Settlement Ecology of the Ancient Americas 1138945560, 9781138945562

In this exciting new volume several leading researchers use settlement ecology, an emerging approach to the study of arc

120 76 8MB

English Pages 320 [283] Year 2016

Report DMCA / Copyright


Polecaj historie

Settlement Ecology of the Ancient Americas
 1138945560, 9781138945562

Table of contents :
Half Title
Title Page
Copyright Page
Table of Contents
List of figures
List of tables
List of contributors
Part I Overview
1 Settlement ecology of the ancient Americas: an introduction
Part II North America
2 The ecology of changing settlement patterns among Piedmont Village Tradition communities in southeastern North America, AD 800–1600
3 Settlement ecology at Singer-Moye: Mississippian history and demography in the southeastern United States
4 Settlement ecology in the precontact North American Southwest
Part III Central America
5 Political-economic strategies and settlement ecology in the Mesoamerican Gulf Lowlands: Olmec, Epi-Olmec, and Classic Period settlement in the El Mesón area of the Eastern Lower Papaloapan Basin, Veracruz, Mexico
6 Climate, ecology, and social change in prehispanic northwestern Mesoamerica
7 Agrarian settlement ecology in the Classic Maya Lowlands: a comparative analysis of La Joyanca (Guatemala) and Río Bec (Mexico)
8 Identifying settlement variability in the Isthmo-Colombian Area: alternative models from the Upper General Valley of the Diquís archaeological subregion
Part IV South America
9 Chanka settlement ecology: disentangling settlement decision-making during a time of risk in the Andean highlands
10 Encountering forgotten landscapes: water, climate, and two millennia of settlement location choices in the Ica-Nasca region of southern coastal Peru
11 The organics of settlement patterns in Amazonia

Citation preview

Settlement Ecology of the Ancient Americas In this exciting new volume several leading researchers use settlement ecology, an emerging approach to the study of archaeological settlements, to examine the spatial arrangement of prehistoric settlement patterns across the Americas. Positioned at the intersection of geography, human ecology, anthropology, economics and archaeology, this diverse collection showcases successful applications of the settlement ecology approach in archaeological studies and also discusses associated techniques such as GIS, remote sensing and statistical and modeling applications. Using these methodological advancements the contributors investigate the specific social, cultural and environmental factors which mediated the placement and arrangement of different sites. Of particular relevance to scholars of landscape and settlement archaeology, Settlement Ecology of the Ancient Americas provides fresh insights not only into past societies, but also present and future populations in a rapidly changing world. Lucas C. Kellett is Assistant Professor of Anthropology at the University of Maine at Farmington, USA. Eric E. Jones is Assistant Professor in the Department of Anthropology at Wake Forest University, USA.

Settlement Ecology of the Ancient Americas

Edited by Lucas C. Kellett and Eric E. Jones

First published 2017 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 711 Third Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2017 Lucas C. Kellett and Eric E. Jones for selection and editorial matter; individual chapters, the contributors The right of the editors to be identified as the authors of the editorial material, and of the contributors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data Names: Kellett, Lucas C., editor of compilation. | Jones, Eric E., editor of compilation. Title: Settlement ecology of the ancient Americas / edited by Lucas C. Kellett and Eric E. Jones. Description: Milton Park, Abingdon, Oxon ; New York, NY : Routledge, 2017. | Includes bibliographical references and index. Identifiers: LCCN 2016015489| ISBN 9781138945562 (hardback: alkaline paper) | ISBN 9781315671284 (ebook) Subjects: LCSH: Indians--Antiquities. | Prehistoric peoples--America--Antiquities. | America--Antiquities. | Land settlement patterns--America--History--To 1500. | Human ecology--America--History--To 1500. | Excavations (Archaeology)--America. | Social archaeology--America. | Landscape archaeology--America. | Environmental archaeology-America. Classification: LCC E61 .S46 2017 | DDC 970.01--dc23 LC record available at

ISBN: 978-1-138-94556-2 (hbk) ISBN: 978-1-315-67128-4 (ebk) Typeset in Bembo by Sunrise Setting Ltd, Brixham, UK

Contents List of figures List of tables List of contributors PART I

Overview 1

Settlement ecology of the ancient Americas: an introduction LUCAS C. KELLETT AND ERIC E. JONES


North America 2

The ecology of changing settlement patterns among Piedmont Village Tradition communities in southeastern North America, AD 800–1600 ERIC E. JONES


Settlement ecology at Singer-Moye: Mississippian history and demography in the southeastern United States STEFAN BRANNAN AND JENNIFER BIRCH


Settlement ecology in the precontact North American Southwest SCOTT E. INGRAM


Central America 5

Political–economic strategies and settlement ecology in the Mesoamerican Gulf Lowlands: Olmec, Epi-Olmec, and Classic Period settlement in the El Mesón area of the Eastern Lower Papaloapan Basin, Veracruz, Mexico MICHAEL L. LOUGHLIN


Climate, ecology, and social change in prehispanic northwestern Mesoamerica MICHELLE ELLIOTT

Agrarian settlement ecology in the Classic Maya Lowlands: a comparative analysis of La Joyanca (Guatemala) and Río Bec (Mexico)




Identifying settlement variability in the Isthmo-Colombian Area: alternative models from the Upper General Valley of the Diquís archaeological subregion ROBERTO A. HERRERA


South America 9

Chanka settlement ecology: disentangling settlement decision-making during a time of risk in the Andean highlands LUCAS C. KELLETT


Encountering forgotten landscapes: water, climate, and two millennia of settlement location choices in the Ica-Nasca region of southern coastal Peru HENDRIK VAN GIJSEGHEM


The organics of settlement patterns in Amazonia FERNANDO OZORIO DE ALMEIDA


Figures 2.1 Map showing the Piedmont region, major river valleys, and PVT settlement locations within each 2.2 Map showing the location of the upper Yadkin River Valley sites used in this study 2.3 Graphs of site-size versus time 2.4 Map showing the location of pre- and post-AD 1200 settlements in the upper Yadkin River Valley 3.1 Singer-Moye site plan, including site core and extent of UGA lands 3.2 Locations of sites discussed in text 3.3 Level IV ecoregions in the lower Chattahoochee River valley 3.4 Interpretive site maps based on gradiometer data and selected squares used for calculations 3.5 Singer-Moye Thiessen polygons 3.6 Extent of Singer-Moye by phase of occupation 3.7 Singer-Moye, early phase of occupation, c.AD 1100 to 1300 3.8 Singer-Moye, middle phase of occupation, c.AD 1300 to 1400 3.9 Singer-Moye, late phase of occupation, c.AD 1400 to 1500 4.1 Study area watersheds in central Arizona 4.2 Droughts in central Arizona from 1200 to 1450 4.3 Scatterplots of drought severity and residential abandonment by rooms in low- and highdensity watersheds 4.4 Scatterplots of drought severity and residential abandonment by riverine proximity 5.1 Gulf Coast with important sites mentioned in the text 5.2 Early Formative settlement in the El Mesón area 5.3 Middle Formative settlement in the El Mesón area 5.4 Late Formative settlement in the El Mesón area 5.5 TZPG complex at El Mesón 5.6 Protoclassic settlement in the El Mesón area 5.7 Early Classic settlement in the El Mesón area 5.8 Standard plan complex at El Mesón South 5.9 Late Classic settlement in the El Mesón area 5.10 Transportation corridors into the southern and western Tuxtla Mountains 6.1 Map of the northern frontier of Mesoamerica with the locations of the archaeological and paleoecological study sites mentioned in the text 6.2 Location of the Malpaso Valley 6.3 Contour map of La Quemada 6.4 Map of the Malpaso Valley showing the location of sites, the Malpaso River, and

6.5 6.6 6.7 6.8 7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8 7.9 8.1 8.2 8.3 8.4 8.5 8.6 8.7 8.8 8.9 9.1 9.2 9.3 9.4 9.5 9.6 9.7 9.8 10.1 10.2 10.3

seasonal streams Results of Elliott’s flow modeling of the Malpaso Valley Synthesis of the results from wood charcoal studies carried out at La Quemada, El Cóporo, and Cerro Barajas Study area of project for Elliott et al.’s paleoenvironmental reconstruction, with individual trenches shown Synthesis of the results of Elliott et al.’s off-site paleoenvironmental reconstruction for the Malpaso Valley Map of the Maya region with principal sites of La Joyanca and Río Bec shown Map of the La Joyanca micro-region Map of the Río Bec micro-region Methods applied to La Joyanca Methods applied to Río Bec Site plan of La Joyanca during the Late Classic period La Joyanca, neighborhoods and agricultural systems during the Late Classic period Site plan of Río Bec during Late Classic period Agrarian production units in Río Bec (50 ha) during the Late Classic period The Isthmo-Colombian Area The Greater Chiriquí archaeological region Plan and proposed layout of El Cholo structures with outlying circular structure in the southwest sector of the site Select examples of commemorative mortuary ritual at El Cholo Later Chiriquí period mortuary features found 900m east of El Cholo Chiriquí period habitation sites demonstrating separation of structures Radiometric dates for El Cholo Map of study zones within the Diquis subregion A rethinking of Greater Chiriquí Zones Simplistic conceptual model showing hierarchy of risk, which can be translated to a set of settlement priorities Regional map showing study area and Chanka ethnic region Andahuaylas Valley showing Chanka Settlement Project area Project area showing spatial distribution of Chanka habitation sites Photo of Toxsama Valley showing local high-relief landscape and domestic structures at the large Chanka phase ridgetop site of Achanchi Map showing the range of residential and subsistence structures located in the project area Map showing viewshed results calculated in GIS as well as defensive features recorded during survey work Map showing spring density in the project area The Ica-Nasca region with sites mentioned in the text Agricultural infrastructure in Quebrada La Yesera View of two successive unfinished canals in Quebrada Campanayoq near the Aja River, Nasca region

10.4 10.5 10.6 10.7 10.8 11.1 11.2 11.3 11.4 11.5

Hypothetical past hydrography of the Ica-Nasca region Settlement survey results in the Cocharcas and Tingue quebradas Well-preserved circular domestic architecture at Chokoltaja Huarangal, Quebrada Tingue: stone-lined canal and aerial view of some agricultural terraces and canals Google Earth™ views of TI-4 and Inka Tambo, and TI-3, a Nasca orthogonal structure Inside an abandoned house Ethnic groups and archaeological contexts cited in text The Mutuca site, guarding the pathway through the serra do Mutuca The Upper Madeira and Middle and Lower Jamari sites The Itapirema site and the two compared areas

Tables 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 3.1 3.2 3.3 3.4 3.5 4.1 4.2 4.3 4.4 4.5 7.1 7.2 8.1 9.1 10.1 11.1 11.2 11.3

Radiocarbon analysis results from the Redtail site Sites with occupations during ad 800–1200 used in this study Sites with occupations during ad 1200–1600 used in this study Environmental and landscape variables measured for each settlement and random point Landscape characteristics measured for the Upper and Lower Great Bend areas Discriminant function analysis results comparing the pre-ad 1200 settlements with the sets of random sites Discriminant function analysis results comparing the post-ad 1200 settlements with the sets of random sites The results of the landscape characterization, comparing the Upper Great Bend area with the Lower Great Bend areas Animal and plant species identified in the archaeological record at Singer-Moye Population estimates from Moundville and Etowah Radiocarbon dates discussed in text Residential settlement size and population estimates at Singer-Moye by phase Estimated population of contemporary sites in the Lower Chattahoochee River valley Droughts and the drought severity index Relationship between drought severity and the percentage of rooms abandoned in lowand high-density watersheds in central Arizona Central Arizona: slopes, intercepts, correlation coefficients, and their probability of equality by rooms located in low- and high-density watersheds Relationship between drought severity and the percent of rooms abandoned by riverine proximity in central Arizona Central Arizona rooms: slopes, intercepts, correlation coefficients, and their probability of equality by rooms and riverine proximity Comparison of household rank and APU surface estimates at Río Bec Comparison of household rank of individual neighborhood and the surface of associated cultivated area at La Joyanca Subregional chronology of Greater Chiriquí Observed vs. expected outcomes for settlement location within the project area General recent chronology for the Ica-Nasca region Features of the Upper Madeira sites Features of the Middle and Lower Jamari sites Comparison in ceramic assemblages between Area 1 and Area 2

Contributors Fernando Ozorio de Almeida, Universidade de São Paulo (University of São Paulo) Jennifer Birch, University of Georgia Stefan Brannan, University of Georgia Michelle Elliott, Université Paris (University of Paris) Roberto A. Herrera, Hunter College, City University of New York Scott E. Ingram, Colorado College Eric E. Jones, Wake Forest University Lucas C. Kellett, University of Maine at Farmington Eva Lemonnier, Université Paris (University of Paris) Michael L. Loughlin, University of Kentucky Hendrik Van Gijseghem, Dickinson College

Part I


1 Settlement ecology of the ancient Americas An introduction Lucas C. Kellett and Eric E. Jones Ancient settlement patterns have long fascinated archaeologists. Since its inception, settlement archaeology has played a crucial role in the comprehension of numerous complex archaeological topics, such as sociopolitical organization and development, state and imperial expansion, peer polity interaction, trade networks, demography, and economic organization, among others. The conceptual shift from a spatially limited and site-specific focus to a landscape-based regional perspective was indeed a watershed moment in archaeology (Willey 1953; see Billman 1999; Blanton et al. 2005). Through its multi-decade course of development, settlement pattern studies has matured into its own sub-field of archaeology where associated methodologies, technologies, and interpretive frameworks have continued to be refined (Kantner 2008; Kowalewski 2008). The critical question in settlement archaeology can be stated as follows: why do people settle in a given place during a specific time and in a particular arrangement? Ostensibly, this appears to be a simple question, yet the corresponding answer often remains frustratingly elusive. This is because a prehistoric settlement pattern is the result of complex decisionmaking in the face of innumerable social, political, and economic factors. As such, analyzing specific rationale for particular settlement decisions presents an especially challenging problem for archaeologists. In a similar vein, mutual causation or equifinality have plagued settlement pattern studies, requiring such studies to remain descriptive rather than explanatory (Stone 1996). Yet, rather than fading away in the modern era of archaeology, settlement pattern studies have done the opposite and witnessed a new renaissance in large part due to the rapid technological advances (e.g. Global Positioning Systems [GPS]; Geographic Information Systems [GIS]; remotely sensed data; and unmanned aerial vehicles [UAV]) in the past two decades. So, a number of readers who encounter this volume may likely first ask: do we really need another volume on ancient settlement patterns? Our answer to this question is yes for several reasons. First, the rapid change in technological innovation related to the recording, analyzing, and modeling of settlement patterns demands that archaeologists periodically address and evaluate how we use and interpret ancient settlement patterns. Second, while technology has rapidly “evolved,” there has been much less attention paid to reconceptualizing and the further development of theories behind settlement pattern analysis. As discussed below, archaeologists still most often describe rather than fully understand or explain settlement patterns and changes to them. A settlement ecology approach seeks to dissect settlement patterns and identify what influencing factors underlie how and why people decide to settle on

a given landscape. Finally, a basic understanding of how prehistoric people settled in a particular area over time is especially relevant in today’s rapidly changing world. Decisions of where to live in the past shape the distribution of our modern settlement patterns. Across the Americas, the distribution of indigenous settlements had a significant impact on the settlement patterns of early European colonizers. Furthermore, in the face of globalization, modernization, climate change, and migration, humans today are still having to making the critical decision of where to live often in the face of complex and difficult circumstances. We hope that this volume can extend our thinking beyond the past and into the present and future to more fully comprehend settlement as an essential part of the human experience (see Moore 2012). We presume a number of readers of this volume will also pose this question: what is settlement ecology? This will likely be followed by two other questions: where did it come from and is it really a new approach in settlement archaeology or just a rehash of previous models outlined decades ago? We argue that it is both. That is, it is both a new and more comprehensive way to think about ancient settlement patterns as well as a more sophisticated and refined synthesis of previous thinking and applications in regional studies and settlement archaeology. This volume attempts to bring together for the first time a group of scholars who are currently using a settlement ecology approach to answer complex archaeological questions across the Americas. In this volume contributors tackle a range of questions that in some way link to ancient settlement patterning, including settlement formation, aggregation, dispersion, abandonment, relocation, fission-fusion, and many others. These broad overlapping settlement phenomena are typically a response to a wide range of physical (environmental) and non-physical (sociocultural) pressures, factors, or priorities that influence settlement decision-making and ultimately help concretize a settlement arrangement in material form. The authors in this volume seek to understand why particular settlement patterns are established and what caused them to change over space and time.

Defining settlement archaeology In the simplest of terms, settlement archaeology is the study of settlement patterns. But how do we define a settlement pattern, and more importantly what does it signify to archaeologists about past cultures and their behaviors? Fish offers a useful definition of a settlement pattern: a settlement pattern is a set of culturally significant locations, each of which occupies a specified position within an array that makes up a coherent distribution … settlement patterns are spatial matrices marking the intersection of human activities and the natural environment. As such they provide a basis for examining the relationship between cultural loci and relevant geographic variables. Settlement patterns simultaneously mark the intersection of human activities and their cultural environment. They encode relationships among spatially distinct elements of societies and reflect the cumulative outcomes of spatially expressed decisions and interactions. (Fish 1999: 203)

This eloquent definition embodies the concept of a settlement pattern, by considering the significant components of space, culture, and geography. Thus, a working definition of settlement archaeology can be stated as the study of past culture through the examination of spatially defined loci of human activity. Studies in settlement archaeology share a number of common characteristics, which set them apart from other approaches in archaeology. Marquardt and Crumley (1987: 1–9) outline several underlying components of such a settlement approach. First, settlement archaeology makes special efforts to understand the archaeological landscape, which is the totality of the archaeological record in a given region, reflecting the interaction between humans, their culture, and their environment. Second, it considers the spatial orientation among different archaeological sites and between archaeological sites and the physical environment. Third, such studies adopt a regional perspective that examines the totality of settlement across a large area. As we discuss below, these basic components of settlement archaeology are important in contemporary studies and approaches, including settlement ecology.

The origins of settlement ecology Although the term settlement ecology was not coined until the mid-1990s, many of its principal tenets were in use well before. Processual archaeology in particular played an important role in how settlement patterns were originally conceived, how they were studied using archaeological methods, and how they were described through the development of middlelevel theory (e.g. Chang 1972; Flannery 1976; Trigger 1967, 1968; Parsons 1971, 1972; see also Billman and Feinman 1999; see Kantner 2008 and Kowalewski 2008 for reviews). Within the processual movement, cultural ecology and systems theory approaches were most commonly used to explain settlement patterns and shifting cultural dynamics (e.g. Binford 1968; Flannery 1968; Plog 1975; Struever 1968). In this context, settlement systems were often seen as adaptive to external stimuli. The processual period also saw the widespread adoption of systematic ground survey in numerous parts of the world, which elevated the value and importance of settlement patterns in large regional archaeological studies (e.g. Adams 1965, 1981; Billman and Feinman 1999; Blanton 1978, 2005; Chang 1972; Flannery 1976; Gumerman 1971; Parsons 1971; Parsons et al. 2000, 2013; Peterson and Drennan 2005; Sanders 1965; Sanders et al. 1979; Willey 1956). The processual era ushered in a shift from just describing broad patterns of settlement to also building deterministic models to explain and predict them. In addition, these models attempted to account for the specific determinants that influenced the formation of ancient settlement patterns (Trigger 1968; Peebles 1978). Most early considerations of settlement determinants included primarily environmental factors (e.g. water, resource availability) as well as population growth in reaction to environmental conditions (e.g. Brown et al. 1978; Flannery 1976; Peebles 1978; Sanders 1981). The majority of studies during this time were also focused on prehistoric hunter-gatherers (e.g. Binford 1980; Kelly 1985; Thomas 1972) much more so than on sedentary agrarian societies. The emphasis on the former was directly related to the rise of behavioral ecology (BE) and “optimal foraging theory” (OFT) within the field of archaeology and their shared focus on “two sets of phenomena: past human behavior

and its material consequences” (Bird and O’Connell 2006: 143; see Schiffer 1987; Smith and Winterhalder 1992). In particular, behavioral ecology aims at modeling past human behavior using a “fitness related landscape” in which to understand human action. While BE has faded in popularity (as originally conceived), it has impacted on approaches to settlement archaeology since the latter often adopts assumptions of economic efficiency (e.g. resource maximization, travel cost reduction) to understand settlement patterns and their changes. Settlement archaeology in the 1970s and 1980s also saw the heavy influence of geographic models (Hagget 1965; Johnson 1977) on the analysis of ancient settlement patterns, including Central Place Theory (CPT) (e.g. Christaller 1966; Crumley 1979; Evans and Gould 1982; Steponaitis 1978, 1981), Site Catchment Analysis (SCA) (e.g. Chisolm 1968; Higgs et al. 1967; Higgs and Vita-Finzi 1972; Vita-Finzi and Higgs 1970), and models of rural land use (Chisolm 1968; Von Thünen (1966) [1826]). Nearly all of these geographic models offer a series of assumptions, which help model human settlement arrangement. First, settlements are located on the landscape intentionally, not randomly, primarily for economic or material reasons. Second, through a detailed understanding of the surrounding physical landscape, one can understand settlement locations based on the spatial correlations between site location and local available resources. Third, since the costs of subsistence production and the related transport costs of goods and people increase as one moves away from a given settlement, people will place their settlements closest to the most critical of resources (otherwise known as the proximity principle). Finally, it is assumed that under normal conditions, the spatial arrangement of settlements is ordered in a predictable and hierarchical fashion (e.g. lattice patterns) to most efficiently produce and move goods among different sites (especially among market systems). While these geographic/spatial assumptions have been critiqued on several grounds (e.g. Crumley 1979; Stone 1996: 12–27), they still form an undeniable part of the continuing development of settlement archaeology, as well as the approach embodied by settlement ecology. The 1980s and 1990s witnessed the increasing frustration of some archaeologists with the trajectory of processual archaeology and thus spawned post-processual theory and approaches. This paradigm heralded a return to a consideration of individual actors and their lived and sensorial experiences in the past. Landscape as a term came to signify something that is or was more culturally constructed, as space had no meaning separate from human actions and thus all places were assumed to have symbolic meaning (Tilley 1994). Settlement archaeology, like other sub-fields, saw a shift in how sites were viewed, experienced, and interpreted with deliberate consideration of use of space and the construction of place (i.e. Ingold 1993; Llobera 1996, 2001; Tilley and Bennett 2001). At the same time, individual monuments, as well as entire settlement patterns, witnessed a more inclusive examination of the ideological, religious, social, and cultural phenomena, which may have helped plan and place sites on the landscape (i.e. Lekson 1999; Thomas 1991). In addition, the increased technological power and accessibility of GIS within archaeology witnessed an explosion in three-dimensional spatial modeling. Visual analyses, such as the popular viewshed analysis, showed some promise when used in an attempt to humanize the landscape and reconstruct the social relationships among groups of people and settlements (e.g. Llobera 2003, 2007; Wheatley 1995; Wheatley and Gillings 2000, 2002). Conversely, the

ability of GIS to analyze the relationships between measurable features of the landscape and settlement patterns can also be given some credit for the survival of processual ideas through the post-processual movement. Studies of the environmental influences on settlement patterns gained convincing empirical support (e.g. Allen 1996; Kvamme 1990), and survey methods like predictive modeling showed real correlations between environmental features and past human behavior (Bevan and Conolly 2002; Kohler 1988; Warren 1990a, 1990b; Wescott and Brandon 2000). Finally, before fully describing settlement ecology we must also briefly discuss the continuing role that landscape archaeology, including historical ecology, has played on settlement archaeology. Landscape archaeology can be broadly defined as the systematic study of how cultural and environmental variables influence the way humans interacted with their landscape (Ingold 1993 Hu, 2011: 80). Landscape archaeology has attempted to move past a site-centered, processual approach to understanding settlement patterns by treating the landscape as a formation of continuous culturally defined spaces in which humans actively create, use, manipulate, and experience landscapes (e.g. Ashmore and Knapp 1999; Crumley and Marquardt 1990; Gillings et al. 1999; Marquardt and Crumley 1987; Tilley 1994; Wagstaff 1987). Landscape archaeology remains a poorly defined term that subsumes a wide array of disparate approaches to landscape studies in archaeology, including more traditional scientific approaches as well as phenomenological and performative approaches (Bruno and Thomas 2010: Hu 2011: 80–81). In addition, while landscape archaeology has embraced geospatial technology (e.g. GIS, remote sensing) with mixed results, it has offered new and creative approaches to better comprehend how ancient peoples experienced their landscapes (Gillings and Goodrick 1996; Gillings et al. 1999; Hu 2011). Historical ecology is a closely related approach to landscape archaeology. It also employs the landscape as a lens to understand the long-term interaction between peoples and their environment (Crumley 1994). More specifically, historical ecology asserts that the landscape is inherently cultural and the result of persistent and intentional human action upon the physical and cultural environment (Balée 1998, 2006; Balée and Erickson 2006). At its core, historical ecology also fervently refutes “environmental determinism” and views settlement patterns and the built environment not as the result of human adaptation to the environments but as the accretional and materialized historical record of complex human environmental interaction. This approach, as with many of those previously discussed, also features strongly in the settlement ecology approach.

What is settlement ecology? While it is still unknown who first coined the term “settlement ecology,” we know that its first thorough consideration was by anthropologist Glenn Davis Stone in his book Settlement Ecology: The Social and Spatial Organization of Kofyar Agriculture (1996). In this book, Stone uses a synthetic approach from anthropology, economics, geography, and ecology to meticulously chart the historical changes to the settlement dynamics of Kofyar agriculturalists of Nigeria. This groundbreaking study offers a powerful interpretive frame through which to comprehend how competing polities, along with the Kofyar, positioned and subsequently

changed their settlement arrangements over time in the face of rapidly changing and interconnected historical, ethnic, environmental, and political conditions and processes. Settlement ecology as envisioned by Stone (1996) remedies a suggested bias in settlement studies towards hunter-gatherer ecology rather than agrarian ecology. He states that, “we have a wide and growing disparity in settlement theory, with our understanding of hunter-gatherer settlement far outstripping what we know about agrarian settlement” and as such he terms settlement ecology as the explicit examination of agrarian settlements or farming communities (Stone 1996: 5). As discussed further below, we do not feel that settlement ecology needs to be agrarian in nature as Stone asserts, but should be inclusive of all types of societies, modes of production, and degrees of mobility. As one can see from the chapters in this volume, we offer a broad perspective of settlement ecology, including semi-sedentary hunter-gatherers, semisedentary and sedentary agriculturalists, and mixed-strategy groups that combine multiple subsistence and settlement regimes. Settlement ecology is best understood as an outgrowth of location studies in geography, which uses an inherently spatial approach to understand the causes of particular settlement strategies. Building on work by Netting (1993), Stone creates a historically contingent agrarian settlement model of the Kofyar and more prominently sets out to make inroads in addressing causality in the formation of settlement patterns—something he believes has been long neglected in anthropology, archaeology, and geography (Stone 1996: 227). In response, Stone proposes that in order to develop a new theory on agrarian settlement, one needs to examine cause-and-effect relationships on settlement arrangements and associated change. Stone argues new settlement studies cannot just describe patterns but must build explanations to understand “[w]hat are the factors that push and pull agrarian settlements?” (Stone 1996: 13). In this context he argues that settlement theory should include the set of rules agricultural populations followed, which in part can address the problem of equifinality in settlement pattern studies (Stone 1996: 7, 13). Since the list of influences that may “determine” settlement patterns is “as long as one cares to make it,” as well as limited by the investigator’s imagination, a set of robust rules or determinants is needed to understand agrarian ecology. However, he is wary about building a set of rules for settlement behavior since real world case studies always deviate and offer exceptions to rule-based models. Instead, Stone (1996: 8) prefers “to think of [settlement rules as] priorities of varying strength.” As archaeologists, we, along with Stone (1996), argue than many scholars have tended to uncritically adopt settlement models often without full consideration of their assumptions and implications. For example, a priori assumptions concerning cost efficiency and settlement arrangements need to be deconstructed and evaluated critically.

Population and agricultural production Since many previous settlement studies derive models from human and economic geography, such work has concentrated more on consumptive and marketing factors rather than on productive and population factors. Stone adeptly makes the following point: There is an important gap in our knowledge of how the productive activities of rural

agricultural settlements affect the location, arrangement, size and duration of those settlements. As a result, archaeologists have come to rely on models that hold agricultural production constant even as population density rises. … Without a better understanding of the factors that actually drive agrarian settlements, it is impossible to adduce general models of agricultural settlement behavior. (Stone 1996: 27) As the quote demonstrates, settlement studies must be wary of holding conditions constant and must consider dynamically changing variables such as population pressure and changing subsistence production (e.g. extensification, intensification) and land tenure. He begins by profiling the migration of the Kofyar within central Nigeria, during the middle of the twentieth century, from the uplands of the Jos Plateau to the broad low-lying Muri Plains almost 30km to the south. His diachronic analysis provides an exceptional understanding of linked changes in settlement strategies and agricultural production. This first phase consisted of “Frontiering” in which low numbers of Kofyar gradually leapfrogged across the fertile piedmont to sandy plains, periodically abandoning settlements in the process. These groups practiced a shifting extensive agricultural pattern in a situation where land was abundant and productive. In the second phase, increasing numbers of Kofyar people settled on the plains causing land scarcity and intensive local agriculture. Demographic pressure co-occurred with the development of spatially defined sociopolitical and/or ethnic units called ungwa, and witnessed the formation of a complex system of labor exchange. From his research on the Kofyar, Stone (1996) offers two important ethnographically based conclusions for new settlement studies. First, the rules of agrarian settlement are embedded in the ecology of agricultural production and such settlement is integrally linked to agricultural intensification (Stone 1996: 181). Second, Kofyar settlement decisions were mediated by numerous sets of priorities, of which certain priorities favored particular locational solutions and the variation in the value of following (or cost of neglecting) each priority (Stone 1996: 182). Stone (1996: 182–184) offers several basic observations concerning agrarian settlement. He argues that land pressure does not automatically lead to intensification or site abandonment, but to a choice between the two (Stone 1996: 182). The decision is based on the effects of marginal work on total agricultural production. In general, increasing labor stimulates population aggregation, except for when inputs are divided among locations, which results in a more dispersed settlement pattern. Stone also argues that labor scheduling has been instrumental in the shaping of Kofyar settlement patterns. Since pooled labor is necessary for success in an intensive agricultural strategy, farm location and shape is defined by proximity to one’s own plots as well as those of neighbors. Stone argues that labor pooling has the effect of formatting the landscape into sociosettlement units (ungwa) containing similar ethnic groups that can manage localized labor exchange. In terms of ecology, Kofyar settlement locations were heavily influenced by the availability of water, which was critical for a successful agricultural regime. The one exception to this

pattern was that in some cases larger-sized plots were more important than locally available water. Finally, over time, Stone observed that with population and agricultural intensification, soil productivity was more important than water availability, especially in the context of decreasing agricultural yields. While Stone does make important contributions to agrarian settlement strategies, he still restates the warning offered by Grossman (1971: 23) that “general laws [of settlement] are meaningless outside the specific cultural and technological context.” Stone stresses the variation in responses that different cultures’ (e.g. subgroups of the Kofyar) settlement systems have to specific ecological constraints and general cultural goals despite access to similar land. In sum, a settlement ecology approach as outlined by Stone is a useful theoretical model for archaeologists. The particularly applicable components are the examples of how settlement changes across time and space and a framework for identifying causative factors and conceptualizing how people weigh those factors in settlement decisions. Moreover, the goal of moving from primarily a descriptive to an explanatory settlement archaeology—with a focus on isolating individual factors and their importance in settlement patterning—is crucial in the development of the field. In addition, Stone (1996) generally argues that settlement ecology is most useful when considered as a system, rather than a set of rules. Furthermore, as Jones (2010: 3) elaborates on Stone’s (1996) approach, the best method is to analyze cultures case-by-case and avoid establishing rules of settlement for any particular subsistence strategy. In each culture, there are too many factors originating from too many sources to establish any rules that incorporate all societies practicing a particular subsistence strategy. In this context, each settlement ecology is unique given a range of contingent factors such as occupational period, local environment, local history, subsistence organization, population, etc. Within this more realistic approach to settlement analysis “the benefits and drawbacks of each factor [are] weighed in each decision and [are] affected by the circumstances at the time” (Jones 2010: 10).

Conceptualizing settlement ecology Based on the above outline of Stone’s settlement ecology, we can identify a number of useful premises. However, we also feel that a more comprehensive and inclusive schema needs defining. In this section we attempt to clearly articulate the underlying conceptual underpinnings and then we will attempt to operationalize a settlement ecology approach. First, we argue as mentioned above that settlement ecology is an inclusive approach that is not limited to agrarian societies as outlined by Stone (1996), but rather can include societies of all types (e.g. hunter-gatherer, agricultural, pastoral) and specific characteristics (e.g. degree of social complexity, mobility/sedentism). Moreover, it is not necessarily limited to prehistoric case studies, but can be useful in understanding modern settlement patterns and changes thereof in numerous areas around the world in the face of rapidly changing conditions and

circumstances (e.g. globalizations, war, migration, urbanization, climate change). Second, we are not offering an entirely new perspective on settlement archaeology, but repacking and synthesizing numerous previous approaches and the ideas of countless archaeologists who have come before us. In addition, settlement ecology is not meant to offer a universal theory or law for human settlement as was the goal of many settlement models of the processual era. Rather it is best understood as a conceptual and methodological approach, which is more comprehensive, robust, and powerful for archaeological interpretation. In addition, we believe that settlement ecology is time and space contingent, and that settlement pattern analysis requires a consideration of primarily specific and local environmental, social, political, economic, ideological, and historical conditions. Although generalization about human settlement can certainly be constructed through cross-cultural comparisons of different settlement ecologies, one must be cautious of overgeneralizing results to create any sort of predictive model of settlement. Third, settlement ecology includes the term “ecology,” which is not accidental. Since ecology can be broadly defined as the “relationships and interactions between entities,” we also believe that this concept is very applicable to a more nuanced and dynamic understanding of human settlement. In archaeology, we can clearly state that a specific settlement strategy is rarely the result of a single phenomenon, but typically and necessarily the result of a total suite of cultural and ecological conditions, needs, pressures, and relationships. It is not just the influence of individual pressures upon settlement decisions, but how they intersect, connect, and impact on one another. For example, a heightened level of warfare may bring people together via local settlement aggregation for communal defense, which may reverberate numerous changes including new forms of political leadership, harden local ethnic divisions, and cause scalar stress while at the same time requiring further travel to fields and other resource areas. In this way the “push and pull” dynamics are complex, coupled, and often contingent, such that the individual strings are tied not just to the settlement, but to one another, resulting in a constantly changing webbed arrangement of settlement pressures, priorities, and values. In this way settlement is usefully considered partly as an adaptation to conditions, but also as an expression of human, cultural, and environmental relationships in a given time and space. Within an archaeological context there have already been a number of successful studies by Elliott (2005), Jones (2010, 2012) Hasenstab (1996), Kohler (1988), Maschner (1996a), Maschner and Stein (1995), and others, who all treat prehistoric settlement as an adaptation by local populations to both natural and cultural environments. For example, in his study of the Haudenosaunee, Hasenstab (1996) argues that settlements’ locations were not random, but strategically placed to have access to local resources and agricultural land, as well as maintain protection from local enemies. More recently, Jones (2010) analyzed 125 Haudenosaunee settlements against a range of landscape variables and found that transportation routes, conditions favorable to agricultural production, and stands of hardwoods heavily influenced the placement of settlements. By isolating those particular settlement factors and/or pressures, one can discern the causes that contributed towards a given culture’s settlement organization. Fourth, we argue that settlement ecology can offer the valuable perspective of a settlement pattern as the result of human decision-making. We emphasize that it is not individual factors

that “determine” a settlement arrangement (cf. Stone 1996), but it is the conscious decisions made by people in the face of these factors that ultimately create a pattern of settlements. This point is important as it maintains the importance of human agency and cultural context in a previously reductionist, anonymous, and often acultural tradition in settlement archaeology. Thus, settlement ecology encapsulates the totality of diverse needs, concerns, and resulting innumerable complex decisions of a society, which together define the size, type, location, and duration of portions or all individual sites of a given settlement pattern. In this way, a settlement pattern is the materialized record of human decision-making over time, which serves as a rich dataset with which to understand the lived experiences of particular cultures over time and space. Finally, we argue that settlement ecology requires a spatial consideration of the settlements in relation to one another. The site/settlement concept and its employment have been critiqued on a number of grounds (and alternatives for a “siteless” archaeology have been proposed: e.g. Ebert 1992; Rossignol and Wandsnider 1992). We contend that despite certain weaknesses, the site concept is still a meaningful analytical unit that can help us understand past human activities. In addition, we do not feel that site/settlement vs. landscape epistemological debate is necessary when aspects of both approaches are valuable in a settlement ecology approach. In large part this is because spatial analysis within GIS can accommodate both bounded entities like defensive forts and springs (e.g. nearest neighbor analysis) as well as continuous entities like elevation and slope (e.g. least cost path analysis). As such, GIS is not only a powerful spatial analytical tool but also a tool that allows for comprehensive examinations of both natural environments and cultural landscapes. This highlights the importance that a spatial analytical framework has in understanding the relative importance (e.g. spatial correlation, regression) that certain influences had on prehistoric settlement decision-making. While the use of GIS or other spatial technologies is not absolutely necessary for a settlement ecology approach, it does offer the best methodological approach through which to unravel the complex nature of prehistoric settlement patterns and the conditions and decision-making which underlie them (e.g. Wheatley and Gillings 2002; Maschner 1996a).

Operationalizing settlement ecology Our focus to this point has been theoretical, so it is critical at this juncture to discuss the ways in which archaeologists have executed settlement ecology studies. Just as it is difficult to pinpoint the origin of the term settlement ecology, it is difficult to identify the first settlement ecology study. A full treatment of the methodology of the approach requires a summary of the history of the analytical methods used (e.g. spatial analysis, GIS) and the steps in the research process necessary to establish an explanation of past settlement behaviors. We will start with the latter. As with all settlement archaeology studies, the results of a settlement ecology study will only be as good as the data collected. Most data for settlement ecology studies come from regional surveys, both past and present. Although smaller-scale analyses of intrasite patterning and even household patterns are starting to fluoresce (e.g. Berman 1994; Creese 2009, 2012), settlement ecology remains largely a regional or subregional venture. With the abundance of

regional settlement data and environmental data available in digital formats today, it is easier to access the required information for building explanations of past settlement behaviors. Settlement pattern data are generally more straightforward, assuming that the locational data are accurate and at the proper level of precision for the scale of analysis, and that the function of a site as a settlement can be established. The more difficult task for any settlement ecology study is the reconstruction of past environments Many studies of the recent past (i.e. focusing on the past 1000 years) rely on modern environmental and landscape data, and assume continuity in basic environmental and landscape features (e.g. Allen 1996; Hasenstab 1996; Maschner 1996a; Jones 2006, 2010; Jones et al. 2012; Jones and Ellis 2016). Although not perfect, the assumption of similarity in most locations without drastic human impacts is not overly problematic for recent cases. In addition, large-scale changes, such as sea-level change or dammed rivers, can easily be accounted for. More problematic are the attempts to characterize or describe environments and landscapes in the deep past, such as during the Pleistocene or Early Holocene. Where paleo-environmental data are not available—with cases of poor preservation or lack of research—this task can be extremely difficult and may require extensive modeling or simulation embedded with many assumptions. Where they are available, reconstructing total environments can still be difficult. For example, knowing the percentages of particular tree species that existed in an area does not tell you exactly where those trees were on the landscape. Brouwer Burg (2013) presents one of the best examples of creating “total” landscapes, or those landscapes reconstructed using a suite of paleo-environmental data and GIS-based spatial simulation software. She uses this method to approximate the location of major bodies of water and their impact on sediments and plant and animal life for the Post-Glacial Netherlands. Other exciting technologies, such as those that can estimate water-table levels using paleo-climatological data (French et al. 2012) or estimate vegetation composition using sediment types (Sipkins 2000), have yet to be applied widely in settlement ecology research but hold enormous potential given the difficulties of reconstructing past surface-water distribution as a result of the modification of wetlands, rivers, and lakes around the world over the past 200 years. As mentioned, the as a result of identification of research as settlement ecology followed about a decade after the establishment of GIS as a productive tool in archaeological research (Maschner 1996b; Maschner and Aldenderfer 1996). However, we must trace the methodological and theoretical roots of settlement ecology back much further to pioneering research that searched for associations between settlements and environmental features. Serious applications of spatial archaeology began 40 years ago, and were solidified by Hodder and Orton’s (1976) work. Included within were methods for determining spatial associations of archaeological remains with particular features of the environment. Similar work continued throughout the 1970s (e.g. Plog and Hill 1971; Thomas and Bettinger 1976), and was made considerably easier with the introduction of GIS technologies. By the late 1980s, GIS facilitated the study of settlement–environment feature associations by making datasets easier to compile, visualize, manipulate, and analyze and by making environmental variables easier to quantify (Kvamme 1999). In particular, GIS made it easier to determine whether observed associations were legitimate spatial correlations or autocorrelations. The latter occur when the observed pattern is a reflection of the background

data not an actual pattern. For example, if 70 percent of settlements exist on loamy sediments, it is not a significant percentage if loamy sediments cover 70 percent of the study area. In fact, that percentage would be expected in a random distribution of settlements with regard to sediment type. Analyses of autocorrelation occurred before the implementation of GIS (see Plog and Hill 1971), but GIS allowed researchers to use computing technology to summarize and measure the features of large geographic areas and to produce site distribution statistics in a fraction of the time it took with orthographic maps (Kvamme 1999). Thus, while the settlement ecologist’s methods existed prior to GIS, the technology expanded the scope of research. GIS allowed for the analysis of larger archaeological spatial datasets and for building easier and more manipulable maps of environments and landscapes. This, in turn, led to expansions in the number of people—because of the ease of obtaining and sharing data as well as the reduced need to understand complex spatial statistics—studying the relationships between settlements and environments and landscapes and in the amount of data that could be analyzed. Several studies throughout the 1990s used a variety of methods to identify significant correlations and explain them. Kvamme (1990) characterized the entire background landscape, comparing his archaeological sample with the sample universe. Allen (1996) and Hasenstab (1996) opted for comparing archaeological samples with random samples (i.e. control groups) across the settled landscape to approximate random settlement behavior. They employed this method to narrow the study area to those places occupied on the landscape. In a similar approach, Kvamme (1996) used Monte Carlo methods to compare archaeological samples with random locations of regular sizes (i.e. 99 or 999) to examine where the samples fall in an arbitrary distribution of points on a landscape. If they fall at the extremes, they represent a tendency away from random patterning. Again, GIS was not doing anything new in these studies; it simply made existing methods more powerful and efficient by making it easier to examine larger samples and larger control groups. With methods for estimating past landscapes and for establishing spatial correlations between archaeological and environmental and landscape data, the stage was set for explaining those patterns. As mentioned, many of the early settlement ecology studies focused on environmental features because the data are easy to obtain and measure. Data concerning past cultural landscapes and perceptions of the landscape are harder to acquire. In addition, the postmodern critique brought along with it questions of the static and utilitarian way in which landscapes were being studied and even inherent biases within GIS methods (Wheatley 1993; Harris and Lock 1995; Gillings and Goodrick 1996; Llobera 1996). Like many other areas of archaeological study that found scientific ways of examining both adaptational and ideational components of past behavior after the postmodern critique, many modern settlement ecology studies integrate environmental and cultural landscape data into traditional settlement location studies (Brannan and Birch, this volume; Jones, this volume; Lemmonier, this volume). Several studies (e.g. Arkush 2011; Borgstede and Mathieu 2007; Haas and Creamer 1993; Maschner 1996a; Jones 2006; Kellett 2010; Sakaguchi et al. 2010) have attempted to understand how settlement behaviors were influenced by intergroup violence and how communities would attempt to increase the defensibility of particular locations in a tumultuous sociopolitical landscape. Others (e.g. Bell and Lockbauer 2000; Llobera 2000; Carballo and Pluckhahn

2007) have examined how mobility and transportation influenced decisions of where to live. Studies of perceptions of the landscape and ideological components of space and place (e.g. Tilley and Bennett 2001; Llobera 2001) with interesting applications to understanding social inequality (Kosiba and Bauer 2013; Wernke 2013) have also been successfully undertaken. In addition, the range of settlement behaviors studies has begun to expand, using a myriad of other statistical and visualization methods. Examples include studies of settlement duration using hazards models (Jones and Wood 2012), studies of mobility and movement using least cost surfaces and incorporating ethnographic data on landscape (Howey 2007, 2011). Furthermore, the use of agent-based modeling is another innovation with enormous potential for helping us learn more about the spatial components of past human behavior and decisionmaking (Kohler and Gumerman 2000; Wurzer et al. 2015). We make the argument here that specialized studies of how particular factors influenced settlement location decisions are important, but more comprehensive studies, which try to approximate the settlement location decision and a range of natural and cultural landscape features that were considered, are even more important. There is no doubt that such studies are much more difficult, but they are necessary. Those that have been completed have produced valuable information from combined examinations of subsistence, sociopolitical, economic, and ideological data and how they made up the “mental balance sheet” of past decision-makers (e.g. Arkush 2011; Bauer and Kellett 2010; Elliott 2005; Hasenstab 1996; Jones 2010; Jones and Ellis 2016). In the long run, these attempts to understand decision-making with regard to settlement behaviors will be the meaningful contribution of settlement ecology to the study of the past. Theories concerning human-settlement patterning and human ecology are critical, but so is the ability to understand how these behaviors were experienced by past people.

Structure of this volume The broad goals of this volume are to usher in the next wave of settlement ecology research in archaeology and to establish it as a viable and stand-alone area of study. In the first 15–20 years of settlement ecology research, archaeologists focused primarily on synchronic regional settlement pattern analyses to produce results on settlement location choice and human interactions with resources and landscapes. This type of research has produced information that has helped us better understand several past behavioral patterns, including semi-sedentary swidden farmers (Allen 1996; Hasenstab 1996; Jones 2010) and the role of visibility in defense (Haas and Creamer 1993; Jones 2006; Maschner 1996a; Sakaguchi et al. 2010). These projects created the base of knowledge on human–landscape interactions and should continue to be an important part of settlement ecology. However, settlement ecology, as we defined it above, is about more than location choice. This volume embodies both the growth of settlement location choice studies as well as explorations into new topics concerning human settlement behavior, including migration patterns, subsistence–settlement dynamics, and the development of sociopolitical complexity. In addition, the research described in the following chapters shows a mix of theoretical orientations, case studies from a variety of geographic areas and time periods, and diverse methods. As such, we believe this volume can be the starting point for a proliferation of settlement ecology research within anthropology and the point of

recognition that this area of research is broadly applicable across a wide variety of research questions and theoretical orientations. The variability in settlement ecology research is on display in the first section of the book, which covers projects conducted in North America. In Chapter 2, Jones examines the environmental and sociopolitical factors that influenced a significant shift in settlement size and geographic range among Piedmont Village Tradition communities in the Southeast during ad 800–1600. This is a blending of the old and new strategies described above because it produces regional settlement location results from sites in the western Piedmont of North Carolina but uses them to examine changes in those preferences over time. In addition, this work incorporates political and economic landscape data that suggest that the formation of Mississippian hierarchical polities, and their participation in political and economic networks in adjacent areas, may have had a significant impact on the observed changes. In Chapter 3, Brannan and Birch examine the spatiotemporal characteristics of demographic change at the Singer-Moye site, a Mississippian center in the Southeast dating to ad 1100–1500. They use ceramic data from survey work and demographic modeling from similar Mississippian centers to reconstruct population sizes and use of space at the site. Using existing archaeological data, landscape characteristics, and an explicit historical ecological approach, they provide an explanation for the growth, decline, and shifts in geographic patterning and monumental construction seen over the life of this settlement. In Chapter 4, Ingram offers an intriguing view by focusing on patterns of settlement abandonment rather than settlement formation. He considers how drought affected the cessation of settlement activities at sites in the Southwest by examining the relationship between population size, proximity to rivers, and drought episodes in central Arizona during ad 1200–1450, a period of notable climatic instability. Using a quantitative approach, he establishes a connection between population density and drought but, interestingly, not one between proximity to rivers and drought. Perhaps more than any other chapter in this work, Ingram clearly displays the utility of archaeological settlement ecology research for modern people and societies as we think about solutions to our own problems with population density, access to water, and unstable climatic conditions. The second section of the book contains studies from Mesoamerica that include detailed examinations of both natural and cultural landscapes. In Chapter 5, Loughlin examines the role a major communication route played in the rise and fall of the Olmec center of El Mesón. Survey results establish the timeline of the site and its connections to other Olmec centers, like La Venta and Tres Zapotes, and an analysis of the stylistic properties of recovered artifacts and the chronology of monumental architecture helps him establish the complex relationship between control over exchange and political power. In Chapter 6, Elliott examines data and results from the Northern Frontier Region of Mesoamerica, particularly as they relate to the formation and collapse of the site of La Quemada, to begin forming an explanation for the observed settlement patterns and processes. This work challenges many of our current notions of how people expand into arid and “harsh” environments. She completes her work by outlining an innovative approach to completing the explanation, which incorporates multiple scales of analysis and several diverse lines of paleo-ecological and archaeological data. In Chapter 7, Lemonnier constructs a new model for Maya agricultural-architectural complexes. She employs settlement data from the La Joyanca and Río Bec sites and estimated agricultural

production values, and uses the results to characterize human–environment and human– landscape interactions. Her model has implications for social organization, as it pertains to the development of elite households in these two communities and how urban areas and monumental landscapes were created and used in different ways over time. In Chapter 8, Herrera describes and explains patterns of sedentism as well as sociopolitical and socioeconomic organization in southern Costa Rica from 300 bc to ad 1550. He summarizes a broad set of current information, ranging from environmental reconstruction to mortuary data to perceptions of landscape, and incorporates new information from the site of El Cholo. In an area assumed to have been occupied by hierarchically organized and sedentary communities, he provides compelling evidence for a settlement system involving semi-sedentary communities collectively contributing to monumental landscapes. The final section of the book includes research projects from South America that are pushing our studies of settlement location choice into increasing levels of detail and complexity. In Chapter 9, Kellett takes a novel approach to settlement ecology by using it as a means to explore the impact of risk on past human behaviors among the Chanka of Peru during the twelfth and thirteenth centuries. Using Maslow’s hierarchy of needs as a hypothesis, he tests whether Chanka communities chose settlement locations based on basic needs before sociopolitical needs. The preference for locations meeting the latter encourages all of us to incorporate social, political, economic, and ideological factors into our settlement ecology models. In Chapter 10, Van Gijseghem re-examines current models of rainfall patterns and water use, and their impact on settlement patterns in the Ica-Nasca region of Peru. He attempts to break down a common assumption that farming and settlement practices, both modern and recent, provide a window into the deep past. In removing his work from this framework, and by incorporating new survey data and recent paleo-climatic and environmental data, he constructs a new model for the spatial organization of agriculture and settlement in the region. His model has broader implications for persisting population-pressure models for sociopolitical development in the region and interesting results pertaining to the politics of water distribution and use. Finally, in Chapter 11, Almeida describes the variability in settlement forms and settlement location choices on several scales, from the household to the micro-regional, in several areas of the Amazon basin. By using a combination of ethnographic, ethnohistoric, and archaeological data, he is able to provide ample evidence for a variety of factors that influenced where people chose to live and in what types of households or communities. These factors range from agricultural to political to ideological. The result is a rich description of the settlement pattern and behavior variability in this region that is a thoughtful complement to many of the quantitative and ecologically based chapters in this volume.

References Adams, Robert McC. 1965 Land Behind Baghdad: A History of Settlements on the Diyala Plain. Chicago, University of Chicago Press. Adams, Robert McC. 1981 Heartland of Cities: Surveys of Ancient Settlement and Land Use on the Central Floodplain of the Euphrates. Chicago, University of Chicago Press. Allen, Kathleen M.S. 1996 Iroquoian Landscapes: People, Environments, and the GIS Context. In New Methods, Old

Problems: Geographic Information Systems in Modern Archaeological Research, edited by H.G. Maschner, pp. 198– 222. Carbondale, Southern Illinois University, Center for Archaeological Investigations. Arkush, Elizabeth 2011 Hillforts of the Ancient Andes: Colla Warfare, Society, and Landscape. Gainesville, University of Florida Press. Ashmore, W., and A.B. Knapp 1999 Archaeologies of Landscape. In Continuities and Changes in Maya Archaeology: Perspectives at the Millennium, edited by C. Golden and G. Borgstede, p. 88. London and New York, Routledge. Balée, William (editor) 1998 Advances in Historical Ecology. New York, Columbia University Press. Balée, William (editor) 2006 The Research Program of Historical Ecology. Annual Review of Anthropology 35(5): 15–24. Balée, William and Clark Erickson (editors) 2006 Time and Complexity in Historical Ecology: Studies in the Neotropical Lowlands. New York, Columbia University Press. Bauer, Brian S. and Lucas C. Kellett 2010 Cultural Transformations of the Chanka Homeland (Andahuaylas, Peru) During the Late Intermediate Period (ad 1000–1400). Latin American Antiquity 21(1): 87–111. Bell, Tyler and Gary Lockbauer 2000 Topographic and Cultural Influences on Walking the Ridgeway in Later Prehistoric Times. In Beyond the Map, edited by G. Lock, pp. 85–100. London, IOS Press. Berman, Marc 1994 Lukurmata: Household Archaeology in Prehispanic Bolivia. Princeton, NJ: Princeton University Press. Bevan, Andrew and James Conolly 2002 GIS, Archaeological Survey, and Landscape Archaeology on the Island of Kythera, Greece. Journal of Field Archaeology 29(1/2); Spring 2002–Summer 2004: 123–138. Billman, Brian R. 1999 Settlement Pattern Research in the Americas: Past, Present, and Future. In Settlement Patterns in the Americas, 50 Years since Virú, edited by Brian R. Billman and Gary M. Feinman, pp. 1–5. Washington, DC, Smithsonian Institution Press. Billman, Brian R. and Gary M. Feinman (editors) 1999 Settlement Patterns in the Americas, 50 Years since Virú. Washington, DC, Smithsonian Institution Press. Binford, Lewis 1968 Post-Pleistocene Adaptations. In New Perspectives in Archaeology, edited by Sally Binford and Lewis R. Binford, pp. 313–341. New York, Aldine. Binford, Lewis 1980 Willow Smoke and Dogs’ Tails: Hunter-Gatherer Settlement Systems and Archaeological Site Formation. American Antiquity 45: 4–20. Bird, Douglas W. and James F. O’Connell 2006 Behavioral Ecology and Archaeology. Journal of Archaeological Research 14(2): 142–188. Blanton, Richard E. 1978 Monte Alban: Settlement Patterns at the Ancient Zapotec Capital. New York, Academic Press. Blanton, Richard E. (editor) 2005 Settlement, Subsistence, and Complexity: Essays Honoring the Legacy of Jeffrey R. Parsons. Los Angeles, Cotsen Institute of Archaeology, University of California. Blanton, Richard E., Mary H. Parsons, Luis Morett Alatorre, and Carla M. Sinopoli 2005 Introduction. In Settlement Subsistence, and Complexity: Essays Honoring the Legacy of Jeffrey R. Parsons, edited by Richard E. Blanton, pp. 1– 18. Los Angeles, Cotsen Institute of Archaeology, University of California. Borgstede, Greg and James R. Mathieu 2007 Defensibility and Settlement Patterns in the Guatemalan Maya Highlands. Latin American Antiquity 18(2): 191–211. Brouwer Burg, Marieka 2013 Reconstructing “Total” Paleo-Landscapes for Archaeological Investigation: An Example from the Central Netherlands. Journal of Archaeological Science 40: 2308–2320. Brown, James A., Robert E. Bell, and Don G. Wyckoff 1978 Caddoan Settlement Patterns in the Arkansas River Drainage. In Mississippian Settlement Patterns, edited by Bruce D. Smith, pp. 169–200. New York, Academic Press. Bruno, David and Julian Thomas (editors) 2010 Handbook of Landscape Archaeology (World Archaeological Congress Research). Walnut, CA, Left Coast Press. Carballo, David M. and Thomas Pluckhan 2007 Transportation Corridors and Political Evolution in Highland Mesoamerica: Settlement Analyses Incorporating GIS for Northern Tlaxcala, Mexico. Journal of Anthropological Archaeology 26: 607– 629. Chang, K.C. 1972 Settlement Patterns in Archaeology. Reading, MA, Addison-Wesley. Chisolm, M. 1968 Rural Settlement and Land Use. 2nd edition. London, Hutchinson. Christaller, W. 1966 Central Places in Southern Germany. Translated by C.W. Baskin. Englewood Cliffs, NJ, Prentice-Hall. Creese, John L. 2009 Post Molds and Preconceptions: New Observations about Iroquoian Longhouse Architecture. Northeast Anthropology 77/78: 47–69. Creese, John L. 2012 The Domestication of Personhood: A View from the Northern Iroquoian Longhouse. Cambridge Archaeological Journal 22(3): 365–386. Crumley, Carole L. 1979 Three Locational Models: An Epistemological Assessment for Anthropology and Archaeology. In Advances in Archaeological Method and Theory, Vol. 3, edited by Michael Schiffer, pp. 141–173. New York, Academic Press. Crumley, Carole L. (editor) 1994 Historical Ecology: Cultural Knowledge and Changing Landscapes. Santa Fe, School of American Research. Crumley, Carol L. and William H. Marquardt 1987 Regional Dynamics: Burgundian Landscapes in Historical Perspective.

London, Academic Press. Crumley, Carole L. and William H. Marquardt 1990 Landscape: A Unifying Concept in Regional Analysis. In Interpreting Space: GIS and Archaeology, edited by Kathleen M. Allen, Stanton W. Green, and Ezra B. W. Zubrow, pp. 73–79. London, Taylor and Francis. Ebert, J.I. 1992 Distributional Archaeology. Albuquerque, University of New Mexico Press. Elliott, Michelle 2005 Evaluating Evidence for Warfare and Environmental Stress in Settlement Pattern Data from the Malpaso Valley, Zacatecas, Mexico. Journal of Anthropological Archaeology 24: 297–315. Evans, Susan T. and Peter Gould 1982 Settlement Models in Archaeology. Journal of Anthropological Archaeology 1: 275– 304. Fish, Suzanne K. 1999 The Settlement Pattern Concept from an Americanist Perspective. In Settlement Patterns in the Americas, 50 Years since Virú, edited by Brian R. Billman and Gary M. Feinman, pp. 203–208. Washington, DC, Smithsonian Institution Press. Flannery, Kent V. 1968 Archeological Systems Theory and Early Mesoamerica. In Anthropological Archeology in the Americas, edited by B.J. Meggers, pp. 67–87. Washington, DC, Anthropological Society of Washington. Flannery, Kent V. (editor) 1976 The Early Mesoamerican Village. New York, Academic Press. French, Kirk D., Christopher J. Duffy, and Gopal Bhatt 2012 The Hydroarchaeological Method: A Case Study at the Maya Site of Palenque. Latin American Antiquity 23(1): 29–50. Gillings, Mark and G.T. Goodrick 1996 Sensuous and Reflexive GIS: Exploring Visualization and VRML. Internet Archaeology 1. Gillings, Mark, D. Mattingly, and J. van Dalen (editors) 1999 Geographical Information Systems and Landscape Archaeology. Oxford, The Alden Press. Grossman, David 1971 Do We Have a Theory for Settlement Geography?—The Case of Iboland. Professional Geographer 3: 197–203. Gumerman, George J. 1971 The Distribution of Prehistoric Population Aggregation. Prescott, AZ, Prescott College Press. Haas, Jonathan and Winifred Creamer 1993 Stress and Warfare Among the Kayenta Anasazi of the Thirteenth Century ad. Fieldiana Anthropology, New Series, No. 21. Chicago, Field Museum of Natural History. Hagget, P. 1965 Locational Analysis in Human Geography. London, Edward Arnold. Harris, Trevor and Gary Lock 1995 Toward an Evaluation of GIS in European Archaeology: The Past, Present, and Future of Theory and Applications. In Archaeology and Geographical Information Systems: A European Perspective, edited by G. Lock and Z. Stančič, pp. 349–366. London, Taylor and Francis. Hasenstab, Robert J. 1996 Settlement as Adaptation: Variability in Iroquois Village Site Selection as Inferred Through GIS. In New Methods, Old Problems: Geographic Information Systems in Modern Archaeological Research, edited by H. Maschner, pp. 223–241. Carbondale, Southern Illinois University, Center for Archaeological Investigations. Higgs, E.S. and C. Vita-Finzi 1972 Prehistoric Economies: A Territorial Approach. In Papers in Economic Prehistory, edited by E.S. Higgs, pp. 27–36. Cambridge, Cambridge University. Higgs, E.S., C. Vita-Finzi, D.R. Hariss, and A.E. Fagg 1967 The Climate, Environment and Industries of Stone Age Greece, Part 3. Proceedings of the Prehistoric Society 331: 1–29. Hodder, Ian and Clive Orton 1976 Spatial Analysis in Archaeology. New Studies in Archaeology. London, Cambridge University Press. Howey, Meghan C.L. 2007 Using Multi-Criteria Cost Surface Analysis to Explore Past Regional Landscapes: A Case Study of Ritual Activity and Social Interaction in Michigan, ad 1200–1600. Journal of Archaeological Science 34: 1830–1846. Howey, Meghan C.L. 2011 Multiple Pathways Across Past Landscapes: Circuit Theory as a Complementary Geospatial Method to Least Cost Path for Modeling Past Movement. Journal of Archaeological Science 38: 2523–2535. Hu, Di 2011 Advancing Theory? Landscape Archaeology and Geographical Information Systems. Papers from the Institute of Archaeology 21: 80–90. Ingold, Tim 1993 The Temporality of the Landscape. World Archaeology 25(2): 152–174. Johnson, G.A. 1977 Aspects of Regional Analysis in Archaeology. Annual Review of Anthropology 6: 479–508. Jones, Eric E. 2006 Using Viewshed Analysis to Explore Settlement Choice: A Case Study of the Onondaga Iroquois. American Antiquity 71(3): 523–538. Jones, Eric E. 2010 An Analysis of Factors Influencing Sixteenth and Seventeenth Century Haudenosaunee (Iroquois) Settlement Locations. Journal of Anthropological Archaeology 29: 1–14. Jones, Eric E. and Peter Ellis 2016 Multiscalar Settlement Ecology Study of Piedmont Village Tradition Communities in North Carolina, ad 1000–1600. Southeastern Archaeology. 35(2): 85–114. Jones, Eric E. and James W. Wood 2012 Using Event-History Analysis to Examine the Causes of Semi-Sedentism Among Shifting Cultivators: A Case Study of the Haudenosaunee, ad 1500–1700. Journal of Archaeological Science 39: 2593– 2603. Jones, Eric E., Madison Gattis, Andrew Wardner, Thomas C. Morrison, and Sara Frantz 2012 Exploring Prehistoric Tribal Settlement Ecology in the Southeast: A Case Study from the North Carolina Piedmont. North American Archaeologist

33(2): 157–190. Kantner, John 2008 The Archaeology of Regions: From Discrete Analytical Toolkit to Ubiquitous Spatial Perspective. Journal of Archaeological Research 16(1): 37–81. Kellett, Lucas C. 2010 Chanka Settlement Ecology: Hilltop Sites, Land Use and Warfare in Late Prehispanic Peru. Unpublished Ph.D. dissertation. Albuquerque, University of New Mexico. Kelly, Robert 1985 Hunter-Gatherer Mobility Strategies. Journal of Anthropological Research 39: 277–306. Kohler, Timothy A. 1988 Predictive Locational Modeling: History and Current Practice. In Quantifying the Present and Predicting the Past: Theory, Method and Application of Archaeological Predictive Modeling, edited by James W. Judge and Lynn Sebastian, pp. 19–59. Denver, US Department of Interior, Bureau of Land Management. Kohler, Timothy A. and G. Gumerman 2000 Dynamics in Human and Primate Societies: Agent-Based Modeling of Social and Spatial Presses. Santa Fe Institute Studies in the Sciences of Complexity. New York, Oxford University Press. Kosiba, Steven and Andrew Bauer 2013 Mapping the Political Landscape: Toward a GIS Analysis of Environmental and Social Difference. Journal of Archaeological Method and Theory 20: 61–101. Kowalewski, Stephen A. 2008 Regional Settlement Patterns. Journal of Archaeological Research 16(3): 225–285. Kvamme, Kenneth L. 1990 One-Sample Tests in Regional Archaeological Analysis: New Possibilities through Computer Technology. American Antiquity 55(2): 367–381. Kvamme, Kenneth L. 1996 Investigating Chipping Debris Scatters: GIS as an Analytical Engine. In New Methods, Old Problems: Geographic Information Systems in Modern Archaeological Research, edited by H. Maschner, pp. 38–71. Carbondale, Southern Illinois University, Center for Archaeological Investigations. Kvamme, Kenneth L. 1999 Recent Directions and Developments in Geographical Information Systems. Journal of Archaeological Research 7(2): 153–201. Lekson, Steve 1999 The Chaco Meridian: Centers of Political Power in the Ancient Southwest. Walnut Creek, CA, Altamira. Llobera, Marcos 1996 Exploring the Topography of Mind: GIS, Social Space, and Archaeology. Antiquity 70: 612–622. Llobera, Marcos 2000 Understanding Movement: A Pilot towards the Sociology of Movement. In Beyond the Map, edited by G. Lock, pp. 65–84. Amsterdam, IOS Press. Llobera, Marcos 2001 Building Past Landscape Perception with GIS: Understanding Topographic Prominence. Journal of Archaeological Science 28: 1005–1014. Llobera, Marcos 2003 Extending GIS Based Analysis: The Concept of Visualscape. International Journal of Geographic Information Science 17(1): 25–49. Llobera, Marcos 2007 Reconstructing Visual Landscapes. World Archaeology 39(1): 51–69. Marquardt, William H. and Carol L. Crumley 1987 Theoretical Issue in the Analysis of Spatial Patterning. In Regional Dynamics: Burgundian Landscapes in Historical Perspective, edited by C.L. Crumley and W. H. Marquardt, pp. 1–18. San Diego, Academic Press. Maschner, Herbert D.G. 1996a The Politics of Settlement Choice on the Northwest Coast: Cognition, GIS, and Coastal Landscapes. in Anthropology, Space and Geographic Information Systems, edited by Herbert D.G. Maschner and M. Aldenderfer. London, Oxford University Press. Maschner, Herbert D.G. (editor) 1996b New Methods, Old Problems: Geographic Information Systems in Modern Archaeological Research. Carbondale, Southern Illinois University, Center for Archaeological Investigations. Maschner, Herbert D.G. and M. Aldenderfer (editors) 1996 Anthropology, Space and Geographic Information Systems. London, Oxford University Press. Maschner, Herbert D. G. and Jeffrey W. Stein 1995 Multivariate Approaches to Site Location on the Northwest Coast of North America. Antiquity 69: 61–73. Moore, Jerry D. 2012 The Prehistory of Home. Berkeley, CA, University of California Press. Netting, Robert McC. 1993 Smallholders, Householders: Farm Families and the Ecology of Intensive Sustainable Agriculture. Stanford, CA, Stanford University Press. Parsons, J.R. 1971 Prehistoric Settlement Patterns in the Texcoco Region, Mexico. Memoirs of the Museum of Anthropology No. 3. Ann Arbor, University of Michigan. Parsons, J.R. 1972 Archaeological Settlement Patterns. Annual Review of Anthropology 1: 127–150. Parsons, Jeffrey R., Charles M. Hastings, and Ramiro Matos M. 2000 Prehispanic Settlement Patterns in the Upper Mantaro and Tarama Drainages, Junín, Peru. Volume 1, Part 1, The Tarama-Chinchaycocha Region. Memoirs of the Museum of Anthropology No. 34. Ann Arbor, University of Michigan. Parsons, Jeffrey R., Charles M. Hastings, and Ramiro Matos M. 2013 Prehispanic Settlement Patterns in the Upper Mantaro and Tarama Drainages, Junín, Peru. Volume 2, The Wanka Region. Memoirs of the Museum of Anthropology No. 53. Ann Arbor, University of Michigan. Peebles, Christopher 1978 Determinants of Settlement Size and Location in the Moundville Phase. In Mississippian Settlement Patterns, edited by Bruce D. Smith, pp. 369–416. New York, Academic Press. Peterson, C.E. and R.D. Drennan 2005 Communities, Settlement, Sites and Surveys: Regional-Scale Analysis of Prehistoric

Human Interaction. American Antiquity 70: 5–30. Plog, Fred 1975 Systems Theory in Archaeological Research. Annual Review of Anthropology 1975(4): 207–224. Plog, F.T. and J.N. Hill 1971 Explaining variability in the Distribution of Sites. In The Distribution of Prehistoric Population Aggregates, Anthropological Reports, No.1, edited by G.J. Gumerman, pp. 7–36. Prescott, AZ, Prescott College Press. Rossignol, J. and L. Wandsnider 1992 Time, Space and Archaeological Landscapes. New York, Plenum Press. Sakaguchi, Takashi, Jesse Morin, and Ryan Dickie 2010 Defensibility of Large Prehistoric Sites in the Mid-Frasier Region on the Canadian Plateau. Journal of Archaeological Science 37: 1171–1185. Sanders, W.T. 1965 The Cultural Ecology of the Teotihuacán Valley. Mimeograph. University Park, Pennsylvania State University, Department of Anthropology. Sanders, W.T. 1981 Ecological Adaptation in the Basin of Mexico: 23,000 bc to the Present. In Archaeology:Supplement to the Handbook of Middle American Indians, edited by J.A. Sabloff, pp. 147–197. Austin, TX, University of Texas Press. Sanders, W.T., J.R. Parsons, and R.S. Santley 1979 The Basin of Mexico: Ecological Processes in the Evolution of a Civilization, 2 vols. New York, Academic Press. Schiffer, M.B. 1987 Formation Processes of the Archaeological Record. Albuquerque, University of New Mexico Press. Sipkins, Penny 2000 GIS Models of Past Vegetation: An Example from Northern England, 10,000–5000 bp. Journal of Archaeological Science 27: 219–234. Smith, E.A. and B. Winterhalder (editors) 1992 Evolutionary Ecology and Human Behavior. Hawthorne, NY, Aldine de Grutyer. Steponaitis, Vincas P. 1978 Locational Theory and Complex Chiefdoms: A Mississippian Example. In Mississippian Settlement Patterns, edited by Bruce D. Smith, pp. 417–453. New York, Academic Press. Steponaitis, Vincas P. 1981 Settlement Hierarchies and Political Complexities in Nonmarket Societies: The Formative Period of the Valley of Mexico. American Anthropologist 83(2): 320–363. Stone, Glenn Davis 1996 Settlement Ecology: The Social and Spatial Organization of Kofyar Agriculture. Arizona Studies in Human Ecology. Tucson, University of Arizona Press. Struever, Stuart 1968 Woodland Subsistence-Settlement Patterns in the Lower Illinois Valley. In New Perspectives in Archaeology, edited by Sally Binford and Lewis R. Binford, pp. 285–312. New York, Aldine. Thomas, David Hurst 1972 Western Shoshone Ecology: Settlement Patterns and Beyond. In Great Basin Cultural Ecology: A Symposium, edited by D.D. Fowler, pp. 87–94. Cambridge, Cambridge University Press. Thomas, David Hurst and Robert L. Bettinger 1976 Prehistoric Piñon Ecotone Settlements of the Upper Reese River Valley, Central Nevada, Anthropological Papers, Vol. 53, No. 3. New York, American Museum of Natural History. Thomas, Julian 1991 Understanding the Neolithic. London, Routledge Press. Tilley, Christopher 1994 A Phenomenology of Landscape: Places, Paths and Monuments. Oxford, Berg. Tilley, C. and Bennett, W. 2001 An Archaeology of Supernatural Places: The Case of West Penwith. Journal of the Royal Anthropological Institute 7(2): 355–362. Trigger, Bruce G. 1967 Settlement Archaeology. Its Goals and Promise. American Antiquity 32(2): 149–160. Trigger, Bruce G. 1968 The Determinants of Settlement Patterns. In Settlement Archaeology, edited by K.C. Chang, pp. 53– 78. Palo Alto, National Press. Vita-Finzi, C. and E.S. Higgs 1970 Prehistoric Economy in the Mt. Carmel Area of Palestine: Site Catchment Analysis. Proceedings of the Prehistoric Society 36: 1–37. von Thünen, Johann Heinrich 1966 Von Thünen’s Isolated State. Translated by Carla M. Wartenburg, edited by P. Hall. London, Permagon Press. Originally published in 1826. Wagstaff, J.M. (editor) 1987 Landscape and Culture: Geographical and Archaeological Perspectives. London, Blackwell. Warren, R.E. 1990a Predictive Modeling of Archaeological Site Location: A Case Study in the Midwest. In Interpreting Space: GIS and Archaeology, edited by K.M.S. Allen, S.W. Green, and E.B.W. Zubrow, pp. 201–215. London, Taylor and Francis. Warren, R.E. 1990b Predictive Modeling in Archaeology: A Primer. In Interpreting Space: GIS and Archaeology, edited by K.M.S. Allen, S.W. Green, and E.B.W. Zubrow, pp. 91–111. London, Taylor and Francis. Wernke, Steven A. 2013 Negotiated Settlements: Andean Communities and Landscapes under Inka and Spanish Colonialism. Gaineseville, University of Florida Press. Wescott, K.L. and Brandon, R. J. (editors) 2000 Practical Applications of GIS for Archaeologists: A Predictive Modelling Toolkit. London, Taylor and Francis/CRC Press. Wheatley, David 1993 Going over Old Ground: GIS, Archaeological Theory, and the Act of Perception. In Computing the Past: Computer Applications and Quantitative Methods in Archaeology, edited by J. Andresen, T. Madsen, and I. Scollar, pp. 133–137. Aarhus, Aarhus University Press. Wheatley, David 1995 Cumulative Viewshed Analysis: A GIS-based Method for Investigating Intervisibility, and its Archaeological Application. In Archaeology and Geographic Information Systems: A European Perspective, edited by G. Lock, and Z. Stančič. London, Taylor and Francis. Wheatley, David and Mark Gillings 2000 Vision, Perception and GIS: Developing Enriched Approaches to the Study of

Archaeological Visibility. In Beyond the Map: Archaeology and Spatial Technologies, edited by G. Lock. Amsterdam, IOS Press. Wheatley, David and Mark Gillings 2002 Spatial Technology and Archaeology: The Archaeological Applications of GIS. New York, Taylor and Francis. Willey, Gordon 1953 Prehistoric Settlement Patterns in the Virú Valley, Peru. Burea of American Ethnology Bulletin No. 155. Washington DC. Willey, Gordon 1956 Prehistoric Settlement Patterns in the New World. Viking Fund Publication in Anthropology No. 23. New York, Wenner-Gren Foundation. Wurzer, Gabriel, Kerstin Kowarik, and Hans Reschreiter (editors) 2015 Agent-Based Modeling and Simulation in Archaeology. Advances in Geographic Information Science. Berlin, Springer-Verlag.

Part II

North America

2 The ecology of changing settlement patterns among Piedmont Village Tradition communities in southeastern North America, AD 800–1600 Eric E. Jones This research uses a combination of traditional settlement pattern and more recent settlement ecology methods to create and explain a model of settlement change among Piedmont Village Tradition communities in the Southeast region of North America during AD 800–1600. Settlement ecology studies—those that attempt to explain settlement patterns through quantitative methods or ethnographic/ethnohistoric analogy—are not as prevalent in the interior Southeast as in other regions of North America (e.g. Haas and Creamer 1993; Allen 1996; Hasenstab 1996; Maschner 1996; Jones 2006, 2010; Sakaguchi et al. 2010). Settlement pattern research—that which focuses on descriptions of settlement patterns—on the other hand, has been as popular in this region as in any other, particularly studies of sociopolitically complex Mississippian societies (e.g. Dickens 1978; Anderson 1994; Jefferies et al. 1996; Pluckhahn and McKivergan 2002). Starting in the 1970s, studies of the settlement patterns of Piedmont Village Tradition societies established them as a viable archaeological culture and produced a wealth of information (Simpkins 1985; Dickens et al. 1987; Woodall 1990; Davis and Ward 1991; Ward and Davis 1993). Some of these studies, particularly Simpkins (1985), began the process of moving beyond describing patterns to explaining them, setting the stage for empirical testing aimed at generating evidence-based explanations for settlement behaviors. To date, settlement ecology research in the Southeast has largely examined regional scales and has not explored change over time (see Jones et al. 2012; Jones and Ellis 2016; Jones 2015). In fairness, most settlement ecology research in North America in general has remained synchronic and on a regional level, my own work included. The research I describe in this chapter is an attempt to address these current limitations. The Piedmont Village Tradition (PVT) is the archaeological remains of non-Mississippian, non-hierarchical societies that resided in the Piedmont regions of the current states of Virginia and North Carolina from 1000 BC to AD 1600. Studies of PVT communities have yielded important information on several broadly applicable topics, including population and demography (Ward and Davis 2001), diet and subsistence amongst mixed subsistence agriculturalists (VanDerwarker et al. 2007), adoption of maize agriculture in the Southeast (Woodall 1990; Ward and Davis 1993), interaction across ethnic and sociopolitical boundaries (Woodall 1999, 2009; Jones and Ellis 2016), and colonial interactions between Native Americans and Europeans (Ward and Davis 1991). Settlement dynamics (Simpkins 1985; Dickens et al. 1987; Davis and Ward 1991; Woodall 1990; Jones et al. 2012) have been at the forefront of most PVT research and have yielded important information on trends, particularly during the Middle Woodland, Late Woodland, and Early Colonial Periods (~AD 1–1700).

Specifically, it appears that after AD 800, populations grew in most areas of the Piedmont, and smaller communities coalesced and began living in larger planned villages (Simpkins 1985; Davis and Ward 1991), mirroring trends seen in the Mid-Atlantic and Northeast regions (Hart 1993; Snow 1995; Birch and Williamson 2013). However, PVT communities in the upper Yadkin River Valley (UYRV) show a different trend. Woodall (1990:83, 92) proposed that communities there reacted to population growth after AD 800 by dispersing into a greater number of smaller communities, to place less strain on local resources. The first step in the research described in this chapter is a testing of Woodall’s model through an examination of the surface artifact scatter sizes at settlement sites in the UYRV, similar to research done in other areas of the Piedmont (Simpkins 1985). Recent site mapping (Jones et al. 2012) provides data from additional sites and new data on older sites that were not available to Woodall. The results support Woodall’s hypothesis but also suggest a more complex picture of demographic and settlement strategy change, particularly after AD 1200. They indicate three possible trends: 1) a possible decline in population; 2) a reduction in community size; or 3) a decrease in sedentism after this date. Any of the possible interpretations show that the UYRV communities experienced different demographic and settlement pattern trajectories than the other PVT areas, and eastern North America in general. After constructing this new settlement model, I then compare the settlement ecology of communities before and after this change. The goal is to move beyond a description of this shift in settlement strategy to explaining it through an examination of how communities interacted with their natural and cultural landscapes. I employ geographic information systems (GIS) to estimate and quantify the characteristics of past landscapes and to collect landscape information related to settlement site locations. I use discriminant function analysis (DFA) to determine which landscape features spatially correlate with settlement locations, approximating the factors that influenced settlement location choices. From these results, I explore why demographic and settlement patterns appear to be very different from those in other areas of the Piedmont at this time. This is a worthwhile case study for several reasons. First, it appears that these PVT communities were undergoing very different demographic and ecological processes than other societies and communities across eastern North America. Second, non-Mississippian societies are an understudied group in the Southeast; however, they played an important role in social (Woodall 1999, 2009) and economic interactions (Thomas 1996) within the macroregion but are often left out of these discussions. Third, PVT communities in the upper Yadkin River Valley are of particular importance because they resided on the edge of the Mississippian world (Jones et al. 2012; Jones and Ellis 2016). The Mississippian emergence in conjunction with the formation of sociopolitically complex societies was one of the most important cultural developments in eastern North America, shaping the landscape for over 500 years. Case studies like this can tell us a significant amount about what it was like for people on the edge of the Mississippian world and how it impacted their interactions with their surrounding natural and cultural landscapes and other communities and societies.


Settlement ecology theory and methods Like many works of archaeological settlement ecology, this one focuses on settlement location choice. Settlement ecology, as presented by Stone (1996), is an approach that strives to reconstruct total settlement systems, including patterns and process. The patterns are often much easier to locate and interpret in cases where only archaeological data are available, which is why location choice is a popular topic. Other processes have been studied, like semisedentism (Jones and Wood 2012), population growth and expansion (Brannan and Birch, this volume), and abandonment (Ingram, this volume), but they often require more than archaeological data either through ethnohistoric information, oral histories, or statistical modeling. I employ a behavioral ecological assumption that spatial correlations are telling us something about cultural preferences or values. Basically, it proposes that our settlement behaviors are not random. If people are settling close to a resource or landscape feature, it is because they value it. Thus, if we can establish spatial correlations using GIS-based or other statistical methods, we have a basis for examining spatial behaviors related to how people interacted with their natural and cultural environments and landscapes. Combined with the proximity principle, as described in the introductory chapter, these two assumptions allow us to rank or relatively examine the influence of various factors based on their proximities to settlements. This is important because Stone (1996) described the process of making any settlement decision, particularly where to live, as a mental balance sheet. Decision-makers consider all of the needs of the community and place the settlement in a location that takes advantage of the most critical resources or locations. No place is perfect. Thus, communities must make compromising decisions, including sacrifices. Maximizing proximity to one resource may mean sacrificing another. The best methods for settlement location choice will attempt to analyze all of the factors that past people took into account and use statistical methods that approximate the mental balance sheet. Studies of settlement location choice in North America have been steadily expanding over the past 20 years (e.g. Kvamme 1990; Allen 1996; Hasenstab 1996; Maschner 1996; Jones 2006, 2010; Sakaguchi et al. 2010; Jones et al. 2012; Jones and Ellis 2016). It is no coincidence that these types of questions increased in popularity with the advent of GIS in archaeology. As outlined in the introduction, the methods existed before, but GIS makes compiling and analyzing large spatial datasets much easier and much more accessible. Most of the studies listed above employ GIS to display past settlement distributions on digitally estimated or reconstructed past landscapes. These representations allow for the measurement of various characteristics of the individual settlements and the surrounding landscapes, again based on the assumption that people are somewhat rational actors settling near valued places and resources. As also mentioned in the introduction, in such studies we must also establish that any observed patterns are not the result of spatial autocorrelation. There are several ways to do this. Many of the examples above have used variations on multiple regression analyses, like discriminant function analysis (DFA) (Hasenstab 1996; Jones 2006, 2010; Jones et al. 2012; Jones and Ellis 2016). By comparing a sample with randomly placed points, a researcher can both establish whether the sample is different from a random distribution (i.e. not autocorrelated) and obtain a ranking of variables that approximates the mental balance

sheet. A secondary check, such as the ability of many statistical programs to comingle the cases and then predict their group membership, helps to firmly establish actual correlations from autocorrelations. History of research in the upper Yadkin River Valley In eastern North America, the traditional chronological designations after 1000 BC (i.e. the end of the Archaic Period) for areas without Mississippian societies are Early Woodland (1000– 200 BC), Middle Woodland (200 BC–AD 800), and Late Woodland (AD 800–1600), collectively referred to as the Woodland Period. A large amount of research in the Piedmont Southeast has focused on this period, so we have a sizeable sample of sites and material culture. Surveys began in the upper Yadkin River Valley in the 1800s. Cyrus Thomas (1887) examined several mound sites at the extreme upstream end of the valley, but surveys of the sections of the valley occupied historically by PVT communities did not commence until the 1940s. Professional archaeological research began with Ned Woodall’s work in the 1970s. He first organized the Lower Great Bend Survey, examining the Yadkin River from where it begins running north– south to approximately 20km downstream (Barnette 1978). That survey established the presence of dozens of permanent Middle and Late Woodland settlements in that section of the valley. During the 1980s and 1990s, Woodall turned his attention toward the portions of the valley immediately upstream in the Upper Great Bend Survey (Woodall 1990). Like the previous work, this research identified dozens of sites through pedestrian surveys and shovel testing. Barnette (1978) compared assemblages from these newly found sites with existing ceramic and projectile-point chronologies and dated almost all of them to 200 BC–AD 1600. Eight sites, located throughout the valley, have been excavated (Figure 2.1), and I provide brief summaries of findings, which will be important for interpreting my results.

Figure 2.1

Map showing the Piedmont region, major river valleys, and PVT settlement locations within each.

The Forbush Creek site resides in the Lower Great Bend area and was excavated in the 1970s as a salvage project. Excavations recovered primarily Uwharrie phase pottery, dating to around AD 1000, with earlier Yadkin phase pottery, dating to 200 BC–AD 500 (Ward and Davis 1999:96). They found burials, storage and refuse pits, and hearths. Findings at the site look similar to Middle Woodland sites in the eastern Piedmont, including ossuary burials and thick mollusk shell deposits (Ward and Davis 1999:97). The overall size of the actual settlement area at the site is not known, but the number of burials suggests a significant population and the dates indicate a long occupation. A significant portion of the entire floodplain was surveyed producing a surface area estimate of 240,000m2. The Donnaha site also lies in the Lower Great Bend area and was the focus of several research projects by Woodall (1984). His surveys and excavations of the site defined a surface area of 72,000m2. Excavations defined a likely main settlement area covering approximately 29,000m2. This work produced five radiocarbon dates: cal 240 BC, 60 BC, AD 1040, AD 1140, AD 1250, and AD 1480 (Woodall 1984:19–21). The second-to-last date was taken from a mollusk shell and is likely skewed to a more recent date (Woodall 1990:21). The ceramic and projectile point styles do not match other sites dating to the fifteenth century, and the site has the characteristic Middle Woodland thick deposits of mollusk shells. This evidence suggests an occupation between 200 BC–AD 1200, similar to the range at Forbush Creek. At the time of excavation, it was unclear whether the site comprised multiple occupations or one large, longterm occupation. Woodall (1990:113) produced convincing evidence that it was the former with the identification of distinct and discontinuously overlapping cultural lenses. Mikell (1987) also showed through botanical analysis that Donnaha was occupied year-round, so the overlapping occupations are not seasonal aggregations. This floodplain was likely a place that PVT communities either moved within or returned to frequently over a 1000-to-1500-year period. Woodall (1990) excavated two smaller sites near Donnaha. The Hardy site is located at the far downstream end of the Upper Great Bend area, just upstream from the bend that divides this area from the Lower Great Bend area. Excavations during the 1970s and 1980s uncovered several pit features with faunal remains and a possible structure defined by an arc of postmolds (Woodall 1990). Radiocarbon dates provided an occupation range of AD 900–1300, but the eastern, central, and western portions all returned non-overlapping dates. As a result, Woodall (1990:53–54) believed the site was occupied three times, each one moving west along the floodplain. The surface artifacts covered an area of 43,000m2, but excavations indicate the site was much smaller, likely less than one-tenth that size. The MacPherson site is located approximately 4km south of Donnaha. Excavations there were limited but identified a 45cm-thick midden and recovered enough ceramics to relatively date the site to the first half of the Late Woodland (i.e. AD 800–1200), agreeing with Barnette’s (1978) relative dating of the site. The Parker site is at the southern end of the Lower Great Bend area and was excavated in the 1970s as part of a master’s thesis project (Newkirk 1978). The number of burials, dense midden deposits, and botanical and faunal remains strongly suggest that the site was a yearround settlement. It is possible that settlement activity was similar to that at Donnaha. Radiocarbon dating indicated an occupation from AD 600 to 1000 (Newkirk 1978: vii). An 80

× 60m area was excavated, but that was not enough to estimate an actual settlement area. The Redtail site is a small settlement site located in the Upper Great Bend section of the UYRV and is the current focus of my research. It was first recorded in 1990, and I surveyed and mapped the site in 2011 and 2012 (Jones et al. 2012). Recent radiocarbon dating of a potential living surface and midden area returned dates that suggest a single occupation during the late AD 1200s to early 1400s (Table 2.1). The overall surface scatter was approximately 10,500m2, but three seasons of excavations suggest a much smaller settlement area of approximately 900m2 that was likely occupied by one or two households. Woodall (1999, 2009) excavated two sites, Porter and T. Jones, in the far upstream end of the Upper Great Bend area in the 1990s. T. Jones dates to AD 1400–1600, and Porter to AD 1500-1600. Both sites appear to have been planned PVT settlements; T. Jones may have even had a fence separating the community from an adjacent swamp. Neither site has a surface area estimate, but excavations identified likely settlement sizes through the identification of pit features and postmolds. No intact cultural layers or middens were found. The settlement area (determined from excavation) of T. Jones is approximately 2500m2, and Porter is 1500m2. Both sites have evidence of interaction with Mississippian communities. Woodall identified a small number of burials at both sites with shaft tomb construction and objects with Southeastern Ceremonial Complex motifs. These burials looked distinctly different from typical PVT burials but similar to those at Mississippian sites (Woodall 1999, 2009). Excavations at Porter recovered ceramics and gaming pieces with Mississippian motifs, suggesting interaction with their Mississippian neighbors to the west (Woodall 1999). Excavations at the T. Jones site, the settlement farthest upriver and thus closest to Mississippian settlements, revealed two components. Both were PVT, but the later occupation had significant Mississippian cultural features, including the aforementioned burials (Woodall 2009). Table 2.1 Radiocarbon analysis results from the Redtail site

PVT culture history These sites and others throughout the Piedmont helped to define the Piedmont Village Tradition, a term popularized by Ward and Davis (1999:79) as a material-culture-based description of the inhabitants of the Piedmont Southeast in Virginia and North Carolina during the Woodland Period. Siouan was a term previously used but it has fallen somewhat out of fashion because of debates over the language spoken by historic Piedmont people (Mooney 1894; Sturtevant 1958; Goddard 2005). PVT is defined by small villages (fewer than a hundred individuals) composed of circular houses, with a mixed subsistence of agriculture and foraging, pottery-making, and settlement focused primarily in the floodplains of major rivers. The four major rivers of the Piedmont are the Yadkin, Dan, Haw, and Eno (shown in Figure 2.2). Extensive surveys in each valley have identified approximately a total of 125 settlement

sites (Barnette 1978; Simpkins 1985; Dickens et al. 1987; Woodall 1990; Ward and Davis 1993), providing a representative look at settlement patterns in those valleys. It should be noted that these four valleys have received the most archaeological attention, but that there are a small number of larger tributaries that have not been surveyed extensively. In the Dan, Eno, and Haw River Valleys, the transition from the Middle Woodland to the Late Woodland was marked by population increase and consolidation into planned villages (Davis and Ward 1991; Ward and Davis 1993). In each valley this consolidation tends to occur in centralized locations relative to previous settlement distributions. There are several studies of settlement change over time in these areas, but they focus mostly on the period AD 1400– 1700 (i.e. Simpkins 1985; Dickens et al. 1987; Davis and Ward 1991; Ward and Davis 1993). What we do know about the time period of AD 800–1600 is that agriculture was adopted but never became a dominant form of subsistence. Botanical remains at several sites show significant use of wild resources in conjunction with domesticated plants throughout this period (Ward and Davis 1993). Trade of non-local shell material was present (Thomas 1996). Warfare was common enough for many settlements to build palisades, particularly after AD 1400 (Dickens et al. 1987; Simpkins 1985; Ward and Davis 1993). In fact, it is proposed that the Eno River Valley was the site of a territorial struggle between PVT communities and Iroquoian-speaking ancestors of the Tuscarora to the immediate east (Simpkins 1985:85). In support of this idea, Eno settlements tend to show a preference for more defensible locations within the valley (Jones and Ellis 2016).

Figure 2.2

Map showing the location of the upper Yadkin River Valley sites used in this study. Sites specifically discussed are labeled.

The Yadkin River PVT appears to have undergone a different cultural trajectory. During the Late Woodland Period, there is no evidence of warfare, larger settlements, or coalescence. In fact, Woodall (1990) suggested an increase in the number of Late Woodland settlements and a

decrease in settlement size. His explanation for this pattern was that the introduction of maize into the agricultural system around AD 800 caused a population increase in the upper Yadkin River Valley that led to the establishment of smaller and dispersed villages. He believed the increase in the number of sites is almost certainly not a result of increased mobility (i.e. shorter occupation times) because they tend to have dense concentrations of artifacts and features. Thus, the current model, which I test in this work, is that UYRV populations rose, resulting in community fission and increased geographic spread throughout the valley. Rogers (1995), building on Woodall’s model, examined the variability in Late Woodland settlement size and proposed a heterarchical model of sociopolitical organization for PVT communities in the UYRV. Households were free to congregate into larger settlements or break down into smaller ones depending on the environmental and cultural climate at any particular time. Social and political organization was fluid and decision-making depended on community structure. When in smaller communities, households may have been almost completely independent, but when in larger communities, particular households and their heads may have had some authority in community-wide decisions. Rogers built this model assuming that most of the settlements throughout the UYRV were contemporaneous.

Methods Prior to my work here, researchers—including myself—have tended to study the late Middle and Late Woodland sites (~AD 1–1600) in the UYRV as a single group. In order to investigate whether we can examine changes over time, I compiled all of the known sites, their absolute or relative dates, and any information about the spatial extent of their surface or subsurface archaeological remains (Tables 2.2 and 2.3). The primary sources for these data were Barnette (1978), Newkirk (1978), Woodall (1984, 1990, 1999, 2009), Ward and Davis (1999), and Jones et al. (2012). For the analysis of settlement size, I used surface scatter areas recorded during previous surveys by Barnette (1978), Woodall (1990), and myself (Jones et al. 2012). These sizes are listed in Tables 2.2 and 2.3. Unfortunately, we do not have enough sites with both surface scatter size and intact subsurface settlement area to work out an equation for the relationship between the two. Surface scatter size is by no means a perfect indicator of settlement size in agricultural and floodplain contexts. For this reason, I examine settlement size in relative rather than absolute terms. The ability to compare the sites is based on the assumption that similar site formation processes have affected them over time. The major site formation processes in this case are: 1) burial by fluvial action; 2) erosion by fluvial action; 3) movement of artifacts by fluvial action; 4) agricultural plowing; and 5) survey methods. Factors 4 and 5 are consistent across the sites. Almost all of the sites have been agricultural fields since the early twentieth century, meaning plowing has impacted the lateral movement of artifacts similarly. All of the sites were surveyed using comparable methods (Woodall 1990; Jones et al. 2012). Based on notes from site files, the amount of ground cover in the fields appears to have been similar. I used differential GPS units to map the sites, but there is no increase in precision between them and using tape measures when measuring the extent of surface scatters.

Table 2.2 Sites with occupations during AD 800–1200 used in this study

Table 2.3 Sites with occupations during AD 1200–1600 used in this study

Factors 1 through 3 are not overly problematic in this case either. The sites are all in floodplains on the same river, meaning similar fluvial processes are at work. That being said, the river does flow through an area of bedrock change, going from harder to softer geologic materials. As a result, floodplains in the Lower Great Bend area tend to be larger than those in the Upper Great Bend area, signaling more deposition in this area of the valley. The increased river action in the Lower Great Bend area could have buried these sites more deeply, which would make them less likely to be brought to the surface by plowing. Thus, the sample of Lower Great Bend sites might be slightly underrepresented. Another consideration is the age of the sites and what that means for how deeply they are buried. In this case, I do not think the age of sites impacts the surface visibility, because the older sites like Donnaha and Forbush Creek tend to be easily located because of surface midden staining and dense artifact scatters. Finally, there is a precedent for using surface areas to examine settlement patterns and change over time. This approach has worked well in other Piedmont river valleys to help examine changes in settlement patterns and what they can tell us about sociopolitical organization (Simpkins 1985; Davis and Ward 1991). For the settlement ecology analysis, I created a shapefile of all of the sites within a GIS. In order to evaluate the significance of spatial correlations through DFA, I needed a set of randomly generated points with which I could compare the settlements. Within ArcMap 10.2, I created three sets of 30 randomly placed points. The area in which they could be located was restricted to floodplains throughout the entire UYRV, including both the Upper and Lower Great Bend areas. This is important because the entire valley seems to have been open for settlement throughout this period. Thus, we must compare the observed patterns with the entire set of possible areas to be settled. I next used the buffer tool to create 2km catchments around each settlement site. I chose this size for catchments based on archaeological, economic, and historic research that found it to be the maximum distance small-scale agriculturalists ranged to

work in fields or gather daily materials (Chisholm 1968; Fecteau et al. 1991:5). In addition, Lawson (1967:52) in the early eighteenth century describes Sapona Town on the lower Yadkin River as being in a clearing 1-mile (1.6km) square. To reconstruct the environment and landscape, I used DEMs and mid-twentieth-century hydrographic, wetland, and soil maps. I also digitized overland trails by combining Mouzon’s (1775) map and Myers’ (1971) historical research. Previous research showed these historic sources are quite similar to least cost paths modeled in GIS (Jones et al. 2012). I measured straight-line distances between settlements and features when examining proximity. The 15 variables measured for each site are shown in Table 2.4. Within SPSS, I then compared the pre-AD 1200 and post-AD 1200 settlement sets individually with the three sets of random points using discriminant function analysis. This analysis is statistically similar to multivariate regression and compares two datasets, indicates whether they are significantly different with regard to their various characteristics, and shows which of those characteristics most distinguish them if they are different, represented by a function value (Poulsen and French 2004; Sokal and Rohlf 1995:679–80). The larger a function value is, the more that variable distinguishes the two datasets. It also indicates the dataset with the higher average value of a particular variable using positive or negative values. In this work, a positive value indicates that the settlements had more of a particular feature within their catchment compared with the random points. This allowed me to determine not only if the PVT settlement locations and catchments were different but also which environmental or landscape features most distinguished them and how. Finally, to compare the landscapes of the Upper and Lower Great Bend areas, I examined all lands 2km away from the river on each side, from the bend to the site farthest away in each direction. I then used zonal statistics tools within GIS to measure several variables, shown in Table 2.5.

Results Tables 2.2 and 2.3 show the sites used in this study, surface and subsurface areas, dating methods, and sources. Based on pottery and projectile-point styles, most of the upper Yadkin River sites were originally dated to the Middle and Late Woodland, or 200 BC–AD 1600 (Woodall 1984, 1990, 1999, 2009; Jones et al. 2012). Barnette (1978) examined the thenknown Lower Great Bend sites relative to Donnaha’s ceramic assemblage and found that they all overlapped stylistically. Five had common characteristics with the earliest occupations at Donnaha, which equates roughly with 200 BC–AD 800. Three of those five also had similar styles to the later Donnaha occupations. This suggests most of the Lower Great Bend sites date between AD 800 and 1200, with a few sites, such as Forbush Creek and Donnaha, being occupied as early as 200 BC until AD 1200 (Woodall 1990; Ward and Davis 1999:96). It should be noted that these dates are based on my interpretation of the Donnaha radiocarbon dates, not Woodall’s. He tended to reject the earlier dates as outliers, whereas I include them and exclude the second-latest date, for reasons discussed above. Thus, my two main assumptions are: 1) that Donnaha was occupied primarily between 200 BC and AD 1200 and 2) that Barnette’s (1978) assessment that several sites have similar assemblages to Donnaha means

they overlap with those same dates. Table 2.4

Environmental and landscape variables measured for each settlement and random point


Landscape activity


Percent loam within catchment


Calculate percentage of 2km buffer covered by loam sediment types

Percent well-drained sediment in catchment


Calculate percentage of 2km buffer covered by well-drained sediment

Average solar radiation in catchment Agriculture/structure placement

Used slope from USGS 10 m DEM and solar radiation tool in ArcGIS; parameters set for annual radiation for year AD 1500

Average slope within catchment


Used slope tool (measured in %) in ArcGIS on USGS 10 m DEM and Zonal Statistics

Average aspect within catchment

Structure, settlement, and field placement

Used zonal statistics on USGS 10 m DEM

Area of good hardwood growth within catchment

Wood resources

Calculate m2 of sediments conducive to tree growth (as defined by NRCS) within 2km buffer

Area of good conifer growth within Wood resources catchment

Calculate m2 of sediments conducive to tree growth (as defined by NRCS) within 2km buffer

Wetlands within catchment


Count of wetlands of which any portion falls within 2km buffer

Largest wetland within catchment


Identification of the largest wetland within 2km buffer

Distance to tributary

Fresh water

Straight-line distance between site and nearest stream (as defined by NWI)

Slope at site

Structure placement

Used slope tool (measured in %) in ArcGIS at site point location on USGS 10m DEM

Aspect at site

Structure placement

Used aspect tool (measured as degree) in ArcGIS at site point location on USGS 10m DEM

Viewshed size

Intergroup relations

Used viewshed tool in ArcGIS using site location and USGS 10m DEM

Distance to overland trail

Intergroup relations

Measured the straight-line distance from sites to nearest trail (digitized from Mouzon 1775 and Myers 1971)

Percent uplands in catchment

Wild resources

Measured percentage of catchment that overlaps with upland environments (i.e. non-floodplain)

Table 2.5

Landscape characteristics measured for the Upper and Lower Great Bend areas

Landscape characteristic Percentage of total area covered by wetlands Average wetland size (m2) Wetland size standard deviation (m2) Percentage of total area covered by forested wetlands

Average forested wetland size (m2) Forested wetland size standard deviation (m2) Average slope Average aspect Average solar radiation Percentage of total area covered by loam Percentage of total area covered by good tree sediment Percentage of total area covered by well-drained sediment

Woodall (1990) used ceramic assemblages to relatively date the Upper Great Bend sites; most of them were dominated by late Dan River styles, suggesting dates of AD 1000–1500 (Woodall 2009:32). Since then, three sites—T. Jones, Porter, and Redtail—have been dated radiometrically. All three were occupied during periods of less than 200 years during AD 1300–1600. Thus, there appears to be a geographic shift in settlement upstream that occurred between AD 1000 and 1200. Given the radiocarbon dates of these three sites and Donnaha, I choose the later date as the dividing date. Based on the above criteria, I dated 23 settlement sites to AD 800–1200 (Table 2.1). The Lower Great Bend area survey identified 123 sites, 58 of which had pottery and thus could be dated broadly to the Woodland Period (200 BC–AD 1600). Thirty-seven of the 58 had 15 or fewer sherds and were thus deemed to be activity sites and not settlements. Of the remaining 21 sites, Barnette (1978) dated 18 to early, middle, and late Donnaha phases based on ceramic analysis, with only three dating solely to the early phase (before AD 800). Thus, 15 sites roughly correlated with AD 800–1200 (my assumed dates for the middle and late Donnaha phases). I did not remove the two early-phase-only sites mentioned above since they did overlap with both Donnaha and Forbush Creek assemblages. I added five more sites, 31Dv25, 31Sr50, 31Fy153, 31Wk155, and 31Yd132, which were surveyed and reported after Barnette’s work (Newkirk 1978; Woodall 1990). There are 19 settlement sites that date to AD 1200 or after (Table 2.3). Two sites came from Barnette’s (1978) analysis of the Lower Great Bend area. I put them in this period, disagreeing with Barnette’s categorization, because the site files described them as having ceramic assemblages with large concentrations of late Dan River style, indicating dates of AD 1000– 1500. The Upper Great Bend Survey (Woodall 1990) identified 43 sites. Eleven sites have either intact subsurface features or visible middens on the surface that suggest they were settlements and had primarily late Dan River style pottery. Woodall’s (1999, 2009) work at T. Jones and Porter dated them to this period. I have since added four more, 31Sr57, 31Sr58, 31Wk26, and 31Yd173, to this list during previous survey work (Jones et al. 2012). I also added two more sites, 31Wk27 and 31Yd24, based on descriptions from site files describing ceramic assemblages dominated by late Dan River styles. With regard to site-size changes over time, Table 2.2 shows the individual surface areas of the pre-AD 1200 settlements. The total surface area is 650,000m2. The average site size is 43,333m2. This, however, contains a rather large estimate for Forbush Creek, of which I am skeptical. The site is large and has a complex arrangement of overlapping occupations, leading

to an estimated site area of 240,000m2, which is more than three times larger than the next largest site. While this estimate seems unlikely when compared with other sites in the valley, it is also unreasonable to remove Forbush Creek as an outlier because we know from excavations that a settlement existed there (Ward and Davis 1999). My solution was to assign it an area of 80,000m2, assuming it is slightly larger than Donnaha. This brings the total surface area to 447,000m2, and the average site size to 29,800m2. In my analyses below, I discuss what it means if my assumptions about Forbush Creek or Donnaha here are wrong. As a precursor, if we remove Donnaha and Forbush Creek based on the possibility their size is more of a reflection of their Middle Woodland occupation, the overall and average site sizes are 338,000 m2 and 18,618 m2. Table 2.3 shows the individual surface areas of the post-AD 1200 settlements. The combined area is 159,500m2, with an average surface size of 14,500m2. With or without the smaller Forbush Creek size estimate, there is a considerable decrease in overall and average settlement site area based on surface site size. The post-AD 1200 settlements average less than half of the size of pre-AD 1200 sites. As previously mentioned, this is likely not a result of differential site formation processes. Actual settlement area results from excavations show that Donnaha was approximately 29,000m2 in size. The three post-AD 1200 sites—Redtail, Porter, and T. Jones—were 900, 1,500, and 2,500m2, respectively. If we look more carefully at the data, we see a complex relationship between time, site size, and number of sites. Plots of site size (Figure 2.3) versus time show that most sites, regardless of age, are less than 30,000m2. All of the sites larger than 30,000m2 are pre-AD 1200. Thus, regardless of date occupied, most sites were similar sized, suggesting similar settlement sizes —most likely a function of occupation duration—throughout the period of interest. However, the lack of larger sites after AD 1200 signifies a change in settlement strategy, which will be discussed more fully below using site-specific data. There is also a slight decrease in the number of sites after AD 1200; however, this small number could be a sampling error or a factor of time. Finally, as alluded to earlier, Figure 2.4 shows a pattern of pre-AD 1200 sites in the Lower Great Bend area and post-AD 1200 sites in the Upper Great Bend area, suggesting a trend in which communities moved upstream over time. The discriminant function analysis results for the pre-AD 1200 sites are shown in Table 2.6. My standard for influential factors on settlement location choice is that a factor must be a highranking variable with similar relationships to the random points (i.e. similar positive or negative value) across all three DFA models. I used natural breaks in the results to qualify variables as high, medium, low, and no influence. The highest-ranking variables for each comparison are shaded depending on their value. The values with shading across all three results and with the same positive or negative value across all three were the variables that showed the most difference between the actual settlements and random points. The more distinguishing the variable, the more that landscape feature or resource likely influenced settlement location decisions, based on the assumptions outlined above. Thus, the results show that pre-AD 1200 settlements sites reside in larger floodplains, are farther from trails, have more loam within their catchments, and contain more forested wetlands within their catchments. The results for the post-AD 1200 sites are shown in Table 2.7, and should be read similarly to Table 2.6. These show that these settlements have more loam in their catchments,

are farther from trails, have smaller wetlands in their catchments, have less slope at the settlement location, and have less wetland area within their catchments. The landscape characterization results (Table 2.8) show that the Lower Great Bend area has more wetland area and larger average wetland size, compared with the Upper Great Bend area. The standard deviation is much larger for the Lower Great Bend area as well. The Upper Great Bend has smaller wetlands and fewer of them. The Lower Great Bend has more forested wetlands (151 to 63) even though it is a smaller area. Those forested wetlands also tend to be larger. Lower Great Bend has lower average slope, likely due to the larger floodplains below the shoals. Average aspect and average solar radiation are very similar in both areas. Both areas have similar percentages of good tree-growing sediments and well-drained sediments. However, they differ with regard to percentage of loam, where the Lower Great Bend area has much more.

Figure 2.3

Graphs of site-size versus time. The upper graph includes Forbush Creek’s original size estimate. The lower graph includes my smaller estimate.

Figure 2.4

Map showing the location of pre- and post-AD 1200 settlements in the upper Yadkin River Valley.

Overall, the results show that larger sites date before AD 1200 and settlements appear to have moved upstream over time. However, across this shift in settlement strategy, communities had similar settlement location preferences. They continued to select high-quality agricultural land and avoid overland trails. Differences occurred with respect to wetlands and tributaries. This consistency occurred even though the Upper and Lower Great Bend areas show important differences with regard to available resources, such as loam sediments and wetlands.

Discussion Modeling settlement change over time The results support Woodall’s (1990) argument for an increase in settlements from the Middle to Late Woodland Periods (after AD 800). Barnette’s (1978) ceramic seriation showed that only five sites overlapped with the earliest occupations at Donnaha (i.e. 200 BC–AD 800), suggesting there may have been very few Middle Woodland settlements, but they may have been larger like Donnaha and Forbush Creek. My compilation of Barnette’s data with more recent work here shows many more Late Woodland sites. Although the evidence for population sizes is scanty, the increase in site numbers is so great that it is difficult to argue with Woodall’s (1990) hypothesis that population increased from the Middle to Late Woodland Periods (200 BC–AD 800). Table 2.6

Discriminant function analysis results comparing the pre-AD 1200 settlements with the sets of random sites. The

shading correlates with high, medium, low, and no distinction between the groups. Original cases were correctly classified into groups across the three results at 90.6%, 86.8%, and 92.5% Variable

Value (R1)

Value (R2)

Value (R3)

Floodplain area




Distance to trail




Area of loam in catchment




Number of forested wetlands within catchment




Number of tributaries within catchment




Average solar radiation in catchment




Slope at site




Viewshed size




Aspect at site




Average slope within catchment




Distance to nearest tributary




Area of good tree growth within catchment




Total wetland area within catchment




Area of well-drained sediment within catchment




Number of wetlands within catchment




Distance to closest contemporaneous site




Largest wetland within catchment




Using the surface area sizes and the data from the excavated sites, we can now begin to construct a model for settlement change over time from AD 800 to 1600. As mentioned, Woodall (1990) concluded, and I believe correctly so, that larger sites like Donnaha were created not by a significantly larger community but by a community reoccupying or moving around in the same location over a long period of time. As such, the settlement strategy during AD 800–1200 appears to have been one in which there were some communities, like at Donnaha, that occupied a particular floodplain for long periods of time. The very thick midden at this site is a good indication of this. They shifted location around the floodplain for what could have been a myriad of reasons: houses needing to be reconstructed due to decomposition, destruction of houses from flooding, build-up of trash in a particular area, etc. At the same time, there were a number of smaller settlements. It is also possible that the larger Donnaha size reflects a more aggregated Middle Woodland population during the earliest occupation dates before AD 800, and the Late Woodland communities dispersed throughout the Lower Great Bend. If MacPherson and Hardy are good examples of AD 800–1200 settlements, they also appear to have been occupied for significant periods of time. Like Donnaha, it appears a community (or series of them) shifted around the Hardy site. At MacPherson, the small area covered in conjunction with the 45cm-thick cultural layer indicates a lengthy occupation by a smaller community. Table 2.7

Discriminant function analysis results comparing the post-AD 1200 settlements with the sets of random sites. The shading correlates with high, medium, low, and no distinction between the groups. Original cases were

correctly classified into groups across the three results at 77.6%, 77.6%, and 85.7% Variable

Value (R1)

Value (R2)

Value (R3)

Floodplain area




Area of loam in catchment




Distance to nearest tributary




Largest wetland within catchment




Total wetland area within catchment




Viewshed size




Slope at site




Average slope within catchment




Distance to trail




Number of tributaries within catchment




Distance to closest contemporaneous site




Area of well-drained sediment within catchment




Average solar radiation in catchment




Number of forested wetlands within catchment




Aspect at site




Number of wetlands within catchment




Area of good tree growth within catchment




Table 2.8

The results of the landscape characterization, comparing the Upper Great Bend (UGB) area with the Lower Great Bend (LGB) areas

Landscape characteristic

UGB area

LGB area

Percentage of total area covered by wetlands



Average wetland size (m2)



Wetland size standard deviation (m2)



Percentage of total area covered by forested wetlands



Average forested wetland size (m2)



Forested wetland size standard deviation (m2)



Average slope



Average aspect (degrees)



Average solar radiation



Percentage of total area covered by loam



Percentage of total area covered by good tree sediment



Percentage of total area covered by well-drained 85.90% sediment


The excavated post-AD 1200 sites appear to have been occupied for shorter periods. Redtail, T. Jones, and Porter either have no intact cultural strata, or it is 5–10cm thick with

moderate densities of artifacts. Plowing has certainly disturbed these layers, but it also has at earlier sites like Donnaha and MacPherson, which still have 45–100cm-thick middens/cultural strata with dense artifact distributions. In addition, radiocarbon dates from these sites suggest occupations spanning 100–200 years compared with earlier sites, which range from 300 to 1,400 years. After AD 1200, these data suggest communities did not stay within particular floodplains for as long as they did before. If we follow Woodall’s conclusion that there is no complete sedentism of these communities, an interesting form of decreased geographic but not temporal sedentism occurred. That is, it is possible that shifts in settlement location occurred at similar frequencies throughout the entire study period. However, it appears that before AD 1200 some communities shifted only very small distances within the same floodplain for several centuries, a behavior not seen after AD 1200. To hypothesize at all about the relationship between site area and population size, we need to use settlement areas determined from excavation because we do not have enough sites to create a spatial relationship between surface and subsurface areas. In addition, if site size is even partially a factor of time, we must account for it. A simple division of the area by time for those sites with excavated settlement areas results in 20m2/year at the pre-AD 1200 Donnaha, and 12.5m2/year at T. Jones, 15m2/year at Porter, and 9m2/year at Redtail, which are all postAD 1200. This information, coupled with the decrease in overall and average surface site size (even with the omission of Donnaha and Forbush Creek) and the number of sites, suggests a population decrease after AD 1200. Site-level data suggest that post-AD 1200 communities were also less economically integrated than earlier communities. Woodall (1990:114–116) found that proportions of locally obtained quartz compared with rhyolite, which is found 50km to the southeast of the valley, correlate positively with distance upstream. Given that the Upper Great Bend sites are almost all post-AD 1200 sites, it suggests not just a geographic trend in decreased access to regional sources but a temporal one as well. They either lost access to rhyolite or lost trade connections that provided them with it. This could show that post-AD 1200 communities were less economically, and by extension possibly socially, interconnected throughout the western Piedmont compared with pre-AD 1200 communities. The pattern of decreased sedentism is opposite to what was happening in other areas of the Piedmont and most of eastern North America at this time. Communities in the Dan, Eno, and Haw River Valleys were building larger and more permanent villages (Simpkins 1985; Davis and Ward 1991). The same trend is seen in the Northeast (Snow 1995; Warrick 2008; Birch and Williamson 2013) and the Mid-Atlantic (George 1983; Hart 1993). In conjunction with these larger and more permanent settlements, communities in these other areas appear to have been more politically, socially, and economically integrated with one another. Again, the opposite trend was occurring among PVT communities in the upper Yadkin River Valley. The ecology of settlement change The discriminant function and landscape characterization results offer an opportunity to explore the relationships between population change, settlement strategy change, and ecology. Overall, the features and places preferred are quite similar for the pre- and post-AD 1200

settlements. The preference for more loam and less slope likely results from placing settlements in floodplains, which would have been good agricultural locations and good foraging locations. These are the best agricultural sediments in the area for growing maize and other crops. The river is nearby for fish, mussels, and turtles, the remains of which are found at most sites in the valley (Woodall 1984, 1990). Many of the floodplains also have backchannels and backswamps that would have also attracted birds and animals. All settlements regardless of date were farther from trails than random locations. This is not surprising for the post-AD 1200 communities that appear to not have been engaged in much regional trade. As mentioned, the pre-AD 1200 communities were acquiring rhyolite, but the trails may have been more trouble than they were worth if they were used by potential enemies. Settling away from trails may have been a defensive tactic. The two changes are in floodplain size and wetlands. Floodplain size became less important. This could have resulted from living in smaller settlements and being more mobile, or from the move upstream where smaller floodplains encouraged smaller communities. Either way, smaller communities do not need as much land for growing crops or foraging. Pre-AD 1200 communities favored forested wetlands, which are preferred deer habitats, while post-AD 1200 settlements were choosing locations with fewer and smaller wetlands. It could be a change in wild resources used. It could be that smaller wetlands could meet the needs of these new smaller communities. The Upper and Lower Great Bend landscapes were different in several ways. The Lower Great Bend area appears to have been the better resource area. I mentioned that the floodplains tend to be larger; the higher percentages of loam and flat land also meant there was more highquality agricultural land. The higher percentage of wetlands, the higher variability in wetlands, and the fact that there were more forest wetlands meant there were greater numbers of these resource areas for foraging. The two areas were similar with regard to sediments with high potential for tree growth, so access to wood resources and nut-producing trees would have been similar. Overall, it appears that the Lower Great Bend area was better for PVT foragerfarmers. This is an important result when we consider the observed pattern of communities moving upstream after AD 1200. It is unlikely that the move upstream was done to increase or maximize agricultural or foraging potential. Climatically, this period does correlate with the onset of the Little Ice Age, but that does not appear to have adversely affected the other PVT communities in nearby valleys. As such, I suggest sociopolitical or economic factors may have been influential in this move. Town Creek was established around AD 1100 (Reid 1965; Oliver 1992; Coe 1995; Boudreaux 2007), and is just over 80km south of the Parker site, which is the southernmost settlement examined here. Town Creek, Teal, Leak, and Payne are thought to represent a Pee Dee migration into the Pee Dee/lower Yadkin River Valley (Oliver 1992). The chronology of these settlements suggests the general movement was north as well. Payne is the most recent settlement and is located over 40km north of Town Creek. It is possible that the settlement of new people in the area either forced or encouraged a move by PVT communities upstream and away from the newcomers. Additionally, these Pee Dee communities would have been larger and had more complex sociopolitical organization. With no evidence of warfare (i.e. skeletal injuries, palisades, or destruction of houses or other structures) at any site in the upper Yadkin

River Valley, we may assume that they avoided conflicts. If the Upper Great Bend area was unoccupied, which it appears it was, and newcomers were encroaching to the south, a move upstream may have been a sacrifice made to avoid conflict with the new people. This could be another reason for avoiding trails as well. Pee Dee peoples were also moving nearby to the rhyolite sources and may have restricted access to them, explaining why the post-AD 1200 settlements have lesser amounts. Hierarchically organized societies with Mississippian traits do not appear in the extreme upper Yadkin River Valley and upper Catawba River Valley until around AD 1400 (Beck and Moore 2002; Moore 2002), so the move upstream would not have put PVT communities closer to them during this move.

Conclusions In this chapter, I proposed a new model for settlement change in the upper Yadkin River Valley of the Piedmont Southeast. I provided evidence that throughout the Late Woodland Period communities likely moved upstream and became less sedentary. There may have been a concurrent decrease in population as well, but at this time it is not clear what would have caused it. This pattern is noteworthy, however, because the general Late Woodland trend is one of increasing population, increasing sedentism, increasing settlement size, and increasing engagement in regional sociopolitical and economic networks. Communities in the upper Yadkin River Valley were trending in the other direction for all of these characteristics. The necessary question is: why did they stray from the larger trends? The settlement ecology analysis suggests that PVT settlement location preferences did not change significantly during this transition. The site-level excavation results support this, showing continuity in subsistence. Whether occupied in AD 800 or AD 1600, PVT communities were relying on a mixture of agriculture and foraging for wild plants and animals. What did change was the landscape into which they moved. It actually had fewer of the resources and features that they preferred. Thus, they were moving into a less desirable landscape. It could be that the decrease in population sizes meant they did not need as much agricultural land. However, all of the results combined —decreasing population, moving to a less desirable landscape, decreased sedentism, and the loss of access to the highest-quality lithic source in the western Piedmont—paint a picture of increasing stress. It is difficult to tease apart the timings of these changes, but a shift upstream to a less desirable landscape may have necessitated a more mobile settlement strategy. The smaller floodplains, poorer sediments, and sparse wetlands may not have been able to sustain the use of any particular location for as long as areas downstream could do. This proposed change necessitates asking why PVT communities would move upstream? They undoubtedly would have recognized it as a less desirable landscape. The answer might be the formation of Mississippian communities into the lower Yadkin River Valley. Whether through migration or in situ development, these communities introduced a hierarchical sociopolitical organization and began building settlements within three days’ travel to the south. For a group of communities that appear to have avoided warfare, which was occurring throughout the rest of the Piedmont, this intrusion of new people may have been enough to encourage a move upstream, leading to the changes outlined above. Within 300 years, hierarchical societies with Mississippian characteristics appear to the immediate west of the

upper Yadkin River Valley. At this point, we see evidence of significant interaction between them and the two closest PVT communities, T. Jones and Porter. Though there is no evidence of violence, it is uncertain how mutually beneficial these interactions were. The combination of traditional settlement pattern and more recent settlement ecology research methods here helped to identify a somewhat unique pattern of demographic and settlement change within eastern North America. Undoubtedly, more site-level work is needed to examine population size, evidence of subsistence change or food stress, and interaction with Mississippian communities. However, this combined approach identified the patterns and offered an empirically based explanation for why the geographic shift upstream was likely not caused by subsistence needs, leading to hypotheses about sociopolitical reasons for the move. These factors need to be explored in more depth, but this identification suggests the move had something to do with their location at the edge of the Mississippian world in the Southeast. The Mississippian frontier is often studied from the point of view of Mississippians. There appear to have been significant impacts on those groups on the outside looking in as well. We need to examine both sides of the frontier—and possibly think about it more as a buffer between political entities in order to accurately examine the lives of those living there—to fully describe and explain the impact of the formation of Mississippian hierarchical societies on neighboring communities and societies, and on the natural and cultural landscapes of eastern North America.

References Allen, Kathleen M.S. 1996 Iroquoian Landscapes: People, Environments, and the GIS Context, in New Methods, Old Problems: Geographic Information Systems in Modern Archaeological Research, Center for Archaeological Investigations, edited by H. Maschner, pp. 198–222. Southern Illinois University, Carbondale. Anderson, David G. 1994 The Savannah River Chiefdoms. University of Alabama Press, Tuscaloosa. Barnette, Karen 1978 Woodland Subsistence-Settlement Patterns in the Great Bend Area, Yadkin River Valley, North Carolina. Unpublished MA thesis. Wake Forest University. Beck, Robin A. Jr. and David G. Moore 2002 The Burke Phase: A Mississippian Frontier in the North Carolina Foothills. Southeastern Archaeology 21(2): 192–205. Birch, Jennifer and Ronald F. Williamson 2013 The Mantle Site: An Archaeological History of an Ancestral Wendat Community. Altamira Press, Lanham, MD. Boudreaux, Edmond A. 2007 The Archaeology of Town Creek. University of Alabama Press, Tuscaloosa. Chisholm, Michael 1968 Rural Settlement and Land Use. Second edition. Hutchinson, London. Coe, Joffre 1995 Town Creek Indian Mound: A Native American Legacy. University of North Carolina, Chapel Hill. Davis, R.P. Stephen and H. Trawick Ward 1991 The Evolution of Siouan Communities in Piedmont North Carolina. Southeastern Archaeology 10(1): 40–53. Dickens, Roy S. Jr. 1978 Mississippian Settlement Patterns in the Appalachian Summit Area: The Pisgah and Qualla Phases, in Mississippian Settlement Patterns, edited by Bruce D. Smith, pp. 115–139. Academic Press, New York. Dickens, Roy S. Jr., H. Trawick Ward, and R.P. Stephen Davis Jr. 1987 The Siouan Project: Seasons I and II. Research Laboratories of Anthropology, Monograph Series No. 1, University of North Carolina, Chapel Hill. Fecteau, Rudy, James Molnar, and Gary Warrick 1991 Iroquoian Ecology. Birdstone 5(1): 1–19. George, Richard L. 1983 The Gnagey Site and the Monongahela Occupation of the Somerset Plateau. Pennsylvania Archaeologist 53(4): 1–97. Goddard, Ives 2005 Indigenous Languages of the Southeast. Anthropological Linguistics 47(1): 1–60. Haas, Jonathan and Winifred Creamer 1993 Stress and Warfare among the Kayenta Anasazi of the Thirteenth Century AD, in Fieldiana: Anthropology, n.s. 21. Field Museum of Natural History, Chicago. Hart, John P. 1993 Monongahela Subsistence-Settlement Change: The Late Prehistoric Period in the Lower Upper Ohio River Valley. Journal of World Prehistory 7(1): 71–120. Hasenstab, Robert J. 1996 Settlement as Adaptation: Variability in Iroquois Village Site Selection as Inferred Through GIS, in

New Methods, Old Problems: Geographic Information Systems in Modern Archaeological Research, edited by H. Maschner, pp. 223–241. Center for Archaeological Investigations, Southern Illinois University, Carbondale. Jefferies, Richard W., Emanuel Breitburg, Jennifer Flood, and C. Margaret Scarry 1996 Mississippian Adaptation on the Northern Periphery: Settlement, Subsistence and Interaction in the Cumberland Valley of Southeastern Kentucky. Southeastern Archaeology 15(1): 1–28. Jones, Eric E. 2006 Using Viewshed Analysis to Explore Settlement Choice: A Case Study of the Onondaga Iroquois. American Antiquity 71: 523–538. Jones, Eric E. 2010 An Analysis of the Factors Influencing Sixteenth- and Seventeenth-Century Haudenosaunee (Iroquois) Settlement Locations. Journal of Anthropological Archaeology 29: 1–14. Jones, Eric E. 2015 Settlement Ecology of Middle-Range Societies in the Western North Carolina Piedmont, AD 1000–1600. North Carolina Archaeology 64: 1–32. Jones, Eric E. and Peter Ellis 2016 Multiscalar Settlement Ecology Study of Piedmont Village Tradition Communities, AD 1000– 1600. Southeastern Archaeology 35(2): 85–114. Jones, Eric E. and James W. Wood 2012 Using Event-History Analysis to Examine the Causes of Semi-Sedentism among Temperate Swidden Agriculturalists: A Case Study of the Haudenosaunee, AD 1500–1700. Journal of Archaeological Science 39(8): 2593–2603. Jones, Eric E., Madison Gattis, Andrew Wardner, Thomas C. Morrison, and Sara Frantz 2012 Exploring Prehistoric Tribal Settlement Ecology in the Southeast: A Case Study from the North Carolina Piedmont. North American Archaeologist 33(2): 157–190. Kvamme, Kenneth L. 1990 One-Sample Tests in Regional Archaeological Analysis: New Possibilities through Computer Technology. American Antiquity 55: 367–381. Lawson, John 1967[1709] A New Voyage to Carolina, edited by Hugh Talmage Lefler. University of North Carolina Press, Chapel Hill. Maschner, Herbert D.G. 1996 The Politics of Settlement Choice on the Northwest Coast: Cognition, GIS, and Coastal Landscapes, In Anthropology, Space, and Geographic Information Systems, edited by M. Aldenderfer and H. Maschner, pp. 175–189. Oxford University Press, New York. Mikell, Gregory A. 1987 The Donnaha Site: Late Woodland Period Subsistence and Ecology. Unpublished master’s thesis. Department of Anthropology, Wake Forest University, Winston-Salem. Mooney, James 1894 The Siouan Tribes of the East. Smithsonian Press, Washington DC. Moore, David 2002 Catawba Valley Mississippian: Ceramics, Chronology, and Catawba Indians. University of Alabama Press, Tuscaloosa. Mouzon, Henry Jr. 1775 An Accurate Map of North and South Carolina, with their Indian Frontiers. Robert Sayer and J. Bennett, London. Myer, William E. 1971 Indian Trails of the Southeast. Blue & Gray Press, Nashville. Newkirk, Judith 1978 The Parker Site: A Woodland Site in Davidson County, North Carolina. Unpublished MA thesis. Wake Forest University, Winston-Salem, NC. Oliver, Billy L. 1992 Settlements of the Pee Dee Culture. Unpublished Ph.D. dissertation, Department of Anthropology, University of North Carolina, Chapel Hill. Pluckhahn, Thomas J. and David A. McKivergan 2002 A Critical Appraisal of Middle Mississippian Settlement and Social Organization on the Georgia Coast. Southeastern Archaeology 21(2): 149–161. Poulsen, John and Aaron French 2004 Discriminant Function Analysis (DA). San Francisco State University. Available at: http://userwww.sfsu.e​du/efc/classes/biol71​0/discrim/discrim.pdf (accessed 26 July 2016). Reid, James 1965 A Comparative Statement on Ceramics from the Hollywood and the Town Creek Mounds. Southern Indian Studies 17: 12–25. Rogers, Rhea J. 1995 Tribes as Heterarchy: A Case Study from the Prehistoric Southeastern United States. Archaeological Papers of the American Anthropological Association 6(1): 7–16. Sakaguchi, Takashi, Jesse Morin, and Ryan Dickie 2010 Defensibility of Large Prehistoric Sites in the Mid-Frasier Region on the Canadian Plateau, Journal of Archaeological Science 37: 1171–1185. Simpkins, Daniel L. 1985 First Phase Investigations of Late Aboriginal Settlement Systems in the Eno, Haw, and Dan River Drainages, North Carolina. Research Laboratories of Anthropology, Research Report No. 3. University of North Carolina, Chapel Hill. Snow, Dean R. 1995 The Mohawk Valley Project: The Sites. Matson Museum of Anthropology, Pennsylvania State University. Sokal, R.R. and F.J. Rohlf 1995 Biometry. Third edition. W.H. Freeman and Company, New York. Stone, Glenn Davis 1996 Settlement Ecology: The Social and Spatial Organization of Kofyar Agriculture. Arizona Studies in Human Ecology. University of Arizona Press, Tucson. Sturtevant, William C. 1958 Siouan Languages in the East. American Anthropologist 60(4): 738–743. Thomas, Cyrus 1887 Work in Mound Exploration of the Bureau of Ethnology. Bureau of Ethnology Bulletin 4. Smithsonian,

Washington DC. Thomas, Larissa 1996 A Study of Shell Beads and their Social Context in the Mississippian Period: A Case from the Carolina Piedmont and Mountains. Southeastern Archaeology 15(1): 29–46. VanDerwarker, Amber M., C. Margaret Scarry, and Jane M. Eastman 2007 Menus for Families and Feasts: Household and Community Consumption of Plants at Upper Saratown, North Carolina, in The Archaeology of Food and Identity, edited by K.C. Twiss, pp. 16–49. Center for Archaeological Investigations, Southern Illinois University, Carbondale. Ward, H. Trawick and R.P. Stephen Davis Jr. 1991 The Impact of Old World Disease on the Native Inhabitants of the North Carolina Piedmont. Archaeology of Eastern North America 19: 171–181. Ward, H. Trawick and R.P. Stephen Davis Jr. 1993 Indian Communities on the North Carolina Piedmont, AD 1000 to 1700. Research Laboratories of Anthropology, Monograph Series No. 2. University of North Carolina, Chapel Hill. Ward, H. Trawick and R.P. Stephen Davis Jr. 1999 Time before History: The Archaeology of North Carolina. University of North Carolina Press, Chapel Hill. Ward, H. Trawick and R.P. Stephen Davis Jr. 2001 Tribes and Traders on the North Carolina Piedmont, AD 1000–1710, in Societies in Eclipse: Archaeology of the Eastern Woodlands Indians, AD 1400–1700, edited by David S. Brose and Robert C. Mainfort Jr., pp. 125–142. Smithsonian, Washington, DC. Warrick, Gary 2008 A Population History of the Huron-Petun, AD 500–1650. Cambridge University Press, New York. Woodall, J. Ned 1984 The Donnaha Site: 1973, 1975 Excavations. North Carolina Archaeological Council Publication No. 22. Raleigh, NC. Woodall, J. Ned 1990 Archaeological Investigations in the Yadkin River Valley, 1984–87. North Carolina Archaeological Council Publication No. 25. Raleigh, NC. Woodall, J. Ned 1999 Mississippian Expansion on the Eastern Frontier: One Strategy in the North Carolina Piedmont. Archaeology of Eastern North America 27: 55–70. Woodall, J. Ned 2009 The T. Jones Site: Ecology and Agency in the Upper Yadkin Valley of North Carolina. North Carolina Archaeology 58: 1–58.

3 Settlement ecology at Singer-Moye Mississippian history and demography in the southeastern United States Stefan Brannan and Jennifer Birch

Conceptual framework Settlement ecology seeks to establish a theoretical basis for understanding human settlement dynamics using principles adapted from geography, anthropology, archaeology, and agricultural economics (Stone 1996). This approach encourages us to consider how settlement patterns reflect relationships between people and their cultural and natural environments. At its core, settlement ecology is based on functionalist principles whereby natural environmental variables, including subsistence resources, other raw materials needed to produce the necessities of life (e.g. to meet shelter, comfort, or health needs), and items for trade or exchange are thought to affect cultural adaptation. While archaeological frameworks informed by settlement ecology take human–environment interaction as a primary organizing principle, its proponents recognize that settlement patterns are also strongly influenced by social and political factors, including but not limited to ethnicity, ideology, belief systems, and placemaking as well as relationships between the natural and cultural realms at multiple scales of analysis (e.g. Anschuetz et al. 2001; Birch and Williamson 2015; Jones and Wood 2012). Our understanding of settlement ecology is in keeping with that outlined in the introduction to this volume. In particular, we recognize that landscapes are a product of long-term human– environment interaction, as advocated by proponents of historical ecology (Balée and Erickson 1998; Crumley 2007; Marquardt and Crumley 1987). Within the framework of an explicitly historical settlement ecology, humans are not passive respondents to environmental and cultural contexts (Marquardt 1992:105–106). Humans act intentionally and unintentionally as a keystone species, playing an active role in the manipulation and construction of environments (Thompson and Waggoner 2013). In some cases, the presence of humans, either continuously or intermittently, creates productive and biodiverse landscapes, prompting permanent investments in the same landscape which can be taken advantage of by succeeding generations (Balée and Erickson 2006; Brookfield 2001). These anthropogenic disturbances can have long-lasting effects on settlement decisions by endogenous and exogenous groups. At the same time, populations respond to internal and external challenges and opportunities based on prevailing social, political, and economic organizations (Marquardt 2013:5). As such, changes in human societies are not caused by environmental or cultural factors alone, but also by the dialectical,

mutually constitutive relationships between societies and the environment (Marquardt 2013: 5). As archaeologists working in the late prehistoric southeastern United States (c.AD 1000– 1540), we appreciate the utility of a conceptual approach grounded in a historical settlement ecology because it allows us to break away from more traditional explanations for changes in settlement patterns which have been dominated by models focused on the development of socio-political complexity and rooted in the chiefdom concept (see, for example, summaries in Anderson and Sassaman 2012; Blitz 2010; Cobb 2003). After c.AD 1000, a suite of commonly co-occurring cultural transformations took place across the southeastern United States that mark the onset of the Mississippian cultural phenomenon. These patterns include the construction of platform mounds, often with associated plazas, which formed the cores of sizable towns (Blitz 2010; Milner 2012; Pauketat 2007). Populations inhabiting multi-mound centers, single-mound centers, and dispersed farmsteads subsisted on maize, as well as eastern agricultural products, supplemented by hunting, fishing, and the collection of a broad range of wild resources (Scarry 1993; Smith 1989). Changes in material culture include shifts in ceramic design and manufacturing sequences as well as elaboration in vessel forms (e.g. Pauketat and Emerson 1991; Steponaitis 2009). Other symbolically charged materials, including copper and shell, were worked and decorated with new iconographic motifs referencing culture-heroes and the cosmos (Muller 1989). Archaeological deposits and ethnohistoric records suggest the existence of leaders who wielded varying degrees of power over the economic, political, and ritual components of society (Anderson 1994; Cobb and King 2005; DePratter 1991; Hally 1996; Knight 1986). These patterns continue, albeit with significant spatiotemporal variation, into the mid-sixteenth century when ethnohistoric descriptions of societies with these characteristics were recorded by early European explorers and soldiers (Clayton et al. 1993; Hudson 1997). A persistent theme of research in the late prehistoric Southeast is explaining the emergence and maintenance of the largest mound centers, often viewed as the centers of regional polities (e.g. King 2003; Knight and Steponaitis 1998; Pauketat and Emerson 1997). Traditional explanations for the regional distribution of mound centers/central places/polities have favored functionalist and materialist arguments based on political centralization (Anderson 1994; Hally 1993, 1996, 2006; Peebles and Kus 1977). Today, archaeologists are increasingly turning away from generalizing approaches and embracing conceptual frameworks based on history and practice, variability in the exercise of agency and power, and the role of ritual and symbolism, as well as traditional concerns with production, consumption, and the ecological dimensions of geopolitical change (e.g. Alt 2010; Beck 2013; Cobb 2003, 2014; Pauketat 2007, 2013; Pauketat and Alt 2005; Pluckhahn 2010). These contemporary approaches acknowledge that socio-cultural change is produced in practice by individuals as they navigate the historically contingent settings of everyday life. Such an approach readily accommodates explaining the archaeological record through the analysis of social and ecological landscapes, and how people navigated each over time. The focus of this chapter is on settlements which Thompson and Pluckhahn (2012:50) call “persistent monumental places.” Such places possessed environmental and economic characteristics as well as natural and cultural features that focused on the occupation and reuse

of a particular place through time (Schlanger 1992). In some locations, the practice of monumentality intersected with the continued occupation of a persistent place and the emplacement of people at a particular locus in the wider social and environmental landscape (Charles and Buikstra 1983; Cobb 2005). Platform mounds were one of the most overt forms of monumentality in the Precolumbian southeastern United States. These elevated spaces facilitated social processes of both integration and differentiation (Lindauer and Blitz 1997). Some supported buildings, including elite residences, temples, or charnel houses. Others served as platforms for public events. The act of mound construction is often cited as an example of communal activity through labor mobilization, which furthered the interests of specific social units: the community as a whole; a sub-set of a community such as kin groups, or sodalities, or the elite; individual leaders and their lineages; or some combination thereof (Byers 2006; Hally 1996, Kidder 2004; Knight 1986, 2010:4-5; Lindauer and Blitz 1997). The act of mound construction represents the fusion of physical and social work involved in community building (Kowalewski 2013). In the same way, the end of a mound’s use-life has been argued to coincide with the fissioning or relocation of local social groups (Blitz 1999). In this chapter we employ a conceptual framework influenced by historical and settlement ecology, together with an approach emphasizing history and practice, in order to understand the settlement history of Singer-Moye, a large multi-mound center located in southwest Georgia, and its shifting place in a set of regional polities in the lower Chattahoochee River valley. Singer-Moye represents one of the largest and most comprehensive data sets for the Mississippian period from the region. In this chapter, we first explore the social and environmental context of Singer-Moye. Second, we discuss recent research aimed at delineating the occupational history of the site, including population estimates for each phase of the site’s occupation. Those observations are interpreted with reference to the wider ecological and regional settlement context. Finally, we produce a set of hypotheses and future research directions related to the historical settlement ecology of the Singer-Moye site and the lower Chattahoochee River valley.

Environmental and cultural context Singer-Moye was occupied between c.AD 1100 and 1500, though perhaps not continuously (Blitz and Lorenz 2002, 2006). The most prominent features of the built environment at the site are the remains of five platform and three dome-shaped mounds arranged around two plazas (Figure 3.1), which together constitute the site core. In 2008, the University of Georgia (UGA) acquired the site. These lands include some 51 ha of property, encapsulating the site core and adjacent lands. Systematic shovel-test survey of the area within the UGA property boundaries revealed 31 ha of occupation (Brannan 2016; Figure 3.1). While approximately 6 ha of this area constituted the site core, we believe that the remaining 25 ha consisted of residential occupation and associated land-use on well-drained upland terraces. All of these lands appear to have been occupied contemporaneously when the town reached its greatest size in the fourteenth century. Additional Mississippian-component occupation is almost certainly present on other well-drained landforms and in the floodplain of Pataula Creek, beyond the survey boundaries.

Figure 3.1

Singer-Moye site plan, including site core and extent of UGA lands.

Between c.AD 1100 and 1200, Mississippian peoples settled in several locations in the lower Chattahoochee River valley, including at Singer-Moye. The rapid appearance of a Mississippian cultural complex without previous local antecedents is suggestive of a site unit intrusion rather than local reorganization and independent innovation (Blitz and Lorenz 2002, 2006; Pauketat 2007). Ceramic motifs resembling those produced by inhabitants of Moundville-related polities in west-central Alabama suggest that these populations came from the west, although the relationship of the local inhabitants to Moundville itself remains unclear, as does the degree of interaction with a “homeland” through time (Jenkins 2009; Knight 2010; Regnier 2014). It is notable that Singer-Moye is the only multi-mound center not located near the major river in the valley. Other central places in the region are situated next to the Chattahoochee River, which marks the boundary between Southeast Alabama and Southwest Georgia and provided the residents with easy access to riverine resources, broad alluvial floodplains, and transportation networks. Singer-Moye, on the other hand, is situated next to a small tributary some 30 km east of the river as the crow flies, or 45 km from the confluence of the stream and the river (Figure 3.2). However, its location on an ecotone and on an upland boundary between two drainages may have been strategic. The community that settled at Singer-Moye situated themselves on the boundary between two neighboring ecoregions—the Hilly Gulf Coastal Plain and the Coastal Plain Red Uplands (Figure 3.3) (Griffith et al. 2001). The former is a heterogeneous region consisting of dissected plains and low hills with broad tops, northward-facing cuestas, and numerous streams with

abundant floodplains. It was covered by an oak-hickory-pine forest, some southern mixed forest, and some southern floodplain forest, which contributed to significant microclimate diversity (Schnell and Wright 1993; Schnell et al. 1981; Wharton 1978). The Coastal Plain Red Uplands ecoregion, in contrast, consists of dissected plains; broad, gently sloping ridges; and interstream divides (Griffith et al. 2001). It was covered by a different forest makeup consisting primarily of southern mixed forest interspersed with oak-hickory-pine (Wharton 1978). Being situated near an ecotone meant that a wider range of resources would have been available to people at Singer-Moye, assuming that they used both environmental regions. Unfortunately, comparative collections from other contemporary sites in the valley are lacking, but zooarchaeological and archaeobotanical analyses at Singer-Moye indicate that numerous terrestrial and riverine faunal resources were used by local inhabitants and that a mixture of agricultural crops and wild resources was incorporated into the local diet (Table 3.1).

Figure 3.2

Locations of sites discussed in text.

Figure 3.3

Level IV ecoregions in the lower Chattahoochee River valley.

Singer-Moye is also located on a cultural frontier. It is the easternmost settlement of appreciable size in this boundary region. It is situated at the height of land between the Chattahoochee and Flint River drainages, a position that may have been strategically important in connecting these two watersheds, each of which was inhabited by peoples engaging in very different cultural practices. While survey coverage east of Singer-Moye is limited, data from sites in the 50-km stretch between Singer-Moye and the Flint River suggest that several different populations inhabited the region, though they remain understudied (GASF 2015; Elliott and Dean 2006). After c.AD 1100, groups in the central lower Chattahoochee River valley were engaged in a fully Mississippian, agriculturally based political economy. Conversely, populations in the Flint River drainage practiced a mix of semi-sedentary settlement-subsistence and year-round sedentary intensive agricultural practices. It seems unlikely that people originating at Singer-Moye or the lower Chattahoochee River settled along

the Flint River. Small, single-mound centers have been identified along the Flint River approximately 100 km to the northeast and southeast of Singer-Moye, but differences in material culture and the timing of their occupations suggest that, while there may have been some minor interaction between these polities and Singer-Moye, their cycles of fluorescence and abandonment do not appear to be connected to polity cycling in the lower Chattahoochee River valley (Chamblee 2006; Worth 1988). Table 3.1

Animal (left) and plant (right) species identified in the archaeological record at Singer-Moye (adapted from Little 2013; Russell and Gordy 2012)

Common Name


Common Name


American beaver

Castor canadensis


Quercus sp.

American bullfrog

Lithobates catesbeianus


Phaseolus sp.




Rubus spp.

Cottontail rabbit

Sylvilagus sp.


Amaranthus sp.

Eastern mud turtle

Kinosternon subrurum


Zea sp.

Fox squirrel

Sciurus niger


Carya sp.

Freshwater mussels



Malva sp.




Phalaris sp.


Didelphis virginiana

Morning glory

Convolvulus sp.

Passenger pigeon

Ectopistes migratorius


Diospyros sp.


Procyon lotor


Passiflora sp.

Red fox

Vulpes vulpes


Cucurbita spp.



River cane

Arundinaria sp.

Southeastern pocket gopher

Geomys pinetis

Sunflower Walnut

Helianthus sp. Juglans sp.


Sciurus spp.

Terrestrial snails




Venus clam

Chione sp.

White-tailed deer

Odocoileus virginianus

Wild turkey

Meleagris gallopavo

Based on our appraisal of the cultural and natural environment at Singer-Moye, we contend that there were few, if any, natural or cultural constraints that limited the initial settlement at the local level: for example, a scarcity of resources or other large, possibly antagonistic groups which would have affected the continued growth and organizational possibilities for the Singer-Moye community. As we discuss in more detail below, the historical trajectory of Singer-Moye is characterized by continued growth after its initial settlement as a central place, and resilience to macroregional changes which affected other local groups and adjacent regions with similarly organized polities.

Reconstructing a settlement history for Singer-Moye The size of a residential population is a key factor in determining the outcomes of long-term human–environment interaction. We recognize that most population estimates based on archaeological data run the gamut from tenuous to wildly speculative. The best population estimates can be reconstructed from settlement pattern data: specifically domestic architecture coupled with estimates of household size (e.g. Snow 1995; Warrick 2008). With cautious optimism, we set out to estimate the population of Singer-Moye, a site where no communitylevel architecture has been identified to date. Our methods involved combining settlement pattern data from extensive geophysical survey of other large multi-mound centers and estimates of the relative density of occupation at Singer-Moye based on the density of artifactual remains from Brannan’s (2016) shovel-test survey. We chose Etowah and Moundville as sites upon which to base our population estimates. These two settlements are more or less contemporary with Singer-Moye. They exhibit similar patterns of monumental, plaza, and residential space, though Moundville is larger in total area than either Etowah or Singer-Moye. Nevertheless, we believe that Singer-Moye contains similar patterns of residential occupation and plaza use. At Etowah (Walker 2009) and Moundville (Davis et al. 2015), large-scale gradiometer surveys have located anomalies which have been interpreted as domestic and special-purpose structures. Some of these interpretations have been confirmed by test excavations (Davis et al. 2015). If we assume that the residential occupation of these other, similar mound centers was comparable to that at Singer-Moye, these data can be used to generate rough population estimates. To obtain an approximate estimate of residential population density within surveyed areas, we selected the structures representative of households based on total roofed area (Steere 2011) and removed larger structures—those that likely served a special purpose—from the data set (Figure 3.4). We then calculated the density of households per hectare for 7 ha of site area at Etowah and 8 ha at Moundville. Using Casselberry’s (1974) ethnographic model of 6 m2 of roofed area per person, we generated estimates of residential population per hectare at each site (Table 3.2). Because settlement density varied within different hectares at both Moundville and Etowah, we calculated a high population density figure based on the average population from all selected hectares and the low population density from the hectare with the minimum. We chose these metrics in an attempt to avoid overestimating our population thresholds. For both Etowah and Moundville, both the high- and low-density population estimates per hectare were similar.

Figure 3.4

Interpretive site maps based on gradiometer data and selected squares used for calculations: (top) Etowah (after Walker 2009: fig. 3.4, reproduced by permission of Adam King); (bottom) Moundville (after Davis et al. 2015: fig. 4, reproduced by permission of the Society for American Archaeology from American Antiquity, 80(1) © 2015). Grid squares = 1 ha.

Although we do not yet know the specific residential footprint for household structures at Singer-Moye, we used our shovel-test data, represented by Thiessen polygons, as a rough proxy for changes in residential settlement size (Figure 3.5). Distributions of chronologically significant ceramic attributes from shovel-tests were mapped across the site’s area and tied to phases of mound use. The occupational history of Singer-Moye is based largely on the local ceramic chronology, originally developed by Knight (1979) and refined by Blitz and Lorenz (2006). The local ceramic chronology was further refined by Brannan (2016) using both ceramic type and attribute data. Mound contexts were dated using both ceramic sequences and radiocarbon assays. Thiessen polygons were associated with high-density occupation when there were ten or more ceramic sherds recovered. Polygons representing nine or fewer sherds

were considered low-density occupation. The resulting calculations allow us to roughly estimate the relative size of high- and low-density occupation per phase at Singer-Moye and calculate the population based on our averages from Etowah and Moundville. Table 3.2

Population estimates from Moundville and Etowah Number of hectares

Number of structures

High-density population (per ha)

Low-density population (per ha)











Figure 3.5

Singer-Moye thiessen polygons.

The site chronology suggests that Singer-Moye’s occupation can be broken down into three distinct phases, as discussed in detail below. In each phase, we see fluctuations in population size, settlement distribution, and the construction and use of monumental architecture (Figure 3.6). The early phase coincides with the initial settlement and limited growth of Singer-Moye, c.AD 1100–1300. The middle phase is identified as the period when the settlement was at its greatest extent, c.AD 1300–1400. The late phase measures Singer-Moye’s final period of occupation, a horizon marked by a significant contraction in utilized space, c.AD 1400– 1450/1500. There are three caveats to our numbers. First, Moundville and Etowah are both palisaded villages and nucleated settlements may have higher population densities than non-nucleated settlements. There is, to date, no evidence of a defensive palisade circumscribing Singer-Moye

(Brannan and Birch 2015; Milner 2000). Second, the early and late phases of occupation at Singer-Moye are completely contained within the footprint of the denser, middle phase of occupation, which may artificially inflate population estimates for the earliest and latest occupations. One drawback to basing population estimates on shovel-test data is that population estimates for each phase are based on the total number of ceramics from each shovel-test, as non-diagnostic sherds could not be identified with any single phase of occupation. Finally, our population estimates for Etowah, Moundville, and Singer-Moye reflect absolute numbers of structures and the sum total of artifactual deposition, respectively, ignoring the fact that each site is a palimpsest. Despite these caveats, we believe that the estimates provide a useful conceptual tool for theorizing about socio-political dynamics as well as differential stress on the local environment. Early phase: c.AD 1100 to 1300 The first phase of occupation at Singer-Moye, beginning at approximately AD 1100, was focused on three loci within the site core (Figure 3.7). One locus is located under and near Mound C, a platform mound constructed during this earliest phase of occupation. The second is located approximately 300 m to the south, near Mound H. Mound H was completely excavated by Russell and Gordy (2012); it comprised a dome-shaped mound dating to later in SingerMoye’s settlement history and was constructed over earlier settlement features, notably, walltrench structures superimposed by a possible elite residence. A third loci was found just north of Mound B. The early phase footprint, with an area of 6 ha, was found completely under latter phases, making the exact boundary difficult to identify. We also suspect that the areas where plazas were constructed may have been occupied during this time period, but the significant modifications to the built environment, at least in the south plaza (Brannan and Bigman 2012), rendered the settlement sequence unclear at the resolution of our current data set.

Figure 3.6

Extent of Singer-Moye by phase of occupation.

Figure 3.7

Singer-Moye, early phase of occupation, c.AD 1100 to 1300.

One AMS and one conventional radiocarbon date from the sub-mound C midden indicates occupation at cal AD 1170 to 1222 and 1209 to 1286 (1-sigma). An occupational stratum located immediately north of Mound H excavated in the 2015 season also recently produced an AMS date of cal AD 1225 to 1256 (1-sigma), which also corresponds to the early phase of occupation (Table 3.3). We suspect that the size of the early settlement never grew above 6 ha and was probably smaller. Based on distributions of early Rood phase shell-tempered pottery and the methods described above, we estimate a maximum population of 445 individuals during this phase of occupation (Table 3.4). Evidence for occupation in other parts of the survey area during the early phase is inconclusive, but we leave open the possibility that households or other buildings were built away from the site core. Table 3.3 Radiocarbon dates discussed in text. All dates calibrated using OxCal version 4.2 (Bronk-Ramsay 2009) and the Intcal13 calibration curve (Reimer et al. 2013)

Table 3.4 Residential settlement size and population estimates at Singer-Moye by phase

As mentioned above, one platform mound, Mound C, was constructed near and over one of the early occupational loci, placing its construction after Singer-Moye’s initial settlement. The ceramics recovered from the lower levels of the mound indicate that the incipient stages were constructed during the early phase of occupation. During this early phase, the community made an investment in the modification of the built environment, emplacing themselves at the site through the construction of a platform mound over remains of some of the earliest activity at the site. However, the precise function of Mound C (e.g. a funerary mound, structural base, viewing platform [see Blitz and Lindauer 1997]) and any related practices are still unknown due to the limited amount of testing conducted on it to date. The only other mound that may have been in use during this early phase of occupation is Mound A (Blitz and Lorenz 2006). The size of Mound A (14 m in height, with a base covering approximately 3,900 m2) suggests a great deal of labor investment, possibly from an early date, though the earliest stages of this mound have not been investigated. There is no evidence of occupation at the site immediately pre-dating the initial Mississippian settlement at Singer-Moye although some Archaic period material (c.8000 to 3000 BP ) was recovered from isolated shovel-tests, including fiber-tempered pottery, steatite bowl fragments, and diagnostic projectile points. The activities of some 400-plus residents practicing an agricultural economy would have significantly modified the natural landscape. In particular, the clearing of forest for settlement, fields, and firewood would have created an anthropogenic mosaic across the landscape that provided expanded habitat for deer and a

range of other animals, especially along forest edges and openings. Residents may have also encouraged the growth of wild plants and cultigens, including both maize and crops of the eastern agricultural complex, in garden plots located near domestic structures and larger field systems located in the floodplain (Scarry and Scarry 2005). Between AD 1100 and 1300, Mississippian peoples settled and constructed mounds at four other known places in the Lower Chattahoochee Valley (Blitz and Lorenz 2006: 33-59). All sites with monumental architecture were spaced between 16 and 36 km apart, close enough for regular interaction but perhaps far enough from each other to avoid the effects of social or environmental circumscription. There are coarse settlement size estimates for two. The first, Cool Branch, was a palisaded village encompassing 4.5 ha occupied between AD 1100 and 1200 (Blitz and Lorenz 2006:48–55). Unfortunately, the area was not systematically sampled before its destruction as the result of reservoir construction, so we are unable to differentiate between high- and low-density areas of occupation. Based on the total area within the palisade adjusted for the percentage of low- and high-density occupation at Singer-Moye at its greatest known extent (60.6 percent of total site area), we calculate that between 141 and 269 people lived at Cool Branch, making it a similar size to Singer-Moye at the same time (Table 3.5). The second settlement, Cemochechobee, contains episodes of mound construction dating to between AD 1200 and 1300 (Blitz and Lorenz 2006:40–42), but the surrounding village may have been occupied for longer, between AD 1200 and 1400 (Schnell et al. 1981). Schnell et al. (1981:2) suggest that the village may have encompassed as many as 61 ha based on surface collections. They qualify their estimate by stating that the whole area may not have been occupied continuously and the low amount of midden accumulation may indicate that households were widely scattered. Again, without systematic survey, we can produce only approximate low and high densities: between 1,912 and 3,650 people. These figures more closely correspond with our estimates for Singer-Moye’s maximal size, outlined below (see also Table 3.3). Less is known about the other two single-mound sites in the valley occupied prior to AD 1300, Mandeville and Purcell’s Landing. Two additional sites, Rood’s Landing and Gary’s Fish Pond, were occupied during this time period and contain mounds, but no definitive evidence for monumental construction dating to AD 1100–1300 has been identified at either (Blitz and Lorenz 2006). However, the complete construction sequence of several mounds at Rood’s Landing has yet to be delineated. At the beginning of the c.AD 1100 to 1300 period, single-mound centers were the most common settlement form in the valley, with three known single-mound sites settled at or soon after AD 1100, and with additional centers developing after AD 1200. Between AD 1200 and 1300, mound construction appears to be concentrated at Cemochechobee, though it is unknown whether construction on the largest mounds at Rood’s Landing and Singer-Moye had commenced. At Cemochechobee, inhabitants initially constructed a single mound and added two more by AD 1300, after which time mound construction at the site ceased. Schnell et al. (1981) suggest that two of the mounds at Cemochechobee functioned as foundations for elite or ceremonial structures and the third was a burial mound. Table 3.5 Estimated population of contemporary sites in the Lower Chattahoochee River valley

Blitz (1999; see also Blitz and Lorenz 2002) has explained shifts in the intensity of population aggregation and mound construction through processes of fission and fusion. Within the fission–fusion model, the active use of a mound is thought to be associated with the efforts of local elites to consolidate authority through appeals to group unity. Conversely, a hiatus in mound use is interpreted as coinciding with the abandonment of the settlement by those same groups. Based on changes in ceramic frequencies from mound contexts, Blitz and Lorenz (2006: Table 5.1) suggest that the Singer-Moye site was abandoned between AD 1200 and 1300, in keeping with the fission–fusion process. However, recent off-mound excavations in the vicinity of Mound H, and the AMS date noted above (Table 3.3), challenge the scenario of cycles of occupation and abandonment that is documented in other regions (Anderson 1994; Beck 2013; Hally 1996, 2006; Williams and Shapiro 1990). Further excavations and additional radiocarbon assays will be required to determine the nature of this occupation and clarify whether or not the site did indeed experience a hiatus in occupation, a hiatus in mound construction, or both. Middle phase: c.AD 1300 to 1400 The middle phase of occupation represents the period of time in which Singer-Moye reached its greatest known geographic extent. Just under 19 ha (or 60.6 percent) of site area produced ceramics dating to c.AD 1300 to 1400, representing an approximate threefold increase in total site area from the previous phase (Figure 3.8). Although the area east of the site core and upstream on Pataula Creek has not been systematically surveyed, we have reason to believe these areas were occupied in a similar fashion to the area downstream of the site core, representing additional occupational areas and population. The artifact distribution beyond the site core is concentrated on areas of well-drained soils, divided by low-lying swampy floodplain suitable for agriculture. Several radiocarbon dates from numerous mound contexts fall within the middle phase of occupation (Table 3.3). Of the 19 ha of site area that were occupied during this phase, 7.2 ha (or 38 percent) were heavily occupied and 11.6 (62 percent) represent lighter occupation (Table 3.4). We stress that the total site boundaries have not yet been established and that these estimates may be adjusted as additional land is surveyed. Based on our calculations, we estimate the population during this time to be approximately 1,321 persons living within the systematically surveyed area. If our assumption that a similar residential expansion existed upstream were true, then this

tentative estimate would increase to around 2,000 inhabitants. Our largest population estimates are contemporary with the period in which the largest changes were occurring to the built environment. Within the site core, two large rectangular structures were constructed near the early phase loci, which were subsequently covered with dome-shaped mounds of earth after the structures were no longer in use, creating Mounds E and H. A platform mound (Mound F) was constructed roughly equidistant between Mounds E and H (Brannan and Bigman 2014). Mound F also demarcates the boundary between the north and southern plazas.

Figure 3.8

Singer-Moye, middle phase of occupation, c.AD 1300 to 1400.

Another platform mound, Mound D, is believed to have been constructed during this phase of occupation. A single radiocarbon date from a pit feature on Mound D suggests it was in use between cal AD 1275 and 1380 (1-sigma) (Table 3.4), though the ceramic assemblage from the mound’s terminal surface indicates a slightly later date, suggesting perhaps a use-life spanning the middle and late phases of occupation. More radiometric dates will be required to build an absolute model of mound use. The extent and nature of construction that took place on Mound A during this period is also unclear. However, its focal place in the natural and built environment of the site—dominating the landform on which the site core was constructed, anchoring the south plaza, and facing Mound D—suggests that its construction continued or began during the middle phase of occupation. All of these special purpose and monumental structures were placed around two plazas, which also appear to have been defined during this phase. The major building programs, including the expansion and definition of a town plan, are suggestive of extensive place-making, socio-political consolidation, and community-building c.AD 1300–

1400. The degree by which the built environment was modified may also reflect changes in the natural environment surrounding Singer-Moye. If people had been living at Singer-Moye continuously since it was originally settled, then their presence did not exhaust the local natural resources nor were there political ramifications to long-term settlement in a single area as suggested by Williams and Shapiro (1990:165–173). The continuing occupation of the area may have contributed to increasing the local biotic diversity (e.g. Balée and Erickson 2006; Brookfield 2001), making it either an attractive location to settle or an environmental niche, which was better able to support larger populations. More data from systematic excavations are needed in order to identify changes in biotic diversity and the presence and use of particular species through time. Within the region, between AD 1300 and 1400, the construction of monuments shifted to two large multi-mound centers, Singer-Moye and Rood’s Landing. Both centers are located north of Cemochechobee, previously the largest mound center in the valley, and are spaced approximately 28 km apart. The only other site with an in-use mound was a small settlement 80 km to the south of Rood’s Landing, at Omussee Creek. At Rood’s Landing, between five and eight mounds were in use during this period (Blitz and Lorenz 2006: Table 5.1), similar to Singer-Moye’s five mounds. Knight and Mistovich (1984:91) have tentatively identified the extent of settlement at Rood’s Landing to be approximately 64 ha. While Rood’s Landing has not been systematically surveyed, if the pattern observed at Singer-Moye holds true, then we tentatively suggest that the population at Rood’s Landing may have been between 2,006 and 3,830 people at its greatest extent (Table 3.5). Singer-Moye and Rood’s Landing are both places that were settled early in the occupational sequence of the valley. The fluorescence of monumental construction at both sites dates to this period, c.AD 1300 to 1400, in which residential populations increased substantially. We do not know for certain what drove this growth (e.g. migration, the fission–fusion process, settlement aggregation). However, it is unlikely to have been related to an increase in conflict. Neither site has revealed evidence for fortifications, nor are they defensively sited. There is a double ditch surrounding the site core at Rood’s Landing, but the village extends at least 700 meters beyond it, which suggests that the ditches served to demarcate space within the community, as opposed to them having a defensive function (Caldwell 1955). Likewise, at Singer-Moye the remains of two palisades have been identified (Kilgore et al. 2015; Russell and Gordy 2012); their relationship to each other is unclear, but their location—surrounding Mounds A and H, respectively—and lack of bastions suggest they functioned to limit access to or visibility of these parts of the site, as opposed to serving as defense against some external threat (Brannan and Birch 2015). We do, however, acknowledge that palisades enclosing chiefly precincts may have also served a defensive function, preventing easy access to ancestor temples and other symbols of chiefly authority contained within (Dye and King 2007). Similarities in ceramic assemblages, monumental architecture and village layout, and population sizes suggest significant interaction between Singer-Moye and Rood’s Landing which may have included emulation, exchange of goods and information, and possibly competition. These patterns resemble models of peer-polity interaction as defined for other medium-complex societies (Renfrew and Cherry 1986).

Late phase: c.AD 1400 to 1500 The late phase of occupation at Singer-Moye included a contraction in the total settlement size to approximately 11 ha, slightly more than two-thirds the size of the previous residential footprint. Late phase occupation is concentrated around the site core and on a small section of an adjacent landform (Figure 3.9). Ceramics recovered from the summits of two of the largest platform mounds, A and D, suggest they were in use at this time (Blitz and Lorenz 2006). Two radiocarbon dates place the terminal occupation of Mound A to between cal AD 1316 and 1430 (1-sigma), and 1494 and 1631 (1-sigma) (Table 3.3). More absolute dates will need to be acquired in order to narrow down this range. For this phase of occupation, just over half (5.90 ha) of the site’s area appears to have been heavily occupied, and just under half (5.13 ha) was lightly occupied (Table 3.4). As with the population estimates above, these estimates are based on total artifact counts from shovel-tests. Populations for the earliest and latest phases of occupation are the most difficult to estimate accurately, due to the fact that both are situated within the middle phase of occupation. Nevertheless, based on our calculations, we estimate the population during this time to be 852. However, this number may overestimate the actual number of people residing within the site core and does not take into account the potential for additional occupation beyond the survey boundaries. Notably, at Singer-Moye and in the wider region, ceramic assemblages thought to post-date AD 1400 are marked by a decrease in local ceramic traditions that had been in place for 300 years in favor of two new styles which originated outside the valley: Fort Walton and Lamar. Lamar is seen primarily as an interior cultural manifestation which was transmitted widely throughout the region while Fort Walton originated along the Gulf Coast of the southeastern United States and was transmitted into the interior via either in-migration or exchange. Both cultural traditions are similar in that they represent a continuation of Mississippian cultural practices that first appeared at the start of the second millennium AD. However, generally speaking, both Lamar and Fort Walton cultural patterns are characterized by the reorientation of interregional interaction, smaller polity sizes, and a dispersed settlement pattern as opposed to population aggregation at large mound centers.

Figure 3.9

Singer-Moye, late phase of occupation, c.AD 1400 to 1500.

It is not known if extant groups abandoned the valley, after which time it was resettled by people producing different ceramics, or if the early fifteenth century represents a period in which local populations adopted new forms of material culture whose origin lay elsewhere in the Deep South. Settlement patterns indicate that the transition started at the end of the valley closest to the point of origin for each style and then spread throughout the valley in a timetransgressive fashion (Blitz and Lorenz 2006). Within the valley itself, site distributions do not suggest that a clear boundary existed between ceramic style zones, and sites with one or both components are intermixed. This shift in ceramic traditions led Blitz and Lorenz to suggest that Singer-Moye may have been the only major multi-mound center occupied across the transition that occurred c.AD 1400. This shift may be due to a hiatus and reoccupation of the site or a macroregional reorganization or reorientation which brought new practices, including ceramic technology, into the valley and to Singer-Moye, providing the illusion of population replacement. Without more systematic excavation to recover depositional histories, we cannot confirm or deny that such an abandonment occurred or whether there was simply a hiatus in mound construction; nor can we confirm whether or not the large population which appears to have resided at SingerMoye c.AD 1300–1400 adversely affected the local environment by depleting natural resources and soil nutrients. The question of whether or not there was an occupational hiatus at either point in the site’s use-life will have important implications with regard to the shifting local settlement ecology. There is evidence that elites were in residence at Singer-Moye after AD 1400. On the summit

of Mound A, excavations revealed a single large structure, which has been interpreted as an elite residence. This structure contained artifacts that have commonly been linked to broad interregional interaction spheres, including copper, a greenstone celt, and pipe fragments. The Mound D summit had been disturbed by historic plowing, but greenstone celt and pipe fragments were also recovered from this context (Blitz and Lorenz 2006:162). Some evidence points to instability in the region at this time. At Rood’s Landing, no evidence for mound use has been identified between c.AD 1400 and 1550. Blitz and Lorenz (2006:81) believe that it was abandoned. The only other site with monumental construction occupied in this period was Gary’s Fish Pond (Blitz and Lorenz 2006:47–48), which has a long history of occupation from AD 1100 to 1450, but the single mound at the site was not constructed prior to AD 1400. The years between AD 1400 and 1450 represent a period of significant change throughout the Deep South. At this time, the large mound centers in adjacent areas were abandoned, including Etowah (King 2003) and Lake Jackson (Scarry 1999). Other regions further afield also saw significant changes: the depopulation of Moundville (Steponaitis 1998), the abandonment of the Savannah River valley (Anderson 1994), and major demographic realignments across the Carolina Piedmont (Beck 2013) and large portions of the mid-continental United States (Williams 1990). In the Georgia Piedmont during the sixteenth century, populations reorganized themselves into a pattern of dispersed farmsteads anchored by small mound centers, which were hubs for religious and political activities (Kowalewski and Hatch 1991). Though these abandonments and reorganizations must have impacted on populations in the lower Chattahoochee River valley, it is difficult to say what these effects were. This region is lacking the sort of extensive survey and excavation required to confirm whether or not populations in the valley dispersed in a similar fashion to populations in the interior, though a preliminary survey of late prehistoric site distributions tentatively suggests a shift to such a dispersed pattern, as opposed to the abandonment of the valley. Climatic instability has often been cited as one of the primary factors which may have influenced the drastic changes in socio-political and settlement systems seen across the southeastern United States at this time. It has been argued elsewhere that declining rainfall and climatic cooling were, in part, responsible for the abandonment of large mound centers as described above (Anderson 1994; Blitz and Lorenz 2006; Meeks and Anderson 2013). While microclimatic data are not available, Singer-Moye’s location on a biotically diverse ecotone may have provided it with a degree of protection from drought-induced economic scarcity. Ultimately the collapse of regional social and symbolic networks as a result of climatic and socio-political instability may have contributed to the site’s abandonment c.AD 1450/1500. However, it appears that Singer-Moye was occupied longer and remained larger than other sites in the valley, perhaps until the end of the fifteenth century.

Summary and conclusions The initial occupation of Singer-Moye, its growth during the fourteenth century, and its continued role as a population center during a period of regional settlement dispersal is due in part to the unique environmental and logistical niche it occupied. The natural and cultural

factors influencing the trajectory of Singer-Moye as a settlement seem to have waxed and waned through time. We hypothesize that Singer-Moye’s unique location a) on an ecotone and b) on a cultural frontier may have contributed to its initial development as a persistent monumental and central place in the lower Chattahoochee River valley. The population which inhabited Singer-Moye may have been able to take advantage of the wide range of resources available both in the Hilly Gulf Coastal Plain and in the Coastal Plain Red Uplands, as their numbers grew as a result of settlement aggregation and natural population growth during the thirteenth and fourteenth centuries. Continued management of the site’s immediate environs may have also made the location more attractive through anthropogenic niche formation. Singer-Moye’s location may have also been advantageous from a socio-political perspective. It was close enough to the settlements of the lower Chattahoochee River proper— particularly Cemochechobee and Rood’s Landing, as well as other, smaller mound sites—to participate in the peer-polity network centered in the valley. At the same time, it may have also had a certain degree of autonomy conferred by its location away from major transportation and interaction routes, such that when these networks collapsed, Singer-Moye may have felt the change less profoundly than more centrally located sites. Singer-Moye was a central place within the region, but not within the larger Mississippian cultural sphere in the way that Etowah or Moundville likely were. Our ongoing research at Singer-Moye will continue to be aimed at better understanding the occupational history of the site, as well as the history of human–environment interaction it encapsulates. Reconstructing the settlement ecology of other sites in the region will also enhance our understanding of the recursive, mutually constitutive relationships between these other societies, the environment, and Singer-Moye.

References Alt, Susan M. 2010 Complexity in Action(s): Retelling the Cahokia Story. In Ancient Complexities: New Perspectives in Precolumbian North America, edited by Susan M. Alt, pp. 119–137. University of Utah Press, Salt Lake City. Anderson, David G. 1994 The Savannah River Chiefdoms. University of Alabama Press, Tuscaloosa. Anderson, David G., and Kenneth E. Sassaman 2012 Recent Developments in Southeastern Archaeology: From Colonization to Complexity. The SAA Press, Washington, DC. Anschuetz, Kurt F., Richard H. Wilshusen, and Cherie L. Scheick 2001 An Archaelogy of Landscapes: Perspectives and Directions. Journal of Archaeological Research 9(2): 157–211. Balée, William A., and Clark Erickson (editors) 1998 Introduction. In Advances in Historical Ecology, edited by William Balée, pp. 1–10. Columbia University Press, New York. Balée, William A., and Clark Erickson (editors) 2006 Time and Complexity in Historical Ecology: Studies in the Neotropical Lowlands. Columbia University Press, New York. Beck, Robin A. Jr. 2013 Chiefdoms, Collapse, and Coalescence in the Early American South. Cambridge University Press, Cambridge. Birch, Jennifer, and Ron Williamson 2015 Navigating Ancestral Landscapes in the Northern Iroquoian World. Journal of Anthropological Archaeology 39: 139–150. Blitz, John H. 1999 Mississippian Chiefdoms and the Fission–Fusion Process. American Antiquity 64(4): 577–592. Blitz, John H. 2010 New Perspectives in Mississippian Archaeology. Journal of Archaeological Research 18: 1–39. Blitz, John H., and Karl G. Lorenz 2002 The Early Mississippian Frontier in the Lower Chattahoochee-Apalachicola River Valley. Southeastern Archaeology 21(2): 117–135. Blitz, John H., and Karl G. Lorenz 2006 The Chattahoochee Chiefdoms. University of Alabama Press, Tuscaloosa. Brannan, Stefan 2016 The Settlement Archaeology of Singer-Moye, a Large 14th-Century Town in the Chattahoochee Valley, Georgia. PhD dissertation. Department of Anthropology, University of Georgia, Athens. Brannan, Stefan, and Daniel P. Bigman 2012 Do Mississippian Plazas Represent Open Spaces or Rich Histories? Paper

presented at the 69th Annual Meeting of the Southeastern Archaeological Conference, Baton Rouge, LA. Brannan, Stefan, and Daniel P. Bigman 2014 Ground Penetrating Radar and Resistivity Results from Mounds D and F at Singer-Moye (9SW2). Early Georgia 42(2): 179–192. Brannan, Stefan, and Jennifer Birch 2015 Palisaded Enclosures and Political Complexity. Paper presented at the 80th Annual Meeting of the Society for American Archaeology, San Francisco, CA. Bronk Ramsay, Christopher 2009 Bayesian Analysis of Radiocarbon Dates. Radiocarbon 51(1): 337–360. Brookfield, Harold 2001 Exploring Agrodiversity. Columbia University Press, New York. Byers, A. Martin 2006 Cahokia: A World Renewal Cult Heterarchy. University Press of Florida, Gainesville. Caldwell, Joseph 1955 Investigations at Rood’s Landing, Stewart County, Georgia. Early Georgia 2(1): 22–49. Casselberry, Samuel E. 1974 Further Refinement of Formulae for Determining Population from Floor Area. World Archaeology 6(1): 117–122. Chamblee, John F. 2006 Landscape Patches, Macroregional Exchanges and Pre-Columbian Political Economy in Southwestern Georgia. Ph.D. dissertation. Department of Anthropology, University of Arizona, Tucson. Charles, Douglas K., and Jane E. Buikstra 1983 Archaic Mortuary Sites in the Central Mississippi Drainage: Distribution, Structure, and Implication. In Archaic Hunters and Gatherers in the Midwest, edited by James L. Phillips and James A. Brown, pp. 117–145. Academic Press, New York. Clayton, Lawrence, A., Vernon James Knight Jr., and Edward C. Moore (editors) 1993 The De Soto Chronicles: The Expedition of Hernando de Soto to North America in 1539–1543. 2 vols. University of Alabama Press, Tuscaloosa. Cobb, Charles R. 2003 Mississippian Chiefdoms: How Complex? Annual Review of Anthropology 32: 63–84. Cobb, Charles R. 2005 Archaeology and the “Savage Slot”: Displacement and Emplacement in the Premodern World. American Anthropologist 107(4): 563–574. Cobb, Charles R. 2014 What I Believe: A Memoir of Processualism to Neohistorical Anthropology. Southeastern Archaeology 33(2): 214–225. Cobb, Charles R., and Adam King 2005 Re-Inventing Mississippian Tradition at Etowah, Georgia. Journal of Archaeological Method and Theory 12(3): 167–193. Crumley, Carol L. 2007 Historical Ecology: Integrated Thinking at Multiple Temporal and Spatial Scales. In The World System and the Earth System: Global Environmental Change and Sustainability Since the Neolithic, edited by Alf Hornborg and Carol L. Crumley, pp. 15–28. Left Coast Press, Walnut Creek, CA. Davis, Jera R., Chester P. Walker, and John H. Blitz 2015 Remote Sensing as Community Settlement Analysis at Moundville. American Antiquity 80(1): 161–169. DePratter, Chester B. 1991 Late Prehistoric and Early Historic Chiefdoms in the Southeastern United States. Garland, New York. Dye, David H., and Adam King 2007 Desecrating the Sacred Ancestor Temples: Chiefly Conflict and Violence in the American Southeast. In North American Indigenous Warfare and Ritual Violence, edited by Richard J. Chacon and Reuben G. Mendoza, pp. 160–181. University of Arizona Press, Tucson. Elliott, Daniel T., and Tracy M. Dean 2006 Flint River Basin Archaeological Survey, Phase 2. LAMAR Institute Report Series, No. 77. Electronic document,, accessed Mar 20, 2015 Georgia Archaeological Site Files (GASF) 2015 Georgia State Site Files. Online Database,, accessed May 10, 2015. Griffith, G.E., J.M. Omernik, J.A. Comstock, S. Lawrence, G. Martin, A. Goddard, V.J. Hulcher, and T. Foster 2001 Ecoregions of Alabama and Georgia (color poster with map, descriptive text, summary tables, and photographs). Reston, Virginia, U.S. Geological Survey (map scale 1:1,700,000). Hally, David J. 1993 The Territorial Size of Mississippian Chiefdoms. In Archaeology of Eastern North America: Papers in Honor of Stephen Williams, edited by James B. Stoltman, pp. 143–167. Mississippi Department of Archives and History, Archaeological Report No. 25, Jackson, MS. Hally, David J. 1996 Platform Mound Construction and the Instability of Mississippian Chiefdoms. In Political Structures and Change in the Prehistoric Southeastern United States, edited by John F. Scarry, pp. 92–127. University Press of Florida, Gainesville. Hally, David J. 2006 The Nature of Mississippian Regional Systems. In Light on the Path: The Anthropology and History of the Southeastern Indians, edited by Thomas J. Pluckhahn and Robbie Ethridge, pp. 26–42. University of Alabama Press, Tuscaloosa. Hudson, Charles 1997 Knights of Spain, Warriors of the Sun: Hernando de Soto and the South’s Ancient Chiefdoms. University of Georgia Press, Athens. Jenkins, Ned J. 2009 Tracing the Origins of the Early Creeks. In Mapping the Mississippian Shatter Zone: The Colonial Indian Slave Trade and Regional Instability in the American South, edited by Robbie Ethridge and Sherri M. Shuck-Hall. University of Nebraska Press, Lincoln. Jones, Eric E., and James W. Wood 2012 Using Event-History Analysis to Examine the Causes of Semi-Sedentism Among Shifting Cultivators: A Case Study of the Haudenosaunee, AD 1500–1700. Journal of Archaeological Science 39: 2593–

2603. Kidder, Tristam R. 2004 Plazas as Architecture: An Example from the Raffman Site, Northeast Louisiana. American Antiquity 69: 514–532. Kilgore, Eli, Emily E. Lew, Justin N. Lynch, Adam C.S. Kazmi, and Jennifer Birch 2015 Palisades and the Segmentation of Space at Singer-Moye. Poster presented at the 72nd Annual Meeting of the Southeastern Archaeological Conference, Nashville. King, Adam 2003 Etowah: The Political History of a Chiefdom Capital. University of Alabama Press, Tuscaloosa. Knight, Vernon J. Jr. 1979 Ceramic Stratigraphy at the Singer-Moye Site, 9Su2. Journal of Alabama Archaeology 25(2): 138– 151. Knight, Vernon J. Jr. 1986 The Institutional Organization of Mississippian Religion. American Antiquity 51(4): 675–687. Knight, Vernon J. Jr. 2010 Mound Excavations at Moundville: Architecture, Elites, and Social Order. University of Alabama Press, Tuscaloosa. Knight, Vernon J. Jr., and Tim S. Mistovich 1984 Walter F. George Lake Archaeological Survey of Free Owned Lands Alabama and Georgia. Office of Archaeological Research, University of Alabama. Submitted to United States Army Corps of Engineers, Contract No. DACW01-83-C-0173. Knight, Vernon J. Jr., and Vincas P. Steponaitis (editors) 1998 Archaeology of the Moundville Chiefdom. Smithsonian Institution Press, Washington, DC. Kowalewski, Stephen A. 2013 The Work of Making Community. In From Prehistoric Villages to Cities: Settlement Aggregation and Community Transformation, edited by Jennifer Birch, pp. 201–218. Routledge, New York. Kowalewski, Stephen A., and James W. Hatch 1991 The Sixteenth-Century Expansion of Settlement in the Upper Oconee Watershed, Georgia. Southeastern Archaeology 10(1): 1–17. Lindauer, Owen, and John H. Blitz 1997 Higher Ground: The Archaeology of North American Platform Mounds. Journal of Archaeological Research 5: 169–207. Little, Maran 2013 Unpublished field notes on the faunal analysis of the 2013 Field Season at Singer-Moye. Manuscript on file, the University of Georgia Laboratory of Archaeology, Athens. Marquardt, William H. 1992 Dialectical Archaeology. In Archaeological Method and Theory, Vol. 4, edited by Michael B. Schiffer, pp. 101–140. University of Arizona Press, Tucson. Marquardt, William H. 2013 The Pineland Site Complex: Theoretical and Cultural Contexts. In The Archaeology of Pineland: A Coastal Southwest Florida Site Complex, AD 50–1710, edited by William H. Marquardt and Karen J. Walker, pp. 1–22. University Press of Florida, Gainesville. Marquardt, William H., and Carol L. Crumley 1987 Theoretical Issues in the Analysis of Spatial Patterning. In Regional Dynamics: Burgundian Landscapes in Historical Perspective, edited by Carol L. Crumley and William H. Marquardt, pp. 1–18. Academic Press, San Diego. Meeks, Scott C., and David G. Anderson 2013 Drought, Subsistence Stress, and Population Dynamics: Assessing Mississippian Abandonment of the Vacant Quarter. In Soils, Climate & Society, edited by John D. Wingard and Sue E. Hayes, pp. 61–84. University Press of Colorado, Boulder. Milner, George R. 2000 Palisaded Settlements in Prehistoric Eastern North America. In City Walls, edited by James D. Tracy, pp. 46–70. Cambridge University Press, Cambridge. Milner, George R. 2012 Mound-Building Societies of the Southern Midwest and Southeast. In The Oxford Handbook of North American Archaeology, edited by Timothy R. Pauketat, pp. 437–447. Oxford University Press, Oxford. Muller, Jon 1989 The Southern Cult. In The Southeastern Ceremonial Complex: Artifacts and Analysis, edited by Patricia Galloway, pp. 11–26. University of Nebraska Press, Lincoln. Pauketat, Timothy 2007 Chiefdoms and Other Archaeological Delusions. AltaMira Press, Plymouth, UK. Pauketat, Timothy 2013 Bundles of/in/as Time. In Big Histories, Human Lives: Tackling Problems of Scale in Archaeology, edited by John Robb and Timothy R. Pauketat, pp. 35–56. School for Advanced Research Press, Santa Fe. Pauketat, Timothy R., and Susan M. Alt 2005 Agency in a Postmold? Physicality and the Archaeology of Culture-Making. Journal of Archaeological Method and Theory 12(3): 213–237. Pauketat, Timothy R., and Thomas E. Emerson 1991 The Ideology of Authority and the Power of the Pot. American Anthropologist 93(4): 919–941. Pauketat, Timothy R., and Thomas E. Emerson 1997 Cahokia: Domination and Ideology in the Mississippian World. University of Nebraska Press, Lincoln. Peebles, Christopher S., and Susan M. Kus 1977 Some Archaeological Correlates of Ranked Societies. American Antiquity 42(3): 421–448. Pluckhahn, Thomas J. 2010 Household Archaeology in the Southeastern United States: History, Trends, and Challenges. Journal of Archaeological Research 18(4): 331–385. Regnier, Amanda L. 2014 Reconstructing Tascalusa’s Chiefdom: Pottery Styles and the Social Composition of Late Mississippian Communities along the Alabama River. University of Alabama Press, Tuscaloosa. Reimer, Paula J., Edouard Bard, Alex Bayliss, J. Warren Beck, Paul G. Blackwell, Christopher Bronk Ramsey, Caitlin E. Buck,

Hai Cheng, R. Lawrence Edwards, Michael Friedrich, Pieter M. Grootes, Thomas P. Guilderson, Haflidi Haflidason, Irka Hajdas, Christine Hatté, Timothy J. Heaton, Dirk L. Hoffmann, Alan G. Hogg, Konrad A. Hughen, K. Felix Kaiser, Bernd Kromer, Sturt W. Manning, Mu Niu, Ron W. Reimer, David A. Richards, E. Marian Scott, John R. Southon, Richard A. Staff, Christian S. M. Turney, and Johannes van der Plicht 2013 IntCal13 and Marine13 Radiocarbon Age Calibration Curves 0–50,000 Years cal BP. Radiocarbon 55(4): 1869–1887. Renfrew, Colin A., and John F. Cherry (editors) 1986 Peer Polity Interaction and Socio-Political Change. Cambridge University Press, Cambridge. Russell, Margaret C. and R. Donald Gordy 2012 The Archaeology of Mound H, Singer-Moye Mound Center, Stewart County, Georgia. Early Georgia 40(2): 127–154. Scarry, C. Margaret 1993 Variability in Mississippian Crop Production Strategies. In Foraging and Farming in the Eastern Woodlands, edited by C. Margaret Scarry, pp. 78–90. University Press of Florida, Gainesville. Scarry, C. Margaret, and John F. Scarry 2005 Native American “Garden Agriculture” in Southeastern North America. World Archaeology 37(2): 259–274. Scarry, John F. 1999 How Great Were the Southeastern Polities? In Great Towns and Regional Polities in the Prehistoric American Southwest and Southeast, edited by Jill E. Neitzel, pp. 59–74. University of New Mexico Press, Albuquerque. Schlanger, Sarah H. 1992 Recognizing Persistent Places in Anasazi. In Space, Time, and Archaeological Landscapes, edited by Jacqueline Rossignol and LuAnn Wandsnider, pp. 91–113. Plenum Press, New York. Schnell, Frank T., and Newell O. Wright Jr. 1993 Mississippi Period Archaeology of the Georgia Coastal Plain. Laboratory of Archaeology Series No. 26. University of Georgia, Athens. Schnell, Frank T., Vernon J. Knight Jr., and Gail S. Schnell 1981 Cemochechobee: Archaeology of a Mississippian Ceremonial Center on the Chattahoochee River. University Presses of Florida, Gainesville. Smith, Bruce D. 1989 Origins of Agriculture in Eastern North America. Science 246: 1566–1571. Snow, Dean 1995 Microchronology and Demographic Evidence Relating to the Size of Pre-Columbian North American Indian Populations. Science 268: 1601–1604. Steere, Benjamin A. 2011 The Archaeology of Houses and Households in the Native Southeast. Unpublished Ph.D. dissertation. Department of Anthropology, University of Georgia, Athens. Steponaitis, Vincas P. 1998 Population Trends at Moundville. In Archaeology of the Moundville Chiefdom, edited by Vernon J. Knight, Jr. and Vincas P. Steponaitis, pp. 26–43. Smithsonian Institution Press, Washington, DC. Steponaitis, Vincas P. 2009 Ceramics, Chronology, and Community Patterns: An Archaeological Study at Moundville. Reprinted. University of Alabama Press, Tuscaloosa. Originally published 1983, Academic Press, New York. Stone, Glenn D. 1996 Settlement Ecology: The Social and Spatial Organization of Kofyar Agriculture. University of Arizona Press, Tuscon. Thompson, Victor D., and Thomas J. Pluckhahn 2012 Monumentalization and Ritual Landscapes at Fort Center in the Lake Okeechobee Basin of South Florida. Journal of Anthropological Archaeology 31: 49–65. Thompson, Victor D., and James C. Waggoner 2013 The Archaeology and Historical Ecology of Small Scale Economies. University Press of Florida, Gainesville. Walker, Chester P. 2009 Landscape Archaeogeophysics: A Study of Magnetometer Surveys from Etowah (9BR1), The George C. Davis Site (41CE19), and the Hill Farm Site (41BW169). Ph.D. dissertation. Department of Anthropology, University of Texas, Austin. Warrick, Gary 2008 A Population History of the Huron-Petun, AD 500-1650. Cambridge University Press, Cambridge. Wharton, Charles H. 1978 The Natural Environments of Georgia. Georgia Department of Natural Resources, Atlanta. Williams, Mark, and Gary Shapiro 1990 Paired Towns. In Lamar Archaeology: Mississippian Chiefdoms in the Deep South, edited by Mark Williams and Gary Shapiro, pp. 163–174. University of Alabama Press, Tuscaloosa. Williams, Stephen 1990 The Vacant Quarter and Other Late Events in the Lower Valley. In Towns and Temples along the Mississippi, edited by David H. Dye and Cheryl A. Cox, pp. 170–180. University of Alabama Press, Tuscaloosa. Worth, John E. 1988 Mississippian Occupation on the Middle Flint River. Master’s thesis. Department of Anthropology, University of Georgia, Athens.

4 Settlement ecology in the precontact North American Southwest Scott E. Ingram The influence of demographic, environmental, and climatic conditions on human decisionmaking can be identified with a settlement ecology approach. The decisions of interest in this chapter are those to remain in place or move. Archaeological data on residential and settlement abandonment are indicators of these past decisions. The methods described in this chapter are new and can be applied in any region with a detailed spatial and temporal settlement, and climate history. Understanding the conditions that influence settlement patterns has a long and analytically productive history in the Americas. As settlement and paleoclimatic data continue to accumulate worldwide, methodological advancements must also continue. Otherwise, the challenges and frustrations of identifying generalizable results, and of separating correlation from causation when settlement pattern and climatic changes co-occur, will persist. The research question asked in the case study presented is: what specific demographic and environmental conditions influenced human decisions to move or remain in place during droughts? The question is addressed with a multi-scalar, comparative, correlation analysis that investigates the long-term (250-year) relationship between changes in drought severity and variation in patterns of settlement abandonment. Here, archaeologically identified residential and settlement abandonment in the Southwest is defined as a record of population movements, not a record of the relinquishment of places or ownership, or the disappearance of a people (Nelson and Schachner 2002: 169). Multiple droughts and repeated periods of settlement abandonment occur during the 1200 to 1450 ce study period in the American Southwest. The long-term aspect of this analysis is essential. If throughout the 250-year period, changes in drought severity were strongly associated with proportionate changes in abandonment, then drought influences on this abandonment may be concluded. Comparing the strength of these relationships among settlements with different demographic and environmental conditions reveals the influence of these conditions on human decisions to remain in place or move to a new location. Archaeologists in the North American Southwest (and likely other arid regions) often rely on common-sense notions and unverified conceptual models to interpret the shifting settlement patterns we observe in the past. For example, in dry climates it is reasonable to expect that agriculturalists living distant from perennial rivers, or in areas of relatively low precipitation, will be most likely to move during drought conditions because they have the least access to reliable water sources. This expectation is logical and intuitively appealing but often lacks systematic empirical scrutiny. Does a long-term settlement history in an arid region demonstrate this relationship between access to water and decisions to move during droughts?

Similarly, what was the influence of settlement or regional-scale population levels on decisions to move or stay in place during droughts? Knowing the answers to these questions (and those similar) is important for current planners and policymakers seeking to prepare for anthropogenic climate change as well as archaeologists attempting to interpret the past. The conditions that influenced human decisionmaking during extreme climate conditions, now or in the past, are not well understood and in some cases are still in dispute (see Cutter et al. 2003; Knight and Jäger 2009; Meyer et al. 1998: 238–243; see Ribot 1995 for discussion of human vulnerability to recent natural hazards). Archaeological results, if they are effectively communicated beyond archaeology, could stimulate modern drought planners and those investigating the environmental dimensions of human migration worldwide (e.g. Fussell et al. 2014) to consider whether similar behavioral patterns are identifiable in dry regions occupied today by millions of smallholder farmers (Ingram 2015: 144). Understanding the conditions influencing settlement location decisions is a window into human decision-making, migration, and human vulnerability to a wide range of conditions. Given the broad audience of this volume, the primary intent in the limited space here is to illustrate the potential of the methods. Researchers interested in the influence of other types of climate extremes (e.g. flooding, warm and cool periods, and periods of high/low temporal and spatial variability) on different human decisions (e.g. settlement founding, aggregation, dispersal) should find the proposed methods widely applicable. Likewise, I present only representative results rather than document the range of actual results or argue the culturalhistorical insights and theoretical challenges these results imply in the Southwest (for such a discussion, see Ingram 2010, 2015).

Study area The central Arizona study area (Figure 4.1, all shaded polygons) includes the low and hot Sonoran desert in the south, a transition zone north to the Colorado Plateau, and the cooler and wetter high mountains of eastern Arizona (Fish and Nabhan 1991; Turner and Brown 1982; Whittlesey and Ciolek-Torrello 1997). Populations throughout the study area were widely distributed throughout this landscape along perennial rivers, intermittent streams, ephemeral washes, and among mountains and mesas distant from perennial rivers. Precipitation levels varied among settlements and watersheds based largely on differences in elevation, topography, and location. Settlements during the period of study were located in areas that historically (c.1900 to 2000 ce) receive an annual average of 8–10 in of precipitation (Sheppard et al. 2002; Western Regional Climate Center 2010). Paleoclimate data from the study area document highly variable rainfall and temperature conditions, and recurrent and persistent drought (Salzer and Kipfmueller 2005). As a result, like many arid and semi-arid regions, the Southwest is considered a “fragile and marginal environment for agriculture” (Diamond 2005: 137) and a “harsh and variable region” (Dean et al. 1994: 86). The people of central Arizona during the period of study were a culturally diverse amalgamation of long-term residents with ties to the Hohokam irrigators of the Lower Salt River Valley (modern-day Phoenix, Arizona) and recent immigrants from the northern

Southwest (Ingram 2014: 31–33 and references therein). Central Arizona farmers cultivated maize (rain-fed and irrigated) and hunted and gathered a wide variety of animals and plants (Cordell 1997; Reid and Whittlesey 1997). Social conditions during the period of study are characterized by dramatic settlement changes including population increases, settlement aggregation, and regional depopulations (Cordell 1997: 365–441). Climatic conditions are among the many factors considered to have influenced these dramatic social changes in central Arizona (e.g. Hill et al. 2004, 2010; Ingram 2010, 2012; Van West and Altschul 1997; Van West and Dean 2000).

Figure 4.1

Study area of watersheds in Central Arizona.

Models, mechanisms, and expectations The drought–population movement model described below requires a thorough investigation of the influence of demographic and environmental conditions on decisions to move or remain in place during droughts. It is important to rely on a regionally appropriate climate–human

behavioral model and identify the mechanisms whereby the independent variables (specific demographic and environmental conditions) might influence the dependent variable (settlement abandonment). Otherwise, settlement pattern studies risk identifying spurious space–time correlations between variables. Identifying influential demographic and environmental conditions can explain shifting settlement patterns and document how people may have managed the endemic risk of food shortfalls in dry climates, such as the American Southwest. The drought-population movement model used in this study (and many other dry lands) links drought-related declines in resource productivity to human population movements (migration). Droughts are expected to stimulate these movements by causing declines in the productivity of resources used for food that trigger subsequent human responses to lessen the risk of food shortfalls (Ingram 2010: 13–22). Droughts decrease resource productivity (wild and cultivated) in arid lands such as the Southwest because water is a primary limiting factor on plant growth (Fischer and Turner 1978) and precipitation levels are typically below the moisture requirements of most cultivated crops such as maize (Muenchrath and Salvador 1995; Shaw 1977). Changes in climate that influence plant growth also impact the animals that rely on plant foods (Bright and Hervert 2005; Osborn 1993) and the humans that rely on these wild herbivores. Population movement allows people to take advantage of the spatial and temporal structure of drought-related resource failure across an arid landscape. People can move away from areas of food scarcity and low productivity to areas of higher productivity (Halstead and O’Shea 1989), thus reducing shortfall risks. Ample ethnohistoric evidence in the Southwest has documented residential abandonment in response to climate-related resource shortfalls (Abruzzi 1989; Slatter 1979). Farmers use a variety of strategies to manage climate-related risks of shortfalls including food storage, trade and exchange, diet diversity, etc. The focus of this study is, however, population movement, archaeologically detectable as residential and settlement abandonment. Studies of the relationship between changes in climatic conditions and residential abandonment are common in the Southwest (e.g. Ahlstrom et al. 1995; Gumerman 1988; Kohler et al. 2010; Van West and Dean 2000, to name only a few). Drought–population movement models rely on an assumption of resource marginality in dry climates. Resource marginality, by definition, occurs where and when resource productivity is low relative to actual or perceived human food needs. In other words, marginality is an inherent condition where this productivity oscillates around a threshold above which there is enough culturally defined food to meet needs and below which there is not. Where resources are marginal, changes in any condition that increases the demand or decreases the supply of resources can increase the risk of food shortfalls and motivate a behavioral response (e.g. population movement). For example, people living in settlement areas of relatively low inherent productivity (e.g. in an area of low precipitation) should be more likely to move during droughts than people living in areas of relatively high inherent productivity. This is because people living in areas with the lowest productivity are assumed to be closest to the threshold above which there is enough food to eat and below which there is not. Thus, when droughts decrease productivity, the risk of shortfalls is assumed to increase in all areas but be most meaningful in areas of low productivity where resource supplies may have declined below a threshold, so that there is not enough food to eat.

Demographic conditions at many spatial scales (e.g. settlement, watershed, region) affect the demand for resources, the rate of consumption of resources, and the extent of labor available to invest in strategies to manage drought-related declines in resource productivity. The demographic condition considered in this chapter is areal population density. High population density in an area can increase the risk of resource shortfalls by limiting mobility as a strategy to manage these risks (Binford 1983; Dean et al. 1994: 85; Minnis 1996; Powell 1988). Population density can limit mobility if productive locations are already claimed, occupied, or hostilities restrict movement. Larger populations might also increase the tilling of relatively less productive plots of land to increase production (Nelson et al. 1994: 61–62) and these plots could quickly become unproductive during droughts. Thus, the hypothesis tested in this study is: people living in areas of high population density will be more likely to move during drought conditions than people living in areas of low population density. Environmental conditions affect the potential productivity of settlement locations and the supply of food resources to which people have access. Access to resources is one factor that affects the ability of people to adapt to and cope with drought conditions. The environmental condition considered in this chapter is a settlement’s location relative to a perennial river. In dry climates, settlement locations adjacent to perennial riverine resources, which include the associated riparian and aquatic resources, are expected to offer greater potential productivity than settlements located distant from perennial riverine resources. Agricultural potential is also greater in riverine areas where irrigated and floodplain agriculture is possible. People living distant from perennial rivers rely on “dry-farming” or “rain-fed” farming and are widely assumed to be among the most vulnerable and sensitive to low precipitation in dry climates (e.g. Liverman and Merideth 2002: 207). Thus, the hypothesis tested in this study is: people living in settlements near perennial rivers were less likely to move during drought conditions than people living in areas distant from perennial rivers. To illustrate the methods in this chapter, I describe the influence of only one demographic condition (areal population density) and one environmental condition (a settlement’s proximity to a perennial river) on decisions to move or remain in place during droughts. To understand settlement location decisions, the influence of many different demographic and environmental conditions (and social conditions) on decisions to move or remain in place must be considered (e.g. Ingram 2010). Investigations of combinations of these conditions, such as the influence of drought severity on people living in high-density watersheds near perennial rivers, will be the most productive. Such analyses allow archaeologists to test carrying capacity and population pressure models. They also provide a nuanced understanding of settlement patterns, a culture history, and advance theoretical understanding of climate–human behavior relationships within a region. For archaeological settlement pattern studies in other regions with different environments and subsistence strategies, other models and assumptions must be relied upon. For example, where resources are not marginal (abundant), changes in demographic and environmental conditions may be decoupled from changes in the drought-related risk of food shortfalls and associated behavioral responses. In other words, where food is abundant, increases in demand or decreases in the supply cannot be directly linked to changes in shortfall risks and the decisions to move or remain in place during drought conditions.

Data and methods Data on demography, residential abandonment, environment, climate, and droughts are required for this investigation. Methods are necessary for evaluating and comparing the statistical relationship between these variables. These comparisons enable the identification of conditions that influenced human decisions to move or remain in place. Space limitations do not allow me to thoroughly argue each aspect of the analysis, the range of variables I have investigated with this approach, the strengths and weaknesses of the data selected, and the limitations of the choices I have made (but see Ingram 2010, 2014 for a full description and argument). Instead, I briefly describe the data and methods so that the approach can be generally understood and attempted in other regions where data, variables, and research questions will differ from those presented here. Demography and spatial units of analysis The case study relies on the settlement history of 440 settlements in five watersheds in central Arizona during the 1200 to 1450 ce period. The data were derived from the Coalescent Communities Database (Wilcox et al. 2003; see Wilcox et al. 2007 for a description of the development of the database). The database contains settlement founding and abandonment dates for all identified settlements in the region. The number of identified or inferred rooms are recorded for each settlement, and in some cases variations within settlements in individual room founding and abandonment dates are also available. I follow Hill et al. (2004: 693) and the Coalescent Communities Database authors (Wilcox et al. 2003) and consider only settlements with at least thirteen rooms. Data on settlements with fewer than thirteen rooms are less complete and reliable due to reduced surface visibility and detection. All settlement data are georeferenced and suitable for spatial analysis within a geographic information system (GIS). This database is the most comprehensive source of settlement data available for the central Arizona study area. Four spatial units of analysis are used in this study: 1) individual rooms, 2) settlements, 3) rooms and settlements within watersheds, and 4) rooms and settlements within the entire central Arizona study area. The individual room scale documents the founding and abandonment dates of all rooms (identified and inferred). The settlement scale documents the founding and abandonment dates of all identified settlements. The watershed scale of analysis aggregates all rooms and settlements within a watershed into groups of proximate settlements and similar resource acquisition zones. I use the smallest watershed units (“cataloging units” or “sub-basins”) identified by the U.S. Geological Survey. The total study area scale aggregates all identified rooms and settlements in the central Arizona study area. This multi-scalar approach captures intraregional demographic variation during five 50-year intervals from 1200 to 1450. The 1200 start date for the study and the 50-year intervals (e.g. 1200 to 1249, etc.) used to identify residential abandonment are based on the strengths of the data and the realities of chronological resolution in the region (Hill et al. 2004). By 1450, archaeological signatures of settlement cease, and the entire study area is assumed to have been depopulated. This multi-scalar, long-term approach attempts to separate the signal

(variables with long-term influences on settlement location decisions throughout the study area) from the noise (variables and factors that were unique to individuals or particular communities). There is no reason to expect uniform human decision-making or sensitivity to droughts given the social and environmental diversity of the region. Thus, identifying patterns in such decision-making will suggest a generalizable result worth considering in other arid regions. Assessing settlement abandonment Variation in settlement abandonment is the dependent variable and an indicator of human decision-making within the study area. It is identified by using the settlement founding and abandonment dates and the number of rooms identified at each settlement. For example, a settlement occupied from 1250 to 1349 with 100 identified rooms would be counted during the 1250 to 1299 and 1300 to 1349 intervals. The number of rooms abandoned during a 50-year interval includes all rooms in settlements with abandonment dates during that interval. Thus, the 100 rooms would also be included in the rooms abandoned count during the 1300 to 1349 interval. These data are calculated at the watershed and total study area scale. The settlement and room abandonment dates in the database were determined using standard archaeological dating procedures (i.e. the presence of well-dated ceramic types and/or radiocarbon dating). The index of residential abandonment used for each 50-year interval from 1200 to 1450 is: total number of rooms abandoned divided by the total number of rooms occupied: thus, the percent of rooms abandoned during each interval. For example, in the central Arizona study area as a whole during the 1250 to 1299 interval, 10,163 rooms were occupied and 4,699 were abandoned. The index of residential abandonment, then, is: 4,699 divided by 10,163 = 46 percent (the percentage of rooms abandoned). I do not consider absolute changes in the total number of rooms abandoned during each interval as absolute changes will vary as total population varies. Using the percentage of rooms abandoned controls for changes in population size. Environment Environmental information for each room and settlement can be assigned or calculated using layers in ArcMap software. Elevations, soil types, vegetation classifications, growing season durations, and many other variables (and combinations of these variables) of interest can be associated with each settlement, depending on the research question. These layers of environmental data are indicators of differences in potential resource productivity among settlements. In the example presented in this chapter, I use a modern hydrographic (streams and rivers) layer to calculate the distance of each settlement to the nearest perennial river. Modern data are appropriate because perennial streams and rivers currently visible were also present during the study period. Climate and droughts I use the San Francisco Peaks (SFP) tree-ring reconstruction of annual precipitation (previous

October to current July) from 570 to 1988 ce to identify droughts in the precipitation record of central Arizona (Figure 4.2). Salzer (2000) and Salzer and Kipfmueller (2005) developed this reconstruction from both living tree and archaeological chronologies with standard procedures developed at the Laboratory of Tree-Ring Research at the University of Arizona (e.g. Fritts 1976; Rose et al. 1981). I identify droughts in the SFP precipitation reconstruction in two steps. First, I smooth year-to-year variation in the reconstruction with a centered nine-yearinterval moving average. Second, I select a threshold value to identify droughts within the reconstruction. The threshold I use is the first quartile value (25th percentile) of nine-yearinterval averages. All intervals and years in the first quartile are classified as drought years. Elsewhere, I have statistically evaluated and argued the suitability of the SFP reconstruction to represent precipitation conditions in central Arizona and have fully described and argued my methods for identifying droughts within a precipitation reconstruction (Ingram 2010, 2014).

Figure 4.2

Droughts in central Arizona from 1200 to 1450.

Table 4.1

Droughts and the drought severity index



Drought duration in years

Total number of Drought severity index drought years (percentage of interval identified in interval as a drought)


































Note: 1 To accommodate this drought, the number of drought years in the 1400 to 1449 interval is divided by a 63-year interval duration (i.e. 1400 to 1462). I use the end date of 1462 because it is a reasonable guess at the termination date of this interval in central Arizona prehistory, which is not well known (see for example, Dean 1991).

Annually resolved precipitation and drought data must be matched for comparative purposes to the coarser 50-year resolution of the settlement data. My solution to this problem involves creating an index of drought severity for each 50-year interval (Table 4.1) Total number of drought years ini Interva). I calculate this index by summing the number of drought years in each 50-year interval and dividing this sum by the 50-year interval duration. I use this percentage to represent differences in drought severity among the five 50-year intervals. Intervals can then be compared on the basis of this severity value. A drought that overlaps interval boundaries is counted in the interval containing the majority of the drought’s years, when human responses are most expected. Creating an index of drought severity and making relative comparisons between intervals is a reasonable approach because it is consistent with the strengths of tree-ring reconstructed precipitation data. These data are the strongest and most reliable when they are used to represent relative changes in climate conditions rather than absolute changes (i.e. year-to-year changes) (Fritts 1976; Meko et al. 1995). Identifying conditions that influenced decision-making during dry conditions This section has, to this point, described the methods used to identify differences in drought severity during the 250-year study period and how each room, settlement, and watershed varied (and is classified) by demographic and environmental conditions. Now, identifying specific demographic and environmental conditions that influenced human decisions to move or remain in place during droughts can be pursued by examining the statistical relationship between drought severity and residential abandonment under different demographic and environmental conditions. If the empirical relationships are substantially different as conditions are varied (e.g. in areas of high vs. low population density, in settlements near vs. far from perennial rivers), then these variations signal the influence of the conditions on decisions to move or remain in place during droughts. Little or no difference in the relationships will indicate a particular condition did not affect decisions to move away from settlements. These relationships can be examined at multiple spatial scales (e.g. among watersheds, along rivers, among clusters of settlements, etc.) to verify the extent of the influence of specific conditions. Statistical analysis I assess the empirical relationship between drought severity and residential abandonment (in all places and under each demographic and environmental condition) with the following methods. 1. Linear regression to identify a best-fit straight line through scatterplots of data values representing drought severity (x) and residential abandonment (y) during each 50-year interval from 1200 to 1450. Non-linear regression and other relationships could be, but are not, considered in this investigation (Ingram 2010: 117–119). 2. The best-fit values of the slope and intercept based on the regression equation and line.

The slope identifies the direction of the relationship (positive or negative) and quantifies the steepness of the line. The slope in this analysis equals the percentage change in residential abandonment (y) for each 1 percent change in drought severity (x). I use the slope to evaluate the extent of sensitivity to drought severity in different places and under different demographic and environmental conditions. For example, if people living in settlements near perennial rivers were less sensitive (or responsive, with residential abandonment) to drought severity than people living far from perennial rivers, then I expect the slope of the line representing the relationship between drought severity and residential abandonment among settlements near perennial rivers will be shallower (less steep) than the slope of the line representing the relationship between drought severity and residential abandonment among settlements far from perennial rivers. Thus, differences in the slopes of the lines are used to determine differences in sensitivity to drought severity among settlements with particular characteristics. The regression equation and line are also used to identify the y-intercept. The intercept identifies the elevation of the line and is the expected mean value of y when x = 0; or, the extent of residential abandonment when the index of drought severity is 0. It suggests a starting value or constant rate of residential abandonment unaffected by droughts. 3. A Pearson’s r correlation coefficient and associated p-levels are calculated for each scatterplot. I also visually inspect each scatterplot for the influence of outliers on the correlation coefficients, slopes, and intercepts. I do not set specific numerical thresholds for the slopes, correlation coefficients, or p-levels to accept or reject the existence of a relationship between the variables, mainly because there is no basis for establishing such a level. Rather, I consider all measures of the relationship to characterize and interpret the influence of drought severity on residential abandonment under different conditions. 4. To evaluate whether differences between relationships are substantial enough to conclude a particular demographic or environmental condition affected the extent of residential abandonment and thus human decisions to move or remain in place, I statistically compare the best-fit slopes, intercepts, and correlation coefficients to identify their “probability of equality.” a. For the slopes, I report a p-value (two-tailed) that answers the question, “If the slopes really were identical, what is the chance that randomly selected data points would have slopes as different (or more different) than those observed?” (Zar 1984; The p-value is the probability that the null hypothesis is correct —that the slopes are identical (the lines are parallel) and sensitivity to droughts under both conditions is similar. b. For the intercepts (also called “elevations”), I also report a p-value for each comparison that answers the question, “If the overall elevations were identical, what is the chance of randomly choosing data points with elevations as different (or more different) than those observed?” (From GraphPad Prism 5 software; c. For the correlation coefficients, I use an interactive on-line calculator (Preacher 2002, where I input the r-values of each correlation to be compared and the n-values for each correlation. The n-values represent the number of 50-year intervals in which there is evidence of settlement occupation under the specific conditions considered. Each correlation coefficient is converted into a zscore using Fisher’s r-to-z transformation. Then, making use of the sample size employed to obtain each coefficient, these z-scores are compared using formula 2.8.5 from Cohen and Cohen (1983: 54). The calculator yields the result of a test of the hypothesis that two correlation coefficients obtained from independent samples are equal. I use and report the p-values associated with a 2-tailed test because there is no reason to expect that one correlation coefficient should be greater than the other. d. The lower the probability of equality of the slopes, intercepts, and correlation coefficients, the greater the probability that differences between them are statistically significant. I do not establish a specific probability level for concluding equality or difference between slopes, intercepts, and coefficients. Rather, I use a general interpretation such as if the probability of equality of two slopes is below 50 percent, then correlations are more likely to be different than not different.

Results These results illustrate the methods and approach by identifying the influence of areal population density and riverine proximity on decisions to move or remain in place as drought severity varied in the study area. Population density To assess the influence of areal population density on residential abandonment, I compare the 250-year relationship between drought severity and residential abandonment in low-density areas with the relationship in high-density areas. Results suggest a stronger, more sensitive relationship between drought severity and residential abandonment among settlements located in high-density areas than among settlements located in low-density areas (Figure 4.3). The slope of the regression line for the high-density scatterplot (m = 3.3) is steeper than the slope for the low-density area scatterplot (m = 2). For every 1 percent change in drought severity, residential abandonment increased 3.3 percent in high-density areas and 2 percent in low-density areas. Thus, residential abandonment in high-density areas was more sensitive to changes in drought severity than that in low-density areas. The slopes cannot, however, be demonstrated as statistically different (50 percent probability of equality), suggesting some necessary caution in our interpretations of difference (Table 4.3). The intercept values from the regression equations (low density −16 percent; high density −44 percent) are also more likely to be the same than different (p = 63 percent). The intercept values are not meaningful as negative percents of residential abandonment, but they suggest that in the absence of drought years, residential abandonment was significantly less. I determine low- and high-density areas by identifying differences in density among the five

watersheds within the central Arizona study area (Ingram 2010: 83–86; calculations not presented here). I sum the number of rooms in each watershed during each of the 50-year intervals and divide it by the number of square kilometers in the watershed (rooms per square kilometer). I then calculate the average population density of each watershed by summing the five density statistics from each watershed and dividing by five. Average watershed population density is a reasonable comparative measure of density because, although absolute density varies over time, relative density (i.e. each watershed’s relative ranking) is mostly consistent throughout the 250-year period. I classify each watershed as having low or high density based on a gap in a histogram of watershed density. All rooms and settlements throughout the central Arizona study area are then classified as located in a high- or low-density area. The resulting demographic and drought severity data at the scale of the total central Arizona study area are presented in Table 4.2 as well as the correlation between these variables. Table 4.2

Relationship between drought severity and the percent of rooms abandoned in low-and high-density watersheds in central Arizona

Figure 4.3

Scatterplots of drought severity and residential abandonment by rooms in low- and high-density watersheds.

Table 4.3

Central Arizona: slopes, intercepts, correlation coefficients, and their probability of equality by rooms located in low- and high-density watersheds

Correlation coefficients and inspection of the scatterplots (Figure 4.3) indicate that the data values from rooms located in high-density areas better fit the regression line (r = .96) than the data values from rooms located in low-density areas (r = .62). The r2 values indicate that 92 percent of the variance in the relationship between drought severity and residential abandonment in high-density areas is explained by changes in drought severity. Only 38 percent of the variance in the relationship between drought severity and residential abandonment is explained in low-density areas. Factors other than drought severity influenced many of the changes in residential abandonment in low-density areas. In sum, decisions to move or remain in place when confronted with drought conditions were moderately influenced by differences in population density throughout the study area. People who lived in areas of high population density were more responsive to changes in drought severity than those who lived in areas with low population density. These results support conceptual models that assert the influence of high population density and the demand for resources on decisions to move or remain in place during droughts. Thus, in the central Arizona study area during the 1200 to 1450 period, settlement pattern change can be explained, at least in part, by variations in population density. A productive next step toward understanding the circumstances that influenced decisions to move or remain in place is to investigate conditions that explain differences in population density among the watersheds. Perhaps variation in the environmental conditions and associated resource productivity in each watershed were the drivers of watershed population density. To test the hypothesis that environmental conditions influenced increases in population density, which resulted in more frequent movements in response to drought, each watershed can be classified by its potential productivity using indicators such as the average annual amount of precipitation each watershed receives, the extent of high-quality arable land, the extent of perennial water sources, etc. Modern environmental data may be appropriate for assigning these classifications. To determine the relationship between potential productivity and density, rank each watershed (low to high) by potential productivity and rank each watershed (low to high) based on density. A scatterplot and Spearman’s rho coefficient can then be calculated to examine the relationship (positive, negative, weak to strong) between productivity and density across all watersheds within a region. Settlement proximity to perennial rivers To assess the influence of a settlement’s location relative to the nearest perennial river on

residential abandonment, I compare the 250-year relationship between drought severity and residential abandonment among rooms located near and far from perennial rivers. To identify differences in riverine proximity, I classify each room as near to (< 2 km) or far from (>2 km) a perennial river using GIS analysis. Table 4.4 presents the number of rooms occupied, the number of rooms abandoned, and the percentage of rooms abandoned by riverine proximity in the total study area during each 50-year interval. I then examine the correlation between the percentage of rooms abandoned in both locations and the drought severity indices. Results indicate that the relationship between drought severity and residential abandonment was not influenced by a room’s location relative to a perennial river (Figure 4.4). The slope of the regression line in the scatterplot of data values representing residential abandonment among rooms far from perennial rivers (m = 2.8) is similar to the slope of the line representing residential abandonment among rooms distant from perennial rivers (m = 3.2). For every 1 percent change in drought severity, residential abandonment increased 2.8 to 3.2 percent indicating sensitivity to drought in both riverine and non-riverine locations. The slopes are statistically more likely the same than different (71 percent probability of equality), suggesting no statistical basis for arguing differences in sensitivity (Table 4.5). The intercept values for settlements far from rivers (b = −27) and near rivers (b = −41) are also more likely the same than different (p = 86 percent). Correlation coefficients (far from rivers r = .88; near rivers r = .96) and the scatterplots (Figure 4.4) indicate a slightly better fit and more explained variance in the near riverine room group but the correlations are statistically more likely the same than different (p = 57 percent). In sum, a settlement’s proximity to a perennial river did not influence decisions to move or remain in place. People who lived both near to and far from perennial rivers were similarly responsive to changes in drought severity. I verified the strength of these results by conducting the same analysis (Ingram 2010: 188–200) with data from each watershed. Thus, results do not support simplified notions of the relationship between access to and availability of water and decisions to move or remain in place during droughts. Table 4.4

Relationship between drought severity and the percentage of rooms abandoned by riverine proximity in central Arizona

Figure 4.4

Scatterplots of drought severity and residential abandonment by riverine proximity.

Table 4.5

Central Arizona rooms: slopes, intercepts, correlation coefficients, and their probability of equality by rooms and riverine proximity

Discussion What has been accomplished in these investigations? A 250-year pattern of ever-shifting settlement locations across a landscape is better understood. People living in areas of higher population density were more likely to move in response to drought than people living in areas of lower population density. And the distance people lived from a perennial river does not appear to have influenced decisions to move or remain in place when they were confronted with drought conditions. What are some implications of these results? The detected influence of population density implies that decisions to settle in or move into less densely populated areas would have been a viable strategy for lessening an endemic risk of drought-related resource shortfalls in dry climates. Decisions to move into more densely populated areas probably required acceptance of increased competition to acquire resources; such decisions also likely required newcomers to settle in areas of relatively lower productivity compared with the early arrivals, given the intraregional variation in productivity. The increased risk of shortfalls for residents of high-density watersheds, indicated by the

stronger drought–movement relationship, also suggests that events outside of high-density areas that stimulated an influx of immigrants would have increased the risk of shortfalls for people living in high-density areas. Controlling immigration into these areas, if this was attempted, would likely have required violence or the threat of violence. Thus, the relationship between the extent of violence and differences in areal population density throughout a region should be tested and might explain variation in the extent of violence over time and space in arid and semi-arid regions. The strong and sensitive relationship between drought severity and residential abandonment in high-density watersheds contradicts arguments that link increases in areal population density to decreases in residential mobility (Cordell 2000: 183; Dean et al. 1994: 85; Minnis 1996; Powell 1988). High population density could limit residential mobility if settlement locations were already claimed or occupied, or hostilities restricted movement. Evidence from this study, however, shows that the percentage of rooms abandoned in high-density watersheds during the five 50-year intervals from 1200 to 1450 is not systematically lower than the percentage of rooms abandoned in low-density watersheds (see Table 4.2). Thus, the idea that high areal population density (watershed or region) is associated with lesser residential mobility, and low areal population density is associated with greater residential mobility, is not supported in the central Arizona study area. Decoupling high population density from decreases in mobility questions arguments that have used this linkage. Similarities in the drought–movement responses among those living near and far from perennial rivers suggest a variety of scenarios and hypotheses to be tested. Perhaps differences in settlement-scale productivity and food shortfall risks were, in effect, equalized by different adaptive strategies. For example, people living far from perennial rivers may have developed a more diverse water management infrastructure (e.g. check dams, terracing, water diversion structures; Ford and Swentzell 2015: 338–350) and a more flexible suite of drought-buffering strategies than those living near perennial rivers, if greater access to water and increased agricultural productivity (e.g. through irrigation) decreased the repertoire of strategies used by riverine agriculturalists to buffer drought-related productivity declines. The lack of influence of the potential resource productivity of a settlement also implies that decisions to move towards perennial rivers might not have been a viable strategy for lessening an endemic risk of resource shortfalls in this dry climate. Analytically, riverine proximity may be a poor proxy for differences in the potential productivity of settlement locations. Social factors (e.g. the existence of food exchange relationships, territoriality that restricted movement) that affect the risk of shortfalls are also not accounted for when local-scale environmental settlement characteristics are emphasized. Whatever the reason, this result challenges simple conceptual models that posit a link between access to perennial riverine resources and less vulnerability to drought. People living near perennial rivers were just as likely to move as drought severity increased as those living far from perennial rivers. This result also suggests a reinvestigation of the assumed challenges and risks of dryland or rain-fed farming in the region compared with the assumed lower risks in irrigated areas. By building an understanding of contributing conditions to decisions to move or remain in place, we build an understanding of climatic influences on human decision-making in particular places and during specific periods. Such an understanding can be used to interpret

significant events such as the depopulation of regions: a classic archaeological problem. In the northern Southwest, for example, the depopulation of the late thirteenth century has been studied for almost 100 years. Among the many factors now considered influential in the depopulation are decreased mobility options due to increased settlement population levels and increases in catchment population density caused by settlement aggregation (Van West and Dean 2000; Varien et al. 1996). This argument is used to explain why the impacts of a specific drought (1276 to 1299) might have been more severe than previous droughts. Such an explanation is necessary because droughts are a common occurrence in the Southwest. While increased population levels and aggregation might explain the differential impact of the drought in the late thirteenth century compared with previous droughts, the argument is partially amenable to testing using a long-term paleoclimatic record of droughts and a demographic history of mobility. That is, do we have any long-term evidence that settlement population levels, catchment population density, and/or settlement aggregation influenced decisions to move at other times in this area? If so, the argument is strengthened; if not, we cannot lessen the possibility that the changing demographic conditions, the drought, and the depopulation were simply space–time coincidences.

Conclusion This chapter has described a new approach using familiar methods to investigate the influence of demographic and environmental conditions on human decisions to move or remain in place as drought severity varies in a dry environment. The emphasis on identifying human decisions and the potential of the approach to enable investigations of a wide variety of social, environmental, and demographic influences on these decisions situates the proposed approach comfortably within the domain of settlement ecology. This chapter has also demonstrated how to lessen the problem of inferring causation from a single space–time coincidence between a drought and a settlement pattern change. Investigation of a long-term (e.g. century-scale) relationship between changes in drought severity and changes in residential abandonment at various spatial scales can be used to identify drought influences on shifting settlement patterns. If, over centuries, changes in the severity of multiple droughts were strongly associated with proportionate changes in residential abandonment, then drought influences on decisions to move or remain in place may be reasonably concluded. Comparing the strength of the drought– population movement relationship among settlements with different demographic and environmental conditions reveals the influence of these conditions on residential decisionmaking. This approach lends itself to investigations using a wide variety of variables at many spatial scales. The approach presented in this chapter could be described as “applied settlement ecology” if the methods and insights were used to contribute to modern studies of human vulnerability to climate extremes associated with contemporary global climate change. Plans for adapting to and mitigating the effects of anticipated climate extremes, such as drought, are a global-scale research effort. Modern drought planners (and/or interested archaeologists) could apply the methods used in this chapter with much more refined data. Instead of tree-ring retrodicted precipitation and inferred drought, modern and historical instrumental weather data could be

used. Instead of coarse-grained settlement abandonment dates, modern census data on population changes and migration could be used. Environmental proxies for resource productivity could be replaced with agricultural statistics from the Department of Agriculture (or similar). Informed by archaeological studies of demographic and environmental conditions that influenced past human decisions to move or remain in place over centuries, modern planners and analysts can determine if these conditions were also influential in recent decades in many locations worldwide. Similarly, archaeologists can investigate conditions these modern planners and analysts have identified as currently influential. Such investigations and results could enrich archaeological understanding of the past and add validity to or question future climate change scenarios, adaptation, and mitigation planning. In these ways, the past and present can work together to inform the future.

References Abruzzi, William S. 1989 Ecology, Resource Redistribution, and Mormon Settlement in Northeastern Arizona. American Anthropologist 91(3): 614–655. Ahlstrom, Richard V. N. Carla R. Van West, and Jeffrey S. Dean 1995 Environmental and Chronological Factors in the Mesa Verde-Northern Rio Grande Migration. Journal of Anthropological Archaeology 14(2): 125–142. Binford, Lewis R. 1983 Long-Term Land-Use Patterning: Some Implications for Archaeology. In Working at Archaeology. Academic Press, New York. Bright, Jill L., and John J. Hervert 2005 Adult and Fawn Mortality of Sonoran Pronghorn. Wildlife Society Bulletin 33(1): 43– 50. Cohen, Jacob and Patricia Cohen 1983 Applied Multiple Regression/Correlation Analysis for the Behavioral Sciences. L. Erlbaum Associates, Hillsdale, NJ. Cordell, Linda 1997 Archaeology of the Southwest. Second edition. Academic Press, San Diego. Cordell, Linda S. 2000 Aftermath of Chaos in the Pueblo Southwest. In Environmental Disaster and the Archaeology of Human Response, edited by Garth Bawden and Richard M. Reycraft, pp. 179–194. University of New Mexico Press, Albuquerque. Cutter, Susan L., Bryan J. Boruff, and W. Lynn Shirley 2003 Social Vulnerability to Environmental Hazards. Social Science Quarterly 84(2): 242–261. Dean, Jeffrey S. 1991 Thoughts on Hohokam Chronology. In Exploring the Hohokam, edited by George J. Gumerman, pp. 61–150. University of New Mexico Press, Albuquerque. Dean, Jeffrey S., William H. Doelle, and Janet D. Orcutt 1994 Adaptive Stress, Environment, and Demography. In Themes in Southwest Prehistory, edited by G.J. Gumerman, pp. 53–86. School of American Research Press, Sante Fe. Diamond, Jared 2005 Collapse: How Societies Choose to Fail or Succeed. Viking, Penguin Group, New York. Fischer, R.A., and Neil C. Turner 1978 Plant Productivity in the Arid and Semiarid Zones. Annual Review of Plant Physiology 29: 277–317. Fish, Suzanne, and Gary Nabhan 1991 Desert as Context: The Hohokam Environment. In Exploring the Hohokam, edited by George J. Gumerman, pp. 29–62. University of New Mexico Press, Albuquerque. Ford, Richard I., and Roxanne Swentzell 2015 Precontact Agriculture in Northern New Mexico. In Traditional Arid Lands Agriculture: Understanding the Past for the Future, edited by Scott E. Ingram and Robert C. Hunt, pp. 330–357. University of Arizona Press, Tucson. Fritts, Harold C. 1976 Tree Rings and Climate. Academic Press, London and New York. Fussell, Elizabeth, Lori M. Hunter, and Clark L. Gray 2014 Measuring the Environmental Dimensions of Human Migration: The Demographer’s Toolkit. Global Environmental Change 28: 182–191. Gumerman, George J. (editor) 1988 The Anasazi in a Changing Environment. Cambridge University Press, Cambridge. Halstead, Paul, and John O’Shea (editors) 1989 Bad Year Economics: Cultural Responses to Risk and Uncertainty. Cambridge University Press, Cambridge. Hill, J. Brett, Jeffery J. Clark, William H. Doelle, and Patrick D. Lyons 2004 Prehistoric Demography in the Southwest: Migration, Coalescence, and Hohokam Population Decline. American Antiquity 69(4): 689–716. Hill, J. Brett, Jeffery J. Clark, William H. Doelle, and Patrick D. Lyons 2010 Depopulation of the Northern Southwest: A Macroregional Perspective. In Leaving Mesa Verde: Peril and Change in the Thirteenth-Century Southwest, edited by Timothy A. Kohler, Mark D. Varien and Aaron M. Wright, pp. 34–52. University of Arizona Press, Tucson.

Ingram, Scott E. 2010 Human Vulnerability to Climatic Dry Periods in the Prehistoric U.S. Southwest. Ph.D. dissertation. Arizona State University, Tempe; University Microfilms, Ann Arbor. Ingram, Scott E. 2012 The Climatic Context of Population Change on Perry Mesa. In Prehistoric Cultures of the Perry Mesa Region, Proceedings of the Perry Mesa Symposium, edited by Will G. Russell and Michael J. Hoogendyk, pp. 241–255. Cultural Resources Committee of the Friends of Agua Fria National Monument, Phoenix. Ingram, Scott E. 2014 Climatic, Demographic, and Environmental Influences on Central Arizona Settlement Patterns. In Alliance and Landscape on Perry Mesa in the Fourteenth Century, edited by David Abbott and Kate A. Spielmann, pp. 23–51. University of Utah Press, Salt Lake City. Ingram, Scott E. 2015 Human Vulnerability to Dry Periods. In Traditional Arid Lands Agriculture: Understanding the Past for the Future, edited by Scott E. Ingram and Robert C. Hunt, pp. 131–160. University of Arizona Press, Tucson. Knight, C. Gregory, and Jill Jäger 2009 Integrated Regional Assessment of Global Climate Change. Cambridge University Press, Cambridge. Kohler, Timothy A., Mark D. Varien and Aaron M. Wright (editors) 2010 Leaving Mesa Verde: Peril and Change in the Thirteenth-Century Southwest. University of Arizona Press, Tucson. Liverman, Diana M., and Robert Merideth 2002 Climate and Society in the US Southwest: The Context for a Regional Assessment. Climate Research 21(3): 199–218. Meko, David, Charles W. Stockton, and W.R. Boggess 1995 The Tree-Ring Record of Severe Sustained Drought. Water Resources Bulletin of the American Water Resources Association 31(5): 789–801. Meyer, William B., Karl W. Butzer, Thomas E. Downing, B.L. Turner II, George W. Wenzel, and James L. Wescoat 1998 Reasoning by Analogy. In Human Choice and Climate Change: Volume Three, The Tools for Policy Analysis, edited by Steve Rayner and Elizabeth L. Malone, pp. 217–289. Battelle Press, Columbus, OH. Minnis, Paul E. 1996 Notes on Economic Uncertainty and Human Behavior in the Prehistoric North American Southwest. In Evolving Complexity and Environmental Risk in the Prehistoric Southwest, edited by Joseph A. Tainter, and Bonnie Bagley Tainter, pp. 57–78. Addison-Wesley Publishing Company, Reading, MA. Muenchrath, Deborah A., and Ricardo J. Salvador 1995 Maize Productivity and Agroecology: Effects of Environment and Agricultural Practices on the Biology of Maize. In Soil, Water, Biology, and Belief in Prehistoric and Traditional Southwestern Agriculture, edited by H.S. Toll, pp. 303–333. New Mexico Archaeological Council Special Publication, Volume 2. New Mexico Archaeological Council, Albuquerque. Nelson, Ben A., Debra L. Martin, Alan C. Swedlund, Paul R. Fish, and George J. Armelagos 1994 Studies in Disruption: Demography and Health in the Prehistoric American Southwest. In Understanding Complexity in the Prehistoric Southwest, edited by George J. Gumerman and Murray Gell-Mann, pp. 59–112. Santa Fe Institute, Studies in the Sciences of Complexity. Addison-Wesley Publishing Company, Reading, MA. Nelson, Margaret C., and Gregson Schachner. 2002 Understanding Abandonments in the North American Southwest. Journal of Archaeological Research 10(2): 167–206. Osborn, Alan J. 1993 Snowblind in the Desert Southwest: Moisture Islands, Ungulate Ecology, and Alternative Prehistoric Overwintering Strategies. Journal of Anthropological Research 49(2): 135–164. Powell, Shirley 1988 Anasazi Demographic Patterns and Organizational Responses: Assumptions and Interpretive Difficulties. In The Anasazi in a Changing Environment, edited by George J. Gumerman, pp. 168–191. Cambridge University Press, Cambridge. Preacher, Kristopher J. 2002 Calculation for the Test of the Difference between Two Independent Correlation Coefficients [Computer software]. (accessed September 1, 2010). Reid, J. Jefferson and Stephanie M. Whittlesey 1997 The Archaeology of Ancient Arizona. University of Arizona Press, Tucson. Ribot, Jesse C. 1995 The Causal Structure of Vulnerability: Its Application to Climate Impact Analysis. GeoJournal 35(2): 119– 122. Rose, Martin R., Jeffrey S. Dean, and William J. Robinson 1981 The Past Climate of Arroyo Hondo New Mexico Reconstructed from Tree Rings. Arroyo Hondo Archaeological Series, Volume 4. School of American Research Press, Santa Fe, NM. Salzer, Matthew W. 2000 Dendroclimatology in the San Francisco Peaks Region of Northern Arizona, USA. Ph.D. dissertation. Department of Geosciences, University of Arizona, Tucson. Salzer, Matthew W., and Kurt F. Kipfmueller 2005 Reconstructed Temperature and Precipitation on a Millennial Timescale from Tree-Rings in the Southern Colorado Plateau, USA. Climatic Change 70(3): 465–487. Shaw, Robert H. 1977 Climatic Requirement. In Corn and Corn Improvement, Volume 18, edited by G.F. Sprague, pp. 591– 623. American Society of Agronomy, Madison, WI. Sheppard, Paul R., Andrew C. Comrie, Gregory D. Packin, Kurt Angersbach, and Malcolm K. Hughes 2002 The Climate of the US Southwest. Climate Research 21: 219–238. Slatter, Edwin D. 1979 Drought and Demographic Change in the Prehistoric Southwest United States: A Preliminary Quantitative Assessment. Ph.D. dissertation. Department of Anthropology, University of California, Los Angeles;

University Microfilms, Ann Arbor, Los Angeles. Turner, Raymond M., and David E. Brown 1982 Sonoran Desertscrub. In Desert Plants 4 (1–4), Volume 4, edited by David E. Brown, pp. 181–221. University of Arizona Press, Tucson. Van West, Carla R., and Jeffrey H. Altschul 1997 Environmental Variability and Agricultural Economics along the Lower Verde River, ad 750–1450. In Vanishing River: Landscapes and Lives of the Lower Verde Valley, The Lower Verde Archaeological Project, Overview, Synthesis, and Conclusions, edited by Stephanie M. Whittlesey, Richard CiolekTorello, and Jeffrey H. Altschul, pp. 337–392. SRI Press, Tucson. Van West, Carla R., and Jeffrey S. Dean 2000 Environmental Characteristics of the ad 900–1300 Period in the Central Mesa Verde Region. Kiva 66(1): 19–44. Varien, Mark D., William D. Lipe, Michael A. Adler, Ian M. Thompson, and Bruce A. Bradley 1996 Southwestern Colorado and Southeastern Utah Settlement Patterns: ad 1100 to 1300. In The Prehistoric Pueblo World, ad 1150 to 1350, edited by Michael A. Adler, pp. 86–113. University of Arizona Press, Tucson. Western Region Climate Center 2010 Historical Climate Information. Western Regional Climate Center (WRCC). Electronic document,, accessed January 2010. Whittlesey, Stephanie M., Richard Ciolek-Torrello, and Jeffrey H. Altschul (editors) 1997 Vanishing River: Landscapes and Lives of the Lower Verde Valley, The Lower Verde Archaeological Project, Overview, Synthesis, and Conclusions. SRI Press, Tucson. Wilcox, David R., William H. Doelle, J. Brett Hill, and James P. Holmlund 2003 Coalescent Communities GIS Database: Museum of Northern Arizona, Center for Desert Archaeology, Geo-Map Inc. On File, Center for Desert Archaeology, Tucson. Wilcox, David R., David A. Gregory, and J. Brett Hill 2007 Zuni in the Puebloan and Southwestern Worlds. In Zuni Origins: Toward a New Synthesis of Southwestern Archaeology, edited by David A. Gregory and David R. Wilcox, pp. 165–209. University of Arizona Press, Tucson. Zar, J.H. 1984 Biostatistical Analysis. Second edition. Prentice-Hall, Englewood Cliffs, NJ.

Part III

Central America

5 Political-economic strategies and settlement ecology in the Mesoamerican Gulf Lowlands Olmec, Epi-Olmec, and Classic Period settlement in the El Mesón area of the Eastern Lower Papaloapan Basin, Veracruz, Mexico Michael L. Loughlin

Introduction In 2003 I initiated the Recorrido Arqueológico El Mesón (RAM) to examine settlement patterns and political and economic organization in the area around El Mesón, a Late Formative period center located in the Eastern Lower Papaloapan Basin (ELPB) of the modern Mexican state of Veracruz. During the course of the survey, approximately 400 individual architectural features and artifact concentrations were identified within the 27km2 survey area. While the focus of the research was on the Formative to Classic period transition, the RAM survey documented an uninterrupted sequence of occupation in the area spanning the Early Formative period through the Postclassic periods. This chapter presents some of the results of the RAM survey. I begin by tracing the development of settlement in the El Mesón area from its Early Formative beginnings through its decline in the Late Classic period. These data indicate that the area expanded from a series of small villages, hamlets, and farmsteads during the Early (1450–1000 BC) and Middle (1000– 400 BC) Formative periods to become a regional center during the Late Formative period (400–1 BC). The area continued to expand through the Protoclassic (AD 1–300) and Early Classic (AD 300–600) periods before declining during the Late Classic period (AD 600–900). By the Postclassic period (AD 900–1520), only remnant populations remained. The reasons for this decline are unclear but they may be related to changes in regional exchange networks and/or climatic changes from drier to wetter conditions. Drawing on settlement ecology, I address questions about the persistence of settlement in the El Mesón area, as well as questions regarding how organizational changes within the El Mesón area affected settlement. I argue that the continuity and evolution of settlement in the El Mesón area was strongly affected by both the area’s location along an important transportation route and the political strategies used by leaders in the area. I suggest that the location of El Mesón made the area attractive for inclusion in the Tres Zapotes regional polity during the Late Formative period, and provided opportunities for local elites to break politically with Tres Zapotes and establish the area’s independence during the subsequent Protoclassic and Early Classic periods. These political shifts were reflected in changes to the architectural programs

of the formal mound groups in the area.

Settlement ecology Settlement ecology is a theoretical approach grounded in anthropology and locational studies in human geography (Stone 1996). This theoretical orientation grew out of a desire to understand the development of settlement patterns in agrarian societies. Of particular concern is not only explaining the “rules” (Stone 1996: 6–7) that guided decisions about settlement location, but also why one particular set of rules was chosen over another. To that end, settlement ecology is focused on the factors that would have impacted people’s decisions about where to locate their settlements on the landscape. These include both the natural environment, as well as cultural factors, such as political organization, ideology, exchange networks, etc. (Jones 2010; Jones et al. 2012). Many studies based in settlement ecology focus on broad-scale regional settlement patterns (e.g. Jones 2010; Jones et al. 2012; Stone 1996). My use of settlement ecology here is slightly different. Rather than focus on regional settlement, my concern here is examining the development and evolution of settlement patterns in and around the Late Formative center of El Mesón. My goal is to examine the influence of the natural and cultural environment on the development and eventual decline of settlement in this relatively small region, and to show the utility of settlement ecology for examining long-term political economic changes in complex societies.

Environmental setting and background El Mesón is located in the Eastern Lower Papaloapan Basin of Mexico’s Southern Gulf Lowlands, just outside the modern town of Ángel R. Cabada (Figure 5.1). The ELPB comprises a broad alluvial plain that extends from the eastern bank of the Papaloapan River to the western slope of the Tuxtla Mountains, a low volcanic massif that rises above the plains of the Papaloapan and Coatzacoalcos River Basins. The site is located on the Arroyo Tecolapan, approximately 13km north of the large center Tres Zapotes. During the Formative period, the ELPB formed the western portion of the Olmec Heartland, Olman. El Mesón is situated in a major ecotone between the alluvial plain and the uplands of the Tuxtla Mountains. A nearby mountain pass provides relatively easy access into and out of the Tuxtlas, and has been an important transportation and communication route since Prehispanic times.

Figure 5.1

Gulf Coast with important sites mentioned in the text.

The climate of the area is hot and moist with mean temperatures of 22–26°C, and high temperatures exceeding 40°C in the summer (Gómez-Pompa 1973: Figure 3; Soto and Gama 1997: Apéndice 1.2). Precipitation in the area ranges between 2000 and 2500mm annually, the majority of which falls during the rainy season, which lasts from June to November (GómezPompa 1973: 81, Figure 6; Soto y Gama 1997: Apéndice 1.2). Soils in the area around El Mesón are fertile luvic phaeozems (FAO-UNESCO 1971–1978). In modern times the area has been an important producer of sugar cane. This environmental setting would have provided a wide range of subsistence resources for the region’s Prehispanic residents. Sediment cores indicate that domesticated maize was present in the Gulf Lowlands as early as 5000 BC (Pope et al. 2001), and by the Early Formative period maize agriculture contributed to the Olmec’s subsistence economy (Borstein 2002; VanDerwarker 2006). Olmec reliance on maize increased during the Middle Formative period (Rust and Leyden 1994: 192–194) and maize has continued to play a critical role as a staple crop throughout Prehispanic and modern times. Other important domesticates, such as beans and squash, were also present by the Early Formative period (Cyphers 1996: 66). Wild plant food resources that have been identified in the archaeological contexts include coyol palmfruit, zapote mamey, and avocado (VanDerwarker 2006: 87–88). Animal resources include a variety of fresh and saltwater fish, turtles and other reptiles, amphibians, and birds, as well as large and small mammals, such as deer, peccary, squirrel, opossum, rabbit, and dog (Pool 2007: 76–77; Pope et al. 2001; Rust and Leyden 1994; Wing 1980; VanDerwarker 2003, 2006). El Mesón first came to the attention of archaeologists when Matthew Stirling visited the site in 1939, identifying Stela 1, a columnar basalt stela (Stirling 1943: 30). Based on his observations of the numerous earthen mounds in the area, Stirling (1943: 30) remarked, “it would appear that El Meson was a very important center in aboriginal times.” Stirling’s

observations are echoed in Coe’s (1965: 679) later description of the number of mounds in the area. Following Stirling’s visit, the focus of interest was the small corpus of stone monuments, particularly Monument 2, the El Mesón Stela, which was drawn and described by Covarrubias (1957) and Drucker (1968). Based on the style of the carving, Covarrubias attributed the monument to the Late Formative period, while Drucker (1968) suggested a Classic date. Prior to the RAM survey, the only collection of artifacts from El Mesón was gathered by John Scott who visited the site in 1975 as part of his study of post-Olmec art (Scott 1977). Scott made a small collection of figurines and ceramic sherds in order to aid in dating the stone monuments. These materials indicate that the site’s primary occupation dated to the Late Formative and Protoclassic periods, with smaller components dating to the Middle Formative and Early Classic periods. Additionally, Scott also provided the first map of the site, which included his reconstructions of the original placements of the stone monuments. The RAM survey methods Given the lack of previous research at El Mesón, the RAM survey was designed to provide temporal and spatial information about settlement in the area. The 27km2 project area was surveyed using a full coverage technique. Survey crew walked the entire survey area in transects with a 20m spacing. This close spacing would ensure that small house platforms would be identified. Because of the possibility for mounds to be densely distributed through the survey area, potentially blurring the boundaries between sites, a “siteless” survey technique (see Dunnell and Dancey 1983; Foley 1981; Gallant 1986) was used, which was based on the method used by Stark (1991) for her survey of the Mixtequilla region of SouthCentral Veracruz. Instead of sites, architectural features (e.g. mounds) and artifact concentrations (many of which are remnants of destroyed mounds) are the basic units of analysis. In cases where there are no clear boundaries between sites, this method allows for the variability in the settlement history to be considered (Stark 1991: 44). When architectural features or artifact concentrations were identified their extents and heights were measured, a GPS point was taken from the center of the structure or concentration, and a collection of artifacts was made. The collection strategy used for the RAM survey was also based on the collection strategy used in the Mixtequilla (Stark 1991). Ceramics were collected until 100 rims or decorated sherds were recovered. Limiting the collections to rims and decorated sherds allowed for manageable assemblages that preserved chronological and functional information. All other artifacts were collected except for large pieces of groundstone (e.g. large metate fragments), which were documented and left in the field. Because of the alluvial setting of the RAM survey area, there is a potential for deposits to be deeply buried. An examination of the exposed bank of the Arroyo Tecolapan indicates that alluvial deposits at least 2m thick cover the floodplain. The consequence of such deposition is that artifacts, and potentially mounds, may not be visible on the surface, thus leading to the underrepresentation of site densities. Indeed at Tres Zapotes, there was minimal evidence for an Early Formative occupation of the site (Pool and Ohnersorgen 2003: 24). However, excavations identified Early Formative deposits at depths greater than 5m below the modern

ground surface in the floodplain (Pool et al. 2010). Moreover, Wendt (2003) indicates that several mounds at Tres Zapotes have also been completely buried by alluvium. Due to its alluvial setting, the El Mesón area has a similar potential for deeply buried deposits, or for low mounds to be completely buried. Given this potential, the reconstruction of the regional settlement pattern most likely does not represent the totality of settlement in the area. Rather, this pattern reflects the minimal settlement in the area.

Settlement history of the El Mesón area A total of 397 architectural features and artifact concentrations (features) were identified during the RAM survey. The overall settlement density is 14.8 features per km2. However, this density is not uniform throughout the survey area. Rather, the area is characterized by zones of relatively dense settlement interspersed with areas of lighter density occupation and areas that were unoccupied. This pattern is intermediate between the relatively continuous distribution of settlement in the Mixtequilla region (Stark 1991) and the more discrete settlements of the Tuxtla Mountains (Killion and Urcid 2001; Kruszczynski 2001; Santley 2007; Santley and Arnold 1996; Stoner 2011). Olmec settlement The earliest evidence for settlement in the El Mesón area dates to the Early Formative Arroyo phase (1250–1000 cal. BC), which is coeval with the San Lorenzo B phase at San Lorenzo (Coe and Diehl 1980; Symonds et al. 2002). Early Formative artifacts were recovered in 29 collections from the El Mesón area (Figure 5.2). Settlement was lightly distributed along the arroyos that flow through the area. The only portion of the survey area where there was some agglomeration of settlement was at El Mesón, where nine mounds (all under 3m in height) and one plaza (34.5 percent of all collections with Early Formative artifacts) yielded Early Formative artifacts. At this time, El Mesón was likely a small village. Outside of El Mesón, no groupings of more than two mounds were located within 100m of each other.

Figure 5.2

Early Formative settlement in the El Mesón area.

During the Middle Formative Tres Zapotes phase (1000–400 cal. BC) settlements expanded. Middle Formative artifacts were recovered in collections from 53 features within the survey area, including 23 of the 29 features with Arroyo phase occupation (Figure 5.3). Like the Early Formative settlement pattern, the Middle Formative settlement also tends to follow the arroyos. In contrast to the Arroyo phase, there is more evidence for settlement clustering during the Tres Zapotes phase. Approximately 30 percent (n = 16) of the Middle Formative features were located at El Mesón. Additionally there is some agglomeration of settlement to the north in the Norte group (n = 8) and on the south side of the Arroyo Tecolapan at El Mesón South. However, there is no indication of civic-ceremonial architecture at this time. With only a few exceptions where earlier artifacts were incorporated into the fill of later structures, all of the features with evidence of Middle Formative occupation were domestic platforms less than 3m in height. Like the Arroyo phase, the Tres Zapotes phase occupation most likely represents a small village at El Mesón with some nearby hamlets and isolated farmsteads. Epi-Olmec settlement Following the decline of the Middle Formative Olmec center of La Venta (c.400 BC), the Southern Gulf Lowlands experienced a dramatic demographic shift as population declined in eastern Olman, the Olmec Heartland, and expanded in western Olman, which includes the ELPB. Moreover, political and economic priority also moved from eastern Olman to the west, and Tres Zapotes became the largest, and most important political and economic center in the region (Pool 2007: 247).

Figure 5.3

Middle Formative settlement in the El Mesón area.

Evidence of this demographic, political and economic shift is clear in the El Mesón area. Late Formative Hueyapan phase (cal. 400–1 BC) artifacts were recovered from 144 collections, an increase of more than 270 percent from the Tres Zapotes phase (Figure 5.4). Approximately 72 percent of the collections with Middle Formative artifacts (n = 38) also had Late Formative artifacts, suggesting some continuity of settlement from the Tres Zapotes phase through the Hueyapan phase. However, where the Tres Zapotes phase occupation was focused on the eastern half of the survey area, Hueyapan phase settlement extends throughout the area. Perhaps the biggest difference between the Late Formative settlement pattern and the earlier occupations is the presence of civic-ceremonial architecture. Artifacts from El Mesón suggest that the site’s civic-ceremonial core was constructed and in use during the Hueyapan phase. This complex comprises an east–west oriented plaza (160m east–west by 90m north–south) bounded by a 2.2m tall long mound to the north, and an 8–9m tall conical mound to the west (Figure 5.5). A low conical mound atop a low platform is located near the centerline of the plaza approximately 32m east of the conical mound. The plaza is closed on the west side by the remnant of a low mound. This architectural layout, which Pool (2003a: 92; 2008: 128) refers to as the Tres Zapotes Plaza Group layout (TZPG), is identical to the architectural programs in the four formal mound complexes at Tres Zapotes. Based on the artifacts recovered from these structures and comparisons with similar structures elsewhere in Mesoamerica, Pool (2008: 128) argues that the tall conical mound served as the base for a temple, the long mound served both elite domestic and administrative functions, and the low mound in the plaza served as a shrine or adoratorio. At Tres Zapotes, the TZPG complexes are interpreted as representing the seats of power for elite faction leaders who shared governance of the center (Pool 2003a, 2003b, 2008). While the use of long and conical mounds in formal complexes is hardly unique, Pool (2008: 147) notes that the specific organization of the TZPG is relatively rare in the Southern Gulf Lowlands, restricted largely to the ELPB region. Of the

known TZPG complexes located outside Tres Zapotes, the TZPG at El Mesón is the largest. I argue that the use of the TZPG at El Mesón indicates that Tres Zapotes had incorporated the area into a regional polity, probably as a secondary center (Loughlin 2012). What is not clear, is whether the El Mesón TZPG was an emulation of Tres Zapotes’s symbols of political power and authority by local elites, or if it represents an extension of one of the governing factions from Tres Zapotes.

Figure 5.4

Late Formative settlement in the El Mesón area.

Figure 5.5

TZPG complex at El Mesón (0.5m contour).

In addition to the TZPG complex at El Mesón, two additional architectural complexes were

also in use by the Hueyapan phase. These complexes, located to the north and south of El Mesón are smaller than El Mesón and do not feature the TZPG architectural layout. Moreover, artifacts recovered from these complexes suggest that they are functionally distinct from El Mesón. In the case of the Norte group, the high density of groundstone artifacts, including production indicators (e.g. basalt flakes, polishing stones, and an anvil), suggests that this complex may have been a barrio for craftsmen engaged in groundstone production. Similarly, at El Mesón South, elevated densities of obsidian cores and debitage, as well as the presence of numerous obsidian drills, suggest that this area was also a craft barrio. These complexes most likely did not serve civic-ceremonial functions and were subject to El Mesón’s political authority during the Late Formative period. Protoclassic and Early Classic settlement Settlement in the El Mesón area peaked during the Protoclassic period Nextepetl phase (AD 1– 300) and the Early Classic Ranchito phase (AD 300–600). The pattern of population expansion that characterized the Formative period occupation of the area continued as the number of collections with temporally diagnostic artifacts grows from 144 during the Hueyapan phase to 247 during the Nextepetl phase, an increase of 172 percent (Figure 5.6). Settlement density increases to 9.1 features/km2, marking the densest occupation in the region. All of the features with evidence of Late Formative occupation were also occupied during the Protoclassic period. At this time settlement is spread throughout the RAM survey area.

Figure 5.6

Protoclassic settlement in the El Mesón area.

In addition to the regional population expansion, the Nextepetl phase also marks the apogee of El Mesón’s political influence. The clearest indication of this prominence was the placement of two stelae at the center. Although their original placement is unclear, Scott (1977:

124) indicates that they were placed to the west of the large conical mound, on the north side of a secondary plaza just west of the main plaza of the TZPG. No other monuments like these have been reported from elsewhere in the RAM survey area. However, the architectural data suggest that El Mesón’s political domination of the area was short-lived, and that by the Early Classic period, it had been eclipsed by other nearby complexes. Nextepetl phase ceramics (e.g. Sandy Fine Orange, a precursor to the Fine Orange of the Tuxtlas) recovered suggest that several new complexes were either under construction or in operation by the end of the Protoclassic period. These new complexes do not conform to the TZPG arrangement. Rather, each complex is unique in terms of its constituent features and arrangement. The only unifying characteristic is the use of large, flat-topped, quadrilateral platforms in each mound group. The largest of these structures measured 76m by 71m at its base and had a height of approximately 14m. In most of the groups no mounds were identified on top of these platforms. In one group, however, the height of the platform was reduced (1.5m), but a conical mound, 6m tall, and a 1.5m platform topped by a long mound, 7.8m tall, were located on its summit.

Figure 5.7

Early Classic settlement in the El Mesón area.

Settlement in the El Mesón area was stable during the Ranchito phase. Diagnostic artifacts were recovered from 246 features, of which 239 were occupied during the preceding Nextepetl phase (Figure 5.7). However, there is clear evidence for a major shift in the organization of the area. The TZPG complex at El Mesón ceased to serve as a political center and the civic-ceremonial complex was largely abandoned, leaving a remnant population outside of the civic-ceremonial core. With the decline of El Mesón, political authority shifted to the El Mesón South mound group, located on the south side of the Arroyo Tecolapan. Here, a new civic-ceremonial complex emerged, possibly as early as the end of the Protoclassic period. This complex, comprising an

east–west oriented plaza bounded by a conical mound to the east, long mounds to the north and south, and a ballcourt to the west, conforms to what has been called the Standard Plan (Daneels 2002; Stark 2003, 2008), a layout associated with political centers in Central and South-Central Veracruz, such as Cerro de las Mesas (Figure 5.8). It is worth noting that large quadrilateral platforms are also more common in Central Veracruz than Southern Veracruz (Daneels 2002, 2008; Stark 2003, 2008). Despite these nonlocal architectural forms, there is little evidence suggesting that the El Mesón area was under the control of Cerro de las Mesas. Rather, the adoption of these architectural forms may reflect increased interaction between Cerro de Las Mesas and the El Mesón area. This interaction is further evidenced by the presence of ceramics of the South-Central Veracruz style in surface collections from the El Mesón area. This adoption could also represent an attempt by local leaders in the El Mesón area to legitimize their political authority by adopting symbols of rulership from the larger center. This organizational shift has parallels with changes at Tres Zapotes. Pool (2008) argues that during the Protoclassic period, and continuing into the Early Classic period, the architectural programs in the major mound groups at the site underwent important changes. New constructions reoriented plazas and new structures were added to mound groups. He suggests that the alteration of the TZPG complexes represents a breakdown of the cooperation between factions for site and polity governance and an intensification of factional competition. However, the adoption of nonlocal architectural forms, such as the large platforms and Standard Plan, is absent, perhaps indicating that Tres Zapotes was not interacting with centers in South-Central Veracruz, or that leaders at Tres Zapotes did not need to adopt foreign symbols to legitimize their political authority.

Figure 5.8

Standard plan complex at El Mesón South (0.5m contour).

Late Classic and Postclassic settlement By the Late Classic period Quemado phase (cal. AD 600–900), settlement in the El Mesón area was in decline. Late Classic artifacts were recovered from 154 features, a decrease of about 37 percent from the Ranchito phase (Figure 5.9). While there was some loss of settlement in

the areas between architectural complexes, the decline in settlement was most pronounced within architectural complexes. Most of the large quadrilateral platforms fell into disuse, and El Mesón was almost completely abandoned. El Mesón South appears to have maintained its political authority at this time; however, occupation around this complex was less intense.

Figure 5.9

Late Classic settlement in the El Mesón area.

By the Postclassic period (AD 900–1520) only a remnant population remained in the El Mesón area. The number of collections with diagnostic artifacts declines to 29. At this time none of the architectural complexes in the area was in use.

Political strategies and settlement ecology The RAM survey documented two uninterrupted millennia of Prehispanic occupation in the El Mesón area. Over the course of this time, the area experienced significant reorganizations in settlement, including the emergence of El Mesón as a center during the Late Formative period, the political break with Tres Zapotes and the subsequent abandonment of El Mesón during the Protoclassic and Early Classic periods, and the decline of the area during the Late Classic period. In the following discussion, I examine how these reorganizations were shaped by the location of the area along an important transportation route and the regional political environment. By the end of the Middle Formative period the Southern Gulf Lowlands were experiencing an important political, demographic, and cultural shift. The large Olmec center of La Venta was in decline, and much of the eastern portion of Olman was depopulated. La Venta’s demise also marked the end of Olmec culture. At the same time, populations grew in western Olman, Tres Zapotes expanded from a local center to become the head of a regional polity (Pool 2003a, 2003b, 2008), and Olmec culture evolved into a new cultural system called Epi-Olmec (Pool

2000). Pool and I (Pool and Loughlin 2015) have argued elsewhere that Tres Zapotes’s survival of this political, demographic, and cultural shift was based, in part, on the ability of political leaders to reorient the center’s political organization: eschewing a centralized system controlled by a single ruler and adopting a decentralized system where political power was shared between faction leaders. The political organization of Early and Middle Formative Olmec centers (including Tres Zapotes) has been characterized as being dominated by exclusionary (also called network) political strategies (Blanton et al. 1996: 8). Blanton et al. (1996; see also Blanton 1998; Feinman 2001) describe these strategies as focusing on the personal prestige of individual leaders. This strategy relies on the ability of a leader to build and maintain a faction of supporters (Brumfiel 1994: 4). Features of this strategy include the establishment of a patrimonial rhetoric through artistic expressions focused on leaders (i.e. portraiture) and the participation of leaders in long-distance exchange of high-value prestige goods that can be used to attract supporters. As a consequence, wealth distinctions between members of a society may be pronounced. For the Olmec, the portrait quality of monumental sculptures, like colossal heads, clearly indicates the importance of network strategies, as does the participation of the Olmec in long-distance trade networks for high-value goods, such as jade. Pool (2003a, 2003b; 2008; Pool and Loughlin 2015) has argued that the political organization of Tres Zapotes during the Late Formative period was characterized by corporate political strategies. In contrast to the patrimonial rhetoric of exclusionary strategies, corporate strategies were underpinned by a cognitive code that emphasized group interdependence and solidarity rather than the personal prestige of the ruler (Blanton et al. 1996: 6; Feinman 2001; Lamberg-Karlovsky 1985). As a result, competition between factions was suppressed, and wealth distinctions between members of a society were not as pronounced. At Tres Zapotes, Pool (2000: 150, 2003b, 2010) cites the thematic shift in monumental art away from depictions of specific leaders, the relative lack of high-value exotic items such as jade, and craft production oriented toward utilitarian rather than prestige goods as indicating the importance of corporate strategies at the large center. Perhaps the clearest indicator of the corporate strategies at Tres Zapotes is the architectural redundancy of the four formal architectural complexes at the site. All of these complexes feature a common architectural layout in the form of the TZPG, which, to recap, comprises an east–west oriented plaza bounded by a conical mound to the west, a long mound to the north, and a low mound that served as an adoratorio along the center line of the plaza. Pool (2008: 140) argues that the redundancy in form indicates that each would have had similar integrative functions, as well as expressing a shared vision of governance. That all of these complexes were in operation at the same time, and that none was clearly dominant over the others, suggests that competition between factions was low and that political power was shared between faction leaders. During the Late Formative period, El Mesón was incorporated into the Tres Zapotes regional polity as a secondary center (Loughlin 2012). Architecturally, the relationship between El Mesón and the larger center is indicated by the construction of a TZPG complex in the architectural core of El Mesón. I have suggested (Loughlin 2012) that the reduced scale of this complex relative to the largest TZPG complexes at Tres Zapotes indicates El Mesón’s

subordinate status. What is unclear is whether Tres Zapotes imposed this architectural form upon El Mesón or if it represents local leaders at El Mesón adopting Tres Zapotes’s symbols of political authority. Despite Tres Zapotes’s political influence on El Mesón, the corporate strategies that underpinned the large center’s governance do not appear to have been as strong at El Mesón. In contrast to the “faceless” nature of Tres Zapotes’s Epi-Olmec sculptural corpus, the Epi-Olmec monuments from El Mesón continue the tradition of portraiture started by the Olmec. The two Epi-Olmec monuments recovered from El Mesón show greater stylistic affinity to other EpiOlmec monuments recovered outside of Tres Zapotes, such as the Alvarado Stela and La Mojarra Stela, than to monuments from Tres Zapotes. The continuation of exclusionary strategies at El Mesón likely reflects differences in local political environments. Unlike Tres Zapotes, there is little evidence suggesting the presence of multiple factions at El Mesón. The TZPG within the site’s core was the only formal architectural group in the area with a political function. Although two other formal complexes were in use at the time, both appear to have been associated with craft production rather than governance. Given this lack of political competition, a shift to corporate strategies may have been unnecessary at El Mesón. This difference illustrates how corporate and exclusionary strategies can be combined at different scales within a polity. One question regarding El Mesón’s rise to prominence is why, of all of the small village sites in the area, did El Mesón emerge as a center? Answering this question is difficult, in part, due to the lack of regional survey in the area. Simply stated, we do not know how many Late Formative period sites were in the region. However, the location of the El Mesón area may provide some clues about why a political center emerged there. El Mesón is situated in an alluvial plain approximately 8km from a mountain pass that provides an access point coming into or out of the Tuxtla Mountains (Figure 5.10). Outside of this pass, the steep slopes would have been a significant obstacle to the movement of people and goods into and out of the mountains. Historically, the Spanish Camino Real ran along the north bank of the Tecolapan River into this pass, and today a modern highway runs through it. Presumably, this area would have also been an important transportation corridor during the pre-Columbian period. I suggest that El Mesón’s importance to the Tres Zapotes polity was, in part, related to this location. El Mesón would have been in a position to control the flows of goods moving into and out of the western Tuxtlas. To the South, Tres Zapotes is also situated near a pass into the southern Tuxtlas. By incorporating El Mesón into its polity, Tres Zapotes would have been able to exert considerable influence over the flows of goods through the region.

Figure 5.10

Transportation corridors into the southern and western Tuxtla Mountains.

There is some evidence which suggests that the importance of this route may have been recognized as early as the Middle Formative period. The identification of two Middle Formative Olmec sculptures near El Mesón (Pool et al. 2010) suggests greater Olmec presence in the area than is indicated by surface artifact scatters. What is particularly intriguing about these monuments is that stylistically both show greater similarities to La Venta, located more than 150km to the east (see Figure 5.1), than to Tres Zapotes (Pool et al. 2010). Specifically, costume elements on these monuments are the same as on similar monuments from the more distant center. While admittedly speculative, the presence of the monuments suggests the possibility of La Venta exerting some influence in the region. Participation in long-distance exchange networks was an important feature of Olmec societies during the Early and Middle Formative periods. These networks provided exotic prestige items—such as jade, serpentine, and polished iron ore—that Olmec rulers used in political and religious displays and offerings. Exchange networks spanned Mesoamerica, linking Olman to regions hundreds of kilometers away, including Central Mexico, the Valley of Oaxaca, and the Motogua Valley of Guatemala. In addition to prestige goods, these longdistance networks would also have been used to import exotic utilitarian goods such as obsidian, which was not locally available in the Gulf Lowlands. Obsidian from La Venta has been chemically sourced to both Guatemalan (San Martín Jilotepeque) and Central Mexican (Paredón, Pachuca, Otumba, and Zaragoza) sources (Doering 2002; Hester et al. 1971; Pool 2007: 149–150). If obsidian and other goods from Central Mexico were moved through the Tuxtlas, then La Venta would have had an interest in making sure that its exchange networks would not be affected by the rise of a potential rival center in Tres Zapotes. While there is no evidence that La Venta politically controlled the El Mesón area during the Middle Formative period, these monuments could have been indicative of La Venta’s interest in the area. Such symbols would

have been important if the growth of Tres Zapotes was viewed as a potential threat. Following La Venta’s decline the incorporation of El Mesón as a secondary center would have strengthened Tres Zapotes’s economic dominance of the region. Over the course of the Protoclassic period, the ELPB region experienced another significant political shift that included an intensification of factionalism. The alliances that held Tres Zapotes together broke down and new construction altered the TZPG layouts (Pool 2008: 146). Although these changes were accompanied by a retraction in the size of the site, Tres Zapotes remained an important political and economic center (Pool 2008; Pool and Loughlin 2015). At this time, the El Mesón area broke politically with Tres Zapotes. The TZPG at El Mesón ceased to function and site governance shifted to the Standard Plan complex at El Mesón South. New architectural complexes, featuring large quadrilateral platforms, proliferated throughout the El Mesón area. It has been considered that such platforms in Central and South-Central Veracruz were palaces or elite estates (Daneels 2002, 2008; Stark 2003, 2008). I argue that these complexes in the El Mesón area were most likely associated with faction leaders who were competing for political and economic control of the area (Loughlin 2012). Leaders in the El Mesón area during the Protoclassic and Early Classic periods may have sought to establish themselves politically through extraregional interactions. Pool et al. (2012) suggest that during the Protoclassic and Early Classic periods, two distinct interaction spheres developed that linked centers in the ELPB and Tuxtla Mountains with groups outside the region. The first linked the El Mesón area and Totocapan, located in the Tepango Valley of the Tuxtla Mountains, with the Mixtequilla region in South-Central Veracruz. Evidence for this interaction includes the adoption of Central Veracruz architectural forms and iconography, as well as the presence of ceramics from Central Veracruz in the El Mesón area and at Totocapan. In the El Mesón area, these Mixtequilla types account for just over 11 percent (n = 416) of the recovered Early Classic period ceramics (Loughlin 2012). Through its location along the transportation corridor, El Mesón would have served as an important linkage for goods moving between the two larger centers. The other interaction sphere linked Tres Zapotes with the large center Matacapan, in the Central Tuxtlas, and Teotihuacán in Central Mexico, as evidenced by the presence of Teotihuacán-style ceramics and artifacts at Tres Zapotes and Matacapan. These artifacts were rare in the El Mesón area and at Totocapan. Instead of utilizing the mountain pass near El Mesón, interactions between Matacapan and Tres Zapotes would have gone through the southern pass near Tres Zapotes. It does not appear, however, that all goods would have flowed through these networks. For example, there is no evidence that suggests differential access to obsidian. By the Late Formative period, the majority of obsidian in the region came from the Zaragoza-Oyameles source located in Central Mexico approximately 250km from the ELPB (see Figure 5.1). The ubiquity of this resource throughout the region from the Late Formative through the Classic period suggests that there was little to no restriction on its availability, and that exchange networks for this good were not constrained by local political boundaries (Pool et al. 2012). For much of the Southern Gulf Lowlands, the Late Classic and Early Postclassic periods were marked by political fragmentation and declining populations. In Central Veracruz, Cerro de las Mesas declined as a regional center and its realm split into three different polities

(Stark 2008: 102). Populations were also in decline in the Central and Western Tuxtla Mountains (Santley 2007: 65–75; Santley and Arnold 1996: 236–240; Stoner 2011). By AD 800, the large center of Matacapan was largely abandoned (Santley 2007: 70; Santley and Arnold 1996: 238–239). Stoner (2008, 2011) notes that although Totocapan remained a regional center during the Late Classic period, settlement intensity at the site dropped. In the ELPB, Tres Zapotes and the El Mesón area were not immune to this trend. At Tres Zapotes settlement declined throughout the Late Classic period, and the site was largely abandoned by AD 900 (Pool 2008). In the El Mesón area, populations declined and many of the large quadrilateral platforms in the formal complexes fell into disuse. After AD 900 only a remnant population remained in the area. The cause of this widespread decline is unclear. Santley (2007: 69–70) has suggested that the disruption of obsidian exchange routes following the decline of Teotihuacán may have been a contributing factor, although Stark (2008) has questioned the role that Teotihuacán played in the exchange networks bringing Zaragoza-Oyameles obsidian into the Gulf Lowlands. However, Stark (2008: 104) does note that there was political and economic turbulence following the decline of Teotihuacán. In the Mixtequilla region, Cerro de las Mesas entered a long decline and the area was reorganized as a series of smaller polities. This shift was accompanied by economic changes that saw an intensification of some craft activities, including obsidian blade production. In the ELPB it is unclear how these shifts affected exchange networks that brought important nonlocal goods to the El Mesón area. One possibility could be that with the decline of Cerro de las Mesas, elites in the El Mesón area were slow to reestablish trade partners. Alternatively the orientation of trade networks may have shifted away from the ELPB and Tuxtlas regions. Other evidence suggests that environmental changes at the end of the Late Classic period may also have profoundly affected regional settlement. A lake core from Lago Verde, located in the ELPB, indicates that beginning about AD 800 lake levels rose and environmental conditions became wetter (Lozano-Garcia et al. 2010). Pollen from that core suggests that these wetter conditions were accompanied by the abandonment of agriculture and declining regional populations. It should be noted that not all areas of the Southern Gulf Lowlands experienced this decline equally. In the Hueyapan region of the southern Tuxtlas, Middle and Late Classic period populations were stable (Killion and Urcid 2001). To the east, in the San Juan and Coatzacoalcos drainages after AD 700, populations were expanding. Borstein (2002: 13) suggests that in the San Juan drainage populations may have exceeded the Olmec occupation of the area. Populations also rebounded in the area around the former Olmec center San Lorenzo (Symonds and Lunagómez 1997: 167). It is unclear how or if this population growth is associated with the depopulation of the ELPB. It is possible that if local agricultural systems collapsed, people may have moved back across the Tuxtla Mountains into the Coatzacoalcos Basin, contributing to population increases there.

Summary and conclusion In this chapter, I have argued that the growth and eventual decline of settlement in the El Mesón

area reflect both the importance of the area’s location, as well as the influence of specific political strategies pursued by political leaders both within and outside El Mesón. Like other settlements in the region during the Early and Middle Formative period, the El Mesón area benefited from being located in a region capable of supporting large populations due to its agricultural productivity, and from the abundance of wild food resources that could have been hunted or gathered. However, unlike many of these other settlements, El Mesón was able to expand from small villages and farmsteads to become a political center during the Late Formative period. I suggest that the area’s expansion after 400 BC was, at least in part, related to its location along an important trade route into the western Tuxtlas Mountains. The importance of this location may have been recognized as early as the Middle Formative period. The presence of Middle Formative Olmec monuments carved in the style of La Venta at least hint that the area may have been part of the exchange networks that linked the Olmec center to other regions of Mesoamerica. Following the collapse of La Venta, this location likely made the El Mesón area attractive for inclusion in the regional polity headed by Tres Zapotes. By incorporating El Mesón into its polity as a regional center, Tres Zapotes would have been able to control the flows of goods into and out of the western Tuxtlas. Given Tres Zapotes’s location near a route into the southern Tuxtlas, the center would have been able to dominate economic exchange in the region. During the Protoclassic and Early Classic periods the El Mesón area broke politically from Tres Zapotes, establishing itself as politically independent. I argue that this trade route would have provided opportunities to local elites to exert local control over trade through the area. Specifically, I suggest that by establishing local control over the importation of obsidian and other exotic goods, elites in the El Mesón South complex may have been able to dominate the area economically. Additionally, this location would have also linked the El Mesón area to growing polities in the Tuxtlas Mountains and Central Veracruz. This economic control may have been a means to achieving political control of the area. Although its location may explain why the El Mesón area developed the way that it did, location alone does not adequately explain how the area was spatially or politically organized, or how that organization changed over time. To fully understand how the area developed requires consideration of both the natural and cultural environment in which it was located and in which it operated. I argue that while the area’s location near an important trade route provided a means for leaders to assert political control, the internal organization of settlement within the El Mesón area reflects both the political strategies pursued by local leaders, as well as the regional political climate.

Acknowledgements The research presented here was funded by the Foundation for the Advancement of Mesoamerican Studies, Incorporated (grant 02058); Lambda Alpha National Anthropology Honor Society; and the Graduate School of the University of Kentucky. All of the fieldwork was conducted with authorization of the Instituto Nacional de Antropología e Historia, and the support of the municipal authorities in Ángel R. Cabada, as well as the ejiditarios and local landowners who allowed us to work in their fields. Special thanks goes out to all of the

local workers who walked the cane fields, and to my field assistant Hugo Huerta for all of his efforts in the field and in the lab. Finally, I would like to thank Lucas C. Kellett and Eric E. Jones for inviting me to participate in this volume.

References Blanton, Richard E. 1998 Beyond Centralization: Steps toward a Theory of Egalitarian Behavior in American States. In Archaic States, edited by Gary M. Feinman and Joyce Marcus, pp. 135–172. School of American Research Press, Santa Fe, New Mexico. Blanton, Richard E., Gary M. Feinman, Stephen A. Kowalewski, and Peter N. Peregrine 1996 A Dual-Processual Theory for the Evolution of Mesoamerican Civilization. Current Anthropology 37(1): 1–15. Borstein, Joshua A. 2002 Tripping over Colossal Heads: Settlement Patterns and Population Development in the Upland Olmec Heartland. Unpublished Ph.D. dissertation. Department of Anthropology, Pennsylvania State University, State College, Pennsylvania. Brumfiel, Elizabeth M. 1994 Factional Competition and Political Development in the New World: An Introduction. In Factional Competition and Political Development in the New World, edited by Elizabeth M. Brumfiel and John W. Fox, pp. 2–13. Cambridge University Press, Cambridge. Coe, Michael D. 1965 Archaeological Synthesis of Southern Veracruz and Tabasco. In Archaeology of Southern Mesoamerica, Part 2, edited by G. R. Willey, pp. 679–715. Handbook of Middle American Indians, Volume 3. R. Wauchope, general editor. University of Texas Press, Austin, Texas. Coe, Michael D., and Richard A. Diehl 1980 In the Land of the Olmec Volume I: The Archaeology of San Lorenzo Tenochtitlán. University of Texas Press, Austin, Texas. Covarrubias, Miguel 1957 Indian Art of Mexico and Central America. Alfred A. Knopf, New York. Cyphers, Anne 1996 Reconstructing Olmec Life at San Lorenzo. In Olmec Art of Ancient Mexico, edited by Elizabeth P. Benson and Beatriz de la Fuente, pp. 61–71. National Gallery of Art, Washington, DC. Daneels, Annick 2002 El patron de asentamiento del periodo clásico en la cuenca baja del Río Cotaxtla, centro de Veracruz. Unpublished Ph.D. dissertation. Instituto de Investigaciones Antropológicos Universidad Nacional Autónoma de México, Mexico City. Daneels, Annick 2008 Monumental Earthen Architecture at La Joya, Veracruz, Mexico. Report submitted to the Foundation for the Advancement of Mesoamerican Studies, Inc., Coral Gables, Florida. www​​reports/0702​1/​ml (accessed 1 December 2015). Doering, Travis 2002 Obsidian Artifacts from San Andrés, La Venta, Tabasco, Mexico. Unpublished MS thesis. Department of Anthropology, Florida State University, Tallahassee. Drucker, Philip D. 1968 The El Mesón Monument at Ángel R. Cabada, Veracruz. Archaeological Research Facility, Contributions 5, pp. 41–57. University of California Department of Anthropology, Berkeley. Dunnell, Robert C., and William S. Dancey 1983 The Siteless Survey: A Regional Scale Data Collection Strategy. In Advances in Archaeological Method and Theory, Volume 6, edited by Michael B. Schiffer, pp. 267–287. Academic Press, New York. Feinman, Gary M. 2001 Mesoamerican Political Complexity: The Corporate-Network Dimension. In From Leaders to Rulers, edited by Jonathan Haas, pp. 151—175. Kluwer Academic/Plenum Publishers, New York. Foley, R. 1981 Off-Site Archaeology and Human Adaptation in Eastern Africa. BAR International Series 97. Oxford University Press, Oxford. Gallant, T.W. 1986 “Background Noise” and Site Definition: A Contribution to Survey Methodology. Journal of Field Archaeology 13(4): 403–418. Gómez-Pompa, Arturo 1973 Ecology of the Vegetation of Veracruz. In Vegetation and Vegetational History of Northern Latin America, edited by Alan Graham, pp. 73–148. Elsevier Scientific Publishing Company, New York. Hester, T.R., R.F. Heizer, and R.N. Jack 1971 The Obsidian of Tres Zapotes, Veracruz, Mexico. Contributions to the University of California Archaeological Research Facility 13: 65–131. Jones, Eric E. 2010 An Analysis of Factors Influencing Sixteenth and Seventeenth Century Haudenosaunee (Iroquois) Settlement Locations. Journal of Anthropological Archaeology 29: 1–14. Jones, Eric E., Madison Gattis, Thomas C. Morrison, Andrew Wardner, and Sara Frantz 2012 Exploring Tribal Settlement Ecology in the Southeast: A Case Study from the North Carolina Piedmont, 800–1600 CE. North American Archaeologist 33(2): 159–192. Killion, Thomas, W., and Javier Urcid 2001 The Olmec Legacy: Cultural Continuity and Change in Mexico’s Southern Gulf Coast Lowlands. Journal of Field Archaeology 28: 3–25. Kruszczynski, Mark 2001 Prehistoric Basalt Exploitation and Core-Periphery Relations Observed from the Cerro el Vigia

Hinterland of Tres Zapotes, Veracruz, Mexico. Unpublished Ph.D. dissertation. Department of Anthropology, University of Pittsburgh, Pittsburgh, Pennsylvania. Lamberg-Karlovsky, C.C. 1985 The Near Eastern “Breakout” and the Mesopotamian Social Contract. Symbols Spring: 8–11, 23–24. Loughlin, Michael L. 2012 El Mesón Regional Survey: Settlement Patterns and Political Economy in the Eastern Papaloapan Basin, Veracruz, Mexico. Unpublished Ph.D. dissertation. Department of Anthropology, University of Kentucky, Lexington, Kentucky. Lozano-García, Socorro, Margarita Caballero, Beatriz Ortega, Susana Sosa, Alejandro Rodríguez, and Peter Schaaf 2010 Late Holocene Paleoecology of Lago Verde: Evidence of Human Impact and Climate Change in the Northern Limit of the Neotropics during the Late Formative and Classic Periods. Vegetation History and Archaeobotany 19(3): 177–190. Pool, Christopher A. 2000 From Olmec to Epi-Olmec at Tres Zapotes, Veracruz, Mexico, In Olmec Art and Archaeology in Mesoamerica, edited by John E. Clark and Mary E. Pye, pp. 137–153. Symposium Papers 35. National Gallery of Art, Washington, DC. Pool, Christopher A. 2003a Centers and Peripheries: Urbanization and Political Economy at Tres Zapotes. In Settlement Archaeology and Political Economy at Tres Zapotes, Veracruz, Mexico, edited by Christopher A. Pool, pp. 90–98. Cotsen Institute of Archaeology Monograph 50. University of California, Los Angeles. Pool, Cristopher A. (editor) 2003b Settlement Archaeology and Political Economy at Tres Zapotes, Veracruz, Mexico. Cotsen Institute of Archaeology Monograph 50. University of California, Los Angeles. Pool, Christopher A. 2007 Olmec Archaeology and Early Mesoamerica. Cambridge University Press, Cambridge. Pool, Christopher A. 2008 Architectural Plans, Factionalism, and the Proto-Classic Transition at Tres Zapotes. In Classic Period Cultural Currents in Southern and Central Veracruz, edited by Philip J. Arnold III and Christopher A. Pool, pp. 121–157. Dumbarton Oaks Research Library and Collections, Washington, DC. Pool, Christopher A. 2010 Stone Monuments and Earthen Mounds: Polity and Placemaking at Tres Zapotes, Veracruz. In The Place of Stone Monuments: Context, Use, and Meaning in Mesoamerica’s Preclassic Past, edited by Julia Guernsey, John E. Clark, and Barbara Arroyo, pp. 97–128. Dumbarton Oaks, Washington, DC. Pool, Christopher A., and Michael L. Loughlin 2015 Tres Zapotes: The Evolution of a Resilient Polity in the Olmec Heartland of Mexico. In Beyond Collapse: Archaeological Perspectives on Resilience, Revitalization, and Transformation in Complex Societies, edited by Ronald K. Faulseit, pp. 287–309. Southern Illinois University Press, Carbondale. Pool, Christopher A., and Michael A. Ohnersorgen 2003 Archaeological Survey and Settlement at Tres Zapotes. In Settlement Archaeology and Political Economy at Tres Zapotes, Veracruz, Mexico, edited by Christopher A. Pool, pp. 7–31. Cotsen Institute of Archaeology Monograph 50. University of California, Los Angeles. Pool, Christopher A., Ponciano Ortíz Ceballos, María del Carmen Rodríguez Martínez, and Michael L. Loughlin 2010 The Early Horizon at Tres Zapotes: Implications for Olmec Interaction. Ancient Mesoamerica 21: 95–105. Pool, Christopher A, Wesley D. Stoner, and Michael L. Loughlin 2012 The Formative–Classic Transition on the Western Tuxtlas Frontier: Continuities and Disjunctures. Paper presented at the 54th International Congress of Americanists, July 15–20, Vienna, Austria. Pope, K.O., M.D. Pohl, J.G. Jones, D.L. Lentz, C.L. von Nagy, F.J. Vega, and I.R. Quitmyer 2001 Origin and Environmental Setting of Ancient Agriculture in the Lowlands of Mesoamerica. Science 292: 1370–1373. Rust, W.F., and B.W. Leyden 1994 Evidence of Maize Use at Early and Middle Preclassic La Venta Olmec Sites. In Corn and Culture in the Prehistoric New World, edited by S. Johannessen and C.A. Hastorf, pp. 181–201. Westview Press, Boulder, Colorado. Santley, Robert S. 2007 The Prehistory of the Tuxtlas. The University of New Mexico Press, Albuquerque, New Mexico. Santley, Robert S., and Phillip J. Arnold III 1996 Prehispanic Settlement Patterns in the Tuxtla Mountains, Southern Veracruz, Mexico. Journal of Field Archaeology 23(2): 225–249. Scott, John F. 1977 El Mesón, Veracruz, and its Monolithic Reliefs. Baessler-Archiv:Beiträge zur Völkerkunde 25: 83–138. Soto, Margarita and Lilly Gama 1997 Climas. In Historia natural de los Tuxtlas, edited by Enrique González Soriano, Rodolfo Dirzo, and Richard C. Vogt, pp. 7–23. Universidad Nacional Autónoma de México, Mexico City. Stark, Barbara L. 1991 Survey Methods and Settlement Features in the Cerro de Las Mesas Region. In Settlement Archaeology of Cerro de Las Mesas Veracruz, Mexico, Institute of Archaeology Monograph 34. University of California, Los Angeles. Stark, Barbara L. 2003 Cerro de las Mesas: Social and Economic Perspectives on a Gulf Center. In Urbanism in Mesoamerica Volume 1, edited by William T. Sanders, Alba Guadalupe Mastache, and Robert H. Cobean, pp. 391–426. Instituto Nacional de Antropología e Historia, Mexico City and Pennsylvania State University, University Park, Pennsylvania. Stark, Barbara L. 2008 Polity and Economy in the Western Lower Papaloapan Basin. In Classic Period Cultural Currents in Southern and Central Veracruz, edited by Philip J. Arnold III and Christopher A. Pool, pp. 85–120. Dumbarton Oaks Research Library and Collections, Washington, DC. Stirling, Matthew W. 1943 Stone Monuments of Southern Mexico. Bureau of American Ethnology, Bulletin 138. Smithsonian Institution, Washington, DC.

Stone, Glenn D. 1996 Settlement Ecology: The Social and Spatial Organization of Kofyar Agriculture. University of Arizona Press, Tucson. Stoner, Wesley D. 2008 Tepango Valley Archaeological Survey: Tuxtla Mountains, Southern Veracruz, México. Report submitted to the Foundation for the Advancement of Mesoamerican Studies, Inc., Coral Gables, Florida. www​​ports/0704​9/​ml (accessed 10 January 2016). Stoner, Wesley D. 2011 Disjuncture among Classic Period Cultural Landscapes in the Tuxtla Mountains, Southern Veracruz, Mexico. Unpublished Ph.D. dissertation. Department of Anthropology, University of Kentucky, Lexington, Kentucky. Symonds, Stacey C., and Roberto Lunagómez 1997 Settlement System and Population Development at San Lorenzo. In Olmec to Aztec: Settlement Patterns in the Ancient Gulf Lowlands, edited by Barbara L. Stark and Philip J. Arnold III, pp. 144– 173. University of Arizona Press, Tucson. Symonds, Stacey C., Ann Cyphers, and Roberto Lunagómez 2002 Asentamiento prehispánico en San Lorenzo Tenochtitlán, Veracruz, Mexico. Instituto de Investigaciones Antropológicas, Universidad Nacional Autónoma de México, Mexico City. VanDerwarker, Amber M. 2003 Agricultural Intensification and the Emergence of Political Complexity in the Formative Sierra de los Tuxtlas, Southern Veracruz, Mexico. Ph.D. dissertation, University of North Carolina, Chapel Hill. UMI, Ann Arbor, MI. VanDerwarker, Amber M. 2006 Farming, Hunting, and Fishing in the Olmec World. University of Texas Press, Austin. Wendt, Carl J. 2003 Buried Occupational Deposits at Tres Zapotes: The Results from an Auger Testing Program. In Settlement Archaeology and Political Economy at Tres Zapotes, Veracruz, Mexico, edited by Christopher A. Pool, pp. 32–46. Cotsen Institute of Archaeology, Monograph 50. University of California, Los Angeles. Wing, E.S. 1980 Aquatic Fauna and Reptiles from the Atlantic and Pacific Sites. In In the Land of the Olmec, edited by Michael D. Coe and Richard A. Diehl, pp. 375–386. University of Texas Press, Austin.

6 Climate, ecology, and social change in prehispanic northwestern Mesoamerica Michelle Elliott

Introduction Mesoamerican archaeologists have long been aware that several major socio-political shifts occur red in multiple regions of the culture area, beginning in the middle Classic period (AD 400–700) and culminating in the Late/Terminal Classic (AD 800–900/1000). These include the collapse of the Teotihuacán state in Central Mexico, as well as the subsequent fall of various political entities in the southern Maya Lowlands. The Late to Terminal Classic period in Mesoamerica correlates with social, political, and economic disruption that may have been provoked, worsened, or at the very least influenced by a number of environmental factors. Indeed, increasingly sophisticated paleoclimatic studies carried out in areas such as Central Mexico and the Yucatan Peninsula indicate that this period was characterized by important fluctuations in annual precipitation that led to massive drought events (Bernal et al. 2011; Haug et al. 2003; Hodell et al. 1995, 2000, 2001; Stahle et al. 2011). These data raise a number of questions regarding the spatial scale at which these climatic phenomena occurred and the specific mechanisms by which they could have provoked social upheaval and contributed to political breakdowns. A lesser-known region of settlement, the northern frontier zone of Mesoamerica, is also implicated in this pattern of cultural collapse toward the end of the Classic period. Although it has not enjoyed the same level of attention from archaeologists, geographers, and paleoecologists as the Basin of Mexico or the Maya Lowlands, the settlement of this region witnessed the spread of Mesoamerican farming economies, exchange networks, and ideologies into territory previously dominated by mobile hunter-gatherer groups, and represented the largest expansion of the Mesoamerican culture area. Furthermore, the uniformity of its abandonment toward the end of the Classic period remains poorly understood. For over 50 years, archaeologists have implicated environmental factors as crucial for understanding the trajectories of cultural growth and collapse in the northern frontier zone (Armillas 1964; Braniff 1989; Braniff and Hers 1998; Coe 1994). Yet, little paleoenvironmental research has been carried out within the region to test these hypotheses. It is only recently that there has been more attention directed at integrating archaeological and paleoecological data at multiple temporal and spatial scales, highlighting possible linking

mechanisms between cultural developments and climatic and ecological phenomena within the region. Despite the relative infrequency of linked socio-paleoenvironmental work in the northern frontier zone compared with other areas of Mesoamerica, the region nevertheless presents an interesting alternative case-study for investigating the widespread changes in settlement patterns that span the Middle to Late Classic throughout Mesoamerica. Furthermore, settlement ecology approaches, which seek to link traditional archaeological settlement data with a wide range of environmental and other cultural data (Stone 1996), appear to have great potential for improving our understanding of how this vast region came to be dominated and then abandoned by Mesoamerican farmers in a span of approximately 700 years. In this chapter, I present an overview of the studies that have been carried out in the northern frontier region to date, beginning with more traditional approaches to understanding settlement change (i.e. studies that consider cultural and ecological factors separately). I then present more recent studies that seek to integrate cultural and environmental data at multiple scales to better understand the dynamics of human–landscape interactions, discussing the new perspectives that this research has provided regarding our understanding of the northern frontier zone. Finally, I conclude with some thoughts on the future of research in the region. From a purely archaeological perspective, the northern frontier offers a promising data set for understanding the complex range of factors that influence human–landscape interactions through time. But this research also has potential for a better understanding of the modern settlement ecology of the region. In particular, it provides long-term data sets that may prove useful for addressing the precarious conditions that exist for subsistence agriculture in the zone today, and which have a significant influence over modern economics, politics, and migration patterns.

The northwest frontier zone The Mesoamerican northern frontier zone is an expanse of roughly 100,000 km²—located principally on the Central Mexican Plateau—whose southern boundary follows the course of the Río Lerma-Santiago (Armillas 1964) (Figure 6.1). The mean elevation is 2,000 m above sea level (masl). This chapter focuses specifically on the western portion of the frontier, roughly following the Sierra Madre Occidental mountain range, whose ecology and cultural developments differ from those found further east, along the Sierra Madre Oriental. Along the southern margin of the region, known as the Bajío (encompassing the modern states of Guanajuato, Querétaro, Aguascalientes, and the Altos of Jalisco), precipitation averages 700–800 mm annually (Secretaría de Programación y Presupuesto 1980). The vegetation consists of grassland and patches of sub-tropical deciduous forest in low-lying areas, and pine-oak forest in the mountains (Armillas 1964:63; Labat 1995; Rzedowski 1981). This southern boundary of the frontier zone marks the modern limit for non-irrigated maize agriculture; in the sixteenth century it was an area of cultural division, with groups of sedentary farmers to the south and mobile societies in the semi-arid north (Armillas 1964:65). Currently, the Bajío is a productive agricultural zone in Mexico.

Figure 6.1

Map of the northern frontier of Mesoamerica with the locations of the archaeological (circles) and paleoecological (stars) study sites mentioned in the text. The darker area represents the northern frontier zone that existed during the Classic period (map by Grégory Pereira).

As one moves northwest toward the mountains of the Sierra Madre Occidental, rainfall gradually decreases to as low as 400 mm annually, and the landscape transitions to semi-arid steppe grassland and desert vegetation (e.g. cactus, acacia, mesquite). Mountainous zones that have not been subject to historic and modern deforestation generally consist of oak (Quercus sp.) woodland at their lower elevations, and pine (Pinus sp.), fir (Abies sp.), and Douglas fir (Pseudotsuga menziesii) at their higher extremes. It is interesting to note that at the northern extreme of the frontier zone (modern-day Zacatecas and southern Durango) environmental crises (particularly drought) are frequent today, and although subsistence farming is widely practiced in this rural landscape, crop failures are very common. Some modern farmers in southern Zacatecas report that they achieve a successful harvest once every five years. Ecological risk for these subsistence farmers has wide-reaching implications, with governmental intervention often necessary in order to supply adequate resources to local populations, particularly notable during times of multi-year drought. Out-migration from modern states like Zacatecas due to environmental factors is a constant issue that further undermines local economies. Observing these dynamics today, and seeing the difficulties faced by local communities (even with modern technology and wide economic networks to rely on), raises a number of questions regarding how pre-industrial farmers with wood- and stone-based technology managed to persist here for at least four centuries.

Initial archaeological and paleoenvironmental research in the northern frontier zone

While mapping, survey, and description of archaeological remains in the northern frontier zone began in the nineteenth century (e.g. Batres 1903; Berghes 1996), it was not until the midtwentieth century that archaeologists began to systematically investigate the region’s settlement patterns, establish and refine the chronological sequences, and consider potential links among prehispanic occupations, climate, and ecological factors. A detailed discussion of the Classic period settlement dynamics of the region is beyond the scope of this chapter, so I present only a brief summary below. A number of other works can be consulted for a more comprehensive discussion (e.g. Armillas 1964; Braniff 1989; Braniff and Hers 1998; Cabrero 1991; Jiménez Betts 1992; Jiménez Betts and Darling 2000; Kelley 1956, 1974, 1985; Medina González 2000; Nelson 1992, 1993, 1995, 1997; Nelson et al. 1992; Trombold 1978, 1990, 1991; Weigand 1978a, 1978b, 1982). Since early on, archaeologists have noted the presence of permanent agricultural settlements north of the Rio Lerma-Santiago that demonstrate clear Mesoamerican characteristics. However, these communities differed from those to the south, in the Mesoamerican heartland of West and Central Mexico, where a gradual evolution of mobile or semi-sedentary groups to settled agricultural villages can be observed archaeologically over the course of the Formative period (2000 BC–AD 200), and larger, urban cites flourish in the Classic period (AD 200–900) (Braniff 1989, 2000; Kelley 1989). In contrast, agricultural settlements in the northern frontier zone appear abruptly as fully developed Mesoamerican complexes starting around the beginning of the Early Classic period. Archaeologists have interpreted this phenomenon as evidence of a colonization that originated in the Mesoamerican core of West and Central Mexico. This northward wave of expansion thus begins around AD 200, peaks between AD 500 and 900, and widespread collapse of frontier settlements occurs by AD 900/1000. After the Classic period abandonment, mobile and semi-sedentary groups once again dominate the region until the arrival of the Spaniards in the sixteenth century (Weigand 1978a, 1978b). The relations between the more recently arrived agricultural Mesoamerican populations and the local hunter-gatherer groups (Chichimeca) are not well understood. Some researchers have suggested they were characterized by conflict and violence (Armillas 1964; Trombold 1991). To support this idea they cite the typical Classic period frontier settlement pattern that consists of apparently fortified centers (monumental architecture located on hilltops) surrounded by networks of smaller sites below. However, other scholars have questioned such oppositional models and propose that relations were more complex, with networks of cultural transmission, exchanges, or complementarity between settled and mobile groups as other possible models for understanding the “Mesoamericanization” of this contact zone (Kelley 2000). The numerous settlement systems in the northern frontier begin in the lacustrine basins and alluvial valleys of the Bajío (on the southern boundary of the region) and initially show ties to the Chupícuaro culture of Formative period West Mexico, through ceramic types and iconography. More settlements appear through time in the series of alluvial valleys that characterize the more northerly reaches of the frontier. These continue to show cultural ties to the Bajío through ceramic types such as Blanco Levantado, Red-on-Buff, and incised/engraved fine grey-black wares (Braniff 2000:39). The presence of architectural elements such as pyramids, colonnaded halls, I-shaped ball courts, and house groups organized around sunken patios (patios hundidos) also link these sites culturally to the Mesoamerican heartland

(Cárdenas García 1999; Jiménez Betts and Darling 2000; Nelson 1993, 1997; Trombold 1991). Despite these general similarities, researchers have also noted intra-regional variability (particularly in ceramic vessels and figurines), with apparently different cultural groups distributed across the zone. Moving from south to north, these include the Bajío groups, the Altos of Jalisco, the Bolaños tradition, the settlements in the Malpaso-Juchipila zone, and the Suchil and Guadiana branches of the Chalchihuites tradition in the extreme north (Braniff 2000; Cabrero 1991; Jiménez Betts and Darling 2000; Kelley 1985). Despite differences in architectural styles and ceramic assemblages, the settlements themselves share a trend toward similarly sized groups of villages and hamlets clustered around larger centers with monumental architecture. Jiménez-Betts (1992) argues that these settlement systems should be understood as a wide network of competing and interacting peer-polities. I present in greater detail below one such typical settlement system, the Malpaso Valley. Archaeologist Pedro Armillas proposed in his seminal 1964 publication that this phenomenon of “Mesoamericanization” in the northern frontier zone was linked directly to fluctuations in precipitation cycles correlated with the Medieval Warm Period and the Little Ice Age. His hypothesis was based on archaeological data and global climate proxies derived from ice cores sampled in Greenland. Armillas (1964:79) argued that the higher sea surfaces associated with the Medieval Warm Period (AD 950–1250) would have increased rainfall in the northern frontier zone and allowed the expansion of Mesoamerican farming groups northward. The onset of the subsequent Little Ice Age around AD 1400 would have decreased regional precipitation and forced the retreat of agricultural communities back toward the Mesoamerican heartland. In this work, Armillas anticipates the emergence of settlement ecology approaches, strongly encouraging his colleagues to undertake interdisciplinary “attacks” to test this hypothesis at study zones within the frontier (Armillas 1964:80). These investigations would involve examining diverse lines of data such as settlement patterns, resource exploitation, vegetation, fauna, and soils. Armillas himself planned a long-term project of environmental archaeology in the Malpaso Valley of southern Zacatecas (at the northern extreme of the frontier), but ultimately was unable to carry out this research. Roy B. Brown was one of the first to take up Armillas’s 1964 exhortation for interdisciplinary study of the northern frontier zone. Brown (1984, 1992) reexamined archaeological excavation data and radiocarbon dating at several sites throughout the frontier region and refined the period of Mesoamerican settlement to AD 600–900/1000, which corresponds to the Classic period that follows the collapse of the Teotihuacán state in Central Mexico. Subsequent radiocarbon dating of sites within the Malpaso Valley of southern Zacatecas (Lelgemann 2000; Nelson 1997; Trombold 1990) conforms to Brown’s (1984, 1992) chronology. This refinement of the chronology of the Mesoamerican occupation in the zone is important not only for dating the phenomenon of settlement expansion, but because it also indicates that Armillas’s (1964) model of linked socio-environmental factors does not correlate well with the timing of the occupation. The climatic phenomena Armillas (1964) invokes occurred well after the abandonment of the northern frontier farming settlements. These results indicate the need to reevaluate the possible climatic factors involved, as well as

the mechanisms by which they would have influenced the settlement strategies of Mesoamerican farmers in the north. In addition to refining the archaeological chronology for the northern frontier zone, Brown (1984, 1992) carried out a palynological study on four sediment cores to test the “Armillas Hypothesis” of increased aridity in the Late Classic. Three of the cores were sampled from locales bordering the frontier zone: two from West Mexico (Laguna de San Pedro, a 6-km² lake in Nayarit; and a 21-km² lake in Jalisco, Lago Guzmán) and one from the Bajío zone (San Nicolás de San Parangueo, a 2.4-km² explosion crater in Guanajuato). The fourth core came from Ojo de San Juan, a spring in San Luís Potosí, at the northern extreme of the frontier. While Brown (1984, 1992) noted evidence of human impact through time, he was unable to find clear evidence of a Late Classic period drought. Other researchers, principally physical geographers and geologists, have carried out a number of multi-proxy sedimentary studies in the lacustrine zone of the Trans-Mexican Volcanic Belt that follows the southern edge of the frontier (Figure 6.1). As this research has been well summarized elsewhere (e.g. Metcalfe 1997; Metcalfe and O’Hara 1997), I present only a few highlights here. Metcalfe et al. (1989:130) find little evidence for Late Classic climate change at the Piscina de Yuriria (a 0.75-km² lake in southern Guanajuato), although they do detect evidence of human impact that correlates chronologically with Brown’s (1984, 1992) Guanajuato sequence. In contrast, a core extracted from a former marsh in the Zacapu Basin of northern Michoacan (74 km southwest of the Yuriria Basin) indicates two major lake level regressions: the first from 2800–2400 BP (c.850–450 BC) and the second around 1100 BP (c. AD 850) (Metcalfe et al. 1989). Continuing 30 km south, to the Pátzcuaro Basin of Michoacan, evidence of lowered lake levels toward the end of the Classic period is detected. O’Hara et al. (1993) argue for a period of drought from 1200 to 850 BP (c. AD 750 to 1100). Fisher et al. (2003:4960) confirm a low lake level in the basin from AD 775 to 1350. Research in the nearby Zirahuén Basin (20 km southwest of Pátzcuaro) indicates low lake levels around AD 1000–1200 and little evidence of erosion, suggesting drought rather than human landscape disturbance (Davies et al. 2004:92). Interestingly, although sedimentary data from Lake Chignahuapan in the Upper Rio Lerma drainage, 250 km east of the Zacapu Basin, support the pattern of Late Classic drought, archaeological data from the same study indicates this period corresponds also to the peak of occupation in the zone (Caballero et al. 2002). However, data derived from the northern edge of the frontier zone tell a different story. Frederick’s (1995) geomorphological study in the Río Lajas region of northern Guanajuato (approximately 100 km north of the Michoacan lakes zone) does not provide evidence of significant changes in annual rainfall patterns for the Classic period, and major human environmental impacts are not detected until the eighteenth century (Frederick 1995:250). In the “Laguna Project,” Butzer et al. (2008) examined alluvial sediments in Durango and Coahuila (just beyond the northern edge of the frontier zone) and their findings indicate that the period AD 1050–1200 is marked by cycles of excessive precipitation that led to extensive erosion and natural landscape degradation. The comparison of these different findings suggests that the landscapes, climate, and human settlement patterns across the frontier zone varied through time and space. While drought conditions and/or human-induced erosion affected the southern edge of the frontier zone during

the Classic period, the northern edge seems instead characterized for the same time period by an overabundance of rainfall and resultant natural erosion processes. However, few data points have been collected for the intervening areas, making it difficult to evaluate how far each of these contrasting climatic tendencies penetrated into the region. Furthermore, the archaeological data suggest that it is not clear that more arid conditions would necessarily increase difficulties for farming groups, as proposed by Armillas (1964). In the lacustrine zones that were a focus for Mesoamerican settlement, particularly toward the southern boundary of the frontier zone, decreased lake levels increased the availability of arable land, and demographic peaks are detected for these drier periods (Caballero et al. 2002; Fisher et al. 2003). These cultural and paleoecological findings suggest that to better understand the complex settlement dynamics of the northern frontier zone in the Classic period, it is crucial to obtain multiple lines of data from a variety of points across this vast region. In the following section, I present data that have been gathered within the frontier zone itself over the past several years with the goal of addressing archaeological questions regarding human–landscape interaction and settlement change.

Figure 6.2

Location of the Malpaso Valley (map by Ben Nelson).

The Malpaso Valley, Zacatecas The majority of the studies I present here were carried out in the Malpaso Valley of southern Zacatecas (Figure 6.2). This valley and its prehispanic settlements have been the focus of archaeological research since the nineteenth century and thus present a well-developed base of cultural data to compare with environmental variables (e.g. Batres 1903; Berghes 1996; Elliott 2005; Kantor 1995; Lelgemann 2000; Millhauser 1999; Nelson 1990, 1992, 1993, 1995, 1997; Nelson et al. 1992; Nelson and Martin 2015; Trombold 1978, 1990, 1991, 2005; Turkon 2004; Wells 2000). Located in the eastern foothills of the Sierra Madre Occidental, the Malpaso Valley lies at 2,140–2,350 masl and receives an average of 500 mm of precipitation annually, principally between the months of June and August (Secretaría de Programación y Presupuesto 1981). Today the valley is a semi-arid landscape, deeply incised with extensive arroyos and associated colluvial fans of transported upland materials. The Malpaso River was the valley’s principal perennial watershed, flowing from north to south, until it was dammed during the Colonial period to form a reservoir just north of the archaeological site of La Quemada. The strip of remnant floodplain that surrounds the non-inundated river channel near the modern village of La Quemada is the current focus for small-scale irrigation agriculture. Dryland farming is practiced away from the floodplain. The region around the Malpaso Valley can be described as Chihuahuan desert grading into low mountain steppe that is characterized by desert scrub, crassicaulescent scrub, chaparral, and pastizal (Matson and Baker 1986; Rzedowski 1981; Turkon 2002). This habitat is dominated by various species of grasses and is generally analogous to North American shortgrass prairie or Andean puna (Rzedowski 1981:215). The composition of pastizal is maintained in part by climatic and edaphic conditions, but disturbance from human farming and animal husbandry is also a component (Rzedowski 1981:215). While some shrubs (e.g. Acacia schaffneri, Opuntia sp., and Prosopis laevigata) occur in the pastizal of southern Zacatecas, they tend to result from local human disturbance (Rzedowski 1981:220). The Malpaso Valley’s principal archaeological site is La Quemada, a large hilltop center made up of more than 60 residential terraces and other monumental construction (Figure 6.3). Ceramics and architecture incorporate several Mesoamerican elements such as the I-shaped ball court, sunken patio complexes, colonnaded halls, and ceramic iconography. Based on ceramic and radiocarbon evidence, the site was founded by AD 500 (Nelson 1997). Scholars debate the precise timing of the abandonment, but after AD 900, La Quemada probably was no longer occupied continuously (Jiménez and Darling 2000; Lelgemann 2000; Nelson 2003). La Quemada is clearly a significant site in the valley. It is visually imposing, as it is one of only a few sites that are located on prominent hilltops. Of these elevated sites, La Quemada (230 m above the floodplain) is the largest by several orders of magnitude (50 ha compared with 5 ha for the next largest group of secondary sites). Armillas (1969:700) describes La Quemada as “a castle-town, an extravagantly fortified settlement on a natural stronghold,” referring to the model of conflict between agriculturalists and hunter-gatherer groups. Yet, as is discussed below, it is not clear that the site’s configuration and position served a primarily defensive purpose. Other researchers have suggested La Quemada existed for ritual reasons.

The presence of a significant civic-ceremonial core that could accommodate large populations helps to support the ritual function hypothesis. The presence of several major causeways that lead into the site also lends credence to the ceremonial function hypothesis, as ritual processions are frequent among modern-day indigenous groups in the region (Medina González 2000). Nevertheless, excavation data indicate that there was also a significant permanent occupation on La Quemada’s many residential terraces, with each the focus of at least one house group organized around a central patio (Nelson 1993, 1997). Nevertheless, we do not understand well the identity of these families. Were they permanent residents, or did they inhabit the house groups for fixed periods of time, perhaps as representatives of residential groups located on the valley floor (Wells 2000)? Did they enjoy a higher status than those families living below the site? Despite its imposing scale, traditional indicators of elevated social status are difficult to detect at the site and the material culture at small and large sites in the valley seems fairly similar throughout the Classic period occupation (Turkon 2002, 2004).

Figure 6.3

Contour map of La Quemada (map by Ben Nelson).

Figure 6.4

Map of the Malpaso Valley showing the location of sites, the Malpaso River, and seasonal streams (adapted from Trombold 1991).

The La Quemada polity also includes more than 200 much smaller contemporaneous settlements in an area of 10 × 12 km (Trombold 1991) (Figure 6.4). Many of these settlements were linked to each other and to La Quemada by an extensive system of ancient roads that total more than 175 km in length. Medina González (2000) demonstrates that the road network’s configuration reflects fundamental cosmological concepts of the modern Huichol Indians who live in the Sierra Madre to the northwest, which makes it likely that the roads were key symbols in ritual activities, such as processions. The majority of sites in the valley are villages or hamlets that measure less than 1 ha. The settlement system is thus maximally three-tiered with estimates for the total valley population ranging between 2,500 and 9,155 (Nelson 1995; Trombold and Israde 2005). Sometime after abandonment by the Classic period occupation, the valley was occupied by small groups of hunter-gatherers and horticulturalists, who were present when the Spaniards arrived (Weigand 1978a, 1978b). Historical documents refer to small encampments of Guachichiles and Zacatecos occupying the desert region (Kelley 1974), while more sedentary groups such as the Cora and Huichol were settled in the mountainous regions that they continue to occupy today. In abandoning the center, the population may have either left the area or reorganized itself into smaller, more mobile groups. The Spanish colonial period that began in the early sixteenth century significantly altered settlement dynamics, economics, and ecology throughout the northwestern frontier zone. Zacatecas quickly became the principal center in New Spain for silver production. Local forests were severely impacted by the mining industry (Bakewell 1971; Medina Gonzalez 2000). This deforestation had long-term consequences for the local landscape, resulting in increasing soil erosion and the loss of mammal and bird species adapted to forested patches. Today the Malpaso Valley is primarily used for subsistence farming (particularly maize and beans) and cattle and pig husbandry. Families forage in the areas of scrub that surround the floodplain fields to collect firewood and some wild plant foods, such as cactus fruits. Livestock is often allowed to graze in these areas as well.

Recent interdisciplinary study in the Malpaso Valley A number of studies have been carried out in the Malpaso Valley in recent years that fall under the rubric of a settlement ecology approach. They share a common goal to better understand the links between settlement patterns and environmental factors through interdisciplinary approaches. Spatial analysis of settlement pattern environmental factors Elliott (2005) investigated the distribution of sites in the valley using quantitative GIS analysis to evaluate hypotheses regarding the occurrence of violent conflict or warfare in the valley, and the importance of access to soils and water necessary for agricultural success. Researchers have suggested that conflict or warfare was prevalent in the northern frontier zone during the Classic occupation, perhaps as a result of tension over access to resources between the indigenous hunter-gatherer population and the newly arrived Mesoamerican farming groups (e.g. Armillas 1964, 1969; Braniff and Hers 1998; Hers 1989; Nelson 2000; Weigand 1978a,1978b). The evidence most often cited to support this theory includes the repeated pattern of hilltop centers with monumental architecture throughout the zone and dense concentrations of human skeletal remains (in mass burials, often sorted by anatomical element) at sites. In the case of the Malpaso Valley, archaeologists have noted La Quemada’s size, apparent system of fortification, abundance of mortuary remains (in public and private contexts), and hilltop location as evidence that it served as a defensive stronghold for the valley’s population (Armillas 1964; Lelgemann 2000; Nelson et al. 1992; Trombold 1991). The results of Elliott’s (2005) spatial analysis indicate that the settlement pattern of the Malpaso Valley is not well explained by expectations for defensive concerns when compared with the size and configurations of sites in other regions, such as Central Mexico and the United States Southwest, known to be associated with sustained conflict or violence (e.g. Armillas 1951; Haas and Creamer 1996; Leblanc 2007). The stronger case can be made against small-scale violence and warfare, aimed at the valley’s general population. The majority of the settlements are small villages and hamlets (< 1 ha) that display no detectable fortifications and are typically located in exposed locations on the valley’s floor. They are located on average 5 km from La Quemada, a distance that likely would have required advance warning if these populations looked to the larger site as a refuge. The possibility of ritualized warfare (preplanned confrontations between declared enemies) remains more of an open question. La Quemada’s large size, elevated location, and monumental architecture would be logical responses to regular aggression. The data are less ambiguous with regard to the second hypothesis proposed: that the valley’s site distribution conforms to the expectations associated with maximizing access to the zones best suited for farming, even when these locations would have reduced the defensibility of the sites. Sixty-seven percent of the sites are located in exposed locations on the Malpaso Valley’s floor, but all are within less than 5 km of high-quality patches of alluvial soil. Furthermore, a hydrological flow model was used to evaluate the agricultural hypothesis based on

observations of traditional farming techniques in other semi-arid regions, such as those used by the Hopi and family-based farms in Oaxaca, Mexico. The results of this spatial modeling show that all sites in the valley are located in close proximity (a distance of 1–0.5 km) to zones that either permit small-scale canal irrigation or receive sufficient seasonal rainfall runoff to allow for crop growth (Figure 6.5). These results caution that hypotheses about the prevalence of warfare and conflict in the valley should be reexamined and refined, and that further study of the systems of the Classic agricultural systems is warranted. Paleoethnobotanical analysis Study of carbonized plant remains at sites throughout the valley has allowed a better understanding of the range of plant resources exploited, their patterning through time, and the potential resource-management strategies that were practiced. Turkon’s analysis of seed and other plant food remains at sites on all three levels of the valley’s settlement hierarchy demonstrates that the inhabitants were equally dependent on domesticated crops (such as maize and beans) as they were on gathered wild plant foods (such as agave, cactus, and wild seed plants) (Turkon 2002, 2004; Weintraub 1992). This mixed economy seems to have remained stable over the duration of the valley’s occupation, suggesting careful management of agricultural and non-cultivated patches of resources by the inhabitants. It may also hint at shorter-term climatic fluctuations that required a diversification of resources to feed the population.

Figure 6.5

Results of Elliott’s (2005) hydrological flow modeling of the Malpaso Valley. Grey patches represent the areas where farming is possible through either small-scale irrigation or rainfall runoff. Prehispanic sites are represented by dots, with the exception of La Quemada, which is represented by a large triangle.

Elliott’s (2000, 2012) analysis of systematically recovered wood charcoal remains from trash middens at La Quemada, and non-systematically collected wood from trash deposits used as construction fill at Los Pilarillos (a second-tier center), demonstrates that the presence and composition of forested patches in the valley has changed significantly since the Classic period. Wood charcoal is generally an abundant ecofact recovered from archaeological sites in semi-arid regions. It provides information related to the vegetative cover and management strategies of woodland zones (Chabal et al. 1999; Asouti and Austin 2005), and can be analyzed much more economically and quickly than other botanical data (e.g. pollen, which also tends to suffer from poor preservation in alluvial settings, such as the Malpaso Valley). The charcoal recovered from the stratified middens at La Quemada indicates that the inhabitants of the valley had access to fuel wood from pine-oak forests, a woody riparian zone, and they also relied on agricultural waste and shrubs associated with agricultural disturbance (e.g. acacia., mesquite, maize stalks) (Figure 6.6). These procurement zones were apparently well-managed by the Mesoamerican occupants of the valley, as the proportions of the taxa

exploited remain stable throughout the occupation. Faunal analysis at sites in the valley, as well as independently derived paleoecological data (discussed below), seem to support this conclusion (Dvorak 2000; Elliott 2007; Elliott et al. 2010). Preliminary study of wood charcoal recovered systematically from similar stratified trash deposits at other contemporaneous northern frontier sites that share the same chronological trajectory of occupation suggests differences in the management and availability of wood fuel resources within the region (Elliott 2012). Charcoal recovered from trash deposits used as construction fill at the Gotas ceremonial complex of the site El Cóporo—located in the Ocampo Valley of Guanajuato, approximately 175 km southeast of La Quemada—indicates the exploitation of pine-oak forest, riparian woods, and open grassland/disturbed zones that may have included cultivated fields. The overall pattern indicates stable patterns of wood use in the early and middle periods. In the late period, pine, oak, and riparian taxa decrease, and an increase in Fabaceae (e.g. acacia, mimosa, mesquite) is detected, suggesting decreased availability of wood from forested zones and an increase in disturbance vegetation. At Los Nogales, a monumental ceremonial complex located on the Cerro Barajas massif of southwest Guanajuato (in the Bajío region that forms the southern limit of the frontier zone), wood charcoal recovered from trash deposits used as fill in the central plaza shows more marked evidence of difficulty in accessing wood resources toward the end of the occupation. Wood is present from a pine-oak zone, as are shrubs indicative of disturbed zones. A high percentage of monocotyledon remains were also recovered (27 percent of the total assemblage) that appear to be agave. A sudden decrease in pine and oak in the late phase is correlated with a strong increase of Fabaceae (acacia, mesquite, mimosa, etc.), which suggests overexploitation of the local forest and subsequent colonization by woody taxa that thrive in disturbed environments. This pattern coincides with the period of peak population growth and construction at Cerro Barajas.

Figure 6.6

Synthesis of the results from wood charcoal studies carried out at La Quemada, El Cóporo, and Cerro Barajas. The sites are represented from northernmost (top) to southernmost (bottom).

Overall, the charred wood data demonstrate a direct relationship between the stability of wood resource exploitation strategies and the degree of aridity, which increases as one moves farther north. The strongest evidence for wood resource stress comes from Cerro Barajas, the site farthest south and with the highest annual rainfall. La Quemada, at the extreme north of the frontier, shows long-term stability, despite its significantly lower annual precipitation. El Cóporo appears to follow a more intermediary pattern: it has some evidence of wood resource stress, but this is much less marked than that for Cerro Barajas.

Multi-proxy sedimentary study in the Malpaso Valley The final research I discuss involves off-site sampling of alluvial sedimentary deposits in the Malpaso Valley carried out to reconstruct landscape evolution through time, and to determine to what degrees both anthropogenic and climatic factors influenced this change (Elliott 2007; Elliott et al. 2010). This study complements the paleoethnobotanical studies carried out at the archaeological sites by providing broader temporal and spatial scales, and examining geoarchaeological and paleobotanical sequences that are less directly influenced by human cultural practices. The study area was an 8-km² portion of the modern floodplain located 3 km south of La Quemada that comprises one of the most productive patches of farmland within the valley, making it a locus of ancient, colonial, and modern settlement. Exposures were studied in 18 trenches excavated by backhoe in three transects and one alluvial cut located 3 km to the north (Figure 6.7). Samples were subjected to phytolith, organic carbon, magnetic susceptibility analyses, and 12 accelerator mass spectrometry (AMS) radiocarbon determinations. Phytoliths were identified and assigned to the categories of open habitat, arboreal and forest patches, and wetland zones. The presence of domesticated taxa was recorded as well. Grass phytolith short cells were used to calculate overall indices of relative aridity and temperature in the valley through time. The data allow us to reconstruct the past 4,000 years of landscape evolution in the valley (Figure 6.8). The results are not consistent with climatic change leading to degradation in the valley or to the Late Classic abandonment. The only fluctuation detected in the sequence is a slight increase in aridity that occurs around 500 BC, and which correlates with the first evidence of maize and land clearance in the valley (approximately 1,000 years before the founding of La Quemada). These data indicate that farming has a much longer history in the valley than previously considered (predating the “Mesoamerican” occupation) and that, contrary to Armillas’s (1964) model, it is associated with drier rather than more humid conditions. A long period of climatic stability follows, characterized by semi-arid conditions similar to those of today. Thus, La Quemada and the other sites in the valley developed during a period of relatively low rainfall.

Figure 6.7

Study area of project for Elliott et al.’s (2010) paleoenvironmental reconstruction with individual trenches shown in each transect (I, II, and III) (map by Christopher Fisher).

An independent study of pollen, phytoliths, and diatoms from agricultural terraces at La Quemada carried out by Trombold and Israde (2005) also indicates stable conditions during the Classic period that resemble the modern warm, semi-arid climate. The discovery of two buried prehispanic irrigation canals indicates one of the strategies likely enacted to ameliorate these conditions (Elliott et al. 2010). The most significant change in the landscape is associated with Spanish colonization in the sixteenth century, when livestock and industrial mining operations were introduced to the region. Massive deforestation and erosion are detected, although climate conditions remain unchanged.

Figure 6.8

Synthesis of the results of Elliott et al.’s (2010) off-site paleoenvironmental reconstruction for the Malpaso Valley.

Discussion The results of the research carried out in the Malpaso Valley indicate that settlement ecology approaches in archaeology have great potential for improving our understanding of human– landscape interactions and the process of “Mesoamericanization” in the northern frontier. The data obtained for the Malpaso Valley allow us to address in particular two of the main hypotheses traditionally advanced to explain settlement dynamics in the region: 1) warfare among communities in competition for subsistence resources, and 2) cyclical climate change. The combination of intra- and inter-site spatial analysis in the Malpaso Valley provides a great deal of data for evaluating the prevalence of violent conflict and for characterizing the types of warfare that are more likely to have been present. The monumental scale of La Quemada compared with other settlements in the valley, as well as its position above the floodplain, has logically raised questions regarding a need for a fortified sanctuary for the valley’s population. Nevertheless, the study presented above indicates that a defensive purpose cannot be unambiguously demonstrated. Although access to certain residential areas of the site is protected by walls and imposing staircases, other areas are clearly intended for public access by large groups and are even served by the complex network of constructed footpaths that links sites of all sizes within the valley. In addition, analysis of the unusually elevated quantity of human remains at La Quemada indicates varied and complex mortuary treatments, some of which seem to have involved careful curation of the remains of probable ancestors

within a lineage, rather than casualties of warfare or sacrifice (Nelson and Martin 2015). Although the size and spatial distribution of the sites in the valley do not seem to strongly support the hypothesis of La Quemada as a shelter from outside invaders, the possibility of ritualized, planned warfare between permanent settlements (a common practice in Postclassic Mesoamerica according to ethnohistoric texts) remains plausible. Such mutually planned interactions could explain the prevalence of human remains and would not be incompatible with the visually imposing nature of La Quemada. The most striking finding of the spatial analysis is the repeated correlation between site location and adequate sources of water for agriculture, whether through the construction of canals or the channeling of natural rainfall runoff. This relationship seems to indicate an acknowledgement of difficult growing conditions in the semi-arid environment, and the development of risk management strategies that served to attenuate variability in climatic conditions from one year to another. Furthermore, the data show that the concern for access to adequately watered fields overrode the desire to place sites in defendable positions. Thus, while competition, conflict, and violence may have been important concerns in the Malpaso Valley, managing the environment to assure consistency in agricultural production was clearly central. The development of strategies to manage local resources and environmental risk in the Malpaso Valley is apparent in other lines of data. The anthracological study of wood fuel use at La Quemada and Los Pilarillos clearly shows that wood resources were more varied and abundant in the Classic period valley than they are today. A number of ecologically distinct zones were exploited (riparian woodland, prairie/grassland, and upland forests) and the use of these various taxa remained stable over the Classic period. This pattern of steady exploitation contrasts with that of contemporary frontier centers farther south. Wood resource stress is apparent toward the end of the occupations of El Cóporo and Cerro Barajas. For the latter site, the change to arid and disturbance-adapted woody taxa is particularly marked. A possible explanation for this pattern is stronger demographic pressure on resources at the more southerly sites toward the end of the Classic period. Cerro Barajas is comparable in scale to La Quemada and El Cóporo, but the density of large centers overall is much higher in the Bajío region than it is for the northern reaches of the frontier zone. Cerro Barajas coexisted with a number of neighboring monumental centers located within only a few kilometers. In contrast, in the semi-arid north, large centers such as El Cóporo and La Quemada were isolated, resulting in lower settlement densities, and thus lower resource demands. Nevertheless, a difference is apparent even between El Cóporo and La Quemada, suggesting that the stable use of wood resources from a variety of habitats at La Quemada may also be due in part to differences in management strategies. Finally, the data from the multi-proxy, off-site paleoenvironmental reconstruction allow us to address the major hypothesis of climate and landscape change in the northern frontier. It becomes clear that climate change is not the likely driver of settlement change, at least in the more arid north. The proposed pattern of increased rainfall at the beginning of the Classic period, followed by significant drought that correlates with the region’s collapse, is not supported for the Malpaso Valley. Furthermore, evidence for maize farming is much more ancient in the valley than previously theorized: it appears in the Formative period, at least 500

years before the occupation of La Quemada and its associated villages. The pattern indicated by the data is, in fact, the reverse of what has been previously proposed. An increase in aridity and temperature to modern levels seems to facilitate the debut of agriculture in the region. A complex, agriculturally based polity made up of permanent settlements develops later under these semi-arid conditions. Finally, its collapse occurs after a span of approximately 400 years, but this abandonment is not associated with changes in precipitation or anthropogenic landscape degradation. The off-site data reinforce the trend toward stability seen in the on-site charcoal studies. What remains to be determined is whether this stability is still visible at a finer temporal scale. It is possible that while the overall longterm trend is stable, there may be finer-scale fluctuations (e.g. cycles of drought lasting 2–4 years) that are not visible in the existing records, but would be problematic for subsistence farmers (such as the variations that occur periodically in the modern valley). These examples of settlement ecology approaches now permit us to revisit the initial question: what factors led to the formation and collapse of northern frontier settlements, such as La Quemada? Although the picture is still not completely clear, we can propose greatly refined models for our example in the northern extreme of the zone. Climate change as a simple prime mover clearly does not explain the development of the La Quemada polity. The Malpaso Valley data do not indicate any significant increase in precipitation correlated with the founding of the Malpaso Valley sites. The only major change in climate is the increase in aridity and temperature during the Formative period, which is associated with more stable streamflow and a decrease in wetlands in the valley. These climatic conditions, which resemble those of the modern day, may have made plant cultivation more feasible, and indeed, the earliest evidence for agriculture is detected at this time (the period traditionally associated with a hunter-gatherer occupation). In light of these results, it seems critical to reconsider the social factors that may have influenced frontier settlement dynamics. The earlier than anticipated presence of maize agriculture in the Malpaso Valley raises questions about the proposed adversarial relations between the indigenous mobile populations and the agricultural societies of the Classic period. While it seems unlikely that the appearance of monumental centers with Mesoamerican iconography and ceramic types can be explained without some population movement out of the Mesoamerican heartland, it now seems premature to assume that a simple model of colonization explains the settlement dynamics. The Middle Classic period is associated with the collapse of the Teotihuacán state in Central Mexico (Cowgill 1992). The breakdown of this far-reaching political power and its influential economic networks provoked significant population displacement, and it is likely that the groups caught up in this diaspora would have moved toward the north, as they sought new territories and opportunities (Diehl and Berlo 1989). It also is plausible to suggest that the mobile or semi-mobile groups already living in the frontier (and practicing small-scale agriculture?) and the newly arriving Mesoamerican groups would find social and economic advantages to cooperating and perhaps even cofounding permanent settlements together that incorporated elements of both cultures. Regarding the abandonment of sites such as La Quemada, more data are clearly needed from across the frontier zone, but even now we can propose new models to explain this phenomenon. Drought does not appear a likely factor for settlement abandonment based on the

data from the Malpaso Valley. Neither does anthropogenic degradation of the landscape. Indeed, the occupants of the valley appear to have been excellent stewards who developed strategies to flourish in a challenging desert/steppe environment. Again, it seems that we must reconsider social and cultural factors to explain the retreat of these populations toward the traditional Mesoamerican heartland. The similarities between Classic period northern frontier settlements and Early Postclassic (AD 1000–1200) sites such as Tula in Central Mexico indicate clear cultural continuity (Diehl 1983; Mastache de Escobar et al. 2002). The formation of the Toltec State at Tula that dominated the Early Postclassic period may have thus exercised a cultural pull toward Mesoamerican populations living farther north, who then moved south to participate in this new, powerful polity. An additional possibility, hinted at by the wood charcoal data presented above, is that environmental crises may have potentially played a partial, uneven role across the frontier. The evidence for the greatest wood resource overexploitation occurs in the southern Bajío zone (rather than in the seemingly more fragile northern territories) and raises the possibility that cultural collapse originated on the southern edge of the frontier due to anthropogenic degradation or localized climate change, and then spread to sites farther north. The abandonment of large settlements like Cerro Barajas, which sit along major corridors for the movement of people, goods, and ideas between Central and West Mexico and regions farther north, would likely have severed key lines of communication between the Mesoamerican core and frontier communities such as El Cóporo and La Quemada. This process would imply a cultural implosion that began in the south and whose socio-political and economic shockwaves would have traveled quickly northward. To confirm or discard this hypothesis will require a great deal more paleoenvironmental study of northern frontier sites along the north–south trajectory, but the potential implications suggest a more nuanced and sophisticated understanding of socio-environmental interactions in the greater region.

Future directions More systematic application of these approaches to other settlement systems throughout the norther frontier region is crucial to improve our understanding of the occupation history and the relations between human populations and their environments. Spatial analysis of settlement patterns in the Malpaso Valley demonstrates that it is possible to evaluate and distinguish between different models of settlement strategies, in this case preference for defensible settlement locations or preference for maximum agricultural production. Other parameters could also be evaluated, such as ease of circulation among sites of different sizes, visibility studies to investigate possible interrelations among specific sites, and the potential of different sectors for a range of arid-adapted agricultural strategies. The cultivation of drought resistant plants (such as cacti and agave) to offset periodic crop failure is often proposed as an important adaptive strategy for the semi-arid northern frontier (e.g. Parsons 2010; Sauer 1963). GIS analysis of landscape variables (such as soil type, aspect, slope, degree of water runoff and infiltration) may aid us to better understand the minimum and maximum extents of possible zones of drought resistant crop cultivation to produce more precise production estimates of this potentially important food resource. Irrigation systems were also likely to have been very

important at frontier settlements and further study of the hydrological potential of frontier landscapes could help to identify and study such systems. More systematic application of paleoethnobotanical analysis at archaeological sites and offsite paleobotanical studies is clearly warranted. Both approaches highlight the vegetation communities that were present and give clues regarding their management by local communities. They also provide distinct yet complementary data that make each essential. While it is true that paleoethnobotanical assemblages recovered from sites are influenced by cultural selection practices, and thus do not directly reflect the “natural” vegetation, they also fill in gaps in the more traditional off-site paleobotanical records (for example, while wood from the Fabaceae family is an important component of on-site charcoal assemblages in the region, Fabaceae pollen grains are unlikely to appear in off-site sediment sequences, as they are produced in small numbers and are evolved to be transported by insects rather than air currents). Furthermore, the study of plant remains from dated archaeological contexts permits archaeologists to examine these trends over time, at temporal scales that correspond to the rhythm of human occupation (which are not always available in the stratigraphy of off-site deposits). Nevertheless, the Malpaso Valley work has also demonstrated the importance of the reconstruction of landscape evolution over the long term using off-site sedimentary sequences. These paleoecological data permit us to understand local environmental and climatic conditions prior to, during, and following a settlement system’s occupation. This bracketing is essential for understanding the conditions that led up to the founding of a site and that may have influenced its abandonment. The low number of regional vegetation reconstructions carried out in the frontier may be due in part to the fact that the alluvial sedimentary records that characterize most of the region provide a poor environment for the preservation of most organic bio-indicators used in environmental archaeology. In particular, pollen, the traditional indicator of past vegetation, is usually too poorly preserved to provide statistically significant results. However, the Malpaso Valley work has demonstrated the utility of phytoliths for reconstructing climate patterns, distinct vegetation communities, and the presence of domesticated crops (Elliott 2007; Elliott et al. 2010). As these microscopic silica particles from plant tissues are highly resistant to the environmental factors that tend to degrade pollen assemblages, they are excellent complements to pollen data (or can even be substitutes for such data, when necessary). Further work must be undertaken to assess the preservation of phytoliths in sediments around other frontier settlement systems and to develop the necessary reference collections to aid in their identification. In particular, it would be interesting to obtain such data around the sites of El Cóporo and Cerro Barajas, to compare with the existing charcoal data, and to see if they support or contradict the pattern of landscape degradation over time. Finally, a question remains regarding whether the stability observed in the off-site sediment sequences and the archaeological charcoal records in the Malpaso Valley truly reflects a lack of climatic change, or whether short-term variability in precipitation may have occurred but is not visible in the temporal resolution of these data. The study of tree rings is a method that allows the reconstruction of climate records with an annual resolution and equally precise dating of archaeological contexts. A dendrochronological approach thus has the potential to

significantly refine our understanding of climate and chronology in the northern frontier zone. Despite the commonly held belief that dendrochronology cannot be applied in Mesoamerica, recent research shows there is, in fact, great potential for such studies. Stahle et al. (2011) have published a 1,200-year tree-ring chronology based on Taxodium mucronatum (Bald cypress, Ahuehuete, Sabino) for Central Mexico. In addition, a pilot study of pines near the Malpaso Valley has demonstrated that these trees are appropriate for dating and climate reconstruction, and that they correspond to the species used in Prehispanic construction at La Quemada (Turkon et al. 2011). Sturt Manning and colleagues at Cornell University have more recently successfully applied “wiggle-match” dating methods to archaeological wood samples from La Quemada; such methods permit floating chronologies to be converted to near-absolute chronologies with an error range that is narrower than that of AMS dating (Turkon et al. 2015). Further development and application of these methods in northern frontier settlements has great potential for understanding fine-scale climate change and human response.

Conclusions The goal of this chapter has been to demonstrate the potential of multi-scalar, interdisciplinary approaches for understanding the dynamics of human–landscape interaction and shifting settlement ecologies in the northern frontier zone of Mesoamerica, as well as to better characterize the phenomenon of settlement expansion in the far north during the Classic period. The past and ongoing work in the Malpaso Valley shows that these approaches, although often relative and simple in their design, have great analytic power to test a number of long-standing hypotheses that have greatly influenced the models and assumptions of several generations of Mesoamerican archaeologists concerning the social and ecological trajectories of the northern extent of the culture area. The next logical step will be to expand this settlement ecology perspective to other settlement systems in the greater northern frontier zone, by carrying out integrated multidisciplinary collaborations involving a variety of specialists to untangle the complex relations that these Classic societies developed with their landscapes and with each other. Although an ambitious and long-term undertaking, this multi-scalar examination of the complexly layered settlement histories will no doubt better inform our understanding of the overall trajectory of occupation in the northern frontier zone, and will also permit us to understand the range of the strategies developed by its inhabitants and the diversity in their outcomes. It seems then that settlement ecology approaches represent a promising future for the archaeology of northern Mesoamerica—one that will allow archaeologists to better integrate and contextualize the northern frontier zone in relation to the social and ecological dynamics of the larger Mesoamerican culture area.

References Armillas, Pedro 1951 Mesoamerican Fortifications. American Antiquity 25: 77–86. Armillas, Pedro 1964 Condiciones Ambientales y Movimientos de Pueblos en la Frontera Septentrional de Mesoamérica. In Homenaje a Fernando Márquez-Miranda, pp. 62–82. Publicaciones del Seminario de Estudios Americanistas y Seminario de Antropología Americana, Universidades de Madrid y Sevilla, Madrid, Spain.

Armillas, Pedro 1969 The Arid Frontier of Mexican Civilization. Transactions of the New York Academy of Sciences, Series II 31(6): 697–704. Asouti, E., and P. Austin 2005 Reconstructing Woodland Vegetation and Its Exploitation by Past Societies, Based on the Analysis and Interpretation of Archaeological Wood Charcoal Macroremains. Environmental Archaeology 10: 1–18. Bakewell, Peter John 1971 Silver Mining and Society in Colonial Mexico: Zacatecas, 1546–1700. Cambridge University Press, Cambridge. Batres, Leopoldo 1903 Visita a los Monumentos Arqueológicos de “La Quemada” Zacatecas. Imprenta de laViuda. de Francisco Diaz de Leon, Mexico City. Berghes, C. 1996 Descripción de las Ruinas de Asentamientos Aztecas Durante su Migración al Valle de México, a Través del Actual Estado Libre de Zacatecas. Compilada Tras Investigaciones y Levantimientos en Este Lugar y Esclarecida por el Manuscrito en Jeroglíficos Aztecas del Museo de México. Estudio int. y notas de Achim Lelgemann, col. Joyas Bibliográficas Zacatecanas II, Gob. del Estado de Zacatecas, Univ. Autónoma de Zacatecas y Centro Bancario del Estado de Zacatecas, Zacatecas, Mexico Bernal, Juan Pablo, Matthew Lachniet, Malcolm McCulloch, Graham Mortimer, Pedro Morales, and Edith Cienfuegos 2011 A Speleothem Record of Holocene Climate Variability from Southwestern Mexico. Quaternary Research 75(1): 104–113. Braniff, Beatriz 1989 Oscilación de la Frontera Norte Mesoamericana: Un Nuevo Ensayo. Arqueología 1: 99–114. Braniff, BeatrÍz 2000 A Summary of the Archaeology of North-Central Mesoamerica: Guanajuato, Querétaro, and San Luis Potosí. In Greater Mesoamerica: The Archaeology of West and Northwest Mexico, edited by Michael S. Foster and Shirley Gorenstein, pp. 35–42. University of Utah, Salt Lake City. Braniff, Beatriz, and Marie-Areti Hers 1998 Herencias Chichimecas. Arqueología 19: 55–80. Brown, Roy B. 1984 The Paleoecology of the Northern Frontier of Mesoamerica. Ph.D. dissertation. Department of Anthropology, University of Arizona, Tucson. Brown, R.B. 1992 Arqueología y Paleoecología del Norcentro de México. Insituto Nacional de Antropología e Historia, Mexico City. Butzer, Karl W., James T. Abbott, Charles D. Frederick, Paul H. Lehman, Carlos E. Cordova, and John F. Oswald 2008 SoilGeomorphology and “Wet” Cycles in the Holocene Record of North-Central Mexico. Geomorphology 101: 237–277. Caballero, Margarita, Beatriz Ortega, Francisco Valadez, Sarah Metcalfe, Jose Luis Macias, and Yoko Sugiura 2002 Sta. Cruz Atizapan: A 22-ka Lake Level Record and Climatic Implications for the Late Holocene Human Occupation in the Upper Lerma Basin, Central Mexico. Palaeogeography, Palaeoclimatology, Palaeoecology 186(3/4): 217–235. Cabrero Garcia, Ma. Teresa 1991 Cultura Arqueológica de Bolaños (Zacatecas y Jalisco): Una Frontera Cultural. Ancient Mesoamerica 2: 193–204. Cárdenas García, Efraín 1999 El Bajío en el Clásico. El Colegio de Michoacán, Zamora, Michoacán. Chabal, L., L. Fabre, J.-F. Terral, and I. Thery-Parisot 1999 L’anthracologie. In La Botanique, Collection “Archéologiques”, pp. 43–104. Editions Errance, Paris. Coe, Michael D. 1994 Mexico. Thames and Hudson, London. Cowgill, George L. 1992 Toward a Political History of Teotihuacan. In Ideology and Precolumbian Civilizations, edited by Arthur Demarest and G. Conrad, pp. 87–114. School of American Research Press, Santa Fe. Davies, S.J., S.E. Metcalfe, A.B. MacKenzie, A.J. Newton, G.H. Endfield, and J.G. Farmer 2004 Environmental Changes in the Zirahuen Basin, Michoacan, Mexico, During the Last 1,000 Years. Journal of Paleolimnology 31(1): 77–98. Diehl, Richard A. 1983 Tula: The Toltec Capital of Ancient Mexico. Thames and Hudson, London. Diehl, Richard A., and Janet C. Berlo (editors) 1989 Mesoamerica after the Decline of Teotihuacan: AD 700–900. Dumbarton Oaks Research Library and Collection, Washington, DC. Dvorak, Sara Anne 2000 Faunal Consumption at La Quemada, Zacatecas, Mexico. Unpublished MA thesis. Department of Anthropology, Arizona State University, Tempe. Elliott, Michelle 2000 Prehispanic Wood Procurement in the Malpaso Valley, Zacatecas, Mexico, AD 500–900. unpublished MA paper. Department of Anthropology, Arizona State University, Tempe. Elliott, Michelle 2005 Evaluating Evidence for Warfare and Environmental Stress in Settlement Pattern Data from the Malpaso Valley, Zacatecas, Mexico. Journal of Anthropological Archaeology 24: 297–315. Elliott, Michelle 2007 Human Occupation and Landscape Change in the Malpaso Valley, Zacatecas, Mexico. Dissertation. Arizona State University, School of Human Evolution and Social Change, Tempe, Arizona. Elliott, Michelle 2012 An Anthracological Approach to Understanding Late Classic Period Cultural Collapse in Mesoamerica’s Northwestern Frontier. In Wood and Charcoal: Evidence for Human and Natural History, edited by Ernestina Badal, Yolanda Carrión, Miguel Macías, and María Ntinou, pp. 217–226. SAGVTVM Extra-13. Departament de Prehistòria i d’Arqueologia, Universitat de Valencia, Valencia. Elliott, Michelle, Christopher T. Fisher, Ben A. Nelson, Roberto S. Molina Garza, Shawn K. Collins, and Deborah M. Pearsall 2010 Climate, Agriculture, and Cycles of Human Occupation over the Last 4,000 Years in Southern Zacatecas, Mexico. Quaternary Research 74(1): 26–35. Fisher, Christopher T., Helen P. Pollard, Isabel Israde-Alcantara, Victor H. Garduno-Monroy, and Subir K. Banerjee 2003 A

Reexamination of Human-Induced Environmental Change within the Lake Patzcuaro Basin, Michoacan, Mexico. Proceedings of the National Academy of Sciences 100(8): 4957–4962. Frederick, C.D. 1995 Fluvial Response to Late Quaternary Climate Change and Land Use in Central Mexico. Ph.D. dissertation. University of Texas, Austin. Haas, Jonathan, and Winifred Creamer 1996 The Role of Warfare in the Pueblo III Period. In The Prehistoric Pueblo World, AD 1150–1350, edited by Michael A. Adler, pp. 205–213. University of Arizona Press, Tucson. Haug, Gerald H., Detlef Gunther, Larry C. Peterson, Daniel M. Sigman, Konrad A. Hughen, and Beat Aeschlimann 2003 Climate and the Collapse of Maya Civilization. Science 299(5613): 1731–1735. Hers, Marie-Areti 1989 Los Toltecas en Tierras Chichimecas. Universidad Nacional Autónoma de México Instituto de Investigaciones Estéticas, Mexico City. Hodell, D.A., J.H. Curtis, and M. Brenner 1995 Possible Role of Climate in the Collapse of Classic Maya Civilization. Nature 375: 391–394. Hodell, David A., Mark Brenner, and Jason H. Curtis 2000 Climate Change in the Northern American Tropics and Subtropics since the Last Ice Age. In Imperfect Balance: Landscape Transformations in the Precolumbian Americas, edited by David L. Lentz, pp. 13–38. Columbia University Press, New York. Hodell, David A., Mark Brenner, Jason H. Curtis, and Thomas Guilderson 2001 Solar Forcing of Drought Frequency in the Maya Lowlands. Science 292(5520): 1367–1370. Jiménez Betts, P. 1992 Una Red de Interacción del Noroeste de Mesoamérica: Una Interpretación. In Origen y Desarollo de la Civilización en el Occidente de México: Homenaje a Pedro Armillas y Angel Palerm, edited by B.B. de Lameiras and P.C. Weigand, pp. 177–204. El Colegio de Michoacán, Zamora. Jiménez Betts, P., and J.A. Darling 2000 The Archaeology of Southern Zacatecas: The Malpaso, Juchipila, and ValparaisoBolaños Valleys. In Greater Mesoamerica: The Archaeology of West and Northwest Mexico, edited by Michael S. Foster and Shirley Gorenstein, pp. 155–180. University of Utah Press, Salt Lake City. Kantor, L. 1995 Lithic Specialization and Exchange at La Quemada, Zacatecas, Mexico. Unpublished MA thesis. Department of Anthropology, State University of New York at Buffalo. Kelley, J.C. 1956 Settlement Patterns in North-Central Mexico. In Prehistoric Settlement Patterns in the New World, edited by G.R. Willey, pp. 128–139. Viking Fund Publications in Archaeology, vol. 23. Wenner-Gren Foundation for Anthropological Research, Inc., New York. Kelley, J.C. 1974 Speculations on the Culture History of Northwest Mesoamerica. In Archaeology of West Mexico, edited by Betty Bell, pp. 19–39. Sociedad de Estudios Avanzados del Occidente de Mexico, A.C., Ajijic. Kelley, J.C. 1985 The Chronology of the Chalchihuites Culture. In The Archaeology of West and Northwest Mesoamerica, edited by M.S. Foster and P.C. Weigand, pp. 269–288. Westview Press, Boulder and London. Kelley, J. Charles 1989 The Retarded Formative of the Northwest Frontier of Mesoamerica. In El Préclasico o Formativo: Avances y Perspectivas: Seminario de Arqueología “Dr. Román Piña Chan”, edited by M. Carmona Garcias, pp. 405– 423. INAH, Mexico City. Kelley, J.C. 2000 The Aztatlán Mercantile System: Mobile Traders and the Northwestward Expansion of Mesoamerican Civilization. In Greater Mesoamerica: The Archaeology of West and Northwest Mexico, edited by Michael S. Foster and Shirley Gorenstein, pp. 137–154. University of Utah Press, Salt Lake City. Labat, Jean-Noël 1995 Végétation du Nord-Ouest du Michoacán Mexique. Instituto de Ecología, Xalapa, Veracruz. LeBlanc, Steven A. 2007 Prehistoric Warfare in the American Southwest. University of Utah Press, Salt Lake City. Lelgemann, Achim 2000 Proyecto Ciudadela de La Quemada, Zacatecas. Report on file with the Instituto Nacional de Antropología e Historia, Mexico City. Mastache de Escobar, Alba Guadalupe, Robert H. Cobean, and Dan M. Healan 2002 Ancient Tollan: Tula and the Toltec Heartland. University Press of Colorado, Boulder. Matson, John O., and Rollin H. Baker 1986 Mammals of Zacatecas. Texas Tech University Press, Lubbock. Medina González, José Humberto 2000 El Paisaje Ritual del Valle de Malpaso. Tésis de Licenciatura en Arqueología, Escuela Nacional de Antropología, Mexico City. Metcalfe, Sarah E. 1997 Palaeolimnological Records of Climate Change in México: Frustrating Past, Promising Future? Quaternary International 43/44: 111–116. Metcalfe, Sarah E., and Sarah O’Hara 1997 The Climate of Mexico since the Aztec Period. Quaternary International 43/44: 25–31. Metcalfe, S.E., F.A. Street-Perrott, R.B. Brown, P.E. Hales, R.A. Perrott, and F. Steininger 1989 Late Holocene Human Impact on Lake Basins in Central Mexico. Geoarchaeology 4: 119–141. Millhauser, John K. 1999 Ritual, Social, and Economic Dimensions of Obsidian Use in the Malpaso Valley, Zacatecas, Mexico, AD 500–900. Unpublished MA thesis. Department of Anthropology, Arizona State University. Nelson, B.A. 1990 Observaciones Acerca de la Presencia Tolteca en La Quemada, Zacatecas. In In Mesoamérica y Norte de México Siglos IX–XI, edited by Sodi Miranda, pp. 521–540. Instituto Nacional de Antropología e Historia, Mexico City. Nelson, B.A. 1992 El Maguey y el Nopal en la Economía de Subsistencia de La Quemada, Zacatecas. In Origen y Desarollo

de la Civilización en el Occidente de México, edited by B.B. de Lameiras and P.C. Weigand, pp. 359–382. El Colegio de Michoacán, Zamora. Nelson, Ben A. 1993 Outposts of Mesoamerican Empire and Architectural Patterning at La Quemada, Zacatecas. In Culture and Contact: Charles C. DiPeso’s Gran Chichimeca, edited by Ann Wooseley and John Ravesloot, pp. 173–190. University of New Mexico Press, Albuquerque. Nelson, Ben A. 1995 Complexity, Hierarchy, and Scale: A Controlled Comparison between Chaco Canyon, New Mexico, and La Quemada, Zacatecas. American Antiquity 60(4): 597–618. Nelson, Ben A. 1997 Chronology and Stratigraphy at La Quemada, Zacatecas, Mexico. Journal of Field Archaeology 24: 85– 109. Nelson, Ben A. 2000 Aggregation, Warfare, and the Spread of the Mesoamerican Tradition. In The Archaeology of Regional Interaction: Religion, Warfare, and Exchange across the American Southwest and Beyond, edited by Michelle Hegmon, pp. 317–337. University of Colorado Press, Boulder. Nelson, Ben A. 2003 A Place of Continued Importance: The Abandonment of Epiclassic La Quemada. In The Archaeology of Site Abandonment in Middle America, edited by Takeshi Inomata and Ronald W. Webb, pp. 77–89. Westview Press, Boulder. Nelson, Ben A., and Debra L. Martin 2015 Symbolic Bones and Interethnic Violence in a Frontier Zone, Northwest Mexico, ca. 500–900 CE. Proceedings of the National Academy of Sciences 112(30): 9196–9201. Nelson, Ben A., J. Andrew Darling, and David A. Kice 1992 Mortuary Patterns and the Social Order at La Quemada, Zacatecas. Latin American Antiquity 3(4): 298–325. O’Hara, S.L., F.A. Street-Perrott, and T.P. Burt 1993 Accelerated Soil Erosion around a Mexican Highland Lake Caused by Prehispanic Agriculture. Nature 362: 48–51. Parsons, Jeffrey 2010 The Pastoral Niche in Prehispanic Mesoamerica. In Pre-Columbian Foodways: Interdisciplinary Approaches to Food, Culture, and Markets in Ancient Mesoamerica, edited by J. Staller and M. Carrasco, pp. 109–137. Springer Press, New York. Rzedowski, Jerzy 1981 Vegetación de México. Editorial Limusa, Mexico City. Sauer, Carl Ortwin 1963 Land and Life: Selections from the Writing of C.O. Sauer. University of California Press, Berkeley. Secretaría de Programación y Presupuesto 1980 Síntesis Geográfica de Guanajuato. Coordinación General de los Servicios Nacionales de Estadística, Geografía e Informática, México, Mexico City. Secretaría de Programación y Presupuesto 1981 Síntesis Geográfica de Zacatecas. Coordinación General de los Servicios Nacionales de Estadística, Geografía e Informática, Mexico City. Stahle, D.W., J. Villanueva Diaz, D.J. Burnette, J. Cerano Paredes, R.R. Heim Jr., F. K. Fye, R. Acuna Soto, M.D. Therrell, M.K. Cleaveland, and D.K. Stahle 2011 Major Mesoamerican Droughts of the Past Millennium. Geophysical Research Letters 38: L05703. Stone, Glenn Davis 1996 Settlement Ecology: The Social and Spatial Organization of Kofyar Agriculture. University of Arizona Press, Tucson. Trombold, C.D. 1978 The Role of Locational Analysis in the Development of Archaeological Research Strategy. Unpublished Ph.D. dissertation. Department of Anthropology, Southern Illinois University. Trombold, Charles D. 1990 A Reconsideration of Chronology for the La Quemada Portion of the Northern Mesoamerican Frontier. American Antiquity 55(2): 308–323. Trombold, Charles D. 1991 Causeways in the Context of Strategic Planning in the La Quemada Region, Zacatecas, Mexico. In Ancient Road Networks and Settlement Hierarchies in the New World, edited by Charles D. Trombold, pp. 145–168. Cambridge University Press, Cambridge. Trombold, Charles D. 2005 A Population Estimate for the Epiclassic Middle Malpaso Valley (La Quemada), Zacatecas, Mexico. Latin American Antiquity 16(3): 235–253. Trombold, Charles D., and Isabel Israde-Alcantara 2005 Paleoenvironment and Plant Cultivation on Terraces at La Quemada, Zacatecas, Mexico: The Pollen, Phytolith and Diatom Evidence. Journal of Archaeological Science 32: 341–353. Turkon, Paula 2002 Social Identity and Food in the Prehispanic Malpaso Valley, Zacatecas, Mexico. Unpublished Ph.D. dissertation. Department of Anthropology, Arizona State University, Tempe. Turkon, Paula 2004 Food and Status in the Prehispanic Malpaso Valley, Zacatecas, Mexico. Journal of Anthropological Archaeology 23(2): 225–251. Turkon, Paula, Michelle Elliott, S.W. Manning, and Carol Griggs 2011 Dendrochronology, Chronological Control, and Climate Reconstruction in Northwestern Mesoamerica. Presented at the 76th Annual Meeting of the Society for American Archaeology, Sacramento, California. Turkon, Paula, S. Manning, Carol Griggs, and Alexandra Jigetts-O’Neill 2015 Dendro-14C-Wiggle-Match Contributions to Northwestern Mesoamerican Chronology. Presented at the 80th Annual Meeting of the Society for American Archaeology, San Francisco. Weigand, Phil C. 1978a The Prehistory of the State of Zacatecas: An Interpretation, Part I. Anthropology 1: 67–87. Weigand, Phil C. 1978b The Prehistory of the State of Zacatecas: An Interpretation, Part II. Anthropology 2(2): 103–117.

Weigand, Phil C. 1982 Mining and Mineral Trade in Prehistoric Zacatecas. In Mining and Mining Techniques in Ancient Mesoamerica, edited by Phil C. Weigand and G. Gwynne, pp. 87–134. Anthropology 6. Special issue. State University of New York at Stony Brook. Weintraub, P.D. 1992 The Use of Wild and Domesticated Plants at La Quemada, Zacatecas, Mexico. Unpublished MA project. Department of Anthropology, State University of New York, Buffalo. Wells, E. Christian 2000 Pottery Production and Microcosmic Organization: The Residential Structure of La Quemada, Zacatecas. Latin American Antiquity 11(1): 21–42.

7 Agrarian settlement ecology in the Classic Maya Lowlands A comparative analysis of La Joyanca (Guatemala) and Río Bec (Mexico) Eva Lemonnier

Introduction Investigations focused on ancient Maya societies for more than 60 years have brought a better understanding of their demographic growth during the Late Classic apogee (AD 600–900). This demographic evolution in the lowland Maya environment, considered as unfavorable and hardly usable, opened one of the most interesting debates of the past two decades on human– environment relationships. How did the Maya produce their means of subsistence? What has been the ecological impact of their territorial expansion? Were their strategies sustainable over the long term, considering both resources and socioeconomic needs? To date, these questions have produced a robust set of archaeological and geographical studies on prehispanic Mayan agriculture (Harrison and Tuner 1978; Killion 1992; Fedick 1996; Gómez-Pompa et al. 2003). Agriculture was at the core of the subsistence economy of the ancient Maya societies and reached its height during the Classic period (AD 300–900) as an infield–outfield agricultural system (Netting 1977) based on the maize-bean-gourd trilogy. Classic Maya cities, characterized by dispersed households (Drennan 1988), were associated with an agricultural territory that included both intensively cultivated spaces close to the dwellings and farther, extensively cultivated spaces. Nevertheless, very few studies reveal precise schemes of the field system organization. In fact, this subject is a methodological challenge on the one hand because of the environmental context of sub-tropical forests (leading to visibility and conservation problems), and on the other hand because of the high variability of such systems between sites. Each city presents a different spatial layout, corresponding with the sociopolitical and economic processes influenced by its own local history. Questions about agricultural systems are raised at every new archaeological site because of inherent difficulties like identifying intra-site plots (location and the nature of exploitation), relating them with dwelling units, and reconstructing the spatio-temporal dynamic of the whole agricultural system. During the 1970s research focused on the diversity of the Maya livelihood strategies, and households are still studied in close relation with the environment and agricultural features. With the development of landscape archaeology in the 1990s, a new energy was given to the

“old” settlement archaeology of the 1950s: “landscapes are both settings for ancient human activity and artifacts in their own right, in the sense that they were altered by the ancient people who lived on and utilized them” (Webster et al. 2000: 13). People not only adapted to these environments but they built them through agrarian and social systems (Guilaine 1991). Therefore many studies have examined settlement factors (e.g. proximity of resources, especially water and soils) and the role of the growing demography of agricultural intensification. A few tried to evaluate the primacy of one factor over another. For example, the foundation of Dos Pilas as the Petexbatún regional center by the Tikal dynasty in AD 625 reflects a political and economic decision: the new city, settled away from water resources and arable lands, probably was used to control the greater inland trade route. As such, it was supplied by the two former cities recognized as the regional centers of the agricultural production since the sixth century (Demarest 2006). Most studies have adopted a regional view to understand the land use associated with a series of ranked settlements. Given the correlation between dispersed settlement and labor-intensive agriculture in the Maya Lowlands, even under population pressure (Drennan 1988), there are few investigations on the chronological and spatial evolution of settlement patterns in relation to the possible changes in agrarian systems. There have been even fewer concerning the organization of agricultural labor and related social groups. The settlement ecology concept represented in this volume is very useful here. It invites investigation of the causative factors that can explain the form of settlements and their evolution over time and space, especially in terms of human decisionmaking and social dynamics. These are obviously studied in Maya archeology but often independently of agrarian issues, while farming practices and social organization are combined in ethnographical and ethnohistorical investigations. After a short synthesis of relevant investigations focused on the human–environment relationship in ancient Maya societies, I will explore in the broader context of settlement ecology two site case studies: La Joyanca and Río Bec. These are lowland Maya sites for which my colleagues and I offer two distinct socioeconomic models, and each provides an opportunity to explore the relationship between social structure and agricultural systems and to evaluate individual/collective influences that shaped their architectural and agricultural landscapes. In brief, La Joyanca is best defined as a co-habitation of intermediate social units, which could be considered as “neighborhoods.” These neighborhoods would have been formed over time, driven by local elite, and would have acted as large agricultural-labor units, similar to their ethnohistoric and ethnographic counterparts. In Río Bec, we identified an agricultural system with field systems enclosing the households. This is the first type of field system like this ever documented in the Maya area. This arrangement suggests the existence of autonomous production units—a distinct form of social and labor organization.

Human–environment interaction in Classic agrarian Maya societies In the Maya area (Figure 7.1), some aspects of the human–environment relationship are evident: the existence of an anthropogenic environment, the great ecological diversity, the adaptation and transformation of the natural setting through an array of technical strategies, and the capacity of managing the landscape to exploit natural resources. The Maya landscape is

perceived as a “managed mosaic” (Fedick 1996): that is to say a structured (or constructed) entity composed of several well-differentiated spatial units, which result from (or have been produced by) earlier agricultural and social systems. However, the ecological impact of prehispanic agriculture is still being debated. Several paleo-ecological studies show that environmental exploitation was the cause for the principal anthropogenic changes of terrestrial and lacustrine settings, with deforestation and agriculture causing soil erosion (maya clay— Deevey et al. 1979; Brenner 1983; Binford et al. 1987; Leyden et al. 1996; Pope et al. 1996; Dunning et al. 1997, 1999), which, in some cases, caused eutrophication of nearby lakes as early as the Preclassic–Classic transition. Permanent lakes and swamps transformed into seasonal swamps close to La Milpa, Yaxha, and Nakbe, and also possibly in El Mirador and Tikal (Dunning et al. 2002; Dunning 2003; Dahlin et al. 1980; Jacob 1994, 1995; Cowgill and Hutchinson 1963; Dahlin and Dahlin 1994). Other paleo-ecological studies in the Petexbatún region and at Copán, Nakbe, La Milpa, Yaxha and La Joyanca demonstrate the existence of agroforestal systems characterized by local and regional dynamics, which were influenced by environmental and climatic factors in addition to social and agricultural ones.

Figure 7.1

Map of the Maya region with principal sites of La Joyanca and Río Bec shown.

The Petexbatún region is one of the best documented from the human–environment relationship perspective, with extensive data from paleo-ecological (e.g. pedology,

sedimentology, palynology) and paleo-agronomical results (e.g. paleo-osteology, paleofauna and paleobotany), as well as the survey and excavations of agricultural features. The most relevant environmental changes are associated with Preclassic human occupation, whereas Late Classic levels showed no evidence of soil exhaustion, strong deforestation, or significant climatic change (Demarest 2006). The Classic population would have inherited an eroded landscape, resulting from millennia of swidden agriculture (Dunning et al. 1998b, 1999, 2002; Beach 1996). However, they managed to stabilize the erosion process and environmental alteration by implementing a diverse set of measures, including terraces, channels, water tanks, and agriculture in natural depressions (i.e. rejolladas). Thus, while some studies favor an ecological theory for the Maya collapse, combining overpopulation, deforestation, erosion, and drought (Abrams and Rue 1988; Culbert 1977, 1988; Santley et al. 1986; Freter 1994; Webster and Freter 1990; Webster 2002; Webster et al. 2000; Hodell et al. 1995; Curtis et al. 1996; Gill 2000; Adams et al. 2004; Braswell et al. 2004), others propose another scenario based on the idea of savoir-faire (Ford 2008) and a society’s ability to adapt and react, especially with regard to resource management related to regular and repetitive drought episodes (Aimers and Hodell 2011; Kennett et al. 2012). Geoarchaeological works permit the observation of a great diversity of land management features (e.g. terraces, ridges, raised and/or drained fields, etc.), thus disposing of detailed typological inventories and excavation data. All of the complex models contain agricultural intensification and diversification as factors (Johnston 2003) that would have highly modified the landscape (Dunning 1996; Dunning et al. 1998a, 1999). In the recent “agrarian city” model (Arnauld 2008; Arnauld and Michelet 2004), agricultural territory and settlement cannot be separated. Fields with staple crops would have surrounded houses. The infield–outfield agricultural system, proposed in the 1970s, hypothesized that intensive cultivation occurred within the cities and their immediate peripheries, along with more extensive cultivations (Killion 1992). In the Maya area, this model is being gradually validated by new data from agricultural and residential structures (Fedick 1994; Dunning et al. 1997; Chase and Chase 1998; Webster et al. 2000; Kunen 2001, 2004; Liendo Stuardo 2002; Robin 2002; Houston and Escobedo in press). Nevertheless, a debate still exists concerning the internal organization of cities and agriculture (Lemonnier 2009), and more specifically whether kitchen gardens (as a reference to the “garden city” concept by Tourtellot et al. 1988) were basic alimentation plots or fields inside the city (i.e. the “agrarian city” concept). The infield agriculture argument is still theoretical. It is based on the existence of structures which we suppose were used to improve agriculture and on a small number of chemical analyses of phosphate signatures that confirm soils were modified and fertilized in the Petexbatún or at Sayil (see Dunning et al. 1997 and Killion et al. 1989). A single study identified maize (13C) in terraces at Caracol (Webb et al. 2004). Thus, the concept of infield agriculture still raises questions about the functioning of its implements, its sustainable nature, and the economic proportion of its production, regarding the outfield component. In addition, it has not been demonstrated that agricultural intensification and diversification were related to demographic pressure. Outfield agriculture during the Classic period was probably restricted to marginal settings and may be harder to identify

because it leaves very few archaeological traces. It also remains poorly documented as a result of the indirect evidence identified in sediment cores—such as maize pollen, charcoal, fungus, and signs of deforestation—which does not allow a complete picture of the general agricultural system. We should explore inter-site areas in order to locate empty spaces as possible farming fields using new technologies such as LIDAR surveying. Furthermore, very few studies have focused on social processes that were involved in agriculture (e.g. specialization and cooperation), in particular the social relationships between agricultural groups, despite existing ethnographic and ethnohistoric data. For example, intermediate units or neighborhoods could have been assimilated into larger co-residences and performed collective economic activities (Alexander 2000; Wilk 1991). In short, this topic needs more multidisciplinary investigations, which would focus on general agricultural systems, including spaces, soils, land features, practices, techniques and cultigens, water resources, and bioclimatic contexts. Research at high resolution on a large scale is still needed. Such work would help to make precise assessments on field system organization. This is critical because this organization, along with the settlement patterns, has been the basis for constructing the territories of ancient Maya cities.

The spatio-temporal contexts of La Joyanca and Río Bec La Joyanca and the Río Bec nuclear zone (Figure 7.1) are located 150 km apart and were defined by surveys associated with the La Joyanca-Peten Northwest Project (Arnauld et al. 2004) and the Río Bec Archaeological Project (Nondédéo et al. 2013). They share nearly equivalent areas (168 and 159 ha) and a contemporary apogee during the Late Classic period. Geographically, the natural settings of La Joyanca and Río Bec are similar (Métailié et al. 2002a, 2002b, 2003; Galop et al. 2004; Vannière et al. 2005, 2006). Both are located on a limestone plateau, 260 m high in Río Bec, and 120 m in La Joyanca, where two principal geomorphological units predominate, though Río Bec’s are more contrasted than La Joyanca’s. The high elevations (interfluves) are separated by deep drainage channels (thalwegs with a southwest–northeast orientation) and some seasonal bajos (Figures 7.6 and 7.7). The rendzina soils that characterize the area are relatively thin (25–50 cm), well-drained, and fertile in upper zones (Ha clay-silt horizon with a rich organic component), while the lower zones’ soils are thick and dominated by clays. On a hydrographic level (karstic net), La Joyanca’s region (Figure 7.2) is more humid as it is surrounded by an almost permanent swamp to the south (cibal), the San Pedro Martir River 4 km to the north, and the Tuspan Lagoon 5 km to the west. The tropical climate alternates between dry and humid seasons. Vegetation corresponds to a subperennifolia tropical rainforest with medium to high height (20–30 m); it is denser at La Joyanca than at Río Bec, and where Sapote trees (Manilkara sapota) dominate. La Joyanca shows a very long occupation sequence from the Middle Preclassic to the Terminal Classic, from 800 BC to AD 1000 (Forné 2006). During the fifth century, a stela was erected, along with the construction of an altar and an elaborate burial, in the largest residential group, Guacamaya, showing that this group could have been the dwelling of this small city’s royal dynasty. La Joyanca reached its apogee during the Late Classic and the beginning of the Terminal Classic, between AD 600 and 900, before occupation ended during a

long and progressive process between AD 950 and 1050 (Forné 2005). On a regional scale, the site is considered a third-level city on a five-level ranking constructed for the Northwest Petén by the La Joyanca Project (Arnauld et al. 2004: 40–42). La Joyanca is a 150-km² geographic unit defined by two natural plateaus surrounded by humid zones (Figure 7.2). It was likely a small urban center in the middle of a dispersed population of small dwelling groups and villages (Arnauld et al. 2004). Río Bec is the eponymous site of a larger Maya region known since the beginning of the twentieth century because of its original and highly decorated monumental architectural style. The region developed between AD 550 and 1000, limited by the Chenes and Puuc regions to the north and the Petén to the south, without any discontinuity in its settlement pattern (Figure 7.3). Inside Río Bec, Preclassic occupation remained minimal until AD 450. Between AD 450 and 600, a transitional period began, clearly visible in architecture and ceramics. This was followed by the apogee of the Río Bec cultural phenomena, marked by the construction of monumental structures that are unique in the Maya area. In addition, its sociopolitical organization distinguishes itself from the sacred royalty model of La Joyanca and many other Classic Maya entities (Michelet et al. 2008). In short, La Joyanca followed a very common path, starting as a Preclassic village and then becoming a Classic urban area with a political and religious center (including pyramids and a local dynasty), surrounded by a large residential zone with several dispersed monumental groups (Figure 7.6). Considering the development of its architecture, Río Bec was similarly prosperous, but without any conglomeration, and apparently without any political structure (or at least, without any identified political structure until today). The settlement was rather rural, marked by the dispersion of its monumental groups and by the presence of several agricultural features, which are still very visible on the surface, particularly as a long-known extensive network of terraces (Turner 1974, 1983) (Figure 7.7).

Figure 7.2

Map of the La Joyanca micro-region (modified from Arnauld et al. 2004: 43).

Figure 7.3

Map of the Río Bec micro-region (modified from Nondédéo et al. 2013: 374).

Testing a new model of agrarian settlement ecology in the Maya Lowlands In La Joyanca, because of the complete absence of any visible agricultural structures on the surface, the hypothesis of intra-residential agriculture during the Classic period was hard to formulate and demonstrate. This is in contrast to Río Bec, where it seems that land tenure was a determinant parameter that structured the settlement pattern. Nonetheless, using the project map available in 1999 (Figure 7.4), a double hypothesis was raised. First, in accordance with the irregular repartition of the households (a repartition that leaves extensive spaces, supposedly free from constructions) and the apparently strong hierarchy between the different households (monumental households versus modest ones), it is proposed that the settlement pattern included a strong proportion of cultivated plots on extensive construction-free spaces between households (i.e. spaces with no visible construction). In addition, those lands were potentially the property of the individuals living in the monumental households, and the subordinated multi-familiar groups who cultivated them. Thus, the investigation focused on the relationships between the architectural remains and environmental components of La Joyanca in order to reconstruct elements of its social and agrarian organization during its peak (Abril 2 phase, AD 750–850).

Figure 7.4

Methods applied to La Joyanca (modified from Lemonnier 2009: figures 4.5, 4.3, 6.2).

In Río Bec, the spatial layout of the site and the presence of probable agricultural structures allow us to offer a more detailed hypothesis, documenting the agrarian system in a more accurate way. The aforementioned structures are stone ridges, or linear stone berms, regionally identified in Becan, Chicanna, Xpujil, and Hormiguero and called camellones (Thomas 1981; Carrasco et al. 1986; Pollock 1967; Peña 1987), agricultural terraces, and water reservoirs. The goal of this study was to construct a model, for the apogee period in Río Bec (Makan phase, AD 700–850) that would reflect the agricultural production and use of resources. On a larger scale, the goal was to determine the spatial and functional relationship between field systems and households.

Methodology A multidisciplinary and multi-scalar strategy was applied at both sites, in collaboration with geographers from each respective project. In La Joyanca, geoarchaeological studies were carried out with Jean-Michel Carozza. Didier Galop and Jean-Paul Métailié provided us with independent data from their own investigation concerning the paleo-environment. At Río Bec, a new geoarchaeological approach for the Maya area was tested by Boris Vannière, combined with conventional excavation and survey methods like those employed in La Joyanca.

Laboratory work, in addition to chronological investigations, included the elaboration of morphological and functional typologies, spatial analysis (Peterson and Drennan 2005), demographic calculation, evaluation of agricultural production, and modeling (with the invaluable support of ethnohistory, ethnography, and ethnoarchaeology as a means to elaborate interpretations). With reference to the set objectives, data were collected at La Joyanca using the following methods (Figure 7.4). A systematic and intensive survey of 80 percent of the residential zone allowed us to complete the original map. Then we were able to establish a household hierarchy based on morphological criteria that would represent the social stratification. Finally, a series of spatial analyses were conducted between households, the civic-ceremonial center, and a series of environmental parameters. These took into account all chronological data and it is estimated that 65 percent of the household were simultaneously occupied in AD 850 (Forné 2006). In addition, a small household (Gavilan Group) was excavated, as an example of the minimal social unit of the city (extended family). This was done so that it could be compared with elite households, which were being excavated by other project members at the same time. Excavating this household unit allowed us a view of its spatio-temporal dynamic over three centuries and its functional organization during its most intensive occupation, circa AD 850, which corresponds to La Joyanca’s apogee. This approach produced chronological, spatial, functional, social, and demographic data, all of which were useful for the larger-scale interpretation of La Joyanca’s settlement ecology. At Río Bec, three zones of study were selected (0.53 ha, 8 ha, and 50 ha) according to the diversity of their environmental components and the morphological and dimensional variations of their structures and intermediate spaces, as well as the location of specific archaeological excavation areas. In addition to B. Vannière’s geomorphological and hydrological study, there was a total of four methodological tools employed (Figure 7.5). A high-resolution microtopographic survey, which was a pioneering cartographic method for the Maya area, was conducted (B. Vannière), providing a “digital terrain model” to understand the anthropogenic management of the landscape. Selected archaeological excavations were undertaken on a sample of the identified structures in order to define categories of land use features and buildings as well as obtain chronological data. Pedological test pits (B. Vannière) were aimed at characterizing soils and identified spaces. An intensive and systematic pedestrian survey, similar to that completed at La Joyanca, allowed for a completed site map and enabled us to integrate our data and refine our interpretations.

Figure 7.5

Methods applied to Río Bec.

The first three approaches were applied to the 8-ha zone (“zone 8”), in the middle of which a higher resolution spatial analysis was completed (0.53 ha, “workshop-zone”). The results of the three combined methods allowed for an accurate reconstruction and spatial model of the workshop-zone, which could be later extrapolated for zone 8. This model was applied to a larger territory, in a third zone of study that measured nearly 42 ha (“zone 33” + 9 ha nearby Group A), where it was validated. The total studied area reflected agricultural production covering 50 ha, or approximately one-third of the size of the nuclear zone of Río Bec (159 ha). Chronological control of structures and features was based on the chronoceramic sequence (elaborated by S. Dzul). The analysis of all the data from La Joyanca and Río Bec, specifically the social and agrarian aspects which help determine settlement strategies as well as local demography and agricultural production, was integrated to define two distinct models for the socioeconomic organization of these sites during their maximum occupation between AD 600 and 900.

Settlement pattern and land use at La Joyanca At La Joyanca (Figure 7.6), the residential zone reached 165 ha and included more than 600 structures around courtyards (5.2 structures/ha, or 1.4 patios/ha). A series of spatial and morphological analyses of the principal site components (Figure 7.7) allow us to determine 11

neighborhoods, which are defined as intermediate social units between extended family and community (Lemonnier 2009, 2011, 2012a,b; Lemonnier and Arnauld 2008; Breuil-Martínez et al. 2004). Criteria for an archaeological identification of the neighborhoods were as follows: 1) each should have a spatial delineation (seasonal swamps and “vacant” spaces); 2) an internal spatial arrangement must be ranked in a hierarchal fashion; 3) the identification of a temporal growth that can be modeled (according to the chronologic sequence developed by Mélanie Forné 2005); 4) the existence of a (probable) cultivated area that should be spatially associated with the unique monumental household within the neighborhood; and finally 5) each neighborhood should represent a concentration of modest households which are grouped into clusters, nucleated around the elite’s residence, with an “agricultural domain” that probably included cultivated lands (between households, and along the seasonal swamps). Together these criteria predict that in social terms each neighborhood would have pertained to a single elite family that controlled lower-ranked subordinated families. On the economical level, it is assumed that the basis of each neighborhood was agricultural. In addition, the presence of large “empty” zones close to elite residences (from 1.5 up to 5 ha, with thick soils conformed by rich organic matter) suggests a collective and more or less intensive exploitation of an agricultural territory, a part of which would be contained within the neighborhood. This preliminary interpretation of the socioagricultural system at La Joyanca is based on an accurate evaluation of demography and agricultural production during La Joyanca’s peak occupational period, c. AD 850. Demographic calculations were based on two coefficients, developed and validated from ethnographic and archaeological research (Becquelin and Michelet 1994), which were applied to all the contemporaneous structures, leading to an estimated population of 1200 inhabitants.

Figure 7.6

Site plan of La Joyanca (168 ha, 635 structures) during the Late Classic period.

In order to calculate agricultural production, we employed indexes developed by previous agronomic studies that were undertaken in the actual villages nearby La Joyanca in which people practice swidden agriculture (Effantin 2000). Those indices estimate that the annual cultivation of a single 1-ha plot could be estimated for a family made up of five individuals. Based on this estimate we could gauge that a population of 1200 required nearly 240 ha of agricultural land. If we extract surfaces occupied by households, rocky outcrops, and bajo zones in La Joyanca, the estimated cultivable land drops by nearly half to 115 ha. In other words 70 percent of the residential zone would be agricultural, or nearly 60–80 percent would be “empty” which has been confirmed at other Maya sites (Smyth et al. 1995: 322). Therefore, considering the hypothesis of intra-site fields (not gardens), almost half of the needed annual production could have been supported by infield agriculture. The other half would have been produced in outfield, probably on the plateau. On this large plateau (10 km²), where the population likely reached nearly 500 inhabitants (still using La Joyanca’s methodology for demographic counting), the land available for agriculture reached 700 ha (excluding La Joyanca)—which could have supported up to 3500 people. However, if we take into account the same agricultural index (1 ha/5 persons) and a ratio cultivation/fallow of 1/2–3 years (which is a minimum), we get to understand that each family would need at least 2 to 3 ha to achieve the minimum rotation of cultivations, that is to say, in total: 700–1000 ha (1700 persons divided into 240 “urban” families and 100 “rural” ones). Although hypothetical, this

simulation indicates that the plateau would have been quickly saturated, which raises the possibility that over-cultivation may have the result of a growing population and more intensive agricultural practices. At present, any demographic density of 10–15 inhabitants/km² implies the use of intensive techniques (J.P. Métailié: pers. comm.). At La Joyanca, Classic demographic density likely reached 170 persons/km² (1700 persons/10 km²); this suggests a probable need to intensify agriculture that would have manifested itself close to the households, which is similar to current farmers in the region (Effantin 2006; also see Drennan 1988; Netting 1977).

Figure 7.7

La Joyanca, neighborhoods and agricultural systems during the Late Classic period (modified from Lemonnier 2009: Figure 7.9; Lemonnier 2011: figures 9, 10).

Land use and settlement at Río Bec Unlike La Joyanca, Río Bec (Figure 7.8) demonstrated that each household had access to one production unit, clearly surrounded by artificial (camellones) or natural topographic boundaries (topography, or swamps). In this arrangement, each agrarian unit was likely distinct from its neighbors and, therefore, probably autonomous. In the 50 ha studied at Río Bec (Lemonnier and Vannière 2013), 26 agricultural production units (APUs) were distinguished (Figure 7.9). Each APU was defined as follows. First, it is composed of a single household, whether

monumental or not; rare cases of multiple households are ranked so that they can be interpreted in terms of subordination relationship. Second, it is defined by camellones, quarries and/or distinct or subtle changes in the local topography. Third, it would have been managed by way of different structures, like terraces, oblong-shaped stone piles (maybe as a result of clearing out of soil or as a reserve for future construction), or water reservoirs (aguadas). Little circular-shaped stone piles, which date after the occupational peak, have been preserved in the Late–Terminal Classic landscape and interpreted as ruins and/or symbolic or territorial markers (Vapnarsky and Le Guen 2006). Land management features contributed to create a notable division of space, including small sections or compartments which were laid out as small plots. In the workshop-zone, they measure from 500 m2 up to 1000 m² and have been interpreted, according to their own characteristics, as specialized agricultural plots, each one being complementary to the others, and intensively cultivated. Fourth, different agricultural techniques have been identified across APUs. In the compartments, soils had been modified, conditioned, and improved. In addition to the sloping surface with terraces (erosion preventing), clayey soils (60–80 cm thick) have been observed in natural depressions and quarries (see Gillot 2010), probably as the result of an anthropic modification of bringing clayey soils for specific cultivations. Also some soils are free of stones and some soils appear to have been fertilized (near absence of ceramic sherds, possibly due to some technique using domestic waste as fertilizer). Regarding general ceramic distribution, we were able to differentiate circulation spaces (absence of material), residential spaces (abundant material), and cultivated spaces (less material). Finally, there was good rainwater management; this was reflected through the presence of camellones—oblong-shaped stone piles, and compartments —which protected some sectors from floods, as they evacuated and directed water. Other camellones were located on the opposite slopes and may testify to the existence of ancient structures such as dams, probably made of perishable materials.

Figure 7.8

Site plan of Río Bec (159 ha) during Late Classic period.

Figure 7.9

Agrarian production units in Río Bec (50 ha) during the Late Classic period (modified from Lemonnier and Vannière 2013: fig. 8).

The data collected from Río Bec indicate that the cultural landscape was highly modified and structured. It also likely that agricultural production and the corresponding work was done at the household level. Each household’s management of several types of land and associated agricultural structures helped shape their individual construction and cultivation. Such autonomy of APUs has to be considered as closely associated with residential construction. For example, the medium size for the smallest APUs is 0.5 ha (1/4–1 ha), which corresponds to intra-settlement plots in others Classic Maya sites, including those more urban than Río Bec (Lemonnier 2009: 85–86). Agricultural productivity could have been high in Río Bec in such managed and probably intensively worked spaces, which were organized as infield agriculture close to the households (Netting 1977). In the workshop-zone (0.5-ha APU), as we apply the same demographic (6.87–8.36 m²/pers.) and agricultural (1 ha per 5 persons) indices as at La Joyanca, we calculated that each house (18.25 m², or 2–3 persons) would have been relatively

self-sufficient with a very small plot (0.5 ha). Unlike La Joyanca, Río Bec indicates that the debate concerning the role of outfield agriculture is still important given the substantive Classic period household remains.

Discussion and conclusion With a rather simple methodology, this chapter has demonstrated that it is possible to construct two models of agrarian settlement ecology, specifically the social and agrarian organizational schemes for La Joyanca and Río Bec during their respective apogee periods. The first model highlights the integration of neighborhoods and infield–outfield agriculture at La Joyanca, while an alternative model can be offered for Río Bec which emphasizes household autonomy and infield intensification. Both systems were deduced by very different settlement patterns from one site to another. From one perspective we see concentrations of commoner dwellings centered around an elite residence and a land property. Conversely the other perspective offers a continuous network of households located in the middle of their own land, forming almost orthonormal units separated from one another by physical limits still visible in the landscape. At La Joyanca, only neighborhoods are delimited by boundaries. The commoner households seem to form several small discrete settlement clusters, but any ancient land management is visible on the surface, between households—except some aguadas scattered in the clusters. These water reservoirs are interesting in that they probably reflect human construction (they did not influence the residential area, but resulted from it) and their location, their quantity in comparison with the number of households, and their size suggest that these structures were shared by members of clusters (Lemonnier 2009: 191, 196–197). There is the same pattern for the co-residential space (clusters and neighborhoods) and maybe for the agricultural space, which seems undifferentiated from one household to another within each neighborhood. Perhaps the only exception is the large and fertile vacant space spatially associated with each elite residence. These possible elite-owned lands allow us to suggest that other plots or fields could have been cultivated within the city and the neighborhoods. Additional arguments in favor of the influence of the agrarian rules in the settlement pattern are the regular distance between households (with a theoretical average of 0.73 ha/household) and maybe their similar orientation within the settlement as a whole. It seems clear in the case of La Joyanca that the deciding factor for location of commoners was the proximity of the elite houses. Did they provide identity and particularly lands through worship of ancestors (McAnany 1995)? The spatial arrangement in households, clusters and neighborhoods is mainly due to the fundamental social structure of Maya societies (a cultural tradition of extended family and lineage—Haviland 1988; Hendon 2000), but probably also due to the exercising of collective activities including labor scheduling (Alexander 2005; Wilk 1988). We must go back in time to better understand the construction of the city of La Joyanca. During the Preclassic period, the first social groups of La Joyanca settled on the edge of the plateau, on the highest part, near good farmlands and the main water source (cibal). On the lakeside located 5 km west of the settlement, paleoenvironmental studies reflect intensive agricultural activity, which suddenly ceased around AD 550 when La Joyanca was just entering

its apogee phase (Fleury et al. 2014). The hypothesis of a migration of rural people to the city (Galop et al. 2004) was supported by a test-pits program undertaken in 2012 within the La Joyanca neighborhoods, which showed that Early Classic and Late Classic occupancy rates jumped from 65 to 95 percent (Arnauld et al. 2013a, 2013b). As high-rank residences generally show longer and earlier chronological sequences than low-rank dwellings, it is assumed that elites attracted a rural population, which resulted in the formation of neighborhoods. Data from the same program also suggest that some of the elite households would have absorbed a number of commoner households. That being said, if we accept the general hypothesis of La Joyanca’s urbanization by migration, we also can accept, on the basis of the internal neighborhood hierarchy, that elites would have controlled lower-rank people and that they would have scheduled economic activities, especially agricultural labor. Ethnographic and ethnohistoric neighborhoods in the Maya area are described as socioeconomic units in several earlier studies (Carrasco 1976a, 1976b; Hayden and Cannon 1982; Wilk 1988; Wilk and Netting 1984). Furthermore, agricultural cooperation (or labor pooling) may have been one of the critical factors explaining the co-residence principle and, conversely, infield agriculture (stimulated by demographic and/or economic pressure) may have been facilitated by social structure. Families living in co-residence within households, clusters, and neighborhoods would have been linked to each other and to the land by a common ancestor via the elite family, to which some of them would have been related by descent and/or marriage. This model may not be appropriate for the polity of Río Bec, where each household appears to have had all the facilities necessary for its production and consumption, and to have been independent from its nearby households, be they large or small. However, when reconstructing the occupational sequence for the closest dwelling units, it has been observed that monumental units tended to absorb modest units and their territory in various ways (marriage? expulsion? colonization? [see Arnauld et al. 2012, 2013c]). This is expressed (as at La Joyanca) by the abandonment of small households located within the boundaries of a large APU when the corresponding household was enlarged and became more monumentalized. Unfortunately, we do not have enough data for Río Bec concerning the occupation levels that preceded the peak, and we have even less information on the agricultural system. What is clear is that landscape management was contemporary with the apogee period and that this period was characterized by a new structuring of space for the purpose of farming, which would have affected households of all ranks. At a broader level, can La Joyanca’s model, with its strong neighborhood socioeconomic integration, be compared with Río Bec’s one, with its small autonomous units living side by side with greater and more powerful social groups? This chapter suggests that such a comparison is warranted and needed to fully understand the variable nature of how Maya societies used and spatially organized the landscape. A first evaluation at Río Bec indicates a strong correlation between the household rank and the land area of its corresponding APU (Table 7.1). We also observe a solid continuum in the surface of APU from rank VI (0.25 ha) to I (4.5 ha). Did the APU’s surface depend on the household’s rank? Is the exploitation of a large land area the mechanism that enabled households to grow? In La Joyanca, a similar correlation is possible between the degree of complexity of one elite’s neighborhood residence, and the associated cultivated area (Table

7.2). In other words, the largest “empty” zones or cultivated areas (and the biggest neighborhoods) are associated with the most developed elite residences (in terms of morphology and construction volume), which are the oldest (with the longest occupational sequences) and the closest to the civic-ceremonial center (Guacamaya, Tepescuintle, Venado, and Ardilla groups). These results, achieved by two independent methods, clearly provide a coherent examination highlighting the relationship between agricultural production, quality of the household, and demography: a pattern well-supported by previous ethnographic work (Wilk 1984, 1985, 1988, 1991). Would agriculture, combined with an important workforce, have been one source for household prosperity? At that point, both models show a trend of increasing numbers of noble houses and the size of their agricultural territories. In Río Bec, this may have happened due to the absorption of households into another residential system with monumental architecture; at La Joyanca, a similar dynamic may have enabled the “integration” of co-resident families which, however, appear to have kept their residential autonomy via more modest architectural units. The exception to this trend appears to be that in some cases these architectural units were either absorbed, like at Río Bec, or ultimately abandoned. Table 7.1

Comparison of household rank and APU surface estimates at Río Bec



Household number

Household rank






Group B

























































Gr B





Gr C & D





Gr B








1.3 min

Gr D




Gr Q

Gr Q



2.5 min

Gr B

Gr H



1.9 min

Sur Gr C & D Gr B Gr C & D Sur Gr B Gr C & D

Gr C & D

Surface APU (ha)

Gr J



2.25 min

Gr C & D

Gr C



3.5 min

Gr A

Gr A




Gr B

Gr B









Table 7.2

Comparison of household rank of individual neighborhood and the surface of associated cultivated area at La Joyanca


Elite household rank

Household number

Structures number

“Empty space” surface





5.1 (+5)















? (in periphery)
















Loro Real

























One of the principal growth factors of the Classic Maya cities was to build an urban space integrating architecture and agriculture (Martínez Hidalgo et al. 1999). Maya agriculture is seen as the foundation of the Maya societies’ growth and prosperity in the long term. The influence of this agriculture on settlement patterns nonetheless remains little studied. The observations on the agrarian and settlement ecology at the Maya sites of La Joyanca and Río Bec reveal that if resources, in particular water and soils, were definitely crucial in the earlier occupations, it is clear that the influences were diverse in the evolution of the spatial layout of each settlement. The spatially defined agrarian landscape would have played a crucial role in shaping the city, but also in the demographic and social organization dynamics. Of course the arrangement of these Maya cities was used as a political “tool” by the elite, who held an important sociopolitical role in all Maya cities throughout the lowland region (Chase and Chase 1992; Inomata and Houston 2001).

References Abrams, E.M., and D.J. Rue 1988 The Causes and Consequences of Deforestation among the Prehistoric Maya. Human Ecology 16(4): 337–396. Adams, Richard E.W., H.R. Robichaux, Fred Valdez Jr., Brett A. Houk, and Ruth Mathews 2004 Transformations, Periodicity, and Urban Development in the Three Rivers Region. In The Terminal Classic in the Maya Lowlands: Collapse, Transition and Transformation, A.A. Demarest, P.M. Rice, and D.S. Rice (eds), pp. 324–341. Westview Press, Boulder,

CO. Aimers, James J., and David A. Hodell 2011 Drought and the Maya. Nature 479: 44–45. Alexander, Rani T. 2000 Patrones de Asentamiento Agregados en el Sudoeste de Campeche: Una Visión desde la Isla Cilvituk. Mesoamérica 39: 359–391. Alexander, Rani T. 2005 Isla Cilvituk and the Difficulties of Spanish Colonization in Southwestern Campeche. In The Postclassic to Spanish-era Transition in Mesoamerica: Archaeological Perspectives, S. Kepecs and R. Alexander (eds), pp. 161–183. University of New Mexico Press, Albuquerque. Arnauld, M. Charlotte 2008 Maya Urbanization: Agrarian Cities in a Preindustrial World (Urbanización Maya: Ciudades Agrarias en un Mundo Pre-Industrial). In Urbanism in Mesoamerica (El Urbanismo en Mesoamerica), Vol. II, A.G. Mastache, R.H. Cobean, A.C. Cook, and K.G. Hirth (eds), pp. 1–36. INAH, Mexico City and Pennsylvania State University, University Park. Arnauld, M. Charlotte, and Dominique Michelet 2004 Nature et Dynamique des Cités Mayas. Annales, Histoire, Sciences Sociales 59: 73–108. Arnauld, M. Charlotte, Véronique Breuil-Martínez, and Erick Ponciano Alvarado 2004 La Joyanca (La Libertad, Guatemala), Antigua Ciudad Maya del Noroeste del Petén. CEMCA, Asociación Tikal, CIRMA, Guatemala. Arnauld, M. Charlotte, D. Michelet, B. Vannière, P. Nondédéo, and E. Lemonnier 2012 Houses, Emulation and Cooperation among the Río Bec Groups. In The Neighborhood as a Social and Spatial Unit in Mesoamerican Cities, M.C. Arnauld, L.R. Manzanilla and M.E. Smith (eds), pp. 202–208. University of Arizona Press, Tucson. Arnauld, M.-Charlotte, Eva Lemonnier, Mélanie Forné, and Didier Galop 2013a The Rise and Fall of a Secondary Polity: La Joyanca (Guatemala). In Millenary Maya Societies: Past Crises and Resilience, M.-Charlotte Arnauld and Alain Breton (eds), pp. 148–168. Electronic document, published online at Mesoweb: www​.me​sow​eb.​com​/pu​bli​cat​ion​s/M​MS/​10_​Arn​aul​d_e​tal​.pd​f (accessed 7 January 2016). Arnauld, M.-Charlotte, Eva Lemonnier, Mélanie Forné, Erick M. Ponciano, Julien Sion, and Didier Galop 2013b Early to Late Classic Population Mobility in La Joyanca, Northwestern Petén, Guatemala. Paper presented at the 78th Annual Meeting for the Society for American Archaeology, Honolulu. Arnauld, M.C., D. Michelet, and P. Nondédéo 2013c Living Together in Río Bec Houses: Co-Residence, Rank and Alliance. Ancient Mesoamerica 24(2): 469–493. Beach, Timothy P. 1996 Estudios de Catenas, Fertilidad y Fosfatos de Suelos. In PPdC: Informe Preliminar, A.A. Demarest, H. Escobedo, and M. O’Mansky (eds.), pp. 86–94. Guatemala City: Instituto de Antropología e Historia, and Nashville: Department of Anthropology, Vanderbilt University. Becquelin, Pierre and Dominique Michelet 1994 Demografía en la Zona Puuc: el Recurso del Método. Latin American Antiquity 5(4): 289–311. Binford, M.W., M. Brenner, T.J. Whitmore, A. Higuera-Gundy, E.S. Deevey, and B.W. Leyden 1987 Ecosystems, Paleoecology and Human Disturbances in Subtropical and Tropical America. Quaternary Science Reviews 6: 115–128. Braswell, Geoffrey E., Joel D. Gunn, María Del Rosario Domínguez Carrasco, William J. Folan, Laraine A. Fletcher, Abel Morales López, and Michael D. Glascock 2004 Defining the Terminal Classic at Calakmul, Campeche. In The Terminal Classic in the Maya Lowlands: Collapse, Transition and Transformation, A.A. Demarest, P.M. Rice, and D.S. Rice (eds), pp. 162–194. Westview Press, Boulder, CO. Brenner, Mark 1983 Paleolimnology of the Peten Lake District, Guatemala, II. Mayan Population Density and Sediments and Nutrient Loading of Lake Quexil. Hydrobiologia 103: 205–210. Breuil-Martínez, Véronique, Eva Lemonnier, and Erick Ponciano Alvarado 2004 Vivir en La Joyanca. In La Joyanca (La Libertad, Guatemala), Ciudad Maya del Noroeste de Petén, M.C. Arnauld, V. Breuil-Martínez, and E.Ponciano Alvarado (eds), pp. 73–93. CEMCA, Asociación Tikal, CIRMA, Guatemala. Carrasco, Pedro 1976a The Joint Family in Ancient Mexico: The Case of Molotla. In Essays on Mexican Kinship, H.G. Nutini, P. Carrasco, and J.M. Taggart (eds), pp. 45–64, University of Pittsburgh Press, Pittsburgh. Carrasco, Pedro 1976b La Sociedad Mexicana Antes de la Conquista. In Historia General de México, Vol. 1, D. Cosio Villegas (ed.), pp. 165–286. El Colegio de México, Mexico City. Carrasco, R., S. Boucher, and A. Peña 1986 Río Bec: Un Modelo Representativo del Patrón de Asentamiento Regional. Boletín de la Escuela de Ciencias Antropológicas de la Universidad de Yucatán 13(78): 20–30. Chase, Arlen F., and Diane Z. Chase 1998 Scale and Intensity in Classic Period Maya Agriculture: Terracing and Settlement at the “Garden City” of Caracol, Belize. Culture and Agriculture 20(2/3): 60–77. Chase, Diane Z., and Arlen F. Chase (eds) 1992 Mesoamerican Elites: An Archaeological Assessment. University of Oklahoma Press, Norman. Cowgill, U.M., and G.E. Hutchinson 1963 E l Bajo de Santa Fe. Transactions of the American Philosophical Society 53 (7): 3–44. Culbert, T. Patrick 1977 Maya Development and Collapse: An Economic Perspective. In Social Process in Maya Prehistory: Studies in Honour of Sir Eric Thompson, N. Hammond (ed.), pp. 509–531. Academic Press, New York. Culbert, T. Patrick 1988 The Collapse of Classic Maya Civilization. In The Collapse of Ancient States and Civilizations, N.

Yoffee and G. Cowgill (eds), pp. 69–101. Arizona University Press, Tucson. Curtis, Jason H., David A. Hodell, and Mark Brenner 1996 Climate Variability on the Yucatan Peninsula (Mexico) during the Past 3500 Years and Implications for Maya Cultural Evolution. Quaternary Research 46: 37–47. Dahlin, B.H., and A. Dahlin 1994 Platforms in the Aklache at El Mirador, Peten, Guatemala and their Implications. Geoarchaeology 9: 203–237. Dahlin, B.H., J.E. Fosse, and M.E. Chambers 1980 Project Akalches: Reconstructing the Natural and Cultural History of a Seasonal Swamp: Preliminary Results. In El Mirador, Petén, Guatemala: An Interim Report. Papers of the New World Archaeological Foundation Publication 45, R.T. Matheny (ed.), pp. 37–58. Brigham Young University, Provo, UT. Deevey, E.S., D.S. Rice, P.M. Rice, H.H. Vaughan, M. Brenner, and M.S. Flannery 1979 Maya Urbanism: Impact on a Tropical Karst Environment. Science 206(19): 298–306. Demarest, Arthur A. 2006 The Petexbatún Regional Archaeological Project—A Multidisciplinary Study of the Maya Collapse. VIMA Series Monograph Vol. 1, University of Vanderbilt, Nashville, TS. Drennan, Robert D. 1988 Household Location and Compact versus Dispersed Settlement in Prehispanic Mesoamerica. In Household and Community in the Mesoamerican Past, R.R. Wilk and W. Ashmore (eds), pp. 273–293. University of New Mexico Press, Albuquerque, NM. Dunning, Nicholas P. 1996 A Reexamination of Regional Variability in the Prehispanic Agricultural Landscape. In The Managed Mosaic: Ancient Maya Agriculture and Resource Use, S.L. Fedick (ed.), pp. 53–68. University of Utah Press, Salt Lake City. Dunning, Nicholas P. 2003 Birth and Death of Waters: Environmental Change, Adaptation, and Symbolism in the Southern Maya Lowlands. In Espacios Mayas: Usos, ((Representaciones, Creencias, A. Breton, A. Monod-Becquelin, and M. Humberto Ruz (eds), pp. 49–76. UNAM, CEMCA, Mexico. Dunning, N.P., T.P. Beach, P. Farrell, and S. Luzzadder-Beach 1998a Prehispanic Agro-systems and Adaptive Regions in the Maya Lowlands. Culture and Agriculture 20(2/3): 87–101. Dunning, Nicholas P., Timothy P. Beach, and David Rue 1997 The Paleoecology and Ancient Settlement of the Petexbatún Region, Guatemala. Ancient Mesoamerica 8(2): 255–266. Dunning, N.P., S. Luzzadder-Beach, T. Beach, J.G. Jones, V. Scarborough, and T.P. Culbert 2002 Arising from the Bajos: The Evolution of a Neotropical Landscape and the Rise of Maya Civilization. Annals of the Association of American Geographers 92(2): 267–283. Dunning, N.P., D. Rue, T. Beach, A. Covich, and A. Traverse 1998b Human-Environment Interactions in a Tropical Watershed: The Paleoecology of Laguna Tamarindito, El Petén, Guatemala. Journal of Field Archaeology 25: 139–151. Dunning, N.P., V. Scarborough, F. Valdez Jr., S. Luzzadder-Beach, T. Beach, and J.G. Jones 1999 Temple Mountains, Sacred Lakes and Fertile Fields: Ancient Maya Landscapes in Northwestern Belize. Antiquity 73: 650–660. Effantin, Rachel 2000 Frontière Agraire et Aire Protégée dans les Basses Terres Mayas (Petén, Guatemala). Les Conditions d’appropriation et d’exploitation des “Terres Neuves” par les Paysans Migrants. DEA ETES, Paris. Effantin, Rachel 2006 De la Frontière Agraire à la Frontière de la Nature: Comment les Migrants Réinventent Leurs Ressources et Leurs Territoires dans la Réserve de Biosphère Maya (Petén, Guatemala). INRA, Paris-Grignon/ABIES, Paris. Fedick, Scott L. 1994 Ancient Maya Agricultural Terracing in the Upper Belize River Area: Computer-Aided Modeling and the Results of Initial Field Investigations. Ancient Mesoamerica 5(1): 107–127. Fedick, Scott L. (ed.) 1996 The Managed Mosaic: Ancient Maya Agriculture and Resource Use. University of Utah Press, Salt Lake City. Fleury, S., B. Malaizé, J. Giraudeau, D. Galop, V. Bout-Roumazeilles, P. Martinez, K. Charlier, P. Carbonel, and M.C. Arnauld 2014 Impacts of Maya Land Use on the Laguna Tuspán Watershed (Petén, Guatemala) as seen through Clay and Ostracod Analysis. Journal of Archaeological Science 49: 372–382. Ford, Anabel 2008 Dominant Plants of the Maya Forest and Gardens of El Pilar: Implications for Paleoenvironmental Reconstructions. Journal of Ethnobiology 28(2): 179–199. Forné, Mélanie 2005 Etude Céramique de Tuspan 2, La Joyanca, Petén, Guatemala. Ms, Projet PNO-La Joyanca. Forné, Mélanie 2006 La Cronología Cerámica de La Joyanca, Noroeste del Petén, Guatemala. British Archaeological Reports International Series 1572, Paris Monographs in American Archaeology 17. Series Editor. Eric Taladoire. Oxford, UK. Freter, AnnCorinne 1994 The Classic Maya Collapse at Copan, Honduras: An Analysis of Maya Rural Settlement Trends. In Archaeological Views from the Countryside: Village Communities in Early Complex Societies, G.M. Schwartz and S.E. Falconer (eds), pp. 160–176. Smithsonian Institution, Washington, DC. Galop, D., E. Lemonnier, J.-M. Carozza, and J.-P. Métailié 2004 Bosques, Milpas, Casay Aguadas de Antaño. In La Joyanca (La Libertad, Guatemala), Antigua Ciudad Maya Del Noroeste Del Petén, M.C. Arnauld, V. Breuil-Martínez, and E. Ponciano Alvarado (eds), pp. 55–71, CEMCA, Asociación Tikal, CIRMA, Guatemala. Gill, Richardson B. 2000 The Great Maya Droughts: Water, Life, and Death. University of New Mexico Press, Albuquerque. Gillot, Céline 2010 Operación IVC: Inventario Arquitectural y Modos de Construcción, Análisis de las Zonas de Extracción. In

Proyecto Río Bec (Campeche México), Informe 8, Octava Temporada de Campo, 2009, D. Michelet et al. CNRS, laboratoire “Archéologie des Amériques” (UMR 8096), Centre d’Études Mexicaines et Centraméricaines (CEMCA), Mexico. Gómez-Pompa, A., M.F. Allen, S.L. Fedick, and J.J. Jiménez-Osornio (eds) 2003 The Lowland Maya Area: Three Millennia at the Human–Wildland Interface. Food Products Press, Haworth Press, New York, London, and Oxford. Guilaine, Jean 1991 Pour une archéologie agraire. Armand Colin, Paris. Harrison, Peter D., and B.L. Turner II (eds) 1978 Pre-Hispanic Maya Agriculture. University of New Mexico Press, Albuquerque. Hayden, Brian, and Aubrey Cannon 1982 The Corporate Group as an Archaeological Unit. Journal of Anthropological Archaeology 1(2): 132–158. Haviland, William A. 1988 Musical Hammocks at Tikal: Problems with Reconstructing Household Composition. In Household and Community in the Mesoamerican Past, R.R. Wilk and W. Ashmore (eds), pp. 121–134. University of New Mexico Press, Albuquerque. Hendon, Julia A. 2000 Having and Holding: Storage, Memory, Knowledge, and Social Relations. American Anthropologist 102: 42–53. Hodell, D.A., J.H. Curtis, and M. Brenner 1995 Possible Role of Climate in the Collapse of Classic Maya Civilization. Nature 375: 391–394. Houston, Stephen, and Hector Escobedo (eds) In press In the Land of the Turtle Lords. University of Oklahoma Press, Norman. Inomata, Takeshi, and Stephen D. Houston (eds) 2001 Royal Courts of the Ancient Maya, 2 vols. Westview Press, Boulder/Oxford. Jacob, John S. 1994 Evidencias para Cambio Ambiental en Nakbe, Guatemala. In VII Simposio de Investigaciones Arqueológicas en Guatemala 1993, J.P. Laporte and H.L. Escobedo (eds), pp. 275–280. Museo Nacional de Arqueología y Etnología, Ministerio de Cultura y Deportes, Instituto de Antropología e Historia, Asociación Tikal, Guatemala. Jacob, John S. 1995 Archaeological Pedology in the Maya Lowlands. In Pedological Perspectives in Archaeological Research: Proceedings of Two Symposia Sponsored by Division S-5 of the Soil Science Society of America in Cincinnati, OH, 8 November 1993, Mary E. Collins (ed.), pp. 51–80. Special Publication 44, Soil Science Society of America, Madison, WI. Johnston, Kevin 2003 The Intensification of Pre-Industrial Cereal Agriculture in the Tropics: Boserup, Cultivation Lengthening, and the Classic Maya. Journal of Anthropological Archaeology 22(2): 126–161. Kennett, Douglas J., Sebastian F.M. Breitenbach, Valorie V. Aquino, Yemane Asmerom, Jaime Awe, James U.L. Baldini, Patrick Bartlein, Brendan J. Culleton, Claire Ebert, Christopher Jazwa, Martha J. Macri, Norbert Marwan, Victor Polyak, Keith M. Prufer, Harriet E. Ridley, Harald Sodemann, Bruce Winterhalder, and Gerald H. Haug 2012 Maya Political Systems in Response to Climate Change. Science 338: 788–791. Killion, Thomas W. (ed.) 1992 Gardens of Prehistory: The Archaeology of Settlement Agriculture in Greater Mesoamerica. University of Alabama Press, Tuscaloosa. Killion, Thomas W., Jeremy A. Sabloff, Gair Tourtellot, and Nicholas P. Dunning 1989 Intensive Surface Collection of Residential Clusters at Terminal Classic Sayil, Yucatán, Mexico. Journal of Field Archaeology 16(3): 273–294. Kunen, Julie L. 2001 Ancient Maya Agricultural Installations and the Development of Intensive Agriculture in NW Belize. Journal of Field Archaeology 28: 325–346. Kunen, Julie L. 2004 Ancient Maya Life in the Far West Bajo: Social and Environmental Change in the Wetlands of Belize. University of Arizona Press, Tucson. Lemonnier, Eva 2009 La Structure de l’Habitat du Site Maya Classique de la Joyanca (Petén Nord-Ouest, Guatemala) dans son Environnement Local. Paris Monographs in American Archaeology 23. BAR International Series. Oxford. Lemonnier, Eva 2011 Des quartiers Chez les Mayas à l’Époque Classique? Journal de la Société des Américanistes 97(1): 7–50. Lemonnier, Eva 2012a Neighborhoods in Classic Lowland Maya Societies: Their Identification and Definition from La Joyanca Case Study (Northwestern Peten, Guatemala), In The Neighborhood as a Social and Spatial Unit in Mesoamerican Cities, M. Charlotte Arnauld, Linda R. Manzanilla, and Michael E. Smith (eds), pp. 181–201. University of Arizona Press, Tucson. Lemonnier, Eva 2012b La Joyanca, Petén Noroeste, Guatemala: Un Caso de Segmentación Interna y su Interpretación Política. In El Poder Compartido. Ensayos Sobre la Arqueología de Organizaciones Políticas Segmentarias y Oligárquicas, Annick Daneels and Gerardo Gutiérrez Mendoza (eds), pp. 285–312. CIESAS, Mexico City. Lemonnier, Eva, and Charlotte Arnauld 2008 Sistemas Clásicos de Asentamientos Jerarquizados en La Joyanca, Guatemala, y Río Bec, México. In XXI Simposio de Investigaciones Arqueológicas en Guatemala 2007, J.P. Laporte, B. Arroyo, and H. Mejía (eds), pp. 227–234. Museo Nacional de Arqueología y Etnología, Ministerio de Cultura y Deportes, Instituto de Antropología e Historia, Asociación Tikal y Fundación Arqueológica del Nuevo Mundo, Guatemala. Lemonnier, Eva, and Boris Vannière 2013 Agrarian Features, Farmsteads and Homesteads in the Río Bec Nuclear Zone.

Ancient Mesoamerica 24: 397–413. Leyden B.W., M. Brenner, T. Whitmore, J.H. Curtis, D.R. Piperno, and B.H. Dahlin 1996 A Record of Long- and Short-Term Climatic Variation from Northwest Yucatán: Cenote San José Chulchacá. In The Managed Mosaic: Ancient Maya Agriculture and Resource Use, S.L. Fedick (ed.), pp. 30–50. University of Utah Press, Salt Lake City. Liendo Stuardo, Rodrigo 2002 The Organization of Agricultural Production at a Classic Maya Center: Settlement Patterns in the Palenque Region, Chiapas, Mexico. University of Pittsburgh Press, Pittsburgh, PA. McAnany, Patricia A. (ed.) 1995 Living with the Ancestors: Kinship and Kingship in Ancient Maya Society. University of Texas Press, Austin. Martínez Hidalgo, G., R. Hansen, J. Jacob, and W. Howell 1999 Nuevas Evidencias de los Sistemas de Cultivo del Preclásico en la Cuenca El Mirador. In XII Simposio de Investigaciones Arqueológicas en Guatemala 1998, J.P. Laporte, H.L. Escobedo, and A.C. de Suasnávar (eds), pp. 327–335. Museo Nacional de Arqueología y Etnología, Ministerio de Cultura y Deportes, Instituto de Antropología e Historia, Asociación Tikal, Guatemala. Métailié, J.-P., J.-M. Carozza, D. Galop, and M.C. Arnauld 2002a Lagos, Bajos y Paleo-Paisajes en el Petén Noroccidental: El Inicio de una Investigación Geográfica y Arqueológica (La Joyanca). In Espacios Mayas: Usos, Representaciones, Creencias, A. Breton, A. Monod-Becquelin, and M. Humberto Ruz (eds), pp. 23–48. UNAM, CEMCA, Mexico. Métailié, J.-P., D. Galop, J.-M. Carozza, G. Selleron, R. Effantin, B. Vannière, G. Bossuet, and M. Véliz Peréz 2003 Dynamiques et Ruptures dans l’Histoire des Basses Teres Mayas. Questions et Méthodes à Partir du Programme “La Joyanca” (Nord-Ouest du Petén, Guatemala). In Des Milieux et des Hommes: Fragments d’Histoire Croisées, T. Muxart, F.D. Vivien, B. Villalba, and J. Burnouf (eds), pp. 139–149. Elsevier, Paris. Métailié, J.-P., G. Selleron, R. Effantin, J.-M. Carozza, and D. Galop 2002b La Fin de la Forêt au Petén? Continuités et Ruptures au Cours des Deux Derniers Millénaires dans les Basses Terres Mayas (Nord-Ouest du Petén, Guatemala). In Équilibres et Ruptures dans les Écosystèmes durant les 20 Derniers Millénaires en Europe de l’Ouest, Septembre 2000, pp. 473–488. Actes du Colloque International de Besançon, Presse Universitaire de Franche-Comté (PUF-C), Besançon. Michelet, Dominique, M.C. Arnauld, S. Dzul, M. Forné, C. Gillot, E. González, A. Lacadena García-Gallo, E. Lemonnier, P. Nondédéo, J. Patrois, G. Pereira, E. Taladoire, and B. Vannière 2008 Río Bec (Basses Terres Mayas): 2002–2008. Nouvelles de l’Archéologie 111–112: 109–115. Netting, Robert McC. 1977 Maya Subsistence: Mythologies, Analogies and Possibilities. In The Origins of Maya Civilization, R.E.W. Adams (ed.), pp. 299–333. University of New Mexico Press, Albuquerque. Nondédéo, Philippe, M. Charlotte Arnauld, and Dominique Michelet 2013 Río Bec Settlement Patterns and Local Socio-Political Organization. Ancient Mesoamerica 24: 373–396. Peña C. Agustín 1987 Hormiguero y Dos Aguadas: análisis comparativo de dos sitios en el sur de Campeche, Tésis profesional, ENAH, Mexico City. Peterson, Christian E., and Robert D. Drennan 2005 Communities, Settlements, Sites, and Surveys: Regional-Scale Analysis of Prehistoric Human Interaction. American Antiquity 70(1): 5–30. Pollock, Harry E.D. 1967 Brainerd y Ruppert en Xpuhil en 1949. Estudios de Cultura Maya 6: 67–80. Pope, Kevin O., Mary D. Pohl, and John S. Jacob 1996 Formation of Ancient Maya Wetland Fields: Natural and Anthropogenic Processes. In The Managed Mosaic: Ancient Maya Agriculture and Resource Use, S.L. Fedick (ed.), pp. 165–176. University of Utah Press, Salt Lake City. Robin, Cynthia 2002 Outside of Houses: The Practices of Everyday Life at Chan Nóohol, Belize. Journal of Social Archaeology 2: 245–268. Santley, R.S., T.W. Killion, and M.T. Lycett 1986 On the Maya Collapse. Journal of Anthropological Research 42(2): 123– 159. Smyth, Michael P., Christopher D. Dore, and Nicholas P. Dunning 1995 Interpreting Prehistoric Settlement Patterns: Lessons from the Maya Center of Sayil, Yucatán. Journal of Field Archaeology 22(3): 321–347. Thomas, Prentice M. Jr. 1981 Prehistoric Maya Settlement Patterns at Becan, Campeche, Mexico. Middle American Research Institute, E.W. Andrews and J.S.H. Brown (eds), Tulane University, Publication 45, New Orleans. Tourtellot, G., J.A. Sabloff, M.P. Smyth, L.V. Whitley, S.T. Walling, T. Gallareta, C. Perez A., G.F. Andrews, and N.P. Dunning 1988 Mapping Community Patterns at Sayil, Yucatan, Mexico: The 1985 Season. Journal of New World Archaeology 7(2/3): 1–24. Turner II, B.L. 1974 Prehistoric Intensive Agriculture in the Maya Lowlands. Science 185: 118–124. Turner II, B.L. 1983 Once beneath the Forest: Prehistoric Terracing in the Río Bec Region of the Maya Lowlands. Westview Press, Boulder, CO. Vannière, Boris, Christine Heinz, Georges Marchand, and Didier Galop 2005 Operación VIII: Uso del Medio Ambiente y Dinámicas Paleo-Ecológicas. In 2005: Proyecto Río Bec (Campeche México). Informe de la Tercera Temporada, del 16 de Febrero al 15 de Mayo de 2004, D. Michelet et al. CNRS, Laboratoire “Archéologie des Amériques” (UMR 8096), Centre d’Études Mexicaines et Centraméricaines (CEMCA), Mexico. Vannière, Boris, Eva Lemonnier, Georges Marchand, Christine Heinz, L. Saligny, and Agnès Stock 2006 Operación VIII:

Antropización y Explotación del Medio Ambiente en Río Bec. In 2006: Proyecto Río Bec (Campeche México). Informe de la Cuarta Temporada, del 8 de Febrero al 6 de Mayo de 2005, D. Michelet et al. CNRS, Laboratoire “Archéologie des Amériques” (UMR 8096), Centre d’Études Mexicaines et Centraméricaines (CEMCA), Mexico. Vapnarsky, Valentina, and Olivier Le Guen 2006 The Guardians of Space and History. Understanding Ecological and Historical Relations of the Contemporary Yucatec Maya to Their Landscape. European Mayan Conférence, Wayeb, Malmö. Webb, Elizabeth A., Henry P. Schwarcz, and Paul F. Healy 2004 Detection of Ancient Maize in Lowland Maya Soils Using Stable Carbon Isotopes: Evidence from Caracol, Belize. Journal of Archaeological Science 31: 1039–1052. Webster, David L. 2002 The Fall of the Ancient Maya: Solving the Mystery of the Maya Collapse. Thames and Hudson, London. Webster, David L., and AnnCorinne Freter 1990 Settlement History and the Classic Collapse at Copan: A Redefined Chronological Perspective. LAAnt 1: 66–85. Webster, David L., AnnCorinne Freter, and Nancy Gonlin 2000 Copán: The Rise and Fall of an Ancient Maya Kingdom. Jeffrey Quilter, Wadsworth-Thompson Learning, Bellmont, CA. Wilk, Richard R. 1984 Households in Process: Agriculture and Domestic Organization among the Kekchi Maya. In Households: Comparative and Historical Studies of the Domestic Group, R. McC. Netting, R.R. Wilk, and R.J. Arnould (eds), pp. 117–244. University of California Press, Berkeley. Wilk, Richard R. 1985 Dry Season Agriculture among the Kekchi Maya and Its Implications for Prehistory. In Prehistoric Lowland Maya Environment and Subsistence Economy, M. Pohl (ed.), pp. 47–57. Papers of the Peabody Museum of Archaeology and Ethnology Harvard University, Vol. 17, Cambridge, MA. Wilk, Richard R. 1988 Maya Household Organization: Evidence and Analogies. In Household and Community in the Mesoamerican Past, R.R. Wilk and W. Ashmore (eds), pp. 135–149. University of New Mexico Press, Albuquerque. Wilk, Richard R. (ed.) 1991 Household Ecology, Economic Change and Domestic Life among the Kechi Maya in Belize. University of Arizona Press, Tucson. Wilk, Richard R., and Robert McC. Netting 1984 Households: Changing Forms and Functions. In Households: Comparative and Historical Studies of the Domestic Group, R. McC. Netting, R.R. Wilk, and R.J. Arnould (eds), pp. 1–28. University of California Press, Berkeley.

8 Identifying settlement variability in the IsthmoColombian Area Alternative models from the Upper General Valley of the Diquís archaeological subregion Roberto A. Herrera

Introduction Encompassing the southeastern portion of Honduras and extending down into northwestern Colombia and Venezuela (Figure 8.1), the Isthmo-Colombian Area (ICA) represents a complex biological and cultural landscape from which archaeologists are forming new hypotheses for interpreting social variability. Current research within one sector of the ICA—the Greater Chiriquí archaeological region (Figure 8.2)—is driving a reconsideration of earlier assumptions regarding site function and its implications for interpreting long-term regional settlement ecology and understanding local vs. macro-regional social and ecological processes. This chapter presents new information from El Cholo, which is located in the northwestern end of the Upper General Valley in the Diquís archaeological subregion of Southern Costa Rica. El Cholo is a mound complex with an occupational history spanning the second to eleventh centuries AD, conforming to late Aguas Buenas (300 BC–AD 800) to Early Chiriquí (AD 800–1550) time periods (Table 8.1).

Figure 8.1

The Isthmo-Colombian Area.

Figure 8.2

The Greater Chiriquí archaeological region with El Cholo in the northwest, contemporaneous sites, modern towns, and geographic features.

This new data provides a useful point of departure for re-evaluating pre-existing ideas about settlement ecology concerning the Aguas Buenas period, with implications for regional variability and how we view the evolution of society in the region. The chapter proposes that Aguas Buenas social systems could be seen as part of a subregional set of initial conditions for later-period hierarchical social groups in the region, providing a more heterogeneous settlement model for Greater Chiriquí, which can offer new insights into the scale of social interaction and its articulation with social stratification and settlement processes.

The use of centralization models in the ICA Historically, research in the area has been dominated by the investigation of mound complexes, spurred on by work on cemeteries throughout Panama and Costa Rica (e.g. Hartman 1901, 1907; Haberland 1957, 1984; Lothrop 1919, 1934; Mason 1942; Stirling 1949). Inferences resulting from these investigations established a pattern that was to have a powerful influence on later studies, laying down a settlement framework that cast a wide shadow across much of Greater Chiriquí and Central Panama. The resulting interpretive schema framed sites within an elite or emergent-elite residential/funerary functional system (Haberland 1984). These inferences were largely derived from sites from the Gran Coclé area in Central Panama and Diquís sites in Southern Costa Rica where significant quantities of interred gold and ceramic offerings were either directly observed or indirectly inferred from secondhand accounts (e.g.

Lothrop 1919, 1934; Merritt 1860). Coupled with lavish burials in cases such as Sitio Conte, this strongly suggested an elevated social status for select individuals. More recently, similar patterns were noted for the nearby sites of El Caño and He-4 (Menzies and Haller 2012), building the case for an integrated mound-centered settlement system within the region: the basic form of hierarchical settlement system that many researchers use as a baseline interpretive model. Table 8.1

Subregional chronology of Greater Chiriquí

This model was further buoyed by discoveries, which in select cases yielded similarities in structural morphology of mortuary complexes, evidence of gold, high-cost ceramics and/or lithic goods such as stone celts or statuary (e.g. Drolet 1984, 1992; Haberland 1984; Lothrop 1963; Laurencich de Minelli and Minelli 1966, 1973; Stone 1943). Supported in part by ethnohistoric accounts, researchers in Greater Chiriquí and its subregions extended the Central Panamanian socio-political system into Costa Rica as well. Elite behavior, emergent or otherwise, evolved to become the de-facto behavioral template for the area, with some scholars arguing for connections and emulation of Mesoamerican social patterns (e.g. Stone 1977) and others proposing a homogenous Isthmian settlement and social pattern drawn from Panamanian models (e.g. Drolet 1984; Haberland 1984; Helms 1979, 1994). Whether oriented to the north or south, an interpretive milieu was put in place where polities ostensibly were seen as competing for territorial control, leveraging their power through the managed exchange of costly goods, as well as labor-intensive residences and ostentatious burials (e.g. Drolet 1992; Lothrop 1963; Laurencich de Minelli and Minelli 1966). While much of the reasoning for the above social and settlement pattern was derived from late prehistoric to pre-Contact period information, it was also suggested that centralized settlement organization extended back into the late Formative period (Briggs 1987; Cooke 2005; Drennan 1991; Ichon 1980; Linares 1977). Support for this idea came largely from work at the site of Barriles (Ichon 1968, 1980; Stirling 1949). Located in the Panamanian sector of the Greater Chiriquí archaeological region, Barriles revealed evidence of a large mound

center ostensibly dating to 300 BC to AD 600 and attributed to hierarchical social processes. More recent research supports the mound site’s central role in the social dynamics of the region (Palumbo 2009), which, along with the presence of ancillary sites such as La Pitahaya and Pitti-Gonzalez, suggests that Barriles was a locus of socio-ritual activity. Nonetheless, whether social dynamics at this relatively early stage of Chiriquí prehistory indicates a hierarchically based nucleated settlement pattern remains to be seen, with scholars acknowledging that social configurations in the area were likely more variable than previously thought (Palumbo 2009: 287). Argued to be a significant influence throughout Greater Chiriquí, Barriles’ potential reach throughout Panama and Southern Costa Rica suggests that it functioned as a Formative period explanatory foundation, implicitly promoting a homogenous centralized settlement pattern extending northwest into the Upper General Valley (e.g. Drolet 1992; Sol Castillo 2013). Diquís subregion mound complexes—ostensibly mimicking Barriles—were effectively interpreted as primate centers, which, it is argued, held sway over districts (Sol Castillo 2013: 44-47) and maintained the special status of their inhabitants through network connections facilitated by the distribution of high-cost portable goods such as stone axes and exotic ceramics. Thus, aspiring elites in the Formative period and later in the Chiriquí period derived their power through connections to macro-regional if not long-distance sources of political prestige (Haller 2004; Helms 1994): an important avenue for enabling aggrandizing behavior and driving social inequality. However, further investigation into specific site function revealed that there is reason to believe that not all sites fit this model and that the reasons underlying the source of site nucleation vary. Even if more modest in scale than that proposed by the Helms model (Helms 1979) of longdistance exchange of prestige goods, I suggest that Greater Chiriquí networks were likely subject to cultural configurations centered on local responses to the surrounding environment. The actual evidence for variability in internal site organization seems to reflect this, demonstrating highly localized preferences in construction and layout that attest to diachronic and synchronic interregional variability (Corrales and Badilla 2011, 2015; Corrales 1988; Drolet 1992; Frost 2009; Herrera 2015; Palumbo 2009, 2013; Quilter 2004; Sol Castillo 2013). Evident even in later Chiriquí period sites, a time when evidence is stronger for centralized distribution and network processes, this seems more apparent during earlier Formative time periods. In the Panamanian Chiriquí, sites with unambiguous indications of higher status dating back to the Formative period are rare. While there exists evidence of statuary depicting purported “master/slave” dynamics (e.g. Hoopes 1996) and high levels of labor investment, sites such as Barriles lack conclusive indications of hierarchy or elevated social status comparable to Central Panamanian counterparts (Palumbo 2009: 201). Moreover, excluding early components found in Central Panama at Sitio Conte that date to around AD 700, most Panamanian sites documented as having ostentatious burials date to later time periods (e.g. Cooke et al. 2000; Menzies and Haller 2012). Southern Costa Rica shows a similar pattern, as mortuary complexes such as Panteón de La Reina and its associated ceremonial settlement of Rivas, while multi-component, largely date

to the twelfth to fourteenth centuries AD (Frost 2009; Quilter 2004). Other locations such as Sitio Zoncho show a similar multi-component nature with a high likelihood that components were intermixed or conflated (e.g. Gómez and Soto 2001; Laurencich de Minelli and Minelli 1973). Although ethnohistoric sources do tell us of the centralized wealth and status of the then well-established chiefdoms in the region (Guardia 1913), it is noted for the early sixteenth century, 200 to 400 years later than the beginning of the Classic Chiriquí period. Thus, evidence for status differentiation and the power-based centripetal effect of settlement centralization, as expressed in elite monumental construction and mortuary offerings, actually only dates to the Chiriquí—and even then, is only unequivocally identified in accounts of Contact period chiefdoms. Earlier sites—such as El Cholo, Bolas, as well as Sitio Cantarero in the Osa Peninsula (Corrales 2015)—albeit complex mound systems, lack definitive evidence of elite or high-status occupation. Interpretive features key to extending a centralized settlement pattern into the earlier time periods remain effectively ambiguous, begging for further investigation into their actual function. In Southern Costa Rica, to say nothing of Greater Chiriquí, sites with clear indications of status differentiation are really restricted to the latter part of Greater Chiriquí prehistory, and even in this time period are tentative. And even if we accept the likelihood of some form of inherent site hierarchization, Chiriquí sites with evidence of monumentality do not readily show differences in status or ranking (e.g. Corrales and Badilla 2011). Elements such as the famous Stone Spheres of the Diquís subregion, seen in some cases as evidence of privileged access, instead may have been markers of group membership and ritual linked to astronomical observations (Corrales and Badilla 2007). While goods such as gold, polychrome, and thinwalled biscuit-ware ceramics arguably were produced and distributed, and used in emerging elite households during the Chiriquí period, some have argued that these items were seen less as representing status and more as related to propagation of a shared belief system. Objects reflexively seen by some as high status in this case are interpretable as more widespread in their usage, adapted to autochthonous social processes that did not directly emulate polities to the north or the south (e.g. Briggs 1993; Hoopes 1991). This less formal distribution arguably existed in a less restricted manner in the periods preceding Chiriquí. This variability is challenging from an interpretive standpoint, when surface indications, specialization, and labor investment in monumental structures are used to infer settlement and social hierarchization. Recent surveys have attempted to address these issues (e.g. Maloof et al. 2011; Sol Castillo 2013). However, significant gaps in our understanding remain, suggesting there may be reason to modify the existing interpretive model and generate other avenues of inquiry. The following section describes in more detail the rationale behind the interpretation of site formation and function as utilized in the specific region of the Upper General and the measures taken by this study to address the larger issue of macro-regional Chiriquí settlement ecology.

Analyzing settlement at site scale: El Cholo as a case study Surveys conducted throughout the Diquís in the 1980s documented multiple mound/platform structures scattered throughout the area of the greater Térraba basin, the name for the entire

General River/Térraba River system (Drolet 1983; Drolet and Markens 1981). Investigators noted a transition from upland terrace to alluvial zone settlement with augmented natural features and platform mounds—often faced with river cobbles—dating to the Aguas Buenas period, typically located on abandoned tertiary alluvial terraces and Chiriquí period constructions found on or near active floodplains (e.g. Drolet 1983, 1984, 1992; Quilter and Blanco-Vargas 1995, Maloof et al. 2011; Maloof 2012). Non-structural ‘hamlet’ sites, ostensibly dating to Aguas Buenas were also recorded in the area surrounding the mounds, identified by their refuse concentrations (Drolet 1984). The presence of non-structural sites, as well as the mounds, formed the basis for a two-tiered settlement hierarchy interpreted as smaller villages clustered around larger nucleated village-like socio-ceremonial centers. It is unclear however, if these socio-ceremonial centers were completely residential, for public ceremonies, funerary, or some combination thereof. To complicate matters, recent survey and excavations in adjacent regions south of the Upper General suggest that some smaller sites did show elements of elaboration (Maloof et al. 2011), but whether these represent mortuary structures, the remains of earlier dispersed farmsteads, or a combination of the two is unclear. Additionally, sites with architectural elaboration in the uplands of the Diquís Delta area—associated with Finca Camaronal in Buenos Aires—seem to date to the Chiriquí period, but their positioning on abandoned terraces follows a pattern usually attributed to earlier Aguas Buenas populations. While there is evidence for smaller populations during Aguas Buenas (Maloof 2012; Sol Castillo 2013), it is unclear whether the sites represented in this area are an indication of sedentary village behavior or the initial processes of funerary landscape demarcation. Nonetheless, the role of these sites, whether they are cemetery plots, hamlets, or “dispersed farmstead” sites, is a crucially under-investigated assumption, overshadowed by the more apparent signatures found at sites proposed to exist at the center of these occupations. While the surrounding community constituting the second supporting tier of settlement remained a vague assumption, the presence of lithic material such as stone axes and ceramic refuse at centrally located structural sites, and the assumed increased effort needed to construct mounds and produce said lithic items, strongly suggested to investigators that some form of elitemanaged craft specialization was in place. This centralized “special residence” pattern (Drolet 1992), arguably managed by emergent chiefs or priest chiefs, draws largely from a market dynamics model where central mounded areas function as distribution nodes for costly goods (i.e. Christaller 1966). Manufactured in large quantities at mound complexes, these goods would have then been used as the socioceremonial currency, to establish and maintain the authority of select persons who resided at these mound centers and participated in a broader, exclusive network social configuration (Drolet 1992; Haberland 1984); this centralized market behavior arguably would have extended into all sectors of the Upper General and Diquís region. If decisions in the Formative and Chiriquí period were indeed managed by emergent elites, the primary expectation would be that the central mounds, as examples of special residences, would likely produce evidence for occupation or interment indicative of unequal/elevated status. This would be an effective first step in supporting or falsifying the idea that a centralized settlement pattern extended throughout the Diquís, and up into the highlands, during

the mid- to late Formative period—while simultaneously evaluating the extant hypothesis as well as gauging the viability of alternatives.

Site-level data Excavations at seven operations within the El Cholo mound complex (Figure 8.3) revealed an extended occupation dating from at least the second century to the eleventh century AD and yielded multiple interments with modest offerings and evidence for ritual consumption of comestibles and destruction of ceramics, as well as production, curation, and ritual deposition of stone artifacts (Herrera 2015). Site-wide examples of ritually burnt and shattered ceramics alongside mortuary features, as well as deposition of waisted stone axes within grave structures, attested to a repeated pattern of funerary ritual. In several cases, interments located in levels dating back to the second and fourth centuries AD yielded extra-local ceramics from the Caribbean coast: evidence for inter-cordilleran contacts during Aguas Buenas times. Another sector of the complex yielded one polychrome sherd from the Guanacaste/Nicoya region of northern Costa Rica stylistically dating to the tenth and eleventh centuries AD, highlighting a pattern of interregional exchange with the north noted by previous scholars (Cooke and Ranere 1992; Hoopes 1996; Lange 1992). Several other examples of Caribbean and Atlantic Watershed style ceramics were also noted but not conclusively identified. These discoveries not only provided evidence of an already known fact that long-distance connections existed with the northwest of Costa Rica, but importantly demonstrated the likelihood that earlier connections existed with Caribbean Region/Atlantic Watershed groups just over the Talamancan cordillera. Lower levels at El Cholo associated with mortuary deposits also revealed lapidary made from blue-green sedimentary rock and highly weathered white sedimentary rock, as well as possible remains of resin ornaments. Evidence for the production of greenstone pendants and polished stone axes was also apparent in the form of broken pieces of greenstone and waste flakes found in non-grave contexts. The presence of such material may have been sufficient to suggest high status. However, minimal differentiation in internment and uniformity in ritual and associated grave goods at the site suggested that the overall distribution of artifacts and tombs did not differ significantly. Rather, the data suggested a generally stable ritual and social uniformity spanning at least two phases of ritual activity over the course of 800 years. Evidence thus suggests that the people interred at El Cholo did not appear to be of overtly high status, using often categorized “costly goods” in a generally egalitarian manner.

Figure 8.3

Plan and proposed layout of El Cholo structures with outlying circular structure in the southwest sector of the site.

Although there have been arguments that status may have been encoded in perishable items (e.g. Quilter 2004), El Cholo, while presumably costing a degree of labor investment to construct, did not seem extensively “rich” in nature, a reality that has often been overshadowed by the apparent richness attributed to finds in the region (Briggs 1993). And, while this apparent modesty in burial goods has been described by some as indicative of wealth variation within an existing network of mounds (Sol Castillo 2013), the fact remains that the general ascription of site hierarchy is untenable. Two-tiered site hierarchies, even by admission of

those attempting to fit sites into the general model, do not capture the variability in built environments. Labor investment does not seem to correlate with site function and funerary centers appear to inhabit a central place where villages were presumed to be located. With the discovery of seemingly contradictory findings, several crucial assertions regarding the centralized nature of Diquís mound sites are called into question. The claim that mound sites, such as El Cholo, functioned as domestic occupations is problematic owing to the lack of “domestic” features such as hearths and refuse middens. This fact opens up the possibility that if this one site—argued to be a primary village—is actually funerary in nature, then other sites may require a re-evaluation and may be more accurately conceived as multifunctional. The proposition that mounds were centers for specialization or potentially elite in nature does not sufficiently account for what potentially is a range of mound types, exemplified by El Cholo with its modest burial goods and relatively equitable distribution and configuration of space and material throughout the site. Indeed, prior to the introduction of an outlying architecturally distinct burial structure around the mid- to late seventh century (Figure 8.4) funerary behaviors at El Cholo are remarkably uniform, continuing to be so—architectural anomalies notwithstanding—up to at least the early eleventh century. This does not lend support to the assertion of elite occupation. It is noted that the later period does see an introduction of arguably newer ceramic complexes, dating to the transition to Early Chiriquí, but it is nonetheless deposited in the same manner as earlier phases, indicating a transition to a newer medium but an unchanged practice. Interestingly, the use of early/transitional Chiriquí material appeared to be the final behavioral sequence at El Cholo, as subsequent Chiriquí period activity moved away downslope from the site, closer to the active alluvial bottomlands of the General River (Figure 8.5). It would seem, then, that El Cholo’s role as a central place was mortuary rather than elite in nature, and that as yet unknown factors prompted a move away from the site at the beginning of the Chiriquí. These intriguing conditions suggest the need for a search for a more compatible explanatory model.

Figure 8.4

Select examples of commemorative mortuary ritual at (clockwise from upper left) operations A, D, G and D at lower levels showing intentional shattering of ceramics and concentrated thermal activity adjacent to features at El Cholo.

Figure 8.5

Later Chiriquí period mortuary features found 900m east of El Cholo.

Primate centers or community burial grounds? Implications for an alternative model of settlement ecology The advent of new information recorded at El Cholo underscores the challenges inherent in interpreting variable site function in the Diquís and Greater Chiriquí in general. How, given this new data and the seeming variability it implies, can we go about defining fundamental units for interpreting the dynamics of settlement ecology in the Aguas Buenas or Chiriquí periods? Sites slightly further south in the Diquís Delta further illustrate this conundrum. Some sites, while appearing to show use of space in a manner similar to that of El Cholo, nonetheless demonstrate what seems to be localized variation. In some Diquís Delta cases, a clear delimitation of space during the transitional period between Aguas Buenas and the Chiriquí period is apparent. In cases such as Finca 6 and Sitio Grijalba (Figure 8.6), rectangular mortuary structures were utilized as separate funerary zones associated with circular house foundations, with significant labor invested into construction of their mound bases and the colocation of stone spheres. Although to date no clear Aguas Buenas component has been recorded at either site (Corrales and Badilla 2011, 2015), their single-component nature suggests examples of breaks from previous Aguas Buenas settlement organization; this parallels observations of Aguas Buenas to Chiriquí transitions in the General Valley (Drolet 1992), where settlement ostensibly moves from the uplands to the alluvial plains. Other sites, such as those at Sitio Batambal and El Silencio, were found to be multi-component in nature, with Chiriquí period material overlying previous Aguas Buenas deposits. This suggests a continual use of space from earlier times—a behavior matching the observed cumulative depositional pattern noted at El Cholo, albeit in reverse, with the larger component at El Cholo

dating to Aguas Buenas and the opposite being true for the Diquís Delta sites. The similarity to and overlap with El Cholo in the depositional history of some sites in the Diquís Delta indicates a mixed settlement strategy that expressed a continuity and evolution of space from earlier Aguas Buenas ceremonial and mortuary practices, if not a direct continuation of meaning and function within said spaces and built environments. The preexisting occupation of some parts of the bottomland during Aguas Buenas times, while poorly known, suggests a shift in settlement strategy similar to that of the adjacent highlands. In the case of sites such as Batambal and El Silencio, earlier occupations, rather than being abandoned, continued to be used and modified; moreover, the nearby rolling hills were still used as domestic and funerary zones well into the Chiriquí period (Maloof 2012). This has been observed at sites such as Bolas and Rivas (Frost 2009; Palumbo et al. 2013; Quilter 2004). At the core of this variable transition of internal site and settlement organization are the changes in the built environment. For some sites in the Diquis, mortuary zones become increasingly separated and are found adjacent to distinct residences characterized by circular ramped mounds (Corrales and Badilla 2011). A similar process also seems to occur around the seventh century at El Cholo. However, the presumed functions of these two circular “houses” differed. For El Cholo the circular structure introduced in the seventh century seems to continue its function as a funerary complex rather than a residence (Herrera 2015: 340). Tombs and an abundance of funerary material were placed within the circular structure, which was incorporated into the existing complex rather than set apart from it. The structure at El Cholo, with its smaller dimensions and evidence of intensive funerary deposition, suggests a concentrated funerary function contrasting with the residential function of the similar circular structures at Finca 6 and Grijalba. The variable use of similar structures as residences or funerary houses, coupled with similar but possibly more socially integrative circular structures at sites such as Rivas in the Upper General (Quilter 2004), speaks to differences in function within similar structures and a variability in social function that arguably distinguished subregions to a significant degree.

Figure 8.6

Chiriquí period habitation sites demonstrating separation of structures (from Corrales and Badilla 2011: 29, 33).

Interestingly, for Finca 6’s rectangular funerary mound, while there is the presence of Tarragó biscuit and polychrome ceramics—used at times as proxies for elite consumption— funerary goods do not show overt levels of ostentation; rather, it may suggest the gradual culmination and separation of emergent elite behavior that is argued for the region (Sol Castillo 2013). That this process seems slow and variable attests to the gradual and heterogeneous nature of ICA social processes in general. In the Diquís subregion, this is underscored by sites such as Batambal (Corrales and Badilla 2011), which seems a closer fit

with the cumulative depositional processes observed at El Cholo. Although the potential for ostentatious goods and the presence of stone spheres may indicate emergent elites for the Chiriquí, extending an elite nucleated pattern to the Aguas Buenas or even the early Chiriquí is tenuous. Specialized items are not isolated enough to suggest overt monopolization, even when speaking of the Chiriquí period. How then can one assert that these sites are “chiefly centers”? To an even more equivocal degree, sites with earlier components—such as Bolas, Monge, and Las Brisas in the Upper General Valley—have also been tentatively presented as elite or chiefdom centers (Drolet 1992; Palumbo et al. 2013; Sol Castillo 2013). Again, this is largely based on construction/labor effort. It remains to be demonstrated whether construction efforts were the product of centrally mobilized endeavors, or an accretional construction process, as the presence of multiple platform floors may attest to repeated building and augmentation (Palumbo et al. 2013). This interpretive problem, along with gathering clear evidence for attached specialization of goods, is an issue currently being addressed by upcoming investigations at Bolas. Southern Costa Rican Aguas Buenas sites, while extensive in size in many cases, do not show clear demarcations of status expected of emergent or established elites, suggesting a more diverse environment. Throughout the area, indicators such as monumental mound construction, ceramic and lithic production—and, in the unique case of the Diquís Delta, stone spheres—have been used to espouse high-status expression, yet little support is available to corroborate the claim that these sites were more than aggregation or nucleation points for the surrounding population. That there was effort is apparent, and centralization seems evident, but the definition of what lies at the center of the pattern, as well as what that implies for the regional settlement ecology are ambiguous. As evidenced by the apparent variability in mound forms and function, and documented by the finds at El Cholo, the prevailing impression increasingly seems that mound structures differ significantly in their internal organization.

Community burials as social landmarks and horizontal social processes Research within the ICA, and beyond, points to the possibility of an alternative perspective, suggesting that we might instead view many of these structures not as instances of elite mobilized intensive construction, but as cumulative/accretional expressions of communityscale, cooperative behavior (Alarcón 2014; Crumley 1995; Giraldo 2010; Hoopes and Chennault 1994: Norr 1986; Sheets 2009). This seems to better fit the available data obtained at El Cholo and is arguably extendable—at the minimum—to sites within the upland areas of the Upper General and the Herradura region, and into the Talamanca. Within the communitybased model, mound structures retain their functional role as central places. However, rather than being elite residences, socio-ceremonial centers function as forums and socially negotiated topographic reference points: social landmarks that hold meaning over multiple generations (Bender 1990, 1993, 2002). Surrounding hamlets, dispersed among the Formative period landscape, would have used a site like El Cholo as a repeatedly visited commemorative location, returning at intervals to inter their deceased, something borne out by initial data from El Cholo and even sites such as Bolas. Evidence of repeated commensal behavior, noted at El Cholo by the evidence of

smashed Quebradas Tazones, rather than having been domestic in nature, would instead have been the material residue of funerary feasts (Bozzoli 1975; Snarskis 1984: 221), related to members within a cyclic power-sharing system not unlike ayllu-style segmental community systems (Bastien 1978). Residues from such horizontal systems would appear less as elite “potlatch” than as community “potluck” behaviors (e.g. Blinman et al. 1989). Unless it is an extreme outlier, it would seem that sites similar to El Cholo from the same time period would have functioned in a manner similar to the above proposition: with ancestral burial grounds serving as socially cohering agents rather than elite centers. Outlying date aside, the secondcentury date (Figure 8.7) for the founding of El Cholo suggests that this is the case for the earlier history of the site. Previous survey results are not mistaking the central function of sites, but rather are reifying one potential interpretation within a range of potential site functions which developed and persisted at a variable rate, depending on the subregion (Figure 8.8). The variety of the site types displayed in Southern Costa Rica, let alone Greater Chiriquí and the Isthmo-Colombian Area in general, suggests the existing dynamic settlement ecology was a variable process consisting of multiple strategies across a mixed landscape, which in turn influenced settlement decisions.

Figure 8.7

Radiometric dates for El Cholo.

Lower population densities, noted for most of the prehistory of the region (e.g. Haller 2004;

Hoopes 1996; Palumbo 2009), while not solely responsible for driving development, constituted a likely factor for this earlier period emphasis on commemorative mound centers. Extended mobility, in a time defined by root crop and mixed subsistence regimes (Corrales 1988; Hoopes 1991)—especially in the more montane parts of the region—arguably facilitated more flexible territorial ranges. While this may be less feasible in the lower alluvial regions of Greater Chiriquí (Piperno et al. 1989; Piperno 2011), it is quite tenable in the highland regions that afford a range of subsistence options. With regard to the creation of these centers and their variable transition to emergent elite centers, I propose that a variable persistence in heterarchical or rotating power strategies fell roughly along environmental zones which enabled flexibility. More mobile upland groups were likely better able to mitigate not only ecological risk but also aggrandizing behavior, a factor suggested for earlier time periods in Costa Rica, which arguably had a more egalitarian cast (Hoopes 1991; Norr 1986; Sheets and Sever 2006).

Figure 8.8

Map of study zones within the Diquis subregion (modified from Corrales 2000).

I submit that the early use of mound sites in the Upper General Valley, if not the Diquís and Greater Chiriquí, was likely related to this more archaic form of settlement, contributing to a heterogeneous settlement ecology in the Diquis, linked to differences in landscape niche utilization as inherited memorial spaces (Basso 1996; Bender 1993; Bernardini 2004). While potentially on an equal footing during the Tropical archaic periods in Costa Rica, relative to their topographical and temporal position in the highlands and down in the Delta/Alluvial regions, mounds likely varied in form and scale, signifying niche differences between groups/communities despite being in such close proximity. The Upper General Valley, with its noted distinctive Quebradas ceramic culture (Corrales 2000; Sol Castillo 2013; Drolet 1992), may be one of several examples indicating a more diverse landscape, one that embodied a

settlement strategy arising from the highlands rather than the plains of Chiriquí, an idea already proposed by earlier authors (i.e. Haberland 1984: 240; Snarskis 1984: 223).

Vertical settlement ecology in the subtropics: mounds as community topographic references Explanations for settlement within Greater Chiriquí generally begin with the introduction of stable horticultural/agricultural villages as the foundation for sedentary and hierarchical social systems, as opposed to less intensive, more flexible silvicultural or shifting agricultural strategies (e.g. Piperno 2011). However, the initial conditions of this regional settlement ecology across the landscape, although better known for parts of Panama, are not fully established for the Talamanca region—and, given limited data, investigations have been restricted to exploratory efforts by local investigators and a few outside Greater Chiriquí (e.g. Cooke 2005; Cooke and Ranere 1992; Corrales 2000; Hoopes 1996). Little is known about the historical and preceding settlement ecology that dominated the Formative period uplands of Southern Costa Rica as well as further south in the Diquís Delta Region. While work from as early as the 1980s suggested different ecological zones within the Diquís subregion of Greater Chiriquí (Figure 8.8), often the practice has been to describe the entire territory in isotropic terms (often out of necessity and lack of data). This chapter therefore proposes an alternative description—which reflects the heterogeneity of these zones and is more firmly based on local conditions of the Upper General—as a basis for revisiting past ideas regarding territoriality of the Diquís region, and building on and expanding upon ideas of semi-sedentism and vertically oriented mobility (Ibarra pers. comm.; Murra 1980, 1985). Based on initial interpretations, the periodically commemorative nature of ritual behavior encountered at El Cholo, in conjunction with the apparent lack of residential occupational refuse, supports earlier suggestions that groups may still have been practicing an earlier form of semi-sedentary subsistence well into Aguas Buenas times (Hoopes 1991; Piperno et al. 1989). The mobile nature of these groups would have benefited from the types of topographic markers and visible landscape references observed throughout the area, such as petroglyphs (Zilberg 1986) and hilltop cemeteries; these landscape features would have been used as topographic mnemonics (e.g. Bender 1993, 2002; Sheets 2009, 2011) to delineate territory, resource boundaries (Buikstra and Charles 1999), and lineal affiliations along a vertically oriented landscape. Moreover, it follows that labor investment would have been focused into areas that were central to the surrounding population; in this way, the center of a given resource range would have been occupied, rather than shifting habitations. The locations of “monumental” mound sites on upper terraces and ridge spurs suggest that placing them along prominent points on the landscape was important from a visual perspective; they were used in later iterations as status markers (Frost 2009), while in earlier times they would have been visual referents for shifting groups, marking the landscape with observable territorial identifiers. Repeated visitation of mortuary features would have gradually cemented these features into the inherited memory of the population, paving the way for later appropriation in

late Chiriquí times (Frost 2009; Quilter 2004; Sheets 2009, 2011). A cursory analysis of the distribution of sites in the uplands, especially as one proceeds further into the highlands, suggests a dynamic mobility (Kantner 1988). Mound sites recorded from initial surveys conducted in the upland Herradura region of the Rio Chirripo Pacifico/General watershed, in the paramo-like region of the western slopes of the Talamanca, suggest that this pattern of multiple niche exploitation was likely prevalent. Mound sites would have provided a ready network of movement and communication devices, through commemorative landmarks that would have facilitated relatively long-distance kin associations —a phenomenon still intact in the Talamancas. Ethnographic data shows that groups such as the Bribri have consistently maintained contact between Caribbean and Pacific Talamancan populations, suggesting this vertical orientation (Bozzoli 1975; Gabb 1875; Kantner 1988; Ibarra pers. comm.; Skinner 1920). Although no Atlantic or Caribbean material was identified at the Herradura sites, it was suggested that the upland sites were “outposts” for contacts to the northwest (Kantner 1988: 59). That trans-cordilleran contacts were in place during Aguas Buenas is supported by the presence of material found at El Cholo, which demonstrates midsecond-century or third-century links to the Caribbean region and Atlantic Watershed (Herrera 2015), as well as evidence of Aguas Buenas material in the Southern Caribbean Sixaola Valley (Corrales pers. comm. 2015). Additionally, the presence of sites along the ridgelines of the Talamanca and the likely presence of considerably more sites (Corrales pers. comm. 2010) in the interior of the Talamancas, within the current Parque de La Amistad, suggests that social configurations were axially montane-oriented. This allows for a provisional redefinition of boundaries to incorporate a cordilleran zone spanning the eastern and the western watersheds, away from the more explored lower areas (Figure 8.9). The verticality sub-hypothesis posits that social conditions in the Upper General reflected the utilization of distinct subregional ecological niches primarily oriented towards the uplands. Variable conditions noted in “monumental” construction of earlier Aguas Buenas sites, in contrast to similar groups further south in the relatively lower elevations, may thus have been restricted to the uplands. This has implications for the manner in which the Upper General communities interacted with the rest of the Diquís and Greater Chiriquí. Unlike what has been suggested for the lower and broader alluvial plains, where stable maize agriculture reduced the risk of less predictable seasonal resources (Piperno et al. 1989; Piperno 2011), vertical topography, with its micro-environments, would likely facilitate rotating subsistence practices of the smaller groups inhabiting the areas. Utilizing this environmental resource ladder, ranging from alluvial to paramo-like conditions (Coates 1999), smaller groups would potentially have been more sustainable over a longer period of time. Given the documented emergence of village life in other parts of the Diquís (Corrales 1988), this seemingly archaic settlement strategy appears to have lingered for far longer than expected. Along with evidence for iterative behavior, the presence of mound sites noted for the Herradura region, the relatively small number of unambiguous nucleated village sites, as well as evidence for dispersed hamlets or farmsteads (Sol Castillo 2013) for the Upper General, suggest that semi-mobile strategies within the piedmont and the upper reaches of the Talamanca were viable strategies existing concomitantly with more fixed agricultural strategies. Considering the potential time-depth and extent of movement throughout the uplands in the

General Valley, the Upper Talamancan region, and the Atlantic/Caribbean region, repeated visitation may have been an early occurrence in the highlands. Moreover, given the area’s relative isolation by means of topography, this pattern could easily have persisted whilst more sedentary villages such Curré (Corrales 1988) exploited more stable and intensive resources.

Figure 8.9

A rethinking of Greater Chiriquí zones.

Over the 1100-year temporal span that is Aguas Buenas, limits in the ability to farm corn, or simply the lack of any urgency to develop large-scale farming in lieu of silvicultural and shifting agricultural options, are arguably partly responsible for why groups did not settle down on the landscape in the standard manner observed for many societies in the Americas. Although lake core data is lacking in the Upper General, data from Anchukaitis and Horn’s assessment for farming and land clearing in the area only shows increased activity around AD 150 (Anchukaitis and Horn 2005). This late date does not even suggest intensification on the landscape. Therefore, it could be, as others have suggested (Hoopes 1991, 1996; Piperno et al. 1989), that the use of corn in this area was not staple but ritual: it was a supplementary crop amid a broad-spectrum subsistence strategy. Thus, the potential for semi-mobile societies during Aguas Buenas and into the Chiriquí period indicates, as others have noted (Hoopes 1991, 1996), that less entrenched social dynamics may have held sway. Settlement dynamics, as suggested above, would have reflected this reality. In this particularly regional case, it would seem that the exceptions rather than the rule were those sites that developed more intensive agricultural systems and subsequent social stratification. Sites such as Curré, Bolas, Finca 6, and Barriles may have all been poised for “growth” while others remained relatively stable in their territorial niches. Interestingly, the lack of any clear distinguishing features at any of these sites, save Finca 6, may suggest that this even persisted in the areas said to house inequality, with aggregation, once again, not a sufficient cause for entrenched hierarchy. While it is admittedly unclear if the proposed pattern existed in the uplands, it seems evident that in the Upper General, if not in the lower Diquís Delta (cf. Baudez 1993), there was a transition from upper terrace “commemorative” sites to aggregate Chiriquí “monumental” sites (Drolet 1983, 1992). The Chiriquí appears to reflect a gradual waning of older Aguas Buenas

social landmarks from the upper terraces of the surrounding terrain, with larger centers arguably concentrating in a new aggregated form in the Diquís bottomland. Similar processes seem to have occurred earlier in Barriles and Bolas, the two large, arguably nucleated sites that rose during Aguas Buenas. However, if the implications of the proposed dynamics for the Upper General highlands hold, the adaptation of populations may not have been so isometric in its distribution, nor nucleated in its full areal extent, especially if we consider topographic variability and movement through the increasingly steep highlands of the Upper General and beyond. It is in these subzones that vertically oriented movement and communication may have had a disproportionate influence on inhabitants in earlier periods. Couple the diversity and increase in topographic relief, the niche-like nature of the Talamanca and coastal ranges further into the main spine of Southern Costa Rica with the value of resource availability and its attendant lack of mobility restrictions, and a possible result is a territorial pattern that diverges from those of lower regions. This is not to deny the influence and contemporaneous interdependence of areas to the south of the Upper General and the highlands. The above only serves to point out the stronger likelihood that this area exercised a greater autonomy than expected, at least for the earlier Aguas Buenas period.

Conclusions Early period investigators, based on settlement configuration gleaned from surface collections, had suggested that site hierarchies likely indicated the beginnings of centralized control and possibly labor mobilization based around a village lifestyle (Corrales 2000; Linares et al. 1975). However, based on recent excavations, central places may instead have been territorial anchor points for semi-mobile groups. A similar point was made for the Eastern Woodlands of the United States (Buikstra and Charles 1999), suggesting a transition from archaic mobile residence patterns to a system of semi-sedentism. Moreover, semi-sedentism arguably persisted in parts of Costa Rica well into the fourth century AD (Murillo-Herrera 2003). A similar dynamic may have been in place in the Upper General, with the upper highlands characterizing a variable settlement strategy facilitated by a vertical style of resource and information acquisition. Ironically, rougher topography may have allowed for a less restricted landscape. Early patterns for the surrounding area show minimal labor investment in hamlets, with perishable materials leaving little trace and only refuse deposits marking their potential positions. The pattern of ephemeral settlement among the highland areas surrounding the General Valley and the placement of mortuary structures amid tertiary terraces suggest a potentially niched settlement orientation of flexible social networks, which were honeycombed in the microenvironments of a broken, “jackknifed” topography (Kantner 1988) and maintained by cemeteries as mechanisms of social cohesion (cf. Sheets 2011: 433–434). The implications of this model for settlement and social organization are varied. This less restricted settlement landscape may correlate with social mobility and greater agency on the part of smaller households. With formative cemeteries not yet co-opted by aspiring elites, hierarchical tendencies and the argued relevant indicators appear minimal. Movement begat social mobility in this model. While the means for ideological control is present in the control

of space and place, it may have taken considerably more time for it to take root in the region as compared with other areas. Therefore, while there is some evidence for hierarchical social environments in the later history of the ICA, as we go back in time and scan the region at finer resolutions, we confront the limitations of homogenously applying a macro-regional settlement model. It is understandable to project the model on to archaeological signatures throughout the ICA, as it is a logically appealing hypothesis, but the use of the model without critical review shoehorns the Diquís into an interpretive system that is likely incomplete. Rather, following along the lines of Hoopes and Fonseca (2003), Greater Chiriquí truly seems as if it functioned as a very “diffuse unity,” with variable implementation of ideological/religious complexes emerging and manifesting at variable rates. Given the flexible reception of ideological complexes and the distribution of mound sites in a potentially more autonomous fashion, we might then see Aguas Buenas as offering the potential conditions for the hierarchical environments that were observable at European contact. Certainly, evidence of labor-intensive constructions such as mounds, and the existence of ceramics and statuary at certain sites in Panama and Costa Rica, suggest a burgeoning inequality; and the rituals expressed in the statuary of Barriles could attest to the development of ascribed inequality from apprentice/assistant and specialist dynamics. Nevertheless, however logical the idea may seem, if patterns are taken at face value, without the aid of finergrained investigation, there can be a tendency for inflexible thinking: viewing the social fabric of the ICA as a static phenomenon and not as the unfolding and variable process it is. It seems clear that the settlement dynamics in many places in the ICA transitioned from more egalitarian to more hierarchical, but they seem now to have been more differentiated across the landscape. To apply one model with assumptions of social domination and hierarchical settlement to both the Aguas Buenas and Chiriquí periods discounts alternatives, which could include heterarchical/horizontal social dynamics: a complementary hypothesis that may gain traction as more information is recovered. While more sites need to be evaluated under this broader rubric, the data obtained from El Cholo suggests that status, in at least the earlier periods of Upper General prehistory, was less defined than is presented by current general models. Moreover, the presence of chronologically overlapping patterns of use of space in the Upper General and the Diquís, and the amorphous boundary of domestic and ritual space as it pertains chiefly to Chiriquí phase settlements, make it difficult to clearly identify evidence for centralized decision-making power or hierarchical status such as that seen in Panama. Notions of ostentation or monumentality and their implications for hierarchy may be incorrectly attributed to earlier components. Thus it remains a challenge to precisely determine when and where attempts at establishing hierarchy actually occurred during Aguas Buenas times. El Cholo is but one of a handful of sites with solid radiometric dates. Distinguishing between the two time periods is key and ongoing, but new research throughout Costa Rica is suggesting that this nebulous transition is not necessarily a methodological problem, but rather an indication of the variable range of behaviors occurring at the time. In the end this is a re-evaluation and modification of the previous settlement model, which came about because of the nagging questions that were generated from new data regarding the

level of social complexity during the Formative. Was social participation closed off and restricted to only a select few, as argued for other parts of Southern Costa Rica and Panama? Was it more open and “multi-vocal”? Given the findings at El Cholo, what are the various motivations behind mound construction, and what did the mounds truly represent? So-called elite items were not highly represented or significantly restricted at El Cholo, varying by grave in an insignificant way. Was El Cholo then a “poor” mound complex as some have suggested (Sol Castillo 2013: 132)? How many “poor” mound complexes were there during Aguas Buenas times, relative to the rich ones? The work at El Cholo suggests that these issues need to be more systematically acknowledged when comparing contemporaneous and later period sites. When the variability seen throughout the ICA is taken into consideration, rates of change and geographic distribution begin to reveal a variable settlement and historical ecology. Similar to the environment of Costa Rica and the ICA itself, which experiences microenvironmental variability, the data from El Cholo suggests that it may be one of many sites that exhibited local “micro-regional” variation within a larger macro-regional context. As more work is carried out, we can further test this very preliminary model.

References Alarcón, Gerardo M. 2014 Estudio de los Límites Espaciales y Temporalis del Sitio Arqueológico Guayabo de Turrialba (C-362 MNG). Segunda Fase. Informe Temporada 2013. University of Costa Rica, San José, Costa Rica. Anchukaitis, Kevin J., and Sally P. Horn 2005 A 2000-Year Reconstruction of Forest Disturbance from Southern Pacific Costa Rica. Palaeogeography, Palaeoclimatology, Palaeoecology 221(1/2): 35–54. Basso, K.H. 1996 Wisdom Sits in Places: Landscape and Language Among the Western Apache. University of New Mexico Press, Albuquerque. Bastien, Joseph William 1978 Mountain of the Condor: Metaphor and Ritual in an Andean Ayllu. Monograph—The American Ethnological Society, 64; Monographs of the American Ethnological Society; 64. West Pub. Co., St. Paul. Baudez, Claude F. 1993 Investigaciones Arqueológicas en el Delta del Diquís. 1. ed. Centro de Estudios Mexicanos y Centroméricanos (Mexico) y la Delegación Regional de Cooperación Científica y Técnica en América Central, México, DF. San José, Costa Rica. Bender, Barbara 1990 Dynamics of Nonhierarchical Societies. In Evolution of Political Systems: Sociopolitics in SmallScale Sedentary Societies, edited by Steadman Upham, pp. 247–263. School of American Research advanced seminar series. Cambridge University Press, Cambridge, UK. Bender, Barbara (editor) 1993 Landscape: Politics and Perspectives. Bloomsbury Academic, Berg, Providence, RI and Oxford. Bender, Barbara 2002 Time and Landscape. Current Anthropology (S43): S103–S112. Bernardini, Wesley 2004 Hopewell Geometric Earthworks: A Case Study in the Referential and Experiential Meaning of Monuments. Journal of Anthropological Archaeology 23(3): 331–356. Blinman, E., W.D. Lipe, and M. Hegmon 1989 Potluck in the Protokiva: Ceramics and Ceremonialism in Pueblo I Villages. In The Architecture of Social Integration in Prehistoric Pueblos, edited by William D. Lipe and Michelle Hegmon, pp. 113– 124. Crow Canyon Archaeological Center, Cortez, CO. Bozzoli de Willie, María Eugenia. 1975 Birth and Death in the Belief System of the Bribri Indians of Costa Rica. Ph.D. dissertation. Department of Anthropology, University of Georgia, Athens. Briggs, Peter S. 1987 Pre-Conquest Mortuary Arts and Status in the Central Region of Panama. Ph.D. dissertation. Department of Art History, University of New Mexico, Albuquerque. Briggs, Peter S. 1993 Fatal Attractions: Interpretation of Prehistoric Mortuary Remains from Lower Central America. In Reinterpreting the Prehistory of Central America, edited by Mark. M. Graham, pp. 141–168. University Press of Colorado, Niwot. Buikstra, Jane, and Douglas K. Charles 1999 Centering the Ancestors: Cemeteries, Mounds, and Sacred Landscapes of the Ancient North American Midcontinent. In Archaeologies of Landscape: Contemporary Perspectives, edited by Wendy Ashmore and A. Bernard Knapp, pp. 201–228. Blackwell Publishers, Malden, MA. Christaller, Walter 1966 Central Places in Southern Germany, edited by C.W. Baskin. Prentice-Hall, Englewood Cliffs, NJ.

Coates, Anthony G. 1999 Central America: A Natural and Cultural History. Yale University Press, New Haven. Cooke, Richard G. 2005 Prehistory of Native Americans on the Central American Land Bridge: Colonization, Dispersal, and Divergence. Journal of Archaeological Research 13(2): 129–187. Cooke, Richard G., and Anthony J. Ranere 1992 The Origin of Wealth and Hierarchy in the Central Region of Panama (12,000–2000 bp), with Observations on its Relevance to the History and Phylogeny of Chibchan-Speaking Peoples in Panama and Elsewhere. In Wealth and Hierarchy in the Intermediate Area: A Symposium at Dumbarton Oaks, 10th and 11th October 1987, edited by Fred Lange, pp. 243–316. Dumbarton Oaks Research Library and Collection, Washington, DC. Cooke, Richard G., Luis A.H. Sánchez, and Koichi Udagawa 2000 An Update Based on Recent Excavations and New Radiocarbon Dates for Associated Pottery Styles. In Precolumbian Gold: Technology, Style and Iconography, edited by Colin McEwan, pp. 154–176. British Museum, London. Corrales, Francisco 1988 Quebradas, Valle del General. Evaluación Arqueológica Inicial. Vínculos 14(1/2): 91–103. Corrales, Francisco 2000 An Evaluation of Long Term Cultural Change in Southern Central America: The Ceramic Record of the Diquís Archaeological Subregion, Southern Costa Rica. Unpublished Ph.D. dissertation. Department of Anthropology, University of Kansas, Lawrence. Corrales, Francisco, and Adrián Badilla 2007 Informe de Investigación Arqueología No. 079-05 Proyecto Investigaciones Arqueológicas en Sitios con Esferas de Piedra, Delta del Diquís Excavaciones en el Sitio P-254 Sector Finca 6 Temporada 2005. Museo Nacional de Costa Rica, San José. Corrales, Francisco, and Adrián Badilla 2011 Investigaciones Arqueológicas en Sitios con Esferas de Piedra, Delta de Diquis Excavaciones en los Sitios Finca 6 (P-254-F6) y Batambal (P-299-Bt) Temporada 2010. Museo Nacional de Costa Rica, San José. Corrales, Francisco, and Adrián Badilla 2015 Asentamientos Precolombinos con Esferas de Piedra en el Delta del Diquis, sureste de Costa Rica. Excavaciones en los Sitios Finca 6 (P-254-F6) y Batambal (P-299-Bt). Vínculos 35: 19–66. Crumley, Carole L. 1995 Heterarchy and the Analysis of Complex Societies. Archeological Papers of the American Anthropological Association 6(1): 1–5. Drennan, R.D. 1991 Pre-Hispanic Chiefdom Trajectories in Mesoamerica, Central America, and Northern South America. In Chiefdoms: Power, Economy, and Ideology, edited by Timothy Earle pp. 263–287. Cambridge University Press, Cambridge, UK. Drolet, Robert P. 1983 Al Otro Lado de Chiriquí, El Díquis: Nuevos Datos para la Integración Cultural de la Región Gran Chiriquí. Vínculos 9(1): 25–76. Drolet, Robert P. 1984 Community Life in a Late Phase Agricultural Village, Southeastern Costa Rica. BAR International Series (212): 123–152. Drolet, Robert P. 1992 The House and the Territory: The Organizational Structure for Chiefdom Art in the Diquis Subregion of Greater Chiriquí. In Wealth and Hierarchy in the Intermediate Area: A Symposium at Dumbarton Oaks, 10th and 11th October 1987, edited by Fred Lange, pp. 207–242. Dumbarton Oaks Research Library and Collection, Washington, DC. Drolet, R., and Robert Markens 1981 Informe Final, Rescate Arqueológico. Proyecto Boruca. Manuscrito en archivo. Departamento de Antropología e Historia, Museo Nacional de Costa Rica, San Jose. Frost, R. Jeffrey 2009 The Ancestors Above, the People Below: Cemeteries, Landscape and Dual Organization in Late PreColumbian Costa Rica. Ph.D. Dissertation. Department of Anthropology, University of Wisconsin, Madison. Gabb, William M. 1875 On the Indian Tribes and Languages of Costa Rica. Proceedings of the American Philosophical Society 14(95): 483–602. Giraldo, Santiago 2010 Lords of the Snowy Ranges: Politics, Place, and Landscape Transformation in Two Tairona Towns in the Sierra Nevada de Santa Marta, Colombia. Ph.D. dissertation. Division of the Social Sciences, Department of Anthropology, University of Chicago. Gómez Belmonte, Luis, and Karel Soto Solórzano 2001 Tras las Huellas de los Antiguos Pobladores del Sitio El Zoncho: Una Aproximaciûn al Modo de la Vida de los Agricultures Especializados, San Vito de Coto Brus. Vínculos 26(1): 145–164. Guardia, Ricardo Fernandez 1913 History of the Discovery and Conquest of Costa Rica. Translated by H.W. Van Dyke. Thomas Y. Crowell Company, New York. Haberland, Wolfgang 1957 Excavations in Costa Rica and Panama. Archaeology 10(4): 258–263. Haberland, Wolfgang 1984 The Archaeology of Greater Chiriquí. In The Archaeology of Lower Central America, SAR Advanced Seminar Series, edited by Frederick Lange and Doris Z. Stone, pp. 233–253; Douglas W. Schwartz, general editor. University of New Mexico Press, Albuquerque. Haller, M.J. 2004 The Emergence and Development of Chiefly Societies in the Rio Parita Valley, Panama. Ph.D. dissertation. Department of Anthropology, University of Pittsburgh. Hartman, C.V. 1901 Archaeological Researches in Costa Rica. Royal Ethnographical Museum, Stockholm. Hartman, C.V. 1907 Archeological Researches on the Pacific Coast of Costa Rica. Carnegie Institute, Pittsburg. Helms, Mary 1979 Ancient Panama: Chiefs in Search of Power. University of Texas Press, Austin. Helms, Mary 1994 Chiefdom Rivalries, Control, and External Contacts in Lower Central America. In Factional Competition

and Political Development in the New World. New Directions in Archaeology, edited by Elizabeth M. Brumfiel and John W. Fox, pp. 55–60. Cambridge University Press, Cambridge, UK. Herrera, Roberto A. 2015 Social and Ritual Dynamics at El Cholo: An Upper General Valley Funerary Village of the Diquís Subregion, Southern Costa Rica. Ph.D. dissertation. Department of Anthropology, University of New Mexico, Albuquerque. Hoopes, John W. 1991 The Isthmian Alternative: Reconstructing Patterns of Social Organization in Formative Costa Rica. In The Formation of Complex Society in Southeastern Mesoamerica, edited by W.R. Fowler Jr., pp. 171–192. CRC Press, Boca Raton, FL. Hoopes, John W. 1996 Settlement, Subsistence, and the Origins of Social Complexity in Greater Chiriquí: A Reappraisal of the Aguas Buenas Tradition. In Paths to Central American Prehistory, edited by Frederick Lange, pp. 15–48. University Press of Colorado, Niwot. Hoopes, John, and Mark Chenault 1994 Excavations at Sitio Bolivar: A Late Formative Village in the Arenal Basin. In Archaeology, Volcanism, and Remote Sensing in the Arenal Region, Costa Rica, edited by P. Sheets and B. McKee, pp. 87–105. University of Texas Press, Austin. Hoopes, John W. and Oscar Fonseca 2003 Goldwork and Chibchan Identity: Endogenous Change and Diffuse Unity in the Isthmo-Colombian Area. In Gold and Power in Ancient Costa Rica, Panama, and Colombia, edited by J. Quilter and J.W. Hoopes. Dumbarton Oaks Research and Library Collection, Washington, DC. Ichon, Alain 1968 Le Problème de la Céramique de Barriles. Boletín del Museo Chiricano del Colegio Félix Olivares 6: 15– 25. Ichon, Alain 1980 Archéologie du Sud de la Péninsule d’Azuero, Panama. Etudes Mésoaméricaines/Estudios Mesoamericanos; Série II Mexico (3): 1–521. Kantner, John 1988 Early Cultural Settlement of High Mountain Zone in the Diquís Subregion of Southern Costa Rica: The Herradura Site Survey. The Colorado College, Associated Colleges of the Midwest. Manuscript of final report in archive. Museo Nacional de Costa Rica, San José. Lange, Frederick 1992 The Search for Elite Personages and Site Hierarchies in Greater Nicoya. In Wealth and Hierarchy in the Intermediate Area: A Symposium at Dumbarton Oaks, 10th and 11th October 1987, edited by Frederick Lange, pp. 109–140. Dumbarton Oaks Research and Library Collection, Washington, DC. Laurencich de Minelli, Laura, and Luigi Minelli 1966 Informe Preliminar Sobre Excavaciones Alrededor de San Vito de Java. Proceedings of the Actas, 36th International Congress of Americanists, Seville [Spain], 1: 416–427. Laurencich de Minelli, Laura, and Luigi Minelli 1973 La Fase “Aguas Buenas” en la Region de San Vito de Java (Costa Rica): Informe Preliminar. Proceedings, 40th International Congress of Americanists, Rome [Italy], 1: 219–224. Linares, Olga F. 1977 Ecology and the Arts in Ancient Panama: On the Development of Social Rank and Symbolism in Central Provinces. Studies in Pre-Columbian Art and Archaeology No. 17. Dumbarton Oaks Research and Library Collection, Washington, DC. Linares, Olga F., Payson D. Sheets, and E.J. Rosenthal 1975 Prehistoric Agriculture in Tropical Highlands. Science 187: 137. Lothrop, Samuel K. 1919 The Discovery of Gold in the Graves of Chiriquí, Panama. Museum of the American Indian, Heye Foundation, New York. Lothrop, Samuel K. 1934 Archaeological Investigation in the Province of Coclé, Panama. American Journal of Archaeology 38(2): 207–211. Lothrop, Samuel K. 1963 Archaeology of the Diquís Delta, Costa Rica. Peabody Museum of Archaeology and Ethnology. Papers, Vol. 5. Peabody Museum, Cambridge, MA. Maloof, George 2012 Informe Sobre las Investigaciones Arqueologicos Relacionadas con el Proyecto Arqueologico Diquis Rosario, Temporada 2011. Instituto Costarricense de Electricidad, Proyecto Hidroelectrico El Diquis, Costa Rica. Maloof, George, M. Brenes, G. Mora-Sierra, and Anayensy Herrera 2011 Informe Final Sobre las Investigaciones del Proyecto Arqueologico Camaronal, Palmar Norte, Puntarenas. Instituto Costarricense de Electricidad, Proyecto Hidroelectrico El Diquis, Costa Rica. Mason, J. Alden 1942 New Excavations at the Sitio Conte, Coclé, Panama. Proceedings of the Eighth American Scientific Congress 2: 103–107. Washington, DC: Department of State. Menzies, Adam C.J., and M.J. Haller 2012 A Macroregional Perspective on Chiefly Cycling in the Central Region of Panama during the Late Ceramic II Period (AD 700–1522). Latin American Antiquity 23(4): 449–466. Merritt, J. King 1860 Report on the Huacals, or Ancient Graveyards of Chiriquí. American Ethnological Society, New York. Murra, John V. 1980 The Economic Organization of the Inka State. 1979 edition. JAI Press, Greenwich, CT. Murra, John V. 1985 The Limits and Limitations of the “Vertical Archipelago” in the Andes. In Andean Ecology and Civilization: An Interdisciplinary Perspective on Andean Ecological Complementarity: Papers from Wenner-Gren Foundation for Anthropological Research Symposium No. 91, edited by Shozo Masuda, Izumi Shimada, and Craig Morris, pp. 3–14. University of Tokyo Press, Tokyo. Murillo-Herrera, Mauricio 2003 Una Reseña y Caracterizacion del Period Tempisque (500 a.C–300 d.C.). Vínculos 28: 135– 147. Norr, Lynette 1986 Archaeological Site Survey and Burial Mound Excavations in the Rio Naranjo-Bijagua Valley. Journal of

the Steward Anthropological Society: Prehistoric Settlement Patterns in Costa Rica 14: 135–156. Palumbo, S.D. 2009 The Development of Complex Society in the Volcán Barú Region of Western Panama. Ph.D. dissertation. Department of Anthropology, University of Pittsburgh. Palumbo, Scott D, Laura Brodie, William A. Locascio, Gina B. Yost, and Francisco Corrales 2013 Early Social Complexity in Southern Costa Rica: New Evidence from Bolas. Antiquity 87(337). Piperno, Dolores R. 2011 The Origins of Plant Cultivation and Domestication in the New World Tropics. Current Anthropology 52(S4): S453–S470. Piperno, Dolores R., David R. Harris, and Gordon C. Hillman 1989 Non-Affluent Foragers: Resource Availability, Seasonal Shortages, and the Emergence of Agriculture in Panamanian Tropical Forests. Foraging and Farming: The Evolution of Plant Exploitation. In Foraging and Farming: The Evolution of Plant Exploitation. Proceedings of the World Archaeological Congress. One World Archaeology Series. 13th edition, edited by David R. Harris and Gordon C. Hillman, pp. 538–554. Unwin Hyman, London and Boston. Quilter, Jeffrey 2004 Cobble Circles and Standing Stones: Archaeology at the Rivas Site, Costa Rica. University of Iowa Press, Iowa City. Quilter, Jeffrey, and Aida Blanco-Vargas 1995 Monumental Architecture and Social Organization at the Rivas Site, Costa Rica. Journal of Field Archaeology 22(2): 203–221. Sheets, Payson D. 2009 When the Construction of Meaning Preceded the Meaning of Construction: From Footpaths to Monumental Entrances in Ancient Costa Rica. In Landscapes of Movement: Trails, Paths, and Roads in Anthropological Perspective. Penn Museum International Research Conferences, Vol.1, edited by James E. Snead, Clark L. Erickson, and J. Andrew Darling, pp. 159–179. University of Pennsylvania Press, Philadelphia. Sheets, Payson D. 2011 Pilgrimages and Persistent Social Memory in Spite of Volcanic Disasters in the Arenal Area, Costa Rica. Ancient Mesoamerica 22(2): 437–448. Sheets, Payson, and Tom Sever 2006 3-D Visualization of Ancient Processional Landscapes in Costa Rica. BAR International Series 1568: 415–420. Skinner, Alanson 1920 Notes on the Bribri of Costa Rica. Indian Notes and Monographs, 6(3). Museum of the American Indian, Heye Foundation, New York. Snarskis, Michael J. 1984 Central America: The Lower Caribbean. In The Archaeology of Lower Central America, SAR Advanced Seminar Series, edited by Fred Lange and Doris Stone, pp. 195–232; Douglas W. Schwartz, general editor. University of New Mexico Press, Albuquerque. Sol Castillo, Felipe 2013 Religious Organization and Political Structure in Prehispanic Southern Costa Rica. Ph.D. dissertation. Department of Anthropology, University of Pittsburgh. Stirling, Matthew William 1949 Exploring the Past in Panama. National Geographic Magazine 95(13): 373–399. Stone, D. 1943 A Preliminary Investigation of the Flood Plain of the Rio Grande de Terrabá, Costa Rica. American Antiquity 9(1): 74–88. Stone, D. 1977 Pre-Columbian Man in Costa Rica. Peabody Museum Press, Harvard University, Cambridge, MA. Zilberg, Jonathan L. 1986 The Diquís Petroglyphs: Distribution, Archaeological Context and Iconographic Content. Journal of the Steward Anthropological Society: Prehistoric Settlement Patterns in Costa Rica 14: 339–359.

Part IV

South America

9 Chanka settlement ecology Disentangling settlement decision-making during a time of risk in the Andean highlands Lucas C. Kellett In this chapter, I explore prehistoric settlement ecology, in particular the process and outcomes of prehistoric decision-making during a period of convergent risks. Using settlement data from highland Peru, I discuss how particular settlement arrangements can serve as a sensitive indicator for a range of environmental, political, and social risks. I argue that through the application of an ad hoc risk analysis, archaeologists can better understand the relative importance that certain environmental, economic, political, and social factors played in the formation of prehistoric settlement arrangements. A geospatial analytical approach offers a promising way of identifying and weighing the individual pressures that influenced settlement decisions in the past. Risk can be generally defined as the unpredictable variation and potential for loss or damage from particular natural or social hazards (Zori and Brant 2012: 403; see Halstead and O’Shea 1989). Early examination of risk stemmed in part from behavioral ecology (Bird and O’Connell 2006; Cashdan 1990; Winterhalder 1986, 1990), which treated it within the rubric of diet choice among hunter-gatherers and simple agricultural societies. Since then risk discourse has expanded and been developed for a broad range of semi-sedentary and sedentary horticultural, agricultural, and pastoral societies (e.g. Browman 1987; Earle and Christenson 1980; Fisher et al. 2009; Goland 1993; Kuznar 2001; Marston 2011). While a general conception of risk is somewhat straightforward to understand, the complex human responses to those risks are not. Especially challenging is unraveling and understanding competing yet connected risks, and their impacts on past societies, given the inherent limitations of the archaeological record. Additionally, one can distinguish between potential (future) and actual (current) risk, yet archaeological indicators often do not discern between the two risk classes. For example, to an archaeologist, a population preparing for the possibility of attack by a neighboring group can often appear very similar to one that experiences ongoing attacks (Leblanc and Rice 2001: 8). Furthermore, while archaeologists recognize the wide and variable range of impacts that risk had on human societies, more vexing are the interpretive problems, since related causes, symptoms, and effects are difficult to distinguish. Equally problematic is that it can be difficult to generalize effects of risk, since they are not experienced by humans (and their social groupings) in a uniform manner. Archaeologists interested in understanding the cultural impacts of risk would ideally be able to identify single risk factors and their isolated effects, but of course that is not the nature of risk. Risk may source from one location or enter the human

experience in a certain manner, but those risks are rarely contained and often transfer into other arenas of the human and cultural experience. For example, while economic risks in the form of food shortages in the past caused malnutrition, general morbidity, and starvation, they also frequently affected social and political organization (i.e. change in leadership), peer polity interactions (i.e. increased raiding of neighbors), and subsistence organization (e.g. shift from intensification to diversification). In addition to risk transfer, risk can be spatially and temporally convergent, and can be compounding (e.g. Moseley 1999), thereby offering a unique interpretive challenge to archaeologists, since the characteristics and impacts of individual risk agents can be masked, or may overlap, thus clouding a clear understanding of the causes and effects of these risk agents. In this way, the construction of a “risk profile” for ancient social groups can be more complicated than reconstructing paleoclimatic or ethnohistoric records for example. Despite these methodological and epistemological challenges, archaeologists have begun to better define risk in a dichotomous fashion with 1) risk serving as the agent of change and 2) the human response as the behavioral outcome of that risk. Most recent studies of risk response have derived from large events of socio-environmental change such as climate and collapse scenarios (Bawden and Reycraft 2000; DeMenocal 2001; Diamond 2011; McAnany and Yoffee 2009; McIntosh et al. 2000), warfare (VanDerwarker and Wilson 2015), and anthropogenic landscape change caused by humans (Fisher et al. 2009; Redman 1999; Redman et al. 2004). The interdisciplinary nature of these studies reflects the complexity as well as the promise and pitfalls of answering large and important anthropological questions tied to risk (e.g. Sandweiss and Kelly 2012). Yet, on a more practical level, such studies may provide important insights into human adaptation in today’s globalized world with regard to the way that humans have responded and can continue to successfully respond to broad categories of risk through time. One promising avenue towards understanding how prehistoric populations experienced risk is by using settlement patterns as a unit of analysis. Settlement patterns have been empirically shown to be sensitive to a diverse suite of risk factors. More specifically, settlement location (e.g. topographic location, elevation, aspect, distance to local resources and neighboring sites) and general characteristics (e.g. size, architectural layout, function) are in part contingent on real, perceived, or potential environmental and social risks. In the Americas, successful studies examining prehistoric settlement and risk, especially from economic (e.g. agricultural production) and political (e.g. internal, external warfare) angles, derive from many areas, including the American Southwest (e.g. Dean 1996; Jones 2010; Schollmeyer 2011; Kohler et al. 2011), Mesoamerica (e.g. Elliott 2005; Nichols 1987; Scarborough 2000), and the Andes (e.g. Arkush and Tung 2013; Conrad 1978; Kellett 2010; Zori and Brant 2012; Zaro 2007; Zaro and Umire-Alvarez 2005). Shifting settlement patterns and their methodological examination through a geospatial approach in particular offer a useful method to begin to construct a risk profile for a given prehistoric population.

A settlement ecology approach to the study of risk One step towards an improved understanding of the prehistoric is the use of settlement patterns

as an indicator of variable risk exposure and response. This approach fits broadly under the term “settlement ecology,” a recent trend in archaeology that is related to the older tradition of cultural ecology, and which holds that settlement is a behavioral adaptation to the social, cultural, and natural environment (e.g. Hasenstab 1996; Jones 2010; Jones et al. 2012; Stone 1996). As outlined in Chapter 1 of this volume, a settlement ecology approach considers settlement as contingent on numerous conditions, factors, and dynamics, which serve to attract or repel, or “push and pull” settlements on a theoretical landscape. These factors are highly localized and can include a range of natural (e.g. water, agricultural land), social (e.g. labor availability), and political (e.g. ethnic territorial boundaries) factors, which can be weighted and integrated into decision-making. This suite of factors is also impacted by a range of additional contingent variables (e.g. occupational period, local environment, history, subsistence organization, population) of a given society in a given time and in a particular region. Ultimately, a settlement ecology may be best understood as the cumulative output of extensive and complex decision-making such that the resulting pattern is the compromise of the costs and benefits of a set of social, political, and environmental dynamics, as well as the negotiation of competing interests by a diverse range of stakeholders (Wernke 2013). Conceptions of risk can be translated into a settlement ecology approach, such that one can envision a spatially defined risk landscape, which helps guide settlement choice. Under this premise we can predict that low-risk locations on the landscape tend to “pull” or attract populations toward settlement, while high-risk locations tend to “push” or repel populations “away from” settlement. It is the overlapping dynamics of “pulling” toward and “pushing” from settlement locations in response to a multitude of risks which generate a spatially defined and variable risk landscape, with “hot” and “cold” spots of risk. Furthermore, it is within this context that the analysis of the risk landscape can offer valuable insights into the role that certain risks play on the design and materialization of a given settlement ecology (e.g. Marston 2011).

Translating risk into settlement priorities It is clear that risk has a salient effect on nearly all societies and the daily lives of populations. In the central Andes, for example, chronic environmental uncertainty in concert with a rapidly changing political landscape made “living with risk” an integral part of life in the region (Dillehay and Kolata 2004; Moseley 1997; Parsons et al. 1997, 2000, 2013). These short-term and long-term risks fluctuate in a dynamic and unpredictable fashion across a range of temporal and spatial scales, and hold numerous implications for ancient settlement strategies. How did early Andean populations, for example, design and implement new settlement strategies in the face of continuous risk, and upon which overlapping variables did they rest? One place to start is to identify those basic needs that human populations required during periods of risk. Decades ago, Abraham Maslow (1943) published a seminal paper entitled, “A Theory of Human Motivation,” which modeled a generalized hierarchy of human needs. Although written for the field of psychology, this early model was one of the first studies to think about how certain human requirements must be met before other more “non-essential” human qualities can be developed. Maslow argued, for example, that physiological (e.g. food,

water) and safety (e.g. family, housing) needs must be met before more psychosocial development can occur (i.e. belonging, esteem, and self-actualization). The value of such a model is that it enables archaeologists to begin to disentangle complex cultural phenomena in a more orderly, hierarchically ordered set of qualities based on issues of need, value, etc. More specifically, I contend that this model offers inroads into how archaeologists might begin to conceive of how past settlement strategies intersect with a range of competing human needs, especially during periods of elevated risk. As such, concurrent social and environmental risks can be translated into a hierarchy of settlement priorities, which can be identified by archaeologists in order to comprehend settlement decision-making processes (Figure 9.1). For example, one would expect that during a period of elevated risk, basic necessities of food and water (and the resources which support them, i.e. distance/access to water, pastoral or agricultural areas) should be given a higher value/priority than less essential qualities such as formal site-planning or quality of ceramics. In this way, archaeologists can construct a set of expectations for settlement (and those factors that influence it) and then examine deviations from those expectations, making it possible to understand the variable and competing human needs that serve to formulate a specific settlement ecology during a period of risk. I now turn to an application and preliminary evaluation of this model by employing archaeological data collected from the south central Andean highlands.

Figure 9.1

Simplistic conceptual model showing hierarchy of risk (left) which can be translated to a set of settlement priorities (right) (after Maslow 1943).

Changes in late highland Andean prehistory While the central Andean highlands (southern Peru and northern Bolivia) experienced at least 4,000 years of complex cultural developments, the late period of this cultural sequence witnessed an extraordinary set of dynamic cultural and environmental changes, especially from AD 800 to 1532. These include the expansive and complex political development of Tiwanaku and Wari during the Middle Horizon (~AD 600–1000) and the long Late Intermediate Period (~AD 1000–1400, hereafter LIP), which saw widespread political decentralization and competition. This period was then followed by the galvanization and rapid expansion of the Inka empire in the Late Horizon (~AD 1400–1532). This chapter examines the Middle Horizon–Late Intermediate transition (~AD 800–1250) in the central highlands, which saw a

shift from the complex political development, general atmosphere of peace, and climate stability of the former period to the political decentralization, growing territorialization, increased conflict, and climate change that characterized the latter period. The Middle Horizon of the central Andes was a period of regional political fluorescence for the large twin polities Tiwanaku and Wari (e.g. Isbell and McEwan 1991; Kolata 1993; Janusek 2008; Schreiber 1992). During this time several dozen large Wari provincial centers were established throughout the northern half of the central highlands (the extent of modern-day Peru) (Jennings and Craig 2001), which were linked to the imperial capital of Huari in the Ayacucho Valley (Isbell 1988). While the extent and nature of control by the empire has been questioned on a number of grounds (Jennings 2006), regional archaeological data reflect a period of population growth, which was stimulated in part by the intensification of maize agriculture linked to terrace agriculture in many valleys (Finucane 2009; Finucane et al. 2006; Hastorf 1993; Schreiber 1992; Williams 2006). In the southern half of the central highlands (northern highland Bolivia), Tiwanaku expanded away from Lake Titicaca and established a series of colonies and enclaves on both sides of the Andes (e.g. Kolata 1993, Janusek 2008, Goldstein 2004). Tiwanaku offered a more ritualbased society (Janusek 2008) in contrast to the more overtly political nature of the Wari empire (Schreiber 1992; Isbell and McEwan 1991). The Tiwanaku were defined by a set of far-flung colonies linked by camelid trading caravans, which were tethered to and supported by the enormous and densely populated Titicaca Basin. The productive potential of the basin was defined by a complementary trifecta or triad of camelid pastoralism, lacustrine resources, and raised field agriculture that supported millions of basin inhabitants for millennia (Janusek 2008; Kolata 1993, 1996). The political and religious capital of the Tiwanaku polity was centered in the large civic-ceremonial center of Tiwanaku, near the southwestern shore of Lake Titicaca, where pilgrimage seemed to play an especially important role. The Middle Horizon in the central Andes was marked by the political and economic rise of Wari and Tiwanaku, and a period of fluorescence, with largely peaceful conditions prevailing in the occupied hinterlands. The LIP, which began c.AD 1000, was ushered in by a shift from a time of relative security to a time of risk, which redefined the socio-environmental landscape. This prolonged and risky era is defined by sustained environmental and socio-political upheaval (e.g. Arkush 2006, 2008; Conlee et al. 2004; Covey 2008; Parsons and Hastings 1988) linked to two principal macro-regional phenomena: one environmental—the prolonged “LIP drought”; and one political—the dissolution of the expansive twin polities of Wari and Tiwanaku. The “LIP drought”, which began c.AD 1000, has been confirmed in a range of proxy records from regional ice (Thompson et al. 1985) and lake cores (Abbott et al. 1997a, 1997b; Binford et al. 1997; Chepstow-Lusty 2011; Chepstow-Lusty et al. 2009; Hillyer et al. 2009; Sublette et al. 2012; Thompson et al. 1985; Valencia et al. 2010; see also Contreras 2010 for a general review for the central Andes). There is a general consensus among scholars that regional precipitation rates likely reduced by as much as 20 percent, and may have lowered water tables by more than 10 m in the Titicaca Basin. While it appears the most severe period of the drought occurred in the thirteenth century AD (~AD 1250), which post-dates the imperial collapse (~AD 1000–1100) of Tiwanaku and Wari (Williams 2002, 2006), it is generally

argued that this prolonged dry spell served as an important proximate cause of collapse. This, in turn, impacted regional subsistence-based economies, especially raised field agriculture for Tiwanaku (Kolata 1993, 1996, 2000; Ortloff and Kolata 1993). It is, however, less clear if this severe drought had impacts on Wari subsistence, but it is likely that it had deleterious effects on irrigation systems (especially for maize-based agriculture; Williams 2006) as well as for simple dryland/rainfall agricultural production (Kellett 2010). While aridity levels are wellconfirmed through multiple proxy records, temperature fluctuations are less well known. While earlier studies argued that the LIP was generally colder (Seltzer and Hastorf 1990), it is now more likely that the first half of the 400-year period was warmer, due to the Medieval Climatic Anomaly (see Arkush 2008), and the latter half was colder, due to the temporal proximity to the Little Ice Age (Thompson et al. 1985). Although temperature fluctuations likely had impacts on crop zonation, frost, planting/harvesting schedules, and broad settlement patterns, more research is still needed to clarify long-term temperature fluctuations and their associated impacts on the region. Although these two convergent phenomena translated into elevated economic and political risk, they also appear to have been interconnected and likely amplified crisis–collapse dynamics between the tenth and twelfth centuries AD (Moseley 1997, 1999; Kellett 2013a; Williams 2002, 2006). More specifically, a range of regional archaeological research shows that LIP populations witnessed subsistence stress and resource competition, while at the same time enduring increased threats of violence and warfare, which appear to have climaxed in the twelfth and thirteenth centuries AD (e.g. Arkush 2006, 2008, 2011; Arkush and Tung 2013; Bauer and Kellett 2010; Covey 2008; D’Altroy and Hastorf 2001; Hastorf 1993; Kellett 2010, 2013a, 2013b; Kurin 2012; Stanish 2003).

Introduction to the Chanka of Andahuaylas The Andahuaylas region of western Apurímac, Peru, was the traditional homeland of the Chanka ethnic group who came to prominence during the Late Intermediate Period. The Chanka were one of dozens, if not hundreds, of similar ethnic groups (etnías, señorios), which functioned as simple chiefdoms and inhabited the central highlands before the arrival of the Inka (cf. Rowe 1946). The Chanka are well known by archaeologists, historians, and colonial writers as the traditional foes of the Inka and are often portrayed as aggressive and warlike. Most infamous was the legendary Chanka–Inka war (believed to have occurred in the early 1400s), in which the Chanka failed to capture the capital city of Cuzco and were defeated by the Inka (e.g. Betanzos 1987 [1551: Ch. 6–10]; Cieza de Leon 1984 [1553: Ch. 44–45]; Garcilaso 1966 [1609]). The research presented here is part of recent survey and excavation work (e.g. Bauer and Kellett 2010; Bauer et al. 2010; Kellett 2010; Kellett et al. 2015a; Kurin 2012) undertaken over the past two decades that treats Andahuaylas as the traditional Chanka homeland (cf. Garcilaso 1966 [1609]; Julien 2002). The Andahuaylas (Chumbao) Valley is one of only a few broad valleys in the region; it is suitable for agriculture due to its moderate elevation range (2,700–4,400 meters above sea level [masl]) and moderate topographical relief. The valley is framed by the large Rio Pampas to the west and north, and by a large expanse of high grasslands (puna) to the south (Figures

9.2 and 9.3). To the northeast it is bordered by the large Laguna Pacucha, which defines a different drainage basin and contributes water to the Rio Pampas to the northwest. The vertical ecology in Andahuaylas can be defined by a series of stacked agro-pastoral production zones, ranging from warm, xeric river bottoms below 2,500 masl to cold grasslands above 4,000 masl. From lowest to highest elevation, these zones include: the yunga (0–2,700 masl), warm and dry valleys suitable for warm weather cultigens (i.e. cotton, coca, avocado); the quechua (2,700–3,500 masl), the breadbasket zone and focus of legume (i.e. tarwi) and maize production; the suni (3,500–3,800 masl), a transitional agro-pastoral zone which is suitable for tuber (i.e. potato, oca, olluca), grains (quinua, kiwicha), and camelid pastoralism (i.e. llama, alpaca); and finally the cold, windy puna (3,800+ masl), a high grassland area suitable for small amounts of tuber and grain cultivation, and a core area for intensive camelid production. Together these zones and the eco-tonal boundaries that distinguished them played an important role in how the Chanka and other populations in Andahuaylas made decisions concerning how to spatially organize their settlement and subsistence regimes.

Figure 9.2

Regional map showing study area and Chanka ethnic region.

Figure 9.3

Andahuaylas Valley showing Chanka Settlement Project area. Inset figure displays relative percentages of ecological zones (yunga, quechua, suni, and puna) within the project area polygon.

Regional settlement data from Andahuaylas As mentioned, two recent overlapping surveys form the basis for the robust settlement database in Andahuaylas: the Andahuaylas Archaeological Project (PAA) (Bauer et al. 2010) and the Chanka Settlement Project (Proyecto Asentamiento Chanka [PAC]) (Kellett 2010). Together, along with recent bioarchaeological research by Kurin (2012), they offer an accurate picture of life during the Middle Horizon–LIP transition (AD 800–1200). Furthermore, a local proxy record from a lacustrine sediment core collected from Laguna Pacucha confirms drought conditions that appear to have commenced c.AD 1000 and climaxed c.AD 1250 (Hillyer et al. 2009; Valencia et al. 2010), reflecting the important role that changing paleoclimatic conditions had on local populations in the valley for millennia (Kellett et al. 2015b). Previous archaeological survey work (Bauer et al. 2010) confirmed that the Wari-occupied Middle Horizon (Wari [AD 600–1000] and Qasawirka [300 BC–AD 1000]) ceramic phases in Andahuaylas were characterized by high-valley populations, which were distributed across hundreds of small (0.5–3 ha) mid- to lower- (suni and quechua zones) elevation (2,700–3,500 masl) hamlets and villages located on moderate slopes and valley bottoms. This settlement pattern generally reflects a relatively low level of internecine conflict with peace predominating under Wari hegemony in Andahuaylas. The Wari empire seems to have employed a very “hands off” strategy of imperial control in

Andahuaylas, despite the region’s close proximity to the capital center. This is reflected in the absence of any regional administrative centers, major terrace infrastructure, or widespread access to locally made and imported imperial ceramics in the valley (Bauer et al. 2010; Kellett 2011). Finally, the dense valley population also suggests an increased dependence on and intensification of valley agriculture, especially maize. This has been confirmed in excavation work in the valley (Kellett 2010, 2011), local survey (Bauer et al. 2010; Kellett 2010), regional survey (Schreiber 1992), and regional stable isotope analysis (Finucane 2009; Finucane et al. 2006). While local survey data from Bauer et al. (2010) clearly show that not all Middle Horizon (Qasawirka and Wari phase) sites were abandoned c.AD 1000–1100, the total number of sites dropped by nearly 58 percent between Qasawirka/Wari (total n = 472) and Chanka phases (n = 202). The settlement ecology of the valley during the LIP (Chanka phase) changed drastically and is characterized by two noticeable changes: 1) the movement and aggregation of local populations into high-elevation (3,600+ masl) ridgetop sites, often fortified and located in defensible locations (Figure 9.5); and 2) the increased dependence on camelid pastoralism by Chanka populations, reflected in a new abundance of corrals at higher elevation (3,600+ masl) (Bauer et al. 2010; Bauer and Kellett 2010; Kellett 2010). In addition to a reduction in the number of sites in the valley, Bauer et al. (2010) also found a general increase in site size by 20 percent (Wari/Qasawirka combined phases, mean site size = 0.60 ha; Chanka phase, mean site size = 0.75), with seven very large sites ( > 5 ha) location in the suni and puna zones (3,500+ masl). The long, four-century Chanka phase in the Andahuaylas region can be divided into two occupational periods, Chanka I (~AD 1000–1250) and Chanka II (~AD 1250–1400), a chronological division that has been offered by other LIP researchers in other areas of the central Andes (e.g. D’Altroy and Hastorf 2001). This is reflected by the fact that most radiometric dates from residential sites in Andahuaylas fit conveniently into Chanka I (see dates from Bauer et al. 2010; Kellett 2010). The dates indicate that the earliest and latest hilltop sites were constructed and settled in the eleventh century AD and late thirteenth century AD, respectively. Where other areas of the region see a temporal gap in the establishment of defensive refuges and fortified hilltop sites (i.e. Arkush 2006, 2008, 2011 in the Titicaca Basin), in Andahuaylas we see populations begin to abandon lower valley areas in the eleventh century, immediately after the collapse of the Wari empire and with the onset of arid conditions. Additionally, we see drastically higher rates of skeletal trauma during Chanka I when compared with previous centuries, indicating the heightened level of internecine conflict in the form of raids (Kurin 2012, 2013; Kurin et al. 2014). The nature of the Chanka II period (AD 1250–1400) is still poorly known, and it is unclear if this is the result of limited radiometric dates from residential hilltop sites or the abandonment of these large sites, with hilltop site populations moving back down into the mid- and lower valleys. The uphill movement and aggregation of Chanka populations into higher elevation, permanent settlements also required a shift in subsistence. I have argued elsewhere that the Chanka I period witnessed a shift from a valley-centered maize-based strategy to a more integrated agro-pastoral subsistence, spatially proximate to these large habitation sites. I posit that this would have functioned as an effective subsistence strategy in the face of heightened

economic and political risk (Kellett 2010, 2013a, 2013b). The increased dependence by local populations on pastoralism during the risky LIP has been documented in other regions of the south central Andean highlands including the Cusco (Covey 2008) and Lake Titicaca regions (Arkush 2011; Stanish 2003).

The Chanka Settlement Project The Chanka Settlement Project (PAC), undertaken in 2005–2006, was a multi-tier archaeological research project focused on understanding settlement and landscape use in the Andahuaylas region during the LIP (Kellett 2010). While preliminary recording of Chanka sites in the valley occurred several years earlier (Bauer et al. 2010), this project was designed to use a more granular, spatial approach to examining the natural and built landscape, in order to understand how the Chanka confronted and successfully adapted to challenging environmental and social conditions between AD 1000 and 1400. This project included two general phases of fieldwork including a full coverage survey (75 km2), as well as a series of test excavations at two neighboring Chanka phase habitation sites, Achanchi and Luisinayoc (~AD 1000–1400) (Kellett 2010). The objective of the survey was twofold: first to identify all Chanka phase structures around a pair of large hilltop sites in order to understand how neighboring micropolities interacted and used the natural and built landscape. The second objective was to complete a site catchment survey to identify all relevant ecological and natural landscape features (e.g. water sources, available building stone) that may have played a role in settlement decision-making. The archaeological survey component of the project recorded all Chanka phase (~AD 1000– 1400) sites (n = 69), which were identified by the presence of diagnostic LIP ceramics and/or limestone (pirka) masonry architecture. A range of Chanka phase sites were recorded, including habitation sites, herding structures (camelid corrals), agricultural terraces (e.g. single or multiple), various defensive structures (i.e. walls, ditches, lookouts), and a range of burial sites (e.g. communal cave tombs [machays], and circular above-ground burial chambers [chullpas]). Chanka habitation sites were the easiest to identify during the survey, given their large size and high density of intact pirka (limestone) architecture. These large aggregated sites, exemplified by the sites of Achanchi (PAA-225) and Luisinayoc (PAA-220), have extensive arrangements of densely packed, agglutinated architecture reflecting hastily constructed sites which served to protect the thousands of residents during the LIP. Approximately 14 percent (n = 10) of the recorded structures from the PAC were habitation sites and these ranged in size from 0.5 to 16 ha, with a mean site size of 6 ha (Figures 9.4 and 9.5). Overall, 60 percent (n = 6) of habitation sites were over 5 ha in size and only a single habitation site was less than 1 ha in size. A statistical analysis between the size and elevation of Chanka habitation sites reveals a significant positive relationship (Pearson’s = 0.68, r² = 0.46, p < 0.05), clearly demonstrating the pattern of population aggregation at higher elevations during the LIP. Of these sites, 0 percent were located in the yunga zone, 40 percent (n = 4) were located in the quechua zone, 20 percent (n = 2) were located in the suni zone, and 40 percent (n = 4) were located in the high puna zone. Compared with the percentages of the project area covered by

different ecological zones (Figure 9.3), this suggests the preference for middle- to highelevation areas for the establishment of large aggregated residential sites. The survey team was successfully able to identify a range of subsistence structures, including numerous isolated corrals (n = 15) and larger corral complexes (n = 13) (Figure 9.6). The widespread presence of herding corrals comprised just over 40 percent of all recorded structures and reflected the dependence of local Chanka populations on camelid pastoralism. Habitation sites made up the next most common site type with 15% (n = 10), followed by terrace/corral complexes and Chanka burial sites, which each contributed almost 12 percent (n = 8) of the site sample. In general, the settlement patterns show a trend of large, permanently settled Chanka habitation sites, which were surrounded by a constellation of corral sites in the adjacent suni and puna zones (3,500+ masl), while fewer isolated terraces and terrace/corral complexes were located in the upper quechua zone (3,200–3,500 masl).

Figure 9.4

Project area showing spatial distribution of Chanka habitation sites.

Figure 9.5

Photo of Toxsama Valley showing local high-relief landscape and domestic structures (lower right) at the large Chanka phase ridgetop site of Achanchi (view to the southeast).

An analysis of Chanka settlement priorities The pronounced shift in settlement ecology between c.AD 800 and 1200 in regions such as Andahuaylas illuminates the sequential stages in settlement decisions that were made by many local populations. The stages include site abandonment, population movement, and the construction of new, large aggregated ridgetop settlements. One can imagine local populations weighing the costs and advantages of leaving the lower valley. Undoubtedly, the decision to move uphill for many was complicated, time and resource consuming, and undoubtedly carried a host of complex and potentially serious social, cultural, and emotional impacts. So, where specifically did Chanka populations decide to move and why? Since it is estimated that these densely occupied residential sites housed between several hundred and several thousand people at any one time (Kellett 2010; see also Arkush 2011; D’Altroy and Hastorf 2001), appropriate site selection must have been a critical decision for Chanka leaders, who were tasked with movement and resettlement of local populations. What criteria or settlement priorities appear and have been most important for the relocation of Chanka populations to higher elevations? This assessment of these settlement factors or priorities requires two steps. The first is the identification of possible factors, while the second is assessing the relative importance of individual factors. In both cases, this presents a challenge to the archaeologist since one is confronted with the entire spectrum or totality of influences that possibly came into play in a given settlement decision. Despite these challenges we can still isolate individual factors and not only examine their fundamental “push” and “pull” power for ancient populations, but also test these assumptions against a data set. The integration of survey data into a geographical information system (or GIS) enables one to assess the relative contribution of differing settlement priorities, and in this case the establishment of Chanka habitation sites in particular

locations. Although simple spatial correlation has its limitations (i.e. autocorrelation) in landscape and settlement analysis, this simple measure can provide valuable insights into prehistoric settlement priorities.

Figure 9.6

Map showing the range of residential and subsistence structures located in the project area.

In this chapter I construct a simple conceptual model, adapted from Maslow (1943), which outlines a set of expectations in the form of ranked settlement priorities that rest on a set of hierarchal needs (Figure 9.1). During the early LIP (Chanka I) we can predict the following settlement priorities, ranging from most to least important: 1) basic access to drinking water; 2) land for agro-pastoral production; 3) access to appropriate building material to create a safe and permanent shelter; 4) adequate site defensibility (which includes defensive topographic locations in concert with constructed fortifications). Based on these expectations, I then compare observed settlement and landscape data to assess the ranking of settlement priorities outlined above (Table 9.1). When we examine the sample of ten habitation sites in the PAC survey area, we see that 100 percent of these population centers have at least one steep slope (> 40°) on at least one side of the site, and typically there are two sides with steep slopes (Figures 9.4–9.6). More specifically, we can see that Chanka populations preferred elevated (3500+ masl) ridgelines and mountain top locations with small flat areas, which were surrounded by steep slopes. Fifty percent (n = 5) of the habitation sites had steep slopes on opposite sides along the long axis of the ridgelines, which allowed for two access points along the ridge. In a similar vein, we observed that 40 percent (n = 4) of habitation sites had clear defensive fortifications (e.g.

ditches, protective walls) associated with them, all of which were located at easier access points to large residential sites along more moderately sloped ridgebacks and flanking slopes (Figure 9.7). Finally, a simple binary viewshed analysis was completed in ArcGIS 10.2 which yielded 31 percent of the project area visible among all habitation sites, but where 100 percent (n = 10/10) of these sites were located in commonly visible areas. While one could argue that this is simply the result of spatial autocorrelation (i.e. mountaintop sites naturally have better viewsheds), it is highly possible that the ability to view neighbors was especially important during higher threats of violent raids. Given the strong masonry tradition of the Chanka and other regional polities, we could expect that building stone may have played an important role in site selection. Based on the model, I predicted that construction material for settlements would have been the second most important settlement priority (after land and water requirements). Not surprisingly, the survey found that 90 percent (n = 9) of the Chanka habitation sites were located on or in close proximity (< 0.25 km) to exposed limestone bedrock. While these rocky exposures spatially correlate along high ridges (Figure 9.5), there are many sections of ridges that do not exhibit them—and in only one case did the project record a large habitation site in an area with little to no available bedrock. As such, easily extractable limestone bedrock for the construction of pirka-style masonry appears to have been an important priority in the site selection process. Table 9.1

Observed vs. expected outcomes for settlement location within the project area

Settlement priority rank (after maslow 1943)

Expected settlement priorities

Observed settlement priorities

Assessment (observed vs. expected)


Close proximity of habitation sites to water

70% of habitation sites located Lower-ranked settlement 0.25–1 km from nearest priority than expected spring


Close proximity of habitation sites to agricultural and pastoral land

86% of all recorded Similarly ranked settlement subsistence sites are < 1 priority as expected hour travel distance from habitation sites, but most pastoral structures located 4–6 km away from population centers


Close proximity of habitation sites to available building stone

90% of all habitation sites and Higher-ranked settlement ~85% of all agro-pastoral priority as expected structures located in close associated with exposed limestone bedrock


Modest pattern of site defensive location/layout

70% of habitation sites have Higher-ranked settlement naturally defensive slopes priority than expected on least 2–3 sides; 40% of habitation sites with clear fortifications; 100% of settlements are located in common viewsheds from one another.

When we consider the nearest available water sources to habitation sites, typically springs and seeps, we see that in no cases do the Chanka have a secure water source within the

boundary of a habitation site. In fact, of the ten habitation sites only two had springs located less than 0.25 km away, while seven of these sites had springs located 0.25–1 km away, and one habitation site was located a little over 1 km from the nearest spring. These data, which are reflected in Figure 9.8, generally reveal an inverse spatial correlation between residential settlement areas and water availability. Implications and potential explanations for this phenomenon are explained below.

Figure 9.7

Map showing viewshed results calculated in GIS as well as defensive features recorded during survey work.

Figure 9.8

Map showing spring density in the project area. Notice the lack of spatial correlation between habitation sites and available springs.

Finally, when we consider the distribution of agro-pastoral sites in relation to Chanka habitation sites, we can see that LIP agro-pastoral production was indeed local in nature. While pastoral activities appear to have been conducted slightly farther away (2–4 km) from population centers than agricultural activities (1–3 km), the general proximity indicates a locally organized subsistence regime (Kellett 2010). Furthermore, when travel time was calculated within a GIS (via Tobler’s walking function), results demonstrated that 86 percent of all recorded Chanka phase subsistence sites were located less than 1 hour’s travel time from habitation sites. If we can use the frequency and the proximity of these subsistence sites as a general proxy for the importance of pastoral and agricultural land in the selection of habitation sites, it appears that it played only a modest role in site selection. One should note that some of the largest of the habitation sites in the survey area were located along Cerro Achanchi along the northern edge of the survey area, which is located at least 5 km from the collection of agropastoral structures clustered in the central and southern sections of the survey region (Figure 9.6). Also, the fact that nearly all agro-pastoral structures, especially camelid corrals, were located on moderately sloped (< 25°) terrain, well to the south of most of population centers, tentatively indicates that severe topography constrained certain subsistence activities and the construction of permanent subsistence structures.

Discussion of results

A review of settlement priorities by Chanka populations in Andahuaylas reveals some interesting insights about decision-making processes during a time of elevated risk. First, earlier I offered a general model for settlement priorities in which basic needs of “food, water, and shelter” should be prioritized over concerns of “safety” according to Maslow (1943). While this model could be critiqued for its simplicity (i.e. one could easily argue that all these human requirements are needed; needs are not necessarily conceptualized or met in a logical order; survival may not always supersede other non-survival needs), its hierarchal layout does offer useful inroads into thinking about settlement decision-making and how those decisions may rest on a host of unequal factors, beliefs or motivations. When we compare these basic expectations with the observed results we can see immediately that site defensibility appears to be a much higher settlement priority than expected, given the ubiquitous presence of steep slopes in close association with the site locations (Table 9.1). The presence of additional defensive site fortifications at the entrance of these large residential sites also supports this conclusion, although we do not see even the majority of habitation sites (only 40 percent) with these defensive structures. Based on the natural and constructed defensive features found in close spatial association with Chanka residential sites, it appears that the concern for protection or “safety” of the population under periodic or chronic threat of potential or actual attack was a first settlement priority over others. A strong spatial correlation (90 percent) between habitation sites and areas of exposed limestone bedrock also suggests that in addition to the presence of adjacent steep slopes, newly constructed hilltop sites required an ample supply of available building stone. While high elevation ridges are of course rockier, the fact that all habitation sites were located where there were sizeable exposures of limestone bedrock suggests that this was an important resource. Other ridge and hillslope locations were not selected for settlement, thus further supporting this proposition. While it is likely the Chanka populations distributed across all agro-ecological zones built with a range of materials including adobe, the Chanka phase pirka architecture found predominantly at higher elevations in the valley likely served multiple functions in regard to site durability in an extreme wet and windy climate and the protection of resident populations with increasing threats of attack. Perhaps the most surprising result was the distance between permanent water sources and habitation sites, which was farther than expected. While water is an undeniable requirement for survival in all cases, it appears that water was not the critical primary factor for site selection. As outlined earlier, the majority of residential sites (70 percent) were located 0.25– 1 km from springs and seeps. Since water was a lower settlement priority than expected, it may indicate that there was less threat to travel for water collection. It could also suggest that there was a lower potential for extended sieges of sites by enemies, which would have necessitated a permanent intramural water source (Arkush 2008; Kellett 2010). Finally, the presence of numerous agricultural and pastoral features in moderate proximity to large habitation sites may indicate that suitable land for these activities could have played an important role in the final site selection for a new population center. As mentioned, the survey recorded a series of large, densely occupied residential sites in the northern part of the survey region along a very narrow knife ridge, which is somewhat distant (5–7 km) from the main

cluster of agro-pastoral subsistence structures (e.g. corrals, terrace/corral complexes, terraces) located in an undulating section of moderately sloped terrain. As such, this may simply reflect that populations lived in a more remote, higher-relief terrain for protection. At the same time, daily agro-pastoral activities appear to have occurred farther away than preferred—largely due to local topography, which could not support concurrent defensive needs (i.e. steep terrain) and subsistence needs (i.e. adequate grazing and farming land).

Conclusion This chapter has employed a settlement ecology approach to disentangle settlement decisionmaking processes of the Chanka. More specifically, I have identified a series of convergent environmental and social risks and then translated them into a set of specific settlement priorities. By comparing a set of settlement expectations and observations, we have been able to clarify those settlement priorities that underlay the settlement ecology of the Chanka during the tumultuous LIP. The analysis reveals that the risks of attack by neighboring groups galvanized the site populations to make defensibility (natural and man-made) one of the priorities—if not the top priority—for new settlement location and construction. A new hilltop settlement location also needed available masonry stone for construction, but to a lesser degree a nearby water source and suitable land for agro-pastoral activities. This analysis in combination with recent skeletal trauma data (Kurin 2012) contrasts with an earlier hypothesis that warfare was less of a concern for the Chanka and that site location decisions for them were linked more directly with climate change and land use (Kellett 2010). In sum, Chanka settlement ecology was designed on a foundation of overlapping and convergent social, political, and environmental risks; however, despite these enduring risks to daily life, the Chanka found the opportunity and the capacity not only to survive, but to care for one another (Jolly and Kurin 2015), and thrive for over four centuries until the arrival of the Inka in the early fifteenth century AD. Finally, it is my hope that this chapter has, in particular, illuminated the important part that the settlement ecology approach can play in understanding prehistoric societies that were in major transition. By using settlement as a dependent variable that is contingent on a multitude of socio-environmental influences, archaeologists can unpack the complex decision-making that occurred in times of cultural transition. This approach also holds promise for better understanding of the “lived experiences” of ancient populations when confronted with periods of settlement change (e.g. abandonment, migration, colonization/resettlement)—an especially timely topic given the major demographic and settlement changes occurring around the world in the face of numerous and rapidly changing cultural, societal, and ecological conditions.

References Abbott, M., M. Binford, M. Brenner, and K. Kelts 1997a A 3500 14C yr High-Resolution Record of Water Level Changes in Lake Titicaca, Bolivia-Peru. Quaternary Research 47: 169–180. Abbot, M., G. Seltzer, K. Kelts, and J. Southon 1997b Holocene Paleohydrology of the Tropical Andes from Lake Records. Quaternary Research 47: 70–80.

Arkush, Elizabeth N. 2006 The Transformation of Warfare in the Late Prehispanic Andean Highlands. In The Archaeology of Warfare: Prehistories of Raiding and Conquest, edited by Elizabeth Arkush and Mark Allen, pp. 286–335. University Press of Florida, Gainesville. Arkush, Elizabeth N. 2008 War, Chronology, and Causality in the Titicaca Basin. Latin American Antiquity 19(4): 339–373. Arkush, Elizabeth N. 2011 Hillforts of the Ancient Andes: Colla Warfare, Society and Landscape. University of Florida Press, Gainesville. Arkush, Elizabeth N. and Tiffany Tung 2013 Patterns of War in the Andes from the Archaic to the Late Horizon: Insights from Settlement Patterns and Cranial Trauma. Journal of Archaeological Research 21: 307–369. Bauer, Brian S. and Lucas C. Kellett 2010 Cultural Transformations of the Chanka Heartland (Andahuaylas, Peru) during the Late Intermediate Period (AD 1000–1400). Latin American Antiquity 21(1): 87–111. Bauer, Brian S., Lucas C. Kellett, and Miriam Aráoz 2010 The Chanka: Archaeological Research in Andahuaylas (Apurimac), Peru. The Cotsen Institute of Archaeology, Los Angeles. Bawden, Garth and M. Reycraft 2000 Environmental Disaster and the Archaeology of Human Response, pp. 163–178. Maxwell Museum of Anthropology Papers No. 7. Betanzos, Juan de 1987 [1551] Suma y Narración de los Incas. Prologo, transcripción y notas por M. del Carmen Martín Rubio. Ediciones Atlas, Madrid. Binford, M., A. Kolata, M. Brenner, J. Janusek, M. Seddon, M. Abbot, and J. Curtis 1997 Climate Variation and the Rise and Fall of an Andean Civilization. Quaternary Research 47: 235–248. Bird, Douglas W., and James F. O’Connell 2006 Behavioral Ecology and Archaeology. Journal of Archaeological Research 14(2): 143–188. Browman, David 1987 Agro-Pastoral Risk Management in the Andes. Research in Economic Anthropology 8: 171–200. Cashdan, Elizabeth A. (editor) 1990 Risk and Uncertainty in Tribal and Peasant Economies. Westview Press, Boulder. Chepstow-Lusty, A. 2011 Agro-Pastoralism and Social Change in the Cuzco Heartland of Peru: A Brief History Using Environmental Proxies. Antiquity 85: 570–582. Chepstow-Lusty, Alex, M.R. Frogley, B.S. Bauer, M.J. Leng, K.P. Boessenkool, C. Carcaillet, A.A. Ali, and A. Gioda 2009 Putting the Rise of the Inca Empire within a Climate and Land Management Context. Climate of the Past 5: 375–388. Cieza de León, Pedro de 1984 [1553] La Crónica del Perú. Instituto Gonzalo Fernández de Oviedo, Madrid. Conlee, C.A., J. Dulanto, C.J. Mackey, and C. Stanish 2004 Late Prehispanic Sociopolitical Complexity. In Andean Archaeology, edited by H. Silverman, pp. 209–236. Blackwell, Malden, MA. Conrad, Geoffrey 1978 Models of Compromise in Settlement Pattern Studies: An Example from Coastal Peru. World Archaeology 9(3): 281–298. Contreras, Daniel A. 2010 Landscape and Environment: Insights from the Prehispanic Central Andes. J. Archaeological Research 18: 241–288. Covey, R. Alan 2008 Multiregional Perspectives on the Archaeology of the Andes during the Late Intermediate Period (c.AD 1000–1400). Journal of Archaeological Research 16: 287–338. D’Altroy, Terence D., and Christine T. Hastorf, (editors) 2001 Empire and Domestic Economy. Kluwer Academic/Plenum Publishers, New York. Dean, Jeffrey S. 1996 Demography, Environment, and Subsistence Stress. In Evolving Complexity and Environmental Risk in the Prehistoric Southwest, edited by J.A. Tainter and B.B. Tainter, pp. 25–56. Addison-Wesley, Reading, MA. deMenocal, Peter D. 2001 Cultural Responses to Climate Change during the Late Holocene. Science 292(5517): 667–673. Diamond, Jared 2011 Collapse. How Societies Choose to Fail or Succeed. Revised edition. Penguin Books, New York. Dillehay, Thomas D., and Alan L. Kolata 2004 Long-Term Human Response to Uncertain Environmental Condition in the Andes. PNAS 101(12): 4325–4330. Earle, Timothy, and Andrew Christenson 1980 Modeling Change in Prehistoric Subsistence Economies. Academic Press, New York. Elliott, Michelle 2005 Evaluating Evidence for Warfare and Environmental Stress in Settlement Pattern Data from the Malpaso Valley, Zacatecas, Mexico. Journal of Anthropological Archaeology 24: 297–315. Finucane, Brian C. 2009 Maize and Sociopolitical Complexity in the Ayacucho Valley, Peru. Current Anthropology 50(4): 535– 545. Finucane, Brian C., Patricia M. Agurto, and William H. Isbell 2006 Human and Animal Diet at Conchopata, Peru: Stable Isotope Evidence for Maize Agriculture and Animal Management Practices during the Middle Horizon. Journal of Archaeological Science 33: 1766–1776. Fisher, Christopher T., J. Brett Hill, and Gary M. Feinman (editors) 2009 The Archaeology of Environmental Change: Socionatural Legacies of Degradation and Resilience. University of Arizona Press, Tucson. Garcilaso de la Vega, Inca 1966 Royal Commentaries of the Incas and General History of Peru, Parts 1 and 2 [1609]. Translated by H.V. Livermore. University of Texas Press, Austin. Goland, Carol 1993 Field Scattering as Agricultural Risk Management: A Case Study from Cuyo Cuyo, Department of Puno, Peru. Mountain Research and Development 13: 317–338.

Goldstein, Paul 2004 Andean Diaspora: The Tiwanaku Colonies and the Origins of South American Empire. University Press of Florida, Gainesville. Halstead, P., and J. O’Shea (editors) 1989 Bad Year Economics: Cultural Responses to Risk and Uncertainty. Cambridge University Press, Cambridge. Hasenstab, Robert J. 1996 Settlement as Adaptation: Variability in Iroquois Village Site Selection as Inferred Through GIS. In New Methods, Old Problems: Geographic Information Systems in Modern Archaeological Research, edited by H. Maschner, pp. 223–241. Center for Archaeological Investigations, Southern Illinois University, Carbondale. Hastorf, Christine A. 1993 Agriculture and the Onset of Political Inequality before the Inka. Cambridge University Press, Cambridge. Hillyer, R., B.G. Valencia, M.B. Bush, M.R. Silman, and M. Steinitz-Kannan 2009 A 24,700-yr Paleolimnological History from the Peruvian Andes. Quaternary Research 71: 71–82. Janusek, John W. 2008 Ancient Tiwanaku. Cambridge University Press, Cambridge. Jolly, Sarah, and Danielle S. Kurin 2015 Surviving Trepanation: Approaching the Relationship of Violence and the Care of “War Wounds” through a Case Study in Prehistoric Peru. Paper presented at the 75 th Annual Meeting of the Society for American Archaeology, San Francisco, CA. Jones, Eric E. 2010 An Analysis of Factors Influencing Sixteenth and Seventeenth Century Haudenosaunee (Iroquois) Settlement Locations. Journal of Anthropological Archaeology 29: 1–14. Jones, Eric E., Madison Gattis, Andrew Wardner, Thomas C. Morrison, and Sara Frantz 2012 Exploring Prehistoric Tribal Settlement Ecology in the Southeast: A Case Study from the North Carolina Piedmont. North American Archaeologist 33(2): 157–190. Isbell, William H. 1988 City and State in Middle Horizon Hauri. In Peruvian Prehistory, edited by R.W. Keatinge, pp. 164–189. Cambridge University Press, Cambridge. Isbell, William H., and Gordon F. McEwan (editors) 1991 Huari Administrative Structure: Prehistoric Monumental Architecture and State Government. Dumbarton Oaks Research Library and Collection, Washington, DC. Jennings, Justin 2006 Core, Peripheries, and Regional Realities in Middle Horizon Peru. Journal of Anthropological Archaeology 25: 346–370. Jennings, Justin, and Nathan Craig 2001 Politywide Analysis and Imperial Political Economy: The Relationship between Valley Political Complexity and Administrative Centers in the Wari Empire of the Central Andes. Journal of Anthropological Archaeology 20: 479–502. Julien, Catherine 2002 Diego Maldonado y los Chancas. Revista Andina 34, enero. Centro de Estudios Regionales Andinos, Bartolomé de Las Casas, Cusco. Kellett, Lucas C. 2010 Chanka Settlement Ecology: Hilltop Sites, Land Use and Warfare in Late Prehispanic Andahuaylas, Peru. Unpublished Ph.D. dissertation. Department of Anthropology, University of New Mexico, Albuquerque, NM. Kellett, Lucas C. 2011 The Domestic Economy of Anccopaccha: The Wari Occupation of the Andahuaylas Region (Apurímac) Perú. Paper presented in the symposium “Domestic Responses, Challenges, and Cooperation in Andean State Development.” 76 th Annual Meeting of the Society for American Archaeology, Sacramento, CA. Kellett, Lucas C. 2013a Collapse, Climate Change and Conflict: A Profile of Crisis and Response in Late Andean Highland Prehistory. In Tough Times: The Archaeology of Crisis and Recovery. Proceedings of the Graduate Archaeology at Oxford Conferences 2010 and 2011, edited by E. Van der Wilt and J.M. Jiménez, pp. 43–55. BAR International Series 2478. Archaeopress, Oxford. Kellett, Lucas C. 2013b Cambios Climáticos, Riesgo y Agropastoralismo en el Territorio Chanka. In Investigaciones Arqueológicas y Antropológicas en los Andes Sud-Centrales: Historia, Cultura y Sociedad, edited by Danielle Kurin and Enmanuel Gomez Choque, pp. 83–109. Fondo Editorial de Investigación, Creación Intelectual y Artística, Universidad Nacional José Maria Arquedas. Andahuaylas, Apurímac, Perú. Kellett, Lucas C., Sarah Jolly, Danielle S. Kurin, and Guni Monteagudo 2015a Life at Achanchi: A High Altitude Burial Site from the Andahuaylas Region of Southern Peru. Poster presented at the 75 th Annual Meeting for the Society of American Archaeology, San Francisco, CA. Kellett, Lucas C., Bryan Valencia, Mark Bush and Brian S. Bauer 2015b Climate Effects on the Long Term Settlement Ecology of the Andahuaylas Region of Southern Highland Peru. Poster presented at the 21 st Annual Meeting of the European Association of Archaeologists, Glasgow, UK. Kohler, Timothy A., Mark D. Varien, and Aaron M. Wright (editors) 2011 Leaving Mesa Verde: Peril and Change in the Thirteenth-Century Southwest. University of Arizona Press, Tucson. Kolata, Alan L. 1993 The Tiwanaku: Portrait of an Andean Civilization. Blackwell Publishers, Cambridge, MA. Kolata, Alan L. 2000 Environmental Thresholds and the “Natural History” of Andean Civilization. In Environmental Disaster and the Archaeology of Human Response, edited by Garth Bawden and M. Reycraft, pp. 163–178. Maxwell Museum of Anthropology Papers No. 7. Kolata, Alan L. (editor) 1996 Tiwanaku and Its Hinterland: Archaeology and Paleoecology of an Andean Civilization,

Volumes 1 & 2. Smithsonian Institution, Washington, DC. Kurin, Danielle S. 2012 The Bioarchaeology of Collapse: Ethnogenesis and Ethnocide in Post-Imperial Andahuaylas, Peru (AD 900–1250). Unpublished Ph.D. dissertation. Department of Anthropology, Vanderbilt University, Nashville, TN. Kurin, Danielle S. 2013 Trepanation in South-Central Peru during the Early Late Intermediate Period (ca. AD 1000–1250). American Journal of Physical Anthropology 152(4): 484–494. Kurin, Danielle S., E.M. Lofaro, D.E. Gomez Choque, and J. Krigbaum 2014 A Bioarchaeological and Biogeochemical Study of Warfare in Andahuaylas, Peru. International Journal of Osteoarchaeology 26: 93–106. Kuznar, Lawrence 2001 Risk Sensitivity and Value among Andean Pastoralists: Measures, Models and Empirical Tests. Current Anthropology 42(3): 432–440. LeBlanc, Steven A., and Glenn E. Rice (editors) 2001 Deadly Landscapes: Case Studies in Prehistoric Southwestern Warfare. University of Utah, Salt Lake City. McAnany, P.A., and N. Yoffee (editors) 2009 Questioning Collapse: Human Resilience, Ecological Vulnerability, and the Aftermath of Empire. Cambridge University Press, Cambridge. McIntosh, Roderick, Joseph A. Tainter, and Susan Keech McIntosh 2000 The Way the Wind Blows: Climate, History and Human Action. Columbia University Press, New York. Marston, John M. 2011 Archaeological Markers of Agricultural Risk Management. Journal of Anthropological Archaeology 302(2): 190–205. Maslow, A.H. 1943 A Theory of Human Motivation. Psychological Review 50(4): 370–396. Moseley, Michael E. 1997 Climate, Culture and Punctuated Change: New Data, New Challenges. The Review of Archaeology 18(1): 19–27. Moseley, Michael E. 1999 Convergent Catastrophe: Past Patterns and Future Implications of Collateral Natural Disasters in the Andes. In The Angry Earth: Disaster in Anthropological Perspective, edited by A. Oliver-Smith and S.M. Hoffman, pp. 59–71. New York, Routledge. Nichols, Deborah L. 1987 Risk and Agricultural Intensification during the Formative Period in the Northern Basin of Mexico. American Anthropologist 89(3): 596–616. Ortloff, Charles R., and Alan L. Kolata 1993 Climate and Collapse: Agro-Ecological Perspectives on the Decline of the Tiwanaku State. Journal of Archaeological Science 20: 195–221. Parsons, Jeffrey R., and Charles M. Hastings 1988 The Late Intermediate Period. In Peruvian Prehistory, edited by Robert W. Keatinge, pp. 190–229. Cambridge University Press, Cambridge. Parsons, Jeffrey R., Charles M. Hastings, and Ramiro Matos M. 1997 Rebuilding the State in Highland Peru: Herder-Cultivator Interaction during the Late Intermediate Period in the Tarama-Chinchaycocha Region. Latin American Antiquity 8(4): 317– 341. Parsons, Jeffrey R., Charles M. Hastings, and Ramiro Matos M. 2000 Prehispanic Settlement Patterns in the Upper Mantaro and Tarama Drainages, Junín, Peru. Volume 1, Part 1, The Tarama-Chinchaycocha Region. Vol. 34, Memoirs of the Museum of Anthropology. University of Michigan, Ann Arbor. Parsons, Jeffrey R., Charles M. Hastings, and Ramiro Matos M. 2013 Prehispanic Settlement Patterns in the Upper Mantaro and Tarama Drainages, Junín, Peru. Volume 2, The Wanka Region. Vol. 53, Memoirs of the Museum of Anthropology. University of Michigan, Ann Arbor. Redman, Charles 1999 Human Impacts on Ancient Environments. University of Arizona Press, Tucson. Redman, Charles, Steven R. James, Paul R. Fish, and J. Daniel Rogers 2004 The Archaeology of Global Change: The Impact of Humans on Their Environment. Smithsonian Books, Washington, DC. Rowe, John H. 1946 Inca Culture at the Time of Spanish Contact. In Handbook of South American Indians, edited by Julian H. Steward, pp. 182–330. Bureau of American Ethnology Bulletin 143. Smithsonian Institution, Washington, DC. Sandweiss, Daniel H., and Alice R. Kelly 2012 Archaeological Contributions to Climate Change Research: The Archaeological Record as a Paleoclimatic and Paleoclimatic and Paleoenvironmental Archive. Annual Review of Anthropology 41: 371– 391. Scarborough, Vernon L. 2000 Resilience, Resource Use, and Socioeconomic Organization: A Mesoamerican Pathway. In Natural Disasters and the Archaeology of Human Response, edited by Richard Reycraft and Garth L. Bawden, pp. 195– 212. University of New Mexico Press, Albuquerque. Schollmeyer, Karen 2011 Large Game, Agricultural Land and Settlement Pattern Change in the Eastern Mimbres Area, Southwest New Mexico. Journal of Anthropological Archaeology 30(3): 402–415. Schreiber, Katherina 1992 Wari Imperialism in Middle Horizon Peru. No. 87, Anthropological Papers. Museum of Anthropology, University of Michigan, Ann Arbor. Seltzer, Geoffrey O. and Christine A. Hastorf 1990 Climatic Change and its Effect on Prehispanic Agriculture in the Central Peruvian Andes. Journal of Field Archaeology 17: 397–414. Stanish, Charles 2003 Ancient Titicaca: The Evolution of Complex Society in Southern Peru and Northern Bolivia. University of California Press, Berkeley. Stone, Glenn Davis 1996 Settlement Ecology: The Social and Spatial Organization of Kofyar Agriculture. Arizona Studies

in Human Ecology. University of Arizona Press, Tucson. Sublette, Mosblech N.A, A. Chepstow-Lusty, B.G. Valencia, and M.B. Bush 2012 Anthropogenic Control of Late Holocene Landscapes in the Cuzco Region, Peru. The Holocene June: 1–13. Thompson, Lonnie G., E. Moseley-Thompson, J.F. Bolzan, and B.R. Koci 1985 A 1500-Year Record of Tropical Precipitation in Ice Cores from the Quelccaya Ice Cap, Peru. Science 229: 971–973. Valencia, B.G., D.H. Urrego, M.R. Silman, and M.B. Bush 2010 From Ice Age to Modern: A Record of Landscape Change in an Andean Cloud Forest. Journal of Biogeography, 9: 1–11. VanDerwarker, Amber M., and Gregory D. Wilson 2015 The Archaeology of Food and Warfare: Food Insecurity in Prehistory. Springer, New York. Wernke, Steven A. 2013 Negotiated Settlement: Andean Communities and Landscapes under Inka and Spanish Colonialism. University of Florida Press, Gainesville. Williams, Ryan Patrick 2002 Rethinking Disaster-Induced Collapse in the Demise of the Andean Highland States: Wari and Tiwanaku. World Archaeology 33(3): 361–374. Williams, Ryan Patrick 2006 Agricultural Innovation, Intensification and Sociopolitical Development: The Case of Highland Irrigation Agriculture on the Pacific Andean Watersheds. In Agricultural Strategies, edited by J. Marcus and C. Stanish, pp. 309–333. Cotsen Institute of Archaeology, UCLA, Los Angeles. Winterhalder, Bruce 1986 Diet Choice, Risk and Food Sharing in a Stochastic Environment. Journal of Anthropological Archaeology 5: 369–392. Winterhalder, Bruce 1990 Open Field and Common Pot: Harvest Variability and Risk Avoidance in Agricultural and Foraging Societies. In Risk and Uncertainty in Tribal and Peasant Economies, edited by E. Cashdan, pp. 67–87. Westview Press, Boulder. Zaro, Gregory 2007 Diversity Specialists: Coastal Resource Management and Historical Contingency in the Osmore Desert of Southern Peru. Latin American Antiquity 18(2): 161–179. Zaro, Gregory, and Adán Umire-Alvarez 2005 Late Chiribaya Agriculture and Risk Management along the Arid Andean Coast of Southern Peru, AD 1200–1400. Geoarchaeology: An International Journal 20(7): 717–737. Zori, Colleen, and Erica Brant 2012 Managing the Risk of Climate Variability in Late Prehistoric Northern Chile. Journal of Anthropological Archaeology 31: 403–418.

10 Encountering forgotten landscapes Water, climate, and two millennia of settlement location choices in the Ica-Nasca region of southern coastal Peru Hendrik Van Gijseghem Introduction The Ica-Nasca region of the south coast of Peru, as part of the Northern Atacama Desert, is one of the most arid locations on the planet. Like most of the central Andean coast, this area is bisected by rivers that seasonally drain the Andean highlands and bring water for coastal irrigation. The arrival of water in these highly variable rivers is therefore vital to people’s livelihood. Today, populations concentrate along these thin strips of fertile land, and it is understandable that settlement surveys exploring past demography have also focused on river valleys and adjacent foothills. The assumption that past populations likewise would have selected river valleys for settlement is reasonable and has been confirmed by extensive surveys over the past half-century. However, the increasingly sophisticated techniques that create environmental reconstructions have recently revealed potentially significant climatic variations in the past that may have affected past hydrological conditions, resources, and human settlement location choices and opportunities. In this chapter I explore this possibility and present survey results that complement recent environmental reconstructions in indicating that the south coastal environment has not been constant and that modern conditions are a relatively recent phenomenon. As a result, south coastal populations have been faced with short- and long-term environmental changes and crises, and these challenges must be considered in examining past socio-political changes. Note that I am not reducing settlement location choices to environmental factors, but suggesting that through time these constituted shifting constraints on a dynamic settlement ecology. Various socio-environmental criteria for choosing settlement locations—in addition to subsistence imperatives—included, at different times: cycles of conflict and alliances, participation in religious movements, culturally constituted landscape perceptions, mass migrations, emergent leadership, and population growth.

South coastal environment and settlement The largest and most productive river in the Ica-Nasca region is the Ica River. It often has some measure of water throughout the year. As a general rule, however, as one moves south into the Nasca region, valleys’ oases become narrower, and the rivers are less reliable and more unpredictable, which presumably would have caused considerable strain on prehispanic

agricultural productivity. Nonetheless, populations have occupied these regions since at least the Archaic period, or the fourth millennium BC (Vaughn and Linares 2006; Isla 1990; Van Gijseghem 2004) (Figure 10.1; Table 10.1). Settlement surveys in the Ica-Nasca region performed in the past few decades reveal similar patterns through time (Reindel 2009; Schreiber and Lancho 2003; Silverman 2002; Massey 1986; Cook 1999). During the Formative Period (sometimes referred to as the Early Horizon, c.800 BC–AD 90), people of the Paracas culture occupied the landscape in relatively low densities, but in the Ica region there was a gradual pattern of population growth and the increasing presence of large population and worship centers in the valley’s several oasis-like basins. In the Nasca region, settlement locations tended to focus on steep, low hills, and the ubiquity of sling stones indicates that this location choice had defensibility as an objective. Settlement patterns suggest that conflict was endemic in the Formative Period (Van Gijseghem 2006; Van Gijseghem and Vaughn 2008), a phenomenon that was widespread in many Andean regions, especially toward the end of the period (Arkush and Tung 2013:325).

Figure 10.1

The Ica-Nasca region with sites mentioned in the text.

The transition between Paracas and Nasca is characterized by a short period of drastic cultural changes alternately known as Initial Nasca, Nasca 1, or Proto-Nasca, at the very end of the Formative Period. During this time, the number of settlements rises markedly in most regions (Reindel 2009; Schreiber and Lancho 2003; Silverman 2002), and settlement patterns are more diverse. These changes are accompanied by the establishment, in the Nasca region, of important civic-ceremonial centers, such as Cahuachi, in the southern Nasca region (Silverman 1993; Orefici 1988, 1996; Bachir 2007). Through its size and architectural elaboration,

Cahuachi became the paramount center on the south coast, but others of smaller size were established in the subsequent Early Nasca period (c. AD 90–325), such as Los Molinos in the Palpa region (Reindel and Isla 2001, 2006) and Cerro Tortolita in the Upper Ica Valley (Massey 1986). The Early Nasca period saw a continuation of settlement pattern changes, away from steep hills and toward flatter terrain alongside the river bottoms, testifying perhaps to a decrease in conflict attributed to new social and ceremonial realities promoted by Cahuachi through pilgrimages and cooperation (Kantner and Vaughn 2012). As the influence of Cahuachi waned during the Middle Nasca period (c. AD 325–440), and increasingly so during the Late Nasca period (c.440–620), a pattern of gradual population nucleation culminated with the presence of one large settlement in each river valley, which has been interpreted as regional political fragmentation (Schreiber and Lancho 2003). It is also during the Middle Nasca period that puquios, aqueducts that bring water from shallow aquifers to the surface and greatly improve conditions for irrigation, began to be constructed along the southern region’s unproductive river basins (Schreiber and Lancho 2003, see also Conlee 2014; Orefici 2011). This innovation allowed populations to circumvent many of the vagaries of seasonal river behaviors and benefit from more reliable water supply. The northern region also sees population movements and high population densities, as testified by the number and size of settlements (Reindel 2009). Table 10.1

General recent chronology for the Ica-Nasca region. The dates are those summarized in Reindel et al. 2009.


Chronological period

South coast culture

AD 1400–1532

Late Horizon


AD 1000–1400

Late Intermediate Period


AD 620–1000

Middle Horizon


AD 440–620

Early Intermediate Period

Late Nasca

AD 325–440

Middle Nasca

AD 90–325

Early Nasca

120 BC–AD 90

Formative Period

800–200 AD

Initial Nasca Paracas

During the Middle Horizon (c.AD 620–1000), a foreign superpower, the highland Wari, encroached on the Nasca region (Schreiber 1999; Edwards 2013; Edwards and Schreiber 2014) and affected population distribution considerably. As a result of circumstances that are not well understood, a near-complete depopulation of the south coast marks the beginning of the Late Intermediate Period (c.1000–1450; Conlee and Schreiber 2006; Conlee et al. 2004). However, in the following centuries, populations once again reached unprecedented numbers (Conlee 2003) and seemingly grew until Inka invasion, followed by the end of the prehispanic period, when they were decimated, mainly by European-borne diseases.

South coastal climatic patterns and subsistence In the past decade, archaeologists and investigators from diverse fields have applied new

technologies and approaches to the reconstruction of the prehistoric central Andean climate and examined how climatic trends affected social complexity and settlement systems (Contreras 2010; Eitel and Mächtle 2009; Eitel et al. 2005; Fehren-Schmitz et al. 2014; Mächtle et al. 2009; Mächtle and Eitel 2013; Unkel et al 2007). The results from these climate, moisture, rainfall, and vegetation reconstructions agree that the past two millennia ushered in the environmental conditions we see today, but that important short- and medium-term variations existed. Data from ice cores (Thompson et al. 1985, 1995), lake sediments (Abbott et al. 1997), and desert loess sediments (Eitel et al. 2005; Mächtle and Eitel 2013) indicate, for instance, that on the south coast moisture was more important during the Early Horizon (Paracas culture), and the first part of the Early Intermediate Period (EIP; Nasca culture), followed by gradual aridification and a brief resurgence of moisture during the Late Intermediate Period (LIP), before the onset of aridity conditions close to those of the present day. In the Nasca region, this water supply from river flooding was complemented by the irrigation galleries known locally as puquios, which channel water from underground aquifers to the surface. Settlement associations with these continually used features indicate that many date from the latter part of the Early Intermediate Period (Schreiber and Lancho 2003), possibly around the time of the onset of modern climate conditions. This method of irrigation has historically been sufficient in most years to insure a reliable yield, given the variability in highland rainfall and consequent coastal flooding. Basically, today, the available water comes from the highlands, whether above or below ground. The practices and processes that enabled ancient populations to produce agricultural goods have seldom been considered in the Ica-Nasca regions (but see Beresford-Jones et al. 2009, 2011). The destruction of ancient fields and irrigation systems by modern intensive agriculture constitutes a serious limiting factor to the study of ancient agricultural practices. By and large, this is because modern agriculture is limited to the very same locations where much ancient farming took place: the flat valley bottoms (Silverman 2002:32). Indeed, south coastal geomorphological conditions limit agricultural potential to inland basins instead of the wide alluvial fans typical of Andean coastal valleys located further north (Moseley and Deeds 1982; Farrington and Park 1978; Ortloff et al. 1982; Ertsen and van der Spek 2009). Therefore most of these basins were continually reoccupied over time, critically altering the remains of more ancient agricultural operations. If during some periods south coastal conditions were more humid, as the data indicate, the questions to be raised concern the impact of this more abundant hydrological regime on agricultural practices. What exactly does more moisture mean? More water flooding the rivers in the austral spring and summer? More fog bringing humidity to low desert hills known as lomas? Is it possible that the annual river floods were not the only potential source of field irrigation? If Nasca populations were using non-river runoff irrigation for agriculture, how common was it? It is an often implicit notion that the challenges involved in bringing water to these basins remained constant through time. This assumption is reasonable since, as Rowe notably pointed out about the Ica river basins: “one good soaking of the land a year was enough to ensure a crop of maize and squash” (Rowe 1963:9). Some form of flood-based, canal-assisted, or

puquio-assisted irrigation has historically been sufficient in most years to insure a reliable yield. It was therefore presumed to have been the main irrigation method in the past, even if the coastal soils—composed of Tertiary terrigenous and sandy marine sediments, and aeolian loess—are relatively poor (Eitel et al. 2005). Reindel (2009:451–452) considers that based on settlement location, Early Nasca society must have developed an important irrigation system in the Palpa region of the northern Nasca region (see also Hesse and Baade 2009). However, according to Reindel no remains of this system survive, nor have they been identified in Silverman’s Ingenio Valley survey (Silverman 2002:150), but fossil canals seemingly dating to the Early Nasca period have been identified in Ica (Beresford-Jones et al. 2009). Few alternatives to a Nasca agriculture based on annual river floods have been considered (but see Mächtle et al. 2009 for evidence of runoff water harvesting during the Late Intermediate Period). However, as I outline below, recent palaeoclimatic reconstructions and remote sensing could reorient scholarly attention toward alternative models and research areas. Using archaeological data, I suggest that the moisture signatures detected by other researchers indicate that during some periods it was regularly raining on the south coast. While the effects of water erosion are evident on the south coast, it is generally thought to be the result of episodic and infrequent El Niño events. These data indicate that as a result of regular instead of purely sporadic rainfall, our vision of the south coastal landscape during these periods of high moisture should be drastically altered. One reason is that, at least implicitly, our perception of past landscapes tends to conform to the present landscape. Accordingly, settlement surveys have traditionally been performed, and justifiably so, along existing river valleys, and typically up to about 100 m on either side of existing channels along the low Andean foothills. However, if we imagine even modest precipitation falling over this landscape rather than it being fed solely by runoff from highland precipitations, the picture changes considerably: basically, new rivers in the desert and the foothills appear on an otherwise dry and sterile land. It follows from this hypothesis that expanding surveys beyond currently occupied areas could complement and significantly alter traditionally held perspectives on prehispanic settlement, landscape, and subsistence organization in the Nasca region. Consequently, I have started paying more attention to those desert quebradas, or dry gullies, which had not necessarily been surveyed. Because of local geomorphological characteristics, some landforms theoretically should have accumulated considerable runoff if precipitations had fallen upon the landscape and should have been productive for human settlement and subsistence.

Archaeological revelations in forgotten landscapes The first discoveries were unintended, while a small team and I were engaged in a survey of ancient mining locations that brought us away from traditionally surveyed river valleys into dry foothills and quebradas (Van Gijseghem et al. 2011; Van Gijseghem et al. 2013). I first present three of those chance encounters, which allowed the establishment of a baseline perspective on the potential of some landforms for runoff irrigation. Accordingly, if a certain watershed

allows for runoff irrigation at certain times, theoretically all watersheds of equal or larger size should also be potentially productive. I then use this information to present more systematic survey results in the subsequent section. Quebrada La Yesera Near the Ica Valley, in a barren but large quebrada east of the valley, some 7 km away from modern and ancient settlements as conventionally understood, my team stumbled across what appeared to be an ancient agricultural complex. It is composed of a total of approximately 5 ha vestigial furrows connected by stone-lined irrigation canals and excavated ditches intended for primary and secondary irrigation (Figure 10.2). Such features do not generally yield many datable artifacts. Some ceramic pieces were encountered among looted tombs whose placement appears to have been meant to avoid the irrigated areas, suggesting their contemporaneity. All such remains date to the Late Intermediate Period (c.AD 1000–1400), but some isolated Nasca 1 sherds were also identified within the quebrada. One of the characteristics of this complex is the presence of a single-intake canal responsible for the irrigation of the entire network. Also significant is the variability in the size and organization of furrows, suggesting that different species of plants were grown here. This includes a small plot, little more than what could be described as a small garden, located near the canal intake, with a dozen tiny, delicate furrows (Figure 10.2b). This is the first place water would have flowed once collected by the feeder canal. Soil analyses are pending, but plant species of some significance may have been grown there. A consideration of the regional context in which we find this agricultural complex reveals that it lies at the convergence of the largest network of quebradas in this immediate massif of the Andean foothills. This location is optimal for harnessing the maximum quantity of rainfall runoff, with the network of quebradas effectively acting as a massive funnel, progressively bringing water toward the point where the canal intake is strategically located. This is a closed catchment area, meaning that it does not have any confluent streams originating in the nearby highlands. It covers an area of approximately 1000 ha, or a little more than 10 sq km. If moderate rainfall could potentially bring sufficient water to the network, in addition to locally falling rain contributing to soil moisture over a reasonably small area, then similar landforms undoubtedly could provide irrigation water to other suitable areas. This indicates that, at times, rain fell on this part of the south coast in enough quantity and with sufficient regularity to warrant the establishment of this field system.

Figure 10.2

Agricultural infrastructure in Quebrada La Yesera. Air photograph of preserved furrows (a); small garden-like plot near the canal intake (b); stone-lined canal (c).

We noticed, however, the complete absence of local settlements associated with this complex. One very badly destroyed structure that may have had two rooms was encountered nearby. It may have been a temporary shelter. This leads me to suggest the possibility that this is not a regular or habitual agricultural infrastructure, but that it may have been established to take advantage of the irregular precipitations caused by the erratic El Niño Southern Oscillation (ENSO) phenomenon. This practice has been documented in the nearby Palpa area during the Late Intermediate Period (Mächtle et al. 2009).

Figure 10.3

View of two successive unfinished canals in Quebrada Campanayoq near the Aja River, Nasca region.

Quebrada Campanayoq In an analog context, found in the nearby Nasca region, a series of agricultural terraces were irrigated by water emerging from the nearby hills through the massive Campanayoq quebrada, a watershed that extends up to the continental divide. Satellite photos indicate that it is replete with ancient settlements. To date, it has yet to be surveyed. Further inspection revealed two successive canals skirting the foothills above the terraces (Figure 10.3). The lower canal is incompletely preserved and it is impossible to trace its entire length. It may have been abandoned as construction on a new canal was begun, at much greater cost, which involved cutting through bedrock, approximately 8 m above the original canal, probably in an attempt to bring water to a greater area. This one was never completed, as it is interrupted in three places by bedrock that had been incompletely modified to accommodate water flow. It is unknown how long it ultimately was, as its lower part was eventually covered by rockslide, but it never in fact reached the terrace system. No datable artifacts were found in direct association with the canal, but some Nasca 1 scatters were identified nearby. From the somewhat unreliable criterion of settlement association, the canals could date to the Formative–EIP transition (Nasca 1) or to the LIP. And if, as is possible, the interruption of canal construction coincided with a nearby site’s abandonment, this would situate it in the middle part of the EIP, or the close of the Early Nasca period, during a time of increasing aridity in the period. Nonetheless, given that the only nearby sites date from the early part of the EIP and to the LIP, settlement association in one case or the other temporally situates this system in one of the humid periods in south coastal history, during which populations attempted, but ultimately failed, to take advantage of runoff from the local quebrada.

Figure 10.4

Hypothetical past hydrography of the Ica-Nasca region with some fossil rivers indicated by dotted lines.

Pataraya Chico A third and final example is located in the upper Tierras Blancas river valley, on a high terrace overlooking the river. Here a small Early Nasca settlement is located on a clifftop with hundreds of low agricultural terraces constructed around its perimeter and on the terrace above (Figure 10.4). The agricultural terraces are fed by two canals that eventually come together from opposite directions and that also drain two different quebrada systems. In contrast to the first two examples, the drainage area feeding these canals is comparatively small, at some 1.8 sq km for one watershed and 1.3 sq km for the other. These anecdotal discoveries imply that if even a moderate amount of rain was to fall on what is today an arid landscape, runoff would accumulate along the foothills and be channeled toward the existing rivers. In that regard, there exists the possibility that under different climatic conditions, dry quebradas may have had the status of rivers or streams in their own right. Under such a regime, water availability, in a sense, may have been more democratic: not flowing from a single source and direction, the highlands, but instead flowing from several directions at once, as tributary quebradas, in addition to the main rivers that drained the highlands. After all, if water was occasionally present in a sufficient quantity to irrigate the systems described above, there must be countless other similar locations. Fortunately, this is easily testable.

Survey in ICA: Quebradas Cocharcas (Yauca) and Tingue A cursory examination of landforms compared with already known cases does reveal large quebradas that would have received runoff water at regular intervals. Starting with newly upgraded Google Earth images from the south coast, I targeted two of these large quebrada systems to test the hypothesis that they may have been productive river valleys under more humid conditions. The selected quebradas are Tingue and Yauca/Cocharcas, both draining in the wide plain of the middle Ica valley. They were in part selected because there are roads leading to the highlands, which would later facilitate the field survey that I carried out the following year. Some results are shown in Figure 10.5. From the outset, it was clear that a large number of unrecorded settlements were present in these former rivers. Even more impressive was the fact that, because they were largely abandoned and rendered unproductive (some limited areas are farmed today with hydraulic pumps), many of the sites were ancient agricultural infrastructures that had not been altered by more modern agriculture and irrigation—as is the norm for the better-known valley bottoms of the south coast.

Cocharcas This is the northernmost quebrada that was surveyed. Only two settlements were encountered, but both are large enough to have been considered important population centers during their

time of occupation. Both show evidence for LIP occupations, with faint evidence for other components ranging from Nasca 1 to later parts of the EIP. In addition, networks of agricultural terracing were also recorded. Chokoltaja This site was originally recorded by Williams and Pazos (1974) in the 1970s. Located near the confluence of quebrada Cocharcas and the middle Ica Valley, it is a hilltop fortress dating to the LIP, although it has occupations from earlier time periods at the bottom of the hill in a nonfortified area (Figure 10.6). The fortified settlement is built atop a ridge surrounded by steep ravines that are essentially unclimbable and reinforced by buttresses. The only access to the site is at the end of a narrow pathway running alongside a steep incline that leaves any intruder entirely exposed and helpless for more than 100 m. The large quantities of sling stones found at the entrance to the site, facing the path that runs perpendicular to it, indicate that this was a strategic setup. For all intents and purposes, this site is impregnable and may testify to LIP internecine conflict or perhaps to momentary local resistance to Inka conquest. The fortified settlement is 1.5 ha in area and the overall site covers more than 17 ha.

Figure 10.5

Settlement survey results in the Cocharcas and Tingue quebradas.

CO-2 This as yet unnamed settlement on the north side of the quebrada Cocharcas covers an area of approximately 13 ha. It is built on the steep foothill slopes where hundreds of architectural terraces with buildings of dry masonry sprawl. Surface material indicates that the site also

dates to the LIP, but a small quantity of Nasca 1 material was encountered, minimally suggesting that there was some occupation of the quebrada during that time, undoubtedly overshadowed by the more imposing and recent LIP settlement. The bottom part of the site is composed of terrace systems that appear ancient and associated with the settlement, although this has yet to be demonstrated.

Figure 10.6

Well-preserved circular domestic architecture at Chokoltaja, a defensive LIP hilltop settlement, located in Quebrada Cocharcas.

CO-3 and CO-4 These two sites are agricultural systems of terraces reminiscent of highland styles of construction. They could not be dated due to the absence of surface material. They are irrigated not by the Cocharcas quebrada itself, but by side watersheds that flow into the quebrada (the watersheds feeding them are very small: CO-3’s is roughly 1 sq km; CO-4 is even smaller, and there may have been a natural spring feeding it at the time, which no longer exists). Interestingly, there are much larger watersheds in the area of CO-3 and CO-4, but none of these seem to have been the object of settlement or agriculture. Once again, it is possible that these

terrace systems were put in place to take advantage of potentially torrential ENSO downpours. This would explain the absence of associated settlements and the selection of comparatively small watersheds to avoid excess runoff that would be destructive to the structures and harder to manage because of water volume and speed. One of the challenges involved in irrigation of arid landscapes consists in slowing down the runoff, allowing it to soak the nearly impermeable desert soil. It may be for this reason that in this case, as well as in other ones surveyed, the larger quebradas may have proven inappropriate due to the large quantity of runoff they generated, especially in times of torrential rains. The presence of agricultural infrastructures at the mouths of somewhat modest quebradas unassociated with habitation settlements may represent such strategies of punctual water management in times of abnormally abundant rainfall.

Tingue Tingue is another large quebrada located 12.5 km to the south of the Cocharcas quebrada. Its chronological profile is similar, and it also reveals considerable archaeological remains in the form of habitation settlements and agricultural infrastructures. TI-1 settlement This is a small village on the north slope of the Tingue quebrada. It comprises approximately 25 stone structures spread over 1 ha. It sits on the side of the dried-up river and dates to the LIP, although here as well a small quantity of Nasca 1 material was encountered. Huarangal This is a network of low terraces, shallow furrows, and stone-lined canals spatially associated with TI-1 and preserved over 21 ha (Figure 10.7). The small amount of diagnostic ceramics recovered here suggests an LIP date. The system is irrigated not by the quebrada itself but by a confluent watershed flowing from the south and covering 14 sq km. The site may have been selected for cultivation because the soil is composed of particularly fine loess not common in other parts of the quebrada. It is unknown at this point, however, whether the site was selected because of the presence of naturally occurring fertile loess, or if the soil accumulated there because of anthropic vegetation cover (e.g. Eitel et al. 2005). In fact, little is known about south coast soil selection in agricultural decision-making and location choices.

Figure 10.7

Huarangal, Quebrada Tingue: stone-lined canal (left) and aerial view of some agricultural terraces and canals (right).

TI-3 This site is an odd orthogonal structure that, according to the few surface sherds, may date to the Early/Middle Nasca period. It is quite small, at a mere 0.15 ha (Figure 10.8). It is composed of small enclosed patios and adjoining rooms. A few round structures that may or may not be contemporaneous, as well as looted tombs, are found associated with it. If it does date to the early part of the EIP, it would be the only site of that period found during our survey. It is also interesting that orthogonal architecture is not common for this time period, except for sites of ceremonial significance (see, for instance, Vaughn 2009; Van Gijseghem and Vaughn 2008; Reindel and Isla 2001, 2006). TI-4 TI-4 (Figure 10.8) is located across the river from TI-3. It dates to the Late Horizon, the time of Inka arrival in the Ica region. Its configuration—composed of enclosed, roughly trapezoidal plazas, low platforms, and what may be a series of storage rooms—would qualify it as an Inka tambo or waystation. Because both Tingue and Cocharcas are today considered to be the most direct routes linking the Ica valley to the highlands, it is possible that the Inka administrative system employed the same thoroughfare. A broad and well-worn path to the north of the site may be a segment of the Inka road system, but was not surveyed by this team. This is one of the only known intrusive Inka sites in the Ica region.

Figure 10.8

Google Earth™ views of TI-4 (top) and Inka Tambo, and TI-3 (bottom), a Nasca orthogonal structure.

To the south of the site is a series of terraces and rooms atop a natural promontory. While their date is uncertain, a few LIP ceramic sherds have been identified, and this sector is likely associated with the Inka site. Other small settlements located up-valley, often on ridgetops or hilltops, have been remotely identified but were not the object of survey. Some appear to be corrals or waystations for caravans, while some are seemingly small permanent farmsteads and villages.

Discussion I have presented archaeological examples of infrastructures meant to irrigate agricultural plots using rainfall runoff in the Ica and Nasca regions. While it is clear that runoff agriculture was possible and was indeed practiced in the Late Formative and LIP periods, and that the lived landscape was starkly different than it is today, it appears that not all quebradas were

occupied in this way during these or any other time periods. Settlement patterns in many areas of the Nasca drainage, for instance, reveal that some very large quebradas, which theoretically would have flowing water for part of the year, were not considered to be ideal locations for settlement. Indeed, river-bottom irrigation based on yearly floods was probably the norm in most time periods, like it is today; but in some cases it was complemented by runoff irrigation originating from tributary quebradas of various sizes, although they often cluster on the smaller end of the spectrum. Even in the very large quebradas of Tingue and Cocharcas, evidence indicates that it was not those waterways themselves that were exploited for irrigation, but rather smaller, confluent watersheds flowing into Tingue and Cocharcas. If some of the largest watersheds were indeed exploited for agriculture, as is the case of Campanayoq, described in the first section of this chapter, others in fact appear to have been avoided, such as the very large confluent of Tierras Blancas, created by the massive obstacle that Cerro Blanco constitutes (Figure 10.1). All the while, some settlements are located at the mouth of fairly modest watersheds, as is the case of Marcaya, the Early Nasca village documented by Vaughn (2009). Many EIP settlements may have prioritized the river-bottom flood-fed irrigation that we have known historically, which is probably less labor-intensive as the infrastructure of water control it requires is not so complex. Indeed, simple ditches and expedient levees appear sufficient to soak valley-bottom lands once the yearly flooding occurs. One possibility is that the very largest watersheds coupled with narrow quebradas may have provided contexts for water management that would have required complex infrastructures because of the volume and flow of their hydrological output. In short, some landforms may have yielded too much water during the rainy season, whether they drained the highlands or funnelled runoff from local rainfall. This is perhaps why all the agricultural complexes in Tingue and Cocharcas take advantage of smaller rather than larger watersheds and appear to avoid the river bottom itself. What are the implications of these observations? First, the data indicate that runoff irrigation was occasionally practiced in the late Formative period and the LIP. However, the geomorphological conditions in which this took place were not uniform. At one point, between the start of the Early Intermediate Period two millennia ago and the historical period, the type of runoff irrigation described here became impossible because of the onset of the arid conditions with which we are familiar, a pattern that was momentarily interrupted during the LIP with the return of moist conditions. All evidence indicates that this gradual process occurred during the Early Intermediate Period, probably at some point between Early and Late Nasca. This evidence includes: (1) palaeoclimatic reconstructions; (2) some settlement associations with canal and terrace systems; (3) broad settlement pattern changes, which include the interruption of monumental building at Cahuachi, toward the Late Nasca general system of “one massive settlement per river valley”; and (4) contemporaneous developments of new puquio-based irrigation technologies focused on river bottoms in Nasca (Schreiber and Lancho 2003), which may have been developed as an alternative to runoff irrigation in the face of the paucity and unpredictability of the southern Nasca rivers. Alternatively, people in Cocharcas and Tingue could rely on the more plentiful and regular Ica river when their respective areas dried up, according to the data: once at the start of the EIP, and once more at the end of the LIP.

In terms of methodology, this study highlights the importance of expanding survey strategy to cover areas located away from river valleys and agriculturally productive areas as they appear today. Past landscapes may have been radically different from today’s with regard to hydrology, vegetation cover, and consequential attractiveness for human groups. The Tingue and Cocharcas quebradas that never appear on archaeological maps, let alone on settlement survey agendas, in fact were prehispanic agricultural breadbaskets, judging from the importance of vestigial irrigation and field systems feeding off watersheds that dried up long ago. The democratization of remote sensing makes possible preliminary surveys of inaccessible locations and can lead to data on which to anchor larger projects and get the proverbial boots on the ground. An intriguing possibility is that the runoff irrigation may have been practiced with varying degrees of intensity and regularity. While some of our data suggest some permanent settlements that relied at least partially on rainfall runoff in their habitual subsistence, in other cases it may have been infrequently practiced during freak meteorological events (i.e. El Niño), and may therefore have been a temporary measure of catastrophe management (Moore 1991; Mächtle et al. 2009; Contreras 2010). Some of the agricultural complexes and irrigation networks of Huarangal in the fossil river of Tingue have flood-control features in the form of natural ditches, which were walled in strategic places in an apparent effort to minimize soil erosion. Such a phenomenon may betray the fact that the runoff flooding was at times turbulent or unpredictable, a phenomenon that is not incoherent with the historical behavior of El Niño events. This would have demanded careful outflow management to minimize damage to the infrastructure. There are two further observations to bolster this point. Firstly, the surveys of Cocharcas and Tingue did not reveal the presence of looted cemeteries (given the fact that ancient south coastal cemeteries are visible and ubiquitous because of extensive looting), save a few graves at TI-3 and Huarangal. Either the cemeteries are there and were not looted, or these landscapes and sites were not the object of permanent settlement. Secondly, there are no clear settlements associated with the 5 ha of vestigial fields in the Quebrada La Yesera. A few isolated structures reminiscent of small farmsteads are present along the quebrada, however, and may represent temporary occupations during times of unforeseen flooding brought on by El Niño. The perceived transition from (permanent or punctual) runoff to valley-bottom irrigation altered more than subsistence patterns. It is likely that it profoundly modified people’s relationship with the natural environment, landscape, and with each other. At the risk of oversimplifying a complex phenomenon, rainfall, even if sporadic, is a “democratic” resource, generally available across the region that receives the precipitations, even allowing for some microclimatic regional variations. When water falls from the sky over a region, it is in theory available to all its inhabitants who benefit from local rainfall and may channel local runoff along diverse landforms and watersheds. The pattern we perceive is a multiplicity of contexts in which people have locally harnessed that resource. The shift from runoff irrigation to riverbottom irrigation changes the dynamic of water acquisition, availability, and control. The water’s origin becomes non-local, its source is singular: the mountains above and the single local drainage, rather than an assortment of quebradas washed in seasonal rains. If its yearly arrival is predictable, its abundance is historically unreliable (Van Gijseghem 2007). It is

therefore tempting to contextualize the population nucleation and political fragmentation distinctive of Late Nasca society, a pattern emerging during Middle Nasca, to at least partially constitute a response to new environmental conditions. Climate change could also provide a context for the pattern of competition or conflict that is attributed to Late Nasca society (Proulx 1994, 2006:44), as well as the near-obsessive focus on water and fertility in Early Nasca iconography emanating from Cahuachi: perhaps these people were undergoing gradual climatic uncertainty characterized by a decrease in rainfall and attempted to mitigate it ritually through worship, and politically through pilgrimages, gatherings, and community ritual. Finally, and this is perhaps at once my most powerful and my most sobering conclusion: according to these data, it appears that during the Early Intermediate Period there was a lot of room to farm in the Ica-Nasca region. According to the subsistence practices we have encountered, even moderate rainfall could have rendered many of these important quebradas fertile and productive, at the cost of establishing more or less labor-intensive hydraulic infrastructures. Yet many quebradas remained unoccupied. For this reason we must admit that any contextualization of Early Nasca society that is based on “Carneirist” notions of population pressure and circumscription remains insufficient (e.g. Carneiro 1988). The current reading of settlement-subsistence data suggests that if Early Nasca people felt any pressure at all, it did not originate from a Malthusian population/territory imbalance, at least during the early part of the EIP. Infrastructures in many large quebradas in the Ica-Nasca region could theoretically have been developed to harness their hydraulic potential, yet they were not. This situation may have changed once rainfall decreased, runoff irrigation became no longer possible, and subsistence options became more limited—conditions mitigated in the Nasca region by the development of puquios—at which time population pressure may have been a factor, although it was also a period of political fragmentation. During the LIP, with the return of relatively moist conditions coupled with maximal population densities, population pressure may once more have been a critical factor and it is at that time that we see the clearer signs of El Niño catastrophe management through water harvesting in places that perhaps would not normally flood (i.e. La Yesera, described above; see also Mächtle et al. 2009) in order to maximize agricultural yields. This is a necessary dimension to consider in examining the Late Nasca settlement pattern of population nucleation and inter-valley political integration.

Conclusion What I have been calling the fossil rivers of Ica hold many secrets yet to be revealed. But the main point is that our vision of settlement systems and occupation of the territory based on modern-day conditions is incomplete, and that many of these vestigial settlement systems remain to be discovered and surveyed. While archaeologists have always been sensitive to changing climate and environmental conditions in the deeper past, particularly with regard to the Late Pleistocene–Holocene planetary changes, we often make implicit assumptions about these conditions for the nearer past: that past climate was roughly similar to that which has been recorded historically is confirmed neither by palaeoclimatic reconstructions nor, as asserted here, by archaeological data. There still are some unanswered questions as to the details of the settlement patterns and

agricultural practices prevalent in different time periods. At the forefront is the relative importance of risk management, whereby people take advantage of unusually large amounts of runoff to be used in punctual irrigation schemes, and habitual, continual settlement of streams and rivers that are now dry, but that would have been reliable on a yearly basis in the past. In the latter case, the gradual aridification of the landscape would have increasingly limited people’s options for irrigation and settlement. In two periods in south coastal prehistory, during the EIP and at the end of the LIP, it seems that people were challenged by changing hydrological conditions, which, on the one hand, may have caused socio-political conflict and, on the other, may have prompted them to develop strategies of water harvesting and management during such infrequent events as El Niño. In any event, climatic fluctuations should now be part of our repertoire of explanations when considering social change in the Ica-Nasca area and, indeed, in other areas of the desert Andean coast. This study introduces an often neglected dimension to the complex settlement ecology of the arid Peruvian south coast. While settlement strategies relied on diverse criteria, these altered through time along side changing political regimes, demographic constraints, religion, and cycles of conflict, alliances, fusion and fission. In this sensitive landscape, however, settlement decisions were also influenced and constrained by even slight variations in rainfall. Consequently, during the first part of the EIP and during the LIP, settlement and irrigation strategies targeted specific landforms. Other criteria hinted at but incompletely understood may have involved certain naturally occurring soil types that were present in conjunction with landforms suitable for runoff irrigation. The present study is a preliminary step in the examination of the archaeological manifestations of changing climates in the Ica-Nasca region, building on the important work that lies at its foundation (e.g. Eitel et al. 2005; Eitel and Mächtle 2009; Hesse and Baade 2009; Mächtle and Eitel 2013). Minimally, it highlights the importance both of broadening survey strategies to incorporate areas lying outside conventionally conceived regions suitable for settlements, and of taking into consideration that landscapes as perceived and used by past populations may have been significantly different than the ones we perceive and use.

References Abbot, Mark B., G.O. Seltzer, K.R. Kelts, and J. Southon. 1997. Holocene paleohydrology of the tropical Andes from lake records. Quaternary Research 47: 70–80. Arkush, Elizabeth, and Tiffiny A. Tung. 2013. Patterns of war in the Andes from the Archaic to the Late Horizon: Insights from settlement patterns and cranial trauma. Journal of Archaeological Research 21: 307–369. Bachir Bacha, Aïcha. 2007. Dinámica y desarrollo de un centro ceremonial andino: El caso de Cahuachi. Arqueología y Sociedad 18: 69–94. Beresford-Jones, D., H. Lewis, and S. Boreham. 2009. Linking cultural and environmental change in Peruvian prehistory: Geomorphological survey of the Samaca Basin, Lower Ica Valley, Peru. Catena 78(3): 234–249. Beresford-Jones, D., O. Whaley, C. Alarcon Ledesma, and L. Cadwallader. 2011. Two millennia of changes in human ecology: Archaeobotanical and invertebrate records from the Lower Ica Valley, south coast Peru. Vegetation History and Archaeobotany 20: 273–292. Carneiro, R. L. 1988. The circumscription theory. American Behavioral Scientist, 31(4): 497–511. Conlee, C.A. 2003. Local elites and the reformation of Late Intermediate Period sociopolitical and economic organization in Nasca, Peru. Latin American Antiquity 14(1): 47–65. Conlee, C.A. 2014. Nasca culture integration and complexity: A perspective from the site of La Tiza. Journal of Anthropological Archaeology 35: 234–247.

Conlee, C.A., and K.J. Schreiber 2006. The role of intermediate elites in the Balkanization and reformation of post-Wari society in Nasca, Peru. In C.M. Elson and A.R. Covey (eds), Intermediate Elites in Pre-Columbian States and Empires, pp. 94– 111. University of Arizona Press, Tucson. Conlee, C.A., J. Dulanto, C.J. Mackey, and C. Stanish. 2004. Late prehispanic social complexity. In H. Silverman (ed.), Andean Archaeology, pp. 209–236. Blackwell, Malden, MA. Contreras, D.A. 2010. Landscape and environment: Insights from the prehispanic central Andes. Journal of Archaeological Research 18: 241–288. Cook, Anita. 1999. Asentamientos Paracas en el Valle Bajo de Ica, Perú. Gaceta Arqueológica Andina 25: 61–90. Edwards, M. 2013. The configuration of built space at Pataraya and Wari provincial administration in Nasca. Journal of Anthropological Archaeology 32: 565–576. Edwards, M., and K.J. Schreiber. 2014. Pataraya: The archaeology of a Wari outpost in Nasca. Latin American Antiquity 25(2): 215–233. Eitel, B., and B. Mächtle. 2009. Man and environment in the Eastern Atacama Desert (Southern Peru): Holocene climate changes and their impact on Pre-Columbian cultures. In Markus Reindel and Gunther A. Wagner (eds), New Technologies for Archaeology: Multidisciplinary Investigations in Palpa and Nasca, Peru, pp. 17–37. Springer Heidelberg, Berlin. Eitel, B., S. Hecht, B. Mächtle, G. Schukraft, G. Wagner, B. Kromer, I. Unkel, and M. Reindel. 2005. Geoarchaeological evidence from desert loess in the Nazca–Palpa region, southern Peru: Palaeoenvironmental changes and their impact on preColumbian cultures. Archaeometry 47(1): 137–158. Ertsen, M.W., and J. van der Spek. 2009. Modeling an irrigation ditch opens up the world: Hydrology and hydraulics of an ancient irrigation system in Peru. Physics and Chemistry of the Earth, Parts A/B/C 34(3): 176–191. Farrington, Ian S., and C.C. Clark. 1978. Hydraulic engineering and irrigation agriculture in the Moche Valley, Peru. Journal of Archaeological Science 5: 255–268. Fehren-Schmitz, L., W. Haak, B. Mächtle, F. Masch, B. Llamas, E. Tomasto C., V. Sossna, K. Schittek, J. Isla C., B. Eitel, and M. Reindel. 2014. Climate change underlies global demographic, genetic, and cultural transitions in pre-Columbian southern Peru. PNAS 111(26): 9443–9448. Hesse, R., and J. Baade. 2009. Irrigation agriculture and the sedimentary record in the Palpa Valley, southern Peru. Catena 77(2): 119–129. Isla C., J. 1990. La Esmeralda: Una ocupación del periodo Arcaíco en Cahuachi, Nasca. Gaceta Arqueológica Andina 20: 67–80. Kantner, J., and K.J. Vaughn 2012. Pilgrimage as costly signal: Religiously motivated cooperation in Chaco and Nasca. Journal of Anthropological Archaeology 31: 66–82. Mächtle, B., and B. Eitel 2013. Fragile landscapes, fragile civilizations: How climate determined societies in the pre-Columbian south Peruvian Andes. Catena 103: 62–73. Mächtle, B., B. Eitel, G. Schukraft, and Katharina Ross. 2009. Built on sand: Climatic oscillation and water harvesting during the Late Intermediate Period. In Markus Reindel and Gunther A. Wagner (eds), New Technologies for Archaeology: Multidisciplinary Investigations in Palpa and Nasca, Peru, pp. 39–46. Springer Heidelberg, Berlin. Massey, Sarah A. 1986. Sociopolitical change in the Upper Ica Valley, 400 BC to AD 400: Regional states on the south coast of Peru. Ph.D. dissertation. Department of Anthropology, University of California, Los Angeles. Moore, J.D. 1991. Cultural responses to environmental catastrophes: Post-El Niño subsistence on the prehistoric north coast of Peru. Latin American Antiquity 2: 27–47. Moseley, Michael E., and Eric Deeds. 1982. The land in front of Chan Chan: Agrarian expansion, reform, and collapse in the Moche Valley. In M. Moseley and K.C. Day (eds), Chan Chan: Andean Desert City, pp. 25–54. University of New Mexico Press, Albuquerque. Orefici, G. 1988. Una expresión de arquitectura monumental Paracas-Nasca: El Templo del Escalonado de Cahuachi. Atti Convegno Internazionale: Archeologia, Scienza y Societa dell’America Precolombiana: 191–201. Orefici, G. 1996. Nuevos enfoques sobre la transición Paracas-Nasca en Cahuachi (Peru). Andes: Boletín de la Misión Arqueológica Andina—Universidad de Varsovia 1: 173–197. Orefici, Giuseppe. 2011. Cahuachi, Capital Teocrática Nasca, Volume I. Universidad de San Martin de Porres, Fondo Editorial, Lima, Peru. Ortloff, C. M., M.E. Moseley, and Robert A. Feldman. 1982. Hydraulic engineering aspects of the Chimu Chicama–Moche intervalley canal. American Antiquity 47(3): 572–595. Proulx, D. A. 1994. Stylistic variation in proliferous Nasca pottery. Andean Past 4: 91–107. Proulx, D.A. 2006. A Sourcebook of Nasca Ceramic Iconography: Reading a Culture through its Art. University of Iowa Press, Iowa City. Reindel, M. 2009. Life at the edge of the desert: Archaeological reconstruction of the settlement history in the valleys of Palpa, Peru. In M. Reindel and G.A. Wagner (eds), New Technologies for Archaeology: Multidisciplinary Investigations in Palpa and Nasca, Peru, 439–462. Springer, Berlin. Reindel, Markus, and Johny Isla Cuadrado. 2001. Los Molinos y La Muña. Dos centros administrativos de la cultura Nasca en

Palpa, costa sur del Perú. Sonderdruck aus Beiträge zur Allgemeinen und Vergleichenden Archáeologie 21: 289–319. Reindel, Markus, and Johny Isla Cuadrado. 2006. Reconstructing Nasca social and political structures: A view from Los Molinos and La Muña. In I. Shimada, H. Baba, K. Shinoda, and O. Masahiro (eds), Nasca, Wonder of the World: Messages Etched on the Desert Floor, pp. 165–173. TBS, Tokyo. Rowe, J.H. 1963. Urban Settlements in ancient Peru. Nawpa Pacha 1: 1–27. Schreiber, K.J. 1999. Regional approaches to the study of prehistoric empires: Examples from Ayacucho and Nasca, Peru. In S.E. Alcock, T.N. D’Altroy, K.D. Morrison, and C.M. Sinopoli. (eds), Empires: Perspectives from Archaeology and History, pp. 70–92. Cambridge University Press, New York. Schreiber, K.J., and J. Lancho Rojas J. 2003. Irrigation and Society in the Peruvian Desert: The Puquios of Nasca. Lexington Books, Oxford. Silverman, H. 1993. Cahuachi in the Ancient Nasca World. University of Iowa Press. Iowa City. Silverman, H. 2002. Ancient Nasca Settlement and Society. University of Iowa Press, Iowa City. Thompson, L.G., E. Mosley-Thompson, J.F. Bolzan, and B.R. Koci. 1985. A 1500-year record of tropical precipitation in ice cores from Quelccaya Ice Cap, Peru. Science 229: 971–973. Thompson, L.G., E. Moseley-Thompson, M.E. Daws, P.N. Lin, K.A. Henderson, J. Cole-Dai, J.F. Bolzan, and K.-B. Liu. 1995. Late glacial stage and Holocene tropical ice core records from Huascaran, Peru. Science 269: 46–50. Unkel, I., A. Kadereit, B. Mächtle, B. Eitel, B. Kromer, G. Wagner, and L. Wacker. 2007. Dating methods and geomorphic evidence of palaeoenvironmental changes at the eastern margin of the south Peruvian coastal desert (14 degrees 30’ S) before and during the Little Ice Age. Quaternary International 175: 3–28. Van Gijseghem, H. 2004. Migration, Agency, and Social Change on a Prehistoric Frontier: The Paracas Nasca Transition in the Southern Nasca Drainage, Peru. Unpublished Ph.D. dissertation. Department of Anthropology, University of California, Santa Barbara. Van Gijseghem, H. 2006. A frontier perspective on Paracas society and Nasca ethnogenesis. Latin American Antiquity 17(4): 417–444. Van Gijseghem, H. 2007. Demography, environment, and punctuated change at the close of the Early Horizon. 72nd Annual Meeting of the Society for American Archaeology, Austin, TX. Van Gijseghem, H., and K.J. Vaughn 2008. Regional integration and the built environment in middle-range societies: Paracas and Early Nasca houses and communities. Journal of Anthropological Archaeology 27(1): 111–130. Van Gijseghem, H., K.J. Vaughn, V.H. Whalen, M. Linares Grados, and J. Olano Canales. 2011. Prehispanic mining in South America: New data from the Upper Ica Valley, Peru. Antiquity (Project Gallery) 085(328). Van Gijseghem, H., K.J. Vaughn, V.H. Whalen, M. Linares G., and C. Olano J. 2013. Economic, social, and ritual aspects of copper mining in ancient Peru: An Upper Ica Valley case study. In N. Tripcevich and K.J. Vaughn (eds), Mining and Quarrying in the Ancient Andes: Sociopolitical, Economic and Symbolic Dimensions, pp. 275–298. Springer, New York. Vaughn, K.J., and M. Linares Grados, 2006. Three thousand years of occupation in upper valley Nasca: Excavations at Upanca. Latin American Antiquity 17: 595–612. Vaughn, K.J. 2009. The Ancient Andean Village: Marcaya in Prehispanic Nasca. University of Arizona Press, Tucson. Williams León, Carlos, and Miguel Pazos Rivera. 1974. Inventario, Catastro, y Delimitación del Patrimonio Arqueológico del Valle de Ica. Instituto Nacional de Cultura, Lima.

11 The organics of settlement patterns in Amazonia Fernando Ozorio de Almeida [I]t is clear that the house is not the analogical image of a living being, but the paradigmatic image of the organic process in general. (Descola, 1996: 121)

Introduction Ismail Kadaré described the stone-walled village in Albania where he grew up as impenetrable and impermeable. Everything was old, and the stones looked as though they had been there forever, and seemed to be naturally cold and harsh. “It was hard to believe that, under this powerful carapace, the tender flesh of life survived and reproduced” (Kadaré, 2007: 1). The traditional tropical forest indigenous settlement would look quite the opposite. Everything was, or had once been, alive: the houses made of wood and palm-tree leaves, the surrounding gardens and adjoining canopy; and the beaten-earth ground, combining earth, ash, waste, ceramic sherds, insects, and possibly some long-ago deposited remains of the dead. Over the centuries, such a combination of rotten matter and imperishable sherds would make the earth go black, leading to the creation of the so-called anthropogenic dark earth (ADE) (e.g. Arroyo-Kalin, 2010; Kämpf and Kern, 2005; Petersen et al., 2001; Schmidt, 2013). These are thus organic contexts. Before the times of the brick houses, quite common nowadays, the houses—and indeed the whole village—would be constantly on the move. A house might fall down only to be rebuilt nearby; a house might grow and change shape, with the building of extensions facilitated by the malleability of the palm-leaf walls—a habit that would be later fully embraced by Brazilian suburban communities under the name of puxadinho. When entering a traditional indigenous Amazonian house, the outsider’s senses are sharply challenged (Figure 11.1). Inside the light is dim, coming only from one or two doors, more or less opposite one another. The air is composed of a thick mixture of moisture and smoke, and there are no rooms or any kind of visible internal divisions. People generally sleep in hammocks (Lowie, 1948), though sleeping mats are also common (e.g. Harner, 1972; Vidal, 1983: 89). Fragile objects, such as ceramics, are often kept high above the floor, away from the children, in racks (e.g. DeBoer and Lathrap, 1979; Steward and Métraux, 1948: 571). Baskets can be hung, while other smaller personal things can be stuffed into the leaf walls (Novaes, 1983: 63; van Velthem, 1983: 188–189). The latter have great thermal properties,

leaving the house cool during the hot day, and maintaining the heat throughout chilly nights (Silva, 1983: 41). Other patterns can be discerned, some of which continue to link the past and present. For example, it is impossible to ignore the strong spatial association between pre-colonial sites and present-day local towns or villages. In Amazonia, one rarely needs to look for sites, since they can frequently be found beneath present-day occupations. This spatial pattern of cooccurring ancient and modern settlement can be observed in larger cities in the Brazilian part of Amazonia (e.g. Manaus, Porto Velho, Marabá, Altamira, and, especially, Santarém), as well as at smaller villages and towns (e.g. Machado, 2010). Hence Amazonia is full of persistent places (see discussion by Moore and Thompson, 2012: 267; Schlanger, 1992: 92; Zedeño and Bowser, 2009: 12). Even when there is no perfect relationship between the archaeological site and recent settlement, it is sometimes possible to find a curious inversion in the use of space. For example, during an archaeological survey in Central Amazonia, on the Paraconi, Parauari, and Apocuitaua Rivers, Guilherme Mongeló and I identified archaeological sites in most of the communities where we stopped. However, in several cases, we found people living over what seemed to be slightly darkened earth (Terra Mulata; Munsell 10YR 4/3) with little or no archaeological material, while their gardens were located in areas with deep (over 50cm) ADEs, filled with ancient ceramic sherds (Almeida et al., 2010). Thus, the village and gardens seem to have switched places through time—an understandable process, since the organic matter created by the ancient villages is highly prized by present-day gardeners—though the settlement complex was roughly in the same place (cf. Stenborg et al., 2012: 239).

Figure 11.1

Inside an abandoned house (Coudreau, 2009: 67 [1897]).

Despite these common features, a general discussion over settlement patterns in Amazonia across space and time is nevertheless complicated, since, if there is a regional pattern, it is one of diversity. In this chapter, using archaeological, ethnohistorical and ethnographic data, I will

address some of the elements that compose this multiplicity. This heterogeneous lifeway or adaptation is based, in part, on riverine and interfluvial variability, which are related to both natural and constructed landscapes.

Tropical forest settlement patterns The concept of “tropical forest culture” used in this chapter is a slightly modified version of the classic definition coined by Lathrap (1970). The latter adopted an economic bias, where tropical forest culture was understood as “a way of life supported by intensive root-crop agriculture” (ibid.: 47). The latter could be, whenever possible, complemented by the exploration of rivers, lakes, and coasts, and secondarily by hunting mammals and birds (ibid.). Lathrap’s definition signified a powerful and positive twist towards Lowie’s (1948) and Meggers’s (1971) culture-ecology taxonomic view, where human cultures had limited possibilities of social evolution in the tropical forest. These limits would be enforced by ecological boundaries, such as poor soils and the unpredictability of rivers for floodplain agriculture and fishing. Lathrap, on the other hand, viewed the floodplains (várzeas) as possessing unlimited potential for agriculture, especially one based on manioc (Manihot esculenta) (cf. Neves, 2008: 364); however, he agreed that the poor soils of the Amazonian uplands would be inappropriate for human occupation and social complexification. Hence, one who follows Lathrap’s view would expect settlements in Amazonia to follow a gradual process of occupation of the várzeas by agricultural groups, demographically increasing and steadily expanding from a core center, such as central Amazonia (Lathrap, 1970: 46–77). Terra firme occupations would be a natural consequence of the major rivers’ saturation. To endure the burden of the uplands’ harshness, inhabitants would have been obliged to associate themselves with the sovereign riverine groups (Lathrap, 1968a). However, several authors (e.g. Carneiro, 1995; Denevan, 1996; Morán, 1990) correctly point out that the várzea/terra firme dichotomy is too simplistic. Both the riverine areas and the uplands were heterogeneous, and so were the possibilities of landscape transformation and cultural adaptations in these environments. Following these authors, I will argue for a gentle inversion in Lathrap’s definition of tropical forest culture. Here, it is the ecotones, the transition or encounter of environments, which mattered most in human decisions for choosing places to settle and explore. This framework is valid both for understanding aquatic environments—i.e. the exploration of rivers, lakes, and coasts—and for upland environments: for example, secondary rivers and the transitional areas between tropical forests and savannas that are constantly sought by upland groups for hunting and gathering activities (discussed below). Consequently, the linear process of colonization, predicted by Lathrap, is broken and substituted by a much more complex set of possibilities (Kellett and Jones, in this volume), including long-distance moves (e.g. over 500km), and frequent reoccupations. In this historical overview, some elements of landscape, such as the best places for horticulture, the existence of abandoned gardens and the presence of ADEs, play dialectical roles in decision-making, as they are an ongoing outcome (Arroyo-Kalin, 2004) of human/environment interactions. Even some “natural” elements of landscape sought by humans

(e.g. lakes and river banks) can suffer changes due to the fluvial dynamics of the lowlands (Lathrap, 1968b). These ongoing transformations of landscape imply that the elements taken into account for “reoccupations” are constantly shifting; they further imply the historical existence of distinct individuals or groups and, consequently, different layers of the significance applied to a determined place in space and time (Zedeño and Bowser, 2009: 10). Therefore, following the proposed theme of this volume, it is my premise that the heterogeneity of landscape and culture was mirrored in the long-term settlement ecology of the Amazon Basin, and that the historical engagement of cultures and environments in Amazonia (cf. Balée, 1992; Balée and Erickson, 2006; Erickson, 2003; Erickson and Balée, 2006; Neves, 2012; Schaan, 2013) produced endless possibilities for places to live, house forms, and village spatial organization, including permanent and temporary structures indoors and out. Even Robert Lowie (1948), despite his cultural ecological background, acknowledged this settlement diversity in his synthesis of the tropical forest cultures. In his view, houses could be made to accommodate single families or many families. A village could be composed of a sole house or a set of houses, surrounded by other specific structures (cf. López-Mazz, 2010). These clusters could be a scattered compound of spaced dwellings or in the form of delicate and balanced geometric arrangements. Soil fertility, distance to water and to other settlements, and safety concerns were all considered (Lowie, 1948: 16). This chapter thus reflects on patterns of variability. I shall first present an excerpted and general overview of the ethnographic and archaeological data (Figure 11.2). Then, I will put forward a specific discussion on the regional archaeological context in the Upper Madeira River, my present research area. The combination of descriptions, comparisons and synthesis will lead us to several issues that are relevant to understanding the organics of settlement patterns in Amazonia.

Ethnographic background Probably the first image one has when envisioning a South American indigenous village is that of a circular form. Ring villages, with a system of pathways leading the community to other areas (e.g. the river, gardens, seasonal camps), are common in both the Upper Xingu (Heckenberger, 2006: 332; cf. Schmidt, 2012 for the archaeology of Central and Lower Amazonia) and in the Central Brazilian Plateau (Novaes, 1983; Silva, 1983; Vidal, 1983). In these regions, this circular pattern is traditionally linked to Arawak-or Macro-Jê-speaking communities, besides groups with isolated languages, such as the Yanomami. The Macro-Jê speakers became famous worldwide mainly because of the portrayal of their ring villages by Levi-Strauss (1976), Maybury-Lewis (1974) and Turner (1992). The village form is central to these groups’ social and cosmological organization. For example, marriages usually happen between a man and women who each come from opposite houses in the circle and belong to the opposite halves (moieties) that comprise the community (Levi-Strauss, 1976). The ring village pattern enabled the Timbira living in Amazonia’s eastern borders to describe themselves as “real Indians” (Ladeira, 1983: 13). Houses are all usually of a similar size and are the same distance apart from each other. The basic spatial difference inside the

village is between the domestic context of the houses, which is the women’s arena, and the central plaza (often the place for the “men’s house”), where men gather to discuss politics, and where the rituals usually take place (Ladeira, 1983; Nimuendaju, 1967; Novaes, 1983). Pathways run circularly, linking the houses, and centripetally, connecting the plaza to the houses and the latter to gardens, hunting and fishing camps, and other villages. Some groups, such as the Xavante and the Kayapó, have summer camps, which are small-scale imitations of the villages. The huts are miniatures of the houses at the village, and the ring disposal— including the order of the houses—is never broken (Silva, 1983; Vidal, 1983). Some Tupi-Guarani-speaking groups living close to the Jê seem to have adopted the latter’s ring village pattern. This seems to have happened to the Tapirapé (cf. Baldus, 1970; Wagley, 1977), the Tenetehara (Ladeira, 1983; Wagley and Galvão, 1961), and possibly the coastal Tupinambá (Staden, 1974 [1557]; cf. Almeida and Garcia, 2008). A similar phenomenon might also have occurred with Tupian groups in Central Amazonia, such as the Munduruku (Arnaud, 1974), who possibly copied the circular pattern from their Arawak neighbors. However, these are exceptions for the Tupi. The usual pattern for the Tupi-Guarani villages is a random disposition of the houses. This can be verified in several groups such as the Araweté (Viveiros de Castro, 1986), the Parakanã (Fausto, 2001; Vidal, 1983), and the Waiãpi (Gallois, 1983, 1986). Viveiros de Castro (1986) argues that this metamorphic character of the Tupi-Guarani can be explained by their cosmological view based on the duo of otherness and predation. Tupian social organization is structured outside the village rather than inside, as we have seen happen in Jê contexts. Cosmological issues thus exercise no pressure, and any kind of village form is possible: ring (e.g. Tupinambá), random (e.g. Waiãpi), linear (e.g. Omágua; cf. Carvajal, 1941[1542]), or single-house types (e.g. Asurini, cf. Silva, 2000). Historical contingency and regional influence will define which will be preferred.

Figure 11.2

Ethnic groups and archaeological contexts cited in text (map adapted from Wikimedia Commons).

A similar metamorphic fluidity can be seen in Carib-speaking groups, which have welldocumented villages with only one house (e.g. the Waiwai; cf. Meggers, 1971), ring villages (e.g. the Kuikuro; cf. Heckenberger, 2005), or randomly scattered villages (e.g. the Wayana; cf. van Velthem, 1983). Van Velthem (1983: 173) also points to a considerable variety of house types (five) for the Wayana groups in the Guiana region. A similar variety was observed for their Tupi-Guarani neighbors, the Waiãpi, including wall-less houses and houses built over stilts, the so-called palafitas (Gallois, 1983: 159). The latter is another indigenous tradition widely adopted by modern local communities throughout the tropical forest. In western Amazonia, along the multi-cultural Montaña region that cuts through Bolivia, Peru and Ecuador, the general pattern—followed by Arawak-, Panoan- and Achuar-speakers—is the dispersed single-house settlement (Steward, 1948; Steward and Métraux, 1948). The Achuar single house, for example, is usually restricted to a nuclear family and is found over high ground to facilitate drainage, and for defensive reasons (Harner, 1972: 41). Since the Achuar settlement does not possess a plaza, it is the house that constitutes the center of both profane and ceremonial life, and it is divided into the men’s half and the women’s half (Harner, 1972). Descola (1996: 109–110) argues that this single-house nuclear family pattern could be reversed, and a unification of households would occur as a response to threats of violence. A similar scenario of household amalgamation in a fortified, strategically located settlement was also portrayed in the Upper Negro River oral history, and was archaeologically tested by Eduardo Neves (2006a; cf. Kellett and Jones, this volume). Similar to the Jê-speaking collectives, most of the Tupi-Guarani living on the border between the tropical forest and the savannas of the Brazilian central plateau used to seasonally split up into smaller groups (e.g. a nuclear family) and abandon their villages during the dry season (roughly between May and November) for trekking activities. In these temporary settlements, only a small shelter (i.e. with low archaeological visibility) was needed (a tapiri). Usually, there is a system of temporary structures for each one of these activities, as well as in the gardens (Assis, 1996). The structures and the pathways connecting them form territorial landmarks that are used to map the surroundings and possess a rotary form: open garden areas used for the new village and the old village used for camping, with its abandoned fruit trees used for gathering activities and as hunting baits (Balée, 1994; Viveiros de Castro, 1986; cf. Almeida and Garcia, 2008). Territorial mapping could also be useful for the foundation of a new competing village, resulting from group fission, a frequent consequence of the political instability that is very common in the Tupi-Guarani groups (e.g. Clastres, 2004; Fausto, 2001; Fernandes, 1963). In the Guiana context, Rivière (2001: 107–108) argues that a frequent cause of political instability and the consequent village fission is demographic growth, for it can lead to the detachment of the family ties used by the chief to maintain social unity. Despite internal village symmetry, hierarchy amongst the settlements was observed on the border between Amazonia and the Brazilian Central Plateau. Wüst (1994, 1998), who worked with the Bororo, argues that “the hierarchical position of the Bororo settlement is considered a function primarily of the presence of shamans and the number of funerals held,” and that there seemed to be a correlation between the presence of the shaman and population size (Wüst,

1994: 321–322). In the Upper Xingu, a region historically referred to for its multi-cultural diversity and long-term (for Amazonia) village occupations (around 15 years; Sá, 1983: 107), Heckenberger (2001: 48; 2006: 333; Heckenberger et al., 1998: 371) identified a hierarchy in both archaeological and ethnographical Xinguano villages; however, archaeological sites were larger (up to 50ha) and had denser occupation lasting long enough to create thick (around 50cm) layers of ADEs—features that allowed the author to discuss the possibility of local urbanism. An indigenous house rarely lasts longer than 15 years (e.g. Descola, 1996: 117), since wood and palm fibers are no match against moisture and insects for long, and sometimes they need to be rebuilt before a time span of five years (e.g. Rivière, 2001). Traditional literature (e.g. Lowie, 1948: 18; Meggers, 1954, 1971) used to propound that the main motive for village abandonment was the exhaustion of the soil, though this argument has been heavily criticized (e.g. Carneiro, 1961). Besides the houses’ short lifespan (Wüst, 1994: 325) and fissions resulting from community quarrels (Neves, 1995; Rivière, 2001), a major and widely acknowledged motive for the abandonment of a settlement is the death of the chief of the village or of the house (Gillin, 1948: 830; Harner, 1972: 4). Further reasons for village abandonment—of huge importance to the understanding of great historical movements—are inhabitants fleeing from enemy raids (e.g. Harner, 1972: 45), witchcraft (Whitehead, 2002), or disease epidemics (Clastres, 1978).

Archaeological background The continental proportion of Amazonia vis-à-vis the insufficient number of researchers, and the presence of sites with areas spread over several hectares, for decades fettered spatial analysis in the region. Still today, it is a complex task to identify house forms amongst the archaeological palimpsests created by this organic world. Postholes, garbage middens, and the low density of material in potential village plazas remain the most common features that enable us to understand aspects of village spatiality (e.g. Barreto, 2015), although sophisticated techniques, such as inferring activity areas using chemical signatures (Schmidt, 2010; Stampanoni, 2016), geophysics (Gomes and Luiz, 2013; Roosevelt, 1991), and GIS (Fonseca, 2013) have also provided new insights. In Brazil, Irmhild Wüst (1994, 1998) pioneered intra-site analysis by means of an interface combination of archaeology and ethnography. Working in the Central Plateau, the author proposed that ring villages, related to pottery of the Uru Tradition, would date back to AD 800 (Wüst, 1994: 332; Wüst and Barreto, 1999: 3). The Uru sites, usually found on “gentle slopes in interfluve areas,” have shallow refuse deposits (20–30cm), and diameters varying between 110 and 410m (Wüst, 1994: 332–333). Wüst (1994: 332; Wüst and Barreto, 1999: 16), also identified sites with three concentric circles, suggesting a superior population density during pre-colonial times. Barreto (2011: 70) discusses possible internal space differences between the ring village sites from the Brazilian Central Plateau, vis-à-vis the ones in Central Amazonia. While the former generally consisted of a cluster of (supposed) houses that were all of very similar sizes, and with the symmetric distance to the plaza also observed in ethnographic research, the latter

was formed by a group of middens with different dimensions and heights and postholes, which suggest residential contexts (Moraes, 2010). People were re-burying their relatives beneath their houses, since primary burials were found in the middens. These burials were interpreted by Py-Daniel (2015: 297) as having patterns related to gender and age, possibly reflecting social differences between the sites’ inhabitants. Hence, we can deduce that the architectural character of the circular village might not always have been a synonym for political internal symmetry throughout indigenous histories (cf. Heckenberger, 2005). The circular pattern of sites in Central Amazonia is related by Moraes (2010; Moraes and Neves, 2012) to the Paredão Phase which, in turn, is potentially associated with ancient Arawak speakers (cf. Lathrap, 1970). This phase was widespread across sites in Central Amazonia between AD 600 and 1200, including extensive sites (e.g. Antônio Galo, with 16 hectares; cf. Moraes and Neves, 2012: 133). The Central Amazonian ring villages seem to vanish from the archaeological record after a period of interaction, and then conflict— materialized by defensive ditches (Neves, 2006b: 65; 2009)—with the immigration of outsiders, producers of Polychrome Tradition ceramics (AD 800–1400), who lived in (perhaps smaller) linear villages. It is possible that these foreign groups, similar to the modern-day riverine communities described in the introduction, were interested in the high fertility of the ADEs resulting from the multi-century occupation of the region (Almeida, 2013; Moraes and Neves, 2012; Rebellato et al., 2009). These large-scale, multi-component sites can often be found in river junctions, such as the one between the Negro and Amazon Rivers. Some of these sites—for example, Hatahara— perfectly fit into Denevan’s (1996) Bluff Model, in which human populations would seek places where they could explore both the fertility and fishing potential of the várzea, and the year-round stability of the drier upland areas (Rebellato et al., 2009). Meggers (1971) viewed these sites with skeptical eyes. Based on ethnographic data provided by low-density population indigenous groups, such as the Wai-Wai and the Sirionó, she defended a sporadic and sparse pre-colonial occupation of both the floodplain and the upland. In her vision, the 500m diameter sites from Central Amazonia (cf. Hilbert, 1968) resulted from the reoccupation processes of low-density populations (Meggers, 1992). This view was challenged by several archaeologists, who dismissed Meggers’s seriations as not being reliable enough to permit such inferences (DeBoer et al., 1996) and who accused her of not looking properly at the historical evidence (Myers, 1973). The Itapirema site, discussed below, will provide further evidence that these vast pre-Columbian sites did truly exist. Besides river junctions, other open areas sought out during Amazonian pre-Columbian times —where multi-cultural archaeological sequences are frequently identified—were the major river islands. Archaeologically, the most renowned of these is Marajó, which is found at the mouth of the Amazon (Meggers and Evans, 1957; Roosevelt, 1991; Schaan, 2004; 2010). This reputation is derived from the combination of the architectonic monumentality of the habitation mounds and the magnificence of the ritual (including mortuary) ceramics from the Marajoara Phase, whose producers occupied the island between AD 400 and 1300 (Barreto, 2008; 2011). The probable existence of craft specialists to manufacture the sophisticated Marajoara ware— which was used as part of religious ceremonies—the apparent differentiation in the treatment of the dead, and the labor organization required to build the mounds suggest the existence of

stratified societies and a complex sociopolitical organization over an area of 20,000km2 (Schaan, 2013: 34–36). Complex sociopolitical organization in wetland contexts was also inferred for Llanos de Mojos, in Bolivia, between AD 500 and 1400 (Erickson, 2006; Jaimes-Betancourt, 2012; Walker, 2012) and the Guianas, between AD 700 and 1000 (Rostain, 2010: 178), where clusters of earthworks, like raised fields, fish weir mounds, middens, causeways, and canals also seem to be linked to labor organization. However, we should not restrict the possibility of hierarchical political organizations to regularly inundated areas. Besides these examples, and the Upper Xingu context that has already been mentioned, one of the most extensive sites of Amazonian pre-Columbian times lies on the dry banks of the junction between the Tapajós and Amazon Rivers, beneath the present-day city of Santarém. Here we can find the Aldeia and Porto sites, which, combined, extend over an area of 50ha (Gomes and Luiz, 2013; Schaan, 2013: 117–119). One might dare say that Santarém, the land of the Tapajó, was an Amazonian city at the time of the contact with the Europeans. Moving upwards, away from the great Amazon Plain towards the river sources, topographical diversity is the general rule: slopes, hills, mountains, rocky outcrops, rock shelters, and caves characterized these higher elevation areas and presented varied landscape for indigenous groups who incorporated, modified, and controlled these areas. Such is the case of the Mutuca site, located in the interfluvial region between the Tocantins and Xingu Basins (Garcia, 2012). The Mutuca site is located in the serra do Mutuca, which divides the waterways going to the Xingu (e.g. the Fresco River) and to the Itacaiúnas River, a strategic place in this formidable connection between the basins. The inhabitants of the site thus controlled a strategic pathway, for this geomorphological gap is the only (easy) way through this natural 50km east–west barrier (Figure 11.3). The Mutuca site’s archaeological remains can be found in an area of over 4ha and are related to two occupations, although just the later one has been dated. The earlier occupation is associated with incised-modeled ceramics, while the late occupation is related to the TupiGuarani Tradition and has seven radiocarbon and thermoluminescence (TL) dates varying from AD 750 to AD 1250 (Garcia, 2012: 105). The site has two ADE concentrations, forming an “L”, measuring 49×8m and 45×10m, which probably represent the houses (ibid.: 94–96). Garcia also identified a few stone axe-blades at the site and three clusters of polishing bowls in the immediate stretch of the Cateté River (ibid.: 98). It could be suggested that during some time of its pre-colonial occupations, the intersection where the Mutuca site can be found could have been some kind of trading post, inside a Xingu-Tocantins interfluvial network, which included the local production of stone axes.

Figure 11.3

The Mutuca site, guarding the pathway through the serra do Mutuca (source: Google Earth™).

We might find a somewhat similar context on the Ji-Paraná River (Axe River, in TupiGuarani), on the Middle Madeira Basin, where local non-indigenous inhabitants are frequently found to possess extensive collections of stone axe-blades collected during the planting activities in their (upland) ranches (Silva, 2015). In this apparent Macro-Tupi context (Almeida, 2013; Suñer, 2015), polishing bowls are also very frequent, as well as a great variety of forms and sizes of axe heads, some over 50cm long, possibly used as adzes. Although this region possesses a deep chronology of occupations by ceramics producers, stretching back to 3000 BC (Miller, 2009; Zimpel Neto, 2009), it is unlikely that all these blades were made for local use. The low availability of quality stone material in most of the Amazon Basin required importation from other places. The existence of trade networks in Amazonia is historically documented in riverine contexts and between the latter and the uplands (e.g. Lathrap, 1973). Chronicles describe the tracks stretching inland from the Amazon River as being beaten through regular use, as well as the existence of trading settlements every three leagues. These places would have had gardens of manioc and corn to supply the travelers. In these descriptions, the riverine groups would be exchanging dry fish and pottery for gold sheets and trinkets (Porro, 1994: 84). Rivière (2001: 116) argues that in the Guianas these “dry” inland networks required that the settlements were never more than a day’s walk from each other. Also regarding the Guianas, Whitehead describes the paths linking the coast and the main river settlements to the inland communities as “arteries of power” (1994: 38), and advocated that they created a balance between the former and latter. Stone items and poisons were examples of valuable elements that enabled the interfluvial groups to obtain this balance (ibid.), an equilibrium that would have been broken in the first centuries of the colonial period when the coastal and riverine groups would gain an advantage as they suddenly became “intermediates” between the Europeans and the mainland (ibid.: 44)

Stone axes and adzes must have been highly valued items for southwestern Amazonian groups, as they were required for the execution of earthworks found in that area, such as the socalled geoglyphs: geometric figures—generally regular or irregular circles or semi-circles and/or squares—usually related to interfluvial contexts (Schaan et al, 2010; Trindade, 2015). It is still unclear if the geoglyphs, which date from 1000 BC up to AD 1400 (Saunaluoma and Schaan, 2012), were intended for defensive (Dias and Carvalho, 2008), ritual (Schaan, 2013), or other purposes. It seems apparent, however, that these sites were not permanent settlements, for the archaeological record is usually scant, and generally there are no ADEs enclosed by the ditches, though these ADEs may appear in nearby sites. Besides the geoglyphs, features occasionally related to ceramic interfluvial sites from southwestern Amazonia also include rock art engravings. The Upper Ji-Paraná region possesses a set of apparent settlement sites close to sandstone rocky outcrops; these sites are filled with geometric and anthropomorphic engravings that are possibly related to rituals (Garcia, 2012), although in some cases the rock art is found in a separate location to the sites, on higher ground. Once again, we find interfluvial contexts with rich landscape alterations: meaningful places resulting from human–environment interactions that have helped to create and maintain territories, and presenting a field of action for practices and emotions related to the concept of territoriality (Suñer, 2015; cf. Zedeño and Anderson, 2010: 12). These rocky territorial landmarks are also very common in a variety of riverine contexts in the Upper Madeira (Almeida, 2013; Kipnis et al., 2013), the Upper and Middle Negro (Valle, 2010), the Central Amazonian Urubu (Cavallini, 2014), and the Lower Amazon (Pereira, 2004).

Indigenous occupation patterns in the Upper Madeira The Madeira River is an important tributary of the Amazon and is formed by the junction of two fluvial complexes: the Guaporé/Mamoré, and the Beni/Madre de Dios. The Beni and Madre de Dios are formed in the Andes. The Guaporé springs at the northwestern borders of the Brazilian Central Plateau, while the Mamoré is formed by the union of the Chaparé, Grande, and Ichilo Rivers, in the Bolivian Lowlands. The Madeira River provides over 50 percent of the sediment load transported to the Amazon River, though it contributes only 15 percent of the water (Tizuka, 2012: 254). I will focus on the Upper Madeira, from the meeting of the Mamoré and Beni Rivers down to its first major tributary below the rapids, the Jamari River. The considerable amount of sites identified in this region, over a hundred with useful data, permits us to discuss the differences between settlements on a major and a secondary river. The Madeira and Jamari Rivers have both experienced a rarely identified (in Amazonia) continuous indigenous occupation, from 7000 BC up to today: Tupian groups (Kawahiwa and Karitiana) continue to inhabit the Upper Jamari, and a cultural–linguistic melting pot can still be found on the right bank of the Guaporé/Mamoré Rivers (Crevels and van der Voort, 2008). This long-term history can be divided into three periods: (1) lithic occupations without ADEs, from 7000 BC–2800 BC, (2) lithic occupations with ADEs, from 2800 BC–800 BC, and (3) ceramic occupations, with or without ADEs, from 800 BC up to today (Miller, 1992). If we compare the data from the Upper Madeira, and the Middle and Lower Jamari (Tables

11.1 and 11.2; Figure 11.4), the first striking aspect is that the frequency of sites with ADEs is higher along the Jamari than along the Madeira River. In both cases the amount of sites with ADEs seems to diminish in the upper segments of the rivers. The Madeira sites seem to be further away (average 208.9m) from the river than the Jamari sites (128.9m), although it can be expected that the former will often be closer to small streams or lakes, since the muddy Madeira waters are inadequate for consumption (cf. Wüst, 1994: 324–325). Table 11.1

Table 11.2

Features of the Upper Madeira sites

Features of the Middle and Lower Jamari sites

Figure 11.4

The Upper Madeira and Middle and Lower Jamari sites (satellite imagery from Google Earth™).

As expected, sites have larger areas in the Upper Madeira (average 45583.6m²) than in the Middle Jamari (average 29884.8m²): generally they are about 50 percent bigger. This set of data is open to challenge, since a considerable number of the sites in both contexts were opportunistically delimited, through the observance of surface material or with a couple of test-pit lines. On the other hand, if there is some truth to this data, it means that although there is a clear difference between major river sites and secondary river sites, this difference is not significant. The average depth of archaeological material is also usually greater in the Madeira sites (average 140.9cm) than in the Jamari (71.9cm), although the latter has a superior quantity of ceramic occupations (average 1.3 occupations vs. 1.13 at the Madeira sites). This could enable us to suppose that the occupations on the banks of the Madeira lasted longer than on the Jamari. Sites like Teotônio (Figure 11.4), on the Madeira, which seem to have been occupied continuously for 6000 years or more (Mongeló, 2015), would strengthen this argument. However, we must bear in mind that the archaeological phases proposed by Miller (1992, 1999) are open to question in both contexts. Furthermore, when we broaden our view on places—for example, the fluvial islands of the Madeira described by Zuse (2014)—rather than only on primary or secondary rivers, we will probably find different and slightly more complex patterns (Almeida, 2013; Zuse, 2014). Such a framework also enables us to understand why these early ADEs are in some locations (i.e. around the downriver rapids and waterfalls) rather than others: these places were consciously sought (Almeida and Neves, 2014). The recent paleobotanic data showing the predominance of non-domesticates (e.g. different palms) in both pre-ceramic and ceramic layers of the Teotônio site (McMichael et al., 2015), and the high year-round fishing yield estimated for the eponymous waterfall (Goulding et al., 1996), seem to provide a different scenario vis-à-vis the standard vision (e.g. Miller et al. 1992) where agriculture (or horticulture) is always a precondition for sedentism (cf. Kelly, 2013; Moraes, 2010). If some places were sought, others were avoided. For example, it was also possible to identify a concentration of sites around the last downriver waterfall of the Jamari River. Below this point, there seems to be an absence of pre-colonial occupations for 40km. These territorial gaps between sites, or clusters of sites, are found all around Amazonia and are known as “buffer zones” (Almeida, 2013; DeBoer, 1981; Myers, 1973; Tamanaha and Neves, 2014). The Madeira presents a similar scenario, with only some scattered sites downriver from the cluster of sites found around the last of its rapids (the Santo Antônio rapid). Apparently, during an imprecise period of the Upper Madeira’s history, possibly around AD 700–1000, the lower Jamari buffer zone kept the producers of ceramics from the Jamari Tradition (500 BC–AD 1700), who occupied the eponymous river, apart from the producers of Polychrome Tradition ceramics (AD 700–AD 1600), who occupied extensive areas along the Madeira River (Almeida, 2013).

Large sites, lakes, abandoned meanders, and the archaeological patterns

The first extensive site to appear on the Madeira, downriver from the cluster of sites found around the last rapids, is the Itapirema site, which accompanies the river for over one kilometer. This site lies precisely in front of the mouth of the Jamari River, and leads us back to another pan-Amazonian settlement characteristic that has already been mentioned: the presence of large archaeological occupations at the meeting of a mainstream and its major tributaries (cf. Lima, 2008; Neves, 2008). The Itapirema site provides insightful data to help us to understand if the great Amazonian sites are large pre-colonial settlements or the outcome of innumerous reduced occupations. Unlike the traditional hard-to-interpret, large, multi-component Amazonian site, Itapirema’s stratigraphy is composed of only one ceramic layer (Almeida, 2013). The chronology obtained for two areas (Figure 11.5; Table 11.3), 400m apart from one another, also suggests a single occupation around AD 1300. If we compare the ceramics of these dated levels from Area 1 and Area 2 (Table 11.3) it is possible to identify a great similarity in the ceramics of both contexts: coiled, with light colors (ocher or orange), high-fired with thin walls (0.6–1cm), tempered with cauixí (a fluvial sponge), with a predominance of rims with simple morphology (i.e. without angles), and occasionally decorated with red paint or incisions. These similarities enable us to infer that in both areas we have the same ceramic industry, related to the Polychrome Tradition (Almeida, 2013). Nevertheless, it is possible to identify a difference in the choices for surface treatments in these areas, with a preference for brown slip in Area 2, and a preference for burnishing inside and red slip outside the vessels of Area 1. The frequency of incised and painted decoration is also higher in Area 2 (Table 11.3). If we compare all the ceramic vessel types inferred for these areas, including all levels of all excavated units, we would identify quite similar patterns of vessels in these places, with the exception of a significantly high frequency of Type 8 vessels in Area 2 (Figure 11.5). This form, an open bowl with inflected morphology, adequate for the consumption of solid food (cf. Rice, 1987), is precisely one of the most decorated vessel types of this site (62 percent in Area 2 have some kind of surface treatment). There are some topographical elements that need to be taken into account when trying to understand these spatial differences. Area 1’s ground is slightly higher than that of Area 2: their altitudes are 55m vs. 53m respectively—a significant difference, considering the annual floods of the Madeira. While Area 1’s terrain is basically plain, Area 2 has a cluster of middens, with an average size of 15×25×0.8m, which provide a lumpy character to this area’s landscape.

Figure 11.5

The Itapirema site and the two compared areas (map by Marco Brito).

Table 11.3

Comparison in ceramic assemblages between Area 1 and Area 2 Area 1 (1×1m)

Area 2 (0.5×2m)








Radiocarbon/AMS date (cal. 2 AD 1310–1360 sigma)

AD 1290–1420 (40–50cm)

AD 1270–1430 (70–80cm)

Analysed sherds



Manufacture technique

Coiled (93%)

Coiled (54%)

Predominant temper

Cauixí (75%)

Cauixí (82%)

Surface color

Orange (36%)

Ocher (29%)


Complete (62%)

Almost complete—thin internal dark line (53%)

Internal slip

Brown (75%)

Burnishing (60%)

External slip

Brown (75%)

Red (40%)

Wall thickness

0.6–1cm (55%)

0.6-1cm (66%)

Rim morphology

Simple (48%)

Simple (58%)


Flat (82%)

Flat (80%—4/5 sherds)

Plastic decoration

Incision (3 sherds)

Incision (90%—9/10 sherds)

Painted decoration

Red (4 sherds)

Red (97%—29/30 sherds)

The huge quantity of ceramics (n = 4023 sherds in 1m³), the fact that the sherds were usually identified disarticulated, often in vertical or transversal position (i.e. they seem to have been piled together), and the similar chronology for the lower and middle part of the midden, permitted us to infer that these middens were deliberately constructed to raise the houses and protect them from the floods (Almeida, 2013). At present, some local residents of the São Carlos community, to be found over the Itapirema site, build palafita houses for the same purpose. If the middens found in Area 2, following the Central Amazonian pattern discussed above, are in fact the remains of house foundations—an issue to be re-addressed in future research—it would be fair to ask the reason for the puzzling higher frequency of decorated material in this area vis-à-vis Area 1. One possibility is that domestic, utilitarian bowls are precisely those vessels with shorter lifespans. When they are not deliberately broken ritually (e.g. DeBoer, 2001) they are quickly worn out by daily use (DeBoer and Lathrap, 1979). Hence, it is likely that these would be the most available sherds to be recycled as house foundations. However, we still lack data to settle the question, since the same argument could be made to defend the suggestion that the middens are simply piles of trash adjacent to the houses. Besides the access to both the Madeira and the Jamari River, another strategic feature of the Itapirema site is the existence of the Cuniã Lake in its backyard: a rich source of fish used by the indigenous occupants of the site and the present community of São Carlos. Cuniã is actually an oxbow lake, and the Itapirema site can be found precisely over the strip of silt and clay deposited by the Madeira waters that cut out the meander from the river. On the other side of the lake, on the former Madeira bank, Eurico Miller (1992) identified archaeological sites (Figure 11.4, sites 4, 14, 32, and 37) with Polychrome ceramics. The relation between these sites and the Itapirema site is still unknown. One hypothesis, based on Lathrap’s (1968b) model linking antique sites to abandoned meanders, is that the inhabitants of the Itapirema site would have followed the river and moved there from the ancient bank (i.e. the sites identified by Miller) when the lake was formed. The existence of older sites in abandoned Upper Madeira meanders, such as the Nova Vida site (cal. AD 80–390, Figure 11.4), favors this hypothesis. Early dates in these unexcavated lake sites would provide further evidence for this argument. An alternative possibility is that people inhabiting the Itapirema site used the opposite lake sites as a temporary camp, probably for fishing, or these sites could have been both habitations and subsequently camps. These lake sites hidden away from the fluvial mainstream, moreover, would present a perfect refuge from indigenous enemy raids and, subsequently, colonial incursions. As we have seen, dynamic strategies of reoccupation are well documented for the southeastern Amazonian Tupi-Guarani-speaking groups. A similar pattern, adapted to riverine areas, could have occurred in the Upper Madeira and elsewhere. This might also have been the case for the nearby Jacarezinho site (Figure 11.4), found beside the homonymous lake, in the Lower Jamari. It was possible to identify two occupations in this site, both of them related to the producers of ceramics from the Jamari Tradition: the early one would be related to two AMS dates of cal. AD 1040–1100 and cal. AD 990–1160, while the late occupation would be

related to an AMS date of cal. AD 1270–1400. If we compare the ceramic density in the levels of these occupations (40–60cm and 0–20cm), it is possible to identify a great difference in the area of each one of them: the earlier population occupying an area of 2ha (i.e. 20,000m²) and the latter occupying over 8ha. Such a spatial difference may indicate different functions in the occupations of the site: the first as a seasonal (fishing?) camp and the second as a permanent village. Otherwise, it could also suggest a population growth of the second occupation vis-àvis the first. Either way it is clear that lakes were desirable places to settle and explore during preColumbian times. Areas such as the Jacarezinho site are extremely attractive because of their fish abundance. This attractiveness is often enhanced by the extensive igapós (seasonable flooded forests) that can be found in the sites’ surroundings. These forests are known for their high density of fish, which are attracted by the great quantity of fruits and nuts that fall from the trees (Latrubesse, 2012: 13), and the decomposing (vegetal and animal) organic material (cf. Weber, 1975: 11). Miller (1992) indicates that the Lower Jamari could also be a fishing camp area for the producers of Polychrome Tradition ceramics from the Upper Madeira River. This could be the case for the Associação Calderita site (Figure 11.4), located on the right bank of the Jamari River, just below the mouth of the Candeias River, and also related to a backyard oxbow lake. The Calderita site seems to have been occupied on at least two occasions, the first related to two AMS dates cal. AD 990–1160 and cal. AD 1020–1200 and the second related to the cal. date AD 1280–1410. Besides being much smaller than the Itapirema site (2.4ha vs. 10ha), and having only one ADE concentration while Itapirema has several, Calderita’s ceramics display much less decoration frequency and variability (basically slip vs. red and white painting, and incisions, in the Itapirema site), and a smaller variety of vessel types (11 versus 15). Still, Calderita could also have been a permanent settlement for a smaller population (in comparison to Itapirema). Although there are fewer vessel types than in the Itapirema site, the number of types still looks excessive for a seasonal site, and so does the existence of ADE. If so, we could be dealing with a single-house site: a rarely identified archaeological phenomenon in the Amazonian Plain. Consequently, following what we have ethnographically identified for Tupian and Carib groups, the Polychrome Tradition would not have had a fixed settlement pattern, since we can identify in the Upper Madeira linear sites (Itapirema), ring sites (the Novo Engenho Velho site studied by Pessoa da Silva, 2015), and (possibly) a singlehouse site (the Associação Calderita).

Discussion and conclusion In the humid tropics water plays a significant role: eroding river banks or creating new ones through transportation of the sediment that results from the formation of meanders and oxbow lakes. Water cuts through mountains; it separates people on opposite banks of a river; it inundates forests which become infested with fish, and fishermen. When it drops from the heights of the waterfalls, water creates great barriers for the fish, which have to swim upwards and make their way through the rocks. Humans were constantly visiting these places and marveled at the exuberance of the falls, and the abundance of fish. When it was simply

dripping from the sky, water grabbed organic content of these human occupations and carried it though almost invisible trails, into dark earth. Moreover, long-term excesses or shortages of precipitation could even make the forest expand or retreat, moving around the borders between the forest and the savanna, places constantly sought for hunting and gathering activities. Rather than opposing this magnificent force, women and men creatively sought the best ways to make use of it. Following Schaan (2010, 2013), it seems that riverine groups constantly used earthworks—mounds, middens, raised fields, artificial lakes—both inside and around their settlements, to deal with water dynamics. Groups could also simply settle over bluffs to explore and avoid water at the same time. Organic materials—such as the stilts for raising houses, sticks to enclose turtle stockyards, palm leaves for roofs, vines for fish traps, and wood and tree bark for canoes—were also important in this aquatic world. Far away, in the uplands, during the rainy season, people would unite their groups in big villages, and avail themselves of water for horticulture. This also laid the foundation for great rituals: initiations, marriages, heavy drinking, trade, and forging of alliances. When the dry season arrived and the smallest streams disappeared, the terrestrial game would concentrate on what was left of the water: this was the time for hunting and for trekkers to leave their villages behind. Furthermore, the areas with transitional vegetation, anthropogenic forests, where resources would be concentrated, would be another option for hunting and gathering (Balée, 1994). Humans often explored the trails that had been elegantly created by the water, and they made their own paths—extending the possibilities of connections between different places—and provisionally marked fluid territories. There could also be long-lasting markers within the network, such as the rock art panels found in rivers and mountains. Even so, such connections would not exclude the necessity for empty spaces—the buffer zones—which separated settlements or clusters of settlements. In this chapter I have sought to present a sample of the enormous variability of indigenous settlement patterns and of the systems surrounding them, over space and time. To find the best paths to follow, maybe Santos-Granero’s argument that “we are well advised that it is more productive to think in terms of culture areas than of language families” (2002: 27) is an appropriate starting point. The advantage of a regional overview seems clear when we look at the multi-cultural ring village patterns around the southeastern Amazonian border (those of the Tupi-Guarani, Jê, Carib, Arawak, and other groups), the scattered disposition of houses in the Guianas (those of the Carib, Arawak, Tupi-Guaranis, and other groups), or the single-house type of the western Montañas (those of the Pano, Arawak, Achuar, and other groups). However, Heckenberger (2010: 32) also seems correct when he points to a general pattern linking the Arawaks to riverine areas, and other groups—such as the Tupi-Guarani-, the Carib, the Jê-, and the Pano-speakers—to the interfluves or secondary rivers. If such a pattern is accurate, it is imperative to emphasize that throughout history all these groups, especially the Tupi-Guarani and the Caribs, have also dominated extensive segments of the major rivers, and a significant number of Arawak groups inhabited (and still inhabit) upland regions. There also seems to be a stronger link between Arawak- and Jê-speakers and ring villages, while the other groups had various possibilities of village disposal. In some places, when space was reduced—e.g. mountainous terrain, fluvial islands, or fluvial sand/silt/clay banks— we can be sure that the choices regarding village disposal did not rely only on tradition:

spheres of action influenced decisions before being transformed by them. Archaeologically, the main aim of this chapter was to provide time depth to ethnography and to present different contexts that are no longer inhabited by indigenous groups. The anthropological data falls particularly short when the discussion approaches mainstream riverine contexts, for these were the most affected by the European colonization. Archaeological research in these areas is thus of paramount importance, and the aim of presenting the data from the Upper Madeira in this chapter was to provide a contribution in this sense. Moving between multiple scales, the displayed analysis prioritized regional contexts, and the archaeology of places. Focusing on specific places in riverine regions, secondary rivers and upland areas helped us to overcome these simplistic analytical divisions, enriching them with the historical remnants of landscape building throughout different contexts of Amazonia. The discussion of the Upper Madeira archaeology, as well as other regions in Amazonia, enabled us to understand that some places are more attractive than others to humans. In the riverine areas, rapids, islands, river junctions, and oxbow lakes were sought-after places where clusters of sites showing long-term settlements are common. In secondary rivers, people also looked for these places and thus similar patterns might be found. Moving towards the interfluves, away from topographical uniformity of the plains, we might expect that geomorphological features gained importance for decisions regarding settlements, as well as the transitional areas between the tropical forest and the savanna, for they imply a bigger variety of game and edible plants. The Upper Madeira also provided us with examples of the great variability found in the tropical forest. This region possibly had linear, circular, and single-house sites. Moreover, the research provided further evidence for the existence of extensive sites, while suggesting that people seemed to be moving their settlements more often as the distance to the mainstreams increased. At the same time, the presented data indicated that the differences between primary and secondary rivers are not as big as is generally assumed. We can hence conclude, amidst the great variability of the organic world, that in the Upper Madeira and throughout Amazonia, a pattern has materialized in these persistent places, which are usually found in ecotones. The historical engagement of humans and landscape in these ecotones provided the dynamics for these patterns. It can be suggested that these are the places upon which we must now concentrate, so as to understand the histories and landscapes of the ancient people of Amazonia.

Acknowledgments I would like to thank Lucas Kellett and Eric Jones for inviting me to contribute to this volume and for their invaluable suggestions for the text. This chapter is part of the Projeto Alto Madeira (Upper Madeira Project) coordinated by Dr. Eduardo Góis Neves (Universidade de São Paulo) and myself. This research received support from Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP). I would also like to thank Susan Pyne and Jaqueline Carou for proofreading, Lorena Garcia for the Mutuca data, and all my friends from Universidade Federal de Rondônia and the ARQUEOTROP lab.

References Almeida, Fernando O. 2013 A Tradição Polícroma no Alto Rio Madeira. Doctoral thesis. MAE-USP, São Paulo. Almeida, Fernando O., Garcia, Lorena G. 2008 Aspectos do Espaço Tupinambá. Revista de Arqueologia, 21(2): 97–119. Almeida, Fernando O., Neves, Eduardo G. 2014 The Polychrome Tradition at the Upper Madeira River. In Antes de Orellana: Actas del 3er Encuentro Internacional de Arqueología Amazón ica. Rostain, Stéphen (ed.). Quito: Instituto Francés de Estudios Andinos; Facultad Latinoamericana de Ciencias Sociales, pp. 175–182. Almeida, Fernando O., Mongeló, Guilherme Z., Py-Daniel, Anne R., Moraes, Claide P. 2010 Diagnóstico Arqueológico na Unidade de Conservação de Maués, AM. Unpublished report. Arnaud, Expedito 1974 Os Índios Munduruku e o Serviço de Proteção aos Índios. Boletim do MPEG, Antropologia, 54: 1–60. Arroyo-Kalin, Manuel 2004 An Ongoing Outcome, a Surrounding World: Materiality, Agency and History. In Rethinking Materiality. Demarrais, E., Renfrew, C., Gosden, C. (eds). Cambridge: McDonald Institute Monographs, pp. 73–84. Arroyo-Kalin, Manuel 2010 A Domesticação da Paisagem: os solos Antropogênicos e o Período Formativo na Amazônia. In Arqueologia Amazônica 2. Pereira, E., Guapindaia, V. (eds). Belém: MPEG, pp. 879–908. Assis, Valéria S. 1996 Da Espacialidade Tupinambá. Masters dissertation. PUC-RS. Baldus, Herbert 1970 Tapirapé: Tribo Tupi do Brasil Central. São Paulo: Brasiliana/Editora Nacional/Edusp. Balée, William 1992 Historical Ecology of Amazonia. In Indigenous Peoples and the Future of Amazonia: An Ecological Anthropology of an Endangered World. Sponsel, L. (ed.). Tucson: University of Arizona Press, pp. 97–110. Balée, William 1994 Footprints of the Forest: Ka´apor Ethnobotany—the Historical Ecology of Plant Domestication by an Amazonian People. New York: Columbia University Press. Balée, William, Erickson, Clark 2006 Time, Complexity and Historical Ecology. In Time and Complexity in Historical Ecology. Balée, W., Erickson, C. (eds). New York: Columbia University Press, pp. 1–20. Barreto, Bruno S. 2015 Cronologia e Dinâmica de Ocupação no Sítio Laranjal do Jari I: Análise Espacial Intra-Sítio de um Assentamento Associado à Cerâmica Koriabo, baixo rio Jari, Amapá. Masters dissertation. DArq-UFS, Laranjeiras-SE. Barreto, Cristiana N.G.B. 2008 Meios Místicos de Reprodução Social: Arte e Estilo na Cerâmica Funerária da Amazônia Antiga. Doctoral thesis. MAE-USP, São Paulo. Barreto, Cristiana N.G.B. 2011 A Construção Social do Espaço: de Volta às Aldeias Circulares do Brasil Central. Habitus, 9(1): 61–79. Carneiro, Robert L. 1961 Slash-and-Burn Cultivation among the Kuikuru and its Implications for Cultural Development in Amazon Basin. In The Evolution of Cultural Systems in Native South America: Causes and Consequences, a Symposium. Wilbert, J. (ed.). Antropologica, supplement Publication, 2. Caracas: Sociedad de Ciencias Naturales La Salle, pp. 47–67. Carneiro, Robert L. 1995 The History of Ecological Interpretations of Amazonia: Does Roosevelt Have it Right? In Indigenous Peoples and the Future of Amazonia: An Ecological Anthropology of an Endangered World. Sponsel, L. (ed.). Tucson and London: Tucson Arizona Press, pp. 45–70. Carvajal Gaspar 1941 [1542] Descobrimento do Rio Amazonas. São Paulo: Companhia Editora Nacional. Cavallini, Marta S. 2014 As Gravuras Rupestres da Bacia do Baixo Rio Urubu: Levantamento e Análise Gráfica do Sítio Caretas, Itacoatiara—Estado do Amazonas. Uma Proposta de Contextualização. Masters dissertation. MAE-USP, São Paulo. Clastres, Hélène 1978 Terra Sem Mal. São Paulo: Brasiliense. Clastres, Pierre 2004 Arqueologia da Violência: Pesquisas de Antropologia Política. São Paulo: Cosac & Naify. Coudreau, Henri A. 2009 [1897] Voyage au Xingú. Cambridge: Cambridge University Press. Crevels, Mily, van der Voort, Hein 2008 The Guaporé-Mamoré as a Linguistic Area. In From Linguistic Areas to Areal Linguistics. Muysken, P. (ed.). Amsterdam/Philadelphia: John Benjamins Publishing Company, pp. 151–180. DeBoer, Warren R. 1981 Buffer Zones in Cultural Ecology of Aboriginal Amazonia: An Ethnoarchaeological Approach. American Antiquity, 46(2): 364–377. DeBoer, Warren R. 2001 The Big Drink: Feast and Forum in the Upper Amazon. In Feasts: Archaeological and Ethnographical Perspectives on Food, Politics and Power. Dietler, M., Hayden, B. (eds). Tuscaloosa: University of Alabama Press, pp. 215–239. DeBoer, Warren R., Lathrap, Donald W. 1979 The Making and Breaking of Shipibo-Conibo Ceramics. In Ethnoarchaeology: Implications of Ethnography for Archaeology. Kramer, C. (ed.). New York: Columbia University Press, pp. 102–138. DeBoer, Warren R., Kintigh, Keith, Rostoker, Arthur G. 1996 Ceramic Seriation and Site Reoccupation in Lowland South America. Latin American Antiquity, 7(3): 263–278. Denevan, William 1996 A Bluff Model of Riverine Settlement in Prehistoric Amazonia. Annals of the Association of American Geographers, 86(4): 654–681. Descola, Philippe 1996 In the Society of Nature: A Native Ecology in Amazonia. New York: Cambridge Studies in Social and Cultural Anthropology. Dias, Ondemar, Carvalho, Eliana 2008 As Estruturas de Terra na Arqueologia do Acre. In Os Geoglífos do Acre. Schaan,

D.P., Ranzi, A., Pärssinen, M. (eds). Belém: Editora UFPA. Erickson, Clark L. 2003 Historical Ecology and Future Explorations. In Amazonian Dark Earths: Origins, Properties, Management. Lehmann, J., Kern, D.C., Glaser, B., Woods, W.I. (eds). Dordrecht, Netherlands: Kluwer Press. Erickson, Clark L., Balée, William 2006 The Historical Ecology of a Complex Landscape in Bolivia. In Time and Complexity in Historical Ecology. Balée, W., Erickson, C. (eds). New York: Columbia University Press, pp. 165–186. Fausto, Carlos 2001 Inimigos Fiéis: História, Guerra e Xamanismo na Amazônia. São Paulo: Edusp. Fernandes, Florestan 1963 A Organização Social dos Tupinambá. São Paulo: Difusão Europeia do Livro. Fonseca, João A. Jr. 2013 Levantamento Regional na Arqueologia Amazônica: o Uso de Sistema de Informação Geográfica e Sensoriamento Remoto. Boletim do Museu Paraense Emílio Goeldi, Ciências Humanas, 8: 675–692. Gallois, Dominique T. 1983 A Casa dos Waiãpi. In Habitações Indígenas. Novaes, S.C. (ed.). São Paulo: Nobel/Edusp, pp. 147–168. Gallois, Dominique T. 1986 Migração, Guerra e Comércio: os Waiãpi na Guiana. Série Antropologia, 15. São Paulo: FFLCH/USP. Garcia, Lorena G. 2012 Arqueologia na Região dos Interflúvios Tocantins-Xingu: os Tupi do Cateté. Masters dissertation. MAE-USP. Gillin, John 1948 Tribes of the Guianas and the Left Amazon Tributaries. In Handbook of South American Indians III. Steward, J. (ed.). Washington, DC: Government Printing Office, pp. 799–857. Gomes, Denise M.C.; Luiz, José G. 2013 Contextos Domésticos no Sítio Arqueológico do Porto, Santarém, Brasil, Identificados com Auxílio da Geofísica por Meio do Método GPR. Boletim do Museu Paraense Emílio Goeldi, Ciências Humanas, 8: 639–656. Goulding Michael; Smith, Nigel; Mahar, Dennis J. 1996 Floods of Fortune: Ecology & Economy Along the Amazon. New York: Columbia University. Harner, Michael 1972 The Jívaro: People of the Sacred Waterfalls. Berkeley, Los Angeles, London: University of California Press. Heckenberger, Michael J. 2001 Estrutura, História e Transformação: a Cultura Xinguano na Longue Durée, 1000–2000 d.C. In Os Povos do Alto Xingu: História e Cultura. Franchetto, B., Heckenberger, M. J. (eds). Rio de Janeiro: UFRJ, pp. 21–62. Heckenberger, Michael J. 2005 The Ecology of Power: Culture, Place, and Personhood in the Southern Amazon, AD 1000–2000. New York: Routledge. Heckenberger, Michael J. 2006 History, Ecology, and Alterity: Visualizing Polity in Ancient Amazonia. In Time and Complexity in Historical Ecology. Balée, W., Erickson, C. (eds). New York: Columbia University Press, pp. 311–340. Heckenberger, Michael J. 2010 Archaeology and Cultural Memory. In Arqueologia Amazônica 2. Pereira, E., Guapindaia, V. (eds). Belém: MPEG, pp. 12–36. Heckenberger, Michael, Neves, Eduardo, Petersen, James 1998 De Onde Surgiram os Modelos? As Origens e Expansões Tupi na Amazônia Central. São Paulo-SP, USP, Revista de Antropologia, 41(1): 69–96. Hilbert, Peter P. 1968 Archäeologische Funde im Oberen Amazonas-Gebiet. Baesslerarchiv, 3: 205–255. Jaimes-Betancourt, Carla 2012 La Ceramica de La Loma Salvatierra. La Paz: KAAL-Plural Editores. Kadaré, Ismail 2007 Chronicle in Stone. New York: Arcade Publishing. Kämpf, Nestor, Kern, Dirse 2005 O Solo como Registro da Ocupação Humana Pré-histórica na Amazônia. Tópicos Ci. Solo, Belém, pp. 277–320. Kelly, Robert L. 2013 The Lifeways of the Hunter-Gatherers: The Foraging Spectrum. Cambridge: Cambridge University Press. Kipnis, Renato, Cipriano dos Santos, Hélder Bruno, Tizuka, Michelle Mayumi, Tavares de Almeida, Miguel Jorge Gomes, de Almeida, Mónica Patrícia, Corga, Silva 2013 Aplicação das Tecnologias de Modelagem 3D Conjugada às Técnicas Tradicionais para o Registro das Gravuras Rupestres do Rio Madeira, Rondônia, Brasil. Boletim do Museu Paraense Emílio Goeldi, Ciências Humanas, 8: 605–619. Ladeira, Maria E. 1983 Uma Aldeia Timbira. In Habitações Indígenas. Novaes, S.C. (ed.). São Paulo: Nobel/Edusp. Lathrap, Donald W. 1968a The Hunting Economies of the Tropical Forest Zone of South America. An Attempt at Historical Perspective. In Man the Hunter. Lee, R., Devore, I. (eds). Chicago: Aldine Publishing Company, pp. 23–29. Lathrap, Donald W. 1968b Aboriginal Occupation and Changes in River Channel on the Central Ucayali, Peru. American Antiquity, 33(1): 62–79. Lathrap, Donald W. 1970 The Upper Amazon. London: Thames & Hudson. Lathrap, Donald W. 1973 The Antiquity and Importance of Long-Distance Trade Relations in the Moist Tropics of PreColumbian South America. World Archaeology, 5(2): 170–186. Latrubesse, Edgardo M. 2012 Amazon Lakes. In Encyclopedia of Lakes and Reservoirs. Bengtsson, L., Herschy, R.W., Fairbridge, R.W. (eds). New York: Springer, pp. 13–26. Levi-Strauss, Claude 1976 Tristes Tropiques. London: Penguin Books. Lima, Helena P. 2008 História das Caretas: a Tradição Borda-Incisa na Amazônia Central. Doctoral thesis. MAE-USP, São Paulo.

López-Mazz, José M. 2010 Aldeas Matis y Paleoaldeas de las Tierras Bajas. In Arqueologia Amazônica 2. Pereira, E., Guapindaia, V. (eds). Belém: MPEG, pp. 825–852. Lowie, Robert H. 1948 The Tropical Forests: An Introduction. In Handbook of South American Indians III. Steward, J. (ed.). Washington, DC: Government Printing Office, pp. 1–56. Machado, Juliana S. 2010 Espaços Antropizados: Entendendo os Processos de Reocupação a Partir de Uma Visão Etnoarqueológica. In Arqueologia Amazônica 1. Pereira, E., Guapindaia, V. (eds). Belém: MPEG, pp. 343–364. McMichael, Crystal, Piperno, D., Neves, E., Bush, M., Almeida, F., Mongelo, G. 2015 Phytolith Assemblages along a Gradient of Ancient Human Disturbance in Western Amazonia. Frontiers in Ecological Evolution, 3: 1–15. Maybury-Lewis, David 1974 Akwe-Shavante Society. Oxford: Clarendon Press. Meggers, Betty J. 1954 Environmental Limitations on the Limitations of Culture. American Anthropologist, 56: 801–824. Meggers, Betty J. 1971 Amazônia: a Ilusão de um Paraíso. Belo Horizonte/São Paulo: Itatiaia/Edusp. Meggers, Betty J. 1992 Prehistoric Population Density in the Amazon Basin. In Disease and Demography in the Americas. Veruno, J., Ubelaker, D.H. (eds). Washington, DC: Smithsonian Institution Press, pp. 197–205. Meggers, Betty J., Evans, Clifford 1957 Archaeological Investigations in the Mouth of the Amazon. Washington, DC: Smithsonian Institution, Government Printing Press. Miller, Eurico T. 1992 Adaptação Agrícola Pré-Histórica no Alto Rio Madeira. In Prehistoria Sudamericana: Nuevas Perspectivas. Meggers, B. J. (ed.). Washington, DC: Taraxacum. Miller, Eurico T. 1999 A Limitação Ambiental como Barreira à Transposição do Período Formativo no Brasil. Tecnologia, Produção de Alimentos e Formação de Aldeias no Sudeste da Amazônia. In Formativo Sudamericano, Una Revelacions. Lederbergerber-Crespo, P. (ed.). Quito: Ediciones Abya-Yala. Miller, Eurico T. 2009 A Cultura Cerâmica do Tronco Tupi no alto Ji-Paraná, Rondônia, Brasil: Algumas Reflexões Teóricas, Hipotéticas e Conclusivas. Revista Brasileira de Linguística Antropológica, 1(1): 35–136. Miller, Eurico T. (ed.) 1992 Arqueologia nos Empreendimentos Hidrelétricos da Eletronorte. Eletronorte: Brasília-DF. Mongeló, Guilherme Z. 2015 O Formativo e os Modos de Produção: Ocupações Pré-Ceramistas no Alto Rio Madeira-RO. Masters dissertation. MAE-USP, São Paulo. Moore, Christopher R., Thompson, Victor D. 2012 Animism and Green River Persistent Places: a Dwelling Perspective of the Shell Mound Archaic. Journal of Social Archaeology, 12(2): 264–284. Moraes, Claide P. 2010 Aldeias Circulares na Amazônia Central: Um Contraste Entre a fase Paredão e a fase Guarita. In Arqueologia Amazônica 2. Pereira, E., Guapindaia, V. (eds). Belém: MPEG, pp. 74–96. Moraes, Claide P., Neves, Eduardo G. 2012 Adensamento Populacional, Interação e Conflito na Amazônia Central. Amazônica – Revista de Antropologia, 4: 122–148. Morán, Emílio 1990 A Ecologia Humana das Populações da Amazônia. Rio de Janeiro: Vozes. Myers, Thomas 1973 Toward the Reconstruction of Prehistoric Community Patterns in the Amazon Basin. In Variation in Anthropology: Essays in Honor of John C. McGregor. Lathrap, D.W., Douglas, J. (eds). Urbana: Illinois Archaeological Survey, pp. 91–105. Neves, Eduardo G. 1995 Village Fissioning in Amazonia: a Critique of Monocausal Determinism. Revista do Museu de Arqueologia e Etnologia, 5: 195–210. Neves, Eduardo G. 2006a Tradição Oral e Arqueologia no alto Rio Negro. In Amazônia, Além dos 500 Anos. Forline, L.C., Murrieta, R.S.S., Vieira, I.C.G. (eds.). Belém: Museu Paraense Emílio Goeldi. Neves, Eduardo G. 2006b Arqueologia da Amazônia. Rio de Janeiro: Jorge Zahar (Coleção Descobrindo o Brasil). Neves, Eduardo G. 2008 Ecology, Ceramic Chronology and Distribution, Long-term History, and Political Change in Amazonian Floodplain. In Handbook of South American Archaeology. Silvernan, H., Isbell, W.H. (eds). New York: Springer, pp. 359– 378. Neves, Eduardo G. 2009 Warfare in Pre-Colonial Amazonia: When Carneiro Meets Clastres. In Warfare in Cultural Context: Practice Theory and the Archaeology of Violence. Nilsen, A., Walker, W. (eds). Tucson: University of Arizona Press, pp. 139–164. Neves, Eduardo G. 2012 Sob os Tempos do Equinócio: Oito Mil Anos de História na Amazônia Central (6.500 AC–1.500 DC). Associate professor thesis. MAE-USP. Nimuendaju, Curt 1967 The Apinayé. Oosterhout: Anthropological Publications. Novaes, Silvia C. 1983 As Casas na Organização Social do Espaço Bororo. Habitações Indígenas. Novaes, S.C. (ed.). São Paulo: Nobel/Edusp, pp. 57–76. Pessoa da Silva, C.G. 2015 Os Contextos Arqueológicos e a Variabilidade Artefatual da Ocupação Jatuarana no alto Rio Madeira. Masters dissertation. UFPA. Pereira, Edithe 2004 Arte Rupestre na Amazônia: Pará. Belém/São Paulo: MPEG/UNESP. Petersen, James B., Neves, Eduardo G., Heckenberger, Michael J. 2001 Gift from the Past: Terra Preta and Prehistoric Amerindian Occupation in Amazonia. In Unknown Amazon: Culture in Nature in Ancient Brazil. Mcewan C., Barreto, C., Neves, E.G. (eds). London: British Museum Press, pp. 86–105. Porro, Antônio 1994 Social Organization and Political Power in the Amazon Floodplain: The Ethnohistorical Sources. In

Amazonian Indians from Prehistory to Present: Anthropological Perspectives. Roosevelt, A.C. (ed.). Tucson and London: University of Arizona Press, pp. 79–94. Py-Daniel, Anne R. 2015 Os Contextos Funerários na Arqueologia da Calha do Rio Amazonas. Doctoral thesis. MAE-USP, São Paulo. Rebellato, Lilian, Woods, William, Neves, Eduardo G. 2009 Pre-Columbian Settlement Dynamics in Central Amazonia. In Amazonian Dark Earths: Wim Sombroek’s Vision. Woods, W., Teixeira, W.G., Lehmann, J., Steiner, C., WinklerPrins, A.M.G.A., Rebellato, L. (eds). New York: Springer. pp. 15–31. Rice, Prudence 1987 Pottery Analysis, a Sourcebook. Chicago: University of Chicago Press. Rivière, Peter 2001 O Indivíduo e a Sociedade na Guiana. São Paulo: Edusp. Roosevelt, Anna C. 1991 Moundbuilders of the Amazon: Geophysical Archaeology on Marajo Island, Brazil. San Diego: Academic Press. Rostain, Stephen 2010 Cacicazgos Guyanenses: Mito o Realidad? In Arqueologia Amazônica 1. Pereira, E., Guapindaia, V. (eds). Belém: MPEG, pp. 163–192. Sá, Cristina 1983 Observações sobre a Habitação de Três Diferentes Grupos Indígenas Brasileiros. In Habitações Indígenas. Novaes, S.C. (ed.). São Paulo: Nobel/Edusp. Santos-Granero, Fernando 2002 The Arawakan Matrix: Ethos, Language, and History in Native South America. In Comparative Arawakan Studies: Rethinking Language Family and Culture Area in Amazonia. Hill, J., Santos-Granero, F. (eds). Urbana and Chicago: University of Illinois Press, pp. 25–50. Saunaluoma, Sanna, Schaan, Denise 2012 Monumentality in Western Amazonian Formative Societies: Geometric Ditched Enclosures in the Brazilian State of Acre. Antiqua, 2(1): 1–11. Schaan, Denise P. 2004 The Camutins Chiefdom: Rise and Development of Social Complexity on Marajó Island, Brazilian Amazon. PhD dissertation. University of Pittsburgh. Schaan, Denise P. 2010 Long-Term Human Induced Impacts on Marajó Island Landscapes, Amazon Estuary. Diversity, 2: 187–206. Schaan, Denise P. 2013 Sacred Geographies of Ancient Amazonia: Historical Ecology of Social Complexity. Walnut Creek: Left Coast Press. Schaan, Denise, Bueno, Mirian, Ranzi, Alceu, Barbosa, Antônia, Silva, Arlan, Casagrande, Edegar, Rodrigues, Alana, Dantas, Alessandra, Rampanelli, Ivandra 2010 Paisagem e Temporalidade na Construção de Espaços Sociais: O Caso dos Geoglifos do Acre. Revista da SAB, 23: 30–41. Schlanger, Sarah H. 1992 Recognizing Persistent Places in Anasazi Settlement Systems. In Space, Time, and Archaeological Landscapes. Rossignol, J., Wandsnider, L. (eds). New York: Plenum Press, pp. 91–112. Schmidt, Morgan 2010 Historical Landscapes in the Neotropics: a Model for Prehistoric Anthrosol (Terra Preta) Formation in the Upper Xingu. In Arqueologia Amazônica 2. Pereira, E., Guapindaia, V. (eds). Belém: MPEG, pp. 853–878. Schmidt, Morgan 2012 Landscapes of Movement in Amazonia: New Data from Ancient Settlements in the Middle and Lower Amazon. Papers of Applied Geography Conferences, 35: 355–364. Schmidt, Morgan 2013 Amazonian Dark Earths: Pathways to Sustainable Development in Tropical Rainforests. Boletim do Museu Paraense Emílio Goeldi, Ciências Humanas, 8(1): 11–38. Silva, Aracy L. 1983 Xavante: Casa–Aldeia–Chão–Terra–Vida. In Habitações Indígenas. Novaes, S.C. (ed.). São Paulo: Nobel/Edusp, pp. 57–76. Silva, Fabíola A. 2000 As Tecnologias e seus Significados: Um Estudo da Cerâmica dos Asurini do Xingú E da Cestaria dos Kayapó-Xikrin sob uma Perspectiva. Doctoral thesis. FFLCH-USP, São Paulo. Silva, Maurício A. 2015 Memórias e Histórias no Sudoeste da Amazônia: o Museu Regional de Arqueologia de Rondônia. Masters dissertation. MAE-USP, São Paulo. Staden, Hans 1974 [1557] Duas Viagens ao Brasil. São Paulo: Edusp. Stampanoni, Fillippo B. 2016 A Maloca Saracá. Uma Fronteira Cultural no Médio Amazonas Pré-Colonial, vista da perspectiva de uma casa. Doctoral thesis. MAE-USP, São Paulo. Stenborg, Per, Schaan, Denise, Lima, Márcio A. 2012 Precolumbian Land Use and Settlement Pattern in the Santarém Region, Lower Amazon. Amazônica: Revista de Antropologia 4: 222–250. Steward, Julian 1948 Tribes of the Montaña: An Introduction. In Handbook of South American Indians III Steward, J. (ed.). Washington, DC: Government Printing Office, pp. 507–533. Steward, Julian, Métraux, Alfred 1948 Tribes of the Peruvian and Equatorian Montaña. In Handbook of South American Indians III Steward, J. (ed.). Washington, DC: Government Printing Office, pp. 535–656. Suñer, Rodrigo A. 2015 Arqueologia Tupi no Médio Ji-Paraná: Teoria do não-Equilíbrio Dinâmico e Abordagem Multifocal dos Processos de Mobilidade Populacional no Sudoeste Amazônico. Doctoral thesis. MAE-USP, São Paulo. Tamanaha, Eduardo K., Neves, Eduardo G. 2014 800 Anos de Ocupação da Tradição Polícroma da Amazônia: Um Panorama Histórico no Baixo Rio Solimões. Anuário Antropológico, 39(2): 45–67. Tizuka, Michelle M. 2012 Geoarqueologia e Paleohifrologia da Planície Aluvial Holocênica do Rio Madeira Entre Porto Velho e Abunã/RO. Amazônica—Revista de Antropologia, 4(1): 252–257.

Trindade, Thiago B. 2015 Geoglifos, zanjas ou earthworks? Levantamento geral dos sÍtios arqueológicos com estruturas de terra em vala no médio rio Guaporé (RO) e análise comparada com os demais sítios no Sudoeste da Bacia Amazônica. Masters dissertation. MAE-USP, São Paulo. Turner, T. 1992 Os Mebengokre Kayapó: História e Mudança Social, de Comunidades Autônomas para Coexistência Interétnica. In História dos Índios no Brasil. Carneiro Da Cunha, M. (ed.). São Paulo: Companhia das Letras, pp. 311– 338. Valle, Raoni 2010 Registros Rupestres do Rio Negro, Amazônia Ocidental—Panorama Preliminar. In Arqueologia Amazônica 1. Pereira, E., Guapindaia, V. (eds). Belém: MPEG, pp. 317–342. van Velthem, Lúcia H. 1983 Onde os Wayana Penduram Suas Redes. In Habitações Indígenas. Novaes, S.C. (ed.). São Paulo: Nobel/Edusp, pp. 169–192. Vidal, Lux 1983 O Espaço Habitado entre os Kaiapó-Xikrim (Jê) e os Parakanã (Tupi) do Médio Tocantins, Pará. In Habitações Indígenas. Novaes, S.C. (ed.). São Paulo: Nobel/Edusp. Viveiros de Castro, Eduardo B. 1986 Araweté: os Deuses Canibais. Rio de Janeiro: Jorge Zahar. Wagley, Charles 1977 Welcome of Tears: The Tapirapé Indians of Central Brazil. New York: Oxford University Press. Wagley, Charles, Galvão, Eduardo 1961 Os Índios Tenetehara: Uma Cultura em Transição. Rio de Janeiro: MEC. Walker, John H. 2012 Regional Association and Ceramic Assemblage from the Fourteenth Century Llanos de Mojos. Andean Past, 10: 241–261. Weber, Ronald L.R. 1975 Caimito: An Analysis of the Late Prehistoric Culture of the Central Ucayali. PhD dissertation. University of Illinois at Urbana-Champaign. Whitehead, Neil L. 1994 The Ancient Amerindian Polities in Amazonia, the Orinoco and the Atlantic Coast: A Preliminary Analysis of Their Passage from Antiquity to Extinction. In Amazonian Indians from Prehistory to Present: Anthropological Perspectives. Roosevelt, A.C. (ed.). Tucson and London: University of Arizona Press, pp. 33–54. Whitehead, Neil L. 2002 Dark Shamans: Kainamà and the Poetics of Violent Death. Durham: Duke University Press. Wüst, Irmhild 1994 The Eastern Bororo from an Archaeological Perspective. In Amazonian Indians from Prehistory to Present. Roosevelt, A. (ed.). Tucson: University of Arizona Press, pp. 315–342. Wüst, Irmhild 1998 Continuities and Discontinuities: Archaeology and Ethnoarchaeology in the Heart of the Eastern Bororo Territory, Mato Grosso, Brazil. Antiquity, 72(277): 663–675. Wüst, Irmhild, Barreto, Cristiana N. 1999 The Ring Villages of Central Brazil: A Challenge for Amazonian Archaeology. Latin American Antiquity, 10(1): 3–23. Zedeño, Maria N., Anderson, Derek 2010 Agency and Politics in Hunter-Gatherer Territory Formation. Revista de Arqueologia, 23(1): 10–29. Zedeño, Maria N., Bowser, Brenda J. 2009 The Archaeology of Meaningful Places. In The Archaeology of Meaningful Places. Bowser, B.J., Zedeño, M.N. (eds). Salt Lake City: University of Utah Press, pp. 1–14. Zimpel Neto, Carlos A. 2009 Na Direção das Periferias Extremas da Amazônia: Estudo da Arqueologia na Bacia do Rio Jiparaná, Rondônia. Masters dissertation. MAE-USP, São Paulo. Zuse, Silvana 2014 Variabilidade Cerâmica e Diversidade Cultural no Alto Madeira. Doctoral thesis. MAE-USP, São Paulo.

Index ‘A Theory of Human Motivation’ 230 access to resources 61, 86, 89, 105, 148 Achanchi (central Andean highlands) 237–8, 240, 246 ADE (anthropogenic dark earth) 278, 280, 281, 285, 286, 289, 290–5, 297, 302 adoratorio 121, 126 agent-based modelling 16 ‘agrarian city’ model 170–1 ‘agricultural domains’ 178 aguadas 182, 184 Aguas Buenas period 196, 197, 201, 202, 206, 208, 212–15, 216–17 Allen, Kathleen M. S. 15 Amazonian settlement patterns: anthropogenic dark earth 278, 280, 281, 285, 286, 289, 290–5, 297, 302; archaeological background 285–90, 303; ethnographic background 282–5, 303; Itapirema site 298–302; Jamari River sites 290, 293–8, 301; modern day settlements 278–80; Mutuca site 287–9; settlement abandonment 285, 301; tropical forest settlement patterns 280–2; Upper Madeira sites 281, 290–2, 296–302, 304; water dynamics 302–3 AMS (accelerator mass spectrometry) dating 69–70, 73, 153, 160, 300, 301–2 Anchukaitis, Kevin J. 214 Andahuaylas (Chumbao) Valley 233–46 ‘applied settlement ecology’ 96 APUs (agricultural production units) 181–3, 185–6 archaeological landscape 5 Armillas, Pedro 141–2, 143, 145, 154 Bajío 138–9, 141–2, 151, 156, 158 Barnette, Karen 32, 34, 37, 38, 39, 42, 43, 47 Barreto, Cristiana 286 Barriles (Greater Chiriquí region) 198–9, 215, 216 behavioral ecology (BE) 6, 227 best-fit values 95, 96 biotic diversity 75 Blanton, Richard E. 126 Blitz, John H. 67, 73, 77 Bluff Model (Denevan) 286 Bolas (Greater Chiriquí region) 200, 206, 208, 209, 215

Borstein, Joshua A. 131 Brannan, Stefan 64, 67 Brouwer Burg, Marieka 14 Brown, Roy B. 142 ‘buffer zones’ 297 Butzer, Karl W. 143 Cahuachi (Ica-Nasca region) 257, 271, 272 camellones 176, 181–3 Carozza, Jean-Michel 176 Casselberry, Samuel E. 64, 66 Cemochechobee 72–3, 75, 79 central Andean highlands 231–3; Chanka settlements see Chanka settlement ecology central Arizona drought study: climate and drought measurement 92–4; correlation coefficients 96–7, 100, 101, 104; data and methodology 90–7; decision-making 85–6, 88, 91–2, 94–7, 100, 101, 104–7; discussion 104–6; drought severity index 94, 98–9, 101–2, 103; drought– population movement model 88–9, 106; environmental information 92; GIS analysis 91, 101; models, mechanisms, and expectations 88–90; population density 89–90, 97–101, 104–6; probability of equality 96–7, 100, 101, 104; resource productivity 88–9, 92, 100, 105, 106; results 97–104; riverine proximity 90, 95, 97, 101–4, 105; settlement abandonment 85–6, 88, 91–2, 95–6, 97–103, 104–5, 106; spatial units of analysis 90–1; statistical analysis 95–7; study area 86–8 Central Place Theory (CPT) 6 ceramic assemblages 38–9, 42–3, 74, 76–7, 116–17, 141, 300 Cerro Barajas (Malpaso Valley) 151, 152, 153, 156, 158, 160 Cerro de las Mesas 115, 123, 130 Chanka settlement ecology 233–5; Chanka Settlement Project 235, 236, 237–40, 241, 243; decision-making 240, 242–8; hierarchy of settlement priorities 230, 242–3, 246; regional settlement data 236–7; settlement abandonment 237, 240, 248; settlement priorities analysis 240, 242–6 Chanka–Inka war (early 1400s) 233 Chattahoochee River Valley 59, 60–1, 62–4, 71–3, 78–9 chiefdoms 58, 199–200, 208, 233 Chiriquí period 197, 199–202, 204–8, 213, 214–15, 217 Chokoltaja (Ica-Nasca region) 264–5, 266 Chupícuaro culture 141 climatic instability 17–18, 78 CO-2/CO-3/CO-4 settlements (Ica-Nasca region) 265–7 Coalescent Communities Database 90–1 Coastal Plain Red Uplands 61–2, 79 Coe, Michael D. 116 Cohen, Jacob 96 Cohen, Patricia 96

community burial grounds (Isthmo-Colombian Area) 206–11 conceptualizing settlement ecology 11–13 Cool Branch 71–2 correlation coefficients 96–7, 100, 101, 104 Covarrubias, Miguel 116 Crumley, Carol L. 5 cultural ecology 5, 229 Curré (Greater Chiriquí region) 214, 215 Dan River Valley 35–6, 42–3, 50 data quality and use 13–16 Davis jnr, R. P. Stephen 35, 37, 38 de Castro, Viveiros 282 decision-making 3–4, 10–11, 12–13, 16; central Arizona drought study 85–6, 88, 91–2, 94–7, 100, 101, 104–7; Chanka settlement ecology 240, 242–8; Piedmont Village Tradition communities 31, 37, 44; and risk 227–9, 230–1, 246–8 defining settlement archaeology 4–5 defining settlement ecology 4, 8–9, 17, 31, 57, 114, 229 dendrochronology 160 Denevan, William 286 Descola, Philippe 278, 284 deterministic models 6 digital terrain models 177 Diquís archaeological subregion 196, 198–9, 200–2, 204, 206, 208, 211–12, 213–15, 216–17 discriminant function analysis (DFA) 30, 32, 40, 44, 47–8 ‘dispersed farmstead’ sites 58, 78, 201 Donnaha site (upper Yadkin River Valley) 33–4, 38–9, 42–4, 46–7, 49 Dos Pilas (Maya Lowland site) 168 drought: central Andean highlands 232; central Arizona study see central Arizona drought study; northern frontier zone of Mesoamerica 137, 139–40, 142–3, 157–9 drought severity index 94, 98–9, 101–2, 103 drought–population movement model 88–9, 106 Drucker, Philip D. 116 ‘dry-farming’ 90 ecotones 281 El Caño (Greater Chiriquí region) 198 El Cholo (Greater Chiriquí region) 196, 200–206, 208–9, 212–13, 217 El Cóporo (Malpaso Valley) 151, 152, 153, 156, 158, 160 El Mesón area study see RAM survey El Niño events 259, 261, 267, 271, 272, 273 El Silencio (Greater Chiriquí region) 206 Elliott, Michelle 12, 148–9, 150–1, 153–5

ELPB (Eastern Lower Papaloapan Basin) region 113, 114, 119, 121, 129–31 ‘empty’ zones 178, 179, 185–6 Eno River Valley 35–6, 50 ENSO (El Niño Southern Oscillation) 261, 267 ‘environmental determinism’ 8 Epi-Olmec settlement (El Mesón area) 118–21 Etowah 64–7, 78–9 Fabaceae 151, 159 Finca 6 (Greater Chiriquí region) 206, 208, 215 Fish, Suzanne K. 4–5 Flint River 63–4 Fonseca, Oscar 216 Forbush Creek site (upper Yadkin River Valley) 33, 34, 38, 39, 42–3, 45, 46, 47, 49 Fort Walton 76 Frederick, C. D. 143 ‘Frontiering’ 10 Galop, Didier 176 Garcia, Lorena G. 289 Gary’s Fish Pond 72, 78 GIS (Geographic Information Systems) analysis: Amazonian settlement patterns 285; central Arizona drought study 91, 101; Chanka settlement ecology 242, 244, 246; and increased access to spatial modelling 7; Malpaso Valley study 148, 159; and viewshed analysis 7; Piedmont Village Tradition communities study 30, 31–2, 40; as powerful spatial analysis tool 13, 14–15, 32; and renaissance in settlement pattern studies 3 globalization 4 González, Medina 147 Gordy, R. Donald 68 GPS (Global Positioning Systems) 3, 39, 116 Greater Chiriquí region (Isthmo-Colombian Area) 195, 198–200, 206, 216; El Cholo case study 200–6, 208–9, 212–13, 217; horizontal social processes 209–11; population densities 210; subregional chronology of 196, 197; vertical settlement ecology 212–15 Grossman, David 10 hamlets 113, 118, 141, 147, 149, 201, 204, 209, 214, 216 Hardy site (upper Yadkin River Valley) 33, 34, 46, 49 Hasenstab, Robert J. 12, 15 Haudenosaunee settlements 12 Haw River Valley 35–6, 50 hazards models 16 Heckenberger, Michael J. 285, 303 Helms model 199

Helms, Mary 199 hierarchy of human needs (Maslow) 230, 242, 243, 246 hierarchy of settlement priorities 230, 242–3, 246 Hilly Gulf Coastal Plain 61, 79 historical ecology 7–8, 17, 57, 217 Hodder, Ian 14 Hoopes, John W. 216 Horn, Sally P. 214 Huarangal (Ica-Nasca region) 265, 267–8, 271–2 Huari (central Andean highlands) 231 ICA (Isthmo-Colombian Area) 195; Aguas Buenas period 196, 197, 201, 202, 206, 208, 212–15, 216–17; centralization models 196, 198–200; Chiriquí period 197, 199–202, 204–8, 213, 214–15, 217; community burial grounds 206–11; El Cholo case study 200–6, 208–9, 212–13, 217; horizontal social processes 209–11; mortuary features 198, 199–200, 201, 202, 204–6, 208, 209, 212–13, 216; population densities 210; vertical settlement ecology 212–15 Ica-Nasca region (southern coastal Peru) 255; field surveys 264–70, 271; as incomplete research area 273–4; puquios irrigation system 257, 258, 259, 271, 273; quebrada gullies 260–72; south coastal climatic patterns and subsistence 258–60; south coastal environment and settlement 255–8 Ica River (southern coastal Peru) 255–6, 259, 265, 273 infield–outfield agricultural system 167, 170–1, 179–80, 183 Israde-Alcantara, Isabel 154 Itapirema site (Amazonia) 298–302 ‘jackknifed’ topography 216 Jamari River sites (Amazonia) 290, 293–8, 301 Jiménez-Betts, P. 141 Jones, Eric E. 11, 12, 37, 39 Kadaré, Ismail 278 Kipfmueller, Kurt F. 92 Knight, Vernon J. Jr. 67, 75 Kofyar agriculturalists (Nigeria) 8–10 Kohler, Timothy A. 12 Kvamme, Kenneth L. 15 La Joyanca (Maya Lowland site) 168, 169, 170, 171–3, 174–6, 177–81, 183–5, 186–7 La Quemada (Malpaso Valley) 145–7, 148–9, 150–1, 152, 153–4, 155–8, 160 La Venta (El Mesón area) 115, 118, 125, 128–9, 131 Laguna Pacucha (Andahuaylas Valley) 233 Lake Titicaca (central Andean highlands) 231–2, 237 Lamar 76

landscape archaeology 7–8, 168 Lathrap, Donald W. 280, 281, 301 Lawson, John 40 least cost surfaces 16 Levi-Strauss, Claude 282 LIDAR (Light Detection And Ranging) surveying 171 linear regression 95 Little Ice Age 51, 141, 232 Lorenz, Karl G. 67, 73, 77 Los Pilarillos (Malpaso Valley) 150–1, 156 Lower Great Bend (upper Yadkin River Valley) 32–3, 34, 39–40, 42–3, 44, 46, 48, 49, 50–1 Lower Salt River Valley 87 lower Yadkin River Valley 51, 52 Lowie, Robert 281 Luisinayoc (central Andean highlands) 237–8 MacPherson site (upper Yadkin River Valley) 33, 34, 46, 49 Malpaso Valley 141–2, 144–8, 161; multi-proxy sedimentary study 153–5, 157, 159–60; paleoethnobotanical analysis 149–53, 156–7, 159; recent studies 148–55; spatial analysis of environmental factors 148–9, 155–6, 159 ‘managed mosaics’ 169–70 Manning, Stuart 160 Marquardt, William H. 5 Maschner, Herbert D. G. 12 Maslow, Abraham 230, 242, 243, 246 Matacapan 115, 129–30 material culture 32, 35, 58, 64, 77, 147 Maya Lowland societies: agricultural production units 181–3, 185–6; agricultural systems 167–8, 170–1, 179–83, 185–6; demographic density 180–1; human–environment interaction 169–71; infield–outfield agricultural system 167, 170–1, 179–80, 183; La Joyanca land use and settlement 178–81; methodology of study 176–8; ‘neighborhood’ units 168, 171, 178, 183–4, 185, 187; new model of agrarian settlement ecology 174–6, 183–7; previous research 167–8; Río Bec land use and settlement 181–3; spatio-temporal contexts of La Joyanca and Río Bec 171–4 Maybury-Lewis, David 282 Medieval Climatic Anomaly 232 Meggers, Betty J. 280, 286 ‘Mesoamericanization’ 141, 155 Métailié, Jean-Paul 176 Metcalfe, Sarah E. 142 microtopographic surveys 177 middle phase (Singer-Moye site) 67, 68, 73–6

Mikell, Gregory A. 34 Miller, Eurico 297, 300–1 Mistovich, Tim S. 75 Monte Carlo methods 15 monumentality 59, 200, 217, 287 Moraes, Claide P. 286 mortuary features (Isthmo-Colombian Area) 198, 199–200, 201, 202, 204–6, 208, 209, 212–13, 216 Moundville 60, 64–7, 78–9 Mouzon, Henry Jr. 40 Mutuca site (Amazonia) 287–9 Myers, William E. 40 ‘neighborhood’ units (Maya Lowland ​societies) 168, 171, 178, 183–4, 185, 187 Netting, Robert 8 Neves, Eduardo 284 Newkirk, Judith 37, 38 northern frontier zone (of Mesoamerica) 138–40; conflict/violence 141, 145, 148–9, 155–6; drought 137, 139–40, 142–3, 157–9; future study ​directions 159–60; increased research into 137–8; initial archaeological and ​paleoenvironmental research 140–4; Malpaso Valley studies see Malpaso Valley; radiocarbon dating 142, 145, 153; settlement abandonment 137–8, 140, 142, 145, 148, 153, 157, 158, 160; subsistence farming 138, 139–40, 148, 157 obsidian 121, 128–9, 130, 131 Olmec settlement (El Mesón area) 117–18 operationalizing settlement ecology 13–16 optimal foraging theory (OFT) 6 origins of settlement ecology 5–8 Orton, Clive 14 PAA (Andahuaylas Archaeological Project) 236 PAC (Chanka Settlement Project) 235, 236, 237–40, 241, 243 paleoclimate data 14, 87, 105–6 paleo-environmental data 14 Panteón de La Reina (Greater Chiriquí region) 199 Parker site (upper Yadkin River Valley) 33, 34, 38, 51 pastizal 145 Pataraya Chico (Ica-Nasca region) 263–4 Pataula Creek (Singer-Moye) 60, 73 Payne (upper Yadkin River Valley) 51 Pazos, Miguel 264 pedological test pits 177, 184

Pee Dee communities (upper Yadkin River Valley) 51 ‘persistent monumental places’ 59 Petexbatún region (Maya Lowlands) 168, 170, 171 petroglyphs 212 phytoliths 153, 154, 160 Piedmont Village Tradition (PVT) communities 29–31; background of study 31–7; conclusions 51–3; culture history 35–7; and decision-making 31, 37, 44; discriminant function analysis 30, 32, 40, 44, 47–8; discussion of study results 46–51; ecology of settlement change 50–1; geographic information systems data 30, 31–2, 40; historic settlement pattern research 29–30, 32–5; landscape variables/characteristics 40, 41–2; location choice 31–2, 44; material culture 32, 35; methodology of study 37–40, 41–2; modeling settlement change over time 46–50; radiocarbon dating 33–5, 42, 49; settlement ecology theory and methods 31–2; study results 40, 42–6; surface scatter data 34, 37–9 pirka architecture 238, 243, 247 platform mounds: El Cholo 200–201; El Mesón area 116, 118, 119–21, 122–4, 129, 130; Singer-Moye 58, 59, 67–71, 74–6 Pluckhahn, Thomas J. 59 Pool, Christopher A. 121, 124, 126, 129 population and agricultural production 9–11 population density: central Arizona drought study 89–90, 97–101, 104–6; Isthmo-​Colombian Area 210; Maya Lowland societies 180–1; Singer-Moye 64, 66–7, 71–2 Porter site (upper Yadkin River Valley) 33, 34–5, 39, 42–4, 46, 49, 52 postmodern critique 15–16 postmolds 34–5 predictive modeling 7, 12 probability of equality 96–7, 100, 101, 104 processual archaeology 5–7, 11 proximity principle 6, 31 puna zone (Andahuaylas Valley) 233, 235, 236, 238–9, 241, 244, 245 puquios irrigation system 257, 258, 259, 271, 273 push and pull factors 9, 12, 229, 242 puxadinho 278 Py-Daniel, Anne R. 286 Quebrada Campanayoq (Ica-Nasca region) 262–3, 270 quebrada gullies 260–72 Quebrada La Yesera (Ica-Nasca region) 260–1, 272 Quebradas Cocharcas (Ica-Nasca region) 264–7, 270, 271 Quebradas Tingue (Ica-Nasca region) 264, 265, 267–70, 271 quechua zone (Andahuaylas Valley) 233, 235, 236, 238–9, 241, 244, 245 radiocarbon dating 33–5, 42, 49, 67, 69–70, 73–4, 76, 142, 145, 153, 289 ‘rain-fed’ farming 90

RAM (Recorrido Arqueológico El Mesón) survey: environmental setting and background 114–17; Epi-Olmec settlement 118–21; exclusionary strategies 126, 127; importance of El Mesón location 113–14, 127–8, 129, 131–2; Late Classic and Postclassic settlement 124–5; monumental art 116, 126–7, 128; Olmec settlement 117–18; persistence of settlement in El Mesón 113–14; platform mounds 116, 118, 119–21, 122–4, 129, 130; political strategies 125–31; Protoclassic and Early Classic settlement 121–4; settlement history of El Mesón area 113–14, 117–25; Standard Plan layout 123–4; survey methods 116–17; Tres Zapotes regional polity 114, 119, 121, 124, 125–9, 131; TZPG complex 119–21, 122, 123, 124, 126–7 Redtail site (upper Yadkin River Valley) 33, 34, 35, 39, 42, 44, 46, 49 regional perspective 3, 5 Reindel, Markus 259 remotely sensed data 3 resource marginality 89 resource productivity 88–9, 92, 100, 105, 106 ring villages 282, 285–6 Río Bec (Maya Lowland site) 168, 169, 171–2, 173–4, 176–8, 181–4, 185–6, 187 Rio Pampas (Andahuaylas Valley) 233 risk 227–31, 242, 246–8 risk management 156, 273 riverine proximity (central Arizona drought study) 90, 95, 97, 101–4, 105 Rivière, Peter 284, 289 Rogers, Rhea J. 37 Rood’s Landing 72–3, 75–6, 78, 79 Rowe, J. H. 259 Russell, Margaret C. 68 Salzer, Matthew W. 92 Santley, Robert S. 130 savoir-faire 170 Schaan, Denise P. 302 Schnell, Frank T. 72, 73 Scott, John 116, 122 semi-sedentism 8, 31, 63, 140, 212, 215–16, 227 settlement abandonment: Amazonian settlement patterns 285, 301; central Arizona drought study 85–6, 88, 91–2, 95–6, 97–103, 104–5, 106; Chanka settlement ecology 237, 240, 248; northern frontier zone of Mesoamerica 137–8, 140, 142, 145, 148, 153, 157, 158, 160 Settlement Ecology: The Social and Spatial Organization of Kofyar Agriculture 8–9 SFP (San Francisco Peaks) precipitation reconstruction 92–4 Shapiro, Gary 75

shovel-test data 59–60, 64, 66–7, 71, 76 Silverman, H. 259 Simpkins, Daniel L. 29 Singer-Moye: AMS dating 69–70, 73; animal and plant species 63, 75; conceptual framework of study 57–9; early phase 67–73; environmental and cultural context 59–64; late phase 67, 68, 76–9; material culture 58, 64, 77; middle phase 67, 68, 73–6; mound centers 58, 59, 61, 63, 64, 72, 75, 76–8; platform mounds 58, 59, 67–71, 74–6; population estimates 64–6, 67, 71–3, 75–6, 78; radiocarbon dating 67, 69–70, 73–4, 76; settlement history reconstruction 64–79; shovel-test data 59–60, 64, 66–7, 71, 76 Site Catchment Analysis (SCA) 6, 238 ‘siteless’ survey technique 116 Sitio Batambal (Greater Chiriquí region) 206, 208 Sitio Cantarero (Greater Chiriquí region) 200 Sitio Conte (Greater Chiriquí region) 198 Sitio Grijalba (Greater Chiriquí region) 206, 208 Sitio Zoncho (Greater Chiriquí region) 199 southern coastal Peru: climatic patterns and subsistence 258–60; environment and settlement 255–8; quebrada gullies 260–72 spatial orientation 5 ‘special residence’ pattern 201–2 Stahle, D. W. 160 Standard Plan layout (El Mesón area) 123–4 Stark, Barbara L. 116, 130 Stein, Jeffrey W. 12 stelae 122 Stirling, Matthew 116 Stone, Glenn Davis 8–11, 31 Stoner, Wesley D. 130 subsistence farming 29, 35–6, 138, 139–40, 148, 157 suni zone (Andahuaylas Valley) 233, 235, 236, 238–9, 241, 244, 245 surface scatter data 34, 37–9 swidden agriculture 170, 179 systems theory 5 T. Jones site (upper Yadkin River Valley) 33, 34–5, 39, 42–4, 46, 49, 52 Talamanca region (Isthmo-Colombian Area) 202, 209, 212–14, 215 Teotihuacán 129–30, 137, 142, 158, 162 terra firme 280 Thiessen polygons 66–7 Thomas, Cyrus 32 Thompson, Victor D. 59 TI-1 settlement (Ica-Nasca region) 265, 267 TI-3 settlement (Ica-Nasca region) 265, 268, 269, 271

TI-4 settlement (Ica-Nasca region) 265, 268–70 Tiwanaku (central Andean highlands) 231–2 TL (thermoluminescence) dating 289 ‘total’ landscapes 14 Totocapan 115, 129–30 Town Creek (upper Yadkin River Valley) 51 Trans-Mexican Volcanic Belt 142 Tres Zapotes 114, 117–21, 124, 125–30, 131 Trombold, Charles D. 154 ‘tropical forest culture’ 280 Tupi-Guarani groups (Amazonia) 282–4, 287 Turkon, Paula 149 Turner, T. 282 TZPG (Tres Zapotes Plaza Group) complex 119–21, 122, 123, 124, 126–7 UAV (unmanned aerial vehicles) 3 UGA (University of Georgia) 59–60 ungwa 10 Upper Great Bend (upper Yadkin River Valley) 32, 34, 39–40, 42, 43, 44, 46, 48, 49, 50–1 Upper Madeira sites (Amazonia) 281, 290–2, 296–302, 304 upper Yadkin River Valley (UYRV) 30, 32–53 US Geological Survey 91 Vannière, Boris 176, 177 várzeas 280 Vaughn, K. J. 270 viewshed analysis 7, 41, 47, 48, 242, 243–4 Ward, H. Trawick 35, 37, 38 Wari period (central Andean highlands) 231–2, 236 Wendt, Carl J. 117 ‘wiggle-match’ dating methods 160 Williams, León 264 Williams, Mark 75 wood charcoal 70, 150–1, 152, 156, 158, 160 Woodall, Ned 30, 32–5, 36–9, 42–3, 46–7, 49–50 Woodland Period 32–4, 35–7, 42, 43, 46–7, 49, 51 Wüst, Irmhild 285–6 yunga zone (Andahuaylas Valley) 233, 235, 238–9, 241, 244, 245 Zacatecas 139–40, 142, 144–5, 148 Zaragoza-Oyameles 115, 130 Zuse, Silvana 297