PHASE I ARCHAEOLOGICAL SURVEY AND PHASE II EVALUATION OF SITE 21PL109 FOR THE COUNTY STATE AID HIGHWAY 1 RECONSTRUCTION AND BRIDGE REPLACEMENT PROJECT IN POLK COUNTY, MN

State Project Number: 060-601-059

Mn/DOT Contract Number: 1002983 and Amendments 1 & 2

MN OSA License Numbers: 16-035 and 16-066

Authorized and Sponsored by: Minnesota Department of Transportation

Prepared by:

Frank Florin, Principal Investigator James Lindbeck, Staff Archaeologist

Florin Cultural Resource Services, LLC N12902 273rd Street Boyceville, WI 54725 Reports of Investigation # 122

December 2017 MANAGEMENT SUMMARY

The Minnesota Department of Transportation (MnDOT) and Polk County are proposing to replace Bridge 5767 over the Red River and reconstruct the adjoining 2.4 miles of Minnesota County State Aid Highway (CSAH) 1 (State Project 060-601-059). Florin Cultural Resources Services, LLC (FCRS) was retained to conduct a Phase I archaeological survey and Phase II evaluation. The project has federal funding, and the archaeological work was conducted under Section 106 of the National Historic Preservation Act. The Federal Highway Administration is the lead agency, and the MnDOT Cultural Resources Unit is the delegated review agent. The project extends to the North Dakota side of the river, and those results are presented in a separate report.

The project area is located in Archaeological Region 6n - Red River Valley North in T147N R149W Sections 23-26 and T147N R148W Sections 19 and 30, Polk County, Minnesota. The archaeological survey corridor along CSAH 1 included an expanded right-of-way of variable width, encompassing approximately 48 acres. The project crosses terraces of the Red River and the Glacial Lake Agassiz plain. The area is in agricultural fields, except for a few rural residences and a small wooded strip on the floodplain. Fieldwork was conducted from June 6 to 23 and October 10 to November 2, 2016. Frank Florin was the principal investigator. The Phase I and II archaeological field methods included pedestrian survey, shovel tests, deep auger tests, and excavation units. A total of 528 shovel tests were dug.

Precontact site 21PL109 was identified and evaluated with ten (1-x-1 meter) excavation units and close-interval shovel tests. A geomorphological investigation was conducted by Strata Morph Geoexploration, Inc. at the site to gain a better understanding of the site’s geomorphology, soils, and stratigraphy. The site is a large multicomponent Archaic, Woodland, and Northeastern Plains Village period habitation along the Red River. Animal kill and processing (mostly bison) appears to be the primary site activity. The site had 389 positive shovel tests, and artifacts were recovered from 0 to 185 cm below surface. Only small portions of the site were deep tested, and it is likely that artifacts extend to three meters or more below surface. Diagnostic artifacts include Northeastern Plains Village ceramic ware and a variety of projectile point types: Besant, Duncan-Hanna, Oxbow, Early Archaic small side-notched, and Plainview. The Plainview point was recovered from the ground surface and is out of context. The Northeastern Plains Village component had radiocarbon dates of 1300 +/- 30 to 1160 +/- 30 to RCYBP. Five separate Early to Middle Archaic occupations were identified in XUs, and they had radiocarbon dates of 6200, 6800, 7100, 7300, and 8000 RCYBP.

Numerous cooking/heating features, dense layers of bone (bone beds), and artifact-laden anthrosols were observed in shovel test profiles. One fire hearth feature was excavated. Artifact density was moderate to high across most of the site. Fauna was by far the most abundant artifact recovered, with notably smaller amounts of ceramics, FCR, lithic debris, stone tools, and cores. Site activities included hunting, animal processing (primarily bison), lithic reduction, stone tool manufacture/ maintenance, and cooking/heating. Some of the faunal material was thermally-altered.

The site is recommended eligible for listing in the National Register of Historic Places (NRHP) under Criterion D because it has integrity and is likely to yield important information on multiple precontact period contexts. The final project design will stay within the existing ROW on the current CSAH 1 alignment and have only shallow subsurface disturbances not to exceed 30 cm on Terraces 2, 3, and 4, thus avoiding the deeper portions of the site that have the potential to answer important research questions. This plan will avoid impacts to the site’s significant archaeological deposits, and there will be no adverse effect, which would alter the site’s integrity or characteristics that qualify it for inclusion in the NRHP. If the project design changes or if other projects will adversely affect the site, then a Phase III data recovery is recommended to mitigate the project’s effects. It is the opinion of FCRS that no historic properties eligible for or listed on the NRHP will be adversely affected by this project. i

TABLE OF CONTENTS

Management Summary ...... i Appendices ...... iv List of Figures ...... iv List of Tables ...... v 1. Project Description ...... 1 1.1 Overview ...... 1 1.2 Project Setting ...... 1 1.3 Project Area and Area of Potential Effect ...... 1 1.4 Curation ...... 2 1.5 Permit and License ...... 2 1.6 Dating Format ...... 2 1.7 Personnel for Lab and Report Tasks ...... 3 2. Research Design ...... 5 2.1 Objectives ...... 5 2.2 Aspects of the Research Design ...... 5 2.3 Eligibility Criteria and Historic Contexts ...... 5 2.3.1 Archaic Contexts (12,500 to 2500 BP) ...... 6 2.3.2 Woodland Period Contexts (2500 to 350 BP) ...... 7 2.3.3 Northeastern Plains Village Context (AD 1200-1650) ...... 10 3. Archaeological Field Methods ...... 12 3.1 Archaeological Field Methods ...... 12 3.1.1 Pedestrian Survey ...... 12 3.1.2 Shovel Tests and Deep Auger Tests ...... 12 3.1.3 Excavation Units (XUs) ...... 13 3.1.4 GPS Data Collection and Site Mapping in ArcView...... 13 3.1.5 Field Documentation ...... 13 4. Archaeological Lab Methods ...... 14 4.1 Artifact Processing ...... 14 4.2 Lithic Raw Material Resources ...... 15 4.3 Lithic Analysis Methods ...... 17 4.3.1 Thermal Alteration ...... 18 4.3.2 Lithic Debris ...... 18 4.3.3 Lithic Tools ...... 22 4.4 FCR Analysis Methods ...... 24 4.4.1 Definition of FCR ...... 24 4.4.2 FCR Background and Previous Studies...... 25 4.4.3 FCR Analytical Methods ...... 29 4.4.4 FCR Morphology ...... 30 4.5 Faunal Analysis Methods ...... 31 4.6 Ceramic Analysis Methods ...... 31 5. Literature Search ...... 33 5.1 Archival and Background Research for Previous Archaeology Sites ...... 33 5.2 Overview of Archaeological Investigations in the Red River Valley ...... 33 5.3 Previous Investigations in the Red River Valley ...... 33 5.3.1 Minnesota ...... 34 5.3.2 North Dakota ...... 34 5.4 Mn/Model Study of the Red River Prairie Subsection ...... 35 5.5 Air Imagery and Plat Map Review ...... 35 6. Culture History ...... 37

ii

6.1 Paleoindian Period (11,300 to 8500 RCYBP / cal 13,200 to 9500 BP) ...... 38 6.1.1 Early Paleoindian (11,300 to 10,500 RCYBP / 13,200 to 12,400 cal BP) ...... 39 6.1.2 Late Paleoindian (10,500 to 8500 RCYBP / 12,400 to 9500 cal BP) ...... 40 6.2 Plains Archaic Period (7700 to 2500 RCYBP / 8500 to 2700 cal BP) ...... 41 6.2.1 Early Plains Archaic Overview (7700 to 5500 RCYBP / 8500 to 6300 cal BP) ...... 43 6.2.2 Regional Early Plains Archaic Sites ...... 43 6.2.3 Middle Plains Archaic (5500 to 3200 RCYBP / 6300 to 3450 cal BP) ...... 55 6.2.4 Late Archaic (4500 to 2500 RCYBP / 5000 to 2700 cal BP) ...... 58 6.3 Woodland Period and Plains Village (2500 to 300 BP / 500 BC to AD 1700*) ...... 60 6.3.1 Early Woodland - Brainerd Complex ...... 62 6.3.2 Middle Woodland Havana-Related Complex ...... 64 6.3.3 The Central Minnesota Transitional Woodland Complex ...... 64 6.3.4 Late Woodland Blackduck-Kathio Complex ...... 65 6.3.5 Late Woodland Psinomani Complex ...... 67 6.3.6 Northeastern Plains Village Period...... 68 6.4 Contact and Historic Period (AD 1700 to 1950) ...... 69 6.4.1 Equestrian Nomadic Tradition ...... 69 6.4.2 Historic Period ...... 70 6.4.3 The French Period ...... 70 6.4.4 The British and American Periods...... 70 6.4.5 Red River Trails ...... 71 6.4.6 Railroad Transportation ...... 71 6.4.7 Local History ...... 72 7. Environmental Background ...... 73 7.1 Modern Environment ...... 73 7.2 Glacial History ...... 73 7.3 Physiography ...... 73 7.4 Lake Agassiz ...... 74 7.5 Hydrology ...... 74 7.6 Vegetation ...... 74 7.7 Fauna ...... 75 7.8 Soils ...... 75 8. Phase I Fieldwork Introduction ...... 76 8.1 Overview and Field Conditions ...... 76 8.2 Deep Auger Testing on Floodplain of the Red River ...... 77 9. Site 21PL109 ...... 80 9.1 Overview ...... 80 9.2 Physical Setting ...... 80 9.3 Soils ...... 80 9.4 Radiocarbon Dates ...... 81 9.5 Phase I Methods and Results for Pedestrian Survey ...... 83 9.6 Phase I and II Methods and Results for Shovel Testing ...... 88 9.7 Phase II XU Methods ...... 120 9.8 XUs 1 and 2 on Terrace 2 ...... 120 9.9 XUs 3 and 4 on Terrace 2 ...... 123 9.10 XUs 5 and 6 on Terrace 4 ...... 126 9.11 XUs 7 and 8 on Terrace 4 ...... 130 9.12 XUs 9 and 10 on Terrace 4 ...... 134 9.13 Phase I and II Artifact Summary ...... 138 9.14 Faunal Analysis by Steven Kuehn ...... 138 9.14.1 Fauna Results ...... 138 9.14.2 Faunal Distribution ...... 144

iii

9.14.3 Faunal Discussion ...... 146 9.15 Lithic Analysis ...... 146 9.16 Ceramic Analysis ...... 153 9.17 FCR ...... 154 9.18 Horizontal Artifact Patterning and Cultural Components ...... 154 9.19 Cultural Stratigraphy ...... 155 9.20 Site Integrity ...... 157 9.21 Radiocarbon Dates of Cultural Complexes ...... 157 9.22 Conclusions ...... 158 10. Summary and Recommendations ...... 182 11. References Cited ...... 183

APPENDICES Appendix A: Geomorphological Investigation by Strata Morph Geoexploration, Inc. Appendix B: Office of State Archaeologist Licenses Appendix C: Artifact Catalogs Appendix D: Radiocarbon Dating Reports from Beta Analytic Inc.

LIST OF FIGURES

Figure 1. CSAH 1 Reconstruction Red River to Nielsville - Archaeology Survey Area and Site 21PL109...... 4 Figure 2. Lithic Resource Regions of Minnesota (adapted from Bakken 2011)...... 16 Figure 3. Location of Project Area on 1930 Air Photo...... 36 Figure 4. Location of Project Area on 1954 Air Photo...... 36 Figure 5. Field Conditions and Shovel Test Areas for CSAH 1 Reconstruction - Red River to Nielsville...... 78 Figure 6. Deep Auger Test Locations on Floodplain of Red River for CSAH 1 Reconstruction. 79 Figure 7. Site 21PL109 Map of Find Spots and FCR...... 163 Figure 8. Site 21PL109 Map of Shovel Tests on Point Bar of Red River with 100 Foot Station Markers Depicted on CSAH 1...... 164 Figure 9. Site 21PL109 Photo of Terraces 2 to 4 on South Side of CSAH 1 Taken from Near XUs 7 and 8, Facing West...... 165 Figure 10. Site 21PL109 Photo of Terraces 2 to 4 on North Side of CSAH 1 Taken from Near XUs 7 and 8, Facing West...... 165 Figure 11. Site 21PL109 XUs 1 and 2 North Wall Profile...... 166 Figure 12. Site 21PL109 Photo XUs 1 and 2 North Wall Profile...... 167 Figure 13. Site 21PL109 XUs 3 and 4 North Wall Profile...... 168 Figure 14. Site 21PL109 Photo XUs 3 and 4 North Wall Profile...... 169 Figure 15. Site 21PL109 XUs 5 and 6 North and XU 6 East Wall Profiles...... 170 Figure 16. Site 21PL109 Photo XUs 5 and 6 North Wall Profile...... 171 Figure 17. Site 21PL109 Photo XU 6 East Wall Profile...... 171 Figure 18. Site 21PL109 Planview Photo of XU 5 Floor with Charcoal, Ash, and Oxidized Soil in Anthrosol at 82 to 86 cmbd above Feature 1...... 172 Figure 19. Site 21PL109 Planview of Feature 1 at 90 cmbd in XUs 5 and 6...... 173 Figure 20. Site 21PL109 Photo of Planview of Feature 1 (Oxidized Soil) at 90 cmbd in XUs 5 and 6...... 174 Figure 21. Site 21PL109 XUs 7 and 8 North Wall Profile...... 175 Figure 22. Site 21PL109 Photo XUs 7 and 8 North Wall Profile...... 176

iv

Figure 23. Site 21PL109 XUs 9 and 10 North and XU 9 West Wall Profiles...... 177 Figure 24. Site 21PL109 Photo XUs 9 and 10 North Wall Profile...... 178 Figure 25. Site 21PL109 Photo XU 10 West Wall Profile (showing road construction cut)...... 179 Figure 26. Site 21PL109 Photos and Illustrations of Projectile Points...... 180 Figure 27. Site 21PL109 Photos and Illustrations of Projectile Points...... 181

LIST OF TABLES Table 1. Descriptive Categories for Artifact Classes in the Catalog...... 14 Table 2. Estimated Primary, Secondary, and Minor Lithic Raw Material Status by Region and Subregion (Bakken 2011)...... 17 Table 3. Definitions of Technological Flake Types (primarily adapted from Root 2004)...... 19 Table 4. Cooking Facilities and Expected Characteristics of FCR Features and Scatters (from Thoms 2008a)...... 27 Table 5. FCR Type Descriptions...... 30 Table 6. Survey Summary CSAH 1 (East to West) - Field Conditions and Survey Results...... 76 Table 7. Deep Auger Tests 1 to 6 Typical Profile on Floodplain of Red River...... 77 Table 8. Shovel Test 180 Soil Profile on Terrace 3...... 81 Table 9. Shovel Test 225 Soil Profile on Terrace 3...... 81 Table 10. Shovel Test 273 Soil Profile on Terrace 3...... 81 Table 11. Radiocarbon Dates from the Sites...... 82 Table 12. Site 21PL109 Summary of Find Spot Artifacts...... 84 Table 13. Site 21PL109 Summary of Find Spot Artifacts...... 84 Table 14. Site 21PL109 Summary of Artifacts from Shovel Tests...... 89 Table 15. Site 21PL109 Features, Bone Beds, and Other Significant Finds in Shovel Tests...... 89 Table 16. Site 21PL109 Summary of Artifacts from Shovel Tests...... 90 Table 17. Site 21PL109 Artifacts by Count from XUs 1 and 2...... 120 Table 18. Site 21PL109 Artifacts by Weight (g) from XUs 1 and 2...... 121 Table 19. Site 21PL109 Fauna counts from XUs 1 and 2...... 121 Table 20. Site 21PL109 Artifacts by Count from XUs 3 and 4...... 123 Table 21. Site 21PL109 Artifacts by Weight (g) from XUs 3 and 4...... 124 Table 22. Site 21PL109 Fauna counts from XUs 3 and 4...... 124 Table 23. Site 21PL109 Artifacts by Count from XUs 5 and 6...... 126 Table 24. Site 21PL109 Artifacts by Weight (g) from XUs 5 and 6...... 127 Table 25. Site 21PL109 Fauna counts from XUs 5 and 6...... 127 Table 26. Site 21PL109 Raw Materials by Depth in XUs 5 and 6...... 129 Table 27. Site 21PL109 Diagnostic Lithic Debris and Tools by Depth in XUs 5 and 6...... 129 Table 28. Site 21PL109 Artifacts by Count from XUs 7 and 8...... 131 Table 29. Site 21PL109 Artifacts by Weight (g) from XUs 7 and 8...... 131 Table 30. Site 21PL109 Fauna counts from XUs 7 and 8...... 132 Table 31. Site 21PL109 Raw Materials by Depth in XUs 7 and 8...... 133 Table 32. Site 21PL109 Diagnostic Lithic Debris and Tools by Depth in XUs 7 and 8...... 133 Table 33. Site 21PL109 Artifacts by Count from XUs 9 and 10...... 134 Table 34. Site 21PL109 Artifacts by Weight (g) from XUs 9 and 10...... 135 Table 35. Site 21PL109 Fauna counts from XUs 9 and 10...... 135 Table 36. Site 21PL109 Raw Materials by Depth in XUs 9 and 10...... 137 Table 37. Site 21PL109 Diagnostic Lithic Debris and Tools by Depth in XUs 9 and 10...... 137 Table 38. Site 21PL109 Summary of Artifacts by Count and (Weight)...... 138 Table 39. Site 21PL109 Faunal Assemblage...... 139 Table 40. Bison Remains by Body Portion and Element...... 140

v

Table 41. Site 21PL109 Distribution of Faunal Remains by Context...... 145 Table 42. Site 21PL109 Lithic Artifacts by Material, Flake, and Tool/Core Types...... 147 Table 43. Site 21PL109 Summary of Diagnostic Flake Types, Technologies, and Reduction Stages...... 148 Table 44. Site 21PL109 Tool Type by Material Type...... 150 Table 45. Site 21PL109 Nonbifacial Flake Tools by Flake Type...... 150 Table 46. Site 21PL109 FCR Count by Material, Size, and Type...... 154

vi

1. PROJECT DESCRIPTION

1.1 Overview

The Minnesota Department of Transportation (MnDOT) and Polk County are proposing to replace Bridge 5767 over the Red River and reconstruct the adjoining 2.4 miles of Minnesota County State Aid Highway (CSAH) 1 (State Project 060-601-059). Florin Cultural Resources Services, LLC (FCRS) was retained to conduct a Phase I archaeological survey and Phase II evaluation. The project has federal funding, and the archaeological work was conducted under Section 106 of the National Historic Preservation Act. The Federal Highway Administration is the lead agency, and the MnDOT Cultural Resources Unit is the delegated review agent. Fieldwork was conducted from June 6 to 23 and October 10 to November 2, 2016. The project extends to the North Dakota side of the river, and those results are presented in a separate report.

1.2 Project Setting

The project is located along CSAH 1 between the Red River and the town of Nielsville, Minnesota. The eastern one-third of the project area is on Red River terraces, and the western two-thirds is on the Glacial Lake Agassiz plain. The area is in agricultural fields, except for a few rural residences and a small wooded strip on the river floodplain. The terrain is fairly level.

1.3 Project Area and Area of Potential Effect

The project area is located in T147N R149W Sections 23-26 and T147N R148W Sections 19 and 30, Polk County, Minnesota (Figure 1). The archaeological survey corridor along CSAH 1 was approximately 2.4 miles long, extending from the Red River to the town of Nielsville. The survey area was variable in size and included approximately 48 acres. The survey parameters are discussed below.

The survey corridor was 90 feet each side of the CSAH 1 centerline in the eastern portion of the project from Nielsville going about 1.5 miles west to the eastern edge of the private residence (Tax Parcel # 39.00031.00, Matthew and Lisa Roudybush) at the river oxbow. This area includes the western .75 mile of Sections 19 and 30 and the eastern .75 mile of Sections 24 and 25.

At the private residence referenced above (approximately .25 mile east of the western edge of Section 24), the survey corridor extended 90 feet south of the CSAH 1 centerline for a distance of approximately 460 feet. West of this parcel to the western edge of Section 24 (a distance of 850 feet), the survey on the south side of the road was limited to the existing ROW (50 feet south of the CSAH 1 centerline) because the landowner denied access. On the north side of CSAH 1 from the oxbow to the western edge of Section 24 (a distance of .25 mile), the survey corridor was variable in width (expanding westwards) and extended a maximum of 366 feet) north of the existing centerline. The landowner denied access for subsurface testing on the north side of the road in the bean field in the SW¼ of Section 24. So, shovel testing in this area was limited to the ROW.

At the west end of the project area in Sections 23 and 26, the survey corridor was 75 feet on the north side of the CSAH 1 centerline. The survey corridor on the south side of CSAH 1 in Sections 23 and 26 was variable in width and extended a maximum of 280 feet south of the existing centerline. Within 400 feet of the Red River on the floodplain, the survey area was 150 feet each side of the CSAH 1 centerline.

1

In order to avoid site 21PL109, the construction limits were reduced and confined to the existing ROW. The final Area of Potential Effect (APE) for the project is the existing ROW, which encompasses the following areas: in the eastern portion of the project in Sections 19 and 30, the ROW extends 50 feet north of centerline and 55 feet south of centerline; in the middle portion of the project in Sections 24 and 25, the ROW extends 60 feet north of centerline and 50 feet south of centerline; and in the western portion of the project in Sections 23 and 26, the ROW extends 50 feet north and south of centerline. The archaeological survey area encompassed the APE and additional areas beyond it, as described above.

A pedestrian survey was conducted for the entire APE, and shovel testing was conducted on the Red River terraces. The vertical APE extends 3.3 feet below the ground surface, except on the floodplain adjacent to the Red River bridge, where the APE extends 10 feet deep within 100 feet of the bridge and 100 feet of the road centerline. The portion of the project area on the Red River terraces has the potential for deeply buried sites, while the potential for sites below 30 to 50 cm on the Glacial Lake Agassiz plain is very low to none. The UTM coordinates along CSAH 1 for the survey area are the following: E660280 N5265945 for the west end and E664030 N5266115 for the east end (1983 Datum, UTM Zone 15). Land ownership included county-owned ROW and private lands beyond the ROW.

1.4 Curation

Copies of project documentation are on file at the FCRS office in Boyceville, Wisconsin. Artifacts recovered from the portion of site 21PL109 in the ROW will be curated at the Minnesota Historical Society (MHS). Artifacts from privately owned lands are being returned to the landowners at their request. Douglas Palm, the landowner in Sections 23 and 26 indicated that he would give his artifacts to the Red River History Museum in Shelly, Minnesota.

1.5 Permit and License

Minnesota Office of State Archaeologist permits 16-035 (Phase I) and 16-066 (Phase II) were obtained for survey on state/county lands. Copies of these permits are in Appendix A.

1.6 Dating Format

Dates in this report are presented in two formats: 1) by their conventional radiocarbon age (uncalibrated) and 2) as calibrated to actual calendar years. The conventional radiocarbon age (measured radiocarbon age corrected for isotopic fractionation) is presented in the format of “RCYBP” (radiocarbon years before present; with “present” by convention being AD 1950). The use of “RCYBP” dates allows for the consistent comparison of dates from sites in previous reports, as this format has been the standard. Radiocarbon dates from older reports may not have been corrected for isotopic fractionation, but this correction is typically small. Dates calibrated to actual calendar years use the convention “cal BP” (for example cal. 8000 BP) to distinguish them from uncalibrated dates (RCYBP).

For various technical reasons, radiocarbon years are not equal to calendar years, and therefore calibration is necessary to assess the actual age of a sample. Radiocarbon years are converted to calendar years by a process called calibration. This process is based on dating samples with a precisely known age, such as wood that can be dated to a calendar year by tree-ring counts. These dates reveal systematic variations between radiocarbon years and calendar years, and allow the statistical estimation of actual calendar age for any given radiocarbon date. Generally speaking, conventional age back to about 3000 RCYBP will be close to the actual calendar (calibrated) age, but beyond that the calendar age becomes progressively older than the radiocarbon age. A date of 2000 2

RCYBP, for example, indicates an age of close to 2,000 calendar years ago, while a date of 10,000 RCYBP indicates a calendar age (calibrated date) of closer to 11,500 years ago. Calibrated dates in this report are 2 sigma calibrations (95% probability).

1.7 Personnel for Lab and Report Tasks

Frank Florin authored all sections of this report, except where noted otherwise. He was also the lab supervisor and conducted the artifact analysis, except for fauna. Zooarchaeologist Steven Kuehn was retained to conduct the faunal analysis. Beth Wergin was the lab manager, and she cataloged artifacts, prepared data tables, and drafted the wall profile illustrations for the report. Sarah Schultz illustrated the projectile points. Amanda Peterson prepared data tables for the report. James Lindbeck conducted background research, edited the report, and authored the Culture History and Literature Search sections and portions of the Environmental Background section. Kent Bakken wrote the Lithic Raw Material Resource Base section. Connie Arzigian and staff at the Mississippi Valley Archaeology Center (MVAC) were retained to conduct the botanical analysis.

3

2. RESEARCH DESIGN

2.1 Objectives

There are several objectives of the Phase I archaeological survey and Phase II site evaluations: 1) to aid project sponsors in complying with Section 106 of the National Historic Preservation Act and 36 CFR 800: Protection of Historic Properties; 2) to identify archaeological sites and assess their eligibility for listing on the National Register of Historic Places (NRHP); 3) to aid in project planning; and 4) to produce a report documenting the archaeological investigations.

2.2 Aspects of the Research Design

The research design was developed to meet project objectives, and it adhered to the research and field method guidelines established by the Minnesota State Historic Preservation Office (MnSHPO), Minnesota Office of State Archaeologist (OSA), and MnDOT. These methods, which included a literature search, fieldwork, analysis of data, and production of a technical report, are summarized below and discussed in greater detail in the following sections.

The literature search provided information on previous investigations, previously recorded sites, potential cultural resources depicted on historic maps, and the environmental setting.

Archaeological fieldwork included pedestrian survey, shovel tests, deep auger tests, and excavation units (XUs). Pedestrian survey was used to identify artifacts or archaeological remains that were present on the ground surface. Shovel tests and deep auger tests were used to identify artifacts that were present below the ground surface, characterize soils at the survey areas and archaeological sites, and provide information on the horizontal and vertical provenience of artifacts. XUs were used to recover artifacts, provide detailed information on artifact provenience and cultural stratigraphy, identify cultural features, assess site integrity, and provide exposures of soil profiles at the sites. Specific details of the field methods are presented in Section 3.

The analysis of artifacts was conducted using current methods appropriate to each artifact class. The analysis was oriented towards identifying specific attributes that would provide useful information for interpreting the function and historic context of the site. Specific analytical methods for each artifact class are discussed in detail in Section 4.

The report documents the results of research, fieldwork, and artifact analysis and provides interpretations of the data and recommendations for the sites and project.

2.3 Eligibility Criteria and Historic Contexts

Recommendations for the NRHP eligibility of site 21PL109 identified for this project are based on the National Register Criteria in 36 CFR Part 60.1 guidelines established by the National Park Service (1991) and Minnesota contexts for the Archaic period, Woodland period, and lithic scatters (Anfinson 1994; Arzigian 2008; Dobbs 1988; Gibbon and Anfinson 2008). Archaeological sites that retain integrity may be eligible for the National Register under the following four criterion:

A) if they are associated with events that have made a significant contribution to the broad patterns of our history; or

B) if they are associated with the lives of persons significant in our past; or

5

C) if they embody the distinctive characteristics of a type, period, or method of construction, or that represent the work of a master, or that possess high artistic value, or that represent a significant and distinguishable entity whose components may lack individual distinction; or

D) if they have yielded, or may be likely to yield, information important in prehistory or history.

Integrity is comprised of seven aspects that include: location, design, setting, materials, workmanship, feeling, and association. Several of these aspects must be possessed for a property to retain sufficient integrity for listing on the NRHP. The three aspects of integrity that are specifically relevant to archaeological sites are location, materials, and association. NRHP Criteria A, B, and C do not apply to the precontact sites identified for this project. The precontact components and sites were evaluated for their NRHP eligibility under Criterion D.

Specific historic contexts for the precontact period in Minnesota have been developed to summarize the extent of knowledge for each context and provide a framework to aid in determining whether a site has the potential to yield information that is considered important to local and regional prehistory. These contexts propose specific research questions and themes that are specifically relevant to each context. In order for the sites to be eligible for the NRHP under Criterion D, they must retain integrity and contain the potential to provide information on relevant research questions and themes that are applicable to the specific historic contexts at the sites. These historic contexts are discussed in detail below.

2.3.1 Archaic Contexts (12,500 to 2500 BP)

Site 21PL109 contains Early, Middle, and Late Archaic components, based on projectile points and radiocarbon dates. Historic contexts and basic research questions for the Archaic Period in Minnesota are presented in multiple documents (Anfinson 1997; Dobbs 1988; Gibbon 2012; Gibbon and Anfinson 2008). The very sparse and limited knowledge of this period requires addressing basic research questions about this culturally and environmentally dynamic period. Site 21PL109 has the potential to provide important information on the Archaic Period and is capable of addressing several important research themes, including:

 What are the ages of the components at the site, and how do they fit within the established chronology of the region?

 What specific complexes are present at the site, and how do these complexes relate to previously defined complexes in the region?

 What was the environment like at the site during the time of occupations and what changes in flora and fauna occurred during the Archaic period and how are these related to climatic changes?

 What internal developments, changes, and adaptations occurred during the Archaic period and how do these relate to climatic and environmental changes occurring at that time?

 What are the diagnostic artifact types (especially projectile points) from the components at the site, and are they similar to named types elsewhere or are there unique types in Minnesota or regional variants of named types in the state?

6

 What specific artifact types occur in the Archaic assemblages at the site and how do the compare with other Archaic groups from the Plains or Woodlands? What kinds of features are present?

 What specific site activities and functions are indicated by the artifact assemblages and features from the various components at the site?

 What were the lifeways, subsistence strategies, and settlement patterns during the Archaic period, as indicated by the Archaic site components and how do they compare with existing regional data? To what extent were they similar or dissimilar to Archaic lifeways in adjacent areas, and how did they change through time?

 What types of lithic technology were employed and are there changes through time or between different Archaic groups?

 What is the pattern of lithic material use and is there evidence for interaction and trade with other Archaic groups from the Plains or Woodlands? Where and how were exotic raw materials (e.g., stone) procured?

 What is the geomorphic context of the Archaic components at the site, and what site-specific environmental changes have occurred during the Holocene with respect to alluviation, soil formation, and site formation processes?

2.3.2 Woodland Period Contexts (2500 to 350 BP)

Undefined Woodland components are likely present at the site based on ceramics, a Besant point, and vertical patterning of artifacts on the terraces. Historic contexts for the Woodland period were initially developed by Dobbs (1988). Updated contexts have been prepared for the National Register of Historic Places Multiple Property Documentation Form (Arzigian 2008). Primary statewide Woodland Tradition research themes, which are relevant to these sites, are presented below (Arzigian 2008:12-16).

Primary Statewide Woodland Research Themes Chronology A fundamental need for understanding Minnesota’s Woodland complexes is an adequate chronology, including absolute dates for the full span of each complex, but particularly for the beginning and end, as well as charting important changes within the complex.

Technology and Material Culture Besides identifying diagnostic artifacts, the full range of material culture for each complex needs to be described. In addition to artifacts typically considered diagnostic, such as rim sherds and projectile points, can other region- or complex- specific cultural items be identified, such as unique pottery designs, bone tools, or patterns of raw material use?

Ceramics Ceramics provide the most sensitive chronological and regional marker for a complex, but many of the typologies are inadequate or outdated. There is a need for refining and updating existing ceramic typologies, developing a better understanding of spatial distribution and regional and temporal variations for ceramics, and conducting detailed attribute analysis. Changes through time and across regions need

7 to be explored. Comparisons also are needed between ceramic types used in Minnesota and those used in nearby regions (for example, how are Late Woodland corded ceramics in the southern part of the state related to the corded-ware horizon found across the Midwest?). Variability within many types of ceramics seems to be great but is also poorly understood. Single-component sites or separable components within stratified sites are needed to identify the range of contemporary ceramic types and varieties and how they change through time. Attribute analysis could generate a database of ceramic characteristics that could be analyzed statistically and modeled in GIS.

Ceramic manufacturing processes and vessel function are in general also poorly known. More detailed technological study of ceramics (e.g., paste, temper) could improve understanding, as could thin-section analysis, X-ray florescence, and diffraction, which can help to identify mineralogical and elemental composition and differentiate locally made vs. imported pottery.

Lithics Much more information is needed on the full range of Woodland lithic artifacts, both tools and manufacturing debris, and the raw materials used, both local and exotic. Lithic typologies need to be refined and their associations with cultural complexes verified. Trait comparison to Archaic, Middle Woodland, and Plains types is essential for distinguishing the points from those of other periods and regions, or for confirming that they are all part of a homogeneous complex. Any temporal changes or specific geographic distributions would be useful.

Lithic tools and debris need to be studied in terms of function, lithic reduction sequences, tool manufacturing, raw material selection, and changes through time in all of these. Can raw material debris profiles be developed to characterize these sites, and possibly to date them even if ceramics are not present? Single-component sites or multicomponent sites with a horizontally or vertically separate component are needed for this research.

More work is needed on the accurate identification of specific lithic sources, and on documentation of changes in the use of particular raw materials through time and space, and for different tool types. Existing collections might then need to be reexamined, and implications drawn for understanding trade and interaction with other regions. Additional data could help to answer questions related to lithic technology and raw-material acquisition and how those might have changed through time.

Further analysis is needed to identify any differences in lithic assemblages (tools, raw materials, etc.) between sites associated with mound construction and other habitation sites, between complexes in different areas, and between sites with different activities represented. What was the effect of the bow and arrow on the rest of the technological tool kit and on hunting practices, settlement, etc.? Where and when was bipolar core technology used?”

Subsistence More detailed information on subsistence is needed for all Woodland complexes in Minnesota. Additional sites with larger samples of subsistence remains are needed from a variety of habitats. Systematic fine-scale recovery from Woodland sites is needed, including flotation to recover plant and animal remains, fine lithic debris, and 8 other small artifacts. Also needed are specialized analyses of these remains, not just superficial analyses such as sorting fauna by class (e.g., fish vs. mammal).

Interpreting the variety of faunal taxa in terms of habitat selection and seasonal availability will be essential to understanding the whole Woodland seasonal round. Extractive strategies must be examined at the site, local area, and regional scales, including changes through time. Patterns will need to be considered with regard to both variable exploitative strategy and taphonomic changes, such as changes in patterns of transport, processing, and/or disposal of animals, and the final deposition of their remains.

Floral analyses need to include wood charcoal as a reflection of both the environment and cultural practices, as well as recovery and identification of macroplant remains such as seeds and nuts, and phytolith and pollen studies. Ceramics can be analyzed for evidence of phytoliths and pollen. Infrared spectrometry and gas chromatography can investigate cooking residues and fatty acids from products cooked in vessels, to identify how the vessels were used and what foods were consumed. The role of wild rice in precontact cultures is a crucial question. When was wild rice first used, and when did it become a prominent part of the economy? How did the use of other resources change? Are there special precontact features used to process wild rice? If so, can they be clearly identified, and can they be distinguished from postcontact ricing features? What cultivated plants were used by Woodland tradition populations in Minnesota? How did the northern limits of corn agriculture change through time? When did corn first appear in various regions? How did people exploit different resources as part of the broader annual round?

In addition to wild rice, where, when, and how were important specialized resources exploited, such as bison or sturgeon? Were sturgeon fisheries occupied for large parts of the year, or only for short periods? What was the nature of bison hunting in various regions, how did it relate to overall way of life, and how did it change through time (including in relation to environmental changes)? Were groups making use of seasonal bison hunts? Which groups, and at what times? Did some groups travel from one region to another as part of a large-scale seasonal round? Was there exchange of bison meat and products, and if so, with whom and in return for what?

Geographic Distribution The boundaries and geographic distribution of individual complexes are poorly known, and the bases on which they were defined are often not explicit.

Modeling (i.e., Mn/Model) Modeling could identify locations along rivers (such as trade routes) that share the characteristics of a complex, to target future field investigations. GIS can be used for site catchment analysis to suggest what resources might have been exploited at individual sites, and how this compares between sites across regions. Site function within the complex’s settlement system can be suggested, and multiple alternative explanations for site location and site function proposed and evaluated. How were ecotones exploited? In particular, what were the effects of the prairie/forest ecotone (and possible changes in this ecotone) on subsistence and settlement systems and movement of peoples across the ecotone? Did some areas, such as ecotonal areas, serve as central points, or trading or culture hubs? Were there regions that were transitional between a number of distinct complexes, and that would have made exposure to or intermarriage with other cultural groups more likely? Evidence of 9

distribution of ceramics or raw materials between different groups might document such patterns of interaction.

What effects did human subsistence and settlement systems have on the environment, including the prairie/forest ecotone? Were people using fires to maintain ecotonal and prairie habitats? Is there evidence of extensive areas of burning (such as in cores obtained from lakes or rivers)? Or evidence of natural resources that are dependent on fire, such as varieties of wood, plants, or animals?

Regional Interaction Research is needed into the full range of interregional interactions within and between peoples of contemporary cultures or complexes, as well as the relationships that helped to shape changes in cultures through time.

Defining the Complexes Finally, after evaluation of the research themes, the definition of each complex needs to be refined. Additional dating and understanding of the regional distribution and changes through time, as well as the relationships to other complexes and other regional populations, will facilitate development of meaningful archaeological phases.

2.3.3 Northeastern Plains Village Context (AD 1200-1650)

A Northeastern Plains Village component was identified at the site, based on ceramics and radiocarbon dates. Historic contexts for the Plains Village period were initially developed by Dobbs (1988) for Great Oasis and Cambria. The Northern and Southern Red River Study Units in northeastern North Dakota also summarize the extent of knowledge and propose research questions for this period (Picha et al. 2016).

Most of the research questions are of a general nature given the overall lack of existing data. The limited knowledge of these archaeological complexes in the northern Red River Valley region also requires addressing basic research questions that are fundamental in understanding the Plains Village period on the Northeastern Plains.

Based on a review of these documents, several general research questions and themes for the Plains Village period in the Red River Valley are proposed below.

 What are the ages of the Plains Village complexes at the site and how do they fit within the established chronology of the Red River Valley and adjacent regions?

 What specific Plains Village complexes are present at the site, and how do these complexes relate to previously defined complexes in the Red River Valley and adjacent areas of the Plains and Woodlands?

 What diagnostic artifact types are present at the site and how do these relate to previously defined diagnostic types in the Red River Valley and adjacent regions?

 What ceramic wares are present at the site, what are their diagnostic attributes, and how do they relate to wares from other Plains Village sites in the Red River Valley and adjacent regions?

 What are the subsistence/land-use strategies and settlement patterns during the Plains Village period in the Red River Valley? 10

 What types of lithic technology were employed and what types of lithic tools are present?

 What is the pattern of lithic material use and is there evidence for regional interaction and trade with cultural groups from the Plains or Woodlands?

 Conduct paleo-environmental research, focusing on stratified lakebed deposits, to provide an updated local climatic history for the Red River Valley.

 Gather baseline data concerning regional stone, ceramic, bone, and shell technologies.

11

3. ARCHAEOLOGICAL FIELD METHODS

3.1 Archaeological Field Methods

The Phase I and II archaeological survey and testing methods adhered to the MnSHPO and OSA guidelines for archaeological fieldwork. Specific field methods were discussed with MnDOT prior to conducting fieldwork. The survey design included an archaeological survey for the entire project APE.

3.1.1 Pedestrian Survey

The goal of the pedestrian survey was to identify and record archaeological sites that could be observed on the ground surface. Pedestrian survey was conducted in the survey area by walking transects parallel to the roadway in intervals not exceeding ten meters. The pedestrian survey was a practical method for identifying archaeological resources that could be observed on the surface such as artifact scatters, pits, earthworks, or historical foundations. Pedestrian survey was particularly useful in agricultural fields that had adequate surface visibility (greater than 25%), and site 21PL109 was identified by this method. Close-interval survey was conducted in two-meter intervals within a ten-meter radius of all find spots at sites in agricultural fields.

3.1.2 Shovel Tests and Deep Auger Tests

Shovel and auger testing was used to identify artifacts and features not visible on the ground surface, characterize soils at survey areas and sites, and provide information on the horizontal and vertical provenience of artifacts at the sites.

Phase I shovel testing was conducted at 15-meter intervals in three areas: 1) on the point bar terraces of the Red River because this landscape has a high potential for stratified archaeological sites below the plow zone; 2) on those portions of the lake plain near the river oxbow that had low surface visibility; and 3) and at locations of site 21PL109 where artifacts were found on the surface in agricultural fields on the lake plain. All areas not shovel tested had adequate surface visibility or low archaeological site potential. Shovel test transects were placed parallel to the roadway.

Shovel tests were 35 to 40 cm in diameter. Shovel tests were typically dug to 100 cmbs on all terraces, 80 cmbs on the lake plain in the bean field, and 90 to 100 cmbs on the lake plain in the yard and grass near the oxbow. Shovel Tests 13 to 44 were typically dug to 80 to 90 cmbs. Soils were dug and screened in 20 to 30 cm increments to provide vertical control of artifact provenience. A small number of tests, which included some of the radials around Shovel Tests 141, 142, 146 to 151, 159, and 160, were augered to 130 cmbs with an 8” auger after they were dug to 100 cmbs with a shovel.

At archaeological sites, Phase II close-interval shovel testing was typically conducted at five-meter intervals in cardinal directions adjacent to positive shovel tests in order to gather additional artifact data, assess site integrity, and define site limits and intrasite artifact patterning. Shovel test data were used to guide the placement of excavation units within portions of the site that have the highest potential to yield data for answering important research questions and evaluating the site.

Because of the potential for deeply buried sites on the Red River floodplain, a Seymour auger with a 20.3-cm (8-inch) diameter bucket was used for deep auger testing below 85 cmbs in each shovel test hole. Each auger sample typically recovers between 15 and 20 cm of soil. A sample of the soil from each auger was collected and laid out in sequence so that the soil profile could be reconstructed. Following the MnDOT protocol for deep-site testing, two deep auger tests were dug at each test location to recover a volume of soil equivalent to a standard shovel test. However, only a single 12 auger was dug in areas with deep fill, historic deposits, peat, or a lack of buried surfaces or cumulic soils. In such cases, the auger test basically provided a soil profile for assessing the potential of the soil for containing archaeological sites. Auger tests were dug to 300 cmbs on the floodplain adjacent to the bridge. All soil, except fill, was screened through 1/4-inch hardware mesh. Clayey soil that was difficult to dry screen was waterscreened in stock tanks, using water pumped from the river. The field crew returned all excavated soil to each test upon completion. All shovel test locations were recorded with a GPS unit.

3.1.3 Excavation Units (XUs)

XUs were 1-x-1 meter in size. XUs were dug and recorded in 10-cm levels below a datum, whose relative elevation was established in relation to the adjacent ground surface. Excavation depths were measured in cm below datum (cmbd). The datum line is illustrated on the XU wall profiles. The extent and types of soil disturbance were recorded for each level to aid in assessing site integrity. Soil was screened through ¼-inch hardware mesh. The units were backfilled after excavation was complete.

3.1.4 GPS Data Collection and Site Mapping in ArcView

GPS data was collected with a Trimble GeoExplorer 6000 for find spots, shovel and auger tests, and XU corners. The data has a typical positional accuracy of 15 to 30 cm after post-processing. This data was then exported to create maps on topographic and aerial imagery. Notes on the field conditions were also recorded for each field parcel.

3.1.5 Field Documentation

A record of daily activities was recorded in a log that documented fieldwork and relevant information on the survey areas and sites. Photographs were taken of archaeological sites, survey areas, and wall profiles of the XUs. A record of the photographs was maintained in a photo log.

Excavation level forms were maintained for each level of an XU and were filled out after the completion of each level. These forms contained information on excavation methods, soils, artifact counts, disturbances, and other relevant observations.

A soil profile was drawn for representative shovel tests, positive shovel tests, and one wall of an XU. Soil colors, textures, horizons, and disturbances were recorded on the profile. Soil colors were described using the Munsell system, and the soils were moistened prior to determining color.

13

4. ARCHAEOLOGICAL LAB METHODS

4.1 Artifact Processing

Artifacts were cleaned, analyzed, and cataloged by FCRS staff. The artifact assemblage consisted of faunal remains, lithic debris, stone tools, cores, ceramics, and fire-cracked rock (FCR).

Artifact catalog numbers are comprised of a provenience bag number and a specimen number, following the MHS system. The provenience bag number is represented in the catalog database by the column titled “Prov.”, and the specimen number is represented by the column titled “Specimen #”. The artifact catalogs for the sites are contained in Appendix C.

Provenience bag numbers were established by FCRS in the lab and consisted of a unique number assigned to each specific provenience by find spot (FS), shovel test (ST), or excavation unit (XU) by depth (“cmbs” for cm below surface). For example, Prov # 1 would represent Shovel Test 1 (ST 1), 0-20 cmbs, and Prov # 2 would represent ST 1, 20-40 cmbs. The specimen portion of the artifact catalog number is a unique sequential number or number range assigned to artifacts within a specific provenience bag number. Individual artifacts were assigned a single number (e.g., 1.1), while artifacts with similar attributes and size grades were grouped together and assigned a sequential specimen number range based on their count (e.g., 1.2-10). Beginning and ending numbers in the range were recorded in one row of the database with attribute data for related artifacts.

Attribute data recorded in the catalog for each artifact, or group of artifacts, included: site number; provenience bag number; specimen number(s); provenience information; artifact class; artifact descriptions; weight (grams); and size grade (inches). Additional artifact information was entered in the “Notes” field of the catalog. The descriptive categories that apply to each artifact class are summarized in Table 1. Specific descriptive attributes recorded for each artifact class are discussed in detail in the following artifact sections. All data was entered in a Microsoft® Access 2010 database. Fields left blank in the database indicate that the attribute does not apply or that the attribute is absent.

Table 1. Descriptive Categories for Artifact Classes in the Catalog. Description Description Description Description Description Description Description Class 1 2 3 4 5 6 7 Lithic Cortex Heat Lithic Debris Flake type N/A N/A material amount treatment Tool Tool flake Lithic Cortex Heat Lithic Tool Tool type category type material amount treatment Flake Platform Lithic Cortex Heat Lithic Core Technology removals modification material amount treatment Fire-cracked Lithic Lithic FCR type N/A N/A N/A N/A rock material Element/ Thermal Faunal Class Portion Modified N/A N/A Side alteration Presence/ Vessel Surface Decoration Absence of Ceramic Temper Condition N/A portion treatment type charred residue Botanical Material Type Portion N/A N/A N/A N/A

14

Gilson standard-testing metal sieves were used for size grading. The following size grades (SG) were used to sort artifacts: >4.0 inch (SG00); <4.0 to >2 inch (SG0); <2 to >1.0 inch (SG1); <1.0 inch to >0.5 inch (SG2); <0.5 inch to >0.233 inch (SG3); and <0.233 inch (SG4). Weight was measured to the tenth of a gram with an electronic scale. Artifacts weighing less than 0.05g were given a weight of “0”.

4.2 Lithic Raw Material Resources by Kent Bakken

The project area lies in the South Agassiz Resource Region (Figure 2) near the border of the Tamarack and Upper Red subregions (Bakken 2011). The portion of these subregions along the Red River, which includes project area, is on the Glacial Lake Agassiz lake plain, which is an area of thick lacustrine deposits that are largely devoid of naturally occurring rocks. In addition, there are no bedrock outcrops in the area. The closest likely toolstone concentrations would have been the Glacial Lake Agassiz beaches located some miles to the east and west. Toolstone availability varies locally within the beaches, but in places cobble concentrations occur (cf. 21RO11, Peterson 1973). It is worth noting that rock does occur in places on the lake plain where it was carried in on floating ice rafts, which then melted and deposited the rock on what would then have been the lake bottom. These are generally not rich sources, however, and it is not clear that there are any in the vicinity of the project area. Very few natural rocks were observed during survey.

The lack of immediate toolstone sources is important for understanding raw material use in the region. In a sense, all toolstone had to be "imported," albeit not always from particularly distant sources. Most toolstone would have come into the lake plain in the form of tools, blanks, or possibly cores. This should mean that there is less early-stage flaking debris in local lithic assemblages, and may also mean that there is less raw material diversity in the assemblages since expedient technologies may not be well represented.

Regional raw materials would be those associated with Des Moines lobe till (Table 2). Most of these materials have geological origins to the north-northwest. The most important raw material would be Swan River Chert, originating in west–central Manitoba and east-central Saskatchewan. Red River Chert, originating in central Manitoba, would also be important but not as common. In the Upper Red subregion, the till is also likely to contain materials of more westerly to northwesterly origin. These were apparently carried into the region by an early and poorly understood glacial advance, the later incorporated and redistributed by a later glacial advance. The most important of these is Tongue River Silica (TRS). Since TRS is largely missing in the Tamarack Subregion, we could expect that it would be available in toolstone sources south of the project area but probably not in sources north of the project area. In addition, occasional pieces of Knife River Flint, silicified wood and related materials might be found in sources to the south of the project area although they would not be common. Finally, these southern sources could include pebbles of agate and other materials originating as far west as the Rocky Mountain (see Bluemle 1972). These were transported east in preglacial rivers as part of the bed load; the river gravels were later incorporated in and redistributed by glacial ice.

Materials of a more northerly to northeasterly origin might also occur in this region. Specifically, this would include Lake of the Woods Rhyolite and related siltstones. Clasts of these materials are probably both larger and more common in the Tamarack Subregion than the Upper Red Subregion, although they are probably not an especially important toolstone in either. Small kipalus, or fragments of oolitic jasper, may have been transported by the same ice masses (Prest et al. 2000). These seem to be generally indistinguishable from classic Jasper Taconite. However, because of their small size they were probably not suitable for use, even for expedient pebble technologies. Because these come from the southern Hudson Bay region, which is also the source of Hudson Bay Lowland 15

Chert (HBLC), it is conceivable that HBLC could also occur in regional sources. However, a program of sampling potential toolstone sources in northwestern Minnesota failed to identify HBLC (Bakken 1985). Quartz should probably be available, although it would be of relatively poor quality and a minor resource. Any other materials would be rare and constitute a very minor resource. The main exotic material in the area is Knife River Flint, which can constitute more than half of some assemblages. Obsidian occurs but is not common, and the same is true of other nonlocal raw materials.

CSAH 1 Project A

Figure 2. Lithic Resource Regions of Minnesota (adapted from Bakken 2011).

16

Table 2. Estimated Primary, Secondary, and Minor Lithic Raw Material Status by Region and Subregion (Bakken 2011). Primary Raw Minor Raw Main Exotic Regions Secondary Raw Materials Materials Materials Raw Materials South Agassiz Resource Region Quartz Tamarack Swan River Chert Border Lakes Tongue River Silica Knife River Flint Subregion Red River Chert Greenstone Group Western River Gravels Group ? Border Lakes Red River Chert Greenstone Group Upper Red Swan River Chert Tongue River Silica Western River Knife River Flint Subregion Quartz Gravels Group Knife River Flint Border Lakes Greenstone Group Western River Tongue River Silica Shetek Gravels Group Knife River Flint Swan River Chert Red River Chert Subregion Knife River Flint Burlington Chert Quartz Fat Rock Quartz Other West Superior materials West Superior Resource Region Quartz Arrowhead Gunflint Silica Border Lakes Hudson Bay Lowland Chert Knife River Flint Subregion Knife Lake Siltstone Greenstone Group Jasper Taconite Lake of the Woods Rhyolite Biwabik Silica Knife Lake Siltstone Gunflint Silica Knife River Flint Quartz Tongue River Silica Swan River Chert Jasper Taconite Hixton Group Subregion Quartz (Fat Rock Kakabeka Chert Burlington Chert and other) Hudson Bay Lowland Chert Lake Superior Agate Pipestone Resource Region Sioux Quartzite Tongue River Silica Swan River Chert ? Quartz Knife River Flint Gulseth Silica ? Red River Chert ? Hollandale Resource Region Cedar Valley Chert Quartz Galena Chert Tongue River Silica Shell Rock Chert ? Hixton Group Grand Meadow Chert Swan River Chert Prairie du Chien Chert Red River Chert

4.3 Lithic Analysis Methods

The analysis of lithics focused primarily on the identification of raw materials, lithic technologies, and specific types of flakes, tools, and cores. Information on site function, lithic economy, lithic technologies, settlement patterns, and regional interaction may be inferred from this data. Raw material, weight, size grade, and presence/absence of cortex were recorded for all lithics. Lithic debris was examined for macroscopic evidence of modification, such as use-wear or retouch. All lithics were examined using a 10x magnification hand lens, which was useful for identifying micro- flaking, lithic material, and other features not visible without the aid of magnification.

17

Frank Florin conducted the lithic raw material identifications, and he has extensive experience in the raw materials of the region, including a lithic comparative collection. Published guides to lithic resources of Wisconsin, Minnesota, and the Upper Midwest were also consulted (Bakken 1997, 2011; Gonsior 1992; Morrow 1984, 1994; Morrow and Behm 1986).

4.3.1 Thermal Alteration

Thermal alteration, commonly known as heat treatment, is the intentional alteration of a lithic material to improve its flakability. Heat treatment produces an increase in surface luster, intensifies ripple marks on flake scars, and creates reddish to orangish color in many cherts and other light- colored materials. In some materials, such as Tongue River Silica, Swan River Chert, and Prairie du Chien Chert, the effects of heat treatment are fairly well-documented and can be discerned with a good degree of accuracy. In the current analysis, materials were classified as heat treated if there was significant and noticeable reddish to orangish color and an increase in luster. If these color and texture traits were subdued, then the piece was coded as “probably heat treated”. The effects of heat treatment on some materials are not well known.

In contrast to heat treatment, burning is defined by excessive heating that often compromises the stone’s flakability. Traits of burning include potlid spalls, crazing, and cracks on the artifact’s surface, and a notable darker color. Burning is interpreted to be unintentional, being caused either by accidental over-heating during the heat treatment process or by discard into a cooking facility.

4.3.2 Lithic Debris

Lithic debris includes flakes, flake fragments, and pieces of shatter that were produced from cobble testing, core reduction, stone tool manufacturing, and stone tool maintenance. The analytical methods used in this report are based on the results of previous lithic studies and experimental replications (Bradbury and Carr 1995; Callahan 1979; Cotterell and Kamminga 1987; Flenniken 1981; Hayden and Hutchings 1989; Inizan et al. 1999; Magne 1985, 1989; Odell 1989; Root 1992, 1997, 2004; Tomka 1989; Yerkes and Kardulias 1993). These studies indicate that lithic-reduction stages and technologies can be inferred from diagnostic flake attributes.

The most promising results are derived from studies that consider a combination of several flake attributes from a large sample of lithic debris. The work of Mathew Root (2004) provides the basis for much of the current analysis because of his extensive lithic replicative studies and their relevance to the current project with regards to cultural context, regional location, comparable raw materials, and lithic technologies. This analytical framework has been used for several large data recovery projects in North Dakota, including Lake Ilo 32DU955A (Ahler et al. 1994), 32RI785 (Root 2001), and Beacon Island 32MN234 (Mitchell and Johnston 2012). Root’s methodology and results are supported by the lithic studies referenced in the preceding paragraph, which tend to focus on more specific aspects of technology and flake attributes. Similar technological approaches based on flake attributes from replicative studies have been developed in other lithic studies (Callahan 1979; Fleniken 1981; Flenniken et al. 1990; Magne 1985; Ozbun 1987). While Root’s work is primarily oriented to bifacial technologies of Knife River Flint, other studies consulted for this analysis provided information on bipolar and nonbifacial technologies.

The lithic analysis utilized for this project assessed multiple flake attributes that were identified as technologically diagnostic in numerous studies. These attributes define the specific flake types used in the current analysis, which are summarized and described in Table 3. The lithic analysis was accomplished by 1) identifying specific flake attributes; 2) comparing the attributes with those defined for specific flake types; and 3) making a determination as to flake type. The lithic analyst,

18

Frank Florin, has moderate experience in lithic replication and has a comparative collection of flake types comparable to the ones used in this study.

Flake attributes examined in this analysis include the following morphological and technological characteristics: presence/absence of cortex; presence/absence of percussion bulb; presence/absence of bulbar scar; extent of platform modifications and preparations (grinding, battering, and faceting); platform size; platform angle; number of dorsal flake scars; flake morphology; flake thickness; and size grade. These attributes have been determined to be diagnostic of specific lithic-reduction technologies and stages.

Decortication flakes are indicative of cobble testing and early-stage core reduction, and in this study are linked to nonbifacial technology. Bifacial technology is indicated by bifacial thinning flakes and shaping flakes, alternate flakes, bifacial cores, and bifacial tools. Bipolar flakes and bipolar cores are indicative of bipolar reduction. Nonbifacial technology is indicated by nonbifacial flakes, decortication flakes, tools made on nonbifacial flakes, and nonbifacial cores. Shatter is most strongly associated with cobble testing, core reduction, and the earlier stages of reduction. Types of lithic debris that are not indicative of specific technologies or reduction-stages include “other size-grade 4” (other SG4) flakes, broken flakes, and unidentified flakes. Some materials, like quartz, which do not have conchoidal fracture properties, are likely to result in greater amounts of nondiagnostic flake types than other materials.

Table 3. Definitions of Technological Flake Types (primarily adapted from Root 2004). Technological Definition Flake Type Decortication flakes have most (>50%) of their dorsal surface covered with cortex. They are associated with raw material testing and the early stages of core and tool reduction (Root 2004). These flakes have a large striking platform and a bulb and bulb Decortication scars that are nearly always quite pronounced as a result of direct percussion with a hard Flakes hammer (Inizan et al. 1999). Other traits of these flakes include: a large flake platform angle (60-90 degree range); whole flakes are typically are SG1 or SG2; typically two or less flake scars on the dorsal surface; and a relatively thick cross-section. These are specialized flakes defined by the presence of 1) parallel or subparallel lateral margins; 2) dorsal flake ridges that are parallel or subparallel with the lateral margins; 3) at least two flake-removal scars evident on the dorsal surface; 4) an axis of applied Blade Flakes force that is approximately parallel with flake’s margins; 5) a length-to-width ratio of at least 2:1; and 6) plano-convex ,triangular, rectangular, or trapezoidal cross sections (Crabtree 1972:42-43; Root 2004; Whittaker 1994:33). These exhibit the following attributes: 1) shattered or pointed platforms with little or no surface area; 2) wedging flake initiations; 3) evidence that force has been applied to both ends of the flake, such as crushing on opposite ends; 4) no bulbs of force (due to Bipolar wedging initiations); 5) pronounced compression rings from compression-controlled Flakes flake propagation; and 6) a generally parallel-sided plan form (Root 2004; see also Flenniken 1981). Flakes classified as bipolar must exhibit most but not all of these attributes. Bipolar flakes do not exhibit positive bulbs of force on opposite ends of the same flake interior surface. Broken flakes are flake fragments that lack a bulb of percussion, platform, or Broken other diagnostic features that would enable a determination of flake type. Such Flakes flakes are typically distal or medial flake fragments. Broken flakes occur in all technologies and are produced during all stages of lithic reduction. Unidentified These flakes do not fit any of the previously described types. Flakes

19

Table 3. Continued. Technological Definition Flake Type These flakes are strongly associated with percussion bifacial thinning (Root 2004). Bifacial thinning flakes without platforms exhibit the following attributes: 1) thin curved long sections; 2) extremely acute lateral and distal edge angles; 3) at least three dorsal flake scars (usually more) that originate from different directions, especially other than the flake itself; 4) 20% or less cortex; and 5) an expanding shape in planview. Bifacial Thinning Flakes with platforms exhibit attributes 1-5 along with 6) a bending initiation and 7) a Flakes – narrow and faceted striking platform without cortex. Proximal flake fragments that (early to consist mainly of a platform are classified as bifacial thinning flakes if they have the middle-stage) above attributes. Flakes with platforms often have a lip at the intersection of the striking platform and the flake ventral surface (caused by a bending flake initiation), and flakes with distal ends usually have feathered terminations.

Soft-hammer percussion with a billet is typically used in the removal of these flakes. The flaking angle is acute, the bulb is diffuse, and there is often abrasion on the overhang (platform) (Inizan et al. 1999). These flakes are usually small, less than < 1/4 inch (SG4), but can be larger (Root 2004). Only flakes SG3 or smaller are classified as bifacial pressure flakes. They are relatively thin with multifaceted and ground platforms, according to Root (2004), but this analysis also includes flakes with single or indeterminate facets that accommodate raw materials in the assemblage. Flakes must retain a platform to be placed in this class. Occasionally, small flakes produced by late-stage percussion bifacial shaping Bifacial possess the defining attributes of pressure flakes. Flakes produced early in the pressure Shaping flaking process have multiple scars on their dorsal surfaces and are curved in long Flakes section and slightly expanding, or petaloid, in planview. by pressure or percussion – Flakes produced during final bifacial pressure flaking have parallel sides. These flakes (late-stage) are generally produced during bifacial pressure flaking. Whether produced by pressure or percussion, these flakes are associated with final bifacial shaping (stage 5 as defined by Callahan [1979]) and bifacial tool maintenance. Because of similar flake traits, this type also may include some unifacial retouch flakes from unifacial cutting tools that have a thin, knife-like edge. This would include flakes derived from retouch flakes and flakes from points made on a uniface. However, such flakes and tools are uncommon in the assemblage compared to bifacial tools. Unifacial flakes from beveled unifacial tools (scrapers) are categorized separately as unifacial retouch flakes (discussed below). Nonbifacial flakes are size-grade SG1 to SG3 and do not have the defining attributes of bifacial or decortication flakes. Diagnostic traits include 1) simple platforms with minimal platform modifications (often with no facets but up to one or two facets); 2) large platform angles (60-90 degree range); 3) generally less than three dorsal flakes scars that are likely to be unpatterned; and 4) may have bulbar scar on ventral side (Andrefsky 2005; Magne 1985, 1989; Odell 1989, 2003:126; Tomka 1989; Yohe 1998). Platform areas may be partially or wholly obliterated from hard hammer percussion. This flake type is comparable to Root’s (2004) “simple flakes”.

Nonbifacial In general, these flakes have relatively thick cross sections, steep lateral edge angles, Flakes and straight or slightly curving profiles. The amount of dorsal surface cortex typically ranges from 0 to 50%. This class contains conchoidal flakes that have a bulb of percussion and bending flakes. While these flakes are produced in biface reduction, particularly the earliest stages, they are most strongly associated with cobble testing, unprepared nonbifacial cores for flake blank production, and the early stages of nonbifacial tool reduction.

Included in this type are flakes classified as “interior flakes”, which are removed from the interior of the core or cobble, with no cortex on their surface (Fleniken et al. 1990; and Yerkes and Kardulias 1993). 20

Table 3. Continued. Technological Definition Flake Type Shatter includes angular, cubical, and irregularly shaped chunks that lack the following: bulbs of force, systematic alignment of fracture scars on faces, striking platforms, and points of flake initiation. Interior (ventral) and exterior (dorsal) surfaces and proximal Shatter and distal ends cannot be determined on these pieces (Root 2004). Shatter may be the result of poor-quality stone with fractures along bedding planes or other material flaws. Shatter is created by most production technologies but is most strongly associated with cobble testing, core reduction, and earlier stages of reduction. Alternate flakes are produced when beveled edges are created from: 1) squared-off or thick edges, such as those on tabular cobbles; 2) the thick margins of flake blanks (especially at the proximal end); 3) margins with stacked-step terminations; and 4) broken flakes or bifaces. The result is the creation of a bifacial (beveled) edge that Alternate prepares it for bifacial thinning or shaping by producing edge angles appropriate for use Flakes as platforms (Flenniken et al. 1990; Root 2004). They are thick in relation to their length and width, are triangular in cross section, have a squared edge (often cortical) adjacent to the platform (this is part of the squared edge of the object piece), have single-faceted platforms, and have a skewed orientation in relation to the axis of percussion. A flake removed from the edge of a flake blank or core to change the angle of the edge Edge to facilitate flaking in order to prepare the blank or core for further reduction (Flenniken Preparation et al. 1990). Bifacial edge preparation flakes usually have thick and wide platforms and Flakes are short in length. A flake expelled from the surface of a lithic artifact by heat-induced differential expansion when overheated in a fire, as opposed removal by the flintknapping process Potlid Flakes (Flenniken et al. 1990). The flake has a flat dorsal surface and a convex ventral surface and is shaped somewhat like the inverted lid of a pot. Other size-grade 4 (SG4) flakes (< 1/4 inch in size) are either too small to be reliably Other identified using the diagnostic attributes of the other defined flake types or they simply Size-Grade 4 lack diagnostic attributes (Root 2004). These are produced in all reduction (SG4) Flakes technologies, including cobble testing. These flakes are likely to be underrepresented in lithic assemblages because their small size makes them less likely to be recovered. These were defined by study of experimental replications of unifacial tool manufacture (see Root 2004). Flakes must be complete or, nearly complete and retain a platform to be placed in this class. Typically, they also have (1) feather terminations; (2) single- Unifacial faceted, noncortical platforms; (3) plano-convex to biconvex cross sections; (4) parallel Retouch to expanding lateral flake margins; (5) long sections that are straight with a slight curve Flakes sometimes present only at the distal end; (6) at least two dorsal flake scars that originate from the same direction as the flake; and (7) an axis (platform) of percussion or pressure that is about 90 degrees. These are typically from unifacial tools with a beveled edge, such as scrapers, as indicated from the platform criteria in (2 and 7) above.

Mass aggregate analysis based on size grades (see Ahler 1989) was deemed not useful for determining lithic technology and reduction stages because SG4 artifacts were not systematically recovered, as the soils were screened through 1/4-inch mesh. The recovery of SG4 debris and large samples is imperative for conducting mass analysis within the established interpretive models. In addition, aggregate analysis draws its inferences from experimental replicative data sets that do not exist for the raw materials at the sites identified in the project area. There are other weaknesses of this method related to the accuracy of separating reduction stages and technologies from sites with mixed components (Andrefsky 2001:5).

21

4.3.3 Lithic Tools

Overview

Stone tools were vital to prehistoric lifeways, and they were used for a variety of tasks: perforating, cutting, sawing, scraping, boring or drilling, graving, whittling or slicing, chopping, pounding, and abrading.

Tool categories were defined by technological attributes (bifacial, unifacial, or pecked/groundstone) and by whether the tool was patterned or unpatterned. Patterned or formal tools include types in which the original shape of the flake blank or raw material has been substantially modified through a systematic sequence of reduction or retouch to produce a specific form that exceeds minimal functional requirements. In patterned tools, the shape of the tool reflects a distinctive style or cultural template. Projectile points, end scrapers, and bifaces are examples of patterned tools. Unpatterned or informal tools include types that were not substantially modified and still largely reflect the original shape of the flake blank or raw material. They lack the complex manufacturing methods of patterned tools and reflect an expedient technology. Flaking is typically restricted to the margin of the artifact. Utilized flakes and retouched flakes are examples of unpatterned tools.

Tool types and their inferred functions (e.g., projectile points, scrapers, cutting tools, etc.) were defined by technological attributes in conjunction with morphological attributes (form), general edge angle, size, and results from micro-wear studies that provide supporting evidence for general tool function (Kooyman 2000:164; Root 2001; Vaughan 1985; Yerkes 1987).

The use-life of a tool is an assessment of its estimated stage of manufacture and reason for discard. Use-life categories include the following: 1) unfinished tools that were not broken; 2) tools that are finished and in working condition; and 3) broken or worn out tools. This information was entered in the “notes” column of the catalog.

Numerous studies indicate that microwear analysis, which uses high-powered magnification to examine the edge of a tool in an attempt to identify the type of material that was worked by the tool and the type of motion with which the tool was used, is necessary to determine a tool’s specific function (Keeley 1980; Odell 2003; Semenov 1976; Vaughan 1985; Yerkes 1987). Microwear studies clearly indicate that there can be a low correlation between tool form and specific function, as tools from different form classes were used for the same task, and a single tool form was often used for multiple functions (Yerkes 1987:128). These studies reveal that there is much more functional variation than is typically assumed from the traditional form-based tool classification.

Microwear studies also indicate that there is some viability to inferring general tool function from the form-based classification, especially for certain tool types. For example, scrapers defined morphologically by a steep working edge often correlate with micro-wear studies that show tools with steep working edges were used for scraping bone, wood, and hide (Kooyman 2000:164; Root 2001; Vaughan 1985; Yerkes 1987).

Of course, without microscopic examination of the edge wear, there is no way to tell what material was scraped. Also, microwear analysis often reveals greater functional variation than can be inferred from typological and technological classification alone (Odell 1996; Vaughan 1985). For example, some “scrapers” were also used for tasks such as cutting, engraving, wedging, shaving, chopping, and shredding. In some cases “scrapers” bear no evidence of use as scrapers. Many projectile points were also used for cutting, shaving, engraving, scraping, and drilling. Other bifacial tools were used to saw bone, antler, or wood as often as they were used for cutting meat (Yerkes 1987:186).

22

Thin, sharp-edged flake and blade tools (such as utilized and retouched flakes) generally correlate with microwear studies confirming their use as cutting implements (Kooyman 2000:164; Odell 1996; Root 2001; Yerkes 1987). Again, the specific material worked or specific use cannot be determined without microscopic examination of wear patterns. Some studies that tested the accuracy of identifying utilized flakes without magnification indicated a low success rate, as the multiple processes (besides use as a tool) that can produce edge wear are not discernible without microscopic analysis (Shen 1999; Young and Bamforth 1990). These processes include wear caused by flake production, artifact trampling, excavation damage, and artifact movement in the soil. The studies show two primary causes of incorrect identification. First, utilized flakes that exhibit no macroscopic wear go unrecognized as tools. Second, use-wear is incorrectly attributed to use as a tool when it is actually created by some other cause.

Despite the benefits of microwear analysis, there are several limitations that hinder its usefulness and practicality. The time and money needed for such analysis is often not available in contract work, few individuals have the necessary training and expertise, and microscopic equipment is not available in most labs. Further, experimental studies have not been conducted on many of the lithic materials that occur in the artifact assemblages in Minnesota. It has also been found that microwear analysis does not necessarily produce conclusive results. Blind tests revealed the accuracy of tool function to be 76 percent for high-power technique and 68 percent for the low-power technique (Yerkes 1987:115). The accuracy of identifying the material worked was 62 percent for high-power technique and 32 percent for low-power technique. Finally, micro-wear analysis may not clearly identify functions of a single tool edge that was used for different tasks, nor may it identify the function use of a tool used for a short time or on very soft materials that do not cause observable wear.

Stone Tool Techno-Morphological Categories and Descriptions

Tool types recovered from sites in the project area are discussed below.

Utilized and retouched flakes are unpatterned flake tools that have a sharp, narrow-angled working edge, which is not beveled. Utilized flakes have no intentional modification but do have a series of micro-flakes (use-wear) that were removed along the working edge during use. Retouched flakes are minimally modified by pressure flaking along the working edge, presumably to shape the edge for optimal use. The micro-flakes on utilized flakes are distinguished from retouch flakes by their smaller size. Use-wear and experimental studies indicate that these are typically light-duty cutting, slicing, scraping, and sawing tools that were used on soft materials (meat, hides, and plant material) or moderately resistant materials (wood and bone). These tools suggest that site activities may have included butchering, animal/plant processing, hide working, and bone and woodworking.

Scrapers are patterned flake tools that have been pressure flaked along a distal or lateral end to form a steeply beveled (wide-angled) edge that is optimum for scraping. End scrapers have a distal working edge that is generally shorter or the same length as the lateral side and may have been hafted. Side scrapers have the working edge along the longest side of a flake and were likely not hafted. Side/end scraper have working edges along both the long and short sides. Scrapers are typically associated with scraping tasks on a variety of soft materials (meat, hides, and plant material) or moderately resistant materials (wood and bone).

Projectile points are bifacial tools with a sharp-pointed distal end and proximal hafting elements. These tools were used for hunting, and larger points may have also been used as cutting tools. Published guides to projectile point types of Minnesota, Iowa, Wisconsin, the Upper Midwest, and the Northeastern Plains were consulted to aid in identifying the points (Alex 2000; Boszhardt 2003; Goldstein and Osborn 1988; Justice 1987; Kehoe 1966, 1973, 1974; Morrow 1984). Projectile points indicate that site activities were associated with the procurement of game animals. 23

Hammerstones are generally rounded stones that have pitting on one or more surface, which resulted from striking a hard material. They were used for flint knapping, processing foods such as acorns, or marrow extraction from animal bones.

Anvilstones are larger stones that have a relatively flat, pitted surface. Generally, the pitting is localized to a specific area and is the result of impact with another stone during bipolar reduction or food processing.

Bifaces are classified into five stages after Callahan (1979), although Callahan’s final stages are condensed in this scheme (cf. Odell 2003; Root 1999). The unfinished bifaces could have been used as tools in an unfinished state, although it is likely that their intended final form would have been projectile points. The bifaces from the current project include broken and whole specimens.

A Stage 1 Biface is a flake blank, a tabular piece of material, or a cobble that was obtained for reduction. Stage 1 bifaces were not identified in the assemblage, as flake blanks are generally classified as primary flakes, and there were no unworked cobbles.

A Stage 2 Biface has initial edging that is characterized by the following: bifacially flaked edges in which relatively widely-spaced scars produce a sinuous outline in lateral view; conchoidal flake scars with cones of force from hard-hammer percussion; minimal shaping; flakes often do not extend to the midline; irregular outline and cross section; and width to thickness ratio ranges from 2:1 to 3:1.

A Stage 3 Biface has primary thinning that is characterized by the following: major projections and irregularities removed edges straightened so they are less sinuous; ridges and humps removed by thinning; production of flakes with bending initiation from billet percussion; lack of cones of force; flakes that often extend to or past artifact midline; edge angles in the 40-60 degree range; and width to thickness ratios of 3:1 to 4:1.

A Stage 4 Biface has secondary thinning and shaping that is characterized by the following: a thin, flat to biconvex cross section; regular edge shape; edges with beveling and grinding; little to no cortex; production of flakes with bending initiation from billet percussion; lack of cones of force; flakes often extend to or past artifact midline; edge angles in the 25-40 degree range; and width to thickness ratios that range from 4:1 to 5:1.

A Stage 5 Biface has undergone final shaping and hafting preparation (partial or full) and is characterized by the following: pressure flaking or light percussion flaking to form a specific shape, especially along margins; edge beveling or grinding; removal of percussion platforms; pressure flaking of notches and stem shape; and basal grinding.

4.4 FCR Analysis Methods

4.4.1 Definition of FCR

Stones used for cooking or heating, referred to here as fire-cracked rock(s) (FCR), are artifacts with distinctive characteristics caused by heating to high temperatures in a fire (House and Smith 1975; Jackson 1998; Latas 1992; Lovick 1983; McParland 1977; Taggart 1981; Thoms 2009). FCR includes both fractured and unfractured rocks that have been thermally-altered and lack other forms of cultural modification, such as flaking, pecking, polishing, or use wear.

Stones used for cooking or heating are generally cobbles of locally available materials that were chosen for their accessibility and predictable thermal qualities. These cobbles, which become FCR after heating, were generally larger than eight cm in diameter (Wentworth 1922). The types of 24 cobbles chosen for heating or cooking were usually coarser than stones used for flintknapping (Lovick 1983) and commonly include quartzite, granite, basalt, sandstone, and limestone. Experimental studies show that igneous rocks are better able to withstand thermal stresses than metamorphic or sedimentary rocks, which explains the predominance of basaltic and granitic rocks in the archaeological record. Quartzite is also common as it one of the metamorphic rocks that can withstand a high degree of thermal stress.

FCR cortical surfaces are often discolored toward pink, red, gray, and/or black hues (Latas 1992; Schalk and Meatte 1988; Taggart 1981). Many pieces retain a high percentage of cortex because of the way FCR fractures. Heating in a fire causes FCR to become more friable (particularly non- basaltic rocks) than unheated stones (House and Smith 1975; McParland 1977). A variety of FCR shapes have been described from experimental studies and archaeological sites, although a correlation between shapes and function is unclear.

FCR is generally recovered either as part of a feature, which is the physical remains of a cooking or heating facility, or in a secondary refuse context where they are no longer in their location of original use. Context is important for the understanding and interpreting FCR and associated subsistence activities at a site.

4.4.2 FCR Background and Previous Studies

The use of heated rocks for cooking, extending back at least 10,000 years, is well-documented ethnographically and archaeologically in North America (Thoms 2009). Cooking stones (FCR) and their associated features have valuable research potential, as is made clear by recent studies that illustrate their significance for interpreting site function and settlement and subsistence patterns (Jackson 1998; Thoms 2007, 2008a, 2009). Ethnographic research has shown that specific cooking and heating facilities were related to specific types of food resources and the seasonality of those resources. Thus, the identification of cooking facilities may indicate the type of food being processed and the seasonality of the site.

Thoms (2008a) notes three important qualities in cooking stones that explain their widespread use. First, the relative non-combustibility and high density of rocks (i.e., heavy per unit volume) enable them to capture and hold heat for longer periods of time than hot coals, allowing extended cooking of foods (particularly roots) to render them readily digestible and nutritious. Second, cooking stones hold heat generated by fire, thus reducing the amount of fuel needed to cook, which is important in areas where wood and other fuels are sparse. Third, cooking stones can be used to boil water and produce greater amounts of steam for longer than would be possible with hot coals alone. Compared to other cooking methods, boiling probably yields a greater proportion of potentially available calories/nutrients from a given piece of food (Wandsnider 1997), especially when the liquid medium is consumed. The heating benefits from rocks are also apparent in their widespread use for sweatbaths and keeping campsites and habitation shelters warm. Crumbled pieces of FCR were also used for temper in pottery.

Cooking-stone facilities and their archaeological byproducts, FCR features, have considerable functional and morphological variation, as they were used to cook a wide array of animal and plant foods (Driver and Massey 1957; Ellis 1997; Thoms 1989, 2007, 2008a; Wandsnider 1997). However, four primary cooking methods are consistently noted (Thoms 2008a): 1) baking in an earth oven with stone heating elements in closed pits and mounds where cook stones may be heated in situ (i.e., in the pit) or on an adjacent surface fire and, once heated, placed in the pit; 2) steaming with stone heating elements in closed pits and mounds where water is added, using cook stones heated in or outside the pit; 3) roasting (stone griddles) on open-air hearths built on an unprepared surface or in shallow pits using stone heating elements; and 4) boiling in open pits and non-ceramic vessels with stones heated 25 on nearby surface hearths/fires. In general, steam cooking takes place over several hours whereas baking often spans several days, but distinctions between hot-rock baking and steaming are often blurred. Hot-rock roasting refers to the use of cook-stone griddles in open-air hearths built on an unprepared surface or in shallow pits.

Jackson (1998:8-10; citing Driver and Massey 1957) provides additional details on the types of cooking facilities that were widespread across North America, which created much of the cooking- related FCR recovered from archaeological contexts:

As this and other ethnographic records indicate, a typical earth oven was usually between 1-3 m in diameter and 30-40 cm deep. The hole was filled with fuel (usually wood) and rocks, and then set ablaze. Once the fire was largely burned down, hot rocks were maneuvered into a flat heating element and then vegetal materials, food packages, more vegetal packing materials, and finally an earth seal were successively added. After sufficient time had passed, usually between twelve and 48 hours, the oven was opened and food was removed; this left a concave basin filled with FCR. Both plant and animal foods were cooked in earth ovens, however, plants were cooked more often (Driver and Massey 1957:233).

The second major type of cooking facility was the rock griddle. It was a type of hearth, used for short-duration cooking, that usually lasted no more than a few hours. It was akin to broiling over a fire or roasting on hot coals (cf. Driver and Massey 1957:233) because it used dry, open-air convection heat to cook food. As such, this cooking facility would have been used most often with animal foods and less often with plants (Driver and Massey 1957:233). In a generic rock griddle, rocks were placed directly in a fire to take on heat; they would release that heat after the fire died down. The fire was usually on a flat surface, enclosed with rocks, or in a shallow basin. A rock griddle was usually about 1 m in diameter. When the fire was mostly burned down, the hot-rocks were spread into a flat or slightly concave platform. Food was placed directly on the platform or placed on skewers directly over the rocks. Rocks would cool in place after the food had been removed, and would not be disturbed as a result of food removal.

Stone boiling, the third cooking facility, occurred when hot stones were immersed in a container of liquid (Driver and Massey 1957:229). It was a common cooking technique across North America, although it was seldom used among groups that had access to pottery.

Ethnographic accounts indicate that a variety of plants, large and small game, fish, and shellfish were cooked using hot-rock facilities. Plant foods, however, predominate in hot-rock cookery, especially those requiring inulin or fructan hydrolysis (Thoms 1989; Wandsnider 1997), with earth ovens being used most commonly for prolonged cooking of root foods (Thoms 2008b). High-lipid and collagen- rich meats that require substantial hydrolysis, which entails prolonged, high-temperature baking, are also well represented in hot-rock cookery (Wandsnider, 1997).

The distinguishing characteristics of primary cooking facilities types on archaeological sites are summarized in Table 4 (Thoms 2008a).

26

Table 4. Cooking Facilities and Expected Characteristics of FCR Features and Scatters (from Thoms 2008a). Hot-rock Expected archaeological Expected archaeological characteristics Cooking characteristics of non-feature of resulting FCR feature Facility FCR Basin-shaped pit, 1–3 m in dia. and 0.1–0.3 m deep, sometimes with rock lining and always with a lens of Scattered FCR in the immediate FCR (i.e., heating element) underlain by and vicinity of remains of earth ovens, Earth oven intermixed with thermally-altered (oxidized, carbon- representing discard and (baking), stained) sediments; FCR (small to large *), typically scavenging activities, and perhaps rocks heated carbon stained and mostly fragments, varies rocks used with oven-top fire; therein considerably in size, whole rocks often found along also other scattered camp debris, edges of heating elements; burned bone (possibly furniture rocks, and unused cook from fuel residue), flakes and tools expected therein stones as discard from routine clean-up activities Large to medium, presumably flattish, rock(s) on or Scattered FCR in the immediate Surface just below the occupation surface, underlain and vicinity of remains of surface oven encompassed by thermally-altered sediment ‘‘ovens” (i.e., open-air griddles) (roasting), (oxidized, perhaps some carbon stained); burned representing discard and rocks heated bone (possibly from fuel residue), flakes and tools scavenging activities; also other therein expected therein as discard from routine clean-up scattered camp debris, furniture activities rock and unused cook stones Basin-shaped pit (ca. 1 m dia. and 0.3 m deep) Scattered FCR in the immediate partially filled or lined with medium and large FCR Steaming vicinity of remains of steaming (typically not carbon stained), or occasionally a large pits; rocks pits, representing discard and flat rock, underlain by thermally-unaltered sediment; heated scavenging activities; also other nearby surface hearths (ca. 1 m dia.) where rocks nearby scattered camp debris, furniture, were heated, represented by ash, charcoal, oxidized and unused cook stones sediments, and a few pieces of FCR Bucket-like (i.e., near-vertical side walls) pits, 0.3– 0.45 m in dia. and 0.15–0.45 m deep, partially filled Comparatively dense, scattered Stone with small, possibly medium-sized, FCR, not FCR in the immediate vicinity of boiling in a typically carbon stained, underlain by thermally- remains of stone-boiling pits or pit; rocks unmodified sediment; nearby surface hearths where concentrations representing heated rocks were heated, represented by ash, charcoal, discard and scavenging activities; nearby oxidized sediments, and a few pieces of FCR, burned also other scattered camp debris, bone (possibly from fuel residue), burned flakes and furniture, and unused cook stones tools discarded in the fire pit Comparatively dense, scattered Surface hearths where rocks were heated, represented Stone FCR in the immediate stone by ash, charcoal, oxidized sediments, and FCR (not boiling in a boiling area, representing discard typically carbon stained); concentrations of discarded container; and scavenging activities; also small- and possibly medium-sized FCR, burned bone rocks heated other scattered camp debris, (possibly from fuel residue), burned flakes and tools, nearby furniture rock, and unused cook possibly discarded in fire pit stones Basin-shaped pit (ca. 1 m dia. and 0.3 m deep) with FCR lens, mostly medium-size large rocks, underlain Scattered FCR in the immediate Open-pit by thermally-unmodified sediment; nearby surface vicinity of remains of open pits, drying hearths (ca., 1 m dia.) where rocks were heated, representing discard and ovens, rocks represented by ash, oxidized sediments, and a few scavenging activities; also other heated pieces of FCR, burned bone (possibly from fuel scattered camp debris, furniture elsewhere residue), flakes and tools expected therein as discard rock, and unused cook stones from routine clean-up activities * Original rock sizes: large rocks, >25 cm in diameter; medium rocks, 10–25 cm in diameter; small rocks,

27

Thoms (2008a) notes that a better understanding of the relationship between cooking methods and cooking requirements allows for a better understanding of the nature of archaeological FCR features. By considering FCR feature characteristics, it should be possible to assess whether FCR represents stone-boiling or oven-baking, estimate the magnitude of activities, suggest what foods may have been cooked there, and fine-tune the search for confirming evidence.

Jackson (1998:45) summarizes the types of information that can be gleaned from collecting basic FCR data:

FCR weights and counts give rough estimates of cooking methods (Taggart 1981:149). In general, large heating elements (i.e. earth ovens) required kilograms of rock to sustain high temperatures for days. While there is considerable overlap between large rock griddles and small earth ovens, rock griddles generally used fewer rocks because they did not need to remain hot for as long as earth ovens. Still fewer rocks were needed for stone boiling in generally small, pot-sized containers.

Rock size is also related to feature function. Large rocks (larger than 10-cm diameter) were preferred in earth ovens and rock griddles (Schalk and Meatte 1988:8.9; Taggart 1981:148- 149) because they stored heat for long periods of time. Small rocks (less than 10-cm diameter) were not preferred in earth ovens because they had a higher ratio of surface area to mass, which caused them to lose heat more rapidly than large rocks (Schalk and Meatte 1988:8.9). This is a bad quality where extended cooking is required. Large rocks should have been preferred for structure heating, be it a sweatlodge or habitation, because of the same heat retention quality. Small rocks were preferred for stone boiling because of better resistance to thermal shock and because they were easier to handle (Schalk and Meatte 1988:8.8; Taggart 1981:148-149).

Ethnographic accounts and archaeological excavations attest to the differential use (preference) of smaller rocks in stone-boiling features and larger rocks in earth ovens. Small rocks <10 cm diameter are good for stone boiling because they have a high surface-to-mass ratio which allows them to store and release heat energy quickly; they are also easy to handle.

Raw material is a critical factor. Certain rock types can be good for certain cooking methods and poor for others (McDowell-Loudan 1983:26; Zurel 1979:5). For example, sandstone reacts well in a rock griddle because it is generally coarse-grained and porous, which makes it elastic and able to deform in response to heating and cooling. It is not very good for stone boiling because it loses individual grains and adds grit to water (Brink et al. 1986:290-292; Jackson 1997); it also absorbs a lot of water because of its high porosity, which requires longer drying periods than fine-grained rock types (Brink et al. 1986:296). Fine-grained rocks were generally preferred for boiling, while coarse grained rocks were preferred for griddle roasting and earth-oven baking. However, some materials like quartzite were preferred whenever available. Homogeneity in mineralogy, grain size, and grain shape, as well as a strong bond make quartzite an all-purpose rock.

Size grade analysis can be used to address these questions. Every time a cooking/heating facility is used, some of the rocks will fracture and/or crack. As the number of times the facility is used increases, the resultant rock sizes become smaller as rocks continue to fracture; the number of fractured rocks increases at the same time. Therefore, size grade analysis can be used to discriminate this thermal weathering process. A relatively small number of large FCR pieces would indicate relatively less use of the rocks than a similar feature containing relatively more FCR that are smaller in size.

28

New lines of research are extending the range of information that can be recovered from FCR through more complex techniques such as analyzing fatty-acid residues to identify remnants of animal fat on FCR, paleo-magnetic testing to reveal whether stones were moved after heating, AMS dating of FCR samples, and examining starch grains, phytoliths, and calcium oxalate crystals on FCR and in features to provide information about plants that were cooked using FCR (Thoms 2008a and 2009).

4.4.3 FCR Analytical Methods

Several criteria were established to provide a consistent method of identifying FCR. The lack of naturally occurring cobble-size rocks within the project area aided the identification of FCR. Data collected for FCR included count, weight, and size grade. In order for a rock to be classified as FCR, it had to meet at least one of the following criteria:

1) The rock is associated with a cooking feature such as fire hearth or cooking pit. Such features may have carbon-stained (blackish) or oxidized (reddish) soil and may be other associated with other materials such as charcoal, ash, and thermally-altered fauna.

2) The rock has distinctive shapes that have been observed at archaeological sites and in ethnographic and experimental studies, such as angular blocky fragments, crenulated or jagged edges, spalls (potlids), or a variety of intermediary shapes. FCR cobbles contain the negative impression where an angular or spall piece detached.

3) The rock’s fracture surfaces are fresh, unweathered, and have fairly sharp edges. The rock also lacks the characteristics of cores and lithic debris from stone knapping, such as bulbs of force, ripple marks, hinge or step terminations, and crushing.

4) The rock is unfractured and whole but has other distinctive thermal stress features such as crazing (surface cracks) or a friable and crumbly surface, especially with granitic rocks and sandstone.

5) Rocks have a reddish, pinkish, or blackish discoloration, particularly the cortical surface.

6) The rock’s grain size is generally too coarse for flaking. Common rock types include granite, basalt and quartzite that originally occur in the local area as rounded cobbles with their source in glacial or outwash deposits.

Some experimental studies appear to have demonstrated that the shape of individual pieces of FCR (spall or angular) results from specific rates and methods of heating and cooling (Homsey 2009: House and Smith 1975; McDowell-Loudan 1983; McParland 1977; Wendt 1988; Zurel 1979). Angular pieces were thought to result from FCR being quickly cooled by immersion in water for stone boiling, while spalls were thought to result from slower cooling around a fire hearth. However, the results of these studies have not produced consistent results. Jackson’s (1998) experimental study suggests that FCR shapes are not related to specific rates and methods of cooling but to rock size and duration of heating. Similar rock shapes can be produced by various types of cooking facilities.

Jackson (1998) conducted microscopic analysis of rock thin-sections subjected to various cooking facilities to examine the mechanical aspects of thermal weathering of rock. The results show that thermal weathering was highest for all rock types in the earth oven and rock griddle plates, while it was lowest in the stone boil and sweatbath plates. The thermal weathering variation is attributed to the length of heat exposure, rather than the rate of cooling. His results indicate that there is valuable research potential for the microscopic study of FCR for understanding cooking facilities and subsistence. In conclusion, additional microscopic and experimental studies need to be conducted before more reliable interpretations can be made. 29

4.4.4 FCR Morphology

Observations of FCR from archaeological sites and experimental studies led to the delineation of three basic FCR shape types (Jackson 1997, 1998; McParland 1977; Schalk and Meatte 1988; Thoms 1986: Zurel 1979, 1982), which are defined as follows: 1) spall types are expansion-fractures that, according to Jackson (1998), “occur because of an internal thermal gradient, where the exterior of a rock becomes hotter and expands more quickly than the interior. When stress becomes too high, a rock releases it by sloughing off curvilinear spalls or convex potlid”; 2) angular types are blocky contraction-fractures that, according to Jackson (1998), “occur because of tension stress where the exterior of a rock cools rapidly and causes cracks to form perpendicular to the surface and at evenly spaced intervals”; and 3) spall/angular types include FCR that is intermediary between the spall and angular types (Jackson 1997; Thoms 1986; Zurel 1979), which represent opposite ends of the typology continuum (McParland 1976, 1977; Thoms 1986).

Despite evidence that cooking methods (rate/methods of heating and cooling) cannot be inferred directly from FCR shapes (Jackson 1998), these shapes are recorded for this analysis because they provide a fair description of the basic shapes and properties of the FCR, are currently in use in the archaeological community, and may someday prove to have more interpretive value. In addition, the FCR analysis for this project also includes other descriptive types that were established to encompass the variety of FCR shapes and conditions that were recovered at the sites. These FCR types are summarized in Table 5.

Table 5. FCR Type Descriptions. FCR Type Description Expansion-fracture, has straight or curvilinear profile following the natural shape of cobble cortical surface (like a section of orange peel), relatively thin in cross-section Spall in relation to the width and length, also includes interior non-cortical pieces that have thin cross-sections, fracture plains are relatively large, smooth, and lack complexity Thick, blocky, and angular pieces with fractures that are generally perpendicular to the exterior surface, sometimes with distinctive serrated or crenulated edges at the Angular exterior surface. The length, width, thickness ratio more approximately equal compared to the relative thinness of spalls. Spall/Angular Intermediary pieces between the Spall and Angular types. Crumbs are small pieces, typically less than 1/2” (SG2) that do not fit other Crumb categories Cobble These are whole cobbles that have cortical discoloration and/or cracks on the surface (Nonfriable) but do not have spall or angular fractures. These are whole cobbles that have a crumbly surface or portion of the surface, which Cobble (Friable) is most common on granitic or sandstone FCR. They do not have spalls or angular fractures. Cobble with Spall These are mostly whole cobbles that have one or more spall fractures. Cobble with These are mostly whole cobbles that have one or more angular fractures. Angular Friable Rounded These are round-shaped FCR with a crumbly surface, which is most common on Piece granitic FCR, classified as crumb if smaller than 1/2” (SG2) Split Cobble Cobble that has split Indeterminate FCR that do not fit any other categories

30

4.5 Faunal Analysis Methods by Steven Kuehn

The faunal analysis was conducted by zooarchaeologist Steven Kuehn. The results of the faunal analysis are presented below. After separation by provenience, the following information was recorded for each specimen larger than 2 mm: element, side of the body (when applicable), section or portion of the element, weight in grams, and taxonomic classification. Relative age (e.g., adult or juvenile) was recorded when it could be reliably determined, based on epiphyseal fusion, tooth eruption, and occlusal wear. Refitting of bone fragments was restricted to specimens recovered from within the same feature, XU, or shovel test. Each specimen was examined for exposure to heat in the form of burned, charred, and calcined bone. Evidence of butchering and cultural modification was recorded when observed.

Because of specimen fragmentation, otherwise unidentifiable pieces of mammal bone are categorized as extra-large, large-sized, medium-sized, or small-sized based on the relative size and thickness of each specimen, and their inferred approximate live weight. The category of extra-large mammal is reserved for indeterminate bison, cattle, elk, and moose remains that have an approximate live weight of greater than 500 lbs (227 kg). The live weight of large-sized mammals is considered to be greater than 50 lbs (23 kg) (bear, deer, wolf, and larger, including extra-large mammal), 11 to 50 lbs (5 to 23 kg) for medium-sized mammals (coyote to adult raccoon size), and less than 10 lbs for small-sized mammals (rabbit, gopher, and vole size). When it was not possible to reliably categorize a specimen based on its size, it is listed simply as mammal of indeterminate size.

The quantitative measure of the number of identified specimens per taxon (NISP) is used throughout this report unless otherwise noted. Minimum number of individuals per taxon (MNI) determinations are based on comparison of repeating or multiple elements, relative age, and overall size, and calculated for the assemblage as a whole. In general, MNI estimates are made only for specimens minimally identifiable to the genus and species level (following Reitz and Wing 1999:198-199). An osteological comparative collection facilitated specimen identification. Habitat information for mammals was taken from Jones and Birney (1988) and Jackson (1961); for birds from Kaufman (1996); for reptiles and amphibians from Phillips et al. (1999); and for fish from Becker (1983) and Smith (1979).

After separation by provenience, the following information was recorded for each specimen: element, side of the body (when applicable), section or portion of the element, weight in grams, and taxonomic classification. Relative age (e.g., adult or juvenile) was recorded when it could be reliably determined, based on epiphyseal fusion, tooth eruption, and occlusal wear. Refitting of bone fragments was restricted to specimens recovered from within the same feature or XU. Each specimen was examined for exposure to heat in the form of burned, charred, and calcined bone. Evidence of butchering and cultural modification was recorded when observed.

4.6 Ceramic Analysis Methods

Data recorded for ceramics included vessel portion (morphology), temper, surface treatment, decoration (if present), and presence/absence of charred residue. The small size and fragmentary condition of most sherds made it difficult to determine vessel portion, and these sherds were classified as body sherds. Thickness was measured for sherds that retained intact internal and external surfaces. The thickness was entered into the catalog “Notes” column. The thickness measurement was useful for establishing the relative age and affiliation of the sherd in the absence of decoration or other diagnostic attributes. For example, Plains Village and Late Woodland wares are thin-walled (< 5.5mm). Temper was determined using a 10x hand lens. Smooth surface treatment is defined by the absence of cord impressions. Some smooth sherds were polished or burnished, and 31 they were easily identified by their shiny, reflective surface compared to the other smooth sherds. Cord-marked surface treatment includes twisted cord impressions (generally finely spaced) that resulted from a cord-wrapped paddle, woven cordage (fabric), or impressions from having been formed in a woven bag. Some cord-marked sherds were partially smoothed, and this condition was included in the catalog “Notes” column. Black-colored, charred organic residue was observed on some sherds (primarily the interior), and the presence of such residue was recorded in the catalog.

32

5. LITERATURE SEARCH

5.1 Archival and Background Research for Previous Archaeology Sites

Archival and background research was conducted to determine whether any previously identified archaeological sites or potential historic sites are located within one mile of the project area. FCRS staff conducted an initial review of sites located near the project area prior to fieldwork in May 2016 at the MnSHPO. Site inventory files, USGS 7.5’ quadrangle site location maps, and research reports were reviewed to provide information on previously recorded archaeological sites and previous investigations within one mile of the project area. Mr. Tom Cinadr, Survey and Information Management Coordinator at MnSHPO, also conducted a search of the site file database and provided a list of sites within one mile of the project area.

There are no recorded archaeological sites within one mile of the project, but there are two previously recorded sites leads from historic documents. These sites include: 1) 21PLy “Old Nielsville” (reported location of original townsite) in T147N R49W NE¼ and NW¼ Section 24, which is at least 0.4 mile north of the project area; and 2) 21NRl “Georgetown Road” (historic trail or road) T147 R48W Section 31, which is at least one mile south of the project area.

5.2 Overview of Archaeological Investigations in the Red River Valley

Archaeological investigations have been conducted in the Red River Valley for more than a century. Theodore Lewis of the Northwestern Archaeological Survey surveyed mound groups and earthworks in the late 1800s (Winchell 1911). Between the 1930s and 1950s, there was an increased interest in the prehistory of the region fueled by the excavation of several mounds by A. E. Jenks and Lloyd Wilford (Jenks 1932; Wilford 1955, 1970). By the 1960s more attention was focused on habitation sites. Elden Johnson’s work in the 1960s revealed the use of the Lake Agassiz beaches in Minnesota from the Paleoindian through the Late Woodland period (Johnson 1964). In the 1970s and 1980s, intensive surveys were undertaken for the Minnesota State Wide Archaeological Survey (Michlovic ca. 1978, 1982).

Since the 1970s, numerous compliance surveys have been undertaken to fulfill the requirements of federal laws. Several corridor surveys in the region were conducted for natural gas pipeline projects. Extensive surveys and site testing have also been conducted in the last five years for the Corps’ flood control project at East Grand Forks, Minnesota and Grand Forks, North Dakota (Florin 2001a, 2001b; Florin et al. 1998, 2001; Florin and Wergin 2002; Ketz and Dolence 1997; Larson and Penny 2000a, 2000b; Lyon et al. 2003a, 2003b; Madigan et al. 2002; Nienow et al. 2003). Surveys conducted on the North Dakota side of the Red River have been summarized in Picha and Gregg (1991a, 1991b) and Larson et al. (1994). In spite of the numerous archaeological investigations in the region, only a small number of sites have been excavated and specific data on the prehistory of the area is relatively sparse.

5.3 Previous Investigations in the Red River Valley

Numerous surveys in the Red River Valley have documented the importance of the region during prehistory. The area has been occupied from the Late Paleoindian through the Late Prehistoric period. Site types include find spots, habitation areas, animal butchering sites, and burial mounds. The survey results indicate that sites are commonly located in riverine settings and may be present in deeply buried contexts. Archaeological investigations have been primarily limited to site identification, and more detailed information on the prehistoric occupants is lacking. The results of the investigations emphasize the lack of knowledge regarding the region’s complex prehistory.

33

5.3.1 Minnesota

The Minnesota State Wide Archaeological Survey (MNSAS) conducted intensive surveys in the Red River Valley south of the project area in Norman, Clay, and Wilkin counties (Michlovic ca. 1978, 1982, 1984). The goal of the surveys was primarily to obtain data on site distributions. Because these surveys were conducted on landscapes similar to those of the project area, the results provide a useful model of prehistoric land-use patterns in the project area.

The MNSAS survey in Clay County, Minnesota consisted primarily of random sampling with additional survey conducted in high potential areas and areas reported by landowners to contain sites (Michlovic ca. 1978). A total of 48 sites were identified. Twenty-one of these sites were identified along tributaries of the Red River. Nine sites were found on the Lake Agassiz plain away from water sources. The other sites were located on glacial beaches, the intersections of streams and glacial beaches, and in areas outside the Lake Agassiz plain.

The survey in Norman County, Minnesota was conducted as a follow up to the MNSAS survey in Clay County, Minnesota (Michlovic 1982). The survey covered a 24-mile stretch along the Red River. A total of 41 sites were identified. The field methods employed in these surveys included 25- meter and 50-meter interval transects. Shovel testing was limited to a depth of 30 to 40 cmbs. It is likely that more sites would have been identified if closer interval transects were used and if deeper testing was conducted. Cultural materials recovered during the surveys provide a record of occupation in the area by Late Paleoindian, Archaic, Transitional Middle to Late Woodland, Late Woodland, and undefined Plains Village-Woodland groups.

A survey along the Red River from Moorhead in Clay County to Breckenridge in Wilkin County, Minnesota was conducted for the Minnesota Historical Society in 1984. The survey consisted of a 25-percent random sampling of the river terraces supplemented with test excavations at selected locations (Michlovic 1984). The total survey area was 12 miles. Each field was surveyed at 25-meter intervals. Two previously recorded sites (21WL01 and 21WL02) were relocated and 26 new sites were identified. Archaeological recommendations resulting from the survey state that archaeological testing should be conducted in alluvial settings for buried deposits along the Red River, even if surface artifacts are not present.

It is also worth noting that two deeply buried sites (more than one meter below the surface) have been identified on the Minnesota side of the Red River in Norman County, approximately 40 miles south of the project area. These sites are the Canning site (21NR9) and Mooney site (21NR29) (Michlovic 1986, 1987a). They have been assigned to the Archaic period based on radiocarbon dates and artifact styles. Other deeply buried sites have been identified near East Grand Forks, Minnesota along the Red River and Red Lake River (Florin et al. 2001; Florin and Wergin 2002; Lyon et al. 2003b).

5.3.2 North Dakota

Surveys conducted on the North Dakota side of the Red River Valley in the Northern Red River Study Unit (NRRSU) have been summarized in Picha et al. (2016). As of 2015, there were 270 archeological sites and 330 archeological site leads and isolated finds recorded in the state computerized site data file for this unit. The sites have the following temporal/cultural affiliations: Paleoindian (n=4), Archaic (n=21), Woodland (n=40), Plains Village (n=3), Late Prehistoric (n=38), and Historic (n=10), and unknown (n=488). Landform location data is available for 348 sites. Most of the surveys were conducted on tributaries of the Red River and on upland plains west of the Red River Valley, and these are the locations where most of the sites are located. There have not been any extensive surveys along the Red River.

34

5.4 Mn/Model Study of the Red River Prairie Subsection

Mn/Model is a statewide GIS-based predictive model for pre-1837 archaeological site locations. The project area is located Mn/Model’s Red River Prairie subsection (Hudak et al. 2002, Section 8.23). This area consists primarily of a level glacial lake plain that drains north via the Red River and its tributaries. Minor landforms include a till plain, beach ridges, sand dunes, and water reworked till. Presettlement vegetation was prairie and wet prairie. Narrow bands of woods were found along the Red River and its major tributaries. Areas of high and medium site potential are located along the Red River and its tributaries.

5.5 Air Imagery and Plat Map Review

Several historic maps were examined to aid in identifying potential historic archaeological resources within the project area. The earliest maps examined were the General Land Office (GLO) survey maps of 1872 and 1873, which were available online (http://www.mngeo.state. mn.us/glo/). Copies of historic plat maps were also examined for 1874, 1902, 1916, and 1930 (Andreas 1874; North West Publishing 1902; Hixson 1916, 1930). The 1874, 1902, and 1930 maps only record public structures and not private dwellings.

The 1873 GLO map depicts a homestead and garden plot with the name Nels Oleson on the line between Sections 24 and 25, which is near the location of the current private residence adjacent to the oxbow and .25 mile east of the western edge of these sections. This location is about 90 meters south of the current survey area. The 1873 GLO map also depicts that the area from .25 mile east of the Section 23 and 24 line to the Red River was wooded. The 1916 map depicts a few farmsteads along CSAH 1 where existing farmsteads are located and also one on the west side of the oxbox that is no longer extant, but no historic artifacts were recovered. None of the other maps depict potential cultural resources within or adjacent to the project area.

Air photos from 1930 and 1954 (Figures 3 and 4) were examined to review land use changes and changing landscape conditions. These air images were obtained online from the Borchert Map Library at the University of Minnesota (http://map.lib.umn.edu/mhapo). In general, historic land use is similar to today. The current alignment of CSAH 1 is approximately the same as that depicted on the 1954 air photo. However, on the 1930 photo the CSAH 1 alignment was much closer to the oxbow and had a different alignment across the Red River point bar in Sections 23 and 26. On the 1954 air photo there appears to be drain tiling of the fields on the north and south sides of CSAH 1 in SE¼ of Section 23 and NE¼ of Section 26, as indicated by a series of unique linear lines running through the fields in this one specific area.

35

Current road alignment shown as yellow dashed line

0 .25 mile N

Figure 3. Location of Project Area on 1930 Air Photo. * note road alignment is different than current alignment

Area of probable drain tiling – white lines

0 .25 mile N

Figure 4. Location of Project Area on 1954 Air Photo. * note possible drain tiling north of oxbow and road alignment is the same as current alignment

36

6. CULTURE HISTORY by James Lindbeck

The following culture history of the precontact period in the project area is derived primarily from Archaeology of Minnesota: Prehistory of the Upper Mississippi Region (Gibbon 2012); Minnesota Archaeology: The First 13,000 Years (Gibbon and Anfinson 2008); the Minnesota Statewide Multiple Property Documentation Form for the Woodland Tradition (Arzigian 2008); and Outline of Historic Contexts for the Prehistoric Period (ca. 12,000 B.P. - A.D. 1700) (Dobbs 1988).

Background on archaeological investigations and culture history specific to the northern Red River Valley is from The Northern Red River Study Unit in the State Historical Society of North Dakota’s Comprehensive Plan for Historic Preservation (Picha et al. 2016), Phase II Archaeological Investigations at the Mooney Site (21NR29) in Norman County, Minnesota (Johnson et al. 1995), and The Red River Valley in the Prehistory of the Northern Plains (Michlovic 1983). The Mooney site is on the Red River about ten miles south of the current project area. The discussion follows the organization of cultural periods used by Gibbon (2012).

The following culture history dates for the Paleoindian and Archaic periods are presented in two formats: 1) by their conventional radiocarbon age (uncalibrated) and 2) as calibrated to actual calendar years. The conventional radiocarbon age (measured radiocarbon age corrected for isotopic fractionation) is presented in the format of “RCYBP” (radiocarbon years before present; with “present” by convention being AD 1950). The use of “RCYBP” dates allows for the consistent comparison of dates from sites in previous reports, as this format has been the standard. Dates calibrated to actual calendar years use the convention “cal BP” (for example cal. 8000 BP) to distinguish them from uncalibrated dates (RCYBP). Dating formats are discussed more in Section 1.6.

The project area is located in northwestern Minnesota in Archaeological Region 6n: Northern Red River Valley and is directly adjacent to the Northern Red River Study Unit (NRRSU), as defined by the State Historical Society of North Dakota (Picha et al., 2016). The Red River of the North, which flows northward along a sinuous course that has meandered over time, is situated in the middle of the broad and flat Glacial Lake Agassiz plain. The main physiographic features of the project area on the lake plain include glacial beach lines, delta deposits, and glacial meltwater trenches, although many of these subtle landscape features have been flattened out and obscured by Euro-American era farming and settlement activities.

Picha et al. (2016:9.9) describe the pre-settlement vegetation in the northern Red River Valley as having been a mixture of tall grass and mixed grass prairie. Floodplain gallery forests would have been present in the bottomlands along the Red River, hosting tree species such as elm, ash, box elder, and bur oak along with an understory of shrubs and forbs. This habitat would have been an important source of nuts, berries, and tubers and would have provided forage for a variety of large and small fauna, including elk, deer, antelope, moose, and caribou. Bison covered the prairie grasslands and other animals attracted to the riparian habitat would have included muskrat, beaver, rabbit, lynx, wolf, fox, ermine, mink, marten, fisher, wolverine, otter, skunk, badger, and raccoon. Several species of ducks and geese along with eagles, hawks, and other raptors were also present in the region. Lithic resources are scarce on the interior portion of the Red River Valley, with even small gravels being rare on glacial lake plain. Beach ridges and glacial deposits near the margins of the valley provided the nearest sources of tool stone. On the interior portion of the lake plain along the Red River, rocks needed for stone tools, cooking, heating, and sweat bathing needed to be transported from beach zones or glacial areas near the margins of the glacial lake plain (Bakken 2011; Picha et al. 2016:9.10).

37

The culture history of the region is complex and reflects its location near the prairie-forest border, with diagnostic artifact types suggesting the presence of several distinct cultural traditions in the region that include Plains and Woodland groups along with archaeological complexes that derived from the interaction of these groups (Michlovic 1990; Picha and Gregg 1991a; Syms 1977, 1985). The complexity is due in part to the shifting of ecological zones during the Holocene in response to climate changes, and the changes in lifeways that occurred in response to these shifts.

6.1 Paleoindian Period (11,300 to 8500 RCYBP / cal 13,200 to 9500 BP)

The Paleoindian period is divided into Early (13,200 to 12,500 BP) and Late (12,500 to 9500 BP) periods, as defined by the use of Fluted (Early Period) or Plano (Late Period) projectile points (spear points) for hunting and also possibly butchering. During the Early Paleoindian period, artifact typologies in Minnesota suggest that the culture was mostly related to the eastern Midwest. During the Late Paleoindian period, the cultural affiliation is clearly more related to the Plains, except in the Mississippi Valley region of southeastern Minnesota.

The Paleoindian period was a time of rapid environmental change as the glaciers retreated from Minnesota (Wright 1974). Substantial changes in vegetation, wildlife, waterways, and the landscape occurred as a result of the ameliorating climate, and Paleoindian lifeways reflect adaptations to these rapidly changing landscapes. The first Paleoindian peoples in the northwestern Minnesota encountered a subarctic environment with no direct parallel in the modern world. The project area would have been covered by Glacial Lake Agassiz during the Early Paleoindian Period, but by the Late Paleoindian, after the draining of the lake, the landscape would have been a tundra-like spruce forest mosaic with prairie openings. As aridity increased, a prairie grassland habitat came to cover most of the region by the end of the Paleoindian Period. It is not known what animals lived in the area at this time, but it can be assumed that mammoths, giant bison, and other now-extinct megafauna were present. Fish would have been present in the newly-formed lakes and rivers soon after the establishment of open water (e.g. Pielou 1991), and plants would have become established on the ice- free landscape.

Only one burial of this period is known, the Browns Valley site (21TR5) in the west-central part of the state. The known sites appear oriented toward current bodies of water, but these locations are also areas that have had a greater amount of archaeological survey. The locations of known sites therefore do not necessarily represent the actual settlement patterns. It is not clear whether the paucity of sites demonstrates that there was a small Paleoindian population in Minnesota, or whether the population was more numerous but the sites have not been identified because they have been destroyed, are deeply-buried, or lack diagnostic artifacts. It is likely that some of the lithic scatter sites found throughout the state belong to this period, but without the recovery of diagnostic artifacts or datable material, it is not possible to determine the cultural affiliation of these sites. Research in other parts of the country, where Paleoindian sites are more common, suggests that the margins of lakes and swamps were preferred habitation locations, and these landscapes were prevalent in the late-glacial and early Holocene periods of central Minnesota.

Picha et al. (2016) identify the known or anticipated Paleoindian occupations in the NRRSU as comprising Clovis, Folsom, and Plano (including Caribou Lake) complexes. While no Folsom components have yet been identified and Clovis has been confirmed only at site at 32PB25 west of the project area along the Pembina River, Plano complex artifacts are much more common and have been recovered from several sites. The authors note that artifacts from the Plano complex (also known as the Reservoir Lakes Phase in Minnesota) are not well-understood and may not always be recognized. All of these sites are located on landforms adjacent to Holocene age river channels and terraces. Many more Paleoindian sites, including Folsom, have been identified north of the project area in southern Manitoba, in areas that would have been at the western edge of Glacial Lake Agassiz. 38

Johnson et al. (1995:16-20) summarize the Paleoindian period in the northern Red River Valley for their background discussion of 21NR29, which is located south of the project area on a terrace of the Red River.

A thorough organization of the Paleoindian period in the northern Red River Valley has been completed by Pettipas and Buchner (1983) and Buchner and Pettipas (1990), based on materials from Manitoba, to the north of the project area. In their discussion, the period is divided into Fluted Point and Plano complexes rather than Early and Late. Fluted (early) occupations are absent but Plano (late) occupations are fairly well-represented, and the evidence suggests that these populations moved into the Lake Agassiz region from the south and west as the first human adaptation to the Lake Agassiz area. Point types include the Agate Basin, Hell Gap, Lusk Scottsbluff, Eden, and Alberta. They also describe the Caribou Lake phase of the Plano culture on the northeastern side of the Lake Agassiz basin, closest to the project area, as a Plano adaptation to the boreal forest.

Only a small number of Paleoindian sites have been documented in the Red River region, and none is in the immediate vicinity of the project area. The Brown's Valley site (21TR5), mentioned earlier, was found by a private landowner during gravel excavation. An adult male skeleton was found along with several lanceolate spear points made of Knife River Flint, which were classified as “Brown’s Valley Points”, but are similar to the Agate Basin type often found at bison kill sites. Radiocarbon dates of 8790 ± 100 BP and 9049 ± 82 BP for bones at the site make them among the earliest documented human remains in North America.

Site 21RO11 (Greenbush), on an eastern beach deposit of glacial Lake Agassiz near the town of Greenbush in Roseau County, included the basal portion of a Scottsbluff point in a heavy scatter of Swan River chert at what was likely a lithic raw material quarry. Johnson et al. (1995:19) note that people in the Red River Valley used Swan River chert for stone tools throughout the precontact period because of its availability in local cobbles, despite the relative difficulty of working the material. Two additional Plano sites (DhLb-1 and EcKx4 [Sinnock]) on eastern Lake Agassiz beaches have been recorded in southeast Manitoba. Little is known about these sites except that they contain Plano style artifacts, including Agate Basin types at Sinnock. A few Paleoindian points are included in private collections from the vicinity of the Red River Valley, including a Clovis point from a private collection at site 32PB25 in Pembina County, but contexts for these artifacts is lacking.

6.1.1 Early Paleoindian (11,300 to 10,500 RCYBP / 13,200 to 12,400 cal BP)

As noted above, artifacts from the Early Paleoindian Period are not expected within the project area, as the region was covered by glacial Lake Agassiz for most of this period. To the south, beyond the margin of the lake, early Paleoindian fluted point types such as Clovis, Folsom, and Gainey of the Early Paleoindian period are very rare, and little archaeological evidence of Early Paleoindian people has been documented thus far. Isolated finds, primarily recovered from the surface of agricultural fields, have been recorded at scattered locations across Minnesota (Anfinson 1997:28-30; Buhta et al. 2011; Higginbottom 1996). These isolated finds are in themselves important contributions to the archaeology of the Early Paleoindians, but it is unfortunate that no other site data are available.

Early Paleoindian people are traditionally thought to have been nomadic big-game hunters, an interpretation derived from the dramatic and defining finds of lanceolate points at megafauna kill sites in the American southwest. These now-famous discoveries at places such as Blackwater Draw and Folsom in New Mexico initially established the antiquity of the Paleoindian tradition and the association of Clovis and Folsom points with mammoths and other extinct megafauna. Mason (1981:97) points out, however, that, “as eastern fluted point sites were found and investigated, and dramatic kill sites eluded discovery… enthusiasm for this idea waned. Because most Paleo-Indian sites east of the Mississippi are unaccompanied by preserved bones, it is now a popular notion that 39 big-game hunting was a western specialization not indulged in by the easterners. But just as it is difficult to argue one way in the absence of evidence, so is it difficult to argue the other way.”

While paleontological finds of extinct megafauna have been made in Minnesota, only the Itasca Bison Kill site (Shay 1971), which contained the extinct bison type Bison occidentalis, also contained cultural materials. The closest known megafauna kill (or possibly scavenging) sites are in Wisconsin, including several on beach ridges of Glacial Lake Michigan. The Boaz Mammoth site in southwestern Wisconsin is the nearest site. The site, which was discovered in the late nineteenth century, contains the remains of a mammoth in apparent association with a Hixton orthoquartzite fluted point (e.g., Overstreet 1993, 1996; Mason 1981, 1997). Anfinson (1997) suggests that Early Paleoindians in the Prairie Lake Region, to the south of the project area, relied on a much wider variety of resources in their boreal environment, such as smaller animals, fish, and vegetal foods, than did the Paleoindians of the southwestern United States.

6.1.2 Late Paleoindian (10,500 to 8500 RCYBP / 12,400 to 9500 cal BP)

The transition from the Early Paleoindian to the Late Paleoindian period is indicated by the appearance of some groundstone tools, such as the adze, and by a variety of large, finely-crafted stemmed and lanceolate projectile point types that lack the distinctive fluted points of the early period. Some of the Late Paleoindian points in Minnesota and the Midwest are smaller and less- finely crafted than those from the Plains, which is perhaps a result of raw material quality and cultural changes through time (Florin 1996). Many of the points from Minnesota are extensively resharpened and reworked so that their original condition is no longer apparent. Another unique feature on points from the Midwest is the presence of basal ears on some specimens, particularly the stemmed forms. Gibbon (2012:73) suggests the Late Paleoindian may have persisted in northern Minnesota until 8000 to 7000 BP and similar late dates have been suggested for northern Wisconsin (Mason 1997).

Two projectile point bases that resemble Agate Basin and an Eden stemmed type were recovered at site 21CR156 in the Minnesota River valley bottom near Shakopee, Minnesota (Florin et. al 2013). A radiocarbon date from calcined bone associated with these points was ca. 7000 RCYBP (cal. 7900 BP), indicating that the Late Paleoindian period overlaps Archaic period, as Gibbon (2012) has suggested. Late Paleoindian points have been recovered in association with Archaic points at several sites in Wisconsin and adjacent areas in the Great Lakes region, confirming they are contemporaneous (Mason 1997; Pleger and Stoltman 2009). Hixton quartzite was used as a raw material throughout the eastern Midwest at this time.

Faunal assemblages from five Late Paleoindian sites in Wisconsin contain a variety of terrestrial and aquatic animal resources, including deer, bear, beaver, muskrat, porcupine, birds, turtle, and fish, indicating a generalized foraging subsistence base (Kuehn 2010). This data contrasts with the outdated concept of Paleoindians being primarily hunters of a few select species of large game animals such as bison, moose, and caribou. The prevalence of wetland and aquatic animals is particularly noteworthy. Faunal material recovered from the Late Paleoindian component at site 21CR156, near Shakopee, conforms to this generalized foraging pattern and the reliance on wetland and aquatic resources.

Numerous Late Paleoindian points have been recorded from private collections and also identified during archaeological investigations across the state (Florin 1996). The point types from Minnesota resemble the stemmed and lanceolate types defined from type sites on the Plains. Late Paleoindian points found in northwestern Minnesota have not typically been recovered from the Glacial Lake Agassiz plain, but from beach ridges and glaciated areas. While no points are reported from Polk County, several have been reported from Kittson, Marchall, and Roseau counties.

40

An important Late Paleoindian site is Bradbury Brook (21ML42) located in Mille Lacs County. The site is a siltstone lithic procurement and initial reduction site associated with the Alberta Complex (Malik and Bakken 1993, 1999). A Phase III data recovery was conducted at the site. One feature was identified, which produced the base of an Alberta point and an associated radiocarbon date of approximately 10,500 BP. It is the oldest radiometrically-dated site in Minnesota, and provides a unique perspective on the Late Paleoindian period in central Minnesota.

The East Terrace site (21BN6) on the Mississippi River near St. Cloud is described as a Plano site that represents an intermittently-occupied location (BRW, Inc. 1994). Diagnostic points recovered included Hell Gap, Alberta, and Scottsbluff, which were extensively reworked.

The Reservoir Lakes Complex of northeastern Minnesota consists of one of the best professionally documented collections in the state. This collection of artifacts recovered from shorelines along a chain of reservoir lakes near Duluth contains a variety of stemmed and lanceolate points (Harrison et al. 1995; Steinbring 1974). Some of these points have basal ears, suggesting an eastern influence. A variety of stone tools also occur, including choppers, bifaces, crescentric blades, adzes, long heavy picks, retouched flakes, scrapers, drills, and asymmetrical knives. The sites are located along lake shores that have been eroded by fluctuating water levels. Because of the deflated nature of the sites, it is not possible to confidently characterize the site components, and some of the assemblages are mixed with later Archaic components.

The Cherokee Sewer site (13CK405) in northwestern Iowa provides some of the best information on the Late Paleodindian and Early to Middle Archaic period in the northeastern plains and adjacent prairie region. The site contained three distinct cultural horizons dating from 8400 to 6400 RCYBP. The earliest component contained points resembling the Hell Gap type that were recovered with bison and other animal bone.

6.2 Plains Archaic Period (7700 to 2500 RCYBP / 8500 to 2700 cal BP)

Prior to the advent of radiocarbon dating, North American archaeologists began to suspect that some of the sites they were excavating came between two more distinctive and easily recognized archaeological entities: Paleoindian with lanceolate points on the early end and Woodland with mounds and pottery on the late end. The advent of radiocarbon dating made it clear that there were several thousand years between Paleoindian and Woodland and that many sites did indeed belong to that interval. The period came to be known as the Archaic, and it was initially defined more by the absence of lanceolate points, mounds, and ceramics, than by the presence of distinguishing characteristics of its own (see McElrath et al. 2009; Emerson and McElrath 2009).

The Archaic period spanned the time when the post-glacial environment of Minnesota continued to moderate, and ecosystems similar to those of modern times evolved by the later portion of the Archaic. During the first half of the Archaic period, the northern hemisphere experienced an episode of warm and dry weather from about 9000 to 6000 BP that is variously referred to as the Altithermal, the Middle Holocene Climatic Optimum, and the Prairie period. The peak of this warming period was reached around 7800 BP. The hot and dry conditions persisted at their maximum for about 1000 years before gradually giving way to a cooler and wetter climate that led to the evolution of ecological communities similar to those of the modern era by about 5000 BP. The climatological and ecological changes had profound impacts on subsistence and settlement patterns. Warming and drying during the period would have been dramatic, with prairie spreading across most of Minnesota, except for the eastern portion.

The dramatic environmental changes of the Altithermal would have caused major shifts in the lifeways of the people, as post-glacial animal species of the forest such as moose, caribou, and deer 41 were replaced by prairie species such as bison. Plant communities also would have changed with the spread of the prairie, and wild rice may have been gathered during this time. Surface water significantly decreased during the Altithermal, as shallow lakes and wetlands dried up or were greatly reduced in size.

It is likely that Archaic period populations engaged in seasonal rounds of resource gathering as the climate stabilized following the retreat of the glaciers. Small bands would have returned to seasonal campsites, and territories may have been relatively limited. With the onset of warmer and drier prairie conditions, however, resources would have become less predictable, and populations would have been pushed into shrinking areas surrounding the larger lakes and streams. The appearance of groundstone milling tools suggests that there was a greater use of seeds and other plant foods. Domesticated dogs, used for transport, suggest that longer-distance travel was required to keep up with migratory bison herds. Group sizes appear to have remained small throughout the Archaic, and known site locations indicate that a high value was placed on a proximity to game, water, and supplies of wood.

The Archaic tradition was originally defined in the eastern United States, where the beginning of the Archaic was dated to around 10,000 RCYBP. Continued research allowed archaeologists to begin defining the Archaic in its own right and not merely by the absence of traits. For example, various styles of stemmed and notched Archaic projectile points were recognized, although many of these were not as distinctive as the older Paleoindian styles and could be difficult to distinguish from one another. Archaeologists were also able to suggest that Archaic groups used a broader spectrum of plant and animal resources than their Paleoindian predecessors, did not cover as much territory, and eventually had longer-term settlements in river valleys. The appearance of groundstone tools was also recognized as a characteristic of the Archaic. Some tools, such as grinding stones, were used for more intensive processing of plant foods.

McElrath et al. (2009) suggest that the division of the Archaic into Early, Middle and Late stages effectively began with Fowler's (1959a, 1959b) reports on the deeply stratified Modoc rockshelter. In these reports, the Archaic was divided into three parts of 2000 years each - more as a pragmatic choice than as a series of divisions discerned in the archaeological evidence. As this trifold division proliferated, there was little consistency in the use of terms, whether for chronology, associated artifacts, subsistence, or social organization. This problem still remains largely unresolved.

Once the idea of the Archaic was established, use of the term was extended west to the Plains where it also applied to sites from the period between the Paleoindian and "Plains Woodland" traditions. On the Plains, however, both the chronology and character of the Archaic turned out to be different than in the east. The Plains Archaic began somewhere around 8000 to 7500 RCYBP, which is on the order of 2000 or more years later than the Eastern Archaic, and the Plains Archaic also persisted later in time. Any changes in subsistence patterns are more subtle, and bison hunting was the dominant subsistence activity beginning before the Plains Archaic and continuing past the Plains Archaic. As in the east, a variety of Archaic notched points have been recognized, and groundstone tools also appeared.

The idea of Early, Middle and Late Archaic was also adopted for the Plains Archaic. As in the east, the use of these terms on the Plains has not been entirely consistent. To the degree that there is consensus for the Central to Northern Plains, however, Early Archaic refers to a fairly distinctive Early Side-Notched Point complex, Middle Archaic refers to the McKean and Oxbow complexes, while Late Archaic includes Pelican Lake and Besant, among others (cf. Frison 1998; Kay 1998). It is important to note that Besant is quite late and in fact is contemporaneous with Plains Woodland complexes. This highlights the use of "Archaic" to mean a cultural tradition rather than a time period. Much of the confusion in Archaic taxonomy may stem not from disagreement over the basic 42 archaeology so much as it does from an inherent conflict in the use of Archaic to mean both a cultural tradition and a time period. Because Canadian sites are discussed in this report, it is also useful to note that the standard Canadian culture-history uses a different nomenclature. In that system Early, Middle, and Late Precontact are generally equivalent to Paleoindian, Archaic and Woodland, and each is similarly subdivided.

6.2.1 Early Plains Archaic Overview (7700 to 5500 RCYBP / 8500 to 6300 cal BP)

The nomenclature used for the Early Plains Archaic varies somewhat. There is general agreement, however, that a widespread cultural complex follows the Paleoindian tradition across the Plains from the eastern periphery to the Rocky Mountains, and from the latitude of Kansas into the southern parts of the Canadian prairie provinces. Although the sites and artifacts are not identical across this broad range, there are strong resemblances. There does not seem to be a single, formal name for this cultural manifestation, but “Early Side-Notched Point complex” is seen with some regularity and that terminology will be used in this report.

The chronology of the Early Side-Notched Point complex is becoming reasonably well known. Widga (2006) reviews about 60 relevant dates for the complex, including a number of more recently- run assays. Most of these were in the range of 7600 to 5500 RCYBP. A small number of older dates had larger margins of error, and come from assays run early in the development of radiocarbon dating. The following discussion includes summaries of information on selected sites that are associated or possibly associated with the Early Plains Archaic.

6.2.2 Regional Early Plains Archaic Sites

21CR141 Site 21CR141 is located in an alluvial fan in the Minnesota River Valley about a mile west (and upstream) of 21CR155, near Chaska in south-central Minnesota (Schoen 2006). The site was found in 2006 during a survey for a proposed bridge replacement on TH 41. Phase I and II investigations at the site included shovel testing and formal excavation of eight m2. Depending on the depositional context, excavation varied in depth from 65 cmbs to more than 4.5 meters. Archaeological deposits were also exposed in a ravine cutbank along the east edge of the site.

A geomorphological evaluation of the site was conducted by Kolb (2006). The ravine cutbank exposure showed a 5Ab soil horizon extending about 30 meters north–south at a depth of around 3.0 to 3.6 meters. Shovel testing indicated that the horizon was truncated on the west by an old drainage channel, and the remnant was estimated to measure five to meters in width east–west. The horizon is described as containing "substantial amounts of bone, charcoal, and chipped stone debris in an intact midden deposit with good preservation" (Schoen 2006:66). Charcoal samples from about 320 cmbs were dated to 6160 +/- 50 and 5760 +/- 70 RCYBP.

Testing of the buried horizon yielded lithic flaking debris, bone, and charcoal. Faunal remains included bison, duck, deer, small mammal, turtle, and bird. No features were found. In addition to the deeply buried deposits, shell-tempered sherds and a flake were found in a shovel test at 35 to 40 cmbs. Further testing failed to find any additional traces of this component, which was interpreted as an isolated find.

21CR155 Site 21CR155 was identified in a deeply-buried setting on alluvial fans in the Minnesota River Valley near the city of Shakopee (Florin et al. 2013; Florin et al. 2015). The Archaic had cultural deposits that included bison and other terrestrial and aquatic remains buried as deep as four meters. The site contained multiple occupations, spanning most of the Holocene from ca. 7100 to 500 RCYBP (8000 43 to 500 cal BP), with a cluster of dates from two areas of the site between about 8000 to 7000 cal BP. The Archaic points include an unnotched "Delong" type and a medium-sized notched type. The site was determined eligible for listing on the NRHP and a Phase III data recovery was conducted.

21NL63 (Fritsche Creek II) Site 21NL63 is a multicomponent site located in a complex landform at the mouth of Fritsche Creek on the eastern edge of the Minnesota River Valley near New Ulm in south-central Minnesota. The landform includes older alluvial fan deposits, substantial colluvium, as well as more recent fluvial deposits related to the modern configuration of the creek. The site was discovered in 1990 during a survey for a proposed bridge replacement over Fritsche Creek. Phase II work in 1992 included surface survey, shovel testing, investigation of a bone bed exposed in a road cut, and excavation of two square meters. This work identified a surficial Late Woodland component at the surface, and a bone bed with associated artifacts at about 85 to 100 cmbs.

A limited Phase III data recovery in 1994 (Roetzel and Strachan 1992; Roetzel et al. 1994) included excavation of three 1 x 2 meter blocks in location where electrical transmission line poles were to be placed. The depth of maximum depth of excavation varied from 130 to 240 cmbs. The report indicated the presence of a ca. 15 cm--thick bone bed in a paleosol, but also indicated that materials were found outside of the bone bed and that there was evidence of substantial rodent activity. The road cut was also re-inspected and a sample of bones and artifacts recovered. Radiocarbon dates were pending when the report was submitted; a later report indicates a single resulting date of 6080 +/- 100 RCYBP (Monaghan et al. 2006:6.1).

The combined Phase II and III investigations recovered a total of 1,425 bone fragments and lithic artifacts from the buried component. Most of the bone consisted of small, unidentifiable fragments, and most of the identifiable bone was bison. It was noted that the bison appeared larger than modern forms, but no specific taxonomic identification was made. Pocket gopher, bird, snake, and mollusc were also noted. The assemblage included an antler tine that may have been a tool. The lithics were limited in number, but included scrapers, utilized flakes, flaking debris, and the midsection of a biface or point that was described as "the midsection of a flint projectile point...... Because the base is missing, it is necessary to simply conjecture as to the shape of the entire specimen. But because the sides of the point are almost completely parallel, it is likely that the point was relatively long in the range of 4+ inches and clearly appears to be of the Plano variety" (Roetzel and Strachan 1994:29).

Further geomorphlogical and archaeological work was conducted at the site in 2004 in connection with the MnDOT Minnesota Deep Testing Protocol study (Monaghan et al. 2006). This investigation included geophysical survey, excavation of trenches to evaluate site stratigraphy and geomorphology, and limited archaeological excavation to recover samples of exposed materials. Archaeological materials recovered during this work came mostly from the upper Woodland component. Recoveries from sub-plowzone contexts were mostly limited to animal bone fragments, many burned or calcined. The geomorphological investigations more fully explained the development and stratigraphy of the site and landform. They provided an additional date of 6570 +/- 40 RCYBP for the previously identified buried component, and also identified an older paleosol in parts of landform. The older paleosol and associated archaeological materials were dated to 8100 +/- 40 RCYBP.

21YM35 (Hildahl No. 3) Site 21YM35 is a multicomponent site situated on a colluvial and alluvial slope on the southern edge of the Minnesota River Valley south of Granite Falls in southwestern Minnesota (Dobbs 1979; Hudak 1978). The site was discovered in 1971, and a controlled surface collection was made in 1976. Two phases of more intensive excavation were undertaken in 1977 and included 108 square meters of excavation. In the first phase, excavation by hand began at the surface; in the second phase, the upper 30 cm was mechanically removed prior to the start of hand excavation. Late Woodland and 44 apparently Middle Woodland components in the upper strata of the site were judged to have poor integrity. Testing focused on a lower component that began around 45 cmbs and continued to about 120 cmbs. The cultural affiliations of the various occupations in the lower component are not clear, but it contains Archaic and probable Late Paleoindian components. Six features were found, including a hearth and five rock concentrations of unknown function.

The faunal sample totaled 954 pieces, 99 percent was likely mammalian. Identifiable specimens included bison, muskrat, raccoon, turtle, fish and birds. The report suggests that some of these may have been noncultural. The sample included a worked and worn cervid antler tine that had been used as a tool. Bone fragments were distributed widely across the site, although concentrations did occur. Few burned or calcined pieces were found.

The lithic sample totaled 725 pieces, most of which were flaking debris. There were also cores, bifaces, scrapers, utilized flakes, and eight projectile points from the Archaic and possible Late Paleoindian components. Five of the points came from a part of the site defined as Area 1, and three were from Area 2. The point descriptions provided below, aside from general morphology, are from Dobbs (1979), as we did not re-examine the collection. Of the five points from Area 1, one is a medium sized side-notched type, three are corner-notched, and one is a broad side-notched or stemmed type. The side-notched point has a concave base that is thinned and ground. The notches are wide and shallow, with the notch on one side being larger than the other. Dobbs notes that the blade has been heavily reworked, so the original blade form is not clear. Based on Dobb's description and the photo, this point is similar to the Early Side-Notched Point complex. One of the corner- notched points is similar to an Early Archaic point presented in Frison (1991:85, Figure 2.46, h). The cultural affiliation of the other points in Area 1 is not clear. There is a resemblance to some points from the Itasca Bison Kill site but the relationship between the various point styles at Itasca is not clear and therefore comparisons to that site are uncertain.

The points from Area 2 include two lanceolate types and a broken point that appears to be side- notched. The lanceolate points have parallel flaking and basal thinning. Both lanceolate points are ground on the base and about a quarter of the way up the side of the blade. Dobbs (1979) compares one of the lanceolate points to specimens from the Late Paleoindian Browns Valley burial, and the other to a point from Horizon II at the Cherokee Sewer site that is similar to the Delong type. These points resemble resharpened Late Paleoindian lanceolate types, such as Agate Basin, based on the slightly convex blade/stem form, basal grinding, reworked blades, and the general size and shape. However, it is possible that they are the Delong type. The side-notched point is asymmetrical in shape and consists of a triangular blade, with evidence of notching where the base is broken off.

21YM47 (Granite Falls Bison Kill) 21YM47 is a multicomponent site located in an alluvial and colluvial landscape on the west edge of the Minnesota River Valley bottom near Granite Falls in southwestern Minnesota. There is no comprehensive report on the site, but the available information is summarized by Anfinson (1997:36) and Kuehn (2000, personal communication 2015). The site was discovered in 1988 by the landowner when he was digging a trash pit. The site was investigated in 1988, 1989 and 1990, with the excavation of 13 square meters and coring to explore stratigraphy and geomorphology. This work documented a minor component at about 1.8 meters below surface (not discussed here), and the main bison bone bed on a sloping paleosurface between about 2.6 and 3.0 meters below surface. The deeper component included a flintknapping feature that consisted of a three to four cm-thick layer of Swan River Chert flaking debris, an anvil stone, and a hammerstone. There are three accepted radiocarbon dates for the lower component. Two charcoal samples returned dates of 7085 +/- 70 and 7035 +/- 80 RCYBP. A bone sample returned a date of 7050 +/- 120.

45

Lewis and Heikes (1989) conducted an analysis of some of the bone from the site, and Kuehn (personal communication 2015) is updating and completing the analysis. The assemblage includes over 3,800 pieces of bison bone, as well as one canid incisor that compares favorably but not conclusively with domestic dog. Analysis indicates a minimum of six adult and six juvenile bison, possibly cows and calves. The kill appears to have been in the late fall to winter. The elements present and the clustering of bones indicate butchering, with the preferred parts of the carcass removed and the less desirable parts abandoned at the site. Examination of the horn cores suggested that this could be Bison occidentalis.

A complete inventory of the lithic assemblage is not available. An approximate inventory gleaned from notes stored with the site materials, however, indicates that the total lithic assemblage includes more than 1,700 pieces, and that about 95 percent of this is Swan River Chert. Other documentation indicates that the assemblage includes, at a minimum, a large ovate biface, two chopping tools, flaking debris, three projectile points found in the bone bed, and one projectile point found in backdirt.

13CK61 (Simonsen) Site 13CK61 is a multicomponent site located on an alluvial terrace along an abandoned meander of the Little Sioux River near Quimby in northwestern Iowa (Agogino and Frankforter 1960; Frankforter 1959b; Frankforter and Agogino 1959a, 1959b, 1960; Widga 2006). The site was investigated between 1956 and 1961. Excavation was relatively extensive, although the total area excavated is not clear. Eight stratigraphic units were identified in about 17 feet (more than five meters) of sediment, and three of these produced cultural material. Zone 3 produced bison and canid bone, a hearth with charred logs, as well as ash and burned earth. Zone 5 produced a projectile point. Zone 7 produced the most substantial remains. Although the report does not discuss features as such, it reports the presence of two large rocks which, based on the pattern of adjacent shattered bone, may have been anvil stones used for breaking bones to get marrow. Cobbles scattered around the larger rocks may have served as hammerstones. A charcoal sample from Zone 7 was dated to 8430 +/- 520 RCYBP. More recently, however, a bone sample from Level 3 (Zone 3) was dated to 6433 +/- 69 RCYBP, and three bone samples from Zone 7 were dated to 7006 +/- 72, 6717 +/- 68, and 6905 +/- 72 RCYBP (Widga 2006:60).

The excavations in Zones 3, 5, and 7 produced abundant bison remains, including several skulls, which were identified as Bison occidentalis. The report notes that numerous articulated bones were observed. The only faunal remains other than bison came from small amphibians and rodents. A recent re-examination of the faunal materials concluded that they represented the actual kill site rather than an associated butchering site. The remains were of a small herd dominated by prime-age males, and that they may have been killed during late summer or fall, although that conclusion is tentative (Widga 2006).

The lithic assemblage is described as containing knives, scrapers, flakes and five projectile points. Agogino and Frankforter (1960:415) describe the general form of the points: "The Simonsen point is triangular, side-notched and concave-based. The side notches are shallow and exhibit moderate grinding and the base has both thinning and grinding. The points are lens-shaped in cross section and generally widest at the shoulder just above the notch. In total length these artifacts range from 25 to 50 mm."

13CK405 (Cherokee Sewer) 13CK405 is a multicomponent site in a series of floodplain and alluvial fan deposits in the Little Sioux River valley (Anderson and Shutler 1978; Anderson and Semken, eds. 1980) near the town of Cherokee in northwestern Iowa. The site was discovered in 1973 during construction of a sewage treatment facility. Six weeks of salvage excavation were conducted in 1973, and additional intensive 46 investigations conducted in 1976. A large area was excavated using a combination of mechanical sediment removal and hand excavation. The work identified three main components, called Horizons I, II, and III that were buried about one, three, and five meters below surface. Although these horizons could be stratigraphically subdivided in parts of the site, the three broader horizons are the basic units for analysis. A suite of 22 radiocarbon dates were used to evaluate geomorphic stratigraphy of the site, as well as to more directly date the archaeological deposits. Based on this and on typological information, Horizon I was classified as Middle Plains Archaic (ca. 6350 RCYBP), Horizon II as Early Plains Archaic (ca. 7200 RCYBP), and Horizon III as Late Paleoindian (ca. 8400 RCYBP). A synopsis of findings from Horizons I and II are presented below.

Horizon I came from a paleosol in an alluvial fan. Features include fire hearths, a bone concentration, and a charcoal concentration. Charcoal and carbonized (but not burned) wood samples were dated to 6300 +/- 90 and 6500 +/- 200 RCYBP. Based on these dates, the investigators proposed an approximate age of 6350 RCYBP for this horizon.

The faunal remains from Horizon I were well-preserved but less abundant than in the other strata. Bison was the most abundant taxon identified, with a minimum of ten animals represented, possibly from a cow-calf herd. Seasonality could not be determined. Canid remains appeared to represent 2 large and 3 small individuals. The large canids were probably domestic dogs, while the smaller canids were probably coyotes. One cut mark was found on the canid bones. Cervids were represented by a section of antler that probably served as a flintknapper's billet. Other worked bone pieces included choppers, awls, a possible pick or punch, a cut bone tube, and a fragment of a flute made from a bird bone. Horizon I is interpreted as a processing site rather than the actual kill site. Chopping marks, cut marks, and spiral fractures indicate intensive processing and marrow extraction. In addition to the animal bone, pieces of human deciduous incisor indicate the present of children at the site.

The lithic assemblage from Horizon I comprised about 400 pieces, including flaking debris, end scrapers, a biface, a chopper, a burin, and five side-notched projectile points. Four of the points are described as well-made, while the fifth is a side-notched but minimally modified flake point. The blades tend to be long but lack parallel lower margins. Bases are straight to subconcave, and only one is ground. All of the notches are ground. Lengths range from about 30 to 60 cm, and widths from about 12 to over 25 mm. Based on drawings of the points, it appears that the longest and widest specimen retains its original form and dimensions, while the shorter and narrower specimens have probably been resharpened or otherwise reworked. Wear patterns on three of the points suggest that they were also used as knives. In addition to the flaked stone artifacts, a number of hammerstones were found in Horizon I.

Horizon II produced the richest set of remains from the site. The materials were concentrated on the toe of the alluvial fan, but also continued onto the adjacent floodplain. The component was contained in a paleosol. Features included hearths, amorphous areas of burned earth, and dense concentrations of artifacts and bone fragments. Charcoal samples were dated to 6800 +/- 190, 7370 +/- 75 and 7370 +/- 100 RCYBP. Based on these dates, the investigators proposed an approximate age of 7200 RCYBP for this horizon.

The faunal remains from Horizon II were generally small and fragmentary, indicating that carcasses were heavily butchered. The bone was also less well preserved than in the other horizons. Bison was the only identified taxon; a minimum of 15 animals were represented. Specific taxon was not determined. The original analysis of the faunal evidence indicated a late winter kill, possibly February or March. However, a re-examination of the remains based on updated techniques (Whittaker 1998) indicated instead a late fall to early winter kill of calves, young adults and old adults. Horizon II is interpreted as a processing site rather than the actual kill site. Chopping marks, 47 cut marks, spiral fractures, and fracturing of most limb bones indicated intensive processing and marrow extraction. Worked bone tools included choppers, awls, and possible hide fleshers. A possible bone pendant was also found. In addition to the animal bone, a human deciduous incisor and a fragment of a deciduous molar were found, indicating the presence of children at the site.

The lithic assemblage from Horizon II included over 10,000 pieces. Of these, more than 9,000 were of Tongue River Silica; this contrasts strongly with the raw material profiles of both the earlier and later components. The assemblage included flaking debris, bifaces, endscrapers, choppers, burins, and 22 projectile points, including 10 side-notched points, nine un-notched points, and three other fragments. The point blades are ovate, except for one specimen that has parallel edges on the lower part of the blade. The notched points have the following attributes: ground bases that tend to be subconcave; point lengths average around 30 mm; and point widths around 20 mm. Unnotched points have the following attributes: bases are generally straight and not ground; point lengths average around 30 mm; and widths around 20 mm. The investigators note that the unnotched forms could have served as preforms for the notched forms, although some of the unnotched forms show wear and even impact damage in one case, suggesting they were used as tools or dart points. In addition to the flaked stone tools, there were hammerstones, an abrader, and a large part of a milling stone made from a slab of Sioux Quartzite that was shaped by pecking, with one face was smoothed by use. Given that they interpreted the site as a winter camp, the investigators speculated that the mill stone may have been used to pulverize meat and hackberries for pemmican.

13ML62 (Hill) 13ML62 is a single-component site in a terrace of Pony Creek in southwestern Iowa (Frankforter 1959a; Widga 2006). The site was discovered during road construction, and salvage excavation was conducted in 1958. In the time between the initial testing and more extensive systematic excavation, flooding eroded most of the archaeological deposits. The main archaeological stratum at the site was buried about 5.7 meters below the modern surface. Features included burned areas and hearths. The site was interpreted as a camp or habitation rather than a processing site or bison kill. At the time of the investigation of the site, a charcoal sample provided a date of 7250 +/- 250 RCYBP. More recently, collagen from bone provided dates of 6592 +/- 70, 6696 +/- 78, and 6484 +/- 58 RCYBP.

The faunal remains were predominantly bison and appear to represent a cow and calf. Canada goose, turtle, deer or pronghorn, and rodents were also identified. All species except the rodents showed butchering marks or burning, indicating use by humans. Modified bone included a cylindrical bead made from bird bone and a grooved bison long-bone fragment.

The lithic assemblage included flaking debris, flake tools, knives, notched and un-notched end scrapers, and five projectile points. Four of the points were side-notched, and one was unnotched. Point lengths ranged from about 20 to 40 mm, and widths from about 15 to 20 mm (cf. Frankforter and Agogino 1960:70, Fig. 5). The side-notched points have ovate blades that curve smoothly from notch to sharp tip. The notches are low on the blade and generally U-shaped. The bases are concave. The site report contains no information on basal thinning or grinding. Widga (2006:315) notes that the unnotched point resembles the Delong type.

21AK11 (Sandy Lake Dam) 21AK11 is a stratified multicomponent site at the outlet of Sandy Lake in northeastern Minnesota. Parts of the site have been previously investigated; the work referenced here consisted of Phase III excavation that took place in 2004, when a 33 square-meter block sampled Archaic deposits (Bradford 2013). Numerous copper artifacts were recovered from this component, indicating an association with the Archaic Old Copper Complex. Calcined bone from a copper-bearing hearth was dated to 5690 +/- 30 RCYBP.

48

A notched projectile point was recovered from this component. It appears to be crudely made, with limited edge retouch on a fortuitously shaped flake. The summary report compares the point to Oxbow, a Plains Archaic type. Inasmuch as the point has a concave base and rough ears, there is a resemblance to Oxbow. However, the date of 5690 +/- 30 RCYBP would place it very early in the oxbow chronology, and it might be argued that the point is so casually made as to preclude meaningful evaluation of typology. The 21AK11 point also resembles the side-notched points from the Early Side-Notched Point complex.

21CE1 (Itasca Bison Kill) 21CE1 is a bison kill site located along Nicollet Creek in Itasca State Park in northwestern Minnesota. The site was first investigated by Jenks and Wilford in 1937 (Jenks 1937). They dug about 440 square meters, with some excavation to about three meters deep. Shay (1971, 1978) conducted additional excavation at the site in 1963, 1964, and 1965. These later investigations covered about 244 square meters, with excavation to a depth of 3.9 meters in parts of the site. Excavation was divided between a bog along Nicollet Creek and a hillside overlooking the bog, with artifacts recovered from both contexts. More recently, Widga (2006) re-examined the faunal materials from 21CE1, reconsidered the interpretation of site formation and integrity, and obtained three additional radiocarbon dates on bone from the site. Itasca is an interesting but problematic site, a fact recognized by the early excavators. Some questions about the site are summarized below.

First, the deposits on the hill and in the bog seem to be related, and some of the materials in the bog may have been redeposited from the hill, but the relationships between the deposits are not clear. Second, it does seem clear that parts of the faunal assemblage in the bog are cultural and parts are not, but it has proven difficult to distinguish one from the other in many cases. Third, and related to the previous point, movement of bone by fluvial processes means that archaeological bone may not be in its primary context while archaeological and non-cultural bone deposits may have been further mixed. Fourth, the original six radiocarbon dates from the site did not clearly date the archaeological deposits. For example, three samples were dated twice and the results for single samples varied by more than 500 years to almost 1,000 years. In addition, the dates were reported in relation to environmental reconstruction, pollen zones, and geomorphological stratigraphy, but their relationship to the archaeological remains is ambiguous. Shay seems to have finally concluded that the site was used between 7,000 and 8,000 years ago. Widga obtained three additional dates on bone samples. These resulting dates of 7112 +/- 69, 7027 +/- 59, and 7552 +/- 86 RCYBP are broadly consistent with the results obtained by Shay. The question remains, however, of what is being dated: is it a bison kill or a natural death assemblage?

Widga (2006:293-294) summarizes his re-examination of the Itasca site: "there was probably at least one small bison kill present in the lower reaches of the Nicollet Creek valley dating to approximately 7800 cal BP and perhaps a second kill around 8300 cal BP.... Both before and after this event, bison, turtles, fish and other animals would have been part of the attritional fossil record in the locality.... Periodic flushing of spring meltwater from small upstream lakes may also have contributed animal bone clasts to the deposits of the Itasca locality...."

Finally, it should be noted that three ceramic sherds were recovered from the hill area of the site during the 1937 excavations. This raises the possibility that some unknown part of the lithic assemblage could be associated with a Woodland component. However, the meager representation of ceramics and the fact that no further ceramics were found during later investigations suggests that this is likely a minor consideration. Given these issues and caveats, a synopsis of the faunal and lithic assemblages is presented below.

Faunal remains were abundant, totaling over 9,000 pieces. Bison was the most common taxon, comprising about 68 percent of the identifiable remains, and probably much of the additional 49 unidentified bone. The bison bone represented at least 16 individuals, mostly immature females, which Shay felt were within the size range of B. occidentalis. Shay identified 39 other taxa, including various mammals (including domestic dog), birds, fish, and turtles. Of these, he felt that bison, dog, and some species of fish and turtles were associated with human activities. Widga, however, notes that Shay's opinion was "primarily based on historically documented patterns of animal-use, inferred economic value, and disproportionate numbers of certain fish taxa, not on systematic analyses of bone modification patterns," and Widga felt that there was no evidence for sustained use of any taxon other than bison.

The lithic assemblage at 21CE1 came mostly from the hill. Shay notes that the bog excavations produced only 18 lithic artifacts, including knives, choppers, and projectile points, and he concluded that many of these had been redeposited, presumably from the nearby hillslope. The total assemblage included lithic debris, preforms, bifaces, end scrapers, side scrapers, various bifacial and flake knives, choppers, perforators, gravers, retouched or utilized flakes, large and sometimes heavily worn tools resembling large end scrapers, and projectile points. The points are somewhat variable, and include specimens that might be called side-notched, corner-notched, and small triangular (or alternatively “small lanceolate”). In addition to the flaked stone artifacts, there were anvil stones surrounded by cores and core fragments.

25BT3 (Logan Creek) 25BT3 is a multicomponent site in an alluvial fan on Logan Creek in eastern Nebraska. The site was investigated from 1957 to 1963. Because these investigations have been minimally published, this synopsis relies on one brief primary report (Kivett 1962), plus a synopsis and new information provided by Widga (2006). The archaeological deposits were estimated to cover an area of more than 100 meters by 30 meters. Features included hearths, pits, shallow basins and post molds. Radiocarbon dates for the Early Archaic component ranged from 6020 +/- 160 to 7350 +/-270 RCYBP. While at least nine occupation levels were identified, Kivett notes that the Logan Creek Complex component, which is an Early Side-Notched Point complex, was found in Zones A, B, C and D; Widga add Zone F. Each of these is briefly reviewed below. Above these levels are one or two poorly defined Plains Woodland components that are not addressed here.

Faunal remains included abundant specimens of bison, but other taxa were also identified, and even common in some strata. Mussel shells were also common in all zones. These other taxa are discussed below by zone. The stratigraphic subassemblages indicate bison kills throughout the year. Modified bone tools included bone spatulas, awls, tubular bone beads, shaft wrenches, needles, flaking tools, and polished bone scraps. There were also mussel shells with serrated edges.

The lithic assemblage included flaking debris, end scrapers, side-notched end scrapers, bifaces, a drill, flake tools, and more than 60 projectile points. Kivett described side-notched points of small to medium size, with straight to concave bases that were often ground; lengths of these points ranged from about 20 to 50 mm. He also described triangular points, rarely with basal thinning, that ranged from about 20 to 25 mm in length. These points may be the unnotched Delong type. In addition to flaked stone artifacts, the excavations produced fragments of groundstone that may represent a metate, appearing " …to be from slabs used as [a] base for grinding stones" (Kivett 1962:3), as well as grinding and polishing stones.

Zone A appears to be associated with late spring bison kills, and possibly selected culling or the accumulation of single-animal kills. Other identified taxa include deer, rabbit and gopher. Zone B produced the largest assemblage of cultural materials. This zone appears to be associated a late spring kill of multiple bison from the same herd. A total of 20 taxa were identified in this stratum, including elk, antelope, deer, badger, beaver, skunk, wild turkey, turtle, and rodents. Zone C appears to be associated with a small number of bison taken in single kills. This zone also contains the 50 highest proportion of non-bison taxa. These include elk, antelope, deer, canids, badger, otter, raccoon, wild turkey, turtle and rodents. Zone D produced abundant bison remains. Other identified taxa include antelope, deer, canids, rabbit, raccoon, bird, turtle and rodents. Zone F produced a small faunal sample consisting primarily of bison. Other identified taxa included elk and deer.

32RI775 (Rustad) Site 32RI775 is a multicomponent site located in alluvial fan and aeolian deposits along the south bank of the Sheyenne River in southeastern North Dakota (Michlovic and Running, eds. 2005). The site was discovered in 1992. Salvage work was conducted at the site in 1992, and additional excavation followed between 1993 and 1998. Excavation totaled about 200 square meters. These investigations identified a substantial Early Archaic component in alluvial fan deposits. There was also a minor Woodland component in the overlying aeolian sediment and a minor Late Paleoindian component in the underlying lacustrine clay. The Early Archaic component was preserved in two paleosols about two meters below the modern surface; these were distinguishable in the central part of the fan, but that merged towards the fan margin. A living floor was identified. The distribution of features, bone, artifacts, and soil patterning on the living floor suggested the existence of a small structure. Other features include hearths, fire pits, and discard piles where debris was concentrated. Five charcoal samples from the Early Archaic component provided dates ranging from 7180 +/- 90 to 7590 +/- 90 RCYBP and six additional dates from soil samples helped clarify the landscape history.

The faunal sample from Rustad included over 50,000 pieces, including over 2,100 pieces that were deemed identifiable. More than 80 percent were bison, representing a minimum of 18 individual bison of an unidentified extinct form. Most seem to have been young adult females, and it was tentatively concluded that several of the animals may have been killed in the late summer or fall and one in the winter, although the evidence was not strong. The bison bone exhibited cut marks, chopping blows, impact fractures and spiral fractures. The assortment of skeletal elements and the breakage patterns indicated heavy butchering of most of the carcass and intensive utilization, including limited marrow extraction. Other taxa identified included canids, skunk, marten, beaver, muskrat, gopher, eagle, grouse, magpie or grackle, turtle, frog, toad, snake, shrew, vole, mouse, and lemming. The canid remains represented both large and small varieties. There was insufficient evidence to indicate whether these were wolves, foxes, or domestic dogs. Some of the canid bone showed spiral fractures, indicating that the animals might have been butchered and eaten. Of the various non-bison taxa, cultural modification was noted only for canids and rabbits, although many of the species were judged to be part of the archaeological assemblage.

The lithic assemblage totaled almost 7,000 pieces, including flaking debris, bifaces, end scrapers (one side-notched), other unifacial tools, cores, bipolar cores, and 30-plus projectile points and fragments. The points include side-notched, corner-notched and unnotched triangular specimens. The side- notched points (n=10) are described as small, about 25 to 35 mm long, with straight to slightly concave bases and seven show basal grinding. The triangular points (n=8) are mostly straight based, and two have basal grinding. These points are similar to the Delong type. Two are said to have similarities to Late Paleoindian forms. One of these has a deeply concave base and converging blade edges; the other is described as having a basal platform that looks as if it was prepared for fluting. The corner-notched points (n=2) include a basal fragment that may have been broken during manufacture, and a flake point. In addition to the flaked stone artifacts, three hammerstones were found.

39BF2 (Medicine Crow) Site 39BF2 is a multicomponent site located on the edge of a high terrace that previously overlooked the valley of the Missouri River, and now overlooks the Big Bend reservoir in central South Dakota. The site was excavated in 1957 and 1958 as part of the Smithsonian Institution River Basin Surveys program, prior to the construction of Big Bend damn. A comprehensive report was completed in 51

1989 and published in 1995 (Ahler and Toom, eds. 1995). The site includes an extensive Plains Village component and, in parts of the site, multiple Archaic components. This synopsis focuses on the Early Archaic component. At the time of the Early Archaic habitation, the river valley was not deeply entrenched and the site would have been near a small lake in the floodplain. Subsequent down-cutting by the river left the site perched on a terrace about 100 feet above the now-submerged floodplain. There are no radiometric dates for the site, and site chronology is estimated based on correlation with dated components from other sites.

The Early Archaic remains at the site are identified with analytical and stratigraphic divisions called C/T Units 5 and 6. C/T Unit 5 occurred in the lowermost part of Holocene aeolian sediment that blankets the landscape, and C/T Unit 6 occurred in the uppermost part of the underlying late Wisconsinan sandy outwash. C/T Unit 5 was the most substantial of the preceramic cultural horizons, and C/T Unit 6 included only materials that were intrusive from the immediately overlying C/T Unit 5. The depth of these deposits varied, and the outwash surface was irregular in contrast to the nearly level modern surface; much of the C/T Unit 5 archaeology was found at a little more than one meter deep. Features included both diffuse and rock-filled hearths, bone concentrations, charcoal concentrations, ocher concentrations, and a burned area. When this evidence is contrasted with the structure and features of the Plains Village components, it is interpreted to mean that "the preceramic occupations were temporary, short-term camping episodes centered around hearths of various forms" (Ahler 1995:132).

The faunal assemblage available for analysis was limited to identifiable and modified pieces, totaling only 41 pieces. Apparently bone fragments that were judged to be unidentifiable had been culled and discarded. Bone from the preceramic components was not well preserved in general, which also hampered the analysis. The available faunal remains indicated the use of bison and antelope.

The lithic assemblage was more substantial, totaling more than 8,300 pieces. These included flaking debris, bifaces, chopping tools, flake tools, core tools, cores, bipolar cores and projectile points. The text concerning description and analysis of the projectile points was extensive, and it was difficult to extract a comprehensive inventory of points from C/T Units 5 and 6, although there were a variety of notched and unnotched forms (Delong). Among the projectile points, eight specimens classified as Simonsen and eight as Delong. One specimen, called "Crude Lanceolate”, may represent the McKean type, although the typological suggestion seems tenuous. In addition to the flaked stone artifacts, there were four large stones that were interpreted as metates and manos.

FaNq-25 (Gowen 1) Gowen 1 is a single-component site located in an alluvial terrace of the South Saskatchewan River, in the city of Saskatoon, south-central Saskatchewan (Walker 1992). It was 70 meters east of Gowen 2 (see below) but not part of the same deposit. The site was discovered in 1977 when it was exposed by heavy equipment borrowing fill in the city landfill. Several weeks of investigations included hand excavation of intact site areas, salvage of artifacts from disturbed site sediment, mechanical exposure and inspection of more peripheral areas of the site, and backhoe excavation of stratigraphic trenches.

The archaeological deposit was a single 10 cm-thick stratum in a paleosol just over one meter below the surface. This stratum was isolated between strata of culturally sterile sediment, so materials from disturbed contexts could be securely associated with the component even though they lacked horizontal provenience. The site was estimated to cover about 300 square meters, and excavation totaled 112 square meters. Features included hearths and small pits of uncertain function, although the latter may have served as smudge pits associated with processing of bison hide. Radiocarbon dates on bone and charcoal ranged from 5670 +/- 135 to 6150 +/- 110 RCYBP (a date of 4725 +/- 130 was treated as a statistical outlier and rejected). Based on the range of artifacts represented, the site was interpreted as a briefly occupied camp or special purpose site rather than a bison kill. 52

Structurally, it was proposed that the site consisted of a bone processing and stone working area that was centered on a large hearth, with possible hide-working focused around the smudge pits and smaller pits in the area.

Faunal remains were abundant at Gowen 1. Nearly all of the bone – representing all of the taxa present – had been broken while fresh, and about 10 percent of the bone was burned. Bison was the most common taxon, comprising at least 95 percent of the sample and including at least nine adult and juvenile individuals; dental data suggested a late summer kill. Other identified species included crow, pocket gopher, gray wolf and unidentified canid, and pronghorn. Worked bone tools included awls, an antelope metacarpal with a drilled and worn hole, fragments of an apparent bone tube, fragments with use polish or abrasion, a possible bone bead blank, miscellaneous cut fragments, and fragments that were deliberately flaked.

The lithic assemblage included flaking debris, block cores, platform cores, bipolar cores, and a substantial number of tools. The tools included projectile points, preforms, larger hafted bifaces, biface knives, end and side scrapers, spokeshaves, other unifaces, scraping planes, perforators, and utilized or retouched flakes. In general, the points had the following traits: convex blade edges that were widest at the proximal end of the blade; relatively shallow and wide side notches low on the blade; straight to concave bases; basal thinning and grinding; and sometimes grinding of the notches. Some appeared to be flake points that were made by minimal marginal retouch of a fortuitously shaped flake. Anvil stones and hammerstones were also recovered.

FaNq-32 (Gowen 2) Gowen 2 is a single-component site located in an alluvial terrace of the South Saskatchewan River, in the city of Saskatoon, south central Saskatchewan (Walker 1992); it was 70 meters west of Gowen 1 (above) but not part of the same deposit. The site was discovered in 1980 when it was exposed by heavy equipment borrowing fill in the city landfill. Subsequent excavation totaled 102 square meters, followed by mechanical stripping of overburden and monitoring for the presence of further artifact concentrations or features. The original site area is roughly estimated at 290 square meters, of which about 120 square meters remained when the presence of the deposits was noted. Features included hearths, a hearth with associated spill piles of burned and unburned bone fragments and FCR, an isolated spill pile, ash concentrations, and a bipolar reduction station with an anvil stone, hammerstone, and fractured pebbles. In addition, there were four postmolds in a rectangle, and adjacent to one postmold a small rectangular stain marked by small charcoal flecks; the function of this association is unclear. Radiocarbon dates one bone and charcoal ranged from 5665 +/- 110 to 6075 +/- 110 RCYBP. It is believed to represent a short term bison hunting camp and a temporary processing area.

Faunal remains were abundant at Gowen 2. Nearly all of the bone had been broken, and about 24 percent of the bone was burned. Bison was the most common taxon, representing at least 14 individuals. Other identified species included muskrat, chipmunk, gopher, deer mouse, fox, coyote, and possibly domestic dog. Worked bone included awls or perforators, knapping tools, a splinter stained with red ocher, a possible netting tool with a polished point and a notch, other fragments with use polish, and fragments with cutmarks unrelated to butchering.

The lithic assemblage included flaking debris and a substantial number of stone tools. The tools included projectile points, preforms, larger hafted bifaces, bifacial knives, end scrapers, other unifaces, scraper planes, perforators, and utilized or retouched flakes. The points resembled those from Gowen 1, although there was more variation at Gowen 2. Two specimens had deeply concave bases and an eared appearance, which were compared to Oxbow points. Several examples were flake points, generally conforming to the style described above, but plano-convex in cross section and executed by marginal retouch of a flake. Anvil stones and hammerstones were also recovered. 53

48CK303 (Hawken) Site 48CK303 is located in an arroyo in the Black Hills of northeastern Wyoming. The site was discovered by artifact hunters and came to the attention of archaeologists in 1971 (Frison et al. 1976; Frison 1978). Extant portions of the site were excavated in 1972. Three strata of bone beds were found, each resulting from a bison kill that followed driving a bison herd up the arroyo. The investigators felt that these represent a single cultural group and were produced over a short period of time. The faunal assemblage indicated winter kills in December or January for all three bone beds. Charcoal samples returned radiocarbon dates of 6470 +/- 140 and 6270 +/- 170 RCYBP.

Faunal remains were abundant, but preservation was not optimal. Bison was the predominant taxon, with nearly 100 individual animals identified. The bison were classified as B. occidentalis, although they showed some characteristics of B. antiquus. Bison long bones were absent or broken open, suggesting intensive harvest of marrow. Other taxa present included wolf, mule deer, and antelope. There was some indication that the wolf may have been butchered, and an antelope metatarsal had been fashioned into a knapping tool. Other worked bone included tibia choppers, humerus tools for skinning and fleshing, and knapping tools made from bison ribs.

The lithic assemblage consisted of flake tools, choppers, hammerstones, flaking debris and nearly 300 projectile points. Because the points were all closely related culturally and functionally, the investigators focused their analysis on a spectrum of changes in morphology over the use life of the points. Non-reworked points had elongated blades that tapered quickly to a very sharp tip, and a regular lenticular cross section along both axes. Blade edges were commonly parallel on the lower part of the blade, although they could be convex. Side notches were placed just above the base and formed sharp corners with the blade, and the base and notches were commonly ground. The basic form was compared to small lanceolate styles seen in terminal Late Paleoindian complexes in the region, but with the addition of notches. The elongated shape facilitated reworking of broken points, whether from basal, medial or distal fragments, and the opportunity to make repairs more than once. Their proposed morphological spectrum showed how the steps of reworking and repair produced considerable variation in morphology, of the sort that might be categorized as more than one point style in a more typical typological analysis.

In 1975 a similar bone bed, Hawken III, was discovered nearby. This was comparable in most features, except that the kill probably happened between mid-March and early April. A date for the site was somewhat more recent, at 6010 +/- 170 RCYBP.

48PA201 (Mummy Cave) Site 48PA201 is a deeply-stratified site in a rock shelter along the Shoshone River in the Absaroka Range of the Rocky Mountains, in northwestern Wyoming (Wedel et al. 1968; Husted and Edgar 2002). The site is remarkable for having 38 separate archaeological strata, covering a span of 9000 years. It is also remarkable for its preservation of organic materials, including worked wood, cordage and basketry, and even a desiccated body buried in a hide robe. The site is helpful in our analysis for recording the transition from Late Paleoindian to Early Archaic. The site was excavated from 1963 to 1966. A manuscript report was completed in 1968 and copies of that report circulated informally, until the report was published in 2002.

The Late Paleoindian component at Mummy Cave is represented in Layers 1 through 15, dating between about 9200 and 8000 RCYBP. These strata produced lanceolate points. In contrast, Layer 16 (7630 +/- 170) produced a variety of notched and stemmed points, and the investigators felt that some of these belonged to the early side-notched point tradition. In Layer 17, the point variability from Layer 16 disappears, and instead there is a set of very similar points which the investigators call Pahaska Site-Notched (Husted and Edgar 2002:47) and describe thus:

54

A medium-size to large point with straight to convex lateral edges and a straight to weakly concave base. The side notches are deep and wide. They may be oriented transversely or at a low angle toward the tip. Normally the notches tend to be parallel with the basal edge. Cross sections are biconvex. Flaking ranges from poor to excellent with a majority exhibiting fair to good workmanship. Lateral edges are retouched. Bases are bifacially thinned. The basal edges are lightly ground. Length 45.0 to 26.0 mm, width 23.0 to 18.5 mm, thickness 6.5 to 4.0 mm...

The investigators note that these points are "identical" to side-notched points from the Simonsen site (Husted and Edgar 2002:23). The same type is found in Layers 18 and 19. Layers 20 through 23 produced few artifacts, and the single point from these strata is damaged and its stylistic affiliations are not entirely clear. Beginning with Layer 24 (4490 +/- 140 RCYBP), a wider variety of point types are present, and the investigators do not call any of them Pahaska Side-Notched.

There are a series of radiocarbon dates for the Mummy Cave stratigraphy, and these allow an estimate for the chronology of the early side-notched complex at the site. Husted and Edgar (2002:23) evaluate the dates for Layers 16 to 24. To summarize, Late Paleoindian points persist to about 8000 RCYBP. The early side-notched points are first seen in Layer 16, which is dated to 7630 +/- 170 RCYBP. Dates for the various strata are ambiguous about the disappearance of the early side- notched points, indicating only that they are gone by 5800 +/- 120 (Layer 20) or perhaps earlier. However, the investigators also note that the side-notched point strata are all under the distinct layer of Mazama ash, which resulted from the eruption of a volcano that subsequently created Crater Lake in Oregon and has been dated to about 6800 to 6700 RCYBP (cf. Fryxell 1965; Mack et al. 1979; Hallett et al. 1997). Evidence from Greenland ice cores puts the eruption event at a calendar age of around 7627 +/- 150 cal BP for the (Zdanowicz et al. 1999).

It is interesting to note that Mummy Cave, as far as could be determined from the site report, did not produce any of the small unnotched points (Delong) that occur with the side-notched forms at the more eastern sites.

6.2.3 Middle Plains Archaic (5500 to 3200 RCYBP / 6300 to 3450 cal BP)

Projectile points are small to medium-sized and generally smaller and less well-made than the points from the Paleoindian period, and there is an increased use of local cherts. These points were most likely attached to atlatl darts rather than spears and were thrown with an atlatl. Diagnostic Middle Plains Archaic point types common to Minnesota are divided into Early and Late Phases (Gibbon 2012), with Early Phase points including Simonsen, Little Sioux, and Oxbow, while Late Phase types including McKean and Table Rock. Many of the Middle Archaic point types continued into the Late Archaic. Other artifacts that were developed in the later portion of this period, and more fully in the Late Archaic, include ground stone tools, such as grooved axes and mauls, manos, metates, and apparatus for the atlatl, including bannerstones, gorgets, and boat stones.

In the Middle Plains Archaic, the partly contemporaneous McKean and Oxbow complexes cover much of the geographic range of the Early Side-Notched Point complex. The Oxbow complex is a northern phenomenon that is found from southern to central Alberta, Saskatchewan, and Manitoba on the north to Montana on the west and North Dakota on the south east (see Green 1998). Walker (1992:144) proposed that Oxbow develops in situ on the northern Plains out of the Early Side- Notched Point complex. That hypothesis is based in part on proposed technological and stylistic similarities in lithic technology and on a proposed gradation of one type into the other through time. The Oxbow point has a concave base and basal ears that can be rounded in outline but often have a distinctive angular shape. The blades tend to be elongated and have concave edges, and they are generally similar in form to blades of the Early Side-Notched Point complex types. The Oxbow 55 people were intensive bison hunters. Their sites often provide rich ceremonial evidence, including cairns, stone alignments, multiple burials associated with dogs and red ocher, and grave goods obtained through long-distance contacts.

The chronology of the Oxbow complex is problematic. There are many dates a bit older than 5000 RCYBP, with a small number of dates as early as 5300 to 5600 RCYBP (see Green 1998:276-280). There are also numerous dates later than 4000 RCYBP, with a few as recent as 3400 or 3300 RCYBP. Eighmy and LaBelle (1996) evaluated the chronological distribution of 28 dates for the Oxbow complex. These have an uncalibrated range of 4700 to 3050 RCYBP. The distribution of the dates is bimodal. Two potential explanations (cf. Green 1998:35-36) of dates are discussed: 1) that the later dates reported by Eighmy and LaBelle (1996) belong to a separate Sandy Creek complex; and 2) that the early dates pertain to the more southern development of Oxbow and the later dates to a northern shift in Oxbow territory.

The McKean complex is found throughout much of the central to northern Plains and into the mountains to the west (Syms 1969; Webster 2004). Although McKean is also closely identified with bison hunting, there is evidence to suggest that regionally some McKean groups made use of a broader range of plant and animal resources. This is reflected in both faunal remains and food processing technologies. There is also potential evidence for intensive marrow extraction, bone grease production, and the manufacture of pemmican. McKean sites produce projectile points of three different styles: McKean Lanceolate, Duncan, and Hanna. Individual assemblages may contain one or multiple types. Since the McKean point type was first proposed, there has been no consensus on the relationship of the types. Some suggest that they represent the range of variation in one point type, some propose they are functionally distinct elements of a single tool kit, while others argue that McKean Lanceolate is early and Duncan-Hanna (one type) comes later.

The McKean complex is partly contemporaneous with Oxbow, although it is often considered to start later. Webster (2004:95-99) provides a useful review of radiocarbon dates for McKean sites and components. His review suggests that McKean originated in the Yellowstone River headwaters region shortly before 5000 RCYBP, spread from there and was well established throughout the Northern Plains by about 4400 RCYBP, and disappeared in most areas by about 3200 RCYBP. Eighmy and LaBelle (1996) evaluated the chronological distribution of 77 dates for the McKean complex. These have an uncalibrated range of 4150 to 3100 RCYBP, and a calibrated range of about 4850 to 3320 cal BP.

As noted, McKean overlaps Oxbow both geographically and chronologically, and a number of hypotheses have been advanced to explain their relationship. Webster (2004) provides a thorough review and evaluation of these. His preferred interpretation is a refinement of a model first proposed by Syms (1969). In this view, sometime around roughly 5500 RCYBP, Oxbow developed in the northernmost range of the Early Side-Notched Point complex, possibly as an in situ development coming out of the Early Side-Notched Point complex. Somewhat later, probably around 5000 RCYBP, McKean developed in the Great Basin and around the headwaters of the Yellowstone River. McKean began to expand fairly rapidly outward from the core area, and was apparently established in the Black Hills sometime around 4700 to 4500 RCYBP. From there McKean continued to spread both north and south. The theory proposes that Oxbow populations retreated to the north in the face of McKean expansion (see also Green 1998:35). The McKean complex persists over a wide territory until about 3200 RCYBP, and may have lasted somewhat longer in a Colorado mountain refugium.

56

Significant regional sites dating to the Middle Plains Archaic include:

21NR9 (Canning) Site 21NR9, on terrraces of the Red River south of the project area, includes a large surface scatter of Sandy Lake pottery and associated artifacts in a cultivated field. Subsurface testing revealed a one- meter-deep sealed Middle Archaic deposit with a discrete layer bison bone, fire stains or thin hearth features, and an assemblage of lithic debris and heavily worn tools with reworked Middle Archaic Hanna type projectile points from the McKean complex. Raw materials included mostly Swan River chert along with smaller amounts of Knife River Flint. Faunal evidence suggests a winter season occupation and bison processing camp.

32CS16 (Forness) Johnson et al. (1995) cite 32CS16, in Cass County, North Dakota as a typical Red River Valley Archaic site. It comprises a large scatter of lithic materials in a cultivated field along a series of oxbow channels that were likely formed as tributaries of the Red River changed course during the Holocene. A test excavation at 32CS16 found Oxbow and McKean complex projectile points but no cultural materials undisturbed below the plow zone, and no Woodland pottery. Forness suggests that such sites dating to the Middle Archaic and earlier in the Red River area may be located along relict river courses. Picha et al. (2016:9.31) state that: “Archaic base camps from all subperiods are suspected to be located along the Red and its major tributaries in areas providing ready access to dependable stores of food, water, fuel, and shelter (cf. Larson and Penny 1990; Michlovic 1986). Temporary camps can be expected in many of these same locations as well as along more ephemeral drainageways throughout the NRRSU.

21NL58 (Altman) Site 21NL58 is a multicomponent site along the eastern edge of the Minnesota River Valley near New Ulm in south-central Minnesota (Terrell et al. 2005). The site is in a secondary alluvial fan formed by an intermittent drainage way (Kolb 2005). The surface component of 21NL58 was originally discovered in 1988 (Justin and Radford 1990), and the deeper components were investigated in 2005 in connection with a proposed reconstruction of TH 14. Phase I investigations included pedestrian survey, coring to evaluate site stratigraphy and geomorphology, and monitored excavation of 10 backhoe trenches. These trenches were dug to investigate stratigraphy and geomorphology, and also to expose buried archaeological deposits. Artifacts were recovered during monitoring, and limited hand excavation recovered a further sample of lithic and faunal materials. This work identified buried surfaces at around 1.50 and 2.15 - 2.30 meters; no archaeological materials were recovered from the lower horizon. Phase II investigations included the excavation of 2 x 2 - meter excavation blocks in two backhoe trenches. Six dates on bone collagen ranged from 5410 +/- 50 to 5170 +/- 40 RCYBP.

The Phase II work yielded over 1,600 artifacts, including faunal remains, lithics, and FCR. No features were found. Artifacts were distributed horizontally throughout the blocks, although density varied. In Block I, artifacts were distributed vertically through about 35 cm of the sediment column, although the majority came from the lower 20 cm. It appears that artifacts may have continued below the maximum depth of excavation. In Block 2, artifacts were vertically distributed through about 60 cm of the sediment column, although artifact density was higher in the lower 35 cm. Within the lower 35 cm there were two peaks in artifact density, suggesting the possibility of multiple site occupations. The report notes a prevalence of rodent runs, which probably contributed to the movement of artifacts

About 90 percent of the 1,891 artifacts recovered during Phase I and II investigations were faunal remains. The majority of these were small, unidentifiable pieces. The pieces that could be identified were from bovid (bison), canid, cervid, turtle, bird, and fish. Most of the bone came from medium to 57 large mammals, with turtle, bird and fish each constituting less than one percent of the assemblage. Butchering, burning, and calcining were noted and one piece of worked bone was found. The report notes striations on the surface, and from a photo it appears that the bone may have been flaked.

The lithic assemblage totaled 146 artifacts. In addition to flaking debris, there were utilized flakes, scrapers, bifaces, a core, and the base of a side-notched point. It is described as having a thinned basal edge. Based on a photo in the report, the 21NL58 point base appears fairly similar to the point base found at 21CR155 in the deeper levels of Area H.

21RW53 (Jackpot Junction J-Squared) Site 21RW53 is a multicomponent site located on a colluvial slope on the southern edge of the Minnesota River Valley near Morton in southwestern Minnesota. The site was first investigated in 1988 and 1989 during survey associated with proposed a highway realignment and bridge replacement work (Justin and Peterson 1991). Shovel testing from the surface and in backhoe trenches identified artifacts and animal bone in two paleosols between 1.5 and 3.0 meters below the surface. A 1991 attempt to excavate formal test units was stymied by a high water table and the related collapse of backhoe trench and excavation unit walls (Mather and Nunnally 1994). In 1994, additional geomorphological research was conducted, and in 1995 Phase II excavation and further geomorphology was undertaken at the site (Hajic 1995; Bower et al. 1996). The 1995 Phase II excavation involved excavating in a backhoe trench, with hand excavation of four square meters beginning around 125 cmbs and continuing to about 265 cmbs.

The combined investigations established that the site landscape was built by three distinguishable episodes of sediment accretion, and that two lower sediment units contained archaeological components. The upper component was sparse and remains undated but seemed to resemble the lower component. The lower component yielded an abundant faunal sample and more modest numbers of lithic artifacts and FCR. No features were excavated, although a concentration of burnt bone, charcoal and FCR was noted in the wall of one excavation unit. Five dates were obtained for the lower component, four from charcoal and one from bone; these are, 5010 +/- 110, 4800 +/- 60, 4730 +/- 110, 4650 +/- 60, and 4580 +/- 60 RCYBP.

The combined sample of faunal remains totaled over 3,500 pieces. Identifiable taxa included bison, raccoon, unidentified mammal, pocket gopher, turtle, bird, fish, and clam. The bone was generally minimally-weathered, and butchering marks were noted on some pieces. Burned bone was abundant, but seemed to be confined to medium-sized mammals. The lithic artifact assemblage from the lower component included a core, a scraper, and flaking debris. The assemblage consisted mostly of locally available raw materials, although Grand Meadow Chert and Knife River Flint were also represented.

6.2.4 Late Archaic (4500 to 2500 RCYBP / 5000 to 2700 cal BP)

The Late Archaic in Minnesota begins around 5000 BP, as a cooler and moister climate ushered in the beginnings of today’s environmental conditions and biomes; a sequence that was completed by around 2500 BP. During this time, smaller lakes that had dried up during the Altithermal once again filled in. Forests in the northern and southeastern part of the state expanded as the prairie retreated west and south. These climatic and environmental changes led to the decrease of bison as the main game animal in reforested areas and the arrival of forest animals into their historical ranges. Bison continued to be a primary species across most of southern Minnesota, except in the southeast. The lifeways of the people during this period in Minnesota were marked by adaptations to the changing environmental conditions and to increasing influences from people and cultures in surrounding regions. It was a time of increasing population numbers and more diverse artifact assemblages, which together with the advent of communal burials and expanded exchange of exotic materials, indicate increased social complexity and changes in subsistence patterns. 58

Late Archaic cultural markers include side-notched and stemmed projectile point types along with groundstone tools (such as manos, matates, mauls, and axes), the use of communal burial sites without mounds (until the period of transition between Late Archaic and Early Woodland), and the increased presence of exotic raw materials (such as native copper and marine shell). Late Archaic point types are divided into regional clusters (Gibbon 2012:79), and the Northern Plains region includes the McKean and Oxbow Clusters, and most Late Archaic remains in the Northern Red River Valley are surface finds of copper artifacts and these diagnostic Oxbow or McKean points. As Gibbon notes, however, some Late Archaic point types overlap with the earlier Middle Archaic and later Initial Woodland occupations, and therefore the dating of Late Archaic occupations based solely on point typology is problematic.

Tools and ornamental items made of native copper are probably the most distinctive material remains of the Late Archaic period. Spear points, fish hooks, harpoons, woodworking tools, and other implements such as awls and drills are found along with bracelets, pendants, and beads. This aspect of the Late Archaic is known as the Old Copper Culture. Picha et al. (2016) note that Old Copper Culture artifacts have been reported from locations within and west of the NRRSU, suggesting the movement of people from the eastern Woodlands onto the prairie, presumably to hunt bison.

Gibbon and Anfinson (2008) use the term Proto-Horticulturalist to describe the addition of garden produce into the resource base of the Late Archaic period, suggesting that this indicates the beginning of a fundamental social transition, although not a heavy reliance on cultivated foods. Fragments of squash (Cucurbit pepo) recovered from a probable Late Archaic context at the King Coulee site near Winona on the Mississippi River is an example of this type of early horticulture from Minnesota (Perkl 1998).

The people during this period likely inhabited a series of relatively stable “base camps” that shifted during the year to access seasonal resources. A variety of smaller special activity areas, such as quarries, butchering, and extraction sites, radiated from these base camps. The communal burials that appear during the Late Archaic period may indicate increasing territoriality associated with greater settlement permanence. Highly-ornamented grave goods have been interpreted as an indication of increasing religious complexity; and the appearance of burial mounds at the transition of the Archaic/Woodland periods is perhaps an indication that it had become more important to make these territorial indicators more visible to outside populations.

The Late Archaic in the Minnesota segment of the Red River valley also includes Larter points, possible Pelican Lake points, and late occurrences of Oxbow points.

The Mooney site (21NR29) is a multi-component Plains Archaic-Woodland site located near 21NR9 on levees of the Red River. The Archaic component occupied a depth of 140-165 cmbs and included a few lithic flakes and a light scatter of bone that was identified as bison. No diagnostic materials were recovered, but charcoal provided a radiocarbon date of about 3400 RCYBP from the Archaic component (Michlovic 1987a). Most of the lithic assemblage comprised Knife River Flint, along smaller amounts of Swan River and unidentified cherts. Michlovic (1987a:59) notes that evidence from the few such Late Archaic sites in the region suggests that bison remained the primary subsistence resource on the Plains during the period when populations to the east were diversifying into a resource base more dependent on fish, woodland animals, and the collection of nuts and seeds. The Woodland component at 21NR29 contained Sandy Lake ware, other unnamed wares from the Northeastern Plains, and a variety of faunal remains that included large and small mammals, fish, birds, and turtle.

Other sites with confirmed or possible Late Archaic components include Pedersen (21LN2), Fox Lake (21MR2), and Mountain Lake (21CO2). Anfinson (1997) has proposed a Mountain Lake phase 59 dating from 5800 to 2200 BP, with 21CO2 as the type-site. Excavations at the site recovered small lanceolate points that more closely resemble forms to the east rather than to the west, and none of the distinctly northern-plains point types such as those of the McKean cluster were found at the site. In the prairies of southwestern Minnesota, the bison-centered lifeway continued until around AD 1000 with the advent of the Plains Village culture. The Pedersen site contained bison bone in all occupation levels, along with remains of other mammals, fish, and bird species. Bison bone is also the main component of the Archaic faunal assemblage at the Mountain Lake site.

6.3 Woodland Period and Plains Village (2500 to 300 BP / 500 BC to AD 1700*) * Because the conventional radiocarbon age and calibrated age are very similar from the Woodland Period and later, dates are expressed as BP and BC/AD following established conventions.

During the late Holocene, from the end of the Archaic period through the Woodland period, the climate and landscape continued to evolve. These changes are well-documented through an extensive series of a series of pollen core studies from across the state and by correlation with other research on vegetation and climate change across the continent. Arzigian (2008:8) summarizes the climate and landscape developments of the Woodland period in Minnesota:

Of greatest significance to the Woodland tradition is a period of cooler temperatures, the Sub-Boreal, that extended through the Early and Middle Woodland periods and was followed by the warmer Neo-Atlantic and Pacific periods, and then the cooler, moister Little Ice Age from about AD 1550 until 1915. During these broader climatic shifts and more local changes, the most noticeable changes would have been the local expansion or contraction of the prairie-forest ecotone and the prairie bison herds. Changes in local lake levels would have affected settlement patterns adjacent to the lakes, with some lakes drying up completely. Fires would have caused changes in the composition and distribution of forests as well as expansion of shrublands and savannas. Fire frequency would have been affected by local and regional climatic conditions, and possibly also by the human population. Starting about AD 1550, the Big Woods expanded at the expense of prairies as a result of changes in fire frequency in the cooler, moister Little Ice Age climate.

The Woodland period for much of Minnesota is most closely associated with cultures to the south and east, and has traditionally been divided into the Early, Middle, Transitional, and Late Woodland traditions (Arzigian 2008) - although Gibbon (2012) prefers to divide the Woodland Period into Initial and Terminal periods in all but the southeastern corner of the state. Until recently, it was thought that Minnesota had no Early Woodland period preceding the cultures that made Malmo ceramics. Although there is still some controversy surrounding dates (i.e., Gibbon and Anfinson 2008), recent research suggests that the Brainerd complex does represent the Early Woodland period in Minnesota (Arzigian 2008).

Closer to the project area, peoples of the Woodland period were likely more closely related to the Plains Woodland traditions, which overlap the time frames of the more eastern traditions and are also divided into Early, Middle, and Late periods. Picha et al. (2016:9.33-34) describe the environmental setting and cultural chronology of the Woodland period in the northern Red River valley:

[The] Environmental history in the Red River valley region during the Woodland period spans some 2,500 years and is a complicated, unresolved affair. Five major climatic episodes have been identified during the period between 500 BC and AD 1600: Sub-Atlantic, Scandic, Neo-Atlantic, Pacific, and Neo-Boreal (cf. Anfinson and Wright 1990). Conditions are suggested to have been generally mesic during the Sub-

60

Atlantic, turning xeric in the Scandic, improving again in Neo-Atlantic, and returning to periodically xeric early in the Pacific…

Early Woodland remains are not well known in the NRRSU. Some of the corner- notched and side-notched dart point forms found at sites could be indicators of Early Woodland use of the prairie-forest ecotone during improved conditions concordant with mesic intervals in the Sub-Atlantic episode. Likewise, early ceramics have not been readily identified…

Middle Woodland components are better represented by both mounds and campsites.… [although] ceramic remains akin to Laurel have not been commonly reported (cf. Lugenbeal 1978b; Stoltman 1973). Many of the reported earthen mounds in the NRRSU are thought to have begun to be used to inter the dead by early Middle Woodland times…

Transitional Middle-Late Woodland (ca. AD 600-900) remains from the Red River valley have been grouped in the Arvilla complex by Johnson (1973). The Arvilla Mounds (32GF1) serve as the namesake for the complex… Contemporaneous Late Woodland Blackduck and Sandy Lake ceramics can be expected to occur at sites in the northern and southern ends of the Study Unit on an AD 1000-1600 time level.

Picha et al. (2016) note that Woodland settlement behavior in the Northern Red River valley in North Dakota has not been investigated in great detail. Habitation sites are expected on glacial Lake Agassiz beaches and in terrace settings along all the major tributaries of the Red River. While not researched to the same extent as are the Woodland cultures farther to the east, subsistence during the Woodland period in the Northern Red River valley seems to have emphasized bison, along with smaller amounts of aquatic resources, such as fish and mussels, and plant foods.

Plains Woodland ceramic and lithic technologies have likewise been studied primarily in relation to Woodland developments to the east (in Minnesota) and to the north (in Manitoba). One of the issues to be researched for the NRRSU is how key traits of the Woodland period were derived from these neighboring regions. Laurel and Malmo ceramic wares are known in the Early to Middle Plains Woodland periods, while Kathio, St. Croix, Blackduck, and Sandy Lake wares are associated with the Late Woodland and subsequent Plains Village periods. Middle Woodland Besant side-notched and St. Croix corner-notched points have been reported from sites throughout the upper Red River valley. Late Woodland projectile points include small side-notched and triangular forms. Bone and shell artifacts are uncommon, although bone and antler points are known from Laurel assemblages to the northeast in the Arvilla Mounds (32GF1).

Trade networks in the Plains Woodland region are indicated by the presence of non-local or exotic materials and artifacts. Picha et al. (2016:9.37) state that “[r]egional Middle Woodland groups are posited to have been linked to an intersocietal network of exchange referred to as the Hopewell Interaction Sphere (HIS). Various materials such as KRF, obsidian, copper, and freshwater and marine shell circulated through the midcontinent as part of the HIS exchange…. [C]orner-notched points recovered from a cultural layer at Mummy Cave in northwestern Wyoming dated to ca. 87 BC [coincide] with the zenith of the HIS activity in the Midwest. These specimens are similar to some regional Middle Woodland forms”.

Johnson et al. (1995:27-28) note that:

The single Woodland archaeological complex presently identified most closely with the Red River Valley is the Arvilla burial mound complex. [Arvilla] … is also known as the Red 61

River Aspect. Arvilla sites are burial mounds located mostly along the eastern Lake Agassiz beach ridges, although the Arvilla site itself is on the western beach deposits in Grand Forks county, North Dakota. The mounds are usually long, linear earthen structures covering single or multiple burials in pits dug below the original ground surface. The burials are often accompanied by grave goods, particularly ceramic vessels, shell beads, and other ornamental artifacts [have] been dated to the Middle-Late Woodland transition, about AD. 600 - 800…. St. Croix pottery is often present at these sites and this ceramic ware has been dated to approximately this period…. St. Croix pottery is present in small quantities along the Red River including at site 21NR28…. Syms (1982) has disputed the early dating of Arvilla and has argued that the mortuary goods associated with the burials reflect Mississippian influences, particularly in the presence of shell from the Gulf Coast. At least some of the Arvilla burials, under this view, should be dated to perhaps AD. 1200 -1400.

Site 32CS29 (Dahnke-Reinke) is the oldest known Plains Woodland site in the Red River Valley (Thompson 1990). The site is located on an alluvial terrace at the confluence of the Red and Sheyenne rivers north of Fargo, North Dakota and dates from between 250 BC to AD. 90. A Late Woodland Sandy Lake component overlies a well-preserved Middle Woodland component containing ceramics similar to Fox Lake and Sonota in addition to Knife River Flint artifacts, at least one piece of Great Lakes copper, and a single Gulf Coast shell bead. The faunal assemblage comprises primarily bison, with smaller quantities of fish and other riverine species.

The Lake Bronson site (21KT1) in Kittson County, Minnesota is located on a Lake Agassiz beach ridge and contains a Middle Woodland Laurel component associated with bison remains. The Lockport site (EaLf-l) in Manitoba contained a stratigraphic sequence of Middle to Late Woodland materials that were assigned to the equivalent of the Laurel and Malmo cultures.

Late Plains Woodland sites are fairly common in cultivated fields along the northern Red River and its tributaries. These sites often contain Sandy Lake pottery, characterized by minimal dowel and cordwrapped dowel net-impressions on the lip and interior rim, cordmarked vessel surfaces, and shell temper (Michlovic 1982).

As noted above, knowledge of the Prairie Woodland period is limited in comparison to the more eastern-focused Woodland occupations in the rest of Minnesota, and much of the analysis of the Prairie Woodland period is made in relation to eastern influences. The following discussion therefore focuses on Woodland period research in other parts of Minnesota.

6.3.1 Early Woodland - Brainerd Complex

The Brainerd Complex of central and northern Minnesota, also known as the Elk Lake culture, spanned the period of approximately 3000 to 1600 BP, and it is generally acknowledged to represent the Early Woodland period in Minnesota. The extent of this time span comes from a series of AMS dates obtained in the 1990s, mostly from residues on ceramics, and from limited stratigraphic evidence. The relative age of the Brainerd Complex is also indicated by the position of many sites on beach ridges associated with higher lake levels from roughly 3000 to 2000 BP. Prior to AMS dating, the Brainerd Complex in northern Minnesota was thought to date to approximately 1400 to 1200 BP in the Middle Woodland period, following Malmo (Arzigian 2008). In the Mississippi Headwaters area, the complex was thought to predate Sandy Lake and possibly Blackduck ceramics and extend back into the early Middle Woodland period.

The net-impressed pottery associated with the Brainerd Complex is found across a wide area extending from north-central Minnesota west into the Plains, and north and west to Lake Winnipeg and Montana (Hohman-Caine and Goltz 1995). The landscape in central Minnesota during this 62 period was a mix of oak and pine forests with oak savannas that provided habitat for elk and bison. Wild rice was also migrating from southern and central Minnesota into the lakes of northern Minnesota by roughly the beginning of the Brainerd Complex. Arzigian (2008) reports that 169 Brainerd Complex sites have been recorded in Minnesota, with most in MnSHPO regions 5c (Central Lakes Coniferous Central) and 4w (Central Lakes Deciduous West).

Despite the many years of excavation and research, Brainerd Complex settlement and subsistence patterns remain poorly understood, due in part to the extensive mixing of archaeological components in the shallow soils of the region. The complex seems to be transitional between the basic hunter- gatherer traits of the Late Archaic period and the more complex subsistence patterns of the Late Woodland. Gibbon and Anfinson (2008) suggest that Brainerd may represent the first appearance of ceramics in the region at the end of the Late Archaic period. It may also represent a spreading of Woodland peoples from the south into the northwest, as they followed the resources of the retreating prairie landscape. Increasing cultural complexity during the period is indicated by burial mounds that became more common during this period and by the presence of Knife River flint at many sites, demonstrating that trade networks extended for a significant distance from the home region.

Although most Brainerd Complex sites lack projectile points or contain points from mixed stratigraphic settings, Hohman-Caine and Goltz (1995) report that a few types have been recovered from well-defined components. With the exception of one small arrow-sized point, all are medium- sized, corner-notched, expanding stemmed or straight stemmed dart point varieties similar to Late Archaic or Early Woodland types such as Pelican Lake, Duncan, Oxbow, Hanna, and Snyders. The most common raw materials are Swan River chert, Tongue River silica, and Knife River flint. Other lithic tools recorded at Brainerd Complex sites include small, medium, and large scrapers, hammerstones, utilized flakes, early- and late-stage bifaces, chopping tools, square to rectangular wedges/chisels believed to have been used for woodworking (sometimes confused with bipolar cores), and debris associated with all stages of lithic reduction. Lower-quality materials such as friable quartz and siltstone are used for many of these other tools.

Sparse subsistence information has been recovered from Brainerd Complex sites. Faunal material, floral remains, and phytoliths from charred residue on ceramics have been recovered from Brainerd components at a few sites. These materials indicate that Brainerd Complex people relied on the hunting of medium to large mammals and that they gathered a wide range of plant foods, including starchy seeds, fruits, and nuts. The subsistence strategies suggested by these foods is consistent with the proposal that the Brainerd Complex is essentially a late Archaic adaptation, with the addition of pottery and wild rice (Gibbon and Anfinson 2008).

Two sites with well-defined Brainerd components, LaSalle Creek (21HB26) and Shingobee Island (21CA28), contained faunal remains of elk, deer, dog, bison, possible caribou, turtle, beaver, otter, and fish. Plant resources recovered from Roosevelt Lake (21CA184) included seeds from edible weedy plants such as chenopod, along with fruit seeds from the raspberry or strawberry family, and shell fragments from acorns and hazelnuts. It has been suggested that early Brainerd vessels were used for starchy non-grass seeds such as chenopods and amaranth, or possibly for nut-oil processing, while by the end of the Brainerd occupation, wild rice had been introduced, although there was no stylistic change in Brainerd ware during this sequence.

The relationships of the Brainerd Complex to other cultures also remain unclear. As Arzigian (2008) explains, Brainerd and Laurel occupations have been found together at sites such as Third River Borrow Pit (21IC176), but it is unclear whether one precedes the other or whether they overlap significantly. Likewise, little is understood of possible relationships between Brainerd ware and other ceramic types such as Malmo and LaMoille in Minnesota, Black Sand in Illinois, Prairie ware in 63

Wisconsin, and ceramics of other Early and Middle Woodland cultures in the region.

6.3.2 Middle Woodland Havana-Related Complex

The Havana-related complex in Minnesota, which spans the period of roughly 2200 to 1700 BP, is represented in Minnesota by the Malmo context. The generally-accepted date range for Malmo is derived from relatively few radiometric dates at mound sites and residue on ceramics, and from inferred relationships to the Havana culture in Illinois. Arzigian (2008) reports that a few dates for the complex range from 2690 to 1400 BP. Dating of the complex in Minnesota, based on relationships to the Havana culture in other areas, is problematic as there is no reason to believe that the complex in Minnesota ended at the same time that it disappeared elsewhere in the Midwest.

Malmo is defined by grit-tempered ceramics, small notched- and stemmed projectile points, and the predominant use of local lithic raw materials supplemented on occasion by exotic raw materials such as obsidian, Hixton silicified sandstone, and Knife River flint. Havana-related peoples used both mound and non-mound burials. Subsistence and settlement practices are not well-understood, but there seems to have been a pattern of seasonal mobility, with larger summer villages and dispersed winter camps. Havana-related sites in Minnesota are generally located in mixed habitats, including riverine settings, areas of wet prairies, and oak openings that are often bordered by mixed deciduous forest.

Malmo ceramics are widely distributed around the Mille Lacs area from Mille Lacs Lake west to the Gull Lake area and from Ottertail County out into the prairies. Sites with Malmo ceramics, including some large burial mounds, are also found in Isanti County. In comparison to the other Havana-related contexts, however, Malmo sites have relatively few ceramics that show general similarities to Hopewell, and researchers in Minnesota prefer to use the term Havana when discussing regional cultural developments in order to avoid a focus on the dramatic expressions of Hopewellian culture during this period.

There is only limited information on Havana-related lithics and lithic assemblages. In general, the lithics from the complex include medium to large, corner-notched projectile points such as Snyders, Norton, and Manker along with small blades and blade cores. Very little evidence of subsistence has been recovered from excavations at Havana-related sites in Minnesota and therefore much of the understanding of Malmo subsistence comes from analogies with cultures in other regions that followed a general hunting-gathering lifeway. Calcined bone fragments from a bear’s paw were recovered from the Van Grinsven site, along with turtle shell. Calcined deer bones, bird bones, and mussel shell are reported from the Brower site. One stone net-weight was also present in that site assemblage. The Gull Lake Dam site contains a large faunal assemblage including bear, moose, and domestic dog. Evidence of floral subsistence has also been recovered during recent excavations and studies, including an analysis of phytoliths in food residues from two Malmo vessels (Thompson 2000). One sample was found to be indicative of wild rice, but the other contained a phytolith assemblage characteristic of maize.

6.3.3 The Central Minnesota Transitional Woodland Complex

The Central Minnesota Transitional Woodland complex, which spans the period of roughly AD 300 to AD 1000, marks the transition between Middle Woodland (Malmo) and Late Woodland (Blackduck-Kathio) with notable shifts in technology, interregional interaction, mortuary practices, subsistence, and settlement. Dating of the complex has been based on a relatively small amount of stratigraphic information, radiocarbon dating, and on similarities to other transitional Woodland sherds, such as Onamia-like ceramics from southwestern Wisconsin.

64

Sites from this complex are concentrated in the Mille Lacs area and along the Snake River drainage of east-central Minnesota. Closely related ceramics and lifeways are found in adjacent areas from northwestern Wisconsin and northeastern Minnesota into the eastern Dakotas. There are connections between the complex and the later Blackduck-Kathio complex in central Minnesota and to contemporaneous cultures in southwestern Minnesota such as Lake Benton. Arvilla burials have been linked to this complex through the presence of St. Croix and Onamia pottery in burial mounds.

Little is known about lithic use or technology in the Central Minnesota Transitional Woodland complex. The Q-Pattern, reflecting heavy reliance on quartz, continues to be prevalent throughout this period (Bakken 2000), suggesting continuity with the preceding Havanna-Malmo Period. Projectile points found with Onamia ceramics are predominantly side-notched, sometimes described as similar to Prairie Side-Notched. Unnotched triangular points have also been recovered.

Despite limited data, Arzigian (2008) outlines a model of subsistence strategy changes proposed by researchers for the Central Minnesota Transitional Woodland complex. An increasing availability of wild rice and local game animals allowed for the sustenance of larger populations than could be achieved by traditional hunting and gathering lifeways. Hohman-Caine (1983) suggests that the rapid stylistic changes that culminated in the development of St. Croix ceramics are related to this increased population density and the shift from diffuse to focused subsistence patterns. The distinctive style of St. Croix ceramics found across a very wide area may also have played a role in maintaining social unity as increasingly large populations began to segment.

Linear earthworks found from the Pine City area east of Mille Lacs to the western prairies and in the Red River Valley are loosely associated with St. Croix ceramics through Johnson’s (1973) definition of the Arvilla Complex. Conical mounds are also known from the complex. Little else is known of the temporal, spatial, and cultural relationships between the Central Minnesota Transitional Woodland complex and earlier, later, and contemporary cultures. Arzigian (2008) discusses possible relationships between St. Croix and early Blackduck bossed ceramics and between Onamia and Lake Benton ceramics to the west. The landscape position of sites is a constant, as all sites are situated on high ground in proximity to water.

6.3.4 Late Woodland Blackduck-Kathio Complex

The Blackduck-Kathio complex (ca. AD 600 to AD 1100) evolved throughout northern Minnesota early in the Late Woodland period. The “Kathio” designation in central Minnesota refers to both a ceramic series and to the Mille Lacs Kathio Phase, which has been interpreted to include earlier Onamia ceramics. Blackduck, Kathio, and related ceramic types are thin-walled and globular and they display a considerable amount of stylistic variation across the chronological and geographical range of the complex. No clear typologies have been established and Arzigian (2008:110) cautions that “…the definition of Blackduck ware has been modified by many researchers and there seems to have been little cross-referencing between researchers….”

Few single-component Blackduck-Kathio sites have been excavated, and therefore many aspects of the complex are poorly-understood. Sites are generally larger and denser compared to earlier complexes in the region, suggesting increased population sizes. The people of the time likely followed an “Ojibwe settlement pattern” (Gibbon and Anfinson 2008) in which populations gathered into large groups in the summer at fishing and ricing camps and then divided into small groups in the winter at hunting camps. Sturgeon spawning grounds associated with the Rainy River may have been significant gathering places during the Blackduck period. Bison hunting in the adjacent prairies was also likely a part of the Blackduck subsistence pattern. Blackduck ceramics occur sporadically in the Minnesota portion of the Red River valley, but Sandy Lake ware is much more common (Arzigian 2008:209-211; Johnson et al. 1995:31). These seasonal migration patterns may have centered on 65 large summer villages that were occupied in the same locations every year. Very sparse evidence of possible house structures gives a tentative indication that sedentism was increasing during the Blackduck period. An increase in the demand for storable foods such as wild rice to feed increasingly large populations may have led to more complex social organization and territorial control. Burials in the Blackduck period are characterized by small mounds built close to habitation areas. Grave goods are present in some Blackduck mounds around the Rainy River, and some burials were intrusions into earlier Laurel mounds.

Lithic technology associated with the Blackduck-Kathio complex includes unnotched and notched triangular points that are typical of other Late Woodland cultures. Prairie Side-Notched and Plains Side-Notched points have been recovered from sites in the eastern Dakotas. Other artifacts recovered from sites with Blackduck pottery include unnotched triangular projectile points; end- and side- scrapers; knives; drills; steatite and clay pipes, bone awls or needles; mammal bone harpoons and spatulas; native copper fishhooks, gorges, awls, and beads; birch bark containers; and red ocher. Many of these artifacts came from multi-component sites, and therefore the specific associations with the Blackduck component cannot always be assured.

Blackduck-Kathio subsistence data is poorly-defined. Arzigian (2008) explains that seasonal exploitation of flora and fauna is assumed to have been a significant part of a subsistence strategy that also likely included the use of wild rice. She cautions, however, that the widely-accepted Blackduck- Kathio emphasis on wild rice is based to a large degree on site location, while the actual presence of wild rice remains is obscured by poor or mixed stratigraphy and a frequent lack of fine recovery techniques. Another complication she cites (2008:116) is that “Blackduck components interpreted as ricing sites are often in the same areas as historic ricing features that are probably being misidentified as precontact features.” Small mammals such as beaver appear to have been hunted at fishing sites while large mammals such as deer, elk, moose, and bison were hunted at other locations.

The social, economic, and political organization of the Blackduck-Kathio complex is inferred mainly from subsistence and settlement patterns. Communal activities such as bison hunting and mound building suggest that relatively large groups came together seasonally on a repeated basis. Evidence for interregional trade is found in the presence of exotic lithic materials such as Knife River flint and Grand Meadow chert and in the presence of ceramics that appear to have been part of larger regional patterns. Arzigian (2008) points out, however, that exotic artifacts are not unique to this complex and that some materials may have persisted from earlier periods in more localized systems of exchange.

The stratigraphic relationships between Blackduck and other complexes are fairly well established, with Blackduck ceramics found above Laurel and beneath Sandy Lake, but the relationships between these and other complexes is not fully understood. Arzigian (2008:119) summarizes some of the main theories regarding these relationships:

Lugenbeal (1979:24) sees a close relationship between Laurel and Blackduck, though Anfinson (in Lugenbeal 1979:24) argues for a closer relationship between Blackduck and Onamia–St. Croix. At the other end, Lugenbeal argues for closest affinity between Late Blackduck and the Selkirk ceramics of Ontario and Manitoba.…[Other researchers have suggested that] Laurel and Blackduck pottery-producing peoples were contemporaneous (Anderson 1979; Dawson 1983; Lenius and Olinyk 1990), suggesting that Blackduck people moved into the area from the south (Dawson 1983). Alternatively, Blackduck pottery might have originated in the Princess Point culture from the western end of Lake Ontario (Buchner et al. 1983).

66

6.3.5 Late Woodland Psinomani Complex

Psinomani (Dakota for “wild rice gatherer,” formerly called Wanikan) is a widespread archaeological complex that generally contains small triangular projectile points and Sandy Lake ceramics. The complex dates from approximately AD 1100 to the beginning of the historic period (ca. AD 1750), in which Sandy Lake ceramics were found in association with French trade goods. Some Psinomani components contain Ogechie ceramics (a locally-made variant of Orr phase Oneota ware). Sandy Lake ceramics have a wide geographical distribution, with the center of the culture in the Mississippi River headwaters region. Sandy Lake ceramics are commonly found in the Red River Valley along the northern portion of the Red River and its tributaries, with the Mooney site representing one of the best documented Sandy Lake sites (Johnson et al. 1995:31). Sandy Lake ceramics in the Minnesota portion of the Red River Valley, and particularly those near the Mooney site, are noted as being almost exclusively shell tempered with a cordmarked surface (Johnson et al. 1995:31). Bison were the primary subsistence remains at the Mooney site.

In general, Sandy Lake ceramics use shell or grit for temper, seldom have surface decoration, and are either cord-marked, smooth, or stamped with grooved paddles on the exterior. In the Red River Valley, the ceramics are characterized by cordmarked vessel surfaces, shell temper, and minimal dowel and cordwrapped dowel impressions on the lip and interior rim (Johnson et al. 1995). Projectile points from the complex are small, unnotched, and triangular in form and are often made of quartz.

Arzigian (2008) describes Sandy Lake ware (after Birk 1979b): “Rims are generally straight, incurved or out-flaring, with flat or rounded cross-sections. Rim body junctures rarely form an abrupt angle or shoulder on the exterior, though the interior junction is often marked by a pronounced thickening. Both mortuary and utilitarian vessels are globular, and somewhat squat (vessel orifices are proportionately wide when compared to body width).” Decoration, while uncommon, includes lip notching, interior punctates, and interior lip notching. The notches vary from a saw-toothed to clustered or evenly spaced impressions created by a variety of tools. Surface treatment includes both vertical cordmarking (Sandy Lake Corded) and plain or smoothed-over cord- marked (Sandy Lake Smooth). Other surface treatments include simple- and check-stamped exteriors. Vessel thicknesses range from three to seven millimeters. Sandy Lake ware used both shell and grit temper, and while there do not appear to be any diagnostic correlations between temper type and other variations within the ware, Arzigian (2008) describes experimental work by Budak (1991) demonstrating that shell-tempered pottery would have been more durable and watertight and would have transferred heat with more efficiency than grit-tempered pottery. .

The origin and extent of Sandy Lake ware and the Psinomani Complex remain uncertain. Gibbon and Anfinson (2008) note that ceramics of very similar style are found throughout the northern Midwest, suggesting that Psinomani is part of a widespread cultural adaptation. It is not clear whether the Psinomani Complex represents the arrival into the region of new populations with new technologies or whether it represents the in situ development of existing populations. Changes through time within Psinomani are not well-documented. Ogechie ceramics appear to come from later components, although some tentative Ogechie radiometric dates from pot residues are fairly early in the sequence. Psinomani sites have been recorded across central and northern Minnesota, west to the Red River Valley and eastern North Dakota, east to the St. Croix River in Wisconsin, and north to the Rainy River and into Manitoba and Ontario. Psinomani sites seem to be located preferentially in areas such as the Mille Lacs Locality, which is near the forest-grassland ecotone on the western edge of the mixed conifer-hardwood forest.

Psinomani subsistence patterns included the gathering of wild rice, hunting of bison and other mammals, and the use of fish and other resources from lakes and rivers. These patterns suggest that 67 the Psinomani peoples engaged in fairly intensive resource extraction from relatively contained areas along with seasonal hunting expeditions. In the more northerly extents of the complex, the use of prairie resources may have been limited, and there appears to have been a much more significant dependence on wild ricing. Peoples of the Psinomani complex traded for corn with Oneota communities to the south and east, although there is no evidence that they grew corn or stored it in the deep storage pits typical of Oneota or Plains Village sites.

Trade connections with Plains Village peoples to the west included seasonal bison hunts and the acquisition of Knife River Flint among other resources. Psinomani peoples buried their dead in both mound and non-mound contexts, although despite increasing population sizes and social complexity, there are few known burials from the period and the scale of mound-building seems to have diminished from those of the preceding cultures in the area.

The range of Ogechie pottery extends from central Minnesota west to the Red River, and Arzigian (2008) cites Michlovic (1983) in arguing that shared ceramic traits, a common environment, chronological overlap, and a likely relationship to historically known Siouan speakers, suggest a link between Sandy Lake and at least one variant of Oneota, to which he suggests the name “Sandyota.” This type is probably under-reported, with assemblages being identified as Oneota rather than the local variant, Ogechie. The Mille Lacs Lake area is likely near the northwestern frontier of the Oneota world, although other Oneota-like archaeological components are known from as far north as Manitoba.

Few discrete Psinomani components have been excavated, and therefore it is difficult to describe a typical lithic assemblage. In general, lithics at Psinomani sites are typical of other Late Woodland and Oneota sites. Some sites appear to have been focused on tool use and reworking, while others involve a more full range of lithic manufacture. Assemblages include small triangular projectile points, end- and side-scrapers, bifaces, wedges or bipolar tools, cores, utilized flakes, hammerstones, anvil stones, double-pointed and bifacially flaked knives, sandstone arrow shaft abraders, milling stones, celts, and pipestone elbow pipes. Raw materials include locally-available quartz, Prairie du Chien chert, Knife Lake siltstone, and Tongue River silica. Exotic materials include Burlington chert, Hixton orthoquartzite, and Knife River Flint.

The historic Dakota people of central Minnesota traveled widely and had many trade connections, suggesting that similar networks may have been present during the Psinomani period (Arzigian 2008). The distribution of Sandy Lake pottery and Knife River flint potentially represents the area of such interaction and movement. Psinomani represents the development of a tribal society with relatively large populations and that grouped together in permanent villages clustered around significant lakes, in contrast to the band-level social organization of earlier cultural groups. These larger and more permanent population centers likely relied on an intensified utilization of grain resources, such as wild rice, to supplement other resources that would have been depleted if they were the sole source of food. The uniformity of Sandy Lake ceramics throughout the region can be seen as an indicator of strong intergroup social relations; although Arzigian (2008) stresses that the Psinomani complex is not coterminous with Sandy Lake and Ogechie ceramics, particularly to the west, where Sandy Lake pottery is found on Plains Village sites. Debate continues on whether Psinomani reflects a population replacement or an in-situ development.

6.3.6 Northeastern Plains Village Period

The Northeastern Plains Village period (ca. AD 1200-1650) developed as conditions in the region became cooler, moister, and more amenable to horticulture during a climatic episode known as the Neo-Boreal, or “Little Ice Age”. Picha et al. (2016) note that during these cooler and moister conditions led to a significant increase in regional floral and faunal population and diversity. 68

Toom (2004:294) suggests that Northeastern Plains Village cultures evolved as an extension of the Cambria complex in Southwestern Minnesota, spreading eventually across the Knife River region of western North Dakota in what he recommends as a distinct taxonomic unit – the Northeastern Plains Village tradition (NEPV). NEPV peoples in the NRRSU are believed to have pursued hunter- gatherer/gardener lifeways. Picha et al. (2016) report that corn has been recovered as far north as the Lockport site in southern Manitoba by AD 1500.

NEPV ceramics were recovered from the Mooney site, and Johnson et al. (1995:37) summarizes the NEPV period as follows:

Archaeological complexes [associated] with the Plains Village period include Oneota, Cambria, Great Oasis, Mill Creek, Stutsman Focus, and Initial Middle Missouri…. The Plains Village period is characterized by a continued focus on bison procurement, small triangular projectile points (arrowheads), globular ceramic vessels with grit or crushed rock temper, smoothed, paddled and stamped surfaces, and decorations consisting of trailing, chevrons, and occasional zoomorphs (Oneota is an exception, with smooth surfaced, shell tempered pottery)…. Village sites, sometimes with fortification ditches are known from various portions of the Northern Plains, and these were occupied by populations sometimes numbering in the hundreds on a seasonal basis for years, or even decades (Picha and Gregg 1991a). Perhaps the most significant feature of the Plains Village period lifeway was the introduction of domestic plants. Maize, squash, sunflower and beans were grown in gardens, which we know from the ethnohistoric record were normally located on the river floodplains (Wilson 1917). Plains Village materials in the Red River Valley are most commonly in the form of potsherds from surface sites. These are classified as "village" ceramics if, they are smooth surfaced, grit tempered, and decorated with trailed or incised lines.

Picha et al. (2016) identify three known Plains Village sites (32GF79, 32PB164, and 32TR402) within the NRRSU, and suggest that some mound sites in the Red River basin are likely to contain burials from the period. NEPV ceramic wares are discussed and defined in Michlovic and Swenson (1998).

One of the most significant Plains Village site known in the Red River valley is 32CS101 (Shea), which dates to ca. A.D. 1470 and is located on a glacial Lake Agassiz beach ridge near the Maple River in the southern portion of the valley. The site is surrounded by a fortification ditch and contains Sandy Lake ware, bison bone, and small amounts of corn that are the first evidence of horticulture in the Red River region. A number of other fortified village sites associated with the Cambria phase have been identified in the area, including, 39R05, 21TR6, and 21TR12, but none of these has been thoroughly excavated. No fortified Plains Village encampments such as these are reported from the upper Red River valley.

6.4 Contact and Historic Period (AD 1700 to 1950)

6.4.1 Equestrian Nomadic Tradition

The Equestrian Nomadic tradition (AD 1700 to 1900) is defined by the lifeways that developed from the use of horses and guns in the protohistoric and early historic periods (Picha and Gregg 1991a). The presence of several tribes has been documented in the Northern Red River Valley region, including the Assiniboine, Dakota, Metis, Ojibwe, and Plains Cree. There are no known sites associated with this tradition, although it is likely that such sites exist.

69

6.4.2 Historic Period

The early history of the European settlement of the project area consisted of a variety of political changes in land possession. French and British traders controlled the Red River region during the 1600s and 1700s. Following the American Revolution, the land remained unorganized and then shifted politically as territories were drawn and redrawn. Settlement in the region relied upon the commerce concentrated on and along the Red River of the North. It was also dependent on the development of transportation networks. As oxcarts and steamboats gave way to railroads and automobiles, the twin cities of Breckenridge, Minnesota and Wahpeton, North Dakota would become the first “twin cities” along the Red.

6.4.3 The French Period

The French (1671-1763) were the first Europeans to establish a presence in the area. Pierre Gaultier de Varennes, Sieur de LaVerendrye, fur trader and explorer for the French government, was the first European in the Red River drainage in 1731 (Ritterbush 1991:22). Evidence of French fur trade activities are concentrated in northern Minnesota and North Dakota along the Canadian border, in the Mississippi Headwaters Region, and on major waterways near the prairie/forest edge. Gilman et al. (1979) suggest that the French, and later the British, did not establish trading posts in the southern stretches of the Red River Valley due to the warfare between the Eastern Dakota and the Ojibwe. At the end of the French and Indian War in 1760 when the British gained control of most of North America, the French largely abandoned fur trading activities in the Red River Valley to the British traders.

6.4.4 The British and American Periods

The British Period in the project area began in 1760 with the triumph of the British over the French in the French and Indian War and ended with the purchase of the Louisiana Territory in 1803 by the United States. Evidence of British activities is found throughout the Red River Valley. Gilman et al. (1979) suggests that independent British trader Robert Dickson established the first permanent British post in the Red River Valley at Lake Traverse in the 1790s. His trading network probably extended west to the James and Missouri rivers and to Devils Lake in the north. By 1797, the North West Company and the Hudson Bay Company were trading with the Ojibwe in the Red River Valley.

In 1797, Charles Chaboillez established a fur post at Pembina for the North West Company, which had been organized in 1783-84 (Kelsey 1951). The Red River subsequently became a valuable highway in a transportation network extending from the forest to the east, westward down tributaries of the Red River to Pembina. In a move to monopolize trade in the Red River Valley, the North West Company set up trading posts along the Red River. Alexander Henry, Jr. established one post on the Park River at its confluence with the Red River in 1800-01, and another at Grand Forks in 1800 at the junction of the Red Lake and Red rivers. The site of Henry’s post has not been relocated.

The North West Company traders at the beginning of the nineteenth century brought with them great quantities of liquor (Robinson 1966:57). The forts at Pembina and upstream at the Park River and at Grand Forks served as a base of operations for fur traders who hunted buffalo on the plains. The buffalo meat was dried into pemmican in order to supply the canoe brigades which carried furs to market (Robinson 1966:58).

70

The Hudson Bay Company continued to gain dominance in this region until the War of 1812, when the United States government began to force traders to either retreat to British territory or to switch citizenship. The period between 1803 and 1837 was characterized by the exploration and mapping of the region and the establishment of the first United States military outposts in North Dakota and Minnesota.

During this period, Thomas Douglas the Earl of Selkirk led the first European settlers into the region. These settlers started an agricultural colony at the confluence of the Pembina and the Red rivers and despite many hardships, the area became a seasonal settlement of the Selkirk colonists and a locus of fur trading activities on the lower Red River (Ritterbush 1991:33). In the 1840s, the American presence in this region was intensified by the establishment of posts by N.W. Kittson and Company (Ritterbush 1991:39).

6.4.5 Red River Trails

Initially used for fur trading, the Red River Oxcart Trails helped to end the Hudson Bay Company’s trading monopoly in the region by carrying commerce to St. Paul:

From the 1830s to 1871, the Red River Trails carried commerce from the Red River region to St. Paul. … Eventually the trails became the major overland thoroughfares for commerce and emigration, and helped to establish St. Paul as the center of trade and travel in the Upper Midwest. The trade founded by the trails ultimately led to the inauguration of wagon roads, steamship lines, and railroads. In essence, the trails laid the foundation for transportation and commerce throughout the northern and western sections of the state. The trails were eventually replaced by the railroads in the 1870s (Hess 1989:El).

The Red River oxcart trails in northwestern Minnesota followed the Lake Agassiz beaches on the eastern edge of the Red River Valley as the interior portion of the lake plain was difficult to traverse (Gilman et al. 1979). Oxcart trails also followed the west bank of the Red River through North Dakota. No oxcart trails are present in the project area. Although traders depended on the Red River oxcarts for the transportation of trading goods, the trails also served as the only land links between small settlements springing up along the Red River in anticipation of steamboat traffic. Anson Northrop built the first Red River steamboat in 1859, using parts from a Mississippi River steamer that were carried across the state. In 1862, the Minnesota Stage Company built the first Red River steamboat from the ground up, naming it the International. Because of the low water levels of the Red River, steamboat traffic was slow to develop (Sylvester 1988; Callan 1937).

6.4.6 Railroad Transportation

The role of the railroad in the settlement and development of commerce in the region is indisputable. The development of a flat-boat building business in the area of the Red River headwaters occurred simultaneously to and in anticipation of the railroad’s arrival. Flat-boats with a capacity of 10 to 50 tons easily navigated the shallow and meandering Red River carrying food, building materials, machinery, and household supplies north. Eventually, railroad traffic would usurp any commerce carried on the river. The railroad companies sold land adjacent to the tracks to encourage farmers to settle and use the railroad for transfer of their crops. The companies also built their own elevators to store wheat. The changes that this system incurred were great, and Bingham and Holcombe (1916) state in regard to the Red River Valley:

71

It is interesting to recall the stages of development of this great Valley. First, we find it a stock country, necessarily so on account of the natural conditions. Transportation was of the crudest kind, mostly by ox teams over poor roads and across unbridged streams. The early settler lived snugly along the river bank, well sheltered by tall timber, in which he had a cluster of log buildings, used as dwelling and stables. He had a large herd of fat, sleek cattle, fed exclusively on prairie hay, which had been gathered on the vast unsettled prairie with a hand scythe and pitchfork…

Then a change came. The prairie began to be settled and opportunities for raising stock began to diminish. Claim shanties began to appear on former meadows and pasture lands. Soon there were seen men driving two ox teams abreast before a breaking plow, turning down the green grass and turning up the black soil, making a field at the best rate of speed then known. Then some lands would be fenced, and soon the hitherto bleak expanse was dotted with shacks and well covered by fenced fields.

6.4.7 Local History

The historical settlement closest to the project area is the town of Nielsville, established in 1872 in what is considered to be one of the best farming sections of the Red River Valley. The area was also known to have been the site of battles between Dakota and Chippewa Indians at the nearby Little River. More families soon settled in the area and a post office was established in 1882, followed by a blacksmith shop in 1883. The Crookston-Moorhead branch of the Great Northern Railroad reached Nielsville in 1896, followed soon thereafter by a growing number of businesses such as a hotel, grain elevator, and a Lutheran church in 1897. The town continued to fill in with grocery and dry goods stores, a lumber yard, and livery stables. At its peak, Nielsville also contained a butcher shop, restaurant, shoemaker, and furniture store. As with many small towns that developed around railway stops, Nielsville began to lose vitality with the increasing use of private and commercial vehicles in the mid-twentieth century and the growth of road networks that allowed residents to more easily access the greater variety and more competitive prices of businesses in nearby larger centers. At the time of the current project, Nielsville housed two local bars and a small number of residences, along with the remains of some of the old buildings in various states of disrepair. Farms to the west of town, including within the project area, are large-scale corn and bean operations.

72

7. ENVIRONMENTAL BACKGROUND

7.1 Modern Environment

The CSAH 1 project area is located in the Red River Valley in northwestern Minnesota near the town of Nielsville, which is at the eastern end of the project and the Red River is at the western end. The primary land use in the survey area is agricultural, with soy beans and wheat being the dominant crops. The project area is in agricultural fields, except for a few rural residences and a small wooded strip on the river floodplain.

7.2 Glacial History

The most recent glacial activity in the region occurred during the Late Wisconsin at the end of the Pleistocene when much of the Upper Midwest was buried beneath glaciers. The Des Moines lobe covered the Red River Valley and much of western Minnesota extending to northern Iowa (Clayton and Moran 1982; Gilbertson 1990). The Des Moines lobe receded and advanced several times between 15,000 and 11,700 years B.P. The final retreat of the ice lobe occurred by 11,700 B.P. from the Big Stone moraine near the junction of North Dakota, South Dakota, and Minnesota. Glacial Lake Agassiz formed as the ice retreated and melt water became ponded between the glacier and moraines to the south.

7.3 Physiography

The project is located in the Red River Valley, which encompasses a broad area in northwestern Minnesota and northeastern North Dakota extending along a 40-mile-wide strip on each side of the Red River. The relatively flat, level terrain of the glacial lake plain characterizes the physiography of the Red River Valley. Primary topographic relief is along stream drainages, which are steeply incised into the lake plain, and beach ridges that are present along the margins of the lake. Moraines border the eastern portion of the Red River Valley in Minnesota. The western portion of the Red River Valley in North Dakota is bordered by the high ground of the Pembina Escarpment.

The Red River Valley is a preglacial lowland that formed over several million years (Bluemle 1977). Numerous glacial advances during the Pleistocene have followed this lowland as they advanced south (Bluemle 1977). The final glacial retreat produced a large glacial lake (Lake Agassiz) that filled the Red River Valley. Sediment deposited in the glacial lake created what is referred to as the Lake Agassiz plain (lake plain). The sediments are finely textured in the deep-water portions of the lake and become coarser near the lakeshores and beaches. As lake levels declined, the interior portion of the lake plain drained through numerous stream channels that ultimately flowed into the Red River.

The west end of the project is situated on a point bar of the Red River, which is inset into the lake plain. Terraces on the point bar formed as sediment was deposited along the river during floods. There appear to be multiple cycles of down cutting (erosion) and filling (deposition) along the Red River that have produced a series of terraces associated with each major fill event (Florin et al. 1998). The point bar morphology indicates a lateral time-transgressive movement from east to west as the river channel migrated during the Holocene. An oxbow (abandoned) channel of the Red River is on the south side of the project area, demonstrating the dynamic and sinuous nature of the river.

The project area is mapped as Holocene alluvium along the Red River, and away from the river it is mapped as lake-modified till (wave-planed, mantled with lake sediments too thin and patchy to map separately) from the Erskine moraine association of the Des Moines lobe (Hobbs and Goebel 1982). There appears to have been a former moraine in the Nielsville vicinity that was not buried as deeply

73 as other low lying areas by Glacial Lake Agassiz sediments, but no topographic remnants of this moraine exists as the topography consists of a nearly level lake plain.

The landscape of the Red River Valley, which is flat and devoid of lakes, influenced prehistoric site locations and transportation routes. Archaeological surveys indicate that sites are concentrated along riverways, and it is likely the rivers were used as transportation corridors on the relatively featureless lake plain.

7.4 Lake Agassiz

Lake Agassiz covered a vast area in the Red River Valley and formed approximately 12,000 years ago, as melt water from the retreating Des Moines lobe was ponded by a recessional moraine near Browns Valley, Minnesota (Ojakangas and Matsch 1982). The total extent of the lake encompassed an area of 320,000 km2 and extended into Minnesota, North Dakota, South Dakota, Ontario, Manitoba, and Saskatchewan. Several stages of the lake are indicated by a series of beach ridges that formed at the lake margin. Water depth was greater than 120 meters in some places. A sparse amount of rocks and cobbles were deposited on the lake plain. These materials were presumably transported on glacial ice rafts that floated out into the lake. As a result of erosional processes, rocks would have been concentrated along river bottoms, particularly at the margins of the lake plain, and beach ridges. Although their availability was very sparse on the lake plain, some of these materials were likely procured and used by prehistoric inhabitants of the region. Lake Agassiz had drained by approximately 9,000 years ago. The area probably became suitable for human habitation shortly after draining.

7.5 Hydrology

The relatively flat, level terrain of the lake plain in the vicinity of the project area is devoid of lakes. Several perennial and intermittent streams form the drainage system on the lake plain. Prairie marshes existed in several areas prior to the construction of drainage ditches. The primary water source near the project area is the Red River. An intermittent drainage, which is partially ditched, flows north of Nielsville. The nearest tributaries of the Red River are the March River, four miles to the south, and the Sandhill River, five miles to the north.

The Red River flows north into Lake Winnipeg in southeastern Manitoba, draining a vast area that encompasses several ecological zones. Wooded moraines occur on the eastern portion of the Red River Valley in Minnesota, and the drift prairie region borders the western portion of the Red River Valley in North Dakota. Aspen parklands and boreal forests border the river in Canada. The drainage network of the Red River provided a link between these ecological zones and was likely a route for the transmission of people, goods, and ideas during the prehistoric period.

7.6 Vegetation

The project area is located in the eastern portion of the North American Grasslands Biome (Barkley 1977). Geographically the area is within the Northeastern Plains, which extends from west-central Iowa northwards through western Minnesota to southern Manitoba and southeastern Saskatchewan. The western limit is bounded by the Missouri River in the Dakotas and the eastern limit is the prairie- forest border in Minnesota. Vegetation in the Red River Valley at the time of European contact consisted of tall grass prairie, which was dominated by big bluestem (Bluemle 1975). Prairie marshes were interspersed throughout the Lake Agassiz plain and included bluejoint grass, cordgrass, cattails, rushes, and sedges (Marschner 1974). Gallery forests existed along the major drainages and included a variety of trees such as elm, ash, box elder, and bur oak, as well as numerous other plant species (Picha and Gregg 1991a). 74

The original vegetation map indicates that the terraces of the Red River were wooded at the time of European settlement, and the lake plain was prairie and wet prairie (Marschner 1974). There is documented use of over 100 plant species by indigenous people in the Red River Valley (Michlovic 1987b). Trees in the river valleys played an important role in survival on the plains as they provided materials for fire and shelter (Ritterbush 1991). The wide variety of plant resources would have provided medicine, food, and utilitarian items to prehistoric people in the region.

Changes in vegetation have been recorded in pollen samples preserved in lake bottom sediments. Evidence from these samples shows that spruce forests dominated the landscape in the eastern Dakotas during the retreat of the glaciers approximately 12,000 years B.P. (Baronsky et al. 1987). Spruce forest was replaced by deciduous forest by 11,000 B.P. Prairie swiftly replaced the deciduous forests by around 10,000 years B.P. During the Altithermal (7,000 to 4,000 years B.P.), the climatic conditions became warmer and drier enabling the prairie to expand eastward into central Minnesota.

7.7 Fauna

A wide variety of animal species were present on the Northeastern Plains when Europeans first entered North Dakota and Minnesota (Ritterbush 1991; Robinson 1966). The range and abundance of species has been altered by the loss of natural habitat and hunting so that some species are no longer present. Early historical accounts of animals include antelope, elk, wolf, lynx, black bear, white- tailed and mule deer, bison, fisher, fox, marten, raccoon, beaver, muskrat, squirrel, woodchuck, rabbit, coyote, and prairie dog. Of particular interest to this project is the abundance of bison that were documented historically along the Red River (Glasrud 1982; Herrick 1892). The Red River Valley is along several major flyways, and a wide variety of waterfowl were present, including ducks, geese, and swans. There was also an abundance of fish in the large drainages of the Red River Valley. Remains of several of these species have been recovered from archaeological sites (Michlovic 1987b). The Canning site (21NR9) contains the butchered remains of at least 20 bison (Michlovic 1986). The wide variety of animal resources in the region would have provided a broad subsistence base to support prehistoric occupations. Bison are overwhelmingly the primary animal remain identified at archaeological sites in the general area (Florin et al. 2001; Florin and Wergin 2002; Lyon et al. 2003b; Michlovic 1986 and 1987b).

7.8 Soils

Soils on the lake plain are mapped as Bygland silty clay loam, Colvin silty clay loam, and Lindaas silty clay loam (Web Soil Survey 2017). These soils formed in silt loam and silty clay loam deposits of Glacial Lake Agassiz. These soils have received minimal sediment deposition during the Holocene, and there is no potential for deeply buried sites. Typical soil profiles consist of the following: Bygland series - Ap, Bt1, Bt2 BC, C1, and C2; Colvin series - Ap, Bkg1, Bkg2, and Cg; and Lindaas series - Ap, A, Bt, Bk, and Cg horizons.

Soils on the extensive upper terraces of the Red River are mapped as Bygland silty clay loam, which is not accurate, as these soils contain a thick sequence of Holocene alluvium and buried soils. Soils on the narrow lower terraces adjacent to the Red River are mapped as Fairdale silt loam and Fluvaquents (frequently flooded-Hapludolls complex). These soils formed in recent alluvium on low terraces and on flood plains. Geomorphological coring at the site provided a detailed record of the soils on the terraces, which is presented in Appendix A.

75

8. PHASE I FIELDWORK INTRODUCTION

8.1 Overview and Field Conditions

Archaeological fieldwork was conducted from June 6 to 23 and October 10 to November 2, 2016. Frank Florin was the principal investigator and field supervisor. The FCRS field crew included Mike Bradford, Val Chapman, Gregg Felber, Frank Koep, Dan Leja, Ryan Letterly, James Lindbeck, Amanda Peterson, Kevin Reider, Jeff Shapiro, Michael Straskowski, Bob Thompson, and Seth Thompson. The locations of the Phase I archaeological survey area and site 21PL109 are presented on a USGS 7.5’ quadrangle map in Figure 1. The field methods are described in Section 3.1. A map of the survey area depicting field conditions, site 21PL109, and shovel test locations is presented on air imagery in Figure 5.

A pedestrian survey was conducted for the entire project area, including an examination of the Red River cut bank along the floodplain, which was about four meters high. Shovel testing was conducted in three areas: 1) on the point bar terraces of the Red River; 2) on those portions of the lake plain near the river oxbow that had low surface visibility; and 3) and at locations of site 21PL109 where artifacts were found on the surface in agricultural fields on the lake plain. All areas not shovel tested had adequate surface visibility or low archaeological site potential. A total of 528 shovel tests were dug. Deep auger testing, discussed below in Section 8.2, was conducted on the floodplain (lowest terrace) for the bridge replacement where new pilings/piers will be placed.

Site 21PL109 was identified during the survey and evaluated. The site is discussed in detail in Section 9. A summary of the survey, including locations, field conditions, survey methods, and results is presented in Table 6, which is organized from east to west by legal location.

Table 6. Survey Summary CSAH 1 (East to West) - Field Conditions and Survey Results. Survey Field Conditions Distance Location Survey Method Results & % Surface Visibility (Meters) T147N Soybeans 90% 1050 R148W Negative Pedestrian Section 19 grass 0% 80 Soybeans 70% 600 T147N Wheat 30% 230 R148W Negative Pedestrian Yard (trees and grass) 0% 170 Section 30 Grass 0% 130 Negative Wheat 30% 960 Pedestrian T147N Soybeans 90%* 670 21PL109 R149W Fallow grass/woods* and yard 0% 250 Pedestrian & Section 24 Ditched/road cut 0% (grass); entire ROW; Shovel Test Negative 400 no access beyond ROW T147N Soybeans 70% 900 R149W Negative Ditched (grass) 0%; to 25 meters south of Pedestrian 100 Section 25 centerline; no access beyond ROW T147N Pedestrian & R149W 21PL109 Wheat 20-60% 600 Shovel Test Section 23 Pedestrian & Wheat 20-60% 680 T147N Shovel Test R149W 21PL109 Pedestrian, Shovel Section 26 Woods 0% (floodplain) 120 Test, and Deep Auger Test

76

*The landowner in the SW¼ of Section 24 denied access for subsurface testing in the bean field on the north side of the road and in the fallow grass/woods on the south side of the road. So, shovel testing in these areas was limited to the ROW.

8.2 Deep Auger Testing on Floodplain of the Red River

Deep auger testing was conducted on the floodplain of the Red River adjacent to the bridge (Figure 6). Six tests were dug in 15-meter intervals to a depth of 300 cmbs. No sites were identified, and the alluvial sediments are interpreted as being young, probably mostly post-settlement age, based on a few lines of evidence: 1) Tests 1, 2, 5 had cement, glass, a round nail, and road gravels at depths ranging from 65 to 130 cmbs; 2) Test 4 contained fill with plastic to 170 cmbs, and gravels found at 230 cmbs in Test 6 are likely road gravels; 3) no buried precontact soils were present below the historic materials, and the soils are poorly-developed; and 4) testing for this project on the floodplain on the North Dakota side of the river also revealed glass and cement to 90 cmbs and dense road gravel to 175 cmbs. Gravels from both sides of the river were found only on the down-river side of the road, leading the author to suspect that they may have eroded off of the road during high flood events that occurred prior to paving the road. Historic maps indicate that the road was present at least as early as 1916. A typical soil profile is presented in Table 7.

Table 7. Deep Auger Tests 1 to 6 Typical Profile on Floodplain of Red River. Depth Below Description Surface (cm) 0-60 Very dark grayish brown (2.5Y 3/2) silty clay; weak to massive, very fine angular structure Very dark gray (2.5Y 3/1) silty clay; moderate to strong very fine angular structure; buried 60-90 historic A horizon Very dark grayish brown or dark grayish brown (2.5Y 3/2 or 4/2) silty clay; moderate to 90-150 strong very fine angular structure, becoming weak with depth Dark grayish brown (2.5Y 4/2) silty clay; weak very fine angular structure, becoming 150-300 massive with depth

77

9. SITE 21PL109

9.1 Overview

Site 21PL109 is a large multicomponent Archaic, Woodland, and Northeastern Plains Village habitation along the Red River. Animal kill and processing (mostly bison) appears to be the primary site activity. Diagnostic artifacts and a suite of radiocarbon dates were obtained from the site, which is in T147N R49W S1/2 SE SW SE & S1/2 SE SE Section 23, SW SW SW Section 24, and NE NE NW & N1/2 NW NE & N1/2 NW NE NE Section 26 (Figure 1) and occupies an area of approximately 1300 by 120 meters, encompassing 27 acres. The UTM coordinates for the site are E660390 N5265965 (West End) and E661650 N5266170 (East End) (1983 NAD Zone 14 North). A map of the site on aerial imagery is presented in Figures 7 and 8. Photos of the site area are included in Figures 9 and 10. All figures, photos, and illustrations are at the end of this chapter.

Initially, the proposed CSAH 1 reconstruction plan shifted the road farther north of the oxbow and then angled it south of the existing alignment before angling back north to the bridge. However, to avoid site 21PL109 these plans were changed so that the CSAH 1 reconstruction would stay on the existing alignment and within the existing ROW at site 21PL109, thus avoiding adverse effects to the site.

9.2 Physical Setting

The site is located on both sides of CSAH 1 and extends across a series of terraces (T1-T4) on a point bar of the Red River and onto the edge of the lake plain (Figure 2 in Appendix A). The site was not identified on the floodplain (T0) despite deep testing to a depth of three meters. A geomorphological investigation was conducted by Strata Morph Geoexploration, Inc. at the site to gain a better understanding of the site’s geomorphology, soils, and stratigraphy (Appendix A). The investigation included extracting a series of cores across the site. Five terraces were defined in the site area (T0-4). Geomorphologist Mike Kolb stated that the horizontal accuracy of the terrace delineations on the maps is 20 meters or less (personal communication, 2017).

The landscape slopes very gently down to the west and is relatively flat, except for subtle swells and swales from old river channels. There is very slight ditching for drainage along the CSAH 1 ROW, and this was mostly noticeable only in the east half of the terrace. Based on LIDAR data, the highest terrace (T4) is approximately 32 feet above the river, and the lowest terrace (T1) is 22 feet, and these elevations indicate a ten foot elevation change from Terrace 1 to 4. Most of the site was in a wheat field, with surface visibility of 20-60% and good surface wash. Fiberglass rods (3/8” diameter) were used to manually separate the wheat rows to improve ground visibility. The east end of the site on the north side of CSAH 1 was in emerging soybeans with 90 percent visibility, and the east end of the site on the south side of CSAH 1 was in a yard with no visibility.

9.3 Soils

Soils at the site are described in the geomorphological report in Appendix A. Soil profiles from the XUs are also presented at the end of this section. In summary, the soils on the terraces are fine textured, formed in stratified alluvium, and often contain paleosols of varying degrees of development. The soils on the lake plain are very fine textured and formed in lacustrine sediments.

Typical soil profiles from shovel tests on Terrace 3 are presented below in Tables 8 to 10 because only two geomorphic cores were recovered from Terrace 3, and the soils in the shovel tests differ from the cores, as a multiple buried paleosols were observed in the upper meter in the shovel tests.

80

The soil profiles recorded in XUs and shovel tests on the other terraces also differed somewhat from the cores, as more buried paleosols were typically observed in the XUs and tests than the cores.

Table 8. Shovel Test 180 Soil Profile on Terrace 3. Depth Below Description Surface (cm) 0-18 Very dark gray (2.5Y 3/1) clay; Ap horizon 18-50 Black (10YR 2/1) silt loam 50-58 Very dark gray (2.5Y 3/1) silt loam; buried paleosol 58-66 Dark grayish brown (2.5Y 4/2) silty clay 66-78 Very dark gray (2.5Y 3/1) silt loam; buried paleosol 78-100 Dark grayish brown (2.5Y 4/2) silty clay

Table 9. Shovel Test 225 Soil Profile on Terrace 3. Depth Below Description Surface (cm) 0-22 Black (10YR 2/1) clay; Ap horizon 22-38 Black (10YR 2/1) silt loam 38-50 Dark grayish brown (2.5Y 4/2) silty clay 50-56 Very dark gray (2.5Y 3/1) silty clay; buried paleosol 56-68 Dark grayish brown (2.5Y 4/2) silty clay 68-80 Very dark gray (2.5Y 3/1) silty clay; buried paleosol 80-88 Dark grayish brown (2.5Y 4/2) silty clay 88-100 Very dark gray (2.5Y 3/1) silty clay; buried paleosol

Table 10. Shovel Test 273 Soil Profile on Terrace 3. Depth Below Description Surface (cm) 0-22 Black (10YR 2/1) silty clay; Ap horizon 22-38 Very dark gray (2.5Y 3/1) silty clay 38-44 Dark grayish brown (2.5Y 4/2) silty clay 44-54 Very dark gray (2.5Y 3/1) silty clay; buried paleosol 54-68 Dark grayish brown (2.5Y 4/2) silty clay 68-75 Very dark gray (2.5Y 3/1) silty clay; buried paleosol 75-95 Dark grayish brown (2.5Y 4/2) silty clay 95-100 Very dark gray (2.5Y 3/1) silty clay; buried paleosol

9.4 Radiocarbon Dates

A total of 16 radiocarbon dates were obtained from faunal samples submitted to Beta Analytic, Inc. (Beta) for AMS dating to determine the age of the components at the site and aid in understanding the natural and cultural stratigraphy. The results are summarized in Table 11, along with their associated cultural contexts. In summary, the dates indicate numerous occupations at the site. Early and Middle Archaic components are well represented in the eastern portion of the site in XUs 5-10, and a Northeastern Plains Village component is in the western portion of the site in XUs 1-4. There are no known contaminants associated with the faunal samples submitted for dating, and the dates overall appear to be accurate, based on their associated artifact types, internal vertical consistency (in XUs 5 to 10), soil contexts, and general correlation with established dates reported from other sites. In a few instances, the dates are older than established dates for diagnostic artifacts (points or ceramics) with which they are presumably associated, but there is no reason to suspect errors in the radiocarbon dates. It may be that the diagnostic artifacts, in some cases, are not actually associated with dated fauna because of artifact displacement or that the diagnostics actually date older than anticipated. 81

The radiocarbon dates are discussed in more detail with the site data below. The dating reports from Beta are included in Appendix D.

Table 11. Radiocarbon Dates from the Sites. 13C/12C Conventional Beta 2 Sigma Calibrated Results Cultural Provenience Material Ratio 14C Age BP Lab No. (95% Probability) Context* (o/oo) RCYBP (92.7%) 5058 - 4897 cal BC Calcined (7007 - 6846 cal BP) ST148 (cremated) (2.4%) 4866 - 4851 cal BC Early-Middle 475506 -19.4 6070 +/- 30 BP 60-100 cmbs bone (6815 - 6800 cal BP) Archaic carbonate (0.4%) 5189 - 5185 cal BC (7138 - 7134 cal BP) (64.4%) 660 - 730 cal AD XU1 Bone (1290 - 1220 cal BP) Northeastern 474606 -18 1300 +/- 30 BP 60-70 cmbd collagen (31%) 736 - 770 cal AD Plains Village (1214 - 1180 cal BP) (83.9%) 764 - 891 cal AD (1186 - 1059 cal BP) XU1 Bone (10.8%) 710 - 745 cal AD Northeastern 474607 -16.5 1210 +/- 30 BP 80-90 cmbd collagen (1240 - 1205 cal BP) Plains Village (0.6%) 695 - 700 cal AD (1255 - 1250 cal BP) (80.8%) 771 - 903 cal AD XU2 Bone (1179 - 1047 cal BP) Northeastern 474608 -16.9 1170 +/- 30 BP 30-40 cmbd collagen (14.6%) 918 - 965 cal AD Plains Village (1032 - 985 cal BP) (87.8%) 765 - 895 cal AD (1185 - 1055 cal BP) Northeastern XU3 Bone (6.2%) 714 - 744 cal AD Plains 474609 -16.4 1200 +/- 30 BP 50-60 cmbd collagen (1236 - 1206 cal BP) Village/Late (1.4%) 928 - 940 cal AD Woodland (1022 - 1010 cal BP) (71.8%) 773 - 906 cal AD (1177 - Northeastern XU3 Bone 1044 cal BP) Plains 474610 -16.5 1160 +/- 30 BP 100-120 cmbd collagen (23.6%) 916 - 968 cal AD 1034 - Village/Late 982 cal BP) Woodland (94.5%) 486 - 360 cal BC Late Archaic XU3 Bone (2435 - 2309 cal BP) 474611 -15.5 2330 +/- 30 BP or Early 140-160 cmbd collagen (0.9%) 271 - 262 cal BC Woodland (2220 - 2211 cal BP) XU5 Bone (95.4%) 5762 - 5648 cal BC 474612 -18.3 6830 +/- 30 BP Early Archaic 80-90 cmbd collagen (7711 - 7597 cal BP) XU5 Bone (95.4%) 6215 - 6060 cal BC 474614 -16.8 7260 +/- 30 BP Early Archaic 110-135 cmbd collagen (8164 - 8009 cal BP) XU5 Bone (95.4%) 6235 - 6084 cal BC 474615 -18.9 7320 +/- 30 BP Early Archaic 135-160 cmbd collagen (8184 - 8033 cal BP) Late XU5 Bone (95.4%) 7047 - 6772 cal BC 474616 -16.7 7980 +/- 30 BP Paleoindian- 160-185 cmbd collagen (8996 - 8721 cal BP) Early Archaic (65.6%) 5184 - 5057 cal BC Calcined (7133 - 7006 cal BP) XU6 (cremated) (18.7%) 5234 - 5192 cal BC Early-Middle 474617 -19.2 6210 +/- 30 BP 40-50 cmbd bone (7183 - 7141 cal BP) Archaic carbonate (11.2%) 5293 - 5240 cal BC (7242 - 7189 cal BP) 82

Table 11. Continued. 13C/12C Conventional Beta 2 Sigma Calibrated Results Cultural Provenience Material Ratio 14C Age BP Lab No. (95% Probability) Context* (o/oo) RCYBP (74.7%) 5306 - 5204 cal BC XU7 Bone (7255 - 7153 cal BP) Early-Middle 474618 -15.6 6240 +/- 30 BP 40-50 cmbd collagen (20.7%) 5167 - 5076 cal BC Archaic (7116 - 7025 cal BP) XU7 Bone (95.4%) 5841 - 5712 cal BC 474619 -19.3 6880 +/- 30 BP Early Archaic 70-80 cmbd collagen (7790 - 7661 cal BP) (75.1%) 6051 - 5973 cal BC XU9 Bone (8000 - 7922 cal BP) 474620 -17.6 7110 +/- 30 BP Early Archaic 100-110 cmbd collagen (20.3%) 5951 - 5916 cal BC (7900 - 7865 cal BP) XU10 Bone (95.4%) 5797 - 5664 cal BC 474621 -14 6850 +/- 30 BP Early Archaic 70-80 cmbd collagen (7746 - 7613 cal BP) * Because cultural context refers to cultural traditions and not time periods and because traditions may temporally overlap, the context designation includes traditions that may be present based on dates from other sites and projectile points recovered in dated levels from this site.

9.5 Phase I Methods and Results for Pedestrian Survey

The site was identified during Phase I pedestrian survey at ten-meter intervals. Surface visibility was fair but adequate. Close-interval survey was conducted at two-meter intervals within a ten-meter radius of all find spots. Find spots within five meters of each other were combined and recorded as a single find. A total of 319 artifact find spots were recovered from the surface, and these are presented in Tables 12 and 13. The distribution of these artifacts is discussed later in this chapter. The artifacts include 286 fauna, 18 lithics, and 15 FCR. Fractured rocks occurred sporadically throughout the field, but in general these were not collected because it was not certain if they were FCR or plow fractured. Some probable FCR were recorded with the GPS, and these are included on the find spot map in Figure 7, but these FCR were not collected or given find spot numbers. In hindsite, given the scarcity of stone in the soils at the site, it is likely that most of the fractured rocks observed in the field was FCR. Some of the fauna was historic (horse) and other large mammal remains could not be distinguished between cattle and bison (Bos/Bison Sp.) Deer remains could be historic or prehistoric.

On Terrace 2, a cluster of fauna find spots (cattle/bison, extra-large mammal, and mammal, including thermally-altered bones), was recovered adjacent to Shovel Tests 201 and 202, which contained precontact artifacts in the plow zone, including fauna, Northeastern Plains Village ceramics, and a lithic. Given the near-surface context of these artifacts, it is possible that some or all of the fauna from the find spots are associated with this component.

Find Spots recovered from Terrace 3 include lithic debris, a biface, a piece of FCR, and fauna (deer, cattle/bison, large mammal, and mammal, including a thermally-altered bone). Shovel tests adjacent to the find spots had lithics, fauna (including bison), and FCR in the plow zone and immediately below the plow zone. Other shovel tests on Terrace 3 had ceramics in the plow zone and immediately below the plow zone. Given the near-surface context of these artifacts from the shovel tests, it is possible that some or all of the fauna from find spots are associated with precontact components. The estimated age of Terrace 3 (see Section 9.19) indicates that surface artifacts would be Woodland or Northeastern Plains Village in age.

Find Spots recovered from Terrace 4 include lithic debris, a stone tool, FCR, and fauna (deer, cattle/ bison, extra-large and large mammal, and mammal, including thermally-altered bones). Shovel tests

83 on the terrace adjacent to the find spots had artifacts in the plow zone and immediately below the plow zone, including lithic debris, a point preform/late-stage biface, FCR, and fauna (including bison). Given the near-surface context of these artifacts, it is possible that some or all of the fauna from find spots are associated with precontact components. The estimated age of Terrace 4 (see Section 9.19) indicates that surface artifacts could be Middle Archaic, Woodland, or Northeastern Plains Village in age.

Find spots recovered from the lake plain include a late-stage biface, FCR, and fauna (deer, cattle/ bison, horse, turtle, extra-large and large mammal, mammal, and shell, including thermally-altered bones). Shovel tests on the lake plain had a piece of lithic debris and fauna immediately below the plow zone. Given the near-surface context of these artifacts, it is possible that some or all of the fauna from find spots are associated with precontact components. The age of the lake plain indicates that surface artifacts could be Late Paleoindian to Late Woodland/Northeastern Plains Village in age.

Table 12. Site 21PL109 Summary of Find Spot Artifacts. Total by % by Artifact Type Count (Weight g) Count (Weight g) Fauna 286 (1,111.2) 90 (31) Lithic 18 (140.8) 6 (4) FCR 15 (2,354.9) 5 (65) Total 319 (3,606.9) - % - 100

Table 13. Site 21PL109 Summary of Find Spot Artifacts. Find Spot Count Artifact Type 1 1 Mammalian, unidentifiable fragment 2 1 Bos/Bison Sp., ulna, left proximal fragment, thermally-altered 3 1 Mammalian, Extra Large, unidentifiable fragment, thermally-altered 1 Mammalian, Large, tooth, enamel fragment 4 1 Mammalian, unidentifiable fragment, thermally-altered 5 1 Equus Caballas (modern horse), metacarpal distal, fragment 1 Mammalian, unidentifiable fragment 6 1 Turtle, peripheral fragment, thermally-altered 1 Odocoileus Virginianus (white-tailed deer), metatarsal proximal fragment 7 3 Mammalian, unidentifiable fragment 1 Mammalian, unidentifiable fragment, thermally-altered 8 1 Mammalian, unidentifiable fragment, thermally-altered 9 1 Mammalian, unidentifiable fragment, thermally-altered 10 1 Equus Caballas (modern horse), tooth, molar, fragment 2 Mammalian, unidentifiable fragment thermally-altered 11 1 Mammalian, Extra Large, phalanx distal, fragment, thermally-altered 12 2 Mammalian, unidentifiable fragment, thermally-altered 13 1 Fire-cracked rock, granitic 14 1 Mammalian, unidentifiable fragment 15 1 Mammalian, Extra Large, long bone shaft, fragment 16 1 Bos/Bison Sp., calcaneous, right fragment 17 1 Mammalian, Extra Large, long bone shaft, fragment 1 Bos/Bison Sp., tooth, molar, fragment 18 1 Mammalian, Extra Large, long bone shaft, fragment

84

Table 13. Continued. Find Spot Count Artifact Type 1 Fire-cracked rock, granitic 2 Mammalian, Extra Large, long bone shaft, fragment 19 1 Mammalian, unidentifiable fragment 1 Molluscan, shell, fragment 20 1 Mammalian, unidentifiable fragment 21 1 Bos/Bison Sp., tooth, molar, fragment 1 Biface, stage 4, Swan River Chert 22 1 Mammalian, unidentifiable fragment 23 1 Mammalian, unidentifiable fragment, thermally-altered 24 6 Molluscan, shell, fragment 25 2 Molluscan, shell, fragment 26 1 Molluscan, valve, left, fragment 27 1 Mammalian, unidentifiable fragment, thermally-altered 28 2 Mammalian, unidentifiable fragment 29 1 Mammalian, tooth, enamel fragment 30 1 Bos/Bison Sp., tooth, molar, fragment 31 1 Odocoileus Virginianus (white-tailed deer), maxilla, right, fragment 32 1 Mammalian, Large, long bone shaft, fragment 33 1 Mammalian, Large, long bone shaft, fragment 34 1 Retouched flake, Knife River Flint 35 2 Mammalian, Extra Large, tooth, enamel, fragment 36 1 Bifacial thinning, Swan River Chert 1 Mammalian, Large, tooth, enamel, fragment 37 7 Mammalian, unidentifiable fragment 38 1 Broken flake, Swan River Chert 39 1 Mammalian, unidentifiable fragment, thermally-altered 1 Mammalian, tooth, enamel fragment 40 1 Mammalian, unidentifiable fragment 1 Mammalian, unidentifiable fragment, thermally-altered 1 Bos/Bison Sp., phalanx distal, fragment 41 2 Mammalian, unidentifiable fragment 8 Mammalian, unidentifiable fragment 3 Mammalian, unidentifiable fragment, thermally-altered 42 1 Nonbifacial, quartzite 1 Broken flake, unidentified chert 1 Shatter, Swan River Chert 1 Bos/Bison Sp., phalanx distal, fragment 43 7 Mammalian, unidentifiable fragment 4 Mammalian, unidentifiable fragment, thermally-altered 1 Fire-cracked rock, granitic 1 Fire-cracked rock, granitic 44 1 Mammalian, Extra Large, tooth, enamel, fragment 2 Mammalian, unidentifiable fragment 3 Mammalian, unidentifiable fragment, thermally-altered 1 Bos/Bison Sp., tooth, enamel fragment 45 1 Mammalian, unidentifiable fragment 6 Mammalian, unidentifiable fragment 46 4 Mammalian, unidentifiable fragment, thermally-altered 2 Fire-cracked rock, granitic 47 1 Mammalian, unidentifiable fragment, thermally-altered 2 Mammalian, unidentifiable fragment

85

Table 13. Continued. Find Spot Count Artifact Type 48 5 Mammalian, unidentifiable fragment 49 1 Mammalian, unidentifiable fragment 50 2 Mammalian, unidentifiable fragment 1 Fire-cracked rock, granitic 51 3 Mammalian, unidentifiable fragment 52 2 Mammalian, unidentifiable fragment 53 1 Mammalian, unidentifiable fragment, thermally-altered 1 Bos/Bison Sp., tooth, molar, fragment 54 1 Bos/Bison Sp., tooth, enamel, fragment 1 Mammalian, Large, long bone shaft, fragment 1 Mammalian, Large, long bone shaft, fragment 55 4 Mammalian, unidentifiable fragment 56 1 Nonbifacial flake, quartzite 1 Odocoileus Virginianus (white-tailed deer), pelvis, left, fragment 57 3 Mammalian, unidentifiable fragment 1 Bos/Bison Sp., tooth, molar, fragment 3 Mammalian, unidentifiable fragment 3 Bos/Bison Sp., long bone shaft, fragment 58* 2 Bos/Bison Sp., tooth, molar, fragment 1 Mammalian, Large, tooth, enamel fragment 10 Mammalian, unidentifiable fragment 1 Mammalian, unidentifiable fragment, thermally-altered 60 1 Mammalian, unidentifiable fragment 2 Mammalian, Extra Large, long bone shaft, fragment 61 5 Mammalian, unidentifiable fragment 10 Mammalian, unidentifiable fragment 62 2 Mammalian, unidentifiable fragment, thermally-altered 2 Mammalian, unidentifiable fragment, thermally-altered 63 1 Mammalian, unidentifiable fragment 1 Bos/Bison Sp., tooth, molar, fragment 64 2 Mammalian, unidentifiable fragment 65 1 Mammalian, unidentifiable fragment 66 1 Mammalian, unidentifiable fragment 67 3 Mammalian, unidentifiable fragment 1 Bifacial shaping, Swan River Chert 68 2 Mammalian, unidentifiable fragment 69 1 Mammalian, unidentifiable fragment 70 1 Mammalian, unidentifiable fragment 71 1 Mammalian, unidentifiable fragment 72 1 Broken flake, Swan River Chert 73 1 Fire-cracked rock, granitic 74 1 Mammalian, Large, unidentifiable fragment, thermally-altered 75 1 Mammalian, Large, unidentifiable fragment 76 1 Bos/Bison Sp., tooth, enamel, fragment 77 2 Mammalian, unidentifiable fragment, thermally-altered 1 Mammalian, tooth, enamel fragment 78 1 Mammalian, unidentifiable fragment 1 Fire-cracked rock, granitic 1 Bos/Bison Sp., tooth, enamel, fragment 79 1 Mammalian, Extra Large, long bone shaft, fragment 1 Mammalian, unidentifiable fragment 1 Mammalian, unidentifiable fragment, thermally-altered

86

Table 13. Continued. Find Spot Count Artifact Type 1 Other G4 flake, Swan River Chert 80 1 Mammalian, Large, unidentifiable fragment, thermally-altered 1 Fire-cracked rock, granitic 81 1 Bos/Bison Sp., tooth, enamel, fragment 82 1 Mammalian, unidentifiable fragment, thermally-altered 1 Mammalian, unidentifiable fragment 83 1 Mammalian, unidentifiable fragment, thermally-altered 84 1 Bos/Bison Sp., tooth, enamel, fragment 1 Bos/Bison Sp., phalanx, fragment 85 1 Bos/Bison Sp., unidentifiable fragment 3 Mammalian, unidentifiable fragment, thermally-altered 1 Bos/Bison Sp., long bone shaft, fragment 86 2 Mammalian, unidentifiable fragment, thermally-altered 2 Mammalian, Extra Large, unidentifiable fragment 87 1 Bos/Bison Sp., long bone shaft, fragment 5 Mammalian, unidentifiable fragment 88 1 Mammalian, unidentifiable fragment, thermally-altered 1 Nonbifacial, quartzite 89 2 Mammalian, unidentifiable fragment 90 1 Mammalian, Medium/Large, unidentifiable fragment 91 1 Mammalian, unidentifiable fragment, thermally-altered 92 1 Mammalian, Large, tooth, enamel, fragment 93 1 Bos/Bison Sp., tooth, incisor, fragment 94 4 Mammalian, unidentifiable fragment, thermally-altered 1 Shatter, Red River Chert 95 1 Mammalian, unidentifiable fragment 96 1 Odocoileus Virginianus (white-tailed deer), vertebra, lumbar centrum, fragment 97 1 Bifacial shaping ,Prairie du Chien Chert 1 Fire-cracked rock, granitic 98 1 Bos/Bison Sp., tooth, enamel, fragment 1 Fire-cracked rock, granitic 99 1 Mammalian, unidentifiable fragment 4 Mammalian, unidentifiable fragment, thermally-altered 100 2 Mammalian, unidentifiable fragment, thermally-altered 5 Mammalian, unidentifiable fragment 101 2 Mammalian, unidentifiable fragment, thermally-altered 1 Mammalian, unidentifiable fragment, thermally-altered 102 1 Bos/Bison Sp., cuneiform, complete 1 Bos/Bison Sp., tooth, enamel, fragment 1 Mammalian, unidentifiable fragment 103 1 Fire-cracked rock, granitic 104 1 Mammalian, unidentifiable fragment, thermally-altered 1 Nonbifacial, quartzite 1 Mammalian, Medium, vertebra, thoracic, fragment 105 1 Bos/Bison Sp., phalanx, complete 4 Mammalian, unidentifiable fragment 107* 1 Biface, stage 3, Swan River Chert 108 1 Bos/Bison Sp., vertebra, cervical, fragment 109 1 Broken flake, igneous 1 Mammalian, unidentifiable fragment 110 1 Bos/Bison Sp., tooth, enamel, fragment 111 1 Mammalian, unidentifiable fragment

87

Table 13. Continued. Find Spot Count Artifact Type 112 1 Bos/Bison Sp., tooth, molar, left, fragment 113 1 Mammalian, unidentifiable fragment 114 1 Mammalian, unidentifiable fragment, thermally-altered 1 Fire-cracked rock, granitic 115 5 Mammalian, unidentifiable fragment 116 1 Nonbifacial, Tongue River Silica 117 1 Mammalian, unidentifiable fragment, thermally-altered 1 Fire-cracked rock, granitic 118 1 Mammalian, unidentifiable fragment 4 Mammalian, unidentifiable fragment, thermally-altered Total 319 * There is no Find Spot 59 or 106

9.6 Phase I and II Methods and Results for Shovel Testing

Phase I shovel testing was conducted at 15-meter intervals parallel to CSAH 1 (Figure 8). The transects nearest the road were placed at 16 meters from the existing centerline. In order to avoid disturbance from previous construction and a slight ditch, the initial transects nearest the road were placed just beyond the ROW, which extends to 50 feet/15 meters each side of centerline. As the County attempted to avoid the site and stay in the existing ROW, subsequent tests were placed in the ROW at a distance of 12 meters from the existing centerline to determine if artifacts were present and assess the extent of disturbance. The tests in the ROW were placed adjacent to the initial tests that were outside the ROW, and they were given a shovel test number that corresponded to the adjacent test but was preceded with an “R” to indicate ROW (e.g., Shovel Test R65). Phase II shovel testing was conducted at 5-meter intervals in areas that contained a high density of artifacts and features along the existing CSAH 1 ROW, which is the area of potential impact for the final design. The plow zone was discarded, except in areas where surface visibility was less than 25% where the plow zone was screened. Typically, part of the soil was dry-screened, and the remaining soil that was difficult to dry-screen was then water-screened. In some tests, the soils were very clayey, and most of this soil was water-screened.

A total of 537 shovel tests were dug in the site area. A total of 12,817 artifacts were recovered from 389 shovel tests, including 12,462 faunal fragments, 215 lithics (755 pieces of debris, 42 tools and two cores), 100 precontact ceramics, and 40 FCR (Tables 14 to 16). Artifacts were recovered from 0 to 130 cm and undoubtedly extend below the depth of testing. The analysis of these artifacts is presented later in this chapter.

Shovel tests extended across a series of terraces to the edge of the lake plain. Delineation of the terraces is approximate and is based on a limited number of cores and interpretations of landscape features. Deep auger tests on the floodplain (Terrace 0) were negative, and these deposits appear to be mostly post-settlement alluvium and young in age. Terrace 1 includes the following positive Shovel Tests: 99 to 101, 205, and 207. Terrace 2 includes XUs 1 to 4 and the following Shovel Tests: Shovel Tests 21 (and all radials), R25, 30, R30, 103 to 118, 188 to 204, 237 to 245, and 272. Terrace 3 includes the following Shovel Tests: 34 to 52 (and corresponding right-of-way “R” tests), 119 to 138, 170 to 189, 216 to 235, 252 to 270, and 273 to 281. Terrace 4 includes XUs 5 to 10 and the following Shovel Tests: 54 to 79 (and corresponding right-of-way “R” tests), 139 to 169 (and corresponding right-of-way “R” tests), 208 to 214, 248 to 251. The lake plain includes the following Shovel Tests: 83, 85, 94, 97, 283, 288, and 291.

88

Table 14. Site 21PL109 Summary of Artifacts from Shovel Tests. Total by % by Artifact Type Count (Weight g) Count (Weight g) Fauna 12,462 (8,873.3) 97 (67) Lithic 215 (315.5) 2 (2) Ceramic 100 (82.7) 1 (1) FCR 40 (3,922.9) <1 (30) Total 12,817 (13,194.4) - % - 100 Table does not include a possible manuport that could be an anvilstone or crushing stone (weight= 11,440 g). The large size of the rock is highly unusual at the site, and no similar large rocks were observed.

A summary of tests with these notable cultural deposits are summarized in Table 15. In several shovel tests, dense layers of bone (possible bone beds), indicative of kill and/or butchering stations, were observed. These bone deposits were recorded on shovel test forms. It is likely that other bone beds are present in other site areas, based on the high quantity of bone in some shovel tests that did not have a discrete identifiable bone bed. Probable cooking or heating features were present in several shovel tests, based on discrete lenses of charcoal, oxidized (heated) soil, ash, and calcined bone. Some tests also contained fauna encrusted with charcoal or fauna in direct association with distinct charcoal-rich lenses that are probably anthrosols.

Table 15. Site 21PL109 Features, Bone Beds, and Other Significant Finds in Shovel Tests. Shovel Depth Description Test R56W5 64-66 Charcoal lens in association with numerous calcined fauna (anthrosol?) Fire hearth with ash, charcoal, and oxidized soil in association with thermally-altered R56E5 74-88 fauna R57 52-55 Bison bone bed, including thermally-altered fauna R63E5 72-80 Fire hearth with ash, charcoal, and oxidized soil in association with thermally-altered fauna R64W5 68-78 Charcoal in association with numerous thermally-altered fauna(anthrosol?) 75-90 Charcoal concentration (anthrosol?)and fired clay 65 80-90 Bison bone bed, including thermally-altered fauna Bison bone bed and fire hearth with charcoal and oxidized soil in association with R65 65-75 numerous thermally-altered fauna and stone tools R65W5 76-86 Bison bone bed, including charcoal Charcoal, thermally-altered fauna, and numerous fauna encrusted with charcoal R66W5 95-110 (anthrosol?) R68W5 60-70 Bison bone bed, including thermally-altered fauna R140 82-88 Bison bone bed 141S5 85-90 Bison bone bed, including thermally-altered fauna 142E5 75-85 Calcined boned deposit in wall profile 144S5 28-46 Calcined boned deposit in wall profile R142E5 60-75 Bone bed (Bison/elk) 145S5 45-60 Bison bone bed – test terminated at 60 cm because of dense bone layer 146S5 85-90 Fire hearth with charcoal and oxidized soil in association with thermally-altered fauna 146W5 30-60 Bison bone bed, including thermally-altered fauna R146W5 25-70 Bison bone bed, including thermally-altered fauna 147 65-70 Bison bone bed, including thermally-altered fauna 147E5 20-80 Bison bone bed, including thermally-altered fauna

89

Table 15. Continued. Shovel Depth Description Test Bison bone bed, including thermally-altered fauna - test terminated at 75 cm because of 148W5 35-75 dense bone layer 103- Charcoal lens (anthrosol ?) with Knife River Flint lithic in association with fauna and 148S5 106 thermally-altered fauna 149 60-80 Possible hearth indicated by fired clay and thermally-altered fauna Fire hearth with charcoal and oxidized soil in association with fauna and thermally-altered 149W5 80-84 fauna 120- Fire hearth with charcoal and oxidized soil in association with fauna and thermally-altered 149S5 128 fauna 102- Oxidized soil (probably from a fire hearth) in association with fauna and thermally-altered 149E5 110 fauna R149 48-52 Charcoal encrusted bison bone (anthrosol?) 150W5 72-76 Charcoal deposit (anthrosol?) with charcoal encrusted fauna (Elk or Wapiti) 150S5 90-100 Possible fire hearth feature 150E5 56-66 Charcoal encrusted fauna and FCR Fire hearth with ash and charcoal in association with large amount of fauna and thermally- 151 80-95 altered fauna; some fauna was encrusted with ash and charcoal 151W5 72-82 Charcoal deposit (anthrosol?) with associated bison 65-80 Bison bone bed R151 87-92 Fire hearth with charcoal lens and oxidized soil R151E5 76-80 Charcoal lens (anthrosol ?) with fauna R152W5 80-85 Charcoal lens (anthrosol ?) with fauna 125- R159E5 Charcoal lens (anthrosol ?) with lithics and fauna 130 Charcoal lens or possible fire hearth in association with large amount of fauna (bison) and R159W5 85-95 thermally-altered fauna 162 75-90 Bone bed (extra large mammal) 164 56-60 Bison bone bed, including thermally-altered fauna 210 85-90 Bison bone bed, including thermally-altered fauna and fired clay Charcoal lens (anthrosol?) or possible fire hearth in association with large amount of fauna 214 76-82 and thermally-altered fauna Oxidized soil (probably from a fire hearth) in association with moderate amount of fired 216 62-72 clay pieces 225 30-50 Bison bone bed 250 85-90 Bison bone bed and charcoal lens (anthrosol?) in association with thermally-altered fauna 260 76-80 Fire hearth with ash, charcoal, and oxidized soil

Table 16. Site 21PL109 Summary of Artifacts from Shovel Tests. Shovel Weight Depth Count Artifact Type Test (g) 45-65 33 11.2 Mammalian, unidentifiable fragment 21 1 - Fire-cracked rock, granitic 65-90 13 8.0 Mammalian, unidentifiable fragment 50-70 2 0.5 Mammalian, unidentifiable fragment 21E2.5 70-90 3 0.3 Mammalian, unidentifiable fragment (1 thermally-altered) 21E5 30-60 2 0.5 Mammalian, unidentifiable fragment 30-60 5 1.4 Mammalian, unidentifiable fragment (2 thermally-altered)) 21N2.5 1 0.1 Bird, phalanx, complete 60-90 1 0.1 Mammalian, unidentifiable fragment

90

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 1 2.4 Large mammalian, long bone, shaft, fragment 21NW3.5 30-60 41 12.7 Mammalian, unidentifiable fragment (1 thermally-altered) 30-60 10 1.0 Mammalian, unidentifiable fragment 21SW3.5 60-80 6 1.9 Mammalian, unidentifiable fragment 50-75 6 0.6 Mammalian, unidentifiable fragment 21W10 75-100 3 0.5 Mammalian, unidentifiable fragment 42 16.2 Mammalian, unidentifiable fragment (1 thermally-altered) 30-60 21W2.5 1 0.1 Tamias striatus (eastern chipmunk), tibia, left, fragment 60-90 4 0.8 Mammalian, unidentifiable fragment 1 - Fire-cracked rock, granitic 30-60 5 0.4 Mammalian, unidentifiable fragment 1 - Ceramic, grit temper, cord marked 21W5 70-90 1 0.4 Extra large mammalian, vertebra, centrum epiphysis, fragment 1 - Utilized flake, unidentified chert 90-100 1 0.1 Mammalian, unidentifiable fragment 21W5N2.5 30-60 6 0.8 Mammalian, unidentifiable fragment 30-60 1 0.3 Mammalian, unidentifiable fragment 21W5NW3.5 60-90 1 0.5 Mammalian, unidentifiable fragment 30-60 2 0.3 Mammalian, unidentifiable fragment 21W5SW3.5 60-80 7 0.5 Mammalian, unidentifiable fragment 80-100 2 0.9 Mammalian, unidentifiable fragment 30-60 7 0.7 Mammalian, unidentifiable fragment 21W7.5 60-80 4 0.6 Mammalian, unidentifiable fragment R21 70-100 5 0.5 Mammalian, unidentifiable fragment R21W5 30-60 1 0.2 Mammalian, unidentifiable fragment 1 0.1 Ictalurus punctatus (channel catfish), cranium, fragment 30-60 R25 1 0.5 Ictalurus punctatus (channel catfish), quadrate, fragment 80-100 4 0.5 Mammalian, unidentifiable fragment (2 thermally-altered) 30 45-65 1 1.0 Mammalian, unidentifiable fragment 34 0-20 1 0.1 Mammalian, unidentifiable fragment 36 25-35 1 0.1 Mammalian, unidentifiable fragment, thermally-altered 38 70-90 1 0.5 Mammalian, unidentifiable fragment 0-20 1 - Other G4 flake, Red River Chert 39 60-80 1 0.5 Poliocitellus franklinii (franklin's ground squirrel), humerus, right, complete 30-70 5 0.6 Mammalian, unidentifiable fragment R39 70-100 1 0.1 Mammalian, unidentifiable fragment 40 60-70 1 0.1 Tamias striatus (eastern chipmunk), maxilla, left, fragment 0-20 1 0.1 Mammalian, unidentifiable fragment 41 70-80 1 0.4 Mammalian, unidentifiable fragment 42 60-70 2 0.3 Mammalian, unidentifiable fragment (1 thermally-altered) R42 0-30 2 0.6 Mammalian, unidentifiable fragment 1 - Broken flake, unidentified material 0-35 1 0.1 Mammalian, unidentifiable fragment 1 - Nonbifacial, Knife River Flint 45 1 7.7 Large mammalian, long bone, shaft, fragment 80-100 10 4.0 Mammalian, unidentifiable fragment (3 thermally-altered) 1 - Utilized flake, Knife River Flint R45 0-30 4 2.0 Mammalian, unidentifiable fragment 46 25-50 9 3.6 Mammalian, unidentifiable fragment 1 0.2 Bird, long bone, shaft, fragment 0-35 28 10.1 Mammalian, unidentifiable fragment (1 thermally-altered) 48 45-60 3 2.4 Mammalian, unidentifiable fragment 90-120 28 6.6 Mammalian, unidentifiable fragment 20-50 3 0.5 Mammalian, unidentifiable fragment R48E5 25-50 2 0.5 Mammalian, unidentifiable fragment (1 thermally-altered)

91

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 2 0.3 Mammalian, unidentifiable fragment 0-30 1 0.1 Ondatra zibethicus (muskrat), mandible, left, fragment 49 30-50 1 0.1 Mammalian, unidentifiable fragment 70-100 1 0.1 Mammalian, unidentifiable fragment 50-75 1 - Broken flake, quartzite R49W5 75-100 2 1.3 Mammalian, unidentifiable fragment 25-40 1 0.2 Mammalian, unidentifiable fragment, thermally-altered 51 80-100 2 0.3 Mammalian, unidentifiable fragment 52 30-40 1 0.1 Mammalian, unidentifiable fragment 54 50-60 33 4.2 Mammalian, unidentifiable fragment R54 30-60 2 0.7 Mammalian, unidentifiable fragment 30-60 2 0.4 Mammalian, unidentifiable fragment R54E5 60-90 1 0.2 Mammalian, unidentifiable fragment 55 0-20 1 0.3 Mammalian, unidentifiable fragment, thermally-altered 0-30 6 2.9 Mammalian, unidentifiable fragment (1 thermally-altered) 1 54 Bison bison (bison), vertebra, lumbar, fragment 30-50 56 60 13.2 Mammalian, unidentifiable fragment (3 thermally-altered) 50-75 15 1.4 Mammalian, unidentifiable fragment (5 thermally-altered) 75-90 9 0.8 Mammalian, unidentifiable fragment (1 thermally-altered) R56 25-50 12 2.0 Mammalian, unidentifiable fragment (4 thermally-altered) 25-50 3 0.5 Mammalian, unidentifiable fragment, thermally-altered R56E5 6 8.9 Large mammalian, long bone, shaft, fragment, thermally-altered 75-100 127 67.6 Mammalian, unidentifiable fragment (110 thermally-altered) 1 0.3 Large mammalian, tooth, enamel, fragment 25-50 2 0.6 Mammalian, unidentifiable fragment (1 thermally-altered) 1 11.8 Extra large mammalian, long bone, shaft, fragment 2 6.1 Extra large mammalian, unidentifiable fragment 50-75 6 8.3 Large mammalian, unidentifiable fragment R56W5 90 21.7 Mammalian, unidentifiable fragment (48 thermally-altered) 1 - Broken flake, unidentified material 1 21.6 Extra large mammalian, long bone, shaft, fragment 75-100 8 1.5 Mammalian, unidentifiable fragment 13 2.8 Mammalian, unidentifiable fragment, thermally-altered 40-60 33 6.2 Mammalian, unidentifiable fragment (7 thermally-altered) 57 60-80 3 0.4 Mammalian, unidentifiable fragment (1 thermally-altered) 80-90 5 1.1 Mammalian, unidentifiable fragment (3 thermally-altered) 1 5.1 Bison bison (bison), phalanx, fragment 1 1.9 Bison bison (bison), tooth, premolar, fragment 25-50 3 0.8 Large mammalian, tooth, fragment 19 11.7 Mammalian, unidentifiable fragment (5 thermally-altered) 1 39.3 Bison bison (bison), metacarpal, distal fragment 1 21.1 Bison bison (bison), scapula, left, fragment 3 1.6 Bison bison (bison), tooth, enamel, fragment R57 3 11.2 Bison bison (bison), tooth, fragment 50-75 1 - Bifacial shaping, unidentified chert 1 - Bipolar flake, quartzite 3 10.4 Extra large mammalian, long bone, shaft, fragment 19 23.5 Large mammalian, unidentifiable fragment (2 thermally-altered) 90 34.7 Mammalian, unidentifiable fragment (23 thermally-altered) 1 5.7 Extra large mammalian, unidentifiable fragment 75-100 21 7.7 Mammalian, unidentifiable fragment

92

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 1 4.4 Bison bison (bison), tooth, fragment 25-50 5 2.8 Mammalian, unidentifiable fragment 1 12.5 Bison bison (bison), cuneiform, complete 1 13.3 Bison bison (bison), magnum, complete 1 15.6 Bison bison (bison), metacarpal, right, proximal fragment

50-75 1 11.7 Bison bison (bison), unciform, complete R57E5 1 10.5 Extra large mammalian, long bone, shaft, fragment 1 5.8 Extra large mammalian, unidentifiable fragment 30 9.7 Mammalian, unidentifiable fragment (4 thermally-altered) 1 0.1 Mammalian, long bone, shaft, fragment 75-100 2 0.5 Mammalian, unidentifiable fragment 25-50 2 0.2 Mammalian, unidentifiable fragment, thermally-altered 1 0.3 Bison/Cervus sp. (bison/elk), tooth, fragment R57W5 50-75 2 2.3 Mammalian, long bone, shaft, fragment 9 1.6 Mammalian, unidentifiable fragment (2 thermally-altered) 75-100 1 1.4 Bison bison (bison), tooth, premolar, fragment 0-15 2 0.6 Mammalian, unidentifiable fragment 58 15-50 1 0.1 Molluscan, shell, fragment 50-70 1 0.4 Mammalian, unidentifiable fragment 1 12.6 Bos/Bison sp. (cattle/bison), tooth, molar, fragment 0-5 1 2.4 Mammalian, unidentifiable fragment R58 1 - Other G4 flake, Prairie du Chien Chert (oolitic) 25-50 8 0.9 Mammalian, unidentifiable fragment (3 thermally-altered) 75-100 1 0.2 Mammalian, unidentifiable fragment, thermally-altered 1 - Fire-cracked rock, granitic 25-50 1 0.3 Mammalian, long bone, shaft, fragment, thermally-altered R58E5 50-75 1 0.1 Mammalian, unidentifiable fragment 75-100 6 0.6 Mammalian, unidentifiable fragment (3 thermally-altered) 25-50 6 2.1 Mammalian, unidentifiable fragment (1 thermally-altered) R58W5 50-75 9 12.4 Mammalian, unidentifiable fragment 75-100 2 0.3 Mammalian, unidentifiable fragment 0-20 11 3.3 Mammalian, unidentifiable fragment (2 thermally-altered) 1 3.2 Large mammalian, unidentifiable fragment 20-40 59 1 1.2 Mammalian, unidentifiable fragment, thermally-altered 40-60 1 0.1 Mammalian, unidentifiable fragment 80-100 4 3.9 Mammalian, unidentifiable fragment 0 1 - Projectile point (Plainview type), Prairie du Chien Chert (oolitic) R59 25-50 10 3.2 Mammalian, unidentifiable fragment (5 thermally-altered) 50-75 1 0.1 Mammalian, unidentifiable fragment R59E5 50-75 12 2.8 Mammalian, unidentifiable fragment (8 thermally-altered) 1 - Nonbifacial, unidentified material 25-50 12 3.9 Mammalian, unidentifiable fragment (4 thermally-altered) 1 1.3 Bison bison (bison), tooth, fragment 50-75 9 2.5 Mammalian, unidentifiable fragment (1 thermally-altered) 3 - Broken flake, Swan River Chert 75-90 1 - Other G4 flake, Swan River Chert R59W5 3 2.3 Mammalian, unidentifiable fragment 1 - Bipolar flake, Swan River Chert 3 - Broken flake, Swan River Chert 1 - Nonbifacial, Swan River Chert 95-105 1 - Other G4 flake, Swan River Chert 1 - Shatter, Swan River Chert 4 0.4 Mammalian, unidentifiable fragment

93

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 0-25 3 0.1 Mammalian, unidentifiable fragment (2 thermally-altered) 25-50 4 0.7 Mammalian, unidentifiable fragment (2 thermally-altered) 60 50-70 18 3.5 Mammalian, unidentifiable fragment (1 thermally-altered) 70-80 1 0.1 Mammalian, unidentifiable fragment 80-100 1 0.1 Mammalian, unidentifiable fragment 25-50 1 0.2 Mammalian, unidentifiable fragment, thermally-altered 50-75 6 1.7 Mammalian, unidentifiable fragment (3 thermally-altered) 4 1.2 Bison bison (bison), tooth, fragment R60 1 1.6 Bison bison (bison), tooth, premolar, fragment 75-100 2 7.6 Large mammalian, long bone, shaft, fragment 11 2.3 Mammalian, unidentifiable fragment (3 thermally-altered) 25-50 2 0.6 Mammalian, unidentifiable fragment 2 - Other G4 flake, Knife River Flint 50-75 1 14.9 Extra large mammalian, long bone, shaft, fragment R60E5 35 13.9 Mammalian, unidentifiable fragment (9 thermally-altered) 19 4.0 Mammalian, unidentifiable fragment (12 thermally-altered) 75-100 1 0.3 Vertebrata, unidentifiable, fragment 20-50 2 0.6 Mammalian, unidentifiable fragment 1 3.8 Extra large mammalian, long bone, shaft, fragment R60W5 50-75 1 0.7 Large mammalian, tooth, incisor, fragment 3 0.7 Mammalian, unidentifiable fragment 1 18.3 Bison/Cervus sp. (bison/elk), scapula, left, fragment 60-80 61 75 15.1 Mammalian, unidentifiable fragment (22 thermally-altered) 80-100 5 1.6 Mammalian, unidentifiable fragment, thermally-altered 25-50 4 0.6 Mammalian, unidentifiable fragment, thermally-altered R61 50-75 1 0.6 Mammalian, unidentifiable fragment 75-100 1 - Bifacial thinning, Grand Meadow Chert 2 0.4 Bison bison (bison), tooth, fragment 1 17.7 Bison bison (bison), tooth, molar, fragment 25-50 R61E5 39 12.9 Mammalian, unidentifiable fragment (7 thermally-altered) 1 - Side scraper, Swan River Chert 50-75 37 6.9 Mammalian, unidentifiable fragment 25-50 2 0.2 Mammalian, unidentifiable fragment, thermally-altered 1 79.3 Bison bison (bison), metatarsal, fragment R61W5 50-70 22 8.2 Mammalian, unidentifiable fragment (2 thermally-altered) 70-100 1 0.3 Mammalian, unidentifiable fragment, thermally-altered 2 0.7 Large mammalian, tooth, enamel, fragment 45-50 1 0.1 Mammalian, unidentifiable fragment 62 1 0.1 Mammalian, unidentifiable fragment, thermally-altered 70-100 4 1.1 Mammalian, unidentifiable fragment 25-50 3 0.3 Mammalian, unidentifiable fragment R62 50-75 1 0.3 Mammalian, unidentifiable fragment, thermally-altered 75-100 1 1.3 Large mammalian, long bone, shaft, fragment 1 - Broken flake, quartzite 25-50 1 - Fire-cracked rock, granitic 3 0.7 Mammalian, unidentifiable fragment R62E5 50-75 7 1.4 Mammalian, unidentifiable fragment (1 thermally-altered) 1 26.6 Bison bison (bison), rib, shaft, fragment 75-100 1 - Fire-cracked rock, igneous 10 13.5 Large mammalian, unidentifiable fragment (9 thermally-altered) 25-50 R62W5 8 1.8 Mammalian, unidentifiable fragment (4 thermally-altered) 75-100 1 0.4 Mammalian, unidentifiable fragment

94

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 0-25 1 0.2 Mammalian, unidentifiable fragment 40-55 8 1.3 Mammalian, unidentifiable fragment (2 thermally-altered) 1 34.8 Bison bison (bison), long bone, shaft, fragment 63 55-75 1 0.4 Carnivora, large, tooth, canine, fragment 17 2.3 Mammalian, unidentifiable fragment (6 thermally-altered) 75-100 2 0.4 Mammalian, unidentifiable fragment 30-50 2 0.4 Mammalian, unidentifiable fragment, thermally-altered R63 75-100 2 0.4 Mammalian, unidentifiable fragment 30-50 5 2.2 Mammalian, unidentifiable fragment (3 thermally-altered) 1 1.9 Bison bison (bison), phalanx, proximal fragment 1 1.1 Large mammalian, long bone, shaft, fragment, thermally-altered 22 6.7 Mammalian, unidentifiable fragment (18 thermally-altered) 50-75 1 0.2 Urocyon cinereoargenteus (grey fox), lunate, complete 1 0.2 Urocyon cinereoargenteus (grey fox), metacarpal, proximal fragment 0.5 Urocyon cinereoargenteus (grey fox), radius, left, distal fragment, thermally- R63E5 1 altered 1 6.2 Extra large mammalian, unidentifiable fragment 72-80 1 0.2 Mammalian, unidentifiable fragment, thermally-altered 1 8.0 Bison bison (bison), phalanx, complete 3 4.6 Large mammalian, unidentifiable fragment, thermally-altered 75-100 1 0.2 Medium mammalian, phalanx, complete, thermally-altered 18 4.6 Mammalian, unidentifiable fragment (12 thermally-altered) 1 - Fire-cracked rock, granitic 25-50 4 0.9 Mammalian, unidentifiable fragment R63W5 50-75 10 1.9 Mammalian, unidentifiable fragment (3 thermally-altered) 75-100 1 5.9 Bison bison (bison), phalanx, distal fragment 1 - Other G4 flake, Tongue River Silica 25-45 64 7 2.9 Mammalian, unidentifiable fragment (3 thermally-altered) 50-70 6 1.2 Mammalian, unidentifiable fragment 25-50 1 0.6 Mammalian, unidentifiable fragment, thermally-altered R64 50-75 1 0.5 Mammalian, unidentifiable fragment, thermally-altered 75-100 5 1.6 Mammalian, unidentifiable fragment (3 thermally-altered) 4 4.1 Large mammalian, unidentifiable fragment (1 thermally-altered) 40-50 2 0.4 Mammalian, unidentifiable fragment R64E5 50-75 7 1.1 Mammalian, unidentifiable fragment (1 thermally-altered) 1 11.5 Bison bison (bison), cuneiform, complete 75-100 14 3.7 Mammalian, unidentifiable fragment (12 thermally-altered) 25-50 1 0.2 Mammalian, unidentifiable fragment, thermally-altered 11 17.2 Large mammalian, unidentifiable fragment, thermally-altered R64W5 50-75 62 17.8 Mammalian, unidentifiable fragment, thermally-altered 1 0.2 Turtle, carapace, fragment, thermally-altered 75-100 18 5.8 Mammalian, unidentifiable fragment, thermally-altered 25-50 4 2.6 Mammalian, unidentifiable fragment (1 thermally-altered) 1 0.7 Bison bison (bison), tooth, cheek, fragment 50-75 R65E5 1 - Broken flake, Swan River Chert 1 5.4 Bison bison (bison), vertebra, centrum epiphysis, fragment 75-100 5 1.1 Mammalian, unidentifiable fragment 50-75 1 34.5 Bison bison (bison), metacarpal, distal fragment 1 15.7 Bison bison (bison), ilium, fragment R65W5 75-100 27 9.2 Mammalian, unidentifiable fragment 80-90 1 220.4 Bison bison (bison), vertebra, cervical, fragment

95

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 1 - Other G4 flake, Swan River Chert 0-35 17 3.7 Mammalian, unidentifiable fragment (8 thermally-altered) 1 - Broken flake, chalcedony 1 0.8 Geomys bursarius (plains pocket gopher), cranium, fragment 35-60 2 0.5 Mammalian, tooth, fragment 9 1.9 Mammalian, unidentifiable fragment (1 thermally-altered) 1 - Utilized flake, Knife River Flint 1 4.9 Bison bison (bison), tooth, premolar, fragment 60-70 - Core, freehand nonbifacial, unpatterned (multi-directional), Tongue River 1 Silica 1 - Broken flake, Swan River Chert 1 - Fired clay 60-80 3 2.3 Large mammalian, unidentifiable fragment 13 2.4 Mammalian, unidentifiable fragment (3 thermally-altered) 10 9.3 Large mammalian, unidentifiable fragment 74-80 65 8 2.6 Mammalian, unidentifiable fragment (1 thermally-altered) 6 77.3 Bison bison (bison), long bone, shaft, fragment 1 3.6 Bison bison (bison), sesamoid, complete 1 23.9 Bison bison (bison), tooth, molar, fragment 1 0.9 Bison bison (bison), vertebra, centrum epiphysis, fragment 2 184.2 Bison bison (bison), vertebra, thoracic, fragment 1 - Bifacial shaping, Knife River Flint 2 - Bifacial shaping, Swan River Chert 80-90 1 - Bifacial thinning, Swan River Chert 3 - Broken flake, Swan River Chert 2 - Other G4 flake, Swan River Chert 1 - Fire-cracked rock, basaltic 16 52.9 Extra large mammalian, unidentifiable fragment 27 21.3 Large mammalian, unidentifiable fragment (3 thermally-altered) 178 42.6 Mammalian, unidentifiable fragment (24 thermally-altered) 90-100 5 1.3 Mammalian, unidentifiable fragment (1 thermally-altered) 25-50 1 0.2 Mammalian, unidentifiable fragment 1 10.2 Bison bison (bison), long bone, shaft, fragment 1 3.1 Bison bison (bison), vertebra, thoracic, fragment 1 - Bifacial shaping, Swan River Chert 1 - Bifacial thinning, Swan River Chert 1 - Broken flake, Red River Chert 50-75 3 - Broken flake, Swan River Chert 1 - Broken flake, Tongue River Silica 1 0.3 Large mammalian, tooth, fragment 24 47.3 Large mammalian, unidentifiable fragment (8 thermally-altered) R65 84 24.0 Mammalian, unidentifiable fragment (37 thermally-altered) 2 - Utilized flake, Swan River Chert 1 14.5 Bison bison (bison), phalanx, complete 65-75 1 8.5 Extra large mammalian, long bone, shaft, fragment 1 11.2 Extra large mammalian, unidentifiable fragment 2 - Bifacial shaping, Swan River Chert 1 - Bifacial thinning, Swan River Chert 75-100 1 - Broken flake, unidentified chert 3 5.3 Large mammalian, unidentifiable fragment 44 12.4 Mammalian, unidentifiable fragment (23 thermally-altered) 20-30 4 0.3 Vertebrata, unidentifiable, fragment 3 1.5 Cervidae (deer), tooth, fragment 66 45-70 9 9.5 Large mammalian, unidentifiable fragment (7 thermally-altered) 60 13.4 Mammalian, unidentifiable fragment (51 thermally-altered) 70-100 17 3.9 Mammalian, unidentifiable fragment (16 thermally-altered)

96

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 30-55 6 3.2 Mammalian, unidentifiable fragment 1 0.6 Bison/Cervus sp. (bison/elk), tooth, fragment R66 55-75 6 7.6 Large mammalian, unidentifiable fragment (3 thermally-altered) 16 5.4 Vertebrata, unidentifiable, fragment (14 thermally-altered) 75-100 40 10.8 Mammalian, unidentifiable fragment (1 thermally-altered) 55-80 6 1.4 Mammalian, unidentifiable fragment (2 thermally-altered) R66E5 75-100 2 1.0 Mammalian, unidentifiable fragment (1 thermally-altered) 25-50 1 3.2 Extra large mammalian, long bone, shaft, fragment 50-75 2 0.4 Mammalian, unidentifiable fragment (1 thermally-altered) 75-110 42 14 Mammalian, unidentifiable fragment (11 thermally-altered) R66W5 10 30.3 Large mammalian, unidentifiable fragment 68 19.4 Mammalian, unidentifiable fragment (13 thermally-altered) 100-110 Wood charcoal (recovered in direct association with fauna that was encrusted 1 - with charcoal) 1 4.2 Extra large mammalian, long bone, shaft, fragment 25-40 7 5.0 Large mammalian, unidentifiable fragment (1 thermally-altered) 67 19 3.5 Mammalian, unidentifiable fragment (1 thermally-altered) 40-60 19 2.8 Mammalian, unidentifiable fragment (1 thermally-altered) 60-80 5 1.0 Mammalian, unidentifiable fragment R67 50-75 2 1.7 Mammalian, unidentifiable fragment, thermally-altered R67E5 50-75 1 6.8 Large mammalian, long bone, shaft, fragment 50-75 1 - Bifacial thinning, Bijou Hills Quartzite R67W5 8 3.8 Mammalian, unidentifiable fragment (1 thermally-altered) 75-100 3 0.2 Vertebrata, unidentifiable, fragment 68 45-70 3 0.7 Mammalian, unidentifiable fragment (1 thermally-altered) 50-75 1 0.1 Vertebrata, unidentifiable, fragment R68 1 2.8 Large mammalian, unidentifiable fragment 75-100 4 0.8 Mammalian, unidentifiable fragment 1 3.7 Extra large mammalian, petrosal, fragment 50-75 6 1.3 Mammalian, unidentifiable fragment 1 - Other G4 flake, Swan River Chert R68E5 3 14.3 Extra large mammalian, long bone, shaft, fragment 75-100 2 7.0 Extra large mammalian, unidentifiable fragment, thermally-altered 30 8.9 Mammalian, unidentifiable fragment (7 thermally-altered) 1 - Projectile point (Oxbow type), Swan River Chert 1 6.2 Extra large mammalian, long bone, shaft, fragment 50-75 24 42.7 Large mammalian, unidentifiable fragment 188 38.9 Mammalian, unidentifiable fragment (22 thermally-altered) 60-65 1 47.8 Bison bison (bison), femur, right, shaft, fragment R68W5 1 - Broken flake, Swan River Chert 1 - Nonbifacial, Swan River Chert 70-100 1 8.4 Extra large mammalian, unidentifiable fragment 28 7.3 Mammalian, unidentifiable fragment (6 thermally-altered) 69 50-70 1 0.1 Mammalian, unidentifiable fragment, thermally-altered 1 2.2 Bos/Bison sp. (cattle/bison), astragalus, fragment 0-25 8 6.1 Large mammalian, unidentifiable fragment 70 18 2.1 Mammalian, unidentifiable fragment 25-50 1 0.5 Mammalian, unidentifiable fragment 0-35 2 0.2 Mammalian, unidentifiable fragment 76 65-100 1 - Other G4 flake, Red River Chert 0-30 1 0.2 Mammalian, unidentifiable fragment 77 30-55 1 0.1 Mammalian, unidentifiable fragment 80-100 1 0.1 Mammalian, unidentifiable fragment 0-30 1 0.2 Large mammalian, tooth, enamel, fragment 78 30-60 1 0.3 Mammalian, unidentifiable fragment, thermally-altered 60-90 2 0.3 Mammalian, unidentifiable fragment (1 thermally-altered)

97

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 3 2.5 Bison/Cervus sp. (bison/elk), tooth, cheek, fragment 79 90-110 12 3.2 Mammalian, unidentifiable fragment (1 thermally-altered) 1 0.1 Bird, long bone, shaft, fragment 83 70-100 1 .03 Mammalian, unidentifiable fragment, thermally-altered 85 30-50 3 0.9 Mammalian, unidentifiable fragment 94 0-20 1 - Decortication, Red River Chert 97 0-30 1 0.4 Mammalian, unidentifiable fragment 1 0.8 Geomys bursarius (plains pocket gopher), mandible, left, fragment 25-50 1 0.3 Small mammalian, long bone, shaft, fragment 99 7 1.0 Mammalian, unidentifiable fragment 100-120 1 0.1 Turtle, carapace/plastron, fragment 1 0.1 Vertebrata, unidentifiable, fragment 100 0-20 2 1.6 Mammalian, unidentifiable fragment 101 0-30 1 0.2 Mammalian, unidentifiable fragment 0-30 1 0.7 Medium/large mammalian, tooth, fragment 103 30-70 1 0.5 Mammalian, unidentifiable fragment 12 11.0 Large mammalian, unidentifiable fragment 20-50 1 0.3 Large mammalian, vertebra, centrum epiphysis, fragment 104 37 6.8 Mammalian, unidentifiable fragment (2 thermally-altered) 50-80 5 1.5 Mammalian, unidentifiable fragment 80-100 1 0.1 Mammalian, unidentifiable fragment 19 4.2 Mammalian, unidentifiable fragment (3 thermally-altered) 30-60 3 0.2 Vertebrata, unidentifiable, fragment 3 2.7 Large mammalian, unidentifiable fragment 104E5 60-80 5 0.7 Mammalian, unidentifiable fragment (2 thermally-altered) 1 0.4 Large mammalian, long bone, shaft, fragment, thermally-altered 80-100 6 0.6 Mammalian, unidentifiable fragment 1 0.3 Bison/Cervus sp. (bison/elk), tooth, cheek, fragment 104E5S2.5 30-60 9 2.1 Mammalian, unidentifiable fragment (1 thermally-altered) 30-60 41 9.6 Mammalian, unidentifiable fragment (2 thermally-altered) 5 5.2 Large mammalian, unidentifiable fragment 60-80 104S2.5 26 4.2 Mammalian, unidentifiable fragment (3 thermally-altered) 3 2.8 Large mammalian, unidentifiable fragment 80-100 3 0.3 Mammalian, unidentifiable fragment 1 9.2 Extra large mammalian, unidentifiable fragment 30-60 9 1.32 Mammalian, unidentifiable fragment 1 0.1 Rodentia, tooth, incisor, fragment 104W5 1 41.7 Bison bison (bison), long bone, shaft, fragment 60-90 12 1.1 Mammalian, unidentifiable fragment 90-100 1 4.9 Extra large mammalian, unidentifiable fragment 30-60 23 5.1 Mammalian, unidentifiable fragment (1 thermally-altered) 104W5S2.5 1 1.7 Bone tool, mammalian, large, unidentifiable fragment, worked/tool 60-90 23 4.4 Mammalian, unidentifiable fragment (1 thermally-altered) 1 - Ceramic, grit temper, smooth 30-60 R104E5 12 1.7 Mammalian, unidentifiable fragment (1 thermally-altered) 60-90 2 0.2 Mammalian, unidentifiable fragment 1 0.8 Large mammalian, long bone, shaft, fragment 50-70 3 0.6 Mammalian, unidentifiable fragment R104W5 1 0.7 Medium/large mammalian, long bone, shaft, fragment 70-90 2 0.2 Mammalian, unidentifiable fragment 1 2.0 Large mammalian, long bone, shaft, fragment 20-50 27 0.7 Mammalian, unidentifiable fragment (2 thermally-altered) 19 4.2 Mammalian, unidentifiable fragment (2 thermally-altered) 50-75 105 1 0.4 Molluscan, shell, fragment 1 0.2 Geomys bursarius (plains pocket gopher), tooth, incisor, fragment 75-100 6 1.8 Mammalian, unidentifiable fragment (1 thermally-altered) 5 0.2 Vertebrata, unidentifiable, fragment

98

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 1 - Broken flake, unidentified material 40-70 7 11.7 Large mammalian, unidentifiable fragment 105E5 40 8.9 Mammalian, unidentifiable fragment 70-100 3 0.3 Mammalian, unidentifiable fragment 30-60 2 0.5 Mammalian, unidentifiable fragment 105E5S2.5 80-100 1 0.2 Mammalian, unidentifiable fragment 30-60 47 10.1 Mammalian, unidentifiable fragment (1 thermally-altered) 105S2.5 60-90 8 1.3 Mammalian, unidentifiable fragment (1 thermally-altered) 30-45 5 1.7 Mammalian, unidentifiable fragment 105W5 45-75 5 1.6 Mammalian, unidentifiable fragment 1 54.2 Bos/Bison sp. (cattle/bison), radius, right, proximal fragment 0-30 1 - Fire-cracked rock, igneous 1 0.8 Large mammalian, unidentifiable fragment 105W5S2.5 30-60 1 0.7 Large mammalian, unidentifiable fragment, thermally-altered 24 4.3 Mammalian, unidentifiable fragment (2 thermally-altered) 80-100 1 0.1 Mammalian, unidentifiable fragment, thermally-altered 30-60 3 0.5 Mammalian, unidentifiable fragment R105E5 80-90 1 0.3 Mammalian, unidentifiable fragment 0-25 3 0.6 Mammalian, unidentifiable fragment (1 thermally-altered) 20-30 2 0.6 Mammalian, unidentifiable fragment 5 0.8 Geomys bursarius (plains pocket gopher), cranium, fragment 1 0.2 Geomys bursarius (plains pocket gopher), humerus, left, distal fragment 1 1.0 Geomys bursarius (plains pocket gopher), mandible, left, fragment 1 1.0 Geomys bursarius (plains pocket gopher), mandible, right, fragment 1 1.2 Geomys bursarius (plains pocket gopher), maxilla, fragment 25-55 1 0.1 Geomys bursarius (plains pocket gopher), scapula, left, distal fragment 1 0.1 Geomys bursarius (plains pocket gopher), scapula, right, distal fragment 40 2.5 Geomys bursarius (plains pocket gopher), unidentifiable, fragment 1 0.1 Geomys bursarius (plains pocket gopher), vertebra, atlas, fragment 106 2 0.1 Geomys bursarius (plains pocket gopher), vertebra, fragment 10 1.7 Mammalian, unidentifiable fragment (1 thermally-altered) 1 0.2 Geomys bursarius (plains pocket gopher), humerus, right, distal fragment 1 0.2 Geomys bursarius (plains pocket gopher), ulna, left, shaft, fragment 1 0.1 Geomys bursarius (plains pocket gopher), ulna, right, shaft, fragment 55-85 3 0.2 Geomys bursarius (plains pocket gopher), vertebra, fragment 1 0.2 Lymnaeidae (pond snails), shell, complete 2 0.2 Mammalian, unidentifiable fragment 70-75 2 12.8 Extra large mammalian, unidentifiable fragment 1 1.3 Large mammalian, long bone, shaft, fragment 85-100 1 0.2 Mammalian, unidentifiable fragment 0-0 1 7.6 Bos/Bison sp. (cattle/bison), phalanx, distal fragment 1 2.0 Large mammalian, long bone, shaft, fragment 20-40 4 0.4 Mammalian, unidentifiable fragment 106E5 40-60 4 0.7 Mammalian, unidentifiable fragment 60-80 1 0.1 Mammalian, unidentifiable fragment 80-100 1 0.3 Mammalian, unidentifiable fragment 0.1 Ictidomys tridecemlineatus (thirteen-lined ground squirrel), mandible, left, 1 fragment 106S2.5 30-60 2 1.4 Large mammalian, unidentifiable fragment (1 thermally-altered) 1 0.2 Mammalian, unidentifiable fragment, thermally-altered 1 - Ceramic, grit temper, smooth 30-60 106W5 1 0.1 Vertebrata, unidentifiable, fragment 60-90 3 0.3 Mammalian, unidentifiable fragment (1 thermally-altered) 30-60 5 0.9 Mammalian, unidentifiable fragment 106W5S2.5 60-70 8 - Fire-cracked rock, igneous 30-60 5 0.4 Mammalian, unidentifiable fragment (2 thermally-altered) R106 60-95 1 1.3 Large mammalian, long bone, shaft, fragment

99

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 30-60 5 1.2 Mammalian, unidentifiable fragment (1 thermally-altered) R106E5 90-100 1 0.3 Mammalian, unidentifiable fragment 30-60 16 3.2 Mammalian, unidentifiable fragment 60-80 7 1.4 Mammalian, unidentifiable fragment R106W5 1 0.6 Large mammalian, unidentifiable fragment 80-100 7 0.7 Mammalian, unidentifiable fragment 1 1.7 Large mammalian, long bone, shaft, fragment, thermally-altered 107 30-50 11 1.4 Mammalian, unidentifiable fragment (1 thermally-altered) 1 1.0 Large mammalian, long bone, shaft, fragment 20-45 7 0.9 Mammalian, unidentifiable fragment 107E5 45-70 12 2.8 Mammalian, unidentifiable fragment 2 2.2 Large mammalian, unidentifiable fragment 70-100 9 1.1 Mammalian, unidentifiable fragment 30-60 1 0.3 Mammalian, unidentifiable fragment, thermally-altered 107E5S2.5 60-90 4 0.3 Mammalian, unidentifiable fragment 1 25.0 Bison bison (bison), tooth, molar, fragment 1 0.2 Geomys bursarius (plains pocket gopher), femur, right, proximal fragment 30-60 25 4.0 Mammalian, unidentifiable fragment 107S2.5 7 0.3 Vertebrata, unidentifiable, fragment 60-80 21 3.0 Mammalian, unidentifiable fragment 80-100 12 1.9 Mammalian, unidentifiable fragment 1 31.9 Bison bison (bison), phalanx, distal fragment 1 - Ceramic, grit temper, smooth 40-60 107W5 2 - Ceramic, grit temper, undetermined 17 2.8 Mammalian, unidentifiable fragment (6 thermally-altered) 60-80 1 0.1 Mammalian, unidentifiable fragment 30-60 3 0.6 Mammalian, unidentifiable fragment 60-80 2 0.3 Mammalian, unidentifiable fragment 107W5S2.5 1 - Other G4 flake, Knife River Flint 80-100 3 0.3 Mammalian, unidentifiable fragment 30-60 1 0.4 Mammalian, unidentifiable fragment R107 60-90 4 1.1 Mammalian, unidentifiable fragment R107E5 90-100 2 0.9 Mammalian, unidentifiable fragment R107W5 30-60 4 0.4 Mammalian, unidentifiable fragment (2 thermally-altered) 0-20 1 0.4 Mammalian, unidentifiable fragment 60-85 8 0.9 Mammalian, unidentifiable fragment 108 1 0.4 Geomys bursarius (plains pocket gopher), maxilla, right, fragment 85-105 1 0.4 Mammalian, unidentifiable fragment 30-60 3 0.8 Mammalian, unidentifiable fragment 108W5 60-90 4 0.4 Mammalian, unidentifiable fragment R108 60-90 4 0.7 Mammalian, unidentifiable fragment (1 thermally-altered) 109 0-30 1 0.4 Mammalian, unidentifiable fragment 110 0-25 1 0.2 Mammalian, unidentifiable fragment 1 2.2 Large mammalian, long bone, shaft, fragment 30-50 112 6 1.1 Mammalian, unidentifiable fragment (1 thermally-altered) 50-70 6 0.4 Mammalian, unidentifiable fragment (3 thermally-altered) 112E5 30-60 2 0.1 Mammalian, unidentifiable fragment 4 0.6 Mammalian, unidentifiable fragment 112W5 30-60 4 0.1 Vertebrata, unidentifiable, fragment 1 0.1 Vertebrata, unidentifiable, fragment, thermally-altered 114 60-70 1 1.3 Geomys bursarius (plains pocket gopher), mandible, right, fragment 5 1.0 Fish, cranium, fragment 25-50 3 0.9 Mammalian, unidentifiable fragment 2 0.3 Vertebrata, unidentifiable, fragment 115 1 0.1 Fish, jaw, fragment 50-75 2 0.2 Mammalian, unidentifiable fragment 1 0.1 Molluscan, shell, fragment

100

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 5 2.7 Mammalian, unidentifiable fragment (1 thermally-altered) 116 70-100 1 0.1 Rodentia, maxilla, fragment R116 30-60 1 - Fire-cracked rock, granitic 30-75 2 0.6 Mammalian, unidentifiable fragment 117 75-100 4 0.6 Mammalian, unidentifiable fragment 1 0.4 Geomys bursarius (plains pocket gopher), mandible, right, fragment 119 75-100 2 1.0 Mammalian, unidentifiable fragment 1 - Decortication, Swan River Chert R119 30-60 1 0 Mammalian, unidentifiable fragment 0-30 1 0.3 Mammalian, unidentifiable fragment 30-40 2 0.6 Mammalian, unidentifiable fragment 120 40-60 4 0.8 Mammalian, unidentifiable fragment, thermally-altered 60-80 1 0.1 Mammalian, unidentifiable fragment, thermally-altered 0-25 2 1.4 Mammalian, unidentifiable fragment 25-50 1 0.4 Mammalian, unidentifiable fragment 122 50-75 2 0.3 Mammalian, unidentifiable fragment 75-100 1 0.2 Mammalian, unidentifiable fragment R122 60-80 1 0.2 Mammalian, unidentifiable fragment, thermally-altered 123 50-75 1 2.0 Castor canadensis (beaver), fibula, distal fragment 0-30 2 0.1 Mammalian, unidentifiable fragment 124 30-40 1 0.1 Mammalian, unidentifiable fragment 80-90 1 49.0 Bison bison (bison), metacarpal, distal fragment R124 60-80 1 0.1 Mammalian, unidentifiable fragment 1 0.2 Canis sp (dogs, wolves, etc…), tooth, incisor, complete 30-75 125 1 0.1 Mammalian, unidentifiable fragment 75-100 1 0.9 Mammalian, unidentifiable fragment 50-70 26 16.5 Cervidae (deer), antler, fragment R125 52-62 13 3.3 Cervidae (deer), antler, fragment 70-80 6 1.0 Mammalian, unidentifiable fragment 126 50-75 1 0.1 Mammalian, unidentifiable fragment 0-30 4 0.8 Mammalian, unidentifiable fragment (1 thermally-altered) 30-60 1 0.3 Mammalian, unidentifiable fragment 3 2.8 Castor canadensis (beaver), vertebra, caudal, fragment 127 30-70 6 7.0 Castor canadensis (beaver), vertebra, fragment 38 6.2 Mammalian, unidentifiable fragment 1 - Fire-cracked rock, granitic 70-100 1 0.3 Mammalian, unidentifiable fragment 0-30 2 0.3 Mammalian, unidentifiable fragment 30-60 6 0.6 Mammalian, unidentifiable fragment 129 7 0.8 Mammalian, unidentifiable fragment 70-100 1 0.1 Rodentia, metapodial, complete 2 0.2 Rodentia, tooth, fragment 0-15 1 0.8 Mammalian, unidentifiable fragment 130 30-40 1 0.1 Mammalian, unidentifiable fragment 50-60 1 18.0 Bison bison (bison), humerus, left, shaft, fragment 131 0-30 3 0.5 Mammalian, unidentifiable fragment 134 100-120 1 0.2 Mammalian, unidentifiable fragment R134 30-60 1 0.2 Mammalian, unidentifiable fragment 136 60-80 1 0.1 Mammalian, unidentifiable fragment 137 30-40 1 0.1 Mammalian, unidentifiable fragment 138 80-100 2 0.4 Mammalian, unidentifiable fragment R138 60-90 1 0.1 Mammalian, unidentifiable fragment 1 0.1 Anaxyrus sp. (toad), ilium, left, fragment 0-30 139 1 0.1 Mammalian, unidentifiable fragment 30-55 22 3.8 Mammalian, unidentifiable fragment

101

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 20-50 3 0.8 Mammalian, unidentifiable fragment (2 thermally-altered) 139E5 50-80 9 2.4 Mammalian, unidentifiable fragment (7 thermally-altered) 80-100 8 2.3 Mammalian, unidentifiable fragment, thermally-altered 40-60 1 1.2 Bison bison (bison), tooth, enamel, fragment 139S5 80-100 1 0.1 Mammalian, unidentifiable fragment, thermally-altered 20-40 1 4.3 Mammalian, unidentifiable fragment 60-80 12 3.6 Mammalian, unidentifiable fragment (4 thermally-altered) 139W5 1 1.1 Mammalian, long bone, shaft, fragment 80-100 10 2.3 Mammalian, unidentifiable fragment (3 thermally-altered) 2 2.6 Large mammalian, unidentifiable fragment R139 30-60 5 1.7 Mammalian, unidentifiable fragment (1 thermally-altered) 1 1.5 Medium/large mammalian, long bone, shaft, fragment, thermally-altered 55-85 24 6.5 Mammalian, unidentifiable fragment (12 thermally-altered) R140 1 28.0 Bison bison (bison), phalanx, complete 1 32.9 Bison bison (bison), tooth, fragment 85-100 1 56.5 Bison bison (bison), tooth, molar, complete 30 6.9 Mammalian, unidentifiable fragment (9 thermally-altered) 1 6.8 Large mammalian, unidentifiable fragment 25-50 1 0.2 Mammalian, unidentifiable fragment, thermally-altered 1 15.3 Bison bison (bison), mandible, left, anterior, fragment R140E5 1 2.3 Bison bison (bison), phalanx, proximal fragment 75-100 1 13.1 Bison bison (bison), tooth, molar, fragment 12 15.2 Mammalian, unidentifiable fragment (2 thermally-altered) 25-50 1 0.3 Mammalian, unidentifiable fragment 50-75 5 0.8 Mammalian, unidentifiable fragment (1 thermally-altered) R140W5 1 0.4 Ictalurus punctatus (channel catfish), cranium, fragment 75-100 2 0.9 Ictalurus punctatus (channel catfish), vertebra, fragment 28 4.1 Mammalian, unidentifiable fragment (1 thermally-altered) 0-25 1 - Bifacial shaping, unidentified chert 141 70-100 26 11.3 Mammalian, unidentifiable fragment (1 thermally-altered) 40-60 4 0.6 Mammalian, unidentifiable fragment (3 thermally-altered) 141E5 60-80 13 1.6 Mammalian, unidentifiable fragment 80-100 29 10.3 Mammalian, unidentifiable fragment 2 35.5 Bison bison (bison), rib, shaft, fragment 80-100 141S5 46 16.1 Mammalian, unidentifiable fragment (19 thermally-altered) 100-130 2 0.3 Mammalian, unidentifiable fragment (1 thermally-altered) 20-40 1 2.1 Mammalian, unidentifiable fragment 141W5 60-80 3 0.3 Mammalian, unidentifiable fragment 1 0.2 Turtle, carapace/plastron, fragment 25-50 1 0.1 Vertebrata, unidentifiable, fragment 1 38.4 Bison bison (bison), tibia, shaft, fragment R141E5 2 2.0 Large mammalian, unidentifiable fragment 50-75 1 0.8 Medium mammalian, vertebra, fragment 11 1.4 Mammalian, unidentifiable fragment (2 thermally-altered) 75-100 1 0.1 Mammalian, unidentifiable fragment 0-30 4 4.5 Mammalian, unidentifiable fragment (2 thermally-altered) 31 35.0 Large mammalian, unidentifiable fragment 30-50 150 46.5 Mammalian, unidentifiable fragment 142 50-70 10 1.1 Mammalian, unidentifiable fragment 1 3.5 Extra large mammalian, unidentifiable fragment 70-100 55 11.4 Mammalian, unidentifiable fragment (3 thermally-altered) 0-30 4 1.4 Mammalian, unidentifiable fragment (3 thermally-altered) 30-50 4 1.4 Mammalian, unidentifiable fragment (2 thermally-altered) 142W5 50-70 3 0.2 Mammalian, unidentifiable fragment (2 thermally-altered) 70-100 3 0.6 Mammalian, unidentifiable fragment 115-125 1 0.2 Mammalian, unidentifiable fragment

102

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 0-30 3 1.4 Mammalian, unidentifiable fragment (1 thermally-altered) 142S5 80-100 2 0.5 Mammalian, unidentifiable fragment, thermally-altered 0-30 3 2.9 Mammalian, unidentifiable fragment (1 thermally-altered) 1 0.1 Bos/Bison sp. (cattle/bison), tooth, enamel, fragment 30-60 2 0.7 Mammalian, unidentifiable fragment 142E5 60-80 24 8.0 Mammalian, unidentifiable fragment, thermally-altered 1 0.7 Castor canadensis (beaver), tooth, incisor, fragment 80-100 22 3.6 Mammalian, unidentifiable fragment (21 thermally-altered) 100-130 2 0.8 Mammalian, unidentifiable fragment 25-50 1 0.1 Mammalian, unidentifiable fragment, thermally-altered 4 0.5 Mammalian, unidentifiable fragment R142 50-75 1 0.1 Vertebrata, unidentifiable, fragment, thermally-altered 75-100 3 0.5 Mammalian, unidentifiable fragment (2 thermally-altered) 18 4.2 Mammalian, unidentifiable fragment (8 thermally-altered) 25-50 2 0.2 Vertebrata, unidentifiable, fragment (1 thermally-altered) 4 1.5 Bison/Cervus sp. (bison/elk), tooth, cheek, fragment 1 - Broken flake, unidentified chert R142E5 2 9.7 Extra large mammalian, long bone, shaft, fragment (1 thermally-altered) 50-75 37 30.9 Large mammalian, unidentifiable fragment (15 thermally-altered) 133 25.3 Mammalian, unidentifiable fragment (69 thermally-altered) 1 0.2 Vertebrata, long bone, shaft, fragment 75-100 4 0.6 Mammalian, unidentifiable fragment (3 thermally-altered) 25-50 1 0.2 Mammalian, unidentifiable fragment R142W5 2 6.3 Large mammalian, long bone, shaft, fragment 50-75 23 4.6 Mammalian, unidentifiable fragment (4 thermally-altered) 0-20 9 4.9 Mammalian, unidentifiable fragment (1 thermally-altered) 0-35 2 0.9 Molluscan, shell, fragment 143 20-50 2 1.1 Mammalian, unidentifiable fragment 1 24.4 Bison bison (bison), rib, fragment 50-75 4 0.6 Mammalian, unidentifiable fragment (3 thermally-altered) 1 0.6 Bison/Cervus sp. (bison/elk), tooth, enamel, fragment 0-20 1 - Broken flake, quartzite 24 8.1 Mammalian, unidentifiable fragment (10 thermally-altered) 143E5 20-40 6 5.1 Mammalian, unidentifiable fragment 1 7.0 Large mammalian, long bone, shaft, fragment 40-60 15 6.8 Mammalian, unidentifiable fragment (1 thermally-altered) 60-80 4 1.6 Mammalian, unidentifiable fragment (2 thermally-altered) 1 5.5 Bison bison (bison), sesamoid, complete 3 - Bifacial shaping, Knife River Flint 1 - Bifacial thinning, Knife River Flint 0-25 1 1.4 Large mammalian, long bone, shaft, fragment 143S5 1 0.2 Mammalian, tooth, enamel, fragment 83 30.0 Mammalian, unidentifiable fragment (52 thermally-altered) 1 - Utilized flake, Knife River Flint 25-50 2 0.7 Mammalian, unidentifiable fragment 50-75 6 3.5 Mammalian, unidentifiable fragment (3 thermally-altered) 0-20 15 5.9 Mammalian, unidentifiable fragment (9 thermally-altered) 1 - Other G4 flake, quartzite 20-40 143W5 5 0.9 Mammalian, unidentifiable fragment (4 thermally-altered) 40-60 18 3.8 Mammalian, unidentifiable fragment (12 thermally-altered) 60-80 4 1.1 Mammalian, unidentifiable fragment (1 thermally-altered) 1 2.0 Extra large mammalian, long bone, shaft, fragment 50-75 R143 4 0.7 Mammalian, unidentifiable fragment (2 thermally-altered) 75-100 1 0.2 Mammalian, unidentifiable fragment, thermally-altered R143E5 75-100 1 0.9 Bison bison (bison), tooth, cheek, fragment 25-50 1 0.1 Mammalian, unidentifiable fragment, thermally-altered R143W5 50-75 2 0.2 Mammalian, unidentifiable fragment (1 thermally-altered)

103

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 0-20 4 0.7 Mammalian, unidentifiable fragment, thermally-altered 20-30 4 1.0 Mammalian, unidentifiable fragment (2 thermally-altered) 144 30-50 2 3.3 Mammalian, unidentifiable fragment 80-100 1 0.2 Mammalian, unidentifiable fragment 1 0.4 Bos/Bison sp. (cattle/bison), tooth, enamel, fragment 0-30 2 0.5 Mammalian, unidentifiable fragment (1 thermally-altered) 144E5 2 5.5 Large mammalian, rib, shaft, fragment 30-50 28 8.0 Mammalian, unidentifiable fragment 50-75 3 0.5 Mammalian, unidentifiable fragment (1 thermally-altered) 0-30 10 2.8 Mammalian, unidentifiable fragment (9 thermally-altered) 1 0.3 Bos/Bison sp. (cattle/bison), tooth, enamel, fragment 144S5 30-60 1 2.7 Large mammalian, rib, shaft, fragment 27 5.7 Mammalian, unidentifiable fragment (5 thermally-altered) 1 - Bifacial thinning, Knife River Flint 1 - Bifacial thinning, Prairie du Chien Chert (oolitic) 0-30 144W5 1 - Broken flake, quartzite 20 4.8 Mammalian, unidentifiable fragment (5 thermally-altered) 30-60 6 1.8 Mammalian, unidentifiable fragment (2 thermally-altered) 1 - Broken flake, Swan River Chert 50-75 R144 2 0.6 Mammalian, unidentifiable fragment (1 thermally-altered) 75-100 1 0.7 Mammalian, unidentifiable fragment 25-50 1 0.4 Large mammalian, unidentifiable fragment 3 2.4 Large mammalian, long bone, shaft, fragment 50-75 4 2.6 Large mammalian, unidentifiable fragment R144E5 12 1.8 Mammalian, unidentifiable fragment (2 thermally-altered) 1 8.8 Large mammalian, long bone, shaft, fragment 75-100 2 0.4 Mammalian, unidentifiable fragment 1 0.1 Vertebrata, unidentifiable, fragment, thermally-altered R144W5 40-50 2 1.2 Large mammalian, unidentifiable fragment 0-10 1 1.8 Mammalian, unidentifiable fragment 10-40 8 2.0 Mammalian, unidentifiable fragment (4 thermally-altered) 145 1 16.8 Bison/Cervus sp. (bison/elk), rib, shaft, fragment 40-70 22 3.5 Mammalian, unidentifiable fragment 70-100 1 1.0 Mammalian, unidentifiable fragment 20-50 8 2.6 Mammalian, unidentifiable fragment (3 thermally-altered) 145E5 50-75 11 2.7 Mammalian, unidentifiable fragment (3 thermally-altered) 75-100 2 0.2 Mammalian, unidentifiable fragment (1 thermally-altered) 1 62.2 Bison bison (bison), calcaneous, right, fragment 145S5 40-60 5 14.9 Bison bison (bison), tooth, fragment 60 116.3 Mammalian, unidentifiable fragment 2 5.4 Large mammalian, long bone, shaft, fragment, thermally-altered 20-40 42 12.1 Mammalian, unidentifiable fragment (5 thermally-altered) 145W5 40-60 22 5.2 Mammalian, unidentifiable fragment (3 thermally-altered) 60-80 11 5.6 Mammalian, unidentifiable fragment 80-100 3 0.4 Mammalian, unidentifiable fragment (1 thermally-altered) 1 1.9 Extra large mammalian, long bone, shaft, fragment R145E5 25-50 1 0.2 Vertebrata, unidentifiable, fragment 1 - Bifacial shaping, Prairie du Chien Chert (oolitic) 5 1.9 Mammalian, unidentifiable fragment (4 thermally-altered) 0-25 - Projectile point (unidentified type) or preform, base fragment, Swan River 1 Chert 146 1 14.8 Bison bison (bison), tooth, premolar, fragment 25-55 6 5.0 Mammalian, unidentifiable fragment, thermally-altered 1 - Utilized flake, Knife River Flint 1 0.3 Large mammalian, tooth, fragment 70-100 6 0.7 Mammalian, unidentifiable fragment (3 thermally-altered)

104

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 1 - Broken flake, rhyolite 35-70 17 17.7 Large mammalian, unidentifiable fragment (2 thermally-altered) 85 12.4 Mammalian, unidentifiable fragment (12 thermally-altered) 1 - Bifacial thinning, Swan River Chert 146E5 1 6.3 Large mammalian, long bone, shaft, fragment 70-100 1 0.9 Large mammalian, tooth, fragment 2 0.2 Mammalian, tooth, fragment (1 thermally-altered) 33 9.8 Mammalian, unidentifiable fragment (2 thermally-altered) 100-130 41 11.3 Mammalian, unidentifiable fragment 1 17.5 Bison bison (bison), tooth, molar, fragment 35-70 2 3.9 Large mammalian, long bone, shaft, fragment, thermally-altered 31 5.6 Mammalian, unidentifiable fragment (5 thermally-altered) 146S5 2 7.2 Extra large mammalian, long bone, shaft, fragment 70-100 102 24.4 Mammalian, unidentifiable fragment (31 thermally-altered) 1 - Side and end scraper, Swan River Chert 100-130 2 0.4 Mammalian, unidentifiable fragment 1 - Bifacial shaping, Knife River Flint 0-30 13 1.7 Mammalian, unidentifiable fragment (11 thermally-altered) 1 56.4 Bison bison (bison), tooth, molar, fragment 1 - Bifacial shaping, Knife River Flint 1 - Other G4 flake, Swan River Chert 30-60 1 5.0 Extra large mammalian, long bone, shaft, fragment, thermally-altered 146W5 9 29.6 Extra large mammalian, unidentifiable fragment 205 41.4 Mammalian, unidentifiable fragment (80 thermally-altered) 1 - Other G4 flake, quartz 60-80 2 0.1 Mammalian, unidentifiable fragment 80-100 4 0.9 Mammalian, unidentifiable fragment (1 thermally-altered) 100-130 1 0.1 Mammalian, unidentifiable fragment, thermally-altered 1 1.2 Extra large mammalian, unidentifiable fragment 25-40 3 0.6 Mammalian, unidentifiable fragment (2 thermally-altered) 40-50 19 5.7 Mammalian, unidentifiable fragment (12 thermally-altered) 1 0.3 Bison/Cervus sp. (bison/elk), tooth, cheek, fragment R146 3 21.2 Extra large mammalian, long bone, shaft, fragment 3 20.3 Extra large mammalian, unidentifiable fragment 50-75 5 3.6 Large mammalian, long bone, shaft, fragment 13 8.2 Large mammalian, unidentifiable fragment (2 thermally-altered) 100 27.3 Mammalian, unidentifiable fragment (23 thermally-altered) 4 3.9 Bison bison (bison), tooth, cheek, fragment 2 14.4 Bison bison (bison), tooth, molar, fragment 1 0.3 Bison/Cervus sp. (bison/elk), tooth, cheek, fragment 2 - Fire-cracked rock, granitic 25-45 4 8.9 Extra large mammalian, long bone, shaft, fragment 3 4.8 Large mammalian, long bone, shaft, fragment (2 thermally-altered) 27 22.6 Large mammalian, unidentifiable fragment (23 thermally-altered) 216 47.2 Mammalian, unidentifiable fragment (165 thermally-altered) 8 8.3 Bison bison (bison), tooth, cheek, fragment R146W5 1 15.1 Bison bison (bison), tooth, molar, fragment 3 0.9 Bison/Cervus sp. (bison/elk), tooth, cheek, fragment 45-70 5 19.1 Extra large mammalian, unidentifiable fragment 8 5.4 Large mammalian, unidentifiable fragment (3 thermally-altered) 73 12.0 Mammalian, unidentifiable fragment (35 thermally-altered) 70-100 2 0.4 Mammalian, unidentifiable fragment (1 thermally-altered)

105

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 0-20 1 0.3 Mammalian, unidentifiable fragment, thermally-altered 20-50 11 1.6 Mammalian, unidentifiable fragment (7 thermally-altered) 50-70 9 2.8 Mammalian, unidentifiable fragment (8 thermally-altered) 1 14.7 Bison bison (bison), femur, left, distal fragment 5 25.7 Bison bison (bison), long bone, shaft, fragment 1 69.2 Bison bison (bison), navicularcuboid, fragment 7 21.7 Bison bison (bison), rib, shaft, fragment 147 1 1.7 Bison bison (bison), tooth, cheek, fragment 60-70 1 22.5 Bison bison (bison), unciform, complete 1 - Shatter, Red River Chert 1 1.2 Extra large mammalian, tooth, enamel, fragment 8 20.2 Extra large mammalian, unidentifiable fragment 135 35.5 Medium/large mammalian, unidentifiable fragment (10 thermally-altered) 1 25.2 Bison bison (bison), tooth, premolar, fragment 80-100 22 6.8 Mammalian, unidentifiable fragment (1 thermally-altered) 1 - Bifacial thinning, Prairie du Chien Chert (oolitic) 20-40 6 5.8 Large mammalian, unidentifiable fragment 86 17.2 Mammalian, unidentifiable fragment (12 thermally-altered) 3 - Bifacial shaping, Prairie du Chien Chert (oolitic) 2 - Other G4 flake, Prairie du Chien Chert (oolitic) 1 - Other G4 flake, quartz 40-60 3 5.1 Extra large mammalian, long bone, shaft, fragment 1 4.6 Extra large mammalian, unidentifiable fragment 45 9.7 Mammalian, unidentifiable fragment (18 thermally-altered) 147E5 1 - Utilized flake, Knife River Flint 1 10.7 Bison bison (bison), mandible, left, fragment 1 2.8 Bison bison (bison), phalanx, proximal fragment 7 34.1 Extra large mammalian, long bone, shaft, fragment (1 thermally-altered) 60-80 1 6.1 Extra large mammalian, petrosal, fragment 2 1.4 Extra large mammalian, tooth, enamel, fragment, , 12 10.0 Large mammalian, unidentifiable fragment (3 thermally-altered) 122 20.0 Mammalian, unidentifiable fragment (33 thermally-altered) 65-70 1 27.2 Bison bison (bison), rib, shaft, fragment 20-40 8 1.6 Mammalian, unidentifiable fragment, thermally-altered 2 - Bifacial shaping, Prairie du Chien Chert (oolitic) 2 - Bifacial shaping, Swan River Chert 40-60 1 - Broken flake, Swan River Chert 147S5 1 - Nonbifacial, unidentified chert 30 3.6 Mammalian, unidentifiable fragment 45-50 1 103.4 Bison bison (bison), tibia, left, distal fragment 60-80 23 3.1 Mammalian, unidentifiable fragment (8 thermally-altered) 80-100 30 3.0 Mammalian, unidentifiable fragment (24 thermally-altered) 20-50 7 1.3 Mammalian, unidentifiable fragment (5 thermally-altered) 1 1.0 Bison bison (bison), tooth, cheek, fragment 50-75 147W5 31 6.8 Mammalian, unidentifiable fragment (26 thermally-altered) 75-100 3 0.5 Mammalian, unidentifiable fragment, thermally-altered 100-130 2 0.3 Mammalian, unidentifiable fragment (1 thermally-altered) 0-25 2 0.6 Mammalian, unidentifiable fragment 1 3.8 Bison bison (bison), tooth, premolar, fragment 1 13.8 Bison/Cervus sp. (bison/elk), long bone, shaft, fragment

1 0.8 Bison/Cervus sp. (bison/elk), tooth, cheek, fragment 25-50 1 - Broken flake, Knife River Flint

1 0.2 Medium/large mammalian, tooth, fragment R147 21 5.7 Mammalian, unidentifiable fragment (15 thermally-altered) 1 0.2 Colubridae (snakes), vertebra, fragment 50-75 6 1.3 Mammalian, unidentifiable fragment (2 thermally-altered)

106

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 1 28.5 Bison bison (bison), tibia, left, distal fragment 3 4.0 Large mammalian, unidentifiable fragment 50-75 5 1.4 Mammalian, unidentifiable fragment (2 thermally-altered) 1 0.1 Vertebrata, long bone, shaft, fragment R147E5 1 0.1 Vertebrata, unidentifiable, fragment, thermally-altered 1 - Bipolar flake, unidentified chert 75-100 1 1.5 Large mammalian, long bone, shaft, fragment 17 3.3 Mammalian, unidentifiable fragment (3 thermally-altered) 2 2.0 Large mammalian, long bone, shaft, fragment, thermally-altered 25-50 20 18.7 Large mammalian, unidentifiable fragment (11 thermally-altered) R147W5 107 21.5 Mammalian, unidentifiable fragment (90 thermally-altered) 1 0.8 Large mammalian, long bone, shaft, fragment, thermally-altered 50-75 20 4.2 Mammalian, unidentifiable fragment (16 thermally-altered) 0-30 1 1.2 Large mammalian, unidentifiable fragment 30-60 1 0.2 Mammalian, unidentifiable fragment 1 1.5 Bison bison (bison), tooth, cheek, fragment 1 - Bifacial shaping, Knife River Flint

1 - Bipolar flake, Knife River Flint

1 6.2 Extra large mammalian, long bone, shaft, fragment 60-100 2 2.4 Large mammalian, long bone, shaft, fragment (1 thermally-altered) 148 8 7.7 Large mammalian, unidentifiable fragment (5 thermally-altered) 116 25.2 Mammalian, unidentifiable fragment (68 thermally-altered) 1 - Side scraper, basaltic 70-90 1 - Projectile point (Besant type), Knife River Flint 1 2.5 Large mammalian, long bone, shaft, fragment 20-50 10 2.6 Mammalian, unidentifiable fragment (6 thermally-altered) 3 0.1 Vertebrata, unidentifiable, fragment, thermally-altered 1 - Bifacial shaping, unidentified chert 148E5 1 - Broken flake, quartzite 50-75 5 17.5 Extra large mammalian, unidentifiable fragment (3 thermally-altered) 139 25.6 Mammalian, unidentifiable fragment (110 thermally-altered) 70-100 14 3.8 Mammalian, unidentifiable fragment (8 thermally-altered) 120-130 2 1.5 Mammalian, unidentifiable fragment 20-40 2 0.4 Mammalian, unidentifiable fragment, thermally-altered 1 9.8 Castor canadensis (beaver), tibia, right, proximal fragment 40-60 31 6.8 Mammalian, unidentifiable fragment 60-80 35 11.4 Mammalian, unidentifiable fragment (8 thermally-altered) 5 9.7 Bison bison (bison), tooth, cheek, fragment 1 6.2 Bison bison (bison), tooth, premolar, complete 1 - Bifacial shaping, Knife River Flint 148S5 80-100 5 6.0 Large mammalian, unidentifiable fragment (1 thermally-altered) 47 10.3 Mammalian, unidentifiable fragment (23 thermally-altered) 1 - Side and end scraper, Knife River Flint 90-95 1 34.5 Bison bison (bison), tibia, right, shaft, fragment 100-105 1 - Bipolar flake, Knife River Flint 1 1.3 Large mammalian, long bone, shaft, fragment 100-130 9 1.0 Mammalian, unidentifiable fragment 1 - Other G4 flake, Swan River Chert 30-50 6 8.8 Large mammalian, unidentifiable fragment (3 thermally-altered) 22 6.0 Mammalian, unidentifiable fragment (17 thermally-altered) R148E5 1 0.4 Bison/Cervus sp. (bison/elk), tooth, fragment 50-75 1 11.9 Extra large mammalian, unidentifiable fragment 10 4.8 Mammalian, unidentifiable fragment (6 thermally-altered)

107

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 20-35 1 0.1 Mammalian, tooth, enamel, fragment 4 2.3 Bison bison (bison), tooth, fragment 4 - Bifacial shaping, Knife River Flint 1 - Broken flake, quartz 35-70 1 - Other G4 flake, Knife River Flint 11 10.2 Large mammalian, unidentifiable fragment (1 thermally-altered) 148W5 1 0.2 Mammalian, tooth, enamel, fragment 118 23.8 Mammalian, unidentifiable fragment (41 thermally-altered) 7 5.7 Bison bison (bison), tooth, fragment 1 49.2 Bison bison (bison), tooth, molar, fragment 70-75 2 7.0 Extra large mammalian, long bone, shaft, fragment 4 5.4 Extra large mammalian, unidentifiable fragment (1 thermally-altered) 64 9.1 Mammalian, unidentifiable fragment (15 thermally-altered) 1 - Nonbifacial, Knife River Flint 5 4.9 Large mammalian, unidentifiable fragment (2 thermally-altered) 25-50 31 8.8 Mammalian, unidentifiable fragment (19 thermally-altered) 2 0.6 Odocoileus virginianus (white-tailed deer), tooth, cheek, fragment 40-50 1 - Manuport, granitic R148 2 9.0 Extra large mammalian, unidentifiable fragment 50-75 7 3.1 Mammalian, unidentifiable fragment (4 thermally-altered) 1 8.2 Extra large mammalian, long bone, shaft, fragment 75-100 1 0.5 Medium/large mammalian, unidentifiable fragment, thermally-altered 90 1 88.9 Bison bison (bison), femur, right, shaft, fragment 1 0.9 Bison bison (bison), tooth, cheek, fragment 5 1.2 Bison/Cervus sp. (bison/elk), tooth, cheek, fragment 2 - Fire-cracked rock, igneous 25-50 1 4.7 Extra large mammalian, long bone, shaft, fragment 5 4.7 Large mammalian, unidentifiable fragment (1 thermally-altered) R148W5 17 3.8 Mammalian, unidentifiable fragment (11 thermally-altered) 2 1.9 Large mammalian, unidentifiable fragment 50-75 24 6.8 Mammalian, unidentifiable fragment (8 thermally-altered) 1 0.1 Bird, long bone, shaft, fragment, thermally-altered 75-100 1 - Bifacial shaping, Knife River Flint 1 0.1 Mammalian, unidentifiable fragment, thermally-altered 1 0.9 Large mammalian, long bone, shaft, fragment, thermally-altered 0-10 3 0.0.8 Mammalian, unidentifiable fragment, thermally-altered 10-30 1 .05 Mammalian, unidentifiable fragment, thermally-altered 2 4.1 Large mammalian, unidentifiable fragment, thermally-altered 30-60 107 28.0 Mammalian, unidentifiable fragment (87 thermally-altered) 149 31 - Fired clay 2 25.1 Extra large mammalian, unidentifiable fragment 60-80 7 12.8 Large mammalian, unidentifiable fragment (2 thermally-altered) 44 11.9 Mammalian, unidentifiable fragment (19 thermally-altered) 80-100 3 0.8 Mammalian, unidentifiable fragment, thermally-altered 20-40 2 0.7 Mammalian, unidentifiable fragment 1 - Fire-cracked rock, basaltic 40-60 1 - Fire-cracked rock, granitic 2 0.3 Mammalian, unidentifiable fragment (1 thermally-altered) 149S5 60-80 44 11.9 Mammalian, unidentifiable fragment (18 thermally-altered) 80-100 2 0.6 Mammalian, unidentifiable fragment (2 thermally-altered) 1 - Other G4 flake, Swan River Chert 120-130 50 15.1 Mammalian, unidentifiable fragment (14 thermally-altered)

108

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 1 - Bifacial shaping, Knife River Flint 1 - Broken flake, Knife River Flint 20-40 1 - Broken flake, Swan River Chert 10 2.2 Mammalian, unidentifiable fragment (6 thermally-altered) 1 0.1 Colubridae (snakes), vertebra, complete 1 - Bipolar flake, igneous 40-60 2 0.7 Mammalian, tooth, enamel, fragment 32 7.8 Mammalian, unidentifiable fragment (17 thermally-altered) 1 38.5 Bison bison (bison), phalanx, complete 40-80 1 6.2 Large mammalian, long bone, shaft, fragment 149E5 36 6.8 Mammalian, unidentifiable fragment (27 thermally-altered) 1 19.8 Bison bison (bison), rib, right, fragment 1 5.3 Bison bison (bison), sesamoid, complete 80-100 1 - Bifacial shaping, Swan River Chert 29 9.0 Mammalian, unidentifiable fragment (3 thermally-altered) 1 - Bifacial shaping, Swan River Chert 1 - Broken flake, Knife River Flint 2 - Nonbifacial, unidentified chert 100-130 2 - Other G4 flake, Swan River Chert 4 18.2 Large mammalian, long bone, shaft, fragment 28 17.5 Mammalian, unidentifiable fragment (3 thermally-altered) 1 0.3 Large mammalian, tooth, enamel, fragment 20-40 1 0.5 Mammalian, tooth, enamel, fragment 89 28.0 Mammalian, unidentifiable fragment (49 thermally-altered) 2 0.3 Mammalian, tooth, enamel, fragment 40-60 149W5 59 12.8 Mammalian, unidentifiable fragment (45 thermally-altered) 60-80 7 2.8 Mammalian, unidentifiable fragment (6 thermally-altered) 1 - Broken flake, Swan River Chert 80-100 35 13.8 Mammalian, unidentifiable fragment (4 thermally-altered) 100-130 28 7.8 Mammalian, unidentifiable fragment 25-50 2 0.4 Mammalian, unidentifiable fragment (1 thermally-altered) 1 28.3 Bison bison (bison), mandible, right, fragment 1 40.6 Bison bison (bison), phalanx, complete 1 5.5 Bison bison (bison), tooth, cheek, fragment R149 50-75 1 3.1 Bison bison (bison), tooth, molar, fragment 2 20.5 Extra large mammalian, unidentifiable fragment 1 1.4 Large mammalian, unidentifiable fragment, thermally-altered 18 2.9 Mammalian, unidentifiable fragment (2 thermally-altered) 1 - Nonbifacial, chalcedony 25-50 2 5.9 Large mammalian, unidentifiable fragment 16 4.0 Mammalian, unidentifiable fragment (1 thermally-altered) R149E5 2 4.2 Large mammalian, unidentifiable fragment 50-75 2 0.1 Mammalian, unidentifiable fragment 5 14.1 Large mammalian, unidentifiable fragment 75-100 16 3.1 Mammalian, unidentifiable fragment 1 14.2 Bison bison (bison), phalanx, proximal fragment 1 - Shatter, Swan River Chert 25-50 2 0.9 Large mammalian, tooth, fragment 6 9.1 Large mammalian, unidentifiable fragment 24 4.5 Mammalian, unidentifiable fragment (2 thermally-altered) R149W5 1 - Broken flake, Swan River Chert 1 - Fire-cracked rock, granitic 50-75 2 1.3 Mammalian, unidentifiable fragment (1 thermally-altered) 1 2.9 Owl, femur, right, shaft, fragment 75-100 8 3.2 Mammalian, unidentifiable fragment (5 thermally-altered)

109

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 0-20 1 1.5 Extra large mammalian, unidentifiable fragment 2 19.2 Extra large mammalian, unidentifiable fragment 20-50 6 2.0 Large mammalian, unidentifiable fragment (1 thermally-altered) 1 66.7 Bison bison (bison), tibia, shaft, fragment 1 - Bifacial shaping, Swan River Chert 150 1 - Broken flake, Swan River Chert 50-80 1 - Other G4 flake, Swan River Chert 17 18.6 Large mammalian, unidentifiable fragment 44 9.6 Mammalian, unidentifiable fragment (4 thermally-altered) 80-100 1 0.2 Mammalian, unidentifiable fragment, thermally-altered 20-40 4 0.8 Mammalian, unidentifiable fragment (2 thermally-altered) 1 1.0 Bison/Cervus sp. (bison/elk), tooth, incisor, fragment 1 - Fire-cracked rock, quartzite 6 70.5 Extra large mammalian, unidentifiable fragment 40-60 120 51.8 Mammalian, unidentifiable fragment (12 thermally-altered) 150E5 1 - Side scraper, basaltic 1 - Utilized flake, Swan River Chert 60-80 20 17.3 Mammalian, unidentifiable fragment (1 thermally-altered) 80-100 4 0.8 Mammalian, unidentifiable fragment 100-110 11 3.4 Mammalian, unidentifiable fragment (2 thermally-altered) 1 31.1 Bison bison (bison), phalanx, fragment 20-40 4 1.3 Mammalian, unidentifiable fragment, thermally-altered 1 2.3 Medium/large mammalian, long bone, shaft, fragment, thermally-altered 40-60 150S5 13 4.8 Mammalian, unidentifiable fragment (8 thermally-altered) 1 - Bifacial thinning, Swan River Chert 80-100 27 6.4 Mammalian, unidentifiable fragment (6 thermally-altered) 100-110 2 0.4 Mammalian, unidentifiable fragment 20-40 7 1.8 Mammalian, unidentifiable fragment (5 thermally-altered) 40-60 2 0.4 Mammalian, unidentifiable fragment (1 thermally-altered) 60-80 33 27.2 Mammalian, unidentifiable fragment 1 5.5 Cerrus elephus (Elk or Wapiti), rib, shaft, fragment 150W5 72-76 1 61.2 Cerrus elephus (Elk or Wapiti), ulna, left, proximal fragment 1 3.3 Extra large mammalian, rib, shaft, fragment 80-100 1 5.7 Mammalian, long bone, shaft, fragment 12 3.6 Mammalian, unidentifiable fragment 25-50 2 0.9 Mammalian, unidentifiable fragment R150 50-75 1 - Broken flake, Swan River Chert 1 - Broken flake, Swan River Chert 1 11.9 Extra large mammalian, rib, shaft, fragment 30-50 20 24.3 Large mammalian, unidentifiable fragment 40 6.8 Mammalian, unidentifiable fragment (2 thermally-altered) R150E5 1 0.3 Molluscan, shell, fragment 1 - Broken flake, Swan River Chert 2 3.2 Large mammalian, unidentifiable fragment 50-75 5 0.8 Mammalian, unidentifiable fragment 3 0.5 Turtle, carapace/plastron, fragment 1 - Fire-cracked rock, granitic 30-50 R150W5 2 1.3 Large mammalian, unidentifiable fragment 75-100 1 0.3 Mammalian, unidentifiable fragment 1 - Decortication, Swan River Chert 1 - Other G4 flake, Swan River Chert 20-40 1 0.1 Mammalian, tooth, fragment 151E5 15 2.5 Mammalian, unidentifiable fragment (9 thermally-altered) 40-60 3 0.2 Mammalian, unidentifiable fragment (1 thermally-altered) 80-100 14 3.3 Mammalian, unidentifiable fragment (2 thermally-altered) 115-120 1 0.8 Mammalian, unidentifiable fragment

110

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 0-10 2 0.5 Mammalian, unidentifiable fragment 2 32.6 Bos/Bison sp. (cattle/bison), cranium, fragment 1 32.9 Bos/Bison sp. (cattle/bison), petrosal, fragment 1 - Bifacial shaping, Swan River Chert 10-40 2 - Broken flake, Swan River Chert 13 24.9 Large mammalian, unidentifiable fragment 66 15.0 Mammalian, unidentifiable fragment (1 thermally-altered) 1 0.8 Bison/Cervus sp. (bison/elk), tooth, fragment 40-70 5 1.1 Mammalian, unidentifiable fragment (2 thermally-altered) 151 2 14.8 Bison bison (bison), rib, proximal fragment 11 36.0 Bison bison (bison), rib, shaft, fragment 1 0.4 Canis sp (dogs, wolves, etc…), fibula, distal fragment 1 - Broken flake, Swan River Chert 1 - Decortication, unidentified material 70-100 3 46.4 Extra large mammalian, long bone, shaft, fragment 16 56.9 Extra large mammalian, unidentifiable fragment 59 59.0 Large mammalian, unidentifiable fragment 280 56.5 Mammalian, unidentifiable fragment (45 thermally-altered) 2 0.4 Turtle, carapace/plastron, fragment 20-40 6 1.9 Mammalian, unidentifiable fragment (4 thermally-altered) 1 30.3 Bison bison (bison), lunate, complete 151S5 40-60 13 7.8 Mammalian, unidentifiable fragment (4 thermally-altered) 60-80 3 0.4 Mammalian, unidentifiable fragment (1 thermally-altered) 1 - Fire-cracked rock, granitic 20-40 10 2.4 Mammalian, unidentifiable fragment (5 thermally-altered) 1 67.0 Bison bison (bison), tibia, left, proximal fragment 151W5 1 - Nonbifacial, Swan River Chert 80-100 2 16.5 Extra large mammalian, unidentifiable fragment 22 10.0 Mammalian, unidentifiable fragment 100-130 3 10.4 Mammalian, unidentifiable fragment 1 4.0 Extra large mammalian, phalanx, distal fragment 25-50 3 4.1 Large mammalian, unidentifiable fragment 5 1.2 Mammalian, unidentifiable fragment 1 15.3 Bison bison (bison), phalanx, complete 30-50 7 8.4 Large mammalian, unidentifiable fragment 6 2.9 Mammalian, unidentifiable fragment (1 thermally-altered) R151 1 5.8 Bison bison (bison), ulna, left, proximal fragment 50-75 3 1.8 Mammalian, unidentifiable fragment (1 thermally-altered) 1 12.3 Bison bison (bison), vertebra, centrum epiphysis, fragment 65-80 1 9.4 Bison bison (bison), vertebra, fragment 6 3.4 Mammalian, unidentifiable fragment 1 4.0 Bison bison (bison), vertebra, centrum fragment 75-100 23 5.1 Mammalian, unidentifiable fragment 30-50 4 1.2 Mammalian, unidentifiable fragment (2 thermally-altered) R151E5 75-100 9 1.7 Mammalian, unidentifiable fragment (1 thermally-altered) 2 10.6 Extra large mammalian, long bone, shaft, fragment 25-50 3 0.6 Mammalian, unidentifiable fragment R151W5 1 5.8 Large mammalian, long bone, shaft, fragment 50-75 23 5.8 Mammalian, unidentifiable fragment (21 thermally-altered) 1 0.1 Vertebrata, unidentifiable, fragment, thermally-altered 0-30 1 0.7 Mammalian, unidentifiable fragment, thermally-altered 1 1.2 Large mammalian, unidentifiable fragment 30-60 20 4.0 Mammalian, unidentifiable fragment (18 thermally-altered) 152 2 1.3 Large mammalian, long bone, shaft, fragment, thermally-altered 60-90 8 2.4 Mammalian, unidentifiable fragment (5 thermally-altered) 90-100 1 1.0 Large mammalian, long bone, shaft, fragment

111

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 20-40 7 3.3 Mammalian, unidentifiable fragment (2 thermally-altered) 152E5 40-60 15 4.8 Mammalian, unidentifiable fragment (9 thermally-altered) 60-70 1 0.7 Mammalian, unidentifiable fragment, thermally-altered 1 - Bifacial shaping, Knife River Flint 20-40 3 0.5 Mammalian, unidentifiable fragment, thermally-altered 40-60 5 1.6 Mammalian, unidentifiable fragment (4 thermally-altered) 152S5 40-80 4 0.5 Mammalian, unidentifiable fragment (2 thermally-altered) 1 1.0 Bifacial shaping, Swan River Chert 60-80 4 - Mammalian, unidentifiable fragment 1 0.1 Mammalian, tooth, enamel, fragment 20-40 22 4.1 Mammalian, unidentifiable fragment (8 thermally-altered) 152W5 40-60 6 1.7 Mammalian, unidentifiable fragment (4 thermally-altered) 60-80 1 0.5 Mammalian, long bone, shaft, fragment 80-100 21 17.6 Mammalian, unidentifiable fragment 25-50 3 1.1 Mammalian, unidentifiable fragment (1 thermally-altered) R152 50-75 1 0.6 Mammalian, unidentifiable fragment 75-100 2 1.0 Mammalian, unidentifiable fragment 25-50 6 1.5 Mammalian, unidentifiable fragment (3 thermally-altered) 1 11.9 Extra large mammalian, unidentifiable fragment

50-75 2 2.0 Large mammalian, long bone, shaft, fragment

5 1.5 Mammalian, unidentifiable fragment R152E5 2 2.3 Large mammalian, long bone, shaft, fragment 75-100 3 0.6 Mammalian, unidentifiable fragment, thermally-altered 25-50 1 0.2 Mammalian, unidentifiable fragment, thermally-altered R152W5 75-100 12 4.0 Mammalian, unidentifiable fragment (3 thermally-altered) 80-85 1 127.6 Bison bison (bison), tibia, right, shaft, fragment 0-20 9 2.2 Mammalian, unidentifiable fragment (8 thermally-altered) 20-30 8 2.8 Mammalian, unidentifiable fragment, thermally-altered 153 50-80 3 6.5 Large mammalian, unidentifiable fragment 80-100 1 1.7 Medium/large mammalian, unidentifiable fragment 153E5 0-35 1 0.1 Mammalian, unidentifiable fragment, thermally-altered 153S5 0-20 1 0.5 Mammalian, unidentifiable fragment 1 0.7 Extra large mammalian, tooth, enamel, fragment, , 0-20 1 0.1 Mammalian, unidentifiable fragment 20-40 3 0.5 Mammalian, unidentifiable fragment 153W5 40-60 29 7.9 Mammalian, unidentifiable fragment (18 thermally-altered) 60-80 1 0.5 Mammalian, unidentifiable fragment 80-100 7 2.9 Mammalian, unidentifiable fragment (3 thermally-altered) 1 - Bifacial thinning, Knife River Flint 25-50 R153 2 0.5 Mammalian, unidentifiable fragment (1 thermally-altered) 50-75 1 0.4 Mammalian, unidentifiable fragment R153E5 25-50 1 0.2 Mammalian, unidentifiable fragment, thermally-altered R153W5 50-75 2 1.4 Mammalian, unidentifiable fragment 0-25 1 0.1 Mammalian, unidentifiable fragment 1 1.5 Large mammalian, unidentifiable fragment, thermally-altered 25-50 3 0.4 Mammalian, unidentifiable fragment 154 50-75 1 0.1 Mammalian, unidentifiable fragment 1 1.2 Large mammalian, long bone, shaft, fragment 75-100 2 0.2 Mammalian, unidentifiable fragment 25-50 3 0.5 Mammalian, unidentifiable fragment (1 thermally-altered) R154 75-100 3 0.4 Mammalian, unidentifiable fragment 25-50 4 0.8 Mammalian, unidentifiable fragment R154E5 50-50 1 24.8 Bison bison (bison), unciform, fragment

112

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 1 - Bifacial shaping, Swan River Chert 2 - Bifacial thinning, Swan River Chert 25-50 2 - Broken flake, Swan River Chert 1 - Other G4 flake, Swan River Chert R154W5 6 0.6 Mammalian, unidentifiable fragment 1 - Bifacial thinning, Swan River Chert 50-75 1 1.5 Large mammalian, unidentifiable fragment 1 - Bifacial shaping, Swan River Chert 75-100 1 0.1 Mammalian, unidentifiable fragment 155 0-30 1 0.2 Mammalian, unidentifiable fragment R155 75-100 1 1.7 Large mammalian, long bone, shaft, fragment 50-75 1 0.1 Mammalian, unidentifiable fragment R155W5 75-100 1 0.9 Large mammalian, unidentifiable fragment 1 - Bifacial shaping, unidentified chert 156 60-90 1 - Other G4 flake, Swan River Chert R156W5 50-75 2 0.5 Mammalian, unidentifiable fragment 158 75-100 1 - Other G4 flake, basaltic 1 86.5 Bison bison (bison), radius, left, shaft, fragment R158 25-35 4 4.2 Mammalian, unidentifiable fragment 1 0.2 Medium/large mammalian, tooth, enamel, fragment 0-30 1 0.4 Mammalian, unidentifiable fragment 159 30-60 12 1.7 Mammalian, unidentifiable fragment (4 thermally-altered) 12 1.6 Mammalian, unidentifiable fragment (4 thermally-altered) 60-90 1 - Biface, stage 4, Swan River Chert 1 - Broken flake, Knife River Flint 20-40 14 3.8 Mammalian, unidentifiable fragment (13 thermally-altered) 1 4.0 Bison bison (bison), sesamoid, fragment 40-60 39 8.3 Mammalian, unidentifiable fragment (13 thermally-altered) 1 - Side scraper, Swan River Chert 2 1.8 Bison bison (bison), tooth, enamel, fragment 159E5 60-80 1 0.7 Large mammalian, unidentifiable fragment 39 9.3 Mammalian, unidentifiable fragment (16 thermally-altered) 80-100 33 6.0 Mammalian, unidentifiable fragment (6 thermally-altered) 4 - Bifacial shaping, Swan River Chert 5 - Broken flake, Swan River Chert 120-130 2 - Other G4 flake, Swan River Chert 2 1.7 Mammalian, unidentifiable fragment 20-40 1 0.3 Mammalian, unidentifiable fragment, thermally-altered 60-80 7 1.8 Mammalian, unidentifiable fragment (4 thermally-altered) 159N5 80-100 4 0.4 Mammalian, unidentifiable fragment (3 thermally-altered) 2 - Bifacial shaping, Swan River Chert 100-130 5 1.2 Mammalian, unidentifiable fragment (3 thermally-altered) 20-40 38 28.1 Mammalian, unidentifiable fragment 40-60 16 5.1 Mammalian, unidentifiable fragment (7 thermally-altered) 159S5 60-80 4 0.8 Mammalian, unidentifiable fragment 80-100 7 3.5 Mammalian, unidentifiable fragment 100-110 3 0.4 Mammalian, unidentifiable fragment 20-40 23 3.3 Mammalian, unidentifiable fragment (2 thermally-altered) 40-60 11 2.9 Mammalian, unidentifiable fragment (5 thermally-altered) 60-80 1 0.6 Mammalian, unidentifiable fragment 2 18.5 Bison bison (bison), phalanx, proximal fragment 159W5 1 20.6 Bison bison (bison), scapula, left, fragment 80-100 1 56.4 Bison bison (bison), tibia, right, proximal fragment 78 18.5 Mammalian, unidentifiable fragment (25 thermally-altered) 100-110 12 2.0 Mammalian, unidentifiable fragment (6 thermally-altered) 110-120 6 1.1 Mammalian, unidentifiable fragment (5 thermally-altered)

113

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 2 0.3 Large mammalian, tooth, fragment 0-20 7 1.5 Mammalian, unidentifiable fragment 1 - Bifacial shaping, Knife River Flint 1 - Bifacial thinning, Knife River Flint 1 - Bipolar flake, Knife River Flint 1 - Other G4 flake, quartz 20-35 1 - Fire-cracked rock, granitic 2 0.4 Large mammalian, tooth, fragment 6 5.2 Large mammalian, unidentifiable fragment (2 thermally-altered) 160 30 6.0 Mammalian, unidentifiable fragment (15 thermally-altered) 1 - Shatter, Swan River Chert 1 0.2 Large mammalian, tooth, fragment 35-65 8 5.4 Large mammalian, unidentifiable fragment 44 6.3 Mammalian, unidentifiable fragment (15 thermally-altered) 65-85 1 0.7 Mammalian, unidentifiable fragment 2 - Bifacial thinning, Swan River Chert 85-105 4 27.6 Extra large mammalian, unidentifiable fragment 9 2.5 Mammalian, unidentifiable fragment (2 thermally-altered) 0-20 1 - Broken flake, quartz 1 - Bifacial shaping, Knife River Flint 40-60 3 0.6 Mammalian, tooth, enamel, fragment 67 18.8 Mammalian, unidentifiable fragment (42 thermally-altered) 1 5.6 Extra large mammalian, long bone, shaft, fragment 60-80 60 30.0 Mammalian, unidentifiable fragment (14 thermally-altered) 160N5 1 5.8 Bison bison (bison), phalanx, distal fragment 42 17.8 Mammalian, unidentifiable fragment (8 thermally-altered) 80-100 22.8 Odocoileus virginianus (white-tailed deer), metacarpal, distal fragment, 1 worked/tool 1 0.1 Mammalian, tooth, enamel, fragment 100-130 12 3.9 Mammalian, unidentifiable fragment 1 - Broken flake, Swan River Chert 0-30 2 - Fire-cracked rock, granitic 161 4 1.1 Mammalian, unidentifiable fragment (1 thermally-altered) 30-60 1 0.1 Mammalian, unidentifiable fragment 90-100 25 3.1 Mammalian, unidentifiable fragment 0-30 1 0.2 Mammalian, unidentifiable fragment 7 16.1 Extra large mammalian, long bone, shaft, fragment 162 80-90 10 24.9 Extra large mammalian, unidentifiable fragment 68 7.5 Mammalian, unidentifiable fragment 80-100 2 0.6 Mammalian, unidentifiable fragment 0-30 1 0.3 Mammalian, unidentifiable fragment, thermally-altered 30-45 3 0.2 Mammalian, unidentifiable fragment, thermally-altered 1 - Broken flake, Knife River Flint 1 5.0 Extra large mammalian, long bone, shaft, fragment 163 45-75 10 6.7 Large mammalian, unidentifiable fragment 47 7.1 Mammalian, unidentifiable fragment (21 thermally-altered) 2 - Utilized flake, Knife River Flint 75-100 14 2.6 Mammalian, unidentifiable fragment (1 thermally-altered) 1 - Bifacial shaping, Knife River Flint 0-30 3 0.5 Mammalian, unidentifiable fragment 1 24.9 Bison bison (bison), metacarpal, proximal fragment 40 127.3 Extra large mammalian, unidentifiable fragment 30-60 164 1 0.2 Mammalian, tooth, enamel, fragment 135 30.3 Mammalian, unidentifiable fragment (49 thermally-altered) 1 - Bifacial thinning, fusilinid chert 60-90 2 2.4 Large mammalian, unidentifiable fragment 9 1.8 Mammalian, unidentifiable fragment (1 thermally-altered)

114

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 30-60 2 0.8 Mammalian, unidentifiable fragment (1 thermally-altered) 165 60-90 1 0.9 Mammalian, unidentifiable fragment 0-30 1 - Broken flake, unidentified chert 30-60 4 0.5 Mammalian, unidentifiable fragment (3 thermally-altered) 166 60-80 2 0.3 Mammalian, unidentifiable fragment 80-100 1 0.2 Mammalian, unidentifiable fragment 167 0-25 1 - Fire-cracked rock, granitic 0-30 1 0.3 Mammalian, unidentifiable fragment, thermally-altered 168 30-60 3 0.4 Mammalian, unidentifiable fragment (1 thermally-altered) 60-80 2 0.6 Mammalian, unidentifiable fragment (1 thermally-altered) 55-65 2 56.4 Bison bison (bison), tooth, molar, fragment 169 85-100 6 0.3 Mammalian, unidentifiable fragment 0-25 1 0.2 Mammalian, unidentifiable fragment, thermally-altered 170 25-50 1 0.1 Mammalian, unidentifiable fragment, thermally-altered 30-60 6 0.1 Mammalian, unidentifiable fragment 171 80-100 4 0.4 Mammalian, unidentifiable fragment, thermally-altered 172 60-80 5 0.7 Mammalian, unidentifiable fragment ( 25-50 3 0.6 Mammalian, unidentifiable fragment 173 50-75 1 0.1 Mammalian, unidentifiable fragment, thermally-altered 174 60-90 9 3.7 Mammalian, unidentifiable fragment 0-30 1 0.1 Mammalian, unidentifiable fragment 175 60-80 6 1.6 Mammalian, unidentifiable fragment (1 thermally-altered) 0-30 3 3.0 Mammalian, unidentifiable fragment 176 75-100 2 - Fire-cracked rock, granitic 0-30 2 0.3 Mammalian, unidentifiable fragment 177 30-60 1 0.1 Mammalian, unidentifiable fragment 0-30 2 0.6 Mammalian, unidentifiable fragment 178 80-100 1 0.4 Mammalian, unidentifiable fragment 2 - Fire-cracked rock, quartzite 179 0-25 1 0.3 Mammalian, unidentifiable fragment 1 3.0 Mammalian, long bone, shaft, fragment 0-30 6 0.5 Mammalian, unidentifiable fragment 30-60 9 0.9 Mammalian, unidentifiable fragment 180 60-80 4 0.4 Mammalian, unidentifiable fragment 1 0.1 Mammalian, unidentifiable fragment 80-100 1 0.3 Mammalian, unidentifiable fragment, thermally-altered 0-30 1 0.4 Mammalian, unidentifiable fragment 181 30-60 2 0.3 Mammalian, unidentifiable fragment, thermally-altered 60-80 10 2.3 Mammalian, unidentifiable fragment (2 thermally-altered) 0-30 1 - Bipolar flake, quartz 182 0-30 2 0.6 Mammalian, unidentifiable fragment 183 60-80 2 0.3 Mammalian, unidentifiable fragment 184 0-30 1 0.4 Mammalian, unidentifiable fragment 0-30 6 1.0 Mammalian, unidentifiable fragment 186 70-100 2 0.3 Mammalian, unidentifiable fragment 187 0-30 1 0.1 Mammalian, unidentifiable fragment 188 0-30 1 0.4 Mammalian, unidentifiable fragment 189 60-90 2 0.1 Mammalian, unidentifiable fragment 190 70-100 4 0.5 Mammalian, unidentifiable fragment 50-75 6 0.7 Mammalian, unidentifiable fragment 193 75-100 2 0.5 Mammalian, unidentifiable fragment 194 30-60 1 0.3 Mammalian, unidentifiable fragment 198 80-100 1 0.1 Mammalian, unidentifiable fragment 60-80 1 0.2 Mammalian, unidentifiable fragment 199 80-100 1 0.2 Mammalian, unidentifiable fragment

115

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 30-60 2 0.1 Mammalian, unidentifiable fragment 200 60-80 1 0.1 Mammalian, unidentifiable fragment 80-100 1 0.2 Mammalian, unidentifiable fragment 2 - Ceramic, grit temper, cord marked (partially smoothed-over) 6 - Ceramic, grit temper, smooth (two are polished) 0-25 2 - Ceramic, grit temper, undetermined 201 1 - Bifacial shaping, Knife River Flint 41 10.7 Mammalian, unidentifiable fragment (2 thermally-altered) 25-50 8 1.1 Mammalian, unidentifiable fragment 80-100 1 0.6 Mammalian, unidentifiable fragment 1 - Ceramic, grit temper, cord marked (partially smoothed-over) 0-30 13 2.2 Mammalian, unidentifiable fragment 202 30-65 12 0.9 Mammalian, unidentifiable fragment (10 thermally-altered) 1 - Ceramic, grit temper, smooth (polished) 70-100 11 3.2 Mammalian, unidentifiable fragment (1 thermally-altered) 0-30 1 0.2 Mammalian, unidentifiable fragment 30-60 9 2.7 Mammalian, unidentifiable fragment 203 60-80 3 0.6 Mammalian, unidentifiable fragment (1 thermally-altered) 80-100 1 - Side and end scraper, Knife River Flint 0-25 1 0.1 Mammalian, unidentifiable fragment 204 25-50 13 1.7 Mammalian, unidentifiable fragment (1 thermally-altered) 50-75 2 0.3 Mammalian, unidentifiable fragment 205 0-35 1 0.2 Mammalian, unidentifiable fragment, thermally-altered 0-30 2 0.2 Mammalian, unidentifiable fragment 1 1.0 Castor canadensis (beaver), metapodial, complete 2 0.3 Castor canadensis (beaver), metapodial, fragment 60-80 1 0.5 Castor canadensis (beaver), phalanx, complete 207 11 2.8 Mammalian, unidentifiable fragment 1 1.2 Castor canadensis (beaver), astragalus, complete 2 1.5 Castor canadensis (beaver), metapodial, fragment 80-100 1 3.3 Castor canadensis (beaver), pelvis, left, fragment 7 0.8 Mammalian, unidentifiable fragment 50-65 2 0.9 Mammalian, unidentifiable fragment, thermally-altered 208 65-100 5 1.5 Mammalian, unidentifiable fragment (4 thermally-altered) 0-30 5 1.1 Mammalian, unidentifiable fragment (3 thermally-altered) 30-60 1 0.1 Medium/large mammalian, tooth, fragment 209 1 4.6 Castor canadensis (beaver), ulna, left, fragment 80-100 4 1.0 Mammalian, unidentifiable fragment (1 thermally-altered) 0-35 3 1.0 Mammalian, unidentifiable fragment (2 thermally-altered) 1 0.4 Lepus/Sylvilagus sp. (hare/rabbit), pelvis, left, fragment 35-70 2 0.9 Mammalian, unidentifiable fragment, thermally-altered 1 151.9 Bison bison (bison), radius, right, proximal fragment 210 1 - Other G4 flake, Swan River Chert 70-100 1 - Fired clay 153 57.4 Mammalian, unidentifiable fragment (32 thermally-altered) 80-90 1 - Fire-cracked rock, granitic 30-60 12 7.0 Mammalian, unidentifiable fragment, thermally-altered 211 60-80 5 0.5 Mammalian, unidentifiable fragment 80-100 1 0.1 Mammalian, unidentifiable fragment, thermally-altered 0-30 1 0.4 Mammalian, unidentifiable fragment, thermally-altered 212 30-60 2 0.7 Mammalian, unidentifiable fragment (1 thermally-altered) 80-100 3 1.1 Mammalian, unidentifiable fragment 1 0.7 Bos/Bison sp. (cattle/bison), tooth, cheek, fragment 0-35 1 - Nonbifacial, quartzite 213 2 0.5 Mammalian, unidentifiable fragment (1 thermally-altered) 35-70 1 0.1 Mammalian, unidentifiable fragment 70-100 18 6.0 Mammalian, unidentifiable fragment (8 thermally-altered)

116

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 0-30 2 0.5 Mammalian, unidentifiable fragment, thermally-altered 30-60 4 2.0 Mammalian, unidentifiable fragment 214 60-80 4 1.0 Mammalian, unidentifiable fragment (1 thermally-altered) 80-100 54 18.9 Mammalian, unidentifiable fragment (31 thermally-altered) 216 60-80 41 - Fired clay 217 30-60 12 2.1 Mammalian, unidentifiable fragment (4 thermally-altered) 30-60 6 0.8 Mammalian, unidentifiable fragment 218 80-100 2 0.3 Mammalian, unidentifiable fragment, thermally-altered 0-35 4 0.8 Mammalian, unidentifiable fragment 1 - Ceramic, grit temper, smooth 219 35-65 2 0.8 Mammalian, unidentifiable fragment (1 thermally-altered) 65-75 2 0.5 Mammalian, unidentifiable fragment 220 0-30 1 0.1 Mammalian, unidentifiable fragment 30-60 5 1.3 Mammalian, unidentifiable fragment 221 60-80 4 1.3 Mammalian, unidentifiable fragment 80-100 1 0.2 Mammalian, unidentifiable fragment 223 0-30 2 0.2 Mammalian, unidentifiable fragment 224 0-30 1 0.9 Mammalian, unidentifiable fragment 30-50 3 100.5 Bison bison (bison), magnum, phalanx, & scaphoid, all are complete 225 50-75 1 0.9 Mammalian, unidentifiable fragment 0-20 3 0.5 Mammalian, unidentifiable fragment 35-60 5 0.8 Mammalian, unidentifiable fragment, thermally-altered 226 60-80 5 0.8 Mammalian, unidentifiable fragment, thermally-altered 1 1.2 Large mammalian, long bone, shaft, fragment 80-100 2 1.8 Mammalian, unidentifiable fragment 1 0.2 Mammalian, unidentifiable fragment 35-65 227 1 0.1 Sphaeriidae (freshwater bivalve molluscs), shell, fragment 65-100 5 0.6 Mammalian, unidentifiable fragment 0-20 1 0.9 Mammalian, unidentifiable fragment 228 20-40 1 0.2 Mammalian, unidentifiable fragment 229 0-30 3 1.0 Mammalian, unidentifiable fragment 0-30 3 1.4 Mammalian, unidentifiable fragment 30-60 2 1.9 Mammalian, unidentifiable fragment, thermally-altered 231 60-80 1 0.1 Mammalian, unidentifiable fragment 80-100 2 1.1 Mammalian, unidentifiable fragment 1 - Ceramic, grit temper, smooth 0-20 1 0.1 Mammalian, unidentifiable fragment 232 1 0.1 Mammalian, unidentifiable fragment 20-45 1 0.6 Ondatra zibethicus (muskrat), vertebra, lumbar, fragment 0-30 2 0.2 Mammalian, unidentifiable fragment 234 30-60 1 0.1 Mammalian, unidentifiable fragment 235 0-20 1 0.1 Mammalian, unidentifiable fragment 237 0-30 1 0.1 Mammalian, unidentifiable fragment 238 80-90 1 7.9 Bison bison (bison), magnum, fragment 239 80-100 2 1.5 Mammalian, unidentifiable fragment 0-30 1 0.3 Mammalian, unidentifiable fragment 240 1 0.3 Fish, cranium, fragment 35-70 1 0.1 Mammalian, unidentifiable fragment 0-30 2 0.1 Mammalian, unidentifiable fragment 241 30-60 2 0.4 Mammalian, unidentifiable fragment 80-100 2 0.2 Mammalian, unidentifiable fragment 40-60 1 0.1 Mammalian, unidentifiable fragment 242 80-100 2 0.3 Mammalian, unidentifiable fragment (1 thermally-altered) 243 35-70 8 8.1 Mammalian, unidentifiable fragment 0-30 1 0.5 Molluscan, shell, fragment 244 70-100 1 0.2 Mammalian, unidentifiable fragment 245 70-100 2 0.1 Mammalian, unidentifiable fragment

117

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 0-30 2 0.4 Mammalian, unidentifiable fragment (1 thermally-altered) 30-50 1 0.2 Mammalian, unidentifiable fragment 248 5 - Fired clay 50-70 1 18.8 Extra large mammalian, unidentifiable fragment 0-30 2 1.4 Mammalian, unidentifiable fragment (1 thermally-altered) 60-80 24 4.8 Mammalian, unidentifiable fragment (2 thermally-altered) 249 1 24.0 Bison bison (bison), metapodial, distal fragment 80-100 10 2.3 Mammalian, unidentifiable fragment (7 thermally-altered) 1 - Utilized flake, Swan River Chert 35-70 1 - Bifacial shaping, unidentified chert 1 12.9 Bison bison (bison), cuneiform, complete 1 16.8 Bison bison (bison), lunate, complete 1 6.8 Bison bison (bison), pisiform, complete 250 1 69.9 Bison bison (bison), radius, right, distal fragment 70-100 1 17.8 Bison bison (bison), scaphoid, complete 1 4.3 Bison bison (bison), ulna, right, distal fragment 1 0.6 Large mammalian, tooth, fragment 30 4.8 Mammalian, unidentifiable fragment (10 thermally-altered) 0-35 2 0.4 Mammalian, unidentifiable fragment (1 thermally-altered) 2 1.5 Large mammalian, unidentifiable fragment 251 10 1.9 Mammalian, unidentifiable fragment (2 thermally-altered) 70-100 Odocoileus virginianus (white-tailed deer), metapodial, distal fragment, 1 1.1 thermally-altered 0-20 2 0.2 Mammalian, unidentifiable fragment 252 80-100 1 - Projectile point (Duncan-Hanna type), Knife River Flint 0-35 2 0.8 Mammalian, unidentifiable fragment (1 thermally-altered) 35-70 1 0.3 Mammalian, unidentifiable fragment 253 50-70 3 0.3 Mammalian, unidentifiable fragment 1 2.3 Medium/large mammalian, unidentifiable fragment 70-100 8 0.9 Mammalian, unidentifiable fragment (1 thermally-altered) 35-70 1 0.1 Mammalian, unidentifiable fragment 254 70-100 5 1.5 Mammalian, unidentifiable fragment 0-30 3 0.6 Mammalian, unidentifiable fragment 255 30-60 10 1.3 Mammalian, unidentifiable fragment 25-55 1 - Fire-cracked rock, quartzite 256 1 - Fire-cracked rock, granitic 55-80 1 0.2 Mammalian, unidentifiable fragment 0-30 1 0.4 Mammalian, unidentifiable fragment 257 30-50 3 0.9 Mammalian, unidentifiable fragment 0-30 1 0.1 Mammalian, unidentifiable fragment 1 0.1 Mammalian, unidentifiable fragment 258 50-75 1 0.1 Rodentia, femur, distal fragment 1 0.1 Rodentia, vertebra, thoracic, fragment 75-100 2 1.2 Mammalian, unidentifiable fragment 259 0-30 5 0.9 Mammalian, unidentifiable fragment 1 - Fire-cracked rock, granitic 0-30 1 0.4 Large mammalian, tooth, fragment 260 4 1.0 Mammalian, unidentifiable fragment 30-60 35 7.8 Mammalian, unidentifiable fragment 60-80 1 0.3 Mammalian, unidentifiable fragment 30-50 7 2.3 Mammalian, unidentifiable fragment 1 - Shatter, Red River Chert 261 75-100 1 0.8 Large mammalian, unidentifiable fragment 2 0.3 Mammalian, unidentifiable fragment (1 thermally-altered) 0-30 3 0.9 Mammalian, unidentifiable fragment 262 30-50 2 0.4 Mammalian, unidentifiable fragment

118

Table 16. Continued. Shovel Weight Depth Count Artifact Type Test (g) 0-30 3 0.5 Mammalian, unidentifiable fragment 263 30-60 1 0.1 Mammalian, unidentifiable fragment 70-100 1 0.2 Mammalian, unidentifiable fragment 25-50 1 0.3 Mammalian, unidentifiable fragment 264 50-75 1 0.3 Mammalian, unidentifiable fragment 30-50 1 0.1 Mammalian, unidentifiable fragment 1 - Nonbifacial, Tongue River Silica 265 50-75 5 2.6 Large mammalian, unidentifiable fragment 7 0.9 Mammalian, unidentifiable fragment (2 thermally-altered) 75-100 3 0.4 Mammalian, unidentifiable fragment 30-50 3 0.6 Mammalian, unidentifiable fragment 267 8 3.8 Large mammalian, unidentifiable fragment 75-100 6 0.7 Mammalian, unidentifiable fragment 1 4.7 Large mammalian, vertebra, lumbar, fragment 0-25 1 1.6 Medium/large mammalian, unidentifiable fragment 2 0.3 Mammalian, unidentifiable fragment 268 1 0.5 Ictalurus punctatus (channel catfish), pectoral spine, shaft, fragment 25-50 2 6.0 Large mammalian, unidentifiable fragment 4 0.8 Mammalian, unidentifiable fragment 0-30 1 0.1 Mammalian, unidentifiable fragment 1 0.3 Geomys bursarius (plains pocket gopher), mandible, left, fragment 269 30-60 1 0.4 Geomys bursarius (plains pocket gopher), mandible, right, fragment 80-100 1 0.1 Vertebrata, unidentifiable, fragment 1 1.1 Large mammalian, unidentifiable fragment 0-30 8 1.1 Mammalian, unidentifiable fragment 270 30-50 4 0.5 Mammalian, unidentifiable fragment 50-75 3 0.2 Mammalian, unidentifiable fragment 272 60-80 1 0.1 Mammalian, unidentifiable fragment 1 4.2 Large mammalian, unidentifiable fragment 273 0-30 14 3.3 Mammalian, unidentifiable fragment (3 thermally-altered) 274 0-30 3 1.0 Mammalian, unidentifiable fragment (1 thermally-altered) 0-30 3 1.6 Large mammalian, unidentifiable fragment 275 50-75 16 3.5 Mammalian, unidentifiable fragment 75-100 1 0.1 Mammalian, unidentifiable fragment 0-30 3 0.5 Mammalian, unidentifiable fragment 276 35-70 1 0.4 Mammalian, unidentifiable fragment 70-100 1 0.2 Mammalian, unidentifiable fragment 1 0.6 Bos/Bison sp. (cattle/bison), tooth, cheek, fragment 0-30 8 1.3 Mammalian, unidentifiable fragment 277 30-50 5 1.0 Mammalian, unidentifiable fragment 50-80 1 0.5 Large mammalian, long bone, shaft, fragment 278 0-35 1 0.4 Mammalian, unidentifiable fragment 279 25-30 1 - Ceramic, grit temper, cord marked (partially smoothed-over) 12 8.7 Large mammalian, unidentifiable fragment 30-50 54 6.9 Mammalian, unidentifiable fragment 1 3.4 Extra large mammalian, tooth, enamel, fragment, , 280 50-75 3 4.0 Large mammalian, unidentifiable fragment 30 3.6 Mammalian, unidentifiable fragment 75-100 14 6.0 Mammalian, unidentifiable fragment 0-30 4 0.7 Mammalian, unidentifiable fragment 281 40-70 1 0.1 Vertebrata, unidentifiable, fragment 30-60 1 - Decortication, rhyolite 283 60-80 1 2.7 Bison bison (bison), sesamoid, complete 288 30-60 1 0.1 Mammalian, unidentifiable fragment 291 30-70 1 0.2 Mammalian, unidentifiable fragment Total - 12,819 8873.3 -

119

9.7 Phase II XU Methods

Ten XUs were placed in site areas that offered the greatest research potential within the ROW, based on the shovel test data (Figure 8). XUs 1 to 4 were placed at the western end of the site one meter beyond the existing ROW, as construction impacts were expected to extend beyond the ROW in this area. XUs 5 to 10 were placed at the eastern end of the site within the existing ROW, as construction impacts will be limited to this area. The XUs were placed in pairs as 1 x 2’s but were dug separately. All XUs were dug below an arbitrary datum established above the ground surface.

9.8 XUs 1 and 2 on Terrace 2

XUs 1 and 2 were contiguous units placed between Shovel Tests 104 and 104W5, which yielded 81 faunal fragments, including a bison long bone fragment and two thermally-altered fragments (Figure 8). The XUs are located 16 meters south of the existing CSAH 1 centerline. Excavation was conducted in 10-cm levels below a unit datum, which was placed about 10 cm above the ground surface. The landscape was fairly level. Excavation was terminated at 100 cmbd, and a shovel test was placed in the base of XU 1 to 180 cmbd to examine the soils and determine if deeply buried artifacts are present. A summary of artifacts recovered in the units is presented in Tables 17 to 19.

Table 17. Site 21PL109 Artifacts by Count from XUs 1 and 2. Faunal Depth Lithic Lithic Faunal Thermally- Ceramic FCR Total % (cmbd) Debris Tool altered 0-25 ------25-30 11 1 - - - - 12 2 30-40 75 10 1 - 4 - 90 11 40-50 141 3 1 - 1 - 146 18 50-60 173 1 3 - 4 - 181 23 60-70 125 6 - - 2 - 133 17 70-80 91 1 - - 1 - 93 12 80-90 58 12 - - - 1 71 9 90-100 62 - 1 1 - - 64 8 100-120* 1 - - - - - 1 <1 Total 737 34 6 1 12 1 791 - % 93 4 <1 <1 2 <1 - 100 * Shovel test

120

Table 18. Site 21PL109 Artifacts by Weight (g) from XUs 1 and 2. Faunal Depth Lithic Lithic Faunal Thermally- Ceramic FCR Total % (cmbd) Debris Tool Altered 0-25 ------25-30 3.1 0.1 - - - - 3.2 1 30-40 19.7 4.7 0.1 - 5.5 - 30.0 8 40-50 32.5 0.6 0.1 - 0.6 - 33.8 9 50-60 74.4 2.8 5.5 - 2.9 - 85.6 23 60-70 43.3 2.3 - - 0.5 - 46.1 12 70-80 21.6 0.4 - - 0.4 - 22.4 6 80-90 76.0 8.0 - - - 11.0 95 25 90-100 23.7 - 0.0 33.9 - - 57.6 15 100-120* 0.3 - - - - - 0.3 <1 120-180* ------0 0 Total 294.6 18.9 5.7 33.9 9.9 11 374 - % 79 5 2 9 3 3 - 100 * Shovel test

Table 19. Site 21PL109 Fauna counts from XUs 1 and 2.

) muck Depth p Total % (cmbd) og/wolf/coyote) (d Rodentia Rodentia Frog/Toad Frog/Toad Vertebrata Vertebrata Mammalian pocket gopher) gopher) pocket Tamias striatus striatus Tamias Molluscan (shell) eastern chi eastern Microtus sp. (vole) sp. (vole) Microtus Bison bison (bison) (bison) Bison bison ( Mammalian (Small) Mammalian (Large) (Large) Mammalian Poliocitellus franklinii Mammalian (Medium) Mammalian (Medium) Geomys bursarius (plains (plains bursarius Geomys Canis sp. Mammalian ( Extra Large) Large) ( Extra Mammalian (Franklin’s ground squirrel) ground (Franklin’s

0-25 ------25-30 ------12 ------12 2 30-40 1 - 3 - - - 79 - - - 1 - - - 1 85 11 40-50 - 1 8 - 1 1 132 ------1 144 19 50-60 5 1 8 - - - 153 2 ------5 174 23 60-70 2 - 7 1 - - 108 2 - 1 1 - 7 1 1 131 17 70-80 - - 9 - - - 80 2 ------1 92 12 80-90 2 - 4 - 1 - 63 ------70 9 90-100 - 1 - - - - 29 24 1 - 3 1 3 - - 62 8 100-120 ------1 ------1 <1 1 Total 10 3 39 1 2 1 657 30 1 1 5 1 1 9 771 - 0 < < % 1 <1 5 <1 <1 <1 85 4 <1 <1 1 1 1 - 100 1 1

121

Artifact Summary and Vertical Distribution for XUs 1 and 2 A total of 791 artifacts were recovered from XUs 1 and 2, including 737 faunal fragments, 34 thermally-altered fauna, six pieces of lithic debris, 12 ceramics, one FCR, and a stone tool (Stage 2 biface) (Table 17). The six pieces of lithic debris include three bifacial shaping flakes, a decortication flake, a broken flake, and a nonbifacial flake. Lithic materials include Knife River Flint (n=3), Swan River Chert (n=3), and unidentified chert (n=1). The lithic debris indicates that lithic activities included the early, middle, and late stages of reduction.

Artifacts were recovered between 25 and 120 cmbd. The greatest artifact density by count occurs in a single zone between 40 and 70 cmbd. By weight, there are two zones of greatest artifact density, a diffuse upper zone from 30 and 70 cmbd and a lower zone from 80 to 100 cmbd. Both count and weight have their highest density from 50 to 60 cmbd, which coincides with the top of a buried A horizon. A moderate amount of rodent runs were observed in the XUs, which caused vertical displacement of artifacts and the diffuse vertical patterning that is observed in the artifact tables. The smaller number of artifacts above and below the primary artifact zones was likely displaced by natural causes, such as freeze-thaw and bioturbation from rodent runs.

The vertical patterning of individual artifact types is similar to the collective pattern, which is based on fauna material because it is the most abundant artifact type. Anomalies to the general vertical pattern include: 1) a single FCR occurs only from 80 to 90 cmbd; 2) thermally-altered fauna have a trimodal distribution, being most abundant by weight from 30 to 40 cmbd, 50 to 70 cmbd, and 80 to 90 cmbd, although the total count is low and the vertical patterning likely reflects random displacement by rodent burrowing; and 3) ceramics have a bimodal distribution, being most abundant from 30 to 40 cmbd and 50 to 60 cmbd, although the total ceramic count is very low and the vertical patterning likely reflects random displacement by rodent burrowing.

Faunal remains include bison, and it is expected that most or all of the extra-large mammal, large mammal, and mammalian remains are bison, based on the identifiable bison remains. Other identifiable cultural remains include dog/wolf/coyote and shell. The vertical distribution of fauna does not show any significant patterning by depth, aside from the primary cluster between 40 and 70 cmbd (Tables 17 to 19).

Three faunal samples from 30 to 90 cmbd were radiocarbon dated to determine the age of the components and cultural stratigraphy. The samples yielded the following dates: 1170 +/- 30 from XU 2 at 30 to 40 cmbd; 1300 +/- 30 RCYBP from XU 1 at 60 to 70 cmbd; and 1210 +/- 30 RCYBP from XU 1 at 80 to 90 cmbd. Calibrated dates for the samples overlap or are within a decade of each other. So, although the faunal samples were recovered between 30 and 90 cmbd, the dates are essentially the same, which suggests that there is vertical displacement of artifacts from a single component. The moderate amount of rodent runs observed in the units was likely the primary cause of the displacement.

The ceramics recovered from 30 to 80 cmbd have very similar attributes, and they also support the interpretation of a single component and do not suggest different wares from multiple components. The vertical distribution of ceramics suggests displacement by natural processes. Nine of the 12 ceramics have sand temper, and three sherds also have grit pieces with sand. It is likely that these are all the same temper, which consists of a sandy paste with very sparse amounts of grit. All the sherds have thin vessel walls, typically between 4.5 and 5.8 mm, and have a smooth surface treatment with no cordmarkings. Two sherds have a polished (burnished) surface. The ceramics are most similar to the Buchanan ware of the Northeastern Plains Village type (Michlovic and Swenson 1998). Given the similar radiocarbon dates and proximity of these units to XUs 3 and 4 and Shovel Tests 201 and 202 (which had polished ceramics), it is likely that these ceramics are from a Northeastern Plains Village Buchanan ware component that occurs in this portion of the site. 122

Soils and Stratigraphy for XUs 1 and 2 The soil horizons from XUs 1 and 2 are depicted in wall profiles and photographs in Figures 11 and 12. The soil profiles consist of stratified alluvium with multiple buried paleosols. Gastropods, charcoal flecking, and charcoal lenses occur through the profile. There are slight increases in charcoal flecking in the paleosols. Aside from the moderate amount of rodent burrows, the soils are fairly undisturbed below the plow zone. Carbonate nodules are common below 70 cmbd in buried Bk horizons. The soils are devoid of natural rocks.

9.9 XUs 3 and 4 on Terrace 2

XUs 3 and 4 were contiguous units placed adjacent to Shovel Test 107W5, which yielded three ceramic sherds and 19 faunal fragments, including a bison phalanx and six thermally-altered fragments (Figure 8). The XUs are located 16.5 meters south of the existing CSAH 1 centerline. Excavation was conducted in 10-cm levels below a unit datum, which was placed about five cm above the ground surface. The landscape was fairly level. Excavation was terminated at 100 cmbd, and a shovel test was placed in the base of XU 3 to 180 cmbd to examine the soils and determine if deeply buried artifacts are present. A summary of artifacts recovered in the units is presented in Tables 20 to 22.

Table 20. Site 21PL109 Artifacts by Count from XUs 3 and 4. Faunal Depth Lithic Lithic Faunal Thermally- Ceramic FCR Total % (cmbd) Debris Tool/Core Altered 0-30 ------30-40 85 9 - - 1 - 95 17 40-50 145 14 - - 3 - 162 29 50-60 126 8 - - 2 - 136 25 60-70 36 6 - - 2 - 44 8 70-80 24 5 - - - - 29 5 80-90 43 1 - - 1 - 45 8 90-100 12 1 - - - - 13 2 100-120* 3 1 - - - - 4 1 120-140* 5 - - - - - 5 1 140-160* 13 1 - - - 1 15 3 160-180* 1 1 - - - - 2 <1 Total 493 47 - - 9 1 550 - % 90 9 0 0 2 <1 - 100 * Shovel test

123

Table 21. Site 21PL109 Artifacts by Weight (g) from XUs 3 and 4. Faunal Depth Lithic Lithic Faunal Thermally- Ceramic FCR Total % (cmbd) Debris Tool/Core altered 0-30 ------30-40 16.3 1.5 - - 0.1 - 17.9 7 40-50 35.1 3.9 - - 1.5 - 40.5 16** 50-60 30.9 1.8 - - 0.8 - 33.5 14** 60-70 6.7 1.4 - - 0.2 - 8.3 3 70-80 4.9 0.5 - - - - 5.4 2 80-90 8.6 0.1 - - 2.2 - 10.9 4 90-100 3.6 2.2 - - - - 5.8 2 100-120* 17.9 0.6 - - - - 18.5 8 120-140* 1.2 - - - - - 1.2 <1 140-160* 5.0 0.1 - - - 98.5 103.6 42 160-180* 0.2 0.8 - - - - 1.0 <1 Total 130.4 12.9 - - 4.8 98.5 246.6 - % 53 5 0 0 2 40 - 100 * Shovel test; ** excluding the FCR from 140-160 cmbd, the percentages would be 27% for 40-50 cmbd and 23% for 50-60 cmbd

Table 22. Site 21PL109 Fauna counts from XUs 3 and 4.

Depth Total % (cmbd) Shrew) (Large) (Large) Rodentia Rodentia (bison/elk) Mammalian Mammalian Mammalian ( Extra Large) Large) ( Extra (channel catfish) catfish) (channel Bison/Cervus sp. sp. Bison/Cervus Molluscan (shell) Geomys bursarius bursarius Geomys Microtus sp. (vole) sp. (vole) Microtus Bison bison (bison) (bison) Bison bison Blarina brevicauda brevicauda Blarina Ictalurus punctatus punctatus Ictalurus (Norther Short-tailed Short-tailed (Norther (plains pocket gopher) 0-30 ------30-40 - - - 12 82 ------94 17 40-50 - - 10 - 146 - - 1 - 2 - 159 29 50-60 - - 1 - 129 3 1 - - - - 134 25 60-70 - - - - 42 ------42 8 70-80 - - - - 29 ------29 5 80-90 - - - 2 40 - - - 1 - 1 44 8 90-100 - 1 - - 11 - - - - - 1 13 2 100-120 1 - - - 3 ------4 1 120-140 - - - - 5 ------5 1 140-160 - - - 6 8 ------14 3 160-180 - - - - 2 ------2 <1 Total 1 1 11 20 497 3 1 1 1 2 2 540 - % <1 <1 2 4 92 1 <1 <1 <1 <1 <1 - 100

124

Artifact Summary and Vertical Distribution for XUs 3 and 4 A total of 550 artifacts were recovered from XUs 3 and 4, including 493 faunal fragments, 47 thermally-altered fauna, nine ceramics, and one FCR (Table 20). No lithic debris, stone tools, or cores were recovered.

Artifacts were recovered between 25 and 180 cmbd. The greatest artifact density by count and weight occurs in a single zone between 40 and 60 cmbd, which coincides with the bottom of the B horizon and top of a buried A horizon (54% by count and 30% by weight, with 50% by weight if the heavy FCR from 140 to 160 cmbd is excluded). A minimal to moderate amount of rodent runs were observed in the XUs, which caused vertical displacement of artifacts and the somewhat diffuse vertical patterning that is observed in the artifact tables. The smaller number of artifacts above and below the primary artifact zones was likely displaced by natural causes, such as freeze-thaw and bioturbation from rodent runs.

The vertical patterning of individual artifact types is similar to the collective pattern, which is based on fauna material because it is the most abundant artifact type. The only anomaly to the general vertical pattern is a piece of FCR that occurs from 140 to 160 cmbd.

Faunal remains include bison, and it is expected that most or all of the bison/elk, extra-large mammal, large mammal, and mammalian remains are bison, based on the identifiable bison remains. Other identifiable cultural remains include fish and shell. The vertical distribution of fauna does show any significant patterning by depth, aside from the primary cluster between 40 and 60 cmbd (Tables 20 to 22).

Three faunal samples from XU 3 were radiocarbon dated from 50 to 160 cmbd to determine the age of the components and cultural stratigraphy. The samples yielded the following dates: 1200 +/- 30 RCYBP from 50 to 60 cmbd; 1160 +/- 30 RCYBP from 100 to 120 cmbd; and 2330 +/- 30 RCYBP from 140 to 160 cmbd. Calibrated dates for the samples from 50 to 60 cmbd and 100 to 120 cmbd overlap. So, although these faunal samples were recovered between 50 and 120 cmbd, the dates are essentially the same, which suggests vertical displacement of artifacts from a single component. The rodent runs observed in the units are likely the primary cause of the displacement. A Late Archaic or Early Woodland component is represented by fauna dating to 2330 +/- 30 RCYBP from 140 to 160 cmbd in a buried Bk horizon.

The ceramics recovered from 30 to 90 cmbd in the XUs and adjacent Shovel Test 107W5 have very similar attributes. They support the interpretation of a single component and do not suggest different wares from multiple components. The vertical distribution of ceramics suggests displacement by natural processes. Ten of the twelve ceramics have grit temper, and two sherds have sand. It is likely that these are all the same temper, with grit occurring very sparsely and not being present in all small- sized sherds (eight of nine are SG3 and SG4 <1/2” in the XUs). Only three sherds retained inner and outer surfaces, and their thickness was 3.6 mm (50 to 60 cmbd), 7.9 mm (80 to 90 cmbd), and 6.8 mm (in Shovel Test 107W from 40 to 60 cmbs). Seven sherds have a smooth surface treatment with no cordmarkings, and the other sherds are exfoliated. The ceramics are similar to the Buchanan ware of the Northeastern Plains Village type and also Sandy Lake ware (Michlovic and Swenson 1998). Given the similar radiocarbon dates and proximity of these units to XUs 1 and 2 and Shovel Tests 201 and 202, it is likely that these ceramics are from a Northeastern Plains Village Buchanan ware component that occurs in this portion of the site.

Soils and Stratigraphy for XUs 3 and 4 The soil horizons from XUs 3 and 4 are depicted in wall profiles and photographs in Figures 13 and 14. The soil profiles consist of stratified alluvium with multiple buried paleosols. Gastropods, 125 charcoal flecking, and charcoal lenses occur through the profile. There are slight increases in charcoal flecking in the paleosols. Aside from the minimal to moderate amount of rodent burrows, the soils are mostly undisturbed below the plow zone. Carbonate nodules are common below 70 cmbd in buried Bk horizons. The soils are devoid of natural rocks.

9.10 XUs 5 and 6 on Terrace 4

XUs 5 and 6 were contiguous units centered on Shovel Test R63E5, which contained a discrete deposit from 72 to 80 cmbs of charcoal, oxidized soil, and thermally-altered fauna (Figure 8). The shovel test yielded 57 faunal fragments, including two bison phalanxes, grey fox remains, and 40 thermally-altered fragments. The XUs are located 12 meters north of the existing CSAH 1 centerline. Excavation was conducted in 10-cm levels below a unit datum, which was placed about five cm above the ground surface. The landscape was fairly level. Excavation was terminated at 110 cmbd, and a shovel test was placed in the base of XU 5 to 180 cmbd to examine the soils and determine if deeply buried artifacts are present. Excavation methods included removing the Ap horizon (the plow zone and the soil above the construction blading - see soil discussion and profiles below). A summary of artifacts recovered in the units is presented in Tables 23 to 25.

Table 23. Site 21PL109 Artifacts by Count from XUs 5 and 6. Faunal Depth Lithic Fired Faunal Thermally- Lithic Tool FCR Total % (cmbd) Debris Clay Altered 0-22 ------22-30 25 8 3 - - - 36 1 30-40 18 12 1 - - - 31 1 40-50 58 26 1 - - - 85 3 50-60 32 27 1 - - - 60 2 60-70 42 30 - - - 1 73 3 70-80 630 403 20 1 2 - 1056 39 80-90 385 475 9 2 3 - 874 32 90-100 239 164 11 2 - - 416 15 100-110 34 37 - - - - 71 3 110-135* 2 - - - - - 2 <1 135-160* 7 3 - - - - 10 <1 160-185* 3 - - - - - 3 <1 Total 1475 1185 46 5 5 1 2717 - % 54 44 2 <1 <1 0 - 100 * Shovel test

126

Table 24. Site 21PL109 Artifacts by Weight (g) from XUs 5 and 6. Faunal Depth Lithic Fired Faunal Thermally- Lithic Tool FCR Total % (cmbd) Debris Clay Altered 0-22 ------22-30 10.9 4.9 2.0 - - - 17.8 1 30-40 7.5 4.4 0.2 - - - 12.1 1 40-50 25.5 5.7 0.1 - - - 31.3 1 50-60 15.0 9.9 0.3 - - - 25.2 1 60-70 69.2 21.7 - - - 1.7 92.6 4 70-80 553.9 196.3 1.7 1.4 33.2 - 786.5 37 80-90 431.9 208.5 3.2 3.5 166.1 - 813.2 38 90-100 158.1 78.6 7.8 2.8 - - 247.3 12 100-110 36.6 16.0 - - - - 52.6 2 110-135* 14.8 - - - - - 14.8 1 135-160* 25.6 1.0 - - - - 26.6 1 160-185* 2.1 - - - - - 2.1 0 Total 1351.1 547 15.3 7.7 199.3 1.7 2122.1 - % 64 26 1 0 9 0 - 100 * Shovel test

Table 25. Site 21PL109 Fauna counts from XUs 5 and 6.

) her p o g Depth Total % Bird

(cmbd) ocket p Rodentia Rodentia (Medium) (Medium) (bison/elk) Vertebrata Vertebrata Mammalian Mammalian Mammalian Mammalian ( Extra Large) Large) ( Extra (Garter snake) snake) (Garter Thamnophis sp. (Medium/Large) (Medium/Large) Bison/Cervus sp. sp. Bison/Cervus lains Geomys bursarius bursarius Geomys Bison bison (bison) (bison) Bison bison Mammalian (Large) (Large) Mammalian (p 0-22 ------22-30 - - - 5 - - 28 - - - - - 33 1 30-40 - - - 1 - - 29 - - - - - 30 1 40-50 - - - 15 - - 66 1 - 1 1 - 84 3 50-60 - - - 9 - - 50 - - - - - 59 2 60-70 2 - - 5 - - 64 1 - - - - 72 3 70-80 11 2 25 89 - - 826 - - 5 1 74 1033 39 80-90 36 - 8 83 1 - 630 - - 3 - 99 860 32 90-100 4 - 9 55 - 1 329 1 1 3 - - 403 15 100-110 1 - 2 4 - - 63 - 1 - - - 71 3 110-135 1 ------1 - - - - 2 <1 135-160 - - - 1 - - 9 - - - - - 10 <1 160-185 - - - 3 ------3 <1 Total 55 2 44 270 1 1 2094 4 2 12 2 173 2660 - % 2 <1 2 10 0 0 79 <1 <1 <1 <1 7 - 100

127

Artifact Summary and Vertical Distribution for XUs 5 and 6 A total of 2717 artifacts were recovered from XUs 5 and 6, including 1475 faunal fragments, 1185 thermally-altered fauna, 46 pieces of lithic debris, five lithic tools (four utilized flakes and a tip fragment from a projectile point), five pieces of fired clay, and one FCR (Table 23).

Artifacts were recovered between 22 and 185 cmbd. The greatest artifact density by count and weight occurs in a single zone between 70 and 90 cmbd (75% by weight and 71% by count), which coincides with a buried B horizon and an anthrosol that occurs in the middle portion of the buried B horizon. Fauna from 80 to 90 cmbd dated to 6830 +/- 30 RCYBP, providing a date for the primary component in these units. It was noted during excavation from 70 to 90 cmbd that most of the artifacts were recovered from the anthrosol, indicating that it was living surface. Its darker color likely caused by charcoal from fires and decay of organics, such as animal remains. The anthrosol is a thin, organically enriched dark-colored lens with lighter gray mottles that is 3-7 cm thick and contains charcoal, oxidized/burned (orangish-colored) soil, lithics, fauna, and thermally-altered fauna. A fire hearth feature (Feature 1) was identified in the anthrosol, confirming that this was a living surface. The component from 70 to 90 cmbd was also identified in XUs 7 to 10, based on similar depths of artifact concentrations, soil context, anthrosols, and radiocarbon dates.

In addition to the primary archaeological deposit from 70 to 90 cmbd, there are other sparse occupations above and below it throughout the soil profile that are represented by much smaller amounts of artifacts, primarily fauna. A small spike in artifacts from 40 to 50 cmbd represents one of these components. These components are inferred from the multiple distinct radiocarbon dates (see below) and vertical patterning of artifacts. A minimal to moderate amount of rodent runs were observed in the XUs, which likely caused some very limited vertical displacement of artifacts.

Five faunal samples were radiocarbon dated from 80 to 185 cmbd to determine the age of the components and cultural stratigraphy. The samples yielded the following dates: 6210 +/- 30 RCYBP from XU 6 at 40 to 50 cmbd; 6830 +/- 30 RCYBP from XU 5 at 80 to 90 cmbd; 7260 +/- 30 RCYBP from XU 5 at 110 to 135 cmbd; 7320 +/- 30 RCYBP from XU 5 at 135 to 160 cmbd; and 7980 +/- 30 RCYBP from XU 5 at 160 to 185 cmbd. The calibrated dates from the samples at 110 to 135 cmbd and 135 to 160 cmbd overlap, indicating that they are from the same occupation. The date of 6830 +/- 30 RCYBP from XU 5 at 80 to 90 cmbd appears to date the concentration of artifacts, anthrosol, and feature at that depth. The other samples have unique ages that represent multiple distinct Early to Middle Archaic components from 40 to 185 cmbd. The age of the sparse component above 40 cmbd is unknown.

The vertical patterning of individual artifact types is similar to the collective pattern, which is based on fauna material because it is the most abundant artifact type. The only anomalies to the general vertical pattern are: 1) a slight spike in lithic debris from 22 to 30 cmbd; 2) fired clay occurs only from 70 to 90 cmbd, which is likely associated with the fire hearth; 3) lithic tools occur only from 70 to 100 cmbd; 4) there is a slight increase in lithic debris from 90 to 100 cmbd; and 5) a single piece of FCR occurs from 60 to 70 cmbd.

Numerous bison remains were identified, and it is expected that most or all of the other mammal remains are bison, based on the identifiable bison remains. Other identifiable cultural remains include bird bone. The vertical distribution of fauna does show any significant patterning by depth, aside from the primary cluster between 70 and 90 cmbd (Tables 23 to 25).

Approximately 90% of the lithics were recovered from 70 to 100 cmbd (Tables 23 and 24). Swan River Chert is the most abundant material, followed by much smaller amounts of Knife River Flint, quartzite, and Tongue River Silica (Table 26). Diagnostic lithic debris consists primarily of bifacial shaping flakes, which are from late-state bifacial tool production and tool maintenance (Table 27). 128

Biface manufacture is represented by a small amount of bifacial thinning flakes. Small amounts of nonbifacial flakes are present, which are from nonbifacial core reduction and flake blank production. A single bipolar flake is also present. Stone tools include four utilized flakes and the tip of a projectile point, which are indicative of animal processing and hunting.

Table 26. Site 21PL109 Raw Materials by Depth in XUs 5 and 6. Depth Swan River Knife River Tongue River Quartzite cmbd Chert Flint Silica 0-22 - - - - 22-30 1 1 1 - 30-40 1 - - - 40-50 1 - - - 50-60 - - 1 - 60-70 - - - - 70-80 19 1 - 1 80-90 11 - - - 90-100 12 1 - - 100-110 - - - -

Table 27. Site 21PL109 Diagnostic Lithic Debris and Tools by Depth in XUs 5 and 6. Bifacial Bifacial Utilized/ Depth Bipolar Nonbifacial Projectile Shaping Thinning Retouch cmbd Flakes Flakes Point Flakes Flakes Flakes 0-22 ------22-30 1 1 - 1 - - 30-40 ------40-50 ------50-60 ------60-70 ------70-80 - 7 - - 1 - 80-90 - 2 1 - 2 - 90-100 - - 2 2 1 1 100-110 ------

Soil samples from the anthrosol in XU 5 from 77 to 80 cmbd (2.4 liters), 81 to 84 cmbd (1.9 liters) and 86 to 90 cmbd (3.8 liters) and XU from 6 74 to 78 cmbd (2.4 liters) were bagged and floated in the lab. Analysis of the botanicals recovered from the light fraction was conducted by Connie Arzigian (paleoethnobotanist) and staff at MVAC. The only botanical material recovered was a small amount of wood charcoal. The artifacts from the heavy fraction are included in artifact Tables 23 and 24, within the 10-cm level that these samples were recovered from.

Soils and Stratigraphy for XUs 5 and 6 The soil horizons from XUs 5 and 6 are depicted in wall profiles and photographs in Figures 15 to 17. Subsurface disturbance from previous road construction was present in the upper portion of these XUs, as expected because they are located in the shallow ROW ditch. The previous construction disturbance is clear in the soil profile and shows truncation of the upper soil horizons to a maximum depth of 35 cmbd on the south side of the units. The soil profiles consist of stratified alluvium with multiple buried paleosols. Aside from the minimal to moderate amount of rodent burrows, the soils are mostly undisturbed below the plow zone and soil truncation from previous road construction.

129

Carbonate nodules are common below 60 cmbd in buried Bk and BCk horizons. The soils are devoid of natural rocks.

Feature 1 in XUs 5 and 6 Feature 1 was identified at 90 cmbd at the base of the anthrosol. The top of the feature seems to actually have been in the anthrosol, and the feature was not identified until the anthrosol was excavated and a circular shape of oxidized soil (orangish colored) was apparent in clear contrast to the adjacent soil. Diffuse deposits of charcoal and ash with oxidized soil were present between 82 and 90 cmbd in the anthrosol directly above the circular oxidized soil discoloration mapped as Feature 1 (Figure 18). The charcoal from the anthrosol was not dated because it was uncertain if it was cultural or natural, and it was found before the feature was defined. Feature 1 is a circular soil discoloration (dark yellowish brown 10YR 3/6) from excessive heat that caused the soil to oxidize and change color. Planviews and profiles of Feature 1 were recorded in illustrations and photos during excavation (Figures 19 and 20). It is likely that Feature 1 consists only of discolored soil oxidized by the heat from a fire above it and is not a pit containing feature fill. The feature is centered in the northeastern quadrant of XU 5 and a small portion extends into the north wall.

All feature soil (14.2 liters) was troweled, and the south half was bagged for flotation, while the north half was screened through ¼” mesh in the field. Analysis of the botanicals recovered from the light and heavy fractions was conducted by Connie Arzigian (paleoethnobotanist) and staff at MVAC. There was no charcoal or other charred botanicals in the feature, only one chenopodium seed that is probably a modern intrusion.

The feature has a diameter of 70 cm and a shallow basin shape that is seven cm deep. The feature is interpreted as the heated-soil remnants of a fire hearth, based its oxidized soil color in conjunction with the ash deposit, charcoal, and fired clay in the anthrosol immediately above it.

A total of 35 artifacts were recovered from the heavy fraction during flotation of Feature 1 soil, including 32 fauna, two fired clay pieces, and one piece of lithic debris. Fauna found in the feature was noted in rodent runs coming from the overlying anthrosol. Fauna included a bird long bone, two Thamnophis sp. (Garter snake) vertebra, five mammalian unidentifiable (four are thermally-altered), 24 Vertebrata unidentifiable (20 are thermally-altered). All the fauna are less than one inch in size. Fauna from 80 to 90 cmbd in XU 5 dated to 6830 +/- 30 RCYBP, providing a likely age for the feature.

9.11 XUs 7 and 8 on Terrace 4

XUs 7 and 8 were contiguous units placed on the north side of Shovel Test R68E5, which contained a Middle Archaic Oxbow projectile point, a piece of lithic debris, and 42 faunal fragments, including six extra-large mammal bone (probably bison) and nine thermally-altered fragments (Figure 8). The XUs are located 12.3 meters north of the existing CSAH 1 centerline. Excavation was conducted in 10-cm levels below a unit datum, which was placed about five cm above the ground surface. The landscape was fairly level. Excavation was terminated at 110 cmbd, and shovel tests were placed in the base of the units to 135 cmbd to examine the soils and determine if deeply buried artifacts are present. Excavation methods included removing the Ap horizon, which included the plow zone and the soil above the construction blading (see soil discussion and profiles below). A summary of artifacts recovered in the units is presented in Tables 28 to 30.

130

Table 28. Site 21PL109 Artifacts by Count from XUs 7 and 8. Faunal Depth Lithic Lithic Faunal Thermally- FCR Total % (cmbd) Debris Tool Altered 0-20 ------20-30 7 - - - - 7 4 30-40 6 2 2 - 1 11 6 40-50 9 36 1 - - 46 23 50-60 8 3 - - 1 12 6 60-70 8 3 - - - 11 6 70-80 77 15 - 1 - 93 47 80-90 5 5 2 - - 12 6 90-100 3 - - - - 3 2 100-110 1 - - - - 1 1 110-135* - - - - - 0 0 Total 124 64 5 1 2 196 - % 63 33 3 <1 1 - 100 *Shovel Test

Table 29. Site 21PL109 Artifacts by Weight (g) from XUs 7 and 8. Faunal Depth Lithic Lithic Faunal Thermally- FCR Total % (cmbd) Debris Tool Altered 0-20 ------20-30 3.3 - - - - 3.3 1 30-40 1.6 0.3 12.3 - 1.6 15.8 7 40-50 13.0 7.0 1.8 - - 21.8 10 50-60 2.6 0.8 - - 6.4 9.8 4 60-70 1.0 1.5 - - - 2.5 1 70-80 154.1 4.3 - 2.4 - 160.8 73 80-90 2.5 2.6 1.6 - - 6.7 3 90-100 0.2 - - - - 0.2 0 100-110 0.4 - - - - 0.4 0 110-135* - - - - - 0 0 Total 178.7 16.5 15.7 2.4 8.0 221.3 - % 81 7 7 1 4 - 100 *Shovel Test

131

Table 30. Site 21PL109 Fauna counts from XUs 7 and 8.

Depth Total % (cmbd) (wolf) (bison) (bison) (Large) (Large) Bison bison Bison bison Canis lupus Canis lupus Mammalian Mammalian Mammalian Mammalian ( Extra Large) Large) ( Extra (Medium/Large) (Medium/Large)

0-20 ------20-30 - - 3 - - 4 7 4 30-40 - - 1 - - 7 8 4 40-50 - - 1 - 8 36 45 24 50-60 - - 2 - - 9 11 6 60-70 - - 1 1 - 9 11 6 70-80 3 7 12 - - 70 92 49 80-90 - - 1 - - 9 10 5 90-100 - - - - - 3 3 2 100-110 - - - - - 1 1 1 Total 3 7 21 1 8 148 188 - % 2 4 11 1 4 79 - 100

Artifact Summary and Vertical Distribution for XUs 7 and 8 A total of 196 artifacts were recovered from XUs 7 and 8, including 124 faunal fragments, 64 thermally-altered fauna, five pieces of lithic debris, a lithic tool (end scraper), and two pieces of FCR (Table 28). A small fleck of ochre (SG4) was also recovered from 70 to 80 cmbd.

Artifacts were recovered between 20 and 110 cmbd. The greatest artifact density by count and weight occurs in a single zone between 70 and 80 cmbd (73% by weight and 47% by count), which coincides with the upper portion of a buried B horizon. A thin charcoal lens, which could be an anthrosol was noted during excavation between 75 and 80 cmbd. Fauna from 70 to 80 cmbd dated to 6880 +/- 30 RCYBP, providing a date for the primary component in these units. The artifact concentration from 70 to 80 cmbd is the same component identified in XU 5 and 6 and 9 and 10, based on similar depths of artifact concentrations, soil context, anthrosols, and radiocarbon dates.

In addition to the primary archaeological deposit from 70 to 80 cmbd that dated to 6880 +/- 30 RCYBP, there is at least one other separate occupation represented by a small spike in artifacts from 40 to 50 cmbd that dated to 6240 +/- 30 RCYBP. A minimal to moderate amount of rodent runs were observed in the XUs, which likely caused some limited vertical displacement of artifacts.

A Middle Archaic Oxbow projectile point was recovered adjacent to XUs 7 and 8 in Shovel Test R68E5 from 75 to 100 cmbs. The Oxbow point could be associated with the component from 70 to 80 cmbd in XUs 7 and 8 that is defined by an artifact concentration and a faunal date of 6880 +/- 30 RCYBP. However, the date is more than 1000 years older than the bulk of established Oxbow dates (Eighmy and LaBelle 1996; Green 1998:276-280), but four Oxbow dates are reported between 6750 to 5600 RCYBP by Green (1998:276-280). While the depth provenience of artifacts from shovel test is not certain because of the lack of well-controlled excavation, it is quite possible that the dated faunal material is associated with the Oxbow point for two reasons: first, most of the fauna from the shovel test was recovered from 75 to 100 cmbs, which likely corresponds with the depth of the primary deposit (mostly fauna) in XU 7 from 70 and 80 cmbd (65 to 75 cmbs) that was dated at 6880 +/- 30 RCYBP; and second, if the Oxbow point was displaced from a shallower depth in the shovel 132 test, its date likely still exceeds the bulk of established Oxbow dates, as a date of 6240 +/- 30 RCYBP was obtained from XU 7 from 40 to 50 cmbd (35 to 45 cmbs).

The vertical patterning of individual artifact types is similar to the collective pattern, which is based on fauna material because it is the most abundant artifact type. The only anomalies to the general vertical pattern are: 1) a notable slight spike in thermally-altered fauna occurs from 40 to 50 cmbd; 2) the sparse lithic debris does not coincide with concentrations of fauna, but occurs in an adjacent level; 3) the single lithic tool occurs in the 70 to 80 cmbd level; 4) there is a slight increase in lithic debris from 90 to 100 cmbd; and 5) two pieces of FCR occur from 30 to 40 and 50 to 60 cmbd.

Faunal remains include bison, and it is expected that most or all of the extra-large mammal, large mammal, and mammal remains are bison, based on the identifiable bison remains. Other identifiable cultural remains include wolf. The vertical distribution of fauna does not show any significant patterning by depth, aside from the primary cluster between 70 and 80 cmbd, except that the wolf bones all cluster from 40 to 50 cmbd (Tables 28 to 30).

Only a very sparse amount of lithics were recovered, including Swan River Chert, unidentified chert, rhyolite, and petrified wood (Table 31). Diagnostic lithic debris consists nonbifacial and decortication flakes, which are from nonbifacial core reduction and flake blank production (Table 32). Stone tools include a scraper that is indicative of animal processing.

Table 31. Site 21PL109 Raw Materials by Depth in XUs 7 and 8. Depth Swan River Unid. Petrified Rhyolite cmbd Chert Chert Wood 0-20 - - - - 20-30 - - - - 30-40 - 1 1 - 40-50 1 - - - 50-60 - - - - 60-70 - - - - 70-80 - - - 1 80-90 2 - - - 90-100 - - - - 100-110 - - - -

Table 32. Site 21PL109 Diagnostic Lithic Debris and Tools by Depth in XUs 7 and 8. Depth Nonbifacial Decortication Scraper cmbd Flakes Flakes 0-20 - - - 20-30 - - - 30-40 1 1 - 40-50 - - - 50-60 - - - 60-70 - - - 70-80 - - 1 80-90 1 - - 90-100 - - - 100-110 - - -

133

Soils and Stratigraphy for XUs 7 and 8 The soil horizons from XUs 7 and 8 are depicted in wall profiles and photographs in Figures 21 and 22. Subsurface disturbance from previous road construction was present in the upper portion of these XUs because they are located in the shallow ROW ditch. The previous construction disturbance is clear in the soil profile and shows truncation of the upper soil horizons to a maximum depth of 40 cmbd on the south side of the units. The soil profiles consist of stratified alluvium with multiple buried paleosols. Aside from the minimal to moderate amount of rodent burrows, the soils are mostly undisturbed below the plow zone and soil truncation from previous road construction. Carbonate nodules are common below 63 cmbd in buried Bk and BCk horizons. The soils are mostly devoid of natural rocks.

9.12 XUs 9 and 10 on Terrace 4

XUs 9 and 10 were contiguous units placed on the north side of Shovel Test R59W5, which contained 12 pieces of lithic debris, and 29 faunal fragments, including a bison tooth and five thermally-altered fragments (Figure 8). The XUs are located 12.35 meters north of the existing CSAH 1 centerline. Excavation was conducted in 10-cm levels below a unit datum, which was placed about five cm above the ground surface. The landscape was fairly level. Excavation was terminated at 120 cmbd, and a shovel test was placed in the base of XU 9 to 195 cmbd to examine the soils and determine if deeply buried artifacts are present. Excavation methods included removing the Ap horizon, which included the plow zone and the soil above the construction blading (see soil discussion and profiles below). A summary of artifacts recovered in the units is presented in Tables 33 to 35.

Table 33. Site 21PL109 Artifacts by Count from XUs 9 and 10. Faunal Depth Lithic Lithic Faunal Thermally- FCR Total % (cmbd) Debris Tool/Core Altered 0-30 ------30-40 9 1 - - - 10 <1 40-50 2 - - - - 2 <1 50-60 9 - - - - 9 <1 60-70 33 12 - - - 45 1 70-80 119 32 - - - 151 4 80-90 27 2 - - - 29 1 90-100 16 4 1 - - 21 1 100-110 1680 815 469 8* - 2972 86 110-120 163 34 23 - - 220 6 120-145** 1 - - - - 1 0 145-170** 2 1 - - - 3 <1 170-195** - - - - - 0 0 Total 2061 901 493 8 0 3463 - % 60 26 14 0 0 - 100 * seven tools and one core; **shovel test

134

Table 34. Site 21PL109 Artifacts by Weight (g) from XUs 9 and 10. Faunal Depth Lithic Lithic Faunal Thermally- FCR Total % (cmbd) Debris Tool/Core Altered 0-30 ------30-40 10.9 0.1 - - - 11.0 <1 40-50 1.2 - - - - 1.2 0 50-60 2.4 - - - - 2.4 <1 60-70 20.3 2.5 - - - 22.8 1 70-80 172.0 9.7 - - - 181.7 7 80-90 4.4 0.4 - - - 4.8 <1 90-100 45.2 8.1 0.1 - - 53.4 2 100-110 1571.6 213.3 330.0 55.6 - 2170.5 83 110-120 139.9 9.3 29.6 - - 178.8 7 120-145** 0.2 - - - - 0.2 0 145-170** 0.3 0.2 - - - 0.5 0 170-195** - - - - - 0 0 Total 1968.4 243.6 359.7 55.6 0 2627.3 - % 75 9 14 2 0 - 100 * seven tools and one core; **shovel test

Table 35. Site 21PL109 Fauna counts from XUs 9 and 10.

Depth Total % (cmbd) (Large) (Large) Rodentia Rodentia (bison/elk) Vertebrata Vertebrata Mammalian Mammalian Mammalian Mammalian Bos/Bison sp. ( Extra Large) Large) ( Extra (Medium/Large) (Medium/Large) Bison/Cervus sp. sp. Bison/Cervus Bison bison (bison) (bison) Bison bison

0-30 ------30-40 - 1 - - 2 - 7 - - 10 <1 40-50 - - 1 - - - 1 - - 2 <1 50-60 ------9 - - 9 <1 60-70 - - - - 13 - 32 - - 45 2 70-80 4 1 - 20 6 1 118 - 1 151 5 80-90 - - - - 2 - 27 - - 29 1 90-100 2 - - - 3 - 13 2 - 20 1 100-110 36 2 - 50 160 2 1511 - 734 2495 84 110-120 1 - - 5 10 - 181 - - 197 7 120-145 ------1 - - 1 <1 145-170 ------3 - - 3 <1 Total 43 4 1 75 196 3 1903 2 735 2962 - % 1 <1 0 3 7 <1 64 <1 25 - 100

135

Artifact Summary and Vertical Distribution for XUs 9 and 10 A total of 3463 artifacts were recovered from XUs 9 and 10, including 2061 faunal fragments, 901 thermally-altered fauna, 493 pieces of lithic debris, seven lithic tools (three late stage bifaces, projectile point, an indeterminate bifacial tool fragment, a scraper, and a retouched flake), and one core (Table 33). The projectile point is a small, side-notched type that is similar to other Early Archaic points. Two small pieces of red ochre were recovered from 100 to 110 cmbd and one small piece was recovered from 110 to 120 cmbd. Also, several small flecks of red ochre were observed in the anthrosols from 100 to 120 cmbd, with some ochre flecks observed on faunal material. It appears that the presence of ochre here was intentional and may represent a customary practice related to the killing and processing of bison. Similar ochre flecking in soil and on probable bison bones was noted at Late Plains Archaic site 21PL57 on a point bar of the Red River near Grand Forks (Florin 2001). The presence of ocher on the bones suggests that spirituality played an important role in bison hunting, and ochre was likely part of hunting and animal processing ritual.

Artifacts were recovered between 30 and 170 cmbd. The greatest artifact density by count and weight occurs in a single zone between 100 and 110 cmbd (83% by weight and 86% by count), which coincides with a buried A horizon and anthrosol. Fauna from 100 to 110 cmbd dated to 7110 +/- 30 RCYBP, providing a date for the primary component in these units and for the small, side-notched point. In addition to the primary archaeological deposit from 100 to 110 cmbd, there are separate occupations above and below it throughout the soil profile that are represented by smaller amounts of artifacts. One of these occupations is defined by an increase in artifacts from 70 to 80 cmbd that is associated with an anthrosol and fauna that dated to 6850 +/- 30 RCYBP. The component from 70 to 80 cmbd is the same component identified in XUs 5 to 8, based on similar depths of artifact concentrations, soil context, anthrosols, and radiocarbon dates. A minimal to moderate amount of rodent runs were observed in the XUs, which likely caused some limited vertical displacement of artifacts.

The vertical patterning of individual artifact types is nearly identical to the collective pattern, which is based on fauna material because it is the most abundant artifact type. There are no notable anomalies. Faunal remains include bison, and it is expected that most or all of the other remains are bison (excluding the Rodentia), based on the identifiable bison remains. The vertical distribution of fauna does not show any significant patterning by depth, aside from the primary cluster at 100 to 110 cmbd and minor cluster from 70 to 80 cmbd (Tables 33 to 35).

Approximately 95% of the lithics were recovered from 100 to 110 cmbd (Tables 33 and 34). Swan River Chert is the most abundant material, followed by much smaller amounts of Tongue River Silica, rhyolite, unidentified chert, Knife River Flint, unidentified material, Red River Chert, and quartzite (Table 36). Diagnostic lithic debris consists primarily of bifacial shaping flakes, which are from late-state bifacial tool production and tool maintenance (Table 37). Biface manufacture is represented by four bifaces and a moderate amount of bifacial thinning flakes. Nonbifacial flakes are also moderately represented, which are from nonbifacial core reduction and flake blank production. The single nonbifacial core and few decortication flakes indicate early stage reduction. A small amount of bipolar flakes are also present. Stone tools include scrapers, utilized flakes, and retouch flakes, indicative of hunting and processing.

136

Table 36. Site 21PL109 Raw Materials by Depth in XUs 9 and 10. Swan Knife Depth Tongue River Unid. Unid. Red River River Rhyolite River Quartzite cmbd Silica Chert Material Chert Chert Flint 0-30 ------30-40 ------40-50 ------50-60 ------60-70 ------70-80 ------80-90 ------90-100 1 ------100-110 365 73 16 8 11 2 1 1 110-120 14 8 - 1 - - - -

Table 37. Site 21PL109 Diagnostic Lithic Debris and Tools by Depth in XUs 9 and 10. Bifacial Bifacial Non- Utilized/ Depth Bipolar Decortication Projectile Shaping Thinning bifacial Bifaces Retouch Scrapers Cores cmbd Flakes Flakes Points Flakes Flakes Flakes Flakes 0-30 ------30-40 ------40-50 ------50-60 ------60-70 ------70-80 ------80-90 ------90-100 ------100-110 3 88 19 25 4 4 1 1 1 1 110-120 - 1 1 1 ------

Soil samples from the anthrosol in XU 9 from 107 to 109 cmbd (4.7 liters) and XU 10 103 to 110 cmbd (9 liters) were bagged and floated in the lab. Analysis of the botanicals recovered from the light fraction was conducted by Connie Arzigian (paleoethnobotanist) and staff at MVAC. The only botanical material recovered was a small amount of wood charcoal, and in XU 10 there were five fragments from one seed of what might be Fabaceae (bean family). The seed has a thin shell and a prominent hilum attachment area, characteristic of the bean family, possibly some type of sumac. The artifacts from the heavy fraction are included in artifact Tables 33 and 34 with the 10 cm level from which these samples were recovered.

Soils and Stratigraphy for XUs 9 and 10 The soil horizons from XUs 9 and 10 are depicted in wall profiles and photographs in Figures 23 to 25. Subsurface disturbance from previous road construction was present in the upper portion of these XUs because they are located in the shallow ROW ditch. The previous construction disturbance is clear in the soil profile and shows truncation of the upper soil horizons to a maximum depth of 65 cmbd on the south side of the units. The soil profiles consist of stratified alluvium with multiple buried paleosols. Aside from the minimal amount of rodent burrows in most levels, with moderate amount in some levels, the soils are mostly undisturbed below the plow zone and soil truncation from previous road construction. Carbonate nodules are common below 63 cmbd in buried Bk and BCk horizons. The soils are mostly devoid of natural rocks.

137

9.13 Phase I and II Artifact Summary

A total of 20,888 artifacts, weighing 22,409 grams, were recovered from the site during the Phase I survey and Phase II evaluation (Table 38). Fauna was by far the most abundant artifact type by count (95%) and weight (66%) with much smaller amounts of lithics, FCR and ceramics.

Table 38. Site 21PL109 Summary of Artifacts by Count and (Weight). Total by % by Artifact Type Count (Weight g) Count (Weight g) Fauna 19,901 (14748.8) 95 (66) Lithic 799 (952.3) 4 (4) FCR 60 (6397.0) <1 (29) Ceramic 128 (310.9) 1 (1) Total 20,888 (22,409)* - % - 100 *does not include possible manuport

9.14 Faunal Analysis by Steven Kuehn

Faunal Summary The faunal assemblage from site 21PL109 contains 19,901 pieces of bone and shell that weigh 14,748.8 grams (g) (Table 39). A total of 5812 display evidence of thermal alteration, in the form of burned and calcined bone. Mammal remains predominate with a limited number of bird, fish, mollusk, and reptile bones also present. Overall, preservation is good although bone fragmentation, thermal alteration, and other taphonomic factors hindered specific identification of many specimens.

9.14.1 Fauna Results

Vertebrata A total of 1006 bone fragments weighing 73.0 g cannot be identified to element or taxon and are listed as taxon indeterminate (Vertebrata). None of the Vertebrata remains exhibit butchery marks or modification evidence, but 445 pieces are thermally-altered (burned or calcined). The majority of Vertebrata bones are from excavation units, with a limited number obtained from shovel test and feature context.

Mammals Mammal remains account for 18,816 bones and bone fragments, with a total weight of 14,631.8 g. Most of the specimens are small, non-diagnostic fragments and are listed as indeterminate mammal. Indeterminate mammal remains account for 16,326 pieces of mammal bone, with a total weight of 4,439.9 g, and 5012 of these bones are thermally-altered (burned or calcined). Butchery marks are present on an indeterminate mammal long bone shaft from Shovel Test 150W5, two indeterminate fragments from Shovel Test 142, and one indeterminate fragment from XU 9 at 100-110 cmbd. A long bone shaft piece from Shovel Test 152W5 has a polished exterior and represents a tool fragment. Two indeterminate mammal bones show evidence of rodent gnawing. Indeterminate mammal remains were recovered from shovel test, excavation unit/feature, and surface contexts.

138

Table 39. Site 21PL109 Faunal Assemblage. Thermally- Taxon NISP Wgt(g) MNI Altered MAMMALS Bison (Bison sp.) 311 5461.8 5 7 Elk (Cervus elaphus) 2 66.7 1 0 White-tailed deer (Odocoileus virginianus) 8 78.1 2 1 Bison/elk, indeterminate (Bison/Cervus sp.) 35 112.0 -- 1 Cattle/bison, indeterminate (Bos/Bison sp.) 46 646.4 -- 1 Elk/deer, indeterminate (Cervidae) 42 21.3 -- 0 Horse (Equus caballus) 2 51.8 1 0 Gray wolf (Canis lupus) 8 11.6 1 0 Dog/wolf/coyote, indeterminate (Canis sp.) 4 15.1 2 0 Gray fox (Urocyon cinereoargenteus) 3 0.9 1 1 Large carnivore, indeterminate (large Carnivora) 1 0.4 -- 0 Beaver (Castor canadensis) 21 34.4 2 0 Muskrat (Ondatra zibethicus) 2 0.7 1 0 Hare/rabbit, indeterminate (Lepus/Sylvilagus sp.) 1 0.4 1 0 Plains pocket gopher (Geomys bursarius) 106 22.3 6 0 Franklin's ground squirrel (Poliocitellus franklinii) 2 0.6 1 0 Thirteen-lined ground squirrel (Ictidomys tridecemlineatus) 1 0.1 1 0 Eastern chipmunk (Tamias striatus) 3 0.3 2 0 Vole, indeterminate (Microtus sp.) 3 0.4 2 0 Northern shrew (Blarina brevicauda) 1 0.1 1 0 Rodent, indeterminate (Rodentia) 17 1.8 -- 1 Extra large mammal 380 1796.9 -- 23 Large-sized mammal 1332 1813.8 -- 297 Medium-large mammal 152 51.3 -- 14 Medium-sized mammal 5 2.3 -- 1 Small-sized mammal 2 0.4 -- 0 Mammal, indeterminate 16,326 4439.9 -- 5012 BIRDS Large owl, indeterminate (cf. Bubo/Strix sp.) 1 2.9 1 0 Bird, indeterminate 17 3.5 -- 6 REPTILES and AMPHIBIANS Turtle, indeterminate 9 1.9 -- 2 Garter snake, indeterminate (Thamnophis sp.) 4 0.3 1 0 Non-venomous snake, indeterminate (Colubridae) 2 0.3 -- 0 Toad, indeterminate (Anaxyrus sp.) 1 0.1 1 0 Frog/toad, indeterminate 1 0.1 -- 0 FISH Channel catfish (Ictalurus punctatus) 7 3.1 2 0 Fish, indeterminate 7 1.4 -- 0 MOLLUSKS AND GASTROPODS Mussel, indeterminate 28 30.1 -- 0 Pea clam, indeterminate (Sphaeriidae) 1 0.1 1 0 Pond snail, indeterminate (Lymnaeidae) 1 0.2 1 0 Indeterminate Taxon indeterminate (Vertebrata) 1006 73.0 -- 445 Total 19,901 14,748.1 37 5812

139

(Table 40). Cranial pieces are the most common with 96 specimens, although 86 of these are individual teeth or tooth fragments. Ten mandible fragments were recognized, but no maxillary portions or crania were observed. Distal limb bones are second in abundance with 72 remains, of which the majority are phalanges, tarsals, and carpals. Fifty-eight elements, primarily rib shaft pieces, are trunk portions. A limited number of upper fore and hind limb bones were recovered, but unrecognized limb elements are likely represented by the 26 bison long bone shaft pieces present in the assemblage.

Bison remains were recovered from shovel tests and excavation units, with the majority found in the former. None of the bison remains show evidence of tool manufacture or similar cultural modification. Cut marks were observed on seven specimens, including an astragalus, magnum, scaphoid, tibia, femur, and two long bone shaft pieces. The location and position of the cut marks are consistent with carcass dismemberment, filleting, and general processing of bison (e.g., Binford 1981). No conclusive impact or spiral fractures, indicative of breakage for marrow extraction, were recognized on any of the bison bones recovered. However, some clearly had spiral-like fractures that could have been caused by human processing activities, such as marrow extraction. Four bison bones display rodent gnawing, one has carnivore gnaw marks, and one shows indeterminate possible gnaw marks. Bison inhabit prairie settings and occasionally occupy open forest and forest edge areas.

Table 40. Bison Remains by Body Portion and Element. Body Portion Element NISP teeth 86 Cranium mandible 10 rib 41 Trunk vertebra 14 pelvis 3 humerus 3 radius 3 Upper forelimb scapula 2 ulna 2 tibia 9 Upper hindlimb femur 5 phalanx 28 carpal 14 tarsal 9 Distal limb sesamoid 8 metacarpal 6 metapodial 4 metatarsal 3 long bone shaft 35 Indeterminate indeterminate 26

One left ulna fragment and a rib shaft piece are classified as elk (Cervus elaphus). The total weight of these specimens is 66.7 g. Both elements were found in Shovel Test 150W5 at a depth of 72-78 cmbs. A single adult elk is indicated. Neither element shows cut marks or evidence of modification. Elk typically occur in forest border and open forest habitats, as well as edge areas in grasslands and prairies.

Eight white-tailed deer (Odocoileus virginianus) remains weighing 78.1 g are present in the assemblage, and one is thermally-altered (burned or calcined). Both adult and young adult elements were observed, indicating a minimum of two individuals. Deer elements include a right maxilla, a left pelvis fragment, a lumbar vertebra centrum, two cheek tooth fragments, and several metapodial

140 fragments. Half of the remains are from shovel tests and half were recovered from surface contexts. One metacarpus fragment from Shovel Test 160N5 has been fashioned into an awl. It is 104.1 mm long, 28.9 mm wide, and 21.2 mm thick, with a honed and polished shaft extending from the intact, unmodified distal end. None of the deer elements display butchery marks. White-tailed deer occur in a variety of habitats but prefer open forest and forest-edge settings.

Three tooth fragments and 39 pieces of antler, weighing a total of 19.8 g, are listed as indeterminate elk/deer (Cervidae). All of the indeterminate cervid remains were recovered from shovel tests. The antler fragments were found in a single shovel test (Shovel Test R125 from 50-70 cmbs) and may be from a single set of antlers. None of the bones display butchery marks or modification evidence, but the antler fragments may represent curated tool stock.

A total of 35 specimens, with a total weight of 112.0 g, are classified as indeterminate bison or elk (Bison/Cervus sp.). One is thermally-altered (burned or calcined). Of these, 29 are teeth or tooth fragments and most of the remainder consist of rib shaft pieces. Twenty-eight elements weighing 60.4 g are from shovel tests, while seven specimens weighing 51.6 g were found in excavation units. None of the indeterminate bison/elk remains display butchery marks or modification evidence.

Forty-six specimens weighing 646.4 g are categorized as indeterminate cattle or bison (Bos/Bison sp.). One is thermally-altered (burned or calcined). The majority of remains are large, non-diagnostic tooth fragments and long bone shaft pieces. None exhibit butchery marks or modification evidence, although one 2nd phalanx has rodent gnaw marks and one radius fragment shows weathering from exposure. Nearly three-quarters of the specimens are from surface contexts, and of the remainder, all but one are from shovel tests.

Two horse (Equus caballus) bones weighing 51.8 g were recovered, and consist of a distal metacarpus fragment and a mandibular molar fragment. Both specimens are from the surface and reflect modern or historic remains. No cut marks or modification evidence was noted on either element, and both specimens are from an adult individual.

Eight gray wolf (Canis lupus) bones with a total weight of 11.6 g are present in the assemblage. All of the wolf remains were found in XU 7 at a depth of 40-50 cmbd. A single adult individual is indicated. The remains consist of a maxilla fragment, five right and left maxillary teeth, and several tooth fragments. None of the wolf bones display butchery marks or evidence of modification.

Four pieces of bone weighing 15.1 g are listed as indeterminate dog/wolf/coyote (Canis sp.). Elements present consist of an axis, the distal epiphysis of a left fibula, a left 2nd maxillary incisor, and a right maxillary 4th premolar. The fibula and incisor are from shovel tests, and the axis and premolar were found in XU 2. At least two individuals are represented based on the presence of adult and juvenile elements. Evidence of periostitis was recognized on the axis, which is from an older, mature animal. Based on the size of the axis, it is likely from a large dog (Canis familiaris) or a gray wolf. The incisor, in contrast, compares favorably with coyote (Canis latrans) but cannot be specifically identified. No butchery marks or evidence of modification was observed on any of the indeterminate dog/wolf/coyote remains. Gray wolves prefer forests and forest-edge settings.

One left radius fragment, a left scapholunar, and the proximal portion of metacarpal, with a total weight of 0.9 g, are classified as gray fox (Urocyon cinereoargenteus). A single individual is represented, and all three bones were found in Shovel Test R63E5 from 50-75cmbs. The radius is burned, and none of the other remains display butchery marks or modification evidence. Gray foxes are found in hardwood and mixed coniferous-hardwood forests, as well as brushy areas and hilly terrain.

141

One canine fragment weighing 0.4 g is categorized as indeterminate large carnivore (large Carnivora). The size of the tooth is consistent with a raccoon-sized or larger carnivore. The specimen was recovered in Shovel Test 63 from 55-75 cmbs. No butchery marks were observed on the tooth, and no modification evidence was recognized.

Twenty-one bones and teeth weighing 34.4 g are identifiable as beaver (Castor canadensis). A minimum of two individuals is indicated, with both adult and juvenile animals represented. None of the beaver remains display butchery marks or modification evidence. All of the specimens were recovered from shovel tests. Beavers inhabit aquatic and riparian settings in proximity to forests and wooded areas.

One left mandible and a lumbar vertebra fragment, with a total weight of 0.7 g, are classified as muskrat (Ondatra zibethicus). Both elements were found in shovel tests. No butchery marks or evidence of modification was observed on either specimen. Muskrats are semiaquatic and inhabit marshes, ponds, lakes, and streams.

One ilium fragment weighing 0.4 g is categorized as hare or rabbit (Lepus/Sylvilagus sp.). The element was recovered from a shovel test and displays no cut marks or modification evidence. White- tailed jack rabbits (Lepus townsendii) are found in grassland settings, while snowshoe hares (Lepus americanus) occupy brushy woodlands and heavy mixed forests. Cottontail rabbits (Sylvilagus floridanus) occur in various settings but are most common in forest-edge and open forest habitats, grasslands, and brushy areas.

A total of 106 specimens, with a total weight of 22.3 g, are identifiable as plains pocket gopher (Geomys bursarius). A minimum of six individuals is indicated, with both adult and juvenile animals represented. Essentially all body portions are present, with mandibles and appendicular elements the most common. None of the pocket gopher elements exhibit butchery marks or evidence of modification. Sixty-nine specimens weighing 12.6 g were found in shovel tests, and 37 remains weighing 9.7 g were recovered from excavation units. In several instances, relatively large concentrations of gopher bones were found together, and in one case a nearly complete juvenile individual was recovered from a rodent burrow. Based on the context of these finds, it is likely that most if not all of the gopher elements represent intrusive or commensal remains. Plains pocket gophers occur in dry meadows, prairies, and grasslands.

A wide variety of other small rodents and insectivores was recognized in the assemblage, including eastern chipmunk (Tamias striatus), Franklin’s ground squirrel (Poliocitellus franklinii), thirteen- lined ground squirrel (Ictidomys tridecemlineatus), northern short-tailed shrew (Blarina brevicauda), indeterminate vole (Microtus sp.), and indeterminate rodent (Rodentia). As with the plains pocket gopher, these small mammal remains likely do not represent dietary items. The taxa recognized occur in a range of habitats including prairies, grasslands, and open forests. None of the remains in this group display butchery marks or evidence of modification. The small rodent and insectivore remains were recovered from shovel tests and excavation units.

The remaining mammal bones are categorized according to approximate live weight. A total of 380 specimens weighing 1,796.9 g are listed as extra-large mammal and likely represent unrecognizable bison or possibly elk elements. A total of 23 are thermally-altered (burned or calcined). These remains were found across the site area and were recovered from the surface, shovel tests, and excavation units. If recovered from the surface or plow zone, the extra-large mammal could also be cattle. Given that there are 311 identifiable bison remains and only two elk, it is likely that most of the subsurface extra-large mammal remains are bison. Butchery marks were noted on one extra-large mammal long bone shaft piece and one indeterminate fragment.

142

Large-sized mammal remains account for 1,332 specimens with a total weight of 1,813.8g. These are wolf-size or larger animals. Given that most of the identifiable remains recovered from the same or adjacent tests are bison, elk, or extra-large mammal, it is likely that most large mammal remains are also bison or possibly elk, with smaller amounts likely to be deer and wolf. The remains were found across the site area and were recovered from the surface, shovel tests, and excavation units. A total of 297 large-sized mammal remains are thermally-altered, and four elements display butchery marks. Two bones show evidence of modification. The first is an indeterminate bone from Shovel Test 151 at 70-100 cmbs weighing 3.0 g that has a polished exterior and smoothed interior surface and edges, which likely represents a tool fragment. The second modified bone, recovered from Shovel Test 104W5S2.5 at 60-90 cmbs weighing 1.7 g is a worked, disk-shaped tool fragment.

A total of 152 bone fragments weighing 51.3 g are categorized as medium-large size mammal, and these remains could be fox, coyote, dog, or beaver to deer and larger-sized animals. A total of 14 specimens are thermally-altered. Most are indeterminate to element, with a limited number of tooth fragments and long bone shaft pieces also present. The remains were found across the site area and were recovered from the surface, shovel tests, and excavation units. None of the medium-large mammal remains display butchery marks or modification evidence.

Only five medium-size mammals were recovered, including three phalanges and two vertebrae. These remains could be beaver, coyote, dog, fox, or similar sized animals. The remains were recovered from the surface, shovel tests, and excavation units. One bone is thermally-altered, but none of the remains display butchery marks or modification evidence.

Two specimens are classified as small-sized mammal, and these remains could be rabbit, gopher, chipmunk, muskrat, squirrel, or vole. One is a long bone shaft piece weighing 0.3 g, and was found in a shovel test. The other is a lumbar vertebra fragment weighing 0.1 g that was recovered from an excavation unit. The lumbar vertebra is from an adult animal. None of the small-sized mammal bones are thermally-altered or exhibit butchery marks or evidence of modification. The remains were found across the site area and were recovered from a shovel test and excavation unit.

Birds Only 18 bird bones are present in the assemblage, and six are thermally-altered. One right femur shaft fragment is classified as a large owl, most likely either a barred owl (Strix varia) or great horned owl (Bubo virginianus). The element weighs 2.9 g and is from an adult owl. Barred owls inhabit woodlands, wooded river bottoms, and forested swamps, and tend to favor dense and thick forests with few clearings. Great horned owls occur in a variety of habitats but are most common in forests, wooded riparian settings, and open areas. It is also possible, however unlikely, that the femur is from a snowy owl (Nyctea scandiaca), a seasonal migrant found in prairies and marshes. No butchery marks or modification evidence was noted on the owl element.

One long bone shaft piece with a weight of 0.2 g is categorized as medium-sized bird, and one tibiotarsus fragment weighing 0.1 g is listed as small-sized bird. Fourteen long bone shaft fragments and a terminal phalanx, with a total weight of 3.2 g, are categorized as indeterminate bird. None of the medium-sized, small-sized, or indeterminate bird bones display cut marks or evidence of modification. Eleven indeterminate bird bones and the small-sized bird tibiotarsus were recovered from excavation units. Three indeterminate bird elements, the owl femur, and the medium-sized bird bone were found in shovel tests. One indeterminate bird long bone shaft piece was obtained from feature context.

Reptiles and Amphibians A small number of reptile and amphibian remains were recovered. Nine turtle carapace and plastron fragments weighing 1.9 g are present, including two pieces that are thermally-altered. All of the turtle 143 bones are from shovel test or surface context. Four trunk vertebrae weighing 0.3 g are identifiable as indeterminate garter snake (Thamnophis sp.) and two trunk vertebrae with a total weight of 0.3 g are listed as indeterminate non-venomous snake (Colubridae). The garter snake elements are from XU 5 and Feature 1 (located in XU 5 and XU 6), and both non-venomous snake bones are from shovel tests. None of the reptiles remains in the assemblage display butchery marks or evidence of modification. Garter snakes, depending on the species, inhabit prairies, marshes, wet meadows, forests, and forest-edge settings.

One left ilium weighing 0.1 g is classified as indeterminate toad (Anaxyrus sp.). One tibiofibula shaft piece with a weight of 0.1 g is listed as indeterminate frog/toad. The toad bone was recovered from a shovel test, while the frog/toad tibiofibula was found in a rodent burrow in XU 1. No modification or butchery marks were observed on either specimen. Toads occur in a variety of aquatic and terrestrial settings, and show greater affinity for the latter than do frogs. These are probably naturally occurring and not subsistence remains.

Fish Fourteen fish elements are present in the assemblage. Seven specimens weighing 3.1 g are identifiable as channel catfish (Ictalurus punctatus). A minimum of two individuals is indicated. Channel catfish are typically found in medium to large rivers but may also occur in lakes and larger streams.

One jaw and six cranial fragments, with a total weight of 1.4 g, are listed as indeterminate fish. One catfish coracoid was found in an excavation unit. None of the channel catfish or indeterminate fish remains show butchery marks or evidence of modification. All of the indeterminate fish and six of the seven channel catfish bones were recovered from shovel tests.

Mollusks and Gastropods Twenty-eight pieces of mollusk shell weighing 30.1 g are classified as indeterminate mussel. One valve fragment from XU 1 compares favorably with the fatmucket/pocketbook (Lampsilis sp.) mussels but cannot be specifically identified. No butchery marks or evidence of modification was noted on any of the freshwater mussel remains. Shell fragments were recovered from shovel tests, excavation units, and surface contexts, and may represent a mix of culturally derived and naturally occurring remains.

Naturally occurring invertebrates observed in the assemblage include a single pea clam (Sphaeriidae) and a pond snail (Lymnaeidae). The specimens were recovered from shovel tests, and none exhibit butchery marks or modification evidence. They are not considered dietary or cultural items.

9.14.2 Faunal Distribution

With regard for quantity and weight, approximately 60 percent of the faunal material was recovered from shovel tests, and over 30 percent was obtained from excavation units, with a small amount of bone found in Feature 1 (Table 41). Relatively little bone and shell is from surface contexts, and it likely represents a mix of precontact, historic, and naturally occurring faunal material.

144

Table 41. Site 21PL109 Distribution of Faunal Remains by Context. Shovel Test XU Surface

Taxon Wgt Wgt Wgt NISP (g) NISP (g) NISP (g) MAMMALS Bison (Bison sp.) 199 3505.3 112 1956.5 -- -- Elk (Cervus elaphus) 2 66.7 ------White-tailed deer (Odocoileus virginianus) 4 24.5 -- -- 4 53.6 Bison/elk, indeterminate (Bison/Cervus sp.) 28 60.4 7 51.6 -- -- Cattle/bison, indeterminate (Bos/Bison sp.) 12 144.2 1 0.5 33 501.7 Deer/elk, indeterminate (Cervidae) 42 21.3 ------Horse (Equus caballus) ------2 51.8 Gray wolf (Canis lupus) -- -- 8 11.6 -- -- Dog/wolf/coyote, indeterminate (Canis sp.) 2 0.6 2 14.5 -- -- Gray fox (Urocyon cinereoargenteus) 3 0.9 ------Large carnivore, indeterminate (large Carnivora) 1 0.4 ------Beaver (Castor canadensis) 21 34.4 ------Muskrat (Ondatra zibethicus) 2 0.7 ------Hare/rabbit, indeterminate (Lepus/Sylvilagus sp.) 1 0.4 ------Plains pocket gopher (Geomys bursarius) 69 12.6 37 9.7 -- -- Franklin's ground squirrel (Poliocitellus franklinii) 1 0.5 1 0.1 -- -- Thirteen-lined ground squirrel (Ictidomys tridecemlineatus) 1 0.1 ------Eastern chipmunk (Tamias striatus) 2 0.2 1 0.1 -- -- Vole, indeterminate (Microtus sp.) -- -- 3 0.4 -- -- Northern shrew (Blarina brevicauda) -- -- 1 0.1 -- -- Rodent, indeterminate (Rodentia) 7 0.7 10 1.1 -- -- Extra large mammal 225 1043.6 140 603.4 15 149.9 Large-sized mammal 775 965.3 546 807.2 11 41.3 Medium-large mammal 146 47.3 5 3.1 1 0.9 Medium-sized mammal 2 1.0 2 0.7 1 0.6 Small-sized mammal 1 0.3 1 0.1 -- -- Mammal, indeterminate 10814 2921.1 5304 1224.2 208 294.6 BIRDS Large owl, indeterminate (cf. Bubo/Strix sp.) 1 2.9 ------Bird, indeterminate 4 0.5 13 3.0 -- -- REPTILES and AMPHIBIANS Turtle, indeterminate 8 1.4 -- -- 1 0.5 Garter snake, indeterminate (Thamnophis sp.) -- -- 4 0.3 -- -- Non-venomous snake, indeterminate (Colubridae) 2 0.3 ------Toad, indeterminate (Anaxyrus sp.) 1 0.1 ------Frog/toad, indeterminate -- -- 1 0.1 -- -- FISH Channel catfish (Ictalurus punctatus) 6 2.4 1 0.7 -- -- Fish, indeterminate 7 1.4 ------Mollusks and Gastropods Mussel, indeterminate 7 2.3 11 11.5 10 16.3 Pea clam, indeterminate (Sphaeriidae) 1 0.1 ------Pond snail, indeterminate (Lymnaeidae) 1 0.2 ------Indeterminate Taxon indeterminate (Vertebrata) 64 9.2 942 63.8 -- -- Total 12,462 8873.3 7153 4764.3 286 1111.2

145

9.14.3 Faunal Discussion

For this analysis, the site 21PL109 assemblage is considered as a whole, which allows for only a general discussion of faunal exploitation at the site. More detailed information on the specific fauna associated cultural components is presented in the sections discussing XU results. The composition of the assemblage suggests a subsistence adaptation centered on bison, with some inclusion of other large and small mammals, birds, turtles, fish, and shellfish. The types of bison elements recovered, in particular the cranial and distal limb bones, indicate that at least some bison were likely killed and processed at or in close proximity to the site area. The recovery of several projectile points, some in association with bison, and stone tools (scrapers and utilized flakes) for butchering and processing supports the interpretation of a kill and processing site. Elk, white-tailed deer, muskrat, and hare remains occur in very small amounts, with slightly more beaver. Other aquatic taxa represented include small amounts of channel catfish, turtle, and shellfish. Several bird remains are present, but with the exception of the owl femur, none are specifically identifiable. Several carnivores (e.g., gray wolf, gray fox, and indeterminate canid) are represented, but these taxa are not usually viewed as dietary items. The wolf and fox remains may reflect animals trapped for their pelts or skeletal remains, possibly for use as decoration, ornamentation, or as ritual paraphernalia. A moderate amount of rodents are present, but most if not all likely represent commensal or natural items, and not dietary refuse. No specific indicators relating to season of occupation were discerned.

The taxa identified demonstrate exploitation of prairie (e.g., bison), forest-edge and forest settings (e.g., elk, deer, gray wolf, gray fox, hare/rabbit), and aquatic and riparian habitats (e.g., beaver, muskrat, channel catfish, turtle, mussel) of the Red River. An exploitation strategy that mirrors the relative abundance of various prey resources is typical of sites located in or near prairie-forest border areas, regardless of time period (e.g., Haury 2005; Mather et al. 1998; Shay 1971, 1978). The site occupants undoubtedly made use of the wide variety of locally available faunal resources, and focused their attention on select species during periods of seasonal abundance or availability.

9.15 Lithic Analysis

The lithic assemblage consists of 799 artifacts, including 755 pieces of lithic debris, 42 stone tools, and two cores (Table 42). A variety of flake types, tools, cores, and lithic materials are present in the assemblage, which is discussed below.

Cortex, Size Grades, and Heat Treatment The amount of cortex on all the lithic artifacts (debris, tools, and cores) is as follows: 0% cortex (n=739; 92%); >0-<50% cortex (n=25; 3%); 50-<100% cortex (n=20; 3%); and 100% cortex (n=15; 2%). Size grade counts for the lithic debris were as follows: SG1 (n=7; <1%); SG2 (n=88; 11%); SG3 (n=308; 39%); and SG4 (n=396; 50%). The lack of cortex and the small size of debris suggest that the lithic assemblage is comprised primarily of tools and middle and late-stage reduction debris. This pattern fits with the scarcity of lithic resources available in the area. Evidence for early stage reduction, which would include larger pieces of lithic debris and great percentages of cortex is conspicuously lacking.

A total of 630 lithic artifacts were likely heat treated, with most of these artifacts being Swan River Chert, but also including small amounts of Prairie du Chien Chert and Tongue River Silica, with one piece each of Knife River Flint, unidentified chert, and Red River Chert. Probable heat treatment was observed on 11 additional lithics in a wide variety of materials, including petrified wood, rhyolite, and Grand Meadow, fusilinid, and unidentified cherts. Seven lithics, including Knife River Flint and unidentified chert, showed evidence of excessive heating, as indicated by crazing and potlid fractures.

146

Table 42. Site 21PL109 Lithic Artifacts by Material, Flake, and Tool/Core Types.

Material Tool/Core Total % Flake Flake Flake Flake Other Other Shatter Shatter Bipolar Bifacial Bifacial Shaping Grade 4 Grade Thinning Nonbifacial Decortication Broken Flake Flake Broken 1 nonbifacial core 1 stage 2 biface 1 stage 3 biface 4 stage 4 bifaces 1 stage 5 biface Swan River Chert 27 4 105 27 1 192 10 151 1 indeterminate 541 68 tool 4 projectile points 1 retouched flake 3 scrapers 7 utilized flakes Tongue River Silica 4 - 12 4 2 50 - 14 1 nonbifacial core 87 11 2 projectile point 2 scrapers Knife River Flint 2 - 23 4 5 13 - 6 66 8 8 utilized flakes 1 retouch flake Unidentified Chert 5 3 6 - 1 - 1 8 1 utilized flake 25 3 Rhyolite - 2 1 3 - 5 - 9 - 20 3 Quartzite 7 - - - 1 1 - 6 - 15 2 Prairie du Chien - - 7 2 - 3 - - 1 projectile point 13 2 Chert* Red River Chert - 1 - - - 2 3 1 1 scraper 8 1 Unidentified 3 1 - - - - - 3 - 7 1 material Quartz - - - - 1 3 - 2 - 6 1 Basaltic - - - - - 1 - - 2 side scrapers 3 <1 Chalcedony 1 ------1 - 2 <1 Igneous - - - - 1 - - 1 - 2 <1 Bijou Hills Quartzite - - - 1 - - - - - 1 <1 Fusilinid Chert - - - 1 - - - - - 1 <1 Grand Meadow Chert - - - 1 - - - - - 1 <1 Petrified Wood ------1 scraper 1 <1 Total 49 11 154 43 12 270 14 202 44 799 - % 6 1 19 5 2 34 2 25 6 - 100 * 12 are Prairie du Chien Chert (oolitic)

Flake Types The wide variety of flake types in the assemblage indicates a range of lithic-reduction technologies and stages. Diagnostic flake types, along with their associated technologies and stages of reduction, are summarized in Table 43. Bifacial technology is well-represented in the assemblage, with smaller

147 amounts artifacts from nonbifacial and bipolar technologies. Most of the lithics are from the middle and late stages of reduction, with a small amount of lithics from the early stage of reduction. Additional supporting evidence for the various technologies includes: 1) two nonbifacial cores and nine tools (scrapers and utilized flakes) made on nonbifacial and decortication flakes are indicative of nonbifacial technology; and 2) five tools (utilized flakes and a scraper) made on bifacial thinning and shaping flakes, eight unfinished bifaces (Stages 2 to 5), and six projectile points are indicative of bifacial technology.

Types of lithic debris that are not indicative of specific technologies or reduction-stages comprise a large percentage of the assemblage and include broken and other SG4 flakes. These nondiagnostic flake types are not included in Table 43.

Table 43. Site 21PL109 Summary of Diagnostic Flake Types, Technologies, and Reduction Stages. Count & Technology Stage of Reduction Flake Type 154 - Bifacial shaping Bifacial Late-stage of reduction (final shaping or tool maintenance) flake 43 - Bifacial thinning Bifacial Early to middle-stage of reduction flake 11 - Decortication Nonbifacial Earliest stage of core reduction and raw material testing flakes Cobble testing, reducing unprepared nonbifacial cores for 49 - Nonbifacial flakes Nonbifacial flake blank production, and the early stages of nonbifacial tool reduction (early to middle-stages of reduction) Mostly from cobble testing, core reduction, and earlier 14 - Shatter N/A stages of reduction 12 - Bipolar flakes Bipolar N/A

Lithic Material Types and Use Lithic materials consisted primarily of Swan River Chert (68%) with substantially smaller amounts of Tongue River Silica (11%), Knife River Flint (8%), and many other materials that each comprises 3% or less. Nearly all of the materials are locally available. The unidentified chert may be local or nonlocal, and the presence of decortication flakes suggest at least some of the unidentified chert is local.

The assemblage contains a small amount of non-local, high-quality materials that were likely acquired through exchange networks or travel, including: 1) Prairie du Chien and Grand Meadow cherts from southeastern and south-central Minnesota; 2) Knife River Flint from western North Dakota; 3) Bijou Hills Quartzite from the middle portion of South Dakota; and 4) Fusilinid Chert from parts of Iowa, Missouri, Kansas and Nebraska.

The lithic data indicates that the raw materials have different debris profiles resulting from availability, differential use, quality of material, and cobble size. The most notable lithic use characteristics are discussed below for materials that have adequate sample sizes of diagnostic flakes. However, many materials lack an adequate sample size.

The small amount of decortication flakes from initial reduction and the small amount of debris with 50 to 100% cortex is because of the absence of raw material sources on the interior of the lake plain. For example, no naturally occurring flakeable-sized rocks were present in the soils at the site, and no rocks were observed in the river cut banks or along the river bottom. The interior portion of the Lake Agassiz plain is characterized by a nearly complete absence of rock, and it is likely there are no abundant sources for several miles around the site. The beach ridges about ten miles to the east are 148 likely the nearest sources of tool stone. In general, most of the raw materials brought to the site were tools or rock fragments that were already reduced into flake and biface blanks.

Swan River Chert occurs in all flake types and technologies, but it is most numerous in bifacial technology as bifacial shaping flakes. It occurs in smaller amounts as bifacial thinning flakes and in unifacial technology as decortication and nonbifacial flakes. It also occurs in small amounts as shatter and a bipolar flake. Although it was used for all stages of lithic reduction and tool production, it is most abundant in late-stage bifacial finishing or tool maintenance. Twelve bifacial tools, eleven flake tools, and a nonbifacial core were manufactured from Prairie du Chien Chert.

Unidentified chert occurs in small amounts in all technologies as nonbifacial and decortication flakes and also as bifacial shaping flakes, shatter, and a bipolar flake. Tongue River Silica occurs primarily as bifacial shaping flakes, with small amounts of bipolar, bifacial thinning, and nonbifacial flakes. Tongue River Silica also occurs as a nonbifacial core. Quartzite occurs primarily as nonbifacial flakes.

Prairie du Chien Chert occurs only as bifacial shaping and thinning flakes and a projectile point. Knife River Flint occurs primarily as bifacial shaping flakes, with small amounts of bipolar, bifacial thinning, and nonbifacial flakes. A notable amount of bifacial and flake tools were made on Knife River Flint. Bijou Hills Quartzite, Fusilinid Chert, and Grand Meadow Chert occur only as single bifacial thinning flakes. The debris profile for these nonlocal materials is nearly exclusively bifacial, with a conspicuous absence of decortication, shatter, and nonbifacial flakes. This use pattern is expected for materials that are transported a long distance from their source, as they were likely brought to the site as bifacial blanks and tools where they were shaped and reworked until no longer useable. Bifacial technology is known to allow for an economical use of raw materials, and it would have been a particularly adaptive strategy for the Lake Agassiz plain, which lacks available lithic materials for a distance of about 30 miles from the site.

Stone Tools A total of 42 stone tools were recovered (Table 44), including 16 utilized flakes, seven projectile points (and fragments), nine scrapers (an end scraper, four side scrapers, and four side/end scrapers), two retouched flakes, one indeterminate biface tool fragment, one Stage 2 biface (early-stage), one Stage 3 biface (middle-stage), four Stage 4 bifaces (late-stage), one Stage 5 biface (late-stage). These tools were recovered from various locations across the site area.

One large and heavy granite probable manuport (21 x16 cm; 11,440 grams) was recovered in Shovel Test R148 from 40 to 50 cmbd. This was the largest rock observed in the survey area. The function of this stone is uncertain, but it could have been used as an anvilstone or crushing stone. It is also possible that the stone is naturally occurring.

Approximately half of all tools were made on Swan River Chert, including unifacial tools, bifacial tools, and early and late stage bifaces. Knife River Flint was used for about a third of tools, including flake tools and projectile points. The amount of Knife River Flint that occurs as tools (31%) notably exceeds its overall amount (8%) in the lithic assemblage, indicating that it was preferentially selected for tools because of its high quality or cultural value. Very small amounts (<2%) of tools were made from a variety of other materials, including basalt, petrified wood, and Prairie du Chien, Red River, and unidentified cherts. Projectile points (and fragments) were made on Knife River Flint, Prairie du Chien Chert, and Swan River Chert. Moderate to high quality raw materials were selected for tool manufacture.

149

Table 44. Site 21PL109 Tool Type by Material Type.

Materials Total % Bifacial Bifacial End Scraper End Scraper Side Scraper Side Scraper Utilized Flake Utilized Flake Indeterminate Biface, Stage 4 Biface, Stage 3 Biface, Stage 2 Projectile Point Retouched Flake Flake Retouched Side/End Scraper Side/End Scraper Swan River Chert 2 - 1 1 7 1 4 5 1 1 23 55 Knife River Flint - - 2 - 8 1 2 - - - 13 31 Basaltic 2 ------2 5 Prairie du Chien Chert* ------1 - - - 1 2 Petrified Wood - 1 ------1 2 Unidentified Chert - - - - 1 - - - - - 1 2 Red River Chert - - 1 ------1 2 Total 4 1 4 1 16 2 7 5 1 1 42 100 *1 Prairie du Chien Chert (oolitic)

The Stage 2, 3, and 4 bifaces may have been manufactured for projectile points and/or could have been used as cutting tools. About half of the bifaces are whole, and the other half are broken. Utilized and retouched lakes are primarily light-duty cutting and slicing tools used on animal remains, wood, and plants. Scrapers are typically associated with scraping tasks on a variety of soft materials (meat, hides, and plant material) or moderately resistant materials (wood and bone). Given the abundance of fauna at the site, it is likely that most of these tools were used for animal processing. These tools suggest that site activities included butchering, animal/plant processing, hide working, and bone and woodworking. Nonbifacial flake tools were mostly made on broken and nonbifacial flakes with smaller amounts also made on bifacial thinning flakes (Table 45).

Table 45. Site 21PL109 Nonbifacial Flake Tools by Flake Type. Type of Flake That the Tool is Made From

Tool Type Total Broken Bifacial Bifacial Shaping Thinning Nonbifacial Nonbifacial Decortication Utilized Flake 1 1 3 1 10 16 Retouched Flake - - - - 1 1 Side/End scraper - 3 - - 1 4 Side scraper - 2 1 - - 3 End scraper - 1 - - - 1 Projectile point - - - - 1 1 Total 1 7 4 1 13 26

150

Projectile Points

A small, side-notched Oxbow Middle Archaic point made of Swan River Chert was recovered in Shovel Test 68E5 from 75 to 100 cmbs on Terrace 4 (Figure 26; Catalog # 98.7). The point is made on a biface, and the faces of both sides are covered with well-executed pressure flakes. The blade is triangular in shape and was likely resharpened. The point is widest at the basal ears, and the base is thinned, concave, and ground. On one face, the basal thinning was accomplished by the removal of a single large flake, while the thinning on the other face consisted of a few smaller flake removals. The point is very thin and lenticular in cross-section. The point is 19.5 mm long, 21.0 mm wide (at basal ears), and 4.6 mm thick. The blade width above the notches is 16.9 mm, and the width between the notches is 16.0 mm. Based on the size and morphological and technological attributes, the point most closely resembles the Oxbow type that dates from 5500 to 3200 RCYBP / 6300 to 3450 cal BP (Eighmy and LaBelle 1996; Frison 1991; Green 1998:276-280; Morrow et al. 2016; Perino 1971). Radiocarbon dates obtained from XU 7, which is adjacent to Shovel Test 68E5, yielded dates of 6240 +/- 30 RCYBP from 40 to 50 cmbd and 6880 +/- 30 RCYBP from 70 to 80 cmbd. The date of the point likely falls somewhere within this range, which is older than the bulk of established Oxbow dates, but within range of some of the earliest dates.

A basal fragment of an Early Archaic small, side-notched point made of Swan River Chert was recovered in XU 9 from 100 to 110 cmbd on Terrace 4 (Figure 26; Catalog # 41.298). A radiocarbon date of 7110 +/- 30 RCYBP was obtained from the same level in XU 9, providing a secure date for the point. The point is broken diagonally and only one basal side is present, with the blade and other half of the base absent. The base is thinned, slightly concave, and ground. The basal ear is squared off. The notch is well shaped and placed low on the base. The basal ear is in line with the blade above the notch. The point is lenticular in cross-section and thin (4.3 mm thick). Because of the breakage, it is not possible to determine the specific technology of manufacture. The point appears to have been made on a flake or thin biface with marginal pressure flaking. The basal remnant is 15.2 mm wide, although it’s full width is indeterminate because of the breakage. Based on the size, morphology, and technology of the point, it most closely resembles the small, Early Side-Notched Point type recovered at other sites in the region, which have similar dates of around 7000 RCYBP, such as 21YM47 Granite Falls Bison Kill (Kuehn 2000; Peterson 1996), 21CE1 Itasca Bison Kill (Shay 1971; Widga 2006), and 32RI775 Rustad (Michlovic and Sather 2005). The variations of the early side-notched points are reflected in the proliferation of various type names. The name Little Sioux was initially proposed for similar points in Iowa by Morrow (1984) and has been more recently replaced by Simonsen (Morrow et al. 2016:162).

A small-sized, projectile point preform or Stage 4 Biface made of Swan River Chert was recovered in XU 10 from 100 to 110 cmbd on Terrace 4 (Figure 26; Catalog # 54.64). A radiocarbon date of 7110 +/- 30 RCYBP was obtained from the same level in adjacent XU 9, providing a secure date for the preform. The preform is made on a biface, with percussion flakes covering both faces and intermittent pressure flaking along the margins. The preform is 31.5 mm long, 20.2 mm wide, and 7.2 mm thick. The preform is similar in dimensions to the Early Archaic Side-Notched Point recovered in XU 9 at 100 to 110 cmbd (discussed above), and is likely an unnotched version of that point type.

A tip of a projectile point made of Swan River Chert was recovered in XU 5 from 90 to 100 cmbd on Terrace 4 (Figure 26; Catalog # 7.83). The point tip appears to be made on a biface, as both faces are completely flaked and well-controlled pressure flaking occurs along the margins. The tip is 4.2 mm thick. It is not possible to determine its age or cultural affiliation because of its fragmentary condition.

A small, stemmed Duncan-Hanna Middle Archaic point made of Knife River Flint was recovered in Shovel Test 252 from 80 to 100 cmbs on Terrace 3 (Figure 27; Catalog # 543.1). The point was 151 recovered in-situ from the wall profile, so the provenience of its depth is well-controlled. The point is made on a flake and has pressure-flaking intermittently along the margins of both faces. The blade form is convex. The base is very slightly concave and not ground, with somewhat asymmetrical basal ears. The point is very thin and lenticular in cross-section. The side notches are prominent and have a somewhat rectangular shape. The point is widest at the blade just above the stem. The point is 22.9 mm long, 12.9 mm wide, and 2.3 mm thick, with a stem width of 8.0 mm. The point appears to be a miniature version of the Duncan-Hanna type (4000 to 3000 BP) (Davis and Keyser 1999; Morrow et al. 2016:121, 178). Tiny versions of Duncan-Hanna have been reported (Davis and Keyser 1999:253), and a study of miniature points indicated that they are relatively common at Plains habitation sites of a various ages (Dawe 1997). The study, which included an ethnographic review, indicated that these tiny points are probably kids’ toys. The age of the point corresponds with the age of the soil anticipated on Terrace 3 at 80 to 100 cmbs.

A medium-sized, side-notched Besant Late Plains Archaic point made of Knife River Flint was recovered in Shovel Test 148 from 70 to 90 cmbs on Terrace 4 (Figure 27; Catalog # 544.1). The point was recovered in-situ from the wall profile, so the provenience of its depth is well-controlled. The point is made on a biface, with percussion flakes present near the middle portion of the point and pressure flaking along the margins. The blade is convex in shape, and the tip is broken off. The point is widest at the blade just above the notches and is thin and lenticular in cross-section. The base is thinned, straight, and ground. The point is 30.9 mm long, 20.2 mm wide, and 6.2 mm thick. The width between the notches is 10.6 mm, and the basal ears are 14.9 mm wide. The point most closely resembles the Besant type, based on the size and morphological and technological attributes (Frison 1991:106; Kehoe 1974; Morrow et al. 2016:122, 198). According to Morrow et al. (2016:198), “The Besant phase or culture, with which the point type is at least in part associated, is generally dated between about 1 and 700 C.E. ( Forbis 1998; Gregg and Davidson 1985:118). Cloutier (2004:16–17) offers a revised age of 500 B.C.E. to 700 C.E.” These points are also described as corner-notched by Kehoe (1974). A radiocarbon date was obtained from fauna in Shovel Test 148 at 60 to 100 cmbs from the same paleosol as the point. The fauna dated to 6070 +/- 30 RCYBP, which is considerably older than Besant and does not appear to provide a date for the point. The radiocarbon date is expected to be accurate, as the date corresponds with other dates on Terrace 4 from XUs 5 to 10. It is not possible to reconcile the discrepancy between the point and dates on associated fauna. The most likely explanation is that the point is not in situ but was vertically displaced downward into an earlier component by rodent burrowing or other natural processes or was displaced by drain tiling that may have occurred in this area.

A large-sized, lanceolate Plainview Late Paleoindian point made of Prairie du Chien Chert (oolitic) was recovered from the surface at Shovel Test R59 on Terrace 4 (Figure 27; Catalog # 43.1). The point is made on a biface, with percussion flakes present near the middle portion of the point and pressure flaking along the margins. Only the basal portion is present, and it is thin and lenticular in cross-section. The base is concave, thinned, and ground. The point is 26.2 mm wide and 5.6 mm thick. Based on its size and morphological and technological attributes, the point most closely resembles the Plainvew type, which dates around 10,000 to 9,000 BP on the southern Plains (Florin 1996; Justice 1987:30). The recovery of the point from the surface clearly indicates that it is not in original context, as an Archaic point was found nearby in XU 9 about a meter below surface and a series of nearby radiocarbon dates suggests that a point of that age would be located deeper than one meter. It is likely that the point was either scavenged and re-used at a much later time or was displaced by drain tiling that may have occurred in this area. Late Paleoindian points have been found on the beaches of Lake Agassiz as well as on the Lake Agassiz plain (Florin 1996).

A corner fragment of a projectile point base or preform made of Swan River Chert was recovered in Shovel Test 146 from plow zone at 0 to 25 cmbs on Terrace 4 (Figure 27; Catalog # 263.2). It appears to be made on a biface, based on its biconvex cross-section and a few percussion flakes. 152

Pressure flaking occurs along the margins, and it is 7.7 mm thick. It appears to be carefully made with well-controlled flaking. The edges are straight and ground. It is not possible to determine its age or cultural affiliation because of its fragmentary condition. The soil context (0 to 25 cmbs) suggests it is probably from the Woodland or Late Archaic traditions.

Cores Two freehand nonbifacial cores with unpatterned flaking and unprepared platforms were recovered. The lithic material of the cores is Swan River Chert and Tongue River Silica.

Lithic Comparison with other Early Archaic Sites The Early and Middle Archaic lithic assemblage from XUs 5 to 10 shows strong similarities to Rustad site (32RI775) (Michlovic and Sather 2005) and site 32RI785 (Root 2001), which are of similar ages and cultural traditions in the Red River Valley in North Dakota. At these sites the most abundant raw material is Swan River Chert, bifaces and bifacial flaking debris is prominent, and stone tools primarily consist of expedient flake tools, scrapers, and side-notched or unnotched projectile points. One difference between these sites and 21PL109 is that they have a wider variety of raw materials and more early stage and nonbifacial reduction debris. These lithic differences are likely because sites 32RI775 and 32RI785 are located on shoreline and delta deposits along Lake Agassiz beaches, where more abundant cobbles were available nearby. Also, these comparative sites had a small amount of Knife River Flint tools and Knife River Flint final stage reduction/tool maintenance debris, which is largely lacking from 21PL109, probably because of the small sample size.

9.16 Ceramic Analysis

A total of 42 ceramic sherds were recovered, with 39 sherds from Terrace 2 at the west end of the site and three sherds from Terrace 3 in the middle portion of the site.

Terrace 2 Ceramics The ceramics on Terrace 2 are most similar to Buchanan and Lisbon wares of the Northeastern Plains Village type (Michlovic and Swenson 1998). The primary diagnostic trait that indicates the sherds belong to this type is a polished surface treatment on some of the sherds, which does not occur on Late Woodland ceramics. In contrast, Sandy Lake ceramics from sites along the Red River are noted as being almost exclusively shell tempered with a cordmarked surface (Johnson et al. 1995:31). Blackduck ceramics are not common, and they also have a cordmarked surface.

A total of 26 sherds have a smooth surface treatment, and these were recovered from XU 1, 2, 4 and Shovel Tests R104E5, 106W5, 107W5, 201, and 202. A total of 18 of these sherds had sparse amount of fine grit temper, and eight sherds had sand. It is likely that these are actually all the same temper, which consists of a sandy paste with sparse amounts of grit. Five smooth sherds recovered from XU 1 and Shovel Tests 201 and 202 had a polished (burnished) surface. Thickness of the sherds is 3.6, 3.8, 4.5, 4.6, 4.8, 5.3, 5.5, 5.6, 5.8, 6.8, 7.0, 7.1, 7.7, and 7.9 mm.

Four sherds have cordmarked surface and grit temper, with three of these sherds having a partially smoothed-over surface. These were from Shovel Tests 201, 202, 279, and 21W5. Thickness of these sherds is 6.6 mm, 5.0 mm, 7.2 mm, and 5.6 mm

Nine sherds have an undetermined surface treatment, and seven of these have grit temper and two have sand. There are exfoliated sherds that lack the exterior.

One rimsherd was recovered in XU 2 from 30 to 40 cmbd. The rim is small (1.5 cm by 2.0 cm). It has a smooth surface, sand temper, and a thickness of 4.8 mm. The lip is undecorated and flat to very slightly rounded. The rimsherd is too small to confidently type, but is assigned to Buchanan and 153

Lisbon wares of the Northeastern Plains Village type based on its association with the polished sherds.

Terrace 3 Ceramics The ceramics on Terrace 3 are from Shovel Test 232 at the west end of the terrace and Shovel Tests 219 and 279 at the east end of the terrace. The sherd from Shovel Test 232 (0-20 cmbs) had grit temper, smooth surface, and an absent interior so that thickness could not be determined. The sherd from Shovel Test 279 (25-30 cmbs) had grit temper, smoothed-over cordmarked surface, and a thickness of 5.4 mm. The sherd from Shovel Test 219 (35-60 cmbs) had grit temper, smoothed surface, and a thickness of 7.8 mm. These ceramics are not diagnostic of specific wares, and they are classified only as Woodland or Northeastern Plains Village ware.

9.17 FCR

A total of 60 pieces of FCR was recovered from find spots, shovel tests, and XUs in scattered locations across the site with no apparent concentrations (Table 46). The small amount of FCR recovered from the site was likely because of the scarcity of stone in the site vicinity. FCR are mostly small-sized pieces, with 90 percent Size Grade 1 or smaller. The angular FCR type (60%) was the most numerous, followed by spalls (22%), crumbs (10%) and smaller amounts of other types. Most of the FCR is granitic (70%) with smaller amounts of igneous rock (20%), quartzite (7%), and basalt (3%). The FCR, in conjunction with thermally-altered fauna and fire hearths, indicates cooking and/or heating activities at the site.

Table 46. Site 21PL109 FCR Count by Material, Size, and Type. Size Grade (SG) FCR Type Total %

Material 00 0 1 2 3

Spall Spall Crumb Angular Cobble with Split Cobble Angular/Spall

Granitic 1 5 7 16 13 21 2 2 1 5 11 42 70 Quartzite - - 1 3 - 4 - - - - - 4 7 Basaltic - - 1 1 - 2 - - - - - 2 3 Igneous (unid.) - - 1 4 7 9 - - - 1 2 12 20 Total 1 5 10 24 20 36 2 2 1 6 13 60 - % 2 8 17 40 33 60 3 3 2 10 22 - 100

9.18 Horizontal Artifact Patterning and Cultural Components

The site is large, measuring 1300 by 120 meters, and extends across a point bar terrace of the Red River. Site testing indicates that there are several vertically discrete components in alluvium that overlap horizontally. The horizontal and vertical delineation of specific site components is only partial and is mostly limited to XUs where radiocarbon dates and diagnostic artifacts were recovered. Shovel test results provide more limited and general data on components. Therefore, the delineation of components at the east and west ends of the site on Terraces 2 and 4, where XUs were placed is better understood than the middle portion of the site on Terrace 3, where only shovel testing was conducted. On Terrace 3 specific data on cultural components is largely lacking because of the sparse amount of diagnostics and absence of radiocarbon dates. The horizontal distribution of artifact

154 classes is biased to some extent by sampling location, as the volume of soil in XUs was much larger than shovel tests, and XUs were placed in locations where specific artifacts were most numerous.

The horizontal patterning and identification of components is also influenced by geomorphic processes and ages of the terraces, whereby the point bar formation occurred progressively from east to west resulting in younger ages of alluvium towards the west, with the effect that older landscapes are progressively more deeply buried or nonexistent to the west. Early and Middle Archaic components were identified in the eastern portion of the site in buried paleo-surfaces on Terrace 4. These components may be very deeply buried on Terraces 2 and 3 in the middle and western portion of the site because of thicker, younger alluvial deposits near the surface, or they may not exist because the landscape is too young, particularly on Terrace 2. Late Archaic to Woodland age components are presumed to occur at or near the surface in the eastern portion of the site on Terrace 4, although no diagnostic artifacts were recovered to confirm their presence. Terrace 4 appears to contain the densest archaeological deposits, as several shovel tests had fire hearth features and bison bone beds.

A Middle Archaic component appears to be present on Terrace 3 at approximately 80 to 100 cmbs, based on the recovery of Duncan-Hanna Middle Archaic point. Northeastern Plains Village and Woodland components were identified in the middle and western portion of the site, extending to nearly one meter in depth on Terraces 1 and 2, and to a depth of up to 65 cm on Terrace 3. No diagnostic artifacts were recovered from the lake plain at the east end of the site.

A large quantity of fauna was recovered across the site area from find spots, shovel tests, and XUs. Particularly dense areas of fauna (bone beds) were identified in some XUs and shovel tests. Fauna was recovered from the surface to the maximum depth of testing and is horizontally and vertically ubiquitous in varying densities across the site, indicating that hunting and processing of game animals, mostly bison, was a primary site activity. Moderate to large amounts of bison bone were recovered from the Early and Middle Archaic and Northeastern Plains Village components, indicating that bison were a valuable subsistence resource through the Holocene.

A small to moderate quantity of lithic debris was recovered across the site area from find spots, shovel tests, and XUs. It was recovered from the surface to near the maximum depth of testing and is horizontally and vertically ubiquitous at the site, indicating that lithic reduction and stone tool manufacture and maintenance were important site activities. Lithic debris was most abundant in the Early and Middle Archaic component and was very sparse in the Northeastern Plains Village component.

Most ceramic were recovered from the western end of the site, with a small amount from the middle portion. A small amount of FCR was recovered from scattered locations across the site with no apparent concentrations.

9.19 Cultural Stratigraphy

The vertical distribution of artifacts ranges from 0 to 185 cmbs. Deep testing was not conducted at the site, except for shovel tests placed in the bottom of XUs because road work was initially planned to extend up to 60 cmbs, and was subsequently reduced to 30 cmbs. The presence of paleosols in the geomorphological cores to a depth of over three meters and artifacts recovered up to 185 cm in the XUs indicate that the archaeological deposits likely extend below the depth of shovel testing.

The vertical patterning of artifacts at the site is directly related to the alluvial history and terrace formation, which is discussed in detail in the geomorphological report in Appendix A. Four terraces are present at the site, as defined from the geomorphic cores, with each terrace containing a unique 155 stratigraphy that is distinct from the other terraces. The age of specific soil strata on Terraces 2 and 4 can be inferred from the radiocarbon dates and diagnostic artifacts recovered from well-controlled contexts in XUs and select shovel tests. The ages of Terrace 2 and 4 are considerably different, with Terrace 4 containing archaeological deposits that are several thousand years older. On Terrace 2, fauna from a Late Archaic or Early Woodland component at 140 to 160 cmbd dated to 2330 +/- 30 RCYBP, whereas on Terrace 4 fauna from an Early Archaic component at 135 to 160 cmbd dated to 7320 +/- 30 BP RCYBP and fauna from 160 to 185 cmbd dated to 7980 +/- 30 RCYBP.

A few shovel tests on Terrace 3 had diagnostic artifacts (ceramics and a projectile point), but no radiocarbon dates were obtained from this terrace. Diagnostic artifacts suggest that the age of the upper meter of Terrace 3 soils are 3000 to 4000 BP and younger, which fits with the general chronological trend established from Terraces 2 and 4 that have progressively younger soils from east to west along the point bar, as would be expected based on the time-transgressive point bar formation processes.

The cultural stratigraphy on Terrace 1 is not defined. Only a very small amount of fauna was recovered from 0 to 120 cmbs. No diagnostic artifacts or dates were obtained. It is assumed from the evidence in XUs 1 to 4 on Terrace 2 that the soils in the upper meter of Terrace 1 are relatively young (probably less than 1000 years old). No paleosols were observed in geomorphic Core 9 on Terrace 1, which shows that the soils consist of an Ap-A horizon over two weakly-developed cumulic AC horizons to a depth of 1.0 m that overlie a series of C horizons to 302 cmbs.

Cultural stratigraphy on Terrace 2 consists of a Northeastern Plains Village component that is primarily contained between 40 and 60 cmbd, with smaller amounts of artifacts displaced to a depth 120 cmbd from natural processes like rodent burrows. A Late Archaic or Early Woodland component is present from 140 to 160 cmbd. In XUs 1 and 2 radiocarbon dates from fauna were as follows: 30 to 40 cmbd dated 1170 +/- 30; 60 to 70 cmbd dated 1300 +/- 30 RCYBP; and 80 to 90 cmbd dated 1210 +/- 30 RCYBP. In XUs 3 and 4 radiocarbon dates from fauna were as follows: 50 to 60 cmbd dated 1200 +/- 30 RCYBP; 100 to 120 cmbd dated 1160 +/- 30 RCYBP; and 140 to 160 cmbd dated 2330 +/- 30 RCYBP. The similar ages of the fauna from 30 to 120 cmbd indicate that fauna from the Northeastern Plains Village component has been displaced by natural processes such as bioturbation and freeze-thaw. Geomorphic coring on Terrace 2 shows that the soils consists of a cumulic soil (A horizon) ranging from 0.63 to 1.0 m thick, overlying a moderately-developed paleosol.

Cultural stratigraphy on Terrace 3 is derived from diagnostic artifacts recovered from shovel tests, as no XUs were placed on this terrace and no radiocarbon dates were obtained. A stemmed Duncan- Hanna Middle Archaic point was recovered in Shovel Test 252 from 80 to 100 cmbs, suggesting that the soils at this depth date to approximately 4000 to 3000 BP. Three ceramic sherds were recovered from the following locations: Shovel Test 219 at 35 to 65 cmbd; Shovel Test 232 at 0 to 20; and Shovel Test 279 at 25 to 30 cmbs. The sherds are not diagnostic and could be from Woodland or Northeastern Plains Village components (2500 to 400 BP). These artifacts suggest that the age of the Terrace 3 soil from 80 to 100 cmbs is approximately 3000 to 4000 BP, with post-3000 BP components (Woodland or Northeastern Plains Village) above 80 cmbs. In contrast, the age of the Terrace 4 soil between 70 and 100 cmbs is approximately 7100 to 6800 RCYBP.

Cultural stratigraphy on Terrace 4 consists of multiple Early Archaic components identified between 40 and 185 cmbd. In XUs 5 and 6 a series of radiocarbon dates were obtained from fauna at various depths as follows: 40 to 50 cmbd dated 6210 +/- 30 RCYBP; 80 to 90 cmbd dated 6830 +/- 30 RCYBP; 110 to 135 cmbd dated 7260 +/- 30 RCYBP; 135 to 160 cmbd dated 7320 +/- 30 RCYBP; and 160 to 185 cmbd dated 7980 +/- 30 RCYBP. In XU 7 radiocarbon dates from fauna were as follows: 40 to 50 cmbd dated 6240 +/- 30 RCYBP and 70 to 80 cmbd dated 6880 +/- 30 RCYBP.

156

In XUs 9 and 10 radiocarbon dates from fauna were as follows: 70 to 80 cmbd dated 6850 +/- 30 RCYBP and 100 to 110 cmbd dated 7110 +/- 30 RCYBP. The dates from these XUs indicate at least five separate Early to Middle Archaic occupations at approximately 6200 RCYBP from 40 to 50 cmbd, 6800 RCYBP from 70 to 90 cmbd, 7100 RCYBP from 100 to 110 cmbd, 7300 RCYBP from 110 to 160 cmbd, and 8000 RCYBP from 160 to 185 cmbd.

The depths and dates of the faunal material from these XUs on Terrace 4 are vertically consistent and correspond well with each other, indicating that some of the components are extensive across the terrace and that the archaeological deposits have good integrity and discrete cultural stratigraphy. Diagnostic artifacts from these components include an Oxbow point recovered adjacent to XUs 7 and 8 in Shovel Test R68E5 from 75 to 100 cmbs and a small, Side-Notched Early Archaic Point recovered from 100 to 110 cmbd in XU 9. The only anomaly was a Besant Late Archaic projectile point from Shovel Test 148 that was recovered in association with fauna that dated to 6070 +/- 30 BP RCYBP.

Artifacts were recovered from a variety of soil horizons. In XUs 1 to 4 the primary archaeological deposits consist of a Northeastern Plains Village component at the base of a B horizon and top of a buried A horizon, with smaller amounts translocated from this deposit by rodent burrowing and natural processes. In XUs 3 and 4 a Late Archaic or Early Woodland component is represented by fauna from 140 to 160 cmbd in a buried Bk horizon.

In XUs 5 to 10 the primary archaeological deposits consist of a series of Early Archaic occupations with undefined overlying components. In XUs 5 and 6, the greatest artifact density occurs in a single zone between 70 and 90 cmbd, which coincides with a buried B horizon and an anthrosol that occurs in the middle portion of the buried B horizon. In XUs 7 and 8, the greatest artifact density occurs in a single zone between 70 and 80 cmbd, which coincides with the upper portion of a buried B horizon and a thin charcoal lens (possible anthrosol?) that was noted during excavation, which occurs in the lower half of the level. In XUs 9 and 10, the greatest artifact density occurs in a single zone between 100 and 110 cmbd, which coincides with a buried A horizon and anthrosol. Multiple other sparser components in XUs 5 to 10 occur throughout the soil profile in A, AB, AC, BC, and ACb/Ck1 horizons. It is evident that the archaeological components occur in a variety of soil horizon contexts.

9.20 Site Integrity

Aside from the near-surface impacts of plowing and previous road construction, the soils at the site are largely undisturbed, and the site has a high degree of integrity. Rodent burrowing was most common in the western portion of the site (perhaps because the soils are less clayey), where the cultural component in XUs 1 to 4 had diffuse vertical patterning. In contrast, there was tight vertical patterning of the multiple occupational deposits in XUs 5 to 10, and fewer rodent runs were observed. Overall, faunal preservation at the site is good. Historic air photos indicated that a portion of Terrace 4 may have some disturbance from drain tiling. However, disturbance from drain tiling was not observed in the XUs.

9.21 Radiocarbon Dates of Cultural Complexes

A total of 16 radiocarbon dates were obtained from faunal samples submitted to Beta for AMS dating to determine the age of the components at the site and aid in understanding the natural and cultural stratigraphy. Most of these dates have an internal stratigraphic consistency and correspond well with established dates for associated diagnostic artifacts. Also, there is good correlation between the bone collagen and carbonate dates. There is no reason to suspect that the dates are erroneous. However, some of the radiocarbon dates obtained from the site challenge or do not fit with the established dates of diagnostic artifacts, which were often recovered from the same provenience as the dated fauna. 157

Ceramics from XUs 1 to 4 and adjacent shovel tests on Terrace 2 most closely resemble Northeastern Plains Village ware, based on the temper, thickness, and smooth or smoothed-over cordmarked surface treatment, including polished sherds, which are not reported for Sandy Lake or Blackduck wares. Radiocarbon dates from fauna found in direct association with the ceramics range from 1160 +/- 30 to 1300 +/- 30 RCYBP, which is up to a few hundred years older than the established dates for Northeastern Plains Village and Psinomani (Sandy Lake) complexes that date about 1,000 to 400 BP.

A Besant Late Plains Archaic type was recovered in Shovel Test 148 from 70 to 90 cmbs. The point was recovered in-situ from the wall profile, so the provenience of its depth is well-controlled. Fauna in Shovel Test 148 at 60 to 100 cmbs had a radiocarbon date of 6070 +/- 30 RCYBP, which is far older than Besant and seems very unlikely to date the point. The radiocarbon date is expected to be accurate, as the fauna was recovered from the same paleosol as the point, and the date corresponds with other dates on Terrace 4 from XUs 5 to 10. It is not possible to reconcile the discrepancy between the point and dates on associated fauna. The most likely explanation is that the point is not in situ but was vertically displaced downward into an earlier component by rodent burrowing, other natural processes, or drain tiling that may have occurred in this area.

A Middle Archaic Oxbow projectile point was recovered from Shovel Test R68E5 from 75 to 100 cmbs. Radiocarbon dates obtained from fauna in XU 7, which was placed adjacent to the shovel test, yielded dates of 6240 +/- 30 RCYBP from 40 to 50 cmbd (35 to 45 cmbs) and 6880 +/- 30 RCYBP from 70 to 80 cmbd (65 to 75 cmbs). The dates are more than 1000 years older than the bulk of established Oxbow dates (Eighmy and LaBelle 1996 and Green 1998:276-280). However, four Oxbow dates are reported between 6750 to 5600 RCYBP by Green (1998:276-280). While the depth provenience of artifacts from shovel tests can’t precisely controlled because of the limitations inherent in shovel testing, it is likely that the dated faunal material is associated with the point for two reasons: first, most of the fauna from the shovel test was recovered from 75 to 100 cmbs, which corresponds with the depth of the primary deposit (mostly fauna) in XU 7 from 70 and 80 cmbd (65 to 75 cmbs) that was dated at 6880 +/- 30 RCYBP; and second, if the Oxbow point was displaced from a shallower depth in the shovel test, its date likely still exceeds the bulk of established Oxbow dates, as a date of 6240 +/- 30 RCYBP was obtained from XU 7 from 40 to 50 cmbd (35 to 45 cmbs).

A fragment of an small, Early Archaic Side-Notched Point was recovered in XU 9 from 100 to 110 cmbd, and a date of 7110 +/- 30 RCYBP was obtained from fauna in XU 9 from 100 to 110. The date corresponds with other Early Archaic dates with similar points in the region.

A series of other dates from Middle to Early Archaic components were obtained from XUs 5 to10 on Terrace 4, but no other diagnostic artifacts were recovered in association with these dates and so their specific cultural context is undetermined.

9.22 Conclusions

Site 21PL109 is a large, multicomponent precontact habitation with a moderately dense to dense artifact scatter on a point bar of the Red River. The site contains Early, Middle, and Late Archaic, undefined Woodland, and Northeastern Plains Village components. A possible Late Paleoindian component may be present as well, based on an isolated surface find. The most prominent components identified at the site, which are partially a result of bias caused by the location of XUs, are the Early to Middle Archaic and Northeastern Plains Village components. The ages and cultural contexts of site components were determined by a suite of 16 radiocarbon dates and several diagnostic artifacts. The radiocarbon dates at the site range from 1160 +/- 30 RCYBP to 7980 +/- 30 RCYBP.

158

The Phase I and II investigations included 389 positive shovel tests and ten 1-x-1 meter XUs. Artifacts were recovered from 0 to 185 cmbs but undoubtedly extend deeper. Artifact density was moderate to high across most of the site. Fauna was by far the most abundant artifact type recovered from the site, with notably smaller amounts of ceramics, FCR, lithic debris, stone tools, and cores. Site activities included hunting, animal processing (primarily bison), lithic reduction, stone tool manufacture/maintenance, and cooking/heating. Some of the faunal material was thermally-altered.

Terrace 2 Components A Northeastern Plains Village component, which occurs in the western portion of the site on Terrace 2 (the 2nd youngest and 2nd lowest terrace), yielded a cluster of five dates between 1300 +/- 30 BP and 1160 +/- 30 RCYBP. The depth of the assemblage, based on ceramic sherds and radiocarbon dates from fauna, extends from the surface to 90 cmbs. The broad vertical distribution of artifacts appears to be caused by rodent burrowing and other natural processes. In XUs, the primary artifact zone occurs from 40 to 70 cmbd. The artifact assemblage and inferred activities are derived from the Terrace 2 XU and shovel test data. Ceramics are relatively thin with sand and/or grit temper, and they have a smooth surface or smoothed-over cord-marked surface, including a small number of sherds with a smooth, polished surface.

A moderate to large quantity of fauna from kill and butchering activities is associated with Northeastern Plains Village component. Most of the remains are probably bison, based on their size and proximity to identifiable bison remains. A small quantity (<10%) of the remains are thermally- altered, and a few of the remains have cut marks from butchering. One worked, disk-shaped tool fragment made from a large mammal bone was recovered. Small amounts of shell, fish, beaver, small mammal, and dog/wolf/coyote remains were also recovered, although their cultural association is not certain. A very sparse amount of lithics were recovered, and they suggest lithic activities included early, middle, and late stages of reduction and tool maintenance. Knife River Flint and Swan River Chert were the primary lithic materials. A scraper, utilized flake, and Stage 2 biface recovered from 70 to 100 cmbs may be associated with the component. A small amount of FCR is present. The FCR and thermally-altered remains suggest that fire hearths are present. A Late Archaic or Early Woodland component is represented by a faunal date of 2330 +/- 30 RCYBP from 140 to 160 cmbd in a shovel test at the base of XU 3, beneath the Northeastern Plains Village component. Artifacts from this component include large mammal and mammal remains, some of which were thermally- altered.

Terrace 3 Components Terrace 3 contains an undefined Woodland or Northeastern Plains Village component, based on a few cord-marked and smooth sherds. A Duncan-Hanna point (Middle Archaic) was recovered from Shovel Test 252 at 80 to 100 cmbs. Faunal material is the primary artifact type, with much smaller amounts of lithics, ceramics, and FCR. No radiocarbon dates were obtained from this terrace. Testing on Terrace 3 only included shovel tests in 15-meter intervals, and site data is more limited than on Terraces 2 and 4 because close-interval testing was not conducted and XUs were not dug.

A moderate to large quantity of fauna from kill and butchering activities was recovered. Most of the remains are probably bison, based on their size and proximity to identifiable bison remains. A small quantity (<7%) of the remains are thermally-altered, and two bison bones have cut marks from butchering. A small amount of deer/elk, fish, beaver, small mammal, beaver, muskrat, and dog/wolf/coyote remains were also recovered, although their cultural association is not certain. Shovel Test 225 had a dense layer of bison bone (bone bed). Overall, the amount of fauna is considerably less than Terrace 4.

A very sparse amount of lithics was recovered, including eight pieces of lithic debris and a utilized flake. The debris contains a variety of diagnostic flake types, including bipolar, nonbifacial, and 159 decortication. Although the sample size is small, it is distinctively different than the bifacial debris that predominates on Terrace 4. The debris suggests lithic activities included the early and middle stages of reduction and nonbifacial technology. A wide variety of material are present, including Knife River Flint, Swan River Chert, quartzite, quartz, Red River Chert, Tongue River Silica, and unidentified material. A few ceramics were recovered from 0 to 60 cmbs, but they are not diagnostic of specific wares and are classified only as Woodland or Northeastern Plains Village ware. FCR were recovered from six shovel tests. The FCR and thermally-altered remains suggest that fire hearths are present. Shovel Test 216 had oxidized soil (probably from a fire hearth) in association with moderate amount of fired clay pieces.

Terrace 4 Components A series of Archaic occupations are present on Terrace 4 that are represented by a suite of radiocarbon dates and several projectile points, including Besant (Late Archaic), Oxbow (Middle Archaic), and a small Side-Notched Early Archaic Point. Nine radiocarbon dates from XUs 5 to 10 indicate at least five separate Middle to Early Archaic occupations at approximately 6200 RCYBP from 40 to 50 cmbd, 6800 RCYBP from 70 to 90 cmbd, 7100 RCYBP from 100 to 110 cmbd, 7300 RCYBP from 110 to 160 cmbd, and 8000 RCYBP from 160 to 185 cmbd. The dates and vertical distribution of artifacts suggest that the Middle and Early Archaic components are the most abundant. A fire heath feature was identified and dates to 6830 +/- 30 RCYBP, based on associated fauna. Other probable fire hearths were observed in several shovel test profiles (see Table 15), and fired clay was recovered from five shovel tests, representing locations where the soil was heated from a fire. Dense layers (bone beds) of bison and large mammal bones were present in some shovel tests (see Table 15), indicating there are multiple animal processing areas.

A moderate to large quantity of fauna from kill and butchering activities define the occupations. Most of the remains are probably bison, based on their proximity to identifiable bison remains. A small quantity (10-25%) of the remains is thermally-altered. Several of the remains have cut marks from butchering. A small amount of elk, deer, turtle, wolf, mollusk shell, fish, beaver, rabbit, small mammal, reptile (snake and toad), bird, and dog/wolf/coyote remains were also recovered, although their cultural association is not certain, except for the deer, fox, turtle, and bird remains which were thermally-altered. A deer awl was also recovered along with two bones used as tools, based on their polished condition. Dog and other canid remains are frequently found at Oxbow sites, where they are the second most common fauna after bison (Green 1998). The assemblage from these components corresponds to the pattern at other Early and Middle Archaic sites in the region, where Bison remains are the most common with notably smaller amounts of a wide range of aquatic and terrestrial animals. Several shovel tests had dense layers of bone (bone beds) that is mostly bison.

Small ochre pieces and flecking were observed in the soils and on some animal remains in the XUs. It appears that the presence of ochre was intentional and may represent a customary practice related to the killing and processing of bison. Similar ochre flecking in soil and on probable bison bones was noted at Late Plains Archaic site 21PL57 on a point bar of the Red River near Grand Forks (Florin 2001). The use of ochre in the Oxbow complex is well documented (Green 1998). The presence of ocher on the bones suggests that spirituality played an important role in bison hunting, and ochre was likely part of hunting and animal processing ritual.

Overall, there was a sparse amount of lithics recovered on Terrace 4, but some locations had a moderate to high density. Lithic debris from the XUs consisted mostly of biface shaping flakes (n=99) and biface thinning flakes (n=23), with much smaller amounts of decortication, nonbifacial, shatter, and bipolar flakes. The relative amounts of specific lithic debris types from shovel tests corresponds with the data from the XUs and consisted mostly of biface shaping flakes (n=49) and bifacial thinning flakes (n=19), with much smaller amounts of decortication, nonbifacial, shatter, and bipolar flakes. Lithic activities appeared to consist primarily of late-stage tool manufacture and 160 maintenance and also middle-stage biface manufacture, with minimal emphasis on early-stage reduction and nonbifacial or bipolar technologies.

Tools in the XUs included three Stage 4 bifaces, three utilized flakes, and a retouched flake. Tools in shovel tests included ten utilized flakes, six scrapers, and a Stage 4 biface, with most recovered below 40 cmbs, indicating they are associated with the Middle and Early Archaic components.

Lithic materials in the Middle and Early Archaic components XUs consisted mostly of Swan River Chert (77%), with much smaller amounts of Tongue River Silica, Knife River Flint, rhyolite, and other materials. In shovel tests, a variety of materials were recovered below 40 cm that are likely associated with Archaic component, including Bijou Hills Quartzite, fusilinid chert, Grand Meadow Chert, Prairie du Chien Chert (oolitic), Knife River Flint, quartzite, quartz, unidentified chert, Tongue River Silica, and Red River Chert. Non-local materials were procured from western and southern sources though long-distance trade networks or possibly travel to source areas.

FCR were recovered from 15 shovel tests. The FCR and thermally-altered remains suggest that fire hearths/ovens are present. Several shovel tests had discrete lenses of charcoal, oxidized (heated) soil, ash, calcined bone, and fired clay, which are indicative of cooking/heating features. Some tests also contained fauna encrusted with charcoal or fauna in direct association with distinct charcoal-rich lenses that may be anthrosols.

Integrity Aside from surficial disturbances caused by plowing and previous road construction in the ROW, soils at the site appear to be relatively undisturbed. Rodent runs are generally minimal but are moderate at east end of the site on Terrace 2. The overall vertical patterning of artifacts, especially in the eastern portion of the site indicates that intact archaeological deposits are present with very minimal artifact displacement. In fact, anthrosols and other cultural features containing the remains of distinct occupational events are preserved, particularly on Terrace 4. At the west end of the site where the vertical artifact patterning is more diffuse from rodent burrowing, the archaeological deposits still retain integrity and can provide important information because a single Plains Village component is present in this area to 90 cmbs. So, there is not mixing of multiple components.

9.23 Recommendation

The site is recommended eligible for listing on the NRHP under Criterion D because it has integrity and the portion of the site on Terraces 2, 3, and 4 is likely to yield important information on the Northeastern Plains Village, Woodland, and Archaic periods. The site contains data that could provide significant information on the following research themes, which are discussed in Section 2.3.2:

 Age of components and relation to regional chronology  Complexes present and relationship to other regional complexes  Climatic and environmental conditions and cultural changes during Holocene  Diagnostic artifacts and artifact assemblages  Lifeways, subsistence strategies, and settlement patterns  Site function  Lithic technology and raw material use  Regional interaction and trade  Site environment  Site formation processes and geomorphology

161

The final project design will stay within the existing ROW on the current CSAH 1 alignment and have only shallow subsurface disturbances not to exceed 30 cm on Terraces 2, 3, and 4, thus avoiding the deeper portions of the site that have the potential to answer important research questions (Criterion D). This plan will avoid impacts to the site’s significant archaeological deposits, and there will be no adverse effect, which would alter the site’s integrity or characteristics that qualify it for inclusion in the National Register.

Compaction at the site from construction activities will be very minimal because: 1) the clayey soils are not very compactable and 2) the site areas within the construction limits previously had road construction equipment on them and have been cultivated with heavy machinery for decades. Therefore, the proposed construction areas already have compacted surface and underlying soils. If the project design changes or if other projects will adversely affect the site, then a Phase III data recovery is recommended to mitigate the project’s effects.

162

Figure 9. Site 21PL109 Photo of Terraces 2 to 4 on South Side of CSAH 1 Taken from Near XUs 7 and 8, Facing West.

Figure 10. Site 21PL109 Photo of Terraces 2 to 4 on North Side of CSAH 1 Taken from Near XUs 7 and 8, Facing West.

165

XU 1 North Wall XU 2 0cmbd Datum Line

10 I I 20

30 II 40 II

50 III 60 III

70

80 IV 90 IV

100 Ground Surface Charcoal lense Rodent Run 110 I Very dark gray (2.5Y 3/1) clay; Ap horizon II Very dark grayish brown (2.5Y 3/2) mottled with dark grayish brown 120 (2.5Y 4/2) silty clay loam V III 130 Shovel Test Very dark gray (2.5Y 3/1) silty clay loam; 2Ab horizon IV Dark grayish brown (2.5Y 4/2) silty clay loam 140 VI V Very dark grayish brown (2.5Y 3/2) silty clay loam; 3Ab horizon VI Dark grayish brown (2.5Y 4/2) silty clay loam 150 VII Grayish brown (2.5Y 5/2) silty clay loam 160 VII

170 0 20 40 cm 180 Figure 11. Site 21PL109 XUs 1 and 2 North Wall Profile.

166

Figure 12. Site 21PL109 Photo XUs 1 and 2 North Wall Profile.

167

XU 3 North Wall XU 4 0cmbd Datum Line

10 I I

20 II 30 II 40

50

60 III III

70

80 IV IV 90

100 Ground Surface Charcoal lense 110 I Very dark gray (2.5Y 3/1) clay; Ap horizon V II Very dark grayish brown (2.5Y 3/2) mottled with dark grayish brown 120 (2.5Y 4/2) silty clay loam III Very dark gray (2.5Y 3/1) silty clay loam; 2Ab horizon Shovel Test 130 IV Dark grayish brown (2.5Y 4/2) silty clay loam VI V Very dark grayish brown (2.5Y 3/2) silty clay loam; 3Ab horizon 140 VI Dark grayish brown (2.5Y 4/2) silty clay loam 150 VII Grayish brown (2.5Y 5/2) silty clay loam

160 VII

170 0 20 40 cm 180 Figure 13. Site 21PL109 XUs 3 and 4 North Wall Profile.

168

Figure 14. Site 21PL109 Photo XUs 3 and 4 North Wall Profile.

169

XU 6 East XU 5 North XU 6 North Datum Line 0 cmbd 0cmbd

10 I 10 I I 20 20 II II Road construction cut

30 III III 30 III

40 40

IV IV 50 IV 50

60 60 PP#2-Bone V V 70 PP#1-Bone 70 V Anthrosol VI

80 AnthrosolIV VI V 80 IV Anthrosol VI VIIa V 90 VI VIIb 90 V Feature 1 VIII VIII 100 VIII 100 IX IX 110 110 Ground Surface Anthrosol Rodent Run IX Light olive brown (2.5Y 5/3) silt loam; unless otherwise noted 120 I Black (10YR 2/1) clay; Ap horizon X Very dark grayish brown (2.5Y 3/2) silty clay loam; Ab horizon II Very dark grayish brown (2.5Y 3/2) clay Olive brown (2.5Y 4/3) silty clay loam 130 IX XI III Very dark gray (2.5Y 3/1) clay; Ab horizon XII Very dark grayish brown (2.5Y 3/2) silty clay loam; Ab horizon IV Dark grayish brown (2.5Y 4/2) silty clay 140 Shovel Test XIII Olive brown (2.5Y 4/3) silty clay loam V Olive brown (2.5Y 4/3) clay X * buried A horizon indicated by bold text 150 VI Anthrosol; very dark grayish brown (10YR 3/2), dark grayish brown (10YR 4/2), dark XI yellowish brown (10YR 3/6), grayish brown (10YR 5/2), and light brownish gray (10YR 6/2) silty loam with charcoal, bone, calcined bone, charcoal-encrusted bone, fired 160 XII clay, and oxidized soil VIIa Feature 1; light gray (10YR 7/1 & 7/2) ash; fire hearth 170 VIIb Feature 1; dark grayish brown (10YR 3/6) silt loam, oxidized soil XIII 0 20 40 cm VIII Very dark grayish brown (2.5Y 3/2) silty clay loam with charcoal flecks and carbonate 180 masses; Ab horizon 185 Figure 15. Site 21PL109 XUs 5 and 6 North and XU 6 East Wall Profiles.

170

Figure 16. Site 21PL109 Photo XUs 5 and 6 North Wall Profile.

Figure 17. Site 21PL109 Photo XU 6 East Wall Profile. 171

ST R63E5

XU 5

Figure 18. Site 21PL109 Planview Photo of XU 5 Floor with Charcoal, Ash, and Oxidized Soil in Anthrosol at 82 to 86 cmbd above Feature 1.

172

Feature 1

GN

I

Bisection A B Line

I II

ST R63E5 Feature 1

II

XU 5 XU 6

A Feature 1 North Wall Profile B 90cmbd II I 100 cmbd

Rodent Run I Feature 1; dark yellowish brown (10YR 3/6) silt loam II Brown (10YR 5/3) silt loam

0 10 20 cm

Figure 19. Site 21PL109 Planview of Feature 1 at 90 cmbd in XUs 5 and 6.

173 ST R63E5

Figure 20. Site 21PL109 Photo of Planview of Feature 1 (Oxidized Soil) at 90 cmbd in XUs 5 and 6.

174

XU 7 North Wall XU 8 Datum Line 0cmbd I 10 I

II II 20 FCR

30 III rock III rock rock

40

IV 50 IV

60

70

V V 80

90

100 VI VI 110 VII VII Shovel Test Ground Surface 120 VII VII I Black (10YR 2/1) clay; Ap horizon 130 II Very dark grayish brown (2.5Y 3/2) clay III Very dark gray (2.5Y 3/1) clay 140 IV Dark grayish brown (2.5Y 4/2) silty clay V Olive brown (2.5Y 4/3) clay Soil Probe 150 VIII VI Very dark grayish brown (2.5Y 3/2) silty clay loam with charcoal flecks and carbonate masses; Ab horizon 160 VII Grayish brown (2.5Y 5/2) silt loam IX VIII Very dark grayish brown (2.5Y 3/2) silty clay loam; Ab horizon IX Olive brown (2.5Y 4/3) silty clay loam 0 20 40 cm * buried A horizon indicated by bold text Figure 21. Site 21PL109 XUs 7 and 8 North Wall Profile.

175

Figure 22. Site 21PL109 Photo XUs 7 and 8 North Wall Profile.

176

XU 9 West XU 9 North XU 10 North Datum Line 0 cmbd 0cmbd

10 10 I I I 20 20

30 30 II II II 40 40 Road construction cut 50 50 III III III

60 60

70 70 IV Anthrosol 1 PP#1-bone Anthrosol 1 IV 80 IV 80 V V 90 V 90 VI VI PP#2-bone & ocher VI 100 PP#5-lithic PP#6-bone PP#3-bone 100 Anthrosol 2 Anthrosol 2 110 VII Anthrosol 3 110 VII Anthrosol 3 VII VII VIII 120 120 VII Charcoal Lense 1 Charcoal Lense 1 Ground Surface Charcoal lense Anthrosol Rodent Run 130 Charcoal Lense 2 Anthrosol 1 Black (10YR2/1) silt loam with charcoal, bone, and oxidized soil I Black (10YR 2/1) clay; Ap horizon II Very dark grayish brown (2.5Y 3/2) clay Anthrosol 2 Black (10YR2/1), very dark gray (10YR 3/1), very dark grayish brown 140 (2.5Y 3/2), and dark grayish brown (2.5Y 4/2) silt loam with ash and III Dark grayish brown (2.5Y 4/2) silty clay VIII oxidized soil, charcoal encrusted bone, red ocher, lithics, organic 150 IV Olive brown (2.5Y 4/3) clay enriched baked soil V Very dark grayish brown (2.5Y 3/2) silty clay loam with charcoal flecks and carbonate masses; Anthrosol 3 Very dark gray (10YR 3/1) mottled with very dark grayish brown Ab horizon (10YR 3/2) silty clay loam with charcoal flecking, organic enriched 160 VI Olive brown (2.5Y 4/3) silt loam * buried A horizon indicated by bold text VII Very dark grayish brown (2.5Y 3/2) silty clay loam; Ab horizon IX 170 VIII Olive brown (2.5Y 4/3) silty clay loam X Shovel Test XI IX Very dark grayish brown (2.5Y 3/2) silty clay loam; Ab horizon 180 X Olive brown (2.5Y 4/3) silty clay loam XII 0 20 40 cm 190 XI Very dark grayish brown (2.5Y 3/2) silty clay; Ab horizon 195 XII Olive brown (2.5Y 4/3) silty clay Figure 23. Site 21PL109 XUs 9 and 10 North and XU 9 West Wall Profiles.

177

Figure 24. Site 21PL109 Photo XUs 9 and 10 North Wall Profile.

178

Figure 25. Site 21PL109 Photo XU 10 West Wall Profile (showing road construction cut).

179

Catalog # 98.7; ST R68E5 (75-100 cmbs); projectile point; Oxbow Middle Archaic, Swan River Chert

Catalog # 7.83; XU 5 (90-100 cmbd); projectile point tip fragment, Swan River Chert

Catalog # 41.298; XU 9 (100-110 cmbd); Early Archaic small side-notched base, Swan River Chert

Catalog # 54.64; XU 10 (100-110 cmbd); stage 4 biface or preform, Swan River Chert

0 Centimeters 5

Figure 26. Site 21PL109 Photos and Illustrations of Projectile Points.

180 Catalog # 263.2; ST 146 (0-25 cmbs); projectile point base or preform, Swan River Chert

Catalog # 544.1; ST 148 (70-90 cmbs); Besant Late Plains Archaic, Knife River Flint

Catalog # 543.1; ST 252 (80-100 cmbs); stemmed Duncan-Hanna Middle Archaic, Knife River Flint

Catalog # 43.1; ST R59 (0 cmbs); Plainview Late Paleoindian base, Prairie du Chien Chert (oolitic)

0 Centimeters 5

Figure 27. Site 21PL109 Photos and Illustrations of Projectile Points.

181 10. SUMMARY AND RECOMMENDATIONS

Site 21PL109 was identified and evaluated during archaeological investigation for the bridge replacement and CSAH 1 reconstruction. The site is a large multicomponent Archaic, Woodland, and Northeastern Plains Village period habitation along the Red River. Diagnostic artifacts include Northeastern Plains Village ceramic ware and a variety of projectile point types: Besant, Duncan- Hanna, Oxbow, Early Archaic small side-notched, and Plainview. The Plainview point was recovered from the ground surface and is out of context. The Northeastern Plains Village component had radiocarbon dates of 1300 +/- 30 to 1160 +/- 30 to RCYBP. Five separate Early to Middle Archaic occupations were identified in XUs, and they had radiocarbon dates of approximately 6200, 6800, 7100, 7300, and 8000 RCYBP. Animal kill and processing (mostly bison) appears to be the primary site activity. The site is recommended eligible for listing on the NRHP under Criterion D because it has integrity and is likely to yield important information on the precontact period.

The final project design will stay within the existing ROW on the current CSAH 1 alignment and have only shallow subsurface disturbances not to exceed 30 cm on Terraces 2, 3, and 4 thus avoiding the deeper portions of the site that have the potential to answer important research questions (criterion D). This plan will avoid impacts to the site’s significant archaeological deposits, and thus there will be no adverse effect, which would alter the site’s integrity or characteristics that qualify it for inclusion in the National Register. If the project design changes or if other projects will adversely affect the site, then a Phase III data recovery is recommended to mitigate the project’s effects. It is the opinion of FCRS that no historic properties eligible for or listed on the NRHP will be adversely affected by this project.

182

11. REFERENCES CITED

Agogino, G. and W. Frankforter 1960 A Paleo-Indian Bison Kill in Northwestern Iowa. American Antiquity 24:414-415.

Ahler, S., M. Root, and E. Feiler 1994 Methods for Stone Tool Analysis. In A Working Manual for Field and Laboratory Techniques and Methods for the 1992-1996 Lake Ilo Archaeological Project, edited by Stanley A. Ahler, pp. 27-121. Quaternary Studies Program, Northern Arizona University, Flagstaff, Arizona. Submitted to the U. S. Department of the Interior, Fish and Wildlife Service, Denver.

Ahler, S. and D. Toom (eds.) 1995 Archaeology of the Medicine Crow Site Complex (39BF2), Buffalo County, South Dakota. Illinois State Museum Reports of Investigations No. 51. Illinois State Museum, Springfield.

Alex, L. 2000 Iowa’s Archaeological Past. University of Iowa Press, Iowa City, Iowa.

Anderson, Dean 1979 Laurel. In A Handbook of Minnesota Prehistoric Ceramics, compiled and edited by Scott Anfinson, pp. 121–135. Occasional Publications in Minnesota Anthropology 5. Minnesota Archaeological Society, St. Paul.

Anderson, Duane and H. Semken, Jr. (editors) 1980 The Cherokee Excavations: Holocene Ecology and Human Adaptations in Northwestern Iowa. Academic Press, New York.

Anderson, D. and R. Shutler, Jr. 1978 The Cherokee Sewer Site (13CK405): A Summary and Assessment. Memoir 14: Bison Procurement and Utilization: A Symposium. Plains Anthropologist 23(82):132-139.

Andreas, A. T., 1874 Northern Minnesota, Minnesota State Atlas 1987. A. T. Andreas.

Andrefsky, W., Jr. (Editor) 2001 Lithic Debitage: Context, Form, and Meaning. University of Utah Press, Salt Lake City.

2005 Lithics: Macroscopic Approaches to Analysis (2nd Edition). Cambridge University Press, New York.

Anfinson, S. 1994 Thematic Context: Lithic Scatter. Draft. Copy on file at the State Historic Preservation Office, St. Paul, Minnesota.

1997 Southwestern Minnesota Archaeology: 12,000 Years in the Prairie Lake Region. Minnesota Historical Society Press, St. Paul.

Anfinson S. and H.E. Wright 1990 Climatic change and culture in prehistoric Minnesota. In “The Woodland Tradition in the Western Great Lakes: Papers Presented to Elden Johnson”, G. Gibbon (ed.). Publications in Anthropology 4. University of Minnesota: Minneapolis; pp. 213–232. 183

Arzigian, C. 2008 Minnesota Statewide Multiple Property Documentation Form for the Woodland Tradition. Mississippi Valley Archaeology Center at the University of Wisconsin-LaCrosse.

Bakken, K. 1985 Lithic Raw Materials in Northwest Minnesota. The Minnesota Archaeologist 44(1):34-46.

1988 A Middle Woodland Beach Ridge Site in Roseau County, Minnesota. The Minnesota Archaeologist 47(1):35-42.

1989 Lithic Reduction Stages and Assemblage Formation Processes. In Experiments in Lithic Technology, edited by Daniel S. Amick and Raymond P. Mauldin, pp. 15-31. BAR International Series 528, Oxford, England.

1994 The Lower Rice Lake Site, 21CE5. MA thesis, University of Minnesota

1997 Lithic Raw Material Resources in Minnesota. The Minnesota Archaeologist 56:51-83.

2000 Lithic Technologies and Raw Material Economies in Mille Lacs Area Archaeological Sites. In The Lake Onamia - Trunk Highway 169 Data Recovery Project, Mille Lacs County, Minnesota, by D. Mather and E. Abel. Loucks Project Report 96506-1. Loucks & Associates, Inc., Maple Grove.

2011 Lithic Raw Material Use Patterns in Minnesota. Unpublished Ph.D. Dissertation, Department of Anthropology, University of Minnesota

Barkley, T. 1977 Atlas of the Flora of the Great Plains. Iowa State University Press, Ames.

Baronsky, C. W., E. C. Grimm, and H. E. Wright 1987 Towards a Postglacial History of the Northern Great Plains: A Review of the Paleoecological Problems. Annals of Carnegie Museum 56:259-273.

Becker, G. 1983 Fishes of Wisconsin. University of Wisconsin Press, Madison.

Binford, L. 1981 Bones: Ancient Men and Modern Myths. Academic Press, Orlando.

Bingham, W. H., and Major R. I. Holcombe (editors) 1916 Compendium of History and Biography of Polk County, Minnesota. W. H. Bingham and Company, Minneapolis.

Birk, D. 1979b Sandy Lake Ware. In A Handbook of Minnesota Prehistoric Ceramics, edited by S. F. Anfinson, pp. 175-182. Occasional Publications in Minnesota Anthropology No. 5. Minnesota Historical Society, Fort Snelling.

Bluemle, J. P. 1972 Pleistocene Drainage Development in North Dakota. Geological Society of America, Bulletin 83:2189-2194.

184

1975 The Prairie: Land and Life. North Dakota Geological Survey.

1977 The Face of North Dakota: The Geologic Story. North Dakota Geological Survey.

Boucher, A. and L. Pettipas 1990 Early Occupations of the Glacial Lake Agassiz Basin In Manitoba; 11,500 To 7,700 B.P. In: Lasca, N. and J. Donahue (eds.) “Archaeological Geology of North America”: Geological Society of America Centennial Special Addition, Volume 4, pp 51-59.

Boszhardt, R. 2003 A Projectile Point Guide for the Upper Mississippi River Valley. University of Iowa Press, Iowa City.

Bower, J., Z. Jing, G. Rapp, Jr., Susan Mulholland, S. Valppu, Stephen Mulholland, and J. Huber 1996 Limited Phase II Investigations at the J Squared Site (21RW53), Redwood County, Minnesota. Archaeometry Laboratory Report No. 96-16. Report submitted to the Minnesota Department of Transportation. Archaeometry Laboratory, University of Minnesota, Duluth.

Bradbury, A. and P. Carr 1995 Flake Typologies and Alternative Approaches: An Experimental Assessment. Lithic Technology 20(2):100-116.

Bradford, M. 2013 Research Note: Radiocarbon Date for the Old Copper Component at Site 21AK11. The Minnesota Archaeologist 72:5-14.

Brink, J., M. Wright, B. Dawe, and D. Glaum 1986 Final Report of the 1984 Season at Head-Smashed-In Buffalo Jump. Alberta. Manuscript Series No.9, Archaeological Survey of Alberta, Edmonton.

BRW, Inc. 1994 Geoarcheaological Data Recovery, East Terrace Site (21BN6) and Gardner Site (21SN14), Benton and Stearns Counties, Minnesota. BRW, Inc., Minneapolis.

Buchner, A., P. Carmichael, G. Dickson, I. F. Dyck, B. J. Haug, T. Jones, D. Joyes, O. Mallory, M. Mallot, D. Meyer, D. Miller, R. Nash, L. Pettipas, C. Shay, E. Syms, M. Tisdale, and J. Whelan 1983 Introducing Manitoba Prehistory. Manitoba Department of Cultural Affairs and Historical Resources, Winnipeg.

Budak, M. 1991 The Function of Shell Temper in Pottery. The Minnesota Archaeologist 50(2):53–59.

Buhta, A., J. Hofman, E. Grimm, R. Mandel, and L. Hannus 2011 Investigating the Earliest Human Occupation of Minnesota: A Multidisciplinary Approach to Modeling Landform Suitability and Site Distribution Probability for the State’s Early Paleoindian Resources. Archeological Contract Series 248. Archeology Laboratory Augustana College, Sioux Falls, South Dakota

Callahan, E. 1979 The Basics of Biface Knapping in the Eastern Fluted Point Tradition: A Manual for Flintknappers and Lithic Analysts. Archaeology of Eastern North America 7:1-180.

185

Callan, F. 1937 Richland County Guide, History of Richland County. Richland County History files at the Leach Public Library, Wahpeton, North Dakota.

Clayton, L., and S. Moran 1982 Chronology of Late Wisconsinan Glaciation in Middle North America. Quaternary Science Reviews 1:55-82.

Cloutier, R. 2004 Testing Contemporaneity: The Avonlea and Besant Complexes on the Northern Plains. Unpublished Master’s thesis, University of Saskatoon, Saskatoon, Saskatchewan.

Cotterell, B., and J. Kamminga 1987 The Formation of Flakes. American Antiquity 52:675-708.

Crabtree, D. 1972 An Introduction to Flintworking. Occasional Papers of the Idaho State Museum, No. 28, Pocatello.

Davis, C., and J. Keyser 1999 McKean Complex Projectile Point Typology and Function in the Pine Parklands. Plains Anthropologist Vol. 44, No. 169, pp. 251-270.

Dawe, B. 1997 Tiny Arrowheads Toys in the Toolkit. Plains Anthropologist Vol. 42, No. 161, pp. 303-318.

Dawson, K. 1983 Prehistory of the Interior Forest of Northern Ontario. In Boreal Forest Adaptations: The Northern Algonkians, edited by A. T. Steegmann Jr., pp. 55–84. Plenum Press, New York.

Dobbs, C. 1979 Archaic Subsistence in Southwestern Minnesota: The View from Granite Falls. M.A. thesis, Department of Anthropology, University of Minnesota, Minneapolis.

1988 Outline of Historic Contexts for the Prehistoric Period (ca. 12,000 B.P. - A.D. 1700). Reports of Investigation Number 37. Institute for Minnesota Archaeology, Minneapolis.

Driver, H. and W. Massey 1957 Comparative Studies of North American Indians. Transactions of the American Philosophical Society, New Series 47(2):165-456.

Eighmy, J. and J. LaBelle 1996 Radiocarbon Dating of Twenty-Seven Plains Complexes and Phases. Plains Anthropologist 41(155):53-69.

Ellis, L. 1997 Hot Rock Technology. In: Black, Stephen L., Linda W. Ellis, Darrell G. Creel, Glenn T. Goode (Eds.), Hot Rock Cooking on the Greater Edwards Plateau: Four Burned Rock Midden Sites in West Central Texas, Studies in Archeology 22. Texas Archeological Research Laboratory, The University of Texas at Austin and Archeological Studies Program, Report 2, Texas Department of Transportation, Environmental Affairs Division, Austin, pp. 43–81. 186

Emerson, T. and D. McElrath 2009 The Eastern Woodland Archaic and the Tyranny of Theory. In Archaic Societies: Diversity and Complexity across the Midcontinent, ed. by T.E. Emerson, D.L. McElrath and A.C. Fortier, pp. 23–38. SUNY Press, Albany.

Flenniken, J. 1981 Replicative Systems Analysis: A Model Applied to the Vein Quartz Artifacts from the Hoko River Site. Washington State University Laboratory of Anthropology. Reports of Investigation No, 59, Pullman.

Flenniken, J., T. Ozbun, C. Fulkerson, and C. Winkler 1990 The Diamond Lil Deer Kill Site: A Data Recovery Project in the Western Oregon Cascade Mountains. Report on file at the State Office of Historic Preservation, Salem, Oregon.

Florin, F. 1996 Late Paleo-Indians in Minnesota and Vegetation Changes from 10,500 to 8,000 B.P. Unpublished Master’s thesis, University of Minnesota, Minneapolis.

2001 Phase I Cultural Resources Investigation and Phase II Testing at the City of East Grand Forks, Polk County, Minnesota. Florin Cultural Resource Services, Beldenville, Wisconsin. FCRS Report of Investigation #4. Submitted to the St. Paul District, U. S. Army Corps of Engineers, Contract No. DACW37-00-M-0122. Copies available at the Minnesota State Historic Preservation Office, St. Paul, Minnesota.

2001a Additional Phase II Testing at 21PL49 Near the City of East Grand Forks, Polk County, Minnesota. Florin Cultural Resource Services, Beldenville, Wisconsin. FCRS Report of Investigation #017. Report submitted to the St. Paul District, U.S. Army Corps of Engineers, Contract No. DACW37-00-M-0122 and Modification P00001. Copy available at the Minnesota State Historic Preservation Office, St. Paul, Minnesota.

2001b Letter Report for Phase I Cultural Resources Investigation of a Proposed Borrow Area Near the City of East Grand Forks, Polk County, Minnesota. Florin Cultural Resource Services, Beldenville, WI. Submitted to the U.S. Army Corps of Engineers, St. Paul District. Copy available at the Minnesota State Historic Preservation Office, St. Paul, Minnesota.

Florin, F., J. Ward, and M. McFaul 1998 Grand Forks/East Grand Forks General Reevaluation: Cultural Resources. Report Submitted to the St. Paul District, U.S. Army Corps of Engineers. Institute for Minnesota Archaeology Reports of Investigation Number 485.

Florin, F., B. Wergin, R. Schirmer and D. Mather 2001 Phase I Cultural Resources Investigation and Phase II Testing at the City of East Grand Forks, Polk County, Minnesota. Florin Cultural Resource Services Report of Investigations 4. Florin Cultural Resource Services, Beldenville, WI.

187

Florin, F., and B. Wergin 2002 Phase I Cultural Resources Investigation and Phase II Site Evaluations for Flood Protection Measures at East Grand Forks, Polk County, Minnesota. Florin Cultural Resource Services, Boyceville, Wisconsin. FCRS Reports of Investigation 028. Report submitted to the St. Paul District, U.S. Army Corps of Engineers, Contract No. DACW37- 01-M-0249 and Modifications P00001 and P00002. Copy available at the Minnesota State Historic Preservation Office, St. Paul, Minnesota.

Florin, F., J. Lindbeck, B. Wergin, and M. Kolb 2013 Phase I Archaeological Survey and Phase II Evaluation of Sites 21CR154, 21CR155, and 21CR156 for the TH101 / CSAH 61 Southwest Reconnection Project in Scott and Carver Counties, Minnesota. Florin Cultural Resource Services, LLC, Boyceville, WI. Reports of Investigation # 107. Prepared for the Minnesota Department of Transportation and Carver County.

Florin, F., K. Bakken, J. Lindbeck, and B. Wergin. 2015 Phase III Data Recovery at Site 21CR155 for the TH101 / CSAH 61 Southwest Reconnection Project in Carver County, Minnesota. Vols. I and II. FCRS Reports of Investigation 115.

Forbis,R. 1998 Besant Phase. In Archaeology of Prehistoric Native America: An Encyclopedia, edited by G. Gibbon, pp. 63–64. Garland Publishing, New York.

Fowler, M. 1959a Summary Report of Modoc Rock Shelter 1952, 1953, 1955, 1956. Report of Investigations 8. Illinois State Museum, Springfield.

1959b Modoc Rock Shelter: An Early Archaic Site in Southern Illinois. American Antiquity 24:257–270.

Frankforter, W.D. 1959a A pre-ceramic Site in Western Iowa. Journal of the Iowa Archeological Society 8(4):47-72.

1959b Excavation at the Simonsen Site (editor). Northwest Chapter of the Iowa Archeological Society, Newsletter 7(4):4.

Frankforter, W. and G. Agogino 1959a Archaic and Paleo-Indian Archaeological Discoveries in Western Iowa. Texas Journal of Science 1(4):482-491.

1959b Recent Pre-Ceramic Archaeological Developments in Western Iowa. Journal of the Iowa Archaeological Society 9:13-19.

1960 The Simonsen Site: Report for the Summer of 1959. Plains Anthropologist 5(10):65-70.

Frison, G. C. 1991 Prehistoric Hunters of the High Plains. Academic Press, New York.

1998 The Northwestern and Northern Plains Archaic. In Archaeology on the Great Plains, ed. by W.R. Wood, pp. 140-172. University of Kansas Press, Lawrence.

188

Frison, G., and D. Stanford (editors) 1982 The Agate Basin Site: A Record of the Paleoindian Occupation of the Northwestern High Plains. Academic Press, New York.

Frison, G., M., and D. Wilson 1976 Fossil Bison and Artifacts from an Early Altithermal Period Arroyo Trap in Wyoming. American Antiquity 41(1):28-57.

Fryxell, R. 1965 Mazama and Glacier Peak Ash Layers: Relative Ages. Science 147(3663):1288–1290.

Gibbon, G. 2012 Archaeology of Minnesota, the Prehistory of the Upper Mississippi River Region. University of Minnesota Press, Minneapolis.

Gibbon, G. and S. Anfinson 2008 Minnesota Archaeology: The First 13,000 Years. Publications in Anthropology No. 6, University of Minnesota, Minneapolis.

Gilbertson, J. P. 1990 Quaternary Geology Along the Eastern Flank of the Coteau Des Prairies, Grant County, South Dakota. Unpublished M.S. thesis, Department of Geology, University of Minnesota, Minneapolis.

Gilman, R., C. Gilman, and D. Stultz 1979 The Red River Trails: Oxcart Routes Between St. Paul and the Selkirk Settlement 1820-1870. Minnesota Historical Society, St. Paul.

Glasrud, C. A. (editor) 1982 Roy Johnson’s Red River Valley. Red River Valley Historical Society, Moorhead, Minnesota.

Green, D'Arcy C. 1998 A Re-Evaluation of the Oxbow Dam Site (DhMn-1): Middle Holocene Cultural Continuity on the Northern Plains. M.A. thesis, University of Saskatchewan, Saskatoon.

Gregg, M., and D. Davidson (editors) 1985 An Overview of the Prehistory of Western and Central North Dakota. Cultural Resources Series No. 1, U. S. Department of the Interior, Bureau of Land Management, Montana State Office, Billings, Montana.

Goldstein, L. and S. Osborn 1988 A Guide to Common Prehistoric Projectile Points in Wisconsin. Milwaukee Public Museum.

Gonsior, L. 1992 Lithic Materials of Southeastern Minnesota. The Platform 4 (1-4). Minnesota Knappers Guild, Duluth, Minnesota.

Hajic, E. 1995 Evaluation of Site Limits and Geological Context of the Jackpot Junction Site (21 RW 52), Redwood Co., Minnesota. Report prepared for the Minnesota Department of Transportation. Foth & Van Dyke, Inc., Eden Prairie. 189

Hallet, Douglas J., L.V. Hills and John J. Clague 1997 New accelerator mass spectrometry radiocarbon ages for the Mazama tephra layer from Kootenay National Park, British Columbia, Canada. Canadian Journal of Earth Sciences 34(9):1202-1209.

Harrison, C., E. Redepenning, C. Hill, G. Rapp, Jr., S. Aschenbrenner, J. Huber, S. Mulholland 1995 The Paleo-Indian of Southern St. Louis Co., Minnesota: The Reservoir Lakes Complex. University of Minnesota, Duluth.

Haury, C. 2005 Analysis of Fauna from the Rustad Site. In Archaeology and Paleoenvironment at the Rustad Site (32RI775), edited by Michael G. Michlovic and Garry L. Running IV. Plains Anthropologist Memoir 37:91-133.

Hayden, B., and W. Hutchings 1989 Whither the Billet Flake? In Experiments in Lithic Technology, edited by Daniel S. Amick and Raymond P. Mauldin, pp. 235-257. BAR International Series 528, Oxford, England.

Herrick, C. 1892 The Mammals of Minnesota. Geological and Natural History Survey of Minnesota, Bulletin 7:1-299. Johnson, Smith, and Harrison, State Printers, Minneapolis.

Hess, D. 1989 Minnesota Red River Trails. National Register of Historic Places Multiple Property Documentation Form, Minnesota Historical Society, St. Paul.

Higginbottom, D. 1996 An Inventory of Fluted Projectile Points from Minnesota. Paper distributed at the 54th Annual Plains Conference, Iowa City, Iowa.

Hixson, W. W., and Company 1916 Plat Book of the State of Minnesota. W. W. Hixson & Co., Rockford, Illinois.

1930 Plat Book of the State of Minnesota. W. W. Hixson & Co., Rockford, Illinois.

Hobbs, H. and J. Goebel 1982 Geologic Map of Minnesota: Quaternary Geology . State Map Series S-1, Minnesota Geological Survey, St. Paul.

Hohman-Caine, C 1983 Normative Typological and Systemic Stylistic Approaches to the Analysis of North Central Minnesota Ceramics. Ph.D. Dissertation, University of Minnesota, Minneapolis.

Hohman-Caine, C. and G. Goltz 1995 Brainerd Ware and the Early Woodland Dilemma. The Minnesota Archaeologist 54: 109-129.

Homsey, L. 2009 The Identification and Prehistoric Selection Criteria of FCR: An Example from Dust Cave, Alabama. Southeastern Archaeology Vol. 28(1). Gainesville, FL.

190

House, J. and J. Smith 1975 Experiments in the Replication of FCR. In The Cache River Archaeological Project: An Experiment in Contract Archaeology. The Arkansas Archaeological Survey, Fayetteville, AR. http://herb.umd.umich.edu, accessed November 2008.

Hudak, J. 1978 A Description of the Early Middle Woodland Ceramics from the Pedersen Site in Southwestern Minnesota. In: “Some Studies of Minnesota Prehistoric Ceramics: Papers Presented at the First Council for Minnesota Archeology Symposium – 1976”, pp. 27-34, A. Woolworth and M. Hall (eds.) Occasional Publications in Minnesota Anthropology, No. 2. Minnesota Historical Society, St. Paul.

1981 Cultural Resource Investigation of the Grand Forks/East Grand Forks Urban Study and the East Grand Forks Flood Control Project. Archaeological Field Services, Inc., Stillwater, Minnesota. Submitted to the St. Paul District, U.S. Army Corps of Engineers, Contract No. DACW37-80-D-0045. Copies available at the Minnesota State Historic Preservation Office, St. Paul, Minnesota.

Hudak, J., E. Hobbs, A. Brooks, C. Sersland, and C. Phillips (editors) 2002 Mn/Model: A Predictive Model of Precontact Archaeological Site Location for the State of Minnesota Final Report 2002. CD version. Minnesota Department of Transportation.

Husted, W. and R. Edgar 2002 The Archaeology of Mummy Cave, Wyoming: An Introduction to Shoshonean Prehistory. Midwest Archaeological Center Special Report No. 4, and Southeast Archaeological Center Technical Reports Series No. 9. United States Department of the Interior, Midwest Archaeological Center, Lincoln.

Inizan, M., M. Reduron-Ballinger, H. Roche, and J. Tixier 1999 Technology and Terminology of Knapped Stone: Followed by a Multilingual Vocabulary Arabic, English, French, German, Greek, Italian, Portuguese, Spanish. Préhistoire De La Pierre Taillée 5. Cercle de Recherches et d’Etudes Préhistoriques, Nanterre, France .

Jackson, H. 1961 Mammals of Wisconsin. University of Wisconsin Press, Madison.

Jackson, M. 1997 Feature and Fire-Cracked Rock Analyses: A Search for Site Function Data from Site in the Tobacco Plains, Middle Kootenai River Valley, Northwest Montana. Manuscript on file at the Center for Environmental Archaeology, Texas A&M University, College Station.

1998 Nature and Fire-Cracked Rock: New Insights from Ethnoarchaeological and Laboratory Experiments. Unpublished Master’s Thesis, Department of Anthropology, Texas A&M University, College Station.

Jenks, A. E. 1932 The Problem of the Culture from the Arvilla Pit. American Anthropologist 34:455-466.

Johnson, C., M. Michlovic, S. Buck, and R. St. George 1995 Phase II Archaeological Investigations at the Mooney Site (21NR29), Norman County, Minnesota. BRW, Inc., Minneapolis. 191

Johnson, E. 1964 Copper Artifacts and Glacial Lake Agassiz Beaches. Minnesota Archaeologist 26(1):5-22.

1973 The Arvilla Complex. Minnesota Prehistoric Archaeology Series No. 9, Minnesota Historical Society, St. Paul.

Jones, J. and E. Birney 1988 Handbook of Mammals of the North-Central States. University of Minnesota Press, Minneapolis.

Justice, N. 1987 Stone Age Spear and Arrow Points of the Midcontinental and Eastern United States. Indiana University Press, Bloomington.

Justin, M. and L. Peterson 1991 Final Archaeological Reconnaissance Report and Research Design for Site Evaluation Testing: Proposed Minnesota Department of Transportation S.P. 6404-26, 6506-26 and 6508-29, Trunk Highways 19 & 71, Reconstruction and Relocation from 1.8 Mile East of Redwood Falls to 0.7 Mile North of E. Jct. T.H. 19 & 71, Redwood and Renville Counties. Report submitted to the Minnesota Department of Transportation, St. Paul. Minnesota Trunk Highway Archaeological Reconnaissance Study, Minnesota Historical Society, St. Paul.

Justin, M. and W. Radford 1990 Archaeological Survey and Site Assessment Report, Proposed Minnesota Department of Transportation S.P. 5202-37 for Upgrading of T.H. 14 from Jct. of T.H. 15 at New Ulm to Jct. of T.H. 99 at Nicollet, Nicollet County. Minnesota Trunk Highway Archaeological Reconnaissance Survey, Archaeology Department, Minnesota Historical Society, St. Paul.

Kaufman, K. 1996 Lives of North American Birds. Houghton Mifflin, Boston.

Kay, M. 1998 The Central and Southern Plains Archaic. In Archaeology on the Great Plains, ed. by W.R. Wood, pp. 173-200. University of Kansas, Lawrence.

Keeley, L. 1980 Experimental Determination of Stone Tool Uses: A Microwear Analysis. University of Chicago Press.

Kehoe, T. 1966 The Small Side-Notched Point System of the Northern Plains. American Antiquity 57:827- 839.

1973 The Gull Lake Site: A Prehistoric Bison Drive in Southwestern Saskatchewan. Milwaukee Public Museum, Publications in Anthropology and History No. 1.

1974 The Large Corner-Notched Point System of the Northern Plains and Adjacent Woodlands. In Aspects of Upper Great Lakes Anthropology: Papers in Honor of Lloyd A. Wilford, edited by Elden Johnson, pp. 103-115. Minnesota Prehistoric Archaeology Series No. 11. Minnesota Historical Society, St. Paul.

192

Kelsey, V. 1951 Red River Runs North! Harper, New York.

Ketz, A. K., and T. Dolence 1997 Phase I Cultural Resources Investigation of the Additional Levee and Floodwall Alignments, a Road Raise, Two Pump Stations, and a Diversion Ditch for English Coulee Proposed for the City of Grand Forks, Grand Forks County, North Dakota. 106 Group Ltd., St. Paul, Minnesota. Submitted to the St. Paul District, U.S. Army Corps of Engineers, St. Paul, Minnesota.

Kivett, Marvin F. 1962 Logan Creek Complex. Mimeographed report, presented at the 20th Plains Anthropology Conference.

Kolb, M. 2006 Geomorphological Investigations in Conjunction with Phase I Archaeological Survey of Five Proposed T.H. 41 Bridge Crossings of the Minnesota River in Scott and Carver Counties, Minnesota. Strata Morph Geoexploration Report of Investigation 133. Strata Morph Geoexploration, Inc., Sun Prairie, WI.

Kooyman, B. 2000 Understanding Stone Tools and Archaeological Sites. University of Calgary Press, Alberta.

Kuehn, S. 2000 The Granite Falls Site (21YM47), A Prairie Archaic Bison Kill in Southwestern Minnesota. Paper presented at the joint 2000 Plains Anthropological and Midwest Archaeological Conference, St. Paul.

2010 Late Paleoindian Strategies in the Western Great Lakes Region. In Foragers of the Terminal Pleistocene in North America, edited by R. B. Walker and B. N. Driskell. University of Nebraska Press.

Larson, T. and D. Penny 1990 The Smilden-Rostberg Site: An Early Archaic Component in the Northeastern Plains. Larson- Tibesar Associates, Inc., Laramie, Wyoming. Submitted to KBM, Inc., Grand Forks, North Dakota.

2000a Results of a Phase I Cultural Resource Investigation for Proposed Levee and Floodwall Alignments at the City of East Grand Forks. Larson-Tibesar Associates, Laramie, Wyoming. Submitted to the St. Paul District, U.S. Army Corps of Engineers, Contract No. DACW37-99-M-0094. Copies available at the Minnesota State Historic Preservation Office, St. Paul, Minnesota.

2000b Results of a Phase I Cultural Resource Investigation for Proposed Levee and Floodwall Alignments at the City of Grand Forks. Larson-Tibesar Associates, Laramie, Wyoming. Submitted to the St. Paul District, U.S. Army Corps of Engineers, Contract No. DACW37-99-M-0095. Copies available at the North Dakota State Historic Preservation Office, Bismarck, North Dakota.

193

Larson, T., D. Penny, and R. Hilman 1994 A Literature Review of the Cultural Resources Within the Red River Basin of North Dakota. Larson-Tibesar Associates, Laramie, Wyoming. Submitted to St. Paul District, U.S. Army Corps of Engineers, St. Paul, Minnesota.

Latas, T. 1992 An Analysis of Fire-Cracked Rock: A Sedimentological Approach. In Deciphering a Shell Midden, edited by J. K. Stein, pp. 211-237. Academic Press, San Diego.

Lenius, B. and D. Olinyk 1990 The Rainy River Composite: Revisions to Late Woodland Taxonomy. In The Woodland Tradition in the Western Great Lakes: Papers Presented to Elden Johnson, edited by G. Gibbon, pp. 77–112. Publications in Anthropology No. 4. University of Minnesota, Minneapolis.

Lewis, S. and P. Heikes 1990 A Preliminary Report on a Paleo-Indian Bison Kill Site (21YM47) Near Granite Falls, Minnesota. IMA Quarterly Newsletter (Institute for Minnesota Archaeology) 5(3):4-5.

Lovick, S. 1983 Fire-Cracked Rock as Tools: Wear-Pattern Analysis. Plains Anthropologist 28(99):41-52.

Lugenbeal, E. 1978b The Laurel Ceramics of the Smith Site (21KC3) and their Implications for Culture History in Northern Minnesota. In Some Studies of Minnesota Prehistoric Ceramics, edited by A. R. Woolworth and M. A. Hall, pp. 35-46. Minnesota Archaeological Society, St. Paul.

1979 Blackduck Ware. In A Handbook of Minnesota Prehistoric Ceramics, edited by S. F. Anfinson, pp. 23-37. Occasional Publications in Minnesota Anthropology No. 5. Minnesota Historical Society, Fort Snelling.

Lyon, M., H. Ekstrom, A. Gronhovd. K. Bakken, S. Arnott, and J. Anderson 2003a Phase I Cultural Resources Investigation of a Tieback Levee Alignment, Two Borrow Areas, and the Valley Golf Course Relocation Area, and Phase II Evaluation of Site 21PL17 at the City of East Grand Forks, Polk County, Minnesota. Summit Envirosolutions, Inc., St. Paul, Minnesota. Submitted to St. Paul District, U.S. Army Corps of Engineers, St. Paul, Minnesota.

Lyon, M., K. Bakken, D. Mather, and J. Anderson 2003b Additional Phase I Cultural Resources Investigations at the Erickson and Hangsleben Borrow Areas, East Grand Forks, Polk County, Minnesota – Supplemental Report. Summit Envirosolutions, Inc., St. Paul, Minnesota. Submitted to St. Paul District, U.S. Army Corps of Engineers, St. Paul, Minnesota.

Mack, R., R. Okazaki and S. Valastro 1979 Bracketing dates for two ash falls from Mount Mazama. Nature 279:228-229.

Madigan, T., R. C. Schirmer, and A. Miller 2002 Testing for Deeply-Buried Archaeological Resources at a Proposed Borrow Area in the City of Grand Forks, Grand Forks County, North Dakota. Hemisphere Field Services, Minneapolis, Minnesota. Report of Investigations No. 664. Copies available at the North Dakota State Historic Preservation Office, Bismarck, North Dakota. 194

Magne, M. 1985 Lithics and Livelihood: Stone Tool Technologies of Central and Southern Interior British Columbia. National Museum of Man, Mercury Series. Archaeological Survey of Canada, Paper No. 133, Ottawa. 1989 Lithic Reduction Stages and Assemblage Formation Processes. In Experiments in Lithic Technology, edited by Daniel S. Amick and Raymond P. Mauldin, pp. 15-31. BAR International Series 528, Oxford, England.

Malik, R. and K. Bakken 1993 Archaeological Data Recovery at the Bradbury Brook Site, 21 ML 42, Mille Lacs County, Minnesota. Bradbury Brook Data Recovery Project, Archaeology Department, Minnesota Historical Society, St. Paul.

1999 The Bradbury Brook Site, 21ML42. The Minnesota Archaeologist 58:134-171.

Mason, R. 1981 Great Lakes Archaeology. Academic Press, New York.

1997 The Paleo-Indian Tradition. The Wisconsin Archeologist 78(½):78-111.

Marschner, F. 1974 Map of the Original Vegetation of Minnesota. U.S. Forest Service, North Central Experiment Station, St. Paul.

Mather, D. and P. Nunnally 1994 Draft Cultural Resources Reconnaissance Survey and Property Evaluation Report, Vol. 1: Technical Report. Properties in the Vicinity of MnDOT S.P. 6404-26, 6506-26, and 6508-29 for Reconstruction Replacement of Four Bridges on T.H. 19 and 71 between Morton and Redwood Falls, Redwood and Renville Counties. Minnesota Trunk Highway Archaeological Reconnaissance Survey, Archaeology Department, Minnesota Historical Society, St. Paul.

Mather, D., T. Olmanson, K. Gragg-Johnson, and L. Schuster 1998 The Washington Creek Site (21ME14) and the Archaeology of the Prairie-Forest Border. The Minnesota Archaeologist 57:99-133.

McDowell-Loudan, E. 1983 FCR: Preliminary Experiments to Determine Its Nature and Significance in Archaeological Contexts. Chesopiean Vol. 21, pp.20-29.

McElrath, D., A. Fortier, and T. Emerson 2009 An Introduction to the Archaic Societies of the Midcontinent. In Archaic Societies: Diversity and Complexity across the Midcontinent, ed. by T.E. Emerson, D.L. McElrath and A.C. Fortier, pp. 3–21. SUNY Press, Albany.

McParland, P. 1977 Experiments in the Firing and Breaking of Rocks. The Calgary Archaeologist, Vol. 5. The University of Calgary, Alberta, Canada.

Meyer, G., L. Swanson, T. Wahl, A. Knaeble 1995 Geologic Atlas of Stearns County, Minnesota, C-10, Part A. Plate 3, Surficial Geology. Minnesota Geological Survey, University of Minnesota, St. Paul. Digital copy, ftp://mgssun6.mngs.umn.edu/map_catalog/pdf/umn22297.pdf, accessed 18 June 2014. 195

Michlovic, M. ca. 1978 An Archaeological Survey of Clay County, Minnesota. On file at Minnesota Historical Society, St. Paul. 1982 Report on the Red River Archaeological Survey in Norman County, Minnesota. Minnesota Archaeologist 41(2):53-69.

1983 The Red River Valley in the Prehistory of the Northern Plains. Plains Anthropologist 28(99):23-31.

1984 Archaeological Studies Along the Upper Red River, Minnesota. Report submitted to the Minnesota Historical Society, St. Paul. Moorhead State University, Moorhead, Minnesota.

1986 The Archaeology of the Canning Site. Minnesota Archaeologist 45(1):3-36.

1987a The Archaeology of the Mooney Site (21NR29). Minnesota Archaeologist 46(2):39-66.

1987b Archaeological Survey and Test Excavations in Cass County, N.D. Report submitted to the State Historical Society of North Dakota, Bismarck. Moorhead State University, Moorhead, Minnesota.

1990 Northern Plains - Woodland Interaction in Prehistory. In The Woodland Tradition in the Western Great Lakes: Papers presented to Elden Johnson, edited by G. E. Gibbon, pp. 45-54. Publications in Anthropology Number 4, University of Minnesota Press, Minneapolis.

Michlovic, M. and G. Running IV (editors) 2005 Memoir 37: Archaeology and Paleoenvironment at the Rustad Site (21RI775). Plains Anthropologist 50(196).

Michlovic, M. and F. Swenson 1998 Northeastern Plains Village Pottery. North Dakota History 65 (2-3):11-25.

Mitchell M. and C. Johnston 2012 Technological Analysis of the Modified Stone Assemblage. In Agate Basin Archaeology at Beacon Island, North Dakota, edited by Mark D. Mitchell. Submitted to the State Historical Society of North Dakota and the U. S. Department of the Interior, National Park Service. Paleocultural Research Group, Research Contribution No. 86, Arvada, Colorado.

Monaghan, G., K. Egan-Bruhy, M. Hambacher, D. Hayes, M. Kolb, S. Kuehn, S. Peterson, J. Robertson, and N. Shaffer 2006 Minnesota Deep Testing Protocol Project. Prepared for the Minnesota Department of Transportation and the Federal Highway Administration. Commonwealth Cultural Resources Group, Jackson, Michigan.

Morrow, T. 1984 Iowa Projectile Points. University of Iowa, Iowa City.

1994 A Key to the Identification of Chipped-Stone Raw Materials Found on Archaeology Sites in Iowa. Journal of the Iowa Archaeology Society 41:108-129.

196

Morrow, T. and J. Behm 1986 Descriptions of Common Lithic Raw Materials Encountered on Wisconsin Archaeological Sites. Paper presented at the Fall Meeting of the Wisconsin Archaeological Survey, Madison.

Morrow, T., S. Anfinson, K. Bakken, G. Gibbon, M. Giller, J. Hahn, D. Higginbottom, and C. Johnson 2016 Stone Tools of Minnesota. Wapsi Valley Archaeology, Inc. Anamosa, Iowa.

National Park Service 1991 How to Apply the National Register Criteria for Evaluation. National Register Bulletin 15. National Register Branch, Interagency Resources Division, National Park Service. U.S. Government Printing Office, Washington, D.C.

Nienow, J., M. Baltus, M. McFaul, and C. Kirchoff 2003 Phase I Cultural Resource Investigations of Two Proposed Borrow Areas and A Ponding Area at East Grand Forks, Minnesota. Schoell and Madson, Minneapolis, Minnesota. Submitted to St. Paul District, U.S. Army Corps of Engineers, St. Paul, Minnesota.

North West Publishing Company 1902 Polk County, Minnesota, 1902. North West Publishing Co., Minneapolis, Minnesota.

Odell, G. 1989 Experiments in Lithic Reduction. In Experiments in Lithic Technology, edited by Daniel S. Amick and Raymond P. Mauldin, pp. 163-198. BAR International Series 528, Oxford, England.

1996 Stone Tools and Mobility in the Illinois Valley. International Monographs in Prehistory, Archaeological Series 10, Ann Arbor, Michigan.

2003 Lithic Analysis. Springer Science+Business Media, Inc., New York.

Ojakangas, R. W., and C. L. Matsch 1982 Minnesota’s Geology. University of Minnesota Press, Minneapolis.

Overstreet, D. 1993 Chesrow: A Paleoindian Complex in the Southern Lake Michigan Basin. Case Studies in Great Lake Archaeology Number 2. Great Lakes Archaeological Press, Milwaukee.

1996 A Tusk Tip from Hebior Mammoth (47 Kn 265), Kenosha County, Wisconsin. The Wisconsin Archeologist 77(1-2):87-93.

Ozbun, T. 1987 Technological Analysis of the Lithic Assemblage from the Buttonhole Rockshelter/Quarry Site, Northeastern New Mexico. Master’s Thesis, Washington State University, Pullman.

Perino, G. 1971 Guide to the Identification of certain American Indian Projectile Points; Special Bulletin No. 3-4 of the Oklahoma Anthropological Society.

197

Perkl, B. 1998 Cucurbita pepo from King Coulee, Southeastern Minnesota. American Antiquity 63(3):279- 288.

Peterson, C. 1996 Three Lithic Artifacts from Bison Kill Site 21YM47 near Granite Falls. In Discovering Our Hidden Past, edited by T. Early. Information Booklet distributed during Minnesota Archaeology Week, 1996. University of Minnesota Printing Services, Minneapolis.

Peterson, L. 1973 An Early Prehistoric Stone Workshop in Northwest Minnesota. The Minnesota Archaeologist 32(3-4):1-57.

Pettipas, L. and A. Buchner 1983 Paleoindian Prehistory of the Glacial Lake Agassiz Region In Southern Manitoba, 11,500 To 6500 B.P. In: Teller, J. and L. Clayton “Glacial Lake Agassiz”, Geological Society of Canada Special Paper 26, pp. 421-451.

Phillips, C., R. Brandon, and E. Moll 1999 Field Guide to Amphibians and Reptiles of Illinois. Maunal 8. Illinois Natural History Survey, Champaign.

Picha, P. and M. Gregg 1991a Northern Red River Study Unit. The North Dakota Comprehensive Plan for Historic Preservation: Archeological Component. State Historical Society of North Dakota, Bismarck.

1991b Southern Red River Study Unit. The North Dakota Comprehensive Plan for Historic Preservation: Archeological Component. State Historical Society of North Dakota, Bismarck.

Picha, P., M. Gregg, A. Bleier, and T. Reed 2016 The Northern Red River Study Unit, in Historic Preservation in North Dakota: A Statewide Comprehensive Plan, State Historical Society of North Dakota, Bismarck.

Pielou, E. 1991 After the Ice Age: The Return of Life to Glaciated North America. The University of Chicago Press, Chicago.

Pleger. T. and J. Stoltman 2009 The Archaic Tradition in Wisconsin. In Archaic Societies: Diversity and Complexity Across the Midcontinent, edited by T. E. Emerson, D. L McElrath, and A. C. Fortier, pp. 697-723. State University of New York Press, Albany.

Prest, V., J. Donaldson, and H. Mooers 2000 The Omar Story: The Role of Omars in Assessing Glacial History of West-Central North America. Géographie physique et Quaternaire 54(3):257-270.

Reitz, Elizabeth J., and Elizabeth S. Wing 1999 Zooarchaeology. Cambridge University Press, Cambridge.

198

Ritterbush, L. W. 1991 The Fur Trade of Northeastern North Dakota: The 1990 Fur Trade Sites Project. Contribution Number 260, Department of Anthropology, University of North Dakota, Grand Forks.

Robinson, E. B. 1966 History of North Dakota. University of Nebraska Press, Lincoln.

Roetzel, K. and R. Strachan 1992 An Archaeological Survey of Two Prehistoric Sites Adjacent to Fritsche Creek, Nicollet County, Minnesota. Impact Services, Incorporated, Mankato.

Roetzel, K., R. Strachan, and C. Broste 1994 An Archaeological Report of a Limited Phase III Mitigation of the Fritsche Creek Bison Kill Site in Nicollet County, Minnesota. Impact Services Incorporated, Mankato, Minnesota.

Root, M. 1992 The Knife River Flint Quarries: The Organization of Stone Tool Production. Ph.D dissertation, Washington State University, Pullman. University Microfilms, Ann Arbor.

1997 Production for Exchange at the Knife River Flint Quarries, North Dakota. Lithic Technology 21:33-50.

1999 Methods and Techniques for Lithic Analysis: Alliance Pipeline Archaeological Project. Draft copy, Ms. on file at Minnesota Archaeology Consulting, Inc., Minneapolis, Minnesota.

2001 Stone Tools and Flake Debris from 32RI785. In Alliance Pipeline L.P.: Excavations at 32RI785, Richland County, North Dakota, edited by Clark A Dobbs. Hemisphere Field Services Reports of Investigation Number 614, Minneapolis.

2004 Technological Analysis of Flake Debris and the Limitations of Size-Grade Techniques. In Aggregate Analysis in Chipped Stone, edited by C. T. Hall and M. L. Larson, Pp. 65–94. University of Utah Press, Salt Lake City.

Schalk, R. and D. Meatte 1988 The Archaeological Features. In The Archaeology of Chester Morse Lake: The 1986-87 Investigations for the Cedar Falls Improvement Project, edited by R. F. Schalk and R. L. Taylor, pp. 8.1 - 8.8. Center for Northwest Anthropology, Washington State University, Seattle Research Unit, Seattle.

Schoen, C. 2006 Phase I and II Archaeological Investigations of Alternative Route Corridors for Trunk Highway 41 Near Chaska, Carver and Scott Counties, Minnesota. Louis Berger Group, Marion, Iowa

Semenov, S. 1976 Prehistoric Technology: An Experimental Study of the Oldest Tools and Artefacts from Traces of Manufacture and Wear. Barnes and Noble Books, Totowa, New Jersey.

199

Shay, C. 1971 The Itasca Bison Kill Site: An Ecological Analysis. Minnesota Historical Society, St. Paul.

1978 Late Prehistoric Bison and Deer Use in the Eastern Prairie-Forest Border. In Bison Procurement and Utilization: A Symposium, edited by Leslie B. Davis and Michael Wilson. Plains Anthropologist Memoir 14:194-212.

Shen, C. 1999 Were ‘Utilized Flakes’ Utilized? An Issue of Lithic Classification in Ontario Archaeology. Ontario Archaeology 68:63-73.

Smith, Philip W. 1979 The Fishes of Illinois. University of Illinois Press, Urbana.

Steinbring, J. 1974 The Preceramic Archaeology of Northern Minnesota. In Aspects of Upper Great Lakes Anthropology: Papers in Honor of Lloyd A. Wilford, edited by E. Johnson, pp. 64-73. Minnesota Prehistoric Archaeology Series Number 11. Minnesota Historical Society, St. Paul.

Stoltman, J. 1973 The Laurel Culture in Minnesota. Minnesota Prehistoric Archaeology Series No. 7. Minnesota Historical Society, St. Paul.

Sylvester, S. 1988 Ruby on the Red: East Grand Forks, Minnesota. In A Meeting of the Reds: East Grand Forks 1887-1987, East Grand Forks Centennial Committee, East Grand Forks, Minnesota.

Syms, E. 1969 The McKean Complex as a Horizon Marker in Manitoba and on the Northern Great Plains. M.A. thesis, Department of Anthropology, University of Manitoba, Winnipeg.

1977 Cultural Ecology and Ecological Dynamics of the Ceramic Period in Southwestern Manitoba. Plains Anthropologist, Memoir 12.

1982 The Arvilla Complex: A Re-Assessment. Journal of the North Dakota Archaeological Association 1:135-166.

1985 Fitting People into the Late Prehistory of the Northwestern Plains: A Need to Return to a Holistic Anthropological Approach. Reprints in Anthropology 31:73-107.

Taggart, D. 1981 Notes on the Comparative Study of Fire-Cracked Rock. In Report of Phase I and II Archaeological Survey 0/ Proposed M-275 Right-of Way Through Western Oakland County, edited by D. Ozker and D. W. Taggart, pp. 142-152. Museum of Anthropology, University of Michigan ,Ann Arbor.

200

Terrell, M., J. Kloss, and M. Kolb 2005 Trunk Highway 14 – New Ulm to North Mankato Archaeological Survey, Nicollet County, Minnesota: Phase I Archaeological and Geomorphological Survey and Phase II Archaeological Testing of 21NL58, 21NL59 and 21NL134. Report prepared for the Minnesota Department of Transportation and the Federal Highway Administration. Two Pines Resource Group, LLC, Shafer.

Thoms, A. 1986 Hot Rocks: The Short-Term Storage of Energy in the Calispell Valley. In Calispell Valley Archaeological Project Interim Report for 1984 and 1985 Field Seasons, edited by A. V. Thoms and G.C. Burtchard, pp. 447-477. Contributions in Cultural Resource Management No. 10, Center for Northwest Anthropology, Washington State University, Pullman.

1989 The Northern Roots of Hunter-Gatherer Intensification: Camas and the Pacific Northwest. Unpublished Ph.D. Dissertation, Department of Anthropology, Washington State University, Pullman.

2007 Fire-Cracked Rock Features on Sandy Landforms in the Northern Rocky Mountains: Toward Establishing Reliable Frames of Reference for Assessing Site Integrity. Geoarchaeology: An International Journal, Vol. 22, No. 5, 477–510.

2008a The Fire Stones Carry: Ethnographic Records and Archaeological Expectations for Hot- Rock Cookery in Western North America. Journal of Anthropological Archaeology Science Vol. 27, pp. 443-460.

2008b Ancient Savannah Roots of the Carbohydrate Revolution in South-Central North America. Plains Anthropologist, Vol. 53, No. 205, pp. 121–136.

2009 Rocks of Ages: Propagation of Hot-Rock Cookery in Western North America. Journal of Archaeological Science Vol. 36, pp. 573-591.

Thompson, R. G. 1990 The Archaeology of the Dahnke-Reinke Site (32CS29). Report submitted to the North Dakota State Historical Society.

Tomka, S. A. 1989 Differentiating Lithic Reduction Techniques: An Experimental Approach. In Experiments in Lithic Technology, edited by Daniel S. Amick and Raymond P. Mauldin, pp. 137-161. BAR International Series 528, Oxford, England.

Toom, D. 2004 Northeastern Plains Village Complex Timelines and Relations. Plains Anthropologist 49(191):281-297.

Vaughan, P. 1985 Use-Wear Analysis of Flaked Stone Tools. University of Arizona Press, Tucson.

Walker, E. 1992 The Gowen Sites: Cultural Responses to Climatic Warming on the Northern Plains (7500 – 5000 B.P.). Archaeological Survey of Canada, Mercury Series Paper 145. Canadian Museum of Civilization, Ottawa, Ontario.

201

Wandsnider, L. 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.

Web Soil Survey 2017 Soil Survey Staff, Natural Resources Conservation Service, United States Department of Agriculture. Web Soil Survey. Available online at http://websoilsurvey.sc.egov.usda.gov/App/WebSoilSurvey.aspx. Accessed 10/29/2015.

Webster, S. 2004 A Re-Evaluation of the McKean Series on the Northern Plains. Ph.D. dissertation, Department of Archaeology, University of Saskatchewan, Saskatoon.

Wedel, W., W. Husted, and J. Moss 1968 Mummy Cave: Prehistoric Record from Rocky Mountains of Wyoming. Science 160(3824):184-186

Wendt, D. 1988 Applications of Shovel Testing and Spatial Analysis to a Disturbed Limited Activity Woodland Site in Anoka County, Minnesota. Minnesota Archaeologist Vol. 47, pp.3-33.

Wentworth, C. 1922 A Scale and of Grade and Class Terms for Clastic Sediments. Journal of Geology 30:377- 392.

Whittaker, W. 1994 The Cherokee Excavations Revisited: Bison Hunting on the Eastern Plains. North American Archaeologist 19(4):293-316.

Widga, C. 2006 Bison, Bogs, and Big Bluestem: The Subsistence Ecology of Middle Holocene Hunter- Gatherers in the Eastern Great Plains. Ph.D dissertation, University of Kansas.

Wilford, L. A. 1955 A Revised Classification of the Prehistoric Cultures of Minnesota. American Antiquity 21-2:130-142.

1970 Burial Mounds of the Red River Headwaters. Minnesota Prehistoric Archaeology Series. Minnesota Historical Society, St. Paul.

Wilson, G. 1917 Agriculture of the Hidatsa Indians: An Indian Interpretation. University of Minnesota, Studies in the Social Sciences 9. Minneapolis.

Winchell, N. H. 1911 The Aborigines of Minnesota. Minnesota Historical Society, St. Paul.

Wright, H. E., Jr. 1974 The Environment of Early Man in the Great Lakes Region. In Aspects of Upper Great Lakes Anthropology: Papers in Honor of Lloyd A. Wilford, edited by Elden Johnson, pp. 8-14. Minnesota Prehistoric Archaeology Series No. 11. Minnesota Historical Society, St. Paul. 202

Yerkes, R. 1987 Prehistoric Life on the Mississippi Floodplain: Stone Tool Use, Settlement Organization, and Subsistence Practices at the Labras Lake Site, Illinois. University of Chicago Press.

Yerkes, R. and P. Kardulias 1993 Recent Developments in the Analysis of Lithic Artifacts. Journal of Archaeological Research 1:89-119.

Yohe, R. 1998 The Introduction of the Bow and Arrow and Lithic Resource Use at Rose Spring (CA-INY - 372). Journal of California and Great Basin Anthropology Vol. 20, No. I, pp. 26-52 (1998).

Young, D. and D. Bamforth 1990 On the Macroscopic Identification of Used Flakes. American Antiquity 55:403-440.

Zdanowicz, C., Zielinski, G., and M. Germani 1999 Mount Mazama Eruption; Calendrical Age Verified and Atmospheric Impact Assessed. Geology 27(7): 621-624.

Zurel, R. 1979 Brief Comments Regarding the Nature of Fire Cracked Rock on Aboriginal Sites in the Great Lakes Area. Working Papers in Archaeology No.3, Laboratory of Archaeology, Oakland University, Rochester, Michigan.

1982 An Additional Note on the Nature of FCR. Paper prepared for distribution at the October 9th Meeting of the Conference on Michigan Archaeology.

203