Counting sheep without falling asleep: using GIS to calculate the minimum number of skeletal...

CAA201421st century Archeaology

concepts, methods and tools

Proceedings of the 42nd Annual Conference on Computer

Applications and Quantitative

Methods in Archaeology

Edited by

F. Giligny, F. Djindjian, L. Costa, P. Moscatiand S. Robert

Archaeopress Archaeology

ArchaeopressGordon House

276 Banbury RoadOxford OX2 7ED

www.archaeopress.com

ISBN 978 1 78491 100 3ISBN 978 1 78491 101 0 (e-Pdf)

© Archaeopress and the individual authors 2015

Cover photograph © Emilie Lesvignes

All rights reserved. No part of this book may be reproduced, stored in retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying or otherwise,

without the prior written permission of the copyright owners.

Printed in England by Oxuniprint, Oxford

This book is available direct from Archaeopress or from our website www.archaeopress.com

Contents

Foreword �� v

Computers and mathematics in Archaeology, anatomy of an ineluctable success! �� 1François DJINDJIAN

Chapter 1 Historiography

Towards a History of Archaeological Computing: An Introduction �� 9Paola MOSCATI

A retrospective on GIS and AIS platforms for Public Archaeology in Italy� Searching backward for roots and looking onwards for new methodological road-maps �� 17Mirella SERLORENZI, Ilaria JOVINE, Giorgia LEONI, Andrea DE TOMMASI, Andrea VARAVALLO

Art History of the Ancient Near East and Mathematical Models� An Overview �� 29Alessandro DI LUDOVICO and Sergio CAMIZ

Archaeology and Computer Applications: the automatic cataloging of Italian archaeological heritage �� 35Alessandra CARAVALE

Chapter 2 Field and laboratory data recording

Practical Assessment of a Multi-Frequency Slingram EMI for Archaeological Prospection �� 43François-Xavier SIMON, Alain TABBAGH, Apostolos SARRIS

Utilizing Magnetic Prospection and GIS to Examine Settlement Organization in Neolithic Southeastern Europe �� 53Alexis NIEKAMP, Apostolos SARRIS

Historic Forest Change: New approaches to Land Use Land Cover �� 65Charlotte E. STANCIOFF, Robert G. PONTIUS Jr, Scott MADRY, Elizabeth JONES

Challenges and Perspectives of Woodland Archaeology Across Europe �� 73Hauke KENZLER, Karsten LAMBERS

Archaeological Mapping of Large Forested Areas, Using Semi-Automatic Detection and Visual Interpretation of High-Resolution Lidar Data �� 81Øivind Due TRIER, Lars HOLGER PILØ

Laser scanning and Automated Photogrammetry for Knowledge and Representation of the Rupestrian Architecture in Cappadocia: Sahinefendi and the Open Air Museum of Goreme �� 87Marco CARPICECI, Carlo INGLESE

Advantages and Disadvantages of Digital Approach in Archaeological Fieldwork �� 95Carlo BIANCHINI, Francesco BORGOGNI, Alfonso IPPOLITO

From Survey to Representation of the Model� A Documentation of Typological and Chronological Sequences of Archaeological Artefacts: Traditional and Innovative Approach �� 107Alfonso IPPOLITO, Luca J. SENATORE, Barbara BELELLI MARCHESINI, Gabriella CEROLI

Archaeology in the Third and Fourth Dimensions: A Case Study of 3D Data Collection and Analysis From Prince Rupert, BC, Canada �� 115Alyssa PARKER, Morley ELDRIDGE

Integrated RTI Approaches for the Study of Painted Surfaces �� 123Eleni KOTOULA, Graeme EARL

Survey, Documentation and Analysis of the Archeological Architecture: the House of the Knights of Rhodes in the Forum of Augustus �� 135Carlo BIANCHINI, Gaia Lisa TACCHI

Digital Research Strategies for Ancient Papyri: A Case Study on Mounted Fragments of The Derveni Papyrus �� 145 Eleni Kotoula, Graeme Earl

i

Chapter 3� Ontologies and Standards

Towards Linked-Data in Numismatics: How the DIANA Approach can Improve the Diachrony Integrating Heterogeneous Pieces of Data �� 157Maria CALTABIANO, Mariangela PUGLISI, Antonio CELESTI, Grazia SALAMONE

Celtic Coins in context, a new database �� 165Katherine GRUEL, Agnes TRICOCHE, Philippe CHARNOTET

Uncertainty handling for ancient coinage �� 171Karsten TOLLE, David WIGG-WOLF

Interoperability of the ArSol (Archives du Sol) database based on the CIDOC-CRM ontology�� 179Emeline LE GOFF, Olivier MARLET, Xavier RODIER, Stéphane CURET, Philippe HUSI

Find the balance - Modelling aspects in Archaeological Information Systems �� 187Franz SCHWARZBACH, Felix F. SCHÄFER, Alexander SCHULZE

Integration of Archaeological Datasets through the Gradual Refinement of Models �� 193Cesar GONZALEZ-PEREZ, Patricia MARTÍN-RODILLA

Linked Open Greek Pottery �� 205Ethan GRUBER, T.J. SMITH

The Digital Archaeological Workflow: A Case Study from Sweden �� 215Marcus J. SMITH

Exploring time and space in the annotation of museum catalogues: The Sloane virtual exhibition experience �� 221Stephen STEAD, Dominic OLDMAN, Jonathan Whitson CLOUD

Building comprehensive management systems for cultural – historical information �� 227Chryssoula BEKIARI, Martin DOERR, Dimitris ANGELAKIS, Flora KARAGIANNI

Managing Time Dimension in the Archaeological Urban Information System of the Historical Heritage of Rome and Verona �� 235Alberto BELUSSI, Sara MIGLIORINI, Piergiovanna GROSSI

Towards an Archaeological Information System: Improving the Core Data Model �� 245Muriel van RUYMBEKE, Cyril CARRÉ, Vincent DELFOSSE, Pierre HALLOT, Michelle PFEIFFER, Roland BILLEN

Chapter 4� Internet and Archaeology

Archaeological open access journals: the case of ‘Archeologia e Calcolatori’ �� 257Alessandra CARAVALE, Alessandra PIERGROSSI

Massive Open Online Opportunity: MOOCs and Internet–Based Communities of Archaeological Practice �� 265Jeffrey P. EMANUEL

Moving Instruction Beyond the Museum’s Walls: Priorities in Online Public Education at the Oriental Institute �� 271Kathryn GROSSMAN, Catherine Kenyon, Megaera LORENZ, Brittany HAYDEN

Crowd- and Community-Fuelled Archaeology� Early Results from the MicroPasts Project �� 279Chiara BONACCHI, Andrew BEVAN, Daniel PETT, Adi KEINAN-SCHOONBAERT

The ACCORD project: Archaeological Community Co-Production of Research Resources �� 289Stuart JEFFREY, Alex HALE, Cara JONES, Sian JONES, Mhairi MAXWELL

Chapter 5� Archaeological Information Systems

12000 years of human occupation, 3 meters deep stratigraphy, 12 hectares… A geographical information system (GIS) for the preventive archaeology operation at Alizay (Normandie, France) �� 299S. MAZET, C. MARCIGN, B. AUBRY, I. COMTE, P. BOULINGUIEZ

Records and spatial representations in the context of a rescue excavation: the case of Quincieux (Rhône-Alpes, France) �� 305Ellebore SEGAIN, Veronique VACHON, Bernard MOULIN, Cécile RAMPONI, Wojciech WIDLAK

ii

Cartography and heritage: past practice and future potential for mapping Scotland’s cultural heritage�� 315Peter MCKEAGUE

Visualization based on the Norwegian University Museum Database �� 323Espen ULEBERG, Mieko MATSUMOTO

An Inventory of Lucanian Heritage �� 333Alain DUPLOUY, Vincenzo CAPOZZOLI, Alessia ZAMBON

Integrating complex archaeological datasets from the Neolithic in a web-based GIS �� 341Kai-Christian BRUHN, Thomas ENGEL, Tobias KOHR, Detlef GRONENBORN

Enhanced 3D-GIS: Documenting Insula V 1 in Pompeii �� 349Giacomo LANDESCHI, Nicolò DELL’UNTO, Daniele FERDANI, Stefan LINDGREN, Anne-Marie LEANDER TOUATI

MAPPA Open Data Metadata� The importance of archaeological background� �� 361Francesca ANICHINI, Gabriele GATTIGLIA

A simple way to formalize the dating of stratigraphic units �� 365 Bruno DESACHY

Recognizing temporalities in urban units from a functional approach: three case studies �� 371 Julie GRAVIER

OH_FET: A Computer Application for Analysing Urban Dynamics Over Long Time Spans �� 381Laure SALIGNY, Ludovic GRANJON, Thomas HUET, Gaël SIMON, Xavier RODIER, Bastien LEFEBVRE

An ‘alphabet’ to describe the spatio-temporal dynamics of settlement systems: a relevant representation of time? �� 393Marie-Jeanne OURIACHI, Frédérique BERTONCELLO, with the collaboration of Pierre NOUVEL, Laure NUNINGER, Elise FOVET, Stephane ALIX

Chapter 6� GIS and spatial analysis

ArkeoGIS, Merging Geographical and Archaeological Datas OnlineLoup BERNARD, Damien ERTLEN, Dominique SCHWARTZ

Counting Sheep Without Falling Asleep: Using Gis to Calculate the Minimum Number of Skeletal Elements (Mne) And Other Archaeozoological Measures At Schöningen 13Ii-4 ‘Spear Horizon’ �� 407Alejandro GARCÍA-MORENO, Jarod M. HUTSON, Aritza VILLALUENGA, Elaine TURNER, Sabine GAUDZINSKI-WINDHEUSER

Looking for the best� A comparison between GIS and PageRank based algorithms for preventive archaeology in urban areas �� 413Dubbini NEVIO, Gabriele GATTIGLIA, Augusto PALOMBINI

Analyses of bone modifications on human remains: a GIS approach �� 423Elsa CIESIELSKI, Hervé BOHBOT

Territorial organisation of the Terramare culture (Bronze Age, Italy): use of GIS methodology to tackle societal issues ��431Julie BOUDRY

From the excavation to the territory: contributions of GIS tools to the study of the spatial organization of the archaeological site of Argentomagus (France, Indre, Saint-Marcel/Argenton-sur-Creuse) �� 443Emeline LE GOFF, Laure LAÜT, Yoann RABASTÉ, Françoise DUMASY

The integration of landscape processes in archaeological site prediction in the Mugello basin (Tuscany/Italy) �� 451Elmar SCHMALTZ, Michael MÄRKER, Hans-Joachim ROSNER, Adrew-Williams KANDEL

The use of Burgundy stone from ancient times to the present day �� 459Stéphane BÜTTNER, Delphine MONTANGE, Laure SALIGNY, Marion FOUCHER with the collaboration of Eric LECLERCQ, Marinette SAVONNET

Exploring Intervisibility Networks: A Case Study From Bronze and Iron Age Istria (Croatia and Slovenia) �� 469Zoran ČUČKOVIĆ

iii

Chapter 7� Mathematics and statistics in archaeology

Intentional Process Modeling of Statistical Analysis Methods �� 481Charlotte HUG, Rebecca DENECKÈRE, Ammar AYMEN

Ancient Mesopotamian Glyptic Products, Statistics and Data Mining: A Research Proposal �� 489Alessandro DI LUDOVICO, Sergio CAMIZ

Intrasite spatial analysis applied to the Neolithic sites of the Paris Basin: from the archaeological feature to global analysis �� 497François GILIGNY

Statistical and mathematical models for archaeological data mining: a comparison �� 509Nevio DUBBINI, Adam LODOEN

Chapter 8� 3D Archaeology and virtual Archaeology

Measuring and describing 3D texture �� 519Vera MOITINHO DE ALMEIDA, Juan Antonio BARCELÓ

Old versus new – introducing image-based 3D modeling into the general documentation workflow of archaeological rescue excavations. Case studies: the Čachtice and Bratislava castles, Slovakia �� 529Jan ZACHAR, Seta ŠTUHEC

Beyond spreadsheets: digitising the archaeological artefact inventory process �� 541Arianna TRAVIGLIA, Stephen WHITE, Andrew WILSON

Potentialities of 3D Reconstruction in Maritime Archaeology �� 549Elisa COSTA, Carlo BELTRAME, Francesco GUERRA

Cultural Heritage Documentation in Cave Environnement Using Low-Cost Means for Archaeologists� Case Study of the Larchant Caves in the Fontainebleau Forest in France �� 557Aurelia LUREAU, Mehdi BELARBI, Pascal RAYMOND, Régis TOUQUET

Forum Romanum: A 3D Model For Self-Service Educational Purposes �� 569Philippe FLEURY, Sophie MADELEINE, Nicolas LEFÈVRE

The Virtual Reconstruction of a Small Medieval Town: The Case of Briviesca (Spain) �� 575Mario ALAGUERO, Andres BUSTILLO, Blanca GUINEA, Lena IGLESIAS

25 Years of Experience in Virtual Reconstructions - Research Projects, Status Quo of Current Research and Visions for the Future �� 585Mieke PFARR-HARFST

Chapter 9� Multi-agent systems and complex system modelling

A Density-Based Simulation Approach for Evaluating Prehistoric Population Fluctuations in Finland �� 595Tarja SUNDELL, Juhana KAMMONEN

Multi-Agent Modelling of the Neolithic LBK �� 595Jean-Pierre BOCQUET-APPEL, Richard MOUSSA, Jérôme DUBOULOZ

Explaining the Adoption of ‘Innovations’ in Western Europe during Bronze Age: Towards an Agent-Based Simulation ��613Juan A. BARCELÓ, Florencia DEL CASTILLO BERNAL, Giacomo CAPUZZO, Berta MORELL, Joan NEGRE

Spatial Interaction Simulation Methods for Ancient Settlement Distributions in Central Italy �� 621Taylor OSHAN, Carson FARMER, Eoin O’DONOGHUE

Interactions and network analysis of a rock art site in Morro do Chapéu, Bahia, Brazil �� 631Carlos ETCHEVARNE, Grégoire VAN HAVRE

iv

Proceedings of the 42nd Annual Conference on Computer Applications and Quantitative Methods in ArchaeologyCAA 2014 - 21st Century Archaeology/F. Giligny, F. Djindjian, L. Costa, P. Moscati, S. Robert (eds.)

Counting Sheep Without Falling Asleep: Using Gis to Calculate the Minimum Number of Skeletal Elements (Mne) And Other

Archaeozoological Measures At Schöningen 13Ii-4 ‘Spear Horizon’

Introduction

In the last decades, the application of Geographical Information Systems (GIS) and computing has led to great developments in Spatial and Landscape Archaeology, Heritage Management, predictive modelling and Virtual Archaeology. Today, GIS capacities allow for an even wider range of applications in Archaeology (Scianna & Villa, 2011). The combination of database management and image or vector entities representation makes GIS a useful tool for documentation and management of archaeological collections over a variety of scales. The multi-scalar capacities of GIS make possible its application to different types of analysis, from continent-wide to regional, local and intra-site comparisons, and even the study of single artefacts.

In this paper, we present a methodology for analysing individual bones and bone fragments in a quantitative and automatic way. The aim of such analysis is to calculate the minimum number of skeletal elements (MNE), defined as ‘the minimum number of skeletal portions necessary to account for the specimens representing that portion’ (Lyman, 1994:102). The calculation of MNE is important as it forms the basis of more complicated measures of abundance, such as minimum number of individuals. Together, these measures estimate the number of animals or portions of animals present at a site and provide one piece of evidence to interpret the taphonomic histories of faunal assemblages and associated human behaviours. One downfall is that the calculation of MNE is often difficult and time consuming when dealing with

large assemblages. Using GIS, these calculations can be made quicker and more precise through a batch process. In addition to measuring skeletal part abundance, this methodology forms a basis for other analyses, including the location and density of bones surface modifications (cut marks, percussion marks, carnivore damage, etc.). As a case study, we applied this methodology to the Middle Pleistocene faunal accumulation from Schöningen 13II-4. However, the methodology explained here can be applied to any faunal assemblage, regardless of chronology and context.

2. Materials and methods

2.1. Case study: the Middle Pleistocene Schöningen 13II-4 ‘Spear Horizon’

Schöningen, located in Lower Saxony, Germany, is considered one of the most significant Palaeolithic sites from Central Europe owing to the discovery well-preserved wooden spears in association with a large assemblage of Middle Pleistocene fauna (Fig. 1). Open-cast lignite mining works led to the discovery of numerous archaeological sites within several erosional ‘channels’ indicating a paleolake environment. The discovery in 1995 of a series of wooden spears, considered the oldest known spears in the world (Thieme, 1997), drove worldwide attention to the so called ‘Spear Horizon’. This horizon was originally dated to around 400 kaBP (Richter & Thieme, 2012), although recent U/Th dating resulted in an age of 290±5 kaBP (Sierralta, Frechen & Urban, 2012). This date places level 13II-4 with Marine Isotope Stage 9, which

AbstractIn this paper, we describe a GIS-based methodology for estimating the minimum number of skeletal elements (MNE) and other archaeozoological measures, such as cut mark distribution and density. As a case study, we present a preliminary application to the Middle Pleistocene site of Schöningen 13II-4, the so-called “Spear Horizon”, where a large and exceptionally preserved faunal assemblage imposes difficulties for quantifying skeletal element abundances. We base our methodology on a series of digital templates introduced in a GIS, where each identifiable bone specimen is drawn and the number of overlapping bone fragments is calculated. This methodology yields a direct and accurate calculation of MNE and provides a foundation for assessing other critical archaeozoological measures. Analysis and interpretation of these measures is essential for understanding Palaeolithic subsistence strategies and hominin behaviour.Keywords: GIS, Archaeozoology, Schöningen, Middle Pleistocene.

Alejandro GARCÍA-MORENO, Jarod M. HUTSON,Aritza VILLALUENGA, Elaine TURNER, Sabine GAUDZINSKI-WINDHEUSER

Archaeological Research Centre and Museum for Human Behavioural Evolution, RGZM, MONREPOS, [email protected], [email protected], [email protected], [email protected], [email protected]

407

A. García-Moreno et al. / Counting Sheep Without Falling Asleep

is consistent with biostratigraphic and environmental data (Urban & Sierralta, 2012). Excavations of the ‘Spear Horizon’, which continued until 2007, extended over an area of ca. 3.900 m2 and yielded roughly 15.000 archaeological remains (Serangeli et al., 2012), including a large, exceptionally preserved faunal assemblage (van Kolfschoten, 2014; Voormolen, 2008).

As part of a research theme regarding hominin adaptations to interglacial environments, ongoing archaeozoological studies conducted by the MONREPOS Archaeological Research Centre and Museum for Human Behavioural Evolution, show that the Schoningen 13II-4 faunal assemblage is overwhelmingly dominated by horse (Equus mosbachensis). Other ungulates, such as aurochs (Bos primigenius), bison (Bison priscus), red deer (Cervus elaphus), giant deer (Megaloceros giganteus), roe deer (Capreolus capreolus) and rhinoceros (Stephanorhinus sp.), are also present in addition to carnivores, small mammals, birds and fishes, suggesting a mosaic interglacial environment around the lakeshore site. The composition and taphonomical analysis of the faunal assemblage suggest an in-situ, intensive exploitation of carcasses by hominin groups, probably during several hunting episodes, followed by secondary scavenging by medium-sized carnivores.

While the exceptional preservation of faunal remains allows for extensive, high-resolution archaeozoological analyses, the thousands of identifiable bones impose difficulties for quantification. This is especially true when estimating MNE, where all bone fragments from the same skeletal element must be considered and compared to each other. For that reason, Schöningen 13II-4 is an excellent site to test and develop a GIS-based methodology for estimating archaeozoological measures, such as MNE.

2.2. Methodology: MNE calculation using GIS

The method presented in this paper is based on the previous work by Marean and colleagues (Marean et al., 2001; Abe et al., 2002) who first presented the possibility of using a GIS to estimate MNE using an image analysis approach using digital templates. This method proved to be very useful, but difficult to implement due to limitations in GIS software since it required extensive preparation of bone templates and GIS processing. For that reason, we decided to develop

a new method by taking advantage of the newest software and tools in order to simplify the calculations. In a basic sense, this method is based on counting how many times cells with the same value overlap within a series of rasters.

As part of the documentation and archaeozoological analysis of the Schöningen 13II-4 faunal assemblage, every identifiable bone or bone fragment was drawn in a digital template. These templates include multiple views (lateral, medial, cranial, caudal, etc.) of every bone (e.g., right and left femur or mandible) for all species represented in the faunal assemblage. We used simple line drawings of each bone that include all major anatomical features for orientation. Our templates were previously created just for reference purposes, but were perfectly suited to our GIS analysis. One advantage of this new methodology is that any templates can be used; scanned images from an anatomical atlas, hand drawings, photographs, etc., can be adapted to accommodate the needs of any archaeozoological analysis. The level of template preparation is up to the user, but the only requirement is that the templates remain unchanged throughout the entire process. Templates were used as a base outline where bone fragments were drawn in their corresponding location within that skeletal element. Using a digitalizing pad and image processing software (Adobe Photoshop CS4 Extended), bone fragments were directly drawn on the blank templates (Fig. 2), using identifiable anatomical features to locate, orient and scale bone fragments and fit them to templates. The resulting drawings were recorded as a .jpg image file, using the bone’s ID number as the file name. That way, an image dataset of bones and bone fragments was created, which not only documents the faunal remains, but also can be integrated directly into a GIS as a raster layer. During the drawing process, special care was taken to represent the bone or bone fragment as accurately as possible. Scaling the background template image with the zoom feature to approximate the size of the bone was helpful to create accurate drawings. The ruler or grid feature included in many graphics programs further aids in drawing bone fragments accurately. Additionally, bone modifications, such as cut marks, percussion notches, carnivore damage, etc., were also drawn. A predefined colour scheme was used in order to ensure that the same colour (defined by Red-Green-Blue values) was used for the same kinds of modifications. Image size and resolution were fixed for

Figure 1: Location of Schöningen, and view of the 13II-4 site.

408 CAA 2014 ParisGIS and spatial analysis

each template so that all raster layers included the same extension once introduced in a GIS.

Once bone fragments were drawn, image files were incorporated in a GIS as raster layers. With the fixed colour scheme, every cell corresponding to ‘bone’ had the same value in all rasters in the dataset, different from any other cells representing blank spaces or bone outlines. Similarly, cells corresponding to bone surface modifications, such as cut marks, had the same value in all rasters. In other words, if a colour value of R: 180, G: 180, B: 180 was used for drawing bone fragments, all cells in the raster dataset corresponding to bone fragment had a value of 180.

As explained previously, MNE is estimated by counting the number of times bone fragments overlap within a faunal assemblage, considering every skeletal element and every species separately. That is, if any portion of two or more bones or bone fragments overlap, they cannot have originated from the same bone. Since all the raster layers corresponding to the same kind of template have the same extension, every cell in the raster shares the same position with another cell in a different raster. Considering that cells representing bone fragments have the same value, overlaps can be calculated easily by counting how many times cells with that value appear in every cell location.

In order to make that calculation, the Equal to Frequency tool from ArcGIS’s Spatial Analyst extension was used.

This tool ‘evaluates on a cell-by-cell basis the number of times the values in a set of rasters are equal to another raster’ 1 (Fig. 3). As an input layer, a raster with the same extension as the templates was created, where all cells had the same value as the cells corresponding to ‘bone fragment’ in the raster dataset.

The resulting raster layer shows how many times the value from the input layer appears in every cell position within the raster dataset. However, since templates have three colour bands (red, green and blue), the Equal to Frequency tool considers every raster three times (the searched value appears three times in every raster, one for every colour band). For that reason, the resulting raster layer must be divided by three (using the Raster Calculation tool). The final raster shows the number of times bone fragments from the template data set overlap (Fig. 4). Each cell records a maximum value that corresponds to the maximum number of times cells corresponding to bone fragments overlap, and therefore to MNE. To avoid negligible overlaps that could overestimate MNE, the raster histogram was checked, and highest values represented in very low frequency (just in a few cells) were ignored (Fig. 5). Ignoring the highest values could turn out in a less precise estimation, although the resulting MNE estimation will be more reliable, since the probability that the estimation is correct is higher when MNE value is lower and its frequency is higher.

1 ArcGIS Help Library, 2010, access 01.07.2014.

Figure 2: Example of three left femur fragments drawn onto a template. For illustrative purposes, each bone was photographed in several views relative to the template (top). Each bone was drawn onto individual templates and later

combined into a single template (bottom). The darker shading represents areas of overlap between the fragments.

F. Giligny, F. Djindjian, L. Costa, P. Moscati, S. Robert (eds.) / Proceedings of CAA 2014 Paris



2.3. Further developments

The method presented here can form the basis for calculating other archaeozoological measures. The use of a predefined colour scheme for drawing every bone modification is used to evaluate the location and frequency of each modification type within the raster dataset. However, with this method, templates’ cells have a different value in every colour band, which impedes the use of the Equal to Frequency tool used to estimate the MNE. Due to this limitation, additional processing of drawings is required.

One solution is to reclassify, using a batch process, templates’ cells values. In this case, values of cells corresponding to modifications, such as cut marks, are changed to ‘1’, while the rest of the cells are given a value of ‘0’. Once all drawings have been reclassified, raster layers can be summed, using ArcGIS’s Raster Calculator Addition function, which sums on a cell-by-cell basis the values of two or more rasters. The resulting raster layer shows the location of bone modifications for each bone, including any locations where modifications overlap (Fig. 6a). Once locations of bone modifications are known, further analyses can be undertaken, such as evaluating the spatial distribution of cut marks across each skeletal element. In this sense, a density analysis will show the

areas where higher concentrations of modifications appear (Fig. 6b), allowing for more accurate and meaningful taphonomical analyses.

Despite the need of previous processing, this method makes possible deeper archaeozoological analysis using a single drawing for every bone fragment, instead of using different templates for every modification. Analysing the preferential locations of bone modifications allows for a better understanding of human behaviour, such as butchering techniques and exploitation of carcasses in the case of cut marks or percussion notches.

Discussion and conclusions

Estimating MNE and other zooarchaeological measures, such as cut mark frequency and location, is a rather complex and time consuming process. This is especially difficult when dealing with very large faunal assemblages where hundreds, or even thousands, of bone fragments must be compared. This is the case for the large and well-preserved faunal assemblage from the Middle Pleistocene site of Schöningen 13II-4.

In order to improve MNE estimation, a GIS-based method was developed and tested with the Schöningen 13II-4 faunal assemblage. This method was based on the use of

Figure 4: Raster layer showing the final number of bone fragments overlapping for Schöningen 13II-4 horse left femora.

Figure 3: Schema of the Equal to Frequency tool calculation process (Image: ArcGIS Help 10.1).


templates, where bone fragments were drawn as part of the documentation process and the archaeological analysis. Those drawings were then introduced as raster layers in to a GIS, and the number of times cells corresponding to bone fragments overlap was calculated. Templates can also be used to evaluate bone modifications, analysing the distribution, concentration and density of alterations such as cut marks.

The application of GIS allows for the calculation of different archaeozoological measures in an easy, batch-like way using any digital template. The method presented here is not entirely automatic since it still requires a few calculations to obtain the final layer showing bone fragment overlaps. In addition, the evaluation of other archaeozoological measures, such as cut marks, needs previous processing of the templates. Despite some additional processing,

Figure 5: Raster’s histogram, showing cells’ value frequency. Negligible frequencies in higher values can be ignored, in order to obtain a more accurate MNE estimation.

Figure 6: Raster’s histogram, showing cells’ value frequency. Negligible frequencies in higher values can be ignored, in order to obtain a more accurate MNE estimation.

F. Giligny, F. Djindjian, L. Costa, P. Moscati, S. Robert (eds.) / Proceedings of CAA 2014 Paris



the method presented here represents an improvement of traditional (manual), extremely time consuming methods and allows for more complex archaeozoological analyses of human subsistence strategies and behaviour. Further development of the methodology presented here could improve the level of automation for this procedure, for example creating a workflow model stringing together the sequence of steps and calculations required. The use of free software with similar capacities and tools as the ones used here would allow the application of this method to any archaeozoological analysis, regardless of the availability of funds. Advancing the method tested on the Schöningen 13II-4 assemblage will improve archaeozoological analyses and will allow for a better understanding of past human behaviour.

Acknowledgements

We would like to thank Curtis W. Marean (Arizona State University) for providing us the details of his previous work, which is the basis for the methodology provided in this paper. The project Subsistenzstrategien mittelpleistozäner Homininen in Schöningen is funded by the Deutsche Forschungsgemeinschaft (DFG; GA6839/-1). We would like to thank the Niedersächsisches Landesamt für Denkmalpflege for their support of the project. AGM attended the CAA2014 Conference thanks to a Deutscher Akademischer Austauschdients.

References

ABE, Y., MAREAN, C.W., NILSSEN, P.J., ASSEFA, Z., STONE, E.C. (2002). ‘The analysis of cutmarks on archaeofauna: a review and critique of quantification procedures, and a new image-analysis GIS approach’, American Antiquity 67, pp. 643-663.

LYMAN, R.L. (1994). Vertebrate Taphonomy. Cambridge : Cambridge University Press.

MAREAN, C.W., ABE, Y, NILSSEN, P.J, STONE, EC (2001). ‘Estimating the minimum number of skeletal elements (MNE) in zooarchaeology: a review and a new image-analysis GIS approach’, American Antiquity 66, pp. 333-348.

RICHTER, D., THIEME, H. (2012), ‘One first chronometric date for the Lower Palaeolithic occupation at Schöningen 13I’ In: K.E. Behre (ed.). The chronological setting of the Palaeolithic sites of Schöningen, Mainz: Römisch-Germanisches Zentralmuseum Mainz, pp. 171-182.

SCIANNA, A., VILLA, B. (2011). ‘GIS applications in archaeology’, Archeologia e Calcolatori 22, pp. 337-363.

SERANGELI, J., BÖHNER, U., HASSMANN, H., CONARD, N. (2012). ‘Die pleistozänen Fundstellen in Schöningen – eine Einführung’, In: K.E. Behre (ed.). The chronological setting of the Palaeolithic sites of Schöningen, Mainz: Römisch-Germanisches Zentralmuseum Mainz, pp. 1-22.

SIERRALTA, M., FRECHEN, M., URBAN, B. (2012). ‘230Th/U dating results from opencast mine Schöningen’, in The chronological setting of the Palaeolithic sites of Schöningen, ed s K.E. BEHRE, Römisch-Germanisches Zentralmuseum Mainz, Mainz, pp. 143-154.

THIEME, H. (1997). ‘Lower Palaeolithic hunting spears from Germany’, Nature 385, pp. 807-810.

URBAN, B., SIERRALTA, M. (2012). ‘New palynological evidence and correlation of Early Palaeolithic sites Schöningen 12 B and 13 II, Schöningen open lignite mine’, in The chronological setting of the Palaeolithic sites of Schöningen, ed s K.E. BEHRE, Römisch-Germanisches Zentralmuseum Mainz, Mainz, pp. 77-96.

VAN KOLFSCHOTEN, T. (2014). ‘The Palaeolithic locality Schöningen (Germany): A review of the mammalian record’, Quaternary International 326-327, pp. 469-480.

VOORMOLEN, B. (2008). Ancient hunters, modern butchers. Schöningen 13II – 4, a kill – butchery site dating from the northwest European Lower Palaeolithic. Universiteit van Leiden, Leiden.


Counting sheep without falling asleep: using GIS to calculate the minimum number of skeletal...

Documents

Transcript of Counting sheep without falling asleep: using GIS to calculate the minimum number of skeletal...