheless, it appears that shortly after the extinction of late Pleistocene megafauna, on which Paleoindians at least partly relied, groups in southern Arizona established an economy that demonstrates a certain reliance on plant processing. A gap (8000-5000 B.P.) is represented by a period in which archaeological 34 Table 1.1 Periods Periodization and chronology of Santa Cruz Valley-Tucson Basin prehistory (adapted from Mabry 1996a). Phases Date Ranges Holocene Hohokam Classic Period Hohokam Sedentary Period Hohokam Colonial Period Tucson phase A.D. 1300--1450? Tanque Verde phase A.D. 1150--1300 Late Rincon phase A.D. 1100--1150 Middle Rincon phase A.D. 100--1100 Early Rincon phase A.D. 950--1000 Rillito phase A.D. 850--950 Canada del Oro phase A.D. 750--850 Late Hohokam Pioneer Period Early Ceramic Period Late Archaic/ Early Agricultural Period Archaic Period Snaketown phase A.D. 700--750 Sweetwater phase A.D. 675--700 Estrella phase A.D. 650--675 Tortolita phase A.D. 550--650 Agua Caliente phase A.D. 150--550 Cienega phase 800 B.C.--A.D. 150 San Pedro phase 1200--800 B.C. Chiricahua phase 3000--1200 B.C. Unnamed phase (gap?) 6000--3000 B.C. Sulphur Springs-Ventana 8500--6000 B.C. Middle Early Middle Early Paleoindian Period 10,000?--8500B.C. 35 materials are not found (Berry and Berry 1986; Huckell 1996b; Irwin-Williams and Haynes 1970), possibly due to loss of older sediments in river valleys (Waters 1986b). Preliminary analysis of data from archaeological site records indicates that the gap in human occupation is real (Mabry et al. 1997). The climatic, chronological, and archaeological implications of this apparent gap are explored further in this dissertation. The Middle Archaic period begins approximately 5000 B.P. and continues to 3500 B.P. (Mabry 1996a; but see Huckell 1996b), at which time the initial transition to agriculture is thought to have taken place. Middle Archaic groups, discussed further in Chapter 2, are denoted by the presence of diagnostic projectile points and ground stone tools that appear to be, on the whole, slightly more sophisticated than Early Archaic ground stone implements. Much of the ground stone found at Middle Archaic sites is more heavily worn, suggesting a reduction in mobility and/or perhaps a frequent return to resource areas. The implications of mobility changes are explored further in Chapter 2. From 3500 to 2000 B.P. (Mabry 1996a), large-scale agricultural occupations are believed to have begun in the American Southwest. This period has been called the Early Agricultural period (Huckell 1995) and is thought to include groups both possessing the ability and incentive to produce domesticated plants (Early Agricultural period groups) and those that remain committed to a hunting and gathering lifestyle (Late Archaic period groups). Soon after early farming villages appear, pottery production begins. The Early Ceramic period (2000 to 1500 B.P.) is defined by the 36 presence of untempered, plain ware pottery and figurines (Mabry 1995). Middle to Late Holocene Sites within the Study Area This dissertation focuses predominantly on the middle to late Holocene or Middle and Late Archaic periods in the Tucson Basin. Evidence for a Late Archaic occupation of the floodplain is abundant, but Middle Archaic sites are found in only a few locations. A possible Middle Archaic component was recorded along Ina Road (Haynes and Huckell 1986) and two others were recorded near Martinez Hill (Haynes and Huckell 1986; Huckell 1996b:338, unpublished field notes, pers. comm. 1997). The excavation of the Los Pozos site (Gregory 1997a) provides a third Middle Archaic locality and suggests that a fourth, at the Cortaro Fan site (Roth 1989), may yet exist 6. As this evidence has mounted, our knowledge of the transition between the Middle and Late Archaic periods has changed significantly. The Middle to Late Archaic Transition Though projectile points dating to the Middle Archaic period are found throughout the Tucson Basin, and hearths and occupation horizons radiocarbon dated to this period have been found buried in floodplain sediments (Haynes and Huckell 1986; Huckell 1996b), relatively little is known about this period. The "transitional" period (ca. 3500-2800 B.P.) is poorly represented in the Santa Cruz floodplain, but is known from other drainages in southern Arizona (Huckell 1990, 1995). Within some of the Middle Archaic and "transitional" features found in the Santa Cruz floodplain is 37 sporadic evidence of cultigens (represented by fragmentary corn cupules). Further investigation of the origin and context of the deposits is necessary in order to accurately assess the importance of these finds. The period following this initial occupation or reoccupation of the river valley is thought to have supported the first agriculturally-based economies in the Tucson Basin. Huckell (1990) has recently suggested that agriculture may have spread from Mexico to the American Southwest along northward trending river valleys. He supports this, in part, from the alluvial record, demonstrating that in the period following 4500-4000 B.P., river valleys in southern Arizona experienced a trend toward net aggradation. Yet, without excavation of these buried Middle Archaic components, it is difficult to determine when and how quickly the shift to an agricultural economy took place. Middle Archaic sites along the Santa Cruz River are situated along the margins of the Pleistocene terrace, in an area that is particularly conducive to the recovery of subsurface data in stratified sites because of its presence away from the active channel, an issue explored in more detail in Chapter 7. Furthermore, the large number of well-documented archaeological sites in southeastern Arizona provide a robust record of the archaeology bracketing the poorly understood Middle Archaic period. Recent study of a buried Middle Archaic component along the Santa Cruz River (Gregory 1997a) has enhanced a rather mundane picture of the Middle Archaic occupation of the Tucson Basin. The Holocene alluvial record in southeastern Arizona has not only preserved 38 the record for middle to late Holocene archaeological sites, but has provided additional data on the varieties of human activities in different physical settings. This, combined with palynological and faunal data available from excavated archaeological and geologic sites, increases the database necessary to reconstruct human activity during this period. At the same time, archaeological sites can also be used as a tool to understand the nature of arid region stream activity, its relationship to climatic variability, and the degree of synchroneity of Holocene stream processes. A portion of the dissertation will address issues critical to understanding the nature of desert stream processes, their relationship to archaeological deposits and human occupation, and will evaluate geologic models of Holocene alluvial activity. A summary of these issues is presented in Chapter 7. ENDNO TES 1. Other types of geoarchaeological study can also be opportunistic. For example, petrography exploits the geologic resources found in artifacts or from which artifacts are made in order to address archaeological issues. 2. Prehistoric cutting and filling is very likely not a product of human intervention; however, seismic events may have played a role. 3. Antevs (1936) believed that historic arroyo cutting was caused by human intervention, but that past episodes of arroyo cutting were climatically induced. 4. Haynes (1968) has developed a method of reducing errors in dating alluvial events; these include sampling from the base of each paleochannel as well as the top of each unit "where it is least eroded" (Haynes 1968:596). He also emphasizes paying close attention to the possibility of redeposition and/or sample contamination. 39 5. Huckell's (1995) primary reason for defining the Early Agricultural period is to acknowledge the quite obvious and well-defined transition to agriculture during the Late Archaic. The term, therefore, reflects an "adaptive" significance rather than a temporal one. 6. The presence of Cortaro points at Los Pozos in Middle Archaic contexts raises the question of whether those same types of points found earlier at the Cortaro Fan site were Late Archaic, as dates from the site suggest, or Middle Archaic, as dates from Los Pozos suggest. Roth and Huckell (1992) have chosen to view them as both Middle and Late Archaic in age. 40 CHAPTER 2 MIDDLE AND LATE HOLOCENE ARCHAEOLOGY IN SOUTHERN ARIZONA INTRODUCTION The term "Archaic" was first used to describe groups in eastern North America (and later in California) with a specific suite of material culture traits. These traits included the presence of shell mounds, "crude" stemmed projectile points (Haag 1942), and in some cases, early pottery (Beardsley 1948). The economic implications associated with these material culture traits included some combination of hunting and collecting (Fairbanks 1942) or hunting-gathering-fishing (Beardsley 1948, Haag 1942). By 1946, increasing desire to relate these groups to their post-Paleoindian and pre-Agricultural counterparts throughout North America led Griffin (1946) to forward a cultural-developmental framework. This framework related groups in the United States to late Paleolithic and early Neolithic groups in Europe, using the terms "Paleo-Indian" and "Neo-Indian," respectively. Willey and Phillips (1958:07) were perhaps the first to attach an explicitly adaptive definition of the Archaic, defining it as "...the stage of migratory hunting and gathering cultures continuing into environmental conditions approximating those of the present." Though they note that the Archaic postdates the extinction of late Pleistocene megafauna, their definition is too broad to be applicable today, for it would encompass (rightly or wrongly) the late Paleoindian period on the High Plains 41 and all historic hunters and gatherers. Although the concept of an Archaic period is basically understood by most North American archaeologists, the definition of it is never quite specific enough to separate it from the preceding Paleoindian period or from contemporaneous groups with different economic and social practices. The problem faced by Willey and Phillips (1958) and by other archaeologists (Caldwell 1965; Cleland 1976; Stoltman and Barreis 1983) who have attempted to segregate the "Archaic" from the preceding "Paleoindian" and numerous antecedant periods— variously defined in terms of increasing sedentism, and the adoption of agriculture and ceramic technology—is that no fixed criteria adequately describe the range of behaviors encompassed during the post-Pleistocene period. In order to segregate the Archaic from the Paleoindian period, Willey and Phillips (1958) defined the Archaic period as existing within an environment that was essentially the same as that of the present. Although the period encompassed by the Archaic (the Holocene) is in geologic terms essentially a "modern" climate, it actually represents a series of conditions, most of which were climatically very distinct from the present (COHMAP Members 1988). Though many significant changes in human activity appear to broadly follow long-term climatic changes, there may actually be some lag time (or response time) which allows significant changes in human behavior to be noticeable after long-term climatic changes are well established. The definition provided by Willey and Phillips also supplies no termination point aside from the general concept of "hunting and gathering cultures." The termination is essentially the successful integration of agriculture and well-established 42 (sedentary) village life. This very broadly defined concept was acceptable in an era when archaeologists relied heavily on cultural-historical frameworks. However, since the late 1950s, American archaeology has undergone a number of significant theoretical changes in the way that changes in human groups are viewed. Prehistoric changes in human economy, mobility, and subsistence-related behavior are viewed very differently by archaeologists in competing theoretical schools (Kelly 1995). Though today many archaeologists still use the broad and inclusive term "adaptation," implying directional evolutionary change in human behavior, most agree that understanding the complexity of those changes requires sophisticated analytical techniques based on a thorough understanding of the environmental and cultural variables about which individuals made choices. Because Archaic archaeologists in the American Southwest have so little data to work with, they are often able to skirt issues of theoretical construct, basing inferences regarding changes in human settlement and subsistence on a database that encompasses a geographically or temporally broad scale and that is comprised of many small fragments of data. As a result, their conclusions are the direct result of where in space or time they have selected to lump each of these small fragments, rather than on changes that are significant at the human scale. The result is a competition between paradigms (often north vs. south) that creates little cohesion in the study of the "Archaic." For instance, working on the Colorado Plateau, Matson (1991) and Geib (1995) would place the Middle Archaic between 6000 and 4000 B.P., while Huckell (1996b), working in southern Arizona, would place it between 5500 and 3500 B.P. 43 Without working from the same set of inferential rules (or at the very least, obvious, but different sets of inferential rules), archaeologists working in these two areas have difficulty communicating the results of their analyses. As a result of historical precedent, definitions of the "Archaic" are often "adaptive, evolutionary, [unidirectional], and temporal" (Huckell 1993). The single criterion which most readily segregates the Archaic from the Paleoindian and Formative (or Early Agricultural) periods is the integration of tools specifically manufactured for more intensive preparation of wild plant foods. This does not mean, as some have implied, that the preceding groups "specialized" in big-game hunting or that wild food processing was abandoned when plants were cultivated. It is important to understand that the Archaic encompasses a range of human behavior in which some degree of hunting and gathering was practiced. THE SOUTHWESTERN "ARCHAIC" Although the concept that preceramic cultures existed in the Southwest was established early in this century (Kidder 1927), the term "Archaic" did not appear in Southwestern literature until the late 1960s (Huckell 1996b; Irwin-Williams 1968a, b). Prior to that time, preceramic cultures in the Southwest were defined by various local artifact assemblages, grouped into small "complexes" or broader "cultures" and "traditions." A more complete outline of the development of these concepts is presented in Huckell (1996b). Those cultural-historical groupings which have had the most lasting impact on Archaic archaeology in southern Arizona include the Desert 44 culture (Jennings 1956, 1957, 1973), the Cochise culture (Sayles 1983; Sayles and Antevs 1941), the PICOSA and Oshara traditions (Irwin-Williams 1967, 1973, 1979), the Southwestern Archaic (Irwin-Williams 1968a), the Pinto-Gypsum complexes or cultures (Campbell and Campbell 1935; Harrington 1933; Rogers 1939) and, to a somewhat lesser extent, the San Dieguito and Amargosa industries (Rogers 1939). Each of these groupings is described briefly below, with particular attention paid to the relationships with the archaeology of southern Arizona. San Dieguito and Pinto-Gypsum Over the course of several decades, Rogers (1929, 1939, 1958) revised what he called the San Dieguito complex based on research he conducted in the Mohave Desert. His studies focused primarily on surface sites and he based the relative ages of three phases, San Dieguito I, II, and III, on presumed changes in artifact attributes and the development of patina on artifacts. A rather thorough review of the changes in Rogers' typological schemes is provided by Warren (1967). In addition to the problems of interpreting assemblages based on surface distributions of artifacts, Warren (1967) cautions against the use of chemical changes on artifacts as temporal markers. To date, the only secure ages that can be assigned to the San Dieguito complex come from Warren's excavation at the Harris site (Warren and True 1961), originally excavated by Rogers in 1938. Three radiocarbon dates place San Dieguito materials at the Harris site between 9000 and 8400 B.P. (Warren 1967). 45 The names originally derived for Rogers' San Dieguito phases II and III were Playa I and Playa II, based on the surface distribution of flaked stone artifacts around playas in the Mohave Desert. At around the same time as Rogers' studies, the Campbells were excavating at sites in the Pinto Basin (Campbell and Campbell 1935) and Lake Mohave (Campbell et al. 1937), and Harrington excavated at Gypsum Cave (Harrington 1933). Each assigned names to projectile points found in these areas: Pinto, Gypsum, and Lake Mohave. The presence of projectile point types at sites in the Mohave Desert, in datable contexts, has given archaeologists in southern Arizona a basis for determining the age of sites with Pinto and Gypsum or Gypsum-like projectile points in Arizona. Amargosa Rogers (1939) defined the Amargosa tradition as the successor to and concurrent with (for the early Amargosa I phase) the Pinto-Gypsum complex. The Amargosa tradition was further refined by Haury (1950) based on excavations at Ventana Cave and has been linked with Middle Archaic occupations there. Difficulty by Haury (1950) in applying Rogers' San Dieguito terminology (review by Huckell and Haynes, 1995) led to subsequent collaboration between Rogers and Haury resulting in a three-phase sequence for the Amargosa tradition. Ventana Cave and the Stahl site yielded stratified Amargosa materials, but no secure radiocarbon dates. In addition to Pinto and Gypsum points, several kinds of corner- and side-notched projectile points are associated with Amargosa II contexts. Ground stone implements, 46 scrapers, "sleeping circles," linear rock alignments, trails, and intaglios' are also associated with the Amargosa tradition. Amargosa terminology is used, predominantly, at the Ventana Cave site, which is discussed in greater detail below. The Cochise Culture A post-Pleistocene human occupation in southern Arizona was first recognized in southeastern Arizona, where Sayles and Antevs in the 1940s utilized geologic and archaeological investigations of the Sulphur Springs and San Pedro valleys and along tributary streams to create a three-part sequence called the Cochise culture. The original report (Sayles and Antevs 1941) suffered from poor distribution and consequently never received the recognition it deserved (Dean 1987). As a consequence, assemblages which may have been recognized as part of the Cochise complex were often subsumed by terms given to other complexes, traditions, or cultures. By the time the updated monograph was published (Sayles 1983), it was already outdated. Nevertheless, its influence on our understanding of preceramic cultural sequences in southern Arizona cannot be ignored. Part one of Sayles' original (Sayles and Antevs 1941) three-part sequence consisted of the Sulphur Spring stage (pre-8000 B.C.), characterized by simple slab grinding stones and a limited chipped stone assemblage; these artifacts were found associated with late Pleistocene faunal remains, interpreted as indicating a mixed economy based on collection and processing of wild plants and on procurement of large game. No projectile points were found with Sulphur Spring stage assemblages, 47 considerably limiting the recognition of this phase in other areas. The subsequent Chiricahua stage (8000 to 3000 B.C.) was characterized somewhat more formally, consisting of basin-shaped metates, hand stones and a more elaborate chipped stone assemblage, including Chiricahua projectile points. Finally, the San Pedro stage (3000 to 500 B.C.) included implements from earlier stages as well as distinct side-notched San Pedro projectile points and features such as hearths and cooking or storage pits. Age estimates for these stages were provided by geologist Ernst Antevs, who correlated his interpreted paleoclimatic record with North American and better age-controlled European glacial and postglacial sequences. Though his ages were remarkably close to subsequently radiocarbon dated sections (Waters 1986a), his paleoclimatic intervals were based on "grossly oversimplified" reconstructions of the relationship between alluviation and precipitation (Dean 1987). Waters' (1986b) reinterpretation of Sayles' three-part chronology placed the Sulphur Spring stage between 8000 and 6000 B.C., the Chiricahua stage between 1500 and 500 B.C., and the San Pedro stage between 500 B.C. and A.D. 1. He also rejected Sayles' subsequently defined (1983) Cazador stage and, instead defined a gap between 6000 and 1500 B.C. for which he argues little evidence for human occupation exists (Waters 1986b). Whalen (1971) also rejected the applicability of the Cazador stage. One of the main themes of Sayles' three-stage culture-history was the idea that Archaic hunter-gatherers became increasingly more reliant on plant foods and that they 48 invested more time in plant processing and associated technology. As they did, they naturally became more sedentary, and eventually began cultivating plants. The Desert Culture Perhaps the most important contribution to the recognition of an Archaic lifeway in western North America was Jennings' (1953, 1957; Jennings and Norbeck 1955) definition of the Desert culture. Jennings recognized that many of the Western Archaic complexes were similar and attempted to equate these with the aridity of the environment. Based on his excavations in the Great Basin, he described a 10,000-year continuum of gathering-hunting-gathering from the end of the Pleistocene to the historic period. This continuum was marked by sparse, small non-sedentary populations living in caves or rockshelters, reliant on hunting, as well as small seed harvesting and grinding. Economically, these groups exploited the environment intensively, but were non-specialized. Although Jennings' interpretation was proposed principally for the Great Basin, its emphasis on wild plant foods caused archaeologists to relate it to regional variants such as the Cochise culture and the San Dieguito and Amargosa industries. PICOSA, Oshara, and the Southwestern Archaic Several bold attempts to synthesize the diverse local assemblages in the Southwest came from Cynthia Irwin-Williams. Her first attempt at synthesizing the pre-ceramic cultural-histories of the Southwest was a sequence she named PICOSA 49 (Irwin-Williams1967). The acronym represented three geographically separate traditions—Pinto, Cochise, and San Jose—which she grouped into "a continuum of similar closely related preceramic cultures" (Irwin-Williams 1967:441). She noted similarities between archaeological traditions from different parts of the Southwest, all of which were being split into individual culture sequences, and which she linked under a single acronym. She also argued that the economy and organization of these groups was similar to the Desert culture of the Great Basin, but preferred to recognize PICOSA and the Desert culture as distinct traditions. Based predominantly on the recognition of projectile point types, her three divisions conform geographically with the spatial distributions of later ceramic cultures (the Mogollon and Hohokam, Anasazi, and Hakataya/Patayan). Irwin-Williams' PICOSA and Jennings' Desert culture, were part of a continental effort to make regional cultural-historical sequences more areally inclusive (Huckell 1996b) and eventually led to the recognition and acceptance of the term "Archaic" (Irwin-Williams 1968a,b). Like many archaeologists today, Irwin-Williams' recognition of a regionally inclusive "Archaic" did not prevent her from continuing efforts to refine regional sequences, one product of which is her Oshara tradition in the northern part of the American Southwest (Irwin-Williams 1973, 1979). THE MIDDLE ARCHAIC ...one can easily count on the fingers of one hand the archaeologists who have devoted more than passing attention to preceramic archaeology in the Southwest,.. .While this situation has been changing 50 over the last 15 years, it is interesting to note that general texts on Southwestern archaeology have had little to say concerning the preceramic period; ... Huckell (1984a:2-3) In the 13 years since Bruce Huckell made this statement, the process of discovery has yielded abundant additional information about the "preceramic" or Archaic Southwest. However, our knowledge of the middle portion of the Archaic period is still sorely lacking. Although comparatively few sites have been found, the diversity of sites found dating to the Middle Archaic hints that we have only scratched the surface of an abundant source of data, particularly in the southern Basin and Range province. Why So Few Middle Archaic Sites? The paucity of archaeological sites dating to the Middle Archaic is owed, in part, to several factors most easily summed up by three words: preservation, identification, and discovery. As in the preceding Paleoindian and Early Archaic periods, the passage of thousands of years can have profound implications for the preservation of Middle Archaic resources. The passage of time affects not only the preservation of the resources themselves, but also those things which act as context for Middle Archaic archaeological remains (including geologic materials, contemporaneous plant and animal assemblages, geographic features, etc.). With the passage of time, it 51 becomes increasingly difficult to reconstruct the environment from proxy data or to reconstruct human behavior from the archaeological record. The temporal dimension also influences our ability to distinguish between Middle Archaic resources and those of other time periods. Certain artifact types, or sometimes sets of archaeological resources such as lithic assemblages or clusters of features, serve as "type fossils" for particular time periods. As time passes, parts of those assemblages or artifacts are lost, destroyed, moved, or transformed. As detail is lost from each site or activity area, we lose the ability to reconstruct the activities that occurred there and to correlate those activities with other sites. For nearly all of preceramic prehistory in North America, the most diagnostic artifacts are projectile points; however, with further examination, features and assemblages can become greater diagnostic tools, simply because they have more "staying power 2 ." In southern Arizona, the ability to distinguish between Middle Archaic and Late Archaic/Early Agricultural temporary campsites is difficult. As will be demonstrated, so-called Middle and Late Archaic projectile points have been recovered together at several sites. Whether these sites represent the accumulation of Middle Archaic projectile points by Late Archaic groups or multi-component occupations at a single locale is relatively immaterial. Either perspective suggests that Middle and Late Archaic groups were utilizing the same environments and locations. One final note: although the Late Archaic in the Tucson Basin has been recently renamed the Early Agricultural period (Huckell 1995), based on growing evidence that the cultivation of domesticated plants was a part of the economy, it is not clear yet whether Late 52 Archaic sites lacking evidence for cultigens represent another behavioral or economic aspect of the same group(s) or whether they represent differing, but contemporaneous groups utilizing the same general area. Finally, with greater time depth, there is greater chance for deep burial of archaeological resources. This can be a blessing, because it preserves not only the archaeological resources but the contextual data used to interpret environmental conditions and human behavior. However, greater time depth and deep burial can also be a curse; older sites, by definition, are susceptible to more periods of erosion than are younger sites and typical archaeological testing rarely exceeds a depth of 2 m (the depth allowed by OSHA for a small trench). Fortunately, recent advances in discovery techniques, including greater use of geoarchaeological prediction of archaeological resource location, have added data to the sparse quantity of Middle Archaic research. Characteristics of the Southwestern Middle Archaic There remain today few ways to segregate Middle Archaic from Early or Late Archaic sites. The primary and most reliable means of separation is through an absolute age estimate. Although estimates can sometimes be obtained by other means (e.g., cation-ratio dating, obsidian hydration, dendrochronology), the technique that is used most frequently is radiocarbon dating. Dates for the Middle Archaic period are not entirely agreed upon, but there appears to be significant evidence to suggest that this period dates to around 5500 to 3500 B.P. (Huckell 1996b). Further examination 53 of the geochronological implications of dating the Middle Archaic period are explored below. These dates are often based on association with the next most frequently used diagnostic tool for assigning age to a site: projectile point typology. Often, there exist certain items that have not independently been established as diagnostic of the time period under study, but that have typically been found with the diagnostic items of that time period. These items should not be used as "diagnostic," but represent the characteristics of the time period. In the following sections, I refer to these items as "hallmarks" of Middle and Late Archaic assemblages. Projectile Points Middle Archaic projectile point types are strongly tied to regional chronologies outlined in the previous section. The recognition and dating of projectile point types at sites across the West has been used as a way of relating sites which may be located in geographically different regions, topographically different contexts, and containing different assemblages of other artifacts and features. Several projectile point types may serve as hallmarks of the Middle Archaic; these include: Chiricahua and other side-notched types, San Jose/Pinto, Gypsum, and Elko. Each of these types may have regional varieties. In addition, certain other projectile point types appear to exist only within restricted regions. In southern Arizona, the Cortaro point may be indicative of a Middle Archaic occupation (Roth and Huckell 1992; Sliva 1997). A long tapering-stemmed variety, which appears to be different from Gypsum, has often been classified as Early Archaic on the basis of radiocarbon dates from other regions 54 (Huckell 1984a); however, their recovery from numerous Middle Archaic contexts in southern Arizona suggests that this point style may be affiliated with the Middle Archaic. Middle Archaic Features Until recently, Middle Archaic features have been thought to be poorly constructed or nonexistent. In the past few years, however, mounting evidence for more formalized features has come from excavation of Middle Archaic sites. Extramural features typically found at Archaic sites include basin-shaped hearths, rock-filled pits, fire-cracked rock clusters, and ground stone caches (Agenbroad 1970; Irwin-Williams 1973; O'Laughlin 1980; Windmiller 1973). A single possible storage feature has been interpreted as part of a Middle to Late Archaic aged occupation near El Paso, Texas (O'Laughlin 1980). The absence of storage features has been used as evidence against an early adoption of agriculture; however, negative evidence cannot be construed as evidence of absence. It is also possible that evidence for storage may be better preserved in floodplain sites; future excavation of such sites may produce storage features. Few inhumations have been found and the few known appear to follow a pattern of burial practices. Two inhumations from the Tucson area were recovered from beneath rock cairns (Dart 1986; Huckell 1996b). Although not all Middle Archaic burials are found beneath these cairns, this trait also appears in Middle Archaic burials in other parts of the Southwest (Mabry 1996b). 55 In addition to these small features, evidence appears to support the presence of domestic structures in Middle Archaic sites. Although not an astounding observation, the recognition of preserved Middle Archaic domestic structures confirms that shelters are not solely a trait of later periods. Simple structures consisting of irregular outlines and posthole patterns were first detected by Irwin-Williams (1973) in west-central New Mexico. Huckell (1984a) has suggested that a rock alignment discovered on the east side of the Santa Rita Mountains near Tucson, which he interpreted as a windbreak, may be Middle Archaic in age. At the Tator Hills site in the Santa Cruz Flats, the remains of a single "brush structure," consisting of a circular depression filled with a dark homogenous, trash-filled sediment and surrounded by several postholes located about the perimeter of the structure, appears to be associated with an Archaic occupation surface on which Chiricahua and tapering-stemmed projectile points have been recovered (Halbirt and Henderson 1993). Radiocarbon dates from other features on this surface have given anomalous ages. Additional evidence for shallow pit structures has been found in the central Rio Grande Valley (Schmader 1996), near El Paso, Texas (O'Laughlin 1980; Whalen 1994), and possibly near Moquino, New Mexico (Beckett 1973). Other Characteristics of the Middle Archaic In addition to these hallmarks, Huckell (1996b) argues that rock art was probably being created during the Middle Archaic. He suggests that Middle Archaic rock art may include "linear geometric and representational elements," as well as 56 anthropomorphic figures. In addition to these forms of artistic or representational expression, he notes that split twig figurines have been found in rockshelters in southeastern California, northwestern Arizona, southern Nevada, and southeastern Utah (Davis and Smith 1981; Emslie et al. 1987; Euler 1984; Janetski 1980; Schroedl 1977). Middle Archaic Sites in Southern Arizona The evidence for a Middle Archaic occupation in southern Arizona is quite spotty and appears to be geographically biased toward bajada sites. Large surveys in this area have identified several sites dating to the Middle Archaic. Areas that have received little coverage include the floodplain and montane regions. By convention, 3 radiocarbon dates cited are expressed as uncalibrated radiocarbon years before 1950 (B.P.) 4 . Calibrated (calendar) ages are cited using the following abbreviations B.C. or A.D. Ventana Cave Ventana Cave is located in the Castle Mountains north of Sells, Arizona. Here, Haury (1950) found a stratified sequence of cultural deposits dating from the early Archaic (Huckell and Haynes 1995) through the present day. At the time of excavation, the age of the deposits was based on geologic assessment of the site deposits by Antevs, Hack, and Bryan (1951; cf. Haury 1950), all of whom had different interpretations, and on correlation between artifacts found in the deposits and 57 their known associative ages at other sites. The "volcanic debris" layer contained the oldest artifacts at the site, the Ventana Complex, and was considered related to Folsom (Haury 1950) or Clovis (Haury and Hayden 1975). Geologic assessment of the ages of subsequent materials in the "red sand" was based on this original correlation and the depth of time represented by an erosional unconformity separating the Ventana complex from the younger Ventana-Amargosa I deposits (containing stemmed projectile points similar to those classified elsewhere as Jay points). Recent radiocarbon dating and correlation of the artifacts with similar assemblages from the Great Basin place this earliest deposit between 10,700 to 8700 B.P. or, narrowing the margin based on evaluation and selection of the most reliable dates, between 9500 and 8700 B.P. (Huckell and Haynes 1995). Overlying the Ventana-Amargosa I deposit was a midden (the "moist midden") containing Middle Archaic (Haury's Chiricahua-Amargosa II) and Late Archaic (Haury's San Pedro) artifacts. The midden was subsequently overlain by a second midden (the "dry midden") containing Hohokam artifacts. Projectile points from the Chiricahua-Amargosa II levels included San Jose/Pinto, and tapered-stemmed (Gypsum) projectile points. The San Pedro levels contained San Pedro and Cienega projectile points. Also found through these deposits were projectile points recently classified as Cortaro (Roth and Huckell 1992). Roth and Huckell (1992) more fully discuss the problems involved in using the Ventana Cave data. Unfortunately, no radiocarbon dates have been derived from the midden deposits, presenting further 58 difficulty in correlating them with assemblages found at other sites of presumably similar age. Ventana Cave was likely used more-or-less continuously as a hunting camp andJor possibly a base camp during the Middle and Late Archaic periods. Bayham (1982) focused on the changing use of the cave from the Archaic to Hohokam periods and concluded that the abundance of artiodactyl remains in more recent levels suggests a change in the transport costs associated with sedentism. In other words, during the Hohokam period, the cave was used as a hunting camp, because the relative cost of transporting an entire deer carcass back to the base settlement would be too high. Picaclzo Reservoir Several Archaic sites are known from work near the Picacho Reservoir in 1983-1986 for the Tucson Aqueduct (Bayham et al. 1986). The sites are located at the northeast edge of a broad flat plain called the Santa Cruz Flats, on the bajada of the Picacho Mountains. A dune field covers a portion of the bajada and a small eroded drainage (the "arroyo") crosses the dunes. At least three of the sites had substantial Middle Archaic components; these include the Buried Dune site [AA:3:16 (ASU)], the Arroyo site [AA:3:28 (ASU)], and the Gate site [AA:3:8 (ASU)]. Each appears to represent a different type of activity, at least partially influenced by prehistoric setting and by the potential for preservation in that setting. The Buried Dune site consisted of several clusters of fire-cracked rock, which have been interpreted as cooking features, and a low density scatter of flaked stone 59 and burned bone. At least some of the flaked stone raw material was acquired in southwestern New Mexico. Projectile points are San Jose/Pinto varieties. The site has been interpreted as a short-term field camp occupied during the winter or early spring. Two radiocarbon dates on features from the Buried Dune site are 4300 ± 310 B.P. (AA-673A) and 4290 ± 330 B.P. (AA-673B). A third radiocarbon date at the base of the younger dune overlying this feature is 4000 ± 220 B.P. (AA-675). A single locus at the Arroyo site consists of a dense midden, near what archaeologists interpreted, based on the presence of cattail (Typha) pollen in the midden (Gish 1986), as a Middle Holocene cienega (Bayham et al. 1986:365). Geomorphologist Mike Waters (Bayham et al. 1986:365; Waters, pers. comm. 1996) has argued that sedimentary and soil characteristics of this deposit do not indicate cienega deposition. Middle Archaic projectile points consisted of Chiricahua and tapering-stemmed points. Radiocarbon dates on features in the Middle Archaic locus yielded dates of 4500 ± 100 B.P. (Beta-12267) and 3910 ± 290 B.P. (Beta-12270). A third date at the base of the Middle Archaic midden yielded an age estimate of 4840 ± 100 B.P. (Beta-12268). Lithic resources and subsistence data point to use of local resources over a relatively lengthy duration. The site has been interpreted as a long-term base camp occupied principally during the summer and fall. The Gate site has been interpreted as a base camp reoccupied repeatedly for short periods of time. The site consists of a small Middle Archaic midden, a large scatter of features and artifacts dating to both Middle and Late Archaic periods. Both Chiricahua and Gypsum projectile points were recovered from the Gate site. 60 Radiocarbon dates are 4910 ± 95 B.P. (Beta-12271) and 4350 ± 90 B.P. (A-2964; Czaplicki et al. 1984:42). A fourth site, AZ AA:3:9 (ASU) yielded a radiocarbon age on a feature of 4140 ± 90 B.P. (Bayham et al. 1986). Based on the presence of non-local raw materials and San Jose/Pinto varieties at the Buried Dune site, Bayham and others (1986) make the argument that two different groups, one regionally mobile and one locally mobile, were utilizing the Picacho Reservoir area during the Middle Archaic period, and that locally mobile groups reused the same sites into the Late Archaic period. Santa Cruz Flats Several Archaic sites are known from work at the southern margin of the Santa Cruz Flats, near where the Aguirre Valley joins the Santa Cruz River (Halbirt and Henderson 1993). The flats are named for the distal fan portion of the Santa Cruz Valley, where outflow from the Tucson reach of the Santa Cruz spreads out over a vast, flat plain. Middle Archaic Pinto and Chiricahua style projectile points were recovered from the Tator Hills site [AZ AA:6:18 (ASM)]. A Chiricahua and a stemmed projectile point were associated with the oldest radiocarbon age estimate from the site of 1890 ± 130 B.P. (Beta-25445)*. The site consisted of a concentration of artifacts in association with the remains of a single "brush" structure, several thermal pits and fire-cracked rock clusters, and likely represents a temporary, seasonal occupation. 61 Several Late Archaic San Pedro and Cienega style projectile points were recovered from the Coffee Camp site [AZ AA:6:19 (ASM)] along with five Middle Archaic Pinto and Chiricahua style projectile points. The majority of radiocarbon dates from this site cluster around 2100 B.P. Three older dates [3120 ± 170 B.P. (Beta-26352)*, 2870 ± 100 B.P. (Beta-27697)*, and 4040 ± 190 B.P. (Beta-27696)1 may represent an earlier occupation of the site; however, many of the dates are run on mesquite charcoal' and dates of different ages are scattered throughout the stratigraphy. A total of 354 features including pithouses, burials and pits were found in three distinct geologic strata 6 . The site appears to represent a "semipermanent camp, whose occupants exploited the resources of the upper and lower bajada, as well as riverine and floodplain areas (Halbirt and Henderson 1993:375)." Middle Archaic projectile points from both Tator Hills and Coffee Camp may represent scavenged artifacts from earlier occupations. Their presence suggests, nevertheless, that a Middle Archaic occupation was present somewhere in the vicinity of the Santa Cruz Flats. The presence of Middle Archaic projectile points at other sites within the study area and a radiocarbon age estimate of 4580 ± 80 B.P.' on a roasting pit at site AZ AA:6:27 (ASM) further supports this possibility. Harquahala Valley Several Archaic sites are known from work in the Harquahala Plain, south of Interstate 10 between Salome and Tonopah, Arizona (Bostwick 1988). The principal drainage flowing through this valley is Centennial Wash, a tributary of the Gila River. 62 Middle Archaic projectile points were recovered from three of the ten sites investigated. The majority of projectile points were Late Archaic in age. Apart from a single corrected date of 7850 ± 400 B.P. (Beta-10959) from the Tortoise Sink site, the only radiocarbon age estimates were Late Archaic in age. These other dates, 3000 ± 120 B.P. (Beta-10957, The Lookout site) and 3290 ± 70 B.P. (Beta-10958, the Apothecary site) date to the San Pedro period of the Late Archaic. Like in the Santa Cruz flats area, the presence of Middle Archaic projectile points does not necessarily indicate site age or duration of site occupation. Their presence does suggest that the general area was utilized sometime during the Middle Archaic period. San Pedro, San Simon, and Sulphur Spring Valleys In a study of the Middle and Late Archaic of the San Pedro Valley, Whalen (1971) concluded that Archaic sites were located along river terraces and on the piedmont. Most of these sites overlooked major stream channels or were near springs or washes. Of his sample of 82 Archaic period sites, only six were Chiricahua (Middle Archaic) phase sites. Recent investigations near Kartchner Cavern reiterate the evidence for repeated seasonal occupation of these sites (Phillips et al. 1993). While historic (Cummings 1927, 1928; Sayles and Antevs 1941) and more recent surveys (Altschul and Jones 1990; Phillips et al. 1993; Whalen 1971, 1975) of river valleys in southeastern Arizona have turned up numerous Late Archaic aged sites, few Middle Archaic sites have been found and fewer yet intensively investigated. 63 The Lone Hill site (Agenbroad 1970, 1978), located in the piedmont west of the San Pedro River, is comprised of multiple hearths, ground stone, and numerous projectile points. Based on the presence of inverted basin metates, suggesting the intent to return, Agenbroad concludes that the site was seasonally reused. One of the metates was completely worn through, suggesting either single, long-term occupation or intensively repeated seasonal occupation of the site. Recent studies (Phillips et al. 1993) of a Middle Archaic occupation near Kartchner Cavern at AZ EE:3:28 (the Kartchner Cavern site) have recovered a single Clovis Paleoindian projectile point, as well as seven tapering-stemmed projectile points, 12 Pinto points, a single Chiricahua point, a single Cortaro point, a handful of later projectile point types, and numerous fragments or unidentifiable point types. The site is interpreted as a lithic workshop that was reused throughout prehistory. Sycamore Canyon and Rosemont Sites Several Archaic sites have been discovered in the foothills and montane areas of the Santa Rita Mountains, southeast of Tucson. Tagg and Huckell (1984) argued for an Archaic component in Sycamore Canyon based on the presence of tapering-stemmed, Pinto, and Cortaro projectile points and on similarities between flaked stone discovered on those sites and at other Archaic sites. The features in Sycamore Canyon all appeared to be much younger. The Sycamore Canyon sites are located on the northwestern edge of the Santa Rita Mountains. 64 The Rosemont sites are located on the eastern side of the Santa Ritas, in a drainage area which supplies Cienega Creek. The McCleary Canyon site (AZ EE:2:102) consists of a small scatter of flaked stone tools and debitage, including San Jose/Pinto style projectile points, a Cortaro point, and other later styles, and five handstones. A shallow firepit was also found at this site. The Wasp Canyon site (AZ EE:2:62) consists of five rock clusters and two additional rock "structures." A radiocarbon date from the floor of a shallow pithouse at the site yielded a date of 1990 I..- 370 B.P. (A-3103). Two Late Archaic corner notched projectile points and fragments of two basin metates were recovered from the immediate vicinity of this structure, further supporting a Late Archaic age for the feature. A second feature consisted of an elliptical alignment of rocks which Huckell interpreted as a simple brush structure or windbreak. A Pinto style projectile point was recovered from the floor of this structure. Tapering-stemmed and Pinto projectile points from the Wasp Canyon site suggest an Early or Middle Archaic age (Huckell 1984a). Tapering-stemmed, Pinto, and Gypsum points were also recovered from the South Canyon site (AZ EE:2:82). Five additional features of clustered rock were found at this site. AZ EE:2:87 yielded a mixed assemblage, including flaked stone and four pieces of ground stone. Projectile points from AZ EE:2:87 include several Pinto points and a "lozenge-shaped" projectile point, similar to the Gypsum form. A single rock cluster was found at this site. 65 Among the most intriguing results of the Rosemont study is the fact that Archaic period sites have been located in montane areas. These areas have seldom been systematically surveyed. Huckell (1984a) also concluded that Early/Middle and Late Archaic sites were differentially distributed. He found that Early and Middle Archaic sites tended to be farther up the small montane canyons, while Late Archaic sites were distributed down valley, closer to the piedmont. He suggested that differences in site location may be due to changes in environmental conditions or in subsistence practices. La Paloma Several Middle Archaic projectile point types (including the Cortaro, Gypsum, and Pinto point styles) were recovered from excavated and surface contexts at the La Paloma site and the nearby Pontatoc site. At La Paloma, Dart (1986) believed he could stratigraphically separate this Middle Archaic component from a subsequent Late Archaic component. Radiocarbon dates were not available for the Middle Archaic component; however, carbonate development in the soils appeared strong enough to support a Middle Archaic age estimate. A stone alignment that may have served as a temporary structure and a circular arrangement of fire-cracked rocks were the only features found associated with the Middle Archaic aged artifacts. Pollen samples from the site suggested that the Middle Archaic environment was roughly equivalent to that of the Late Archaic. Chenopods, amaranths, spiderling, and wild buckwheat were present in these samples 66 and may have served as plant foods for this Middle Archaic activity area. Evidence for seasonality was inconclusive, but Dart proposed that the site may have been occupied during the early spring through autumn. Some lithic materials recovered were obtained at considerable distance (60-100 km) from the site, suggesting either mobility outside the Tucson Basin or interaction with groups outside the basin. Dart concluded that the occupation at La Paloma represented "impermanent annual settlement." Although he suggests that repeated use of the site during the Middle Archaic may have been possible, the relative dearth of evidence for features and extensive amounts of artifact debris directly related to the Middle Archaic suggest that seasonal reoccupation may have not been so intensive. Flying V Ranch During the mid-1980s, three Archaic sites were found in the Ventana Canyon area (Douglas and Craig 1986) during a survey of the Flying V Ranch. Middle and Late Archaic projectile points were discovered at each of the three sites. The diagnostic artifacts from all three sites were collected; however, only one of the two larger sites was excavated and extensively surface collected. At least 19 projectile points or fragments were recovered from this site (AZ BB:9:139, ASM). Based on analysis of this single site, the area appeared to have been seasonally reoccupied by Archaic groups since the Middle Archaic period. The Middle Archaic projectile points recovered from the three sites include both Gypsum and Cortaro styles. 67 Tortolita Mountains Sites Four Middle Archaic sites were found in the upper bajada of the Tortolita Mountains during the Tucson Basin survey (Roth 1988, 1989). A fifth, previously discovered by Hewitt and Stephen (1981), had been recorded prior to the survey. Projectile points on these sites included Elko, stemmed, Pinto, Pelona, Chiricahua, and Cortaro point types. Two Pinto points were recovered from the surface of another Late Archaic site (AZ AA: 12:84; Roth 1995). These sites have been classified as limited activity and small multiple activity sites (Roth 1989) and appear to have been used by later occupations, as well. RiliC012 Mountains Sites Cortaro projectile points have been recovered from the upper bajada of the Rincon Mountains (Simpson and Wells 1983). The sites tend to be located near potential water sources. No features have been associated with these lithic scatters. A vra Valley Hundreds of Middle Archaic projectile points have been recovered from the Avra Valley area (Czaplicki 1983; Dart 1987; Downum et al. 1986; Swartz 1987). During a survey for the Central Arizona Project (CAP) Aqueduct, Downum and others (1986) recorded a number of sites with Middle Archaic projectile points in dunes located at the base of the Tucson Mountains. These areas also appear to have been utilized during the Late Archaic. Middle Archaic projectile points have occasionally 68 been discovered eroding from alluvium exposed in entrenched portions of Brawley Wash. Huckell (ASM site files) suggested that Middle Archaic artifacts eroding from alluvium in Brawley Wash at the LaBoissiere site were reworked from the floodplain surface. These conclusions have been supported by recent investigations (Lindeman and Freeman 1996). Santa Cruz River Floodplain Sites Prior to excavation of the Los Pozos site (discussed in detail in Chapter 6), firm evidence for a Middle Archaic occupation of the Santa Cruz River was extremely scarce. Chiricahua style projectile points were recovered from AZ AA:12:86, a predominantly Late Archaic site on the Santa Cruz floodplain. Haynes and Huckell (1986, Huckell 1996b:338, pers. comm. 1997) have recorded a Middle Archaic site near the Santa Cruz River south of Martinez Hill in Brickyard Arroyo (AZ BB: 13:70). This site consists of several rock clusters, a small lithic scatter, an immature bison skull, and a fragmentary Chiricahua projectile point. A single radiocarbon date from the site has yielded a date of 4320 ± 120 B.P. (AA-2139). They have also reported a possible Middle Archaic occupation along Ina Road, near the interstate, at AZ AA:12:111, and a Middle Archaic component at the Joe Ben site (AZ BB:13:11), just north of AZ BB: 13:70 in Brickyard Arroyo. A radiocarbon date on the lowest cultural stratum at AZ AA: 12:111 yielded a date of 4260 ± 140 (A-2234); the stratum also contained animal bone and fire-cracked rocks. The lowest cultural bearing stratum at the Joe Ben site yielded three dates: 4850 ± 90 (A-1854), 3980 ± 100 (A-1783), and 69 4400 ± 220 (A-1853); however, the only diagnostic artifact found at the site was a San Pedro (Late Archaic) projectile point (Haynes and Huckell 1986). At each of these sites, strata containing flaked stone, fire-cracked rock, and animal bone have been found. Hearths or other types of pit features are occasionally exposed. The Cortaro Fan site (AZ AA: 12:486) is located at the distal portion of an alluvial fan protruding from the Tortolita Mountains into the Santa Cruz floodplain. During excavation of the site, Roth (1992) found numerous small surface or subsurface features, including roasting pits, hearths, middens and clusters of fire-cracked rock. Maize was recovered from some of the features and yielded radiocarbon dates of 2790 ± 60 B.P. (AA-2782) and 2594 ± 90 B.P. (AA-2783). Three additional dates on mesquite charcoal in roasting pits associated with maize are 2290 ± 240 B.P. (A-4727), 2270 ± 50 B.P. (A-4728), and 2300 ± 100 (Beta-29803). In addition to numerous other Middle Archaic projectile points recovered from this site are several examples of the Cortaro projectile point type, which appears to date to the Middle and Late Archaic periods (Roth and Huckell 1992; Sliva, 1997). Status of the Middle Archaic in Southern Arizona Our knowledge of the Middle Archaic period in southern Arizona has changed a great deal during the past 10 years. Although the number of new discoveries of Middle Archaic sites has not been large, the significance of those discoveries has revolutionized the understanding of the Middle Archaic period. Radiometric dating of previously discovered sites has confirmed the presence of Middle Archaic features and 70 artifacts in Santa Cruz floodplain sediments. These discoveries have conveyed a more dynamic record characterized by the recognition of additional hallmarks, site types, and site distributions. New Hallmarks of the Southern Arizona Middle Archaic As more sites are excavated new hallmark traits of the Middle Archaic are discovered. With additional research, split twig figurines found in rockshelters and caves on the Colorado Plateau may be found in other well-preserved contexts and refinement of dating techniques may be used to determine the difference between rock art from Middle Archaic sites and that from later periods. Among the most intriguing discoveries, however, are the recognition of Middle Archaic features and additional projectile point types. The presence of paleobotanical remains in Middle Archaic features and the ability to recognize additional surface distributions of Middle Archaic sites may, in the future, enhance our knowledge of the Middle Archaic and floodplain resource areas. Middle Archaic Features Several criteria have been used to determine sedentism in the archaeological record. These criteria include: larger site size, thick occupation deposits, and high artifact densities; the presence of formal architecture, including ceremonial structures; and access to critical resources, particularly water (Rafferty 1985). Investment in large features, such as dwellings, is often used as evidence of permanence or 71 semi-permanence of residence (Eder 1984; Rafferty 1985; Stark 1981). The degree of investment in dwellings must be considered (Whitelaw 1991), however, as brush structures or circular rock "wind breaks" require far less investment in time and energy than do formal pit structures. Storage and caching are also important components when measuring the mobility of foraging groups (Binford 1980; Gilman 1987). These types of features typically indicate repeated use of sites. Although few dwellings have been recognized from Middle Archaic sites in southern Arizona, the presence of these dwellings in other parts of the southern Basin and Range may signify the potential for discovery of additional features of this type in the Tucson Basin. New discoveries in the floodplain are likely to yield more of these features due to the area's potential for preservation of the remains of these features. Future discoveries may also be able to glean "hallmark" data from other Middle Archaic features. Certain types of other features, such as pits, may become more recognizable and their functions more interpretable. In addition, these features yield important subsistence and temporal data that may be used to recognize the differences between Early, Middle, and Late Archaic economy and mobility. Middle Archaic Projectile Points It is clear that additional research has yielded new "hallmark" projectile points for the Middle Archaic period. The appearance of tapering-stemmed (thought to be Early Archaic) and Cortaro (thought to be Middle or Late Archaic; cf. Roth and Huckell 1992) projectile points repeatedly at sites dating between 5500 and 4000 B.P. 72 may signal the presence of two additional diagnostic types. The presence of the Cortaro point in undeniably Middle Archaic contexts at the Los Pozos site appears to confirm its legacy as a Middle Archaic hallmark. Projectile points may also indicate the presence of groups practicing two distinctly different patterns of mobility. The research conducted by Bayham and others (1986) at the Picacho Reservoir, by Roth (1995, 1996) in the Tortolitas, and by Huckell (1984a) in the Santa Rita Mountains seems to suggest that certain Middle Archaic groups were practicing a more limited pattern of mobility. These groups may have already been tethered to smaller regions, accessible to both floodplain and upper bajada environments, a pattern that continued into the Late Archaic/Early Agricultural period. Middle Archaic Site Types and Site Distributions The idea that some Middle Archaic sites in southern Arizona mark groups tethered to certain resources is further supported by studies of subsistence data from Middle Archaic sites (Bayham et al. 1986; Huckell 1984a; W6cherl 1997). Although numerous Middle Archaic limited activity sites have been discovered, additional sites supporting repeated use of sites and site areas have been discovered. The presence of sites in floodplain localities which would offer abundant wild plant foods signals an important change in the relationship between environment and human occupation. The discoveries at Los Pozos and Cortaro Fan may suggest that the fortuitous discoveries of past decades are but a minor manifestation of a potential 73 larger-scale occupation of the floodplain. Future discoveries along channel margins and in distal fan reaches where entrained sediment from incised reaches is allowed to accumulate are likely to reveal additional Middle Archaic floodplain occupations. Some Middle Archaic sites may also be preserved at deeper stratigraphic levels; however, new archaeological testing and data recovery techniques will have to be developed before these deeper deposits can be explored. Just five years ago the number of Late Archaic floodplain sites was very small, but today we recognize an almost continuous occupation of the floodplain by Early Agricultural populations during that period. The environment of the Santa Cruz River may have provided the initial conditions to support reduced mobility during the Middle Archaic. The presence of archaeological sites in the floodplain, a pattern of repeated site use, and tethering to certain resource areas create a pattern of site distribution and site use that is amplified during the Late Archaic/Early Agricultural period. This pattern has important implications for the transition to agriculture in the Tucson Basin. THE LATE ARCHAIC/EARLY AGRICULTURAL PERIOD Volumes have been written on the Late Archaic period in southern Arizona (e.g., Eddy and Cooley 1983; Gregory 1997b; Huckell 1995; Mabry 1996a; Sayles 1983). The following section highlights features that distinguish it from the preceding period. For the purposes of this discussion, I follow the chronological scheme presented by Huckell (1995) and utilized by Mabry (1995), which classifies sites with 74 the earliest evidence for agriculture as the "Early Agricultural Period." Investigations of Late Archaic/Early Agricultural Sites in Southern Arizona The Late Archaic, late preceramic, late preagricultural, and Early Agricultural periods, as they are variously defined, are encompassed under the San Pedro phase of the Cochise Culture. The same periods from nearby regions include the Fresnal and Hueco phases of the Chihuahua tradition in Mexico (MacNeish 1993) and the Armijo and En Medio phases of the Oshara tradition (Irwin-Williams 1973). Early Investigations The San Pedro phase was first defined on the basis of artifacts and features found at the Fairbank and Charleston sites in the San Pedro Valley (Sayles and Antevs 1941; Sayles 1983). The Fairbank site contained only small storage and cooking features and provided no evidence for agriculture. Initial studies of this phase suggested that Late Archaic groups were mobile hunters and gatherers following a seasonal pattern of repeated site occupation. Pit structures were found at the Charleston site after publication of the original report (Sayles 1983). Evidence of agriculture came later, with Eddy and Cooley's excavation of sites in the Cienega Valley (Eddy 1958; Eddy and Cooley 1983). In addition to maize, the features found at the San Pedro Cochise 8 phase AZ EE:2:35 (ASM), Donaldson and Los Ojitos sites included small round pit structures, hearths, cooking and storage pits, and middens. 75 Defining the Early Agricultural Period As more Late Archaic sites were discovered, increasing evidence existed that sites dating to this time period fell into two distinct groups: those having evidence for agriculture and those without evidence for agriculture. Although some non-ceramic sites are difficult to place in time, the securely dated Late Archaic sites in the Tucson Basin were remarkably consistent in the presence of maize (Fish et al. 1990; Roth 1989). Moreover, the ubiquity of maize at these sites often exceeded that of the later Hohokam period. Few of these sites in the upper portion of the bajada have been excavated and only one of them has exhibited evidence of agriculture. A pollen wash on a mano found at the La Paloma site revealed evidence of maize pollen. Roth (1995) has suggested that sites on the bajada surrounding the Tucson Basin may represent use of the area during logistical trips from central habitation sites in the floodplain. Of the many sites dating to this time period, perhaps one of the best examples of an apparently non-agricultural site is the Coffee Camp site (see above, Halbirt and Henderson 1993). Coffee Camp epitomizes the problem of distinguishing between a Late Archaic "base camp" and an Early Agricultural period "logistical camp." Huckell (1995) recognized the "ecological" problem inherent in classifying sites of clearly agricultural economy together with sites that lacked evidence of domesticates, and proposed the resurrection of a concept that was proposed more than 40 years prior (Martin and Rinaldo 1951; Woodbury 1993). The Early Agricultural period, as Huckell (1995) conceived it, was comprised of two phases: San Pedro 76 (1500-1200 B.C. to 500 B.C.) and Cienega (500 B.C. to A.D. 200). Huckell based his assignment of a new period on the mounting evidence for Early Agricultural populations in the Tucson Basin at the Milagro site (Huckell 1988, Huckell et al. 1994), the Cortaro Fan site (Roth 1992), the Clearwater site, 9 the sites along the Santa Cruz River, 10 and at other sites with evidence for early agriculture. He also based much of this new period on reinvestigation of the Fairbank site in the San Pedro Valley (Huckell 1990) and the Donaldson and Los Ojitos sites in Matty Canyon (Huckell 1995). New Discoveries along the Santa Cruz River In the past two years, additional Early Agricultural sites have been found along the Santa Cruz River, within the study area. Nearly all of these sites are located on the Holocene terrace (t2, Figure 2.1). Those that are directly related to geoarchaeological investigations within the study area are discussed in Chapter 4, and include the cluster of sites along Ina Road (AZ AA: 12:103, 111/688, and 503), and the Rillito Fan (AZ AA:12:788) and Clearwater (AZ BB:13:6) sites. Information on archaeological sites within the focus area is presented in Chapters 5 and 6. Because the analyses of Square Hearth (AZ AA:12:745), Stone Pipe (AZ BB:13:425), and Santa Cruz Bend (AZ AA:12:746) have been presented in a synthetic volume, the three sites are discussed together in Chapter 5, along with archaeological testing at the Juhan Park site (AZ AA: 12:44), located on the opposite bank of the river, in this area. 77 AA:12:91-12:90 t AA:12:85 AA:12:74f) .• 7 13B:13425 f Ireerna cer--arcelsoratil KEY I vs= ar. =1 WAIIII 1 acr- ch. ch. y t1 68:13:17i 142. O. or arbt difil Archaeological site 0 1 Kilometers II Desert Arch000logy. i;gi , Figure 2.L Distribution of archaeological sites in the study area and their location on the Holocene (Qt2) terrace. For descriptions of geologic units (after McKittrick 1988), see Appendix A. 78 A discussion of the Los Pozos (AZ AA:12:91) and Wetlands (AZ AA:12:90) sites is presented in Chapter 6. Florescence of the Early Agricultural Period Many of the excavated sites dating to the Early Agricultural period are large and contain numerous pit structures and extramural pit features. Few of the pit structures overlap, suggesting that the sites were repeatedly occupied over a short time-span, and many of the pit structures form what appear to be house groups (Mabry 1996a). Possible communal structures have been found at the Santa Cruz Bend site (Mabry 1996a) and the Wetlands site (Freeman 1997). Maize ubiquity values (frequency of contexts in which maize is present) at these sites are consistently very high (see Table 2.1), indicating that maize constituted a substantial portion of the prehistoric diet. Table 2.1 Presence of Carbonized Maize Remains at the Middle Santa Cruz River sites. Site No. of Samples Presence Value Santa Cruz Bend* 41 0.83 Early Agricultural Stone Pipe* 23 0.87 Early Agricultural 13 0.93 Early Ceramic 6 0.83 Early Ceramic Los Pozos** 140 0.88 Early Agricultural Wetlands** 46 0.87 Early Agricultural Square Hearth* * Analysis by L. Huckell (1996). ** Analysis by M. Diehl (1997a, 1997b). Time Period 79 The majority of sites dating to this period fall between 2500 and 2000 B.P., although a San Pedro-type pit structure was recently discovered at the Wetlands site (Freeman 1997)." There appear to be significant changes in site structure over the Cienega phase (800 B.C.-A.D. 150) that may reflect temporal changes in site use and that appear to correspond with seasonality of occupation (Freeman 1997); however, additional excavation will be needed determine whether these patterns hold. In addition to these Early Agricultural localities, a number of sites in the Santa Cruz floodplain display evidence for early ceramic technology. Sites dating to this later time period have been designated using the term "Early Ceramic period," and Mabry (1996a) has modified the cultural chronology to reflect those changes (see Table 1.1). Recent excavations have demonstrated, however, that the earliest evidence for ceramic production extends into the early Cienega phase (Diehl 1996b; Freeman 1997). Altogether, these new discoveries suggest that the Early Agricultural period floresced during the Cienega phase. During the early part of this phase it appears that use of sites (such as Clearwater and Wetlands) was more highly seasonal with occupations during the late summer or early fall (Freeman 1997), while occupations later at Santa Cruz Bend, Stone Pipe, and Los Pozos sites may reflect a year round occupation (Gregory 1997b; Mabry 1996a).I2 80 GEOCLIMATIC MODELS FOR ARCHAIC PERIODIZATION Long-term changes in climate are presumed to have had a great effect on geologic systems, particularly rivers. The changes in climate reflected by Holocene alluvium also mark changes in the environment to which human groups sometimes responded. Geoclimatic intervals, therefore, such as early, middle, and late Holocene, have often been used as criteria for defining boundaries between long-term changes in human activities. Unfortunately, these geoclimatic periods are also poorly-defined and there is a deplorable lack of consensus about how to interpret and designate these intervals. The period marked by Antevs' (1948, 1953) Altithermal appears to represent a significant set of changes in geologic and climatic parameters. In southern Arizona, the onset of this period seems to be marked by a complete lack of human occupation (Berry and Berry 1986; Irwin-Williams and Haynes 1970; Waters 1986b). A recent inventory of Paleoindian and Archaic sites has indicated, in fact, that a "middle Holocene" occupation of southern Arizona is virtually nonexistent (Mabry et al. 1997). As climate apparently ameliorated toward the end of this period, the first "Middle Archaic" occupations occur. If an Altithermal abandonment took place, then there is apparently a lack of continuity between Early and Middle Archaic occupations in southern Arizona. Southern Arizona could have been re-occupied by groups with no previous knowledge of the area or by groups that occupied the area thousands of years before or both. Each of these scenarios may involve groups with similar or different economies, therefore, it is important to consider both origin and economy in any analysis of post-Altithermal human response. The following section addresses the 81 ways in which archaeologists have interpreted human response to the Altithermal drought and the implications this has for examining post-Altithermal groups in southern Arizona during the Middle to Late Archaic transition. Antevs' Altithermal Antevs (1955) defined the Altithermal as a hot, dry period lasting from approximately 7000 to 4500 B.P. He based this interval on data supporting lake desiccation, stream entrenchment, eolian erosion, and soil formation in the Great Basin and the American Southwest (Grayson 1993). Today, we recognize that climatic change during this period following the onset of the Holocene affords more complex interpretation than Antevs' original definition and that the timing of these changes may not be concurrent across elevational and latitudinal gradients. For example, Davis and others (Davis 1984; Davis et al. 1986) have suggested that multiple thermal maxima explain the differences in timing of vegetation changes for low versus high elevation sites. During these maxima, summer temperature probably increased by 2 degrees C. The first real contest to Antevs' hot, dry Altithermal was advanced by Martin (1963) who used palynological data from southern Arizona and modern atmospheric circulation patterns to advance the idea of a warmer, but wetter Altithermal. Mehringer (1967) subsequently challenged Martin's results, demonstrating that evidence for increased effective precipitation came from strata younger than 5400 B.P. However, Martin's proposal that increased solar radiation promoted intensified monsoonal circulation has retained some credibility (Van Devender 1987; Van 82 Devender et al. 1984; but see Spaulding 1991). What remains at question, at least for the American Southwest, is whether or not intensified monsoonal circulation of precipitation results in increased effective moisture. The idea of a hot, dry Altithermal has received support from palynology (cf., Hall 1985), alluvial cutting and filling events (Haynes 1968; Waters 1986a), lake desiccation (Waters 1989), eolian deposition and soil formation (Holliday 1989; Monger 1995), and packrat middens (Spaulding 1991). Moreover, human use of the landscape, as well as humanly-produced water control features, seem to indicate that the middle Holocene was dry in the American Southwest and perhaps elsewhere in North America (Waters and Kuehn 1996). Erosion of small streams in western Iowa (Bettis and Hajic 1995; Bettis and Thompson 1981, 1982; Thompson and Bettis 1982) provides potential geologic evidence for widespread climatic changes during this period, but it also removes or fails to preserve the archaeological records of human response to these changes that can be found in certain landscapes. Similar erosive forces have been cited as the cause of missing archaeological site records for the Middle Holocene in southern Arizona (Waters 1986b). Correlate Human Responses Meltzer (1991, cf., Evans 1951) has recently made a case for a correlate human response to Altithermal climatic change. In West Texas, he discovered more than 60 prehistoric water wells, dug from an "Altithermal surface" to the receding water table below. These wells and their stratigraphy provide evidence that the warm, dry period 83 during the Altithermal was pronounced enough to create drought conditions. And, although Holliday (1985) suggested early in his investigations of soils on the southern High Plains that there were two droughts during the Altithermal, he more recently recanted (1989) this earlier hypothesis, claiming that the Altithermal period is more complex than originally suspected. Evidence for the onslaught of the Altithermal comes from a marl at the Mustang Springs site, near Midland, Texas; diatoms in the marl suggest slow, shallow, slightly saline water. Although the top of the marl unit ("Altithermal surface") is highly eroded, marl deposition seems to end concurrently with the beginning of eolian deposition, around 6900-6800 BP. Well-digging behavior probably began shortly after. A similar record of well-digging behavior has been recorded by Haynes (1995) at Blackwater Draw. Utilizing geologic records associated with the Blackwater Draw wells, Haynes describes the erosional and hydrologic evidence supporting this interpretation. One of the most important aspects of archaeological research that both Haynes and Meltzer bring to light is how archaeologists make decisions about whether humans are actually responding behaviorally to climatic stress: ...before concluding that the presence of Altithermal-age site is evidence of an adaptation to drought, one must determine,.. .the degree (if any) of ecological and climatic stress.. .Then, those selective factors must be linked directly to human adaptive responses detected archaeologically. Too often efforts to link Alithermal climates with a human response do not go beyond showing that certain cultural patterns are roughly "compatible" with a model of severe drought...(Meltzer 1991:237). During the Altithermal, there is evidence for bison population decline on the southern High Plains (Dillehay 1974; MacDonald 1981). Although Reeves (1973) infers, on the 84 basis of data from only a few sites on the northern Plains, that there is little evidence for a reduction in the number of bison. Additionally, it has been suggested that bison underwent evolutionary change during the Altithermal, with the resultant form (smaller Bison bison) surviving this period (Frison 1991; Reher 1979). Benedict and Olson (1978) propose that the Mount Albion complex of Colorado may hold the answer to where these bison and the people subsisting on them went during the Altithermal. They argue that the timberline environment may have been a "refuge" for both humans and animals during this drought period. Dillehay (1974) claims that bison population on the southern Plains declined again from A.D. 500-1200. Altithermal Abandonments? Following Antevs' definition of a hot, dry Altithermal, several authors (Berry and Berry 1986; Huckell 1996b; Irwin-Williams and Haynes 1970; Mabry 1996a; Waters 1986b; Waters and Kuehn 1996) have suggested that human groups in southern Arizona may have abandoned the area at the onset of the Altithermal. Preliminary analyses of site distribution data for identified and recorded Early and Middle Holocene indicates that there is a significant reduction in the number of sites dating to the Middle Holocene and that most of the Middle Holocene sites identified in Arizona are located on the Colorado Plateau (Mabry et al. 1997). This could, in part, be due to the lack of ease in identifying assemblages dating to the Middle Holocene, though a scan of the available data suggests this is unlikely (Mabry, pers. comm. 1997). Furthermore, the distribution of archaeological sites is highly dependent on the 85 geomorphic processes that have altered the landscape during the Holocene period (Waters and Kuehn 1996). A similar argument was made for an abandonment of the High Plains during the same time period, before all the data were in (Frison 1991). Are we just not recognizing Middle Archaic sites dating prior to 5500 B.P? Were occupations present, but the remains of them swept away during erosion of river channels? Or, has the density of hunter-gatherers in southern Arizona always been so light on the landscape that a gap is apparent rather than real? Whether or not there is a gap in occupation of southern Arizona during the Middle Holocene, there appears to be sufficient reason to believe that humans would have been influenced by climatic changes. The current evidence for southern Arizona suggests that human occupations changed rather drastically. This is further supported by studies of modern hunter-gatherers under drought conditions» Modeling Forager Response to Altithermal Drought The general !Kung strategy is to camp in an area where a mix of resources - including water, plant, and animal foods - is readily available. Shifts in campsite reflect changes in food preference, the availability of new vegetable resources, or new knowledge about the location of wide-ranging and constantly moving large game... .Of course, in all such moves, water is the major, and most limited factor. (Yellen 1976:56-58) Water is, indeed, one of the most limiting variables in forager strategies, particularly in arid and semi-arid settings (Meltzer 1995). Forager responses to severe droughts can include range shift or expansion, aggregation at permanent water sources, 86 or abandonment (Meltzer 1995). Each of these strategies should have distinctive archaeological consequences. In an effort to provide explanation to the evidence he gained from examination of Altithermal archaeological sites on the Southern High Plains, Meltzer (1995) developed a four-part model of Altithermal "adaptive strategies," based on examination of modern forager response to drought. He assigned names to the four types of human groups which would practice such strategies: hardscrabblers, collectors, wayfarers, and expatriates. Hardscrabblers are groups who remain in the region, despite the scarcity of resources. Collectors abandon the region, but annually or intermittently return to the region for certain resources. Wayfarers represent groups that abandon the region, only passing through it to get to another place (sensu Meltzer 1995) or may represent groups who have never lived in the region, but pass through it on the way to another place. Expatriates represent groups who, over time, abandon the region in favor of another, more suitable, territory. Two of these strategies, wayfarer and expatriate, would result in ephemeral sites with few, if any, identifying characteristics. The most identifiable characteristic would be a few water wells, at irregular depths. Collectors and hardscrabblers, on the other hand, leave more identifiable traces of their presence on the landscape. For both collectors and hardscrabblers, wells should be more numerous and appear in clusters. Other features and artifacts should also be more numerous. Though erosion could account for some site loss, it is likely that the lack of sites dating to the middle 87 Holocene in southern Arizona is the result of poor identification or erosion of ephemeral site types produced by wayfarer or expatriate strategies. Modeling Post-Altithermal Foraging Strategies Post-Altithermal forager response, presumably what has been called the Middle Archaic, is partly dependent on landscape use during the Altithermal drought period. Hardscrabblers and collectors, whose knowledge of the formerly drought-stricken region would be superior to those who either abandoned the region or who travel across it ephemerally, would presumably be quicker to identify favorable changes in the region, while wayfarers and expatriates would require longer time to become acquainted with the landscape and its available resources. Hardscrabblers and collectors would be well aware of raw materials located near their settlements. Wayfarers and expatriates, on the other hand, may spend a considerable amount of time exploring the region, resulting in a scattered distribution of sites. Exotic raw materials would comprise a significant portion of their assemblages. A lag in human response by any of the four idealized groups is unlikely to be geologically visible. The forces that drive human groups not living in a drought-stricken region (every group except "hardscrabblers") to claim or reclaim these territories under more favorable (wetter) climatic conditions is critically important to the evaluation of their post-drought strategies. A model that evaluates one such force was proposed by Huckell (1990) to account for the Middle to Late Archaic transition. 88 MIDDLE TO LATE ARCHAIC TRANSITION Huckell's (1995:16) defined Early Agricultural period is marked by evidence he cites for the "initial appearance of agriculture between 1500 B.C. and 1200 B.C." The growth of this conceptual scheme is first expressed best in his dissertation (Huckell 1990). Although he argues that evidence for the earliest appearance of agriculture in southern Arizona (the San Pedro phase of the Early Agricultural period, 1200-800 B.C.) portrays a mixed farming-hunting-gathering economy, he also cites evidence from architecture and material culture to suggest that this stage "appears to represent a surprisingly advanced stage in the transition from hunting and gathering to agriculture (Huckell 1990)". Huckell (1990)' 4 proposed that environmental conditions along river valleys began to improve at the end of the Holocene Altithermal, and that these improved conditions eventually led to the settlement and exploitation of these environments by immigrant groups of farmer-foragers. He further suggests that northward trending river valleys from northern Mexico into the Southwest would have provided a "natural corridor" for such migration. Huckell (1990) acknowledged the need to "establish the exact nature of the conditions along these rivers between 3,000 and 5,000 years ago." The distribution and recognition of Middle Archaic archaeological sites in the floodplain and their material evidence from those sites suggest that Huckell's argument may not be as simple as originally proposed. Middle Archaic groups occupying the floodplain immediately following the end of the Holocene Altithermal signifies recognition of this resource area by the first (or indigenous) groups in southern 89 Arizona. The nature of this evidence is presented in Chapters 6 and 7, and has profound implications for understanding the Middle to Late Archaic transition. ENDNOTES 1. Also called "geoglyphs," these are alignments and/or clearings of rock that are formed to create geometric, anthropomorphic, or zoomorphic designs. 2. Flaked and ground stone assemblages have greater frequency on the landscape, while features are more likely to be preserved. Additional examination of these two types of archaeological data, with a keen eye for trends in the co-occurrence of feature types, artifact types, or the distribution of these materials, could create patterns that are characteristic of certain time periods. The preservation of botanical or faunal materials in features could then be used to establish radiometric ages for such patterns. 3. Convention established by the Twelfth International Radiocarbon Conference, Trondheim, Norway, 1985. Departing from this convention, I have chosen to use the terms B.C. and A.D. rather than cal B.C. and cal A.D. and do not list the calibration curve used when citing the work of previous authors who have not reported this information. 4. Ages from Beta Analytic, Inc. are cited using the age estimate after correction for 813C. In cases where the original site report did not give both measured and 8 13C-corrected age estimates from Beta Analytic, Inc., it is assumed that the original publication has cited the 6 13C-corrected age estimate; however, these ages are marked with a star (*). Readers should consult either the original publication, the author of the volume from which the dates are cited, or laboratory-published date lists for radiocarbon date information. 5. Schiffer (1986) has demonstrated that mesquite charcoal can provide erroneous age estimates. Mesquite that has been dead for several hundred or even 1,000 years can be found on the desert floor. This old wood may have been particularly handy as firewood since green mesquite is very tough, making it difficult to remove branches from a live tree. A radiocarbon sample would date the death of the tree, rather than the cultural event, making the age of the charcoal erroneously old. 6. Geological investigations of the Coffee Camp site were carried out by Keith Katzer, but the relationships between geologic strata and pit features was established 90 by the archaeologists after fieldwork was complete by projecting feature elevations across the site and "correcting" those elevations based on the presumed conformability of geologic strata. However, the geologic strata need not be conformable or continuous across the site. It is difficult to tell whether the mixed radiocarbon ages of some strata are due to radiocarbon date error or error in stratigraphic placement of the dates. 7. The original report (Halbirt and Henderson 1993) gives neither lab number nor information on the nature of this date. It is assumed that the date was run at Beta Analytic, Inc. and that the reported age is corrected for •313C. 8. Huckell (1995) would later define the Cienega phase of the Early Agricultural period on the basis of excavations at these sites. 9. The Clearwater site was at the time comprised of the San Agustin Mission site, the Brickyard site, and the Mission Road site. For the purposes of simplicity and distinction from historic occupations of the same area, the Early Agricultural component of the site has been renamed "Clearwater" (Doelle 1996). 10. At the time of Huckell's publication, only Santa Cruz Bend, Stone Pipe, and Square Hearth had been excavated (Mabry and Clark 1994; Mabry 1995). Square Hearth is not included in the Early Agricultural Period. 11. A radiocarbon date on this feature yielded an age estimate of 2790 ± 50 B.P. (Freeman 1997), falling within the expected range of dates for the San Pedro phase. The remainder of the features at this site were early Cienega phase in age. 12. Each of these later Cienega phase sites shows evidence for occupation in spring, summer, and fall. There are no winter indicator plants, making it difficult to absolutely assign these sites to year-round use. 13. Meltzer (1995) has conducted a rather thorough review of the anthropological and ethnoarchaeological literature, surveying specifically for studies of how modern foragers respond to drought conditions. Some of the key points are highlighted in the following section. 14. The highlights of Huckell's (1990) argument are presented in this discussion. Those interested in this interpretation should refer to his dissertation for a more complete discussion. 91 CHAPTER 3 HISTORIC STREAM BEHAVIOR AND HYDROLOGIC PROCESSES OF THE SANTA CRUZ RIVER INTRODUCTION ...the present is the key to the past... This simple statement, often cited as the definition of uniformitarianism, once caused great anguish in fluvial geomorphology, particularly among those who observed that extreme flood events did not occur in historical records (Baker 1988a, b). However, "uniformitarianism in its twentieth century form... [proposes only that an hypothesized event]...obeys the laws of physics [or of human behavior] and is consistent with the field evidence (Baker 1988a, b)." This dictum is useful both in archaeology and geosciences. In order to utilize this principle, the geologist or archaeologist must first understand the laws or rules under which natural or human systems operate. These laws are defined on the observation of modern and historic stream processes and on the documentation of the effects of those processes preserved in historic and prehistoric alluvium. Desert stream systems create particularly dynamic stratigraphic records. Fluvial processes preserve archaeological sites and geologic deposits, but can also be erosive, removing parts of the archaeological and geological records that may have been preserved otherwise. In the course of its history, natural processes of the Santa Cruz River have exposed preserved portions of this record. Archaeologists and 92 geologists have recorded these natural exposures as well as exposures artificially revealed by modern construction activities. We now have a good record of archaeological and geologic events on the Santa Cruz River over the past 8,000 years. In addition to the prehistoric sedimentary record, we also have a historic record of meteorological events and their effects on the character of the Santa Cruz River. Like archaeology, geology uses analogy to determine what prehistoric landscapes were like. From analogical reasoning, geologists are able to develop models based on the physical characteristics of modern systems. Because the known conditions of the Santa Cruz do not always reflect the possible past conditions of the stream, it is sometimes necessary to model those conditions using a number of different sources of data. The purpose of this chapter is to outline the laws or rules under which desert streams operate, and to examine the known hydrologic processes that act on the Santa Cruz River specifically. Five types of paleohydrologic data are typically used to reconstruct river behavior: (a) historical reconstructions, which use non-systematic historical records to reconstruct flow conditions; (b) regime-based reconstructions, which use properties of the drainage network or channels to infer past flow conditions; (c) palaeocompetence reconstructions, which relate characteristics of channel sediments to flow parameters; (d) geobotanic methods, in which the characteristics of vegetation along the channel are used to reconstruct flow history; and (e) paleostage indicators, which record the stage of individual flows (Wohl and Enzel 1995:24). 93 This dissertation utilizes data from most of the above parameters as well as from the archaeological record to interpret the sedimentary and stratigraphic record left by the Santa Cruz River. PALEOCOMPETENCE RECONSTRUCTIONS Paleocompetence reconstructions focus on the relationship between streamflow and sediment characteristics. Typically such reconstructions involve calculation of discharge estimates based on a number of variables including channel width and depth, and valley slope. Discharge estimates are dependent, in part, on local variables such as sediment source, transport mechanisms, resisting forces such as vegetation and geology, and precipitation. Because of the incomplete nature of prehistoric alluvial records, it is generally inappropriate to calculate discharge estimates. Paleocompetence can be inferred, however, based on knowledge of the local variables that affect a particular stream and on flow conditions necessary to produce given sediment characteristics. What follows, therefore, is a general discussion of flow conditions in desert streams and the kinds of sediment characteristics that are used in southern Arizona to make inferences about paleocompetence. Desert Stream Processes Desert streams operate in a unique manner (Graf 1988a). The main difference between streams in humid and dry regions is, of course, precipitation. Precipitation can work together with a number of other variables, including vegetation, topography, 94 lithology, temperature, geologic structure, and human behavior in determining the operation of stream systems. Precipitation and temperature affect the kind and number of plants supported. The small number of plants near desert streams allows precipitated moisture to flow (as runoff) virtually unhindered into the stream system, so that water received by the stream system arrives with a higher energy to entrain and transport sediment (Parker 1995). The kind of vegetation supported by desert precipitation can also have an effect on sediment entrainment. Desert plants produce less debris than plants in humid regions, inhibiting pedogenesis (Bull 1991). As a result, unconsolidated sediments near stream channels are more likely to be entrained by the energetic desert stream system. Flow conditions also differ between streams in dry and humid regions. Desert streams experience intermittent flow in some or all reaches and tend to react quickly to extreme flow events, enhancing the spatial and temporal variability of streamflow. The great variability in discharge in dry region streams (Baker 1977; Chippen and Bue 1977) promotes channel instability (Burkham 1981; Graf 1988b), and slow rates of channel recovery (Wolman and Gerson 1978). Desert streams are characterized as having unstable channels (Burkham 1981; Graf 1988b), yielding abundant sediment (Langbein and Schumm 1958) that is entrained or stored depending on streamflow conditions. Flashy discharge caused by localized meteorological events is transmitted quickly through a terminable portion of the stream system (Burkham 1970). Streamflow in humid regions, on the other hand, is more regular and often differs little during higher streamflow events. Channel dimensions are regulated by previous 95 streamflows (Schumrn 1977) and are determined by sediment load, discharge, and valley slope (Waters 1992). Because streamflow is regular, channels in humid regions are more stable than those in dry regions. The regularity of flow that characterizes humid region streams promotes channel recovery after extreme events (Costa 1974). Thus, models for channel change that have been developed for perennial streams in more humid environments are not adequate for explaining the operation of desert streams. Sediment Characteristics Exhaustive discussions of the processes of sedimentary deposition and the characteristics of sediments in alluvial environments have been compiled by numerous authors (e.g., naming only a few, Boggs 1987; Church 1981; Miall 1978, 1992; Morisawa 1985; Ritter 1986; Waters 1992) and to prepare another such discussion would be pointless. However, a quick review of the factors that contribute most significantly to streamflow reconstruction and depositional environments along the focus reach will facilitate later discussions of Santa Cruz River stratigraphy. A number of factors contribute to sediment characteristics, including stream morphology, depositional regime, and velocity of streamflow. Each of these factors creates distinct sedimentary structures, facies, and sediment architecture (features within a deposit), which can be used to recognize the factors that created them (Miall 1992). The current morphology of the focus reach is that of a meandering stream. In the past, however, it is possible that the stream had a different morphology. In fact, 96 Waters (1987, 1988) has suggested that the Santa Cruz was a braided stream during the middle Holocene. Braided Streams Braided streams typically contain numerous bars that represent temporary storage of sediment deposited during periods of reduced streampower. Braided streams are characterized by highly variable discharge, allowing for storage of sediment that the stream is incompetent to handle (Leopold and Wolman 1957). Either lateral or vertical accretion of sediments is possible under a braided stream regimen; each of these depositional processes creates vertical sequences that reflect fluctuations in bedload and discharge (Boggs 1987). Meandering Streams The typical model for a meandering stream begins with a basal lag deposit of gravel overlying an erosional surface and is followed by an upwardly-fining sequence, terminating in fine overbank deposits typically comprised of mud and/or silt (Allen 1970). Sand and gravel in the basal unit typically decrease in size toward the top of the pile and display trough cross-bedding features'. These features are created by a complex set of processes related to low-flow and high-flow conditions. During low-flow conditions, the river hugs the outer (concave) part of the meander, but during high-flow conditions it takes a straighter path. Lateral shifting of the current causes the water to circulate in a strong spiral pattern known as helical flow. This type of 97 flow carries sediment across the stream channel to the inner bank, releasing it under lower velocity conditions in point bars. Turbulent, high velocity flow along the outer bank results in deposition of only the coarsest sediments (Boggs 1987). Overbank flow occurs during flood stage, leaving behind a deposit of fine silt and mud on the floodplain. Summary Inferences regarding paleocompetence can be derived from sedimentary characteristics displayed in prehistoric alluvial sequences. These inferences are based on a long history of documentation of the processes that create these characteristics. It is important to recognize that desert streams will operate in a manner unique to streams in more humid regions; therefore, historical reconstructions of the mechanisms under which the Santa Cruz operates can provide additional data to be used in inferential modeling. HISTORICAL RECONSTRUCTIONS Historic records of the Santa Cruz River, compiled predominantly by Betancourt and Turner (1990), provide examples of the types of flow conditions possible under channel parameters which do not exist today, but which may have existed at different times prehistorically. Historic records prior to incision of the Santa Cruz (see below) consist predominantly of cadastral surveys, newspaper reports, personal diaries or journals, and photographic records. 98 For the early part of the historic period, personal diaries and journals and historic cadastral surveys have provided some of the most detailed descriptions of the river. Occasionally, these records provide accounts of the vegetation present along the banks and on the floodplain of the river, the depth and width of the channel, the presence of water or of waterlogged areas, the agricultural potential of floodplain soils, the techniques used in agriculture, and the nature of sediments within the channel. Each of these types of data can be used to infer the flow conditions under certain channel parameters. For example, the depth and width of the channel, when combined with accounts of the frequency and extent of flooding, can yield important data about the nature of sedimentation that potentially occurs under those channel parameters. Historic records also provide examples of the way in which cause-and-effect relationships operate to produce channel changes. Historic records of the incision of the Santa Cruz channel (discussed in greater detail below) have revealed the multiplicity of factors that caused the Santa Cruz to incise to its present depth and the reasons for its extensive channel widening (Betancourt and Turner 1990; Parker 1995). Historic records also provide information about key landscape and hydrologic features that may repeatedly be the source of threshold changes in certain river systems. Stable landscape features such as mountains, bedrock features, and other impenetrable surfaces can act as boundaries to surface and groundwater flow and can influence hydrologic and sedimentary input into the stream system. Together with application of the known laws under which rivers operate, these historic data can be powerful interpretive tools. The following descriptions are used as 99 a basis for inferring the flow conditions possible under channel parameters in prehistory that are similar to those during the historic period. The following accounts of the historic Santa Cruz River are derived predominantly from data presented in Betancourt and Turner (1990). Historic Descriptions of the Santa Cruz River The Santa Cruz River is a deeply entrenched drainage much like many of the drainages found in southern Arizona. The river today exhibits steep, nearly vertical channel walls. However, before the turn of the century, historic accounts and photographs portray the Tucson portion of the river as a reliable source of water, shallowly incised, with perennial reaches that supported cottonwood groves and wet meadows called cienegas. These cienegas were located historically in at least two places, at the base of A-Mountain and at Punta de Agua, near San Xavier. Historic accounts of cienegas are supported by the presence of clay-rich sediments exposed in the stream banks 2 . These same historic sources also tell us that the Santa Cruz was a chain made of diverse microenvironmental links. Perennial reaches present in historic accounts covered only 20 percent of the entire river course. Other reaches of the Santa Cruz were intermittent and mesquite bosques were present in some parts of the valley. Although historic cadastral surveys rarely mentioned depth of the channel, short entrenched reaches were present by 1849 upstream of Tucson and by 1871 near San Xavier. 100 By 1882, the discontinuous arroyo near Tucson had incised three meters. Hydrologic changes related to an earthquake in 1887 created a new arroyo, Spring Branch, in a cienega near Martinez Hill. Natural processes were exacerbated by human impacts and by 1890 a combination of ill-designed diversion ditches, a declining water table, and a series of large floods caused headward erosion (cutting upstream) to form or deepen these entrenched reaches. Additional winter floods in 1905 and 1915 resulted in the continuously entrenched, only occasionally flooding, river with which we are familiar today. Historical records from specific reaches within the valley portray the individual character of the Santa Cruz throughout parts of the historic period. Though each reach changed over time, a fact that is highlighted by in-depth examination of historic records over several years (Betancourt and Turner 1990), the portrayals noted below have been gleaned from historical records because of their applicability to the study area and their apparent consistency over time. San Xavier and Martinez Hill Perennial flow during the historic period at San Xavier is ascribed to a buried dyke of flat-lying basalt. This dyke blocks flow into northern reaches of the Santa Cruz and draws groundwater to the surface. A spring called Punta de Agua is known from the San Xavier area and is attributed to the development of a cienega in this part of the river. 101 As early as 1699, when Father Kino visited San Xavier, the mission was well-endowed with extensive agricultural fields, supplied by numerous irrigation ditches. South of San Xavier, these historic accounts describe "plains and meadows covered with pasture" (Manje 1954 in Betancourt and Turner 1990:43); however, there were also miles with no perennial flow in this area. Springs are also reported at Agua de la Mission, near Martinez Hill and are attributed to yet another cienega. It is this cienega that was cut by the earthquake in 1887. A-Mountain/Sentinel Peak Historically, groundwater was forced to the surface by an impenetrable stratum in the narrow part of the valley surrounding A-Mountain (Hinderlider 1913 as cited in Betancourt and Turner 1990:58). Some historic accounts describe springs near the base of A-Mountain (Betancourt and Turner 1990:50), though most historic records place these springs only on the south side of the mountain (Betancourt and Turner 1990:107). Historic photographs and records indicate that the bottomlands (floodplain) surrounding Sentinel Peak were about a mile wide and were crossed by "irrigating canals in every direction" (Bartlett 1854:292-302, as cited in Betancourt and Turner 1990:50). 102 Focus Reach: Grant Road to Ruthrauff Road Few historic records are available for the area surrounding the Los Pozos site during the historic period; however, the accounts that are available portray a river that was a relatively reliable, though small stream for about 10 mi north of A-Mountain. The channel was rarely more than 3 m wide and often no more than 1-2 m deep. Several early accounts suggest that mesquite bosques were present intermittently along the course of the channel for several miles south of Point of the Mountains. A later survey places a thick mesquite grove just south of Los Pozos where the river becomes a narrow channel after occupying an approximately 10 m width. Rillito Creek Although not directly part of the focus area, sedimentary input from Rillito Creek has significant influence on the Santa Cruz River; its historic character is, therefore, summarized in the following discussion. As the Santa Cruz reaches Rillito Creek, historic accounts portray it as broadening slightly. Rillito Creek today exhibits a relatively straight channel pattern with nearly vertical walls. However, historic records suggest that the Rillito was once a wide, shallow, braided stream: [In 1858] The entire valley was an unbroken forest, principally of mesquite, with a good growth of gramma [sic] and other grasses between the trees. The river course was indefinite, - a continuous grove of tall cottonwood, ash, willow, and walnut trees with underbrush and sacaton and galleta grass, and it was further obstructed by beaver dams (Smith 1910:98). 103 This account of the Rillito channel has been confirmed by examination of paleoflood records, aerial photographs, and survey records (Slezak-Pearthree and Baker 1987). Changes in land use, a series of large floods, and the incursion of large, obstructive linear constructions such as roads, diversion dams, canals, and the Southern Pacific Railroad caused channel changes that occurred near the turn of the century. As a tributary channel to it, the Rillito responded to historic incision of the Santa Cruz by adjusting to a new base level. Headward erosion of the Rillito from its confluence with the Santa Cruz resulted in entrenchment of the Rillito in several reaches. Though little documentation of these changes is available, historic accounts of early floods along the Rillito suggest that the Rillito "[flowed] on an upper surface of alluvial fill [T-2]" (Graf 1984) and that during high-magnitude floods, water rose over the floodplain. However, by the 1890s, entrenchment and lateral erosion of channel banks had begun (Hastings 1958). By 1930, entrenchment was complete (U.S. Army Corps of Engineers 1986). Smith (1910) suggests that entrenchment of the Rillito and creation of its now nearly vertical banks was caused in part by overgrazing, cutting of floodplain grasses for hay production, concentration of runoff in cattle trails, and summer floods. Most of these effects were felt after the U.S. Army post was moved to Fort Lowell in 1872. Canada del Oro to Point of the Mountains Abundant historic survey data has been compiled from the area north of Cariada del Oro to the northernmost edge of the Tucson Mountains (Point of the 104 Mountains). Historically, the river was very broad here and exhibited a very shallow channel, rarely more than 1 m deep. Few other historic accounts are known from the area, however, and most of these are vague in placement. Tucson is described as "the lowest line of constant running water" (Emory 1857 in Betancourt and Turner 1990:51) and the river "never succeeds in reaching the Rio Gila" (Bell 1869 in Betancourt and Turner 1990:56). REGIME-BASED RECONSTRUCTIONS Utilizing data from historic period flows along the Santa Cruz River, Parker (1995) developed a model for channel change along the Santa Cruz River. Although his purpose was primarily related to predictive modeling of channel change for flood hazard assessment and planning purposes, he also employed the results of his research to develop a conceptual model for prehistoric channel changes. Parker employed geomorphic concepts developed by fluvial geomorphologists over the past 50 years, and utilized characteristics of the channel, the drainage network, and geomorphic surfaces in his model. Channel pattern, dimensions, and aerial photographs were used to reconstruct the lateral processes involved in channel change. Meteorological events (storms) and discharge records were used to examine the effects of spatially and temporally variable precipitation on hydrologic processes of the Santa Cruz River. He concluded that spatial variability of channel change is controlled by major landscape elements. The nature of channel change within a single reach is, therefore, predictable. 105 Parker (1995) has divided the portions of the Santa Cruz River flowing through Tucson into three reaches: San Xavier (Pima Mine Road to 22nd Street/Sentinel Peak), Tucson (Sentinel Peak to the Rillito River), and Cortaro (Rillito Cortaro Road). The study area of this dissertation encompasses all River to of the Tucson reach and parts of the Cortaro reach. Results of previous work in the San Xavier reach is examined in Chapter 4 as a reference point from which to understand new research within the study area. The focus area is contained within Parker's (1995) Tucson reach. Using Parker's (1995) dissertation as a guide, the geologic and topographic controls that affect channel change within these reaches are reviewed below. San Xavier Reach Previously confined by intramontane basin fill and the Quaternary Jaynes terrace, the Santa Cruz River enters the San Xavier reach where the basin widens. Here, floodwater is able to spread widely, depositing large volumes of fine-grained material. The reach upstream from Pima Mine Road is characterized by high sediment production and mobility; these large quantities of sediment are delivered to the San Xavier reach. As a consequence, the San Xavier reach is characterized by sediment storage, creating cienega conditions and aggrading floodplains with shallowly-incised or unincised channels. Coarser sediments are transported through the reach during intermittent periods of discontinuous arroyo formation. Although the reach has 106 experienced few periods of continuous arroyo entrenchment, only a fraction of the stored sediment was removed during these periods. Tucson Reach As the river enters Tucson, it is again constricted into a narrow valley, formed on the west side by alluvial fans emanating from the Tucson Mountains and on the east by the Jaynes, Cemetery, and University terraces. Because sediments are trapped south of A-Mountain, the Tucson reach receives little to no sedimentation from the San Xavier reach. Minor sedimentary input is derived from the Tucson Mountains and from erosion of the Quaternary terraces, contributing a rather coarse sediment load.' Low sedimentary input, coupled with sediment storage in upstream reaches, can cause sediment storage in the Tucson reach. Parker characterizes this stored sediment as "a wedge...[that] moves downstream through the Tucson reach" (Parker 1995:175). The steep valley slope causes frequent sediment removal at the downstream margin of this wedge. This zone of "sediment evacuation" serves as the threshold point for rare episodes of continuous arroyo formation extending upstream through the entire San Xavier reach. Headward erosion during periods of continuous arroyo formation and subsequent arroyo widening substantially increase sediment delivery to the Tucson and San Xavier reaches. Because storage space is low in the Tucson reach, the reach acts as a zone dominated by sediment transport. 107 Cortaro Reach At the Cortaro reach, the valley widens again and is bounded by alluvial fan sediments, derived on the west from the Tucson Mountains and on the east and north from the Santa Catalina Mountains. The resemblance [to the San Xavier reach] does not extend to channel morphology and the nature of channel change, however, because of differences in sediment source and hydrologic differences that are at least partly related to landscape controls. The proximity of the Cortaro reach to significant sediments sources, particularly the piedmont of the Catalina Mountains and the upper watersheds of Taupe Verde Creek and Pantano Wash result in high sediment delivery rates (Parker 1995:175). Channel widening, shifting channel positions, and frequent, low-magnitude changes in channel elevation are caused by a combination of increased discharge and coarser sediments in this reach. High rates of lateral erosion result in short sediment storage times, but high sediment input have made the Cortaro reach one of long-term aggradation. As a result, "the residence times of sediments stored more than a few meters below the flood plain is relatively high" (Parker 1995:176). Downstream of the Cortaro reach, the river becomes completely unconfined, eventually emptying into the area known as the Santa Cruz Flats. Parker (1995) does not address the effects of short-term sediment storage and gentler valley slope in the Cortaro reach on the upstream Tucson reach. Changes in sediment delivery, sediment storage, and valley slope within the Cortaro reach should have significant effects on the Tucson reach. Because flood flows on the southern portion of the Santa Catalina Mountains are significantly affected by the frequency, 108 size, and type of storms (Martinez-Goytre 1993; Martinez Goytre et al. 1994), the interaction of climatological processes and hydrologic response within the Cortaro reach would have played a key role in the removal or storage of sediments in the Tucson (focus) reach. According to Parker (1995), tropical and frontal storms produce the kind of intensity and extent to cumulate energy in downstream reaches. MODELING PREHISTORIC CHANNEL BEHAVIOR Utilizing both regimen-based reconstruction and paleocompetence reconstruction as a guide, Parker (1995) suggests that prehistoric periods of channel filling were the result of sustained periods of low to moderate discharge during which tropical and frontal storms would not occur or would occur with lower frequency than the present. In the absence of scouring floods created by these storm types, sediment is removed from or entrained through reaches of long-term sediment storage by shallowly-incised, meandering channels or through the incision and excavation of discontinuous arroyos. Parker (1995) cites Waters' (1988) units V 2 and V 3 , as examples of such channel filling episodes. Parker (1995) further contends that such periods would be dominated by monsoon-caused flood, incapable of producing storms of a magnitude or frequency to sustain or intensify streampower in downstream reaches. Monsoonal storms may, however, have an effect on smaller tributary basins. Parker (1995) also suggests that continuous and deep channel entrenchment produced by tropical and frontal storms would have been dependent on the character of the pre-entrenchment channel. Discontinuous, entrenched reaches would have been 109 impacted by tropical and frontal storms to a greater effect than indistinct, unincised channels in vegetated cienega environments. Under the right pre-entrenchment conditions, these discontinuous, entrenched reaches would have coalesced into a single continuously entrenched channel. It is important to recognize that neither Waters (1988) nor Parker (1995) were equipped with prehistoric records from the Tucson reach to use as a guide for reconstruction of the prehistoric behavior of the Santa Cruz River. Chapters 4, 5, and 6 (below) describe the prehistoric record in this reach as a to test of Parker's (1995) model on that prehistoric record and an explanation for the processes that might have created the record documented for the focus reach. ENDNOTES 1. Variations from the fining-upward model do exist (cf., Collinson 1978); therefore, this model should only be used as an ideal and not the rule. 2. It must be noted that clay-rich sediments, in and of themselves, are not necessarily indicative of cienega conditions. This term has been misused in much of the archaeological literature for the Santa Cruz River. Bruce Huckell (pers. comm., 1995) has suggested that cienega soils would likely display a mollic epipedon, rather than a histic epipedon, probably indicating hydrophytic plant growth. These soils appear to be richer in organic matter than other alluvial deposits. Other characteristics of note appear to be the thickness of the deposit and the presence of a permeable underlying layer in which evidence for perennial distribution of water is often found. 3. Parker (1995) attributes the coarse sediment load to short transport distance and coarseness of the source material. 110 CHAPTER 4 GEOARCHAEOLOGICAL BACKGROUND OF THE SANTA CRUZ RIVER WITHIN THE TUCSON BASIN Today the Santa Cruz River and Rillito Creek are deeply incised, ephemeral streams. The Santa Cruz originates just north of the U.S.-Mexico border in the Canelo Hills, running south to Santa Cruz Sonora before turning north and re-entering Arizona east of Nogales. The reach of the middle Santa Cruz encompassing the project area is influenced predominantly by large and small drainages originating in the Tucson Mountains, the Santa Catalina Mountains, and on the surrounding bajada surfaces. Rillito Creek, which is a tributary to the Santa Cruz, is served by a series of large and small drainages originating predominantly in the Santa Catalina and Rincon mountains and farther upstream by Pantano Wash, Cienega Creek, and other drainages originating in the Santa Rita, Whetstone, and Empire mountains. As a rule, the differing sources deliver distinct materials to the streams. In addition to drainage contribution, the surrounding mountains and terrace surfaces can affect the hydrologic system of the Santa Cruz in other ways. Small volcanic hills such as Sentinel Peak and Martinez Hill, along with impenetrable petrocalcic horizons in surrounding terraces, often act as a barrier to surface flow and occasionally force groundwater to the surface. The mountains also can have an effect on atmospheric processes, favoring precipitation in some drainages, rather than others (Martinez-Goytre 1993). 111 As outlined in the previous chapter, the current character of the Santa Cruz River is not indicative of its past historic status. During the prehistoric period, the Santa Cruz was a dynamic stream system, at times entrenched and other times certain reaches would flow perennially. Where not covered by soil cement, sections exposed in the walls of today's arroyo provide access to this prehistoric record. Geologists and archaeologists examining the walls of the present arroyo and excavating trenches in the floodplain have described pieces of this record in reaches extending from an area south of the San Xavier reservation to Ina Road north of Tucson's city center. The purpose of this chapter is to describe previous and current research in each of these reaches (excepting the focus reach) as a background for the present study, in order to synthesize work conducted over 30 years by several scholars. OVERVIEW OF THE SANTA CRUZ RIVER TERRACES Using a combination of aerial photometric mapping and field reconnaissance, the Arizona Geologic Survey mapped the geology of the entire Tucson metropolitan area (Jackson 1989; McKittrick 1988; Pearthree et al. 1988). They defined five Quaternary terraces of the Santa Cruz River (Qt 1 through Qt5). Throughout this chapter, the Arizona Geological Survey's units are referred to in locational maps and in geological referencing of site areas. Four of these terraces were first described in the early twentieth century by Smith (1938) and have been correlated with the surfaces defined by the Arizona Geologic Survey and with sedimentary units identified by 112 Haynes and Huckell (1986, see below). Table 4.1, below, shows the relationship between geologic units described by different researchers. Table 4.1. Correlation between geologic units defined by different researchers. Smith (1938) McKittrick (1988) Haynes and Huckell (1986) 'Bottomlands'' Qt1 and Qt2 A3 through D Jaynes terrace Qt3 A, Cemetery terrace Qt4 A, University terrace Qt5 Qt5 and Qt4 (the University and Cemetery terraces) The thick petrocalcic horizon (a layer of hard "caliche," calcium carbonate) capping the oldest and highest terrace (Qt5 or University terrace), suggests an early to middle Pleistocene age (Anderson 1987). Soils forming the upper surface of the next lower terrace (Qt4 or Cemetary terrace) exhibit a well developed clay horizon and varying degrees of carbonate accretion, indicating that deposition terminated sometime during the middle Pleistocene. Except where buried by more recent deposits, the Cemetery (Qt4) and University (Qt5) terraces are of little concern archaeologically. These terraces were deposited during the middle Pleistocene or earlier. To date, no archaeological material has been discovered in North or South America that predates the late Pleistocene. 113 Qt3 (the Jaynes terrace) The third terrace (Qt3 or Jaynes terrace) is preserved in only a few places. A radiometric date on groundwater-derived carbonates from a zone 2.5 m below the Jaynes terrace surface yielded a minimum age estimate 18,400 B.P. (late Pleistocene) for the beginning of deposition (Haynes and Huckell 1986). Because the Jaynes terrace has a complex history that is not fully understood, it is not clear if it could contain archaeological deposits that are contemporaneous with its formation. Although this terrace has been identified in numerous reaches of the Santa Cruz River (Haynes and Huckell 1986, Jackson 1989, McKittrick 1988), the Jaynes terrace remains poorly defined. McKittrick (1988) and Jackson (1989) defined the surface of Qt3 as comprised, in part, of moderately to weakly developed argillic (clay) horizons, which often indicate Pleistocene age. However, Haynes and Huckell (1986) suggest that, in places, this terrace is buried by younger deposits of the Santa Cruz River, and that it merges with a lower terrace (Qt2), thereby forming a compound terrace. Therefore, sediments of post-Pleistocene age can be found on terraces mapped as Qt3; these sediments can potentially preserve subsurface cultural materials. Nevertheless, the Pleistocene terrace was probably completely abandoned by 8,000 years ago (cf., Waters 1987, 1988). Qt2 (the Holocene terrace) The next lowest terrace (Qt2 or Smith's  "bottomlands") is comprised of unconsolidated sediments of Holocene age, between at least 8000 and 100 years old'. 114 These sediments demonstrate that the Santa Cruz experienced at least five major periods of deposition and four downcutting episodes during the last 5,000 years. These alluvial cycles are bracketed in time by estimated ages on buried archaeological remains and on the development of calcic horizons, and are further supported by 71 radiocarbon-dates from organic materials incorporated into the sediments (Haynes and Huckell 1986). In his study of alluvial strata exposed in the arroyo wall south of Martinez Hill, near the San Xavier Mission, Waters (1987, 1988) extended the range of this alluvial sequence by at least 2,000 years and refined the chronology with an additional 27 radiocarbon dates, eight archaeomagnetic dates, and temporally diagnostic pottery sherds. Pre-8000 B.P. deposits along the San Xavier reach of the river consist of a unit of channel gravel. Deposition of this gravel is followed by a lengthy period of channel erosion and widening between 8000 and 5000 B.P. and later by valley aggradation between 5500 and 2500 B.P. This aggradational episode is followed by four epicycles of cutting and filling occurring during the last 2,500 years in this part of the valley (Waters 1987, 1988). The oldest and deepest Holocene fill identified north of Martinez Hill is comprised of slope wash deposits (Waters 1987, 1988), yielding radiocarbon dates between 4800 and 3900 B.P. and containing Middle Archaic artifacts, isolated features, and archaeological sites (Haynes and Huckell 1986). A weakly developed paleosol caps the alluvial deposit overlying this unit and has yielded radiocarbon dates from about 3700 to 2600 B.P. Buried Middle and Late Archaic artifacts and features have also been found on, or excavated into, the top of 115 this unit. Non-diagnostic lithic artifacts and a hearth have been found within the next higher fill, comprised of alluvial silts and sands and separated by three or more bands of clayey sediments. Radiocarbon dates within this unit range from 2500 to 2000 B.P. Fluvial sands, separated by bands of clay form the next alluvial unit. Late Archaic hearths and occupational horizons found within this unit probably represent artifacts redeposited from older surfaces. Colonial and Sedentary period Hohokam ceramics and habitation sites found within the upper portion of this unit have yielded radiocarbon ages between 1800 and 1000 B.P. (ca. AD 225-1100). Sedentary and early Classic period sites are found on the surface of sand dunes, and are partly covered by alluvial overbank deposits. These alluvial deposits are also found in a floodplain facies, capped by a weak soil, and containing Sedentary and Classic period archaeological resources. Radiocarbon dates on these deposits range from 1000 to 190 B.P. (ca. AD 1100-1760). This fill potentially contains protohistoric resources, which can be difficult to segregate from the earliest ceramic period and historic period remains. Qtl (the Historic terrace) More than a meter of mud containing recent trash has been deposited on the pre-twentieth century terrace by the most recent overbank floods in 1977 and 1983 (Mabry 1993; Freeman 1995, unpublished field notes). A 2- to 3-meter high terrace of slackwater flood deposits also formed in the mouths of tributary arroyos along the Tucson reach of the river during these floods, where mud and gravel were spread 116 across the wide, active floodplain near Marana. Qt1 represents this youngest and most recent floodplain activity along the Santa Cruz River. Often this terrace is buried or otherwise modified by historic and modern activities such as mining and dumping. It is rarely identified along the Tucson reach of the Santa Cruz River. Since its turn-of-the-century entrenchment, the channel of the Santa Cruz has reached a gradient on a more resistant bed or on older tougher sediments, limiting the potential for further downcutting. Channel erosion is now more lateral than vertical. For this reason, cement-stabilization of the banks is being implemented. During extreme flood events, which may be associated with an intensification in monsoonal precipitation, the Santa Cruz River overflows and erodes its banks, creating a hydrological planning nightmare. SAN XAVIER Beginning in the late 1960s and continuing through the late 1970s and late 1980s, Vance Haynes and students from his classes at the University of Arizona began mapping alluvium exposed in several arroyos between Pima Mine Road and the Hughes Access Road (south of Martinez Hill). They also mapped exposures north of Martinez Hill at the San Xavier gravel pit, at the Airport Wash and in Avra Valley. Together with Bruce Huckell (then at the Arizona State Museum), Haynes had the opportunity to map exposures of archaeological features and artifacts along Ina Road. Haynes and Huckell (1986) identified five major (Units A-E) and nine minor episodes (A 2 through E) of sedimentation from the late Pleistocene through the historic period. 117 During the late 1980s, Stafford (1986) and Waters (1987, 1988) conducted geologic investigations of the San Xavier reach of the Santa Cruz River in conjunction with archaeological research in the private sector. Waters' (1987, 1988) reports of the reach between Pima Mine Road and Martinez Hill are to date the most inclusive discussions of these radiocarbon dated units. He identifies seven major geologic units within this reach (units I through VII). However, because many of the reaches, described hereafter, will refer primarily to the report by Haynes and Huckell (1986), I have included their units in parentheses following the Waters' (1987, 1988) unit designation. Unit I Unit I (unidentified by Haynes and Huckell) is comprised of sands and gravels, overlain by a black "cienega" clay, on which pedogenesis has taken place. A radiocarbon date from the upper portion of Unit lb gave an age estimate of 7970 ± 130 B.P. (Beta 14537) run on organic carbon extracted from the clay. This is the earliest age on Holocene sediments within the Santa Cruz Valley. Unit II Unit II (Haynes and Huckell unit B 2 ), which unconformably overlies this black clay, is distinctly younger. Charcoal from a hearth in the upper portion of Unit II yielded a corrected age estimate of 2570 ± 210 B.P. (Beta 13707). The hiatus that follows deposition of Unit I appears to be relatively continuous across the valley; 118 however, older dates recovered by Haynes and Huckell (1986) indicate that deposition of this unit probably began as early as 4000 B.P. In the Brickyard Arroyo, Haynes and Huckell (1986) also identified a unit near the floodplain margin that appeared to have been derived from slopewash from the nearby bajada (Waters' unit IIsw, Haynes and Huckell unit B 1 ). Radiocarbon dates on these sediments have yielded ages ranging from 5500 to 4500 years B.P. The section from San Xavier is missing this Middle Holocene unit, that is well-preserved in downstream reaches. Unit III Unit III (Haynes and Huckell 1986, unit C 1 ), a series of sands, gravels and silts follows deposition of a second (brown) "cienega" clay (Unit IIc). Dates on this unit range from approximately 2400 to 1900 B.P. Unit IV Unit IV (Haynes and Huckell 1986, unit C 2 ) consists of channel fill and is comprised of sand and silty sand with discontinuous silt and clay lenses (Haynes and Huckell 1986; Waters 1987). Radiocarbon dates on this unit range from approximately 1800 to 1000 B.P. Unit V Unit V (Haynes and Huckell 1986, unit C 3 ) appears to be comprised of overbank and/or slackwater sediments (predominantly silts and clays), although 119 Stafford (1986) has identified a channel component of this unit. Radiocarbon dates on this unit range from approximately 900 to 600 B.P. Unit C 3 is also underlain by a series of vegetated dunes in a portion of the valley (Haynes and Huckell 1986). Unit VI Unit VI (Haynes and Huckell 1986, units C 3 and D) are comprised of channel fill and overbank deposits of Unit C 3 and modern silty clay and sandy clay overbank deposits of Haynes and Huckell's unit D. Radiocarbon dates on the older component range from roughly 450 to 190 B.P. and the younger dates are essentially modern in age. Unit VII Unit VII (Haynes and Huckell unit E) is described by Waters (1987) as modern clay and silt overbank deposition following historic entrenchment of the Santa Cruz River. The radiocarbon content of charcoal found within this deposit yields very young age estimates or exceeds the modern value. Radiocarbon values in excess of 100 percent modern are due to an increase in atmospheric radiocarbon produced by nuclear testing (Cain and Suess 1976). RIO NUEVO SOUTH / A-MOUNTAIN The Rio Nuevo South property is located south of Congress Street and west of the Santa Cruz River within a highly urbanized portion of the Tucson Basin. The 120 property crosses a series of deposits representing former channels, terraces, and floodplains of the Santa Cruz River. Here, the Santa Cruz River is diverted around the Sentinel Peak-Tumamoc Hill volcanic complex, a series of low hills comprised primarily of basaltic andesites, porphyrys, tuffs, and lavas laid down during the Miocene and Oligocene. Tumamoc Hill and Sentinel Peak (A-Mountain) were uplifted, faulted, and tilted southeastward sometime after 20 million years ago (Mayo et al. 1968; Phillips 1976). Today, both the hills and the river provide important sources for sedimentary input into the project area. The floodplain and former channel of the Santa Cruz River are well preserved here due to the presence of Sentinel Peak, a setting which provides protection of the west bank from high-velocity flows. Archaeological resources found on the Rio Nuevo South property include a system of historic and prehistoric canals at the base of A-Mountain (the A-Mountain Canal System, AZ BB:13:481, ASM) and a series of historic and prehistoric components including the Tucson Pressed Brick Company and the Clearwater site (AZ BB:13:6, ASM). The prehistoric component on the Rio Nuevo South property is referred to as the Clearwater site. This site name, proposed by Doelle (1996) also encompasses prehistoric components found on Spruce Street and Brickyard Lane, formerly known as the Brickyard site (Smiley et al. 1953). Clearwater Site (AZ BB:13:6) The highest density of prehistoric archaeological resources on the Rio Nuevo South property is represented by an Early Agricultural period component, located in 121 the southern portion of the Rio Nuevo South property, south of Congress Street and west of the river (Diehl 1996b). Another relatively high density of features is located along Brickyard Lane, south of the Rio Nuevo property (Elson and Doelle 1987). A few isolated features have been found in trenches along Spruce Street (Diehl 1996b). The area encompassed by these features suggests that the Clearwater site is quite large, however few individual features have been excavated. Because the area excavated is relatively small, it is difficult to determine whether the individual loci forming the defined Clearwater site is one large site or a series of smaller sites. Because the geologic investigations described below were conducted only in conjunction with the portion of the Clearwater site located on the Rio Nuevo property, the results of those investigations are presented here. Earlier investigations also identified Early Agricultural period pit structures (Elson and Doelle 1987; Smiley et al. 1953). Testing was conducted on the Rio Nuevo property in spring and fall 1995 (Diehl 1996a; Thiel 1995a, 1995b). A portion of this testing was conducted because a storm drain, following the alignment of Spruce Street to the southern portion of the Rio Nuevo property was to be installed. Excavation of Early Agricultural features in the southern portion of the property was conducted as part of that storm drain project. Two pit structures, thirteen extramural pits, a trash midden, and a single, flexed human burial were excavated within floodplain sediments comprising a narrow strip of land in the southern portion of the Rio Nuevo property (see below). The storm drain right-of-way cuts through most of that narrow strip of land. The Early Agricultural 122 settlement potentially covers a roughly 7500 sq m area. Only a portion of this area was excavated. Buildings and other features related to operation of the Tucson Pressed Brick Company directly overlie most of the Early Agricultural features, removing or disturbing any prehistoric sediments that may have postdated the Early Agricultural period features; some Hohokam artifacts were found in overlying the Early Agricultural features. Both pits and pit structures were filled with trash containing animal bone, flaked and ground stone, and the remains of plants, including maize. Radiocarbon dates on annuals found in these features are presented in Table 4.2. These dates produced an average age estimate of 2471 ± 16 B.P. The date corresponds well with the presence of Cienega (n=12), San Pedro (n=2) and Cortaro (n=2) projectile points found at the site. The date also places the site on the earlier part of the Cienega phase and roughly contemporaneous with the Wetlands site (AZ AA: 12:90), described later is this chapter. On the basis of plant ubiquities, Diehl (1996b) suggests that maize farming, small seed harvesting, and cactus fruit harvesting all played a role in the economy of these early inhabitants. Seasonality was not addressed in the site report. The Clearwater site also had a smaller, but more diverse faunal assemblage, but like at many of the Early Agricultural period sites in the floodplain appeared to be focused on lagomorphs and artiodactyls. The site also contained Cienega phase pottery. 123 Table 4.2 AMS radiocarbon dates from A-Mountain mitigation features. Corrected Calibrated Age Age (B.P.) b(1 S.D.) Calibrated Age (2 S.D.) 2580 ± 60 2500 ± 60 805 - 770 B.C. 785 - 505 B.C. 825 - 525 B.C. 800 - 405 BC,. Maize cupule 2600 ± 50 2390 ± 50 810 - 780 B.C. 505 - 395 B.C. 825 - 560 B.C. 760 - 380 B.C. Maize cupule 2420 ± 50 745 - 400 B.C. 770 - 390 B.C. 780 - 515 B.C. 380 - 205 B.C. 795 - 410 B.C. Sample a Feature Material B-90225 1006.02 Maize cupule B-90226 1006.03 Maize cupule B-90227 1040.04 Wood B-90228 1016 B-90229 1029 B-90230 insufficient carbon to obtain date B-90231 1040.02 Prosopis seed 2500 ± 50 B-90232 1032 Prosopis seed B-92617 1009 Prosopis seed 2250 ± 50 2390 ± 70 B-92618 1022 Prosopis seed B-92619 370 Maize cupule B-92620 1014 B-92621 B-92622 Mean` 2440 ± 60 525 - 390 B.C. 760 - 405 B.C. 395 - 180 B.C. 775 - 365 B.C. 785 - 390 B.C. Prosopis seed 2430 ± 60 2510 ± 50 760 - 400 B.C. 785 - 525 B.C. 780 - 385 B.C. 800 - 415 B.C. 1020 Prosopis seed 2440 ± 60 760 - 405 B.C. 785 - 390 B.C. 371 Prosopis seed 2480 ± 50 775 - 425 B.C. 2471 ± 16 760 - 415 B.C. 790 - 405 B.C. 765 - 410 B.C. a Piii samples are from Beta-Analytic, Inc. Corrected using the C13/C12 ratio; in years before present. `Calculated mean excludes sample B-90232. S.D. = standard deviation b On the basis of a high proportion of bifaces at the site, Sliva (in Diehl 1996b) suggests that the site represents a short-term camp associated with a high-degree of residential mobility. The assemblage pattern runs counter to other analyses, including the high ubiquity of maize in feature contexts, and also runs counter to other Early Agricultural period assemblages. For this reason, Sliva suggests the possibility that the tool-type patterning reflected in the site assemblage may represent a task-specific 124 function. Given the small proportion of site excavated, its presence on the outer edge of the known boundaries of the Clearwater site, and the relatively few number of features excavated, assemblage bias must be considered a possibility. Project Area Background The Rio Nuevo south property is located in an area mapped by the Arizona Geologic Survey as representing Qt2 (Figure 4.1). This terrace represents a complex suite of geologic events and encompasses a large part of the Holocene epoch (postPleistocene) in which most of the evidence for prehistoric occupation of the Santa Cruz Valley is found. Further confounding the geology of the project area is that the Sentinel Peak-Tumamoc Hill complex, in protecting sedimentary deposition along the western side of the Santa Cruz River, has preserved sediments and promoted the development of soils not likely to be found elsewhere along the Santa Cruz River. The geomorphology of the project area and its deposits has been previously studied by Ahlstrom et al. (1994) and Katzer (1987). Because Katzer's (1987) examination of the project area was limited to a few widely dispersed trenches, he was able to identify only one of the five geomorphic units described in this report. Ahlstrom et al. (1994) were similarly limited by the scope of their project, but they were able to identify what are essentially included in the following section as allostratigraphic Units 2, 3, 4, and 5 (though Units 3 and 5 were lumped as a single 125 3565000m fki 3563000m N Figure 4.1. Location of Clearwater site showing surficial geologic deposits and relevant Rio Nuevo and Alameda Street cross-sections, A-A', B B', and C-C'. UTM coordinates indicate the position of map on the USGS 7.5 topographic quadrangle, Tucson, AZ. - unit defined only as "channel deposits"). The fifth unit described in this chapter, allostratigraphic Unit 1, is critical to the identification and evaluation of the prehistoric 126 component on the property and is temporally and geomorphically separate from later deposits. The geomorphic units defined in this report (allostratigraphic units 1-5) were summarily covered in the report and the data recovery plan for testing on the Rio Nuevo South property (Diehl and Freeman 1996). Unit boundaries are defined on the basis of their geomorphological origin and associated ages in order to provide a model for recovery of archaeological resources on the property. The following discussion of these allostratigraphic units utilizes data recovered during all phases of archaeological research as well as data from previous reports (Ahlstrom et al. 1994; Katzer 1987), to elaborate on the relationships between these units. Methods Backhoe trenches excavated for archaeological testing were utilized to define the relationships between geologically distinct units. When unique geologic features were exposed, trenches were extended deeper, where necessary, to better define the boundaries between stratigraphically distinct geologic units. The relationships between distinct stratigraphic and geomorphic units were identified and recorded using a modified adaptation of the USDA and Folk classification systems (Folk 1954, 1974; Soil Survey Staff 1951, 1975). A 4.5-m-deep stratigraphic trench was also excavated to examine the relationship between prehistoric channels of the Santa Cruz River and project area features, including a historic canal. The location of trench profiles, soil 127 profiles, and the stratigraphic pit (referred to in descriptions below) are shown in Figure 4.2. Stratigraphy The project area is generally defined by five distinct episodes of sedimentation and soil formation: Unit 1, an early series of alluvial deposits, is truncated by Unit 2, a younger series of alluvial deposits, the upper portion of which forms what has been previously called a "cienega." These two deposits are likely followed by an incision (as yet obscured by modern mining and dumping) and deposition of Unit 3, a sequence of younger alluvial sediments, including a prehistoric channel of the Santa Cruz River, and Unit 5, a second channel. The surface distribution of these deposits is represented in Figure 4.2. The project area is covered in places by dumping and potentially by historic channels created when diversion ditches were further eroded by natural processes (Unit 4). Allostratigraphic Unit 1: Early Agricultural Period Floodplain Alluvium Unit 1 is comprised of a series of alluvial deposits consisting primarily of tan silts and very fine sands interrupted by thick, dark brown, silty clay to clay bands (Figure 4.3, units I-V; Figure 4.4, unit I). These fine-grained deposits are the result of slow deposition of overbank or slackwater sediments on the floodplain of the Santa Cruz River. Three strongly developed paleosols and two weakly developed paleosols 128 UNIT 5 UNIT 3 -.'% ' A—Mountain Testing Stratigraphic Pit EXPLANATION Tri —Tr7 Desert Archaeology backhoe trenches San Agustin Project (1986) Desert Archaeology backhoe trench A—Mountain Testing Project (1995) 112.04 Rio Nuevo South Testing Project (1995) trench datum ALLOSTRAT1GRAPHIC UNITS Early Agricultural Period UNIT 1 floodplain deposits UNIT 2 Agua Caliente Phase and later clenega deposits UNIT 3 Rincon Phase and later channel deposits of the Santa Cruz River UNIT 4 deposits modified or created by historic or modern processes UNIT 5 prehistoric or historic channel margin deposits of the Santa Cruz River RESULTS OF FALL 1995 ARCHAEOLOGICAL TESTING ALLOSTRATIGRAPHIC UNITS OF SURFICIAL DEPOSITS ON THE Qt2 TERRACE NEAR A-MOUNTAIN 0 0 50 250 100 m 500 ft Digital cartography by GEO-MAP . Inc. 1997 Figure 4.2 Location of allostratigraphic units identified at the Clearwater site and relevant trench locations referred to in text. 129 W -Surface (.76 m1 .09 m • "'Datum 112.04 3r0 • • -.22 m .30 m -'37m =1). J. .... • • • . • • • O ui • • • . . . . • -.30 m • . • -.97m -1.33m • • • . • • . • . • • • . . . . . • • • • Unit II: Dark grayish brown, silty clay loam. Strongly effervescent. Numerous fine carbonate filaments distributed throughout this unit. Fine, subangular blocky peds. Upper contact is abrupt and flat. Erosional disconformity is present between units II and III. Unit III: Brown, loamy sand (Ma), grading upward to a sandy silt (111b). Structure is medium to fine, subangular blocky (IIIb) to massive in the lower portions of the unit (IIIa). Effervescence is strong in the upper portions of the unit (IIIb), moderate in the center of the unit, and non-effervescent in the lower portion of the unit (1Ila). Many fine carbonate filaments are present in the upper portion of the unit. The upper contact is abrupt to clear and flat. Erosional disconformity is present between units II and III. Soil development is moderate at this contact. Unit IV: Brown, massive fine sand (IVa) grading upward to a dark grayish brown to very dark grayish • brown, silty clay loam (IVb). Some horizontal bedding still present in unit IVa. Very fine, subangular blocky structure in upper portion of unit (IVb). Effervescence is strong (IVb) to moderate (IVa). Moderate soil development represented by the presence of some fine carbonate filaments. Upper contact is clear, flat, and - Bottom of conformable. trench (-233m) Unit V: Dark grayish brown to brown massive silt (Va) to grayish brown loam (Vb) with very fine, subangular, blocky structure. Effervescence is strong to moderate. Some fine carbonate filaments are present throughout the unit. Upper contact is clear, flat, and conformable. - ..• Unit I: Brown, coarse sand grading upward to fine, sandy clay loam (la), fining upward to a brown, silty clay loam (lb). Diffuse pedogenic boundary between la and lb. Effervescence is strong (lb) to moderate (la). Many fine carbonate filaments are found in upper portion of the unit (lb). Some evidence of pedogenic reworking (clods of sediment from unit II distributed in upper unit I) between units II and I. Incipient soil formation in the upper portion of unit I is somewhat masked by soil development from unit II; the upper contact represents a compound soil. Upper boundary is flat, clear, and appears conformable. Feature 365 appears to be excavated from the top of this unit. 1.37 m Figure 4.3. Stratigraphic column, south wall of Trench 112. Unit VI: Grayish brown, massive sandy loam. Moderately effervescent. Upper surface is disturbed by modern activities. 130 allostratigraphic Unit 1 -0 ô 7 0 3 E. 3 1.) . e c 7 o ... -:: 0 cal .... .0 P .. .99o CI. ,i.) ._ -0 0 -0 a 00 co 0 ,... a.) C ,. E . a.) 3 c* ea . . c " oo _, ••'. . 8 0 -0 0 t) .0 C.> -4.=; - c".5 = .... 0 ... .... 9 .tri 0 __c - . . cl t... y ,... (.1 ›, .a ta° 0.. C 'V Cu eu , '0 el .0 0 0 70 a P 6,... .M mi c. -0 o a) cet 0 74 E •E' E = 0, 0 2 ,..) - = . -;,E2.9 ..0 0 , • 7., to c 0 a c ,,, a) 0 c 17 3 10 tr, =E .c a 6- >6 0 -. C. a.) t.... n5.0 • c .y. = .... c Y .,,g c.) 9 c 6. 4) 4 7,"'t 0) -6 .... cu 60 c.) 0. 0 0 0. U 17 8 ..Sa 0 >6 c C ggc.) -Y a 2 ...) a.) -C2 '0 7" .0 t) _ P. E >, oo `-' g = ..+4 . 8 >, a z -Ng .0 c ea n-• .-3' cn 0 . C *-e . . es >6 .0 u . C. f10 E 0 cl.) 0 40 ,.., - •0 0. -Na . i..) a) . = .c.) a ) 2 cp IDA os c - 2= -E ri a. 10 0 = E wo -0 .... -7 C CL) c 0 a .. ... c c a) 0, ‘„ c..., .0 .. 0 . . ... .0 a u ... le . )... _c tt; xu =E = 0. n.. ir. y ms 0 ,.. Lu. 2. c3 .c. -a E c . ,, t..1 c _ c ›... a a )6 . 2 73 .. d 46 0 ,.) 2 0 I) c) c c) c.) 2 a . ta .0 U' 6... ,..... =2 1 • 7 c.o.) E .c . 7 0 _ 7 0 = g 3 u, Q 7,1 0 o-.... c .E. j. .0 c•-• = , E 0 c •0 a• . . -0 "V .- 3 a . = 0.. tu) „, E '7r. a -0 = E = a ,z. to ›. .,= Z LE. ,-.. 1-. ea u 7.) c . 73 74 cn c ,:::, 1.) c.) •o-... c. a .) 4... el a 0 0. E 0 c. .... a.) . gg C P. Ir .0 7 0 .V. 5 40 ai "0 .0 --. = 0 t.... A. .0 U 0 Ir2 O.) 01) I.. C'' ...;.... = .0 = ....=,-.. E c.) 7, •"a' o c ‘... c 4_,.) .y 1:J 2..)... a . = . .o = ooo a )... tu .=.:7 ••• C 4.) "0 .0 Q a „ 75 E oo E -Y 0 6. 1 .,_:• g 7, ,..- . 0 „, 7, .0 c* To. ..., 0 6. 15 t,0 C ° ..0'tc : f. , 6. ..2 >6 5 .0 CZ P.3s >... 6. 0 .c) a) > -0 0 .5 9 ,„: 12 'R C 6. co 0 c 6 4e E ,•••• u c = 'LL-. t) 2 ,-, V' ° >7. <3 ,-. y 0C! 7. > 1 0>. ......) I I ). : 1.4 ,-)a >6 ••-• c.) . .0 . ,.... va 0 ca C M .3 8 0 7„; 4= _C C :-. ‘1 = U '.'" .0 '" VV_ 3 >6 0 = ••-• tt) >, e4 7 C ..; a) 0 g 2, 1.--: > 3 7 --- 1c Ou 0 15. ›. . >, 3 >., ,o 's ,,,,, '"" .0 ,.. 2E,-1 • ‘ - 02 . >6 9 > "1 ›.., D., -c,t> ' " -0 CI CI -0 . = .1.) m 0 C '0 2 _4 -0 c 0.0 0 ..0 C 0 >6 ... c„,-• .1 ca. - t.---C• 7 .14 ,3 > •-• ca 7 = ›. 2 o . ..c o c 0 0 >, C -0 ca >, ,?. 3 ,.; E .= ..0. 3 ... to a .0 CI C 'F'g e .0 C 2 u "C) cgt g 2 2 ..Ne .. 2 E oo. .-. .0 > .. 3 .e .02a.E to _ ,... ,... "' ea ta. 0 ,,‘‘, ..., ...-,.. TA ,_, .c 0 -0 .0 E'. 0 .0 EL. c .. 3 0 .5;15 a c c _ _• ,,, 6-73 1 = 0e..-F„ . . ., ,6 0 > >.„-. o )_„--•)... -Nt j. >6 0 0 a 6. --6 0 la ,,, , c,.., .I.) . • 2 E .t1 a.) .0 ti.) >6 13 . •co , 0 . 0 I-. ,... .... ct) cl.) ..0 0 1>'-. 1'5 12 t‘:70 '2 C) tO .t7 P. " 6, 6, -'1! "?..g t.' -i5. " c., •-ii 74 -' ,.., c a.) as a) co ea es O. 1:4 E-. 2C En.a.C12acZi>.M>M00 es .0 e 3Ndei6p1IS011e a-ci" JI - 131 are formed within these dark clay bands. Pithouses that date to the Early Agricultural period are found within the upper three paleosols (Figure 4.3, units IIIb, IVb, and Vb). Other cultural materials found within the upper portion of this unit include a Cienega phase (800 B.0 - A.D. 150) projectile point and Agua Caliente (A.D. 150 - 550) phase sherds. A radiocarbon date on charcoal from the Early Agricultural period occupation of 2520 ± 40 B.P. (Beta 85405) fits with the expected date for the Cienega phase projectile point and provides a minimum age for the upper portion of Unit 1. An additional 13 radiocarbon age estimates were run on charcoal from archaeological features within this deposit (Beta 90225-90232, 90230 excluded, and Beta 92617-92622). These dates produced an average age estimate for the prehistoric occupation of Unit 1 of 2471 ± 16 B.P. (see Table 4.2). Historic features related to the Tucson Pressed Brick Company overlie the Early Agricultural period pit features. No radiocarbon estimates have been run on charcoal recovered from Feature 365, located in the lowermost paleosol identified on the property. A significant hiatus marked by a disconformity follows deposition of the sediment on which the Early Agricultural paleosol is formed and deposition of Unit 2 to the west and north. It is possible that the Early Agricultural period occupation of Unit 1 (and paleosol formation) is concurrent with the initial deposition of cienega sediments within the area marked Unit 2. Unit 1 is truncated on the east side by an erosional unconformity that may be related to channel incision following the Early Agricultural occupation at sites downstream from the project area (Freeman 1997; 132 Huckell 1996a). Erosion of this paleochannel is followed by deposition of Unit 3 (Figure 4.4, Units II and III). Allostratigraphic Unit 2: Early Agricultural to Historic Period Cienega Unit 2 is a series of younger alluvial deposits located across the majority of the parcel. The lowest-recognized unit of this deposit is a medium to coarse sand and fine gravel. This sand and gravel is overlain by a very thick (in places, exceeding 1.5 m depth), dark brown, silty clay deposit, which has been described previously as a cienega (Ahlstrom et al. 1994; Katzer 1987). This deposit is generally 1.5 to 2 m thick, and sedimentation throughout the deposit is continuous and uninterrupted. Katzer (1987) examined seven trenches on the Rio Nuevo South property (Figure 4.2, tri through tr7). He identified this unit as extending south to the portion of the Clearwater site excavated by Elson and Doelle (1987) near Brickyard Lane. His profile (Figure 4.5) demonstrates the variability in the thickness of this deposit over the length of those seven trenches and 11 additional trenches along Brickyard Lane. Both the cienega and underlying sand and gravel are considered part of Unit 2. The upper surface of Unit 2 has been modified significantly by historic plowing. However, the structure of the soil formed within, the thickness of the deposit, and indications of significant groundwater seepage in the form of upwelling features and oxidized mineral deposits at the boundary between the dark brown silty clay and the underlying sand and gravel suggest that this deposit would support the 133 growth of wetland grasses, sedges, and other plants (cf. Hendrickson and Minckley 1984). The presence of a cienega in this area is known from historic accounts (Betancourt and Turner 1990), but most accounts place the cienega south of Sentinel Peak. It is possible that groundwater availability and protection of the area from erosion, or the presence of additional spring seeps, allowed a similar feature to extend north of Sentinel Peak. The bedrock outcrop formed by Sentinel Peak may have even promoted sediment accumulation downstream of this source area that would not be eroded by natural alluvial processes. North South SECTION 1 SECTION 2 Congress Street Mission Lone ersiTrench 2 Met Trench 7 4 0 --,Y V P - ---,........4 jli„ p e i t,_iv . Trench Trench/ 10 18 Disturbed Cress-Section of Cien ego f 4 ,11 li Medium-to-Coarse Sand and Fine Gravel 2 Figure 4.5 100 260 3i:/0 460 560 1 600 70:1)0 Meters Profile of trenched area showing a north-to-south cross-section of the alluvial soils (from Elson and Doelle 1987). Both Cienega and underlying sand and gravel are considered part of allostratigraphic unit 2. Archaeological materials located Unit 2 postdate the Early Agricultural period pithouse occupation, with the exception of a single Cortaro point from a disturbed context. Canada del Oro phase ceramics and some Agua Caliente phase ceramics are found in features within the cienega-like sediments in the upper portion of Unit 2. These archaeological materials suggest a minimum date for Unit 2 sediments of at 134 least A.D. 750 to 850 and possibly earlier (A.D. 150 - 550, based on dates for the Agua Caliente phase). Historic Papago plainware ceramics are also found in the upper portion of Unit 2. A canal (Canal B, not shown on figures) crosses Unit 2 and Unit 1. On the basis of geometry and placement, the canal is probably prehistoric, and likely dates to Hohokam Colonial or later periods. Allostratigraphic Unit 3: Pre-Rincon Period Alluvium Unit 3 is a series of young alluvial deposits representing a prehistoric channel of the Santa Cruz River postdating Unit 1, and which also appears to postdate Unit 2. The relationship between this deposit and older sedimentary units has been obscured by historic and modern modification of the ground surface that prevents examination of the contact between these units. An erosional contact between Unit 1 and what is probably Unit 3 was recognized in trench 12 (Figure 4.4). The cross-section for trench 12 is marked on Figure 4.1 as B-B'. The fill of that erosional contact is marked here by a package of silts and clays (Units II and III) representing a sequence of overbank and floodplain deposits. Clay deposits within this package of sediments are thickly developed and may form a facies relationship with Unit 2. Occasional fire-cracked rocks are found within the upper portion of this deposit, probably representing the sedimentary reworking of prehistoric Early Agricultural deposits. The portion of Unit 3 identified in the stratigraphic pit Figure 4.6 represents bank margin 134 least A.D. 750 to 850 and possibly earlier (A.D. 150 - 550, based on dates for the Agua Caliente phase). Historic Papago plainware ceramics are also found in the upper portion of Unit 2. A canal (Canal B, not shown on figures) crosses Unit 2 and Unit 1. On the basis of geometry and placement, the canal is probably prehistoric, and likely dates to Hohokam Colonial or later periods. Allostratigraphic Unit 3: Pre-Rincon Period Alluvium Unit 3 is a series of young alluvial deposits representing a prehistoric channel of the Santa Cruz River postdating Unit 1, and which also appears to postdate Unit 2. The relationship between this deposit and older sedimentary units has been obscured by historic and modern modification of the ground surface that prevents examination of the contact between these units. An erosional contact between Unit 1 and what is probably Unit 3 was recognized in trench 12 (Figure 4.4). The cross-section for trench 12 is marked on Figure 4.1 as B-B'. The fill of that erosional contact is marked here by a package of silts and clays (Units II and III) representing a sequence of overbank and floodplain deposits. Clay deposits within this package of sediments are thickly developed and may form a facies relationship with Unit 2. Occasional fire-cracked rocks are found within the upper portion of this deposit, probably representing the sedimentary reworking of prehistoric Early Agricultural deposits. The portion of Unit 3 identified in the stratigraphic pit Figure 4.6 represents bank margin Page 135 view looking S 2 3 -33 E . canal • F179 canal ..f., •4•"y •-n Vb ;14";" ° ei4t.,. ' UPPER TIER 111a 4 Illaf Ma bottom of 2 stratigraphic pit Ma i 11b 2 LOWER TIER Ila 2 Ila i la SS-359 SS-358 SEDIMENTARY UNITS PIMA COUNTY — TUCSON, ARIZONA AZ BB: 13: 6 ASM pebble to cobble grovel medium sand silt, channel fades slump silty cloy, prehistoric floodplain deposits coarse sand to cobble gravel very fine sand separated by cloy band Mof modern trash deposits muddy sand (with groundwater —derived? carbonates) fine pale brown sand, overbonk flood ladies of lia medium sand silt overbonk facies of lia pole brown coarse to medium sand coarse sand and pea gravel overbonk flood fades related to either Unit lia f2 Unit or 11b 1 dark brown silty clay with soil formation at the upper boundary reddish brown coarse sand and cobble gravel laminated coarse to medium sand, magnetite placer deposits, low ongle trough to horizontal cross—bedding pale brown sand and grovel pale yellowish—brown coarse to medium sand muddy silt and reddish brown medium sand wet bank margin fades of Unit Ilb i pole reddish—brown sand and grovel dark brown cloy coarse sand and pea to small cobble grovel (2-6 cm) dark brown compact silty clay dark reddish—brown coarse to medium sand pole brown muddy silt, wet bank margin facies of Unit 116 2 Unit 1162 1 and Unit Ilb i f grade to fine sand and coalesce to west pedo genesis STRATIGRAPHIC PIT SOUTH WALL CROSS-SECTION cloy bonds fine sand medium sand Digital cartography by CEO—MAP, Inc. 1997 coarse sand silt overbonk facies of Unit 1c3 fine pole brown muddy sand that grades westward from a coarse to medium sand series of dark brown cloy and silt bands In overbank and slockwa ter deposits with incipient soil formation at upper boundary brown muddy sand grading westward from coarse to medium sand which coalesces with Unit Ilc coarse sand and pea grovel ;;:"4..•ti. sand, coarse sand, gravel, and cobbles cobble grovel Profile of the Clearwater site stratigraphic pit numbers in bold represent elevation in meters above sea level (MASL) vertical exoggeralion — 2X HORIZONTAL SCALE: in meters along profile face petrographic sample Figure 4.6 VERTICAL SCALES: 136 and channel fill deposits that are either contemporaneous with or postdate the sediments identified in trench 12. The 4.5-m stratigraphic pit (Fig. 4.6) was excavated to clarify the relationship between a historic canal (Canal D, Feature 180) and the alluvial deposits underlying the canal. An unexpected bonus was the ability to identify channel margin deposits adjacent to the alluvial fill into which the canal [now part of the A-Mountain canal system, AZ BB: 13:481 (ASM)] was excavated. The south profile of the stratigraphic pit is shown in Figure 4.6. A simplified profile demonstrating the inferred relationship between deposits in trench 12 and in the stratigraphic pit is shown in Figure 4.7. The location of the cross sections represented by letters A to A' and B to B' is shown on Figure 4.1. Sandy alluvial deposits (Unit II, allostratigraphic unit 3) underlying the canal (Feature 180) contained ceramics and shell dating to the Early and Middle Rincon subphases. These ceramics suggest a minimum age for the deposit of A.D. 950 to 1100. Cultural materials found in the upper portion of this deposit postdate the prehistoric occupation of the area. Most material on the surface overlying the channel deposit dates from the Protohistoric to modern periods. Another unit of alluvium (Unit III, allostratigraphic Unit 5), consisting of bank slump and channel fill deposits of a later channel, is also found underlying the canal and adjacent to Unit 3. These deposits, which are described in greater detail below, may represent historic incision of the river or may predate that incision. C 1... 9 R R 0 R 0 .7: . R. L 1Rin oludag34•490119 138 Allostratigraplzic Unit 4: Historic and Modern Trash and Land Modification Unit 4 is a series of historic and modern trash deposits representing either the fill of mined areas or the accumulation of historic and modern trash on the property. Modification of Units 1, 2, 3 and 5 by historic and modern processes, such as mining, dumping, and plowing, is included within Unit 4. Because plowing covers the entire property, Unit 4 is mapped only where it exceeds approximately 37 cm, the depth of most plowzones. Allostratigraphic Unit 5: Protohistoric or Historic Channel Alluvium. Unit 5 consists of a package of sediments exhibiting soft sediment deformation and low angle trough to horizontal cross-bedding structures. These sediments are typical of bank slump (channel margin) and channel fill deposits and represent an incision of the Santa Cruz River postdating deposition of Unit 3. The deposit predates construction of the historic canal (Canal D, Feature 179). A former channel is visible in some aerial photos and may be related to Unit 3 or Unit 5. Correlation By using age estimates on archaeological materials found within each of these units and radiocarbon dates, where available, and by examining the general lithology and stratigraphy of the units within the project area, the following correlations can be made with Haynes and Huckell's (1986) units. The ages of archaeological material 139 found within allostratigraphic Unit 1 correlate with the ages of the upper Unit B2 of Haynes and Huckell. The geomorphic, lithologic, and archaeological context of these deposits is consistent with their data. Similar preceramic archaeological materials have been found in these deposits upstream. The ages of cultural materials found within Unit 2 correspond with Haynes and Huckell's (1986) Units C 1 and C 2 . Although cultural materials from a variety of periods, including the historic period, are found within this deposit, deposition of cienega sediments probably occurred between 2,500 and 1,000 years ago. A similar cienega soil is found in the Olberg-Schanck reach of the Santa Cruz River (upstream of the project area) during this period. As suggested by Ahlstrom et al. (1994), standing water need not have been perennially present for such deposits to form. It is possible that the thick, dark brown clay to silty clay deposits are the result of shallow groundwater and perhaps small spring-fed pools surrounded by wetland grasses and other hydrophytic vegetation. The presence of Early and Middle Rincon ceramics in portions of Unit 3 suggests that this channel was active either during or after the period from A.D. 950 to 1100. Rincon phase ceramics were found in dunes overlying Unit C 2 west of Brickyard Arroyo. Haynes and Huckell (1986) suggest that dune formation may be related to Unit C3 channel incision. Slopewash deposits of Unit C3 are found upstream at the Drexel Road site and along parts of Airport Wash. These deposits contain Tanque Verde phase (A.D. 1150 - 1300) and later archaeological materials. It is 140 possible that Unit 5 is related to channel activity that was responsible for the deposition of these slopewash deposits upstream. Unit 4 can be correlated with Unit D and is essentially historic or modern in age. Summary The succession of sedimentary cutting and filling events in the A-Mountain project area generally fits that found by Haynes and Huckell (1986) elsewhere in the Santa Cruz Valley; however, further examination of the relationship between deposits is necessary to confirm whether or not these sedimentary events are concurrent across the entire valley. Cienega deposits in the A-Mountain project area are potentially promoted and preserved by the Sentinel Peak volcanic intrusion and by the presence of shallow, spring-fed groundwater sources. Some of the units identified in the project area appear to have formed contemporaneously with other geomorphic units, making the boundaries between time-geomorphic units less obvious. These relationships have been further obscured by historic modification of the ground surface. The prehistoric occupation found at the eastern end of the A-Mountain drainage system is remarkably well preserved and may be due to a number of geomorphic factors. Overbank or slackwater sedimentation seems to have accumulated on the downstream side of A-Mountain, allowing preservation of archaeological deposits to occur. The Early Agricultural period settlement was restricted to a narrow band of 141 preserved floodplain, which formed a small rise surrounding what would become or what already was a prehistoric cienega. Periodic inundation by floodwaters of the Santa Cruz River would have made the cienega an ideal place to exploit wetland vegetation or to cultivate crops. Both the cienega and the floodplain have undergone significant modification from prehistoric and historic channel changes of the Santa Cruz River and from historic and modern human activities, leaving only a tiny fragment of the prehistoric land surface and the archaeological resources preserved on that surface. ALAMEDA STREET During summer of 1995, Desert Archaeology, Inc. excavated a series of trenches along the entire length of Alameda Street west of the Santa Cruz River (Figure 4.1). The trenches extended from the Santa Cruz Riverpark to Tumamoc Hill. Exposed in the trenches were prehistoric and historic artifacts and features, including canals or ace quias which may have been part of an extensive system at the base of A-Mountain and recorded in historic documents. According to the Arizona Geological Survey (McKittrick 1988), the portion of the project area east of Grande is Qt3 and the portion west of Grande is Qt2; however, data recovery efforts found Holocene sediments across the entire project area. Because the trenches were not deeply cut, it is difficult to determine whether these Holocene sediments are identical to those which Haynes and Huckell (1986) describe 142 as forming a compound terrace with the Jaynes terrace (Qt3) or whether the entire project area represents the alluvium that comprises the Holocene terrace (Qt2). Stratigraphy The Menlo Park Storm Drain testing project provides a unique opportunity to glance at a very long profile of surface (< 2 m deep) sediments of the Santa Cruz River Valley. The following section provides a general description of sediments found along Alameda Street and generalized in Figure 4.8 and represented by the cross-section indicated by C to C on Fig. 4.1. The entire section of sediments exposed during testing consists of alluvial deposits of the Santa Cruz River, with the exception of minor evidence for colluvial deposition of clasts coming off the Sentinel Pealc/Tumamoc Hill area in the westernmost portion of the project area. The alluvial deposits recognized in trenches are comprised of a single large deposit of cienega-like clays underlain and cut into by sandier deposits representing channels (both large and small) of the Santa Cruz River. In places, a silty overbank deposit is preserved on top of this clay band. Unit] Unit 1 consists of a package of sediments that include finer-grained silts and clays and coarser-grained channel features. Lenses of former channels are found within the deposit in trenches 8, 9, 10, 11 and 12. This unit may be part of the 143 C.) — 144 Holocene sediments covering the valley floor in this location or may be Pleistoceneaged sediments which form the Jaynes terrace. These sediments are covered by younger, Holocene-aged sediments, suggesting that if the Pleistocene Jaynes terrace (Qt3) is present in this area, it forms a compound terrace with the younger Qt2. Unit 2 Unit 2 represents a series of laminated silts and sands in trenches 2 and 3. These may be equivalent in age to sediments defined as unit 1 or are younger. Due to the thickness of overlying units in trenches to the west, the extent of this unit could not be determined. The bottom portion of the channel contained two thin laminae of abundant charcoal. Charcoal from the bottom most of these laminae, yielded an age estimate of 3650 ± 60 B.P. (Beta 85537). A possible pit structure is located in the upper portion of this deposit and found in trench 1. Unit 3 Unit 3 is comprised of very dark brown cienega-like clays that are found in the bottoms of the trenches west of trench 3. These clay loam to silty clay loam deposits are usually marked by abundant carbonates and angular to subangular blocky structure. Most of the prehistoric archaeological features found along Alameda St. are dug from the top of this deposit. If the age of features elsewhere in the area can be used as a guide, then these prehistoric features probably date to the earlier part of the Hohokam 145 period. This deposit appears to be a Holocene floodplain sediment overlying deposits of the Jaynes terrace formation. Unit 4 Unit 4 represents a silty overbank deposit that overlies the dark brown cienega clays of unit 3. The unit becomes thicker to the west. Toward the west end of the project this unit is very thick and interfingers with sheetwash/debris flows off the Sentinel Peak-Tumamoc Hill complex. Unit 5 Unit 5 represents a former channel of the Santa Cruz River, identified in trench 2. Three possible canals are found within this channel. The channel cuts through units 2, 3, and 4. This channel may be a part of a ditch excavated by Sam Hughes in 1888 that began headcutting the following year (Betancourt and Turner 1990). Unit 6 In few places does an intact "plowzone" (unit 6) appear, indicating that post-agricultural processes have altered or destroyed evidence for agriculture. In other words, plowzones were removed or altered to the effect that they are no longer visible in the soil profile. Since this is unusual in natural floodplain settings, we can assume that the lack of a plowzone (known to exist from historic records) is due to 146 post-agricultural destructive processes inhibiting its preservation. The most obvious process, other than removal or redistribution of the plowzone sediments is compaction by modern travel and the construction of roadways used for such travel. In places, it is obvious that the sediments which may have in part formed the plowzone, are highly compacted. Where sherds are present, the sediments have been pressed to conform around the shape of the sherd body, suggesting that these sediments became compact during wetter conditions. It is probable, that when the dirt road that became Alameda Street was in use, it occasionally became muddy and forced compaction of these sediments. Further compaction may have occurred when the road was actually paved. Nevertheless, this upper sediment was deposited sometime after historic use of the ace quias and before the current construction of Alameda Street. RILLITO FAN The project area is located at the confluence of the Santa Cruz River and Rillito Creek on a finger of land comprising the former floodplains of these streams (Figure 4.9). The following discussion of the geomorphology will refer to the two portions of the project area separately. The portion of the project area that encompasses the Sunset Mesa Ruin (AZ AA:12:10) and the old Sunset Dairy is referred to as the northern portion of the project area. The portion of the project south of Sunset Road that encompasses the Rillito Fan site (AZ AA:12:788) is referred to as the southern portion of the project area. 147 Figure 4.9. Surficial geology of the Rillito Fan project area (after McKittrick 1988) showing location of the Rillito Fan site and trenches described in text. UTM coordinates place the map in its position on the USGS 7.5' quadrangle for Jaynes, Arizona. Rillito Fan Site (AZ AA:12:788) Although the confluence of the Santa Cruz River and Rillito Creek is better known for the Rincon-age Sunset Mesa Ruin (AZ AA:12:10), a recent Late Archaic 148 component has been discovered by Desert Archaeology, Inc. (Tucson, AZ) south of the Sunset Mesa Ruin on the site known as Rillito Fan (AA:12:788; Wallace 1996). The site was surface collected and tested during Fall 1995. Only 853 m of trenching was placed over a 160 acre area. Twenty-three cultural features, including pit structures dating to both the preceramic and ceramic periods, clusters of fire-cracked rock, a homo (pit oven), and various small pits. Most of the features are thought to date to the Early Agricultural or Early Ceramic period. Although most of the Early Agricultural component is shallowly buried under the Holocene terrace in what appears to be sediments from Rillito Creek, the finer-grained materials underlying these sediments may be derived from the Santa Cruz. A date of 2860 ± 40 B.P. was derived from charcoal in a pit eroding from the western portion of the site in the lowest identified cultural layer. Methods Prior to fieldwork, aerial photographs, topographic maps, and surficial geologic maps were consulted to determine which areas would be most likely to contain cultural remains. Test trenches located in these areas were examined for geologic study. Sedimentary characteristics and stratigraphic relationships were recorded in order to verify the origin of the deposits and the cultural materials within those deposits. Soils were described using USDA soil classification criteria, further elucidating the origin and transformation of sediments within the project area. 149 Because the project area was potentially influenced by two different drainages, sediment samples were examined from key localities within both parts of the project area and compared with reference samples to determine from which drainage sediments were most likely to originate. The southern wall of the Rillito arroyo, which was cut during the 1983 flood caused by tropical storm Octave and which preserves an excellent exposure of the deposits, was examined for additional pertinent information. Because of the low probability of recovering intact archaeological or historic resources within T-1 and because previous projects within the Santa Cruz River valley have demonstrated a high likelihood of recovering archaeological resources from T-2, test excavations for this project were focused on T-2. Results of Geomorphic Study Representative stratigraphie descriptions from the northern and southern portions of the project are provided in Tables 4.3 and 4.4. These are referred to in the discussion below where stratigraphic units are described. Generalized Stratigraphy Both portions of the project area are comprised of a loose, coarse-grained component (unit III) underlain at depth by fine-grained sands, silts and clays typical of overbank, slackwater, and cienega deposits (units I and II). Detailed analysis of these 150 Table 4.3. Stratigraphic desccriptions from the south wall of Trench 10, AZ AA:12:788. Depth (m below Unit datum) 9.73 IV 9.73-10.11 Description Ground surface. Plowzone; lower boundary is abrupt and undulating. Occasionally as great as 50 cm. in depth. Compact, light grayish brown slightly sandy silt loam. Abundant very fine rootlets present, some bioturbation, and rare fragments of fire-cracked rock. nib 10.11-10.97 Pale brown silt with some very fine carbonates, grades to a heavy silt loam to silty clay loam at lower boundary; massive. A thin lens (<3 cm thick) of silty clay forms an abrupt lower boundary with the underlying deposit. Extensive bioturbation and moderate quantities of rootlets are present throughout the deposit. Illfac 10.61-10.75 Moderately compact channel deposit which forms a facies within Unit The lens is comprised of a fine sand in the upper portion of the deposit and grades vertically to a fine gravel at the base of the deposit. The sand and gravel are pale brown in color, and are comprised primarily of granitic particles and abundant flecks of muscovite. The deposit extends at least 4.3 m horizontally. lila 10.97-11.38 Channel sand. The upper portion of the deposit is comrpised of a lightly compact pale brown fine loamy sand and grades to a very pale brown coarse sand at the base of the deposit. The sand is granitic and contains abundant muscovite flecks. Abundant rootlets and fine flecks of charcoal are present throughout the loose sand at the base of this deposit. lIc 11.38-11.59 Pale brown silt loam to silty clay loam; massive to granular structure. Rare fine to medium flecks of charcoal are present throughout. IIb 11.59-11.86 Dark grayish brown, very compact clay loam; medium subangular blocky structure. Abundant fine to medium-sized flecks of charcoal are present throughout. Carbonates are present throughout as coarse sand to very fine gravel-sized nodules. Cultural material mixed in includes charcoal, firecracked rock, and flaked stone. The concentration of this material here was labelled Feature 17 and dated 2860 ± 40 B.P. (Beta 90318). lIa 11.86-12.37 Very compact slightly sandy pale brown silt with a few fine carbonates. 12.37-12.56+ Dark grayish brown clay loam, very compact; medium subangular blocky structure. Abundant charcoal fine to medium-sized flecks of charcoal are present throughout the deposit. Carbonates are present throughout in coarse sand to very fine gravel-sized nodules. No artifacts were observed; however, only a very small exposure examined. This unit extends beyond 12.56 m below datum but it is unknown to what depth in this locality. At other sites, similar bands rarely exceed 50 cm in thickness. 151 Table 4.4. Stratigraphic descriptions from the north wall of Trench 14, AZ AA:12:10. Depth Deposit (m below datum) Description 14.26 Ground surface. a 14.26-15.40 Zone of surface disturbance; compact pale brown silty sand. No carbonates visible, abundant fine and coarse rootlets. Muscovite mica flecks are abundant. Extensive bioturbation. 15.40-15.54 Loose, pale brown silty sand. No carbonates, same composition as uppermost stratum. Some fine gravel. Muscovite mica flecks are abundant. Extensive bioturbation, abundant fine rootlets. 15.54-15.69 Pale brown silty clay loam; laminated. d-g, m 15.69-16.01 Pale to dark brown clay loam, laminated with lenses of pale brown silty loam. Laminated sedimentary structure still present in places but varies considerably both horizontally and vertically. Moderate amount of fine carbonates and rare fine to medium flecks of charcoal are present with the basal five cm bearing the highest frequency. h 16.01-16.36 Moderately compact, pale brown silt loam. No carbonates, bioturbated. Rare fine to medium flecks of charcoal. Grades vertically into clay loam below. 16.36-16.61 Massive brown silty clay loam with few, very fine carbonates and rare medium flecks of charcoal. Fine to medium subangular blocky structure. 16.61-16.74 Pale brown compact fine loamy sand lens. No carbonates or charcoal present. 16.74-16.92 Dark grayish brown, very compact clay loam; medium subangular blocky structure. Abundant fine to medium-sized flecks of charcoal are present throughout though their density varies considerably along the length of Trench 14. Carbonates are present throughout in coarse sand to very fine gravel-sized nodules. May be equivalent to Unit 2B1 south of Sunset Road. 1 16.92-17.24 Grayish brown (top) grading to pale brown (bottom) compact heavy silt loam. Some medium-sized mottles of sediment from stratum above are present in the upper portion of this unit. Small flecks of charcoal and carbonates in the form of coarse sand to very fine gravel-sized nodules are present throughout. o 17.24-17.43 Pale brown compact coarse sandy loam. Rare small flecks of charcoal are present throughout. At the base of the unit there is a concentration of dark gray heavy minerals, including magnetite. 17.43-17.61+ Very compact pale brown sandy silt. Extends to an unknown depth. 152 two components is difficult, given the very limited proportion of ground exposed during testing; however, some general interpretations can be proposed. The upper component is composed of massive, poorly sorted sand and gravel units that are typical of those found on alluvial fans. These massive units are occasionally punctuated by well- sorted sand and/or silt lenses typical of overbank or slackwater deposition and other less well- sorted sand channels. The strata within this upper component cannot be traced over long horizontal distances and are typical of a braided stream regimen characterized by a network of channels that join and separate and that are separated by islands or bars (Leopold et al. 1964). While the individual channels comprising the network may be narrow, the network itself is normally wide and shallow (Leopold and Wolman 1957). In this system, large flow events such as floods would have the opportunity to spread out widely across the valley floor. The presence of a coarse-grained upper component suggests that the braided channel morphology of Rillito Creek known from historic record also may have existed during the prehistoric period. The lower component is comprised of at least three dark brown clay bands, which exhibit considerable pedogenic development in the form of soil structure and abrupt to very abrupt boundaries occasionally marked by a darker, organic-rich horizon in the upper 5 cm. Although these clay bands are similar to cienega deposits, they generally do not exhibit the darker color and other characteristics of soils that would support the growth of sedges and other riparian plants. These clay bands, and the silts 153 and fine to very fine sandy silts that separate them, are typical of overbank and slowmoving or slackwater deposits that are characteristic of an aggrading stream. Elsewhere along the Santa Cruz River, similar deposits have been discovered. The abrupt to very abrupt upper contacts of each clay band are often indicative of erosional or channel-cutting episodes. The upper, organic-rich horizon within each clay band probably represents the formation of a grassy swale on sediments that have filled a former river channel. Provenance Apart from proximity to the current channel of the Santa Cruz River and general similarities among sediments comprising the lower component of the Rillito Fan site, including the presence of abundant fire-cracked rock and chipped stone within the uppermost of these clay bands (unit Ilb), there was no particular reason to conclude that either the upper or the lower strata could be correlated with sediments found at other sites along the Santa Cruz River. Sourcing of the sediments found within the project area was necessary to determine which stream(s) most influenced past sedimentation within the project area. Because the project area is situated at the confluence of the Santa Cruz River and Rillito Creek, sediments could potentially originate from either one or both of these streams. The project area is also near the confluence with Cafiada del Oro Wash, adding yet another source. In fact, examination of sands from the south bank of the Rillito (T-1) near the northern project 154 area revealed sands originating in the Canada del Oro system and probably representing slackwater deposition during the period when the Rillito met the Santa Cruz near the Catiada del Oro Wash (see Figure 4.10). —— 18 72 1880 MO 1912 1918 ME 1937 c — -- 1941 _— 1954 MIDI 1960 - 196 7 ME 1980 o 1 o Linsberlest Drive Figure 4.10. Historic channel locations along Rillito Creek (after Graf 1984). 155 Fine- and coarse-grained sand samples from the upper component (unit IV) were examined by petrographers Beth Miksa and Michael Wiley of Desert Archaeology. By using a set of reference samples taken from various localities within the Tucson Basin, they were able to determine that the upper strata in both the northern and southern portions of the project area are comprised of sands that originate within the Rillito Creek system. Provenance is more difficult to assess on fine-grained materials (i.e., clays, silts); however, some portion of coarser material is nearly always present in finegrained sediments. Occasionally, within the uppermost clay band (Unit lib), somewhat discrete features can be identified. Two such features (Features 17 and 24) were identified on the eastern and western edges of the Rillito Fan site (AZ AA:12:788). The coarse components recovered from flotation samples derived from these features were examined by petrographer Jim Heidke of Desert Archaeology. Examination of the coarse component recovered from those features exhibited abundant carbonate nodules and fragments of volcanic rocks. No metamorphic rock fragments were identified; therefore, the sediments are likely derived from the Santa Cruz River. Correlation and Dating A single radiocarbon age estimate on a feature (Table 4.3, Feature 17) within unit IIb yielded a date of 2860 ± 40 B.P. (Beta 90318). This date correlates well with the upper part of Haynes and Huckell's unit B 2 . Correlation between this site and the 156 Los Pozos site [AZ AA:12:91 (ASM)] suggests that deposition of this unit corresponds with the Early Agricultural horizon at the Los Pozos site. This deposit, at Los Pozos, is followed by a significant period of channel incision. If the correlations between the two sites are correct, changes in channel morphology and direction of the Santa Cruz, which may have resulted from this episode, may explain the incursion of Rillito Creek-derived deposits (unit IV) at the Rillito Fan site. Additional exploration of exposures along the Santa Cruz could potentially resolve any questions regarding the origin and correlation of these sediments. Correlation between deposits in the northern and southern portions of the project area is much more difficult to assess. The northern portion of the project area is highly influenced by its location at the confluence of two streams. In the deepest exposure (trench 14) on the northern portion of the project area, thin bands of sedimentation (Table 4.4, deposits b-g and m) in the vertical center of the exposure likely signal the transition between Santa Cruz and Rillito deposition, and they may even represent interfingering deposition from these two streams. In this location, transition between the two systems likely occurred over a longer period of time than the fairly abrupt sequence displayed in trenches on the southern portion of the project area. 157 Summary Prehistoric Geologic Setting The generalized stratigraphy suggests that sedimentation within the project area occurred in two phases. First, slow-moving aggradational deposits of the Santa Cruz River were emplaced within the project area. Toward the end of this sequence, the remains of a substantial prehistoric occupation were deposited within a clay band (unit lib). These people were probably living on or near a grassy swale floodplain of the Santa Cruz River. Either during or shortly after this occupation, the Santa Cruz channel changed course, eventually allowing Rillito Creek sediments to be emplaced within the project area. A substantial hiatus probably occurred before the project area was again inhabited. Clearly, a major change occurred in the physical environment of the project area. The incursion of Rillito Creek derived sediments into the project area represents a markedly different depositional system. Rillito Creek was a wide, shallow braided stream during the later prehistoric occupation. The site area itself was likely part of a wide floodplain that occasionally flooded during larger storm events. The river was in a state of dynamic equilibrium, meaning that aggradation and degradation (small scale) occurred in several reaches at the same time. Large base-level changes, which occur during periods of net degradation or channel incision, were clearly not the norm. The channel itself probably migrated widely in its course; however, the presence of prehistoric cultural material within the project area and evidence for incipient soil 158 formation suggest that the site area itself was part of a wide fan of material being deposited and eroded only in flood-scale events. Landscape Stability Although channel morphology can have some effect on the selection of a site for habitation or agricultural activities, a more important criterion is landscape stability. Braided streams can be easily erodible, but the presence of carbonates and organic matter typical of soil formation can indicate that stable surfaces existed in the prehistoric past. In both the northern and southern portions of the project area, loose, coarse-grained sands are often separated by finer-grained silts in the upper strata. These silts are weakly to moderately cemented with carbonates, forming an incipient soil and representing a surface on which people are more likely to conduct relatively short-term activities. Many of the features discovered in the southern portion of the project area appear to be cut into these compact silts. Though interrupted by periodic incisions, the landscape comprising the lower component is clearly a more stable environment. Soil development on each of the dark brown clay bands is clearly stronger and represents a longer period of landscape stability. Upstream of the project area, the later prehistoric (Late Archaic) occupation consists of large pithouse villages, which were probably occupied over a substantial period of time. Another important factor related to landscape stability is the preservation of cultural remains. The abandonment of the T-2 terrace provides an ideal setting for the 159 preservation of cultural materials. Since the landform is not being actively eroded by the river, few other natural processes are available to remove cultural material from its buried context in the floodplain. However, cultural factors such as the historic use of the land for farming can move or remove subsurface cultural materials. The surface of the southern portion of the project area is particularly flat, suggesting that some land leveling took place prior to plowing. This fact, combined with variable preservation of stratigraphic units across the field, suggests that small rises and dips that naturally form in a bar and swale topography were leveled to form a single surface, thus mixing stratigraphic units of different ages. This is particularly evident in the southwest portion of the project area. Finally, location within the stream system affects the amount of active erosion or deposition that may have taken place in the past. The Sunset Mesa site appears to have been geomorphically active for a longer period of time than the Rillito Fan site, which is perhaps why the occupations found on that portion of the terrace date to a later time period. INA ROAD Ina Road forms the northern boundary of the study area. Three sites are located along the portion of Ina Road from the railroad tracks just east of Interstate-10 to Silverbell Road. These archaeological sites, AZ AA:12:111, 130, 503, and 688 160 have been reported by several different individuals or companies and should probably be lumped under one or two archaeological sites. AZ AA:12:111/688 Site AZ AA: 12:111 was discovered during subsurface trenching for a sewer line, and was further defined by Bruce Huckell and C. Vance Haynes (1979, ASM site files). The site consisted of at three buried occupation horizons associated with chipped stone, fire-cracked rock, charcoal, and animal bone. The subsurface horizontal extent of the site covered about 800 sq m. Radiocarbon dating of these horizons placed the use of the area between approximately 2,800 and 4,300 B.P. (Huckell 1988:65). The western locus of AA: 12:111 is within the boundaries of AA: 12:688, which was identified during subsurface testing in 1988 (Bontrager and Stone 1989). Four deeply buried (8-10 ft) archaeological features were identified in over 150 linear ft of trenching. A radiocarbon date of 2775 ± 45 B.P. was obtained from carbonized wood found within a cultural feature (Bontrager and Stone 1989:Table 1). The site was recommended for National Register of Historic Places nomination. AZ AA:12:130 AZ AA:12:130 has also been reported by Haynes and Huckell (1986). Cultural remains are known to extend to 4.3 (14 feet) below the modern ground surface. 161 Radiocarbon dates of 3260 ± 100 B.P. (A-3145), 3730 ± 110 B.P. (A-3104), 3240 ± 50 B.P. (A-3147), and 3140 ± 90 (A-3146), were recovered from limited work at the site. A small portion of the site was tested by SWCA Environmental, Inc. (Tucson, AZ) in Fall 1996. A few possible features and isolated artifacts were found. A short examination by the author confirmed the sequence of sedimentary deposits previously reported by Haynes and Huckell (1986); no additional sedimentary profiles were drawn. AZ AA:12:503 AZ AA: 12:503 was originally recorded as a low density artifact scatter containing abundant shipped stone debitage and tools, fire-cracked rock, ground stone fragments, and occasional sherds covering a 13 acre parcel on the floodplain east of the Santa Cruz River and north of Ina Road. Diagnostic ceramics included one Canada del Oro red-on-brown, one Santa Cruz red-on-brown, and a few Rincon red-on-brown sherds (ASM site files). Surface disturbance, due to modern development, was heavy and the exact boundaries of the sites remain undefined. In 1987, AZ AA:12:503 was tested with 235 m of backhoe trenches by Desert Archaeology, Inc. (then the Tucson branch of the Institute for American Research; Doelle 1987). One cultural feature, unidentified as to type at that time, was identified. Given current information, that features may have been a burned pit structure dating prior to A.D. 1. Statistical Research, Inc. (Tucson, AZ) has recently completed 162 excavations in an area north of the current project area (Ezzo and Deaver 1996). These excavations identified a total of 208 cultural and natural features. One pit structure was exposed. Radiocarbon dates recovered during the course of these excavations suggested that the site was occupied during the San Pedro phase (1200-800 B.C.; Ezzo and Deaver 1996). During Spring 1995, Desert Archaeology, Inc. (Tucson, AZ) tested a linear alignment along the south side of Ina Road from Silverbell Road to a point approximately 0.2 km west of the Interstate. A few possible pit structures were found. The author was able to confirm, by only cursory examination of trench sediments, that the sedimentary deposition along Ina Road, exposed in the trenches, was similar to that noted by Haynes and Huckell (1986) and Ezzo and Deaver (1996) elsewhere in the area. Geologic Background Surficial deposits within the project area were mapped by the Arizona Geologic Survey in 1988 (McKittrick 1988, Figure 4.11). The project area comprises the alluvial fan deposits (M2) at the very western edge of the project area and the Holocene floodplain of the Santa Cruz River (Qt2). Testing did not extend far enough to reach the Pleistocene terrace (Qt3) of the Santa Cruz River. The historic floodplain, labelled Qt 1 by the Arizona Geologic Survey, was not encountered in any of the trenches; however, significant modern disturbance is present near the channel 163 thalweg where the Ina Road bridge was installed. Interfingering between alluvial fan sediments and Holocene terrace sediments occurs in trench 3. co co \to ...... co ,0 (;) .... 411 CV- Ill' D .1 INES M2 71-. 3578000 m N ah,h, /.1 t2 M1 A VY ss, -4 .:••• M2 ch QTbf QTbf Qtbf Figure 4.11. Surficial geology in the Ina Road area (after McKittrick 1988) showing relevant archaeological sites and geologic cross-sections. UTM coordinates place the map in its position on the USGS 7.5' quadrangle for Jaynes, Arizona. 164 The project area is located immediately downstream from the confluence of the Santa Cruz River and the Canada del Oro. The influx of sediments from the Canada del Oro significantly affects sedimentation in this part of the Santa Cruz River. In addition to providing different source materials for sedimentation in this part of the valley, the geomorphic processes acting on a reach where a significant drainage adds it's discharge to the total discharge of the valley produces distinct depositional results (effects). In the case of this reach of the Santa Cruz, sediment deposition appears to be fairly continuous with little chance for incipient soil development. Previous Geochronologic Work in the Project Area In 1979 and 1980, Bruce Huckell and Vance Haynes examined prehistoric features exposed in a trench excavated for a sewer line and excavation for the Ina Road landfill. They derived several dates on geologic units exposed in the excavated trenches and have correlated these units with their geochronologic investigations elsewhere in the Santa Cruz Valley (Haynes and Huckell 1986). The two upper units, which extend at least five feet below the surface along most of Ina Road and pinches out on the east side of I-10, date to between 2846 ± 73 (average of 4 dates) and 1400 ± 220 (A-3141) years B.P. Weak soil development separates these silty sand deposits from the underlying geologic deposits. These underlying deposits date approximately 4260 ± 140 (A-2234) and 3222 ± 43 (average of three dates). Continued excavations 165 in 1980 for the Ina Road landfill revealed a number of preceramic fire hearths 3.5 to 4.5 meters below the modern floodplain. At least one of these hearths consisted of a 2 m diameter mixture of cobbles and charcoal (Haynes, personal communication, 1997). Radiocarbon dates on these features yielded dates between 3700 and 3100 years B.P. (Haynes and Huckell 1986). Recent Excavation At AA:12:503 (ASM) by Statistical Research During summer and fall 1995 Statistical Research conducted testing and data recovery at AA: 12:503, named the Costello-King site. They identified four strata, each containing cultural material. The uppermost deposit is predominantly a clay and contains Rincon phase ceramics. Underlying this horizon is a rather thick deposit variably classified as silty slackwater deposits and "fossil channels" which they characterize as the distal portion of the Canada del Oro alluvial fan (Ezzo and Deaver 1996). Artifacts and features within this deposit date to roughly 2700 B.P. This unit is underlain by another deposit of clay, which was not extensively investigated. Recent Excavation at AA:12:130 (ASM) by SWCA In November of 1996, SWCA Environmental, Inc. excavated five trenches east of the Santa Cruz River and south of Ina Road. The first three trenches were placed at an angle approximately parallel to the current channel of the river. Trenches 4 and 5 followed the proposed effluent pipe realignment to the south. Trench placement was 166 fortuitously advantageous for compiling a cross-section of prehistoric and historic sedimentation. The trenches were located on a relatively flat parcel of land near a large pit excavated as part of the landfill operation. Tree growth in the pit indicates that the pit has been present for several years, perhaps decades. The upper meter (approximately) of sediment on the flat parcel where the trenches were located is comprised of modern fill. Two prehistoric sedimentary units were recognized in SWCAs trenches. The westernmost sediments are young in age, and represent a late prehistoric to protohistoric channel of the Santa Cruz River. This channel (or set of channels) is present upstream and has been identified at the Los Pozos (AZ AA:12:91) and Wetlands (AZ AA: 12:90) sites (Freeman 1997). Bedding structures are still visible in both sandy and silty sediments indicating that they were once part of channel filling events and that these events have not been subsequently altered by soil forming processes. At the easternmost end of trench 2 and continuing in trenches 3 through 5 were a series of clay and silt bands. The boundary between these floodplain deposits and the younger channel contained several ephemeral features of charcoal and fire-cracked rock. Plain ware sherds were found underlying these features. The uppermost of these clay bands probably represents the pre-incision floodplain margin of the Santa Cruz River, sometime during the Hohokam period. 167 The lower series of clay bands (which are best represented in trenches 3 through 5) represent slightly older floodplain sediments. Carbonates appear in few, fine filaments and there is little other evidence for long episodes of pedogenesis. A single shallow channel was present within this package of sediments. It is possible that these units are equivalent in age to the units described by Haynes and Huckell (1986) nearby; however, they are probably slightly younger, given the relative position and apparent age of the sediments (probably somewhere in the range of 1400-2700 yrs B.P.). Results of Testing Along Ina Road Excavation of backhoe trenches along Ina Road revealed a 1500 m profile of alluvial sediments. West of the Santa Cruz River, at AZ AA:12:315, the sediments are characterized by alluvial fan sediments and Holocene alluvium derived from the Santa Cruz River. Neither the archaeological features nor the geologic strata provided significant distinguishing characteristics to assess the age of these deposits. East of the Santa Cruz River, at AZ AA: 12:503, the sediments are characterized predominantly by overbank and slackwater deposits of the Santa Cruz River. Shallow channels, similar to those identified by Statistical Research at the Costello-King site and by the author at AZ AA: 12:130 were encountered in some trenches. The succession of sediments was relatively homogeneous across the trenched area, indicating that the profile drawn by Haynes and Huckell (1986; Figure 168 4.12) is representative of the general stratigraphy. Although it is difficult to assess the reliability and significance of a single radiocarbon date, it appears that stratigraphy containing intact artifacts and features dating to approximately 2400 yrs. B.P. is preserved in at least a part of the project area. D Silver Bell Road Santa Cruz River Gravel Pit 01. Brook sachon 3140±90(A-3146) 3240 t 50(A-3147) 3730 t 110 (A-3104) 3260 t100(A- 3145) 1970 t 310(AA- 319)- Ina Land Fill Ina Rood ••••• ..... .•• f A 1ft 2700t 130 (A-2453) 1-1400 220 (A-3141) 1-2820±2130 (A-2237) 2720±210 (A-2452) 2970±100 (4-2451) 42601140 (A-2234) I-I0 B I C?) ... 2150 - Figure 4.12. Generalized diagram of alluvial stratigraphy along Ina Road (from Haynes and Huckell 1986). Correlation Correlation between the work conducted by Haynes and Huckell in 1979 and 1980 and the present excavations suggests that the present excavations recovered data from only the up per units described in their report. The radiocarbon age estimate . suggests that features recovered during testing along Ina Road represented an age equivalent to the middle portion of Haynes and Huckell's (1986) unit C2. Their 169 limited testing did not recover artifacts and features dating to this period in this portion of the Santa Cruz River Valley. Summary Although only limited excavation and testing has been conducted along Ina Road, some preliminary conclusions can be suggested regarding the nature of stream behavior along this portion of the Santa Cruz River. The project area appears to be aggrading through the entire prehistoric sequence represented in this reach, beginning around 4000 B.P. Sometime during the late prehistoric or protohistoric period, the river has incised, forming the boundary between the two units identified in trenches excavated by SWCA. This "sediment storing" behavior is atypical of upstream reaches of the Santa Cruz River, where periodically throughout the prehistoric sequence, the river incised, sometimes to depths of at least 2 or 3 meters. This process, which involves aggradation and degradation simultaneously occurring in different reaches of the same river system, is known as "complex response" (Schumm 1973; Schumm and Parker 1973). The river forms a sort of distal alluvial fan in this reach of the river. During those periods of incision, it appears that the stream forms a braided system, probably as a result of high levels of sedimentary input from all parts of the fluvial system upstream (the Santa Cruz River, Rillito Creek, and the Canada del Oro Wash). 170 SILVERBELL ROAD Sites AZ AA:12:502 and AZ AA:12:750 are located along the margin at which alluvial fans originate in the Tucson Mountains enter the Santa Cruz River Valley (Figure 4.11). This margin is nearly perfectly matched with the northwest to southeast trending Silverbell Road. Older alluvial fans are typically found on the west side of Silverbell Road. The floodplain and terraces of the Santa Cruz River are typically found on the east side of Silverbell Road. The sites are located on a portion of the floodplain that has been alternately mapped by the Arizona Geologic Survey as either the youngest terrace of the Santa Cruz River or alluvial veneers that mantel older fan surfaces (McKittrick 1988). However, neither of these descriptions accurately reflect the origin of the sediments found in archaeological test trenches. Both sites are bounded northwest and southwest by older alluvial surfaces which no longer exhibit fan morphology. These surfaces (labelled Qtbf by AGS) are dissected by interfluves composed of younger alluvium (labelled M1 and M2). The examined archaeological test trenches were predominantly located on the east side of Silverbell Road in alluvium that was transported by the Santa Cruz River. The sediments exposed in archaeological test trenches consisted predominantly of finegrained silts and clays representing overbank or slackwater deposits of the Santa Cruz River. The location of the archaeological test trenches at the boundary between the interfluves and relatively young alluvium probably promoted the deposition of these fine-grained materials. 171 Archaeological features at site AZ AA:12:750 appear to be cut from or overlying a weakly developed paleosol. The features are preserved by deposition of additional bands of fine-grained silts and clays. Subtle evidence for pedogenesis is found within these younger bands. A fairly well-developed paleosol laden with abundant large stringers of calcium carbonate is found well below archaeological features discovered at AZ AA:12:502. The exception, a burned surface is located below that paleosol and likely represents a natural fire. This feature appears to have been formed atop another, older paleosol; however, the sediment on which the feature was discovered was not exposed adequately in the trenches to determine its pedogenetic properties. Archaeological features at both sites are preserved in floodplain sediments of the Santa Cruz River. These sediments are appear to be similar in age to those found within the Holocene terrace of the Santa Cruz River (labelled t2 by AGS). Numerous sites dating to the Early Agricultural and Early Ceramic periods have been found these deposits elsewhere along the floodplain of the Santa Cruz River. SUMMARY Recent excavation between A-Mountain and Ina Road has begun to fill in a critical portion of the alluvial record. However, the records gained demonstrate a critical aspect of alluvial stratigraphy, not all records are equal. Many of sequences described represent different portions of the channel system. Some deposits represent 172 the fill of paleochannels or the channel margins, while other parts of the record represent the floodplain component of the record. The relationship between the paleochannel and the floodplain components are rarely observed in a single reach, as archaeological testing normally attempts to excavate the areas in which archaeological sites will be preserved and erosion of the current channel often exposes the younger, poorly-consolidated sediments. Utilizing the radiocarbon record or the presence of archaeological materials as a guide, these records can be chronologically correlated. The numerous geologic sections presented also demonstrate that although chronological correlation is possible, the records from different reaches may exhibit different fluvial processes. The differences between these reaches are important in interpreting the way in which prehistoric people could have used the river and the manner in which channel changes took place. ENDNOTES 1. In places, these unconsolidated sediments appear to be inset against a buried late Pleistocene terrace (Qt3). 173 CHAPTER 5 ALLUVIAL STRATIGRAPHY, GEOCHRONOLOGY, AND GEOMORPHOLOGY OF THE SANTA CRUZ BEND AND JUHAN PARK SITES INTRODUCTION The Santa Cruz Bend (AZ AA:12:746, ASM) and Juhan Park (AZ AA:12:44, ASM) sites are located in a sharp bend of the Santa Cruz River, just west of Interstate 10 near the Miracle Mile interchange (Figure 5.1). The abrupt turn of the river here has exposed prehistoric alluvium on both banks of the arroyo. Archaeological investigations conducted at both sites between 1993 and 1996 and geologic investigations of Holocene alluvium at both sites have contributed a high-resolution record of human use of the floodplain as it transformed during the middle to late Holocene. The Santa Cruz Bend, Square Hearth (AZ AA:12:745, ASM), and Stone Pipe (AZ BB: 13:425) sites were investigated by Desert Archaeology, Inc. between 1993 and 1995. Large-scale excavations were carried out at these sites as a part of cultural resource management investigations of the Interstate 10 corridor. The vast horizontal extent of these early farming villages literally changed the face of Tucson Basin archaeology. The three sites produced radiocarbon dates spanning the period from 2500 to 1600 B.P. Radiocarbon dates from the Santa Cruz Bend site fall at the early end of 174 Square Hearth stratigraphic pit Figure 5.1 Surficial geology of the Juhan Park and Santa Cruz Bend site areas showing relevant geologic cross-sections and stratigraphic profiles. UTM coordinates place the map on the USGS 7.5 topographic quadrangle for Jaynes, Arizona. the spectrum, dates from the Square Hearth site fall at the late end of the spectrum, and dates from the Stone Pipe site are bracketed by dates from the other two sites. 175 The sites exhibit characteristics that appear to reflect a trend toward sedentism (Mabry 1996a). Although the foci of the geologic studies described in this chapter are the Santa Cruz Bend and Juhan Park sites, Stone Pipe and Square Hearth are also described briefly below. Santa Cruz Bend (AZ AA:12:746) The Santa Cruz Bend site is located just east of the Santa Cruz River and west of the Miracle Mile interchange. The exposed (excavated) portion of the site covered roughly one hectare; however, the site is estimated to cover approximately eight hectares. Because the right-of-way was not restricted to a linear exposure, more detailed information regarding site structure could be gleaned from this site. Sixty-three Cienega phase pit structures and numerous extramural pits were completely or partially excavated at the site. One large structure, which may have been used for communal or ceremonial activities, was found in the northwestern portion of the site. Numerous other features were recorded in trenches or exposed during backhoe stripping but unexcavated. It is estimated that a total of 2,400 features may have been present prehistorically. Mabry (1996a) defined five classes of pit structures based on size of the structure and types of features inside the structure. He believes that these features may have served different functions. Some of the pit structures appear to form circular clusters, which may have been the domain of an extended family. All pit structures at the site consist of shallowly-excavated, round subsurface 176 depressions. Within these depressions are small and large intramural pits. Formal intramural hearths are rare. Those pit structures thought to be used for storage contain numerous large (1 m diameter, 1 m deep) intramural pits and tend to have a smaller surface area, while habitation structures contain shallow, informal depressions that may have served as resting places for household items and have a larger surface area. Some storage structures were found in the center of household clusters. No formal entries are present within these pit structures, however, larger gaps in posthole spacing may have served as an entryways. Stone Pipe (AZ BB:13:425) The Stone Pipe site was excavated in a single narrow strip along the east side of Interstate 10 between Speedway Boulevard and Grant Road. A small portion of the site is located on the opposite side of the freeway. The excavated site area covers 250 m by 25 m, though the highest concentration of structures is located in an area one-half that size. Twenty-five of the dated pit structures at the Stone Pipe site date to the Cienega phase, while 15 of the dated structures date to the Agua Caliente phase of the Early Ceramic period. Numerous other structures were not dated. Some of the Agua Caliente phase structures were rectangular in shape, distinguishing them from the round, Cienega phase structures. Some of these later structures also exhibited features that indicate greater labor investment in building the structure, such as plastered hearths and walls, and formal entryways flanked by adobe pillars (Mabry 1996a). 177 Square Hearth (AZ AA:12:745) The Square Hearth site is located on both sides of Interstate 10, just north of Miracle Mile and may simply be an extension of the Santa Cruz Bend site. The exposed portion of the site measures approximately 40 m by 40 m. The site is much lower in feature density than the Santa Cruz Bend site. Radiocarbon dates on pit structures from the site differentiate it from Santa Cruz Bend, placing it firmly within the Agua Caliente phase of the Early Ceramic period. Four round structures and one rectangular structure were found. Although the sizes of individual structures vary, the average structure found at the Square Hearth site was generally larger and deeper than the average pit structure found at Santa Cruz Bend, and most contained raised square hearths and entryways flanked by adobe pillars. At least two of the structures had plastered floors and walls. Both the depth of the structure and plastered features indicate a substantial increase in labor investment over the structures at Santa Cruz Bend. Juhan Park (AZ AA:12:44) Originally discovered and recorded in 1957, this site is located directly across the Santa Cruz River from Santa Cruz Bend. A flexed burial was discovered at 2 m depth along the river bank. During subsequent surveys, Rillito Red-on-brown and Salado and Gila polychrome sherds were discovered on the bank surface along with several pieces of flaked stone; artifacts dating to the Hohokam period are still found at the surface along the northwestern portion of the site. Several roasting pits are also 178 visible on the surface. During Spring 1996, Desert Archaeology conducted archaeological testing of the parcel. Few archaeological features were found; however, both the surface distribution of artifacts and the few subsurface features identified indicate that an Early Agricultural occupation is present in at least part of the project area. Implications of Excavations at the Santa Cruz Bend, Stone Pipe, and Square Hearth Sites A variety of other evidence from the sites has been used to support the trend toward sedentism. These include the initial production of pottery, the establishment of trade networks, and tool recycling (Mabry 1996a). Although the presence of numerous structures does not necessarily imply sedentary settlement, site structure and structural features indicate that some degree of permanency of settlement was practiced. For instance, few of the pit structures overlap, which may indicate either memory of former structures or the present remains of those structures. Large storage features and increased investment in construction during the Early Ceramic period favors year-round occupation and over-wintering at the site. The presence of large communal or integrative structures may be evidence that some form of community organization (e.g., ceremonial activities, meetings, etc.) was taking place during this period. As demonstrated in Table 2.1 (Chapter 2), maize ubiquities at these sites are very high. This, combined with the very high capacity for storage, suggests that Early 179 Agricultural and Early Ceramic period floodplain sites received a major portion of their diet from maize agriculture. Unfortunately, there are no winter indicator plants that might confirm the presence of people at the sites during this season (L. Huckell 1996); however, the remainder of the plant assemblage suggests a nearly year-round occupation of the sites. In addition to maize, the Santa Cruz floodplain sites produced evidence for beans, squash, cotton, and tobacco. The rapid introduction of additional cultivated plants provides further proof that these groups were well on their way to becoming agriculturally dependent and that by the Early Ceramic period, the plants that formed the basis for Hohokam subsistence were already in place in southern Arizona. While numerous wild plants were also exploited, most could be found in surrounding biotic communities formed on the bajadas of the Tucson and Santa Catalina Mountains (L. Huckell 1996). Evidence from palynology at the sites supports the presence of an environment favorable to plant cultivation. The presence of cattail (Typha) pollen suggests that perennially-wet sediments were present near the site. Mesquite (Prosopis), willow (Salix), hackberry (Celtis), and sedge all indicate at least seasonally damp soils in the vicinity of the sites. The status of the floodplain is better understood by examination of the geology of the site area. GEOLOGICAL INVESTIGATIONS Geologic investigations at the Santa Cruz Bend and Square Hearth sites were conducted by Bruce Huckell in 1993 and 1995 and published in a synthetic volume 180 dealing with those sites (Mabry 1996a). In addition to two relatively narrow but deep stratigraphic pits at the Square Hearth and Santa Cruz Bend sites, Huckell (1995) mapped a 140 m long exposure along the right bank of the Santa Cruz River. During spring 1996, the left bank of the river directly across from the Santa Cruz Bend bank exposure was mapped as part of archaeological investigations at the Juhan Park site (AZ AA:12:44, ASM). The Santa Cruz River takes a sharp bend to the west here, exposing a vertical wall of sediments in places 6 m high. A 600 m profile of the left bank of the river was mapped using electronic mapping techniques, supplemented by hand-drawn profiles. These mapped localities are within 3 km of the Los Pozos site. Because of their close proximity to the Los Pozos deposits and the quantity of exposed and documented sediments, correlation between geologic events in the Santa Cruz Bend site area and the Los Pozos site area (the focus reach) should reflect the relationship between intrinsic geomorphic processes and the preserved results of those processes more accurately than correlation between geologic events over the entire Tucson reach (the study area). The first part of this chapter summarizes the results of Huckell's research in this area. The second part describes the results of mapping efforts along the Juhan Park side of the river. Description of the Area The Juhan Park and Santa Cruz Bend sites are located on surfaces defined by the Arizona Geological Survey (Jackson 1989; McKittrick 1988) as representing Qt 1, 181 Qt2, and Qt3 (the Holocene and Pleistocene terraces; see Figure 5.1). Though the defined boundaries of these mapped units would probably be changed on the basis of intensive subsurface investigation, all three terraces are represented in this general area. Because all three terraces are present, the banks of the river, which cut through these surfaces, reveal a remarkable history of alluvial cutting and filling events. The Santa Cruz Bend Profile Huckell (1996a) defined six stratigraphic units in the right bank profile and stratigraphic pits excavated at the Square Hearth and Santa Cruz Bend sites dating from at least the Middle Holocene. Summary descriptions are given below'. Unit 1 is comprised of a sandy coarse gravel, firmly cemented with calcium carbonate. The contact between units 1 and 2 along the river bank was sharp and erosional, while in the two stratigraphie pits this same contact was gradational. Huckell (1996a) attributes this to the small exposure of the unit in the two stratigraphic pits and the well preserved nature of the contact and erosional features in the bank profile. Unit 2 is comprised of an arkosic sand. Dispersed calcium carbonate occurs throughout the deposit and is prominent in the upper 50 cm. In the two stratigraphic pits, a thin clay-rich B-horizon overlies this sandy unit. Carbonate concentrations in the two stratigraphic pits rise sharply beneath the clay horizon and peak at 60 cm below the clay. The upper contact in the two stratigraphic pits is sharp whereas at the riverbank the contact is gradational. Huckell attributes this to possible erosion of the clay B-horizon at the riverbank profile. 182 Unit 3 is a pinkish brown silty clay. This unit is hard in consistency and contained abundant small nodules of calcium carbonate, particularly in the upper portion of the unit. In both stratigraphic pits, the upper portion of the unit contains abundant organic matter constituting an A-horizon and no evidence for erosion of the paleosol. Microstratigraphy, in the form of thin (2-10 cm) dark gray organic clay bands separated by lighter-colored bands, is present along the riverbank, suggesting that it received sedimentation from floodwater. Unit 4 is comprised of a series of clay and silt bands which vary in number and thickness. In the riverbank, the lower portion of this series of bands forms two clay-filled channels, which Huckell calls charcos. He suggests that the two channels are either backwater sloughs of abandoned channels or filled crevasse-splays (upstream channel breaches). He favors the latter interpretation, noting that channel cutting and filling does not appear to be of long duration. Small calcium carbonate nodules were found dispersed throughout the channel and manganese and iron oxide stains were found on ped faces. A few scattered pieces of fire-cracked rock and a single piece of flaked stone were discovered in the riverbank profile. Unit 5 is a light brown sandy silt to silty sand. Early Agricultural period artifacts and features were present within this sedimentary unit. Two zones of pedogenic carbonate were noted in the stratigraphic pits. The top of unit 5 is eroded and truncated and unit 6 sediments are deposited in a channel fill and on top of unit 5. Both historic trash and prehistoric artifacts are found within unit 6 and Huckell believes that the cutting event may have taken place 183 as late as the Rincon phase of the Hohokam Sedentary period (ca. A.D. 1100-1150). Radiocarbon Dates and Correlation Radiocarbon dates on the charcoal from the riverbank profile and the stratigraphic pit are reported and compared in Table 5.1. Many of the radiocarbon dates were run on dispersed charcoal which must be used carefully in Table 5.1. Unit 5 Radiocarbon dates on alluvial stratigraphic units at the Santa Cruz Bend and Square Hearth sites (as reported by Huckell 1996a). Riverbank Profile 2360 ± 60 (Beta 81057) AA:12:746 Strat. pit 1770 ± 60 (Beta 80152)* AA:12:745 Strat. pit 410 ± 60 (Beta 81049)** 2440 ± 70 (Beta 87067) 3600 ± 60 (Beta 87068) 5b 3800 ± 50 (Beta 81054)* 4 3 2 Note: 2290 ± 60 (Beta 81053)** 3830 ± 60 (Beta 81047) 4610 ± 50 (Beta 81056) 3740 ± 60 (Beta 81050) 4380 ± 60 (Beta 81048) Asterisks denote date considered questionable (*) or probably in error (**). All dates are uncorrected AMS assays on wood charcoal. alluvial stratigraphy. Criteria for accepting or rejecting radiocarbon dates are established by evaluating the context and pretreatment results of each sample. Huckell (1996a) has rejected several of these dates and his reasons for doing so are outlined below. A radiocarbon date on charcoal from unit 2 of the Square Hearth stratigraphic pit yielded an age estimate of 4380 ± 60 B.P. Overlying unit 3 along the riverbank 184 yielded an age estimate of 4610 ± 50 B.P. The presence of abundant evidence for rodent disturbance in the stratigraphic pit caused Huckell (1996a) to favor the radiocarbon date from the riverbank profile and to suggest that the date from the Square Hearth stratigraphic pit is "somewhat too young." Huckell (pers. comm. 1997) suggests that older evidence of bioturbation, which may be unrecognizable, may have transported the charcoal in unit 2 from one of the younger units above. Two radiocarbon assays from the base of the organic part of the unit 3 soil yielded identical age estimates of 3740 ± 60 B.P. Huckell suggests that these dates also may be just slightly too young. He suggests that both flecks may derive from a single tree root that burned and was incorporated into the unit 3 paleosol. A similar date of 3830 ± 60 B.P. was determined from charcoal in overlying unit 4 approximately 10 cm above the unit 3 paleosol. Also in unit 4 of the Santa Cruz Bend stratigraphic pit was a sample yielding an age of 2290 ± 60 B.P. Because the late preceramic archaeological site was located 1.5 m higher in the section and because dates on features at that site were between 2,000 and 2,400 years old, Huckell rejects the date, again proposing that the charcoal is intrusive into the unit from an imperceptible rodent hole. Four age estimates were recovered from unit 5. The first date of 2360 ± 60 was obtained on charcoal from the vicinity of the cluster of fire-cracked rock in the riverbank profile. Another date on a small channel in the riverbank profile, which appears to be cut from a vague soil boundary at the level of this fire-cracked rock concentration, yielded an age of 2480 ± 70 B.P. The channel may have been a 185 cultural feature, such as a drainage ditch (Mabry 1996a), or may be a natural erosional feature. These dates conform with the Early Agricultural occupation at the site. A date of 410 ± 60 B.P. on dispersed charcoal from the Square Hearth stratigraphic pit and 1770 ± 60 B.P. from a wood charcoal within a pit structure cut into this unit at the Santa Cruz Bend site appear too young. Additional Evidence Supporting Radiocarbon Date Rejection The similarity of sedimentary properties and elevational data from the two stratigraphic pits and the riverbank indicate that Huckell's correlation of stratigraphic units is in proper position. Furthermore, the sediments documented tend to indicate vertical accretion of floodplain sediments (Huckell, pers. comm. 1997), rather than accumulation in a paleochannel. Combined with the potential for intrusion of younger charred material into older units through bioturbation or other means, these data support rejection of radiocarbon dates as indicated by Huckell (1996a). A diagram of Huckell's correlation of geologic units is presented in Figure 5.2. Alternative Interpretation of Santa Cruz Bend Stratigraphy An alternative interpretation of the stratigraphy at the Santa Cruz Bend and Square Hearth sites is supported by acceptance of the radiocarbon dates and somewhat different view of interpretation of sedimentary and soil characteristics. This interpretation holds radiocarbon age estimates more important than elevational data, particularly across a 950 m stretch of floodplain. 186 z 700 11 ca. 150 m ca. 800 m E' 1 2440 70 (Beta 87067) 699 N \ \ \ \ \ 1770 60 2360 t 70 (Beta 81057) (Beta 80152) Untt 5 Érn rj= – 410 *60 (Beta 81049) 3600-260 (Beta117068) 3800*50 (Beta 810547 .... 698 5 5 TrriTiTTTTTIT Watt 4 4 2290* 60 (Beta 81053) 697 1.1111111 .11 4. 3 3740 *60 (Beta 81050r87089) — 696 r. en, • r• ; • 695 run* • • r", UNt 1 r• , 694 VZ • ..b qe55ij ec;:ip.• lat: OD°F . .0 .. • • I. ... 693 692 Elevation (masl) Figure 5.2. Generalized stratigraphic columns at the Santa Cruz Bend and Square Hearth sites, showing inferred correlation of stratigraphic units by Huckell (1996). 187 Floodplains accrete sediments in a generally conformable manner on top of older surfaces which may or may not be flat. These older surfaces may represent the filled thalwegs of paleochannels, the near-bank margins or the outer margins of former floodplains, abandoned meander scars, or filled lateral arroyos. Therefore, the floodplain itself may not be flat during certain periods of time. Often floodplain surfaces will slope toward the channel present at that particular time. The floodplain will also tend to slope downstream, conforming roughly to the slope of the channel (though not necessarily at exactly the same angle). Because the sample of examined sediments within the floodplain is so small, this alternate explanation focuses on the radiocarbon dates first, with some consideration for the sedimentology and soil formation within the deposits, and then attempts to find a geologic explanation for the radiometrically-dated succession of sediments found in different parts of the floodplain. It should be noted that channel deposits were not found in any of the profiles above the upper portion of unit 2 and below unit 5. Therefore, all profiles exhibit sedimentation within the floodplain. Former channels are expected to be either washed away in the present thalweg or somewhere to the east of the present thalweg. This alternate explanation acknowledges that Huckell's (1996) unit 2 appears to be traceable across all profiles. The exhibition of characteristics typical of an arkosic origin of the sediments and soil formation on this unit in both of the stratigraphic pits suggests that these two units are, indeed equivalent in age. The problem then remains that a date from the channel wall profile in unit 3 of 4610 ± 50 exceeds a date on unit 188 2 of 4380 ± 60 by several hundred years. The clay B-horizon present in the stratigraphie pits is apparently missing in the upper portion of unit 2 in the riverbank profile while a series of pedogenically-altered clay bands are present unit 3 in the riverbank profile, but not documented in either stratigraphic pit. Huckell (1996a) interprets the missing clay B-horizon in the riverbank profile as the product of erosion. In a similar vein, Huckell (1996a) has proposed that pedogenically-altered clay bands in the riverbank profile are a product of overbank deposition, while clay missing in the stratigraphie pits is unaccounted for. Since there was apparently no erosion of the unit 3 paleosol in the stratigraphic pits, one must assume that the absence of these clayey sediments in the stratigraphic pits is a product of non-deposition, rather than erosion. While this is a possibility, an alternate explanation can be derived that accounts for both the radiocarbon record and the stratigraphic sequence of sediments in the riverbank profile. This alternate explanation places the second band of clay sediments in the riverbank profile (unit 3) within the upper portion of unit 2 in the stratigraphic pits, forming a compound soil on unit 2 in the stratigraphic pits; notation of a compound soil is shown in Huckell's figures of the stratigraphic pit profiles. The lack of a strong hiatus between unit 2 and unit 3 in the riverbank profile is perhaps better explained under this interpretation. Although soil formation is present in unit 2, sedimentation continues, virtually uninterrupted throughout unit 3 at the riverbank and is followed by a second episode of soil formation. Under this interpretation, sedimentation in the stratigraphie pits continues following erosion and soil formation at the top of unit 2; 189 this episode of sedimentation is considered equivalent to unit 4 deposition along the riverbank. The remainder of the radiocarbon dated strata fall in line with the possible exception that Huckell (1996a) segregates clay bands found above unit 3 in all profiles from unit 3 and that he also assigns clay bands found in the riverbank profile to unit 5. It seems possible that the series of clay bands and silt units in the riverbank profile, including the two charcos, might correlate with the lower set of clay bands in the Square Hearth stratigraphic pit, thereby explaining the older age on this first clay band. This alternate explanation is presented in Figure 5.3. Re-ordering of the radiocarbon dates according to this alternate stratigraphic interpretation is presented in Table 5.2. Additional Evidence for Alternate Interpretation In the fall of 1995, I had the opportunity to examine a pit for a pipeline underlying part of new construction associated with Interstate 10 east of the riverbank profile and south of the Santa Cruz Bend stratigraphie pit. The first observation recorded from this pit is that the gravels and sands near the bottom of the pit did not exhibit the heavy cementation found in cobble gravels along the riverbank profile, suggesting that the cobbles found in the riverbank, which may be Pleistocene in origin, are unrelated to the cobbles found in the bottom of the stratigraphie pits, which may be Holocene in origin. There are also a greater number of soils and sedimentary units found in this profile than were exhibited in any of the stratigraphie pits, indicating that 190 ca. 150 m 700 ca. 800 m 2440 t 70 (Beta 87067) 2360 t 70 (Bata 81057) 1770 a 60 (Beta 80152) 699 5 698 2290 *60 (Beta 81053) 897 3740 a 60 (1?ata 81050, 87069)/ 4610 50 (Bata 81056). 'MT3U 3 . 695 694 ?.. ebae •, ; 0* • :** • .• •_ • OL' elo, VC?•••-ce - t. 4%' ...o .o.0 : ° ;?*(1.•:o‘P *9 -• P °693 692 Elevation (mast) Figure 5.3. Generalized stratigraphic columns at the Santa Cruz Bend and Square Hearth sites, showing alternate interpretation of correlation between stratigraphic units. 191 some of those units have either been eroded in the other profiles or have formed compound soils. Table 5.2. Alternate division of alluvial stratigraphic units at the Santa Cruz Bend and Square Hearth sites. AA: 12:746 Unit Riverbank Profile Strat. pit 6 5 1770 ± 60 (Beta 80152) 410 ± 60 (Beta 81049)* 2360 ± 70 (Beta 81057) 2440 ± 70 (Beta 87067) 4b AA: 12:745 Strat. pit 2290 ± 60 (Beta 81053)* 3800 ± 50 (Beta 81054) 3830 ± 60 (Beta 81047) 3600 ± 60 (Beta 87068) 4 3740 ± 60 (Beta 81050) 3740 ± 60 (Beta 87069) 3 4610 ± 50 (Beta 81056) 4380 ± 60 (Beta 81048) 2 Note: Asterisks denote date considered questionable (*) or probably in error (**). All dates are uncorrected AMS assays on wood charcoal. Radiocarbon date shown in unit 3 in the AA:12:745 strat. pit is actually within a compound soil formed from units 2 and 3, and is at the top of the unit therefore in unit 3. Sediments Revealed in Juhan Park Archaeological Testing The Juhan Park site is located on a surface defined by the Arizona Geologic Survey (Jackson 1989; McKittrick 1988) as Qt3 (the Jaynes or Pleistocene terrace). Although some of the sediments revealed during archaeological testing displayed weakly developed carbonate accumulation in the form of fine filaments, the carbonate accumulation does not appear to be sufficiently developed to support a 192 Pleistocene age. Clay present in some sedimentary units is due to sedimentation, rather than post-depositional alteration and no other attributes of soil development were present that would support a Pleistocene age for the sediments uncovered during testing. Therefore, the entire exposed section of project area is Holocene in age. In fact, the presence of discontinuous carbonate filaments and weak carbonate coatings suggest that the sediments are only 1,000 to 2,000 years old (Gile 1975). The southern and western portions of the project area are situated on a small rise which slopes abruptly to the north and east toward the Santa Cruz River. Although some of the slope at the northern end of the project area is natural, the surface has been significantly modified by modern human activity. A large deposit of trash located in the northern and eastern portions of the project area covers what was probably the natural cut of the channel around the turn of the century. Sediments underlying the trash deposit in this area (and exposed in trenches 5, 6, 7, 8, and 9; Figure 5.4) are composed of a massive homogeneous deposit of fine sand and silt. A massive deposit of sterile fill sediment has been placed on top of the high portion of the property to the south and west, possibly in an effort to level the surface for construction of the industrial park to the south. These sterile sediments cover Holocene alluvium of the Santa Cruz River which contain buried features and artifacts: Additional buried features are found in preserved Holocene sediments near the base of this rise (in the vicinity of trench 3; Figure 5.4). Sediments revealed in test trenches and in the exposed walls of the channel nearby are comprised predominantly of fine- to medium- grained deposits of Santa Page 193 CITY OF 'TUCSON RESULTS OF JUHAN PARK ARCHAEOLOGICAL TESTING AZ AA:12:44 (ASM) contour interval — 1 foot o 25 o 100 50 m 200 ft Computer cartography by GEO—MAP, Inc. 1996 EXPLANATION inhumation pit 0 cg? trench sb trash concentration ---- dirt rood existing chain link fence ..... proposed chain link fence proposed site development PROPOSED COPPER STREET ............ ... .............. „ ,„, ,„ ........ ...... Figure 5.4 Location of excavated trenches at the Juhan Park site .... HUACHUCA DRIVE 221 modem building area to be archaeologically monitored 194 Cruz River alluvium. Although a large portion of the sediments is silt and clay deposits that represent slow-moving overbank and slackwater deposits, some fine-to medium-grained sand deposits were also present; these were probably deposited by either small tributary or auxiliary channels of the river. The natural channel cut which forms the slope of the property and the sequence of deposits located within the walls of today's Santa Cruz River channel (to the northwest) indicate that much of the property is located near the former channel of the Santa Cruz. The Juhan Park Profile In order to complement the archaeological testing conducted at Juhan Park and to further investigate the exposure of sediments along the left (Juhan) bank of the river portions of the bank exposure along this stretch of the river were mapped. Six locations were mapped extending over 600 m of bank exposure. Profile 1 (Figure 5.5) is located directly beneath the Juhan Park site test excavations. Profile 6 (Figure 5.6) is located at the northernmost portion of the exposed bank, just south of the area marked by Arizona Geological Survey (Jackson 1989; McKittrick 1988) as Qt 1. The exposure covers middle to late Holocene alluvium of the Santa Cruz River and likely represents the margin of these channel deposits. The upper surface across the entire bank exposure has been eroded by modern flooding, most obvious near profile 6 where the upper soil horizon is truncated, forming a shallow saddle. 195 Unit 8: Gravel and historic trash, Qt1. Unit 7: Brown sand, upper boundary has been eroded and is truncated by unit 8. Unit 6: Pale to medium brown slightly loamy sand. Forms an abrupt, conformable boundary with unit 7, above. Unit 5b: Silt. Upper 40 cm contains abundant calcium carbonate filaments. Forms an abrupt conformable boundary with unit 6 above. Unit 5a: Silt and clay bands, overbank deposit. Forms a clear to abrupt boundary with unit 5b. Unit 4: Dark brown clay. Forms an abrupt boundary with unit 5a. Unit 3b: Silt, grading upward to very fine sand. Forms an abrupt boundary with unit 4. Unit 3b: Silt, grading upward to very fine sand. Forms an abrupt boundary with unit 4. Approximately 20 m north of this profile, this unit is conformably overlain by the post-3900 B.P. channel. Unit 3a: Silt and clay bands, overbank deposit. Forms a clear boundary with unit 3b.. Unit 2b: Medium brown silty clay. Forms an abrupt with unit 3a. Unit 2a: Fine loamy sand. Forms a clear boundary with unit 2b. Unit 1: Hard gritty muddy silt, heavily indurated with calcium carbonate. Forms and abrupt, erosional boundary with unit 2a. Figure 5.5 Stratigraphic column for Juhan Park, profile 1. Page 196 cross—section B EXPLANATION PIMA COUNTY 9 • radiocarbon sample CROSS-SIECTION KEY 696 TUCSON, ARIZONA WEST BANK OF SANTA CRUZ RIVER WEST OF MIRACLE MILE INTERCHANGE radiocarbon dote 697 — poorly developed poleosol pea grovel and coarse sand well developed poleosol coarse sand and grovel silt pebble grovel clay bonds pebble to cobbel gravel fine sand cobble gravel 0 10 m 0 695 50 ft Digital cartography by GEO—MAP. Inc. 7997 694 VER77CAL SCALES: numbers in bold represent elevation 693 medium sand //1 slump in meters above sea level (MASL) vertical exaggeration — 2X HORIZONTAL SCALE: in meters along profile face 692 Figure 5.6 coarse sand Juhon Pork profile 6. Inset is o higher resolution drawing of the bank prior to bank failure which resulted in the retreat of the bank by 6 meters. The inset shows the location of radiocarbon samples recovered prior to bank retreat. 197 Methods Hand mapping was conducted at each profiled area from a level line. These level-line profiles were connected by mapping the datum nails with an electronic total station surveying instrument by Geo-Map, Inc (Tucson, AZ). Profiles were tied together using a computer-generated graphics package and additional field research was conducted to fill in any gaps or unknowns. Profiles 2-5 were not intensively examined. The purpose of these intermediate profiles was to provide stratigraphic evidence for the succession of geologic events occurring between profiles 1 and 6 only. Profile I Profile 1 exhibits a series of Holocene soils representing former floodplains of the Santa Cruz River. In places, overbank deposition remains on top of the soils indicating that the post-erosional surface received some sedimentation after abandonment of each surface. Historic trash and cobble gravel form an inset terrace adjacent to this bank exposure. Stratigraphy, Geochronology, and Correlation The lowermost deposit (unit 1) is comprised of a silt, heavily indurated with calcium carbonate. This unit is most likely equivalent to Huckell's (1996a) unit 1 or intermediate between units 1 and 2. The amount of carbonate suggests that the sediment is likely Pleistocene in age (Gile 1975), although carbonate formation can 198 be the result of groundwater infiltration. Unit 2 is comprised of fine loamy sand, interrupted by a thin band of clay and followed by a medium brown silt to silty clay. This sediment is probably middle Holocene in age. The boundary between unit 2 and unit 1 represents the abrupt boundary between late Pleistocene/early Holocene sediments and middle Holocene floodplain deposits. Unit 2 is overlain by unit 3, a series of silt and clay bands representing overbank deposits of the Santa Cruz River, and a A second overbank deposit of silt, grading upward to a very fine sand. Either unit 2 or unit 3 likely represent Huckell's unit 2. Unit 3 is comprised of a thick, dark brown clay. This unit appears to be equivalent to Huckell's unit 3 in the two stratigraphic pits and the riverbank profile. It is also equivalent to a series of channels documented in profile 6. Charcoal recovered from a paleosol formed at the base of one of these channels yielded a radiocarbon date of 3810 ± 60 (CAMS-33965). These channels are truncated on both the north (profile 6) and south (profile 1) sides by a series of younger channels comprising the channel component of unit 4. Unit 4 is comprised of two silt and clay bands forming overbank deposits on unit 3, followed by a fairly thick deposit of silt with a moderately well-developed zone of carbonate forming the boundary between this deposit and overlying unit 5. This unit is almost certainly the same deposit as is represented in Huckell's units 4 and 5. As stated previously, the channel component of this deposit, which is 199 obliquely cut by the current bank profile forms the sediments intermediate between profile 1 and profile 6. The northern end of the channel is illustrated by the unconformity at the southern edge of profile 6 (left side, Figure 5.6) Unit 5 is comprised of a pale to medium brown slightly loamy sand. This deposit appears to be eroded in Huckell's (1996a) profiles, but may represent a portion of his unit 6. A fairly well-developed B-horizon, represented by the downward movement of organic matter, is formed at the top of this unit. The upper boundary of this soil is eroded and capped by a thin layer of gravel. Unit 6 is comprised of a well-developed A to A/B horizon soil formed on a silt. Historic or modern artifacts are located within this deposit, indicating that it represents the accumulation of trash and sediment moved by recent natural or human processes. Profile 6 Profile 6 (Figure 5.6) represents the obliquely cut Middle Holocene channel of the Santa Cruz River. It appears to be roughly equivalent in age to the lowermost unit in the Los Pozos west side stratigraphic pit (see Chapter 6). The profile was mapped twice by hand and once electronically. Unfortunately, bank failure occurred due to undercutting of the bank by the channel thalweg during the summer and fall thunderstorms, between each mapping period. 200 Stratigraphy, Geochronology and Correlation Units A and B are sand and gravel deposits of the Santa Cruz River that are at a minimum, middle Holocene in age. These deposits fine upward to unit C, a series of channel sands. These sand and gravel deposits are truncated forming the boundary between one set of channel deposits and the obliquely cut channel (Figure 5.6). Unit D is the fill of a middle Holocene channel. Two thin bands of clay, forming a compound soil in places, line the bottom of the channel. The channel is most likely obliquely cut, given the changes observed between hand-mapped and electronically-mapped profiles. This channel and the clay bands are equivalent to unit 3 of profile 1. Unit D is conformably overlain by sands (unit E, unit 4 of profile 1). These sands form the near channel floodplain deposits, while the intermediary sand and gravel sediments located south of profile 6 represent the channel itself. Unit F is the uppermost sediment represented in this profile. It is equivalent to unit 5 of profile 1. The upper portion of this sand and silt is eroded forming a shallow saddle on which ceramic period artifacts are found. SUMMARY The Juhan Park/Santa Cruz Bend reach of the river exhibits a record of floodplain and near-margin sediments of the Santa Cruz River during the middle to late Holocene. Their poor correlation in some parts of the project area demonstrate 201 the variability that can be represented by floodplain sediments that may form on uneven surfaces and the difficulties in interpreting the differences between depositional facies when the topography of the river and its paleochannels are not known. Sediments younger than approximately 1700 B.P. are poorly represented in any part of the profile. Channel sediments represented in the left bank profile are obliquely cut indicating that at least a portion of the late Holocene channel was near the current channel thalweg. The middle Holocene channel is probably represented elsewhere and may have been partially removed by the current channel. Because of the close proximity to Los Pozos, the floodplain sediments provide further information that can be utilized when interpreting the channel and floodplain sediments there and that are explored further in the next chapter. ENDNOTES 1. For additional detail, consult the original report (Huckell 1996a). 202 CHAPTER 6 ALLUVIAL STRATIGRAPHY, GEOCHRONOLOGY, AND GEOMORPHOLOGY OF THE LOS POZOS SITE INTRODUCTION The excavation of a multiple-component Middle Archaic and Early Agricultural period site along the floodplain of the Santa Cruz River has presented an opportunity to explore the period spanning the shift from hunting and gathering to agriculture. • Two exposures of middle Holocene alluvium provide indirect evidence of the environments surrounding the river during these two periods. Correlation between the geoarchaeological stratigraphy found at this site and excavations in other portions of the Tucson Basin allow us to explore desert stream processes and the effects those processes had on prehistoric settlement, subsistence, and site preservation. The purpose of this chapter is not only to place the Los Pozos deposits in their stratigraphic, geomorphic, and geochronologic context, but also to assess the prehistoric behavior of the river in relation to its known historic behavior. As an analog, historic situations are not always appropriate for evaluating the prehistoric behavior of a river. However, by combining historic records with the known dynamics of desert streams, the nature and regularity of streamflow can be assessed, providing additional information about the environment in which prehistoric people lived. 203 THE LOS POZOS SITE (AZ AA:12:91, ASM) The Los Pozos site (AZ AA:12:91) is located on the Holocene floodplain (Qt2) east of the Santa Cruz River between Prince and Ruthrauff roads within the City of Tucson (Figure 6.1). A number of Late Archaic projectile points were discovered on the site surface by an amateur archaeologist in the late 1940s (Morris 1951). Since that time, additional archaeological surveys (Fritz 1974) and inspection of remains found in berms and trenches excavated for activities associated with the Roger Road Sewage Treatment Plant (Huckell 1992) have yielded abundant evidence of prehistoric and historic cultural resources. During investigation of a trench excavated by an employee of the Treatment Plant in 1978, archaeologists working at the Arizona State Museum recorded a Late Archaic pit structure and burial preserved 40 to 50 cm below the ground surface. Wood charcoal from the pit structure yielded a radiocarbon date of 1780 ± 80 B.P. In 1991, additional bone fragments and teeth of a juvenile human were discovered near this location. Excavation of these remains demonstrated that they were located in secondary context. Samples of carbonized maize and charred wood from a charcoal lens were found about 2.5 m below the surface in a river channel deposit; the wood yielded a radiocarbon date 3230 ± 70 B.P. and the maize date was a few hundred years younger (Bruce Huckell, personal communication 1997). While not associated with any artifacts, the maize date is one of the oldest indications of agriculture in the American Southwest. 204 205 Investigations of the portion of the site within the Interstate 10 corridor were carried out in 1995 by Desert Archaeology, Inc. (Tucson, Arizona). On the west side of the freeway, portions of 42 pit structures dating to the Early Agricultural period were uncovered. On the east side of the freeway, a small Middle Archaic component was excavated. The two components are separated stratigraphically by several units of overbank deposition and horizontally by the interstate. The Middle Archaic Component The Middle Archaic component was restricted to a small strip of excavated area of roughly 400 m 2 and consists of a number of small pit features, flaked stone tools, ground stone implements, and botanical and faunal remains. Although few Middle Archaic archaeological resources were discovered in trenches beyond the boundary of the site, Middle Archaic features are often small and difficult to find. Therefore, the site may extend beyond the boundaries of the excavated area, but is obviously much smaller in size than later, Early Agricultural period sites. Archaeological resources at the site were found in a silty clay alluvial unit representing overbank and slackwater deposits of the Santa Cruz River and dating around 3800 B.P. (see Table 6.1). Projectile points found at the site include both Pinto (n = 4) and Cortaro (n = 14) styles as well as several other unattributable fragments. The Cortaro point is considered by Sliva (1997) to be an expedient point form. Although the Pinto and Cortaro point types appear to differentiate themselves 206 spatially and stratigraphically, secure radiocarbon dating places the Cortaro-style projectile point in the Middle Archaic period. The spatial and stratigraphic distributions of raw material sources suggest that at least two groups occupied the site and that these groups exploited different raw material sources. Other evidence from the flaked stone indicates that activities at the site included toolkit maintenance and refurbishing. Table 6.1. Radiocarbon dates from the Middle Archaic component at Los Pozos (after Gregory 1997). Sample Number Feature Number B-88145 none Sample Context Material Dated culture bearing natural stratum wood charcoal C13/C12 Conventional C14 Age -25.4 3700 +/- 70 (Leguminosae) B-95634 none culture bearing natural stratum wood charcoal (mesquite) -25.2 3820 +/- 70 B-88144 919 extramural pit wood charcoal (mesquite) -25.9 3820 +/- 80 B-95633 none culture bearing natural stratum wood charcoal -24.2 3880 +/- 100 B-95635 924 extramural pit wood charcoal (mesquite) -24.0 3900 +/- 80 B-81329 none culture bearing natural stratum wood charcoal (mesquite) -25.1 3950 +/- 100 B-88148 923.00 extramural pit wood charcoal (mesquite) -26.1 4250 +/- 60 Flaked stone analysis indicates that the site was a short-term hunting or "gearing-up" area, where flaked stone tools were maintained or repaired. On the basis of both the exotic origin of much of the flaked stone found at the site, and the 207 incongruity between raw material used in tools and debitage, Sliva (1997) infers that the site was used episodically by logistic groups of hunters. The faunal data support this argument, as do botanical and ground stone data. Fauna found at the site represent species that could have been obtained nearby. Although W6cherl (1997) attributes a relatively high artiodactyl count (compared to other Middle Archaic sites) to subsistence choice, no other excavated Middle Archaic site yielding a substantial sample of faunal remains has been located in such an environmental setting. The nearest approximations are sites in small ephemeral drainages such as the Harquahalla Valley. It is possible, therefore, that the environment created by the Santa Cruz River in this particular locale was favorable for both humans and their large artiodactyl prey. Additional bone tools were recovered, including three awl fragments, the tip of a bone flaker, and three indeterminate worked bone implements. Ground stone was infrequent, but not inconsistent in form with other Middle Archaic assemblages. One whole mano, two mano fragments, and three indeterminate fragments were found together with a single basin metate that had been destroyed by chipping a large hole in the bottom. Both bone and ground stone implements exhibited restricted distribution in the excavated area, and flaked stone debitage, including microdebitage, appeared to cluster around features, suggesting that the distribution of tools may represent work areas utilized by the different groups exploiting the site (Gregory 1997). Paleobotanical evidence was typical of a riverine grassland. Plants that are indicators of disturbance have a high ubiquity (Diehl 1997). Occasional trees, such as cottonwood or mesquite, may have been present nearby. A single maize cupule was 208 recovered from the Middle Archaic context. The flowering period of plant taxa represented at the site indicates that occupation of the site is most likely placed between September and November, although a possible May-June occupation may also be likely. As is typical with many floodplain sites, palynological sampling unfortunately yielded little information. The Early Agricultural Period Component The Late Archaic settlement of the Los Pozos site is typical of other Early Agricultural/Early Ceramic period settlements in the Santa Cruz floodplain. It consists of numerous pit structures and extramural pit features. Radiocarbon dates on the Early Agricultural period component have produced an average age estimate of 2097 ± 15 B.P.' (Table 6.2), falling on the late end of the Cienega phase. The report on the Early Agricultural period component has not yet been finalized; however, preliminary analyses indicate that the site is very similar to other Early Agricultural period sites on the floodplain of the Santa Cruz River and comparable to the Santa Cruz Bend site (AZ AA: 12:746) in terms of size and age. The exposed portion of the site covers a 5,200 m 2 area and consists of a linear right-of-way excavated along the west side of Interstate 10. The site is expected to cover the entire area between the Santa Cruz River and the interstate and between Sweetwater and Ruthrauff roads. The Wetlands site (AZ AA: 12:90) may be an extension of the Early Agricultural period Los Pozos site, representing an extramural activity area, or it may be a separate site. 209 Table 6.2. Radiocarbon dates from the Early Agricultural component at Los Pozos (after Gregory 1997b). Sample Number Feature Number Sample Context Material Dated a C13/C12 "C Age B-88139 825.00 burned structural elements grass stems -25.2 1940 +/- 60 B-88140 866.26 intramural pit fill mesquite (?) seed -21.4 1980 +/- 60 B-88141 327.01 intramural pit fill mesquite pod -24.2 2020 +/- 50 B-91143 318.01 intramural pit fill* maize cupule -7.5 2050+/- 50 B-95631 815.00 pit structure fill* maize cupule -11.6 2060 +/- 80 B-91146 389.00 pit structure fill* maize cupule -8.3 2090 +/- 60 B-95632 861.01 intramural pit fill* maize cupule -11.3 2090 +/- 80 B-91145 333.01 intramural pit fill* maize cupule -11.6 2110 +/- 50 B-95630 840.00 pit structure fill* maize kernel -10.7 2110 +/- 80 B-91141 305.01 intramural pit fill* maize cupule -8.2 2120 +/- 60 B-91140 898.00 pit structure fill* maize cupule -2.5 2140 +/- 60 B-91148 417.00 pit structure fill* maize cupule B-88142 407.00 burned structural elements grass stems B-91142 337.14 intramural pit fill* B-91149 812.23 B-95629 -9.6 2140 +/- 50 -17.2 2150 +/- 50 maize cupule -7.6 2150 +/- 50 intramural pit fill* maize cupule -6.7 2150 +/- 60 819.00 pit structure fill* maize cupule -11.0 2150 +/- 80 B-91144 355.10 intramural pit fill* maize cupule -8.2 2170 +/- 60 B-95628 352.01 intramural pit fill* maize cupule -11.3 2190 +/- 80 B-91147 416.01 intramural pit fill* maize cupule -8.0 2240 +/- 60 a All samples dated by AMS analysis. * Material recovered from flotation samples. 210 Wetlands Site First identified during a 1973 survey (Fritz 1974), an undisturbed, 50-m-long, 18cm-thick archaeological deposit was found at a depth of about 40 cm below the disturbed ground surface during later subsurface testing (Kinkade and Fritz 1975). Artifacts in this deposit included ground stone metate fragments; flaked stone projectile points (including a Late Archaic point), cores, a drill, plain ware pottery sherds, and decorated sherds identified as Tanque Verde Red-on-brown, Gila Polychrome, and Papago red ware. Segments of possible prehistoric canals were also found at a depth of about 40 cm. A historic canal and the adobe foundations of a historic homestead occupied between 1911 and 1913 were also found. The northern part of the site lay in plowed cotton fields, the middle section was disturbed by the excavation of a wastewater canal, and the south end had been extensively graded for irrigation. Along a high berm covering the sewage pipe, charred human bones, lithic debitage, shell artifacts, sherds, heat fractured rocks, and ashy soil were found during a 1976 survey (Betancourt 1978). Archaeological testing was conducted in 1986 by Desert Archaeology, Inc. (then the Tucson branch of the Institute for American Research). Six prehistoric pits filled with fire-cracked rocks, charred wood, stone flakes, and plain ware sherds were found between .5 and 1 m below the surface (Bernard-Shaw 1986). Two historic canals were also found. Recent excavation by the author (Freeman 1997) revealed a cluster of Early Agricultural pits in a formal extramural area (Figure 6.2). Near the center of the cluster was a large ground stone cache containing two manos, three metates, two 211 S.,'771 ;NG BASIN 52 5' 77 7 INC: SiN SZ limit of mechanical stripping POLISHING BASIN PI SWEETWATER ROAD PROPOSED RECHARGE/WETLANDS AREA EXPLANATION . :=.— RESULTS OF ARCHAEOLOGICAL EXCAVATIONS LOCUS 2 Oackhoe trench pithouse or pit structure 0 pit • burial AZ AA:12: 90 (ASA4) o o 25 100 50 ri 200 ft Computer cartography by CEO—MAP. Inc. 1996 Figure 6.2. Map of Wetlands site archaeological excavations. Burials are indicated by solid black dots. 212 pestles, a stone ball, and a netherstone (Adams 1997). The cache may have been left and intended for future use (Freeman 1997) or may have been part of a burial ritual, which Mabry (1996a,b) has argued is the extension of a Middle to Late Archaic pattern of cairn burials. Several pit structures, including one large, possibly ceremonial, structure were also discovered in the area. Analysis of the material remains from the site are currently under investigation, but preliminary analyses suggest that maize was a dominant part of the diet. In addition to maize, paleobotanical remains indicate that plants with the highest ubiquity values (frequency of contexts in which a taxon is present) are Chenopodium/Amaranthus (.73), Prosopis (.43), Camegia (.3), Gramineae (.3), and Echinocereus (.23)(Diehl 1997). Interpreting the significance of these ubiquity values is difficult. Both Chenopodium and Amaranthus thrive in disturbance contexts, but the seeds may also be ground into meal and consumed (Adams 1988). Furthermore, the differences in ubiquity values between plant taxa are dependent on whether plants were used and on the ways in which plants were processed, used, and disposed of. With the exception of a single date of 2790 ± 50 B.P. on a San Pedro phase pit structure, the radiocarbon dates on the pit features cluster between 2400 and 2500 B.P. (Table 6.3). Radiocarbon dates from the Wetlands site place it at the early portion of the Cienega phase, suggesting that it may indeed be a separate site from the Los Pozos component. The pit cluster was later reused as a cemetery during the Early Agricultural period. 213 Si S•1 ," _, •-•I 1.-. 1n1 q o. CD .—. ,—. MS CI 8 8 8en c'' r-: .0 S.—. en .--. Si 0 .... .—n c . C. o C) CI) ›-. 0:1 c c cl) o c:D E E E CO CO 0 I-Q N d dd d d cd d tri tri tri tri tri v-) +I +i ±I g cg al) cg cv ---. Ln ir tri tri ..ct- -1.1- r-- Nsi N N si si N tr) 8 8 ON ON -0 0C 0 v-i tri tri +1 +1 +1 8 en SI 0 v.-, 0 0 Cg S... Cm, •Zr si Nsi Si SI 00 a. 3 1 c,'s Cs N ON C.° S. .... .—. ....enICI 0 v-) ?,' 00 sr? .17:5 I I 0 tri 0 tri "8 0? ck.3 o 0. O ô 4". L,I . .., = 8 8 cd tri tri cg cg 00 •--. LO 00 tri tri si si C•••• en 7 0 Ire cv 0 tri 0 -zi- a.) a.) OP P:1 cg -ci- .1- DI- si C•4 tri toS oo Ven Ri a) Ir) g Cg,cg, enen Si SI Si SI SI SI OP 0 Cg SI § § § § § § gel en d .2 +1 N en At a.) -c-zC a.) 6. ..N4 po cn , .. co) CO co) • -. Ci3CI ECl) `.'4CL ct) E ' CO ors c. o tn d = o C.1 cr) >-. CCI cf, "0 L•• " ci.) a.) a) cl) ti.) ..- ..4 0.) cn••3 c co, Cf) co, >, >-% con - •-• CO --6. CO N 3 4". en vz, - ,,, ',.:1 6 o c SI SI Cl SI SI a) 0:1 a.) Ri <0 CL) CU PP PC1 PO IY1 SI 214 METHODS Two large exposures of middle to late Holocene alluvium (Qt2) were excavated on the west and east sides of the Interstate 10 corridor. The east-side excavation exposed the Middle Archaic archaeological component. Trenches were placed adjacent to the stripped area in order to expose additional geologic and archaeological features which were present beneath the known Middle Archaic component. Sketch profiles were drawn of all trench walls as well as the vertical walls of the stripped area (Figure 6.3, see Appendix B for descriptions). The west-side stratigraphic trench was excavated in three tiers to expose a series of Santa Cruz River channels to a depth of 6.5 m. The vertical faces of the south side of this stratigraphic trench were mapped in detail (Figure 6.4, located in the map packet). Additional detail on bedding and other sedimentological features was derived from evidence in the east and north walls of the stratigraphic pit and included in field notes. Geologic features were described in detail using a combination of USDA and Folk classification systems (Folk 1954, 1974; Soil Conservation Service 1951; 1975). Descriptions for Figure 6.4 are found in Appendix C. Radiocarbon samples were recovered and assayed where possible. ALLUVIAL STRATIGRAPHY Excavation of two areas (Figure 6.1, see also Figure 1.2) provided the opportunity to observe geologic features within two separate contexts. The east side excavation provided a deep exposure of alluvial sediments forming the prehistoric floodplain of the Santa Cruz River. Archaeological materials dating from the Middle Archaic Page 215 TRENCH 2 TRENCH 33 TRENCH 13 MASL MBD 1.00 690 2.00 689 3.00 688 burned vegetation (similar to F918) 4.00 687 trench bottom Scale: 1=3 meters Vertical exaggeration — 2X 5.00 686 Cross—sections perpendicular to Trench 12 are schematic and not in true perspective. They do, however, accurately reflect the stratigraphic relationships revealed in the respective exposures. trench bottom Surveying and digital cartography by GEO—MAP, Inc., Tucson, Arizona 1997 N 1060 N 1050 N 1090 N 1080 N 1070 CROSS-SECTION KEY un excavated o .171 Tench N1050 E400 N1060 E400 N1090 E390 060 E390 N1070 E390 N1060 E390 N1050 E390 Ni N1100 E390 disturbance drainage trench clay sand unexcavated fine gravel 1111111111111L medium gravel 12 N1070 E400 NI OBO II" MI11111111 E400 burned clay poorly developed paleosol PLAN VIEW Figure 6.3 Profile of east side excavation Ift well developed paleosol 216 through the Early Agricultural periods were found within these sediments. The west side stratigraphic trench provided a record of several former channels and channel margins of the Santa Cruz River, dating from the Middle Archaic through the historic periods. The two areas represent different facies of the same sequence, and used in tandem they can reveal much about the nature of prehistoric channel change. West Side Stratigraphie Trench The 6.5 m deep west side stratigraphic trench provided an exposure of several incisions and minor channels of the Santa Cruz River (Figure 6.4). Although not all channel incisions were as severe as the historic incision of the river, the alluvial stratigraphy provides a record of severe erosion as well as a general trend toward net aggradation during the Holocene. The generalized stratigraphy of the west side stratigraphic trench is represented in Figure 6.5. Unit I Sedimentation within unit I represents an upwardly fining sequence of alluvial materials beginning with sands and gravels deposited in channel bars and the former stream thalweg to a series of fine-grained silt and clay bands, representing overbank and slackwater deposits associated with a shift in the channel (unit 2). Subunit Ta (Figure 6.5) is comprised predominantly of small to medium cobble gravel and medium- to coarse-grained sands exhibiting features typical of channel bar deposits (Figure 6.4, deposits 1-5) and consistent with the modem bedload of the Santa Cruz 217 fa] 218 River. The basal portion of this subunit was probably entrained during flood events. As the channel aggraded, the river could no longer transport the heavier deposits. As it did so, it accumulated and stored these coarse deposits in channel bars. Although the base of this channel was not recovered in excavation of the stratigraphic pit, this subunit is probably close to the base of the channel. A radiocarbon date of 3990 ± 60 B.P. (CAMS 33961, Table 6.4) was obtained from a fragment of charcoal within the upper portion of this subunit. This charcoal may have been washed in from the Middle Archaic surface, presumably located to the east and possibly recognizable in the bottom of trenches 90-92. Table 6.4. Additional radiocarbon dates. Unit/ Deposit (5 13 C Radiocarbon Age CAMS-33961 Unit Ia-top channel alluvium wood charcoal -25.0 %o 3990±60t CAMS-34923 Deposit 510 overbank alluvium maize cupule -10.0 %o 4050±50* CAMS-34924 Deposit 506 Feature 901 maize cupule -10.0 %o I990±50* Sample No. Sample Context Material Dated Samples prepared and age estimates provided by INSTAAR-Laboratory for Radiocarbon Research, University of Colorado. Measured radiocarbon content run at Lawrence Livermore Laboratory for AMS Research. tSample contained sufficient mass to measure 4513C. These gasses are currently being run. *Sample did not contain enough mass to measure . The (5 I3C reported for this sample is based on an estimated value of approximately -10 per mil, typical of the values expressed by maize in other contexts. Subunit lb is comprised of a few coarse-grained materials, but predominantly finer-grained deposits (Figure 6.4, deposits 6-8). This subunit exhibits features typical of the upper flow regime. This subunit is followed by deposition of very fine-grained 219 silts and clay (subunit Ic) typical of overbank or slackwater sedimentation (Figure 6.4, deposits 9a-9d). The channel appears to have incised to a depth of at least 3 m below the former floodplain and shifted slightly to the west during and after deposition of this subunit. An inceptisol is formed on the upper boundary of this subunit and may represent a grassy swale present near the thalweg. A more organic surface horizon is missing at the upper boundary of this deposit, indicating that portions of the floodplain sediments near the former channel were removed by erosion. Unit II Sedimentation within unit H includes both channel and a floodplain facies, which can be correlated laterally across the profile. The bottommost portion of this channel was not reached by backhoe excavation, but appears to have incised at least 3 m below the former floodplain. Apart from the lowermost gravel, subunit Ha (channel fill) appears to represent the upper portion of the flow regime. Sand and gravel deposits in subunit Ha are similar to deposits at the base of subunit Ia; however, the gravel is finer-grained, composed of generally small cobble to pebble gravel and indicating that the stream experienced diminishing competence over time. Sediments in the upper portion of subunit Ha vary from very coarse to very fine. The channel, although truncated by subsequent periods of erosion, appears to be parabolic and is wider than it is deep. The floodplain facies of subunit Ha is comprised of a massive medium sand with some weak horizontal laminae (Figure 6.4, deposit 10). This facies represents an 220 episode of overbank flooding and is capped by a very thin carbonate band which can be traced into the channel facies (Figure 6.4, deposit 11). Subunit IIb is comprised of a fine- to medium-grained finely-laminated sand (Figure 6.4, deposit 12). This deposit contains laminae of magnetite (placer deposits) and displays very fine low-angle trough cross bedding. The deposit represents a minor channel fill and overbank deposits. An entisol is formed at the top of the deposit, indicated by bioturbation and some carbonate accumulation. The boundary between unit II and overlying unit III is marked by a slight depositional hiatus (diastem) and a very minor erosional unconformity. Unit III Unit III is comprised of a series of overbank and slackwater deposits which fill a minor channel formed at the unit II/III boundary. Subunit IIIa is comprised of a massive silt with strong carbonate accumulation (Figure 6.4, deposit 13). A depositional hiatus between subunits Ma and IIIb and the accumulation of carbonate filaments may be pedogenic, but could also be derived from groundwater. This hiatus is followed by deposition of very fine-grained silt and clay bands of subunit IIIb. These bands are found both in channel fill and overbank facies (Figure 6.4, deposits 14 and 15). A radiocarbon sample was recovered from the base of deposit 14 (subunit IIIb) but not assayed. On the basis of correlation with trenches to the east and south of the stratigraphic pit, it is probable that unit III is the Early Agricultural period floodplain. Minor 221 episodes of soil formation may have been continuous throughout this deposit, and have resulted in the translocation of younger sediments into the underlying silt. A very abrupt non-depositional or possibly erosional unconformity forms the boundary between subunits IIIb and Inc. Subunit IIIc is again comprised of a series of silt and clay bands (Figure 6.4, deposit 16) followed by a single, massive silt (Figure 6.4, deposit 17). These two deposits probably represent a soil formed during the Early Agricultural period occupation of the Los Pozos site. The erosional unconformity at the top of this sequence is marked by severe channel incision. The channel filled by unit III sediments is clearly much shallower than previous channels and also parabolic in form. Deposits within the channel facies of unit III are as fine-grained as their floodplain components and represent a clear change in streamflow conditions. Incision at the upper boundary of unit II was likely less than 1 m in depth and is followed by 2 m of aggrading sediments. After aggradation of the unit III sediments, the channel experiences an incision greater than any in its Holocene history and probably only equaled by incision during the late nineteenth century. Unit IV Though the bottom of the channel was not reached, incision following unit III was probably at least 4 m below the former floodplain. Evidence for a vertical-walled arroyo is present near the bottom of the exposed channel, as is evidence for mass-wasting of the arroyo walls (Figure 6.4, subunit IVa, deposit 20d). Gleyed well-sorted fine sands surrounded by concentric, circular clay rings at the bottom of 222 the stratigraphic pit and near the erosional boundary may mark the presence of a spring conduit near the base of the channel. Soft sediment deformation characterizes the sediments overlying the spring and may have been caused by sapping at the bank margin. Subunit IVb is comprised of channel bar and fill deposits. The lowermost portion of this subunit is comprised of small- to medium-sized cobble gravels. The uppermost gravels are imbricated. This unit is overlain by a series of sands, silts, and clays. A floodplain facies of this deposit (Figure 6.4, subunit IVc, deposit 20) is characterized by bands of silt and clay. The uppermost clay was very dark grayish brown in color and contained numerous carbonate filaments. These characteristics are typical of an organic A-horizon, and may be the result of a high water table or slackwater deposition in the channel thalweg and subsequent soil formation. Abundant charcoal is present in the floodplain facies as well as parts of the slumpblock. Unit V Unit V is characterized by several minor channels and channel fill episodes. These channels were cut no more than 1.5 m below the former floodplain. The channels are filled predominantly with coarse- to medium-grained sands and occasional silt bands. A floodplain facies is recognizable in a small area near the bank margin, but is incorporated into the plowzone to the east. 223 Unit VI Unit VI is again comprised of minor channels and channel fill episodes. Again the channels were cut no more than 1.5 m below the former floodplain. These channels are also filled with coarse to medium sands and silt bands. The floodplain of unit VI component is incorporated into the historic plowzone. East Side Excavation In contrast to the west-side stratigraphic pit, the east side excavation displays a long sequence of overbank and slackwater deposits of the Santa Cruz River floodplain (Figure 6.3). For the most part, erosional boundaries are recognizable only by documentation of a weathered soil profile and correlation with known erosional deposits in the west side stratigraphic pit. Less time was available to document the northern portion of the excavated deposits, some of which appear to be older than the southern portion of the excavation. Deposits 533-537 Deposits 533-537 are located in the northern portion of the trench. This column of sediments received less geologic attention than other parts of the excavation, due to time constraints. The lowest deposit is a gravel (Figure 6.3, deposit 537). This gravel did not exhibit the carbonate accumulation typical of Pleistocene gravels documented elsewhere in the floodplain, but did display a reddish hue and accumulation of secondary clays typical of other Pleistocene soils. It is possible, therefore, that these 224 gravels are late Pleistocene or early Holocene in age. Overlying the gravels were a series of clays and silts, typical of overbank or slackwater deposits. Oxidized mineral stains on these sediments are probably a result of capillary movement of groundwater. These sediments are also probably early Holocene in age. Channel deposit (520s) and Deposits 531-532 Deposit 533 is truncated by a channel filled predominantly with coarse- to medium-grained sands and gravels. An overbank facies of this channel appears to conformably overlie the previously discussed deposits. This channel is then cut by a second minor channel (Figure 6.3, deposit 517), representing a braided middle Holocene stream. Deposits 514-517 Deposits 514-517 are located in the southern portion of the excavation area. The majority of geologic attention on the east side was focused on two soil profiles in this area, one of which forms a complete stratigraphic section with overlying deposits 501-509. Deposits 514-517 represent the fill of a low middle Holocene channel. The lowermost portion of the channel fill is comprised of pale reddish brown matrix of coarse sand surrounding very small- to medium-sized cobble gravels (Figure 6.3, deposit 517). This deposit is overlain by a pale brown gravelly loam (Figure 6.3, deposit 516). The pale brown gravelly loam is typical of slopewash deposits formed by reworked bajada sediments. These slopewash sediments are overlain by a grayish 225 brown sandy clay with some fine pebble gravel (Figure 6.3, deposit 515), which is in turn overlain by a series of silts and clays (Figure 6.3, deposit 514). In places deposit 514 is sandier. Deposits 514-517 appear to be the fill of a low middle Holocene channel and shift to floodplain overbank and slackwater deposition. As this near-margin channel was filled, the thalweg shifted somewhere to the west of this profile and this location became part of the floodplain. Soil formation can be observed at the upper boundary of deposit 514, where color is a dark grayish brown and a moderate number of carbonate filaments are present. The accumulation of oxides, carbonates, and very dark grayish brown color in subunit 514c is likely a result of close proximity to groundwater prehistorically during initiation of this sedimentary sequence. Descriptions for these units in nearby trenches appear to represent the floodplain facies of this sequence (G. Huckleberry field notes). Radiocarbon dates on cultural feature 917 (a feature of unknown use cut from deposit 515) support a middle Holocene age for the deposit (Table 6.1, Beta-85204 and Beta-88147). A single radiocarbon date on an in situ band of burned vegetation at the upper boundary of deposit 514 further supports this middle Holocene age (Table 6.1, Beta-88143) around 4500 B.P. (corrected, uncalibrated). The average of these three dates is 4471 ± 50 B.P. The upper portion of this sequence of deposits may extend to the northern portion of the trench (represented by deposit 531). 226 Deposits 512-513 Deposits 512-513 are located in the southern portion of the excavation area and may be compressed into deposit 531 in the northern portion of the excavation. These sediments represent a sequence of overbank siltation, followed by additional overbank or slackwater deposition and soil formation. The same basic sedimentation pattern is repeated throughout the remainder of the sequence. Deposits 510-511 Deposits 510-511 are located in the southern portion of the excavation area and are part of this repeated pattern of sedimentation occurring in previous deposits. These sediments also extend to the northern portion of the excavation. Deposits 510-511 represent a sequence of overbank siltation, followed by additional overbank or slackwater deposition and soil formation. Deposit 510 was the focus of most of the archaeological excavation on the east side of I-10 and produced numerous flaked stone artifacts as well as some animal bone. Four radiocarbon dates (Beta-81329, -88145, -95633, and -95634) were derived from the dispersed charcoal in natural stratum (Figure 6.3, deposit 510) from which Middle Archaic archaeological resources were recovered, and three additional dates were recovered from charcoal in cultural features within this natural stratum (Beta-88144, -88148, and -95635). The six youngest of these dates can be combined for an average age of 3829 ± 35 B.P. (see Table 6.1, corrected, uncalibrated) and confirm a middle Holocene age for these deposits. A 227 single corn cupule recovered from this natural stratum yielded an age estimate of 4050 ± 50 B.P. (CAMS-34293, Table 6.4). Deposit 510 is comprised of a series of laminated sedimentary deposits ranging from a very dark grayish brown clay to grayish brown to light grayish brown silty clay in the center of the deposit. The upper portion is again darker, indicating greater organic content near the surface of the soil. Organic content at the base and top of the soil and few, very fine oxides indicate close proximity to groundwater. Many fine carbonate filaments and small nodules are also present in the soil. The upper surface of the deposit has been eroded and weathered. The deposit is similar to cienega soils, but appears to have received overbank sedimentation throughout its depositional history. Deposits 502-509 These deposits were defined predominantly during archaeological testing and observed in a single profile during excavation. Deposits 503 through 509 represent the continuous, rapid sequence of overbank and slackwater sedimentation typical of floodplain deposits. Although these deposits are roughly 1 m in total thickness, there is little evidence for pedogenesis within this series of deposits and no strong erosional hiatuses, apart from the hiatus at the top of deposit 510. A coarser-grained very fine sand (Figure 6.3, deposit 507) within the sequence displayed low-angle trough cross-bedding, typical of sedimentary structure that has been unmodified by pedogenesis. Although this type of structure can be found in older sediments, the thin 228 nature of each deposit should exhibit pedogenetic activity if the deposit had been continuously building up since ca. 3800 B.P. The weak, medium, angular blocky structure displayed by the clayier sediments is probably due to the dry content rather than pedogenetic alteration. Some weak to moderate carbonate accumulation is present in deposits 505 through 509. A single radiocarbon date on a corn cupule recovered from an archaeological feature in unit 506, yielded an age estimate of 1990 ± 50 B.P. (CAMS-34294, Table 6.4), basically contemporaneous with the Early Agricultural component on the west side of the interstate, averaging 2097 ± 15 B.P. (Gregory and Baar, 1997; Table 6.2). This average represents 18 of 19 dates from the Early Agricultural period of the site; a single date of 2240 ± 60 B.P. from the Early Agricultural period component was not averageable. Deposit 502 (Figure 6.3) represents a more slowly accumulating cienega-like deposit. This deposit is darker in color than most of the clayier sediments in the profile suggesting a higher percentage of organic matter. The lower portion of the deposit is even darker in color than the upper portion of the deposit. A Hohokam canal was found excavated into this deposit; however, Gregory and Baar (1997) correlate this deposit with the Early Agricultural period occupation excavated on the west side of Interstate 10. Deposit 501 Deposit 501 (Figure 6.3) represents the accumulation of additional floodplain deposits, most of which have been disturbed by historic plowing or other historic and 229 modern processes. Modern construction episodes are highly visible in parts of this deposit near the interstate. GEOCHRONOLOGY AND CORRELATION Correlation between deposits within and outside the Los Pozos site are necessary to assess the geomorphology of the site area, the nature of stream dynamics and the environment in which prehistoric people lived. Intersite correlation provides the necessary background to examine the relationship between the stream and the floodplain at a single locality, providing the environmental context of the project. Intrasite correlation provides a broader perspective of how those individual environments interrelate and whether the river responded to changes in the same manner in each reach. Intrasite Geochronology and Correlation The lateral distribution of the east side excavation and the west side stratigraphic pit across the project area must be correlated before attempting intersite correlation. Because of the nature of facies relationships, certain deposits in the channel fills cannot be directly related to a corresponding floodplain component. Floodplain deposits typically demonstrate subtle changes across wide areas. Single bands of clay will split or several will combine together to form a single band. Soils can be eroded in one area and perfectly preserved in another. Because depositional facies of the Santa Cruz River will exhibit varying lithologie and pedogenetic characteristics and 230 because floodplain deposits exhibit similar lithologic and pedogenetic characteristics, chronostratigraphic correlation is emphasized. Secure stratigraphic correlation can only be accomplished with numerous radiometric age estimates or diagnostic cultural horizons if strata cannot be traced from one site to another. Deposits with distinct pedogenic or sedimentary features can sometimes be used to support these correlations. The series of clay bands at the upper portion of unit I (Figure 6.5, subunit Ic) in the west side stratigraphic pit are probably slightly younger than deposit 510 (east side) and probably represent a late middle Holocene or early late Holocene deposits. A radiocarbon sample on dispersed charcoal from sands (upper subunit Ta) underlying these clay and silt bands produced an age estimate of 3990 ± 60 B.P. (CAMS-33961) and probably represents charcoal reworked from the middle Holocene deposit (deposit 510) exposed in trenches to the east of the west side stratigraphic pit (Figure 6.4). Under this scheme, unit I represents a migration of the middle Holocene channel to the west and incision after the middle Holocene occupation of the project area. A series of deposits on the floodplain extending from the Early Agricultural period excavation area to the west-side stratigraphic pit appears to correlate well with the upper portion of unit III. Although radiometric evidence has not been used, archaeological features in these floodplain deposits appear to satisfy independent correlation between deposits of similar characteristics. Because subtle depositional changes can occur over the short distance between trenches, secure correlation between the excavated Early Agricultural component and deposits in the east-side excavation is more difficult. Characteristics of the floodplain depositional regime during the Early 231 Agricultural occupation are found throughout deposits 502-509. The entire sequence of overbank floodplain deposits (502-509) may represent the same sequence seen in subunits Illb through Hid in the west-side stratigraphic pit. Alternately, floodplain deposition during the Early Agricultural period may begin somewhere within the sequence of sediments represented by deposits 502-509. The former hypothesis is supported by radiometric evidence in unit 506. Aggradation rates, estimated below, can be used as a rough gauge to support this hypothesis and the overall depositional history of the floodplain. Ag gradation Rates and the Estimated Age of Paleosols Rough aggradation rates for similar channels have been calculated by Parker (1995) based on cross sections of the San Xavier reach (Waters 1988). By applying these aggradation rates to channels in the west-side stratigraphic pit, estimates of the ages of deposits in the west-side stratigraphic pit can be proposed. Units III and IV of Waters' (1988) report are the closest in age (although younger), size, and sedimentary content to the middle to late Holocene channels represented in the west-side stratigraphic pit. Estimates calculated for these channels were .68 and .4 cm/yr respectively. Following the curvature of the channel, where present, a bottom depth of each channel in the west-side stratigraphic pit was estimated. The oldest channel represented was assumed to be close to its bottom depth, based on the size of cobble clasts in the southeastern corner of the profile. A beginning date of approximately 3800 B.P. was attributed to the bottom of unit I based on charcoal found within the 232 unit and probably reworked from the adjacent bank deposit. The 3800 B.P. estimated date is probably older than unit I and does not reflect intervening time between deposition of the adjacent bank deposit and erosion of the paleochannel filled by unit I sediments. However, the charcoal may have been produced (by either natural or cultural agents) on that surface any time between final deposition of the adjacent bank deposit and erosion of the paleochannel. The 3800 B.P. estimate is, therefore, a very rough estimate and probably the maximum possible age for the base of the oldest paleochannel. The upper surface of unit III was attributed to an age estimate of 2000 B.P. based on the average age of cultural materials found in the deposit as it extends to the excavated site area. Two methods were used to calculate the age estimates in Table 6.5. Beginning with a date of 3800 B.P. on the bottom of unit I, estimated ages for the top of each aggradational episode were calculated by subtracting the number of estimated years of aggradation from the previous age. The slower aggradation rate provides the minimum top age of the deposit and the faster aggradation rate provides the maximum. Unit IV deposit ages were calculated using this method and a starting date of ca. 2000 B.P. Retrograded estimates were calculated using the opposite method. Beginning with a date of 2000 B.P. on the top of unit III, estimated ages for the bottom of each incision were calculated by adding the number of estimated years of aggradation from the previous age. The slower aggradation rate provides the maximum age of the bottom of each deposit and the faster aggradation rate provides a minimum age. Neither hiatuses, which are obviously present in the profile, nor 233 erosion of paleosols are accounted for in the age estimates. Also, it is important to remember that the channel deposits represent only the last flow event and may have experienced several decades of scour and fill prior to final deposition. Finally, channel geometry and sedimentation have a great impact on the rate of aggradation and probably have had the most significant impact on the aggradation rate used for unit III. As a consequence, these estimated dates should be used only as a rough scale from which to understand the sequence of deposition and develop testable models of prehistoric floodplain use. Table 6.5. Estimated age of stratigraphic units in the west side stratigraphic pit at Los Pozos (AZ AA:12:91). Bottom min. age Top max. age Bottom Top Rate max. age min. age (cm/yr.) (est.rcybp) (est. rcybp) cm/yr. (est.rcybp) (est.rcybp) 3.5 0.68 3800/3325 3285 0.4 4250 2925 H 3.5 0.68 2810 2770 0.4 3375 2050 III 2 0.68 2295 2475 0.4 2500 2000/1550 IV 4.5 0.68 1340 0.4 Estimated Unit no. aggradation (m) Rate 875 Given the age of the wood charcoal in subunit Ia, the maximum age estimates are probably closer to the true age of the paleosol within this profile. Using these age estimates, the series of clay bands within subunit Ic probably date to ca. 3,200-3,300 years ago and the clay bands within subunits Illb-d probably date between 2,700-2,500 and 2,000 years ago. 234 Calculation of aggradation rates for deposits in the east-side stratigraphic pit would be unfounded, as deposition within a floodplain facies is dependent not only on streamflow conditions but also on position of the channel relative to the documented floodplain sequence and the differential potential for preservation of floodplain sediments across the site. Larger floods and floods of channels close to the examined profile would deposit greater concentrations of sediment, while floods that produced less extensive overbank flows and in channels at greater distance from the profile may not deposit sediments at the examined profile. The absence of strong hiatuses between deposits 509 and 502 indicate that sedimentation within these units was relatively continuous and rapid. Intersite Geochronology and Correlation Intersite correlation is usually quite difficult. The two most reliable methods of correlating deposits across such a wide expanse are radiocarbon dating and associated cultural materials. Using excavated and natural exposures, Haynes and Huckell (1986) have attempted to correlate Holocene deposits of the Santa Cruz River from Pima Mine Road to Ina Road. Most of their data focused on the area between Airport Wash (just north of San Xavier) and Pima Mine Road. They relied predominantly on radiocarbon dating and associated archaeological materials to establish correlations between deposits separated by lengthy distances. Later, Huckell (1996) produced an alluvial stratigraphy for deposits at the nearby Santa Cruz Bend site (AZ AA: 12:746, ASM). For conformity with other excavations, the designations used by Haynes and 235 Huckell (1986) will be used here; however, some discussion of the relationship of this excavation to excavations conducted in other parts of the Interstate 10 corridor will be addressed using Huckell's (1996) report as well as additional data presented in chapter 5. Based on existing radiocarbon dates (Haynes and Huckell 1986), sediments in deposit 514 of the east side excavation appear to be roughly equivalent in age to the upper portion of Haynes and Huckell's unit B 1 . Sediments in deposit 510 (Figure 6.3) are roughly equivalent in age to either the upper portion of unit B / or the lower portion of B2. Deposits 503-509 are either equivalent in age to the upper portion of unit B2 or the lower part of C I . Deposit 502 is roughly equivalent in age to the upper portion of unit C I or unit C2. The age estimates for older deposits in the east side excavation area (the 520 and 530 series deposits) are based on relationships to younger deposits and conjecture from previous studies. Waters (1987, 1988) notes that the Santa Cruz River was a shallow, braided stream prior to 8,000 years ago. He observes a hiatus between 7000 and 2500 B.P., however, evidence from Haynes and Huckell (1986) suggests that the hiatus only lasted to roughly 4500 B.P. Waters (personal communication 1996) has suggested that the discrepancy between his data and Haynes and Huckell (1986) may be due to complex response. He notes that the ca. 4500 B.P. dates from Haynes and Huckell's work are predominantly from slopewash deposits, and surmises that, in the San Xavier area, the river was incising while at the same time at Ina Road, those entrained sediments were being deposited and stored. 237 1989; Monger 1995), from palynology (Hall 1985), and from packrat middens (Spaulding 1991) support the interpretation of this period as one of erosion and drought conditions. Around 5500 years ago, the Santa Cruz River began to aggrade, forming at first a shallow, braided stream. The presence of abundant iron and manganese oxides/hydroxides in units 514 through 517 of the east side excavation indicates a resurgence of groundwater, absent in the 520 series deposits. The river was at this time located closer to the margin of the Jaynes terrace (Qt3) and at least one of the braids of that channel was present where excavations on the east side of the present interstate were undertaken (520 deposits). Evidence appears to indicate a water table relatively close to the surface. Again, in both the San Xavier and Los Pozos areas slopewash forms at least a portion of this component. Based on the known patterns of channel change in the Santa Cruz River, instrinsic factors do not appear to account for channel changes around 5,000-4,500 years ago. An extrinsic factor, such as climate change, seems a more likely cause of the change in stream competence that set the Santa Cruz River on an aggradational trend. Because frontal and tropical moisture produces large flood discharges in the Santa Catalina Mountains with enough energy to continue downstream (see Chapter 3), tributary valleys of the Santa Cruz (e.g., Rillito Creek, Canada del Oro) and their points of confluence with the Santa Cruz were most likely the first areas affected by changes in climate. The resulting geomorphic changes in the relationship between the Santa Cruz River and its tributary valleys signal that climatic change. The proximity of Los Pozos to Rillito Creek 238 enhanced the ability of this reach to store sediment. As it aggraded, the thalweg, which was a single narrow and shallow channel, shifted to the west. The near absence of deposition during this time period in other reaches of the Santa Cruz River is probably due to the lag in response to changing climatic conditions. Overbank flooding from each successive channel deposited a series of fine-grained silts and clays on the floodplain. The first of these fine-grained materials was deposited around 4,000 years ago and formed the context for the Middle Archaic occupation on the east side of the interstate. Previous channels may be indicated by the slope of overlying deposits conforming to depressions underlying them, possibly exhibited in trenches intermediate between the east side excavation and the west side stratigraphic pit (Figure 6.4). Shortly after the Middle Archaic period occupation, the stream underwent incision and the channel shifted to the west. Unfortunately, no datable samples were found at the upper boundary of this aggradational episode, making it impossible to predict the possible age of incision. This post-Middle Archaic incision was probably 3-3.5 m in depth (unit I). After channel incision, the aggrading stream became increasingly less competent until series of fine-grained silts and clays were deposited on the floodplain (unit Ic). This may represent some kind of early Late Archaic/Early Agricultural deposit and may have again been the setting of floodplain activities. A second incision (this time to a depth of at least 3 m) followed, and the channel again shifted westward. During these two periods of incision, the stream was underfit to its channel. Floods in the channel rarely exceeded bankfull discharge and were not 239 extensive across the floodplain. However, particularly rich or attractive habitats may have been locally available to populations utilizing the floodplain during short periods at the end of each aggradational episode (represented by unit Ic in the west-side stratigraphic pit). One of these aggradational periods, followed by a minor depositional hiatus, appears to form the context for the florescence of the Early Agricultural period (probably between 2500 and 2000 B.P., see Gregory 1997). During this period, the smaller size and depth of the channel allowed numerous extensive episodes of overbank flooding. The floodplain was stable for a relatively lengthy period in which early farmers comfortably made the transition to a more sedentary lifestyle (Gregory 1997). A third, and this time the deepest (> 4 m) incision of the Santa Cruz River followed this Early Agricultural period fluorescence. During this time, early farmers may have accessed the water table by constructing wells like those found at both the Los Pozos (Gregory 1997) and Sweetwater Wetlands sites (Freeman 1997). The channel formed a steep-sided, almost vertical-walled arroyo, much like that created by the historic incision of the river. This channel also aggraded over a several hundred year period, during which time the river was again underfit to its channel and an even less reliable source of water. Farmers in the Tucson Basin during this time must have adapted new methods of farming, including ak-chin irrigation on the bajadas. Aggradation of this underfit channel probably occurred until sometime between 1,340 and 875 estimated years ago (or roughly A.D. 600 to 1100) and is followed by minor 240 channel shifts until the historic period incision, which caused the channel to cut a continuous trench to a depth of 10 m in some reaches. Environmental Implications of Channel Changes The trend toward net aggradation demonstrated by river alluvium after at least 4,500 years ago appears to be evidence for a long-term change in climate. The increased effective precipitation noted after 5,500 years ago has been attributed to an increase in monsoonal-type precipitation (Markgraf 1985). A high frequency of this type of storm activity would be consistent with a trend toward aggradation. During this aggradational trend of at least 2,500 years, episodes of severe degradation or nondeposition also occurred. These degradational episodes are discontinuous through the stream system and may have resulted from either short-term changes in the frequency of different types of storms or response to intrinsic geomorphic factors. Short-term changes in the types of storms can create perturbations in the geomorphic system that result in intrinsic responses. For instance, the input or loss of sediment at the confluence of tributary streams (i.e., Rillito Creek or Canada del Oro), resulting from short-term changes in the frequency of tropical and frontal storms, would potentially cause a perturbation in the trunk stream (i.e., the Santa Cruz) and would require an adjustment in the behavior of the trunk stream. While the initial perturbation was caused by an extrinsic factor (climate), the trunk stream would be responding to an intrinsic factor (change in gradient). 241 An adjustment in the behavior of the reach of the Santa Cruz upstream (south) of its confluence with Rillito Creek is particularly likely due to confinement of the valley in this reach (see Chapter 3) and is highlighted by the discontinuous nature of channel change reflected in this reach. A change in gradient might result in a series of localized cutting and filling events that would migrate through this reach, but would not form a continuously entrenched arroyo. These events are probably asynchronous for at least short reaches, and are typified by incision, entrainment, and storage of sediments in different parts of the same system. During the period from 4500 to 2800 or 2500 B.P., the stream system was characterized by complex response to perturbations created at the confluence of tributary streams. In areas where sediment storage was the dominant process, local microenvironments were created. Historic records suggest that these areas might have formed local wet meadows (cienegas) or more extensive mesquite bosques and grassy swales. The discontinuous nature of these microenvironments was altered after a short period of non-deposition 2,800 to 2,500 years ago when the channel became a small shallow drainage. During the Early Agricultural period (ca. 2800 or 2500 to 2000 B.P.) sediment storage and overbank deposition were dominant processes throughout the focus reach, extending these microenvironments across a wide area. It is difficult to say what processes dominated channel change along the river after the Early Agricultural period. The west-side stratigraphic pit exposes a portion of the record for this interval, but without more detailed information about the stream 242 behavior during the periods that followed the environmental implications of channel change during this period is more sketchy. Implications for Human Settlement and Site Preservation The trend toward net aggradation created localized favorable riparian habitats that migrated upstream as each reach of the river successively responded to this change. By 4,000 years ago, a favorable localized environment was in place at the Los Pozos site for Middle Archaic populations to exploit. As the stream responded to changes in its system, the specific location of this favorable localized environment would have migrated upstream and the location utilized by prehistoric occupants of the Los Pozos site might have become less favorable for a period until it again became favorable around 3,500 to 3,300 years ago. This pattern of incision and deposition continued for approximately 1,500 years (until 2800 to 2500 B.P.), at which time a longer-term favorable environment was created. The dynamic processes of erosion and deposition continuing from the Middle Archaic into Late Archaic times created a series of localized favorable habitats during the Middle Archaic and Early Late Archaic/Early Agricultural periods. During this time, settlement of the floodplain was restricted to local environmental niches which progressively moved upstream over the course of several generations. By the florescence of the Early Agricultural period (2,500-2,000 years ago), the floodplain had attained greater stability. Though progression of these favorable niches probably still occurred, there appears to have been a significant period, during which prehistoric 243 people could rely on a favorable floodplain environment conducive to the long-term cultivation of plants. Following the Early Agricultural period farmers were likely drawn away from the river. A period of incision would have made the river a poor place for floodwater farming. The few Early Ceramic period sites found in the floodplain appear to be located slightly closer to the Pleistocene terrace. Certainly, by the Rillito phase, the floodplain was a more reliable source of water once again. SUMMARY The current model of floodplain development created by geoarchaeological stratigraphers in the Santa Cruz Valley focuses on the apparent synchroneity of geologic events (particularly incision) within the valley itself and across southern Arizona (Haynes 1968); however, both the cause of incision and the synchroneity of these events are the subject of much debate in fluvial geomorphology and climatology. Furthermore, those models cannot account for the subsistence-level decisions employed by human groups over periods of several generations. By employing Parker's (1995) model that accounts for sedimentational and hydrologic controls on the river system, the apparent synchroneity of geologic events and the subtle differences in depositional history can be documented in different parts of the river. These processes have a tremendous effect on the way in which prehistoric people utilized the floodplain environment and on what resources may have been available for them to exploit. The model also accounts for gaps in the archaeological record that 244 may be present in different reaches of the river or at different times. The Los Pozos stratigraphic record, including both a floodplain and a channel component, has been instrumental in developing this model, which can be tested in other parts of the Santa Cruz River. ENDNOTES 1. This age estimate is based on 18 of 19 radiocarbon dates from the site. A date that did not fit within the average was 2240 ± 60 B.P. 245 CHAPTER 7 THE IMPACT OF MIDDLE HOLOCENE STREAM DYNAMICS ON HUMAN SETTLEMENT, THE TRANSITION TO AGRICULTURE, AND ARCHAEOLOGICAL SITE PRESERVATION Utilizing a method that details both the alluvial stratigraphy and site-reach specific hydrologic factors influencing the Holocene floodplain record, this dissertation offers a new interpretation of the middle to late Holocene archaeological record. Site and reach specific data are critical to addressing questions beyond the level of explanation that encompasses several hundred or even thousands of years. The defining characteristics of Middle to Late Archaic transition in southern Arizona include the transition to agriculture and changes in settlement and subsistence patterns. Yet, the current application of geologic stratigraphy to this archaeological record of change does little more than convey general trends in the human use of riverine environments. New data on the Middle Archaic in the Tucson Basin, utilize a high-resolution geologic record to demonstrate the importance of this level of alluvial interpretation to understanding human use of the floodplain during this period. SUMMARY OF MIDDLE HOLOCENE STREAM DYNAMICS By utilizing a site- and reach-specific sedimentological record in this thesis, the research presented operates at a scale that is better able to address the changes occurring in human land use during the transition to agriculture in southern Arizona. The thesis employs stratigraphic records that cannot be correlated across the entire 246 Tucson portion of the Santa Cruz River, but that are specific to the area surrounding the Santa Cruz Bend and Los Pozos sites. Historic and hydrologic records indicate that the area from A-Mountain to the Rillito confluence would have responded separately to small-scale changes (possibly as a result of intrinsic geomorphic factors). Two topographic controls isolate the river, Rillito Creek and A-Mountain. A-Mountain is in a part of the floodplain that is affected by local restrictions on hydrologic processes. An impermeable stratum, created by the merging of Pleistocene terraces with hard bedrock, limited downcutting and forced groundwater to the surface, potentially creating a location favorable for sediment storage in the A-Mountain area. This narrow portion of the valley also created an area that, during more active flooding, would have transported sediments through the reach. Under a meandering stream condition, like that of the middle to late Holocene, flood events would occasionally erode the banks of the river. Another important topographic influence on the focus reach is the presence of Rillito Creek. Because the source of its drainage is the Santa Catalina Mountains, and because the Santa Catalinas receive significant flood flows during periods with a high incidence of tropical and frontal storms, subtle changes in the intensity of these storms could promote sediment storage at the Rillito confluence. Contribution to the hydrologic and sediment budgets in the area surrounding the Rillito confluence would potentially have impacts on the upstream and downstream reaches of the Santa Cruz. The internal record between A-Mountain and Rillito Creek, two locations that act as thresholds of stream behavior, is representative, therefore, of the phenomena affecting 247 this reach during its Holocene history. When the alluvial records from the study area are combined with records from the area south of A-Mountain (near San Xavier Mission) and north of the Rillito (and Canada del Oro) confluence (near Ina Road), it is obvious that there are geologic units that can be correlated across this long stretch of the river and that there are geologic units that cannot be correlated. Geologic records that can be correlated across the entire Tucson portion of the river, including trends toward net aggradation, are interpreted as the product of either significant climatic or hydrologic changes and, predominantly, the effect of extrinsic processes. Geologic records that can be correlated across shorter reaches are the product of either intrinsic geomorphic processes or localized hydrologic changes (e.g., intensive localized storms or local drops in the water table). Long-Term Climatic Changes and the Geologic Record of the Santa Cruz River Several geologic units appear to reflect long-term climatic or hydrologic changes affecting the Santa Cruz River. An apparent, rather lengthy hiatus appears to have preceded middle Holocene deposition of sediments on the Santa Cruz River. Although none of the pre-middle Holocene sediments in the study reach have been radiocarbon dated, the high degree of argillic accumulation in them suggests a late Pleistocene or early Holocene age. Holocene sedimentation in the study reach apparently begins between approximately 5,500 and 4,500 years ago. These sediments are accompanied by evidence for a relatively high water table, suggesting that the 248 trend toward net aggradation that occurs during the middle Holocene is the result of an increase in the hydrologic budget of the Santa Cruz River. This apparent wet trend probably occurs in response to a long-term climatic change as reflected in palynological and lacustrine sediments elsewhere in southern Arizona and across the Southwest. Middle Holocene sedimentation occurs in the Santa Cruz River record and is represented by Haynes and Huckell's (1986) unit B, dating between ca. 5000 and 2500 B.P. The erosional contact represented between Haynes and Huckell's (1986) units B and C (ca. 2500-2000 B.P.) is marked by an insignificant depositional hiatus in the focus reach. A shift in depositional regime resulting from a decrease in channel size and depth and causing increased floodplain sedimentation at the Los Pozos site (Figure 6.5, units Ilb and Ilia) occurs after this depositional hiatus. The small, shallow channel created during deposition of unit ilia in the Los Pozos west side stratigraphic pit would have supported large quantities of floodwater and sedimentation to the floodplain, creating a favorable setting for agricultural economies. Erosion of Haynes and Huckell's unit C appears to be quite significant, affecting most of the Tucson portion of the river. In the study reach, this period is represented by significant downcutting, creating a very deep and steep-sided arroyo. The effect of this incision is felt upstream in the San Xavier Mission area and probably reflects either climatic change or a breach of a significant geomorphic threshold. Wells found at the Los Pozos and Wetlands sites appear to indicate a drop in the availability of surface water during this period, perhaps reflecting the hydrologic 249 effects of a change in climate. Because records for the study reach are incomplete following this period, there is no basis from which to correlate post-entrenchment conditions with areas upstream and downstream of the study reach. Short-Term Climatic Changes or Small-Scale Threshold Breaches During the period between 4500 and 2800 to 2500 B.P., the study reach displays a record of channel cutting and filling events not represented elsewhere in the Santa Cruz River geologic record. Although these events could be present, but not within the confines of the profiles examined by Haynes and Huckell (1986), Stafford (1986) or Waters (1987, 1988), there is no evidence for similar changes occurring in the area around A-Mountain, suggesting that A-Mountain acted as a threshold to stream behavior. Stream behavior affected by this threshold would have created different records upstream and downstream of A-Mountain. Because these cutting and filling events are locally restricted, they represent exceeded internal thresholds. Although such thresholds could be exceeded by short-term changes in the frequency of tropical and frontal storms, they do not constitute long-term climatic change. It is possible that sediment storage occurs at the Rillito Creek confluence during short-term changes in the freqency of tropical and frontal storms. Changes in the longitudinal profile of the stream caused by sequential erosion and deposition would create localized microenvironments where sediment storage was the dominant process, creating a succession of events representing what Schumm (1973; Schumm and Parker 1973; Womack and Schumm 1977) has called 250 "complex response." Complex response refers to the lag time in the response of individual reaches of a stream channel to a perturbation in the stream system. The stream can be simultaneously incising in one reach while aggrading in another. Previously, this concept has been viewed in stark contrast to the apparent synchroneity of arroyo incision across the southern Southwest during both the historic (Bryan 1941; Cooke and Reeves 1976) and prehistoric periods (Haynes 1968). However, utilizing Parker's (1995) model, which focuses attention on the cause-and-effect relationships between climatic/geomorphic perturbations and hydrologic response, synchronous incision and asynchronous stratigraphic records can be the result of the scale of perturbations to the stream system. Slackwater and overbank deposits at the top of each aggradational column would support the growth of native grasses and other foodstuffs near the river; these plant taxa are represented in paleobotanical samples from the Los Pozos site. Overbank deposition would also be conducive to small-scale agriculture. These conditions are possible in reaches dominated by sediment storage. As the position of stored sediment moves through the focus reach, favorable microenvironments would not have been laterally extensive, and would have migrated upstream and downstream over periods of 500 to 1,000 years (the time it takes for one incised channel to aggrade completely). Over a space of 6 km this is roughly 15 m/year, and may influence the distribution of archaeological resources across the landscape. The scale at which these local microenvironments would migrate up and downstream would be imperceptible to human groups utilizing the area and is likely 251 not capable of being detected through radiocarbon dating. The small size of these local habitat areas, detectable today only by intensive stratigraphic tracing of subtle lenses of overbank deposition and probably not exceeding 25 ha in area, would be enough to influence human population or seasonality of use of the area. During this time populations moving into the area would need to rely on the surrounding bajada areas to support subsistence needs, establishing a pattern of settlement that is continued during episodes of more consistently reliable environmental conditions (the period from 2800 or 2500 to 2000 B.P.). HUMAN SETTLEMENT Chapter 2 demonstrated that Middle Archaic use of the floodplain has been underestimated and that, as a result, interpretation of Middle Archaic occupations in southern Arizona has overemphasized use of the bajada and montane environmental zones by Middle Archaic groups (Roth 1988, 1989; Huckell 1990). Preservation could certainly be part of the difficulty, but recent research in the Santa Cruz floodplain has illustrated the potential for Middle Archaic use of the area. Though occupation of the floodplain is not as apparently intensive as during the Late Archaic period, it reflects a pattern of seasonal land use that could be carried into periods of greater agricultural intensification. Though human groups may not have been able to detect subtle changes in the location of microenvironments from year-to-year, over several generations it would have affected the location of their settlements. The discontinuous nature of these 252 microenvironments during the Middle Archaic would not have supported the large settlements present during the more favorable Late Archaic period. During the Middle Archaic period, the focus reach is dominated by complex response, meaning that sediment storage, overbank deposition, and sediment entrainment migrate through the stream system. Microenvironments would be present only in the study area and not in other portions of the river dominated by sediment entrainment. Middle Archaic populations cannot be expected to have used areas where discontinuous gullying was the dominant sedimentary process. THE TRANSITION TO AGRICULTURE Over the years, a number of models have been developed to account for early agriculture in the American Southwest. In 1962, Haury proposed that maize entered the Southwest by means of a "highland corridor." At the time, the earliest ages on domesticated plant materials were maize from Bat Cave, New Mexico. Subsequent refinement of the chronology of early maize (Berry 1982, 1985; Smiley and Parry 1990; Wills 1988b) provided the impetus for additional models of the spread of agriculture. Models of the Transition to Agriculture Following on the highland theme offered by Haury (1962), Ford (1981) defined what he called the "Upper Sonoran Agricultural Complex." His revision of Haury's model would emphasize the use of these crops by hunter-gatherers living in these 253 highland regions. The agricultural crops both supplemented their diet of edible resources and provided the disturbed environment favored by wild plant resources, further enhancing these resources (Ford 1984). Minnis (1985, 1992) favored a similar low-level integration of cultivated plants into the subsistence-system of foragers. He further suggested that both opportunistic and stress-based adoption of agriculture may have been involved. Stress-based models have been adopted by a number of archaeologists (Hard 1986; Hunter-Anderson 1986; MacNeish 1992; Wills 1988a, 1988b, 1990) who argue that either population growth and/or increasing environmental uncertainty caused foragers to adopt agriculture to enhance resource predictability. Other archaeologists have proposed that agriculture arrived in the Southwest by means of immigrant populations from the south (Berry 1982; Berry and Berry 1986; Huckell 1990, 1992a). Both models suggest that environment played a significant role in the spread of agriculture from Mexico into the American Southwest. Basing his model predominantly on the excavation of a number of San Pedro-aged sites, which he views as distinctive in material culture from earlier groups, Huckell (1990, 1992a) suggested that improved environmental conditions along rivers and streams in the southern Southwest. Huckell further proposes that these immigrant populations practiced a mixed fanning-foraging economy. Results of the recent excavation of Los Pozos, and the addition of early radiometric ages on maize throughout the Southwest, suggest that these models should be reevaluated in light of new evidence. 254 New Chronology for Maize One of the most evident results of refinement of the chronology for maize has been that the dramatic difference in time between highland adoption of agriculture and the adoption of maize by desert groups was no longer remarkable (Smiley and Parry 1990). Since the first application of AMS technology to early maize dates, more evidence for early agriculture has appeared in questionable or possibly contaminated contexts. New evidence from the Los Pozos site, along with these early dates on maize, may record the introduction of agriculture in the American Southwest. Over the past seven years new dates on maize have exceeded the earliest dates from caves, rockshelters, and alluvial sites by several hundred years (Table 7.1). The date on maize from the Los Pozos site threatens to break that boundary by several hundred years more, but, when viewed as a group, the dates are less remarkable. Figure 7.1 illustrates the calibrated radiocarbon age ranges on early maize dates in the American Southwest. It must be noted that most of these individual samples have been questioned for one reason or another. The early date from Stone Pipe was located in an unusually stained sediment that was interpreted as possible chemical contamination, lending suspicion to the date. Similar questions of contamination orcontext have been advanced for early maize dates in other parts of the Southwest. 255 Table 7.1. Earliest radiocarbon dates of cultigens in the Southwest (2500 B.P. for maize; 1700 B.P. for other cultigens, after Mabry 1996a). Radiocarbon Site Date (B.P.) Lab No. Reference 3370 ± 60 Beta-86544 D. Seymour, P.C. maize maize 3610 ± 170 2880 ± 140 Beta-2725 Beta-26271 Smiley and Parry Lukachukai maize maize maize 3445 ± 45 3135 ± 45 3050 ± 50 AA-9317 AA-9319 AA-9321 Gilpin 1994 Gilpin 1994 Gilpin 1994 Sheep Camp squash 2900 ± 230 A-3388 Simmons 1986 LA 18091 maize 2720 ± 265 UGa-4179 Simmons 1986 Salina Springs maize 2630 ± 45 AA-9318 Gilpin 1994 3740 ± 70 3120 ± 70 3060 ± 110 3010 ± 150 2980 ± 120 2780 ± 90 2690 ± 90 2630 ± 90 2140± 110 A-4187 A-4188 A-4189 A-4167 A-4186 A-4166 A-4185 A-4182 A-4184 Wills 1988 Wills 1988 Wills 1988 Wills 1988 Wills 1988 Wills 1988 Wills 1988 Wills 1988 Wills 1988 bean 2470 ± 250 squash 1900 ± 70 A-4179 A-4178 Wills 1988 Wills 1988 GX-12720 Upham et al. 1987 AA-6402 AA-6409 Tagg 1996 Tagg 1996 Cultigen Upper Rio Grande Valley LA 10577 maize Colorado Plateau Three Fir Shelter Mountain Transition Zone maize Bat Cave maize maize maize squash maize maize squash bean Tularosa Cave Southern Basin and Range Province 3175 ± 240 maize Tornillo Fresnal Shelter maize maize 2945 ± 55 2880 ± 60 256 Radiocarbon Site Cultigen Date (B.P.) Lab No. Reference maize bean bean bean 2540 ± 200 2085 ± 60 2015 ± 65 1955 ± 55 A-3070 Carmichael 1982 AA-6407 AA-6405 AA-6404 Tagg 1986 Tagg 1986 Tagg 1986 Milagro maize maize maize maize maize 2930 ± 45 2915 ± 45 2910 ± 45 2780 ± 90 2775 ± 60 AA-12055 AA-12056 AA-12053 AA-1074 AA-12054 B. Huckell et al. B. Huckell et al. B. Huckell et al. B. Huckell 1988 B. Huckell et al. Solar Well maize 2835 ± 85 AA-6641 unpublished Fairbank maize maize maize 2815 ± 80 2800 ± 140 2590 ± 75 AA-4457 AA-4458 AA-4459 B. Huckell 1990 B. Huckell 1990 B. Huckell 1990 Cortaro Fan maize maize 2790 ± 60 2595 ± 70 AA-2782 AA-2783 B. Roth, p.c. B. Roth, p.c. West End maize maize 2735 ± 75 2675 ± 80 AA-4810 AA-4811 B. Huckell 1990 B. Huckell 1990 Clearwater maize maize maize 2580 ± 60 2520 ± 40 2500 ± 60 Beta-90225 Beta-85405 Beta-90226 M. Diehl, 1996a M. Diehl, 1996a M. Diehl, 1996a Charleston maize 2565 ± 75 AA-4809 B. Huckell 1990 Matty Canyon maize 2505 ± 55 AA-13125 B. Huckell 1995 Eagle Ridge cotton 1725 ± 65 AA-13690 Elson et al. 1995 257 9-40228 • Denotes floodplain site 1 MOM 10 0 I MN G C11:11 o LA -0276,3 8-89862 0 MEE Direct dates on maize from Tucson Basin sites (>2500 B.P. RCY). 8-90225 AA -12054 WEI I AA III - 01074 1 AA-02782 003111111113 • M MEW III 0 LA-06641 • 8-90.318 AA-12053 11 AA-12069 AA -12055 0 Earliest direct dates on maize from the American Southwest. I II Fresnal Shelter I Lukachukai OO Luchachuki AA-09321 AA-093111 IMMI:1 LA 10577 9- 86544 l• II0 Luchachuki Anomalous direct date on maize -H ON I 1 • Square Hearth from the Tucson Basin. • ON I 3-88145 I 9-95634 •a Los Pozos Middle Archaic-Dates on wood charcoal. B-66144 9-956.3.3 I III 8-95635 NMI 8-81329 F 2500 AA-13257 MI I M I 0 AA-01317 • 10 1 AA-06402 2000 ..._ Los Pozos Middle Archaic date on maize. 1500 1000 CAMS - 34923 500 CALIBRATED YEARS BC Figure 7.1. Calibrated radiocarbon age ranges on early maize in the American Southwest (after Gregory 1996a). 258 In 1992, maize cupules were recovered from an alluvial lens at the Los Pozos site (Huckell personal communication 1997). An unpublished radiocarbon date on the maize is a few hundred years older than a wood charcoal date from the same deposit. Because the maize age and charcoal from the alluvium were out of sequence and roughly 300 years apart, he assumed that the maize date must be in error (Huckell, personal communication 1997). The addition of the 4050 ± 50 B.P. date from the Los Pozos site in a stratum dating roughly the same age, raises the question of whether these early dates could represent the migration of agriculturally capable groups into the area, establishing a presence after the Altithermal warm period. Both dates are in a more secure context and lack the potential indicators of contamination that are present in other early Southwestern samples. During the succeeding Late Archaic period, groups living in southern Arizona moved toward an agriculturally dependent, at least semi-sedentary society. By 2500 B.P., maize agriculture along the Santa Cruz River was fairly well-developed. The earliest of these large sites (early Cienega phase, ca. 2600-2400 B.P.) typically displays a number of extramural pit features and a few pit structures, as demonstrated by archaeological excavations at the Wetlands and Clearwater sites. By 2400 B.P., large villages with hundreds of pit structures are present along the floodplain of the Santa Cruz. Initial research suggests that these large villages are only present within the study area; however, intensive research such as has been conducted over the past four years has not been conducted in other parts of the Santa Cruz River Valley. Additional planned cultural resource management of portions of the interstate corridor 259 north of Ruthrauff Road and south of A-Mountain will determine whether the pattern of occupation present at sites in the study area is repeated elsewhere in the valley. The Relationship between Stream Changes and the Adoption of Agriculture The record of stream changes recorded at the Los Pozos site and elsewhere in the study and focus reaches provides a mechanism for the adoption of agriculture by groups in southern Arizona. Huckell (1990, 1992a) has suggested that immigrant groups, arriving from Mexico following the Altithermal hiatus, responded to valley aggradation in southern Arizona by moving into the area. In his model, valley aggradation begins roughly 4,500 years ago, but human groups capable of agriculture do not enter the area for about 1,000 years. Their sudden appearance displaces or overtakes the Middle Archaic population already living in the area. The record of stream changes suggests, however, that small microenvironments, similar to those found during the Early Agricultural period, are present in parts of the river immediately following middle Holocene aggradation. The possibility that these microenvironments are occupied by Middle Archaic groups with the capacity to cultivate maize suggests that either the microenvironments and their preservation were too small to leave a record as robust as that found during the Early Agricultural period or that the areas were too small to allow large groups of people to exploit this part of the landscape. During the Early Agricultural period, stream changes provide the impetus for these same groups to enhance their use of the floodplain. Increased sedimentation 260 over larger geographical areas provided a large, favorable riverine environment, capable of supporting large-scale agriculture. For approximately 500 years, this Early Agricultural population thrived in a floodplain environment that contributed large quantities of fresh sediment and water to the floodplain, creating an environment conducive to agricultural production. ARCHAEOLOGICAL SITE PRESERVATION Prehistoric processes acting on the Santa Cruz River have preserved Middle Archaic sites in some localities. Individual areas or reaches may experience different processes either enhancing sediment entrainment or sediment storage. Along the river, there exist features which have acted historically as controlling mechanisms, preventing thresholds from being exceeded in reaches upstream or downstream of those features. Only during significant events are critical thresholds surrounding these features also exceeded. Preservation between each of these features should be predictable, but may not be the same as preservation in reaches downstream or upstream of that feature. In the San Xavier Mission area, middle Holocene sediments appear to be preserved only along the valley margins as slopewash deposits. Although the middle Holocene sediments at the Los Pozos site are also located along the valley margin, they appear to be floodplain rather than slopewash sediments, indicating the potential for additional preservation of middle Holocene sediments in this reach. At the Clearwater archaeological site, near A-Mountain, sediment storage is a dominant 261 process throughout the middle Holocene and may have the potential to preserve Middle Archaic archaeological sites. A 3,000-year old feature uncovered in archaeological trenches that exceeded the OSHA-regulated depth may be the clue of this earlier preservation. It is also possible that additional Middle Archaic sites will be discovered in the area between Ruthraff and Ina Roads where sediment input from the Rillito and Catlada del Oro washes may be high, enabling much of the area to preserve, rather than erode middle Holocene sediments. Both the preservation of Middle Archaic aged charcoal at the Ina Road railroad locality (Haynes and Huckell 1986) and the general uninterrupted aggradational sequence in the Ina Road area documented by additional studies in the Ina Road area, suggest that preservation may be the dominant process in this area. LEVELS OF ARCHAEOLOGICAL AND ENVIRONMENTAL INTERPRETATION The interpretation of human response to environmental changes is highly dependent on the scales of the two records being observed. This interpretation works best when short-term catastrophic events such as volcanism, earthquakes, or flooding both preserve the archaeological record in the geologic matrix and demonstrate a direct human response to sudden changes. The second most common level of archaeo-environmental interpretation involves long-term changes in climate (i.e., the Pleistocene to Holocene transition, the Altithermal drought, etc.), that are often tied to 262 long-term changes in human "adaptive" patterns (i.e., large scale migrations, dramatic changes in human subsistence, such as hunting-gathering to agriculture). This broader level of analysis, although useful as a framework for archaeological research, is at too gross a scale to address questions about subtle shifts in human use of the landscape and even more subtle changes in mobility and subsistence. Most archaeologists today recognize that our knowledge of Archaic hunter-gatherer groups has moved beyond the level of reconstructing "adaptations" to reconstructing prehistoric phenomena at the scale of differences between residential and logistical mobility or differences between two scales of logistical mobility. Environmental data, derived from an alluvial stratigraphic record that measures only large-scale changes in hydrology, can offer little assistance in understanding these more fine-grained aspects of the archaeological record. An approach that considers the implications of small-scale geologic changes as they are reflected in that stratigraphie record can provide resolution of the human-environmental interface at a scale that is appropriate to the research questions of the future. 263 APPENDIX A KEY TO GEOLOGIC MAP UNITS IN FIGURE 4.1 The following descriptions of geologic map units are from McKittrick (1988): cha The most active portion of the main drainage channels. Washes commonly contain coarse to fine-grained sand exhibiting bar-and-swale topography. The channel position is unstable and subject to rapid migration within the finergrained floodplain deposits that include terraces 1 and 2. It is the topographically lowest unit in the map area, and is frequently too young to support dense vegetation. These areas are flooded frequently. ch Active and recently active channel deposits that are associated with incised channels, except in Avra Valley, where main axial drainage is wide, aggradational, and unconfined. The unit includes a complex of low terraces, active channels, gravel bars, and floodplains. The average height of the lower terraces above the active channels is about 1 m. These areas are subject to occasional frequent flooding and sediment transport. ti Youngest and lowest terrace that has been recently abandoned. Soil development is very weak to non-existent. t2 Flat, well-preserved terraces associated with modern floodplains of the Santa Cruz River, Rillito Creek, and Pantano Wash. Lies topographically above ti but below t2. Surfaces are generally well preserved and lacking in erosional modification. Soil development consists of weakly indurated Entisols. t3 Generally narrow, poorly defined terraces that are intermediate in height between t2 and t4. Soils appear to contain moderately developed color and structural argillic horizons with little carbonate accumulation. Corresponds to the Jaynes terrace of Smith (1938). t4 Broad terrace covering a large aerial extent in the central Tucson Basin area. Soils typically contain a fairly well-developed argillic horizon with varying degrees of secondary carbonate accumulation. Corresponds to the Cemetery terrace of Smith (1938). t5 Highest, oldest terraces in the map area. These terraces form elongate ridges that may represent the surface of a former level of maximum alluvial fill in the Tucson Basin; they correspond to the University terrace of Smith (1938). 264 Y Active or recently active alluvial deposits. These deposits commonly form thin veneers that mantle older map units. Fan surfaces often contain a concentration of coarser pebbles at the surface that mantle underlying silt, although this surface armor is loose and not continuous. Surface clasts show no rock varnish. The unit is typically lower in relief than older fan surfaces. It covers a portion of Avra Valley but is primarily confined to small fans close to the mountain fronts in the rest of the map area. Gullies that originate on fan surfaces are usually less than 0.5 m deep and may be either erosional or distributary. Soils are very weakly developed, if present at all, reflecting the young age and recent activity of the unit. This unit should be considered to be potentially subject to flooding and sediment transport. M2 The youngest alluvium abandoned by active depositional processes. This unit is older than Y and generally younger than Ml, but it ranges between these two units in age and surficial characteristics. A slight pavement occurs on many of the fan surfaces, with surface clasts typically averaging 2 cm in diameter. Varnish is rare, although surface clasts can have a slight pinkish hue in some locations. The surface of this unit is generally intermediate in height between M1 and Y. Interfluves are usually flat to slightly rounded. Gullies originating on fan surfaces range from less than 0.5 m to 4 m in depth. Surface expression varies from being fairly flat and smooth to gently undulating. Soil development varies from slight development (Entisol) with a brownish surface color to moderately well-developed argillic horizons (Typic Haplargid) and even petrocalcic horizons in a few areas, although the presence of a petrocalcic horizon usually indicates the occurrence of a buried older deposit. Variations in depth and density of dissection cause portions of this unit to be susceptible to flooding during larger events. Ml Relatively old geomorphic unit of wide aerial extent. Surfaces tend to have a slight pavement development where concentrations of fairly well-sorted clasts averaging 2 to 3 cm in diameter partially armor fan surfaces. A reddish rock varnish is common on surface clasts. The fan surface is typically hummocky and has well-rounded interfluves. Channels heading on fan surfaces are broad and V-shaped, contain sandy floors, and range up to 5 m in depth. Soils commonly contain a petrocalcic horizon with or without an overlying red, clay-rich argillic horizon (Paleargid or Paleorthid). These surfaces are isolated from active fluvial processes, and only gullied areas are subject to flooding. 265 0 Highest and oldest alluvium in the map area that retains a preserved geomorphic surface. This unit encompasses all surfaces that are higher in relief than M1 and may range widely in age. Fans are frequently cored by bedrock on all of the mountain fronts. Surfaces have slight pavement development consisting of scattered clasts overlying silt. A reddish rock varnish is common on the surface clasts in some location, although absent in other. Interfluves are very rounded, broad, and V-shaped. Gullies heading on fan surfaces range from 1 m to 15 m in depth. Soil development consists of thick petrocalcic horizons at or near the fan surface (Paleorthid) unless removed by erosion. Argillic horizons are not common, although they do occur locally in particularly stable areas. Flooding is restricted to gullies. Qtbf Alluvium that does not exhibit a preserved geomorphic surface. This unit is usually higher in relief and probably older than Ml, although wherever fan surfaces are absent, the surface age cannot be determined. Flooding is restricted to gullies. Br Bedrock. 266 APPENDIX B SEDIMENT DESCRIPTIONS FOR FIGURE 6.3 The following sedimentary descriptions are compiled from unpublished fieldnotes taken by Huckleberry in 1995 and Freeman in 1996. 502. Dark grayish brown (10YR 4/2) silty clay; weak, medium, angular blocky structure; many, fine carbonate filaments; diffuse, smooth boundary; overbank alluvium modified by pedogenesis. 503. Pale brown (10YR 6/3) silt; massive; two thin, discontinuous, silty clay layers; moderately bioturbated, very few, very fine oxide stains; abrupt wavy boundary; overbank alluvium. 504. Dark grayish brown (10YR 4/2; 10YR 3/2 to 3/3 when moist) silty clay; weak, medium, angular blocky structure; contains silt laminae and charcoal; slightly bioturbated; abrupt wavy boundary; overbank alluvium. 505. Pale brown (10YR 6/3) silt; massive; few fine carbonate filaments; abrupt wavy boundary; overbank alluvium. 506. Grayish brown (10YR 4/2; 2.5Y to 10 YR 3/2 when moist) silty clay and pale brown (10YR 6/3) silt; weak, medium, angular blocky structure; common, fine, carbonate filaments; slightly bioturbated; abrupt wavy boundary; overbank alluvium,; contains cultural features and frequent charcoal flecks and small chunks. 507. Pale brown (10YR 6/3) very fine sand with low angle, trough cross beds; massive; few, fine, soft carbonate masses; abrupt wavy boundary, smooth in places.; overbank alluvium. 508. Grayish brown (10YR 5/2) silty clay and pale brown (10YR 6/3) silt; weak, medium, angular blocky structure; common, fine, carbonate filaments; slightly bioturbated; abrupt wavy boundary; overbank alluvium. 509. Pale brown (10YR 6/3) silt to very fine sand; massive; few carbonate masses; abrupt smooth boundary; overbank alluvium. 510. Brown (7.5YR 5/2) silty clay; weak, medium, angular blocky structure; common, fine, carbonate filaments and medium nodules; contains artifacts and charcoal; diffuse, smooth boundary.; overbank alluvium modified by pedogenesis. 267 510a. Dark grayish brown (10YR 3.5/2) clay loam; moderate carbonate coatings on fine subangular blocky ped faces; probably a Btk horizon; forms a clear boundary with the underlying 510b; ca 4-5 cm thick. 510. Grayish brown to light grayish brown (10YR 5.5/2 to 10YR 6/2) clay loam with many grayish brown (10YR 5/2) mottles and very few, very fine Mn/Fe and oxide stains; few carbonate stringers; peds are fine to medium subangular blocky; forms a clear boundary with the underlying 510c; ca 14 cm thick. 510c. Light brownish gray (10YR 6/2) and grayish brown (10YR 5/2) silty clay loam with many medium to fine dark grayish brown (10YR 4/2) mottles; very few, very fine carbonate stringers and very few flecks of charcoal; structure is massive to very fine subangular blocky; forms a clear boundary with the underlying 510d; ca 17 cm thick. 510d. Very dark to dark grayish brown (10YR 3/2 to 4/2) clay loam with light brownish gray (10YR 6/2) mottles; many very fine Mn/Fe and oxides (orangish to reddish brown); some intrusion of Unit 510c through bioturbation; forms abrupt boundary with underlying Unit 511. 511. Light gray (10YR 4/2) to very pale brown (10YR 7/2.5) silt; compact, massive, and relatively unmodified, but with some platy sedimentary structure still present); contains some "rootlets" which may actually be the remains of organic material covered as this unit was deposited over the top of Unit 512; upper boundary with Unit 510 abrupt, lower boundary with Unit 512 very abrupt; marks the onset of sedimentation after a hiatus marked by soil formation at the top of underlying Unit 512; appears to be part of the same series of deposition as the overlying Unit 510, but the absence of pedogenic qualities seen in Unit 510 makes this a distinct depositional episode. 512. Dark grayish brown (10YR 4/2) to grayish brown/light grayish brown (10YR 5.5/) clay loam; the colors are mottled in roughly equal amounts, although the soil takes on the general appearance of the darker color in the cross section; some silt is present throughout the unit, and is probably the source of the lighter color; some banding is dimly visible but diffuse, and this sediment was probably originally a series of bands, but pedogenic processes, bioturbation, and compression have made it difficult to distinguish distinct sedimentary episodes; a thin silt band is slightly more predominant in the lower (ca. 5 cm) in places (Unit 268 513. forming a sedimentary horizon marked by an abrupt to clear boundary with the upper portions of the unit; some very fine Mn/Fe and oxide stains and abundant rootlets throughout; structure of the upper half is platy in appearance, while the lower half is fine subangular blocky; abrupt boundary with the overlying Unit 511 is marked by soil formation (Btk) at the top of Unit 512. 514a 1. Dark grayish brown (10YR 4/2) clay loam with light brownish gray (10YR 6/2) mottles of silt to silty clay loam; moderate, fine carbonate filaments; structure medium to fine subangular blocky; forms an abrupt, smooth boundary with Unit 512 (513) above and a clear to gradual boundary with 514a 2 below. 514a 2 . Grayish brown (10YR 4.5/2 to 5/2) clay loam to silty clay loam, finely mottled with the two colors; moderate, fine carbonate filaments and few very fine Mn/Fe and oxides; gradual boundary with Unit 514b below. 514b. Light brownish gray (10YR 6/2 and 10YR 4/3.5) silt loam; compact and massive to blocky structure; bioturbation bringing in materials from upper and lower units; gradual boundary with Unit 514c below. 514e. Very dark grayish brown (10YR 3/2 to 3/3) clay loam with few, fine Mn/Fe and oxides; moderate to many carbonate filaments; structure medium subangular blocky; gradual boundary with underlying Unit 515 515. Very dark grayish brown (10YR 3/2 to 3/3) when moist and dark grayish brown (10YR 4/3 to 4/2) when dry, sandy clay loam with some fine pebble gravel (ca. <2 cm), possibly representing reworked materials from the north or from Unit 516 below; massive, very compact when dry; lower boundary is clear to abrupt and forms an erosional surface. 516. Very gravelly loam, approx. 10 YR 6/2, pebble gravel (pea-sized) with some oxide staining. 517. Reddish, grayish brown coarse sand and very fine pea gravel with small cobbles (6-10 cm in diameter); abrupt boundary with Unit 516 above. 520. Channel deposits. Medium to coarse sands and occasional clay laminae. 531. Brown (10YR 4/3) sandy silt; gradual boundary with Unit 532 below. 532a. Yellowish brown (10YR 5/4) silt 532b. Brown (10YR 4/3) silty clay; some fine iron oxides stains. 269 532e. Dark yellowish brown (10YR 4/4) silt. 533. Brown to dark grayish brown (10YR 4/2 to 4/3) silty clay. 534. Dark yellowish brown (10YR 3/4) silt. 535. Reddish brown clay. 536. Yellowish brown silt. 540. Reddish brown muddy gravel. 270 APPENDIX C SEDIMENT DESCRIPTIONS FOR FIGURE 6.4 The following sediment descriptions are taken from unpublished fieldnotes taken by Andrea Freeman and David Gregory in 1996. Depositional units are labelled in arabic numerals (e.g., 1-69), while stratigraphic units are labelled with roman numerals (e.g., I, lia, etc.). TRENCH 91 1. Modern disturbance. Pale brown sandy clay. 2. Pale brown silty clay; moderately compact, blocky; overbank alluvium. 3. Pale brown silt; compact; numerous carbonate filaments; overbank alluvium. 4. Brown silty clay; overbank or slackwater alluvium. 5. Pale brown silt; overbank alluvium. 6. Dark brown clay; overbank or slackwater alluvium. 7. Pale brown silt; overbank alluvium. 8. Two clay dark brown bands separated by a small thin band of silt; boundaries are wavy and indistinct; overbank alluvium. 9. Pale brown sandy silt; massive; overbank alluvium. 10. Dark brown clay band; overbank of slackwateralluvium. 11. Pale brown silt; overbank alluvium. 12. Dark brown clay; overbank or slackwater alluvium. TRENCH 92 1. Modern disturbance 2. Pale brown silty sand; moderately compact; overbank alluvium. 271 3. Pale brown silt; compact; numerous carbonate filaments; overbank alluvium. 4. Brown silty clay; overbank or slackwater alluvium; weak pedogenic alteration in the upper portion of this unit; underlying the thin upper band of clay is a medium brown silt, followed (underneath) by a clay drape. This clay drape becomes thicker at both ends of the trench and pinches out near the center of the trench to a very thin (ca. 1-3 cm) band. 5. Pale brown silt; overbank alluvium. 6. Grayish brown silty clay; overbank or slackwater alluvium. 7. Pale brown silt band; overbank alluvium. 8. Dark brown clay; overbank or slackwater alluvium. 9. Medium brown (when damp) sandy silt; massive; overbank alluvium. 10. Two clay dark brown bands separated by a small thin band; boundaries wavy and indistinct; charcoal flecks at contact with 11, highly oxidized at lower boundary; overbank alluvium. 11. Pale brown sandy silt; massive; overbank alluvium. 12. Medium brown silt (ca. 40 cm thick) and dark grayish brown silty clay; overbank or slackwater alluvium. TRENCH 93 1. Modern disturbance. 2. Pale grayish brown sandy silt, compact; overbank alluvium. 3. Pale brown silt and sand; many fine carbonates in bottom third; overbank alluvium. 4. Dark grayish brown silty clay; overbank alluvium. 5. Pale brown silt; overbank alluvium. 6. Dark brown silty clay; overbank alluvium. 272 7. Pale brown silt; overbank alluvium. 8. Dark brown silty clay; overbank alluvium. 9. Pale brown silt; overbank alluvium. 10. Dark brown silty clay; overbank alluvium. TRENCH 94 (east segment) 1. Modern disturbance 2. Grayish brown silt; compact; some carbonate filaments; overbank alluvium; contains several cultural features (possibly Hohokam). 3. Grayish brown to brown silt; more compact than 2 and more carbonate filaments; overbank alluvium. 4. Dark grayish-brown silty clay; overbank alluvium. 5. Pale brown silt; overbank alluvium. 6. Dark brown silty clay; overbank alluvium. 7. Pale brown silt; overbank alluvium. 8. Dark brown silty clay; overbank alluvium. 9. Pale brown silt; overbank alluvium. 10. Two dark brown silty clay bands separated by a thin, pale brown silt band; overbank alluvium. 11. Pale brown sandy silt and sand; overbank alluvium. 273 CHANNEL DEPOSITS AND ASSOCIATED UNITS Unit Ia 1. Massive gravel (ca. 2-6cm.) and very coarse sand (la); larger cobbles (ca. 6-15 cm) form a facies at the very eastern corner of the profile (lb); poorly sorted, clast-supported; forms an abrupt, conformable and flat boundary with 2. 2. Gravel (ca. <2 cm., few 2-6 cm.) and very coarse sand; some sedimentary structure present (laminated bedding); forms a very abrupt, conformable and flat to slightly undulating (at west end of trench) boundary with 3. 3. Medium to coarse sand; lamellar-horizontal bedding (beds slope downward to the east); some oxidation in spots (ca. 5 cm. diam.); abrupt boundary with 4, a facies which cuts 3 to the west; at the eastern edge of the exposure, 4 has deposited a gravel lag atop 3 and below deposit 5, forming a very abrupt boundary. 4. Coarse to very coarse poorly sorted sand and pea gravel; includes several lenses of sand and gravel separated by medium to coarse, well-sorted sand lenses; occasional large cobbles in upper coarse sand and gravel lens; forms an abrupt, conformable, flat boundary deposit 5 (above); some magnetite placer deposits present. 5a. Coarse sand (5a); lamellar-horizontal sedimentary structure; abundant large oxidation stains, especially in upper part of deposit. 5h. Pale brown (I0YR 6/3) medium to fine sand and silt; massive, no sedimentary structure; upper portion is primarily fine sand and silt with carbonates, the latter consisting of long strings of carbonates; some magnetite placer deposits present. Unit lb 6a. Pale gray (10YR 7/2) coarse sand (6a); some weak horizontal bedding still present; thin carbonate formed at top of the unit; forms an abrupt, flat and conformable boundary with underlying deposits; coarse substrate of 6a correlates with dipping to north visible in the east and north walls of the excavation; 6a fines upward into 6b. 6b. Pale gray (10YR 7/2) to very pale brown (7/3) fine to medium sand; forms a clear, unconformable boundary with overlying deposit 7. 274 7. Pale brown to very pale brown (10 YR 6.5/3) silt and carbonates, with many large carbonate stringers; massive sedimentary structure; clear, conformable boundary with overlying 8. 8. Pale brownish gray to pale brown (10YR 6/2.5) medium sand and carbonates; massive sedimentary structure; coarser sand and bedding to the north (opposite trench wall) with abundant iron and manganese oxide/hydroxide staining; forms a very abrupt, conformable boundary with 9. Unit Ic 9a. Dark brown (10 YR 6/1 to 6/2) clay with many very fine, dark grayish brown (4/2) to brown (7.5 YR 5/2) manganese and iron oxide/hydroxide stains; fine subangular blocky structure; approximately 5 cm thick; strongly effervescent. 9b. Dark brown (10 YR 6/2) silt to silty clay; fine to very fine subangular blocky structure; approximately 10 cm thick. Possible grass fire at lower boundary noted in east wall consisting of charcoal and oxidized sediment. 9c. Grayish brown (10 YR 5/2) clay loam with many very fine, very dark gray (10 YR 3/1) stains and pale brownish gray (10 YR 6/2) silts; medium subangular blocky; approximately 15 cm thick. 9d. Pale brownish gray (10 YR 6/2) massive silt; approximately 5-6 cm thick. 9e. Dark gray to gray (10 YR 4.5/1) clay loam with charcoal; very fine subangular blocky structure. Unit Ha 10. Medium sand (10a) with a facies of coarse sand and small gravels (10b) to the west at the bottom. 11. Silt band; moderately compact. 12. Bands of clay, silt, and sand with some small gravels. 13. Fine sand. 14. Dark brown clay band; thin (ca 1 cm). 275 15. Silt to clayey silt with some weak banding. 16. Pale brownish gray (10 YR 6/2) medium sand, but with some organic matter; massive with occasional silt/clay nodules; some very weak sedimentary structure visible; very abrupt, unconformable boundary with underlying 9; very abrupt boundary with overlying 11; no effervescence. Unit IIb 17. Pale brownish gray (10 YR 6/2) silt band; slackwater deposit, very thin (ca. 1-2 cm); conformable boundary over 10 and over other Channel I deposits; moderately effervescent. 18. Fine to medium sand with some cross-bedding structures still visible; expansion of channel or channel margin/ wavy band of silt/clay and carbonates; shows up in center of deposit and appears to separate two separate deposits of cross-bedding; a silt deposit caps deposit 12, which begins formation of a series and silt and clay bands; this finely laminated sand also has magnetite laminae present, some micro crossbedding, subsequently disturbed by bioturbation and carbonate formation; no effervescence except where carbonates in upper deposit have translocated downward. Unit III 19. Massive silt with strong carbonates; forms an abrupt but wavy boundary with 14; this silt may actually be deposit 12; strongly effervescent. 20. Sand. 21. Dark clay band; thin (ca 1 cm). 22. Sand. 23. Clay loam to silty clay loam and silt loam occurring in a series of wavy bands; lowermost clay contains some charcoal and many fine oxides; upwelling of this deposit through the silt below (deposit 13) is visible in places (deposit is ca. 15-20 cm. thick); fine to medium subangular blocky; an abrupt, conformable boundary separates this deposit from 15 (above) in the east end of the trench; to the west this boundary is clear; some effervescent bands; medium subangular blocky to columnar structure; charcoal present in lowest darkest clay band (10YR 5/2). 276 24. Clay loam (10YR 5/2) to silty clay loam (10 YR 6/2) (dark in upper 3-4 cm); medium subangular blocky structure (20-30 cm thick); forms a very abrupt, conformable, and flat boundary with 16 (above); strongly effervescent in center of deposit and toward bottom; no effervescence in upper 5-10 cm; fine columnar structure, very abrupt upper boundary. 25. A series of clay (10 YR 5/2) and silt (10 YR 7/2) bands; wavy but conformable; forms a wavy, conformable, abrupt boundary with the silt above (17), but is part of the same depositional deposit; massive to medium subangular blocky structure; moderate to weak effervescence throughout. 26. Silt (10YR 7/2), thicker to east of profile (more yellow, no carbonates compared to deposit 13); massive; forms an abrupt, wavy, conformable boundary with 18 (above); moderately effervescent. 27. Clay band (10 YR 5/1 to 5/2)(dark), pinches out to east where the silt (below) is thicker; medium subangular blocky structure; forms a clear, wavy but conformable boundary with 19 (above); no effervescence. 28. Paler colored sandy clay loam with abundant carbonates. Fine to medium subangular blocky structure. Forms a very abrupt unconformable boundary with deposit 20 (above). Medium brown (10 YR 6/2) silty clay; thin band of sand follows, forming a very abrupt upper contact with that sand; moderate effervescence; fine subangular blocky to columnar structure Unit IVa l 29. Coarse sand Unit IVa2 30. Silt and fine sand with numerous carbonate filaments. Unit IVa3 31. Silt and fine sand with numerous carbonate filaments. 277 Unit IVb 32. Small to medium sized cobble gravels and coarse sand Unit IVc 33. Floodplain facies; silt and clay bands. 34/21. Medium sand with some carbonates and organic matter in upper portion of the deposit. 35. Compact silt; some carbonate filaments toward top. 36. Fine sand and silt; bedded. 37. Compact silt. 38. Very dark brown sandy clay; many carbonate filaments and abundant evidence for organic matter. 39. Very fine sand and silt 40. Dark brown silt; abundant organic matter. Unit V 41. Compact silt. 42. Fine sand; magenite placer deposits present in laminae. 43. Coarse sand and pea gravel. 44. Clay 45. Silt 46. Medium sand 47. Coarse sand/small gravels 278 48. Silt 49. Coarse sand 50. Silt 51. Sand 52. Sand 53. Coarse sand with some small gravels. 54. Two dark clay bands separated by a silt band Unit VI 55. Medium and coarse sand; laminar bedding. 56. Clayey silt band. 57. Fine sand, well sorted; prominent iron or manganese oxide mottling toward the west near flood channel (57a); common charchoal flecks; becomes progressively finer to the east, grading into finely bedded sandy silt and silt (57b). 58. Clayey silt band. 59. Dark clay bands; thin (ca 1 cm). 60. Medium sand. 61. Silty clay, progressively siltier toward the east; prominent iron or manganese oxide mottling near flood channel (62) 62. Medium to coarse sands (62a), moderately well-sorted; grades into medium, moderately well-sorted sand to the west (62b); a high energy flood deposit. 63. Clay band with charcoal flecks at top. 64. Silty clay; this unit may represent modern disturbance or possibly a cultural feature. 279 65. Dark clay band; thin (ca 1-2 cm.) Other Deposits 66. Coarse sand with some medium to coarse gravels at the top in places (lag gravels ?); poorly sorted; faint iron or manganese oxide mottling except adjacent (ca. 1 m) to flood channel (62), where mottling is prominent. 67. Three clayey silt bands (ca 5-7 cm) separated by sands; strong to medium, prismatic to blocky structure; prominent iron or manganese oxide mottling near flood channel (62; ca. 1.5 m). 68. Fine sand; well-sorted; more compact near the top with few carbonate filaments. 69. Modern disturbance 280 REFERENCES Adams, Jenny L. 1997 Groundstone assemblage. In Archaeological Investigations of the Wetlands Site, AZ AA:12:90 (ASM), edited by Andrea K.L. Freeman. Technical report 97-5, Center for Desert Archaeology, Tucson. Adams, Karen 1988 The Ethnobotany and Phenology of Plants in and Adjacent to Two Riparian Habitats in Southeastern Arizona. Ph.D. dissertation, Department of Ecology and Evolutionary Biology, University of Arizona. University Microfilms, Ann Arbor. Agenbroad, Larry 1970 Cultural Implications from the Statistical Analysis of a Prehistoric Lithic Site in Arizona. Unpublished Master's thesis, University of Arizona, Tucson. 1978 Cultural Implications from the Distributional Analysis of a Lithic Site, San Pedro Valley, Arizona. In Discovering Past Behavior: Experiments in the Archaeology of the American Southwest, edited by Paul Grebinger, pp. 55-72. Gordon and Breach, New York. Ahlstrom, Richard V. N., Heidi Roberts, and Gary Huckleberry 1994 Archaeological and Geonzorphological Overview and Testing Plan for the Rio Nuevo Property, Tucson, Pima County, Arizona. SWCA Archaeological Report No. 94-174, Tucson, Arizona. Allen, J.R.L. 1970 Studies in fluviatile sedimentation: A comparison of fining-upward cyclothems, with special reference to coarse-member composition and interpretation. Journal of Sedimentary Petrology, 40:298-323. Altschul, Jeffrey H., and Bruce A. Jones 1990 Settlement Trends in the Middle San Pedro: A Cultural Resources Sample Survey of the Fort Huachuca Military Reservation. Technical Series No. 19. Statistical Research, Inc., Tucson. Anderson, S. R. 1987 Cenozoic Stratigraphy and Geologic History of the Tucson Basin, Pima County, Arizona. U.S. Geological Survey Water Resources Investigations Report 87-4190. 281 Antevs, Ernst 1936 Dating Records of Early Man in the Southwest. The American Naturalist 70:331-336. 1948 The Great Basin, with Emphasis on Glacial and Postglacial Times. University of Utah Bulletin 38:168-191. Salt Lake City. 1952 Arroyo Cutting and Filling. The Journal of Geology 60 (4):375-385. 1953 Geochronology of the Deglacial and Neothermal Ages. Journal of Geology 61:195-230. 1955 Geologic-Climatic Dating in the West. American Antiquity 20:317-335. 1962 Late Quaternary Climates in Arizona. American Antiquity 20:193-198. Baker, Victor R. 1977 Stream-channel Response to Floods, with Examples from Central Texas. Geological Society of America Bulletin 88:1057-1071. 1988a Overview. In Flood Geomorphology, edited by V. R. Baker, R. Kochel, and P. Patton, pp. . 1988b Geological Fluvial Geomorphology. Geological Society of America Bulletin 100:1157-1167. Baker, Victor R., and C. R. Twidale 1991 The Reenchantment of Geomorphology. Geomorphology 4:73-100. Bayham, Frank E. 1982 A Diachronic Analysis of Prehistoric Animal Exploitation at Ventana Cave. Ph.D. dissertation, Department of Anthropology, Arizona State University, Tempe. Bayham, Frank E., Donald H. Morris, and M. Steven Shackley 1986 Prehistoric Hunter-Gatherers of South Central Arizona: The Picacho Reservoir Archaic Project. Anthropological Field Studies No. 13, Office of Cultural Resource Management, Department of Anthropology, Arizona State University, Tempe. Beardsley, Richard K. 1948 Culture Sequences in Central California. American Antiquity 14(1):1-28. 282 Beckett, P. H. 1973 Cochise Culture Sites in South Central and North Central New Mexico. Unpublished Master's thesis, Department of Anthropology, Eastern New Mexico University, Portales. Benedict, J. B. 1979 Getting Away From It All: A Study of Man, Mountains, and the Two Drought Altithermal. Southwestern Lore 45:1-12. Benedict, J. B., and B. L. Olson 1978 The Mount Albion Complex: A study of Prehistoric Man and the Altithermal. Center for Mountain Archaeology, Ward, Colorado. Bernard-Shaw, Mary 1986 Archaeological Testing of Three Sites along the El Rio-Starr Pass Reclaimed Water Line, Tucson, Arizona. Technical Report No. 86-1. Institute for American Research, Tucson. Berry, Claudia F., and Michael S. Berry 1986 Chronological and Conceptual Models of the Southwestern Archaic. In Anthropology of the Desert West: Essays in Honor of Jesse D. Jennings, edited by C. J. Condie and D. D. Fowler, pp. 253-327. University of Utah Anthropological Papers 110, University of Utah Press, Salt Lake City. Berry, Michael S. 1982 Cultural Resources Overview: The Middle Gila Basin, An Archaeological and Historical Overview. Northland Research, Flagstaff, Arizona. 1985 The Age of Maize in the Greater Southwest: A Critical Review. In Prehistoric Food Production in North America, edited by Richard I. Ford. Anthropological Papers, Museum of Anthropology, University of Michigan. 75: 279-307. Betancourt, Julio L. 1978 Cultural Resources within the Proposed Santa Cruz Riverpark Archaeological District. Archaeological Series No. 125. Arizona State Museum, University of Arizona, Tucson. 1990 Tucson's Santa Cruz River and the Arroyo Legacy. Unpublished Ph.D. dissertation, Department of Geosciences, University of Arizona, Tucson. 283 Betancourt, Julio L., and Raymond M. Turner 1990 Tucson's Santa Cruz River and the Arroyo Legacy. University of Arizona Press, Tucson, in press. Bettis, E. Arthur III, and Edwin R. Hajic 1995 Landscape Development and the Location of Evidence of Archaic Cultures in the Upper Midwest. In Archaeological Geology of the Archaic Period in North America, edited by E.A. Bettis, III, pp. 87-113. Geological Society of America Special Paper 297, Boulder, Colorado. Bettis, E. Arthur, III, and D. M. Thompson 1981 Holocene Landscape Evolution in Western Iowa: Concepts, Methods, and Implications for Archaoelogy. In Current Directions in Midwestern Archaeology: Selected Papers from the Mankato Conference, edited by S. F. Anfinson, pp. 1-14. Occasional Publications in Minnesota Anthropology No. 9. Minnesota Archaeological Society, St. Paul. 1982 Interrelations of Cultural and Fluvial Deposits in Northeast Iowa. Association of Iowa Archaeologists Field Trip Guidebook, Vermillion, South Dakota. Binford, L. R. 1980 Willow Smoke and Dog's Tails: Hunter-gatherer Settlement Systems and Archaeological Site Formation. American Antiquity 45(1):4-20. Boggs, Sam, Jr. 1987 Principles of Sedbnentology and Stratigraphy. Merrill Publishing, Columbus Ohio. Bontrager, Daniel R., and Lyle M. Stone 1989 Archaeological Testing and Evaluation of Sites AZ AA:12:688 (ASM) and AZ AA:12:111 (ASM), North of Tucson, Pima County, Arizona. Arizona Department of Transportation Project H-0152-04C, Project N-900-325. Archaeological Research Services, Inc., Tempe. Bostwick, Todd (editor) 1988 An Investigation of Archaic Subsistence and Settlement in the Harquahalla Valley. Northland Research, Flagstaff, Arizona. Brackenridge, G. Robert 1980 Widespread Episodes of Stream Erosion During the Holocene and Their Climatic Cause. Nature 283:655-656. 284 Brackenridge, G. Robert 1981 Late Quaternary Sedimentation Along the Pomme de Terre River, Southern Missouri. Quaternary Research 15:62-76. Bradley, Raymond S. 1985 Quaternary Paleoclimatology. Chapman & Hall, London, England, 472 p. Bryan, Kirk 1925 Date of Channel Trenching (arroyo cutting) in the Arid Southwest. Science 62:338-344. 1927 Channel Erosion of Rio Salado, Socorro County, New Mexico. U.S. Geological Survey Bulletin 790:17-19. 1928 Historic Evidence on Changes in the Channel of Rio Puerco, A Tributary of the Rio Grande in New Mexico. Journal of Geology 36:265-282. 1940 Erosion in the Valleys of the Southwest. New Mexico Quarterly Review 10:227-232. 1941 Pre-Columbian Agriculture in the Southwest, as Conditioned by Periods of Alluviation. Annals of the Association of American Geographers 31(4): 219-242. Bull, William B. 1979 Threshold of Critical Power in Streams. Geological Society of America Bulletin 90:453-464. 1988 Floods—Degradation and Aggradation. In Flood Geomorphology, edited by V. R. Baker, R. C. Kochel, and P. C. Patton, pp. 157-165. John Wiley, New York. 1991 Geomorphic Responses to Climatic Change. Oxford University Press, 326 p. Burkham, D. E. 1970 Depletion of Streamflow by Infiltration in the Main Channels of the Tucson Basin, Southeastern Arizona. U.S. Geological Survey Water Supply Paper 1939-B. 1981 Uncertainties Resulting from Changes in River Form. American Society of Civil Engineers, Journal of the Hydraulics Division 107(HY5):593-610. 285 Cain, W. F., and H. E. Suess 1976 Carbon-14 in Tree-rings. Journal of Geophysical Research 81:3688-3694. Caldwell, Joseph R. 1958 Trend and Tradition in the Prehistory of the Eastern United States. American Anthropological Association. Memoir 88. 1965 Primary Forest Efficiency. Southeastern Archaeological Conference Bulletin 3:66-69. Campbell, Elizabeth W. Crozer, and William H. Campbell 1935 The Pinto Basin Site: An Ancient Aboriginal Camping Ground in the California Desert. Paper No. 9. Southwest Museum, Los Angeles. Campbell, Elizabeth W. Crozer, William H. Campbell, Ernst Antevs, Charles A. Amsden, Joseph A. Borbieri, and Francis D. Bode 1937 The Archaeology of Pleistocene Lake Mohave. Paper No. 11. Southwest Museum, Los Angeles. Chippen, J. R. and C. D. Bue 1977 Maximum floodflows in the Conterminous United States. U.S. Geological Survey Water-Supply Paper 1887. Church, M. 1981 Reconstruction of the hydrological and climatic conditions of past fluvial environments. In L. Starkel and J.B. Thornes (eds), Palaeohydrology of River Basins. BGRB Technical Bulletin No. 28. Cleland, C. E. 1976 The Focal-diffuse Model: An Evolutionary Perspective on the Prehistoric Cultural Adaptations of the Eastern United States. Midcontinental Journal of Archaeology 1:59-76. COHMAP Members 1988 Climatic Changes of the Last 18,000 Years: Observations and Model Simulations. Science 241:1043-1052. Collinson, J.D. 1978 Alluvial sediments. In Sedimentary Environments and Facies, edited by H.G. Reading, pp. 15-60. Elsevier, New York. 286 Cooke, Ronald U., and Richard W. Reeves 1976 Arroyos and Environmental Change in the American Southwest. Clarendon Press, Oxford, England. Costa, John E. 1974 Response and Recovery of a Piedmont Watershed from Tropical Storm Agnes, June 1972. Water Resources Research 10:106-112. Cummings, Byron 1927 Cochise's Yesterday. The Tombstone Epitaph 449(4):12, 15. 1928 Cochise's Yesterday. Arizona, Old and New 1(4):9-10, 26-28. Czaplicki, Jon S. 1983 Paleo-Indian and Archaic Period Cultural Resources. In An Archaeological Assessment of the Middle Santa Cruz River Basin, Rillito to Green Valley, Arizona, for the Proposed Tucson Aqueduct Phase B, Central Arizona Project, by J. S. Czaplicki and J. D. Mayberry, pp. 15-26. Archaeological Series No. 164. Arizona State Museum, University of Arizona, Tucson. Czaplicki, Jon S. (compiler) 1984 Class III Survey of the Tucson Aqueduct Phase A Corridor Central Arizona Project. Archaeological Series No. 165. Arizona State Museum, University of Arizona, Tucson. Dart, Allen 1986 Archaeological Investigations at La PaIonia: Archaic and Hohokam Occupations at Three Sites in the Northeastern Tucson Basin, Arizona. Institute for American Research Anthropological Papers 4, Tucson. 1987 Archaeological Studies of the Avra Valley, Arizona, for the Papago Water Supply Project: Vol. I. Class III Archaeological Surveys on the Tohono Oiodham Indian Reservation. Anthropological Papers No. 9. Institute for American Research, Tucson. Davis, C. A., and G. A. Smith 1981 Newberry Cave. San Bernardino County Museum Association, Redlands, California. Davis, Owen K. 1984 Multiple Thermal Maxima During the Holocene. Science 225:617-619. 287 Davis, Owen K., J. C. Sheppard, and S. Robertson 1986 Contrasting Climatic Histories for the Snake River Plain, Idaho, Resulting from Multiple Thermal Maxima. Quaternary Research 26:321-339. Davis, William M. 1902 Base-level, Grade, and Peneplain. Journal of Geology 10:77-111. Dean, Jeffrey S. 1987 The Archaic of Southern Arizona. Quarterly Review of Archaeology 8(4):1, 2, 11-14. Diehl, Michael 1996a Further Archaeological Investigations of the Rio Nuevo South Property, City of Tucson, Arizona. Technical Report No. 96-5. Center for Desert Archaeology, Tucson. 1996b Archaeological Investigations of the Early Agricultural Period Settlement at the Base of A-Mountain, Tucson, Arizona. Technical Report No. 96-21. Center for Desert Archaeology, Tucson. 1997 Paleobotanical Remains. The Middle Archaic Occupation at Los Pozos, AZ AA:12:91 (ASM) (draft). Anthropological Papers No. 20. Center for Desert Archaeology, Tucson, in preparation. Diehl, Michael, and Andrea K L. Freeman 1996 Results of Testing. In Further Archaeological Investigations of the Rio Nuevo South Property, City of Tucson, Arizona. Technical Report No. 96-5. Center for Desert Archaeology, Tucson. Dillehay, Thomas D. 1974 Late Quaternary Bison Populations Changes on the Southern Plains. Plains Anthropologist 19:180-196. Doelle, William H. 1985 The Southern Tucson Basin: Rillito-Rincon Subsistence, Settlement, and Community Structure. In Proceedings of the 1983 Hohokam Symposium, edited by Alfred E. Dittert, Jr., and Donald E. Dove, pp. 183-198. Occasional Paper No. 2. Arizona Archaeological Society, Phoenix. 1996 Preface. In Further Archaeological Investigations of the Rio Nuevo South Property, City of Tucson, Arizona. Technical Report No 96-5. Center for Desert Archaeology, Tucson. 288 Douglas, John E., and Douglas B. Craig 1986 Investigations of Archaic and Hohokam Sites on the Flying V Ranch, Tucson, Arizona. Pima Community College Archaeological Report No. 13, Tucson, Arizona. Downum, Christian E., Adrianne Rankin, and Jon S. Czaplicki 1986 A Class III Archaeological Survey of the Phase B Corridor, Tucson Aqueduct, Central Arizona Project. Archaeological Series No. 168. Arizona State Museum, University of Arizona, Tucson. Eddy, Frank W. 1958 A Sequence of Cultural and Alluvial Deposits in the Cienega Creek Basin, Southeastern Arizona. Unpublished Master's thesis, Department of Anthropology, University of Arizona, Tucson. Eddy, Frank W., and Maurice E. Cooley 1983 Cultural and Environmental History of the Cienega Valley, Southeastern Arizona. Anthropological Paper No. 43, University of Arizona, Tucson. Eder, J. 1984 The Impact of Subsistence Change on Mobility and Settlement Pattern in a Tropical Forest Foraging Economy: Some Implications for Archaeology. American Anthropologist 86:837-853. Elson, Mark D. 1995 Assessment of Chronometric Methods and Dates. In The Roosevelt Community Development Study: New Perspectives on Tonto Basin Prehistory, edited by M.D. Elson, M.T. Stark, and D.A. Gregory, pp. 39-60. Anthropological Paper No. 15, Center for Desert Archaeology, Tucson. Elson, Mark D., and William H. Doelle 1987 Archaeological Assessment of the Mission Road Extension: Testing at AZ BB:13:6 (ASM). Technical Report No. 87-6. Institute for American Research, Tucson. Ely, Lisa L. 1992 Large Floods in the Southwestern United States in Relation to Late Holocene Climatic Variations. Ph.D. dissertation, University of Arizona, Tucson, Arizona. Emslie, S. D., R. C. Euler, and J. I. Mead 1987 A Desert Culture Shrine in Grand Canyon, Arizona, and the Role of Split-twig Figurines. National Geographic Research 3:511-516. 289 Euler, Robert C. (editor) 1984 The Archaeology, Geology, and Paleobiology of Stanton's Cave, Grand Canyon National Park, Arizona. Monograph 6, Grand Canyon Natural History Association. Euler, Robert C., George J. Gumerman, Thor N. V. Karlstrom, Jeffrey S. Dean, and Richard H. Hevly 1979 The Colorado Plateaus: Cultural Dynamics and Paleoenvironment. Science 205(4411):1089-1101. Evans, G. L. 1951 Prehistoric Wells in Eastern New Mexico. American Antiquity 17:1-9. Ezzo, Joseph A., and William L. Deaver 1996 Data Recovery at the Costello-King Site (AZ AA:12:503IASMD, A Late Archaic Site in the Northern Tucson Basin. Draft report prepared for Waste Management of Southern Arizona, Statistical Research, Tucson. Fairbanks, C. H. 1942 The Taxonomic Position of Stallings Island, Georgia. American Antiquity 7:223-231. Fish, Suzanne K., Paul R. Fish, and John Madsen 1985 A Preliminary Analysis of Hohokam Settlement and Agriculture in the Northern Tucson Basin. In Proceedings of the 1983 Hohokam Symposium, Part I, edited by A. E. Dittert and D. E. Dove, pp. 75-100. Occasional Papers 2, Arizona Archaeological Society, Phoenix. Fish, Suzanne K., Paul R. Fish, and John Madsen 1990 Sedentism and Settlement Mobility in the Tucson Basin Prior to A.D. 1000. In Perspectives on Southwestern Prehistory, edited by P. E. Minnis and C. L. Redman, pp. 76-91. Westview Press, Boulder, Colorado. Folk, Robert 1954 The Distinction Between Grain Size and Mineral Composition in Sedimentary-rock Nomenclature. Journal of Geology 62:344-59. 1974 Petrology of Sedimentary Rocks. Hemphill Publications, Austin, Texas. Ford, Richard I. 1981 Gardening and Farming Before A.D. 1000: Patterns of Prehistoric Cultivation North of Mexico. Journal of Ethnobiology 1:6-27. 290 Ford, Richard I. 1984 Ecological Consequences of Early Agriculture in the Southwest. In Selected papers on the archaeology of Black Mesa, vol. 2, edited by Stephen Plog and Shirley Powell, pp. 127-138. Southern Illinios University Press, Carbondale and Edwardsville. Freeman, Andrea K. L. 1995 Archaeological Testing of Soils Test Pits for Proposed Development of a Wetlands and Recharge Basin at AZ AA:12:90. Technical Report No. 95-3. Center for Desert Archaeology, Tucson. 1996 The Wetlands/Recharge Project: Summary Results of Archaeological Testing and a Plan for Data Recovery at AZ AA:12:90 (ASM). Center for Desert Archaeology, Tucson. 1997 Archaeological Investigations of the Wetlands Site, AZ AA:12:90 (ASM). Technical Report No. 97-5. Center for Desert Archaeology, in preparation. Frison, George C. 1991 Prehistoric Hunters of the High Plains (2nd edition). Academic Press, New York. Fritz, Gordon L. 1974 The Tucson Sewage Project: An Archaeological Survey of a Sewage Interceptor Route. Ms. on file, Arizona State Museum, Tucson. Gile, LeLand H. 1975 Holocene Soils and Soil-Geomorphic Relations in an Arid Region of Southern New Mexico. Quaternary Research 5:321-360. Gilman, P. A. 1987 Architecture as Artifact: Pit Structures and Pueblos in the American Southwest. American Antiquity 52:538-564. Gilpin, Dennis 1994 Lukachukai and Salinas Springs: Late Archaic/Early Basketmaker Habitation Sites in the Chinle Valley, Northeastern Arizona. Kiva 60(2):203-218. Gish, Jannifer 1986 Palynology of Picacho Reservoir, Arizona. In Prehistoric Hunter-Gatherers of South Central Arizona: The Picacho Reservoir Archaic Project, edited by F.E. Bayham, D.H. Morris, and M. Steven Shackley, pp. 279-298. 291 Gilpin, Dennis 1994 Lukachukai and Salinas Springs: Late Archaic/Early Basketmaker Habitation Sites in the Chinle Valley, Northeastern Arizona. Kiva 60(2):203-218. Graf, William 1983 The Arroyo Problem--Paleohydrology and Paleohydraulics in the Short Term. In Background to Paleohydrology, edited by K. J. Gregory, pp. 279-302. John Wiley, New York. 1984 A Probabilistic Approach to the Spatial Assessment of River Channel Instability. Water Resources Research 20:953-962. Graf, William 1988a Fluvial Processes in Dryland Rivers. Springer-Verlag, Berlin. 1988b Definition of Flood Plains Along Arid-region Rivers. In Flood Geomorphology, edited by V. R. Baker, R. C. Kochel, and P. C. Patton. John Wiley, New York. Grayson, Donald K. 1993 The Desert's Past: A Natural Prehistory of the Great Basin. Smithsonian Institution Press, Washington, D.C. Gregory, David A. 1995 Interstate 10 Frontage Road Project, Ruthrauff Road to Prince Road: Results of Archaeological Testing and a Plan for Data Recovery at AZ AA:12:91 (ASM). Technical Report 95-8. Center for Desert Archaeology, Tucson. 1997a Data Integration and Synthesis. In The Middle Archaic Occupation at Los Pozos, AZ AA:12:91 (ASM) (draft). Anthropological Papers No. 20. Center for Desert Archaeology, Tucson. 1997b The Early Agricultural Occupation at Los Pozos, AZ AA:12:91 (ASM) (draft). Anthropological Papers No. 21. Center for Desert Archaeology, Tucson, in preparation. Gregory, David A., and Sam Baar 1997 Stratigraphy, Chronology, and Characteristics of the Cultural Features and Deposits. In The Middle Archaic Occupation at Los Pozos, AZ AA:12:91 (ASM) (draft). Anthropological Papers No. 20. Center for Desert Archaeology, Tucson. 292 Gregory, Herbert E. 1917 Geology of the Navajo Country. USGS Professional Paper No. 93. Griffin, James B. 1946 Cultural Change and Continuity in Eastern United States Archaeology. In Man in Northeastern North America, vol. 3, edited by Frederick Johnson, pp. 37-95. Papers of the Robert S. Peabody Foundation for Archaeology, Andover, Massachusetts. Grove, G. T. 1962 Rillito Creek Floodplain Study, Tucson, Pima County. City-County Planning Department, Tucson. Haag, William G. 1942 Early Horizons in the Southeast. American Antiquity 7(3):209-222. Hack, John T. 1939 The Late Quaternary History of Several Valleys of Northern Arizona: A Preliminary Announcement. Museum of Northern Arizona Notes 11:67-73. Halbirt, Carl D. and T. Kathleen Henderson 1993 Archaic Occupation on the Santa Cruz Flats. Northland Research, Flagstaff, Arizona. Hall, Steven A. 1985 Quaternary Pollen Analysis and Vegetational History of the Southwest. In Pollen Records of Late-Quaternary North American Sediments, edited by V. M. Bryant, Jr., and R. G. Holloway, pp. 95-123. American Association of Stratigraphic Palynologists Foundation, Dallas. Hard, Robert J. 1986 Ecological Relationships Affecting the Rise of Farming Economies: A Test from the American Southwest. Unpublished Ph.D. dissertation, Department of Anthropology, University of New Mexico, Albuquerque. Harrington, Mark R. 1933 Gypsum Cave, Nevada. Southwest Museum Papers No. 8. Southwest Museum of Man, Los Angeles. Hastings, J.R. 1958 Vegetation Change and Arroyo Cutting in Southeastern Arizona during the Past Century: A Historical Review. In Arid Lands Colloquia, pp. 24-39. University of Arizona, Tucson. 293 Haury, Emil W. 1950 The Stratigraphy and Archaeology of Ventana Cave, University of Arizona Press, Tucson, and University of New Mexico Press, Albuquerque. 1960 Association of Fossil Fauna and Artifacts of the Sulphur Spring Stage, Cochise Culture. American Antiquity 25: 609-610. 1962 The Greater American Southwest. In Courses Toward Urban Life: Some Archaeological Considerations of Cultural Alternates, edited by Robert J. Braidwood and Gordon R. Willey, pp. 106-131. Viking Fund Publications in Anthropology no. 32. New York. Haury, Emil W., and Julian D. Hayden 1975 Preface 1975. In The Stratigraphy and Archaeology of Ventana Cave (2nd Printing), edited by E. W. Haury, pp. v-vi. University of Arizona Press, Tucson. Haynes, C. Vance, Jr. 1968 Geochronology of Late-Quaternary Alluvium. In Means of Correlation of Quaternary Successions. Vol. 8, pp. 591-631. Proceeding VII Congress International Association for Quaternary Research. University of Utah Press. 1987 Curry Draw, Cochise County, Arizona: A Late Quaternary Stratigraphic Record of Pleistocene Extinction and Paleo-Indian Activities. Geological Society of America Centennial Field Guide–Cordilleran Section, 1987:23-28. 1990 The Antevs-Bryan Years and The Legacy for Paleoindian Geochronology. In Establishment of Geologic Framework for Paleoanthropology, edited by L.F. Laporte, pp. 55-68Geological Society of America, Special Paper 242, Boulder, CO. 1995 Geochronology of Paleoenvironmental Change, Clovis Type Site, Blackwater Draw, New Mexico. Geoarchaeology 10(5):317-388. Haynes, C. Vance, Jr., and Bruce B. Huckell 1986 Sedimentary Successions of the Prehistoric Santa Cruz River, Tucson, Arizona. Open file report, Arizona Bureau of Geology and Mineralogy, University of Arizona, Tucson. Hendrickson, Dean A., and W.L. Minckley 1984 Cienegas—Vanishing Climax Communities of the American Southwest. Desert Plants 6(3). 294 Hereford, Richard 1984 Climate and ephemeral-stream Process: Twentieth-century geomorphology and alluvial stratigraphy of the Little Colorado River, Arizona. Geological Society of America Bulletin 95:654-668. Hewitt, James M., and David Stephen 1981 Archaeological Investigations in the Tortolita Mountains Region, Southern Arizona. Archaeological Field Report No. 10. Pima Community College, Tucson, Hirschboeck, Kathryn K. 1985 Hydroclimatology of Flow Events in the Gila River Basin, Central and Southern Arizona. Doctoral dissertation, University of Arizona, Tucson, Arizona, 335 p. 1987 Catastrophic Flooding and Atmospheric Circulation Anomalies. In Catastrophic Flooding, edited by L. Mayer and D. B. Nash, pp. 23-56. Allen and Unwin, Boston. Holliday, Vance T. 1985 Archaeological Geology of the Lubbock Lake Site, Southern High Plains of Texas. Geological Society of America Bulletin 96:1483-1492. 1989 Middle Holocene Drought on the Southern High Plains. Quaternary Research 31:74-82 Huckell, Bruce B. 1984a Archaic Occupation of the Rosemont Area. ASM Archaeological Series No. 147, Vol. I. University of Arizona, Tucson. 1984b The Paleo-Indian and Archaic Occupation of the Tucson Basin: An Overview. The Kiva 49:133-145. 1988 Late Archaic Archaeology of the Tucson Basin: A Status Report. In Recent Research on Tucson Basin Prehistory: Proceedings of the Second Tucson Basin Conference, edited by William H. Doelle and Paul R. Fish, pp. 57-80. Anthropological Papers No. 10. Institute for American Research, Tucson. 295 Huckell, Bruce B. 1990 Late Preceramic Farmer-Foragers in Southeastern Arizona: A Cultural and Ecological Consideration of the Spread of Agriculture into the Arid Southwestern United States. Unpublished Ph.D. dissertation, Arid Lands Resource Sciences, University of Arizona, Tucson. 1992 Archaeological Investigations at the City of Tucson Reclaimed Water Treatment Plant. Technical Report No. 91-6. Center for Desert Archaeology, Tucson. 1992 Review of "The Origins of Southwestern Agriculture." Science 257:15711572. 1993 The Southwestern Archaic: Scale and Perception of Preceramic Hunter-Gatherers. Paper presented in the symposium titled: The Current Status of Research in the American Southwest. 58th Annual Meeting of the Society for American Archaeology, St. Louis, Missouri, 14-18 April. 1995 Of Marshes and Maize: Preceratnic Agricultural Settlements in the Cienega Valley, Southeastern Arizona. Anthropological Papers No. 59. University of Arizona Press, Tucson. 1996a Alluvial Geomorphology and Geochronology. In Archaeological Investigations of Early Village Sites in the Middle Santa Cruz Valley: Analyses and Synthesis. Part I, edited by Jonathan B. Mabry. Anthropological Papers No. 19. Center for Desert Archaeology. 1996b The Archaic Prehistory of the North American Southwest. Journal of World Prehistory 10(3):305-373. Huckell, Bruce B., and C. Vance Haynes 1995 The Ventana Complex: New Dates and New Ideas on its Place in Early Holocene Western Prehistory. Paper presented at the 60th Annual Meeting of the Society for American Archaeology, Minneapolis, Minnesota. Huckell, Bruce B., Lisa W. Huckell, and Suzanne K. Fish (eds.) 1994 Investigations at Milagro, A Late Preceramic Site in the Eastern Tucson Basin. Technical Report No. 94-5. Center for Desert Archaeology, Tucson. 296 Huckell, Lisa W. 1996 Macrobotanical Remains. In Archaeological Investigations of Early Village Site in the Middle Santa Cruz Valley: Analyses and Synthesis, edited by Jonathan B. Mabry. Anthropological Papers No. 19. Center for Desert Archaeology, Tucson. Hunter-Anderson, Rosalind L. 1986 Prehistoric Adaptation in the American Southwest. Cambridge University Press. Cambridge, England. Huntington, E. 1914 The Climatic Factor as Illustrated in Arid America. Carnegie Institution of Washington Publication 192. Irwin-Williams, Cynthia 1967 PICOSA: The Elementary Southwestern Culture. American Antiquity 32:441- 457. 1968a The Reconstruction of Archaic Culture in the Southwestern United States. In Archaic Prehistory in the Western United States, edited by C. Irwin-Williams, pp. 19-23. Eastern New Mexico University Contributions in Anthropology 1(3), Eastern New Mexico University, Portales. 1968b Archaic Culture History in the Southwestern United States. In Early Man in Western North America, edited by Cynthia Irwin-Williams, pp. 48-54. Eastern New Mexico University Contributions in Anthropology 1(4), Eastern New Mexico University, Portales. 1973 The Oshara Tradition: Origins of Anasazi Culture. Eastern New Mexico University Contributions in Anthropology 5(1). Eastern New Mexico University, Portales. 1979 Post-Pleistocene Archeology, 7000-2000 B.C. In Southwest, vol. 9, Handbook of North American Indians, edited by Alfonso Ortiz (volume editor) and William C. Sturtevant (general editor). Smithsonian Institution, Washington, D.C. Irwin-Williams, Cynthia, and C. Vance Haynes, Jr. 1970 Climatic Change and Early Population Dynamics in Southwestern United States. Quaternary Research 1:59-71. 297 Jackson, Garrett 1989 Surficial Geologic Maps of the Northeastern, Southeastern, and Southwestern Portions of the Tucson Metropolitan Area. Arizona Geological Survey, OpenFile Report 89-2, Tucson. Janetski, J. C. 1980 Wooden and Reed Artifacts. In Cowboy Cave, edited by J. D. Jennings, pp. 75-95. University of Utah Anthropological Papers 104, University of Utah Press, Salt Lake City. Jennings, Jesse D. 1953 Danger Cave: A Progress Summary. El Palacio 60(5):179-213. 1956 The American Southwest: A Problem in Cultural Isolation. In Archaeology, 1955, edited by R. Wauchope. Memoirs of the Society for American Archaeology, 11, Society for American Archaeology, Salt Lake City. Jennings, Jesse D. 1957 Danger Cave. University of Utah Anthropological Papers No. 27. Salt Lake City. 1973 The Short Useful Life of a Simple Hypothesis. Tebiwa 16:1-9. Jennings, Jesse D., and Edward Norbeck 1955 Great Basin Prehistory: A Review. American Antiquity 21(1):1-11. Katzer, Keith 1987 Geological Reconnaissance of the Mission Road Extension. In Archaeological Assessment of the Mission Road Extension: Testing at AZ BB:13:6 (ASM), by Mark D. Elson, and William H. Doelle. Technical Report 87-6. Institute for American Research, Tucson. Kelly, Robert L. 1995 The Foraging Spectrum. Smithsonian Institution Press, Washington, D.C. Kidder, Alfred V. 1927 Southwestern Archaeological Conference. Science 66(11):489-491. Kinkade, Gay M., and Gordon L. Fritz 1975 The Tucson Sewage Project: Studies at Two Archaeological Sites in the Tucson Basin. Archaeological Series No. 64. Arizona State Museum, University of Arizona, Tucson. 298 Knox, J. C. 1983 Responses of River Systems to Holocene Climates. In Late Quaternary Environments of the United States, vol. 2, The Holocene, edited by H.E. Wright, pp. 26-41. University of Minnesota Press, Minneapolis. Knox, J. C. 1995 Fluvial Systems Since 20000 Years BP. In Global Continental Paleohydrology, edited by K. J. Gregory, L. Starkel, and V. R. Baker. John Wiley and Sons. Chichester, England. Kochel, R. C. 1988 Geomorphic Impact of Large Floods; Review and New Perspectives on Magnitude and Frequency. In Flood Geomorphology, edited by V. R. Baker, R. C. Kochel, and P. C. Patton. John Wiley, New York. Kochel, R. C., V. R Baker, and P. C. Patton 1982 Paleohydrology of Southwestern Texas. Water Resources Research 18:1165-1183. Langbein, W. B., and S. A. Schumm 1958 Yield of Sediment in Relation to Mean Annual Precipitation. Transactions of the American Geophysical Union 39:1076-1084. Leopold, Luna B. 1951 Rainfall Frequency—An Aspect of Climatic Variation. Transactions of the American Geophysical Union 32:347-357. Leopold, Luna B., and J. P. Miller 1954 A Post-glacial Chronology for Some Alluvial Valleys in Wyoming. U.S. Geological Survey Water-Supply Paper 1261. Leopold, Luna B., and M. Gordon Wolman 1957 River Channel Patterns—Braided, Meandering, and Straight. U.S. Geological Survey Professional Paper 282-B. Leopold, Luna B., M. G. Wolman, and J. P. Miller 1964 Fluvial Processes in Geomorphology. Freeman, San Francisco. Leopold, Luna B., W. W. Emmett, and R. M. Myrick 1966 Channel and Hillslope Processes in a Semiarid Area, New Mexico. U.S. Geological Survey Professional Paper 352-G. 299 Lindeman, Michael, and Andrea K. L. Freeman 1996 Archaeological Survey and Monitoring Results and a Proposed Testing Plan for Recharge Basins, Central Avra Valley Storage and Recovery Recharge Project, Pima County, Arizona. Technical Report No. 96-16. Center for Desert Archaeology, Tucson. Love, David W. 1983 Quaternary Facies in Chaco Canyon and Their Implications for Geomorphic-sedimentologic Models. In Chaco Canyon Country, edited by S. G. Wells, D. W. Love, and T. W. Gardner, pp. 207-219. American Geomorphological Field Group, Albuquerque. Mabry, Jonathan 1993 Interstate 10 Frontage Road Project, Results of Phase 1 Data Recovery at AZ AA:12:746 (ASM), Tucson, Pima County, Arizona (Draft Report). Technical Report No. 93-13. Center for Desert Archaeology, Tucson. Mabry, Jonathan 1995 Archaeological Investigations of Early Village Sites in the Middle Santa Cruz Valley: Descriptions of the Santa Cruz Bend, Square Hearth, Stone Pipe, and Canal Sites. Anthropological Papers No. 18. Center for Desert Archaeology, Tucson. 1996a Archaeological Investigations of Early Village Sites in the Middle Santa Cruz Valley: Analyses and Syntheses (draft). Anthropological Papers No. 19. Center for Desert Archaeology, Tucson. 1996b Archaic and Early Agricultural Period Mortuary Patterns in the Southwest. Paper presented at the Conference on the Archaic Prehistory of the North American Southwest, Albuquerque, NM, October 24-26. Mabry, Jonathan B., and Jeffery J. Clark 1994 Early Village Life on the Santa Cruz River. Archaeology in Tucson 8(1):1-5. Mabry, Jonathan B., Andrea K.L. Freeman, and Michael K. Faught 1997 Early Arizonans: Contexts for Investigating and Preserving Paleoindian and Archaic Sites in Arizona. Technical Report No. 97-7. Center for Desert Archaeology, Tucson. Maddox, R.A., F. Canova, and L.R. Hoxit 1980 Meterological Characteristics of Flash Flood Events over the Western United States. Monthly Weather Review 108: 1866-1877. 300 MacDonald, J. 1981 North American Bison: Their Classification and Evolution. University of California Press, Berkeley. MacNeish, R.S. 1992 The Origins of Agriculture and Settled Life. University of Oklahoma Press, Norman. MacNeish, R.S. (ed.) 1993 Preliminary Investigations of the Archaic in the Region of Las Cruces, New Mexico. Historic and Natural Resources Report 9. Cultural Resources Management Program, Directorate of the Environment, United States Army Air Defense Artillery Center, Fort Bliss, Texas. Markgraf, Vera, J. Platt Bradbury, R. M. Forester, G. Singh, and R.S. Sternberg 1984 San Agustin Plains, New Mexico: Age and Paleoenvironmental Potential Reassessed. Quaternary Research 22:336-343. Martin, Paul Schultz 1963 The Last 10,000 Years, University of Arizona Press, Tucson. Martin, Paul Schultz, and John B. Rinaldo 1951 The Southwestern Co-Tradition. Southwestern Journal of Anthropology 7:215-219. Martinez-Goytre, Juan 1993 Spatial Variability of Paleoflood Magnitudes in Small Watersheds, Santa Catalina Mountains, Arizona. Master of Science, pre-publication manuscript, Department of Geosciences, University of Arizona, Tucson. Martinez-Goytre, Juan, P. Kyle House, and Victor R. Baker 1994 Spatial Variability of Small-Basin Paleoflood Magnitudes for a Southeastern Arizona Mountain Range. Water Resources Research 30(5):1491-1501. Matson, R.G. 1991 The Origins of Southwestern Agriculture. University of Arizona Press, Tucson. Mayo, E. B., Paul E. Damon, and D. L. Bryant 1968 Stratigraphie and Volcanic Geology, Tucson Mountains: Field trip V, Second Day. In Guidebook III, Geological Society of America Cordilleran Section 64th Annual Meeting, pp. 345-350. 301 McKittrick, Mary Anne 1988 Suificial Geologic Maps of the Tucson Metropolitan Area. Open File Report 88-18. Arizona Geological Survey, Tucson. Mehringer, Peter J. 1967 Pollen Analysis and the Alluvial Chronology. The Kiva 32:96-101. Meltzer, David J. 1991 Altithermal Archaeology and Paleoecology at Mustang Springs on the Southern High Plains of Texas. American Antiquity 56:236-267. 1995 Modelling the Prehistoric Response to Altithermal Climates on the Southern High Plains. In Ancient Peoples and Landscapes, edited by E. Johnson. Museum of Texas Tech University, Lubbock. Miall, Andrew D. 1978 Fluvial sedimentology. Canadian Society of Petroleum Geologists, Memoir 5, Calgary, Canada. 1992 Alluvial Deposits. Facies Models: Response to Sea Level Change, ed. by Roger G. Walker and Noel P. James, pp. 119-142. Geological Association of Canada: Department of Earth Sciences. Memorial University of Newfoundland. St. Johns' Newfoundland, Canada. Minnis, Paul 1985 Domesticating Plants and People in the Greater American Southwest. In Prehistoric food production in North America, edited by Richard I. Ford, pp. 309-340. Anthropological Papers no. 75. Museum of Anthropology, University of Michigan, Ann Arbor. 1992 Earliest Plant Cultivation in the Desert Borderlands of North America. In The Origins of Agriculture, An International Perspective, edited by C. W. Cowan and P. J. Watson, pp. 121-141. Smithsonian Institution Series in Archaeological Inquiry, Smithsonian Institution Press, Washington. Monger, H. C. 1995 Pedology in Arid Lands Archaeological Research: An Example from Southern New Mexico-Western Texas. In Pedological Perspectives in Archaeological Research. Soil Science Society of America Special Publication 44. Madison, Wisconsin. Morisawa, Marie 1985 Rivers: Form and Process. Longman, London. 302 Morris, George D. 1951 Report on the Silverbell Road Sites. Manuscript on file, Arizona State Museum, Tucson. O'Laughlin, T. C. 1980 The Keystone Dam Site and Other Archaic and Formative Sites in Northwest El Paso, Texas. Publications in Anthropology 8, El Paso Centennial Museum, University of Texas at El Paso. Parker, John T. C. 1995 Geomorphology and Hydrology of the Santa Cruz River, Southeastern Arizona. Unpublished Ph.D. dissertation, Department of Geosciences, University of Arizona, Tucson. Pearthree, Phillip A, Mary Anne McKittrick, Garrett W. Jackson, and Karen A. Demsey 1988 Geologic Map of Quaternary and Upper Tertiary Deposits, Tucson 1 ° x 2 ° Quadrangle, Arizona. Open file report 88-21. Arizona Geologic Survey, Tucson Phillips, David A. , Mark C. Slaughter, and Susan B. Bierer 1993 Archaeological Studies at Kartclzner Caverns State Park, Cochise County, Arizona. SWCA Archaeological Report No. 93-26, SWCA, Inc., Environmental Consultants, Tucson. Phillips, M. P. 1976 Geology of Tumamoc Hill, Sentinel Peak, and Vicinity, Pima County, Arizona. Unpublished Master's thesis, Department of Geosciences, University of Arizona, Tucson. Rafferty, J. 1985 The Archaeological Record of Sedentariness: Recognition, Development, and Implications. In Advances in Archaeological Method and Theory, vol. 8, edited by M. B. Schiffer, pp. 113-156. Academic Press, New York. Reeves, Brian O. K. 1973 The Concept of an Altithermal Hiatus in Northern Plains Prehistory. American Anthropologist 75:1221-1253. Reher, Charles A. 1979 Adaptive Process on the Shortgrass Plains. In For Theory Building in Archaeology, edited by L. Binford, pp. 13-40. Academic Press, New York. 303 Ritter, Dale F. 1986 Process Geomorphology. 2d ed. W.C. Brown, Dubuque, Iowa. Rogers, Malcolm J. 1929 Report of an Archaeological Reconnaissance in the Mohave Sink Region. Paper No. 1. San Diego Museum of Man. 1939 Early Lithic Industries of the Lower Basin of the Colorado River and Adjacent Desert Areas. Paper No. 3. San Diego Museum of Man. 1958 San Dieguito Implements from the Terraces of the Rincon-Pantano Wash and Rillito Drainage System. The Kiva 24:1-23. Roth, Barbara J. 1988 Recent Research on the Late Archaic Occupation of the Northern Tucson Basin. In Recent Research on Tucson Basin Prehistory: Proceedings of the Second Tucson Basin Conference, edited by W. H. Doelle and P. R. Fish, pp. 81-86. Anthropological Papers No. 10. Institute for American Research, Tucson. 1989 Late Archaic Settlement and Subsistence in the Tucson Basin. Ph.D. dissertation, Department of Anthropology, University of Arizona, Tucson. 1992 Sedentary Agriculturalists or Mobile Hunter-Gatherers? Recent Evidence on the Late Archaic Occupation of the Northern Tucson Basin. Kiva 57(4):291314. 1995 Late Archaic Occupation of the Upper Bajada: Excavations at AZ AA: 12:84 (ASM), Tucson Basin. Kiva 6(2):189-207. Roth, Barbara J., and Bruce B. Huckell 1992 Cortaro Points and the Archaic of Southern Arizona. Kiva 57:353-370. Sayles, E. B. 1983 The Cochise Cultural Sequence in Arizona. Anthropological Papers of the University of Arizona, No. 42. University of Arizona Press, Tucson. Sayles, E. B., and Ernst Antevs 1941 The Cochise Culture. Medallion Papers No. 29. Gila Pueblo, Globe, Arizona. Schiffer, Michael B. 1986 Radiocarbon Dating and the "Old Wood" Problem: The Case of the Hohokam Chronology. Journal of Archaeological Science 13:13-30. 304 Schmader, M. F. 1996 Archaic Structures of the Middle Rio Grande Valley, New Mexico. Paper presented at the 61st Annual Meeting of the Society for American Archaeology, New Orleans. Schroedl, A. R. 1977 The Grand Canyon Figurine Complex. American Antiquity 42:254-265. Schumm, Stanley A. 1973 Geomorphic Thresholds and Complex Response of Drainage Systems. In Fluvial Geomorphology, edited by Marie Morisawa, pp. 299-310. Publications in Geomorphology, 4th Annual Meeting, State University of New York at Binghampton. 1977 The Fluvial System. Wiley Interscience, New York. Schumm, Stanley A., and R. S. Parker 1973 Implications of Complex Response of Drainage Systems for Quaternary Alluvial Stratigraphy. Natural Physical Science 243:99-100. Simmons, Alan H. 1986 New Evidence for the Early Use of Cultigens in the American Southwest. American Antiquity 51(1):73-89. Simpson, Kay, and Susan J. Wells 1983 Archaeological Survey in the Eastern Tucson Basin, Saguaro National Monument, Rincon Mountain Unit, Tan que Verde Ridge, Rincon Creek, Mica Mountain Areas. Publications in Anthropology No. 22(1). Western Archaeological and Conservation Center, National Park Service, Tucson. Sleazak-Pearthree, Marie, and Victor R. Baker 1987 Channel Change along the Rillito Creek System of Southeastern Arizona 1941 Sliva, Jane 1997 Flaked Stone Artifacts. The Middle Archaic Occupation at Los Pozos, AZ AA:12:91 (ASM) (draft), edited by David A. Gregory, pp. 39-56. Anthropological Papers No. 20. Center for Desert Archaeology, Tucson, in preparation. Smiley, F.E., and William J. Parry 1990 Early, Intensive, and Rapid: Rethinking the Agricultural Transition in the Northern Southwest. Paper presented at the 55th Annual Meeting of the Society for American Archaeology, Las Vegas. 305 Smiley, Terah, Henry F. Dobyns, Bonnie Jones, and James T. Barter 1953 San Jose de Tucson, Its History and Archaeological Exploration. Ms. on file, Arizona State Museum Library Archives (Folder A-271), University of Arizona, Tucson. Smith, W. 1986 The Effects of Eastern North Pacific Tropical Cyclones on the Southwestern United States. NOAA Technical Memorandum, NWS WS-197, 299 pp. Smith, G. E. P. 1938 The Physiography of Arizona Valleys and the Occurrence of Ground Water. Agricultural Experiment Station Technical Bulletin No. 77, University of Arizona, Tucson. Soil Survey Staff 1951 Soil Survey Manual. United States Department of Agriculture Handbook No. 18. U.S. Government Publications Office, Washington, D.C. 1975 Soil Taxonomy: A Basic System of Soil Classification for Making and Interpreting Soil Surveys. U.S. Department of Agriculture Handbook No. 436, Soil Conservation Service, U.S. Government Publications Office, Washington, D.C. Spaulding, W. Geoffrey 1991 A Middle Holocene Vegetation Record from the Mojave Desert of North America and its Paleoclimatic Significance. Quaternary Research 35:427-437. Stafford, Thomas W., Jr. 1986 Quaternary Alluvial Stratigraphy Reconnaissance of the Santa Cruz River, Tucson, Arizona. Report submitted to Cultural and Environmental Systems. Tucson, Arizona. Stark, Barbara 1981 The Rise of Sedentary Life. In Supplement to the Handbook of Middle American Indians: Archaeology, edited by V. Bricker, pp. 345-372. University of Texas Press, Austin. Stein, Julie K. 1993 Scale in Archaeology, Geosciences, and Geoarchaeology. In Effects of Scale on Archaeological and Geoscientific Perspectives, edited by Julie K. Stein and Angela R. Linse. GSA Special Paper No. 283, The Geological Society of America, Boulder, pp. 1-10. 306 Stoltman, J., and D. Baerreis 1983 The Evolution of Human Ecosystems in the Eastern United States. In The Holocene, edited by H. E. Wright Jr., pp. 252-268. Late Quaternary Environments of the United States, vol. 2, edited by H. E. Wright. University of Minnesota Press, Minneapolis. Swartz, Deborah L. 1987 Archaeological Studies of the Avra Valley, Arizona for the Papago Water Supply Project, Vol. 2: Archaeological Site Descriptions. Anthropological Papers No. 9. Institute for American Research, Tucson. Tagg, Martyn D. 1996 Early Cultigens from Fresnal Shelter, Southeastern New Mexico. American Antiquity 61(2):311-324. Tagg, Martyn D., and Bruce B. Huckell 1984 The Sycamore Canyon Sites. In Miscellaneous Archaeological Studies in the ANAMAX-Rosemont Land Exchange Area, by M. D. Tagg, R. G. Ervin, and B. B. Huckell, pp. 61-106. Archaeological Series No. 147(4). Arizona State Museum, University of Arizona, Tucson. Thiel, Homer 1995a Archaeological Testing along the A-Mountain Drainage System. Technical Report 95-6, Center for Desert Archaeology, Tucson. 1995b Archaeological Testing on the Rio Nuevo South Property, Tucson, Arizona. Technical Report 95-11, Center for Desert Archaeology, Tucson. Thompson, D. M., and E. Arthur Bettis, III 1982 Out of Sight, Out of Planning: Assessing and Protecting Cultural Resources in Evolving Landscapes. Contract Abstracts and CRM Archeology 2(3):16-22. Thornthwaite, C. W., C. F. S. Sharpe, and E. G. Dosch 1942 Climate and Accelerated Erosion in the Arid and Semi-arid Southwest, with Special Reference to the Polacca Wash Drainage Basin, Arizona. U.S. Department of Agriculture, Technical Bulletin 808. Upham, Steadman, Richard S. MacNeish, Walton C. Galiant and Christopher M. Stevenson 1987 Evidence Concerning the Origin of Maiz de Ocho. American Anthropologist 89 (2):410-419. 307 Van Devender, Thomas R. 1987 Vegetation History of the Deserts of Southwestern North America: The Nature and Timing of the Late Wisconsin-Holocene Transition. In North America and Adjacent Oceans during the Last De glaciation, edited by W. F. Ruddiman and H. E. Wright, Jr., pp. 323-352. The Geology of North America K-3. The Geological Society of America, Boulder, Colorado. Van Devender, Thomas R., Julio L. Betancourt, and Mark Wimberly 1984 Biogeographic Implications of a Packrat Midden Sequence from the Sacramento Mountains, South-Central New Mexico. Quaternary Research 22:344-360. Wallace, Henry D. 1996 Archaeological Assessment of a Portion of the Knapp Parcel Southeast of the Confluence of the Rillito and Santa Cruz Rivers, Tucson, Arizona. Technical Report 96-2. Center for Desert Archaeology, Tucson. Warren, Claude N. 1967 The San Dieguito Complex: A Review and Hypothesis. American Antiquity 32(2):168-185. Warren, Claude N., and D.L. True 1961 The San Dieguito Complex and Its Place in California Prehistory. In Archaeological Survey Annual Report 1960-1961, pp. 1-105. Department of Anthropology and Sociology, University of California, Los Angeles. Waters, Michael R. 1983 The Late Quaternary Geology & Archaeology of Whitewater Draw, Southeastern Arizona. Ph.D. Thesis. Department of Geosciences, University of Arizona, Tucson. 1985 Late quaternary alluvial stratigraphy of Whitewater Draw, Arizona: Implications for regional correlation of fluvial deposits in the American Southwest. Geology 13:705-708. 1986a The Geoarchaeology of Whitewater Draw. Anthropological Papers of the University of Arizona, No. 45. University of Arizona Press, Tucson. 1986b The Sulphur Spring Stage and Its Place in New World Prehistory. Quaternary Research 25:251-256. 308 Waters, Michael R. 1987 Holocene Alluvial Geology and Geoarchaeology of AZ BB: 13:14 and the San Xavier Reach of the Santa Cruz. In The Archaeology of San Xavier Bridge Site (AZ BB:13:14), Tucson Basin, Southern Arizona, edited by John C. Ravesloot, pp. 39-60. Archaeological Series No. 171. Arizona State Museum, University of Arizona, Tucson. 1988 The Impact of Fluvial Processes and Landscape Evolution on Archaeological Sites and Settlement Patterns along the San Xavier Reach of the Santa Cruz River, Arizona. Geoarchaeology 3(3):205-219. 1989 Late Quaternary Lacustrine History and Paleoclimatic Significance of Pluvial Lake Cochise, Southeastern Arizona. Quaternary Research 32:1-11. 1992 Principles of Geoarchaeology. The University of Arizona Press, Tucson, Arizona. Waters, Michael R., and David Kuehn 1996 The Geoarchaeology of Place: The Effect of Geological Processes on the Preservation and Interpretation of the Archaeological Record. American Antiquity 61(3):483-497. Webb, R. H. 1985 Late Holocene Flooding on the Escalante River, South-central Utah. Unpublished Ph.D. dissertation, Department of Geosciences, University of Arizona, Tucson. Webb, R.H., and J.L. Bentacourt 1992 Climatic variability and flood frequency of the Santa Cruz River, Pima County, Arizona, U.S. Geological Survey Water Supply Paper 2379(40). Whalen, Norman M. 1971 Cochise Culture Sites in the Central San Pedro Drainage, Arizona , Unpublished Ph.D. dissertation, University of Arizona, Tucson. 1975 Cochise Site Distribution in the San Pedro Valley. The Kiva 40:203-211. 1994 Moving Out of the Archaic on the Edge of the Southwest. American Antiquity 59:622-638. 309 Whitelaw, Todd 1991 Some Dimensions of Variability in the Social Organisation of Community Space Among Foragers. In Ethnoarchaeological Approaches to Mobile Campsites, edited by C. S. Gamble and W. A. Boismier, pp.139-188. International Monographs in Prehistory, Ethnoarchaeological Series 1, Ann Arbor, Michigan. Wilcox, David R., and Charles Sternberg 1983 Hohokam Ballcourts and Their Interpretation. Archaeological Series No. 160. Arizona State Museum, University of Arizona, Tucson. Willey, Gordon R., and Philip Phillips 1955 Method and Theory in American Archaeology II: Historical-developmental Interpretation. American Anthropologist 57(4): 723-819. 1958 Method and Theory in American Archaeology. The University of Chicago Press, Chicago. Wills, W. H. 1988a Early Prehistoric Agriculture in the American Southwest. School of American Research Press, Santa Fe, New Mexico. 1988b Early Agriculture and Sedentism in the American Southwest: Evidence and Interpretations. Journal of World Prehistory 2(4): 171-214. 1990 Cultivating Ideas: The Changing Intellectual History of the Introduction of Agriculture in the American Southwest. In Perspectives on Southwestern Prehistory, edited by Paul E. Minnis and Charles L. Redman. Boulder: Westview Press. Windmiller, Ric 1973 The Late Cochise Culture in the Sulphur Springs Valley, Southeastern Arizona: Archaeology of the Fairchild Site. The Kiva 39(2):131-169. Wiicherl, Helga 1997 Faunal Remains. In Excavations in the Santa Cruz River Floodplain: The Middle Archaic Component at Los Pozos (draft), edited by David A. Gregory. Anthropological Papers No. 20, Center for Desert Archaeology, Tucson, pp. 71-88. Wohl, Ellen E., and Yehouda Enzel 1995 Data for Paleohydrology. In Global Continental Paleohydrology, edited by K. J. Gregory, L. Starkel, and V. R. Baker. John Wiley and Sons. 310 Wolman, M. G., and R. Gerson 1978 Relative Scales of Time and Effectiveness of Climate in Watershed Geomorphology. Earth Surface Processes and Landforms 3:189-208. Womack, W. R., and Stanley A. Schumm 1977 Terraces of Douglas Creek, Northwestern Colorado: An Example of Episodic Erosion. Geology 5:72-76 through 1983: Implications for Floodplain Management. Special Paper No. 6. Arizona Bureau of Geology and Mineral Technology, Geological Survey Branch, Tucson. Woodbury, Richard B. 1993 Sixty Years of Southwestern Archaeology: A History of the Pecos Conference. University of New Mexico Press, Albuquerque. Yellen, John E. 1976 Settlement Patterns of the !Kung: An Archaeological Perspective. In Kalahari Hunter-Gatherers: Studies of the !Kung San and their Neighbors, edited by Richard B. Lee and Irven DeVore. Harvard University Press, Cambridge, Massachusetts.
* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project