Chapter 1: IntroductionProblem Vast areas of the central Idaho important to the Shoshone-Bannock, Nez Perce and other native peoples have been barely surveyed by archaeologists; yet eff
Introduction
Vast areas of central Idaho, which hold significant cultural importance to the Shoshone-Bannock, Nez Perce, and other Native peoples, remain largely unexplored by archaeologists Effective cultural resource management is essential to protect these sites and ensure compliance with Section 106 of the National Historic Preservation Act Preserving these cultural landscapes is crucial for honoring indigenous heritage and preventing potential damage from development or neglect Increased archaeological surveys and proactive preservation efforts are needed to safeguard Idaho’s rich Native history for future generations.
The National Historic Preservation Act (NHPA) emphasizes the importance of understanding the distribution of archaeological sites across landscapes This thesis introduces an innovative approach to predicting site locations by analyzing the caloric costs of traversing different terrains—such as favoring lowlands—and the caloric benefits of nearby food resources through economic modeling Archaeological sites are defined as specific locations containing collections of artifacts or features, and this method aims to enhance our ability to identify these sites by integrating ecological and economic factors.
Current site location prediction methods evaluate whether a site meets specific parameters such as slope, proximity to water, and presence of critical species, increasing the likelihood of identifying a suitable site with more criteria met Incorporating an economic model of costs and benefits enhances the understanding of prehistoric land and resource utilization This thesis employs the Huff Model, originally designed to analyze shopping behavior, to predict site attractiveness based on travel caloric expenditure and the caloric yield from resources within the area, providing a more refined approach to site prediction.
This research examines the Frank Church River of No Return Wilderness (FC-RONRW), spanning over 9,000 square kilometers in central Idaho and managed by four national forests The wilderness area contains a documented total of 1,279 archaeological sites, highlighting its rich cultural history Due to the rugged terrain of the region, access and excavation pose significant challenges to archaeological research and preservation efforts.
Limited federal funding and challenging terrain have restricted archaeological surveys in wilderness areas, primarily concentrating on river corridors Few studies have explored higher elevations, leaving significant gaps in understanding upland site distributions within the FC-RONRW (Canaday 2012) There is an urgent need for targeted archaeological investigations in mountain uplands to develop more accurate models of site locations The methodology introduced here aims to improve the efficiency of these efforts Scholars emphasize the importance of focusing on mountain habitats, which contain unique seasonal resources such as whitebark pine (Pinus albicaulis) that are more prevalent at higher elevations, highlighting the ecological significance of upland environments for archaeological research.
The Frank Church River of No Return is a significant site for high elevation archaeology, highlighting the importance of exploring ancient activities in mountainous regions Although previous research in neighboring areas has contributed valuable insights, most studies have concentrated on specific resources or practices, leaving broader archaeological understanding limited Expanding investigations in this region can deepen knowledge of past human activities at high elevations, making it a vital focus for future archaeological research.
Effective acorn storage practices, such as using Morgan's (2012) methods or whitebark pine nuts, are important for understanding site use; however, focusing narrowly on these aspects can overlook the influence of variability between sites, terrain, and seasons within broader cultural contexts (Stirn, 2012; Bettinger, 1991b) Seasonal changes, including plant blooming cycles and animal migrations, significantly alter which areas are desirable at different times of the year, complicating the process of locating and interpreting archaeological sites without extensive fieldwork.
Archaeologists agree that food resources significantly influence settlement patterns among hunter-gatherers, sparking ongoing research and debate This relationship is extensively studied through Optimal Foraging Theory (OFT), which offers models explaining how hunter-gatherers make subsistence decisions While OFT provides valuable insights into their decision-making processes, it has limited practical application for managing and protecting prehistoric sites because it lacks integration of spatial aspects.
The aspatial nature of Opportunity Fee Theory (OFT) limits its practical usefulness for cultural resource managers focused on locating and protecting archaeological sites This research builds on OFT’s core premise—that decisions are driven by maximizing benefits and minimizing costs—and integrates spatial economic modeling to predict potential archaeological site locations By adding a spatial component to traditional OFT models, this approach enhances the ability of cultural resource managers to identify and safeguard previously unknown archaeological sites effectively.
This thesis aims to develop and evaluate methods for predicting archaeological site locations, aiding upland survey efforts It investigates the relationship between terrain, resource use, and prehistoric seasonal settlements to better understand settlement patterns Additionally, it explores how the Huff model can assess how seasonal resource availability influences settlement desirability within canyon corridors The study also evaluates whether GIS-based models of seasonal settlements and resource utilization can effectively predict future archaeological site locations Figure 2 illustrates the approach taken to address these research questions comprehensively.
This study involves compiling and evaluating current site inventories alongside analyzing environmental data using GIS to develop a predictive model for prehistoric site locations in the FC-RONRW uplands By integrating cultural and environmental information, the research assesses seasonal resources and subsistence strategies to accurately predict where archaeological sites are likely to be found Building on Hackenberger's (1984) work, the methodology is refined to emphasize that Native American sites tend to be located in accessible areas offering higher caloric resources, highlighting the importance of resource availability and ease of access in settlement patterns.
Implementing this methodology involves identifying the locations and caloric values of upland resources like roots, berries, game, and pine nuts, as well as lowland resources such as anadromous fish runs By estimating the calorie yield per square kilometer based on dominant land cover, we can predict the resource's overall value.
Understanding how terrain and resource distribution influence site selection requires analyzing resource use across different elevations throughout seasonal cycles High elevation areas, defined as above 2,100 meters (6,888 feet), experience significant changes in vegetation productivity, particularly for species like ponderosa pine (Pinus ponderosa) and Douglas-fir (Pseudotsuga menziesii) These elevations also mark the arrival of certain plants such as whitebark pine and spruce (Picea glauca) Incorporating terrain variability into the analysis enhances the accuracy of predictive models for settlement and resource exploitation.
Accessing resources involves traversing terrain with varying costs, such as steep slopes versus flat valley bottoms Using a formula from the American College of Sports Medicine (ACSM) combined with elevation data from the FC-RONRW, GIS can accurately estimate the caloric expenditure for different routes across the landscape I utilize GIS to generate a cost surface model that displays the travel costs between two specific points, enabling comparison of caloric expenditure with the potential benefits gained from resource harvesting along those routes.
The caloric value of available resources combined with spatial access costs is used to create an index that predicts potential archaeological site locations By adapting the Huff Model (Huff, 1964), this index helps identify high-probability sites, facilitating targeted archaeological exploration and research.
Literature Review
This article begins with essential background on the FC-RONRW and its inhabitants, providing context for understanding the region’s archaeological significance It summarizes key archaeological studies conducted in the FC-RONRW, highlighting important findings and insights An overview of resource use models is included to illustrate how ancient populations utilized available resources in the area The discussion on high elevation archaeology sheds light on the unique challenges and adaptations of communities living at higher altitudes Finally, the literature review synthesizes how previous research informs and supports the objectives of my current research project, emphasizing its relevance to ongoing studies in the region.
The Frank Church River of No Return Wilderness
The FC-RONRW is a vast wilderness area spanning over 9,000 square kilometers in central Idaho, known for its dynamic and imposing landscapes The region features significant elevation changes, ranging from 919 meters above sea level at the mouth of the Middle Fork Salmon River to 3,148 meters at the summit of General, one of the tallest peaks in the Salmon River Mountains This diverse topography creates a variety of environments, including alpine tundra, river valleys, and canyons, making it a unique natural habitat.
The geology of the area is primarily shaped by the Idaho Batholith, featuring mountains with alpine ridges, cirques, and expansive U-shaped valleys (McNab and Avers, 1994) Vegetation is predominantly composed of grand-fir (Abies grandis), Douglas-fir, western spruce, and ponderosa pine (Pinus ponderosa), supporting a diverse forest ecosystem The soils in this region are generally shallow to moderately deep, with volcanic deposits contributing to their high fertility, which promotes healthy plant growth.
Designated as a wilderness area, the FC-RONRW serves as an ideal location for ecological and cultural studies due to its protected status, which shields it from many detrimental disturbances (Canaday, 2012) However, this designation also presents challenges, as certain activities like mining and cattle grazing are permitted in specific circumstances, potentially posing threats to the area's untouched environment (Wilderness.net, 2014).
Access to the area is challenged by restrictions such as the prohibition of 12-wheeled transport (except wheelchairs), making logistics difficult The scarcity of roads and trails helps protect many sites but also limits access, especially since most of the wilderness lies outside the river corridors The most cost-effective way to reach the FC-RONRW is by boat, which has led many studies to focus on the river (Canaday, 2012) However, this means that much of the wilderness remains difficult to access and understudied.
The Forest Service has conducted cultural resource surveys in accordance with the National Historic Preservation Act (NHPA), documenting a total of 1,279 cultural sites as of 2012 (Canaday, 2012) These surveys have identified various prehistoric sites, including lithic scatters, pictographs, rock shelters, and house pits, primarily in lowland areas As noted earlier, the surveys are less focused on upland regions The primary purpose of these assessments is to evaluate the potential impacts of expanding facilities such as campgrounds, boat ramps, and trails within the FC-RONRW, as well as monitoring how ongoing operations affect archaeological sites, following guidelines established in studies like Knudson et al (1981).
The early prehistory of central Idaho remains poorly understood, though archaeological evidence provides insights into the region's ancient past The earliest human occupation in the FC-RONRW area dates back to between 12,000 and 8,000 BP, based on four fragmentary fluted points indicative of the Paleo-Indian period Understanding this archaeological context is essential before discussing the tribes that historically inhabited and utilized this region.
Davis et al (2014) document the presence of Paleo-Indian deposits at the Cooper’s Ferry Site along the Salmon River, highlighting its significance in early human history in the region The Paleo-Indian period at this site predates the Archaic Period, which began around 8,000 years ago and marked a crucial transition in prehistoric occupation and cultural development.
Between 13,000 and 4,000 years Before Present (BP), the region's archaeological record shows that big game hunting remained a central aspect of the culture, with technological advancements such as the introduction of the atlatl dart leading to the development of smaller, notched points During the Proto-Historic Period, from around 4,000 BP until the 1800s, there is increased archaeological evidence of cultural changes, including the appearance of the Nez Perce and Shoshone-Bannock peoples in the record, coinciding with the arrival of Euro-American settlers.
The archaeology of the FC-RONRW is primarily influenced by Shoshone-Bannock and Nez Perce material cultures, with the Nez Perce historically settling along the Snake, Salmon, and Clearwater rivers in central Idaho, as well as parts of eastern Oregon and Washington Within this region, they focused their settlements along the main Salmon River and its major tributaries, while also utilizing interior resources Their lifestyle was heavily oriented toward river environments, featuring elaborate fishing systems and settlements along waterways, although they also exploited terrestrial resources The Nez Perce shared cultural traits with other Columbia Plateau tribes, such as architectural styles and a strong reliance on salmon, which accounted for approximately 50% of their diet, highlighting the significance of river resources in their subsistence and culture.
The Shoshone-Bannock are closely connected to Plains cultures, reflecting their cultural and historical ties (Murphy and Murphy, 1986) Traditionally based in southern Idaho, their territory encompasses key regions such as the Sawtooth Mountains, Lemhi Valley, Salmon River Valley, and Snake River.
The Shoshone-Bannock adopted a highly mobile lifestyle, relying less on permanent settlements and instead exploiting seasonal and variable food sources, unlike the Nez Perce who used semi-permanent winter villages revisited annually This mobility allowed them to adapt efficiently to changing resource availability across their territory However, in specific areas such as the Lemhi Valley, Middle Fork Salmon River, and upper Salmon River, they established semi-permanent villages, highlighting regional variations in settlement patterns.
Shoshone-Bannock villages traditionally housed populations during winter months, with family groups dispersing into smaller bands in summer (Steward, 1938) Their subsistence relied on game hunting and seed gathering, highlighting a primarily hunter-gatherer lifestyle (Murphy & Murphy, 1986; Walker, 1973) Notable subgroups of the Shoshone-Bannock include the Tukudika.
The Sheepeaters, also known as the Agaidika or Salmon Eaters, were primarily based in central Idaho, inhabiting the northern canyons of the Snake River Unlike most Shoshone-Bannock tribes, they often journeyed into the Yellowstone area in Montana to hunt buffalo (Bison bison) Due to the rugged terrain of their homeland, they did not adopt the horse as readily as other Native American tribes, influencing their mobility and hunting practices.
As the name suggests, they were known for the consumption of bighorn sheep (Ovis
The Lemhi were based in the Lemhi River Valley south of the FC-RONRW and unlike the Sheepeaters, they readily adopted the horse, which enhanced their ability to hunt buffalo They regularly undertook trips south into Utah or east into Montana, where buffalo herds were more abundant than in the mountainous regions of central Idaho (Madsen, 1979; Steward, 1938).
Within the FC-RONRW there was a wide variety of biotic resources that could be used by Native North Americans Animals that were important include mule deer
(Odocoileus hemionus), elk (Cervus canadensis), bighorn sheep (Ovis canadensis), salmon (Oncorhynchus spp.), and pacific lampreys (Lampetra tridentate) (Hackenberger
Methods and Analysis
To achieve my research objectives, I first identify upland resources based on their caloric value to assess their attractiveness Next, I estimate the costs involved in accessing these resource areas Using this data, I develop a predictive model with ArcMap GIS software to apply an economic analysis across a broad spatial area I then validate and refine the model by testing it against a previously surveyed region Finally, I produce survey maps that outline how future field research can be designed to test and validate the model effectively.
Economists and geographers have developed numerous models to understand how people interact with resources, many of which are applicable to archaeology (Wilson, 2012) The Huff Model (Huff, 1964), originally designed to predict trade areas for shopping centers, is particularly relevant to my research because it evaluates the attractiveness of a location relative to the effort needed to reach it, estimating the likelihood that an individual will utilize that center While these models assume an established road network—appropriate for modern travel—they do not fully account for the travel patterns of prehistoric Native Americans, who likely used different means of movement.
The Huff Model, although not initially designed to predict subsistence strategies, proves to be an effective tool for measuring attraction across space In my research, I will adapt this model to better capture hunter-gatherer strategies, providing valuable insights into spatial behavior and resource distribution.
34 analyzing caloric costs and payoffs As adapted for this study, the Huff Model is expressed as follows:
Figure 4 The Huff Model (webhelp.esri.com 2015b)
Pij represents the probability that a hunter-gatherer at point i will travel to resource location j, considering both resource size and travel cost Wi measures the energy content of the resource, while Dij indicates the caloric expenditure required for the journey The parameter α reflects how distance diminishes a site’s attractiveness, highlighting that resources closer or more valuable are more likely to be visited The model’s numerator captures the specific desirability of location i based on resource quality, whereas the denominator sums the attractiveness of all potential sites, making P the proportion of total resource appeal at location i To implement this model, ArcMap GIS software was utilized for data extraction, and a Visual Basic program was developed to run the calculations effectively.
The Huff Model, despite its distinct conceptual origins, shares significant similarities with the Optim Product of Transport (OFT) model, particularly the core assumption that individuals aim to maximize returns while minimizing effort This approach aligns closely with central place foraging theory, where a central “home base” is the starting point for activities across landscape patches Both models assume that foraging or resource-gathering occurs in a series of different locations, emphasizing efficient decision-making to optimize resource acquisition.
35 amount of energy that can be harvested from each patch modified by the energy it takes to reach each patch
The Huff Model differs from other optimal foraging theory (OFT) models by not ranking individual species; instead, it evaluates patches solely based on total caloric content Unlike the diet breadth model, which ranks prey species according to preference and considers selection order, the Huff Model compares patches within the broader environment to determine which are chosen, akin to the marginal value theorem but focused on patch selection rather than abandonment Although it appears similar to the patch choice model, the key distinction is that the Huff Model incorporates spatial factors, whereas the patch choice model primarily emphasizes prey rankings without spatial considerations.
Different Native American groups exhibit seasonal adaptations that may influence the accuracy of the Huff Model in predicting archaeological sites Currently, the Huff Model assumes all resources are equally valued, treating calories from different sources as interchangeable This simplification may lead to inaccuracies for groups like the Nez Perce, who favored specific resources such as salmon, resulting in site distributions that deviate from model predictions Incorporating the diet breadth model could improve predictions by accounting for resource preferences, but this approach requires further dedicated research.
36 testing this hypothesis would require extensive analysis of artifact, faunal and floral remains of archaeological sites to fully address and is beyond the purview of this thesis
The Huff Model was selected for this research due to its inherent spatial component and alignment with the assumptions of the Optimizing Formal Theory (OFT) Economists and marketers have extensively applied the Huff Model using Geographic Information Systems (GIS) over the past decades, highlighting its effectiveness in analyzing spatial consumer behavior and market dynamics.
2005) This research takes their methods and applies them to archaeology
Four areas within the Salmon-Challis National Forest, designated as Control 1, 2, 3, and 4, were selected to evaluate the Huff Model’s effectiveness in predicting optimal site locations These regions were chosen based on survey coverage and the presence of prehistoric sites, providing relevant data to assess the model's accuracy in site location prediction within forested landscapes.
The largest test area, Cover 1, spans 619 square kilometers and is home to 61 prehistoric sites, making it a significant archaeological zone Northeast of Cover 1, Control 2 covers 219 square kilometers and contains 85 prehistoric sites, highlighting its rich prehistoric heritage Control 3 encompasses 250 acres and features 21 prehistoric sites, further contributing to the region's archaeological importance.
224 square kilometers and contains 55 prehistoric sites I chose these areas to test the Huff Model because they contain a sufficient sample of archaeological sites, as well as adequate survey coverage
Due to limited survey coverage of the wilderness, the study areas are situated outside the FC-RONRW Accurate model testing requires a sufficiently large sample size, ideally 30 or more samples for reliable statistical analysis (McGrew et al., 2014) The FC-RONRW lacks such areas except along the rivers, which poses challenges since river corridors cover only a small part of the landscape and offer a limited range of resources, mainly fish Outside the wilderness, there are regions with adequate survey data, enabling more comprehensive model evaluation.
37 coverage, a wide variety of food sources, and sufficient sites for a sample Figure 5 shows the locations of these control areas in relation to the FC-RONRW
Figure 5: Control areas used to test the Huff Model
Applied to these four study areas, the Huff Model provides a set of predictions
This article explores the potential locations of Native American sites by analyzing areas with high Huff Model values, suggesting that these zones are more likely to contain such sites Conversely, sampling regions with low Huff Model scores is expected to yield few or no Native American sites, validating the model’s predictive power The study's findings indicate that while the Huff Model effectively explains site distribution in one of the four study areas, incorporating additional variables—such as the distribution of caloric resources and the costs associated with accessing them—could enhance future analysis This approach emphasizes the importance of multiple factors in understanding the spatial patterns of Native American archaeological sites.
To run the model effectively, three separate spreadsheets are required for each study area, which is divided into 200m x 200m grid cells One spreadsheet details the caloric value of food resources available in each cell, while another provides the caloric costs associated with traversing these cells The third spreadsheet specifies which cells should be evaluated Using GIS data from the Forest Service that maps vegetation within the study unit, I estimate the potential caloric harvest from each cell following methods outlined by Hackenberger (1984) These estimates are based on harvestable food resources within various vegetation types, such as whitebark pine and bighorn sheep (Canaday, 2012).
Calories are estimated based on the availability and abundance of plants and animals within specific vegetation zones These calculations consider the type and quantity of food resources at each location, tailored for different seasons and elevation levels Hackenberger classified high elevations as areas above 2100 meters, medium elevations between 1500 and 2100 meters, and low elevations below 1500 meters This approach helps to understand the energy intake potential within varying environmental conditions.
Calorie values for each vegetation and animal source are estimated by assessing the proportion of each food source within vegetation units and multiplying this by the total harvestable calories, then summing these amounts to determine overall food availability A similar calculation is applied to game animals, where the density of game in each vegetation unit is multiplied by the total calories they provide, including species like deer, elk, and bighorn sheep Fish contributions are incorporated in areas with major streams or rivers capable of supporting fishing activities Table 1 presents calorie data per vegetation zone, adapted from Hackenberger (1984), noting that calorie availability varies with changes in elevation.
Table 1: Calories by Vegetation Zone and Elevation
Calories from Game (Without Snowpack)
Calories from Game (With Snowpack)
Table 1: Calories by Vegetation Zone and Elevation
Table 1: Calories by Vegetation Zone and Elevation (Continued)