Citation
A Comparison of portable and bedrock ground stone technology among hunter-gatherers at a trinchera cave in southeastern Colorado

Material Information

Title:
A Comparison of portable and bedrock ground stone technology among hunter-gatherers at a trinchera cave in southeastern Colorado
Creator:
Baker, Ryan Christopher
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English

Thesis/Dissertation Information

Degree:
Master's ( Master of arts)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Anthropology, CU Denver
Degree Disciplines:
Anthropology
Committee Chair:
Stone, Tammy
Committee Members:
Hodgkins, Jamie M.
Kent, Jonathan

Notes

Abstract:
In 2007, the Chaquaqua Plateau Ground Stone Project (CPGSP) initiated a long-term study of the ground stone features found on the Chaquaqua Plateau in Southeastern Colorado. The purpose of the multi-disciplinary project is to describe, study, and standardize research into these understudied features since minimal research into the nature of the bedrock ground stone features has been conducted. We know very little about their function, site distribution, and relationship to the regional archaeological picture. The features were used as food-grinding areas to process local wild resources, while others believe they were used to process corn that was grown in the canyon floodplain. Based on ethnographic data from other regions, some researchers have suggested that these features may have been used for a number of different functions. Missing from current research for the region is an understanding of how portable and bedrock ground stone are related to regional hunter-gatherer adaptation and if these relationships changed through time. The purpose of my research is twofold: 1) to determine if there is a functional difference between bedrock ground stone and portable ground stone and 2) to determine when bedrock ground stone is preferred over portable ground stone within the mobility and subsistence system seen at Trinchera Cave. This research may lend insight to how the prehistoric inhabitants of the area articulated with the landscape through their subsistence system. The marginal value theorem and the technological investment model have direct relevance to this research. Building on this study, future research within the Chaquaqua Plateau should include refining assumptions and applying the models to additional sites in the region.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
Copyright Ryan Christopher Baker. Permission granted to University of Colorado Denver to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.

Downloads

This item has the following downloads:


Full Text
A COMPARISON OF PORTABLE AND BEDROCK
GROUND STONE TECHNOLOGY AMONG HUNTER-GATHERERS AT TRINCHERA CAVE IN SOUTHEASTERN COLORADO by
RYAN CHRISTOPHER BAKER B.A., Metropolitan State University of Denver, 2015
A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Arts Anthropology Program
2019


11
©2019
RYAN CHRISTOPHER BAKER
ALL RIGHTS RESERVED


This thesis for the Master of Arts degree by Ryan Christopher Baker has been approved for the Anthropology Program by
Tammy Stone, Chair Jamie M. Hodgkins
Jonathan Kent


IV
Baker, Ryan Christopher (MA, Anthropology Program)
A Comparison of Portable and Bedrock Ground Stone Technology Among Hunter-Gatherers at Trinchera Cave in Southeastern Colorado Thesis directed by Professor Tammy Stone
ABSTRACT
In 2007, the Chaquaqua Plateau Ground Stone Project (CPGSP) initiated a long-term study of the ground stone features found on the Chaquaqua Plateau in Southeastern Colorado. The purpose of the multi-disciplinary project is to describe, study, and standardize research into these understudied features since minimal research into the nature of the bedrock ground stone features has been conducted. We know very little about their function, site distribution, and relationship to the regional archaeological picture. The features were used as food-grinding areas to process local wild resources, while others believe they were used to process corn that was grown in the canyon floodplain. Based on ethnographic data from other regions, some researchers have suggested that these features may have been used for a number of different functions.
Missing from current research for the region is an understanding of how portable and bedrock ground stone are related to regional hunter-gatherer adaptation and if these relationships changed through time. The purpose of my research is twofold: 1) to determine if there is a functional difference between bedrock ground stone and portable ground stone and 2) to determine when bedrock ground stone is preferred over portable ground stone within the mobility and subsistence system seen at Trinchera Cave. This research may lend insight to how the prehistoric inhabitants of the area articulated with the landscape through their subsistence system. The marginal value theorem and the technological investment model have direct


V
relevance to this research. Building on this study, future research within the Chaquaqua Plateau should include refining assumptions and applying the models to additional sites in the region.
The form and content of this abstract are approved. I recommend its publication.
Approved: Tammy Stone


VI
DEDICATION
This thesis is dedicated to my wife Laura and my sons Ethan and Caleb. Thank you to the guest services staff at the Denver Museum of Nature & Science for allowing me to take the necessary time off to pursue this. I want to thank everyone that has helped me along the way in unwavering support of my dream, including other family members, friends, and colleagues. Never stop giving up on your dreams. You can become anything you want to be if you work hard enough.


ACKNOWLEDGEMENTS
I would like to offer my sincerest gratitude to everyone that has been a part of this process. I would like to thank my academic advisor and thesis chair, Dr. Tammy Stone, and my committee members Dr. Jamie Hodgkins and Dr. Jonathan Kent for taking the time to read and comment on my thesis. I am extremely thankful to Dr. Stone for our many conversations during my time spent at UCD. Your comments and suggestions on my thesis drafts were very much appreciated. My thesis was much improved as a result of your guidance. To my cohorts in the UCD anthropology department, thank you for your inspiration and encouragement. I wish to thank the remaining faculty members at UCD’s anthropology department for pushing me to succeed and do better in each one of their respective classes. I also wish to thank Loretta Martin and the Louden-Henritze Archaeology Museum at Trinidad State Junior College in Trinidad, Colorado for allowing me to conduct my research on the ground stone collections housed there.
Many other individuals helped me along the way. They are: Philip Baca and Buford Garcia, Baca Revocable Trust; Kevin Black, Office of the State Archaeologist (retired); Dr. Linda Cummings and R.A. Varney, PaleoResearch Institute, Inc.; Lauren Fuka, University of Michigan Museum of Anthropological Archaeology; Bea Gallegos, Colorado State Land Board; Dr. Michele Koons, Denver Museum of Nature & Science; Dr. Elizabeth Lynch, University of Wyoming; Jeff Noblett, Colorado College; Brooke Rohde, University of Denver Museum of Anthropology; Rebecca Simon, Office of the State Archaeologist; Connie Turner, University of Colorado Denver; Dr. Mary Van Buren, Colorado State University, and the Karen S. Greiner Endowment; Leah Zavaleta, University of Denver; and Christian Zier, Centennial Archaeology, Inc. Finally, I would like to thank all my family, friends, and colleagues without whom this thesis could not have been completed.


Vlll
TABLE OF CONTENTS
CHAPTER
I. INTRODUCTION.............................................................1
Arrangement of the Thesis.................................................3
II. THEORY...................................................................5
Interpretative Framework..................................................5
Human Behavioral Ecology.............................................5
Optimal Foraging Theory..............................................6
Marginal Value Theorem.........................................8
Technological Investment Model.................................8
Middle Range Theory.................................................10
Forager/Collector Model.............................................11
Case Study...............................................................13
III. BACKGROUND..............................................................15
Introduction.............................................................15
Early Researchers at Trinchera Cave......................................19
Etienne Renaud (1930 - 1941)........................................20
Haldon Chase (1949 - 1950)..........................................22
Herbert W. Dick (1954 - 1957).......................................25
Colorado Archaeological Society (1966 and 1969).....................26
Caryl Wood Simpson (1974)...........................................27
Contemporary Researchers at Trinchera Cave ..............................28
Office of the State Archaeologist (1997 - 1999)
28


ix
Michael Nowak (1999 - 2001)............................................29
Colorado College (2002 - 2003).........................................30
Christian J. Zier (2013 - 2015)........................................33
IV. GROUND STONE TECHNOLOGY.....................................................35
Portable Ground Stone.......................................................35
Bedrock Ground Stone........................................................38
Grinding Behaviors..........................................................40
Design and Manufacture.................................................40
Use....................................................................41
Use-Wear Analysis......................................................43
Wear...................................................................43
Wear Rates and Patterns................................................45
Resources..............................................................46
Translating Portable Ground Studies to Bedrock Ground Stone.................47
V. METHODOLOGY.................................................................49
Introduction................................................................49
Data Limitations............................................................49
Inter-observer Variability.............................................49
Other Limitations......................................................49
Sample Selection............................................................50
Pollen Analysis........................................................53
Artifact Analysis......................................................56
Archaeobotanical Analysis..............................................56


X
ED-XRF (Energy Dispersive X-Ray Fluorescence) Analysis.................60
Results................................................................61
Data Collection Techniques on Portable Ground Stone.........................62
Data Collection Techniques on Bedrock Ground Stone..........................68
VI. RESULTS.....................................................................73
Introduction................................................................73
Metate Sample Description...................................................73
Comparison of Metates to Bedrock Grinding Surfaces.....................78
Summary of Results.....................................................79
Mano Sample Description.....................................................80
Comparison of Manos to Grinding Surfaces...............................80
Summary of Results.....................................................83
VII. CONCLUSION..................................................................85
Summary.....................................................................85
Suggestions for Future Research.............................................87
Site Protection........................................................87
Museum Collections.....................................................88
Future Research Questions..............................................89
REFERENCES
91


XI
LIST OF TABLES
TABLE
3.1: Radiocarbon Dates from Trinchera Cave..................................31-32
5.1: Pollen Types Recovered in Samples from Site 5LA1057, Las Animas County,
Colorado..................................................................62
6.1: T-test results comparing the length and width of bedrock ground stone to
portable ground stone.....................................................78
6.2: Histogram results comparing mano length, width, and thickness to bedrock
and portable grinding surfaces............................................84


xii
LIST OF FIGURES
FIGURE
2.1: Graphic Representation of the Marginal Value Theorem.................................9
3.1: Relief Map of Trinchera Cave in Southeastern Colorado...............................16
3.2: Trinchera Cave Site Map.............................................................18
3.3: Map of Trinchera Cave Showing the Locations of Excavation Units, 1949 - 2001 .......19
3.4: Overview of Trinchera Cave..........................................................20
4.1: Grinding Strokes and the Subsequent Wearing of Two Adjacent Surfaces Against a
Metate Surface......................................................................45
5.1: Different Views of Mano Used for Analysis...........................................55
5.2: Photos of Mano Taken at 15X - 20X Magnification.....................................57
5.3: Pollen Diagram for a Mano, 5LA1057, Las Animas County, Colorado.....................59
5.4: Spectral Image for Area of Mano Exhibiting Blue Stain...............................63
5.5: Blank Artifact Analysis Sheet.......................................................65
5.6: Photos of Bedrock Grinding Surfaces Taken at Trinchera Cave.........................69
5.7: Trinchera Cave Base Camp............................................................71
6.1: Bar Graph of Raw Material Types.....................................................74
6.2: Bar Graph of Artifact Types.........................................................74
6.3: Box Plot Illustrating Length Range for Bedrock and Portable Grinding Surfaces of
Each Raw Material Type..............................................................76
6.4: Box Plot Illustrating Width Range for Bedrock and Portable Grinding Surfaces of
Each Raw Material Type..............................................................76
6.5: Box Plot Illustrating Length Range for Each Artifact Type...........................77
6.6: Box Plot Illustrating Width Range for Each Artifact Type............................77
6.7: Hi stogram Illustrating Mano Length.................................................81


Xlll
6.8: Hi stogram Illustrating Mano Length....................................................81
6.9: Histogram Illustrating Mano Thickness..................................................82
6.10: Bar Graph of Wear Pattern Occurrence...................................................82
6.11: Bar Graph of Use-Wear Occurrence.......................................................83


1
CHAPTERI INTRODUCTION
In Southeastern Colorado, bedrock ground stone features are primarily concentrated along the canyon sidewalls of the Apishapa, Chacuaco, Cuchara, and Purgatoire Rivers and their tributaries (Campbell, 1969; Loendorf, 2008; Lynch, 2010; Owens, 2007). Archaeological research on the function and meaning of these features is in the beginning stages in this region. Ethnographic evidence for these features in other geographic regions indicates a number of possible functions - individual and/or group production of food resources for daily consumption and ceremonial activities such as harvest festivals or initiation rituals for women (Barrett and Gifford, 1933; Bartlett, 1933; Claassen, 2011; Curtis, 1924; Frisbie, 1967; Gayton, 1948a,
1948b; Gifford, 1932; Jackson, 1991; Kluckhohn et al., 1971). An important aspect of these grinding features is the social significance of the area where shared labor was organized and where social expectations were expressed between generations (Dick-Bassonnette, 1998;
Gayton, 1948a, 1948b; Jackson, 1991). Lynch (2014) suggests that bedrock ground stone features vary depending on where they appear on the landscape and because of sociocultural beliefs, values, and traditions. Bedrock ground stone features have the potential to enrich our understanding of past social dynamics between prehistoric peoples and their environment.
The purpose of this research is twofold: 1) to determine if there is a functional difference between bedrock ground stone and portable ground stone and 2) to determine when bedrock ground stone is preferred over portable ground stone within the residential mobility and subsistence system seen at Trinchera Cave (5LA1057). This research may lend insight to how the prehistoric inhabitants of the area articulated with the landscape through their subsistence system. Archaeology in the region suggests a highly complex set of mobility and settlement


2
patterns based on the cultural remains of architecture, ceramics, lithics, and perishables. Trinchera Cave is an especially important site because of its exceptionally well-preserved archaeological materials, geographic location, rock art panels, and a continued occupation spanning from the Late Archaic Period (1050 BC - 100 AD) to the Late Prehistoric stage (AD 100 - 1725) (Crocket, 2002). The perishable remains of bark, bone, feather, fiber, seeds, shell, skin, and wood are unmatched for the region (Simpson, 1976). Additionally, this is the only site in southeastern Colorado that has yielded the three domesticated crops - beans, maize, and squash (Black, 2000).
Large sections of the rock shelter were excavated by avocational and professional archaeologists during several projects between 1949 and 2001, with the greatest amount of work occurring during the 1950s. Most of the artifacts and other excavated materials discovered during these excavations are housed at the Louden-Henritze Archaeology Museum at Trinidad State Junior College (TSJC) in Trinidad, Colorado. Unfortunately, artifact inventories, field notes, and survey manuscripts from excavations prior to the 1970s are exceedingly rare and often times inadequate (Zier, 2015). A good deal of ambiguity appears to surround the earlier projects. The only project for which we have substantial documentation on the site comes from four field school sessions between 1999 and 2001 that was led by Michael Nowak of Colorado College. A comprehensive map of the site was created between 2013 and 2015 by Chris Zier of Centennial Archaeology, Inc. A considerable amount of looting has taken place at the site, resulting in the loss of archaeological materials and information. Despite this, the materials housed at the Louden-Henritze Archaeology Museum and the features at the site have the potential to address questions relating to ground stone use in the area.


3
Arrangement of the Thesis
Chapter 1 outlines my thesis argument regarding the ground stone technology used by hunter-gatherers at Trinchera Cave in southeastern Colorado through the analysis of portable and bedrock grinding surfaces using human behavioral ecology.
Chapter 2 discusses the different theoretical models that can help us to understand how Trinchera Cave fits into a broader use of the prehistoric landscape. Specifically, the collection of bedrock ground stone features in the region is technology directly tied to the subsistence system. This chapter focuses on the relationship between subsistence practices and settlement patterns in the region. Constructing early hunter-gatherer mobility patterns and settlement systems are crucial in our understanding of the archaeological record.
Chapter 3 provides a background on the early and contemporary researchers that have worked at Trinchera Cave and the roles they each played within the framework of a research design at the site.
Chapter 4 highlights the importance of ground stone technology and how it can potentially contribute to our understanding of past lifeways. The role of portable and bedrock ground stone is examined here. Different grinding behaviors such as a tool design, manufacture, and utilization are covered. Raw material availability can directly affect ground stone tool design, manufacture, use, and abandonment patterns.
Chapter 5 explains the methodology employed during the course of this study. Portable ground stone recovered from Trinchera Cave and bedrock ground stone present at the site are used as the two sub-groups for analysis. A pollen, starch, and XRF analysis was conducted on one of the portable artifacts. With analysis carried out by PaleoResearch Institute, Inc., testing determined the substances that were being processed by the grinding tool. Data collection


4
techniques of the portable and bedrock ground stone are covered. IBM SPSS Statistical Software (V. 25) is used to record and manage these data sets.
Chapter 6 provides the results from my analysis as well as a discussion suggesting there is a relationship between artifact type and median length and width. A bimodal distribution was discovered for mano length in particular. These portable and bedrock grinding surfaces show strong evidence of specific wear patterns that can possibly be tied in with mano wear patterns. Results indicate that the portable grinding surfaces may have been predominantly used at Trinchera Cave, with the bedrock grinding surfaces being used as a complementary form of technology.
Finally, Chapter 7 summarizes the thesis research and discusses insights into human behavior and adaptation in the region. Recommendations are then given on the site, museum collections, and future research.


5
CHAPTER II THEORY
Interpretative Framework
Researchers interested in explaining technological change as a result of adaptive decision making have gravitated toward two theoretical approaches: Human Behavioral Ecology (HBE) and Optimal Foraging Theory (OFT). While these two approaches may have different intellectual beginnings, OFT and HBE share many fundamental beliefs and goals.
• They both wish to understand the flexibility of human behavior in regards to economic constraints.
• They each believe that time and energy are key currencies to understanding complex technological behavior.
• They each believe that optimal behaviors become commonplace, keeping conditions constant.
• Both assume that changes in technology are not subject to natural selection. Instead, technological change reflects plasticity in the behavioral phenotype.
• Both approaches have addressed questions of material design and technological investment.
Human Behavioral Ecology
Human behavioral ecology (HBE) is one of the most widely used theoretical frameworks in the archaeology of hunter-gatherers. HBE examines the flexibility in human behavior and uses formal hypotheses to explore the adaptive responses of individuals to environmental pressures (Bird and O’Connell, 2006; Broughton and Cannon, 2010; Codding and Bird, 2015; Lupo, 2007; Nettle et al., 2013; Winterhalder and Smith, 2000), especially those related to


6
questions on prehistoric subsistence practices (Broughton, 2002; Cannon, 2003; Madsen, 1993; Madsen and Schmitt, 1998; Simms, 1987; Ugan, 2005; Whelan et al., 2013).
Optimal Foraging Theory
Optimal Foraging Theory (OFT) was first developed by behavioral ecologists to help them better understand the foraging behavior of animals (Emlen, 1966; MacArthur and Pianka, 1966; Pulliam, 1974; Stephens and Krebs, 1986). OFT models address such issues as habitat movement, the selection of specific resources, and time allocation (Winterhalder and Smith, 2000). Food selection and resource utilization are regulated by various factors including predation, resource availability, seasonality, and territory size (Keene, 1983). By employing a behavioral ecology framework, we can often predict the optimal foraging behavior of animals. Keene (1983: 139) recognized a cost-benefit component associated with this approach. While food may provide an animal with energy, other factors are known to influence foraging behavior such as the time and energy costs associated with searching, collecting, and preparing the food. OFT suggests which resources will be collected and which will be ignored (Cannon, 2003; Kelly, 2013) based on these factors. The foraging strategies that offer the greatest benefit at the lowest cost are commonly used (Keene, 1983).
Researchers have observed similarities in behavior among animals and humans. Studies on animal foraging behavior can improve our understanding of different mobility strategies (Keene, 1983; Kelly, 2013). OFT models include several features, including a goal, a unit of currency, a set of constraints, and a range of options (Kelly, 2013; Smith et al., 1983; Winterhalder and Smith, 2000). Often times, the goal is to be as efficient as possible while foraging (described as energy gained over energy spent) (Kelly, 2013). Currency is typically a measure of energy, such as calories gained per unit of time (Smith et al., 1983). However, other


7
measures of resource value can include material requirements, monetary exchange, ornament or prestige items, or protein capture (Winterhalder and Smith, 2000). The set of constraints are all the possible factors that may limit the amount of time an individual or group can spend foraging or processing resources. One important example of this is caring for children, which may have played a significant role for hunter-gatherer’s in general (Homsey-Messer, 2015; Whelan et al., 2013) and at Trinchera Cave in particular. Lastly, the range of options refers to the potential food resources that are available in the area. Resources have specific characteristics such as ritual or ceremonial importance, harvesting and processing times, medicinal value, and nutritional content (Kelly, 2013: 47; Nowak and Gerhart, 2002: 35 -36).
These characteristics influence which resources will be harvested and which will be overlooked (Cannon, 2003; Kelly, 2013). This is also referred to as the diet-breadth model (DMB). DMB is a traditional optimal foraging model from behavioral ecology that ranks food resources according to their potential caloric benefits and search, procurement, and processing costs (Kelly, 2013; Madsen, 1993; Madsen and Schmitt, 1998). According to this model, foragers will select specific resources that maximize foraging return rates (Kelly, 2013; Madsen, 1993; Madsen and Schmitt, 1998; Whelan et al., 2013). Ethnoarchaeological studies of modern and prehistoric hunter-gatherer groups validate the importance of this model (Bird and O’Connell, 2006; Smith, 1991; Smith et al., 1983; Hill and Hawkes, 1983). DMB takes into account how long it takes to find a particular resource and then how long it takes to collect and process it once it is located (Kelly, 2013). By considering the foraging strategies of different animals, we can begin to understand the complexities about human behavior.


8
Marginal Value Theorem
The marginal value theorem (MVT) was first proposed by Eric Charnov (1976). MVT is a widely used optimality model that describes the optimal foraging behavior of an individual or group in a system where resources are geographically dispersed in discrete patches (Calcagno et al., 2014; Charnov, 1976; Kuhn and Miller, 2015) (Figure 2.1). These patches are often separated by areas with no viable resources. MVT is also useful in those situations where individuals face declining return rates while foraging (Kelly, 2013: 65). Earlier research has shown that in order to capitalize on foraging return rates, foragers will leave a resource patch once the harvest rate reaches the average rate of the remaining resource patches (Charnov, 1976; Kelly, 2013). A patch can be something as small as a single food source or as large as natural habitat zone (Kuhn and Miller, 2015). Foragers tend to move out of a resource patch before the return rate out of that patch has dropped to zero (Kelly, 2013).
Technological Investment Model
Archaeologists are particularly drawn to foraging models, because the evidence of past subsistence practices and the social use of space are critical components of the archaeological record. Early foraging models do not typically take into account the effects of subsistence technology (Bright et al., 2002). Nevertheless, the technological investment model does just that (Bettinger et al., 2006; Kelly, 2013). The model estimates how much time an individual and/or group should invest in any specific type of technology. Most technologies are manufactured to help minimize resource handling time (Kelly, 2013). Ethnographic examples have shown that the initial construction of more complex tools and technologies can be huge. For example, the up-front costs of fish weirs and hunting nets require a great deal of time to manufacture (Olson, 1936; Bailey and Aunger, 1989). Hunting nets can take months to build (Satterthwait, 1987).


9
<------------ time to new patch time in patch---------------->
Figure 2.1. Graphic representation of the marginal value theorem.
Figure from Kuhn and Miller (2015: 175).
That time increases with any raw materials that are collected for the rope and the time necessary to create it (Lindstrom, 1996; Olson, 1936; Satterthwait, 1987).
Complex tools and technologies also require regular maintenance once they are constructed. When compared to the cost of initial construction, routine maintenance is a lesser expense since it is often carried out during down time (Kelly, 2013). Manufacturing complex technology often requires large investments of time that could potentially be spent acquiring or processing food, though with a less efficient form of technology (Kelly, 2013). Any increases with a slightly more complex form of technology is expected to be small, while major differences in technology likely represent a more significant increase in return rates (Kelly, 2013), such as with the difference between a fishing spear and gill net (Bettinger et al., 2006). The fishing spear offers a lower return rate than the gill net but requires less time to manufacture. Handling time and return rates are directly associated with manufacturing time (Kelly, 2013).


10
By using the technological investment model, we can determine when a specific technology will be enhanced or when a more complex form of technology replaces another (Bettinger et al., 2006; Buonasera, 2015). Using the earlier example, a gill net can bring in a nice haul but at a lower return rate than the fishing spear if we factor in the manufacturing time. The technological investment model offers several insights (Bettinger et al., 2006): when technology changes it is believed to do so (a) quickly, (b) extensively across a population, and (c) usually permanently. Changes in technology can often be explained with the help of this model. When two coexisting forms of technology are used to acquire or process the same resource, they should have similar rates of return (Kelly, 2013). Coexisting forms of technology are often used for different purposes. This model is used to explore the functional differences between bedrock ground stone and portable ground stone found on the Chaquaqua Plateau in Southeastern Colorado and will be covered in a subsequent chapter.
Middle Range Theory
In the 1960s, the term middle range theory was first applied to archaeology (Pierce, 1989). Middle range theories provide researchers with the ability to connect human behavior and natural processes to material remains (Raab and Goodyear, 1984). Many of these theories come directly from experimental research and ethnoarchaeological studies. Material remains uncovered in the present allow archaeologists to make inferences about past human behavior. Hunter-gatherer behavior is highly variable, and so we can expect to see a variety of signature patterns in the archaeological record. Ethnoarchaeological studies often attempt to relate human behavior with material remains (O’Connell, 1995; Kelly, 2013).
The spatial and temporal distribution of hearths, flaked debris, and/or other material remains found at a site are used to reconstruct group size and/or how long a site may have been


11
occupied (Kelly et al., 2005, 2006). Similarly, the patterns and distribution of bedrock ground stone features can aid in our understanding of prehistoric technology, landscape use, and subsistence practices (Lynch, 2017c). Hunter-gatherers use different mobility strategies to schedule the collection and processing of resources. Binford (1980) argued that individual and group movement is predicated on different variables including demographics, ecology, and resource availability. He distinguished logistical mobility (the movement of specially organized task groups on short trips from a residential base) from residential mobility (the movement of all members of a residential base from one location to another) (Binford, 1980, 1990; Fitzhugh and Habu, 2002). Based on these differences, Binford identified these two basic settlement patterns that allow researchers to better understand subsistence practices.
Forager/Collector Model
Archaeologists have studied hunter-gatherer mobility as a continuum with residential mobility on one end and logistical mobility on the other (Binford, 1980). The study of early hunter-gatherer mobility patterns and settlement systems is crucial in our understanding of the archaeological record. When discussing resource availability, seasonal mobility among hunter-gatherers can be expectedly modeled (Binford 1980, 5 - 6). Binford’s forager/collector model is based on residential and logistical mobility. There are key differences between these two basic settlement patterns. Foragers move their group to the resources, while collectors move the resources to the group (Binford, 1980). Resource distribution often influences residential mobility patterns (Binford, 1980; Grove, 2009; Kelly, 1983; Kelly and Todd, 1988; Kent, 1992). Whallon (2006) argues that establishing and maintaining a social network among foraging groups is especially important during times of resource scarcity.


12
In a residential mobility system, individuals tend to forage for short periods of time within the general vicinity of their base camp (Kelly, 2013; Whelan et al., 2013). Hawkes et al. (1995) notes that during these foraging trips, it is typically easier to bring their children along with them. Foragers in a logistical mobility system often use residential bases for an extended period of time (Whelan et al., 2013). Women in particular are forced to make longer trips to acquire resources, since those nearby become exhausted. However, these extended foraging trips are energetically costly because they increase the cost of transport with the accompanying children (Surovell, 2000). Native American women often had to choose between bringing their children along with them on their extended foraging trips or leaving them behind under the care of others for an undisclosed amount of time (Whelan et al., 2013). Consequently, the opportunity costs of foraging are lower for women in residentially mobile systems (Whelan et al., 2013). Collectors, on the other hand, have primary residences based on resource distribution. Instead, resources are logistically moved to the group (Binford, 1980). The recovery of food storage pits at a site frequently indicates the presence of a primary residence (Binford, 1980: 5). Collectors often move great distances to acquire resources, increasing the energetic costs of collecting.
By studying mobility strategies in these terms, archaeologists are able to better understand the conditions where resources and hunter-gatherers move in relation to one another (Binford 1980; Kelly 1983). Archaeologists often look for evidence in the archaeological record (flora and fauna, settlement patterns, and distribution of site features and material remains) to help them determine which of these mobility strategies a group is using. Foragers and collectors have distinct behaviors that set them apart. Food procurement strategies is one such example. Foragers typically gather their food daily and do not usually store their foods (Binford, 1980: 5).


13
Unlike foragers, collectors store their food for part of the year and use specially organized task groups (Binford, 1980: 10). They regularly venture out to acquire specific resources, as opposed to generalized searching. The forager/collector mobility systems represent the extremes of a wide range of possible settlement patterns. Most groups exist in an in-between zone, with a complex mix of behaviors.
Case Study
These theoretical models can help us to understand how Trinchera Cave fits into a broader use of the landscape. In particular, the large collection of bedrock ground stone features in the area is technology directly tied to the subsistence system. Bedrock surfaces have been used throughout Arizona, California, Kentucky, Nevada, New Mexico, and Texas to process a variety of food resources (Basgall, 1987; Buonasera, 2016; Claassen, 2011; Gayton, 1948a, b; Madsen, 2003; Mohr, 1954; True, 1993; Wallace andHolmlund, 1983; Webb and Funkhouser, 1929). The bedrock grinding surfaces found in southeastern Colorado were probably used for similar reasons. The mobility strategies of hunter-gatherers in the area are indicated not only by food exploitation but also a possible food storage pit at the site. Resource procurement often involves planning the exploitation and storage of certain foods.
Nowak and Gerhart (2002: 12) suggest that Trinchera Cave was used on a temporary, but regular basis by people living in the region. The introduction of maize and squash sometime between AD 500 and AD 1000, may indicate that people wanted to diversify the food resources in the area, allowing them to spend more time at the site (Nowak and Gerhart, 2002: 12). The presence of food storage pits at the site would support such an argument. Simpson (1976: 23) references a single storage pit that was five inches deep and 20 inches in diameter. She suggests
that:


14
“This pit was probably used for storage since there was no indication of burning other than the oxidation of the plaster. Since the pit is located inside the structure adjacent to the wall, it seems likely that the oxidation occurred at the time the structure burned. ”
While this may have been a food storage pit, Simpson does not say anything about the
possible contents (1976: 12). Nowak and Gerhart (2002: 79) suggest the picture is not as clear
cut. They report finding three pockets of Hackberry seeds at the site. Rodents often collect
seeds for the purpose of storage (Brown and Davidson, 1977; Chaney, 1936), and this is a
possible reason for the pit described by Simpson. Ground disturbance at the site may be one
reason why Nowak and Gerhart (2002) did not locate any storage pits (Nowak and Gerhart,
2002: 12). However, it may also represent the residential mobility patterns of the inhabitants.


15
CHAPTER III BACKGROUND Introduction
Indigenous peoples have occupied the tributaries and side canyons of the Purgatoire River in Southeastern Colorado for thousands of years. Evidence reveals habitation by early woodland groups, horticulturists, and plains hunter-gatherers (Black, 2000; Lynch, 2010). An unusual feature of this region is the presence of bedrock grinding areas appearing at different elevations and with different site assemblages. While these features have been found throughout region, they are especially concentrated in the canyons of the Chaquaqua Plateau (Lynch, 2010, 2014, 2017a, 2017b, 2017c). Bedrock ground stone features are an important part of the native landscape in northeastern New Mexico and southeastern Colorado. The highest concentration of these features seems to be alongside the canyon walls of the Apishapa, Chacuaco, Cuchara, and Purgatoire Rivers and their respective tributaries (Campbell, 1969; Loendorf, 2008; Lynch, 2010; Owens, 2007). Bedrock ground stone features were frequently constructed along the sandstone canyon walls, although some have been found in other locations (Andrefsky, 1990; Campbell, 1969; Chomko and De Vore, 1990; Gunnerson et al., 1989; Hartley and Vawser, 2003; Loendorf, 2008; Renaud, 1931).
Trinchera Cave (5LA1057) is a prehistoric site located about 48 km east of Trinidad in Las Animas County, Colorado (Figure 3.1). It is in fact a naturally occurring rock shelter and not a cave in the geological sense. Some early researchers have referred to the site as Trinchera Shelter. However, the name Trinchera Cave has dominated the literature (Zier, 2015). For continuity purposes, the site is referred to as Trinchera Cave in this thesis. The site is situated near the confluence of Trinchera Creek and the Purgatoire River, roughly thirty-two kilometers


16
IVValsenburg
iTnncab]
jSOCOR!\DOM
Tnncrera Cave (5LA1057)
mm
NEWiMEXieO
WYOMING
Trinehei a Cave (5LA1057)
ssr419"
^COLORADO
MEXICO
• Site Location
50
Figure 3.1. Relief map showing the location of Trinchera Cave in Southeastern Colorado. Figure from Zier (2015: 4).
KANSAS


17
north of the New Mexico border. Trinchera Cave is located on school district land (Simpson, 1976) and is administered by the Colorado State Land Board through their Alamosa District Office (Black, 2000). The land is currently leased for cattle grazing to Mr. Antonio Baca. Recently, efforts have been made to consolidate the bulk of predominantly unpublished work that has been recorded at the site since its initial discovery in the late 1940s (Black, 2000;
Gerhart, 2001; Zier, 2015, 2017). Little research has focused on the function and relationship of the site to the regional archaeology. In fact, no research has been conducted on the ground stone materials discovered at Trinchera Cave. Archaeological discovery represents only a fraction of the overall picture.
Zier (2017) estimated the sheltered area of the cave to be around 500 m2 (Figure 3.2). However, the actual habitable area is calculated to only be around 280 m2 due to the presence of several large sandstone blocks that broke off from the cliff face and fell into the northeastern half of the shelter (Zier, 2017). This area was apparently not used by the inhabitants to any considerable degree based on the lack of cultural deposition (Zier, 2017). Between 1949 and 2001, different archaeologists excavated within the shelter (Figure 3.3). Trinchera Cave lies within a fairly small canyon (20 m deep x 75 m wide) (Figure 3.4). The shelter is some 25 - 30 feet above the stream bed and is only 6.5 km away from the Purgatoire River. The shelter is cool in the summer and warm in the winter based on its southeasterly exposure (Simpson, 1976). The somewhat secluded environment is close to major sources of water, making it possible to access other nearby regions. Rivers often serve as travel routes, helping to connect people (Noble,
2000: 23). Habitat corridors enhance the population capability of hunter-gatherers (Beier and Noss, 1998). This accessibility may have allowed for trade with other regions. The presence of imported ceramics and marine shells supports this argument. Cattle frequent the canon bottom


18
Figure 3.2. Trinchera Cave site map. Figure from Zier (2015: 11).
where the site is situated, and have impacted the surface of the shelter. The development of arroyos in the area are thought to be contemporaneous with the development of ranching, and are believed to be caused by roaming cattle (Duce, 1918). Cattle have affected the land by wearing the trails down and destroying the local vegetation, resulting in stronger opportunities for soil erosion and weaker opportunities for absorption (Duce, 1918).
Simpson (1976) believes that the site represents a combination of Plains and Southwestern cultures. Nowak and Gerhart (2002: 104) suggest that Trinchera Cave was a cultural melting pot of peoples. This belief is based on a number of pottery sherds, shell fragments, imported calcium carbonate, obsidian tools, and other diagnostic materials that represent trade with these different cultures. Simpson (1976: 203 - 204) uncovered three


19
TRINCHERA CAVE (5LA1057) NT"
1 â–  JiiMKiiBBirad
HAD IMSU1MZona 13N hr
0 IS 30 45 SO 75_
ii
0 5 10 1 20 25
A Si« Eamm
Tim
Contour
Oaecva] ~ 0.5 m)
x ^ BKiWiI
css
,»♦' Capita* CreABfri
^^Boddsc
Espc;«d Bedrock
â–¡ Chaia rxcivacca
(1950)_______
l—l (1955-57)
â–¡ Sirrywnn
Eacsrakn (1974)
â–¡ Nowak Excarocoo
(1999-2001)
Figure 3.3. Map of Trinchera Cave showing the locations of Chase, Dick, Simpson, and Nowak excavation units, 1949 - 2001. Figure from Zier (2015: 50).
different types of dateable ceramics including Sante Fe Black-on-White (AD 1200 - 1350), stamper Cordmarked (AD 1176 - 1485), and Taos Plain (AD 1100 - 1300). One Sante Fe Black-on-White sherd was retrieved from Level II in Area D, as well as thirty-eight Taos Plain sherds from multiple levels (Simpson, 1976: 155 - 156). Nowak and Gerhart (2001, 2002) also found direct evidence of Sante-Fe Black-on-White and Sopris Plain pottery from the same occupational level.
Early Researchers at Trinchera Cave
Information is provided not only on individuals who worked at the site, but also in the broader area to provide a more nuanced background and archaeological context.


20
Figure 3.4. Overview of Trinchera Cave, taken June 2018, view to the northwest from the opposite canyon rim. Trinchera Creek is visible at the bottom.
Etienne Renaud (1930 - 1941)
Etienne B. Renaud was a faculty member at the University of Denver (DU) between 1920 and 1948. He came to the United States to teach French, and had no previous knowledge of anthropology (Nelson et al., 2001). However, Renaud proved himself to be a hard worker and an avid reader. He founded the anthropology department at DU in 1922. During his time there, he developed new archaeological survey methods and was one of the first people to attempt systematic excavations in Southeastern Colorado (Renaud, 1931 - 1947). Together with his students, Renaud actively searched for prehistoric sites in Colorado and other nearby states.
They keenly sought out farmers and ranchers to find out what might be on their land. In 1929,


21
Renaud conducted a field expedition to the Lindermeier Site in northeastern Larimer County, CO for the Colorado Museum of Natural History (now known as the Denver Museum of Nature and Science) (Renaud, 1931 - 1947). Paleoindian artifacts were discovered at the Lindermeier Site by Judge Claude C. Coffin and his son, A. Lynn Coffin five years earlier (Coffin, 1937; Wilmsen and Roberts Jr, 1978). When Renaud arrived at the site, he identified several of the artifacts as Folsom points. An extraordinary collection of bison bones and Folsom points were uncovered during the 10 years of site excavations (Coffin, 1937).
He later conducted different archaeological surveys of the Great Plains. Some of the areas he surveyed included: Eastern Colorado (1930 - 1933), Eastern Wyoming (1931), Western Nebraska (1933), Northeastern New Mexico (1934 - 1935), Southern Wyoming (1935 - 1939), and Southern Colorado (1940 - 1943) (Renaud, 1931 - 1947). Renaud documented hundreds of different prehistoric sites in Eastern Colorado (Renaud, 1931, 1935, 1931 - 1947; Renaud and Chatin, 1943), with 30 of those sites being in Las Animas County (Black, 2000). Renaud was an avid explorer, documenting many of the sites he came across. His field notes, journal articles, unpublished survey manuscripts with artifact sketches and original site map drawings, photographs, postcards, microfilm, site forms, and early newspaper clippings are stored at the University of Denver.
In 1930, the first descriptions of the Snake Blakeslee Archaeological Site were published (Gunnerson et al., 1989). Renaud first came across the site earlier that same year when he surveyed different parts of Eastern Colorado. Reconnaissance surveys were primarily conducted during this time (Lintz, 1999; Black, 2000). He was guided to the Cramer and Snake Blakeslee sites by R. D. Mutz, a man from the nearby town of Fowler (Renaud, 1931, 1935, 1931 - 1947; Lintz, 1999). In 1931, Renaud returned to the same sites and again a decade later. He worked


22
on writing new descriptions of the Cramer and Snake Blakeslee sites. After his investigation, Renaud believed that the two sites each had a ceremonial function based on his discoveries (Renaud, 1931 - 1947). Preliminary excavations were later conducted at Trinchera Cave as a part of the project (Renaud, 1931). In 1948, Renaud chose to retire from the University of Denver.
The University of Denver Anthropology Museum (DUMA) currently stores the portable ground stone artifacts Renaud discovered during his field expeditions in the Apishapa Canyon (Rohde, personal communication, 2018). However, upon examination of the collection, I learned that they only had one definitive piece of ground stone from Renaud’s 1931 survey of the Cramer site, and that was a slab metate fragment made out of sandstone. There were no materials from Trinchera Cave present at DUMA. When going through the minimal amount of paperwork they did have from those earlier excavations, I learned from a small note in Renaud’s collection that two manos and one metate from an undisclosed site were thrown out at some point by Arnold Withers, a doctoral candidate from Columbia University. It is unknown at this time why or where they were discarded or even if they were from any of the sites mentioned here.
Haldon Chase (1949 - 1950)
In 1949, Haldon Chase was a graduate student at Columbia University. Chase directed a project in the area with the primary goal of understanding early Apache sites on the plains of Southern Colorado (Gunnerson et al., 1989; Lintz, 1999). The High Plains Columbia Expedition (HPCE) of 1949 was a joint expedition with Columbia University and the University of Denver (Gunnerson et al. 1989; Lintz 1999; Zier 2015, 2017). The funding for the expedition came largely from the Missouri River Basin Project (Lintz, 1999). Robert (Bob) Stigler joined the


23
expedition early on. He was also a graduate student from Columbia University. As a back-up strategy, Ferdinand (Ferd) Okada was also associated with the project (Okada, 1949a, 1949b).
He was a doctoral candidate at the time and had previously earned his Bachelor’s and Master’s degrees in California (Lintz, 1999). The three men made their way to Apishapa Canyon in Southeastern Colorado (Gunnerson et al., 1989; Lintz, 1999).
Hal Chase was selected as field director for the project, even though he had just recently graduated from Columbia with his BA (Lintz, 1999). This decision was made because he was native to Colorado and had originally envisioned the project (Lintz, 1999) and had fieldwork experience in Colorado. However, the HPCE was under the general direction of Arnold Withers from Columbia University (Lintz, 1999). The three men (Chase, Okada and Stigler) spent more than five weeks excavating in the canyon during their 1949 expedition (Colorado Encyclopedia, 2016; University of Denver Anthropology Museum, 2018). The men made a brief stop at the Cramer site sometime in mid-July. Most of their time was spent excavating at the Snake Blakeslee site, located near the mouth of the Apishapa Canyon (Encyclopedia Staff, 2016; University of Denver Anthropology Museum, 2018). At the time, Trinchera Cave was little more than an afterthought during the project. Chase decided to excavate at the site for a single day on August 21, 1949 (Zier, 2017), having primarily worked at the Snake Blakeslee site. He was joined by Jack Gilstrap, Abe Mason (local informant), Ferdinand Okada (Ph.D. candidate), Morris Taylor (history professor at TSJC), Richard and Willard Louden (land owners), and an amateur archaeologist named Stubby Ischam (Lintz, 1999). Together, they worked in the eastern most portion of the cave, referred to now as Area A (Zier, 2017).
According to the University of Denver Anthropology Museum, Hal Chase and Bob Stigler spent a majority of their time out at the site. However, Christopher Lintz (1999) paints a


24
different story. Stigler was allegedly forced to abandon the project after only 53 days (Lintz, 1999). He left for Arkansas to be with his family because his father was having an emergency surgery (Lintz 1999). After the High Plains Expedition wrapped up, Chase later returned to the area by himself on September 10, 1949 (Lintz, 1999). He spent 11 days re-investigating the Snake Blakeslee site (Lintz, 1999), which is located due north of Trinchera Cave. Chase’s research laid the groundwork for future archaeological work in the region, including Trinchera Cave.
All the archaeological materials that were discovered as a result of the HPCE were transferred to Columbia University in 1949 (Lintz, 1999). It is believed that some of the larger ground stone items may have been left in the field (Okada, 1949c; Lintz, 1999: 20). This seems plausible due to the weight of some of the metates from the later expeditions currently being housed at the Louden-Henritze Archaeology Museum. In 1955, the materials were reportedly shipped from Columbia University to the University of Denver (Zier, 2017). The location of those excavated materials from the expedition is currently unknown. Sometime between September 10 and September 21, 1949, Chase ventured to Trinchera Cave to take photographs of the site. However, the location of these black and white photographs is unclear. They are most likely housed at the Louden-Henritze Archaeology Museum in Trinidad, Colorado (Martin, personal communication, 2018a), but have not been located to date.
Trinchera Cave was first noticed by William Louden in 1948 during a reconnaissance flight (Black, 2000). Ruth Henritze, William Louden and his brother Richard, later followed up at the site by conducting a test excavation. While there, they discovered a “twig box,” bark mats made out of juniper, and unfired clay figurines (Louden and Louden, 1998). These materials are currently stored at Louden-Henritze Archaeology Museum. The Loudens chose to include


25
Haldon Chase in their work. Chase was teaching at TSJC at the time, where he had established their anthropology program (Anonymous, 1952; Cassells, 1997). Chase began more formal excavations at Trinchera Cave in the early 1950s (Black, 2000). Between June 19 - July 1,
1950, Chase and his crew excavated blocks in Areas A and C at the shelter (Zier, 2017). After the 1950 excavation had ended, Chase sent all the botanical remains to the University of Michigan Ethnobotanical Laboratory (Zier, 2017). The results are discussed later in this thesis. The recovered faunal remains were reportedly sent to the Colorado Museum of Natural History for identification (Zier, 2017). However, according to the archivist and collections data manager at the Denver Museum of Nature & Science, Sam Schiller (personal communication, 2018), no records of these materials could be found.
After conferring with a Navajo expert in La Jolla, California, Chase began to believe that all the materials recovered at Trinchera Cave were more than 1,500 years old (Chase, 1950; Lintz, 1999). Little information about this meeting exists. Chase remained with TSJC until 1953. Prior to his departure from the school, he helped select Herbert Dick to replace him at the college (Lintz, 1999). Dick picked up where Chase left off, by continuing the excavations at the cave during the 1954 - 1956 field seasons (Black, 2000). Very little information exists from the actual fieldwork conducted by Haldon Chase and Herbert Dick, with the exception of a brief summary in Simpson’s thesis (1976: 5). A short newspaper article published by the Rocky Mountain News (RMN) is also known to exist (Gavin, 1955). The short article discusses the archaeological finds at Trinchera Cave and Chase’s involvement with the project.
Herbert W. Dick (1954 - 1957)
In 1953, Dr. Herbert W. Dick replaced Haldon Chase as a faculty member at TSJC. He remained with the school until 1962 when he moved to Alamosa, Colorado to teach at Adams


26
State College (Zier, 2017). Dick began his excavations at Trinchera Cave in 1954, soon after he was hired. Working predominantly with students from TSJC, Dick was able to carry out large-scale excavations at the site. Based on photographs and limited field records, Zier (2017) determined that Dick had eight people working with him on any given day. They excavated blocks in Areas A, B, and C in the shelter, with Area B being taken all the way down to bedrock. However, the greatest amount of work was done in Area C (Zier, 2017). All the excavated materials were processed at a field laboratory that Dick had established at his base camp (Zier, 2017: 28). When Dick left TSJC, he brought many of the artifacts that were recovered from the shelter along with him (Zier, 2017). However, he never analyzed any of the materials or wrote up a report on them.
When Caryl Wood Simpson began excavating a part of the cave in the 1970s, she and William Louden were able to recover all of the materials in Dick’s possession and return them to Trinidad (Zier, 2017). All of Dick’s excavated materials are currently being housed at the Louden-Henritze Archaeology Museum. Simpson was also able to obtain a few pages of notes, a sketch map of the site, and some black and white photographs from Dick (Zier, 2017). Based on Dick’s few pages of notes, she produced an artifact inventory of the retrieved collection (Simpson, 1976: 179 - 199). However, there is no way to know for sure if Dick gave Simpson and Louden all the materials he took from Trinchera Cave. Zier (2017) reports that besides a schematic sketch from 1956, no known site map is known to exist from Dick’s excavations. Colorado Archaeological Society (1966 and 1969)
Further excavations at Trinchera Cave were carried out by the Trinidad Chapter of the Colorado Archaeological Society in the 1960s. However, not much information is currently known about these excavations. The Trinidad Chapter of the Colorado Archaeological Society


27
no longer exists. They disbanded in 1971 (Martin, personal communication, 2018b). Most of their old newsletters are currently stored at the Louden-Henritze Archaeology Museum in Trinidad, Colorado. Martin (personal communication, 2018b) believes they may have collected surface finds and that is what was recorded in 1966.
Caryl Wood Simpson (1974)
Caryl Wood Simpson was attending the University of Wyoming as a graduate student when she carried out her excavations at Trinchera Cave (Zier, 2017) from April 15-26 and again from June 12 - July 19, 1974 (Wood, 1974). Later that fall, she wrote a preliminary report on the site, its excavation procedures, and its contents (Simpson, 1976). Two years later, she completed her master’s thesis based on the results of her excavations and that of Herb Dick’s work from the 1950s. Simpson primarily worked in Area D of the shelter, southwest of where Dick had previously excavated. Between April 15 -26, 1974 Simpson managed an archaeological field school for high school students from the Denver area. She later revisited the site between June 2 - July 19 with students from TSJC to conduct further excavations.
Simpson (1976: 156) retrieved an assortment of various ceramic artifacts from Trinchera Cave. A number of unfired ceramic objects were uncovered from Level I inside of a suspected ‘jacaT structure (Simpson, 1976: 156). Several small bowls no larger than one to two cm in depth and 5 cm in diameter were among them. All the retrieved bowls are reportedly quite fragile (Martin, personal communication, 2018a) and are currently housed at the Louden-Henritze Archeology Museum in Trinidad, CO. Although their function is unknown at the present time, they may have belonged to children because of their small size and fragility (Simpson, 1976: 156). Simpson (1976: 23) noted the presence of a pit and timbers from a collapsed structure. A layer of matting comprised of grasses, small juniper branches, and other


28
plant material was discovered at the shelter and may represent roof fall from several collapsed structures (Nowak and Gerhart, 2002: 9). The additional protection likely meant that the site was used during the winter. Much of the portable ground stone was uncovered under the remains of the collapsed structures. By the end of her 1974 field season, Simpson uncovered a minimum of four different cultural levels (Simpson, 1976). Simpson’s (1976) chronology suggests that the shelter was utilized on a regular basis from the Late Archaic Period (1050 BC -100 AD) to the Late Prehistoric stage (AD 100 - 1725). Simpson left behind a set of field notes and artifact inventories. Her master’s thesis is considered the first formal documentation of archaeological research conducted at the site and it contains the first known partial map of the site (Zier, 2017). In 2013, Centennial Archaeology, Inc. created a more thorough map of Trinchera Cave and its immediate surroundings using a Gowin TKS electronic total station (Zier, 2017).
Contemporary Researchers at Trinchera Cave Office of the State Archaeologist (1997 - 1999)
The Program for Avocational Archaeological Certification (PAAC) is an educational program designed for avocational and professional archaeologists alike. PAAC was first established in 1978 by the Office of the State Archaeologist of Colorado (OS AC) and the Colorado Archaeological Society (CAS). The program was designed to allow CAS members and those interested the opportunity to work at an actual archaeological site and gain expertise rather than pursuing an academic degree. Between 1997 and 1999, Kevin Black and CAS volunteers completed a comprehensive survey of nearly 650 acres in the Trinchera Cave Archaeological District (TCAD) through the PAAC program (Black, 2000). The volunteers


29
created a new map of the shelter, recording the different rock art panels that were present. 53 indigenous sites within the district were also documented.
Eighteen isolated finds and 57 sites were recorded for the first time, including Native American and non-Native American materials (Black, 2000). The non-Native American materials and historic sites are not discussed in this thesis and can be found in Black’s (2000) comprehensive report. The presence of ceramic artifacts, metal artifacts, and rock art (the latter being found solely at Trinchera Cave) suggests more recent use in the area by the Apache, Comanche, and/or Kiowa groups (Black, 2000). Human occupation in the TCAD appeared to be fairly intense. Very little research has focused on the function and relationship of Trinchera Cave to other regional sites.
Michael Nowak (1999 - 2001)
Michael Nowak was a faculty member at Colorado College (CC) when he conducted extensive excavation work at Trinchera Cave over the course of several years (Nowak and Budnick, 2000; Nowak and Gerhart, 2001, 2002). The primary goal of the project was to discover what else remained at the site (Nowak, 2009). Specifically, the purpose of Nowak’s research was to locate any unexcavated areas of the site, determine the stratigraphy prior to earlier excavations, and to obtain radiocarbon dates from site sediments (Nowak and Budnick, 2000).
Nowak and his team excavated blocks in Areas A, C, and D. Since Dick had excavated Area B to bedrock, there was presumably no need to dig there again (Zier, 2017: 29, 42).
Nowak conducted most of his work inside the shelter, similar to the earlier researchers.
However, he was the first to extend test units just beyond the dripline to help uncover any exterior deposits that may have been present (Zier, 2017). Additional units were established


30
below the shelter to help determine how the archaeological materials had moved downslope. Nowak elected to extend some units while abandoning others based on what was unearthed.
This work produced 12 radiocarbon dates from organic materials at the site (Zier, 2017) (See Table 3.1). Nowak uncovered evidence of a midden directly under the dripline. Besides Trinchera Cave, middens containing abundant and diverse cultural artifacts were only found at one other site in the region, suggesting that most site occupations were relatively short-lived (Black, 2000: 66). Many of the portable ground stone fragments were found in this midden and under the collapsed structures (Nowak and Gerhart, 2002).
Archaeologists frequently rely on stratigraphy to help reconstruct site history - before, during, and after site occupation. Nowak’s interpretations are reasonable, taking into account each of the previous excavations as well as heavy looting at the site. Chase noted that looting may have begun as early as 1949 (Gerhart 2001). According to the site foreman Buford Garcia (personal communication, 2018), more serious looting has been a problem at Trinchera Cave since the early 1970s. The looting got so bad that in the late 1990s a warning sign was posted on the northwest side of Trinchera Creek to deter future activity (Garcia, personal communication, 2018). The sign has since been removed by looters. More protection is certainly needed for this site. It is recommended that Trinchera Cave be periodically monitored with updated signage to help deter future looting and vandalism.
Colorado College (2002 - 2003)
After Nowak’s field seasons had wrapped up at the site, CC’s work continued. In 2002, CC student Kylie Crocket cataloged between 75 - 80% of Haldon Chase and Herbert Dick’s collections housed at TSJC (Crocket, 2002; Martin, personal communication, 2018a). In 2003, CC student Laura McCarthy and visiting assistant professor Carl-Georg (Charly) Bank carried


Table 3.1. Radiocarbon Dates from Trinchera Cave.
Table from Zier (2015: 52-53).
Lab No. (TSJCNo.) GreamzatioD Original Reference 5ample Description (Provenience) Conventional Age 2-Sigma Calibration Cultural Stage,1Period Comments,- Contest Interpretation
3eta-39S5S0 (TR1-D40E-7) Caitemaial Airtapnlngy? Inc- Partially burned port sample frou: probable structure; semes Jimiptnis - juniper .Area D, Lend I, Grid 4QE 000 -K- 30 BP Cal AD 005-1O5O (Cal 055-0O0 BP); Cal .AD I0S5-1125 (CalSdi —S25BP);Cal AD 1140 — 1150 (Cal BIO -200 BP) L ate Prehistoric Developmental. -Divers ificahon period boundary Excavated by Simpson
3eta-39S5Sl (Trl-D-45E-5} Centennial Archaeology, Inc. Partially burned post sample from.probable structure; gams 3uniper.l; —jutcinec Area D. Lend I, Grid 45E, depth 15" 040+7-30 BP Cal .AD 1020 -1155 (Cal 03O-7S5 BP) L ate Prehistoric Diversification paiod Excavated by Simpson
3eta-39S5S2 (Trl-D-40E-5) Centennial Archaeology. Inc. Pollen sample taken front fiiepil Area D. Grid40E. depth IS" 1550 -h'-SOEP Cal AD 420-575 (Cal 1530-1375 EP) Late Prehistoric DevelofHnentnl period Excavated by-Simpson
3eta-30S5S3 (Trl-D-KJE-IE TEU-D-40E-:a TR1-D-4QE-21) Centennial Ajchaeoiogy Inc. Three pollen samples taken from single column .Area D, Grid 40E, depth;: of 54” , 50” . and tS" 4d20 +'- 30 B? Cal 3500 - 3430 BC (Cal 5450-53 B0 BP}; Cal 33 SO — 3350 BC (Cal 5330-5300 BP) .Archaic. Early Archaic period Three sauries combined into a single sample for bating analysis: excavated by-Simpson
Beta-KX)732 (Trl-IM5F-ll) Catemial Archaeology, Inc. Tied yucca leases Area D. Grid 45F. depot: ■ 0 -33” JPre-LeveHY) 01O+1- 30 BP Cal .AD 1030-1210 (Cal 02O-740 BP) Late Prehistoric Drversifrcation per.od Excavated by-Simpson
Beta-400733 JTA202) [CC cat# 1O57.D.2090] Gafflanial Archaeology. Inc. Bone head fragment -Cottontail tibia shaft Compartment 1,220W - 2GN1 Depth 155 an 12SD+/-30BP Cal ADddo -775 (Cal 12*5-1175 BP) L ate Prehistoric Developnental period Excavated by Chase


Table 3.1 cont. Radiocarbon Dates from Trinchera Cave.
Table from Zier (2015: 52-53).
Lab No. (TSJC'No.) Organization OripTinl Reference Sample Description (Provenience) Conventional Age 2-Sigma Calibration Cultural St age,‘Period Comments C cofcxt Interpretation
B*a-Wa734 (TR.l-C-936) [CCcat# 1057.0.3-010] Centennial Archaeology, Inc. 3 one bead - small mammal Long bone fra^uant Shelter C. Grid 5A depth 72 - 7fT 1530-w-SOB? Cal AD 3S0 -435 (Cal 1570-1515 BP); Cal AD 450 - 455 (Cal 1490-I4E5 BP); Cal AD 490 - 535 (Cal 1450- 1415 BP) Late Prehistoric Developmental period Excavated by Diet
Beta-KXPss (TR1-A-711) [CC cat# 1057.D.277C] Centennial Archaeology, Inc Tied yucca Leaves 3heller A, Grid 4A-B, Depth 24 - 3ffn 900 -H- SO BP Cal .AD 1035-1215 (Cal 915-735 BP) Late FTehistorir Diversification per.od Excavated by Diet
Beta-400737 (TR.I-B-339) [CC cat = ID57.0.2SB0] Centennial Archaeology, Lac S-trust, Z-piy cordage made from probable PoacitK (Gras; family) stem Shelter B. Grid iA-2A-3A, Depth IS-24" 119D +f- 30 BP Cal AD 725 - 740 (Cal 1225-1210 BP); Cal AD 770 - S95 (Cal 11 SO- 1055 BP); Cal AD 925 - 940 (Cal 1025-1010 BP) Late Prehistoric Developmental period Excavated by Diet
Beta-40G73S fna-A-sw) [CC cat# 1057.0.2CS3] Centennial Archaeology, Inc Three beans [Pbaseolns vulgaris] -one Anasazi beau and rivo common beans, freon small sn-cca-ried grass pouch Shelter A Chid 6B. in pad: rat nest depth IT’ 790 -b'- SO BP Cal .AD 1210-1275 (Cal 740-575 BP) Late Prehistoric Driersificatian per.od Excavated by Diet
3eta-39S5S+ (TRI-C-910) Centennial Archaeology, Inc. Soil sample Shelter C, Grid 3A Depth 75 - £4" 9250+,'-40 B? Cal S610- £320 BC (Cal 10,550-10,270 BP) Paleoindtan Folsom period Excavated by Diet
Beta-39S5S5 (None) Centennial Archaeology, Inc Soil sample Shelter C, Gnd 7E. Depth 90 -11CT 9730+-40 BP Cal 9275 - 9150 BC (Cal 11,225- 11.110 BP) Paleoindtan Clovis period Excavated by Diet


33
out a preliminary geophysical study with remote sensing at Trinchera Cave. Their primary goal was to locate the foundation walls of the jacal structure Simpson (1976: 23) had mentioned in her thesis. They conducted seismic, geodetic, magnetic and resistivity surveys (McCarthy and Bank, 2003). Although recent excavations failed to turn up any evidence of the structure, resistivity profiling provided the most reliable results (McCarthy and Bank, 2003). A structure was located in the southwest portion of the cave which they interpreted as the foundation walls (McCarthy and Bank, 2003).
Christian J. Zier (2013 - 2015)
Between 2013 and 2015, Chris Zier of Centennial Archaeology, Inc. (CAI) created a more comprehensive map of Trinchera Cave. Using available sources, Zier and CAI reconstructed the locations of previously worked areas in the shelter. These sources included field notes, early newspaper clippings, an incomplete catalog of previously excavated artifacts, markings on the rear wall of the shelter, Simpson’s (1976) master’s thesis, a personal interview (with Chase), photographs, profile sketches and original site map drawings, and various published reports (Zier, 2015, 2017). Zier (2015) assessed the archaeological potential of the different collections from the site that are housed at the Louden-Henritze Archaeology Museum. Recent radiocarbon dates from the Trinchera Cave collections yielded two Paleoindian/Folsom period dates of 9275 - 9160 BC (Cal. 11,225 - 11,110 BP) and 8610 - 8320 BC (Cal. 10,560 -10,270 BP) (Zier, 2015: 54 - 55) (See Table 3.1). Both deep soil samples, 90 - 110” and 78 -84” respectively, were originally excavated by Dick. These analyses potentially indicate that the artifacts uncovered at those levels also date to the Paleoindian/Folsom period.
However, it is possible that these radiocarbon dates may have been affected by “old wood.” Most of the wood located in the Trinchera Cave Archaeological District is Juniperus sp.


34
(juniper) (Black, 2000; Nowak and Gerhart, 2002; Zier, personal communication, 2018). Zier (personal communication, 2018) suggests that the actual dates could be much younger than the radiocarbon assays. Samples often provide misleading results if materials of different ages accumulate in the same archaeological context.


35
CHAPTER IV
GROUND STONE TECHNOLOGY
Ground stone tools represent an important aspect of the archaeological record and can contribute much to our understanding of past lifeways. They can offer insights into areas such as diet, division of labor, food processing techniques, mobility patterns, settlement systems, and other specialized activities (Ebeling and Rowan, 2004). These grinding tools are used to process food items such as amaranth seeds, beans, cacao, chiles, maize kernels, squash seeds, sunflower seeds, and tomatoes as well as clay, pigments, and temper (Searcy, 2011: 5). They can also be used to process items such as ceremonial and medicinal herbs. Ground stone tools can come in different shapes and sizes. Different methods have been developed to help determine the specific uses of these implements.
Portable Ground Stone
“Ground stone” is the broad term commonly used by archaeologists to refer to stone artifacts that are modified by or used to modify other materials through different reduction techniques such as abrasion, pecking, or polishing (Adams, 2014a). This includes abraders, hammerstones, manos and metates, mortars and pestles, and polishers as well as the variety of artifacts that are shaped using these techniques. Food-grinding tools have played a vital role in the description of prehistoric subsistence strategies (Adams, 1999). Experimental work carried out by Jenny Adams (2014a) on different metate types similar to those found in the archaeological record of the American Southwest provides a model for understanding the tool design, manufacture, and use of metates all over the world. Adams (2014a: 69) argued that vesicular materials provide a much better working surface on which to grind large-grained food resources because it shears apart the grain while leaving little gravel behind during processing.


36
Vesicular basalt is one of the most popular choice for grinding seeds and dried corn kernels today (Searcy, 2011: 83).
Adam’s (2014a) research also investigated the efficiency of different tool designs. She looked at different metate types based on those discovered in the southwest; including basin, concave/flat, and trough metates. Her research demonstrated that concave/flat and trough metates were each effective in processing-soaked com kernels (Adams, 2014a: 68 - 69). Adams (2014a: 106) also found that trough metates could grind more grain at a time than concave and flat metates. Looking at a number of ethnographic studies on the relationship of mano size and agricultural dependence, Hard (1990) found that manos with an increased surface area could process more grain than smaller designs. Diehl (1996: 106) combined macrobotanical and ground stone data to support this relationship.
The processing of corn and other large-grained food resources became more important over time, resulting in changes to ground stone tool morphology (Stone, 1994). Grinding corn with manos and metates is very time-consuming (Cushing, 1920; Hard, 1986; Lancaster, 1983; Spier, 1933: 79). Ethnographic studies have shown that efficiency is improved by increasing the number of ground stone tools used for reciprocal grinding and the size of the grinding surfaces (Hard, 1986, 1990; Horsfall, 1987; Mauldin, 1991, 1993). A strong correlation exists between the length of the grinding surface, as indicated by mano length, and the importance of agriculture to the subsistence economies of modern groups (Hard, 1986, 1990; Mauldin, 1991, 1993). By increasing the size of the grinding surface, processing speed increases and it allows for more grain or meal to be processed at a time (Mauldin, 1993). Hard (1990) identified a strong correlation between mano length and the level of agricultural dependence. His study indicated that manos smaller than 19 cm often have a low reliance on agriculture, and those between 17


37
and 25 cm in length tend to have a moderate reliance on agriculture (Hard, 1990; Stone, 1994). Manos that range between 24 and 31 cm on average generally have a high reliance on agriculture (Hard, 1990; Stone, 1994).
In Southeastern Colorado, other researchers have identified two types of manos: the one-handed mano (less than 18 cm), and the two-handed mano (with a length greater than 18 cm) (Campbell, 1969: 56; Ireland, 1974: 83; Simpson, 1976: 105). It is believed that the larger two-handed manos may be more prevalent in the region after AD 1050 (Nowak and Gerhart, 2002: 127). As corn became more important, Lancaster (1983) argues that food grinding tools became more specialized. Experimental research led by Adams (1993b) indicated that ground stone tool morphology can be used to draw conclusions about prehistoric grinding efficiency. In areas of the southwest that were heavily dependent upon corn, there was an increase in the number of flat and trough metates (Adams, 1993b) and an increase in size of individual grinding surfaces as indicated by mano length (Hard, 1986, 1990; Mauldin, 1991, 1993). Using appropriate types of raw material is shown to be just as important as tool morphology and grinding surface size for those dependent upon agriculture.
Early attempts were made to identify specific tool types that can differentiate between the processing of wild foods and domesticated foods and to use those differences to determine when agriculture was first introduced to a particular region (Adams, 1999). Variability in tool size was once assumed to have been associated with the processing differences between wild, cultivated, and domesticated foods (Adams, 1999). However, research has shown that tool morphology is more closely associated with different processing techniques than to resource procurement (Adams, 1999). Using ethnographic data, Wright (1994: 242) argued that
“tool morphology is not a reliable guide to highly specific functions or dietary emphasis, but it can suggest overall strategies offood processing... ”


38
The different types of stone tools used in food processing do not sufficiently reflect the range of food resources being processed (Adams, 2014a: 55). The same food item can be processed using more than one type of tool, and the same tool can be used to process different food types. Tool morphology cannot be used to establish what type of food resource was being ground (Adams, 1999; Horsfall, 1987; Wright, 1994). Pollen and starch analyses performed on prehistoric ground stone strengthens this argument (Greenwald, 1993: 348 - 349; Lancaster, 1984: 257).
The study of ground stone technology includes not only the tools themselves, but also the activities, behaviors, knowledge, and social contexts related to their manufacture, use, and abandonment (Adams, 1993b, 1995a; Bleed, 1986; Dobres and Hoffman, 1994; Kingery, 1989; Lemonnier, 1986; Nelson, 1996; Schiffer, 1992; Schiffer and Skibo, 1987). Technological development can be affected by new information on tool efficiency or design, newly discovered resources to use on existing tools, new methods to process the same resources, and raw material availability (Adams, 1993b: 332).
Bedrock Ground Stone
Indigenous groups have occupied the tributaries and canyon sidewalls of the Purgatoire River in southeastern Colorado for several thousand years. Archaeological evidence reveals habitation by horticulturalists, Plains hunter-gatherers, and Early Woodland groups (Black, 2000; Lynch, 2010). A distinctive feature of this region is the presence of numerous bedrock grinding areas appearing at different elevations and with different site assemblages. Bedrock grinding areas represent an important part of the native landscape in northeastern New Mexico and southeastern Colorado. Documented in site descriptions since Renaud (1931), the highest concentration seems to be located along the canyon sidewalls of the Apishapa, Chacuaco,


39
Cuchara, and Purgatoire Rivers and their tributaries (Campbell, 1969; Loendorf, 2008; Lynch, 2010; Owens, 2007). Bedrock Grinding Surfaces (BGS) were often constructed along the sandstone canyon walls, though others have been discovered among rock outcrops (Andrefsky, 1990; Campbell, 1969; Chomko and De Vore, 1990; Gunnerson et al., 1989; Hartley and Vawser, 2003; Loendorf, 2008; Renaud, 1931).
Archaeology in southeastern Colorado suggests an array of lifeways and subsistence practices based on the remains of architecture, ceramics, grinding tools and lithics (Campbell, 1969; Eighmy, 1984; Zier et al., 1999). However, BGS are difficult to put into an archaeological context (Campbell, 1969; Eighmy, 1984; Zier et al., 1999). There is little ethnographic and archaeological evidence of their use in the region. Unlike their portable counterparts, BGS features are a permanent part of the landscape (Lynch, 2010). This leaves them unprotected and susceptible to a variety of erosional forces (Walker and Fratt, 1991).
There are currently several ways to directly date bedrock grinding surfaces and to determine functionality through pollen or starch analysis (Cummings, personal communication, 2018; Lynch, 2010). Dating ground stone pulled from the archaeological record can be accomplished through direct study, or may be deduced through association with nearby features such as hearths, rock art, and structures (Campbell, 1969; Chomko and De Vore, 1990;
Loendorf, 2008). Dating also can be determined using diagnostic materials found in associative context. The small number of bedrock grinding surfaces that have been dated in the region through radiocarbon methods or by associative context with other features are believed to date from the Late Archaic Period (2000 RCYBP - RCY500 BP) (Campbell, 1969; Chomko and De Vore, 1990; Loendorf, 2008; Lynch, 2010; Owens, 2007). There are no historical records about the use of these grinding features in the southeastern Colorado, and research suggests that they


40
stopped being used by the time Athapaskan-speaking peoples began filtering into the region during the late 14th to early 15th century (Campbell, 1969; Gunnerson et al., 1989; Hammond and Rey, 1940; Owens, 2007; Stoffle et al., 1984; Zier et al., 1999).
Grinding Behaviors
Design and Manufacture
The first step to understanding technological processes is to consider how an object was designed and manufactured, why the specific raw material was chosen, and what modifications were made to prepare the selected material for the object’s intended function (Adams, 2014a:
21). The design stage begins with choosing a material of suitable size and texture (Adams et al., 2009: 44 - 46; Delgado-Raack et al., 2009; Horsfall, 1987: 340). Raw material variability includes rock size and granularity. Comfort features, such as grooves or handles, can make a tool easier to use and are often included in the manufacturing process (Adams, 2014a: 21). Raw materials can be selected from a variety of different locations including the landscape, a nearby streambed, or from bedrock outcrops. Material size and weight are common selection criteria for certain tools. However, Stone (1994) notes that size choices are often restricted by resource availability for portable grinding stones.
Raw materials have a natural granularity (fine-grain to course-grain) to them that is perceived as texture (smooth, rough, and so forth) (Adams, 2014a: 21). Texture is a critical attribute to consider while manufacturing tools. A fine-grain rock and a course-grain rock may be chosen for different processing tasks because of their respective granularity. A fine-grained rock can be used as a polisher. Meanwhile, a course-grained rock can be fashioned into an abrader since it is rough enough to break down the surface material of another object (Adams, 2014a: 21). The abrader’s texture becomes smoother through continued use. However, the


41
initial rougher texture can be restored through certain restorative techniques such as pecking (Adams, 2014a). Durability is another important factor to consider when selecting raw material type and refers to the ability of the material to withstand wear (Adams, 2014a). Especially when grinding food, it is important that the grinding tool be durable enough to not erode into the food.
The behavioral constructs of tool design can be thought of in terms of socioecological systems. A tool is considered to have an expedient design if the natural shape of the raw material was modified through use or just enough shaping to become functional (Adams, 2014a: 22). Modifications that make a tool easier to handle, improve efficiency, or achieve a shape not related to function suggest a strategic design (Adams, 2014a: 22). By looking at tool design for an assemblage as a whole, archaeologists can often determine whether expediently designed objects are used or regarded differently from those of strategic design.
Use
The way a tool was designed, manufactured, and utilized can be assessed in terms primary and secondary uses (Adams, 1994, 1995b, 2014a: 24). Primary use is the task for which the tool was first intended. Portable ground stone tools often serve a single function (Adams, 2014a). Secondary use refers to an additional function on top of a tool’s primary use. Minimal research into the nature of BGS has been conducted. They may have served multiple functions. Adams (1994, 1995b, 2014a, 2014b) defined several use categories that a tool may go through - single use, reused, redesigned, multiple use, and recycled for portable ground stone that can inform research into BGS.
Grinding efficiency is another important concept from portable ground stone studies that can be applied to BGS analysis. It is typically expressed as the amount of material that can be processed or the number of hours spent completing a specific task like food processing (Adams,


42
2014a, 2014b). Different tool designs may be more or less efficient than one another. For grinding tools in particular, efficiency can be measured by the size of the grinding surface and the overall weight of the tools (Adams, 2014a: 30; Hard, 1990). Like tool design, efficiency is a behavioral construct. Certain attributes of ground stone tools might make one more efficient than another. Hard et al. (1996) discusses the size of grinding surface as it relates to the amount of food processed. Studies have shown that by increasing the size of grinding surfaces, grinding efficiency increases as well (Buonasera, 2015; Hard et al., 1996; Mauldin, 1993). The concepts of use intensity, use efficiency, and wear- management strategies can be used to interpret the different technological processes of designing, manufacturing, and maintaining tool efficiency when viewed together (Adams, 2014a: 30, 2014b: 136- 137).
Ground stone tools also have ritual or ceremonial significance, especially when associated with inhumations (Adams, 2014a: 27; Carle, 1941). Those deposited with burials are not usually modified, but those placed into funeral pyres show distinct patterns of fire-cracking. Ground stone tools found near burials were often associated with food-processing activities (Adams, 2014a: 27), illustrating their possible significance to the user. As noted above, women have played a vital role in assigning meaning to rock shelters in many cultures. This is illustrated by the fact that in the American Midsouth, female burials are believed to outnumber male burials in sheltered spaces (Carmody and Hollenbach, 2013; Homsey-Messer, 2010). Burials are used to form ancestral ties to the past (Johnson and Earle, 2000). Carmody and Hollenbach (2013: 50) note that burials may have also been used by the local inhabitants to “mark their claim to a territory and usufruct rights to nearby resource patches. ” Hollenbach (2009: 215) suggests that bedrock mortars may have belonged to a specific family, serving a similar purpose. At Trinchera Cave in Trinidad, Colorado, the remains of a young


43
adult female were uncovered (Simpson, 1976: 198). There are currently no other burials known to exist in the Trinchera Cave Archaeological District (Black, 2000). Human dentition was also discovered at the site and they are known to have been deciduous teeth (Nowak and Gerhart, 2002: 262). It is unknown at this time whether they belonged to male or female children. Use-Wear Analysis
Use-wear analysis is an archaeological method used to determine tool function by examining their work surfaces. Close examination may allow researchers to understand how tools are altered during use (Adams, 1988, 1989a, 1989b, 1993a, 2014b; Adams et al. 2009). When researching ground stone technology, four different mechanisms typically are used to describe and recognize how specific damage patterns are formed - abrasive wear, adhesive wear, surface fatigue, and tribochemical wear (Adams, 2014a, 2014b). Each of the mechanisms interact with one another when two surfaces come into contact. Use wear on certain items should be compared against an unused area of the stone to help determine the stone’s natural state (Adams, 2014a, 2014b). Surface wear results from the motion of one stone item coming into direct contact with another (Adams, 1988: 310,1993a: 63,2014a: 28,2014b: 130; Czichos and Dowson, 1978: 98; Szeri, 1980: 35; Teer and Amell, 1975: 94).
Wear
Wear is the gradual loss of surface material of a stone item as a consequence of the motion between it and a secondary contact surface (Adams, 1988: 310,1993a: 63,2014a: 28, 2014b: 130; Czichos and Dowson, 1978: 98; Szeri, 1980: 35; Teer and Arnell, 1975: 94). The amount of wear an item sustains is discernible using qualitative variables. Items that are unused may have sustained surface damage as a part of the design and manufacturing process, but there is no discernible damage from actual use. Light wear leaves behind very little evidence that can


44
be seen with the naked eye (Adams, 2014a: 28,2014b: 136). Conversely, moderate wear leaves noticeable damage, it does not actually change the basic tool shape. (Adams, 2014a: 28, 2014b: 136). Finally, heavy wear alters the natural shape of the tool itself (Adams, 2014a: 28, 2014b: 136).
Wear-management strategies are recognized for both manos and metates. As a mano is being used, it becomes less efficient and harder to handle. Users developed certain wear-management strategies to help them cope with the wear-produced inefficiencies (Adams, 2014a: 28 - 30, 2014b: 136). If a rougher texture is needed, pecking the individual grinding surface can help accomplish that. Another method involves turning the mano over and using the opposite side as a secondary grinding surface. This way the user does not have to stop quite as often with two usable surfaces. A third strategy is to lift the distal edge of the mano, while pushing down with the proximal edge so that only part of the surface is ever in contact with the metate (Adams, 2014a: 29). On the return stroke, the user tilts the distal edge of the mano forward while lifting the proximal edge so a secondary surface is being used during the grinding process (Figure 4.1).
The intensity of tool usage can be difficult to interpret in the absence of a time frame (Adams, 2014a). Intensity can be assessed in terms of intensive and extensive uses. An intensively used tool is one that it is utilized for a longer period of time. For example, four hours a day for 30 days, resulting in 120 hours of wear. An extensively used tool is one that is utilized for a shorter period of time at any one sitting; one hour a day for 120 days, also resulting in 120 hours of wear. From an analytical standpoint, the amount of wear looks to be the same -moderate. One way to distinguish between intensive and extensive use in the archaeological record is through an evaluation of tool design (Adams, 2014a). If comfort features are present on a moderately worn tool, such as grooves or handles, it was designed to be held comfortably for


45
Downward Stroke
Figure 4.1. Grinding strokes and the subsequent wearing of two adjacent surfaces against a metate surface. Figure from Adams (1993b: 335). Redrawn by Ron Beckwith, based on
Figure 8 from Barlett (1933).
longer periods of time (Adams, 2014a). A moderately worn tool that does not have any comfort features may not have been designed for extensive use (Adams, 2014a).
Wear Rates and Patterns
Mona Wright (1993) made an effort to quantify the experimental wear rates of grinding tools in order to develop a model for prehistoric wear rates. This model allows researchers to determine the life history and abandonment processes of grinding tools (Adams, 2014a: 69 -70). Using replicated manos and metates, Wright (1993) concluded that certain raw material types do wear at quantifiable rates and that manos wear down much faster than metates. She also discovered that differences in individual grinding skill is an important factor to consider in wear rates. However, when Wright (1993: 353) compared her results to prehistoric wear rates she found that it was not possible


46
“to determine the initial weight and thickness of prehistoric manos and metates before they were used to grind maize... ”
Jenny Adams (2014a, 2014b) looked at time and use wear with a large number of experiments. She examined the use wear patterns of different contact surfaces - stone-against-stone contact, stone-against-wood contact, stone-against-bone contact, and stone-against-hide contact. Using replicated stone tools as a control, she questioned whether it was possible to distinguish the damage patterns caused by different intermediate substances such as amaranth seeds, clay, maize kernels, and sunflower seeds. She determined that the amount of wear an item sustains is a contributing factor. On lightly used surfaces, it was not possible to differentiate between the different substances. However, it is possible to recognize distinctive use-wear patterns on moderately and heavily used tools (Adams, 2014a: 36 - 44).
Resources
Raw material availability directly affects ground stone tool design, manufacture, use, and abandonment patterns (Stone, 1994). Early ground stone tool designs were based on notions of optimization and energy efficiency (Hard, 1986, 1990; Lancaster, 1983; Mauldin, 1991, 1993). Stone (1994) argues that modem studies using archaeological assemblages to reconstruct mobility and subsistence practices should take raw material availability into account. Natural topography may affect the procurement of certain raw materials and ground stone size (Stone, 1994). Differences in mano and metate size may be a direct result of new or greater access to larger raw materials in the region. Lithic studies have shown that abundance, characteristics, and spatial distribution of raw materials can affect the design and production of stone tool technology (Andrefsky, 1994, 2005; Bamforth, 1986; Barton, 1988, 1990; Gould, 1980; Odell, 1977, 1989). Gould (1980) studied how the abundance of raw materials had an impact on stone tool production for the Aborigines of Australia. He discovered that when certain raw materials were


47
available near the habitation base-camp, the Aborigines used those raw materials in the
production of different tool types, including expediently and strategically designed tools
(Andrefsky, 2005: 235). Gould (1980: 134) noted that
“Whenever random factors of geography play sources of usable stone, whether in the form of quarries or nonlocalized in nature, at or in close proximity to a water source where a habitation base-camp will occur, ease of procurement will outweigh other factors and unusually high percentages of artifacts of these locally available stones will be made, used, and discarded at such campsite. ”
Other researchers have studied raw material abundance in relation to transportation costs (Kuhn, 1991) and land-use and mobility patterns (Daniel, 2001; Dobosi, 1991; Goodyear, 1993; Meltzer, 1985; Seeman, 1994; Wiant and Hassen, 1985). Raw material choice is also affected by the processing constraints of different food resources (Greenwald, 1990; Horsfall, 1987). The material being ground should be hard and durable enough to effectively process food resources without the material wearing down too fast. The grinder should also make sure that sand or gravel is not being added to the food being processed. Furthermore, variations in grain, nut, or seed size and durability place processing constraints on portable and bedrock ground stone technology (Hard, 1986; Lancaster, 1983).
Translating Portable Ground Studies to Bedrock Ground Stone Differences in the design and manufacturing process are separate technological traditions that could reflect distinct methods of constructing and operating food-processing equipment (Adams, 2014b: 131). Technological knowledge is transferred by the migration of individuals and/or small groups who then continued their learned traditions in relocated areas (Adams,
2010). Increased frequencies of a specific metate design in a region previously dominated by a different type of design may represent the migration of a larger social unit. It may be possible to recognize the natural development of permanent grinding stations. The Ancestral Puebloans and


48
Mogollon, for example, developed two different metate designs based on the concept of permanent grinding stations to better confine the product during processing (Adams, 2014b:
131). Both traditions managed to increase the number of grinders that could work together. Ground stone technology varies across time and space not only in the American Southwest, but in southeastern Colorado as well. Trinchera Cave represents a combination of Plains and Southwestern cultures (Simpson, 1976), and may have been a melting pot of peoples (Nowak and Gerhart, 2002: 104). Technical knowledge may have been transferred during migration into the area.


49
CHAPTER V METHODOLOGY Introduction
Missing from contemporary research for the region is an understanding of how portable and bedrock ground stone are connected to regional hunter-gatherer adaptation and if these relationships changed through time. Bedrock ground stone represent an important aspect of human behavior, landscape adaptation, mobility and subsistence strategies, and symbolic ideology (Lynch, 2017b).
Data Limitations
Inter-observer Variability
Archaeologists that document bedrock grinding surfaces in the field typically record the length (longest axis), width (second longest axis), and basin depth as standard units of measurement (Lynch et al., 2012). The amount of training and field experience recording bedrock grinding surfaces varies greatly among archaeologists. Irrespective of training, individuals will have somewhat different interpretations of grinding surface length, width, and depth. With this level of inter-observer variability, any errors in measurement have the potential to be problematic for comparative studies (Lynch et al., 2012). A certain level of caution is suggested when measuring bedrock grinding surfaces as well as using measurement data in the archaeological record.
Other Limitations
Time and budget constraints prohibited the ability to analyze every piece of portable ground stone housed at the Louden-Henritze Archaeology Museum. Variations in the types of measuring devices and instrument accuracy in the field can add to differences in observer


50
measurements. Differences in length, width, and depth can result from instrument variation (Lynch et al., 2012). A total of 211 items were selected to examine the nature of the ground stone activities that occurred at Trinchera Cave.
Sample Selection
A sample of (n = 179) portable ground stone recovered from Trinchera Cave in southeastern Colorado is used in this study. This sample is the accumulation of multiple archaeological digs from the Trinchera Cave site and includes (n = 150) manos and (n = 29) metates. The metates are represented by (n = 22) slab metates, (n = 6) basin metates, and (n = 1) trough metate. This sample includes portable ground stone housed at the Louden-Henritze Archaeology Museum at TSJC in Trinidad, Colorado. Each of the portable materials from this data set were measured using plastic Ohaus calipers and a Kobalt tape measure. Identification of the ground stone materials was done by Herbert Dick, Caryl Wood Simpson, and Michael Nowak. Most of the materials recovered from Haldon Chase’s and Herbert Dick’s excavations were cataloged into the Colorado College database by Kylie Crocket, with help from Michael Nowak, Heather Gerhart, and Loretta Martin (Nowak and Gerhart, 2002: 99). Colorado College maintains a 75 - 80% digital inventory of the materials recovered from the site. All of the materials, field notes, and analysis sheets are currently housed at TSJC.
The portable manos and metates recovered by Herbert Dick and Caryl Wood Simpson were recorded from May 13th through May 18th, 2018 using artifact analysis sheets. No materials from Haldon Chase and the High Plains Columbia Expedition are curated at TSJC. Most of the materials from their joint 1949 expedition were reportedly shipped to Columbia University, and were later transferred to the University of Denver in 1955 (Zier, 2017: 26). However, after contacting the anthropology departments at both Columbia University and the University of


51
Denver, it was learned that any materials collected from Trinchera Cave are not currently being housed at either location. Martin (personal communication, 2018a) also has no idea of their whereabouts. Lintz (1999: 20) believes that some of the larger ground stone items may have initially been left behind following Chase’s excavation. This seems reasonable due to the weight of some of the metates currently being housed at the Louden-Henritze Archaeology Museum.
No portable ground stone from Michael Nowak and Colorado College’s excavations (1999 - 2001) were recorded while at the museum. Nowak and his team excavated trenches and test units in all areas of the shelter, except for one (Zier, 2017: 29; Martin, personal communication, 2018a). Between 1954 and 1957, Herbert Dick excavated all the way down to bedrock at Area B (Zier, 2017: 29), so re-excavating there was not necessary. Nowak and Gerhart (2002: 8) approached Trinchera Cave primarily as a salvage project. All of the materials recovered from Nowak’s excavations did not contain provenience information. This may be due in part to Nowak’s team having recovered a significant amount of the materials from backfill (Martin, personal communication, 2018a). The extent of bioturbation is unclear. Nowak (2002: 8) states that
“In most situations, we don’t know where the undisturbed/disturbedfill boundaries are.
Fill uniformity and the paucity of good stratigraphic markers contribute to the problem;
the lack of earlier field notes creates most of it... ”
Due to the lack of provenience information, this collection could not inform us on the research question regarding changes through time. Even without the provenience information, his materials were examined for qualitative purposes. There are several boxes of mano and metate fragments present at the museum. The different raw material types are consistent with the portable ground stone that was provenienced from the earlier excavations: Dakota sandstone, granite, conglomerate, and vesicular basalt. Additional ground stone materials were found to be


52
manufactured from argillite, mudstone, and siltstone suggesting they may have been used for activities other than food processing. All of these materials are local. In other words, the raw material used for the unprovenienced ground stone is similar to that found in the other collections. Since these materials have no context within the shelter, they are excluded from the analysis.
A Temporary Access Permit (No. 111861) for Trinchera Cave was approved by the Colorado State Land Board on May 31st, 2018. The authorized dates for temporary access were June 22nd - June 24th, 2018 or June 29th - July 1st, 2018. The site is located in section 16, range 59 west, and township 33 south in the sixth prime meridian (Black, 2000: v). Section 16 is currently being leased from the State of Colorado by the Baca Revocable Trust (Baca, personal communication, 2018). The owner is Philip (Phil) Baca and he primarily resides in Tucson, Arizona. The Baca property resides on roughly 17,000 acres of deeded land. After speaking with Baca about gaining access to the site, he approved both proposed dates. He asked that any future communication go through his foreman, Buford Garcia, when time came to visit the site. The first set of dates were chosen.
The second data set represents the (n = 32) bedrock grinding surfaces present at Trinchera Cave. They were recorded between June 22nd and June 24th, 2018. Surface dimensions (length, width, and depth) were measured once by each analyst to try and accommodate for interobserver error (Lynch et al., 2012). Laura Baker assisted in the field by collecting the secondary set of measurements of the bedrock grinding surfaces. A Kobalt tape measure and a 16” x 24” Aluminum Professional Square from Empire were used while at Trinchera Cave. Artifact form, raw material, and use wear were also recorded. The individual variables are detailed below.


53
IBM SPSS Statistical Software (V. 25) is used to record and manage the two data sets. The software is available for use at the University of Colorado Denver.
Pollen Analysis
While recording the portable ground stone at the Louden-Henritze Archaeology Museum, it was decided that a pollen, starch, and XRF analysis should be conducted on one of the artifacts. A pollen analysis can help provide archaeologists with information on the plants and other resources processed on ground stone. Pollen and starches have been found to accumulate on ground stone surfaces through repeated use (Cummings and Varney, 2009). With analysis, we may be able to determine the intermediate substances that were being processed by grinding tools. Inadvertent pollen deposits like pollen rain or wind-blown pollen may have occurred in the region. Data from pollen samples are frequently examined in order to distinguish foodprocessing activities against a backdrop of naturally occurring pollen rain (Brush, 2001; Bryant and Holloway, 1983; Pearsall, 2000). Archaeologists often strive to understand how environmental change affected hunter-gatherer behavior, and what socioeconomic factors influenced how people adapted to these new conditions. Pollen profiles are especially useful as they record the ways people may have shaped their landscape.
Pollen related to food procurement and processing is generally transported by people, while airborne and/or water transported pollen is frequently non-food in nature (Mercuri et al., 2010). Windblown pollen is often found at archaeological sites and we use it for environmental reconstruction. Researchers want to know changes in forest make up to understand long-term environmental change. Windblown pollen may not affect what we think is going on with food, but it is definitely important in understanding landscape. Pollen evidence of food at archaeological sites is predominantly from local instead of regional sources (Davis, 1994;


54
Pearsall, 2000). Reconstructing prehistoric diet often relies on the belief that pollen mainly comes from the resource patch visited and exploited for subsistence purposes. Mercuri et al. (2006) refers to this patch as the area of influence for each archaeological site. Breeding, cultivation, exploitation, and settlement patterns all affect the environment resulting in changes to pollen diagrams (Faegri et al., 1989). Variables that are associated with different activities, such as trade or cultivation, become part of the pollen spectra deposited directly on tools because they more accurately reflect human behavior than the environment (Kelso and Good, 1995; Li et al., 2008; Mercuri, 2008).
Plant taxa are identified by quantity and condition of mineral elements by provenience (Cummings and Varney, 2018). The condition of mineral elements is important to matters of possible contamination. Rodents and insects are both known to harvest and store seeds (Brown and Davidson, 1977; Chaney, 1936), resulting in the post-occupational deposit of seeds into different archaeological settings. The presence of rodent pellets and insect remains in soil samples can provide an indication of contamination potential. Carbonization, or charring, is a means of preserving prehistoric animal remains and plant material (Gleichman, 2002). Plant material that is uncharred may or may not be prehistoric.
Prehistorically introduced seeds preserved through burning are not automatically indicative of culturally used plants. They may be naturally dispersed seeds that were inadvertently burned. There are many ways seeds can be burned and integrated into the archaeological record. Burned seeds representing eight different taxa were identified from soil samples at Trinchera Cave (Gleichman, 2002: 95): Celtis reticulata (hackberry), Chenopodium sp. (goosefoot), Helianthus sp. (sunflower), Juniperus sp. (juniper), Portulaca sp. (purslane), Oryzopsis hymenoides (rice grass), and Scirpus sp. (bulrush). Burned seeds from the grass


55
Figure 5.1. Different views of mano used for analysis, family (Poaceae) were also identified. All of these plants are considered edible. When
compared with the material culture recovered from Trinchera Cave, archaeobotanical remains,
phytoliths, pollen, and starches can become potential use indicators (Cummings and Varney,
2018: 2).
After careful consideration, a small, one-handed sandstone mano (10.3 cm in length x 6.7 cm in width x 4.8 cm in thickness) from Herbert Dick’s excavation was selected for analysis (Figure 5.1). This specific artifact was selected for its unique triangular shape, the presence of


56
blue pigment, and several small black hairs embedded on one of the grinding surfaces. Tool morphology cannot be used to determine what was ground (Adams, 1999; Horsfall, 1987;
Wright, 1993). In order to understand the functional differences between portable ground stone and bedrock ground stone, we need to establish what the inhabitants of Trinchera Cave were grinding. This is best achieved through archaeobotanical analysis. The specimen bag that the mano came in initially had three different artifacts in it. Each of the three items were bagged separately and placed into the larger specimen bag. All three were previously given the catalog number Tr. 1 - 932. After communicating with the museum director, Loretta Martin, we separated the artifacts into three distinct bags. The artifacts were labeled as Tr. 1 - 932a, Tr. 1 -932b, and Tr. 1 - 932c. The catalog number for the selected mano became Tr. 1 - 932a.
Artifact Analysis
Prior to taking the artifact to PaleoResearch Institute for an archaeobotanical analysis, it was photographed and examined under a microscope. Initial photographs of the mano were taken using a Cannon PowerShot SX530 HS. It was later examined in the archaeology lab at Metropolitan State University using a Leica EZ4 W Microscope. The low staging area allowed for the mano to be fully placed under the objective lenses. Pictures of the artifact were taken using the microscope at 15X - 20X magnification, providing high resolution images (Figure 5.2). After microscopic examination was complete, the mano was rebagged and transported to the nearby lab.
Archaeobotanical Analysis
Using funds awarded from the 2018 Karen S. Greiner Endowment, a pollen, starch, and XRF analysis was conducted on the sandstone mano by PaleoResearch Institute, Inc. (PRI) in Golden, Colorado. Palynologist and paleoecologist, R. A. (Robert) Varney, processed the mano using standard lab protocols. Specifically, all visible dirt was first removed with regular tap


57
Figure 5.2. Photos of mano taken at 15X - 20X magnification, water and gentle hand pressure to eliminate any modem contaminants. A small piece of the
grinding surface was tested using a 10% dilution of hydrochloric acid (HC1) to help detect the
presence of any calcium carbonates (Cummings and Vamey, 2018). These carbonates were
eliminated using an additional solution of HC1. The grinding surface was then washed with an
0.5% solution of Triton X-100 (Johnson, 2018) to recover any pollen and starch grains that were


58
present. The grinding surface was carefully scrubbed with an ultrasonic toothbrush and then rinsed using reverse osmosis deionized (RODI) water (Varney, personal communication, 2018). Each sample was filtered through 250-micron mesh to remove any large particles that may have come off during the washing process (Cummings and Varney, 2018).
This pollen-rich organic sample was rinsed and then received a 25-minute bath in hot hydrofluoric acid (HF) to eliminate the remaining inorganic particles (Cummings and Varney, 2018). The sample was acetylated for 10 minutes to eliminate any unnecessary organic matter (Varney, personal communication, 2018). The sample was then washed again using RODI to achieve a neutral pH level. Next, several drops of potassium hydroxide (KOH) were added to the sample which was then stained with safranin (Cummings and Varney, 2018), coloring all the cell nuclei red. The sample was then centrifuged at high speeds, causing the minute organic debris to travel towards the bottom of the tube.
A light microscope at 500X magnification was used to count the pollen. The preservation of pollen in the sample ranged from poor to good (Cummings and Varney, 2018). Comparative reference material collected at the University of Colorado Herbarium and the Intermountain Herbarium at Utah State University was used to help identify pollen types to the family, genus, and species level (Cummings and Varney, 2018). Microscopic pieces of charcoal were recorded during a portion of the pollen count. The estimated abundance of charcoal was determined through computer extrapolation and is presented on the pollen diagram (Figure 5.3). Indeterminate pollen refers to pollen grains that are folded, mutilated, or are otherwise unrecognizable (Cummings and Varney, 2018).
Starch granules can be retrieved from sediments, but are seldom preserved (Kooyman, 2015). While susceptible to enzymatic attack, they can survive for an extended period of time in


59
Figure 5.3. Pollen diagram for a mano, 5LA1057, Las Animas County, Colorado. Figure from Cummings and Varney (2018: 8).
a dry, stable environment. For example, they can survive in dental calculus and in the microcracks of ground stone tools (Hardy, 2007). Starches can also come from residue analysis, including carbonized residues (Kooyman, 2015). Pollen extraction from ground tools often retains those granules. Starches were recorded during the pollen count on the mano. Starch granules are an important energy reserve in plants. Starches are found in various seeds, such as corn, rice, and wheat. It is also found in starchy roots and tubers. The main categories of starches include the following: with or without a visible hilum, a centric or eccentric hilum, hila


60
patterns (cracked, dot, or elongated), and starch shape (angular, circular, ellipse, or lenticular) (Cummings and Varney, 2018). Some of these starch categories are relatively common and occur in many different plant species, while others are unique to specific plants.
ED-XRF (Energy Dispersive X-Ray Fluorescence) Analysis
A small quantity of blue pigment observed on the surface of the mano was the focus of the X-ray fluorescence (XRF) analysis. XRF is a non-destructive analytical technique used to assess the elemental composition of a wide range of different geological materials. A Bruker Tracer 5i (900F4197) was used with an Rh anode and XFlash® silicon drift detector (SDD) with a resolution of < 140 eV, full width height maximum (FWHM) of the manganese (Mn) Ka peak (Cummings and Varney, 2018). Assays for trace elements and metals were collected at 50 kV with a current of 10.3 pA and a green filter composed of 150 pm Cu, 25 pm Ti, and 300 pm A1 to help eliminate any secondary emission of bremsstrahlung radiation below 13 keV (Cummings and Varney, 2018).
For elemental analysis, an empirical calibration reference set was used composed of 41 clay standards that were collected and analyzed by the University of Texas at Arlington (UTA) following Rowe et al.’s (2012) methodology (Cummings and Varney, 2018). The use of empirical standards is essential for data reliability and interlaboratory comparisons (Speakman and Shackley, 2013), especially when accuracy is at the parts-per-million (ppm) level for most elements (Cummings and Varney, 2018).
ART AX and S1PXRF software were used to help determine the elemental composition of the provided samples. Although present in the resulting spectrum, rhodium (Rh) and palladium (Pd) are not listed in the discussion below because of their association with the instrument (Cummings and Varney, 2018). Geological samples, like this mano, are often


61
difficult to evaluate because of their heterogeneous composition (Cummings and Varney, 2018). While these samples appear consistent, they still represent a combination of different materials that were not homogenized before testing (Cummings and Varney, 2018). ED-XRF analysis of the data from separate points in a heterogeneous sample represents the structure of that area. The area displaying the blue pigment was smaller than the size of the Tracer, indicating that the mano itself is also reflected in the spectrum collected (Cummings and Varney, 2018).
Results
Archaeobotanical analysis on the mano revealed the presence of several composites, members of the amaranth family, and grasses (all of which are potential food sources). Specifically, analysis revealed the presence of Amaranthaceae, Artemisia, High-spine Asteraceae, Liguliflorae, Melilotus, and Poaceae (Figure 5.3, Table 5.1), representing plants from the goosefoot family, sagebrush, plants from the sunflower family, members of the chicory tribe of the sunflower family, sweet clover, and grasses (Cummings and Varney, 2018). Charcoal and other indeterminate starches were also found. Amaranthaceae, High-spine Asteraceae, and Poaceae were the most plentiful, representing plants from the goosefoot family, sunflower family, and grasses. All of these families include plants with edible portions (Cummings and Varney, 2018). Seeds from plants in each of these families are believed to have been ground.
The recovery of a large, lenticular starch exhibiting heavy edge damage, suggesting cooking, indicates that seeds from a large-seeded grass species, such as barley grass (Hordeum pusilim), native wheatgrass (Agropyron), or ryegrass (Elymus), was ground (Cummings and Varney, 2018). All of these cool-season grasses have seeds that ripen during the late spring to early summer. Pollen recovered from the samples offers a look at the plants available to the shelter’s inhabitants. Seeds from a member of the goosefoot and/or sunflower families may have


62
Table 5.1. Pollen Types Recovered in Samples from Site 5LA1057,
Las Animas County, Colorado. Table from Cummings and Varney (2018: 7).
Scientific Name Common Name
NON-ARBOREAL POLLEN:
Amaranthaceae Amaranth family (now includes Chenopodiaceae, these two families were combined based on genetic testing and the pollen category "Cheno-ams")
Asteraceae: Sunflower family
Artemisia Sagebrush
High-spine Includes Aster, Rabbitbrush, Snakeweed, Sunflower, etc.
Liguliflorae Chicory tribe, includes Dandelion and Chicory
Melilotus Sweet clover
Poaceae Grass family
Indeterminate Too badly deteriorated to identify
STARCHES:
Lenticular starch Typical of starches produced by grass seeds such as those from wheat grass (Agropyron), ryegrass {Elymus), or barley grass (Hordeum)
OTHER:
Microscopic charcoal Microscopic charcoal fragments
Total pollen concentration Quantity of pollen per cubic centimeter (cc) of sediment
been ground using this mano. The XRF analysis generated numerous peaks (Figure 5.4); including, in order of diminishing amplitude - silicon, calcium, aluminum, iron, magnesium, potassium, sulfur, phosphorus, and titanium (Cummings and Varney, 2018). All the other peaks are considered minor in comparison. None of the elements stand out to explain the small amount of blue pigment on the surface of the mano (Cummings and Varney, 2018).
Data Collection Techniques on Portable Ground Stone Portable ground stone artifacts from Trinchera Cave are used in this study. The sample represents multiple excavations that have occurred since 1949 when the Louden brothers first


63
Figure 5.4. Spectral image for area of mano exhibiting blue stain.
Figure from Cummings and Varney (2018: 9).
discovered the site. Manos and metates housed at the Louden-Henritze Archaeology Museum in Trinidad, Colorado were examined between May 13th - May 18th, 2018. All of the portable materials included in this data set were measured individually using a Kobalt tape measure and plastic Ohaus calipers. Measurements were then compared against catalogs, field notes, and artifact inventories present at the museum. If a discrepancy arose, the mean was taken between the two measurements and then recorded on the artifact analysis sheet.
Artifact analysis sheets were created prior to conducting research. They were devised to help collect necessary data from both the portable ground stone and the bedrock ground stone. The artifact analysis sheets asked the following information - site name, catalog number, year of excavation, provenience, artifact type, material type, shape, condition (whole, < half, > half), location of wear (long surface, both ends, both long surface and ends, or none), pecked surface (yes or no), polished (yes or no), portable (yes or no), use-wear (light, moderate, heavy, unused),


64
length and width, and wear patterns (circular, reciprocal, pounding, unknown, none) (Figure 5.5). Each of these variables is detailed below.
Site Name: Site name for all the ground stone materials refers to Trinchera Cave, site 5LA1057. Catalog Number: The catalog number indicates the cataloging system used by the different excavation teams: Chase (TA #), Dick (TR.l #), Simpson (TR1-D- #), or Nowak (1057.0. #). Year of Excavation: The year of excavation represents what year the artifact was collected: Chase (1949 - 1950), Dick (1954 - 1957), Simpson (1974), or Nowak (1999 - 2001). Provenience: Multiple excavations have taken place at the shelter over the years, with each using a different cataloging system. Chase’s artifacts have a TA # printed on them and are all listed in Colorado College’s database (Nowak and Gerhart, 2002: 103). All his measurements were taken in cm. Dick’s artifacts have a TR.l - # printed on them and are all listed in same database. All of Dick’s recovered artifacts have a depth recorded in inches and a number and letter provenience associated with them (Nowak and Gerhart, 2002: 105). The number and letter refer to his grid system. Unfortunately, details of how the grid system was set up are not present in the archived field notes. Since artifact depth was recorded in inches, it is presumed that he established a grid system using feet. Simpson’s artifacts have a TR1-D- # printed on them. All of her measurements were taken in cm. Nowak’s artifacts have a 1057.M (or O)., then a # printed on them and are all listed in Colorado College’s database. All of his artifacts were measured in cm and have a letter provenience associated with them (Nowak and Gerhart, 2002: 105).
Artifact Type: Artifact type refers specifically to the type of ground stone that was examined: slab metate, basin metate, trough metate, undetermined metate fragment, bedrock metate, or mano. Although the term “ground stone” includes such items as abraders, hammerstones, manos


65
Catalog#___________________ Artifact Type:____________________________________________________
Site: Trinchera Cave 5LA9555 Portable: Yes No
Excavation: 1954-57 1974-75 1999-01 Material Type:___________________
Provenience: Yes________________No Fragment: Whole < Half > Half
Metrics:
Length (mm) Width (mm) Depth (mm) Perimeter (mm) Surface Area (mm2) Weight (g)

Use-Wear: Light Moderate Heavy Unused Wear Patterns: Reciprocal Rotary Pounding
Pecked Surface: Yes No Location of Wear: Long Surface Ends Both
Polished: Yes No
Shape:________
Photograph ft's:
Recorder:
Additional Comments:
Figure 5.5. Blank artifact analysis sheet.
and metates, mortars and pestles, and polishers (Adams, 2014a), only manos and metates were examined during the course of this study.
Material Type: Material type was determined through visual inspection and by going through catalogs, field notes, and artifact inventories. The different material types represented in this collection are Dakota sandstone, granite, conglomerate, and vesicular basalt.
Shape: Shape (circular, oval, square, rectangular, triangular, and undefined) of the ground stone artifacts is particularly subjective. The shape of the grinding surfaces has the potential to inform us of a variety of behavioral processes used to process material (circular, reciprocal, or pounding) (Lynch, 2017a).
Condition: This variable records whether the artifact is whole or a fragment. A fragment refers to a broken part of an artifact. The artifact analysis sheet included fragment type: whole, less than half, or more than half.
Location of Wear: The location of wear (long surface, both ends, both long surface and ends, or none) refers to where use occurred on the grinding tool.


66
Pecking: Pecked surface (yes or no) indicates whether or not the grinding surface of the artifact shows visible evidence of pecking. Pecking was determined by the presence of multiple peck marks on the grinding surface and by individual marks that continue past the grinding area. Battering is not included in this analysis. Battering results from use - as in a pestle - and is seen in a crushing of the grains on the grinding surface as opposed to pecking which is used to refurbish (as on a mano or metate surface) and is evident as individual marks that extend beyond the surface into the body of the specimen.
Taphonomic agents can interfere with the analyses of ground stone artifacts, including carnivores, scavengers, and humans (Blumenschine, 1988). Other taphonomic processes have the potential to affect ground stone artifacts, such as fluvial action, gravity, and weathering (Lyman, 1994; Rick, 1976; Stodder, 2008). Rock shelters are often subjected to a wide range of different taphonomic processes. Each portable artifact was examined for any identifiable taphonomic signatures.
Polish: Polished (yes or no) refers to whether or not the grinding surface shows visible evidence of sheen.
Portable: Next on the artifact analysis sheet is portable (yes or no). Chase, Dick, Simpson, and Nowak’s collections are all portable, meaning they can be moved from one spot to another. Use-Wear: Use-wear (light, moderate, heavy, unused) is the next attribute and was determined by using catalogs, field notes, and artifact inventories when possible. Each piece of ground stone was carefully inspected. Light wear was selected when very little evidence of use wear was visible with the naked eye. Moderate wear was selected when evidence of use wear left behind noticeable damage and did not alter the basic shape of the tool. Selecting heavy wear indicated


67
that the use wear was so intense that it changed the natural shape of the tool. Finally, unused was selected when there was no visible use wear on the tool.
Length and Width: For the manos, length and width were determined by measuring the entire artifact. For the metates however, length and width were determined by measuring the portion of the artifact that contains wear, not the entire artifact itself. On the metates, width is represented by the second longest axis.
Wear Pattern: Use-wear analysis can reveal the presence of abrasion, impact fractures, and/or sheen (Adams, 2014a). Behavioral inferences are often possible through use-wear analysis and ethnographic descriptions. Wear patterns (circular, reciprocal, pounding, unknown, none) indicate how an intermediate substance was being ground. It is often possible to identify distinctive use-wear patterns on moderately and heavily used tools (Adams, 2014a: 36 - 44).
All of the artifacts were carefully inspected for wear patterns. A circular stroke causes visible striations to appear in various directions across the surface of the stone. A reciprocal stroke entails moving the tool back and forth across the stone’s surface. The pounding stroke uses the tool’s size and weight to reduce the intermediate substance being ground. The crushing of grains on the individual surface is more closely associated to battering (Adams, 2014a: 132). Pounding is more of a high-impact action resulting in impact fractures that are much different than those caused by crushing (Adams, 2014a: 32). Depending on the intermediate substance being ground, the two adjacent surfaces may not come into contact with each other reducing impact damage to both tools (Adams, 2014a: 132). Unknown was selected when a wear pattern was present, but could not be determined. If no wear pattern was found, then none was selected and
was coded as absent in SPSS.


68
Research of the portable materials took place between May 13th and May 18th, 2018 at the Louden-Henritze Archaeology Museum in Trinidad, Colorado. This marks the first time any ground stone materials have been studied from Trinchera Cave (Martin, personal communication, 2018a). Comparable information was collected on the BGS surfaces as was collected on the portable ground stone. Depth was added as an additional variable.
Data Collection Techniques on Bedrock Ground Stone The bedrock grinding surfaces present at Trinchera Cave were recorded between June 22nd and June 24th, 2018 (Figure 5.6). Surface dimensions (length, width, and depth) were measured twice to accommodate for inter-observer error (Lynch et al., 2012). All measurements were taken in cm and then later converted into mm. The first set of measurements were collected using a 16” x 24” Aluminum Professional Square from Empire and a Kobalt tape measure. A secondary set of measurements were then collected by Laura Baker to insure inter-observer consistency in the data. The mean of each measurement was calculated and recorded on an artifact analysis sheet. IBM SPSS Statistical Software (V. 25) is used to record and manage this data set.
As indicated earlier, artifact analysis sheets were made prior to conducting research (Figure 5.4). Data on portable ground stone and bedrock ground stone (BGS) were collected using the same paperwork. Comparable information was collected on the portable ground stone as was collected on the BGS.
Site Name: The site name used for all the bedrock grinding surfaces Trinchera Cave, site 5LA1057.
Catalog Number: The catalog number indicates the cataloging system used. The BGS found at the shelter have not been previously cataloged. A cataloging system was created for them for


69
Figure 5.6. Photos of bedrock grinding surfaces taken at Trinchera Cave, this study: 1057.R. #, based on Nowak’s system of 1057.M. (or O) #.
Year of Data Recording: For the portable ground stone, the year of excavation represents what year the artifact was collected. However, for the BGS the year of excavation refers to the year the surfaces were analyzed. In this case, the year is 2018.
Provenience: Provenience information indicates the depth the bedrock features were analyzed at (surface, 0 - 6”, and so on). All of the bedrock features examined are at the surface level.


70
Artifact Type: Artifact type refers to the type of ground stone that was examined, bedrock metates. Zier (2015: 19) established a primary datum point opposite the cave entrance during his investigation of the site. A datum point is a point of reference from known coordinates and is often established using an electronic total station. When the site was visited between June 22nd and June 24th, 2018, the datum point set up by Zier was no longer present. Zier (2015: 19) also established a permanent datum point when he was mapping the site. A large nail was embedded in a block of concrete with the site number “5LA1057” etched nearby (Zier, 2015: 19).
However, the permanent site datum could not be located. It may have been pulled or become buried over time. Zier’s (2015: 11) detailed map of the site is used as a reference for provenience (Figure 5.7).
Material: The different material types represented by the portable artifacts housed at the Louden-Henritze Archaeology Museum include Dakota sandstone, granite, conglomerate, and vesicular basalt. Permanent grinding stations are often carved into unmovable rocks and includes those found in caves, shelters, rocky outcrops, and structure walls (Woodbury, 1954: 117). All of the grinding surfaces found at Trinchera Cave are carved into large boulders of Dakota sandstone.
Shape: Shape is the next attribute on the artifact analysis sheet. Determining shape (circular, oval, square, rectangular, triangular, or undefined) of the grinding surfaces is largely subjective. Each surface was individually inspected for an approximate shape. The shape of BGS surfaces may be able to inform us of a range of behavioral processes used to process intermediate substances (circular, reciprocal, or pounding) (Lynch, 2017a).
Condition: Fragment (whole, < half, > half) refers primarily to the portable materials. All of the
bedrock features are considered intact.


71
Figure 5.7. Trinchera Cave base camp. Figure from Zier (2015: 11).
Location of Wear: The location of wear (base only, up the sides) was then determined for each
of the grinding surfaces. Wear was not determined for the bedrock feature, just the individual grinding surfaces. Examination was completed in bright sunlight and with the naked eye. Pecking: A pecked surface (yes or no) indicates whether or not the BGS surfaces shows evidence of pecking. Pecking is particularly useful for resharpening the grinding surfaces of manos and metates (Adams, 2014a: 45). Battering is not included in this analysis. Pecking is used to refurbish (as on a mano or metate surface) and is represented by individual marks that continue on past the grinding surface into the body of the specimen.
Polish: A polished surface (yes or no) shows visible evidence of sheen.


72
Use-Wear: Use-wear (light, moderate, heavy, unused) was determined through careful inspection.
Length. Width, and Depth: For the bedrock grinding surfaces, length and width were determined by measuring the portion of the surface that contains wear. Width is represented by the second longest axis. The depth of each grinding surface was measured in cm: 0.1 - 1.1 cm (light), 1.2-2.1 cm (moderate), and 2.2 cm and greater (heavy). To determine the depth of the grinding surfaces, a 16” x 24” Aluminum Professional Square was laid across the top. A Kobalt tape measure was then used to determine depth in the center, the proximal end, and the distal end.
The three depth measurements taken were then averaged and subsequently recorded. The second analyst then repeated the entire procedure. The mean of those two figures was then calculated for the final depth of each grinding surface.
Wear Patterns: Wear patterns (circular, reciprocal, pounding, unknown, or none) suggest how a substance was processed. All of the BGS were carefully inspected for wear patterns. It is unclear how the weathering of these surfaces may have affected the grinding surfaces. A circular stroke causes visible striations to appear in various directions across the surface of the stone. A reciprocal stroke entails moving the tool back and forth across the stone’s surface. The pounding stoke uses the tool’s size and weight to reduce the intermediate substance being ground. Often times, impact fractures are generated by the pounding. These fractures are usually deeper and more jagged than those caused by crushing (Adams, 2014a: 132). Depending on the material being, the opposing surfaces may not even come into contact with one another. This results in less impact damage to both tools (Adams, 2014a: 132). Unknown was selected when a wear pattern was present, but could not be determined. If no wear pattern was found, then none was
selected and was coded as absent in SPSS.


73
CHAPTER VI RESULTS Introduction
The portable and bedrock ground stone data sets presented in this study are representative of an area that has had substantial human activity. All information gathered comes from the Trinchera Cave (5LA1057) archaeological site in Las Animas County, Colorado. The first data set used in this study is represented by (n = 179) portable ground stone artifacts recovered from Trinchera Cave in Southeastern Colorado. The second data set represents the bedrock grinding surfaces present at Trinchera Cave (n = 32).
Metate Sample Description
The different raw materials represented by the ground stone technology includes Dakota sandstone, granite, conglomerate, and vesicular basalt. A bar graph is used to determine the quantity of each raw material type (Figure 6.1). Dakota sandstone represents the largest percentage in the assemblage (75.8%). This percentage is to be expected given the site is located in an area dominated by Dakota sandstone. The other raw materials are locally available. A second bar graph is used to help determine the quantity of each specific artifact type in the archaeological assemblage (Figure 6.2).
The frequency of artifact types recovered at the site varies. This sample is the accumulation of various archaeological digs conducted at the site and includes manos and mano fragments (n = 150), slab metates (n = 22), basin metates (n = 6), and one trough metate. Manos were often cached for later use, while metates may have been removed once a site was abandoned or were later scavenged (Adams, 1998; Schlanger, 1991). Broken manos are often cached because they can be used later as hammer stones, supports, and as structural fill (Clark,


Count Count
74
Dakota Sandstone Vesicular Basalt Conglomerate Granite Argillite
Raw Material
Figure 6.1. Bar graph of raw material types.
Slab Metate Basin Metate Trough Metate Bedrock Metate Mano
Artifact Type
Figure 6.2. Bar graph of artifact types.


75
1988: 94; Hayden, 1987: 191; Searcy, 2011: 98). Manos and metates are both heavily valued by family members over many generations (Searcy, 2011). Nowak uncovered evidence of a midden directly under the shelter’s dripline. Many of the portable ground stone fragments were discovered in this midden (Nowak and Gerhart, 2002). The high concentration of ground stone indicates that this area was likely a deposit zone for grinding tools after they had broken. Seed and plant processing likely took place further inside the shelter.
Bedrock metates are the most prevalent metate type found at the site, followed by slab metates (Figure 6.2). If Trinchera Cave was used on a temporary, but regularly used basis as Nowak and Gerhart (2002: 12) suggest, then people living in the region may have transported specific metates with them to the shelter. Once they left, they may have taken their metates with them. It is also possible that the bedrock ground stone may have been preferred over portable ground stone at Trinchera Cave.
The portable artifacts are described through three primary continuous variables: length, width, and thickness. The basic unit of measurement is metric with each of these variables being measured in millimeters (mm). The bedrock grinding surfaces are also described through three primary continuous variables: length, width, and depth. Due to the size of the grinding surfaces and the limitations of the measuring tools, each of the variables were initially measured in centimeters (cm) and were later converted to mm for analysis. Several box plots are used to determine:
1) median length while controlling for raw material type (Figure 6.3).
2) median width while controlling for raw material type (Figure 6.4).
3) median length while controlling for artifact type (Figure 6.5).
4) median width while controlling for artifact type (Figure 6.6).


Width (mm) Length (mm)
76
600.00
*
1
400.00
200.00
.00
172
£---
9
Dakota Vesicular Basalt Conglomerate Granite Argillite
Sandstone
Raw Material
Figure 6.3. Box plot illustrating length range for bedrock and portable grinding surfaces of each raw material type.
400.00
9
*144
22
Dakota Vesicular Basalt Conglomerate Granite Argillite
Sandstone
Raw Material
Figure 6.4. Box plot illustrating width range for bedrock and portable grinding surfaces of each raw material type.


Width (mm) Length (mm)
77
600.00
★
1
400.00
200.00
.00
172
174
X
O
9
Slab Metate Basin Metate Trough Metate Bedrock Metate
Artifact Type
Figure 6.5. Box plot illustrating length range for each artifact type
400.00
300.00
144
-W---
172
*
*
9
200.00
100.00
192
O
159
X
X
.00
Slab Metate
Basin Metate Trough Metate Bedrock Metate Artifact Type
I
i
Mano
Mano
Figure 6.6. Box plot illustrating width range for each artifact type.


78
Table 6.1. T-test results comparing the length and width of bedrock ground stone to portable ground stone.
Variable Degrees of t-score p-value Difference in
freedom means (mm)
Length 52 -3.655 0.001 89.5
Width 52 -1.707 0.094 27.2
Comparison of Metates to Bedrock Grinding Surfaces
The first research question asks if there is a functional difference between bedrock ground stone and portable ground stone. Several statistical tests were run to compare sub-groups within the assemblage from Trinchera Cave. Specifically, independent T-tests were run to compare length and width - to determine if there is a statistically significant difference in the mean and variation around the mean for these variables. For analysis of the base stones, slab metates are compared to the bedrock grinding surfaces. In both cases, the analyses indicate the null hypothesis of no difference should be rejected. The bedrock grinding surfaces are, on average, 89.5 mm longer than metate grinding surfaces. In addition, the bedrock grinding surfaces are, on average, 27.2 mm wider than metate grinding surfaces. These results are illustrated in Table 6.1.
Chi-square tests were then ran on several nominal variables. Chi-square tests are commonly used to determine if two nominal variables are related in any way. While the tests are used to determine whether or not an association is present, they cannot determine the strength of that relationship. If a relationship is present, then a Cramer’s V test is used to determine the strength of that relationship. For this analysis, artifact type is compared against two different nominal variables (wear patterns and use-wear) to see if an association exists. Wear patterns are


79
described through five nominal-type variables: reciprocal, rotary, pounding, unknown, or none. Use-wear is described through four nominal-type variables: light, moderate, heavy, or unused.
The first chi-square test was used to determine if an association exists between artifact type and wear patterns. After the initial test, SPSS software noted that 13 cells (65.0%) had expected values of less than 5. The minimum expected count is 0.01. Artifact types were then recoded into different variables. Slab metates, basin metates, trough metates, and undetermined metate fragments were combined together as portable base stones. Bedrock grinding surfaces were compared to these. For wear patterns, the variables “unused” and “none” were excluded from this analysis. A chi-square test was then run a second time with the new coded variables. This time, 0 cells (0.00%) had an expected outcome of less than 5. The minimum expected count is 7.13. The analysis indicates there is not a relationship present between artifact type and wear pattern (df = 2, p-value = 0.405).
A chi-square test was then used to determine if an association exists between artifact type and use-wear. The artifacts were recoded the same as noted above. Use-wear variables were left alone. A chi-square test was then ran again with the newly coded variables. This time, 0 cells (0.00%) had an expected outcome of less than 5. The minimum expected count is 7.13. There is not a relationship present between artifact type and use-wear (df = 2, p-value = 0.405).
Summary of Results
IBM SPSS Statistical Software (V. 25) is used to record and manage the data sets. Multiple box plots are used to determine median length while controlling for raw material type (Figure 6.3), median width while controlling for raw material type (Figure 6.4), median length while controlling for artifact type (Figure 6.5), and median width while controlling for artifact type (Figure 6.6). Based on the results illustrated by the box plots, the greatest median length


80
and width for raw material type is Dakota sandstone. The greatest median length and width for artifact type are bedrock metates. This means that on average, the bedrock grinding surfaces are longer and wider than their portable counterparts.
Two independent sample t-tests were then ran to address the first research question on whether or not there is a functional difference between bedrock ground stone and portable ground stone. The first independent sample t-test was ran to determine if there is a statistical difference in length between bedrock and portable grinding surfaces and the second on width. Testing revealed a significant difference in mean length (p = 0.001) and width (p = 0.094). Conversely, the chi-square tests indicate that wear patterns and type of use-wear do not differ between portable and non-portable base stones (p = 0.405 for both).
Mano Sample Description
The manos and mano fragments are described through three primary continuous variables: length, width, and thickness. The basic unit of measurement is metric with each of these variables being measured in millimeters (mm). Histograms are used to determine:
1) mean and standard deviation for length (Figure 6.7).
2) mean and standard deviation for width (Figure 6.8).
3) mean and standard deviation for thickness (Figure 6.9).
A bar graph is used to determine the occurrence of each wear pattern type (Figure 6.10). Mano wear patterns are largely undetermined. A second bar graph is used to help determine the occurrence of each use-wear type (Figure 6.11). Mano use-wear is predominantly heavy. Comparisons of Manos to Grinding Surfaces
Several statistical tests were run on the manos and mano fragments. Specifically, histograms were run to compare length, width, and thickness - to determine if there is a normal


Frequency Frequency
81
SO .00 100.00 150.00
Length (mm)
Figure 6.7. Histogram illustrating mano length.
20
Width (mm)
Figure 6.8. Histogram illustrating mano width.
Manos Manos


Count Frequency
82
40
30
20
10
0
20.00 40.00 60.00 80.00
Thickness (mm)
Figure 6.9. Histogram illustrating mano thickness.
Pounding Undetermined
Wear Patterns
Figure 6.10. Bar graph of wear pattern occurrence.
Manos Manos


83
120 100
80
c â–¡ o
60 40 20 0
Moderate Heavy
Use-Wear
Figure 6.11. Bar graph of use-wear occurrence, or bimodal distribution. The portable and bedrock grinding surfaces have varying lengths and
widths. Through analysis, the manos may be matched with one type or the other. A bimodal distribution exists for mano length (Figure 6.7). Manos shorter than the mean length of 90.0 mm, may have been used for the portable grinding surfaces. Those longer than the mean may have been used for the bedrock grinding surfaces. Meanwhile, a normal distribution exists for mano width and thickness (Figures 6.8 and 6.9).
Summary of Results
Histograms are used to determine the mean and standard deviation for mano length (Figure 6.7), mano width (Figure 6.8), and mano thickness (Figure 6.9). Based on the results illustrated by the histograms (Table 6.2), there is a bimodal distribution present solely for mano length. A normal distribution is present for mano width and thickness. Two bar graphs are used
Manos


84
Table 6.2. Histogram results comparing mano length, width, and thickness to bedrock and portable grinding surfaces.
Variable Mean (mm) Standard Distribution
Deviation (mm) Type
Length 90.0 29.20 Bimodal
Width 71.2 19.96 Normal
Thickness 42.1 10.73 Normal
to determine the prevalence of specific wear patterns and use-wear. Mano wear patterns are largely undetermined without the aid of a microscope. Future research may be able to shed some light in this area. Both the portable metates and bedrock grinding surfaces show strong evidence of reciprocal wear patterns. Mano use-wear is predominantly heavy. Portable grinding surfaces show mostly moderate wear (44.8%), followed by heavy wear (31.0%), and light wear (24.2%). Bedrock grinding surfaces show mostly moderate wear (43.8%), followed by light wear (37.5%), and heavy wear (18.7%). The portable grinding surfaces may have been predominantly used at Trinchera Cave, with the bedrock grinding surfaces being used as a supplemental form of technology.


85
CHAPTER VII CONCLUSIONS Summary
The purpose of my research is twofold: 1) to determine if there is a functional difference between bedrock ground stone and portable ground stone and 2) to determine when bedrock ground stone is preferred over portable ground stone within the mobility and subsistence system seen at Trinchera Cave. Several statistical tests were run to compare sub-groups within the assemblage from Trinchera Cave. Specifically, independent T-tests were run to compare metate length and width. Statistics revealed that bedrock grinding surfaces are, on average, 89.5 mm longer than portable grinding surfaces. Furthermore, bedrock grinding surfaces are, on average, 27.2 mm wider than portable grinding surfaces.
Ethnographic studies have demonstrated that efficiency is improved by increasing the number of ground stone tools used for reciprocal grinding and the size of the grinding surfaces (Hard, 1986, 1990; Horsfall, 1987; Mauldin, 1991, 1993). The portable and bedrock grinding surfaces from Trinchera Cave show strong evidence of reciprocal wear patterns, though mano wear patterns remain largely undetermined. A positive correlation exists between the length of the grinding surface, as illustrated by mano length, and the importance of agriculture to the subsistence economies of contemporary groups (Hard, 1986, 1990; Mauldin, 1991, 1993). By increasing the size of the grinding surface, processing speed increases and it allows for more grain to be processed at a time (Mauldin, 1993). The length and width of the bedrock grinding surfaces are larger on average than the portable grinding surfaces from Trinchera Cave. These two variables represent a functional difference between the two sub-groups.


86
Paleoenvironmental records are not extensive for southeastern Colorado, though some relevant research has been completed. Zier (1989), for example, discusses the environmental conditions over the last 4500 - 4000 years in southeastern Colorado. Reviewing local and regional data, Zier (1989: 41 - 42) determines that the period 2000 - 1000 BP was fairly cool and moist with greater effective moisture than what we see today. Warmer temperatures and drier conditions followed around 1000 - 450 BP, while the Little Ice Age (LIA) between 450 -150 BP saw a return to cooler temperatures and a wetter climate (Zier, 1989: 41 - 42). Within the last 150 years, the region has seen much warmer and drier conditions. Painter et al. (1999) provides a more comprehensive synthesis.
Nowak and Gerhart (2002: 12) suggest that maize and squash were introduced to the area sometime between AD 500 and AD 1000. This may indicate that people wanted to diversify the food resources in the area, allowing them to spend more time at the site. During this period, the region would have been relatively cold and wet. Bellorado (2011: 41) argues that in the San Juan Basin, it is very difficult for people to successfully produce mature maize and squash harvests during these colder periods. Most varieties of corn require a minimum growing season of between 110 - 120 days, and a minimum of 2200 - 2500 com growing degree days (Black, 2000: 8). Corn growing degree days (GDD) are determined by subtracting the plant’s threshold temperature from the region’s average daily temperature (NDawn Center,
2019). GDD is between 2579 and 2800 for Las Animas County. The TCAD currently fits these temperature and agricultural requirements (Black, 2000: 8), but may not have when maize and squash were first introduced to the region.
The number of frost-free days also plays an important part during the growing season. According to the Old Farmer’s Almanac (2019), between 1981 and 2010 the average number of


87
frost-free days in Trinidad, Colorado was 123. Unpredictable growing seasons may have caused the inhabitants to respond by complementing their daily processing needs. Bedrock ground stone may have been used in addition to the portable ground stone during this time period, rather than as a preference. Starch and pollen analysis conducted on the small mano revealed the presence of wild resources - by extension, this indicates that portable metates may have been used for wild resources and the bedrock for maize - but it’s a small sample and requires further research.
No single environmental factor influenced where different activities would take place in the region. Rather, the combination of construction/tool material and fuel in the Pinon-juniper woodlands, abundant game animals, economically useful plants, natural rock shelters present in the side canyons, and permanent water sources all contributed to the appeal of the greater Trinchera Creek basin (Black, 2000: 67). The presence of rock art and architectural elements may be an indication that spiritual motivations also played a role in some site locations; most notably, Trinchera Cave.
Suggestions for Future Research
Site Protection
Different warning signs have been posted at the shelter over the years. In the early 1990s, looting got so bad that a sign was posted on the northwest side of Trinchera Creek to help prevent against the illegal removal of artifacts (Garcia, personal communication, 2018). That sign was removed at some point by looters. New signage was established on the slope below the entrance to the shelter in 2013. However, that sign was discovered to be missing in 2018. More protection is desperately needed for this site. It is recommended that Trinchera Cave be periodically monitored by the Office of the State Archaeologist, both to help deter future looting and vandalism and to ensure that visible signage remains in place (Zier, 2015: 96). Regular


Full Text

PAGE 1

A COMPARISON OF PORTABLE AND BEDROC K GROUND STONE TECHNOLOGY AMONG HUNTER GATHERERS AT TRINCHERA CAVE IN SOUTHEASTERN COLORADO b y RYAN CHRISTOPHER BAKER B.A., Metropolitan State University of Denver, 2015 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Arts Anthropology Program 2019

PAGE 2

ii © 201 9 RYAN CHRISTOPHER BAKER ALL RIGHTS RESERVED

PAGE 3

iii This thesis f or the Master of Arts degree by Ryan Christopher Baker has been approved for the Anthropology Program by Tammy Stone, Chair Jamie M. Hodgkins Jonathan Kent Date: July 27 , 2019

PAGE 4

iv Baker, Ryan Christopher (MA, Anthropology Program) A Comparison of Portable and Bedrock Ground Stone Technology Among Hunter Gatherers at Trinchera Cave in Southeastern Colorado Thesis directed by Professor Tammy Stone ABSTRACT In 2007, the Chaquaqua Plateau Ground Stone Project (CPGSP) initiated a long term study of the ground stone features found on the Chaquaqua Plateau in Southeastern Colorado. The purpose of the multi disciplinary project is to describe, study, and standard ize research into these understudied features since minimal research into the nature of the bedrock ground stone features has been conducted. We know very little about their function, site distribution, and relationship to the regional archaeological pict ure. The features were used as food grinding areas to process local wild resources, while others believe they were used to process corn that was grown in the canyon floodplain. Based on ethnographic data from other regions, some researchers have suggeste d that these features may have been used for a number of different functions. Missing from current research for the region is an understanding of how portable and bedrock ground stone are related to regional hunter gatherer adaptation and if these relationships changed through time . The purpose of my research is twofold: 1) to determine if there is a functional difference between bedrock ground stone and portable ground stone and 2) to determine when bedrock ground stone is preferred over portable ground stone within the mobility and subsistence system seen at Trinchera Cave . This research may lend insight to how the prehistoric inhabitants of the area articulated with the landscape through the ir subsistence system. The m arginal value theorem and the technological investment model have direct

PAGE 5

v relevance to this research. Building on this study, f uture research within the Chaquaqua Plateau should include refining assumptions and applying the model s to additional sites in the region. The form and content of this abstract are approved . I recommend its publication. Approved: Tammy Stone

PAGE 6

vi DEDICATION This thesis is dedicated to my wife Laura and my sons Ethan and Caleb. Thank you to the guest services staff at the Denver Museum of Nature & Science for allowing me to take the necessary time off to pursue this. I want to thank everyone that has helped me along the way in unwavering support of my dream , including other family members, friends, and colleagues. Never stop giving up on your dreams. You can become anything you want to be if you work hard enough .

PAGE 7

vii ACKNOWLEDGEMENTS I would like to offer my sincere st gratitude to everyone that has been a part of this process. I would like to thank my academic advisor and thesis chair, Dr. Tammy Stone, and my committee members Dr. Jamie Hodgkins and Dr. Jonathan Kent for taking the time to read and comment o n my thesis. I am extremely thankful to Dr. Stone for our many conversations during my time spent at UCD . Your comments and suggestions on my thesis drafts were very much appreciated. My thesis was much improved as a result of your guidance. To my coho rts in the UCD anthropology department, thank you for your inspiration and encouragement . I wish to thank the remaining faculty members succeed and do better in each one of their respective classes. I al so wish to thank Loretta Martin and the Louden Henritze Archaeology Museum at Trinidad State Junior College in Trinidad, Colorado for allowing me to conduct my research on the ground stone collections housed there. Many other individuals helped me along the way. They are: Philip Baca and Buford Garcia, Baca Revocable Trust; Kevin Black, Office of the State Archaeologist (retired); Dr. Linda Cummings and R.A. Varney, PaleoResearch Institute, Inc.; Lauren Fuka, University of Michigan Museum of Anthropolog ical Archaeology; Bea Gallegos, Colorado State Land Board; Dr. Michele Koons, Denver Museum of Nature & Science; Dr. Elizabeth Lynch, University of Wyoming; Jeff Noblett, Colorado College; Brooke Rohde, University of Denver Museum of Anthropology; Rebecca Simon, Office of the State Archaeologist ; Connie Turner, University of Colorado Denver; Dr. Mary Van Buren, Colorado State University, and the Karen S. Greiner Endowment; Leah Zavaleta, Uni versity of Denver; and Christian Zier, Centennial Archaeology, Inc. Finally, I would like to thank all my family, friends, and colleagues without whom this thesis could not have been completed.

PAGE 8

viii TABLE OF CONTENTS CHAPTER I. INTRODUCTION ................................ ................................ ................................ ............... 1 Arrangement of the Thesis ................................ ................................ ................................ ... 3 II. THEORY ................................ ................................ ................................ ............................. 5 Interpretative Framework ................................ ................................ ................................ ............ 5 Human Behavioral Ecology ................................ ................................ ....................... 5 Optimal Foraging Theory ................................ ................................ ........................... 6 Marginal Value Theorem ................................ ................................ ................ 8 Technological Investment Model ................................ ................................ ... 8 Middle Range Theory ................................ ................................ ............................... 1 0 Forager/Collector Model ................................ ................................ ........................... 1 1 Case Study ................................ ................................ ................................ ......................... 1 3 III. BACKGROUND ................................ ................................ ................................ ............... 1 5 Introduction ................................ ................................ ................................ ....................... 1 5 Early Researchers at Trinchera Cave ................................ ................................ ................ 19 Etienne Renaud (1930 1941) ................................ ................................ ................. 2 0 Haldon Chase (1949 1950) ................................ ................................ .................... 2 2 Herbert W. Dick (1954 1957) ................................ ................................ ................ 2 5 Colorado Archaeological Society (1966 and 1969) ................................ .................. 2 6 Caryl Wood Simpson (1974) ................................ ................................ .................... 2 7 Contemporary Researchers at Trinchera Cave ................................ ................................ . 2 8 Office of the State Archaeologist (1997 1999) ................................ ...................... 2 8

PAGE 9

ix Michael Nowak (1999 2001) ................................ ................................ ................. 29 Colorado College ( 2002 2003) ................................ ................................ ............... 3 0 Christian J. Zie r (2013 2015) ................................ ................................ ................. 3 3 IV. GROUND STONE TECHNOLOGY ................................ ................................ ................ 3 5 Portable Ground Stone ................................ ................................ ................................ ....... 3 5 Bedrock Ground Stone ................................ ................................ ................................ ...... 3 8 Grinding Behaviors ................................ ................................ ................................ ............ 4 0 Design and Manufacture ................................ ................................ ........................... 4 0 Use ................................ ................................ ................................ ............................ 4 1 Use Wear Analysis ................................ ................................ ................................ ... 4 3 Wear ................................ ................................ ................................ .......................... 4 3 Wear Rates and Patterns ................................ ................................ ........................... 4 5 Resources ................................ ................................ ................................ .................. 4 6 Translating Portable Ground Studies to Bedrock Ground Stone ................................ ....... 4 7 V. METHODOLOGY ................................ ................................ ................................ ............ 49 Introduction ................................ ................................ ................................ ........................ 49 Data Limitations ................................ ................................ ................................ ................. 49 Inter observer Variability ................................ ................................ .......................... 49 Other Limitations ................................ ................................ ................................ ...... 49 Sample Selection ................................ ................................ ................................ ................ 5 0 Pollen Analysis ................................ ................................ ................................ ......... 5 3 Artifact Analysis ................................ ................................ ................................ ....... 5 6 Archaeobotanical Analysis ................................ ................................ ....................... 5 6

PAGE 10

x ED XRF (Energy Dispersive X Ray Fluorescence) Analysis ................................ .. 60 Results ................................ ................................ ................................ ....................... 6 1 Data Collection Techniques on Portable Ground Stone ................................ .................... 6 2 Data Collection Techniques on Bedrock Ground Stone ................................ .................... 6 8 VI. RESULTS ................................ ................................ ................................ .......................... 7 3 Introduction ................................ ................................ ................................ ................................ .. 7 3 Metate Sample Description ................................ ................................ ................................ ........ 7 3 Comparison of Metates to Bedrock Grinding Surfaces ................................ ............ 7 8 Summary of Results ................................ ................................ ................................ .. 7 9 Mano Sample Description ................................ ................................ ................................ .. 80 Comparison of Manos to Grinding Surfaces ................................ ............................ 80 Summary of Results ................................ ................................ ................................ .. 8 3 VII. CONCLUSION ................................ ................................ ................................ .................. 8 5 Summary ................................ ................................ ................................ ............................ 8 5 Suggestions for Future Research ................................ ................................ ....................... 8 7 Site Protection ................................ ................................ ................................ ........... 8 7 Museum Collections ................................ ................................ ................................ . 8 8 Future Research Questions ................................ ................................ ....................... 8 9 REFERENCES ................................ ................................ ................................ .............................. 9 1

PAGE 11

xi LIST OF TABLES TABLE 3.1: Radiocarbon Dates from Trinchera Cave ................................ ................................ .. 31 32 5.1: Pollen Types Recovered in Samples from Site 5LA1057, Las Animas County, Colorado. ................................ ................................ ................................ ............................ 6 2 6.1: T test results comparing the length and width of bedrock ground stone to portable ground stone. ................................ ................................ ................................ ........ 7 8 6.2: Histogram results comparing mano length, width, and thickness to bedrock and portable grinding surfaces ................................ ................................ ........................... 8 4

PAGE 12

xii LIST OF FIGURES FIGURE 2.1: Graphic Representation of the Marginal Value Theorem ................................ .................... 9 3.1: Relief Map of Trinchera Cave in Southeastern Colorado ................................ .................. 1 6 3.2: Trinchera Cave Site Map ................................ ................................ ................................ ... 1 8 3.3: Map of Trinchera Cave Showing the Locations of Excavation Units, 1949 2001 ......... 19 3. 4 : Overview of Trinchera Cave ................................ ................................ .............................. 2 0 4.1: Grinding S trokes and the S ubsequent W earing of T wo A djacent S urfaces A gainst a M etate S urface ................................ ................................ ................................ ................... 4 5 5.1: Different Views of Mano Used for Analysis. ................................ ................................ .... 5 5 5.2: Photos of Mano Taken at 15X 20X Magnification. ................................ ........................ 5 7 5. 3 : Pollen D iagram for a M ano, 5LA1057, Las Animas County, Colorado ........................... 59 5.4: Spectral I mage for A rea of M ano E xhibiting B lue S tain ................................ ................... 63 5. 5 : Blank Artifact Analysis Sheet ................................ ................................ ............................ 6 5 5. 6 : Photos of Bedrock Grinding Surfaces Taken at Trinchera Cave. ................................ ...... 6 9 5. 7 : Trinchera Cave Base Camp ................................ ................................ ............................... 7 1 6.1: Bar Graph of Raw Material Types ................................ ................................ ..................... 7 4 6.2: Bar Graph of Artifact Types ................................ ................................ .............................. 7 4 6.3: Box Plot Illustrating Length Range for Bedrock and Portable Grinding Surfaces of Each Raw Material Type. ................................ ................................ ................................ .. 7 6 6.4: Box Plot Illustrating Width Range for Bedrock and Portable Grinding Surfaces of Each Raw Material Type ................................ ................................ ................................ ... 7 6 6.5: Box Plot Illustrating Length Range for Each Artifact Type ................................ .............. 7 7 6.6: Box Plot Illustrating Width Range for Each Artifact Type ................................ ............... 7 7 6.7: Histogram Illustrating Mano Length ................................ ................................ ................. 8 1

PAGE 13

xiii 6.8: Histogram Illustrating Mano Length ................................ ................................ ................. 8 1 6. 9 : Histogram Illustrating Mano Thickness ................................ ................................ ............. 8 2 6.10: Bar Graph of Wear Pattern Occurrence ................................ ................................ ............. 8 2 6. 11 : Bar Graph of Use Wear Occurrence ................................ ................................ .................. 8 3

PAGE 14

1 CHAPTER I INTRODUCTION I n Southeaster n Colorado, bedrock ground stone features are primarily concentrated along the canyon sidewalls of the Apishapa, Chacuaco, Cuchara, and Purgatoire Rivers and their tributaries (Campbell, 1969; Loendorf, 2008; Lynch, 2010; Owens, 2007). Archaeological resea rch on the function and meaning of these features is in the beginning stages in this region. Ethnographic evidence for these features in other geographic regions indicates a number of possible functions individual and/or group production of food resourc es for daily consumption and ceremonial activities such as harvest festivals or initiation rituals for women (Barrett and Gifford, 1933; Bartlett, 1933; Claassen, 2011; Curtis, 1924; Frisbie, 1967; Gayton, 1948a, 1948b; Gifford, 1932; Jackson, 1991; Kluckh ohn et al., 1971). An important aspect of these grinding features is the social significance of the area where shared labor was organized and where social expectations were expressed between generations (Dick Bassonnette, 1998; Gayton, 1948a, 1948b; Jacks on, 1991). Lynch (2014) suggests that bedrock ground stone features vary depending on where they appear on the landscape and because of sociocultural beliefs, values, and traditions. Bedrock ground stone features have the potential to enrich our understa nding of past social dynamics between prehistoric peoples and their environment. The purpose of this research is twofold: 1) to determine if there is a functional difference between bedrock ground stone and portable ground stone and 2) to determine when bedrock ground stone is preferred over portable ground stone within the residential mobility and subsistence system seen at Trinchera Cave (5LA1057) . This research may lend insight to how the prehistoric inhabitants of the area articulated with the landscape through the ir subsistence system. Archaeology in the region suggests a highly complex set of mobility and settlement

PAGE 15

2 patterns based on the cultural remains of architecture, ceramics, lithics, and peri shables. Trinchera Cave is an especially important site because of its exceptionally well preserved archaeological materials, geographic location, rock art panels, and a continued occupation spanning from the Late Archaic Period (1050 BC 100 AD) to the Late Prehistoric stage (AD 100 1725) (Crocket, 2002). The perishable remains of bark, bone, feather, fiber, seeds, shell, skin, and wood are unmatched for the region (Simpson, 1 976). Additionally, this is the only site in southeastern Colorado that has yielded the three domesticated crops beans, maize, and squash (Black, 2000). Large sections of the rock shelter were excavated by avocational and professional archaeologists du ring several projects between 1949 and 2001, with the greatest amount of work occurring during the 1950s . Most of the artifacts and other excavated materials discovered during these excavations are housed at the Louden Henritze Archaeology Museum at Trini dad State Junior College (TSJC) in Trinidad, Colorado. Unfortunately, artifact inventories, field notes, and survey manuscripts from excavations prior to the 1970s are exceedingly rare and often times inadequate ( Zier, 2015) . A good deal of ambiguity appears to surround the earlier projects. The only project for which we have substantial documentation on the site comes from four field school sessions between 1999 and 2001 that was led by Michael Nowak of Colorado College. A comprehensive map of the site was created between 2013 and 2015 by Chris Zier of Centennial Archaeology, Inc. A considerable amount of looting has taken place at the site, resulting in the loss of archaeological materials and information . Despite this, the materials housed at the Louden Henritze Archaeology Museum and the features at the site have the potential to address questions relating to ground stone use in the area.

PAGE 16

3 Arrangement of the Thesis Chapter 1 outlines my thesis argument regarding the ground stone technology used by hunter gatherers at Trinchera Cave in southeastern Colorado through the analysis of portable and bedrock grinding surfaces using human behavioral ecology. Chapter 2 discusses the different t heoretical models that can help us to understand how Trinchera Cave fits into a broader use of the prehistoric landscape. Specifically , the collection of bedrock ground stone features in the region is technology directly tied to the subsistence system. T his chapter focuses on the relationship between subsistence practices and settlement patterns in the regio n . Constructing early hunter gatherer mobility patterns and settlement systems are crucial in our understanding of the archaeological record. Chapte r 3 provides a background on the early and contemporary researchers that have worked at Trinchera Cave and the roles they each played within the framework of a research design at the site . Chapter 4 highlights the importance of ground stone technology and how it can potentially contribute to our understanding of past lifeways. The role of portable and bedrock ground stone is examined here. Different grinding behaviors such as a tool design, manufacture, and utilization are covered. Raw mat erial availability can directly affect ground stone tool design, manufacture, use, and abandonment patterns. Chapter 5 explains the methodology employed during the course of this study. Portable ground stone recovered from Trinchera Cave and bedrock grou nd stone present at the site are used as the two sub groups for analysis. A pollen , starch , and XRF analysis was conducted on one of the portable artifacts . With analysis carried out by PaleoResearch Institute, Inc. , testing determine d the substances that were being processed by the grinding tool. Data collection

PAGE 17

4 techniques of the portable and bedrock ground stone are covered. IBM SPSS Statistical Software (V. 25) is use d to record and manage these data sets. Chapter 6 provides the results from my analysis as well as a discussion suggesting there is a relationship between artifact type and median length and width. A bimodal distribution was discovered for mano length in particular . These portable and bedrock grinding surfaces show strong evidence of specific wear patterns that can possibly be tied in with mano wear patterns. Results indicate that the portable grinding surfaces may have been predominantly used at Trinch era Cave, with the bedrock grinding surfaces being used as a complementary form of technology. Finally, Chapter 7 summarizes the thesis research and discusses insights into human behavior and adaptation in the region. Recommendations are then given on th e site, museum collections, and future research.

PAGE 18

5 CHAPTER II THEOR Y I nterp retative Framework Researchers interested in explaining technological change as a result of adaptive decision making have gravitated toward two theoretical approaches: Human Behavioral Ecology (HBE) and Optimal Foraging Theory (OFT) . While these two approaches may have different intellectual beginnings, OFT and HBE share many fundamental beliefs and goals . They both wish to understand the flexibility of human behavior in regards to economic constraints. They each believe that time and energy are key currencies to understanding complex technological behavior. They each believe that optimal behaviors become co mmonplace , keeping conditions constant. Both assume that change s in technology are not subject to natural selection. Instead, technological change reflects plasticity in the behavioral phenotype. Both approaches have addressed questions of material design and technological investment. Human Behavioral Ecology Human behavioral ecology (HBE) is one of the most widely used theoretical frameworks in the archaeology of hunter gatherers . HBE examines the flexibility in human behavior and uses formal hypotheses to explore the adaptive responses of individuals to environmental pressures , 2006; Broughton and Cannon, 2010; Codding and Bird, 2015; Lupo, 2007 ; Nettle et al., 2013; Winterhalder and Smith, 2000 ) , especially th ose related to

PAGE 19

6 questions o n prehistoric subsistence practices (Broughton, 2002; Cannon, 2003; Madsen, 1993; Madsen and Schmitt, 1998; Simms, 1987; Ugan, 2005; Whelan et al., 2013) . Optimal Foraging Theory Optimal F oraging T heory (OFT) was first developed by behavioral ecologists to help them better understand the foraging behavior of animals (Emlen, 1966; MacArthur and Pianka, 1966; Pulliam, 1974; Stephen s and Krebs, 1986). OFT models address such issues as habitat movement , the selection of specific resources , and time allocation (Winterhalder and Smith, 2000) . Food selection and resource utilization are regulated by various factors including predation, resource availability, seasonality, and territory size (Keene, 1983). By employing a behavioral ecology framework , w e can often predict the optimal foraging behavior of animals. Keene (1983: 139) recognized a cost benefit component associated with this approach. While food may provide an animal with energy, other factors are known to influence foraging behavior such as the time and energy costs associated with searching, collecting, and prepar ing the food. OFT suggests which resources will be collected and which will be ignored ( Cannon, 2003; Kelly, 2013) based on these factors . T he foraging strategies that offer the greatest benefit at the lowest cost are commonly used (Keene, 1983) . Researchers have observed similarities in behavior among animals and humans. Studies on animal foraging behavior can improve our understanding of different mobility strategies (Keene , 1983; Kelly , 2013). OFT models include several features , including a g oal, a unit of currency, a set of constraints, and a range of options (Kelly, 2013; Smith et al., 1983 ; Winterhalder and Smith, 2000 ). Often times, the goal is to be as efficient as possible while foraging ( described as energy gained over energy spent) (K elly, 2013). Currency is typically a measure of energy, such as calories gained per unit of time (Smith et al., 1983). However, o ther

PAGE 20

7 measures of resource value can include material requirements , monetary exchange , ornament or prestige items , or protein capture (Winterhalder and Smith, 2000) . The set of constraints are all the possible factors that may limit the amount of time an individual or group can spend foraging or processing resources . One important example of this is caring for children, which may have played a significant role for hunter (Homsey Messer, 2015; Whelan et al., 2013) and at Trinchera Cave in particular . Lastly, the range of options refers to the potential food resources that are available in the area. Resources have specific characteristics such as ritual or ceremonial importance, harvesting and processing times, medicinal value, and nutritional content (Kelly, 2013: 47; Nowak and Gerhart, 2002: 35 36). These characteristics influence which resour ces will be harvested and which will be overlooked ( Cannon, 2003; Kelly, 2013). This is also referred to as the diet breadth model (DMB). DMB is a traditional optimal foraging model from behavioral ecology that ranks food resources according to their potential caloric benefits and search, procurement, and processing costs (Kelly, 2013 ; Madsen, 1993; Madsen and Schmitt, 1998 ). According to this model, foragers will select specific resources that maximize foraging return rate s (Kelly, 2013 ; Madsen, 1993 ; Madsen and Schmitt, 1998 ; Whelan et al., 2013 ). Ethnoarchaeological studies of modern and prehistoric hunter gatherer groups validate the importance of this model ( Bird and . DMB t akes into account how long it takes to find a particular resource and then how long it takes to collect and process it once it is located (Kelly, 2013) . By considering the foraging strategies of different animals, we can begin to understand the complexiti es about human behavior.

PAGE 21

8 Marginal Value Theorem The marginal value theorem (MVT) was first proposed by Eric Char nov (1976). MVT is a widely used optimality model that describes the optimal foraging behavior of an individual or group in a system where resources are geographically dispersed in discrete patches (Calcagno et al., 2014; Charnov, 1976; Kuhn and Miller, 2015) (Figure 2.1) . These patches are often separated by areas with no viable resources. MVT is also useful in those situations where individuals face declining return rates while foraging (Kelly, 2013: 65). Earlier r esearch has shown that in order to capitalize on foraging return rates , foragers will leave a resource patch once the harvest rate reaches the average rate of the remaini ng resource patches (Charnov, 1976; Kelly, 2013). A patch can be something as small as a single food source or as large as natural habitat zone (Kuhn and Miller, 2015). Foragers tend to move out of a resource patch before the return rate out of that patch has dropped to zero (Kelly, 2013). Technological Investment Model Archaeologists are particularly drawn to foraging models, because the evidence of past subsistence practices and the social use of space are critical components of the archaeological record. Early foraging models do not typically take into account the effects of subsistence technology (Bright et al., 2002). Nevertheless, the technological investment model does just that (Bettinger et al., 2006; Kelly, 2013). The model estimates how much time an individual and/or group should invest in any specific type of technology. Most technologies are manufactured to help minimize resource handling time (Kelly, 2013). Ethnographic examples have shown that the initial construction of more complex tools and technologies can be huge. For example, the up front costs of fish weirs and hunting nets require a great deal of time to manufacture (Olson, 1936; Bailey and Aunger, 1989). Hunting nets can take months to build (Satterthwait, 1 987).

PAGE 22

9 That time increases with any raw materials that are collected for the rope and the time necessary to create it ( Lindström , 1996; Olson, 1936; Satterthwait, 1987). Complex tools and technologies also require regular maintenance once they are constructed. When compared to the cost of initial construction, r outine maintenance is a lesser expense since it is often carried out during down time (Kelly, 2013). Manufacturing complex technology often requires large investments of time that coul d potentially be spent acquiring or processing food, though with a less efficient form of technology (Kelly, 2013). An y increases with a slightly more complex form of technology is expected to be small, while major differences in technology likely represen t a more significant increase in return rates (Kelly, 2013) , such as with the difference between a fishing spear and gill net (Bettinger et al., 2006). The fishing spear offers a lower return rate than the gill net but requires less time to manufacture. Handling time and return rates are directly associated with manufacturing time (Kelly, 2013). Figure 2.1. Graphic representation of the marginal value theorem. Figure from Kuhn and Miller ( 2015: 175).

PAGE 23

10 By using the technological investment model, we can determine when a specific technology will be enhanced or when a more complex form of technology replace s another (Bettinger et al., 2006 ; Buonasera, 2015 ). Using the earlier example, a gill net can bring in a nice haul but at a lower return rate than the fishing spear if we factor in the manufacturing time. The technological investment model offers several insights (Bettinger et al., 2006): when technology changes it is believed to do so ( a ) quickly , ( b ) extensivel y across a population , and (c) usually permanently . C hanges in technology can often be explained with the help of this model. When two coexisting forms of technolog y are used to acquir e or process the same resource , they should have similar rates of return (Kelly, 2013). Coexisting forms of technology are often used for different purposes . This model is used to explore the functional differences between bedrock ground stone and portable ground stone found on the Chaquaqua Plateau in Southeastern Colorado and will be covered in a subsequent chapter. Middle Range Theory In the 1960s, the term middle range theory was first applied to archaeology (Pierce, 1989) . Middle range theories provide researchers with the ability to connect human behavior and natural processes to material remains (Raab and Goodyear , 1984). Many of t hese theories c o me directly from experimental research and ethnoarchaeological studies. Material remains uncovered in the present allow archaeologists to make inferences about past human behavior. Hunter gatherer behavior is highly variable, and so we can expect to s ee a variety of signature patterns in the archaeological record. Ethnoarchaeological studies often attempt to relate human The spatial and temporal distribution of hearths, flaked debris, and/ or other material remains found at a site are used to reconstruct group size and/ or how long a site may have been

PAGE 24

11 occupied (Kelly et al., 2005, 2006). Similarly, the patterns and distribution of bedrock ground stone features can aid in our understanding of prehistoric technology, landscape use, and subsistence practices (Lynch, 2017 c ). Hunter gatherers use different mobility strategies to schedule the collection and processing of resources. Binford (1980) argued that individual and group movement is predicated on different variables including demographics, ecology, and resource availability . He distinguished logistical mobility (the movement of specially organized task groups on short trips from a residential base) fr om residential mobility (the movement of all members of a residential base from one location to another) (Binford, 1980, 1990 ; Fitzhugh and Habu, 2002 ) . Based on these differences, Binford identified these two basic settlement patterns that allow researchers to better understand subsistence practices . Forager / Collector Model A rchaeologists have studied hunter gatherer mobility as a continuum with residential mobility on one end and logistical mobility on the other (Binford, 1980) . The study of ea rly hunter gatherer mobility patterns and settlement systems is crucial in our understanding of the archaeological record. When discussing resource availability, seasonal mobility among hunter gatherers can be expectedly modeled (Binford 1980, 5 6). Bi based on residential and logistical mobility . There are key differences between these two basic settlement patterns. Foragers move their group to the resources , while collectors move the resources to the group (Binford, 1980). Resource distribution often influences residential mobility patterns (Binford , 1980; Grove , 2009; Kelly , 1983; Kelly and Todd , 1988; Kent , 1992). Whallon (2006) argues that establishing and maintaining a social network among foraging gr oups is especially important during times of resource scarcity.

PAGE 25

12 I n a residential mobility system, individuals tend to forage for short periods of time within the general vicinity of their base camp (Kelly, 2013 ; Whelan et al., 2013). Hawkes et al. (1995) notes that during these foraging trips, it is typically easier to bring their children along with them. Foragers in a logistical mobility system often use residential bases for an extended period of time (Whelan et al., 2013). W omen in particular are fo rced to make longer trips to acquire resources, since those nearby become exhausted . However, these extended foraging trips are energetically costly because they increase the cost of transport with the accompanying children (Surovell, 2000). Native American women often had to choose between bringing their children along with them on their extended foraging trips or leaving them behind under the care of others for an undisclosed amount of time (Whelan et al., 2013). Consequently , the opportunity cost s of foraging are lower for women in residentially mobile system s (Whelan et al., 2013). Collectors, on the other hand, have primary residences based on resource distribution . Instead, r esources are logistically moved to the group (Binford, 1980). The r ecovery of food storage pits at a site frequently indicates the presence of a primary residence (Binford, 1980: 5). Collectors often move great distances to acquire resources , increasing the energetic costs of collecting. By studying mobility strategies in these terms , archaeologists are able to better understand the conditions where resources and hunter gatherers move in relation to one another (Binford 1980; Kelly 1983). Archaeologists often look for evidence in the archaeological record (fl ora and fauna, settlement patterns, and distribution of site features and material remains) to help them determine which of these mobility strategies a group is using. Foragers and collectors have distinct behaviors that set them apart . F ood procurement strategies is one such example . Foragers t ypically gather their food daily and do not usually store their foods (Binford , 1980 : 5).

PAGE 26

13 Unlike foragers, collectors store their food for part of the year and use specially organized task groups (Binford, 1980: 10). They regularly venture out to acquire specific resources , as opposed to generalized searching. The forager/collector mobility systems represent the extrem es of a wide range of possible settlement patterns. Most groups exist in an in between zone, with a complex mix of behaviors. Case Study These theoretical models can help us to understand how Trinchera Cave fits into a broader use of the landscape. In p articular, the large collection of bedrock ground stone features in the area is technology directly tied to the subsistence system. Bedrock surfaces have been used throughout Arizona, California, Kentucky, Nevada, New Mexico, and Texas to process a variety of food resources (Basgall, 1987; Buonasera, 2016; Claassen, 2011; Gayton, 1948a, b; Madsen, 2003; Mohr, 1954; True, 1993; Wallace and Holmlund , 1983; Webb and Funkhouser, 1929). The bedrock grinding surfaces found in southeastern Colorado were pr obably used for similar reasons. The m obility strategies of hunter gatherers in the area are indicated not only by food exploitation but als o a possible food storage pit at the site . Resource procurement often involves planning the exploitation and stora ge of certain foods. Nowak and Gerhart (2002 : 12 ) suggest that Trinchera Cave was used on a temporary, but regular basis by people living in the region. The introduction of maize and squash sometime between AD 500 and AD 1000, may indicate that people wanted to diversify the food resources in the area , allowing them to spend more time at the site (Nowak and Gerhart, 2002: 12). The presence of food stor age pits at the site would support such an argument. Simpson (1976 : 23 ) references a single storage pit that was five inches deep and 20 inches in diameter. She suggests that :

PAGE 27

14 ning other than the oxidation of the plaster. Since the pit is located inside the structure adjacent to While this may have been a food storage pit, Simpson does not say anything about the possible contents (1976: 12) . Nowak and Gerhart (2002: 79) suggest the picture is not as clear cut. They report finding three pockets of Hackberry seeds at the site . R odents often collect seeds for the purpose o f storage ( Brown and Davidson, 1977; Chaney, 1936) , and this is a possible reason for t he pit described by Simpson . Ground disturbance at the site may be one reason why Nowak and Gerhart (2002) did not locate any storage pits (Nowak and Gerhart, 2002: 12). However, it may also represent the residential mobility patterns of the inhabitants.

PAGE 28

15 CHAPTER III BACKGROUND Introduction Indigenous peoples have occupied the tributaries and side canyons of the Pu rgatoire River in Southeastern Colorado for thousands of years. E vidence reveals habitation by early woodland groups, horticulturists, and plains hunter gatherers ( Black, 2000; Lynch , 2010). A n un usual feature of this region is the presence of bedrock gr inding areas appearing at different elevations and with different site assemblages . While these features have been found throughout region , t hey are especially concentrated in the canyons of the Chaquaqua Plateau (Lynch, 2010, 2014, 2017a, 2017b, 2017c) . Bedrock gr ound stone features are an important part of the native landscape in northeastern New Mexico and s outheastern Colorado. The highest concentration of these features seems to be alongsid e the canyon walls of the Apishapa, Chacuaco, Cuchara, and Purgatoire Rivers and their respective tributaries (Campbell, 1969; Loendorf, 2008; Lynch, 2010; Owens, 2007). Bedrock ground stone features were frequently constructed along the sandstone canyon walls , al though some have been found in other locations (Andrefsky, 1990; Campbell, 1969; Chomko and De Vore, 1990; Gunnerson et al. , 1989; Hartley and Vawser, 2003; Loendorf, 2008; Renaud, 1931). Trinchera Cav e (5LA1057) is a prehistoric site located about 48 km east of Trinidad in Las Animas County, Colorado (Fig ure 3. 1 ) . It is in fact a naturally occurring rock shelter and not a cave in the geological sense. S ome early researchers have referred to the site as Trinchera Shelter . However, the name Trinchera Cave has dominated the literature (Zier, 2015). For continuity purposes, the site is referred to as Trinchera Cave in this thesis. The site is situated near the confluence of T ri nchera Creek and the Purgatoire River , roughly t hirty tw o kilometers

PAGE 29

16 Figure 3.1. Relief map showing the location of Trinchera Cave in S outheastern Colorado. Figure from Zier (2015: 4).

PAGE 30

17 north of the New Mexico border. Trinchera Cave is located on school district land (Simpson, 1976) and is administered by the Colorado State Land Board through their Alamosa District Office (Black, 2000). The land is currently leased for cattle grazing to Mr. Antonio Baca. Recently, efforts have been mad e to consolidate the bulk of predominantly unpublished work that has been recorded at the site since its initial discovery in the late 1940s (Black, 2000; Gerhart, 2001; Zier, 2015, 2017). Little research has focused on the function and relationship of th e site to the regional archaeology. In fact, no research has been conducted on the ground stone materials discovered at Trinchera Cave. Archaeological discovery represents only a fraction of the overall picture. Zier (2017) estimated the sheltered area of the cave to be around 500 m² (Figure 3.2). However, the actual habitable area is calculated to only be around 280 m² due to the presence of several large sandstone blocks that broke off from the cliff face and f ell into the northeastern half of the shelter (Zier, 2017). This area was apparently not used by the inhabitants to any considerable degree based on the lack of cultural deposition (Zier, 2017). Between 1949 and 2001, different archaeologists excavated w ithin the shelter (Figure 3.3). Trinchera Cave lies within a fairly small canyon (20 m deep x 75 m wide) (Figure 3.4). The shelter is some 25 30 feet above the stream bed and is only 6.5 km away from the Purgatoire River. The shelter is cool in the su mmer and warm in the winter based on its southeasterly exposure (Simpson, 1976). The somewhat secluded environment is close to major sources of water, making it possible to access other nearby regions. Rivers often serve as travel routes, helping to conn ect people (Noble, 2000: 23). Habitat corridors enhance the population capability of hunter gatherers (Beier and Noss, 1998). This accessibility may have allowed for trade with other regions. The presence of imported ceramics and marine shells supports this argument. Cattle frequent the canon bottom

PAGE 31

18 where the site is situated, and have impacted the surface of the shelter. The development of arroyos in the area are thought to be contemporaneous with the development of ranching, and are believed to be caused by roaming cattle (Duce, 1918). Cattle have affected the land by wearing the trails down and destroying the local vegetation, resulting in stronger opportunities for soil eros ion and weaker opportunities for absorption (Duce, 1918). Simpson (1976) believes that the site represents a combination of Plains and Southwestern cultures. Nowak and Gerhart (2002: 104) suggest that Trinchera Cave was a cultural melting pot of people s. This belief is based on a number of pottery sherds, shell fragments, imported calcium carbonate, obsidian tools, and other diagnostic materials that represent trade with these different cultures. Simpson (1976: 203 204) uncovered three Figure 3.2. Trinchera Cave site map. Figure from Zier (2015: 11).

PAGE 32

19 different types of dateable ceramics including Sante Fe Black on White (AD 1200 1350), stamper Cordmarked (AD 1176 1485), and Taos Plain (AD 1100 1300). One Sante Fe Black on White sherd was retrieved from Level II in Area D , as well as thirty eight Taos Plain sherds from multiple levels (Simpson, 1976: 155 156). Nowak and Gerhart (2001, 20 02) also found direct evidence of Sante Fe Black on White and Sopris Plain pottery from the same occupational level. Early Researchers at Trinchera Cave I nformation is provided not only on individuals who work ed at the site, but also in the broader area to provide a more nuanced background and archaeological context . Figure 3. 3. Map of Trinchera Cave showing the locations of Chase, Dick, Simpson, and Nowak excavation units, 1949 2001 . Figure from Zier (2015: 50 ).

PAGE 33

20 Etienne Renaud (1930 1941) Etienne B. Renaud was a faculty member at the University of Denver (DU) between 1920 and 1948. He came to the United States to teach French, and had no previous knowledge of anthropology (Nelson et al. , 2001). However, Renaud proved himself to be a hard worker and an avid reader. He founded t he anthropology department at DU in 1922. During his time there, he developed new archaeological survey methods and was one of the first people to attempt systematic excavations in Southeastern Colorado ( Renaud , 1931 1947 ) . Together with his students, Renaud actively searched for prehistoric sites in Colorado and other nearby states. They keenly sought out farmers and ranchers to find out what might be on their land. In 1929, Figure 3. 4 . Overview of Trinchera Cave, taken June 2018, view to the northwest from the opposite canyon rim. Trinchera Creek is visible at the bottom.

PAGE 34

21 Renaud conducted a field expedition to the Lindermeier Site in northeastern Larimer County, CO for the Colorado Museum of Natural History (now known as the Denver Museum of Nature and Science) (Renaud , 1931 1947) . Paleoindian artifact s were discovered at the Lindermeier Site by Judge Claude C. Coffin and his son, A. Lynn Coffin five years earlier ( Coffin, 1937; Wilmsen and Roberts Jr, 1978) . When Renaud arrived at the site, he identified several of the artifacts as Folsom points. An extraordinary collection of bison bones and Folsom points were uncovered during the 10 years of site excavations (Coffin, 1937). He later conducted different archaeological surveys of the Great Plains. Some of the areas he surveyed included: Eastern Colorado (1930 1933), Eastern Wyoming (1931), Western Nebraska (1933), Northeastern New Mexico (1934 1935), Southern Wyoming (1935 1939), and Southern Colorado (1940 1943) (Renaud , 1931 1947). Renaud documented hundreds of different prehistoric sites in Eastern Colorado (Renaud , 1931, 1935, 1931 1947; Renaud and Chatin , 1943) , with 30 of those sites being in Las Animas County (Black , 2000). Renaud was an avid explorer, documenting many of the sites he came across. His field notes, journal ar ticles, unpublished survey manuscripts with artifact sketches and original site map drawings, photographs, postcards, microfilm, site forms, and early newspaper clippings are stored at the University of Denver . In 1930, the first descriptions of the Snake Blakeslee Archaeological Site were published (Gunnerson et al. , 1989). Renaud first came across the site earlier that same year when he surveyed different parts of Eastern Colorado. Reconnaissance surveys were primarily conducted during this time ( Lintz , 1999; Black , 2000). He was guided to the Cramer and Snake Blakeslee sites by R. D. Mutz, a man from the nearby town of Fowler (Renaud , 1931, 1935, 1931 1947 ; Lintz , 1999 ) . In 1931, Renaud returned to the same sites and again a decade later. He worked

PAGE 35

22 on writing new descriptions of the Cramer and Snake Blakeslee sites. After his investigation, Renaud believed that the two sites each had a ceremonial function based on his discoveries (Renaud , 1931 1947) . Preliminary excavations were later cond ucted at Trinchera Cave as a part of the project (Renaud , 1931). In 1948, Renaud chose to retire from the University of Denver. The University of Denver Anthropology Museum (DUMA) currently stores the portable ground stone artifacts Renaud discovered during his field expeditions in the Apishapa Canyon (Rohde, pers onal comm unication , 2018). However, upon examination of the collecti on, I Cramer site, and that was a slab metate fragment made out of sandstone. There were no materials from Trinchera Cave present at DUMA. W hen going through the minimal amount of paperwork they did have from those earlier excavations, I learned from a small note collection that two manos and one metate from an undisclosed site were thrown out at some point by Arnold Withers, a doctoral candidate f rom Columbia University. It is unknown at this time why or where they were discarded or even if they were from any of the sites mentioned here. Haldon Chase (1949 1950) In 1949, Haldon Chase was a graduate student at Columbia University. Chase directed a project in the area with the primary goal of understanding early Apache sites on the plains of Southern Colorado (Gunnerson et al. , 1989; Lintz , 1999). The High Plains Columbia Expedition (HPCE) of 1949 was a joint expedition with Columbia University and the University of Denver ( Gunnerson et al. 1989; Lintz 1999; Zier 201 5 , 2017 ) . The funding for the expedition came largely from the Missouri River Basin Project (Lintz , 1999) . Robert (Bob) Stigler joined the

PAGE 36

23 expedition early on. He was also a graduate student from Columbia University. As a back up strategy, Ferdinand (Ferd) Okada was also associated with the project (Okada , 1949a, 1949b). He was a doctoral degrees in California (Lintz , 1999). T he three men made their way to Apishapa Canyon in Southeastern Colorado (Gunnerson et al. , 1989; Lintz , 1999) . Hal Chase was selected as fie ld director for the project, even though he had just recently graduated from Columbia with his BA (Lintz , 1999) . This decision was made because he was native to Colorado and had originally envisioned the project (Lintz , 1999) and had fieldwork experience in Colorado . However, the HPCE was under the general direction of Arnold Withers from Columbia University (Lintz , 1999). The three men (Chase, Okada and Stigler) spent more than five weeks excavating in the canyon during their 1 949 expedition (Colorado Encyclopedia, 201 6 ; University of Denver Anthropology Museum, 2018) . The men made a brief stop at the Cramer s ite sometime in mid July. Most of their time was spent excavating at the Snake Blakeslee site, located near the mouth o f the Apishapa Canyon (Encyclopedia Staff , 2016; University of Denver Anthropology Museum , 2018). At the time, Trinchera Cave was little more than an afterthought during the project . Chase decided to excavate at the site for a single day on August 21, 1949 (Zier , 2017), having primarily worked at the Snake Blakeslee site. He was joined by Jack Gilstrap, Abe Mason (local informant), Ferdinand Okada (Ph.D. candidate), Morris Taylor (history professor at T SJC ), Richard and Willard Louden (land owners) , an d a n amateur archaeologist n amed Stubby Ischam (Lintz , 1999). Together, they worked in the eastern most portion of the cave, referred to now as Area A (Zier , 2017) . According to the University of Denver Anthropology Museum , Hal Chase and Bob Stigler spen t a majority of their time out at the site. However, Christopher Lintz (1999) paints a

PAGE 37

24 different story. Stigler was allegedly forced to abandon the project after only 53 days (Lintz , 1999). He left for Arkansas to be with his family because his father w as having an emergency surger y (Lintz 1999) . After the High Plains Expedition wrapped up, Chase later returned to the area by himself on September 10, 1949 (Lintz , 1999). He spent 1 1 days re investigating the Snake Blakeslee site (Lintz, 1999) , which is located due north of research laid the groundwork for future archaeological work in the region , including Trinchera Cave . All the archaeological materials tha t were discovered as a result of the H PCE were transferred to Columbia University in 1949 (Lintz, 1999) . It is believed that some of the larger ground stone items may have been left in the field ( Okada , 1949c; Lintz , 1999: 20). This seems plausible due to the weight of some of the metates from the later expeditions currently being housed at the Louden Henritze Archaeology Museum. In 1955 , t he materials were reportedly shipped from Columbia University to the Univers ity of Denver (Zier , 2017). The location of those excavated materials from the expedition is currently unknown. Sometime between September 10 and September 21, 1949, Chase ventured to Trinchera Cave to take photographs of the site. However, the location of these black and white photographs is unclear . They are most likely housed at the Louden Henritze Archaeology Museum in Trinidad , Colorado (Martin, personal communication, 2018a), but have not been located to date. Trinchera Cave was first noticed by William Louden in 1948 during a reconnaissance flight (Black , 2000). Ruth Henritze, William Louden and his brother Richard, later followed up at the site by conducting a test excavation. While there, they discovered a made out of juniper, and unfired clay figurines (Louden and Louden , 1998) . These materials are currently stored at Louden Henr it ze Archaeology Museum. T he Loud ens chose to include

PAGE 38

25 Haldon Chase in their work . Chase was teaching at TSJC at the time, where he had established their anthropology program ( Anonymous , 1952; Cassells , 19 97 ) . Chase began more formal excavations at Trinchera Cave in the early 1950s (Black , 2000). Between June 19 July 1, 1950, Chase and his crew excavated blocks in Areas A and C at the shelter (Zier, 2017). After the 1950 excavation had ended, Chase sent all the botanical remains to the University of Michigan Ethnobotanical Laboratory (Zier, 2017 ). The results are discussed later in this thesis. The recovered faunal remains were reportedly sent to the Colorado Museum of Natural History for identification (Zier, 2017). However, according to the archivist and collections data manager at the Denve r Museum of Nature & Science, Sam Schiller (personal communication, 2018), no records of these materials could be found . After conferring with a Navajo expert in La Jolla, California, Chase began to believe that all the materials recovered at Trinchera Cave were more than 1,500 years old (Chase , 1950; Lintz , 1999). Little information about this meeting exists. Chase remained with TSJC until 1953. Prior to his departure from the school , he helped select Herbert Dick to replace him at the coll ege (Lintz , 1999) . Dick picked up where Chase left off, by continuing the excavations at the cave during the 1954 195 6 field seasons (Black, 2000) . Very little information exists from the actual fieldwork conducted by Haldon Chase and Herbert Dick, with the exception of a brief summary in thesis (1976 : 5 ) . A short newspaper article published by the Rocky Mountain News (RMN) is also known to exist (Gavin , 1955). The short article discusses the archaeological finds at Trinchera Cave Herbert W. Dick (1954 1957) In 1953, Dr. Herbert W. Dick replaced Haldon Chase as a faculty member at TSJC. He remained with the school until 1962 when he moved to Alamosa, Colorado to teach at Adams

PAGE 39

26 State Coll ege (Zier, 2017). Dick began his excavations at Trinchera Cave in 1954, soon after he was hired. Working predominantly with students from TSJC, Dick was able to carry out large scale excavations at the site . Based on photographs and limited field records, Zier (2017) determined that Dick had eight people working with him on any given day. They excavated blocks in Areas A, B, and C in the shelter, with Area B being taken all the way down to bedrock. However, t he greatest amount of work was done in Area C (Zier, 2017) . All the excavated materials were processed at a field laboratory that Dick had established at his base camp (Zier, 2017: 28). When Dick left TSJC , he brought m any of the artifacts that were recovered from the shelter along with him (Zier, 2017) . However, he never analyzed any of the materials or wrote up a report on them. When Caryl Wood Simpson began excavating a part of the cave in the 1970s, she and William Louden were able to recover housed at the Louden Henritze Archaeology Museum. Simpson was also able to obtain a few pages of notes, a sketch map of the site, and some black and white photographs from Dick (Zier, 2017). Based on few pages of notes, she produced an artifact i nventory of the retrieved collection (Simpson, 1976: 179 199) . However, t here is no way to know for sure if Dick gave Simpson and Louden all the materials he took from Trinchera Cave. Zier (2017) reports that besides a schematic sketch from 1956, s . Colorado Archaeological Society (1966 and 1969) Further excavations at Trinchera Cave were carried out by the Trinidad Chapter of the Colorado Archaeologica l Society in the 1960s. However, not much information is currently known about these excavations. The Trinidad Chapter of the Colorado Archaeological Society

PAGE 40

27 no longer exists. They disbanded in 1971 (Martin, personal communication , 2018 b ). Most of their old newsletters are currently stored at the Louden Henritze Archaeology Museum in Trinidad, Colorado. Martin (personal communication, 2018b) believes t hey may have collected surface finds and that is what was recorded in 1966 . Caryl Wood Simpson (1974) Caryl Wood Simpson was attending the University of Wyoming as a graduate student when she carried out her excavations at Trinchera Cave (Zier, 2017) from April 15 26 and again from June 12 July 19, 1974 (Wood, 1974). Later that fall , she wrote a preliminary report on the site, its excavation procedures, and its contents (Simpson, 1976) . Two years later, s he completed work from the 1950s. Simpson primarily worked in Area D of the shelter , southwest of where Dick had previously excavated. Between April 15 26, 1974 Simpson managed an archaeological field school for high school students from the Denver area. She later revisited the site be tween June 2 July 19 with students from TSJC to conduct further excavations. Simpson (1976: 156) retrieved an assortment of various ceramic artifacts from Trinchera Cave. A number of unfired ceramic objects were uncovered from Level I inside of a suspected (Simpson, 1976 : 156 ) . Several small bowls no larger than one to two cm in depth and 5 cm in diameter were among them. All the retrieved bowls are reportedly quite fragile (Martin, personal communication, 2018a) and are curren tly housed at the Louden Henritze Archeology Museum in Trinidad, CO. Although their function is unknown at the present time, they may have belonged to children because of their small size and fragility (Simpson, 1976: 156). Simpson (1976: 23) noted the presence of a pit and timbers from a collapsed structure. A layer of matting comprised of grasses, small juniper branches, and other

PAGE 41

28 plant material was discovered at the shelter and may represent roof fall from several collapsed structures (Nowak and Ger hart, 2002: 9). The additional protection likely meant that the site was used during the winter. Much of the portable ground stone was uncovered under the remains of the collapsed structures. By the end of her 1974 field season, Simpson uncovered a min suggests that the shelter was utilized on a regular basis from the Late Archaic Period (1050 BC 100 AD) to the Late Prehistoric stage (AD 100 1725). Simpson left behind a set of field notes and artifact inventories . Her thesis is considered the first formal documentation of archaeological research conducted at the s ite and it contains the first known partial map of the site (Zier, 2017) . In 2013, Centennial Archaeology, Inc. created a more thorough map of Trinchera Cave and its immediate surroundings using a Gowin TKS electronic total station (Zie r, 2017). Contemporary Researchers at Trinchera Cave Office of the State Archaeologist (1997 1999) The Program for Avocational Archaeological Certification (PAAC) is an educational program designed for avocational and professional archaeologists alike. PAAC was first established in 1978 by the Office of the State Archaeologist of Colorado (OSAC) and the Colorado Archaeological Society (CAS). The program was designed to allow CAS members and those interested the opportunity to work at an actual archaeological site and gain expertise rather than pursuing an academic degree. Between 1997 and 1999, Kevin Black and CAS volunteers completed a comprehensive survey of nearly 650 acres in the Trinchera Cave Archaeological District (TCAD) through the PAA C program (Black, 2000) . The volunteers

PAGE 42

29 created a new map of the shelter , recording the different rock art panels that were present. 53 indigenous sites within the district were also documented. Eighteen isolated finds and 57 sites were recorded for the first time, including Native American and non Native American materials (Black, 2000). The non Native American materials and historic sites are (2000) comprehensive report . T he presence of ceramic artifacts , metal artifacts, and rock art (the latter being found solely at Trinchera Cave) suggests more recent use in the area by the Apache, Comanche, and/or Kiowa groups (Black, 2000). Human occupation in the TCAD appeared to be fairly intense. Very little research has focused on the function and relationship of Trinchera Cave to other regional sites. Michael Nowak (1999 2001) Michael Nowak was a faculty member at Colorado College (CC) when he conducted extensive excavation work at Trinchera Cave over the course of several years (Nowak and Budnick, 2000; Nowak and Gerhart, 2001, 2002). The primary goal of the project was to discover what else remained at the site (Nowak , 2009). Specifically, t he purpose of research wa s to locate any un excavated areas of the site , determine the stratigraphy prior to earlier excavations , and to obtain radiocarbon dates from site sediments (Nowak and Budnick, 2000) . Nowak and his team excavated blocks in Areas A, C, and D. Since Dick had excavated Area B to bedrock , there was presumably no need to dig there again (Zier, 2017: 29, 42) . Nowak conducted most of his work inside the shelter, similar to the earlier researchers. However, h e was the first to extend test units just beyond the dripline to help uncover any exterior deposits that may have been present (Zier, 2017). Additional units were established

PAGE 43

30 belo w the shelter to help determine how the archaeological materials had moved d ownslope. Nowak elected to extend some units while abandoning others based on what was unearthed . This work produced 12 radiocarbon dates from organic materials at the site (Zier, 2017) (See Table 3.1). Nowak uncovered evidence of a midden directly under the dripline. Besides Trinchera Cave , middens containing abundant and diverse cultural artifacts w ere only found at one other site in the region, suggesting that most site occupations were relatively short lived (Black, 2000: 66) . Many of the portable ground stone fragments were found in this midden and under the collapsed structures (Nowak and Gerhart, 2002). Archaeologists frequently rely on stratigraphy to help reconstruct site history before, during, and after site occupation. Nowa each of the previous excavations as well as heavy looting at the site. Chase noted that looting may have begun as early as 1949 (Gerhart 2001). According to the site foreman Buford Garcia (personal communication, 2018), more serious looting has been a problem at Trinchera Cave since the early 1970s. The looting got so bad that in the late 1990s a warning sign was posted on the northwest side of Trinchera Creek to deter future activity (Garcia, perso nal communication, 2018). The sign has since been removed by looters. More protection is certainly needed for this site. It is recommended that Trinchera Cave be periodically monitored with updated signage to help deter future looting and vandalism. Col orado College ( 2002 2003) In 2002, CC student Kylie Crocket cataloged between 75 collections housed at TSJC (Crocket, 2002; Martin, personal communication, 2018a). In 2003, C C student Laura McCarthy and visiting assistant professor Carl Geo rg (Charly) Bank carried

PAGE 44

31 Table 3.1. Radiocarbon Dates from Trinchera Cave. Table from Zier (2015: 52 53).

PAGE 45

32 Table 3.1 cont. Radiocarbon Dates from Trinchera Cave. Table from Zier (2015: 52 53).

PAGE 46

33 out a preliminary geophysical study with remote sensing at Trinchera Cave. Their primary goal was to locate the foundation walls of the jacal structure Simpson (1976: 23) had mentioned in her thesis. They conducted seismic, geodetic, magnetic and resistivity surveys (McCarthy and Bank , 20 03). Although recent excavations failed to turn up any evidence of the structure, resistivity profiling provided the most reliable results (McCarthy an d Bank, 2003). A structure was located in the southwest portion of the cave which they interpreted as t he foundation walls (McCarthy and Bank, 2003). Christian J. Zier (2013 2015) Between 2013 and 2015, Chris Zier of Centennial Archaeology, Inc. (CAI) created a more comprehensive map of Trinchera Cave. Using available sources, Zier and CAI reconstructed the locations of previously worked areas in the shelter. These sources included field notes, early newspaper clippings, an incomplete catalog of previ ously excavated artifacts, markings on the rear wall of the shelter, s , a personal interview (with Chase), photographs, profile sketches and original site map drawings , and various published reports (Zier, 2015, 2017). Zier (2015) assessed the archaeological potential of the different collections from the site that are housed at the Louden Henritze Archaeology Museum. Recent radiocarbon dates from the Trinchera Cave collections yielded two Paleoindian/Folsom period dates of 9275 9160 BC (Cal. 11,225 11,110 BP) and 8610 8320 BC (Cal.10,560 10,270 BP) (Zier, 2015: 54 55) (See Table 3.1). Both deep soil samples, 90 te that the artifacts uncovered at those levels also date to the Paleoindian/Folsom period. is Juniperus sp.

PAGE 47

34 (juniper) (Black, 2000; Nowak and Gerhart, 2002; Zier, personal communication, 2018). Zier (personal communication, 2018) suggests that the actual dates could be much younger than the radiocarbon assays. Samples often provide misleading results if materials of different ages accumulate in the same archaeological context.

PAGE 48

35 CHAPTER I V GROUND STONE TECHNOLOGY Ground stone tools represent an important aspect of the archaeological record and can contribute much to our understanding of past lifeways. They can offer insights into areas such as diet, division of labor, food processing techniques , mobility patterns, settlement systems, and othe r specialized activities (Ebeling and Rowan, 2004). These grinding tools are used to process food items such as amaranth seeds, beans, cacao, chiles, maize kernels, squash seeds, sunflower seeds, and tomatoes as well as clay, pigments, and temper (Searcy, 2011 : 5). They can also be used to process items such as ceremonial and medicinal herbs. Ground stone tools can come in different shapes and sizes. Different methods have been developed to help determine the specific uses of these implements. Portable Ground Stone broad term commonly used by archaeologists t o refer t o stone artifacts that are modified by or used to modify other materials through different reduction techniques such as abrasion , pecking, or polishing (Adams, 2014 a ) . This includes abraders, hammerstones, manos and metates, mortars and pestles, and polishers as well as the variety of artifacts that are shaped using these techniques . Food grinding t ools ha ve played a vital role in the description of prehistoric subsistence strategies (Adams, 1999) . Experiment al work carried out by Jenny Adams (20 14a ) on different metate types similar to those found in the archaeological record of the American Southwest provi des a model for understanding the tool design, manufacture, and use of metates all over the world . Adams (20 14a : 69) argued that vesicular materials provide a much better working surface on which to grind large grained food resources because it shears ap art the grain while leav ing little gravel behind during processing.

PAGE 49

36 Vesicular basalt is one of the most popular choice for grinding seeds and dried corn kernels today (Searcy, 2011: 83). (20 14a ) research also investigated the efficiency of different tool designs. She looked at different metate types based on those discovered in the southwest; including basin, concave / flat, and trough metates. Her research demonstrated that concave / flat and troug h metates were each effective in processing soaked corn kernels (Adams, 20 14a : 68 69). Adams (20 14a : 106) also found that trough metates could grind more grain at a time than concave and flat metates. Looking at a number of ethnographic studies on th e relationship of mano size and agricultural dependence, Hard (1990 ) found that manos with an increased surface area could process more grain than smaller designs. Diehl (1996: 106) combined macrobotanical and ground stone data to support this relationsh ip. The processing of corn and other large grained food resources became more important over time, resulting in changes to ground stone tool morphology (Stone, 1994). Grinding corn with manos and metates is very time consuming (Cushing, 1920; Hard, 1986; Lancaster, 1983 ; Spier, 1933: 79 ). Ethnographic studies have shown that efficiency is improved by increasing the number of ground stone tools used for reciprocal grinding and the size of the grinding surfaces (Hard, 1986, 1990; Horsfall, 1987; Mauldin, 1 991, 1993). A strong correlation exists between the length of the grinding surface, as indicated by mano length, and the importance of agriculture to the subsistence economies of modern groups (Hard, 1986, 1990; Mauldin, 1991, 1993). By increasing the si ze of the grinding surface, processing speed increases and it allows for more grain or meal to be processed at a time (Mauldin, 1993). Hard (1990) identified a strong correlation between mano length and the level of agricultural dependence. His study ind icated that manos smaller than 19 cm often have a low reliance on agriculture, and those between 17

PAGE 50

37 and 25 cm in length tend to have a moderate reliance on agriculture (Hard, 1990; Stone, 1994). Manos that range between 24 and 31 cm on average generally h ave a high reliance on agriculture (Hard, 1990; Stone, 1994). I n Southeastern Colorado , other researchers have identified two types of manos : the one handed mano ( less than 18 cm ) , and the two handed mano (with a length greater than 18 cm ) ( Campbell, 19 69 : 56; Ireland, 1974: 83; Simpson, 1976: 105) . It is believed that the larger two handed manos may be more prevalent in the region after AD 1050 (Nowak and Gerhart, 2002: 127). As corn became more important, Lancast er (1983) argues that food grinding tools became more specialized. Experimental research led by Adams (1993b) indicated that ground stone tool morphology can be used to draw conclusions about prehistoric grinding efficiency. In areas of the southwest tha t were heavily dependent upon corn, there was an increase in the number of flat and trough metates (Adams, 1993b) and an increase in size of individual grinding surfaces as indicated by mano length (Hard, 1986, 1990; Mauldin, 1991, 1993). Using appropriat e types of raw material is shown to be just as important as tool morphology and grinding surface size for those dependent upon agriculture. Early attempts were made to identify specific tool types that can differentiate between the processing of wild food s and domesticated foods and to use those differences to determine when agriculture was first introduced to a particular region (Adams, 1999) . V ariability in tool size was once assumed to have been associated with the processing differences between wild, cultivated, and domesticated foods (Adams, 1999) . However, research has shown that tool morphology is more closely associa ted with different processing techniques than to resource procurement (Adams, 1999) . Using ethnographic data, Wright (1994: 242) argued that

PAGE 51

38 The diffe rent types of stone tools used in food processing do not sufficiently reflect the range of food resources being processed (Adams, 2014 a : 55). The same f ood item can be processed using more than one type of tool, and the same tool can be used to process different food types. Tool morphology can not be used to establish what type of food resource was being ground (Adams, 1999; Horsfall, 1987; Wright, 1994). Pollen and starch analyses performed on prehistoric ground sto ne strengthens this argument (Greenwald, 1993: 348 349; Lancaster, 1984: 257). The study of ground stone technology includes not only the tools themselves , but also the activities, behaviors, knowledge, and social contexts related to their manufacture, use, and abandonment (Adams, 1993b, 1995 a ; Bleed, 1986; Dobres and Hoffman, 1994; Kingery, 1989; Lemonnier, 1986; Nelson, 1996; Schiffer, 1992; Schiffer and Skibo, 1987). Technological development can be affected by new information on tool efficiency or design , new ly discovered resources to use on existing tools , new methods to process the same resources, and raw material availability (Adams, 1993b : 332 ) . Bedrock Ground Stone Indigenous groups have occupied the tributaries and canyon sidewalls of the Purgatoire River in s outheastern Colorado for several thousand years . Archaeological evidence reveals habitation by horticulturalists , P lains hunter gatherers, and E arly W oodland groups (Black, 2000; Lynch, 2010). A distinctive feature of this region is the presence of numerous b edrock grinding areas appear ing at different elevations and with different site assemblages. Bedrock grinding areas represent an important part of the native landscape in northeastern New M exico and southeastern Colorado . Documented in site descriptions since Renaud (1931) , the highest concentration seems to be located along the canyon side walls of the Apishapa, Chacuaco,

PAGE 52

39 Cuchara, and Purgatoire Rivers and their tributaries (Campbell, 1969; Loendorf, 2008; Lynch, 2010; Owens, 2007). Bedrock Grinding Surfaces ( B GS ) were often constructed along the sandstone canyon walls , though others have been discovered among rock outcrops (Andrefsky, 1990; Campbell, 1969; Chomko and De Vor e, 1990; Gunnerson et al., 1989; Hartley and Vawser, 2003; Loendorf, 2008; Renaud, 1931). A rchaeology in southeastern Colorado suggests a n array of lifeways and subsistence practices based on the remains of architecture, ceramics, grinding tools and lithics (Campbell, 1969; Eighmy, 1984; Zier et al., 1999) . However, BGS are difficult to put into an archaeological context (Campbell, 1969; Eighmy, 1984; Zier et al., 1999). There is little ethnographic and archaeological evidence of their use in the re gion. Unlike their portable counterparts, BGS features are a permanent part of the landscape (Lynch, 2010). This leaves them unprotected and susceptible to a variety of erosional forces ( Walker and Fratt, 1991). There are currently several ways to directly date bedrock grinding surfaces and to determine functionality through pollen or starch analysis (Cummings, personal communication, 2018; Lynch, 2010). Dating ground stone pulled from the archaeological record c an be accomplished through direct study , or may be deduced through association with nearby features such as hearths, rock art, and structures (Campbell, 1969; Chomko and De Vore, 199 0; Loendorf, 2008). Dating also can be determined using diagnostic materi als found in associative context. The small number of bedrock grinding surfaces that have been dated in the region through radiocarbon methods or by associative context with other features are believed to date from the Late Archaic Period (2000 RCYBP RC Y500 BP) (Campbell, 1969; Chomko and De Vore, 1990; Loendorf, 2008; Lynch, 2010; Owens, 2007). There are no historic al record s about the use of these grinding features in the southeastern Colorado , and research suggests that they

PAGE 53

40 stopped being used by the time Athapaskan speaking peoples began filtering into the region during the late 14 th to early 15 th century (Campbell, 1969; Gunnerson et al. , 1989; Hammond and Rey, 1940; Owens, 2007; Stoffle et al. , 1984; Zier et al., 1999) . Grinding Behaviors Design and Manufactur e The first step to understanding technological processes is to consider how an object was designed and manufactured , why the specific raw material was chosen , and what modifications were made to prepare the selected material for the function (Adams, 2014 a : 21). The design stage begins with choosing a material of suitable size and texture (Adams et al., 2009: 44 46; Delgado Raack et al., 2009; Horsf all, 1987: 340). Raw material variability includes rock size and granularity. Comfort features, such as grooves or handles, can make a tool easier to use and are often included in the manufacturing process (Adams, 2014 a : 21). Raw materials can be sele cted from a variety of different location s including the landscape, a nearby streambed, or from bedrock outcrops . Material size and weight are common selection criteria for certain tools. However, Stone (1994) notes that size choices are often restricted by resource availability for portable grinding stones. Raw materials have a natural granularity (fine grain to course grain) to them that is perceived as texture (smooth, rough, and so forth ) (Adams, 2014 a : 21). Texture is a critical attribu te to consider while manufacturing tools. A fine grain rock and a course grain rock may be chosen for different processing tasks because of their respective granularity. A fine grained rock can be used as a polisher . Meanwhile, a course grained rock can be fashioned into an abrader since it is rough enough to break down the surface material of another object (Adams, 2014 a use. However, the

PAGE 54

41 initial rougher texture ca n be restored through certain restorative techniques such as pecking (Adams, 2014 a ). Durability is another important factor to consider when selecting raw material type and refers to the ability of the material to withstand wear (Adams, 2014 a ). Especiall y when grinding food, it is important that the grinding tool be durable enough to not erode into the food . The behavioral constructs of tool design can be thought of in terms of soci oecological systems . A tool is considered to have an expedient design if the natural shape of the raw material was modified through use or just enough shaping to become functional ( Adams , 2014 a : 22). Modifications that make a tool easier to handle, improve efficiency, or achieve a shape not relate d to function suggest a strategic design ( Adams , 2014 a : 22). By looking at tool design for an assemblage as a whole , archaeologists can often determine whet her expediently designed objects are used or regarded differently from those of strategic design. Use The way a tool was designed, manufactured, and u tilized can be assessed in terms primary and secondary uses ( Adams , 1994, 1995b, 2014 a : 24) . Primary use is the task for which the tool was first intended . Portable g round stone tool s often serve a single function (Adams, 2014 a ) . S econdary use refers to a n addition al function on top of a primary use . M inimal research into the nature of BGS has been conducted . They may have served multiple functions. Adams (1994, 1995 b , 2014 a , 2014b ) defined several use categories that a tool may go through single use, reused, redesigned, multiple use, and recycled for portable ground stone that can inform research into BGS . Grinding efficiency is another important concept from portable ground stone studies that can be applied to BGS analysis. It is typically expressed as the amount of material that can be processed or the number of hours spent completing a specific task like food processing (Adams,

PAGE 55

42 2014 a , 2014b ). Different tool designs may be more or less efficient than one another. For grinding tools in particular, efficiency can be measured by the size of the grinding surface and the overall weight of the tools (Adams, 2014 a : 30 ; Hard, 1990 ). Like tool design, efficiency is a behavioral construct. Certain attributes of ground stone tools might make one more efficient than another. Hard et al. (1996) discusses the size of grinding surface as it relates to the amount of food processed . Studies have shown that by increasing the size of grinding surfaces , grinding efficiency increases as well ( Buonasera, 2015; Hard et al., 1996; Mauldin, 1993). The concepts of use intensity, use efficiency, and wear management strategies can be used to interpret the different technological processes of designing, manufacturing, and maintaining tool efficiency when viewed together (Adams, 2014a: 30, 2014b: 13 6 137). Ground stone tools also ha ve ritual or ceremonial significance, especially when associated with inhumations ( Adams, 2014 a : 27; Carle, 1941). Those deposited with burials are not usually modified, but those placed into funeral pyres show distin ct patterns of fire crackin g . Ground stone tools found near burials were often associated with food processing activities (Adams, 2014 a : 27), illustrating their possible significance to the user . As noted above, w omen have played a vital role in assigning meaning to rock shelter s in many cultures . This is illustrated by the fact that in the American Midsouth, female burials are believed to outnumber male burials in sheltered spaces (Carmody and Hollenbach, 2013 ; H omsey Messer, 2010). Burials are used to form ancestral ties to the past (Johnson and Earle, 2000) . Carmody and Hollenbach (2013: 50) note that burials may have also been used by the local inhabitants to Hollenbach (2009: 215) suggests that bedrock mortars may have belonged to a specific family, serving a similar purpose. At Trinchera Cave in Trinidad, Colorado, the remains of a young

PAGE 56

43 adult female were uncover ed (Simpson, 1976: 198). T here are currently no other burials known to exist in the Trinchera Cave Archaeological District (Black, 2000). H uman dentition was also discovered at the site and they are known to have be en deciduous teeth (Nowak and Gerhart, 2002: 262) . It is unknown at this time whether they belong ed to male or female children. Use Wear Analysis Use wear analysis is an archaeological method used to determine tool function by examining their work surfaces. Close examination may allow researchers to understand how tools are altered during use (Adams, 1988, 1989a, 1989b, 1993a , 2014b; Adams et al. 2009). When researching ground stone technology, four different mechanisms typically are used to describe and recognize how specific damage patterns are formed abrasive wear, adhesive wear, surface fatigue, and tribochemical wear (Adams, 2014a , 2014b ). Each of the mechanisms interact with one another when two surfaces come into contact. Use wear on certain items should be compared against state (Adams, 2014a , 2014b ). Surface wear results from the motion of one stone item coming into direct contact with another (Adams, 1988: 310, 1993a: 63, 2014 a : 28 , 2014b: 130 ; Czichos and Dowson, 1978: 98; Szeri, 1980: 35; Teer and Arnell, 1975: 94). Wear Wear is the gradual loss of surface material of a stone item as a consequence of the motion between it and a secondary contact surface (Adams, 1988: 310, 1993 a : 63 , 2014 a : 28 , 2014b: 130 ; Czichos and Dowson, 1978: 98; Szeri, 1980: 35 ; Teer and Arnell, 1975: 94). The amount of wear an item sustains is discernible using qualitative variables . Items that are unused may have sustained surface damage as a part of the design and manufacturing process, but there is no discernible damage from actual use. Light wear leaves behind very little evidence that can

PAGE 57

44 be seen with the naked eye (Adams, 2014 a : 28 , 2014b: 136 ). Conversely, moderate wear leaves noticeable damage, it does not actually change the basic tool shape. (Adams, 2014a: 28 , 2014b: 136 ). Finally, heavy wear alters the natural shape of the tool itself (Adams, 2014a: 28 , 2014b: 136 ). Wear management strategies are recognized for both manos and metates. As a mano is being used, it becomes less efficient and harder to handle. Users developed certain wear management strategies to help them cope with the wear produced inefficiencies (Adams, 2014a: 28 30 , 2014b: 13 6 ). If a rougher texture is needed, pecking the individual grinding surface can help accomplish that. Another method involves turning the mano over and using the opposite side as a secondary grinding surface. This way the user does not have to stop quit e as often with two usable surfaces. A third strategy is to lift the distal edge of the mano, while pushing down with the proximal edge so that only part of the surface is ever in contact with the metate (Adams, 2014a: 29). On the return stroke, the use r tilts the distal edge of the mano forward while lifting the proximal edge so a secondary surface is being used during the grinding process (Figure 4.1). The intensity of tool usage can be difficult to interpret in the absence of a time frame (Adams, 201 4a). Intensity can be assessed in terms of intensive and extensive uses. An intensively used tool is one that it is utilized for a longer period of time. For example, four hours a day for 30 days, resulting in 120 hours of wear. An extensively used too l is one that is utilized for a shorter period of time at any one sitting; one hour a day for 120 days, also resulting in 120 hours of wear. From an analytical standpoint, the amount of wear looks to be the same moderate. One way to distinguish between intensive and extensive use in the archaeological record is through an evaluation of tool design (Adams, 2014a). If comfort features are present on a moderatel y worn tool, such as grooves or handles, it was designed to be held comfortably for

PAGE 58

45 longer periods of time (Adams, 2014a). A moderately worn tool that does not have any comfort features may not have been designed for extensive use (Adams, 2014a). Wear Rates and Patterns Mona Wright (1993) made an effort to quantify the experimental wear rates of grinding tools in order to develop a model for prehistoric wear rates. This model allows researchers to determine the life history and abandonment processes of grinding tools (Ad ams, 2014 a : 69 70). Using replicated manos and metates, Wright ( 1 993) concluded that certain raw material types do wear at quantifiable rates and that manos wear down much faster than metates . She also discovered that differences in individual grinding skill is an important factor to consider in wear rates. However, when Wright (1993: 353) c ompar ed her results to prehistoric wear rates she found that it was n ot possible Figure 4.1. Grinding strokes and the subsequent wearing of two adjacent surfaces against a metate surface. Figure from Adams (1993b: 335). Redrawn by Ron Beckwith, based on Figure 8 from Barlett (1933).

PAGE 59

46 initial weight and thickness of prehistoric manos and metates before they were used to grind maize Jenny Adams (2014 a , 2014b ) looked at time and use wear with a large number of experiments. She examined the use wear patterns of different contact surfaces stone against stone contact, stone against wood contact, stone against bone contact, and stone against hide contact. Using replicated stone tools as a control, she questioned whether it was possible to di stinguish the damage patterns ca used by different intermediate substances such as amaranth seeds, clay, maize kernels, and sunflower seeds. She determined that the amount of wear an item sustains is a contributing factor. On lightly used surfaces, it was not possible to differentiate b etween the different substances. However, it is possible to recognize distinctive use wear patterns on moderately and heavily used tools (Adams, 2014 a : 36 44). Resources Raw material availability direct ly affects ground stone tool design, manufacture, use, and abandonment patterns (Stone, 1994). Early ground stone tool designs were based on notions of optimization and energy efficiency (Hard, 1986, 1990; Lancaster, 1983; Mauldin, 1991, 1993). Stone (1994) argues that modern stud ies using archaeological assemblages to reconstruct mobility and subsistence practices should take raw material availability into account. Natural topography may affect the procurement of certain raw materials and ground stone size (Stone, 1994 ). Differences in mano and metate size may be a direct result of new or greater access to larger raw materials in the region. Lithic studies have shown that abundance, characteristics, and spatial distribution of raw materials can affect the design and p roduction of stone tool technology (Andrefsky, 1994, 2005; Bamforth, 1986; Barton, 1988, 1990; Gould, 1980; Odell, 1977, 1989). Gould (1980) studied how the abundance of raw materials had an impact on stone tool production for the Aborigines of Australia. He discovered that when certain raw materials were

PAGE 60

47 available near the habitation base camp , the Aborigines used those raw materials in the production of different tool types , including expediently and strategically designed tools (Andrefsky, 2005: 235). Gould (1980: 134) noted that form of quarries or nonlocalized in nature, at or in close p roximity to a water source where a habitation base camp will occur, ease of procurement will outweigh other factors and unusually high percentages of artifacts of these locally available stones will Other researchers have studied raw material abundance in relation to transportation costs (Kuhn, 1991) and land use and mobility patterns (Daniel, 2001; Dobosi, 1991; Goodyear, 1993; Meltzer, 198 5 ; Seeman, 1994; Wiant and Hassen, 1985). Raw material choice is also affected by the processing constraints of different food resources (Greenwald, 1990; Horsfall, 1987). The material being ground should be hard and durable enough to effectively process food resources without the material wear ing down too fast. The grinder should also make sure that sand or gravel is not being added to the food being processed . Furthermore, variations in grain , nut, or seed size and durability place processing constraints on portable and bedrock ground stone technology (Hard, 1986; Lanca ster, 1983). Translating Portable Ground Studies to Bedrock Ground Stone Differences in the design and manufacturing process are separate technological traditions that could reflect distinct methods of constructing and operating food processing equipment (Adams, 2014 b : 131). Technological knowledge is transferred by the migration of individuals and/or small groups who then continue d their learned traditions in relocated areas (Adam s, 2010). Increased frequencies of a specific metate design in a region previously dominated by a different type of design may represent the migration of a larger social unit. It may be possible to recognize the natural development of permanent grinding stations. The Ancestral Puebloans and

PAGE 61

48 Mogollon , for example, developed two different metate designs based on the concept of permanent grinding stations to better confine the product during processing (Adams, 2014 b : 131). Both traditions managed to incre ase the number of grinders that could work together. Ground stone technology varies across time and space not only in the American Southwest, but in s outheastern Colorado as well. Trinchera Cave represents a combination of Plains and Southwestern culture s (Simpson, 1976), and may have been a melting pot of peoples (Nowak and Gerhart, 2002: 104). Technical knowledge may have been transferred during migration into the area.

PAGE 62

49 CHAPTER V METHODOLOGY Introductio n Missing from contemporary research for the region is an understanding of how portable and bedrock ground stone are connected to regional hunter gatherer adaptation and if these relationships changed through time . Bedrock ground stone represent an important aspect of human behavio r, landscape adaptation, mobility and subsistence strategies, and symbolic ideology (Lynch , 2017b). Data Limitations Inter observer Variability Archaeologists that document bedrock grinding surfaces in the field typically record the length (longest axis), width (second longest axis), and basin depth as standard units of measurement (Lynch et al., 2012). The amount of training and field experience recording bedrock grinding surfaces varies greatly among archaeologists. Irrespective of train ing, individuals will have somewhat different interpretations of grinding surface length, width, and depth. With this level of inter observer variability, any errors in measurement have the potential to be problematic for comparative studies (Lynch et al. , 2012). A certain level of caution is suggested when measuring bedrock grinding surfaces as well as using measurement data in the archaeological record. Other Limitations Time and budget constraints prohibited the ability to analyze every piece of porta ble ground stone housed at the Louden Henritze Archaeology Museum. Variations in the types of measuring devices and instrument accuracy in the field can add to differences in observer

PAGE 63

50 measurements. Differences in length, width, and depth can result from instrument variation (Lynch et al., 2012). A total of 211 items were selected to examine the nature of the ground stone activities that occurred at Trinchera Cave. Sample Selection A sample of (n = 179 ) portable ground stone recovered from Trinchera Cave in s outheastern Colorado is used in this study. This sample is the accumulation of multiple archaeological digs from the Trinchera Cave site and includes (n = 1 50 ) manos and (n = 29 ) metates . The metates are represented by (n = 22) slab metates, (n = 6) basin metates, and (n = 1) trough metate. This sample includes portable ground stone housed at the Louden Henritze Archaeology Museum at TSJC in Trinidad, Colorado. Each of the portable materials from this data set were measured using plastic Ohaus calipers and a Kobalt tape measure. Identification of the ground stone materials was done by Herbert Dick, Caryl Wood Simpson, and Michael Nowak . Most of the materials recovered from Haldon C were cataloged into the Colorado College database by Kylie Crocket, with help from Michael Nowak, Heather Gerhart, and Loretta Martin (Nowak and Gerhart, 2002: 99). Colorado College maintains a 75 80% digital inven tory of the materials re covered from the site. All of the materials, field notes, and analysis sheets are currently housed at TSJC. The portable manos and metates recovered by Herbert Dick and Caryl Wood Simpson were recorded from May 13 th through May 18 th , 2018 using artifact analysis sheets. No materials from Haldon Chase and the High Plains Columbia Expedition are curated at TSJC. Most of the materials from their joint 1949 expedition were reportedly shipped to Columbia University, and were later transferred to the University of Denver in 1955 (Zier , 2017: 26). However, after contacting the anthropology departments at both Columbia University and the University of

PAGE 64

51 Denver, it was learned that any materials collected from Trinchera Cave are not currently being housed at either location. Martin (personal communication, 2018a) also has no idea of their whereabouts. Lintz (1999: 20) believes that some of the larger ground stone items may have initially been left behind excavation . This seems reasonable due to the weight of some of the metates currently being housed at the Louden Henritze Archaeology Museum. (1999 2001) were recorded while at the museum. Nowak and his team excavated trenches and test units in all areas of the shelter, except for one (Zier , 2017: 29; Martin , personal communication , 2018 a ). Between 1954 and 1957, Herbert Dick excavated all the way down to bedrock at Area B (Zier , 2017: 29) , so r e excavating there was not necessary. Nowak and Gerhart (2002: 8) approached Trinchera Cave primarily a s a salvage project. A ll of the backfill (Martin , personal communication , 2018 a ). The extent of bioturbation is unclear. Nowak (2002: 8) states that Fill uniformity and the paucity of good stratigraphic markers contribute to the problem; the lack of earlier Due to the lack of provenience information, this collection could not inform us on the research question regarding changes through time. Even without the provenience information, his materials were examined for qualitative purposes. There a re several boxes of mano and metate fragments present at the museum . The different raw material types a re consistent with the portable ground stone that was provenienced from the earlier excavations: Dakota sandstone, gr anite, conglomerate, and vesicular basalt . Additional ground stone materials were found to be

PAGE 65

52 manufactured from argillite, mudstone, and siltstone suggesting they may have been used for activities other than food processing . All of these materials are lo cal. In other words, the raw material used for the unprovenienced g round stone is similar to that found in the other collections. Since these materials have no context within the shelter, they are excluded from the analysis. A Temporary Access Permit (No. 111861) for Trinchera Cave was approved by the Colorado State Land Board on May 31 st , 2018. The authorized dates for temporary access were June 22 nd June 24 th , 2018 or June 29 th July 1 st , 2018. The site is located in section 16, range 59 west, and township 33 south in the sixth prime meridian (Black, 2000 : v ) . Section 16 is currently being leased from the State of Colorado by the Baca Revocable Trust (Baca , personal communication , 2018). The owner is Philip (Phil) Baca and he primarily resides in Tucson, Arizona. The Baca property resides on roughly 17,000 acres of deeded land. After speaking with Baca about gaining access to the site , he approved both proposed dates. He asked that any future communicat ion go through his foreman, Buford Garcia, when time came to visit the site . T he first set of dates were chosen . The second data set represents the (n = 32) bedrock grinding surfaces present at Trinchera Cave. They were recorded between June 22 nd and June 24 th , 2018. Surface dimensions (length, width, and depth) were measured once by each analyst to try and accommodate for inter observer error (Lynch et al., 2012). Laura Baker assisted in the field by collecting the secondary set of measurements of the bedrock gr ind ing surfaces . Aluminum Professional Square from Empire were used while at Trinchera Cave. A rtifact form, raw material, and use w ear w ere also recorded . The individual variables are detailed below.

PAGE 66

53 IBM SPSS Statisti cal Software (V. 25) is used to record and manage the two data sets. The software is available for use at the University of Colorado Denver. Pollen Analysis While recording the portable ground stone at the Louden Henritze Archaeology Museum, it was decided that a pollen , starch , and XRF analysis should be conducted on one of the artifacts. A pollen analysis can help provide archaeologists with information on the plants and other resources processed on ground stone . Pollen and starches have been fou nd to accumulate on ground stone surfaces through repeated use (Cummings and Varney, 2009) . With analysis, we may be able to determine the intermediate substances that were being processed by gr inding tools. Inadvertent pollen deposits like pollen rain o r wind blown pollen may have occurred in the region. Data from pollen samples are frequently examined in order to distinguish food processing activities against a backdrop of naturally occurrin g pollen rain ( Brush, 2001; Bryant and Holloway , 1983; Pearsall, 2000) . Archaeologists often strive to understand how environmental change affected hunter gatherer behavior, and what socioeconomic factors influenced how people adapted to these new conditions. Pollen profiles are especially useful as they rec ord the ways people may have shaped their landscape. Pollen related to food procurement and processing is generally transported by people, wh ile airborne and/ or water transported pollen is frequently non food in nature (Mercuri et al., 2010). Windblown p ollen is often found at archaeological sites and we use it for environmental reconstruction. Researchers want to know changes in forest make up to understand long term environmental change. Windblown pollen may not affect what we think is going on with f ood, but it is definitely important in understanding landscape. Pollen evidence of food at archaeological sites is predominantly from local instead of regional sources (Davis, 1994;

PAGE 67

54 Pearsall, 2000). Reconstructing prehistoric diet often relies on the belief that pollen mainly comes from the resource patch visited and exploited for subsistence purposes. Mercuri et al. (2006) refers to this patch as the area of influence for each archaeological site. Breeding, cultivation, exploitation, and settlement patterns all affect the environment resulting in changes to pollen diagrams (Faegri et al., 1989). Variables that are associated with different activities, such as trade or cultivation, become part of the pollen spectra deposited directly on tools because they more accurately reflect human behavior than the environment (Kelso and Good , 1995; Li et al. , 2008; Mercuri , 2008). Plant taxa are iden tified by quantity and condition of mineral elements by provenience (Cummings and Varney, 2018) . The condition of mineral elements is important to matters of possible contamination. R odents and insects are both known to harvest and store seeds (Brown and Davidson, 1977; Chaney, 1936), r esulting in the post occupational deposit of seeds into different archaeological settings . The presence of rodent pellets and insect remains in soil samples can provide an indication of contamination potential. Carbonization, or charring, is a means of p reserving prehistoric animal remains and plant material (Gleichman, 2002). Plant material that is uncharred may or may not be prehistoric. Prehistorically introduced seeds preserved through burning are not automatically indicative of culturally used plants. They may be naturally dispersed seeds that were inadvertently burned. There are many ways seeds can be burned and integrated into the archaeological record. Burned seeds representing eight diff erent taxa were identified from soil samples at Trinchera Cave (Gleichman, 2002: 95): Celtis reticulata (hackberry), Chenopodium sp. (goosefoot), Helianthus sp. (sunflower), Juniperus sp. (juniper), Portulaca sp. (purslane), Oryzopsis hymenoides (rice gr ass), and Scirpus sp. (bulrush). Burned seeds from the grass

PAGE 68

55 A B C D family (Poaceae) were also identified. All of these plants are considered edible. When compared with the material culture recovered from Trinchera Cave, archaeobotanical remains, phytoliths, pollen, and starches can become potential use indicators (Cummings and Varney, 2018: 2). After careful consideration, a small , one handed sandstone mano (10.3 cm in length x 6.7 cm in width x 4.8 cm in thickness) (Figure 5.1) . This specific artifact was selected for its unique triangular shape, the presence of Figure 5.1. Different views of mano used for analysis.

PAGE 69

56 blue pigmen t, and several small black hairs embedded on one of the grinding surfaces. Tool morphology cannot be used to determine what was ground (Adams, 1999; Horsfall, 1987; Wright, 1993). In order to understand the functional differences between portable ground stone and bedrock ground stone, we need to establish what the inhabitants of Trinchera Cave were grinding. This is best achieved through archaeobotanical analysis. The specimen bag that the mano came in initially had three different artifacts in it. Eac h of the three items were bagged separately and placed into the larger specimen bag. All three were previously given the catalog number Tr. 1 932. After communicating with th e museum d irector, Loretta Martin, we separated the artifacts into three disti nct bags. The artifacts were labeled as Tr. 1 932a, Tr. 1 932b, and Tr. 1 932c. The catalog number for the selected mano became Tr. 1 932a. Artifact Analysis Prior to taking the artifact to PaleoResearch Institute for an archaeobotanical analysi s, it was photographed and examined under a microscope. Initial photographs of the mano were taken using a Cannon PowerShot SX530 HS. It was later examined in the archaeology lab at Metropolitan State University using a Leica EZ4 W Microscope. The low s taging area allowed for the mano to be fully placed under the objective lenses. Pictures of the artifact were taken using the microscope at 15X 20X magnification, providing high resolution images (Figure 5.2). After microscopic examination was complete , the mano was rebagged and transported to the nearby lab. Archaeobotanic al Analysis Using funds awarded from the 2018 Karen S. Greiner Endowment, a pollen , starch, and XRF analysis was conducted on the sandstone mano by PaleoResearch Institute, Inc . (PRI) in Golden, Colorado. Palynologist and paleoecologist, R.A. (Robert) Varney, processed the mano using standard lab protocols. Specifically, all visible dirt was first removed with regular tap

PAGE 70

57 A B C D water and gentle hand pressure to eliminate any modern contaminants. A small piece of the gr inding surface wa s tested using a 10% dilut ion of hydrochloric acid (HC1) to help detect the presence of any calcium carbonates (Cummings and Varney, 2018) . These carbonates we re eliminated using an additional solution of HC1 . T he gr inding surface wa s then washed with a n 0.5% solution of Triton X 100 (Johnson, 2018) to recover any pollen and starch grains that were Figure 5. 2 . Photos of mano taken at 15X 20X magnification.

PAGE 71

58 present . The grinding surface was carefully scrubbed with a n ultrasonic toothbrush and then rinsed using reverse osmosis deionized (RODI) water (Varney, personal communication, 2018) . Each sample wa s filtered through 250 micron mesh to remove any large particles that m ay have come off during the washing process (Cummings and Varney, 2018) . This pollen rich organic sample was rinsed and then received a 25 minute bath in hot hydrofluoric acid (HF) to eliminate the remaining inorganic particles (Cummings and Varney, 2018) . The sample was acetylated for 10 minutes to eliminate any unnecessary organic matter (Varney, personal communication, 2018) . The sample was then washed again using RODI to achieve a neutral pH level. Next, several drops of potassium hydroxide (KOH) were added to the sample which was then stained with safranin (Cummings and Varney, 2018) , coloring all the cell nuclei red. The sample was then centrifuged at high speeds , causing the minute organic debris to travel towards the bottom of the tube. A light microscope at 500X magnification was used to count the pollen. The preservation of pollen in the sample ranged from poor to good (Cummings and Varney, 2018) . Comparative reference material collected at the University of Colorado Herbarium and the Intermountain Herbarium at Utah State University was used to help identify pollen types to the family, genus, an d species level (Cummings and Varney, 2018). Microscopic pieces of charcoal were recorded during a portion of the pollen count. The estimated abundance of charcoal was determined through computer extrapolation and is presented on the pollen diagram (Figu re 5.3). Indeterminate pollen refers to pollen grains that are folded, mutilated, or are otherwise unrecognizable (Cummings and Varney, 2018). Starch granules can be retrieved from sediments, but are seldom preserved (Kooyman, 2015). While susceptible to enzymatic attack, they can survive for an extended period of time in

PAGE 72

59 a dry, stable environment. For example, they can survive in dental calculus and in the micro cracks of ground stone tools (Hardy, 2007). Starches can also come from residue analysis, including carbonized residues (Kooyman, 2015). Pollen extraction from ground tools often retains those granules. Starch es were recorded during the pollen count on the mano. Starch granules are an important energy reserve in plants. Starches are found in various seeds, such as corn, rice, and wheat. It is also found in starchy roots and tubers. The main categories of st arches include the following: with or without a visible hilum, a centric or eccentric hilum, hila Figure 5. 3 . Pollen diagram for a mano, 5LA1057, Las Animas County, Colorado. Figure from Cummings and Varney (2018: 8).

PAGE 73

60 patterns (cracked, dot, or elongated), and starch shape (angular, circular, ellipse, or lenticular) (Cummings and Varney, 2018). Some of these starch catego ries are relatively common and occur in many different plant species, while others are unique to specific plants. ED XRF (Energy Dispersive X Ray Fluorescence) Analysis A small quantity of blue pigment observed on the surface of the m ano was the focus of the X ray f luorescence (XRF) analysis . XRF is a non destructive analytical technique used to assess the elemental composition of a wide range of different geological materials. A Bruker Tracer 5i (900F4197) was used with an Rh anode and XFlash ® silicon d rift detector (SDD) with (Cummings and Varney, 2018). Assays for trace elements and metals were collected at 50 kV with a current of 10.3 and a green filter compos ed of , , and Al to help eliminate any secondary emission of bremsstrahlung radiation below 13 keV (Cummings and Varney, 2018). For elemental analysis, an empirical calibration reference set was used composed of 41 clay standards that were collected and analyzed by the University of Texas at Arlington (UTA) empirical standards is essential for data reliability and interlaboratory compariso ns (Speakman and Shackley, 2013), especially when accuracy is at the parts per million (ppm) level for most elements (Cummings and Varney, 2018). ARTAX and S1PXRF software were used to help determine the elemental composition of the provided samples. Alth ough present in the resulting spectrum, rhodium (Rh) and palladium (Pd) are not listed in the discussion below because of their association with the instrument (Cummings and Varney, 2018). Geological samples, like this mano, are often

PAGE 74

61 difficult to evaluat e because of their heterogeneous composition (Cummings and Varney, 2018). While these samples appear consistent, they still represent a combination of different materials that were not homogenized before testing (Cummings and Varney, 2018). ED XRF analys is of the data from separate points in a heterogeneous sample represents the structure of that area. T he area displaying the blue pigment was smaller than the size of the Tracer, indicating that the mano itself is also reflected in the spectrum collected (Cummings and Varney, 2018). Results Archaeobotanical analysis on the mano revealed the presence of several composites, members of the amaranth family, and grasses (all of which are potential food sources). Specifically, analysis revealed the presence of Amaranthaceae, Artemisia , High spine Asteraceae, Liguliflorae, Melilotus , and Poaceae (Figure 5. 3, Table 5.1 ), representing plants from the goosefoot family, sagebrush, plants from the sunflower family, members of the chicory tribe of the sunflower family, sweet clover, and grasses (Cummings and Varney, 2018). Charcoal and other indeterminate starches were also found. Amaranthaceae, High spine Asteraceae, and Poaceae were the most plentiful, representing plants from the goosefoot family, sunflower family, and grasses. All of these families include plants with edible portions (Cummings and Varney, 2018). Seeds from plants in each of these families are believed to have been ground. The recovery of a large, lenticular starch exhibiting heavy edge damage, suggesting cooking, indicates that se eds from a large seeded grass species, such as barley grass (Hordeum pusilim), native wheatgrass (Agropyron), or ryegrass (Elymus), was ground (Cummings and Varney, 2018). All of these cool season grasses have seeds that ripen during the late spring to ea rly summer. Pollen recovered from the samples offers a look at the plants available to the

PAGE 75

62 Table 5. 1 . Pollen Types Recovered in Samples from Site 5LA1057, Las Animas County, Colorado. Table from Cummings and Varney (2018: 7). been ground using this mano. The XRF analysis generated numerous peaks (Figure 5.4); including, in order of diminishing amplitude silicon, calcium, aluminum, iron, magnesium, potassium, sulfur, phosphorus, and titanium (Cummings and Varney, 2018). All the other peaks are considered minor in comparison. None of the elements stand out to explain the small amount of blue pigment on the surface of the mano (Cummings and Varney, 2018). Data Collection Techniques on Portable Ground Stone Portable ground stone artifacts from Trinchera Ca ve are used in this study. The sample represents multiple excavations that have occurred since 1949 when the Louden brothers first

PAGE 76

63 discovered the site. Manos and metates housed at the Louden Henritze Archaeology Museum in Trinidad, Colorado were examined between May 13 th May 18 th , 2018 . All of the portable materials included in this data set were measured individually using a Kobalt tape measure and plastic Ohaus calipers. Measurements were then compared against catalogs, field notes, and artifact inventories present at the museum. If a discrepancy arose, the mean was taken between the two measurements and then recorded on th e artifact analysis sheet. A rtifact analysis sheets were created prior to conducting researc h . They were devised to help collect necessary data from both the portable ground stone and the bedrock ground stone. The artifact analysis sheets asked the following information site name, catalog number, year of excavation, provenience, artifact type, material type, shape, condition (whole, < half, > half), location of wear ( long surface, both ends, both long surface and ends, or none) , pecked surface (yes or no), polished (yes or no), portable (yes or no), use wear (light, moderate, heavy, unused), Figure 5.4 . Spectral image for area of mano exhibiting blue stain. Figure from Cummings and Varney (2018: 9).

PAGE 77

64 len gth and width, and wear patterns ( circular, reciprocal, pounding, unknown, none) (Figure 5.5) . Each of these variables is detailed below. Site Name : Site name for all the ground stone materials refers to Trinchera Cave, site 5LA1057. Catalog Number: The catalog number indicates the cataloging system used by the different excavation teams: Chase (TA #), Dick (TR.1 #), Simpson (TR1 D #) , or Nowak (1057.O. #) . Year of Excavation: The year of excavation represents what year the artifact was co llected: Chase (1949 1950), Dick (1954 1957), Simpson (1974), or Nowak (1999 2001). Provenience: Multiple excavations have taken place at the shelter over the year s, with each u sing a different cataloging system. # printed on them and are all listed in same database. letter provenience associated with them (Nowak and Gerhart, 2002: 105). The number and letter refer to his grid system. Unfortunately, details of how the grid system wa s set up are not present in the archived field notes. Since artifact depth was recorded in inches, it is presumed D # printed on them. All of her measurements were taken in cm measured in cm and have a letter provenience associated with them (Nowak and Gerhart, 2002: 105). Artifact T ype: Artifact type refers specifically to the type of ground stone that was examined: slab metate, basin metate, trough metate, undetermined metate fragment, bedrock metate, or erstones, manos

PAGE 78

65 and metates, mortars and pestles, and polishers (Adams, 2014a), only manos and metates were examined during the course of this study. Material T ype: Material type was determined through visual inspection and by going through catalogs, field notes, and artifact inventories. The different material types represented in this collect ion are Dakota sandstone, granite, conglomerate, and vesicular basalt . Shape: Shape (circular, oval, square, rectangular, triangular, and undefined) of the ground stone artifact s is particularly subjective . The shape of the grinding surfaces has the potential to inform us of a variety of behavioral processes used to process material (circular, reciprocal, or pounding) (Lynch, 2017a). Condition : This variable records whether the artifact is whole or a fragment. A fragment refers to a broken part of an artifact. The artifact analysis sheet included fragment type: whole, le ss than half, or more than half. Location of W ear : The location of wear (long surface, both ends, both long surface and ends , or none) refers to where use occurred on the grinding tool . Figure 5. 5 . Blank artifact analysis sheet.

PAGE 79

66 Pecking: Pecked surface (yes or no) indicates whether or not the grinding surface of the artifact shows visible evidence of pecking. Pecking was determined by the presence of multiple peck marks on the grinding surface and by individual marks that continue past the grinding area . Battering is not included in this analysis. Battering results from use as in a pestle and is seen in a crushing of the grains on the grinding surface as opposed to pecking which is used to refurbish (as on a mano or metate surface) and is evident as individual marks that extend beyond the surface into the body of the specimen . T aphonomic agents can interfere with the analyses of ground stone artifacts , including carnivores, scavengers, and humans (Blumenschine , 1988). Other taphonomic processes have the potential to affect ground stone artifacts, such as fluvial action, gravity, and weathering (Lyman , 1994; Rick , 1976; Stodder , 2008) . Rock shelter s are often subjected to a wide range of different taphonomic processes . Each portable artifact was examined for any identifiable taphonomic signatures. Polish: Polished (yes or no) refers to whether or not the grinding surface shows visible evidence of sheen. Portable: Next on the artifact analysis sheet is portable (yes or no). Chase, Dick, Simpson, and , meaning they can be moved from one spot to another . Use W ear: Use wear (light, moderate, heavy, unused) is the next attribute and was determined by using catalogs, field notes, and artifact inventories whe n possible. Each piece of ground stone was carefully inspected. Light wear was selected when very little evidence of use wear was visible with the naked eye. Moderate wear was selected when evidence of use wear left behind no ticeable damage and did not alter the basic shape of the tool. Selecting heavy wear indicated

PAGE 80

67 that the use wear was so intense that it changed the natural shape of the tool. Finally, unused was selected when there was no visible use wear on the tool. Len gth and Width: For the manos, length and width were determined by measuring the entire artifact. For t he metates however , length and width were determined by measuring the portion of the artifact that contains wear, not the entire artifact itself. On the metates, width is represented by the second longest axis. Wear Pattern: Use wear analysis can reveal the presence of abrasion, impact fractures, and/or sheen (Adams, 2014 a ). Behavioral inferences are often possible through use wear analysis and ethnographic descriptions. Wear patterns (circular, reciprocal, pounding, unknown, none) indicate how an intermediate substance was being ground . It is often possible to identify distinctive use wear patterns on moderately and heavily used tools (Adams, 2014 a : 36 44). All of the artifacts were carefully inspected for wear patterns. A circular stroke cause s visible striations to appear in various directions across the surface of the stone. A reciprocal stroke entails moving the tool back and forth ac ross the . The pounding stroke uses the The crushing of grains on the individual surface is more closely associated to battering (Adams, 2014a: 132) . Pounding is more of a high impact action resulting in impact fractures that are much different than those caused by crushing (Adams, 2014a: 32). Depending on the intermediate substance being ground, the two adjacent surfaces may not come into contact wit h each other reducing impact damage to both tools (Adams, 2014a: 132). Unknown was selected when a wear pattern was present, but could not be determined. If no wear pattern was found, then none was selected and was coded as absent in SPSS.

PAGE 81

68 Research of the portable materials took place between May 13 th and May 18 th , 2018 at the Louden Henritze Archaeology Museum in Trinidad, Colorado . This marks the first time any ground stone materials have been stud ied from Trinchera Cave (Martin, personal communicati on, 2018a). Comparable information was collected on the BGS surfaces as was collected on the portable ground stone. Depth was added as an additional variable. Data Collection Techniques on Bedrock Ground Stone The bedrock grinding surfaces present at Tr inchera Cave were recorded between June 22 nd and June 24 th , 2018 (Figure 5.6). Surface dimensions (length, width, and depth) were measured twice to accommodate for inter observer error (Lynch et al., 2012). All measurements were taken in cm and then later converted into mm. The first set of measurements were collected secondary set of measurements were then collected by Laura Baker to insure inter observer consistency in the data. The mean of each measurement was calculat ed and recorded on an artifact analysis sheet. IBM SPSS Statistical Software (V. 25) is used to record and manage this data set. As indicated earlier, artifact analysis sheets were made prior to conducting research (Figure 5.4). Data on portable ground stone and bedrock ground stone (BGS) were collected using the same paperwork. Comparable information was collected on the portable ground stone as was collected on the BGS. Site Name: The site name used for all the bedrock grinding surfaces Trinchera Ca ve, site 5LA1057. Catalog Number: The catalog number indicates the cataloging system used. The BGS found at the shelter have not been previously cataloged. A cataloging system was created for them for

PAGE 82

69 A B C D Year of Data Recording: For the portable ground stone, the year of excavation represents what year the artifact was collect ed. However, for the BGS the year of excavation refers to the year the surfaces were analyzed. In this case, the year is 2018. Provenience: Provenience information indicates the depth the bedrock features were analyzed at (surface, 0 examined are at the surface level. Figure 5. 6 . Photos of bedrock grinding surfaces taken at Trinchera Cave.

PAGE 83

70 Artifact Type: Artifact type refers to the type of ground stone that was exami ned , bedrock metates. Zier (2015: 19) established a primary datum point opposite the cave entrance during his investigation of the site. A datum point is a point of reference from known coordinates and is often established using an electronic total stat ion. When the site was visited between June 22 nd and June 24 th , 2018 , the datum point set up by Zier was no longer present. Zier (2015: 19) also established a permanent datum point w hen he was mapping the site. A large nail was embedded However, the permanent site datum could not be located. It may have been pulled or become buried over time. Zi provenience (Figure 5. 7 ). Material: The different material types represented by the portable artifacts housed at the Louden Henritze Archaeology Museum include Dakota sandstone, granite, conglomerate, and vesicular basalt. Permanent grinding stations are often carved into unmovable rocks and includes those found in caves, shelters, rocky outcrops, and structure walls (Woodbury, 1954: 117). All of the grinding surfaces found at Trincher a Cave are carved into large boulders of Dakota sandstone. Shape: Shape is the next attribute on the artifact analysis sheet. Determining shape (circular, oval, square, rectangular, triangular, or undefined) of the grinding surfaces is largely subjective . Each surface was individually inspected for an approximate shape. The shape of BGS surfaces may be able to inform us of a range of behavioral processes used to process intermediate substances (circular, reciprocal, or pounding) (Lynch, 2017a). Condition : Fragment (whole, < half, > half) refers primarily to the portable materials. All of the bedrock features are considered intact.

PAGE 84

71 Location of W ear: The location of wear (base only, up the sides) was then determined for each of the grinding surfaces. Wear was not determined for the bedrock feature, just the individual grinding surfaces. Examination was completed in bright sunlight and with the nake d eye. Pecking: A pecked surface (yes or no) indicates whether or not the BGS surfaces shows evidence of pecking. Pecking is particularly useful for resharpening the grinding surfaces of manos and metates (Adams, 2014a: 45). Battering is not included in this analysis. Pecking is used to refurbish (as on a mano or metate surface) and is represented by individual marks that continue on past the grinding surface into the body of the specimen. Polish: A polished surface (yes or no) shows visible evidence of sheen. Figure 5. 7 . Trinchera Cave base camp . Figure from Zier (2015: 11).

PAGE 85

72 Use W ear: Use wear (light, moderate, heavy, unused) was determined through careful inspection. Length, Width, and Depth: For the bedrock grinding surfaces, length and width were determined by measuring the portion of the surface that contains wear. Width is represented by the second longest axis. The depth of each grinding surface was measured in cm: 0.1 1.1 cm (light ), 1.2 2.1 cm (moderate), and 2.2 cm and greater (heavy). To determine the depth of the grinding surface s , measure was then used to determine depth in the center, the proxi mal end, and the distal end. The three depth measurements taken were then averaged and subsequently recorded. The second analyst then repeated the entire procedure. The mean of those two figures was then calculated for the final depth of each grinding s urface . Wear Patterns: Wear patterns (circular, reciprocal, pounding, unknown, or none) suggest how a substance was processed. All of the BGS were carefully inspected for wear patterns. It is unclear how the weathering of these surfaces may have affected the grinding surfaces. A circular stroke causes visible striations to appear in various directions across the surface of the stone. A reciprocal stroke entails moving the tool back and times, impact fractures are generated by the pounding. These fractures are usually deeper and more jagged than those caused by crushing (Adams, 2014a: 132). Depending on the material being, the opposing surfaces may not even come into contact with one another. This results in less impact damage to both tools (Adams, 2014a: 132). Unknown was selected when a wea r pattern was present, but could not be determined. If no wear pattern was found, then none was selected and was coded as absent in SPSS.

PAGE 86

73 CHAPTER V I RESULTS Introduction The portable and bedrock ground stone data sets presented in this study are representative of an area that has had substantial human activity. All information gathered comes from the Trinchera Cave (5LA1057) archaeological site in Las Animas County, Colorado. The first data set used in this study is represented by (n = 179) portable ground stone artifacts recovered from Trinchera Cave in Southeastern Colorado. The second d ata set represents the bedrock grinding surfaces present at Trinchera Cave (n = 32) . Metate Sample Description The different raw materials represented by the ground stone technology includes Dakota sandstone, granite, conglomerate, and vesicular basalt . A bar graph is used to determine the quantity of each raw m aterial type (Figure 6.1). Dakota sands tone represents the largest percentage in the assemblage (75.8%). This percentage is to be expected given the site is located in an area dominated by Dakota sandstone. The other raw material s are locally available. A second bar graph is used to help determine the quantity of each specific artifact type in the archaeological assemblage (Figure 6.2). The frequency of artifact types recovered at the site varies. This sample is the accumulation of various archaeological digs conducted at the site and in cludes manos and mano fragments (n = 150), slab metates (n = 22), basin metates (n = 6), and one trough metate. Manos were often cached for later use, while metates may have been removed once a site was abandoned or were later scavenged (Adams, 1998; Schl anger, 1991). Broken manos are often cached because they can be used later as hammer stones, supports, and as structural fill (Clark,

PAGE 87

74 Figure 6.1. Bar graph of raw material types. Figure 6.2 . Bar graph of artifact types.

PAGE 88

75 1988: 94; Hayden, 1987: 191; Searcy, 2011: 98). Manos and metates are both heavily valued by family members over many generations (Searcy, 2011). Nowak uncovered evidence of a midden directly under the discovered in this midden (Nowak and Gerhart, 2002). The high concentration of ground stone indicates that this area was likely a deposit zone for grinding tools after they had broken. Seed and plant processing likely took place further inside the shelter. Bedrock metates are the most prevalent metate type found at the site, followed by slab metates (Figure 6.2). If Trinchera Cave was used on a temporary, but regularly used basis as N owak and Gerhart (2002: 12) suggest, then people living in the region may have transported specific metates with them to the shelter . Once they left, they may have taken their metates with them . It is also possible that the bedrock ground stone may have been preferred over portable ground stone at Trinchera Cave. The portable artifacts are described t hrough three primary continuous variables: length, width, and thickness. The basic unit of measurement is metric with each of these variables being measu red in millimeters (mm). The bedrock grinding surfaces are also described t hrough three primary continuous variables: le ngth, width, and depth. Due to the size of the grinding surfaces and the limitations of the measuring tools, e ach of the variables were initially measured in centimeters (cm) and were later converted to mm for analysis. Several b ox plots are used to determine: 1) median length while controlling for raw material type (Figure 6.3). 2) median width while controlling for r aw material type (Figure 6.4). 3) median length while controlling for artifact type (Figure 6.5). 4) median width while controlling for artifact type (Figure 6.6).

PAGE 89

76 Figure 6.3 . Box plot illustrating length range for bedrock and portable grinding surfaces of each raw material type. Figure 6.4 . Box plot illustrating width range for bedrock and portable grinding surfaces of each raw material type. .

PAGE 90

77 Figure 6.5 . Box plot illustrating length range for each artifact type. Figure 6.6 . Box plot illustrating width range for each artifact type.

PAGE 91

78 Table 6 . 1 . T test results comparing the length and width of bedrock ground stone to portable ground stone. Variable Degrees of freedom t score p value Difference in means (mm) Length 52 3.655 0.001 89.5 Width 52 1.707 0.094 27.2 Comparison of Metates to Bedrock Grinding Surfaces The first research question asks if there is a functional difference between bedrock ground stone and portable ground stone . Several statistical tests were run to compare sub groups within the assemblage from Trinchera Cave. Specifically, independent T t ests were run to compare length and width to determine if there is a statistically significant difference in the mean and variation around the mean for these variables. For analysis of the base stones, slab metates are compared to the bedrock grinding s urfaces . In both cases, the analyses indicate the null hypothesis of no difference should be rejected. The bedrock grinding surfaces are, on average, 89.5 mm longer than metate grinding surfaces. In addition, the bedrock grinding surfaces are, on averag e, 27.2 mm wider than metate grinding surfaces . These results are illustrated in Table 6.1. Chi square tests were then ran on several nominal variables. Chi square tests are commonly used to determine if two nominal variables are related in any way. W hile the tests are used to determine whether or not an association is present, they cannot determine the strength of that relationship. If a relationship is present, then a used to determine the strength of that relationship. For this analysis, artifact type is compared against two different nominal variables (wear patterns and use wear) to see if an association exists. Wear patterns are

PAGE 92

79 described through five nominal type variables: reciprocal, rotary, pounding, unknown, or none. U se wear is described through four nominal type variables: light, moderate, heavy, or unused. The first chi square test was used to determine if an association exists between artifact type and wear patterns. After the initial test, SPSS software noted tha t 13 cells (65.0%) had expected values of less than 5. The minimum expected count is 0.01. Artifact types were then recoded into different variables . Slab metate s , basin metate s , trough metate s , and undetermined metate fragment s were combined together as portable base stones . Bedrock grinding surfaces were compared to these . For wear patterns, the variables unused and none were excluded from this analysis. A chi square test was then ru n a second time with the new coded var iables . This time, 0 cells (0.00 %) had an expected outcome of less than 5. Th e minimum expected count is 7.13 . The analysis indicates t here is not a relationship present between a rtifact type and wear pattern (df = 2, p value = 0.405). A chi square test was then used to determine if an association exists between artifact type and use wear. The artifacts were recoded the same as noted above. Use wear variables were left alone. A chi square test was then ran again with the newl y coded var iables. This time, 0 cells (0.00 %) had an expected outcome of less than 5. The minimum expected count is 7.13 . There is not a relationship present between artifact type and use wear (df = 2, p value = 0.405). Summary of Results IBM SPSS Statistical Sof tware (V. 25) is used to record and manage the data sets . Multiple box plots are used to determine median length while controlling for raw material type (Figure 6.3), median width while controlling for raw material type (Figure 6.4), median length while c ontrolling for artifact type (Figure 6.5), and median width while controlling for artifact type (Figure 6.6). Based on the results illustrated by the box plots, the greatest median length

PAGE 93

80 and width for raw material type is Dakota sandstone. The greatest median length and width for artifact type are bedrock me tates. This means that on average, the bedrock grinding surfaces are longer and wider than their portable counterparts. Two independent sample t tests were then ran to address the first research question on whether or not there is a functional difference between bedrock ground stone and portable ground stone. The first independent sample t test was ran to determine if there is a statistical difference in length between bedrock and portab le grinding surfaces and the second on width . Testing revealed a significant difference in mean length (p = 0.001) and width (p = 0 .094). Conversely, the chi square tests indicate that wear patterns and type of use wear do not differ between portable and non portable base stones (p = 0 .405 for both). Mano Sample Description The manos and mano fragments are described through three primary continuous variables: length, width, and thickness. The basic unit of measurement is metric with each of these varia bles being measured in millimeters (mm). H istograms are used to determine: 1) mean and standard deviation for length (Figure 6.7). 2) mean and standard deviation for width (Figure 6.8). 3) mean and standard deviation for thickness (Figure 6.9). A bar graph is used to determine the occurrence of each wear pattern type (Figure 6.10). Mano wear patterns are largely undetermined. A second bar graph is use d to help determine the occurrence of each use wear type (Figure 6.11). Mano use wear is predo minantly heavy. Comparisons of Manos to Grinding Surfaces Several statistical tests were run on the manos and mano fragments. Specifically, histograms were run to compare length, width, and thickness to determine if there is a normal

PAGE 94

81 Figure 6.7 . Hi stogram illustrating mano length. Figure 6.8 . Histogram illustrating mano width.

PAGE 95

82 Figure 6.9 . Histogram illustrating mano thickness. Figure 6.10 . Bar graph of wear pattern occurrence.

PAGE 96

83 or bimodal distribution. The portable and bedrock grinding surfaces have varying lengths and widths. Through analysis, the manos may be matched with one type or the other. A bimodal distribution exists for mano le ngth (Figure 6.7). Manos shorter than the mean length of 90.0 mm, may have been used for the portable grinding surfaces. Those longer than the mean may have been used for the bedrock grinding surfaces. Meanwhile, a normal distribution exists for mano wi dth and thickness (Figures 6.8 and 6.9). Summary of Results Histograms are used to determine the mean and standard deviation for mano length (Figure 6.7 ), mano width (Figure 6.8 ), and mano thickness (Figure 6.9). Based on the results illustrated by the histograms (Table 6.2) , there is a bimodal distribution present solely for mano length. A normal distribution is present for mano width and thickness. Two bar graphs are used Figure 6.11 . Bar graph of use wear occurrence.

PAGE 97

84 Table 6 . 2 . Histogram r esults compa ring mano length , width , and thickness to bedrock and portable grinding surfaces. Variable Mean (mm) Standard Deviation (mm) Distribution Type Length 90.0 29.20 Bimodal Width 71.2 19.96 Normal Thickness 42.1 10.73 Normal to determine the prevalence of specific wear patterns and use wear. Mano wear patterns are largely undetermined without the aid of a microscope. Future research may be able to shed some light in this area. Both the portable metates and bedrock grinding surfaces show strong evidence of reciprocal wear patterns. Mano use wear is predominantly heavy. Portable grinding surfaces show mostly moderate wear (44.8%), followed by heavy wear (31.0%), and light wear (24.2%). Bedrock grinding surfaces show mostly moderate wear (43.8%), followed by light wear (37.5%), and heavy wear (18.7%). The portable grinding surfaces may have been predominantly used at Trinchera Cave, with the bedrock grinding surfaces being used as a supplemental form of technology.

PAGE 98

85 CHAPTER VI I CONCLUSIONS Summary The purpose of my research is twofold: 1) to determine if there is a functional difference between bedrock ground stone and portable ground stone and 2) to determine when bedrock ground stone is preferred over portable grou nd stone within the mobility and subsistence system seen at Trinchera Cave. Several statistical tests were run to compare sub groups within the assemblage from Trinchera Cave. Specifically, independent T tests were run to compare metate length and width . Statistics revealed that bedrock grinding surfaces are, on average, 89.5 mm longer than portable grinding surfaces. Furt hermore , bedrock grinding surfaces are, on average, 27.2 mm wider than portable grinding surfaces. Ethnographic studies have demonstrated that efficiency is improved by increasing the number of ground stone tools used for reciprocal grinding and the size of the grinding surfaces (Hard, 1986, 1990; Horsfall, 1987; Mauldin, 1991, 1993). The portable and bedrock grinding surfaces from Trinchera Cave show strong evidence of reciprocal wear patterns, though mano wear patterns remain largely undetermined. A po sitive correlation exists between the length of the grinding surface, as illustrated by mano length, and the importance of agriculture to the subsistence economies of contemporary groups (Hard, 1986, 1990; Mauldin, 1991, 1993). By increasing the size of t he grinding surface, processing speed increases and it allows for more grain to be processed at a time (Mauldin, 1993). T he length and width of the bedrock grinding surfaces are larger on average than the portable grinding surfaces from Trinchera Cave. T hese two variables represent a functional difference between the two sub groups.

PAGE 99

86 Paleoenvironmental records are not extensive for southeastern Colorado, though some relevant research has been completed. Zier (1989), for example, discusses the environmental conditions over the last 4500 4000 years in southeastern Colorado. Reviewing local and regional data, Zier (1989 : 41 42 ) determines that the period 2000 1000 BP was fairly cool and moist with greater effective moisture than what we see today. Warmer temperatures and drier conditions followed around 1000 450 BP, while the Little Ice Age (LIA) between 450 150 BP saw a return to cooler temperatures and a wetter climate (Zier, 1989: 41 42). Within the last 150 years, the region has seen much warmer and drier conditions. Painter et al. (1999) provides a more comprehensive synthesis. Nowak and Gerhart (2002: 12) s uggest that maize and squash were introduced to the area sometime between AD 500 and AD 1000. This may indicate that people wanted to diversify the food resources in the area, allowing them to spend more time at the site. During this period, the region w ould have been relatively cold and wet. Bellorado (2011: 41) argues that in the San Juan Basin, it is very difficult for people to successfully produce mature maize and squash harvests during these colder periods . Most varieties of corn require a minimum growing season of between 110 120 days, and a minimum of 2200 2500 corn growing degree days (Black, 2000: 8) . C orn growing degree days (GDD) are determined by subtracting the rature (NDawn Center, 2019). GDD is between 2579 and 2800 for Las Animas County. The TCAD currently fits these temperature and agricultural requirements (Black, 2000: 8), but may not have when maize and squash were first introduced to the region . The n umber of frost free days also plays an important part during the growing season.

PAGE 100

87 frost free days in Trinidad, Colorado was 123. U npredictable growing season s may ha ve caused the inhabitants to respond by complementing their daily processing needs. Bedrock ground stone may have been used in addition to the portable ground stone during this time period, rather than as a preference. Starch and pollen analysis conducte d on the small mano revealed the presence of wild resources by extension, this indicates th at portable metates may have been used for wild resources and the bedrock for maize requires further research. No single environmental factor influenced where different activities would take place in the region. Rather , the combination of construction/tool material and fuel in the Piñon juniper woodlands , abundant game animals, economically useful plants, natural rock shelt er s present in the side canyons , and permanent water sources all contributed to the appeal of the greater Trinchera Creek basin (Black, 2000: 67). The presence of rock art and architectural elements may be an indication that spiritual motiv ations also played a role in some site locations ; most notably , Trinchera Cave. Suggestions for Future Research Site Protection Different warning signs have been posted at the shelter over the years. In the early 1990s, looting got so bad that a sign was posted on the northwest side of Trinchera Creek to help prevent against the illegal removal of artifacts (Garcia, personal communication, 2018). T hat sign was removed at some point by looters. New signage was est ablished on the slope below the entrance to the shelter in 2013. However, that sign was discovered to be missing in 2018. More protection is desperately needed for this site. It is recommended that Trinchera Cave be periodically monitored by the Office of the State Archaeologist, both to help deter future looting and vandalism and to ensure that visible signage remains in place (Zier, 2015: 96). Regular

PAGE 101

88 security patrols should be implemented at the site, especially during the summer months when visits to the nearby springs are at the highest (Garcia, personal communication, 2018). In order to help protect the cultural deposits still present at the site, fill should be brought in from an outside source to refill the shelter. Use of the back dirt on the slope below the shelter is not recommended because the slope may become unstable and more susceptible to erosional processes (Zier, 2015: 96) . Additionally, in situ cultural remains may be present. With the formal excavation of the slope outside the Museum Collections All of the known collections discovered from archaeological excavations between 1949 and 2001 are currently housed at the Louden Henritze Archaeology Museum in Trinidad, Colorado. While some of the artifacts have disappeared fro m the facility long ago, a collection of items representing site occupation from the Late Archaic through the La te Prehistoric remains. Although different aspects of the collection have been analyzed to varying degrees, no comprehensive examination has been attempted (Zier, 2015 : 96 ). The collection contains a wide range of remains including architectural element s, bone and faunal remains, ceramics, dental remains, ground stone, lithics, macrobotanical materials, perishables, phallic fragments, and worked/unworked shells. Radiocarbon dates should be obtained from artifacts excavated in all areas of the shelter an d at varying levels (Zier, 2015) . This could be supplemented by ceramic and projectile point analyses. This entire collection is long overdue for analysis. Prior to a comprehensive analysis, it is suggested that the entire collection be combined into a single cataloging system. CC student Kylie Crocket cataloged between 75 80% of Whether th is work is

PAGE 102

89 continued or starts anew is debatable. Future a nalysis should not begin until the entire collection is inventoried and cataloged (Zier, 2015) . In addition, each artifact should be repackaged separately with new, museum quality materials. Rath er than being completed by a single individual, it is recommended that a team work together to complete the much needed improvements. A project of this magnitude would need to be supported by grant money and would need to be completed in stages. Trincher a Cave has received too little s sporadic history of investigation , it has proven to have the potential to inform us about the prehistory of southeastern Colorado . Future Research Questions Findings at this site should be compared against other sites in the region to better understand mobility patterns and compare those components to those th at date to another period to better understand change through time. The presence of basketry, beads, bedding, bone awls, cordage, feathers, fibers, ground stone, imported calcium carbonate, Olivella sp. shells, phallic fragments, and the abundance of pote ntial medicinal plants suggest that Trinchera Cave may have been used for ceremonial activities, including initiation ceremonies for women. Shells are often interpreted as symbols of fertility (Claassen, 2009), as are phallic shaped objects (Howe, 1997; M eighan, 1997). The remains of a young adult female were uncovered in Area A (Simpson, 1976: 198). The woman had been partially cremated with only a few fragments remaining unburned (Simpson, 1976: 198). Black human hair was discovered entangled in coc klebur. Oil extracted from the burrs can be used for paints and varnishes (Curtin, 1947; Krochmal et al., 1954). Ancestors of the Zuni are known to have used the common cocklebur for different purposes (Stevenson, 1915). The chewed seeds were often used in ceremonies, and can

PAGE 103

90 also be used as a bandage when applied to wounds (Stevenson, 1915). Human dentition was also uncovered at the site in the form of deciduous teeth, revealing the presence of infants and young children (Nowak and Gerhart, 2002: 262) . Fibers and yucca leaves were used extensively in the production of a wide variety of objects (Simpson, 1976: 180). Area C produced a tumpline made from whole yucca leaves braided together (Simpson, 1976: 181). Tumplines were often used by Native Ame rican groups to help carry heavy loads; such as baskets, firewood, and infants (Hedges, 1997). A variety of yucca and grass bundles tied together with cordage were also recovered. One of the bundles was soiled and may have been used as a menstrual pad (S impson, 1976: 181). Chase excavated a second possible menstrual pad constructed from cedar and yucca fibers in Area B (Simpson, analysis. These materials give a strong indicatio n that Trinchera Cave was a place used by women for various purposes. The gendered interpretation of the site should also be investigated and would require a lot of additional evidence to substantiate this claim; such as DNA or blood residue analysis, comparative work, and so on. periods of birthing, healing, sickness, and menstruation (Claassen, 2011).

PAGE 104

91 REFERENCES Adams, J. (1988). Use We ar Analysis of Handstones Used to Grind Corn and Process Hides. Journal of Field Archaeology , 15 (3), 307 315. Adams, J. (1989a). Methods for Improving G round Stone Artifacts Analysis: Experiments in Mano Wear Patterns. In Experiments in Lithic Technology (Vol. 528, pp. 259 276). BAR International Series 528 , British Archaeological Reports , Oxford. Adams, J. (1989b). Experimental Replication of the Use of Ground Stone Tools. Kiva , 54(3), 261 271. Adams, J. ( 1993a). Mechanisms of Wear on Ground Stone Surfaces. Pacific Coast Archaeological Society Quarterly , 29 (4), 61 74. Adams, J. (1993b). Toward Understanding the Technological Development of Manos and Metates. Kiva , 58 (3), 331 344. Adams, J. (199 4 ). The Development of Prehistoric Gr inding Technology in the Point of Pines Museum Anthropology , 19 (3), 17 29. Adams, J. (1995 a ). The Development of Prehistoric Grinding Technology in the Point of Pines Museum Anthropology , 19 (3), 17 29. Adams, J. (1995b). The Ground Stone Assemblage: The Development of a Prehistoric Grinding Technology in the Eastern Tonto Basin. In The Roosevelt Community Development Study , 1 , 43 114. Adams, J. (1998). Ground S tone A rtifacts. In Archaeological I nvestigations of E arly V illage S ites in the M iddle Santa Cruz Valley: A nalysis and S ynthesis. Anthropological Papers , 19 , 35 7 422. Adams, J. (1999 ). Refocusing the Role of Food Grinding Tools as Correlates for Subsistence Strategies in the US Southwest. American Antiquity , 64 (3), 475 498. Adams, J. (2010). Engendering H ouseholds T hrough T echnological I dentity. In Engendering H ouseholds in the P rehistoric S outhwest , 208 228. Adams, J. (2014a). Ground Stone Analysis: A Technological Approach . University of Utah Press. Adams, J. (2014b ). Ground Stone Use Wear Analysis: A Review of Terminology and Experimental Methods. Journal of Archaeological Science , 48 , 129 138.

PAGE 105

92 Adams, J., Delgado, S., Dubreuil, L., Hamon, C., Plisson, H., and Risch, R. (200 9 ). Functional A nalysis of M acro L ithic A rtefacts : A Focus on Working Surfaces . In Non Flint Raw Material Use in Prehistory: Old Prejudices and New Directions , 43 66. Archaeopress. Andrefsky , W. (1990 ). An Introduction to the Archaeology of Pinon Canyon, Southeastern Colorado. Larson Tibesar Associates. Andrefsky, W . (1994). Raw Material Availability and the Organization of Technology. American Antiquity , 59, 21 25. Andrefsky, W. (2005). Lithics: Macroscopic Approaches to Analysis (Cambridge Manuals in Archaeology) . Cambridge University Press. Anonymous. (1952). Additions to TJC Faculty Staff Announced by President Baird. The Trojan Tribune (student paper of Trinidad State Junior College) 9 October: 1. Trinidad, Colorado. Baca, P. (2018, June 11). Personal communication. Bailey, R., and Aunger, R. (1989). Net Hunters Vs. Archers: Variation in Women's Subsistence Strategies in the Ituri Forest. Human Ecology , 17 (3), 273 297. Bamforth, D. (1986). Technological E fficiency and T ool C uration. American A ntiquity , 51 (1), 3 8 50. Barrett, S., and Gifford, E. (1933). Miwok material culture . Yosemite National History Association. Bartlett, K. (1933). Pueblo M illing Stones of the Flagstaff Region and T heir Relation to Others in the Southwest: A Study in Progressive Efficiency , 3. Northern Arizona Society of Science and Art. Barton, C. (1988). Lithic V ariability and Middle Paleolithic B ehavior: N ew E vidence from the Iberian Peninsula . BAR International Series 40 8 , British Archaeological Reports , Oxford. Barton, C. (1990). Beyond Style and Function: A View from the Middle Paleolithic. American Anthropologist , 92, 57 72. Basgall, M. (1987). Resource intensification among hunter gatherers: acorn economies in prehistoric California. Research in Economic Anthropology , 9 (198), 21 52. Beier, P., and Noss, R. (1998 ). Do Habitat Corridors Provide Connectivity?. Conservation B iology , 12(6), 1241 1252.

PAGE 106

93 Bellorado, B. (2011). Pushing the Limits and Tormenting Corn Seeds Cultural Adaptations and Climatic Chance in the Upper San Juan during the Basketmaker II Pueblo and Beyond. Southwestern Lore , 77(2), 33 47. Bettinger, R., Winterhalder, B., and McElreat h, R. (2006). A Simple Model of Technological Intensification . Journal of Archaeological Science , 33 (4), 538 545. Binford, L . , ( 1965 ) . Archaeological Systematics and the Study of Culture Process. American Antiquity , 203 210. Binford, L. (1980). Willow S moke and D T ails: H unter G atherer S ettlement S ystems and Archaeological Site Formation. American Antiquity , 45 (1), 4 20. Binford, L. (1990). Mobility, Housing, and Environment: A Comparative Study. Journal of Anthropologi cal Research , 46 (2), 11 9 152. Journal of Archaeological Research , 14 (2), 143 188. Black, K. (2000). Archaeological Survey and PAAC Training in the Trinchera Cave Area, Las Animas County, Colorado . Colorado Historical Society , Office of the State Archaeologists of Colorado. Bleed, P. (1986). The Optimal D esign of H unting W eapons: M aintainability or R eliability. American A ntiquity , 51 (4), 737 747. Blumenschine, R. ( 1988 ) . An experimental model of the timing of hominid and carnivore influence on archaeological bone assemblages. Journal of Archaeological Science 15 ( 5 ) , 483 502. Bohrer, V. (1954). Addendum to Report No. 307. Ethnobotanical Laboratory, Museum of Anthropology, University of Michigan, Ann Arbor. Bright, J., Ugan, A., and Hunsaker, L. (2002). The Effect of Handling Time on Subsistence Technology. World Archaeology , 34 (1), 164 181. Broughton, J. (2002). Prey S patial S tructure and B ehavior A ffect A rchaeological T ests of O ptimal F oraging M odels: E xamples from the Emeryville Shellmound V ertebrate F auna. World A rchaeology , 34 (1), 60 83. Broughton, J., and Cannon, M. (2010). Evolutionary Ecology and Archaeology: Applications to Problems in Human Evolution and Prehistory . University of Utah Press.

PAGE 107

94 Brown, J., and Davidson, D. (1977). Competition Between Seed Eating Rodents and Ants in Desert Ecosystems. Science , 196 (4292), 880 882. Brush, G. (2001). Forests before and after the C olonial E ncounter. Discovering the Chesapeake: T he H istory of an Ecosystem . John Hopkins University Press, 40 59. Bryant, V., and Holloway, R. (1983). The Role of Palynology in Archaeology. In Advances in archaeological method and theory , 191 224. Buonasera, T. (2015). Modeling the costs and benefits of manufacturing expedient milling tools. Journal of Archaeological Science , 57 , 335 344. Buonasera, T. (2016). Lipid residues preserved in sheltered bedrock features at Gila Cliff D wellings National Monument, New Mexico. Journal of Lithic Studies 3(3), 78 101. Calcagno, V., Mailleret, L., Wajnberg, É., and Grognard, F. (2014). How optimal f oragers should respond to habitat changes: a reanalysis of the Marginal Value Theorem. Journal of M athematical B iology , 69 (5), 1237 1265. Campbell, R. (1969). Prehistoric Panhandle Culture on the Chaquaqua Plateau, Southeast Colorado . Doctoral Dissertation, University of Colorado. Carmody, S., and Hollenbach, K. (2013). The r ole of g athering in Middle Archaic s ocial c omplexity in the M id South: a d iachronic p erspective. In Barely Surviving or More Than Enough , 2 9 57. Cassells, E. (1997). The Archaeology of Colorado . Johnson Books, Boulder. Cannon, M. (2003). A M odel of C entral P lace F orager P rey C hoice and an A pplication to F aunal R emains from the Mimbres Valley, New Mexico. Journal of Anthropological Archaeology , 22 (1), 1 25. Carle, P. (1941). Burial Customs of the Indians of the Southwest . Doctoral D issertation, Texas Tech University. Chaney, R. (1936). The Food of Peking Man. Scientific American , 154 (1), 14 16. Charnov , E. (1976 ). Optimal foraging, the Marginal Value Theorem. Theoretical Population Biology , 9, 129 136. Chase, H. (1950). Unpublished letter to Morris Taylor dated August 7, 1950. Letter on file, Louden Henritze Archaeology Museum, Trinidad, Colorado. Chomko, S., and De Vore, S. (1990). Sorenson Prehistoric Fortifications (Site 5LA330 and 5LA331): An Apishapa Village on the Purgatoire River . Interagency Archaeological Services, Rocky Mountain Regional Office, National Park, Denver, Co.

PAGE 108

95 Claassen, C. (2009). Shell S ymbolism in North America. In Early Human Impact on Megamolluscs , edited by Andrzej Antczak and Roberto Cipriani, pp. 37 43. BAR International Series S1865, British Archaeological Reports, Oxford. American Antiquity , 76(4), 628 641. Clark, J. (1988). The L ithic A rtifacts of La Libertad, Chiapas, Mexico: A n E conomic P erspective , 52. Brigham Young University. Codding, B., and Bird, D. (2015). Behavioral ecology and the future of archaeological science. Journal of Archaeological Science , 56 , 9 20. Colorado Encyclopedia . (2016, August 25). Snake Blakeslee Archaeological Site. Colorado Encyclopedia . Retrieved from Http://coloradoencyclopedia.org/article/snake blakeslee archaeological site . Coffin, R. (1937). Northern Colorado's First Settlers. Co lorado State College. Cook, W. (1987). The Wen, the Botany, and the Mexican Hat: the adventures of the first women through Grand Canyon on the Nevills Expedition . Borgo Press. Crocket, K. (2002). Trinidad State Junior College Collections, Louden Henritze Archaeology Museum, Recovered from 5LA1057. Appendix B in Archaeological Investigations in Southeastern Colorado , by Michael Nowak and Heather J. Gerhart, pp. 164 261. Colorad o College Publications in Archaeology, No. 22. Colorado Springs, Colorado. Cummings, L. (2018, May 25 ). Personal communication. Cummings, L., and Varney, R. (2009). Pollen, Starch, and Organic Residue (Ftir) Analysis of Bedrock Mortars from Colorado. PRI Technical Report. Cummings, L., and Varney, R. (2018). Pollen, Starch, and XRF Analysis from Site 5LA9555, Las Animas County, Colorado. PRI Technical Report. Curtin, L. (1947). Healing Herbs of the Upper Rio Grande. Laboratory of Anthropology, Santa Fe, New Mexico . Curtis, E . (1924). The North American Indian, Vol. 14: The Kato, the Wailaki, the Yuki, the Pomo, the Wintun, the Maidu, the Miwok, the Yokuts. ES Curtis, Seattle . Cushing, F. (1920). Zuni B readstuff , 8 . Museum of the American Indian, Heye Foundation. Czichos, H., and Dowson, D. (1978). Tribology: A Systems Approach to the Science and Technology of Friction, Lubrication and Wear . Tribology Series No. 1. Elsevier Scientific Publishing Company.

PAGE 109

96 Daniel, I. (2001). Stone R aw M aterial A vailability and Early Archaic S ettlement in the S outheastern United States. American Antiquity , 66 (2), 237 265. Davis, O. (1994). Aspects of Archaeological Palynology: Methodology and Applications. AASP Contributions , 29, 1 5. Delgado Raack, S., Gómez Gras, D., and Risch, R. (2009). The Mechanical Properties of Macrolithic Artifacts: A Methodological Background for Functi onal Analysis. Journal of Archaeological Science , 36 (9), 1823 1831. Dick Bassonnette , L. (1998). Gender and Authority among the Yokuch, Mono, and Miwok of Central California. Journal of Anthropology , 54(1), 49 72. Diehl, M. (1996). The I ntensity of M aize P rocessing and P roduction in U pland Mogollon P ithouse V illages A . D . 200 1000. American Antiquity , 61 (1), 10 2 115. Dobosi, V. (1991). Economy and R aw M aterial: A C ase S tudy of T hree Upper Paleolithic S ites in Hungary. In Raw Material Economies A mong Prehistoric Hunter Gatherers , 197 203. Dobres, M., and Hoffman, C. (1994). Social Agency and the Dynamics of Prehistoric Technology. Journal of A rchaeological M ethod and T heory , 1 (3), 211 258. Duce, J. (1918). The Effect of Cattle on the Erosion of Canon Bottoms. Science , 47(1219), 450 452. Ebeling, J., and Rowan, Y . (2004). The Archaeology of the Daily Grind: Ground Stone Tools and Food Production in the Southern Levant. Near Eastern Archaeology , 67 (2), 108 117. Eighmy, J. (1984). Colorado Plains Prehistoric Context . Colorado Historical Society, Office of Archaeology and Historic Preservation, Denver . Emlen, J. (1966). The Role of Time and Energy in Food Preference. The American Naturalist , 100 (916), 611 617. Evans, C. (2012). The Saint Charles River Project: An Exploration of Prehistoric Trade, Thesis, Department of Anthropology, Colorad o State University, Fort Collins. Faegri, K., Kaland, P., and Krzywinski, K. (1989). Textbook of pollen analysis . John Wiley & Sons . Fenneman, N. (1917). Physiographic Subdivision of the United States. Proceedings of the National Academy of Sciences , 3 (1), 17 22.

PAGE 110

97 Fitzhugh, B., and Habu, J. (Eds.). (2002). Beyond foraging and collecting: evolutionary change in hunter gatherer settlement systems (Vol. 1). Springer Science & Business Media. Fonner, R . (1950). Report No. 307. Ethnobotanical Laboratory, Museum of Anthropology, University of Michigan, Ann. Arbor. Frisbie, C. (1967). Kinaalda: A S tudy of the Navaho G irl's P uberty C eremony . Univ ersity of Utah Pr ess. Fuka, L. (2018, October 8). Personal communication. Garcia, B. (2018 , June 2 3 ). Personal communication. Gavin, T. (1955, August 28). Finds at Trinidad Disclose Ancient Culture. Rocky Mountain News , 12. Gayton, A. (1948a). Yokuts and Western Mono Ethnography : Tulare Lake, Southern Valley , and Central Foothill Yokuts . University of California Press. Gayton, A. ( 1948 b ). Yokuts and W estern Mono E thnography : N orthern Foothill Yokuts and W estern Mono . University of California Press. Gerhart, H. (2001). Chasing the Lost Field Record of Trinchera Cave. I n Archaeological Investigations in Southeastern Colorado , 115 131. Gifford, E. (1932). The Northfork Mono. University of California Publications in American Archaeology and Ethnology , 31, 15 65. Gleichman, P. (2002). Botanical and Small Scale Remains, Trinchera Cave (5LA1057) 2000/01 Excavations. In Archaeological Investigations in Southeastern Colorado. Colorado College Publications in Archaeology, No. 2 1 . The Colorado College, Colorado Springs, Colorado . Goodyear, A. (1993). Tool K it E ntropy and B ipolar R eduction: A S tudy of I nterassemblage L ithic V ariability among Paleo Indian S ites in the N ortheastern United States. In North American Archaeologist , 14 (1), 1 23. Gould, R. (1980). Living Archaeology . Cambridge University Press. Greenwald, D. (1990). A F unctional E valuation of Hohokam F ood G rinding S ystems. Doctoral D issertation, Northern Arizona University. Greenwald, D. (1993). Ground Stone Artifacts from La Ciudad de los Hornos. In In the Shadow of South Mountain: The Pre Classic Hohokam of La Ciudad de Los Hornos, 1991 1992 Excavations, Part I , 317 358. Archaeological Report No. 93 30. SWCA Environmental Consultants, Inc.

PAGE 111

98 Grove, M. ( 2009 ) . Hunter gatherer movement patterns: Causes and constraints. Journal of Anthropological Archaeology 28 (2), 222 233 Gunnerson, J., Hamblin, N., Cummings, L., and Schaafsma, C. (1989). Apishapa Canyon Archeology: Excavations at the Cramer, Snake Blakeslee and Nearby Sites . J & L Reprint Company. Hammond, G., and Rey, A. (1940). Narratives of the Coronado Expedition, 1540 1542 , (2). University of New Mexico Press. Hard, R. (1986). Ecological Relationships Affecting the Rise of Farming Economies: A Test from the American Southwest . Doctoral Dissertation, University of New Mexico. Hard, R. (1990). Agricultural Dependence in the Mountain Mogollon. In Perspectives on Southwestern Prehistory . Westview Press, 135 149. Hard, R., Mauldin, R., and Raymond, G. (1996). Mano Size, Stable Carbon Isotope Ratios, and Macrobotanical Remains as Multiple Lines of Evidence of Maize Dependence in the American Southwest. Journal of Archaeological Method and Theory , 3 (3), 253 318. Hardy, K. (2007). Survival, E xtraction and I dentification of S tarch G ranules at Kaman Kalehöyük, Turkey. AAS XVI , 189 194. Hartley, R., and Vawser, A. (2003). Rockshelters, Rock Art, and Grinding Activity: A Preliminary Assessment of Relationships in Picket Wire Canyonlands, Comanche National Grasslands . Midwest Archaeological Center National Park Service. Hawkes, K., O'Connell, F., and Jones, N. (1995). Hadza Children's Foraging: Juvenile Dependency, Social Arrangements, and Mobility among Hunter Gatherers. Current Anthropology , 36 (4), 688 700. Hayden, B. (1987). Past to P resent U ses of S tone T ools in the Maya Highlands , 160 234. University of Arizona Press. Hedges, K. (1997). Fibers & Forms: Native American Basketry of the West . Kiva Publishing. Hill, K., and Hawkes, K. (1983). Neotropical H unting A mong the Ache of E astern Paraguay. In Adaptive R esponses of N ative Amazonians , 139 188. Hollenbach, K. (2009). Foraging in the Tennessee River Valley: 12,500 to 8,000 years ago . University of Alabama Press. Homsey Messer, L. (2010). The Hunter Gatherer Use of Caves and Rockshelters in the American Midsouth: A Geoarchaeological and Spatial An alysis of Archaeological Features at Dust Cave . Archaeopress.

PAGE 112

99 Homsey Messer, L. (2015). Revisiting the Role of Caves and Rockshelters in the Hunter Gatherer Taskscape of the Archaic Midsouth. American Antiquity , 80 (2), 332 352. Horsfall, G. (1987 ). Design Theory and Grinding Stones. In Lithic Studies Among the Contemporary Highland Maya , 332 377. Howe, C. (1997). Fertility Symbols of the Western Indians . Graphic Press. Ireland, S., Dick, H., Henritze, R., and Wood, C. (1974). Trinidad Reservoir Salvage Archaeology, 1972 . Department of Anthropology, Trinidad State Junior College. Irwin Williams, C., and Irwin, H. (1966). Excavations at Magic Mountain: A Diachronic Study of Plains Southwest Relations . Denver Museum of Nature and Science, Denver. Jackson, T. (1991). Pounding A corn: W P roduction as S ocial and E conomic F ocus. Engendering A rchaeology: Women and P rehistory . Wiley Blackwell, 301 325. Johnson, M. (2018). Detergents: Triton X 100, Tween 20, and More. Material and Methods , 3, 163 172. Johnson, A., and Earle, T. (2000). The Evolution of Human Societies: From Foraging Group to Agrarian S tate . Stanford University Press. Kalasz, S., and Shields, W. (1997). Report of the 1994 / 1996 grid block archaeological excavations at the Magic Mountain Site (5JF223) in Jefferson County, Colorado. Report prepared for The State of Colorado Historical Fund Project , 96 01. Keene, A. (1983). Biology , behavior, and borrowing: a critical examination of optimal foraging theory in archaeology. In Archaeological H ammers and T heories , 137 155. Kelly, R. ( 1983 ) . Hunter G atherer M obility S trategies. Journal of A nthropological Research 39 ( 3), 277 306 . Kelly, R. (2013). The Lifeways of Hunter Gatherers: The Foraging Spectrum . Cambridge University Press. Kelly, R., Poyer, L., and Tucker, B. (2005). An E thnoarchaeological S tudy of M obility, A rchitectural I nvestment, and F ood S haring among Madagascar's Mikea. American Anthropologist , 107 (3), 40 3 416. Kelly, R., Poyer, L., and Tucker, B. (2006). Mobility and Houses in Southwestern Madagascar: Ethnoarchaeology Among the Mikea and Their Neighbors. Archaeology and Ethnoarchaeology of Mobility , 75 107. Kelly , R., and Todd, L. (1988). Coming into the Country: Early Paleoindian Hunting and Mobility. American Antiquity 53(2), 231 244.

PAGE 113

100 Kelso, G., and Good, I. (1995). Quseir Al Qadim, Egypt, and the P otential of A rchaeological P ollen A nalysis in the Near East. Journal of F ield Archaeology, 22(2) , 19 1 202. Kent, S. (1992). Studying Variability in the Archaeological Record: An Ethnoarchaeological Model for Distinguishing Mobility Patterns. American Antiquity 57(4), 635 660. Kingery, W. (1989). Ceramic M aterials S cience in S ociety. Annual Review of Materials Science , 19 ( 1), 1 21. Kluckhohn, C., Hill, W., and Kluckhohn, L. (1971). Navaho material culture . Belknap Press. Kooyman, B. (2015). Starch Granules: Preparation and Archaeological Extraction. In Plant Microtechniques and Protocols , 525 540. Krochmal, A., Paur, S., and Duisberg, P. (1954). Useful Nativ e Plants in the American Southwestern Deserts. Economic Botany , 8 (1), 3 20. L ithic raw material economy in the Mousterian of W est C entral Italy. Journal of Anthropological Archaeology , 10 (1), 7 6 106. Kuhn, S., and Miller, D. (2015). Artifacts as Patches: The Marginal Value Theorem and Stone Tool Life Histories. Lithic Technological Systems and Evolutionary Theory , 172 197. Lancaster, J. (1983). An Analysis of Manos and Metates from the Mimbres Valley, New Mexico . Doctoral Dissertation, University of New Mexico. Lancaster, J. (1984). Ground Stone Artifacts. In The Galaz Ruin: A Prehistoric Mimbres Village in Southwestern New Mexico , 247 262. University of New Mexico Press. Lemonnier, P. (1986 ). The Study of Material Culture Today: Toward an Anthropology of Technical Systems. Journal of Anthropological Archaeology, 5(2), 147 186. Li, Y., Zhou, L., and Cui, H. (2008). Pollen indicators of human activity. Chinese Science Bulletin , 53 (9), 128 1 1293. Lindström , S. (1996). Great Basin Fisherfolk: Optimal Diet Breadth Modeling the Truckee River Aboriginal Subsistence Fishery. Prehistoric H unter G atherer F ishing S trategies , 114 179. Lintz, C. (1999). Haldon Chase, the Snake Blakeslee Site, and the Archaeology of Southeastern Colorado : 1949 to 1955. Southwestern Lore , 65(2) , 5 31. Loendorf, L. (2008). Hunting, Grinding, and Dancing: Petroglyph Sites on the J.E. Canyon Ranch, Southeastern Colorado . Colorado Rock Art Association, Colorado Historical Society.

PAGE 114

101 Loendorf, L. and Loendorf, C. (1999). Archaeological Sites in Welsh Canyon, Las Animas County, Colorado . Midwest Archaeological Center National Park Service. Louden, R., and Louden, W. (1998). A Recounting of Cer tain Archaeological Excavations of Southeastern Colorado from 1949 to the Present. Paper presented at the annual meeting of the Colorado Archaeological Society. Pueblo, C O . Lupo, K. (2007). Evolutionary F oraging M odels in Z ooarchaeological A nalysis: R ecen t A pplications and F uture C hallenges. Journal of A rchaeological Research , 15 (2), 143 189. Lyman, R. ( 1994 ) . Vertebrate Taphonomy . Cambridge University Press. Lynch, E. (2010). Bedrock Metates along the Chaquaqua Drainage: Building a Conceptual Framework of Prehistoric Landscape Knowledge. Paper presented to the Society of American Archaeology Annual Meeting. Tucson, AZ. Lynch, E. (2014). Prehistoric Grinding Landscapes of the Southern Plains: The Archaeology of Bedrock Ground Stone F eatures on the Chaquaqua Plateau in Southeastern Colorado . University of Wyoming. Lynch, E. (2017a). Ancient Shapes, Modern Measures: A Quantitative Method to Describe Bedrock Ground Stone Shape. Journal of Archaeological Science: Reports , 13 , 211 221. Lynch, E. (2017b). Bedrock Ground Stone Features on Chacuaco Creek, Southeastern Colorado. Plains Anthropologist , 62 (243), 219 246. Lynch, E. (2017c). Unraveling the Enigma of Prehistoric Bedrock Ground Stone Features on the Chaquaqua Plateau, Usi ng Close Range Photogrammetry. Quaternary International , 439 , 50 68. Lynch, E., Holthus, L., and S imons , S . (2012). Does size really matter? An investigation of inter observer error rate in the field recording of bedrock grinding surfaces in S outheastern Colorado. Poster presented to the Society of American Archaeology Annual Meeting, Memphis, Tennessee. Lync h, E., Holthus, L., and Sanchez, K. (2013). Southern Plains Grinding Landscapes: Can a Local Wood Rat Midden Inform Us About Prehistoric Human Grinding Work Spaces and Behavior at a Colorado Rockshelter? . Quercus , 12 (6), 13. MacArthur, R., and Pianka, E. (1966). On Optimal Use of a Patchy Environment. The American Naturalist , 100 (916), 603 609. Madsen, D. ( 1993). Testing Diet Breadth Models: Examining Adaptive Change in the Late Prehistoric Great Basin. Journal of Archaeological Science , 20 (3), 321 329.

PAGE 115

102 Madsen, J. (2003). Cultural Resources of the Santa Rita Experimental Range. US Department of Agriculture, Forest Service, Rocky Mountain Research Station , 30, 68 79. Madsen, D., and Schmitt, D. (1998). Mass Coll ecting and the Diet Breadth Model: A Great Basin Example. Journal of Archaeological Science , 25 (5), 445 455. Martin , L. (2018 a , May 14 18). Personal communication. Martin, L. (2018b, November 26 December 4). Personal communication. Mauldin, R. (1991). Thumping and Grinding: Ethnographic Observations on Ground Stone from the Andes. In 56 th Annual Meeting of the Society for American Archaeology, New Orleans . Mauldin, R. (1993). The R elationship B etween G round S tone and A gricultural I ntensification in W estern New Mexico. Kiva , 58 (3), 317 330. McCarthy, L., and Bank, C. (2003, December). A Preliminary Geophysical Study Involving Remote Sensing at the Archaeological Site Trinchera Cave, Colorado. In AGU Fall Meeting Abstracts . Meighan, C. (1997). Foreword. In Fertility Symbols of the Western Indians . Graphic Press, 1 2. Meltzer, D. (1985). On Stone Procurement and Settlement Mobility in Eastern Fluted Point Groups. In North American Archaeologist , 6 (1), 1 24. Mercuri, A. (2008). Archeopalinology and cult contexts: pollen as a witness to ritual practices . In Men, plants, and animals in the dimension of the sacred , 14 7 159. Mercuri, A., Accorsi, C., Bandini Mazzanti, M., Bosi, G., Trevisan Grandi, G., Card a ll eri, G., et al. (2006). Cereal fields from the Middle Recent Bronze Age, as found in the Terramara di Montale, in the Po Plain (Emilia Romagna, Northern Italy), based on pollen, seeds/fruits and microcharcoals. In The archaeolog y of crop fields and gardens , 251 270. Mercuri, A., Florenzano, A., Massamba N'siala, I., Olmi, L., Roubis, D., and Sogliani, F. (2010). Pollen from archaeological layers and cultural landscape reconstruction: Case studies from the Bradano Valley (Basil icata, southern Italy). Plant Biosystems , 144 (4), 888 901. Mohr, A. (1954). The Deep Basined Metate of the Southern California Coast. American Antiquity , 19 (4), 394 396. NDawn Center . (2019). Corn Growing Degree Days. North Dakota Agricultural Weather Network . Retrieved from Https://Ndawn.ndsu.nodak.edu/help corn growing degree days.html .

PAGE 116

103 Nelson, M. (1996). The Value of a Technological Strategies Approach: Introductory Remarks. In Interpreting Southwestern Diversity: Underlying Principles and Overarching Patterns . Arizona Stat e University, 185 187. Nelson, S., Berry, K., Carrillo, R., Clark, B., Rhodes, L., and Saitta, D. (2001). Denver: An Archaeological History . University of Pennsylvania Press. Nettle, D., Gibson, M., Lawson, D., and Sear, R. (2013). Human behavioral ecology: current research and future prospects. Behavioral Ecology , 24 (5), 1031 1040. Noble, D. (2000). Ancient Colorado: An Archaeological Perspective . Colorado Council of Professional Archaeologists. Nowak, M. (2009). Trinchera Cave: A Study in Salvage Archaeology. Manuscript on file, Department of Anthropology, Colorado College, Colorado Springs, Colorado. Nowak, M., and Budnick , N. (2000). Archaeological Investigations in Southeastern Colorado. Colo rado College Publications in Archaeology, No. 2 0 . The Colorado College, Colorado Springs, Colorado. Nowak, M., and Gerhart, H. (200 1 ). Archaeological Investigations in Southeastern Colorado. Colorado College Publications in Archaeology, No. 2 1 . The Colorado College, Colorado Springs, Colorado. Nowak, M., and Gerhart, H. (2002). Archaeological Investigations in Southeastern Colorado. Colorado College Publications in Archaeology, No. 22. The Colorado College, Colorado Springs, Colorado. O'Connell, J. (1995). Ethnoarchaeology Needs a General Theory of Behavior. Journal of Archaeological Research , 3 (3), 205 255. Odell, G. (1977). The Application of Micro Wear Analysis to the Lithic Component of an Entire Prehistoric Settlement: Methods, Problems, and Functional Reconstructions . Doctoral Dissertation, Harvard University. Odell, G. (1989). Fitting A nalytical T echniques to P rehistoric P roblems with L ithic D ata. Archeological Papers of the American Anthropological Association , 1 (1), 15 9 182. Okada, F. (1949a). Unpublished Letter to Arnold Withers dated May 30, 1949. Letter on file, Department of Anthropology, University of Denver, Denver. Okada, F. (1949b). Unpublished Letter to Arnold W ithers dated June 15, 1949. Letter o n file, Department of Anthropology , University of Denver, Denver.

PAGE 117

104 Okada, F. (1949c). Unpublished Letter to Arnold Withers dated November 22, 1949. Letter on file, Department of Anthropology, University of Denver, Denver . Https://www.almanac.com/gardening/frostdates/CO/Trinidad# . Olson, R. (1936). The Quinault Indians . University of Washington Publications in Anthropology 6(1). Owens, M. (2007). A Model for Predicting Late Prehistoric Architectural Sites at the Pinon Canyon Maneuver Site in Southeastern Colorado . Directorate of Environmental Compliance and Management, Fort Carson and the National Park Service, Midwest Archaeological Center. Ft. Belvoir Defense Technical Information Center. Painter, M., Holmes, A., McFaul, M., and Zier, C. (1999). Environmental Setting. In Colorado Prehistory: A Context for the Arkansas River Basin, 5 24. Colorado Council of Professional Arch ae ologists . Pearsall, D. (2000). Paleoethnobotany: A Handbook of Procedures. Academic Press. Pierce, C. (1989). A C ritique of M iddle R ange T heory in Archaeology. Unpublished paper: University of Washington. Http://vixra. org/pdf/1201.0059 v1. pdf . Pulliam, H. (1974). On the Theory of Optimal Diets. The American Naturalist, 108 (959), 59 74. Raab, M., and Goodyear, A. (1984). Middle Range Theory in Archaeology: A Critical Review of Origins and Applications . American Antiquity 49(2), 255 268. Renaud, E. (1931). Archaeological Survey of Eastern Colorado . University of Denver, Denver. Renaud, E. (1935). The Archaeological Survey of Colorado, Fourth Report, Seasons 1933 and 1934. Ms. On file, Colorado Historical Society, Office of Archaeology and Historic Preservation, Denver. Renaud, E. (1931 1947). Reports and Surveys by Renaud. Etienne B. Renaud Papers, M015.05. University of Denver Libraries, Special Collections and Archives. Https://Duarchives.coalliance.org/repositories/2/archival_objects/1309. Accessed September 05, 2018. Renaud, E., and Chatin, J. (1943). Archaeological Sites of the Cuchara Drainage, Southern Colorado . University of Denver, Department of Anthropology. Rick, J. ( 1976 ) . Downslope Movement and Archaeological Intrasite Spatial Analysis. American Antiquity , 41 (2), 133 144.

PAGE 118

105 Rohd e , B. (2018 , September 10 12 ). Personal communication. Rowe, H., Hughes, N., and Robinson, K. (2012). The Quantification and Application of Handheld Energy Dispersive X Ray Fluorescence (ED XRF) in Mudrock Chemostratigraphy and G eochemistry. Chemical Geology , 324 , 122 131. Satterthwait, L. (1987). Socioeconomic Implications of Australian Aboriginal Net Hunting. Man , 613 636. Schiffer, M. (1992). Technological Perspectives on Behavioral Change . University of Arizona Press. Schiffer, M., and Skibo, J. (1987). Theory and Experiment in the Study of Technological Change. Current Anthropology , 28 (5), 595 622. Schiller, S. (2018, September 10). Personal communication. Schlanger, S. (1991). On M anos, M etates, and the H istory of S ite O ccupations. American Antiquity , 56 (3), 46 0 474. Searcy, M. (2011). The Life Giving Stone: Ethnoarchaeology of Maya Metates . University of Arizona Press. Seeman, M. (1994). Intercluster L ithic P atterning at Nobles Pond: A C D P rocurement among E arly Paleoindian S ocieties. American Antiquity , 59 (2), 273 288. Simms, S. (1987). Behavioral Ecology and Hunter Gatherer Foraging: An Example from the Great Basin (Vol. 381). BAR International Series 381, British Archaeological Reports, Oxford. Simpson, C. (1974). Excavations at Trinchera Cave: A Preliminary Report. Southwestern Lore , 40, 53 56. Simpson, C. (1976 ). Trinchera Cave: A Rock Shelter in Southeastern Colorado University of Wyoming. Smith, E. (1991). Inujjuamiut Foraging Strategies: Evolutionary Ecology of an Arctic Hunting Economy . Aldine de Gruyter. Smith (1983). Anthropological Applications of Optimal Foraging Theory: A Critical Review [and comments and reply]. Current Anthropology, 24 (5), 625 651. Speakman, R., and Shackley, M. (2013). Silo Science and Portable XRF in Archaeology: A Response to Frahm. Journal of Archaeological Science , 40 (2), 1435 1443.

PAGE 119

106 Spier, L. (1933). Yuman T ribes of the Gila River . University of Chicago Press. Stephens, D., and Krebs, J. (1986). Foraging theory . Princeton University Press. Stevenson, M. (1915). Ethnobotany of the Zuni Indians , 1. U.S. Government Printing Office, Washington, D.C. Stodder, A. ( 2008 ) . Taphonomy and the Nature of Archaeological Assemblages. Biological Anthropology of the Human Skeleton , 2, 73 115. Stoffle, R., Dobyns, H., Evans, M., and Stewart, O. (1984). Toyavita Piavuhuru Koroin = Canyon of Mother Earth: Ethnohistory and Native American Religious Concerns in the Fort Carson Pinon Canyon Maneuver Area . University of Wisconsin. Stone, T. (1994). The Impact of Raw Material Scarcity on Ground Stone Manufacture and Use: An Example from the Phoenix Basin Hohokam. American Antiquity , 59 (4), 680 694. Surovell, T. (2000). Early Paleoindian Women, Children, Mobility, and Fertility. American Antiquity , 65 (3), 493 508. Surovell, T., Toohey, J., Myers, A., LaBelle, J., Ahern, J., and Reisig, B. (2017). The End of Archaeological Discovery. American Antiquity , 82 (2), 288 300. Szeri, A. (1980). Tribology: Friction, Lubrication, and Wear . Hemisphere Publishing. Teer, D., and Arnell, R. (1975). Wear. In Principles of Tribology . MacMillan Press, 94 107. True, D. (1993). Bedrock M illing E lements as I ndicators of S ubsistence and S ettlement P atterns in N orthern San Diego County, California. Pacific Coast Archaeological Society Quarterly , 29 (2), 1 26. Ugan, A. (2005). Does Size Matter? Body Size, Mass Collecting, and Their Implications for Understanding Prehistoric Foraging Behavior. American Antiquity , 70 (1), 75 89. University of Denver Anthropology Museum (2018). Snake Blakeslee Apishapa Canyon Archaeological Site. Retrieved fro m Https://Www.Du.edu/ahss/ anthropology/museum/ exhibits/virtualexhibits/blakeslee/index.html. Varney, R.A. (2018, May 25). Personal communication. Walker, W., a nd Fratt, L. (1991). Under the Weather: A Systematic Study of Weathering on Groundstone Artifacts Recovered from Survey , Southwestern Anthropological Association meeting, Tucson / The Society for American Archaeology meeting, New Orleans.

PAGE 120

107 Wallace, H., and Holmlund, J. (1983). The M ortars, P etroglyphs, and T rincheras on Rillito Peak. The Kiva , 48, 137 246. Webb, W., and Funkhouser, W. (1929). The so American Anthropologist , 31 (4), 701 709. Whallon, R. (2006). Social networks and information: Non gatherers. Journal of A nthropological A rchaeology 25(2), 259 270. Whelan, C., Whitaker, A., Rosenthal, J., . . . Wohlgemuth, E. (2013). Hunter G atherer S torage, S ettlement, and the O pportunity C osts of W F oraging. American Antiquity , 78 (4), 66 2 678. Wiant, M., and Hassen, H. (1985). The R ole of L ithic R esource A vailability and A ccessibility in the O rganization of T echnology. In Lithic Resource Procurement: Proceedings from the Second Conference on Prehistoric Chert Exploitation, Occasional Paper , Occasional Paper 4 , 101 114. Wilmsen, E., and Robert Jr, F. (1978). Lindenmeier, 1934 1974: Concluding report on investigations. Smithsonian Contributions to Anthropology . Winterhalder, B . , and Smith, E. (2000). Analyzing Adaptive Strategies: Human Behavioral Evolutionary Anthropology: Issues, News, and Reviews , 9 (2), 51 72. Woodbury, R. (1954). Prehistoric S tone I mplements of N ortheastern Arizona , 34. Periodicals Service Company. Wright, M. (1993). Simulated Use of Experimental Maize Grinding Tools from Southwestern Colorado. Kiva , 58 (3), 345 355. Zier, C. (1989). Archaeological Excavation of Recon John Shelter (5PE648) on the Fort Carson Military Reservation, Pueblo County, Colorado. Ms. # PE.NP.R3 on file, Colorado Historical Society, Office of Archaeology and Historic Preservation, Denver. Zier, C. (201 5 ). Reconstructing Trinchera Cave: An Examination of the Excavation History, Chronolo gy, and Stratigraphy of Site 5LA1057, Las Animas County, Colorado . Submitted to Kinder Morgan, Inc. and History Colorado State Historical Fund by Centennial Archaeology , Inc., Fort Collins, Colorado. Zier, C. (2017). The Archaeology of Archaeology: Rec onstructing the Excavation History of Trinchera Cave. Southwestern Lore , 83(2). Zier, C., Kalasz, S., and Painter, M. (1999). Colorado Prehistory: A Context for the Arkansas River Basin . Colorado Council of Professional Arch ae ologists.