A MULTIPROXY ANALYSIS OF CULINARY, TECHNOLOGICAL, AND ENVIRONMENTAL INTERACTIONS IN THE NORTHERN GREAT LAKES REGION By Susan M. Kooiman A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Anthropology—Doctor of Philosophy 2018 A MULTIPROXY ANALYSIS OF CULINARY, TECHNOLOGICAL, AND ENVIRONMENTAL INTERACTIONS IN THE NORTHERN GREAT LAKES REGION ABSTRACT By Susan M. Kooiman A novel combination of analytic methods is used to address the decades-long debate about diachronic subsistence, settlement, and social pattern changes during the Woodland period (AD 1 – 1600) in the northern Great Lakes of North America. While some have argued for dietary continuity throughout the regional Woodland, others maintain that certain specific resources—including fish, wild starchy plants, and/or maize—were more intensively exploited over time in reaction to various technological, social, and/or environmental factors. The Cloudman site (20CH6), located on Drummond Island off Michigan’s Upper Peninsula in Lake Huron, is a multicomponent habitation site with two millennia of Middle Woodland, early Late Woodland, and late Late Woodland occupations, as well as a late precontact component characterized by Ontario Iroquois pottery. The ceramic assemblage is therefore ideal for diachronic assessment of alterations in diet and technology in the context of dynamic natural and social environments and is employed as a case study for the multiproxy approach. Ceramic typological classification and AMS dating of pottery residues are used to reconstruct an occupational history of the Cloudman site by which change over time can be evaluated. Functional pottery analysis of technical properties and use-alteration traces reveals that ceramic technology and cooking techniques evolved to facilitate new subsistence and processing needs. Absorbed lipid residue analysis, and microbotanical and stable isotope analysis of adhered carbonized food residue are used in tandem to construct a chronological sequence of culinary practices, which are characterized by both continuity of certain subsistence traditions, such as acorn and aquatic resource consumption, and transformative food choice in response to social and environmental change, including variable exploitation of maize and wild rice. The diversity of the information captured and produced by each method highlights the importance of multiproxy dietary analyses in foodways studies for improving interpretive outcomes. Cooking and pottery technology lend further insight into adaptive decision-making and cultural tradition, and interpretations of past cuisine are further supported and enhanced through comparisons with ethnographic and ethnohistoric accounts of local indigenous cooking and diet. The rich data resulting from the complementary nature of these diverse methods demonstrates a complex interplay of technology, environment, and culturally-based decisions, and underscores the potential applications of such an analytic suite to long-standing problems in the northern Great Lakes and other archaeological contexts worldwide. Copyright by SUSAN M. KOOIMAN 2018 For Mom and Dad “Things we lose have a way of coming back to us in the end, if not always in the way we expect.” – J.K.R. v ACKNOWLEDGEMENTS Although this dissertation bears only my name, it is the result of the effort and support of many others, without whom this body of work would not have been possible. First and foremost, I would like to thank my adviser, William Lovis, for his support, wisdom, patience, and guidance throughout this process. I am also grateful for the constant flow of support, encouragement, and good advice from my committee members, Jodie O’Gorman, Mindy Morgan, and Ryan Tubbs. Jim Skibo has continued to be an inspiring mentor and collaborator, and Lynne Goldstein has been a valued mentor to me during my tenure at MSU. I am indebted to Rebecca Albert for all her hard work in contributing to this project, and I am even more thankful for her friendship. The prior works of Sean Dunham and Chris Stephenson were essential building blocks for my research and saved me a great deal of work, so I am extremely grateful for these intelligent and gracious colleagues and friends. I owe deep gratitude to Eric Drake and Mary Malainey for their invaluable contributions, and I must also thank Timothy Figol of the Residue Analysis Laboratory at Brandon University and Shari Effert- Fanta of the Illinois State Geological Survey for their vital contributions to this research. I am grateful to the MSU Department of Anthropology for laboratory space for my analysis, and to the MSU Museum for access to the Cloudman pottery assemblage. My research would not have been possible without funding and support from the National Science Foundation, the MSU Graduate School, the MSU College of Social Science, and the MSU Alumni and Friends of Archaeology Fund. I would be nothing without the constant love and support of my family and friends. To my siblings—Scott, Keith, Julie, Dan, and Joe—we have been through a lot together and I am vi ever grateful for your presence in my life, along with Monica, Holly, Jake, Abby, Emma, and many other loving family members who are numerous to name here. I am grateful to Erin Beachey for a lifetime of friendship and for her graphic design skills that created the beautiful maps in this manuscript. I was lucky enough to be able to move through this process alongside Caitlin Vogelsberg, Nicole Geske, and Kate Frederick from the very beginning, and their friendship kept me sane and grounded. Amy Michael, Mari Isa, Jack Biggs, Micca Metz, Emma Meyer, Lisa Bright, Josh Burbank, and Jeff and Autumn Painter have made graduate school a bearable (and even enjoyable!) experience and for that I am forever thankful. Finally, I would like to thank my parents, Calvin and Elaine, who instilled in me the strength, determination, and a work ethic required to make it through the demands of achieving my goals, and who always supported my dreams. This dissertation is dedicated to their memory. vii TABLE OF CONTENTS LIST OF TABLES xii LIST OF FIGURES xv CHAPTER 1 1 INTRODUCTION 1 Introduction 1 Research Context and Problem 2 Research Objectives 4 An Integrated Theoretical Framework 5 Pottery Function 6 Foodways 9 Pottery Style 13 Research Questions 15 Dissertation Organization 17 CHAPTER 2 19 ENVIRONMENTAL AND CULTURAL HISTORY/BACKGROUND OF THE NORTHERN GREAT LAKES 19 Introduction 19 Paleoclimate and Environmental Background 20 Cultural Background: Overview of Upper and Northern Great Lakes Archaeology 22 Pre-Woodland Periods 22 Woodland Period 23 Iroquoian 29 The Protohistoric/Contact Period 32 Conclusion 35 CHAPTER 3 37 SUBSISTENCE AND TECHNOLOGY IN THE NORTHERN GREAT LAKES 37 Introduction 37 Research Problem 38 Case Study: Cloudman Site (20CH6) 45 Research Expectations 50 Conclusion 56 CHAPTER 4 57 METHODS AND DATA COLLECTION 57 Introduction 57 The Cloudman Pottery Assemblage 57 Principals of Pottery Function 58 Functional Analysis of Cloudman Pottery 59 viii Intended Function 59 Actual Function 61 Ceramic Taxonomic Classification 63 Microbotanical Analysis 64 Stable Isotope Analysis 69 Lipid Residue Analysis 71 AMS Dating 73 Soil Samples 75 Conclusion 78 CHAPTER 5 79 REGIONAL CERAMIC TAXONOMY AND CHRONOLOGY, AND THE OCCUPATIONAL HISTORY OF THE CLOUDMAN SITE 79 Introduction 79 Pottery Taxonomy 79 Taxonomic Classification 81 Middle Woodland Ceramic Subassemblage 82 Laurel Ware 82 North Bay 83 Untyped 84 Middle Woodland/Late Woodland Transitional Subassemblage 84 Early Late Woodland Ceramic Subassemblage 86 Mackinac Ware 86 Blackduck Ware 87 Bowerman Ware 88 Untyped 88 Early/Middle Late Woodland Transition 88 Middle Late Woodland Ceramic Subassemblage 89 Bois Blanc Ware 89 Late Late Woodland Ceramic Subassemblage 90 Juntunen Ware 90 Traverse Ware 91 Untyped 92 General Late Woodland Ceramic Vessels 93 Ontario Iroquoian Ceramic Subassemblage 93 Early Ontario Iroquoian 93 cf. Huron Incised 94 cf. Ripley Plain 94 cf. Lawson Ware 95 Untyped 96 Unidentified Affiliation 96 Miniature Vessels 96 Middle Woodland 97 Early Late Woodland 98 Late Late Woodland 98 Ontario Iroquoian 99 ix Pottery Age 99 AMS Dates 99 Relative Dating 101 Conclusion 105 CHAPTER 6 106 POTTERY FUNCTION 106 Introduction 106 Technical Properties and Intended Function 107 Temper Size 107 Rim Diameter 109 Vessel Thickness 112 Synchronic Technical Variation 117 Intended Function Summary 120 Use-Alteration Traces and Actual Function 121 Exterior Sooting 122 Exterior Carbonization 122 Interior Carbonization 123 Habitual Cooking Behaviors 124 Vessel Fill Levels 124 Interior Carbonization Patterns 126 Synchronic Variation of Interior Carbonization Patterns 133 Actual Function Summary 134 Discussion 135 Conclusion 138 CHAPTER 7 139 FOOD SELECTION AND COOKING AT THE CLOUDMAN SITE 139 Introduction 139 Carbon and Nitrogen Stable Isotope Analysis 139 Nitrogen Isotopes 140 Carbon Isotopes 149 Lipid Residue Analysis 150 Microbotanical Analysis 157 Tandem Dietary Analysis Results 165 Seasonality at the Cloudman Site 167 Methodological Considerations 168 Aquatic Resources, Acorns, Lipid Residue Analysis, and Stable Isotope Analysis 169 Maize, Stable Isotope Analysis, and Microbotanical Analysis 171 Discussion 172 Conclusion 177 CHAPTER 8 179 ETHNOGRAPHIC AND ETHNOHISTORIC ACCOUNTS OF DIET AND COOKING 179 Introduction 179 x Fish 180 Acorns 182 Maize 183 Wild Rice 184 Squash 186 Other Foods 188 Cooking and Cuisine 191 Conclusion 195 CHAPTER 9 196 CONCLUSIONS 196 Introduction 196 Context and Chronology of the Cloudman Site 196 Research Questions and Results 198 Methodological Importance 211 Future Research 213 Conclusion 215 APPENDICES 217 218 Lipid Residue, and Stable Isotope Analyses 243 Illinois State Geological Survey (ISGS) Reports 249 APPENDIX D: Lipid Residue Analysis Report 252 APPENDIX E: Microbotanical Analysis Data 293 APPENDIX F: Stable Isotope, Microbotanical, and Lipid Residue Analysis Results by Vessel 296 APPENDIX G: Select Vessels from the Cloudman Pottery Assemblage 300 REFERENCES 334 APPENDIX C: Carbon and Nitrogen Stable Isotope Analysis: Summary of APPENDIX A: Cloudman Pottery Data APPENDIX B: Cloudman Site Pottery Residue Samples for Microbotanical, xi LIST OF TABLES Table 1.1: Northern Great Lakes Chronology 5 Table 4.1: Vessels Sampled for Microbotanical Analysis, Lipid Residue Analysis, Stable Isotope Analysis, and AMS Dating 66 Table 4.2: Pottery Vessels Sampled for AMS Dating of Carbonized Residue 74 Table 4.3: Soil Samples from the Cloudman Site Selected for Stable Isotope Analysis 77 Table 4.4: Soil Samples from the Cloudman Site Selected for Lipid Residue Analysis 78 Table 5.1: Cloudman Pottery Vessels by Socio-Temporal Association 81 Table 5.2: Middle Woodland Vessels by Type 83 Table 5.3: Miscellaneous Woodland and Unknown Vessels 85 Table 5.4: Early Late Woodland Vessels by Type 88 Table 5.5: Late Late Woodland Vessels by Type 91 Table 5.6: Ontario Iroquoian Vessels by Type 94 Table 5.7: Miniature Vessels by Type 97 Table 5.8: AMS Dates from Carbonized Pottery Residue Samples 100 Table 5.9: Occupational History of the Cloudman Site, Derived from Relative and Direct Dating of Pottery 104 Table 6.1: Mean Temper Size by Subset 109 Table 6.2: Temper Size Relationships (Welch’s Unpaired T-Test) 109 Table 6.3: Mean Rim Diameter by Subset 110 Table 6.4: Rim Diameter Relationships (Welch’s Unpaired T-Test) 111 Table 6.5: Vessel Wall Thickness by Subset 114 Table 6.6: Neck Thickness Relationships (Welch’s Unpaired T-Test) 114 xii Table 6.7: Shoulder Thickness Relationships (Welch’s Unpaired T-Test) 115 Table 6.8: Body Thickness Relationships (Welch’s Unpaired T-Test) 115 Table 6.9: Corrected Thickness (Thickness/Rim Diameter) 115 Table 6.10: Corrected Vessel Neck Thickness Relationships (Welch’s Unpaired T-Test) 116 Table 6.11: Corrected Average Neck + Shoulder Thickness Relationships (Welch’s Unpaired T-Test) 116 Table 6.12: Technical Properties of Vessels by Type/Ware 120 Table 6.13: Frequency of Use-Alteration Traces by Subset 123 Table 6.14: Interior Carbonization Pattern Frequency by Subset 130 Table 6.15: Primary Interior Carbonization Pattern Frequency by Subset 131 Table 6.16: Interior Carbonization Pattern Relationships (Kruskal-Wallis) 133 Table 6.17: Interior Carbonization Pattern Frequency by Type/Ware 134 Table 7.1: Mean δ15N and δ13C Values of Cloudman Pottery Residues by Subset 140 Table 7.2: δ15N Values, Cloudman Soil Samples 142 Table 7.3: Relationship between δ15N Values of Cloudman Pottery Residue Samples and Other Archaeological and Biological Samples (Unpaired T-test) 144 Table 7.4: Frequencies of Lipid Categories by Subset 151 Table 7.5: Vessel Clusters by Lipid Content (Jaccard’s Coefficient) 156 Table 7.6: Cloudman Site Soil Sample Lipid Content 157 Table 7.7: Number of Vessels Containing Maize, Wild Rice, and Squash Microbotanicals by Subset 158 Table 7.8: Frequencies of Microbotanical Species by Subset 161 Table 7.9: Microbotanical Frequency Relationships between Subsets (Kruskal-Wallis) 162 Table 7.10: Vessel Clusters by Microbotanical Species Content (Jaccard’s Coefficient) 164 xiii Table 7.11: Vessel Cluster by Lipid and Microbotanical Content (Jaccard’s Coefficient) 167 xiv LIST OF FIGURES Figure 2.1: Distribution of Cultural Groups ca. 1630 (adapted from Trigger 1976:92) 33 Figure 3.1: Location of the Cloudman (20CH6) Site and Other Woodland Sites 46 Figure 4.1: Example Interior Carbonization Pattern Categorizations (Kooiman 2012, 2016) 62 Figure 4.2: Terracing at the Cloudman Site (20CH6) 76 Figure 6.1: Rim Diameter Frequencies of the Cloudman Pottery Assemblage 110 Figure 6.2: Interior Carbonization Patterns for the Cloudman Site (20CH6) Pottery Assemblage: a) Type 1 (boiling); b) Type 2 (stewing); c) Type 3 (boiling + stewing); d) Type 4 (possible boiling or stewing; e) Type 5 (no discernable pattern) 127 Figure 6.3: Interior Carbonization Pattern 1 (Boiling) 128 Figure 6.4: Interior Carbonization Pattern 2 (Stewing) 128 Figure 6.5: Interior Carbonization Pattern 3 (Boiling + Stewing) 129 Figure 6.6: Proportions of Interior Carbonization Patterns by Subset 132 Figure 7.1: Plot of δ15N/δ13C Values of Cloudman Pottery Residues 141 Figure 7.2: Plot of δ15N/δ13C Values of Cloudman Pottery Residues vs. Soil Samples 142 Figure 7.3: Plot of δ15N/δ13C Values of Archaeological Faunal Samples, Kelly-Campbell Site, Ontario (Katzenberg 1989) 144 Figure 7.4: Plot of δ15N/δ13C Values of Cloudman Pottery Residues vs. Archaeological Fish Remains from Southern Ontario (Van der Merwe et al. 2003) and Belgium (Fuller et al. 2012) 145 Figure 7.5: Plot of δ15N/δ13C Values of Cloudman Pottery Residues vs. Modern Fish Samples from Lake Michigan (Turschack 2013) 146 Figure 7.6: Plot of δ15N/δ13C Values of Cloudman Pottery Residues vs. Human Bone Collagen of Woodland & Iroquoian Individuals with Aquatic Resource-Rich Diets (Brandt 1996; DeWar et al. 2010; Schwarcz et al. 1985; Vander Merwe et al. 2003) 147 xv Figure 7.7: Plot of δ15N/δ13C Values of Cloudman Pottery Residues by Subset 149 Figure 7.8: Primary Lipid Categories by Subset 152 Figure 7.9: Dendrogram Showing Average Linkage (Between Groups) of Vessels by Lipid Content using Jaccard’s Coeffient (vessel numbers: left column; distance: right column) 155 Figure 7.10: Microbotanical Frequencies of Maize, Wild Rice, and Squash by Subset 161 Figure 7.11: Dendrogram Showing Average Linkage (Between Groups) of Vessels by Microbotanical Content using Jaccard’s Coefficient (vessel numbers: left column; distance: right column) 163 Figure 7.12: Dendrogram Showing Average Linkage (Between Groups) of Vessels by Lipid and Microbotanical Content using Jaccard’s Coefficient (vessel numbers: left column; distance: right column) 166 xvi CHAPTER 1 INTRODUCTION Introduction Food and cooking are vital components of human survival and culture. Understanding subsistence-related behaviors and attendant technologies is important for unveiling the lifeways of past societies because of their close association to adaptive decisions, identity, social and political relationships, and ideologies. Food remains and pottery are both widely-studied artifact categories, and new analytic techniques for extracting increasing amounts of information from these data are constantly being developed and refined. Many such methods are used independently without employing them together, despite the oft-complementary nature of the data each yields. Expansion of routine pottery and dietary analyses to include a variety of analytic methods holds potential to improve interpretations of sites with limited preserved archaeological materials and amplify evidence for adaptive and social behaviors at archaeological sites across the globe. The intersection of foodways and pottery is a promising arena for multidimension research that could construct more robust interpretations about the past. The need for multidimensional analysis of this kind is exemplified in the northern Great Lakes (i.e., Lake Superior, and northern Lake Michigan-Huron) of North America, where mobile groups left behind limited archaeological remains and preservation of organic materials is generally poor. While some scholars have argued for a transition from broad spectrum hunting- gathering to intensification of certain wild and/or cultivated resources during the Woodland period (200 BC – AD 1600), others have argued for greater or complete continuity in settlement and subsistence patterns throughout the Woodland period. The dissertation examines the issue of 1 hypothesized changing settlement and subsistence patterns from the perspective of food processing technology, food/resource selection, and cooking methods based on a ceramic assemblage from the multicomponent Cloudman (20CH6) site near the Straits of Mackinac in the northern Lake Huron basin. The results will contribute to the growing body of data about precontact northern Great Lakes dietary behaviors and provide clarification about diachronic behavioral change in the context of social and environmental factors. They will also demonstrate the effectiveness of a novel combination of methods for examining ancient cuisine and culinary technology. Research Context and Problem For decades, researchers have hypothesized, based on an increasing body of data, that subsistence and settlement strategies in the northern Great Lakes underwent significant changes from the Middle Woodland (200 B.C. – A.D. 500/600) to the Late Woodland (A.D. 500/600 – A.D. 1600) periods. This includes apparent intensification of aquatic resources (Cleland 1982; Smith 2004) and certain starchy resources, such as acorns and wild rice (Dunham 2014) or maize (O’Shea 2003). Others have argued for continuity of subsistence regimes throughout the entire Woodland period (Martin 1989). Observed diachronic changes in cooking habits, evident through carbonized food patterns on ceramic cooking vessels, could corroborate the hypothesized subsistence shifts in the Late Woodland period (Kooiman 2012, 2015, 2016). If a change in diet occurred, the resources that became the foci of intensification, as well as the timing of this intensification, remain unclear. Many have also connected the Late Woodland intensification of resources to social changes simultaneously expressed through pottery style (Brose and Hambacher 1999; Dunham 2 2014; McHale Milner 1998; O’Shea and McHale Milner 2002). Increasing stylistic heterogeneity throughout the Late Woodland period signal increasingly bounded group identity and localization, requiring resource intensification in the face of decreased geographic ranges of exploitation. The formation and expansion of Iroquoian groups to the east and later arrival of Europeans into the area further complicate the sociocultural and dietary history of the region. Precontact northern Great Lakes subsistence has rarely been discussed from the perspective of foodways, which contextualizes subsistence behaviors within broader social, political, and ideological behavior (Twiss 2012). This perspective requires a wide range of inference drawn from and situated within a wide array of complementary evidence. While this study focuses on ceramic technology and the physical and proxy evidence of foods found in direct context with this technology, these evidences will be discussed within the broader social context of the northern Great Lakes, primarily through comparisons with ceramic typologies. The intersection of subsistence and ceramic technology has long been discussed by archaeologists working in the northern Great Lakes, but recent work has explored the topic from the perspective of pottery function (Kooiman 2012, 2016; Skibo et al., 2009). Elsewhere in the Eastern Woodlands, alterations in pottery construction and composition have been connected to shifts in subsistence strategies (Braun 1983; Hart 2012; Pierce 2005). The initial adoption and subsequent alterations in vessel form and ceramic paste recipes (Chivis 2016, Stoltman 2001) could therefore be useful for distinguishing technological change enacted to accommodate new cooking and dietary requirements. 3 Research Objectives This dissertation explores the long-standing issues of Woodland and Protohistoric northern Great Lakes subsistence while demonstrating the effectiveness of examining variation and change in subsistence habits and social relationships from the perspective of food processing technology, resource selection, and cooking methods. The overarching question of this research is: Do pottery technology, pottery use, diet, and cooking habits change over time, and if so, how do these changes relate to hypothesized transitions in subsistence, settlement, and social patterns among pottery-making groups in the northern Great Lakes region? Food residues extracted from pottery are analyzed both microscopically and chemically. Microbotanical, lipid residue, and stable isotope analyses are employed to examine various dimensions of the food types and recipes prepared in ceramic cooking vessels. Functional pottery analysis, which includes assessment of both technical properties and use-alteration traces, is used to assess both technological adaptations related to food cooking requirements and the methods of cooking employed by pottery-producing precontact and possible protohistoric groups. Results of the dietary and technological analyses are compared to taxonomic pottery classifications to evaluate the possible relationship of food and cooking with social identity, and then considered in context with observed culinary behaviors of local indigenous groups detailed in ethnographic and ethnohistoric accounts to enhance interpretations of past cuisine. This unique combination of ceramic and dietary analytic methods will be applied to the ceramic assemblage from the multicomponent Cloudman site, located on Drummond Island in Lake Huron. Periodically occupied during the Middle Woodland, Late Woodland, and possible Protohistoric periods (Table 1.1), it is the ideal site for making observations on a range of variables in one place over a long period of time, thereby minimizing the effects of distance and 4 social, political, or subsistence variation on the observed changes and allowing diachronic changes in pottery function to be related to broader adaptive and social changes. Excavation of the Cloudman site also produced a substantial pottery assemblage of suitable size for this type of analysis. Table 1.1: Northern Great Lakes Chronology Period Middle Woodland Late Woodland early Late Woodland middle Late Woodland late Late Woodland Protohistoric/Contact Dates 200 BC – AD 500/600 AD 500/600 – 1600 AD 500/600 - 1000 AD 1000 - 1200 AD 1200 – 1600 AD 1600 - ca. 1700 Examining the ceramic assemblage from the Cloudman site from the perspectives of foodways and technology demonstrates the effectiveness of multidimensionally constructed research for providing clarification to long-standing questions using existing archaeological collections. This research, therefore, demonstrates the importance of analyzing old data from different perspectives to bring to light new information. An Integrated Theoretical Framework This research employs multiple methods to answer a set of diverse yet related questions, thereby necessitating the integration of several theoretical perspectives. In this case, pottery function, foodways, and pottery style are the overarching theoretical frameworks upon which the data will be laid and the interpretations constructed. 5 Pottery Function Ceramic analysis has traditionally been focused on style because of its perceived social and temporal symbolism. Linton (1944) was among the earliest proponents of viewing pottery not for its symbolic and socio-ideological meaning, but instead for its role as a utilitarian household item. He states that an effective cooking pot must have certain physical characteristics to properly serve its functions, such as a mouth large enough to prevent violent overboiling and to allow access, but small enough to prevent all liquid from evaporating during the boiling process. However, Linton’s contemporaries mostly ignored his argument and carried on with traditional stylistic analyses, largely to serve as frameworks for constructing chronologies and cultural spatiotemporal units prior to the advent of absolute dating. Another mid- 20th century archaeologist, Anna O. Shepard, agreed that typological categories were the creation of archaeologists rather than of the culture being studied, and it was therefore “strange that pottery should be studied without considering its relations to the people who made it” (1968:309-310). She encouraged attention to the materials and construction techniques used to make pottery and championed the use of temper and paste properties for its classification, which she believed were better indicators of social relationships than were subjective categories. Shepard began a long tradition of compositional studies of ceramic assemblages worldwide. Her area of expertise, petrographic analysis, was practiced more commonly in Europe than in North America but become more popular in the latter following Stoltman’s (1989) revised methodology (Rice 1996b). Compositional studies have now expanded to include a variety of geochemical methods for identifying the composition of both clays and inorganic temper inclusions, such as x-ray fluorescence (e.g., Ligman 2013; Morgenstein and Redmount 2005; Tykot 2016), neutron activation (e.g., Falabella et al. 2013; 6 Glascock and Neff 2003; Wallis et al. 2016), and laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) (e.g., Druc et al. 2017; Duwe and Neff 2007; Stoner 2016). A handful of archaeologists began investigating ceramic vessel function following Binford (1965), who distinguished between primary function (the specific use made of the vessel) and secondary function (the by-product of the social context of a vessel’s manufacturer). Hally (1983) suggested that primary function could be explored through physical properties, archaeological contexts, and the alteration of vessels through use. He cites early explorations of vessel wall absorption of phosphorous and fatty acids (Condamin et al. 1976), surface accumulations of carbonized food residues (McPherron 1967), and the breakdown of vessel surfaces (Griffiths 1978; Matson 1965), which he synthesizes as explorations of ceramic vessel “use alteration.” Braun (1983) helped introduce the most recent wave of interest in pottery function, calling for archaeologists to begin to view “pots as tools,” made by people to be used for a variety of functions beyond symbolizing social identity. Focused study of pottery can reveal manufacturing processes, and Braun urged the examination of mechanical performance characteristics to recognize why a vessel was constructed. He emphasized that the physical properties of a pot could have been controlled by potters in order to achieve certain desired performance characteristics; in other words, the technical characteristics of a pot have the potential to reveal what functions it was designed to fulfill. A series of studies in the interceding decades have tested the performance characteristics of various physical properties. Rye’s (1976) initial investigation into the effects of temper on various stages of manufacture set the stage for other temper studies. Bronitsky and Hamer (1986) tested the effects of tempering materials for impact and thermal shock resistance, finding all 7 smaller-grained temper improved ceramic wall strength, and that shell temper was better for shock resistance than grit temper. Feathers (2006) found that although shell temper increases workability and overall vessel strength, its adoption in Eastern North America was not widespread until widescale agriculture was practiced, limiting fuel for fire and requiring pottery that could be fired at lower temperature, which shell tempering improves. Upton et al. (2015) found through experimental work that shell tempering did not have a nixtamalizing effect on contents cooked in vessels, as had been hypothesized by some due to its proliferation alongside maize agriculture. Schiffer and Skibo (1987) adopted and expanded Braun’s ideas under the “behavioral archaeology” theoretical framework. They refer to decisions made by potters as “technical choices,” which cater to and gravitate towards the most desirable performance characteristics, sometimes at the expense of other characteristics. Performance characteristics for ceramic vessels include ease of manufacture, cooling effectiveness, heating effectiveness, portability, impact resistance, thermal shock resistance, and abrasion resistance, which can be influenced by physical characteristics such as vessel size, wall thickness, vessel shape, paste composition, temper density, temper type, etc. (Schiffer and Skibo 1987). Behavioral archeologists conducted a set of experimental studies to explore performance characteristics of multiple technical choices. They found that surface treatment, such as slips and resins applied to the surfaces of vessels, increase heating effectiveness (Schiffer 1990) but have variable effects on abrasion resistance (Skibo et al. 1997). Organic temper, the earliest type used across the world, is thought to have been used for expedient pottery manufacture, but was later replaced by inorganic tempers as a result of its poor performance characteristics (Skibo et al. 1989). 8 The physical characteristics imbued by a potter onto a vessel determine what Skibo (1992, 2013) calls “intended function” (see also Schiffer and Skibo 1987, 1997). Skibo argues that while knowing a vessel’s intended function is useful, archaeologists must also consider its actual function, or the ways in which people made use of a vessel, regardless of its intended function and physical capabilities. This is best achieved by looking at Hally’s (1983) “use- alteration traces.” According to Skibo, these include sooting, carbonization (charred food residue), attrition, and absorbed residues. These traces provide a more refined perspective of pottery use, allowing inferences about use over fire, cooking practices, cuisine, and more. Foodways The study of foodways has its roots in folklore studies and is a growing topic of cross- disciplinary interest in anthropology. Within archaeology, food-related research was traditionally concerned with “diet” and “subsistence,” focusing on human acquisition of necessary nutrients to ensure survival. The terms “foodways” and “cuisine” set past food production, preparation, and consumption in context with politics, ideologies, and economies (Twiss 2012:357). Examining food from an adaptive, environmental perspective has been the traditional avenue for dietary studies. Perhaps the most famous model for hunter-gatherer adaptive subsistence behavior is Binford’s (1980) foraging/collecting spectrum. He identifies two major poles of subsistence-settlement strategies: foraging, which entails high residential mobility (frequent residential relocation based on resource abundance/availability) and daily procurement of food resources; and collecting, which employs logistical mobility (lower residential mobility with base camps from which task groups disperse on logistical forays) and some form of food storage. Foraging and collecting represent two extreme ends of a continuum of strategies 9 employed by hunter-gatherers, and where groups land on this spectrum (both inter- and intra- annually) is often dependent to a large degree on environment (particularly resource distribution across time and space) and competition for resources (Binford 1980). In temperate zones, seasonal fluctuations in resource availability require regular short- term predictive risk-buffering to counteract potential caloric and nutritional shortfalls. Speth and Spielman (1983) found that diets consisting primarily of lean protein caused heightened metabolism, requiring increasing amounts of food to maintain proper energy intake. This is important in the lean time of late winter/early spring, when plant food sources are low and wild animals are at their leanest, the consumption of the latter leading to what some hunter-gatherer groups call “rabbit starvation” (Speth and Spielman 1983). Carbohydrates and fats both counteract this effect but carbohydrates are more effective. The role of carbohydrates is critical in predictable short-term food shortage risk buffering, and explains the importance of selecting foods for storage. The Sami of subarctic Scandinavia use the inner bark of Scots pine to add carbohydrates and fiber to their diet throughout the year and counteract the effects of protein starvations (Bergman et al. 2004). Early use of seed crops in the Eastern Woodlands has also been linked to carbohydrate storage for consumption during lean periods (Gremillion 1996, 2004). Not only does a balance of time, energy, and efficiency influence food choice, but so does nutritional balance. Maize and beans were grown and consumed together by many late prehistoric (post-A.D. 1300) North American indigenous societies. The two crops grow well together, and beans provide protein otherwise lacking in maize-based diets (Hart 2008; Hart and Scarry 1999; Hart et al. 2002; Monaghan and Parker 2014). Maize is present rather early in portions of the Great Lakes (ca. cal 200 BC; Albert et al. 2018), and may be related to an 10 intensification of wild rice exploitation, which like beans, is relatively rich in amino acids and would have also been a good complement to the nutritionally devoid maize (Hart and Lovis 2013). The selection of food for nutritional needs is clearly a complex process involving a mosaic of decisions, and it is further complicated by the addition of cultural factors. The intersection between biological food requirements and cultural food preferences is best summarized by the “omnivore’s paradox,” which states that humans have incredible freedom and adaptability in food choices, but they also mistrust new foods. A social/cultural group’s traditions of food choice and preparation (i.e., cuisine) work to resolve an omnivore’s anxiety by limiting options and setting parameters for preparation (Fischler 1988:277-279). The parameters are culturally defined and vary based on accessible resources, cultural traditions, and acceptable behaviors. Food is a culturally defined term and therefore what is conceptually accepted as food varies from group to group. Consequently, cuisine, or food culture, is used to differentiate between economic classes, ethnicity, gender, and religious beliefs, serving as a proxy representation of social diversity and identity (Twiss 2012). Food is also an integral part of everyday life. It is an avenue by which to socialize and unite families and communities, making food a physical need fulfilled in social contexts (Atalay and Hastorf 2006). Food can serve “diametrically opposed semiotic functions” that act to both unify and divide; they can create a sense of belonging or emphasize group differences, both of which solidify identity (Appadurai 1981; M. Smith 2006). It can serve as both a cultural mediator but also as a device for “othering” (Montanari 2006), especially between groups defined by ethnicity (Jones 1997; Kalčik 1984; e.g., Barrett et al. 2001; Egan-Bruhy 2014; Scott 1996). Even when common foods are shared, food preparations and cooking methods can be used to distinguish 11 members of a community from those of other groups (Beoku-Betts 1995) or from others of differing ethnicities or classes within their own communities (Chase 2012) as an act of dietary identity formation. Like food selection, cooking is “related in complex and varied ways to issues of gender, work, politics, economic life, and social differentiation” (Rodríguez-Alegría and Graff 2012). Cooking techniques are often aimed at maximizing digestibility and nutritional value of foodstuffs (Wandsnider 1997), but cultural traditions or taste preferences are not always synchronous with nutritional maximization (Rodríguez-Alegría and Graff 2012). Cooking is directly related to food production, food collection, and the manufacture of cooking tools, all factors critical to the overall economy (Rodríguez-Alegría and Graff 2012), and the tools involved in cooking can also be symbolic of identity (MacLean and Insoll 1999). This situates cooking as a useful lens through which to view the intersection of resource selection and ceramic technology. Changes in food selection and subsistence strategies are traditionally couched in terms of adaptation to new or changing environments, climatic changes, or increasing population pressure (Binford 1968; Childe 1936; Flannery 1973), but some archaeologists have begun to investigate dietary shifts from social, political, and ideological perspectives. Spielmann (2002) has argued that the mechanism driving the intensification of food production in small-scale societies is neither economic nor political, but communal ritual activity, specifically feasting and craft specialization. She claims that demand for items critical for communal ritual participation common among many small-scale societies was more critical for the development of agricultural intensification than subsistence provisioning. Hastorf and Johannessen (1994) argue that the timing and patterning of the intensification of maize production and consumption in both North 12 and South America was not advantageous economically but instead linked to the cosmological importance of maize. In their view, maize only increased in use alongside rapid elaboration of social hierarchy, possibly because it was used by political officials for its ritual significance. Food is a biological need, selected because of its availability, taste, abundance, and nutritional value. Yet these choices are often restricted and/or influenced by cultural factors, such as identity, cultural tradition, and ideology. This is important to consider even when studying small-scale societies for whom markers of social affiliation and cosmological beliefs can be difficult to discern. Pottery Style Pottery function analysis and food analyses are applied within a broader foodways framework, and as such they must be discussed in the context of social behavior including various scales and kinds of interactions. Among the mobile hunter-gatherer (-fisher) societies of the northern Great Lakes, this is often achieved via comparisons of pottery style (see Brose 1970; Dorothy 1978, 1980; Janzen 1968; McPherron 1967; Richner 1973; Stoltman 1973). Style has long been used in time-space systematics, in which style denotes both temporal affiliation and geographic spread (Sackett 1977). Given that style is connected to the time and space within which human groups behave, it is often considered symbolic, whether actively or passively, of ethnicity or identity and has been used to separate groups into “cultures” or “ethnicities” based upon ceramic stylistic differences (Peelo 2011; Rice 1996a). The conflation of archaeologically defined stylistic types and past identities/ethnicities has been criticized over the decades (Rice 1996; Sackett 1977; Shepard 1968), although geographic variation in ceramic vessel style does 13 suggest that style is influenced by social interaction (McHale Milner 1998; O’Shea and McHale Milner 2002). Wobst (1977) states that style, independent of technomic function (Binford 1962), serves the function of symbolic communication and information exchange. Symboling, a learned behavior, allows individuals to interact with their social environment through the medium of artifacts, often in the expression of identity through style (Wobst 1977:320). Items which are the most visible and used in arenas of interaction with other groups are the most likely to be used to communicate group affiliation, while less visible items, particularly those used within the home, are less likely to be used to carry such messages and will instead display clinal variation. Weissner (1983) identifies two types of style: emblemic, which transmits a clear message about group affiliation or identity; and assertive, which is formal variation that carries information about personal identity, either consciously or unconsciously, but which does not directly symbolize identity. Within small-scale societies with somewhat fluid, cooperative risk- buffering relationships, the need to distinguish between groups symbolically (using emblemic style) is not as strong as in larger-scale societies with increased competition for resources. Instead, hunter-gatherer artifacts are more likely to transmit assertive style (Weissner 1983:258). Her study of San metal projectile points demonstrated stylistic differences only between non- interacting linguistic groups, which she attributes not to conscious distinction between groups but to non-overlapping social spheres. Thus, with many utilitarian items, such as projectile points and domestic pottery, style is best for interpreting spheres of interaction and less useful for more fine-grained relationships among cooperative small-scale societies. McHale Milner (1998) and Carroll (2013) found that among pottery from the Straits of Mackinac and Southwest Michigan (respectively), the degree of visibility of stylistic elements 14 transmit information in accordance with social distance. Low-visibility elements, such as lip shape, were shared among local groups, while higher visibility elements, such as rim shape and decoration, were shared among geographically broad but interacting groups. In both cases, stylistic similarities of high- vs. low-visibility items correlated with multiple scales of group interaction based on geographic/social distance. According to Wobst (1977) and Weissner (1983), utilitarian pottery, because of its use in the household and subsequent low public visibility, is a medium unlikely to carry strong, emblemic messages about identity. In small-scale societies, media displaying style, such as pottery and lithics, are more likely to carry information about breadth of social interaction. In regions and time periods with high social fluidity, pottery is even less likely to convey strong or distinct markers of identity. Within societies with more restricted social boundaries and/or increased resource competition, pottery styles might become more geographically restricted and more diverse within a given region (Carroll 2013; McHale Milner 1998; Wobst 1977). In these contexts, pottery style might not be a consciously created symbol of identity but rather an assertive representation of the interactive group, in which craft traditions are shared with little outside influence, such as within the matrilineal longhouses in late precontact Iroquoia (Hart and Brumbach 2009). Research Questions Although disparate in their theoretical foci, these perspectives can be used together to effectively answer how pottery technology, pottery use, diet, and cooking habits change over time and explore their relationship to social and environmental transformations in the precontact northern Great Lakes. Pottery function plays an integral role in understanding cooking and 15 cuisine, while pottery style can facilitate recreation of ancient social environments and build interpretations of ancient foodways. Applied to the specific problem of evolving Woodland and late precontact lifeways in the northern Great Lakes of North America, the multi-faceted perspective can extract additional data from limited archaeological remains and bring to light new information about past human behavior. Given this framework, the specific questions to be addressed in this research are as follows: 1) Are there differences in the technical properties (i.e., thickness, rim diameter, volume, etc.) between Middle Woodland, Late Woodland, and Protohistoric period pottery? 2) Are there diachronic changes in ceramic vessel use and cooking habits evident through use- alteration traces? 3) Are there diachronic changes in subsistence strategies (and possible attendant changes in cooking habits) detectable through lipids, stable isotopes, and microbotanical remains extracted from pottery? 4) Is there synchronic variation in ceramic vessel use, subsistence strategies, and cooking habits evident through use-alteration studies or detectable through lipids, stable isotopes, and microbotanical remains extracted from pottery of differing typological categories? 5) How do ethnographic and ethnohistoric accounts of indigenous diet and cooking in the Great Lakes inform interpretations of ancient cuisine generated from the archaeological data? Data generated to answer these questions will contribute valuable new insights into the ongoing discussion of resource intensification, technological adaptation, and social transformation in the precontact northern Great Lakes. Research objectives will be addressed using a series of diverse but complementary methods, including functional ceramic analysis (of 16 technical properties and use-alteration traces), typological ceramic analysis, identification of starches and phytoliths from burned food residue, lipid residue analysis, and stable isotope analysis. Interpretations of the outcomes will be enhanced by ethnographic comparisons. The combined data will demonstrate the efficacy of a multidimensional approach to studying subsistence and technology that can be applied to archaeological assemblages from a wide variety of contexts, while also resulting in a fuller yet more nuanced picture of the dynamism of northern Great Lakes Woodland foodways. Dissertation Organization This dissertation is organized into nine chapters. Following this introduction to the research problem, questions, and theoretical orientation, Chapter 2 provides an overview of the archaeology and cultural history of the northern Great Lakes in the context of trends seen in Eastern North America. Chapter 3 is a discussion of the research problem and provides a background of the Cloudman site, which demonstrates its suitability for investigating the research questions. Chapter 4 will review the methods employed for this research project, including taxonomic pottery classification and functional pottery analysis, as well as lipid residue, stable isotope, and microbotanical analysis of absorbed and adhered food residues obtained from pottery. Chapter 5 will reconstruct the occupational history of the Cloudman site using taxonomic classification of the pottery assemblage and AMS radiocarbon dates. Chapter 6 reviews the functional analysis of pottery from the Cloudman site, while Chapter 7 presents the results of the various subsistence-related analyses. Chapter 8 places the research outcomes in context with ethnographic and ethnohistoric accounts of local indigenous culinary behaviors. The ninth and final chapter will provide a detailed discussion of the results, their implications for 17 understanding northern Great Lakes subsistence and technology, and the effectiveness of this multi-tiered approach to studying past behavior. 18 CHAPTER 2 ENVIRONMENTAL AND CULTURAL HISTORY/BACKGROUND OF THE NORTHERN GREAT LAKES Introduction The northern Great Lakes region, although often overlooked in the archaeological literature, has a rich, millennia-long cultural history characterized by complex and dynamic human interactions with both their social and natural environments. Situated on the perimeter of the Midwestern United States, the archaeological history of the northern Great Lakes is integrated into the overall cultural trajectory of the region. However, a number of phenomena distinguish the northern Great Lakes from the rest of the Midwest, from environment to sociocultural trends. It is within this unique microcosm that questions concerning foodways and technological change will be considered. This chapter will detail the broader ecological and social contexts of the research problem. Terminological clarification of geographic terms is required to properly convey the scope of the trends discussed below. The Great Lakes region constitutes land surrounding Lakes Superior, Michigan, Huron, Erie, and Ontario in northeastern North America, including portions of modern-day United States and Canada. The upper Great Lakes refers to the subset of this region lying north and west of Detroit, encompassing lands surrounding Lakes Huron and Superior, and the much of Lake Michigan. The focus of this study, the northern Great Lakes, encompasses Lake Superior and the northern portions Lake Huron and Lake Michigan and includes the Upper Peninsula of Michigan, northern Lower Michigan, parts of northern Wisconsin and Minnesota, and parts of southern coastal Ontario. 19 Consistent communication of dates is also required for clarity given the diachronic nature of the study. Henceforth, dates related to paleoclimatic trends will be reported as “BP” (before “present”, aka 1950). Sociocultural dates and time periods will be reported as calendrical years “BC” and “AD”, while direct radiocarbon dates will be reported in calibrated form (“cal BC” or cal AD) correlating with calendrical years, rather than in radiocarbon years. Paleoclimate and Environmental Background The onset of human occupation in northern Lower Michigan and the Upper Peninsula of Michigan occurred later than in other parts of the Midwest because of glacial ice coverage during the retreat of the Wisconsin ice sheet (Cleland et al. 1988, Larson and Schaetzl 2001). Michigan was deglaciated between approximately 20,000 and 9,000 BP. This deglacial phase was time transgressive, with the southern parts of Lower Michigan uncovered initially and northern Upper Michigan the final area to become ice free (Blewett et al., 2009). The first evidence of human occupation in southern Lower Michigan appears between 12,000 and 10,000 BP, after the forest cover became more extensive thus creating an environment more amenable to human habitation (Kapp 1999; Lovis 2009; Shott and Wright 1999:61). During this time, known as the Paleoindian period, the northern Michigan environment was composed of tundra and boreal forest (Hupy and Yansa 2009) and was occupied by large herbivores such as caribou and the last of the megafauna (Holman and Brandt 2009). Although glacial retreat from the Upper Peninsula had occurred by the beginning of the Younger Dryas, the Marquette Re-advance again covered the area with ice ca. 11,500 BP, and retreated, for the final time, ca. 9900 BP (Larson and Kincare 2009; Pregitzer et al. 2000). By this time, all of Michigan except the south shore of Lake Superior in the Upper Peninsula was 20 ice-free (Kapp 1999). However, a long period of sparse vegetation development in northern Lower Michigan and Upper Michigan following deglaciation precluded occupation of these areas prior to 10,500 BP (Cleland et al. 1988). What followed was a period of overall warming but frequent fluctuations in rainfall and lake levels (Kapp 1999; Larsen 1999), affecting vegetational and faunal composition (such as the extinction of megafauna) as well as human habitation patterns (e.g., Lovis et al. 2005). At this time, the eastern Upper Peninsula was covered by pine-dominated forests (Hupy and Yansa 2009). By 8000 BP, forest zones across the Eastern Woodlands were relatively similar to modern composition, with hemlock and birch trees as major components of the forest (Crawford 2011; Hupy and Yansa 2009). For a time, the climate shifted to cooler and moister conditions, causing a decrease in pine and an increase in hemlock ca. 6800 BP (Hupy and Yansa 2009) and the rise in lake levels, resulting in the Nipissing maximum phase of Great Lakes level by ca. 5000 BP (Kincare and Larson 2009; Robertson et al. 1999). However, between ca. 4000 and 3000 BP, lake levels dropped to levels known as the Algoma Stage. This was due to a very warm, dry period known as the Mid-Holocene Climatic Optimum, which again altered the vegetation composition across the landscape (Hupy and Yansa 2009; Kapp 1999; Robertson et al. 1999). Major ecological and biotic changes took place across the state beginning ca. 3000 BP, when the climate become moister and cooler, resulting in widespread marsh formation in the Upper Peninsula of Michigan (Kapp 1999:57). Between 3000 and 2000 BP, the forest in the eastern half of the peninsula stabilized to the hemlock-dominated forest (composed of birch, cedar, beech, and maple) that covers the region today (Hupy and Yansa 2009). Terrestrial and aquatic faunal species in the Midwest likewise became more similar to those present in the 21 region under modern conditions around this time (Shott 1999:72). However, fluctuations in lake levels and forest compositions influenced the abundance of and accessibility to certain species throughout time, and the human advent of new technologies, such as the bow and arrow and fishing nets, also affected the species exploited (Styles 2011). The Medieval Climatic Optimum (ca. AD 900 – AD 1100), also known as the Medieval Warm Period, has been connected to widespread demographic and cultural change across North America (Foster 2012; Lovis et al. 2012), and corresponds to technological and social shifts seen across Michigan (Brashler et al. 2000). Changes in climate and environment over such deep time segments are important for understanding past behavior, as many of the major and minor ecological changes referenced above can be correlated to settlement-subsistence patterns evident in the archaeological record. However, environment and climate are not determinate factors of sociocultural change or decision-making, but rather provide parameters that constrain variation. Although groups living in the northern Great Lakes were most certainly reactive to their environment, interactions with other groups and the development of new ideas and technologies played a part in the overall trajectory of northern Great Lakes history. Cultural Background: Overview of Upper and Northern Great Lakes Archaeology Pre-Woodland Periods The first humans moved into modern-day Michigan between 10,000 to 8,000 BC (Shott & Wright 1999), commencing the Paleoindian Period. Sites dating to this period are sparse throughout Michigan. Paleoindians were highly mobile hunter-gatherers reliant on megafauna, large herbivores such as caribou, fish, and plants (Holman and Brandt 2009; Lovis 2009). 22 The subsequent Archaic Period (8000 BC – 1000 BC)—typically divided into Early, Middle, and Late subperiods (McElrath et al. 2009)—is an important time interval in the trajectory of foodways in Eastern North America. Four indigenous seed-bearing plants (squash, sunflower, marshelder, and chenopod), known as the Eastern Agricultural Complex (EAC), were brought under domestication in the midlatitudes of Eastern North America during the Archaic period, between 3000 and 1800 BC (Smith and Yarnell 2009). Evidence for the contemporaneous cultivation of several cultigens and cultivars were found at the Riverton Site (1800 cal BC) in southern Illinois, which was occupied by a partially mobile, small-scale society (Smith and Yarnell 2009). Thus, the development of earliest crop complex in the Eastern Woodlands was likely not the result of population pressure but instead a long-term response to a resource-rich environment (Smith and Yarnell 2009). However, there is little, if any, evidence of the EAC or any other domesticates in the northern Great Lakes region during the Archaic Period, despite their presence as far north as the Saginaw drainage basin. Woodland Period The Woodland period, since its inception in McKern’s (1939) Midwest Taxonomic Method of classification, is generally recognized as a “stage marker” representing a suite of cultural developments between ca. 1000 BC and AD 1000/1600 in the Eastern Woodlands of North America (Anderson and Mainfort 2002). The Woodland period is most regularly associated with the advent of ceramic technology, and is typically divided into three subperiods: Early, Middle, and Late. The beginning of the Early Woodland period is generally signaled by the appearance of pottery; by ca. 500 BC, when sites containing technologically rudimentary and thick-walled “Marion” or “Schultz Thick” vessels appear in parts of the Midwest (Garland and 23 Beld 1999). The subsequent Middle Woodland (200 BC – AD 500/600) period is dominated in the much of the Midwest by the appearance of Hopewell, a “diverse sets of specific Middle Woodland societies, each internally bound through several diverse spheres of alliance” (Abrams 2009:172). The northern Great Lakes of North America constitutes a unique region within the Eastern Woodlands of North America relative to the normative Woodland construct applied further to the south. Some groups occupying this northern area adopted pottery nearly a thousand years after other Midwestern societies, and therefore in the absence of an “Early Woodland,” (Mason 1970) the Middle Woodland (200 BC – AD 500/600) and Late Woodland (AD 500/600 – 1600) remain the primary temporal divisions for this area (Brose and Hambacher 1999). While groups to the south were transitioning to characteristic Woodland lifeways, such as the manufacture and use of pottery and construction of burial mounds, those in the north persisted in Archaic lifeways until the Middle Woodland period despite undoubted centuries of contact with pottery-making neighbors to the south. The Hopewell phenomenon that prevailed over much of the Midwest during the Middle Woodland had little effect on the groups living in the northern Great Lakes besides their participation in long-distance trade or down the line exchange of goods (Brose and Hambacher 1999; Martin 1999b). Southern Middle Woodland ceramics have been found as far north in Michigan as the northwestern lower peninsula (Holman 1978; Lovis 1971; Lovis et al. 1998). The Middle Woodland period in the northern Great Lakes (where it is sometimes referred to as the Initial Woodland, particularly among Canadian researchers) witnessed the first adoption of pottery in the region. Prior assessments dated the outset of the Middle Woodland ca. AD 1, but recent dates of early pottery in the northern Great Lakes suggests the adoption of pottery as 24 early as 200 BC (see Albert et al. 2017). The lack of information concerning the lifeways of preceding Archaic populations that occupied the Upper Peninsula of Michigan impedes understanding of the nature of the transition to Middle Woodland settlement and social patterns. There is a lack of clear geographic attribute clustering of pottery types throughout the Middle Woodland period, which reflects a fluidity of regional populations rather than stable and formally geographically bounded social relationships (Brose and Hambacher 1999). This fluidity was necessitated by the mobility of the populations, who still relied on a broad spectrum of resources spread over large geographic ranges. Middle Woodland groups seemed to have increased aggregation at coastal villages for procurement of spring-spawning fish species; they also came together along lakeshores for limited fall fishing or along inland streams and lakes for wild-rice gathering (Brose and Hambacher 1999; Cleland 1982; Smith 2004). Middle Woodland social fluidity is demonstrated through the large geographic distribution of Laurel ware, which is found in southern Manitoba, northern Minnesota and Wisconsin, the Upper Peninsula of Michigan, and Ontario (Janzen 1968; Stoltman 1973; Wright 1967). Ceramic types found in adjacent regions include North Bay pottery, found in the Door Peninsula of Wisconsin and the western Upper Peninsula of Michigan (Mason 1966, 1967); Goodwinian Middle Woodland wares, local to the northeastern lower peninsula of Michigan (Fitting et al. 1969); and Point Peninsula wares of southeastern Ontario (Mason 1981; Ritchie 1969). All traditions utilized coil-constructed pottery with sub-conoidal bases, and stylistic variation occurs in a continuum across the region with little regional restriction, suggesting high levels of group movement and interaction (Brose and Hambacher 1999). The Laurel culture is typically defined by a mobile hunting-gathering-fishing lifeway, a lithic industry dominated by end-scrapers and stemmed or notched projectile points, bone-antler 25 tools and harpoons (Stoltman 1973). Laurel pottery is grit tempered with smoothed, rarely cordmarked exterior surfaces, and upper rim surfaces often decorated with a variety of dentate stamps (Janzen 1968; Stoltman 1973). Most sites are located on islands or sandy beaches along lakeshores or river mouths, places accessible by canoes or dugouts, and most residential sites suggest occupancy by a single family or small groups of relatives (Brose and Hambacher 1999:191). Major Middle Woodland sites in the northern Great Lakes include Naomikong Point (Janzen 1968), Winter (Bianchi 1974; Richner 1973), Gyftakis (Fournier 2007), Summer Island (Brose 1970), Cloudman (Branstner 1995), Portage (Lovis, Rajnovich, and Bartley 1998) and Arrowhead Drive (McPherron 1967), as well as Pic River (Wright 1966), Heron Bay, Michipicoten Harbor, and Sand River sites (Wright 1967) in Ontario. These sites are or were once located along the shoreline, and several are the earliest components of multi-component sites, speaking to the increased interest in occupying prime fishing locales beginning in the Woodland period. The adoption of pottery at the outset of the Middle Woodland in the northern Great Lakes is likely indicative of some social or economic change, the nature of which is still under debate. Skibo et al. (2009) posited the adoption of pottery was prompted by the need for improved efficiency of acorn processing, as demonstrated by lipid residue analysis of both fire-cracked rock and early pottery on Grand Island. However, lipid residue analysis of Middle Woodland pottery from the nearby Naomikong Point site did not contain nut lipids (Kooiman 2012, 2016; Malainey and Figol 2015). Interior carbonization patterns found in pottery from across the Upper Peninsula suggest that vessels were routinely filled to the top during cooking (Kooiman 2015, 2016), a habit possibly carried over from a long tradition of cooking in organic vessels, which 26 can be placed over fire without burning if filled with liquid (Speth 2015). The use of organic vessels fulfilled the needs of upper Great Lakes cooks for thousands of years, and reasons for transitioning to clay cooking pots may have varied across the region (Skibo et al. 2016). The Late Woodland period (AD 500/600 – AD 900/1600) may be described as a time of significant economic and cultural change that varied considerably across Eastern North America; changes possibly spurred by increased populations from the Middle Woodland to the Late Woodland (Anderson and Mainfort 2002; Holman and Brashler 1999). The shared artifact styles seen among many areas during the Middle Woodland gave way to increasing heterogeneity of ceramic styles throughout the period, suggesting greater social distinction and territoriality as resources were in more demand (Braun and Plog 1982; Holman and Brashler 1999; McElrath et al. 2000). The Late Woodland was a time of technological and subsistence change in the Eastern Woodlands. The bow and arrow made its first appearance during this time, while cultivated crops became increasingly important (McElrath et al. 2000). The adoption and dietary incorporation of domesticated plant species was gradual throughout the Woodlands, and most groups remained semi-mobile until the intensification of maize cultivation after AD 800/900. Throughout much of the Eastern Woodlands, Late Woodland period traditions were followed by a variety of new cultural manifestations, such as Mississippian, Upper Mississippian/Oneota, Fort Ancient, and Iroquoian, all marked by maize agriculture and increased sedentism (Schroeder 2004). In contrast, Late Woodland-period sociocultural practices and lifeways generally persisted until European contact in much of the upper and northern Great Lakes, although they did not remain static (Schroeder 2004). 27 Major Late Woodland sites in the northern Great Lakes include Sand Point (Dorothy 1978, 1980), Scott Point, Juntunen (McPherron 1967), Wycamp Creek (Lovis, Arbogast and Monaghan 2012), O’Neill (Lovis 1973), and Cloudman (Branstner 1992, 1995), mostly large coastal sites interpreted as seasonal aggregation fishing sites. However, according to Dunham (2014), smaller interior sites were just as important for the seasonal subsistence rounds, but are less visible due to their size and location in less-developed areas (e.g. the Inland Waterway sites [Lovis 1978]). While some argue that populations grew over the course of the Late Woodland period, leading to new food procurement and mobility strategies (Cleland 1982), others believe populations and settlement/subsistence strategies remained steady throughout (Martin 1989, 1999a). Unlike the preceding Laurel Middle Woodland, Late Woodland pottery was formed by slab construction instead of coiling and frequently cordmarked. Ceramic wares from across the northern Great Lakes continued to share many stylistic attributes in the early Late Woodland period (AD 600-1000; Holman and Brashler 1999). Bowerman and Skegemog wares common to the Traverse corridor in the northeastern lower peninsula of Michigan bare resemblance to wares further to the south (Hambacher 1992; Holman and Brashler 1999). Juntunen sequence vessels, most closely associated with the Straits of Mackinac, predominate in the eastern Upper Peninsula throughout the Late Woodland period (Martin 1999a). The sequence includes Mackinac, Bois Blanc, and Juntunen wares (McPherron 1967). The earliest, Mackinac ware, is common between AD 750 – AD 1000 (Lovis 2014; McPherron 1967), and while distinct from most lower Michigan wares, it bears some resemblance to Skegemog pottery (Holman and Brashler 1999). Blackduck ware, the stylistic descendant of Laurel ware, is found across and beyond the northern Great Lakes, from Saskatchewan to Quebec, and from northern Minnesota to the Upper 28 Peninsula of Michigan across variable durations of time (Hamilton et al. 2007; Lugenbeal 1978; McHale Milner 1998; McPherron 1967). Blackduck closely resembles Bois Blanc ware, the second of the Juntunen sequence wares, which was manufactured in the Straits region between AD 1000-1200 (Lovis 2014; McPherron 1967). Pottery style generally became more heterogeneous during the late Late Woodland (Holman and Brashler 1999), reflecting the social localization proposed by McHale Milner (1991, 1998). Traverse wares replace Bowerman and Skegemog pottery in the Traverse corridor, while Juntunen wares, the last of the Juntunen sequence, predominate in the Straits and eastern Upper Peninsula after AD 1200 (McPherron 1967; McHale Milner 1998). Juntunen wares display stylistic attributes that suggest interactions with Iroquoian groups of Ontario (Brashler et al. 2000; Holman and Brashler 1999; McHale Milner 1998), using similar decorative motifs but employing drag-and-jab decoration techniques rather than incising. Upper Mississippian wares also make an appearance at certain sites across the Upper Peninsula (Dorothy 1978, 1980; Holman and Brashler 1999; McPherron 1967). The increase in distinctive local pottery wares during the late Late Woodland also more frequently co-occur at coastal aggregation sites, suggesting both an increased sense of group identity and the importance of extralocal relationships during this time (Holman and Brashler 1999; McHale Milner 1991, 1998; O’Shea and McHale Milner 2002). Iroquoian Iroquoian-speaking people occupied portions of modern-day New York and southeastern Ontario. Iroquoian groups settled in modern-day Ontario during the late precontact and protohistoric periods include the Huron, Tionnontati (Petun), Neutral, and St. Lawrence 29 Iroquoians (Warrick 2000). Although the emergence of sedentary agriculturalists with pottery and a language system quite distinct from surrounding Algonquian groups caused many to believe that the Iroquoians migrated into the region, most archaeologists now believe that these groups developed in situ from Woodland populations (Smith 1990; Warrick 2000). Late Woodland Princess Point groups in southern Ontario (AD 600-1000) have been interpreted as being the direct ancestors of later Iroquoian groups based on the continuity of material culture and settlement-subsistence patterns. They were among the first to experiment with maize agriculture in northeastern North America (Fox 1990a). Princess Point people lived in small year-round villages with warm season camps for hunting and gathering, since wild resources still made up most of their diet (Snow 1980a; Warrick 2000). Immediate descendants of Princess Point are referred to as “Early Iroquois” (AD 1000- 1300 [Warrick 2000; Williamson 1990]). Wright (1973) divided the early Iroquoian stage of Ontario into two distinct regional variants: the Glen Meyer branch in southwestern Ontario, and the Pickering Branch of southeastern Ontario. While the two branches are no longer considered discrete sociocultural entities, they are still recognized as distinct regional populations (Williamson 1990). Glen Meyer and Pickering peoples lived in unstable, semi-sedentary villages comprised of small longhouses and began to rely more heavily on maize (Bamann et al. 1992; Kuhn and Funk 2000). The subsequent Middle Ontario Iroquoian stage (AD 1300-1400) is marked by the growing intensification of horticulture subsistence (particularly maize cultivation). This was accompanied by the establishment of stable, year-round residences, which gradually became populated with longhouses of increasing size (Bamann et al. 1992; Kuhn and Funk 2000). Households were located in small to medium-sized villages that were widely 30 dispersed across the landscape and were not yet incorporated into region-wide identities, as signaled by ceramic motifs (Hart et al. 2017). The emergence of Huron populations (distinct from other Ontario Iroquoian groups, such as the Neutral and St. Lawrence Iroquois) arose ca. AD 1400, in south-central Ontario, between the Trent River, the Niagara Escarpment and Lake Ontario, an area historically known as “Huronia” (Ramsden 1990). The Huron developed as population increases solidified regional identities. Villages were autonomous and highly organized, with each matrilineage contained within a longhouse; both village and house sizes were larger than in the Middle Iroquoian period (Warrick 2000). Along with villages, farming hamlets and logistical resource extraction camps associated with the Huron have also been found (Ramsden 1990). The population migrations, social restructuring, and coalescence of smaller villages into larger ones that occurred after A.D. 1500 have been attributed to the indirect effects of European contact, although this has been disputed in recent years (Ramsden 1990). While there is some stylistic overlap between “Huron- Petun” and “Neutral” pottery, these groups are also associated with specific ceramic typologies emblematic of distinct social identities (Wright 1973). Proto-Iroquoian and Iroquoian populations witnessed the gradual intensification of cultivated foods. Although maize first appeared in southern Quebec by 391-209 cal BC (St. Pierre and Thompson 2015), northern New York by cal AD 1 (Hart et al. 2007), and southern Ontario between cal AD 260 and AD 660, Princess Point groups only began to experiment with maize horticulture (Fox 1990a). Maize was not a significant part of the diet until the Early Iroquoian period (AD 1000-1300), when it still only constituted between 20-30% of the diet among Ontario occupants, who continued with predominantly hunter-gatherer subsistence strategies (Warrick 2000). Evidence for beans, squash, and sunflower does not surface until the 31 middle of the 11th century; the “Three Sisters” subsistence regime (maize, beans, and squash) became dominant between AD 1300-1525, and sunflower and tobacco were also grown (Bamann et al. 1992; Kuhn and Funk 2000). Although the Huron were sedentary agriculturalists by this time, their diets were supplemented by wild game, fishing, and some wild plant resources (Williamson 1990). They may have also traded maize in exchange for fish and wild game, often via the Odawa, with northern hunting-gathering Algonquian groups as a form of risk-buffering against periodic crop failure (O’Shea 1989; Smith 1996). The Protohistoric/Contact Period Europeans arriving in and traveling across North America in the 16th century encountered resident Native populations and began to trade with them (Figure 2.1). The Anishinaabeg people, which includes the Ojibwe, Potawatomi, and Odawa, are believed by scholars to have originated from east of the Great Lakes, migrating west in the late 16th or early 17th century (Danziger 1978:7). This is corroborated by Anishinaabe oral tradition, which recounts their origins in the North Atlantic coast and subsequent 500-year westward migration along the St. Lawrence River, culminating in their arrival at the western end of Lake Superior by the 17th century, settling among resident Algonquian groups (Benz and Williamson 2005). When the French arrived in Ontario in 1615, they encountered Algonquian-speaking bands of mobile hunter-gathering living adjacent to Huron territory (Fox 1990b). These people were identified as the Odawa. At the time, these groups occupied the Bruce Peninsula, Manitoulin Island, and other parts of the coast of Georgian Bay (Fox 1990b). Early historical accounts and archaeological excavation of protohistoric Odawa sites indicate that they were mobile hunter-gatherers who heavily fished and practiced marginal horticulture (Fox 1990b). 32 Figure 2.1: Distribution of Cultural Groups ca. 1630 (adapted from Trigger 1976:92) Fox and Garrad (2004) argue that substantial exchange between precontact Huron-Petun groups in southern Ontario and Algonquian groups occupying the south shore of Georgian Bay (assumed to be the Odawa), commenced in the 14th century AD, and that Iroquoian- manufactured and imitation Iroquoian pottery located along the Georgian Bay region is evidence of this exchange. Archaeological sites from this time have been identified as Odawa based on Algonquian lithics and ritual signatures (e.g., dog burials), despite the Huron pottery and pipes (or locally produced copies), which are argued to be the result of trade with the Huron (Fox 1990b; Ross 1975). Proponents of this model claim hunter-gatherer Odawa groups exchanged fish for maize with the Huron (Fox 1990b, 2004). Alternatively, Warrick (2000:442) interprets 33 these findings as evidence of early Iroquoian occupation of south Georgian Bay, mapping Middleport (AD 1330-1420) Iroquoian sites in the area. Interactions between the Huron-Petun and the Odawa may have been hindered by linguistic differences. Northeastern North America is not considered a “strong linguistic area,” or a region in which small linguistic communities have long been in constant contact and individuals are often multilingual (Mithun 1999:314). Groups living in the western Upper Great Lakes region spoke Algonquian languages in the Algic language family, while the Huron- Wyandot language belongs to the Iroquoian family. These languages would not be mutually intelligible (Mithun 1999), making communications between the Odawa and Huron difficult. Smith (1996), however, found evidence of an exchange network centered on the exchange of food items, where northern Algonquian groups, primarily the Ojibwe, traded meat to middlemen Odawa for their own consumption and for trade with Huron and Petun groups in southern Ontario, who in turn provided maize for distribution to the north. The Odawa themselves provided fish, reed mats, and other products to the system but primarily served the role of transporters of food items (Smith 1996:278). Groups known historically as the Ojibwe (aka Ojibay, Ojibwa, Chippewa) settled in the Upper Great Lakes primarily around all shores of Lake Superior, although some migrated and settled as far as the Western Plains of Minnesota and the Dakotas (Danziger 1978:8). Odawa is a dialect of the Ojibwe language, the mutual intelligibility of which would have facilitated trade among the different Anishinaabeg groups (Mithun 1999). Most Ojibwe lived a mobile, hunter- gatherer lifestyle, following seasonal economic cycles based around subsistence activities such as fishing, hunting, and gathering, much as resident Woodland populations (Danziger 1978:9, 34 11-13). Anishinaabeg people in the Great Lakes retained a flexible social structure that frustrated the European traders who attempted to trade and make treaties with them (Witgen 2012). As the Europeans moved further to the continental interior, various Iroquoian-speaking groups in the east banded together into what became known as the Iroquois Confederacy or League (Hunt 1940; Morgan 1922), which sought to protect its territory and expand it into new areas (Parmenter 2010). The Iroquois Five Nations/Confederacy, occupying modern-day New York, instigated the dispersal of the Huron from Huronia after 1649 through a series of attacks on their neighbors (Ramsden 1990). Therefore, post-1650 northern Great Lakes demographics are even more complicated due to migration and new identity formation following increased territorial pressures. Refugee Huron groups either fled east, integrated into the Five Nations Iroquois, or joined with the Tionantati (Petun) (Mithun 1999:421; White 1991). Remnants of the Huron, Tionantati, Erie, and Neutral, all defeated by the Five Nations, were collectively known at the Wyandot and together fled westward to settle near Detroit (Mithun 1999:421). Some refugee Huron groups fled further west to the lower peninsula of Michigan and Wisconsin, settling among local Anishinaabeg populations (White 1991) and creating mixed communities that are visible archaeologically (Mason 1986). Eventually, the long-distance trade and communication networks that developed facilitated the irregular spread of Western diseases, depleting Indigenous populations and interrupting the balance of power throughout Eastern North America (Milner et al. 2001). Conclusion The sociocultural history of the Upper Great Lakes is complex, despite the superficial appearance of simplicity when compared to the cultural histories of neighboring regions. 35 Precontact indigenous societies in this region had to contend with harsh climatic and environmental fluctuations and conditions. While these challenges sometimes delayed or precluded the types of cultural developments that occurred elsewhere in the southern Great Lakes, riverine Midwest and southeastern Canada, it also led to local and regionally successful adaptations and behaviors that resulted in the long-term endurance of groups occupying the region and a cultural dynamism reflected in the material culture. The paucity of sites and artifacts resulting from low populations density and relatively poor preservation of archaeological materials contributes to the initial appearance of the Upper Great Lakes as a less interesting region of scholarly inquiry. It also makes interpretations of past behavior in the northern Great Lakes difficult. However, the collective results of the many decades of research in this region, when examined closely, reveal the complex nature of northern Great Lakes societies through time. Yet, many questions remain. The application of new analytic techniques has the potential to provide new data and inform different or standing unanswered questions, highlighting both the intriguing history of the northern Great Lakes and the scope of human flexibility and creativity in the face of social and environmental change. 36 SUBSISTENCE AND TECHNOLOGY IN THE NORTHERN GREAT LAKES CHAPTER 3 Introduction The spread of domesticates from the midsouth and Mesoamerica initially had little impact on the occupants of the Midwestwestern United States and southeastern Canada, a testimonial to the long-term stability of hunting-gathering-fishing strategies. Domesticates did eventually become the focus of subsistence for various groups, such as the Oneota, Mississippians, Fort Ancient groups, and Iroquoian societies after AD 900. Societies in the northern Great Lakes, too, initially incorporated domesticates into their diets on a limited basis. Around the time their neighbors to the south became increasingly involved in agriculture, these groups may have also sought out more productive subsistence strategies. Whether such a change occurred and, if so, which strategies were employed to increase resource productivity are the core of one of the most lingering and long-debated problems in northern Great Lakes archaeology. Long-term subsistence change can only be investigated through diachronic comparisons of archaeological data. Prior inquiries have used various types of evidence to address the problem, but inconsistent spatial and temporal data, scales of inquiry, and limited preservation of organic materials in the northern Great Lakes have precluded a firm resolution. To avoid these issues, this study looks at change through time at a single location, the Cloudman site (20CH6), using a new suite of techniques never before employed together in the northern Great Lakes. Following the context of the history of the research problem and the background of the Cloudman site, expected outcomes to the research questions are outlined at the end of this chapter. 37 Research Problem The settlement-subsistence patterns of the Woodland occupants of the northern Great Lakes has long been a topic of archaeological inquiry, and therefore much debated. Various models for Upper Great Lakes (including the northern Great Lakes) Woodland settlement- subsistence patterns have been proposed. The most prominent model for Middle Woodland subsistence centers around seasonal aggregation at coastal sites, where groups took advantage of spring-spawning fish in shallow waters using, according to Cleland (1982), the seine net. Summer Island was interpreted by Brose (1970) as a spring coastal site with evidence of a return to a generalized mixed economy following the spawning runs. These groups likely moved to different coastal sites, or to interior lakes or streams, during the remaining warm season to exploit various other seasonal resources before retreating to small interior hunting encampments for the summer (Brose 1970; Brose and Hambacher 1999). Cleland (1982) proposed that the invention of the gill net in the Late Woodland period allowed inhabitants to exploit deep-water spawning fish, leading to a series of widespread changes marking the Middle Woodland-Late Woodland transition, including increased populations, larger and denser shoreline sites, and subsistence strategies that relied on exploitation of seasonally abundant fish and plant materials. Seasonal aggregation at coastal sites therefore also occurred in the fall, when deep-water spawning fish (such as whitefish and lake trout) could be harvested en masse. Martin (1989) disputed Cleland’s “inland shore fishery” model of change over time, citing a lack of evidence for increasing population size, instead attributing perceived site size increase to the palimpsest effect of repeated occupations. Questioning Cleland’s evidence for the invention of the gill net at the outset of the Late Woodland period, Martin presents data 38 suggesting some deep-water fishing occurred during the Middle Woodland period. She also disputes alteration of settlement patterns resulting from new seasonal fishing habits in the Late Woodland, arguing instead for subsistence and residential continuity throughout the Woodland period. In rebuttal to Martin, Cleland (1989) reiterates evidence supporting the intensified exploitation of fall-spawning fish during the Late Woodland and claims Martin failed to demonstrate the use of gill nets, which are necessary for deep-water fishing, during the Middle Woodland period. In a reexamination of this issue and debate, Smith (2004) uses multiple lines of evidence to demonstrate that fall-spawning fish were exploited in the Middle Woodland, but the intensive use of gill nets to procure them in large quantities did not become common until the Late Woodland period, particularly after AD 1100 in the Lake Michigan basin and after AD 1400 throughout the rest of the Upper Great Lakes. Drake and Dunham (2004) likewise advocate for greater continuity in the use of coastal sites between the Middle and Late Woodland periods than was suggested by Cleland, but acknowledge broader shifts in social, subsistence, and settlement patterns moving forward into the Late Woodland period. Most recently, Dunham (2014) highlights alterations in settlement, subsistence, and social structure over the course of the Late Woodland period in the Eastern Upper Peninsula of Michigan, the approximate center of the northern Great Lakes region. While early Late Woodland people were more residentially mobile and produced more homogenous ceramic styles (suggesting social homogeneity similar to that seen during the Middle Woodland), people in the late Late Woodland period (post-AD 1000) were more logistically mobile, occupied sites with greater resource diversity, exploited greater amounts of starchy resources with interior (i.e., wild rice) or patchy (i.e., acorns) geographic distribution, increased storage, and displayed 39 greater social heterogeneity. Dunham hypothesizes these changes were in response to more dynamic lake levels resulting from the Medieval Climatic Optimum post-AD 900 (Lovis et al. 2012). Environmental and resource instability prompted the aforementioned reactions—all economic or social mechanisms of risk buffering. Reliance primarily on the exploitation of spring and fall spawning fish, as posited by Cleland, was too risky, and therefore both wild rice and acorns were exploited and stored as buffering resources against periodic fish shortages. Dunham did not find maize to have been included as a major component of even the late Late Woodland buffering mechanisms in this portion of the northern Great Lakes. An alternative view of Late Woodland subsistence regimes states that the ceramic heterogeneity seen in the late Late Woodland is indicative of the formation of horticulturalist coastal societies (such as groups using Juntunen wares) and inland groups, who remained foragers (O’Shea 2003). This is a classic ethnographic forager-farmer interaction model often applied to the Mesolithic of Europe (Zvelebil 1986). The Great Lakes model entails coastal societies who were maize agriculturalists occupying large or major aggregation sites used for rituals and exchange of maize and other coastal resources to inland foraging groups. This interpretation is based on the finding that maize constituted 18-20% of the Juntunen diet (Brandt 1996) and the discovery of earthwork enclosures with associated storage pits in northern Lower Michigan (Howey and O’Shea 2002). The role of domesticates and horticulture in the northern Great Lakes during the Middle and Late Woodland periods remains unclear. Whether certain Late Woodland occupants of the northern Great Lakes post-AD 1000 intensified their horticultural practices (per O’Shea 2003), or instead intensified exploitation of specific wild resources (per Cleland 1982; Dunham 2014; Holman and Brashler 1999) is still debated. Macrobotanical evidence of maize in the northern 40 Great Lakes is limited, confined to three Late Woodland sites located in the mildest climatic coastal areas (zones with at least 140 frost-free days required for sufficient maize growth; Yarnell 1964) (Dunham 2014:199; Egan-Bruhy 2007; Lynott 1974; McPherron 1967). There is, however, increasing evidence for the incorporation of maize into the diet, either grown locally or exchanged in from elsewhere, obtained from microbotanical food remains in carbonized residue on pottery. Some of this evidence supports the entry of maize into the region much earlier than the macrobotanical evidence suggests. Maize starches and phytoliths have been identified in Middle Woodland contexts in northern New York (Hart, Brumbach, and Lustek 2007), southern Quebec (St-Pierre and Thompson 2015), boreal Manitoba and Ontario (Boyd et al. 2014), northern Minnesota (Burchill 2014), and the Saginaw basin of Michigan (Raviele 2010). While residues containing maize microbotanicals were dated to as early as 390 cal BC in Minnesota (Burchill 2014), Albert et al. (2017) recently found the earliest evidence of maize in the northern Great Lakes in residues from pottery at the Winter site (20DE17), which was dated to as early as cal 200 BC (Lovis et al. 2012). Tracing the rate of maize adoption is another current avenue of inquiry, since it is demonstrated across the Americas that maize intensification often occurred centuries after its initial introduction into a region (Hastorf and Johannessen 1994). Only trace amounts of maize were found in carbonized food remains from Laurel pottery at the Winter site (20DE17), and maize was not ubiquitous among vessels examined, suggesting only casual use of the cultigen during the Middle Woodland period (Albert et al. 2017). Maize consumption at the Late Woodland Frazer-Tyra site (AD 900-1200) in the Saginaw Valley of lower Michigan varied based on age and sex (Muhammad 2010). This suggests that occupants of the site were trading for maize rather than growing it, and its acquisition and consumption was rooted in social 41 networks and obligations rather than subsistence needs, even almost one thousand years after the arrival of maize into Michigan (Muhammad 2010). Evidence of other domesticated cultigens in the Great Lakes region and surrounding areas is much more limited, particularly in the Middle Woodland period. Evidence of squash and beans in boreal Manitoba and Ontario was encountered only in Late Woodland contexts, later than evidence for maize in the same region (Boyd et al. 2014). Beans arrived in the upper Midwest/Great Lakes area by cal AD 1200 via the Plains, and spread into the northeastern/New England regions shortly thereafter (Hart et al. 2002; Monaghan et al. 2014). Squash, one of the cultigens domesticated in North America, was present in lower Michigan by 2300 cal BC (Monaghan et al. 2006), but does not seem to have been a significant food source in the Great Lakes until the Late Woodland period, although it may be underrepresented in the microfossil record due to problems with identification (Boyd et al. 2014). Other North American cultigens, such as marshelder, chenopod, and sunflower, which along with squash are known as the Eastern Agricultural Complex (Smith 2006), are rare in the northern Great Lakes throughout the precontact period. Wild rice was another important starchy resource for occupants of the northern Great Lakes and surrounding regions. While not a domesticate, wild rice had to be harvested from wild stands in early fall, much like a cultivated crop (Vennum 1988). Evidence for wild rice in archaeological contexts appears in the Middle Woodland period in the Great Lakes region (Arzigian 2000; Lovis et al. 2001) and in boreal Ontario (Boyd et al. 2014). Many scholars have noted an increase in wild rice exploitation during the Late Woodland period across the Great Lakes (Dunham 2014; Moffat and Arzigian 2000), although the drier climate characteristic of the Late Woodland and subsequent reduction of wetlands at its outset may have reduced wild rice 42 harvesting opportunities in some areas, as seen at the Schultz site in the Saginaw Valley (Lovis et al. 2001). Several have found wild rice in close association with maize following the latter’s arrival into the region (Boyd et al. 2014; Hart and Lovis 2013; Raviele 2010), possibly due to the nutritionally complementary nature of the two foods (Hart and Lovis 2013). The intensification of starchy food exploitation, whether wild or cultivated, in the Late Woodland period may be supported by ceramic use-alteration traces. Kooiman (2012, 2015, 2016) found that cooking habits varied between the Middle Woodland and Late Woodland periods in the northern Great Lakes. Patterning of interior carbonization on Middle Woodland ceramic vessels from the Naomikong Point (20CH2) and Winter (20DE17) sites indicated the prevalence of stewing, while Late Woodland vessels from the Sand Point site (20BG14) were more frequently marked with carbonization patterns indicative of boiling practices. The observed change in cooking habits has been hypothesized to signal broader dietary shifts, possibly connected to the intensification of starchy foods in the Late Woodland, which require long-term boiling or cooking to be made digestible (Braun 1983; Wandsnider 1997). However, lipid residue analysis of vessels from Naomikong Point and Sand Point revealed few differences in the foods cooked in Middle Woodland and Late Woodland vessels (Malainey and Figol 2015). At present, the role of starchy foods and the timing of the incorporation of maize, wild rice, and other cultigens in regional Woodland subsistence regimes is unknown. Subsistence, as detailed by foodways theory (Fischler 1988; Rodríguez-Alegría and Graff 2012; Twiss 2012), is inextricably related to the specifics of settlement systems and patterns of social organization and interaction. The Middle Woodland-period populations in the northern Great Lakes were highly mobile and maintained fluid social relations and group compositions, necessitated by their reliance on a broad spectrum of resources spread over large geographic 43 ranges (Brose and Hambacher 1999). People did, however, aggregate at coastal villages for procurement of spring-spawning fish species; they also came together along lakeshores for limited fall fishing or along inland streams and lakes for wild-rice gathering (Brose and Hambacher 1999; Cleland 1982; Smith 2004). The Late Woodland period appears to have been a time of increasing localization in the northern Great Lakes, in which subsistence efforts were concentrated within decreased geographic ranges on productive but localized resources (including possible maize production, fishery exploitation, and certain wild resources) (McHale Milner 1991:42-43). Using ethnohistorical information and stylistic ceramic analysis of Juntunen phase (AD 1200 – AD 1450) sites in the Upper Peninsula, McHale Milner (1991) argued that Late Woodland peoples developed increasingly bounded social groups at the local level while also maintaining long- distance, extralocal social ties for risk buffering purposes in times of local food shortages. The extralocal, long-distance relationships were confirmed by the unexpected presence of “foreign” stylistic elements on Juntunen pottery (McHale Milner 1998). According to McHale Milner these social relationships were formed, maintained, and solidified through integrative events at seasonal aggregations sites. The localization occurring in the Late Woodland may have, therefore, not only changed the nature of social relationships and subsistence behaviors but may have simultaneously forced the formation of long-distance relationships and trade networks as an alternate form and scale of spatially nested or hierarchical risk-buffering (Carroll 2013). Overall, the evidence suggests that northern Great Lakes Woodland groups increasingly employed a series of risk-buffering techniques, including shifting intensification of targeted resources (fish and starchy foods) and 44 integrative social events, over time in response to environmental instability, particularly during the late Late Woodland. Amelioration of risk by alterations in social, settlement, and subsistence regimes in the northern Great Lakes would have been further complicated by the entry of Europeans into Eastern North America. Movement of both Algonquian- and Iroquoian-speaking groups into the Great Lakes region required new social negotiations and relationships, which may have included further localization, the formation of bounded social identities, and a greater need for trade relationships between groups. The timing of group identity formation, group movements, and the level of trade between groups are unresolved issues requiring further archaeological inquiry. Ultimately, the ways in which technology, diet, cooking, and social relationships relate to one another and change from the Middle Woodland through the Protohistoric period is unclear. Pottery occupies a unique position at the crossroads of subsistence and society and offers the ideal pathway by which to explore these issues together. Case Study: Cloudman Site (20CH6) The Cloudman site is located on Drummond Island, MI, in the Eastern Upper Peninsula of Michigan near the southern outlet of the St. Mary’s River into Lake Huron (Figure 3.1). It lies on the northern bank of the Potagannissing River, occupying a series of river terraces. The site covers approximately 30,000 square meters and contains four occupational components that are both vertically stratified and horizontally separated (Branstner 1995). Branstner (1995), using relative dating of artifact styles, interpreted the timing of occupation for each component of the site: Middle Woodland (AD 0 – AD 400); Late Woodland (AD 800 – AD 1500); Protohistoric (AD 1500 – AD 1650); and two historic period homesteads (AD 1880 and 1920). 45 Figure 3.1: Location of the Cloudman (20CH6) Site and Other Woodland Sites The Cloudman site also includes a large mound (possibly associated with the Middle Woodland occupation), which was the focus of amateur excavations in the early twentieth century. Hinsdale (1931) was the first to officially report the site, but a thorough assessment of the site was not conducted until 1974, when John Franzen assessed the locale as part of a survey of Chippewa County. He initially reported Cloudman as containing Middle Woodland and Historic-period components (Franzen 1975). The St. Mary’s River Archaeological Project, under the auspices of Michigan State University, visited the site in 1990 and 1991 to identify its components and spatial extent. The site was excavated under the supervision of Charles Cleland and Christine Branstner of Michigan State University in 1992 and 1994, during which 102 46 square meters were excavated and thirty-three features uncovered (Branstner 1995). An additional unit was also excavated in 1995 by Cleland, although a report of the results from this work was not completed (Christine Stephenson, personal communication). Gradual lowering of lake levels over time led to the formation of three distinct river terraces at the site, which roughly correspond to each of the three primary occupations (Branstner 1995:23). Middle Woodland materials are largely confined to the upper terrace (181m and above); the middle (179 – 181m) and lower (177 – 179m) terraces contain mostly Late Woodland and Protohistoric materials, respectively (Branstner 1995:23). There was only minimal spatial overlap between the different temporal occupations (Branstner 1995). Branstner (1995) conducted preliminary analysis of the ceramic assemblage from all components of the Cloudman site, which was limited to typological categorization of pottery vessels. She sorted and identified vessels from the Middle Woodland, Middle Woodland/Late Woodland transitional, early Late Woodland, late Late Woodland, and Contact/Protohistoric periods. There were also a number of miniature vessels which were not described. Identified Middle Woodland vessels at Cloudman are almost exclusively varieties of Laurel ware, while those vessels identified by Branstner (1995) as early Late Woodland are primarily Mackinac ware varieties. In accordance with McHale Milner (1991, 1998), late Late Woodland vessels demonstrate great stylistic diversity. Branstner was unable to assign types to 31 (72%) of the late Late Woodland vessels, which she attributed to hyper-localized styles not yet identified in the existing literature. Identifiable types included Juntunen and Algoma wares. According to Cleland (1999), the Protohistoric component of the Cloudman site is one of the earliest sites in Michigan with evidence of European contact, with trade goods dating to approximately 1615-1630 (Branstner 1995). Branstner and Cleland interpret this component as 47 an Odawa occupation, based on Trigger’s (1976:294) mapping of the Bruce Peninsula, Manitoulin Island, and Drummond Island as “Ottawa” territory ca. 1615, which also shows trade routes between the Odawa territory and the Huron settled on the eastern shore of Lake Huron. Branstner categorized pottery in this component as non-local Ontario Iroquoian/Huron vessels as well as locally-made “Huron imitation pots” (Branstner 1995:12). More recently, 16 cobalt blue glass trade beads from the Cloudman site were sampled with LA-ICP-MS and identified as part of a pre-1700 Mg-low-P compositional group (Walder 2018), and two additional analyses of white glass beads identified the opacifier arsenic, which means that the manufacture of the bead likely post-dates c. 1670 (Hancock et al. 1997; Walder personal communication 2018). This calls into question the timing of occupation of the site during the Protohistoric period. Drummond Island lies within the Eastern Upper Michigan regional landscape ecosystem, which is forested by northern conifers, northern hardwoods, and patches of oak (Albert, Denton, and Barnes 1986). This would have supported a variety of wild plants and animals that could be used for human exploitation. Although the region is cool with rather sandy soil, it has a relatively long growing season due to climatological amelioration from Lakes Superior, Michigan, and Huron (Albert et al. 1986). This means that the growing season would have just been long enough for maize at the Cloudman site. Additionally, the site is located close to modern wild rice stands, although it is difficult to determine the antiquity of these stands (Dunham 2014). Some dietary and seasonality data is available for the Cloudman site based on floral and faunal remains, although it is limited. Flotation samples were taken from most features, and analysis of macrobotanical remains extracted from these samples revealed a great diversity of plant species (Egan-Bruhy 2007). Cloudman occupants ate a variety of fruits (including strawberry, raspberry, elderberry, and wild plum), nuts (such as hazelnuts, walnuts, and acorns), 48 and seeds/cultigens such as chenopod, maize, and wild rice (Egan-Bruhy 2007). These resources would have been available during the summer and fall seasons. Faunal remains from only two features were analyzed. Feature 26, attributed to the Late Woodland period, contained woodchuck and beaver, spring and early summer-spawning fish (northern pike, sucker, catfish, bass, perch, walleye), and fall-spawning fish (whitefish and drum) (Cooper 1996). Feature 27, associated with the Protohistoric period, contained a dog burial and most of a snapping turtle, but also contained portions of beaver, muskrat, black bear, caribou, common loon, spring/summer-spawning fish (sturgeon, pike, sucker, catfish, perch, and walleye) and fall-spawning fish (whitefush and drum). Sample sizes of various species in both features were too small to assign strict occupational seasonality, although the large proportion of sucker (39% of fish elements) from Feature 27 suggests targeted spring-spawning species exploitation. Additional and more refined data is required for more accurate interpretation of diet and site seasonality. Large, multicomponent sites available for study in the northern Great Lakes, and particularly in the Upper Peninsula, are rare. The Arrowhead Drive/Juntunen site, Summer Island, Scott Point, and Cloudman are the only sites with substantial ceramic assemblages attributed to Middle Woodland, Late Woodland, and/or assumed Protohistoric periods; the Cloudman assemblage is the only assemblage with pottery from all three periods. This makes Cloudman an excellent if not ideal site for the investigation of the problem of changing subsistence strategies and cooking techniques throughout the Woodland Period and the impact of culture contact on foodways. 49 Research Expectations Expectations for the results of the analyses were formulated prior to research. These expectations were based on background information and previous research and are outlined below: 1) Are there differences in technical properties (i.e., thickness, temper size, rim diameter) among Middle Woodland, early Late Woodland, late Late Woodland, and assumed Protohistoric period pottery from the Cloudman site? Braun (1983) first proposed that the increased consumption of starchy seeds from the Middle Woodland to Late Woodland in the Midwest prompted technological alterations in cooking vessels. The palatability and digestibility of starchy seeds is improved by cooking them to a point of gelatinization in liquid broth, although this process requires both longer cooking times and higher temperatures (Braun 1983:116). Late Woodland vessels from western Illinois and eastern Missouri have thinner walls, smaller temper particles, and a more globular shape, which improve thermal conductivity and thermal shock resistance. Braun concluded that technical alterations to Late Woodland vessels were directly related to the incorporation of more starchy foods in the diet. Dried maize also requires long periods of boiling to make it edible and digestible. Using bulk δ13C analysis, Hart (2012) also found a direct correlation between an increase in water-based maize cooking and wall thinning in Woodland vessels in central New York. Given Dunham’s hypothesized intensification of starchy food use in the Late Woodland of the eastern Upper Peninsula, it is expected that the trend of wall thinning and temper size reduction throughout the Woodland period will be reflected in the Cloudman assemblage. However, there was no significant statistical difference between vessel wall thickness and temper 50 size of Laurel Middle Woodland vessels from Naomikong Point and the Late Woodland vessels from Sand Point (Kooiman 2012, 2016), the possible result of the geographic difference between the sites. The Cloudman site assemblage will allow an investigation of vessel wall thickness and temper size through time at a single locale. The trend of wall thinning is expected to continue into the Protohistoric period, when greater reliance on starchy foods, particularly maize, likely continued or increased, as demonstrated by ethnohistoric accounts of the Huron (Tooker 1991), Ojibwe (Densmore 1979; Hilger 1951) and Iroquois (Waugh 1973). Hart (2012) found Iroquois pottery wall thickness decreased steadily up to A.D. 1600 in New York state. Vessel size (or vessel volume) is a less understood aspect of formal variability. Size variation has been connected to utilitarian function (Rice 1987:299), household size (Nelson 1981; Tani 1994; Turner and Lofgren 1966), and to utilitarian versus ceremonial functions (Blitz 1993; Kooiman 2012, 2016; Potter 2000). Neither McHale Milner (1998) nor Kooiman (2012, 2016) found any correlation between vessel function and vessel size (inferred from rim diameter). While there was no statistical difference in vessel size between Laurel Middle Woodland vessels (from both the Winter site and Naomikong Point site) and Late Woodland vessels originating from the habitation area at the Sand Point site, all three subsamples were significantly smaller than vessels found in mound fill contexts at Sand Point (Kooiman 2015, 2016). Vessels buried in the Sand Point mounds may have played a role in ceremonial gatherings, where larger vessels may have been required to feed large numbers of people. While a mound once existed at the Cloudman site, the location of the materials excavated from the mound is unknown, and the ceramic assemblage used for this study originated from the general occupation area of the site. Therefore, it is not expected that there will be any significant increases in vessel size from the Middle Woodland to the Late Woodland period. 51 The social affiliation of the Protohistoric occupants of the Cloudman site is still up for debate. Based on early dates of the trade goods from this component, Branstner (1995) and Cleland believed the site to have been occupied by Odawa groups ca. 1630; however, if the occupation was later the site may have been inhabited by Huron groups migrating westward. If the latter is true, then households may have been larger in the tradition of the longhouse (Warrick 2000), and pot sizes may be larger than in prior periods to accommodate the feeding of a great number of people. 2) Are there diachronic changes in ceramic vessel use and cooking habits evident through use-alteration traces? Prior research has found that large proportions of the Naomikong Point and Sand Point pottery vessels were used over a fire, based on high frequencies of certain use-alteration traces, specifically exterior sooting and interior carbonization (Kooiman 2012, 2016). The most prevalent interior carbonization pattern among the Middle Woodland vessels of Naomikong Point was an even veneer of residue along the interior surface, indicative of stewing, while there was a significant increase in carbonization patterns representing boiling behaviors among the Late Woodland vessels at Sand Point. The changing cooking styles over the course of the Woodland Period may have been necessitated by alterations in settlement and subsistence patterning suggested by Cleland (1982), Smith (2004), and Dunham (2014). The intensified exploitation of various starchy resources that require intensive boiling prior to consumption would result in these changing patterns. The Cloudman assemblage allows comparison of use-alteration patterns through time in one location with a more refined timeline. It is expected that the frequency of vessels with carbonization 52 patterns indicative of boiling will increase throughout the post-Middle Woodland occupations of the site. 3) Are there diachronic changes in subsistence strategies (and possible attendant changes in cooking habits) detectable through lipids, stable isotopes, and microbotanical remains extracted from pottery? Lipid residues, stable isotopes, and microbotanical remains present in absorbed and carbonized food residues provide direct evidence of foods cooked in ceramic vessels. While they may not be representative of the diet as whole, these evidences are optimal for exploring technical change in pottery because the vessels might be altered to suit the components of the diet processed in ceramic vessels. It is predicted that maize may appear in small amounts in the Middle Woodland period, along with other starchy foods, such as wild rice, based on previous findings in northeastern North America (Hart and Lovis 2013). The amount of starchy food is expected to increase throughout the course of the Late Woodland period, along with a possible increase in lipids indicating fatty nuts (i.e., acorns). Given that the site is positioned close to the lakeshore and on the banks of the Potagannissing River, Cloudman occupants likely had access to acorns and wild rice and may have been able to grow maize (Dunham 2014). Evidence of all three resources is expected in the microbotanical, lipids, and stable isotope results, particularly in the Late Woodland period. Maize is expected to be the predominant botanical component of food residues found in context with Iroquoian ceramic vessels given the association of contact-period Huron and Odawa with maize consumption (Smith 1996; Waugh 1973). 53 4) Is there synchronic variation in ceramic vessel use, subsistence strategies, and cooking habits evident through use-alteration studies or detectable through lipids, stable isotopes, and microbotanical remains extracted from pottery of differing typological categories? If the nature of social relationships is reflected in ceramic style, then pottery vessels can be used to examine the relationship between diet/cooking and identity. The social fluidity apparent in the Middle Woodland resulted in a broad cline of indistinct pottery styles across the Upper Great Lakes that have thwarted attempts by archaeologists to isolate spatial groupings (Brose and Hambacher 1999). Per Branster (1995), all the identified Middle Woodland vessels from Cloudman are variations of Laurel ware. Late Woodland peoples, on the other hand, seemed to have become more localized, based on both pottery styles (Milner 1991) and settlement and subsistence patterns (Cleland 1982; Dunham 2014; Smith 2004). Although pottery types may be more differentiated and geographically restricted, the need for increased extra-local interaction for risk-buffering purposes often results in a mixture of these distinct types at seasonal aggregation sites (Carroll 2013; McHale Milner 1998; see also Dorothy 1978; McPherron 1967). Early Late Woodland types at Cloudman are mostly variants of Mackinac Ware, along with a few Blackduck vessels, while the late Late Woodland assemblage is more diverse (Branstner 1995). This seems to support the increase of both localization and seasonal cooperative aggregation of distinct groups at Cloudman. I predict that variation in technical properties and use among Middle Woodland vessels at Cloudman will be low. I also predict that results of the lipid residue, stable isotope, and microbotanical analyses will be relatively uniform across all Middle Woodland vessels, with little variation between the Laurel subtypes. If distinct identities were not being expressed through ceramic traditions, it is unlikely that there were distinct cuisines during this time as well. 54 I do predict an increased level of variation in the evidence for technology, cooking techniques, and diet between ceramic types used during the Late Woodland period (as compared to the Middle Woodland). Whereas these differences may not be conscious expressions of identity through cooking and food choice, they may reflect the effect of localized and nested spheres of interaction, where subsistence traditions might be more restricted in scope, and access to certain resources may be restricted as well. As Dunham (2014) points out, the availability of different starchy foods—maize, acorns, and wild rice—is restricted based on location, and individual annual yields may be highly variable. Although residential groups may have been coming together at Cloudman for coordinated large-scale fish exploitation, they also may have maintained some of their unique food and cooking preferences while inhabiting the site. Therefore, I predict increasing variation in technical properties, cooking techniques, and diet among vessels throughout the Late Woodland period. The exact social affiliation and composition of the assumed Protohistoric occupants of the Cloudman site remains unclear, therefore predictions concerning expressions of identity are made difficult. Branstner (1995) identified Huron vessels, “imitation” Huron vessels (with poorer construction), and a number of vessels not attributable to a defined type. It is predicted that ceramic vessel technical attributes may vary, but if the social composition of the occupants was homogenous, then little variation in diet and cooking techniques is expected. If variation in diet and cooking occurs, it could be indicative of a heterogeneous population at the Cloudman site during this late occupation. Outcomes can then be compared to ethnographic and ethnohistoric accounts of group-specific culinary habits to evaluate potential ethnic associations with specific ceramic types. 55 5) How do ethnographic and ethnohistoric accounts of indigenous diet and cooking in the Great Lakes inform interpretations of ancient cuisine generated from the archaeological data? This final question is auxiliary to and derived from the primary research questions but serves an important role in the overall study. Results of the archaeological analyses will be examined in the context of ethnographic and ethnohistoric accounts of foodways and cooking of societies that historically occupied regions geographically adjacent to the Cloudman site, including the Ojibwe to the west and Iroquoian groups to the east. Reference to actualistic observations of human interaction with a similar environment and suite of resources will be used to enhance the interpretive outcomes of the study. Methods of food preparation and food combinations practiced by Ojibwe and Iroquoian cooks will strengthen our understanding of the data derived from the use-alteration and dietary analyses applied to the Cloudman pottery assemblage. They may also delineate culinary differences between ethnic groups and enable recognition of identity through archaeological food remains. Conclusion This study follows a strong, decades-long tradition of research on subsistence and ceramics in the upper/northern Great Lakes. Although previous studies have created a substantial base of knowledge on these topics, questions about the nature of dietary, technological, and social change throughout the late precontact and contact periods remain. The use of new perspectives and techniques applied to a pottery assemblage from the multicomponent Cloudman site will enhance interpretive outcomes about the ways in which precontact peoples negotiated changing natural and social environments. 56 CHAPTER 4 METHODS AND DATA COLLECTION Introduction The multiproxy approach of this study applies diverse yet complementary analytic methods to ceramic vessels from the Cloudman site. This includes functional and stylistic ceramic analysis, and residue analysis, the latter including microbotanical, stable isotope, and lipid analyses. To evaluate the potential for taphonomic contamination of the residues, stable isotope and lipid residue analyses of soil samples taken from the matrix for the analyzed materials at the Cloudman site, were conducted for comparative purposes. The procedures carried out for each analysis are detailed below. A majority of the analysis of the Cloudman pottery assemblage took place in the Consortium for Archaeological Research Laboratories, Department of Anthropology and MSU Museum, Michigan State University. This included the recordation of technical attributes and use-alteration traces; assessment and selection of samples for residue analysis; and collection of adhered residues for microbotanical analysis, stable isotope analysis, and AMS dating. Microbotanical analysis, conducted by Rebecca Albert, also took place in the laboratories. The Cloudman Pottery Assemblage Branstner (1995) grouped and identified a total of 177 vessels, excavated between 1990 and 1994, during her initial analysis of the assemblage. These vessels were numbered 1-190, with vessel numbers 68, 84, 107, 111, 119, and 92-99 unused. An additional 28 vessels were discovered in 1995 during a brief and limited excavation conducted by Cleland; these were sorted, grouped, and given vessel numbers 191-218, but never attributed to stylistic categories 57 (Christine Stephenson, personal communication). Rim and body sherds of each vessel were boxed and bagged together following the original analysis, reducing the chance of evaluating or sampling the same vessel twice. During re-analysis, it was discovered that Branstner’s Vessels 196 and 203 were from the same vessel, as were Vessels 40 and 153, so they have henceforth been grouped together (as Vessel 196/203 and Vessel 40/153). The Cloudman assemblage consists of a total of 202 discrete identified vessels, which served as the primary unit of analysis for this research. Principals of Pottery Function Analysis of pottery function is detailed by Skibo (1992, 2013). He divides function into two types: intended and actual. Intended function is rooted in the premise that “all pots are designed to be used” (Skibo 2013:27). More specifically, potters are intentional in the decisions they make about vessel form and composition, decisions that Skibo terms technical choices. These choices determine the formal properties/physical characteristics of a ceramic vessel, such as shape, size, thickness, surface treatments, and paste composition (clay type and chemistry; temper type, size, and density). Each technical property affects one or more performance characteristics of a vessel, or how well a vessel can perform in various circumstances. Performance characteristics include thermal shock resistance, heating effectiveness, impact resistance, permeability, and even gripability. Those most important for cooking vessels are thermal shock resistance (the ability of a vessel to resist breakage when heat is applied to it) and heating effectiveness. Some technical properties enhance certain performance characteristics while detracting from others; for example, 58 adding temper to a paste will increase a vessel’s thermal shock resistance while simultaneously decreasing its impact resistance. Actual function entails the ways in which pottery was used, regardless of its technical properties. Actual function is inferred from use-alteration properties, which include: exterior sooting, exterior carbonization, interior carbonization, attrition, and absorbed residues. Exterior sooting forms from contact of smoke with the pottery surface, indicating the use of a vessel over fire. Exterior and interior carbonization refer to burned food residues and provide direct evidence not only of cooking but cooking methods. Various cooking methods (e.g., roasting, stewing, boiling) result in different patterns of interior carbonization. Attrition refers to the removal ceramic material through abrasive or nonabrasive processes and can occur through contact with other objects/surfaces, water vaporization, salt crystallization, and fermentation processes occurring within the vessel (Skibo 2013:122). Absorbed residues are analyzed primarily through lipid residue analysis, which examines fatty acid ratios and biomarkers (Malainey 2011). Functional Analysis of Cloudman Pottery Analysis of pottery function was carried out in accordance with the standards established by Skibo (1992, 2013). Distinct vessels (rim and body sherds as grouped by Branstner [1995]) were the unit of analysis. One set of measurements was taken per vessel. Intended Function Intended function was assessed through physical characteristics. Branster (1995) originally recorded rim diameter, paste type, paste hardness, temper size, temper type, manufacture type, thickness, lip form, lip profile, lip eversion, rim thickness, rim height, rim 59 profile, rim eversion, surface treatment, decoration locations, and the presence of residues. The physical attributes chosen as the focus for this study were selected based on the results of prior studies of morphological changes attending alteration of cooking techniques in Eastern North American (Braun 1983; Hart 2012) and on attributes which proved significant in prior studies of pottery assemblages from the northern Great Lakes (Kooiman 2012, 2016). These properties include rim diameter, wall thickness, and temper size. Wall thickness and temper size were chosen because of their previously determined sensitivity to detecting functional change in relation to changing cooking requirements (Braun 1983; Hart 2012). Rim diameter was chosen for its relationship to overall vessel size (Blitz 1993, Kooiman 2015a; Potter 2000) and for its standard recordation in ceramic analysis. The function of rim diameter is variable and its connection to cooking effectiveness is undetermined, and therefore its role in this study is also exploratory. Since 28 vessels were not part of Branstner’s original analysis, chosen characteristics were re-measured on all vessels for analytic consistency. Rim diameter was measured using an Archmat Ceramic Radius Template. Only rim sherds large enough to comprise 5% or more of the total rim circumference of their respective vessel were recorded, resulting in rim diameter data for only 47% (94 of 202) of the total vessels. Wall thickness was measured at the lip, rim, neck, shoulder, and body, when possible. Vessels comprised of only exfoliated sherds were excluded from this data set. Temper size was averaged from three measurements per vessel. Wall thickness and temper particle size were measured using Mitutoyo Digimatic CD-6” digital calipers. 60 Actual Function Assessment of actual function of pottery vessels is based on the presence of use-alteration traces. Each vessel was examined for exterior sooting, exterior carbonization, interior carbonization, and attrition. Exterior sooting can be difficult to distinguish from staining from contact with the burial environment or fireclouding that forms during the firing process; however, fireclouding occurs infrequently on vessels in the northern Great Lakes. Sooting was marked as present only when it could be positively distinguished and identified. Exterior carbonization—burned food residue occurring on the exterior surface of the vessel—sometimes has the appearance of sooting; therefore, it was only categorized as carbonization when it was located on the rim (where sooting infrequently occurs) or in cases where there was continuity with carbonization on the interior surface. Attrition, which is rare among coastal northern Great Lakes Woodland pottery assemblages (Kooiman 2012, 2015b), was not observed in the Cloudman assemblage. Interior carbonization was recorded as present or absent, and its distribution patterning was also categorized. Presence was recorded only in cases where carbonization could be confidently identified. The categorization scheme of interior carbonization patterns from the Naomikong Point and Sand Point sites is depicted in Figure 4.1 (Kooiman 2012, 2016); however, these categories proved unsuitable for statistical analysis. These categories were condensed into three categories: Types 1 and 2, which indicate boiling; Types 3 and 4, in which the vessel was not filled to the top, but ambiguous as to whether boiling or stewing occurred; and Type 5, indicative of stewing (Kooiman 2016). For the Cloudman pottery assemblage, the patterns could be categorized most meaningfully by the types depicted in Figure 6.2 (see Chapter 6): Type 1, which represents a 61 distinct band of carbonization at or near the rim of the vessel, indicative of boiling; Type 2, in which carbonization covers a significant portion of the interior vessel wall, indicative of stewing (a long term, liquid reduction process); Type 3, representing the presence of a thick ring of residue around the rim of the vessel with lighter carbonization along the body of the vessel, indicating both boiling and stewing processes; Type 4, in which there is definitive carbonization on the rim but the lower extent of the carbonization along the interior is unclear; and Type 5, which signifies the presence of interior carbonization with no discernable pattern. Figure 4.1: Example Interior Carbonization Pattern Categorizations (Kooiman 2012, 2016) 62 Ceramic Taxonomic Classification Branstner (1995) was able to assign 92 of 177 identified vessels from the 1992 and 1994 excavations to existing typologies. Another 49 vessels, most of which she attributed to the late Late Woodland period, were categorized as “untyped” due to their failure to fit into established categories. Another 36 vessels were classified only by their decoration (e.g., “crosshatched/impressed lip’). Branstner attributed these vessels to Middle Woodland (n=31), Middle Woodland/Late Woodland transitional (n=6), early Late Woodland (n=48), late Late Woodland (n=45), Late Woodland/Protohistoric transitional (n=4), and Protohistoric (n=34) periods. She also identified 9 miniature vessels from the Late Woodland and Protohistoric periods. An additional 28 vessels, excavated by Charles Cleland in 1995 but never officially reported, were initially identified and grouped but never assigned to stylistic typologies (Branstner-Stephenson, personal communication). Upon review, most of Branstner’s initial identifications were corroborated, and typological assignations for only 19 vessels were altered or modified. Next, the remaining untyped vessels were re-examined to determine if they fit into any established typologies. This assessment revealed the affinity of many of these untyped vessels to types from Ontario. Given the proximity of the Cloudman site to this region, an extensive perusal of literature from assemblages from Ontario was undertaken (e.g., Kenyon 1970; MacNeish 1952; Wright 1973). This deeper investigation revealed many of these untyped vessels resembled Ontario ceramic wares, particularly various Iroquoian (Huron/Wendat-Petun) wares. Additional sources (e.g., Hambacher 1992; Lovis 1973; Lugenbeal 1978) for ceramics not falling into the local Juntunen sequence yet exhibiting traditional Woodland characteristics were also consulted. 63 Following the reassessment of the Cloudman pottery and incorporation of miniature vessels and those vessels excavated in 1995, it was found the assemblage was composed of Middle Woodland (n=36), Middle Woodland/Late Woodland transitional (n=5), early Late Woodland (n=65), middle Late Woodland (n=5), late Late Woodland (n=47), and Ontario Iroquois (n=37) vessels. There are also five (5) general Late Woodland vessels, and two (2) not attributable to any socio-temporal style/period. A total of 29 vessels remain untyped, with all other vessels attributable to at least a ware category or marked as resembling a type or ware. A more detailed discussion of the results of the taxonomic classification will take place in the following chapter. Microbotanical Analysis Microbotanical analysis has recently emerged as an increasingly common method for investigating past diet and environment. Microbotanical remains include pollen, starches, and phytoliths, and may be incorporated into various forms of residue, adherance or accretion. While pollen is useful for reconstructing ancient environments, it is not typically found in contexts that can inform human food choice and consumption. Phytoliths and starches can likewise be found in the soil and used for environmental reconstruction, but they can also be collected from artifacts such as pottery and grinding stones, connecting them directly to food processing (Pearsall and Hastorf 2011). Identifiable to plant species, phytoliths and starches can tie the processing of a specific type of plant to a ceramic vessel, providing a higher level of specificity than is possible with other types of cooking residue evidence, such as lipids or stable isotopes. While not all plants yield starches or phytoliths that preserve archaeologically, many plants 64 important for understanding subsistence in the Eastern Woodlands (i.e., wild rice, maize, and squash) have been identified in archaeological residues (see Raviele 2010; Simon 2011). Starch and phytolith analyses are highly complementary and are increasingly used in tandem. Raviele (2010) analyzed phytoliths and starches contained in adhered burned food residue in Middle and Late Woodland ceramic vessels from Michigan. The earliest residues (ca. cal 200 BC) contained maize starch, which comes from the kernel, but not maize phytoliths, which usually derive from cobs. She hypothesized this indicated a lack of cultivation activities and the use of dried maize traded from the south. Samples for microbotanical analysis were collected from 48 discrete pottery vessels (Table 4.1). Selection of vessels for sampling was based on two criteria: presence of sufficient interior carbonized residue, and the socio-temporal association of the vessel. While 51% (n=104) of the Cloudman vessels had traces of interior carbonization, some only had thin layers of residue or residue staining. Based on standards set forth by Raviele (2010), a minimum of 0.001g was required for the collection of residue, but a minimum of 0.002g was preferred. This excluded some vessels from consideration. Selection of samples was also based on the associated time period and stylistic typology of the vessels. Preserved adhered residue was present on Middle Woodland, early Late Woodland, late Late Woodland, and Iroquoian pottery in ample amounts for testing, in relative proportion to the size of the subassemblages. Within these larger categories, sampling from vessels of different typological categories (e.g., Mackinac Banded vs. Mackinac Punctate) were targeted to capture vessel function across the range of varieties. Residue sampling was primarily restricted to the rim and neck; collection of residue from body sherds was only conducted if there was insufficient residue from rim/neck sherds of the 65 Table 4.1: Vessels Sampled for Microbotanical Analysis, Lipid Residue Analysis, Stable Isotope Analysis, and AMS Dating Vessel Period Type Micro- Botanical Stable Isotope Lipids AMS 1 MW 4 MW 5 MW 6 MW 10 MW 12 MW 20 MW 22 MW 23 MW 28 MW 109 MW 112 MW Laurel Pseudo-scallop Shell Laurel Dentate Stamped Laurel Pseudo-scallop Shell Laurel Dentate Rocker Stamped Laurel Pseudo-scallop Shell Laurel Dentate Stamped (oblique) Laurel Dentate Stamped (oblique) Laural Pseudo-scallop Shell Laurel Banked Linear Stamped Laurel Pseudo-scallop Shell Laurel Banked Linear Stamped Laurel Banked Linear Stamped Laurel Banked Linear Stamped North Bay Linear Stamp 114 MW 131 MW 35 MW/LW Late Laurel (cross-hatched) 118 MW/LW 8 34 41 46 50 55 76 80 81 88 100 103 ELW ELW ELW ELW ELW ELW ELW ELW ELW ELW ELW ELW Untyped (cordmarked/undecorated) Mackinac Ware Mackinac Ware Mackinac Ware Mackinac Ware Mackinac Banded Mackinac Ware Mackinac Undecorated Mackinac Punctate Blackduck Banded Blackduck Banded Mackinac Punctate Mackinac Banded 66 X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X Table 4.1 (cont’d) Vessel Period Type Micro- Botanical Stable Isotope Lipids AMS 105 ELW 120 ELW 122 ELW 124 ELW 132 ELW 173 ELW 174 ELW 175 ELW 191 ELW 193 ELW 42 MLW 215 MLW 24 LLW LLW 25 26 LLW 43 LLW 75 101 102 150 152 204 205 216 36 LLW LLW LLW LLW LLW LLW LLW LLW IRO 40/153 IRO IRO IRO 70 77 146 IRO Mackinac Punctate Mackinac Banded Mackinac Ware (cf. Punctate) Mackinac Banded Mackinac Ware (cf. Punctate) Mackinac Banded Mackinac Ware (cf. Punctate) Mackinac Punctate Mackinac Punctate Blackduck Banded Bois Blanc Ware Bois Blanc Ware "proto-Juntunen" Ware (plain) Juntunen Ware Juntunen ware Traverse Decorated v. Punctate Traverse Plain v. Scalloped (mini) Juntunen Ware Juntunen Drag-and-Jab Traverse Plain v. Scalloped Juntunen Ware (cf. O'Neil Curvilinear) Juntunen Linear Punctate Juntunen Linear Punctate Untyped cf. Huron Incised cf. Lawson Opposed or Methodist Point Group 7 cf. Huron Incised Untyped Early Ontario Iroquoian (incised) X X X X X X X X X X X X X X X X X X X X X X 67 X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X Table 4.1 (cont’d) Vessel Period Type Micro- Botanical Stable Isotope Lipids AMS 162 164 166 179 IRO IRO IRO IRO cf. Sidey Notched or Lawson Incised cf. Huron Incised cf. Huron Incised cf. Huron Incised TOTAL X X 48 X X 50 X X X 30 X 5 same vessel. The residue was collected from the interior surface of the sherd and gently scraped onto weigh paper from each sherd (representing one vessel) with disposable scalpels. After each residue sample was weighed on a Cole-Parmer® Symmetry™ scale and placed into 15 mL plastic centrifuge tubes with screw caps and the weigh paper discarded. Samples were transported to the MSU Department of Geography Quaternary Landscape Research Group Laboratory for processing. The processing procedure is as follows: The chemical extraction technique for extracting microbotanicals from ceramic residues follows the protocols developed by the University of Wisconsin–Duluth Archaeometry Laboratory (Archaeometry Laboratory 1989). This procedure uses a consistently heated solution of potassium chlorate (KClO3) and nitric acid (HNO3) to dissolve non-silica materials. Materials left after this procedure are phytoliths, starches, diatoms, silicified tissues, and other various non-diagnostic organic and inorganic materials. All samples were centrifuged, and thoroughly rinsed with distilled water (dH2O) to remove any remaining potassium chlorate and nitric acid. These samples were stored in ethanol (C2H6O) in ½ dram glass vials. A few drops of each sample was placed on slides and were then placed on a slide 68 warmer to evaporate the ethanol. The sample was then mixed with PermountTM on the slide, and a slide cover was placed over the PermountTM. Samples were viewed and analyzed using Leica DM 2500™ compound microscope at 200X magnification. The Hoya polarizers were generally set very high, and a green Hoya filter was occasionally used to diminish the light intensity and search for starches. Diagnostic objects (see materials and methods) identified on the slides were photographed and their location on the slide was marked, along with a morphological description (Albert et al. 2018). Potential contamination of the samples during processing was controlled and monitored. The samples were prepared in a sterile environment, and powder-free (starch-free) nitrile gloves were used during preparation and analysis. Control slides placed in the processing laboratory demonstrated the lack of contaminates in the collection and processing laboratories. Stable Isotope Analysis Stable isotope analysis in this study focused on isotopes of the standard elements Carbon and Nitrogen, which enables measurement of their ratios in adhered or absorbed food residues (Evershed et al 1999). Resultant ratios facilitate the separation of plants into three groups: legumes, non-legumous C3 plants, and C4/CAM plants (Hastorf and DeNiro 1985). Most plants in temperate climates are C3 pathway plants, while tropical cultigens, such as maize, are C4 pathway plants. This makes it possible to detect the presence of maize in human diets, especially in temperate zones such as the Midwestern United States (Schoeninger 1995). Hastorf and DeNiro (1985) were the first to apply stable isotope analysis to pottery residues for assessing 69 past vessel contents, and it has since an increasingly common method for investigating past diet (e.g., Hart 2012; Lovis 1990; Morton and Schwarcz 2004; Taché and Craig 2015). Samples for stable isotope analysis were collected from 50 discrete pottery vessels (see Table 4.1). Selection of vessels for sampling was based on two criteria: presence of sufficient interior carbonized residue, and the socio-temporal association of the vessel. Although 51% (n=104) of the Cloudman vessels had traces of interior carbonization, some only had thin layers of residue or residue staining. A minimum sample of 0.0015g was required by ISGS for analysis. Residue for microbotanical analysis and stable isotope analysis was collected from the same vessels simultaneously. An additional two samples (from Vessels 100 and 191) were collected for isotope analysis but not microbotanical analysis due to time and budgetary restrictions. Selections for sampling were also based on the associated time period and taxonomy of each vessel. Preserved adhered residue was present on Middle Woodland, early Late Woodland, late Late Woodland, and Ontario Iroquois pottery in ample amounts for testing, in relative proportion to the size of the subassemblages. Within these larger categories, sampling from vessels of different regional archaeological types (e.g., Mackinac Banded vs. Mackinac Punctate) were targeted for sampling to capture vessel function across the range of varieties. Procedures for sampling were the same for collection of residues for microbotanical analysis (see previous section). Residue samples were then sent to the Illinois State Geological Survey for bulk δ13C and δ15N isotope analysis. Samples were analyzed using a CE Instruments NC 2500 Elemental Analyzer in series with a ConFlo IV universal interface coupled to a Delta V Advantage Isotope Ratio Mass Spectrometer. Samples were compared with NIST and IAEA Standard Reference Materials (USGS 40, USGS 41, IAEA NO-3) for precision and accuracy, while Methionine and 70 USGS 40 were used as the standard reference materials for elemental composition. Analytical precision for δ13C was + 0.05 ‰ and is reported relative to Vienna Pee Dee Belemnite (VPDB). Analytical precision for δ15N was + 0.13 ‰ and is reported relative to Air (Shari Effert-Fanta, personal communication). Lipid Residue Analysis Lipids, or fatty acids, are present in both plant and animal tissue and provide direct evidence of food processing for consumption (Malainey 2011). They are effective for ancient dietary assessment because they are present in almost all human food, they maintain high stability with increased temperatures, and decomposition is minimal when compared to carbohydrates or proteins (Röttlander 1990). Absorbed lipid residues can be extracted from archaeological materials associated with cooking and food processing, such as ceramic vessel walls or cooking and grinding stones. They can be analyzed and assessed based on fatty acid composition/ratios, the presence of certain biomarkers, or by measuring the stable isotope values of lipid components. Components are generally separated with some form of gas chromatography: gas chromatography with a flame ionization detector (GC), gas chromatography with mass spectrometry (GC/MS), or gas chromatography-combustion-isotope ratio analysis (GC-C-IRMS) (Malainey 2011:201). Once the components are separated, the origins of the lipids present can be identified. Sampling for lipids can be affected by three factors: accumulation, preservation, and contamination (Malainey 2011). Lipids are differentially accumulated into the ceramic material of a cooking pot based on vessel portion. Accumulation is generally low in the base of the vessel, while lipids are most plentiful near the rim of the vessel, or just below the level to which the 71 vessel was regularly filled (Charters et al. 1993; Charters et al. 1997). Preservation of lipids is also optimal in the upper portions of a vessel (Malainey 2007). Contamination derives either from the burial environment or from archaeological processing (Malainey 2011:205). The densest accumulation of lipids occurs in the interior of a pottery vessel, where contamination from the burial environment is generally negligible (Condamin et al. 1976; Röttlander 1990). Standard processing of pottery for lipid analysis therefore involves the removal of exterior surfaces and sampling from the interior portion of the wall only. A total of 30 pottery sherds were selected from the Cloudman site assemblage for lipid residue analysis (see Table 4.1). Criteria for selection included: sherd weight, vessel portion, the presence of adhered residue on the vessel, and temporal association of the vessel. Analysis of fatty acid ratios requires sherds with a minimum weight of 7g for the purposes of sufficient sampling. As noted above, lipid accumulation and preservation are optimal in the upper portions of vessels; therefore, sherds originating from the neck or shoulder portions of the vessels were favored in the selection of samples. Curatorial concerns also called for the selection of neck and shoulder sherds containing no portion of the rim, although in a few cases the only sherd available for analysis were rim sherds. Sherds from vessels exhibiting adhered residue were favored over vessels exhibiting no visible residue. Adhered residue is a certain indicator that the vessel was used for cooking and increases the likelihood of abundant absorbed residue. Vessels lacking visible residues may have been employed in food processing and may contain absorbed lipids in cases where food was cooked in the vessel but never burned, or where the vessel was cleaned either by the original user or washed during archaeological lab processing. However, some vessels may have never been involved in cooking foods, either used for storage or simply breaking during the firing process and entering the archaeological record before it was used. 72 Vessels associated with each occupation period of the Cloudman site were sampled. However, differential preservation of vessels from each time period affected the sampling. Early Late Woodland vessels were the most numerous and the best preserved, with ample numbers of sherds of sufficient size from the upper portions of the vessels. Middle Woodland vessels, based on age and decoration, generally had smaller sherds, many of which did not meet the minimum size for lipid analysis. Additionally, these vessels did not have as many body sherds associated with the rim sherds used for vessel categorization when compared to other subassemblages. Selected sherds were sent to the Archaeological Residue Analysis Laboratory at Brandon University in Brandon, Manitoba, for analysis. Processing and analysis was conducted by Mary Malainey and Timothy Figol. HT-GC and HT GC-MS analysis on a Varian 3800 gas chromatograph and a Varian 4000 mass spectrometer was used to determine fatty acid ratios used in identification criteria (C12:0, C14:0, C15:0, C16:0, C16:1, C17:0, C18:0, C18:1ω9, C18:1ω11 and C18:2). Ratios from the Cloudman site samples were then compared against known ratios of plants and animals common to northern North America. Detailed preparatory and analytic procedures of the lipid residues are outlined in Appendix C. AMS Dating Accelerator mass spectrometry (AMS) dating allows detection of carbon-14 in small samples of organic matter. This makes it ideal for directly dating adhered pottery residues (Lovis 1990). Although some have argued that the processing of freshwater fish in ceramic vessels could introduce old carbon into food residues (aka the Freshwater Reservoir Effect), thereby affecting dating outcomes (e.g., Fischer and Heinemeier 2003; Philippssen et al. 2010), recent 73 studies have demonstrated that the effect is minimal (Hart et al. 2013; Heron and Craig 2015; Lovis and Hart 2015). Carbonized residue from five pottery vessels from the Cloudman site were selected for direct AMS dating. Sample selections were based on vessel type to clarify both the occupational chronology of the site and the regional ceramic chronology (Table 4.2; see Table 4.1). A Laurel Dentate Stamped vessel (V4) was selected because of the lack of late Laurel dates in the northern Great Lakes, while a Laurel Pseudo-Scallop Stamped vessel (V1) was anticipated to provide additional information about the timing of the Middle Woodland occupation at Cloudman. The timing of the occurrence of Blackduck pottery in the eastern northern Great Lakes is unclear, so one Blackduck vessel (V193) was sampled. While regional chronologies for the late Late Woodland pottery types are well established (Lovis 2014; McHale Milner 1998), the early Late Woodland chronology is less clear. A Mackinac Banded vessel (V103) was selected to situate this type within the regional ceramic chronology and to clarify the timing of the early Late Woodland occupation of the Cloudman site. Finally, an Iroquoian vessel (V162) closely resembling either Lawson Opposed or Sidey Notched vessels was selected to date the presence of Iroquoian/Wendat-Petun occupants or trade vessels at the Cloudman site. Prior dating of this most recent occupation of the site was based solely on relative dating of various trade goods. Table 4.2: Pottery Vessels Sampled for AMS Dating of Carbonized Residue Vessel No. 1 4 103 193 162 Period Type Laurel Pseudo-Scallop Stamp Laurel Dentate Stamped MW MW ELW Mackinac Banded ELW Blackduck Banded IRO cf. Lawson Opposed or Sidey Notched 74 Wt. (g) 0.0185 0.0073 0.0044 0.0142 0.0154 Collection of residues from the selected vessels followed the same procedures as for stable isotope and microbotanical analyses (see above). Samples were required to weigh at least 0.004g, with ideal weight of 0.005-0.010g preferred. The five Cloudman site samples were graphitized in the Carb, Water, and Soils Lab of the USDA-FS Northern Research Station in Houghton, Michigan. The radiocarbon measurements were conducted at the Keck Carbon Cycle AMS Facility, Earth System Science Department, University of California Irvine. Soil Samples The sole field component of this study involved collection of soil samples from the Cloudman site. A primary concern with testing pottery residues is contamination from the soil/burial environment. The purpose of the soil samples was to assess the potential of contamination of absorbed and adhered food residues on pottery via the burial environment. Morton and Schwarcz (2004) found the humic contamination of stable isotope assays negligible, but 15N levels can be elevated by both natural manure fertilizers or synthetic denitrified fertilizers (Chang et al. 2002), the latter of which has been found to elevate 15N levels in modern-day fish samples in rivers contaminated with fertilizer runoff (Brugam et al. 2017). As fertilizer spread on the ground could likewise contaminate buried artifacts, the chance for a false negative for aquatic resources is possible. Stable isotope analysis of soil samples from archaeological sites can serve as indicators of contamination of the burial environment and artifacts found within it. Contamination of absorbed lipid residues from the burial environment is generally of less concern, as Condamin et al. (1976:199) found it to be negligible. However, soil samples for comparative lipid residue testing were collected for analytic rigor and to affirm the reliable use of lipid residue analysis without concern for contamination. 75 I visited the Cloudman site on July 21, 2017, when Gary Cloudman, the landowner, showed me where the excavations had taken place. Based on his outline of the limits of excavation and the site description provided by Branstner (1995), I identified the three terraces descending towards the Potagannissing River (Figure 4.2). Using a soil probe with a core sampler, I took small core soil samples at three points across each terrace, spanning the limits of the prior excavations. The area was never formally farmed and therefore did not have a plowzone, and most of the cultural material excavated at the site was contained within the uppermost two soil horizons. Samples from both strata were taken at each sampling locale, in accordance with the stratum descriptions detailed by Branster (1995). Each sample was wrapped in tin foil and placed in a labeled plastic canister. Figure 4.2: Terracing at the Cloudman Site (20CH6) 76 A total of 12 soil samples were selected for submission to ISGS for carbon and nitrogren stable isotope analysis (Table 4.5). Samples were selected from across the site and from different strata, with some preference for subsoil. Both strata of the easternmost edge of Terrace 3 were selected because of their proximity to a modern-day garden bed, in which, according to the land owner, maize is occasionally grown. These samples were therefore the most likely to be contaminated from both modern-day corn signatures, as well as possible nitrogen contamination from fertilizers. Table 4.3: Soil Samples from the Cloudman Site Selected for Stable Isotope Analysis Sample No. T2-S1-2 T2-S2-2 T3-S1-1 T3-S2-1 T1-S2-3 T1-S2-2 T3-S2-3 T2-S2-3 T1-S1-2 T1-S2-1 T3-S2-2 T2-S2-1 Context Terrace 2, Stratum 1, Center Terrace 2, Stratum 2, Center Terrace 3, Stratum 1, East Terrace 3, Stratum 2, East Terrace 1, Stratum 2, West Terrace 1, Stratum 2, Center Terrace 3, Stratum 2, West Terrace 2, Stratum 2, West Terrace 1, Stratum 1, Center Terrace 1, Stratum 2, East Terrace 3, Stratum 2, Center Terrace 2, Stratum 2, East Wt (g) 2.3927 2.7298 2.1282 3.6903 3.0748 2.9083 3.5073 3.1991 3.2733 2.7410 5.7370 4.7282 An initial four samples were sent unprocessed to ISGS, where the samples were dried and pulverized. The remaining eight samples selected for isotope analysis were laid out to dry in the MSU Anthropology and Archaeology Laboratories. They were then taken to the labs at the MSU Department of Geography, Environment, and Spatial Sciences and pulverized in a Fritsch Pulverisette 0 vibratory micro mill and sieved through a No. 60 geological sieve, placed in 15mL 77 centrifuge tubes with screw caps and weighed. Equipment was carefully cleaned between the processing of each sample. Another three soil samples were selected for lipid residue analysis (Table 4.6). Samples of a minimum of 10g were sent to the Archaeological Residue Analysis Laboratory at Brandon University for processing and analysis (see Appendix C). A selection of four soil samples will be sampled and processed for microbotanical analysis in forthcoming research. Table 4.4: Soil Samples from the Cloudman Site Selected for Lipid Residue Analysis Sample No. Context T2-S1-2 Terrace 2, Stratum 1, Center T2-S2-2 Terrace 2, Stratum 2, Center T3-S1-1 Terrace 3, Stratum 1, East Wt. (g) 13.59 17.03 12.14 Conclusion The methods detailed above are either historically common or are becoming increasingly popular in archaeological practice. Stylistic pottery analysis, functional pottery analysis, microbotanical analysis, stable isotope analysis, and lipid residue analysis are strong methods for analyzing past human diet- and technology-related behaviors. However, they are most frequently employed independently, providing discrete sets of data that limit interpretations. The use of a wide variety of pottery analyses can expand interpretive potential and provide a more holistic view into the past. The next three chapters will detail the results of the aforementioned analyses. Data yielded from all methods will then be discussed together, demonstrating the efficacy of multiproxy research for investigating past foodways and technology. 78 CHAPTER 5 REGIONAL CERAMIC TAXONOMY AND CHRONOLOGY, AND THE OCCUPATIONAL HISTORY OF THE CLOUDMAN SITE Introduction Exploration of diachronic change requires situating units of analysis in time. In this case, the units of analysis are Cloudman pottery vessels and their residues. Construction of an occupational history of the Cloudman site allows identification and isolation of the pottery assemblage subsets for use in comparative and statistical analyses of technical properties, use- alteration traces, and chemical and microscopic food signatures. While adhered residues from pottery vessels can be directly dated using AMS, this method is cost prohibitive. Relative dating of vessels using spatio-temporal taxonomic typologies based on stylistic properties provides a cost-effective and time-tested means of creating temporal groupings that facilitate examination of diachronic change. This analysis relies primarily on relative dates provided by taxonomic categorizations of pottery, although a small set of AMS-dated vessels provides absolute dates for anchoring the relative chronologies. Pottery Taxonomy Stylistic properties and taxonomic categorization of Cloudman site pottery vessels have been discussed at length by Branstner (1995). The goal of the present study was to review original classifications in comparison to more recent literature and create the most accurate portrayal of the occupational history of the Cloudman site and the social relationships of the people who lived there. Many of the classifications below are a reiteration of the work done by Branstner (1995) and points of classificatory divergence are distinguished. Appendix A presents both Branster’s categorizations and the final categorizations used in this study. 79 One challenge encountered during the taxonomic assessment of the Cloudman ceramic assemblage was the admixture of Ontario Iroquoian or Iroquoian-like wares into the otherwise Woodland/Algonquian assemblage. At the outset of taxonomic re-assessment, the chronological association of the Iroquoian pottery was unclear. As discussed in Chapter 3, Branstner (1995) initially associated Ontario Iroquoian wares with a protohistoric Odawa occupation relatively dated to ca. AD 1630. However, an AMS date confirms the presence of Ontario Iroquoian pottery at the Cloudman site in the 1400s (see below), and Iroquoian wares associated with earlier periods were identified throughout the classification process. This suggested a potential temporal overlap between Ontario Iroquoian and certain Woodland/Juntunen-sequence vessels. Aside from chronological significance, pottery style is also symbolic of sociocultural affiliations. Stylistic differences between Iroquoian and Woodland pottery are, in most cases, quite distinct, despite their temporal overlap, evidence of production by people of distinct social or cultural traditions. Although most of Cloudman pottery vessel subsets, such as “Middle Woodland,” are applied here as spatiotemporal categories, it is recognized that the category “Ontario Iroquoian/Iroquoian” conflates sociocultural with spatiotemporal associations, resulting in what is here called a “socio-temporal” categorical subsets. However, given the clear stylistic differences between late Late Woodland and Iroquoian vessels, they will be grouped and described separately. Another challenging factor was variation in vessel completeness. Some vessels are represented by a single rim sherd (or, in some cases, a highly distinctive neck/shoulder sherd), while others are represented by dozens of sherds comprising large portions of the vessels (although there are no whole vessels in the assemblage). This variation often affected taxonomic categorization. Early Late Woodland vessels generally had the most complete vessels rims and 80 the most associated body sherds, while vessels from other time periods were often represented by small rim sherds or rim sections. The latter limited the specificity of typological categorization. Taxonomic Classification The Cloudman assemblage consists of 202 identified, distinct minimal vessels (Table 5.1; see Appendix G for images of select vessels). The largest subassemblage is comprised of vessels associated with the early Late Woodland period, with sizable Middle Woodland, late Late Woodland, and Iroquoian subassemblages. The remaining vessels either possess stylistic attributes that appear transitional between primary subassemblages or lack distinctive attributes that associate them with a specific subassamblage. Table 5.1: Cloudman Pottery Vessels by Socio-Temporal Association Socio-Temporal Association Middle Woodland (MW) Middle/Late Woodland (MW/LW) Early Late Woodland (ELW) Middle Late Woodland (MLW) Late Late Woodland (LLW) General Late Woodland (LW) Ontario Iroquoian (IRO) Unknown Total Ct. 35 7 62 5 49 5 37 2 202 Regional Dates 200 BC - AD 500/600 AD 500/600 - AD 700 AD 700 - AD 1000 AD 1000 - AD 1200 AD 1200 - AD 1600 AD 600 - AD 1600 AD 1200 - AD 1650 N/A A number of miniature vessels (n=11) are included in the Cloudman assemblage. Although often disregarded as “children’s pots,” several of the miniature vessels from Cloudman are very finely constructed and show clear stylistic elements of established typological 81 categories. Miniature vessels are included in vessel counts for each time period, but the miniature vessel subassemblage as a whole is summarized and described at the end of this section since they were excluded from Branstner’s original vessel descriptions. The assemblage is described below, presented in chronological order of socio-temporal affiliations and then grouped by type. The following descriptions are not meant to be comprehensive and merely highlight the distinguishing characteristics of the vessels by which they were categorized. More detailed morphological and compositional details of Cloudman pottery types and vessels can be found in Branstner (1995). Middle Woodland Ceramic Subassemblage The Middle Woodland pottery subassemblage (n=35) is largely homogeneous, consisting mostly of Laurel ware vessels and two North Bay vessels (Table 5.2; Figures G1-G14). Laurel Ware (n=32) Varieties of Laurel ware at the Cloudman site include Banked Linear Stamped (Vessels 23, 109, 110, 112, 113, 114), which are decorated with oblique stamps created by a plain tool. A single Dentate Rocker Stamped (Vessel 6) is distinguished by oblique dentate stamping on the interior and exterior rim with horizontal rows of dentate rocker stamping on the body. Dentate Stamped vessels (4, 12, 13, 17, 19, 20, 59) exhibit vertical, horizontal or oblique stamping with a notched tool on the exterior rim surface. Pseudo-scallop Shell vessels (1, 2, 5, 7, 9, 10, 14, 15, 18, 22, 28, 169) are decorated along the entire upper body using push-pull stamping, a technique resulting in a pattern resembling scallop shell stamping, and many have a single row of punctates along the rim. Laural Plain vessels (3 and 91) are undecorated, while a single Laurel Trailed vessel (Vessel 16) is decorated with a vertical trailed motif. The above type descriptions are based on the original definitions by Janzen (1968). 82 Table 5.2: Middle Woodland Vessels by Type Type Laurel Banked Linear Stamped Laurel Dentate Rocker Stamped Laurel Dentate Stamped Laurel Plain Laurel Pseudo-scallop Shell Laurel Trailed Laurel Ware, untyped North Bay Cordmarked North Bay Linear Stamp Untyped Total Ct. 6 1 7 2 13 1 2 1 1 1 35 Re-evaluation was consistent with Branster’s (1995) classifications except in a few cases. Vessel 30 was originally categorized as Laurel Incised, but it was instead classified as a general late Laurel vessel (see below). Vessels 39 and 144 were both originally untyped. Vessel 144 displays hallmarks of coil manufacture with a fine, sandy paste. It has a smoothed rim and oblique tool impressions on a rounded lip, which allows it to be broadly categorized as Laurel Ware. Vessel 39 is a miniature vessel also classified as Laurel Ware, described below with other miniature vessels. North Bay (n=2) The Middle Woodland subassemblage includes two North Bay vessels (131 and 142). These were both originally categorized as untyped. North Bay pottery was first identified at the Heins Creek and Mero sites on the Door Peninsula of Wisconsin and is “usually thick, heavy, and extremely crude” (Mason 1966), distinguishing it from its more finely-made Laurel counterparts. Vessel 131 is North Bay Linear Stamped, with notched tool stamping on the rim and the lip. Vessel 142 is North Bay Cordmarked, with deep vertical cordmarking on a vertical rim with a squared lip. 83 Untyped (n=1) Vessel 201 is an untyped miniature vessel and is described below with the other miniature vessels. Middle Woodland/Late Woodland Transitional Subassemblage A small subset of vessels (n=7) appear transitional between Middle Woodland and Late Woodland types (Table 5.3; Figures G15-G17). Three vessels (30, 35, and 170) are Late Laurel wares. Vessels 30 and 35 have broader lips than classic Laurel with straighter rim profiles than is common among Mackinac wares, exhibiting characteristics which straddle the temporal divide. Vessel 30 has a row of vertical incised lines directly below the exterior lip, a row of possible fingernail impressions on the interior rim, and incised crosshatching on the squared lip. Vessel 170 has three rows of oblique linear stamping with crosshatching on the body below. Vessel 35 has cord-wrapped stick impressions on a broad lip, a slightly everted rim, a row of double punctates (possibly made with the distal end of a phalanx of a small mammal), and crosshatched incising over a cordmarked body. Incising is present in low quantities at many Laurel sites (Brose 1970; Janzen 1968; Stoltman 1973; Wright 1967) and some vessels exhibiting this motif are classified as Laurel Incised, a type that generally occurs late in the Laurel sequence (Stoltman 1973). Vessel 35 appears more similar to Mackinac in overall style, but also displays the cross- hatched incising, more evidence in support of the late occurrence of this decorative element. Vessels 33, 118, and 171 are untyped but possess characteristics that merge both Middle and Late Woodland features. Vessel 118 displayed a thin, rounded lip, vertical rim, and cordmarking on the exterior, defying established taxonomic categories and straddling the characteristics between typical Middle Woodland and Late Woodland wares. Vessel 171 has vertical punctate decoration on the exterior and vertical cord-wrapped stick impressions on the 84 interior rim, exhibiting a slightly everted rim but with a thin, rounded lip, again suggesting a time of manufacture during the Middle-Late Woodland transition and prohibiting taxonomic classification. Vessel 197 has straight rim profile, a square, flattened lip extruded to the exterior, which was then smoothed flat over a cordmarked exterior surface. It bears strong resemblance to North Bay Cordmarked wares, but is composed of a coarser paste than typical of Middle Woodland pottery. Table 5.3: Miscellaneous Woodland and Unknown Vessels Temporal Affiliation Type Middle/Late Woodland Transition Middle Late Woodland General Late Woodland Late Laurel Untyped Total Blanc Blanc Ware Untyped (cf. Bois Blanc Ware) Total Untyped, ELW/MLW Untyped Total Unknown Untyped Ct. 3 4 7 4 1 5 1 4 5 2 The most intriguing of these transitional vessels is Vessel 33, which has a vertical rim, a flattened lip, large grit particles, smoothed-over vertical textile impressed exterior surface, double cordwrapped-object punctates, and a thin, folded-over flap resembling an incipient collar. The paste, grit, and rim orientation are more in alignment with Middle Woodland manufacture, while the decorative elements are more common in the Late Woodland. Overall, the vessel most closely resembles Blackduck or Sandy Lake wares from northern Minnesota (Jill Taylor- Hollings, personal communication). The vessel does not fit into any established typological 85 categories, but may be incipient Blackduck ware. The aforementioned vessels post-date the other Laurel types present in the assemblage, representing an ephemeral human presence at the Cloudman site after the primary Middle Woodland occupation. Early Late Woodland Ceramic Subassemblage The early Late Woodland pottery vessels (n=62) comprise 31% of the overall Cloudman assemblage, representing the largest subassemblage. The category includes Mackinac, Blackduck, and Bowerman wares (Table 5.4; Figures G18-G40). Although the early Late Woodland subassemblage is slightly more stylistically diverse than the Middle Woodland subassemblage, the variation is limited. Mackinac Ware (n= 56) Mackinac wares predominate within the early Late Woodland assemblage. Mackinac Punctate vessels (21, 53, 80, 100, 105, 123, 139, 175) were identified by Branstner (1995), and Vessels 191 and 196/203, not included in her analysis, have also been placed in this category. Mackinac Punctate vessels have outflaring rims with exterior decoration restricted to a single or double row of punctates (McPherron 1967). A number of vessels (106, 122, 132, 137, 138, 141, 174) closely resembled Mackinac Punctate but were not complete enough for confident classification. Branstner (1995) identified nine Mackinac Banded vessels (50, 78, 103, 108, 120, 121, 124, 172, 173). Vessel 192, which was not included in Branstner’s original analysis, was also categorized as Mackinac Banded. This type is defined by complex designs in horizontal or diagonal bands of decoration bordered by rows of punctuations at the top and bottom (McPherron 1967). An additional vessel (V137) closely resembled Mackinac Banded. 86 Mackinac Undecorated, which have the everted rim and splayed lip hallmarks of Mackinac but lack exterior decoration (McPherron 1967) includes Vessel 76, as identified by Branstner, and Vessel 52, originally untyped, which is a miniature vessel and is described later in this chapter. Eight vessels (116, 125, 128, 129, 147, 168, 176, 198) closely resembled Mackinac Undecorated but are not complete enough for confident classification. Another fifteen full-size vessels (8, 2, 32, 41, 46, 49, 55, 61, 72, 87, 126, 127, 130, 181) and two miniature vessels (83 and 202) were broadly categorized as Mackinac ware. Vessels 189 and 194 represent late Mackinac ware (ca. AD 1000). Vessel 189 is relatively thin and has a lip that is an amalgamation of the thickened lip of Mackinac wares and the braced rim of Bois Blanc. The rim is almost vertical, like Bois Blanc, with a hint of Mackinac-like eversion. Vessel 194 displays the thickened lip, rim eversion, and punctates of Mackinac but has multiple bands of decoration, with thinner and more rounded lip than is typical of Mackinac ware, suggesting a time of manufacture late in the sequence. Blackduck Ware (n=4) The early Late Woodland assemblage also includes four (4) Blackduck Banded vessels, characterized by horizontal bands of cord-wrapped stick impressions under a row of punctates (McPherron 1967). Branstner (1995) identified two Blackduck Banded vessels (81, 88) and more tentatively classified Vessel 117 as either Laurel or Blackduck. This vessel has cord-wrapped stick impressions on a broad lip and exterior decoration of a row of round punctates superior to three bands of horizontal cord-wrapped stick impressions, characteristics more closely aligned with Blackduck Banded. Vessel 193, which was not included in the 1995 analysis, clearly displays the hallmarks of Blackduck. 87 Table 5.4: Early Late Woodland Vessels by Type Type Mackinac Banded Mackinac Punctate Mackinac Undecorated Mackinac Ware (cf. Banded) Mackinac Ware (cf. Punctate) Mackinac Ware (cf. Undecorated) Mackinac Ware, untyped Late Mackinac Ware Blackduck Banded Bowerman Plain v. Cordmarked Untyped Total Ct. 10 10 2 1 6 8 17 2 4 1 1 62 Bowerman Ware (n=1) Vessel 199 is Bowerman Plain variety Cordmarked, with the straight rim, flat lip, and fine vertical cordmarking that are the hallmarks of the type (Hambacher 1992). This type was identified at the Skegemog Point site in northern lower Michigan and is considered generally contemporaneous with Mackinac wares (Hambacher 1992: 91). Untyped (n=1) Vessel 63 is a miniature vessel that appears to have been manufactured during the Early Late Woodland period but remain untyped. It is detailed later in the chapter in the section on miniature vessels. Early/Middle Late Woodland Transition (n=1) Vessel 200 is an untyped vessel with a flared, everted rim with deeply-impressed oblique cord-wrapped stick impressions on the lip, vertical cord-wrapped stick impressions on the interior rim, unusually small, round punctates on the exterior rim, and smoothed-over cordmarking on the body. These elements are an amalgamation of Mackinac and Bois Blanc characteristics, suggesting it was manufactured in the early Late Woodland to middle Late Woodland period (see Table 5.3; Figure G41). 88 Middle Late Woodland Ceramic Subassemblage The initial assessment of the Cloudman site occupational history excluded a middle Late Woodland component (Branstner 1995). Re-evaluation of some of the originally untyped vessels identified a small number of vessels (n=5) with strong affinities to Bois Blanc ware, which are associated with the middle Late Woodland period at the Juntunen site (AD 1000-1200; McPherron 1967). Bois Blanc Ware (n=5) Five vessels are categorized as or show strong affinities to Bois Blanc ware (see Table 5.3; Figure G42-G43). Bois Blanc is characterized by thickened rims (by folded-over lips or by the addition of a strip/fillet), castellations, and cord-wrapped object decoration (McPherron 1976:104). These vessels described below all possess the characteristic thickened rims that are the hallmark of Bois Blanc but lack cord-wrapped object impressions or apparent castellations (McPherron 1967). However, the rim morphologies are clearly Bois Blanc despite the non-traditional decorative elements. Vessel 42 has a slightly inverted rim with a square lip, two horizontal bands of cord impressions below the lip, a fillet with oblique cord impressions, and at least one more horizontal band of cord impression below the fillet. Vessel 158 has a slightly inverted braced rim with a lip displaying widely spaced vertical linear tool impressions over a smoothed exterior rim. Vessel 215 has an inverted, wedge-shaped rim with two faint horizontal bands of cord impressions on both the interior and exterior rim surfaces, and possible vertical cord-wrapped stick impressions just below the interior lip. Vessel 73 has the characteristic fillet/braced rim of Bois Blanc, the surface of which appears fabric impressed or impressed with a paddle wrapped loosely with cord, a surface treatment occasionally seen on Bois Blanc vessels (McPherron 1967:106). A fifth 89 vessel, Vessel 31, strongly resembles Bois Blanc ware, displaying a braced rim, but the paste and manufacture quality are poor, possibly representing an expedient vessel. Late Late Woodland Ceramic Subassemblage The greatest stylistic variation occurs among the late Late Woodland ceramic subassemblage (n=49). While the majority of the subassemblage is comprised of Juntunen ware, it also includes Traverse ware and a number of vessels of such stylistic singularity that they remain untyped (Table 5.5; Figures G44-G55). Juntunen Ware (n=25) Branstner (1995) initially categorized only three vessels (101, 102, 143) as Juntunen ware, but a re-assessment demonstrates that many of the late Late Woodland vessels that were left untyped displayed some characteristics of Juntunen ware. Furthermore, a number of vessels from the 1995 excavation, which were not included in Branstner’s report, were clearly Juntunen vessels. Juntunen ware is characterized by the nearly universal presence of true collars and decoration of linear punctates or drag-and-jab techniques, with little to no use of cord impressions or incising (McPherron 1967:111). A “proto-Juntunen vessel” (Vessel 24), likely manufactured ca. AD 1200, has “peaking” on the rim (an incipient castellation common among Bois Blanc) with a vertical rim shape, but has a tall, thin collar and a smoothed exterior surface, which is common later among Juntunen wares. Juntunen Linear Punctate vessels (204, 205, 206, 209, 211, 212) are characterized by smoothed surfaces, true collars, and decoration by closely spaced but separate punctates (McPherron 1967:111). Juntunen Drag-and-Jab vessels (44, 45, 89, 102, 143) are distinguished by vertical rims, collars, castellations, and push-pull decorative motifs (McPherron 1967:113). 90 Another twelve (12) vessels (25, 26, 47, 60, 86, 101, 115, 133, 210, 213, 214, 218) could be classified as Juntunen ware but were not complete enough for further categorization. Table 5.5: Late Late Woodland Vessels by Type Type Juntunen Linear Punctate Juntunen Drag-and-Jab Juntunen Ware (cf. Jab-and-Drag) Juntunen Ware (cf. Linear Punctate) Juntunen Ware (cf. Plain) Juntunen Ware Proto-Juntunen Ware Late Juntunen Ware (cf. O'Neil Curvilinear) Traverse Decorated v. Punctate Traverse Plain v. Scalloped Traverse Ware Untyped (cf. Juntunen Ware) Untyped (cf. O'Neil cup) Untyped Total Ct. 6 5 1 2 1 8 1 1 3 5 2 1 1 12 49 One “Late Juntunen” vessel bears affinity to pottery from the O’Neil site in northern lower Michigan. Vessel 152 is a late variety of Juntunen ware (post-1400) resembling O’Neil Curvilinear, a style late in the Juntunen sequence characterized by curved cord impressions superior to punctate decoration (Lovis 1973). Traverse Ware (n=10) The greatest disparity between the original analysis by Branstner (1995) and the present study was in the attribution of Late Woodland Algonquian pottery not belonging to the Juntunen sequence. Branstner originally believed ten (10) vessels resembled Algoma ware, a type established by Thor Conway in Ontario but never fully described in the 91 literature. His most detailed description of the type likens it to “Dumaw Creek ware,” which, along with Algoma ware, display “scalloped lips and various modes of decoration” (Conway 1977:21). The vagueness of this type description made it an inadequate category for type attribution. Algonquian wares with scalloped lips have been more recently categorized and described in detail by Hambacher (1992) as Traverse ware. This type is based on the assemblage from the Skegemog Point site in northern lower Michigan and is generally contemporaneous with Juntunen ware, dating to AD 1100-1550/1600 (Hambacher 1992). Therefore, all of the formerly categorized Algoma vessels have been recategorized as Traverse ware except for Vessel 177, which was recategorized as untyped. Five (5) vessels (69, 75, 104, 150, 190) are categorized as Traverse Plain variety Scalloped. They are characterized by a scalloped lip, straight to everted rim profiles, and smoothed exterior surfaces (Hambacher 1992:176). Three (3) vessels are classified as Traverse Decorated variety Punctate (Vessel 43, 56, 67). They are characterized by smooth exteriors, flat lip forms and profiles, smooth lip surfaces, and slightly everted collared rims with simple punctate exterior decoration (Hambacher 1992:198-199). Vessels 149 and 165 were classified as Traverse ware because of their scalloped lips but were not complete enough for more specific typological attribution. Untyped (n=14) Despite the review of existing literature, fourteen vessels either remained outside of existing typological categories or did not have enough vessel material present to be properly categorized. These vessels displayed sufficient paste and morphological characteristics to conclude that they were manufactured during the late Late Woodland period. Of particular note among these is Vessel 54, a miniature vessel that closely resembles a cup from at the O’Neil site. As with Vessel 152 above, the cup V54 resembles is associated with late precontact/early 92 contact period trade items (Lovis 1973). This cup is described below in the section detailing miniature vessels in the assemblage. General Late Woodland Ceramic Vessels Four pottery vessels (65, 85, 180, 184) show hallmarks of manufacture during the Late Woodland period, such as everted rims, cord-wrapped stick impressions, smoothing, square and flattened lips, and, in one case, brushing, but they lack visible characteristics associating them with early, middle, or late Late Woodland groups (see Table 5.3). Ontario Iroquoian Ceramic Subassemblage Branstner (1995) originally identified eighteen (18) Huron pottery vessels, four (4) of which she believed were of Algonquian/Odawa manufacture, attempts to mimic Huron forms. An additional two (2) vessels were believed to resemble Lalonde High Collar. Reassessment of these vessels confirmed their identification as Ontario Iroquoian or Iroquoian-like (n=37; Table 5.6; Figures G56-G65). Additionally, a number of vessels previously left untyped or thought to be late Juntunen wares were found to more closely resemble Ontario Iroquoian wares. The nature of the presence of such vessels at the Cloudman site is contentious, and therefore definitive categorization of vessels into established typologies was avoided. Early Ontario Iroquoian (n=2) Vessels 145 and 146 were originally categorized as possible late Juntunen varieties, but upon re-examination were found to be decorated with incising rather than the drag-and-jab method associated with Juntunen wares. Vessel 146 has opposed incising on a collar-less, everted rim, with a row of punctates on the rim/shoulder margin and incised decoration on the lip. The shoulder of the vessel is cordmarked. Vessel 145 is identified from a 93 smaller sherd lacking the shoulder and shoulder margin, but it is otherwise identical to V146 besides the absence of lip incising. The combination of opposed incising and cordmarking on the body suggests these vessels fall in the range of Early Ontario Iroquoian wares, likely manufactured between AD 1200 and 1300 (William Fox, personal communication). Table 5.6: Ontario Iroquoian Vessels by Type Type Early Ontario Iroquoian ware cf. Huron Incised cf. Ripley Plain cf. Lawson Opposed cf. Lawson Opposed or Methodist Point Ware cf. Lawson Incised or Huron Incised cf. Lawson Incised or Sidey Notched Untyped Total Ct. 2 22 3 1 2 1 1 5 37 cf. Huron Incised (n=22) The majority of the Iroquoian vessels can be categorized as closely resembling Huron Incised (Vessels 36, 62, 70, 74, 79, 82, 155, 156, 159, 160, 161, 163, 166, 167, 178, 179, 182, 183, 185, 186, 187). Huron Incised vessels are characterized by oblique or vertical incised line decoration on short, outflaring collars with straight or convex interior rim surfaces (MacNeish 1952:34). This type is generally associated with Wendat-Petun groups. Vessels 167 and 182 are miniature vessels detailed later in the chapter. cf. Ripley Plain (n=3) Vessels 64, 135, and 151, previously untyped, closely resemble Ripley Plain. Ripley Plain is notable for its smooth surfaces and lack of decoration, and it is generally associated with the Wendat,-Petun and the Neutral of southern Ontario (MacNeish 1952). 94 cf. Lawson Ware (n=5) Several vessels bear close resemblance to Lawson wares, specifically Lawson Incised (V136, V162) and Lawson Opposed (V154, V40/153, V157). Vessel 162, originally categorized as Huron Incised, displays a concave rim interior and incising on the lip that are not typical of that type. Incised collars with concave rim interiors are more common of Lawson Incised (Birch and Williamson 2012; MacNeish 1952), while the castellation, closely- spaced oblique incised lines, well-defined collar, and incised lip are more characteristic of Sidey Notched (MacNeish 1952). Residue from V162 was AMS dated to cal AD 1420-1446. Sidey Notched pottery, while usually associated with Wendat-Petun groups living in closer proximity to the Cloudman site, generally dates to AD 1550-1650 (William Fox, personal communication), while Lawson Incised, a type associated with the Neutral, have been associated with materials dating to AD 1400-1500 in New York (Fink 2013:71). Vessel 136 displays broadly-spaced, vertical incising, resembling either Lawson Incised and Huron Incised; however, it lacks both the distinct channeled or concave interior of Lawson (instead possessing a crudely interior-extruded lip) and the defined collar of Huron Incised. Vessels 40/153 and 157 lack intact lips required for accurate typological classification, but both are collared and incised. Possibly deriving from the same vessel, both are decorated with opposed oblique incising on the collar, two horizontal incised lines along the bottom edge of the collar, and a row of small punctates just below the collar. Without the horizontal incising, these could be categorized as Lawson Opposed (Macneish 1952), but with these lines they are a closer visual match with vessels from the historic Huron/Wendat site of Methodist Point (Kenyon 1970). Vessel 154 also strongly resembles Lawson Opposed, with a thin collar and opposed oblique incising on the collar underscored by punctates. However, the lip is not present, preventing identification of the diagnostic channeled/concave rim. 95 Untyped (n=5) Several vessels (38, 48, 68, 77, 188) had characteristics that associated them with Iroquoian manufacture, such as incising, collars, smoothed exterior surfaces, and squared lips, but lacked enough definitive characteristics to be further categorized. Unidentified Affiliation Two vessels could not be confidently attributed to a specific time period or cultural affiliation. Vessel 195 consists of a single rim sherd too small to display characteristics associating it with a period of manufacture. Vessel 148, however, is a unique vessel that is likely quite late in the occupational sequence of the Cloudman site, but remains a stylistic anomaly. It is smoothed and peaked with a row of punctates with at least three rows of cord impression below on the rim and a very fine cording or fabric impression on a squared lip. The paste is very hard and contains a low density of temper. Its style fits neither fully with Juntunen nor Iroquoian and may represent some amalgamation of both Algonquian and Iroquoian decorative traditions. Miniature Vessels Branstner (1995) identified miniature vessels in the Cloudman assemblage but excluded them from the original assemblage descriptions. Miniature vessels are often classified as practice pots made by children, but several of the Cloudman miniature vessels show craftsmanship and decorative organization equal in quality to their full-sized counterparts. A total of 11 miniature vessels were identified in the Cloudman assemblage (Table 5.7; Figures G69-G79). Although included in the type counts and summaries above, a more detailed account of these vessels was warranted. Vessels were categorized as “miniature” if their orifice radius was less than 10 cm. 96 Middle Woodland (n=2) Vessels 39 and 201 display hallmarks of manufacture during the Middle Woodland period. Vessel 39 is Laurel ware, with horizontal bands of oblique stamping and a narrow, squared lip with a stamped lip. It is not as finely made as other Laurel vessels, its interior surface crudely smoothed, but it contains a fair amount of grit temper, suggesting it may have been constructed for a purpose rather than as a training vessel for a child. Vessel 201 is untyped, made of fine paste with cross-hatched incising on a thickened exterior rim with a rounded lip. It displays evidence of coil manufacture, placing it solidly in the Middle Woodland. It is a very small, shallow vessel, and may be a local imitation of a Hopewellian Middle Woodland vessel more common to the south in Michigan. Table 5.7: Miniature Vessels by Type Socio-Temporal Association Type Middle Woodland Early Late Woodland Late Late Woodland Iroquoian Laurel Ware Untyped, Hopewellian Subtotal Mackinac Punctate Mackinac Undecorated Mackinac Ware Untyped Subtotal Traverse Plain v. Scalloped Untyped (cf. O'Neil cup) Subtotal cf. Huron Incised Subtotal TOTAL 97 Ct. 1 1 2 1 1 2 1 5 1 1 2 2 2 11 Early Late Woodland (n=5) Almost half of the miniature vessel assemblage is associated with the early Late Woodland period. Two vessels could be attributed to specific types: Vessel 53 is Mackinac Punctate; Vessel 52 is Mackinac Undecorated. Both have the hallmark everted rims and exterior cordmarking of Mackinac ware, and both are thin-walled and very finely made. Vessel 53 also has faint staining along the top of the interior rim, suggesting it may have been used for cooking. An additional two vessels, 83 and 202, can be generally categorized as Mackinac ware. Vessel 83 is thin-walled with an everted rim, two rows of punctates on the exterior and one row on the interior. Although thin, it is rather crude and may represent a child’s pot. Vessel 202 is very fine and thin, with what appears to be fabric impressions on the exterior surface and along the exterior lip. The rim is everted, and fine, round punctates on the interior rim that were pushed through to create exterior bosses. The delicate nature of the decoration suggests it was made by an experienced potter. Vessel 63 is untyped but associated with the early Late Woodland period because of its splayed lip with oblique cord-wrapped stick impressions and punctates on the exterior rim. However, it appears rather crudely made and lacks much visible temper, and therefore may be a child’s pot. Late Late Woodland (n=2) Miniature vessels manufactured in the late Late Woodland period include Vessel 54, which is 8 cm in diameter and 8 cm tall with straight sides and a rounded bottom. There is no exterior surface treatment or decoration, but there are small punctates pressed downwards into the anterior surface of the lip. This vessel closely resembles a cup found at the O’Neil site in context with late prehistoric/early protohistoric artifacts (Lovis 1973). Vessel 75 is Traverse Plain variety Scalloped Lip. It has a smoothed, everted rim with cordmarking beginning at the neck, creating the neck/body zoning typical of Traverse wares 98 (Hambacher 1992). Vessel 75 also has a ring of thick interior carbonization around the rim and was therefore used for cooking. Its manufacture and use suggest construction by a skilled potter. Ontario Iroquoian (n=2) Two miniature vessels were categorized as Iroquoian. Vessel 167 has a very short (<1cm), vertical collared rim with oblique incising. It is thick, crude, and heavily tempered. Vessel 182 is collared with faint vertical incising and has no visible temper. Both most closely resemble Huron Incised and may represent pots made by children. Pottery Age The diachronic nature of the research questions necessitates both absolute and relative dating of pottery from the Cloudman site. AMS dating of carbonized food residue adhered to pottery permitted direct association of foods cooked in vessels to specific dates, while relative dating of vessels based on style creates a more robust chronology of site occupation. Drawing on both types of data, a general history of the Cloudman site is constructed and presented. AMS Dates Carbonized residue was collected from five (5) vessels from the Cloudman site and submitted for AMS dating (Table 5.8). Two Middle Woodland vessels, a Laurel Dentate (V4) and a Laurel Pseudo-scallop vessel (V1), produced similar age ranges, but the radiocarbon dates are significantly different at 95% confidence (Stuiver et al. 2018) and demonstrate that there were at least two distinct Middle Woodland occupations: the first, represented by a Laurel Dentate Stamped vessel, occurred ca. cal AD 87; the second, represented by a Laurel Pseudo- scallop Stamped vessel, took place around cal AD 127. The Cloudman site may have been used as a short-term resource extraction camp to which small groups returned repeatedly throughout 99 the early decades of the Middle Woodland period. The radiocarbon results also indicate that Laurel Dentate Stamped pottery precedes Pseudo-scallop Stamped pottery in the local taxonomic chronology. Radiocarbon age ranges for a Mackinac Banded vessel (V103) and a Blackduck vessel (V193) closely overlapped, proving statistically significant at 95% confidence with a mean pooled age of 1092 ± 11 cal BP, placing the primary early Late Woodland occupation ca. cal AD 957 (Stuiver et al. 2018). The date for the Blackduck vessel is particularly important since this type is found from Saskatchewan to Ontario, and from Minnesota to Michigan, and has been variably dated to ca. AD 900-AD 1500 across this geographic expanse (McPherron 1967:97). The date obtained from Vessel 193 pinpoints the timing of the use of Blackduck wares in the eastern Upper Peninsula and supports McPherron’s (1967:100) hypothesis that Blackduck was present at the nearby Juntunen site between AD 800-1100. Table 5.8: AMS Dates from Carbonized Pottery Residue Samples Sample No. UCIAMS# Vessel No. Type Radiocarbon Age (14C BP) Calibrated Age (2σ)* Median Probability Cloudman 1 187416 4 Cloudman 4 187417 103 Cloudman 6 190514 162 Cloudman 7 190515 1 Cloudman 8 190516 193 *Stuiver et al. 2018 Laurel Dentate Stamped Mackinac Banded cf. Sidey Notched/ Lawson Incised Laurel Pseudo- Scallop Stamp Blackduck Banded 100 1915±15 1085±15 AD 59-126 (1.0) AD 941-997 (0.688) cal AD 87 cal AD 968 475±15 AD 1421- 1445 (1.0) cal AD 1433 1870±15 AD 80-180 (0.895) cal AD 127 1100±15 AD 938 - 987 (0.577) cal AD 946 Finally, V162, which is an Iroquoian vessel resembling either Lawson Incised or Sidey Notched, dated to cal AD 1433. This pre-dates the normal time span for Sidey Notched vessels, which were commonly produced between AD 1550-1650 (William Fox, personal communication). Lawson wares were produced somewhat earlier, and Lawson Incised has been associated with materials dating to AD 1400-1500 in New York (Fink 2013:71). Ontario Iroquoian vessels (including those associated with the Wendat, Petun, and Neutral) consistently appear west of Ontario after AD 1400, overlapping with the Late Juntunen subphase (AD 1400 – AD 1600) (McHale Milner 1998:214; see also McPherron 1967). Relative Dating Relative dating based on stylistic relationships corroborates and enriches the Cloudman site occupational timeline. There is little evidence to support a human occupation of the site prior to the Laurel Dentate Stamped vessel AMS median age of cal AD 87. The assemblage does contain two North Bay vessels, and residues from a North Bay Plain vessel at the Winter site on the Garden Peninsula produced an AMS date with a median probability of cal BC 113, significantly earlier than the Cloudman Laurel dates. However, North Bay Cordmarked and North Bay Punctate vessels from the Winter site produced median probability ages of cal AD 82 and cal AD 155, respectively (Albert et al. 2018; Lovis et al. 2012; Richner 1973; Stuiver et al. 2017), which are consistent with the Cloudman Laurel occupations. Three late Laurel ware vessels at the site, along with four other vessels that appear transitional between Middle and Late Woodland, suggest an additional, if brief, occupation sometime between cal AD 500-700 (Lovis 2014). It is unclear whether the site was occupied between AD 200 and AD 500. 101 A second intensive occupation of the site took place in the Early Late Woodland period. The ceramic assemblage includes both Mackinac Punctate and Mackinac Banded vessels. Mackinac Punctate was generally manufactured between AD 700-900 in the Upper Great Lakes, while Mackinac Banded is commonly dated to AD 850-1000 (Lovis 2014). Dates from the Mackinac Banded and Blackduck Banded tightly overlap and suggest a substantial occupation between AD 900 and AD 1000 (Lovis 2014), although the presence of Mackinac Punctate vessels at the site could represent an additional and somewhat earlier occupation. Only five small Bois Blanc rim sherds represent the middle Late Woodland, and although there was no intensive occupation of the Cloudman site between AD 1000-1200, the site was still utilized during this period, if but briefly. The Late Precontact period (AD 1200 – 1600), comprises the most complicated span of the Cloudman site occupation, and includes both late Late Woodland and Ontario Iroquoian pottery. A “proto-Juntunen” vessel from ca. AD 1200 is the earliest of these vessels. Juntunen Linear Punctate vessels (n=6) were commonly manufactured between AD 1200-1300, while Juntunen Drag-n-Jab (n=5) generally dates to AD 1300-1400 (Lovis 2014). Although the late Late Woodland pottery included in this analysis are most closely associated with the time period between AD 1200 and AD 1400, these varieties were also manufactured in small quantities afterwards (Lovis 2014; McHale Milner 1998; McPherron 1967). Two late Juntunen vessels closely resembling vessels from the Late Precontact/Early Contact component of the O’Neil site, suggest a late, post-AD 1500 presence of Algonquian peoples and ceramic vessels on the site. Traverse ware (n=10) is largely contemporaneous with Juntunen wares, dating to AD 1100 – AD 1550/1600 (Hambacher 1992). The Cloudman site was frequently and repeatedly occupied by 102 local groups from AD 1200 until the arrival of Europeans, although the intensity of Algonquian occupation post-AD 1400 appears limited. Further complicating the Late Precontact history of the Cloudman site, Ontario Iroquoian vessels also begin to appear at the site post-AD 1200. Two Early Ontario Iroquoian vessels represent the earliest of these “foreign” wares, which were likely manufactured between AD 1200-1300. Vessel 162, which resembles either Lawson Incised or Sidey Notched, was AMS dated to cal AD 1420-1445. Four other vessels resembled Lawson ware, which has been associated with materials dated to AD 1400-1500 in New York (Fink 2013:71), although the timing of their presence west of Ontario is unclear. Huron Incised, to which a substantial number of Cloudman vessels bear resemblance (n=22), normally date between AD 1500-1650 (William Fox, personal communication). Overall, the presence of Ontario Iroquoian vessels at the site generally post-dates AD 1400. Although most of the Juntunen vessels at the site likely pre-date most of the Iroquoian vessels at the site, there is temporal overlap between the two assemblages. The two Early Ontario Iroquoian vessels are contemporaneous with Juntunen Linear Punctate, while at least two vessels represent late Juntunen wares manufactured post-AD 1400. The distinct styles and cultural affiliations of Juntunen and Iroquoian wares, and their general temporal separation validates their treatment as distinct analytic subsets. In summary, primary occupations of the Cloudman site, represented by the largest proportions of the overall pottery assemblage include the earlier part of the Middle Woodland period (AD 0-200), the early Late Woodland period (AD 800-1000), the late Late Woodland period (AD 1200-1400), and Ontario Iroquoian occupation (AD 1400-1600). Again, it is important to note that it is still unclear whether the post-AD 1400 occupants of the site would 103 have identified as Iroquoian or whether they were Algonquians (i.e., proto-Odawa) using Iroquoian pottery. Smaller, secondary occupations of the site in interstitial periods and are indicated in Table 5.9. Pottery from the primary occupations will be used as the primary data subsets in the analyses of subsequent chapters because the secondary subassemblages would yield very small sample sizes. Table 5.9: Occupational History of the Cloudman Site, Derived from Relative and Direct Dating of Pottery Century 100-0 BC AD 0-100 AD 100-200 AD 200-300 AD 300-400 AD 400-500 AD 500-600 AD 600-700 AD 700-800 AD 800-900 AD 900-1000 AD 1000-1100 AD 1100-1200 AD 1200-1300 AD 1300-1400 AD 1400-1500 AD 1500-1600 Woodland Iroquoian ? X X* ? ? ? x x x X X* x x X X x x - - - - - - - - - - - - - x ? x* X X= primary occupation x= secondary occupation ?= possible but unknown period of occupation *associated AMS date 104 Conclusion The re-evaluation of taxonomic categorization of the Cloudman pottery assemblage was largely in accordance with the original analysis conducted by Branstner (1995). The primary discrepancy between the former and present studies was in the identification and timing of the later site occupations. Ontario Iroquoian pottery was originally associated with an early contact- period occupation of the site (ca. AD 1630) by Odawa traders, but both relative and AMS dating indicate many of these vessels were manufactured and used at the site prior to this time, mostly between AD 1400 and AD 1600, but also as early as AD 1200. The uncertain identity of the manufacturers and users of these wares and their possible overlap with late Late Woodland Juntunen pottery requires further investigation to clarify the late occupational history of the Cloudman site. For the purposes of this study, the late Late Woodland and Ontario Iroquoian subassamblages will be treated as separate analytic subsets. Contextualization of the Cloudman vessels within established chronologies for regional ceramic taxonomies permitted construction of a solid occupational history of the site, supported and detailed by a small set of direct AMS dates. Pottery from the most intensive occupations (Middle Woodland, early Late Woodland, late Late Woodland, Ontario Iroquoian) identified through the work presented in this chapter will serve as the primary analytic subsets for explorations of pottery function (Chapter 6) and cuisine (Chapter 7) at the Cloudman site. 105 CHAPTER 6 POTTERY FUNCTION Introduction The establishment of a site chronology through traditional ceramic taxonomic classification of the Cloudman assemblage permitted the implementation of a second approach to pottery analysis emphasizing vessel function. Although occasionally constructed for symbolic or ceremonial use, most ceramic vessels were made for the utilitarian functions of cooking, serving, and storage. Pottery function can be evaluated through technical properties, a vessel’s characteristics chosen by the potter to suit its intended purpose, and use-alteration traces, which reveal how a vessel was used. Properties related to function can be sensitive to changes in food- processing requirements and cooking methods, reflecting diachronic alterations in pottery use and dietary habits when set against the site chronology established in the previous chapter. Synchronic variation of vessel function can signal communication of group identity through food selection and cooking styles in addition to pottery style. This chapter will evaluate both technical and use-alteration properties of the Cloudman site pottery vessels, followed by discussions of intended and actual ceramic function. Function will be assessed synchronically to explore the relationship between pottery use and identity, and diachronically to answer research questions concerning pottery technology and use in relation to culinary transformations. 106 Technical Properties and Intended Function The three technical properties chosen as the focus of this study were temper size, rim diameter, and vessel thickness. Temper size and vessel thickness have proven sensitive measures of technical variability in relation to foodways transformations (Braun 1983; Hart 2012). Rim diameter has demonstrated variability in the northern Great Lakes and beyond (Blitz 1993; Kooiman 2012, 2016; Potter 2000), although the significance of vessel size in relation to function is complex and unclear. Examination of vessel size in relation to other technical properties and cooking requirements could clarify its functional role among Woodland vessels; it therefore serves an exploratory purpose in this study The temporal overlap in construction and use of late Late Woodland and Ontario Iroquoian vessels (as discussed in Chapter 5) creates an analytical issue. These subassemblages clearly represent two distinct pottery traditions and social/cultural groups, but it is unclear if the Ontario Iroquoian pottery represents the physical presence of Iroquoian people at the Cloudman or if they were Iroquoian trade items used by Woodland/Algonquian groups. If the latter, then there may be little if any difference in the way vessels from both taxonomic categories were used. Temper size Temper size influences the performance of a pottery vessel. Both vessel strength and thermal shock resistance of fired vessels increased with the decrease in temper particle size (Bronitsky and Hamer 1986). Cooking vessels are expected to be constructed with smaller temper particles to increase their use-life over the fire. Temper size has been found to decrease over time in parts of the Eastern Woodlands in correlation with increased processing of starchy 107 foods (Braun 1983), including maize (Hart 2012), because these foods require high-temperature, long-term boiling to become palatable (Wandsnider 1997). However, a prior study in the northern Great Lakes found no significant difference in temper size between Middle Woodland from the Naomikong Point site and the Late Woodland Sand Point site (Kooiman 2016). These sites are geographically separated by considerable distance, and the Cloudman assemblage provides an opportunity to explore the potential for diachronic temper size change at a single site. Table 6.1 shows mean temper size among the primary analytic subsets. Average temper size shows a small initial increase in early Late Woodland, followed by a decrease in late Late Woodland and Iroquoian vessels. A Welch’s unpaired T-test, a non-parametric analysis used to compare independent samples of unequal variances and sample sizes, was used to determine the statistical relationships between average temper size of vessels (Table 6.2). The slight increase in mean temper size between Middle and early Late Woodland vessels proved insignificant. Despite the 500-year time gap between the primary Middle Woodland occupation and the early Late Woodland occupation at the Cloudman site, temper size remained relatively consistent. However, in a comparatively short span of time between the early Late Woodland and the Late Late Woodland occupations, temper underwent a significant size reduction. Early Late Woodland vessel temper is significantly larger than late Late Woodland and Iroquoian temper, and Middle Woodland temper is similarly significantly larger than Iroquoian temper. This abrupt change in temper size after a millennia of consistency signals a major shift in temper selection in the middle of the Late Woodland period. Following this shift, temper size remains consistently small after AD 1200. 108 Table 6.1: Mean Temper Size by Subset Subset Middle Woodland Early Late Woodland Late Late Woodland Ontario Iroquois Mean (mm) 1.49 1.58 1.34 1.26 n 33 58 48 35 Table 6.2: Temper Size Relationships (Welch’s Unpaired T-Test) Subset Middle Woodland Middle Woodland Early Late Woodland Late Late Woodland Ontario Iroquois p* --- 0.2842 0.0755 0.0045 df --- 84 66 62 Early Late Woodland p --- df --- Late Late Woodland p --- --- 0.0030 0.0001 --- 103 93 --- --- 0.1907 df --- --- --- 76 *two-tailed α=0.05 Rim Diameter A total of 88 vessels, 43.6% of the total vessel assemblage, had rim segments sufficient in size for diameter measurements. A plot of rim diameters shows a roughly normal distribution (Figure 6.1) with an overall mean of 19.27 cm. The average rim diameter for Middle Woodland vessels was 17cm, smaller than the averages for early Late Woodland, late Late Woodland, and Iroquoian vessels (Table 6.3). A Welch’s t-test confirmed that Middle Woodland rim diameters were significantly smaller than early Late Woodland rim diameters (Table 6.4). Meanwhile, early Late Woodland, late Late Woodland, and Iroquoian assemblages were almost statistically identical. The results demonstrate a significant increase in orifice diameter during the Middle Woodland-Late Woodland transition. 109 Figure 6.1: Rim Diameter Frequencies of the Cloudman Pottery Assemblage Table 6.3: Mean Rim Diameter by Subset Subset Mean (cm) Middle Woodland Early Late Woodland Late Late Woodland Ontario Iroquois 17 20 19 20 n 16 41 17 14 110 Table 6.4: Rim Diameter Relationships (Welch’s Unpaired T-Test) Subset Middle Woodland p* --- 0.0448 0.1651 0.0822 df --- 30 23 22 Middle Woodland Early Late Woodland Late Late Woodland Ontario Iroquois *two-tailed α=0.05 Early Late Woodland p --- --- 0.8396 0.9439 df --- --- 21 20 Late Late Woodland p --- --- --- 0.8226 df --- --- --- 25 Vessel orifice (rim) diameter is an aspect of pottery morphology that is not fully understood. Rim diameter is often used as a proxy for vessel volume/size, but only a handful of studies in North America have demonstrated the relationship between the two measures (Blitz 1993; Parker and Kennedy 2010; Potter 2000; Shapiro 1984). Statistical relationship between rim diameter and vessel volume in Late Woodland vessels from the Upper Great Lakes have been previously demonstrated (Kooiman 2015a), but a similar test among Middle Woodland and Iroquoian vessels has yet to be carried out. Although overall morphologies of these three categories of ceramic vessel differ, the disparities are not drastic enough to preclude using rim diameter as a proxy for size/volume for Middle Woodland and Iroquoian vessels. Still, the significance of vessel volume/size is largely unknown and likely variable through space and time. Size may be related to utilitarian function (Kobayashi 1994; Rice 1987:299), social function (Blitz 1993; Potter 2000), and/or household size (Nelson 1981; Tani 1994; Turner and Lofgren 1966). The pottery assemblage from the Laurel Middle Woodland site of Naomikong Point was statistically significantly smaller than Late Woodland vessels from the site of Sand Point; however, at Sand Point, Late Woodland vessels from habitation contexts were significantly smaller than vessels from mound contexts at the site and statistically equivalent to 111 the Middle Woodland vessels from Naomikong Point (Kooiman 2016). Ethnoarchaeological studies have also shown that there is a clear size distinction between daily cooking pots and ceremonial cooking pots (Kobayashi 1994). Therefore, context of use (utilitarian vs. ceremonial) may be one reason for size variation among pottery vessels in the northern Great Lakes. A burial mound was once present at the Cloudman site, but it was excavated in the early 20th century and the whereabouts of associated artifacts are unknown (Branstner 1995). All of the measured pottery vessels derived from the primary habitation zone, and the function of the site is undifferentiated between the components. Therefore, the diachronic size variation at Cloudman is probably not associated with context of use. The increase in vessel size over time could be linked to several factors: new cooking techniques and/or foods may have been better accommodated by larger vessels; increased group size my have necessitated larger vessels to accommodate feeding more people; or decreased mobility may have allowed for greater energy investment in manufacturing larger vessels that no longer needed to be carried from site to site. At present, the reason for diachronic vessel orifice/vessel size variation remains unclear. Vessel Thickness Pottery wall thickness was measured on the lip, neck, shoulder, and body (Table 6.5). Lip thickness increases, as expected, from the narrow-lipped Middle Woodland vessels, to the splayed-lip early Late Woodland vessels, and then decreases slightly with the collared, squared- lipped late Late Woodland and Iroquoian vessels. Neck thickness steadily increases through time, likely correlated with the proliferation of collared wares post-AD 1200. Shoulder thickness varies greatly through time, initially decreasing from the Middle Woodland to the early Late Woodland, then increasing in the Late Late Woodland, and decreasing among Iroquoian vessels, 112 which had the thinnest average shoulders overall. Body thickness decreases from the Middle Woodland to the Late Woodland, then increases again among Iroquoian vessels. These trends run counter to predictions that vessel wall thickness would steadily decrease over time. Sample sizes for both shoulder and body thickness were small, however, with few vessels complete enough to include these portions for measurement, affecting the overall accuracy of thickness comparisons. Statistical analysis was used to determine the significance of observed thickness variations. Lip thickness was considered less sensitive to change based on cooking needs, so it was excluded from statistical scrutiny. Miniature vessels were excluded from thickness calculations because of their disproportionate distribution through time and the potential for their inclusion to skew the data. A Welch’s T-test was employed to examine the relationships between analytic subsets. The most significant differences manifested in neck thickness (Table 6.6). Middle Woodland and early Late Woodland vessel necks have statistically identical thicknesses, while vessels from both subsets were significantly thinner than Late Late Woodland and Iroquoian vessels. Late Late Woodland and Iroquoian vessel neck thickness was also significantly different, despite their similar mean thicknesses. Shoulder thickness and body thickness revealed fewer significant differences between analytic subsets (Tables 6.7 and 6.8) Hart (2012) recorded vessel thickness through time among Iroquoian vessels in New York. He notes that vessel thickness is often affected by vessel size and controlled for this variation by dividing thickness by diameter. Following Hart’s procedures, neck thickness and averaged neck/shoulder thickness were corrected by rim diameter (Table 6.9), which reduced sample sizes by restricting inclusion to only vessels with rim diameter measurements. In both 113 cases, thickness decreased between the Middle Woodland and early Late Woodland, then increased moving into the late Late Woodland, and decreased again among the Iroquoian vessels. Table 6.5: Vessel Wall Thickness by Subset Lip Neck Subset Mean (cm) Middle Woodland 4.43 Early Late Woodland Late Late Woodland 8.51 7.2 Ontario Iroquois 6.98 n 35 61 46 30 Mean (cm) 6.1 6.36 7.85 7.84 N 33 57 45 27 Shoulder Mean (cm) 6.79 6.43 8.02 5.94 N 9 25 19 7 Body Mean (cm) 7.4 6.11 6.7 9.12 n 6 9 7 4 Table 6.6: Neck Thickness Relationships (Welch’s Unpaired T-Test) Subset Middle Woodland Early Late Woodland Late Late Woodland Iroquoian *two-tailed α=0.05 Middle Woodland Early Late Woodland Late Late Woodland p* --- df --- 0.2527 84 P --- --- 0.0001 0.0001 68 97 0.0004 0.0001 df --- --- 72 53 p --- --- --- 0.0344 df --- --- --- 68 114 Table 6.7: Shoulder Thickness Relationships (Welch’s Unpaired T-Test) Subset Middle Woodland Early Late Woodland Late Late Woodland p* --- df --- 0.6920 17 p --- --- df --- --- 0.0617 25 0.0150 27 p --- --- --- df --- --- --- Middle Woodland Early Late Woodland Late Late Woodland Ontario Iroquois 0.7246 8 0.6758 7 0.1104 9 *two-tailed α=0.05 Table 6.8: Body Thickness Relationships (Welch’s Unpaired T-Test) Subset Middle Woodland p* --- 0.4722 0.4803 0.2707 df --- 9 8 7 Middle Woodland Early Late Woodland Late Late Woodland Ontario Iroquois *two-tailed α=0.05 Early Late Woodland Late Late Woodland P --- --- 0.9068 0.0830 Df --- --- 7 5 p --- --- --- 0.1002 df --- --- --- 6 Table 6.9: Corrected Thickness (Thickness/Rim Diameter) Subset Neck + Shoulder Neck Middle Woodland Early Late Woodland Late Late Woodland Ontario Iroquois n 8 24 13 6 Mean 0.40 0.37 0.49 0.41 n 16 41 17 14 Mean 0.45 0.37 0.52 0.37 115 Table 6.10: Corrected Vessel Neck Thickness Relationships (Welch’s Unpaired T-test) Subset Middle Woodland Middle Woodland Early Late Woodland Late Late Woodland Ontario Iroquois p* --- 0.5589 0.2349 0.8987 df --- 23 30 27 Early Late Woodland p df Late Late Woodland p --- --- 0.0496 0.4497 --- --- 23 20 --- --- --- 0.2785 df --- --- --- 28 *two-tailed α=0.05 Table 6.11: Corrected Average Neck + Shoulder Thickness Relationships (Welch’s Unpaired T-Test) Subset Middle Woodland Early Late Woodland p df Late Late Woodland p p* --- df --- 0.5656 8 --- --- 0.5718 0.6427 11 11 0.0546 0.9786 --- --- 17 5 --- --- --- 0.2567 df --- --- --- 8 Middle Woodland Early Late Woodland Late Late Woodland Ontario Iroquois *two-tailed α=0.05 Corrected neck and neck/shoulder thicknesses were subjected to Welch’s t-tests to evaluate significant differences between analytic subsets (Tables 6.10 and 6.11). After correction, it was apparent that there was no longer a significant difference in neck thickness between Middle Woodland vessels and later vessels. The critical factor, therefore, is vessel 116 size—Middle Woodland vessels are significantly smaller than Late Woodland and Iroquoian vessels and therefore thinner in accordance with size. Overall, these results do not follow predicted trends of thinning over time. The only significant increase in vessel thickness occurs during the late Late Woodland period, when, according to predictions, vessels should have become thinner to accommodate increased processing of starchy foods. Because neck thickness yielded the largest samples sizes, this might be an outcome of the advent of collars. Collars become more common among Juntunen and Iroquoian pottery vessels. Although neck thickness was measured below collars, the need for structural support of the collar would manifest in thicker vessel necks. Neck thickness of Traverse ware vessels, which lack collars, were significantly thinner than contemporaneous Juntunen ware vessels (p=0.0005, df=25) but were not significantly different from early Late Woodland vessels (p=0.461, df=10). The lack of body sherds associated with identified vessels, which were the focus of this study, affected a true assessment of vessel thickness of the areas that would be most influenced by a potter’s actions to improve heating effectiveness. Temporal attribution of body sherds not associated with identified vessels is dubious at best, so this was not pursued. The level to which neck and shoulder thickness relate to overall vessel thickness should be a topic of future inquiry. Synchronic Technical Variation Technical properties within the primary analytic subsets were investigated to assess synchronic variability. Contemporaneous taxonomic types could represent separate social groups with distinct stylistic traditions. Functional variability could mirror stylistic variability, signifying different pottery construction traditions and perhaps even different functional 117 requirements if culinary habits vary according to group identity. Rim diameter, thickness (averaged from neck and shoulder thickness measurements), and temper (averaged from three measurements per vessel) of pottery vessels from each typological category were compared to investigate the question of synchronic intended function variability. Only types comprising large portions of their respective subassemblages were chosen for comparison; other types with small sample sizes are excluded from the discussion below. Three major distinct pottery types were present in the Middle Woodland subassamblage (Table 6.12). Variation among all three technical properties among the types was low. This is unsurprising given that all three types are variations of Laurel ware. Relationships of these properties between types were assessed using Welch’s T-test, which revealed no statistical differences. Pottery construction techniques appear consistent across all Middle Woodland taxonomic types. Variation in technical properties slightly increases between types in the early Late Woodland subset. AMS dates from a Mackinac Banded and a Blackduck Banded vessel revealed extremely close contemporaneity (see Table 5.8), although Mackinac is a local ware and Blackduck has a broader distribution and is more common to regions north of the Upper Peninsula. Most apparent is the difference in mean temper size, with Blackduck temper smaller on average than Mackinac Banded. However, this difference was not statistically significant. Mackinac Punctate ware, which generally pre-dates but also temporally overlaps with Mackinac Banded, had a smaller mean temper size. However, comparisons of all physical properties between various early Late Woodland types demonstrated no statistically significant differences. Juntunen ware was the most common among the late Late Woodland subset, although vessels allowing for more specific categorization were limited in number. Therefore, all Juntunen 118 vessels (including Drag-and-Jab and Linear Punctate) were grouped together in a single analytical set, and compared against Traverse ware vessels, which also comprised an analytical set (see Table 6.12). These two wares are generally contemporaneous but differ greatly in form and style. Although rim diameter and temper size of Juntunen and Traverse vessels are not statistically different, there is a significant difference in thickness (p=0.0004, df=25). As discussed above, thickened necks and shoulders may have been required to support the thick, collared rims that are the hallmark of Juntunen ware. Traverse vessels, on the other hand, lack collars and would therefore not require thickened upper body walls. Finally, Iroquoian vessels were grouped by vessels resembling Huron Incised and those resembling various Lawson wares (see Table 6.12). These were the largest groupings within the subset, and represent wares associated with disparate Iroquoian groups (Wendat-Petun vs. Neutral). Rim diameter could not be compared because the Lawson group had a sample size of one. Vessel thickness and temper size were statistically identical. Overall, synchronic variation of technical properties within all time periods was low. Middle Woodland vessels are, as predicted, particularly consistent, but technical variability also remained low among later pottery subassemblages. Thickness measures of late Late Woodland Juntunen and Traverse wares was the sole significantly different factor, a result most likely related to the distinct rim styles characteristic of these wares. Although the small sample sizes available for most of these comparisons may have affected the statistical analyses, there is an apparent synchronic consistency in the ways pottery vessels at the Cloudman site were constructed, regardless of decoration and style. 119 Table 6.12: Technical Properties of Vessels by Type/Ware Subset Pottery Type Rim Radius Mean n Thickness* Mean n Temper Mean n 6 7 Laurel Banked Linear Stamped Middle Woodland Laurel Dentate Stamped Laurel Pseudo- scallop Shell Blackduck Banded Mackinac Banded Mackinac Punctate Juntunen Ware Traverse Ware cf. Huron Incised cf. Lawson Ware Early Late Woodland Late Late Woodland Ontario Iroquois 22 19.5 15 22.5 20 21.1 21.2 16 21.8 27 *average neck+shoulder Intended Function Summary 2 4 5 4 8 10 9 4 8 1 5.41 6.25 6 8 1.5 1.41 6.31 13 1.46 13 6.52 4 1.37 4 6.94 11 1.82 11 6.78 8.47 6.05 7.79 7.82 15 22 9 17 3 1.56 1.39 1.29 1.22 1.25 15 23 10 20 5 Middle Woodland pottery vessels were overall the most technologically distinct among the Cloudman assemblage. Middle Woodland pots were generally small, thin pots with large temper particles relative to thickness. Although vessel size significantly increased in the early Late Woodland, thickness and temper size remained relatively consistent with Middle Woodland properties. Late Late Woodland potters continued to construct larger but thicker pots with significantly smaller temper particles. Iroquoian vessels, although displaying very different decorative techniques, were statistically identical to late Late Woodland vessels in their technical properties, aside from a slight reduction in thickness. Late Woodland and Iroquoian potters preferred larger vessels, although the functional advantage of greater vessel volume is still 120 unclear. Further exploration of vessel use, cooking, and diet may provide answers to this question. After AD 1200, there is clear evidence for intentional use of significantly smaller temper particles, a decision that could reflect new food processing requirements entailing more intensive cooking techniques. The diachronic trends in technical variation contrasts with synchronic consistency of the same properties, suggesting that contemporary social groups did not enact social identity through distinctive pottery construction techniques. Use-Alteration Traces and Actual Function The functions fulfilled by a pottery vessel in the past can be accessed through use- alterations traces, or the physical and chemical changes that indicate the various processes in which the vessel was involved throughout its life history (Schiffer and Skibo 1987, 1997). Use- alteration traces include exterior sooting, exterior carbonization, interior carbonization, attrition, and absorbed residues (see Chapter 4). Sooting and carbonization are direct evidence of a vessel’s use over fire, from which the function of cooking can be inferred. Food residues are absorbed in the vessel walls during food processing and storage and show the direct relationship between specific foods and individual pottery vessels. Attrition involves removal of the pottery surface through various physical and chemical processes; however, this was not detected in any of the Cloudman pottery assemblage and will not be further discussed. Absorbed food residues will be summarized and discussed separately in Chapter 7. The Cloudman pottery assemblage was assessed for the presence of exterior sooting, exterior carbonization, and interior carbonization. Patterning of interior carbonization was also categorized and recorded to identify distinct cooking techniques. Presence and patterning of use- alteration traces are compared both diachronically and synchronically to identify trends and 121 changes in pottery use at the Cloudman site. Together, these data provide insight into the actual function of Cloudman pottery vessels. Exterior Sooting Sooting on the exterior of a pottery vessel forms when smoke from a fire adheres to the ceramic surface and is an important indication of a vessel’s involvement in cooking. Sooting most commonly forms around the bottom and lower sides of a vessel, where the vessel comes in contact with smoke from the fire (Skibo 1992, 2013). Only 10% of the overall Cloudman pottery assemblage displayed exterior sooting (Table 6.13). Sooting was most frequent on early Late Woodland vessels, the subset containing the most complete and partial vessels. As previously discussed, vessels from other subsets are often represented only by rim or upper body sherds, where exterior sooting is unlikely to be present. Therefore, sooting is probably underrepresented among the identified vessels in the Cloudman assemblage, which overall have few associated body sherds and even fewer basal sherds. While a review of body sherds not associated with identified vessels would likely reveal a higher frequency of sooting at the site, it would be difficult to associate these sherds with a specific occupation and would therefore not contribute to the study of diachronic pottery function. Exterior Carbonization Burned food residue on the exterior surface of a pottery vessel is another indicator of involvement of a vessel in cooking. Foodstuffs burned onto the outside of a cooking pot can be the result of pouring/spilling or boiling-over. Exterior carbonization was positively identified on approximately 15% of all vessels (see Table 6.13). Exterior carbonization is most likely to 122 appear on the upper portions of a vessel and is, therefore, likely not underrepresented. However, only a small proportion of cooking events would be expected to result in exterior carbonization, so frequencies observed within the Cloudman assemblage align with expectations. Table 6.13: Frequency of Use-Alteration Traces by Subset Subset Exterior Sooting Exterior Interior Carbonization Carbonization Middle Woodland Middle/Late Woodland Early Late Woodland Middle Late Woodland Late Late Woodland Ontario Iroquois Total Ct. 0 0 14 1 2 4 21 % 0.0 0.0 22.6 20.0 4.1 10.8 10.4 Ct. 2 1 10 0 11 7 31 % 5.7 14.3 16.1 0.0 22.4 18.9 15.3 Ct. 16 4 34 3 28 21 106 % 45.7 57.1 54.8 60.0 57.1 56.8 53.0 Interior Carbonization Food residue burned onto the interior surface of the pot is the most important evidence of the involvement of a vessel in cooking. The mere presence of interior carbonization is direct proof that resource processing took place within the vessel, and the patterning (discussed below) and microbotanical/chemical composition (see Chapter 7) of the residue is vital to reconstructing cooking and food selection habits. Over half of all identified vessels from the Cloudman site displayed interior carbonization (see Table 6.13), suggesting that a high proportion of vessels were involved in resource processing. Interior carbonization is consistently present through time, occurring on 45% to 60% 123 of vessels from each subset. Cooking appears to be the primary function of pottery vessels during all occupations of the Cloudman site. Despite its relatively high frequency, interior carbonization is still likely to be underrepresented in the assemblage. Differential preservation of residues on pottery surfaces is evident on vessels with portions that refit; in many cases, one sherd is stained with carbonization while an adjacent sherd is not. Discrepancies could result from variations in post-depositional taphonomic processes or archaeological lab processing. Ancient pottery cleaning habits may also affect the preservation of adhered interior residues. Some vessels may have never been used at all. The fail rate of firing pottery in open fire can be as high as 100% (Rice 1987:173), so vessels without use-alteration traces may represent vessels that broke during manufacture, causing overrepresentation of vessels lacking carbonization. However, it is plausible that pots lacking visible interior residues were used as storage vessels. Habitual Cooking Behaviors Vessels with interior carbonization were further categorized based on the patterning of residue along the vessel wall. Carbonization location and patterning inform interpretations of past cooking habits and methods and are rooted in ethnoarchaeological observations (see Kobayashi 1994; Skibo 1994). Both the level at which vessels are filled and the patterning of the interior carbonization are useful indicators cooking traditions, such as habitual vessel use and modes of cooking. Vessel Fill Levels The presence of carbonization on the uppermost portions of the interior surface is almost ubiquitous among Cloudman site pottery vessels. In many cases, residue extended all the way to the lip; in others there was a half- to one-inch buffer between the residue 124 line and the lip. This pattern indicates high fill-levels during cooking. This is consistent with patterning found in other northern Great Lakes sites (Kooiman 2012, 2016), and stands in contrast to habits found in other regions of the world, where cooking pots are often filled two- thirds full to prevent boiling-over (Kobayashi 1994; Skibo and Blinman 1999), a costly mistake that would waste food and douse the hearth fire. Several ethnographic sources note that organic skin or birchbark vessels can be placed directly over the fire without burning as long as they are kept full (Densmore 1979; Wallis and Wallis 1955; Waugh 1973), potentially representing a habitual behavior extended to pottery vessels, which were adopted in the region after millennia of cooking in organic vessels. In this case, boil-overs may not have been as costly a mistake as in other regions, such as Colorado (see Skibo and Blinman 1999), where population density is higher and access to wood resources is more limited than in the northern Great Lakes. Another explanation for observed filling behavior is technology-based. According to Erik Vosteen, a local Michigan potter who experiments with making pottery of local clays in the style of Woodland vessels, low-fired pottery vessels also tend to potlid along the water line because of the temperature differential between the liquid and the air above it. Filling the vessel could have reduced potlidding along the critical shoulder zone of the vessel and reduced vessel breakdown and/or prevented vessel fracture, prolonging the life of the vessel. Potlidding was not frequently observed in the Cloudman pottery assemblage, so this technique may have functioned well for local cooks. Considering that many vessels in the Cloudman assemblage are represented solely by rim sherds, it is possible that fill lines along the shoulder of the vessel are present but not observable 125 in the sample. However, defined fill lines at the vessel shoulder were not observed in any of the partial vessels in the assemblage. Interior Carbonization Patterns Although cooks at the Cloudman site consistently filled vessels to the top during the cooking process, the cooking techniques employed were varied. Figure 6.2 illustrates the five patterns observed in the Cloudman assemblage, which signify different modes of cooking. Pattern 1 (Figures 6.2a, 6.3) represents residue restricted to a distinct ring around the rim of the vessel. This pattern is consistent with boiling, a wet-cooking mode in which the aqueous nature of the vessel contents prevent adherence of food particles the ceramic surface, except at the water line, where both small starchy and fatty food particles accumulate and burn in this high-temperature zone (Skibo 1992, 2013). Pattern 2 (Figures 6.2b, 6.4) represents residue distributed over most or all of the interior surface of the pot, which most likely represents stewing, a long-term liquid reduction process. Stewing gradually removes much of the water from the food mix, allowing more food particles to come into contact with the vessel wall and become charred when exposure is prolonged. Dry mode cooking techniques, such as parching or roasting, can also leave thick residue deposits across the vessel surface, although this is usually restricted to either the very bottom or to one side of the pot (Skibo 1992, 2013). Roasting often takes place with a vessel placed on its side and neither interior carbonization nor exterior sooting patterns observed in the Cloudman assemblage are consistent with this behavior. However, given the lack of complete vessels from the Cloudman site, it cannot be stated that pottery vessels were never used for roasting/parching, although this pattern was not observed at other northern Great Lakes sites with more complete or partially-reconstructed vessels (Kooiman 2012, 2016). 126 Figure 6.2: Interior Carbonization Patterns for the Cloudman Site (20CH6) Pottery Assemblage: a) Type 1 (boiling); b) Type 2 (stewing); c) Type 3 (boiling + stewing); d) Type 4 (possible boiling or stewing; e) Type 5 (no discernable pattern) 127 Figure 6.3: Interior Carbonization Pattern 1 (Boiling) Figure 6.4: Interior Carbonization Pattern 2 (Stewing) 128 Figure 6.5: Interior Carbonization Pattern 3 (Boiling + Stewing) Interior carbonization Pattern 3 (Figures 6.2c, 6.5) represents a ring of thick carbonization around the rim with a thinner layer of residue covering the remaining interior surface, indicative of a single vessel involved in both boiling and stewing of foods. This pattern has not been previously observed in the northern Great Lakes (Kooiman 2012, 2016). Vessels displaying only one type of carbonization pattern could signify the designation of vessels for specific purposes (e.g., boiling pots vs. stew pots), or that stewing events obscured evidence of prior boiling episodes. Vessels could have also been washed and scrubbed between uses, the pattern present archaeologically representing only the last mode of cooking for which the pot was used. The newly observed Pattern 3, however, is a clear indicator of multiple cooking events taking place in the same vessel: one or more intensive boiling episodes resulting in thick residue deposition, distinct from thinner residues representing less-intensive prior or subsequent stewing events. Pattern 4 (Figure 6.2d) was used to categorize cases in which interior carbonization is visible along the rim of the vessel but the extent of the residue below the rim is indeterminate. 129 This pattern could represent either boiling or stewing. In most cases there is not enough of the rim present to determine whether or not the residue distribution distinctly stops at or near the base of the rim (boiling) or continues further down the side of the interior vessel surface (stewing). Pattern 5 (Figure 6.2e) represents patchy interior carbonization on the vessel surface with no discernable pattern. Table 6.14 shows the frequency of each interior carbonization pattern by primary analytic subset. The most frequent pattern among all subsets is Pattern 4, largely due to the fragmentary nature of the assemblage, which precludes proper assessment of the distributional extent of residues on many vessels, especially those represented by small rim sherds. Unfortunately, Pattern 4 has very little analytic meaning aside from demonstrating the prevalence of vessel filling behaviors. Pattern 5 likewise holds little analytical significance. Subsequent discussion will therefore focus on Patterns 1-3 (Table 6.15, Figure 6.6), which more clearly represent specific cooking behaviors. Pattern Category 1 2 3 4 5 Total Table 6.14: Interior Carbonization Pattern Frequency by Subset Middle Woodland Early Late Woodland Late Late Woodland Iroquoian Totals Ct. 2 5 0 4 5 16 % 13 31 0 25 31 100 Ct. 8 4 4 14 4 34 % 24 12 12 41 12 100 Ct. 6 2 4 10 3 25 % 24 8 16 40 12 100 Ct. 4 2 4 9 2 21 % 19 10 19 43 10 100 Ct. 20 13 12 37 14 96 % 21 14 13 39 15 100 130 Among Middle Woodland vessels, Pattern 2 (stewing) is the most frequent interior carbonization pattern. This is consistent with prior assessments of Middle Woodland cooking, where stewing predominates and boiling is less frequently represented (Kooiman 2012, 2016). However, the fragmentary nature of the Middle Woodland vessels resulted in a small sample size (n=7) of vessels with clear carbonization patterns, thereby affecting analytical outcomes. Pattern 3 was not observed among Middle Woodland vessels, and clear signatures for boiling were observed in only two vessels. Pattern 2 is much less frequent among the early Late Woodland, late Late Woodland, and Iroquoian subsets, in which Patterns 1 and 3 predominate. Both boiling and boiling/stewing methods increased in frequency at the outset of Late Woodland and carried on until the contact period. As previously observed in the northern Great Lakes, boiling seems to become a more frequent cooking technique during the later periods, although stewing habits continue through all occupations. Table 6.15: Primary Interior Carbonization Pattern Frequency by Subset Pattern Category 1 2 3 Total Middle Woodland Early Late Woodland Late Late Woodland Ontario Iroquois Ct. 2 5 0 7 % 29 71 0 100 Ct. 8 4 4 16 % 50 25 25 100 Ct. 6 2 4 12 % 50 17 33 100 Ct. 4 2 4 10 % 40 20 40 100 131 Figure 6.6: Proportions of Interior Carbonization Patterns by Subset To assess the statistical significance of patterns suggested in the raw frequencies, the data was subjected to a Kruskal-Wallace test. Kruskal–Wallis is a non-parametric, one-way analysis of variance by ranks that evaluates whether k independent samples are from different populations (Siegel 1956). The results of the Kruskal-Wallis show no significant differences in the distributions of interior carbonization patterns between analytic subsets (Table 6.16). The small sample sizes may have influenced this outcome, and although none of the differences are statistically significant, it can be observed that the Middle Woodland subset is statistically more different from early Late Woodland, late Late Woodland, and Iroquoian subsets than they are from each other. In fact, these latter subsets are close to statistically identical. The evidence is still indicative of an overall shift in cooking behaviors after AD 500. 132 Table 6.16: Interior Carbonization Pattern Relationships (Kruskal-Wallis) Subset Middle Woodland Early Late Woodland Late Late Woodland Middle Woodland Early Late Woodland Late Late Woodland Ontario Iroquois (p*) --- 0.094 0.183 0.072 (p) --- --- 0.838 0.732 (p) --- --- --- 0.710 *two-tailed α=0.05, df=2 Synchronic Variation of Interior Carbonization Patterns Variability of cooking styles between contemporary pottery types could indicate distinct culinary traditions associated with social or cultural identity. As with synchronic variation of technical properties (see above), assessment of synchronic variation is limited by small sample sizes. Table 6.17 shows the proportions of vessels within the major typological categories of each primary analytic subset exhibiting interior carbonization Patterns 1, 2, and 3. The most drastic difference is apparent in the Middle Woodland subset, in which two-thirds of the Laurel Banked Linear Stamped vessels with clearly patterned interior carbonization exhibited Pattern 1 (boiling), whereas Pattern 2 (stewing) was the only pattern present among Laurel Dentate Stamped and Laurel Pseudo- Scallop Shell vessels. The samples sizes, however, are too small for statistical analysis and interpretations about specialized vessel use. Fewer differences between types are apparent in early Late Woodland, late Late Woodland, and Iroquoian subsets. Again, sample sizes are too small to determine whether cooking styles varied significantly between stylistic types. Conclusions about synchronic variation in cooking and pottery use are therefore limited. 133 Table 6.17: Interior Carbonization Pattern Frequency by Type/Ware Subset Pottery Type n Ct Pattern 1 Pattern 2 Pattern 3 % Ct % Ct % Laurel Banked Linear Stamped 3 2 66.7 1 1 2 0 1 1 4 1 0 2 33.3 100 100 0 25 20 50 25 0 67.7 0 0 0 1 1 1 2 2 2 1 0 0 0 100 25 20 25 50 67.7 33.3 Middle Woodland Laurel Dentate Stamped Laurel Pseudo- scallop Shell Blackduck Banded Mackinac Banded Mackinac Punctate Early Late Woodland Late Late Woodland Juntunen Ware Traverse Ware cf. Huron Incised Ontario Iroquois 1 0 2 0 1 0 0 0 0 4 2 50 5 8 4 3 3 2 1 1 60 25 25 33.3 cf. Lawson Ware 3 0 0 Actual Function Summary Use-alteration traces present on the Cloudman site pottery vessels allow several insights into the actual function of ancient pottery. Over half of all vessels contained interior carbonization, strong evidence that the primary function of pottery at the Cloudman site was for cooking. Although the proportion of the assemblage with exterior sooting, another good indicator of cooking function, is low, this may be the result of the low frequency of identified vessels with associated body sherds, where sooting is most likely to occur. As with other Woodland vessels found throughout the northern Great Lakes, cooking vessels at the Cloudman site were filled 134 nearly to the top, a routine potentially carried over from cooking with organic vessels, or a conscious choice for limiting potlidding during the cooking process. Interior carbonization patterns, while not significantly different between analytic subsets, still change in accordance with previous observations of cooking change in the Middle and Late Woodland periods (Kooiman 2012, 2016). Clear boiling signatures are proportionally low in the Middle Woodland, a period during which there is also a complete absence of clear evidence that the same vessels were employed in both boiling and stewing. Larger proportions of early Late Woodland, late Late Woodland, and Iroquoian vessels displayed boiling and dual-use patterns. Changes in patterning frequencies among Cloudman pottery vessels supports the hypothesis that cooking styles changed over time, and that boiling became an increasingly important means of food processing during the Late Woodland/Late Precontact period. As with technical properties, synchronic variation between distinct types within the same analytic subsets was low. Sample sizes were too small for detectable or meaningful distinctions. The matter of differential pottery use based on group identity will require more data for accurate assessment. Discussion Results of functional analysis of pottery from the Cloudman site demonstrate the tension between tradition and innovation among Woodland and Late Prehistoric peoples. Pottery construction, pottery use, and cooking methods were altered to suit new needs and new generations, albeit slowly and subtly. Middle Woodland vessels were relatively small with large temper in relation to thickness. Early Late Woodland pots maintained thicknesses similar to their Middle Woodland predecessors, but gained in overall size/volume. The upper portions of late 135 Late Woodland vessels grew thicker while grit temper particles became significantly smaller. The same properties also characterize Ontario Iroquoian vessels. Vessel size variation could be associated with several factors. Middle Woodland groups were the first to use pottery vessels despite being generally mobile with malleable social structure (Brose and Hambacher 1999). Their mobile nature may have constrained the construction of vessels to sizes manageable during travel. As attraction and attachment to “persistent places,” or locales repeatedly occupied and used for intensive resource extraction, increased throughout the Late Woodland period (see Dunham 2014, 2017), vessels may have been constructed for in situ use and stored for future use at the site. This would have given potters the freedom to make larger pots. If, as Cleland (1982) claimed, occupation of coastal aggregation sites increased in size and duration during the Late Woodland, the larger vessel size could have also grown according to needs. The early Late Woodland vessels assemblage is the largest at Cloudman (n=49), suggesting either larger or more prolonged occupations after AD 700/800. Potters continued to construct larger vessels post-AD 1200, but they also began incorporating significantly smaller temper into ceramic pastes. Smaller temper improves both impact resistance and thermal shock resistance (Bronitsky and Hamer 1986), the latter integral for maintaining vessel integrity during heat-intensive and/or long-term cooking events. Smaller temper particles also provide greater green phase (pre-firing) ceramic strength (Chu 1968; Rice 1987:362), facilitating the often more complex decorations and collars common to Juntunen and Iroquoian vessels. Thinner walls allow for greater heating effectiveness during cooking. Middle Woodland vessels are significantly thinner than Late Woodland and Iroquoian vessels, but when thickness 136 was corrected for rim diameter (a proxy for vessel size), it was clear that the thin walls of Middle Woodland vessels was a consequence of their overall smaller dimensions. Neck/shoulder thickness significantly increased among late Late Woodland vessels compared to their early Late Woodland predecessors. Body thickness would be the most important factor in heating effectiveness, and an underrepresentation of body sherds among the identified Cloudman vessels affected the sample size for effectively assessing the evolution of this characteristic through time. The relationship between neck/shoulder thickness and body thickness is unclear. The transition from the everted-rim profile of early Late Woodland vessels to the collared wares characteristic of both late Late Woodland and Iroquoian vessels would require thickened neck and shoulder to support the thick, heavy rims. Therefore, observations of variation in thickness within the Cloudman assemblage seem more closely related to size and style rather than cooking function. Changes in cooking techniques are apparent, if subtle. Middle Woodland vessels were less frequently engaged in boiling-only cooking events, and signs of using Middle Woodland vessels for both boiling and stewing are completely absent. Clear signatures for boiling and boiling/stewing cooking techniques increase in relative frequency among early Late Woodland vessels and remain consistent throughout subsequent occupations of the site. Small sample sizes preclude statistical support of the observed frequencies. Synchronic variation of both technical properties and use-alteration traces was limited. This is also due in part to the small sample sizes, and a region-wide survey of northern Great Lakes pottery assemblages would be required to obtain sufficient data. It must also be noted that although many types within the primary subsets overlap in time, they are not entirely synchronic. Mackinac Punctate, for example, generally occurs earlier (AD 700-900) than Mackinac Banded (AD 850-1000) (Lovis 2014), and therefore do not represent full contemporaneity. Rather, they 137 likely represent evolving stylistic techniques among related descendant groups rather than separate social groups occupying the same space and time. Conclusion The evolution of pottery technology and use at the Cloudman site both aligns and contrasts with outlined expectations. The greatest distinctions fall between Middle Woodland and Late Woodland/Iroquoian pottery. Although vessel size was not predicted to change, Middle Woodland vessels were smaller, with vessel size increasing in size in the early Late Woodland and remaining consistent thereafter. These smaller vessels were employed in stewing more frequently than boiling, a cooking style that, as predicted, increases in frequency through time. Although boiling becomes more frequent and vessel size increases around the same time, it is presently unclear whether these two factors are connected by intentional choice to fulfill a particular function, or if both are the results of greater social and behavioral changes occurring at the outset of the Late Woodland period. A significant decrease in temper size, which has been associated with increased processing and consumption of starchy foods, including maize, does not appear in vessels until after AD 1200, subsequent to changes in both vessel size and cooking habits. Technological variability of pottery is often associated with changes in subsistence, as processing needs change in relation to food types. The next chapter will address questions of subsistence and cooking at the Cloudman site with a discussion of pottery residue composition. The combination of lipid residue, stable isotope, and microbotanical analyses provide direct evidence of foods cooked in pottery vessels, allowing closer associations between dietary, technological, and cooking style changes. 138 CHAPTER 7 FOOD SELECTION AND COOKING AT THE CLOUDMAN SITE Introduction Holistic depictions of past food choice and processing techniques derived from multiple lines of archaeological evidence allow for reconstructions of ancient cuisine and a nuanced view of past identity and adaptive decision-making. Archaeological food residues, both adhered and absorbed, link specific foods directly to processing tools and techniques, informing interpretations of synchronic and diachronic variation in food choice and cooking. This chapter assesses diet and cooking behaviors at the Cloudman site using stable isotope, lipid residue, and microbotanical analyses of adhered and absorbed food residues associated with ceramic cooking vessels. The results will be used to assess the degree of diachronic dietary change, especially in relation to hypotheses arguing for intensified use of aquatic (Cleland 1982), wild starchy (Dunham 2014), and/or domesticated (O’Shea 2003) resources in the northern Great Lakes throughout the Woodland period. Synchronic variation in food choice in relation to group identity will also be discussed. Finally, the outcomes of each method will be evaluated in context with each other to identify strengths and shortcomings of each, formulating best practices for evaluating diet from pottery residues. Images of all vessels included in residue analyses can be found in Appendix G. Carbon and Nitrogen Stable Isotope Analysis Intensified exploitation of both fish and maize at coastal sites but primarily autumn season coastal sites during the Late Woodland period have been proposed by Cleland (1982) and 139 O’Shea (2003), respectively. Nitrogen and carbon stable isotope ratios have been used to identify aquatic resources and maize in carbonized food residues (e.g. Craig et al. 2013; Hart et al. 2003; Lovis 1990; Morton and Schwarcz 2004; Taché and Craig 2015). Stable isotope analysis is thus well-suited for informing questions about diachronic variation in aquatic and terrestrial fauna, trophic levels, and maize exploitation. Carbonized food residue was collected from the interior rims of 50 identified vessels, selected from all periods of occupation of the Cloudman site. The samples were sent to the Illinois State Geological Survey Prairie Research Institute for analysis of bulk stable isotope values of δ15N and δ13C (Appendix B). Nitrogen Isotopes The results of stable isotope analysis of Cloudman site residue samples revealed consistently enriched δ15N values (Table 7.1, Figure 7.1), with a mean of 11.69‰. δ15N values higher than 9‰ are associated with processing of high-trophic-level aquatic products, as observed in experimental cooking residues (Craig et al. 2007; Craig et al. 2013; Morton and Schwarcz 2004). Only residues from two vessels (V88, a Blackduck Banded vessel, and V179, an Iroquoian vessel) fell below this threshold. Table 7.1: Mean δ15N and δ13C Values of Cloudman Pottery Residues by Subset Subset Mean 15NAir 13CVPDB Middle Woodland Middle/Late Woodland Early Late Woodland Middle Late Woodland Late Late Woodland Iroquoian 11.87 13.28 11.74 10.96 11.98 10.72 140 -25.78 -21.78 -25.60 -24.54 -26.01 -25.93 n 11 1 20 2 10 6 Figure 7.1: Plot of δ15N/δ13C Values of Cloudman Pottery Residues Diagenesis associated with the surrounding matrix has been invoked as a possible source of error in isotope studies. Potential contamination of the soil from nitrogen rich fertilizers was considered a possible factor in these results, but an interview with Gary Cloudman, whose family has owned the land surrounding the Cloudman site for over one hundred years, revealed that intensive agriculture has never taken place on the property. Stratigraphic profiles from excavation corroborate this account, demonstrating the lack of a plow zone (Ap horizon) at the site (Branstner 1995). To further evaluate the potential for contamination from the soil matrix, a total of 12 soil samples were taken from across the site and subjected to stable isotope analysis. These samples were consistently less δ15N enriched than samples deriving from pottery, averaging 8.41‰ (Table 7.2, Figure 7.2). Only one sample reached levels of nitrogen enrichment similar to the pottery residues samples. When subjected to an unpaired T-test, δ15N levels of the 141 pottery residue samples and the soil samples proved extremely significantly different (p=0.0001, df=60). Nitrogen contamination of pottery residues from the site matrix is, therefore, unlikely. Table 7.2: δ15N Values, Cloudman Soil Samples Sample T1-S1-2 T1-S2-1 T1-S2-2 T1-S2-3 T2-S1-2 T2-S2-1 T2-S2-2 T2-S2-3 T3-S1-1 T3-S2-1 T3-S2-2 T3-S2-3 Context d15NAir Terrace 1, Stratum 1 Terrace 1, Stratum 2 Terrace 1, Stratum 2 Terrace 1, Stratum 2 Terrace 2, Stratum 1 Terrace 2, Stratum 2 Terrace 2, Stratum 2 Terrace 2, Stratum 2 Terrace 3, Stratum 1 Terrace 3, Stratum 2 Terrace 3, Stratum 2 Terrace 3, Stratum 2 8.70 8.79 8.86 7.47 8.89 8.12 7.25 7.32 8.26 9.16 5.80 12.27 ) l i m r e p ( N 5 1 a t l e d 16.00 14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00 -35.00 -30.00 -25.00 -20.00 -15.00 -10.00 -5.00 0.00 delta 13C (per mil) Pottery Residues Soil Samples Figure 7.2: Plot of δ15N/δ13C Values of Cloudman Pottery Residues vs. Soil Samples 142 Eliminating the possibility of contamination, nitrogen enrichment of the Cloudman pottery residues can be directly associated with foods processed in the vessels. Aquatic resources produce higher δ15N values than terrestrial resources, and nitrogen isotope ratios also display a “trophic-level effect,” where a consumer will be enriched 2 to 4‰ higher that the source of dietary protein (Shoeninger 1985; Schwarcz and Schoeninger 1991). Nitrogen levels of aquatic vertebrates can be 6 to 8‰ higher than terrestrial vertebrates of equivalent trophic levels (Schoeninger 1995:85), and therefore high trophic-level fish (fish that consume other fish) will produce among the highest δ15N values (Van der Merwe et al. 2003). The enriched nitrogen isotope signatures from the Cloudman samples appear indicative of high-trophic-level aquatic resources, which in the Upper Great Lakes would include spring-spawning fish such as pike, perch, and walleye, and fall-spawning species such as whitefish and lake trout (Vander Zanden et al. 1997). The location and high proportion of faunal remains derived from fish (Cooper 1996) suggest that the Cloudman site was used as a base for fishing activities. In this context, the presence of chemical signatures that may represent fish in pottery food residues is unsurprising. However, prior research found that habitual processing of fish in pottery vessels across the upper Great Lakes was not common (Kooiman 2016; Lovis and Hart 2015; Malainey and Figol 2015; Skibo et al. 2016), although others have argued the opposite (Taché and Craig 2015). To further evaluate the association of elevated δ15N values of the Cloudman residue samples with high-trophic-level fish, isotope data from other studies with isotope measures for various remains and resources were sought for comparison. Katzenberg (1989) assessed the δ15N and δ13C values of bone collagen collected from archaeological faunal samples from the Kelly- Campbell site in southern Ontario (Figure 7.3). The plotted values for fowl and mammals, 143 including bear, deer, and dog, fell below the nitrogen delta values observed in the Cloudman samples. Closest to the Cloudman δ15N values were fish, including walleye and the lower-trophic level species pickerel and sucker, as well as two raccoon samples. Raccoons are omnivores known to eat large amounts fish, which could lead to δ15N enrichment of bone collagen. Figure 7.3: Plot of δ15N/δ13C Values of Archaeological Faunal Samples, Kelly-Campbell Site, Ontario (Katzenberg 1989) Table 7.3: Relationship between δ15N Values of Cloudman Pottery Residue Samples and Other Archaeological and Biological Samples (Unpaired T-test) Sample Archaeological fish remains (SW Ontario) Modern fish remains (Lake Michigan) p 0.2232 0.101 df 55 58 144 ) l i m r e p ( N 5 1 a t l e d 16.00 14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00 Cloudman samples Archaeological Fish Remains -35.00 -30.00 -25.00 -20.00 -15.00 -10.00 -5.00 0.00 delta 13C (per mil) Figure 7.4: Plot of δ15N/δ13C Values of Cloudman Pottery Residues vs. Archaeological Fish Remains from Southern Ontario (Van der Merwe et al. 2003) and Belgium (Fuller et al. 2012) Archaeological remains of high trophic-level fish (including lake trout, whitefish, pike, and walleye) from sites in southern Ontario (Van der Merwe et al. 2003) and Belgium (Fuller et al. 2012) also lie within the range of δ15N values from the Cloudman pottery samples (Figure 7.4) and are statistically undifferentiated from Cloudman pottery residue δ15N values (Table 7.3). Isotope values taken from modern-day perch and trout samples from Lake Michigan also have equivalent δ15N values (Turschack 2013; Figure 7.5; see Table 7.2) and strongly suggest that the δ15N values of the Cloudman residues represent fish. Royer et al (2017) found that boiling fish can actually increase δ15N values by 2‰, which would bring the range of the of the fish from the referenced isotopic studies even more closely in line with the Cloudman residue values. Finally, experimentally-processed high-trophic-level aquatic resources have observed δ15N values of 9 ‰ or greater (Craig et al. 2007; Craig et al. 2013; Morton and Schwarcz 2004), a threshold exceeded by 96% of the Cloudman residues. 145 Human bone collagen samples from both Woodland and Iroquoian archaeological sites in southern Ontario with δ15N values between 11.8 - 12.5‰ have been interpreted as indicating diets comprising significant amounts of high trophic-level protein sources, most likely fish (DeWar et al. 2010; Schwarcz et al. 1985; Vander Merwe et al. 2003). The δ15N values of Cloudman pottery residues are similar to those derived from bone collagen of individuals from Ontario and average δ15N values for individuals from the nearby late Late Woodland Juntunen site, located in the Straits of Mackinac, as depicted in Figure 7.6 (Brandt 1996; DeWar et al. 2010; Schwarcz et al. 1985; Vander Merwe et al. 2003). Human bone collagen is typically between 2-4‰ enriched in nitrogen compared to dietary components (including cooked food remnants) due to the fractionation factor between a consumer and its food (Ambrose and DeNiro 1986; Morton and Schwarcz 2004; Shoeninger 1985; Schwarcz and Schoeninger 1991), so the ) l i m r e p ( N 5 1 a t l e d 16.00 14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00 Cloudman samples Modern Lake Michigan Fish -35.00 -30.00 -25.00 -20.00 -15.00 -10.00 -5.00 0.00 delta 13C (per mil) Figure 7.5: Plot of δ15N/δ13C Values of Cloudman Pottery Residues vs. Modern Fish Samples from Lake Michigan (Turschack 2013) 146 ) l i m r e p ( N 5 1 a t l e d 16.00 14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00 Cloudman samples Human remains (SW Ontario) Human remains (Juntunen site) -35.00 -30.00 -25.00 -20.00 -15.00 -10.00 -5.00 0.00 delta 13C (per mil) Figure 7.6: Plot of δ15N/δ13C Values of Cloudman Pottery Residues vs. Human Bone Collagen of Woodland & Iroquoian Individuals with Aquatic Resource-Rich Diets (Brandt 1996; DeWar et al. 2010; Schwarcz et al. 1985; Vander Merwe et al. 2003) residues from Cloudman may represent food mixes comprised of even greater levels of dietary protein derived from high-trophic-level sources in comparison to the average diets of the individuals sampled for bone collagen. Aquatic plants can also yield high δ15N values, although these values vary based on geographic location. In the San Fransisco Bay of California, δ15N ranged from -3.4 to 17.4‰ and δ13C ranged from -32 to -12.4, clustering at levels slightly less enriched in both carbon and nitrogen than the Cloudman residues (Cloern et al 2002). In Spain, δ15N values freshwater aquatic plants fell in the range of 5.2 to 20.1‰ (with a median of 3.5‰ [Chappuis et al. 2017]), while freshwater plants in subarctic Canada produced δ15N values in the range of -5.4 to 7.4‰ (Milligan et al. 2010). The δ15N values from the subarctic are especially low compared to those of the Cloudman residues, showing geographic variability that may also be influenced by the geomorphological type of the body of water in which plants grow (Chappuis et al. 2017). It 147 cannot be ruled out that processing of aquatic plants may have also contributed to the δ15N enrichment of the Cloudman samples, although no aquatic plant remains were present in the microbotanical assemblage at the site (Egan-Bruhy 2007). Further exploration of isotope values for aquatic plants in the northern Great Lakes is required to clarify future isotopic interpretations (see Methodological Considerations below). Overall, the stable isotope values of the Cloudman samples display considerable consistency at the site through time. Figure 7.7 shows the lack of distinguishable clustering of isotope ratios by subassemblage, a similarity reflected in the mean δ15N values of each subset (see Table 7.1). The ubiquity of elevated delta values in 48 of 50 total vessels sampled, across all subassemblages and without visible clustering is solid evidence for a long tradition processing of aquatic resources, possibly high-trophic-level fish, in pottery throughout all occupations of the Cloudman site. Cleland’s (1982) fall fishery model includes increased exploitation of deep-water fall spawning fish species, such as lake trout and whitefish (high trophic-level species), during the Late Woodland period. To support Cleland’s hypothesis, an observed increase, steady or of shorter duration, in pottery residues with elevated δ15N values over time would be necessary. However, high trophic-level fish were exploited in both the spring and fall (Cleland 1966), so species, and therefore seasonality, cannot be attributed based on the isotope data alone. However, the consistency of the nitrogen isotope data demonstrates that acquisition, processing, and consumption of aquatic resources, including high-trophic-level fish, were important activities throughout the 1400-year occupational history of the Cloudman site. 148 ) l i m r e p ( N 5 1 t l e d 16.00 14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00 -35.00 -30.00 -25.00 -20.00 -15.00 -10.00 -5.00 0.00 delta 13C (per mil) Middle Woodland Middle/Late Woodland Early Late Woodland Middle Late Woodland Late Late Woodland Iroquoian Figure 7.7: Plot of δ15N/δ13C Values of Cloudman Pottery Residues by Subset Carbon Isotopes Intensification of maize horticulture by groups occupying coastal zones of the northern Great Lakes in the late Late Woodland period has been hypothesized by O’Shea (2003). Dunham (2014) found that maize was consumed but was not a significant starchy resource for late Late Woodland groups in the eastern Upper Peninsula of Michigan. Instead, they focused on acorn and wild rice exploitation to obtain starch, a critical component of diet. Past standards interpret δ13C values above -22‰ as indicative of the presence of maize in adhered food residues (Hastorf and DeNiro 1985), while more current standards associate maize with δ13C values closer to -17.4‰ (Katzenberg and Pfeiffer 1995). As shown in Table 1, average δ13C values for all but one of the analytic subsets are well below that threshold. Bone collagen from archaeological human remains of Late Woodland and Iroquoian groups in southwest Ontario show both elevated nitrogen and carbon delta values (see Figure 7.6), which are 149 reflective of diets heavy in high-trophic-level fish and maize, respectively (DeWar et al. 2010; Schwarcz et al. 1985; Vander Merwe et al. 2003). The δ15N values of these samples are roughly congruent with those of the Cloudman samples, while the δ13C values are quite distinct, falling in the range between -10 to - 17‰. The mean δ13C values of human bone collagen samples from the Juntunen site lie between those of the Ontario and Cloudman samples. Brandt (1996:70) suggests, based on observed δ13C values, that maize made up only 18% of the Juntunen occupants’ diet. The carbon enrichment of the Cloudman pottery residue samples are not high enough to indicate intensive maize processing; however, the results do not entirely preclude the presence of maize in pottery residues (see Methodological Considerations below). Lipid Residue Analysis Detection of a wide array of both plants and animals processed in pottery cooking vessels is best attained through lipid analysis of absorbed food residues. Sherds from a total of 30 vessels were submitted to the Archaeological Residue Analysis Laboratory at Brandon University in Brandon, Manitoba for lipid residue analysis (see Table 4.3). Full results of the analysis can be found in Appendix C (Malainey and Figol 2018). Frequencies of food/content category in vessels by analytic subset are summarized in Table 7.4. Missing from this table are V35, a Late Laurel vessel from the Middle Woodland/Late Woodland transition, which contained only decomposed nut oil, and V215, a Bois Blanc vessel from the Middle Late Woodland period, which contained decomposed nut oil, low fat-content plants, and animal product. Nut processing was common at the site throughout all occupations (Figure 7.8, see Table 7.4). However, there is an increase in the frequency of vessels used for nut processing over the course of the Middle Woodland, early Late Woodland, and late Late Woodland, during which 150 Table 7.4: Frequencies of Lipid Categories by Subset Lipid Category Middle Woodland Early Late Woodland Late Late Woodland Iroquoian Ct % Ct % Ct % Ct % Decomposed Nut Oil Large Herbivore Moderate High Fat Food Medium Fat Content Food Low Fat Content Plants Animal Product Possible Animal Product Possible Plant Product Conifer Product Possible Conifer Product 3 0 1 1 2 3 1 0 0 0 60.0 0.0 20.0 20.0 40.0 60.0 20.0 0.0 0.0 0.0 7 2 0 1 5 5 0 1 1 2 77.8 22.2 0.0 11.1 55.6 55.6 0.0 11.1 11.1 22.2 7 0 0 2 4 5 1 1 1 1 87.5 0.0 0.0 25.0 50.0 62.5 12.5 12.5 12.5 12.5 3 1 0 1 4 3 2 0 1 2 50.0 16.7 0.0 16.7 66.7 50.0 33.3 0.0 16.7 33.3 87.5% of sampled vessels contained nut lipids, with a sudden decrease in frequency (50%) among Iroquoian vessels. A Kruskal-Wallis test, however, determined there was no significant differences in these frequencies (p=0.430, df=3), underlining the consistent processing of nuts during all occupations of the Cloudman site. However, acorns may have increased in importance during the Late Woodland period and became somewhat less important during later occupations of the site based on the raw frequencies of nut oils. Ethnographic accounts of Ojibwe groups consistently report the use of acorns, and rarely reference other types of nuts (Densmore 1979; Hilger 1959). Densmore (2005:307) lists hazelnut (Corylus americana) in a table of plants used as foods but does not discuss them further in text, as she does with acorns. Hazelnut, butternut/white walnut (Juglans cinerea), and acorn are 151 referenced as nuts consumed by the Ojibwe in Yarnell (1964:63, 67). Acorns are the only nut mentioned by Tooker (1991:62) as a common food among the Wendat/Huron, and Waugh (1973:123) describes Iroquois utilization of a variety of nuts (such as hickory, walnut, butternut, hazelnut, beechnut, chestnut) but emphasizes acorns. Figure 7.8: Primary Lipid Categories by Subset Acorns and hazelnuts and are the most common nuts in the eastern Upper Peninsula (Comer et al. 1995; Dunham 2009, 2014; Voss and Resnicek 2012). Macrobotanical remains from Late Woodland and “Protohistoric” features at Cloudman reveal that acorn remains account for 84% of total nut specimens and 75% of total nut weight (Egan-Bruhy 2007), but small amounts of hazelnut and butternut were also present. Presettlement vegetation at the Cloudman 152 site consisted of spruce, fir, and cedar forest, suggesting that these nuts were not necessarily gathered in the immediate vicinity of the site (Comer et al. 1997). Across to the south side of the Potagannissing River (opposite the Cloudman site), the vegetation transitions to aspen-birch forest, where hazelnut (part of the birch family) would have been available. Beech-sugar maple- hemlock forest dominate significant portions of the interior site, where acorn-producing oaks (of the beech family) would have been abundant (Comer et al. 1995, 1997). These evidences suggest that the decomposed nut oil in the lipid residues most likely represent acorns, although other nuts, particularly hazelnut, may have contributed to the signatures. The Cloudman site was either a very important acorn/nut processing locale, or processed acorns/nuts were an overall integral component of Woodland cuisine. Frequencies of the remaining food/content categories are relatively consistent across all analytic subsets, suggesting relative dietary consistency through time. Signatures for large herbivores (such as white-tailed deer) are relatively infrequent, appearing in residue from only three vessels. Moderate high-fat foods were only clearly detectable in a single Middle Woodland vessel. This signature is indicative of medium-sized mammals (Malainey and Figol 2018). Animals of this category consumed by historic-period Ojibwe and Iroquoian groups include beaver, porcupine, skunk, woodchuck, muskrat, and hare (Hilger 1959; Rogers 1962; Waugh 1973). While Ojibwe cooks sometimes boiled meat, spit-roasting was also common (Densmore 1979; Hilger 1959; Roger 1962), which may have been the primary methods of meat processing at the Cloudman site. Medium fat content foods include maize and fish. These were identified in low frequencies in every analytic subset (see Table 4), detected in only 17% of vessels sampled. The category of “low fat content plants” includes berries, roots, and greens. Berries found in macrobotanical remains at the Cloudman site include hawthorn, strawberry, cherry, wild plum, 153 raspberry, elderberry, and grape (Egan-Bruhy 2007). Other berries/fruits commonly used by historic-period Ojibwe groups include juneberry, bearberry, cranberry, currant, blackberry, and blueberry (Densmore 2005:307). Roots common in Ojibwe diets were wild ginger, wild bean, Jerusalem artichoke, and bugleweed, as well as aquatic roots like arrowhead and bulrush (Densmore 2005:307). Leaves of aster, creeping snowberry, wintergreen, and hemlock were also used (Densmore 2005:307). Low fat content plants would also include wild rice, which was recovered in the macrobotanical remains from Late Woodland features at the site (Egan-Bruhy 2007). Wild rice was central to northern Great Lakes subsistence economies during the historic period (Densmore 1979, 2005; Hilger 1959, Vennum 1988), but it cannot be distinguished from other low fat content foods in lipid residue analysis, requiring other methods of detection. The sole non-food category apparent in the lipid signatures is conifer product. Probable or possible conifer product appears in eight total vessels. Signatures for conifer product are completely absent from Middle Woodland vessels and appear minimally in every other subset. Pine resin may have been applied to the interior of the vessel to reduce permeability, as observed in precontact pottery from New York (Reber and Hart 2008) and among modern pottery- producing societies with low-fired, unglazed wares (Aronson et al. 1994; Kobayashi 1994; Skibo 2013). Cluster analysis was used to explore associations between vessels with similar lipid content. Jaccard’s coefficient of association was used because it omits joint absences of a variable to calculate similarity (Aldenderfer and Blashfield 1984). The variables included for this test include: nut lipids, large herbivore, moderate-high fat animal, medium fat content food, low fat plants, animal product, and conifer product. The categories of possible animal product, possible plant product, and possible conifer product were excluded. Figure 7.9 displays 154 Figure 7.9: Dendrogram Showing Average Linkage (Between Groups) of Vessels by Lipid Content using Jaccard’s Coefficient (vessel numbers: left column; distance: right column) associations between all sampled Cloudman vessels according to chosen variables, revealing three distinct vessel clusters. The clusters show no time-trends and cluster vessels with primarily non-specific food groups (Table 7.5). Processing of some combination of nuts, plants, and animals appears common, although whether foods were processed together in the same cooking events or in separate cooking episodes in the same vessel is not clear. 155 Table 7.5: Vessel Clusters by Lipid Content (Jaccard’s Coefficient) Vessel Subset Type Lipid Content ELW LLW IRO ELW ELW MLW LLW LLW IRO MW ELW ELW LLW LLW CLUSTER 1 Mackinac Ware Traverse Plain v. Scalloped Early Ontario Iroquoian CLUSTER 2 Mackinac Punctate Blackduck Banded Bois Blanc Ware Juntunen Ware Traverse Decorated v. Punctate cf. Huron Incised CLUSTER 3 Laurel Banked Linear Stamped Mackinac Undecorated Mackinac Banded Traverse Plain v. Scalloped Juntunen Linear Punctate 55 75 146 100 193 215 25 43 70 109 76 173 150 204 Nut oil, Low fat content plants Nut oil, Low fat content plants, Animal product Nut oil, Animal product To test for potential contamination of pottery lipid residue samples via the burial matrix, three soil samples collected from the Cloudman site were submitted for lipid residue analysis (see Malainey and Figol 2018, Appendix C). One sample deriving from the upper stratum of the lowest terrace yielded signatures typical in natural soils (Table 7.6). Samples from the middle of the central terrace, located in the heart of the site, contained lipids deriving from potential cultural activities—specifically, the processing of high-fat foods. High-fat lipid saturation, indicated by relative C18:1 isomer content, was highest in the sample from the second soil stratum, or subsoil. Malainey and Figol (2018) suggest this could be indicative of an activity area where high fat foods were processed. This sample, 17KMS32, contained a relative C18:1 composition of over 53%, much higher than ratios in any of the pottery samples. Relative C16:0 156 composition, which is indicative of decomposed nut oil, was low in the soil samples relative to those in the pottery samples. When C18:0 degrades, it turns into C16:0, so the high-fat content of the pottery sherds suggests the lipid content is from cooked/processed fatty foods rather than from exposure to unprocessed fats in the surrounding soil matrix. Past studies have found that contamination of pottery lipids by the burial environment is generally negligible (Condamin et al. 1976; Röttlander 1990), and the results from Cloudman lend additional support to this conclusion. Table 7.6: Cloudman Site Soil Sample Lipid Content Sample No. 17KMS31 17KMS32 17KMS33 Context Lipid Types Terrace 2, Stratum 1, Center Terrace 2, Stratum 2, Center Terrace 3, Stratum 1, East natural & possible cultural cultural (high-fat foods), natural natural Outcomes of the lipid residue analysis reveal that nut processing was very important at the Cloudman site during all occupations. Various terrestrial mammals and low fat content plants were also important components of the Cloudman residents’ diets. Medium fat content foods, such as fish and maize, were present in lipid signatures but infrequent. Animal and plant products were routinely cooked in the same vessels, so there were no obvious food-specific cooking functions for ceramic cooking pots at the Cloudman site. Microbotanical Analysis Although limited to the identification of certain plants, microbotanical analysis provides the greatest degree of specificity of all the food identification methods used in this study, 157 allowing classification at the species level. Adhered food residues sampled from 48 Cloudman pottery vessels were subjected to microbotanical analysis (see Table 4.1). Analysis was conducted by Rebecca Albert, the full results of which are reported in Appendix D. Phytoliths from maize, wild rice, and squash and starches from maize and squash were identified in the Cloudman samples (Table 7.7). All three food groups were represented in nearly all analytic subsets, processed and consumed at the site consistently or periodically throughout the entire 1500-year occupational history of the site. The results underscore the deep historical importance of these foodstuffs to indigenous groups in the northern Great Lakes and distinguish the Cloudman site as an important exploitation or processing locale for all three resources. Table 7.7: Number of Vessels Containing Maize, Wild Rice, and Squash Microbotanicals by Subset Subset Middle Woodland Middle/Late Woodland Early Late Woodland Middle Late Woodland Late Late Woodland Iroquoian n 12 2 19 2 9 4 Maize Wild Rice Squash 5 1 5 0 0 2 2 1 1 0 4 4 2 0 3 0 3 2 Total 48 13 12 10 The presence of maize, wild rice, and squash at the site from the outset of occupation (ca. AD 100) is a significant discovery. Maize has been reported in the Upper Peninsula of Michigan by cal 200 BC (Albert et al. 2018), in Minnesota, central New York, and southern Quebec by cal 300 BC (Burchill and Boyd 2015; Hart, Brumbach and Lusteck 2007; St-Pierre and Thompson 2015), and in the Saginaw drainage of lower Michigan by AD 1 (Raviele 2010). At the 158 Cloudman site, maize starch was encountered in residues from Vessel 4, a Laurel Dentate Stamped vessel which produced an AMS date of cal AD 60-125, yet another line of evidence supporting the early use of maize in the northern Great Lakes region. The microbotanical results also offer some of the earliest evidence for regional exploitation of wild rice and squash. Although neither plant was detected in either of the Middle Woodland vessels subjected to AMS dating, they occur in vessels of the same taxonomic typologies. A wild rice phytolith was detected in V5, identified as a Laurel Pseudo-scallop Shell vessel; a similar vessel dated to cal AD 80-214. Squash phytoliths were identified in residue from Vessel 12; a nearly identical vessel produced an associated AMS date of cal AD 60-125. Although wild rice appears in the paleoecological record in northeastern Minnesota by 7000 BC and northwestern Ontario by 4100 BC (Boyd et al. 2013; Huber 2001), evidence for widespread human exploitation of wild rice is not apparent until the Middle Woodland period, when archaeobotanical remains appear at sites from Minnesota to Ontario (Arzigian 2000; Boyd and Surette 2010; Boyd et al. 2014; Burchill and Boyd 2015; Hart and Lovis 2013; Surette 2008). Wild rice seeds from the Laurel Big Rice Lake site in northeastern MN have been dated to as early as cal 172 BC (1σ; Valppu 2000:36). In the lower peninsula of Michigan, wild rice remains were found in association with materials dated to cal AD 90-383 (1σ) at the Schultz site in the Saginaw basin (Lovis et al. 2001), and at the Dunn Farm site, near the Leelanau Peninsula (Ford and Brose 1975), where a related burial was dated to cal AD 534-635 (1σ; Brose and Hambacher 1999; Stuiver et al. 2018). Few other early contexts for wild rice in the region have been directly dated. The dates from the Cloudman site are therefore among the earliest for wild rice in the northern Great Lakes. 159 Squash was an unexpected component of the food mixes at the Cloudman site. Archaeological squash remains have been found in Pennsylvania dating to cal 5064-4336 BC (2σ), lower Michigan circa cal 2300 B.C. (Monaghan et al. 2006), southeastern Minnesota from 580 BC (Perkl 1998), and the central Mississippi River valley circa AD 434-613/681-889 (Hart, Brumbach, & Lusteck 2007). Squash has rarely been encountered in the northern Great Lakes. Cucurbit microbotanicals were found in residues from Vessel 12, a Laurel Dentate-Stamped vessel. Another Laurel Dentate-Stamped vessel (V4) produced an AMS date of cal AD 60-125, providing some of the only and earliest evidence for squash exploitation in the Upper Peninsula of Michigan. Besides the early and persistent presence of maize, wild rice, and squash at the Cloudman site, the frequency at which each species is present varies between occupations (Table 7.8; Figure 8.10). Contrary to expectations, the frequency of maize reduces over time. Present in 42% of Middle Woodland vessel residues, it occurs in only 26% early Late Woodland residues and is absent from late Late Woodland samples. Although there is evidence of moderate maize consumption among late Late Woodland populations in lower Michigan (Brandt 1996; Muhammad 2010) and casual maize consumption by late Late Woodland occupants of the nearby Juntunen site (Brandt 1996), maize does not seem to have been an important component of diet at the Cloudman site at this time. Maize does, however, reappear in Ontario Iroquois vessel residues, occurring in half of the samples. Conversely, the frequency of both squash and wild rice increase through time. Wild rice underwent the greatest change in frequency, represented in only 5% of early Late Woodland samples but present in over half of the late Late Woodland and Ontario Iroquoian samples. Frequencies of maize, wild rice, and squash were compared between the primary analytic subsets 160 using a Kruskal-Wallis analysis of variance (Table 7.9). The only significant difference was between the early Late Woodland and the late Late Woodland subsets. Maize was more common than wild rice in the early Late Woodland samples, while maize was absent, supplanted by wild rice in the late Late Woodland period, signaling a drastic change in starchy foods utilized between these two occupations. Table 7.8: Frequencies of Microbotanical Species by Subset Middle Species Woodland Early Late Woodland Late Late Woodland Iroquoian Maize Wild Rice Squash Ct 5 2 2 % 42 17 17 Ct 5 1 3 % 26 5 16 Ct 0 4 3 % 0 44 33 Ct 2 4 2 % 40 80 40 Figure 7.10: Microbotanical Frequencies of Maize, Wild Rice, and Squash by Subset 161 Table 7.9: Microbotanical Frequency Relationships between Subsets (Kruskal-Wallis) Subset Middle Woodland Early Late Woodland Late Late Woodland (*p) --- 0.777 0.069 0.409 (p) --- --- 0.044 0.218 (p) --- --- --- 0.368 Middle Woodland Early Late Woodland Late Late Woodland Iroquoian *two-tailed α=0.05, df=2 Cluster analysis was used to explore associations between vessels with similar microbotanical content. Jaccard’s coefficient of association was applied to the data to evaluate within-group similarity (Figure 7.11). Unlike the lipid residue clusters, which were generalized and not temporally significant, the microbotanical clusters are distinct and show diachronic trends of resource exploitation and cooking, particularly when the groupings are displayed with corresponding interior carbonization patterns (Table 7.10). Cluster 1 consists of vessels with residues containing only maize microbotanical remains. These include only Middle Woodland vessels with interior carbonization patterns representing stewing, and early Late Woodland vessels with interior carbonization pattern indicative of boiling. This supports the observation that maize may have been more important at the Cloudman site during the Middle and early Late Woodland periods than in later occupations. Within Cluster 1, there is an apparent diachronic shift in the way that maize is cooked. Within Middle Woodland vessels, maize processing is associated with stewing; it may have been used primarily as flour/meal added to soups and stews to thicken them. In the early Late 162 Woodland, the vessels used for processing maize show signatures for boiling, suggesting that cooks at this time may have begun engaging in nixtamalization, a processing technique in which Figure 7.11: Dendrogram Showing Average Linkage (Between Groups) of Vessels by Microbotanical Content using Jaccard’s Coefficient (vessel numbers: left column; distance: right column) maize is boiled with hard wood ash, breaking down the pericarp and improving grain palatability (Lovis et al. 2011). This process can also mask maize signatures in stable isotope assays, explaining the lack of δ13C enrichment of Cloudman samples (Lovis et al. 2011). Cluster 2 includes vessels with residues containing only wild rice phytoliths, which is restricted to late Late Woodland and Ontario Iroquoian vessels, additional evidence for the increased importance of wild rice to the subsistence regime at the Cloudman site post-AD 1200. 163 As in Cluster 1, there is an associated diachronic shift in wild rice cooking styles. Among late Late Woodland vessels, only stewing was employed in processing of the rice (and other Table 7.10: Vessel Clusters by Microbotanical Species Content (Jaccard’s Coefficient) Vessel Period Type Maize Wild Rice Squash IC Pattern 1 1 1 1 1 1 0 0 0 0 0 0 0 0 1 1 1 1 0 0 0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1 0 0 0 0 0 0 0 0 0 0 1 1 1 1 0 0 1 1 2 2 1 1 1 1 2 2 3 3 1 3 3 3 2 2 2 3 Cluster 1 Cluster 2 4 MW 109 MW Laurel Dentate Stamped Laurel Banked Linear Stamped 120 ELW Mackinac Banded 122 ELW Mackinac Ware (cf. Punctate) 173 ELW Mackinac Banded 174 ELW Mackinac Ware (cf. Punctate) 101 LLW Juntunen Ware 204 LLW 146 IRO 162 IRO 23 MW Juntunen Linear Punctate Early Ontario Iroquoian cf. Sidey Notched or Lawson Incised Laurel Banked Linear Stamped Mackinac Undecorated Cluster 76 ELW 3 105 ELW Mackinac Punctate 75 LLW Traverse Plain v. Scalloped (mini) Cluster 4 5 8 MW Laurel Pseudo- scallop Shell ELW Mackinac Ware Cluster 40 IRO 5 70 IRO cf. Lawson Opposed or Methodist Point Group 7 cf. Huron Incised 164 contents), which is surprising given that wild rice requires extensive boiling before it is edible. Users of Iroquoian vessels, on the other hand, processed food mixes containing wild rice through a combination stewing/boiling method. Clusters 3 and 4 include vessels involved in the processing of squash or a combination of maize and wild rice, respectively. There are no obvious diachronic trends within these clusters besides the emergence of a culinary tradition combining maize and wild rice, a nutritionally- complementary food mix that emerged in the Middle Woodland, in accordance with previous observations (Boyd et al. 2014; Hart and Lovis 2013; Raviele 2010). This combination appears associated with stewing (Cluster 4) while the processing of squash is associated with boiling/stewing practices throughout the Late Woodland occupations (Cluster 3). Cluster 5 may be connected to an important historical culinary development. This cluster includes the only vessels with microbotanical evidence for maize, wild rice, and squash. The traditional “Three Sisters” subsistence regime characteristic of Iroquoian groups includes maize, squash, and beans, and is believed to have emerged post-AD 1300 (Bamann et al. 1992; Hart 2008; Kuhn and Funk 2000). Boyd et al. (2014) found that in areas where wild rice was available, beans tended to be less important because the two foods are nutritionally similar. The use of maize, squash, and wild rice together in vessels dating to post-AD 1400 signals the emergence of the Three Sisters culinary tradition at the Cloudman site with the opportunistic substitution of local wild rice for beans. Tandem Dietary Analysis Results In total, residues from 61 vessels were subjected to one, two, or all three dietary analyses (Appendix E). Twenty (20) vessels were subjected to all three analyses, resulting in one of the 165 largest and richest data sets of its kind. Residues from only 5 vessels lacked any identifiable foods remains, while an additional 14 contained signatures for only one food category (aquatic resources, in the case of 12 of these samples), and all 17 of these samples were subjected to only one or two of the analyses. The 20 samples subjected to all three analyses tested positive for at least two food groups, attesting to the combined analytic strength of these methods. It also highlights the multipurpose nature of the vessels used at the Cloudman site, which were employed to process a variety of foods. Figure 7.12: Dendrogram Showing Average Linkage (Between Groups) of Vessels by Lipid and Microbotanical Content using Jaccard’s Coefficient (vessel numbers: left column; distance: right column) 166 Table 7.11: Vessel Cluster by Lipid and Microbotanical Content (Jaccard’s Coefficient) Vessel Period 25 LLW Type Foods Juntunen Ware Traverse Decorated v. Punctate LLW 43 215 MLW Bois Blanc Ware 193 ELW Blackduck Banded Nut oil, Low fat content plants, Animal products, Aquatic resources To evaluate possible clustering of food remains inferred from microbotanical and lipid residue analyses, Jaccard’s coefficient of association was again used to evaluate within-group similarity (Figure 7.12). Stable isotopes were excluded from this evaluation because of the uniformity of the results. The analysis identified a single cluster of vessels from varying time periods containing generalized, rather than specific, food categories (Table 7.11). The test did not associate maize, squash or wild rice with any other major food group. Although diachronic trends in the use of different wild and cultivated crops are apparent, there are no associated trends for identifiable or specific food combinations. Seasonality at the Cloudman Site Cumulative results of the three dietary analyses provide a wealth of information about subsistence at the Cloudman site. The food categories most frequently identified at the site include high-trophic-level fish, nuts/acorns, wild rice, squash, and maize. As discussed in Chapter 3, both spring- and fall-spawning fish species were identified in faunal remains from Late Woodland and “Protohistoric” features (Cooper 1996), although the high-trophic levels of the fish as implicated by elevated δ15N values are more closely associated with deep-water, fall- spawning fish, such as whitefish. Fishing probably took place at the site during both the spring and the fall. Hilger (1959:125) observed that in the fall, fishing “became a seasonal occupation” 167 where groups relocated to fish specifically for winter storage. Fish “seemed to come near shore in November, just before the lakes froze” (Densmore 1979:125), the most intensive period of fishing routinely taking place in late fall. Archaeological plant remains indicate intensive use of the site during the late summer and fall. Hazelnut, butternut, and acorns were all present in the macrobotanical remains (Egan- Bruhy 2007) and nut lipids featured prominently in lipid signatures from the majority of vessels sampled. Hazelnut is harvested in August and September, acorn in September and October, and butternut in October (Yarnell 1964). Wild rice is harvested in late August or early September (Hilger 1959:147; Vennum 1988). Maize and squash are also harvested in the fall. Foods represented in pottery residues suggest intensive fall occupation of the Cloudman site, where residents lived from late August until November, using the site as a central residential locale from which logistical forays for hunting, fishing, and gathering activities could be conducted. Small-scale horticulture involving cultivation of maize and squash may have also taken place at the site. This does not preclude a spring occupation, as many species of fish and animals could be acquired in the spring, and nuts, maize, and wild rice are all easily stored for spring consumption and crops would require planting. Methodological Considerations The results of the chemical and microbotanical analyses of adhered and absorbed residues from Cloudman site pottery were highly complementary. Each of the three methods provides unique dietary data that are critical to the study, but inconsistencies between results highlight potential shortcomings of each method. The greatest disparities in outcome originate from lipid and isotopic signatures for fish, and microbotanical and isotopic signatures for maize. 168 Aquatic Resources, Acorns, Lipid Residue Analysis, and Stable Isotope Analysis The greatest discrepancy between the results of the lipid residue analysis and the stable isotope analysis lies in the detection of signatures for fish. As previously observed, the high nitrogen enrichment of the Cloudman pottery residues correlates well with signatures for high- trophic-level fish. Malainey and Figol (2018) identify the presence of medium fat content foods by a C18:0 level less than or equal to 25% and a C18:1 isomer level between 15% and 27.5% (see Appendix C). Both maize and fish are medium fat content foods, although fish usually display higher levels of C14:0 and lower levels of C12:0 and C:15. Only 17% (5 of 30) of vessel residues tested positive for medium fat foods (maize or fish), while 96% (48 of 50) of residue samples tested positive for aquatic resources through stable isotope analysis. Seventeen (17) vessels contained lipid biomarkers (specifically, cholesterol) for general animal product, which could be indicative of fish but cannot be positively associated with aquatic (or any animal) species. A total of 19 vessels with elevated δ15N values tested negative for medium fat lipids. However, fish are often smoked and/or dried and later reconstituted in soups (Densmore 1979), a process that may affect the lipid content of the fish upon cooking. High δ15N values could also be associated with aquatic plants (Chappuis et al. 2017; Cloern et al. 2002), which may correspond to low fat content plants detected in the lipid residues. An alternative method for identifying fish lipids was employed by Taché and Craig (2015), who use a suite of biomarkers, including isoprenoid alkanoic acids and ω-(o- alkylphenyl)alkanoic acids containing a minimum of 20 carbon atoms, as indicators for aquatic resources in absorbed lipid residue. They found these markers in approximately half of their samples of Early Woodland Vinette I pottery from northeastern North America. They note, however, that these biomarkers are vulnerable to degradation through exposure to the burial 169 environment and underrepresent the presence of aquatic resources in lipid residues. Fish may be difficult to capture in lipid signatures and may therefore be underrepresented by current methods of lipid residue analysis employed in North America. An additional discrepancy between the stable isotope and lipid residue analyses lies in the detection of acorns. Acorns produce very low δ15N values (Eerkens et al 2013), and high δ15N in human bone collagen and archaeological residues have been interpreted as diets dominated by fish and lacking beans, another low-15N foodstuff (Katzenberg et al. 1995; Morton and Schwarcz 2004; Schwarcz et al. 1985). The lipid residue analysis of the Cloudman pottery residues indicated the presence of nut/acorn lipids in 67% of the pottery vessels, the presence of which did not appear to affect the overall δ15N values of the adhered residues. The presence of nuts/acorns in the lipids adds another layer of complication to the analysis. Acorns produce very low which may be expected to lower the overall nitrogen enrichment of the Cloudman residues. This, however, did not appear to manifest in the stable isotopes. However, δ15N values are more heavily influenced by foods with higher proportions of dietary protein (Philips and Koch 2002; Schwarcz et al. 1985), and therefore δ15N values from high-protein sources, such as fish, would likely be more strongly expressed than those of lower protein foods, such as acorns (Boyd et al. 2008; Craig et al. 2007). In consideration of these factors, the effectiveness of stable isotope analysis for evaluating pottery residues and cooked food warrants further exploration. Only a small number of studies have investigated the stable isotope values of carbonized food mixes. Morton and Schwarcz (2004) conducted the primary study used for evaluating stable isotopes values of mixes containing maize, beans, and fish. Royer et al (2017) evaluated raw versus cooked fish and meat for the effects of various cooking techniques on the stable isotope levels, although these 170 were not evaluated in the context of food mixes. Hart et al. 2007 found that a variety of factors, including food mixes and preparations, can effect δ13C yields in food residues. Similar studies focusing on the effects of these variations on nitrogen stable isotopes have not been conducted, nor have residue experiments containing acorns and aquatic plants. Future experimental work investigating isotope values of cooked residues containing acorns, aquatic plants, and fish of varying preparations and food mixes is required to fully evaluate the expected stable isotope values of residues representing these foodstuffs. Until further investigation of these issues can be conducted, tandem application of the lipid residue analysis and stable isotope analysis may be the best method for identifying a broad range of resources in food residues. The effectiveness of using lipid residue analysis alongside stable isotope analysis for investigating aquatic resource processing in pottery has been demonstrated by Taché and Craig (2015) and Anderson et al. (2017). If, ultimately, lipid residue analysis underrepresents fish and stable isotope analysis underrepresents nut/acorns, as the results of this study suggest, then their corroborative employment will be necessary for future investigations of ancient diet. Maize, Stable Isotope Analysis, and Microbotanical Analysis The presence of maize in pottery residues is also obscured in the stable isotope analysis results. The mean δ13C of the Cloudman pottery residues is -26.65‰, and the highest delta value of any samples was -22.65‰ (see Appendix B). Hastorf and DeNiro (1985) interpreted δ13C values above -22‰ as indicative of the presence of maize in carbonized food remains, and all but one (V35 at -21.78‰) of the Cloudman samples fell below that threshold. More recent studies correlate maize with δ13C values closer to -17.4‰ in human bone collagen (Katzenberg and 171 Pfeiffer 1995; Muhammad 2010). Considering fractionation, expectations for maize in carbonized residues should be around -20‰, higher than the Cloudman residue values. Hart et al. (2003) also encountered adhered residues producing both low δ13C enrichment and maize phytoliths in pottery from the Finger Lakes region of New York. Hart et al. (2007) demonstrated that while the ratio of C4 to C3 foods affects delta values, the amount of carbon released by different foods during the cooking process is highly variable. They also found that dried maize and green maize release different amounts of carbon. This variability makes it difficult to accurately quantify a delta value threshold for residues containing maize, and Hart et al. (2007) therefore recommend against using stable isotope analysis as an independent measure for the presence of maize in archaeological residues. The results of the Cloudman residue analysis support this conclusion. A total of 48 pottery residues were tested for both stable isotopes and microbotanical remains. The average δ13C value of residues lacking maize microbotanicals (n=35) is -25.59‰; the average δ13C value of residues containing maize microbotanicals (n=13) is -25.60‰, a nearly identical value. In the case of the Cloudman pottery assemblage, the amount of maize in food mixes was not sufficient to elevate carbon stable isotope values which are therefore not predictive of maize presence or absence. Discussion The application of three different analytic methods for evaluating foods cooked in ancient pottery provides a robust depiction of precontact indigenous cuisine while highlighting the shortcomings of each method. Important foodstuffs for occupants of the Cloudman site include acorns/nuts, high-trophic-level fish, maize, wild rice, and squash, in addition to various wild 172 plants and terrestrial animals. Exploitation of acorns and aquatic resources appears intensive and consistent throughout the occupational history of the site, while the intensity of maize, wild rice, and squash consumption varied over time. The most unexpected finding was the timing and intensity of maize consumption at the site. Maize was most frequent in residues from Middle Woodland cooking vessels and continued to be common through the early Late Woodland period, but it is absent from all food signatures yielded by late Late Woodland vessels. The intensification of maize at coastal sites during the northern Great Lakes late Late Woodland period has been posited by O’Shea (2003), but evidence from the Cloudman site does not support this hypothesis at the local level. The lack of maize at the Cloudman site during the late Late Woodland is not representative of overall regional trends—several studies have found evidence of moderate maize consumption in late Late Woodland diets in the Upper Great Lakes (Brandt 1996; Muhammad 2010). Instead, it signals a possible change in the use of the Cloudman site rather than a macroregional dietary shift. Conversely, the presence of wild rice microbotanicals in pottery residues increases in frequency over time. Although present in the Middle Woodland samples, it is not as abundant as maize within the same subset. Wild rice microbotanicals all but disappear in the early Late Woodland period, occurring in only 5% of samples. Among pottery constructed after AD 1200, the frequency of wild rice in residues dramatically increases, present in nearly half of the late Late Woodland samples and 80% of the Iroquoian samples. The Cloudman site is located at the mouth of the Potagannissing River and down river from contemporary wild rice patches, and although the antiquity of these patches is unknown, the microbotanical remains indicate that wild rice has grown in the area for some time. 173 The early Late Woodland period occupation of the Cloudman site occurred during the Medieval Climatic Optimum (AD 900-1000), a warm period during which lake levels dropped, causing riverine fluctuation and flooding (Lovis et al. 2012; Monaghan and Lovis 2005). Wild rice is very sensitive to climate and water levels (Boyd et al. 2013; Vennum 1988), so water- level fluctuations may have disrupted local rice beds while the warmer climate would have been conducive to maize horticulture. The end of the Optimum brought cooler weather and caused higher lake levels and streams grading to higher elevations, thereby reducing riverine fluctuations and allowing stable aquatic environments required for the re-establishment of productive wild rice beds (Lovis et al. 2012; Monaghan and Lovis 2005). During the late Late Woodland period, the Cloudman site, located downriver from productive modern wild rice paddies, would have been a prime base locale for logistical forays exploiting the renewed abundance of this resource. At the same time, the site may not have remained a viable locale for continued maize cultivation, or the expense of energy invested in growing maize or acquiring it through social relationships was too great. Instead, late Late Woodland occupants of Cloudman site supplemented their diet with nuts/acorns, which appear in nearly 90% of lipid signatures of the late Late Woodland vessels sampled. The focus on wild rice exploitation may also signal a shift in food-related identities at a time of increased group localization and identity, as manifested by the proliferation of distinct pottery styles during the late Late Woodland period (McHale Milner 1991, 1998; O’Shea and McHale Milner 2002). While maize may have grown poorly at the Cloudman site at this time, its availability and/or appeal may have also changed in the altering social landscape. Wild rice is closely related to the identity of modern Ojibwe identity (Densmore 1979, 2005; Scott 1996; 174 Vennum 1988), so as a locally abundant resource, late Late Woodland groups may have begun to root both their subsistence and identity to this sacred food. Maize microbotanicals appear again in Iroquoian vessel residues, as expected given the centrality of maize in Iroquoian diets. Whether the vessels present at Cloudman were used by Ontario Iroquois groups or by Late Woodland/proto-Odawa groups engaging in intensive trade relationships with the Ontario Iroquois, maize would have been a component of their diet. However, even among the Iroquoian vessels, wild rice was more common than maize. Wild rice is not mentioned in ethnohistoric accounts of Iroquoian/Huron groups (Tooker 1991; Waugh 1973), nor is it generally associated with the Odawa (Scott 1996; Smith 1996). Wild rice must have been abundant enough at the site for it to become an important resource for all local groups using the site post-AD 1200. In aggregate, the microbotanical and lipid evidence ultimately supports Dunham’s (2014) hypothesis that wild rice and acorns grew in importance in the late Late Woodland period, and that maize was not as significant a resource during the same period for groups occupying the Eastern Upper Peninsula. Synchronic culinary trends associated with identity were not evident from the results of the dietary analyses. Cluster analyses of food types inferred from lipid and microbotanical data did not reveal groupings of foods or food combinations within analytic subsets. However, the only vessels in which maize, wild rice, and squash were found together were in Ontario Iroquois vessels. This could be a local variety of the Three Sisters food tradition that later formed the foundation of Iroquoian cuisine, which, although is important to modern Anishinaabe/Algonquian groups, has been historically more closely associated with the Iroquois, and is the strongest evidence for cuisine-related identity at the Cloudman site. 175 The results of the dietary analyses also have important methodological implications. Lipid analysis of Cloudman pottery residues detected medium fat foods (indicated by C1:0 isomer ratios) in 17% of samples, while stable isotope analysis detected aquatic resources in 96% of sampled vessels. These high delta nitrogen values may represent aquatic plants but fall closer to values expected from high-trophic-level fish. Smoking or drying fish prior to cooking in pottery vessels may affect its lipid content, so this method of lipid analysis may therefore underrepresent fish, as do other methods using lipid biomarkers (Taché and Craig 2015). If fish are anticipated in food signatures from archaeological pottery residues, lipid residue analysis may best be employed in cohort with stable isotope analysis. Stable isotope analysis of charred food residue, however, has also been found problematic when independently employed for evaluating maize processing. The use of maize in varying amounts was predicted at the Cloudman site, where maize macrobotanicals were identified from Late Woodland features (Egan-Bruhy 2007). Microbotanical analysis identified maize starches and phytoliths in pottery residues from the Middle Woodland, early Late Woodland, and Iroquoian pottery subsets, while stable isotope analysis yielded δ13C values well below those expected for residues containing maize. As Hart et al. (2007) demonstrated, stable isotope analysis is not a reliable indicator of maize in archaeological food residues, yielding false negatives unless large amounts of maize are present in food resource mixes. Acorns may similarly become masked in stable isotopes results; nuts, which are low in nitrogen, were not represented in the high δ15N values of the Cloudman pottery residues, possibly because protein- rich foods contribute more to nitrogen values than resources with less protein (Philips and Koch 2002; Schwarcz et al. 1985). Additional experimental replications are required to further clarify isotopic values of cooked fish, aquatic plants, and acorns in carbonized food residue. The results 176 of this study demonstrate that tandem employment of lipid residue, stable isotope, and microbotanical analyses of pottery residues is not only recommended but necessary for accurate depictions of ancient diet. Conclusion Results of microbotanical and chemical (stable isotope and lipid residue) analyses of adhered and absorbed food residues from ancient pottery contribute rich insight into dietary and culinary habits of groups occupying the Cloudman site. Nuts (mostly acorns) and aquatic resources (likely high-trophic-level fish) were important dietary staples of people living at the site from its earliest occupation (ca. AD 100) until the contact period. Maize, wild rice, and squash were also consumed throughout the occupational history of the site, although in varying intensity. Maize was more frequently present in food residues associated with the earlier occupations of the site (AD 100-AD 1000), all but disappearing from residues in the late Late Woodland, in a near-inverse relationship with wild rice, which increases in frequency through time. Whether external social relationships, environmental change, or culinary preferences changed (or some combination of these reasons) is unknown, although the environmental effects of the Medieval Climatic Optimum post AD-900 may have affected resource abundance and selection. A local variant of the “Three Sisters” cuisine may have developed at the Cloudman site post-AD 1200, when maize, wild rice, and squash are found together in residues from Ontario Iroquoian vessels. These results support Dunham’s (2014) finding that exploitation of starchy foods, particularly wild rice and acorns, was intensified in the Eastern Upper Peninsula of Michigan in the late Late Woodland period. However, if Cloudman is considered representative of the larger 177 region, there is little evidence to support intensive maize agriculture at northern Great Lakes coastal sites during the late Late Woodland period (O’Shea 2003). There is also no clear evidence in support of an increased reliance on deep-water spawning (high-trophic-level) fish during the Late Woodland period per Cleland (1982); although the exploitation of aquatic resources, likely including high-trophic-level fish, appears consistent at the Cloudman site throughout its occupational history, the data is not fine-grained enough for evaluation of the relative dietary proportion of these resources at any given time. The novel combination of lipid residue, stable isotope, and microbotanical analyses for assessing foods cooked in ancient pottery has provided greater inferential resolution than could any one method alone. Each method has proven important for providing unique and vital information about ancient diet. There are also inherent and acknowledged limitations in the types of foods each method can identify, but their use together has highlighted additional shortcomings in their applications to evaluating residue content. This study should be used as the basis for developing guidelines and best practices for inferring ancient diet from pottery residues. The archaeological data summarized in this chapter has substantially increased our understanding of culinary habits of people who lived at the Cloudman site have been. However, it can be further enhanced through examination of ethnographic and ethnohistoric accounts of food selection and cooking. The next chapter will review observed culinary traditions of historic Native American groups living in the Upper Great Lakes region and discuss them in context with the outcomes of the archaeological data, discursive comparisons that will augment the inferential strength of interpretations of past cuisine. 178 CHAPTER 8 ETHNOGRAPHIC AND ETHNOHISTORIC ACCOUNTS OF DIET AND COOKING Introduction Interpretations of archaeological data can be supported and enhanced by comparisons with traditions and practices of historic or modern peoples by forming analogical connections between culinary behavior and its physical (i.e., archaeological) manifestations. This chapter reviews ethnographic and ethnohistoric accounts of Ojibwe and Iroquoian culinary practices and discusses dietary and cooking data from the Cloudman site within the context of observed indigenous traditions. Together, the ethnographic and archaeological information will be used to reconstruct precontact cuisine and highlight potential ethnic culinary differences in food preparation methods and recipes. Ethnographic analogy as a middle-ranging method for interpreting archaeological data has limitations and is here approached with recognition of its limits and biases (Johnson 2010). As the archaeological evidence demonstrates, diet and cooking are not static behaviors and are instead subject to change over time, although subsistence behaviors are also often conservative, with certain culinary traditions remaining constant through long stretches of time (Twiss 2012). Historic-period lifeways of indigenous communities may therefore resemble those of precontact groups inhabiting similar environments with comparable resources. These accounts also capture Native American life only after contact with Europeans and subsequently cannot fully reflect precontact lifeways, whether explicitly acknowledged by the ethnographer (e.g., Rogers 1962) or not (e.g., Densmore 1979; Hilger 1959). For example, most of the groups observed no longer manufactured or used pottery, which prevents analogous observation of cooking habits, as metal 179 cooking pots have different performance characteristics and ranges of functions than their ceramic predecessors. Access to local, traditional foods may have been limited to communities inhabiting reserve lands, while the introduction of Western (Euro-American) foods after and during contact with Europeans would have also impacted local foodways. Historical and ethnographic accounts are likewise biased, both in the information recorded by the observers, and the ways in which the information is conveyed. Despite these concerns, ethnographic data is informative when carefully examined in context with archaeological data. Ethnographic and ethnohistoric accounts of the historic-period Algonquian and Iroquoian groups provide additional details and support the archaeological data. Accounts of Ojibwe (Baraga 1976; Densmore 1979, 2005; Hilger 1959; Rogers 1962; Vennum 1988) and Huron/Iroquoian (Tooker 1991, Waugh 1973) lifeways were used for assessment of culinary habits and behaviors. Culinary habits as inferred from the archaeological data are compared to and re-assessed in context with behaviors observed in the ethnographic and ethnohistoric sources. Fish The evidence for possible habitual fish processing in pottery vessels from the Cloudman site was unexpected given the results of prior ethnographic and lipid residue analyses (and modern distaste for boiled fish). Previous ethnographic surveys of Ojibwe, Cree, Innu, and Iroquoian ethnographic and ethnohistoric literature concluded that boiling was among the least common methods of fish preparation (Kooiman 2012, 2016; Lovis and Hart 2015), supporting the lack of fish lipids in previous sampling of pottery from the south shore of Lake Superior (Malainey and Figol 2015; Skibo et al. 2009). The results of the stable isotope analysis of 180 Cloudman site pottery residues necessitated a more in-depth and broadly framed review of ethnographic and ethnohistoric accounts of Ojibwe and Iroquoian groups to better capture the range of traditional methods for fish preparation. One of the earliest accounts of Ojibwe life, from Rev. Frederick Baraga in 1847, claims that the L’Anse Chippewa of the Upper Peninsula “have no particular skill in boiling fish” (1976:64). In Chippewa Customs, Densmore (1979:42) notes that fish were either eaten fresh or stored by drying or freezing. When eaten fresh, the fish were commonly spit roasted, fried, or flaked and packed with sugar. However, “the head of fresh fish, especially suckers, were boiled and greatly liked,” and “fresh fish were boiled and the broth used;” when dried or smoked fish were needed for food, they were also boiled (Densmore 1979:42). Hilger notes in Chippewa Child Life that both fresh and smoked fish were important in early Chippewa diets, but that meat, fish and fowl “were boiled with cultivated vegetables, such as beans, corn, squash, and pumpkin, and with native ones, such as wild rice, wild potatoes, and tips of certain plants” (Hilger 1959:144). The Round Lake Ojibwe in northern Ontario mostly dried or smoked fish, but “fish head and intestine provide oil; boiled in water, oil rises to surface and skimmed off with a wood spoon” (Rogers 1962:C49). Species of fish other than whitefish were infrequently used as food (Rogers 1962:C47-48), indicating the taste preference for the high-trophic-level species reflected in the stable isotope values of pottery residues at the Cloudman site. Iroquoian groups also relied heavily on fish and processed them with a variety of cooking techniques. F. W. Waugh recounts how early explorers encountered “great kettles of Indian corn soup…with dried eels and other fish boiled in it” (1973:136). Among the preparations for fish observed directly by Waugh include boiled fish, fish soup, and fish and potato soup, although 181 fish were also fried, roasted, and dried (1973:137). Among the Huron of Ontario, fish was commonly dried in the sun or smoked but was also used “as a relish for their soup, especially in the winter,” and the “biggest and fattest fish” were boiled to extract the fat (Tooker 1991:64). According to these sources, boiling fish was practiced among the Ojibwe and Ontario Iroquois and is not in opposition to an interpretation of the stable isotope analysis results indicating habitual fish processing in pottery at the Cloudman site. Acorns The lipid analysis of absorbed pottery residue suggests that nut processing was a regular if not important function of pottery vessels at the Cloudman site. Among some Ojibwe groups, acorns of white oak (red oak was considered too bitter) were “boiled in hulls, cooled, hulled, and dried in the sun, [and] when needed they were crushed or pounded to meal, boiled with meat, and served as thick soup” (Hilger 1959:145). Densmore (2005) reported that other Ojibwe groups had an affinity for Quercus macrocarpa, or burr oak, which produce large acorns. These were gathered in the late fall and buried for use in winter or spring, but they were also cooked for immediate consumption, when they were roasted in ashes, boiled and mashed, or boiled, split open and “eaten like a vegetable” (Densmore 2005: 320). According to Waugh (1973:123), hickory nuts were the most esteemed nuts among historic-period Iroquoian groups, but they also routinely consumed walnut, butternut, hazelnut, beechnut, chestnut. Acorns were also popular, prepared by “first boiling them in lye made from ashes, in order to take from them their excessive bitterness (pp. 122-123). Nut meats were “pounded, boiled slowly in water, and the oil skimmed off into a bowl,” the oil then mixed with a variety of foods, used in ceremonial foods of the False Face society, or used as mosquito 182 repellent (Waugh 1973:124). Nut meats were also ground, sifted, and added to corn soup to “make it rich” (Waugh 1973:90, 124). Tooker (1991:62) mentions that among the Huron, acorns were boiled several times to “take away the bitter taste” and then consumed. Although acorns are never referred to as primary staples of either the Ojibwe or Iroquois/Huron, their ubiquity in lipid signatures and macrobotanical remains demonstrate their importance to the occupants of the Cloudman site. Acorns were cooked and eaten independently but also added to soups and stews as a thickener. As Hilger (1959) noted, acorns were boiled to leech the tannins before they were stored or consumed, so Cloudman may have served as a seasonal residential locale to which logistical parties of gatherers brought acorns for large-scale processing and/or consumption. Maize According to the microbotanical evidence, there is a long history of maize consumption at the Cloudman site. The Ojibwe considered maize a primary food staple, which they roasted in the husks, parched in hot kettles, dried and boiled, or incorporated into soup (Densmore 1979:39; Densmore 2005:319). Maize was also boiled in its shucks prior to ripening, then braided, and hung to dry; the dried maize was then ground into meal and used primarily for thickening soups, although it was occasionally made into breads (Hilger 1959:145). The Ojibwe also made hominy by boiling maize in hardwood ashes (nixtamalization), rinsing, and boiling again (Densmore 2005:319). Iroquoian groups developed a maize-centric cuisine over the course of several centuries, resulting in over 40 methods of corn preparation (Waugh 1973). The Huron (Wendat) preferred boiling maize to roasting it, and commonly used it as the primary ingredient in soups, which 183 constituted many of their meals (Tooker 1991:68). Waugh (1973) lists pages upon pages of Iroquoian maize preparations. Maize was routinely nixtamalized, rinsed, hulled, and pounded into a flour. Corn bread was common and was boiled rather than baked; the resulting broth was then used to make corn soup (Waugh 1973:84). Both maize nixtamalization and cooking processes are enhanced by extensive boiling, the preferred cooking method for the grain. During nixtamalization, maize would have been the sole content of a pottery vessel, but in all other preparations, corn was cooked with other ingredients, most commonly in soups and stews. Of the thirteen vessels from the Cloudman site associated with residues containing maize microbotanical remains, maize always co-occurs with another food or food group. This tradition of cooking maize primarily with other foods may have extended back to its earliest incorporation into the northern Great Lakes subsistence regime. Wild Rice Wild rice, or manoomin, has long been a primary dietary staple of Ojibwe groups of the northern Great Lakes. An ideologically and ceremonially important food, it has transformed into a marker of Ojibwe identity (Vennum 1988) and has even been used to distinguish their cuisine from those of neighboring Odawa groups in archaeological studies (Scott 1996). Densmore (1979, 2005), Hilger (1959), and Vennum (1988) all identify wild rice as the primary staple of Ojibwe diet. Following harvest, wild rice was dried. Both Hilger and Densmore discuss the two primary methods of drying: air/fire drying, or parching. Hilger (1959:14) claims the divide in practices can be attributed to technology—that prior to European contact, wild rice was routinely spread on sheets of birchbark to dry in the sun, while after contact it was instead parched in 184 metal kettles. Densmore does not link parching to metal kettles, but does comment that drying has greater antiquity than parching: “The second [method] is undoubtedly the oldest process, and produced what was known as “hard rice”. This was greenish black in color, much darker than parched rice and requiring longer to cook. This rice could be kept indefinitely, and could be used for seed. In preparing “hard rice,” a frame was made similar to that on which berries were dried. It was covered by a layer of hay on which the rice, either on stalks or in the husk, was spread to a depth of about 3 inches. A slow fire was kept burning beneath the frame. In this manner the rice was dried as vegetables or berries are dried (Densmore 2005:315). After drying or parching, which loosened the husks, the rice was pounded and stored in birchbark makuks (Densmore 1979:148). To prepare it for consumption, wild rice was boiled in water or broth (Densmore 1979:39) or boiled in soups (Hilger 1959:148). It was also boiled in water and eaten “with or without maple sugar” or boiled with meat (Densmore 2005:319). Sometimes meat or fish broth was poured over fresh parched rice and allowed to “steam” until softened, while so-called “hard rice” was stored with dried blueberries in the winter and cooked together in the spring (Densmore 2005:319). Some historical accounts record contact period groups preparing wild rice in the form of a gruel, which required a greater liquid-to-rice ratio than nonglutinous (fluffy) rice (Vennum 1988:47). Rice was also “used thicken broths including venison, bear, fish, and wildfowl,” cooked into a bread-like paste, or pounded into flour (Vennum 1988:48). Wild rice is cited by 185 Vennum (1988) as being incorporated into a variety of stews with venison, small game, duck, or tassimanonny, a dish of boiled wild rice, corn, and fish. Wild rice is not mentioned in ethnohistoric accounts of the Iroquoian groups, suggesting it was not a primary resource for these groups post-European contact, but it is also not generally perceived of as an Iroquoian food staple, despite the fact that it grows regularly throughout traditional Iroquois territory (Boyd et al. 2013; Terrell et al. 1997). The high proportion of Iroquoian vessels containing traces of wild rice implicates opportunistic foraging habits of the users of these vessels (be they Odawa or Iroquoian, neither of whom are associated with wild rice consumption). Considering the proximity of the Cloudman site to wild rice stands, occupants of the site, despite cultural identity, wisely exploited the seasonal abundance of this resource. Like maize, wild rice phytoliths were only identified in vessels containing signatures for other foods. Although not entirely clear, ethnographic accounts seem to indicate that wild rice was sometimes cooked independently, as we cook rice today, but was more often cooked with a variety of other foods. Pottery vessels, then, were not constructed specifically for rice cooking, as is seen in other societies (i.e., Kobayashi 1994; Skibo 1994), but instead were all-purpose cooking pots. Regardless, wild rice requires long-term boiling to become palatable, necessitating cooking vessels with sufficient heating effectiveness and thermal shock resistance to withstand these processing demands. Late Late Woodland and Ontario Iroquois pottery from the Cloudman site with small temper particles would be technologically ideal for rice processing. Squash The persistent and consistent presence of squash was the most unexpected result of the food residue analyses at the Cloudman site. This is not because squash was unimportant for 186 indigenous groups historically, but because the antiquity of its use in the northern Great Lakes is relatively unknown and rarely discussed. Squash and pumpkins were in common use by 1847, when Frederick Baraga recorded them as one of the few crops grown by the L’Anse Ojibwe. The Ojibwe ate pumpkins and squash fresh, or they dried them for use later in the winter (Densmore 1979, 2005; Hilger 1959). When consumed fresh, they were generally baked in coals (Hilger 1959:144). Neither Hilger nor Densmore mentions boiling fresh squash. Squash was cut into strips to dry, and dried squash and pumpkin were boiled with meat or maple sugar (Densmore 2005: 319; Hilger 159:144). If dried squash were boiled, it might gradually break down and act as a thickening agent, potentially resulting in the dual boiling/stewing interior carbonization patterns associated with squash phytoliths in Cloudman cooking vessels. The Iroquois incorporated squash into a more complex cuisine. They, too, dried, boiled, or baked squash and pumpkins in coals (Tooker 1991:71; Waugh 1973:114). Squash/pumpkin and corn were “frequently combined in the preparation of food” (Waugh 1973:87), including corn and squash or pumpkin bread (into which berries were also mixed), dried pumpkin hominy porridge, and a sort of boiled pudding made of pumpkin/squash and hominy. Like maize and wild rice, squash was never the sole foodstuff cooked in the sampled vessels from the Cloudman site. Although squash has long been associated with maize as two of the “Three Sisters,” it is not found in context with maize except in two Iroquoian vessels, suggesting a relatively late adoption of the Three Sisters culinary tradition in the region, or signaling a switch in the cultural (and culinary) identity of the occupants of the Cloudman site alongside the new, distinct pottery tradition post-AD 1400. 187 Other Foods Also integral to the Ojibwe and Iroquoian diets were meat, berries, some vegetables, and maple sugar, although the degree to which these foods could be distinguished in the food residues is variable. Lipid signatures for large herbivores and medium-sized mammals were present in a few vessels. Most vegetables/greens (including aquatic plants) and berries would be considered low fat content plants but cannot be identified to a more specific level. Maple sugar is not detectable by any of the methods employed in this study. Nonetheless, these foods were all important components of Ojibwe and Iroquoian diets and may have been incorporated into the cuisine of precontact Cloudman site occupants. The Ojibwe prepared meat much in the same way as fish: it was boiled; cut in pieces and spit-roasted on the fire; or cut in thin slices, dried in the sun or smoked over a low fire, and stored for later use, when it was typically boiled (Densmore 1979:43; Hilger 1959:148). Fresh meat was often cooked with green vegetables, dried berries, and/or wild rice (Hilger 1959:148). Moose meat and rabbit bones were boiled to render fat and grease (Densmore 1979:44). The Iroquois employed similar yet slightly different methods of cooking meat. Large game was boiled twice, then removed from the pot and fried in grease (Waugh 1973:134). Meat and bones of bear, raccoon and porcupine were rendered for their grease, which was retained for medicinal purposes. Meat was also spit roasted or dried and rehydrated by boiling (Waugh 1973:134). Non- mammal species consumed by Iroquoian groups include frogs, snakes, and turtles (Waugh 1973:138). Berries were integral to Ojibwe and Iroquoian diets. They were regionally abundant (Baraga 1976:10) and added flavor to many dishes. Berries were eaten fresh but also extensively collected and dried, then boiled with other foods (Densmore 1979:40; Hilger 1959:144; Waugh 188 1973:126). Fresh berries were also boiled down and spread into little patches on birch barks sheets, leaving storable and transportable concentrated berry cakes that could be eaten raw or easily added into a pot of food (Densmore 1979:127; Rogers 1962:C52; Waugh 1973:127). Among the Iroquois, dried berries were often incorporated into cornbread (Waugh 1973:80). Blackberries and thimbleberries were combined with maple sugar and used in longhouse ceremonies (Waugh 1973:145). Vegetables are not emphasized in the ethnographic literature, but still played an important role in ancient diet nonetheless (Yarnell 1964). Roots, bark, and lichen are referred to as Ojibwe starvation foods by Baraga (1976:65), and Rogers (1962:C47) claims the collection of vegetal foods is of “limited importance” among the Round Lake Ojibwe. However, a wide array of vegetables is mentioned in other sources. Milkweed flower, woodbine bark, and white pine moss are “unusual vegetable foods” of the Ojibwe (Densmore 1979:40). Milkweed shoots and tips of ferns were boiled and flavored with grease by the Lac Courte Oreilles Ojibwe (Hilger 1959:146). Wild potatoes were eaten by both the Ojibwe and Iroquois, and were typically boiled (Densmore 2005:319; Waugh 1973:119). The Iroquois extensively used the vegetative parts of various trees and shrubs, which were “cooked like spinach” (Waugh 1973:117). They also used roots such as pepper root, burdock, and artichoke, which were boiled or fried, and bark was pounded and made into bread (Waugh 1973:118-120). A number of other vegetable foods present in archaeological and ethnographic record are detailed by Yarnell (1964). Aquatic plants may have contributed to the high δ15N values detected in the stable isotope analysis of the Cloudman pottery residues. Those consumed by the Ojibwe include arrowhead root, bulrush (cattail), and hog peanut (Densmore 1979, 2005; Yarnell 1964). The Iroquois were known to consume marsh marigold, watercress, yellow pond-lilly roots, and skunk cabbage 189 (Waugh 1973). Aquatic plants were exploited and consumed but do not appear to have been a significant component of indigenous diet in the Great Lakes region. While they may have been available on Drummond Island prior to contact, none of these species were present in the macrobotanical remains from the Cloudman site (Egan-Bruhy 2007). Maple sugar was an important component of historic Ojibwe diet. It served as one of the primary food seasonings, used with fruits, vegetables, cereals (including wild rice) and fish (Densmore 1979:123). It was also considered a snack, as “all forms of the sugars were extensively eaten as a delicacy” (Densmore 1979:39). Even various Iroquoian groups enjoyed maple sugar, adding it to hominy to make “parched corn meal” or sometimes even fermenting the sap for use as an intoxicant (Waugh 1973:141, 147). Unfortunately, standard lipid residue, stable isotope, or microbotanical analyses cannot distinguish maple sugar in pottery residues. The antiquity of maple sugaring has been hotly debated (Holman 1984; Holman and Egan 1985; Mason 1986; Mason and Holman 2000), so the timing of its incorporation into regional cuisine is unclear. Many of the foodstuffs present in the residues are late summer/early fall foods, suggesting the Cloudman site was used extensively in the fall, while maple sap is collected in the spring. Still, faunal and macrobotanical remains indicate some spring occupation of the site (Cooper 1996; Egan-Bruhy 2007), so some occupations may have taken place during sugaring season. Maple sugar is also easily storable, and, as an important flavoring agent and source of carbohydrates, it may have been brought to the site for consumption during the fall. Even if pottery vessels at Cloudman were not involved in maple sap processing, maple sugar may have been incorporated into recipes and meals cooked in the pots, leaving chemical signatures in the residues. Biomolecular analysis of organic 190 residues may be a line of future study for detecting maple sugar in archaeological pottery (see McGovern and Hall 2016). Cooking and Cuisine The historical focus of archaeological investigations into foodways has been the individual components of diet. Cuisine, or food culture, encompasses not only food choice but also food combinations and cooking styles (Twiss 2012). Patterning of carbonized food residues on Cloudman pottery vessels (and pottery from other sites in the northern Great Lakes) show some degree of diachronic variation, suggesting a change in cooking styles over time (see Chapter 6). Microbotanical analysis revealed a diachronic shift in maize and wild rice processing, yet connections between transformations of diet and cooking styles remain speculative. The archaeological food residues also prove that different types of foods were routinely cooked in the same vessels, but the dietary analyses cannot distinguish whether processing of distinct foods occurred together or in separate, sequential cooking events. Ethnographic observations of cooking behaviors, coupled with actualistic experiments, can inform and enhance inferential connections between diet and cuisine. Among both the Ojibwe and Iroquois, the following are specified as foods that are habitually boiled: corn, squash/pumpkin, fish, berries, meat, acorns, wild potatoes, roots, greens, and maple sap. Wild rice was also boiled by the Ojibwe. Among the Round Lake Ojibwe, “all food is prepared basically in one of two ways, either boiling or roasted on a spit beside a fire; boiling is more common and practically all species of mammal, bird, and fish are cooked in this way” (Rogers 1962:C53). The ubiquity of boiling as a cooking practice among post-contact 191 Ojibwe and Iroquois/Huron appears rooted in cooking traditions extending back to the Middle Woodland period (and probably earlier—see Speth 2015). “Stewing” was only mentioned as a cooking technique by Vennum (1988) in his discussion of wild rice preparations, although several other sources mention adding ingredients such as acorn flour, pumpkin blossoms, and corn silk to soups as “thickening agents” (Densmore 1979; Hilger 1959; Waugh 1973). Stewing, a cooking process involving water-content reduction, was practiced at the Cloudman site throughout its occupational history, but it was most commonly employed during the Middle Woodland occupation. Groups with distinct cultural identities and traditions may consume the same foods as other groups but maintain unique methods of food preparation. For the Ojibwe, “a typical meal comprised meat or fish, broth, rice with maple sugar, and dried berries prepared in some way” (Densmore 1979:40). Meat and fish were “boiled with cultivated vegetables, such as beans, corn, squash, and pumpkin, and with native ones, such as wild rice, wild potatoes, and tips of certain plants” (Hilger 1959:144). Foods were boiled, often multiple types together, but it is not clear if food was extracted from the cooking water and served, or if components of each dish were served together as soups. Stews incorporating wild rice were common (Vennum 1988). Among the Iroquois, “a very large proportion of… foods were evidently of liquid nature – numerous references to soups and broths made from ripe and unripe corn, beans, squashes, meats, and other materials” (Waugh 1973:79). Corn soup was particularly common, and modern iterations tend to be rather aqueous. There are reports of “great kettles of Indian corn soup, or thin hominy, with… fish boiled in it” (Waugh 1973:136). It is then of little surprise that evidence of boiling was commonly observed in Ontario Iroquois vessels. 192 However, another Iroquoian dish, sagamité, could be characterized as a stew or thick soup made by “pounding two or three handfuls of raw pounded [corn] meal which had not had the hull removed; put in an earthen pot full of water, boiled very clear and stirred to prevent meal from sticking to pot and burning; if available a small quantity of fish… or meat was added, sometimes pumpkin, too. If fish had been added, it was taken out and pounded very fine, without removing the bones, scales or entrails, and put back into the pot” (Tooker 1991:68). In this process, maize was first boiled, then other foods, such as squash/pumpkin and pounded fish, were added, which would thicken the soup. Sagamité preparation could result in the combined interior carbonization pattern (Pattern #3) indicative of both boiling and stewing, which was slightly more prevalent among the Iroquoian pottery than among late Late Woodland vessels (see Table 6.14). Both Ojibwe and Iroquoian cuisines included varieties of boiled foods and dishes. Although foods can be boiled in organic containers by either adding hot stones to the contents or by filling containers full and putting them directly over the fire (Densmore 1979; Holman and Egan 1985; Speth 2015; Wallis and Wallis 1955; Waugh 1973), the advent of pottery at the outset of the of the Woodland period would have facilitated greater incorporation of boiling into everyday culinary routines. The eventual adoption of metal cooking pots by Native American groups following the establishment of trade with Europeans may have led to an increase in boiling practices in the historic period, resulting in overrepresentation of the cooking method in the ethnographic record. Nevertheless, interior carbonization patterns present in the Cloudman pottery assemblage demonstrates that boiling practices were employed as early as the Middle Woodland period and became increasingly common through time. 193 Stewing, or long-term, low-heat cooking, was employed by the Ojibwe for certain preparations of wild rice and squash, which could result in interior carbonization patterns corresponding to stewing or boiling/stewing. In Middle, early Late, and late Late Woodland vessels, wild rice phytoliths are associated only with stewing interior carbonization patterns (see Table 7.10). Among Ontario Iroquois pottery vessels, wild rice is associated with the dual boiling/stewing pattern. Ojibwe and Iroquoian cuisines share many similarities while remaining distinct. Soup was a common dish among the Iroquois; Ojibwe cooks boiled food and consumed their meals with broth, ethnographers only occasionally refer to their dishes as “soups,” and wild rice stews were popular. Overall, cooking styles appear similar, as are cooking styles inferred from interior carbonization patterns of late Late Woodland and Ontario Iroquois pottery vessels. The Iroquois diet centered on maize, while wild rice was more important to the Ojibwe. However, both maize and wild rice were more frequent in residues from Ontario Iroquois vessels than in late Late Woodland vessels (see Table 7.7). The Three Sisters culinary tradition, including maize, beans, and squash, may be represented at the Cloudman site with wild rice serving as a substitute for beans. Maize, wild rice, and squash co-occur in residues only from Ontario Iroquois vessels, possibly representing a local expression of Three Sisters cuisine. However, because the identity of the group using the Iroquoian vessels at the Cloudman site is unclear, such an interpretation is merely speculative, as the Ojibwe and other Algonquian groups also routinely consumed these foods. 194 Conclusion The ethnographic and ethnohistoric records largely support the archaeological data for the food types selected and modes of cooking employed at the Cloudman site. Most of the primary food staples consumed by post-European contact Ojibwe and Iroquoian groups were represented in the pottery food residues, with the exception of maple sugar, which is not detectable through conventional residue analyses. Common cooking methods, particularly boiling, reflect interior carbonization patterns encountered in the ceramic cooking vessels, particularly those from later occupations of the site. Stewing as a cooking technique may have gradually fallen out of favor with northern Great Lakes cooks over time and was likewise a less common cooking method observed by ethnographers. Culinary distinctions are apparent between historic Ojibwe and Iroquois, although these differences cannot definitively distinguish ethnic identities of the groups that occupied the Cloudman site. Overall, the ethnographic and ethnohistoric evidence supports and enhances interpretations of the archaeological data presented in this study. Historic Ojibwe and Iroquoian resource selection closely parallel the archaeological food remains and chemical signatures discovered at the Cloudman site. Actualistic observations of indigenous preparations of the same suite of resources has proved a useful tool for connecting cooking styles with specific foodstuffs and food mixes. Reinforced by these data, interpretations of the results of this study can be reviewed and final conclusions summarized. 195 CHAPTER 9 CONCLUSIONS Introduction Application of a novel combination of analytic methods to the pottery assemblage from the Cloudman site has produced a robust body of data about precontact northern Great Lakes indigenous pottery technology, pottery use, and cuisine. This study employed taxonomic categorization and functional analysis of the Cloudman pottery assemblage in tandem with vessel-specific AMS dating, stable isotope analysis, lipid residue analysis, and microbotanical analysis of food residues associated with pottery. This is the largest sample subjected to this range of analytic approaches in the Great Lakes region, and the results, enhanced by ethnographic analogy, inform local site history, long-standing regional questions in the northern Great Lakes region, and methodological considerations for evaluating ancient diet through food residues associated with pottery. Context and Chronology of the Cloudman Site The occupational history of the Cloudman site was established through the taxonomic classification of pottery vessels and AMS dating of adhered pottery residues. Diachronic comparisons of the ceramic technical and use-alteration properties and outcomes of dietary analyses could not have been achieved without an established chronological framework. Initial taxonomic classifications of the Cloudman assemblage by Branstner (1995) were revisited and refined with accumulated new data. Middle Woodland, early Late Woodland, late Late Woodland, and Ontario Iroquois pottery types represented the most substantial occupations of 196 the site. Small subassemblages from the Middle Woodland/Late Woodland transition and the middle Late Woodland period were identified but were not large enough for inclusion in most analytic comparisons. Reassessment of the assemblage supported a majority of Branstner’s identifications and temporal assignations, and AMS dating allowed further refinement of the occupational history of the site. A Middle Woodland period occupation is represented by several varieties of Laurel ware. Two Laurel vessels yielded statistically different median AMS ages—cal AD 87 and cal AD 127—representing at least two distinct occupations at the site during this period. Several late Laurel and Middle/Late Woodland transitional vessels (ca. AD 500-700) signify brief uses of the site after the primary Middle Woodland occupations. The early Late Woodland occupation is represented by AMS dates from individual Mackinac Banded and Blackduck Banded vessels, which produced a mean pooled median age of cal AD 957. Five Bois Blanc vessels represent a small middle Late Woodland occupation. The late Late Woodland occupation is represented primarily by Juntunen wares typically manufactured between AD 1200-1400 (Lovis 2014) and Traverse wares associated with the period between AD 1100-1550 (Hambacher 1992). Reassessment revealed that the most recent occupation, originally considered a “Protohistoric” Odawa occupation of the site ca. AD 1630, may instead represent a more complex habitation history. Much of the Ontario Iroquoian pottery assumed to have been brought to Drummond Island by the Odawa likely pre-dates 1630, supported by an AMS date of cal AD 1433 for one Ontario Iroquois vessel. Trade beads associated with this early Protohistoric occupation were likely made in the latter half of the 17th century (Heather Walder, personal communication), meaning that trade items initially associated with Iroquoian pottery may instead date later than AD 1630. Most of the Ontario Iroquoian pottery present at the Cloudman site was 197 manufactured between AD 1400-1650 and represents a distinct occupation subsequent to that of late Late Woodland peoples, although some overlap may have occurred. European-manufactured trade items at the site are likely indicative of distinct occupations dating to post-AD 1650. Research Questions and Results The questions used as a framework for this research (detailed in Chapter 3) were explored using a unique suite of analytic methods. Interpretations of the data for each question are summarized below. A synthesis of the findings is provided in response to the overarching research question posed in Chapter 1, which includes a diachronic summative account of the social and subsistence sequences observed through pottery and food remains at the Cloudman site. Question 1: Are there differences in technical properties (i.e., thickness, temper size, rim diameter) among Middle Woodland, early Late Woodland, late Late Woodland, and Ontario Iroquoian pottery from the Cloudman site? If regional reliance on starchy foods increased during the Late Woodland period, as proposed by Dunham (2014), it was predicted that there would be changes in pottery vessel thickness and temper size to accommodate new food processing needs, specifically, intensive boiling to make foods like acorn, wild rice, and maize palatable. Tethered dietary and technological changes have been observed elsewhere in the Eastern Woodlands (Braun 1983; Hart 2012). Observed technical changes include thinner pottery walls to increase heating effectiveness, and/or temper size reduction to improve thermal shock resistance. 198 Contrary to expectations, average vessel wall thickness increased through time at the Cloudman site. Middle Woodland vessels were, on average, the thinnest subassemblage, although once thickness was corrected for vessel size, Middle Woodland vessels were no longer significantly thinner than pots from later time periods. Vessel necks and shoulders were thickest among late Late Woodland vessels, a technical choice more likely associated with the need to support the heavy collared rims characteristic of the time rather than heating effectiveness, a structural or engineering decision rather than one associated with food preparation. Body thickness would be the most sensitive indicator of technical decisions to increase heating effectiveness, but this property did not significantly alter over time. Sample sizes for body thickness were small because of the low incidence of body sherds that could be confidently associated with identified vessels, affecting statistical outcomes. Temper size proved the most significant and informative technical property of the Cloudman pottery vessels. Particle size remained relatively constant among Middle Woodland and early Late Woodland pottery but underwent significantly reduction among late Late Woodland and Ontario Iroquoian vessels. Decreased temper size can improve the overall strength of a vessel, a property appealing to potters constructing the collared wares characteristic of both late Late Woodland and Iroquoian societies, but it also increases thermal shock resistance, a durability required for vessels employed in intensive and/or long-term cooking episodes required for processing starchy foods, such as wild rice, acorns, and maize. Prior diachronic studies of ceramic technology have not included vessel size as a factor, and its use in the study was largely exploratory. Using rim diameter as a proxy for size, Middle Woodland vessels were found to be significantly smaller than those constructed after AD 700. Since people in the Middle Woodland were more mobile and may have traveled in smaller 199 groups (Brose and Hambacher 1999), the diminutive vessel size may be connected to group size or the need for more transportable vessels. If later occupations included larger aggregates of people and their durations of stay were more prolonged, construction of larger vessels would have been warranted. The increase in vessel size also co-occurs with the proliferation of boiling signatures, so vessel capacity may also be associated with some aspect of cooking effectiveness, although the connection is unclear at present and requires experimental work for clarification. Question 2: Are there diachronic changes in ceramic vessel use and cooking habits evident through use-alteration traces? Pottery use at the Cloudman site is characterized by both consistency and change. Pottery vessels from all occupations were primarily involved in cooking over fire, evidenced by high frequencies of vessels with interior carbonization across all subassemblages. Distribution of interior carbonization patterns also demonstrated that pots from all time periods were routinely filled to the top during the cooking process, a culinary habit seen at Woodland sites across the Upper Peninsula of Michigan (Kooiman 2012, 2015b, 2016). Interior carbonization patterning, however, varied over time, representing alterations in cooking behaviors. Patterns indicative of stewing were most frequent among the Middle Woodland pottery assemblage, while boiling patterns increased in frequency among early Late Woodland, late Late Woodland, and Ontario Iroquoian subassamblages. This corroborates previously observed trends among Woodland pottery assemblages in the northern Great Lakes (Albert et al. 2018; Kooiman 2012, 2016). Differences in interior carbonization pattern frequencies between the Cloudman subassemblages were not statistically significant, but the small sample size may have affected the outcome. 200 An interior carbonization pattern not previously observed in the northern Great Lakes is present within the early Late Woodland, late Late Woodland, and Ontario Iroquoian subassemblages (see Figures 6.2 and 6.5). This pattern represents distinct signatures for both boiling and stewing in the same vessel. It was not observed in any Middle Woodland vessels and is therefore a cooking habit apparently adopted by northern Great Lakes peoples after AD 600. The boiling/stewing pattern could indicate boiling of starchy foods followed by water-reduction cooking (as with cooking rice) or later incorporation of other foods to create a thick soup or stew. The overall increase in the frequency in boiling and boiling/stewing patterns, when contextualized in data from other sites in the northern Great Lakes, signals a shift in cooking styles following the end of the Middle Woodland period. Boiling became a more common cooking technique, and while stewing was still employed, it diminished in importance. Question 3: Are there diachronic changes in subsistence strategies (and possible attendant changes in cooking habits) detectable through lipids, stable isotopes, and microbotanical remains extracted from pottery? Results of lipid residue and stable isotope analyses revealed that some aspects of diet remained consistent throughout the occupational history of the Cloudman site. Lipid residue analysis revealed that a majority (67%) of ceramic cooking vessels sampled from all subassemblages were involved in nut processing. Based on the proximity of the Cloudman site to precontact oak stands (Comer et al. 1995, 1997), the prevalence of acorns in microbotanical remains at the site (Egan-Bruhy 2007), and a preference for acorns among historic Ojibwe groups (Densmore 1979; Hilger 1959; Yarnell 1964), the nut lipids most likely represent acorns (although hazlenut and butternut were also probably consumed at the site). Nut lipids were 201 increasingly present in pottery residues throughout the Late Woodland period, appearing in almost 90% of late Late Woodland samples, but decreased in frequency among Ontario Iroquois pottery. Stable isotope analysis of Cloudman pottery residues yielded high δ15N values, representing aquatic resources, either fish or plants. While consumption of fish at the Cloudman site was expected, the processing of fish in pottery vessels was not predicted based on the absence of fish lipids in absorbed pottery residues from other sites in the northern Great Lakes (Kooiman 2016; Malainey and Figol 2015; Skibo et al. 2009). Enriched δ15N values were encountered in 96% of sampled vessels from the Cloudman assemblage, including vessels from all occupations. These values most closely resemble those of high-trophic-level fish, which would include species like whitefish and lake trout. The practice of boiling fish is chronicled in several ethnographic and ethnohistoric sources, corroborating this interpretation of the stable isotope results. Aquatic plants may have also contributed to the nitrogen enrichment of the pottery residues, although these were less important components of historic-period indigenous diet and are absent from the microbotanical remains recovered from the site. Whether floral or faunal, aquatic resources were consistently processed in pottery vessels at the Cloudman site throughout all occupations. Microbotanical analysis revealed diachronic disparities in foods processed in ceramic vessels. Maize was present in adhered food residues in Middle and early Late Woodland vessels with relatively greater frequency than in late Late Woodland and Ontario Iroquoian vessels. Conversely, wild rice increased in frequency over time. All but absent in the early Late Woodland, wild rice is present in over half of all late Late Woodland and 80% of Iroquoian pottery residues sampled. Differences in microbotanical frequencies between the early Late 202 Woodland and late Late Woodland samples were statistically significant, signaling a drastic shift in subsistence habits between AD 1000 and AD 1200. These shifts in resource selection could be associated with the Medieval Climatic Optimum (AD 900-1000), which caused both an overall warmer climate and lower lake levels, contributing to increased riverine fluctuations and flooding episodes that could destabilize established wild rice beds (Lovis et al. 2001; Lovis et al. 2012; Monaghan and Lovis 2005). Higher temperatures would have increased productivity of horticultural efforts while water level fluctuations adversely affected the availability of wild rice during the early Late Woodland occupation of the Cloudman site. Climatological cooling concurrent with the late Late Woodland period may have inhibited productivity of maize horticulture, and re-stabilization of riverine environments would have facilitated regrowth of wild rice beds, causing a reversal in resource availability and selection (Lovis et al. 2012). Maize starches and phytoliths, completely absent from late Late Woodland pottery residues, reappear among Ontario Iroquoian vessels, although only in context with squash and wild rice. Squash was processed and consumed throughout the occupational history of the site, but is most abundant in pottery residues associated with later occupations. Previous studies argued that exploitation of high-trophic-level, deep-water-spawning aquatic resources (Cleland 1982), maize (O’Shea 2003), and wild rice and acorns (Dunham 2014) intensified in the Late Woodland period, particularly post-AD 1200. Results of this study support the increased importance of wild rice but do not support the increased importance of maize at the Cloudman site during the late Late Woodland period. The results presented here also indicate consistent exploitation of aquatic resources and nuts/acorns from AD 100 to AD 1600. Increased intensification of aquatic resources, including high-trophic-level fish, during the Late 203 Woodland is neither supported nor disputed by data from the Cloudman site. Nut/acorn lipids are most frequent in late Late Woodland vessels, in accordance with Dunham’s (2014) findings, but nut lipids are present in over half of all samples during all time periods and were therefore important at the Cloudman site throughout all occupations. Question 4: Is there synchronic variation in ceramic vessel use, subsistence strategies, and cooking habits evident through use-alteration studies or detectable through lipids, stable isotopes, and microbotanical remains extracted from pottery of differing typological categories? Synchronic variation of technology and diet was difficult to assess given the limited size of the Cloudman assemblage. As hypothesized, the number of distinct taxonomic pottery types increased over the span of the Woodland occupations. The Middle Woodland subassemblage was dominated by Laurel wares, for which typological variation appears related to diachronic disparities rather than expression of group identity, according to the statistically distinct AMS dates between the Laurel Dentate Stamped and Laurel Pseudo-scallop Stamped vessels. Two North Bay vessels were the only non-Laurel ceramics in the subassemblage. The early Late Woodland subassemblage was only slightly more varied, comprised primarily of Mackinac wares, but also including Blackduck ware common to the north and Bowerman ware, which is local to the northern lower peninsula of Michigan. The late Late Woodland assemblage was the most varied, with Juntunen wares, a substantial number of Traverse wares, and numerous vessels that did not fit into existing taxonomic typologies. This supports the narrative that the social fluidity of Middle Woodland groups was gradually replaced with increased social localization of Late Woodland groups, with socially-distinct groups (using distinct pottery types) interacting at 204 seasonal aggregation sites during the Late Woodland period (Brose and Hambacher 1999; Carroll 2013; Cleland 1982; Dorothy 1978; McHale Milner 1998; McPherron 1967). Variation in technical and use-alteration properties and food residue content between contemporary pottery types was explored to identify unique dietary and cooking habits associated with social identity. However, given the relatively small sample sizes for each taxonomic type, clear patterns of pottery use, cooking, and diet were not apparent. A broader, regional survey of taxonomic types is required to accurately assess the relationship between cuisine and identity among precontact groups in the northern Great Lakes region. Question 5: How do ethnographic and ethnohistoric accounts of indigenous diet and cooking in the Great Lakes inform interpretations of ancient cuisine generated from the archaeological data? Culinary habits of historic Ojibwe and Iroquoian groups reflect many of the food traditions inferred from the archaeological record, particularly those of later occupations. Boiling foods was a common cooking mode among both the Ojibwe and Iroquoians, a practice that emerged as the predominant cooking technique during the Late Woodland period, as evidenced by interior carbonization patterns on pottery from Cloudman and other sites in the northern Great Lakes. Foods were generally cooked and/or served together as soups or multi-component dishes. The twenty Cloudman pottery vessels subjected to all three dietary analyses (lipid residue, stable isotope, microbotanical) all produced signatures for more than one food category, reflecting the multi-use nature of pottery and culinary traditions of groups occupying the site. Both the Ojibwe and Iroquois commonly consumed fish and nut/acorns, much as Cloudman residents did throughout history. Despite prior oversight of the prevalence of boiling 205 fish, review of the ethnographic and ethnohistoric sources revealed that fish were regularly boiled, reflecting the patterns of possible fish processing in pottery vessels used at the Cloudman site throughout precontact history. Nuts, and particularly acorns, were boiled for leaching purposes but also ground into flour and added to soups as a thickening agent. Maize was commonly consumed by both historic Ojibwe and Iroquois, although it was much more central to Iroquoian cuisine. Based on the observed frequencies of starches and phytoliths, maize may have been more common during the earlier occupations of the Cloudman site (AD 100-1000) than in post-AD 1200 occupations. By the Historic period, however, maize is ethnographically listed as one of the primary foods for both societies. Squash, which was grown and consumed by both the Ojibwe and Iroquois, was consistently present at the Cloudman site during all occupations. Wild rice was considered a staple food among the Ojibwe, although it does not seem to have been a staple among groups at the Cloudman site until the late Late Woodland period, although it was processed and consumed in small amounts during the earlier occupations. Wild rice was not reported as a component of historic-period Iroquoian cuisine. The presence of wild rice in food residues adhered to Ontario Iroquois pottery poses an intriguing question about the identity of the users of these vessels. Wild rice is not a foodstuff generally associated with the Odawa, either (Scott 1996; Smith 1996), so if occupants of the site post-AD 1400 were Iroquoian or proto-Odawa, they were opportunistically exploiting a locally abundant resource that may not have been part of their traditional cuisine. Residues from these same vessels also yielded a unique combination of maize, wild rice, and squash microbotanicals not observed in residues associated with earlier occupations, possibly reflecting a local variant of the Three Sisters, with locally abundant wild rice replacing the nutritional contribution of beans in the culinary triad. 206 Overarching Research Question: Do pottery technology, pottery use, diet, and cooking habits change over time, and if so, how do these changes relate to hypothesized transitions in settlement, subsistence, and social patterns among pottery-making groups in the northern Great Lakes region? Data derived from ceramic taxonomic classifications, pottery function analysis, dietary analyses, and ethnographic analogy successfully informed the discrete archaeological queries posed in this study, but when evaluated together they create a dynamic narrative of life at the Cloudman site. Middle Woodland Laurel peoples occupied the Cloudman site on at least two separate occasions, around AD 87 and AD 127. The highly mobile, socially fluid Middle Woodland groups made small pottery vessels which served the primary function of stewing foods. Pottery vessel size may have been restricted to facilitate mobility/transportability, to feed smaller groups of people, or to fulfill other cooking-related requirements. Diet consisted of aquatic resources, nuts/acorns, various wild animal and plant foods, including wild rice, and cultivated food such as maize and squash. While boiling was employed as a cooking method at this time, it was not as prevalent as stewing, a habit observed at other Middle Woodland sites in the eastern Upper Peninsula of Michigan (Albert et al. 2018; Kooiman 2012, 2016). Maize macrobotanicals are associated with stewing patterns of adhered food residues; maize at this time may have been incorporated into the diet primarily as a ground meal incorporated into soups as a thickening agent. Peoples inhabited the site after the primary Middle Woodland occupations, as represented by a few Late Laurel vessels (ca. AD 500-700), although use-alteration traces and food residues reflect general continuity with prior food-related behaviors. 207 At the outset of the early Late Woodland period, pottery vessels grew larger and were employed more frequently for boiling food than stewing. The larger vessels may reflect larger groups gathering at “persistent places,” locales repeatedly visited for extended occupations to exploit abundant nearby resources (Schlanger 1992; Thompson 2010), a trend observed among Late Woodland sites in the eastern Upper Peninsula of Michigan (Dunham 2014). The new social and settlement pattern accommodated manufacture and use of less-transportable materials that could be stored at a site for future use, aka “locale provisioning” (Lovis et al. 2005). The increase in vessel size is also concurrent with the proliferation of boiling practices, so expanded volume may have provided a functional advantage for this method of cooking. The variety of foods cooked during this period did not significantly differ from those processed and consumed during the Middle Woodland, so the impetus for transforming cooking habits lies elsewhere. Among early Late Woodland vessels, maize microbotanicals are associated with interior carbonization patterns for boiling, potentially representing the emergence of nixtamalization (boiling whole kernels in alkaline solutions) as a maize processing method in this region. Although early Late Woodland vessels increased in capacity, temper particle size did not decrease, as seen in New York, where thermal shock resistance became a premium physical characteristic because of increased dependence on maize (Hart 2012). However, early Late Woodland vessels are distinguished from all other subassamblages by a one morphological characteristic: constricted necks (and everted rims). Vessel neck constriction would increase the heating effectiveness of a vessel, allowing the larger vessels to heat up more quickly, remain hot by limiting the flow of air and steam in and out of the vessel, and, as a consequence, more effectively boil contents for prolonged periods of time. 208 People occupying the Cloudman site during the late Late Woodland period continued to manufacture large vessels (relative to Middle Woodland pottery). Distinct social groups, represented by greater ceramic stylistic diversity, aggregated at the site for intensive seasonal resource extraction. Potters selected temper particles that were significantly smaller than observed in earlier pottery. Smaller temper size, which improves increased thermal shock resistance, may have been a technical choice related to requirements for vessel durability during prolonged high-temperature cooking events. Manufacturers of Juntunen wares had abandoned the constricted necks characteristic of earlier Mackinac pottery for straight rim profiles and distinctive collars, a stylistic change that may have required a technological trade-off—the loss of temperature regulation provided by neck constriction required greater external heat application and the ability of pottery vessels to maintain integrity in these conditions. Frequency of wild rice phytoliths in carbonized residues increases during this same period. Wild rice requires long-term cooking to be made palatable and would require vessel built to withstand prolonged use over fire. Maize is unexpectedly absent from late Late Woodland pottery residues. The amelioration of the Medieval Climatic Maximum at this time would have led to a cooler climate with riverine stabilization, a reversal that would have decreased maize productivity while creating an advantageous growing environment for wild rice on Drummond Island (Lovis et al. 2012; Monaghan and Lovis 2005). A new abundance of wild rice, which grows downriver from the Cloudman site, may have undercut the value of maize output given the energy required to cultivate the crop. The late Late Woodland period also signals a time of greater social localization (McHale Milner 1991, 1998; O’Shea and Milner 2002). Distinct group identities manifested through pottery style may have also been expressed through food selection, 209 representing a conscious choice by late Late Woodland Cloudman occupants to exploit and associate themselves with locally-abundant wild rice. Microbotanical remains and lipid residues provide evidence for selection of wild rice and acorns in lieu of maize by late Late Woodland occupants of the Cloudman site. Wild rice is also found in context with stewing and boiling/stewing interior carbonization patterns on cooking pots. As boiling became a popular cooking method, the continuation of stewing practices may be connected to various modes of wild rice processing. The manufacturers of the Ontario Iroquoian (or Iroquoian-like) vessels also followed the straight-rimmed, small-tempered pottery formula of late Late Woodland vessels, while employing distinct methods of decoration application. Again, these were vessels built to hold a reasonable amount of food that could withstand long periods over the fire without breaking. The users of these vessels (which may or may not have been the same people who manufactured them) employed cooking techniques and cooked foods similar to those of the late Late Woodland occupants of the site. However, maize reappears in association with Iroquoian pottery. The onset of the Little Ice Age after AD 1300 may have affected wild food availability, driving groups to increased reliance on cultivated maize and squash. The choice may also be connected to identity- related culinary choice. This is the first reported subassemblage in which wild rice, maize, and squash microbotanical remains co-occur in pottery residues. The Three Sisters cuisine (maize, squash, and beans) that became synonymous with Iroquoian cuisine emerged after AD 1300 (Bamann et al. 1992; Hart 2008; Kuhn and Funk 2000). Wild rice has similar nutritional value to beans (Boyd et al. 2014), and the combination observed at the Cloudman site may represent a local variation of Three Sisters cuisine, with occupants taking advantage of abundant wild rice 210 stands. Ultimately, neither functional nor dietary analysis could clarify the identity of groups occupying the site post-AD 1400, a line of inquiry requiring further investigation. Methodological Importance Aside from contributions to the scholarship on regional food processing technology, diet, and cooking in the northern Great Lakes, this study yielded critical insights concerning methods for exploring ancient diet using food residues associated with pottery. At the outset of the study, it was clear that each method had defined limits for the types of data it could provide. Carbon/nitrogen stable isotope analysis of adhered carbonized food residues can detect terrestrial vs. aquatic food contents and distinguish C4 plants (i.e., maize) present in food mixes, but cannot identify other food categories or species. Lipid analysis of absorbed residues can identify general food groups, both plant and animal, but rarely allows for identification at the level of species. Microbotanical analysis of adhered residues can identify only a limited number of plant species known to preserve well in archaeological remains. This suite of analytic methods was complementary from the outset, but their application to the Cloudman pottery assemblage suggests that not only is it beneficial to use these methods in cohort, it may also be necessary. The lipid residue analysis results came back mostly negative for medium-fat foods, a category inclusive of both maize and fish. The presence of medium-fat signatures in only 17% of samples is low, particularly compared to the 27% frequency of maize in microbotanical samples and the presence of enriched δ15N values, possibly representing high-trophic-level fish, in 96% of residues sampled for stable isotope analysis. The lipid residue analysis used by Malainey and Figol (2018) and Taché and Craig 2015 may both underrepresent fish. Lipid residue analysis is otherwise a rich resource for dietary and cooking data, but if aquatic resources are expected, it 211 should be employed in tandem with stable isotope analysis, a relatively inexpensive method that can enhance explorations of diet at a variety of sites. Stable isotope analysis, however, should also be applied to ceramic assemblages with caution. Only a single sample (out of 50 total) yielded δ13C values high enough to be interpreted as containing maize by some standards (Hastorf and DeNiro 1985), although even this sample is not considered enriched enough to indicate maize by more current standards (Katzenberg and Pfeiffer 1995; Muhammad 2010). This is well below the frequency of samples containing maize microbotanicals (27%). Hart et al. (2007) have already noted that carbon yield of cooked foods is highly variable and that maize can be underrepresented in the isotope values of food mixes, yielding false negatives. The results presented here support their conclusion that stable isotope analysis should not be used as the sole measure for maize in carbonized residues. Additionally, acorns evident in the lipid residue analysis did not appear to affect the stable isotope results. Acorns are low 15N plants and the δ15N values of the Cloudman residues were consistently very high, so the interplay of carbon and nitrogen contributions to the stable isotope compositions of food residues requires further exploration, as does the effect of cooking on isotope values. This method is therefore most gainfully employed in tandem with microbotanical and lipid analyses. Finally, soil samples from the Cloudman site were subjected to both lipid residue analysis and stable isotope analysis to test for potential contamination of the pottery residue samples. Stable isotope signatures of the soil samples yielded low δ15N values relative to the level of nitrogen enrichment characterizing the Cloudman pottery residue samples, proving the site matrix did not contaminate the archaeological remains. Lipid residue analysis provided similar results; the sole sample with high “cultural” lipid saturation contained different ratios of fatty acids than those yielded by absorbed pottery residues (Malainey and Figol 2018, Appendix C). 212 Soil samples have also been collected for examination using microbotanical analysis, the results of which are forthcoming. The completed tests established that contamination of adhered and absorbed archaeological food residues from the burial environment and taphonomic processes is negligible. Future Research This study has laid the groundwork for future research by revealing aspects of the archaeological record requiring additional clarification. The most obvious issue affecting analysis of the Cloudman pottery assemblage was the ambiguity of the timing and nature of the site’s later, more recent, occupations. The Cloudman site occupies a unique geographical crossroads between the Upper Peninsula of Michigan, traditionally occupied by Woodland Algonquian groups, and southwestern Ontario, the territory of Iroquoian groups such as the Wendat, Petun, and Neutral (Trigger 1976). The presence of Ontario Iroquois pottery in areas outside of their traditional territory is commonly interpreted as the result of trade between local Algonquian groups and the Iroquoians, rather than the result of Iroquoian occupation of these sites (Fox and Garrad 2004; Guindon 2009). Such is the case at the Cloudman site, where the use of Ontario Iroquois pottery was attributed to Odawa traders occupying the site ca. AD 1630 (Branstner 1995). However, two Early Ontario Iroquoian vessels, likely dating to AD 1200-1300, and an AMS date of AD 1433 from another Iroquoian vessel, demonstrate an earlier presence of Iroquoian wares at the site than previously believed (although most of the Iroquoian pottery was probably manufactured and used after the late Late Woodland occupation of the site). Functional and dietary analyses revealed very few differences between the late Late Woodland and the Ontario Iroquoian assemblages, except for an absence of maize in late Late Woodland 213 carbonized food remains and its presence in combination with wild rice and squash in Iroquoian pottery residues. Whether this disparity is a function of time or identity-related cuisine cannot be distinguished. The exact timing of the late Late Woodland occupations requires further investigation by obtaining AMS dates from late Late Woodland Juntunen and Traverse wares and additional Iroquoian vessels to examine the time-depth and span of occupation(s) associated with both subassamblages. Compositional studies of the pottery itself could reveal clay and temper sources and determine whether all or some of the Ontario Iroquois vessels were manufactured locally or elsewhere. A mixture of foreign and locally manufactured vessels could support Branstner’s (1995) hypothesis that the Late Precontact occupants were Odawa (or proto-Odawa) who used traded Huron/Wendat pottery alongside their own imitation Iroquoian vessels. Revisiting original excavation notes could provide additional contextual clues about the Iroquoian pottery in relation to Woodland pottery and various protohistoric trade goods. The Cloudman assemblage was not large enough to provide useful insights about the relationship between cuisine and identity. Distinct culinary identities have been identified between archaeologically distinct groups and historic period tribes through archaeological and ethnographic data (Egan-Bruhy 2014; Scott 1996; Smith 1996). A regional survey of pottery and food residue contents of common local ceramic taxonomic types, particularly from the late Late Woodland and protohistoric periods, would facilitate more accurate examination of food choice and cuisine in relation to group identity. Expansion of the ethnographic and ethnohistoric survey, and the inclusion of sources chronicling the culinary habits of the Odawa, would be useful for refining the details of group-specific cuisine. The resulting data could be used to create standard 214 dietary patterns/signatures for use in exploring possible multi-ethnic protohistoric and historic period sites in the northern Great Lakes and surrounding regions. Actualistic experiments could answer lingering questions related to cooking techniques, pottery function, and interpretations of dietary analyses. Most interpretations of vessel size variation relate it primarily to the processing of different types of foods (Kobayashi 194; Rice 1987), social/ideological functions (Blitz 1993; Kooiman 2016; Potter 2000) or group/household size (Nelson 1981; Tani 1994; Turner and Lofgren 1966). Experiments testing the heating efficacy and other cooking-related capabilities of vessels of varying capacity could allow further insight into the function of vessel size in food processing. The effects of cooking on the stable isotope yields of different foods and food mixes has also been minimally explored and requires further investigation to improve interpretive accuracy. Cooking wild rice, maize, squash, fish, aquatic plants, and acorns in different forms and combinations and recording the resulting interior carbonization patterns and carbon and nitrogen stable isotope values could yield important information about the interpretation of cooking, diet, and ceramic use-alterations traces in the northern Great Lakes and beyond. Although re-analysis of the Cloudman ceramic assemblage has informed several questions about northern Great Lakes lifeways and methodological efficacy, it has ultimately created a greater number of avenues of future inquiry. Conclusion A multi-decade debate about the nature of Woodland settlement and subsistence has persisted in northern Great Lakes archaeological research, and this study has contributed new information to this topic. Archaeological remains from the Cloudman site revealed the deep 215 historical importance of fish, acorns, maize, wild rice, and squash to groups occupying the northern Great Lakes. Fluctuations in the intensity of wild rice and maize exploitation and variations in cooking methods and pottery technology reflect adaptive reactions to changes in the natural and/or social environments, reinforcing both the maintenance of food traditions and the adaptability and dynamism of human response to social and environmental change. This study has also demonstrated the efficacy of a multiproxy approach to pottery and culinary research. Taxonomic and functional pottery analyses intertwine time, identity, and pottery use patterns, resulting in a deeper understanding of ancient social and human- environment relationships The novel collaborative application of lipid residue, stable isotope, and microbotanical analysis, further informed by macrobotanical information and ethnographic research, to a large pottery sample culminated in a comprehensive picture of ancient diet and cooking and serves as a model for future research initiatives exploring past foodways in contexts across the world. 216 APPENDICES 217 APPENDIX A: Cloudman Pottery Data 218 Vessel Period Final Type Dates Original Type (Branstner 1995) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 MW MW MW MW MW MW MW Laurel Pseudo-scallop Shell Laurel Pseudo-scallop Shell Laurel Plain Laurel Dentate Stamped Laurel Pseudo-scallop Shell Laurel Dentate Rocker Stamped Laurel Pseudo-scallop Shell ELW Mackinac Ware MW MW MW MW MW MW MW MW MW MW MW MW Laurel Pseudo-scallop Shell (oblique) Laurel Pseudo-scallop Shell Laurel Pseudo-scallop Shell (oblique) Laurel Dentate Stamped (oblique) Laurel Dentate Stamped Laurel Pseudo-scallop Shell Laurel Pseudo-scallop Shell (oblique) Laurel Trailed Laurel Dentate Stamped (oblique) Laurel Pseudo-scallop Shell Laurel Dentate Stamped (oblique) Laurel Dentate Stamped (oblique) ELW Mackinac Punctate MW MW LLW LLW LLW Laurel Pseudo-scallop Shell Laurel Banked Linear Stamped "proto-Juntunen" Ware (plain) Juntunen Ware Juntunen Ware cal AD 80-214 (AMS) Laurel Pseudo-scallop Shell AD 100-300 AD 100-300 Laurel Pseudo-scallop Shell Laurel Plain cal AD 60-125 (AMS) Laurel Dentate-stamped AD 100-300 AD 100-300 AD 100-300 AD 700-1000 AD 100-300 AD 100-300 AD 100-300 AD 100-300 AD 60-125 AD 100-300 AD 100-300 AD 100-300 AD 60-125 AD 100-300 AD 60-125 AD 60-125 AD 700-900 AD 100-300 AD 100-300 AD 1200 AD 1200-1600 AD 1200-1600 219 Laurel Pseudo-scallop Shell Dentate Rocker Stamped Laurel Pseudo-scallop Shell Mackinac Phase/untyped Laurel Oblique/Pseudo-scallop Shell variety Laurel Pseudo-scallop Shell Laurel Oblique/Pseudo-scallop Shell variety Laurel Oblique/Dentate Stamped Laurel Dentate-stamped Laurel Pseudo-scallop Shell Laurel Oblique/Pseudo-scallop Shell variety Laurel Trailed Laurel Oblique/Dentate Stamped Laurel Pseudo-scallop Shell Laurel Oblique/Dentate Stamped Laurel Oblique/Dentate Stamped Mackinac Punctate Laural Pseudo-scallop Shell Laurel Banked Linear Stamped untyped untyped untyped Vessel Period Final Type Dates Original Type (Branstner 1995) LLW MW ELW untyped (late) Laurel Pseudo-scallop Shell Mackinac Ware MW/LW Late Laurel (cf. Laurel Incised) MLW cf. Bois Blanc Ware (expedient?) ELW Mackinac Ware AD 1200-1600 untyped AD 100-300 AD 700-1000 AD 500-700 AD 1000-1200 AD 700-1000 Laurel Pseudo-scallop Shell Mackinac Phase/untyped Laurel Incised untyped Mackinac Phase/untyped MW/LW Untyped (incipient Blackduck? MN type) AD 500-700? untyped ELW Mackinac Ware AD 700-1000 Mackinac Phase/untyped MW/LW Late Laurel (cross-hatched, cf. Laurel Incised) AD 500-700 Cross-hatched/Impressed lip cf. Huron Incised AD 1450-1700 Huron Incised IRO LLW IRO MW untyped untyped Laurel Ware (mini) cf. Lawson Opposed or Methodist Point Group 7 27 28 29 30 31 32 33 34 35 36 37 38 39 41 42 43 44 45 46 47 48 49 50 52 53 54 40/153 IRO ELW Mackinac Ware MLW Bois Blanc Ware NA NA AD 100-300 AD 1600-1700 AD 700-1000 AD 1000-1200 untyped untyped Untyped untyped/chevron; cf. Lalonde High Collar Mackinac Phase/untyped untyped LLW LLW LLW ELW LLW IRO ELW ELW ELW ELW LLW Traverse Decorated v. Punctate AD 1100-1550/1600 cf. Algoma ware/scalloped lip Juntunen Drag-and-Jab Juntunen Drag-and-Jab Mackinac Ware Juntunen Ware untyped Mackinac Ware Mackinac Banded Mackinac Undecorated (mini) Mackinac Punctate (mini) AD 1300-1400 AD 1300-1400 AD 700-1000 AD 1400-1500 NA AD 700-1000 AD 897-995 AD 800-1000 AD 700-900 untyped untyped Mackinac Phase/untyped untyped untyped Mackinac Phase/untyped Mackinac Banded Untyped Mackinac Punctate untyped (cf. O'Neill site cup) AD 1300-1650 Untyped 220 Vessel Period Final Type Dates Original Type (Branstner 1995) 55 56 57 58 59 60 61 62 63 64 65 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 ELW LLW LLW LLW MW LLW ELW IRO ELW IRO LW LLW IRO LLW IRO LLW ELW Mackinac Ware AD 700-1000 Mackinac Phase/untyped? Traverse Decorated v. Punctate AD 1100-1550/1600 cf. Algoma ware/scalloped lip untyped (cf. Juntunen Ware) AD 1200-1400 untyped Laurel Dentate Stamped (oblique) Juntunen Ware Mackinac Ware cf. Huron Incised untyped (mini) cf. Ripley Plain untyped (brushed) NA AD 100-300 AD 1200-1600 AD 700-1000 AD 1450-1700 AD 800-1000 AD 1000-1700 NA untyped untyped Laurel Oblique/Dentate Stamped untyped Mackinac Phase/untyped Huron Incised Untyped untyped untyped Traverse Decorated v. Punctate AD 1100-1550/1600 cf. Algoma ware/scalloped lip untyped NA untyped Traverse Plain v. Scalloped AD 1100-1550/1600 cf. Algoma ware/scalloped lip cf. Huron Incised untyped Mackinac Ware AD 1450-1700 Huron Incised NA untyped AD 700-1000 Mackinac Phase/untyped MLW Bois Blanc Ware AD 1000-1200? untyped IRO LLW ELW IRO ELW IRO ELW ELW IRO ELW cf. Huron Incised AD 1450-1700 Huron Incised Traverse Plain v. Scalloped (mini) AD 1100-1550/1600 cf. Algoma ware Mackinac Undecorated untyped Mackinac Banded cf. Huron Incised Mackinac Punctate Blackduck Banded cf. Huron Incised Mackinac Ware (mini) AD 800-1000 AD 1000-1700 AD 897-995 Mackinac Undecorated untyped Mackinac Banded AD 1450-1700 Huron Incised AD 700-900 895-988 AD AD 1450-1700 AD 800-1000 221 Mackinac Punctate Blackduck Banded untyped Mackinac Phase Vessel Period Final Type Dates Original Type (Branstner 1995) untyped Juntunen Ware (cf. Plain) Mackinac Ware Blackduck Banded Jununen Drag-and-Jab untyped Laurel Plain Mackinac Punctate Juntunen Ware Juntunen Drag-and-Jab Mackinac Banded NA AD 1200-1600 AD 700-1000 895-988 AD AD 1300-1400 NA AD 100-300 AD 700-900 AD 1200-1600 AD 1300-1400 untyped untyped Mackinac Phase/untyped Blackduck Banded untyped untyped Laurel Plain Mackinac Punctate jab-drag/cw object; untyped jab-drag; untyped cal AD 897-995 (AMS) Mackinac Banded Traverse Plain v. Scalloped AD 1100-1550/1600 cf. Algoma ware/scalloped lip Mackinac Punctate Mackinac Ware (cf. Punctate) Mackinac Banded Laurel Banked Linear Stamped Laurel Banked Linear Stamped Laurel Banked Linear Stamped Laurel Banked Linear Stamped Laurel Banked Linear Stamped AD 700-900 AD 700-900 AD 897-995 AD 100-300 AD 100-300 AD 100-300 AD 100-300 AD 100-300 Mackinac Punctate Mackinac Phase/untyped/punctate Mackinac Banded Laurel Banked Linear Stamped Laurel Banked Linear Stamped Laurel Banked Linear Stamped Laurel Banked Linear Stamped/Incised? Laurel Banked Linear Stamped? Juntunen Ware AD 1200-1600 Cross-hatched/Impressed lip LW LLW ELW ELW LLW LLW MW ELW LLW LLW ELW LLW ELW ELW ELW MW MW MW MW MW LLW ELW ELW 85 86 87 88 89 90 91 100 101 102 103 104 105 106 108 109 110 112 113 114 115 116 117 118 120 121 122 123 Mackinac Ware (cf. Undecorated) Blackduck Banded MW/LW untyped (cordmarked/undecorated) ELW ELW ELW ELW Mackinac Banded Mackinac Banded Mackinac Ware (cf. Punctate) Mackinac Punctate AD 800-1000 895-988 AD AD 500-700? AD 897-995 AD 897-995 AD 700-900 AD 700-900 222 Mackinac Phase/undecorated similar to Laurel and Blackduck banded Cordmarked/undecorated Mackinac Banded Mackinac Banded Mackinac Phase/punctate? Mackinac Punctate? Vessel Period Final Type Dates Original Type (Branstner 1995) 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 ELW ELW ELW ELW ELW ELW ELW MW ELW LLW LLW IRO IRO ELW ELW ELW LLW ELW MW LLW MW IRO IRO Mackinac Banded Mackinac Ware (cf. Undecorated) Mackinac Ware Mackinac Ware Mackinac Ware (cf. Undecorated) Mackinac Ware (cf. Undecorated) Mackinac Ware North Bay Linear Stamp Mackinac Ware (cf. Punctate) Juntunen Ware untyped cf. Ripley Plain (notched lip) cf. Lawson Incised or Huron Incised Mackinac Ware (cf. Banded) Mackinac Ware (cf. Punctate) Mackinac Punctate untyped Mackinac Ware (cf. Punctate) North Bay Cordmarked Juntunen Drag-and-Jab Laurel Ware Early Ontario Iroquoian (incised) Early Ontario Iroquoian (incised) ELW Mackinac Ware (cf. Undecorated) AD 897-995 AD 800-1000 AD 700-1000 AD 700-1000 AD 800-1000 AD 800-1000 AD 700-1000 AD 100-300 AD 700-900 AD 1200-1600 NA AD 1000-1700 AD 1450-1700 AD 800-1000 AD 700-900 AD 700-900 NA AD 700-900 AD 100-200 Mackinac Banded? Mackinac Phase/undecorated? Mackinac Phase/untyped Mackinac Phase/untyped Mackinac Phase/undecorated Mackinac Phase/prob. Undecorated Mackinac Phase/untyped untyped Mackinac Phase/punctate? untyped untyped untyped untyped Mackina Phase/banded Mackinac Phase/punctate Mackinac Punctate untyped Mackinac Phase/M. Punctate? untyped AD 1300-1400 Juntunen Phase/jab-drag AD 100-300 AD 1200-1300 AD 1200-1300 AD 800-1000 untyped Juntunen Phase/late? Juntunen Phase/late?/chevron Mackinac Phase/undecorated NA LLW LLW IRO untyped (late, peaked, smoothed, punctates) NA untyped Traverse Ware (lost) Traverse Plain v. Scalloped cf. Ripley Plain (trailed) AD 1100-1550/1600 cf. Algoma ware/scalloped lip AD 1100-1550/1600 AD 1000-1700 untyped untyped 223 Vessel Period Final Type Dates Original Type (Branstner 1995) 152 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 LLW Juntunen Ware (cf. O'Neil Curvilinear) IRO IRO IRO IRO cf. Lawson Opposed cf. Huron Incised cf. Huron Incised cf. Lawson Opposed or Methodist Point Group 7 MLW Bois Blanc Ware (cf. Braced Rim) IRO IRO IRO IRO IRO IRO cf. Huron Incised cf. Huron Incised cf. Huron Incised cf. Sidey Notched or Lawson Incised cf. Huron Incised cf. Huron Incised AD 1400-1700 AD 1400-1700 AD 1450-1700 AD 1450-1700 AD 1600-1700 AD 1000-1200 AD 1450-1700 AD 1450-1700 AD 1450-1700 cal AD 1420-1446 (AMS) AD 1450-1700 AD 1450-1700 untyped untyped "imitation" Huron incised Huron Incised untyped/chevron; cf. Lalonde High Collar untyped "imitation" Huron incised Huron Incised Huron Incised Huron Incised Huron Incised Huron Incised LLW Traverse Ware AD 1100-1550/1600 cf. Algoma ware/scalloped lip IRO IRO ELW MW cf. Huron Incised cf. Huron Incised (mini) Mackinac Ware (cf. Undecorated) Laurel Pseudo-scallop Shell MW/LW Late Laurel (cf. Laurel Incised) MW/LW untyped (vertical punctate) ELW ELW ELW ELW ELW LLW IRO IRO Mackinac Banded Mackinac Banded Mackinac Ware (cf. Punctate) Mackinac Punctate Mackinac Ware (cf. Undecorated) untyped cf. Huron Incised cf. Huron Incised AD 1450-1700 AD 1450-1700 AD 800-1000 AD 100-300 AD 500-700 AD 500-700? AD 897-995 AD 897-995 AD 700-900 AD 700-900 AD 800-1000 AD 1200-1600 AD 1450-1700 AD 1450-1700 224 similar to Huron incised Huron Incised Mackinac Phase/undecorated Laurel Pseudo-scallop Shell cross-hatched/linear stamped/incised vertical punctate Mackinac Banded Mackinac Banded Mackinac Phase/punctate Mackinac Punctate Mackinac Phase/undecorated? cf. Algoma ware/scalloped lip Huron Incised similar to Huron incised Vessel Period Final Type Dates Original Type (Branstner 1995) untyped Mackinac Ware cf. Huron Incised (mini) cf. Huron Incised untyped cf. Huron Incised cf. Huron Incised cf. Huron Incised untyped Late Mackinac Ware NA AD 700-1000 AD 1450-1700 AD 1450-1700 NA AD 1450-1700 AD 1450-1700 AD 1450-1700 AD 1000-1700 AD 1000 untyped Mackinac Phase/untyped Huron Incised Huron Incised untyped Huron Incised Huron Incised Huron Incised untyped pseudo-collared/exterior beveled lip Traverse Plain v. Scalloped AD 1100-1550/1600 cf. Algoma ware/poss. shell tempered Mackinac Punctate Mackinac Banded Blackduck Banded Late Mackinac ware AD 700-900 AD 897-995 Mackinac???? Mackinac Banded? cal AD 895-988 (AMS) NA AD 900-1000 Mackinac Banded? untyped, unknown time period NA NA NA NA NA NA NA LLW Mini Vessel? ? Mini Vessel? NA NA NA NA NA 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 LW ELW IRO IRO LW IRO IRO IRO IRO ELW LLW ELW ELW ELW ELW NA ELW ELW LW MW ELW LLW 197 198 199 200 201 202 204 205 206 207 208 196/203 ELW Mackinac Punctate MW/LW untyped Mackinac Ware (cf. Undecorated) cf Bowerman Plain v. Cordmarked AD 900 untyped (Generic Woodland, ELW/MLW) untyped (mini) (Hopewellian) Mackinac Ware (mini) Juntunen Linear Punctate LLW Juntunen Linear Punctate LLW LLW LLW Juntunen Linear Punctate untyped untyped AD 700-900 AD 500? AD 800-1000 NA NA AD 700-1000 AD 1200-1300 AD 1200-1300 AD 1200-1300 AD 1200-1700 AD 1200-1700 225 Vessel Period Final Type Dates Original Type (Branstner 1995) 209 210 211 212 213 214 215 216 217 218 LLW LLW LLW LLW LLW LLW Juntunen Linear Punctate Juntunen Ware Juntunen Linear Punctate Juntunen Linear Punctate Juntunen Ware (cf. Jab-and-Drag) Juntunen Ware (cf. Linear Punctate) MLW Bois Blanc Ware LLW LLW LLW untyped untyped AD 1200-1300 AD 1200-1700 AD 1200-1300 AD 1200-1300 AD 1300-1400 AD 1200-1300 AD 1000-1200 post-AD 1200 NA NA NA NA NA NA NA NA NA NA NA Juntunen Ware (cf. Linear Punctate) AD 1200-1300 226 Vessel Ct. Rim Radius (cm) % rim Lip Neck Thickness Thickness Shoulder Thickness 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 5 2 5 8 1 103 1 1 3 3 8 3 1 3 1 2 9 1 1 12 5 1 1 7 10 1 13 na 13 20 na 24 na na na na 14 18 na na 15 na na na na 16 17 na na 19 na 22 6 na 8 6 na 22 na na na na 10 8 na na 5 na na na na 23 21 na na 17 na 7 3.9 3.69 2.96 4.28 4.71 4.81 5.85 6.01 3.36 3.12 2.74 5.42 2.53 3.07 3.13 5.64 2.63 4.49 3.89 3.72 10.4 2.92 5.02 9.1 8.94 8.25 na na na 6.69 9.11 9.52 na 6.58 na 6.1 na na na 5.81 5.78 na na na na 6.5 5.65 na na 10.16 10.89 10.02 5.57 7.17 5.07 6.69 7.62 8.09 7.14 6.43 5.84 5.93 5.34 7.78 6.35 4.59 5.12 6.99 4.91 5.88 4.44 6.43 6.24 6.55 4.93 9.01 10.28 6.63 227 Body Temper Temper Temper Thickness 1 9.74 na na na 9.02 10.12 na na na na na na na na na na 4.62 na na 5.56 5.79 na na 8.41 8.62 na 0.58 1.1 1.23 2.45 1.18 3.03 1.32 1.7 1.05 1.37 2.39 1.46 1.44 1.8 1.09 1.74 0.88 1.56 na 1.48 1.47 2.21 1.51 1.58 1.32 1.14 2 1.38 1.15 2.46 1.08 1.12 1.34 1.38 1.39 1.69 2.21 1.94 1.43 1.08 2.49 1.45 1.69 1.36 0.7 na 1.96 0.78 1.9 1.21 1.03 1.78 1.12 3 1.37 0.98 1.27 1.99 1.28 1.83 1.14 0.98 0.9 1.61 1.16 1.16 1.21 0.99 1.15 1.12 0.93 1.26 na 1.14 2.27 1.2 1.64 1.61 1.14 0.84 Vessel Ct. Rim Radius (cm) % rim Lip Neck Thickness Thickness Shoulder Thickness 27 28 29 30 31 32 33 34 35 36 37 38 39 40/153 41 42 43 44 45 46 47 48 49 50 52 53 54 4 1 1 1 1 1 4 6 56 1 1 1 1 1 12 1 4 3 19 1 3 1 1 3 1 2 9 18 19 na na na na 24 27 22 17 na na na na 32 20 22 30 26 na na 15 14 25 5 9 8 7 7 na na na na 8 25 50 9 na na na na 20 6 7 7 7 na na 9 8 21 25 25 50 9.63 4.4 4.42 8.43 7.01 10.94 10.69 7.45 9.17 3.8 8.41 na 5.58 na 14.4 7.15 7.65 8.08 8.74 10.45 6.89 6.52 8.17 11.29 3.91 5.84 5.92 11.14 12.84 na na na na na 8.09 7.89 6.36 na na na 6.75 9.15 na na na na 7.75 na 9.41 na 4.3 5.44 4.93 3.21 7.13 8.37 3.11 8.04 6.85 6.98 8.62 5.25 5.5 6.96 9.61 na 6.2 na 6.62 4.31 7.09 12.03 14.62 5.66 14.5 7.64 3.83 6.06 3.89 6.39 5.22 228 Body Temper Temper Temper Thickness 1 na na na na na na na 8.06 6.59 na na na na 8.41 na na na na na na na na na na 4.14 na 5.32 0.9 2.08 2.02 2.36 1.66 1.16 1.33 1.2 2.52 0.71 1.16 1.06 3.07 1.32 0.89 1.26 1.03 1.58 1.07 1.06 1.19 1.27 1.29 1.4 na 2.17 0.64 2 3.11 3.76 0.59 1.35 1.08 1.52 1.53 0.96 1.11 0.73 na 1.75 1.22 2.11 2.64 1.1 1.22 2.32 1.56 1.12 2.4 0.95 0.62 1.77 na 1.4 1.22 3 1.39 1.41 0.92 0.98 1.81 0.76 1.91 2.25 1.61 0.52 na 1.42 1.4 1.34 1.65 1.42 1.15 1.11 1.6 0.82 1.15 1.02 1.08 0.82 na 1.49 0.63 Vessel Ct. Rim Radius (cm) % rim Lip Neck Thickness Thickness Shoulder Thickness 55 56 57 58 59 60 61 62 63 64 65 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 105 1 2 1 2 2 1 3 1 2 14 1 1 1 7 1 1 1 5 7 31 8 34 1 22 1 1 12 30 na na na na na na na 8 28 15 na na na na na 16 na 25 8 24 na 24 12 25 28 28 9 21 na na na na na na na 10 8 na na na na na 5 na 7 20 25 na 6 5 24 8 5 20 10.27 na na 7.14 3.39 5.15 5.84 na 9.38 6.33 9.36 8.05 7.04 7.76 8.34 7.41 9.35 4.88 8.43 5.29 8.8 9.8 8.4 3.45 14.03 11.15 5.33 5.03 7.67 na 6.85 8.97 4.55 4.22 5.67 na 6.76 6 11.46 6.09 7.62 7.77 7.77 5.56 5.46 8.92 8.86 5.48 8.06 8.93 7.04 5.02 6.01 7.87 7.03 4.38 229 na na 6.49 na na na na na na 5.03 11.29 na na na 6.16 na na na na 3.54 6.02 na 6.98 na 5.93 na 0 3.49 Body Temper Temper Temper Thickness 1 2.8 na 9.9 na na na na na na 0 na na na na na na na na na na 7.72 11.37 8.46 na na na 0 na 1.01 0.97 1.11 1.43 0.91 0.75 2.06 1.64 na 1.86 2.02 0.96 2.16 1.2 1.58 1.23 1.38 1.15 1.21 0.86 1.99 1.91 3.06 0.88 3.7 0.96 0.85 na 2 1.24 1.35 1.36 1.07 0.62 0.91 1.62 1.51 na 0.99 0.94 1.83 1.39 3.37 1.26 1.3 0.99 1.87 1.83 1.35 0.83 1.22 1.46 0.76 1.8 0.85 1.45 na 3 1.95 1.13 1.74 1.02 na 1.53 na 1.35 na 1.35 1.35 0.82 na 1.5 0.76 1.16 1.15 2.07 1.18 1.02 2.62 1.01 2.15 na 1.3 0.79 1.26 na Vessel Ct. Rim Radius (cm) % rim Lip Neck Thickness Thickness Shoulder Thickness 85 86 87 88 89 90 91 100 101 102 103 104 105 106 108 109 110 112 113 114 115 116 117 118 120 121 122 123 1 2 1 1 1 1 1 88 2 5 6 1 4 1 6 6 1 2 1 1 1 4 2 3 3 19 5 1 na 20 na 30 na na na 19 na na 14 13 19 na na 25 >18 na na na na 14 12 na 20 18 25 25 na 16 na 6 na na na 9 na na 10 10 20 na na 10 na na na na na 9 7 na 19 20 7 5 5.75 8 7.63 11.17 7.07 6.53 3.48 8.41 na 8.81 9.95 5.88 6.47 8.16 na 4.55 5.49 5.06 4.42 4.82 7.22 6.3 9.03 4.51 8.02 7.14 9.58 9.53 8.98 9.67 9.04 6.77 7.63 5.84 5.1 9.14 9.86 7.2 8.06 5.18 3.83 6.86 6.3 4.5 6.36 4.45 5.84 6.21 9.79 4.29 5.01 4.23 7.79 6.84 7.14 5.68 230 7.89 na na na na na na 10.89 na na na 7.1 3.57 8.63 6.05 4.85 na na na na na na 5.02 na 5.53 5.49 na na Body Temper Temper Temper Thickness 1 na na na 4.36 na na na na na na na na na na na 5.33 na na na na na na na na na na na na 0.74 1.46 3.18 2.62 1.13 1.37 1.17 3.02 1.09 1.9 1.85 1.04 1.58 1.25 2.42 2.18 1.5 0.77 2.06 1.96 2.48 2.32 1.37 1.09 2.32 5.58 1.18 1.9 2 1.06 1 0.89 1.42 0.82 0.89 1.54 0.66 2.72 1.46 1.22 0.9 1.33 2.02 1.75 1.75 1.4 1.15 1.42 0.95 1.63 1.03 1.35 1.68 1.05 2.92 0.82 1.14 3 1.02 1.48 2.15 1.84 1.01 0.84 1.74 1.53 1.42 1.93 0.78 1.58 1.12 1.21 0.96 1.82 1.19 1.82 1.5 1.16 1.41 2.01 1.24 1.02 1.45 1.97 0.88 1.18 Vessel Ct. Rim Radius (cm) % rim Lip Neck Thickness Thickness Shoulder Thickness 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 1 4 3 2 11 3 1 19 2 23 1 2 1 2 1 6 1 1 5 1 1 3 4 3 1 1 21 1 na 13 25 na 24 na 26 25 16 na na na na na 20 na na 19 20 na na 17 18 17 17 na 21 na na 6 6 na 7 na 5 5 10 na na na na na 9 na na 6 9 na na 6 16 17 16 na 15 na 10.64 5.28 10.42 8.58 9.25 8.82 10.19 7.6 8.71 9.11 9.89 5.35 7.3 6.62 10.63 9.24 5.81 9.65 9.16 na 5.96 8.13 8.52 5.75 8.06 7.92 6.56 8.15 6.44 5.38 7.62 6.24 4.29 4.08 6.08 7.79 6.31 8.31 6.67 7.39 6.68 7.49 9.18 7.45 6.07 7.52 7.36 na 7.28 7.76 7.21 4.02 6.46 5.83 6.7 7.57 231 na na 7.48 na na na na na na na na na na na na na na 6.56 6.75 na na na 5.92 na na na 5.69 na Body Temper Temper Temper Thickness 1 na na na na na na na na na na na na na na na na na na na na na na na na na na 3.89 na 5.86 1.28 1.52 2.63 1.87 0.95 2.45 1.33 1.33 1.81 1.1 1.22 1.08 1.47 1.69 1.16 1.34 2.55 1.36 1.37 1.51 0.86 1.76 1.37 1.38 1.32 1.28 2.3 2 1.41 1.02 1.72 2.14 1.1 1.39 3 0.87 1.05 2.25 1.17 1.13 1.1 1.5 3.66 1.01 3.43 3.15 3.4 1.81 1.08 1.77 0.94 1.45 1.23 1.41 0.87 1.19 3 0.98 1.9 1.68 1.18 1.4 1.31 1.04 1.64 1.31 1.94 na 0.74 1.71 1.28 1.35 1.2 1.48 1.36 1.41 1.24 1.26 1.37 1.8 1.23 1.22 1 1.34 1.05 Vessel Ct. Rim Radius (cm) % rim Lip Neck Thickness Thickness Shoulder Thickness 152 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 2 2 3 3 1 1 1 3 2 5 1 14 1 48 1 1 1 8 1 2 12 6 19 1 2 1 1 na na 18 na na na na na na 27 na 21 na 28 na na 14 15 na 26 17 na na na na na 25 na na 10 na na na na na na 20 na 10 na 20 na na 5 5 na 7 23 na na na na na 7 7.35 na 13.06 6.51 na 5.95 6.62 7.1 5.26 6.8 7.39 7.62 4.52 6.22 5.74 5.45 3.68 3.49 4.34 10.31 10.49 3.92 9.52 8.43 7.82 5.83 8.06 7.95 na 8.63 8.48 na 8.17 8.57 8.36 6.92 7.63 10.01 8.9 5.86 6.6 6.61 5.12 5.15 4.12 5.58 7.97 8.01 4.06 3.44 4.26 6.56 5.29 9.2 232 na na 0.19 na na 0 na na na 7.63 na na 7.73 7.51 na na na 5.55 na 6.72 8.45 7.47 3.08 na na na na Body Temper Temper Temper Thickness 1 na na na na na 0 na na na 6.66 na na na 10.03 na na na na na na na na 6.38 na na na na 1.32 0.89 1.59 2.34 1.15 0.79 1.3 1.88 1.23 1.45 1.56 2.17 1.61 1.65 na 4.11 1.19 1.37 1.23 1.43 1.43 1.18 1.45 1.94 1.16 1.06 1.13 2 1.25 0.75 1.07 1.54 0.97 1.28 2.02 0.79 1.11 1.72 1.3 1.8 1.91 1.53 na 1.56 1.27 0.99 0.93 1.36 2.15 1.46 1.25 1.16 0.74 0.89 0.9 3 1.58 0.78 0.71 1.27 0.82 1.1 0.66 1.42 0.8 1.52 1.1 1 1.12 1.11 na 1.04 1.17 1.73 1.42 1.26 1.16 0.68 1.73 1.74 1.28 1.03 1.64 Vessel Ct. Rim Radius (cm) % rim Lip Neck Thickness Thickness Shoulder Thickness 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196/203 197 198 199 200 201 202 204 205 206 207 208 1 2 1 2 1 1 1 1 1 1 1 9 114 82 1 1 2 3 9 13 36 3 4 19 2 4 1 1 12 na 7 na na na na na na na na 26 16 20 19 na na na 26 na 10 5 6 26 16 30 na na 5 na 14 na na na na na na na na 23 21 22 7 na na na 6 na 18 20 27 19 20 6 na na 5.57 10.34 4.19 6.6 8.89 9.23 na 6.59 na 3.38 5.31 12.44 10.49 9.4 5.01 na 10.75 8.89 7.63 6.91 6.03 5.76 4.15 8.91 7.64 7.77 7.01 3.98 3.93 10.13 6.69 12.56 5.71 7.97 na na na 5.25 4.07 7.91 7.7 6.81 4.74 na 8.75 6.98 6.83 4.94 6.64 5.57 4.52 5.98 8.55 8.35 6.88 6.89 233 na na 5.7 na na na na na na na na 7.48 6.35 6.03 na na na na 7.23 na 9.43 6.74 2.78 4.15 5.19 na na 5.43 Body Temper Temper Temper Thickness 1 na na na na na na na na na na na na 5.27 6.14 na na na na na na 7.79 5.15 3.71 4.2 na na na na 1.45 0.77 0.74 1.04 1.27 1.35 1.29 1.57 0.8 2.77 1.2 1.89 1.32 0.84 1.02 1.98 2.21 0.98 2.18 1.51 3.47 0.95 na 1.38 1.05 1.67 1.2 1.52 2 0.79 1.6 na 0.64 1.01 1.58 0.76 1.47 0.65 1.31 na 1.75 1.81 1.21 1.05 0.68 0.87 2 1.12 1.3 3.13 na na 1.9 1.58 0.78 0.87 0.82 3 0.69 2.22 na 1.13 1.24 0.69 0.87 1.1 1.03 1.26 na 1.85 0.77 1.93 0.81 1.45 1.1 1.48 1.73 1.03 2.64 na na 1.07 1.17 1.37 1.27 1.29 Vessel Ct. Rim Radius (cm) % rim Lip Neck Thickness Thickness Shoulder Thickness 209 210 211 212 213 214 215 216 217 218 1 1 1 1 1 7 3 2 1 1 13 na na na na 8 na 30 na na 8 na na na na 10 na 5 na na 6.21 4.4 6.9 8.47 5.47 7.86 6.45 7.3 9.72 8.03 9.12 6.49 8.35 8.44 6.56 8.73 8.65 8.61 6.04 6.58 11.23 9.15 5.65 na 7.67 7.46 11.26 8.37 na na Body Temper Temper Temper Thickness 1 na na na na na na na 6.59 na na 1.29 1.45 3.18 1.18 1.22 1.21 1.19 1.47 0.82 2 2 1.39 0.7 0.81 1.31 0.67 0.83 1.51 1.31 1.24 1 3 1.44 0.86 1.16 1.28 0.88 0.9 1.13 1.33 0.85 1.64 234 Vessel Exterior Sooting Exterior Interior Carbonization Carbonization IC Type Attrition 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 2 na na 2 2 5 na 2 na 4 na 4 na na na na na na na 5 4 4 1 4 2 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 0 0 1 1 1 0 1 0 1 0 1 0 0 0 0 0 0 0 1 1 1 1 1 1 1 235 Vessel Exterior Sooting Exterior Interior Carbonization Carbonization IC Type Attrition 27 28 29 30 31 32 33 34 35 36 37 38 39 40/153 41 42 43 44 45 46 47 48 49 50 52 53 54 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 1 0 1 1 0 1 0 0 0 0 0 0 0 1 1 1 1 0 0 0 1 1 1 0 na 0 1 1 1 1 1 1 1 0 0 0 1 0 1 0 4 4 4 4 na na na 2 5 1 na na na 2 4 4 2 2 1 4 na na na 2 na 1 na 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 236 Vessel Exterior Sooting Exterior Interior Carbonization Carbonization IC Type Attrition 55 56 57 58 59 60 61 62 63 64 65 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 1 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 1 0 0 0 na 1 0 1 0 0 0 0 1 0 0 0 0 1 0 0 1 1 1 1 1 0 0 1 1 1 1 na na 5 na 5 na na na na 1 na na na na 3 na na 4 3 3 3 1 na na 5 4 4 1 0 na 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 237 Vessel Exterior Sooting Exterior Interior Carbonization Carbonization IC Type Attrition 85 86 87 88 89 90 91 100 101 102 103 104 105 106 108 109 110 112 113 114 115 116 117 118 120 121 122 123 0 1 1 1 0 0 0 1 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 1 0 0 0 1 0 0 0 0 0 1 0 0 1 0 0 0 0 1 0 1 0 0 0 1 0 0 1 0 1 1 0 1 1 1 0 0 1 1 1 1 1 0 0 1 0 1 0 1 0 1 0 1 1 0 1 1 1 4 na 5 4 4 na na 2 4 4 1 3 na na 2 na 1 na 5 na 5 na 4 1 na 1 5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 238 Vessel Exterior Sooting Exterior Interior Carbonization Carbonization IC Type Attrition 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 1 1 0 1 0 0 0 1 0 0 1 1 1 1 0 1 0 0 1 0 0 0 na 0 1 1 1 0 1 1 0 4 na na na 4 na na 5 4 4 5 na 4 na na 4 na na na na na 4 3 1 na 4 3 na 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 na 0 0 0 0 0 0 0 0 239 Vessel Exterior Sooting Exterior Interior Carbonization Carbonization IC Type Attrition 152 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 1 1 1 1 0 0 na 0 1 1 1 1 1 1 0 1 1 0 0 0 0 1 1 1 1 1 1 1 1 240 1 2 na na na na 4 4 4 3 4 4 na 5 1 na na na na 4 1 1 1 4 4 4 5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Vessel Exterior Sooting Exterior Interior Carbonization Carbonization IC Type Attrition 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196/203 197 198 199 200 201 202 204 205 206 207 208 0 0 0 0 0 0 0 0 1 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 na 1 1 0 0 0 1 0 1 1 0 0 1 na na na na na na na na na na na 2 3 3 4 na 4 5 na na na 2 na 2 3 na na 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 241 Vessel Exterior Sooting Exterior Interior Carbonization Carbonization IC Type Attrition 209 210 211 212 213 214 215 216 217 218 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 0 1 0 0 0 1 1 0 1 1 0 1 na na na 5 1 na 4 1 na 4 0 0 0 0 0 0 0 0 0 0 242 APPENDIX B: Cloudman Site Pottery Residue Samples for Microbotanical, Lipid Residue, and Stable Isotope Analyses 243 Table B1. Pottery Residue Samples Collected for Microbotanical Analysis (by Vessel) Vessel No. 1 4 5 6 10 12 22 23 28 109 112 35 8 34 41 50 76 80 81 88 103 105 120 122 124 132 173 174 175 193 42 215 25 26 43 75 Type Laurel Pseudo-scallop Shell Laurel Dentate Stamped Laurel Pseudo-scallop Shell Laurel Dentate Rocker Stamped Laurel Pseudo-scallop Shell Laurel Dentate Stamped (oblique) Laural Pseudo-scallop Shell Laurel Banked Linear Stamped Laurel Pseudo-scallop Shell Laurel Banked Linear Stamped Laurel Banked Linear Stamped Period MW MW MW MW MW MW MW MW MW MW MW MW/LW Late Laurel Ware ELW Mackinac Ware ELW Mackinac Ware ELW Mackinac Ware ELW Mackinac Banded ELW Mackinac Undecorated ELW Mackinac Punctate ELW Blackduck Banded ELW Blackduck Banded ELW Mackinac Banded ELW Mackinac Punctate ELW Mackinac Banded ELW Mackinac Ware (cf. Punctate) ELW Mackinac Banded ELW Mackinac Ware (cf. Punctate) ELW Mackinac Banded ELW Mackinac Ware (cf. Punctate) ELW Mackinac Punctate ELW Blackduck Banded MLW Bois Blanc Ware MLW Bois Blanc Ware LLW LLW LLW LLW Juntunen Ware Juntunen ware cf. Traverse Ware cf. Traverse Ware (mini) 244 Wt (g) 0.0110 0.0074 0.0061 0.0110 0.0026 0.0030 0.0027 0.0024 0.0064 0.0053 0.0026 0.0054 0.0106 0.0031 0.0047 0.0061 0.0052 0.0023 0.0103 0.0034 0.0043 0.0073 0.0049 0.0032 0.0064 0.0030 0.0077 0.0058 0.0055 0.0278 0.0066 0.0061 0.0053 0.0058 0.0041 0.0056 101 102 150 152 204 205 36 40 70 146 162 179 Table B1 (cont’d) Juntunen Ware Juntunen Drag-and-Jab Traverse Undecorated v. Scalloped Juntunen Ware (late, corded) Juntunen Linear Punctate Juntunen Linear Punctate cf. Huron Incised cf. Lawson Opposed or Methodist Point Huron Incised Early Ontario Iroquoian (incised) cf. Lawson Incised or Sidey Notched cf. Huron Incised LLW LLW LLW LLW LLW LLW IRO IRO IRO IRO IRO IRO 0.0165 0.0077 0.0232 0.0062 0.0041 0.0079 0.0051 0.0101 0.0072 0.0111 0.0133 0.0056 245 Table B2: Pottery Residue Samples Collected for Stable Isotope Analysis (by Vessel) Vessel No. 1 4 5 6 10 12 22 23 28 109 112 35 8 34 41 50 76 80 81 88 100 103 105 120 122 124 132 173 174 175 191 193 42 215 25 26 Period MW MW MW MW MW MW MW MW MW MW MW Type Laurel Pseudo-scallop Shell Laurel Dentate-stamped Laurel Pseudo-scallop Shell Laurel Dentate Rocker Stamped Laurel Pseudo-scallop Shell Laurel Dentate Stamped (oblique) Laural Pseudo-scallop Shell Laurel Banked Linear Stamped Laurel Pseudo-scallop Shell Laurel Banked Linear Stamped Laurel Banked Linear Stamped MW/LW Late Laurel Ware ELW Mackinac Ware ELW Mackinac Ware ELW Mackinac Ware ELW Mackinac Banded ELW Mackinac Undecorated ELW Mackinac Punctate ELW Blackduck Banded ELW Blackduck Banded ELW Mackinac Punctate ELW Mackinac Banded ELW Mackinac Punctate ELW Mackinac Banded ELW Mackinac Ware (cf. Punctate) ELW Mackinac Banded ELW Mackinac Ware (cf. Punctate) ELW Mackinac Banded ELW Mackinac Ware (cf. Punctate) ELW Mackinac Punctate ELW Mackinac Punctate ELW Blackduck Banded MLW Bois Blanc Ware MLW Bois Blanc Ware LLW LLW Juntunen Ware Juntunen ware 246 Wt. (g) 0.0132 0.0065 0.0099 0.0271 0.0045 0.0027 0.0028 0.0033 0.0099 0.0103 0.0026 0.0074 0.0034 0.0026 0.0026 0.0051 0.0043 0.0022 0.0045 0.0031 0.0070 0.0028 0.0034 0.0056 0.0023 0.0062 0.0030 0.0037 0.0051 0.0048 0.0108 0.0148 0.0048 0.0108 0.0028 0.0068 43 75 101 102 150 152 204 205 36 40 70 146 162 179 Table B2 (cont’d) cf. Traverse Ware cf. Traverse Ware (mini) Juntunen Ware Juntunen Drag-and-Jab Traverse Undecorated v. Scalloped Juntunen Ware (late, corded) Juntunen Linear Punctate Juntunen Linear Punctate cf. Huron Incised cf. Lawson Opposed or Methodist Point Huron Incised Early Ontario Iroquoian (incised) cf. Lawson Incised or Sidey Notched cf. Huron Incised LLW LLW LLW LLW LLW LLW LLW LLW IRO IRO IRO IRO IRO IRO 0.0036 0.0051 0.0137 0.0058 0.0067 0.0800 0.0056 0.0104 0.0045 0.0054 0.0098 0.0104 0.0069 0.0047 247 Table B3: Pottery Residue Samples Collected for Lipid Residue Analysis (by Vessel) Vessel No. 1 6 20 109 131 35 34 55 76 80 100 173 175 191 193 215 24 25 43 75 101 150 204 216 70 77 146 162 164 166 Period Type Wt. (g) MW MW MW MW MW Laurel Pseudo-scallop Laurel Dentate Rocker Stamped Laurel Dentate Stamped (oblique) Laurel Banked Linear Stamped North Bay Linear Stamp MW/LW Late Laurel Ware ELW Mackinac Ware ELW Mackinac Ware ELW Mackinac Undecorated ELW Mackinac Punctate ELW Mackinac Punctate ELW Mackinac Banded ELW Mackinac Punctate ELW Mackinac Punctate Blackduck Banded ELW MLW Bois Blanc Ware "Proto-Juntunen" Ware (plain) LLW Juntunen Ware LLW LLW cf. Traverse Ware cf. Traverse Ware (mini) LLW Juntunen Ware LLW LLW Traverse Undecorated v. Scalloped Juntunen Linear Punctate LLW Untyped LLW IRO Huron Incised Untyped IRO Early Ontario Iroquoian (incised) IRO cf. Lawson Incised or Sidey Notched Huron Incised cf. Huron Incised IRO IRO IRO 6.56 43.38 9.64 14.95 6.20 7.14 23.45 16.84 16.48 12.44 23.68 13.86 10.16 36.72 15.94 7.21 16.77 24.27 11.86 5.71 13.62 10.35 21.24 31.23 39.94 13.21 9.02 11.04 7.73 21.73 248 APPENDIX C: Carbon and Nitrogen Stable Isotope Analysis: Summary of Illinois State Geological Survey (ISGS) Reports 249 Pottery Residue Analysis Results Vessel Job # ISGS # Mass (mg) 15NAir 1 4 5 6 8 10 12 22 23 25 26 28 34 35 36 40 41 42 43 50 70 75 76 80 81 88 100 101 102 103 105 109 112 120 122 124 132 146 150 1261 1261 1292 1261 1292 1292 1261 1261 1292 1261 1292 1292 1261 1261 1292 1292 1261 1292 1292 1292 1261 1261 1261 1261 1261 1292 1292 1261 1261 1261 1261 1261 1261 1292 1261 1292 1292 1261 1261 EA008538 EA008539 EA008887 EA008540 EA008888 EA008889 EA008541 EA008542 EA008890 EA008543 EA008891 EA008892 EA008544 EA008545 EA008893 EA008894 EA008546 EA008895 EA008896 EA008900 EA008547 EA008551 EA008552 EA008553 EA008554 EA008901 EA008909 EA008555 EA008556 EA008557 EA008558 EA008559 EA008560 EA008902 EA008564 EA008903 EA008904 EA008565 EA008566 1.294 1.295 1.296 1.291 1.298 1.314 1.304 1.282 1.288 1.291 1.320 1.315 1.323 1.288 1.287 1.299 1.302 1.315 1.284 1.282 1.318 1.307 1.292 1.308 1.296 1.288 1.308 1.299 1.300 1.316 1.282 1.304 1.298 1.280 1.282 1.294 1.299 1.280 1.302 9.63 12.83 12.81 10.24 11.93 12.98 11.49 11.98 10.93 11.50 11.65 13.44 11.26 13.28 10.29 11.43 11.85 10.20 11.53 12.81 12.51 10.33 11.67 10.58 12.66 5.17 13.76 13.77 13.37 12.35 13.33 11.90 12.31 10.22 11.96 12.88 10.28 10.32 10.57 250 % N 4.70 5.24 6.56 5.87 7.47 5.61 5.47 5.07 3.35 2.84 3.22 5.61 2.94 6.67 3.83 3.63 4.38 3.93 5.76 7.58 3.57 4.36 5.71 3.04 6.50 3.60 8.28 6.20 3.97 7.63 5.40 3.17 2.39 3.85 3.41 5.86 6.17 2.72 5.13 13CVPDB -26.12 -26.36 -26.1 -23.21 -23.77 -26.53 -24.45 -26.66 -25.9 -26.65 -26.3 -26.8 -24.74 -21.78 -23 -26.46 -25.68 -26.3 -23.66 -25.01 -26.41 -24.49 -22.65 -22.99 -27.02 -28.36 -25.18 -27.04 -27.33 -24.65 -27.2 -24.03 -27.43 -26.84 -26.63 -26.56 -24.43 -29.13 -24.8 % C 47.05 51.17 57.76 46.80 61.40 52.72 45.68 48.24 51.29 36.43 49.39 54.97 33.08 44.24 47.22 43.80 36.95 55.05 45.53 54.24 40.59 46.70 51.75 22.25 55.70 39.10 53.18 54.75 45.19 51.03 49.32 25.80 39.34 52.22 40.42 61.81 49.37 54.03 56.66 Vessel Job # ISGS # Mass (mg) 15NAir 152 162 173 174 175 179 191 193 204 205 215 1261 1261 1261 1292 1292 1261 1292 1261 1261 1292 1261 EA008567 EA008568 EA008569 EA008905 EA008906 EA008570 EA008908 EA008571 EA008572 EA008907 EA008573 1.283 1.302 1.308 1.300 1.317 1.314 1.307 1.294 1.303 1.319 1.309 11.84 11.01 12.93 12.78 13.27 8.74 10.58 12.52 13.57 11.69 11.71 Soil Sample Analysis Results % N 4.53 3.69 5.19 6.57 8.04 4.73 6.63 5.66 4.58 1.93 8.45 13CVPDB -25.03 -25.31 -27.25 -27.34 -22.88 -25.29 -27.29 -25.61 -25.94 -28.89 -22.78 % C 43.56 46.78 50.53 60.69 57.86 47.84 60.24 52.92 34.89 51.07 48.97 Mass (mg) 29.997 44.994 29.992 30.009 15.005 15.002 29.997 29.991 10.002 30.001 60.002 15NAir 8.89 7.25 8.26 9.16 7.47 8.86 12.27 7.32 8.70 8.79 5.80 8.12 % N 0.27 0.06 0.30 0.16 0.86 1.10 0.74 0.41 1.50 0.76 0.06 0.07 13CVPDB % C 5.02 -22.91 -4.66 -18.08 -11.47 -23.46 -25.42 -25.08 -21.78 -25.67 -24.37 -5.72 -5.93 3.20 5.07 4.00 13.78 15.36 9.42 6.51 21.02 11.19 2.90 3.92 Sample ID Job # ISGS # T2-S1-2 T2-S2-2 T3-S1-1 T3-S2-1 T1-S2-3 T1-S2-2 T3-S2-3 T2-S2-3 T1-S1-2 T1-S2-1 T3-S2-2 T2-S2-1 1278 1278 1278 1278 1292 1292 1292 1292 1292 1292 1292 1292 EA008610 EA008611 EA008605 EA008606 EA008932 EA008933 EA008915 EA008916 EA008934 EA008918 EA008935 EA008936 60.000 251 APPENDIX D: Lipid Residue Analysis Report 252 Analysis of Lipid Residues Extracted from Archaeological Material from the Cloudman site, 20CH6. Prepared for Susan Kooiman Department of Anthropology Michigan State University 655 Auditorium Drive East Lansing, MI U. S. A. 48824 by M. E. Malainey, Ph.D. and Timothy Figol Department of Anthropology Brandon University 270-18th Street Brandon, MB Canada R7A 6A9 253 Thirty pottery sherds and three soil samples from the Cloudman site were submitted for Introduction analysis. Exterior surfaces were ground off the pottery to remove any contaminants then crushed. Absorbed lipid residues were extracted from the powdered sherds and loose soil with organic solvents. Lipid extracts were analyzed using gas chromatography (GC), high temperature GC (HT- GC) and high temperature gas chromatography with mass spectrometry (HT-GC/MS). Residue identifications were based on fatty acid decomposition patterns of experimental residues, lipid distribution patterns and the presence of biomarkers. Procedures for the identification of archaeological residues are outlined below; analytical procedures and results are then presented. The Identification of Archaeological Residues Identification of Fatty Acids Fatty acids are the major constituents of fats and oils (lipids) and occur in nature as triglycerides, consisting of three fatty acids attached to a glycerol molecule by ester-linkages. The shorthand convention for designating fatty acids, Cx:yz, contains three components. The “Cx” refers to a fatty acid with a carbon chain length of x number of atoms. The “y” represents the number of double bonds or points of unsaturation, and the “z” indicates the location of the most distal double bond on the carbon chain, i.e. closest to the methyl end. Thus, the fatty acid expressed as C18:19, refers to a mono-unsaturated isomer with a chain length of 18 carbon atoms with a single double bond located nine carbons from the methyl end of the chain. Similarly, the shorthand designation, C16:0, refers to a saturated fatty acid with a chain length of 16 carbons. Their insolubility in water and relative abundance compared to other classes of lipids, such as sterols and waxes, make fatty acids suitable for residue analysis. Since employed by Condamin et al. (1976), gas chromatography has been used extensively to analyze the fatty acid component of 254 absorbed archaeological residues. The composition of uncooked plants and animals provides important baseline information, but it is not possible to directly compare modern uncooked plants and animals with highly degraded archaeological residues. Unsaturated fatty acids, which are found widely in fish and plants, decompose more readily than saturated fatty acids, sterols or waxes. In the course of decomposition, simple addition reactions might occur at points of unsaturation (Solomons 1980) or peroxidation might lead to the formation of a variety of volatile and non- volatile products which continue to degrade (Frankel 1991). Peroxidation occurs most readily in fatty acids with more than one point of unsaturation. Attempts have been made to identify archaeological residues using criteria that discriminate uncooked foods (Marchbanks 1989; Skibo 1992; Loy 1994). The major drawback of the distinguishing ratios proposed by Marchbanks (1989), Skibo (1992) and Loy (1994) is they have never been empirically tested. The proposed ratios are based on criteria that discriminate food classes on the basis of their original fatty acid composition. The resistance of these criteria to the effects of decompositional changes has not been demonstrated. Rather, Skibo (1992) found his fatty acid ratio criteria could not be used to identify highly decomposed archaeological samples. In order to identify a fatty acid ratio unaffected by degradation processes, Patrick et al. (1985) simulated the long-term decomposition of one sample and monitored the resulting changes. An experimental cooking residue of seal was prepared and degraded in order to identify a stable fatty acid ratio. Patrick et al. (1985) found that the ratio of two C18:1 isomers, oleic and vaccenic, did not change with decomposition; this fatty acid ratio was then used to identify an archaeological vessel residue as seal. While the fatty acid composition of uncooked foods must be known, Patrick et al. (1985) showed that the effects of cooking and decomposition over long periods of time on the fatty acids must also be understood. 255 Development of the Identification Criteria As the first stage in developing the identification criteria used herein, the fatty acid compositions of more than 130 uncooked Native food plants and animals from Western Canada were determined using gas chromatography (Malainey 1997; Malainey et al. 1999a). When the fatty acid compositions of modern food plants and animals were subject to cluster and principal component analyses, the resultant groupings generally corresponded to divisions that exist in nature (Table 1). Clear differences in the fatty acid composition of large mammal fat, large herbivore meat, fish, plant roots, greens and berries/seeds/nuts were detected, but the fatty acid composition of meat from medium-sized mammals resembles berries/seeds/nuts. Samples in cluster A, the large mammal and fish cluster had elevated levels of C16:0 and C18:1 (Table 1). Divisions within this cluster stemmed from the very high level of C18:1 isomers in fat, high levels of C18:0 in bison and deer meat and high levels of very long chain unsaturated fatty acids (VLCU) in fish. Differences in the fatty acid composition of plant roots, greens and berries/seeds/nuts reflect the amounts of C18:2 and C18:33 present. The berry, seed, nut and small mammal meat samples appearing in cluster B have very high levels of C18:2, ranging from 35% to 64% (Table 1). Samples in subclusters V, VI and VII have levels of C18:1 isomers from 29% to 51%, as well. Plant roots, plant greens and some berries appear in cluster C. All cluster C samples have moderately high levels of C18:2; except for the berries in subcluster XII, levels of C16:0 are also elevated. Higher levels of C18:33 and/or very long chain saturated fatty acids (VLCS) are also common except in the roots which form subcluster XV. Secondly, the effects of cooking and degradation over time on fatty acid compositions were examined. Originally, 19 modern residues of plants and animals from the plains, parkland and forests of Western Canada were prepared by cooking samples of meats, fish and plants, alone or 256 combined, in replica vessels over an open fire (Malainey 1997; Malainey et al. 1999b). After four days at room temperature, the vessels were broken and a set of sherds analysed to determine changes after a short term of decomposition. A second set of sherds remained at room temperature for 80 days, then placed in an oven at 75C for a period of 30 days in order to simulate the processes of long term decomposition. The relative percentages were calculated on the basis of the ten fatty acids (C12:0, C14:0, C15:0, C16:0, C16:1, C17:0, C18:0, C18:1ω9, C18:1ω11, C18:2) that regularly appeared in Precontact Period vessel residues from Western Canada. Observed changes in fatty acid composition of the experimental cooking residues enabled the development of a method for identifying the archaeological residues (Table 2). It was determined that levels of medium chain fatty acids (C12:0, C14:0 and C15:0), C18:0 and C18:1 isomers in the sample could be used to distinguish degraded experimental cooking residues (Malainey 1997; Malainey et al. 1999b). Higher levels of medium chain fatty acids, combined with low levels of C18:0 and C18:1 isomers, were detected in the decomposed experimental residues of plants, such as roots, greens and most berries. High levels of C18:0 indicated the presence of large herbivores. Moderate levels of C18:1 isomers, with low levels of C18:0, indicated the presence of either fish or foods similar in composition to corn. High levels of C18:1 isomers with low levels of C18:0, were found in residues of beaver or foods of similar fatty acid composition. The criteria for identifying six types of residues were established experimentally; the seventh type, plant with large herbivore, was inferred (Table 2). These criteria were applied to residues extracted from more than 200 pottery cooking vessels from 18 Western Canadian sites (Malainey 1997; Malainey et al. 1999c; 2001b). The identifications were found to be consistent with the evidence from faunal and tool assemblages for each site. 257 Work has continued to understand the decomposition patterns of various foods and food combinations (Malainey et al. 2000a, 2000b, 2000c, 2001a; Quigg et al. 2001). The collection of modern foods has expanded to include plants from the Southern Plains. The fatty acid compositions of mesquite beans (Prosopis glandulosa), Texas ebony seeds (Pithecellobium ebano Berlandier), tasajillo berry (Opuntia leptocaulis), prickly pear fruit and pads (Opuntia engelmannii), Spanish dagger pods (Yucca treculeana), cooked sotol (Dasylirion wheeler), agave (Agave lechuguilla), cholla (Opuntia imbricata), piñon (Pinus edulis) and Texas mountain laurel (or mescal) seed (Sophora secundiflora) have been determined. Experimental residues of many of these plants, alone or in combination with deer meat, have been prepared by boiling foods in clay cylinders or using sandstone for either stone boiling (Quigg et al. 2000) or as a griddle. In order to accelerate the processes of oxidative degradation that naturally occur at a slow rate with the passage of time, the rock or clay tile containing the experimental residue was placed in an oven at 75C. After either 30 or 68 days, residues were extracted and analysed using gas chromatography. The results of these decomposition studies enabled refinement of the identification criteria (Malainey 2007). Using Lipid Distribution and Biomarkers to Identify Archaeological Residues Archaeological scientists working in the United Kingdom have had tremendous success using high temperature-gas chromatography (HT-GC) and gas chromatography with mass spectrometry (HT-GC/MS) to identify biomarkers. High temperature gas chromatography is used to separate and assess a wide range of lipid components, including fatty acids, long chain alcohols and hydrocarbons, sterols, waxes, terpenoids and triacylglycerols (Evershed et al. 1990, Evershed et al. 2001). The molecular structure of separated components is elucidated by mass spectrometry (Evershed 2000). 258 Triacylglycerols, diacylglycerols and sterols can be used to distinguish animal-derived residues, which contain cholesterol and significant levels of both triacylglycerols, from plant- derived residues, indicated by plant sterols, such as β-sitosterol, stigmasterol and campesterol, and only traces of triacylglycerols (Evershed 1993; Evershed et al. 1997a; Dudd and Evershed 1998). Barnard et al. (2007), however, have recently suggested that microorganisms living off residues can introduce β-sitosterol into residues resulting from the preparation of animal products. Waxes, which are long-chain fatty acids and long-chain alcohols that form protective coatings on skin, fur, feathers, leaves and fruit, also resist decay. Evershed et al. (1991) found epicuticular leaf waxes from plants of the genus Brassica in vessel residues from a Late Saxon/Medieval settlement. Cooking experiments later confirmed the utility of nonacosane, nonacosan-15-one and nonacosan- 15-ol to indicate the preparation of leafy vegetables, such as turnip or cabbage (Charters et al. 1997). Reber et al. (2004) recently suggested n-dotriacontanol could serve as an effective biomarker for maize in vessel residues from sites located in Midwestern and Eastern North America. Beeswax can be identified by the presence and distribution of n-alkanes with carbon chains 23 to 33 atoms in length and palmitic acid wax esters with chains between 40 and 52 carbons in length (Heron et al. 1994; Evershed et al. 1997b). Terpenoid compounds, or terpenes, are long chain alkenes that occur in the tars and pitches of higher plants. The use of GC and GC/MS to detect the diterpenoid, dehydroabietic acid, from conifer products in archaeological residues extends over a span of 25 years (Shackley 1982; Heron and Pollard 1988). Lupeol, α- and β-amyrin and their derivatives indicate the presence of plant materials (Regert 2007). Eerkens (2002) used the predominance of the diterpenoid, Δ–8(9)- isopimaric acid, in a vessel residue from the western Great Basin to argue it contained piñon resins. Other analytical techniques have also been used to identify terpenoid compounds. Sauter et al. 259 (1987) identified the triterpenoid, betulin in Iron Age tar to confirm the tar was produced from birch. Azelaic acid is a short chain dicarboxylic acid is associated with the oxidation of unsaturated fatty acids (Regert et al. 1998). Unsaturated fatty acids are most abundant in seed oils; its presence may indicate the residue reflects the processing of plant seeds. The data obtained by HT-GC and HT-GC/MS analysis is useful for distinguishing plant residues, animal residues and plant/animal combinations. As noted above, the sterol cholesterol is associated with animal products; β-sitosterol, stigmasterol and campesterol are associated with plant products. The presence and abundance of triacylglycerols (TAGs) also varies with the material of origin. When present, amounts of TAGs tend to decrease with increasing numbers of carbon atoms in plant residues (Malainey et al. 2010, 2014, in press). The peak arising from C48 TAGs is largest and peak size (and area) progressively decreases with the C54 TAG peak being the smallest. A line drawn to connect the tops of the C48, C50, C52 and C54 TAG peaks slopes down to the right. This pattern is due to the preponderance of triacylglycerols with fatty acids having carbon chains ranging between 12 and 16 in length; C46 TAG peaks may also be detected. In animal residues, amounts of TAGs tend to increase with carbon numbers, with the C52 or C54 TAG peaks being the largest (Malainey et al. 2010, 2014, in press). A line drawn to connect the tops of the C48, C50, C52 and C54 TAG peaks either resembles a hill or the line slopes up to the right. A parabola-like pattern, such as the shape of a “normal distribution,” can also occur in the residues of oil seeds that contain high levels of C18:1 isomers (Malainey et al. 2010, 2014, in press). This pattern is due to the abundance of triacylglycerols composed of fatty acids with mostly chain lengths of 16 or 18 carbons. 260 Methodology Descriptions of the various samples are provided in Tables 4, 5 and 6. Possible contaminants were removed from pottery by grinding off exterior surfaces with a Dremel® tool fitted with a silicon carbide bit. Immediately thereafter, the sample was crushed with a hammer mortar and pestle and the powder transferred to an Erlenmeyer flask. The loose soil did not require any additional preparation. Lipids were extracted using a variation of the method developed by Folch et al. (1957). The powdered sample was mixed with a 2:1 mixture, by volume, of chloroform and methanol (2 × 25 mL) using ultrasonication (2 × 10 min). Solids were removed by filtering the solvent mixture into a separatory funnel. The lipid/solvent filtrate was washed with 13.3 mL of ultrapure water. Once separation into two phases was complete, the lower chloroform-lipid phase was transferred to a round-bottomed flask and the chloroform removed by rotary evaporation. Any remaining water was removed by evaporation with 2-propanol (1.5 mL); 1.5 mL of chloroform- methanol (2:1, v/v) was used to transfer the dry total lipid extract to a screw-top glass vial with a Teflon®-lined cap. The sample was flushed with nitrogen and stored in a -20C freezer. Preparation of FAMES A 100-400 L aliquot of the total lipid extract solution was placed in a screw-top test tube and dried in a heating block under nitrogen. Fatty acid methyl esters (FAMES) were prepared by treating the dry lipid with 3 mL of 0.5 N anhydrous hydrochloric acid in methanol (68°C; 60 min). Fatty acids that occur in the sample as di- or triglycerides are detached from the glycerol molecule and converted to methyl esters. After cooling to room temperature, 2.0 mL of ultrapure water was added; FAMES were recovered with petroleum ether (2 × 1.5 mL) and transferred to a vial. The solvent was removed by heat under a gentle stream of nitrogen; the FAMES were dissolved in 75 µL of iso-octane then transferred to a GC vial with a conical glass insert. 261 Preparation of TMS derivatives A 100-200 L aliquot of the total lipid extract solution was placed in a screw-top vial and dried under nitrogen. Trimethylsilyl (TMS) derivatives were prepared by treating the lipid with 70 L of N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA) containing 1% trimethylchlorosilane, by volume (70ºC; 30 min). The sample was then dried under nitrogen and the TMS derivatives were redissolved in 100 L of hexane. Solvents and chemicals were checked for purity by running a sample blank. Traces of fatty acid contamination were subtracted from sample chromatograms. The relative percentage composition was calculated by dividing the integrated peak area of each fatty acid by the total area of fatty acids present in the sample. In order to identify the residue on the basis of fatty acid composition, the relative percentage composition was determined first with respect to all fatty acids present in the sample (including very long chain fatty acids) and second with respect to the ten fatty acids utilized in the development of the identification criteria (C12:0, C14:0, C15:0, C16:0, C16:1, C17:0, C18:0, C18:1ω9, C18:1ω11 and C18:2) (not shown). The second step is necessary for the application of the identification criteria presented in Table 2. It must be understood that the identifications given do not necessarily mean that those particular foods were actually prepared because different foods of similar fatty acid composition and lipid content would produce similar residues (see Table 3). It is possible only to say that the material of origin for the residue was similar in composition to the food(s) indicated. High temperature gas chromatography and high temperature gas chromatography with mass spectrometry is used to further clarify the identifications. 262 Gas Chromatography Analysis Parameters The GC analysis was performed on a Varian 3800 gas chromatograph fitted with a flame ionization detector connected to a personal computer. Samples were separated using a VF-23 fused silica capillary column (30 m × 0.25 mm I.D.; Varian; Palo Alto, CA). An autosampler injected the sample using a split/splitless injection system. Hydrogen was used as the carrier gas with a column flow of 1.0 mL/min. Column temperature was increased from 80°C to 140°C at a rate of 20°C per minute then increased to 185°C at a rate of 4oC per minute. After a 4.0 minute hold, the temperature was further increased to 250°C at 10°C per minute and held for 2 minutes. Chromatogram peaks were integrated using Varian MS Workstation® software and identified through comparisons with external qualitative standards (NuCheck Prep; Elysian, MN). High Temperature Gas Chromatography and Gas Chromatography with Mass Spectrometry Both HT-GC and HT GC-MS analyses were performed on a Varian 3800 gas chromatograph fitted with a flame ionization detector and a Varian 4000 mass spectrometer connected to a personal computer. For HT-GC analysis, the sample was injected onto a DB-1HT fused silica capillary column (15 m × 0.32 mm I.D.; Agilent J&W; Santa Clara, CA) connected to the flame ionization detector, using hydrogen as the carrier gas. The column temperature was held at 50°C for 1 minute then increased to 350°C at a rate of 15°C per minute and held for 26 minutes. For HT-GC/MS analysis, samples were injected onto a DB-5HT fused silica capillary column (30 m × 0.25 mm I.D.; Agilent J&W; Santa Clara, CA) connected to the ion trap mass spectrometer in an external ionization configuration using helium as the carrier gas. After a 1 minute hold at 50°C, the column temperature was increased to 180°C at a rate of 40°C per minute then ramped up to 230°C at a rate of 5°C per minute and finally increased to 350°C at a rate of 15°C per minute and held for 27.75 minutes. The Varian 4000 mass spectrometer was operated in electron-impact ionization 263 mode scanning from m/z 50-700. Chromatogram peaks and MS spectra were processed using Varian MS Workstation® software and identified through comparisons with external qualitative standards (Sigma Aldrich; St. Louis, MO and NuCheck Prep; Elysian, MN), reference samples and the National Institute of Standards and Technology (NIST) database. Results of Archaeological Data Analysis Fatty acid compositions of the residues were determined by using the area under the chromatographic peak of a given fatty acid, as calculated by the Varian MS Workstation® software minus the solvent blank. Lipid compositions of all extracted pottery residues, except residue 17KMS 25, are presented in Table 4. Insufficient fatty acids were present in residue 17KMS 25 to attempt identification but lipid biomarkers were detected (Table 5). Compositions of the lipid residues extracted from the soil samples are given in Table 6. The term “Area” represents the area under the chromatographic peak of a given fatty acid, as calculated by the Varian MS Workstation® software minus the solvent blank. Hydroxide or peroxide degradation products can interfere with the integration of the C22:0 and C22:1 peaks; these fatty acids were excluded from the analysis. Decomposed Nut Oil Residue 17KMS 5 is characterized by extremely high levels of the fatty acid C16:0, 71.07%, and relatively low levels of other fatty acids. No biomarkers were detected; analysis using HT- GC/MS indicates that the C48 and C50 TAGs may be present. This residue does not conform to the criteria outlined in Table 2, but its identity can be established with confidence. High and very high fat content foods produce residues dominated by greatly elevated levels of C18:1 isomers, moderate levels of C16:0 and low levels of other fatty acids. When these residues are highly degraded, the level of monounsaturated C18:1 isomers drops and relative levels of the stable saturated fatty acids 264 increase. The result is a residue with very high levels of C16:0 and low to moderate levels of other fatty acids. The experimental cooking residue resulting from the preparation of crushed acorns, residue MQ 19D, can be used to illustrate how this occurs (Table 7; see also Malainey 2007). This crushed acorn cooking residue was prepared by J. Michael Quigg in Texas and sent to the Brandon University lab where it was stored in an oven at 75°C for 30 days. Lipids were extracted with organic solvents; FAMEs were derived and analyzed using gas chromatography. The first column of Table 7 shows the relative fatty acid composition of the partially degraded residue MQ 19D after 30 days of oven storage. The next two columns show how the relative amounts of stable fatty acids would change if the level of C18:1 isomers dropped to 3.73%, which is the amount in 17KMS 5. The relative percentage composition of 17KMS 5 is given in the fourth column. The changes in relative fatty acid composition extrapolated for highly degraded residue of cooked crushed acorns compares very favourably with the archaeological residue 17KM 5. For this reason, residue 17KMS 5 can be identified as a decomposed nut oil. The level of C18:0 is above 14%; this is slightly higher than expected and may indicate traces of animal products are also present. Decomposed Nut Oil Dominates Residues 17KMS 19 and 17KMS 27 are characterized by extremely high levels of the fatty acid C16:0, 68.65% and 69.95%, respectively. As outlined above, the very high levels of C16:0 can be attributed to the decomposition of high and very high fat content foods, such as nut oil, that originally had high levels of C18:1 isomers, moderate levels of C16:0 and low levels of other fatty acids. Levels of the fatty acid C16:0 in residues 17KMS 19 and 17KMS 27 are just slightly lower than those observed in residue 17KMS 5 and there is evidence that other foods were prepared in the vessels, as well. The level of medium chain saturated fatty acids (sum of the fatty acids C12:0, 265 C14:0 and C15:0) is 15.14% in residue 17KMS 19. In North America, similarly high levels of medium chain saturated fatty acids indicate the preparation of low fat content plants, such as roots, greens and certain berries. The level of the fatty acid C18:0 in residue 17KMS 27 is 16.32%, which suggests the possible presence of animal products. The occurrence of animal products in residue 17KMS 27 is further substantiated by the probable presence of the animal sterol cholesterol. The distribution of triacylglycerols in the residues attests to the presence of both plant and animal products. The area of the C48 TAG in residue 17KMS 27 was larger than C50 TAG peak; the C52 and C54 TAGs were also present. There is only weak evidence for the presence of cholesterol in residue 17KMS 19 and only traces of TAGs were detected. Decomposed Nut Oil and Other Foods Nineteen residues, 17KMS 1, 17KMS 2, 17KMS 3, 17KMS 7, 17KMS 8, 17KMS 9, 17KMS 11-15, 17KMS 17, 17KMS 18, 17KMS 20, 17KMS 22, 17KMS 23, 17KMS 26, 17KMS 29 and 17KMS 30, are identified as combinations of decomposed nut oil and other foods. The fatty acid C16:0 appears in all foods and archaeological food residues. The mean and standard deviation of C16:0 levels in 600 archaeological residues previously identified as food was determined and found to be 31 ± 9%. Levels of the fatty acid 16:0 in the residues discussed in this section are outside of the expected range of 22% and 40%. Levels of the fatty acid 16:0 in all but one of the residues discussed in this section range exceed 49% but are less than 67%, i.e., they are more than two standard deviations from the mean, but within the four standard deviations of the mean value. At 48.91%, the C16:0 level in 17KMS 7 is just outside of the two standard deviation range of 49%. Elevated levels of C16:0 in these residues are likely due to presence of decomposed nut oil but 266 various other foods, such as animal products, low fat content plants and/or medium fat content foods, were also prepared in these vessels. Decomposed Nut Oil and Low Fat Content Plant Products, Animal Products Present or Probably Present: Residues 17KMS 9, 17KMS 13, 17KMS 14, 17KMS 15, 17KMS 18, 17KMS 20, 17KMS 22, 17KMS 23, 17KMS 26 and 17KMS 29 are similar in that the level of the fatty acid 16:0 is high, ranging between 49.30% and 64.19%, which indicates the presence of decomposed nut oil. Levels of medium chain saturated fatty acids (the sum of C12:0, C14:0 and C15:0) exceed 10% in all of these residues. As noted above, similarly high levels of medium chain saturated fatty acids in North American archaeological lipid residues indicate the preparation of low fat content plants, such as roots, greens and certain berries. The residues are likely a combination of decomposed nut oil and low fat content plants. Elevated levels of the fatty acid C18:0 indicate the presence of animal products in all of these residues as well. In most residues, the level of this fatty acid ranges between 13% and 16%. The level of the fatty acid C18:0 is close to or exceeds 20% in residues 17KMS 9, 17KMS 13, 17KMS 14 and 17KMS 22, which suggests the animal component of some of these complex residues may be due in part to the presence of large herbivore products, such as deer, moose or bison flesh. The animal sterol cholesterol was detected in residues 17KMS 9, 17KMS 15, 17KMS 26 and 17KMS 29. Dehydroabietic acid may occur in residue 17KMS 15 and 17KMS 26; this biomarker indicates the presence of conifer products, which may have been introduced from pine nuts, firewood, resins or other conifer products. The C48 TAG peak appears in residue 17KMS 14, 17KMS 15 and 17KMS 18, which suggests that plant products dominate these residues. Both the C48 and C50 TAGs may appear in residues 17KMS 9 and 17KMS 13. Only traces of TAGs appear in the other samples. 267 Decomposed Nut Oil with Low Fat Content Plants and Medium Fat Content Foods: Residue 17KMS 17 has a high level of C16:0, 55.49%, which suggests the presence of decomposed nut oil. As with the residues described above, the sum of the levels of medium chain saturated fatty acids C12:0, C14:0 and C15:0 is 11.32%, which indicates the presence of low fat content plant foods, such as greens, roots or certain berries. The level of C18:1 isomers is 20.34%, which indicates the presence of medium fat content foods. As indicated in Table 3, both plant and animal foods can produce these types of residues. Examples of medium fat content plant foods include corn, mesquite and cholla. Freshwater fish, terrapin, Rabdotus snail and late winter, fat-depleted elk are examples of medium fat content animal foods. The biomarker for conifer products dehydroabietic acid may occur in this residue. Only traces of TAGS appear in this residue. Decomposed Nut Oil, Animal Products present or possibly present: Residues 17KMS 2, 17KMS 3, 17KMS 7, 17KMS 11, 17KMS 12 and 17KMS 30 are similar in that 1) the levels of C16:0 are high, 2) levels of medium chain saturated fatty acids are elevated but do not exceed 10% and 3) there is evidence of animal products. Levels of C16:0 in most of these residues range between 54.69% and 62.24%; the level in 17KMS 7 is 48.91%. High levels of C16:0 suggest the presence of decomposed nut oil in all residues; the somewhat elevated levels of medium chain saturated acids may be due to the nuts themselves. Evidence of animal products is stronger in some residues than in others. The animal sterol cholesterol occurs in residue 17KMS 11 and probably occurs in residue 17KMS 7 and possibly in 17KMS 30. Levels of the fatty acid C18:0 are close to, or exceed, 20% in residues 17KMS 3, 17KMS 7, 17KMS 12 and 17KMS 30, which suggests the presence of animal products. The C18:0 level is somewhat elevated in residue 17KMS 2, 17.19%, which indicates animal products may occur. Dehydroabietic acid may occur in residues 17KMS 7 and 17KMS 30, which indicates the possible presence of conifer products. Analysis by HT-GC/MS shows that both 268 the C48 and C50 TAGs occur in residue 17KMS 12. These TAGs may be present in residues 17KMS 2, 17KMS 3 and 17KMS 11; the C48 TAG may appear in residue 17KMS 7. Traces of TAGs were detected in residue 17KMS 30. Decomposed Nut Oil with Animal and High C18:0 (Large Herbivore): Residue 17KMS 8 has a high level of C16:0, 59.03%, which indicates the presence of decomposed nut oil. It also has a high level of C18:0, 28.44%; similar levels of C18:0 result from the preparation of large herbivores, such as bison, deer, moose, fat elk meat or other bovines or cervids; but javelina meat and tropical oil seeds also produce residues high in C18:0 and must be considered as potential sources where available. No biomarkers occur; the C48 and C50 TAGs may be present. Decomposed Nut Oil with Medium Fat Content Foods: As with the other residues described in this section, residue 17KMS 1 has a high level of C16:0, 51.32%, which suggests the presence of decomposed nut oil. It differs, however, in that the level of C18:1 isomers is 15.07%, which indicates the presence of medium fat content foods. As noted above, both plant and animal foods can produce these types of residues. Examples of medium fat content plant foods include corn, mesquite and cholla. Freshwater fish, terrapin, Rabdotus snail and late winter, fat-depleted elk are examples of medium fat content animal foods. The biomarker for conifer products dehydroabietic acid occurs; the C48 and C50 TAGs may be present. High C18:0 level “Large Herbivore” Both residue 17KMS 24 and 17KMS 28 are characterized by high levels of the fatty acid C18:0, 46.19% and 38.45% respectively. Similarly high levels of C18:0 result from the preparation of large herbivores, such as bison, deer, moose, fat elk meat or other bovines or cervids; but javelina meat and tropical oil seeds also produce residues high in C18:0 and must be considered as potential sources where available. The level of medium chain saturated fatty acids 269 in residue 17KMS 28 is just below 10%, which suggests low fat content plant products are probably present and the level of the fatty acid C17:0 is 5.53%. Elevated levels of this particular fatty acids are associated with plant roots. No lipid biomarkers were detected in either residue. The C48: C50:C52:C54 TAG ratio in residue 17KMS 24 is 1.0: 3.1: 4.3: 2.2, which is consistent with the identification of animal products. Only traces of TAGs were detected in 17KMS 28. Low Fat Content Plants Dominate Residues 17KMS 4 and 17KMS 10 are similar in that the levels of medium chain saturated fatty acids (sum of C12:0, C14:0 and C15:0) are very high, 33.72% and 40.15%, respectively. As these vessels were recovered from a site in North America, they were probably used in the preparation of low fat content plant foods. The level of C18:0 is elevated in residue 17KMS 4 and the animal sterol cholesterol may occur, which strongly suggests the presence of animal products. The biomarker for conifers, dehydroabietic acid, may occur in residue 17KMS 4. The level of C18:1 isomers is 18.15% in residue 17KMS 10, which indicates the presence of medium fat content foods. Both plant and animal foods can produce medium fat content residues; but in this case, the absence of the fatty acid C18:0 indicates medium fat content plant foods are the most likely source. Examples of medium fat content plant foods include corn, mesquite and cholla. No biomarkers were detected in residue 17KMS 10; the C48 and C50 TAGs may occur in both residues. Low Fat Content and Medium Fat Content Residue 17KMS 6 is characterized by high levels of medium chain saturated fatty acids, 18.67%, which probably indicates the presence of low fat content plant foods, such as greens, roots and certain berries. The level of C18:1 isomers in this residue is 24.85%, which indicates the presence of medium fat content foods. While both plant and animal foods can produce 270 medium fat content residues, the presence of the animal sterol cholesterol suggests that an animal source is more likely. Freshwater fish, terrapin, Rabdotus snail and late winter, fat-depleted elk are examples of medium fat content animal foods. The biomarker for conifers, dehydroabietic acid, may occur in this residue; the C48 and C50 TAGs may also occur. Medium Fat Content with Animal Products and Low Fat Content Plants The level of C18:1 isomers in residue 17KMS 21 is 15.20%, which indicates the presence of medium fat content foods. While both plant and animal foods can produce medium fat content residues, the possible presence of the animal sterol cholesterol and elevated levels of the fatty acid C18:0, 18.88%, suggests that an animal source is more likely. Freshwater fish, terrapin, Rabdotus snail and late winter, fat-depleted elk are examples of medium fat content animal foods. The level of medium chain saturated fatty acids is just under 10%, which suggests that low fat content plant foods, such as greens, roots and certain berries, are probably present. The biomarker for conifers, dehydroabietic acid, may occur in this residue. Traces of the C48, C50, C52 and C54 TAGs appear. Moderate-High Fat Content The level of C18:1 isomers in residue 17KMS 16 is 28.61%, which is a moderate-high level. Foods known to produce moderate-high fat content residues include Texas ebony seeds and the fatty meat of medium-sized mammals, such as beaver. The origin of the residue is somewhat ambiguous. The level of the fatty acid C18:0 in this residue is 13.30%; which suggests the possible presence of animal products. The presence of cholesterol derivatives further indicates that the processing of an animal gave rise to this medium-high fat content residue. The presence of plants is indicated by the probable occurrence of the plant sterol β- 271 sitosterol. It is possible that the residue was due to a combination of moderate-high fat animal and moderate-high fat content plants. Only traces of TAGs were detected in this residue. Biomarkers Detected: Insufficient Fatty Acids Insufficient fatty acids were recovered to attempt a characterization of residue 17KMS 25 but two biomarkers may occur (Table 5). The animal sterol cholesterol may occur; the biomarker for conifers, dehydroabietic acid may occur. This suggests the possible presence of animal products and conifer products. No triacylglycerols were detected in the residue. Soil Samples Please note that only fatty acids that elute (i.e., emerge from the GC column) in the typical range of fatty acids found in foods are discussed. The fatty acid compositions of the residues extracted from the soil samples are very different from those extracted from the pottery. Levels of polyunsaturated fatty acids (those with multiple double bonds) in the soil residues are much higher than those observed in archaeological pottery residues. The composition of soil residue 17KMS 33 is most different from the archaeological pottery residues and may reflect only natural soil lipids. If this is correct, the other two soil residues may be combinations of natural and cultural lipids. The animal sterol cholesterol and the plant sterols stigmasterol and β- sitosterol occur in all soil residues. Triacylglycerols appear in all residues but we do not have information on the typical distribution of TAGs in soils. If the distribution is similar to that in foods, it suggests they were primarily derived from plant products. The fatty acid composition of soil residue 17KMS 33 is characterized by high levels of very long chain polyunsaturated fatty acid C22:4 (19.04%) and high levels of two polyunsaturated fatty acids that elute simultaneously, a C20:3 isomer and C20:5 (14.06%). The level of the long chain polyunsaturated fatty acid C18:2 is 5.05%. The level of the fatty acid 272 C16:0 is 16.85% and the level of the fatty acid C18:1 is 12:84%. A wide variety of other polyunsaturated, monounsaturated and saturated fatty acids were also detected. Soil residue 17KMS 33 may have been derived primarily from natural soil lipids. Soil residue 17KMS 31 is quite similar to 17KMS 33 but levels of the fatty acid C22:4 and the C20:3/C20:5 combination are somewhat lower, 16.86% and 8.59%, respectively. The level of the fatty acid C18:2 in soil residue 17KMS 31 is slightly higher, 5.83%. The major difference is the significantly higher level of C18:1 isomers in residue 17KMS 31. At 28.03%, the level of C18:1 isomers in soil residue 17KMS 31 is more than twice that in soil residue 17KMS 33. If the fatty acid composition of residue 17KMS 33 reflects only natural soil lipids, the elevated level of C18:1 isomers may signify the presence of some cultural lipids in this residue. Levels of the polyunsaturated fatty acid C22:4 (5.08%) and the C20:3/C20:5 combination (2.79%) are much lower in soil residue 17KMS 32 compared to the other soil residues. The level of the polyunsaturated fatty C18:2 is somewhat higher, 8.57%; the level of C18:1 isomers is much higher, 53.67%. The context of the soil sample should be carefully evaluated; it may indicate the location of an activity area where high fat foods were processed. REFERENCES CITED Barnard, H., A. N. Dooley and K. F. Faull 2007 Chapter 5: An Introduction to Archaeological Lipid Analysis by GC/MS. In Theory and Practice of Archaeological Residue Analysis, edited by H. Barnard and J. W. Eerkens, pp.42-60. British Archaeological Reports International Series 1650. Oxford, UK. Charters, S., R. P. Evershed, A. Quye, P. W. Blinkhorn and V. Denham 1997 Simulation Experiments for Determining the Use of Ancient Pottery Vessels: Te Behaviour of Epicuticular Leaf Wax during Boiling of a Leafy Vegetable. Journal of Archaeological Science 24: 1-7. 273 Condamin, J., F. Formenti, M. O. Metais, M. Michel, and P. Blond 1976 The Application of Gas Chromatography to the Tracing of Oil in Ancient Amphorae. Archaeometry 18(2):195-201. Dudd, S. N. and R. P. Evershed 1998 Direct demonstration of milk as an element of archaeological economies. Science 282: 1478-1481. Eerkens, J. W. 2002 The Preservation and Identification of Pinon Resins by GC-MS in Pottery from the Western Great Basin. Archaeometry 44(1):95-105. Evershed, R.P. 1993 Biomolecular Archaeology and Lipids. World Archaeology 25(1):74-93. Evershed, R. P. 2000 Biomolecular Analysis by Organic Mass Spectrometry. In Modern Analytical Methods in Art and Archaeology, edited by E. Ciliberto and G. Spoto, pp. 177-239. Volume 155, Chemical Analysis. John Wiley & Sons, New York. Evershed, R. P., C. Heron and L. J. Goad 1990 Analysis of Organic Residues of Archaeological Origin by High Temperature Gas Chromatography and Gas Chromatography-Mass Spectroscopy. Analyst 115:1339-1342. Evershed, R.P., C. Heron and L.J. Goad 1991 Epicuticular Wax Components Preserved in Potsherds as Chemical Indicators of Leafy Vegetables in Ancient Diets. Antiquity 65:540-544. Evershed, R. P., H. R. Mottram, S. N. Dudd, S. Charters, A. W. Stott, G. J. Lawrence, A. M. Gibson, A. Conner, P. W. Blinkhorn and V. Reeves 1997a New Criteria for the Identification of Animal Fats in Archaeological Pottery. Naturwissenschaften 84: 402-406. Evershed, R. P., S. J. Vaugh, S. N. Dudd and J. S. Soles 1997b Fuel for Thought? Beeswax in Lamps and Conical Cups from Late Minoan Crete. Antiquity 71: 979-985. Evershed, R. P., S. N. Dudd, M. J. Lockheart and S. Jim 2001 Lipids in Archaeology. In Handbook of Archaeological Sciences, edited by D. R. Brothwell and A. M. Pollard, pp. 331-349. John Wiley & Sons, New York. Folch, J., M. Lees and G. H. Sloane-Stanley 1957 A simple method for the isolation and purification of lipid extracts from brain tissue. Journal of Biological Chemistry 191:833. Frankel, E. N. 274 1991 Recent Advances in Lipid Oxidation. Journal of the Science of Food and Agriculture 54:465-511. Heron, C., and A.M. Pollard 1988 The Analysis of Natural Resinous Materials from Roman Amphoras. In Science and Archaeology Glasgow 1987. Proceedings of a Conference on the Application of Scientific Techniques to Archaeology, Glasgow, 1987, edited by E. A. Slater and J. O. Tate, pp. 429- 447. BAR British Series 196 (ii), Oxford. Heron, C., N. Nemcek, K. M. Bonfield, J. Dixon and B. S. Ottaway 1994 The Chemistry of Neolithic Beeswax. Naturwissenschaften 81: 266-269. Loy, T. 1994 Residue Analysis of Artifacts and Burned Rock from the Mustang Branch and Barton Sites (41HY209 and 41HY202). In: Archaic and Late Prehistoric Human Ecology in the Middle Onion Creek Valley, Hays County, Texas. Volume 2: Topical Studies, by R. A. Ricklis and M. B. Collins, pp. 607- 627. Studies in Archeology 19, Texas Archaeological Research Laboratory, The University of Texas at Austin. Malainey, M. E. 1997 The Reconstruction and Testing of Subsistence and Settlement Strategies for the Plains, Parkland and Southern boreal forest. Unpublished Ph.D. thesis, University of Manitoba. Malainey, M. E. 2007 Chapter 7: Fatty Acid Analysis of Archaeological Residues: Procedures and Possibilities. In Theory and Practice of Archaeological Residue Analysis, edited by H. Barnard and J. W. Eerkens, pp.77-89. British Archaeological Reports International Series 1650. Oxford, UK. Malainey, M. E., M. Álvarez, I. Briz i Godino, D. Zurro, E. Verdún i Castelló and T. Figol 2014 The Use of Shells as Tools by Hunter-Gatherers in the Beagle Channel (Tierra del Fuego, South America): An Ethnoarchaeological Experiment. Archaeological and Anthropological Sciences. DOI: 10.1007/s12520-014-0188-1. Malainey, M. E., P. J. Innes and T. J. Figol 2010 Taking a Second Look: Results of the Re-analysis of Archaeological Lipid Residues from North America and Beyond, Paper presented at the 75nd Annual Meeting of the Society for American Archaeology, St. Louis, MO. Malainey, M. E., P. Innes and T. Figol in press Taking a Second Look: A Functional Analysis of Burned Rock Features from Eight Sites in Texas and Arizona. Chapter prepared for a volume edited by H. Hoekman-Sites and M. Raviele to be published by the University of Colorado Press (40 pages). Malainey, M. E., K. L. Malisza, R. Przybylski and G. Monks 2001a The Key to Identifying Archaeological Fatty Acid Residues. Paper presented at the 34th Annual Meeting of the Canadian Archaeological Association, Banff, Alberta, May 2001. 275 Malainey, M. E., R. Przybylski and B. L. Sherriff 1999a The Fatty Acid Composition of Native Food Plants and Animals of Western Canada. Journal of Archaeological Science 26:83-94. 1999b The Effects of Thermal and Oxidative Decomposition on the Fatty Acid Composition of Food Plants and Animals of Western Canada: Implications for the Identification of archaeological vessel residues. Journal of Archaeological Science 26:95-103. 1999c Identifying the former contents of Late Precontact Period pottery vessels from Western Canada using gas chromatography. Journal of Archaeological Science 26(4): 425-438. 2001b One Person’s Food: How and Why Fish Avoidance May Affect the Settlement and Subsistence Patterns of Hunter-Gatherers. American Antiquity 66(1): 141-161. Malainey, M. E., R. Przybylski and G. Monks 2000a The identification of archaeological residues using gas chromatography and applications to archaeological problems in Canada, United States and Africa. Paper presented at The 11th Annual Workshops in Archaeometry, State University of New York at Buffalo, February 2000. 2000b Refining and testing the criteria for identifying archaeological lipid residues using gas chromatography. Paper presented at the 33rd Annual Meeting of the Canadian Archaeological Association, Ottawa, May 2000. 2000c Developing a General Method for Identifying Archaeological Lipid Residues on the Basis of Fatty Acid Composition. Paper presented at the Joint Midwest Archaeological & Plains Anthropological Conference, Minneapolis, Minnesota, November 2000. Marchbanks, M. L. 1989 Lipid Analysis in Archaeology: An Initial Study of Ceramics and Subsistence at the George C. Davis Site. Unpublished M.A. thesis, The University of Texas at Austin. Patrick, M., A. J. de Konig and A. B. Smith 1985 Gas Liquid Chromatographic Analysis of Fatty Acids in Food Residues from Ceramics Found in the Southwestern Cape, South Africa. Archaeometry 27(2): 231-236. Quigg, J. M., C. Lintz, S. Smith and S. Wilcox 2000 The Lino Site: A Stratified Late Archaic Campsite in a Terrace of the San Idelfonzo Creek, Webb County, Southern Texas. Technical Report No. 23765, TRC Mariah Associates Inc., Austin. Texas Department of Transportation, Environmental Affairs Division, Archaeological Studies Program Report 20, Austin. Quigg, J. M., M. E. Malainey, R. Przybylski and G. Monks 2001 No bones about it: using lipid analysis of burned rock and groundstone residues to 276 examine Late Archaic subsistence practices in South Texas. Plains Anthropologist 46(177): 283-303. Reber, E. A., S. N. Dudd, N. J. van der Merwe and R. P. Evershed 2004 Direct detection of maize in pottery residue via compound specific stable carbon isotope analysis. Antiquity 78: 682-691. Regert, M., H. A. Bland, S. N. Dudd, P. F. van Bergen and R. P. Evershed 1998 Free and Bound Fatty Acid Oxidation Products in Archaeological Ceramic Vessels. Philosophical Transactions of the Royal Society of London, B 265 (1409):2027-2032. Regert, M. 2007 Chapter 6: Elucidating Pottery Function using a Multi-step Analytical Methodology combining Infrared Spectroscopy, Chromatographic Procedures and Mass Spectrometry. In Theory and Practice of Archaeological Residue Analysis, edited by H. Barnard and J. W. Eerkens, pp.61-76. British Archaeological Reports International Series 1650. Oxford, UK. Sauter, F., E.W.H. Hayek, W. Moche and U. Jordis 1987 Betulin aus archäologischem Schwelteer. Z. für Naturforsch 42c (11-12):1151-1152. Shackley, M. 1982 Gas Chromatographic Identification of a Resinous Deposit from a 6th Century Storage Jar and Its Possible Identification. Journal of Archaeological Science 9:305-306. Skibo, J. M. 1992 Pottery Function: A Use-Alteration Perspective. Plenum Press, New York. Solomons, T. W. G. 1980 Organic Chemistry. John Wiley & Sons, Toronto. 277 List of Tables Table 1. Summary of average fatty acids compositions of modern food groups generated by hierarchical cluster analysis. Table 2. Criteria for the identification of archaeological residues based on the decomposition patterns of experimental cooking residues prepared in pottery vessels. Table 3. Known food sources for different types of decomposed residues. Table 4. Sample descriptions and lipid compositions of Cloudman site pottery residues. Table 5. Sample description and lipid biomarkers in residue 17KMS 25. Table 6. Sample descriptions and lipid compositions of Cloudman site soil lipids. Table 7. Experimental cooking residue of crushed acorn with extrapolation of further degradation compared to the relative fatty acid composition of residue 17KM 5. 278 Table 1. Summary of average fatty acid compositions of modern food groups generated by hierarchical cluster analysis. Cluster A B C Subcluster I II III IV V VI VII VIII IX X XI XII XIII XIV XV Type Mammal Fat and Marrow C16:0 19.90 Large Fish Fish Berries Mixed Herbivore Meat 19.39 16.07 14.10 and Nuts 3.75 Roots Seeds Mixed Greens Berries Roots Greens Roots Seeds and Berries 12.06 7.48 19.98 7.52 10.33 18.71 3.47 22.68 24.19 18.71 C18:0 7.06 20.35 3.87 2.78 1.47 2.36 2.58 2.59 3.55 2.43 2.48 1.34 3.15 3.66 5.94 C18:1 56.77 35.79 18.28 31.96 51.14 35.29 29.12 6.55 10.02 15.62 5.03 14.95 12.12 4.05 3.34 C18:2 C18:3 VLCS VLCU 7.01 0.68 0.16 0.77 8.93 2.91 4.04 41.44 35.83 54.69 48.74 64.14 39.24 18.82 29.08 26.24 16.15 15.61 2.61 4.39 3.83 1.05 3.66 1.51 7.24 5.49 19.77 35.08 39.75 9.64 17.88 3.42 0.32 0.23 0.15 0.76 4.46 2.98 8.50 5.19 3.73 6.77 9.10 15.32 18.68 43.36 4.29 39.92 24.11 0.25 2.70 1.00 2.23 0.99 2.65 1.13 0.95 2.06 0.72 1.10 VLCS- Very Long Chain (C20, C22 and C24) Saturated Fatty Acids VLCU - Very Long Chain (C20, C22 and C24) Unsaturated Fatty Acids 279 Table 2. Criteria for the identification of archaeological residues based on the decomposition patterns of experimental cooking residues prepared in pottery vessels. Identification Large herbivore Large herbivore with plant OR Bone marrow Plant with large herbivore Beaver Fish or Corn Fish or Corn with Plant Plant (except corn) Medium Chain C18:0 C18:1 isomers  15% low  15% low low  15%  10%  27.5%  15%  25% 15%  X  25%  25% Low no data  25%  25% 15%  X  27.5%  25% 15%  X  27.5%  27.5%  15% Table 3. Known food sources for different types of decomposed residues. Decomposed Residue Identification Large herbivore Large herbivore with plant OR Bone marrow Low Fat Content Plant (Plant greens, roots, berries) Medium-Low Fat Content Plant Medium Fat Content (Fish or Corn) Moderate-High Fat Content (Beaver) High Fat Content Plant Foods Known to Animal Foods Known To Produce Produce Similar Residues Similar Residues Tropical seed oils, Bison, deer, moose, fall-early winter including sotol seeds fatty elk meat, Javelina meat Jicama tuber, buffalo gourd, yopan leaves, biscuit root, millet Prickly pear, Spanish dagger Cooked Camel’s milk None Corn, mesquite beans, Freshwater fish, Rabdotus snail, cholla Texas ebony terrapin, late winter fat-depleted elk Beaver and probably raccoon or any other fat medium-sized mammals High fat nuts and seeds, Rendered animal fat (other than large including acorn and pecan herbivore), including bear fat Very High Fat Content Very high fat nuts and Freshly rendered animal fat (other seeds, including pine nuts than large herbivore) 280 Table 4. Sample descriptions and lipid compositions of Cloudman site pottery residues. Fatty acid 17KMS 1 17KMS 2 17KMS 3 Area Rel% Area Rel% Area Rel% C12:0 C14:0 C14:1 C15:0 C16:0 C16:1 C17:0 C17:1 C18:0 C18:1s C18:2 C18:3s C20:0 C20:1 C24:0 C24:1 Total Biomarkers Triacylglycerols Identification Vessel No. Catalogue No. Sample Mass 25739 89947 1280 46171 897233 68704 29529 2334 282682 263520 3772 3751 23130 4920 4456 1137 1748305 Dehydroabietic acid 1.47 5.14 0.07 2.64 51.32 3.93 1.69 0.13 16.17 15.07 0.22 0.21 1.32 0.28 0.25 0.07 100.00 Possibly C48 and C50 TAGS Medium fat content; Decomposed nut oil; Plant products present; Animal product may be present; Conifer products present 1 17827 323400 0 245908 4884437 135756 220596 30616 1360199 564478 17691 3998 71520 4812 33303 0 7914541 0.23 4.09 0.00 3.11 61.71 1.72 2.79 0.39 17.19 7.13 0.22 0.05 0.90 0.06 0.42 0.00 100.00 None detected Possibly C48 and C50 TAGS 1.17 5.76 0.81 2.59 54.69 1.04 2.73 0.49 23.87 3.74 0.03 1.04 1.68 0.02 0.26 0.06 100.00 76814 377191 53031 169748 3578739 68330 178638 32025 1561883 244748 1714 67818 110197 1409 17247 3818 6543350 None detected Possibly C48 and C50 TAGS Decomposed nut oil; Animal products may be present Decomposed nut oil; Animal products probably present 6 109 7282.169.2.01.01 7282.195.06.02 7282.770.05.02 6.563 g 11.788 g 10.271 g 281 C12:0 C14:0 C14:1 C15:0 C16:0 C16:1 C17:0 C17:1 C18:0 C18:1s C18:2 C18:3s C20:0 C20:1 C24:0 C24:1 Total 84787 92001 15319 93688 199973 6601 15656 6058 145136 93107 21809 7544 14808 0 1628 3939 802054 10.57 11.47 1.91 11.68 24.93 0.82 1.95 0.76 18.10 11.61 2.72 0.94 1.85 0.00 0.20 0.49 100.00 27690 351963 56244 154868 6565301 11101 222725 9901 1320023 344797 6121 24060 80343 7328 53867 1193 9237525 Rel% 0.30 3.81 0.61 1.68 71.07 0.12 2.41 0.11 14.29 3.73 0.07 0.26 0.87 0.08 0.58 0.01 100.00 Biomarkers possibly Dehydroabietic None detected Possibly Cholesterol; 17KMS 6 Area Rel% 27435 56122 0 19246 229683 15846 0 2247 24316 136856 17301 0 10398 6188 3502 1525 550667 4.98 10.19 0.00 3.50 41.71 2.88 0.00 0.41 4.42 24.85 3.14 0.00 1.89 1.12 0.64 0.28 100.00 Cholesterol; possibly Dehydroabietic acid Table 4 cont’d. Sample descriptions and lipid compositions of Cloudman site pottery residues. Fatty acid 17KMS 4 Area Rel% 17KMS 5 (dil) Area Triacylglycerols Identification Vessel No. Catalogue No. Sample Mass acid Possibly C48 and C50 TAGS Low fat content plant dominates; Animal products; Conifer products may occur Possibly C48 and C50 TAGS Possibly C48 and C50 TAGS Decomposed nut oil; possible traces of animal products Low fat content plant and Medium fat content food; Animal products; Conifer products may occur 131 35 34 7282.770.03.02 7282.479.04.03 7282.479.03.01 6.187 g 7.138 g 8.184 g 282 Table 4 cont’d. Sample descriptions and lipid compositions of Cloudman site pottery residues. Fatty acid 17KMS 7 17KMS 8 (dil) 17KMS 9 Area Rel% Area Rel% Area Rel% C12:0 C14:0 C14:1 C15:0 C16:0 C16:1 C17:0 C17:1 C18:0 C18:1s C18:2 C18:3s C20:0 C20:1 C24:0 C24:1 Total 116245 648828 110588 261115 5358484 25621 232927 39212 2294192 1079260 142203 83604 472498 10949 80326 0 10956052 Probably Cholesterol; 1.06 5.92 1.01 2.38 48.91 0.23 2.13 0.36 20.94 9.85 1.30 0.76 4.31 0.10 0.73 0.00 100.00 19626 246229 53083 163802 8576741 8663 407807 13228 4131903 423166 9231 25571 276696 0 41460 131800 14529006 0.14 1.69 0.37 1.13 59.03 0.06 2.81 0.09 28.44 2.91 0.06 0.18 1.90 0.00 0.29 0.91 100.00 34181 390212 45668 154672 3164006 57011 169064 27459 1179750 447253 11625 2063 55331 4915 23679 1368 5768257 0.59 6.76 0.79 2.68 54.85 0.99 2.93 0.48 20.45 7.75 0.20 0.04 0.96 0.09 0.41 0.02 100.00 Biomarkers possibly Dehydroabietic None detected Cholesterol acid Triacylglycerols Possibly C48 TAG Possibly C48 and C50 TAGS Possibly C48 and C50 TAGS Decomposed nut oil; Identification Animal products Large Herbivore; probably present; Conifer Decomposed nut oil products may occur Animal; Low fat content plant; Decomposed nut oil Vessel No. Catalogue No. Sample Mass 76 80 193 7282.880.02.04 7282.932.04.02 7282.619.03.01 11.223 g 12.438 g 7.998 g 283 Table 4 cont’d. Sample descriptions and lipid compositions of Cloudman site pottery residues. Fatty acid 17KMS 10 17KMS 11 (dil) Area Rel% Area Rel% 15.39 19.66 1.83 5.09 27.11 4.03 2.82 1.78 0.00 18.15 1.62 0.00 1.03 1.14 0.35 0.00 100.00 134205 171467 15986 44399 236428 35137 24575 15496 0 158310 14091 0 8940 9970 3011 0 872015 None detected Possibly C48 and C50 TAGS 31613 692023 85011 188585 6454880 7668 186144 42796 1722293 730294 6898 0 82981 16715 117573 5117 10370591 0.30 6.67 0.82 1.82 62.24 0.07 1.79 0.41 16.61 7.04 0.07 0.00 0.80 0.16 1.13 0.05 100.00 Cholesterol Possibly C48 and C50 TAGS Low fat content plant dominate; Medium fat content plants Decomposed nut oil; Animal products present 17KMS 12 (dil) Area Rel% 0.17 7.49 0.53 1.16 57.21 0.06 1.29 0.05 20.79 8.00 0.10 0.18 1.15 0.41 0.27 1.13 100.00 23454 1008727 71555 156103 7705492 7713 173689 7391 2799966 1076937 13716 23893 154636 55356 36972 152403 13468003 None detected C48 and C50 TAGS Decomposed nut oil; Animal products probably present C12:0 C14:0 C14:1 C15:0 C16:0 C16:1 C17:0 C17:1 C18:0 C18:1s C18:2 C18:3s C20:0 C20:1 C24:0 C24:1 Total Biomarkers Triacylglycerols Identification Vessel No. Catalogue No. Sample Mass 101 150 204 7282.787.03.03 7282.939.03.02 7282.619.04.03 14.098 g 10.813 g 10.461 g 284 Table 4 cont’d. Sample descriptions and lipid compositions of Cloudman site pottery residues. Fatty acid 17KMS 13 17KMS 14 17KMS 15 Area Rel% Area Rel% Area Rel% C12:0 C14:0 C14:1 C15:0 C16:0 C16:1 C17:0 C17:1 C18:0 C18:1s C18:2 C18:3s C20:0 C20:1 C24:0 C24:1 Total 103892 1576100 184746 419886 10334127 9648 379951 55212 3615596 583078 115855 109920 331836 5776 117573 5117 17948313 0.58 8.78 1.03 2.34 57.58 0.05 2.12 0.31 20.14 3.25 0.65 0.61 1.85 0.03 0.66 0.03 100.00 511414 12123650 931544 2442276 55042282 355035 1039082 2084691 20122359 5206328 124905 68268 475534 692948 7776 0 101228092 0.51 11.98 0.92 2.41 54.37 0.35 1.03 2.06 19.88 5.14 0.12 0.07 0.47 0.68 0.01 0.00 100.00 158478 2455800 230086 541595 12042466 137713 128874 58471 3018355 1548211 91499 156049 182079 13075 165064 5333 20933148 0.76 11.73 1.10 2.59 57.53 0.66 0.62 0.28 14.42 7.40 0.44 0.75 0.87 0.06 0.79 0.03 100.00 Biomarkers None detected None detected Cholesterol; possibly Dehydroabietic acid Triacylglycerols Possibly C48 and C50 TAGS C48 TAG C48 TAG Identification Vessel No. Catalogue No. Sample Mass Decomposed nut oil; Low fat content plant; Decomposed nut oil; Low fat content plant; Animal products Animal products present 25 present 70 Decomposed nut oil; Low fat content plant; Animal products present; Conifer products may occur 162 7282.479.03.02 7282.2.07.02 L3 7282.787.01.03 9.935 g 9.892 g 11.487 g 285 Table 4 cont’d. Sample descriptions and lipid compositions of Cloudman site pottery residues. Fatty acid 17KMS 16 17KMS 17 17KMS 18 Area Rel% Area Rel% Area Rel% C12:0 C14:0 C14:1 C15:0 C16:0 C16:1 C17:0 C17:1 C18:0 C18:1s C18:2 C18:3s C20:0 C20:1 C24:0 C24:1 Total 8957 32622 0 13242 410092 1596 12509 1643 122064 262576 17937 3664 9998 15313 5641 0 917854 0.98 3.55 0.00 1.44 44.68 0.17 1.36 0.18 13.30 28.61 1.95 0.40 1.09 1.67 0.61 0.00 100.00 20688 48199 0 16351 417884 13213 9152 1385 37519 153170 17947 3361 2631 8343 3188 0 753031 2.75 6.40 0.00 2.17 55.49 1.75 1.22 0.18 4.98 20.34 2.38 0.45 0.35 1.11 0.42 0.00 100.00 13698 53277 0 38513 522847 0 34625 3953 152667 83981 9278 0 24000 4678 4013 0 945530 1.45 5.63 0.00 4.07 55.30 0.00 3.66 0.42 16.15 8.88 0.98 0.00 2.54 0.49 0.42 0.00 100.00 Biomarkers present; β-sitosterol Cholesterol derivative probably present Possibly dehydroabietic acid None detected Triacylglycerols Traces of TAGs Traces of TAGs C48 TAG Identification Sample No. Catalogue No. Sample Mass Moderate-high fat content; Animal and plant products present 20 Decomposed nut oil, medium fat content; low fat content plants; conifer products may be present 24 Decomposed nut oil; Low fat content plant, Animal products present 43 7282.481.06.03.01 7282.479.03.01 7282.567.02.03 9.569 g 7.972 g 9.564 g 286 Table 4 cont’d. Sample descriptions and lipid compositions of Cloudman site pottery residues. Fatty acid C12:0 C14:0 C14:1 C15:0 C16:0 C16:1 C17:0 C17:1 C18:0 C18:1s C18:2 C18:3s C20:0 C20:1 C24:0 C24:1 Total 17KMS 19 (dil) 17KMS 20 17KMS 21 Area Rel% Area 22254 995328 0 145750 5275432 121792 94402 28387 636099 318151 2055 0 24660 6262 14016 0 7684588 0.29 12.95 0.00 1.90 68.65 1.58 1.23 0.37 8.28 4.14 0.03 0.00 0.32 0.08 0.18 0.00 100.00 35613 334540 0 155045 3074762 22446 135898 21452 711362 249618 5761 0 29180 3409 11134 0 4790220 Rel% 0.74 6.98 0.00 3.24 64.19 0.47 2.84 0.45 14.85 5.21 0.12 0.00 0.61 0.07 0.23 0.00 100.00 Area Rel% 11346 39238 0 9470 285917 20470 9391 1880 114865 92464 4547 0 6797 9383 2518 0 608286 1.87 6.45 0.00 1.56 47.00 3.37 1.54 0.31 18.88 15.20 0.75 0.00 1.12 1.54 0.41 0.00 100.00 Biomarkers Possibly Cholesterol None detected possibly Dehydroabietic Possibly Cholesterol; acid Triacylglycerols Traces of TAGs Traces of TAGs Traces of TAGs Identification Decomposed nut oil; Low fat content plants Sample No. Catalogue No. Sample Mass 55 (exfoliated) 7282.567.02.003 7.504 g Decomposed nut oil; Medium fat content; Low fat content plants; Animal products may be Animal products, Low Fat Content plants probably present; conifer products present 75 may occur 77 7282.880.03.04 7282.889.03.01 5.706 g 13.212 g 287 Table 4 cont’d. Sample descriptions and lipid compositions of Cloudman site pottery residues. Fatty acid 17KMS 22 17KMS 23 Area Rel% Area Rel% 17KMS 24 (dil) Area Rel% C12:0 C14:0 C14:1 C15:0 C16:0 C16:1 C17:0 C17:1 C18:0 C18:1s C18:2 C18:3s C20:0 C20:1 C24:0 C24:1 Total Biomarkers 62823 85303 0 24135 688206 1340 39024 4643 332632 123275 10467 1553 12084 5656 4679 0 1395820 4.50 6.11 0.00 1.73 49.30 0.10 2.80 0.33 23.83 8.83 0.75 0.11 0.87 0.41 0.34 0.00 100.00 Possibly Dehydroabietic acid 43238 84902 0 23950 457996 5748 14832 2827 111618 84015 5990 0 6241 2317 1736 0 845410 5.11 10.04 0.00 2.83 54.17 0.68 1.75 0.33 13.20 9.94 0.71 0.00 0.74 0.27 0.21 0.00 100.00 26934 329217 0 179630 4596852 92851 496245 25924 6093217 1025651 76158 3818 229863 4780 10718 0 13191858 0.20 2.50 0.00 1.36 34.85 0.70 3.76 0.20 46.19 7.77 0.58 0.03 1.74 0.04 0.08 0.00 100.00 None detected None detected C48:C50:C52:C54 Triacylglycerols Traces of TAGs Traces of TAGs ratio Identification Sample No. Catalogue No. Sample Mass Decomposed nut oil; Animal products; Low fat content plants; Conifer products may be present 100 7282.918.05 9.084 g Decomposed nut oil; Low fat content plants; Animal products may be present 146 1.0: 3.1: 4.3: 2.2 High C18:0 Large Herbivore 164 7282.823.03.03 7282.770.02.1 9.001 g 7.727 g 288 Table 4 cont’d. Sample descriptions and lipid compositions of Cloudman site pottery residues. Fatty acid C12:0 C14:0 C14:1 C15:0 C16:0 C16:1 C17:0 C17:1 C18:0 C18:1s C18:2 C18:3s C20:0 C20:1 C24:0 C24:1 Total 17KMS 26 17KMS 27 17KMS 28 Area Rel% Area Rel% Area 38880 91696 0 32639 877764 2991 36044 6314 216346 91211 10908 0 10926 0 8054 0 1423773 2.73 6.44 0.00 2.29 61.65 0.21 2.53 0.44 15.20 6.41 0.77 0.00 0.77 0.00 0.57 0.00 100.00 12450 155177 0 59107 2540048 50849 72316 26068 592690 83905 1218 0 20587 2300 14661 0 3631376 0.34 4.27 0.00 1.63 69.95 1.40 1.99 0.72 16.32 2.31 0.03 0.00 0.57 0.06 0.40 0.00 100.00 54391 130565 0 127847 1145591 75290 176717 10525 1227835 140712 11362 0 74439 3938 12868 1117 3193197 Rel% 1.70 4.09 0.00 4.00 35.88 2.36 5.53 0.33 38.45 4.41 0.36 0.00 2.33 0.12 0.40 0.03 100.00 Biomarkers Stigmasterol; Probably Cholesterol None detected Cholesterol; possibly Dehydroabietic acid C48 TAG larger than Triacylglycerols Traces of TAGs C50 TAG; traces of C52 Traces of TAGs and C54 TAG occur Decomposed nut oil; Low fat content plants; Animal product; Conifer products present 175 Decomposed nut oil; Animal products present Large herbivore; low fat content plant roots may be present 173 191 7282.000.033 10.314 g 7282.989.031.04 7282.880.04.03 10.158 g 13.843 g 289 Identification Sample No. Catalogue No. Sample Mass Table 4 cont’d. Sample descriptions and lipid compositions of Cloudman site pottery residues. Fatty acid 17KMS 29 17KMS 30 Area Rel% Area Rel% C12:0 C14:0 C14:1 C15:0 C16:0 C16:1 C17:0 C17:1 C18:0 C18:1s C18:2 C18:3s C20:0 C20:1 C24:0 C24:1 Total 36211 154294 0 86053 1060792 15778 60611 12836 285590 131465 13667 1454 19437 7120 12621 0 1897929 1.91 8.13 0.00 4.53 55.89 0.83 3.19 0.68 15.05 6.93 0.72 0.08 1.02 0.38 0.66 0.00 100.00 75215 424413 0 109879 4235558 79234 139478 24653 1312648 356759 20890 2164 89568 12443 35651 0 6918553 1.09 6.13 0.00 1.59 61.22 1.15 2.02 0.36 18.97 5.16 0.30 0.03 1.29 0.18 0.52 0.00 100.00 Biomarkers Cholesterol Possibly Cholesterol; probably Dehydroabietic acid Triacylglycerols Traces of TAGs Traces of TAGs Identification products; Low fat content Decomposed nut oil; Animal Sample No. Catalogue No. Sample Mass plants 215 7282.619.041.04 7.179 g Decomposed nut oil; Animal products; Low fat content plants may be present; Conifer products probably present 216 7282.219.04.02 10.913 g Table 5. Sample description and lipid biomarkers in residue 17KMS 25. Biomarkers Possibly Cholesterol; possibly Dehydroabietic acid Triacylglycerols None detected Identification Sample No. Catalogue No. Sample Mass Animal products may be present; Conifer products may be present 166 7282.029.06.2 11.791 g 290 Table 6. Sample descriptions and lipid compositions of Cloudman site soil lipids. Fatty acid C12:0 C14:0 C14:1 C15:0 C16:0 C16:1 C17:0 C17:1 C18:0 C18:1s C18:2 C18:3s C20:0 C20:1 C20:2 C20:3 C20:4 C20:3/C20:5 C22:0 C22:1 C22:2 C22:4 C22:5 C24:0 C22:6 C24:1 Total 17KMS 31 17KMS 32 17KMS 33 Area Rel% Area Rel% Area 9326 39594 1558 19153 467886 289759 9897 30491 121865 1092745 227452 10533 126476 12194 24418 126042 21697 335021 3022 8049 85410 657241 42601 3767 95820 36465 3898482 0.24 1.02 0.04 0.49 12.00 7.43 0.25 0.78 3.13 28.03 5.83 0.27 3.24 0.31 0.63 3.23 0.56 8.59 0.08 0.21 2.19 16.86 1.09 0.10 2.46 0.94 100.00 4224 21475 1758 17110 553842 80474 7022 21917 104298 2071376 330916 22096 54270 16304 37708 17696 13344 107544 16848 2875 57832 196051 3578 5103 73928 19990 3859579 0.11 0.56 0.05 0.44 14.35 2.09 0.18 0.57 2.70 53.67 8.57 0.57 1.41 0.42 0.98 0.46 0.35 2.79 0.44 0.07 1.50 5.08 0.09 0.13 1.92 0.52 100.00 6736 48301 4513 28181 555104 160486 17520 21828 257871 422748 166269 50039 215590 14231 899 4275 31798 463143 22009 26070 3660 627153 26850 13192 105109 0 3293575 Rel% 0.20 1.47 0.14 0.86 16.85 4.87 0.53 0.66 7.83 12.84 5.05 1.52 6.55 0.43 0.03 0.13 0.97 14.06 0.67 0.79 0.11 19.04 0.82 0.40 3.19 0.00 100.00 Biomarkers Cholesterol; Stigmasterol; β- Cholesterol; Stigmasterol; β- sitosterol; possibly Dehydroabietic acid sitosterol; possibly Dehydroabietic acid Cholesterol; Stigmasterol; β- sitosterol Triacylglycerols Plant products Plant products Plant products Identification Possible Combination of Cultural and Natural Soil Lipids; Conifer products may occur Combination of High Fat Content Cultural Lipids and Natural Soil Lipids; Conifer products occur Natural Soil Lipids Dominant Sample No. Sample Mass T2-S1-2 9.211 g T2-S2-2 10.070 g T3-S1-1 10.105 g 291 Table 7. Experimental cooking residue of crushed acorn with extrapolation of further degradation compared to the relative fatty acid composition of residue 17KM 5. Fatty acid MQ 19D Rel% Recalculated MQ 19D (C18:1s = 3.73%) Area Rel% 17KMS 5 Rel% C12:0 C14:0 C14:1 C15:0 C16:0 C16:1 C17:0 C17:1 C18:0 C18:1s C18:2 C18:3s C20:0 C20:1 C24:0 C24:1 Total 0.30 3.66 0.03 0.18 28.27 0.17 0.71 0.00 4.64 56.04 3.78 0.14 1.01 0.43 0.65 0.00 100.00 0.30 3.66 0.03 0.18 28.27 0.17 0.71 0.00 4.64 1.54 0.00 0.00 1.01 0.00 0.65 0.00 41.14 0.72 8.90 0.07 0.43 68.71 0.41 1.73 0.00 11.27 3.73 0.00 0.00 2.45 0.00 1.57 0.00 100.00 0.30 3.81 0.61 1.68 71.07 0.12 2.41 0.11 14.29 3.73 0.07 0.26 0.87 0.08 0.58 0.01 100.00 Cooked crushed acorns after 30 Extrapolation of Decomposition: Decomposed nut oil; Identification days decomposition in 75°C oven Crushed acorns when C18:1 isomers drop to possible traces of animal products 3.73% 292 APPENDIX E: Microbotanical Analysis Data 293 Vessel No. Phytoliths Starch Maize Wild Rice Squash Maize Squash 1 4 5 6 8 10 12 22 23 25 26 28 34 35 40/153 41 43 46 50 70 75 76 80 81 88 101 102 103 105 109 112 114 118 120 122 124 132 146 150 152 162 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 2 1 0 0 0 0 0 0 0 0 1 1 0 0 0 0 1 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 1 1 0 9 0 0 0 0 0 0 3 0 1 0 0 0 0 0 1 0 0 0 0 1 1 1 0 1 0 0 1 0 1 0 0 0 0 0 0 0 0 0 1 0 0 294 0 2 1 4 1 0 0 2 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Vessel No. 173 174 175 179 193 204 215 Phytoliths Starch Maize Wild Rice Squash Maize Squash 1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 2 2 0 0 0 0 0 0 0 0 0 0 0 0 295 APPENDIX F: Stable Isotope, Microbotanical, and Lipid Residue Analysis Results by Vessel 296 297 13CVPDB15NAirMaizeWild RiceSquashNut OilLarge HerbivoreMod-High FatMedium Fat (maize, fish)Low Fat PlantsAnimal ProductPossible AnimalPossible PlantConifer Product1-26.129.630001001000004-26.3612.83100nananananananana05-26.112.81110nananananananana06-23.2110.241101000001008-23.7711.93110nananananananana010-26.5312.98000nananananananana012-24.4511.49001nananananananana020nanananana00101100022-26.6611.98100nananananananana023-25.910.93001nananananananana024nanananana10010000025-26.6511.5000010001100026-26.311.65000nananananananana028-26.813.44000nananananananana034-24.7411.2600000011000035-21.7813.2811010000000036-2310.29nanananananananananana040/153-26.4611.43111nananananananana041-25.6811.85000nananananananana042-26.310.20nanananananananananana043-23.6611.5300010001100046nana000nananananananana050-25.0112.81000nananananananana055nanananana10001000070-26.4112.5111110001100075-24.4910.3300110001010076-22.6511.67001100001000MICROBOTANICALSLIPIDSSTABLE ISOTOPESVESSEL 298 13CVPDB15NAirMaizeWild RiceSquashNut OilLarge HerbivoreMod-High FatMedium Fat (maize, fish)Low Fat PlantsAnimal ProductPossible AnimalPossible PlantConifer Product77nanananana00011100080-22.9910.5800011000000081-27.0212.66001nananananananana088-28.365.17000nananananananana0100-25.1813.76nanana100011000101-27.0413.77010000110000102-27.3313.37011nananananananana0103-24.6512.35000nananananananana0105-27.213.33001nananananananana0109-24.0311.90100100001000112-27.4312.31000nananananananana0114nana000nananananananana0118nana000nananananananana0120-26.8410.22100nananananananana0122-26.6311.96100nananananananana0124-26.5612.88000nananananananana0131nanananana000011000132-24.4310.28000nananananananana0146-29.1310.32010100010100150-24.810.57011100001000152-25.0311.84000nananananananana0162-25.3111.01010100011001164nanananana010000000166nanananana000000100173-27.2512.93100100001000174-27.3412.78100nananananananana0175-22.8813.27000100011001179-25.298.74000nananananananana0VESSELSTABLE ISOTOPESMICROBOTANICALSLIPIDS 299 13CVPDB15NAirMaizeWild RiceSquashNut OilLarge HerbivoreMod-High FatMedium Fat (maize, fish)Low Fat PlantsAnimal ProductPossible AnimalPossible PlantConifer Product191-27.2910.58nanana010000010193-25.6112.52000100011000204-25.9413.57010100001000205-28.8911.69nanananananananananana0215-22.7811.71000100011000216nanananana100001011VESSELSTABLE ISOTOPESMICROBOTANICALSLIPIDS APPENDIX G: Select Vessels from the Cloudman Pottery Assemblage 300 MIDDLE WOODLAND VESSELS Figure G1. Laurel Pseudo-scallop Shell Figure G2. Vessel 5, Laurel Pseudo-scallop Shell 301 Figure G3. Vessel 10, Laurel Pseudo-scallop Shell Figure G4. Vessel 22, Laurel Pseudo-scallop Shell Figure G5: Vessel 28, Laurel Pseudo-scallop Shell 302 Figure G6. Vessel 4, Laurel Dentate Stamped Figure G7: Vessel 12, Laurel Dentate Stamped (oblique) Figure G8. Vessel 20, Laurel Dentate Stamped (oblique) 303 Figure G9. Vessel 23, Laurel Banked Linear Stamped Figure G10. Vessel 109, Laurel Banked Linear Stamped Figure G11. Vessel 112, Laurel Banked Linear Stamped 304 Figure G12. Vessel 114, Laurel Banked Linear Stamped Figure G13. Vessel 6, Laurel Dentate Rocker Stamped Figure G14: Vessel 131, North Bay Linear Stamped 305 MIDDLE WOODLAND/LATE WOODLAND TRANSITION VESSELS Figure G15. Vessel 35, Late Laurel (cf. Laurel Incised or Mackinac Banded) Figure G16. Vessel 118, Untyped (Middle/Late Woodland Transition) 306 Figure G17. Vessel 33, Untyped (incipient Blackduck?) 307 EARLY LATE WOODLAND VESSELS Figure G18. Vessel 80, Mackinac Punctate Figure G19. Vessel 191, Mackinac Punctate 308 Figure G20. Vessel 100, Mackinac Punctate Figure G21. Vessel 105, Mackinac Punctate Figure G22. Vessel 175, Mackinac Punctate 309 Figure G23. Vessel 50, Mackinac Banded Figure G24. Vessel 103, Mackinac Banded Figure G25. Vessel 124, Mackinac Banded 310 Figure G26. Vessel 120, Mackinac Banded Figure G27. Vessel 173, Mackinac Banded 311 Figure G28. Vessel 76, Mackinac Undecorated Figure G29. Vessel 55, Mackinac Ware 312 Figure G30. Vessel 8, Mackinac Ware Figure G31. Vessel 34, Mackinac Ware Figure G32. Vessel 41, Mackinac Ware 313 Figure G33. Vessel 46, Mackinac Ware Figure G34. Vessel 122, Mackinac Ware (cf. Punctate) Figure G35. Vessel 132, Mackinac Ware (cf. Punctate) 314 Figure G36. Vessel 174, Mackinac Ware (cf. Punctate) Figure G37. Vessel 81, Blackduck Banded Figure G38. Vessel 88, Blackduck Banded 315 Figure G39. Vessel 193, Blackduck Banded Figure G40. Vessel 199, cf. Bowerman Plain v. Cordmarked 316 EARLY LATE/MIDDLE LATE WOODLAND TRANSITIONAL AND MIDDLE LATE WOODLAND VESSELS Figure G41. Vessel 200, Untyped (ELW/MLW Transition) Figure G42. Vessel 42, Bois Blanc Ware Figure G43. Vessel 215, Bois Blanc Ware 317 LATE LATE WOODLAND VESSELS Figure G44. Vessel 24, “Proto-Juntunen” Ware (plain) Figure G45. Vessel 102, Juntunen Drag-and-Jab 318 Figure G46. Vessel 205, Juntunen Linear Puncate Figure G47. Vessel 204, Juntunen Linear Punctate 319 Figure G48. Vessel 213, Juntunen Ware (cf. Drag-and-Jab) Figure G49. Vessel 25, Juntunen Ware 320 Figure G50. Vessel 26, Juntunen Ware Figure G51. Vessel 101, Juntunen Ware Figure G52. Vessel 152, Late Juntunen Ware (cf. O’Neil Curvilinear) 321 Figure G53. Vessel 43, Traverse Decorated v. Punctate Figure G54. Vessel 150, Traverse Plain v. Scalloped Figure G55. Vessel 216, Untyped 322 ONTARIO IROQUOIS VESSELS Figure G56. Vessel 146, Early Ontario Iroquoian Figure G57. Vessel 40/153, cf. Lawson Opposed or Methodist Point Group 7 323 Figure G58. Vessel 162, cf. Sidey Notched or Lawson Incised Figure G59. Vessel 64, cf. Ripley Plain 324 Figure G60. Vessel 70, cf. Huron Incised Figure G61. Vessel 74, cf. Huron Incised 325 Figure G62. Vessel 36, cf. Huron Incised Figure G63. Vessel 155, cf. Huron Incised Figure G64. Vessel 156, cf. Huron Incised 326 Figure G65. Vessel 164, cf. Huron Incised Figure G66. Vessel 166, cf. Huron Incised 327 Figure G67. Vessel 179, cf. Huron Incised Figure G68. Vessel 77, Untyped 328 MINIATURE VESSELS Figure G69. Vessel 39, Laurel Ware (Middle Woodland) Figure G70. Vessel 201, Untyped (cf. Hopewellian, Middle Woodland) 329 Figure G71. Vessel 52, Mackinac Undecorated (Early Late Woodland) Figure G72. Vessel 53, Mackinac Punctate (Early Late Woodland) 330 Figure G73. Vessel 83, Mackinac Ware (Early Late Woodland) Figure G74. Vessel 202, Mackinac Ware (Early Late Woodland) Figure G75. Vessel 63, Untyped (Early Late Woodland) 331 Figure G76. Vessel 75, Traverse Plain v. Scalloped (Late Late Woodland) Figure G77. Vessel 54, Untyped (cf. O’Neill site cup; Late Late Woodland) 332 Figure G78. Vessel 167, cf. Huron Incised (Ontario Iroquois) Figure G79. Vessel 182, cf. Huron Incised (Ontario Iroquois) 333 REFERENCES 334 REFERENCES Abrams, E. M. 2009 Hopewell Archaeology: A View from the Northern Woodlands. Journal of Archaeological Research 12(2): 169-204. Albert, Dennis A. Shirley R. Denton, and Burton V. Barnes 1986 Regional Landscape Ecosystems of Michigan. School of Natural Resources, University of Michigan, Ann Arbor. Albert, Rebecca, Susan M. Kooiman, Caitlin Clark, and William A. Lovis 2018 Earliest Microbotanical Evidence for Maize in the Northern Lake Michigan Basin. American Antiquity 83(2):345-355. Aldenderfer, Mark S., and Roger K. Blashfield 1984 Cluster Analysis. SAGE Publications, Newbury Park, CA. Ambrose, Stephen H., and Michael J. DeNiro 1986 Reconstruction of African Human Diet Using Bone Collagen Carbon and Nitrogen Isotope Ratios. Nature 319:321-324. Anderson, David G. and Robert C. Mainfort, Jr. 2002 An Introduction to Woodland Archaeology in the Southeast. In The Woodland Southeast, edited by D. G. Anderson and R. C. Mainfort, Jr., pp. 1-19. The University of Alabama Press, Tuscaloosa. Anderson, David G., and Kenneth E. Sassaman 2004 Early and Middle Holocene periods, 9500 to 3750 B.C. In Handbook of North American Indians—Southeast, edited by Raymond D. Fogelson, pp. 87-100. Smithsonian Institution Press, Washington. Anderson, Shelby L., Shannon Tushingham, and Tammy Y. Buonasera 2017 Aquatic Adaptations and the Adoption of Arctic Pottery Technology: Results of Residue Analysis. American Antiquity 82(3):452-479. Appadurai, Arjun 1981 Gastro-Politics in Hindu South Asia. American Ethnologist 8(3):494-511. Aronson, Meredith, James M. Skibo, and Miriam T. Stark 1994 Production and Use Technologies in Kalinga Pottery. In Kalinga Ethnoarchaeology: Expanding Archaeological Method and Theory, edited by William A. Longacre and James M. Skibo, pp. 83-112. Smithsonian Institution, Washington, D.C. 335 Arzigian, Constance 2000 Middle Woodland and Oneota Contexts for Wild Rice Exploitation in Southwestern Wisconsin. Midcontinental Journal of Archaeology 25:245-268. Atalay, Sonya and Christine A. Hastorf 2006 Food, Meals, and Daily Activities: Food Habitus at Neolithic Catalhoyuk. American Antiquity 71(2):283-319. Bamann, Susan, Robert Kuhn, James Molnar, and Dean Snow 1992 Iroquoian Archaeology. Annual Review of Anthropology 21:435-460. Baraga, Frederic 1976 Chippewa Indians, as Recorded by Rev. Frederick Baraga in 1847. Studia Slovenica, New York. Barrett, James H., Roelf P. Beukens, and Rebecca A. Nicholson 2001 Diet and Ethnicity During the Viking Colonization of Northern Scotland: Evidence from Fish Bones and Stable Carbon Isotopes. Antiquity 75:145-54. Barton, Huw, Robin Torrence, and Richard Fullagar 1998 Clues to Stone Tool Function Re-examined: Comparing Starch Grain Frequencies on Used and Unused Obsidian Artefacts. Journal of Archaeological Science 25:1231-1238. Benz, Charmaine M. and R. Todd Williamson (editors) 2005 Diba Jimooyung, Telling Our Story: A History of the Saginaw Ojibwe Anishinabek. Saginaw Chippewa Indian Tribe of Michigan and the Ziibiwing Cultural Society, Mt. Pleasant, MI. Beoku-Betts, Josephine 1995 We Got Our Way of Cooking Things: Women, Food, and Preservation of Cultural Identity among the Gullah. Gender and Society 9(5):535-555. Bergman, Ingela, Lars Ostlund, and Olle Zackrisson 2004 The Use of Plants as Regular Food in Ancient Subarctic Economies: A Case Study Based on Sami Use of Scots Pine Innerbark. Arctic Anthropology 41(1):1-13. Bianchi , Thomas 1974 Description and Analysis of the Prehistoric Ceramic Materials Recovered on the Winter Site. Master’s thesis, Department of Anthropology, Western Michigan University, Kalamazoo. Binford, Lewis R. 1962 Archaeology as Anthropology. American Antiquity 28(2):217-225. 1965 Archaeological Systematics and the Study of Culture Process. American Antiquity 31(2):203-210. 336 1968 Post-Pleistocene Adaptations. In New Perspectives in Archaeology, edited by Sally Binford and Lewis R. Binford, pp. 313-341. Aldine, Chicago. 1980 Willow Smoke and Dogs’ Tails: Hunter-Gatherer Settlement Systems and Archaeological Site Formations. American Antiquity 45:4-20. Birch, Jennifer, and Ronald F. Williamson 2012 The Mantle Site: An Archaeological History of an Ancestral Wendat Community. AltaMira, New York. Blewett, William L., David P. Lusch, and Randall Schaetzl 2009 The Physical Landscape: A Glacial Legacy. In Michigan Geography and Geology, edited by R. Schaetzl, J. Darden, and Danita Brandt, pp. 249-273. Custom Publishing, New York. Blitz, John H. 1993 Big Pots for Big Shots: Feasting and Storage in a Mississippian Community. American Antiquity 58(1):80-96. Boyd, M., and C. Surette 2010 Northernmost Precontact Mazie in North America. American Antiquity 75(1):117-133. Boyd, M., Surette, C., Lints, A, and Hamilton, S. Wild Rice (Zizania spp.), the Three Sisters, and the Woodland Tradition in Western and Central Canada. Midwest Archaeological Conference Inc. Occasional Papers 1: 7-32. Boyd, M., C. Surette, J. Surette, I. Therriault, and S. Hamilton 2013 Holocene Paleoecology of a wild rice (Zizania sp.) Lake in Northwestern Ontario, Canada. Journal of Paleolimnology 50:365-377. Boyd, M., T. Varney, C. Surette, and J. Surette 2008 Reassessing the Northern Limit of Maize Consumption in North America: Stable Isotope, Plant Microfossil, and Trace Element Content of Carbonized Food Residue. Journal of Archaeological Science 35:2545-2556. Brandt, Kari L. 1996 The Effects of Early Agriculture on Native North American Populations: Evidence from the Teeth and Skeleton. Unpublished Ph.D. dissertation, The University of Michigan, Ann Arbor. Branstner, Christine N. 1992 National Register of Historic Places Archaeological Testing of 20CH6: A Multicomponent Site on Drummond Island, Michigan, 1991 Investigations. Report prepared for the Bureau of History, Michigan Department of State, Lansing, MI. 337 1995 Archaeological Investigations at the Cloudman Site (20CH6): A Multicomponent Native American Occupation on Drummond Island, Michigan, 1992 and 1994 Excavations. Report on file at the Consortium for Archaeological Research, Michigan State University, East Lansing. Brashler, Janet G., Elizabeth B. Garland, Margaret B. Holman, William A. Lovis, and Susan R. Martin 2000 Adaptive Strategies and Socioeconomic Systems in Northern Great Lakes Riverine Environments: The Late Woodland of Michigan. In Late Woodland Societies: Tradition and Transformation Across the Midcontinent, edited by T.E. Emerson, D.L McElrath, and A.C. Fortier, pp. 543-579. University of Nebraska Press, Lincoln. Braun, David P. 1983 Pots as Tools. In Archaeological Hammers and Theories, edited by J.A. Moore and A.S. Keene, pp. 107-134. Academic Press, New York. Braun, David P. and Stephen Plog 1982 Evolution of “Tribal” Social Networks: Theory and Prehistoric North American Evidence. American Antiquity 47(3):504-525. Bronitsky, Gordon and Robert Hamer 1986 Experiments in Ceramic Technology: The Effects of Various Tempering Materials on Impact and Thermal-Shock Resistance. American Antiquity 51(1):89-101. Brose, David S. 1970 The Archaeology of Summer Island: Changing Settlement Systems in Northern Lake Michigan. Anthropological Papers No. 41. Museum of Anthropology, University of Michigan, Ann Arbor. Brose, David S. and Michael J. Hambacher 1999 The Middle Woodland in Northern Michigan. In Retrieving Michigan’s Buried Past: The Archaeology of the Great Lakes State, edited by John R. Halsey, pp. 173- 192. Cranbrook Institute of Science, Bloomfield Hills, Michigan. Brugam, R. B., K. Little, L. Kohn, P. Brunkow, G. Vogel, and T. Martin 2017 Tracking Change in the Illinois River Using Stable Isotopes in Modern and Ancient Fishes. River Research and Applications 33:341-352. Burchill, Alexandra 2014 Plant Microfossil Analysis of Middle Woodland Food Residues, Northern Minnesota. Unpublished MA thesis, Dept. of Environmental Studies, Lakehead University, Thunder Bay, Ontario. Burchill, Alexandra, and Matthew J. Boyd 2015 Analysis and Dietary Implications of Plant Microfossils on Middle Woodland Food Residues, Northern Minnesota. Minnesota Archaeologist 74:107-121. 338 Carroll, Jon W. 2013 Simulating Springwells: A Complex Systems Approach Toward Understanding Late Prehistoric Social Interaction in the Great Lakes Region of North America. Ph.D. dissertation, Department of Anthropology, Michigan State University, East Lansing. Chang, Cecily C.Y., Carol Kendall, Steven R. Silva, William A. Battaglin, Donald H. Campbell 2002 Nitrate Stable Isotopes: Tools for Determining Nitrate Sources Among Different Land Uses in the Mississippi River Basin. Canadian Journal of Fisheries and Aquatic Sciences 59:1874-1885. Chappuis, Eglantine, Vanesa Seriñá, Eugénia Martí, Enric Ballesteros, and Esperanca Gacia 2017 Decrypting Stable-Isotope (δ13C and δ15N) Variability in Aquatic Plants. Freshwater Biology 62:1807-1818. Chase, Brad 2012 Crafting Harappan Cuisine on the Saurashtran Frontier of the Indus Civilization. In The Menial Art of Cooking, edited by Sarah R. Graff and Enrique Rodríguez- Alegría, pp. 145-172. University Press of Colorado, Boulder. Childe, V. Gordon 1936 Man Makes Himself. New American Library of World Literature, New York (1951 reprint). Chivis, Jeff 2016 The Introduction of Havana-Hopewell in West Michigan and Northwest Indiana: An Integrative Approach to the Identification of Communities, Interaction Networks, and Mobility Patterns. Ph.D. dissertation, Department of Anthropology, Michigan State University, East Lansing. Chu, G. P. K. 1968 Microstructure of Complex Ceramics. In Ceramic Microstructures: Their Analysis, Significance, and Production, edited by R.M. Fulrath and J.A. Pask, pp. 828-62. John Wiley, New York. Cleland, Charles E. 1966 The Prehistoric Animal Ecology and Ethnozoology of the Upper Great Lakes Region. Anthropological Papers, Museum of Anthropology, University of Michigan No. 29. The University of Michigan, Ann Arbor. 1982 The Inland Shore Fishery of the Northern Great Lakes: Its Development and Importance in Prehistory. American Antiquity 47(4):761-784. 1989 Comments on “A Reconsideration of Aboriginal Fishing Strategies in the Northern Great Lakes Region” by Susan R. Martin. American Antiquity 54(3):605-609. 339 1999 Cultural Transformation: The Archaeology of Historic Indian Sites in Michigan, 1670-1940. In Retrieving Michigan’s Buried Past: The Archaeology of the Great Lakes State, edited by John R. Halsey, pp. 279-290. Cranbrook Institute of Science, Bloomfield Hills, Michigan. Cleland, Charles E., Margaret B. Holman, and J. Alan Holman 1988 The Mason-Quimby Line Revisited. The Wisconsin Archeologist 79(1):8-27. Cloern, James E., Elizabeth A. Canuel, and David Harris 2002 Stable Carbon and Nitrogen Isotope Composition of Aquatic and Terrestrial Plants of the San Francisco Bay Estuarine System. Limnology and Oceanography 47(3):713-729. Comer, P., D. Albert, H. Wells, B. Hart, J. Raab, D. Price, D. Kashian, and R. Corner 1995 Michigan’s Presettlement Vegetation, as Interpreted from the General Land Office Surveys 1816-1856. Michigan Natural Features Inventory, Lansing, Michigan. Comer, P., D. Albert, H 1997 Vegetation circa 1800 of Chippewa County, Michigan, East Part: An Interpretation of the General Land Office Surveys. Michigan Natural Features Inventory, Lansing, Michigan. https://mnfi.anr.msu.edu/data/veg1800/chippewa_east.pdf (accessed 3/7/18). Condamin, J., F. Formenti, M. O. Metais, M. Michel, and P. Blond 1976 The Application of Gas Chromatography to the Tracing of Oil in Ancient Ampohorae. Archaeometry 18(2):195-201. Conway, Thor A. 1977 Whitefish Island: A Remarkable Archaeological Site at Sault Ste. Marie, Ontario. Ministry of Culture and Recreation, Historical Planning and Research Branch, Ottawa, Ontario Cooper, Janet 1996 Cloudman Site (20CH6), Drummond Island, Michigan, Features 26 and 27, 1992 Excavations. On file at the Consortium for Archaeological Research, Department of Anthropology, Michigan State University, East Lansing. Craig, O. E., M. Forster, S. H. Andersen, E. Koch, P. Crombé, N. J. Milner, B. Stern, G. N. Bailey, and C. P. Heron 2007 Molecular and Isotopic Demonstration of the Processing of Aquatic Products in Northern European Prehistoric Pottery. Archaeometry 49(1):135-152. Craig, O.E., H. Saul, A. Lucquin, Y. Nishida, K. Taché, L. Clarke, A. Thompson, D.T. Altoft, J. Uchiyama, M. Ajimoto, K. Gibbs, S. Isaksson, C.P. Heron, and P. Jordan. 2013 Earliest Evidence for the Use of Pottery. Nature 496:351-354. 340 Crawford, Gary W. 2011 People and Plant Interaction in the Northeast. In Subsistence Economies of Indigenous North America, edited by Bruce D. Smith, pp. 431-448. Smithsonian Institution Scholarly Press, Washington, D.C. Danziger, Edmund J. 1978 The Chippewas of Lake Superior. University of Oklahoma, Norman. Densmore, Frances 1979 Chippewa Customs. Minnesota Historical Society Press, St. Paul. 2005 Strength of the Earth: The Classic Guide to Ojibwe Uses of Native Plants. Minnesota Historical Society Press, St. Paul. Dewar, G., J.K. Ginter, B.A.S. Shook, N. Ferris, H. Henderson 2010 A Bioarchaeological Study of a Western Basin Tradition Cemetery on the Detroit River. Journal of Archaeological Science 37:2245-2254. Dorothy, Lawrence G. 1978 The Ceramics of the Sand Pont Site (20BG14) Baraga County, Michigan: A Preliminary Description. Unpublished Master’s Thesis, Department of Anthropology, Western Michigan University, Kalamazoo. 1980 The Ceramics of the Sand Point Site (20BG14) Baraga County, Michigan: A Preliminary Description. The Michigan Archaeologist 26(3-4):39-90. Drake, Eric C. and Sean B. Dunham 2004 The Woodland Period Occupation of Grand Island. Midcontinental Journal of Archaeology 29(2):133-165. Druc, Isabelle C., Kinya Inokuchi, and Laure Dussubieux 2017 LA-ICP-MS and Petrography to Assess Ceramic Interaction Networks and Production Patterns in Kuntur Wasi, Peru. Journal of Archaeological Science: Reports 12:151-160. Dunham, Sean B. 2009 Nuts about Acorns: A Pilot Study on Acorn Use in Woodland Period Subsistence in the Eastern Upper Peninsula of Michigan. The Wisconsin Archeologist 90(1-2):113-130. 2014 Late Woodland Settlement and Subsistence Patterns in the Eastern Upper Peninsula of Michigan. Ph.D dissertation, Department of Anthropology, Michigan State University, East Lansing. 2017 Location, Location, Location: A Foray into Persistent Places in da UP. Paper presented at the 61st Annual Meeting of the Midwest Archaeological Conference, Indianapolis, IN. 341 Duwe, Samuel and Hector Neff 2007 Glaze and Slip Pigment Analyses of Pueblo IV Period Ceramics from East-Central Arizona using Time of Flight-Laser Ablation-Inductively Coupled Plasma-Mass Spectrometry (TOF-LA-ICP-MS). Journal of Archaeological Science 34:403-414. Eerkens, Jelmer W., and Eric J. Bartelink 2013 Sex-Based Weaning and Early Childhood Diet Among Middle Holocene Hunter- Gatherers in Central California. American Journal of Physical Anthropology 152:471-483. Egan-Bruhy, Kathryn C. 2007 20CH6. Cloudman Site Floral Table. On file at the Consortium for Archaeological Research, Department of Anthropology, Michigan State University, East Lansing. 2014 Ethnicity as Evidenced in Subsistence Patterns of Late Prehistoric Upper Great Lakes Populations. Midwest Archaeological Conference Inc. Occasional Papers 1: 7-32. Evershed, Richard P., Stephanie N. Dudd, Stephanie Charters, Hazel Mottram, Andrew W. Stott, Anthony Raven, Pim F. van Bergen, and Helen A. Bland 1999 Lipids as Carriers of Anthropogenic Signals from Prehistory. Philosophical Transactions of the Royal Society B 354(1379):19-31. Falabella, F., L. Sanhueza, I. Correa, M.D. Glascock, T.J. Ferguson, and E. Fonseca 2013 Studying Technological Practices at a Local Level: Neutron Activation and Petrographic Analyses of Early Ceramic Period Pottery in Central Chile. Archaeometry 55(1):33-53. Feathers, James K. 2006 Explaining Shell-Tempered Pottery in Prehistoric Eastern North America. Journal of Archaeological Method and Theory 13(2):89-133. Fink, Andrea K. 2013 Late Woodland Communities of Practice: An Analysis of Pottery from the Engelbert Site, Nichols, NY. Master’s thesis, Graduate School, Binghamton University, State University of New York, Binghamton. Fischer, Anders and Jan Heinemeier 2003 Freshwater Reservoir Effect in 14C Dates of Food Residue on Pottery. Radiocarbon 45:449-466. Fischler, Claude 1988 Food, Self and Identity. Social Science Information 27(2):275-292. Fitting, James E., David S. Brose, Henry T. Wright, and James Dinerstein 1969 The Goodwin-Gresham Site, 20IS8, Iosco County, Michigan. The Wisconsin Archeologist 50:125-183 342 Flannery, Kent V. 1973 The Origins of Agriculture. Annual Review of Anthropology 2:217-310. Ford, Richard I. and David S. Brose 1975 Prehistoric Wild Rice from the Dunn Farm Site, Leelanau County, Michigan. The Wisconsin Archeologist 56:9-15. Foster, William C. 2012 Climate and Culture Change in North America AD 900-1600. University of Texas Press, Austin. Fournier, Michael R. 2007 The Gyftakis Site: A Reevaluation of a Middle Woodland Site After 30 Years. Unpublished M.A. thesis, Department of Anthropology, Western Michigan University, Kalamazoo. Fox, William A. 1990a The Middle Woodland to Late Woodland Transition. In The Archaeology of Southern Ontario to A.D. 1650, Occasional Publication of the London Chapter, OAS Number 5, edited by Chris J. Ellis and Neal Ferris, pp. 171-188. Ontario Archaeological Society Inc., London. 1990b The Odawa. In The Archaeology of Southern Ontario to A.D. 1650, Occasional Publication of the London Chapter, OAS Number 5, edited by Chris J. Ellis and Neal Ferris, pp. 457-474. Ontario Archaeological Society Inc., London. Fox, William A. and Charles Garrad 2004 Hurons in an Algonquian Land. Ontario Archaeology 77/78:121-134. Franzen, John 1975 An Archaeological Survey of Chippewa County, Michigan. Michigan History Division, Michigan Department of State, Archaeological Survey Reports, Number 5, Lansing. Fuller, Benjamin T., Gundula Muldner, Wim Van Neer, Anton Ervynck, and Michael P. Richards 2012 Carbon and Nitrogen Stable Isotope Ratio Analysis of Freshwater, Brackish and Marine Fish from Belgian Archaeological Sites (1st and 2nd Millennium AD). Journal of Analytical Atomic Spectrometry 27:807-820. García-Granero, Juan José, Carla Lancelotti, and Marco Madella 2017 A Methodological Approach to the Study of Microbotanical Remains from Grinding Stones: A Case Study in Northern Gujarat (India). Vegetation History and Archaeobotany 26:43-57. Garland, Elizabeth B. and Scott G. Beld 1999 The Early Woodland: Ceramics, Domesticated Plants, and Burial Mounds Foretell the Shape of the Future. In Retrieving Michigan’s Buried Past: The 343 Archaeology of the Great Lakes State, edited by John R. Halsey, pp. 125-146. Cranbrook Institute of Science, Bloomfield Hills, Michigan. Glascock, Michael D. and Hector Neff 2003 Neutron Activation Analysis and Provenance Research in Archaeology. Measurement Science and Technology 14:1516-1526. Grace, R. 1996 Use-Wear Analysis: The State of the Art. Archaeometry 38:209-229. Gremillion, Kristen J. 1996 Early Agricultural Diet in Eastern North America: Evidence from Two Kentucky Rockshelters. American Antiquity 61(3):520-536. 2004 Seed Processing and the Origins of Food Production in Eastern North America. American Antiquity 69(2):215-233. Griffiths, Dorothy M. 1978 Use-Marks on Historic Ceramics: A Preliminary Study. Historical Archaeology 12: 68-81. Hally, David J. 1983 Use Alteration of Pottery Vessel Surfaces: An Important Source of Evidence for the Identification of Vessel Function. North American Archaeologist 4(1):3-26. Hambacher, Michael J. 1992 The Skegemog Point Sie: Continuing Studies in the Cultural Dynamics of the Carolinian-Canadian Transition Zone. Ph.D. dissertation, Department of Anthropology, Michigan State University, East Lansing. Hamilton, Scott, James Graham, and B.A. Nicholson 2007 Archaeological Site Distributions and Contents: Modeling Late Precontact Blackduck Land Use in the Northeastern Plains. Canadian Journal of Archaeology 31(3):93-136. Hancock, Ronald V., Susan Aufreiter, and Ian Kenyon 1997 European White Glass Trade Beads as Chronological and Trade Markers. In Materials Issues in Art and Archaeology V, edited by P. Vandiver, J. R. Druzick, J. F. Merkel, and J. Stewart, pp. 181-191. Materials Research Society, Symposium Proceedings Vol 462, Pittsburgh, PA. Hart, John P. 2008 Evolving the Three Sisters: The Changing Histories of Maize, Bean, and Squash in New and the Greater Northeast. In Current Northeast Paleoethnobotany II, edited by John P. Hart, pp. 87-99. New York State Museum Bulletin 512. University of the State of New York, Albany. 344 2012 Pottery Wall Thinning as a Consequence of Increased Maize Processing: A Case Study from Central New York. Journal of Archaeological Science 39:3470-3474. Hart, John P. and Hetty Jo Brumbach 2009 On Pottery Change and Northern Iroquoian Origins: An Assessment from the Finger Lakes Region of Central New York. Journal of Anthropological Archaeology 28:367-381. Hart, John P. and C. Margaret Scarry 1999 The Age of Common Beans (Phaseolus vulgaris) in the Northeastern United States. American Antiquity 64(4):653-658. Hart, J., Asch, D. Scarry, C. and Crawford, G. 2002 The Age of Common Bean in the North Eastern Woodlands. American Antiquity 76, 377-385. Hart, John P., Hetty Jo Brumbach, and Robert Lusteck 2007 Extending the Phytolith Evidence for Early Maize (Zea mays ssp. Mays) and Squash (Cucurbita sp.) in Central New York. American Antiquity 72(3):563-583. Hart, John P. and William A. Lovis. 2013 Reevaluating What We Know About the Histories of Maize in Northeastern North America: A Review of Current Evidence. Journal of Archaeological Research 21(2):175-216. Hart, John P., William A. Lovis, Janet K. Schulenberg, and Gerald R. Urquhart 2007 Paleodietary Implications from Stable Carbon Isotope Analysis of Experimental Cooking Residues. Journal of Archaeological Science 34:804-813. Hart, John P., William A. Lovis, Gerald R. Urquhart, and Eleanora A. Reber 2013 Modeling Freshwater Reservoir Offsets on Radiocarbon-dated Charred Cooking Residues. American Antiquity 78(3):536-552. Hart, John P, Robert G. Thompson, and Hetty Jo Brumbach 2003 Phytolith Evidence for Early Maize (Zea Mays) in 2003 the Northern Finger Lakes Region of New York. American Antiquity 68(4), 619-640. Hart, John P., Termeh Shafie, Jennifer Birch, Susan Dermarkar, and Ronald F. Williamson 2017 Nation Building and Social Signaling in Southern Ontario: A.D. 1350-1650. PLoS ONE 11(5): e0156178. doi:10.1371/journal.pone.0156178 Haslam, Michael 2004 The Decomposition of Starch Grains in Soils: Implications for Archaeological Residue Analyses. Journal of Archaeological Science 31:1715-1734. Hastorf, Christine A. 1999 Recent Research in Paleoethnobotany. Journal of Archaeological Research 7(1):55-103. 345 Hastorf, Christine A. and Michael J. DeNiro 1985 Reconstruction of Prehistoric Plant Production and Cooking Practices by a New Isotopic Method. Nature 315:489-491. Hastorf, Christine A. and Sissel Johannessen 1994 Becoming Corn-Eaters in Prehistoric America. In Corn and Culture in the Prehistoric New World, edited by Sissel Johannessen and Christine A. Hastorf, pp. 427-443. University of Minnesota Publications in Anthropology No. 5. Westview Press, Boulder, CO. Heron, Carl and Oliver Craig 2015 Aquatic Resources in Food Crusts: Identification and Implication. Radiocarbon 57: 707-719. Hilger, Inez 1951 Chippewa Child Life and Its Cultural Background. Smithsonian Institution Bureau of American Ethnology Bulletin 146. Governement Printing Office, Washington, D.C Hinsdale, W. 1931 Archaeological Atlas of Michigan. Michigan Handbook Series, X No. 4. University of Michigan Press, Ann Arbor. Holman, Margaret B. 1978 The Settlement System of the Mackinac Phase. Ph.D. dissertation, Department of Anthropology, Michigan State University, East Lansing. 1984 The Identification of Late Woodland Maple Sugaring Sites in the Upper Great Lakes. Midcontinental Journal of Archaeology 9(1):63-89. Holman, J. Alan and Danita Brandt 2009 Pleistocene Fauna. In Michigan Geography and Geology, edited by R. Schaetzl, J. Darden, and Danita Brandt, pp. 106-114. Custom Publishing, New York. Holman, Margaret. B. and Janet. G. Brashler 1999 Economics, Material Culture, and Trade in the Late Woodland Lower Peninsula of Michigan. In Retrieving Michigan’s Buried Past: The Archaeology of the Great Lakes State, edited by John R. Halsey, pp. 212-220. Cranbrook Institute of Science, Bloomfield Hills, MI. Holman, Margaret B. and Kathryn C. Egan 1985 Processing Maple Sap with Prehistoric Techniques. Journal of Ethnobiology 5(1):61-75. Howey, Meghan C. L. and John O’Shea 2002 Thinking Outside the Circle: New Research at Michigan’s Missaukee Earthworks. Paper presented to the 48th Midwest Archaeological Conference, Columbus, Ohio. 346 Huber, James K. 2001 Palynological Investigations Related to Archaeological Sites and the Expansion of Wild Rice (Zizania aquatica L.) in Northeast Minnesota. Masters thesis, Graduate School, University of Minnesota, Minneapolis. Hunt, George T. 1940 The Wars of the Iroquois: A Study of Intertribal Trade Relations. University of Wisconsin Press, Madison. Hupy, Christina M. and Catherine Yansa 2009 The Last 17,000 Years of Vegetation History. In Michigan Geography and Geology, edited by R. Schaetzl, J. Darden, and Danita Brandt, pp. 91-105. Custom Publishing, New York. Janzen, Donald E. 1968 The Naomikong Point Site and the Dimensions of Laurel in the Lake Superior Basin. Anthropological Papers No. 36. Museum of Anthropology, University of Michigan, Ann Arbor. Johnson, Matthew 2010 Archaeological Theory: An Introduction. Wiley-Blackwell, Oxford, UK. Jones, Siân 1997 The Archaeology of Ethnicity. Routledge, London and New York. Kalčik, Susan 1984 Ethnic Foodways in America: Symbol and the Performance of Identity. In Ethnic and Regional Foodways in the United States: The Performance of Group Identity, edited by Linda Kelly Brown and Kay Mussell, pp. 37-65. The University of Tennessee Press, Knoxville. Kapp, Ronald O. 1999 Michigan Late Pleistocene, Holocene, and Presettlement Vegetation and Climate. In Retrieving Michigan’s Buried Past: The Archaeology of the Great Lakes State, edited by John R. Halsey, pp. 31-58. Cranbrook Institute of Science, Bloomfield Hills, MI. Katzenberg, M. Anne 1989 Stable Isotope Analysis of Archaeological Faunal Remains from Southern Ontario. Journal of Archaeological Science 16:319-329. Katzenberg, M. Anne and Susan Pfeiffer 1995 Nitrogen Isotope Evidence for Weaning Age in a Nineteenth Century Candian Skeletal Sample. In Bodies of Evidence: Reconstructing History Through Skeletal Analysis, edited by Anne L. Grauer, pp. 221-236. Wiley-Liss, New York. 347 Katzenberg, M. Anne, Henry P. Schwarcz, Martin Knyf, and F. Jerome Melbye 1995 Stable Isotope Evidence for Maize Horticulture and Paleodiet in Southern Ontario, Canada. American Antiquity 60:335-350. Kenyon, Walter A. 1970 Methodist Point. Royal Ontario Museum Art and Archaeology Occasional Paper 22. The Royal Ontario Museum, Toronto. Kincare, Kevin, and Grahame Larson 2009 Evolution of the Great Lakes. In Michigan Geography and Geology, edited by R. Schaetzl, J. Darden, and Danita Brandt, pp. 174-190. Custom Publishing, New York. Kobayashi, Masashi 1994 Use-Alteration Analysis of Kalinga Pottery: Interior Carbon Deposits of Cooking Pots. In Kalinga Ethnoarchaeology: Expanding Archaeological Method and Theory, edited by William A. Longacre and James M. Skibo, pp. 127-168. Smithsonian Institution, Washington, D.C. Kooiman, Susan M. 2012 Old Pots, New Approaches: A Functional Analysis of Woodland Pottery from Lake Superior’s South Shore. Master’s thesis, Department of Sociology and Anthropology, Illinois State University, Normal. 2015a Sizing up the Past: Evaluating the Relationship between Rim Diameter and Volume in Late Woodland Upper Great Lakes Ceramic Jars. Paper presented at the 59th Annual Midwest Archaeological Conference, Milwaukee, WI, November 2015. 2015b Pottery Function, Cooking, and Subsistence in the Upper Great Lakes: A View from the Middle Woodland Winter Site in Northern Michigan. Paper presented at the 80th Annual Meeting of the Society for American Archaeology, San Francisco, CA. 2016 Woodland Pottery Function, Cooking, and Diet in the Upper Great Lakes of North America. Midcontinental Journal of Archaeology 41(3):1-25. Kuhn, Robert D. and Robert E. Funk 2000 Boning Up on the Mohawk: An Overview of Mohawk Faunal Assemblages and Subsistence Patterns. Archaeology of Eastern North America 28:29-62. Larson, Curtis E. 1999 A Century of Great Laeks Levels Research: Finished or Just Beginning? In Retrieving Michigan’s Buried Past: The Archaeology of the Great Lakes State, edited by John R. Halsey, 1-30. Cranbrook Institute of Science, Bloomfield Hills, MI. 348 Larson, Grahame J. and Kevin Kincare 2009 Late Quaternary History of the Eastern Mid-Continent Region, USA. In Michigan Geography and Geology, edited by R. Schaetzl, J. Darden, and Danita Brandt, pp. 69-90. Custom Publishing, New York. Larson, Grahame and Randall Schaetzl 2001 Origin and Evolution of the Great Lakes. Journal of Great Lakes Research 27(4): 518-546. Ligman, Michael S. 2013 “Put That in Your Pipe and Smoke It”: An Exploratory Study of Native American Ceramic Tobacco Pipes at the James Fort Site in Virginia Using Portable X-Ray Fluorescence. Unpublished masters thesis, Historical Archaeology Program, University of Massachusetts Boston Linton, Ralph 1944 North American Cooking Pots. American Antiquity 9(4):369-380. Lovis, William A. 1971 The Holtz Site (20AN26), Antrim County, Michigan: A Preliminary Report. The Michigan Archaeologist 17(2):49-64. 1973 Late Woodland Cultural Dynamics in the Northern Lower Peninsula of Michigan. Ph.D. dissertation, Department of Anthropology, Michigan State University, East Lansing. 1978 A Case Study of Construction Impacts on Archaeological Sites in Michigan's Inland Waterway. Journal of Field Archaeology 5(3):357-360. 1990 Curatorial Considerations for Systematic Research Collections: AMS Dating a Curated Ceramic Assemblage. American Antiquity 55(2):382-387. 1999 The Middle Archaic: Learning to Live in the Woodlands. In Retrieving Michigan’s Buried Past: The Archaeology of the Great Lakes State, edited by John R. Halsey, pp. 83- 94. Cranbrook Institute of Science, Bloomfield Hills, MI. 2009 Between the Glaciers and Europeans: People from 12,000 to 400 Years Ago. In Michigan Geography and Geology, edited by R. Schaetzl, J. Darden, and Danita Brandt, pp. 389-401. Custom Publishing, New York. 2014 An Up North Fishing Trip: Reinvestigating the Absolute Dated Archaeological Chronology of Northern Lake Michigan. Paper presented to the Museum of Anthropology, University of Michigan, Ann Arbor, MI. 3 April 2014. 349 Lovis, William A., Alan F. Arbogast, and G. William Monaghan The Geoarchaeology of Lake Michigan Coastal Dunes. Environmental Research Series, No. 2, J. A. Robertson series editor. Michigan Department of Transportation and Michigan State University Press, East Lansing. Lovis, W. A., R. E. Donahue and M. B. Holman. 2005 Long Distance Logistic Mobility as an Organizing Principle among Northern Hunter- Gatherers: A Great Lakes Middle Holocene Settlement System. American Antiquity 70(4):669-693. Lovis, William A., Kathryn C. Egan-Bruhy, Beverly A. Smith, and G. William Monaghan 2001 Wetlands and Emergent Horticultural Economies in the Upper Great Lakes: A New Perspective from the Schultz Site. American Antiquity 66(4):615-632. Lovis, William A. and John P. Hart 2015 Fishing for Dog Food: Ethnographic and Ethnohistoric Insights on the Freshwater Reservoir in Northeastern North America. Radiocarbon 57:557-570. Lovis, William A., Grace Rajnovich, and Aryn Bartley 1998 Exploratory Cluster Analysis, Temporal Change, and the Woodland Ceramics of the Portage Site at L’Arbre Croche. Wisconsin Archeologist 79:9-112. Lovis, William A., Gerald R. Urquhart, Maria E. Raviele, and John P. Hart 2011 Hardwood Ash Nixtamalization May Lead to False Negatives for the Presence of Maize by Depleting Bulk δ13C in carbonized Residues. Journal of Archaeological Science 38: 2726-2730. Lugenbeal, Edward 1978 The Blackduck Ceramics of the Smith Site (21KC3) and Their Implications for the History of Blackduck Ceramics and Culture in Northern Minnesota. Midcontinental Journal of Archaeology 3(1):45-68. Lynott, Mark J. 1974 The Ferrier Site. The Michigan Archaeologist 20(3-4):139-170. Malainey, Mary 2007 Fatty Acid Analysis of Archaeological Residues: Procedures and Possibilities. In Theory and Practice of Archaeological Residue Analysis, edited by Hans Barnard and Jelmer W. Eerkens, pp. 77-89. Archaeopress, Oxford. 2011 Lipid Residue Analysis. In A Consumer’s Guide to Archaeological Science, M. Malainey, pp. 201-218. Springer, New York. 350 Malainey, Mary E. and Timothy Figol 2015 Methodology and Results of Lipid Residues Extracted from Naomikong Point and Sand Point Pottery. Report prepared for Susan Kooiman, Department of Anthropology, Michigan State University, East Lansing, Michigan. 2018 Analysis of Lipid Residue Extracted from Archaeological Material from the Cloudman site, 20CH6. Report prepared for Susan Kooiman, Department of Anthropology, Michigan State University, East Lansing, Michigan. MacLean, Rachel and Timothy Insoll 1999 The Social Context of Food Technology in Iron Age Gao, Mali. World Archaeology 31(1):78-92. MacNeish, Richard S. 1952 Iroquois Pottery Types: A Technique for the Study of Iroquois Prehistory. National Museum of Canada Bulletin No. 124. Department of Resources and Development, National Museum of Canada, Ottawa. Martin, Susan R. 1989 A Reconsideration of Aboriginal Fishing Strategies in the Northern Great Lakes Region. American Antiquity 54(3):594-604. 1999a A Site for All Seasons: Some Aspects of Life in the Upper Peninsula during Late Woodland Times. In Retrieving Michigan’s Buried Past: The Archaeology of the Great Lakes State, edited by John R. Halsey, pp. 221-227. Cranbrook Institute of Science, Bloomfield Hills, MI. 1999b Wonderful Power: The Story of Ancient Copper Working in the Lake Superior Basin. Wayne State University Press, Detroit, MI. Mason, Carol I., and Margaret B. Holman 2000 Maple Sugaring in Prehistory: Tapping the Sources. In Interpretations of Native North American Life: Material Contributios to Ethnohistory, edited by M. S. Nassaney and E. S. Johnson. University Press of Florida, Gainesville. Mason, Ronald J. 1966 Two Stratified Sites on the Door Peninsula of Wisconsin. Anthropological Papers No. 26. Museum of Anthropology, University of Michigan, Ann Arbor. 1967 The North Bay Component at the Porte des Morts Site, Door County, Wisconsin. The Wisconsin Archeologist 48:267-345. 1970 Hopewell, Middle Woodland, and the Laurel Culture: A Problem in Archeological Classification. American Anthropologist 72(4):802-815. 1981 Great Lakes Archaeology. Academic Press, New York. 351 1986 Rock Island: Historical Indian Archaeology in the Northern Lake Michigan Basin. Kent State University Press, Kent, Ohio. Matson, F. R. 1965 Ceramic Ecology: An Approach to the Study of Early Cultures of the Near East. In Ceramics and Man, edited by F. R. Matson, 202-218. Aldine, Chicago. McElrath, Dale L., Thomas E. Emerson, and Andrew C. Fortier 2000 Social Evolution or Social Response? A Fresh Look at the “Good Gray Cultures” After Four Decades of Midwest Research. In Late Woodland Societies: Tradition and Transformation across the Midcontinent, edited by T. Emerson, D. McElrath, and A. Fortier, 543-579. University of Nebraska Press, Lincoln. McElrath, Dale L., Andrew C. Fortier, and Thomas E. Emerson 2009 An Introduction to the Archaic Societies of the Midcontinent. In Archaic Societies: Diversity and Complexity Across the Midcontinent, edited by A. C. Fortier, Dale L. McElrath, and T. E. Emerson, pp. 3-21. State University of New York Press, Albany, NY McGovern, Patrick E. and Gretchen R. Hall 2016 Charting a Future Course for Organic Residue Analysis in Archaeology. Journal of Archaeological Method and Theory 23:592:622. McHale Milner, Claire 1991 Localization in Small-Scale Societies: Late Prehistoric Social Organization in the Western Great Lakes. In Between Bands and States, Occasional Paper No. 9, edited by Susan A. Gregg, pp. 35-57. Center for Archaeological Investigations, Southern Illinois University, Carbondale. 1998 Ceramic Style, Social Differentiation, and Resource Uncertainty in the Late Precontact Upper Great Lakes. Unpublished Ph.D. dissertation, Department of Anthropology, University of Michigan, Ann Arbor. McKern, W. C. 1939 The Midwestern Taxonomic Method as an Aid to Archaeological Culture Study. American Antiquity 4(4):301-313. McPherron, Alan L. 1967 The Juntunen Site and the Late Woodland Prehistory of the Upper Great Lakes Area. Anthropological Papers No. 30. Museum of Anthropology, University of Michigan, Ann Arbor. Meltzer, David J. Pleistocene Overkill and North American Mammalian Extinctions. Annual Review of Anthropology 44:33-53. 352 Milligan, Heather E., Troy D. Pretzlaw, and Murray M. Humphries 2010 Stable Isotope Differentiation of Freshwater and Terrestrial Vascular Plants in Two Subarctic Regions. Écoscience 17(3):265-275. Milner, George, D. Anderson, and M. Smith 2001 The Distribution of Eastern Woodlands Peoples at the Prehistoric and Historic Interface. In Societies in Eclipse: Archaeology of the Eastern Woodlands Indians, A.D. 1400-1700, edited by D. Brose, C.W. Cowan, and R. Mainfort, pp. 9-18. Smithsonian Institution Press, Washington, D.C. Mithun, Marianne 1999 The Languages of Native North America. Cambridge University Press, Cambridge, UK. Moffat, Charles R. and Constance M. Arzigian 2000 New Data on the Late Woodland Use of Wild Rice in Northern Wisconsin. Midcontinental Journal of Archaeology 25(1):49-81. Monaghan, G. William and William A. Lovis 2005 Modeling Archaeological Site Burial in Southern Michigan: A Geoarchaeological Synthesis. Michigan State University Press, East Lansing. Monaghan, G. William, William A. Lovis, and Kathryn C. Egan-Bruhy 2006 Earliest Cucurbita from the Great Lakes, Northern USA. Quaternary Research 65: 216-222. Monaghan, G. William, Timothy M. Schilling, and Kathryn E. Parker 2014 The Age and Distribution of Domesticated Beans (Phaseolus vulgaris) in Eastern North America: Implications for Agricultural Practices and Group Interactions. Midwest Archaeological Conference Inc. Occasional Papers 1: 33-52. Montanari, Massimo 2006 Identity, Exchange, Traditions, and “Origins”. In Food is Culture, edited by Carole Counihan and Penny Van Esterik, pp. 133-137. Columbia University Press, New York. Morgan, Lewis Henry 1922 League of the Ho-De’-No-Sau-Nee. Dodd, Mead and Co., New York. Morgenstein, Maury and Carol A. Redmount 2005 Using Portable Energy Dispersive X-ray Fluorescence (EDXRF) Analysis for On-Site Study of Ceramic Sherds at El Hibeh, Egypt. Journal of Archaeological Science 32: 1613-1623. Morton, June D. and Henry P. Schwarcz 2004 Paleodietary Implications from Stable Isotopic Analysis of Residues on Prehistoric Ontario Ceramics. Journal of Archaeological Science 31:503-517. 353 Muhammad, Allison June 2010 A Bioarchaeological Study of Late Woodland Population from Michigan: Frazer-Tyra Site (20SA9), Conclusion, pp 85-103. Ph.D. dissertation, Department of Anthropology, Wayne State University, Detroit, MI. Nelson, Ben A. 1981 Ethnoarchaeology and Paleodemography: A Test of Turner and Lofgren’s Hypothesis. Journal of Anthropological Research 37(2):107-129. O’Shea, John M. 1989 The Role of Wild Resources in Small-Scale Agricultural Systems: Tales from the Lakes and the Plains. In Bad Year Economics: Cultural Responses to Risk and Uncertainty, edited by Paul Halstead and John O’Shea, pp. 57-67. Cambridge University Press, Cambridge. 2003 Inland Foragers and the Adoption of Maize Agriculture in the Upper Great Lakes of North America. Before Farming: The Archaeology and Anthropology of Hunter-Gatherers 1(3):68-83. O’Shea, John M. and Clare McHale Milner 2002 Material Indicators of Territory, Identity, and Interaction in a Prehistoric Tribal System. In The Archaeology of Tribal Societies, edited by W.A. Parkinson, pp. 200- 226. International Monographs in Prehistory, Ann Arbor, MI. Parker, Bradley J. and Jason R. Kennedy 2010 A Quantitative Attribute Analysis of the Ubaid-Period Ceramic Corpus from Kenan Tepe. Bulletin of the American Schools of Oriental Research 358:1-26. Parmenter, Jon 2010 The Edge of the Woods: Iroquoia, 1534-1701. Michigan State University Press, East Lansing. Pearsall, Deborah M. 2002 Maize is Still Ancient in Prehistoric Ecuador. Journal of Archaeological Science 29: 51-5. Pearsall, Deborah M. and Christine A. Hastorf 2011 Reconstructing Past Life-Ways with Plants II: Human-Environment and Human-Human Interactions. In Ethnobiology, edited by E.N. Anderson, D. Pearsall, E. Hunn, and N. Turner, pp. 173-187. Wiley-Blackwell, Hoboken, New Jersey. Peelo, Sarah 2011 Pottery-Making in Spanish California: Creating Multi-scalar Social Identity through Daily Practice. American Antiquity 76(4):642-666. 354 Perkl, Bradley E. 1998 Cucurbita pepo From King Coulee, Southeastern Minnesota. American Antiquity 63(2):279-288. Philippsen, Bente, Henrik Kjeldsen, Sönke Hartz, Harm Paulsen, Ingo Clausen, and Jan Heinemeier 2010 The Hardwater Effect in AMS 14C Dating of Food Crusts on Pottery. Nuclear Instruments and Methods in Physics Research B 268:995-998. Phillips, D. L., and P. L. Koch 2002 Incorporating Concentration Dependence in Stable Isotope Mixing Models. Oecologia 130:114-125. Pierce, Christopher 2005 Reverse Engineering the Ceramic Cooking Pot: Cost and Performance Properties of Plain and Textured Vessels. Journal of Archaeological Method and Theory 12(2):117-157. Piperno, Dolores R. 2003 A Few Kernal Short of a Cob: On the Staller and Thompson Late Entry Scenario for the Introduction of Maize into northern South America. Journal of Archaeological Science 30:831-836. Potter, James M. 2000 Pots, Parties, and Politics: Communal Feasting in the American Southwest. American Antiquity 65(3):471-492. Pregitzer, Kurt S., David D. Reed, Theodore J. Bornhorst, David R. Foster, Glenn D. Mroz, Jason S. McLachlan, Peter E. Laks, Douglas D. Stokke, Patrick E. Martin, and Shannon E. Brown 2000 A Buried Spruce Forest Provides Evidence at the Stand and Landscape Scale for the Effects of Environment on Vegetation at the Pleistocene/Holocene Boundary. Journal of Ecology 88:45-53. Ramsden, Peter G. 1990 The Hurons: Archaeology and Culture History. In The Archaeology of Southern Ontario to A.D. 1650, Occasional Publication of the London Chapter, OAS Number 5, edited by Chris J. Ellis and Neal Ferris, pp. 361-384. Ontario Archaeological Society Inc., London. Raviele, Maria E. 2010 Assessing Carbonized Archaeological Cooking Residues: Evaluation of Maize Phytolith Taphonomy and Density through Experimental Residue Analysis. Ph.D. Dissertation, Department of Anthropology, Michigan State University, East Lansing. 2011 Experimental Assessment of Maize Phytolith and Starch Taphonomy in Carbonized Cooking Residues. Journal of Archaeological Science 38:2708-2713. 355 Reber, Eleanor A., and John P. Hart 2008 Pine resins and Pottery Sealing: Analysis of Absorbed and Visible Pottery Residues from Central New York State. Archaeometry 50:999–1117. Rice, Prudence 1987 Pottery Analysis: A Sourcebook. University of Chicago Press, Chicago 1996a Recent Ceramic Analysis 1: Function, Style, and Origins. Journal of Archaeological Research 4(2):133-163. 1996b Recent Ceramic Analysis 2. Composition, Production, and Theory. Journal of Archaeological Research 4(3):165-202. Richner, Jeffrey J. 1973 Depositional History and Tool Industries at the Winter Site: A Lake Forest Middle Woodland Cultural Manifestation. Unpublished Master’s thesis, Department of Anthropology, Western Michigan University, Kalamazoo. Ritchie, William A. 1969 The Archaeology of New York State (Revised Edition). The Natural History Press, Garden City, New York. Robertson, James A., William A. Lovis, and John R. Halsey 1999 The Late Archaic: Hunter-Gatherers in an Uncertain Environment. In Retrieving Michigan’s Buried Past: The Archaeology of the Great Lakes State, edited by John R. Halsey, pp. 95-124. Cranbrook Institute of Science, Bloomfield Hills, MI. Rodríguez-Alegría, Enrique and Sarah R. Graff 2012 Introduction: The Menial Art of Cooking. In The Menial Art of Cooking, edited by Sarah R. Graff and Enrique Rodríguez-Alegría, pp. 1-18. University Press of Colorado, Boulder. Rogers, Edward S. 1962 The Round Lake Ojibwa. Occasional Paper 5, Art and Archaeology Division, Ontario Department of Lands and Forests for the Royal Ontario Museum, Toronto. Ross, W. A. 1975 Leslie M. Frost Natural Resources Centre Archaeological Resource Inventory. Manuscript on file, Ontario Ministry of Culture and Communications, Toronto, Ontario. Röttlander, Rolf C.A. 1990 Lipid Analysis in the Identification of Vessel Contents. MASCA Research Papers in Science and Archaeology 7:37-40. 356 Royer, Aurélien, Valérie Daux, Francois Fourel, and Christophe Lécuyer 2017 Carbon, Nitrogen and Oxygen Isotope Fractionation During Food Cooking: Implications for the Interpretation of the Fossil Human Record. American Journal of Physical Anthropology 163:759-771. Rye, O.S. 1976 Keeping Your Temper under Control: Materials and the Manufacture of Papuan Pottery. Archaeology and Physical Anthropology in Oceania 11(2):106-137. Sackett, James R. 1977 The Meaning of Style in Archaeology: A General Model. American Antiquity 42(3):369-380. Schiffer, Michael B. 1990 The Influence of Surface Treatment on Heating Effectiveness of Ceramic Vessels. Journal of Archaeological Science 17:373-381. Schiffer, Michael B. and James M. Skibo 1987 Theory and Experiment in the Study of Technological Change. Current Anthropology 28(5):595-622. 1997 The Explanation of Artifact Variability. American Antiquity 62(1):27-50. Schlanger, S. H. 1992 Recognizing Persistent Places in Anasazi Settlement Systems. In Space, Time, and Archaeological Landscapes, edited by J. Rossignol and L. Wandsnider, pp. 91-112. Plenum Press, New York. Schoeninger, Margaret J. 1995 Stable Isotope Studies in Human Evolution. Evolutionary Anthropology 4(3):83-98. Schroeder, Sissel 2004 Current Research on Late Precontact Societies of the Midcontinental United States. Journal of Archaeological Research 124:311-372. Schwarcz, Henry P., Jerry Melbye, M. Anne Katzenberg, and Martin Knyf 1985 Stable Isotopes in Human Skeletons of Southern Ontario: Reconstructing Palaeodiet. Journal of Archaeological Science 12:187-206. Schwarcz, Henry P., and Margaret J. Schoeninger 1991 Stable Isotope Analyses in Human Nutritional Ecology. American Journal of Physical Anthropology 34:283-321. Scott, Elizabeth M. 1996 Who Ate What? Archaeological Food Remains and Cultural Diversity. In Case Studies in Environmental Archaeology, edited by Elizabeth J. Reitz, Lee A. Newsom, and Sylvia J. Scudder. Plenum Press, New York. 357 Shapiro, Gary 1984 Ceramic Vessels, Site Permanence, and Group Size: A Mississippian Example. American Antiquity 49(4):696-712. Shepard, Anna O. 1968 Ceramics for the Archaeologist. Carnegie Institution of Washington, Washington, D.C. Shott, Michael J. 1999 The Early Archaic: Life after the Glaciers. In Retrieving Michigan’s Buried Past: The Archaeology of the Great Lakes State, edited by John R. Halsey, pp. 71-82. Cranbrook Institute of Science, Bloomfield Hills, MI. Shott, Michael J. and Henry T. Wright 1999 The Paleo-Indians: Michigan’s First People. In Retrieving Michigan’s Buried Past: The Archaeology of the Great Lakes State, edited by John R. Halsey, pp. 59-70. Cranbrook Institute of Science, Bloomfield Hills, MI. Siegel, Sidney 1956 Nonparametric Statistics for the Behavioral Sciences. McGraw-Hill, New York. Simon, Mary L. 2011 Evidence for Variability among Squash Seeds from the Hoxie Site (11CK4), Illinois. Journal of Archaeological Science 38:2079-2093. Skibo, James 1992 Pottery Function: A Use-Alteration Perspective. Plenum, New York. 1994 The Kalinga Cooking Pot: An Ethnoarchaeological and Experimental Study of Technological Change. In Kalinga Ethnoarchaeology: Expanding Archaeological Method and Theory, edited by William A. Longacre and James M. Skibo, pp. 113-126. Smithsonian Institution, Washington, D.C. 2013 Understanding Pottery Function. University of Utah Press, Salt Lake City. Skibo, James M. and Eric Blinman 1999 Exploring the Origins of Pottery on the Colorado Plateau. In Pottery and People: A Dynamic Interaction. University of Utah Press, Salt Lake City. Skibo, James M., Tamara C. Butts, and Michael B. Schiffer 1997 Ceramic Surface Treatment and Abrasion Resistance: An Experimental Study. Journal of Archaeological Science 24:311-317. Skibo, James M., Mary E. Malainey and Eric C. Drake 2009 Stone Boiling, Fire-Cracked Rock and Nut Oil: Exploring the Origins of Pottery Making on Grand Island. The Wisconsin Archeologist 90(1-2):47-64. 358 Skibo, James M. Mary E. Malainey, and Susan M. Kooiman 2016 Early Pottery in the North American Upper Great Lakes: Exploring Traces of Use. Antiquity 90:1226-1237 Skibo, James M., Michael B. Schiffer, and Kenneth C. Reid 1989 Organic-Tempered Pottery: An Experimental Study. American Antiquity 54(1):122-146. Smith, Beverly A. 1996 Systems of Subsistence and Networks of Exchange in the Terminal Woodland and Early Historic Periods in the Upper Great Lakes. Ph.D. dissertation, Department of Anthropology, Michigan State University, East Lansing. 2004 The Gill Net’s “Native Country”: The Inland Shore Fishery in the Northern Lake Michigan Basin. In An Upper Great Lakes Archaeological Odyssey: Essays in Honor of Charles E. Cleland, edited by William A. Lovis, pp. 64-84. Wayne State University Press, Detroit. Smith, Bruce D. 2006 Eastern North America as an Independent Center of Plant Domestication. Proceedings of the National Academy of Science 103(33): 12223-12228. Smith, Bruce D., and Richard A. Yarnell 2009 Initial formation of an indigenous crop complex in eastern North America at 3800 B.P. PNAS 106(16):6561-6566. (Open Access www.pnas.org/cgi/doi/10.1073/ pnas.0901846106). Smith, David G. 1990 Iroquoian Societies in Southern Ontario: Introduction and Historical Overview. In The Archaeology of Southern Ontario to A.D. 1650, Occasional Publication of the London Chapter, OAS Number 5, edited by Chris J. Ellis and Neal Ferris, pp. 279-290. Ontario Archaeological Society Inc., London. Smith, Monica L. 2006 The Archaeology of Food Preference. American Anthropologist 108(3):480-493. Snow, Dean R. 1980 Early Horticultural Period. In The Archaeology of New England, Dean R. Snow, pp. 261-306. Academic Press, New York. Speth, John D. 2015 When Did Humans Learn to Boil? PaleoAnthropology 2015: 54-67. Speth, J. and K. Spielmann 1983 Energy Source Protein Metabolism and Hunter Gatherer Subsistence Strategies. Journal of Anthropological Archaeology 2:1-31. 359 Spielmann, Katherine A. 2002 Feasting, Craft Specialization, and the Ritual Mode of Production in Small-Scale Societies. American Anthropologist 104(1):195-207. Stoltman, James B. 1973 The Laurel Culture in Minnesota. Minnesota Prehistoric Archaeology Series No. 8. Minnesota Historical Society, St. Paul. 1989 A Quantitative Approach to the Petrographic Analysis of Ceramic Thin Sections. American Antiquity 54(1):147-160. 2001 The Role of Petrography in the Study of Archaeological Ceramics. In Earth Sciences and Archaeology, edited by Paul Goldberg, Vance T. Holliday, and C. Reid Ferring, pp. 297-326. Kluwer Academic/Plenum Publishers, New York. Stoner, Wesley D. 2016 The Analytical Nexus of Ceramic Paste Composition Studies: A Comparison of NAA, LA-ICP-MS, and Petrography in the Prehispanic Basin of Mexico. Journal of Archaeological Science 76:31-47. St-Pierre, Christian G. and Robert G. Thompson 2015 Phytolith Evidence for the Early Presence of Maize in Southern Quebec. American Antiquity 80:408-415. Stuiver, Minz, Paula J. Reimer, and Ron W. Reimer. 2017 CALIB 7.1. Electronic program, http://calib.org, accessed November 27, 2017. Stuiver, Minz, Paula J. Reimer, and Ron W. Reimer. 2018 CALIB 7.1. Electronic program, http://calib.org, accessed March 14, 2018. Styles, Bonnie W. 2011 Animal Use by Holocene Aboriginal Societies of the Northeast. In The Subsistence Economies of Indigenous North American Societies, edited by Bruce D. Smith, pp. 449-482. Smithonian Institution Scholarly Press, Washington, D.C. Surette, Clarence 2008 The Potential of Microfossil Use in Paleodiet and Paleoenvironmental Analysis in Northwestern Ontario. Master’s thesis, Department of Geology, Lakehead University, Thunder Bay, Ontario. Taché, Karine and Oliver E. Craig 2015 Cooperative Harvesting of Aquatic Resources and the Beginning of Pottery Production in North-Easter North America. Antiquity 89(343):177-190. 360 Tani, Masakazu 1994 Why Should More Pots Break in Larger Households? Mechanisms Underlying Population Estimates from Ceramics. In Kalinga Ethnoarchaeology: Expanding Archaeological Method and Theory, edited by William A. Longacre and James M. Skibo, pp. 51-70. Smithsonian Institution, Washington, D.C. Terrell, E. E., P. M. Peterson, J. L. Reveal, and M. R. Duvall 1997 Taxonomy of North American Species Zizania (Poaceae). Sida 17:533-549. Thompson, V. 2010 The Rhythm of Space-Time and the Making of Monuments and Places during the Archaic. In Trend, Tradition, and Turmoil: What Happened to the Southeastern Archaic, edited by D. Thomas and M. Sanger, pp. 217-228. American Museum of Natural History, New York. Tooker, Elisabeth 1991 An Ethnography of the Huron Indians, 1615-1649. Syracuse University Press, Syracuse, New York. Trigger, Bruce G. 1976 The Children of Aataentsic: A History of the Huron People to 1660. McGill- Queen’s University Press, Montreal and London. Turner, Christy G. and Laurel Lofgren 1966 Household Size of Precontact Western Pueblo Indians. Southwestern Journal of Anthropology 22(2):117-132. Turschak, Benjamin A. 2013 Changes in the Lake Michigan Trophic Structure: As Revealed by Stable C and N Isotopes. Unpublished MS thesis, Dept. of Freshwater Sciences and Technology, University of Wisconsin-Milwaukee. Twiss, Katheryn 2012 The Archaeology of Food and Social Diversity. Journal of Archaeological Research 20:357-395. Tykot, Robert H. 2016 Using Nondestructive Portable X-ray Fluorescence Spectometers on Stone, Ceramics, Metal, and Other Materials in Museums: Advantages and Limitations. Applied Spectroscopy 70(1):42-56. Upton, Andrew J., William A. Lovis, and Gerald R. Urquhart 2015 An Empirical Test of Shell Tempering as an Alkaline Agent in the Nixtamalization Process. Journal of Archaeological Science 62:39-44. 361 Valppu, Seppo H. 2000 Paleoethnobotanical Investigations at the Big Rice Site: Laurel Culture Use of Wild Rice (Zizania aquatica L.) and Associated Radiocarbon Dates. In Wild Rice Research and Management, Proceedings o the Wild Rice Research and Management Conference, edited by L.S. Williamson, L.A. Dlutkowski, and McCammon Soltis, pp. 27-39. Great Lakes Indian Fish and Wildlife Commission, Carlton, PA. Van der Merwe, Nikolaas J., Ronald F. Williamson, Susan Pfeiffer, Stephen C. Thomas, and Kim O. Allegretto 2003 The Moatfield Ossuary: Isotopic Dietary Analysis of and Iroquoian Community, Using Dental Tissue. Journal of Anthropological Archaeology 22:245-261. Vander Zanden, M. Jake, Gilbert Cabana, and Joseph B. Rasmussen 1997 Comparing Trophic Position of Freshwater Fish Calculated Using Stable Nitrogen Isotope Ratios (delta 15N) and Literature Dietary Data. Canadian Journal of Fisheries and Aquatic Science 54:1142-1158. Vennum, Thomas 1988 Wild Rice and Ojibway People. Minnesota Historical Society Press, St. Paul. Voss, E. G. and A. Reznicek 2012 Michigan Flora: A Guide to the Identification and Occurrence of the Native and Naturalized Seed-Plants of the State (2 volumes). Cranbrook Institute of Science and the University of Michigan Herbarium, Bloomfield Hills, Michigan. Walder, Heather 2018 Small Beads, Big Picture: Assessing Chronology, Exchange, and Population Movement through Compositional Analyses of Blue Glass Beads from the Upper Great Lakes. Historical Archaeology, https://doi.org/10.1007/s41636-018-0100-4 Wallis, Neill J., Thomas J. Pluckhahn, and Michael D. Glascock 2016 Sourcing Interaction Networks of the American Southeast: Neutron Activation Analysis of Swift Creek Complicated Stamped Pottery. American Antiquity 81(4):717-736. Wallis, Wilson D. and Ruth Sawtell Wallis 1955 The Micmac Indians of Eastern Canada. University of Minnesota Press, St. Paul. Wandsnider, LuAnn 1997 The Roasted and the Boiled: Food Composition and Heat Treatment with Special Emphasis on Pit-Hearth Cooking. Journal of Anthropological Archaeology 16:1-48. Warrick, Gary 2000 The Precontact Iroquoian Occupation of Southern Ontario. Journal of World Prehistory 14(4):415-466. 362 Waugh, F. W. 1973 Iroquois Foods and Food Preparation. Anthropological Series No. 12, Memoir 86, Canada Department of Mines, Geological Survey, Ottawa. White, Richard 1991 The Middle Ground: Indians, Empires and Republics in the Great Lakes Region, 1650- 1815. Cambridge University Press, Cambridge. Wiessner, Polly 1983 Style and Social Information in Kalahari San Projectile Points. American Antiquity 48(2):253-276. Williamson, Ronald F. 1990 The Early Iroquoian Period of Southern Ontario. In The Archaeology of Southern Ontario to A.D. 1650, Occasional Publication of the London Chapter, OAS Number 5, edited by Chris J. Ellis and Neal Ferris, pp. 291-320. Ontario Archaeological Society Inc., London. Witgen, Michael 2012 An Infinity of Nations: How the Native New World Shaped Early North America. University of Pennsylvania Press, Philadelphia. Wobst, H. M. 1977 Stylistic behavior and information exchange. In Papers for the Director: Research Essays in Honor of James B. Griffin, edited by C. E. Cleland, pp. 317-342. Anthropological Paper 61. University of Michigan, Museum of Anthropology, Ann Arbor, MI. Wright, J. V. 1966 The Pic River Site. National Museum of Canada, Bulletin 206. National Museum of Canada, Dept. of the Secretary of State, Ottawa. 1967 The Laurel Tradition and the Middle Woodland Period. National Museum of Canada Bulletin 217, Anthropological Series No. 79. National Museum of Canada, Dept. of the Secretary of State, Ottawa. 1973 The Ontario Iroquois Tradition (Facsimile Edition). National Museum of Canada Bulletin 210. National Museum of Canada, Ottawa. Yang, XiaoYan, JinCheng Yu, HouYuan Lu, TianXing Cui, JingNing Guo, and QuanSheng Ge 2009 Starch Grain Analysis Reveals Function of Grinding Stone Tools at Shangzhai Site, Beijing. Science in China Series D: Earth Sciences 52(8):1164-1171. Yarnell, Richard A. 1964 Aboriginal Relationships Between Culture and Plant Life in the Upper Great Lakes Region. Anthropological Papers No. 23. Museum of Anthropology, University of Michigan, Ann Arbor. 363 Zvelebil, Marek and Peter Rowley-Conwy 1986 Foragers and Farmers in Atlantic Europe. In Hunters in Transistion: Mesolithic Societies of Temperate Eurasia and the Transition to Farming, edited by Marek Zvelebil, pp. 67-93. Cambridge University Press, Cambridge. 364