Human Interactions with the Geosphere: The Geoarchaeological Perspective
The Geological Society of London Books Editorial Committee Chief Editor
Bob Pankhurst (UK) Society Books Editors
John Gregory (UK) Jim Griffiths (UK) John Howe (UK) Rick Law (USA) Phil Leat (UK) Nick Robins (UK) Randell Stephenson (UK) Society Books Advisors
Mike Brown (USA) Eric Buffetaut (France) Jonathan Craig (Italy) Reto Giere´ (Germany) Tom McCann (Germany) Doug Stead (Canada) Gonzalo Veiga (Argentina) Maarten de Wit (South Africa)
Geological Society books refereeing procedures The Society makes every effort to ensure that the scientific and production quality of its books matches that of its journals. Since 1997, all book proposals have been refereed by specialist reviewers as well as by the Society’s Books Editorial Committee. If the referees identify weaknesses in the proposal, these must be addressed before the proposal is accepted. Once the book is accepted, the Society Book Editors ensure that the volume editors follow strict guidelines on refereeing and quality control. We insist that individual papers can only be accepted after satisfactory review by two independent referees. The questions on the review forms are similar to those for Journal of the Geological Society. The referees’ forms and comments must be available to the Society’s Book Editors on request. Although many of the books result from meetings, the editors are expected to commission papers that were not presented at the meeting to ensure that the book provides a balanced coverage of the subject. Being accepted for presentation at the meeting does not guarantee inclusion in the book. More information about submitting a proposal and producing a book for the Society can be found on its web site: www.geolsoc.org.uk.
It is recommended that reference to all or part of this book should be made in one of the following ways: Wilson, L. (ed.) 2011. Human Interactions with the Geosphere: The Geoarchaeological Perspective. Geological Society, London, Special Publications, 352. Heinzel, C. & Kolb, M. 2011. Holocene land use in western Sicily: a geoarchaeological perspective. In: Wilson, L. (ed.) Human Interactions with the Geosphere: The Geoarchaeological Perspective. Geological Society, London, Special Publications, 352, 97– 107.
GEOLOGICAL SOCIETY SPECIAL PUBLICATION NO. 352
Human Interactions with the Geosphere: The Geoarchaeological Perspective
EDITED BY
L. WILSON University of New Brunswick in Saint John, Canada
2011 Published by The Geological Society London
THE GEOLOGICAL SOCIETY The Geological Society of London (GSL) was founded in 1807. It is the oldest national geological society in the world and the largest in Europe. It was incorporated under Royal Charter in 1825 and is Registered Charity 210161. The Society is the UK national learned and professional society for geology with a worldwide Fellowship (FGS) of over 10 000. The Society has the power to confer Chartered status on suitably qualified Fellows, and about 2000 of the Fellowship carry the title (CGeol). Chartered Geologists may also obtain the equivalent European title, European Geologist (EurGeol). One fifth of the Society’s fellowship resides outside the UK. To find out more about the Society, log on to www.geolsoc.org.uk. The Geological Society Publishing House (Bath, UK) produces the Society’s international journals and books, and acts as European distributor for selected publications of the American Association of Petroleum Geologists (AAPG), the Indonesian Petroleum Association (IPA), the Geological Society of America (GSA), the Society for Sedimentary Geology (SEPM) and the Geologists’ Association (GA). Joint marketing agreements ensure that GSL Fellows may purchase these societies’ publications at a discount. The Society’s online bookshop (accessible from www.geolsoc.org.uk) offers secure book purchasing with your credit or debit card. To find out about joining the Society and benefiting from substantial discounts on publications of GSL and other societies worldwide, consult www.geolsoc.org.uk, or contact the Fellowship Department at: The Geological Society, Burlington House, Piccadilly, London W1J 0BG: Tel. þ 44 (0)20 7434 9944; Fax þ 44 (0)20 7439 8975; E-mail:
[email protected]. For information about the Society’s meetings, consult Events on www.geolsoc.org.uk. To find out more about the Society’s Corporate Affiliates Scheme, write to
[email protected] Published by The Geological Society from: The Geological Society Publishing House, Unit 7, Brassmill Enterprise Centre, Brassmill Lane, Bath BA1 3JN, UK (Orders: Tel. þ 44 (0)1225 445046, Fax þ 44 (0)1225 442836) Online bookshop: www.geolsoc.org.uk/bookshop The publishers make no representation, express or implied, with regard to the accuracy of the information contained in this book and cannot accept any legal responsibility for any errors or omissions that may be made. # The Geological Society of London 2011. All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No paragraph of this publication may be reproduced, copied or transmitted save with the provisions of The Copyright Licensing Agency Ltd, Saffron House, 6–10 Kirby Street, London EC1N 8TS UK. Users registered with the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923, USA: the item-fee code for this publication is 0305-8719/11/$15.00. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. ISBN 978-1-86239-325-7 Typeset by Techset Composition Ltd, Salisbury, UK Printed by MPG Books Ltd, Bodmin, UK Distributors North America For trade and institutional orders: The Geological Society, c/o AIDC, 82 Winter Sport Lane, Williston, VT 05495, USA Orders: Tel. þ 1 800-972-9892 Fax þ 1 802-864-7626 E-mail:
[email protected] For individual and corporate orders: AAPG Bookstore, PO Box 979, Tulsa, OK 74101-0979, USA Orders: Tel. þ 1 918-584-2555 Fax þ 1 918-560-2652 E-mail:
[email protected] Website: http://bookstore.aapg.org India Affiliated East-West Press Private Ltd, Marketing Division, G-1/16 Ansari Road, Darya Ganj, New Delhi 110 002, India Orders: Tel. þ 91 11 2327-9113/2326-4180 Fax þ 91 11 2326-0538 E-mail:
[email protected] Contents Preface WILSON, L. The role of geoarchaeology in extending our perspective
vii 1
TRAPEZNIKOVA, O. N. Environmental limitations on agricultural development of the forest zone of the East European Plain (Russian Federation)
11
KRAFT, J. C., RAPP, G., BRU¨KNER, H. & KAYAN, ˙I. Results of the struggle at ancient Ephesus: natural processes 1, human intervention 0
27
VATTUONE, M. M. S. & NEDER, L. Quaternary landscape evolution and human occupation in northwestern Argentina
37
GILLMORE, G. K., STEVENS, T., BUYLAERT, J. P., CONINGHAM, R. A. E., BATT, C., FAZELI, H., YOUNG, R. & MAGHSOUDI, M. Geoarchaeology and the value of multidisciplinary palaeoenvironmental approaches: a case study from the Tehran Plain, Iran
49
GILLMORE, G. K. & MELTON, N. Early Neolithic sands at West Voe, Shetland Islands: implications for human settlement
69
LIU, Z., SUN, H., LI, H. & WAN, N. d13C, d18O and deposition rate of tufa in Xiangshui River, SW China: implications for land-cover change caused by climate and human impact during the late Holocene
85
HEINZEL, C. & KOLB, M. Holocene land use in western Sicily: a geoarchaeological perspective
97
HILL, C. L., RAPP, G. & JING, Z. Alluvial stratigraphy and geoarchaeology in the Big Fork River Valley, Minnesota: human response to Late Holocene environmental change
109
RAPP, G. & JING, Z. Human –environment interactions in the development of early Chinese civilization
125
RAAB, A., BRU¨TZKE, W., CHRISTOPHEL, D., VO¨LKEL, J. & RAAB, T. Reconstruction of the fire history in the Siedlungskammer Burgweinting (Bavaria, Germany) in relation to settlement and environmental history
137
WILSON, L. Raw material economics in their environmental context: an example from the Middle Palaeolithic of southern France
163
ROBERTSON, E. C. Reassessing Hypsithermal human– environment interaction on the Northern Plains
181
Index
195
Preface Human impact on our environment, resulting in climate change, deforestation, desertification, soil erosion and other effects, is not a new phenomenon. For millennia, humans have been coping with, or provoking, environmental change. We have exploited, extracted, over-used but also in many cases nurtured the resources that the geosphere offers. Geoarchaeologists study the traces of human interactions with the geosphere dating back to ancient times, as well as up to and in the present. Geoarchaeological investigations provide the key to recognizing landscape and environmental change, and human impacts as a result of such things as subsistence and resource exploitation activities, settlement location, and local and regional landuse patterns. This approach also allows us to determine the effects of environmental change on human societies. This volume is a collection of papers from around the world, including both case studies and broader reviews of regions or of geoarchaeology as a whole, covering the time period since before modern human beings came into existence up to the present day. The papers look at how human land use has affected the environment, and how
environmental characteristics have affected human land use, as well as how geoarchaeology itself can help us elucidate these interactions. To understand ourselves, we need to understand that our world is constantly changing, and that change is dynamic and complex. Geoarchaeology provides an inclusive and long-term view of human–geosphere interactions, and serves as a valuable aid to those who try to determine sustainable policies for the future. This book has greatly benefited from the hard work of many people, including the ever-helpful staff at the Geological Society of London Publishing House. The editor wishes particularly to thank all of the contributors for giving us their ‘good’ stuff, and the multitude of peer reviewers, a very diverse group of people, who uniformly provided thoughtful, careful, and very detailed reviews of the papers. The early stages of preparation of this book benefited greatly from the hard work and careful eye of Pamela J. Dickinson, who acted as Volume Editor for five papers. Pam had to withdraw from the project owing to time constraints and other commitments, but her contribution is greatly appreciated, and her collaboration always makes projects more enjoyable. Thanks, Pam.
The role of geoarchaeology in extending our perspective LUCY WILSON Department of Biology, University of New Brunswick in Saint John, PO Box 5050, 100 Tucker Park Road, Saint John, N.B. E2L 4L5, Canada (e-mail:
[email protected]) Abstract: Specialists and the general public alike are very aware of human impacts on our environment. Climate change, deforestation, desertification, soil erosion and other topics are currently much in the news, but human influence on the environment is not a new phenomenon. Geoarchaeologists study the traces of human interactions with the geosphere dating back to ancient times, as well as up to and in the present. Geoarchaeological investigations provide the key to recognizing landscape and environmental change within a region, as well as reconstructing ancient landscapes and palaeoclimatic regimes. Such an interdisciplinary approach makes it possible to interpret the ways that humans affect the geosphere, through such things as subsistence and resource exploitation activities, settlement location, and local and regional land-use patterns. This approach also allows us to determine the effects of environmental change on human societies. For millennia, humans have been coping with, or provoking, environmental change. We have exploited, extracted, over-used but also in many cases nurtured the resources that the geosphere offers. In the geoarchaeological perspective, human life has never been separate from nature. Geoarchaeology can thus provide a more inclusive and longer-term view of human– geosphere interactions, and serve as a valuable aid to those who try to determine sustainable policies for the future.
The world seems to be changing: the weather is messed up, climate is different, sea level is rising, glaciers are melting, new plants and animals are showing up in places where they never used to be . . . A lot of people are worried. They feel as if we are in uncharted waters. For many people, there is an underlying assumption that the environment is ‘supposed to be’ stable. The mere fact of it changing is frightening. They feel that we have never had to deal with such changes before, and they do not know how we will be able to handle it. This perspective is understandable, as what most people know of the world is based on their own experiences, with a smattering of history tacked on. Our own experiences are naturally very limited, no more than one lifetime (plus what our parents and grandparents tell us). The perspective that history can provide is not much better. History, which depends on written documents, is in fact very short: at best no more than a few thousand years, and most of that very sketchy. For the most part we have only a few snippets of information about any one time, not an overall view. In addition, if we want to learn about the environment, historical documents tend to focus on the wrong details. They tell us about kings and wars and trade amounts, and they give us only the part of the story that one small subset of participants wanted to tell. We have to work hard to extract information about the environment from such sources. History also has the drawback of having taken place during relatively recent times, when climate
was in fact fairly stable. The little information that it does give us is, therefore, misleading when it comes to understanding the real behaviour of the environment. As geologists and archaeologists know, climate and the environment are anything but stable: change is the real constant. So although historians can tell us that during the Little Ice Age glaciers in the Alps advanced many kilometres down valley, and that during the preceding, warmer, period the Romans grew grapes and made wine in Britain, what they cannot tell us is that even those changes were minor, compared with what went on in prehistoric times. Archaeology extends our knowledge of human activities back through time, well beyond the limits of history. It uses the physical traces of those activities to reconstruct past behaviours and cultures. However, although archaeologists are generally well aware of the need to place their finds in an environmental context, their main focus is culture and culture change, not environment and environmental change. Geology on the other hand studies the planet Earth, and therefore tells us how climate and environments have changed through time, but not how those changes have affected people nor how people have reacted to them. Geoarchaeology bridges the gap, and studies human behaviour in its dynamic environmental context, over the full time range of the existence of human beings. It shows us that we have, in fact, lived through many major climatic and environmental changes, some of them very rapid. We have caused
From: Wilson, L. (ed.) Human Interactions with the Geosphere: The Geoarchaeological Perspective. Geological Society, London, Special Publications, 352, 1 –9. DOI: 10.1144/SP352.1 0305-8719/11/$15.00 # The Geological Society of London 2011.
2
L. WILSON
some of them, we have exploited some of them, and we have suffered from some of them. Yes, we should be concerned about our current situation, but we have faced instability before. Geoarchaeological case studies, such as those in this book, show us that we can survive, as long as we are willing to adapt.
What is geoarchaeology? Geoarchaeology (sometimes called archaeological geology), at its most basic, can be described as the application of geological techniques to answer archaeological questions. This is an insufficient definition, however, as it makes geoarchaeology sound like the dependent child of its two parents: merely a set of tools, with no independent existence, no theory of its own, no questions of its own to answer. In fact, geoarchaeology today is developing into a coherent discipline with dedicated practitioners: it is no longer ‘geologists who help out archaeologists’, nor ‘archaeologists with an interest in geology’. It is not something that lies between two disciplines, and therefore is neither: it is a discipline of its own that bridges its two parents and draws on aspects of both. Geoarchaeologists devote themselves to deciphering both the natural world and the ways in which humans in the past interacted with it. The papers in this book show the current state of the discipline. Some of the papers (those by Gillmore et al., Heinzel & Kolb, Liu et al., Raab et al., Sampietro & Neder and Trapeznikova) were presented at one of several geoarchaeological sessions at the 33rd International Geological Congress in Oslo, Norway, in August 2008, whereas the others have been contributed since that meeting. Geoarchaeological sessions can also be found at archaeological conferences, and at solely geoarchaeological conferences such as those of the Developing International Geoarchaeology (DIG) series (Wilson et al. 2007, and see also the proceedings of DIG 2007, published as volume 78 of Catena). The roots of geoarchaeology can be traced back to the beginnings of archaeology itself. As Butzer (1982, p. 4) noted, ‘environmental archaeology is one of the oldest interdisciplinary bridges in the field. Archaeologists have always been conscious of environmental context.’ As an identified discipline, however, geoarchaeology is generally considered to have developed starting in the 1960s (Rapp & Hill 2006; Rapp 2007), as part of the growing perception that the data and perspectives of the sciences had to be more systematically incorporated within archaeology. A classic publication of the time is that of Clarke (1968), who pointed out (p. 636) that ‘the relationship between
analysts and their data may be as much enlightened by simple changes in viewpoint as by direct augmentation of the quantity of data. Archaeologists have concentrated far too much upon increasing the quantity of their data and far too little upon increasing the quality of their conceptual apparatus.’ In the early days geoarchaeology probably still was the dependent child, contributing to adding more data and not to changing the conceptual apparatus. By 1982 Butzer (1982, p. 4) was bemoaning the fact that ‘What remains poorly articulated is the equally fundamental environmental dimension . . . [There is] . . . substantial empirical input from those involved in the applied sciences, who nevertheless have little impact on the dominant intellectual currents within archaeology. Perhaps the environment is taken for granted.’ Substantial empirical input remains essential, of course: understanding the place of cultures within their environments requires accurate environmental reconstruction (French 2003; Brooks 2006; Rapp & Hill 2006; Rech et al. 2007; Heinzel & Kolb). But we are now also moving towards fulfilling the rest of Butzer’s plea. Most geoarchaeologists probably agree with Butzer (1982, p. 4) that ‘the concept of environment should not be considered synonymous with a body of static, descriptive background data. The environment can indeed be considered as a dynamic factor in the analysis of archaeological context.’ It is that concept of the environment as a dynamic and ever-changing participant in human lives and cultures, and vice versa, that geoarchaeology can bring to the forefront. In our work we help to eliminate the false dichotomy between ‘nature’ and ‘culture’ that Head (2008) decried: false, because culture has never existed divorced from nature, and there is not now (and has not been for a very long time) any such thing as pure nature, unaffected by human activity. Humans are not outside nature. We and nature are intimately involved with each other, so, as Head (2008, p. 376) stated, we need to think about and explain our world ‘relationally, in terms of associations rather than separations’. In addition, the geoarchaeological perspective involves an environment that is seen not only as that existing in a single location at a single time, but as one that extends both spatially and temporally outwards in all directions from the location and age of the remains we study. This was well expressed by Rapp (2007, p. 4), who stated that ‘human social frameworks and their natural environments have co-evolved through time. A thorough understanding of culture and culture change is not possible without an appreciation of the environmental context. Humans wandered and worked across, and made an impact on, a continuous landscape. It is with the continuous
INTRODUCTION
space– time landscape that geoarchaeologists must and archaeologists should deal.’ Geoarchaeology thus emerges from the interplay between the human-centred discipline of archaeology, with its concern for culture and cultural evolution, and the more empirical, nature-focused science of geology. It has as its central tenet the belief that human life is not separate from the natural world, and cannot be separated from it. This is not geographical determinism: we do not claim that humans behave the way they do because the environment makes them do it. But it does mean that the factors influencing human behaviour include natural, environmental, ones. In common with our parent discipline of geology, geoarchaeologists tend to believe that these natural factors can be measured and understood, although they may be fluctuating and complex. We also believe that for much of the time of human existence, the natural environment has, in turn, been influenced by continuously varying human cultural influences. We take heed of Butzer’s (1982) warning, and aim to chart and understand those interactions, through time and in each area and case, without treating either (human culture or the environment) as a static constant. If we can ultimately extend that understanding into widely applicable generalities, or even ‘laws’, so much the better, although we are not yet close to achieving that goal. The recent literature, conference presentations, and the contents of this book exemplify the current state of the discipline. Nowadays some geoarchaeological studies include very detailed, small-scale technical analyses of samples from archaeological sites, such as sediments (e.g. Rapp & Hill 2006; Hill 2007a; Gillmore & Melton) and micromorphology thin sections (e.g. French 2003; Guttmann et al. 2006, 2008; Simpson et al. 2006; Huisman et al. 2009; Sageidet 2009). Others range in scale up to examinations of entire regions over long time periods (e.g. Goman et al. 2005; van der Leeuw et al. 2005; Brooks 2006; Rapp & Jing). They may be concerned with archaeological sites that are tens to hundreds of thousands of years old (e.g. Hill 2007a, b; Rech et al. 2007; Wilson), from only hundreds of years old up to the present (e.g. Guttmann et al. 2006; Sandor et al. 2007), or any age range between them. Geoarchaeologists are also aware of the importance of information from sources beyond archaeological sites: regional geomorphology, non-cultural stratigraphic sequences, and so on (e.g. French 2003; Rapp & Hill 2006; Wilson 2007a, b; Carson 2008; Kraft et al.; Liu et al.). To determine the context of past human lives we study climatic conditions, and also such aspects as soils, geomorphology, hydrology, and the nature
3
and distribution of natural resources. A quick glance through some recent papers reveals work being done, for instance, on how changing lake levels relate to the distribution of archaeological sites in Maine, USA (Pelletier et al. 2007) and around the Aral Sea (Boroffka et al. 2006). Past changes to sea level have been shown to have had an impact on the disappearance of the Lapita culture in New Caledonia (Carson 2008), on the distribution of Italian Neolithic sites (Lorusso 2007) and prehistoric American sites (Leach & Belknap 2007), on Chinese civilizations (Zhang et al. 2005), and on the use of alluvial gold deposits in prehistoric Ireland (Moore 2006). Problems of increasing aridity have received attention in many parts of the world, including North America (French 2007; Huckleberry & Duff 2008), South America (Neme & Gil 2009; Sampietro & Neder), and Europe and the Middle East (Brooks 2006; French 2007; Gillmore et al. 2007; Gillmore et al.). River courses are also notoriously changeable, and have had their effect on cultures through time and around the world (e.g. Howard et al. 2003; Morozova 2005; Adelsberger & Kidder 2007; Hill et al.; Rapp & Jing). Work on natural resources ranges from strict analysis of, for instance, the petrography and geochemistry of lithic material, metals or ceramics (e.g. Luedtke 1992; Rapp et al. 2000; Moody et al. 2003) to determine the natural resources used for their manufacture, into aspects that have some bearing on geological questions of petrogenesis, and also into more archaeological aspects concerning territories and economic strategies (Black & Wilson 1999; Wilson 2007a, b). In this domain, too, geoarchaeology’s contribution to elucidating past human behaviour is essential. In addition, geoarchaeology can contribute to deciphering how past resource use has had an impact on the natural environment itself. Archaeologists are good at determining past behaviour, usually (but not always) at specific locations (sites). Geoarchaeology is and has been good at reconstructing environments and determining what resources were available and used, so that those sites can be put in their environmental context. We are aware that it is important to consider not only what we find, but where we find it, if we want to understand the interactions between people and nature. The study by Robertson provides a good example of the importance of this. We try to understand what aspects of the environment and which resources were desirable, and how they were being exploited. For instance, studies can aim at determining which landforms in a region were considered suitable for settlements (e.g. Oetelaar 2004; Evans et al. 2007; Trapeznikova) or can explain, through the reconstruction
4
L. WILSON
of past geomorphology or hydrology, why we find sites where we do (e.g. Rømer et al. 2006; Rech et al. 2007). Based on such data, we can extend our research into other areas and try to find more sites, which will give us more information about what people were doing and how they and their environment were mutually influencing each other at various times. Although many studies still restrict themselves to presenting the geological data, without making suggestions as to the human side of things, geoarchaeology has gradually been moving beyond its simple descriptive phase. As many of the papers presented here show, we are starting to develop a body of ‘middle range’ theory: theory that links the reconstructed environment and resources to the cultural remains that we find, and allows us to interpret past behaviour. For instance, my own work stems from studying the provenance of raw materials used for stone tools. It is based on geological theory, in the belief that rocks that formed in different places and/or at different times will have some identifiable physical or chemical difference from each other. Finding those differences allows us to trace archaeological materials back to their source outcrops. This feeds into archaeological interpretation, in the context of theory about human mobility patterns, resource procurement strategies, and so on. It is, however, peculiarly geoarchaeological in that it considers those raw material sources not just as spots on a map, but as locations that need to be understood in the context of all the other spots on the map: as part of a complete landscape. The sources are located at a certain distance from other spots that people used (such as the site where the tools were found), and the intervening spaces, as well as the sources themselves, have topographic characteristics (terrain that is difficult or easy to cross, for instance) and the potential to provide other resources, or hazards to be avoided. Developing geoarchaeological theory thus allows us to interpret past behaviour in a broader way, better grounded in the environment in which people lived. Much of that theory is still implicit in our work, and not explicitly formulated and stated. It is hoped that publications such as this book, and conferences such as the DIG series, will stimulate a more deliberate and conscious effort to cultivate such theory, allowing us to move more securely into that middle ground of nature – behaviour interactions.
Human interactions with the geosphere Because it derives from a discipline devoted to deciphering the interactions of cultures and nature over the entire span of human existence, the geoarchaeological perspective clearly can be of great help in
providing context and even guidance in society’s present situation of rapid climate and environmental change. There are lessons to be learned. The record of the past shows that sometimes we caused problems, sometimes we suffered from changes beyond our control, and sometimes we found ways of living successfully for many generations in ways that did no harm and (from the human point of view) may even have improved the situation. It might help us now to have some idea of what we did right and what we did wrong in the past, although we are in most cases very unlikely to adopt the actual techniques used. However, we can also learn a salutary lesson from the ways in which past cultures reacted to the mere fact of change: those that resisted, suffered. Those that found new ways of coping generally did better. None came through unchanged. One of the first major ways that humans started affecting the natural environment was through the use of fire. There is some debate about when the earliest deliberate use of fire was, but it is clear that hominids controlled fire at least several hundred thousand years ago (Goudie 2006). Fire provided humans with tremendous advantages: cooking fires provided heat and light as well as more digestible food; vegetation could be burned to open up the landscape, making hunting easier, and to direct herds of animals to slaughter. Fire could be used for protection from predators, for eliminating vermin and noxious insects, for improving the quality of some lithic raw materials . . . the list goes on. In turn, however, regular use of fire resulted in major changes to the environment: the more open landscape now contained a different suite of vegetation, including in many cases pyrophytic species that tolerate fire, or even require it to permit germination. The extent of forest was changed, as was the distribution of animal species. It is fair to say that one of the fundamental factors creating the state of the world today, and even in prehistory, has been the use of fire by humans (see, e.g. Raab et al.). This has been demonstrated, for instance, by considerable work on Aboriginal fire use in Australia, and by pre-contact cultures in New Zealand, North America, and other parts of the world. Goudie (2006) and Williams (2006) have provided interesting summaries of such work. The subject of fire leads us straight to the next major way that humans have influenced the environment: agriculture. The use of fire not only changes the vegetation in an area, but enriches the soil temporarily, which is of course the basic reason for the widespread adoption of slash-and-burn agriculture (e.g. Trapeznikova). Many other agricultural techniques have also been developed, and a great deal of geoarchaeological work concerns agricultural landscapes. These studies can reveal additions to soils
INTRODUCTION
(e.g. Bintliff et al. 2006; Guttmann et al. 2006, 2008; Simpson et al. 2006), or other techniques such as terracing (Krahtopoulou & Frederick 2008) or runoff irrigation (Sandor et al. 2007), which create or improve arable land, or otherwise allowed communities to manage risk (Zaro & Alvarez 2005). Human land-use practices can have unintended consequences, however, which can be either positive or negative. The negative consequences are unfortunately relatively common (e.g. Wiseman 2007), and often take the form of predisposing a landscape to subsequent erosion, such as in the situations described by Bintliff et al. (2006) and Ayala & French (2005). This latter study described a case in Sicily where land-use practices (pastoralism v. agriculture) caused some areas to pass a stability threshold, resulting not only in erosion but also in a landscape that was resistant to regeneration after abandonment. The problems that we cause can thus have long-term consequences. On the other hand, work by McCoy & Hartshorn (2007) in Hawaii demonstrated that deterioration of some areas as a result of land use serendipitously resulted in wind erosion-derived enrichment of soils in nearby areas, allowing later reuse of the land. The interactions of a dynamic landscape and human cultures through time are never simple and straightforward: they are complex, ever-changing, with people ‘winning’ in some respects while ‘losing’ in others, both at any one time and overall (e.g. Kraft et al.). Morozova’s (2005) work on the Tigris and Euphrates rivers provides a good example. Periodic avulsions of the river channels would have had negative effects on settlements along those channels, but Mesopotamian farmers were also able to use the floodplain and channel morphology to their own advantage, for gravity irrigation and for access to better drained soils that were less susceptible to flooding and salinization. For this and other reasons, the periodic changes in channel morphology in fact created a beneficial situation for people, as well as a risky one. Furthermore, people were capable of influencing the avulsions, and engaged in activities that could ‘significantly alter, delay, or accelerate natural avulsion processes’ (Morozova 2005, p. 417). These activities included enlarging levee breaks, diverting channels for irrigation, diverting or blocking channels to control water flow to downstream city-states, building walls and levees, and so on. Morozova (2005, p. 418) concluded that ‘Because of the enormous scale of human activity in the Tigris – Euphrates plains over the last 7000 years, it is impossible to ascribe avulsions to natural causes alone . . . [H]igh-risk, low-yield irrigation agriculture, unpredictable river flow, frequent population migrations due to changes in river courses, and constant wars
5
and tensions between city-states over water supply contributed to instability, population flux, and pessimism of early lower Mesopotamian civilization, as well as to some of its greatest achievements.’ From this perspective, then, we humans are not only subject to natural environmental changes, but are also agents of change in our own right, and our actions often have unintended consequences. What may seem for a time or in one area to be beneficial may be revealed as a problem in the longer term or the larger scale, whereas what seems to be a problem in one area may well have beneficial consequences somewhere else. It all depends on your point of view. The work of Goman et al. (2005) provides an excellent example of this. They documented changes along the Rı´o Verde and the Oaxaca coast in Mexico over several centuries during the Formative period (before c. 2000 BP). The initiation of agriculture in the highland areas resulted in erosion, which resulted in increased runoff and sediment load in the river. This in turn affected channel morphology, changing the river from a meandering to a braided system, with flashier flow and more floods. This may sound like a negative scenario, but it had at least two arguably positive effects. First of all, the finer sediments were deposited on an expanding floodplain in the lower reaches of the river, thus greatly increasing the area suitable for agriculture. Second, the excess sediments that arrived at the coast were carried by longshore currents and caused the build-up of bay barriers. Behind those barriers, what had previously been a high-energy, wave-dominated coast became a protected estuarine environment, and estuaries are among the most productive marine habitats known. ‘Archaeological and ethnohistoric data show that pre-Hispanic populations exploited the estuaries for key resources such as fish, shellfish, waterfowl, salt, and ornamental shell . . . Given the high-energy wave environment along the Oaxacan Coast, it is possible that exploitation of marine resources would have been difficult until the estuaries formed’ (Goman et al. 2005, p. 256). These environmental changes were followed by a population expansion in the lower Rı´o Verde valley and along the coast (Goman et al. 2005, p. 257). Human agency can thus have both deleterious and beneficial effects, over time and across space. Geoarchaeological studies can help show us the larger picture, rather than restricting our view to one small slice of time, one small area, or for that matter one particular group of people. Climate and environmental change can certainly result in societal upheaval, however. Rapp & Jing show that river floods and avulsions resulted in the displacement of entire cities during the Shang period in China. Brooks (2006) discussed the fact that although it has traditionally been thought that
6
L. WILSON
social complexity and urbanization developed during benign times, there is considerable evidence to suggest that such development was often associated with environmental deterioration, especially times of increased aridity. Brooks made two major points about this. First of all, urbanization is associated with increased social inequality and violence, so it perhaps should be seen as a ‘suboptimal’ adaptation (p. 46). Second, developing urbanization was not the only way of reacting to increasing aridity: increasing mobility, for instance through pastoralism, also worked. Pastoralism in fact seems to have been more resilient and flexible than the fragile early urban civilizations, as it is still practised today. As Brooks (2006, p. 45) noted, ‘The resilience of responses based on mobility holds important lessons for the emerging fields of adaptation research and policy, which are developing in response to concerns about current and future anthropogenic climate change . . . Any measures to enhance resilience and promote adaptation must take account of local contexts and recognize that imported developmental models may be inappropriate; indigenous livelihood strategies have often emerged from centuries or even millennia of linked environmental and social change.’ Sandor et al. (2007) also discussed the implications of their geoarchaeological study for our current situation. They looked at runoff agriculture, which has for millennia allowed crops to be grown successfully in arid lands, as it not only increases water supply to the crops but also results in enriched soils owing to the influx of nutrient-rich organic matter and sediments. The fields that they studied in the Zuni area of the USA have been farmed for at least 1000 years, ‘making them among the oldest identified fields in the United States. The age and continuity of these sites make it possible to greatly extend the time perspective on the impacts of agriculture on soil resources’ (p. 360). This is the time perspective, they point out, that is ‘envisioned in the concept of agricultural sustainability’ (p. 361). The traditional Zuni method can be seen as much more beneficial to the land and the society using it than the more ‘modern’ methods that have been imposed by the government over the last century. On the other hand, as Morrison (2007) explained using examples from India, modern degradation of the environment is not always in contrast to past stability. Earlier populations were also capable of making major, detrimental, changes to soils, vegetation, and watercourses. The need for a long time-perspective was repeated by van der Leeuw et al. (2005), who decried the fact that ‘most of the research on environmental degradation is concerned with very short time-spans, of up to a century at most. As a
result, the approach is inadequate in a number of different ways’ (p. 11). Because degradation is often long term, involving multiple processes with multiple time scales, and intersecting with each other both spatially and in terms of dynamics, ‘We need to place the study of degradation in the context of other phases of landscape formation, such as soil formation, expansion of the vegetation, and human behavior that does not result in degradation. Doing that on the basis of data concerning the present alone is impossible’ (p. 11).
Dealing with current and future change The thesis of this present book is that geoarchaeology, by providing us with a longer time perspective on human–nature interactions, can give us a better understanding of how the changes that we currently face will affect us. It can also show us that some adaptations have been less deleterious, even more beneficial, than others over that longer time span. Can it help us plan for the future? There will never be a simple answer to that. The papers in this book show that the possibilities offered by nature, such as the distribution of amenable soils or landforms in climatically suitable areas (e.g. Trapeznikova), constrain the possible societal responses. They also show that human activities (e.g. Liu et al.) can have an impact on seemingly unrelated aspects of the environment. We are involved in a complex and highly dynamic system. It will not be easy, and may not even be possible, to predict what changes are coming nor how best to respond to them. As Brooks (2006 p. 46) stated, ‘The archaeological record emphasizes that adaptation has in the past been associated with great social upheavals that could not have been foreseen by those who were undertaking the adaptation. The consequences of adaptation were unplanned and unpredictable, arising from the ad hoc responses of a variety of actors to environmental change.’ Brooks went on to state that we would be naive to think that we can neutralize the effects of abrupt climate change through planned adaptation strategies. A similar viewpoint was expressed by van der Leeuw et al. (2005, pp. 25 –26), who pointed out that ‘many so-called “environmental crises” are in actual fact just as much societal as environmental . . . such problems are in effect instances of the inadequacy of the mechanisms by which societies deal with the dynamics of their surroundings . . . As a society interacts with its environment according to its world-view, it will rely upon, and impact on, certain aspects of that environment. The society’s impact on its environment is, therefore, in those domains where the society is the most
INTRODUCTION
dependent upon that environment. That is where the society is the most vulnerable, and where it finds adapting to changes occurring in its environment most difficult.’ For these researchers, any attempt to maintain stability is ultimately doomed to failure. Over the last couple of centuries, our society has achieved a semblance of stability by becoming highly dependent on technology, but ‘the problems facing us cannot be dealt [with] in the time-frame available to us’ (p. 26). This pessimistic view may well be correct, but it seems to me that it is better to try to solve our problems, even if the job sometimes seems hopeless. (In fact, van der Leeuw & Redman (2002) apparently agreed, as they called on archaeologists to actively engage in addressing problems relevant to contemporary society.) In our current situation we are facing potentially great changes, and societal upheaval is arguably already under way. However, this is in fact nothing new: societies have always had to change. The more we know of past environmental and societal changes, the more context we have, and the better our chances for dealing with present and future changes. Geoarchaeology can play a leading role in the present and the future by showing us a more accurate picture of the past, by emphasizing the inescapable interweaving of society and nature, and by refuting the dual myths of constant stability and constant progress. Twenty years ago Bruce Trigger (1989, pp. 410 – 411) described the contribution that archaeology can make to social change. His words are still pertinent today, and, I would say, even more so if you substitute ‘geoarchaeology’ for ‘archaeology’, and ‘environmental’ for ‘social’: ‘The fact that archaeology can provide a growing number of insights into what has happened in the past suggests that it may constitute an increasingly effective basis for understanding social change. That in turn indicates that in due course it may also serve as a guide for future development, not in the sense of providing technocratic knowledge to social planners but by helping people to make more informed choices with respect to public policy. In a world that has become too dangerous for humanity to rely on trial and error, archaeologically derived knowledge may even be important for human survival. If archaeology is to serve that purpose, archaeologists must strive against heavy odds to see the past as it was, not as they wish it to have been.’
References Adelsberger, K. A. & Kidder, T. R. 2007. Climate change, landscape evolution and human settlement in the lower Mississippi Valley, 5500–2400 Cal BP. In: Wilson, L., Dickinson, P. & Jeandron, J.
7
(eds) Reconstructing Human–Landscape Interactions. Cambridge Scholars, Newcastle, 84– 108. Ayala, G. & French, C. 2005. Erosion modeling of past land-use practices in the Fiume di Sotto di Troina River Valley, North–Central Sicily. Geoarchaeology: An International Journal, 20, 149– 167. Bintliff, J., Farinetti, E., Sarri, K. & Sebastiani, R. 2006. Landscape and early farming settlement dynamics in central Greece. Geoarchaeology: An International Journal, 21, 665– 674. Black, D. W. & Wilson, L. A. 1999. The Washademoak Lake chert source, Queen’s County, New Brunswick, Canada. Archaeology of Eastern North America, 27, 81–108. Boroffka, N., Oberha¨nsli, H. et al. 2006. Archaeology and climate: settlement and lake-level changes at the Aral Sea. Geoarchaeology: An International Journal, 21, 721 –734. Brooks, N. 2006. Cultural responses to aridity in the Middle Holocene and increased social complexity. Quaternary International, 151, 29– 49. Butzer, K. W. 1982. Archaeology as Human Ecology. Cambridge University Press, Cambridge. Carson, M. T. 2008. Correlation of environmental and cultural chronology in New Caledonia. Geoarchaeology: An International Journal, 23, 695 –714. Clarke, D. L. 1968. Analytical Archaeology. Methuen, London. Evans, C. P., Aitken, A. E. & Walker, E. G. 2007. A GIS approach for archaeological site distribution analysis by physiographic elements in the Lake Diefenbaker Region, Saskatchewan, Canada. In: Wilson, L., Dickinson, P. & Jeandron, J. (eds) Reconstructing Human– Landscape Interactions. Cambridge Scholars, Newcastle, 68– 83. French, C. 2003. Geoarchaeology in Action: Studies in Soil Micromorphology and Landscape Evolution. Routledge, London. French, C. 2007. Sustaining past and modern landscape systems in semi-arid lands. In: Wilson, L., Dickinson, P. & Jeandron, J. (eds) Reconstructing Human– Landscape Interactions. Cambridge Scholars, Newcastle, 252– 272. Gillmore, G. K., Coningham, R. A. E., Young, R., Fazeli, H., Rushworth, G., Donahue, R. & Batt, C. M. 2007. Holocene alluvial sediments of the Tehran Plain: sedimentation and archaeological site visibility. In: Wilson, L., Dickinson, P. & Jeandron, J. (eds) Reconstructing Human–Landscape Interactions. Cambridge Scholars, Newcastle, 37– 67. Goman, M., Joyce, A. & Mueller, R. 2005. Stratigraphic evidence for anthropogenically induced coastal environmental change from Oaxaca, Mexico. Quaternary Research, 63, 250–260. Goudie, A. 2006. The Human Impact on the Natural Environment, 6th edn. Blackwell, Oxford. Guttmann, E. B., Simpson, I. A., Davidson, D. A. & Dockrill, S. J. 2006. The management of arable land from prehistory to the present: case studies from the Northern Isles of Scotland. Geoarchaeology: An International Journal, 21, 61–92. Guttmann, E. B., Simpson, I. A., Nielsen, N. & Dockrill, S. J. 2008. Anthrosols in Iron Age Shetland: implications for arable and economic activity.
8
L. WILSON
Geoarchaeology: An International Journal, 23, 799– 823. Head, L. 2008. Is the concept of human impacts past its use-by date? Holocene, 18, 373– 377. Hill, C. L. 2007a. Surficial processes and Pleistocene archaeology: context, landscape evolution and climate change. In: Wilson, L., Dickinson, P. & Jeandron, J. (eds) Reconstructing Human–Landscape Interactions. Cambridge Scholars, Newcastle, 6– 36. Hill, C. L. 2007b. Geoarchaeology and late glacial landscapes in the western Lake Superior region, central North America. Geoarchaeology: An International Journal, 22, 15– 47. Howard, A., Passmore, D. & Macklin, M. (eds) 2003. The Alluvial Archaeology of North-West Europe and the Mediterranean. Balkema, Rotterdam. Huckleberry, G. & Duff, A. I. 2008. Alluvial cycles, climate, and puebloan settlement shifts near Zuni Salt Lake, New Mexico, USA. Geoarchaeology: An International Journal, 23, 107– 130. Huisman, D. J., Jongmans, A. G. & Raemaekers, D. C. M. 2009. Investigating Neolithic land use in Swifterbant (NL) using micromorphological techniques. Catena, 78, 185– 197. Krahtopoulou, A. & Frederick, C. 2008. The stratigraphic implications of long-term terrace agriculture in dynamic landscapes: polycyclic terracing from Kythera Island, Greece. Geoarchaeology: An International Journal, 23, 550– 585. Leach, P. A. & Belknap, D. F. 2007. Marine geophysics and vibracoring applied to refining the search for submerged prehistory in the Damariscotta River, Maine, USA. In: Wilson, L., Dickinson, P. & Jeandron, J. (eds) Reconstructing Human–Landscape Interactions. Cambridge Scholars, Newcastle, 141–158. Lorusso, P. 2007. Settlement and territory in the Early and Middle Apulo-Lucan Neolithic (southeast Italy): A geoarchaeological approach. In: Wilson, L., Dickinson, P. & Jeandron, J. (eds) Reconstructing Human–Landscape Interactions. Cambridge Scholars, Newcastle, 159–176. Luedtke, B. E. 1992. An Archaeologist’s Guide to Chert and Flint. Archaeological Research Tools, 7. Institute of Archaeology, University of California, Los Angeles. McCoy, M. D. & Hartshorn, A. S. 2007. Wind erosion and intensive prehistoric agriculture: a case study from the kalaupapa field system, Moloka’i Island, Hawai’i. Geoarchaeology: An International Journal, 22, 511–532. Moody, J., Robinson, H., Francis, J., Nixon, L. & Wilson, L. 2003. Ceramic fabric analysis and survey archaeology: the Sphakia survey. Annual of the British School at Athens, 98, 37– 105. Moore, K. R. 2006. Prehistoric gold markers and environmental change: a two-age system for standing stones in Western Ireland. Geoarchaeology: An International Journal, 21, 155 –170. Morozova, G. S. 2005. A review of Holocene avulsions of the Tigris and Euphrates rivers and possible effects on the evolution of civilizations in lower Mesopotamia. Geoarchaeology: An International Journal, 20, 401–423.
Morrison, K. 2007. Making places and making states: agriculture, metallurgy, and the wealth of nature in south India. In: Boomgaard, P. & Bankoff, G. (eds) The Wealth of Nature: How Natural Resources Have Shaped Asian History, 1600– 2000. Palgrave MacMillan, Basingstoke, 81–99. Neme, G. & Gil, A. 2009. Human occupation and increasing mid-Holocene aridity. Current Anthropology, 50, 149–163. Oetelaar, G. A. 2004. Landscape evolution and human occupation during the Archaic period on the Northern Plains. Canadian Journal of Earth Sciences, 41, 725–740. Pelletier, B. G. Jr., Hall, B. & Robinson, B. R. 2007. Evaluating Palaeoindian settlement in relation to Pleistocene lake levels at Munsungan Lake, Northern Maine, USA. In: Wilson, L., Dickinson, P. & Jeandron, J. (eds) Reconstructing Human–Landscape Interactions. Cambridge Scholars, Newcastle, 125–140. Rapp, G. 2007. Prologue: The organisation, development, and future of Geoarchaeology. In: Wilson, L., Dickinson, P. & Jeandron, J. (eds) Reconstructing Human– Landscape Interactions. Cambridge Scholars, Newcastle, 1– 5. Rapp, G. & Hill, C. L. 2006. Geoarchaeology: The EarthScience Approach to Archaeological Interpretation, 2nd edn. Yale University Press, New Haven, CT. Rapp, G., Allert, J., Vitali, V., Jing, Z. & Henrickson, E. 2000. Determining Geologic Sources of Artifact Copper: Source Characterization Using Trace Element Patterns. University Press of America, Lanham, MD. Rech, J. A., Quintero, L. A., Wilke, P. J. & Winer, E. R. 2007. The lower Paleolithic landscape of ’Ayoun Qedim, al-Jafr Basin, Jordan. Geoarchaeology: An International Journal, 22, 261–275. Rømer, S., Breuning-Madsen, H., Balstrøm, T. & Jensen, A.-E. 2006. Mapping Quaternary deposits as a method for explaining the distribution of Mesolithic sites in reclaimed landscapes: an example from Va˚lse Vig, Southeast Denmark. Geoarchaeology: An International Journal, 21, 113–124. Sageidet, B. M. 2009. Late Holocene land use at Orstad, Jæren, Southwestern Norway, evidence from pollen analysis and soil micromorphology. Catena, 78, 198–217. Sandor, J. A., Norton, J. B. et al. 2007. Biogeochemical studies of a native American runoff agroecosystem. Geoarchaeology: An International Journal, 22, 359–386. Simpson, I. A., Guttmann, E. B., Cluett, J. & Shepherd, A. 2006. Characterizing anthropic sediments in north European Neolithic settlements: an assessment from Skara Brae, Orkney. Geoarchaeology: An International Journal, 21, 221–235. Trigger, B. G. 1989. A History of Archaeological Thought. Cambridge University Press, Cambridge. van der Leeuw, S. & Redman, C. L. 2002. Placing archaeology at the center of socio-natural studies. American Antiquity, 67, 597– 605. van der Leeuw, S. E. & THE ARCHAEOMEDES RESEARCH TEAM. 2005. Climate, hydrology, land use, and environmental degradation in the lower
INTRODUCTION Rhoˆne Valley during the Roman period. Comptes Rendus Ge´oscience, 337, 9– 27. Williams, M. 2006. Deforesting the Earth, From Prehistory to Global Crisis: An Abridgment. University of Chicago Press, Chicago, IL. Wilson, L. 2007a. Terrain difficulty as a factor in raw material procurement in the middle Palaeolithic of France. Journal of Field Archaeology, 32, 315–324. Wilson, L. 2007b. Understanding prehistoric lithic raw material selection: application of a gravity model. Journal of Archaeological Method and Theory, 14, 388–411.
9
Wiseman, J. 2007. Environmental deterioration and human agency in ancient Macedonia: a case study. Geoarchaeology: An International Journal, 22, 85–110. Zaro, G. & Alvarez, A. U. 2005. Late Chiribaya agriculture and risk management along the arid Andean coast of Southern Peru´, A.D. 1200– 1400. Geoarchaeology: An International Journal, 20, 717– 737. Zhang, Q., Zhu, C., Liu, C. L. & Jiang, T. 2005. Environmental change and its impacts on human settlement in the Yangtze Delta, P.R. China. Catena, 60, 267– 277.
Environmental limitations on agricultural development of the forest zone of the East European Plain (Russian Federation) OLGA N. TRAPEZNIKOVA RAS Institute of Environmental Geoscience, 13, Ulanskii pereulok, Moscow, Russia, 101000 (e-mail:
[email protected]) Abstract: Intensive agricultural development of the forest zone of the East European Plain started in the second part of the first millennium AD. Although the majority of the mediaeval population were peasants, archaeological study of ancient rural settlements is much less developed than that of ancient towns. The analysis of interrelationships between environmental conditions and the agricultural pattern across space, including the corresponding pattern of rural settlements, helps us to delimit the spatial frame in which it is possible to find rural settlements of different historical epochs, even if they have since disappeared. Five areas with different historical types of agricultural landscapes were revealed, based on their geological and climatic characteristics. Another analysis essential for archaeology deals with the age of contemporary agricultural landscapes and rural settlements along with the factors and laws that control their changes through time and space.
The forest zone occupies a large area. The whole northern part of the East European Plain (EEP) is covered by forest, except for the most northerly permafrost strip along the Arctic Ocean. The contemporary forest zone consists of a number of latitudinal vegetation belts, such as northern, middle and southern taiga (coniferous forests), mixed forest and broad-leaved forest. The forest zone was even wider in the middle of the Holocene, when it extended far to the south and possibly linked up with the Caucasus forests along river valleys. It is now thought that during the 3500 years of the Bronze Age the southern boundary of the forest zone retreated 500 km to the north under both anthropogenic and climatic impacts (Markova et al. 1995; Smirnova et al. 1995). The single continuous Eurasian steppe belt appeared and Asian martial nomadic tribes had free access to southern European settled grain growers. The nomadic invasion caused the ‘great migration of peoples’, which took place in the middle of the first millennium AD. As a result the grain-growing tribes left their lands and some of them, namely Slavs in the centre and west of the EEP and Finno-Ugric tribes in the east, migrated to the north under the protection of impenetrable forests. Although settled Finno-Ugric and Slavic tribes suffered greatly from nomad incursions, only in the 8th–9th centuries did climatic warming allow them to reach their northern limits, which have been preserved as the boundary of agricultural development until today (Table 1). In fact, the forest zone was already populated at the time of the migrations. Different Finno-Ugric tribes lived in the central and eastern parts of the forest zone (the Ananianskaya and Dyakovskaya
archaeological cultures), and Balts occupied the western part of the plain, on the southeastern coast of the Baltic Sea. They pursued different variants of non-plough agriculture, but their main occupations were hunting and fishing. Thus intensive agricultural development of the contemporary forest zone of the EEP started in the second part of the first millennium AD, and this was the most agriculturally developed area of the EEP until the end of the 18th century. This may seem rather strange, because the natural conditions (climate and soil) of the northern wooded part of the EEP are less favourable for agriculture than those of the middle steppe part of the EEP. The area that is naturally the most favourable for agriculture occupies the central part of the plain within the forest –steppe and steppe zones, with the famous chernozem, one of the most fertile soils in the world, and a warm climate. According to archaeological data, plough agriculture appeared first in the southern part of the EEP, at least as early as the first part of the first millennium AD (Sedov 2005). Less than 1000 years later tribes who had mastered the plough technique in areas naturally favourable to agriculture started to cultivate the more difficult forest zone where only the southern taiga, mixed and broad-leaved forest belts are suitable for agricultural development. Although the majority of the mediaeval population were peasants, archaeological study of ancient rural settlements is much less developed than that of ancient towns because of the scarcity of traces left when such villages disappear: rural settlements were usually very small and existed for shorter periods than towns. It is therefore very
From: Wilson, L. (ed.) Human Interactions with the Geosphere: The Geoarchaeological Perspective. Geological Society, London, Special Publications, 352, 11– 26. DOI: 10.1144/SP352.2 0305-8719/11/$15.00 # The Geological Society of London 2011.
12
Table 1. Initial agricultural development of the East European Plain (EEP) Period Southern EEP (contemporary steppe and forest–steppe)
Northern EEP (contemporary forest zone)
Processes
Bronze and Iron Ages (6000 years ago to 1500 years ago)
Cattle-breeding and agriculture: from hoe–mattock to tillage
Forest contraction to the north: climate desiccation, soil disturbance
Rare husbandry sites: cattle-breeding and slash-and-burn agriculture
1500 years ago to 1000 years ago
The great migration of peoples: end to tillage
Formation of fertile chernozem soils under steppe vegetation
1000 years ago to 300 years ago: establishment of centralized state
Nomadic cattle-breeding
A number of well-developed agricultural landscapes based on tillage and organic fertilizing of soils Russian peasant colonization from west and south to east and north
Processes
Years
Dt (8C)
Gradual degradation of biodiversity, increase in open areas, formation of forest belts Deforestation of the most suitable landscapes: intensive fluvial erosion
600 BC AD 200
22.0 þ2.0
AD 500 AD 1000
22.5 þ1.0
Ecological balance of majority of agricultural landscapes
AD 1250 AD 1500 AD 1700
21.0 20.2 22.0
O. N. TRAPEZNIKOVA
Characteristic of agricultural development
Characteristic of agricultural development
Summer average temperature oscillations in comparison with 2000 (Shpolyanskaya 2003)
ENVIRONMENTAL LIMITATIONS ON AGRICULTURE
difficult to find traces of abandoned rural settlements within the forest zone of the EEP, but the historical scene is incomplete without them. To help us find them, we can restrict the search area by reconstructing to some extent the corresponding agricultural landscapes and the settlement patterns connected with them. The analysis of the interrelationships between environmental conditions and the agricultural pattern across space, including the corresponding pattern of rural settlements, helps us to delimit the spatial frame in which we should be able to find rural settlements of different historical epochs, even if they have since disappeared. Another analysis essential for archaeology deals with the ages of present-day agricultural ecosystems and rural settlements, along with the factors and laws that control their changes through time and space.
Objectives The goal of our research is to find areas of old agricultural landscapes, their pattern across space and corresponding rural settlement patterns, based on the interrelationships of the natural environment, social situations and agrarian technologies, using the forest zone of the EEP as an example. To accomplish this task we examine the following questions. (1) What are the historical agricultural landscapes and when did they appear? (2) What are the relationships between the historical types of agricultural landscapes and the environment, including nature, society and agrarian technologies? (3) What is the relationship between the historical types of spatial patterns of agricultural landscapes and the parameters of corresponding rural settlements, including location and size? (4) How are age and previous pattern reflected in contemporary agricultural landscapes, and is it possible to use the contemporary spatial pattern of agricultural lands for historical research?
Methods This research is based on a complex of heterogeneous data concerning different historical periods. The research combines three levels of generalization. The first one deals with the whole forest zone of the EEP and general laws of agricultural development within it. On the second level historical types of agricultural development are revealed and analysed by their location, age, and general spatial and settlement patterns. The third, most detailed, level is used to study the internal structure of certain agricultural landscapes, relationships
13
between their spatial and settlement patterns, and the historical development of their physiographic and social environment. Such an approach requires specific methods integrating conventional and new techniques. The methods used in the research include computer processing and landscape interpretation of remote sensing data, statistical analysis, diachronous and synchronous historical geographical analysis, and a geographical information system (GIS), allowing remote sensing data and maps of different scales and projections to be put into conformity and link statistical data concerning administrative units (which changed from period to period) with agricultural landscapes and the natural environment. The GIS allows all three levels of the research to be combined. The initial materials of the research include remote sensing data of medium and high resolution for different dates (since the second half of the 20th century); topographical and land-use maps from General Land Surveying maps of the 18th century until the end of the 20th century; the results of archaeological and historical geographical research concerning agriculture (e.g. Anon. 1997, 2001; Sedov 2005); statistical data about agriculture and demography from the end of the 19th century to the present; historical sources, such as mediaeval land cadastres (known as Pistzovaya kniga); and environmental data, including climate, geological maps and others. The main approach used in this research is a comparative historical geographical analysis. A historical GIS was designed for the forest zone of the EEP, which united the available historical cartographic data with geological, climatic and other geographical maps including remote sensing data and modern topography. Analyses of remote sensing data and land-use maps, along with field trips, were aimed at finding specific types of agricultural landscapes. Historical geographical analysis detected their age and peculiarities of development. Our research (Trapeznikova 2007) showed that the strict natural limitations for agriculture together with the specific Soviet agricultural model, which mainly conserved land use of the previous epoch, made the contemporary distribution of arable lands generally correspond to that of at least the 17th –18th centuries, and the settlement patterns essentially changed during the 20th century. Since ‘perestroika’ at the end of the 20th century, agricultural land use in the forest zones has changed dramatically (Nefedova & Loffe 1998); nevertheless, we can use topographical maps, airborne and space images from before 2000. That is why we started the analysis of the agricultural pattern of the forest zone of EEP from the contemporary data (before 2000) and then turned to historical
14
O. N. TRAPEZNIKOVA
maps and other sources within the key agricultural landscapes.
Theoretical basis of the agricultural landscape analysis Many researchers, starting with Odum (1994), have examined the differences between natural landscapes and agricultural ones, including their spatial organization. The main difference between agricultural and natural landscapes results from the fact that agricultural landscapes are not selfgoverning systems like natural landscapes, but governed subsystems of a more complicated selfgoverning general social system. Therefore we cannot regard agricultural ecosystems as simple modifications of natural units. The main agents for governing agricultural landscapes are additional sources of energy and changes to material streams (Table 2). At the same time, at any level of agricultural techniques, natural conditions appear to be a limiting factor for agriculture. They limit the spatial distribution of agricultural ecosystems, providing for both a specific spatial agricultural landscape pattern and the substantive content of any agricultural landscape. These restrictions can be more or less strict, and can be very important when they act together with other factors. On the other hand, the increasing capability of human society weakens the effects of natural limitations, so that more types of landscapes can be involved in agricultural development with more intensive
technological loading. Another important social factor involves the environmental aspects of land use and land ownership. In simple terms, we should take into account two contrasting types of land use and land ownership (common and private), with all possible combinations of these at the same time. An agricultural landscape is a major material part of a traditional cultural landscape. Agricultural landscapes always connect with settlements. The settlements are the centres of corresponding agricultural landscape units. They are the kernel of the inner pattern of any agricultural landscape, although this pattern is determined by the natural environment and agricultural techniques. For example, hoe –mattock agriculture used by the Anan’inskii tribes in the Iron Age determined their settlement pattern so that the settlements were located on high river banks with wide flood terraces beneath (Anon. 1997). These terraces were the only suitable area for this type of agriculture. They were more or less woodless and always fertile because of repeated flooding. Subsequent peoples used slash-and-burn agriculture, which was preserved in some regions of Russia until the 19th century. This type of agriculture does not depend upon the distribution of woodless flood plains, which make up a very small part of the total area. Restoration of soil fertility is provided by giving up a plot after several years of use and returning to it after many years when the natural fertility would be restored. These peoples are also characterized by a special settlement pattern and specific type of agricultural landscape. Their settlements were large, not only
Table 2. Comparison of characteristics for organization and functioning of natural and agricultural landscapes Characteristics
Natural landscapes
Energy sources Role of natural conditions Role of man-made conditions Boundaries
Natural Organizing Disturbing Discrete and diffuse natural boundaries
Spatial organization
Natural combination of morphological subunits according to their genesis
Hierarchy
Determined by mass distribution and natural energetic streams
Sustainability
Based on the possibilities of self-organization Components of different age; the lithogenous component is the oldest and most sustainable one
Age
Agricultural landscapes Natural and anthropogenic Limiting Governing Man-made discrete boundaries; sometimes they coincide with discrete natural boundaries Elementary agricultural ecosystems (tillage, hayland, pasture and others) surrounding their governing centre: a settlement Determined by hierarchy within the settlement pattern and infrastructure Depends on effective governing Historical
ENVIRONMENTAL LIMITATIONS ON AGRICULTURE
for defensive purposes but also because slashand-burn agriculture required intense efforts by large numbers of people over short periods of time. At the same time, the absence of the plough put limitations upon which sites were suitable for cultivation. Only sandy and sandy – loam soil could be used. Such soil can be found in Quaternary fluvioglacial valleys, most of which were inherited by modern rivers. The next stage of agricultural development, plough agriculture, once again changed the agricultural landscape and the corresponding settlement pattern. Permanent arable land with more or less regular crop rotation appeared. As a result the area of arable land necessary for one person decreased, and the anthropogenic impact changed and increased. Using draught horses and ploughs allowed people to cease tillage of sandy soil, which is very poor. Obligatory conditions for persistent tillage in the forest zone include more or less regular crop rotation (usually the three-fields system) and fertilizing (manuring) of the soil. This change in agricultural technique resulted in a change in the settlement pattern. Minimizing distance to the settlement appeared to be a very important factor for decreasing labour expenditure, as is shown in Table 3. Small settlements consisting of 1–3 farms appeared, starting at the end of the first millennium AD, instead of the large villages of the Iron Age. Archaeologists (Bader 1951; Sedov 2005) have recorded the dispersal of large patriarchal families (clans) into a great number of small settlements (of 1–3 houses) at the beginning of the second millennium AD. Thus we see how the agricultural technique influenced settlement preferences. As society developed and the state appeared, social and economic factors also became very important. Hence our assessment of settlement patterns for different epochs should be based on the analysis of production management and techniques, social and economic conditions, and the natural environment. It should also be taken into account that temperature changes in the last 2000–3000 years are essentially less than temperature variations within the colonized area of the forest zone (Tables 1 & 4). Hence the importance of temporal climate variations is relatively low in comparison with spatial ones.
15
Spatio-temporal agricultural landscape analysis of the forest zone of the EEP Initially the whole forest zone of Europe was characterized by an amazing uniformity of agriculture and monotony of economic systems. As was stressed in the Agrarian History of Northwestern Russia (Anon. 1978, p. 372), the state of agriculture in northwestern and northeastern Russia in the 15th century did not differ essentially from that in other countries situated in the non-chernozem zone of Eastern and Central Europe: there was the same prevalence of the three-field system, a similar set of crops, similar agricultural tools, and, at least, the same instability and discontinuity of yields and their similar values. According to Slicher van Bath (1963), the average production of the main cereals was at a ratio of about one measure of seeds planted to 4.3 measures of grain yielded in France at the end of the 15th century, and in Germany, the Scandinavian countries, Poland, Latvia, Lithuania and Estonia this ratio was 1:4.1 and 1:4.2 even in the 16th century. The same was true in the forest zone of Russia. At first sight such uniformity of agriculture within a large area of northern Europe clashes with the diversity of the natural conditions, including climate (Table 4). Differences can, however, be found in the peculiarities of selective spatial agricultural land use that resulted in different historical types of agricultural landscapes in different regions, which stand in close relationship to settlement spatial patterns. Hence, determining and analysing agricultural spatial landscape patterns is the first step in detecting extinct rural settlements of the forest zone of the EEP. It is common to perceive the natural environment as stable, compared with dynamically changing anthropogenic factors. However, landscapes do change through time, and in the EEP landscape age is a strong determinant of potential production capacity. The majority of the forest zone landscapes are very young; they appeared in the Quaternary once the area became ice-free after the glacial period, so the youngest southwestern landscapes are not older than 10 000 years. Because there were a number of glaciations, which covered the areas within the forest zone to greater or lesser extents, they created a heterogeneous cover of
Table 3. Working days necessary for soil manuring depending on distance from a field (Anon. 1898) Average distance from settlement to field (km) Number of days necessary for manure transportation per 1 ha
0
1
2
3
4
5
5.9
9.9
13.9
17.9
22.0
26.0
16
Table 4. Contemporary climate parameters in the forest zone of the East European Plain Zone
Central temperate zone of older landscapes Northeastern cold zone of older landscapes (Perm district)
Northwestern cold zone of young landscapes
Latitude (N)
Longitude (E)
Average temperature (8C) January
July
Per year
Sum of effective temperatures (.10 8C)
Precipitation per year (mm)
Snow thickness (cm)
Frost-free period (days)
Ryazan’
548000
408000
28.0
19.0
4.2
2200
500 – 575
30
140
Novgorod
588300
318150
28.0
18.0
4.0
1950
500 – 550
30
140
Valday Moscow
588000 558450
338150 378400
210.5 210.0
16.0 18.0
2.5 3.0
1550 2000
750 – 800 550 – 600
45 30
125 130
Kudymkar, a centre of Invenskoe porech’e Vereschagino, a centre of Obvinskoe porech’e Kargopol
598000
568400
215.7
17.4
1
1737
450 – 470
43
108
588000
548400
215.5
17.8
1.3
1819
440 – 460
40
120
618300
388550
214.0
16.0
1.5
1550
300 – 500
65
100
The named settlements are mostly old towns, which appear to have been centres of agricultural development since their origin. Their locations are shown in Figure 1.
O. N. TRAPEZNIKOVA
Southern warm zone (opol’e) Western temperate zone of young landscapes
Settlements
ENVIRONMENTAL LIMITATIONS ON AGRICULTURE
deposits forming a great number of geomorphological units. The age of landscapes increases from the NW (the youngest and most diverse) to the SE, where the forest zone extends beyond the formerly glaciated area and the landscapes are formed upon the Neogene –Palaeogene erosion surface. Thus, different sectors of the forest zone are characterized by different geological conditions, which greatly influence their agricultural potential. The east and SE of the region belong to the periglacial area, which was never covered by glaciers during the Pleistocene. The central part of the region is covered by deposits of the middle Pleistocene Moscow glacier, which corresponds to the Riss Ice Age in Alpine terminology, whereas the drift deposits in the western sector of the region formed under the last late Pleistocene Valday glacier, corresponding to the Wu¨rm. The area of the Valday glacier is characterized by the most contrasting mosaic of geological and geomorphological conditions, with young, slightly developed and usually boggy river valleys unsuitable for cultivation. At the same time, climate harshness increases from the SW to the NE of the forest zone (Table 4). The analysis shows that the climatic conditions are favourable for plant agriculture only in the southern part of the forest zone, whereas the northern and eastern parts are much less favourable and their usefulness for agriculture is usually due to local microclimates, which depend on geological and geomorphological peculiarities. The duration of the frost-free period is one of the most important parameters, and this is determined to a great extent by local topography and hydrology. Ground frosts are more likely in small hollows in hilly areas and less likely in river valleys and in the vicinity of large lakes. Thus microclimate is a natural limiting factor for agriculture in different natural units. In fact, it is not a single factor but a result of specific combinations of geological, topographical and other factors that control the distribution of arable land and hence of rural settlements. Thus we can describe the change in environmental conditions in terms of two main natural trends: the climatic trend of increasing harshness from SW to NE and the geological trend of increasing landscape age from NW to SE. These trends form five zones (Fig. 1), each of which is characterized by a set of historical types of agricultural landscapes, which are discussed in this paper. The boundaries of these zones are rather conventional and include transition belts. One can find two cold and two temperate zones (of correspondingly younger and older natural landscapes), and a southern warm zone. The boundary between the cold and temperate agricultural zones is tentatively taken as the boundary between the cold and temperate climatic zones, according to the ratio of the sums
17
for activeP(.10 8C) and negative P (,0 8C) temperatures: (j tactive . 108j _ j t , 08j) per year as suggested by Uglov (1987). The cold climatic zone is characterized by a ratio of less than unity, whereas the temperate climatic zone is characterized by a ratio of more than unity. The boundary between older and younger natural landscapes is taken as the boundary of the latest Valday (Wu¨rm) glacial deposits.
The southern warm zone The analysis shows that the southern part of the forest zone, where the climatic conditions are favourable for agriculture and do not limit it, nevertheless looks like a mosaic of regions, with tillage percentages varying from 15 to 80%. This mosaic mainly corresponds to areas locally known as opol’e and poles’e, shown in Figure 1. In Russian, ‘opol’e’ comes from a word meaning field; ‘poles’e’ comes from a word meaning wood. Opol’e areas are located in regions with high tillage (50 –80%), whereas low tillage is characteristic of poles’e areas. The terms opol’e and poles’e correspond to two types of landscape, which in the aggregate form an opol’e –poles’e structural morphological landscape belt. This belt extends from west to east through the central part of the EEP in the periglacial zone of the Moscow (Riss) glaciation, taking in the whole southern part of the forest zone. The opol’e– poles’e morphological landscape belt was revealed as a separate landscape unit by the Russian geographer Milkov (1964) and was later studied in detail by many researchers. These two types of landscape occur in the same climatic conditions but opol’e areas are characterized by very high tillage whereas poles’e areas have always preserved their forests. Poles’e areas tend to have elongate shapes because they follow landforms formed by glacial melt-water streams, and consist of well-washed fluvioglacial sand, which produces very poor soil. In addition, because of this origin such areas represent low topographic features and hence are affected by repeated wetting, making them boggy. On the other hand, the opol’e areas were high topographic units (islands) between fluvioglacial streams where loess-like sediment developed during the last glaciation. Their current form is a result of centuries of intensive agricultural development. It was formerly believed that these landscapes were always naturally woodless and were northern steppe relicts. However, it has recently been proven (Smirnova et al. 1995) that they lost their forests after agricultural development. Their high topographic level provides for good drainage and the most fertile soil within the forest zone develops on loess-like sediment. It should be stressed that the
18 O. N. TRAPEZNIKOVA
Fig. 1. EEP zones of agricultural development and the climatic and geological trends.
ENVIRONMENTAL LIMITATIONS ON AGRICULTURE
fertile soil of the opol’e is a result of their long periglacial development, in contrast to the younger Valday landscapes. Opol’e landscapes are the most productive and stable agricultural ecosystems within the forest zone, whereas under the same climatic conditions poles’e landscapes are hardly used for agriculture and are still covered with forest, which determines the southern boundary of the forest zone of the EEP. Opol’e landscapes are characterized by total tillage except in areas of steep slopes and ravines owing to high fluvial erosion caused by tillage. The 20th century pattern of settlements preserves that existing since their development, and consists of rather large villages situated near rivers or other water sources (ponds, wells).
Northeastern cold zone of older landscapes NE of the opol’e–poles’e belt climate harshness rapidly increases (Fig. 1, Table 4) and its influence dominates in the distribution of arable land and rural settlements. The influence of the temperature regime is both direct, on prevailing crops and agricultural patterns, and indirect, by imposing limitations on the natural units suitable for tillage. One of the most common northern agricultural landscape types is known locally as porech’e, which means an area near a river (Trapeznikova 2005). Although everywhere in the forest zone rivers were a natural and practically the only route for distribution of new population, in most areas settlers soon developed the interfluvial plains. However, the interfluvial plains in the northeastern cold zone of older landscapes were preferentially excluded from agricultural development (Fig. 2). The short vegetation period allows sustainable agriculture to be possible only in river valleys, where there is a warmer microclimate. Another important factor is that in the case of loam soils, areas within the valley slopes and terraces become suitable for tillage much sooner after snow melt in the spring than do the flat interfluvial plains. Loam soils are more fertile than sandy ones and give a better response to fertilizer, which is necessary in areas of constant tillage. Our detailed research into this type of agricultural landscape took place within the southern taiga belt at the east of the EEP, on the west bank of the river Kama, between latitudes 578300 and 598200 N (Fig. 2). This is a region of insecure agriculture, characterized by barren soil and a rather severe temperature regime (Table 4). In spite of this, it is a stable centuries-old agrarian area, with agricultural land currently covering from 21 to 53% of the region, whereas the forested area makes up from 35 to 67%. Interfluvial drainable plains, relict ridges and hills, and river valleys are the most common natural landscapes
19
here (Chazov 1961). Most of the region is divided between two basins of large tributaries of the Kama: the Inva basin to the north and the Obva basin to the south. In spite of the difference in geographical position, the two basins have similar geological and geomorphological conditions and every type of landscape can be found in each basin. On the other hand, their geographical positions (the Inva is located about 200 km north of the Obva) do result in major climatic differences. All agricultural landscapes have a similar internal spatial pattern, formed historically by exploitation of arable lands surrounding numerous small villages along river valleys (Figs 3 & 4). Each unit of the agricultural landscape had a settlement in the centre surrounded by a continuous cultivated area, which needed regular fertilization. This region was first colonized by Finno-Ugric tribes, who lived along the Kama tributaries such as the Inva and the Obva starting at the end of the first millennium, surrounded by impenetrable dense forest. Travel within the habitable region was possible only by river. That is why earthen and wooden fortresses were built at the river mouths, protecting the routes to the populated areas. This culture was named ‘Rodanovskaya’ by archaeologists, after the name of one of the fortresses, ‘Rodanovskoye Gorodishe’. The Rodanovskaya culture was characterized by the dominance of arable farming and the origin of the fallow system. If we compare the extent of the area of the Rodanovskaya culture with the current area occupied by their descendants, the Komi-permyaks, we see only insignificant changes, such as a contraction of the area in the south, west and east. However, the northern boundary remained stable and still coincides with the northern boundary of intensive agricultural development in the near Urals region. Thus, every porech’e developed starting in the 9th century as a growing agricultural landscape, and preserved the same spatial pattern until the beginning of the 20th century. The valleys of the Kama’s tributaries differ from the sandy valley of the lower and middle Kama owing to their hard loam soil, which could be cultivated only once new techniques using horses and ploughs were introduced. The plough technique does not need as many people as slash-and-burn agriculture does, and thus the settlement pattern changed: instead of widely distributed large settlements, people lived in numerous small ones (of 1–3 farms). The 16th and 17th centuries were a period of spontaneous peasant colonization of this region by Russians, after the ‘Troubles’ (1584–1613) and the development of a deep and unbridgeable cleavage between different aspects of Orthodoxy in Russia (after 1650). During that period the population of the central and western parts of Russia
20
O. N. TRAPEZNIKOVA
Fig. 2. Satellite image of the Inva and Obva basins. Black frame shows the area of Figure 3. Dark tones correspond to forests; light tones are arable land.
critically diminished whereas the population of the near Urals area increased by 2–3 times. The number of settlements also increased during this period. However, there was a difference between the Obva basin and the Inva basin because this colonization coincided with the so-called Little Ace Age of the 14th–19th centuries, a climatic minimum after the Middle Ages, when the average summer temperature was 2 8C less than now (Table 1). Newcomers therefore preferred the southern Obva basin as more suitable for agriculture. As a result, the Komi-permyaks remained in the Inva basin (nowadays it is the Komi-permyak Autonomous Okrug) whereas the Obva basin’s
Komis were totally assimilated by the 19th century. Thus the number of settlements and the agricultural landscape grew more rapidly in the Obva basin than in the Inva basin, as a result of this climatic factor, whereas the settlement pattern remained the same in both basins. It is important to stress that owing to the conversion of the Komi-permyaks to Christianity, the peaceful nature of the colonization, and the total uniformity of agriculture within the forest zone, Russian colonization did not change the agricultural and settlement patterns. Moreover, comparison of current maps and space images with old maps demonstrates the historical inheritance of the
ENVIRONMENTAL LIMITATIONS ON AGRICULTURE
21
Fig. 3. Settlement pattern of porech’e in the Obva basin: continuous-line circles show 1 km zone around present settlements; dashed circles show 1 km zone around extinct settlements of the 18th to the first half of the 20th century. Rivers are shown with black continuous lines. Dark tones of the background satellite image correspond to forests; light tones are arable land.
arable lands despite major changes to agricultural techniques in the 20th century (Fig. 3). In the 20th century large-scale collective farming based on agricultural machinery and engineering, with livestock specialization, replaced small subsistence farming. Both the rural population and the number of settlements were reduced, but the lack of natural grassland and its poor productivity together with a long stabling period in winter made people use tillage for growing fodder. This explains why the total arable area remained almost the same, in contrast to the settlement pattern. The historical and geological similarities together with their internal resemblance, and the difference in geographical position between the Obva porech’e and the Inva porech’e, make these two agricultural landscapes a useful source of proxy data for the indication of landscape fluctuations caused by climatic changes. The Obva porech’e is situated at the southern border of the south taiga zone, whereas the Inva porech’e is situated at the northern border of the south taiga zone (Table 4). These locations allow this pair of agricultural landscapes to represent both the local features of agricultural landscapes of this size and general
features because they mark differences between two climatically determined vegetation zones: middle taiga and southern taiga. Modern-day agricultural landscapes were studied in detail using both statistical data and space images. In particular, a spectral space image made on 10 May 1997 by scanner MSU-SK from satellite Resurs-01-3 was processed with the ENVI software package and arable lands were classified on the image using training classes. The map of arable lands resulting from ENVI processing was analysed with the MapInfo software package by using the index of squareness K, given by pffiffiffi pffiffiffi P 4 S ¼ 0:25P= S K¼ : S S where P is perimeter and S is area. Although each field has a more or less rectangular shape, together they form aggregations (tracts) with very indented boundaries (Figs 2 & 3). The K value indicates how greatly each tract is indented in comparison with a rectangle of the same area. If we compare these tracts by eye they all seem more or less equally indented. However, if we use the K
22
O. N. TRAPEZNIKOVA
Fig. 4. Arable land areas of the Inva basin (Inva porech’e) and the Obva basin (Obva porech’e) with index of squareness values (K).
value we see differences (Fig. 4), which can explain the impact mechanisms of the natural limitations. It was pointed out above that the internal spatial and settlement patterns of both basins are the same. Hence, the difference of K value demonstrates the difference in selection of arable lands under natural limitations. These limitations are the result of microclimate, including the frost-free period and the sum of effective temperatures, which are under the control of topography. The values of K obtained vary from 1.02 (very close to the minimal possible value of unity) to
13.84. The tracts were divided into four ranks according to K with a step equal to four (Fig. 4). The pattern analysis according to the K value shows different spatial agricultural landscape patterns in the Obva porech’e and the Inva porech’e. The Inva porech’e is characterized by large arable tracts with high K (8–10). In the centre of the porech’e with its main settlement, the town of Kudymkar, K is much lower (4 –6). In addition, large tracts of arable land with low and medium K values are located within the wide and flat river terraces of the downstream Inva. The Obva porech’e is
ENVIRONMENTAL LIMITATIONS ON AGRICULTURE
larger and is characterized by a more diverse pattern of K values. The most indented tracts (with K . 13) are situated there. The other tracts of arable lands have a range of K values from 1.02 to 13. In general, large tracts within river terraces and the rather flat interfluvial drainable plains are characterized by lower K values. Small, tilled river valleys in the northern part of the Obva porech’e are characterized by high K values, similar to those of the Inva porech’e. The most indented tracts (K . 13) are located within highly eroded hilly landscapes. The reason is that the better climatic conditions in the Obva porech’e allow more types of landscapes, including hilly ones, to be used in farming. The pattern analysis of the Obva porech’e and the Inva porech’e shows that in spite of similar geological, geomorphological, soil condition and historical backgrounds their agricultural landscape patterns differ considerably. The agricultural landscape pattern of the Inva porech’e is much poorer and more uniform, whereas the better climatic conditions of the Obva porech’e allow more types of landscapes to be used in farming. Agricultural landscapes of this type (porech’e) are situated mainly in the north and NE of the EEP within the area with severe climatic conditions but well-developed old river valleys. We can regard them as relict agricultural landscapes, as their spatial agricultural and sometimes settlement patterns have been preserved since the introduction of cultivation more than a thousand years ago. The main reasons for conservation of the agricultural spatial patterns in the NE of the EEP include the initially isolated geographical position of porech’e agricultural ecosystems, which saved the area from war and other disturbances; centuries-old extensive agricultural development; and the ecological optimum of the spatial pattern of porech’e for the taiga zone. In addition, the complicated and contrasting environmental conditions rigidly limited the possibilities for agricultural land use and prevented repeated tillage and disturbances of the ecological balance within the area.
Western temperate zone of young landscapes The western part of the forest zone of the EEP was cultivated very rapidly. If the northeastern porech’e agricultural landscapes grew until approximately the 20th century, according to the land cadastre of the 15th century (Pistzovaya kniga), in the west all areas suitable for agriculture were cultivated by the end of the 15th century. This was a result of rapid population growth owing to the presence of river routes for west– east and north –south transit, which crossed the area. One of the most ancient Russian towns, Novgorod, appeared there as a centre of a state starting at the end of the 8th
23
century. At the same time the natural conditions were rather poor. Although the environmental limitations were less strict than in the NE because of a milder climate (Table 4), geological factors played a greater role here. The area affected by the Valday (Wu¨rm) glacier is characterized by the most contrasting and mosaic-type geological and geomorphological conditions, with young poorly developed and usually boggy river valleys. Hence the Novgorod medieval republic suffered from a lack of food and depended on grain imports from southern and central areas of the EEP. This was one of the reasons for Novgorod losing its independence in the 15th century. However, the first agricultural landscapes were productive and large. They are called poozer’e (which in Russian means an area near a lake), and they extend over vast flat plains formed by sandy– loam and loam limnoglacial deposits. The centre of these plains is occupied by large lakes of glacial origin. In spite of a tendency to overwetting (bogging) these limnoglacial plains have many advantages, including better microclimate, potentially fertile deposits and excellent natural fertilizer, the lacustrine mud sapropel. One of the most wellknown poozer’e agricultural landscapes with Novgorod as its centre is situated around Lake Ilmen’. Together with opol’e, poozer’e are the most ancient agricultural landscapes formed by the Slavic population in the forest zone of the EEP. In spite of its fertile soil the poozer’e is a young landscape developed only within the Valday and Moscow glaciation areas. Older lakes of glacial origin have been drained by rivers and nowadays there are no natural lakes (except small karst ones) in the east of the EEP. Unfortunately, the limnoglacial flat plains make up only a small area in the zone, and young morainal hilly landscapes are predominant (the Valday highland). In the course of agricultural development of the Valday highland, tillage did not cover large areas (as in the opol’e and poozer’e) or tracts (as in the porech’e) but instead appeared as small scattered patches (Fig. 5a). All of these were located within moraine loam sandy kames instead of in boggy young river valleys. Until the 16th century the settlement pattern was the same as in the east: small villages (1–3 farms) surrounded by arable lands at a short distance. General Land Surveying maps from the 18th century show that at the end of the 18th century the rural settlement pattern changed in the western zone (in contrast to the northeastern zone). According to these maps, villages were situated on the tops of moraine kames, with an average distance of about 3 –5 km between them (Fig. 5a). At the same time, tillage was spread over the entire territory without any conformity with the contemporary settlement pattern. The distribution of tillage reflected the
24
O. N. TRAPEZNIKOVA
Fig. 5. Distribution of arable land within the Valday moraine highland in (a) the 18th century, (b) the 19th century, and (c) the 20th century. Legend: 1, kame hills formed by glacial loamy sand with layers of clay and sand; 2, river and lake terraces formed by alluvial loamy sand and loam deposits; 3, small stream valleys and flood terraces formed by alluvial loamy sand and loam deposits; 4, interhill depressions with bogs; 5, tillage; 6, lakes.
ENVIRONMENTAL LIMITATIONS ON AGRICULTURE
previous settlement pattern. This is proven by the analysis of land cadastres, which recorded a sharp decrease in the number of settlements within the Novgorod region during the 16th–17th centuries. This was due to many social disasters, including war (1558– 1583), political and administrative terror inflicted by Ivan IV (the Oprichnina, 1565– 1584), and ‘Troubles’ (1603–1613). As a result, the number of settlements decreased by 5–10 times. After the end of the ‘Troubles’ economic renewal took place but the previous settlement and spatial agricultural patterns were never restored. Instead of living in numerous small settlements peasants lived in larger but fewer villages, at first near fortresses, where they were protected from enemies and numerous ‘evil people’ as a result of the ‘Troubles’. In addition, the government needed soldiers and it gave new landlords land and peasants as a service award. These new landlords had to protect and control their peasants, and large villages were more suitable for this task. Besides safety reasons, increasing village size was useful because it allowed community members to graze cattle collectively and help each other. At the same time, in large villages peasants were doomed to lack of nearby arable land, accessible for regular fertilizing. As a result, the allotment of every peasant family consisted of separate close and distant strips with ‘good’ and ‘bad’ land. This made forced crop rotation obligatory for every member of a community. The majority of arable land near neglected villages turned to wasteland and the agricultural pattern changed (Fig. 5b). Before this change arable lands were located mainly within moraine loam sandy kames, whereas in the 18th–19th centuries the population increased and had to cultivate all of the area, even where the land was not good for agriculture (Fig. 5b). This caused environmental disturbance, soil erosion, bog development and hence decreasing crop production. As a result, in the 20th century many people left for towns and the tillage area decreased (Fig. 5c). Only tillage within kame hills and near villages remained.
Central temperate zone of older landscapes The central zone is a transitional one (Fig. 1) and it is characterized by agricultural patterns inherent to the other zones, including poozer’e, interfluvial (similar to opol’e) and valley (similar to porech’e) agricultural landscapes. The change in the settlement pattern within this zone followed that of the western zone but with a time lag. Unfortunately, compared with the Novgorod area, fewer medieval historical documents (especially cadastres) describing the central zone remain today. Generally, the zone was characterized by larger agricultural plots
25
than in the west and their pattern changed several times as a result of non-strict natural limitations and high social pressures.
Northwestern cold zone of young landscapes Generally, this is the zone with the worst combination of natural conditions within the area of intensive agricultural development, as it is characterized by a severe climate and young morainic natural landscapes (Fig. 1, Table 4). This is a zone of very selective spatial agricultural patterns, where populations could not earn their living without other activities, such as handicraft, seasonal work, fur trading, hunting and fishing. Real agricultural landscapes are located only in special, most favourable natural environments, such as limnoglacial plains (poozer’e of Onega and others) and well-drained flat limestone plateaux with their fertile calcareous soil (Kargopol’ sush: in Russian, sush means dry land).
Conclusions Our research shows intimate interrelationships between natural conditions and agricultural and settlement patterns. In fact, the strictness of natural limitations determines possible spatial agricultural patterns under certain social, economic and technological circumstances. Two natural trends, climatic and geological, determine agricultural development within the forest zone of the EEP, forming five zones with different types of spatial agricultural patterns (types of historical agricultural landscapes). The most common types of agricultural landscapes formed under the various natural conditions have their own historical names: ‘opol’e’, ‘poozer’e’, ‘porech’e’ and ‘sush’. The agricultural spatial pattern in the east and south of the EEP (porech’e and opol’e) has been maintained since the beginning of agricultural development more than 1000 years ago. On the other hand, the spatial agricultural pattern in the west and centre of the forest zone repeatedly changed following corresponding displacements of the population. Our research shows the inherited pattern of those agricultural landscapes, such as porech’e, which have developed under the most severe natural conditions. The pattern of porech’e has remained the same through the centuries and one can use it when looking for ancient rural settlements. On the other hand, under more benign natural conditions (at Valday highland, for example) the agricultural pattern was less stable and could change greatly in the course of its development because social pressure was more important than natural limitations. So the comtemporary pattern of these agricultural
26
O. N. TRAPEZNIKOVA
landscapes cannot be taken as a basis for research into ancient rural settlements. Nevertheless, even in this case agricultural lands still prevail only within certain geological units and thus we can determine the spatial framework for extinct settlements in different natural conditions. This research is supported by the Russian Foundation for Basic Research (project 08-05-00755).
References ANON. 1898. Materials for land assessment of Perm gubernia since 1898 to 1900, Vol. 1. Okhanskii uezd, Perm [in Russian]. ANON. 1978. Agrarian history of the Northwestern Russia. 16th century. General results for development of the Northwest. Nauka, Leningrad [in Russian]. ANON. 1997. Archeology of the Republic of Komi. DiK, Moscow [in Russian]. ANON. 2001. General parameters of development for towns and districts of Permskaya Oblast. Statistical Report. Permoblstat, Perm [in Russian]. Bader, O. N. 1951. Ancient settlements of Vetluga and Undja. Materialy i issledovaniya po archeologii SSSR, 22, 110–158 [in Russian]. Chazov, B. A. 1961. Physiographic zoning of Permskaya Oblast. Voprosy geografii, 55, 55– 67 [in Russian]. Markova, A. K., Smirnov, N. G., Kazantseva, N. E., Kozharinov, A. V. & Simakova, A. N. 1995. Late Pleistocene distribution and diversity of mammals in Northern Eurasia. Paleontologia i Evolucio, 28–29, 5 –143. Milkov, F. N. 1964. About origin of opol’e in the Russian plain. Voprosy regional’nogo landshaftovedeniya i geomorfologii SSSR, 8, 20– 27 [in Russian].
Nefedova, T. & Loffe, G. 1998. Continuity and Change in Rural Russia. A Geographical Perspective. Westview, Boulder, CO. Odum, H. T. 1994. Ecological and General Systems: an Introduction to Systems Ecology. University Press of Colorado, Boulder, CO. Sedov, V. V. 2005. Slavs. Ancient Russian People. Historical Archeological Research. Znak, Moscow [in Russian]. Shpolyanskaya, N. 2003. Big Climatic Dispute, ESCO, No. 3. World Wide Web Address: http://esco-ecosys. narod.ru/2003_3/art82.htm. Slicher van Bath, B. H. 1963. Yield Ratios, 810–1820. Afdeling Agrarische Geschiedenis, Wageningen, Bijdragen, 10. Smirnova, O. V., Popadyouk, R. V. et al. 1995. Current state of coniferous– broad-leaved forests in Russia and Ukraine: historical development, biodiversity, dynamic. Pushchino. World Wide Web Address: http://oaks.forest.ru/eng/publications/smirnova/ index.html. Trapeznikova, O. N. 2005. Indicative Properties of Agricultural Landscapes for Studying Global and Local Changes of Environment within Taiga Zone of the Russian Plain. Understanding Land-Use and LandChange in Global and Regional Context. IGU–LUCC. Science Publishers, Enfield, NH, 47– 60. Trapeznikova, O. N. 2007. Relict agricultural ecosystems of the East European plain. 25 years of landscape ecology: scientific principles in practice. In: Bunce, R. G. H., Jongman, R. H. G., Hojas, L. & Weel, S. (eds) Proceedings of the 7th IALE World Congress, Part 2, Wageningen, Netherlands. IALE Publication Series, 4, 737. Uglov, V. 1987. Climate. Agrinatural and agricultural zoning of nonchernozem zone of European part of Russian Federation. Publishing House of Moscow University, Moscow, 28– 42 [in Russian].
Results of the struggle at ancient Ephesus: natural processes 1, human intervention 0 ¨ KNER3 & I˙LHAN KAYAN4 JOHN C. KRAFT1, GEORGE RAPP2*, HELMUT BRU 1
Department of Geological Sciences, University of Delaware, PO Box 250, Schwenksville, PA 19473, USA
2
Department of Geological Sciences, University of Minnesota Duluth, Duluth, MN 55812, USA 3
Department of Geography, University of Marburg, D-35032 Marburg, Germany
4
Department of Geography, Ege University Izmir, TR-35100 Bornova-Izmir, Turkey *Corresponding author (e-mail:
[email protected]) Abstract: Coastal areas have been prime locations for habitation and commerce. Early authors such as Pausanias (second century CE), and Strabo (64 or 63 BCE–24 CE) noted the impacts of shoreline changes. Geomorphological and subsurface geological data, combined with archaeological excavation and ancient texts, indicate a long interplay between natural processes of estuarine infilling by sediments from the Ku¨c¸u¨k Menderes River (ancient Cayster River) and multiple attempts of human intervention to preserve the harbours of Ephesus. Strabo noted that harbour engineering efforts there, such as the construction of a mole to prevent siltation, instead created a sediment trap that made things worse. The pre-Holocene river valley was inundated by Holocene sea-level rise that formed the ancient Gulf of Ephesus. Extensive palaeogeographical studies, based on sediment coring, geomorphology, archaeology and history, have provided details of the problems the inhabitants faced in keeping vital harbours in operation. Dating and analysis of sedimentary deposits has allowed the description of shifting river courses, floodplain changes, human intervention, and anthropogenic deposits at Ephesus. During and following Classical times sediment deposition rapidly began to fill in the embayment, requiring the inhabitants to regularly shift the harbours westward. Ultimately, it was to no avail.
Coastal areas host a large percentage of the world’s population. These areas have high energy available from wind and waves, so they are frequently sites of geomorphological change. In addition, sites that also are located in estuaries are in jeopardy of being overrun by alluvial deposition from riverine sources. Ancient Ephesus, on the Aegean coast of Turkey, is one such site. In this paper we describe the human reaction to the continual progradation of the ancient Cayster River delta floodplain past ancient Ephesus. Many efforts were made to interrupt or slow geological processes that were leading to the demise of the city and its important harbour. The subsurface stratigraphic sections in river valleys and estuaries contain the fossil and sediment record of past environments and their geographies, as well as the record of local relative sea-level rise. From the study of these sediments and their included fossils we can determine with precision the loci of ancient river channels, floodplains, backswamps, deltas, lagoons, barriers, and other features commonly associated with the coastal zones of Greece and Anatolia. We have spent over 20 seasons and drilled over 200 cores to determine
the palaeogeography of ancient Ephesus. This paper summarizes the data contained in approximately 20 of our publications, the most salient of which are referenced. Dating was provided by radiocarbon, historical texts, or in some cases archaeological artefacts. Our work at Ephesus was carried out ¨ sterreichisches Archa¨olounder the aegis of the O gisches Institut (Austrian Archaeological Institute) excavations there. Ancient Ephesus was located along the southern flank of a Neogene –Quaternary-age graben. Major archaeological excavations began there in the 1860s and have been continued for most of the past 100 years by researchers from the Austrian Archaeological Institute and their Turkish colleagues. The main archaeological sites occupy an irregular area roughly 6 km by 3 km. The long-term work at Ephesus by the Austrian Archaeological Institute indicates that archaeological remains vary in age from the Neolithic period to modern times. Throughout the early periods of occupation, the inhabitants had to adapt to the ever-changing morphologies of the ancestral Gulf of Ephesus, especially the overwhelming impact of the progradation of the ancient Cayster River. Here can be
From: Wilson, L. (ed.) Human Interactions with the Geosphere: The Geoarchaeological Perspective. Geological Society, London, Special Publications, 352, 27– 36. DOI: 10.1144/SP352.3 0305-8719/11/$15.00 # The Geological Society of London 2011.
28
J. C. KRAFT ET AL.
seen evidence of people adapting to ever-changing environments over the past 7000 years. The geomorphologically stable areas of Ephesus and the Derbent River (Fig. 1) (the ancient Marnas and Selinus rivers) were located along the southern flank of the graben where there were the limestone (marble) horsts of Panayirdag˘ (ancient Mount Pion) and Bu¨lbu¨ldag˘ (ancient Mount Preon), along with isolated smaller horsts. Approximately 3000 years ago the Amazons are said to have dedicated a sacred site to the goddess Artemis, resulting in the construction of the earliest Artemision (a temple to the Greek goddess Artemis; Roman name Diana) on the shoreline of the ancient Gulf of Ephesus (Callimachus Hymn III, Callimachus 1973). From the time of the construction of the earliest temple of Artemis (c. 800 BCE) on a shoreline in ancient Ephesus to its destruction a millennium later by an earthquake, the inhabitants of Ephesus were engaged in a continuous struggle with the effects of colluvium derived from the uplands and, more seriously, alluvium from the prograding Marnas and Selinus rivers (Kraft et al. 2001; Bru¨ckner et al. 2008). In the two millennia since its demise the as-yet unexcavated city surrounding the temple (and later its ruins) was buried by 6 m of alluvium and colluvium (Bru¨ckner et al. 2008). The Artemision has now been excavated into the surface of the floodplain. Figure 1 shows the present-day distribution of the various geomorphological elements of the
lower Ku¨c¸u¨k Menderes River delta floodplain and the lesser impact of alluvium from the flanking rivers along the northern and southern sides of the area. In the lower regions of the delta are areas of seasonal swamp as well as marsh in the immediate coastal region. The abundant isolated meander loops indicate that the river has repeatedly changed course. The floodplain geometry was permanently altered by the construction of the Ku¨c¸u¨k Menderes Canal to alleviate flooding conditions as well as facilitate irrigation to the abundant agricultural areas now occupying the lower delta floodplain. Sandy barriers now occur along the shoreline of the Aegean Sea. The earliest evident remnants of sandy barriers as the delta prograded seaward are also shown. At the immediate coastal area the shore face drops relatively rapidly into the deeper and clearer marine waters of the Aegean Sea. Along the northern flank are two freshwater lakes surrounded by marsh that were originally marine embayments of the ancestral Gulf of Ephesus, as shown by marine fossils in our sediment cores. Floodplains, backswamps, marshes, lakes, lagoons, barrier accretion plains, talus slopes, alluvial fans, and other such features are only the surface expression of 3D sedimentary environments of deposition. From floral and faunal elements we determined the nature of the sedimentary environments of deposition in three dimensions. To do this we drilled over 200 cores to determine the
Fig. 1. Physiographic features of the Ephesus region of the Ku¨c¸u¨k Menderes floodplain and delta.
DEMISE OF THE EPHESUS HARBOURS
distribution and change of marine, coastal, and fluvial environments of deposition over the long term. Carbon dating provided us with the time frame. We used five coring units capable of extracting sediments from depths of 5– 30 m. From our palaeogeographical studies of the delta floodplain of the Ku¨c¸u¨k Menderes River (ancient Cayster River) we have been able to show that the earliest known archaeological sites (Neolithic, 7000 BP) in the region of ancient Ephesus lay along the southern flank of a relatively deep (30 m) clear water embayment of the ancestral Gulf of Ephesus. The initial Holocene sediments at the time of the Mid-Holocene sea-level highstand (c. 1 m above present sea level) consisted of sand with an abundant clear water marine molluscan fauna. Evidence from water wells across the plain nearly always shows an abundant and highly varied marine molluscan fauna. We identify two Neolithic sites (N) in Figure 2. The northern site lies on a pre-Holocene gravel fan with scattered shells of edible molluscs of varied clear water marine species. They are accompanied by fragments of red pottery that have been dated tentatively to the Neolithic. Along the southern flank, inland, and at higher elevation on a Neogene gravel fan between the two valleys of the Marnus and Selinus rivers, lies a newly discovered Chalcolithic site that was
Fig. 2. Mid-Holocene Neolithic marine embayment.
29
being excavated in 2008 by the Austrian Archaeological Institute. From drill-core evidence it appears that the site was initially occupied in Neolithic time. We do not have precise knowledge of the eastern landward extent of the ancestral Gulf of Ephesus in Neolithic time. However, it appears from the occurrence of bottom-set delta facies to have inundated the Quaternary graben valley to at least 18 km from the sea. Figure 3 is a west to east geological cross-section from the present barrier accretion plain along the coastline of the Aegean Sea to the Belevi Gorge (Fig. 1). It should be noted that the greater portion of the marine sediment that has filled the valley consists of marine muds, as indicated by the molluscan assemblage, derived from the Ku¨c¸u¨k Menderes River. Of far greater importance are rates of deposition. In the first 2600 years of deposition, starting from 7000 years BP, there was clearly a much lower rate of deposition of marine muds than in the 200 years from Classical –Archaic time to the Hellenistic period. The rates of deposition since Hellenistic time appear to be much lower as the sea bottom of the ancestral Gulf of Ephesus has migrated seaward into the Aegean Sea. The river floodplain alluvium, as expected, thickens in a landward direction. One can speculate from this that human impact on the marine environment has
30
J. C. KRAFT ET AL.
Fig. 3. Profile of the Ku¨c¸u¨k Menderes floodplain and the Neolithic embayment.
been profound since Neolithic time. The Neolithic expansion of agriculture across Anatolia and the even more intensive farming activities from Classical– Archaic time onwards to the present time have provided a driving mechanism for the high deposition rates. Figure 4 illustrates shoreline and harbour configurations from c. 300 to 0 BCE. During the Hellenistic and early Roman time there was a prograding lower deltaic sequence between varied birdsfoot sandy distributaries of the ancient Cayster River. Such deltaic formations were rapidly changing. Between Mount Pion and Mount Preon lay an area of relatively deep water in Hellenistic times, with the strand or shoreline in the time of Lysimachus (360–281 BCE) lying very close to the immediate base of the flanks of the two mountains. The extent of the city of Lysimachus is shown. One of the more impressive structures in the entire history of the occupancy of ancient Ephesus is the massive defensive wall that stretches along the top of Mount Preon to the SE and then to the NE across Mount Pion and around to the northern flank where, based on more than 20 cores, we posit a series of sacred harbours that were forced progressively westward as the delta shoreline rapidly prograded. The sacred harbour of the Artemision, which at the time of construction would
have been built in the immediate coastal zone, was thus gradually forced westward. This is perhaps the first series of major adaptations to the environment required of the people occupying the immediate coastal zone of the Artemision and of Hellenistic and earliest Roman Ephesus. Furthermore, in Hellenistic and Roman times the city was built into the floodplain surrounding the later Artemision, to be in turn buried under 5 m of alluvium until rediscovered a century and a half ago. In the time of the earlier Lydian city (seventh – sixth centuries BCE), located in the valley between Mount Pion and Mount Preon, and through the initial occupancy of the city that Lysimachus built, the harbours were forced to move from the vicinity of the Artemision to this area, where ships could approach the shoreline near the base of the steep slopes along the flanks of Mount Pion and Mount Preon. However, as Hellenistic time progressed, rapid shoaling of this harbour area began as the prograding delta of the Cayster River continued to move to the west. By the time of Attalus II Philadelphus (159 –138 BCE) the great mole (a breakwater built to enclose and protect the harbour) along the northern side of the harbour of Ephesus had been established. We have little knowledge of the early Lysimachan harbour other than that it would have started with ship access directly to the foot of the
DEMISE OF THE EPHESUS HARBOURS 31
Fig. 4. Ephesus: Hellenistic –Early Roman times.
32
J. C. KRAFT ET AL.
two mountains. By the time of Attalus II Philadelphus human impact plus that of the Cayster River on the environment of the harbour area had greatly accelerated. Indeed, as early as 190 BCE Livy (1997, Book 37, 14 –15) commented on the nature of the entrance to the harbour of Ephesus as already having only a narrow entrance and that it was ‘full of shoals’. Livy (59 BCE –17 CE) noted that a Roman, Rhodian and Pergamene allied fleet had tried to attack the fleet of the Seleucid King Antiochus III (242–187 BCE) that was trapped in the harbour of Ephesus. The Roman commander of the fleet proposed to sink ships to block the harbour. However, he was persuaded not to as it would have destroyed a great harbour; Livy noted ‘because the mouth of the harbour was like a river: long and narrow and full of shoals’. We have schematically presented possible configurations of shoals with narrow channels in Figure 4; however, we have not yet been able to precisely delineate the configurations as described by Livy. We can perhaps better understand a major environmental impact created by the engineers of Attalus II Philadelphus. Strabo wrote (Strabo 1924, Book 14, 1, 24), ‘The city has both an arsenal and a harbour. The mouth of the harbour was made narrower by the engineers, but they, along with the king who ordered it, were deceived as to the result. I mean Attalus Philadelphus; for he thought that the entrance would be deep enough for large merchant vessels–as also the harbour itself, which formerly had shallow places because of silt deposited by the Cayster River –if a mole were thrown up at the mouth, which was very wide, and therefore ordered that the mole should be built. But the result was the opposite, for as the silt, thus hemmed in, made the whole of the harbour, as far as the mouth, more shallow. Before this time the ebb and flow of the tides would carry away the silt and draw it to the sea outside.’ This quote describes a major engineering mistake, which created a quiescent settling basin out of what was earlier a deep-water major harbour. From then on, from Roman Imperial into Byzantine times, the great harbour had to be dredged continually. From our drill cores in the harbour we determined that the greater portion of the sediments were beach anoxic silts. This confirmed that the harbour was cut off totally from the river, which distributed new silt with each spring flood. The great mole was extended to allow ships to approach the great harbour (see Fig. 5). There also were at least five attempts to clean and maintain the great harbour of Ephesus in the times of Roman Emperor Nero (54 –68 CE) and through the third century CE (Kraft et al. 2000). Apparently, numerous laws were passed to prevent using the harbour as a sewer, disposal area or dump. Many
legal methods were tried to stop negative environmental actions. At times there were massive infusions of money to dredge the harbour and maintain the ever-extending narrow channel to the sea. The negative impacts of the works of Attalus II Philadelphus were thereafter compounded by three centuries or more of dumping of industrial and human waste into the harbour with major impact. In the middle of the third century CE the politician Valerius Festus enlarged the harbour. However, by this time, what had formerly been the open harbour of Lysimachus, with excellent access of ships to the shore, had now been filled with dumped materials and construction debris so that the harbour area progressively moved westward (Fig. 5). During Roman Imperial times the Cayster River delta with its distributaries and marshes continued to prograde several hundred metres to the west. This necessitated continual dredging of the channel to the open sea and the construction of a large mole to prevent the channel from being filled by deltaic river deposits. Also during Roman Imperial times, in the early second century CE, Hadrian had a dam built to divert the Cayster River, to stop the sediment flow into the harbour. The location of this dam has not been determined. By High Byzantine time maintaining the channel to sea from the great harbour of Ephesus became extremely difficult. There are many references (see Foss and Meric¸, cited by Kraft et al. 2000) to transferring cargoes from platforms near the sea to shallow draught barges or boats that could continue up the channel to the great harbour. Figure 5 shows possible locations of such platforms, but we cannot know of these features in any detail. Continued shoaling and the advance of the delta distributaries westward occurred. By Late Byzantine time the first sandy coastal barriers had formed and the great harbour at Ephesus was no longer in use. There is speculation that later harbours may have utilized areas between the mountains along the southern flank of the graben. As noted above, by Roman time massive amounts of fill were being placed in the area of the former harbour of Lysimachus so that the lower city of Ephesus could be built. The great harbour of Ephesus was moved westward. Kraft et al. (2000), Bru¨ckner et al. (2008), and comments in the ancient literature have provided details of the environmental problems. These problems were compounded by Roman efforts to fill in the harbour area of Lysimachus to provide a foundation for the construction of public buildings and marble streets and columns from the vicinity of the great theatre of Ephesus to the harbour basin. Figure 6 delineates the nature and timing of the fill materials from an excavation trench. By 400 BCE coastal swamp sediments contained a metre of fill with
DEMISE OF THE EPHESUS HARBOURS 33
Fig. 5. Palaeogeographical map of the environs of Ephesus during Roman Imperial to Byzantine times showing the shoreline positions at three periods.
34 J. C. KRAFT ET AL. Fig. 6. Schematic diagram of the sediments encountered in an excavation trench, and a core from the bottom of the trench allowing us to locate the position of the shoreline in the time of Lysimachus.
DEMISE OF THE EPHESUS HARBOURS
abundant sherds and other debris extending into the edge of the sea. By the first century CE massive amounts of fill are indicated, and this continued throughout the second, third, and fourth centuries. During this time foundation walls were constructed to support a large number of public buildings in the lower city and its adjacent port facility. In addition, a temple and other public buildings were constructed along the northern mole of the great harbour. Archaeological excavation was able to delineate and date over 4 m of fill in the lower city. This contrasted with the centuries of effort to prevent people filling in the harbour as well as the dredging to clear the harbour to allow continuation of shipping.
35
The harbour of the Lydian (seventh –sixth centuries BCE) city at Ephesus as well as the harbour of Lysimachus (in the same location) in Hellenistic time was a natural deep-water harbour sheltered from westerly, southwesterly, easterly and northerly winds (Kraft et al. 2005). On the other hand, the sacred harbour of the Artemision for the Greek city of Androclus (c. 1000 BCE) as well as the Classical Greek city was located in the area of the modern city of Selc¸uk Ephesus. This was between Aya Suluk and the Artemision and the eastern slopes of Mount Pion, B in Figure 7, whereas A shows a possible but not proven early harbour. Throughout the earliest times the site of the Artemision had to be defended from inundation by
Fig. 7. Evolution of the harbours of Ephesus and the Artemision over two millennia from our core data, geomorphology studies, archaeological excavation and literature sources.
36
J. C. KRAFT ET AL.
alluvium from the Marnas, Selinus and Serenc¸i Rivers, as well as colluvium from the slopes of Aya Suluk. Indeed, the site of the Artemision throughout the history of its use was subject to flooding. C and D in Figure 7 show locations of later sacred harbours for the Artemision as the delta floodplain of the Cayster River continued to prograde seaward. E shows the location of the Lysimachus harbour, F locates the great harbour of Hellenistic and Roman times, and G and H are suggested locations of possible harbours of Late Byzantine and Venetian time. I, J and K indicate the locations of open sea harbours in later times. L indicates the possible location of the Panormus harbour, which was the last attempt at a commercial harbour for Ephesus but was no longer considered to be important. The excavated marble-block road to Panormus Harbour was constructed from the ruins of Ephesus in Selc¸uk time (1071–1299 CE). This was the last attempt by the inhabitants to utilize a harbour within the ancient embayment.
Conclusions Over the past seven millennia humans have occupied the southern flank of the ancestral Gulf of Ephesus. Over time, they were forced to continually adapt to ever-changing coastal configurations created by the colluvial and alluvial sedimentary processes. As the main delta of the ancient Cayster River continued to fill in the ancestral gulf, people occupying the coastal zone of the various loci of the ancient cities of Ephesus had to continually adapt and change their patterns of occupancy. From the time of construction of the first Artemision, c. 1000 BCE, buildings, roads and harbour facilities were affected by the dynamic nature of the coastal environment. However, by the time of the Hellenistic construction of the greater city of Ephesus by Lysimachus the human actions came into direct conflict with the natural processes of progradation and aggradation of the Cayster River floodplain and delta. In the case of the Artemision, the sacred harbour was of necessity moved several times to the west along the base of Mount Pion. Indeed, in excavations at the Feigengarten and the Via Sacra(e) at the base of Mount Pion, deep-water conditions were shown to be rapidly filled in by the prograding deltaic sediments. The excavations by the Austrian Archaeological Institute indicate that two roads running from the Artemision to the lower city by the great harbour were buried by up to 5 m of a composite of colluvium, alluvium and structural debris. The area of the Artemision ruins was buried under up to 6 m of
colluvium from Aya Suluk and alluvium of the Marnas and Selinus Rivers as well as the Cayster River delta floodplain. In these areas natural processes of deposition dominated (Bru¨ckner et al. 2008). On the other hand, at the foot of the western flank of Mount Pion and the northern flank of Mount Preon, in the area of the various harbours, over a period of five or six centuries human impact on the environment clearly dominated changes in landforms. In essence, from Classical times until High Byzantine times human actions and structures predominated. The struggle between the peoples of Ephesus and the natural alluvial and colluvial processes continued for over one and a half millennia. However, in the end, the natural processes of sedimentation prevailed. The struggle ended by Byzantine times with the score: natural processes 1, human intervention 0. Currently, the nearest harbour to the city of Selc¸uk Ephesus is the resort harbour of Kus¸adasi far to the SW along the rocky coast of the Aegean Sea. All seven figures in this paper are adapted and modified from Kraft et al. (2000). We thank C. Kubeczko, who adapted the figures, and H. Englemann for his intensive analysis of the ancient literature. Also, we are indebted to two anonymous reviewers, who made many worthwhile suggestions.
References Bru¨ckner, H., Kraft, J. C. & Kayan, I. 2008. Vom Meer Umspu¨lt, vom Fluss begraben Zur Pa¨laogeographie des Artemisions. In: Muss, U. (ed.) Die Archa¨ologie der ephesischen Artemis, Gestalt und Ritual eines Heiligtums. Phoibos, Wien, 21–31. CALLIMACHUS. 1973. Calimachus. Loeb Classical Library, Cambridge. Kraft, J. C., Kayan, I., Bru¨ckner, H. & Rapp, G. R. 2000. Geologic Analysis of Ancient Landscapes and the Harbors of Ephesus and the Artemision in Anatolia. ¨ sterreichischen Archa¨ologischen Jahreschefte Des O Institutes in Wien, 69, 125– 231. Kraft, J. C., Kayan, I. & Bru¨ckner, H. 2001. The Geological and Paleogeographical Environs of the Artemision. Der Kosmos der Artemis von Ephesos, ¨ sterreichischen herausgegeben von Ulrike Muss. O Archa¨ologischen Institute, Sonderschriften, 37, 123–133. Kraft, J. C., Bru¨ckner, H. & Kayan, I. 2005. The Sea under the City of Ancient Ephesus. In: Brandt, B., Gassner, V. & Ladsta¨tter, S. (eds) Synergia. Festschrift fur Friedrich Krinzinger, 1. Phoibus Verlag, Wien, 147–156. Livy, T. 1997. History of Rome. Loeb Classical Library, Cambridge. STRABO. 1924. Geography. Loeb Classical Library, Cambridge.
Quaternary landscape evolution and human occupation in northwestern Argentina M. M. SAMPIETRO VATTUONE1* & L. NEDER2 1
INGEMA – CONICET, Espan˜a 2903, 4000 San Miguel de Tucuma´n, Argentina
2
INGEMA, Juan Luis Nougue´s 1363, 4000 San Miguel de Tucuma´n, Argentina *Corresponding author: (e-mail:
[email protected])
Abstract: Our study area is located in northwestern Argentina. It is a semiarid valley in which developed agricultural pre-Columbian settlements were located. The objectives of our research were to establish the geomorphological characteristics of the area, its relative chronological development, and the relationships between geomorphological development and pre-Columbian settlements. Pre-Quaternary lithologies are represented by a metamorphic basement that is commonly exposed on slopes and belongs to the Precambrian and Cambrian periods. Tertiary sediments from several formations are exposed over an extensive surface forming cuesta relief landforms. Quaternary landscape units were classified according to their genesis into structural – denudational landforms (denudational slopes and structural scarps), denudational landforms (covered glacis), fluvio-alluvial landforms (alluvial fans, fluvial fans, and fluvial terraces) and aeolian forms (stabilized dunes). Archaeological sites belonging to the Formative (500 BC– AD 1000) and Regional Development (AD 1000– 1500) periods were identified. The main archaeological sites are located on the surfaces of debris-flow deposits and some covered glacis. They are characterized by the presence of residential units together with agricultural structures (terraces and irrigation channels). The earlier settlements (Formative period) are restricted to alluvial fan landforms (debris-flow deposits), where present hydrological supply is lower than in the rest of the study area. Later settlements (Regional Development period) are juxtaposed with earlier settlements in the south of the area, where present hydrological supply is higher owing to larger river catchments and moistureladen winds from the SE.
This research was carried out on the western hillsides and piedmont of the Calchaquı´es Summits (Santa Marı´a Valley, Tucuma´n Province, northwestern Argentina). The study area’s limits are the Amaicha River to the south (in Tucuma´n province), the Campo La Hoyada to the north (in Salta Province), the watershed over the Calchaquı´es Summits to the east, and the Santa Marı´a River to the west (Fig. 1). The Santa Marı´a Valley is a deep depression with a planar floor oriented north –south. It extends between the Quilmes Range to the west and the Calchaquı´es Summits and Aconquija Ranges to the east. Water flows from south to north, in the opposite direction to the normal regional inclination. The altitude at the bottom of the valley varies between 1900 and 1600 m above sea level (Strecker 1987). The regional climate is desert-type, with an average annual temperature of 18 8C, and average annual precipitation of less than 250 mm (most falling during the summer season) (Perea 1994). Precipitation varies according to hillside orientation, and because moisture-laden winds regularly
come from the east and the SE, our study area is the drier sector of the valley. Wind action produces desiccation and hardening of exposed surfaces, and also the movement of fine particles from the surface, which constitute the material for the dune fields of the region.
Antecedents Geological antecedents The geology of the Santa Marı´a Valley consists of medium- and high-grade metamorphic basement rocks, overlain discordantly by Early Tertiary layers (Salta Group) of fluvial sedimentary rocks. Over them, also discordant, lie Miocene and Pliocene (Santa Marı´a Group) fluvial and lacustrine sediments (Gonza´lez et al. 2000). Quaternary deposits are of either fluvial or torrential origin, with coarse sand textures close to the summits and sandy clay textures in the centre of the valley (Gonza´lez et al. 2000). According to Ruiz Huidobro (1972), during the Quaternary, fluvial cycles determined the formation of piedmont
From: Wilson, L. (ed.) Human Interactions with the Geosphere: The Geoarchaeological Perspective. Geological Society, London, Special Publications, 352, 37– 47. DOI: 10.1144/SP352.4 0305-8719/11/$15.00 # The Geological Society of London 2011.
38
M. M. SAMPIETRO VATTUONE & L. NEDER
Fig. 1. Study area location.
levels over older pediments. Strecker (1987) established the presence of five pediment levels formed by fanglomerates. He also observed the large extent of old coalescent alluvial fans, both landforms being of Early Pleistocene age. The Holocene is represented by debris-flow deposits covering the piedmonts of the Aconquija Ranges and the
Calchaquı´es Summits, with fanglomerates, sands and fluvial silts in the centre of the valley. Petrocalcic horizons and dunes are associated with arid periods (Sayago et al. 1998). The Middle Holocene period had wetter conditions at a regional level, and is linked to the generation of thick clastic sequences associated with alluvial fans or debris flows
LANDSCAPE AND OCCUPATION IN ARGENTINA
produced by increased glacial and periglacial activity in summit areas (Sayago et al. 1998). At a regional level, from the palaeoenvironmental point of view, the Pleistocene period is characterized by the presence of loess –palaeosol sequences representative of environmental conditions changing between cold–dry phases (periods of loessic deposition) and warm –wet phases (with pedological development) (Sayago et al. 1998). In the Tucuma´n plain and piedmont, Early and Middle Holocene environments are represented by oscillating dry and wet sequences (Sayago et al. 1998). The oldest palaeosol in the plain has been dated to 6290 + 120 BP (14C years). Its characteristics suggest that the climate was wetter than at present (Sayago et al. 1998). Contemporaneously, in the northwestern area of the Southern Sub-Andean Ranges, the climate was semiarid with wetter phases. Humidity increased progressively through the Late Holocene. The transition from Middle to Late Holocene coincided with a loessic deposition cycle and the development of soils. In the western Chaco plain a palaeosol was dated to 3780 + 40 BP (14C years) (Sayago et al. 1998). In the Tafı´ Valley, close to Santa Marı´a Valley, a sequence of alternating dry and wet periods was identified. A palaeosol reflecting wet conditions was dated to 2480 + 110 BP (14C Tafí Valley
875 ± 20 BP Pollen profile data Garralla (1999)
years) (Sampietro Vattuone 2002). This period was contemporary with the development of Formative Tafı´ pre-Columbian settlements, and lasted until 875 + 20 BP (14C years), when pollen (Garralla 1999) and pedological (Sampietro Vattuone 2002) evidence shows that a drier period was established. During that phase Tafı´ culture collapsed. After that a slight environmental recovery was inferred from pedological evidence. In the Santa Marı´a Valley, Strecker (1987) found sand layers enriched with organic matter on the western side of the Santa Marı´a River. They were dated to 2190 + 530 BP (14C years) and 1470 + 50 BP (14C years). They may correspond to a period with wetter climatic conditions. Other layers located above showed that after 1100 + 70 BP (14C years) drier conditions were established (Strecker 1987) (Fig. 2).
Archaeological antecedents The Santa Marı´a Valley has many archaeological remains from the pre-Columbian period, which we use to outline the following cultural sequence, based on Gonza´lez & Pe´rez (1972). We consider that although this is not the most recent cultural framework proposed (see Olivera 2001) it is the most satisfactory at a regional level.
Santa María Valley
Dry Wet
Dry Wet
1000 BP
1100 ± 70 BP Dryer conditions Strecker (1987) 1470 ± 50 BP Sands with organic matter Strecker (1987)
2000 ± 50 BP Pollen profile data Garralla (1999)
2000 BP
2480 ± 110 BP Paleosol data (Sampietro Vattuone 2002)
39
2190 ± 530 BP Sands with organic matter Strecker (1987)
Inferred data from pedological evidence 3000 BP
Fig. 2. Palaeoenvironmental data from the Santa Marı´a and Tafı´ valleys.
Radiocarbon dated pedological evidence
40
M. M. SAMPIETRO VATTUONE & L. NEDER
Two main cultural phases can be distinguished: (1) pre-ceramics, characterized by the presence of hunter–gatherer populations covering the time period from the arrival of the first humans in the region to 2500 BP; (2) the ceramic period (2500– 500 BP) characterized by the appearance of ceramics, agriculture and sedentary settlements (Gonza´lez & Pe´rez 1972). Within the pre-ceramic stage it is possible to distinguish the Palaeoindian period (? to 7000 BP), which is poorly represented in the regional archaeological record and is characterized by the presence of hunters of large mammals; and the Archaic (7000–2500 BP), with specialized hunter–gatherers with high mobility and an excellent use of natural resource diversity (Gonza´lez & Pe´rez 1972). The Pre-Columbian ceramic period can be divided into three well-differentiated periods: (1) the Formative period (500 BC–AD 1000), during which sedentary settlements were established and agricultural practices started, together with ceramic manufacture and camelid domestication; (2) the Regional Development period (AD 1000– 1500), during which populations grew, defensive structures were built and agricultural systems improved (especially in terms of irrigation structures); (3) the Inca period (AD 1400–1500), during which the Inca Imperium expanded over wide areas of northwestern Argentina, generating changes in the power system (Gonza´lez & Pe´rez 1972). The study area comprises several archaeological settlements representative of the cultural stages mentioned above, especially the Formative and the Regional Development periods. Settlements belonging to the pre-ceramic period were observed in the Amaicha River basin and adjacent areas, in the south of our study area (Cigliano 1961, 1968; Hocsman et al. 2003; Somonte 2007). Archaeological material appeared in open-air workshops, without stratification or associated buildings. Therefore the exact chronology is difficult to establish. Earlier ceramic sites, belonging to the Formative period, were identified in the southern sector of the study area. Buildings from this period are still visible on the surface, although they are fragmentary and obscured by later structures (Sosa 1996– 97, 1999; Aschero & Ribotta 2007; Somonte 2007). The typical settlement pattern consists of circular rooms, similar to those described by Sampietro Vattuone (2002) in the Tafı´ Valley. Walls are built of stones without mortar or foundations. The rooms are dispersed over terraced agricultural fields. Radiocarbon dates obtained for them are 900 + 70 BP (accelerating mass spectrometry (AMS), UGA 8359), 1180 + 40 BP (AMS, UGA 8360) and 1130 + 40 BP (AMS, UGA 8361) (Aschero & Ribotta 2007).
Sosa (1996–97) used the visual interpretation of aerial photographs from the east side of Santa Marı´a Valley, especially the Amaicha del Valle and its surroundings, to identify Formative archaeological settlements. He identified the presence of six occupational settings located along the Amaicha River. Later settlements, located around Los Cardones gulch, belong to the Regional Development period. Radiocarbon dates obtained for these settlements are 460 + 60 BP (LP 1484), 570 + 60 BP (LP 1573) and 930 + 70 BP (LP 1495) (Rivolta 2007). The settlement pattern can be characterized as a semi-urban village; the areas where slopes are more pronounced have terraced structures. Potsherds are of the Santa Marı´a bicolour and crude types (Rivolta 2005). Downhill from this archaeological site, next to Los Zazos, Formative potsherds were collected (cie´naga, vaquerı´as and tafı´ types); they were associated with circular structures and agricultural terraces (Rivolta 2005). The materials and structures showed the superimposition of settlements from the two periods in this sector (Rivolta 2005). Related to this point, Sampietro (1992) commented on the presence of architectural and ceramic features belonging to the Formative and Regional Development periods juxtaposed in the area of the present-day Ampimpa settlements (Fig. 3).
Methods We started by making geomorphological interpretations from 1:50 000 aerial photographs taken by Spartam Air Service in 1971 (there are no more recent photographs available). Thematic cartography of the area involved the production of a lithogeomorphological map, containing the morphogenetic units in chronological order (Fig. 4), and an archaeological map, to represent the location of archaeological sites in landscape context (Fig. 5). At this scale it was impossible to draw an accurate archaeological map, so we decided to delimit archaeological areas instead. The geomorphological interpretation was performed following the approach proposed by the International Institute for Geo-Information Science and Earth Observation (ITC, Netherlands) (Van Zuidam 1976). This approach takes into consideration the genetic factors that make it possible to characterize each landform unit, together with lithology and topography. We considered the determination of geomorphological units to be especially important because: (1) their genesis and temporal evolution is homogeneous over the entire landform surface; (2) they have spatial homogeneity owing to the recurrence of endogenous morphogenetic elements such as lithology, sedimentary composition,
LANDSCAPE AND OCCUPATION IN ARGENTINA
41
Fig. 3. Simplified topographic map with archaeological settings.
stratigraphy, etc.; (3) as these factors (endogenous morphogenetic elements) are common to the whole landscape unit, they permit the extrapolation of palaeopedological, palaeoclimatic, lithostratigraphic and geochronological features to the entire landform unit; (4) they make it easier to have a dynamic and integrated vision of the area’s palaeoecological evolution through the analysis of the evolutionary schemes of each unit (Sayago & Collantes 1991). From an archaeological perspective all of these features allow us to make a better correlation between settlement characteristics and their possible relationships to natural resource exploitation (Sampietro Vattuone 2002). Thematic maps were ground-truthed by field survey, and a geographic information system (GIS) was constructed using ILWIS 3.4 software (Integrated Land and Water Information System). Subsequently, we performed systematic field research in the areas where archaeological sites
were identified. Over these landforms we surfacecollected potsherds (there was no other diagnostic material available on the surface) and surveyed archaeological evidence along one longitudinal transect across each landscape unit. We dug 13 archaeological test pits (1 m 1 m) in agricultural sectors. As structure features are different according to the cultural period, these features were compared across the various surveyed areas. Although they were very scarce, surface-collected potsherds were classified by comparison with pre-existing typologies, using qualitative and diagnostic criteria following the 1st National Convention on Anthropology (Anon. 1966), to establish the agricultural periods represented.
Results and discussion The lithogeomorphological characterization of each landform was based on its genetic characteristics
42
M. M. SAMPIETRO VATTUONE & L. NEDER
Fig. 4. Geomorphological map of the study area. Landform units classified by their origins: 1, structural– denudational; 2, denudational; 3, fluvio-alluvial; 4, aeolian.
and its lithology. From the morphogenetic point of view we were able to distinguish landforms with structural–denudational, denudational, fluvio-alluvial and aeolian origins (Fig. 4).
Archaeological settlements were identified by photo-interpretation and field survey. Pedestrian surveys allowed us to surface-collect potsherds but the scarcity of such material made it impossible to
LANDSCAPE AND OCCUPATION IN ARGENTINA
Fig. 5. Small structures set into the side of highland fertile valleys (called ‘vegas’). Calchaquı´ summits, denudational slopes geomorphological unit.
undertake ceramic seriation. Nevertheless, it was possible to use these fragments as diagnostic materials by comparison with previous typologies from the region. The landforms of structural– denudational origin are relief forms developed by the action of seismic movements that generate scarps of various sizes. These have then been affected by denudational processes, giving superficial morphologies that vary according to lithology, climate and tectonics (Neder & Busnelli 2003). In the study area we have identified (1) denudational slopes, (2) erosion scarps, and (3) cuesta relief landforms. Denudational slopes (on the western side of the Calchaquı´es Summits), developed over igneous –metamorphic basement rock. They extend from the watershed of the Calchaquı´es Summits to the upper limit of the piedmont. These slopes are very steep, with slopes around 45%, and have extant fault scarps. The drainage network is sub-dendritic, and in some sectors subparallel. The fluvial regime is seasonally torrential. It was impossible to establish the presence of archaeological sites by photo-interpretation, although it is highly probable that this area was used for camelid grazing and hunting. By ground survey it was possible to find small structures set into the side of highland fertile valleys (called ‘vegas’) (Fig. 6). Erosion scarps are present in the apical and middle sector of the piedmont, developed around the earliest levels of covered glacis (Covered Glacis L1). Scarps are indicative of tectonic activity and were generated by degradation processes. Cuesta relief landforms developed in the piedmont over Tertiary formations; they have gentle to moderate slopes (15– 45%). Glacis landform units are of denudational origin, forming as a result of mass movement processes. The changes in their relief reflect the morphogenetic action of rapid climate change alternating with stable periods (Van Zuidam 1976). As a
43
result, it is possible to observe smooth surfaces that required a long time to form. As they are climate and time dependent, their incorporation in the Quaternary stratigraphic record, as chronological and palaeoenvironmental indicators, is very important. Birot (1960), studying the dominant geomorphological processes of arid regions, proposed the use of the term ‘covered glacis’ for those surfaces that are covered by a thin layer of detrital deposits. On the western slope of the Calchaquı´es Summits it is possible to distinguish three levels of covered glacis of Pleistocene age, starting at the limit between hillside and piedmont and extending downhill to the west. They are formed by mass movements of rock fragments weathered from slopes and the igneous– metamorphic basement of the hillsides, and can be lithologically defined as texturally coarse debris-flow deposits. The fragments are in a dominantly clastic sandy matrix. The three levels correspond to separate accumulation cycles that also show differences in topography and altitude. The first, and highest, level of covered glacis (Covered Glacis L1) corresponds to the oldest level. It is found at the limit between hillside and piedmont, and has been strongly incised by subsequent morphodynamic processes. In some cases it still has its original triangular shape, principally in the central and northern parts of the study area. Photo-interpretation showed that there are no archaeological structures visible on this level of covered glacis. The next covered glacis level (Covered Glacis L2) is the most extensive in the study area, extending from the apical to middle piedmont and covering Tertiary deposits. This glacis level presents a large quantity of archaeological structures, especially in the central and southern sector of the study area. Circular and quadrangular structures are present. Both are dispersed among agricultural terraces in several areas. Although it was not possible to see by photo-interpretation at the scale we worked at, one semi-urban settlement has been described in the southern sector of the area, at Los Cardones gulch, by Rivolta (2000). Photointerpreted features (such as changes in the texture of the surface by the accumulation of sand with specific patterns or the growth of vegetation around the collapsed walls) allowed us to infer the juxtaposition of buildings from the Formative and Regional Development periods. Pedestrian surveys and collection of archaeological potsherds corroborated this observation. Formative potsherds are represented by the ceramic types Cie´naga incised, polished grey, a few fragments of Aguada, and polished light red. Regional Development period potsherds are represented by the types bicolour Santa Maria, tricolour Santa Maria, crude coarse
44
M. M. SAMPIETRO VATTUONE & L. NEDER
Fig. 6. Archaeological areas related to geomorphological landforms. Landform units classified by their origins: 1, structural – denudational; 2, denudational; 3, fluvio-alluvial; 4, aeolian.
red, and unpolished grey. Finally, the third covered glacis level (Covered Glacis L3) is chronologically the youngest and is located in the lower piedmont. It is composed of texturally finer (silty sand)
materials. It is more extensive in the south of our study area. Archaeological settlements are scarce and are located next to those on Covered Glacis L2.
LANDSCAPE AND OCCUPATION IN ARGENTINA
After the development of the covered glacis levels, landforms of fluvio-alluvial origin were formed. The earliest landforms of this type in our study area are alluvial fans (Early Holocene), followed by fluvial fans (Middle and Late Holocene) and fluvial terraces (Late Holocene). Three variables affect the development of alluvial and fluvial fans: topography, lithology and climate. Topography influences the contribution of sediments, because on very steep slopes mass movement processes generate dense flows with the materials generated by weathering. When slopes are gentler, fluvial activity dominates. Lithology conditions the type and size of rocks broken by weathering (Gutie´rrez Elorza 2001). Climate is the third factor that regulates fan development. Water determines weathering processes, the transportation of materials and plant growth. In semiarid environments, large volumes of clastic sediments, which form alluvial fans, are transported by running water during storms, owing to the short duration and high intensity of precipitation. In the study area we were able to distinguish, in chronological order, first, alluvial fans restricted to the SE sector of the area, formed by debris-flow deposits transported from the hillside downstream by currents and deposited in the apical piedmont. They are texturally finer than those that formed covered glacis. They were formed under wetter conditions than at present. In these areas it was possible to identify residential structures, mainly of circular shape, dispersed among agricultural terraces. Surface-collected potsherds have Formative characteristics as described above. After that, three cycles of fluvial fans were identified. The oldest fluvial fans (C1) were formed of transported Tertiary sediments. They are coalescent and much dissected by subsequent erosive processes. They are located in the central and southern sectors of the area and between the distal piedmont and the bottom of the valley. The next fluvial fans (C2) are restricted to the northern sector of the area; as relicts of large fans buried by subsequent deposits (C3), they appear only in the distal sectors of C3. The final fluvial fans (C3) correspond to later fans of wide extent, and are restricted to the same sector as the previous ones. Their drainage network has a radial pattern. In no case was it possible to distinguish archaeological structural features, and no potsherds were surface-collected. The accumulation landforms described above are the result of fluvial processes, which depend on the volume of precipitation, the amount of sediments available, and the transport capacity of each stream belonging to the Santa Marı´a river basin. These rivers have a torrential seasonal regime. During the winter the river beds are dry, whereas in the summer season they transport large amounts
45
of sediments to the lower sectors, where the sediments are deposited. These landforms have a strong relationship with past and present climatic conditions. Thus we were able to distinguish fluvial terraces (T1 and T2), secondary fluvial terraces (ST), and the present river bed and flood plain of the Santa Marı´a River, all belonging to the Late Holocene. Fluvial terrace T1 is developed on the right side of the Santa Marı´a River. It is more extensive in the middle and northern sectors of the study area, and in the south its development is more restricted. The next terrace level is fluvial terrace T2. It is younger than T1, and shows the same tendencies as the previous terrace, owing to decreasing river surface runoff and shrinking of the present river bed. Secondary terraces (ST) are located in the apical and middle piedmont, and were developed by the tributary streams of the Santa Marı´a River. They are smaller in the northern and central parts of the study area than in the south. All of them are of elongated form. Finally, we were able to distinguish landforms of aeolian origin, represented by stabilized dunes that are located in the northern section of the study area, among the present-day fluvial fans and fluvial terrace level 1 (T1). They are composed of fine sediments, with some vegetation growth over them, and are currently being incised by intermittent streams. At the bottom of the valley, close to the Santa Marı´a River, there are mobile dunes composed of very fine materials generated by the dominant dry winds. Their shapes are elongated, with a north –south orientation, and they are not visible on the aerial photographs. There is no evidence of archaeological occupation in this sector.
Conclusions Geomorphological analysis allowed us to identify three levels of covered glacis developed over Tertiary deposits. These could correspond to the five levels of pediments proposed by Strecker (1987) through the south of our study area over the Aconquija Ranges, with the difference in numbers being due to the intensive neotectonic processes that affected the southern part of Strecker’s study area. These covered glacis levels are located in different topographic positions, reflecting changes in climatic conditions during their formation. They were formed during the Pleistocene period by currents of high erosive power that eroded existing deposits, generating thick accumulations downstream. During the Early Holocene, alluvial fans were formed by more fluid currents that formed debrisflow deposits. After that, during the Middle and Late Holocene, fluvial fans were formed, in
46
M. M. SAMPIETRO VATTUONE & L. NEDER
several cycles, reflecting periods of torrential water availability. During the Middle and Late Holocene, streams from the Calchaquı´es Summits hillsides formed the secondary terraces of the Santa Marı´a River tributaries, excavating old deposits and forming several cycles of fluvial fans. During the Late Holocene fluvial terraces were developed on the Santa Marı´a River. Finally, dunes were formed then stabilized. Against this background, it is possible to identify archaeological sites with different superficial characteristics, composed of residential and agricultural structures. Residential units appear dispersed on agricultural fields, and have different shapes (quadrangular and circular), allowing us to differentiate two occupational periods, Formative (500 BC–AD 1000) and Regional Development (AD 1000–1400). The only semi-urban settlement identified was that already cited by Rivolta (2005). Archaeological surveys showed that alluvial fans are the only landscape unit with only Formative period structures. The other geomorphological units with archaeological remains (slopes and covered glacis (L2)) had structures from the Formative and Regional Development periods juxtaposed. The construction characteristics of the agricultural terrace walls vary. Agricultural structures from the Formative period, which are associated with circular rooms and early ceramic types, are made with boulders without mortar, whereas the structures from the Regional Development period, associated with rectangular structures and Santa Maria potsherds, are made with flagstones and mud. We relate this differential distribution of archaeological settlements to the general process of aridization that the area was affected by during the Quaternary period. According to the geomorphological evidence, water availability diminished through time. The sequence of landscape unit evolution, together with the sizes of such units, and the decreasing transport energy of the entire system through time support this interpretation. In general terms, the mass movement processes that accumulated the original material of the covered glacis required the action of transport agents of low fluidity and high competence, generating solifluxion processes over the materials crushed during glacial periods in the high mountains. Those materials are characteristic of fluvio-glacial depositional environments (Sayago & Collantes 1991). After that, torrential environments dominated, first forming alluvial fans in a very restricted area and then fluvial fans. Alluvial fans are formed by texturally coarse debris flows, which represent greater transport energy than the later landforms. Finally, fluvial terraces are restricted to present-day rivers, especially in the valley bottom where water flow
is higher, and are of limited extent compared with the previously described landforms. From the archaeological point of view, it is possible to observe a differential distribution of settlements by period. Alluvial fans, formed in the smallest river basins, were occupied only during the Formative period, whereas in the south and north of this area, where river basins are larger, Formative settlements are juxtaposed with Regional Development period structures. This could be related to the differences in water availability that have already been established in the Tafı´ Valley, close to Santa Marı´a Valley, by Sampietro Vattuone (2002). The climatic and geomorphological evidence thus combines to suggest that during Formative times the climate was wetter and occupation of the smaller river basins (as well as the larger ones) was possible. By the Regional Development period, the drier climate caused people to restrict their settlements to the larger river valleys, as only there was there an adequate supply of water. Even today, water availability is better in these areas, reinforcing this argument. Although this is a first approach to the subject in this region, we believe that the results of our work demonstrate that the integrated interpretation of archaeological features and landscape evolution systems is a profitable approach to understanding human occupation evolution through time.
References ANON. 1966. Primera Convencio´n Nacional de Antropologı´a. Facultad de Filosofı´a y Humanidades, Instituto de Antropologı´a, Universidad Nacional de Co´rdoba. Publicaciones, 1. Aschero, C. A. & Ribotta, E. E. 2007. Usos del espacio, tiempo y funebria en El Remate (Los Zazos, Amaicha del Valle, Tucuma´n). In: Arenas, P., Manasse, B. & Noli, E. (eds) Paisajes y Procesos Sociales en Tafı´ del Valle. Magna, Tucuma´n, 79– 94. Birot, P. 1960. Le cycle d’e´rosion sous les diffe´rents climats. Centro de Pesquisas de Geograficas do Brasil, Rio de Janeiro. Cigliano, E. 1961. Noticia sobre una nueva industria precera´mica en el valle de Santa Marı´a (Catamarca): el Ampajanguense. Anales de Arqueologı´a y Etnologı´a. Buenos Aires, 16, 169 –179. Cigliano, E. 1968. Panorama general de las industrias precera´micas en el Noroeste Argentino. In: XXXVII Congreso Internacional de Americanistas, 3. CIA, Buenos Aires, 339–344. Garralla, S. 1999. Ana´lisis polı´nico de una secuencia sedimentaria en el Abra de El Infiernillo, Tucuma´n, Argentina. In: 1st Congreso Argentino de Cuaternario y Geomorfologı´a, Sociedad Argentina de Geomorfologı´a, La Pampa, 11. Gonza´lez, A. R. & Pe´rez, J. A. 1972. Historia: argentina indı´gena, vı´speras de la conquista. Paido´s, Buenos Aires.
LANDSCAPE AND OCCUPATION IN ARGENTINA Gonza´lez, O., Tchilinguiria´n, P., Mon, R. & Barber, E. 2000. Hoja Geolo´gica 2766-II, San Miguel de Tucuma´n. Programa Nacional de Cartas Geolo´gicas de la Repu´blica Argentina, 1:250.000, Boletı´n, 245. Gutie´rrez Elorza, M. 2001. Geomorfologı´a clima´tica. Omega, Barcelona. Hocsman, S., Somonte, C., Babot, M. P., Martel, A. R. & Toselli, A. 2003. Ana´lisis de materiales lı´ticos de un sitio a cielo abierto del a´rea valliserrana del NOA: Campo Blanco (Tucuma´n). Cuadernos Universidad Nacional de Jujuy, Argentina, 20, 325– 350. Neder, L. & Busnelli, J. 2003. Geomorfologı´a de la ladera oriental de la Sierra del Campo (sector austral), Dpto Burruyacu´. Tucuma´n—Argentina. In: II Congreso Argentino de Cuaternario y Geomorfologı´a. Magna, Tucuma´n, 165– 169. Olivera, D. F. 2001. Sociedades pastoriles tempranas: el Formativo Inferior del Noroeste argentino. In: Berbaria´n, E. E. & Nielsen, A. E. (eds) Historia Argentina Prehispa´nica. Brujas, Co´rdoba, 83– 126. Perea, C. 1994. Mapa de vegetacio´n del Valle de Santa Marı´a, sector oriental (Tucuma´n, Argentina). Lilloa, 37, 2. Rivolta, G. 2000. Conformacio´n y articulacio´n espacial en un poblado estrate´gico defensivo: Los Cardones. BSc thesis, Universidad Nacional de Co´rdoba. Rivolta, G. M. 2005. Nuevos avances en las prospecciones arqueolo´gicas en la Quebrada de Los Cardones. Cuadernos Universidad Nacional de Jujuy, Argentina, 29, 81–94. Rivolta, G. M. 2007. Diversidad cronolo´gica y estructural en los diferentes sectores de la Quebrada de Los Cardones: sus espacios y recintos (valle de Yocavil, Tucuma´n). In: Arenas, P., Manasse, B. & Noli, E. (eds) Paisajes y Procesos Sociales en Tafı´ del Valle. Magna, Tucuma´n, 95– 110. Ruiz Huidobro, O. 1972. Descripcio´n geolo´gica de la Hoja 11e, Santa Marı´a (Prov. de Catamarca y Tucuma´n). Boletı´n del Servicio Nacional Minero Geolo´gico, Buenos Aires, 134, 1 –72.
47
Sampietro, M. M. 1992. Una prospeccio´n arqueolo´gica en Ampimpa. Dto. Tafı´ del Valle, Tucuma´n. Revista Pangea, Facultad de Ciencias Naturales e Instituto Miguel Lillo, Tucuma´n, 1, 3– 4. Sampietro Vattuone, M. M. 2002. Contribucio´n al conocimiento geoarqueolo´gico del valle de Tafı´, Tucuma´n (Argentina). PhD thesis, Universidad Nacional de Tucuma´n. Sayago, J. M. & Collantes, M. M. 1991. Evolucio´n paleogeomorfolo´gica del valle de Tafı´ (Tucuma´n, Argentina) durante el Cuaternario Superior. Bamberger Geographische Schriften, Bamberg, 11, 109– 124. Sayago, J. M., Collantes, M. M. & Toledo, M. A. 1998. Geomorfologı´a. In: Gianfrancisco, M., Puchulu, M. E., Durango de Cabrera, J. & Acen˜olaza, G. F. (eds) Geologı´a de Tucuma´n. Cole´gio de Ge´ologos, Tucuma´n, 241–258. Somonte, C. 2007. Espacios persistentes y produccio´n lı´tica en Amaicha del Valle, Tucuma´n. In: Arenas, P., Manasse, B. & Noli, E. (eds) Paisajes y Procesos Sociales en Tafı´ del Valle. Magna, Tucuma´n, 47–78. Sosa, J. 1996–97. Teledeteccio´n arqueolo´gica en Amaicha del Valle (Tucuma´n): la ocupacio´n Formativa. Cuadernos Instituto Nacional de Antropologı´a y Pensamiento Latinoamericano, 17, 275–292. Sosa, J. 1999. Teleprospeccio´n arqueolo´gica en Amaicha del Valle (Dpto. Tafı´ del Valle, Tucuma´n). In: Diez Marı´n, C. (ed.) XII Congreso Nacional de Arqueologı´a Argentina, La Plata, 3. Universidad Nacional de La Plata, 358 –365. Strecker, M. R. 1987. Late Cenozoic landscape in Santa Marı´a valley, Northwestern Argentina. PhD thesis, Cornell University, Ithaca, NY. Van Zuidam, R. 1976. Geomorphological development on the Zaragoza region, Spain: processes and landforms related to climatic changes in a large Mediterranean river basin. International Institute for Aerial Survey and Earth Sciences (ITC), Enschede.
Geoarchaeology and the value of multidisciplinary palaeoenvironmental approaches: a case study from the Tehran Plain, Iran G. K. GILLMORE1*, T. STEVENS2, J. P. BUYLAERT3, R. A. E. CONINGHAM4, C. BATT5, H. FAZELI6, R. YOUNG7 & M. MAGHSOUDI8 1
Centre for Earth and Environmental Sciences Research (CEESR), Kingston University, Penrhyn Road, Kingston-upon-Thames KT1 2EE, UK
2
Department of Geography, Royal Holloway, University of London, Egham TW20 0EX, UK 3
Nordic Laboratory for Luminescence Dating, Department of Earth Sciences, Aarhus University, Risø DTU, DK-4000 Roskilde, Denmark
4
Department of Archaeology, University of Durham, South Road, Durham DH1 3LE, UK 5
Archaeological Geographical and Environmental Sciences, University of Bradford, Bradford BD7 1DP, UK 6
Iranian Cultural Heritage, Tourism and Handicrafts Organisation, Tehran, Iran
7
Department of Archaeology and Ancient History, University of Leicester, University Road, Leicester LE1 7RH, UK 8
Faculty of Geography, University of Tehran, Tehran, Iran
*Corresponding author (e-mail:
[email protected]) Abstract: Tepe Pardis, a significant Neolithic–Chalcolithic site on the Tehran Plain in Iran, is, like many sites in the area, under threat from development. The site contains detailed evidence of (1) the Neolithic–Chalcolithic transition, (2) an Iron Age cemetery and (3) how the inhabitants adapted to an unstable fan environment through resource exploitation (of clay deposits for relatively large-scale ceramic production by c. 5000 BC, and importantly, possible cutting of artificial water channels). Given this significance, models have been produced to better understand settlement distribution and change in the region. However, these models must be tied into a greater understanding of the impact of the geosphere on human development over this period. Forming part of a larger project focusing on the transformation of simple, egalitarian Neolithic communities into more hierarchical Chalcolithic ones, the site has become the focus of a multidisciplinary project to address this issue. Through the combined use of sedimentary and limited pollen analysis, radiocarbon and optically stimulated luminescence dating (the application of the last still rare in Iran), a greater understanding of the impact of alluvial fan development on human settlement through alluviation and the development of river channel sequences is possible. Notably, the findings presented here suggest that artificial irrigation was occurring at the site as early as 6.7+0.4 ka (4300–5100 BC).
Although archaeological exploration has shown that irrigation was utilized during the Ubaid period in Mesopotamia (c. 5900–4200 BC), far less is known about the early manipulation of water than in later historical periods. Rescue excavations at Tepe Pardis, a tell site on the Tehran Plain, Iran (excavated in 2003–2007; see Coningham et al. 2004; Fazeli et al. 2007; Fig. 1) highlighted a small channel-like feature in an exposed section of a brick quarry cutting through the site. The age
and origin of this feature are significant in understanding the use of irrigation in the early Chalcolithic in a region adjacent to Mesopotamia (but separated from it by hundreds of kilometres and significant orographic barriers). Furthermore, the site shows evidence of occupation spanning the Neolithic–Chalcolithic transition. This crucial transition to metallurgy again originated in the Fertile Crescent but it is not clear how quickly this technology found its way to
From: Wilson, L. (ed.) Human Interactions with the Geosphere: The Geoarchaeological Perspective. Geological Society, London, Special Publications, 352, 49– 67. DOI: 10.1144/SP352.5 0305-8719/11/$15.00 # The Geological Society of London 2011.
50
G. K. GILLMORE ET AL. ARM
TURKEY
ENI
AZERBAIJAN A
Land above 2,000 metres
Lake Urmia
Al
bo
Tehran
rz Mountains Cheshmeh-Ali
Hamadan
N
TURKMENISTAN
Caspian Sea
Tabriz
Mashhad
Tepe Pardis
Z a
IRAQ
g ro s
IRAN
AFGHANISTAN
M
Esfahan
o
u n ta
in s
KU W AI T
PA K
SAUDI ARABIA
0
IST AN
Persian Gulf
400 km
Fig. 1. Location map of the Tepe Pardis site in Iran, and associated sites.
rather distant locations such as those found on the Tehran Plain to the east, peripheral to the main locus of cultural development in the region. In addition, the transition to the Chalcolithic in this area was marked by shifts away from egalitarian to more hierarchical political structures. The associated changes in resource exploitation are evident at the site, with shifts to large-scale ceramic production at c. 5000 BC and, as highlighted above, the potential use of irrigation through cutting water channels. Although it is not clear whether site occupation continued relatively uninterrupted from this period, an Iron Age cemetery has also been located at the site (Fazeli et al. 2007). As outlined below, this site is located on a highly unstable alluvial fan environment and local populations would have had to readily adapt to these ever-changing conditions. Thus the site contains evidence for human –geosphere interaction over a variety of time scales and spanning some key transitions in the development of more modern hierarchical political structures, irrigation and metallurgy in a region peripheral to but potentially closely tied to Mesopotamia. As such, this study employs a variety of sedimentological and palaeoenvironmental techniques to better understand human development at the site, and to tie this into the geological processes operating over the time scales
relevant to the occupation. A detailed and independent time scale for the site is developed using radiometric radiocarbon and optically stimulated luminescence (OSL) dating.
Environmental setting of the Tepe Pardis site Tepe Pardis (358270 N, 518360 E, c. 990 m above sea level) is located on the Tehran Plain (Fig. 2); a piedmont basin adjacent to the Alborz Mountains that experiences dry continental desert type climate and a winter precipitation maximum. Precipitation in the basin is less than a few hundred mm a21, therefore most water at the site will have come from rivers draining the highlands to the north and west where maximum discharge is associated with spring snow melt (Beaumont 1972). These conditions would have ensured that the use of this water source through irrigation was extremely valuable in agriculture. The site is located within the lower stretches of an alluvial fan fed by the Jajrud, a river that has a catchment area of more than 2500 km2 (Beaumont 1972). This catchment extends from the southern margins of the Alborz Mountains, through the densely settled Tehran Plain, to salt desert or kavir
GEOARCHAEOLOGY AND MULTIDISCIPLINARITY
51
Alborz Mountains
N Pakdasht Tepe Pardis Qarchak
Jajr
ud
Varamin
Pishva Pishva Hills 0
5 km
Fig. 2. Aerial photograph of the Tepe Pardis site and the Jajrud alluvial fan with associated river channels indicated.
(Beaumont 1972). Associated with the closure of the Tethys Ocean, the Alborz Mountains to the north and east of the Tehran Plain are undergoing active uplift and denudation. This ensures a very abundant supply of gravels and sands to the alluvial fans of the rivers draining these mountains, creating a highly unstable geomorphological environment where river channels are in constant flux and episodes of sedimentation and erosion are highly variable through space and time. The exact timing of alluvial fan sedimentation (alluviation) in the area is not completely clear. Sedimentological and geomorphological evidence such as gullying and desert varnishing suggests that over at least the last 750 years the alluvial fans have been relatively stable. This prompted Beaumont (1972) to suggest that the entire Holocene may have been a time of erosion rather than deposition and that cold periods in the Pleistocene may have provided optimal conditions for fan formation. It is certainly clear that these sands and gravels have been accumulating at least since the early Pleistocene (Abbassi & Farbod 2009), and that debris-flow and flash flood discharge events may be typical transport events (Beaumont 1972). However, uncertain dating and the probability that the last 750 years does not represent the majority of the Holocene make it unclear just how dynamic this landscape was during human occupation. Certainly, the fans around Tehran appear to contain Holocene-age deposits (Abbassi & Farbod 2009). Understanding the human element is an important
contribution to understanding this landscape, and the interactions between sedimentation, climate and humans are as yet poorly understood. A rescue excavation was conducted at the site and consisted of a step trench excavated down the north face of the tell, which revealed a 10.5 m deep sequence (Figs 3 & 4, Table 1), dated as late Neolithic to Chalcolithic, based on Buff ware Neolithic pottery and radiocarbon dates (Coningham et al. 2006; Fazeli et al. 2007; Gillmore et al. 2009). The digging of this trench was aided by the fact that the tell site was on the edge of an active quarry, the clay being excavated for brick making. Indeed, the presence of the quarry was the initial reason for the archaeological excavation: to retrieve information before it had been lost to quarry activity. The quarry excavation provided a 3.5 m deep sequence of cultural features and old land surfaces around a significant part of the tell that could be utilized for sedimentological analysis or logging purposes (Gillmore et al. 2007). Later, horizontal trenches were extended on either side from the step trench to expose structures dating to the Transitional Chalcolithic associated with ceramic manufacture production. The hypothesized irrigation channel noted above was stratigraphically linked to Late Neolithic levels in the tell (c. sixth to seventh millennium BC). Channel morphology was triangular in profile (Figs 5 & 6), possessed a very different fill from the surrounding sediments, and ran at right angles to a number of other apparently natural channels
52
G. K. GILLMORE ET AL.
2
3
Selected radiocarbon dates
36 Unexcavated brick foundation
7
Log from site GSF (see figure 6 for detailed log) Graphic lithology Clay Silt Sand f m c Clays & silts Sand, silt bands Clay
995.38
11 12 19 18 15=16 25 30 1002 1003 1005 1007 Trench II
8
4
50 BC 120 AD
4
Trench I
2
3
36 Unexcavated brick foundation
4050-3920 BC (Context 5) 3960-3770 BC (Context 8)
5 6
Trench I
5 8 7
Trench II
2
6
9 9
10
4830-4680 BC
10
11 12
13 14 17
14
4900-4780 BC 4920-4800 BC
31=22 1001
17 19
18 33 32
4950-4820 BC 1008
28
25 20
24=29 35
1000 1001 1003
1002
1006 1013 1011 1014 1015 1016 1017
100.000 MSL
13
5000-4860 BC 1013
5280-5050 BC 5220-4990 BC (top of irrigation channel) 5100-4300 BC OSL date 5310-5050 BC
1008 1010 1018 1011 1015 1016
1004 997.556 1005 1006 1007 1009 1012 1014
1017
Fig. 3. North–south archaeostratigraphical section and northern elevation of Tepe Pardis, with a simplified lithological log and position of the irrigation channel and its relative elevation (see Fig. 6 for a detailed log). The archaeological contexts are shown in Table 1. Selected radiocarbon dates and OSL dates are also highlighted.
in the sequence. This feature was compared with these natural channels, which have a very different gross morphology (Fig. 7), and which were logged and sampled for granulometric analysis, pollen and radiocarbon dating. OSL samples were taken from the channel sands (see Fig. 8), including the potential irrigation channel, which was also dated via radiocarbon and pottery remains. The river sediments usually had no associated charcoal, hence
Fig. 4. The step trench down the northern face of the tell.
the focus on OSL to try and ascertain temporal relationships. In 2006, whilst preparing for the second season of excavation at Tepe Pardis, grey ware ceramic sherds were excavated to the SW of the tell, within the surrounding quarry (Fazeli et al. 2007). These sherds were attributed to the Iron Age, and were found 1.5 m below the modern surface. In 2007 a rescue excavation recorded human remains
GEOARCHAEOLOGY AND MULTIDISCIPLINARITY
53
Table 1. Summary of archaeological contexts illustrated in Figure 3 Archaeological context 2 3 4 5 6 7 8 9 10 11 12 13 14 15 ¼ 16 17 18 19 20 24 ¼ 29 ¼ 34 25 28 30 31 ¼ 32 33 ¼ 35 36 1000 1001 1002 ¼ 1004 1003 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017
Comment Loose clay mixed with fragments of modern bone, plastic, and ancient and modern ceramics Weathered material. Contaminated; contained fragments of bone, charcoal, burnt clay patches, ceramics and plastic. Between 0 and 0.34 m thick Unconsolidated unbaked mudbrick, presumably loosened from mudbrick foundation. Presence of bone, plastic, ash and ceramics. Mixed deposit. 0– 0.93 m thick Clay old land surface containing ‘in situ’ charcoal, ceramics, bones and some fragments of unbaked mudbrick. Contained chert blades and sherds of Early Chalcolithic pottery. 0.18 –0.28 m thick Old land surface dipping to the north. Ashy wash. Large fragments of bone, ceramic sherds, charcoal flakes, burnt clay and flakes and blades. 0.10 – 0.40 m thick Dipping towards the edge of the mound in a northerly direction. Significant amounts of charcoal, one flake, 12 blades. 0.13 –0.20 m thick Earlier phase of occupation. Unbaked mudbrick within a clay matrix Northerly dipping clay. 0.13– 0.40 m thick Northerly dipping layer of clay and burnt clay fragments. One flake, three blades, and the first Transitional Chalcolithic ceramics. 0.22– 0.73 m thick Uniformly thick horizon, clay, unbaked mudbrick, stone fragments. Perhaps representing a paved area. Two flakes and seven blades. 0.10 m thick Horizontal context. Clay containing larger fragments of charcoal. Four blades. 0.10 –0.16 m thick Horizontal layer. Compact clay and ash. Single piece of debitage. 0.22 m– 0.40 m thick Gently dipping old land surface, mixed clay, ash, charcoal, ceramic sherds. Four flakes, six blades. 0.10 –0.44 m thick Mixed fill (kiln waste plus other material) of burnt clay, ash, ceramic sherds and charcoal. 18 flakes, beads and part of a terracotta figurine. Is on top of old land surface context 30 Mixed layer of ceramics, lithic fragments, burnt clay, unbaked mudbrick fragments and ash. 0.20 –0.27 m thick Ash containing very large ceramic sherds and charcoal. Waste from kiln? 0.60 m thick A thin layer of burnt clay with kiln waste within depression 20 Shallow depression. Filled with clay Horizontal old land surface, 1.20 m thick. Linked Trench I to the sequence of Trench II. Same context as 1001 Mudbrick wall. 3.94 m long, 0.30 –0.50 m wide, and to a height of five bricks. Part of a mudbrick kiln Current land surface Old land surface Context 32, a 0.04 m thick layer of ash and charcoal Old land surface Unexcavated brick foundation Exposed ‘in situ’ archaeology, contaminated with modern domestic rubbish Old land surface Thin layer of light grey clay dipping to the north. 0 – 0.16 m thick Old land surface. Dipping to the north. Charcoal, 25 blades, one terracotta spindle whorl Dipping clay, ash, charcoal, single flake. 0 – 0.06 m thick. Ash dump? Old land surface. 0.14– 0.28 m thick White clay, ash, charcoal, single bead, 0– 0.14 m thick Undulating old land surface. Four lithic fragments, a stone block with a smoothed central depression. 0.06– 0.48 m thick Ash deposit. Contained charcoal, ash, grit, three flakes Old land surface, 0.26 m thick clay. 13 flakes, single bead Old land surface devoid of special finds. 0.10 – 0.61 m thick Ashy dump, four lithic fragments. 0 –0.06 m thick Unbaked mudbrick wall. Greater than 0.60 m high. Over 1 m thick. Varied between three and four courses Old land surface, single flake. Northward dipping, maximum 0.36 m thick Horizontal clay, no special finds. 0.22– 0.29 m thick Horizontal old land surface. Fragments of bone, charcoal, ceramics. 0.36 –0.38 m thick Clay containing charcoal, two lithic fragments and a Late Neolithic sherd above natural clay. 0.34 m thick
54
G. K. GILLMORE ET AL.
5220–4990 BC 6.7 ± 0.4 ka (5100–4300 BC) 5220–5030 BC
Fig. 5. The sand-filled irrigation channel at log section GS-F (OSL sample GG4). Dates for radiocarbon results in black and OSL in white.
and associated vessels. These remains consisted of single and double flexed burials oriented east – west with grave goods, examples being grey ware tripod dishes and a red ware sieve (see Fazeli et al. 2007, who described eight graves). Bronze objects were also noted in the graves. The skeletons and grave goods had been compressed and damaged by quarry machinery, and the nature of the surrounding material made it difficult to identify burial cuts, or stratigraphic relationships between the burials. OSL dating was undertaken to see how the dates obtained might (or might not) match with the pottery dates and hence what the chronostratigraphic relationships are between overlying sediment at the Iron Age cemetery.
Methods Sedimentary analysis Sediment characteristics and sedimentary structures, such as bedding, lamination and ripples, were recorded in the field using standard approaches, including graphic logging (after Tucker 2001, 2003) with 10 detailed graphic logs being taken within the Tepe Pardis quarry site. The position of each logged section was located using global positioning system (GPS) (see Fig. 9). In addition to the field observations, sediment texture was assessed in the laboratory. This involved an analysis of grain size and grain-size parameters, grain morphology, grain surface texture and
sediment fabric (Tucker 2001, 2003). These textural characteristics are referred to as textural maturity and the texture of a sediment is largely a reflection of depositional processes (Tucker 2001), which are greatly influenced by climate. When deciding which techniques to use, comparisons were made in the literature between laser particle-size analysis, dry sieving and the sieve –pipette method. A laser particle analyser can provide a very accurate grain measurement of particle sizes; however, a number of researchers (e.g. Beuselinck et al. 1998) have suggested that laser diffractometry may not provide accurate results for fine platy materials. Because much of the material to be examined in this survey was fine-grained, it was decided to apply the classic dry-sieving method for sand-sized materials, with the sieve – pipette method (using settlement columns and a constant temperature water bath) for finer materials. The methods followed for dry sieving have been described by Friedman & Johnson (1982), Prothero & Schwab (1996) and Jones et al. (2000). The samples were split to produce representative 50 –100 g subsamples. They were washed to remove soluble salts and oven dried (at 65 8C). Samples were then weighed and the mass was recorded. Standard dry-sieving methods were applied, as were measures of central tendency and dispersion (see Folk 1966; Walford 1995). GRADISTAT 5.0 was utilized to plot these (e.g. skewness, kurtosis; see Blott & Pye 2001). The results were then plotted as frequency histograms and on
GEOARCHAEOLOGY AND MULTIDISCIPLINARITY OPERATOR
GKG
LOCALITY &/OR GRID REF
DATE
29/07/04
FORMATION/MEMBER
SHEET NUMBER BED NO
TEPE PARDIS SECTION GS F
WEATHER
55 BRIGHT, SUNNY
1 of 2
GRAPHIC LITHOLOGY
Sedimentary Structures (Graphic Symbols)
GRAIN SIZE
COLOUR
FOSSILS
ADDITIONAL NOTES
FACIES No
SAND
Scale
cl si
f
m c
g
No laminations
7.5 YR 7/3
7
No laminations
7.5 YR 7/2 - 7/3
6
No laminations
5 YR 6/3
No obvious laminations
7.5 YR 7/3
8
p
Bed 6 contains pink patches with ash/charcoal Bed undulates
5
Fine clay
p Tooth Lighter silt horizon
Clay cast at the top of Bed 4 is 5 YR 7/4
7.5 YR 7/2
Irrigation channel
10 YR 6/3
4
No laminations except at the top of Bed 3
3
2
Bed 2 Top of Bed 1
c
feature ?
Cattle? bone
Channel has a smooth base (water-worn?)
7.5 YR 7/3 - 7/2 7.5 YR 6/4
2.5 YR 8/1
c 1
Structureless Base not seen Key Clay
c
Charcaol fragments
Silt
Clay flakes
Sand
Rip-up casts
p Ceramic sherd
Occasional pebbles
Parallel laminations
Undulating boundary
Thin charcoal bands
Bone/teeth fragments
Fig. 6. Detailed lithological log of section GS-F, including the irrigation channel.
cumulative frequency curves (see Gillmore et al. 2009, e.g. plots and further details).
Pollen analysis Five samples of sediment have been analysed from the Tepe Pardis site and one from a comparative section on the banks of the Karaj River. The sediments mostly comprise calcareous clay, silt and sands. Only a few samples were collected because the primary objective of the exercise was to ascertain if the samples contained microscopic pollen grains of suitable quantity and quality for further palynological assessments.
The subsamples were tipped into five 125 ml Pyrex beakers and macerated in 3 mol HCl to break down the carbonates in the sediment. Forty ml of 10% potassium hydroxide (KOH) and 5 mg of sodium pyrophosphate (Na4P2O7) were added to each sample. The containers were placed on a sand bath and heated for 20 min. The contents of each beaker were poured through a 140 mm sieve onto a Perspex ‘swirling’ dish. Mains supply water filtered through a Whatman polycap HD disposable filter capsule was used to swill the macrofossils contained within the sieve, with the liquid running onto the swirling dish. The macro remains were collected for microscopic
56
G. K. GILLMORE ET AL.
Fig. 7. Natural channels, log section GS-K (OSL sample GG9).
examination. The liquid contained within the swirling dish was then agitated gently until a ring of sediment had formed on the base of the dish. The swirling dish was then tipped into a 6 mm sieve, leaving the majority of the sediment on the base of the swirling dish to be run to waste. The 6 mm sieve was then filled with distilled mains supply water and rinsed to remove all trace of the alkalines (KOH and Na4P2O7). Following the final rinse two drops of Safranin 0 solution 1% was added to the solution in the sieve (c. 2 ml) to stain the pollen grains. The stained solution was then pipetted into phials for storage prior to analysis. Six microscope slides were prepared and four transects from each slide were examined using a Leica Galen III microscope.
Radiocarbon and OSL dating of sequences OSL samples were processed for luminescence dating in the Nordic Laboratory, Roskilde, Denmark. Full details of analytical procedures will form the basis of another publication and so will not be covered in great detail here. However, the essential methodological information is presented below. OSL samples were collected in c. 25 cm long metal tubes hammered into the cleared and cleaned section face. Samples were wrapped in light-tight black plastic bags immediately after being removed from the sections and were processed under subdued red light. Sunlight-exposed ends of tube samples were excluded from equivalent dose (De) determinations but retained for radioisotope measurements
using laboratory gamma spectrometry (to determine dose rate). Water contents were calculated from subsamples taken in airtight pots after oven drying and reweighing. Carbonates and organic matter were removed from the De (unexposed) fractions using 0.1M HCl and 15% H2O2. Medium- to finegrained sand (150–212 mm) was isolated by wet sieving. Quartz was isolated by immersion in concentrated HF for 1–2 h, with a subsequent 0.1M HCl wash to remove fluorite precipitates. Equivalent dose (De) values were measured using the single-aliquot regenerative-dose (SAR) procedure (Murray & Wintle 2000) performed on a Risø TL-DA TL/OSL reader (Fig. 10). A blue LED (l ¼ 470 + 20 nm) stimulation source was used on samples and the OSL signal was measured using a 9235QA photomultiplier tube filtered by 6 mm of Hoya U340 (Bøtter-Jensen et al. 2000). All growth curves were fitted using a saturating exponential plus linear, a saturating exponential, or a linear function. The signal was generally integrated from the first 0.6 s of stimulation minus a background estimated from the last 6 s of stimulation, or the closest possible alternative, to isolate the fast component as best as possible. Aliquots yielding recycling ratios (Murray & Wintle 2000) or IR ratios (Duller 2003) differing from unity by more than 10% were rejected. Recuperation was normally negligible and always below 2%. The uncertainty (one standard error) on single De values was estimated using Monte Carlo simulation and a weighted mean De was calculated for each sample (typically .12 aliquots). Beta attenuation
GEOARCHAEOLOGY AND MULTIDISCIPLINARITY
57
Fig. 8. OSL core sample collection from sandy sediments (OSL sample GG1).
was obtained using the calculations of Mejdahl (1979) and dose rate conversion factors were taken from Adamiec & Aitken (1998). Uncertainties are based on the propagation, in quadrature, of individual errors for all measured quantities, which if unknown are taken as 10%. In addition to uncertainties calculated from counting statistics, errors owing to beta source calibration (3%) (Armitage & Bailey 2005), radioisotope concentration (3%), dose rate conversion factors (3%) and attenuation factors (3%) (Murray & Olley 2002) were included. Dose rates were calculated from sample activity measured during laboratory gamma spectrometry (Murray et al. 1987) and calculated cosmic dose (Prescott & Hutton 1994). Both small (2 mm mask) and standard aliquots were analysed for many samples, and varying preheats were tested before assignment of the most appropriate (usually a combination of 260 and 220 8C). Dose recovery
tests were also applied to samples and analyses of De with signal integration time was performed. Based on this and small aliquot analysis, samples showed no evidence of partial bleaching and recovered laboratory doses to within 10%. Dates and dosimetry are given in Table 2 and an example of a typical decay curve and growth curve is shown in Figure 11. The preheats and type of aliquot used are also shown. As pointed out by Tucker (2001), fossils are uncommon in fluvial sediments and are mostly of plant material or skeletal fragments of freshwater and terrestrial animals. However, some macrofossils were identified and sampled for radiocarbon dating. Charcoal and animal bone fragments were collected near the channel feature and from layers stratigraphically above and below this channel, to place it within a date range. Fifteen accelerating mass spectrometry (AMS) radiocarbon age
58
G. K. GILLMORE ET AL.
Sample point GS-K Log section Natural channel Artificial channel
N
Brick quarry
GG9 GS-K ? ?
GS-J GS-I GS-H GS-G
Bronze Age burial site
?
GG4 GS-A GS-F GS-A1 GS-E GS-B ? GS-B1 GS-D GS-B2 GS-C
Tepe Pardis
Road
0
100 m
Fig. 9. Site map of log sections, GS-A to GS-K, plotted using GPS. The irrigation channel was recorded at GS-F whereas small natural channels discussed in the text appear at GS-K. The GG numbers refer to OSL dated material. The radiocarbon and OSL approaches were both applied at GS-F (see Tables 2 & 3), although OSL was mostly focused where no material was available for radiocarbon dating. The directions of both the natural and irrigation channels (shaded areas) are postulated from limited exposure data.
determinations (carried out by the Oxford Research Laboratory for Archaeology and the History of Art) were used to provide a radiocarbon framework for the site (see Coningham et al. 2006) and place, for example, the ceramic typology into an absolute dating sequence. Calibration of the dates was carried out using OxCal V3.10 (Bronk Ramsey 1995), based on the internationally agreed calibration curve of Reimer et al. (2004).
Results Field observations The results from field observations have been presented by Gillmore et al. (2009). At Log Section
GS-F (see Figs 6 & 9), on the north face of the tell, the following sequence was observed. Unit 1. Youngest. Clay of 1.35 m thickness with no obvious laminations, with Chalcolithic pottery and charcoal. Thin silty horizons within clay, the latter having no laminations. Occasional small pebbles or granules and bone. Unit 2. Maximum thickness 0.24 m. Triangularshaped channel feature with smoothed base (water worn?). Infill deposits parallel laminated. Gravel (rare) to sand and silt with some clay. At least three fining upwards sequences with rip-up clasts at base of beds, with clay flakes and lighter silt horizons. Unit 3. Oldest. Thickness 0.32 m. Clays (mostly structureless and base not seen) with charcoal
GEOARCHAEOLOGY AND MULTIDISCIPLINARITY
Fig. 10. Modified SAR protocol used in study.
fragments and bands. Parallel laminations in thin beds near the base of the overlying channel. To summarize, the sediments observed in the quarry were a mixture of mostly fine silts or clays (colour often 7.5 YR 7/3), many appearing to be structureless. However, occasionally clear ripple structures (the preserved ripple tops could be seen with small-scale cross-laminations) were noted. Associated with these fine sediments were coarser sandy filled structures, occurring either as horizontal beds, or mostly as cross-sections through various scoop-shaped channels. They varied considerably in size, from a few metres across, to one to the north of the tell, but still within the quarry, that was around 25 m across. However, the latter represented a migrating channel (with epsilon crossbedding showing point bar development), which at times was braided, with gravel-rich layers. Alongside what were clearly natural river channel sediments and structures, the small (c. 0.2 m deep and 2 m wide) triangular (in cross-section) channel was noted, with a different infill, which consisted of rip-up clasts or clay flakes, sands and silts, with fining upwards sequences.
Laboratory analysis Samples taken from the irrigation channel (GPS site GS-F; see Fig. 9 for location of log sections) suggested a unimodal poorly sorted sediment (after the method of Folk & Ward 1957, logarithmic mean of 1.043 f), with a coarse skew (20.231 f) and a very leptokurtic distribution (1.878 f; in
59
other words, a high peak, a normal curve having a value of 1.0 according to King (1975)), with a dominant grain size (mean) of fine sand (2.193 f). Poor sorting (or the spread of the grain sizes around the average) of these sediments (GS-F) suggests that they may have been deposited quickly. The skew towards coarser values is perhaps unusual for a river deposit because much fine material is usually trapped between larger grains (King 1975). The particular characteristics of sample GS-F suggest flood events with coarse material being generally deposited and finer material being carried elsewhere before deposition (or with some winnowing by wind action). This sample belongs to the textural group ‘Slightly Gravelly Sand’ according to Blott & Pye (2001) after Folk (1954), being 1% fine gravel, 97.6% sand (mostly fine) and 1.4% ‘mud’. Some samples from Tepe Pardis showed a bimodal distribution with cumulative frequency plots showing two inflection points. Sample GS-K from a river channel (see Fig. 9) was bimodal, poorly sorted, with a mean of 0.942 f (coarse sand), sorting of 1.303 (poorly sorted), a symmetrical distribution (0.088) and platykurtic (having a flat-topped peak). The cumulative frequency curve is steep compared with that of GS-F. The existence of bimodal sediment implies modes of behaviour in entrainment and transport that are distinctive from those of unimodal sediments (see Wilcock 1993; Evans & Benn 2004). Very high values of kurtosis suggest sorting in a region of high energy, then transportation without a change in the character of that material to another environment, where it is mixed with another sediment, possibly in a low-energy environment (King 1975), producing a bimodal distribution.
Pollen The assessment of the sediment from Tepe Pardis quarry indicates that pollen (and other useful microfossils) has been preserved in the sediment from the excavated site, although its occurrence is very sparse. As a result of the harsh background matrix, specifically clastic sediments that are lime-based and subject to regular episodes of flooding and subsequent drying, it is not surprising that pollen analysis has yielded limited results. Although findings are preliminary, the local environment at Tepe Pardis was in part slow flowing or standing water, as indicated by the presence of dinoflagellate cysts, algal cysts and diatoms. The presence of grasses within the burnt plant material is notable, as is that of vesicular arbuscular mycorrhizal and soil fungi, both of which indicate soil erosion. Although very preliminary, the results that have been obtained to date offer a valuable adjunct to our other environmental data, and certainly warrant
60
Table 2. Optical dating results Depth (m)
U (ppm)
Th (ppm)
K (%)
Cosmic (Gy ka21)
Preheats (8C)/aliquot
Dose rate (Gy ka21)
De (Gy)
n
Age (ka)
GG1 GG3 GG4 GG5 GG6 GG7 GG8 GG9 GG10 GG11
2.25 0.85 3.50 1.75 2.25 2.00 1.10 2.40 3.00 3.00
2.52 + 0.08 2.12 + 0.07 1.84 + 0.06 2.48 + 0.08 2.14 + 0.07 2.16 + 0.07 2.21 + 0.07 2.00 + 0.07 2.33 + 0.08 2.25 + 0.09
8.37 + 0.30 7.01 + 0.27 6.70 + 0.24 8.90 + 0.29 7.80 + 0.31 7.21 + 0.28 7.82 + 0.26 6.99 + 0.28 8.38 + 0.31 8.24 + 0.49
2.02 + 0.08 1.77 + 0.07 1.89 + 0.07 2.17 + 0.08 2.04 + 0.08 1.83 + 0.07 2.01 + 0.07 1.91 + 0.08 1.89 + 0.08 2.08 + 0.09
0.17 + 0.01 0.20 + 0.01 0.16 + 0.01 0.18 + 0.01 0.17 + 0.01 0.17 + 0.01 0.20 + 0.01 0.16 + 0.01 0.15 + 0.01 0.15 + 0.01
260– 220/SA 260– 220/SA 260– 220/LA 240– 200/LA 260– 220/LA 260– 220/LA 240– 200/LA 260– 220/SA 260– 220/LA 260– 220/LA
3.05 + 0.12 2.69 + 0.10 2.67 + 0.11 3.22 + 0.12 2.95 + 0.12 2.73 + 0.10 2.97 + 0.11 2.75 + 0.11 2.87 + 0.11 3.02 + 0.13
18.42 + 0.74 4.91 + 0.40 17.87 + 0.65 3.76 + 0.17 18.69 + 1.26 8.06 + 0.41 3.94 + 0.43 17.85 + 0.62 8.21 + 0.27 8.23 + 0.23
37 16 16 11 17 22 13 17 12 17
6.05 + 0.34 1.82 + 0.17 6.68 + 0.37 1.17 + 0.07 6.34 + 0.50 2.96 + 0.17 1.33 + 0.06 6.50 + 0.35 2.86 + 0.15 2.73 + 0.14
Ages and radioisotope concentrations are quoted to one standard deviation (or 10% if dispersion data are unknown), cosmic dose with 5% error and De to one standard error. n, number of aliquots. SA, small aliquot, 2 mm mask. LA, large aliquot, 10 mm mask.
G. K. GILLMORE ET AL.
Sample
GEOARCHAEOLOGY AND MULTIDISCIPLINARITY
Fig. 11. A typical OSL decay curve and growth curve. Lx/Tx is luminescence intensity normalized to response from test dose.
further investigation and the collection of larger samples during future investigations.
Geochronology The pottery sherds from this site have been assigned to Late Neolithic, Transitional Chalcolithic (c. 5300–4300 BC), Early Chalcolithic (c. 4300– 4000 BC) and Middle Chalcolithic (c. 4000–3700 BC) (see Fazeli et al. 2004) and this has formed the basis of the site’s chronology. However, to obtain a more highly resolved and independently calibrated age model, further radiometric dating has been necessary. This has provided some direct dating of the geomorphological –sedimentological features of the site in places that provide an insight into the role of the environment in determining activity at the site and vice versa. To this end, both radiocarbon and optically stimulated luminescence (OSL) dating have been employed. The 15 radiocarbon measurements were carried out on charcoal and bone fragments and are presented in Table 3. In addition to the radiocarbon determinations, there were stratigraphic records from which the relationships between the contexts and their assorted radiocarbon samples could be determined. To utilize the radiocarbon age determinations in calendar years, use was made of the calibration and analysis program OxCal (Bronk Ramsey 1995). The resulting percentages are an index of how well the chronological model agrees with the
61
dating evidence; in some cases the agreement is better than expected and is greater than 100%, in other cases it is poorer. We are able, therefore, to suggest an end of the Late Neolithic and a beginning of the Transitional Chalcolithic at c. 5300 BC, and an end of the Transitional Chalcolithic and a beginning of the Early Chalcolithic at c. 4600 BC at Tepe Pardis. Unfortunately, as our sequence does not extend before the Late Neolithic or after the Early Chalcolithic we are unable to propose further boundaries based upon our data from Cheshmeh-Ali, which is located in the south of Tehran. The site was first excavated by Schmidt during 1933 and later excavated in 1997 (see Fazeli 2001). Optically stimulated luminescence dating has only rarely been applied in Iran (e.g. Fattahi et al. 2006) and seldom in an archaeological context. The technique is powerful in that it directly dates the last exposure of sediment to sunlight, hence it is a depositional chronometer. As with previous work in northern Iran (Fattahi et al. 2006), the luminescence characteristics of the sediments analysed appear not to be optimum, with relatively weak signals (Fig. 11) and in some instances, an apparently weak fast component. A full discussion of these issues is beyond the scope of this paper but it is important to note that reproducible results could be obtained for all but one sample (Table 2). All samples passed internal tests such as dose recovery, feldspar contamination, reproducible growth curves and recuperation tests. The major problems encountered were relatively large inter-aliquot variability (over-dispersion) and in some aliquots under certain low-temperature preheat combinations, decreasing equivalent dose with increased size of initial integral used for signal integration. The former problem of wide equivalent dose frequency distributions leads to large errors on the final OSL date and can be indicative of partial exposure of the sediments to light (partial bleaching; whereby the luminescence level prior to last exposure is not fully removed by subsequent optical bleaching). However, this effect should become more pronounced when aliquot size (and hence the number of grains per aliquot) is reduced (Duller 2004). When smaller aliquots were examined, no evidence of a significant difference between the standard deviations of small (2 mm width) and large (10 mm width) aliquots was identified, suggesting that partial bleaching was not significant in these samples. The cause of aliquot over-dispersion remains unclear but was also noted in studies by Fattahi et al. (2006). The latter problem of changing calculated De with integration limits can also be indicative of partial bleaching (Singarayer et al. 2005). The decay of quartz OSL at a constant stimulation
62
Table 3. A summary of radiocarbon dating results, Tepe Pardis
Context
Material
Identification details
Reference code
14
d13C
C determination
Date at 95% probability
Date after statistical inference at 95% probability
Bone
Thoracic spine, cattle OxA-14736
215.4
1967 + 31
3 11
Bone Charcoal
Long bone fragment Tamarix sp.*
OxA-14738 OxA-14737
215.2 225.1
5050 + 35 5156 + 37
10 12
4 14
Bone Charcoal
OxA-14739 OxA-14740
217.6 225.4
5894 + 37 6004 + 38
Trench I Trench I Trench I
14 17 20
15 5 18
Charcoal Bone Charcoal
OxA-14741 P16748 P16749
223.9 Fail Fail
5928 + 35
4910 – 4710 BC
4900 – 4780 BC
Trench I
18
17
Charcoal
OxA-14742
225.0
5978 + 38
4990 – 4770 BC
4920 – 4800 BC
Trench II Trench II
1001 1002
21 22
Bone Bone
Trench II
1003
31
Charcoal
Trench II Trench II
1008 1014
24 34
Bone Charcoal
Calcaneus, sheep Fragments too small to identify Populus sp.* Long bone fragment Fragments too small to identify Fragments crumbled when tried to split for identification Vertebrae, sheep? Articulation fragment, sheep? Fragments too small to identify Long bone fragments Fragments too small to identify
Trench I
4
2
Trench I Trench I
8 5
Trench I Trench I
P16951 P16952
50 BC – AD 90 (94.3%) 50 BC – AD 120 AD 100 – 120 (1.1%) Not included in analysis 3960 – 3760 BC 3960 – 3770 BC 4050 – 3930 BC (79.4%) 4050 – 3920 BC (93.1%) 3880 – 3800 BC (16%) 3860 – 3820 BC (2.3%) 4850 – 4680 BC 4830 – 4680 BC 5000 – 4790 BC 4880 – 4740 BC
Fail Fail
OxA-14743
224.2
5976 + 36
4980 – 4770 BC
4950 – 4820 BC
OxA-14744 OxA-14745
216.2 225.0
6000 + 38 6100 + 39
4990 – 4790 BC 5210 – 4910 BC
5000 – 4860 BC 5200 – 5170 BC (1.8%) 5130 – 4930 BC (93.6%)
G. K. GILLMORE ET AL.
Part of site
Sample no.
Trench II
1015
27
Bone
Bone fragments, bird OxA-14746
215.3
6226 + 37
Trench II
1017
29
Bone
217.3
6230 + 45
Quarry
G1
37
Bone
Long bone fragment, OxA-14747 small mammal Bird OxA-14748
218.3
1018 + 29
AD 900 – 920 (1.0%) AD 970 – 1050 (89.1%) AD 1090 – 1120 (4.4%) AD 1140 – 1150 (1.0%)
900 – 1150 BC
Quarry Quarry
G4 G6
40 42
Bone Bone
Quarry – irrigation channel Quarry – irrigation channel Quarry – irrigation channel Quarry – irrigation channel
QX
55
Bone
Long bone fragment Fail Long bone fragment, Fail large mammal Long bone fragments Fail
IX
52
Bone
Long bone fragments Fail
NX
54
Bone
Sheep teeth, young animal
OxA-14749
216.1
6152 + 40
5220 – 4990 BC
5220 – 4990 BC
DX
51
Bone
Long bone fragment, OxA-14750 cattle size?
219.0
6153 + 38
5220 – 4990 BC
5220 – 5030 BC
5280 – 5050 BC 5310 – 5080 BC
GEOARCHAEOLOGY AND MULTIDISCIPLINARITY
*Neither identification is for a particularly long-lived type.
5310 – 5190 BC (49.2%) 5180 – 5060 BC (46.2%) 5310 – 5050 BC
63
64
G. K. GILLMORE ET AL.
through time comprises a number of exponential decay components. Each has a different optical stability, with the initial fast component being one of the most sensitive and quickest to decay. Thus, when sediment is partially bleached, slower components of the OSL decay, when integrated into the De calculation, will show higher equivalent doses and this should be reflected when integration limits are increased to include these slower components. However, the opposite pattern (i.e. a reduction in equivalent dose) was shown in the samples using initial preheats of up to between 220 and 240 8C. This reinforces the probability that the sediments are not partially bleached, but still leaves the cause of this change in De with integration limits unclear. However, when initial preheats were raised to between 240 and 260 8C (dependent on the sample), with second preheats (Fig. 10) tracking at 40 8C lower, this effect was minimized. The dates that passed internal tests using protocols that minimized over-dispersion and change of De with integration limits are shown in Table 2. All dates are in approximate accordance with the site’s pottery-based age model and support the radiocarbon framework published by Coningham et al. (2006). Furthermore, radiocarbon dates from above and below the irrigation channel overlap with the direct OSL date of sediment exposure in the irrigation channel to 1s. This and other important findings arising from this dataset are outlined below.
Discussion Grain-size properties are related to the dynamic conditions under which transport and deposition occur. Trends in grain size can be used, for example, to suggest the direction of sediment dispersal. With conglomerates, it is useful to measure the maximum clast size and bed thickness, as this will give an indication of the competence of the flow (Tucker 2001). However, there were few gravelsized clasts encountered in the bulk of the sediment at Tepe Pardis, but those found were examined for size, shape and roundness characteristics. At the same level as the late Neolithic ceramics and mudbrick structures, a cross-section was also exposed that had the characteristics of a small channel-like structure that ran directly north– south and perpendicular to what appear to be later natural channels higher in the sequence. Identifying the anthropogenic origin of this ancient channellike feature clearly has great significance for interpreting agricultural developments and land use in this archaeologically important region from the late Neolithic to the early Chalcolithic. The independent dating presented here has significant implications for interpretation of the site.
First, as stated above, the OSL chronology broadly supports the radiocarbon chronology published by Coningham et al. (2006). This in turn reinforces some key conclusions concerning occupation at the site. Furthermore, the OSL ages allow inferences to be made concerning sediment aggradation and channel activity at the site. The earliest OSL dates (GG1, 6 and 9) all overlap each other to 2s and are centred on c. 6.4 ka (4400 BC). These dates come from fluvial channels and may indicate an apparent period of enhanced fluvial activity at the site. Whether this enhanced fluvial activity is due to enhanced precipitation or snow melt from the Alborz Mountains is unclear. However, a speleothem record from the monsoonal Qunf cave in Oman (Fleitmann et al. 2003) indicates higher summer monsoon precipitation between 8 and 6 ka (6000–4000 BC). Whether this extended as far north as the Tehran Plain is unclear. A further possibility is zonal moisture transport from the Mediterranean or enhanced summer snow melt. Both possibilities would be tied more closely to westerly circulation, and reconstructions do suggest that higher temperatures (and probably precipitation) occurred around 6 ka (4000 BC) in the North Atlantic (Kaplan & Wolfe 2006). More evidence is needed to differentiate between these possible causes of enhanced fluvial activity at Tepe Pardis and adjacent sites to gain a regional record of activity. In association with these natural channels is the proposed anthropogenic irrigation canal (GG4). This is dated by OSL at 6.7 + 0.4 ka (5100– 4300 BC) and the sediments surrounding are dated at 5220–4990 BC by calibrated radiocarbon methods. This result confirms the antiquity of the feature and demonstrates that it was formed at about the same time as (or is even older than) the palaeo-river channels. It is also significant that the two independent dating methods agree within errors on the age of the feature. That it was formed at around the same time as the natural channels suggests that it was constructed as a response to the availability of water from natural channels at this point, and may be considered as designed to harness this resource. The age of this channel is close to those proposed for the first irrigation channels in Mesopotamia, such as the mid-sixth millennium BC dates for the channel at Choga Mami (Oates 1982) and on the Deh Luran plain (Hole 1977), and suggests a rapid diffusion of this skill to the more arid, peripheral regions to the east. Potentially strong links between Mesopotamia and the Tehran Plain allowed this diffusion to take place, and this finding emphasizes the potential importance of the Tehran Plain in understanding the diffusion of cultural ideas from core centres to adjacent peoples.
GEOARCHAEOLOGY AND MULTIDISCIPLINARITY
Evidence can be found in the area for ceramic production while this diffusion of ideas about irrigation appears to have been occurring. The site of Pardis, on the central Iranian plateau, was excavated in 2006 and, in trench VII, 50 cm of Late Neolithic layers, followed by 2 m of both Late Neolithic ceramics and Transitional Chalcolithic ceramic assemblages were recorded. Although the trench size was small (2 m 2 m) remarkable evidence of craft tools indicates that this site, from the beginning of its occupation, was used for ceramic production (H. Fazeli, pers. comm.). The site was therefore a centre of resource use around this period, with significant human interaction with the environment. After c. 6 ka (4000 BC), the OSL dates show a hiatus in the fluvial record of the site. The absence of river channels suggests that the source of water available around 6.5 ka (4500 BC) disappeared until at least 3 ka (1000 BC). Whether this trend is tied to the gradual cooling and/or aridification of the summer monsoon and the North Atlantic is unclear. However, occupation began again at around 2.9 ka (900 BC) and is marked by fluvial channel development, where the OSL ages of dated channels all overlap to 1s (GG7, 10, 11). This phase of fluvial activity does not correspond to any changes in the Oman speleothem record (Fleitmann et al. 2003), nor to any obvious large-scale North Atlantic changes (Kaplan & Wolfe 2006). Finally, a red soil horizon (GG3) formed in sediments deposited at c. 1.8 ka (c. AD 200), followed by the deposition of fluvial sediments at 1.3 ka (c. AD 700; OSL sample GG8), and clay above a cemetery at 1.2 ka (c. AD 800; sample GG5; Fig. 12, Table 2). The presence of the Iron Age cemetery possibly suggests increased activity at the site during this time (1500 BC onwards is often taken as the development of the Iron Age in this region, with the onset of Iron Age I), although
65
there is no evidence of sedentary occupation at the tell site. However, it is possible that erosion may have removed Iron Age levels, prior to deposition of later material, or that quarrying activity has removed key evidence. Whether irrigation occurred at the site in later times is unclear, but further fluvial activity is recorded in the Umayyad –Abbasid Dynastic periods (GG8), perhaps indicating the onset of a wetter phase, when the site could have been used as a village and for craft production. The AD 200 date for GG3 is interesting, in that there were Parthian (commonly dated as being between 247 BC and AD 224) sherds in the quarry machine spoil cleared from the disturbed layers above Iron Age graves, together with Parthian material from the top of the Tepe. This coincides with the reoccupation of the Tepe when a major mudbrick structure was built on the weathered archaeological site of which the lower foundation of 4 m still remains. It has been suggested that the cemetery would have been in use at Tepe Pardis from 1550 BC to 1300 BC (H. Fazeli, pers comm.). The younger OSL ages from sediments above the cemetery appear to confirm this chronology. In relation to the lack of Holocene alluviation proposed by Beaumont (1972), the results above suggest that this hypothesis may not apply to all of the Holocene. At a minimum of three times over the Holocene (centred on c. 6.5 ka, 2.9 ka and 1.8 ka), potentially by means of increased temperature or precipitation, more water may have been available at the site. This was particularly the case at around 6.5 ka, which broadly corresponds to the so-called ‘Holocene Optimum’ as recorded in both the Arabian monsoon (Fleitmann et al. 2003) and the North Atlantic (Kaplan & Wolfe 2006). Although the Tepe Pardis site contains no direct evidence of large-scale alluvial fan alluviation, the record does suggest that around 6.5 ka conditions were different enough that deposition and the build-up of fan deposits may have been possible. Such a possibility deserves further investigation as it will have implications for understanding longterm Holocene climate change in this arid region where small changes in available moisture are hugely significant both for the environment and human habitants.
Conclusions
Fig. 12. The Iron Age cemetery, Tepe Pardis.
Radiocarbon and OSL chronologies are broadly in agreement within errors and reinforce the site’s chronology. Furthermore, the OSL dates show no evidence of partial bleaching, as often reported for sediments deposited under fluvial conditions, and appear to be in good stratigraphic order, agreeing with the broad site stratigraphy. Table 2 shows the
66
G. K. GILLMORE ET AL.
OSL dates for the site and Table 3 shows the radiocarbon dates. Together, the dates demonstrate that: (1) artificial irrigation appears to have been occurring at the same time as or prior to natural river channel development at the late Neolithic – Transitional Chalcolithic interface; (2) this was just prior to or coincident with large-scale pottery production, utilizing the sand and clay deposits on the fan during the Transitional Chalcolithic period; (3) an Iron Age cemetery discovered in the last season of excavation (see Fazeli et al. 2007) was sited close to the old tell considerably later. These multidisciplinary results have implications for the environments around the Tepe both at the time of occupation and later, and allow us to state that there is clear evidence at this site of some form of water management or environmental exploitation and manipulation taking place in the Neolithic –Chalcolithic. Thus, the site provides important evidence for the interaction of environment and occupation or land use in the development of the settlement of Tepe Pardis, and suggests a strong and rapid diffusion of irrigation to the east after development in Mesopotamia. Furthermore, the findings suggest that Holocene climate and alluviation in the area were highly variable, reinforcing the idea of instability in the landscape in this area. This work was made possible by grants from the Iranian Centre for Archaeological Research, the British Academy and the British Institute of Persian Studies. The authors would also like to thank M. Manuel for field and technical assistance, and G. Rushworth for laboratory work in support of this project. Thanks also go to C. Ivison for cartographic assistance.
References Abbassi, M. R. & Farbod, Y. 2009. Faulting and folding in Quaternary deposits of Tehran’s piedmont (Iran). Journal of Asian Earth Sciences, 34, 522– 531. Adamiec, G. & Aitken, M. J. 1998. Dose rate conversion factors: update. Ancient TL, 16, 37– 50. Armitage, S. J. & Bailey, R. M. 2005. The measured dependence of laboratory beta dose rates on sample grain size. Radiation Measurements, 39, 123– 127. Beaumont, P. 1972. Alluvial fans along the foothills of the Elburz Mountains, Iran. Palaeogeography, Palaeoclimatology, Palaeoecology, 12, 251– 273. Beuselinck, L., Govers, G., Poesen, J., Degraer, G. & Froyen, L. 1998. Grain-size analysis by laser diffractometry: comparison with the sieve–pipette method. Catena, 32, 193–208. Blott, S. & Pye, K. 2001. GRADISTAT: A grain size distribution and statistics package for the analysis of unconsolidated sediments. Earth Surface Processes and Landforms, 26, 1237– 1248. Bøtter-Jensen, L., Bulur, E., Duller, G. A. T. & Murray, A. S. 2000. Advances in luminescence
instrumentation. Radiation Measurements, 32, 523–528. Bronk Ramsey, C. 1995. Radiocarbon calibration and analysis of stratigraphy: the OxCal program. Radiocarbon, 37, 425–430. Coningham, R. A. E., Fazeli, H., Young, R. L. & Donahue, R. E. 2004. Location, location, location: a pilot survey of the Tehran Plain in 2003. Iran, 42, 13–23. Coningham, R. A. E., Fazeli, H. et al. 2006. Socioeconomic transformations: settlement survey in the Tehran Plain and excavations at Tepe Pardis. Iran, 44, 33–62. Duller, G. A. T. 2003. Distinguishing quartz and feldspar in single grain luminescence measurements. Radiation Measurements, 37, 161– 165. Duller, G. A. T. 2004. Luminescence dating of Quaternary sediments: recent advances. Journal of Quaternary Science, 19, 183– 192. Evans, D. J. A. & Benn, D. I. (eds) 2004. A Practical Guide to the Study of Glacial Sediments. Edward Arnold, London. Fattahi, M., Walker, R. et al. 2006. Holocene slip rate on the Sabzevar thrust fault, NE Iran, determined using optically stimulated luminescence (OSL). Earth and Planetary Science Letters, 245, 673–684. Fazeli, H. 2001. Social complexity and craft specialisation in the Late Neolithic and Early Chalcolithic period in the Central Plateau of Iran. PhD thesis, University of Bradford. Fazeli, H., Coningham, R. A. E. & Batt, C. M. 2004. Cheshmeh-Ali revisited: towards an absolute dating of the Late Neolithic and Chalcolithic of Iran’s Tehran Plain. Iran, 42, 13–23. Fazeli, H., Coningham, R. A. E., Young, R. L., Gillmore, G. K., Maghsoudi, M. & Raza, H. 2007. Socio-economic transformations in the Tehran Plain: final season of settlement survey and excavations at Tepe Pardis. Iran, 45, 267– 286. Fleitmann, D., Burns, S. J., Mudelsee, M., Neff, U., Kramers, J., Magini, A. & Matter, A. 2003. Holocene forcing of the Indian monsoon recorded in a stalagmite from southern Oman. Science, 300, 1737– 1739. Folk, R. L. 1954. The distinction between grain size and mineral composition in sedimentary-rock nomenclature. Journal of Geology, 62, 344– 359. Folk, R. L. 1966. A review of grain-size parameters. Sedimentology, 6, 73–93. Folk, R. L. & Ward, W. C. 1957. Brazos River bar: a study in the significance of grain size parameters. Journal of Sedimentary Petrology, 27, 3– 26. Friedman, G. M. & Johnson, K. G. 1982. Exercises in Sedimentology. Wiley, New York. Gillmore, G. K., Coningham, R. A. E., Young, R., Fazeli, H., Rushworth, G., Donahue, R. & Batt, C. M. 2007. Holocene alluvial sediments of the Tehran Plain: sedimentation and site visibility. In: Wilson, L., Dickinson, P. & Jeandron, J. (eds) Reconstructing Human–Landscape Interactions. Cambridge Scholars, Newcastle, 37– 67. Gillmore, G. K., Coningham, R. A. E., Fazeli, H., Young, R. L., Maghsoudi, M., Batt, C. M. &
GEOARCHAEOLOGY AND MULTIDISCIPLINARITY Rushworth, G. 2009. Irrigation on the Tehran Plain, Iran: Tepe Pardis—the site of a possible Neolithic irrigation feature? Catena, 78, 285–300. Hole, F. 1977. Studies in the Archeological History of the Deh Luran Plain: the Excavation of Chagha Sefid. Museum of Anthropology, University of Michigan, Ann Arbor, MI. Jones, A., Duck, R., Reed, R. & Weyers, J. 2000. Practical Skills in Environmental Science. Prentice–Hall, London. Kaplan, M. R. & Wolfe, A. P. 2006. Spatial and temporal variability of Holocene temperature in the North Atlantic region. Quaternary Research, 65, 223–231. King, C. A. M. 1975. Techniques in Geomorphology. Edward Arnold, London. Mejdahl, V. 1979. Thermoluminescence dating: betadose attenuation in quartz grains. Archaeometry, 21, 61– 72. Murray, A. S. & Olley, J. M. 2002. Precision and accuracy in the optically stimulated luminescence dating of sedimentary quartz: a status review. Geochronometria, 21, 1 –16. Murray, A. S. & Wintle, A. G. 2000. Luminescence dating of quartz using an improved single-aliquot regenerative-dose procedure. Radiation Measurements, 32, 57–73. Murray, A. S., Marten, R., Johnston, A. & Martin, P. 1987. Analysis for naturally occurring radionuclides
67
at environmental concentrations by gamma spectrometry. Journal of Radioanalytical and Nuclear Chemistry, 115, 263– 288. Oates, J. 1982. Choga Mami. In: Curtis, J. (ed.) Fifty Years of Mesopotamian Discovery. British School of Archaeology in Iraq, London, 22–29. Prescott, J. R. & Hutton, J. T. 1994. Cosmic ray contributions to dose rates for luminescence and ESR dating: large depths and long term variations. Radiation Measurements, 23, 497–500. Prothero, D. R. & Schwab, F. 1996. Sedimentary Geology: An Introduction to Sedimentary Rocks and Stratigraphy. Freeman, San Francisco, CA. Reimer, P. J., Baillie, M. G. L., Bard, E. et al. 2004. Intcal04 Terrestrial Radiocarbon Age Calibration, 0– 26 Cal Kyr BP. Radiocarbon, 46, 1029– 1058. Singarayer, J. S., Bailey, R. M., Ward, S. & Stokes, S. 2005. Assessing the completeness of optical resetting of quartz OSL in the natural environment. Radiation Measurements, 40, 13– 25. Tucker, M. E. 2001. Sedimentary Petrology, 3rd edn. Blackwell, Oxford. Tucker, M. E. 2003. Sedimentary Rocks in the Field, 3rd edn. Wiley, Chichester. Walford, N. 1995. Geographical Data Analysis. Wiley, Chichester. Wilcock, P. R. 1993. Critical shear stress of natural sediments. Journal of Hydraulic Engineering, 119, 491– 505.
Early Neolithic sands at West Voe, Shetland Islands: implications for human settlement G. K. GILLMORE1* & N. MELTON2 1
Centre for Earth and Environmental Science Research, School of Geography, Geology and the Environment, Kingston University, Kingston-upon-Thames KT1 2EE, UK 2
Archaeological, Geographical and Environmental Sciences, University of Bradford, Bradford BD7 1DP, UK *Corresponding author (e-mail:
[email protected]) Abstract: In 2002 and 2004–2005 archaeological investigations were undertaken on middens exposed by coastal erosion at West Voe in the south of Mainland Shetland, UK. This work established that the site dated from c. 4000 cal BC to c. 3250 cal BC and was of major importance for two reasons: (1) as the first of Mesolithic date to be found on Shetland; (2) as the first site to be found in the Northern Isles that spanned the Mesolithic– Neolithic transition. This paper describes investigations into the origin of sands deposited around 3500 cal BC and their potential effect on human settlement. The sands in question lie between two midden deposits, the lower of which accumulated over the period 4000– 3500 cal BC and the upper 3500– 3250 cal BC. The sands, therefore, dated to the period shortly after the adoption of agriculture on the archipelago, represented in the lower midden by the appearance of domesticated species and ceramics at around 3700– 3600 cal BC, and represented a disruption in human occupation at a critical point in the development of a changing use of the landscape.
Preliminary investigations in 2002 on a shell midden exposed by coastal erosion at West Voe, in the south of Mainland Shetland (see Fig. 1; Melton & Nicholson 2004), established that the site consisted of two middens, the lower of which appeared to be predominantly composed of oysters and sealed a thin black sandy Holocene layer, typically 0.10 m thick, that overlay glacial till. This lower midden was sealed by c. 0.4 m of sand, the subject of this paper. Above the sand was a second, upper, midden that consisted entirely of cockles. This midden butted a substantial wall, and was sealed by a sequence of dune sands in excess of 6 m thickness. A single radiocarbon date was obtained from each of the middens and these indicated that the lower midden dated from the late fifth millennium BC and the upper from the second quarter of the fourth millennium BC, and thus that the midden sequence spanned the Mesolithic– Neolithic transition. The West Voe site, therefore, was the first of Mesolithic date to be identified in Shetland and was significant as it had generally been accepted that coastal sites of this period in Shetland, the outermost point in the Mesolithic colonization of the North Atlantic, would have been lost to Holocene sea-level rise (Mykura 1976, pp. 110 –111; Firth & Smith 1993; Fojut 2006, p. 8); hence the interest in assessing the sands in question for palaeoenvironmental data and proximity to the shoreline.
The 2004– 2005 excavations at West Voe A 14 m length of the cliff face was examined in 2004– 2005. The section examined in 2004 incorporated the preliminary investigations carried out in 2002 and was positioned to permit examination of both the lower and the upper middens and the wall that the latter butted (see section Figs 2 & 3). The thick deposits of unstable sands that sealed the middens restricted the investigations to the preparation and recording of a vertical section, exposing a narrow strip of the lower midden, typically 0.5 m wide. A narrow, 2 m long, step was also cut at the level of the ground surface associated with the upper midden, which was column sampled. To ensure maximum recovery of artefactual and economic information from the limited area of midden available, a total sampling policy was adopted, with all excavated midden deposits being processed using a Siraf-style flotation tank. The midden deposits within this strip were recorded in plan. The same approach was used in 2005 when the section was extended to the north of that examined in 2004 (Fig. 1) to confirm the northern extent of the lower midden and to determine whether evidence for human activity was present to the north of the middens (Melton & Nicholson 2007; Melton 2008, 2009). The excavations revealed that the lower midden contained a succession of shellfish species (see
From: Wilson, L. (ed.) Human Interactions with the Geosphere: The Geoarchaeological Perspective. Geological Society, London, Special Publications, 352, 69– 83. DOI: 10.1144/SP352.6 0305-8719/11/$15.00 # The Geological Society of London 2011.
70
G. K. GILLMORE & N. MELTON
Fig. 1. Plan showing the location of the West Voe midden excavations in 2004– 2005, regional map, and map showing the relationship of the study site to other sites mentioned in the paper.
Fig. 4). At the base of the midden was a layer of large oyster shells that had been pressed into, and become intercalated with, the underlying black sandy Holocene layer recorded in 2002. The midden deposits sealed a possible archaeological feature with vertical sides and a flat base that contained coarse angular gravel with no evidence for human activity (Fig. 2). Overlying the oysters was a layer of limpet shells that contained pockets of seal and sea-bird bones. On the upper surface
of this layer was an activity surface, formed from trampled mussel shells and with uncrushed mussels lying on top of it. Bone fragments from large terrestrial ungulates and sherds of pottery were recovered from the activity surface and the layers immediately above it. Evidence for offmidden activity was present in the form of a 0.8 m wide band of trampled shell overlain by uncrushed mussels that was located 2.5 m to the north of the midden’s edge.
EARLY NEOLITHIC SANDS, SHETLAND ISLANDS
71
Fig. 2. West Voe: midden section excavated in 2004.
Sealing the lower midden were the sands. They varied between 0.3 and 0.45 m in thickness and had a number of pits and linear features cut into them (see Figs 2 & 5). Within the sands, a thin band of charcoal with shells, which was present over a distance of c. 1 m, possibly represented the remains of a bonfire, indicating that during their deposition there had been at least one period of stability in which humans had been present. The sands were sealed by the upper, cockle, midden that butted a 0.6 m thick wall that survived to a
height of 0.6 m (see Fig. 6). To the north of the wall a cow tooth and fragment of sheep skull were found in a thin layer of shells present in the sand that had accumulated against the wall, and represented the final evidence for a prehistoric human presence on the site (Fig. 2). Radiocarbon dates on material recovered in 2004– 2005 indicated that the oyster phase of the lower midden dated to c. 4000–3700 cal BC; the limpet phase to c. 3700–3600 cal BC; the activity surface to c. 3700–3600 cal BC; and the final phase,
Fig. 3. The wall and cockle midden sequence illustrated in Figure 2.
72
G. K. GILLMORE & N. MELTON
Fig. 4. The lower midden exposed by coastal erosion at West Voe. Fig. 5. Sample section through the sand sequence at West Voe.
the deposits of complete mussels, to c. 3650–3500 cal BC. The upper, cockle, midden was dated to c. 3500–3250 cal BC. The detailed analysis of the radiocarbon dates from the site will be discussed in a separate paper, but details of those that relate to the deposition on the sands examined here are given in Table 1. They indicate that the sands between the middens had been laid down within a relatively short period in the middle of the fourth millennium BC.
Geological context The east-facing lithological sequence (running parallel to north –south) at West Voe consisted of a solid geology in the form of well-bedded buff– light red Old Red Sandstone (ORS) with depositional sedimentary structures in the form of small-scale cross-bedded (millimetres thickness) and ripple laminations, overlain by a horizon of fragmented, contorted and folded ORS material. This was overlain by a diamict containing large (e.g. 120 mm long by 100 mm breadth) to small (e.g. 20 mm length by 14 mm breadth) clasts with a matrix of reddish clay (sandy in parts). Sealing this was the sequence of middens and sands described above.
The local and regional settings of the West Voe middens in Mesolithic and Early Neolithic Shetland The Shetland Isles were separated from the British mainland throughout the Holocene (Warren 2005, figs. 17 and 18) and it was long considered that the sea crossing involved would have prevented the Mesolithic colonization of the archipelago. The first indication that Mesolithic colonizers had the necessary seafaring ability to make this crossing was the finding of a Mesolithic core-tool on Fair Isle in the 1940s (Cumming 1946). Subsequent to this, scatters of Mesolithic lithic artefacts have been found on Orkney (Saville 2000, pp. 95– 97; Wickham-Jones 2006, pp. 23 –24) and a Mesolithic human presence on Shetland had been inferred from vegetation changes seen in pollen cores taken in west and north Mainland at Dallican Water (Bennett & Sharp 1993), Cata Ness (Bennett et al. 1992), Gunnister Water (Bennett et al. 1993) and Loch of Brunnawatt (Edwards & Moss 1993). Mesolithic sites in the Northern Isles, however, remained elusive in the archaeological record and it was generally considered that as sites of this period would probably have been close to the coast
EARLY NEOLITHIC SANDS, SHETLAND ISLANDS
73
Fig. 6. Neolithic sands sealed by cockle midden with associated wall.
to optimize subsistence strategies, they would have been lost to Holocene sea-level rise. In Shetland the few sites that have provided Early Neolithic dates are restricted to a boundary dyke at Ward of Shurton, near Lerwick (Whittington 1979), the earliest phases of the agricultural settlement excavated at Scord of Brouster in West Mainland (Whittle 1986), and the skeletal remains found in a cist uncovered during construction works
at Sumburgh Airport, only 400 m to the north of the West Voe middens (Hedges & Parry 1980). A date of c. 3500 cal BC has been suggested for the start of the Neolithic period on Shetland, on the basis of the local development of pottery and architectural forms (Fojut 2006, pp. 13–27). The West Voe middens are located in the NW of the voe, directly opposite the multi-period site of Jarlshof (see Fig. 1). Excavations at Jarlshof in the
Table 1. Radiocarbon dates from contexts above and below the Early Neolithic sands Laboratory code OxA-14161 OxA-14180
Material
Marine shell (Cerastoderma edule) Marine shell (Cerastoderma edule)
Early Neolithic sand deposit SUERC-9011 Marine shell (Patella vulgata)
All dates calibrated using Calib 5.0.2.
d13C
Radiocarbon age (years)
Calibrated date (95.4% confidence)
1.4
4955 + 32
3479 – 3275 BC
Context 407. Upper (cockle) midden
1.3
4945 + 65
3499 – 3107 BC
Context 420. Fill of pit cut into Early Neolithic sands and sealed by wall and upper midden
1.1
5135 + 35
3634 – 3476 BC
Context 570. Lens of shells, uppermost layer in lower midden. Sealed by Early Neolithic sands
Comments
74
G. K. GILLMORE & N. MELTON
twentieth century revealed a sequence of occupation that ran from the Early Bronze Age to PostMedieval times (Hamilton 1956), although evidence that was thought to suggest an earlier transient human presence was also noted and comprised occasional mussel shells and stones in basal deposits above a clean sand that sealed the natural clay and rockhead (Hamilton 1956, p. 9). The evidence for an early human presence at Jarlshof has been confirmed by recent investigations (Dockrill et al. 2005) and in the vicinity of the West Voe middens is further indicated by the Early Neolithic multiple burial discovered in 1977 during construction works at Sumburgh Airport (Hedges & Parry 1980). The discovery of the West Voe middens has, therefore, provided the first opportunity to examine Late Mesolithic subsistence strategies in Shetland and to date the appearance of ceramics and cattle and sheep on the archipelago. In addition to the detailed economic evidence and dates from the West Voe site, the recently published pollen core from the Loch of Gards (Edwards et al. 2009) on the Scat Ness peninsula, some 0.5 km to the SW of the midden site (see Fig. 1), provides a complementary, broader-scale view of the local Holocene landscape. The picture that emerges from the Loch of Gards pollen analysis is one of a partial covering of woodland, with species that included birch, hazel and willow, that extended into the Mesolithic. A gradual reduction in woodland cover occurred during the period c. 6350– 6000 cal BC, with the minimum woodland cover being between c. 6050 and 5890 cal BC. There was evidence for a regeneration of woodland cover prior to a second phase of reduction that commenced around 3910 cal BC. The suggestion from the pollen record, therefore, is that there were two Mesolithic clearance events and that the commencement of deposition of the West Voe middens coincided with the later event recorded in the pollen core.
et al. (2011). For ease and adequate accuracy of analysis, each sandy bed was dry sieved using a standard set of sieves at half-phi intervals with a pan at the base (see Friedman & Johnson 1982; Prothero & Schwab 1996; Jones et al. 2000, for comments on the technique), using a mechanical sieve shaker for at least 10 min per sample. Representative subsamples were taken by splitting the bulk sample using a sample splitter to produce samples of 50–100 g. Where appropriate, the samples were washed to remove soluble salts and dried on a tray using an oven at a low temperature (65 8C). The samples were weighed and the mass was recorded. Standard measures of central tendency and dispersion were calculated; that is, phi mean, median, sorting (standard deviation), skewness and kurtosis (Folk 1966; Walford 1995). GRADISTAT 5.0 (see Blott & Pye 2001) was used to plot the data. The sand sequence examined at West Voe has also been assessed for carbon content. The samples were heated to burn off organic material before dry sieving. To measure this loss on ignition, the samples were dried at 105 8C for 24 h and then heated to 550 8C for 2 h. Weight loss on ignition was calculated as a percentage of the dried sample weight. Single sand (and rock fragment) grains were also viewed using an SEM to assess grain shape and roundness, and surface texture (see Fig. 7). Micropalaeontology. All sieved fractions were examined for microfaunal content under an Olympus SZ30 binocular reflected light microscope. Specimens were picked out using a OOO sable hair brush and placed into a microfaunal slide. A target of 200–250 specimens was set to statistically represent in situ microfaunas for each sample. As a result of abundance issues this target was not always met. The specimens were later mounted onto an SEM stub and viewed in a Zeiss EVO55 Extended Variable Pressure SEM (see Figs 8 & 9).
The geoarchaeology of the Early Neolithic sands
Results
Methods: field and laboratory
Sediments
Sediments. Graphic logs were constructed of the sections at West Voe and samples were removed for laboratory-based grain size, shape, mineralogical and micropalaeontological assessment. When deciding which analytical techniques to use comparisons have been made in the literature between laser particle-size analysis, dry sieving and the sieve –pipette method. Brief discussions on the benefits or otherwise of these techniques have been given by Beuselinck et al. (1998) and Gillmore
The underlying diamict consisted of mostly local ORS clasts, but also contained granitic and gneissic clasts in places. The thickness of this sediment varied but at the main log section was 0.8 m. In the field it was observed that the sands above the diamict, within the ‘target’ sequence, contained well-rounded comminuted shell fragments plus clear quartz grains and rock fragments (probably ORS sandstone and shale). The sand was divided up into many (c. 20) thin beds several centimetres
EARLY NEOLITHIC SANDS, SHETLAND ISLANDS
75
Fig. 7. SEM image of a sand grain, sample WV05.
thick (e.g. 2–20 cm) above a horizon of flat stones in sand. They appear to be dipping between 108 and 48 towards the east (0808). Within those beds few laminations can be observed and most appear to be structureless, except for grading. The tops of some of the horizons were uneven with the beds being of variable thickness. The difference between
the beds was highlighted primarily by subtle colour changes, some appearing to be darker and others more grey, together with subtle grain-size variation. The quartz grains had a fresh appearance (this was confirmed under stereomicroscopic examination in the laboratory). Some of the sand grains are angular, but many appear to be subrounded to
Fig. 8. SEM image of selected foraminiferal species (E. crispum).
76
G. K. GILLMORE & N. MELTON
Fig. 9. SEM image of selected foraminiferal species (C. lobatulus).
subangular, with some being closer to rounded. A few grains are well rounded. Grain size was variable but towards the base of the sequence examined it was a medium to coarse sand (375 –750 mm) in hand specimen. More detailed examination of grain size was undertaken in the laboratory to assess distribution. A comparative sample from the modern beach at West Voe was examined for grain characteristics in the field and laboratory. The results of sedimentary analysis were plotted as frequency histograms and on cumulative frequency curves. Over 30 samples were analysed using this approach. Figures 10 and 11 illustrate one example; here the sediment can be named as a ‘moderately well-sorted medium sand’, with 70% of the grain-size distribution being within medium sand, 12% in coarse sand, over 16% fine sand and 0.6% very fine sand. The distribution is unimodal, with a mean of 1.538 (medium sand), a sorting of 0.509, skewness of 0.021 (symmetrical) and a kurtosis value of 0.991 (mesokurtic) (statistics reported based on Folk & Ward 1957 logarithmic phi values). The modern beach sand that was also analysed was generally coarser than the sand beds at the excavation site (with 52.6% coarse sand and 1.3% gravel) but the nature of the distribution was similar in many respects, being moderately sorted with a skewness of 0.005 (symmetrical) and a kurtosis value of 1.030 (mesokurtic). Sample WV05 1, which was the first sand horizon above the lower midden, contained more shell fragments and fossil foraminifera than the beds above, and
not surprisingly had a slightly different sedimentological profile, with less fine material.
Micropalaeontology Sand horizons 1 and 2 (sample numbers WV05 1 and WV05 2) have been examined in detail to see what mineral and biogenic materials they contained. Under a stereoscopic microscope it is clear that quartz grains are common with rock fragments in the smaller size fractions, but there are many biogenic grains particularly in the larger fractions, as hand lens examination in the field suggested. These biogenic clasts are mainly of fragmented eroded shells (bivalves and gastropods) together with broken sea urchin spines and bryozoan remains. Microfossils in the form of calcareous benthonic foraminifera appear to be common, particularly in horizon 1 (which contained more biogenic clasts than the overlying horizons). These benthic foraminifera are dominated by Cibicides lobatulus (mostly in a reasonable condition, although some have broken chambers or eroded surfaces) and C. refulgens, with Eponides repandus, poorly preserved elphidiids and rare Elphidium crispum. Comparisons were made with the modern beach sands, which also contain similar benthonic foraminiferal assemblages (e.g. C. lobatulus, but in higher numbers and with better preservation) but additionally contain agglutinated benthonic foraminifera. These results are presented in Table 2 as a simple presence or absence table, and are discussed below in detail.
EARLY NEOLITHIC SANDS, SHETLAND ISLANDS
77
Fig. 10. Example plot of a sedimentary analysis frequency histogram.
Discussion Sediments Beach sediments tend to have an excess of coarse material in the ‘tail’ of the distribution because the finer components are carried away by wave action (see Tucker 2001). Dune sands, on the other hand, may show a high kurtosis value and a positive skewness (with a fine tail material) owing to the small amount of fine silt in the sediment (King 1975). They also tend to be better sorted than river sands. In general, sediment becomes more negatively skewed (and finer grained) along its sediment transport pathway.
Work by Lovell (1979) to assess how the composition of rocks on a land source area is reflected in the petrography of reworked marine sands in an area of recent glaciation and marine transgression may have some relevance to this study. Lovell (1979) noted that the principal characteristic of onshore and nearshore sands was the predominance of igneous rock fragments. Offshore sands, however, were dominated by quartz and biogenic carbonate (as the sands of West Voe appear to be). Sources for the offshore sands included glacial sediments that underlie Holocene sands (Lovell 1979). There are always difficulties in assessing provenance however, owing to transgressions. Swift (1969) has highlighted that modern
78
G. K. GILLMORE & N. MELTON
Fig. 11. Example plot of a sedimentary analysis cumulative frequency curve.
marine shelves rely on their Pleistocene substrates as a source of sediment. Where Holocene sands are concerned, diagenesis often modifies texture and composition considerably. Biogenic carbonate sands may be preserved as calcareous cement for example, or much of the carbonate may be removed after lithification (Lovell 1979; Murray 1991). Basaltic rock fragments in sands examined by Lovell (1979) from Mull beaches were in an advanced stage of chemical alteration and physical disintegration. The lack of clear sedimentary structure within the sand horizons is notable. Both siliciclastic and carbonate sands deposited on a beach of moderate to high wave activity are often characterized by very low-angle planar cross-stratification arranged into truncated sets, with the low-angle flat bedding
directed offshore (Tucker 2003). The lack of structure within these beds suggests deposition possibly as sand sheets. Aeolian sediments often also show cross-sets but on a much larger scale. The beds themselves may show reverse grading from avalanching (grain-flow deposits) and normal grading (as in the beds in this study) from traction flow. Coarser beds may often alternate with finer beds deposited from suspension (grain-fall deposits). Aeolian material is typically well-sorted, wellrounded, medium-grained sand. Coarse to finer graded bedding may be generated through deposition from decelerating sediment-laden currents. On clastic shelves and carbonate ramps, graded beds (tempestites) are deposited from waning storm currents (Tucker 2001); storm beds, however, are generally poorly sorted.
Table 2. Presence or absence data for foraminiferal analysis from the modern beach at West Voe (WV05 BS) and from samples taken from the sand column above the lower midden (WV05 1 and 2) Sample WV05 BS X X X X X X X X X X
WV05 1
WV05 2
Species names
X X X X X
X X
Cibicides lobatulus Cibicides refulgens Elphidium crispum Eponides repandus Haynesina orbiculare (?) Trifarina angulosa Quinqueloculina lata Quinqueloculina sp. Gaudryina rudis Textularia saggitula group
X
X X
EARLY NEOLITHIC SANDS, SHETLAND ISLANDS
It is interesting to note that these sands may date to a time when tsunami deposits were laid down elsewhere in the Shetland Islands (e.g. 3550 cal BC; see Bondevik et al. 2005). Tsunami deposits are well preserved in coastal lakes inundated by tsunamis (Bondevik et al. 1997) and Shetland has around 2500 lochs. Bondevik et al. (2005) suggested that peat is a possible depository for such sandy deposits. The sands at West Voe investigated by this study were fairly evenly laid down and they did not contain rip-up clasts. However, Bondevik et al. (2005) did note that some of the tsunami deposits they observed were structureless (as some of the sands at West Voe are, although most appear to show some graded bedding). Ideally, one would wish to explore these sand horizons in more detail in three dimensions, that is, explore their bed thickness variations inland through coring. This may be an option for future studies and, indeed, has begun already (by the authors). Comparative material was collected from known tsunami deposits in the north and east of the Shetlands and subjected to grain-size analysis and microscopic examination. In the field, these tsunami deposits are often very fragmented, in that they may be dominated by angular clasts. The c. 3550 cal BC tsunami event noted by Bondevik et al. (2005) at Garth Loch in eastern Shetlands (currently the most southerly recording of tsunamigenic deposits; see Fig. 1), contains lamina of medium to fine sand overlying lacustrine gyttja deposits, with clasts of gyttja incorporated, and fine sandy –silt lamina that are possibly graded at the upper boundary. Bondevik et al. (2005) stated that high tide level at this time was at least 2 m below the present high tide. Deposits of the same age were noted nearby at the Loch of Benston at a height of 1.6 m above high tide, which suggests a minimum run-up of 3.6 m. However, the sea level at 3550 cal BC has been suggested by several researchers to be between 7 and 12 m below the present high tide based on sealevel curves (see Mykura 1976; Lambeck 1993; Peltier et al. 2002). This would give a run-up for the Garth tsunami event of more than 10 m (Bondevik et al. 2005). The lower midden at West Voe occurs at 3.8 m above Ordnance Datum; hence the interest in this study to ascertain the nature of the sands at West Voe. If these were associated with a tsunami event the implications for human settlement would be significant.
Micropalaeontology The benthonic foraminiferal microfauna noted in this study are typical of high-energy shallow-marine environments found in the inner neritic zone (water depths 0 –50 m) according to Murray (1991; see also Gillmore et al. 2001). Murray (1991) suggested
79
that although Cibicides can be found in a wide variety of marine environments it is noted in modern-day environments around the Shetlands. Similarly, Eponides and Elphidium have been noted as part of the Cibicides –Eponides microfaunal assemblage in shallow waters around the Shetlands (Murray 1991). A comparative study was made of the modern beach sand at West Voe. The faunal assemblage included abundant shell fragments (mostly bivalve but also with abundant gastropods), abundant sea urchin spines and fragments, rare sponge spicules and common bryozoan fragments. Foraminifera were much more common than in sample WV05 1, but the suite was similar, being dominated by Cibicides lobatulus and C. refulgens, with Eponides repandus, Elphidium (specifically E. crispum), rare miliolids (e.g. Quinqueloculina lata), but with the addition of agglutinated benthonic foraminifera. The modern beach sediments (WV05 BS) also contained rare Trifarina angulosa. This species has a preference for muddy sandy substrates, within a temperature range in the Labrador Sea of 1–8 8C (Murray 1991). Q. lata has also been noted on sandy substrates around the UK at depths of 20 – 120 m (Murray 1991). Out of over 200 specimens of foraminifera picked from the modern beach sample (WV05 BS), only 10 species were recorded (although this was a higher species diversity than for WV05 1). The alpha diversity index can be calculated as two, possibly suggesting a hypersaline lagoonal depositional environment, and the summary triangular plot after Murray & Wright (1974) also suggests a lagoonal setting. However, both these plots could also represent one end of the normal salinity marine environmental depositional setting. No planktonic foraminifera were recorded, which confirms the shallow-water nature of this deposit. E. crispum is a shallow-water, epifaunal species that harbours algal chloroplasts as adults and throughout ontogeny (Castignetti & Manley 1998). Castignetti & Manley (1998) also referred to this species as having typical epifaunal morphology. Mendes et al. (2004) highlighted that this species typically occurs in muddy sandy substrates. Haynes (1981) suggested that Elphidium occurs in inner shelf marine environments together with Cibicides. Elphidium is particularly abundant in marginal marine environments and the turbulent zone (Haynes 1981). Keeled species (such as E. crispum) usually occur in normal marine conditions (Haynes 1981). The dominant foraminifera by a considerable margin, however, in both the ancient and modern deposits examined in this study, is the genus Cibicides. This species typically occurs, according to Haynes (1981), in muddy sand or silt facies,
80
G. K. GILLMORE & N. MELTON
whereas Sen Gupta (1971) pointed out that C. lobatulus is not found living on substrate that is finer than sand, with a preference for sandy, gravelly or rocky bottoms. Sen Gupta (1971) went so far as to suggest that the distribution of this species is dependent on grain size of the substrate rather than other environmental factors. The morphology of the test can be seen to vary (noted particularly in examples from the modern beach), caused by the character of the substrate on which this sessile species lives, and by variation in the chamber numbers that the schizont develops as it grows (Sen Gupta 1971). C. refulgens forms a C. refulgens dominated benthic association noted in the shelf and deep sea between Antarctica and the subtropical converge at c. 40–508S (Murray 1991). Here temperatures are of the order of 21.9 to þ1.5 8C, with coarse sand and gravel bottom conditions, and water depths of 136 –950 m, and C. refulgens occurs together with species such as C. lobatulus. C. refulgens is also found around Norway at depths of 215– 255 m and in the Biscay area at depths of 230 m. Murray (1991) noted C. refulgens and C. lobatulus as having similar modes of life, being attached to, for example, the substrate, with C. refulgens being a herbivore, a suspension feeder and parasitic. C. refulgens is a common additional species in a C. lobatulus association (which includes T. angulosa, T. saggitula, E. repandus and G. rudis) at temperatures of 0–30 8C, with gravel, sand, mud and seaweed substrates, at depths of 0–900 m and salinities of 32 –36.5‰, in regions including those around Spitsbergen, the Barents Sea, Tromsø (Norway), Møre (Norway), Orkney –Lincolnshire, the Iceland –Faeroe Ridge, west of Scotland, the Celtic Sea, the English Channel, northern Biscay, Portugal and NW Africa. Between Scotland and the Shetlands there are areas of temperate carbonate accumulation that consist of biogenic sand under strong wave action (Murray 1991). The Fair Isle area is influenced by strong currents and a C. lobatulus assemblage has been noted living on hydroids (Murray 1991). West of Scotland on the shelf region this species can be found living on bivalve shells (these are certainly abundant in the modern beach sediment from the Shetlands). One significant difference in the microfauna between the modern beach sand and the ancient sands is the lack of agglutinated benthonic foraminifera in WV05 1 and 2. WV05 BS, on the other hand, contains multiserial forms (Gaudryina rudis) together with biserial flattened textulariids (Textularia saggitula group of Murray 1971). This may be a preservational feature of the older sands, but Gaudryina is typical of warm temperate conditions in shelf to upper bathyal depths (50–500 m), whereas Textularia occurs in cold to warm
conditions, at depths of 0–500 m (lagoonal to bathyal; Murray 1991). One significant caveat that should be applied when interpreting the foraminiferal assemblages in both the modern beach and the ‘fossil’ sedimentary material is the fact that erosional, transportational and depositional processes will entrain and deposit foraminifera as grains. These grains may be derived from older sediments, and indeed reworking of older materials in Quaternary sediments is perhaps more common than is currently recognized (J. Whittaker, pers. comm.). At this site, there is some glacial material, which could be a possible source for derived fossils, although the similarity between the modern beach fauna and the sands in section would suggest that the foraminifera are probably not derived.
Summary of sedimentary and foraminiferal analysis Palaeoenvironmental analysis suggests that the sand beds were derived from beach sands that may have experienced wind action, having a narrower grainsize distribution than the modern beach, generally a finer grain size and better sorting. The faunal elements of these sands bear a striking similarity to modern local beach sediments, but are less diverse and lack agglutinated benthonic forms. The microfauna of the archaeological section sands fits into a shallow-marine interpretation, with the organisms living on a sandy, gravelly or rocky substrate (probably the last). Given the variety of grain shapes and broken shell fragments, this is likely to be a series of storm deposits. The preservation of the foraminifera suggests that these are not typical of wind-blown sediments.
Conclusions: implications for human settlement The lower midden, which, on the evidence of the section excavated in 2004–2005, continued in use from c. 4000 cal BC until around 3500 cal BC, contained a succession of selected species of shellfish that has been described as exhibiting the characteristics of resource depletion consistent with a colonization event (Anderson 2007, p. 198). This interpretation appears to accord with the evidence from the Loch of Gards pollen core (Edwards et al. 2009), where two distinct phases of tree cover reduction suggesting colonization events were noted, the first around 6000 cal BC and the second coinciding with the start of deposition of the lower midden around 4000 cal BC.
EARLY NEOLITHIC SANDS, SHETLAND ISLANDS
The fragments of pottery and bones from domesticated species that were found in small numbers on and above the trampled activity surface that was dated to 3700–3600 cal BC push back the date for the appearance of ceramics in Shetland, making it closer to that of mainland Scotland, which Ashmore (2004, p. 130) has stated occurred c. 3800–3700 cal BC. They indicate that additional cultural contacts occurred sometime shortly after 3700 cal BC. Isotopic studies of human skeletal material have indicated that a rapid change in diet accompanied the shift from Mesolithic hunter–gatherer to Neolithic agriculturalist lifestyles in the west of Scotland (Schulting & Richards 2002). This contrasts with the evidence from the uppermost deposits in the lower midden at West Voe in which mussels are present and, especially, from the upper midden, which was dated to 3500– 3250 cal BC and is composed entirely of cockles, indicating a continued reliance on marine resources after humans returned to the site well after adoption of agriculture in Shetland’s marginal island environment. This continued use of marine resources into the Neolithic accords with recently published evidence from Prestatyn in north Wales (Schulting & Gonzalez 2007), where, in a local setting in which a number of Mesolithic– Neolithic transition shell middens have been found, isotope analyses of a female skeleton, radiocarbon dated to 3750– 3525 cal BC, have indicated a continued use of marine resources in a mainly terrestrial-based diet (Schulting & Gonzales 2007). This study has shown that, at around 3500 cal BC, the settlement process at West Voe was disrupted by the rapid deposition of a series of sands. A similar sequence of sands has been observed in Quendale Bay, some 3.5 km to the NW of West Voe, implying that a series of fairly large-scale events had occurred, probably storm surges resulting from southerly gales. The inundation of the landscape by these sands would have had a significant impact on the newly established agricultural economy in the south of Shetland, forcing a temporary abandonment of areas of land that had been laid waste. These effects can be paralleled by the sand blows that occurred in the seventeenth century AD that devastated considerable areas of agricultural land in the south of Shetland and resulted in the loss of the Sinclair estate at Brow near Quendale (Kay 1680; Low 1879, pp. 184 –185). The sand deposition events also changed the offshore environment in the Early Neolithic, resulting in conditions favourable to the cockles that were exploited by the people who moved back into the area. It may well be that continued exploitation of marine resources was a necessity for these people,
81
as they battled to re-establish the soils necessary for the agricultural economy in the area. That subsequent reoccupation of the area seems to have been extensive; in addition to the upper midden and structure at West Voe there is the evidence for a local population in the form of the burials in the Sumburgh cist and in the earliest layers at Jarlshof (Hamilton 1956; Dockrill et al. 2005). It is this period that also seems to mark the widespread Neolithic settlement of the Shetland landscape, which Fojut (2006, p. 14) has suggested probably occurred around 3500 cal BC. Was this a separate colonization? This seems unlikely, especially given the evidence of a continued human presence in the vicinity of West Voe. The distinctive Neolithic cultures of the neighbouring Shetland and Orcadian archipelagos need to be addressed. The Shetland Isles were the most remote outpost of Neolithic Britain and it is not surprising that they developed their own cultural traits, but was this process exacerbated by a period of increased isolation at the beginning of the Neolithic, with the sands studied in this paper marking a fairly short-lived episode of increased storminess that would have made an already difficult sea crossing even more hazardous? Whether this was the case or not, reoccupation of the site at West Voe was relatively short-lived, for around 3250 cal BC it was again overwhelmed by sand deposition. Optically stimulated luminescence dates have shown that the sands overlying the upper midden extend throughout the remainder of the Neolithic and into the Bronze Age (Cullen 2005). Although there is evidence of Late Neolithic and Early Bronze Age activity, including the creation of plaggen soils and ard-cultivation, at the nearby sites at Jarlshof (Hamilton 1956; Dockrill et al. 2005) and Sumburgh (Downes & Lamb 2000, pp. 8–9), at West Voe there was no subsequent resettlement. Thanks go to J. Murray and J. Whittaker for comments on foraminiferal species recovered, and to E. K. Newman for technical assistance. This work was carried out with the support of Historic Scotland, NERC award NER/B/S/ 2003/00779, and the Prehistoric Society. The midden lies within the boundary of Sumburgh Airport and the authors are grateful to Scotair Properties and to N. Flaws, Airport Manager, for permission to carry out the investigations, and for providing indoor space for finds processing and storage.
References Anderson, A. 2007. Discussion: middens of the sea peoples. In: Milner, N., Craig, O. E. & Bailey, G. N. (eds) Shell Middens in Atlantic Europe. Oxbow, Oxford, 196 –202.
82
G. K. GILLMORE & N. MELTON
Ashmore, P. 2004. Absolute chronology. In: Shepherd, I. A. G. & Barclay, G. J. (eds) Scotland in Ancient Europe. Society of Antiquaries of Scotland, Edinburgh, 125– 136. Bennett, K. D. & Sharp, M. J. 1993. Holocene environmental history at Dallican Water, NE Mainland, Shetland. In: Girnie, J., Bennett, K. & Hall, A. (eds) The Quaternary of Shetland: a Field Guide. Quaternary Research Association, London, 77– 82. Bennett, K. D., Boreham, S., Sharp, M. J. & Switsur, V. R. 1992. Holocene history of environment, vegetation and human settlement on Cata Ness, Lunasting, Shetland. Journal of Ecology, 80, 241–273. Bennett, K. D., Boreham, S., Hill, K., Packman, S., Sharp, M. J. & Switsur, V. R. 1993. Holocene environmental history at Gunnister, north Mainland, Shetland. In: Girnie, J., Bennett, K. & Hall, A. (eds) The Quaternary of Shetland: A Field Guide. Quaternary Research Association, London, 83– 98. Beuselinck, L., Govers, G., Poesen, J., Degraer, G. & Froyen, L. 1998. Grain-size analysis by laser diffractometry: comparison with the sieve–pipette method. Catena, 32, 193–208. Blott, S. & Pye, K. 2001. GRADISTAT: A grain size distribution and statistics package for the analysis of unconsolidated sediments. Earth Surface Processes and Landforms, 26, 1237– 1248. Bondevik, S., Svendsen, J. L. & Mangerud, J. 1997. Tsunami sedimentary facies deposited by the Storegga tsunami in shallow marine basins and coastal lakes, western Norway. Sedimentology, 44, 1115–1131. Bondevik, S., Mangerud, J., Dawson, S., Dawson, A. & Lohne, Ø. 2005. Evidence for three North Sea tsunamis at the Shetland Islands between 8000 and 1500 years ago. Quaternary Science Reviews, 24, 1757–1775. Castignetti, P. & Manley, C. J. 1998. Benthonic foraminiferal depth distribution within the sediment in a modern ria. Terra Nova, 10, 37–41. Cullen, V. L. 2005. New evidence for the early occupation of Shetland. BSc dissertation, University of Bradford. Cumming, G. A. 1946. Flint core axe found on Fair Isle, Shetland. Proceedings of the Society of Antiquaries of Scotland, 80, 146 –148. Dockrill, S. J., Bond, J. M. & Batey, C. E. 2005. Jarlshof, Shetland: an economic, environmental and chronological reappraisal. Unpublished Data Structure Report, University of Bradford. Downes, J. & Lamb, R. (eds) 2000. Prehistoric Houses at Sumburgh in Shetlands: Excavations at Sumburgh Airport 1967–1974. Oxbow, Oxford. Edwards, K. J. & Moss, A. G. 1993. Pollen data from the Loch of Brunnawatt, West Mainland. In: Birnie, J. F., Gordon, J. E., Bennett, K. D. & Hall, A. M. (eds) The Quaternary of Shetland: Field Guide. Quaternary Research Association, London, 126–129. Edwards, K. J., Schofield, J. E., Whittington, G. & Melton, N. D. 2009. Palynology ‘On the Edge’ and the archaeological vindication of a Mesolithic presence? The case of Shetland. In: Finlay, N., McCartan, S., Milner, N. & Wickham-Jones, C. (eds) From Bann Flakes to Bushmills. Prehistoric Society Research Paper, 1, 113–123.
Firth, C. R. & Smith, D. E. 1993. Holocene sea level changes in Shetland. In: Birnie, J., Gordon, J. & Hall, A. (eds) The Quaternary of Shetland. Quaternary Research Association, London, 77–82. Fojut, N. 2006. Prehistoric and Viking Shetland. Shetland Times Ltd, Lerwick, 13– 27. Folk, R. L. 1966. A review of grain-size parameters. Sedimentology, 6, 73–93. Folk, R. L. & Ward, W. C. 1957. Brazos River Bar: a study in the significance of grain size parameters. Journal of Sedimentary Petrology, 27, 3– 26. Friedman, G. M. & Johnson, K. G. 1982. Exercises in Sedimentology. Wiley, New York. Gillmore, G. K., Kjennerud, T. & Kyrkjebø, R. 2001. The reconstruction and analysis of palaeowater depths: a new approach and test of micropalaeontological approaches in the post-rift (Cretaceous to Quaternary) interval of the northern North Sea. In: Martinsen, O. J. & Dreyer, T. (eds) Sedimentary Environments Offshore Norway—Palaeozoic to Recent. NPF Special Publication, 10, 365–381. Gillmore, G. K., Stevens, T. et al. 2011. Geoarchaeology and the value of multidisciplinary palaeoenvironmental approaches: a case study from the Tehran Plain, Iran. In: Wilson, L. (ed.) Human Interactions with the Geosphere: The Geoarchaeological Perspective. Geological Society, London, Special Publications, 352, 49– 67. Hamilton, J. R. C. 1956. Excavations at Jarlshof, Shetland. HMSO, Edinburgh. Haynes, J. R. 1981. Foraminifera. Macmillan, London. Hedges, J. W. & Parry, G. W. 1980. A Neolithic multiple burial at Sumburgh Airport, Shetland. Glasgow Archaeological Journal, 9, 15– 26. Jones, A., Duck, R., Reed, R. & Weyers, J. 2000. Practical Skills I Environmental Science. Prentice–Hall, London. Kay, J. 1680. A description of Dunrossness. In: Bruce, J. (ed.) Description of Ye Countrey of Zetland. James Skinner, Edinburgh. King, C. A. M. 1975. Techniques in Geomorphology. Arnold, London. Lambeck, K. 1993. Glacial rebound of the British Isles—I. Preliminary model results. Geophysical Journal International, 115, 941 –959. Lovell, J. P. B. 1979. Composition of Holocene sands of Mull and adjacent offshore areas: a study of provenance. Report of the Institute of Geological Sciences, 79/2. Low, G. 1879. A Tour Through the Islands of Orkney and Schetland. William Peace, Kirkwall. Melton, N. D. 2008. West Voe: a Mesolithic– Neolithic transition site in Shetland. In: Noble, G., Poller, T., Raven, J. & Verril, L. (eds) Scottish Odysseys: The Archaeology of Islands. Tempus, Stroud, 23–36. Melton, N. D. 2009. Shells, seals and ceramics: an evaluation of a midden at West Voe, Sumburgh, Shetland, 2004–2005. In: McCartan, S. & Woodman, P. (eds) Mesolithic Horizons: Papers Presented at the Seventh International Conference on the Mesolithic in Europe, Belfast 2005. Oxbow, Oxford, 184–189. Melton, N. D. & Nicholson, R. A. 2004. The Mesolithic in the Northern Isles: the preliminary evaluation of an
EARLY NEOLITHIC SANDS, SHETLAND ISLANDS oyster midden at West Voe, Sumburgh, Shetland, U.K. Antiquity, 78, 299. Melton, N. D. & Nicholson, R. A. 2007. A late Mesolithic– early Neolithic midden at West Voe, Shetland. In: Milner, N., Craig, O. E. & Bailey, G. N. (eds) Shell Middens in Atlantic Europe. Oxbow, Oxford, 94– 100. Mendes, I., Gonzalez, R., Dias, J. M. A., Labo, F. & Martins, V. 2004. Factors influencing recent benthic foraminifera distribution on the Guadiana shelf (Southwestern Iberia). Marine Micropalaeontology, 51, 171–192. Murray, J. W. 1971. An Atlas of British Recent Foraminiferids. Heinemann, London. Murray, J. W. 1991. Ecology and Palaeoecology of Benthic Foraminifera. Longman, Harlow. Murray, J. W. & Wright, C. A. 1974. Palaeogene Foraminiferida and Palaeoecology, Hampshire and Paris Basins and the English Channel. Special Papers in Palaeontology, 14. Mykura, W. 1976. British Regional Geology: Orkney and Shetland. HMSO, Edinburgh. Peltier, W. R., Shennan, I., Drummond, R. & Horton, B. 2002. On the postglacial isostatic adjustment of the British Isles and the viscoelastic structure of the Earth. Geophysical Journal International, 148, 443–475. Prothero, D. R. & Schwab, F. 1996. Sedimentary Geology: An Introduction to Sedimentary Rocks and Stratigraphy. Freeman, San Francisco, CA. Saville, A. 2000. Orkney and Scotland before the Neolithic Period. In: Ritchie, A. (ed.) Neolithic Orkney in its European Context. McDonald Institute of Archaeological Research, Cambridge, 91–100.
83
Schulting, R. & Gonzalez, S. 2007. ‘Prestatyn Woman’ reconsidered. In: Bell, M. (ed.) Prehistoric Coastal Communities: The Mesolithic in Western Britain. Council for British Archaeology Research Report, 149, 303– 305. Schulting, R. J. & Richards, M. P. 2002. The wet, the wild and the domesticated: the Mesolithic –Neolithic transition on the west coast of Scotland. European Journal of Archaeology, 5, 147–189. Sen Gupta, B. K. 1971. The Benthonic Foraminifera of the Tail of the Grand Banks. Micropalaeontology, 17, 69– 98. Swift, D. J. P. 1969. Evolution of the shelf surface, and the relevance of modern shelf studies to the rock record. In: Stanley, D. J. (ed.) The New Concepts of Continental Margin Sedimentation. American Geological Institute, Washington, DC. Tucker, M. E. 2001. Sedimentary Petrology, 3rd edn. Blackwell Science, Oxford. Tucker, M. E. 2003. Sedimentary Rocks in the Field, 3rd edn. Wiley, Chichester. Walford, N. 1995. Geographical Data Analysis. Wiley, Chichester. Warren, G. 2005. Mesolithic Lives in Scotland. Tempus, Stroud. Whittington, G. 1979. A sub-peat dyke on Shurton Hill, Mainland, Shetland. Proceedings of the Society of Antiquaries of Scotland, 109, 30–35. Whittle, A. W. R. (ed.) 1986. Scord of Brouster: an Early Agricultural Settlement on Shetland. Oxford University Committee Archaeological Monograph, 9. Wickham-Jones, C. 2006. Between the Wind and the Water: World Heritage Orkney. Windgatherer, Macclesfield.
d13C, d18O and deposition rate of tufa in Xiangshui River, SW China: implications for land-cover change caused by climate and human impact during the late Holocene ZAIHUA LIU1,2*, HAILONG SUN1, HONGCHUN LI3 & NAIJUNG WAN3 1
The State Key Laboratory of Environmental Geochemistry, Institute of Geochemistry, Chinese Academy of Sciences, 550002 Guiyang, China
2
Karst Dynamics Laboratory, Ministry of Land and Resources, 541004 Guilin, China
3
Department of Earth Sciences, National Cheng Kung University, 701 Tainan, Taiwan *Corresponding author (e-mail:
[email protected]) Abstract: Measurements of d13C and d18O values of the riverine tufa samples dated by the accelerating mass spectrometry-14C method have been used for discussion of land-cover change in the catchment area of the Xiangshui River, SW China during the late Holocene. The results show that tufa was deposited in this short river only between c. 4280 and c. 110 years BP. Based on the characteristics of d13C values of the complete tufa profile in the river, three stages of land-cover conditions in the headwater area could be identified. The earliest, Stage I, contained the most extensive vegetation cover with mainly C3 plants, as shown by the lightest d13C value, whereas the latest, Stage III, had the least land cover, reflected by the heaviest d13C value. By comparison of speleothem and historical records, it was found that the land cover in Stage I was controlled mainly by climate change, whereas the land-cover changes in the later two stages were most probably related to major human disturbance (land use), especially since the Qin Dynasty and Ming Dynasty, in the headwater area of the river.
Tufa and travertine appear to be significant indicators of palaeoenvironmental and palaeoclimatic changes (Thorpe et al. 1980; Kronfeld et al. 1988; Pazdur et al. 1988, 2002; Goudie et al. 1993; Andrews et al. 1994; Pentecost 1995, 2005; Ford & Pedley 1996; Frank et al. 2000; Auler & Smart 2001; Pentecost & Zhang 2001; Liu et al. 2003, 2006; Kano et al. 2004; Andrews 2006; Candy & Schreve 2007 Li et al. 2008; Valero Garce´s et al. 2008). These deposits reflect the combined effects of karst processes, controlled to a great extent by climatic factors (e.g. temperature and humidity) and biological factors (such as vegetation and soil). Freshwater tufas in the continental realm, deposited by physicochemical and/or biochemical processes, are considered to be reliable recorders of palaeoclimatic and environmental change (Andrews 2006). Stable isotopes such as 18O and 13C, trace elements, deposition rates, and frequency distribution of U/Th and 14C dates, have frequently been used as tools for analysis of sedimentological processes and reconstruction of palaeoclimatic conditions (Soligo et al. 2002; Ihlenfeld et al. 2003; Kele et al. 2006), as well as for stratigraphic purposes. Measurements of d18O and d13C in tufa samples dated by the 14C method have been used to reconstruct Holocene climatic changes in
southeastern Poland and Eastern India (Pazdur et al. 2002). However, very few studies have addressed the use of tufa as a tool for reconstruction of landcover change (Preece & Bridgland 1999). This information is very important to understanding the response of land cover to climate change and/or human activities, and thus to sustainable land-use management in a catchment area. In this paper, we have attempted to make the first investigation of land-cover change in the headwater area of the Xiangshui River, Libo, SW China through the stable isotopic signatures and deposition rates of the riverine tufa. To reconstruct the time record of palaeoenvironmental conditions of sedimentary processes, it is necessary to reconstruct the time scale of tufa deposition. Radiocarbon and U/Th dating methods are often used for this (Srdoc et al. 1980; Pazdur & Pazdur 1986; Krajcar-Bronic et al. 1992; Szabo et al. 1996; Genty & Massault 1997; Genty et al. 1999, 2001; Meyrick & Karrow 2007); however, the high-precision accelerating mass spectrometry (AMS)-14C dating method (Yi et al. 2004) was adopted in this study because of the presence of occasional clasts in the tufa, which could cause problems when using the U/Th dating method (Garnett et al. 2004).
From: Wilson, L. (ed.) Human Interactions with the Geosphere: The Geoarchaeological Perspective. Geological Society, London, Special Publications, 352, 85– 96. DOI: 10.1144/SP352.7 0305-8719/11/$15.00 # The Geological Society of London 2011.
86
Z. LIU ET AL.
Regional setting The Xiangshui River (Fig. 1), only 1 km long, is the last surface part of the Huanghou Underground River, which thereafter reaches the Zhang River (Fig. 2). Bounded by Wolong Dam, the catchment area of the Huanghou Underground River can be divided into two parts: the small (c. 10%) eastern recharge area with Xiangshui karst virgin forest cover on cone karst (Fig. 1), and the large (c. 90%) western headwater area characterized by the presence of strong karst rocky desertification (KRD: a process of land degradation involving vegetation clearing, serious soil erosion, extensive exposure of bedrock, and drastic decrease in soil productivity under major human impacts on the vulnerable karst eco-geoenvironment; Wang et al. 2004) mainly caused by agricultural land use (Figs 2 & 3). The headwater area, 700 –900 m in elevation, is in Dushan County of Guizhou Province, SW China. This area has a subtropical Asian monsoonal climate, with a mean annual air temperature of 16 8C and annual precipitation of 1346 mm. The rainy season, with 75% of the annual precipitation, is from
May to October, and the dry season is from November to April. The bedrock in the area is mainly (.95%) Carboniferous and Permian limestones. The lower eastern recharge area, the Xiangshui River Valley, is located in the southwestern part of Libo County of Guizhou Province, SW China, with elevations of 300–700 m (Fig. 2). With beautiful karst mountains, waters, forests, caves and tufa terraces, this area is one of the key sights in Maolan Karst Forest Natural Preserve, Libo (part of the natural World Heritage site ‘South China Karst’ declared by UNESCO in 2007). The annual precipitation (1750 mm) here is significantly higher than that (1346 mm) in the headwater area, perhaps owing to the microclimatic effect of the largest continuous virgin forests on cone and tower karst areas in the world (.200 km2). The main passage of the Huanghou Underground River is 56.5 km long, and its catchment area is 445 km2. The discharge of the river ranges from 1.55 to 60 m3 s21 (Li et al. 1988). Although the Xiangshui River is only 1 km long, its water chemistry presents the final stage of mixing of waters from the headwater area and the lower recharge
Fig. 1. The tufa sampling profile with surrounding virgin karst forests in the Xiangshui River valley, Libo, Guizhou, SW China.
d13C OF TUFA AND LAND COVER CHANGE
Fig. 2. Hydraulic connections between the Xiangshui River, Huanghou Underground River, Wolong reservoir, Yuanyang Lake and Zhang River.
Fig. 3. The karst rocky desertification in the headwater area of Xiangshui River, Libo, Guizhou, SW China.
87
88
Z. LIU ET AL.
area (Fig. 2). However, the headwater area is much larger than the lower recharge area (100:1). Therefore, the tufa deposits formed in the Xiangshui River primarily record changes in the water chemistry and isotopic composition of the large headwater area. After the Xiangshui River joins the main river, the Zhang River, the water chemistry is dominated by waters from a much larger catchment basin.
Methods Analysis of river water samples River water samples were analysed to understand their potential for tufa deposition at present. Temperature, pH and specific conductivity of water were measured in situ with a WTW (WissenschaftlichTechnische-Werkstaetten) Technology MultiLine 340i multi-parameter meter. The meter was calibrated using pH (7 and 10) standards before use (Liu et al. 2007b). In situ titration was used to measure seasonal 2þ variations in the [HCO2 3 ] and [Ca ] content of water, using the Aquamerck Alkalinity Test and Hardness Test. The resolutions are 6 and 1 mg l21, respectively (Liu et al. 2007b). To understand the general chemistry of other major ions in the area and the difference in 2þ [HCO2 3 ] and [Ca ] between in situ and laboratory measurements, river water samples at Wolong Dam and the tufa outcrop were collected by syringes with 0.45 mm Minisartw filters and analysed in the laboratory at a seasonal interval for one hydrological year. The analytical methods used were standard þ titration for HCO2 3 , atomic absorption for K and Naþ, titration with EDTA for Ca2þ, Mg2þ and 2 (Liu et al. SO22 4 , and the Mohr titration for Cl 2007b). The full hydrochemical datasets, including in situ measured temperature, pH, and concentrations of Ca2þ and HCO2 3 , and laboratory measured concentrations of Kþ, Naþ, Mg2þ, Cl2 and SO22 4 , were processed using the program WATSPEC (Wigley 1977), which calculates CO2 partial pressure (PCO2) and calcite saturation index (SIc). SIc is calculated from 2þ (Ca )(CO2 3 ) SIc ¼ log Kc where activities are denoted by parentheses, and Kc is the temperature-dependent equilibrium constant for calcite dissociation. If SIc . 0, water is supersaturated with respect to calcite, and tufa may deposit; if SIc , 0, water is aggressive to calcite, and tufa dissolution could happen; and if SIc ¼ 0, equilibrium is reached.
Tufa sampling and measurements To take fresh tufa samples, the surface part of the tufa profile was removed. It was found that the tufa has a horizontal laminar structure, indicating a pool environment when the tufa was deposited. It was also found by micrography that there has been no recrystallization or post-depositional disturbance of the deposits. We sampled the tufa outcrop with a spatial interval of c. 5 cm from bottom to top (Fig. 1). A total of 52 tufa samples (c. 30 mg in weight and about 5 mm in diameter for each sample) for stable isotope analysis were collected. A c. 10 mg powdered sample of each was put into a reaction vial of a Kiel-III automatic carbonate device connected to a Finnigan DELTA XP Plus isotope ratio mass spectrometer (IRMS) at National Cheng Kung University, Taiwan, to react with pure H3PO4 at 70 8C. The CO2 generated was purified and sent to the IRMS for d18O and d13C analyses. The results were reported against the VPDB standard at 25 8C. The IRMS was calibrated with NBS standards of NBS-19 (limestone), NBS-18 (carbonatite) and LSVEC (lithium carbonate). The IRMS working condition was monitored by a working standard, Ultiss (limestone), for every seven samples. Based on the measurements of NBS-19 (n ¼ 590), the external analytical uncertainty is 0.045‰ for d13C and 0.090‰ for d18O (1s standard deviation). Based on the depositional features, seven samples out of the 52 samples along the tufa profile (Fig. 1) were chosen for determination of the ages of the tufa deposits. Sample preparation was carried out in the School of Archaeology and Museology of Peking University following a standard procedure. A CN analyser was used for CO2 preparation and CO2 was reduced to graphite by H2 and using Fe powder as a catalyser. Samples were analysed with a new compact AMS system at the Institute of Heavy Ion Physics of Peking University (Liu et al. 2007a).
Results and discussion The hydrochemistry of present-day river water and its potential for tufa deposition Table 1 lists the hydrochemical results for Xiangshui River water at the tufa site and its source waters, including Wolong Reservoir, Yuanyang Lake and the Huanghou Underground River, in wet (July and September) and dry (March and November) months. Only minor differences (,10%) were found between in situ and laboratory measurements 2þ of [HCO2 3 ] and [Ca ]. Based on Table 1, the hydrochemical characteristics can be summarized as follows.
Table 1. Hydrochemical and carbon isotopic characteristics of the Xiangshui River water and its source water Sampling site, year and month
Water temperature (8C)
pH
Wolong River 2002.07 2002.09 2002.11 2003.03 Mean
19.7 – 19.1 19.6 19.5
Yuanyang Lake 2002.07 2002.09 2002.11 2003.03 Mean
Major ions (mg l21) Kþ
Naþ
Ca2þ
Mg2þ
Cl2
SO22 4
HCO2 3
7.52 7.56 7.53 8.00 7.65
0.34 0.44 0.34 0.50 0.41
0.62 0.73 0.42 0.70 0.62
63.16 63.06 64.21 63.18 63.40
0.27 2.59 2.00 2.27 1.78
5.21 3.48 2.61 3.55 3.71
13.76 13.64 6.91 4.98 9.82
170.42 191.21 191.21 191.02 185.97
21.3 – 17.4 21.8 20.2
7.82 8.09 8.04 8.44 8.1
0.51 0.54 0.54 0.47 0.52
0.68 0.95 0.63 0.75 0.75
63.16 64.01 65.86 58.60 62.91
0.27 2.01 0.50 2.02 1.20
4.34 3.48 2.61 4.43 3.72
10.59 14.77 6.91 7.97 10.06
166.27 174.58 180.82 164.36 171.51
Underground river 2002.07 2002.09 2002.11 2003.03 Mean
19.9 – 18.9 18.4 19.1
8.16 8.16 8.06 7.90 8.07
0.28 0.30 0.37 0.37 0.33
0.45 0.52 0.44 0.42 0.46
62.28 59.74 62.98 59.02 61.01
0.27 2.30 1.25 2.02 1.46
6.08 2.61 2.61 3.55 3.71
11.64 12.50 7.89 8.97 10.25
164.19 168.35 174.58 166.58 168.43
Tufa profile 2002.07 2002.09 2002.11 2003.03 Mean
21.0 – 18.5 19.1 19.5
8.33 8.18 8.19 8.12 8.20
0.28 0.29 0.29 0.23 0.27
0.41 0.52 0.34 0.02 0.32
62.28 59.27 61.74 65.67 62.24
0.80 1.15 2.25 3.78 2.0
5.21 2.61 2.61 3.55 3.50
9.53 14.77 7.89 6.97 9.79
170.42 157.95 180.82 202.12 177.83
d13CDIC (‰)
SIc*
pCO†2 (ppmv)
0.05 – 0.11 0.57 0.24
4680 – 5130 580 3463
0.36 – 0.57 0.93 0.62
2340 – 1450 1860 1883
0.66 – 0.58 0.37 0.54
1020 – 1350 540 970
0.56 – 0.70 0.71 0.66
590 – 1020 740 783
210.73
– 11.41
210.3
d13C OF TUFA AND LAND COVER CHANGE
210.05
*Calcite saturation index in water (SI ¼ log IAP/K, where IAP is ionic activity product and K is the calcite equilibrium constant). If SI . 0, supersaturation occurs and travertine may be deposited; if SI , 0, water is aggressive to calcite; and if SI ¼ 0, equilibrium is reached. † Calculated CO2 partial pressure of water by WATSPEC (Wigley 1977).
89
90
Z. LIU ET AL.
2þ (1) HCO2 are respectively the major 3 and Ca anion and cation in the waters, and all waters are HCO3 –Ca type, showing the control of lithology (limestone) on hydrochemical type. (2) There is apparently no seasonal change in 2þ the concentrations of HCO2 3 and Ca , as a result of the great damping effect of the large karst aquifer system. (3) Compared with the Guilin Experimental Site under a similar subtropical climate and similar lithology (Liu et al. 2004b), the CO2 partial pressure 2þ in (PCO2) and concentrations of HCO2 3 and Ca the Xiangshui River are much lower. The low 2þ are PCO2 and concentrations of HCO2 3 and Ca due to the strong karst rocky desertification (KRD) in the headwater area (Figs 2 & 3) at the present day. KRD decreases vegetation and soil coverage, hence reducing soil CO2 production and limestone dissolution, which is evidenced by the higher d13C values of the dissolved inorganic carbon (d13CDIC, Table 1) than those of waters in the areas with virgin forest cover (213 to 215‰, Zhou et al. 2002) and high canopy cover percentage (Finlay 2003). (4) From upstream to downstream, the concen2þ are stable, showing trations of HCO2 3 and Ca that calcium carbonate (tufa) is not being deposited at present. Although the SIc in the river water was greater than zero, there was no apparent tufa deposition in the river. This is confirmed by the observation that there are no encrustations on present plant remains in the river bed. The reason for this is two-fold (Dreybrodt 1988): the low concen2þ as stated above, and trations of HCO2 3 and Ca the low SIc values (generally less than 0.80, Table 1). Many researchers have found that calcite does not precipitate until SIc . 0.80, because of lack of free energy to create new surface areas, unavailability of reactive calcite to act as nucleation 2þ and organic sites, and inhibition of PO32 4 , Mg ligands (Jacobson & Usdowski 1975; Reddy 1977; Suarez 1983; Troester & White 1986; Herman & Lorah 1987; Dreybrodt et al. 1992; Lebron & Suarez 1996). It has been suggested that an SI . 0.8 is a critical factor necessary to induce inorganic carbonate precipitation (Herman & Lorah 1987; Merz-Preiss & Riding 1999). All the above hydrochemical characteristics indicate that the small karst virgin forest in Xiaoqikong (Figs 2 & 3) has little influence on the hydrochemistry of the Xiangshui River. In other words, the presence or absence of tufa deposits in the Xiangshui River is mainly an indication of environmental change in the large headwater area.
Tufa profiling The height difference of the Xiangshui River between the source water (power station No. 4)
and the Zhang River is about 40 m in a length of 1 km flow path (Fig. 2). Along the flow path are tens of small waterfalls. However, tufa deposits are found only starting at a distance of at least c. 300 m from the source water, and are mainly located at the downstream waterfalls. The tufa profile studied is in the lowest part of the Xiangshui River (Fig. 1), just before it empties into the Zhang River (Fig. 1). The tufa profile here has a thickness of c. 255 cm, which is the thickest deposit along the river (Fig. 1). The tufa has a horizontal laminar structure, indicating a pool environment when it was deposited. At the base of the tufa profile, fluvially deposited bedrock (limestone) pebbles separate the bedrock and the tufa deposit, and the top of the tufa is under water. That means that the oldest tufa has been uncovered, and that this tufa profile might be the most complete one, including both the oldest and the newest tufa deposits. From the colour and compactness, three main stages of tufa deposition (corresponding to three stages of land cover; see below) were distinguished. The first stage (Stage I), with a yellowish colour and porous structure, is from the base (0 cm) to c. 35 cm; the second (Stage II), from 35 to c. 140 cm, has a brown colour and dense structure; and the third (Stage III), from c. 140 cm to the top (255 cm), has a red –brown colour and dense structure. It seems that tufa is no longer being deposited. Moreover, dissolutional features (Fig. 4) were seen on the top of the tufa profile. These observations suggest that the tufa deposits are the remains of ancient environments.
Ages and deposition rates of the tufa deposits Dating tufa or speleothems using 14C is difficult because of the dead carbon effect of carbonate rocks and endogenic CO2 (Srdoc et al. 1980; Pazdur & Pazdur 1986; Krajcar-Bronic et al. 1992; Genty & Massault 1997; Genty et al. 1999, 2001; Kattan 2002; Meyrick & Karrow 2007). As it is not possible to accurately determine the ‘dead carbon fraction’ (DCF) in tufas of this type without dating coeval organic matter (e.g. plant remains) that grew in equilibrium with atmospheric (compared with aqueous) CO2, the chronology has no absolute basis. However, many researchers (Srdoc et al. 1980; Pazdur & Pazdur 1986; Genty & Massault 1997) have found that DCF values generally amount to about 15% for both tufa and speleothem. That means that an apparent age (defined here as the difference between corresponding 14C dates of carbonate and associated organic matter) of 1300 years can be used to correct the 14C dating results of tufa. It will be seen that this treatment is acceptable because the corrected tufa ages
d13C OF TUFA AND LAND COVER CHANGE
91
Fig. 4. The dissolutional features of tufa in the Xiangshui River bed, Libo, SW China, showing that no tufa deposition is occurring at present, owing to the karst rocky desertification in the headwater area of Xiangshui River.
correspond well to results obtained from historical records. Table 2 lists the AMS-14C dating results and the corrected ages of the tufa deposits from the Xiangshui River profile. It can be seen that deposition of the tufa started c. 4280 years BP and ended c. 110 years BP. We intend to use this latter age to represent the modern period. From the thickness and age determinations, the deposition rates of each deposition stage were obtained: c. 0.02, 0.08 and 0.19 cm a21 for stages I, II and III, respectively (Fig. 5a). The overall deposition rate is 0.06 cm a21.
Figure 5b and Table 3 show the variations in d13C and d18O of the fossil tufa deposits along the tufa profile formed during the period from 4280 to 110 years BP. The details of these variations and their implications for land-cover change are discussed below.
the tufa samples from Stage I (Fig. 5b, Table 3), possibly indicating the greatest land cover. In Stage I, vegetation was mainly C3 plants, and the soils were mature and contributed isotopically negative carbon from the decay of organic matter (Andrews et al. 1994). According to Dykoski et al. (2005), there was a mid- to late Holocene stepwise decrease in Asian monsoon intensity at about 3550 + 59 years BP. Therefore, the change in d13C trends from decreasing to increasing during Stage I (Fig. 5b) may be related to this climate change. During Stage I, the mean tufa d18O value of 29.33‰ is lower than that (27.8‰) of the stalagmite D4 found in nearby Dongge Cave (680 m above sea level; Dykoski et al. 2005), possibly showing the effect of elevation on oxygen isotopes. The variation of 1‰ in tufa d18O is also lower than that (2.0‰) of the stalagmite (Fig. 5b, Table 3; Dykoski et al. 2005), possibly indicating the greater buffering effect of the large headwater area than that of the small stalagmite recharge area on the stable oxygen isotopic composition.
Stage I (4280–2130 years BP). The lowest d13C values, from 210.52‰ to 211.56‰, are found for
Stage II (2130–680 years BP). There was an abrupt increase in d13C from Stage I to Stage II (Fig. 5b),
Tufa d13C and d18O values and their implications for land-cover change
92
Z. LIU ET AL.
Table 2. AMS-14C dating results and the corrected ages of the tufa deposits from the Xiangshui River profile, Libo, Guizhou, SW China Lab. code BA06566 BA06567 – BA06568 BA06569 BA06571 BA06572 BA06573 –
Sampling code
Distance to tufa bottom (cm)
AMS-14C ages (14C years BP)*
Calibrated age (cal years BP)†
Corrected ages accounting for DCF effect (years BP)‡
LX-01 LX-03 – LX-12 LX-19 XL-40 XL-49 XL-51 XL-52
0–1 10 35 55 90 195 240 250 255
4750 + 35 4205 + 35 – 2300 + 35 2085 + 35 1990 + 35 1580 + 30 1535 + 45 –
5580 4820 3400 2330 2050 1930 1520 1430 –
4280 3520 2100§ 1030 750 630 220 130 110k
A 14C half-life of 5568 years was used when calculating the ages. *The year relative to 1950. † Calibrations were carried out following Reimer et al. (2004) (IntCal04). ‡ DCF (dead carbon fraction) ¼ 15% (Srdoc et al. 1980; Genty & Massault 1997). § Liu et al. (2004a). k Extrapolated value according to deposition rate.
Fig. 5. Variations in deposition rates (a), and carbon and oxygen isotopes (b) of the tufa profile in the Xiangshui River valley, Libo, Guizhou, SW China, showing that there were three stages of tufa deposition at the site, which corresponded to the land-cover change from virgin forest to karst rocky desertification.
d13C OF TUFA AND LAND COVER CHANGE
93
Table 3. Stable carbon and oxygen isotopic characteristics of tufa deposits at different stages in the Xiangshui River, Libo, Guizhou, SW China Isotopic composition Stage I (4280 –2130 years BP) Stage II (2130 –680 years BP) Stage III (680 –110 years BP) The whole duration (4280 –110 years BP)
D13C (‰, PDB)
D18O (‰, PDB)
211.09* (211.56 to 210.52)† 28.72 (29.22 to 28.25) 27.96 (28.64 to 27.30) 28.75 (211.56 to 27.30)
29.33 (29.77 to 28.77) 29.22 (210.64 to 28.41) 29.41 (29.83 to 28.76) 29.32 (210.64 to 28.41)
*Mean value. † Minimum– maximum.
indicating notable land-cover change, from richly vegetated soil to karst rocky desertification. According to Dykoski et al. (2005), there was no remarkable climate change during this period. Therefore, the land-cover change was very probably due to significant human activities (large-scale lumbering and farming) in the headwater area of the Xiangshui River starting in the Qin Dynasty (2171– 2157 years BP) (Han 2006; Zhou 2006). The larger variation in d18O during Stage II (Fig. 5b) may be related to land-cover changes in the headwater area, which controlled the evapotranspiration in the area. Stage III (680 –110 years BP). The highest d13C values, greater than 29.0‰, are found in the tufa samples from Stage III (Fig. 5b, Table 3), possibly indicating the least land cover, where karst rocky desertification (i.e. soils are thin and poorly vegetated, or absent) might occur and contribute isotopically heavy carbon from the dissolution of carbonate rock and/or the atmosphere (Andrews 2006). According to historical records (Han 2006; Zhou 2006), not only have there been large-scale lumbering and farming activities in the area since the Ming Dynasty (582– 306 years BP), but also C4 plants, mainly maize, were introduced and have been cultivated in the area since that time. Therefore, the large-scale cultivation of maize, with higher d13C (Deines 1986), may explain why the tufa d13C in Stage III is higher than that in Stage II, when maize had not been introduced into the catchment area. During Stage III, there is a decreasing trend in tufa d18O (Fig. 5b), which is similar to the results obtained by Dykoski et al. (2005), showing some increase in the Asian monsoon intensity. This climate change may be the cause for the decreasing trend in tufa d13C during Stage III (Fig. 5b) although land use and land-cover change raised the stage’s tufa d13C values as a whole. Therefore, the tufa d13C values were shown to be controlled by the combined action of land-cover change and climate change.
Conclusions River water samples were taken in an attempt to understand their potential for tufa deposition at present. It was found that tufa is not being deposited now because of low concentrations of Ca2þ and HCO2 3 , which result from the lower soil CO2 production induced by karst rocky desertification in the headwater area, and the lower calcite saturation index, which is related to the short river flow path and the lower CO2 pressure difference between the river water and the atmosphere (both being disadvantageous to reaching the critical level of calcite supersaturation by degassing of CO2 from the river water to the atmosphere). Measurements of d18O and d13C in the tufa samples dated by the AMS-14C method, along with deposition rates, have been used to reconstruct the land-cover (vegetation and soil) change in the headwater area of the Xiangshui River during the Holocene. Results show that there are no tufa deposits at all at the site dating to before c. 4280 years BP and after c. 110 years BP. Tufa was deposited mainly during the period from c. 4280 years BP to c. 110 years BP. Based on the characteristics of d13C, d18O and deposition rates of the calcareous tufa profile, three stages (Stage I: 4280–2130 years BP; Stage II: 2130–680 years BP; Stage III: 680–110 years BP) of land cover could be distinguished. Stage I was the period of the greatest land cover, with vegetation consisting mainly of C3 plants, resulting in the lowest d13C values and deposition rate. Because it has the least amount of land cover (or more karst rocky desertification with maize cultivation), Stage III shows the highest d13C and deposition rate, and Stage II attests to intermediate land cover. In the modern period, the absence of tufa deposits is mainly related to anthropogenically induced karst rocky desertification in the headwater area of the river, which has resulted in a decrease in soil CO2 production in the area, and thus diminished the supply of Ca2þ and HCO2 3 for tufa deposition.
94
Z. LIU ET AL.
The findings above have both theoretical and practical implications. Theoretically, the mechanisms for tufa deposition have to be considered more extensively; that is, that tufa deposition is controlled not only by climatic factors (temperature, humidity), but also by biological factors (soil, vegetation), the latter being influenced by both climate and anthropogenic activities (such as land use). This contributes critical insights into the interpretation of environmental proxies in ancient tufa. Practically, these findings will provide a scientific basis for managing land use in the large headwater area of the Xiangshui River in favour of the protection of karst features and landscape at the new natural World Heritage site, and also have implications for understanding tufa decline in Europe during the late Holocene (Goudie et al. 1993; Pentecost 2005). This work was supported by the Hundred Talents Program of the Chinese Academy of Sciences, the Foundation of the Chinese Academy of Sciences for Innovation (Grant No. kzcx2-yw-306), the National Natural Science Foundation of China (Grant No. 40872168), and the Ministry of Science and Technology of China (Grant No. 2005DIB3J067). Special thanks are given to L. Wilson and the two anonymous reviewers for their comments and language modifications, which considerably improved the paper.
References Andrews, J. E. 2006. Palaeoclimatic records from stable isotopes in riverine tufas: synthesis and review. EarthScience Reviews, 75, 85–104. Andrews, J. E., Pedley, H. M. & Dennis, P. F. 1994. Stable isotope record of palaeoclimatic change in a British Holocene tufa. Holocene, 4, 349–355. Auler, A. S. & Smart, P. L. 2001. Late Quaternary paleoclimate in semiarid Northeastern Brazil from U-Series dating of travertine and water-table speleothems. Quaternary Research, 55, 159–167. Candy, I. & Schreve, D. 2007. Land–sea correlation of Middle Pleistocene temperate sub-stages using highprecision uranium-series dating of tufa deposits from southern England. Quaternary Science Reviews, 26, 1223–1235. Deines, P. 1986. The isotopic composition of reduced organic carbon. In: Fritz, P. & Fontes, J. C. (eds) Handbook of Environmental Geochemistry: The Terrestrial Environment. Elsevier, Amsterdam, 331–406. Dreybrodt, W. 1988. Processes in Karst Systems. Springer Series in Physical Environment. Springer, Heidelberg. Dreybrodt, W., Buhmann, D., Michaelis, J. & Usdoswski, E. 1992. Geochemically controlled calcite precipitation by CO2 outgassing: Field measurements of precipitation rates to theoretical predictions. Chemical Geology, 97, 287– 296. Dykoski, C. A., Edwards, R. L. et al. 2005. A highresolution, absolute-dated Holocene and deglacial
Asian monsoon record from Dongge Cave, China. Earth and Planetary Science Letters, 233, 71– 86. Finlay, J. C. 2003. Controls of streamwater dissolved inorganic carbon dynamics in a forested watershed. Biogeochemistry, 62, 231–252. Ford, T. D. & Pedley, H. M. 1996. A review of tufa and travertine deposits of the world. Earth-Science Reviews, 41, 117–175. Frank, N., Braum, M., Hambach, U., Mangini, A. & Wagner, G. 2000. Warm period growth of travertine during the last interglaciation in southern Germany. Quaternary Research, 54, 38– 48. Garnett, E. R., Gilmour, M. A., Rowe, P. J., Andrews, J. E. & Preece, R. C. 2004. 230Th/234U dating of Holocene tufas: possibilities and problems. Quaternary Science Reviews, 23, 947–958. Genty, D. & Massault, M. 1997. Bomb 14C recorded in laminated speleothems—part 1: dead carbon proportion calculation. Radiocarbon, 39, 33–48. Genty, D., Massault, M., Gilmour, M., Baker, A., Verheyden, S. & Kepens, E. 1999. Calculation of past dead carbon proportion and variability by the comparison of AMS 14C and TIMS U/Th ages on two Holocene stalagmites. Radiocarbon, 41, 251– 270. Genty, D., Baker, A., Massault, M., Proctor, C., Gilmour, M., Pons-Branchu, E. & Hamelin, B. 2001. Dead carbon in stalagmites: Carbonate bedrock paleodissolution vs. ageing of soil organic matter. Implications for 13C variations in speleothems. Geochimica et Cosmochimica Acta, 65, 3443– 3457. Goudie, A. S., Viles, H. A. & Pentecost, A. 1993. The late-Holocene tufa decline in Europe. Holocene, 3, 181–186. Han, Z. 2006. Exploitation of Guizhou Province during the Yongzheng Reign-Period and its effect on the rocky desertification in this area. Fudan Journal (Social Sciences), 2, 120–127 [in Chinese, with English abstract]. Herman, J. S. & Lorah, M. M. 1987. CO2 outgassing and calcite precipitation in Falling Spring Creek, USA. Geochimica et Cosmochimica Acta, 62, 251–262. Ihlenfeld, C., Maas, R., Norman, M. C., Gagan, M. K., Drysdale, R. N. & Webb, J. 2003. The climatic significance of trace element and stable isotope variations in a modern freshwater tufa. Geochimica et Cosmochimica Acta, 67, 2341– 2357. Jacobson, R. L. & Usdowski, E. 1975. Geochemical controls on a calcite precipitating spring. Contributions to Mineralogy and Petrology, 51, 65–74. Kano, A., Kawai, T., Matsuoka, J. & Ihara, T. 2004. High-resolution records of rainfall events from clay bands in tufa. Geology, 32, 793– 796. Kattan, Z. 2002. Effects of sulphate reduction and geogenic CO2 incorporation on the determination of 14C groundwater ages—a case study of the Palaeogene groundwater system in north-eastern Syria. Hydrogeology Journal, 10, 495– 508. Kele, S., Korpa´s, L., Deme´ny, A., Kova´cs-Pa´lffy, P., Bajno´czi, B. & Medzihradszky, Z. 2006. Palaeoenvironmental evaluation of the Tata Travertine Complex (Hungary), based on stable isotopic and petrographic studies. Acta Geologica Hungarica, 49, 1– 31.
d13C OF TUFA AND LAND COVER CHANGE Krajcar-Bronic, I., Horvatincic, N., Srdoc, D. & Obelic, B. 1992. Experimental determination of the 14 C initial activity of calcareous deposits. Radiocarbon, 34, 593–601. Kronfeld, J., Vogel, J. C., Rosenthal, E. & Weinstein-Evron, M. 1988. Age and paleoclimatic implications of the Bet Shean travertines. Quaternary Research, 30, 298– 303. Lebron, I. & Suarez, D. L. 1996. Calcite nucleation and precipitation kinetics as affected by dissolved organic matter at 25 8C and pH . 7.5. Geochimica et Cosmochimica Acta, 60, 2276–2765. Li, H., Xu, X., Ku, T., You, C., Buchheim, H. P. & Peters, R. 2008. Isotopic and geochemical evidence of palaeoclimate changes in Salton Basin, California, during the past 20 kyr: 1. D18O and d13C records in lake tufa deposits. Palaeogeography, Palaeoclimatology, Palaeoecology, 259, 182–197. Li, J., Mao, J., Yan, G. & Mei, Z. 1988. Distribution of karst groundwater in southern part of Dushan, Guizhou Province and its exploitation and management. Carsologica Sinica, 7, 279– 285 [in Chinese, with English abstract]. Liu, K., Ding, X., Fu, D., Pan, Y. & Wu, X. 2007a. The AMS measurement of VIRI samples at Peking University. Quaternary Sciences, 27, 469–473 [in Chinese, with English abstract]. Liu, Z., Zhang, M., Li, Q. & You, S. 2003. Hydrochemical and isotope characteristics of spring water and travertine in the Baishuitai area (SW China) and their meaning for paleoenvironmental reconstruction. Environmental Geology, 44, 698–704. Liu, Z., Dai, Y. & Lin, Y. 2004a. Paleoenvironmental reconstruction based on hydrochemistry and tufa stable isotopes: case study of Xiangshui river, Libo, Guizhou. Quaternary Sciences, 24, 447–453 [in Chinese with English abstract]. Liu, Z., Groves, C., Yuan, D., Meiman, J., Jiang, G. & He, S. 2004b. Hydrochemical variations during flood pulses in the southwest China peak cluster karst: Impacts of CaCO3 – H2O–CO2 interactions. Hydrological Processes, 18, 2423–2437. Liu, Z., Li, H., You, C., Wan, N. & Sun, H. 2006. Thickness and stable isotopic characteristics of modern seasonal climate-controlled sub-annual travertine laminas in a travertine-depositing stream at Baishuitai, SW China: implications for paleoclimate reconstruction. Environmental Geology, 51, 257–265. Liu, Z., Li, Q., Sun, H. & Wang, J. 2007b. Seasonal, diurnal and storm-scale hydrochemical variations of typical epikarst springs in subtropical karst areas of SW China: CO2 and dilution effects. Journal of Hydrology, 337, 207– 223. Merz-Preiss, M. & Riding, R. 1999. Cyanobacterial tufa calcification in two freshwater streams: ambient environment, chemical thresholds and biological processes. Sedimentary Geology, 126, 103–124. Meyrick, R. A. & Karrow, P. F. 2007. Three detailed, radiocarbon-dated, Holocene tufa and alluvial fan mollusc successions from southern Ontario: The first in northeastern North America. Palaeogeography, Palaeoclimatology, Palaeoecology, 243, 250– 271.
95
Pazdur, A. & Pazdur, M. F. 1986. 14C dating of calcareous tufa from different environments. Radiocarbon, 28, 534 –538. Pazdur, A., Pazdur, M. F., Starkel, L. & Szulc, J. 1988. Stable isotopes of Holocene calcareous tufa in southern Poland as paleoclimatic indicators. Quaternary Research, 30, 177–189. Pazdur, A., Dobrowolski, R., Durakiewicz, T., Piotrowska, N., Mohanti, M. & Das, S. 2002. 13C and 18O time record and palaeoclimatic implications of the Holocene calcareous tufa from south-eastern Poland and eastern India (Orissa). Geochronometria, 21, 97– 108. Pentecost, A. 1995. The Quaternary travertine deposits of Europe and Asia minor. Quaternary Science Reviews, 14, 1005–1028. Pentecost, A 2005. Travertine. Springer, Berlin. Pentecost, A. & Zhang, Z. 2001. A review of Chinese travertines. Cave and Karst Science, 28, 15–28. Preece, R. C. & Bridgland, D. R. 1999. Holywell Coombe, Folkestone: A 13,000 year history of an English Chalkland Valley. Quaternary Science Reviews, 18, 1075–1125. Reddy, M. M. 1977. Crystallization of calcium carbonate in presence of trace concentration of phosphoruscontained anions: I. Inhibition of phosphate and glycerophorus ions at pH 8.8 and 25 8C. Journal of Crystal Growth, 41, 287 –295. Reimer, P. J., Baillie, M. G. L. et al. 2004. Intcal04 terrestrial radiocarbon age calibration, 0–26 cal kyr BP. Radiocarbon, 46, 1029– 1058. Soligo, M., Tuccimei, P., Barberi, R., Delitala, M. C., Miccadei, E. & Taddeucci, A. 2002. U/Th dating of freshwater travertine from Middle Velino Valley (Central Italy): paleoclimatic and geological implications. Palaeogeography, Palaeoclimatology, Palaeoecology, 184, 147– 161. Srdoc, D., Obelic, B., Horvatincic, N. & Sliepcevic, A. 1980. Radiocarbon dating of calcareous tufa: How reliable data can we expect? Radiocarbon, 22, 858– 862. Suarez, D. L. 1983. Calcite supersaturation and precipitation kinetics in the lower Colorado River, All American-Canal and East Highland Canal. Water Resource Research, 19, 653–661. Szabo, B. J., Bush, C. A. & Benson, L. V. 1996. Uranium-series dating of carbonate (tufa) deposits associated with Quaternary fluctuations of Pyramid Lake, Nevada. Quaternary Research, 45, 271– 281. Thorpe, P. M., Otlet, R. L. & Sweeting, M. M. 1980. Hydrological implications from 14C profiling of UK tufa. Radiocarbon, 22, 897– 908. Troester, W. T. & White, W. B. 1986. Geochemical investigations of three tropical karst drainage basins in Puerto Rico. Ground Water, 24, 475– 482. Valero Garce´s, B. L., Moreno, A. et al. 2008. The Taravilla lake and tufa deposits (Central Iberian Range, Spain) as palaeohydrological and palaeoclimatic indicators. Palaeogeography, Palaeoclimatology, Palaeoecology, 259, 136– 156. Wang, S., Liu, Q. & Zhang, D. 2004. Karst rocky desertification in southwestern China: geomorphology, land
96
Z. LIU ET AL.
use, impact and rehabilitation. Land Degradation and Development, 15, 115–121. Wigley, T. M. L. 1977. WATSPEC: a computer program for determining equilibrium speciation of aqueous solutions. British Geomorphological Research Group Technical Bulletin, 20, 1– 48. Yi, C., Liu, K., Cui, Z., Jiao, K., Yao, T. & He, Y. 2004. AMS radiocarbon dating of late Quaternary glacial landforms, source of the Urumqi River Tien Shan—a pilot study of 14C dating on inorganic carbon. Quaternary International, 121, 99– 107.
Zhou, Y. 2006. The historic progress and sustained development of man–earth relationship of Guizhou Province. Journal of Taiyuan Normal University (Natural Science Edition), 5, 107– 111 [in Chinese, with English abstract]. Zhou, Y., Zhang, P., Pan, G., Xiong, Z. & Ran, J. 2002. Carbon fluxes in the interfaces of soil– air– water of epikarst ecosystem in short time scale: an example of daily dynamics in Autumn as monitored in Maolan Karst Park, Guizhou. Quaternary Sciences, 22, 258– 265.
Holocene land use in western Sicily: a geoarchaeological perspective CHAD HEINZEL1* & MICHAEL KOLB2 1
Department of Earth Science, University of Northern Iowa, Latham Hall, Cedar Falls, IA 50614-0335, USA 2
Anthropology Department, Northern Illinois University, DeKalb, IL 60115, USA *Corresponding author (e-mail:
[email protected]) Abstract: Geoarchaeological research within Sicily continues to characterize the effects of anthropogenic and geological processes upon the island’s Holocene alluvial landscape developments. Interdisciplinary approaches have been used including geomorphological mapping, archaeological survey and excavations to characterize land-use practices though the mid- to late Holocene. Landscape development changes are recorded in the alluvial sediments as a consequence of land use by the indigenous and Roman settlers of Sicilian valleys in the Nebrodi and Polizzo Mountains. A marked change in erosion has been identified during the late Roman occupation of Sicily, probably as a product of intensive pastoralism and land clearing. Sedimentation during indigenous hilltop occupation of north– central and western Sicily was dominated by coarse-grained (cobble or boulder) deposits attributed to flash-flooding. Sedimentation that temporally coincided with the Greek and later Roman occupation of the adjacent valleys is marked by fine-grained deposits. These data continue to support the geological and archaeological interpretations of human–landscape interactions in Sicily. Furthermore, such geoarchaeological data may be used in models to strengthen our present and future landscape conservation methods.
The geoarchaeological sciences provide a useful series of tools and applications that facilitate a deeper understanding of human –landscape interactions through time. As the Earth’s current populations continue to face droughts, floods, earthquakes, shortages in natural resources, and other population-forcing factors, geoarchaeology offers an opportunity to view how past natural and anthropogenic events shaped our modern civilizations and diagnostic cultural developments. Western Sicily’s Holocene sediments contain a substantial record representing an important segment of the Mediterranean region’s prehistoric and historical cultural progression (Leighton 1999; Tusa 1999). Sicily’s strategic centralized location within the Mediterranean Sea and abundant natural resources provided the necessary means for survival and the development of productive trade networks. This paper seeks to delineate land-use changes in western Sicily from recent geoarchaeological data (Fig. 1).
Geoarchaeological data in western Sicily Investigations regarding human –landscape interactions within Mediterranean landscapes have advanced greatly over the past 40 years. The data obtained have led to an increased understanding of the relative roles that anthropogenic and geological processes have had upon the region’s landscape
developments (e.g. Vita-Finzi 1969; Kraft et al. 1977; Pope & Van Andel 1984; Bintliff 1992; Barker 1995; Macklin et al. 2006). Recent geoarchaeological research conducted by the University of Cambridge (the Troina Project) in north–central Sicily and the Universities of Northern Illinois and Northern Iowa (the Monte Polizzo Project) in western Sicily have begun to characterize prehistoric and historical human settlement patterns and landscape changes in Sicily. Sicily was a recurrent sphere of contact, exchange, and dispute between indigenous and colonizing peoples during the Late Bronze Age to Early Iron Age transition (Leighton 1999; Tusa 1999; Serrati 2000). Movement of goods such as foodstuffs and exotic items was frequent; movement of people perhaps just as much, given the repeated human expansions from Italy, Iberia, Phoenicia, Attica, and North Africa (Knapp & Blake 2005). The rise and fall of local political centres and later colonization by Phoenicians, Greeks, and Romans testify to this area’s importance as an agricultural breadbasket in many key periods of history. Western Sicily contains a group of Iron Age proto-urban hilltop sites that may provide insight into the relationships between the development of indigenous political centres and 8th to 6th century BC Greek and Phoenician colonization in the Mediterranean. A group of hilltop sites (Segesta, Entella, and Eryx) were developed by indigenous Iron Age
From: Wilson, L. (ed.) Human Interactions with the Geosphere: The Geoarchaeological Perspective. Geological Society, London, Special Publications, 352, 97– 107. DOI: 10.1144/SP352.8 0305-8719/11/$15.00 # The Geological Society of London 2011.
98
C. HEINZEL & M. KOLB
Fig. 1. Map of western Sicily including examples of the region’s prominent archaeological sites and rivers.
people known as the Elymi (Leighton 1999; Kolb & Tusa 2001; Morris et al. 2002). The exact origins of the Elymi are indistinct. According to Thucydides (c. 460–c. 395 cal yr BC) and Diodorus Siculus (c. 90 to c. 30 cal yr BC), the Elymi had established at least three major strongholds in western Sicily [Segesta, Entella (Giustolisi 1985), and Eryx] by 800 cal yr BC. The Elymi appear to be an amalgam of Anatolian or Italic immigrants and an indigenous ethnic group known as the Sicans (Leighton 1999; Spatafora & Vassallo 2002). As a third political power the Elymi were caught between the strengthening coastal colonies of Motya (Phoenician, established 720 cal yr BC) and Selinus (Greek, established 628 cal yr BC). Northwestern Sicily has been investigated since 1998 by a large multi-national team, exploring Mediterranean Iron Age (EIA) proto-urban settlements under the pressures and influences of Greek and Phoenician colonialism. As very little was known regarding the hinterlands of these protourban settlements, Northern Illinois University was charged with conducting a regional archaeological survey. Extensive archaeological survey data based on 4 m surface-spacing revealed a substantial Neolithic to Medieval human presence within the
Chuddia River Valley. A geographic information system (GIS) analysis of the distribution of recovered ceramic material revealed a number of large ceramic concentrations as well as other areas of ‘off-site’ activity. The Chuddia Valley survey recovered 42 901 non-diagnostic artefacts. These non-diagnostic artefacts (body shards) were visually attributed to general (century) chronologies, but did not possess distinctive rims, handles, or slips or designs that could identify the artefact’s function. These bulk artefacts were sorted into major chronological categories, and counted in the field. An additional 3066 diagnostic artefacts, including chronologically (sub-century) precise diagnostic features such as ceramic base, rim, handle, and lamp fragments, were collected from the field and analysed in a laboratory. Other artefacts indicative of site function were also documented, such as bone, ceramic wasters, mosaic tiles, metal slag, glass, abraders, and hammer and grinding implements. The variability of these concentrated plough-soil ceramic zones and their unique suites of cultural materials are interpreted as traceable links to each archaeological site’s probable size, and function and occupation. When these survey data are viewed in conjunction
GEOARCHAEOLOGY IN SICILY
with time, regional variability, and geomorphological processes it is also possible to begin interpreting important archaeological developments such as settlement patterns and interaction between distinct cultural groups (trade). Archaeological data recovered from Monte Polizzo (713 m above sea level) suggest it was a significant Iron Age settlement along with Segesta, Entella, and Eryx. The existence of a complex trade network between western Sicily’s Iron Age hilltop settlements has emerged from ceramic analyses (Kolb & Speakman 2005; Polito 2006; Heinzel et al. 2009). The Elymi at Monte Polizzo also appear to have been influenced by the Greeks through the adaptation of ceramic styles and writing, but still remained a distinctive ethnic group despite centuries of coastal Greek and Phoenician colonization and the political conflicts that ensued. Recent archaeological excavations in western Sicily have documented an influx of Greek and Greek-imitation ceramic styles within Monte Polizzo’s and other local proto-urban cultural material assemblages at c. 600 BCE (Morris et al. 2002; Spatafora & Vassallo 2002; Kolb & Speakman 2005). A documented set of inscriptions excavated at Segesta (a known Elymian stronghold) was written using Greek styled-script, but has been interpreted as an example of the non-Greek Elymian script (La Rosa 1996).
Landscape analysis The University of Northern Iowa was charged with characterizing the late Holocene landform sediment-assemblage developments of Monte Polizzo’s hinterlands and delineating their interrelationships with western Sicily’s evolving Late Bronze to Roman settlement patterns. The hinterlands surrounding Monte Polizzo contained a discontinuous sedimentary record dating from the late Triassic (215 Ma) to the present day. Common bedrock types include limestone, shale, siltstone, sandstone, conglomerate, and gypsum assemblages that reveal deep to shallow Mesozoic marine to Cenozoic nearshore transitional depositional environments. The Quaternary alluvial setting contains prominent alluvial fans that have been dissected by a fluvial drainage system, the Chuddia River. Thirty-four alluvial (13 fan and 21 fluvial) stratigraphic sections were investigated from natural exposures or excavated using a track-hoe (1 m bucket). Each stratigraphic section was measured, described (Eyles et al. 1983), and sampled. Each sample was given a label that included a section number, year sampled, and height of sample above the stratigraphic base (e.g. CHS21-02-35 cm; CHS stands for Chuddia River section). Lithofacies logs
99
were measured from the base upward. Three of the fluvial stratigraphic sections were discarded because they had been compromised by modern agricultural practices.
Alluvial fans A thick assemblage of coalescing alluvial fan deposits (.40 m in thickness) produced the southern slope of Montagna Grande. Modern quarrying for road aggregate has exposed many outcrops. Three alluvial fans were investigated and named after the nearest landowners: the Verme, Armata and Lentini Fans (Fig. 2). Poorly cemented gravel, moderately well- to well-cemented gravel, and clay-rich deposits are the three primary depositional components of these fan-sediment assemblages. As a representative example the Lentini Fan contained a complex sequence of eight stacked Holocene channels (Fig. 3). Each channel cut and fill sequence contained angular to sub-angular limestone gravel. These overlapping channels were incised within well-imbricated pre-Holocene fan gravel deposits. Cross-cutting channel relationships were observed. The channels were distinguished from earlier fan sedimentation by: (1) irregular concave channel shapes; (2) a decrease in cementation within the cross-cutting channels; (3) an overall increase in relative particle size from the underlying preHolocene gravel; (4) a prominent increase in organic-rich matrix. These stacked Holocene channels are archaeologically significant. They contain artefacts including ceramics, snail shells, lithic debitage, bone, and charcoal. The diagnostic ceramics are attributed to the Neolithic and Copper Ages (5000 BC to 3100 cal yr BC). In addition to these geological deposits, a large buried anthropogenic stone wall, an artefact-bearing compacted clay floor and a concentrated layer of ash– charcoal–bone were identified adjacent to the Holocene channel complex’s eastern boundary. Clay-rich deposits (5YR 4/4, dry) with varying degrees of pedogenic development are interbedded within each fans’ braided channel networks. Particle size analyses of the sampled sediments revealed that these deposits contained an average of 43% clay, 32% silt, and 25% sand. In some places the reddish clay-rich units within the Montagna Grande fans appear to be highly oxidized and preserved soil horizons. Modern soils developed within the carbonate-rich (limestone) uplands of Montagna Grande often have a characteristic reddish brown terra rossa colour, and are similar in texture and composition to the clay-rich units. The clay-rich fan units appear to influence the local hydrological conditions. The transition from
100
C. HEINZEL & M. KOLB
Fig. 2. Map of the study area surrounding Monte Polizzo (archaeological site outlined) and Montagna Grande exhibiting the Armata Fan –Palaeosol, the Lentini Fan, the Verme Fan, and the locations of the fluvial stratigraphic sections (CHS 9, 16, 21) discussed within the Chuddia River valley.
surficial to vadose to groundwater flow is directly influenced by the local geological settings. At present the internal cemented gravel and clay deposits act as aquicludes that redirect flowing vadose and groundwater to leave fan-sediment assemblages as natural springs at various points throughout the fan surfaces. The cemented alluvial deposits were produced during past stable landscapes with prolonged atmospheric exposure and interactions with seasonal high water tables. A carbonate-rich setting produced cements through dissolution and precipitation. This increased cementation has locally reduced the infiltration capacity of the fans. Increased runoff and runoff velocity during intense precipitation events have led to channel incision. A well-developed palaeosol was preserved in the mid-fan portion of the Armata Fan. This palaeosol indicated a period of fan surface stability. Organic material from the A-horizon of the Armata Palaeosol was radiocarbon dated to 6084–5837 cal yr BC (2s). Stratigraphic description of this palaeosol revealed that it was developed within the coarsegrained limestone fan particles and was rapidly buried by a well-imbricated, 150 cm thick, gravel deposit. Rapid sheet-wash sedimentation is typical of arid to semi-arid depositional environments
(Nemec & Postma 1993) and the burial depth of greater than 150 cm has lessened the possibility of modern root contamination of the radiocarbon sample from the palaeosol.
The Chuddia River The two sediment source areas for the Chuddia Valley are Monte Polizzo (to the south) and Montagna Grande (to the north). Monte Polizzo was produced during an uplift of the Terravecchia Formation (11.2–6.3 Ma) and provides quartzite pebbles and boulders, sand, and clay. Montagna Grande to the north contributed limestone to metalimestone particles (very fine pebble to boulder size) to the Chuddia drainage network. Variable percentages of sandstone and limestone particles have produced a bimodal mixture of sediment within the valley. The fluvial stratigraphic contacts (transitions) between deposition and erosion episodes ranged from 2 to 5 cm thick. This range of stratigraphic contact thicknesses indicates either episodic sedimentation, a change in the environmental conditions (discharge variability), or a deposition hiatus leading to landscape stabilization and soil
GEOARCHAEOLOGY IN SICILY
101
Fig. 3. Stratigraphic units characterizing the younger (Holocene) cut and fill deposits. The channel contained imbricated fining-upward sequences of limestone gravel. The channels contained a high concentration of pottery (P), bone (B), and charcoal (C). Adjacent to the channel an anthropogenic structure or tomb (T) featured a compacted clay floor, ash layer (20 cm), and a variety of artefacts.
development. The 21 fluvial stratigraphic sections contained a variety of basal deposits including clay, a clast-supported, quartzite-cobble conglomerate or an imbricated gravel. The fine-grained unit (average texture 5% sand, 37% silt and 58% clay) was characterized by a massive blue–grey (7.5YR 5/1) clay. This deposit contained a relatively wide range of textural end members, up to 28% sand and 82% clay, but never contained coarse fragments. The thickness of this fine-grained basal unit ranged between 15 and 120 cm. Preliminary pollen analyses of these fine-grained sediments indicate the presence of abundant Dinophyceae (marine algae cysts). Charcoal obtained from two of these clay units (CHS16, Unit 1, 80 cm above base) produced an average radiocarbon date of 544 –381 cal yr BC (2s). The second unit that could be correlated throughout the Chuddia River system was a clast-supported (quartzite-cobble) conglomerate. The clast distribution was predominantly bimodal, with rounded to well-rounded coarse-grained clasts set within a fine- to medium-grained matrix of 56% sand, 14%
silt and 30% clay. Snail shells (Theba pisana) recovered from a correlative stratigraphic unit (CHS9, Unit 1, 33 cm above the stratigraphic base) gave a radiocarbon date of 1895 –1616 cal yr BC, which, when corrected for the ingestion of ‘old’ carbon from the surrounding Eocene carbonate bedrock, placed this shell between 595 and 316 cal yr BC, well within the statistical range of the other charcoal radiocarbon dates. The coarsegrained depositional Unit 1 of CHS9 contained the first occurrence (inclusion) of ceramic artefacts (Fig. 4). The artefacts recovered from this stratigraphic unit dated exclusively to Sicily’s Iron Age period (900 –600 cal yr BC). Each successive stratigraphic unit contained chronologically mixed artefact assemblages. The third primary fluvial stratigraphic unit was classified as an imbricated, clast-supported gravel. CHS21, Unit 2 (60 cm thick) contained a representative angular limestone gravel deposit (Fig. 5). The average (n ¼ 50) clast size was 3.5 cm (a-axis), 2.3 cm (b-axis), and 0.7 cm (c-axis). These gravel deposits did not contain measurable
102
C. HEINZEL & M. KOLB
Fig. 4. CHS9 lithofacies column (1:10 cm), This drawing includes the locations of ceramic concentrations and charcoal used to obtain radiocarbon dates (black circles with diagonal white lines). Stratigraphic nomenclature (Eyles et al. 1983) was used to characterize the valley’s fluvial sediments: Dmsc (D, diamicton; m, matrix supported; s, smooth contact; c, current reworking); Sm (S, sand; m, matrix-supported); Sg (S, sand; g, graded); Fm (F, Fine–clay; m, massive); Dmm (D, diamicton; m, matrix-supported; m, massive). The relative particle size from unit to unit, and the locations and orientations of clasts are also shown.
amounts of matrix. The stratigraphic contact between Units 2 and 3 is abrupt (0.5–2 cm) and inclined at an angle of 58 to the west. The three distinctive fluvial basal deposits continued at depth, but each section was terminated when it intersected with the current water table. The fluvial stratigraphic units were correlated by combining lithofacies data, the succession of
artefact assemblages and radiocarbon data, which facilitated reliable physical and temporal correlations. The alluvial deposits overlying the clay represent an assemblage of floodplain and channel deposits of varying thickness. The channel deposits (medium- to coarse-grained sand) often contain small trough cross-beds and variable amounts of clay. Colluvial limestone gravel was commonly
GEOARCHAEOLOGY IN SICILY
103
Fig. 5. CHS21 western side wall (1:20 cm). Unit 1 contains coarse-grained quartzite cobble to boulder fluvial sediments. Unit 2 contains imbricated alluvial fan (Verme Fan) limestone gravels from Montagna Grande; the arrows represent average palaeocurrent directions based on 100 clast measurements. Unit 3 contains massive colluvial and sand deposits.
mixed within the coarse-grained channel deposits. The angular nature of the limestone gravel confirmed Montagna Grande as the proximal source area. These stratigraphic and sedimentological data also verified that the alluvial toe slopes of the Verme Fan came into direct contact with the Chuddia Valley’s main fluvial channel.
Discussion Geoarchaeological research is truly an interdisciplinary effort. Geoarchaeology integrates multiple subdisciplines of the archaeological, geological, and biological sciences in an attempt to delineate the interrelationships of an area’s causative natural processes and the anthropogenic influences upon a landscape’s development. These data are also valuable when interpreting a landscape’s effects upon an evolving human population. Defining the scale of a geoarchaeological investigation is critically important given the varying temporal resolutions of geological processes and human life cycles (Stein 1993). Geoarchaeological investigations within Sicily are in their infancy. The scales of current investigations have yet to achieve resolutions capable of linking specific (decadal to annual) climatic variance to past human lives. This does not mean that these data are not important, it simply necessitates that care be taken when attempting to directly correlate climatic and landscape variables to specific human events. For example, archaeological data suggest that the Elymi thrived on Monte Polizzo beginning around 575 BC and ending abruptly at
475 BC (Morris et al. 2002). The reason for the Elymi’s abandonment of Monte Polizzo remains a mystery. Was site abandonment a product of sociopolitical change, warfare, sickness, drought, or depletion of natural resources? The Chuddia Valley’s landform sediment assemblages do record distinct periods of landscape destabilization during and/or immediately after the Neolithic and Iron Age human occupation periods. Interpreting the causative factor or factors leading to a settlement’s collapse is often equally important and difficult. The archaeological and physical landscapes from the Chuddia River Valley provide a glimpse into the possible environmental and anthropogenic roles leading to the rise and fall of the Elymian settlement at Monte Polizzo. The Chuddia Valley’s mid-Holocene (Early to Mid-Neolithic) record begins with the presence of the well-developed Armata Palaeosol (6100– 5808 cal yr BC). This palaeosol suggests the midHolocene pediments of Montagna Grande were stable and facilitated the development of an organicrich (10YR 3/1 and 1.81% organic carbon) soil. The dark (3/1) soil colour facilitates an A-horizon designation according to USDA (US Department of Agriculture) criteria. Lacustrine core data from Lago di Pergusa (central Sicily) and Gorgo Basso (southwestern Sicily) indicate a period of increased precipitation and coastal forest expansion (Sadori & Narcisi 2001; Zanchetta et al. 2007; Tinner et al. 2009). These data provide supporting evidence that there was enough precipitation for plant material to grow and decay, supplying organic debris to soil development.
104
C. HEINZEL & M. KOLB
The coastal Neolithic populations of Sicily were probably forced to search for new farmland in the interior hills of Sicily during a climatically induced period of increased rainfall and expansion of evergreen coastal forests (Tinner et al. 2009). The Aramata Palaeosol, archaeological survey data and the presence of a Neolithic (5500– 4000 cal yr BC) artefact assemblage within the Chuddia Valley serve as records of western Sicily’s migrating (coastal to interior) Neolithic population (Kolb & Tusa 2001; Heinzel 2004). Slope instability processes have continued to shape the Chuddia Valley’s landforms since the mid-Holocene. Three primary indications of slope instability were identified within the valley: (1) the burial of the Armata Palaeosol by imbricated fan gravel; (2) the reactivation of the Lentini Fan and channel; (3) the stratigraphic interrelationships between the alluvial fans of Montagna Grande and the main fluvial channel of the Chuddia River. MidHolocene landscape destabilization is first recorded with the valley’s burial of the Armata Palaeosol, after 5000 BC, by imbricated alluvial fan gravel. A radiocarbon date of 5349– 5171 cal yr BC from Unit 2 of the Lentini complex also marks this midHolocene landscape instability through the erosion of an older late Pleistocene fan surface through incision and sedimentation within a younger midHolocene fan channel. A Late Copper Age (3000–2500 BC) anthropogenic structure was identified adjacent to the Lentini channel’s eastern end. Two charcoal pieces collected from opposite ends of the structure’s compacted clay floor gave radiocarbon dates between 2679–2468 and 2886 –2620 cal yr BC. This site contained multiple artefacts including lithic (flint) tools, flint debitage, hearth – ash accumulation, and a variety of ceramics. The ceramics associated with this site and the adjacent alluvial channel ranged from Neolithic mobile cooking ware to Late Copper Age house tile and cooking ware with a characteristic red burnish (Heinzel 2004). Since the mid- to late Holocene Sicily has experienced a general trend of increasing aridity. Vegetative types and concentrations appear to have remained relatively constant until the 850 – 650 cal yr BC collapse of western Sicily’s coastal forests. The extensive and rapid nature of this vegetative change has been attributed to the influx of Greek colonists rather than the trend towards aridity (Tinner et al. 2009). Archaeological excavations and surveys have documented an extensive Iron Age (850–475 cal yr BC) indigenous artefact assemblage believed to be linked to an Elymian polity at Monte Polizzo (Prescott et al. 1998; Kolb & Tusa 2001; Morris et al. 2002). A corresponding Elymian artefact assemblage was recovered from the valley’s fluvial stratigraphic units. Charcoal
and ceramic artefacts recovered from the massive clay or clast-supported quartzite conglomerate layers of the Chuddia River’s sediments provided radiocarbon dates between 600 and 380 cal yr BC, which appears to strengthen the probability of a temporal link between the Elymi at Monte Polizzo (height of site occupation 575–475 BC) and erosion of the valley’s upland landscapes and subsequent infilling of the Chuddia River. The massive grey clay unit (554–383 cal yr BC) is common to seven fluvial stratigraphic sections (n ¼ 21) in the river’s headwater area. Well-sorted clay-sized sediments are commonly associated with low-energy depositional environments. The Late Iron Age clay deposit suggests two possibilities: (1) adequate precipitation levels and topographic controls led to a lacustrine depositional setting or (2) the presence of an alluvial fan or anthropogenically produced dam created a lowenergy depositional environment (Heinzel 2004). The specific contributing climatic and anthropogenic variables present during the formation of the Chuddia Valley’s massive clay remain unclear. The only other climatic evidence is offered by the Lake Pergusa core data, which document a return to a lacustrine carbonate setting at 2800 BP, suggesting a slight increase in precipitation relative to the previous 500 years (Sadori & Narcisi 2001). The specific reasons for the Elymi’s abandonment of the Monte Polizzo site continue to be mysterious. However, there is a clear geological change from fine- to coarse-grained sedimentation within the Chuddia River while the Elymi occupied Monte Polizzo or immediately following the site’s abandonment (Heinzel 2004). This increase in particle size is documented by the clast-supported quartzite conglomerate and the successive coarse-grained fluvial stratigraphic layers (quartzite conglomerates or imbricated limestone gravel). In addition to the introduction of coarse fragments there is also a notable change in the texture of the fluvial matrix. The massive clay has an average texture of 5% sand, 37% silt and 58% clay whereas the coarsegrained units contain an average of 56% sand, 14% silt and 30% clay (Heinzel 2004). The archaeological evidence documents that the site was abandoned near 475 cal yr BC. Excavations have not uncovered any evidence of on-site warfare or earthquakes. Recent geoarchaeological studies in southern Europe have characterized complex palaeoenvironmental and archaeological records (e.g. Sadori & Narcisi 2001; Ayala 2004; Heinzel 2004; Magny et al. 2007; Tinner et al. 2009). Interpreting the relationships between the indigenous peoples of Sicily, Greek and Phoenician colonialism, Holocene climatic variability, and their cumulative effect on human land-use patterns or geomorphological processes is complicated. Using
GEOARCHAEOLOGY IN SICILY
the current data, it is not possible to explicitly state the mechanisms that led to the recorded changes in the Chuddia Valley’s fluvial stratigraphy. Identifying the causative variables that produced the identified landscape changes will require additional well-dated and high-resolution geoarchaeological investigations. Recent palaeobotanical investigations are beginning to shed light upon the mid- to late Holocene relationships between humans and western Sicily’s vegetation history (Sadori & Narcisi 2001; Stika et al. 2008; Tinner et al. 2009). There is still a great need for further geoarchaeological investigations that delineate the relationships between climatic variability and anthropogenic settlement patterns. One method that is capable of providing these climatic proxy data is carbon and oxygen isotopic studies. Western Sicily contains at least two promising sources for obtaining a local highresolution climate record of temperature and precipitation: speleothems and snail shells. Given the highly variable distribution of precipitation throughout the Holocene and the carbonate bedrock of western Sicily an isotopic investigation may provide the temporal and physical resolution needed to identify the factors leading to the abandonment of Monte Polizzo. The transition from indigenous to colonial (Greek, Phoenician and Roman) settlement in Sicily is clearly preserved within the island’s sedimentary record. The primary difference recorded in sedimentation patterns is shifts in particle size (coarse fragments and matrix). This change in particle size has been recorded in western Sicily, in the Chuddia River Valley, with a change from fine to coarse following the Iron Age (Heinzel 2004). Ayala (2004) also documented a particle size change within Sicily’s fluvial record, from coarse to fine grained during the fourth to second centuries BC in north –central Sicily (Troina Valley). Recent attempts using GIS and a variation of the Universal Soil Loss Model Equation (USLE) have also documented the adverse effects of Roman land-use practices in north–central Sicily (Fitzjohn 2003; Ayala & French 2005). These data have shown a significant period of landscape destabilization during the Roman occupation of north–central and western Sicily (as seen in the change from coarse-grained flash-flood sedimentation to fine-grained slopewash). Ayala & French (2005) have attributed this sedimentation change to extensive land clearance during a period of intensive Roman pastoralism. The continuing geoarchaeological investigations within Sicily are using geological, archaeological, historical (ancient literature) and biological data in an attempt to provide a comprehensive characterization of the interaction between humans and the natural world.
105
Conclusion Geoarchaeology is an interdisciplinary science that constructs a bridge between the Earth’s natural processes and human endeavours past, present and future. The construction of such bridges is complicated by incomplete stratigraphic records, disturbed archaeological material cultures and contrasting temporal scales. The road to a comprehensive understanding of human–environmental interrelationships requires the construction of such bridges no matter how difficult. Recent geoarchaeological data from Sicily and elsewhere are beginning to provide insight into the positive and negative ramifications humankind has upon the Earth. The Chuddia Valley contains a geoarchaeological record that represents nearly 8000 years of human activity and landscape processes. Archaeological and palaeoenvironmental investigations are beginning to delineate the complex interrelationships between humans and their environment. Three periods of landscape destabilization have been documented in association with the Neolithic, Iron Age and Roman occupation periods in Sicily: (1) the reactivation of alluvial fan surfaces and burial of fertile Neolithic soil horizons near Montagna Grande (Heinzel 2004); (2) a dramatic increase in particle size within the fluvial sediments of the Chuddia River linked to the Elymi’s use and/or abandonment of the Monte Polizzo site (Heinzel 2004); (3) a substantial decrease in particle size and increase in rate of sediment supply to the Troina River as a product of intensive land use during the Roman period (Fitzjohn 2003; Ayala 2004; Ayala & French 2005). Although the specific anthropogenic or climatic roles that influenced Sicily’s landform sediment assemblages still require further investigation, these sediments and associated archaeological remains provide a means to learn from the past. Striving to understand such interrelationships not only facilitates our understanding of the rise and fall of human populations, but also provides vital data that may be used to lessen the severity of future climatic fluctuations. For example, Tinner et al. (2009) have used their geoarchaeological data from Gorgo Basso (southwestern Sicily) and models from the Intergovernmental Panel on Climate Change (IPCC 2007) to predict the expansion of drought-adapted treeless vegetation and desertification in Sicily if global temperatures continue to rise. Whether human interactions with the environment are recorded by changes in local sediment patterns as in the catchments of Sicily (Ayala 2004; Heinzel 2004) or are recorded within regional to global depositional systems (Ruddiman 2003), interdisciplinary, regional and site-specific studies are critically important when attempting to advance
106
C. HEINZEL & M. KOLB
our knowledge of the Mediterranean’s prehistoric and historical landscapes, and to enhance our understanding of the causative roles that humans and climate have upon one another and the Earth’s dynamic landscapes. Funding for this research came from the University of Northern Iowa, the Geological Society of America, the National Science Foundation (G1A-62128), and the Soprintendenza BB. CC. AA. Di Trapani, Sezione per i Beni Archaeologici. Our work would not be possible without the dedication and assistance of our Salemi associates: S. Conforto, N. Spagnolo, G. Scememi, S. Cassa, B. Terranova, N. Bascone, V. Scalisi, A. Puma, N. Puma, the Bascone family, and many other wonderful people in and around Salemi.
References Ayala, G. 2004. Landscape/land use change in north– central Sicily: A geoarchaeological approach. Doctoral dissertation, University of Cambridge. Ayala, G. & French, C. A. I. 2005. Erosion modeling of past land-use practices in the Fiume di Sotto di Troina River Valley, north–central Sicily. Geoarchaeology, 20, 149–167. Barker, G. 1995. The Biferno Valley Survey: The Archaeological and Geomorphological Record. Leicester University Press, London. Bintliff, J. 1992. Erosion in the Mediterranean lands: a reconsideration of pattern, process and methodology. In: Bell, M. & Boardman, J. (eds) Past and Present Soil Erosion: Archaeological and Geographical Perspectives. Oxbow Monograph, 22, 125– 131. Eyles, N., Eyles, C. H. & Miall, A. D. 1983. Lithofacies types and vertical profile models; an alternative approach to the description and environmental interpretation of glacial diamict and diamictite sequences. Sedimentology, 30, 393– 410. Fitzjohn, M. P. 2003. Spatial investigations of interaction between the indigenous and colonial populations in Sicily during the first millennium B.C. Doctoral dissertation, University of Cambridge. Giustolisi, V. 1985. Nakone ed Entalla. Assessorato dei Beni Culturali e Ambientali e della Pubblica Istruzione, Stampatori Tipolitografi Associati, Palermo. Heinzel, C. E. 2004. Greek, Phoenician, and Roman colonization v. Holocene landscape development: Environmental implications on a developing indigenous society, western Sicily. PhD dissertation, Northern Illinois University, DeKalb. Heinzel, C. E., Montana, G. et al. 2009. Resource production capabilities of an Elymian Polity, western Sicily. Geological Society of America, Abstracts and Programs, 41, 65. IPCC 2007. Working Group I ‘The Physical Science Basis’. Cambridge University Press, Cambridge. Knapp, A. B. & Blake, E. 2005. Prehistory in the Mediterranean: the connecting and corrupting sea. In: Knapp, A. B. & Blake, E. (eds) The Archaeology of Mediterranean Prehistory. Blackwell Studies in Global Archaeology. Blackwell, Oxford.
Kolb, M. J. & Speakman, R. J. 2005. Elymian regional interaction in Iron Age western Sicily: a preliminary neutron activation study of incised/impressed tablewares. Journal of Archaeological Science, 32, 795–804. Kolb, M. J. & Tusa, S. 2001. The Late Bronze Age and Early Iron Age Landscape of Interior Western Sicily. Antiquity, 75, 503 –504. Kraft, J. C., Aschenbrenner, S. E. & Rapp, G. Jr. 1977. Paleogeographic reconstructions of coastal Aegean archaeological sites. Science, 195, 941– 947. La Rosa, I. 1996. The Impact of the Greek colonists on the Non-Hellenic Inhabitants of Sicily. In: PuglieseCarratelli, G. (ed.) The Western Greeks. Thames & Hudson, London. Leighton, R. 1999. Sicily Before History: An Archaeological Survey from the Palaeolithic to the Iron Age. Cornell University Press, Ithaca, NY. Macklin, M. G., Benito, G. et al. 2006. Past hydrological events reflected in the Holocene fluvial record of Europe. Catena, 66, 145– 154. Magny, M., Vanniere, B. et al. 2007. Early-Holocene climatic oscillations recorded by lake-level fluctuations at Lake Accesa (Tuscany, Italy). Quaternary Science Reviews, 26, 1951–1964. Morris, I., Jackman, T. et al. 2002. Stanford University excavations on the acropolis of Monte Polizzo, Sicily, III: Preliminary Report on the 2002 Season. Memoirs of the American Academy in Rome, 47. Nemec, W. & Postma, G. 1993. Quaternary alluvial fans in southwestern Crete: sedimentation processes and geomorphic evolution. In: Marzo, M. & Puigdefabregas, C. (eds) Alluvial Sedimentation. International Association of Sedimentologists, Special Publications, 17, 235– 274. Polito, A. 2006. Analisi petrografiche e chimiche della ceramic indigena proveniente da Monte Saraceno di rav anusa e da altri siti. Assessorato dei Beni Culturali ed Ambientali e della Pubblica Istruzione, Agrigento. Pope, K. & Van Andel, T. 1984. Late Quaternary alluviation and soil formation in the Southern Argolid: its history, causes, and archaeological implications. Journal of Archaeological Science, 11, 281– 306. Prescott, C., Muhlenbock, C. & Englund, E. (eds) 1998. Sicilian– Scandinavian Archaeological Project Annual Report 1998. University of Oslo, Oslo. Ruddiman, W. F. 2003. The anthropogenic greenhouse era began thousands of years ago. Climate Change, 61, 261– 293. Sadori, L. & Narcisi, B. 2001. The postglacial record of environmental history from Lago di Pergusa, Sicily. Holocene, 11, 655 –670. Serrati, J. 2000. Sicily from pre-Greek times to the fourth century. In: Smith, C. & Serrati, J. (eds) Sicily from Aeneas to Augustus: New Approaches in Archaeology and History. Edinburgh University Press, Edinburgh. Spatafora, F. & Vassallo, S. 2002. Sicani Elimi e Greci: Storie di contatti e terra di frontier. Assessorato dei beni culturali ed ambientali e della pubblica, S.F. Flaccovio, Palermo. Stein, J. K. 1993. Scale in archaeology, geosciences, and geoarchaeology. In: Stein, J. K. & Linse, A. R. (eds) Effects of Scale on Archaeological and Geoscientific
GEOARCHAEOLOGY IN SICILY Perspectives. Geological Society of America, Special Papers, 283, 1– 10. Stika, H. P., Heiss, A. G. & Zach, B. 2008. Plant remains from the early Iron Age in western Sicily: differences in subsistence strategies of Greek and Elymian sites. Vegetation History and Archaeobotany, 17, 139– 148. Tinner, W., Van Leeuwen, J. F. N. et al. 2009. Holocene environmental and climatic changes at Gorgo Basso, a coastal lake in southern Sicily, Italy. Quaternary Science Reviews, 28, 1498–1510.
107
Tusa, S. 1999. La Sicilia nella preistoria. Sellerio editore, Palermo. Vita-Finzi, C. 1969. The Mediterranean Valleys: Geological Changes in Historical Times. University of Cambridge Press, Cambridge. Zanchetta, G., Borghini, A., Fallick, A. E., Bonadonna, F. P. & Leone, G. 2007. Late Quaternary palaeohydrology of Lake Pergusa (Sicily, southern Italy) as inferred by stable isotopes of lacustrine carbonates. Journal of Palaeolimnology, 38, 227–239.
Alluvial stratigraphy and geoarchaeology in the Big Fork River Valley, Minnesota: human response to Late Holocene environmental change CHRISTOPHER L. HILL1, GEORGE RAPP2* & ZHICHUN JING3 1
Department of Anthropology and Environmental Studies Program, Boise State University, Boise, ID 83725-1950, USA
2
Department of Geosciences, University of Minnesota Duluth, Duluth, MN 55812, USA 3
Department of Anthropology, University of British Colombia, Vancouver, BC V6T 1Z1, Canada *Corresponding author (e-mail:
[email protected])
Abstract: The Late Quaternary geomorphology and stratigraphy of the Big Fork River valley, within the Rainy River basin of northern Minnesota, reveals evidence of prehistoric human interaction with late Holocene riverine environments. By 11 000 14C BP, deglaciation made the region inhabitable by human groups using Clovis artefacts. Human habitation would also have been possible during the Moorhead low-water stage of glacial Lake Agassiz, starting at 10 500 14 C BP. Near its confluence with the Rainy River, the valley floor of the Big Fork valley consists of a floodplain complex and two terraces. The multi-component stratified Hannaford site is situated within the active floodplain. Overbank deposits contain artefacts in primary context, whereas artefacts within the point bar deposits are in secondary archaeological context; these deposits are associated with changing alluvial settings as the river moved eastward. Aggradation of the valley fill beneath the lowest surface (T0, floodplain complex) began by 3000 years ago and is associated with human activities focused on seasonal fishing and the use of riparian resources from 1300 to 650 14C BP.
Human– environmental interactions include the role that landscape contexts and resources play in determining where human activities take place, and the ways in which humans respond to changing environmental situations. Quaternary fluvial landforms and stratigraphic sequences provide information on the long-term consequences of environmental change (Benito et al. 1998; Gregory & Benito 2003). Although landform evolution often is the result of Late Quaternary natural climatic variability it can also result from shifts in biomes, base levels, or other factors that influence how humans adapt to environmental conditions and how the archaeological record is subsequently modified (Knox 1983, 1993; Schumm & Brakenridge 1987; Bull 1991). For instance, in some formerly glaciated areas, floodplains can be relatively low-energy systems that have been subjected over the last 10 000 – 8000 years to slow lateral migration and overbank sedimentation. Environmental change also has been implicated in the formation of terrace sequences as a result of down-cutting during and after the Pleistocene–Holocene transition. On some scales, river valleys become entrenched and aggradational terraces are the result of the combination of surface
uplift and fluctuating climate (Bridgland & Westaway 2008). For example, climate conditions leading to reduced hydrological base levels along with isostatic rebound can facilitate retrenchment within a river valley. This study documents the Late Quaternary geomorphological evolution of a river valley in a deglaciated area of central North America, evaluates the role of alluvial processes in interpretation of the prehistoric archaeological record, and uses this information to examine how humans have been influenced by and responded to environmental change, with an emphasis on the late Holocene. Although changes in the regional environment (site macroenvironment; Butzer 1982) are examined, mesoenvironmental (topographic setting and landforms) and microenvironmental (site strata and local environmental elements) landscape contexts are emphasized. The focus is on the lower drainage of the Big Fork River in northern Minnesota, along the US– Canada border within the Rainy River basin, where the valley fills contain a postglacial environmental record that can be directly related to archaeological assemblages reflecting prehistoric human activities.
From: Wilson, L. (ed.) Human Interactions with the Geosphere: The Geoarchaeological Perspective. Geological Society, London, Special Publications, 352, 109– 124. DOI: 10.1144/SP352.9 0305-8719/11/$15.00 # The Geological Society of London 2011.
110
C. L. HILL ET AL.
Regional setting The Big Fork River is a meandering stream that originates in north– central Minnesota (Fig. 1) and flows northward for about 266 km before it enters Rainy River. The study area is in the lower reach of the Big Fork River at its junction with the Rainy River, about 27 km west of Fort Francis– International Falls. Near the confluence with the Rainy River, the channel is about 100 –200 m wide and the valley, as marked by the 335 m (1100 ft) contour, is about 1 km wide. In this vicinity there are several important archaeological localities including the Smith Mound complex adjacent to the Rainy River and east of the Big Fork, and the Hannaford site situated immediately west of the Big Fork (Fig. 2). Artefacts from these sites are local variants of assemblages taxonomically assigned to the Late Woodland period (Lenius & Olinyk 1990). At the Smith site, Late Woodland assemblages have been classified into several groups chiefly distinguished and subdivided on the basis of ceramic typology. Radiocarbon dates and stratigraphic contexts suggest that Smith site Laurel artefacts date from about 1500 to 1250 14C
Fig. 1. Regional map.
BP whereas Early Blackduck pottery dates to around 1000 14C BP (Lugenbeal 1978). Assemblages containing Late Blackduck ceramics are assigned to the Rainy River Composite or Rainy River Late Woodland Complex and date from about 900 to 650 14C BP (Lugenbeal 1978; Lenius & Olinyk 1990; Arzigian 2008). Before the development of the Big Fork River, this region of northern Minnesota and northwestern Ontario was covered by late Wisconsin glacial ice. Final deglaciation occurred around 11 000 14C BP when the ice margin was in the vicinity of the Rainy River moraine, north of the Big Fork. The area was exposed during the Moorhead low-water stage of glacial Lake Agassiz and then reflooded until about 9500 14C BP. Since then the geomorphology and stratigraphy of the area have been controlled by the fluvial regimes of the Rainy and Big Fork rivers.
Materials and methods The data for this study consist of published and unpublished proposals and reports, field notes,
HUMAN RESPONSES TO A FLUVIAL REGIME
111
Fig. 2. Geomorphological setting.
photographic slides and excavation forms, maps and profiles, air photographs, USGS (US Geological Survey) topographic maps, sediment cores, and laboratory analyses of sediments (Hill et al. 1995; Rapp et al. 1995). We first examined published maps and reports, and interpreted aerial photographs, satellite images, and 1:24 000 topographic maps. These provided an understanding of the basic landforms and were used to place the lower Big Fork drainage within the context of the regional deglaciation and glacial Lake Agassiz chronological framework. Field studies allowed construction of a geomorphological map, which defined and characterized the floodplain and terraces along the river, and the creation of a generalized cross-section across the valley. The maps and cross-sections illustrated the relation of the recorded archaeological occurrences with landform sediment assemblages in the lower Big Fork drainage. The sedimentary sequences beneath the floodplain complex provided information on the stratigraphic relationships between artefact assemblages and on the role of natural processes in patterning the archaeological record. The main study area is located on the west bank of the Big Fork River, c. 0.5 km south of the confluence with the Rainy River (Fig. 2). A topographic map with 10 cm contour intervals was created for the lower terrace in the vicinity of the Hannaford archaeological
site. Archaeological excavations reached depths of 2 m below the surface. Stratigraphic profiles from these excavations were used in conjunction with the detailed analysis of 19 cores. Fifteen cores reached a maximum depth of 3.1 m (1-92, 1A-92, 1B-92, 1C-92, 2-92, 2a-92, 3-92 to 11-92). An additional three cores (1-94, 2-94 and 3-94) were made available by the Minnesota Department of Transportation, and a core was drilled from a platform on an oxbow lake on the east side of the river. The stratigraphic profiles and sediment cores of Big Fork alluvial deposits were used to interpret the record of past river activity associated with the formation of the floodplain and the geoarchaeological context of Late Woodland (late Holocene) artefactual occurrences. Stratigraphic exposures (archaeological sections) and cores were described in the field and selected sediment samples from the cores and sections were analysed for grain-size distribution and loss-on-ignition to determine per cent organic matter and per cent carbonate. To place the landforms and archaeological materials in chronological context, 33 samples were submitted to Beta Analytic, Inc., for accelerator mass spectrometry (AMS) radiocarbon dating. Ten samples are from the cores, 22 from the excavation units, and one is from the oxbow lake on the east side of the river (Table 1). All dates are reported as years 14C BP.
112
C. L. HILL ET AL.
Table 1. Radiocarbon measurements Radiocarbon age (years)
1s standard deviation
Laboratory number
Material
1240 1460
60 80
Beta-61879 Beta 54638
Charcoal Charcoal
1470 1330 980 1330 1650 1160 1350 930 1460 1240 670 720 1280 1020 1090 1160 1150 1020 1420 1310 1290 1360 2130
60 70 70 80 60 80 50 60 60 60 90 80 60 60 70 70 60 60 60 60 60 60 60
Beta-61890 Beta-61891 Beta-61888 Beta-54637 Beta-61878 Beta-54639 Beta-74334 Beta-61874 Beta-61886 Beta-61887 Beta-61889 Beta-61881 Beta-61882 Beta-61875 Beta-61880 Beta-61883 Beta-61884 Beta-61885 Beta-61876 Beta-61877 Beta-61872 Beta-61873 Beta-78699
Carbon crust Charcoal Charcoal Charcoal Carbon crust Charcoal Charcoal Charcoal Carbon crust Carbon crust Charcoal Charcoal Carbon crust Charcoal Charcoal Charcoal Charcoal Charcoal Charcoal Carbon crust Charcoal Wood Charcoal
1340
50
Beta-78701
Charcoal
1750
40
Beta-78702
Charcoal
1470
60
Beta-77958
Charcoal
5400 2140 2980
60 40 50
Beta-77959 Beta-77960 Beta-77961
Charcoal Charcoal Charcoal
2290 1830
60 50
Beta-77962 Beta-78697
Charcoal Charcoal
Results Results of this study document the presence of three geomorphological surfaces in the lower Big Fork River Valley. The relationship of these surfaces is depicted in Figures 2 and 3. The oldest and highest of these surfaces (T2) is c. 335 m above sea level (m a.s.l.). Test borings on the upper terrace on the east side of the Big Fork River revealed alluvial deposits overlying glacial lake deposits including c. 1.8 m of silt clay and varved clay (Birk & George 1976). The lowest surface (T0) is the present-day floodplain; it consists of a complex of surfaces ranging in elevation from 329 to 331 m a.s.l. A scarp separates the floodplain
Sample number and stratigraphic context 30-1, Area A, stratum 2, Blackduck artefacts 28-1, Area A, stratum 3 (below Blackduck artefacts) 92-1, Area A, stratum 4, not used 93-1, Area A, stratum 4, Early Blackduck 84-1, Area A, stratum 6a, Blackduck hearth 12-1, Area A, Early Blackduck 12-2, Area A, Early Blackduck 41-1, Area B, 45 – 40 cm, Blackduck hearth 80-31, Area B, 42 – 37 cm, top series 2-1, Area C, top series 53-1, Area D, stratum 23, not used 53-2, Area D, stratum 22, not used 88-1, Area D, stratum 20 (late Blackduck) 47-1, Area D, stratum 20 (late Blackduck) 47-2, Area D, stratum 19, not used 5-1, Area D, stratum 16 (Middle Blackduck) 46-1, Area D, stratum 16 (Middle Blackduck) 50-1, Area D, stratum 13 (Middle Blackduck) 51-1, Area D, stratum 4 (Early Blackduck) 51-2, Area D, strata 1 – 2 (Early Blackduck) 10-1, Area E, Late Blackduck (zone 1a) 10-2, Area E, Early Blackduck (zone 3) Core 3-92, top series Core 5-92, 300 – 275 cm, slackwater facies Core 9-92, west of Area A, 100.5 cm, bottom series Core 11-92, SE of Area B, 63 cm, bottom series Core 11-92, SE of Area B, 151.5 cm, bottom series Core 1-94, northwest of Area D, 200 cm, bottom series Core 2-94, 80 cm, bottom series (redeposited) Core 2-94, 238 – 236 cm, bottom series Core 2-94, 250 – 248 cm, bottom series (redeposited) Core 2-94, 268 – 266 cm, bottom series Core from oxbow lake, 95 – 94 cm
from the 335 m surface (T2) on the east side of the Big Fork. At the Hannaford and Smith Mound sites, the floodplain is underlain by strata that contain artefacts. On the east side of the Big Fork River there are several floodplain surfaces. For instance, the Smith site earthen mounds are on the 329 m surface. Also, there are distinct meander scars on the T0 surface delineating the locations of abandoned channels (Fig. 2). On the west side of the river, the Hannaford site lies at an elevation of c. 331 m a.s.l. Local swales are clearly visible on air photographs and satellite images. This topography continues from the edge of the Big Fork River westward for about 487 m with a general NW trend. Directly to the SSW of
HUMAN RESPONSES TO A FLUVIAL REGIME
113
Fig. 3. Cross-section A–B across the Big Fork River.
the Hannaford site is another set of ridges and swales that have a changing alignment. Point bar migration on T1 is to the NW (see Fig. 2). The ridge-and-swale topography is the geomorphological expression of an eastward lateral movement of the Big Fork River. To the west, a scarp separates the floodplain complex (T0) ridge-and-swale topography from the lowest terrace (T1, starting at c. 332 m a.s.l.). T1 contains clearly visible meander scars. The youngest channel of the middle terrace is marked by contours at an elevation of 332.23 m a.s.l. Meander scars indicate that a former channel cut laterally into the fill of a higher terrace (T2). The T2 surface has an elevation of c. 335 m a.s.l. The meander scar pattern west of the Hannaford site on the lower terrace (T1) is clearly older than the ridge-and-swale topography on the floodplain at the Hannaford site. The meander channels seem to have been cut by a river moving westward. Detailed studies of the strata that form the valley fill provide a basis for evaluating the alluvial processes active in forming the present floodplain. Based on the archaeological stratigraphic profiles and drill cores, we have developed a model of late Holocene river behaviour for the floodplain complex (T0). As a river channel shifts laterally, alluvium aggrades in a floodplain creating a bottom stratum or lateral accretion deposits. The top
stratum or vertical accretion deposits aggrade when discharge inundates the plain. Lateral accretion deposition is within and adjacent to the channel whereas vertical accretion sediments aggrade beyond the channel. Our model takes into consideration the variation in lithofacies to evaluate both synchronous and diachronous patterns of depositional environments. This is based on the concept that alluvium aggrades at different locations within a floodplain creating alluvial facies that can signify variability of depositional environments at a particular time or, for a stratigraphic sequence, changes that take place over time. The sedimentary architecture consists of ‘bottom series’ coarse-grained sediments and ‘top series’ fine-grained sediments. The bottom series is dominated by sediments formed by the process of lateral movement of the channel (such as point bar and channel bed deposits) under normal river conditions. The top series is composed primarily of fine-grained sediments formed only during the bank-full stage (higher than normal stage) and is exposed above the water level during most of the time, particularly those sediments on or near channel banks. The Hannaford site is divided into five excavation areas, A –E (Fig. 4). Area E is not shown in any figure, because of its scale. Area E was the
114
C. L. HILL ET AL.
Fig. 4. Location of excavated area, cores, and cross-sections (see Fig. 2).
site of a small excavation 50 m north of Area B. The stratigraphic context was not studied but there are two radiocarbon dates from this area. Area A is situated on the west side of the Hannaford site where stratum 1 is interpreted as a point bar deposit, the product of lateral accretion of the Big Fork River (Fig. 5). It is topographically represented by a subsurface ridge-and-swale stretching NW –SE across the excavated area. The sedimentary deposits in the swale became parallel to the modern surface with increasing height, indicative of the gradual filling of overbank flooding sediments after the formation of stratum 1. Strata 2–7 in Figure 5 are primarily the result of suspended-load deposition on the ridge-and-swale topography of the pre-existing point bar. This inference is based on several lines of evidence, including the relatively fine-grained sediments above stratum 1. Thus, strata 2 –7 can be interpreted as a top series in contrast to stratum 1, which appears indicative of lateral accretion and is thus designated as a bottom series. The top series strata contain Early and Middle Blackduck artefacts and also fish remains (consisting solely of lake sturgeon, Acipenser fulvescens). Fish remains dominate the faunal assemblage. Another indication of a subsistence focus on riparian resources is the presence of beaver, muskrat, and river otter remains
in contrast to very rare occurrences of deer, elk, and moose. The high number of harpoons in the Area A assemblage supports the contention that spearing sturgeon was a major activity. Table 1 presents a summary of stratigraphic contexts associated with the radiocarbon samples. According to the available radiocarbon dates and diagnostic artefacts, the bottom series deposits at Area A might have formed before c. 1400 14C BP (Fig. 5). The depositional sequence in Area B (Fig. 4) shows a similar pattern to that seen in Area A, consisting of both top series and bottom series sediments. The top series is composed of the top 60 cm of sediments formed by suspended-load deposition. Below the top series are layers of relatively coarse-grained sediments including laminated fine to medium and sandy silt. These constitute the bottom series and are a product of lateral accretion. Two radiocarbon dates from Area B indicate that top series sediments were deposited starting around 1300 14C BP (Table 1). Early and Middle Blackduck artefacts were recovered from the top series and fish dominate the faunal assemblage. Area C is situated on the top of the modern river levees on the east of the site (Fig. 4), where the top series might have started forming about 900 BP or later (Table 1). Area D is also situated on the
HUMAN RESPONSES TO A FLUVIAL REGIME
115
Fig. 5. Area A stratigraphic section.
116
C. L. HILL ET AL.
eastern side of the site (Fig. 4), where it is topographically about 30 cm higher than Areas A and B owing to its levee setting. Twenty-three strata were recorded (Fig. 6). Ages of 1020 and 1150 14 C BP (Table 1) from the bottom series (layers 1–3) could be from material in a secondary context. Eight radiocarbon dates are assigned to the top series in Area D (Table 1). Three dates ranging from 1160 to 1020 14C BP are associated with Middle Blackduck artefacts found in an organic-rich zone (layers 12–16). A very thin bed of sand (layer 15) contains no artefacts and probably represents a single flood. Layer 18 is composed of a series of discontinuous lenses of sand and contains no artefacts; it represents another large flooding event. Late Blackduck artefacts were recovered from layers 19 to 23. Layer 20 dates to about 700 14 C BP (Table 1). Thus the top series might have started forming after about 900 14C BP in the vicinity of Area D, where fish, especially redhorse suckers (Moxostoma macrolepidotum), dominate the faunal assemblage, indicating that spring fishing was a major activity. The riparian setting is also reflected by the presence of beaver, porcupine, muskrat, and river otter.
Fig. 6. Area D stratigraphic profile.
The locations of the cores in relation to the archaeological units are shown in Figure 4. Stratigraphic information from the cores is more detailed than the data available from the archaeological excavations, so it is the primary basis for our geomorphological interpretations. Core 9-92, about 5 m west of Area A, has a top series subdivided into several layers (Fig. 7). The bottom series consists mainly of fine to medium sands interbedded with silt and clayey silts dating to 2130 14C BP (Table 1). Cross-bedding structure and a finingupward sequence are present in the lower part of the coarse-grained deposits. Core 11-92, about 4 m SE of Area B (Figs 7 & 8), has a bottom series with dates of 1750 and 1340 14C BP (Table 1). The lower 1.2 m of the sequence is composed of fine to medium cross-bedded sand with several thin layers of silt and clayey silt. Overlying the bottom series of coarse-grained deposits is 10 cm of clay with three 1–2 cm thick dark lenses that may be organic-rich flood drapes (Mandel & Bettis 2001). On the top of this stratum is a 20 cm silty clay or clayey silt with blocky structure typical of pedogenically unmodified, organic-rich flood drapes with high clay content.
HUMAN RESPONSES TO A FLUVIAL REGIME
117
Fig. 7. Cross-section C–D.
Radiocarbon dates from both Core 11-92 and the excavation units suggest that the development of the top series started around 1300 14C BP in Area B (Table 1), about 100 years later than in Area A.
Fig. 8. Cross-section E– F.
Core 1-94 was drilled NW of Area D (Fig. 4). A layer of dark grey silt and clayey silt with some plant remains represents the top series deposits. Underlying the silt and clayey silt are the relatively coarsegrained sediments of the bottom series. These are
118
C. L. HILL ET AL.
gravelly sand and sandy gravel separated by 30 cm of massive silty clay with many wood and root remains interpreted as a slackwater facies that accumulated around 1470 14C BP (Table 1). Core 3-92 lies between Areas B and C (Fig. 4) and contains a relatively thin top series composed of clays dated to 1290 14C BP (Fig. 7 and Table 1). The bottom series consists of medium sand interlayered with very thin beds of silty clay that grades downward into cross-bedded coarse sand. Wood in the nearby Core 5-92 dated to 1360 14C BP (Fig. 7 and Table 1) was recovered from a 40 cm thick deposit overlain by medium sand that grades upward into fine sand and silt. The fine sand and silt could have been deposited on a floodplain by vertical accretion during a moderate or high-magnitude flood event. Thus, we interpret the layer containing wood remains as a slackwater facies prior to 1300 14C BP. Core 2-94 is situated on a 0.5 m high ridge SW of the archaeological excavations, 43.5 m from the excavation datum (Fig. 2). It has the same sedimentary architecture as seen within the site, including both top series and bottom series (Fig. 8). The lowest sediments are coarse sand and gravelly sand. They grade upward into fine to medium sand interbedded with thin beds of silty or sandy clay and silt showing cross-bedding structure and a fining-upward sequence. The gravelly sand and sand layers constitute the bottom series. Four radiocarbon ages were determined on charcoal from the bottom series in Core 2-94. A charcoal sample dated to 5400 14C BP may have been redeposited; the date appears unreasonably old in terms of the stratigraphic context. Three other dates range from 2980 to 2140 14C B.P (Table 1). Above the bottom series is 10 cm of dark greyish brown clayey silt containing few very thin organic lenses that grades upward to silty clay. The top series of relatively finegrained deposits and associated soils can be correlated with the stratigraphic sequence within the Hannaford site. Figures 7 and 8 are cross-sections based on the cores studied on the floodplain (T). In general, the depths of the organic-rich layers gradually increase eastward toward the present location of the river. The formation of the organic-rich strata can be attributed to inundation proximal to the channel with little if any contribution resulting from incipient pedogenesis or prehistoric human activity. After being deposited on the floodplain, the surface of the organic-rich layers would have been available for human activities. The sediments below the organic-rich deposits are composed mainly of fine to coarse sand interlayered with clay and are interpreted as the bottom series. The deposits above the organic-rich layers are always silty clay and silt; these fine-grained sediments are assigned to the
top series. The same bipartite sedimentary architecture occurs in the off-site Cores 2-94 and 3-94 and can be correlated to the archaeological excavations and on-site cores (Fig. 8). Based on the sedimentary sequence and stratigraphic correlation shown in excavation units and drill cores the sedimentary stratigraphy of the floodplain complex consists of two major facies, the bottom series and top series. The bottom series was created primarily by the process of lateral channel movement of the Big Fork River. It is composed of relatively coarse-grained sediments, including gravelly sand and coarse to fine sand with thin beds of silt and silty clay, showing a finingupward sequence. The sediments are topographically expressed as a point bar. The top series deposits consist of relatively fine-grained sediments, mostly silty clay and clay, deposited by overbank flooding. Deposition of these overbank sediments led to vertical accretion at the site. Artefacts and features in primary contexts are restricted stratigraphically to the top series. As the fine-grained sediments slowly accumulated on the point bar, humans were present on the floodplain and left a stratified archaeological record in the top series. The top series deposits become thicker toward the channel of the Big Fork River because a natural levee was built up immediately adjacent to the channel. Owing to the decreasing age of the underlying bottom series from SW to NE, the top series sediments show a decrease in age not only from bottom to top (vertically) but also from SW to NE (horizontally).
Discussion The Rainy River region of northwestern Ontario and northern Minnesota was affected by two late Wisconsin age glacial advances: the Rainy lobe from the NE and the St. Louis sublobe of the Red River –Koochiching lobe from the NW (Hill 2007). The Rainy lobe retreated to the NE, forming a succession of moraines including the Vermilion, Rainy River, and Eagle–Finlayson moraines. The area between the Vermilion moraine to the south and the Eagle –Finlayson and Brule Creek moraines to the north was deglaciated before 10 400 14C BP (Lowell et al. 2009). Following the retreat of the Rainy lobe, the St. Louis sublobe of the Red River –Koochiching lobe expanded to the SE. The glacial margin was located in the vicinity of the Rainy River moraine immediately north of the Big Fork area c. 11 000 14C BP (Bajc 1991), which generally corresponds to the time Clovis artefacts occur throughout North America. Melting of the St. Louis sublobe led to the formation of glacial Lake Agassiz.
HUMAN RESPONSES TO A FLUVIAL REGIME
Lake Agassiz initially formed around 12 000 14C BP and had a low-water stage (Moorhead phase) from about 10 500 to 9300 14C BP (Bajc et al. 2000; Fisher et al. 2008) when the lower Big Fork area was exposed to subaerial conditions. Human populations using Folsom artefacts are generally dated to about 10 500 14C BP (c. 10 800 –10 300 14 C BP); thus, the Moorhead phase represents a time in landscape evolution when Folsom or younger human populations could have been present. On the north side of the Rainy River, an exposure east of the confluence of the La Vallee River with the Rainy River (Fig. 1) shows the regional sequence (Nielson et al. 1982). Sands of the Moorhead phase overlie clays deposited in an earlier phase of Lake Agassiz. The lake clays are above calcareous till. Reddish brown clay with varves overlies the Moorhead sands and represents Emerson phase Lake Agassiz deposits, which continued to about 9500 14C BP (Bajc et al. 2000). South of the Rainy River, exposures in gravel pits show Rainy lobe till overlain by laminated lake clays and silts. The latter underlie outwash attributed to the St. Louis sublobe or Koochiching lobe. Deposits of glacial Lake Agassiz have been designated as the Little Fork Formation (Ojakangas & Matsch 1982). The upper terrace on the east side of the Big Fork River contains lacustrine deposits at 329 –330.8 m a.s.l., buried by alluvium (Birk & George 1976). After drainage of Lake Agassiz by about 9500 14C BP (Bajc 1991) stream channels were formed on the old lake plain and alluvial deposition began. The initial post-glacial course for the Big Fork River may have followed the lower reaches of the Black River to the west, when the Black River flowed into the Rainy River about 7 km downstream from the present confluence. A Paleoindian site, the Plummer site (Magner 2001), is situated on a terrace of this earlier combined Big Fork–Black River drainage. The Little Fork River also entered the Rainy River downstream from its present location sometime in the past. This study examined post-glacial alluvial processes and related landforms in the vicinity of the Hannaford archaeological site. Results indicate that three major sets of Holocene terraces are present in the lower reach of the Big Fork River near its confluence with the Rainy River. These terrace surfaces reflect stream responses to Holocene environmental change. The earliest or ‘upper’ terrace (T2) is underlain by Lake Agassiz sediments (Fig. 3). Locally, a treeless tundra-like landscape and spruce parkland existed adjacent to the retreating ice margin and Lake Agassiz. By the time the ice margin was at the Hartman moraine, 130 km NE of the Rainy River basin, closed spruce forests had expanded into northwestern Ontario from northern Minnesota (Janssen 1968).
119
Following the interval of entrenchment that left the early Holocene floodplain as the T2 terrace, the Big Fork River went through an episode of aggradation. This episode was followed by entrenchment that created the T1 terrace, which has numerous palaeomeanders on its surface. A prominent scarp separates the T1 terrace from the floodplain complex. Channel scars and ridge-and-swale topography characterize the floodplain surface. The valley fill beneath the floodplain complex had begun to aggrade by 3000 14C BP. Changes in the dynamics of Holocene fluvial processes have been related to variations in climate. Middle Holocene dry climates (‘Hypsithermal’; see McAndrews 1982) have been linked to cases of river entrenchment and valley alluviation. For instance, middle Holocene aridity and climate instability have been connected to reduced vegetation and aggradations in the Upper Mississippi Valley (Knox 1972). Middle Holocene aridity led to the eastward and northward expansion of the prairie. East of the Big Fork River at Cayou Lake, Voyageurs National Park, spruce dominated to around 9000 14C BP and pine became dominant until 6000 14C BP, when climate became its warmest and driest (Davis et al. 2000). It appears that regionally there was a significant change in effective moisture during the middle Holocene, resulting in lower lake levels (Laird & Cumming 2008). Intensified lateral migration has been attributed to higher flooding frequency and cooler and moister conditions between about 6000 and 4500 14 C BP (Knox 1985). In contrast, higher lateral channel migration and floodplain construction occurred during a period of warmer –drier climate along the Red River with no intensification of fluvial activity after the shift to a cooler –moister climate (Brooks 2003). In other situations where there apparently was a change in forest composition, periodic flooding may be connected with persistent downcutting (Arbogast et al. 2008). After a period of floodplain deposition there was another major period of incision, this time into the T1 surface. This period of downcutting appears to have occurred prior to 3000 years ago. The downcutting led to the set of surfaces that are associated with the present-day floodplain (grouped together as T0). During this stage, the Big Fork continued to meander, resulting in a series of point bar deposits. Assuming a constant rate of deposition, the radiocarbon date of 1850 14C BP (Table 1, recovered at a depth of 95 cm for the 1.48 m core) from the oxbow lake adjacent to the Smith site implies that valley filling began before c. 3000 14C BP; this is similar to the radiocarbon age of redeposited charcoal in the lower section of Core 2-94 on the west side of the river (Fig. 8). Palaeoclimate evidence from the region suggests that climate became
120
C. L. HILL ET AL.
cooler and moister after about 5000–3000 14C BP. Myrtle Lake, 70 km to the SE, was dominated by pine after 3000 BP (Janssen 1968). The shift to wetter climates may have led to a rise in the regional water table, leading to the spread of peatlands and increasing the headward erosion of streams (Glaser et al. 1997). The hydroclimatic record from Hole-in-the-Bog peatland south of the Big Fork drainage shows extreme drought events associated with fluctuations during the Medieval Warm Period or Climatic Anomaly (Laird et al. 1996; Booth et al. 2006). Isostatic depression by the Laurentide Ice Sheet has resulted in post-glacial rebound. This is demonstrated by the differential tilting of Lake Agassiz shorelines. How has the rate of uplift affected the Big Fork stream gradient? Was it initially very fast, leading to incision? Did later slower rebound contribute to floodplain stability? The gradients and base levels of river systems affected by differential uplift in areas deglaciated during the late Quaternary include river incision and terrace formation as well as valley aggradation (Brooks et al. 2005). Differential uplift on the Big Fork might mirror the Red River to the west where a 60% loss of gradient has occurred in the last 8000 years. The Lower Campbell beach shows that from the southern outlet of Agassiz to Lake Winnipeg there has been uplift of 64 m. Long-term changes in fluvial processes associated with uplift may be important to archaeological site locations (settlement patterns) along low-gradient rivers (Brooks et al. 2005). In the Big Fork drainage, based on the difference between isobase 5 and 6 (Teller 2001), vertical displacement by uplift-rebound may be of the order of c. 40 m since the formation of the Upper Campbell beach around 9500 14C BP. The distribution of artefacts appears to reflect changing human use of parts of the late Holocene floodplain and site microenvironment. The earliest documented human presence at the Hannaford site is associated with Early Blackduck artefacts starting at c. 1300 14C BP, found on the west side of the Hannaford site. These artefacts are confined to Area A, which at the time was adjacent to the Big Fork River; the sediments at present east of Area A had not yet been deposited because that part of the floodplain was probably the location of the river (Figs 8 & 9). Some possible Early Blackduck artefacts have been recorded in the eastern part of the site in deeply buried but artefact-poor deposits along the current river bank; these could be the result of secondary deposition resulting from fluvial activity. By c. 1050 14C BP Middle Blackduck artefacts are present at the site. The most extensive Middle Blackduck presence was also in Area A, but some river bank areas to the east may have also been used. It is possible to divide the Middle Blackduck
Fig. 9. Four-stage model of landscape evolution.
assemblages into two temporal groups. The first relates to the extensive use of Area A and minimal use of the river bank. The second interval involves a shift in emphasis from Area A to Area D, closer to the present-day river; during this time Area A appears to have received less use than the river bank. Most Middle Blackduck zones along the river are separated from the Late Blackduck artefact
HUMAN RESPONSES TO A FLUVIAL REGIME
zone by sediments without artefacts. Late Blackduck (¼ Rainy River Late Woodland Complex) artefacts are found along the river bank at c. 900 14 C BP. The most recent dated Lake Blackduck phase artefacts occur in Area D at c. 650 14C BP. The sedimentary sequence at Hannaford records a bipartite pattern of lateral channel accretion and vertical overbank accretion over the past 3000 years or more. This pattern produced a changing microenvironmental landscape context that in turn influenced site selection and intensity of use, and affected the distribution and preservation of archaeological occurrences. On the basis of sedimentary stratigraphy, artefact assemblages, and radiocarbon dates, a four-stage model can be proposed showing the interrelationships between sedimentation and human activity that reflect the processes of site formation (Fig. 9). Before about 1500 14C BP, prior to the onset of the Medieval Climate Anomaly, the locus of the Hannaford site was still within the channel of the Big Fork River (Fig. 9, Stage I). The river was migrating eastward and southward by means of lateral channel movement (i.e. concave bank erosion and convex bank (point bar) deposition). Along the river, the convex banks form the locations of deposition on gently sloping accretionary landforms (point bars). The Hannaford site was located on the northernmost point bar on the east side of the Big Fork River. Because of lateral accretion, the point bar was gradually increasing in height and migrating toward the NE. Lateral accretion resulted in the deposition of relatively coarse sediments including sandy gravel, and coarse to fine sand showing a fining-upward sequence. The relatively coarse-grained sediments constitute the bottom series of the stratigraphic sequence. As the point bar deposition continued, the subaqueous point bar surface emerged gradually from SW to NE. The topography of the emerging point bar consisted of a series of roughly concentric ridges and swales. Around 1400 14C BP the point bar surface in the west side of the Hannaford site (Area A) was emerging from the channel as a result of lateral accretion. Human use began to take place on the west part of the site around 1400 14C BP after the sedimentary environment had gradually shifted from primarily lateral to vertical accretion. After subaerial emergence, the point bar remained relatively stable and started receiving vertical accretion sediments created by intermittent overbank flooding. The floodplain would have been dry most of the year and intermittently inundated during relatively brief periods of overbank flooding. The site would have been occupied during the period of normal stream flow; however, human activity was not feasible when the floodplain was inundated by floodwaters.
121
While the vertical accretion deposits were accumulating upward, the point bar continued to migrate gradually to the east by deposition in the channel bed. By about 1300 14C BP (Fig. 9, Stage II), the channel had migrated to the east of Area B. Early Blackduck artefacts from Areas A and B are associated with an organic-rich zone consisting of a series of thin layers in the lower part of top series deposits. Sometime after 1100 14C BP, roughly corresponding to the beginning of the Medieval Climatic Anomaly (Mann et al. 2009), a relatively large flooding event covered the site with a thin (3–4 cm) bed of fine sand, designated layer 5 in Area A (Fig. 6). Like other meandering rivers, the Big Fork River inundated the floodplain periodically. Sediments suspended in the floodwaters accumulated on the upper portion of the point bar. Early Blackduck artefacts in the lower strata in Areas C, D, and E are believed to be in secondary context. This inference is supported by stratigraphic position, sedimentary context, and inverted radiocarbon dates. The Early Blackduck artefacts are embedded in the bottom series sediments that were created by point bar deposition on the channel bed. Some pottery fragments found at the depth of 1.98 m in Core 5-92 could be in a similar situation. These redeposited artefacts might be derived from the early phase of human presence in Areas A and B during the Early Blackduck period when the river channel bank close to Area B and Area D was within the river channel. The Early Blackduck artefacts at 85 –100 cm depth in Area C are probably also in secondary context. As the coarse sediments continued to accumulate on the channel bed by point bar deposition, the bank of the river moved to its current position (east of Areas D and C), probably around 900 14C BP (Fig. 9, Stage III). The water table was probably lower during this time as a result of decreasing stream discharge. The point bar became stabilized owing to the development of a younger point bar system south of the site. Vertical accretion deposits began to accumulate on the stabilized point bar platform. As a result, a natural levee started to form and a relatively thick sequence of vertical accretion sediments was deposited. The newly formed bank became a location for human activity. Humans were there for the riparian resources and the human–environment interactions were chiefly the human responses to changing conditions along a dynamic river bank. In the mean time, the previous locations of occupation were still utilized during normal stream flow periods. In addition to the normal-magnitude overbank flooding, relatively major flooding occurred occasionally; at least two large-magnitude floods occurred. The flooding deposited a thin layer or lens of fine sand over the
122
C. L. HILL ET AL.
occupied area near the channel. Continued use of the river bank areas is reflected by the presence of younger artefacts. The youngest dated Blackduck artefacts (Rainy River Late Woodland Complex) are in Area D at c. 650 14C BP. The vertical accretion deposits are the locations of primary archaeological records at the Hannaford site. As a result of flooding, artefacts and features associated with the human activities on the floodplain surface were buried as fine-grained sediments settled from suspension during the waning stages of stream flow. After flooding, the site was used again. Sometimes the site was covered with a relatively thick layer of sediments produced by a single major flooding event. Thus, the new activity surface could be separated by a layer of sediment from the set of older artefacts. As a result of repeated intermittent periods of landscape stability and human use of the floodplain from about 1500 to 600 14C BP, stratified layers containing artefacts were incorporated within the top series sediments. The site is still subject to the episodic deposition of overbank flooding (Fig. 9, Stage IV). The four-stage model of site formation presents a picture of how sedimentary environments and associated deposits dynamically affected the distribution and preservation of artefact assemblages on the floodplain of the Big Fork River. The alluvial model has implications for interpreting the archaeological records at other sites with similar geological settings. In a meandering point bar setting, artefacts in primary contexts are normally found in association with vertical accretion sediments instead of lateral accretion sediments of the floodplain. Artefacts could be incorporated into lateral accretion deposits, but usually only in secondary contexts. Erosion of archaeological materials can occur along stream banks of the river.
Conclusions This study provides a reconstruction of environmental events along the lower Big Fork River, relates the sedimentological processes to the archaeological record, and examines human activities related to a dynamic riparian environment and its resources. By the time that people were using Clovis artefacts, around 11 000 14C BP, glaciers had melted north of the study area to the Rainy River moraine. At the time human groups were using Folsom artefacts, around 10 800 –10 300 14C BP, the Big Fork area was available for human habitation as a result of the Moorhead low stage of Lake Agassiz. The geomorphological and stratigraphic records reflect the fluvial dynamics of the Big Fork River once it became established on the lake sediments of glacial Lake Agassiz after c. 9500 14C BP. The
river has since eroded a valley that contains a floodplain complex and two terraces near its confluence with the Rainy River. The valley fill beneath the oldest and highest terrace (T2) contains Archaic artefacts and overlies lacustrine deposits of glacial Lake Agassiz. The T1 fill represents an aggrading floodplain; this terrace surface consists of distinct meanders. The youngest valley fill underlies the floodplain complex; aggradation is estimated to date to at least 3000 14C BP. The curvilinear pattern of ridge-and-swale topography across T1 and T0 indicates that these floodplains are the result of lateral migration of the stream channel. After a period of middle Holocene entrenchment and drier climates, the lower terrace appears to be linked to overall wetter climate conditions punctuated by intervals of drought. This latest phase of aggradation appears to have lasted several thousand years during which the dynamics of the river have been influenced by more modest fluctuations in climate. The Hannaford site is a late Holocene, multicomponent, stratified locality situated on the west bank of the lower Big Fork River in northern Minnesota. Its geological setting in the active floodplain of a meandering river produced a series of discrete sedimentary strata that comprise two general groups. The lower, coarse-grained strata (bottom series) are primarily the result of point bar deposition when the river was farther west of its present position. The upper, finer-grained sediments (top series) represent deposition from various overbank flooding events. This boundary is timetransgressive; it developed as the river channel migrated east, and so it is older in the west. Contained within these strata are Late Woodland artefacts that were used by prehistoric hunter – gatherers. Most of the artefacts are located within the upper sediments and are in primary context, although affected to some degree by the flooding events that buried them. Artefacts within point bar deposits are more likely to be secondary deposits. Radiocarbon dates from the floodplain imply that aggradational processes have dominated at least since 3000 14C BP. The alluvial geomorphological setting provided a microenvironmental context associated with fishing and aquatic mammals. There appears to have been a shift in the types of fish over time; there is a change from the procurement of lake sturgeon by spearing (as represented on the west side of the site in Area A) to taking bottom-feeding fish such as redhorse (on the east side of the site at Area D). These fishing stations may have been seasonal to take advantage of the spring fishing runs. A more permanent human presence would have been difficult in a riparian setting characterized by flooding. The radiocarbon dates indicate that some reworking (erosion and redeposition) of artefacts
HUMAN RESPONSES TO A FLUVIAL REGIME
has occurred. The cross-sections show a series of aggrading river deposits. Based on these crosssections we propose that there was a progressive expansion of land toward the east and north associated with the human prehistoric presence at Hannaford from about 1300 to 650 14C BP. Regionally, extreme drought events at 1000, 800, and 700 BP within this time period of human presence are associated with climate fluctuations during the Medieval Climatic Anomaly. The transition from the early part of the sequence characterized by Early and Middle Blackduck artefacts to the late part of the sequence containing Rainy River Late Woodland Complex artefact assemblages along the Big Fork River roughly coincides with the beginning of the Medieval Climatic Anomaly at 1100–1050 14C BP. The youngest part of the archaeological sequence dates to before the onset of the Little Ice Age. This research was conducted partially under Minnesota Department of Transportation Agreement 70974, project manager G. J. Hudak. Laboratory analyses for the project were conducted by personnel at the University of Minnesota’s Archaeometry Laboratory with contributions by S. C. Mulholland, S. L. Mulholland, D. E. Stoessel, S. H. Valppu, J. K. Huber, and C. Elmgren. Significant work on the manuscript was done by C. Kubeczko and C. Wofford-Hill. Two anonymous reviewers made excellent suggestions that improved the paper.
References Arbogast, A., Bookout, J., Schrotenboer, B., Landsdale, A., Rust, G. & Bato, V. 2008. Post-glacial fluvial response and landform development in the upper Muskegon River valley in North– Central Michigan, U.S.A. Geomorphology 102, 615–623. Arzigian, C. 2008. Minnesota statewide multiple property documentation form for the woodland tradition. University of Wisconsin, Mississippi Valley Archaeological Center Report, 735. Bajc, A. 1991. Glacial and Glaciolacustrine History of the Fort Frances– Rainy River Area, Ontario, Canada. University of Waterloo, Waterloo, Ont. Bajc, A., Schwert, D., Warner, B. & Williams, N. 2000. A reconstruction of Moorhead and Emerson Phase environments along the eastern margin of glacial Lake Agassiz, Rainy River basin, northwestern Ontario. Canadian Journal of Earth Sciences, 37, 1335–1353. Benito, G., Baker, V. & Gregory, K. (eds) 1998. Palaeohydrology and Environmental Change. Wiley, Chichester. Birk, D. & George, D. 1976. A woodland survey strategy and its application to a late archaic locus of the Smith site (21KC3), Koochiching County, Minnesota. Minnesota Archaeologist, 35, 1 –30. Booth, R., Notaro, M., Jackson, S. & Kutbach, J. 2006. Widespread drought episodes in the western
123
Great Lakes region during the past 20 000 years: geographic extent and potential mechanisms. Earth and Planetary Science Letters, 242, 415–427. Bridgland, D. & Westaway, R. 2008. Climatically controlled river terrace staircase: a worldwide Quaternary phenomenon. Geomorphology, 98, 285– 315. Brooks, G. 2003. Holocene lateral channel migration and incision of the Red River, Manitoba, Canada. Geomorphology, 54, 197– 215. Brooks, G., Thorleifson, L. & Lewis, C. 2005. Influence of loss of gradient from postglacial uplift on Red River flood hazard, Manitoba, Canada. Holocene, 15, 347– 352. Bull, W. 1991. Geomorphic Responses to Climatic Change. Oxford University Press, New York. Butzer, K. 1982. Archaeology as Human Ecology: Method and Theory for a Contextual Approach. Cambridge University Press, Cambridge. Davis, M., Douglas, C., Calcote, R., Cole, K., Winkler, M. & Flake, R. 2000. Holocene climate in the Western Great Lakes National Parks and Lakeshores: implications for future climate change. Conservation Biology, 14, 968– 983. Fisher, T., Yansa, C., Lowell, T., Lepper, K., Hajdas, I. & Ashworth, A. 2008. The chronology, climate, and confusion of the Moorhead Phase of glacial Lake Agassiz: new results from the Ojata Beach, North Dakota, USA. Quaternary Science Reviews, 27, 1124– 1135. Glaser, P., Siegel, D., Romanowicz, E. & Shen, Y. 1997. Regional linkages between raised bogs and the climate, groundwater, and landscape of northwestern Minnesota. Journal of Ecology, 85, 3– 16. Gregory, K. & Benito, G. (eds) 2003. Palaeohydrology: Understanding Global Change. Wiley, Chichester. Hill, C. 2007. Geoarchaeology and Late Glacial landscapes in the western Lake Superior region, central North America. Geoarchaeology, 22, 15–47. Hill, C., Rapp, G., Valppu, S. & Jing, Z. 1995. Geoarchaeology and geochronology of the Hannaford site. Archaeometry Laboratory Report, 95-7. Janssen, C. 1968. Myrtle Lake: a late- and post-glacial pollen diagram from northern Minnesota. Canadian Journal of Botany, 46, 1397 –1408. Knox, J. 1972. Valley alluviation in southwestern Wisconsin. Annals of the Association of American Geographers, 62, 401– 410. Knox, J. 1983. Responses of river systems to Holocene climates. In: Wright, H. E. (ed.) Late Quaternary Environments of the United States, Volume 2. University of Minnesota Press, Minneapolis, 26– 41. Knox, J. 1985. Responses of floods to Holocene climatic change in the Upper Mississippi Valley. Quaternary Research, 23, 287–300. Knox, J. 1993. Large increases in flood magnitude in response to modest changes in climate. Nature, 361, 430– 432. Laird, K. & Cumming, B. 2008. Reconstruction of Holocene lake level from diatoms, chryosophytes and organic matter in a drainage lake from the Experimental Lakes Area (northwestern Ontario, Canada). Quaternary Research, 69, 292–305. Laird, K., Fritz, S., Maasch, K. & Cumming, B. 1996. Greater drought intensity and frequency before A.D.
124
C. L. HILL ET AL.
1200 in the northern Great Plains, U.S.A. Nature, 384, 552– 554. Lenius, B. & Olinyk, D. 1990. The Rainy River Composite: Revisions to Late Woodland taxonomy. In: Gibbon, G. (ed.) The Woodland Tradition in the Western Great Lakes. University of Minnesota, Publications in Anthropology, 4, 77– 112. Lowell, T., Fisher, T., Hajdas, I., Glover, K., Loope, H. & Henry, T. 2009. Radiocarbon deglaciation chronology of the Thunder Bay, Ontario, area and implications for ice retreat patterns. Quaternary Science Reviews, 28, 1597–1607. Lugenbeal, E. 1978. The Blackduck ceramics of the Smith site (21KC3) and their implications for the history of Blackduck ceramics and culture in northern Minnesota. Mid-Continental Journal of Archaeology, 3, 45– 68. Magner, M. 2001. The Lake of the Woods/Rainy River Late Paleoindian Complex: an initial assessment from the Rainy River Region in Minnesota. Minnesota Archaeologist, 60, 87– 98. Mandel, R. & Bettis, E. 2001. Use and analysis of soils by archaeologists and geoscientists: A North American perspective. In: Goldberg, P., Holliday, V. & Ferring, C. (eds) Earth Sciences and Archaeology, Kluwer– Plenum, New York, 173– 204.
Mann, M., Zhang, Z. et al. 2009. Global signatures and dynamical origins of the Little Ice Age and Medieval Climatic Anomaly. Science, 326, 1256– 1260. McAndrews, J. 1982. Holocene environment of a fossil bison from Kenora, Ontario. Ontario Archaeology, 37, 41–51. Nielsen, E., McKillop, W. & McCoy, J. 1982. The age of the Hartman moraine and the Campbell beach of Lake Agassiz in northwestern Ontario. Canadian Journal of Earth Sciences, 19, 1933– 1937. Ojakangas, R. & Matsch, C. 1982. Minnesota’s Geology. University of Minnesota Press, Minneapolis. Rapp, G., Mulholland, S. C. et al. 1995. Final Report: Hannaford Data Recovery Project, Koochiching County, Minnesota. University of Minnesota, Archaeometry Laboratory Report, 95-31. Schumm, S. & Brakenridge, R. 1987. River responses. North America and adjacent oceans during the last deglaciation. In: Ruddiman, W. & Wright, H., Jr. (eds) The Geology of North America, Vol. k-3. Geological Society of America, Boulder, CO, 221–240. Teller, J. 2001. Formation of large beaches in an area of rapid differential isostatic rebound: the threeoutlet control of Lake Agassiz. Quaternary Science Reviews, 20, 1649– 1659.
Human–environment interactions in the development of early Chinese civilization GEORGE RAPP1* & ZHICHUN JING2 1
Department of Geological Sciences, University of Minnesota, Duluth, MN 55812, USA
2
Department of Anthropology, University of British Columbia, Vancouver, BC V6T 1Z1, Canada *Corresponding author (e-mail:
[email protected]) Abstract: Beginning with the earliest organized habitation sites the options provided by the regional environment have largely or partially governed the location and relocation of human settlements. The settlement system in second millennium BCE Henan Province, China, evolved during a period of significant climatic change and shifting river courses but relative soil stability. Human–environment interactions across the landscape have left ample remains for investigation by scholars of social and cultural change and by natural scientists. The social effects of climate and geomorphological change during this period are complex and only partially understood. It is well documented that long-term soil stability before and during the second millennium BCE gave rise to the development of good agricultural soils, without which population expansion probably could not have taken place. This paper summarizes some of the recent research in climate change and, from two of our own projects, in geomorphology and ecology that underlie environmental impacts on the evolving state-level societies, especially related to settlement location and relocation. For example, the Shang possibly relocated one or more capital sites in response to disastrous floods.
We have been engaged in investigating aspects of the co-evolution of Shang society (c. 1550–1046 BCE) with environmental factors since 1990 in two distinct projects: the first in the Shangqiu area, the second in the Anyang area (Fig. 1), both in Henan Province. This paper weaves some of our work into the emerging picture of the changing environment in these areas during the period c. 2200–800 BCE. Although our studies focused on the Shang, we start at 2200 BCE because of the global-scale climatic event called the ‘4.2 K Event’. It also represents the approximate beginning of the period of transition from the Neolithic to the Bronze Age in China and the emergence of the earliest cities and states. The natural environment provides an underpinning for the cultural landscape. This paper deals with some aspects of the natural environment that possibly supported or, conversely, restricted the forces of social and cultural development of the Three Dynasties. The Shang had the first literate society in East Asia. Their core area was located on the North China Plain in Henan Province north and south of the Yellow River in the region where the river turned to the east (Fig. 1). However, Shang cultural influence reached far beyond the area their rulers controlled directly. The interaction of natural and human factors is a complex sequence of feedback processes, but we know that Shang society evolved during a period of climatic and geomorphological change. This
paper considers the natural environment from the perspective of Earth science, including geology, geomorphology, pedology and climatology, with some admixing of ecology. In a human time frame the earth-surface environment has static components such as mountains and geological deposits. However, many environmental factors are dynamic and change over time. These include climate, aspects of the landscape including shifting river courses, erosion and deposition, shifting patterns of vegetation, and soil development and degradation. The natural environment both supported and vexed the Shang. Among the objectives of their practice of divination was the assessment of environmental threats. The changing climate and the evolving landscape influenced a series of social decisions regarding habitation sites, transportation routes, diet and food supply, and exploitation of or exchange for raw materials such as jade, copper, tin, stone, salt and other resources. As the Neolithic merged into the Bronze Age the rise of state formation and generally increasing population put greater pressure on forest, food, water, soil and metal resources. Nearly every major human event during Shang times had some concrete interaction with the environment. People are geomorphological agents, and human activities modify almost every component of the landscape. In terms of settlements and population in this region, there was a remarkable reduction following the Neolithic Longshan period and into the
From: Wilson, L. (ed.) Human Interactions with the Geosphere: The Geoarchaeological Perspective. Geological Society, London, Special Publications, 352, 125– 136. DOI: 10.1144/SP352.10 0305-8719/11/$15.00 # The Geological Society of London 2011.
126
G. RAPP & Z. JING
Fig. 1. Map showing the location of Shang capitals at Zhengzhou and near Anyang.
Predynastic Shang. This phenomenon apparently was widespread (deMenocal 2001). However, by Early Shang there was again an increase in population and settlements that continued through the Late Shang.
The climate Climate change can be considered in terms of normal fluctuations, long-term trends, and highimpact change as well as the geographical scope of the impact. The 4.2 K Event was felt on a global scale. In Egypt it coincided with the fall of the Old Kingdom c. 2180 BCE (Hassan 1997). There is good evidence that the 4.2 K Event was felt in China, where the monsoonal belts were depressed southward, yielding a drier and cooler episode for several centuries, as described by Redman et al. (2007). Those researchers suggested that among dry-field farming and stock-breeding people of the Yellow River region this climatic perturbation seriously affected agricultural productivity and pastoral economics, with a drop in population and an increase in population mobility. Environmental and social transformations marked the collapse of Neolithic cultures in central China during the late third millennium BCE when severe climatic changes occurred across much of China (Wu & Liu 2004). Yu et al. (2000) have provided data on the role of climate in the fall of Neolithic cultures on the Yangtze Delta. Keightley (1999a, 2000) has summarized our knowledge of the Shang climate up to 2000 BCE. Palaeoclimatologists such as Bryson (1993) divide the Holocene into global climatic episodes. The reign of the Shang fits neatly into the SubBoreal II period from 2100 to 950 BCE.
New information on second millennium BCE climate change in eastern China continues to emerge. Yancheva et al. (2007) have presented a high-resolution look at the strength of the East Asian winter monsoons over the last 16 000 years using magnetic properties and the titanium content of sediments from Lake Huguang Maar in SE China. Those workers have shown that major changes in the Chinese dynasties (including the Shang) occurred when the winter monsoon was strong. Although we do not suggest that climate change was the major factor in the fall of the Shang Dynasty, environmental stresses can contribute significantly to political and social crises. Dykoski et al. (2005) presented a continuous record of the Asian monsoon over the last 16 000 years from d18O measurements of stalagmite calcite from Dongge Cave, China. Their data show a decline of monsoon intensity in mid- to late Holocene characterized by an abrupt positive shift in d18O that occurs about 1550 years BCE, almost exactly at the time of the transition from the Xia to the Shang as the dominant dynasty. Feng et al. (1993) presented data showing a cool episode beginning about 1000 BCE, at the transition from the Shang to the Zhou, and argued that prior to this time there was a warm, moist, ‘optimal’ period (i.e. during the Shang Dynasty). A multi-disciplinary research effort in the southern Loess Plateau of China (Huang et al. 2002) determined an increase in climatic aridity at 1100 BCE, leading to a degradation of land resources. Huang et al. suggested that severe drought and a famine were the causes of social instability and eventual collapse of the Shang Dynasty. We agree that severe changes in the climatic regime probably increased social stress but doubt that this was a leading cause of Shang collapse. We are also aware that there is, at best, only a tenuous relationship between correlation and causation. In another paper Huang et al. (2007a) posited that research into human–environment interactions during past global changes at a landscape scale offers a way to integrate diverse evidence and allow the changes to be traced through time. Based on their work on Holocene pedogenic change along with climatic and anthropogenic factors they recorded rapid climatic change from 2200 to 1800 BCE, consistent with the work of others. Their data show accelerated erosion and redeposition from 1100 BCE onward as a result of intensified cultivation along with dense settlement and climatic decline. This is consistent with our pollen data presented below. Mayewski et al. (2004) have studied Holocene climate variability and recorded six periods of rapid climate change including 2200–1800 and 1500–500 BCE, periods of focus for this paper. Their work indicates that most of the climate
HUMAN AND ENVIRONMENT IN EARLY CHINA
change events are characterized by polar cooling, tropical aridity, and major atmospheric circulation changes. High-impact climate change occurs on all time scales from years to millennia. For example, emissions of sulpur dioxide and dust from volcanoes adversely affect cultivated crops for 1 or 2 years but unless persistent they only rarely cause significant social changes. Volcanic emissions travel globally, so the lack of local volcanism does not protect an area from impact. Archaeologically we find food storage silos dating back thousands of years, built to maintain reserves to respond to short-term drought and to crop failure in general. One climatic change during the second millennium BCE was in the distribution of the monsoon belts in eastern China, leading to drought in the north and flooding in the south. In the North Atlantic realm there was a cold spell centred on 2000 BCE, usually referred to as the ‘Holocene Event 3’ in the Bond cycle (Bond et al. 1997). A similar change also can be identified from geological records such as the d18O curve from the Dunde Ice Core of Mt. Qilian and the pollen and micropalaeontological data from drill cores in the Tibetan Plateau and the South China Sea. These data indicate a cooler and drier episode between 2000 and 1500 BCE; its effect on human populations was widespread. In fact, the climate deterioration between 2000 and 1000 BCE seems to be worldwide (Shi et al. 1993; Winkler & Wang 1993; deMenocal 2001). Data on the Near East have been provided by Weiss et al. (1993). The forcing mechanism for this widespread event is being debated. The reduced solar output has been suggested as a possible cause (van Geel et al. 1996). The Shang climate was controlled by the East Asian monsoons, which brought summer rains from the SE (An et al. 2000). In the winter the Siberian high-pressure system dominated. Compared with other regions of the Earth at the same latitude, the monsoon precipitation in China shows the largest variability (Ding 1994). Although not yet understood in any detail, the social responses to these severe climatic impacts are complex and undoubtedly varied by region. The rise of complex societies in China appears to have occurred in a period of environmental stress. A major climatic regime change came at or near the time of collapse of the Neolithic cultures and the rise of state-level society (Liu 2000). It should be noted that the impacts of climate change and human activities are always superimposed on one another; sorting this out requires both data and diligence. The reconstruction of palaeoclimates involves largely the specification of the distributions and amplitudes of precipitation and temperature over
127
time and space. The variation and trends for precipitation are more complex than for temperature. We derive prehistoric precipitation and temperature regimes from vegetation proxies (especially from pollen data), isotope variations, and faunal associations. That the elephant and the rhinoceros once lived in northern Shang territory is now well known, indicating that the climate was warmer. It should be noted that trees trade off the effects of temperature and precipitation in terms of the growth rings that are used as climate-change proxies: colder and drier have the same effect as warmer and wetter. Based on the migration of vegetation zones, midHolocene precipitation was more plentiful than at present, but precise rainfall data are lacking. We do know that the deciduous forests in the Shang heartland in 2000 BCE would have required about 60 cm of rainfall per year. Dependable rainfall amounts dominate agricultural outcomes, such as millet v. rice harvests. The average temperature in the lower Yellow River area from 3000 to 1100 BCE was about 2 8C warmer, and the winter temperature was 3–5 8C higher than today, as described by Shi et al. (1993). Those workers used the Dunde Ice Core records to report a 1000 –900 BCE higher temperature event followed by a period of severe fluctuation with decreasing temperature and general environmental deterioration. Elvin (1993) also reported a shift to colder weather around the beginning of the first millennium BCE, the time of the Shang collapse. The Shang heartland may have been relatively short of precipitation during most of Shang times. In an early review of climate-related divination inscriptions, Wittfogel (1940) pointed out that interest in rain was most intense during the early months of the year. River water would have been available most of the time, but it should be noted that it takes well-organized social and command structures to accomplish major irrigation projects. Li (1977, p. 196) interpreted a system of ditches at Yinxu (the Late Shang capital near Anyang) as irrigation canals. In contrast, Keightley (1999b) noted that there is no archaeological evidence for large-scale irrigation works and the ditches at Xiaotun (a part of the Yinxu site) may have been for drainage. Our own view is that (1) insufficient subsurface work has been done to reveal irrigation systems in agricultural field areas, and (2) other civilizations at a comparable stage of evolution developed irrigation systems. In times of severe drought when rivers and streams run dry, shallow wells in river beds could supply potable water for humans and livestock but not enough for irrigation. Ice core data from Greenland demonstrate that the Holocene climate was prone to significant fluctuations. In the Indian Ocean laminated cores indicate
128
G. RAPP & Z. JING
variations in monsoonal circulation, salinity, and sea-surface temperature of the order of one dry phase every 100 –1000 years (Doose-Rolinski et al. 2001). Excess rainfall not only leads to flooding but also makes floodplain roads impossibly muddy, interrupting military and commerce movements. It can ruin crops as easily as drought does. Within Shang territory and nearby lands there were natural landscape determinants of climate variation; for example, the orographic effects of mountains and the effect of the sea on coastal areas. Climate change often has a severe effect on agricultural productivity, which translates directly into wealth and power. Again, as noted by Liu (1996), ‘The occurrences of climatic fluctuations and the changing course of the Yellow River often coincided with the development of archaeological cultures.’ In the next section we consider the geomorphological environment and the importance of the Yellow River.
The Shang landscape Landscape changes and associated geomorphological processes are preserved in the sediments, soils, and erosional contacts that make up regional stratigraphy. Sediments record depositional environments and processes; palaeosols provide critical evidence of landscape stability; and erosional unconformities record episodes of landscape degradation. We have used coring (see below) to determine the sedimentology and stratigraphy of landscape change in second millennium BCE Henan Province. China’s topography is marked by a series of steps, with decreasing altitude from the Qinghai– Xizang plateau in west –central China with an
altitude of 4000 m to the Yellow River plain with a mean altitude of less than 200 m. The Huai River valley forms a north–south divide that coincides with the dividing line between areas of rice and millet cultivation in ancient China. All the earliest major civilizations developed along the world’s great river systems: the Nile, the Tigris –Euphrates, the Indus, the Yangtze and the Yellow. The Shang heartland was the broad plain of the Yellow River and its tributaries. All floodplains are continuously evolving, not to any specific endpoint, but changing in response to factors such as rainfall and sediment input. Rivers are in a delicate balance of inputs, outputs, and the storage of sediments. On floodplains rivers are highly mobile; they meander across the plain and also move vertically. Significant changes in the courses of major rivers would lead to changes in transportation routes and relocation of settlements. The evolution of large urban centres requires communication and transportation systems to move resources and people. Any major relocation will have a large effect on social institutions. Our geoarchaeological fieldwork in the Anyang region provides an example of river channel migration (Fig. 2). Rivers flowing on broad floodplains both preserve archaeological sites through burial and destroy them through lateral migration. The earliest extraordinary floods mentioned in the Chinese historical texts occurred shortly before the beginning of the second millennium BCE. These texts refer to trying to control the floods (Pang 1987). In his paper Pang has attempted to date major floods and climate change during the second and early first millennium BCE using historical texts and astronomical data. We suggest he
Fig. 2. Archaeological sites usually are located on water courses operating at the time of habitation. The current course of the Huan River east of Anyang lies far to the north of the early sites in the region. An old river course lies between two outlined clusters of sites that undoubtedly delineate ancient courses of the river.
HUMAN AND ENVIRONMENT IN EARLY CHINA
has been only partially successful. In addition, we have a healthy scepticism about the accuracy and completeness of the very early historical record. By breaking through the natural levees during major flooding events the Yellow River has changed its channel location tens, if not hundreds, of times in the last four millennia. In 2600 BCE the Yellow River shifted its course from northeastward, debouching into the Bo Hai, to southeastward, flowing to the Yellow Sea. According to Liu (1996) the Yellow River flowed to the Yellow Sea from 2600 to 2000 BCE but this is questionable. Keightley (1999a) reported that the Yellow River shifted its main course in the North China Plain at least twice during Neolithic times. Major channel change can involve a dramatic change in the entire floodplain system. Yellow River flooding has been the focus of concern for thousands of years. Slack-water deposition from flood waters on alluvial plains was episodic with dust accumulation and soil formation throughout the Holocene, as described by Huang et al. (2007b). Loess – soil sequences of overbank deposits in the SE part of the middle reaches of the Yellow River drainage basin investigated by those workers allowed the reconstruction of Holocene flooding events. Six episodes of flooding were recorded over the alluvial plain. The third episode came at about 1620–1520 BCE and devastated a Xia settlement. Xu (2001) has documented 407 bank-breaching disasters on the Yellow River during the 2117 years from 168 BCE to 1949 CE. Because of human construction of artificial levees these flood frequencies probably were greater during this interval than during the second millennium BCE. Slackwater deposits form in restricted areas of reduced velocity. They can indicate heights reached by the floodwaters. Using evidence from slack-water deposits Yang et al. (2000a) have postulated that the largest Yellow River flood occurred about 5362 BCE. Their research indicated catastrophic floods at about 6200, 4100 and 400 BCE, but did not show any comparable floods during the second millennium BCE. The Yellow River flows from the easily eroded loess plateau and carries vast quantities of silt and clay sediments to form the large alluvial plain in its lower reaches. This broad, flat floodplain slopes about 1% to the east. The surface is composed of overbank silt and clay deposits of the Yellow River and its tributaries. In the lower reaches the Yellow River is aggrading, and with its easily eroded banks it floods often, breaches, and shifts its course. For the Yellow River we can consider two types of floods: the normal, probably seasonal, with the overbank flooding not affecting channel stability; and the catastrophic, breaching the levees with the formation of an entirely new channel.
129
An abandoned channel of the Yellow River, active from about 1550– 1850 CE, is located 20 km north of Shangqiu and rises above the surrounding plain. Visiting this elevated former channel one can see how easily, in flood stage, the river would destroy its levees, pour down onto the plain, and establish a totally new channel, leaving abandoned channels across the landscape (Fig. 3). The lower Yellow River has radically relocated its main channel many times through avulsion following catastrophic flooding. At times it has discharged northward into the Bo Hai as it does currently, but periodically it has shifted southward to the Jiangsu coast south of the Shandong Peninsula. Figure 4 illustrates some of these radical shifts in the course of the river.
Soils and vegetation Settlement farming, an increase in population during the Shang Dynasty, the use of bronze and wooden agricultural implements, the use of animal labour to grow millet, rice and vegetables, and the raising of domestic animals led to an increase in the disturbance of soils and vegetation (Hsiung et al. 1995). Foxtail millet was the principal crop during the Shang Dynasty. Wheat became a significant crop between 1600 and 1300 BCE (Lee et al. 2007).
Fig. 3. Aerial photograph showing abandoned channels of the Yellow River and its tributaries. The major east– west former channel near the top of the photograph operated between about 1550 and 1855 CE. This fairly recent channel remains a depression so accumulates rainwater. The area covered by this photo is approximately 45 km east–west and 48 km north–south.
130
G. RAPP & Z. JING
Fig. 4. Some of the many ancient courses of the Yellow River, marked as follows: (0) abandoned in 602 BCE; (1) 602 BCE to 11 CE; (2a) 11 to 70 CE; (2b) 70 to 1048 CE; (3) 893 to 1099 CE; (3a) 1060 to 1099 CE; (4) 1048 to 1194 CE; (5) 1194 to 1288 CE; (6) 1288 to 1324 CE; (7) 1324 to 1855 CE; (8) 1855 to present date; (8a) 1887 to 1889 and 1938 to 1947. Data are from Needham et al. (1971).
Pollen studies reveal large changes in Holocene vegetation in the area north of the Yangtze River. The diversity of tree species reached a maximum somewhat after 4000 BCE then decreased, especially on the Yellow River plain, because of farming (Ren & Beug 2002). In the Shang heartland the dense forests of the period before 2000 BCE were denuded for wood products for fuel and construction and land clearing for agriculture. The oracle bones describe the clearing of new fields for cultivation (Chang 1980, p. 223). Mencius wrote of the degradation of once-productive lands. To the best of our knowledge the Shang had mountain gods and river gods but no forest gods.
Methods The reconstruction of climatic and geomorphological environments has provided an increasingly clear picture of the landscapes and habitats of the North China Plain during the second millennium BCE. In reconstructing ancient environments, Earth scientists and archaeologists depend on mapping and stratigraphy to provide the records of the past. Archaeologists use walk-over surveys to pinpoint sites. However, even intensive surveys lack two
features: (1) they are 2D rather than 3D, covering only the surface of the land; (2) they recover only certain types of strictly archaeological information, with limited attention paid to ecological and geomorphological components. To overcome the limited 2D information obtained by traditional archaeological survey methods, which cannot locate deeply buried sites, we have developed a systematic 3D survey by utilizing coring to explore the subsurface dimension, including the stratigraphy and palaeopedology. For more than two decades, from 1970 to 1994, the senior author was involved with determining Holocene coastal change at archaeological sites in the eastern Mediterranean region (e.g. Rapp & Kraft 1994; Kraft et al. 2000). This research was based on intensive core drilling with detailed analyses of the cores to provide the sequence of coastal depositional environments and associated chronologies. In the eastern Mediterranean region three types of coring instruments were used: a vibracorer, a truck-mounted rotary drill, and a manual corer (the Dutch auger). In 1990 work shifted to Henan Province, China, with a chronological focus on the second millennium BCE, the time of the Shang Dynasty.
HUMAN AND ENVIRONMENT IN EARLY CHINA
131
Fig. 7. The first author using a Chinese Luoyang spade which retrieves a short core of about 15 cm.
Fig. 5. The Dutch corer can be operated with a metre-long split-spoon core barrel or a small auger head when rocks are encountered; extension rods are shown at the left.
Two types of coring devices were used in the two projects described below. The Dutch corer (Figs 5 & 6) was the work-horse of projects in the eastern Mediterranean and was the instrument used during the initial phases of both the Shangqiu and Anyang projects. Largely because this coring device could
only drill two or three 6–12 m holes per day we switched to the use of multiple Luoyang spades (Fig. 7) in the hands of skilled Chinese workers. In regions of major alluvial sedimentation many or most early archaeological sites are covered to a depth that makes the sites invisible to surface archaeological survey. In some cases ground-based geophysical exploration methods (electrical resistivity, magnetometer, or ground-penetrating radar) have proved useful, but we found magnetometer and ground-penetrating radar were not useful in our work at Shangqiu. Coring proved to be universally successful in understanding the subsurface dimension. The drill cores provided direct evidence of palaeo-vegetation patterns from macrobotanic, pollen, and phytolith remains. Environmental magnetic techniques (anhysteretic remanent magnetism and low-field magnetic susceptibility) provided additional parameters for characterizing site formation and anthropogenic information.
The Shangqiu Project
Fig. 6. Crew removing unconsolidated sediments from a Dutch corer split-spoon barrel.
In the early 1990s, we used core drilling to determine Neolithic, Bronze Age, and later landscape evolution in the Shangqiu area (Fig. 1), part of the Yellow River floodplain in southeastern Henan Province, China (Jing et al. 1995, 1997). At Shangqiu we initially used the Dutch auger but switched to the Luoyang spade (the traditional coring device used to detect buried cultural remains in China). Figure 8 is a cross-section showing diagrammatically the drill cores, sites, and major features of a portion of the research area. The layers marked ‘PS’ are palaeosols, and the layers marked ‘A’ are anthropogenic sediments. In the Shangqiu area the early Shang layers are buried c. 10 m below the surface. The results of
132
G. RAPP & Z. JING
Fig. 8. A diagrammatic cross-section of the stratigraphy in the Shangqiu area. The positions of some of the many cores drilled are illustrated. The Units are lithostratigraphic layers. Layers marked A are anthropogenic sediments; layers marked PS are paleosols.
the field work that we undertook showed that there are up to 12 m of silt deposited since early Shang times. This means that the Yellow River has deposited on average 33 cm per 100 years, but the greatest deposition occurred in the last 1000 years. The record from the drill cores was crucial for evaluating and interpreting, as well as predicting, buried archaeological sites. The archaeological importance of the Shangqiu region lay in its position as the centre of Predynastic Shang cultures. Coring in this region led to the discovery of a major Eastern Zhou Dynasty city, buried many metres below the surface. For the establishment of major city sites and for the development of agricultural soils one needs long-term landscape stability. This was achieved in the Shangqiu region during the Neolithic to the Han period. The cultural strata for this entire period are thus compressed into a very thin deposit. Beginning in the early twelfth century CE rapid sedimentation has led to the deep burial of early Shang levels (Jing et al. 1995). The formation of the first urban societies in China had much to do with politics and war, but it also had something to do with the soils and landscape stability needed for a productive agriculture. If a single flood deposited 20 cm or more of sediment over hundreds of square kilometres it would destroy not only the current crops but also the whole soil regime and limit agricultural production for years. From the mythological founding of the Predynastic Shang (by Xie), flood control has been a preoccupation of the people in the Yellow River region. Late Prehistoric and early historical capitals and other major cities were located on topographic highs, especially near the confluence of two rivers. Chang (1980, p. 141) has pointed out that place names in the oracle bones indicate that Yinxu (the Late Shang capital near Anyang) was located on top of a hill or mound. All floodplain rivers change their courses over time, sometimes rapidly
and dramatically; settlements in lowland areas are perpetually at risk. One or more Shang capitals were moved following destruction by catastrophic floods. Post-Shang records state that one Shang capital city (Geng) was destroyed by a flood and the capital moved (Chang 1980, p. 7). Historical records written much later than the Shang Dynasty list many Shang capitals. Most have not been confirmed by archaeological evidence. We are sure of only three: the Early Shang (1600–1350 BCE) capital at Zhengzhou; the Middle Shang (1350–1250 BCE) capital at Huanbei; and the Late Shang (1250–1046) capital at Yinxu. Zhengzhou lies south of the Yellow River (see Fig. 1); Huanbei (discovered by our project in 1997– 1998) lies near the modern city of Anyang, just north of the Huan River. (Huanbei means Huan North.) The Late Shang capital at Yinxu also lies near the modern city of Anyang, but the central district of Yinxu lies south of the Huan River. The Yellow River and its tributaries drain thick loess deposits in Shaanxi Province and have deposited thick sequences of silt. These rivers have a very high concentration of suspended sediment. One of the tributaries, the Dali River, currently has a mean sediment load that is the highest on Earth. Soil and sediment erosion on the Yellow River plain began at least by 3000 BCE when cultivation became a major practice. Soil is one of the most vulnerable of natural resources. The greatest impact on soil is from agriculture. Deforestation, beginning in the Neolithic, caused increased erosion and sediment load in the major rivers, increasing the flooding potential. Xu (1998) commented that because the Yellow River plain has been subsiding continuously during the Holocene, erosion should not have been a problem. However, this neglects the erosion caused by lateral migration of rivers and by runoff from deforested terrain. Finally, the flatness of the Yellow River plain must have resulted in the formation of marshes
HUMAN AND ENVIRONMENT IN EARLY CHINA
that probably had a major impact through diseases such as malaria.
The Anyang Project Despite 80 years of excavation at Yinxu, the last capital of the Shang near Anyang, the environmental context of this urban centre has been poorly understood. With this in mind, since 1997 we have used core drilling as a systematic exploration method, which led to the critical discovery of the Middle Shang site of Huanbei Shang City at Anyang (Tang et al. 2000). The Huanbei site is located north of the Huan River in the northern suburb of the modern city of Anyang, Henan Province. The average elevation of the Huan River floodplain near Anyang is about 70 m above sea level (see map, Fig. 2). From 1997 to 2003 this was a joint project undertaken by the Institute of Archaeology of the Chinese Academy of Social Sciences and the Archaeometry Laboratory at the University of Minnesota, Duluth. Since 2003 the North American component has shifted to the Department of Anthropology, University of British Columbia. To explore for Middle and Late Shang sites we depended on subsurface archaeological surveys based on coring. It became obvious that the Dutch auger method was too slow to be useful. However, the use of multiple Luoyang spades by well-trained Chinese operators was extremely successful. It was while surveying with the Luoyang spade that we discovered Huanbei Shang City. Over 2000 cores were drilled. Intensive coring revealed that this large urban settlement site was buried about 2.5 m beneath the modern surface over an area of at least 470 ha. The nature and origin of the sediment or soil matrix are critically important in understanding human –environment relationships. The soil cover of the Earth is the principal land carbon reservoir. Organic carbon is perhaps the best indicator of Holocene palaeosols, which are indicative of a stable land surface where habitation was possible. The changing locations of major rivers certainly had impacts on social structures and perhaps even on social stability. During Shang times the increase in state control required an increase in communication and transportation (i.e. roads and watercourses). The Shang heartland encompassed a largely flat landscape with only rivers, bogs and marshes providing major landscape challenges. More attention needs to be paid to understanding the Shang communication and transportation network and its evolution over time. Oracle bones show boats, indicating that the river networks were used. West of Anyang, in the Taihang Mountains, during the period 6000–3000 BCE tree pollen accounts for 50% or more of the pollen record, indicating a
133
closed forest landscape. Then people began to clear the forests. By 500 BCE the forests were mostly gone, as described by Yang et al. (2000b). Natural changes may have augmented the effects of human disturbance. The end of the Holocene temperature optimum about 2500 BCE led to the contraction of deciduous forests and the expansion of coniferous forests and grasslands. The significant reduction in oak along with the increase in pine at about 2000 BCE probably reflects the initiation of widespread human interference on the landscape. There are three ancient lake beds in the Anyang region: Wangjiadian, Guangrenpo, and Caidianpo. Our cores from Wangjiadian produced a good pollen profile, shown in Figure 9. This profile indicates massive clearing of pine forests around 1050 BCE, at the end of the Shang Dynasty. Although there is no archaeological evidence for a rapid rise in population at this time, the corresponding increase in Artemisia and chenopods implies human disturbance of the vegetation. We should note that excavated buildings at Huanbei have pine beams. Deforestation has two direct effects on flooding: (1) it increases direct rainfall runoff into water courses; (2) it leads to increased erosion, which builds up the river beds, raising them above the level of the surrounding floodplains, thereby increasing the susceptibility to major flooding. Deforestation affects more than erosion. It affects climate through changes in local albedo (the reflecting power of a surface), latent heat flux, evapotranspiration, moisture, soil formation, and even air currents. Through loss of habitat it affects biodiversity. Likewise, associated land-use change involves more than landscape change. It leads to an alteration of the biogeochemical systems that underpin the whole ecology. In terms of fauna it is well known that Shang royalty hunted elephants, tigers, and rhinos, indicating an environment very different from that of today. The population increase would have required at least some shift away from animal husbandry to cereal agriculture. Intensive agriculture actually reduces the overall biomass to concentrate on the most desirable plants.
Discussion and conclusions General human impacts on the environment such as deforestation leading to erosion, invasion of weedy species, changes in seed dispersal, and destabilizing of predator –prey dynamics are well known and need no detailed discussion here. What we know is that the climate deteriorated during 2200–800 BCE but perhaps not severely enough to result in major social impacts, except possibly making some contribution to the demise of the Xia and Shang Dynasties. What we cannot quantify with any confidence for this period is the population
134
G. RAPP & Z. JING
Fig. 9. Pollen diagram of the Wangjiadian Profile, Anyang.
and the population growth that affect the magnitude of the human impacts. Studies of human– environment interactions require a full understanding of the three dimensions of space and one dimension of time, across the entire geoarchaeological landscape, not just at single sites. Our understanding of the buried soils and landscape evolution rapidly expanded with our coring at Huanbei. This site sits on the old floodplain, a stable surface. This challenges the traditional presumption that major settlements always rested on highlands. We now have for the Anyang region a fairly good picture of environmental evolution for the period 2200–800 BCE. As excavation continues and analyses proceed, societal changes will become clearer. Special studies of palaeo-DNA and stable isotopes are producing critical palaeopathological, palaeogenetic, and palaeodietary data on the changing characteristics of the Shang population over time. The continuing evolution of the floodplains of the Yellow and Huan rivers in China during the second millennium BCE profoundly influenced the location and stability of major habitation sites. The example of the Yellow River floodplain was the more dramatic. Major flooding and channel migration (not conducive to hosting major city sites) co-acted with extended periods of minimal alluviation (leading to a stable landscape). The much smaller Huan River floodplain witnessed similar but smaller changes. Our data, obtained in our search to locate major Shang sites buried up to 12 m in floodplain sediments, are consistent with the changing environmental conditions leading to societal responses such as relocation of city sites.
Although radical shifts in the course of the lower Yellow River shown in Figure 4 go back only to 602 BCE, such shifts, as shown by other workers referenced in this paper, were common throughout the Holocene. Hence it is our view that radical changes in the course of the Yellow River, accompanied by severe flooding, may have contributed to the relocation of Shang capitals. What environmental factors were most influential in actions taken by Shang kings? It appears that the two that caused the most concern were floods and the lack of sufficient rainfall. We would suggest that perhaps the most dramatic and far-reaching decision made by Shang royalty was moving the capital over significant distances. There undoubtedly were political, economic, and even personal reasons for moving a capital, but the seriousness of catastrophic floods must be considered as consequential. Ancient texts say the Shang had royal capitals at Po (Bo), Hsiao (Ao), Hsing (Xing) (also called Geng), Pi, Yen (Yan) and Yin (Chang 1980). Other sites such as Huanbei probably were royal capitals. We know that Geng was destroyed by a flood and the capital was moved (Chang 1980, p. 7). Major floods (even if less than catastrophic) were likely to have been highly disruptive of agriculture and transportation. Twenty centimetres of flood deposits force soil-forming processes to start anew, resulting in the loss of much of the built-up soil fertility. The rivers the Shang depended on were flood prone and, even lacking devastating floods, the rivers migrated broadly over the North China Plain. As population increased and forests were
HUMAN AND ENVIRONMENT IN EARLY CHINA
cut down erosion led to the elevation of river beds, increasing the number and severity of major floods. In terms of precipitation the spring rains were largely problematical. Throughout their reign the Shang had to accommodate a climate that was slowly deteriorating. Finally, the focus of historical studies for two millennia has been the rise and fall of nations with little attention to the environmental context. This has been changing in recent years, as can be seen from this volume and the references accompanying each paper. As scientific methods for reconstructing ancient environments become increasingly robust, human –environment interaction studies will become more commonplace. S. Yu made valuable suggestions that improved this paper. C. Kubeczko assisted with the figures, and N. Nelson provided constructive editing. We benefited from the reviews of the first draft by two anonymous reviewers. We received support for this paper from the graduate School of the University of Minnesota and the University of Minnesota Retirees Association.
References An, Z., Porter, S. et al. 2000. Asynchronous Holocene optimum of the East Asian monsoon. Quaternary Science Reviews, 19, 743– 762. Bond, G., Showers, W. et al. 1997. A pervasive millennial-scale cycle in North Atlantic Holocene and glacial climates. Science, 278, 1257–1266. Bryson, R. 1993. Simulating past and forecasting future climates. Environmental Conservation, 20, 339–346. Chang, K.-C. 1980. Shang Civilization. Yale University Press, New Haven, CT. deMenocal, P. 2001. Cultural responses to climate change during the Late Holocene. Science, 292, 667–673. Ding, Y. 1994. Monsoon over China. Kluwer, Dordrecht. Doose-Rolinski, H., Rogalla, U. et al. 2001. High resolution temperature and evaporation changes during the late Holocene in the northeastern Arabian Sea. Paleoceanography, 16, 358–367. Dykoski, C., Lawrence, R. et al. 2005. A highresolution, absolute-dated Holocence and deglacial Asian monsoon record from Dongge Cave, China. Earth and Planetary Science Letters, 233, 71–86. Elvin, M. 1993. Three thousand years of unsustainable growth: China’s environment from Archaic times to the present. East Asian History, 6, 7 –46. Feng, Z., Thompson, L., Mosely-Thompson, E. & Yao, T. 1993. Temporal and spatial variations of climate in China during the last 10 000 years. Holocene, 3, 174–180. Hassan, F. 1997. Nile floods and political disorder in early Egypt. In: Dalfes, H., Kukula, G. & Weiss, H. (eds) Third Millennium BC Climate Change and Old World Collapse. NATO ASI, 149, 1– 23. Hsiung, W., Yang, S. & Tao, Q. 1995. Historical development of agroforestry in China. Agroforestry Systems, 30, 277–287.
135
Huang, C., Pang, J. & Li, P. 2002. Abruptly increased climatic aridity and its social impact on the Loess Plateau of China at 3100 BP. Journal of Arid Environments, 52, 87– 99. Huang, C., Pang, J., Su, H., Yang, Q. & Jia, Y. 2007a. Climatic and anthropogenic impacts on soil formation in the semiarid loess tablelands in the middle reaches of the Yellow River, China. Journal of Arid Environments, 71, 280– 298. Huang, C., Peng, J., Zha, X., Su, H., Jia, Y. & Zhu, Y. 2007b. Impact of monsoonal climatic changes on Holocene overbank flooding along Sushui River, middle reach of the Yellow River, China. Quaternary Science Reviews, 26, 2247– 2264. Jing, Z., Rapp, G. & Gao, T. 1995. Holocene landscape evolution and its impact on the Neolithic and Bronze Age sites in the Shangqui Area, Northern China. Geoarchaeology, 10, 481– 513. Jing, Z., Rapp, G. & Gao, T. 1997. Geoarchaeological aids in the investigation of early Shang civilization on the floodplain of the lower Yellow River, China. World Archaeology, 29, 36–50. Keightley, D. 1999a. The environment of Ancient China. In: Loewe, M. & Shaughnessy, E. (eds) The Cambridge History of Ancient China: From the Origins of Civilization to 221 B.C. Cambridge University Press, Cambridge, 30–36. Keightley, D. 1999b. The Shang. In: Loewe, M. & Shaughnessy, E. (eds) The Cambridge History of Ancient China: From the Origins of Civilization to 221 B.C. Cambridge University Press, Cambridge, 232– 291. Keightley, D. 2000. The Ancestral Landscape: Time, Space, and Community in Late Shang China. University of California, Berkeley, CA. Kraft, J., Kayan, I., Bruckner, H. & Rapp, G. 2000. A geologic analysis of ancient landscapes and the harbors of Ephesus and the Artemesion in Anatolia. Jahreshefte des Austerreichischen Archaologischen Instituts in Wien, 69, 175–233. Lee, G.-A., Crawford, G., Liu, L. & Chen, X. 2007. Plants and people from the Early Neolithic to Shang periods in North China. Proceedings of the National Academy of Sciences of the USA, 104, 1087– 1092. Li, C. 1977. Anyang. University of Washington Press, Seattle, WA. Liu, L. 1996. Settlement patterns, Chiefdom variability, and the development of early states in North China. Journal of Anthropological Archaeology, 15, 237– 288. Liu, L. 2000. The development and decline of social complexity in northern China: some environmental and social factors. Indo-Pacific Prehistory Association Bulletin, 4, 14–34. Mayewski, P., Rohling, E. et al. 2004. Holocene climate variability. Quaternary Research, 62, 243– 255. Needham, J., Wang, L. & Lu, G.-D. 1971. Science and Civilization in China. Vol. 4, Part III. Civil Engineering and Nautics. Cambridge University Press, Cambridge. Pang, K. 1987. Extraordinary floods in Early Chinese history and their absolute dates. Journal of Hydrology, 96, 139 –155.
136
G. RAPP & Z. JING
Rapp, G. & Kraft, J. C. 1994. Holocene coastal change in Greece and Aegean Turkey. In: Kardulias, P. (ed.) Beyond the Site: Regional Studies in the Aegean Area. University Press of America, Lanham, MD, 69–90. Redman, C., Crumely, C. et al. 2007. Group report: millennial perspectives on the dynamic interaction of climate, people, and resources. In: Costanza, R., Graumlich, L. & Steffen, W. (eds) Sustainability or Collapse: An Integrated History and Future of People on Earth. MIT Press, Cambridge, MA, 115– 148. Ren, G. & Beug, H.-J. 2002. Mapping Holocene pollen data and vegetation of China. Quaternary Science Reviews, 21, 1395–1422. Shi, Y., Kong, Z. et al. 1993. Mid-Holocene climates and environmental change in China. Global and Planetary Change, 7, 219– 233. Tang, J., Jing, Z. & Rapp, G. 2000. The largest walled Shang city located in Anyang, China. Antiquity, 74, 479– 480. van Geel, B., Buurman, J. & Waterbolk, H. T. 1996. Archaeological and palaeoecological indications of an abrupt climate change in the Netherlands, and evidence for climatological teleconnections around 2650 BP. Journal of Quaternary Science, 11, 451– 460. Weiss, H., Courty, M.-A. et al. 1993. The genesis and collapse of third millennium North Mesopotamian civilization. Science, 261, 995–1004. Winkler, M. & Wang, P. 1993. The late-Quaternary vegetation and climate of China. In: Wright, H., Jr.,
Kutzbach, J., Webb, T., III, Ruddiman, W., StreetPerrott, F. & Bartlien, P. (eds) Global Climates since the Last Glacial Maximum. University of Minnesota Press, Minneapolis, MN, 221–264. Wittfogel, K. 1940. Meteorological records from the divination inscriptions of Shang. Geographical Review, 30, 110–133. Wu, W. & Liu, T. 2004. Possible role of the ‘Holocene Event 3’ on the collapse of Neolithic cultures around the Central Plain of China. Quaternary International, 117, 153 –166. Xu, J. 1998. Naturally and anthropologically accelerated sedimentation in the Lower Yellow River, China, over the past 13 000 years. Geografiska Annaler, 80, 67–78. Xu, J. 2001. Historical bank-breachings of the lower Yellow River. as influenced by drainage basin factors. Catena, 45, 1– 17. Yancheva, G., Nowaczk, N. et al. 2007. Influence of the intertropical convergence zone on the East Asian monsoon. Nature, 445, 74–77. Yang, D., Yu, G., Xie, Y., Zhan, D. & Li, Z. 2000a. Sedimentary records of large Holocene floods from the middle reaches of the Yellow River, China. Geomorphology, 33, 73–88. Yang, X., Sun, L., Xu, Q., Zhao, H. & Liang, W. 2000b. Vegetation changes of the Taihang Mountains since the last glacial. Chinese Geographical Science, 10, 261–269. Yu, S., Zhu, C., Song, J. & Qu, W. 2000. Role of climate in the rise and fall of Neolithic cultures on the Yangtze Delta. Boreas, 29, 157– 165.
Reconstruction of the fire history in the Siedlungskammer Burgweinting (Bavaria, Germany) in relation to settlement and environmental history ¨ LKEL1 & T. RAAB3 ¨ TZKE2, D. CHRISTOPHEL2, J. VO A. RAAB1*, W. BRU 1
Brandenburg University of Technology, Research Centre Landscape Development and Mining Landscapes, PO Box 101344, Konrad-Wachsmann-Allee 6, D-03013 Cottbus, Germany
2
Technische Universita¨t Mu¨nchen, Wissenschaftszentrum Weihenstephan (WZW), Department of Ecology and Ecosystem Management, Hans-Carl-von-Carlowitz-Platz 2, D-85354 Freising-Weihenstephan, Germany 3
Brandenburg University of Technology, Faculty of Environmental Sciences and Process Engineering, PO Box 101344, Konrad-Wachsmann-Allee 6, D-03013 Cottbus, Germany *Corresponding author (e-mail:
[email protected]) Abstract: Palaeoenvironmental investigations were carried out in the Siedlungskammer (prehistoric settlement area) Burgweinting (Regensburg, Bavaria, Germany) to reveal past settlement conditions and human impact on the environment. Two sequences were obtained from the Islinger Mu¨hlbach Fen, in close proximity to the archaeological excavation site in Burgweinting, which documents an almost continuous settlement history since the Neolithic Period. The analyses of the sequences comprise stratigraphic, geochemical and microscopic charcoal analyses. For chronological information, radiocarbon dating was conducted on a total of 10 samples. Thus, the first longterm fire record was reconstructed for the investigation area, and the results were correlated, based on radiocarbon dating, with the available environmental information and settlement history in the Siedlungskammer Burgweinting. The fire record reveals an almost continuous, but alternating fire history. Furthermore, it shows that fire played an important role in the Siedlungskammer Burgweinting and that most probably as early as the Mesolithic hunterer– gatherers deliberately used fire.
Fire is, beyond a doubt, an important abiotic factor affecting the environment. Biomass burning affects, in particular, the vegetation cover and woodland composition. During prehistoric times, fire was used for land management, wildlife management, and domestic use (Bowman et al. 2009). The anthropogenic use of fire is, in turn, associated with activities such as agriculture, firewood gathering, etc., which affect ecosystems in a diverse manner. To reveal the fire history of a prehistoric settlement area is therefore a vital part of palaeoecosystem research with regard to human –landscape interaction. The occurrence of fire is directly evidenced by particulate charcoal incorporated in soils, sediments and peat (e.g. Whitlock & Larsen 2001; Conedera et al. 2009). Because of the resistance of charcoal to microbial attack, the particles persist over thousands of years and can be used for the reconstruction of fire history (e.g. Tolonen 1986). Numerous studies using charcoal analyses have been carried out that are related to the issue of fire– climate– vegetation relationships (e.g. Beaty & Taylor 2009), human impact on the environment and an archaeological context (e.g. Marguerie & Hunot
2007; Marinova & Thiebault 2008). As a consequence, many different methodological approaches, especially concerning laboratory and counting procedures as well as data processing, have been applied, and there has been controversy related to these in the literature (e.g. Rhodes 1998; Carcaillet et al. 2001; Whitlock & Larsen 2001; Conedera et al. 2009). Concerning the interpretation of the records, many questions still remain unanswered. The charcoal records possibly comprise both charcoal derived from natural lightning-induced fires and charcoal produced by anthropogenically ignited fires (e.g. Edwards & Whittington 2000). The incidence of natural fires depends on the availability of combustible material (presence of wood, wood species, grass) and on climatic conditions. Based on the assumption that wildfires in Central Europe are rare because it is thought that the woodland composition is not flammable (Ellenberg 1996), the charcoal records are interpreted to reflect the deliberate use of fire by humans. The study area, Siedlungskammer (prehistoric settlement area; clearing for settlement surrounded by woodland) Burgweinting (central Bavaria,
From: Wilson, L. (ed.) Human Interactions with the Geosphere: The Geoarchaeological Perspective. Geological Society, London, Special Publications, 352, 137– 161. DOI: 10.1144/SP352.11 0305-8719/11/$15.00 # The Geological Society of London 2011.
138
A. RAAB ET AL.
Germany), offers an ideal setting to study past environmental conditions and the human impact on the landscape, as a result of the assemblage of archaeological evidence in combination with the presence of geoarchives. The Siedlungskammer Burgweinting is part of the Regensburger Altsiedelland, one of the oldest settlement areas in Central Europe, where settlement history can be traced back to the Palaeolithic period (Torbru¨gge 1984; Schier 1985; Paetzold 1992) (Fig. 1). The physiogeographical conditions of a moderate climate and very productive soils, combined with a close proximity to the river Danube traffic route, have attracted human populations ever since. An archaeological excavation in the Siedlungskammer Burgweinting has been continuing since 1994, and has delivered excellent insight into the spatio-temporal distribution of prehistoric cultures from the Linear Pottery Culture (Neolithic Period) to the Middle Ages (Stadt Regensburg et al. 2004). An area of c. 40 ha has been excavated, and the site has become one of the outstanding excavation areas in Central Europe. In this paper we present the first long-term fire history record for the Regensburg region. We investigate the question of whether the different settlement periods evidenced by the archaeological excavations are reflected in the geoarchives. Therefore, multi-proxy investigations comprising stratigraphic, geochemical and microscopic charcoal analyses as well as radiocarbon dating were carried out on the Islinger Mu¨hlbach Fen, in close proximity to the excavation site. In this paper, we present the results of two profiles: a long peat profile (profile 7038-302) from the centre of the fen, and a peat –colluvia sequence from the margin (profile 7038-306). The final aim of this study is to combine the palaeoenvironmental data with the outcome of the archaeological excavations.
Location and physiogeographical setting The study area Siedlungskammer Burgweinting (central Bavaria, Germany) is located south of the river Danube on the southeastern outskirts of Regensburg (Figs 1 & 2). The landscape surrounding Regensburg is characterized by the presence of four distinct natural landscape units: the Franconian Alb NW of Regensburg; the Bavarian Forest in the NE; the Lower Bavarian Tertiary Hills in the south; and the gently undulating loess-covered broad plain of the Regensburg –Straubing basin, the so-called Dungau (Fig. 1). The geology of the study area is characterized by Cretaceous, Tertiary and Quaternary rocks and deposits. In the Weintinger Holz SE of the study area, an outcrop of Großberger Sandstein, a Cretaceous rock, is present. To the south and west of the study area, Tertiary sediments of the Upper Freshwater Molasse appear, consisting of Ho¨henhofer Schotter, Tertiary feldspar sands, clays and marl (Oschmann 1958; Unger & Doppler 1996). These cover the Upper Jurassic bedrock of the Franconian Alb, which dips in an eastern and southeastern direction. Quaternary sediments are loess, loess loam and gravels of the Danube. The lower terrace, which is present on both sides of the river, consists of micaceous sands with single gravel interlayers (Oschmann 1958; Buch 1988). In contrast, the gravel of the Rissian high terrace is covered by Wu¨rmian loess and loess loam. At the rearward rim of the high terrace, the Lower Bavarian Tertiary Hills border, the Tertiary sediments are also widespread and covered by loess and loess loam. Additionally, Oschmann (1958) mapped alluvial sediments in the area of the investigated fen. The dominant soil types in the study area are Luvisols developed on loess and loess loam, and Cambisols developed on feldspar sands. As a result of the long history of agricultural
Fig. 1. Location of the study area Burgweinting, situated on the southeastern outskirts of Regensburg. RMD-Channel, Rhein–Main–Donau Channel.
FIRE AND SETTLEMENT HISTORY
139
Fig. 2. Study area, Burgweinting. The map shows the location of the archaeological excavation area, the Islinger Mu¨hlbach Fen (rectangle) and the location of the investigated profiles 7038-302 and 7038-306 [base map: topographic map (TK) 1:25 000, Bad Abbach, Bayerisches Landesvermessungsamt 2002].
use, the soil profiles are often truncated and degraded (e.g. Brunnacker 1958; Leopold 2003). The regional climate of the study area is characterized by its location in the transition zone between oceanic and continental climates. The annual mean air temperature ranges between 7 and 8 8C, and the annual precipitation averages between 650 and 750 mm (BayFORKLIM 1996). The potential natural vegetation in the Lower Bavarian Tertiary Hills would be deciduous woodlands of Galio– Carpinetum typicum. Between the merging of the Aubach system into the Danube and the rearward rim of the high terrace to the Tertiary Hills, a Querco –Ulminetum minoris would be present (Seibert 1969). However, the study area is an intensively used agricultural landscape, with little woodland cover composed of spruce forests, mixed forests and oak –hornbeam forests. As a result of the land-use pressure caused by building activities and road construction, the area of agricultural land has been reduced. The Aubach system, a fluvial system that consists of small streams originating in the Lower Bavarian Tertiary Hills, drains the surface towards the Danube. The surface catchment is c. 36 km2 in size. The hydrogeology is determined by the
Upper Jurassic bedrock, dipping below the molasse sediments, which represents a continuous aquifer (Bayerisches Landesamt fu¨r Umwelt 2007). In the study area, a spring containing small amounts of hydrogen sulphide (H2S) exists, which was produced by groundwater drilling in 1925. The study site is a fen [R: 4509000–4509200; H: 5427100; 342 m above sea level (a.s.l.)], located along the Islinger Mu¨hlbach stream (Fig. 2), which will hereafter be referred to as the Islinger Mu¨hlbach Fen. The peat deposit, with a maximum thickness of 5.50 m, has developed in a topogenic depression situated at the southern limit of the high terrace, at the border with the Lower Bavarian Tertiary Hills. The Islinger Mu¨hlbach Fen (c. 5.4 ha in size) is classified as a Kalkniedermoor (calcareous fen). It is hydrologically fed by mainly carbonate-rich ground water. Oschmann (1958) mapped several sources in the fen area. Additionally, precipitation, surface runoff, and most probably lateral strata water, from the Tertiary sediments, contribute to the water balance. Today, the fen is a protected landscape component, but it was formerly drained by several canals. The Islinger Mu¨hlbach marks the northern boundary of the fen, and the limit of the southern spatial extension of
140
A. RAAB ET AL.
the fen is marked by a track. Analyses of historical maps showed that the Islinger Mu¨hlbach was relocated after the Second World War and the real river course should be c. 200 m north of the fen (Raab et al. 2008). According to radiocarbon dating of the bottom peat (profile 7038-302, 337 –338 cm, Erl-7516), peat growth commenced during the Late Glacial Period (13408 –11878 a cal. BC, 2s). This is in accordance with comparable ages for the beginning of peat growth from the Kirchenmoos Fen near Poign, c. 10 km south of Regensburg (Vo¨lkel et al. 2002; Raab et al. 2005). Detailed plant macrofossil analyses have been performed on profile 7038-305, derived from the central part of the fen, to study the local mire vegetation. Only a few fen species are identifiable, owing to the high degree of decomposition. Only decay-resistant plant remains, such as plant tissue, rootlets, Amblystegiaceae stems (commonly known as ‘brown mosses’), seeds, etc. are present. The major peat constituents are rootlets (Radizellen), and therefore the peat is classified as sedge peat. Furthermore, the occurrence of fungal bodies such as Cenococcum geophilum indicates periods when the peat surface was well aerated (e.g. Langdon et al. 2003). Detailed results are being considered for publication elsewhere.
Archaeological background An archaeological rescue excavation has been continuing in the Siedlungskammer Burgweinting since 1994, conducted by the Office of Archive and Monument Protection of the City of Regensburg (Amt fu¨r Archiv und Denkmalpflege der Stadt Regensburg) in co-operation with the Bavarian Office for the Protection of Monuments (Bayerisches Landesamt fu¨r Denkmalpflege) (Amt fu¨r Archiv und Denkmalpflege 2004). An area of more than c. 40 ha has been excavated, and structures and remains from five millennia have been documented (Zuber 2006). The Siedlungskammer Burgweinting is situated on the Rissian high terrace of the Danube, near the edge of the lower terrace. Such locations were preferred settlement sites for millennia because the location between the wet lowland and the wooded hinterland is an effective ecological junction of different natural economic types (Torbru¨gge 1984; Dallmeier 2004). Generally, the whole Danube region has been characterized by a constant and relatively dense population since Early Neolithic times (Torbru¨gge 1984; Paetzold 1992; Dallmeier 2004; Zuber 2006). The earliest remains found in the Siedlungskammer Burgweinting belong to the Neolithic
Period (5500–2300 or 2200 a cal. BC). The presence of a Neolithic population is evidenced by ceramic shards from the oldest Linear Pottery Culture (sixth millennium BC), c. 100 m north of the excavation site (Lu¨ning 1991; Dallmeier 2004). There are single graves from the Corded Ware Culture (third millennium BC) (Kirpal 2005; Zuber 2006), and several graves and settlement evidence from the subsequent Bell Beaker Culture (third millennium BC, End-Neolithic Period) (Schro¨ter 2004; Zuber 2006). The latter finds are interpreted to suggest the existence of small hamletlike settlements and the first rural population in the Siedlungskammer Burgweinting (Zuber 2006). In contrast, the absence of finds for the cultural periods from the End-Neolithic (c. 2200 a cal.) to the Urnfield Period (c. 1300 a cal. BC, Bronze Age) points to a hiatus in settlement activities (Siedlungsruhe) lasting for about 1000 years (Zuber 2006). A remarkably intensive settlement period in the Siedlungskammer Burgweinting is the Urnfield Period (1300–800 a cal. BC). An extended Urnfield burial ground with about 400 cremation graves, and traces of two associated villages with a total of 135 post constructions for houses have been excavated so far (Zuber 2002, 2004a). These document a c. 300 years continuous settlement (Zuber 2006). At the shift from the Urnfield Period to the Iron Age, a decline in population is inferred from the decrease in the numbers of finds. The Iron Age (800 –50 a cal. BC) is subdivided into the Hallstatt Period (800–500 a cal. BC) and the La Te`ne Period (500 –50 a cal. BC). For the Hallstatt Period, ditch complexes suggest the presence of so-called Herrenho¨fe, a characteristic South Bavarian settlement pattern. During the following Early La Te`ne Age Period the settled area increased again. In the southeastern part of the excavation area, millstones, spindle whorls and c. 40 storage pits (Erdkeller) are dated to the fourth century BC (Zuber 2004a). Bronze slags suggest the processing of bronze in the settlement. The settlement complex was oriented towards the brook rather than to the edge of the terrace (Zuber 2006). For the Middle La Te`ne Age Period, graves from the fourth century BC were found, and from the Younger La Te`ne Age Period, remnants of a kiln are present. An immediate succession from the La Te`ne Period to the oldest Roman settlement is not detectable (Zuber 2006). This is consistent with the archaeological results of the Regensburg region (Schier 1985; Zuber 2006). The Roman Empire is the second prominent settlement period in the Siedlungskammer Burgweinting. The development of the rural population in Burgweinting began at the same time as the first Roman military base in Regensburg was established in AD 79, the legionary cohort fort Kumpfmu¨hl (Boos 2004). In
FIRE AND SETTLEMENT HISTORY
AD 179, under the Emperor Marcus Aurelius, the ‘Castra Regina’ (fort by the river Regen) was completed for the III Italica Legion in the area of the old town of Regensburg. In Burgweinting, four villae rusticae found in an area of c. 25 ha attest to a dense population during this time. The selfsustaining villae rusticae consisted of residential buildings, stables, barns, storehouses and workshops. Primarily subsistence agriculture was practised, and the troops based at Regensburg and the civilian urban population were provided with the surplus of agricultural products (Moosbauer 2004). Moosbauer (2004) assumed that the area of agricultural land for each villa rustica was between 50 and 120 ha, and that the villae rusticae in Burgweinting had rather larger areas. The villae rusticae were built in the borderland between wet and dry ground on gently inclined slopes. Whereas the pastures and meadows were located in the areas near the stream, the agricultural fields were established uphill. About 20 Roman wells were found near the stream, which served as the source of water. During the 4th century AD, the villae rusticae were abandoned, possibly owing to the insecure political circumstances in the border region of Germania libera (Zuber 2006). Further settlement of the Burgweinting area is documented for the Early Middle Ages (end of the fifth century AD).
Material and methods Prior to profile coring, an auger survey of the fen was carried out to find suitable core locations. One location was chosen in the centre of the mire (profile 7038-302), where the disturbance of the peat sequence was expected to be minimal, and the other one at the margin of the fen (profile 7038-306), where the peat has interbedded colluvium layers (peat –colluvia sequence). Profile 7038-302 is in total 455 cm long. It was recovered with a Russian peat corer, which produces undisturbed half-cores (section length 50 cm, diameter 6 cm). The uppermost peat (0– 55 cm) was not sampled because this section was disturbed. Profile 7038-306 was gained by percussion drilling (Atlas Copco Co.). The mineral interbeddings in the sequence ruled out coring with a Russian peat corer. The uppermost 50 cm were cored with an open steel drill pipe. The subsequent 2 m were obtained with a closed probe using a Plexiglas tube. Unfortunately, the coring technique caused a compaction of the peat and sediments in the Plexiglas tube and c. 50 cm of the tube was empty, although the drilling reached 250 cm in depth. Therefore, the complete sequence is actually 199 cm long. After recovery, both cores were stored in a cooling chamber at þ4 8C until processing. The
141
cores were opened in the laboratory, and the visible stratigraphy was described in terms of changes of colour (Munsell Soil Color Charts, Anon. 1994), peat or sediment composition, carbonate contents and macrofossil remains. The description followed Ad-Hoc AG Boden (2005) and the Troels –Smiths sediment description system was used (see Birks & Birks 1980). Prior to sampling, the profiles were documented by digital photography. The profiles were sectioned into continuous 1 cm slices (bulk samples). From the centre of each slice, a 1 cm3 subsample was extracted for potential pollen analyses. Additionally, macro-remains samples were collected for species determination and radiocarbon dating. All samples were dried in a laboratory-type drying cabinet at þ40 8C. For the analyses, one part of the bulk sample was carefully crushed using a mortar, and the other part was pulverized using a bullet mill. To characterize major variations in peat stratigraphy, geochemical analyses were conducted on both profiles. The analyses were carried out at 1 cm intervals. Geochemical analyses on profile 7038-302 were carried out from 57 to 454 cm depth, but only the results from 57 to 225 cm depth are given in this paper. Profile 7038306 was analysed from 51 to 164 cm depth. The determination of total carbon, nitrogen and sulphur contents, sequential loss on ignition and colorimetric humification were carried out on the ground bulk samples. Total carbon (wt%), total nitrogen (wt%) and total sulphur (wt%) were determined with a CNS auto-analyser (Vario EL III, Elementar Analysensysteme GmbH). The sequential determination of loss on ignition (LOI) was carried out to estimate organic matter values (OM), CaCO3 and ignition residue following the recommendations of Heiri et al. (2001). For that, 0.25 g homogenized bulk samples (duplicates) were sequentially combusted at 550 8C (OM) and 950 8C (CaCO3). For the calculation of the proportion of organic carbon (Corg), the percentage of OM is divided by two, as it is assumed that the proportion of organic carbon is roughly 50% of the OM (Ad-Hoc AG Boden 2005). The proportion of CaCO3 is calculated by multiplying the loss on ignition after combustion at 950 8C by a factor 1.36. Finally, the ignition residue is calculated according to Lawson et al. (2004). The determination of colorimetric humification was exclusively performed on profile 7038-302 for the segment from 57 to 344 cm depth according to a laboratory protocol slightly modified from Chambers (2006). The samples were treated with an alkali extraction procedure (NaOH 6%). Afterwards, the transmission (expressed as percentage transmission, TM %) of the solution was measured at a wavelength of 540 nm on a spectral photometer (Lambda
142
A. RAAB ET AL.
25 UV/VIS Perkin Elmer). High TM values correspond to slightly humified peat, and low TM values to highly humified peat. For microscopic charcoal analyses, the method described by Rhodes (1998) was chosen, because this is a gentle processing method in which little particle fragmentation occurs. Several test runs showed that, after preparation of sample material from profile 7038-302, the abundant material in the Petri dishes hindered the counting of the charcoal particles. Therefore, the sample size was reduced to 0.1 g. In contrast, the test runs of profile 7038-306 showed that a sample size of 0.2 g was acceptable and therefore this sample size was used. Counting of the microscopic charcoal particles was carried out with a stereo binocular microscope (Zeiss Type 475052-9901, Zeiss Stemi DV4) with incident light at 20– 40 magnification, using a manufactured grid (1 cm 1 cm) for orientation. The samples were completely counted, but not identified. All particles with the characteristic attributes (black, completely opaque, angular, with a silvery lustre in incident light) were counted as charcoal fragments (Swain 1978; Patterson et al. 1987; Clark 1988; Rhodes 1998; Enache & Cumming 2006). For the comparison of the results, the absolute counts were converted to parts per gram (ppg). The pre-examination of the samples in the Petri dishes, which involved measuring the longitudinal axis of the fragments with a micrometre eyepiece, showed that the fraction sizes of the charcoal particles and the fraction size distribution differ in the two profiles, and therefore different counting strategies were applied. In total 75 samples from profile 7038-302 were analysed at 1 cm intervals. Additionally, the section from 82 to 69 cm depth was continuously counted. The microscopic charcoal particles of this profile were classified into three classes of 50–500, 500 –1000 and .1000 mm. From profile 7038-306, the microscopic charcoal analyses were conducted for the section from 51 to 164 cm depth, at 1 cm intervals (n ¼ 55). No analyses were carried out on the samples from 155–156 and 113 –114 cm, because the sample size was too small. The microscopic charcoal particles of this profile were classified into the classes of 50 –150, 150– 250 and .250 mm. Finally, for each profile the microscopic charcoal sum was calculated. For chronological control, a total of 10 14C samples (wood fragments, undetermined plant macroremains and peat bulk samples) were selected from profiles 7038-302 and 7038-306 for 14CAMS (accelerator mass spectrometry) dating at the AMS C14-Labor Erlangen of the FriedrichAlexander-Universita¨t Erlangen –Nu¨rnberg. The 14 C dates were converted to calibrated ages (a cal. BC or AD) using the calibration dataset of Reimer et al. (2004).
Results and interpretation Stratigraphy and geochemistry of profile 7038-302 The peat profile 7038-302 from the centre of the fen (Fig. 2) is in total 455 cm long and can be visually subdivided into three units. The first unit is loess at the base (455 –361 cm depth). This is overlain by a transition horizon from loess to peat (361 – 341 cm depth), which represents the second unit. The third unit is a 286 cm thick minerotrophic peat (341 –55 cm depth), which includes an interbedded mineral layer at 329– 322.5 cm. The peat from 55 to 0 cm depth was not sampled, because it was disturbed. This study focuses on the peat from 225 to 55 cm depth, as it was expected to include the settlement periods. A detailed profile description of this section is shown in Figure 3. The minerotrophic peat has a brownish black to black colour and is amorphous to highly humified, and hence only few plant macro-remains are present and identifiable. The peat is mainly composed of rootlets and is therefore classified as sedge peat. Wood fragments are present, which are often strongly decomposed. At 224–225 cm depth, some seeds of Menyanthes trifoliata were found, indicating wet conditions and mineral enrichments (Blundell et al. 2008). Furthermore, charcoal fragments and mollusc shells are present (Fig. 3). The peat is moderately to highly calcareous and carbonate precipitates are visible in some sections of the peat. This resulted in pH values between 6.5 and 7.1 and indicates slightly acidic to neutral conditions (Ad-Hoc AG Boden 2005). Based on the geochemical data (total carbon (TC) wt%, total nitrogen (TN) wt%, total sulphur (TS) wt%, TM %, Corg/TN ratio, organic matter (OM) wt%, CaCO3 wt%, ignition residue wt%), the peat from 255 to 55 cm depth is subdivided into eight sections (Fig. 4). Section 8 (225 –198 cm). The peat is characterized by relatively high TC (39.8 –46.7 wt%), TN (1.9– 2.4 wt%) and TS (3.5–5.3 wt%) values. The Corg/ TN ratio is between 17.8 and 23.0. The TM values range between 26.8 and 54.2%. The OM values are high (73.9 –85.9 wt%). The CaCO3 values are between 0.7 and 8.1 wt%, with increasing values towards the top. The proportions of the ignition residue of 12.9–17.7 wt% are comparatively low and run parallel to the CaCO3 curve. Section 7 (198–184 cm). The TC (29.4–41.5 wt%), TN (1.5–2.2 wt%) and TS (2.2–4.2 wt%) values decrease towards the top. The Corg/TN ratio (14.4– 17.7) is lower than in Section 8. The TM values (41.3– 60.8%) increase towards the top,
FIRE AND SETTLEMENT HISTORY 143
Fig. 3. Profile 7038-302; stratigraphy, horizons and description from profile depth 225 to 55 cm.
144
A. RAAB ET AL.
Fig. 4. Profile 7038-302; geochemical parameters (TC, TN, TS, TM, Corg/TN ratio, loss on ignition) and sections distinguished (1– 8).
whereas the OM values (44.7–74.7 wt%) decrease in the same direction. Both CaCO3 (5.2– 25.7 wt%) and ignition residue values (16.3 –29.5 wt%) increase towards the top. Section 6 (184 –168 cm). Whereas the TC values (27.7–31.4 wt%) show only minor fluctuations, TN (1.0–1.5 wt%) decreases towards the middle part of this section, and then increases towards the top. TS (2.0–3.7 wt%) initially remains at about 2.0 wt%, and then increases towards the top. The Corg/TN ratios are between 15.9 and 24.0, with a maximum at 175– 176 cm depth. The TM values fluctuate between 56.6 and 64.0%. OM varies between 42.6 and 49.0 wt%. The CaCO3 (22.7– 26.6 wt%) and ignition residue (27.2–32.5 wt%) contents run parallel and show only minor fluctuations. Section 5 (168 –158 cm). The TC (32.3– 35.2 wt%) and TN (1.6–1.9 wt%) values are higher than in Section 6. TS (2.2–4.9 wt%) shows a peak at 165– 166 cm depth, which is followed by decreasing values. The Corg/TN ratio is between 15.3 and 18.0. The TM values are between 51.6 and 55.2%, and the OM contents are between 51.3 and 59.9 wt%. The CaCO3 values of 15.1– 23.7 wt%, and the ignition residue of 22.2–25.7 wt% are lower than in Section 6, and also run parallel to each other.
Section 4 (158 –154 cm). This 4 cm thick section has lower values of TC (27.2 –27.3 wt%), TN (1.2 wt%) and TS (1.8 wt%). The Corg/TN ratio is between 15.3 and 19.0. The TM value is higher (68.6– 69.5%) than below. The OM decreases sharply (37.3–45.2 wt%), whereas the CaCO3 (31.3– 32.2%) and ignition residue (23.4–30.5%) show higher values. Section 3 (154–128 cm). The TC (18.7–31.7 wt%), TN (0.6– 1.5 wt%) and TS (0.8–2.5 wt%) decrease towards the middle part, and then increase towards lower profile depths. The Corg/TN ratio (14.0 – 17.4) maintains low values. The TM values rise strongly (58.8–87.9%) towards the middle part of this section, as do the OM values (16.9 – 50.6 wt%). In contrast, CaCO3 (24.5 –46.9 wt%) increases towards the middle part, as do the values for ignition residue (24.9 –36.1 wt%), which run parallel to CaCO3, but are less pronounced. Section 2 (128 –84 cm). All geochemical parameters show highly fluctuating values. The TC values are between 23.0 and 34.8 wt%, and TN values are between 0.7 and 1.4 wt%. The TS values (0.9– 3.5 wt%) show a decrease towards the top. The Corg/TN ratios vary between 12.4 and 21.6. The variations in TM are also high (49.7– 85.6%). The OM values range between 22.0 and 50.4 wt%. CaCO3 (24.2–43.4 wt%) and ignition
FIRE AND SETTLEMENT HISTORY
residues (24.5– 36.1 wt%) values also show high fluctuations. Section 1 (84–55 cm). TC (22.5–39.7 wt%) and TN (0.8–2.0 wt%) values rise towards the middle part of this section and then show a decrease. The decline in TS (0.8–1.5 wt%) values continues. The Corg/TN ratio (14.9 –28.5) shows a peak at 83– 84 cm. The TM values range between 40.5 and 80.0%, with the highest values in the middle part of the section. The OM values (43.6 – 72.1 wt%) also are highest in the middle part of the section. In contrast, CaCO3 (10.2– 30.1 wt%) and ignition residue (17.5 –31.3 wt%) decrease towards the middle part and increase towards the top. The uppermost part of the profile, at ,65 cm depth, has decreasing values for TC (22.5– 18.4 wt%), TN (0.7–0.9 wt%) and TS (0.5– 0.4 wt%). The Corg/TN ratio (13.7– 14.0) is low. The TM values are between 73.8 and 78.5%. The OM values (18.8– 26.0 wt%) decrease, whereas the CaCO3 values (38.4–44.3 wt%) and ignition residues (33.7–36.8 wt%) increase. Most probably this uppermost profile section has been disturbed by humans in the form of past drainage measures and use as a meadow. The results of the analyses of the geochemical parameters are highly variable throughout the peat profile, suggesting changing peat compositions and peat formation conditions. The TC values throughout the profile are between 18.4 and 46.7 wt%. The TN values in the peat profile are between 0.6 and 2.5 wt% with the highest values at 191 –224 cm, 166 –161 cm and 78 –73 cm, which could suggest an altered peat composition. Overall, the TN values are typical for minerotrophic peat (see Grosse-Brauckmann 1990). The TS values are between 0.4 and 5.3 wt% and reach a maximum for the profile in the bottom section at 224 –193 and 168 –165 cm. Between 165 and 135 cm, with the exception of a small peak at 155 cm, the TS values decrease. From 135 to 105 cm, the TS values show an increase. From 105 cm to the top of the profile, TS values decrease. The TS values represent pyrite-bound S, as pollen and macrofossil analyses have shown that pyrite framboids (FeS2) are present in pollen grains and plant tissue. The sulphur is derived from organic decay and most probably from the groundwater. The Corg/TN ratios vary between 12.4 and 28.5 and are typical for minerotrophic fen peat (Grosse-Brauckmann 1990). Close Corg/TN ratios indicate strong decomposition and decay of organic material under relatively anaerobic conditions. Colorimetric humification (TM %) was ascertained for peat stratigraphic investigation. The TM values fluctuate throughout the profile. The TM values run parallel to the organic matter curve. Therefore, highly
145
humified peat corresponds to high organic matter contents and vice versa. The different degrees of humification might reflect the conditions during peat formation; however, the humification is also influenced by the decay resistance of the peatforming plants (Yeloff & Mauquoy 2006). The organic matter (OM), CaCO3 and ignition residue values are determined by sequential LOI. The OM values are highly variable, between 16.9 and 86.4 wt%. The calcium carbonate (CaCO3) values of the peat are between 0.7 and 46.9 wt%, and prove that the peat is generally highly calcareous. CaCO3 values of ,1 wt% are present only in the lower part of the profile at 224– 213 cm depth. Carbonate precipitates, the so-called Almkalk (synonyms are Wiesenkalk or Wiesenmergel) are visible in the peat. Carbonate precipitation is related to spring water rich in calcium bicarbonate (Jerz 1983; Grosse-Brauckmann 1990; Niller 1998). Additionally, a minor proportion of CaCO3 comes from mollusc shells. The ignition residue throughout the peat profile is between 12.9 and 36.8 wt%, and therefore is comparatively high. According to Grosse-Brauckmann (1990) the ash contents of common minerotrophic peat are between 5 and 15 wt%. The ignition residue represents the contents of mineral components, which are primarily silicates in the form of easily transportable grain-size fractions of silt and clay. Horizontally and vertically moving groundwater is linked with the transport of allochthonous material in a dissolved state. Mineral input by surface runoff may deliver further material. The CaCO3 and the ignition residue depth curves show coinciding variabilities, suggesting a relationship between CaCO3 and ignition residue. Overall, both parameters reflect the geochemistry of the carbonaceous groundwater connected with the input of minerals.
Stratigraphy and geochemistry of profile 7038-306 Profile 7038-306 is from the margin of the fen (Fig. 2). The peat –colluvia sequence is 199 cm long and shows a complex stratigraphy. Percussion drilling of the core caused an artificial compaction of the material (see Methods). The peat –colluvia sequence is built up by loess at the base (199 –180 cm), which is overlain by loess loam (180 –159 cm). A transition horizon from loess loam to peat is present at 159–152 cm. This is followed by amorphous to highly humified peat, composed of herbaceous plants with some wood fragments (152–117 cm). From 117 to 68 cm depth, the profile was subdivided into three peat –colluvium complexes (117 –100, 100–95 and 95 –68 cm). A 3 cm thick peat –colluvium
146
A. RAAB ET AL.
transition horizon at 68 –65 cm depth leads to the 65 cm colluvial cover on top of the sequence. A detailed profile description is shown in Figure 5. For the profile at 164– 51 cm depth, the following geochemical parameters were established: TC (wt%), TN (wt%), TS (wt%), Corg/TN ratio, OM (wt%), CaCO3 (wt%), and ignition residue (wt%), and the profile was subdivided into eight sections accordingly (Fig. 6). Section 8 (199 –159 cm). The loess at the base (199–180 cm) is covered by loess loam (180– 159 cm). Geochemical analyses were carried out from a 164 cm profile depth towards the top, and therefore are available for the upper part of the loess loam only. TC values (3.6–3.7 wt%) and TS values (0.1–0.4 wt%) are low, whereas the TN proportions (3.2–3.7 wt%) are comparatively high. The Corg/TN ratio is 1.3–1.4. The OM values of 9.0–9.3 wt% and the CaCO3 values (1.9– 2.8 wt%) are low. In contrast, the ignition residue values (87.8–89.0 wt%) are very high. Section 7 (159 –152 cm). This is the transition horizon from the mineral substrate to peat. The TC values (4.9–8.6 wt%) are relatively low. The TN values are between 2.5 and 3.2 wt%. The TS values are below ,1 wt%. The Corg/TN ratios (2.4–2.7) result from the low OM values and indicate a high degree of decomposition. The substrate becomes increasingly rich in OM (11.9– 17.5 wt%) towards lower profile depths. The CaCO3 values (2.9–3.1 wt%) are low. The substrate is mainly mineral, as indicated by the high amount of ignition residue (79.8–85.2 wt%). Towards the top, the ignition residues show a strong decrease. Section 6 (152–117 cm). This 34 cm thick peat shows TC values that increase at the bottom (13.2–46.2 wt%) and remain high overall. The TN values (2.6–7.8 wt%) are highly variable, ranging around 3 and 4 wt%. Noticeably higher TN values of 6–8 wt% are measured in the profile section from 131 to 117 cm depth, with the exception of one sample at 123 –124 cm depth with 2.6 wt%. The TS values (0.4–3.2 wt%) also show an increase towards lower profile depths. The range of variation of the Corg/TN ratios (4.5–16.1) is wide and is contrary to the TN curve shape. The OM (25.7– 85.6 wt%) values sharply increase and remain high throughout this peat section. The CaCO3 values (0–3.0 wt%) are low. The ignition residues (13.9– 71.4 wt%) decrease and then remain at a lower level throughout the profile. Section 5 (117 –100 cm). The TC (21.8 –35.0 wt%), TN (1.5–2.5 wt%) and TS (0.4–1.4 wt%) values decrease. The Corg/TN ratios fluctuate between 7.0 and 15.3. The OM values (30.2–67.0 wt%)
decrease towards lower profile depths, whereas CaCO3 (1.8–3.7 wt%) remains at a low level. In contrast, ignition residue (30.0 –66.4 wt%) increases towards lower profile depths. The analyses show that the substrate is composed of organic and mineral substrates in approximately equal parts. Section 4 (100 –95 cm). The TC (8.8–20.5 wt%), TN (0.6–1.5 wt%) and TS (0.1–0.3 wt%) values decrease towards lower profile depths. The Corg/ TN ratios (13.5–14.6) are stable. OM values (18.4– 40.0 wt%) decrease, whereas the ignition residues (57.7–80.3 wt%) sharply rise. The CaCO3 values (1.3–2.0 wt%) are low. The analyses show that this part of the profile is mainly composed of mineral substrate mixed with organic material. Section 3 (95–68 cm). The TC contents are between 11.0 and 21.8 wt%. The TN contents are low (generally less than 2%), with the exception of one sample at 71 –72 cm with 5.6 wt%. The TS values (0.1– 0.2 wt%) are generally low. The Corg/TN ratio is about 15, except in the layer with increased TN where the Corg/TN ratio is four. The OM values fluctuate (22.5 –44.6 wt%), as do the ignition residue values (52.6–76.0 wt%). The CaCO3 (1.4– 2.9 wt%) is low. The analyses show that this section is mainly composed of mineral material with a higher proportion of organic matter than below, and that the proportions of mineral substrate and organic matter alternate. Section 2 (68–65 cm). This section represents the transition from organic-rich mineral substrate to the colluvial cover. The TC (8.7– 12.6 wt%), TN (0.6–0.8 wt%) and TS (0.1–0.2 wt%) values decrease. The Corg/TN ratios are between 15.2 and 16. OM values are between 17.9 and 25.5 wt%, and the ignition residue is high (72.6–80.4 wt%). The CaCO3 (1.7–1.9 wt%) values are low. Section 1 (65– 50 cm). This section is the lowermost part of the colluvial cover. The TC (0.7–2.0 wt%) and TN (0.1–0.2 wt%) values are low. TS was not detected in this sediment. The Corg/TN ratios are between 16.5 and 36.1. The OM (3.4–6.2 wt%) values are low, as are the CaCO3 values (1.7– 3.1 wt%). In contrast, the ignition residues rise sharply (92.4–95.6 wt%). The geochemical parameters show that the substrate is mainly composed of mineral material with a small amount of organic material. In conclusion, the results of the geochemical analyses are consistent with the stratigraphic description of the sequence. The total carbon contents (TC) consist mainly of organic carbon, as shown by the high OM values. CaCO3 contributes only minimally to the TC contents, as the values are low. The TN values fluctuate throughout the
FIRE AND SETTLEMENT HISTORY 147
Fig. 5. Profile 7038-306; stratigraphy, horizons and description from 199 to 50 cm profile depth.
148
A. RAAB ET AL.
Fig. 6. Profile 7038-306; geochemical parameters (TC, TN, TS, Corg/TN ratio, Loss on ignition) from 164 to 50 cm profile depth and sections distinguished (1– 8).
profile, but are exceptionally high at the 131 – 117 cm and 71– 72 cm profile depths. These high TN percentages could result either from nitrogenrich plants or from anthropogenic nitrogen input. The TS percentages are highest in the peat section (Section 6), and they show that pyrite-forming conditions (anaerobic conditions, S and Fe availability) were present in this section. The Corg/TN ratios are highly variable throughout the profile. The widest range in Corg/TN occurs in Section 1 as a result of very low TN and low OM percentages in the mineral substrate. In general, the proportions of OM and ignition residues reflect the profile stratigraphy. From 117 cm depth to the top, changing proportions of OM and ignition residue indicate peat development with colluvial interbeddings. Three peat –colluvium layers (117–100, 100– 95 and 95 –68 cm depth) can be determined based on varying proportions of organic matter and mineral substrate. However, as a result of the artificial compaction caused by the drilling technique, distinct colluvial layers cannot be distinguished.
Microscopic charcoal analyses Microscopic charcoal analyses of profile 7038-302 were conducted exclusively on the peat substrate (193–57 cm profile depth). Charcoal particles are continuously present throughout the profile in variable quantities. The minimum microscopic charcoal particle sum is 219 ppg and the maximum is
29 435 ppg (mean: 7700 + 4977 ppg). The charcoal fragments were subdivided into three size classes: 50 –500 mm, 500–1000 mm and .1000 mm. In addition, the total sum of microscopic charcoal particles was calculated (Fig. 7). The particle size classification shows that the majority of the particles belong to the small size class of 50– 500 mm (219 –28 647 ppg), followed by the medium size class of 500–1000 mm (0–1671 ppg). Furthermore, large particles .1000 mm (0– 199 ppg) are also present (Fig. 7). The depth curves of the small and medium size fractions show approximately comparable shapes, whereas the depth curve of the large particle fraction has a different shape. Based on the curve shape of the microscopic charcoal sum, nine fire episodes (FE 1–9) have been identified. The lowermost profile section (193 –164 cm) is defined as FE 1, with comparatively low microscopic charcoal counts. In the next section (164 –145 cm), noticeably higher microscopic charcoal counts are present and this section is therefore classified as FE 2. Between 145 and 132 cm, the microscopic charcoal counts are at a lower level (FE 3). The profile section from 132 to c. 98 cm (FE 4) is a segment with relatively high microscopic charcoal particle counts extending over a c. 34 cm long section, with a peak at 115– 116 cm. Between c. 98 and 79 cm (FE 5) the charcoal counts tend to decrease, but vary in quantity. FE 6 (79 –76 cm) represents a 3 cm thick layer with increased counts and a peak at 78 –77 cm.
FIRE AND SETTLEMENT HISTORY
149
Fig. 7. Charcoal record of profile 7038-302, which shows the different charcoal size classes: 50– 500 mm (ppg), 500–1000 mm (ppg), .1000 mm (ppg) and the charcoal sum (ppg).
From 76 to 71 cm, the lowest charcoal particle abundance throughout the profile was found (FE 7). In the section from 71 to 64 cm (FE 8) the microscopic charcoal counts increase, with the highest counts throughout the profile at 70– 69 cm. From 64 to 57 cm the microscopic charcoal counts decrease towards the top of the profile (FE 9). On profile 7038-306, microscopic charcoal analyses were undertaken on the three substrate types of loess, peat and colluvial material (profile section 164 –51 cm depth). Microscopic charcoal particles are present throughout the profile. The minimum charcoal sum is 173 ppg and the maximum 5576 ppg (mean: 1153 + 1123 ppg). The microscopic charcoal fragments were subdivided into three size classes: small (50 –150 mm), medium (150 –250 mm) and large (.250 mm) (Fig. 8). The dominant particle class is the small particle size fraction with 137 –4263 ppg. This is followed by the medium size fraction with 5–865 ppg, and then the large fraction with 0 –783 ppg. Based on the shape of the charcoal sum depth curve, seven fire episodes (FE 1–7) are identifiable. FE 1 corresponds to the profile depth from 164 to c. 151 cm. Initially, there is an increase in charcoal abundance up a depth of 154 –153 cm, followed by a decrease in abundance. The profile section between 153 and 145 cm is classified as FE 2, with the highest microscopic charcoal abundance
throughout the profile. In the section from 145 to 127 cm depth (FE 3) considerably lower numbers of charcoal fragments were counted. The next section, between 127 and 119 cm depth (FE 4), contains even lower numbers of charcoal fragments than FE 3. In the section from 119 to 105 cm depth (FE 5), the charcoal counts increase slightly. Subsequently, in the section from 105 to 87 cm depth (FE 6), the microscopic charcoal counts are low. In the section from 97 to 51 cm, the charcoal fragments are again frequent (FE 7), and at almost the same level as in FE 3. The two charcoal records show several similarities as well as discrepancies. Beginning with the similarities, microscopic charcoal is commonly recorded in both profiles, and therefore documents the occurrence of fire in the environment. The shapes of the curves of the microscopic charcoal sums and the varying abundance of charcoal fragments indicate a continuous but alternating fire history. With respect to the discrepancies in the charcoal records, there are four major aspects to discuss. The first aspect concerns the different core locations within the Islinger Mu¨hlbach Fen, which resulted in different sequences of stratigraphy and furthermore, in different fire records. The different core locations and substrates also involve different transport and deposition processes. Although the
150
A. RAAB ET AL.
Fig. 8. Charcoal record of profile 7038-306, which shows the different charcoal size classes: 50–150 mm (ppg), 150– 250 mm (ppg), .250 mm (ppg) and the charcoal sum (ppg).
charcoal particles of the central profile most probably derive from aerial transport, the marginal profile can also include charcoal transported surficially by water and/or charcoal incorporated into colluvial material. Both terrestrial transport processes allow a secondary redeposition of charcoal fragments. Furthermore, a disintegration of larger charcoal particles by saltation is possible, as charcoal is brittle. This would also explain the absence of larger charcoal fragments in profile 7038-306 from the margin of the fen. The second aspect concerns the abundance of charcoal particles, which is much higher in the centre of the fen (7038-302) than in the peat –colluvia sequence at the margin (7038-306). The central site probably represents an integrating sink for both regional and local charcoal input, thus receiving more total charcoal. The third aspect concerns the different charcoal particle sizes and the size-class distribution. There is a longstanding discussion about charcoal particle sizes and their interpretation with respect to transport distance (e.g. Clark et al. 1998; Froyd 2006). It is generally assumed that the aeolian transport of larger particles is possible only over short distances (Patterson et al. 1987; Clark 1988), and that small particles, transported over long distances, indicate fire at regional or intercontinental sources (Clark 1988). Both charcoal records are dominated by
small particles, suggesting charcoal particles from aerial fallout, and therefore reflecting the regional fire history. However, there are also larger particles in the record of profile 7038-302, which hint at burning of local biomass, such as the probable burning of local mire vegetation. Although the charcoal particles were not identified, a large number appear to derive from herbaceous plants, which supports this assumption. Finally, the most striking aspect is the different charcoal depth curves of the profiles. Therefore, both fire records have to be evaluated with consideration of the chronological information.
Chronology The chronology of the peat profiles is based on 10 14 C AMS ages. The radiocarbon ages, the analysed materials, and d13C (‰) values are presented in Table 1. The calibrated radiocarbon ages (a cal. BC or AD, 2s confidence interval) are correlated to the chronozones as defined by Mangerud et al. (1974). In general, the five 14C ages from profile 7038-302 show a consistent succession from older to younger ages with movement from greater depth to the top. At a depth of 153–152 cm, an age of 6426–6236 a cal. BC (2s, Erl-13006) was
Table 1. Radiocarbon ages from the Islinger Mu¨hlbach Fen sequences d13CPDB (‰)
Laboratory code
Horizon
Profile 7038-302 74–75 91–92 112 –113 116 –117 152 –153
Erl-7289 Erl-7290 Erl-11878 Erl-11879 Erl-13006
III nH IV nH V nH V nH VIII nH
Wood Wood Plant remains Plant remains Bulk peat sample
3478 + 44 3551 + 46 6663 + 50 6931 + 49 7461 + 52
1916 – 1687 BC 2017 – 1745 BC 5562 – 5491 BC 5971 – 5723 BC 6426 – 6236 BC
230.6 231.2 226.9 226.0 227.4
MSB MSB MA EA – MA EA
Profile 7038-306 78–79 102 –103 114 –115 130 –131 144 –145
Erl-11886 Erl-13014 Erl-13015 Erl-13016 Erl-11887
III nH-M V nH-M V nH-M VI nH VI nH
Bulk sample Bulk sample Bulk sample Bulk peat sample Bulk peat sample
1988 + 43 3761 + 46 4346 + 48 6735 + 54 6667 + 54
94 BC – 123 AD 2335 – 2031 BC 3091 – 2891 BC 5729 – 5559 BC 5666 – 5489 BC
228.0 228.0 228.0 27.2 225.8
ESA MSB ESB EA EA
C a BP (1s)
a cal. BC or AD (2s)
Chronozone FIRE AND SETTLEMENT HISTORY
Material analysed
14
Sample depth (cm)
The dates were calibrated using the calibration dataset of Reimer et al. (2004). Calibrated ages are rounded off (by 10) where standard deviation is 50 years. The calibrated 14C ages are correlated to the chronozones defined by Mangerud et al. (1974). EA, Early Atlantic; MA, Middle Atlantic; ESB, Early Subboreal; MSB, Middle Subboreal; ESA, Early Subatlantic.
151
152
A. RAAB ET AL.
determined, which corresponds to the Early Atlantic Period. Two 14C ages from 117 –116 cm (5971– 5723 a cal. BC, 2s, Erl-11879) and 113 –112 cm (5562–5491 a cal. BC, 2s, Erl-11878) are dated to the Early Atlantic Period, specifically to the transition from the Early Atlantic to Middle Atlantic Periods. The age determinations from 92 –91 cm (2017–1745 a cal. BC, 2s, Erl-7290) and 75– 74 cm (1916–1687 a cal. BC, 2s, Erl-7289) display overlapping error ranges; however, the ages still correlate to the Middle Subboreal Period. In conclusion, the profile section from 153 to 74 cm represents the cultural periods from the Late Mesolithic Period to the Early Bronze Age. For profile 7038-306, the 14C ages from the base at 145 –144 cm depth (5666– 5489 a cal. BC, Erl-11887) and at 131 –130 cm (5729–5559 a cal. BC, Erl-13016) show an age inversion. Both age determinations correspond to the Early Atlantic Period. The subsequent three radiocarbon ages are consistent, with age determinations at 115 – 114 cm (3091–2891 a cal. BC, Erl-13015), at 103– 102 cm (2336–2031 a cal. BC, Erl-13014) and at 79 –78 cm (94 a cal. BC to 123 a cal. AD, Erl-11886). The ages correspond respectively to the Early Subboreal, Middle Subboreal and Early Subatlantic Periods. Therefore, the profile section from 145 to 78 cm represents the cultural periods from the Late Mesolithic Period to the transition from the La Te`ne Period (Iron Age) and to the Roman Empire. Although the 14C ages are generally consistent, the age determinations with overlapping error ranges of profile 7038-302 and the age inversion at the base of profile 7038-306 indicate inconsistencies in the chronology. These problems are most probably associated with the material dated. In profile 7038-302, the ages with the apparent overlapping error ranges from the top were determined from wood. It is possible that the wood is root wood and therefore dates from the same root could be an explanation. Concerning the age inversion at the base of profile 7038-306, bulk samples mainly composed of rootlets (sedge peat) were dated in both cases. Root penetration from above could be the cause of the age inversion and result in the younger ages at the base (e.g. Charman 2002). A further explanation could be contamination by mobile humic acids moving downwards and/or laterally, which might result in ages which are too young. However, with the knowledge of possible contamination by humic acids, the 14C samples were chosen from several centimetres above the peat base. Taking these factors into account, we assume that the ages are most probably too young rather than too old. In summary, these inconsistencies make it difficult to construct accurate age–depth models, and
the calculation of accumulation rates was therefore omitted. Also, the correlation of the radiocarbon ages with distinct cultural settlement periods is problematic because of the wide error ranges of the calibrated 2s ages. Nevertheless, the 14C ages provide a chronological framework for the reconstruction of environmental history and human impact.
Reconstruction of the fire history in the Siedlungskammer Burgweinting in relation to settlement and environmental history Combining the results from stratigraphic, geochemical, and microscopic charcoal investigations with the chronological information from the two sequences (7038-302 and 7038-306) derived from the Islinger Mu¨hlbach Fen and the available palaeoenvironmental and archaeological information, the following fire history and implications for the environment in the Siedlungskammer Burgweinting could be reconstructed (Fig. 9).
Late Mesolithic (c. 8000 – 5500 a cal. BC) In Central Europe during the Early Atlantic period the environmental conditions were generally stable (Kalis et al. 2003). The vegetation was natural and the landscape was covered with woodland. There are only a few traces of human occupation, although the presence of a Mesolithic population is documented by archaeological finds and by charcoal fragments in lake sediments. However, no major human impact on the environment has been found (Kalis et al. 2003). In the surroundings of the Siedlungskammer Burgweinting, during the Late Mesolithic Period (Early Atlantic Period), the Islinger Mu¨hlbach Fen was an already established landscape component. Peat growth continued and the spatial spread of the fen is indicated by the radiocarbon ages from the basal peat at the margin of the fen (profile 7038-306) dating to the Early Atlantic Period. It is assumed that Mesolithic people were present in the Regensburger Altsiedelland. Taking the problems of the radiocarbon ages discussed above into account, FE 1– 4 in profile 7038-302 should correspond to the Late Mesolithic to the Early Neolithic transition. Two phases with lower fire activity (FE 1 and FE 3) and two phases with enhanced fire activity (FE 2 and FE 4) can be distinguished (Fig. 9). The same time interval most probably correlates to FE 1 –3 or 1–4 in profile 7038-306. Charcoals were found in the upper part of the loess loam and the loess–peat transition horizon, representing FE 1. This is followed by FE 2 with the highest microscopic charcoal
FIRE AND SETTLEMENT HISTORY Fig. 9. Reconstructed fire history (charcoal records of profiles 7038-302 and 7038-306) and development of the Islinger Mu¨hlbach Fen correlated with the settlement history in the Siedlungskammer Burgweinting. 153
154
A. RAAB ET AL.
particle counts throughout the profile. The two 14C ages above FE 2 show an age inversion; however, FE 2 most probably corresponds to the Late Mesolithic Period. Significantly lower microscopic charcoal counts are present in FE 3, which is followed by still lower counts in FE 4. In summary, the depth curve shows the highest fire activity at the base of the sequence followed by decreasing fire activity. Although the charcoal sum curves of the two profiles show different patterns, the high abundance of charcoal particles in the profile sections correlating to the Late Mesolithic Period to the Neolithic transition document the relevance of fire in the environment. However, the origin of the charcoal particles, natural lightning-induced fire v. deliberate use of fire by people, has been much discussed. Some researchers (e.g. Ellenberg 1996; Kaal et al. 2008) have stated that temperate deciduous forests would rarely burn naturally because the woodland composition is not flammable. Thus, charcoal particles incorporated in peat and sediments have been interpreted to indicate human-induced fire. However, in the Regensburg region there is evidence that the woodland included a high proportion of pine. Charcoal analyses of a Linear Bandceramic site (Early Neolithic Period) near Mintraching yielded an unusually high proportion of charcoal derived from coniferous trees, and therefore forests composed of oak and pine are inferred for the Early Neolithic Period (Kreuz 1990). In addition, pollen analyses of a sediment sequence from a palaeo-meander near Sarching and of a peat – colluvium sequence of the Kirchenmoos Fen support this inference (Knipping 2005; Raab et al. 2005). It is believed that prior to and during the Early Neolithic Period, a similar woodland composition existed in the study area. Consequently, the possibility of the occurrence of lightning-induced fires cannot be completely excluded, as conifers may catch fire spontaneously (Kalis et al. 2003). The deliberate use of fire by Mesolithic hunter– gatherers is still debated (Kaal et al. 2008), although many studies in various European regions have proven the use of fire by Mesolithic people to open the landscape, increase vegetable and animal food resources, and facilitate the mobility of human groups (e.g. Clark et al. 1989; Ro¨sch 1996; Moore 2000; Urz 2000; Bos & Urz 2003; Gerlach et al. 2006; Andricˇ 2007; Kaal et al. 2008). It has been assumed that fire was used to aid hunting of big game by keeping the landscape free of dense woodland cover (e.g. Bos & Urz 2003). The charcoal record of the Islinger Mu¨hlbach Fen presumably derives partly from herbaceous plants, suggesting the burning of mire vegetation. Reed-swamp burning by Mesolithic people was suggested by Bos et al. (2005) for the Mesolithic sites near Zutphen (The Netherlands) and by Dark
(1998) for the site at Star Carr (North Yorkshire, UK). Also, Innes & Simmons (2000) concluded that substantial fires occurred on and at the edge of peat mires during the Early Mesolithic to Iron Age at North Gill (North York Moors, UK). The use of fire by Mesolithic hunter–gatherers in the Siedlungskammer Burgweinting should therefore be considered. Additional evidence that supports our assumption and proves the anthropogenic impact on the landscape during the Mesolithic Period in the Regensburg region is colluvium deposits containing charcoal-rich layers, which are present in the valley of the Kleine Laaber c. 20 km SE of Regensburg (Niller 2001).
Neolithic Period (5500– 2300 or 2200 a cal. BC) The Neolithic Period (5500–2300 or 2200 a cal. BC) is correlated to the Atlantic chronozone, the time interval between 8000 and 5000 a BP, also known as the Holocene climatic optimum or MidHolocene Hypsithermal (Meyers & Lallier-Verge´s 1999; Kalis et al. 2003). According to Pott (1992), the vegetation of oak –mixed Atlantic forests was mainly composed of Quercus, Ulmus, Tilia, Fraxinus and Alnus, but their proportions varied in space and time. In the loess areas south of Regensburg, deciduous forests also existed (Bakels 1992). Pollen analyses from the Kirchenmoos Fen showed that during the Late Neolithic Period, a more or less closed primeval deciduous forest existed, including oak, elm, lime, maple and beech (Raab et al. 2005). Pine was an important woodland component (Kreuz 1990; Knipping 2005; Raab et al. 2005). There is abundant archaeological evidence for Neolithic population in the Regensburger Altsiedlland. Several findings of ceramic shards north of the excavation site in Burgweinting document the presence of Neolithic people in this Siedlungskammer. The settlements were present on the lower terrace of the Danube; for example, the Linear Bandceramic site at Mintraching (Kreuz 1990). This indicates a lower groundwater level for this time period. The introduction of the Neolithic subsistence system involved the establishment of permanent settlements and the introduction of agriculture and animal husbandry, and therefore comprises a variety of related activities that would have an impact on the environment in the settlement surroundings. Nevertheless, it is assumed that the human impact on the woodland was comparatively low. Lu¨ning & Kalis (1992) concluded that only small and disjointed areas around the settlements were cleared of forests. Behre (2000) assumed that
FIRE AND SETTLEMENT HISTORY
open grazed forests existed around the settlements. It is estimated that 5–6% of the woodland in South and West Germany was thinned out (Lu¨ning & Kalis 1992). According to Bakels (1992) the people of the Linear Bandceramic culture, settling on the loess soils in the Netherlands, Belgium and Northern France, did not open large areas of the forests, but as a result of their impact the woodland composition changed. In the Kirchenmoos sequence, a colluvium deposited on the mire during the Neolithic Period suggests intensified human activity (Raab et al. 2005). In profile 7038-302 from the centre of the Islinger Mu¨hlbach Fen, the section from c. 111 to 92 cm depth correlates to the Neolithic Period based on radiocarbon dating. Peat formation continued, although the degree of humification increases from highly humified peat to amorphous peat. Furthermore, radiocarbon dating of this section indicates a decline in peat growth (Fig. 9). It is assumed that a change in the hydrological balance (e.g. a lowered local water table resulting in increased aeration of the upper peat layer) could be responsible. The reason for the enhanced decomposition, whether caused by natural climatic conditions or by human impact, is debatable. In the charcoal record of profile 7038-302, the Neolithic Period is represented between FE 4 and FE 5 (Fig. 9), with comparatively high but decreasing charcoal abundance towards lower profile depth. In profile 7038-306, the Neolithic Period is included in the section from c. 130 to c. 105 cm. In this section, a stratigraphic change from peat to organicrich colluvium is present at 117 cm, indicating that peat formation was interrupted by colluvial deposition. Colluvium formation is a direct evidence for deforestation and agricultural use of the adjacent slopes to the south. In the same profile section, an abrupt rise in total nitrogen content is detected (Fig. 6). This could indicate nitrification by human activity at the margin of the fen, adjacent to an agricultural field or a track along the fen. A high degree of nitrification caused by local human presence was also found by Urz (2000) and interpreted from macrofossil analyses. The Neolithic Period most probably corresponds to FE 4 and 5. Whereas during FE 4 (peat), the microscopic charcoal counts are comparatively low, the abundance of charcoal particles increases in FE 5 (organic-rich colluvium). During the Neolithic Period several sources for microscopic charcoal production have to be considered, such as woodland clearance for agricultural use or domestic fireplaces. Concerning charcoal transport, other than aerial transport and deposition, at the margin of the fen, water transport by surface runoff is possible, as well as incorporation into colluvial material and subsequent deposition.
155
Near-surface transport can also account for the disintegration of the charcoal particles (Umbanhowar & McGrafth 1988; Clark et al. 1998). In summary, the Neolithic Period is reflected in profile 7038-302 by a decreasing charcoal abundance. This is in accordance with the results of the charcoal record from profile 7038-306. However, comparing the charcoal peak curves, it becomes obvious that pre-Neolithic burning was more important than later biomass burning. This could be connected with the availability of combustible material in the vicinity of the fen. It is also possible that slash-and-burn practices became less important because they were no longer needed, and instead fire was used only to keep the agricultural fields open.
Bronze Age (2300 or 2200 – 800 a cal. BC) At the beginning of the Bronze Age at 3500 a BP, stable oceanic climatic conditions existed. In general, according to Magny (1982), drier phases persisted during the Early Bronze Age (2200– 1800 a cal. BC) and during the Urnfield Period (1200– 800 a cal. BC), and between these periods, the climate was slightly wetter. The relatively dry climatic conditions during the Urnfield Period are inferred from lower lake levels throughout Europe. In the Siedlungskammer Burgweinting, the absence of finds for the cultural periods from the End-Neolithic (c. 2200 a BC) to the Urnfield Period (c. 1200 a cal. BC) (Zuber 2006) suggests a hiatus in settlement activities (Siedlungsruhe) lasting for about 1000 years. Likewise, in the Federsee region (West Germany), the Early Bronze age is not documented by archaeological finds, and therefore a period of Siedlungsruhe is inferred (Maier & Vogt 2007). In contrast, pedological and mire investigations by Maier & Vogt (2007) proved that the Early Bronze age period is reflected by particularly high sedimentation rates and a striking abundance of charcoal fragments. We therefore suggest that further archaeological research in this area is required. At the beginning of the Urnfield Period, a noticeable rise in population is recorded in the Siedlungskammer Burgweinting and in the surroundings of Regensburg. So-called Urnfield hilltop settlements were established; for example, at the Weltenburger Frauenberg, the Bogenberg near Straubing, and the Schloßberg above Kallmu¨nz (Schauer 1998; Rind 1999; Neudert 2003; Sandner 2005). The sediment sequence from the palaeomeander of the river Danube near Sarching shows intensified soil erosion and deposition of alluvial clay, proving an increased human impact (Knipping 2005). This is also reflected in the pollen diagram with an alteration of the woodland composition; however, pine remains an important forest
156
A. RAAB ET AL.
constituent (Knipping 2005). In the Kirchenmoos area, the forest-free areas most probably did not greatly increase, and the woodland composition remained similar to that during the Neolithic Period with the addition of hornbeam (Raab et al. 2005). For the first time, the pollen diagram from the sequence near Sarching shows a closed curve of cereal-type pollen, suggesting a continuous agricultural practice (Knipping 2005); also, cereal cultivation is documented in the Kirchenmoos area (Raab et al. 2005). According to Ku¨ster (1995), during the Early Bronze Age (Chamer Gruppe) a clear change in the economic basis of agriculture took place, with a decreasing importance of wheat (T. monocuccum) and increasing importance of spelt (T. spelta) and barley (Hordeum vulgare). Furthermore, wheats (T. aestivum, T. monocuccum and T. dicoccum) and millets (Panicum miliaceum and Setaria italica) were cultivated. From that point forward, no substantial change in cultivated plants occurred in Bavaria until the Middle Ages (Ku¨ster 1995). In the peat profile 7038-302, the section between 92 and 74 cm probably corresponds to the Bronze Age. The overlapping error ranges of the radiocarbon ages exclude a correlation with distinct settlement periods. Based on the charcoal record, FE 5–8 represent the Bronze Age. Following the charcoal peak at the Mesolithic –Neolithic transition, a decrease in charcoal abundance is visible towards FE 5. The samples between 82 and 69 cm depth were continuously analysed to gain more detailed information. A short-term increase in charcoal abundance is visible in FE 6. Subsequently, during FE 7 the lowest charcoal abundance throughout the profile is detected, which is followed by the highest charcoal abundance (FE 8) throughout the profile. It is assumed that FE 7 might correspond to the Early to Middle Bronze Age settlement gap with a lack of finds in the archaeological excavation, even though this 5 cm thick section might represent less than 1000 years. FE 8 could then represent the Urnfield period (1200–800 a cal. BC) with very high charcoal abundance. In profile 7038-306, the radiocarbon age from 102 to 103 cm depth shows that this profile section belongs to the Bronze Age. Organic-rich colluvia are present, which cannot be better defined because of the compaction problem associated with the coring technique, but the mineral substrate proves soil erosion. This section corresponds to FE 6 with comparatively low charcoal abundance, which could also reflect the Early to Middle Bronze Age break in settlement activity. The following Urnfield Period could be represented by the increasing charcoal abundance towards FE 7 in the core. The charcoal peak in the record does not seem adequate to reflect the intensive use of the
Siedlungskammer Burgweinting during the Urnfield Period, documented in the archaeological excavations (Zuber 2004b, 2006). An explanation might be that the landscape was already opened and forest clearance by fire was no longer practised. Furthermore, there must have been a high need for firewood, and wood for construction and other settlement activities and use in burial customs.
Bronze Age (2300 or 2200 – 800 a cal. BC) to Iron Age transition, Iron Age (800– 15 a cal. BC) and Roman Empire (15 BC – AD 476) According to Dark (2006) a major climate downturn occurred at 850 a cal. BC, at the Bronze Age to Iron Age transition, causing settlement abandonment in western Europe. A global cooling around 2800 a cal. BP (850 a cal. BC) has also been described by Denton & Karle`n (1973), Bond et al. (1997, 2001) and van Geel et al. (1999). According to the pollen diagram from Sarching (Knipping 2005), an intensified anthropogenic impact on the vegetation is noticeable, with a decrease in the arboreal pollen percentages indicating that the woodland is strongly reduced. Pine remains the dominant woodland constituent, but the QM (Quercetum Mixtum) species (Quercus, Ulmus, Tilia, Acer, Fraxinus) and Fagus as well as Salix percentages decrease. The presence of Plantago lanceolata pollen documents pastures and increased cereal pollen percentages indicate increased agricultural use. No further information is available from this profile, because the pollen analytical processing of younger sediments is not possible. The results of the archaeological excavation in the Siedlungskammer Burgweinting indicate a significant local population decline following the Urnfield Period. According to Brunnacker (1994), during Celtic times prior to the Roman invasion, agriculture was comparatively highly developed, as indicated by soil erosion. Intensive soil erosion induced by agricultural use is documented by studies at the Celtic square enclosure (Viereckschanze) Poign (Leopold 2003; Leopold & Vo¨lkel 2007). The charcoal record of profile 7038-302 shows a decreasing charcoal abundance following the Bronze Age until the end of the analysed profile at 55 cm depth (FE 9). In contrast, profile 7038-306 shows an increase in charcoal abundance towards 78 –79 cm depth, which is dated to 94 a cal. BC – AD 123 a cal. (Erl-11886, 2s) corresponding to the transition from the La Te`ne Age to Roman Empire. In profile 7038-306 the La Te`ne Age to Roman Empire transition corresponds to the peat –colluvia complex, documenting soil erosion induced by agricultural use. The charcoal record shows a steady
FIRE AND SETTLEMENT HISTORY
increase of charcoal abundance (FE 7), which remains high with a slight bias to decreasing abundance towards the top. Based on archaeological research, no immediate succession from the La Te`ne Age to the Roman Empire has been detected in the Burgweinting excavation or in the Regensburg region. This is in contrast to the charcoal record of profile 7038306, which shows that charcoal is continuously present during the corresponding time interval, with increasing charcoal abundance towards the Roman Empire times. The settlement conditions during the time of the Roman Empire were certainly supported by the favourable climatic conditions. Based on archaeological finds, the settlement by a Roman rural population in the Siedlungskammer Burgweinting starts contemporaneously with the establishment of the legionary fort at Kumpfmu¨hl (c. AD 79). Four villae rusticae present in an area of 25 ha document an intensive use of the Siedlungskammer Burgweinting. Assuming an estimated agricultural land of 50 –120 ha per villa rustica, the woodland must have been strongly reduced during the Roman Empire because of the enormous need for wood for domestic use and constructions by the rural population, as well as the civil urban population and the military base. Information from the charcoal records for Roman Empire times is available only from profile 7038-306 corresponding to FE 7, with a high charcoal abundance in the peat –colluvia substrate, which documents agricultural use of the adjacent field south of the fen. The top of the profile is formed by colluvium with low organic content, indicating that peat formation at the margin of the fen was completely stopped by colluvium deposition continuing most probably to the Middle Ages.
Conclusions (1) Charcoal fragments that are present in the investigated profiles of the Islinger Mu¨hlbach Fen indicate the occurrence of fire in the environment. Although the possibility that the charcoal particles derive from wildfires cannot be completely ruled out, the charcoal records are interpreted to be most probably produced by primarily anthropogenically induced fires, as the presence of humans in the study area for thousands of years is documented by archaeological evidence. (2) The two profiles investigated, one from the centre of the fen and one from the margin, represent different sinks for palaeoenvironmental data. However, the records provide overlapping and complementary information about environmental history and human impact, using the profile
157
stratigraphy, geochemical analyses and microscopic charcoal analyses as a fire proxy. (3) The chronological control of the cores is problematic, but the ages determined, in combination with the charcoal records, suggest the deliberate use of fire since Mesolithic times. Mesolithic hunter –gatherers could have practised reed-swamp burning to open the vegetation for hunting. For further evidence, the determination of the charcoal species would be helpful. Based on the high abundance of charcoal particles in the profile sections correlating to the Mesolithic Period, fire played a prominent role in the environment. The Mesolithic population is barely documented by archaeological evidence or by palaeoecological data such as pollen records, but is indicated by the charcoal records. The deliberate use of fire by the Mesolithic hunter –gatherers would furthermore imply that the Neolithic farmers did not enter a primeval environment. (4) The continuous occurrence of charcoal particles in changing quantities indicates continuous settlement activities. The fire episodes distinguished extend over several centimetres or decimetres of the records, which represent hundreds to thousands of years. Because of the wide error ranges of the radiocarbon ages, the fire episodes cannot be correlated to distinct settlement periods. (5) In the charcoal record, there is a single section with low charcoal abundance, which might reflect a general decline in the intensity of human activity in the catchment. The absence of finds for the settlement periods from the End-Neolithic to Urnfield Period, which is interpreted as a Siedlungsruhe, could correspond to this decline in charcoal abundance. According to the archaeological information, this time interval extends over c. 1000 years. In contrast, in the record this period is represented by a peat section that is only 5 cm thick, suggesting either that the duration of this Siedlungsruhe might be shorter than 1000 years or that peat accumulation was very low during this time interval. (6) The prominent role of the Urnfield period in the settlement history of the Siedlungskammer Burgweinting, lasting for about 300 years, is not as markedly represented in the charcoal records as one might assume. A possible explanation could be that the woods were already opened and forest clearance by fire for arable land was not necessary. Sources for charcoal production are domestic use, processing of bronze and cultural activities. Most probably because of the high population density, there was great a demand for firewood, wood for construction, tools and other related settlement activities. (7) Further information about the human– landscape interrelationship, and especially the
158
A. RAAB ET AL.
relationship between fire and vegetation, is expected from pollen analyses, which are in progress. This study is part of the research project ‘Torfe, Kolluvien und Bo¨den als Archive zur Rekonstruktion der Pala¨oumwelt im Regensburger Altsiedelland (Burgweinting, Stadt Regensburg)’ (RA 1129/2-1), funded by the Deutsche Forschungsgemeinschaft (DFG). We thank L.-M. Dallmeier, J. Zuber (both Office of Archive and Monument Protection of the City of Regensburg) and S. Codreanu-Windauer (Bavarian Office for the Protection of Monuments, Regensburg), for providing archaeological information and making the data from the excavation at RegensburgBurgweinting available.
References Ad-Hoc Ag Boden 2005. Bodenkundliche Kartieranleitung. Schweizerbart, Stuttgart. Amt Fu¨r Archiv und Denkmalpflege, Amt Fu¨r Wirtschaftsfo¨rderung & Museen der Stadt Regensburg (eds) 2004. Von der Steinzeit bis zum Mittelalter. 10 Jahre Fla¨chengrabung in RegensburgBurgweinting. Schmidt, Neustadt an der Aisch. Andricˇ, M. 2007. Holocene vegetation development in Bela krajina (Slovenia) and the impact of first farmers on the landscape. Holocene, 17, 763– 776. Anon. 1994. Munsell Soil Color Charts, revised edition. Gretag Macbeth, New Windsor, NY. Bakels, C. C. 1992. Research on land clearance during the Neolithic in the loess regions of The Netherlands, Belgium, and Northern France. In: Frenzel, B. (ed.) Evaluation of Land Surfaces Cleared from Forests by Prehistoric Man in Early Neolithic Times and the Time of Migrating Germanic Tribes. Fischer, Stuttgart, 47–55. Bayerisches Landesamt Fu¨r Umwelt 2007. Tiefengrundwasserleiter Malm unter der Molasse. World Wide Web Address: http://www.lfu.bayern.de/geologie/daten/hydrogeologie/tiefengrundwasserleiter/ doc/tiefengrundwasserleiter_malm_unter_ molasse.pdf. Bayforklim 1996. Klimaatlas von Bayern. Bayerischer Klimaforschungsverbund. Meteorologisches Institut Mu¨nchen, Mu¨nchen. Beaty, R. M. & Taylor, A. H. 2009. A 14 000 year sedimentary charcoal record of fire from the northern Sierra Nevada, Lake Tahoe Basin, California, USA. Holocene, 19, 347– 358. Behre, K.-E. 2000. Der Mensch o¨ffnet die Wa¨lder— zur Entstehung der Heide und anderer Offenlandschaften. In: Bayerische Akademie der Wissenschaften (ed.) Entwicklung der Umwelt seit des letzten Eiszeit, 18. Rundgespra¨che der Kommission ¨ kologie. Dr Friedrich Pfeil Verlag, Mu¨nchen, fu¨r O 103– 116. Birks, H. J. B. & Birks, H. H. 1980. Quaternary Palaeoecology. Edward Arnold, London. Blundell, A., Charman, D. J. & Barber, K. 2008. Multiproxy late Holocene peat records from Ireland: towards a regional palaeoclimate curve. Journal of Quaternary Science, 23, 59– 71.
Bond, G., Showers, W. et al. 1997. A pervasive millennial-scale cycle in North Atlantic Holocene and Glacial climates. Science, 287, 1257–1266. Bond, G., Kromer, B. et al. 2001. Persistent solar influence on North Atlantic climate during the Holocene. Science, 294, 2130–2136. Boos, A. 2004. Die Ergebnisse von 10 Jahren Fla¨chengrabung als Ausstellung. In: Amt Fu¨r Archiv und Denkmalpflege, Amt Fu¨r Wirtschaftsfo¨rderung & Museen der Stadt Regensburg (eds) Von der Steinzeit bis zum Mittelalter. 10 Jahre Fla¨chengrabung in Regensburg-Burgweinting. Schmidt, Neustadt an der Aisch, 9 –12. Bos, J. A. A. & Urz, R. 2003. Late Glacial and early Holocene environment in the middle Lahn river valley (Hesse, central–west Germany) and the local impact of early Mesolithic people—pollen and macrofossil evidence. Vegetation History and Archaeobotany, 12, 19–36. Bos, J. A. A., Van Geel, B., Groenewoudt, B. J. & Lauwerier, C. G. M. 2005. Early Holocene environmental change, the presence and disappearance of early Mesolithic habitation near Zutphen (The Netherlands). Vegetation History and Archaeobotany, 15, 27– 43. Bowman, D. M. J. S., Balch, J. K. et al. 2009. Fire in the earth system. Science, 324, 481–484. Brunnacker, K. 1958. Die Bo¨den. In: Oschmann, F. (ed.) Erla¨uterungen zur Geologischen Karte von Bayern 1:25 000 Blatt Nr. 7038 Bad Abbach. Bayerisches Geologisches Landesamt, Mu¨nchen, 169– 170. Brunnacker, K. 1994. Bodenkunde und Besiedlungswesen no¨rdlich der Alpen (in den ro¨mischen Rhein– Donauprovinzen vom 1. bis 5. Jh. n. Chr.). In: Bender, H. & Wolff, J. (eds) La¨ndliche Besiedlung und Landwirtschaft in den Rhein–Donau-Provinzen des Ro¨mischen Reiches. Marie Leidorf, Espelkamp, 1– 5. Buch, M. W. 1988. Spa¨tpleistoza¨ne und holoza¨ne fluviale Geomorphodynamik im Donautal zwischen Regensburg und Straubing. Regensburger Geographische Schriften, 21. Carcaillet, C., Bouvier, M., Fre´chette, B., Larouche, A. C. & Richard, P. J. H. 2001. Comparison of pollen-slide and sieving methods in lacustrine charcoal analyses for local and regional fire history. Holocene, 11, 467– 476. Chambers, F. 2006. Humification protocol. World Wide Web Address: http://www2.glos.ac.uk/accrotelm/ humoproto.html. Charman, D. 2002. Peatlands and Environmental Change. Wiley, Chichester. Clark, J. S. 1988. Particle motion and the theory of charcoal analysis: source area, transport, deposition and sampling. Quaternary Research, 30, 67–80. Clark, J. S., Merkt, J. & Mu¨ller, H. 1989. Post-glacial fire, vegetation, and human history on the Alpine forelands, south-western Germany. Journal of Ecology, 77, 897–925. Clark, J. S., Lynch, J., Stocks, B. J. & Goldammer, J. G. 1998. Relationships between charcoal particles in air and sediments in west–central Siberia. Holocene, 8, 19–29. Conedera, M., Tinner, W., Neff, C., Meurer, M., Dickens, A. F. & Krebs, P. 2009. Reconstructing
FIRE AND SETTLEMENT HISTORY past fire regimes: methods, applications, and relevance to fire management and conservation. Quaternary Science Reviews, 28, 555– 576. Dallmeier, L.-M. 2004. Ein gutes Jahrhundert archa¨ologische Forschungen in Burgweinting. Ein Ort mit Jahrtausende alter Vergangenheit. In: Amt Fu¨r Archiv und Denkmalpflege, Amt Fu¨r Wirtschaftsfo¨rderung & Museen der Stadt Regensburg (eds) Von der Steinzeit bis zum Mittelalter. 10 Jahre Fla¨chengrabung in Regensburg-Burgweinting, Schmidt, Neustadt an der Aisch, 13– 21. Dark, P. 1998. The lake centre sequence: results. In: Mellars, P. & Dark, P. (eds) Star Carr in Context: New Archaeological and Palaeoecological Investigations at the Early Mesolithic Site of Star Carr, North Yorkshire. McDonald Institute for Archaeological Research, Cambridge, 163– 178. Dark, P. 2006. Climate deterioration and land-use change in the first millennium BC: perspectives from the British palynological record. Journal of Archaeological Science, 33, 1381–1385. Denton, G. H. & Karle`n, W. 1973. Holocene climatic variations—their pattern and possible cause. Quaternary Research, 3, 155–205. Edwards, K. J. & Whittington, G. 2000. Multiple charcoal profiles in a Scottish lake: taphonomy, fire ecology, human impact and inference. Palaeogeography, Palaeoclimatology, Palaeoecology, 164, 67–86. Ellenberg, H. 1996. Vegetation Mitteleuropas mit den ¨ kologischer, Dynamischer und Historischer Alpen in O Sicht. Ulmer, Stuttgart. Enache, M. & Cumming, B. 2006. Tracking recorded fires using charcoal morphology from sedimentary sequence of Prosser Lake, British Columbia (Canada). Quaternary Research, 65, 282– 292. Froyd, C. A. 2006. Holocene fire in the Scottish Highlands: evidence from macroscopic charcoal records. Holocene, 16, 235– 249. Gerlach, R., Baumewerd-Schmidt, H., van den Borg, K., Eckmeier, E. & Schmidt, M. W. I. 2006. Prehistoric alteration of soil in the Lower Rhine Basin, Northwest Germany—archaeological, 14C and geochemical evidence. Geoderma, 136, 38– 50. Grosse-Brauckmann, G. 1990. Stoffliches. Ablagerungen der Moore. In: Go¨ttlich, K.-H. (ed.) Moor- und Torfkunde. Schweizerbart, Stuttgart, 175– 230. Heiri, O., Lotter, A. F. & Lemcke, G. 2001. Loss on ignition as a method for estimating organic and carbonate content in sediments: reproducibility and comparability of results. Journal of Paleolimnology, 25, 101–110. Innes, J. B. & Simmons, I. G. 2000. Mid-Holocene charcoal stratigraphy, fire history and palaeoecology at North Gill, North York Moors, UK. Palaeogeography, Palaeoclimatology, Palaeoecology, 164, 151–165. Jerz, H. 1983. Kalksinterbildungen in Su¨dbayern und ihre zeitliche Einstufung. Geologisches Jahrbuch, A71, 291–300. Kaal, J., Cortizas, A. M., Eckmeier, E., Casais, M. C., Este´vez, S. & Boado, F. C. 2008. Holocene fire history of black colluvial soils reveald by pyrolysisGC/MS: a case study from Campo Lameiro (NW Spain). Journal of Archaeological Science, 35, 2133–2143.
159
Kalis, A. J., Merkt, J. & Wunderlich, J. 2003. Environmental changes during the Holocene climatic optimum in central Europe—human impact and natural causes. Quaternary Science Reviews, 22, 33–79. Kirpal, U. 2005. Neue schnurkeramische Funde aus Burgweinting, Stadt Regensburg. Beitra¨ge zur Archa¨ologie in der Oberpfalz und in Regensburg, 7, 31–38. Knipping, M. 2005. Pollenanalytische Untersuchungen an einem Pala¨oma¨ander der Donau bei Sarching (Lkr. Regensburg). Hoppea, 66, 495–502. Kreuz, A. M. 1990. Die ersten Bauern Mitteleuropas— Eine archa¨obotanische Untersuchung zu Umwelt und Landwirtschaft der A¨ltersten Bandkeramik. University of Leiden, Leiden. Ku¨ster, H. 1995. Postglaziale Vegetationsgeschichte Su¨dbayerns. Geobotanische Studien zur pra¨historischen Landschaftskunde. Akademie, Berlin. Langdon, P. G., Barber, K. E. & Hughes, P. D. M. 2003. A 7500-year peat-based palaeoclimatic reconstruction and evidence for an 1100-year cyclicity in bog surface wetness from Temple Hill Moss, Pentland Hills, southeast Scotland. Quaternary Science Reviews, 22, 259– 274. Lawson, I., Frogley, M., Bryant, C., Preece, R. & Tzedakis, P. 2004. The Lateglacial and Holocene environmental history of the Ioannina basin, northwest Greece. Quaternary Science Reviews, 23, 1599– 1625. Leopold, M. 2003. Multivariate Analyse von Geoarchiven zur Rekonstruktion eisenzeitlicher Landnutzung im Umfeld der spa¨tlate`nezeitlichen Viereckschanze von Poign, Lkr. Regensburg. Regensburger Beitra¨ge zur Bodenkunde, Landschaftso¨kologie und Quarta¨rforschung, 2. World Wide Web Address: http:// www.bibliothek.uni-regensburg.de/opus/volltexte/ 2003/233/. Leopold, M. & Vo¨lkel, J. 2007. Quantifying prehistoric soil erosion—a review of soil loss methods and their application to a celtic square enclosure (Viereckschanze) in southern Germany. Geoarchaeology, 22, 873 –889. Lu¨ning, J. 1991. Fru¨he Bauern in Mitteleuropa im 6. und 5. Jahrtausend v. Chr. Jahrbuch des Ro¨mischGermanischen Zentralmuseums Mainz, 35, 27– 93. Lu¨ning, J. & Kalis, A. J. 1992. The influence of Early Neolithic settlers on the vegetation of the Lower Rhinelands and the determination of cleared areas based on archaeological and palynological criteria. In: Frenzel, B. (ed.) Evaluation of land Surfaces Cleared from Forests by Prehistoric Man in Early Neolithic Times and the Time of Migrating Germanic Tribes. Fischer, Stuttgart, 41–46. Magny, M. 1982. Atlantic and Sub-boreal: dampness and dryess. In: Harding, A. (ed.) Climatic Change in Later Prehistory. Edinburgh University Press, Edinburgh, 33–43. Maier, U. & Vogt, R. 2007. Pedologisch-moorkundliche Untersuchungen zur Landschafts- und Besiedlungsgeschichte des Federseegebiets. Stuttgarter Geographische Studien, 138. Mangerud, J., Andersen, S. T., Berglund, B. E. & Donner, J. J. 1974. Quaternary stratigraphy of Norden, a proposal for terminology and classification. Boreas, 3, 109– 128.
160
A. RAAB ET AL.
Marguerie, D. & Hunot, J.-Y. 2007. Charcoal analysis and dendrology: data from archaeological sites in north-western France. Journal of Archaeological Science, 34, 1417–1433. Marinova, E. & Thiebault, S. 2008. Anthracological analysis from Kovacevo, southwest Bulgaria: woodland and its use during the earliest stages of the European Neolithic. Vegetation History and Archaeobotany, 17, 223– 231. Meyers, P. A. & Lallier-Verge´s, E. 1999. Lacustrine sedimentary organic matter records of Late Quaternary paleoclimates. Journal of Palaeolimnology, 21, 345– 372. Moore, J. 2000. Forest fire and human interaction in the early Holocene woodlands in Britain. Palaeogeography, Palaeoclimatology, Palaeoecology, 164, 125– 137. Moosbauer, G. 2004. Die ro¨mischen Gutsho¨fe. In: Amt Fu¨r Archiv und Denkmalpflege, Amt Fu¨r Wirtschaftsfo¨rderung & Museen der Stadt Regensburg (eds) Von der Steinzeit bis zum Mittelalter. 10 Jahre Fla¨chengrabung in RegensburgBurgweinting. Schmidt, Neustadt an der Aisch, 64– 69. Neudert, C. 2003. Archa¨ologische Untersuchungen im Umland des Frauenberges, Ndb. Regensburger Beitra¨ge zur Pra¨historischen Archa¨ologie, 11. Niller, H.-P. 1998. Pra¨historische Landschaften im Lo¨ßgebiet bei Regensburg—Kolluvien, Auenlehme und Bo¨den als Archive der Pala¨oumwelt. Regensburger Geographische Schriften, 429. Niller, H.-P. 2001. Wandel pra¨historischer Landschaften. Kolluvien, Auenlehme und Bo¨den: Archive zur Rekonstruktion vorgeschichtlicher anthropogener Landschaftsvera¨nderungen im Lo¨ßgebiet bei Regensburg. Erdkunde, 55, 32–49. Oschmann, F. 1958. Erla¨uterungen zur Geologischen Karte von Bayern 1:25 000, Blatt Nr. 7038 Bad Abbach. Bayerisches Geologisches Landesamt, Mu¨nchen. Paetzold, D. 1992. Bemerkungen zum Siedlungsverhalten neolithischer bis late`nezeitlicher Bevo¨lkerung zwischen Regensburg und Deggendorf. Bayerische Vorgeschichtsbla¨tter, 57, 77– 102. Patterson, W. A. I., Edwards, K. J. & Maguire, D. J. 1987. Microscopic charcoal as a fossil indicator of fire. Quaternary Science Reviews, 6, 3– 23. Pott, R. 1992. The impact of Early Neolithic agriculture on the vegetation of Northwestern Germany. In: Frenzel, B. (ed.) Evaluation of Land Surfaces Cleared from Forests by Prehistoric Man in Early Neolithic Times and The Time of Migrating Germanic Tribes. Fischer, Stuttgart, 57– 72. Raab, A., Leopold, M. & Vo¨lkel, J. 2005. Vegetation and land-use history in the surroundings of the Kirchenmoos (Central Bavaria, Germany) since the late Neolithic period to the early middle ages. Zeitschrift fu¨r Geomorphologie, 139, 35–61. Raab, A., Kunz, S. & Bru¨tzke, W. 2008. Rekonstruktion der Pala¨oumwelt am Beispiel der Siedlungskammer von Burgweinting, Stadt Regensburg. In: Schauer, P. (ed.) Festgabe 40 Jahre Lehrstuhl fu¨r Vor- und Fru¨hgeschichte der Universita¨t Regensburg 1968– 2008, 20. Universita¨tsverlag, Regensburg, 367– 393.
Reimer, P. J., Baillie, M. G. L. et al. 2004. IntCal04— terrestrial radiocarbon age calibration, 0 –26 cal kyr BP. Radiocarbon, 46, 1029–1058. Rhodes, A. N. 1998. A method for the preparation and quantification of microscopic charcoals from terrestrial and lacustrine sediment cores. Holocene, 8, 113– 117. Rind, M. 1999. Der Frauenberg oberhalb Kloster Weltenburg I, Teilband 1. Regensburger Beitra¨ge zur Pra¨historischen Archa¨ologie, 6. Ro¨sch, M. 1996. New approaches to prehistoric land-use reconstruction in south-western Germany. Vegetation History and Archaeobotany, 5, 65–79. Sandner, R. 2005. Siedlungsarcha¨ologische Untersuchungen auf dem Schloß-, dem Kirchen- und dem Hirmesberg oberhalb Kallmu¨nz, Lkr. Regensburg/Opf. Regensburger Beitra¨ge zur Pra¨historischen Archa¨ologie, 14. Schauer, P. 1998. Umweltbedingungen und Siedelverhalten zur Urnenfelderzeit: Das Fallbeispiel der befestigten Großsiedlung auf dem Bogenberg, Lkr. Straubing-Bogen, Niederbayern. In: Ku¨ster, H., Lang, A. & Schauer, P. (eds) Archa¨ologische Forschungen inurgeschichtlichenSiedlungslandschaften – Festschrift fu¨r Georg Kossack zum 75. Regensburger Beitra¨ge zur Pra¨historischen Archa¨ologie, 5, 317– 354. Schier, W. 1985. Zur vorro¨mischen Besiedlung su¨do¨stlich von Regensburg mit einem Beitrag von J. Boessneck & J. Scha¨ffer. Bayerische Vorgeschichtsbla¨tter, 50, 9– 90. Schro¨ter, P. 2004. Gra¨ber der Glockenbecherkultur. In: Amt Fu¨r Archiv und Denkmalpflege, Amt Fu¨r Wirtschaftsfo¨rderung & Museen der Stadt Regensburg (eds) Von der Steinzeit bis zum Mittelalter. 10 Jahre Fla¨chengrabung in RegensburgBurgweinting. Schmidt, Neustadt an der Aisch, 25–30. ¨ bersichtskarte der natu¨rlichen VegSeibert, P. 1969. U etation von Bayern 1:500 000 mit Erla¨uterungen. Schriften fu¨r Vegetationskunde, 3. Swain, A. M. 1978. Environmental changes during the last 2000 years in north– central Wisconsin: analysis of pollen, charcoal, and seeds from varved lake sediments. Quaternary Research, 10, 55–68. Tolonen, K. 1986. Charred particle analysis. In: Berglund, B. E. (ed.) Handbook of Palaeoecology and Palaeohydrology. Wiley, New York, 485 –496. Torbru¨gge, W. 1984. Die Landschaften um Regensburg in vor- und fru¨hgeschichtlicher Zeit. In: RieckhoffPauli, S. & Torbru¨gge, W. (eds) Regensburg– Kelhein– Straubing I. Fu¨hrer zu archa¨ologische Denkma¨lern in Deutschland, 5, 28–121. Umbanhowar, C. E. J. & Mcgrafth, M. 1988. Experimental production and analyses of microscopic charcoal from wood, leaves and grasses. Holocene, 8, 341–346. Unger, H. J. & Doppler, G. 1996. Ju¨ngste tertia¨re Ablagerungen im Molassebecken (Schotter, plioza¨n bis a¨ltestpleistoza¨n), G. In: Bayerisches Geologisches Landesamt (ed.) Erla¨uterungen zur Geologischen Karte von Bayern 1:500 000. Bayerisches Geologisches Landesamt, Mu¨nchen, 185– 187. Urz, R. 2000. Begraben unter Auelehm: Fru¨hmesolithische Siedlungsspuren im mittleren Lahntal. Archa¨ologisches Korrespondenzblatt, 30, 33– 43.
FIRE AND SETTLEMENT HISTORY Van Geel, B., Raspopov, O. M., Renssen, H., Van Der Pflicht, J., Dergachev, V. A. & Meijer, H. A. J. 1999. The role of solar forcing upon climate change. Quaternary Science Reviews, 18, 331– 338. Vo¨lkel, J., Leopold, M. & Weber, B. 2002. Neue Befunde zur Landschaftsentwicklung im niederbayerischen Donauraum wa¨hrend der Zeitenwende (Keltisches Oppidum von Manching und Viereckschanze von Poign bei Bad Abbach). Zeitschrift fu¨r Geomorphologie, N.F., 128, 47–66. Whitlock, C. & Larsen, C. 2001. Charcoal as a fire proxy. In: Smol, J. P., Birks, H. J. B. & Last, W. M. (eds) Tracking Environmental Change Using Lake Sediments. Volume 3 Terrestrial, Algal, and Siliceous Indicators. Cambridge University Press, Cambridge, 75– 97. Yeloff, D. E. & Mauquoy, D. 2006. The influence of vegetation composition on peat humifications for palaeoclimatic studies. Boreas, 35, 662–673.
161
Zuber, J. 2002. Ein Friedhof der spa¨ten Bronze- und der Urnenfelderzeit in Burgweinting. Das Archa¨ologische Jahr in Bayern 2002, 42– 45. Zuber, J. 2004a. Gra¨ber und Siedlungen der Bronzeund Eisenzeit. In: Amt Fu¨r Archiv und Denkmalpflege, Amt Fu¨r Wirtschaftsfo¨rderung & Museen der Stadt Regensburg (eds) Von der Steinzeit bis zum Mittelalter. 10 Jahre Fla¨chengrabung in Regensburg-Burgweinting. Schmidt, Neustadt an der Aisch, 31– 56. Zuber, J. 2004b. Bemerkungen zum Stand der Ausgrabungen in Burgweinting. Denkmalpflege in Regensburg, 6, 9. Zuber, J. 2006. Alle Jahre im April. 12 Jahre archa¨ologische Ausgrabung in Burgweinting. In: Trapp, E. & Dallmeier, L.-M. (eds) Denkmalpflege in Regensburg. Beitra¨ge zur Denkmalpflege in Regensburg fu¨r die Jahre 2003 bis 2005, Universita¨tsverlag, Regensburg, 10, 31–59.
Raw material economics in their environmental context: an example from the Middle Palaeolithic of southern France LUCY WILSON Department of Biology, University of New Brunswick in Saint John, 100 Tucker Park Road, Saint John, NB E2L 4L5, Canada (e-mail:
[email protected]) Abstract: To understand the human behaviour reflected in stone tool assemblages, we must take into account the characteristics of the lithic resources, their distribution across the landscape, the characteristics of the landscape itself, the distribution of other resources such as water and food, and human strategies of mobility and resource exploitation. The assemblage from one layer of a Middle Palaeolithic rock shelter site, the Bau de l’Aubesier, shows that raw materials from different areas were used in different ways: they are more or less common in the assemblage, and they are more or less likely to have been brought in as raw material and knapped in situ. Various factors may have influenced this pattern. Measures of terrain difficulty and energy expenditure, the raw material quality, and characteristics of the sources are woven together to determine the attractiveness of each source. This is then placed in the context of the geology and geography of the area to distinguish a ‘main’ or core territory from a more extended territory visited during longer trips. The results show the value of taking a geoarchaeological perspective, which sees nature and culture as inextricably intertwined.
Middle Palaeolithic hominids, like all humans before and since, were not separate from nature but instead were intimately involved with it. The environment offers both challenges and opportunities to any human group, which constrain the group’s cultural responses, and at the same time the group’s cultural or behavioural repertoire must constrain their ways of reacting to the environment’s characteristics. The lithic assemblage that we find at any site is thus the expression of this intimate involvement. The specific tool types and proportions of knapping debris, and so on, in an assemblage are not just the reflection of behavioural choices, they result from the characteristics of the resource landscape at the time. In the same way, the proportions of different raw materials present in an assemblage are not the strict product of the distribution of resources in the region of the site: they must also be due to behavioural choices. Resource use starts with acquisition. Which resources were acquired was influenced by many factors (discussed by Wilson 2007a –c), such as abundance, quality, and the size of the ‘packages’ available (Kuhn 1991; Ataman et al. 1992). Availability also includes the ease of extraction of the resource from its enclosing matrix (Ataman et al. 1992; Elston 1992b), and the difficulty of reaching the source area, which may vary depending on the steepness of the terrain around the source (Wilson 2007b), the type and thickness of vegetation, the strength of water courses, and so on. These geographical and geological factors, which Elston (1992a) called extrinsic cost factors, make up the
lithic landscape or lithic terrane (Gould & Saggers 1985; Elston 1992a). They will have varied through time, but some attempt can be made at reconstructing them. Other factors are also important, however, and these may be much more difficult, or impossible, to evaluate: they are the ‘human’ factors. Personal contingencies such as fatigue and the desire for a certain variety of resources must have combined with larger societal or cultural factors such as technological needs, mobility strategies, group organization, and territorial limits to make one resource more or less available or desirable than another in any given situation (Binford 1989; Torrence 1989; Raven 1992; Kuhn 1995, 2004). It is generally agreed (Geneste 1989; Kuhn 1995) that for any group of foragers, subsistence activities are crucial, and other activities, such as making tools, are likely to be organized around the demands of subsistence. Tool materials will thus be acquired ‘along the way’; this is generally called embedded procurement (Binford 1980, 1989; Binford & Stone 1985; but see also Gould & Saggers 1985). Procurement may be expedient (also called ‘encounter’), where raw materials are picked up and used only when needed, but in most cases it probably involved some degree of planning, and curation of the materials and the tools, so that they were carried around and were available when the need arose (Binford 1977; Bamforth 1986; Roebroeks et al. 1988; Ataman et al. 1992; Elston 1992a; Elston et al. 1992). The degree of planning or anticipation will have influenced the procurement
From: Wilson, L. (ed.) Human Interactions with the Geosphere: The Geoarchaeological Perspective. Geological Society, London, Special Publications, 352, 163– 180. DOI: 10.1144/SP352.12 0305-8719/11/$15.00 # The Geological Society of London 2011.
164
L. WILSON
strategy adopted, as will other factors such as the distance of transport (Renfrew 1977; Newman 1994), the topography and the mode of transport (Ericson 1977; Findlow & Bolognese 1982; Reid 1986; Wilson 2007b), the time available (Torrence 1983, 1989; Elston 1992b), the weight of material to carry (Elston 1992b; Metcalfe & Barlow 1992; Beck et al. 2002), and the distribution and ubiquity or otherwise of raw material sources (Geneste 1988; Kuhn 1995; Wilson 2007c). Given all those factors, it is clear that every human group will vary its behaviour from one situation to the next. Still, many attempts have been made to find generalities and develop models that will allow us to gain some understanding of technological and subsistence organization in the past. Kuhn (1991, 1995, 2004), for instance, has worked on the concept of provisioning of individuals (such that each person carries what he or she needs) versus provisioning of places (bringing materials to a home base, and in some cases storing or even caching them there for future use). These concepts feed into models of social and economic organization, such as optimal foraging. In that perspective, foragers try to maximize the benefits of their activities while minimizing costs, by adjusting their mobility, their choices of resource exploitation locations (preferring areas where multiple resources are available), and their settlement locations (Shott 1986; Jeske 1992; Raven 1992; Brantingham 2003, 2006). In central-place foraging (Hodder & Orton 1976), a group will range out from a home base, perhaps using logistical excursions, where part of the group goes farther afield to acquire and bring back certain resources. In this case, the group may show low residential mobility, staying at their home bases for long periods of time, and using their territory in a predictable way (which, as Kuhn (1995) pointed out, promotes a strategy of provisioning of places). In contrast, more mobile groups, with high residential mobility, move through their territory in a sequence of short stays in a series of locations, and are more likely to provision individuals (Kuhn 1995), especially in areas where sources of suitable raw materials are widely scattered. A good example of this is the Middle Palaeolithic site of La Combette in southeastern France, about 25 km south of the Bau de l’Aubesier. Almost no suitable raw material was locally available, and small groups of hunters arrived with ready-made toolkits. After a few days at the site, during which they killed some horses and tanned their hides, they left; presumably on their way back to areas where raw materials were available, because they left the still-usable tools behind (Texier et al. 1996, 1998, 2003). Understanding the behaviour that led to the creation of archaeological assemblages is a challenge,
and it is greatly complicated by the fact that any major deposit at any archaeological site is likely to be a palimpsest of remains left by different groups, and, even if formed by a single group, of activities organized according to a variety of different strategies and contingencies (Bailey 2007). It may be very difficult to sort out the various traces of those different activities. For instance, if there is a large proportion of a certain raw material in an assemblage, does that mean that people went to its source many times, or went there only once and brought back a lot of material? Can a lithic assemblage accurately reflect the full extent of a group’s territory? Probably not, as it will be most influenced by the source that was most recently visited (Ingbar 1994). If a site was occupied for a long time, should we find a lot of local raw material (Wilson 1988), or a more diverse assemblage including a variety of raw materials from relatively distant sources (Wilson 1998), or both? To complicate matters further, just as a site will be a palimpsest of traces left by different activities, it will also give us only part of any group’s suite of activities. The more logistically organized the group was, the more the tasks that needed to be accomplished will have been highly differentiated, and will have been carried out in separate places and at separate times (home base tasks such as tool maintenance, compared with raw material extraction tasks, or with initial butchery tasks, for instance). That degree of logistical organization clearly already existed during the Middle Palaeolithic (Geneste 1985, 1988, 1989; Roebroeks et al. 1988; Fe´blotAugustins 1993; Kuhn 1995). Understanding the ways of life of prehistoric peoples is thus a complex endeavour. The present paper approaches the task from a geoarchaeological perspective, by exploring the middle ground where nature and culture intertwine. Study of the tool assemblage from layer IV of a Middle Palaeolithic rock shelter site in southern France, the Bau de l’Aubesier, shows that raw materials from different geographical areas were used in different ways. This paper suggests why this was so, by taking into account both the environmental context of the site (its surrounding ‘resource landscape’), and the behaviour of the hominid groups that created the assemblage.
The study site and region The Bau de l’Aubesier (hereafter called the Bau) is a large rock shelter located in the Vaucluse region of southern France (Fig. 1). This is a region of limestone, strongly dissected by gorges, hills and valleys due to karst processes and to tectonism associated with the creation of the Alps (to the NE
RAW MATERIAL ECONOMICS
165
Fig. 1. Location of the Bau de l’Aubesier.
of the region). From north to south the major topographic features include the Mont Ventoux (1902 m); the hilly Monts de Vaucluse and the gorges of the Nesque river, within which the site is located; the valley of the Calavon river; and the Luberon mountains. The region is bounded to the east and south by the Durance river, which comes down from the Alps and swings westward south of the Luberon mountains, to join the Rhoˆne at Avignon. Although it is drained by well-established river systems, the region is at present relatively dry, as the karst processes have created abundant fissures, solution channels, sinkholes and other ways for water to disappear underground. The Nesque River itself disappears at the entrance of the gorge, so the bottom of the gorge is mainly dry except during times of flood owing to heavy rainfall or snowmelt. The water does reappear at various small springs, and above all in the SW of the region at Fontaine de Vaucluse, source of the Sorgue river. The Bau contains sedimentary deposits c. 13 m thick, within which several archaeological layers have been identified. The site was first dug in the earliest part of the twentieth century (Moulin 1903, 1904), and studied briefly in the mid-1960s (de Lumley 1971), but serious scientific excavation was only undertaken starting in 1987, by a multidisciplinary international team led by Serge Lebel of the Universite´ du Que´bec a` Montre´al, Canada (Lebel 2000a; Wilson 2007a). The site has produced
very abundant lithic and faunal assemblages, as well as several Neanderthal and pre-Neanderthal teeth and a jaw fragment (Trinkaus et al. 2000; Lebel et al. 2001; Lebel & Trinkaus 2002; Fernandez 2006; Wilson 2007a). Located in the upper part of the site, layer IV is a highly burnt, very black layer containing innumerable tiny burnt bone and charcoal fragments, as well as an abundant but fragmentary faunal assemblage and many thousands of lithic pieces, including tools, flakes, debris, and so on. It has also yielded five Neanderthal teeth (Trinkaus et al. 2000). The layer varies in thickness from a few centimetres to more than 20 cm, and has been excavated over more than 40 m2. It clearly continues into unexcavated areas, so it represents a substantial surface area. Based on the faunal assemblage (Fernandez 2006) and on U/Th and electron spin resonance (ESR) dating (Blackwell et al. 2001), the layer dates to Oxygen Isotope Stage 5d, or roughly 110 000 BP, although it may have taken anywhere from years to thousands of years to accumulate. According to Fernandez (2006), the identifiable fragments and teeth in the faunal assemblage include Bos (aurochs, 31.8%), Capra (goat, 19.2%), Equus (horse, 19.1%), and Cervus (red deer, 17.2%), with smaller quantities of Dama (fallow deer, 6.9%) and Capreolus (roe deer, 3.8%), and traces of Megaloceros (Irish elk, 0.8%), Sus (wild boar, 0.8%), and Rupicapra (chamois, 0.4%). Despite their fragmentary nature,
166
L. WILSON
Fernandez (2006) was able to determine that they are due to human, not carnivore, activity, and that they result from selective hunting strategies and not from scavenging. The proportions of the different species therefore reflect some combination of their abundance in the environment and human hunting choices. The faunal remains attest to a fairly rigorous, cold and dry climate, with a mainly open landscape but some wooded areas (where wild boar and the deer species would be more likely to be found), perhaps in more protected or better watered areas near springs, and on south-facing slopes. Argant (2000) found that the pollen was mainly steppic, but with a surprising variety of trees (perhaps a mixed sample?), and that the climate was probably cold and fairly dry. The geoarchaeological study of the lithic materials making up the stone tool assemblage from the Bau has broadened into a regional-scale investigation, the Vaucluse raw material project, which has so far involved 20 years of field seasons and considerable laboratory and analytical work (Wilson 2007a). The assemblage from the Bau includes almost 40 000 pieces, categorized into 46 types, almost entirely flint. There is also one piece of volcanic rock (porphyritic andesite), and a few pieces of quartzite, but as none of these were found in layer IV, they will not be discussed further here. The assemblage from layer IV is the largest at the Bau, with more than 15 000 pieces catalogued by raw material type. However, this number includes many pieces that are either completely covered by a secondary patina (as a result of weathering within the site), or are so highly burnt that their original rock type is obscured. Eliminating them, and a few pieces for which the source of the rock has not yet been identified, leaves us with a total of 4481 pieces for which both the flint variety and the geological source are known. This is still the largest number of pieces from a single layer at the Bau, which is why this assemblage was chosen for this study. The tool assemblage is Typical Mousterian, rich in sidescrapers (Lebel 2000b), as is shown in Table 1.
Sources of the rock types The Vaucluse raw material project has included examination of archaeological material from several Middle Palaeolithic sites, primarily the Bau and the younger site of La Combette, which has several layers dating to between about 50 000 and 60 000 BP (Texier et al. 1996, 1998, 2003), and a long-term intensive and extensive prospecting campaign, which has resulted in the location of 352 potential sources of flint or other useful lithic raw materials. The region has now been thoroughly
Table 1. Summary data on major tool type classes as per cent of identified pieces, summarized from Lebel (2000b) Typology
Layer IV
Levallois cores Non-Levallois cores Levallois flakes Blades and bladelets Levallois points Mousterian points Simple sidescrapers Other sidescrapers Endscrapers Natural-backed knives Other Late Palaeolithic types Denticulates Notches
4.3 13.4 28.9 17.4 3.0 0.6 15.2 6.5 2.7 2.6 1.3 2.3 2.0
Number of pieces
2457
explored, and although it is always possible that a few more sources will be found, the basic pattern (shown in Fig. 2) is clear. There are areas where flint sources are abundant, and areas where they are scarce or absent. Sources include, in some cases, specific outcrops, and in some cases secondary (alluvial or colluvial) deposits. Samples collected from each source have been described, catalogued, and analysed in a variety of ways. Details are available at http://pizza.unbsj. ca/~lwilson/. Photographs of specimens will be added eventually, as time and funding allow. All sources have been described first of all in terms of macroscopic criteria, including the colour or colours of the raw materials available at the source, texture (grain size, homogeneity, inclusions), the characteristics of the cortex (if applicable), and any visible fossils. In most cases, samples from the source have been thin-sectioned and studied under the polarizing microscope, and the minerals, granulometry, microstructures (zoning, bands, etc.) and any microfossils recorded. In a smaller number of cases samples have been analysed using geochemical methods and the scanning electron microscope, or for their oxygen isotope ratios. So far a combination of microfossils and macroscopic characteristics has proven to be the most useful for distinguishing the provenance of the samples (Wilson 2007a). Most of the flint is Early Cretaceous (Bedoulian stage) in age and occurs as nodules in fine-grained bioclastic limestones that formed as a shallow marine platform (Rouire 1975; Masse 1993). The second most common flint sources are of Oligocene age, where flint nodules or beds are found in finegrained limestones that formed in lakes or lagoons (Rouire 1975). Some of these were hypersaline
RAW MATERIAL ECONOMICS
167
Fig. 2. Potential raw material sources in the Vaucluse and surrounding areas, categorized as Cretaceous, Tertiary (mainly Oligocene and Eocene), or Mixed (including secondary deposits).
whereas others were freshwater features. In either case, their fossil content is very different from that of the normal marine setting where the Cretaceous limestones formed. The Oligocene rocks tend to contain some combination of gastropods, ostracods and/or charophytes (a type of freshwater algae), whereas the Bedoulian ones have a typical marine assemblage of various foraminifera, sponge spicules, red and green algae, and occasional bryophytes, coral fragments, and so on. There are also a few sources that are of other ages, primarily including other stages of the Cretaceous, and Eocene deposits. All told, the list of possible sources includes 202 Cretaceous, 74 Tertiary (Oligocene and Eocene), and 76 secondary or mixed deposits (Fig. 2). The archaeological lithic assemblage from the Bau de l’Aubesier has been catalogued into 46 types of raw materials, mainly varieties of flint, based on visual criteria such as colour, cortex, structure and fossils. The types were defined as narrowly as possible, in the awareness that many of them would turn out to be variations of each other, rather than coming from distinct sources. At least one representative of each type was selected for a petrographic reference series. In a few cases of very rare types, the pieces were too small for thin sectioning, but most have been sectioned and analysed under the polarizing microscope, so that they can be compared with the samples from the surrounding environment. To date, of the 46 lithic types, 18 have been located to precise sources, based on macroscopic criteria confirmed by
microscopic structure or fossils; nine have been more broadly located to a source area, based on macroscopic criteria confirmed by microfossils; five have been identified to age of formation, based on fossil content (of these, four are Tertiary and one is Cretaceous); seven types are not identifiable because they are too burnt or patinated; and the sources of seven types are still unidentified. Because in some cases one source or source area can provide more than one lithic type, the 32 types for which we have some indication of provenance have been further grouped into a total of 15 sources and source areas, described in Appendix 1. Twelve of those sources or source areas were used for the assemblage from layer IV. The discussion below will therefore refer only to those 12 sources or source areas, not to the overall list of 46 lithic types, nor to all potential sources. Table 2 lists the sources and source areas used, and the numbers of pieces from each in the layer IV assemblage. Figure 3 shows the locations of the sources or source areas, superimposed on the map of all potential sources. Some of the source areas delimited are much larger than others. This is because in some cases a specific outcrop (such as near Roussillon) has been identified as a source, whereas in others several outcrops and secondary deposits in an area may have been used, such as at Murs. The very large area labelled ‘Tertiary’ in Figure 3 includes many potential sources of Oligocene age. This does not mean that all of those sources were used. Rather, this is meant to show
168
L. WILSON
Table 2. Sources and source areas used for layer IV Source area name Local Sault St. Trinit St. Jean Gue´rin Tertiary Roussillon Bedoulian Les Sautarels Bezaure Murs Treille
Direction from Bau
Distance (km)
Number of small pieces
Number of large pieces
S NE NE E SE S S SW SW SW SW W
0.6 9.8 13.9 4.9 31.9 14.5 18.2 4.5 10.8 11.3 12.4 5.6
120 110 7 5 2 1 1 1140 39 26 896 9
16 127 17 2 1 2 2 1143 20 18 761 16
that there are many potential sources of Oligocene flint in that overall area, any one or more of which may have been used to provide the three lithic pieces from layer IV that are attributed to that area. The attribution is based on overall appearance and on fossil content, but it has not yet been possible to determine exactly which source may have been used. Table 2 also gives the approximate direction from the Bau to the source, and the distance in kilometres from the Bau to the source or source area. The distance is not that from the Bau to the closest border or to the centre of a source area: it is the distance to a single source that has been chosen to represent the source area in the calculations below.
The representative source for a source area is the closest source to the Bau that can provide the greatest quantity of material that most closely resembles the material found at the Bau. This is not meant to imply that raw material could not have been obtained at other outcrops or deposits within the source area.
Results: differences in use of the rock types Table 1 lists numbers of pieces grouped into various typological categories (cores, sidescrapers, etc.), according to Lebel (2000b). Unfortunately, Lebel
Fig. 3. Raw material sources and source areas represented in the assemblage from layer IV.
RAW MATERIAL ECONOMICS
did not provide catalogue numbers for the pieces on his list, so it is not at present possible to correlate them with the raw material catalogue. Determination of the proportions of raw materials used for specific tool types will therefore have to wait for a future publication. However, the petrographic catalogue does make it possible to examine differences in use of raw materials in two ways: the total numbers of pieces of each raw material found in layer IV, and the proportions of each that represent individually coordinated pieces, which (as will be explained below) can allow us to draw some conclusions as to the stage or stages of the chaıˆne ope´ratoire present for materials from each source area. Tables 2 and 3 show that there are major differences in the numbers of pieces of rocks from different areas, ranging from a total of 2283 pieces of Bedoulian flint from the area to the SW of the site, down to three pieces from several other sources. The most used sources are located to the SW of the site; in fact, all SW sources together account for a total of 4043 pieces, or 90% of the entire assemblage. The next most represented area is the NE, with 261 pieces (6% of the total). Sources in other directions are even less used, and it is worth noting that some sources that were used for the assemblages from other layers at the Bau are not represented at all in the assemblage from layer IV. Those sources are located farther to the west, south and NE than the sources used in layer IV, suggesting that layer IV’s assemblage results from concentrated exploitation of a core territory. Tables 2 and 3 also show that the proportions of small versus large pieces from each area vary widely. The large pieces are those that were individually coordinated at the time of excavation. When we excavated layer IV, the sheer number of small lithic and faunal pieces made it impossible to individually coordinate every single piece before
169
removing it from the site. We therefore coordinated only the larger ones, generally those more than about 3– 5 cm in length. All smaller pieces were catalogued by subsections of square metres: everything from a certain area was bagged, and screened and sorted out later in the labratory. All lithic pieces from that bag were given the same catalogue number. Therefore, where the raw material type catalogue shows multiple pieces with one artefact number, they are relatively small pieces. Where the artefact number represents a single piece of rock, it is a larger piece. Small pieces are most likely to result from in situ flint knapping or retouch of tools. Larger pieces can also be byproducts of flint knapping, but may also be utilizable flakes, tools or cores. The basic concept of the chaıˆne ope´ratoire describes the process that raw materials go through as they become tools. In very simple terms, the raw material is obtained at a source and then knapped, either there or somewhere else, with the products possibly further transported before being used as tools; those tools may then be retouched and modified once or several times before ultimately being abandoned. Knapping results in the rapid accumulation of many small pieces, especially in the earliest stages of preparation of cores. Therefore, if rocks were brought in to the Bau as nodules or cores and knapped on site, we should find the by-products of knapping: many small flakes or chips, and perhaps a few tools and cores. An assemblage dominated by small pieces would therefore include the early part of the chaıˆne ope´ratoire. Another consideration is the quantity of material from a source: large numbers of pieces (small and large) result from importation of raw material followed by knapping at the site. On the other hand, if we find mainly larger flakes and tools, we are looking at the later part of the chaıˆne ope´ratoire: a working toolkit
Table 3. Proportions of small versus large pieces from each source area Source area name Local Sault St. Trinit St. Jean Gue´rin Tertiary Roussillon Bedoulian Les Sautarels Bezaure Murs Treille
Direction from Bau
Distance (km)
Total number
Proportion of small pieces
Proportion of large pieces
Early or late stages
S NE NE E SE S S SW SW SW SW W
0.6 9.8 13.9 4.9 31.9 14.5 18.2 4.5 10.8 11.3 12.4 5.6
136 237 24 7 3 3 3 2283 59 44 1657 25
0.88 0.46 0.29 0.71 0.67 0.33 0.33 0.50 0.66 0.59 0.54 0.36
0.12 0.54 0.71 0.29 0.33 0.67 0.67 0.50 0.34 0.41 0.46 0.64
Early Late Late Early Early Late Late c. Early Early Early Early Late
170
L. WILSON
has been transported in to the site, and the knapping was done elsewhere. This is especially likely if we find small numbers of pieces from a given source. Tables 2 and 3 therefore demonstrate that different source areas are represented by different stages of the chaıˆne ope´ratoire. The numbers of small and large pieces for each source area are given in Table 2, and the proportions of each are listed in Table 3, along with an attribution of each source to either ‘early stages’ or ‘late stages’ of the chaıˆne ope´ratoire. The cut-off point for the attributions is at 0.50, where there are equal numbers of small and large pieces. Three of the source areas are represented by only three pieces each, and one by seven, so for them the proportions of small versus large are probably more accidental than significant. If we look at the others, however, we see that there is an interesting pattern: sources to the south and SW have higher proportions of small pieces: they were transported in and knapped on site. On the other hand, sources to the west and NE have higher proportions of large pieces: they tend to have been knapped away from the site. The source with the highest proportion of on-site knapping is the local one, located a few hundred metres to the south of the site, on the plateau above the gorge. Only 12% of the 136 pieces in this rock type were large enough to be individually coordinated. The highest proportion of large pieces is found in the 24 pieces of material from St. Trinit, 13.9 km to the NE of the Bau, of which 71% were individually coordinated. The relative distances of these areas from the site can not be the reason for these different proportions, however, as the sources around the village of Murs provided 1657 pieces, of which only 46% were big enough to be individually coordinated. Murs, therefore, which is more than 12 km to the SW of the Bau, was the source of a considerable amount of material that was brought to and knapped at the site.
Discussion The different proportions of raw materials in the layer IV assemblage result from the interplay of prehistoric behaviours and the characteristics of the resource landscape. How far away the sources were, how easy or difficult they were to reach, how good the raw material was, what other resources were available there and along the way, what each group considered its territory to be, and other human factors, all combined to influence how resources were acquired, transported, used and abandoned. The first factor to look at is the distance to each source, and what that can tell us about territory size. There is a longstanding debate in prehistoric archaeology concerning prehistoric mobility, and
the limits of what can be considered ‘local’ versus ‘exotic’ (e.g. Metcalfe & Barlow 1992; Newman 1994; Gamble 1999; Beck et al. 2002; Brantingham 2003, 2006). The local area is generally considered to be one that the prehistoric people used on a daily or habitual basis, and corresponds to what Gamble (1999) has termed the ‘landscape of habit’. Exotic raw materials would have been brought in as a result of longer hunting trips, travel in from other territories, or trade. For the people who created layer IV at the Bau, 12.4 km to Murs seems to have been local (much used), whereas 14.5 km and farther seems to have been ‘beyond local’ (scarce pieces). Setting a limit in kilometres is a misleading exercise, however, because the limits between such areas must surely have depended on criteria other than just distance. In fact, although material from the closest source was the most likely to have been knapped on site, and the three least represented sources are the three most distant ones, amongst the other sources there is no simple pattern of the proportions of small pieces decreasing with distance, nor in fact any statistically significant correlation between distance and numbers of pieces, small or otherwise. In addition, distance should not really be measured ‘as the crow flies’, or as a straight line marked on a map. What matters to a person walking through any landscape is not so much how far they have gone, but how much effort they have expended, and that depends on both distance and topography. Walking up a steep hill requires more caloric expenditure than walking the same distance on flat land. Wilson (2007b) presented a way of measuring the caloric expenditure required to walk across varying topography, based on slope angles. That paper reported on the use of that method to determine the difficulty of routes from raw material sources to two of the archaeological sites in the Vaucluse region, and looked at whether the difficulty of each route may have had an influence on the proportions of use of each source for the archaeological assemblages. The results indicated that path difficulty may well have had some influence (Wilson 2007b). In current and future work, we intend to investigate alternative ways of understanding terrain, such as Tobler’s (1993) hiking algorithm (which uses slope to calculate walking velocity, thereby converting distance to time– distance, or the ‘time path’ to the destination), the methods of determining anisotropic cost-of-passage described by Bell & Lock (2000) and Conolly & Lake (2006), and the accessibility index concept as described by Llobera (2000). These will be compared and contrasted to find the most effective technique. Table 4 gives the route difficulty values for the sources considered here, in kilocalories (Cal) per kilometre, and according to broad categories of
RAW MATERIAL ECONOMICS
171
Table 4. Route difficulty from each source (or representative source in source area) to the Bau de l’Aubesier Source area name Local Sault St. Trinit St. Jean Gue´rin Tertiary Roussillon Bedoulian Les Sautarels Bezaure Murs Treille
Direction from Bau
Distance (km)
Cal km21
Difficulty
Number of pieces
Early or late stages
S NE NE E SE S S SW SW SW SW W
0.6 9.8 13.9 4.9 31.9 14.5 18.2 4.5 10.8 11.3 12.4 5.6
62.87 68.33 73.98 86.46 79.56 92.00 81.50 93.81 83.77 81.20 87.93 129.31
Easy Easy Easy Medium Easy Medium Easy Medium Easy Easy Medium Hard
136 237 24 7 3 3 3 2283 39 44 1657 25
Early Late Late Early Early Late Late c. Early Early Early Early Late
Difficulty is broadly categorized based on the complete range of routes calculated, from all known sources to the Bau de l’Aubesier. The easiest third is ‘easy’ (,85.02 Cal km21), the middle third is ‘medium’ (85.02 –107.17 Cal km21), and the most difficult third is ‘hard’ (.107.17 Cal km21).
‘easy’, ‘medium’, and ‘hard’. Those categories were based on the total range of difficulties determined for all of the routes from over 300 potential sources to the Bau: the overall range was evenly divided into three subsets, with the easiest third of the values (,85.02 Cal km21) categorized as ‘easy’, the middle third as ‘medium’, and the most difficult third (.107.17 Cal km21) as ‘hard’. (For comparison, walking on flat land at 3 km h21 requires roughly 50 Cal km21 (Wilson 2007b), so the ‘hard’ category really does require a substantial caloric expenditure.) The results show that walking from the western source area to the Bau is considerably more difficult than travel in any other direction, because it necessitates crossing the gorges of the Nesque river, which are between 100 and 300 m deep in most places. This may well explain why a comparatively small amount of material from this relatively nearby area has been found at the site. The most abundantly used areas are not necessarily easy to reach (Murs and Bedoulian both being categorized as ‘medium’), but, interestingly, the terrain between the Bau and the most distant sources used is relatively easy to cross. If distance and terrain difficulty explain only part of the pattern, what other factors should we consider? Wilson (2007c) laid out the next approach, which uses a cost –benefit equation to determine the attractiveness of each source; that is, how likely would someone be to choose to go to that source, rather than another? The benefits of the source are the raw material quality (suitability for knapping), the extent or size of the source, and the size of the raw material pieces (nodules, etc.) available at the source. The costs are the terrain difficulty, the difficulty of extraction of raw material
from the source, and the scarcity of usable material at the source, which influences both abundance and time spent searching. The attractiveness equation, and the values used to calculate the attractiveness of the sources (or representative sources for the larger source areas) used for layer IV, are given in Appendix 1. Table 5 lists the attractiveness values obtained for the sources. Higher values mean that the source is more attractive. The most attractive source is the one that is representative of the Bedoulian outcrops to the SW of the Bau, and that is also the source area that was used the most for layer IV. Its attractiveness is due to its abundant nodules of good quality flint. The second and third most attractive sources provided the third and second most abundant rock types. Raw material use is not solely dependent upon geographical or geological factors: it also depends upon human choices. The attractiveness equation allows us to lump together the geological factors, and in a sense eliminate them, to make clear where the human factors must be more or less important. It also allows us to map out the area within which each source would be used: its ‘area of influence’ (Wilson 2007c). An attractive source will have a larger area of influence than a less attractive source: people will be inclined to go to it from a greater distance, because it is worth the trip. Sources in areas where they are widely dispersed will also have larger areas of influence than those in areas where there are many sources close to each other. On the other hand, people would be unlikely to go to unattractive sources, and would use them only if they were already in that vicinity for some other reason. The source at Roussillon (Fig. 3) provides a good example of this. Its attractiveness value is
172
L. WILSON
Table 5. Attractiveness values for each source (or representative source in source area) Source area name
Direction from Bau
Local Sault St. Trinit St. Jean Gue´rin Tertiary Roussillon Bedoulian Les Sautarels Bezaure Murs Treille
S NE NE E SE S S SW SW SW SW W
Distance (km)
Attractiveness value
Number of pieces
0.6 9.8 13.9 4.9 31.9 14.5 18.2 4.5 10.8 11.3 12.4 5.6
167.01 224.40 63.36 185.06 62.85 114.13 3.07 351.77 69.64 92.36 219.77 69.60
136 237 24 7 3 3 3 2283 39 44 1657 25
only just over three, because it provides only scarce, very small pieces of mediocre quality, and yet it was used. There are three pieces from this source in the assemblage from layer IV, and a few more in other layers. As the rock itself cannot have been the attraction, there must have been something else that was bringing people to that area, and while they were there, if they did see a suitable rock just when they needed one, they picked it up. An overall explanation of lithic use must, therefore, take into account the distribution and characteristics of potential raw material sources, the topography of the region (including the locations of natural obstacles and barriers, and natural pathways or corridors that encourage travel), and the distribution of other resources such as fauna, flora, water, and shelter. In addition, the interaction of human behaviour with this environmental set-up has to be accounted for, as much as possible. For instance, humans may choose to exploit resources that are more easily accessible because they are located along natural corridors of travel, and neglect other resources that are harder to reach. As a consequence, use of those natural corridors may became entrenched in human behaviour, so that they become paths that people will be more and more inclined to follow, as behaviour becomes habit, reinforcing itself through repetition over time. The ‘comfort of walking’ (Helbing et al. 1997) along such trails will increase as they become more well-trodden. This will increase the perceived attractiveness of the resources that are commonly used (as long as they remain sufficiently abundant), and decrease the likelihood that other resources will be exploited. From this perspective, it is necessary to study an area not only as a landscape but also as a resource landscape that changes through time, naturally, anthropogenically,
Early or late stages Early Late Late Early Early Late Late c. Early Early Early Early Late
and in the way that humans perceive it (e.g. Church et al. 2000). The size of the faunal and lithic assemblages from layer IV, and the evidence for importation of raw material for in situ knapping, suggests that the deposits result from a palimpsest of occupations by separate groups, at least some of which demonstrated low residential mobility (long length of stay) and a strategy of provisioning of place (Kuhn 1995). The numbers of pieces of different rock types show that the area to the SW of the Bau was by far the predominant source of raw materials for the assemblage found in layer IV. Over 90% of the pieces come from the local source, the Bedoulian outcrops, Les Sautarels, Bezaure, or the Murs area, and much of that material was probably knapped at the Bau. Obviously, then, there was a lot of travel to and from the SW, and it was considered close enough that bringing raw material to the site for knapping was worth while. This is therefore the primary territory exploited by the Neanderthals who created the layer IV assemblage, yet the Bau is not in the centre of it, it is in the NE corner. We can therefore ask ourselves why they used this territory, and also, why did they occupy the Bau? First of all, why the Bau? Because it has many attractions. It is a particularly large rock shelter near a water source, the Nesque river. Nowadays water in that territory is scarce; other than at the Nesque, water can be found only at a few springs. During a dry and cold period in the past, that may also have been true. The Bau is, furthermore, in an area of highly diverse topography and therefore diverse habitats for animals and plants, ranging from the protected, possibly watered and wooded areas within the gorge (where wild boar and chamois may have been hunted, and where fuelwood
RAW MATERIAL ECONOMICS
may also have been found), to the plateau and hillsides around the gorge, where herds of large mammals roamed. It is also near a source of useful lithic raw material. There are other rock shelters in the area, but the Bau was certainly prime real estate for a group of hunter–gatherers. A combination of factors may account for the fact that the area to the SW was the main territory exploited. First of all, geologically and geographically, that area is similar to the area around the Bau: rugged terrain with diverse habitats, some protected, some open; occasional springs for water; and numerous lithic sources. The area corresponds to the southwesterly extension of the Cretaceous limestones that make up the Monts de Vaucluse (with occasional pockets of Oligocene rocks as a result of downfaulting of small basins). Therefore the range of flora, fauna, rocks and terrain are similar throughout the territory, and presumably that combination of characteristics suited the Neanderthals: it provided the diversity of resources that they needed and wanted. Table 6 shows the percentages of identified faunal remains grouped into their likely habitat categories. The smallest, but still substantial (19.6%) proportion comes from animals that are likely to have been hunted in the gorge around the site itself. A further 28.7% of the remains are from animals that preferred wooded habitats, such as might have been found in the gorge or on protected south-facing slopes in this highly dissected landscape. The majority, 50.9%, are from large herbivores that lived in herds the open areas, such as the plateaux and plains. There is plenty of room for such herds in the areas to the SW and to the NE of the Bau. Furthermore, the area around Murs is one of the best for finding really good quality, abundant, raw materials, so it was worth going there, and also using the territory between Murs and the Bau. In fact, although the route from Murs to the Bau is categorized in Table 4 as being of medium difficulty, it is located along a fairly straightforward
173
natural corridor, surrounded by areas that are more difficult to cross. From the Bau to Murs, the route goes through the ‘local’ source area, and proceeds across a slightly hilly area to the SW Bedoulian sources. From there, the easiest thing is to follow a narrow valley that leads to the vicinity of Les Sautarels and Bezaure, avoiding some very steep slopes and cliffs. From those two sources, the Murs area is just across a low ridge. Taken in the other direction, Les Sautarels, Bezaure, the Bedoulian outcrops and the local source are all ‘on the way home’ from Murs. Thus it was relatively easy to move around in this territory, and we find the reflection of that in the lithic assemblage, with abundant pieces and evidence of knapping on site. The next most abundantly used sources were to the NE, in the vicinity of the present town of Sault. That area is also easy to reach: once out of the gorge the route crosses a plateau, comes down into a broad valley and follows that until it continues up onto the plateau behind Sault. The area was reasonably abundantly used, but the difference is that there is a higher proportion of large pieces in this raw material, suggesting that less of it was knapped at the Bau. Continuing in the northeasterly direction across the plateau, we come to St. Trinit, and the material from there has the highest proportion of large pieces, therefore probably representing the later stages of the chaıˆne ope´ratoire. Pieces from this area were more likely to be already part of a working toolkit, and less likely to be unknapped raw material when they arrived at the Bau. It seems that the northeastern area was not part of normal ‘day trip’ territory, but rather was covered on the outward leg of longer trips. Material collected there was less likely to be carried to the Bau in nodule or core form because the Bau was in fact farther away than it seems: there were other places to go to first. The source to the east was nearby but not much used: only seven pieces, mainly very small. This
Table 6. Faunal assemblages by main habitat type Habitat type
Faunal type
% in layer IV
Gorge Gorge Total gorge Woods Woods Woods Woods Total woods Plains Plains Total plains
Capra Rupicapra
19.2 0.4 19.6 3.8 17.2 6.9 0.8 28.7 31.8 19.1 50.9
Capreolus Cervus Dama Sus Bos Equus sp.
174
L. WILSON
may represent only one piece brought to the Bau and knapped or retouched there. The source is well within day trip range, but seems not to have been used much at all for lithic material. The sources to the west, as we saw above, are hard to reach because they are across the gorge. That material has a low proportion of small pieces, suggesting that it represents the later stages of the chaıˆne ope´ratoire, and may, again, have been obtained during longer trips, rather than being exploited directly from the Bau. As for the more outlying areas, the geology and geography of those areas are different from those of the Monts de Vaucluse, and clearly they were not part of the habitual territory of the groups when they were based at the Bau. The very few pieces of these raw materials must be evidence of longer trips, or of travel in from other territories. Finally, there is the very small, unattractive source at Roussillon. The geology of Roussillon is very different from that of the Monts de Vaucluse. It consists of thick deposits of ochres, rather than massive limestones, and the raw material is not flint, but rather consists of fragments of a silicified crust from the upper layers of the ochre deposits. The soils in this area are different (more acidic) and the plants that grow on them are also different (more pine, less oak, at least at present). Different plants mean different habitats, and presumably different animals too. This area may therefore have provided some unusual floral or faunal resource that the Neanderthals appreciated. There is also a slight chance that they were exploiting the ochre itself, although we have found only very small, occasional, traces of that at the Bau, and none in layer IV (where in any case it would be invisible or destroyed in the black, highly burnt sediments).
Conclusions The main goal of archaeology is to understand past human behaviour. From the geoarchaeological perspective, understanding past behaviour requires understanding past environments, because the material remains that we find and study are the reflection of the interaction of two interconnected, not separate, factors: nature and culture. Humans do not now and never have lived entirely within some abstract cultural world, nor, for many millennia, has there been any such thing as a human-less nature. Our real world is found in the middle ground between those two abstract worlds. This paper demonstrates this through the examination of one archaeological assemblage from one Middle Palaeolithic site, placed in its environmental context. In reconstructing past human lifestyles, resources should be defined and qualified in terms of human use. A rock is not just a rock: it is more or less suitable as a raw material for tool-making. The slope of
the terrain is not just an angle: it requires more or less energy for humans to walk up or down it. Plants are not just present or absent, nor are they just potential food sources: they are also sources of fuel, sources of raw material for tools, and, as food sources for animals, a controlling factor in the distribution of fauna. Furthermore, it is important to pay attention to scale. People live on a relatively small scale of landscape. Water may be obtained from a spring that measures a few metres across, rocks from an outcrop that is not much bigger. Adding in the complexity of human behaviour makes the job even more difficult. Still, it seems to be worth the effort. As Church et al. (2000, p. 149) noted: ‘No one is denying the complexity of the task, but that is the challenge, not an excuse.’ Layer IV from the Bau de l’Aubesier provided a large assemblage of lithic material, along with faunal and other remains that allow us to date the layer and determine the climate and landscape type at the time. We know that over the course of the accumulation of these deposits, the surrounding area was mainly fairly open, with wooded pockets, and the climate was generally dry and cold. These factors naturally influenced the types of animals that were available to be hunted, which of course influenced human subsistence. Beyond climate and biota, however, the landscape itself needs to be seen as a participant in fundamentally shaping human behaviour, not as a blank sheet upon which activities happened. Resources are not all available in the same place or at the same time: it takes effort and planning to obtain them. Evaluating factors such as the quality of the raw materials, the topography and therefore energy expenditure required to move around in this landscape, the likely locations of other desirable resources, and the mobility patterns that contributed to the creation of the assemblage gives us a much more nuanced and wide-ranging view of those subsistence and resource-use strategies. All in all, then, the Bau de l’Aubesier shows us Middle Palaeolithic groups that were mobile, well organized, well aware of the resources in their territory, and well able to exploit them in a variety of ways. While occupying the Bau, they concentrated their efforts primarily along a SW –NE axis, corresponding to the main geographical orientation of the topography, with their core territory extending about 12 or 13 km to the SW. They also left us a few traces of a larger territory, however, in the form of small quantities of lithic material from sources outside that main area, some from more than 30 km away. Thus we have evidence of strategies of low residential mobility, provisioning of place, and concentration upon exploitation of resources from a core area, as well as of a logistical
RAW MATERIAL ECONOMICS
strategy involving longer excursions into a broader territory extending farther and in other directions. This is a pattern that is not unusual in the Middle Palaeolithic, although in every specific case the extent of the territories must be conditioned partly by the distribution of resources within them. Geneste (1985, 1988, 1989), for instance, found that at Mousterian sites in the Dordogne, in western France, the vast majority of lithic material always came from within about 5 km of the site, and that sources more than 30 km away were commonly represented but in small numbers, with the artefacts consisting of retouched (and much used) tools. Geneste’s (1989) conclusion is that the groups that created the Mousterian deposits in the Dordogne were very mobile, and that their mobility was complex and varied, and no doubt adapted to particular economic situations. They were capable of selecting the best strategies to suit their needs, within their environmental settings. The same is true in the Vaucluse. The lithic pieces that make up the layer IV assemblage thus result from the strategies used to obtain and use the resources that the landscape offered, to meet human needs for subsistence and survival. In keeping with the theme of this book, can we learn any broader lessons from this? Neanderthals were greatly influenced by their natural environment, as are we. Neanderthals are generally not given much credit for intelligence, but ‘even’ Neanderthals were flexible in their strategies, using apparently similar resources (flint from two different places) in somewhat different ways, depending on other circumstances such as what other resources were available in the vicinity, or where they wanted to go next. These results make it clear that behavioural flexibility with respect to interactions with external (environmental) resources and challenges is a deeply human trait. The geoarchaeological perspective thus makes it possible for us to better perceive the evolutionary depth of that trait. Humans have evolved by working out new ways of dealing with nature’s offerings as they have changed, and as we have changed them, through time. It is often said that culture mitigates environmental constraints, and that as we have evolved ever more sophisticated cultural adaptations, we have become more and more removed from nature. I doubt this. If anything, by pushing the immediate constraints away, we may be increasing their consequences. For instance, the local constraint on where we can live has been pretty much removed: people can spend the winter in Antarctica, if they have the resources. However, the consequence of our methods of removing those local constraints turns out to be global climate change. We are not Neanderthals and we are not living in the Palaeolithic, but we are still just as intimately
175
intertwined with our natural environment: we shape it and it shapes us. The dynamism and the ever-increasing consequences of that interaction require that we draw ever more strongly upon our human trait of behavioural flexibility. This paper is the result of many years of work and many discussions with colleagues, students, friends and family: too many people to name. Thanks are due to all of them, and particularly to S. Lebel, P.-J. Texier, N. Moore, C. Browne, G. Rapp, to the University of New Brunswick, and to the Social Sciences and Humanities Research Council of Canada.
Appendix 1: Raw material source areas The following were used for the lithic assemblage from layer IV. The approximate locations of these source areas are shown in Figure 3.
Local This source extends to the south and SW of the Bau de l’Aubesier, on the plateau above the site and on the same side of the gorge. Abundant good quality brown to grey flint of Early Cretaceous (Bedoulian stage) age is found here, in nodules that generally range between 20 and 40 cm in diameter. The source is most abundant within about 1 km of the site. Three lithic types are attributed to this source. One is the local Bedoulian (Early Cretaceous) biomicritic limestone; flakes of this occasionally show up in the site assemblage. The others are flint varieties, attributed on the basis of colour, texture and, where present, cortex colour, which differentiate them from the Bedoulian flints available farther to the east (area of St. Jean de Sault) and SW (the ‘Bedoulian’ area described below). The microfossil content of the type specimens corroborate a Bedoulian age, consisting of fossils such as benthic foraminifera, typical of the marine platform environment that existed in the area at the time (Rouire 1975; Masse 1993). The fossil assemblage is, however, not sufficient to distinguish these flints from others of Bedoulian age in this region.
Sault Approximately 10 km NE of the Bau de l’Aubesier, on the plateau behind the town of Sault, are very abundant nodules of good quality opaque brown or blue flint, usually with a grey cortex. These are of Bedoulian age, and primarily contain sponge spicules and calcareous algae, with some other marine fossils. Specimens of one lithic type from the Bau de l’Aubesier are assigned to this source area on the basis of appearance (good quality, opaque, brown or blue colour), as the fossil content is not distinctive.
St. Trinit Roughly 14 km to the NE of the Bau de l’Aubesier, in the vicinity of the village of St. Trinit, there are more outcrops
176
L. WILSON
of the nearly ubiquitous Bedoulian flint, source of one raw material type. In this case, the flint has a characteristic soft appearance owing to a higher than usual percentage of very finely disseminated clays, calcite and/or iron. Its fossil assemblage includes sponge spicules and foraminifera, but is not distinctive.
layer that formed long ago at the surface of the ochres. Under the microscope, this rock resembles a ferruginous quartzose sandstone with silica cement, not a flint or chert. It is highly distinctive because of the presence of iron-rich spots, giving it a red-spotted appearance (sometimes spots of other colours are also present). One lithic type is attributed to this source.
St. Jean de Sault Five kilometres to the east of the Bau de l’Aubesier, around the village of St. Jean de Sault, pale grey coloured Bedoulian flint is abundant, but is coarse textured and of mediocre quality. It is particularly rich in marine microfossils, including a large variety of benthic and possibly planktonic foraminifera, red and green algae, sponge spicules, shell fragments, and so on. The colour and texture are the basis for assigning three lithic types from the Bau de l’Aubesier to this source area; the fossil suite in the type specimens corroborates this identification.
Gue´rin The area marked ‘Gue´rin’ in Figure 3 is located over 30 km to the SE of the site, along the Largue river. This area is well known to French prehistorians as a source of raw material, especially in the Neolithic. The opaque, cafe´-au-lait-coloured flint found as one lithic type at the Bau de l’Aubesier contains gastropod fossils, as would be expected for these Oligocene rocks, but the attribution of the archaeological specimens to this source is really based on their distinctive colour and overall macroscopic appearance.
Tertiary The large area labelled ‘Tertiary’ in Figure 3 includes potential raw material sources of both Oligocene and Eocene age, to the south of the site. Some lithic types from the Bau de l’Aubesier contain charophytes, a freshwater alga, and/or gastropods typical of the Oligocene flints in the region, rather than the marine fossil assemblages typical of Cretaceous-aged flints. These fossils tend to be visible to the naked eye, which allows four lithic types to be distinguished and attributed to the area where such flints occur. I am as yet unable to prove which source(s) were used, so I am merely showing the entire large area as a potential source of those four raw material types. (There also remains the possibility that at least one Eocene source was used for one of the types whose provenance has not yet been traced.) Further work may allow more precise sites to be identified.
Roussillon This area is famous to tourists in Provence as the site of spectacular red, orange and yellow ochre deposits. In this vicinity (about 18 km south of the site) it is still possible to find some remnants of a highly weathered siliceous
Bedoulian Bedoulian and other Early Cretaceous biomicritic limestones crop out over much of this region, but they do not all contain flint nodules. The ‘Bedoulian’ area shown in Figure 3 extends from about 3 to 10 km to the SW of the Bau de l’Aubesier. Flint is patchily but abundantly available within that overall area. It can be of poor to very good quality, is generally blue–grey in colour, and may have a reddish cortex. It tends to be richer in sponge spicules than the flint from other areas, and relatively impoverished in other fossils. The macroscopic appearance and the fossil content form the basis for assigning seven lithic types from the Bau de l’Aubesier to this area.
Les Sautarels This small area close to 11 km SW of the Bau de l’Aubesier is the source of a distinctive opaque black or black and yellowish brown flint, often with a thin yellowish cortex. The flint is of Bedoulian age, but contains no or very few fossils. It is of excellent quality, although nodules greater than 15 or 20 cm in size are hard to find. Two raw material types come from this source.
Bezaure Near Les Sautarels and Murs there are small Oligocene outcrops that provide a medium quality, rough-textured, grey–black, somewhat spotted flint containing abundant charophytes. One lithic type from the Bau de l’Aubesier is assigned to this source.
Murs The area around the village of Murs, about 12– 13 km SW of the Bau de l’Aubesier, is famous among French prehistorians as a source of very good quality flint. It was used not only in the Middle Palaeolithic but also through to the Neolithic and even into historical times for gunflints. This flint is again Bedoulian in age, relatively rich in sponge spicules, and generally brown to grey in colour; it very often shows a distinctive dark grey or brown sub-cortical band. This distinctive appearance allows specimens of three lithic types from the Bau de l’Aubesier type list to be assigned to this source area. The microfossil content of the type specimens corroborates this attribution.
RAW MATERIAL ECONOMICS
Treille (NW Bedoulian) The Bedoulian-age limestones extend on the other (northwestern) side of the gorges of the Nesque river, across from the Bau de l’Aubesier and on the slopes rising up to the Mont Ventoux. In this area these formations provide mediocre to fairly good quality grey–brown flint, sometimes striped, with few fossils. Although these outcrops start just over 2 km from the site, flint is most abundant at a greater distance. The sources were not much used, and provided only two raw material types, assigned to this area on the basis of macroscopic appearance corroborated by a relatively impoverished fossil suite in the type specimens. In addition, other layers within the Bau de l’Aubesier contain material from the following sources. As these were not used for layer IV, they are not shown in Figure 3.
Late Cretaceous rocks of this area, they do not seem to have been used at the Bau de l’Aubesier. However, in this area one can also find fossiliferous glauconitic fineto medium-grained quartzose sandstone (‘greensand’), which in terms of mineralogy and granulometry is a close match to the dark coloured, also glauconitic, quartzose sandstone found at the Bau de l’Aubesier. This area is thus the source of one raw material type in the assemblage. It should be noted, however, that the type specimen from the Bau de l’Aubesier assemblage is not clearly fossiliferous, except for some unidentifiable ‘blobs’ that might once have been fossils.
Appendix 2: Attractiveness of the sources The equation
Durance This is a major river that flows southward down from the Alps, forms the eastern boundary of the Vaucluse, then curves to flow in a westerly direction along the southern side of the Luberon mountains, until it reaches the Rhoˆne river at Avignon. At its closest, it is about 30 km south of the Bau de l’Aubesier; it is farther away to the east of the site. The alluvial deposits of this river provided the one piece of a greenish porphyritic andesite (with phenocrysts of plagioclase, and often altered or weathered hornblende and pyroxene) that was found at the Bau de l’Aubesier, as is confirmed by petrographic analysis of thin sections of that specimen and of a piece that I found in the modern alluvial deposits of the Durance south of the Petit Luberon. The likely original source of this rock lies near Brianc¸on, at the headwaters of the Durance, in the French Alps, but the rock was most probably initially eroded by a glacier before being transported by the river.
Mormoiron To the west of the Bau de l’Aubesier, the Early Cretaceous limestone hills slope down into a large plain covered with Quaternary alluvial deposits, and occasional outcrops of a variety of ages from Cretaceous to Miocene. Near the village of Mormoiron the Eocene and the nearby alluvial deposits provide chunks of a very fine-grained and homogeneous siliceous rock, ‘chert’ in the broadest sense of the term, probably a silcrete in origin. Its very pale colour and high degree of translucency, almost to transparency, make it very distinctive in appearance. None of the specimens that I have examined contained any fossils. This is the source of one raw material type in the Bau de l’Aubesier assemblage.
North Aurel Although some very good quality yellow and red flint nodules can be found in ancient (possibly Oligocene?; Goguel 1964) alluvial deposits in pockets eroded into the
177
A(s) ¼
(quality)(extent)(100) size (terrain difficulty)(extraction) scarcity
is used to calculate the attractiveness of each source (or representative sources for the larger source areas) used for layer IV, based on the values listed in the table below. (See Wilson (2007c) for a full explanation of the equation and its use.) A(s) stands for attractiveness (source). Quality is given as a numerical value, subjectively determined, with higher numbers indicating better quality for tool-making. The determination of quality is based on the material’s suitability for stone tool-making, and includes its homogeneity, granulometry, and the presence or absence of cracks, inclusions, and so on. The scale ranges from very poor (given a value of 0), to poor (1), fair (2), good (4), very good (8) and excellent (16). Where a range of qualities is available, an intermediate score can be attributed. Extent of the outcrop or deposit is judged on a size scale, from 1 (small: less than 10 m in diameter) to 2 (medium: 10– 50 m), 3 (extensive: 50–100 m) and 4 (very extensive: greater than 100 m in diameter). Values are multiplied by 100 to eliminate unnecessary decimal points. Size refers to the size of the nodules or pieces available, in centimetres. In most cases the maximum nodule or fragment size is used, but where a large range of sizes is available, the equation can handle multiple sizes and their applicable scarcity or abundance values (see Wilson 2007c). Terrain difficulty refers to the value for energy expenditure in kilocalories per kilometre (Cal km21) calculated for the route from the source to the Bau de l’Aubesier. (See Wilson (2007b) for more details.) Extraction refers to the ease or difficulty of removing usable material from its matrix. At all of these sources, material is exposed at the surface. The Bau is a Middle Palaeolithic site, and although quarrying has been reported from some Palaeolithic sites, including cases dating as far
178
L. WILSON
back as the Early Palaeolithic (Vermeersch 2002; Barkai et al. 2006; Paddaya et al. 2006; Sampson 2006), we assume that all pieces were collected as loose surface material. The extraction value is therefore taken as 1 in all cases. Scarcity, the inverse of abundance, is judged on a numerical scale from 1 (very abundant: more than 50% of the surface area of the source consists of potential raw
Source area name
Local Sault St. Trinit St. Jean Gue´rin Tertiary Roussillon Bedoulian Les Sautarels Bezaure Murs Treille
material) to 2 (abundant: 25–50%), 3 (medium: 5 –25%) and 4 (scarce: less than 5% of surface area). Each size value is associated with a corresponding scarcity value, as is shown in the table below. The source numbers given in the table below are the catalogue numbers attributed to each source in the overall list of 352 potential sources identified by the Vaucluse raw material project.
Source number
Quality
Extent
Terrain difficulty (Cal km21)
Extraction
Nodule size (cm)
Scarcity
A(s)
3 85 38 48 96 87 62 9 8 81 13 203
3 3 2.5 2 5 3 2 4.5 10 3 8.5 3
4 4 3 4 2 4 1 4 2 2 3 4
62.87 68.33 73.98 86.46 79.56 92.00 81.50 93.81 83.77 81.20 87.93 129.31
1 1 1 1 1 1 1 1 1 1 1 1
15, 20 20, 25 25 20 20 15, 60 5 15, 30 5, 20 30, 40 5, 10, 15 30
2, 4 2, 3 4 1 4 3, 4 4 1, 3 3, 4 3, 4 1, 2, 4 4
167.01 224.40 63.36 185.06 62.85 114.13 3.07 351.77 69.64 92.36 219.77 69.60
References Argant, J. 2000. Le Bau de l’Aubesier (Vaucluse): Pre´le`vements 1999 pour l’analyse pollinique et premiers re´sultats. In: Lebel, S. (ed.) Le Bau de l’Aubesier, Monieux –Vaucluse: Fouilles 1998–1999–2000. Unpublished manuscript, 71– 71g. Ataman, K., Carambelas, K. R. & Elston, R. G. 1992. The economics of toolstone extraction and processing. In: Elston, R. G. & Raven, C. (eds) Archaeological Investigations at Tosawihi, a Great Basin Quarry, Part 3: A perspective from locality 36. Report prepared for the Bureau of Land Management, Elko Resource Area, Nevada. Inter-mountain Research and Bureau of Land Management, Silver City, NV, 233–250. Bailey, G. 2007. Time perspectives, palimpsests and the archaeology of time. Journal of Anthropological Archaeology, 26, 198–223. Bamforth, D. B. 1986. Technological efficiency and stone tool curation. American Antiquity, 51, 38–50. Barkai, R., Gopher, A. & Laporta, P. C. 2006. Middle Pleistocene landscape of extraction: quarry and workshop complexes in northern Israel. In: Goren-Inbar, N. & Sharon, G. (eds) Axe Age: Acheulian Toolmaking from Quarry to Discard. Equinox, London, 7– 44. Beck, C., Taylor, A. K., Jones, G. T., Fadem, C. M., Cook, C. R. & Millward, S. A. 2002. Rocks are heavy: transport costs and Paleoarchaic quarry behavior in the Great Basin. Journal of Anthropological Archaeology, 21, 481–507.
Bell, T. & Lock, G. 2000. Topographic and cultural influences on walking the Ridgeway in later prehistoric times. In: Lock, G. (ed.) Beyond the Map: Archaeology and Spatial Technologies. IOS, Oxford, 85–100. Binford, L. R. 1977. Forty-seven trips: A case study in the character of archaeological formation processes. In: Wright, R. V. S. (ed.) Stone Tools as Cultural Markers. Australian Institute of Aboriginal Studies, Canberra, 24–36. Binford, L. R. 1980. Willow smoke and dogs’ tails: Hunter– gatherer settlement systems and archaeological formation processes. American Antiquity, 45, 4 –20. Binford, L. R. 1989. Debating archaeology. Academic Press, San Diego, CA. Binford, L. R. & Stone, N. 1985. ‘Righteous rocks’ and Richard Gould: some observations on a misguided ‘debate’. American Antiquity, 50, 151– 153. Blackwell, B. A. B., Skinner, A. R., Blickstein, J. I. B., Lebel, S. & Leung, H. Y. M. 2001. ESR Isochron dating analyses at Bau de l’Aubesier, Provence, France: Clues to U uptake in fossil teeth. Geoarchaeology, 16, 716 –761. Brantingham, P. J. 2003. A neutral model of stone raw material procurement. American Antiquity, 68, 487– 509. Brantingham, P. J. 2006. Measuring forager mobility. Current Anthropology, 47, 435– 459. Church, T., Brandon, R. J. & Burgett, G. R. 2000. GIS applications in archaeology: method in search of theory. In: Westcott, K. L. & Brandon, R. J. (eds)
RAW MATERIAL ECONOMICS Practical Applications of GIS for Archaeologists: A Predictive Modeling Kit. Taylor & Francis, London, 144–166. Conolly, J. & Lake, M. 2006. Geographical Information Systems in Archaeology. Cambridge Manuals in Archaeology. Cambridge University Press, Cambridge. De Lumley, H. 1971. Le Pale´olithique infe´rieur et moyen du Midi me´diterrane´en dans son cadre ge´ologique. In: Gallia-Pre´histoire, tome 1, Ligurie–Provence, Vie`me supple´ment. CNRS, Paris. Elston, R. G. 1992a. Modeling the economics and organization of lithic procurement. In: Elston, R. G. & Raven, C. (eds) Archaeological Investigations at Tosawihi, a Great Basin Quarry, Part 1: The Periphery, Volume 1. Report prepared for the Bureau of Land Management, Elko Resource Area, Nevada. Inter-mountain Research and Bureau of Land Management, Silver City, NV, 31– 47. Elston, R. G. 1992b. Economics and strategies of lithic production at Tosawihi. In: Elston, R. G. & Raven, C. (eds) Archaeological Investigations at Tosawihi, a Great Basin Quarry, Part 1: The Periphery, Volume 1. Report prepared for the Bureau of Land Management, Elko Resource Area, Nevada. Inter-mountain Research and Bureau of Land Management, Silver City, NV, 775– 801. Elston, R. G., Ingbar, E., Leach, M., Raven, C., Carambelas, K. & Ataman, K. 1992. Research design. In: Elston, R. G. & Raven, C. (eds) Archaeological Investigations at Tosawihi, a Great Basin Quarry, Part 1: The Periphery, Volume 1. Report prepared for the Bureau of Land Management, Elko Resource Area, Nevada. Inter-mountain Research and Bureau of Land Management, Silver City, NV, 49–70. Ericson, J. E. 1977. Egalitarian exchange systems in California: a preliminary view. In: Earle, T. K. & Ericson, J. E. (eds) Exchange Systems in Prehistory. Academic Press, New York, 109–126. Fe´blot-Augustins, J. 1993. Mobility strategies in the Late Middle Palaeolithic of Central Europe and Western Europe: elements of stability and variability. Journal of Anthropological Archaeology, 12, 211– 265. Fernandez, P. 2006. Etude pale´ontologique des ongule´s du mouste´rien du Bau de l’Aubesier (Vaucluse, France): Morphome´trie et contexte biochronologique. Documents du Laboratoire de Ge´ologie de Lyon, 161. Findlow, F. J. & Bolognese, M. 1982. Regional modeling of obsidian procurement in the American Southwest. In: Ericson, J. E. & Earle, T. K. (eds) Contexts for Prehistoric Exchange. Academic Press, New York, 53– 81. Gamble, C. 1999. The Palaeolithic Societies of Europe. Cambridge World Archaeology. Cambridge University Press, Cambridge. Geneste, J.-M. 1985. Analyse lithique d’industries mouste´riennes du Pe´rigord: Une approche technologique du comportement des groupes humains au pale´olithique moyen. The`se de Doctorat, Universite´ de Bordeaux I. Geneste, J.-M. 1988. Syste`mes d’approvisionnement en matie`res premie`res au pale´olithique moyen et au pale´olithique supe´rieur en Aquitaine. In: Otte, M. (ed.) L’Homme de Ne´andertal, Vol. 8: La mutation. ERAUL 35. Universite´ de Lie`ge, Lie`ge, 61–70.
179
Geneste, J.-M. 1989. Economie des ressources lithiques dans le mouste´rien du sud-ouest de la France. In: Otte, M. (ed.) L’Homme de Ne´andertal, Vol. 6: La subsistance. ERAUL 33. Universite´ de Lie`ge, Lie`ge, 75–98. Goguel, J. 1964. Carte ge´ologique a` 1:50,000 et notice explicative, feuille de Se´deron (XXXII-40). BRGM, Orle´ans. Gould, R. A. & Saggers, S. 1985. Lithic procurement in Central Australia: a closer look at Binford’s idea of embeddedness in archaeology. American Antiquity, 50, 117 –136. Helbing, D., Keltsch, J. & Molna´r, P. 1997. Modelling the evolution of human trail systems. Nature, 388, 47–50. Hodder, I. & Orton, C. 1976. Spatial Analysis in Archaeology. Cambridge University Press, Cambridge. Ingbar, E. 1994. Lithic material selection and technological organization. In: Carr, P. J. (ed.) The Organization of North American Prehistoric Chipped Stone Tool Technologies. International Monographs in Prehistory, 7, 45–56. Jeske, R. J. 1992. Energetic efficiency and lithic technology: an Upper Mississippian example. American Antiquity, 57, 467–481. Kuhn, S. L. 1991. ‘Unpacking’ reduction: Lithic raw material economy in the Mousterian of West –Central Italy. Journal of Anthropological Archaeology, 10, 76–106. Kuhn, S. L. 1995. Mousterian Lithic Technology: An Ecological Perspective. Princeton University Press, Princeton, NJ. Kuhn, S. L. 2004. Upper Palaeolithic raw material economies at Uc¸agizli cave, Turkey. Journal of Anthropological Archaeology, 23, 431–448. Lebel, S. 2000a. Monieux: Bau de l’Aubesier. In: Bilan Scientifique 1999. Service Re´gional De l’Arche´ologie, DRAC PACA, Aix-en-Provence, 179– 180. Lebel, S. (ed.) 2000b. Le Bau de l’Aubesier, Monieux – Vaucluse: Fouilles 1998-1999– 2000. Unpublished manuscript. Lebel, S. & Trinkaus, E. 2002. Middle Pleistocene human remains from the Bau de l’Aubesier. Journal of Human Evolution, 43, 659–685. Lebel, S., Trinkaus, E. et al. 2001. Comparative morphology and paleobiology of Middle Pleistocene human remains from the Bau de l’Aubesier, Vaucluse, France. Proceedings of the National Academy of Sciences of the USA, 98, 11097–11102. Llobera, M. 2000. Understanding movement: a pilot model towards the sociology of movement. In: Lock, G. (ed.) Beyond the Map: Archaeology and Spatial Technologies. IOS, Oxford, 65– 84. Masse, J.-P. 1993. Valanginian–Early Aptian carbonate platforms from Provence, Southeastern France. In: Simo, J. A. T., Scott, R. W. & Masse, J.-P. (eds) Cretaceous Carbonate Platforms. AAPG Memoirs, 56, 363 –374. Metcalfe, D. & Barlow, K. R. 1992. A model for exploring the optimal trade-off between field processing and transport. American Anthropologist, New Series, 94, 340 –356. Moulin, F. 1903. L’abri du Bau de l’Aubesier. Bulletin de l’Acade´mie du Var, 1, 1– 84.
180
L. WILSON
Moulin, F. 1904. L’abri mouste´rien du Bau de l’Aubesier. Bulletin de la Socie´te´ Pre´historique de France, 1, 14–20. Newman, J. R. 1994. The effects of distance on lithic material reduction technology. Journal of Field Archaeology, 21, 491– 501. Paddaya, K., Jhaldiyal, R. & Petraglia, M. D. 2006. The Acheulian quarry at Isampur, Lower Deccan, India. In: Goren-Inbar, N. & Sharon, G. (eds) Axe Age: Acheulian Tool-making from Quarry to Discard. Equinox, London, 45–73. Raven, C. 1992. The natural and cultural landscape. In: Elston, R. G. & Raven, C. (eds) Archaeological Investigations at Tosawihi, A Great Basin Quarry, Part 1: The Periphery, Volume 1. Report prepared for the Bureau of Land Management, Elko Resource Area, Nevada. Inter-mountain Research and Bureau of Land Management, Silver City, NV, 7– 30. Reid, P. 1986. Models for prehistoric exchange in the Middle Great Lakes’ Basin. Ontario Archaeology, 46, 33– 44. Renfrew, C. 1977. Alternative models for exchange and spatial distribution. In: Earle, T. K. & Ericson, J. E. (eds) Exchange Systems in Prehistory. Academic Press, New York, 71–90. Roebroeks, W., Kolen, J. & Rensink, E. 1988. Planning depth, anticipation and the organization of Middle Palaeolithic technology: the ‘archaic natives’ meet Eve’s descendants. Helinium, XXVIII, 17– 34. Rouire, J. 1975. Carte ge´ologique a` 1:100,000 et notice explicative, feuille de Carpentras (941). BRGM, Orle´ans. Sampson, C. G. 2006. Acheulian quarries at hornfels outcrops in the Upper Karoo region of South Africa. In: Goren-Inbar, N. & Sharon, G. (eds) Axe Age: Acheulian Tool-making from Quarry to Discard. Equinox, London, 75– 107. Shott, M. 1986. Technological organization and settlement mobility: an ethnographic examination. Journal of Anthropological Research, 42, 15– 51. Texier, P.-J., Lemorini, C., Brugal, J.-P. & Wilson, L. 1996. Une activite´ de traitement des peaux dans l’habitat mouste´rien de La Combette (Bonnieux, Vaucluse, France). Quaternaria Nova, VI, 369– 392. Texier, P.-J., Brugal, J.-P., Lemorini, C. & Wilson, L. 1998. Fonction d’un site du Pale´olithique moyen en marge d’un territoire: l’abri de La Combette (Bonnieux, Vaucluse). In: Economie pre´historique: Les comportements de subsistance au Pale´olithique. APDCA, Sophia-Antipolis, 325– 348.
Texier, P.-J., Brugal, J.-P., Desclaux, E., Lemorini, C., Lopez Saez, J. A., Thery, I. & Wilson, L. 2003. La Combette (Bonnieux, Vaucluse, France): a Mousterian sequence in the Luberon mountain chain, between the plains of the Durance and Calavon rivers. Preistoria Alpina, 39, 77– 90. Tobler, W. 1993. Three Presentations on Geographical Analysis and Modeling. National Center for Geographic Information and Analysis, University of California, Santa Barbara, Technical Report, 93-1. Torrence, R. 1983. Time budgeting and hunter–gatherer technology. In: Bailey, G. (ed.) Hunter–Gatherer Economy in Prehistory: A European Perspective. Cambridge University Press, Cambridge, 11– 22. Torrence, R. (ed.) 1989. Time, Energy and Stone Tools. Cambridge University Press, Cambridge. Trinkaus, E., Lebel, S. & Bailey, S. E. 2000. Middle Paleolithic and recent human dental remains from the Bau de l’Aubesier, Monieux (Vaucluse). Bulletin et Me´moires de la Socie´te´ d’Anthropologie de Paris, 12, 207– 226. Vermeersch, P. M. (ed.) 2002. Palaeolithic Quarrying Sites in Upper and Middle Egypt. Egyptian Prehistory Monographs, 4. Wilson, L. 1988. Petrography of the Lower Palaeolithic Assemblage at the Caune de l’Arago, France. World Archaeology, 19, 376–387. Wilson, L. 1998. Mousterian raw material strategies in a regional context in Southern France. In: Milliken, S. & Peresani, M. (eds) Lithic Technology: From Raw Material Procurement to Tool Production. Workshop No. 12, XIII Congress of the Union of Pre- and Protohistorical Sciences; Forlı`, Italy, September 1996, Universita´ degli Studi di Ferrara, Dipartimento di Scienze Geologiche e Paleontologiche. M.A.C. srl., Forlı`, 55– 63. Wilson, L. 2007a. The Vaucluse raw material project: artifact provenance and landscape context in the Middle Palaeolithic of southern France. In: Wilson, L., Dickinson, P. & Jeandron, J. (eds) Reconstructing Human–Landscape Interactions. Cambridge Scholars, Newcastle, 234–251. Wilson, L. 2007b. Terrain difficulty as a factor in raw material procurement in the Middle Palaeolithic of France. Journal of Field Archaeology, 32, 315–324. Wilson, L. 2007c. Understanding prehistoric lithic raw material selection: application of a gravity model. Journal of Archaeological Method and Theory, 14, 388–411.
Reassessing Hypsithermal human –environment interaction on the Northern Plains ELIZABETH C. ROBERTSON Department of Archaeology and Anthropology, University of Saskatchewan, Archaeology Building, 55 Campus Drive, Saskatoon, Saskatchewan, Canada, S7N 5B1 (e-mail:
[email protected]) Abstract: Palaeoenvironmental records from the Northern Plains of North America attest to an extended period of Middle Holocene warming and drying, making this a useful region and period for research on long-term human response to marked climate change. However, archaeological perspectives on human– environment interaction during this episode have remained preoccupied with a refugial model that incorporates limited latitude for dynamic human adaptation. In part, this situation reflects the challenging geomorphological and typological obstacles faced by those studying this period. However, this paper argues that our failure to develop new perspectives also reflects a longstanding and continued conservatism that casts Northern Plains lifeways as inflexible and unchanging, rather than dynamic and adaptable.
Traditionally, archaeologists who study the precontact cultures of the Northern Plains of North America tend to highlight bison hunting as a particularly fundamental and defining feature of these cultures (see Bryan 2005, for an example of a synthesis). Even the earliest portion of the Northern Plains’ archaeological sequence, with its Clovis groups and their reliance on diverse non-bison megafaunal species, is interpreted largely in terms of the centrality of big game hunting to these groups’ lifeways (e.g. Frison 1991; Hofman & Graham 1998; Dixon 1999). Certainly, current research on the past cultures of the Northern Plains has become increasingly concerned with developing a better understanding of the role of other animal and plant resources in precontact subsistence practices, as well as a more multidimensional view of these cultures that extends beyond subsistence and the closely linked variable of environment (e.g. Duke & Wilson 1995). However, the concept of Northern Plains precontact cultures as bison-focused hunters in much the same mould as the region’s historical-period aboriginal populations has retained its centrality, despite the issues attached to imposing an inflexible and ‘timeless’ view of precontact North American cultures based on historical-period accounts (Trigger 1980). The problems generated by continuing adherence to traditional views of precontact cultures, not only on the Northern Plains, but across the larger region of the Great Plains have, of course, been commented on previously, perhaps most effectively by Duke & Wilson in their introduction to the aptly named volume Beyond Subsistence: Plains Archaeology and the Postprocessual
Critique (Duke & Wilson 1995) (see Fig. 1 for an illustration of the archaeological subdivisions of the Great Plains). None the less, the persistence of an interpretative outlook focused not merely on subsistence but on the primacy of bison in subsistence and culture has tacitly continued to operate as a fundamental paradigm in Northern Plains archaeology, a situation that has resulted in a continuing tendency to closely link human ecology to bison ecology and to regard this relationship as playing a major role in shaping the region’s precontact groups (e.g. Morgan 1980; Vickers 1991; Malainey & Sherriff 1996; Epp & Dyck 2002; Peck 2004). This assumption, in turn, has somewhat limited efforts to develop more nuanced, complex and dynamic understandings of the long trajectory of human–environment interaction on the Northern Plains, despite recent efforts to look at additional dimensions of human subsistence, ecology and culture on the Great Plains (e.g. Kornfeld & Osborn 2003). I would argue that this problem is perhaps most pronounced and least questioned among Northern Plains archaeologists looking at human response to the Hypsithermal, a roughly 3000 year period of elevated temperature and aridity that affected much of North America in the Middle Holocene (Antevs 1948, 1955; Deevey & Flint 1957; Webb et al. 1983; Bartlein et al. 1984; Ritchie & Harrison 1993). Whereas archaeologists working in other parts of the continent have posited dynamic processes of cultural and subsistence change during this time frame, those working on the Northern Plains have asserted that the region’s residents remained largely focused on bison (e.g. Doll 1982, pp. 62–64; Van Dyke & Stewart 1985; Walker
From: Wilson, L. (ed.) Human Interactions with the Geosphere: The Geoarchaeological Perspective. Geological Society, London, Special Publications, 352, 181– 194. DOI: 10.1144/SP352.13 0305-8719/11/$15.00 # The Geological Society of London 2011.
182
E. C. ROBERTSON
Fig. 1. The Great Plains cultural region, showing its archaeological subdivisions. It should be noted that its boundaries differ from those of the Great Plains physiographic region.
REASSESSING HYPSITHERMAL
1992, pp. 97– 109; Frison 1998, pp. 148– 152). The basis for this continuing emphasis is often not explained, although some have suggested unpalatability and/or scarcity of alternative plant and animal food sources (e.g. Forbis 1992, pp. 40, 55). The sustained focus on bison is linked, sometimes explicitly and sometimes implicitly, to the apparent changes in settlement pattern on the Northern Plains during this period. Specifically, the contemporaneous shift to low numbers of sites clustered in the region’s valleys and along its borders is read as reflecting a retreat to adjacent regions and/or to well-watered refugia within the Northern Plains, apparently necessitated by an inflexible reliance on bison that made following them to such areas the only option in the face of Hypsithermal climate change (Fig. 2, Table 1; e.g. Mulloy 1958; Hurt 1966; Wedel 1978; Buchner 1980; Greiser 1985; Forbis 1992; Walker 1992). This suite of ideas remains fundamental in current archaeological understanding of this period on the Northern Plains. However, as archaeological, geoarchaeological and palaeoenvironmental knowledge of this period and region accumulates, it appears increasingly likely that these assumptions need to be explicitly confronted and their basis questioned, to permit an empirically based interpretation of what may very well have been an entirely more complex human interaction with long-term climatic change.
Background Across much of North America, Holocene climate change can be broadly characterized as a transition from to the relatively cool, wet conditions of the immediate postglacial period to a period of significantly elevated warmth and aridity in the Middle Holocene, followed by a shift to the intermediate temperature and precipitation levels of the current climatic regime. This pattern was first described by Antevs (1948, 1955), who named the hot, dry period the Altithermal and associated it with dates of 7000–4500 BP. Currently, many scholars prefer to refer to this period as the Hypsithermal, which, with more broadly defined dates of 9000–2500 BP (Deevey & Flint 1957), better accommodates the time-transgressive fashion in which increased temperature and aridity moved from NW to SE across the continent (Barnosky et al. 1987; Schweger & Hickman 1989, p. 1832; Ritchie & Harrison 1993, p. 411; Vance et al. 1995, p. 94). Of course, this is an extremely generalized view of Holocene climate change in North America, and it fails to incorporate the continuing accumulation of proxy evidence for shorter periods of more or less elevated temperature and precipitation within this
183
time frame (e.g. Vance et al. 1992, p. 881, 1993, p. 117; Grimm 2001, p. 59; Clark et al. 2002, pp. 598–599; Yansa 2007, p. 130). Unfortunately, building this kind of more refined understanding is a continuing process for the Hypsithermal on the Northern Plains, as many of the region’s lakes dried during this period, limiting the availability of high-resolution lacustrine proxy records covering 8500–5500 BP, the approximate time frame during which northwestern North America experienced elevated temperatures and aridity (Ritchie 1985, p. 348; Beaudoin 1993, p. 93; Vance et al. 1995, p. 91; Yansa 1998, p. 430). None the less, there are rare lake pollen records from within the Northern Plains, and these are complemented by alternative forms of proxy data from wetland and glacial meltwater channel sequences and by the more abundant lacustrine pollen records from the aspen parkland and boreal forest zones that border the region (e.g. Watts & Bright 1968; Ritchie 1976, 1985; Vance et al. 1983; Hickman et al. 1984; MacDonald & Ritchie 1986; Barnosky 1989; Schweger & Hickman 1989; Sauchyn 1990; Sauchyn & Sauchyn 1991; Ritchie & Harrison 1993; Laird et al. 1996; Last et al. 1998; Yansa 1998, 2007; Dean & Schwalb 2000; Grimm 2001; Clark et al. 2002; Robertson 2006; Yansa et al. 2007). Collectively, these records show peak Hypsithermal warming and drying on the Northern Plains between about 8500 and 6000 BP, in large part through evidence for northward shifts of the adjacent belts of aspen forest and boreal forest vegetation. There are a number of confirmatory records from within the mixed grass prairie of the Northern Plains, including the macrobotanical, pollen and sedimentological data from Chappice Lake in southeastern Alberta (Vance et al. 1992, 1993); the mineralogical, stable isotope and pollen data from Guardipee and Lost Lakes in Montana (Barnosky 1989); and the macrobotanical and pollen data from prairie pothole sites on and near the Missouri Coteau upland of southern Saskatchewan and North Dakota (Yansa 1998, 2007; Yansa et al. 2007). Additionally, the Cypress Hills, a semiforested upland within the Northern Plains that is currently cooler and wetter than the surrounding mixed grass prairie, have also yielded evidence for Hypsithermal warming and drying through terrestrial phytolith and stable isotope records, and lacustrine pollen, plant macrofossil, diatom, ostracod, mineralogical and sedimentological data (Sauchyn 1990; Sauchyn & Sauchyn 1991; Last & Sauchyn 1993; Porter et al. 1999; Robertson 2006). Despite these changes and their implications for an altered resource base on the Northern Plains, the region’s limited number of excavated sites from the Hypsithermal have generally been interpreted as
184 E. C. ROBERTSON
Fig. 2. Distribution of Early Middle Prehistoric sites on the Northern Plains and adjacent areas. (See Table 1 for site names and designations.)
REASSESSING HYPSITHERMAL
185
Table 1. Early Middle Prehistoric sites on the Northern Plains* Map number
Site designation
1 2 3
FdPe-4 ElPa-1 FdOt-1 FdOt-31
4 5
EhPv-8 EgPs-48 EgPs-51 EgPt-6 EgPt-17 EgPm-3 EgPm-179 EgPn-87 EgPn-146 EgPn-230 EgPn-377 EgPn-480 EgPn-625 EgPn-633 EgPn-700 EhPd-55 EfPs-3 EgPr-2 EdPc-1
6
7 8 9 10 11 12 13
14 15 16 17 18 19 20
21 22 23 24 25 26
EeOv-68 DlPo-20 DjPp-11 DjPq-1 DjPn-16 DjPn-47 DjPn-66 DjPn-90 DjPo-9 DjPo-47 DjPo-49 DjPo-63 DjPo-81 DjPp-8 DjPl-47 DjPm-36 DkPj-1 DjOn-26 DjOm-18 FhNg-25 FfNk-7 FbNp-10 FbNp-17 FbNp-24 FbNp-56 FaNq-25 FaNp-32 EcNx-1 DgMr-1 DhMn-1 FbMi-5 DiMe-27
Site name(s) Boss Hill Scapa Ribstone Anderson Hardisty Bison Pound Vermilion Lakes n.a. Spring Kill n.a. n.a. Mona Lisa Hawkwood n.a. Wimpey n.a. Tuscany n.a. Gooseberry Kill Snack Everblue Springs n.a. n.a. Sibbald Creek Majorville Medicine Wheel Boy Chief Gap n.a. n.a. Sara n.a. Michalsky Jensen Spring n.a. Maple Leaf n.a. n.a. n.a. n.a. n.a. Welsch Head-Smashed-In Bison Jump Stampede East Battle Creek Below Forks St. Louis Redtail Amisk Dog Child Norby Gowen I Gowen II Gray Burial Long Creek Oxbow Dam Swan Valley Atkinson
Map number
Site designation
Site name(s)
27 28 29 30 31 32 33 34
n.a. n.a. 24CA74 24BW626 n.a. n.a. 24PA504 24PA302 24PA401 24SW264 24CB4 24CB5 24CB221 24CB202 24CB84 24CB86 24BH406 24BH1726 48PA201 48PA29 48BH206 48BH330 48BH657 n.a. 48CK7 48WA324 48SH301
Shoup Rockshelters Birch Creek Sun River Indian Creek Mammoth Meadow Barton Gulch Myers– Hindman Carbella Rigler Bluffs Cremer Pretty Creek
35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52
n.a. 48CK303 48FR308 48BH345 48BH499
53 54 55 56
48BH364 48WA363 48WA304 48HO301
57
48WA323
58 59 60
48WA365 48FR1484 48CR122 48NA67
61 62 63 64 65 66 67 68 69 70 71
48GO305 48SW3039 48SW5175 48SW5809 48SW2590 48SW1455 48CR2353 48AB1 48PL68 48LA312 5LR284 5WL45
Magnus Sorensen False Cougar Cave Bobcat Rockshelter Kobold Benson’s Butte Mummy Cave Horner Bottleneck Cave Granite Creek Eagle Shelter BA Cave McKean Carter Cave Bentzen – Kaufmann Cave Spanish Point Hawken Lookingbill Laddie Creek Medicine Lodge Creek Southsider Cave Rice Cave Leigh Cave Wedding of the Waters Cave Little Canyon Creek Cave Bush Shelter Split Rock Ranch Shoreline Dunlap – McMurry Burial Hell Gap n.a. Sweetwater Creek Bozner Maxon Ranch Deadman Wash Medicine House China Wall Patten Creek Pine Bluffs Lightning Hill Wilbur– Thomas (Continued)
186
E. C. ROBERTSON
Table 1. Continued Map number
Site designation
72 73 74
n.a. 5ST85 5BL120
75
5BL170 5BL67 6BL73
76 77 78 79 80 81 82 83 84 85
N5BL70 n.a. 32SL100 32DU605 32DU452 32MN101 32GF123 32RI775 21CE1 39HN570
Site name(s) Yarmony House Vail Pass Camp Fourth-of-July Valley Ptarmigan Hungry Whistler Albion Boardinghouse n.a. Hutton –Pinkham Pretty Butte Tysver–Olson Benz Moe Smilden–Rostberg Rustad Itasca Licking Bison
Map number
Site designation
Site name(s)
86 87 88
n.a. 39WW203 39CU779
89
39BF2 39BF224 39BF225 39BF233 39BF270 25SF17 25FT31 25CC28 25BT3 13ML62 13ML224 n.a. 13CK61 13CK405
Reva Walth Bay Beaver Creek Shelter Medicine Crow Truman Sitting Crow Side Hill McBride n.a. Spring Creek Walker – Gilmore Logan Creek Hill Lungren A. C. Banks Simonsen Cherokee Sewer
90 91 92 93 94 95 96 97 98
*The Early Middle Precontact sites listed in this table and shown in Figure 2 were compiled, for the most part, from Walker (1992), Kornfeld et al. (2010) and Peck (2011). Sites were included if: (1) they yielded at least one radiocarbon date that was accepted by the excavator and that falls within the Hypsithermal and the Early Middle Precontact periods (c. 8500– 4500 B.P.); (2) they yielded diagnostic points that are clearly Early Middle Precontact types; and/or (3) they produced stratigraphic evidence that they fall within the Hypsithermal and the Early Middle Precontact periods, such as nondiagnostic cultural material between layers with Palaeoindian points and the Mazama tephra, a volcanic ash band deposited across much of the Northern Plains at c. 6730 BP (Hallett et al. 1997, p. 1207; Zdanowicz et al. 1999, p. 623). Some sites accepted by previous researchers as Early Middle Precontact were not incorporated in Table 1 and Figure 2 because they failed to clearly meet at least one of these criteria. n.a., not applicable.
indicating that its precontact populations retained an economy focused around bison hunting throughout this time frame (e.g. Gryba 1975, p. 155; Doll 1982, pp. 62 –64; Van Dyke & Stewart 1985; Walker 1992, pp. 97 –109; Frison 1998, pp. 148 – 152). For this reason, the term ‘Archaic’ has been rejected by many Northern Plains archaeologists, specifically because it is used in many other parts of North America to mark the shift from Palaeoindian cultures focused on big game hunting to Hypsithermal-period cultures with subsistence strategies reliant on a much broader variety of plant and animal species (e.g. Forbis 1968, p. 39, 1992, p. 29; Reeves 1985, p. 14; Vickers 1986, pp. 9–10). Instead, Northern Plains cultures of this time frame are commonly referred to as ‘Early Middle Precontact’, thereby avoiding the implication of an ‘Archaic’ subsistence strategy and highlighting the continued reliance on large game that has been postulated for these groups (Reeves 1973, p. 1237, 1985, p. 14; Vickers 1986, pp. 9– 10; Walker 1992, p. 121). This is not to say that these cultures are not believed to have been affected by Hypsithermalinduced environmental change. In fact, the apparent
scarcity of Northern Plains sites dating to this time frame was, in the past, thought to have indicated that the region became unsuitable for occupation by its bison-dependent occupants, resulting in its desertion (e.g. Wedel 1978, p. 199). Although the continuing discovery and excavation of occasional Hypsithermal-period sites on the Northern Plains have largely discredited this model, many researchers continue to subscribe to the related but less radical idea that environmental change in this time frame compelled these groups to retreat to wellwatered valley or upland refugia within or adjacent to the region (e.g. Hurt 1966, p. 111; Buchner 1980, p. 205; Greiser 1985, p. 114; Walker 1992, p. 125; Sheehan 1995, p. 268); although it is often not made explicit, these models appear to assume that, rather than flexibly altering their economy to cope with changing conditions, Early Middle Precontact groups on the Northern Plains altered their location to maintain a subsistence pattern focused on bison exploitation. These ideas are also predicated on the belief that the relative scarcity of Early Middle Precontact sites on the Northern Plains accurately reflects radically decreased population levels and a resultant
REASSESSING HYPSITHERMAL
lack of archaeological site formation at this time; however, these assumptions have been challenged based on two major lines of thinking. First, it has been suggested that, across much of the Great Plains, the Hypsithermal’s hot, dry conditions generated a geomorphological regime that encouraged high levels of erosion, precluding the formation and preservation of sites in most places or burying sites extremely deeply in the rare localities where eroded sediment accumulated (Reeves 1973, p. 1243; Wilson 1983, pp. 420 –421, 1986, pp. 72 – 76, 1990, p. 77; Mandel 1992, p. 93, 1995, p. 60; Artz 1995, p. 67, 1996, p. 384). Second, it has been noted that, although there are still many questions about Northern Plains projectile point typologies for the Early Middle Precontact Period, the medium-sized side-notched Mummy Cave Complex points considered most characteristic of this time frame are not unlike the much more common, medium to small side-notched points typical of the Late Middle and Late Precontact Periods (Reeves 1973, p. 1246; Frison 1991, p. 88; see Kornfeld et al. 2010, pp. 110 –112, 126, 133, for figures that illustrate this phenomenon). Thus, Early Middle Precontact points, particularly small examples found in surface contexts or in unstratified and undated sites, may very well be misidentified as representing later cultures. The apparent preponderance of later sites in the archaeological inventory of the Northern Plains also predisposes analysts toward such attributions, creating circumstances conducive to a continuing pattern of systematic underreporting of Early Middle Precontact sites. It has been noted that the best way to address the first of these hypotheses is to undertake studies designed to identify localities in which Hypsithermalperiod deposits were able to accumulate without subsequent disturbance by erosion, and then assess them for sites (Wilson 1983, pp. 415 –416, 1986, pp. 85– 87; Mandel 1992, p. 94, 1995, pp. 60–61; Artz 1995, p. 83, 1996, p. 384; Robertson 2006, p. 3). Unfortunately, this strategy has not been widely employed, although occasional geoarchaeological studies of the Northern Plains have confirmed that there are contexts in which deposits of Hypsithermal age can be reliably found and that it is not unusual for such deposits to contain archaeological material (Wilson 1983, pp. 415 –416, 1986, pp. 85 –86; Robertson 2006, pp. 256– 257). Also, the high numbers of cultural resource management projects necessitated by recent resource development and population growth in areas such as southern Alberta have resulted in the identification of increasing numbers of Early Middle Precontact sites, often in deeply buried strata in areas such as glacial meltwater channels and alluvial terraces, where local geomorphological characteristics have
187
favoured extensive Holocene sediment accumulation (e.g. Oetelaar 2004; Vivian & Dow 2007; Vivian & Blakey 2009; Vivian et al. 2009). As a result, there is increasing evidence that geomorphological influences are significant in the difficulties archaeologists have encountered in their efforts to identify Early Middle Precontact sites on the Northern Plains and that these difficulties should not be purely attributed to a lack of Early Middle Precontact human activity in this region. Still, Early Middle Precontact sites continue to be rarely identified on the Northern Plains, making the excavation of such sites by both cultural resource management and academic research projects an important source of valuable data. Such projects have, in turn, provided some support for the hypothesis that Early Middle Precontact occupations have been underreported owing to misidentification of Early Middle Precontact projectile points as later types. Specifically, these excavations have yielded clear sequences of radiocarbon dates and/or diagnostic tools from multiple components, allowing definite assignment of a growing number of points to the Early Middle Precontact Period and revealing that many of them do not display the combination of side notching and medium size traditionally considered diagnostic of the Mummy Cave Complex and of this time frame on the Northern Plains (e.g. Gryba 1975, p. 134; Wilson 1980; Doll 1982, pp. 40 –41, 147; Van Dyke & Stewart 1985, pp. 23, 219; Walker 1992, p. 72; Vivian et al. 1998, 2008; Vivian & Dow 2007). For example, excavations of well-dated Early Middle Precontact strata at the Stampede Site, a multicomponent locality in southern Alberta, have produced relatively small side-notched points that, if found in an undated context, would be extremely easy to erroneously classify as Late Middle Precontact Besant or perhaps even Late Precontact Prairie or Plains side-notched points (Gryba 1975, p. 134; Vivian et al. 2008). Furthermore, the proliferation of securely dated Early Middle Precontact points with features atypical of the Mummy Cave Complex, such as basal concavities and corner notches (e.g. Wilson 1980; Doll 1982, pp. 40–41, 147; Walker 1992, p. 72; Vivian et al. 1998, 2008; Vivian & Dow 2007), suggests that many previous Early Middle Precontact finds from undated contexts may have been deemed unclassifiable or erroneously identified as Late Middle Precontact basally concave or corner-notched types, like McKean, Oxbow and Pelican Lake (see Vickers (1986), Kornfeld et al. (2010) and Peck (2011), for more information on and illustrations of these point types). In fact, increasing recognition of the marked level of variability in Early Middle Precontact points on the Northern Plains has encouraged some much-needed reassessment of the point
188
E. C. ROBERTSON
typology for this period (Peck 2010, 2011), which, if successful, will greatly facilitate accurate identification of undated surface and single-component sites of Hypsithermal age. Like geomorphologically informed efforts to locate Early Middle Precontact sites by looking at landforms with deposits of appropriate age, this typological work is critical in establishing if the apparent rarity of Early Middle Precontact sites on the Northern Plains truly reflects limited human activity in this region during the Hypsithermal or is at least in part a product of factors that hinder identification of such sites.
So what’s the problem? Geomorphological and typological studies will sort it out, right? Certainly, archaeologists with an interest in clarifying Early Middle Precontact point typologies are aware that our understanding of this period may very well need adjustment, particularly if they work in areas such as southern Alberta, where the extensive cultural management programmes necessitated by current levels of resource development and population growth are generating new finds on a regular basis. However, many Northern Plains archaeologists, including those interested in using geoscience approaches to better understand the region’s Early Middle Precontact record, continue to base their discussions of human response to the Hypsithermal around the desertion and the refugium models, with their attendant assumption of significant depopulation. I would argue that this tendency reflects an underlying conservatism that has less to do with the utility of these models in explaining the still-sparse data than with a failure to clearly assess the assumptions on which the desertion and refugium models were developed, something for which the time is ripe, particularly given the current potential to develop valuable new perspectives on this time frame. For example, in framing a recent geoarchaeological research project on the archaeological potential of the meltwater channels of southeastern Alberta’s Cypress Hills region, I noted how their geomorphological characteristics favoured high levels of Holocene deposition that was likely to span the Hypsithermal, but I also strongly emphasized how their unusual physiographic features favour relatively cool, wet conditions that might have drawn human groups, particularly during the Hypsithermal (Robertson 2006, pp. 3–8, 252 – 262). I argued that the combination of these influences was responsible for the documented presence of multiple stratified Early Middle Precontact sites in these meltwater channels, and I used this rationale to support my assertion that my proposed
programme of deep subsurface testing in these landforms would result in the identification of additional Early Middle Precontact sites. Ultimately, both the initial design of the project and my interpretation of its results were predicated on a tacit and uncritical acceptance of the refugial view of Hypsithermal human response on the Northern Plains, with the Cypress Hills cast in the role of refugium. A similar tendency is evident in Yansa’s recent synthesis of Northern Plains Hypsithermal palaeoenvironmental records with respect to the implications of these records for Early Middle Precontact lifeways (Yansa 2007). Although an exceptionally valuable and comprehensive piece of work, its consideration of human activity during this period assumes an increased reliance on and concentration in well-watered areas, specifically termed ‘oases’ or ‘refugia’. Thus, the refugium hypothesis once again occupies a central but unexamined role in a study that otherwise represents an extremely sophisticated effort to bring new insights to archaeological understanding of this difficult time frame. Does the refugium hypothesis really need to be examined when invoked in such contexts, or is it resilient and useful to the point that it can and should be used in such a paradigmatic way in discussions of the Early Middle Precontact Period? Certainly, it appears to be a reasonable model for human response to environmental change in a period of increasing temperature and aridity, particularly when these trends extend to the point of significantly reducing available sources of surface water, a situation with important implications for patterns of human mobility. Furthermore, the Northern Plains during the Hypsithermal appear to have experienced significant surface water loss; in fact, a major obstacle in palaeoenvironmental research on this period is gaps in lacustrine sediment records owing to widespread drying of lake basins (Ritchie 1985, p. 348; Beaudoin 1993, p. 93; Vance et al. 1995, p. 91; Yansa 1998, p. 430). However, do such conditions necessitate retreat to refugial areas retaining accessible water? The adaptive strategies of ethnographically documented hunter– gatherer groups in arid and semi-arid regions such as the Great Basin, the Kalahari Desert and Central Australia actually appear to have emphasized relatively high mobility of small groups (e.g. Steward 1938; Tonkinson 1978; Lee 2003), rather than population aggregation in favoured localities. Of course, the irregular distribution of surface water at limited numbers of easily exhausted sources typically tethers the base camps of such groups to a series of particular locales. However, this pattern does not entail a refugial restriction of activities to these locales, but instead often encourages increasingly extended food acquisition trips into
REASSESSING HYPSITHERMAL
surrounding areas until the base camp water supply is used up (Kelly 1995, pp. 126– 127). Alternatively, it can encourage an expansive mobility pattern that involves frequent moves between large numbers of short-term base camps for opportunistic exploitation of ephemeral but reliable water sources. For example, the G/wi of southern Africa’s Kalahari Desert dealt with scarcity of surface water during the dry season by deriving their water from melons and the rumen of ruminant prey animals, a strategy that necessitated frequent moves over considerable terrain (Tanaka 1980; Silberbauer 1981). A somewhat parallel strategy for dealing with water stress can be posited for the winter season on the Hypsithermal-period Northern Plains, when at least some water was presumably widely available in the form of snow cover, even if its accumulation was significantly reduced by the increased temperature and aridity. A pattern that exploited this resource through frequent moves and short stays would, in turn, produce archaeological sites of limited durability and visibility, a situation that, perhaps not coincidentally, is highly consistent with the apparently impoverished record of Hypsithermal-period occupation on the Northern Plains. Regardless of the specifics of the water acquisition strategy of various groups, it is also important to note that the ethnographic data strongly suggest that water scarcity does not encourage hunter– gatherers to increase their population in or restrict their movement to refugia. In fact, the only hunter–gatherers who appear to experience the kind of population concentration and limited mobility implicit in the refugial model are complex hunter–gatherers or fisher– hunter–gatherers (Sassaman 1995). This category includes cultures such as those of the Northwest Coast, the Danish Mesolithic and the Late Archaic Eastern Woodlands, all of whom characteristically lived in conditions where abundant water allowed subsistence to include a substantial proportion of freshwater or marine resources. This phenomenon suggests that, although the refugial model makes intuitive sense in relation to the palaeoenvironmental and archaeological evidence available for the Early Middle Precontact period on the Northern Plains, it is not consistent with the body of data on archaeological and ethnographic hunter– gatherer mobility strategies and needs to be explicitly considered in their light, with particular attention to the nature and archaeological implications of settlement patterns in contexts characterized by water scarcity. Even though such an approach might support the idea that the survival and visibility of Early Middle Precontact period sites is limited by a mobility pattern involving frequent moves and short stays, it cannot be denied that, insofar as such sites have
189
been identified across the Northern Plains, they generally occur in landforms that are or once were close to perennial water sources, such as meltwater channels or alluvial terraces (e.g. Gryba 1975, p. 6; Wilson 1983, p. 222; Walker 1992, p. 1; Oetelaar 2004, p. 726; Johnston 2005, pp. 37, 43; Cyr 2006, p. 5; Vivian & Dow 2007; Vivian & Blakey 2009; Vivian et al. 2009). In fact, this pattern appears to be growing more defined as cultural resource management programmes in areas of growth and development, such as southern Alberta, continue to find occasional sites dating to this period. Does the archaeological evidence for concentration of Early Middle Precontact sites in well-watered areas therefore suggest that the refugial view of Hypsithermal human adaptation in this region retains some validity? I would argue that this is not the case, given that this pattern is more easily explained in terms of Hypsithermal geomorphological influences and ambiguous point typologies than in terms of the observed behaviour of hunter–gatherer groups in regions of water scarcity. Specifically, as has been observed by a number of researchers (e.g. Reeves 1973; Artz 1995), the apparently restricted occurrence of such sites in depositionally active environments such as river valleys and meltwater channels may merely reflect that these are the only contexts in which the burial processes necessary to create and to protect subsurface archaeological strata persisted throughout the Hypsithermal. Thus, occurrence of these sites in areas in close proximity to perennial water may not reflect the clustering of Early Middle Precontact human groups around this resource, but may merely be because this water provided a depositional agent capable of burying nearby archaeological debris in a landscape and at a time when such depositional regimes were scarce. Of course, restriction of Hypsithermal deposition to relatively well-watered areas does not preclude the formation of surface sites on depositionally inactive landforms in nonrefugial areas. In fact, surface sites are commonly found on the depositionally stable Late Pleistocene glacial, glaciofluvial and glaciolacustrine landforms that cover much of the Northern Plains. Such sites are generally relatively recent; however, these kinds of contexts occasionally yield Middle and Early Precontact finds (e.g. Gryba 1983), indicating that archaeological material of considerable time depth can be preserved in the exposed contexts that dominate this region. Thus, if Early Middle Precontact groups employed subsistence and mobility strategies extending beyond heavy reliance on wellwatered refugia, archaeologists should be able to find some evidence of sites reflecting these strategies. However, because most surface finds on the Northern Plains are still dated based on the region’s point typology and given the proliferating
190
E. C. ROBERTSON
questions surrounding its validity for the Early Middle Precontact Period (e.g. Peck 2010b), it appears increasingly likely that any Early Middle Precontact surface sites yielding putatively diagnostic artefacts have been and, for the foreseeable future, will probably continue to be misassigned to the wrong period.
So where does that leave us, and what do we need to do about it? Although both the refugial hypothesis and the geomorphological–typological hypotheses present a number of problems, both also offer explanations for the sparseness of the Early Middle Precontact archaeological record that remain compatible with the data that are currently available for this period on the Northern Plains. However, we cannot effectively assess or confidently subscribe to either without first dealing with the site formation and identification biases identified by the geomorphological–typological hypotheses; only then will we be able to move toward any kind of rigorous and meaningful understanding of human –environmental relations on the Northern Plains during the Hypsithermal. For this reason, it is essential to identify and to implement research strategies designed to deal with these issues. Current efforts to develop a more accurate point typology for this period (e.g. Peck 2010, 2011) are extremely timely and useful in this regard, as they will allow us to more accurately determine if and to what extent Early Middle Precontact occupation is represented at surface sites beyond refugial areas. However, geoarchaeological efforts to identify and examine depositionally active landforms in areas lacking perennial water also need to be pursued to determine if such landforms integrate Hypsithermal deposits; once this has been established, archaeological investigation of these deposits can be conducted to ascertain whether the Northern Plains truly lack evidence for Early Middle Precontact human activity beyond refugial areas, and, if they do not, the nature and extent of this activity. This will not be an easy task, as much of the region beyond its uplands and valleys is dominated by glacial till that has been depositionally inactive since the Late Pleistocene. However, continuing studies of Northern Plains geomorphology have made it increasingly apparent that the region incorporates features such as Holocene sand hills and loess (e.g. Wolfe et al. 2002), which have considerable potential to provide stratified deposits suitable for archaeological site formation and investigation. Of course, these kinds of strategies have been previously recommended by archaeologists interesting in developing a more rigorous and systematic
understanding of human response to the Hypsithermal on the Great Plains. For example, Artz (1996, p. 384) has persuasively argued for the importance of finding and assessing Hypsithermal deposits for sites before passing premature judgement on the extent and nature of Early Middle Precontact cultural activity in this region. However, as noted above, archaeologists dealing with this time frame continue to uncritically revert to cultural interpretations of its poorly understood record, most frequently invoking the refugial model. So why do many archaeologists continue to treat this model as effectively paradigmatic, despite the fact that current shortcomings in our understanding of Hypsithermal geomorphology and typology make it inappropriate to do so? I would argue that, fundamentally, this may represent an example of the continued influence of the nonprogressive view of North American aboriginal cultures described by Trigger (1980). Specifically, Trigger noted that, for an array of historical and cultural reasons, Euroamerican archaeologists up to the time at which he wrote had persisted in maintaining a view of North American aboriginal groups as unchanging over both the short and long terms, despite considerable archaeological and ethnohistorical evidence to the contrary. His analysis and critique of this situation proved timely, becoming a foundational work in subsequent postcolonial approaches geared toward building more sensitive and nuanced understandings of the North American archaeological record. However, such approaches have not yet become entirely commonplace, and an internalization of the new perspectives that they promote is perhaps most challenging in regions such as the Great Plains, which has provided some of the most enduring elements of traditional Euroamerican ideas about the nature of North American aboriginal lifeways. The image of Plains Indians as economically and ideationally tied to mobile bison hunting is certainly an iconic one, and its continuing and pervasive influence can be difficult to escape, even for well-educated and well-informed members of the Euroamerican population, such as archaeologists. Euroamerican ideas regarding the environmental and climatic characteristics of the Great Plains may also exert an influence in this situation. The region was the focus of the Dustbowl of the 1930s, a period of elevated temperature and aridity that proved incompatible with the intensive agriculture regime that Euroamerican settlers established. As a result, many people who were reliant on this lifeway were forced to retreat from the Great Plains. It could be argued that the retreat necessitated by the convergence of an inflexible agricultural adaptation with a short but marked climatic episode has been transposed to archaeological
REASSESSING HYPSITHERMAL
interpretations of the Hypsithermal, based on the tacit assumption that, if the Euroamerican system of land use on the Great Plains was inflexible and unsuitable in the face of increased temperature and decreased precipitation, Early Middle Precontact lifeways were similarly vulnerable and only able to respond in the same fashion as the farmers of the Dustbowl. Thus, these ideas may be a factor in our continuing and uncritical adherence to the refugial view of Northern Plains human response to the Hypsithermal. For if we conceptualize the region’s residents as consistently and inflexibly focusing on mobile bison hunting as the foundation of their culture, we require them to respond to marked environmental and climatic change not by flexibly adjusting their economy but by finding alternative ways to maintain a lifeway that we regard as conservative and unchanging. As noted above, whether retreat to refugia truly would have allowed a continued reliance on such a lifeway is questionable; however, if we do not examine this hypothesis closely, it allows us to maintain the assumption that these groups were unable or disinclined to adjust their lifeways to changing conditions and therefore had to change their location to support these lifeways. It can be argued that the limited subsistence data available from Early Middle Precontact sites on the Northern Plains support this model by indicating that bison remained an important prey species for human groups in this region (Forbis 1968, p. 39; Reeves 1983, p. 1; Vickers 1986, pp. 9–10; Frison 1991, pp. 20 –21; Walker 1992, p. 119); however, the relatively limited number of well-analysed sites from this period and the high degree of fragmentation observed in many of their faunal assemblages (e.g. Doll 1982, p. 61; Walker 1992, pp. 103, 107), suggests that we may be overinterpreting the available evidence in order to fit the region’s Early Middle Precontact groups into the mould of the highly bison-focused Plains Indians of the historical period. However, is this really a problem or does it simply represent an accepted archaeological perspective that we can and will continue to modify as new data become available? This is the manner in which the advancement of knowledge often occurs, but, in this instance, I would argue that the relatively complex and intertwined network of assumptions at play have slowed substantive efforts to critically reassess and thoughtfully rework our understanding of Northern Plains human response to the Hypsithermal. Only by explicitly confronting the reasons for our willingness to uncritically invoke and promote the refugial perspective can we recognize what we need to do to effectively test its validity and develop alternative models. Challenging ourselves to fundamentally reconsider
191
how we see this period and opening ourselves to new interpretative possibilities may not be an entirely comfortable experience; however, given that the Early Middle Precontact Period on the Northern Plains offers a valuable opportunity to better understand human–environment interaction over an extended period of climatic change and instability, it is essential for archaeologists involved in these questions to take on these important and worthwhile tasks. I am deeply indebted to B. Vivian and T. Peck for sharing their valuable insights on current debates in Early Middle Precontact archaeology on the Northern Plains. I would also like to thank C. Westman and A. Korejbo for taking the time to discuss this paper with me during its long gestation, and L. Wilson for her immense patience with that long gestation. Thanks are also extended to two peer reviewers who significantly improved this paper through their helpful and insightful feedback.
References Antevs, E. 1948. Climatic changes and pre-White man. Bulletin of the University of Utah, 38, 168 –191. Antevs, E. 1955. Geologic– climatic dating in the west. American Antiquity, 20, 317– 335. Artz, J. A. 1995. Geological contexts of the Early and Middle Holocene archaeological record in north Dakota and adjoining areas of the Northern Plains. In: Bettis, E. A. III (ed.) Archaeological Geology of the Archaic Period in North America. Geological Society of America, Special Papers, 297, 67–86. Artz, J. A. 1996. Cultural response or geological process? A comment on Sheehan. Plains Anthropologist, 41, 383– 393. Barnosky, C. W. 1989. Postglacial vegetation and climate in the northwestern Great Plains of Montana. Quaternary Research, 31, 57– 73. Barnosky, C. W., Grimm, E. C. & Wright, H. E. Jr. 1987. Towards a Postglacial history of the northern Great Plains. Annals of the Carnegie Museum, 56, 259– 273. Bartlein, P. J., Webb, T. III & Fleri, E. 1984. Holocene climatic change in the northern Midwest: pollen-derived estimates. Quaternary Research, 22, 361– 374. Beaudoin, A. B. 1993. A compendium and evaluation of Postglacial pollen records in Alberta. Canadian Journal of Archaeology, 17, 92–112. Bryan, L. 2005. The Buffalo People: Pre-Contact Archaeology on the Canadian Plains. Heritage House, Surrey, BC. Buchner, A. P. 1980. Cultural responses to Altithermal (Atlantic) climate along the Eastern Margins of the North American Grasslands, 5500 to 3000 B.C. Archaeological Survey of Canada, Mercury Series Paper, 97. Clark, J. S., Grimm, E. C., Donovan, J. J., Fritz, S. C., Engstrom, D. R. & Almendinger, J. E. 2002. Drought cycles and landscape responses to past
192
E. C. ROBERTSON
aridity on Prairies of the northern Great Plains, USA. Ecology, 83, 595– 601. Cyr, T. J. 2006. The Dog Child Site (FbNp-25): a 5500-year-old multicomponent site on the Northern Plains. Master’s thesis, University of Saskatchewan. Dean, W. E. & Schwalb, A. 2000. Holocene environmental and climatic change in the northern Great Plains as recorded in the geochemistry of sediments in Pickerel Lake, South Dakota. Quaternary International, 67, 5– 20. Deevey, R. F. Jr. & Flint, R. F. 1957. Postglacial Hypsithermal interval. Science, 125, 182– 184. Dixon, E. J. 1999. Bones, Boats and Bison: Archeology and the First Colonization of Western North America. University of New Mexico Press, Albuquerque. Doll, M. F. V. 1982. The Boss Hill Site (FdPe-4) Locality 2: Pre-Archaic manifestations in the Parkland of Central Alberta, Canada. Provincial Museum of Alberta Human History Occasional Paper, 2. Duke, P. & Wilson, M. C. (eds) 1995. Beyond Subsistence: Plains Archaeology and the Postprocessual Critique. University of Alabama Press, Tuscaloosa. Epp, H. & Dyck, I. 2002. Early human–bison population interdependence in the plains ecosystem. Great Plains Research, 12, 323–337. Forbis, R. G. 1968. Alberta. In: Caldwell, W. W. (ed.) Northwestern Plains: A Symposium. Rocky Mountain College, Billings, Montana, Center for Indian Studies Occasional Paper, 1, 37– 44. Forbis, R. G. 1992. The Mesoindian (Archaic) period in the Northern Plains. Revista Arqueologia Americana, 5, 27– 70. Frison, G. C. 1991. Prehistoric Hunters of the High Plains, 2nd edn. Academic Press, San Diego, CA. Frison, G. C. 1998. The northwestern and Northern Plains Archaic. In: Wood, W. R. (ed.) Archaeology on the Great Plains. University Press of Kansas, Lawrence, 140– 172. Greiser, S. T. 1985. Predictive models of hunter gatherer subsistence and settlement strategies on the central high plains. Plains Anthropologist Memoir, 20. Grimm, E. C. 2001. Trends and palaeoecological problems in the vegetation and climate history of the northern Great Plains, U.S.A. Biology and Environment: Proceedings of the Royal Irish Academy, 101B, 47–64. Gryba, E. M. 1975. The Stampede Site (DjOn-26): a deeply stratified, multi-component site in the Cypress Hills Provincial Park of Alberta. Report on file with the Alberta Parks Planning Division, Parks Branch, Alberta Department of Parks, Recreation and Wildlife, Edmonton, Alberta. Gryba, E. M. 1983. Sibbald Creek: 11,000 years of human use of the Alberta Foothills. Archaeological Survey of Alberta Occasional Paper, 22. Hallett, D. J., Hills, L. V. & Clague, J. J. 1997. New accelerator mass spectrometry radiocarbon ages for the Mount Mazama Tephra Layer from Kootenay National Park, British Columbia, Canada. Canadian Journal of Earth Sciences, 34, 1202–1209. Hickman, M., Schweger, C. E. & Habgood, T. 1984. Lake Wabamun, Alta.: a paleoenvironmental study. Canadian Journal of Botany, 62, 1438–1465.
Hofman, J. L. & Graham, R. W. 1998. Paleoindian cultures of the great plains. In: Wood, W. R. (ed.) Archaeology on the Great Plains. University Press of Kansas, Lawrence, 87–139. Hurt, W. R. 1966. The altithermal and the prehistory of the Northern Plains. Quaternaria, 8, 101– 113. Johnston, J. S. 2005. The St. Louis Site (FfNk-7) and the Below Forks Site (FhNg-25): the faunal analysis of two Mummy Cave Series and Oxbow complex sites in Central Saskatchewan. Master’s thesis, University of Saskatchewan, Saskatoon. Kelly, R. L. 1995. The Foraging Spectrum: Diversity in Hunter– Gatherer Lifeways. Smithsonian Institution Press, Washington, DC. Kornfeld, M. & Osborn, A. J. (eds) 2003. Islands on the Plains: Ecological, Social and Ritual Use of Landscapes. University of Utah Press, Salt Lake City. Kornfeld, M., Frison, G. C. & Larson, M. L. 2010. Prehistoric Hunter–Gatherers of the High Plains and Rockies. Left Coast Press, Walnut Creek, CA. Laird, K. R., Fritz, S. C., Grimm, E. C. & Mueller, P. G. 1996. Century-scale paleoclimatic reconstruction from Moon Lake, a closed-basin lake in the Northern Great Plains. Limnology and Oceanography, 41, 890–902. Last, W. M. & Sauchyn, D. J. 1993. Mineralogy and lithostratigraphy of Harris Lake, southwestern Saskatchewan, Canada. Journal of Paleolimnology, 9, 23–39. Last, W. M., Vance, R. E., Wilson, S. & Smol, J. P. 1998. A multi-proxy limnologic record of rapid EarlyHolocene hydrologic change on the northern Great Plains, Southwestern Saskatchewan, Canada. Holocene, 8, 503–520. Lee, R. B. 2003. The Dobe Ju/’hoansi, 3rd edn. Nelson Thomson Learning, Toronto, Ontario. MacDonald, G. M. & Ritchie, J. C. 1986. Modern pollen spectra from the western interior of Canada and the interpretation of Late Quaternary vegetation development. New Phytologist, 103, 245–268. Malainey, M. E. & Sherriff, B. J. 1996. Adjusting our perceptions: historical and archaeological evidence of winter on the plains of western Canada. Plains Anthropologist, 41, 333– 357. Mandel, R. D. 1992. Soils and Holocene landscape evolution in central and southwestern Kansas: implications for archaeological research. In: Holliday, V. T. (ed.) Soils in Archaeology. Smithsonian Institution Press, Washington, DC, 41– 100. Mandel, R. D. 1995. Geomorphic controls of the Archaic record in the Central Plains. In: Bettis, E. A. III (ed.) Archaeological Geology of the Archaic Period in North America. Geological Society of America, Special Papers, 297, 37–66. Morgan, R. G. 1980. Bison movement patterns on the Canadian Plains: an ecological analysis. Plains Anthropologist, 25, 143– 160. Mulloy, W. T. 1958. A preliminary historical outline for the northwestern Plains. University of Wyoming Publications in Science, 22. Oetelaar, G. A. 2004. Landscape evolution and human occupation during the Archaic period on the Northern Plains. Canadian Journal of Earth Sciences, 41, 725–740.
REASSESSING HYPSITHERMAL Peck, T. P. 2004. Bison ethology and Native Settlement Patterns during the Old Women’s Phase on the Northwestern Plains. British Archaeological Reports International Series, 1278. Peck, T. P. 2010. Grey matter: revising Alberta Plains prehistory. Paper presented at the 43rd Annual Meeting of the Canadian Archaeological Association, Calgary, Alberta, April 28 to May 1, 2010. Peck, T. P. 2011. Light from Ancient Campfires: Archaeological Evidence for Native Lifeways on the Northern Plains. Athabasca University Press, Edmonton, Alberta. Porter, S. C., Sauchyn, D. J. & Delorme, L. D. 1999. The ostracode record from Harris Lake, southwestern Saskatchewan: 9200 years of local environmental change. Journal of Paleolimnology, 21, 35–44. Reeves, B. O. K. 1973. The concept of an Altithermal cultural hiatus in Northern Plains prehistory. American Anthropologist, 75, 1221– 1253. Reeves, B. O. K. 1983. Culture change in the Northern Plains: 1000 B.C.–A.D. 1000. Archaeological Survey of Alberta Occasional Paper, 20. Reeves, B. O. K. 1985. Northern Plains culture historical systematics. In: Burley, D. V. (ed.) Contributions to Plains Prehistory: The 1984 Victoria Symposium. Archaeological Survey of Alberta, Occasional Paper, 26, 3– 21. Ritchie, J. C. 1976. The Late-Quaternary vegetational history of the western interior of Canada. Canadian Journal of Botany, 54, 1793– 1818. Ritchie, J. C. 1985. Quaternary pollen records from the western interior and the Arctic of Canada. In: Bryant, V. M. & Holloway, R. G. (eds) Pollen Records of Late-Quaternary North American Sediments. American Association of Stratigraphic Palynologists Foundation, Dallas, TX, 327– 352. Ritchie, J. C. & Harrison, S. P. 1993. Vegetation, lake levels, and climate in Western Canada during the Holocene. In: Wright, H. E. Jr., Kutzbach, J. E., Webb, T. III, Ruddiman, W. F., Street-Perrott, F. A. & Bartlein, P. J. (eds) Global Climates since the Last Glacial Maximum. University of Minnesota Press, Minneapolis, 401–414. Robertson, E. C. 2006. Late Quaternary landform development, paleoenvironmental reconstruction and archaeological site formation in the Cypress Hills of southeastern Alberta. PhD dissertation, University of Calgary. Sassaman, K. E. 2004. Complex hunter–gatherers in evolution and history: a North American perspective. Journal of Archaeological Research, 12, 227–280. Sauchyn, D. J. 1990. A reconstruction of Holocene geomorphology and climate, Western Cypress Hills, Alberta and Saskatchewan. Canadian Journal of Earth Sciences, 27, 1504–1510. Sauchyn, M. A. & Sauchyn, D. J. 1991. A continuous record of Holocene pollen from Harris Lake, southwestern Saskatchewan, Canada. Palaeogeography, Palaeoclimatology, Palaeoecology, 88, 13– 23. Schweger, C. E. & Hickman, M. 1989. Holocene paleohydrology of Central Alberta: testing the GeneralCirculation-Model climate simulations. Canadian Journal of Earth Sciences, 26, 1826–1833.
193
Sheehan, M. S. 1995. Cultural responses to the Altithermal or inadequate sampling? Plains Anthropologist, 40, 261 –270. Silberbauer, G. 1981. Hunter and Habitat in the Central Kalahari Desert. Cambridge University Press, Cambridge. Steward, J. H. 1938. Basin– Plateau Aboriginal Sociopolitical Groups. Smithsonian Institution Bureau of American Ethnography Bulletin, 120. Tanaka, J. 1980. The San Hunter–Gatherers of the Kalahari: A Study in Ecological Anthropology. University of Tokyo Press, Tokyo. Tonkinson, R. 1978. The Mardudjara Aborigines: Living the Dream in Australia’s Desert. Holt, Rinehart & Winston, New York. Trigger, B. G. 1980. Archaeology and the image of the American Indian. American Antiquity, 45, 662–676. Vance, R. E., Emerson, D. & Habgood, T. 1983. A MidHolocene record of vegetative changes in central Alberta. Canadian Journal of Earth Sciences, 20, 364– 376. Vance, R. E., Mathewes, R. W. & Clague, J. J. 1992. 7000 year record of lake-level change on the Northern Great Plains: a high-resolution proxy record of past climate. Geology, 20, 879– 882. Vance, R. E., Clague, J. J. & Mathewes, R. W. 1993. Holocene paleohydrology of a hypersaline lake in southeastern Alberta. Journal of Paleolimnology, 8, 103– 120. Vance, R. E., Beaudoin, A. B. & Luckman, B. H. 1995. The paleoecological record of 6 ka BP climate in the Canadian Prairie Provinces. Ge´ographie Physique et Quaternaire, 49, 81–98. Van Dyke, S. & Stewart, S. 1985. Hawkwood Site (EgPm-179): a multicomponent Prehistoric Campsite on Nose Hill. Archaeological Survey of Alberta Manuscript Series, 7. Vickers, J. R. 1986. Alberta Plains Prehistory: a review. Archaeological Survey of Alberta Occasional Paper, 27. Vickers, J. R. 1991. Seasonal round problems on the Alberta Plains. Canadian Journal of Archaeology, 15, 55– 72. Vivian, B. C. & Blakey, J. 2009. Historical resource mitigative studies Crestmont Stage 4 development area (SW 31-24-2-W5M and NW 30-24-2-W5M) Final Report (Permit 2008-159). Report on file with the Archaeological Survey of Alberta, Edmonton, Alberta. Vivian, B. C. & Dow, A. 2007. Historical resources conservation excavations of the Tuscany Stage 9 Subdivision Development (Permits 2005-112 and 2006-163). Report on file with the Archaeological Survey of Alberta, Edmonton, Alberta. Vivian, B. C., Bosch, K. & Reeves, B. O. K. 1998. Historical resources conservation excavations EgPn-230 Final Report (ASA Permit 97-83). Report on file with the Archaeological Survey of Alberta, Edmonton, Alberta. Vivian, B. C., Meyer, D., Roe, J. & Blakey, J. 2008. Historical resource excavations at the Stampede Site (DjOn-26) Final Report (Permit 2007-382). Report on file with the Archaeological Survey of Alberta, Edmonton, Alberta.
194
E. C. ROBERTSON
Vivian, B. C., Roe, J. & Blakey, J. 2009. Historical resource mitigative studies Crestmont Stage 3 development area (NE Section 30-24-2-W5M) Final Report (Permit 2007-141). Report on file with the Archaeological Survey of Alberta, Edmonton, Alberta. Walker, E. G. 1992. The Gowen Sites: cultural responses to climatic warming on the Northern Plains (7500– 5000 B.P.). Archaeological Survey of Canada, Mercury Series, 145. Watts, W. A. & Bright, R. C. 1968. Pollen, seed, and mollusk analysis of a sediment core from Pickerel Lake, northeastern South Dakota. Geological Society of American Bulletin, 79, 855–876. Webb, T. III, Cushing, E. J. & Wright, H. E. Jr. 1983. Holocene changes in the vegetation of the Midwest. In: Wright, H. E. Jr. (ed.) Late-Quaternary Environments of the Midwest. Volume 2: The Holocene. University of Minnesota Press, Minneapolis, 142– 165. Wedel, W. R. 1978. Prehistoric Plains. In: Jennings, J. D. (ed.) Ancient Native Americans. Freeman, New York, 183– 219. Wilson, M. C. 1980. Archaeological excavations, Mount Royal Village (Mona Lisa Site, EgPm-3, Locality C), Calgary, Alberta. Report on file with the Archaeological Survey of Alberta, Edmonton, Alberta. Wilson, M. C. 1983. Once upon a river: archaeology and geology of the Bow River Valley at Calgary, Alberta, Canada. Archaeological Survey of Canada, Mercury Series, 114. Wilson, M. C. 1986. Late Quaternary landscape modification in the Calgary–Cochrane Area of the Bow
Valley and its influence on archaeological visibility. In: Ronaghan, B. (ed.) Eastern Slopes Prehistory: Selected Papers. Archaeological Survey of Alberta, Occasional Paper, 30, 61– 90. Wilson, M. C. 1990. Archaeological geology in Western Canada: techniques, approaches, and integrative themes. In: Lasca, N. P. & Donahue, J. (eds) Archaeological Geology of North America. Geological Society of America, Centennial Special Volume, 4, 61–86. Wolfe, S. A., Ollerhead, J. & Lian, O. B. 2002. Holocene eolian activity in south– central Saskatchewan and the southern Canadian Prairies. Ge´ographie Physique et Quaternaire, 56, 215–227. Yansa, C. H. 1998. Holocene paleovegetation and paleohydrology of a prairie pothole in southern Saskatchewan, Canada. Journal of Paleolimnology, 19, 429–441. Yansa, C. H. 2007. Lake records of Northern Plains Paleoindian and Early Archaic environments: the ‘park oasis’ hypothesis. Plains Anthropologist, 52, 109–144. Yansa, C. H., Dean, W. E. & Murphy, E. C. 2007. Late Quaternary paleoenvironments of an ephemeral wetland in North Dakota, USA: relative interactions of ground-water hydrology and climate change. Journal of Paleolimnology, 38, 441–457. Zdanowicz, C. M., Zielinski, G. A. & Germani, M. S. 1999. Mount Mazama Eruption: calendrical date verified and atmospheric impact assessed. Geology, 27, 621–624.
Index Page numbers in italic refer to Figures. Page numbers in bold refer to Tables. accessibility index 170 Acipenser fulvescens 114 Aegean Coast see Ephesus aeolian landforms, Argentina 45 Agassiz, Lake 110, 119, 122 agriculture 4, 6 development on East European Plains 11, 12 methods of research 13– 14 spatio-temporal analysis 15– 17, 18 central temperate zone 25 northeastern cold zone 19–23 northwestern cold zone 25 southern warm zone 17, 19 western temperate zone 23–25 theory 14–15 terracing in Argentina 40, 44 Alborz Mountains 51 alluvial deposition 27 alluvial fans of Holocene Argentina 42, 44, 45 Tepe Pardis 51 Western Sicily 99–100 alluvial stratigraphy Big Fork River Valley 110, 111 methods of analysis 110– 111 results deposits 112–118 radiocarbon dating 112, 116, 117, 118 results discussed 118–122 summary 122– 123 Ephesus 27–29 sedimentation record in Gulf 29– 30, 34 Hellenistic– Early Roman 30, 31 late Byzantine 32 Roman Imperial–Byzantine 32, 33, 35–36 summary history 36 Altithermal see Hypsithermal Amazons 28 Ananianskyay culture 11 Ananinskii tribe 14 Archaic artefacts 122 Archaic period Argentina 40 Northern Plains 186 Argentina, Tucuma´n Province see Santa Marı´a Valley aridity, impact of 3 Armata Palaeosol 103, 104 Artemis and the Artemision 28, 29 Attalus II Philadelphus 32 attractiveness equation 171– 172, 177– 178 aurochs bone remains 165, 173 Balt culture 11 Bau de l’Aubesier (France) Middle Palaeolithic site 164– 165 bone remains 165– 166 raw materials for tools project rock sources 166 –168 rock uses 168– 170
significance of resource landscape 170–174 summary 174–175 tool assemblage 166 Bavaria see Burgweinting Siedlungskammer beaver remains 114, 116 Bedoulian 176 Bell Beaker Culture 140 Bezaure 176 Big Fork River Valley (USA) 110 geomorphology 111 Holocene stratigraphy methods of analysis 110–111 results alluvial deposits 112–118 radiocarbon dating 112, 116, 117, 118 results discussed 118 –122 summary 122–123 bison hunting 181 Blackduck artefacts 110, 114, 116, 120, 121, 123 Bos bone remains 165, 173 Bronze Age Bavaria 140, 155 –156 China 125 East European Plains 11, 12 Shetland site 73–74 Buff ware 51 Bu¨lbu¨ldag˘ 28 Burgweinting Siedlungskammer biological setting and vegetation 139 climate 139 geological setting 138– 139 history 137–138 hydrological setting 139–140 peat 140 rescue excavation 140–141 charcoal study methods 141 –142 results chronology 150– 152 geochemistry 145, 146, 148 microscopy 148– 150 stratigraphy 142 –145, 145 –146, 147 results discussed Bronze Age 155 –156 Iron Age 156 Mesolithic 152– 154 Neolithic 154 –155 Roman Empire 156– 157 Byzantine, sedimentation at Ephesus 32, 33 d13C record tufa in SW China methods of analysis 88 results 93–94 results discussed 93– 94 14 C dating see radiocarbon Canada see North American Northern Plains Capra bone remains 165, 173 Capreolus bone remains 165, 173
196 cattle bones record 62, 63, 71 breeding 12 Cayster River see Ku¨c¸u¨k Menderes River ceramic period, Argentina 40 Cervus bone remains 165, 173 chaıˆne ope´ratoire 169–170 Chalcolithic sites Ephesus 29 Sicily 99, 104 Tepe Pardis 49– 50, 61, 62, 63, 64– 65 chamois bone remains 165 Chappice Lake, Hypsithermal record 183 charcoal in sedimentary record Big Fork River Valley 112 Burgweinting Siedlungskammer 137 methods of analysis 141– 142 results chronology 150–152 geochemistry 145, 146, 148 microscopy 148–150 stratigraphy 142– 145, 145– 146, 147 results discussed Bronze Age 155– 156 Iron Age 156 Mesolithic 152–154 Neolithic 154– 155 Roman Empire 156–157 Tepe Pardis 62 chernozem soil 11 China Henan Province Shang society– environment interaction 125 climate 126–128 landscape and geomorphology 128 –129 soils and vegetation 129– 130 stratigraphic survey methods 130– 131 results Anyang Project 133 Shangqiu Project 131–132 results discussed 133–135 Shang Period 5 –6 Xiangshui River tufa record 86, 87, 88 methods of analysis 88 results hydrochemistry 88, 89, 90 isotopic composition 91– 93 profiling 90 radiocarbon dating 90– 91 summary of results 93– 94 Cibicides spp. 76, 78, 79–80 Classical–Archaic time, sedimentation at Ephesus 29, 30 climate and climate change 5– 6 4.2 K Event 125, 126 forest zone of East European Plains 16, 17 Holocene tufa record of SW China 86, 87, 88 methods of analysis 88 results hydrochemistry 88, 89, 90 isotopic composition 91– 93 profile 90 radiocarbon dating 90– 91
INDEX summary 93– 94 see also Hypsithermal Clovis groups and artefacts 122, 181 coastal sites, hazards of 27 colonial v. indigenous settlement, Sicily 104–105 Copper Age see Chalcolithic Corded Ware Culture 140 cost-benefit equation 171 cost-of-passage 170 crop rotation 15 Cypress Hills, Hypsithermal record 183, 188 Dama bone remains 165, 173 Derbent River 28 Dongge Cave stalagmites 126 Dunde Ice Core 127 Durance 177 Dustbowl (Great Plains of America) 190 Dyakovskaya culture 11 East European Plain, forest zone 11 agricultural development 11, 12 agricultural landscape analysis methods of research 13– 14 spatio-temporal data 15– 17, 18 central temperate zone 25 northeastern cold zone 19– 23 northwestern cold zone 25 southern warm zone 17– 19 western temperate zone 23–25 theory 14–15 Egypt, Old Kingdom fall 126 Elphidium spp. 76, 78, 79–80 Elymi 98, 99, 103, 104 embedded procurement 163 Entella 97, 99 Ephesus Chalcolithic sites 29 floodplain geomorphology 28–29 geographical setting 27–28 harbour development 35– 36 Neolithic sites 29 sedimentation record in Gulf 29–30, 34 Hellenistic–Early Roman 30, 31 late Byzantine 32 Roman Imperial– Byzantine 32, 33 summary history 36 Equus bone remains 165, 173 erosion 5 impact on site preservation 187, 189 Eryx 97, 99 estuarine sites, hazards of 27 Euphrates, River, effects of 5 Fair Isle, Mesolithic 72 fallow deer bone remains 165, 173 fertilizer in agriculture 15 Finno-Ugric tribes 11 fire use 4 see also charcoal fire history of Burgweinting Siedlungskammer 137–138 biological setting and vegetation 139 climate 139 geological setting 138– 139
INDEX hydrological setting 139– 140 peat 140 rescue excavation 140– 141 charcoal study methods 141–142 results chronology 150–152 geochemistry 145, 146, 148 microscopy 148–150 stratigraphy 142–145, 145– 146, 147 results discussed Bronze Age 155– 156 Iron Age 156 Mesolithic 152–154 Neolithic 154–155 Roman Empire 156–157 flooding see river courses fluvial fans, Holocene of Argentina 42, 44, 45 Folsom artefacts 122 foraging strategy 164 foraminifera use in environmental record on Shetland methods of analysis 74 results 76 results discussed 79–80 forest clearance 130, 133 forest zone of East European Plains agricultural development 11, 12 agricultural landscape analysis methods of research 13– 14 spatio-temporal data 15– 17, 18 central temperate zone 25 northeastern cold zone 19–23 northwestern cold zone 25 southern warm zone 17– 19 western temperate zone 23–25 theory 14–15 Formative period, Argentina 40 foxtail millet 129 France see Bau de l’Aubesier Gaudryina rudis 78, 80 geoarchaeology, defined 1, 2 geochronology see optically stimulated luminescence dating also radiocarbon dating geomorphological mapping, Argentina methods 40– 41 results 41– 45 Germany see Burgweinting Siedlungskammer glaciation, Eastern European Plains 17, 18 glacis landform units 43–44 goat bone remains 165, 173 grain growing 11, 15 great migration 11, 12 Greeks, colonists on Sicily 104, 105 grey ware 52, 54 Guardipee, Hypsithermal record 183 Gue´rin 176 Hadrian’s dam 32 Hallstatt Period 140 Haynesina sp. 78 Hellenistic Period, sedimentation at Ephesus 29, 30, 31
197
Herrenho¨fe 140 hiking algorithm 170 hoe-mattock agriculture 14 Holocene climatic optimum see Hypsithermal East European forest zone 11 land cover change in China 86, 87, 88 tufa record methods of analysis 88 results hydrochemistry 88, 89, 90 isotopic composition 91–93 profile 90 radiocarbon dating 90–91 summary 93– 94 landform evolution in Argentina 45–46 landform evolution in Western Sicily alluvial fans 99–100 Chuddia River 100–103 landscape evolution in Big Fork River Valley (Minnesota) 110, 111 methods of analysis 110–111 results radiocarbon 112, 116, 117, 118 stratigraphy 112–118 results discussed 118 –122 summary 122–123 sea-level highstand 29 Holocene Event (3) 127 horse bone remains 165, 173 Huan River impact on settlement 128, 133, 134 Huguang Maar, Lake, climate record 126 human impacts see agriculture; fire; sedimentation Hypsithermal climatic optimum 65, 181, 183 impact on precontact Northern Plains culture 183, 186 –191 Inca period, Argentina 40 Inva Basin 19, 20, 22 Iran see Tepe Pardis Irish elk bone remains 165 Iron Age East European Plains 12, 14, 15 settlement in Bavaria 140, 156–157 settlements in Sicily 97– 98, 104 Tepe Pardis cemetery 50, 54, 65 sherds 52 irrigation 5, 127 Chalcolithic 49–50 Late Neolithic 51–52, 64, 65 Kama River 19, 20 knapping debris, Palaeolithic 163 Komi-permyaks 19, 20 Ku¨c¸u¨k Menderes River 28, 29 Hadrian’s dam 32 profiles 30 sedimentation rate 29–30 La Combette (France) 164, 166 La Te`ne Period 140 lacustrine record, Hypsithermal problems 183, 188 lake levels, impact of 3
198
INDEX
landform mapping Chuddia River (Sicily) 100– 103 Santa Marı´a Valley (Argentina) methods 40–41 results 41–45 Big Fork River Valley (Minnesota) 110, 111 methods of analysis 110– 111 results radiocarbon 112, 116, 117, 118 stratigraphy 112– 118 results discussed 118–122 summary 122–123 Western Sicily, alluvial fans 99–100 Laurel artefacts 110 Les Sautarels 176 Linear Pottery Culture 138, 140 lithic landscape, Palaeolithic 163 loess, East European Plains 17 Longshan period 125 Lost Lakes, Hypsithermal record 183 Lysimachus 30, 32
West Voe (Shetland) midden sites 69 chronology 71– 72, 73 excavations 69– 71 geographical setting 72– 74 geological setting 72 sedimentological study methods 74 results micropalaeontology 76 particle size data 74– 76 results discussed micropalaeontology 79–80 particle size data 77– 79 summary and conclusions 80– 81 North American Northern Plains archaeological sites 184, 185 –186 cultural divisions 182 Hypsithermal 181, 183 interpreting the impact 183, 186–191 precontact cultures 181 North Aurel 177 Northern Plains see North American Northern Plains
malaria 132 manuring in agriculture 15 Marnas River 28 Medieval Climatic Anomaly 121, 123 Megaloceros bone remains 165 Mesolithic Bavaria 152– 154 Fair Isle 72 North Wales 81 Orkney 72 Shetland Islands 69, 72 Mexico, Formative Period 5 micropalaeontology see foraminifera midden sites, Shetland Islands 69 millet cultivation 129 monsoon belts China 127 Indian Ocean 127–128 Monte Polizzo Project 97, 98, 99 landscape analysis 99 alluvial fans 99–100 Chuddia River 103–104 proto-urban settlements 98–99 site abandonment 103–104 Mormoiron 177 Moscow glacier 17 Mousterian tools 166 Moxostoma macrolepidotum 116 Mummy Cave Complex 187 Murs 176 muskrat remains 114, 116
d18O record Mt Qilian ice core 127 stalagmites in Dongge Cave 126 tufa in SW China methods of analysis 88 results 91– 93 results discussed 93–94 oasis see refugium Oaxacan Coast 5 Obva Basin 19, 20, 21, 22 opol’e (field) areas, East European Plains 17, 18, 19 optically stimulated luminescence (OSL) dating Tepe Pardis methods 56– 57 results 60, 61 oracle bones 132, 133 Orkney, Mesolithic 72 otter remains 114, 116
natural resources see resource exploitation Neanderthal remains 165 Neolithic sites Bavaria 140, 154–155 China 125, 126, 127, 132 Ephesus 29 Sicily population 98, 99, 104 Tepe Pardis, transition to Chalcolithic 49– 50, 51, 61, 62, 63, 64–65
Palaeoindian period, Argentina 40 Palaeolithic, Middle see Bau de l’Aubersier palaeosols 128, 133 Panayirdag 28 Pion, Mount 28, 30 plough agriculture 11, 15 poles’e (wood) area, East European Plains 17, 18, 19 pollen record China Shang Dynasty 130, 133, 134 Shetland 74 Tepe Pardis methods of analysis 55– 56 results 59 porcupine remains 116 porech’e (river) areas, East European Plains 19, 21 Preon, Mount 28, 30 procurement strategy 163 projectile point typologies, Northern Plains 187 provisioning strategy 164 Qilian, Mount ice core 127 Quinqueloculina spp. 78, 79
INDEX radiocarbon dating Big Fork River sediments 112, 116, 117, 118 Burgweinting Siedlungskammer 150– 152 Santa Marı´a Valley (Argentina) 40 Shetland midden 71– 72 Tepe Pardis methods 57–58 results 61, 62– 63, 64 tufa in SW China methods 85 results 90, 92 red deer bone remains 165, 173 redhorse remains 116 refugium hypothesis 188, 191 Regensburg see Burgweinting Siedlungskammer Regional Development period, Argentina 40 resource exploitation 3, 4, 163 Bau de l’Aubesier (France) 164–165 bone remains 165–166 raw materials for tools project rock sources 166– 168 rock uses 168–170 significance of resource landscape 170–174 summary 174–175 tool assemblage 166 river courses, impact on landscape and society 3 Ephesus 27–28 floodplain geomorphology 28– 29 harbour development 35 Neolithic sites 29 sedimentation record in Gulf 29– 30, 34 Hellenistic– Early Roman 30, 31 late Byzantine 32 Roman Imperial–Byzantine 32, 33 summary history 36 Shang Period and society 5– 6, 125 effect of climate on 126 –128 landscape and geomorphology 128– 129 soils and vegetation 129–130 stratigraphic survey methods 130–131 results Anyang Project 133 Shangqiu Project 131– 132 results discussed 133–135 Rodanovskaya culture 19 roe deer bone remains 165, 173 Roman Empire evidence in Bavaria 140–141, 156–157 evidence in Sicily 105 Roman Period, sedimentation at Ephesus 30, 31, 32, 33 Roussillon 176 runoff agriculture 6 Rupicapra bone remains 165, 173 Russian Federation see East European Plain St Jean de Sault 176 St Trinit 175– 176 Santa Marı´a Valley (Argentina) cultural phases 40 environmental setting Holocene 39 Pleistocene 39 geographical setting 37
199
geological setting 37– 39 landform and settlement mapping methods 40–41 results 41, 42 aeolian 45 denudational 43– 44 fluvio-alluvial 45 structural–denudational 43 summary 45– 46 Sault 175 sea level 3 mid-Holocene 29 sedimentation impact on site preservation 187, 189 use in environmental record Big Fork River Valley (USA) 110, 111 methods of analysis 110–111 results alluvial deposits 112– 118 radiocarbon dating 112, 116, 117, 118 results discussed 118– 122 summary 122– 123 Ephesus 27–29 sedimentation record in Gulf 29– 30, 34 Hellenistic–Early Roman 30, 31 late Byzantine 32 Roman Imperial–Byzantine 32, 33, 35– 36 summary history 36 Shetland methods 74 results 74– 76 results discussed 77–79 Sicily, Monte Polizzo Project landscape analysis 99 alluvial fans 99–100 Chuddia River 100– 103 proto-urban settlements 98–99 site abandonment 103–104 Tepe Pardis methods 54– 55 results 58– 59 Segesta 97, 98, 99 Selinus River 28 settlement patterns 14– 15 Shang Period and society 5– 6, 125 effect of climate on 126– 128 landscape and geomorphology 128–129 soils and vegetation 129–130 stratigraphic survey methods 130–131 results Anyang Project 133 Shangqiu Project 131– 132 results discussed 133 –135 sheep bones, record of 62, 71 shellfish, record in Shetland middens 69, 70, 71, 72 Shetland Islands Jarlshof excavation 73– 74 pollen record 74 Scord of Brouster excavation 73 Sumburgh Airport excavation 73 West Voe midden sites 69 chronology 71–72, 73 excavations 69–71
200 Shetland Islands (Continued ) geographical setting 72–74 geological setting 72 sedimentological study methods 74 results micropalaeontology 76 particle size data 74–76 results discussed micropalaeontology 79– 80 particle size data 77–79 summary and conclusions 80–81 Sicily Monte Polizzo Project 97 landform– sediment analysis 99 alluvial fans 99– 100 Chuddia River 100–103 proto-urban settlements 98–99 site abandonment 103 –104 Troina Project 97 Siedlungskammer see Burgweinting Siedlungskammer slash-and-burn agriculture 14 Slav tribes 11 steppe zone 11 sturgeon remains 114 Sus bone remains 165, 173 Tehran Plain see Tepe Pardis Tepe Pardis (Iran) environmental setting 50– 51 alluvial fan 51 rescue excavation 51–52 geochronology methods 56– 58 results 60, 61, 62–63, 64 pollen analysis methods 55– 56 results 59 sedimentary analysis methods 54– 55 results 58–59 stratigraphy 52, 53 summary of results 64– 66 terracing 5 fluvial terraces in Argentina 42, 44, 45 three-field system 15 Textularia spp. 78, 79–80 Tigris, River, effects of 5 tool types Bau de l’Aubesier, assemblage 166 Palaeolithic 163 tree growth rings and climate 127 Treille 177 Trifarina spp. 78, 79–80
INDEX Troina Project 97 tufa record of environmental history in China 86, 87, 88 methods of analysis 88 results hydrochemistry 88, 89, 90 isotopic composition 91– 93 profile 90 radiocarbon dating 90– 91 summary 93– 94 Turkey see Ephesus Ubaid period, Mesopotamia 49 ungulates, record of 70 urban society, development in China 132, 133 Urnfield Period 140 USA Great Plains cultural regions 182 Zuni agriculture 6 see also Big Fork River Valley; North American Northern Plains Valday glacier 17 Valerius Festus 32 Vaucluse region see Bau de l’Aubesier vegas 43 Verde, Rı´o 5 villae rusticae 141 water acquisition strategies 188–189 wheat cultivation 129 wild boar bone remains 165, 173 wildfires 137 see also fire history Woodland artefacts 110, 122 Xia dynasty 126, 129, 133 Xiangshui River setting 86, 87, 88 tufa record methods of analysis 88 results hydrochemistry 88, 89, 90 isotopic composition 91–93 profiling 90 radiocarbon dating 90–91 summary of results 93–94 Yangtze delta, Neolithic culture 126 Yellow River 126, 127, 128, 129, 130, 134 Shang Dynasty stratigraphy 131– 132 Zhou dynasty 126, 132 Zuni agriculture 6