Proceedings of the First Zooarchaeology Conference in Portugal: Held at the Faculty of Letters, University of Lisbon, 8th-9th March 2012 9781407313047, 9781407342696

Table of contents :
Front Cover
Title Page
Copyright
Table of Contents
FOREWORD: The Zooarchaeology in Portugal – 150 years of bones…
Exploitation of bone and antler in the Upper Palaeolithic of Portugal
So many rabbits! Small animal’s dietary role in the Mesolithic shellmidden of Cabeço da Amoreira (Muge, Central Portugal) – preliminary results
Study of the mammals recovered in Cabeço da Amoreia (Muge – Portugal): State of the art
Mesolithic and Neolithic shell middens in Western Algarve: Issues in ecology, taphonomy and economy
The inclusion of faunal remains in Bronze Age funerary practices in Southern Portugal. Montinhos 6 - a case study
In death as in life. Ties between man and animals in the recent prehistory of lower Alentejo: two case studies from Alto de Brinches 3 and Torre Velha 3 (Serpa, Alentejo, Portugal)
Mammal remains from the Governor's House (Belém Tower, Lisbon) and Rua dos Correeiros (Baixa, Lisbon) in the context of fish processing factories in Lusitania
Animal bones from the Roman site of Tróia (Grândola, Portugal): mammal and bird remains from the fish-salting workshop 2 (2007/08)
The faunal assemblage from a roman well in the villa of São Miguel de Odrinhas (Sintra, Portugal). A preliminary view on the archaeological context
A contribute to know the dietary habits in Tavira. From the Iron Age to the Modern Period
What did the Romans and Moslems eat in Conimbriga (Portugal)? The animal bones from the 1990’s excavations
Zooarchaeological perspective of the Islamic sites in Algarve: Current State of Knowledge
At table with the nuns: the mammals of 17th century Santa-Clara-a-Velha Monastery (Coimbra, Portugal)
Early European knowledge and trade of Neotropical mammals: a review of literary sources between 1492 and the first two decades of the 16thcentury
On the origin of the Iberian Chameleons
On archaeofauna experts and commercial archaeology in Portugal: present scenario and considerations

Citation preview

Proceedings of the First Zooarchaeology Conference in Portugal Held at the Faculty of Letters, University of Lisbon, 8th–9th March 2012 Edited by

Cleia Detry Rita Dias

BAR International Series 2662 2014

Published in 2016 by BAR Publishing, Oxford BAR International Series 2662 Proceedings of the First Zooarchaeology Conference in Portugal © The editors and contributors severally and the Publisher 2014 by Joel Marteleira depicts a Roman pit containing faunal remains from Odrinhas (Sintra, Portugal). This find/context is described by Alexandre Gonçalves in this volume. COVER IMAGE

The authors' moral rights under the 1988 UK Copyright, Designs and Patents Act are hereby expressly asserted. All rights reserved. No part of this work may be copied, reproduced, stored, sold, distributed, scanned, saved in any form of digital format or transmitted in any form digitally, without the written permission of the Publisher.

ISBN 9781407313047 paperback ISBN 9781407342696 e-format DOI https://doi.org/10.30861/9781407313047 A catalogue record for this book is available from the British Library BAR Publishing is the trading name of British Archaeological Reports (Oxford) Ltd. British Archaeological Reports was first incorporated in 1974 to publish the BAR Series, International and British. In 1992 Hadrian Books Ltd became part of the BAR group. This volume was originally published by Archaeopress in conjunction with British Archaeological Reports (Oxford) Ltd / Hadrian Books Ltd, the Series principal publisher, in 2014. This present volume is published by BAR Publishing, 2016.

BAR

PUBLISHING BAR titles are available from:

E MAIL P HONE F AX

BAR Publishing 122 Banbury Rd, Oxford, OX2 7BP, UK [email protected] +44 (0)1865 310431 +44 (0)1865 316916 www.barpublishing.com

Table of Contents

Foreword. The Zooarchaeology in Portugal – 150 years of bones........................................................ 1 Cleia Detry & Rita Dias Exploitation of bone and antler in the Upper Palaeolithic of Portugal.................................................. 5 Marina Almeida Évora So many rabbits! Small animal’s dietary role in the Mesolithic shellmidden of Cabeço da Amoreira (Muge, Central Portugal) – preliminary results................................................. 11 Rita Dias & Cleia Detry Study of the mammals recovered in Cabeço da Amoreia (Muge – Portugal): State of the art...................................................................................................................................... 19 Alexandra Pereira Mesolithic and Neolithic shell middens in Western Algarve: issues in ecology, taphonomy and economy....................................................................................... 23 Maria João Valente The inclusion of faunal remains in Bronze Age funerary practices in Southern Portugal. Montinhos 6 - a case study............................................................................... 33 Cláudia Costa & Lidia Baptista In death as in life. Ties between man and animals in the recent prehistory of lower Alentejo: two case studies from Alto de Brinches 3 and Torre Velha 3 (Serpa, Alentejo, Portugal)................................................................................................................... 47 Eduardo Porfírio & Miguel Serra Mammal remains from the Governor's House (Belém Tower, Lisbon) and Rua dos Correeiros (Baixa, Lisbon) in the context of fish processing factories in Lusitania................................................................................................... 57 Silvia Valenzuela-Lamas Animal bones from the Roman site of Tróia (Grândola, Portugal): mammal and bird remains from the fish-salting workshop 2 (2007/08)............................................. 69 Mariana Nabais The faunal assemblage from a roman well in the villa of São Miguel de Odrinhas (Sintra, Portugal). A preliminary view on the archaeological context.............................................................................. 77 Alexandre Gonçalves A contribute to know the dietary habits in Tavira. From the Iron Age to the Modern Period............. 87 Jaquelina Covaneiro & Sandra Cavaco What did the Romans and Moslems eat in Conimbriga? The animal bones from the 1990’s excavations................................................................................... 97 Cleia Detry, João Luís Cardoso & Virgílio H. Correia Zooarchaeological perspective of the Islamic sites in Algarve. Current State of Knowledge........... 111 Vera Pereira

At table with the nuns: the mammals of 17th century Santa-Clara-a-Velha Monastery (Coimbra, Portugal)........................................................................ 117 Cleia Detry, Lígia Inês Gambini & Artur Corte-Real Early European knowledge and trade of Neotropical mammals: a review of literary sources between 1492 and the first two decades of the 16th century................. 129 Marco Masseti & Cecilia Veracini On the origin of the Iberian Chameleons........................................................................................... 139 Octávio S. Paulo On archaeofauna experts and commercial archaeology in Portugal: present scenario and considerations................................................................................................... 145 Nelson Almeida & Cláudia Costa

FOREWORD The Zooarchaeology in Portugal – 150 years of bones… Cleia Detry1 & Rita Dias2 1 2

UNIARQ – Centro de Arqueologia da Universidade de Lisboa Núcleo de Arqueologia e Paleoecologia – FCHS, Universidade do Algarve

Introduction In 1999 the Archaeosciences lab (former CIPA from IPA) was formally established, today belonging to DGPC, with its main contribute having been the construction of an extensive osteological reference collection of mammals, birds and fish which is open to students and researchers making it essential to Zooarchaeologists work. This lab team also produced an extensive number of publications and zooarchaeological studies (Moreno-Garcia et al. 2003, Davis & Moreno-Garcia, 2007).

Zooarchaeology, the study of faunal materials recovered from archaeological sites, had a slow beginning among Portuguese archaeologists. Nevertheless, over the last two decades substantial advances have been made, resulting in exponential increases in the number of sites sampled and scholarly reports/papers published on archaeofauna in Portugal. This development was made by the individual efforts of scholars such as Professor JL Cardoso, with the creation of the Archaeosciences Laboratory (DGPC, formerly IGESPAR) that included a comprehensive faunal bone reference collection, and by the introduction of courses on zooarchaeology in several universities (e.g. U. of the Algarve), thus making archaeologists better trained on the subject.

Over the last decade numerous young students and researchers have dedicated their work to animal bone analysis and consequently, archaeologists are also giving increased importance to animal remains. The zooarchaeology’s future is promising and Portugal is benefiting from the surging of new minds dedicated to this cause and the organization of the International Conference on Portuguese zooarchaeology in 2012 is a proof of that.

One of the most cited early publications on animal bones is Pereira da Costa’s (1865) paper in which he mentions remains recovered by Carlos Ribeiro at Cabeço da Arruda excavations (Muge shell middens). Pereira da Costa used the faunal species to interpret the chronology of the site at a time where direct dating, such as C14, was far from being discovered, and the species present provided a good proxy for the age of the site. He further made Paleoecological and Paleoconomical interpretations of these past populations through animal bones.

Contributions to this volume This volume is the result of the First International Conference in Portugal held at the Faculty of Letters of the University of Lisbon during 8 and 9 March of 2012. It brought together several persons that have dedicated their works to Zooarchaeological studies in Portugal, both foreigners that came for several years to collaborate in Portuguese projects, and Portuguese researchers that have dedicated or began to dedicate their research to this area of expertise. During the two days of the conference a wide range of Portuguese research was presented, not only exclusively by zooarchaeolgy students and/or senior researchers but encompassing techniques such as ancient DNA. The resulting meeting was highly enriching, as can be seen from the articles presented in this volume.

Portuguese archaeology benefited from the newly established scientific community in Portugal, particularly through the efforts of Carlos Ribeiro, a notorious and visionary geologist who established the Geological Commission of the Portuguese Kingdom in 1857. He further promoted geological surveys throughout Portugal which lead to the discovery of numerous important archaeological sites. He also promoted the publication of the results in international conferences such as the IX Session of the International Congress for Prehistoric Anthropology and Archaeology, held in Lisbon (Ribeiro, 1884, Delgado, 1884). Significantly, Ribeiro also collected and documented zooarchaeological materials from the sites he studied.

The conferences, as well as the papers in this volume, discuss faunal remains recovered in Portuguese sites, and include the work of specialists and non-specialists which deepened the interpretations of the sites. Archaeologists also participated giving their view of animal bones interpretation and showing how important the interpretation of this particular kind of remains is important to their readings, something that in Portuguese archaeology is clearly innovative.

Unfortunately, during the 20th century, zooarchaeological research was more or less forgotten except in a few studies made by foreigners (eg. Harlé 1910-11). Since the 1980s two authors – MT Antunes (1985, 1987) and JL Cardoso (1993) have been almost the sole researchers working in this subfield. 1

Covaneiro & Cavaco do an extended description of the animal remains recovered from several sites in Algarve from the Iron Age to the Modern period, giving a comprehensive guide to the work done in this area of southern Algarve. Also in the Algarve, V. Pereira presents a broad analysis of Islamic sites and the use of animals by these populations.

This volume is organized by the chronology of the sites in order to allow a comprehensive and quick finding of the articles. The later chapters discuss works that either look at broader time span or use research techniques that are different from orthodox zooarchaeology. We begin with the paper of M. Évora on the worked bone industries from Upper Palaeolithic sites in Portugal, which summarizes earlier and recent experimental research on this subject and the amount of work already done in the area of experimental research dedicated to animal bones.

Back to the north of Portugal Detry, Cardoso & Correia study the food refuse remains deployed near a roman amphitheater of the well-known roman town of Conimbriga. The remains studied are from roman and mostly medieval levels. Also in Coimbra, but from modern period Santa Clara-a-Velha monastery, Detry, Gambini & Corte-Real reveal the use of animals by 17th century nuns and how the animal remains could reveal more than written documents on the very high economic status of the people that produced this waste.

Following that we have two articles on the Muge shellmidens, one from A. Pereira summing the Zooarchaeology research on these sites and Dias & Detry on the role of rabbits on past human populations. The first is a synthesis, and the second analyses the use of small mammals such as rabbit and compares them with the nutritional value of other species.

Masseti & Veracini, explain how the Neotropical mammals circulated after the discoveries in the modern period, in a time were the globalization started to influence animal commerce and transportation.

M. J. Valente studied the Mesolithic and Neolithic sites in Western Algarve with an economic, ecological and Taphonomic detailed analysis of these sites, sheading some light on the shell middens of these geographical area comparatively much less studied than the Tagus’ and Sado’s Rivers ones.

O. Paulo shows one of the most expressive areas in the presentations: genetics applied to the study of animals in the past and the movements of these animals namely by the influence of humans. In this case, he studies the introduction of the chameleon in the Iberian Peninsula.

Costa & Baptista as well as Porfírio & Serra, papers analyze animal remains from the Bronze age of Southern Alentejo in sites with similar contexts, that have human and animal burials, animal limbs and scattered remains in halls or silos. Costa & Baptista studied the event in terms of Taphonomic aspects and Porfírio & Serra describe the archaeological context in more detail, reflecting these authors' background.

Finally an important study by Almeida & Costa on the details of the rise in researchers’ and students’ dedicated to archaeofaunal studies in Portugal, demonstrating that zooarchaeology is still a young scientific area in Portugal, which developed enormously over the past few years. Acknowledgments

After these, we look at articles that study animal remains from historical archaeological sites in Portugal.

Great thanks to UNIARQ (Archaeology Center at the University of Lisbon), Faculty of Letters of the University of Lisbon and NAP (Univ. of Algarve), for supporting this event. Further thanks to members of the DGPC (formerly IGESPAR) Archaeosciences Lab, particularly Simon Davis, Sónia Gabriel and Carlos Pimenta. Funding was provided by the Foundation for Science and Technology (FCT) through FACC program. Finally our sincerest gratitude to the participants and volunteers that helped this event happen.

Valenzuela studied the mammal remains from the excavations at Casa do Governador and Núcleo Arqueológico at Rua dos Correeiros (NARQ-BCP, Lisbon) both in Lisbon, from levels dated of the 1st to 5th century AD. Both assemblages were created by fish salting factories but in the abandonment phase. The author also makes interesting comparisons with other roman sites from southern Iberia. M. Nabais studies the mammals and birds from Tróia, also from a context of a fish-salting workshop, between the 1st and 5th cent AD.

References

A. Gonçalves an archaeologist that excavated a well in the roman town of Odrinhas, were there is a Museum dedicated to this period, described the animal remains and mainly its archaeological context, more profound zooarchaeological and biological study these remains is still lacking. The assemblage is very interesting not constituting a normal food refuse accumulation but probably having some ritual implication.

Antunes, M. T. (1985) – Sciurus vulgaris no Cabeço da Arruda, Muge. Presença e extinção em Portugal. Arqueologia. 12: 71-84. Antunes, M. T. (1987) – O Povoado fortificado do Calcolítico do Monte da Tumba. IV – Mamíferos (Nota preliminar). Setúbal Arqueológica. 8: 103-144. 2

Cardoso, J. L. (1993) – Contribuição para o conhecimento dos grandes mamíferos do Plistocénico Superior de Portugal. Tese de Doutoramento. Universidade Nova de Lisboa. Oeiras: Câmara Municipal de Oeiras. Davis, S. & Moreno-Garcia, M. (2007) - Of metapodials, measurements and music – eight years of miscellaneous zooarchaeological discoveries at the IPA, Lisbon. O Arqueólogo Português, 25(4): 9-165. Delgado, M. J. (1884) – La grotte de Furninha a Peniche. IX ème session Congrès International d’Anthropologie et Archéologie Préhistorique (Lisboa 1880), Lisboa, pp. 207-278. Harlé, E. (1910/11) – Les mammifères et oiseaux quaternaires connus jusqu’ici en Portugal. Comunicações dos Serviços Geológicos de Portugal. 8: 22-85. Moreno-garcía, M.; Davis, S. & Pimenta, C. (2003) – Arqueozoologia: estudo da fauna no passado. In J. E. Mateus & M. Moreno-García (eds.), Paleoecologia Humana e Arqueociências. Um programa multidisciplinar para a Arqueologia sob a tutela da Cultura, pp. 191-234. Trabalhos de Arqueologia. Pereira da Costa, F. A. P. da (1865) – Da existência do Homem em epochas remotas no valle do Tejo. Lisboa: Imprensa Nacional.

3

Exploitation of bone and antler in the Upper Palaeolithic of Portugal Marina Almeida Évora1,2 1 2

Núcleo de Arqueologia e Paleoecologia – FCHS, Universidade do Algarve Fundação para a Ciência e a Tecnologia / UNIARQ – Centro de Arqueologia da Universidade de Lisboa

Abstract In the Upper Paleolithic contexts where organic remains were preserved, we see that the artisans of this period used all the hard animal materials that were available (bone, antler, tooth, ivory, tortoise-shell, shell, egg shell, horn, etc.) for the manufacture of utilitarian or / and symbolic objects. The choice and the differentiated use for each of these materials appears to have been directly related with the cognitive capacity of these populations and their subsequent recognition of the different mechanical and aesthetic properties of each raw material. The analysis of faunal collections, in addition to providing information about the means of subsistence of these communities, also enables us to infer about the procedures of exploitation of osseous materials and its use in the manufacture of tools. The production of the osseous artefacts may have had different procedures that, as will be seen, can be inferred by the characteristic marks left on the bone surface. We present here the techniques of transformation and modification of the red deer antler and of mammal bones, since these were the raw materials used in the Upper Paleolithic in Portugal for the manufacture of tools like wedges, awls, bevelled tools, fishhooks or spear heads. Keywords: Upper Paleolithic; osseous industry, processing, modification Introduction To undertake the technological analysis of the bone tools assemblages it is necessary to collect all bone fragments, even the smallest ones, for it can hold specific stigmata that permit to identify a specific fracturing or shaping technique.

In Portugal, the Upper Paleolithic bone tools assemblages, that are known to date, came from 12 archaeological sites, maily located in Estremadura (Buraca Grande, Lagar Velho Rockshelter, Lapa do Picareiro, Lapa dos Coelhos, Caldeirão Cave, Casa da Moura Cave, Lapa Furada, Bocas Rockshelter, Lapa da Rainha and Salemas Cave), in Alto Alentejo (Escoural Cave), and in Western Algarve (Vale Boi) (Évora, 2008). The assemblage’s composition is not homogenous, that is, the older assemblages have less quantities of artefacts, with no debris reported, then the more recent ones, with several categories of tools.

Exploitation of bone and antler Mechanical properties In Upper Paleolithic (UP) of Portugal, the hard animal materials, known until now, used in the production of bone tools were mammal bone and Red deer antler. These raw materials were known by their mechanical properties, in the sense that, by using the two raw materials Paleolithic artisans, in their dairy life, would accumulate the knowledge concerning the utility of the two in a diversity of purposes (Knecht, 1991). Red deer antler is composed by 60% of minerals and 40% of organic compounds. Its mechanical properties vary for fresh antler (which is obtained with the hunted animal) or shed antler (which has fallen off the stag naturally) and depends also on the animal’s age, on the environmental and living conditions to which the animal was exposed and the conditions the antler suffered after it was shed (Goutas, 2004). Also, antler has larger dimensions than some types of mammal bones, but has limits on manufacturing methods due to its curvature. Due to its large elasticity, has greater resistance to fracture and impact than mammal bone (Otte, 1974), because it absorbs the energy from direct impacts and thereby becomes more resistant to fractures

In the last century, the archaeologists who reported the presence of bone tools in the archaeological record of Upper Paleolithic sites presented their analysis always based on typological descriptions and giving importance to the technological analysis of lithic assemblages (Cardoso e Gomes, 1994; Gomes, Cardoso and Santos,1990; Salvado, 2004). The exceptions were Aubry and Moura (1994), and Chauviere (2002) when publishing the bone tools assemblages from Buraca Grande and Caldeirão Cave, respectively. One of the reasons for that to happen could be the bad preservation of some faunal assemblages, making necessary to deal with the preservation of a few bone tools and debitage debris or even with the total absence of them. This low frequency of bone tools in the UP faunal record could be due to taphonomic alterations or, most likely, with the excavations techniques, housing and selection of bone fragments still in the field and later in the laboratory (Évora, 2008; Marks et al, 1994). 5

Figure 1. Fracturing mammal long bone by direct percussion (experimental work, photo TECHNOS 2010).

Figure 2. Sectioning antler by direct percussion (experimental work, photo TECHNOS 2010).

(Goutas, 2005). Antler’s acquisition may have been dependent on Red deer hunting or on seasonality, as they are cast and regenerated annually (Knecht 1991). On the other hand, mammal bone combines approximately 1/3 of organic matter (collagen) and 2/3 of inorganic matter in its composition giving it hardness, stiffness, elasticity and resilience (Brothwell 1981; Davis 1987, Évora, in press). For this reason it would be a preferred raw material in the production of several kinds of tools and its compacta was most exploited by huntergatherer groups to manufacture their hunting and fishing toolkits (Évora, 2008). The mammal long bone is strong, has elasticity and is harder than antler, but it more easily fractures (Knecht, 1991) because it has a higher degree of mineralization than antler. It has a fibrous structure due to the presence of collagen which gives it great strength facing tractions and pushups, but it also causes naturally pointed fractures. Its hardness is due to the presence of minerals in its composition. The diversity of forms offered by bone anatomy of the hunted animals allows a wide choice of materials of different sizes and shapes (Otte, 1974). Bone would be easy to acquire since huntergatherers could choose to collect certain kinds of bone fragments after the animal carcasses were butchered. The fractured bone flakes could be transformed into a variety of shapes and sizes, until the desired tool was produced.

consequence of animals, sediments weight, ice… This technique is the most elementary and ancient one, but is also the most uncontrollable one because it is not possible to predict the size of the blanks in the case of mammal bones. And it poses the problem of recognizing if the fracturing of the bones is due to marrow extraction only or if it is to manufacture tools too. In the case of antler, the fracturing is a debitage technique, because antlers were not consumed, they were only used to produce tools (fig.1). b) Sectioning By sectioning we mean a way to obtain a support with its full section in which part of its morphology is intact, this technique can be applied transversally to the long axis of the bone or antler, to separate the antler tines and creating a support. In mammal bones it can be used to separate the epiphyses from the diaphyses, thus creating a support. This technique can be applied by direct percussion, by sawing and flexion or by combining them (figs.2, 3, 4, and 5). c) Longitudinal grooving and bipartition By this we mean creating a longitudinal groove parallel to the long axis of the bone with a lithic tool, making a uni / bidirectional movement leaving a cross-section in U or V, depending on the active edge of the lithic tool, when the groove is deep enough an intermediate tool (made of bone or antler with a bevel as an active part (fig.6) or a lithic splintered piece) is used to help separate both parts of the bone or antler (figs.7 and 8).

Fracturation techniques Until now there are few fracturing and shaping techniques recognized on the archeological record of the UP in Portugal and all were used along this time period. These techniques could have been combine between them, and also may have been with other fracturing techniques that we don’t have record of, because the shaping process that took place after fracturing erased those stigmata.

Shaping a) Scraping

a) Fracturing

By scraping we mean the action of using of a lithic tool, which can be retouched or nonretouched, in the surface of

By fracturing we mean the anthropic action of breaking a bone or antler, by opposition to fragmentation which is a 6

Figure 3. Antler debris and stigmata resulted from sectioning by direct percussion (experimental work, photo TECHNOS 2010).

Figure 4. Sectioning mammal long bone by sawing (experimental work, photo TECHNOS 2010).

Figure 5. Antler tine sectioned by sawing and flexion (Vale Boi gravettian artefact).

Figure 7. Bipartition on a Red deer metacarpus, stigmata left by a wedge (Vale Boi gravettian artefact).

Figure 6. Distal unibevel made on antler tine sectioned by direct percussion (Lapa do Picareiro magdalenian artefact).

Figure 8. Longitudinal grooving (experimental work). 7

Figure 9. Stria made by scraping with an unretouched lithic tool (Buraca Grande gravettian artefact, 20x).

Figure 10. Stria made by scraping with a retouched lithic tool (Buraca Grande gravettian artefact, 20x).

Figure 11. Stria left by an abrasive (Lapa dos Coelhos gravettian artefact, 20x). Figure 12. Diagonal incisions (Buraca Grande magdalenian artefact, 20x).

the bone tool to remove portions of the material until obtaining the desired shape. Depending on the active part of the lithic tool, different pattern of strias will be left on the bone surface (figs.9 and 10).

Conclusion In the Upper Palaeolithic archaeological sites of Portugal where bone tools were recovered, we find that, until now, the main raw materials used were Red deer antler and mammal bones. This fact is related to food activities where mammal bones were available to use after the butchering of the animal carcasses and their fracturing to extract bone marrow, and antler could be gathered in the wild or obtained when hunting Red deer. Both were used to manufacture several utensils, but their mechanical properties were somehow known by the artisans to produce specific tools, as projectile points and wedges were preferably made of antler which is more suitable to resist to direct impacts, and awls were made of mammal bone that is harder than antler. Both antler and bone were suitable to manufacture spear heads and that is seen on the Portuguese UP record.

b) Abrasion By abrasion we mean the action of eliminate small particles of material by rubbing it on an abrasive, which can be a soft stone, leather, sand, sand and water or sand and leather, and according to the abrasive that was used different stria will be left on the bone surface. This technique was usually applied after scraping and overlaps on to it (fig.11). c) Incision By incision we mean a line that is rectilinear, transversal or diagonal to the longitudinal axis of the artifact, made on the bone surface by a lithic tool, usually its crosssection is V or |/ and it’s not as deep as a groove (fig.12). They are present in several types of tools, like spear heads (decorated or not), appearing in sets of lines that have been interpreted as having a functional or decorative role, or both, depending on the tool.

These techniques of fracturing and shaping were continuous along the UP in Portugal, all of them leaving specific stigmata and stria on the bone and antler artefacts surfaces. These can be recognized in the faunal record, which is why all bone fragments should be recovered during excavation, some are small fragments but can still preserve the identification marks. 8

Abbaye de Senanque, Editions de l’Université de Provence, pp. 120-133. Salvado, M. C. (2004) - Apontamentos sobre a utilização do osso no Neolítico e Calcolítico da Península de Lisboa. As colecções do Museu Nacional de Arqueologia, M.A. Dissertation, Faculdade de Letras da Universidade de Lisboa, suplemento nº 2 a O Arqueólogo Português, Lisboa.

Acknowledgments I want to thank all the Institutions and persons who gave me the opportunity to study the bone tools assemblage from the 12 Upper Paleolithic archaeological sites: Nuno Bicho, J. L. Cardoso, T. Aubry, F. Almeida, L. Raposo (Museu Nacional de Arqueologia), M. M. Ramalho (Museu Geológico), J. Zilhão and M. V. Gomes. References Gomes, M. V., Cardoso, J. L., Santos, M. F. (1990) Artefactos do Paleolítico superior da Gruta do Escoural (Montemor-o-Novo), Almansor, 9: 15-36. Aubry, T., Moura, M. H. (1994) - Paleolítico da Serra de Sicó, Separata das Actas dos Trabalhos de Antropologia e Etnologia, vol. XXXIV – Fasc. 3-4: 43-60. Bicho, N., Stiner, M., Lindly, J. (2004) - Shell ornaments, bone tools and long distance connections in the Upper Paleolithic of Southern Portugal, La Spiritualité, M. Otte (ed.) ERAUL 106, Liège, pp. 71-80 Brothwell, D. R. (1981) - Digging up bones. The excavations, treatment and study of human skeletal remains. British Museum of Natural History, Cornell University Press, Ithaca, New York, 3ª edicão. Cardoso, J. L., Gomes, M. V., 1994. Zagaias do Paleolítico Superior de Portugal, Portvgalia, Nova Série, vol. XV: 07-31. Chauviere, F. X. (2002) - Industries et parures sur matières dures animales du Paléolithique Supérieur de la grotte de Caldeirão (Tomar, Portugal), Revista Portuguesa de Arqueologia, vol. 5(1): 05-28. Davis, S. J. M. (1987) - The archaeology of animals, Yale University Press, New Haven and London. Évora, M. 2008 - Artefactos em haste e em osso do Paleolítico Superior Português, Promontória, Ano 6 (6): 9-50. Évora, M. A. (in press) - Raw material used in the manufacture of osseous artefacts during the Upper Paleolithic in Portugal. From These Bare Bones: Raw materials and the study of worked osseous materials, edited by A. M. Choyke and S. O’Connor, Oxbow Books, Oxford, UK. Goutas, N. (2005) - Caractérisation et Evolution du Gravettien en France par l’approche techno-économique des industries en matières dures animales (étude de six gisements du Sud-Ouest), Ph.D. Thesis, Université de Paris I – Pantheon – Sorbonne, 2 vols. Knecht, H. D. (1991) - Technological innovation and design during the Early Upper Paleolithic: A study of organic projectile technologies, Ph.D. Thesis, New York University, UMI Dissertation Services, Michigan, USA Marks, A. E., Bicho, N., Zilhão, J., Ferring, C. R. (1994) - Upper Pleistocene prehistory in Portuguese Estremadura: results of preliminary research, Journal of Field Archaeology, vol. 21(1) - 53-68. Otte, M. (1974) - Caractéristiques inhérentes à l’analyse par attributs de l’outillage osseaux, Premier Colloque International sur L’Industrie de l’os dans la PréHistoire, 9

10

So many rabbits! Small animal’s dietary role in the Mesolithic shellmidden of Cabeço da Amoreira (Muge, Central Portugal) – preliminary results Rita Dias1 & Cleia Detry2 1 2

Núcleo de Arqueologia e Paleoecologia – FCHS, Universidade do Algarve UNIARQ – Centro de Arqueologia da Universidade de Lisboa

Abstract Studies about Palaeolithic subsistence normally refer to large game hunting as primary diet source for hunter-gatherer populations. During the Mesolithic, the general idea points to an intensification of aquatic resources consumption. Nevertheless, recent investigations suggest that not only there is a considerable regional variability in hunter-gatherer subsistence but aquatic resources have been consistently consumed from well before. Following the same idea, there has been a devaluation of small animals’ exploitation, not only in the Paleolithic but also in the Mesolithic. Recent theories, such as Nutritional Ecology, have valued the role of small animals in human diet, such as rabbits and hares. This paper aims to discuss the role of rabbit consumption in the diet during the Mesolithic, based on new data from the Cabeço da Amoreira Mesolithic shellmidden. Keywords: Small Game; Subsistence; Cabeço da Amoreira research, it’s well-documented that that major changes on hunter-gatherers ecology do not only occur with the beginning of the Mesolithic, but instead, this new technocomplex continued the changes launched in the Late Upper Paleolithic record, more so in areas where this alterations occurred earlier and slowly, like Southern Europe. In fact, the emergence of the Mesolithic in SW Iberia is related with the beginning of the Holocene, around 10.300 Years BP, when the post-glaciar climatic conditions are in full (Araújo, 1998, 2003; Bicho et al., 2010; Bicho, Pereira, Umbelino, et al., in press).

Environmental and climatic background It is well known that environmental and climatic changes during the Pleistocene-Holocene transition made significant shifts on biomass and territories occupied by prehistoric hunter-gatherers (Clark and Straus, 1986; Bicho, 1999, 2009; Schriek, 2004; Schriek et al., 2007; Bicho and Haws, 2008; Harrison and Sanchez Goñi, 2010). In Southern Europe these climatic oscillations began with the onset of the Alleröd climatic phase, when part of Europe was already re-forested (Naughton et al., 2007; Fletcher et al., 2010). This means that human adaptations to the new climatic conditions also started well before, still in the Late Upper Palaeolithic (Jochim, 2011). By the end of Dryas III (11,5 ka BP), these climatic oscillations were characterized by a fast global temperature rise, that led to the polar caps retreat and therefore to a significant rise of mean sea level (Bicho et al., 2011, Jochim, 2011). Paleoenviromental data from Southwestern Europe was, until recently, very scarce, with the exception of some marine cores from the Portuguese coast (Bicho et al., 2011), and there are still very limited knowledge about the impact of Holocene climatic phenomenon and its variability. Nevertheless, although less dramatic than in the last glaciation cycle, the Holocene climatic changes were more intense than previously thought (Soares and Dias, 2006). In Southwestern Iberian Peninsula the beginning of the impact of Holocene climatic shifts are connected with the onset of Epipaleolithic techno-complex, at c. 10.500 Years BP, and result in significant changes regarding in human settlement patterns. The archaeological data shows that the number of sites increases and these occupations are associated with intensification in logistical mobility, which might probably display a significant demographic increase, and therefore profoundly influencing hunter-gatherer subsistence (Araújo, 2003; Umbelino, 2006; Detry, 2007; Bicho and Haws, 2008). In this scenario, during the last decade of

All these new climatic changes were sufficiently profound and broad to affect both humans and ecosystems, and therefore in human subsistence models (Mayewski et al., 2004). In this scenario, small game plays as an important role in the new human ecological models, since they are, particularly in the case of rabbit and hare, easily accessible, abundant and fast reproducing (Brambell, 1957; Poole, 1960; Flux, 1965). Resource intensification and subsistence: The importance of rabbit (Oryctolagus cuniculus) and Hare (Lepus sp.) in human diet Regarding to faunal remains in the archaeological record, the consistent numerous quantities of rabbit and hare remains reveal their importance in past human diet. This high rank in human subsistence does not resume to the Holocene period alone, but consistently increased during the Palaeolithic (Hockett and Bicho, 2000). The increase and broad diversity in animal consumption, such as lagomorphs, according to some authors, like Binford (Binford, 1968) and Burney & Flannery (Burney and Flannery, 2005) suggest that this resulted from the demographic growth and territorial circumscription 11

Rabbit

Rabbit facts at a glance gestation period

28–30 days

age at mating

3–4 months

litter size average

4–6 (up to 10)

breeding season

all year

time from giving birth to remating

½–2 hours

Nutrient

Unit

Value per 100 g.

Proximates

Table 1. Source: http://www.dpi.qld.gov.au/documents/Biosecurity_Envir onmentalPests/IPA-Rabbit-Control-In-QueenslandSection-2. together: Broad Spectrum Revolution model (BSR). Although used mainly to frame Pleistocene huntergatherer subsistence, the BSR model is still used by some researchers to consubstantiate later (Mesolithic) huntergatherer subsistence behavior. Earlier works (Binford, 1968; Flannery, 1969; Clark and Straus, 1986), and some recent ones (Araújo, 1998, 2003), described the Pleistocene-Holocene transition as a period of resource intensification, diversification and specialization, scenario triggered by imbalances in the populationresources balance, mostly consequence of considerable human population growth and post-glacial climatic changes (Hockett and Haws, 2009). Some of these researches suggest that changes in the type of small animals consumed were related with the balance between population and resources from much earlier, across the Middle-Upper Paleolithic transition in southern Europe, when shifts in different types of small game hunted were directly correlated with population pulses: quicker, fast reproducing animals, like rabbits, gained importance as slower, slow-reproductive animals, like shellfish, were smaller and less abundant due to population pressures (Stiner, 1999, 2001; Stiner and Munro, 2002). This linked BSR to an evolutionary ecology framework connecting diet choices to human demography phenomenon (Hockett and Haws, 2009). Recent research correlates human demography and diet from a Nutrional Ecology point of view (Hockett and Haws, 2003, 2009), taking into account that there were regional variations in the role of small game in subsistence.

Water

G

61.37

Energy

Kcal

Protein

G

33.02

Total lipid (fat)

G

3.51

Carbohydrate, by difference Fiber, total dietary

G

0

G

0

173

Calcium, Ca

mg

18

Iron, Fe

mg

4.85

Magnesium, Mg

mg

31

Phosphorus, P

mg

240

Potassium, K

mg

343

Sodium, Na

mg

45

Zinc, Zn

mg

2.38

Vitamin C, total ascorbic acid Thiamin

mg mg

0.02

Riboflavin

mg

0.07

Niacin

mg

6.4

Vitamin B-6

0.34

Folate, DFE

8

Vitamin B-12

6.51

Vitamin E (alphatocopherol) Vitamin D (D2 + D3)

0.41 0

Vitamin D

0

Vitamin K (phylloquinone)

1.5

Vitamin A, R A E

mcg_RAE

0

Vitamin A, IU

IU

0

Lipids

That being said, it is of course a known fact that behaviors that improve intake of important nutrients will improve childhood survivorship and brain development, rabbit meat, not only provide those critically important nutrients, but actually have a good ratio of those for 100 grams of its cooked meet, not falling behind (actually overcoming it) the same amount of red deer cooked meet (see table 2 and 3). Although daily requirements for protein are quite modest, they are, nevertheless, constant, especially for children and pregnant woman. Rabbits and hares are, as previously referred easy to catch and

Fatty acids, total saturated

G

1.05

Fatty acids, total monounsaturated Fatty acids, total polyunsaturated Cholesterol

G

0.95

G

0.68

mg

123

Table 2. U.S. Nutrients present per 100g of Rabbit meat. Department of Agriculture, Agricultural Research Service. USDA National Nutrient Database for Standard Reference, Release 22. Nutrient Data Laboratory Home Page:http://www.nal.usda.gov/fnic/foodcomp/search/ …under a PD license. 12

available all year round, not requiring a huge outlay of available energy, not implying any risks, and catching them could easily be done by children, and pregnant woman.

Red Deer Nutrient

Unit

Value per 100 g.

Proximates Water

G

Energy

Kcal

Protein

G

30.21

Total lipid (fat)

G

3.19

Carbohydrate, by difference Fiber, total dietary

G

0

G

0

Calcium, Ca

mg

7

Iron, Fe

mg

4.47

Magnesium, Mg

mg

24

Phosporus, P

mg

226

Potassium, K

mg

335

Sodium, Na

mg

54

Zinc, Zn

mg

2.75

Vitamin C, total ascorbic acid Thiamin

mg

0

mg

0.18

Riboflavin

mg

0.6

Niacin

mg

6.71

Vitamin A, R A E

mcg_RAE

0

Vitamin A, IU

IU

0

One would except that that would be a greater variety in Mesolithic subsistence. Albeit that variety, small game role, namely rabbits, does not decrease or dilute in that reality, but grows in importance during the Final Pleistocene and Early Holocene (Haws, 2003; Bicho and Haws, 2008; Hockett and Haws, 2009; Bicho et al., 2011). In Southwestern Iberian Peninsula the Muge shellmiddens, are in that context, not an exception. Actually, rabbit remains are, has archaeological data shows, largely superior in number to every other mammal remains. This paper focuses on those remains as a mean to confirm the great importance of that resource in Mesolithic hunter-gatherer diet.

65.23 158

Minerals

The Muge Shell Middens – the case of Cabeço da Amoreira In the archaeological site of Cabeço da Amoreira the Oryctolagus cuniculus remains represent more than 90% of the total lagomorphs recovered from old excavations, actually mildly increasing its representation on the older layers (Detry, 2007). This irrefragably shows that rabbits are not intrusive, but an intensely explored resource, in fact, if they were a high number of intrusive elements, one would expect to find a large number of complete, well preserved skeletons or, at least, a very high number of skeleton parts by individual, which in fact it is not verifiable. The increase of the number of Lepus sp. on the more recent archaeological levels could mean an over exploitation of Oryctolagus cuniculus, that lead to the increase of hare procurement. Nevertheless, the progressive and very slight increase of its representation does not allow space for definitive explanations. Also, both MNI (Minimum Number of Individuals) and NISP (Number of Identified Specimens) confirm the same tendency (Detry, 2007). The very slight difference between the number and type of Lagomorph remains in the several levels means that there has been a similar manipulation of this resource throughout the occupation of this site, which recent data shows to have been over only around 200 years (Table 4) (Bicho et al., in press).

Vitamins

Lipids Fatty acids, total saturated Fatty acids, total monounsaturated Fatty acids, total polyunsaturated Cholestrol

G

1.25

G

0.88

G

0.62

mg

112

Different exploitation dynamics of rabbit and hare between the several Muge shellmiddens is not clear. Lagomorph remains are numerous in all shellmiddens, showing that it was consistently explored as a food resource. Rabbit distribution on all of them is very similar to that of wild boar and red deer. Albeit that fact, measurements of humeri distal breath on Oryctolagus cuniculus remains from several of these shellmiddens show some differences (fig. 3) (Detry, 2007). These differences are not sufficient to present definitive interpretations, but could suggest a more intensive exploitation of rabbits at Cabeço da Amoreira putting

Table 3. Nutrients present per 100g of Red Deer meat.Source: U.S. Department of Agriculture, Agricultural Research Service. USDA National Nutrient Database for Standard Reference, Release 22. Nutrient Data Laboratory Home Page:http://www.nal.usda.gov/fnic/foodcomp/sear ch/…under a PD license. 13

complex reality were there are a variety of sources in subsistence – terrestrial (small and large game) and aquatic (table 5). As discussed, lagomorphs are permanent resources that do not imply risks or considerable energy loss to capture, have a very rapid life cycle and provide a good intake of essential nutrients, therefore making them a very attractive resource, as shown by the number of its remains in almost all prehistoric sites raging from Middle Paleolithic to Mesolithic periods.

5000 4000 3000 2000 1000 0

Superficial Medium layer layer

Profound layer

Total

Most of the Cabeço da Amoreira data here discussed are from old excavations having, therefore, some limitations in terms of its stratigraphic origin. Nonetheless, data from the new excavations confirms the major importance of rabbits for the hunter-gatherer communities of Muge, and has the potential to further inform us about their importance, through spatial analysis since all complete anatomical elements, fragments of elements bigger than 2 cm are piece plotted, as well as materials resulting from sieving, since all buckets are also piece plotted, making it possible to have, at least, a volumetric information of its origin.

Figure 1. Number of Identified Specimens (NISP) of Oryctolagus cuniculus recoverd at Cabeço da Amoreira, based on Detry (2007).

1000 800

600 400

200 0

Superficial Medium layer layer

Profound layer

Seasonality, on the contrary, is harder to infer about, since, as previously said, rabbit’s life cycle is extremely rapid. Also, although most of the rabbit remains belong to an adult, which has to be examined with care because more fragile juvenile bones can be underrepresented.

Total

Figure 2. Number of Identified Specimens (NISP) of Cervus elaphus recoverd at Cabeço da Amoreira, based on Detry (2007).

However further study of the remains from recent excavations is still needed it is clear that Rabbits consumption resulted from their availability and ease of capture and not from stressed hunters hunting prey they would otherwise ignore - although, undoubtedly, biomass, territory and demography changed.

100 95

90 85 80

References

75

Araújo, A.C. (1998) - O concheiro de Toledo (Lourinhã) no quadro das adaptações humanas do Pós-Glaciar no litoral da Estremadura. Revista portuguesa de Arqueologia, 1: 19–38. Araújo, A.C. (2003) - Long term change in Portuguese early Holocene settlement and subsistence, in: Larsson, L., Kindgren, H., Knutsson, K., Loeffer, D., Akerlund, A. (Eds.), Mesolithic on the Move. Oxbow Books, Oxford, pp. 569–580. Bicho, N. (1999) - Portugal 10,000 years ago: Human Ecology at the End of the Pleistocene, in: World Archaeological Congress 4. University of Cape Town 10th-14th January 1999. University of Cape Town, Cape Town. Bicho, N. (2009) - Sistemas de povoamento, subsistência e relações sociais dos últimos caçadores-recolectores do Vale do Tejo. Estudos Arqueológicos de Oeiras, 17: 133– 156.

70 C. Am.

C. Ar.

M. S.

Mod-IP

Figure 3. Confidence Intervals (95%) Oryctolagus cuniculus of the distal Breadth (Bd) of the humerus (Detry, 2007). C. Am. – Cabeço da Amoreira; C. Ar. – Cabeço da Arruda; M.S. – Moita do Sebastião; Mod-IP – Modern reference skeletons from the Iberian Peninsula. sufficient pressure on the resource to result in size differences (Davis, 1981; Davis and Moreno-García, 2007). Discussion and ongoing issues Data shows that rabbits cannot be considered only as a complementary food resource but a part of a more 14

15

Area José Rolão José Rolão José Rolão José Rolão José Rolão Shellmdidden Shellmdidden Shellmdidden West cut West cut 2 North cut West cut West cut West cut West cut West cut West cut West cut West cut West cut West cut West cut West cut West cut West cut West cut West cut West cut 2 North cut West cut West cut West cut Layer 2 Jean Roche

Code Wk-26796 TO-10225 Sac-2023 Sac-2080 Sac-2079 Wk-26797 Wk-26798 Wk-31142 Wk-28050 Wk-30672 Wk-30674 Wk-28049 Wk-28048 Wk-28047 Wk-28046 Wk-28045 Wk-28044 Wk-28043 Wk-28042 Wk-28041 Wk-28040 Wk-28039 Wk-28038 Wk-28037 Wk-28036 Wk-28035 Wk-28034 Wk-30671 Wk-30673 Wk-28033 UAGMS-7197 UGAMS-7196 Wk-32143 TO-11819-R

Material Homo Homo Shell Shell Shell C. edule C. edule Homo S. plana S. plana S. plana S. plana S. plana S. plana S. plana S. plana S. plana S. plana S. plana S. plana S. plana S. plana S. plana S. plana S. plana S. plana S. plana S. plana S. plana S. plana S. plana Coal Homo Homo

Date BP 6329±40 6550±70 7260±60 7080±80 7050±45 7291±35 7145±377 6910±40 7377±33 7360±34 7356±33 7193±33 7445±33 7376±34 7368±39 7315±35 7311±34 7273±34 7323±48 7384±48 7305±48 7395±48 7365±49 7307±48 7251±48 7395±48 7370±48 7417±37 7406±32 7479±48 7450±30 6990±30 7132±41 7300±80

Cal BP 2σ 6847-7146 7329-7570 7460-7740 7270-7590 7300-7540 7510-7730 7410-7590 7425-7570 7591-7819 7574-7805 7570-7800 7435-7631 7658-7898 7589-7819 7579-7816 7548-7766 7545-7761 7499-7710 7537-7792 7580-7841 7514-7770 7588-7854 7569-7826 7517-7772 7463-7704 7588-7854 7573-7828 7620-7865 7610-7850 7668-7929 7663-7899 7735-7930 7590-7785 7690-8010

Cal BC 1σ 5069-4936 5611-5471 5714-5581 5591-5428 5534-5415 5715-5620 5602-5504 5600-5495 5812-5694 5791-5672 5784-5668 5632-5536 5878-5752 5812-5692 5804-5680 5735-5631 5730-5630 5709-5608 5754-5631 5830-5694 5735-5619 5838-5703 5806-5670 5737-5620 5702-5585 5838-5703 5812-5676 5845-5725 5835-5718 5924-5777 5882-5759 5971-5841 5775-5670 6005-5844

Cal BC 2σ 5197-4898 5623-5374 5789-5508 5643-5323 5593-5354 5784-5565 5642-5461 5620-5475 5870-5642 5852-5627 5846-5624 5682-5486 5949-5709 5870-5640 5867-5630 5817-5599 5812-5596 5771-5544 5843-5588 5892-5631 5821-5565 5905-5639 5877-5620 5823-5568 5755-5514 5905-5639 5879-5624 5912-5672 5897-5664 5980-5719 5950-5714 5981-5786 5835-5640 6064-5742

Context Cairn Cairn Cairn Cairn Cairn Cairn Cairn Shellmidden Shellmidden Shellmidden Shellmidden Shellmidden Shellmidden Shellmidden Shellmidden Shellmidden Shellmidden Shellmidden Shellmidden Shellmidden Shellmidden Shellmidden Shellmidden Shellmidden Shellmidden Shellmidden Shellmidden Shellmidden Shellmidden Shellmidden under under under under

Reference Bicho et al. Rocksandic Martins et al. Martins et al. Martins et al. Bicho et al. Bicho et al. Not published Bicho et al. Bicho et al. Bicho et al. Bicho et al. Bicho et al. Bicho et al. Bicho et al. Bicho et al. Bicho et al. Bicho et al. Bicho et al. Bicho et al. Bicho et al. Bicho et al. Bicho et al. Bicho et al. Bicho et al. Bicho et al. Bicho et al. Bicho et al. Bicho et al. Bicho et al. Bicho et al. Bicho et al. Not published Rocksandic

Table 4. Radiocarbon dates from Cabeço da Amoreira (short life samples with known stratigraphic provenance). Calibration wiith Calib, 6.0 software (Stuiver et al., 2005) with use of IntCal09 (Reimer, et al., 2009). Marine reservoir correction of 0±40 based on Soares (Soares, 2005; Soares and Dias, 2006) and Martins et al. (2008).

Stratigraphic origin Multiple burial CAM-01-01 Multiple burial CAM-01-01 Layer 3(=Layer 1b) Layer 3(=Layer 1b) Layer 3(=Layer 1b) Layer 1b Layer 1b Layer 2b Layer B Layer 22 Layer 22 Layer C Layer D Layer F Layer G Layer H Layer H Layer I Layer J Layer K Layer K Layer L Layer M Layer N Layer O Layer P Layer Q Layer 2 Layer 2 Level 1 Main Pit Hearth 1 Burial 2 - 2011 Burial CAM-00-01

Site

Provenance

Cal BP 2σ

Cabeço da Amoreira Cabeço da Amoreira Cabeço da Amoreira Cabeço da Amoreira Cabeço da Amoreira Cabeço da Amoreira Cabeço da Amoreira Cabeço da Amoreira

?

76908010 74307560 71807430 73207570

Esqueleto 7 Base Níveis 2 e 3 (topo) Esqueleto 7

Age

Esqueleto 6

Adult (indeterminate) Adult (indeterminate) Middle-aged adult

Esqueleto 8

Middle-aged adult

Esqueleto 4

δ13C (‰) -16.3

δ15N (‰) -

%Marine diet

-16.5

11.9

48

-17.1

-

42

-20.1 -21.8 -16.5

8.2 4.9 11.9

0 0 50

-15.7

12.7

59

-14.8

12.5

69

-15.6

12

60

50

Table 5. Isotope results for several human remains from Cabeço da Amoreira, based on Umbelino, 2006.

Clark, G.A., Straus, L.G. (1986) - Synthesis and conclusions Part I: Upper Paleolithic and Mesolithic huntergatherer subsistence innorthern Spain, in: La Riera Cave: Stone Age Hunter-Gatherer Adaptations in Northern Spain. ASU Anthropological Research Papers, 36. Arizona State University, Arizona, pp. 351–266. Davis, S., Moreno-García, M. (2007) - Of metapodials, measurements and music-eight years of miscellaneous zooarchaeological discoveries at the IPA, Lisbon. O Arqueólogo Português, (IV) IV: 9–165. Davis, S.J.M. (1981) - The effects of temperature change and domestication on the body size of Late Pleistocene to Holocene mammals of Israel. Paleobiology, 7: 101– 114. Detry, C. (2007) - Paleoecologia e Paleoeconomia do Baixo Tejo no Mesolítico Final : O contributo do estudo dos mamíferos dos concheiros de Muge Paleoecologia e Paleoeconomia do Baixo Tejo no Mesolítico Final . PhD Thesis. University of Salamanca. FLUX, J.E.C. (1965) - Timing of the breeding season in the hare, Lepus europaeus pallas, and rabbit, oryctolagus cuniculus (l.). Mammalia, 29: 557–562. Flannery, K.V. (1969) - The Domestication and Exploitation of Plants and Animals, in: Ucko, P.J., Dimbleby, J.W. (Eds.), The Domestication and Exploitation of Plants and Animals. Chicago, pp. 77–100. Fletcher, W.J., Sánchez Goñi, M.F., Allen, J.R.M., Cheddadi, R., Combourieu-Nebout, N., Huntley, B., Lawson, I., Londeix, L., Magri, D., Margari, V., Müller, U.C., Naughton, F., Novenko, E., Roucoux, K., Tzedakis, P.C. (2010) - Millennial-scale variability during the last glacial in vegetation records from Europe. Quaternary Science Reviews, 29: 2839–2864.

Bicho, N., Haws, J. (2008) - At the land’s end: Marine resources and the importance of fluctuations in the coastline in the prehistoric hunter–gatherer economy of Portugal. Quaternary Science Reviews, 27: 2166–2175. Bicho, N., Haws, J., Almeida, F. (2011) - Hunter–gatherer adaptations and the Younger Dryas in central and southern Portugal. Quaternary International, 242: 336–347. Bicho, N., Pereira, T., Cascalheira, J., Marreiros, J., Gonçalves, C., Dias, R., n.d. - The chronology of the Mesolithic occupation of the Muge valley, central Portugal: the case of Cabeço da Amoreira. Quaternary International. Bicho, N., Pereira, T., Umbelino, C., Jesus, L., Marreiros, J., Cascalheira, J. Gonçalves, C., Detry, C., n.d. The construction of a shellmidden: the case of Cabeço da Amoreira, Muge (Portugal), in: The Eighth International Conference on the Mesolithic in Europe Will Be Held in Santander from 13th to 17th September 2010, Cantabrian International Institute for Prehistoric Research. Santander. Bicho, N., Umbelino, C., Detry, C., Pereira, T. (2010) - The Emergence of Muge Mesolithic Shell Middens in Central Portugal and the 8200 cal yr BP Cold Event. The Journal of Island and Coastal Archaeology, 5: 86–104. Binford, L.R. (1968) - Pos-Pleistocene adaptations, in: Binford, S.R., Binford, L.R. (Eds.), New Perspectives in Archaeology. Chicago, pp. 313–341. Brambell, F.W.R. (1957) - Reproduction of the wild rabbit, Oryctolagus cuniculus (L.). New Zealand Journal of Science and Technology B, 114: 1–45. Burney, D. a, Flannery, T.F. (2005) - Fifty millennia of catastrophic extinctions after human contact. Trends in ecology & evolution, 20: 395–401.

16

Harrison, S.P., Sanchez Goñi, M.F. (2010) - Global patterns of vegetation response to millennial-scale variability and rapid climate change during the last glacial period. Quaternary Science Reviews, 29: 2957–2980. Haws, J.A. (2003) - An Investigation of Late Upper Paleolithic and Epipaleolithic Hunter- Gatherer Subsistence and Settlement Patterns in Central Portugal. PhD Thesis. University of Wisconsin-Madison. Hockett, B., Haws, J. (2003) - Nutritional ecology and diachronic trends in Paleolithic diet and health. Evolutionary Anthropology: Issues, News, and Reviews, 12: 211–216. Hockett, B., Haws, J. (2009) - Continuity in animal resource diversity in the Late Pleistocene human diet of Central Portugal. Before Farming, 2: 1–14. Hockett, B.S., Bicho, N.F. (2000) - The Rabbits of Picareiro Cave: Small Mammal Hunting During the Late Upper Palaeolithic in the Portuguese Estremadura. Journal of Archaeological Science, 27: 715–723. Jochim, M., (2011) - The Mesolithic, in: Milisauskas, S. (Ed.), European Prehistory. A Survey. Springer, pp. 125– 152. Monge Soares, A.M., Alveirinho Dias, J.M. (2006) Coastal upwelling and radiocarbon-evidence for temporal fluctuations in ocean reservoir effect off Portugal during the holocene. Radiocarbon, 48: 45–60. Mayewski, P. a., Rohling, E.E., Curt Stager, J., Karlén, W., Maasch, K. a., David Meeker, L., Meyerson, E. a., Gasse, F., van Kreveld, S., Holmgren, K., Lee-Thorp, J., Rosqvist, G., Rack, F., Staubwasser, M., Schneider, R.R., Steig, E.J. (2004) - Holocene climate variability. Quaternary Research, 62: 243–255. Naughton, F., Sanchez Goñi, M.F., Desprat, S., Turon, J.L., Duprat, J., Malaizé, B., Joli, C., Cortijo, E., Drago, T., Freitas, M.C. (2007) - Present-day and past (last 25000 years) marine pollen signal off western Iberia. Marine Micropaleontology, 62: 91–114. Poole, W. (1960) - Breeding of the of wild rabbit, Oryctolagus cuniculus (L.), in relation to the environment. CSIRO Wildlife Research, 5: 21. Schriek, T. van der (2004) - Holocene envrionmental change and the alluvial geoarchaeology of Mesolithic settlement-subsistence in the Muge and Magos Valleys, lower Tagus basin, Portugal. PhD Thesis. University of Newcastle upon Tine. Schriek, T.V.D., Passmore, D.G., Anthony, C., Rolão, J. (2007) - The palaeogeography of Mesolithic settlementsubsistence and shell midden formation in the Muge valley, Lower Tagus Basin, Portugal. The Holocene, 3: 369– 385. Soares, A. (2005) - Variabilidade do “upwelling” costeiro durante o holocénico nas margens atlânticas ocidental e meridional da Península Ibérica. Phd Thesis. University of Algarve. Stiner, M.C. (1999) - Paleolithic Population Growth Pulses Evidenced by Small Animal Exploitation. Science, 283: 190–194.

Stiner, M.C. (2001) - Thirty years on the “broad spectrum revolution” and paleolithic demography. Proceedings of the National Academy of Sciences of the United States of America, 98: 6993–6996. Stiner, M.C., Munro, N.D. (2002) - Approaches to Prehistoric Diet Breadth, Demography , and Prey Ranking Systems in Time and Space. Time 9. Stuiver, M., Reimer, P., Reimer, R. (2005) - CALIB 5.0. Umbelino, C. (2006) - Outros Sabores do Passado. As análises de oligoelementos e de isótopos estáveis na reconstituição da dieta das comunidades humanas do Mesolítico Final e do Neolítico Final / Calcolítico do território português. PhD Thesis. University of Coimbra.

17

Study of the mammals recovered in Cabeço da Amoreia (Muge – Portugal): State of the art Alexandra Pereira1 1

Universidade do Algarve – FCSH.

Abstract: This paper summarises the current state of investigations of the mammals recovered in the shellmidden of Cabeço da Amoreira, Muge, referring the studies resulting from 150 years of archaeological research. The paper focus on such issues as climate change occurred at the beginning and during the Mesolithic, shellmidden geographic and archaeological context, species and taphonomic processes observed in the collections of medium and large mammals of this archaeological site. Keywords: Muge; Cabeço da Amoreira; Shellmiddens; Mesolithic; Mammals. Cabeço da Amoreira shellmidden is on the left side of the Muge River (Fig. 1), near the present day village of Muge, Salvaterra de Magos (Arnaud, 1987; Umbelino, 2006). Cabeço da Amoreira was one of the earliest of the Muge shellmiddens discovered, and has its own cultural identity, described by Ribeiro (1867). Following the discovery of the Muge shellmiddens, numerous excavations were made, and monographs and articles produced (e.g. Bicho 2010 and 2011; Cardoso, 1999/2000; 2008; Rolão, 1999).

Introduction The study of archaeological faunal remains can provide unique insights into the human past, particularly about the types of animals that inhabited a region and the dietary choices of local human populations. The present paper examines the dietary choices and settlement subsistence systems of early Holocene hunter-gatherers of the Iberian Atlantic regions during a period of changing climatic conditions. The paper reviews the 150 years of research on the Muge shellmiddens, focusing on the hunting of medium- and large-bodied animals and what these faunal assemblages can tell us about palaeoconomic activities at the Muge site of Cabeço da Amoreira.

The Cabeço da Amoreira shellmound was excavated in 1930, 1931, 1933 by Mendes Correa team; in 1962, 1963/64 by Roche and Veiga Ferreira; and in 1999 and 2000 by Rolão. From 2008 to the present, renewed excavations, directed by Bicho have brought together an interdisciplinary team to investigate the Mesolithic activities that formed the midden (Bicho et al., 2010). The first phase of field work opened an area of 6 x 12 m in the centre of the shellmidden. This area is divided into 1 m squares and excavated by artificial levels of 5 cm.

The Portuguese Mesolithic emerged during the transition from the Pleistocene to the Holocene, between 12,000 to 10,000 BP. During this period of improving climatic and environmental conditions, increasing temperatures and humidity lead to sea level rises and coastline losses and increasing plant and animal biodiversity, including the formation of dense forests (Bicho, 2009). The cold event of 8200 cal BP, in combination with a decrease in intensity of upwelling off the Portuguese coast, resulted in further environmental and geomorphologic changes, among them the formation of large river estuaries along the Atlantic coast, created by the rapidly rising waters and low sediment accumulation (Bicho, 2009; Bicho et al., 2010). Epipalaeolithic hunter-gatherer populations, formerly living along the Atlantic coast, moved into these river estuaries including those of Muge and Sado where they established new patterns of subsistence, settlement, hunting strategies and dietary choice (Bicho et al., 2010; Van der Shriek, et al., 2002/2003).

Up to the present, Bicho and his team have excavated the top three layers. Layer 1 is characterized by the presence of pebbles and fire-cracked rock which in fact protected the site and archaeological remains. Abundant flint and quartz instruments were found in this level (Bicho et al., 2010). Layer 2 is characterized by grayish colored sands. Diverse lithic materials and animal species, mainly mamalogical and estuarine fauna, were collected. Layer 3 corresponds to a thick layer of the accumulation of shells and different species of fauna. Bicho and his team have identified two phases of occupation, Mesolithic and Neolithic. Approximately 1200 years of occupation by hunter-gatherers, from 8000 cal BP to 6800 cal BP, are evident (Bicho et al., 2012). From the amount of marine and terrestrial fauna and lithic, bone and stone adornments, this site appears to have served as a camp site, a hunting grounds and, in its final phase, a necropolis (Bicho, 2011).

Geographic and archaeological context The Mesolithic Muge shellmiddens, 13 in all, are located along the Muge tributary to the Tagus. They were discovered in 1863 by Carlos Ribeiro during surveys carried out in the Tagus Valley (Ribeiro, 1867). The 19

Figure 1. The Muge complex shell-middens in the Tagus Valley (Célia Gonçalves).

horse and rabbit in the Mesolithic communities of Cabeço da Amoreira (Bicho, 2009; Lentacker, 1986).

Mammalian fauna Despite the long history of investigations and numerous publications about the Muge middens, up until recently the study of mammals was merely descriptive (Detry, 2007; Bicho, 2010). Only small summary listings of the identified species were published (Arnaud, 1987). The first exhaustive faunal studies of Cabeço da Amoreira were the doctoral dissertations of. Lentacker (1991) and Detry (2007). Using taxonomic, biometric, tafonomics, paleoecological and paleoeconomic analyses Lentacker and Detry obtained more accurate and detailed information about the animals that were hunted. Moreover, these authors identified hunting strategies used by the Muge communities. (Detry, 2007).

According Lentacker (1986), in Cabeço da Amoreira the Cervus Elaphus is the most important food resource, followed by Sus scrofa and Bos primigenius. Cervus elaphus is a typical temperate mammal, which supports the argument for a transitional climatic period of climate improvement. Many elements of deer bone were identified, of both adult and juvenile. They were likely abundant in the Tagus valley during the Mesolithic (Detry, 2007). Sus scrofa was also hunted for food and the bones of juveniles and adults may also have been used to manufacture adornments. Wild boar prefers to inhabit closed forests, often near small lakes or marshes where food is abundant (Lentacker, 1986). The Oryctolagus cuniculus is the most abundant and is present in all layers. The data indicates that the species has been identified identically throughout the occupation and that was a key element to the diet of the people there (Detry, 2007).

Artiodactyls, perissodactyls, lagomorphs and carnivores were found in the Cabeço da Amoreira faunal assemblages. The predominant hunted species were deer (Cervus elaphus), boar (Sus scrofa) and rabbit (Oryctolagus cuniculus). Other mammals found at the site are aurochs (Bos primigenius), roe deer (Capreolus capreolus), horse (Equus sp.), hare (Lepus), fox (Vulpes vulpes), dog (Canis familiaris), otter (Lutra lutra), badger (Meles meles) and lynx (Lynx pardinus) (Detry, 2007). Significant changes occurred in the relative importance of individual taxa over time, particularly deer, boar, aurochs,

The distribution of the skeleton parts reveals information about species use (Detry, 2007). The last phases in Cabeço da Amoreira are marked by a decrease in animals that inhabit open areas (e.g. auroch) and an increase in 20

1 cm

Figure 2. Mandible of boar (Sus scrofa) with vertical cut marks (Cleia Detry’s foto).

species adapted to dense forests (e.g., roe deer). This suggests a temporal change in ecology.

relationship; therefore the identification of dog has important implications about the lives of the Mesolithic hunters and gatherers that bred them. Likewise, burials of dogs within the middens suggest that the human residents of Cabeco da Amoreira may have conferred symbolic value to their dogs (Detry, 2010).

Lentacker (1986) argued that Mesolithic diet was broader spectrum than in the past, with economic activities expanding to include shellfish collecting, fishing and the catching of faster-moving animals such as birds and other small vertebrates.

Most of the animals found at Cabeço da Amoreira were hunted near the camp (Detry, 2007). Most are nonmigrating species, ie. lived in the area all year round. Nevertheless seasonality would have been a factor as some animals are hunted more efficiently in certain periods of the year. Red deer, for example, was hunted from January to March and wild boar from November to March. These patterns suggest year-round human occupation of the site (Lentacker, 1986).

At Cabeço da Amoreira there is a greater diversity of species compared to the faunal assemblages of the other Muge shellmiddens. Detry (2010) identified domestic dog but not excluding the presence of wolf among the faunal materials from Cabeço da Amoreira although Lentacker (1986) previously identified wolf only. This discrepancy may have been due to the loss or mixture of faunal collections from early excavations at the Museum (Detry, 2010) and the fact that the elements Lentacker identified as wolf were probably identified as dog by Detry. Humans and dogs have a unique inter-species

The large deposit of shells at the Muge shellmiddens, created geochemical conditions favourable to the preservation of faunal remains and facilitated the identification of taphonomic processes. Among these processes, those caused by anthropogenic agents are among the most visible, including how they processed and disarticulated the carcasses after capturing and killing the animal. Cut marks, fractures and fire marks provide evidence of human processing activities. In particular, the mammal bones are highly fragmented, suggesting marrow extraction. The extraction of fat (marrow) from bone is a subsistence activity known to our evolutionary lineage since the lower Palaeolithic. Some marks on the skull, shoulder blade and humerus indicate that the animal was hunted with bow and arrow (Detry, 2007).

1 cm An animal carcass can be used in its entirety for fat, meat, bone, skin, teeth, tendons, organs and the bones themselves, and by analysis of fractures and cut marks can be identified the processes of butchering and its effective use. Cut marks result from sharp instruments, in this case produced by stone tools, with a V-shaped asymmetric section. The marks vary in character, depending on tool type, the raw material that it was made from, the technique used by the hunter or butcher, the

Figure 3. Phalanx I of red deer (Cervus elaphus) with cut marks (Cleia Detry’s foto).

21

objective of the activity and the condition of the bone itself.

trabalhos arqueológicos efectuados. Estudos Arqueológicos de Oeiras, 8: 83-240 David, F. (1994) - L'action des carnivores dans les gisements pléistocènes d'Europe. Rappel de quelques étapes de la recherche en Europe occidentale. Artefacts, 9, p. 77-82. Detry, C. (2007) - Paleoecologia e Paleoeconomia do Baixo Tejo no Mesolítico Final: O contributo do estudo dos mamíferos dos concheiros de Muge, Tese de doutoramento em História – Variante de Arqueologia, Facultad de Geografia e História, Universidade de Salamanca, Salamanca. Detry, C.; Cardoso, J. L. (2010) – On some remains of dog (Canis Familiaris) from the Mesolithic Shellmiddens of Muge, Portugal, in Journal of Archaeological Science, 37, pp. 2762-2774. Driesch, A. (1976) A guide to the measurement of animal bones from archaeological sites, Bulletin 1, Cambridge, MA: Peabody Museum Press, Harvard University. Fisher, J. W. (1995) - Bone Surface Modifications in Zooarchaeology. In Journal of Archaeological Method and Theory, Vol 1. Lentacker, A. (1986) - Preliminary results of the fauna of Cabeço de Amoreira and Cabeço de Arruda (Muge, Portugal). Trabalhos de Antropologia e Etnologia, vol. 26: 9-26. Lentacker, A. (1991) – Archeozoologish onderzoek van Laat-Prehistorische vindplaasten uit Portugal. Tese de doutoramento, Laboratorium voor Paleontologie. Faculteit Wetenschappen. Rijksuniversiteit. Gent. Pickering, T. R.; Egeland, C. P. (2006) – Experimental patterns of hammerstone percussion damage on bones: implications for inferences of carcass processing by humans, in Journal of Archaeological Science, Nº33, pp. 459-469. Ribeiro, C. (1867) – Note sur le terrain quaternaire du Portugal. Bulletin de la société Géologique de France. XXIV (2): 692-717. Van der Shriek, et al. (2002/2003) - The Holocene environmental history and geoarchaeology of the Mesolithic cultures in the Muge valley, Lower Tagus basin, Portugal. Muge - Estudos Arqueológicos, 1: 185198. Zilhão, J. (1998) A passagem do Mesolítico ao Neolítico na costa do Alentejo. In Revista Portuguesa de Arqueologia, Vol.1, Nº1, pp. 27-43.

The patterning in the fractures and cut indicate a temporal continuity in carcass processing methods throughout the Mesolithic occupation of the site. Conclusions Faunal materials analysed from the Cabeço da Amoreira show that the Mesolithic people that settled on this area adapted well to the new environmental conditions and benefitted from resources they had at their disposal. The mammals were included in the diet of these populations, notably the deer and the wild boar. These species appear to have been plentiful at this time and, consequently, were commonly hunted and used for food and other purposes. Most of the examined bones are highly fragmented and/or have anthropogenic processing marks, indicating that the animal parts were used for human consumption as well as some parts being altered for artisanal activity and/or damaged by other elements such as being subsequently gnawed by other animals (Lentacker, 1986). The predominance of boar and deer bones suggest a dietary preference for these animals (Detry, 2007). The types of mammals hunted by these Mesolithic Muge communities are typical of a temperate climate where dense forests interspersed by open spaces are part of a high biodiversity ecosystem (Detry, 2007). References Arnaud, J. M. (1987) Os concheiros mesolíticos dos vales do Tejo e Sado: Semelhanças e diferenças. Arqueologia, 15, Porto, pp. 53-63. Bicho, N. (2009) - Sistemas de Povoamento, Subsistência e Relações Sociais dos Últimos Caçadores-Recolectores do Vale do Tejo. Estudos Arqueológicos de Oeiras, 17: 133-156 Bicho, N. et al (2010) – Cabeço da Amoreira, Muge: resultados dos trabalhos de 2008 e 2009, in Os últimos caçadores-recolectores e as primeiras comunidades produtoras do sul da Península Ibérica e do norte de Marrocos. pp. 11-17. Bicho, N., et al (2010b) - The Emergence of Muge Shell Middens in Central Portugal and the 8200 cal yr BP Cold Event. Journal of Island & Coastal Archaeology, 5: 86104. Bicho, N., et al (2011) – The 2008-2010 excavations of Cabeço da Amoreira, Muge, Portugal. Bicho, N. et al (En prelo) - The chronology of the Mesolithic occupation of the Muge valley, central Portugal: the case of Cabeço da Amoreira. Cardoso, J. L., Rolão, J. (1999/2000) - Prospecções e escavações nos concheiros mesolíticos de Muge e Magos (Salvaterra de Magos): contribuição para a história dos 22

Mesolithic and Neolithic shell middens in Western Algarve: issues in ecology, taphonomy and economy Maria João Valente1 1

Universidade do Algarve – FCHS, Portugal; [email protected]

Abstract In Western Algarve there are several archaeological sites with abundant shell deposits (Castelejo, Barranco das Quebradas, Rocha das Gaivotas, Vale Santo I, Alcalar 7 and Ribeira de Alcantarilha). Most of these sites date back to the Mesolithic and/or Early Neolithic—from the Preboreal to the Atlantic periods—and are considered very similar to each other regarding function and deposited materials. Nonetheless, these same sites and their deposits also show noticeable differences in relative abundance of mollusk species, evidences of their processing as food, and post-depositional activity. It is our intent to critically organize the available data and discuss issues such as: (1) differences on the relative abundances of species and their possible causes; (2) food processing techniques, through time and in between contemporaneous occupations; and (3) post-depositional processes and their effects. Interpretative hypotheses regarding the economy and diet of the last hunter-gatherer and first productive communities (Mesolithic to Early Neolithic) of Western Algarve will also be proposed. Keywords: Shell middens, Mesolithic, Algarve. Introduction In this study, using zooarchaeological data as working frame, we argue that these same sites and their deposits also show noticeable differences in relative abundance of mollusk species, evidences of their processing as food, and post-depositional activity. On this regard it is our intent to critically organize the available data and address issues such as:

Western Algarve is located in the southwestern area of Portugal, and is known by its several archeological sites dating back to the Mesolithic and/or Early Neolithic— from the Preboreal to the early Atlantic periods (10.000– 5.600 BP or 9.400–4.600 cal BC). Many of these sites have excellent preservation, sometimes with important shell deposits, mostly due to favorable post-depositional factors which include limited modern anthropic activities. In fact, part of this area, located along the western Atlantic, is classified as a natural park—Parque Natural do Sudoeste Alentejano e Costa Vicentina—thus being subjected to less intense human modification.

– Are relative abundances of species mirroring local ecological availability or are they the product of human selective choices? Or should both factors be considered? – What kind of food processing techniques were used? Are these techniques (and their traces) perceivable through time or in between sites from same age?

Amongst these sites, the most significant ones are Rocha das Gaivotas and Armação Nova (Carvalho, Valente & Dean, 2010), Barranco das Quebradas (Valente, 2010), Vale Santo (Carvalho, 2008), Cabranosa (Cardoso, Carvalho & Norton, 2001), Castelejo (Silva & Soares, 1997), Padrão (Carvalho, 2008), Vale Boi (Dean & Carvalho, 2011), Alcalar 7 (Carvalho, 2008), Ibn-Amar Cave (Carvalho, 2008) and Ribeira de Alcantarilha (Stiner et al., 2003) (Figure 1). During the last decades, some of these sites have been subject of extensive archaeological studies (see above references; also Dean et al. (2011) and Valente et al., in press). Despite that, the available data for each site is varying, depending on the amount of materials in the collections and analyses exhaustiveness. In general, the sites have been considered very similar regarding their function (specifically as temporary human occupations aiming the capture and processing of mollusks and/or collection of lithic raw material in the surrounding area) and deposited materials (mostly mollusks, with limited number of lithic materials and rare utensils) (e.g. Bicho, 2004; Soares & Silva, 2004; Valente & Carvalho, 2009).

Figure 1. Western Algarve’s archaeological sites mentioned in text (Mesolithic and Neolithic). Legend: (1) Rocha das Gaivotas & Armação Nova; (2) Barranco das Quebradas; (3) Vale Santo & Cabranosa; (4) Castelejo; (5) Padrão; (6) Vale Boi; (7) Alcalar 7; (8) Ibn-Amar Cave; (9) Ribeira de Alcantarilha. 23

– What are the obvious post-depositional processes? How do they affect our perception of the archaeological context?

the site location: in the western rocky cliffs areas we mostly find hard substrates inhabitants, such as mussels, whelks, topshells and limpets; in the southern zone, where sites are located by fresh or brackish water, other species are found, like cockles and clams.

From the obtained framework we will propose some interpretative hypotheses regarding the economy and diet of the last hunter-gatherer (Mesolithic) and first productive communities (Early Neolithic) of Western Algarve.

Table 1 presents a list of the main marine invertebrate taxa found in the Mesolithic and Neolithic sites located in Western Algarve. The following descriptions were obtained consulting several printed publications (Saldanha, 1995; Riedl, 2000; Borges, 2007) and online database (WoRMS and MarLIN).

Western Algarve sites The sites featured in this study are located on Algarve’s western area (Figure 1), which is marked by three distinct sub-areas that are longitudinally developed. From north to south there are higher interior mountains composed largely of schist (Serra), a middle area called Barrocal with limestone and clay soil, and a lower littoral zone made of sandy soils.

The most abundant species within the Gastropoda are: – Monodonta lineata, da Costa 1778 (=Osilinus lineatus; thick topshell). With a turbinated shell, this species is frequent in rocky shores, in the intertidal zones. It can reach up to 3cm in height and is thicker and heavier than other species found in the area. Easy to distinguish from other Trochidae, given its prominent 'tooth' or bulge on the nacreous inner shell layer.

Most of the archaeological sites are located in the littoral zone, either on the west or the south sectors of the Atlantic coast. This territory is regarded as one of the richest areas in Portugal in terms of marine animal communities, having a high concentration of fishes and rock-dwelling mollusks and crustaceans, such as limpets and gooseneck barnacles.

– Patella spp. (limpets). Several species with small to mid-size conical shells (up to 6cm diameter). Found in suitable hard substratum, on rocks, stones and in rock pools. Abundant in all degrees of wave exposure, from the high shore to the sublittoral fringe. In archaeological collections it is difficult to distinguish individual species; however, in Barranco das Quebradas (where some of the specimens are well preserved), it was possible to identify the following species:

On the west side of this coast, the geological substrate is largely sandy or sandy-clayish, with areas of schist and limestone, sometimes including a variety of siliceous bodies, thus being a good region to collect flint. The shore has been exposed to strong marine and aeolian abrasion, resulting in an irregular coast featuring very high cliffs intercalated by small sandy beaches lying at the base of minor ravines. The site of Castelejo is located in one of these beaches. Barranco das Quebradas, on the other hand, is located in one of the many ravines of the area. On the tops of the cliffs there are several fossilized dunes (Pereira & Angelucci, 2004), some with deposited archaeological materials, like Rocha das Gaivotas and Armação Nova. All these archaeological sites are within a short distance from the sea, but some other sites, such as Vale Santo and Cabranosa, are located in more interior sand dunes.

– Patella vulgata, Linnaeus 1758: usually not abundant on shores with a dense growth of seaweed; the highest densities coincide with wave exposed conditions. – Patella intermedia, Pennant 1777 (=P. depressa): found on exposed, wave-beaten rocky shores from the middle to the lower shore. Distinguished from Patella vulgata by its smaller, flatter shell. – Patella ulyssiponensis, Gmelin 1791 (=P. aspera): common on wave exposed rocky shores, avoiding extremely sheltered areas and fresh water. Found on rocks on the lower shore, also in shallow rock pools on the middle shore, on overhanging rocks.

The south littoral zone has a similar geological substrate, but presents a smaller degree of abrasion. This particular region contains several small rivers and estuaries, the most important being Alvor and Arade, as well as lagoons. Nearby these freshwater sources, we find the sites of Padrão, Vale Boi, Alcalar 7, Ibn-Amar Cave and Ribeira de Alcantarilha.

– Patella caerulea, Linnaeus 1758: nowadays with distribution restricted to the Mediterranean (Poppe & Goto, 1991), is a species frequent in intertidal zones. – Thais haemastoma, Linnaeus 1767 (= Stramonita haemastoma; southern oyster drill or purpura). As for Bivalvia, the most abundant species are:

Marine invertebrate abundance

taxa

and

their

relative – Mytilus galloprovincialis, Lamarck 1819 (mussel). This species can grow to 140mm, in densely populated communities. Normally found in the infra to midlittoral zones of rocky coasts. Although very similar to Mytilus edulis, with which is easily mistaken, the

The mollusks and crustaceans taxa found in the archaeological site of Algarve, although relatively homogeneous in each period, tend to differ according to 24

25

Goose barnacle Limpet Common or thick topshell Southern oyster drill Mussel Common cockle Clam

Pollicipes pollicipes

Patella spp. Monodonta lineata Thais haemastoma

Mytilus sp.a Cerastoderma edule Ruditapes decussata +

++

++ + +

++

RGV

++

++

++

AN*

+

++ ++ +

+

BQ

++

++ + ++

++

VS

+

+

+

CAB

++

++ ++

CST*

+ ++ ++

+

++

+

PAD

Table 1. List of main taxa found in Western Algarve’s mesolithic and early neolithic sites.

Common name

Taxonomic name

+

+

+ +

+

VB

++

AL7*

+

IBN*

+ + ++

+

+

+

RBA

8.8 – 8 ky 8 – 7.6 ky 7.6 – 6.5 ky 6.6 – 5.6 ky

Early Boreal

Late Boreal

Early Atlantic (I)

Early Atlantic (II)

Pollicipes & Mytilus > Patella (rocky shores) Ruditapes (sandy river beds)

Patella & Pollicipes > Mytilus

Patella & Mytilus

Patella & Mytilus

Mollusk abundance (taxa) Monodonta > Patella

Table 2. Relative abundance of species (or taxa) through time.

Dates BP 10 – 8.8 ky

Period Preboreal

Sites: RGV – Rocha das Gaivotas; AN – Armação Nova; BQ – Barranco das Quebradas; VS – Vale Santo; CAB – Cabranosa; CST – Castelejo; PAD – Padrão; VB – Vale Boi; AL7 – Alcalar 7; IBN – Ibn-Amar Cave; RBA – Ribeira de Alcantarilha. (sites without quantitative data are marked with an asterisk). Taxa: + = present; ++ = abundant. a Probably M. galloprovincialis, given its geographical dispersion (see text).

CLASS / Family CIRRIPEDIA Pollicipedidae GASTROPODA Patellidae Trochidae Muricidae BIVALVIA Mytilidae Cardiidae Veneridae

galloprovincialis species dominates the areas located south of the English Channel.

however, Silva & Soares (1997) and Soares & Silva (2004) mention that the most significant taxa are limpets and topshells, followed by mussels.

– Cerastoderma edule, Linnaeus 1767 (common cockle). With a mid sized shell, up to 5cm, this species typically appears in mobile (normally sandy) substrates. Mostly found in the circalittoral zone.

During the first phase of the Early Atlantic period (7.6– 6.5 ky BP) we have three sites with deposits of shells. Full quantification data for Rocha das Gaivotas (middle layers) shows a majority of limpets, followed by gooseneck barnacles and then by mussels. We also have the relative abundance data for the other two sites: in Armação Nova, barnacles seems to be the most abundant species, with mussels and limpets coming in second; as for the middle layer of Castelejo, the species with higher percentage are limpets and then mussels.

– Ruditapes decussata, Linnaeus 1767 (= Tapes decussatus, Venerupis decussata; clam). Animal with a midsized shell, which can reach 7cm. Found in the lower shore and shallow infralittoral, in sand, muddy gravel, or clay substrates, in estuaries or lagoons. Within the crustaceans, the species to be noted is the Pollicipes pollicipes, Gmelin 1789 (= Pollicipes cornucopia; gooseneck barnacle). This animal is found in groups attached to rocks on exposed rocky shores and cliffs (infralittoral zone). It’s distinguished by its long neck (sometimes reaching more than 5cm), where the edible part is located. Its shell, or capitulum, is made up of 18 small plates, easily separated post-deposition.

There are many more sites dating from the second phase of the Early Atlantic (6.5–5.6 ky BP), some of them located nearby rocky sea cliffs, others by river beds. All these contain Early Neolithic occupations. The first group—Cabranosa (small amount of shell materials), Castelejo (upper layers) and Vale Santo 1— yielded a bigger amount of mussels, followed by gooseneck barnacles or limpets. The other rocky shore neighboring sites show more heterogeneous collections: Rocha das Gaivotas (top layers) is dominated by barnacles, and then limpets, followed by mussels; Padrão 1 has more limpets, and then two other species typical of different, sandy, environments (cockle and clam); Vale Boi’s small shell collection has a predominance of gooseneck barnacle, venus and limpets.

Within the less abundant species, the most distinctive are Haliotis tuberculata (abalone), Diodora graeca (keyhole limpet), Gibbula umbilicalis (flat topshell), Charonia lampas (trumpet shell), Nucella lapillus (dog whelk), Glycymeris insubrica (dog cockle) and Pecten maximus (great scallop). Relative abundance of species

In the sites located near brackish river water—Alcalar 7, Ribeira de Alcantarilha and Ibn-Amar Cave—clams are the most common, followed by oysters in Ibn-Amar.

Gathering all the available data, we can distribute the relative abundance of species through time (summarized data by period can be seen on Table 2). In this analysis, when possible, we used MNI (Minimal Number of Individuals). MNI diminishes the errors of intra and intersite differences in shell fragmentation, not controlled when using NISP (Number of Identified Specimens; Lyman, 2008, pp. 43-44) and of intersite variable deposit characteristics that affect the weight of the remains (idem, pp. 102-108). It also overcomes problems of intertaxonomic variation in the number of identifiable elements per individual (idem, p. 44).

Summarizing (Table 2), we can conclude that during the Preboreal there is a majority of topshells and then limpets. During Early and Late Boreal a bigger amount of limpets and/or mussels. And, finally, during the Atlantic topshells’ quantity is very reduced and the shell deposits are dominated by limpets and/or gooseneck barnacles, followed by mussels; exception due in sites located by brackish waters and sandy substrates where clams are in bigger amounts.

The oldest anthropic shell deposit know in Western Algarve is Barranco das Quebradas 1. Its base layer, dated from the Preboreal period (10–8.8 ky BP), displays a predominance of topshell and limpets.

The final question concerning this pattern is whether the relative abundances of species mirror local ecological shellfish availability or if they are the result of human selective resource gathering. If we take into account that most of the contexts have similar ecological background (i.e. rocky slope areas), and that the data is constant within periods, we accept that this pattern—e.g. abundance of topshells in the earlier period; limpets and gooseneck barnacles in the later one—supports local ecological availability. Selective cultural choices may have existed but, in our opinion, they played a lesser role.

From the Early Boreal (8.8–8 ky BP) we have all the other Barranco das Quebradas contexts (Barranco das Quebradas 1 [top], 3, 4 and 5), yielding the same previous abundances and, later on, a large percentage of mussels and limpets. In the oldest occupations of Rocha das Gaivotas (lower layers), the most abundant species is also mussel. Later Boreal (8–7.6 ky BP) is only known in the inferior layer of Castelejo. Its specific quantification is unknown, since full zooarchaeological studies were never done; 26

Species

Morphology

Breakage Patterns

Agents / Causes

Monodonta lineata

Exoskeleton: very thick. Effort on removal of meat without breakage: high.

Variable data. Mostly human breakage to extract meat in different contexts. Some post-depositional breakage.

Patella spp.

Exoskeleton: thin. Effort on removal of meat without breakage: low.

Mytilus sp.

Exoskeleton: very thin. Effort on removal of meat without breakage: low.

Thais haemastoma

Exoskeleton: thick Effort on removal of meat without breakage: high

Variable degree of fragmentation. 10% – BQ1 35% – BQ3 60% – BQ4 >95% – BQ5 100% – RGV inferior High degree of fragmentation, namely along concentric growth rings. 80-90% – all BQ 90-100% – all RGV Very high degree of fragmentation. 100% – all BQ 100% – all RGV Very high degree of fragmentation. >95% – BQ1 100% – BQ3 100% – RGV inferior 98,5% – RGV middle 90-95% – RGV top

Regular data. Mostly postdepositional. Possible breakage during food processing. Very regular data. Mostly post-depositional. Possible breakage during food processing. Regular data. Mostly human breakage to extract meat. Some post-depositional breakage.

Table 3. Summary of breakage patterns: Barranco das Quebradas (BQ) and Rocha das Gaivotas (RGV).

Also to be noticed is a significant decrease in foraging efficiency across the Mesolithic/Neolithic transition in the area. The most significant of these decreases occurs with the Early Neolithic, when the small gooseneck barnacle dominates assemblages and shell-midden layers are sparser and more ephemeral than those seen in the Late Mesolithic. This shift is global to southwestern Portugal, one of the earliest areas of the Atlantic coast to experience the transition to Neolithic and agriculture. The roots of this declining efficiency appear in the Late Mesolithic when valued resources, like topshell and purpura, dropped out of the diet. This also represents an adaptation toward the intensive use of mass-collected species (like mussels and goose barnacles), at the expense of these gastropods (Dean & Carvalho, 2011; Dean et al., 2011).

in between sites with similar age deposits. For this we can only use data from fully studied contexts, like Barranco das Quebradas and Rocha das Gaivotas. The comparative data will focus breakage patterns, specially natural versus anthropic caused fractures, and fire processing marks, mostly burning and eventually boiling. Different breakage patterns are mainly seen across species, although there are also some contrasting preservation between sites (Table 3). Limpets (Patella spp.) and mussels (Mytilus sp.) are animals whose exoskeleton structure is thin or very thin. In these animals obtaining the soft edible part requires a minor effort, therefore it comes as no surprise that their shells present a high degree of fragmentation (80–100%). In our opinion, this is largely due to non-anthropic post-depositional causes, even if we cannot rule out some human caused breakage during food processing.

The same trend can be seen on throughout the Neolithic layers at Rocha das Gaivotas, where there is a significant decrease in the average size of both limpets and gooseneck barnacles (see above references for further development).

Whelks (Thais haemastoma) also have regular breakage patterns, with a high degree of fragmentation (90–100%; Figure 2). However their exoskeleton is much thicker than the one from limpets and mussels. The effort to remove meat without shell breakage is also quite larger. Hence, it is probable that most fragmentation is human caused while processing the animal for meat extraction. Naturally, some post-depositional breakage is possible as well.

Processing techniques Our intent is also to evaluate food processing techniques and compare the observable evidences through time and

27

Figure 2. Examples of breakage patterns. Legend: A – Monodonta lineata (Barranco das Quebradas 3); B - Thais haemastoma (Barranco das Quebradas 4). Finally, topshells (Monodonta lineata) display a very variable degree of fragmentation (10–100%; Figure 2) across deposits. Their shells are very thick and the effort required for removal of meat without breaking the exoskeleton is high. Therefore, we consider that most of the fracturing is caused by human food processing. In this regard, we can also point out that in some deposits, like in Barranco das Quebradas 5, there are some rocks or pebbles that can be interpreted as anvils (Figure 3).

Figure 3. Pebble interpreted as anvil (Barranco das Quebradas 5). Note: the center of the pebble shows percussion marks. dependable on the context: materials collected within or nearby fireplaces show a much higher percentage of carbonization (e.g. c. 29% in Fireplace 5; Table 4).

As for carbonization and other fire evidences, in Barranco das Quebradas the percentage of unequivocal processing by fire is low (0.6–2.4%). Nonetheless, some other evidences, like the presence of charcoal and thermo-fractured rocks, are important to corroborate the idea that some of the animals were indeed processed by direct exposition to fire (by contact with burning coal or braziers). The usual gray color displayed by these burned shells also validates this (Chernokian, 1986).

As conclusion, we can say that shellfish food processing shows a clear variation between species—whelks and topshells seem to be the most affected. The data collected shows no obvious change of patterns along the time, with exception of more structured fireplaces during the Late Mesolithic (Rocha das Gaivotas), which are clearly related with a higher percentage of burned materials. The most observed processing methods are breakage for meat extraction and burning (this last shown both by direct and indirect evidences).

The brown color displayed by some shells is more elusive. Some authors mention the possibility of liquid boiling with mixed seaweed leaving a brown color in the shells (see Dupont, 2003); however, the color of the sediments in which the Barranco das Quebradas’ materials were found is also red-brownish, therefore dithering the liquid boiling hypothesis.

Post-depositional activities One of the biggest challenges of the archaeologist is to assess the post-depositional processes that operated in a site, which is fundamental to control non-anthropic interferences in the archaeological context and its

As for Rocha das Gaivotas, the amount of remains showing burned marks is highly variable and clearly Site Barranco das Quebradas

Rocha das Gaivotas

Direct evidences BQ1: 1.2% of the remains display color alteration, mostly gray, occasional brown. BQ3/4: 0.6-1.3% altered remains (mostly Thais and Monodonta). BQ5: 2.4% remains display gray color. Variable % of remains displaying gray color (higher in fireplaces).

Indirect evidences Possible fireplaces, but so far these were not found preserved. Thermo fractured rocks. Charcoal in BQ5.

Agents / Causes Gray color = contact with burning coal or hot braziers. Brown color = eventual liquid boiling, particularly when mixing seaweed.

Fireplaces. Thermo fractures rocks. Charcoal.

idem

Table 4. Summary of carbonization and other fire evidences: Barranco das Quebradas (BQ) and Rocha das Gaivotas (RGV). 28

Figure 5. Quantification as control method in Barranco das Quebradas 1 (NISP and MNI): general data by artificial level. Note: Particularly valid for species that are more susceptible to fragmentation by diagenetic factors, like compression, abrasion, etc (e.g. Mytilus). A closer value between NISP and MNI is indicative of a well preserved taxa or context (i.e. optimal balance). We should, however, take into account intertaxonomic variation in the number of identifiable elements per individual (ex. Bivalvia versus Gastropoda) and that a higher number of identifiable elements normally equals a higher NISP.

Figure 4. Vertical dispersion in Rocha das Gaivotas (Sector 1, Layer 2c). Notes: Modified from Carvalho, 2008. Ceramics: total weight; mollusks: total weight from squares B29 and C29. materials. During this analysis, we will, again, focus in Barranco das Quebradas and Rocha das Gaivotas, given the need of full and liable data. Collected data demonstrate that the most obvious postdepositional alterations are related with flotation and vertical dispersion of materials, something connected with their weight and the nature of the surrounding sediments. This is most obvious in Rocha das Gaivotas where we have sand deposits, which are more susceptive to movement, namely by eolian action. Preliminary interpretation of Level 2a (from Sector 1; Carvalho & Valente, 2005) considered it as an unique neolithic occupation; later on, after further analysis to verify the weight of the materials (mollusks and ceramics), two different occupations were distinguished (Figure 4): one from the Early Neolithic, with a deposition peak of ceramic materials in artificial level 4; another dating from the Late Mesolithic, with a maximum deposition of shells in level 6.

This demonstrates how post-depositional evidences are a highly variable proxy that should be meticulously addressed to avoid biased interpretations, including those regarding relative abundance of species in the archaeological contexts. Also, special attention should be given to the context sediments and general morphology (e.g. presence of rocks, pebbles and their size). As an example, sand sediments with a high post-depositional perturbation rate normally show an intense fragmentation or alteration of remains (including vertical and horizontal movement). Final remarks on ecology, taphonomy and economy Mesolithic and Neolithic shell deposits in Western Algarve show a clear dependence on local resources. On the other hand, species natural availability seems to have changed through time, with topshells being one of the most abundant species in the earliest Mesolithic occupations, being replaced by limpets and gooseneck barnacles during the Later Mesolithic and Early Neolithic. However, since the oldest occupation sites are all located in the rocky western coast, data from this period is biased towards hard substrate habitat species. Sites located nearby rivers or estuaries date only from the Early Neolithic, yielding other taxa like cockles and clams.

An additional method to control post-depositional processes is related with the shell fragmentation intensity, specifically by comparing the NISP/MNI of each of the main species or taxa (Figure 5). This is particularly valid for species that are prone to be fragmented by diagenetic factors, such as compression or abrasion (e.g. mussels) or fractures by human action (e.g. whelk). Another factor to take into account is the presence of animals with more anatomical elements, as bivalves (two shells instead of one) or crustaceans. In Barranco das Quebradas 1, NISP results inflate the abundance of the most fragmented species, such as limpets, whelks or mussels (Figure 6). If we compare the NISP quantification with the one obtained for MNI, we observe the decrease of mussels or whelks, with a clear abundance of topshells, a species less affected by fragmentation.

Data show that mollusk food processing was mostly done by breaking gastropods for meat extraction purposes: specially purpura and occasionally topshells. Fire exposure is moderately used during the earliest period, being more pronounced during the Late Mesolithic, when

29

pre- and post-deposition phenomena are sometimes intense and, if disregarded, will lead to misleading interpretations. Online databases Appeltans, W.; Bouchet, P.; Boxshall, G. A.; De Broyer, C.; de Voogd, N.J.; Gordon, D. P.; Hoeksema, B. W.; Horton, T.; Kennedy, M.; Mees, J.; Poore, G. C. B.; Read, G.; Stöhr, S.; Walter, T. C. & Costello, M. J. (Eds.) (2012) – World Register of Marine Species (WoRMS). Accessed at http://www.marinespecies.org on 2012-0130. MarLIN (Marine Life Information Network) (2009) – Marine Life Information Network. Plymouth: Marine Biological Association of the United Kingdom. [cited 01/01/09]. Accessed at http://www.marlin.ac.uk on 201201-30. References Bicho, N. F. (2004) – As comunidades humanas de caçadores-recolectores do Algarve Ocidental. In: A. A. Tavares, M. J. F. Tavares & J. L. Cardoso (Eds.) – Actas do congresso evolução geohistórica do litoral português e fenómenos correlativos: Geologia, história, arqueologia e climatologia. Lisboa: Universidade Aberta. Pp. 359-396. Borges, T. C. (Ed.) (2007) – Biodiversidade nas pescas do Algarve (Sul de portugal) / Biodiversity in the fisheries of Algarve (South of Portugal). Faro: Universidade do Algarve. Cardoso, J. L.; Carvalho, A. F. & Norton, J. (2001) – A estação do Neolítico Antigo de Cabranosa (Sagres, Vila do Bispo): Estudo dos materiais e integração cronológico-cultural. O Arqueólogo Português, 16, pp. 55-96. Carvalho, A. F. (2008) – A neolitização do Portugal Meridional. Os exemplos do Maciço Calcário Estremenho e do Algarve Ocidental. Faro: Universidade do Algarve. Carvalho, A. F.; & Valente, M. J. (2005) – Novos contextos conquíferos pré-históricos na Costa Vicentina. In: Actas do 2.º Encontro de Arqueologia do Algarve (Xelb, 5). Silves: Câmara Municipal de Silves. Pp. 9-26. Carvalho, A. F.; Valente, M. J. & Dean, R. M. (2010) – O Mesolitico e o Neolitico Antigo do concheiro da Rocha das Gaivotas (Sagres, Vila do Bispo). In: Actas do 7.º Encontro de Arqueologia do Algarve (Xelb, 10). Silves: Câmara Municipal de Silves. Pp. 39-53. Chernokian, R. (1986) – Caractéres specifiques et modalités d’étude des amas coquilliers anthropiques. Travaux Du LAPMO, 1986, pp. 1-20. Dean, R. M. & Carvalho, A. F. (2012) – Surf and turf: The use of marine and terrestrial resources in the early Neolithic of coastal Southern Portugal. In: N. F. Bicho; J. A. Haws & L. G. Davis (Eds.) – Trekking the shore. Changing coastlines and the antiquity of coastal settlement. New York: Springer. Pp. 291-302. Dean, R. M.; Valente, M. J. & Carvalho, A. F. (2011) –

Figure 6. Quantification as control method in Barranco das Quebradas 1 (NISP and MNI): main species. Note: NISP values display a regular inflation of the most fragmented species, such as Patella, Thais and Mytilus, or of species with more anatomical elements, as bivalves or crustaceans. MNI controls fragmentation and variable number of anatomical elements. structured fireplaces were constructed (as seen in Rocha das Gaivotas). There is no clear evidence supporting global overexploitation of species in the Mesolithic. However, throughout the Neolithic layers at Rocha das Gaivotas, there is a significant decrease in the average size of both limpets and goose barnacle, which may indicate that foraging efficiency decrease occurred in the Mesolithic– Neolithic transition. Observed decrease in the average size of both limpets and gooseneck barnacles during the Neolithic supports this trend. Neolithic times also seem to imply new areas with human occupation, closer to rivers and away from rocky shores, where clams are the most consumed species. At the same period, shell deposits in the western rocky coast are far thinner and more ephemeral than those seen in the Late Mesolithic. Finally, we would like to point out the need to attentively address taphonomic processes, since the signatures of 30

The Mesolithic/Neolithic transition on the Costa Vicentina, Portugal. Quaternary International, 264, pp. 100–108. Dupont, C. (2003) – La Malacofaune de sites mésolithiques et néolithiques de la façade atlantique de la France: contribuition à l'économie et à l'identité culturelle des grupes concerné. Unpublished Ph.D. Thesis. Paris: Université Paris I (Sorbonne). Lyman, R. L. (2008) – Quantitative paleozoology. Cambridge; New York: Cambridge University Press. Pereira, A. R. & Angelucci, D. (2004) – Formações dunares no litoral português, do final do Plistocénico e inícios do Holocénico, como indicadores paleoclimáticos e paleogeográficos. In: A. A. Tavares, M. J. F. Tavares & J. L. Cardoso (Eds.) – Actas do congresso evolução geohistórica do litoral português e fenómenos correlativos: Geologia, história, arqueologia e climatologia. Lisboa: Universidade Aberta. Pp. 221-256. Poppe, G. T., & Goto, Y. (1991) – European seashells: Polyplacophora, caudofoveata, solenogastra, gastropoda (Vol. 1). Wiesbaden: Verlag Christa Hemmen. Riedl, R. (2000) – Fauna y flora del mar mediterraneo. Barcelona: Ediciones Omega. Saldanha, L. (1995) – Fauna submarina atlântica: Portugal continental, Açores, Madeira. Mem Martins: Publicações Europa-América. Silva, C. T. & Soares, J. (1997) – Economias costeiras na pré-história do Sudoeste português: O concheiro de Montes de Baixo. Setúbal Arqueológica, 11-12, pp. 69108. Soares, J. & Silva, C. T. (2004) – Alterações ambientais e povoamento na transição Mesolítico-Neolítico na Costa Sudoeste. In: A. A. Tavares, M. J. F. Tavares & J. L. Cardoso (Eds.) – Actas do congresso evolução geohistórica do litoral português e fenómenos correlativos: Geologia, história, arqueologia e climatologia. Lisboa: Universidade Aberta. Pp. 397-423. Stiner, M. C.; Bicho, N. F.; Lindly, J. & Ferring, R. (2003) – Mesolithic to Neolithic transitions: new results from shell-middens in the Western Algarve, Portugal. Antiquity, 77(1), pp. 75-86. Valente, M. J. (2010) – O Barranco das Quebradas (Vila do Bispo, Portugal) no contexto dos concheiros mesolíticos do Sudoeste Português. In: Actas do 7.º Encontro de Arqueologia do Algarve (Xelb, 10). Silves: Câmara Municipal de Silves. Pp. 15-38. Valente, M. J. & Carvalho, A. F. (2009) – Recent developments in early holocene hunter-gatherer subsistence and settlement: a view from South-Western Iberia. In S. McCartan; R. Schulting; G. Warren & P. Woodman (Eds.) – Mesolithic horizons: papers presented at the Seventh international conference on the Mesolithic in Europe, Belfast 2005. Vol. 1. Oxford: Oxbow Books. Pp. 312-317. Valente, M. J.; Dean, R. & Carvalho, A. F. (in press) – Shell-Middens in Western Algarve (Southern Portugal) during the Mesolithic and Early Neolithic: functionality, subsistence and material culture. In M. Roksandic; S. Mendonça; S. Eggers; M. Burchell & D. Klokler (Eds.) – The cultural dynamics of shell middens and shell mounds: a worldwide perspective. Gainesville: University Press of Florida. 31

The inclusion of faunal remains in Bronze Age funerary practices in Southern Portugal. Montinhos 6 - a case study Cláudia Costa1 & Lidia Baptista2 1 2

Universidade do Algarve, FCSH; [email protected]. FLUP – CEAUCP-CAM. PhD scholarship from FCT. [email protected]

Abstract Montinhos 6 is located in Serpa (Beja, south Portugal) and was excavated under the direction of one of the authors (LB) during an emergency intervention within the Alqueva Project. The site was characterized by negative structures of different types, chronologies and fills. In this paper the authors wish to focus on the structures related to the burial practices from the Bronze Age: both hypogea and pits with burials from one individual to a maximum of five. Only six hypogea revealed faunal remains associated with funerary rituals, consisting mainly of bones from the forelimbs of Bos taurus and Ovis/Capra and a partial skeleton of Oryctolagus cuniculus. Most parts of the faunal assemblage were deposited in articulation, but isolated radius are also present. An analysis of how these elements are used in these practices can contribute some information to our understanding of the burial practices of the Bronze Age. Key-words: Bronze Age; funerary practices; faunal remains. hercynian substrate partially covered by quaternary and tertiary deposits which form the Alentejan peneplain, an extensive flat area with gentle undulations (Duque & Almeida 1998). The closest mountainous areas are the Serra de Serpa to the south, the Serra de Portel to the north and the Serra de Ficalho to the east.

Introduction The site Montinhos 6 was identified and excavated under the auspices of the “Project for the minimization of Impact on cultural heritage arising from the execution of the Brinches-Enxoé irrigation block – Work Phase” (Projeto de minimização de Impactes sobre o Património Cultural decorrentes da execução do Bloco de Rega Brinches-Enxoé – Fase de Obra) undertaken by the company EDIA, SA. The execution of the project, which focused on an extensive area of the county of Serpa, and specifically the parishes of Brinches and Salvador, involved the opening of irrigation channels and in some case more ample areas (where reservoirs or pumping stations were to be installed). Excavation was carried out by teams from the companies Arqueologia e Património, Lda., and Histórias e Tempus, Lda, under the coordination of Lidia Baptista. Various excavations were carried out which allowed the identification and study of contexts that contributed to a greater understanding of the past of the county of Serpa. The results obtained in the different emergency excavations allowed us to convert a series of negative impacts on our cultural heritage into an opportunity to know and study many archaeological sites (Baptista & Gomes 2012). Of the 38 sites identified just Montinhos 6, which revealed important archaeological contexts from various chronologies was extensively excavated in terms of area, as its location coincided with the future site of a reservoir.

In terms of lithology, this is an area characterised by the presence of metamorphic, sedimentary and igneous rock aged to between the Precambrian and the late Palaeozoic period, and large quantities of igneous rock, such as diorites, gabbros, granite and porphyries can be found in the area (Hidroprojecto et alli 1998: 5). Montinhos 6 is situated on two small hills separated by a light depression where the substrate is made up of caliço (a form of limestone) of a whitish colour which corresponds to a decomposition of gabbro-diorites which are very easy to carve. The site of Montinhos 6 has a line

Summary of the excavation results Location Montinhos 6 (Baptista & Gomes 2011; Rodrigues 2011) is made up exclusively by negative structures of varying morphologies and chronology (from the Chalcolithic until Late Antiquity). The site is situated in the catchment area of the river Guadiana, which geologically is made up of a

Figure 1. Localization of Montinhos 6 in the Iberian Península and in the county of Serpa.

33

Figure 2. Distribution of the negative structures over the two hills.

Figure 3. Distribution of the negative structures in the northeast hill. 34

Figure 4. Distribution of the negative structures in the southwest hill. categories of artefact participate. Ceramics are the most commonly recurring element, sometimes associated with faunal remains and stone elements and the identification of some refits between several of the ceramic fragments from different deposits and structures has sparked discussion about the biographies of the artefacts. For example, there exist various cases in which separate halves of the same vessels have been deposited in different areas within the same pit and the refitting of these fragments has allowed us to consider temporal relations between structures and to think of the filling of the structures themselves as architecture (Garrow et alli 2005; McFadyen 2006).

of site to the distant horizon in the north, west and south and the surrounding landscape is dotted by gentle hills cut in places by small ravines. The view is somewhat limited to the east where there are some slightly bigger hills. This expanse gives us a view of the catchment area of both the Ribeira do Enxoé to the south and the Ribeira de Grafanes to the north (figure 1). Structures and chronology The site was identified during the construction of a reservoir which entailed the stripping of an area of 77713 m². In this area more than 200 negative structures were discovered, without any type of ditch or trench delineating the area of this large number of structures (figures 2, 3 e 4).

Funerary contexts” of the Bronze Age in Montinhos 6 Hypogea

Dating to the 3rd millennium BC, two architectural types have been identified: Pits with a subcircular form in plan and 55 structures which in plan roughly take the form of a long bone (Baptista & Gomes 2013). It is however the Bronze Age contexts that are the most numerous. This category is made up of 14 hypogea, 130 pits of subcircular plan, 2 small pits/depressions and an “occupation area”. Human burials were found in all of the hypogea, in the 2 small pits/depressions and in 7 of the pits.

The human burials associated with the Bronze Age occur in 14 of the hypogea and 7 of the pits. The structures referred to here as hypogea are diverse in orientation and dimension, but are defined as composite structures dug out of the substrata and consisting of a oval/circular or square/rectangular antechamber with one or two funerary chambers of circular plan (figure 5). When we consider the relations between the fill, the burial levels, and the different spaces that make up the hypogea, there are several thing to note: 1) the majority of the fill deposits from the hypogea show similar characteristics to the substrata in which the structures

In addition to the funerary contexts, other acts of deposition exist within the pits in which various 35

were opened, do not normally contain any artefacts, and when they do they are very fragmented and eroded; 2) the antechambers were intentionally filled in within a short time, probably with the earth taken from the excavation of the structure; 3) in the majority of cases, the roofs of the funerary chambers had fallen in, probably during prehistory. This hypotheses is supported by the exclusive presence of manual ceramic fragments, the majority of which were badly eroded in the sediments that covered the in inhumations; 4) when the roofs of the chambers fell in without allowing sediment to enter the chamber we can see a different and adverse effect on the individual skeletons present in the chamber in the form of more pronounced fragmentation; 5) beneath some antechambers pre-existing pits were discovered which were incorporated in the construction of the hypogea themselves; 6) the characteristics of the deposits which fill the pre-existing pits are very distinct from the fill of the hypogea and contain some artefacts; 7) the function of these pits, some of considerable depth (the average depth is 2,07m, the deepest being 2,42m deep and the shallowest 1,06m) perhaps allowed an evaluation of the substrata with a view to constructing the hypogeum. In other words, these pits could have been used as test pits to determine the opening of a funerary chamber; 8) the entrance to the funerary chambers were found untouched.

adults and just 4 sub-adults buried in the hypogea. The analysis of the positions of the skeletons found in primary position revealed that 10 of the skeletons were found lying on their right side, 7 on their left side, two in a supine position with their limbs bent to the left side and just one skeleton lying prone. Given that in a total of 14 hypogea, there are 17 funerary chambers, the number of individuals per chamber is as follows: 9 chambers contained individual inhumations, the other 8 holding between 2 and 5 individuals. In 4 chambers funerary offerings were not recovered, those were the chamber of H102, which contained 5 individuals, chamber 2 of H118, which contained one sub-adult, the chamber of H156, containing one adult, and chamber 2 of H159 which contained 3 individuals. In the other 13 chambers various artefactual elements, such as ceramic vessels and metallic objects were found. Regarding the animal remains that constituted offerings, they corresponded to specific anatomical parts of a unique animal, in some cases the bones being found complete and in others still articulated. Ceramics were the most common offering, represented by a collection of 31 morphologically diverse vessels, with open and closed forms. The metal objects found were awls and daggers, probably in copper.

Regarding the burial levels themselves (figure 6), a first look at the data shows that the primary burials are of 18 adults: 5 males, 9 females, 4 individuals of undetermined sex; and 2 sub-adults. The secondary burials are of 12 adults: 2 males, 4 females, 6 individuals of an undetermined sex; and 2 sub-adults. In total there were 30

Faunal remains were detected in just 6 chambers associated with single burials, with the exception of chamber H59 where 2 associations were identified. One was a human skeleton in primary position, the second and

Figure 5. Types of hypogea identified in Montinhos 6.

36

Figure 6. Summary of all burials in Hypogea (all drawing are of the same scale, oriented to the north). ossuary. Besides this, a fragment of human bone, a right ulna (the proximal half) of a sub adult was found associated with an adult in chamber 1 of H118. The fragment belonged to a sub-adult which was found in another chamber of the same hypogeum.

associated complete ceramic and metallic objects. The faunal remains correspond to complete parts of the forelimbs of sheep/goats and domesticated cattle, the sheep represented by their right side, the cattle, the left, with the exception of the partial skeleton of the rabbit. Individual and collective burials of numerous combinations were also found

The faunal remains were made up of the forelimbs of Bos taurus (domestic cattle) and Ovis/Capra (sheep/goat) and the partial skeleton of a rabbit. With the exception of the skeleton in H169, the faunal remains were mostly associated with female skeletons. The spatial associations of the faunal remains in relation to the human elements are diverse and they may appear at the feet of a human skeleton (as in H59), with the human skull (as in H155 and H169) or between the thoracic region of the human skeleton and the wall opposite the entrance of the hypogeum (as in chamber 2 of H118 and chamber 2 of H159). There is also an association of faunal remains and an ossuary in H59, making it the only hypogeum where within the same chamber two faunal associations were found. The remains of the rabbit were found in the interior of chamber H153, along with the other bones, the majority of which belonged to an adult female. These bones were found disarticulated and scattered over the base of the chamber indicating post-mortem events that are difficult to explain.

Pits Out of the large number of 130 pits, just 7 contain human burials from which 9 individuals were exhumed (figure 7). Within the group of pit burials (FE) the occurrence of one or two individuals in varied positions was noted. The dimensions of the pit (diameter of the opening, maximum diameter and depth) are variable; all have sub-circular plans and have open and closed forms, and in the majority of cases have cylindrical bodies and a flat base; in 3 examples they contain small niches. Just 3 pits with burials have faunal remains in their interior (FE77, FE144 and FE152), which correspond to the pits with niches, and FE77 which is the only closed structure made up of fragments of quern stone. Out of the structures containing burials, just FE152 contains a context in which we can consider the association of fauna as a funerary practice. The first burial, corresponding to a child of 7-10 years old at the time of death, was found in the niche. Given the small size of the space within the niche, the corpse (specifically the torso) was deposited partially in the area

In summary, the specific architecture of the hypogea is defined through funeral practice. The majority of human remains correspond to adult individuals with which are 37

Figure 7. Sections of the pits with human burials. of the pit. In the top of the deposit which covered the deposition a concentration of faunal remains which we considered to be in association with the burial was identified (figure 14). The second burial is a sub-adult of 14 years of age at the time of death and was found in the area of the pit about 30 cm above the first burial, against the opposite wall occupied by the previous inhumation. Associated with this burial was a bone bead and 4 perforated shells.

the sheep/goats, and radii in articulation with the ulnae and respective carpus in the case of the Bos taurus. The radii of the sheep/goat were found associated with just two human skeletons in hypogea H59 and H159, while the bones of the bovine were found among the votive artefacts of the ossuary of H59 and the skeletons of H118 and H155. Both the remains of Bos taurus and the Ovis/Capra show cut marks at the joints suggesting intentional disarticulation of those parts of the bodies, all belonging to sub-adult animals. Lastly, the partial skeleton of a rabbit was found associated with the remains of a female individual placed in the interior of the funerary chamber of hypogeum H153.

Faunal remains in Bronze Age burial contexts Hypogea

The arrangement found in H153 reveals some particularities. On the one hand, the remains of a single skeleton of an adult female were recovered grouped into small clusters. Just some parts, the left shoulder blade, thoracic and lumbar vertebrae and the sacrum were found in articulation (Rodrigues 2011: 43). In the centre of this, a partial rabbit skeleton was recovered, made up mainly of elements of the upper and lower limbs and some vertebrae. As the ceiling of this chamber had collapsed, we should not rule out the idea that the presence of the rabbit is intrusive (and the dispersion of the human remains is due to post depositional processes), but in reality there was no stratigraphic evidence to unequivocally support this hypotheses. Considering the fact that the osteological remains, of human and animal, were found at the same height in the base of the funerary chamber (almost two meters from the surface), the sediment which fills the pit provided little archaeological

Of the 14 hypogea identified, just 7 faunal assemblages were recovered with a total of 53 elements, exclusively of mammals. The fauna found in the interior of the deposits was found directly associated with human burials, with 2 exceptions, one an unidentified animal of medium size, the other a sheep/goat which was found in the bottom of the antechamber of hypogeum H102, along with 2 elements of rabbit found in the sediment that covers the burials in the interior of the funeral chamber of H155. The range of faunal remains associated with the human burials contains just 3 species: domestic cattle (Bos Taurus), sheep/goat (Ovis/Capra) and rabbit (Oryctolagus cuniculus) not including the bones from an unidentified animal which was comparable in size to a sheep or goat. The types of bones appearing were radii, in the case of Species

H59

H102

H118

H153

H155

H159

H169

Total

Bos taurus

2

-

6

-

6

-

-

14

Ovis/Capra

5

1

-

-

-

1

-

7

Oryctolagus cuniculus

-

-

-

28

2

-

-

30

1

-

-

-

-

1

2

2

6

28

8

1

1

53

Midsize animal Total

7

Table 1. Frequency of faunal remains associated with hypogeal.

38

Species Bos taurus

H59

H118

H153

H155

H159

H169

X

X

-

X

-

-

Left ulna

X

-

X

-

-

Left scaphoide

X

-

X

-

-

X

-

X

-

-

Left pyramidal

X

-

X

-

-

Left pisiform

X

-

X

-

-

Left radius

Left lunar

Ovis/Capra

Midsized animal O. cuniculus

Partial skeleton

X

Und carpal

X

-

-

-

-

-

Right radius

X

-

-

-

X

-

Right ulna

X

-

-

-

-

-

Right scaphoide

X

-

-

-

-

-

Radius

-

-

-

-

-

X

-

X

-

-

-

-

Table 2. Association of faunal remains to the burials contained in hypogea. material, and all the bones exhibit the same degree of erosion on their surfaces (pointing to the contemporaneity of the deposition of human and animal), it seems pertinent to accept the possibility that this was an intentional association.

The case of hypogeum H59 is also strange insofar as in the interior of the pit that pre-dates the hypogeum (found under the antechamber) 32 fragments of dog remains from a single animal, along with 2 fragments of Ovis/Capra were found. The specific bones of the dog found were the metatarsals and metacarpals, first and second phalanges of both sides, a caudal vertebrae, skull bones and several teeth. Clearly without anatomical articulation, the bones were integrated in the deposit complete and without marks of post mortem manipulation, and therefore the intentional nature of their presence also cannot be ruled out, although it is difficult to interpret. The dog is a species which only appears in Montinhos 6 in a group of articulated bones in pit F115 (Costa 2013). But, in the context of structured depositions or groups of articulated bones identified in the other pits of Montinhos 6, it seems that there is a special relevance to the deposition of limbs and/or elements of the paws of deposited animals, as the existence of the front and hind limbs of a dog in the fill of a pit predating a hypogeum is not exceptional in the context of Montinhos 6. Effectively, the incorporation of faunal elements in a pit predating the construction of a hypogeum has a parallel in pit F49, which predates hypogeum H49. Faunal material in this pit was not found associated to human remains, but some elements of Ovis/Capra and remains of an animal of similar size were present.

As for the two elements of rabbit which were found in the deposit covering the burials in the interior of the funerary chamber of hypogeum H155, their presence is more difficult to explain. On one hand, the two elements are loose and without any anatomical connection, suggesting perhaps that they had formed part of the deposit that filled the chamber and perhaps their presence was not intentional. Also, following the caving in of the roof of the chamber the cavity filled naturally with sediment and this could also explain some foreign elements in the deposit. On the other hand, as we have already seen, rabbit is a species which appears to be associated to funeral rites in hypogeum H153 so the intentionality of the association of these remains should not be ruled out. The case of H102 is slightly different from the other hypogea because in the interior of the funerary chamber various alternating levels of ossuarys and an articulated human skeleton were found. As for faunal material, a fragment of metacarpal bone from an Ovis/Capra and a fragment of a medium size undetermined animal were found in the deposit at the base of the antechamber, however, they are not unequivocally part of the funerary ritual, as they may have just made part of the deposit. Effectively, the funerary chamber was reutilised successively and if these remains made part of offerings, after repeat reuse of the space, could have been removed from their original position.

Given the above, and limiting ourselves to the finds unequivocally associated with human remains in the hypogea, it seems as if only the Bos taurus and Ovis/Capra forelimbs, and the partial skeleton of the rabbit do not leave us with any doubts as to their participation in funerary practice (figures 8 to 13). A 39

Figure 8. Hypogeum 59 - Two votive sets with faunal remains: Bos Taurus associated with an ossuary (on the left); Ovis/Capra associated with a female skeleton (on the right). summary of the faunal associations can be found in tables 1 and 2.

is related to the burial of an non adult individual who was deposited in a fetal position in the niche and the group of artifacts present was composed of elements of Ovis/Capra, of at least three young animals, constituted by elements of the upper apendicular skeleton and some vertebrae of an unidentified animal of similar size to Ovis/Capra, grouped and deposited near the abdominal area of the human skeleton.

Pits Of the seven pits which contained primary burials, faunal remains were only recovered in three: FE77, FE144 and FE152. In the first two cases, the faunal remains were not found directly associated with the burials, but in the case of FE152, they seem to be associated with one of the burials. The faunal assemblage contains 17 remains made up of Ovis/Capra and an undetermined animal of medium size, represented by the bones of its forepaws.

Final remarks “In other words, identity is not something that people have, an unchanging set of qualities; rather, it is an ongoing act of production – an inherently fluid set of properties under continual construction and revision” (Bruck 2004, 311).

Because of this the case of FE152 is unusual, not just because it contains the burial of 2 individuals, but also because the faunal remains found in the stratigraphic unit UE15207 have a stratigraphic position that suggests a ritual association to the second burial (figure 14 and table 3). The levels UE15204 and UE15205, which underlies the first burial level, both contained faunal remains. However the majority of the finds from UE15204 was composed of unclassifiable fragments. The same also goes for UE15205, which is another heterogeneous deposit whose faunal contents should not be regarded as part of the funeral ritual. However, the deposit UE15207,

The presence of fauna within burial contexts, principally in hypogea, make up part of a series of items, like ceramics or metal, that compose the different scenarios of burial. Although the sample in Montinhos 6 is small, of the 6 cases in which fauna is associated with primary burials (two cases of Bos taurus, two of Ovis/Capra, one rabbit and an undetermined medium sized animal), 5 are associated with female adults. The Bos taurus remains are limbs from the left side, and from the Ovis/Capra the 40

Figure 9 - Hypogeum 118 – Radius, ulna and carpals, with cut marks, associated to a female skeleton.

Figure 10 - Hypogeum 153 - Partial skeleton of a rabbit among bones from a female skeleton, mostly disarticulated. 41

Figure 11 - Hypogeum 155 – Bos, with cut marks, associated to a female skeleton.

Figure 12 - Hypogeum 159 – Ovis/Capra radius, with cut marks. Associated with a female skeleton. 42

Figure 13 - Hypogeum 169 –Radius diaphysis (of a mid-size animal) associated with a human skeleton of undetermined sex, probably a male.

Figure 14 - Pit FE152 – Burial of a sub-adult individual placed in lateral left decubitus, inside a small wall niche (age at death 10 years old). With faunal remains. 43

limbs correspond to the right side of the body. The intentional choosing of these parts of specific (and preferentially young) animals, many of them deposited with soft tissue intact, reveals an action that clearly had a pattern. Bearing in mind the limits of the sample, we hope that more work of this kind are undertaken that allows a larger analysis at a regional level.

faunal remains, of Ovis/Capra, disarticulated and from a minimum of 3 different animals. The pit burials have parallels in sites of the same county (Torre Velha 3, Outeiro Alto 2 e Torre Velha 12) as in other areas in Alentejo (Antunes et alli 2012), like Horta do Albardão 3, Monte da Cabida 3, Casarão da Mesquita 3, Casarão da Mesquita 4, Pedreira de Trigaches 2 and Horta de Jacinto. Of the sites listed, only in pit 53 of Outerio Alto 2 (Costa & Cabaço 2012) were faunal remains identified of the species Sus sp. and Bos sp. on the lower back of a human skeleton. However, the chronology of this structure was not determined. In Horta de Jacinto, in Beringel (Baptista et alli 2012) we also see the association of an almost complete animal (Sus sp.) and a human burial, but this context also has several peculiarities, which so far, make it exceptional.

A pattern of association with the same characteristics has been identified in other archaeological sites of the same region from the Bronze Age. Belmeque (Oliveira 1994; Soares 1994) and more recently, Torre Velha 3 (Alves et alli 2010; Porfírio & Serra 2010), Outeiro Alto 2 (Costa & Cabaço 2012; Valera & Filipe 2010) and Torre Velha 12 (Rodrigues et alli in press), in which the study of faunal material is still ongoing. Nevertheless, more than a dozen of these types of site where hypogea have been identified have been located in the county of Serpa. The disclosure of these sites would contribute immensely to the study of the Bronze Age population in the south east of the Iberian Peninsula.

The growth in the amount of fieldwork in the area has shown that the funeral practices attributed to the 2nd millennium BC are very diverse, both architecturally and in the treatment of the human skeleton itself and the artifacts and animal bones that accompany it.

The similarities between the funerary traditions present in this region, seen as part of the Argaric world, with regard to “funerary fauna” in hypogea apparently show a standardization of practices which were seen by most Iberian authors as the result of rituals of commensality which were in one way an animal sacrifice, in another, a way to consume meat, the faunal deposit being a representation of the participation of the deceased in the ritual meal (Alves et alli 2010, Aranda Jimenez & Esquivel Guerrero 2006, 2007; Porfírio & Serra 2010; Sanchéz Romero et alli 2007; Soares et alli 2009). Associated with this practice of feasting was the production of ceramic vessels especially for the occasion, with specific shapes, namely open formed vessels for solid food, closed form vessels for liquids, and cups, well made, and of higher quality, specifically in terms of decoration (which included the intense burnishing of surfaces). The different items, ceramics, metal artifacts, and faunal remains made part of a ritual which worked as a symbolic expression of social cohesion in an asymmetrical and unequal society. Feasting rituals are seen to have been one the practices related to the exercise of power in the Argaric world (Aranda Jimenez 2008, 2011).

Recently, several essays have been addressed to this theme (Baptista 2013; Baptista et alli 2012; Valera & Costa 2013), following lines of research that value the idea of fragmentation/segmentation as social practice, based on the numerous empirical contexts in which humans and/or animals, and other materialities, can arise in an archaeological context in a complete state or in fragments. The site of Montinhos 6 (Baptista 2013; Baptista et alli 2012), stresses that the influences on the Argaric “world” had been as a result of the contact between different supra-regional groups in a network of circulation/manipulation of things (goods, people, ideas) and were not just a direct importation of social organizational models. And, that these influences reflected the web of actions in which they themselves shared and organized similarities and produced rituals in which to negotiate them. Furthermore, the differences in the funerary deposits revealed the fluidity and adaptability of the funeral ritual, which although being standardized, cannot be disassociated from its participants. A fluidity and adaptability which makes us return to the quote with which we began this section, looking at the different relations between the items which show the funerary contexts of Montinhos 6 as invocations of the relational character of the identity of all the participants.

However, neither Belmeque, Torre Velha 3 nor Outeiro Alto 2 or Montinhos 6 (sites which have hypogea that fit into the same chronological period and have information about faunal remains available) have provided bioanthropological or archaeological evidence which permits us to support the hypotheses that different faunal associations can effectively show us social differentiation. However, we should not forget that of all the hypogea excavated in Montinhos 6, a total of 14, just 6 (around 40%) gave evidence for the association of faunal remains.

References Alves, C., Costeira, C., Estrela, S., Porfírio, E., Serra, M., Soares, A. M. M., Moreno-Garcia, M. (2010) “Hipogeus funerários do Bronze Pleno da Torre Velha 3 (Serpa Portugal). O Sudeste no Sudoeste?!”. Zephyrus LXVI : 133-153.

Regarding the pit burials, of the 7 registered, just one showed us evidence of an indubitable association of 44

Antunes, A.S.; Deus, M.; Soares, A.M.; Santos, F.; Arêz, L.; Dewulf, J.; Baptista, L.; Oliveira, L. (2012) “Povoados Abertos do Bronze Final na Médio e Baixo Guadiana” Sidereum Ana II – El valle del Guadiana en el Bronce Final, editado por J. Jiménez Ávila. CSIC, Instituto de Arqueología de Mérida, Anejos de Archivo Español de Arqueología, 2012: 277-308. Aranda Jiménez, G. (2008) - “Cohesión y distancia social. El consumo comensal de bóvidos en el ritural funerário de las sociedades argáricas”, CPAG, 18: 107-123. Aranda Jiménez, G. (2011) - “Nuevos actores para viejos escenarios. La sociedad argárica”, Memorial Luis Siret I Congreso de Prehistoria de Andalucía La tutela del património prehistórico, Junta de Andalucia: 249-270 Aranda Jiménez, G., Esquível Guerrero, J. A. (2006) “Ritual funerário y comensalidad en las sociedades de la edad del bronce del sureste peninsular: la cultural de el Argar”, Trabajos de Prehistória, 63, 2: 117-133. Aranda Jiménez, G., Esquível Guerrero, J. A. (2007) “Poder y prestigio en las sociedades de la cltura de el Argar. El consumo comunal de bóvidos y ovicapridos en los rituales de enterramiento”, Trabajos de Prehistoria, 64, 2: 95-118. Baptista, L. (2013) - “A Idade do Bronze no concelho de Serpa: um primeiro esboço de um conhecimento em construção”, Actas do VI Encontro de Arqueologia do Sudoeste Peninsular, em 4 e 6 de Outubro de 2012, Villafranca de los Barros (Badajoz). Baptista, L., Gomes, S. (2011) - Bloco de Rega de Brinches-Enxoé. Intervenção Arqueológica em Montinhos 6, Arqueologia & Património Lda. Baptista, L., Gomes, S. (2012) - Relatório Final Global. Minimização de Impactes sobre o Património Cultural decorrentes da execução do Bloco de Rega de BrinchesEnxoé – Fase de Obra, Arqueologia & Património Lda. Baptista, L., Gomes, S., Costa, C. (2012) - “As dinâmicas de deposição no sítio pré-histórico de Horta de Jacinto (Beringel, Beja)”, Actas do V Encontro de Arqueologia do Sudoeste Peninsular, Novembro de 2010, Município de Almodôvar: 585-596. Baptista, L.; Gomes, S. (2013) - “Modalidades de construção do espaço das “estruturas em osso”, Actas do VI Encontro de Arqueologia do Sudoeste Peninsular, em 4 e 6 de Outubro de 2012, Villafranca de los Barros (Badajoz). Baptista, L., Pinheiro, R., Rodrigues, Z. (2012) “Espacialidades dos cadáveres em Montinhos 6: contributos para uma compreensão das práticas funerárias da Idade do Bronze no Sudoeste Peninsular”, Actas do V Encontro de Arqueologia do Sudoeste Peninsular, Novembro de 2010, Município de Almodôvar: 149-170. Brück, J. (2004) - “Material metaphors: The relational construction of identity in Early Bronze Age burials in Ireland and Britain”, Journal of Social Archaeology, 4: 307-333. Costa, C. (2013) - Tafonomia em contexto pré-histórico A zooarqueologia como recurso para a compreensão das “estruturas em negativo” da Pré-história Recente. Tese de Doutoramento apresentada à Universidade do Algarve. Texto Policopiado. Costa, C., Cabaço, N. (2012) - “Associação de restos de

animais vertebrados a contextos funerários da pré-história recente: o caso do Outeiro Alto 2”, Apontamentos de Arqueologia e Património, 8: 43-47. Duque, J., Almeida, C. (1998) - “Caracterização hidroquímica do sistema aquífero dos gabros de Beja”, 4º Congresso da Água, FIL, Lisboa, CD-164, consultado em linha em 15 de Janeiro de 2013, http://www.aprh.pt/congressoagua98/files/com/164.pdf Garrow, D., Beadsmoore, E., Knight, M. (2005) - “Pit clusters and the Temporality of Occupation: an Earlier Neolithic Site at Kilverstone, Thetford, Norfolk”, Proceedings of the Prehistoric Society 71: 139-157. McFadyen, L. (2006) - “Material culture as architecture Neolithic long barrows in Southern Britain”, Journal of Iberian Archaeology 8: 91-102. Oliveira J. C. (1994) - “Estudo do Espólio Ósseo de sepulturas do Bronze do Sudoeste.” Atas das V Jornadas Arqueológicas da Associação dos Arqueólogos Portugueses, Lisboa, 1993: 185-186. Porfírio, E. M. B., Serra, M. A. P. (2010) - “Rituais funerários e comensalidade no Bronze do Sudoeste da Península Ibérica: Novos dados a partir de uma intervenção arqueológica no sítio da Torre Velha 3 (Serpa)” Estudos do Quaternário, 6: 49-66. Rodrigues, Z. (2011) - Minimização de Impactes sobre o património cultural decorrentes da implementação do Bloco de Rega de Brinches-Enxoé Intervenção em Montinhos 6. Relatório Antropológico. Texto policopiado Rodrigues, Z., Baptista, L., Gomes, S. (2013) “Contextos de inumação da Idade do Bronze da Torre Velha 12”, Actas do VI Encontro de Arqueologia do Sudoeste Peninsular, em 4 e 6 de Outubro de 2012, Villafranca de los Barros (Badajoz). Sánchez Romero, M., Aranda Jiménez, G., Alarcón García, E. (2007) – “Gender and age identities in rituals of commensality. The argaric societies.”, Interpreting household practices, Barcelona, 21-24 november 2007,Treballs d’Arqueologia, 13: 69-89. Santos, F. J. C., Soares, A. M. M., Rodrigues, Z., Queiroz, P. F., Valerio, P., Araújo, F. (2009) – “A Horta do Albardão 3: um sítio da Pré-História Recente, com fosso e fossas, na Encosta do Albardão (S. Manços, Évora)”, Revista Portuguesa de Arqueologia, 12, 1: 5371. Soares, A. M. (1994) - “O Bronze do Sudoeste na margem esquerda do Guadiana. As necrópoles do concelho de Serpa”, Atas das V Jornadas da Associação dos Arqueólogos Portugueses, Lisboa, 1993: 179-197 Soares, A. M. M.; Santos, F. J. C., Dewulf, J., Deus, M., Antunes, A. S. (2009) - “Práticas Rituais no Bronze do Sudoeste – Alguns dados”, Estudos Arqueológicos de Oeiras, 17: 433-456. Valera, A. C., Costa, C. (2013) - “Animal limbs in funerary contexts in southern Portugal and the segmentation problem”, Anthropozoologica, 48(2), in press. Valera, A. C., Filipe, V. (2010) - “Outeiro Alto 2 (Brinches, Serpa): Nota preliminar sobre um espaço funerário e de socialização do Neolítico Final à Idade do Bronze”, Apontamentos de Arqueologia e Património, 5: 49-56. 45

Other Sources Hidroprojecto; COBA; Hidrotécnica Portuguesa; WS AtKins; ConsulGal; GIBB Portugal (1999) Plano da Bacia Hidrográfica do Rio Guadiana, Volume III.1 – Caracterização Geral [Em linha] INAG. [Consult. Outubro 2012]. Disponível em http://www.inag.pt/inag2004/port/a_intervencao/planeam ento/pbh/pbh04.html, Ministério da Ambiente.

46

In death as in life. Ties between man and animals in the recent prehistory of lower Alentejo: two case studies from Alto de Brinches 3 and Torre Velha 3 (Serpa, Alentejo, Portugal) Eduardo Porfírio1 & Miguel Serra1 1

Palimpsesto - Estudo e Preservação do Património Cultural, Lda. [email protected] & [email protected] Abstract The sites of Alto de Brinches 3 and Torre Velha 3 were excavated by Palimpsesto Ltd. teams, under the ambit of impact minimization associated with the Alqueva dam project (EDIA.S:A). In this paper we will stress two pit structures, one from each of those archaeological sites, with animal osteological remains in anatomical connection. The archaeological excavations carried out by Palimpsesto Ltd at Alto de Brinches 3 and Torre Velha 3 (Serpa, Alentejo, Portugal) were conducted under a specific program designed to assess the archaeological impacts caused by the construction of Alqueva dam (EDIA SA) irrigation systems. The fieldwork was developed in both sites revealing a large number of archaeological contexts which are dated to different historic and prehistoric periods. In each site, pits were found containing articulated animal bones. At Alto de Brinches 3 the pit [664] revealed at least two different episodes of animal deposition. Firstly, a canid was deposited and subsequently a wider range of articulated animal bones, e.g. a pig and another canid. At Torre Velha 3, the analysis of the filling sequence of pit [2411] revealed a human burial in flexed position together with a human ossuary and, in an upper layer, a suid burial. The aim of this paper is to present these two cases, currently under investigation, where the analysis of the faunal remains will play a central role, both in the understanding of the deposition conditions and the symbolic component involved with these practices. Lastly, a brief framework of similar occurrences identified in neighbouring sites from recent prehistory. Case studies like this from Alto de Brinches 3 and Torre Velha 3 contribute to better characterize the relationships between human communities and their animals. Ties like the ones reported here show a high complexity level. A more traditional economist's point of view sees animals linked exclusively as a source of food, raw materials and means of transport. The study of ritual manifestations of this kind, combined with the analysis of associated artifactual assemblages plays an important role to characterize the symbolic practices and the daily life of these prehistoric communities. Keywords: Recent Prehistory from Southern Portugal; depositions of animals; Pits; Ritual. schist soils found in the Serpa’s Hills, located to the south and with the residual reliefs from the Ficalho’s Hill in the NW. Besides these two geomorphological accidents the homogeneity of the landscape is only interrupted by the Guadiana River valley and some of its affluent such as the Enxoé Stream (AAVV, 2002).

Introduction The archaeological sites of Alto Brinches 3 and Torre Velha 3 were excavated by various teams of archaeologists from Palimpsesto Ltd 1 in the ambit of assessment and minimization of impacts on cultural heritage required due to the construction works associated with the Alqueva dam under the responsibility of EDIA S. A. Both sites are located in Alentejo’s hinterland, in the peneplain on the left bank of Guadiana River, in Serpa’s county. The region’s relief is formed by an almost unending succession of low hills with smooth slopes, with a maximum high around 200m. The high agricultural potential of these clay soils, spread out around the city of Serpa and its surroundings, has made this area known as Serpa’s fields in opposition with the

The archaeological contexts identified in Torre Velha 3 occupied part of the summit and especially the gentle rising slopes with a maximum of about 180m height. In general the slopes are not very prominent, except those facing north and west that have greater gradients, although without enough expression to counter the open nature of this region. This hill does not stand out particularly from the surrounding countryside, because it is surrounded in the south and east by higher points, which reduce its visual field into the immediate surroundings. The area where the archaeological intervention was carried out is located on the slope facing the left edge of the Brook of Laje, which encircles the site in the north and east quadrants (Alves et al. 2012).

1 The archaeological excavation in Alto de Brinches 3 was directed by the authors in association with Catarina Alves and Susana Estrela, in Torre Velha3 the direction team was formed by the names above mentioned and also Catarina Costeira. 47

Figure 1. Torre Velha 3 and Alto de Brinches 3 location in the Iberian Peninsula. The human occupation identified on Alto Brinches 3 occupied the top of a smooth hill, with a maximum altitude of 190m, whose slopes are crossed by several watercourses, which drain into the Retorta and Carelinha brook’s. These watercourses, with a quite irregular flow, are both subsidiaries of Guadiana River. The visual dominance of this site is limited to the south and east by the elevations where actually are sited the city of Serpa and the geodesic tower of Atalaia da Torre (Alves et al., in press b).

Of this large universe two structures will be highlighted in detail one from each archaeological site, characterized by the presence of osteological remains in anatomical connection belonging to various animals. It should also be noted that the context of Torre Velha 3 also has a human burial, held at an early phase of the process of filling of the structure.

Torre Velha was already cited in the Portuguese archaeological bibliography due to its numerous finds from the Roman period correlated with a villa, the same happened with the site of Alto de Brinches with the occurrence of various materials from recent prehistory (Lopes et al., 1997, numbers 129 and 148). However, the archaeological excavations carried out in those two sites demonstrated the existence of an occupational dynamic characterized by a long diachrony, witnessed mainly by contexts dated to the Chalcolithic and the Bronze Age, and others from proto-historic and historic periods (Alves et al., 2010, in press a and b). The archaeological contexts identified correspond mainly to structures excavated in the geological substratum, characterized by a great morpho-typological and functional diversity, with dwelling and funeral character.

Alto de Brinches 3 – Structure [664]

Archaeological contexts

When structure [664] was identified it had already been partially affected by the excavation of a contemporary ditch, which caused the removal of part of the filling located near the west wall of the pit. Thus, except for the first deposit which was fully preserved, none of the other was preserved in its integrity. Structure [664] showed nearly vertical walls, with a depth ranging between 120 and 150cm. The contact zone between, the pit walls and the base had a slight curvature, although the transition from the South wall to the base was characterized by some irregularity. Both the upper opening and the base have a sub-circular shape, in the first case with a diameter of 221.5 cm and in the latter 228.57 cm. The base is slightly flattened. 48

Figure 3. Location of Alto de Brinches 3 in the Ordinance Survey Map, Sheet n.º 532, Scale1:25 000.

Figure 2. Torre Velha 3 Location in the Ordinance Survey Map, Sheet n.º 523, Scale 1:25 000.

Torre Velha 3 – Structure [2411]

The filling process of this structure began with the deposition of [644] a reddish-brown sandy-silty deposit with some granite blocks arranged horizontally, particularly concentrated near the pit walls. Immediately underneath this layer was identified the deposition of a canid in anatomical connection [642], his skull was placed next to the West pit wall with the rest of the body stretching South. The animal was placed in right lateral decubitus following a West-East orientation.

The structure of Torre Velha 3 shares some morphological similarities with that from Alto de Brinches 3, which are especially evident at the level of the sub-circular shape of the upper opening, the tendency for verticality of the pit walls and a certain flattening of the base. Regarding its dimensions, the structure [2411] has a preserved depth of 160cm, whereas the diameter of the top opening is 173cm. As in Alto de Brinches 3, the depositions of animals were not done directly on the base of the structure, but above a preparatory level. However, the similarities stop here, because the filling sequence of the structure [2411] shows a more complex reality, which brings together in the same space, but in different moments, two burials, one human and another of a suid.

The burial was covered by [633] a brownish-orange sandy-clay deposit with many granite pebbles and stones placed horizontally. Under this sediment there’s a second episode of animal deposition, more numerous and complex. A cluster of osteological remains in anatomical connection was identified [607], and the better preserved bones are from a canid and a suid, both deposited in right lateral decubitus. The first animal was placed in a WestEast orientation and the later was deposited in a SouthNorth alignment.

The filling process began with the deposition of layer [2404], a grey sandy sediment, which was followed by an orange-brown sandy-silty deposit, with numerous granite rocks of small and medium size [2380]. Above this layer was placed the burial [2336], which corresponds to an adult woman, above 30 years old, buried without any associated grave goods. The body was placed near the south wall of the pit in foetal position, following a WestEast orientation, with the skull resting on the right side (Ferreira, 2009).

Maintaining the pattern observed in the first deposition recorded in this structure, the bodies of the animals were also covered with a reddish-brown, compact earth with multiple blocks of granite, [529]. The remaining stratigraphic sequence consists of the deposits [512] and [116]. The first consists of a light brown clay sediment with numerous chalky nodules, which differed markedly from the second layer, much more gritty and with a dark brown coloration (Alves et al. 2010).

Shortly after the inhumation, since the osteological elements of the individual [2336] present evidence that the decomposition process was conducted in a closed space (Ferreira, 2009), the burial was covered with a brown-orange sandy-silty sediment, very compact [2302], with numerous chalky nodules and some small granite rocks. Subsequently, the burial was sealed almost entirely with a structured group of large granite slabs, whose 49

Figure 4. Structure [664] of Alto de Brinches 3.

Figure 5. Canid deposition [642]. Alto de Brinches 3.

interstices were filled by a reddish sandy-clay very compact, [2265]. The filling process of the structure continued with the deposition of a layer of compact black clay, with much granite rubble and chalky nodule [2197]. On this clay level was identified an ossuary [2173] consisting essentially of various osteological elements from a human skull, namely part of the cranial vault. The ossuary was covered by clay sediment mixed with rubble and chalky nodules [2096], above which were deposited a suid, [2095]. Despite the bones poor state of preservation and their high fragmentation and clay compression it seems believable that it corresponds to the whole animal.

Figure 6. Cluster of animal depositions [607]. Alto de Brinches 3.

The structure was sealed with the deposition of two layers with a clay matrix [2058] and [458] (Alves et al. 2009).

fragments, linked to its low morpho-typological representativity, if not completely invalidates any chronological assignment, often contributes to the inclusion of a specific context within temporally large chrono-cultural periods. This situation applies to structures [664] and [2411], whose archaeological material only enables its chronological insertion in a period corresponding to the recent pre-history. However, it should be noted that in relation to the structure of Torre Velha 3 its absolute dating is planned and will be published as soon as possible. Moreover, the lack of a zooarchaeological study of the faunal remains exhumed limits our knowledge of the pre and post-depositional circumstances of the presented animal depositions. It is not possible to present an exhaustive characterization of the species found and of their skeletal representation, as well as information regarding the minimum number of individuals present, age, cause of death, sex and other osteological descriptors. Therefore, all information given is based primarily on observations recorded during the fieldwork and during some office work, giving this paper a preliminary character.

Discussion Prior to the discussion of the archaeological contexts presented, it is necessary to consider several elements, which somehow curtail a more comprehensive approach to these realities. Firstly, the absence both in the Torre Velha 3 and in Alto de Brinches 3 of preserved stratigraphic sequences outside the negative structures, that allow interrelating them spatially and chronologically. It should also be noted that these sites were re-occupied, with or without chronological gaps over several cultural-historical periods, a situation which is often embodied in an high density of structures which intersect each other, a situation that if on one side contributes to its relative dating, on the other, leads to the disruption of contexts and the remobilization of archaeological materials. Due to the absence of absolute dating, all the negative structures assigning chronology depends on the archaeological materials collected in the fillings. In many cases the small quantity of artefacts, namely ceramic 50

Figure 8. Human ossuary [2173].Torre Velha 3.

Figure 7. Structure [2411] of Torre Velha 3, stratigraphy, and burial plan [2336]. Despite the constraints mentioned above, we believe that the structures [664] and [2441] are very interesting case studies that deserve to be presented, thus this became the aim of this work.

Figure 9. Suid deposition [2095]. Torre Velha 3. In its turn, the stratigraphic sequence of structure [664] reveals a universe of very different characteristics, defined primarily by the deposition, spaced in time and differentiated, of various animals. The artefactual component proves to be quite scarce, registering only 23 ceramic fragments spread across five layers. The distribution of the fragments indicates a degree of randomness, which probably reflects the fact that its presence is due to their inclusion in the sediments that filled the structure.

The intentional nature of the animal’s depositions held in structures [664] and [2411] is, in our view, clearly evident in the careful disposal of the animal bodies, namely in the grouping two by two of the hind limbs, but also in a certain horizontality, which characterizes the disposal of carcasses. Another element to be taken into account relates to the fact that the animal were never buried directly on the pits floors, but always on a filling layer previously deposited. On the other hand with the exception of the suid [2095] from Torre Velha 3 all burials of animals were later covered by pebbles and gravel, a more or less dense stony agglomeration that materialize an intentional act of sealing the deposition. In a similar way, the human burial [2336] in Torre Velha 3 was also deposited on a preparatory filled level, which in turn was closed by a pavement with large slabs of granite.

In Alto de Brinches 3 two other structures revealed animal depositions, unfortunately this study is actually in a very early stage. But it deserves to be mentioned, even briefly. In structure [275] was identified a partial deposition of a canid (probably) and a cervid. The latter, presents parts of the antler, of the spine and of long bones in articulation. The partial bodies of both animals were deposited on a level of granite stones, ceramic sherds and river pebbles, and were covered by three layers of sandysilty and clayey sediments (Alves et al. in press b).

In terms of archaeological remains it was verified that the high frequency of pottery sherds collected in the structure [2411], clearly distanced them from the remaining prehistoric funerary contexts of Torre Velha 3 held in pits, all of them poor in archaeological remains (Alves et al. 2009). This seems to make structure [2411] quite exceptional, a fact that certainly must be related to its impressive burial record.

Structure [586] exhibit a human burial and an animal deposition. Both depositions took place after the deposition of two layers of sandy-silty sediments characterized by the high frequency of stones, pebbles 51

and ceramic sherds, intercalated with two layers of sandy and sandy-silty sediments which present reddish and orange colours (Alves et al. 2010). The human body was placed near the north wall of the pit in foetal position. Despite the poor preservation of the bones, the osteometric analysis point to the fact that the skeletal remains belong to a middle age or even an elderly woman (Rodrigues et al. 2012: 78-79). In clear contrast to the human burial, the faunal remains (probably belonging to a canid) occupied the center of the pit, and the most remarkable is the fact that it was encircled by granite stones conforming an authentic burial structure (Alves et al. 2010; Inocêncio, 2013; Inocêncio and Porfírio, in press). It seems that the traditional and supposed roles assigned to men and animals are reversed. Unfortunately the poor preservation of the faunal remains, and the absence of their study constraints the discussion of these contexts, for this time (Alves et al. 2010; Inocêncio, 2013; Inocêncio and Porfírio, in press).

The presence of isolated animal bones in funerary contexts is also recorded in the Chalcolithic site of Carrascal 2 (Ferreira do Alentejo), in tomb 5 of Sobreira de Cima (Pedrogão, Vidigueira) and for the hypogea of Outeiro Alto 2 (Pias, Serpa) dating from the Late Neolithic (Valera, 2009; Valera and Filipe, 2012; Valera et al., in press a). In hypogeum 1 of Carrascal 2 the existence of human burials was associated with isolated faunal remains (long and short bones of various animals). In the ditch/atrium located in front of Hypogeum 2 was excavated an ossuary with human bones belonging to adults and non-adults, associated with faunal remains and a ceramic container (Valera et al., in press a). At the site of Sobreira de Cima in the funerary hypogea 1 and 5 was recorded the presence of goat/sheep phalanges (Valera, 2009), which replicates the situation found in the ritual of tombs 4/5 of Outeiro Alto 2 (Valera and Filipe, 2012). Regarding the deposition of partial animal skeletons in funerary contexts, various cases were identified such as of Porto Torrão and Monte do Cardim 6, both located in Ferreira do Alentejo and Monte das Covas 3 (São Matias, Beja). In the first case, there is a preliminary reference to a deposition in a pit, corresponding to the spine bones of a large mammal (Valera and Costa, in press). In the Middle/Final Chalcolithic tholos of Monte do Cardim 6 was identified under the layer corresponding to the falling of the stony cover, part of a young goat/sheep, with skull bones, cervical vertebrae and a forelimb in articulation (Valera et al. in press). In pit [704] of Monte das Covas 3 were identified osteological remains of 16 individuals (MNI) which were associated with various faunal remains, some of which in connection, including the partial skeleton of a canid (Miguel and Godinho, 2009).

Typologically the ceramic sherds from structure [275] and [586] are consistent with the Chalcolithic period (Rodrigues et al. 2012: 78, Inocêncio, 2013; Inocêncio and Porfírio, in press). In the study area2, contexts like these presented here and other with similar features, have occurred mainly at sites characterized by the presence of pits. These sites have been identified during the implementation of heritage safeguard measures resulting from projects associated with Alqueva Dam or the Highway of Baixo Alentejo. The publication of the results from numerous archaeological interventions has contributed to change, in a revolutionary way (Valera et al., in press b), the state of our knowledge about the communities that occupied the Alentejo during recent prehistory.

It has been already mentioned that the deposition of a total animal skeleton in the same burial container used for human burials was rare (Valera and Costa, in press), and thus beyond the structure [2411] of Torre Velha 3 and the structure [586] of Alto de Brinches 3, it was also identified at another site, Horta do Jacinto, located in Beringel, near Beja (Baptista et al., 2010). In the latter site on the base of structure number 1 had been deposited a pig over a ring of stones, subsequently, after the deposition of a layer consisting essentially of chalky debris was carried an human burial of a juvenile (9-12 years old at the time of death). The body was laid out sitting on the feet, in a pit dug in one of the layers of the structure n. º 1. The burial was then covered by a thick deposit of stones. The archaeological remains collected consisted of ceramic fragments and lithic elements, in the last case, two elements of the same grindstone had been identified, although found in different layers. The faunal analysis of the pig bones revealed that it was a juvenile (9-16 months old) missing both ends of the forelimbs. Cut marks were not identified, although the bone’s preservation does not favour their identification. Typologically the pottery and the lithic elements collected in the fillings of structure 2 from the same site, are consistent with the 2nd Millennium AD (Baptista et al., 2010).

In the Lower Alentejo there are other cases where in the same pit, human and animals were buried. A first survey of these cases was conducted by António Carlos Valera and Claudia Costa. This phenomenon is characterized by a certain variability including various situations: such as the existence of animal bones alone or mixed with human remains, through the more frequent presence of partial skeletons in tandem to the rarer case which relates to the deposition of the whole animal (Valera and Costa, in press). Regarding the first case we can, therefore, refer the frequent presence of the distal ends of animals legs such as those identified in the Bronze Age hypogea of Torre Velha 3 (Alves et al., 2010), and Belmeque (Soares, 1994), or in Outeiro Alto 2 (Valera and Filipe, 2010; Filipe et al. 2013) and Montinhos 6 (Baptista et al., 2012). These realities emphasize the existence of burial practices involving ritual feasting similar to what happens in the Argaric world (Alves et al. 2010). 2

Although we use in this paper the designation “Lower Alentejo”, our study area corresponds more precisely to Beja district. 52

However, there are depositions of faunal remains that are not associated with human burial environments, among these situations are included not only the burial of animals but also the deposition of isolated bones or partial skeletons of animals. Associated with the first case is pit 5 of Corça 1 (Brinches, Serpa), where in one of the fillings was identified the skeleton of a canid in anatomical connection (Valera et al. 2010). In the same level of the animal body were also identified manual ceramic fragments, various lithic elements including one of an axe, as well as other animal’s osteological remains. Regarding the latter, the zooarchaeological study identified a dog, the presence of deer, domestic ox, pigs and sheep/goat. The osteological remains of the dog burial belong to a young individual (between two and 11/12 months) (Valera et al. 2010). In Ourém 7 (Pias, Serpa) the complete skeleton of a small dog was excavated, deposited on the initial filling of a pit. The few fragments collected may not allow a categorical assignment of this context to the Bronze Age, however, other structures intervened on this site provided materials characteristic of that chronology (Valera et al., in press b).

variable number, which can easily reach hundreds. These are morphologically varied structures (often referred to as pits, silos, cabin bases) that are conspicuous by the absence of vertical stratigraphy, which relate them directly to each other. Flat areas near water lines and good agricultural soils are favoured for settling. Chronologically this type of settlement is known since the Neolithic and with smaller or larger gaps occupation endured until the Middle Ages, reaching the highest numerical expression during the Bronze Age (Blasco Bosqued, 2004: 350). Another characteristic of the pit settlements present since the Neolithic times relates to the relatively uncommon presence of contexts with faunal remains, corresponding both to partial or complete skeletons (Liesau, 2012). In the Meseta these realities have been studied regularly for some time ago and three major groups that summarize the main features of the deposits with faunal remains from the Chalcolithic and the Bronze Age have been defined. Thus, non funerary deposits have been recognized, containing one or more species, associated with funerary deposits and votive deposits characterized by the presence in combination, or not, of isolated bones from animal, human bone and other artefacts with special characteristics (Liesau, 2012: 224).

More sparse and of a preliminary nature are the references of the identification of faunal remains of total or partial skeletons of animals in several open settlements of the Bronze Age in the Lower Alentejo region, as Pedreira de Trigaches 2 (Trigaches, Beja) were it was recorded the deposition of a deer (Antunes et al. 2012). In Salsa 3 (Santa Maria, Serpa), pit 2 revealed the osteological remains of a bovid associated with several ceramic fragments belonging to a large storage container (Deus et al. 2010). At the site of Cadavais (Brinches, Serpa) it was identified in trench 3 a pit that contained the deposition of a skull of a large animal and other faunal remains in the levels close to its base. Previously to the filling of this structure a large slab, stopped with stone blocks was placed vertically along its wall. The few materials collected in association with the faunal deposit, consisted of a fragment belonging to a flat bottom conical container, which may indicate a chronology of the Bronze Age (Valera et al., in press b).

Concerning the species present in these contexts, there is a predominance of domestic animals, including cattle, dogs, pigs and a smaller percentage of sheep/goat (Liesau, 2012: 224). The interpretations advanced for the deposits of animals in pit settlements have accentuated the ritual character of these manifestations, based mainly on the reduced numerical expression of these contexts when compared with other structures from these sites, but also for its frequent association with deposits with numerous remains of coal or greyish sediments. Another feature very highlighted is the distribution of the animal skeletal remains within the pits, preferentially occupying half of the space available (Liesau, 2012: 245 and frame n. º 2, p. 229-232). The similarities between the open settlements from the Guadiana basin and the designated pit settlements have been noted on several occasions (Alves et al., 2010: 133; Antunes et al., 2012; Soares et al. 2009: 440 and 446). These similarities are recognized, for example, at the level of site location that privileges flattened topographical areas along water lines, but also the very conditions of preservation of the archaeological record, since in the overwhelming majority of cases only the structures excavated in the geological substratum were preserved. Chronological similarities, as this model of settlement started during the Neolithic, was repeatedly reused during other prehistoric periods and even historic times until the Late Antiquity (Alves et al. 2012).

If within Alentejo the presence of such sites and contexts was a novelty at the time of their identification, the same does not occur in the neighbouring country, where pit settlements have marked for a long time presence in Andalusia and especially in the Meseta. Sites with the same characteristics are also beginning to be known in Extremadura, as showed with the archaeological intervention conducted in El Carrascalejo, Badajoz (Enríquez Navascués et al., 2007). The great frequency and geographical spread of this type of habitat transforms it into an authentic paradigm within the settlement schemes of the Late Prehistory not only in the Iberian Peninsula (Blasco Bosqued, 2004: 350), but also in other European areas (Valera et al., 2010: 12).

Another aspect seems to relate the sites in Alentejo with their spanish counterparts, the presence, very residual until now, of ceramics that can be associate with the Cogotas or Proto-Cogotas universe such as a fragment

As with the sites in Alentejo that we have been addressing the pit settlements are characterized mainly by structures excavated in the geological substrate in 53

collected at Casarão da Mesquita 3 (São Manços, Évora), which is part of a carenated container decorated with incised lines and printed circles correlated with decorative schemes of the Proto-Cogotas tradition. The best parallel for this piece can be found on the above mentioned site of El Carrascalejo (Santos et al., 2008: 73). Another example, albeit in Late Bronze contexts, comes from the settlement of Passo Alto (Vila Verde de Ficalho, Serpa) and corresponds to a ceramic fragment associable with the formal and decorative Cogotas universe (Soares et al., 2010: 547).

Construção do Reservatório Serpa-Norte (Serpa). Relatórios Palimpsesto 2010. Alves, C.; Estrela, S.; Porfírio, E.; Serra, M., (in press b) – Intervenção Arqueológica no sítio de Alto de Brinches 3 (Reservatório Serpa – Norte): Resultados Preliminares, Actas do IV Colóquio de Arqueologia do Alqueva - O Plano de Rega (2002-2010), Beja 2010. Antunes, A. S.; Deus, M. de; Soares, A. M.; Santos, F.; Arêz, L.; Dewulf, J.; Baptista, L.; Oliveira, L. (2012) Povoados abertos do Bronze Final no Médio Guadiana, Sidereum Ana II – El río Guadiana en el Bronce Final, Anejos de Archivo Español de Arqueologia, LXII, p. 277308. Baptista, L. (2010) – The late Prehistory of the watershed of the Ribeiras do Pisão e Álamo (Beja, South Portugal): a research programme, Journal of Iberian Archaeology, 13, p. 69-84. Baptista, L.; Gomes, S.; Costa, C. (2010) - As dinâmicas de deposição no sítio pré-histórico de Horta de Jacinto Beringel, Beja), in DEUS, M. de, Actas do V Encontro de Arqueologia do Sudoeste Peninsular (Almodôvar, 18 a 20 de Novembro de 2010), Município de Almodôvar, Almodôvar, p. 585-595. Baptista, L.; Pinheiro, R.; Rodrigues, Z. (2012) – A espacialidade dos cadáveres em Montinhos 6: contributos para uma compreensão das práticas funerárias da Idade do Bronze do Sudoeste Peninsular, in Deus, M. de, Actas do V Encontro de Arqueologia do Sudoeste Peninsular (Almodôvar, 18 a 20 de Novembro de 2010), Município de Almodôvar, Almodôvar, p. 149-170. Blasco Bosqued, C. (2004) – Los poblados ribereños de «hoyos» en el entorno madrileno. Un modelo de asentamiento de la Edad del Bronce peninsular, in García Huerta, M. del R. and Morales Hervás, J., La Península Ibérica en el II Milenio A.C.: poblados t fortificaciones, Ediciones de la Universidad de Castilla-La Mancha, Cuenca, p. 349-387. Deus, M. de; Antunes, A. S.; Soares, A. M. M. (2010) – Salsa 3 (Serpa) no contexto dos povoados abertos do bronze Final do Sudoeste, in Pérez Macías, J. A. and Romero Bomba, E., IV Encuentro de Arqueología del Suroeste Peninsular, (Aracena, 27 a 29 de Novembro de 2008), Universidad de Huelva Publicaciones, Huelva, p. 514-523. Enríquez Navascués, J. J. and Drake Garcia, B. (2007) – El campo de hoyos de la Edad del Bronce del Carrascalejo (Badajoz), Memorias de Arqueología Extremeña, 7, Junta de Extremadura, Mérida. Ferreira, M. T. (2009) – Torre Velha 3 (Barragem da Laje, Serpa), Relatório dos trabalhos de Antropologia Biológica desenvolvidos no âmbito da minimização de impactes no sítio da Torre Velha 3. Filipe, V.; Godinho, R.; Granja, R.; Valera, A. C. (2013) – Espaços funerários da Idade do Bronze no Outeiro Alto 2 (Brinches, Serpa, Portugal): a necrópole de hipogeus, Zephyrus, 71, p. 107-129. Inocêncio, J. (2013) – Contextos e práticas funerárias Calcolíticas no baixo Alentejo Interior (Sudeste Alentejano), Tese de mestrado, Instituto de Ciências Sociais, Universidade do Minho.

The preliminary study of the Alto Brinches 3 and Torre Velha 3 contexts advises some caution in its interpretation, at least until the zooarchaeological study of the faunal remains is completed, and the study of the archaeological material and its dating is done, only then we can properly frame the realities presented with what is being gradually known on this issue. However, we think that these contexts demonstrate above all the importance that the livestock held for the Recent Pre-history communities, reflecting the socio-economic weight that they represented in the livelihood strategies and how these communities were socially, economically and symbolically structured. On this last point in particular, the archaeological remains seem to witness a deep relationship between this human groups and their animals, a relationship that is more dynamic and complex than the traditional economic view, which regards animals merely as a source of raw materials, food resource and as means of transportation. In this way we need to go beyond animal protein and calories to understand the prehistoric communities (Russel, 2012). References AAVV, (2002) ‐ Contributos para a identificação e caracterização da Paisagem em Portugal Continental – Grupos de unidades de paisagem – Alentejo Central a Algarve, Colecção Estudos 10. Lisboa: Direcção Geral do Ordenamento do Território e Desenvolvimento Urbano. Vol. V. Alves, C.; Costeira, C.; Estrela, S.; Porfírio, E.; Serra, M. (2009) – Torre Velha 3. Relatório final (2.ª fase). Minimização de Impactes sobre o Património Cultural decorrentes da Construção da Barragem da Laje (Serpa). Relatórios Palimpsesto 2009. Alves, C.; Costeira, C.; Estrela, S.; Porfírio, E.; Serra, M. (2012) – Torre Velha 3 (Serpa): Dados preliminares, Almadan, adenda electrónica, II série, n.º 17, tomo 1, p. 3138. Alves, C.; Estrela, S.; Costeira, C.; Porfírio, E.; Serra, M.; Soares, A. M. and Moreno-García, M. (2010) – Hipogeus funerários do Bronze Pleno da Torre Velha 3 (Serpa, Portugal). O Sudeste no Sudoeste?!, Zephyrus, 66, p. 133-153. Alves, C.; Estrela, S.; Porfírio, E.; Serra, M., (2010) – Alto de Brinches 3. Relatório Final. Minimização de Impactes sobre o Património Cultural decorrentes da 54

Inocêncio, J. and Porfírio, E. (in press) – A interacção Homem-Animal no Calcolítico do Sudoeste de Portugal. Reflexões a propósito de um enterramento de Alto de Brinches 3, Serpa, Beja, Estudos do Quaternário. Liesau Von Lettow-Vorbeck, C. (2012) – Depósitos con ofrendas de animals en yacimientos Cogotas I: antecedents y características, in Rodriguez Marcos, J. A. and Fernández Manzano, J. Cogotas I, una cultura de la Edad del Bronce en la Península Ibérica, Universidad de Valladolid, Valladolid, p. 219-257. Lopes, M. da C.; Carvalho, P. C.; Gomes, S. M. (1997) – Arqueologia do concelho de Serpa, Serpa, Câmara Municipal de Serpa. Miguel, L. and Godinho R. (2009) – Notícia do sítio arqueológico do Monte das Covas 3 (Beja), Apontamentos de Arqueologia e Património, 4, p. 23-24. Morris, J. (2008) – Associated bone groups; one archeologit’s rubbish is another’s ritual deposition, in Davis, O.; Sharples, N.; Waddington, K., Changing Perspectives on the first millennium BC. Oxford, Oxbow, p. 83-98. Rodrigues, Z.; Estrela, S.; Alves, C.; Porfírio, E.; Serra, M. (2012) - Os contextos funerários do sítio de Alto de Brinches 3 (Serpa): dados antropológicos preliminares, in Actas do V Encontro de Arqueologia do Sudoeste Peninsular (Almodôvar, 18, 19 e 20 de Novembro de 2010), Município de Almodôvar, Almodôvar, p. 73-83. Russel, N. (2012) – Social Zooarchaeology: humans and animals in prehistory, Cambridge, Cambridge University Press. Sánchez Pollo, A. (2012) – Algo más que animales de compañia: la deposición ritualizada de perros en hoyos en el solar de Cogotas I, in Rodriguez Marcos, J. A. and Fernández Manzano, J., Cogotas I, una cultura de la Edad del Bronce en la Península Ibérica, Universidad de Valladolid, Valladolid, p.449-468. Santos, F. J.C.; Arez, L.; Soares, A. M. M.; Deus, M. de; Queiroz, P. F.; Valério, P.; Rodrigues, Z.; Antunes, A. S.; Araújo, M. F. (2008) - O Casarão da Mesquita 3 (S . Manços , Évora): um sítio de fossas “ silo ” do Bronze Pleno/Final na Encosta do Albardão, Revista Portuguesa de Arqueologia, Vol. 11, n.º 2, p. 55-86. Soares, A. M. M. (1994) – O Bronze do Sudoeste na margem esquerda do Guadiana. As necrópoles do concelho de Serpa, in Actas das V Jornadas Arqueológicas (Lisboa, 1993), Associação dos Arqueólogos Portugueses, Lisboa, vol. 2, p. 179-197. Soares, A. M. M.; Santos, F.; Dewulf, J.; Deus, M. de; Antunes, A. S. (2009) – Práticas rituais no Bronze do Sudoeste, alguns dados, Estudos Arqueológicos de Oeiras, 17, p. 433-456. Soares, A. M. M.; Antunes, A. S.; Queiroz, P. F.; Deus, M. de; Soares, R. M. G. M.; Valério, P. (2010) – A ocupação sidérica do Passo Alto (V. V. de Ficalho, Serpa) in Pérez Macías, J. A. and Romero Bomba, E., IV Encuentro de Arqueología del Suroeste Peninsular, (Aracena, 27 a 29 de Novembro de 2008), Universidad de Huelva Publicaciones, Huelva, p. 544-575. Valera, A. C. (2009) – Estratégias de identificação e recursos geológicos: o anfibolito e a necrópole da

sobreira de Cima, Vidigueira, in Bettencourt, A. M. S. e Alves, L. B., Dos montes das pedras e das águas. Formas de interacção com o espaço natural da pré-história à actualidade, CITCEM/APEQ, p. 25-36. Valera, A. C.; Santos, H.; Figueiredo, M.; Granja, R. (in press a) – Contextos funerários na periferia de Porto Torrão: Cardim 6 e Carrascal 2, Actas do IV Colóquio de Arqueologia do Alqueva - O Plano de Rega (2002-2010), Beja 2010. Valera, A. C. and Costa, C. (in press) – Animal paws in funerary contexts in southern Portugal and the segmentation problem, 11th ICAZ Conference, Paris 2010. Valera, A. C. and Filipe, V. (2010) – Outeiro Alto 2 (Brinches, Serpa): Nota preliminar sobre um espaço funerário e de socialização do Neolítico Final à Idade do Bronze, Apontamentos de Arqueologia e Património, 5, p. 49-56. Valera, A. C. and Filipe, V. (2012) – A necrópole de hipogeus do Neolítico Final de Outeiro Alto 2 (Brinches, Serpa), Apontamentso de Arqueologia e Património, 8, p. 29-41. Valera, A.C.; Godinho, R.; Calvo, E.; Moro Berraquero, J.; Filipe V.; Santos, H. (in press b) – “Um mundo em negativo” – fossos, fossas, e hipogeus entre o Neolítico Final e a Idade do Bronze na margem esquerda do Guadiana (Brinches, Serpa), Actas do IV Colóquio de Arqueologia do Alqueva - O Plano de Rega (2002-2010), Beja 2010. Valera, A. C.; Nunes, T.; Costa, C. (2010) – Enterramentos de canídeos no Neolítico: a fossa 5 de Corça 1 (Brinches, Serpa), Apontamentos de Arqueologia e Património, 5, p. 7-17.

55

Mammal remains from the Governor's House (Belém Tower, Lisbon) and Rua dos Correeiros (Baixa, Lisbon) in the context of fish processing factories in Lusitania Silvia Valenzuela-Lamas1,2 1

IEF-Marie Curie, University of Sheffield UNIARQ (Centre of Archaeology of the U. of Lisbon) and GRACPE (U. Barcelona)

2

Abstract This article presents the results of the taphonomical, taxonomical and biometrical analysis of the mammal remains recovered in the fish salting factories Casa do Governador da Torre de Belém (Lisbon) and Núcleo Arqueológico da rua dos Correeiros (NARQ-BCP, Lisbon). The results are contextualized with other synchronic sites in Lusitania. Sheep and goats are the most abundant species in 4th -5th c. AD, and there is a high diversification between sites in terms of meat consumption. Biometrical analysis suggests that cattle and sheep size was not improved after the Roman conquest in this region, which contrasts with other parts of the Iberian Peninsula. Keywords: Mammals, Roman period, Portugal. the economic circuits of the Empire collapsed. Both sites are located in present-day Lisbon (Portugal; figure 1) and were part of the complex of fish factories in the Tejo estuary (Bugalhão, 2001). The fish factories developed together with a number of amphoric workshops to transport the fish conserves all over the Empire (Lagóstena, 2005, Dias et al., 2010).

Introduction The transition between the Iron Age and the Roman period in Lusitania and the integration of this provincia in the Roman Empire is a complex phenomenon with significant differences between regions and sites (Fabião, 2001). How the economic change imposed by the Romans translated in the diet and husbandry practices of human populations is still barely known, as the number of zooarchaeological studies on Roman Portuguese material is still limited.

The sites were object of commercial excavations in the frame of local works affecting them (Era Arqueologia and IPPC/IPPAR – Dept. de Arqueologia). Both provide diachronic evidence spanning the 1st to the 5th c. AD.

The study of the faunal record from Casa do Governador (Governor’s House, Belém, Lisbon) and Núcleo Arqueológico da Rua dos Correeiros (NARQ-BCP, Lisbon) was developed in the frame of a post-doctoral fellowship funded by the Fundação para a Ciência e a Tecnologia (FCT). The main objective of the project was to characterize the meat diet and the husbandry practices in Roman Lusitania, and to compare them with those dating from the Iron Age, prior to the Roman invasion. This article presents the results obtained on these two Roman assemblages, and contextualizes them in the frame of other sites of the same region and chronology (1st – 4th c. AD).

- NARQ-BCP The site of NARQ-BCP is located in the Baixa Quarter, in Lisbon. In Roman times, this area was at the edge of

Materials and Methods The sites The sites in Núcleo Arqueológico da Rua dos Correeiros (thereafter NARQ-BCP) and Casa do Governador da Torre de Belém (thereafter CG) are fish factories producing conserves for the Empire market (Bugalhão, 2001, Filipe and Fabião, 2006-2007). They were both built in the first half of the 1st c. AD, and then abandoned at the end of the 4th – beginnings of the 5th c. AD, when

Figure 1. Location of the sites named in the text and approximate boundary of ancient Lusitania. 57

Figure 3. Plan of the fish processing factory in Casa do Governador site (Filipe and Fabião, 2006-2007). Animal bones were recovered from all areas, and especially come from the abandon layers, when the tanks were not in use (phases 3 to 5). This is important, as the bones will not reflect the diet or commercial activity of the fish factory, but will reflect the rubbish coming from the surrounding area of the abandoned factory. The results of these abandon layers (phase 3 to 5) will be presented all together, as the Chi square test indicated the absence of significant differences between them. Figure 2. Plan of the fish processing factory in NARQ-BCP site (Bugalhão, 2001).

- Casa do Governador (CG) The site of CG is located in the hinterland of Roman Olisipo, in present-day Belém (Lisbon). The excavation unearthed 34 tanks for fish processing (Filipe and Fabião, 2006-2007), discovering an extensive fish factory in the Tejo estuary (figure 3). Numerous samples were taken for wet sieving, which produced a huge icthyological assemblage mainly composed of whole sardines (Sardina pilchardus; Gabriel et al. 2009). Unfortunately, no mammal bones were available from these samples, and only the hand-collected material is presented here.

the Roman city, in the harbour zone located on the left margin of the Tejo river, and very close to the commercial forum (Bugalhão, 2001). The area concentrated several fish processing factories. The excavated remains comprise more than 30 tanks for fish processing (Bugalhão, 2001), organized in groups and separated by open spaces (patios) to allow the circulation of people and goods (figure 2).

During the excavation, individualised:

During the excavation, 5 different chronological phases were identified:

several

phases

were

Phase 1 – Building of the fish factory and the tanks (0-50 AD)

Phase 1 – refurbishment in the fish factory (75-100 AD). Phase 2 – partial abandon of the factory (3rd c. AD).

Phase 2 – Abandon layers of the Roman factory (late 4 th c. AD)

Phase 3 – abandon (3rd – 5th c. AD).

Phase 3 – levels dating from the 15th-16th centuries.

Phase 4 – abandon (late 4th – early 5th c. AD).

Phase 4 and 5 – modern levels

Phase 5 – abandon (early 5th c. AD).

The animal bone assemblage.

The animal bone assemblage.

The bone assemblage coming from phases 1 and 2 from CG consisted of 112 identified specimens (table 2). This

The bone assemblage from NARQ-BCP consisted of 1,114 identified specimens (table 1). All the assemblage was hand collected during the commercial excavation. 58

NARQ-BCP Bos Sus Ovis Capra Ovis/Capra Cervus Equus Oryctolagus Lepus Canis Felis Shells Ichtyo birds Human TOTAL

R1 0 3 0 0 2 0 0 0 0 0 0 6 0 0 0 11

R2 2 1 0 0 1 0 0 0 0 0 0 0 0 0 0 4

R3 2 1 5 1 12 3 0 2 1 0 0 1 1 1 1 30

R4 9 15 12 2 17 3 1 2 0 0 0 6 0 1 0 68

R5 16 128 80 4 194 11 2 31 1 31 2 395 40 65 0 1000

TOTAL 29 148 97 7 226 17 3 35 2 31 2 408 41 67 1

Table 1. Number of Identified Specimens (NISP) by species and period from NARQ-BCP site. Bos= Bos taurus, cattle; Sus= Sus domesticus, pig; Ovis= Ovis aries, sheep; Capra= Capra hircus, goat; OC= Ovis/Capra, sheep/goat; Cervus= Cervus elaphus, red deer; Equus= Equus sp., equid (horse/donkey); Oryctolagus= Oryctolagus cuniculus, rabbit; Lepus= Lepus sp., hare; Canis= Canis familiaris, dog; Felis= Felis catus, cat.

Bos Sus Ovis/Capra Cervus Oryct. Equus Canis Birds Other TOTAL

Phase 1 NRD 0 0 4 0 0 0 0 0 0 4

of IGESPAR and a diversity of atlas and catalogues of determination (Schmid, 1972; Barone, 1980; Boessneck et al., 1964; Boessneck, 1980; Payne, 1985; Halstead et al., 2002). The epiphyseal stage and the teeth wear were recorded on the basis of the works of Barone (1980) and Gardeisen (1997). Butchery marks were recorded basing on the work of Vigne (1988), and the biometric work followed Driesch (1976) as the main reference, and the works of Davis (1996 and 2002) for specific measurements. All the measurements were taken with a manual calliper and expressed in millimetres.

Phase 2 NRD %NRD 18 16,7 13 12,0 55 50,9 0 0,0 3 2,8 2 1,9 8 7,4 6 5,6 3 2,8 108

The degree of preservation of the bone cortical and the reconstruction of the mortality profiles followed the methodology explained in Valenzuela (2008). Only the identifiable specimens (bones, teeth and molluscs preserving the apex) were counted and included in the analysis.

Table 2. Number of Identified Specimens (NISP) by species and period from Casa do Governador site. Bos= Bos taurus, cattle; Sus= Sus domesticus, pig; Ovis/Ca= Ovis/Capra, sheep/goat; Cervus= Cervus elaphus, red deer; Oryctolagus= Oryctolagus cuniculus, rabbit; Equus= Equus sp., equid (horse/donkey); Canis= Canis familiaris, dog.

The quantification units considered were the Number of Identified Specimens (NISP), the simple Minimum Number of Individuals (MNI, White 1953), and the Percentage of Representation (PR, Dodson and Wexlar, 1969).

limits the possibilities of inference about the diet and husbandry practices in the surrounding area of the factory in the Roman period, but still provided some information about the species present and some osteometrical data.

Results Taphonomy In NARQ-BCP site, the level of bone preservation is good or medium in most cases (figure 4), most of them displaying little areas or a maximum of the half bone cortical damaged by post-depositional agents. Among them, root etching was the main agent of alteration (table

Methods The faunal analysis focused on the taxonomical and anatomical determination using the reference collection 59

Table 3. NISP affected by taphonomical agent from NARQ-BCP site. Roots= root etching; Carn= carnivore; desq/fiss= fissures and desquamation; fire= burning. 3), followed by desquamation and fissures, probably related to the seasonal changes in the humidity conditions in the soil. Pitting marks were recorded in 3.2% of the these agents identified bones, which suggest a moderate incidence of these agents.

Among the marine species, the majority of shells were identified as Ostrea cf. edulis (table 6), both the upper (224 fragments; 58.2%) and the lower shell (161 fragments; 41.8%). Other species present are the mussel (Mytillus galloprovincialis), saltwater clams (Venerupis sp.), seashells (Fasciolaria lignaria), limpets (Patella sp.), and scallops (Pecten jacobaeus, Chlamys varia). Concerning the terrestrial molluscs, seven garden snails (Helix aspersa) were recovered in the excavation.

Concerning CG site, the bone preservation is medium (figure 4), with most bones displaying half the bone cortical with some kind of alteration (degree 2). This probably has not compromised the post-depositional conservation of any bone remain. Concerning the taphonomical agents (table 4), the root etching has been the main factor of alteration, followed by fissures, probably related to a seasonal phenomenon of hydration / dehydration of the bones. There is very low proportion of bones displaying pitting marks, which suggest that carnivores did not heavily attack the bone assemblage.

Concerning CG assemblage, the phase 1 is represented for sole 4 sheep bone fragments with some butchery marks (US 1060). Their presence in the construction levels of the tanks is probably occasional and cannot be used to make any inference. In the phase 2 (abandon layers, late 4th c. AD) sheep and goats are the best represented species, followed by cattle and pig (table 2). Other mammal species in presence are the dog, the rabbit and an equid, together with six bird remains and other remains corresponding to fragments of turtle carapace and reptilian bones. The low number of remains only allows to testify the presence of these taxa in the abandon layers of the fish factory (4th c. AD), but does not allow to make any economic or dietary inference. We did not have access to the mollusc remains from this site.

Species Sheep and goats are the most abundant species in NARQBCP site (table 1 and table 5 - detail cetarias), followed by pigs and leporids (rabbit or hare). Dogs, cattle, red deer, equids and cats are also present in the abandon layers. Other than mammals, the presence of bird bones is notable (6% NISP), as well as the abundance of marine shells. In that regard, as mentioned in the methods section, only the fragments preserving the apex were counted, in order to avoid their over-representation. 60

Phase 2 Bos Sus

Roots 15 11

Carn. 0 0

Desq 2 0

Fiss. 7 5

Fire 0 2

NISP 18 13

Ovis Capra

12 4

1 0

0 0

5 1

0 0

13 4

Ovis/Capra

37

0

1

18

0

38

Canis Equus

8 0

0 0

0 0

5 1

0 0

8 2

Cervus Oryct.

0 3

0 0

0 0

0 0

0 0

0 3

TOTAL

90

1

3

42

2

99

Table 4. NISP affected by taphonomical agent from CG site. Roots= root etching; Carn= carnivore; desq/fiss= fissures and desquamation; fire= burning.

Bos taurus Sus domesticus Ovis aries Capra hircus OC Canis familiaris Felis catus Cervus elaphus Equus Oryct. cuniculus Lepus sp. Icthyo Birds Shells Human TOTAL

1st c. AD Cet. 30 0 3

3rd c. AD Cet. 17 2 1

3rd-5th c. AD

4th-5th c. AD

5th c. AD

Patio 5 0 0

Cet. 1 1 0

Cet. 2 0 0

Cet. 4 1 1

Patio 2 1 1

Cet. 18 5 6

Cet. 19 0 2

Cet. 21 1 0

Cet. 22 0 0

Cet. 26 0 1

Cet. 32 2 5

Cet. 12 5 36

Cet. 23 11 92

0 0

0 0

0 1

4 0

0 0

1 0

0 0

8 0

1 2

0 0

0 2

0 0

2 0

29 0

51 2

2 0

1 0

2 0

6 0

0 0

4 0

0 0

7 0

5 0

0 0

0 0

0 0

5 0

29 8

165 23

0 0

0 0

0 3

0 0

0 0

0 0

0 0

0 3

0 0

0 0

0 0

0 0

0 0

2 0

0 11

0 0

0 0

0 0

0 1

0 0

0 1

1 0

0 0

0 0

0 0

0 0

0 0

0 2

0 4

2 27

0 0 0 6 0 11

0 0 0 0 0 4

1 0 0 0 0 7

0 1 1 0 0 14

0 0 0 0 1 1

0 0 0 1 0 9

0 0 0 3 0 6

0 0 1 3 0 33

0 0 0 0 0 10

0 0 0 0 0 1

0 0 0 0 0 2

0 0 0 0 0 1

0 0 0 0 0 16

0 24 9 9 0 155

1 16 56 386 0 843

Table 5. NISP recovered from each tank (cetaria) and patio of the NARQ-BCP site. Anatomical representation

and fire marks on the bones suggest that they all correspond to domestic waste, in which no evidence of industrial or systematic butchery pattern has been observed.

Table 7 gives the detail of the anatomical representation of the main taxa and levels, both in NARQ-BCP and CG sites.

Concerning the red deer, the bones correspond to different elements (mainly distal tibia and tarsal bones), and the skull is only present by an unworked fragment of antler. This also corresponds with the expected composition of domestic waste. Concerning the dogs, the

In NARQ-BCP, the main domesticates (cattle, pigs and caprines) are present with all the anatomical parts. The lack of a specific anatomical representation, together with the breakage pattern and the presence of some butchery 61

Species Venerupis sp. Fasciolaria lignaria Helix aspersa Mytilus galloprovincialis Ostrea cf. edulis Patella sp. Pecten jacobaeus, Chlamys varia TOTAL

In CG, most of the species are represented by a low number of remains, which makes difficult to assess their significance. In the case of sheep and goats, all the anatomical parts (skull, anterior member, posterior member) are present by some elements, and only the smallest bones (carpals and tarsals, third phalanxes) are completely absent.

NR 3 2 7 4 385 1 4 406

Kill-off patterns Figures 9 to 11 display the reconstructed mortality profiles for the main domesticates in the abandon layers of NARQ-BCP site (4th-5th c. AD). The bar width corresponds to the duration of each age class. Therefore, the bar height indicates the incidence of killings per age class (the ‘rhythm’ of killings), while the area indicates the proportion of killings per age class, also displayed with the survival curve. The vertical lines indicate the variation range of the estimations (two standard deviations; 95% confidence). The ages are expressed in months.

Table 6. NISP by species of shell remains. Only fragments preserving the apex were counted bones correspond to a MNI of three dogs with some bones in anatomical connexion each, which suggest that they correspond to animals that died or were buried in the abandoned factory. Figures 5 and 6 display the %NISP for the species represented by more than 80 bones (in this case, pigs and caprines). Figures 7 and 8 display the percentage of representation (PR) for these species, calculated following Dodson and Wexlar (1969). Concerning the caprines, both meaty parts and low meat parts are well represented (figure 7). The low number of isolated teeth, carpal and tarsal bones can be explained because their small size, but also because the low breakage degree of the mandibles and maxilla. Concerning vertebrae and ribs, they have not been identified to the species level with some exceptions (i.e: atlas or axis). In the case of pigs, the PR values (figure 8) indicate that meaty anatomical parts are well represented, especially the anterior member (scapula, humerus, radius, but also tibia and mandibles display a good representation). Feet remains and isolated teeth are poorly present.

The number of remains with ageing information for cattle (figure 9) is too little to make any solid inference (n=26). The results suggest that, in this urban context, the killings concentrate between the 2 and 4 years of life, and then before the 6 years of life. Concerning pigs (figure 10; n= 130), the majority of killings occur between a year and a half and 3 years of life, despite a high incidence of killings in younger ages, especially between 6 and 12 months. As pigs are devoted to meat production and their productivity rapidly decreases after 2 years of life, no animals older than 3 years were registered in NARQ-BCP site.

Degree of preservation of the cortical bone 80 70 60

%

50

NARQ-Phase1

40

NARQ-Phase2

30

NARQ-Phase3/5

20

CG-Phase2

10 0 Degree 0

Degree 1

Degree 2

Degree 3

Degree 4

Figure 4. Degree of alteration of the bone cortical in a scale from 0 to 4, in which 0 is the absence of alteration, and 4 is a bone cortical completely destroyed

62

skull mand teeth scapula humerus radius ulna carpal MC ribs vertebrae pelvis femur tibia fibula tarsal MT MP P1 P2 P3 TOTAL

skull mand teeth scapula humerus radius ulna carpal MC ribs vertebrae pelvis femur tibia fibula tarsal MT MP P1 P2 P3 TOTAL

CG Phase 2 (4th c. AD) Sus Ovis Capra OC Cervus Equus Oryct. 0 0 0 0 0 0 0 0 6 0 3 0 0 2 2 2 1 0 0 1 0 1 3 1 3 0 0 0 0 0 0 5 0 0 0 3 0 0 4 0 0 0 2 0 0 1 0 0 0 0 0 0 0 0 0 0 0 1 1 2 0 0 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 6 0 0 1 0 0 0 5 0 0 0 1 0 0 6 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 0 0 0 1 0 0 0 0 1 0 1 0 0 2 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 14 12 3 40 0 2 3 NARQ-BCP abandon layers (3rd-5th c.AD) Sus Ovis Capra OC Cervus Equus Oryct. 5 12 1 1 1 0 0 25 30 0 16 0 1 2 6 10 0 3 0 1 0 14 3 0 9 2 0 0 14 3 0 22 0 0 11 9 0 0 22 0 0 0 7 0 0 5 1 0 0 0 0 0 0 0 0 0 8 18 2 19 0 0 0 0 0 0 0 0 0 0 3 2 1 4 0 0 0 6 0 0 16 0 0 7 1 0 0 26 0 0 7 21 0 0 39 7 0 6 5 0 0 0 0 0 0 6 2 0 12 4 0 0 4 17 1 12 0 0 0 3 0 0 2 0 0 0 7 0 0 16 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 144 97 5 224 16 2 33 Bos 1 0 1 0 2 0 1 0 0 0 0 1 1 1 0 2 0 0 6 2 0 18

Table 7. Anatomical representation (NISP values) by species and site

63

Canis 0 0 0 0 0 1 3 0 0 0 0 0 3 1 0 0 0 0 0 0 0 8 Canis 1 3 0 2 4 3 3 0 0 0 1 1 3 2 0 2 0 6 0 0 0 31

Figure 5. Proportion of the skeletal elements (%NISP) and location of the butchery marks observed on sheep and goat bones.

Figure 6. Proportion of the skeletal elements (%NISP) and location of the butchery marks observed on pig bones.

OC-NARQ BCP (abandon layers)

Sus NARQ-BCP (abandon) 100 80 60 40 20 0

60

PR

PR

80 40 0

skull mand teeth scapula humerus radius ulna carpal MC ribs vertebrae pelvis femur tibia tarsal MT P1 P2 P3

20

Skull Mand Teeth Scapula Humerus Radius Ulna Carpal MC Ribs Vertebrae Pelvis Femur Tibia Fibula Tarsal MT MP P1 P2

100

Figure 7. Percentage of representation of sheep and goat remains from the abandon layers in NARQ-BCP site.

Figure 8. Percentage of representation of pig remains from the abandon layers in NARQ-BCP site.

Figure 9. Mortality profile (histogram) and survival curve basing on cattle long bones and teeth from NARQ-BCP site. Ages expressed in months.

Figure 10. Mortality profile (histogram) and survival curve basing on pig long bones and teeth from NARQ-BCP site. Ages expressed in months.

In sheep and goats (figure 11; n= 292), the killings concentrate between 1 and 2 years of life, in the optimum time for meat production (Payne 1975, Helmer and Vigne 2004). These are followed in frequency by animals between 2 and 4 years. Younger animals are also consumed in the abandoned factory or its surroundings

(1-12 months of life), and only a small proportion of bones could be attributed to animals older than 4 years. No foetal bones were identified for any of the main taxa.

64

The Student Test indicated the inexistence of significant differences between the two periods (t = 1.5107, df = 89.627, p-value = 0.06719; mean of Roman= 14.1; mean of Iron Age 13.8; standard deviations= 0,13 and 0,15 respectively). Discussion This work presents the main results obtained on two zooarchaeological assemblages (NARQ-BCP and Casa do Governador) from two fish processing factories in Lisbon area (Portugal). Both factories have a similar building plan: they consist in a number of tanks for fish processing organised in groups surrounding patios that act as spaces of circulation and distribution (figures 2 and 3). From a taphonomical point of view, the two assemblages come from domestic waste from the abandon layers (4th c. AD) and do not reflect the diet or the industrial activity of the factories when they were in use.

Figure 11. Mortality profile (histogram) and survival curve basing on sheep and goat long bones and teeth from NARQ-BCP site. Ages expressed in months. Biometry Only cattle and sheep provided enough individual measurements to be analysed on their own. The results from NARQ-BCP and CG were compared to those of other contemporary sites and also with the data available for the Iron Age to see if any change could be observed. Concerning cattle, no differences between periods nor sites were attested. Figure 12 exemplifies this on the lateral length of cattle talus (GL). The number of data for both NARQ-BCP and CG is very low, but, other than this, it can be seen that the means between periods and sites clearly overlap.

The assemblages from NARQ-BCP and CG were compared to the published data for other sites in Lusitania as Alcáçova de Santarem (Davis, 2006), Mesas

Sheep measurements showed a similar pattern. The most abundant measurement and that displaying larger differences between periods was the height of the distal trochlea (HTC; figure 13). A slight increment of the mean can be appreciated between the Iron Age and the Roman period, although the confidence intervals clearly overlap.

Figure 12. Histogram displaying the total length (GL) of the cattle talus in different Iron Age and Roman sites from Lusitania. The line indicates the mean in each period with and the 95% interval of confidence (2 times the standard deviation). 65

Figure 13. Histogram displaying the height of the humerus throclea (HTC) of sheep in different Iron Age and Roman sites from Lusitania. The line indicates the mean in each period with and the 95% interval of confidence (2 times the standard deviation).

do Castelinho (Valenzuela and Fabião 2012) and Torre de Palma (MacKinnon 1999/2000), which preserved more than 100 remains. The assemblages of Ilha do Pessegueiro (Cardoso, 1993) and Quinta do Marim (Antunes and Mourer-Chauviré, 1992) were not included because their small sample size.

less wealthy position of the populations from the Late Roman period living in this site. This phenomenon finds a parallel in the 4th and 5th contexts of NARQ-BCP and CG, where sheep and goats are also the main species consumed. In the case of NARQ-BCP, located at the edge of the Roman capital – Olissipo - pig remains are abundant (28,3%NISP), while cattle and red deer are poorly represented. Concerning CG site, located about 5 Km from the Roman city, the results suggest a higher proportion of cattle and a lower proportion of pig in this area, maybe reflecting a less urban landscape or less wealthy position.

Figure 14 displays the %NISP of the four main taxa in the considered sites. Sheep and goat are the most abundant species in terms of NISP, generally followed by pigs. Red deer is present in almost all the sites, and in two cases –Mesas do Castelinho and Torre de Palma– is more abundant than cattle, with %NISP higher than 20%. In general, cattle bones represent more than the 10%NISP, with the only exception of NARQ-BCP site, which displays a considerably lower proportion (5.5%).

None of the sites considered reflects of the Italic diet pattern, which displays a very high proportion of pig remains (King, 1999). The high proportion of red deer remains suggests a specificity of Portuguese assemblages in Alentejo, contrasting with the faunal records further North, in Gallaecia (Fernández Rodríguez 2003), and also the rest of the Iberian Peninsula (King 1999) and even with other parts of present day Portugal (Moreno and Valera 2007). The abundance of red deer in faunal contexts from Alentejo is already attested in precedent chronologies (Chalcolithic and Iron Age; Moreno and Valera 2007, Valenzuela and Fabião, 2012).

Table 8 gives the results of the Chi square test performed to assess if the sites could come from the same distribution or there are significant differences between them. It can be seen that all of them have very significant differences between them, with the only exception of the comparison between Torre de Palma and Mesas do Castelinho, both corresponding to Roman villae in present day Alentejo. The urban assemblages of Santarém and also the two fish factories display a varied composition in terms of %NISP. This suggest that the meat consumption in each site is adapted to the wealth level of people, local availability of products, or the local food culture, and that there is not any uniform pattern in terms of meat consumption between sites, maybe with the only exception of the Roman villae in Alentejo region.

The high proportion of sheep and goats is characteristic of the 4th-5th c. all over the Mediterranean (King, 1999), and can be a reflection of the changes in the economy after the crisis of the 3rd c. AD and the ulterior crises in the 4th and 5th centuries. Concerning biometrical data, no significant differences were observed between periods and sites on cattle and sheep remains. This contrasts with other parts of the Iberian Peninsula, where size increase in cattle bones has been attested in Roman times (Altuna, 1980; Colominas et al. 2013), but finds a parallel in Gallaecia, where no change in animals size has been attested (Fernández Rodríguez 2003). This suggests that Romans did not perform any significant improvement in the livestock herded in the Western part of Iberian Peninsula.

From a diachronic point of view, there is a notable increase of sheep and goat remains in Alcáçova de Santarém between the 1st – 2nd c. AD (phase R3; figure 14) and the 4th-5th c. AD (phase R5; figure 14), together with a decrease of pig and red deer. This could reflect a %NISP Lusitania Bos

Dealing with the mortality profiles, NARQ-BCP displays profiles compatible with a meat-centred kill-off patterns, which agrees with the urban and presumably consumer character of the site. In that respect, all the anatomical parts seem to be present on the site for the main domesticates. More data is needed from other sites in order to make any inference or characterisation of the husbandry practices in Roman times. In that regard, it would be interesting to compare urban contexts with villae and small rural sites.

Sus

Torre de Palma (n=1368)

NARQ-BCP (NISP=523)

Casa do Governador (NISP=86)

Santarem R5(NISP=176)

Santarem R3(NISP=238)

OC Mesas do Castelinho…

70 60 50 40 30 20 10 0

Cervus

Conclusion

Figure 14. %NISP of the four main species in different Roman sites from Lusitania. Bos= Bos taurus, cattle; Sus= Sus domesticus, pig; OC= Ovis/Capra, sheep/goat; Cervus= Cervus elaphus, red deer).

The present study contributes to the knowledge of the Roman world in one of the occidental limes of the Roman Empire. The results indicate that sheep and goats are the 66

Chi Values (dof=3) NARQ CasaGovernador MesasCastelinho Santarem.R3. Santarem.R5. TorrePalma

NARQ

CasaGovernador

MesasCastelinho

Santarem.R3.

Santarem.R5.

TorrePalma

0 30,1463481 137,897538 88,7340304 77,8268365 199,486113

30,1463481 NA 42,7728812 26,2140299 11,1356426 48,9205962

137,897538 42,7728812 0 12,1429104 55,904958 5,15162366

88,7340304 26,2140299 12,1429104 0 32,7141822 8,50989895

77,8268365 11,1356426 55,904958 32,7141822 0 75,19174

199,486113 48,9205962 5,15162366 8,50989895 75,19174 0

P values NARQ CasaGovernador MesasCastelinho Santarem.R3. Santarem.R5. TorrePalma

NARQ 1 1,29E-06 1,07E-29 4,10E-19 8,98E-17 5,45E-43

CasaGovernador 1,29E-06 NA 2,75E-09 8,60E-06 0,01101457 1,36E-10

MesasCastelinho 1,07E-29 2,75E-09 1 0,0069093 4,40E-12 0,1610253

Santarem.R3. 4,10E-19 8,60E-06 0,0069093 1 3,70E-07 0,03656924

Santarem.R5. 8,98E-17 0,01101457 4,40E-12 3,70E-07 1 3,30E-16

TorrePalma 5,45E-43 1,36E-10 0,1610253 0,03656924 3,30E-16 1

Table 8. Chi and P values resulting from the Chi square test between sites. most abundant species in terms of NISP in 4th-5th c. AD, and that there is a high diversification between sites in terms of meat consumption, maybe with the only exception of two Roman villae in Alentejo (Mesas do Castelinho and Torre de Palma; figure 14 and table 8), that display a high proportion of red deer remains.

Barone, R., (1980) - Anatomie comparée des mammifères domestiques, Ed. Vigot, Paris. Boessneck, J., Müller, H.H., Teichert, M. (1964) Osteologische Unterschiedungsmerkmale zwischen schaf Ovis aries Linné und ziege Capra hircus Linné, Kühn. Archiv, 78 (1-2): 1-129. Boessneck, J. (1980) - Diferencias osteológicas entre las ovejas (Ovis aries Linné) y las cabras (Capra hircus Linné), Brothwell, D.R. and Higgs, E.S.(Eds.) Ciencia en Arqueología, Fondo de Cultura Económica, Madrid, 331358. Bugalhão, J. (2001) - A indústria romana de transformação e conserva de peixe em Olisipo. Núcleo Arqueológico da Rua dos Correeiros, Trabalhos de Arqueologia 15, Lisbon. Cardoso, J. L. (1993) Restos de grandes mamíferos da Ilha do Pessegueiro. Contribuição para o conhecimento da alimentação na época romana, in: Tavares da Silva, C., Soares, J., Ilha do Pessegueiro. Porto romano da costa alentejana, Instituto de Conservação da Natureza, 205215. Colominas, L., Schlumbaum, A., Saña, M. et al (2013) The impact of the Roman Empire on animal husbandry practices: study of the changes in cattle morphology in the north-east of the Iberian Peninsula through osteometric and ancient DNA analyses, Archaeological and Anthropological Sciences, D.O.I: 10.1007/s12520013-0116-9. Davis, J.M.S. (1996) - Measurements of a Group of Adult Female Shetland Sheep Skeletons from a Single Flock: a Baseline for Zooarchaeologists, Journal of Archaeological Science, 23: 593–612. Davis, J.M.S. (2002) - The mammals and birds from the Gruta do Caldeirão, Portugal, Revista Portuguesa de Arqueologia, 5: 29-98. Davis, J.M.S. (2006) - Faunal remains from Alcáçova de Santarém, Portugal, Trabalhos de Arqueologia, 43, 144p. Dias, M.I., Prudêncio, M.I., Gouveia, M.A., Trindade, M.J., Marques, R., Franco, D., Raposo, J, Fabião, C.,

The mortality profiles obtained on the material from NARQ-BCP site indicate that the cattle, sheep and goats and pigs consumed there (or its surroundings) were killed mainly for their meat, in the moment when the animals could have reproduced at least once (figures 9 to 11). Biometrical analysis suggests that no improvement in the size of livestock occurred after the Roman conquest both in cattle and sheep (figures 12 and 13). This contrasts with other parts of the Iberian Peninsula, where a significant increase on cattle bones size has been attested after the Roman conquest (Altuna, 1980; Colominas et al. 2013), and agrees with other data from the Western part of the Iberian Peninsula but further north, in Gallaecia province (Fernández Rodríguez, 2003). The number of published faunal studies on Iron Age and Roman assemblages is still very limited in Portugal, and therefore more research is needed before reaching any strong conclusion about the changes (or the lack of them) in Lusitania between these two periods. References Altuna, J. (1980) - Historia de la domesticación en el País Vasco desde sus orígenes hasta la Romanización, Munibe, 32. Antunes, M.T., Mourer-Chauviré, C. (1992) - The Roman site (2nd to 5th centuries AD) at Quinta do Marim near Olhão (Algarve, Portugal): vertebrate faunas, Setúbal Arqueológica, IX-X: 375-382. 67

Guerra, A., Chemical tracers of Lusitanian amphorae kilns from the Tagus estuary (Portugal), Journal of Archaeological Science, 37: 784–798. Dodson, P., Wexlar, D. (1979) - Taphonomic investigations of owl pellets. Paleobiology, 5(3): 275– 284. Driesch, A. (1976) - A Guide to the Measurement of Animal Bones from Archaeological Sites, Peabody Museum, Harvard. Fabião, C. (2001) - Mundo indígena, romanos e sociedade provincial romana: Sobre a percepção arqueológica da mudança, (ERA) arqueologia, 3: 108131. Fernández Rodríguez, C. (2003) - Ganadería, caza y animales de compañía en la Galicia romana: estudio arqueozoológico, Brigantium 15. Filipe and Fabião (2006-2007) - Uma unidade de produção de preparados de peixe de época romana na Casa do Governador da Torre de Belém (Lisboa): uma primeira apresentação, Arqueologia & História, 58/59: 103-120. Gabriel, S., Fabião, C., Filipe, I. (2009) - Fish remains from the Casa do Governador - a Roman fish processing factory in Lusitania, "Fishes - Culture - Environment Through Archaeoichthyology, Ethnography & History", The 15th Meeting of the ICAZ Fish Remains Working Group (FRWG), Poland. Gardeisen , A. (1997) - Exploitation des prélèvements et fichiers de spécialité (PRL, FAUNE, OS), Lattara, 10: 251-278. Halstead, P., Collins, P., Isaakidou, V. (2002) - Sorting the sheep from the goats: morphological distinctions between the mandibles and mandibular teeth of adult Ovis and Capra, Journal of Archaeological Science, 29: 545-553. Helmer, D., Vigne, J.-D. (2004) - La gestion des cheptels de caprinés au Néolithique dans le midi de la France, Approches fonctionnelles en Préhistoire. Actes du XXVe Congrès Préhistorique de France, 397-407. King, A. C. (1999) - Diet in the Roman world: a regional inter-site comparison of the mammal bones, Journal of Roman Archaeology, 12: 168-202. Lagóstena, L. G. (2005) - Pesquerías en la Hispania altoimperial. Reflexiones y perspectivas para su estudio, in: Molina, J., Sánchez, M.J. (Eds.) III congreso internacional de estudios históricos en el Mediterráneo: la cultura del mar y la sal, 77-88. MacKinnon, M. (1999/2000) - O papel dos animais na economia rural da Lusitânia romana: zooarqueologia de Torre de Palma. A Cidade. Revista Cultural de Portalegre, 13/14: 129–140. Moreno, M., Valera, A.C. (2007) - Os restos faunísticos de vertebrados do sítio do Mercador (Mourão), Vipasca Arqueologia e história, 2: 139-152. Payne, S. (1973) - Kill-off Pattern in Sheep and Goats: the mandibles from Asvan kale, Journal of the British Institute of Archaeology at Ankara, 23: 281-303. Payne, S. (1985) - Morphological distinction between the mandibular teeth of young sheep, Ovis and goats Capra, Journal of Archaeological Science, 12: 139-147.

Schmid, E. (1972) - Atlas of Animal Bones, Elsevier, London. Valenzuela, S. (2008) - Alimentació i ramaderia al Penedès durant la protohistòria (segles VII-III aC), Premi d’Arqueologia, Memorial Josep Barberà i Farràs, Societat Catalana d’Arqueologia, Barcelona. Valenzuela, S., Fabião, C. (2012) - Ciervos, ovejas y vacas: el registo faunístico de Mesas do Castelinho (Almodôvar) entre la Edad del Hierro y Época Romana, V Encontro de Arqueologia do Sudoeste Peninsular, Almodôvar, 413-432. Vigne, J.-D. (1988) - Les Mammifères post-glaciaires de Corse, étude archéozoologique, XXVI supplément Gallia Préhistoire, CNRS Editions, Paris, 1988. White, T. (1953) - A Method of Calculating the Dietary Percentage of Various Food Animals Utilized by Aboriginal Peoples, American Antiquity,18 (4): 396-398.

68

Animal bones from the Roman site of Tróia (Grândola, Portugal): mammal and bird remains from the fish-salting workshop 2 (2007/08) Mariana Nabais1 1

University College London (UCL), [email protected]

Abstract The Roman site of Tróia is located at about 50 km south of Lisbon, Portugal. It is positioned on the NE side of Tróia peninsula facing the river Sado estuary. Tróia is known since the 16 th century and several archaeological interventions have been conducted until now. The great importance of this site is intimately related to the production of large amounts of salted fish and fish sauces that were sold throughout the West Roman Empire, from the 1 st to the 5th century AD. From the excavation campaigns of 2007 and 2008 in workshop 2, a total of 663 mammal and bird bones were recovered and studied. They demonstrate that some of the Tróia’s population diet was based on domestic caprines, pigs and chicken that might have been kept within small family groups. No large mammals were kept maybe due to their high needs of maintenance which contrast with the easy care of medium size animals. Hunting activities seemed to have taken place, being lagomorphs an important element in the diet. Finally, patterns of animal consumption in Tróia appear to be very similar to other Portuguese Roman sites. Keywords: Portugal; Tróia; Roman; salt-fishing tanks; Zooarchaeology preservation, which is essentially due to the dry environment caused by the sand dunes that cover the whole site.

Introduction The Roman site of Tróia is located at about 50 km south of Lisbon (Figure1). It is placed in Tróia peninsula which is oriented NW-SE. The East side of the peninsula faces the river Sado estuary, while the West part stares at the Atlantic Ocean. The site is positioned on the East side of the peninsula, close to a natural and shallow bay called Caldeira.

The faunal assemblage is mainly composed by medium size macrofauna (65%), small size macrofauna (20%) and birds (9%). Curiously, large size macrofauna has a very low representation (6%), such as the microfauna (1%) (Figure 3). Most of the bone remains were recovered from tanks 1, 5, 7c and the NW area, contrasting with the very low contribution of tanks 8, 9 and Pátio 2B (Figure 2).

The geographical location of the Roman site Tróia is highly favourable to fishing activities. Therefore, during its occupation, from the 1st to the 5th century AD, Tróia was specialized in the production of salted fish and fish sauces that were sold throughout the West Roman Empire. So far, they have been identified 25 fish-salting workshops in Tróia (Pinto, Magalhães & Brum 2011). The present study, however, will only be focused on the faunal material recovered from the fish-salting workshop 2 (Figure1) from the excavation campaigns of 2007 and 2008. The excavation occurred only in untouched contexts, such as the fish-salting tanks 1, 5, 7c, 8, 9, the Pátio 2B and the NW area. In chronological terms, tanks 8 and 9 date to the end of the 2nd century, beginning of the 3rd century. All the others belong to the abandonment levels of the first half of the 5th century. Species Found Tróia’s workshop 2 faunal assemblage is composed by 663 mammal and bird remains (Table1). Most of the bones were handpicked during excavation or collected from a screen with 0,5 cm mesh. The material has a good

Figure1. (A) Location of Tróia in Portugal. (B) Map of the Roman site of Tróia. (C) Workshop 2 and its different compartments. 69

NISP

%NISP

MNI

%MNI

MAU

%MAU

Equids

3

0.46

2

3.08

1

2.63

Cattle

8

1.21

4

6.15

2

5.26

Deer

2

0.30

1

1.54

0.5

1.32

Pig

76

11.23

10

15.38

5

13.16

Caprine

102

15.48

12

18.46

8.5

22.37

Dog

3

0.46

2

3.08

1

2.63

Cat

14

2.12

2

3.08

1

2.63

Lagomorphs

36

5.46

6

9.23

4.5

11.84

Chicken

43

6.53

10

15.38

5.5

14.47

Aquatic Birds

11

1.66

6

9.24

3.5

9.21

Birds undetermined

6

0.91

1

1.54

0.5

1.32

Rodents

7

1.06

2

3.08

1

2.63

Amphibians

1

0.15

1

1.54

0.5

1.32

Large Macrofauna

24

3.49

1

1.54

0.5

1.32

Medium Macrofauna

247

37.33

3

4.62

2

5.26

Small Macrofauna

80

12.14

2

3.08

1

2.63

Total

663

100

65

100

38

100

Table 1. NISP, MNI and MAU for the species found in the excavated compartments of Troia’s workshop 2 (2007/08). Equids

meat was certainly consumed, even though there is only one chopped radius.

Only three fragments of equids were identified (Table1): one metacarpus IV in tank 8 and two distal radii in the NW area. All fragments were identified to genus (Equus sp.) but not to species. The identification to species is more successful when teeth are present but that is not the case of this assemblage.

Although none of the cattle elements were identified to species, it is assumed they all belonged to Bos taurus. This assumption is not only due to the cultural and economic context we are dealing with but also due to the conviction that wild cattle, or aurochs (Bos primigenius), were already extinct in Iberia since the Iron Age (Davis 2006:23).

The very low presence of equids is expected since all the other Portuguese Roman sites show the same trend (Cardoso & Detry 2005:376; Davis 2004:2; Davis 2007:5). Nevertheless, horsemeat seems to have been consumed in Tróia, which is supported by the presence of a chopped radius.

Red Deer Cervids are represented by only two bones (Table1): one distal tibia and one upper premolar 2. The tooth belongs to an adult animal and it was identified as Cervus elaphus. Therefore, the same identification is considered to be possible for the distal tibia.

Cattle Cattle are represented in Tróia by only eight bones (Table1): four of them are from the NW area, two from tank 1 and the remaining ones from tanks 5 and 7c. Their

Red deer has a very low representation in Portuguese Roman sites but its presence is somehow frequent, since it is a source of meat. Moreover, other products can be used from this animal, such as its antlers. Red deer is known to have ever been domesticated, which is then a 70

clear proof of hunting activities. However, it is not yet clear where hunting took place since there is no woodland environment in the surroundings of Tróia.

fusing. This individual is, therefore, considered as a subadult. Cat Felis catus is represented by 14 bones (Table 1), being 13 from tank 1 and only one from tank 5. All bones are complete and all anatomic elements are different, which suggests they might have belonged to the same individual. Based on cat’s epiphyseal fusion (Habermehl 1975), it is concluded the animal was younger than 10 months.

Pig Pig is the second most frequent species in Tróia (Table1) and it is present in all the excavated compartments. Generally, wild boar is larger than domestic pig. However, the Iberian wild boar appears to be of a relatively small size (Davis 2006:26). No distinctions between pig and wild boar were then possible. None of the Sus sp. bones seemed to show an irregular large size to be considered as wild boar. Nevertheless, it cannot be ruled out the hypothesis of wild boars being part of Tróia’s bone assemblage.

A fracture was identified on a right femur. However, this pathology was cured and it was not the cat’s cause of death. This suggests that the animal might have been a pet, whose owner would have taken good care of.

From all pig bones only one could be identified as being from a male animal, which identification was based on a right mandibular canine. Through epiphyseal fusion of the long bones, it was possible to determine that 67% of pig early fusing bones (Reitz & Wing 2008:72) were fused, which indicates that the majority of these animals were killed before they were one year old. This is related to the consumption of piglets that are known to have much tender meat. Only one mandible was identified as being of a juvenile.

Lagomorphs Lagomorphs are the fourth most represented in Tróia’s assemblage with 36 bones identified (Table1). Although no distinctions were made between rabbit and hare, the most probable is that rabbits mainly composed the assemblage. This suggestion is supported by the size of humeri and calcanei. According to Callou (2003, cited by Davis 2007:5), rabbits were only domesticated in Medieval times. The idea that the rabbits in this assemblage were wild is also corroborated by the animals’ size, essentially based on measurements such as the breadth of the distal end (Bd) and the greatest length from the caput (GLC) of the humeri.

Caprine Caprines (sheep and goat) are the most numerous mammals represented (Table 1). Identification to species is very difficult and it was only possible for 20 bones out of 102, which were classified as Ovis aries. No goats were identified. Caprines were found in all Tróia’s compartments but tanks 8 and 9, which are the only ones dating from the 1st and 2nd centuries. So far, it is not possible to determine if this is just a coincidence or if in the oldest periods of occupation caprines were not consumed at all.

Although no cut marks were identified, three femuri, a metatarsus III and a pelvis had been chopped. No carnivore marks were detected so it is assumed the presence of lagomorphs is due to human consumption and not the result of later intrusions or carcasses brought in by other predators.

Most caprines were killed at an early stage of their lives, since 72% of the early fusing bones (Reitz & Wing 2008:72) were fused and one mandible was identified as a juvenile. It can, therefore, be suggested that caprines were mainly explored for their meat and not for secondary products (such as milk and wool).

Rats In Tróia’s workshop 2 assemblage, seven bones were identified as being of Rattus sp (Table1): three humeri, two pelvis, one femur and one tibia. Since all the humeri are from the left side, the MNI for this group is of three. All bones were found in tank 1, except one humerus recovered from tank 7c. Although identification to species was not possible, it is probable that these remains belonged to Rattus rattus (black rat).

Dog Dog was identified as Canis familiaris. The possibility of these bones belonging to a wolf is excluded due to the remains’ small size and the urban context where they were found. Dog is represented by three bones that are possibly from one same individual (Table 1): one left scapula, one left metatarsus IV and one pelvis. All elements are fused but the proximal metatarsal that is

Chicken A total of 43 bones were identified as Gallus sp. (Table 1). The most probable is that they were domesticated for food purposes, which is supported by the identification of four chopped marks and two cut marks. Since only one 71

coracoid is identified as a juvenile, it can be assumed that the main trend was to consume adults.

macrofauna bones, chiefly in long bones, metapodials and phalanges.

Aquatic Birds Burning A total of 11 bones were identified as being from aquatic birds (Table 1).

Burning was only detected in three bones from the whole assemblage. A brown stain was identified on one lagomorph’s distal humerus, whereas a black stain was registered on one acetabulum of a large size mammal and on a proximal phalanx of a medium size mammal. As such, it can be assumed that not many roasts might have taken place since, if that was the case, many other bones would present burning stains. Consequently, it is assumed that most of the food might have been boiled, in order to make stews and soups.

From the Alcidae family (auks), one carpometacarpus was found in Pátio 2B. No identification to species was possible. In the NW area another carpometacarpus was discovered, which probably is from an Alca torda (razorbill). The Anatidae family (ducks, geese, swans) is represented by eight elements: three from tank 7c and 5 from tank 9. Only two bones from tank 9 were identified to the genus Anas sp. and it can possibly be from an Anas platyrhynchos (mallard). Ducks could have been used as a food resource since two butchery marks (a chop mark and a cut mark) were identified in two bones, one tibiotarsus and a humerus.

Gnawing Only 2% of the bones present carnivore activity. This indicates that most of the bones were rapidly disposed off after their meat consumption and promptly buried.

Finally, the Sulidae family (gannets) is represented by one proximal humerus of Morus bassanus (gannet) that was collected in tank 7c.

Only one amphibian element was found in the whole assemblage (Table 1). It is a femur from a species of the Anura order, which means a tailless amphibian. This frog bone was found in tank 1.

Gnawing was only recorded for 10 bones. They were gnawed by carnivores; no evidence of rodent gnawing was found. As expected, gnawing happened on long bones’ ends since they are easier to chew and contain large amounts of marrow. The majority of gnawing marks might have been done by dogs, which is suggested by the size and shape of the puncture marks recorded in three elements. It then seems plausible to assume that, after meat consumption, some of the bones were discarded to the dogs.

Taphonomy

Other Modifications

From the 663 bones recorded, 37% present some sort of taphonomic modification. It must be noted that the majority of modified bones are from mammals and only 13 are from birds.

About 20% of Tróia’s assemblage shows evidence of other modifications. The majority (18%) of these modifications concern insect activity, which is recognized by the presence of a speckled surface on bones. This might be the result of termite gnawing of the bone, which happens more intensively in fresh bone than in old bones. Root-etching was detected in only 2% of the assemblage.

Amphibians

Butchery Only 20% of the bone assemblage presents butchery marks. 15% of these marks are registered as chopped marks and are related to the breakage of bones with a cleaver, in order to dismember the animals. Once the animal was cut into pieces it would be easier to transport and cook it. The greatest incidence of chop marks is on medium and large size macrofauna.

Body Part Representation Body part representation was only studied for the most abundant species: pig, caprines, lagomorphs and chicken. To what concerns pig (Figure 4A), there is a significantly greater presence of forelimbs to hindlimbs. This trend does not seem to be related with differential bone density since low-density bones (such as the ulna) are present, whereas dense bones (like the femur) are completely absent. Consequently, it is valid to conclude that preferential selection of meat joints was taking place. Quite clearly, the data suggests a preference for the consumption of pig’s shoulder and front legs. Tróia’s

Cut marks are the second most represented marks (4%). In Tróia, they are mainly present on the distal ends of leg bones and are generally transversal to the bones’ orientation. Finally, marrow extraction was recorded in 2% of the bones. They were all recorded in medium size 72

Figure 3. Distribution of %NISP per animal groups from the excavated compartments in Tróia’s workshop 2 (2007/08).

Figure 2. Distribution of %NISP through the excavated compartments in Tróia’s workshop 2 (2007/08).

The chicken skeleton appears to be relatively well represented in Tróia since almost all elements were found (Figure 4D). Just like lagomorphs, the femuri and humeri are the most frequent elements, which is mainly due to their high bone density and also to the large amount of meat they carry. An identical body part representation was found in Quinta das Longas, where it should be noticed the absence of exactly the same elements as in Tróia. These results support the theory that these absences are due to low-density levels of those bones.

pattern of pig’s body part representation is very similar to the one of Quinta das Longas and to other Portuguese Roman sites, where it is evident the predilection for shoulder, front and back legs. On the other hand, the scarcity of metapodials and phalanges might be closely related to the little meat they bear. The caprine body parts reflect a different pattern to that of pig, since almost all skeletal parts are present (Figure 4B). Overall, there seems to be a more-or-less equal selection of hindlimbs and forelimbs. The humerus and the radius are the most represented limbs. Other meat bearing bones, such as the pelvis and the tibia, are very high, while the femur is quite low. Although not containing much meat, mandibles are the most represented elements. Such an even representation of skeletal parts suggests that complete animals were present on site. Considering other Roman sites, it is evident the high percentage of long bones and pelvis, which is closely related to the great amount of meat they bear.

Discussion and Conclusions Domestic Animals Domestic animals form the great majority of Tróia’s faunal assemblage. They were mainly kept for food purposes but also as pets. Equids and cattle are almost absent of the record. Due to their large body size, these animals are not easy to maintain and the food they provide can be found in other smaller species. In addition, the scarcity of these large mammals indicates that no agricultural activities were taking place in Tróia, at least in such an intensive way that would require the use of traction. This suggestion is corroborated by the absence of pathologies.

Lagomorphs are mainly represented by femuri and humeri (Figure 4C). This can be explained by the high density of these elements and to the fact that both bones bear larger amounts of meat. It is curious to note the complete absence of metacarpals when compared with the metatarsals that are relatively well represented. Since all metapodials have the same amount of flesh and similar density, their preservation rates are alike. Therefore, the explanation of a better representation of the metatarsals over the metacarpals might be related to dietary preferences. When looking at other Portuguese Roman sites, an identical pattern is found: good representation of femuri and humeri and complete absence of metapodials.

Pigs and caprines are the best-represented species in Tróia’s workshop 2. As they were mainly kept for meat, they were slaughtered before they were two years old in order to get tenderer meat. As such, only few animals were kept to older ages for reproduction purposes. Therefore, and for what caprines are concerned, the exploration of secondary products (like wool or milk) might have been low. Contrasting with large size animals, pigs and caprines are very easy to maintain within a small family group. Caprines are not very selective in the bushes they eat, whereas pigs can be easily fed with human meals’ leftovers. Consequently, the small effort put in the maintenance of these medium size animals has a very high outcome.

73

74

C

B

A

Figure 4. Body part representation of the most abundant species in Tróia’s workshop 2 faunal assemblage: (A) Pig. (B) Caprines.(C) Lagomorph. (D) Chicken.

D

The same can be applied to chicken that can survive off food scraps. Since only one chicken bone was identified as juvenile, it can be assumed that chickens were mainly kept for their eggs and then killed at an older age for their meat. The presence of butchery marks on some of the bones confirms the consumption of adult animals.

References Albarella, U. (2005) - Alternate fortunes? The role of domestic ducks and geese from Roman to Medieval times in Britain. In G. Grupe & J. Peters (eds), Feathers, Grit and Symbolism. Documenta Archaeobiologiae III, pp. 249-258. Cardoso, J. & Detry, C. (2005) - A lixeira baixo-imperial da uilla da Quinta das Longas (Elvas): análise arqueozoológica e significado económico-social. Revista Portuguesa de Arqueologia, 8, pp. 369-386. Davis, S. (2004) - Animal remains from Roman and Medieval Tomar. Trabalhos do CIPA nº68. Lisboa: Instituto Português de Arqueologia. Davis, S. (2006) - Faunal Remains from Alcáçova de Santarém, Portugal. Trabalhos de Arqueologia 43. Lisboa: Instituto Português de Arqueologia. Davis, S. (2007) - Mammal and bird remains from the Iron Age and Roman periods of Castro Marim, Algarve. Trabalhos do CIPA nº107. Lisboa: Instituto Português de Arqueologia. Étienne, R., Makaroun, Y. & Mayet, F. (1994) - Un Grand Complexe Industriel a Tróia (Portugal). Paris: Diffusion E. de Boccard. HabermehL, K-H. (1975) - Die Alterbestimmung bei haus und Labortieren. Berlin: Parey, pp. 177. Pinto, I., Magalhães, A.P. & Brum, P. (2011) - O complexo industrial de Tróia desde os tempos dos Cornelii Bocchi. In J.L. Cardoso & M. Almagro-Gorbea (eds) Lucius Cornelius Bocchus. Escritor Lusitano da Idade de Prata da Literatura Latina. Colóquio Internacional de Tróia 6-8 de Outubro de 2010. LisboaMadrid: Academia Portuguesa da História e Real Academia de la Historia, pp. 133-167. Reitz, E. & Wing, E. (2008) - Zooarchaeology. Cambridge: Cambridge University Press.

The identification of dog and cat remains is a good indicator that some animals might have been kept as pets. The evidence of one cat’s healed femur fracture supports this idea. Wild Animals The presence of wild mammals in a faunal assemblage is generally interpreted as the result of hunting activities. This is the interpretation for deer and lagomorph remains since the evidence of cut and chop marks prove that their presence is not due to the activity of any other predator but humans. However, the same interpretation should not be applied for rodent bones. In fact, where human’s dwell, rodents will not be far behind surviving off the rubbish. Wild birds are represented by aquatic species from estuarine environments, such as ducks, auks and gannets. It is interesting to notice the complete absence of terrestrial birds, such as partridges or bustards, which are generally found in Roman sites. Although it cannot be assessed if these aquatic wild birds were being hunted, there is evidence of butchery marks on two elements of duck. Even though some Roman texts refer to duck husbandry, it seems that the Romans were not yet able to fully domesticate these animals (Albarella 2005:254). Therefore, their consumption must be interpreted as occasional. Food Provisioning In general, it is assumed that a meat producer site exports its best meat pieces to other sites. Conversely, a meat consumer is characterized by importing the best meat pieces. Hence, in the latter, skeletal parts with less meat bearing (like mandibles, for instance) are absent. Nevertheless, such a linear approach cannot be always followed and that is the case of Tróia. Based on the body part representation of caprines, pigs and chicken, which are of extreme completion, it is clear that those animals were kept on site. However, Tróia should not be considered as a producer site since it seems more likely that those animals were kept within small family groups in order to assure their subsistence and not for commercial purposes. Besides, any kind of stables that would facilitate a larger production were not yet found.

75

The faunal assemblage from a roman well in the villa of São Miguel de Odrinhas (Sintra, Portugal). A preliminary view on the archaeological context Alexandre Gonçalves1 1

Museu Arqueológico de São Miguel de Odrinhas, Sintra, Portugal.

Abstract We present in this article the faunal remains recovered during the excavation of a roman pit associated with the villa of São Miguel de Odrinhas (Sintra, Portugal). I describe the contexts where this assemblage came from and disclose a first classification of the bones. The assemblage is mainly composed of complete skeletons that were deposited intentionally, although a large number of bones belonging to rodents, small insectivores, reptiles, amphibians and small birds may have an intrusive origin. Those skeletons that were probably intentionally deposited by humans in the pit for reasons of disposal or even ritual belong mainly to sheep and dogs; there were also some cases of isolated crania of cattle and one skeleton of a cat. It was concluded that most of the remains do not represented disposal of food left-over’s. Keywords: Animal bones; roman pit; villa. subject of this preliminary study. Still in the same decade, excavations directed by Catarina Coelho identified a possible sewer of the roman villa, as well as some archaeological materials possibly connected to a late occupation and several graves of the medieval and modern graveyard (Coelho, 2007). Subsequently, new excavations directed by this author identified the well, with its faunal remains described herein (Figure 2). The complete series of excavations allow us to ascertain a human presence in São Miguel de Odrinhas from the second half of the 1st century BC, followed by an important roman occupation between the 1st century AD and the beginning of the following century, in correspondence with the construction of the primitive villa of São Miguel de Odrinhas.

The archaeological site of São Miguel de Odrinhas The archaeological site of São Miguel de Odrinhas is located in the municipality of Sintra, whose territory was in Antiquity a part of the ager of the city of Olisipo, and a part of the wider network of roman settlements in this region (Figure 1). This network consisted of somewhat urbanized towns (vici) and, mainly, large farms (villae), not counting smaller rural properties (Alarcão, 1998). The western area of this vast territory, where São Miguel de Odrinhas is included, forms a coherent cultural, social and economic unit, as the local epigraphy can attest (Cardim Ribeiro: 1982-83: 156). Among the archaeological remains still extant in São Miguel de Odrinhas, the northern apse of the church was an immediate focus for the attention of investigators. However, the first archaeological intervention was made in 1949, by Camarate França. Later work was carried on by Fernando de Almeida in 1957, exposing many of the now visible structures, such as the complete foundations of the apse and its connection to the adjacent buildings, including a room with a mosaic pavement and rooms with remnants of similar flooring. It was during the interventions of the 1950s and 60s that a large part of the medieval graveyard was excavated. This necropolis developed around the medieval church dedicated to São Miguel, on top of the roman ruins. Further preventive interventions took place during restoration and other works. These were directed by José Cardim Ribeiro and Élvio Sousa in 1988, and they revealed several divisions of the roman house, including a room with brick structures that seem to indicate the existence of a hypocaust. During the 1990s, new works took place, due to the building of the new Museum of Odrinhas, during which several more structures were identified, including a water conduit in opus signinum located a few metres away from the well that is the

Figure 1. Location of the villa from São Miguel de Odrinhas in the agri of the Olisiponense Municipium. (Drawing of: Joel Marteleira, MASMO).

77

Figure 2. Plan of the villa with the location of the well. (Drawing of the plan: Joel Marteleira, MASMO). Eventually the main house underwent alterations, being converted into an Áulica villa. Witness to this profound remodelling is the apse with adjoining structures, as well as the mosaic pavement (Figure 3). This probably happened during the end of the 3rd century AD or the beginning of the following century (Cardim Ribeiro, 1993). The proposed dating for the best preserved mosaic is the first half of the 4th century AD (Caetano, 1997: 6382).

naturally associated with the Christian temple. This problematical change is the subject of much discussion – see for example (Pereira, 1914; Correia: 1928, Almeida, 1958 e 1962; Lambrino, 1952; Palol, 1967; Almeida, 1986; Maciel e Baracho, 1992; Maciel, 1999; Alarcão 1994; Cardim Ribeiro, 1993 e 1994). Central to the question has been the chronology and function of the apse. Several propositions have been put forward as to the date of its construction and its original function. Dates vary widely from the early 4th century to the 6th or even 7th century AD.

The roman occupation of the site must have continued at least until the beginning of the 5th century AD or even the early 6th century, judging from the presence of African red slip ware D and late roman C ware recovered in 1997 (Coelho, 2007). Besides these imported Mediterranean ceramics, excavations have also revealed other materials that indicate, albeit indirectly, a later human presence, namely from the 7th and 10th centuries (Coelho, 2007: 134).

The first modern interpretations concerning the monument were made prior to archaeological excavations, when the apse is one of few visible elements, north of the church of São Miguel. In that context Virgílio Hipólito Correia correlates the structure with a primitive Christian baptistery, a proposition considered feasible by Scarlat Lambrino (1952), who nevertheless did not adopt it. This perspective distances itself from the oldest, by Félix Alves Pereira, who in 1914 considers the edification a mausoleum.

The evolution of the site beyond this later stage is still poorly documented, and there are many uncertainties concerning the transition from the Roman period to the early Middle Ages until the building of the presently standing church dedicated to São Miguel. For this discussion particular relevance has been given to the transformation of the site into a cult area and of the habitation structures transformed into a cemetery,

After the initial excavations at the site, Fernando de Almeida (Almeida, 1958 and 1962) interpreted the structures discovered as belonging to a palaeochristian temple, built in the transition from the 4th to the 5th

78

Figure 3. View of the villa, with the mosaic and the abs. century. This is accepted by Pedro de Palol (1967) who, however, postpones the chronology to the 6th or even 7th century AD.

time, adapted into a palaeochristian temple (Idem, 1993: 110-111).

Carlos Alberto Ferreira de Almeida (1986) suggested additionally that the apse was a totally new construction built upon the pre-existing roman walls, a part of a new 6th century edification that integrated the older structures.

Archaeological context of the faunal assemblage The well of the villa of São Miguel de Odrinhas is located to the southwest of the main habitation structures, apparently peripheral to the central apse of the main house. A few metres away from the well the mentioned construction in opus signinum had been identified in the 1990s, and it may have been connected to the withdrawal of water from the well. Some metres to the north, next to the habitation structures, are the previously mentioned brick structures, possible evidence for a suspensura and the existence of a hypocaust (cf. Figure 2), whose function is not evident, but which may be connected to the existence of a heated room or even the baths of the villa (Coelho, 2007: 122).

Justino Maciel and Carlos Baracho (Maciel and Baracho, 1992) oppose this view. According to them the erection of the apse was contemporary with the immediately adjacent structures, forming an architectural entity that would have been a mausoleum with an atrium, datable to the 4th or 5th century AD. According to these authors, this previous burial tradition would have been in the origin of the medieval graveyard, and possibly of the transformation of the site into a place for worship. These propositions for the apse structure that identify it with a palaeochristian temple or a mausoleum have been met with scepticism by some investigators, who in the context of a villa believe it to simply be a room of the main house (Hauschild, 1986: 152), possibly its triclinium (Alarcão, 1994: 62) or its oecus (Caballero Zoreda e Sánchez Santos, 1990: 449). This opinion is shared by Cardim Ribeiro (1994: 83), although nothing prevents those structures from having been, in a later

The well has a circular shape with a diameter of around 4 meters, with the remains of a part of the stone structure that originally surrounded it. This was formed by a wall of limestone blocks with a mortar of sand and lime. Its profile is irregular; being wider at the top and it narrows down progressively from 2 metres deep, having a width of approximately one metre at the bottom (Figure 5).

79

Figure 4. View of well during excavation. It was filled in with building material, namely limestone blocks of varying sizes, as well as fragments of brick, tile and mortar. Some deposits are almost exclusively formed by small materials, such as limestone gravel and fragments of bricks and tiles. Were also collected many tesselae in those deposits, some still with mortar attached, indicating the possible destruction of a pavement of mosaic. Among the building materials, some ceramic fragments were also recovered belonging to containers for transport and storage (dolia and amphorae) and domestic ware (common ceramics and terra sigillata), and also a considerable quantity of faunal remains. Here I present the contexts of the negative structure filling in an attempt to contextualize them within the occupation of the roman villa of São Miguel de Odrinhas. A thorough analysis of the artefact assemblage is indispensible for the correct characterization of the contexts and their cultural attribution. At present some of these materials are still being studied, and for this reason a selection of artefact categories was done for a more precise definition of the chronology of the fill, namely the fragments of amphorae and terra sigillata. Besides these specific materials, two complete recovered items are presented, a pot and a jar. The faunal characterization

assemblage:

a

preliminary

The recovered faunal assemblage was located through the different deposits, mixed with the construction materials, and different situations have been registered according to the nature of the depositions.

Figure 5. Profile of the well with the indication of the different levels. (Drawing of: Joel Marteleira, MASMO.

The assemblage is mainly composed of complete skeletons, which in some cases may correspond to intentional depositions, whereas others may have an intrusive origin, which will account for the high quantity of bones belonging to rodents, small insectivores, reptiles, amphibians and even small birds (Figure 6).

80

Figure 6. A skeleton in the well of a goat or sheep.

Figure 7. skull of a goat (left) and skull of an ox (right). Amongst the assemblage of skeletons that may document an intentional human action, we found those of sheep and goats, dogs and probably also cats. The presence of these skeletons inside the well may reveal different events, including a possible utilization for disposal of food leftovers. However, at the present stage of the study of these contexts, we cannot exclude other possibilities. In fact, a preliminary study of the faunal assemblage reveals no concrete evidence that we can clearly associate to their preparation for consumption.

placed at the same level, facing opposite directions, without the rest of their skeletons (Figure 7). Considering that a large part of the faunal assemblage is still unstudied, it will be indispensable to undertake an exhaustive work that may allow us to form a more detailed characterization of its composition, to understand the nature of the deposit. Chronological integration

One of the aspects we can consider that may potentially reveal a particular nature for the faunal remains in the well was the presence of two skulls, of an ox and a goat,

Among finds recovered from the well there are two ceramic pieces that were recovered in their entirety, even

81

Figure 8. [39] - Gaulish terra sigillata (Drag. 24/25); [50] - African red slipe ware D (Hayes 67); [50] - African culinary pottery type-A (Hayes 181). Scale 1/3. (Drawing of Ana Neves, MASMO).

Figure 9. Amphoras: Keay XVI from the level [43] and Almagro 51 C from the level [52]. Scale 1/3. (Drawing of Ana Neves, MASMO). if very fragmented, namely a jar, placed at the bottom of the well, and a pot, in one of the last deposits, near the top. Besides these two elements, there is no evidence for other complete pieces. Other remains of pottery derive from differing productions and chronologies, and mixed with the rest of the elements of the fill.

of water has been found. The oldest material belongs to the gaulish terra sigillata a shard of the Dragendorff 24/25 shape (Figure 8). Hispanic productions are present in a fragment belonging to the Dragendorff 27 shape, as well as a late production fragment of the Dragendorff 37.

Contrary to what might be expected in a well over seven metres deep, other than the jar at the bottom no other concrete evidence for items connected with the collection

North African productions make up the majority of the recovered artefact assemblage, being present since the earliest stages in African red slip ware A of the Hayes 14/17 shape and African C of the Hayes 52. The African D types are foremost in this assemblage, shapes Hayes 52, 58-A, 59 e 67 being found (Hayes, 1972). With the same geographic origin there is a dish fragment of the Hayes 181 shape in African culinary pottery (Idem). The fact that there is a majority of African ceramics in these contexts fits the known commercial patterns for the coastal regions of Lusitania – as already documented for the villa of São Miguel de Odrinhas (Coelho, 2007: 132). Besides the imported tableware ceramic fragments previously mentioned, some elements of transport and storage containers have been recovered. The Lusitanian productions are documented in the Lusitanian 3, Keay

Figure 10. The pot collected on the top of the well (scale1/4). (Drawing of Ana Neves, MASMO).

82

Figure 11. The jar placed on the bottom of the well (scale 1/4). (Drawing of Ana Neves, MASMO). XVI and Almagro 51 C amphorae, and a Dressel 20 indicates the importation of olive oil from Baetica (Figure 9). The recovered pot from one of the final debris deposits (Figure 10) does not provide a very safe chronology, since it belongs to a common ceramics production used over an extended period. The other complete piece, the jar recovered from the bottom of the well (Figure 11), is also local or regional in origin, but it does have a few typological features, including the manufacture, that seems to point to a later time, consistent with the dating for some of the finer chronology materials.

relevant, this paper should be viewed as departure point towards a necessary and more complete study that allows for a more solid assessment. However, with this in mind, we believe that with the information already available a few propositions may be suggested concerning the process that led to the de-activation of this well. According to the data presented here, we conclude that the content of the well is a mixture of materials from a wide range of periods from the 1st century AD and the 4th to early 5th century AD. The archaeological contexts indicate the presence in the same strata of material with very different dates, with even some of the more ancient elements in the upper levels. This evidence seems to document a concrete action of termination of the well, by means of filling it to the top with whatever debris might be found in the surrounding area. This could explain the chaotic mixture of building materials with objects belonging to the different phases of the villa’s activity. This seems to diverge from the perspective of a slower process with episodical actions during a more extended time. From the analysis of the artefactual assemblage, and judging from the later materials, we can only deduce that this action may have occurred at least from the first quarter of the 4th century AD onwards. The difficulty in dating the end of the well may partly be resolved by the dating of the recovered organic materials, not just fauna, but also some small pieces of wooden plank found at the bottom.

In spite of this wide interval between earliest and latest materials, their stratigraphic position did not allow a definition of sequential phases for the process of in filling of the well with chronological correspondences. The characteristics of the deposits and the placement of the materials seem to indicate a relatively quick filling, not one in different stages along a longer span - possibly between the second half of the 4th century and the beginning of the 5th. As noted, this initial approach to the recovered material assemblage does not consider an analysis of the more generalized common ceramics, since a complete and detailed study of these has yet to be undertaken. Therefore, these first few data must be interpreted with some caution, until a more comprehensive study has been done.

In the context of the dynamics of the site itself, particularly in late imperial times, it is possible that the filling of the well was connected to an episode of abandonment of some habitation structures or, even, the re-founding of the edifices, in an action of deep transformation of some buildings. As previously seen,

Discussion and future perspectives This being a preliminary approach to an archaeological context in which the faunal remains are particularly

83

there remain many questions concerning the alterations that will have taken place at the villa in a later time, including the question of the chronology and function of the apse or its hypothetical transformation into a palaeochristian temple. Besides the diversity of purposes that have been suggested, the supposed chronologies differ widely, between the early 4th century and the 7th century AD. The recent identification of 7th to 10th century materials from São Miguel de Odrinhas seems to refute the possibility of a complete and effective abandonment of the roman structures in the transition from the 5th to the 6th centuries AD, although it is still difficult to trace a human presence in the transition to the early Middle Ages (Coelho, 2007: 134).

Alarcão, Jorge de (1998) – Três níveis de aglomerados populacionais romanos. O Arqueólogo Português. Lisboa, Série IV, 16, p. 175-186. Almeida, Carlos Alberto Ferreira de (1986) – «Arte paleocristã da época das invasões», História da Arte em Portugal, Vol. 2. Lisboa/Barcelona: Publicações Alfa, pp. 9-37. Almeida, Fernando de (1958) – «Escavações em Odrinhas», Comunicações dos Serviços Geológicos de Portugal, XXXIX, Lisboa, pp. 11-25. Almeida, D. Fernando de (1962) – «Arte Visigótica em Portugal. A basílica de Odrinhas», O Arqueólogo Português, II Série, IV, pp. 113-118. Bonifay, Michel (2004) - Études sur la céramique romaine tardive d’Afrique. Oxford: BAR Int. series 1301 Caballero Zoreda, L. Sánchez Santos, J. (1990) – Reutilizaciones de material romano en edificios de culto cristiano, Congreso sobre Cristianismo y Aculturación en tiempos del Imperio romano Universidades de Murcia y Complutense de Madrid (Madrid 1988) Antigüedad y Cristianismo, 7, pp. 431-493. Caetano, Maria Teresa (1997) – Musivária Olisiponense. Estudo dos Mosaicos Romanos de Olisipo e da “Zona W” do Ager (Dissertação de Mestrado apresentada à Faculdade de Ciências Sociais e Humanas da Universidade Nova de Lisboa). Lisboa. Texto policopiado. Cardim Ribeiro, José (1982-83) – «Estudos Históricoepigráficos em torno da figura de L. IVLIVS MAELO CAVDICVS», Sintria, I-II. Sintra: Gabinete de Estudos de Arqueologia, Arte e Etnografia, pp. 151-476. Cardim Ribeiro, José (1993) – «A villa romana de São Miguel de Odrinhas (Sintra, Lisboa): novos dados, novas interpretações», Livro-guia do I Congresso de Arqueologia Peninsular (Porto 12-18 de Outubro). Porto: Sociedade Portuguesa de Antropologia e Etnologia, pp. 110-111. Cardim Ribeiro, José (1994) – «FELICITAS IVLIA OLISIPO. Algumas considerações em torno do catálogo Lisboa Subterrânea», Al-Madan, IIª série. Almada: Centro de Arqueologia de Almada, pp. 75-95. Coelho, Catarina. (2007) - Ruínas arqueológicas de São Miguel de Odrinhas: a propósito da campanha de 1997. Arqueologia e História: Revista da Associação dos Arqueólogos Portugueses. Vol. 58/59 , pp. 119-142. Correia, Vergílio (1914) – «No Concelho de Sintra – Escavações e Excursões», O Archeólogo Português, XIX. Lisboa, pp. 200-216. Correia, V. (1928) – «Arte Visigótica», História de Portugal, Vol. I. Barcelos. Gorges, Jean-Gerard (1979) – Les Villas Hispanoromaines. Paris: Editións du Boccard. Hauschild, Theodor (1986) – «Arte Visigótica em Portugal», História da Arte em Portugal, Vol. 1. Lisboa/Barcelona: Publicações Alfa, pp. 149-169. Hayes, John W. (1972) – Late Roman Pottery. London: The British School at Rome. Keay, S. (1984) - Late Roman amphore in the western Mediterranean. A tipology and economic study: the Catalan evidence. Oxford: BAR International Series, 196.

The well and the trench which is thought to be a sewer of the villa (Coelho, 2007), are located in an area south of the main structures of the main house. We should considered that the cession of use of these two structures may have coincided with a phase of contraction of the habitational area in late antiquity, although that does not mean that the site has been completely abandoned. The presence of African red slip ware D and late roman C ware in the contexts of closure of the likely sewer indicate it must have taken place after the mid-5th century, perhaps up until the middle of the following century. Those later materials are absent in the deposits that fill the well, that may have a slightly earlier date, between the first quarter of the 4th century AD and the first quarter of the following century. The differences in the artefactual composition of both assemblages may reveal a dynamic process of abandonment of those structures, in which the well is closed on an earlier stage, before the utilization of the possible sewer as a dumping area. However, we must consider the small quantity of elements recovered during the excavation of the well that can offer a fine chronology. For that reason, we should not exclude definitively that both contexts were created almost at the same time. In face of the different possibilities for explaining the process of filling the well, it becomes clear there is the need for an exhaustive study of all the archaeological materials, including fauna, to more clearly explain its formation. One of the most relevant actions will be to search for occasional evidence of a handling of the bones in a context of domestic consumption, or an attempt to identify possible causes of death, mainly of the complete skeletons. References Alarcão, J. (1994) - Lisboa romana e visigótica. Lisboa Subterrânea. Lisboa: Instituto Português de Museus, p. 58-63.

84

Lambrino, Scarlat (1952) – «Les inscriptions de S. Miguel d’ Odrinhas», Bulletin des Études Portugaises, XVI. Coimbra: Coimbra Editora, pp. 134-176. Maciel, Justino (1983) – S. Miguel de Odrinhas (Sintra): Arquitectura Romana ou Paleocristã?, dissertação policopiada. Universidade Nova, Lisboa. Maciel, Justino & Baracho, Carlos (1992) – O Monumento Absidal de Odrinhas (Sintra). Actas da III Reunió d’ Arqueologia Cristiana Hispânica (Mao, 1988). Barcelona, pp. 93-103. Maciel, Justino (1999) – A Antiguidade Tardia do «Ager» Olisiponense: O mausoléu de Odrinhas. Porto: Centro de Estudos de Ciências Humanas. Mayet, F. (1984) - Les céramiques sigillées hispaniques: contribution à l’histoire économique de la Péninsule Iberique sous l’Empire Romain. Bordeaux: Publications du Centre Pierre Paris, (Collection de la Maison des Pays Ibériques, 21). Palol, Pedro de (1967) – Arqueologia cristiana de la España Romana. Madrid/Valladolid. Paz Peralta, J. A. (2008) – Las producciones de terra sigillata hispânica intermedia y tardía, In: BERNAL Cassola, D.; Ribera Lacomba, A. (eds.). Cerámicas hispanorromanas. Un estado de la questión. Cádiz, pp. 497-540. Peakock, D.P.S. & Williams, D.F. (1986) – Amphorae and the Roman economy. London- New-York: Lomgman. Pereira, Félix Alves (1914) – «Por caminhos da Ericeira». O Archeólogo Português, vol. XIX. Lisboa: Imprensa Nacional, pp. 324-362.

85

A contribute to know the dietary habits in Tavira. From the Iron Age to the Modern Period Jaquelina Covaneiro1 & Sandra Cavaco1 1

Câmara Municipal de Tavira, Portugal.

Abstract In the past few years several archaeological interventions were completed in Tavira, particularly in the historical centre. In consequence, it has been possible to gather more data about the evolution of this town and its past inhabitants. As a result of the study of the fauna materials from Convento das Bernardas, Convento da Graça, Bela Fria and Parque de Festas it is possible to know the domesticated and wild species, the slaughter/butcher patterns, the method of flesh cutting and other taphonomic phenomena. Key-words: Tavira; Mammalogical fauna. periods: Modern Period (MP), Islamic Period (IP) and Iron Age Levels (IAL).

Introduction The town of Tavira is located in the South of Portugal, on the left bank of the Gilão River’s mouth. The fauna analysed in the present article is from four archaeological sites, located in the city’s historical centre.

Fauna materials from Bela Fria and Convento da Piedade were also found during town hall’s interventions while conducting the reconstruction of the sites.

Convento da Graça, Parque de Festas and Bela Fria are located in Santa Maria hill, while Convento da Piedade is in an area called Campo da Atalaia (Figure 1).

Archaeological works in Bela Fria revealed the existence of an Islamic occupation, dating from the Almoad period. The study of the ceramic materials dated this occupation from circa the 1st half of the 13th century (Covaneiro and Cavaco, 2012). In the beginning of the 14th century the site is thought to have been abandoned; its usage became agricultural, and the area was then known as Horta da Bela Fria.

The archaeological intervention at Convento da Graça was made by the town hall’s team and took place between 2002 and 2006. The fauna analysed is from two negative structures - a ceramic container and a garner, located close to each other - in the inferior floor of the cloister. The excavation of these structures revealed that after an initial utilization as storage they were sealed with domestic waste materials (Lopes et al., 2005: 313).

In the case of Convento da Piedade it was possible to confirm the hypothesis that the site was immediately occupied after is construction (1509) and remained so until the contemporary period. The materials analysed are from archaeological contexts dating from the 16th and 17th centuries (Covaneiro and Cavaco, 2010: 639).

The study of the ceramic materials contained in it concluded that these dated from the 15th and/or 16th centuries, when the city of Tavira experienced a substantial urban growing period and kept strong commercial trades with several areas in the Iberian Peninsula.

Methods All the bones were examined according to the CIPA´s osteological collection, used here as a standard reference. The measurements of skeleton parts (Desse et al., 1986) were obtained using vernier callipers and according to parameters defined by von den Driesch (1976).

In 2009, the fauna materials from Parque de Festas were recovered during an archaeological intervention carried on by Era, Arqueologia enterprise. The following analyses concluded that this particular area was occupied from the Final Bronze Age to the Iron Age. The levels with Islamic occupation displayed profound alterations as result of medieval and modern constructions, namely the construction of Palácio da Galeria. The final occupation was from the modern period (Pinto and Simão, 2009).

The ribs and vertebras were included in the nondetermined specimens. Regarding isolated teeth, we analysed both their wear and length of the occlusal surface. In sheep and goats’ third molar, we adopted Altuna’s (1980) method. For the height of goat, sheep and cattle’s dental crown, we considered the differential point of the vestibular roots (David and Payne, 1993). For swine’s third molar, we measured the length of the tooth in its maximum point (DMD) and the width of the first lobe (DVL).

From this intervention three out of 12 test excavations were analysed. The materials resulted from more than 100 stratigraphic units and were aggregated in three main occupational 87

3 1

2

4

Figure 1. Location of the archaeological sites in Tavira. 1. Convento da Graça; 2: Parque de Festas; 3: Bela Fria; 4: Convento das Bernardas. Additional measures were taken for Ovis aries and Capra hircus metatarsal/metacarpal (Davis, 1996; Bartosiewicz et al. 1993).

The alterations caused by weathering, the action of natural and biological agents (root growing and bone compression), humans, rodents and carnivore animals were considered, in order to analyse the taphonomical changes presented by the osteological remains.

The morphological distinction between goat and sheep remains was done with mandible and maxilla elements with partial dental series, specifically with the deciduous (milk) fourth premolar (dp4). The identified remains are from two individuals less than 24 months old, given that the dp4 is substituted by the third molar (M3) around two years of age.

Species present in fauna samples Table 1 presents the total number of specimens (TN), specifying the correspondent value to the non-determined specimens (ND) and to the number of identified specimens (NISP).

The chronology of the dental eruption and substitution, the level of wear of definitive dentition and the state of epiphysial fusion were considered, in order to establish the percentage of animals slaughtered before a determined age (Altuna, 1980; Carine Tomé and JeanDenis Vigne, 2003; Chaix and Méniel, 2001; Reitz and Wing, 1999; Stiner, 1994).

Consequently, it is possible to verify that the majority of the remains are from Parque de Festas, though the samples from Bela Fria provided a higher number of taxonomical identifications. The Convento das Bernardas

88

C. Graça

Bela Fria

P. Festas

B

CAH

O

CEE

S

F

L

ORC

631

2522

22

1

51

3

14

1

2

7

ND

714

C. Bernardas 2656

NISP

143

808

388

956

TN

857

3464

1019

3478

Tot al 101

Table 2. NISP - Parque de Festas, Iron Age. B- Bos taurus; CAH – Capra hircus; O – Ovis capra; CEE – Cervus elaphus; S – Sus sp.; F - Felis sp.; L – Lepus sp.; ORC – Oryctolagus cuniculus.

Table 1. TN - Total number of specimens. provided the second most numerous sample, followed by Bela Fria and Convento da Graça.

In what concerns the location of human modifications, in general these were subtle and must be related with disarticulation processes.

Iron Age

None of the remains showed fire alterations. Modifications by carnivore or rodents were not relevant, only being identified in four remains.

Both Convento da Graça and Parque de Festas presented occupational levels from the Iron Age. However, only Parque de Festas had its material (partially) analysed, being the one presented in this paper (Table 2).

The marks produced by root growing and action of associated microorganisms were reduced.

The archaeological work allowed the identification of a diversified set of residential structures. The study of ceramic materials infers that the site occupation during the Iron Age was around the 7th and 5th centuries B.C. (Pinto and Simão, 2009).

c) Deer (Cervus elaphus) Only three remains of this species were identified. The probability of belonging to one individual slaughtered before reaching adult age and with reduced bone alteration is high.

Mammals a) Cattle (Bos taurus)

d) Pig/Boar (Sus sp.) Only 14 remains of this species were registered. Nine are elements from the cranial skeleton with a high rate of loose teeth.

Most of the identified cattle remains belong to the appendicular skeleton (73%). From other parts of the skeleton, namely the cranial, we identified six teeth, which present intermediate wear.

Anatomical segments were represented by fused elements from the anterior appendicular skeleton. The dental elements did not present a high level of wear; thus, it is possible to consider that both individuals have been slaughtered at the verge of adult age.

The analysis of the epiphysis fusion state of long bone showed that all of them were fuse, thus suggesting that three of the identified individuals were slaughtered around five years old.

Cut-marks produced by humans were infrequent, as well as marks produced by carnivores or rodents. On the other hand, modifications by microorganisms associated to plant roots were prominent.

Some of the remains presented marks caused by human usage, in particular in the distal parts of epiphysis, metatarsal/metacarpal and in the feet elements. The later seem to be related with the disarticulation of the feet bones.

e) The Carnivores. Cat (Felis sp.)

The taphonomic changes caused by rodents and carnivore animals are mainly located in feet bone elements. b) Sheep/Goats (Capra hircus and Ovis aries)

Carnivores were only represented by one cat remain. It is possible that this animal was responsible for some of the bite marks displayed by some bones.

Only one dental remain of deciduous tooth was identified as belonging to goat.

e) The Lagomorpha. Rabbit (Oryctolagus cuniculus) and hare (Lepus sp.)

Analysis of the sample revealed that 45% of the elements unfused or showed the suture line, still being possible to see milk teeth. Consequently, it is possible to determine that all the individuals (seven) were slaughtered before reaching adult age. There were three infants and one juvenile.

There were identified two hare and two rabbit remains. In general the bones show slightly altered surface; some remains have marks produced by microorganisms associated to plant roots.

89

P. Festas

Bela Fria

Ursus arctos

1

1

Bos taurus

13

41

Capra hircus

7

2

Ovis aries

_

3

156

136

Cervus elaphus

8

5

Equus caballus

_

13

Sus sp.

13

47

Sus scrofa

_

1

Felis sp.

12

4

Canis familiaris

1

3

Martes foina

_

1

Lepus sp.

13

3

Oryctolagus cuniculus

405

128

Ovis aries/capra hircus

Islamic period The archaeological intervention in Bela Fria identified an Islamic suburb (Covaneiro and Cavaco, 2012). The diversification of exhumed objects showed that, in this location, several activities have taken place (spinning, pottery, metallurgy, etc.). Parque de Festas site excavation identified a diversified set of residential structures, even though the Islamic levels showed a strong deterioration (Simão and Pinto, 2009). The study of ceramic materials from both sites determined that the Islamic occupation occurred between the middle of 12th and beginning of 13th centuries. Mammals a) Cattle (Bos taurus) Bovines were identified in Bela Fria and in Parque de Festas, with appendicular skeleton elements prevailing in both sites (Table 3). It was equally observed, that the number of isolated teeth was reduced, which may indicate that the disarticulation process could have been done in a different location, and that less valuable parts (meat wise) remained in slaughterhouse.

Table 3. NISP - Parque de Festas e Bela Fria. Islamic Period. Some considerations The fauna collection was characterized by a reduced number of remains. However, it is possible to state that the remains represent a large fauna spectrum, with eight different species, both wild and domestic.

Unfused bones were not recorded and only a few elements still presented the suture line. As a result, the three individuals identified in Parque de Festas were subadults, while in Bela Fria there were four subadults and one adult.

Despite the contribution of wild species to this Iron Age population’s diet, domestic species remained a fundamental animal resource. Cattle, sheep and goats played the most important roles.

Human cut-marks were mostly located on the distal parts of the humerus epiphysis, metatarsal/metacarpal and foot elements. These cut-marks are related with the disarticulation of the several foot bones. They feature fine traces, maybe produced by metallic instruments such as knives.

The data relating to cat are scarce, so we can not determine their incorporation in the diet. On the other hand, the cat is a companion animal and is therefore likely their presence in these contexts.

Marks by carnivores or rodents are more abundant in the sample from Bela Fria.

Regarding the distribution of anatomical segments, it was showed that the anterior appendicular skeleton’s elements are predominant, with isolated teeth being the most represented within the cranial skeleton.

b) Sheep/Goats (Capra hircus and Ovis aries) Goat remains were common in both samples, while the sheep were only present in Parque de Festas.

Slaughter patterns seem to imply a diversity of species’ exploitation. Sheep, goats, cattle and pigs were used as meat supply. These animals were slaughtered once their physical growth reached a certain stage.

There was a high percentage of unfused elements or with the suture line still visible, as well as a high number of isolated milk teeth. Out of the 13 individuals identified in Bela Fria, four were slaughtered before reaching two years old. In Parque de Festas out of 10 identified individual, seven did not reached the 24 months old.

In general, the fauna revealed a high degree of bone fragmentation, caused by humans during food consumption.

In general, both sites fauna displayed human cut-marks on long bones distal portions (humerus, radius, ulna or the metatarsal/metacarpal). 90

Taphonomic analysis showed that remains with fire marks were residual in both sites.

Carnivore and rodent action is, again, more accentuated in the remains from Bela Fria.

Both samples included remains with carnivores and rodents’ modifications. Bela Fria presented more evidences related with the action of microorganisms of plant roots.

f) The carnivores. Dog (Canis familiaris), Cat (Felis sp.), Weasel (Martes foina) and Grizzly Bear (Ursus arctos) In general, the number of carnivores in the samples was sparse. Dog, cat, and bear were identified in both sites. The weasel was only identified in Bela Fria.

c) Horse (Equus caballus) All five horse remains registered were from Bela Fria [In the article “Proibições e tolerâncias. Os hábitos alimentares das populações em época islâmica (Tavira)” to be published in Arqueologia Medieval n.º 13 magazine, it is mentioned that the number of Equus caballus is five, instead of thirteen in reality] and must correspond to the domesticated species.

Bear presence was attested by the presence of two metacarpals, one in each site. The bones were fused and the muscle marks were well defined, thus suggesting the presence of adult individuals. It is not possible to state if the bears were part of the alimentary diet. Both samples have dog and cat’s remains, probably the animals responsible for the bite marks identified in the samples.

In one of the individuals, its long bones epiphysis’ level of fusion allowed an estimated age of death of three and a half years old.

Weasel was only identified in Bela Fria sample with one remain. Its presence could be random, however it is possible to consider other hypotheses.

None of the remains displayed human cut-marks; however, its usage as food should not be excluded. It is also possible that this horse served as draft animal.

Cut-marks were only present in two distal epiphyses of cat’s long bones (humerus and femur). These remains may belong to Felis silvestris, the wild cat; in this case, the cut-marks should be related with fur extraction process. It is not possible to determine if this cat was used as part of alimentary diet.

d) Deer (Cervus elaphus) There were identified 13 deer remains. Four were loose teeth. At Bela Fria one individual was slaughtered before reaching 12 months years old, and the other before adult age. Parque de Festas showed a similar pattern.

g) Lagomorpha. Rabbit (Oryctolagus cuniculus) and hare (Lepus sp.)

Although deer remains with cut-marks are rare, it is probable that these animals were used as food. Nonetheless, the usage of their skin as leather or some other elements as tools (e.g., antlers) should also be considered.

In Bela Fria rabbit’s remains corresponded to 33% of identified specimens, and in Parque de Festas to 64%. In Parque de Festas the number of remains corresponds to an accentuated presence of isolated teeth. This large number could be related with the method used to clean and store the fauna materials, which could lead to a detachment of the dental elements from the mandible or maxilla.

e) Pig/boar (Sus sp.) and boar (Sus scrofa) Bela Fria yielded 48 remains of swine: one was identified as Sus scrofa (boar) and all the others were impossible to identify to species. In Parque de Festas 13 remains were of Sus sp.

The epiphyses analysis suggests that nine out of 42 individuals identified in Parque de Festas were slaughtered before nine months old, while in Bela Fria only two out of 16 were slaughtered before nine months old.

Isolated teeth are more numerous in Parque de Festas and the anterior appendicular skeleton remains are prevalent. The Bela Fria sample revealed that four out of six individuals were slaughtered before reaching adulthood. Parque de Festas showed the same slaughter pattern (three juveniles). The boar identified in Bela Fria is an adult. This shows a diversified age of slaughter for this taxon: both juveniles and full developed animals were killed.

The hare is represented by three remains in Bela Fria and by 13 remains in Parque de Festas. The number of this animal in the sample is residual, a fact that complicates further observations. The remains present some alterations due to human cutmarks or to the action of carnivores or rodents. The small dimension of the lagomorphs could be responsible for a smaller presence of cut-marks, since they could be cooked in one piece.

Cut-marks are more numerous in the remains from Bela Fria, and they are probably related with dismemberment and disarticulation processes. 91

C. da Graça

P. de Festas

C. das Bernardas

B

20

37

52

CA H OV A O

7

1

2

9

2

99

99

On the other hand, the axial elements were not strongly represented. This fact leads us to consider that the animals could be slaughter in another site. However, this site, until today, has not been found in any archaeological work.

309 2

CE E EQ Q S

1 2

19

183

F

3

3

2

2

The location of this type of sites is given by historical registers. Consequently, for more recent periods, it is possible to know their location in particular after the Royal Crown inclusion of Tavira under its sovereignty, the “fangas” and slaughterhouses were part of the goods given to the King (Macieira and Manteigas, 2008: 36). According to the literature these site were “in the river bank, where Praça da Ribeira and the royal “teracenas” were located (ibidem: 36).

3

CA F L OR C

The distribution of the anatomical segments presents a prevalence of the anterior appendicular skeleton while the posterior elements are less frequent.

2

6

8

48

245

Table 4. NISP – C. da Graça, Parque de Festas e C. das Bernardas, Modern Period.

The Parque de Festas size 2 class remains (Uerpmann, 1973) correspond to 37% of the non-determined specimens, while at Bela Fria they correspond to 51%. Both samples present a considerable percentage of remains that must belong to medium size animals (sheep, goat, etc.), but whose high fragmentation prevented their anatomical and taxonomic determination.

B- Bos taurus; CAH – Capra hircus; O – Ovis capra; CEE – Cervus elaphus; EQQ – Equus sp.; S – Sus sp.; F - Felis sp.; CAF – Canis familiaris; L- Lepus sp. ; ORC – Oryctolagus cuniculus. Due to their small corporal dimensions, hare and rabbit would provide a reduced meat supply to the population. However, they would have been essential as protein suppliers, in particular during the periods were there was a real lack of food.

Medieval/Modern Period Parque de Festas archaeological excavation identified a medieval/modern occupation, which can be dated to a period between the end of 16th century and the 17th century. The archaeological structures functioning as a residential complex are related, partially, with Palácio da Galeria.

Some considerations The fauna materials from Bela Fria and Parque de Festas highlight the domesticated species contribute (cattle, sheep and goats) as food supply to the community. Nevertheless, the record of some wild species reveals the importance of hunting, as complementary activity to these human communities, especially in periods were there was a shortage of food.

The materials from Convento da Graça were analysed together, since they have the same functional context and similar chronologies (15th and 16th centuries). The Convento das Bernardas materials were mainly from litter disposal areas of the convent. The study of the ceramic materials showed that they date from the 16th to 18th centuries.

However, it is impossible to determine if some of these wild species were included in the human diet. Such are the cases of the carnivores. The age at death of sheep and goats demonstrates a diversified exploitation of these species. These animals were used as meat resources, and as suppliers of secondary products, as wool or milk.

Mammals a) Cattle (Bos taurus) All the sites had bovine remains, represented mostly by appendicular skeleton elements (Table 4).

In the case of cattle it was possible to observe a slaughter pattern of subadults or adult individuals.

The dental elements were absent from Convento das Bernardas materials, but present in the other two sites.

Swine remains displayed cut-marks, which can raise questions about the religious rituals related with their consumption. Islamic populations, in periods of lack of food, could consume forbidden products like swine; or they could be a result of consumption by “moçárabes” populations living in Tavira.

In Convento das Bernardas and Convento da Graça there was an accentuated presence of distal foot elements, unlike Parque de Festas.

92

After the analysis of long bones fusion state, it was possible observe that in all the samples the slaughter occurred in subadult or adult animals. There was only one individual with less than 24 months old.

In Convento das Bernardas the slaughter pattern was more diversified, since two of 10 individuals did not have more than one year old when killed, and one did not even reach two years old.

Bone surface observations revealed carnivores or rodents action, as well as the action of microorganisms associated with plant roots.

The samples of Convento das Bernardas and Parque de Festas suffered taphonomic alterations, as a result of microorganisms associated with plant roots and carnivores and rodents bite marks.

b) Goat (Capra hircus) and Sheep (Ovis aries) g) The carnivores. Dog (Canis familiaris) and cat (Felis sp.)

Goat was identified in all the archaeological sites, and sheep was only present in Convento da Graça and Parque de Festas.

There were identified cat remains in all archaeological sites, while dog remains were only identified in Convento das Bernardas.

c) Goat/Sheep (Capra hircus or Ovis aries) These species were identified in all the archaeological sites. In Convento das Bernardas anterior appendicular skeleton elements prevail, while in the other two sites it was observed a balance between the anterior and posterior skeleton elements.

The number of remains of these two species is residual. Nevertheless, the presence of bite marks, especially in the distal parts of long bones of other species, leads us to draw the hypothesis that these two species were abundant in the Convent, despite the lack of direct evidences in the archaeological materials.

The number of dental elements is reduced in Convento das Bernardas.

h) The Lagomorpha. Rabbit (Oryctolagus cuniculus) and hare (Lepus sp.)

The analysis of the slaughter patterns of the animals revealed an elevated percentage of individual killed before reaching 12-24 months old. In Parque de Festas, five of the 10 identified individuals were killed before attaining one year old, and two before two years old.

Hare was only identified in Parque de Festas and Convento das Bernardas, always in reduced number. Rabbit was present in all the sites, its presence being more significant in Convento das Bernardas.

Post-mortem alterations are present in all the samples. The remains were affected by microorganism associated to plants roots, as well as by bites inflicted by carnivores and rodents, especially in the distal parts of long bones.

The epiphysis state in the hare long bones suggests the presence of individual adults. The majority of rabbit remains tells us that they were killed in adult age.

d) Deer (Cervus elaphus)

Some considerations

This specie was only identified in Convento das Bernardas, a mandible with part of its dental series and one loose tooth. The wear level suggests that the animal was killed before reaching three years old.

If we consider the number of remains identified, Convento das Bernardas has the most significant sample, and a more diverse fauna spectrum. Domesticated species prevail over the wild ones as main providers of meat to human populations.

e) Horse (Equus sp.) There were identified a few horse remains in Convento das Bernardas and Convento da Graça.

Parque de Festas medival/modern occupation displays a bigger number of bovine and pig remains, and a reduction of sheep, goats and lagomorphs remains, especially rabbit when compared to the recorded values for the Islamic period. At the moment, it is not possible to determine how deep was the change regarding human diet behaviour; only more and bigger samples could clarify this aspect.

f) Pig/boar (Sus sp.) These two species were identified in all the archaeological sites, prevailing the anterior appendicular skeleton elements. Isolated dental elements are abundant. In Convento das Bernardas and Parque de Festas they represent almost half of the identified remains.

Conclusions

Adult animal prevailed both in Convento da Graça and Parque de Festas.

As a result of the archaeological interventions that have been done in Tavira, the knowledge about the diet of its inhabitants throughout the centuries has been enriched. It is possible to know that these populations had included in 93

their diets many different animals, like gallinaceous, whale, tuna, oyster, cockle, clams, conches, amongst others. This diversity may indicate that there was a good economic exploitation of the local natural resources.

Acknowledgments We would like to thank the elements of the Archaeology Team of Tavira’s Municipality, without them our work would have not been accomplished (Ana Sofia Vieira, Celso Candeias e Susana Gonçalves).

The analysed samples revealed a significant fauna spectrum, where the domesticated species prevailed over the wild one, in particular in meat supply. However, it is impossible to determine if some of those wild species including carnivores were included in the diet.

References Altuna, J. (1980) - Historia de la domesticacion en el Pais vasco. Munibe. 32: 1-2. Sociedad de Ciencias Aranzadi. San Sebastian. Barone, R. (1976) - Anatomie comparée des mammiféres domestiques. Tome Premier. Osteologie. Vigot Freres Editeurs. Bartosiewicz, L.; Neer, W. Van e Lentacker, A. (1993) Metapodial asymmetry in draft catle. International Journal of Osteoarchaeology. 3: 69-75. Boesseneck, J. (1969) - Osteological differences between sheep (Ovis aries, L.) and goat (Capra hircus, L.). In D. Brothwell, E. Higgs e G. Clark [eds.] – Science in Archaeology. London. Thames and Hudson. pp. 331-358. Carine-Tomé & Vigne, J-L. (2003) - Roe deer (Capreolus capreolus) age at death estimates: new methods and modern reference data for tooth eruption and wear, and for epiphyseal fusion. Archeofauna. 12: 157-173. Chaix, L. Méniel, P. (2001) - Archéozoologie. Les animaux et l´archéologie. Paris. Editions Errance. Covaneiro, J. & Cavaco, S. (2012) - Relatório Final dos Trabalhos Arqueológicos realizados na Bela Fria. Covaneiro, J. & Cavaco, S. (2010) - Relatório Final dos Trabalhos Arqueológicos realizados no Convento das Bernardas. Davis, S. (1989) - La Arqueología de los animales. Barcelona. Ediciones Bellaterra. Davis, S. & Payne, S. (1993) - A barrow full of cattle skulls. American Antiquity. 67 (254): 12-22. Davis, S. (1996) - Measurements of a group of adult female shetland sheep skeletons from a single flock: a baseline for zooarchaeologists. Journal of Archaeological Science. 2: 593-612. Desse, J.; Chaix, L. & Desse-Berset, N. (1986) - Ostéo. Base-Réseau de données osteómetriques pour l´Archéozoologie. Paris. CNRS. Gautier, A. (1987) - Thaphonomic groups: How and Why? Archaeozoologia.1(2): 47-52. Letow-Vorbeck, C. L (1998) - El Soto de Medinilla: faunas de mamiferos de la Edad del Hierro en el valle del Duero (Valadollid, España). Archeofauna. 7: 17-166. Lopes, G.; Covaneiro, J. and Cavaco, S. (2005) - Claustro do Convento da Graça. Análise dos materiais cerâmicos e faunísticos provenientes de dois contextos fechados. Xelb. 6(1): 313-326. Macieira, I.& Manteigas, R. (2008) - “A ribeira de Tavira: dízimas... cordas, mastros, remos... especiarias... carnes e versas.” In. Tavira, Patrimónios do Mar. Câmara Municipal de Tavira. Museu Municipal de Tavira. Palácio

In the age of death determination it was observed that the slaughter pattern of several domestic species (sheep, goat cattle and pigs) is maintained during a long course of time. One of the most represented species in the samples is the rabbit; however, its meat supply is relatively smaller comparing to other species. The elevated number of remains of rabbits leads us to consider the hypothesis that this species was raised in the city interior, during the Islamic period. Consequently, its presence in the archaeological record should not be related uniquely to hunting. Deer quantity is smaller in one of the sites from the medieval/modern period. Only more data could tell us if there is any relation with the reduction of wooded area around Tavira in more recent periods, or if this could be a result of a change of patterns in the populations’ diet. The analysis of taphonomic phenomena shows that there is a slight reduction of the bone surface alteration, as well as less activity of carnivores and rodents. The modifications produced by microorganisms associated with plant roots shows that the remains were used as resources for plants. This action is more clear in the samples obtained in Bela Fria and Convento das Bernardas, where the sediments are humus rich, as result of a continuous practise of agriculture. Fire modifications in the remains are minimal. In Convento da Graça and Convento das Bernardas there are no evidences that the remains were burn. In Bela Fria there were identified three burned remains and in Parque de Festas only one. The culinary preferences by these populations could have been casseroles, stews and soups. It is also possible that after eating the animals they were thrown in the domestic litter and not into fires. After the analysis of cut-marks it is possible to conclude that the faunistic remains under study are a result of alimentary usage. Thus, the sample can be compared to Group 1, defined by Gautier (1987) - the remains correspond to domestic waste, which result from human cottage, food preparation, and food leftovers.

94

da Galeria. pp.35-46. Pinto, M. & Ferreira, Â. (2008) - Intervenção arqueológica no Convento das Bernardas (Tavira). Relatório Final. Exemplar policopiado. Reitz, E. & Wing, E. S. (1999) - Zooarchaeology. Cambridge. University Press. Simão, I. and Pinto, M. (2009) - Sondagens Arqueológicas de Diagnóstico. Antigo Parque de Festas, Tavira. Relatório dos Trabalhos Arqueológicos. Stiner, M. C. (1994) - Honor among thieves. A zooarchaeological study of Neanderthal ecology. Princeton. University Press. Uerpmann, H-P. (1973) - Animal bone finds and economic archaeology: a critical study of “osteoarchaeological” method. World Archaeology. 4(3): 30722.

95

What did the Romans and Moslems eat in Conimbriga (Portugal)? The animal bones from the 1990’s excavations Cleia Detry1, João Luís Cardoso1,2 & Virgílio H. Correia3,4 1

UNIARQ - Centro de Arqueologia da Universidade de Lisboa. Universidade Aberta, Lisboa. 3 Museu Monográfico de Conimbriga, Condeixa-a-Velha. 4 Centro de Estudos Arqueológicos das Universidades de Coimbra e Porto. 2

Abstract The 1992 and 1993 excavations of Roman to Medieval layers of Conimbriga, in central Portugal, uncovered a Roman amphitheatre that pre-dated the Late Roman wall. Layers dated to the late Roman (3rd - 4th Cent. AD), late Antiquity (6th-7th Cent. AD) and Islamic (7th-11th Cent. AD) periods were also exposed. Almost 3000 animal bones were recovered. Because faunal assemblages are scarce in central Portuguese Roman to Medieval sites, particularly for the Moslem period, these bones presented a unique opportunity to compare Roman and Moslem dietary preferences, the non-food uses of animals, and livestock breeding practices. Mammal bones, particularly cattle, dominated the assemblage, followed by pig and sheep. The presence of wild boar within Moslem contexts suggests that the religious dietary rules were not strictly applied. The presence of wild boar as well as rabbit and red deer, further provide evidence for hunting. Keywords: Zooarchaeology; Roman; Islamic period; Faunal remains; Conimbriga. Adriaan De Man (2006) and the publication of radiocarbon dates for these and other samples from other parts of the site by De Man & Soares (2007).

Introduction Roman and Moslem contexts of Conimbriga (Condeixaa-Velha, Portugal) were excavated in 1992 and 1993 under the direction of Virgílio H. Correia. The excavated area, of three squares (4x4m), was aligned along a Late Roman city wall to extend over an area where the southern limits of the Roman amphitheatre were thought to be located. The amphitheatre was known to have been demolished during the Late Roman occupation to facilitate the construction of a city wall (Fig. 1).

Because our analysis was initiated so many years after the excavation, we relied heavily on Correia’s (1994b) excavation report for our interpretations of the stratigraphic and spatial distributions of the faunal remains. The presence of both Roman Late Antiquity and Medieval levels was significant for understanding the history of occupation of the site. Its abandonment had traditionally, and perhaps incorrectly, been assigned to

We (CD and JLC) recently studied the assemblage of c. 3000 animal bones from Correia’s excavations. Animal bones are rarely found in great abundance in sites from Roman to Medieval periods in central Portugal, particularly from Moslem contexts. Therefore the recovery of such a large faunal assemblage from Conimbriga presented an exceptional opportunity to study Roman and early Medieval animal uses in the more northerly regions of the Iberian peninsula. It also provided an opportunity to investigate the degree of influence that the Romans and Muslims had upon local dietary and husbandry practices, to confirm the presence of expected domesticated livestock such as sheep, goat, and pig and identify hunting of wild taxa such as red deer and rabbit. Archaeological context and chronology

Figure 1. In the superior left part of the map, the squares excavated at the roman amphitheatre beside the late roman wall of Conimbriga. Adapted from Reis, De Mann & Correia (2011).

The Conimbriga faunal assemblage was first described by Cardoso (1995) who also studied a number of individual taxa, e.g. the camelid (Cardoso 1992).The excavations themselves were first described by Correia (1994b), followed by the publication of an analysis of the finds by 97

Figure 2. Profile one to three of the excavation of the Conimbriga roman amphitheatre. From Correia (1994b). early post-barbarian invasions dates (cf. Alarcão 2004). Correia’s excavation was the first to consider post-Roman remains since the Luso-French excavations undertaken some 40 years ago.

times. 2 – Second occupation in the south area, following the construction of the Late Roman Wall. The calibrated radiocarbon dates (Sac 1995, 1996, 1999 and 2003) span 895 - 1160 cal AD (1σ) and 785 - 1210 cal AD (2σ).

Correia partially and superficially excavated three squares (4 by 4 m each) in 1992 (Layer 1 was not included in the stratigraphic profile in fig. 2). The stratigraphic layers of the Late Roman and Medieval Periods were recorded in arbitrary levels in 1993 as well as the stratigraphy of the south profile (see fig.2).

3 – Silo 1. Two 14C dates were obtained (Sac 1997 and 2000) which calibrated at 1σ are between 775 - 1020 cal AD, and at 2σ between 710 - 1025 cal AD. 4 and 5 – Silos 4 and 5were linked with the second occupation. One C14 date of 905 and 1160 cal AD. (2σ) was obtained (Sac 1998).

Following the removal of the superficial layer (Layer 1), Layer 2, was found to comprise of broken blocks of mortar and large stones in which cracks and spaces were filled with soil. This layer represents the period of destruction of the wall and land-clearance for agricultural purposes.

6 – This horizon overlays the amphitheatre demolition; it is the first occupation after the demolition of the amphitheatre and construction of the city wall in Late Antiquity.

Profiles one to three (fig.2), show the following stratigraphic horizons:

7 – Period in which the amphitheatre demolition and corresponding city wall construction took place, dating from the last years of the 3rd century to the first quarter

1 – Surface (soil) and other layers acquired in recent Roman period

Late antiquity

Islamic period

Omiads period

3rd-4th cent.

6th-7th cent.

7th-11th cent.

8th-9th cent.

Almoravids period 9th-11th cent.

N

%

N

%

N

%

N

%

N

%

Bos sp.

36

15%

4

2%

33

8%

41

9%

84

5%

Ovis/Capra

19

8%

10

6%

24

6%

17

4%

48

3%

Sus sp.

19

8%

4

2%

30

7%

21

4%

42

3%

C. elaphus

2

1%

1

1%

5

1%

9

2%

32

2%

C. capreolus

0

0%

0

0%

11

3%

0

0%

0

0%

Artiodactyla

4

2%

5

3%

5

1%

7

1%

18

1%

Equus sp.

1

0%

0

0%

0

0%

0

0%

2

0%

O. cuniculus

5

2%

7

4%

8

2%

31

7%

41

3%

Macrofauna

98

40%

30

17%

119

29%

113

24%

325

20%

Mesofauna

58

24%

17

10%

70

17%

64

14%

292

18%

Undetermined

0

0%

100

56%

100

25%

165

35%

741

46%

TOTAL

242

178

405

468

1625

Table 1. Number of indentified specimens, by taxonomic species and undetermined size groups from the Conimbriga Amphitheatre.

98

Figure 3. Comparison of the Number of Identified Specimens of the mammals in the several periods, since the the 3rd century to the 11th century. of the 4th century AD.

counted all bones. Those that could not be identified to species were classified into two size groups: macro (cow and horse size) and mesofauna (sheep, goat, pig size). In cases where it was not possible to classify bones by their size, they were classified as "undetermined". Most of the macrofaunal remains probably belonged to cattle, although a few may be of equids, since most of the identified macrofauna remains belonged to cattle (see table 1 and figure 3).

8 – Amphitheatre construction. Dated to the time of Nero (1st century AD). Based on Correia’s (1994b) excavation report, as well as the site profiles and previous analyses of the materials (De Man 2006) , and 14C dates De Man & Soares (2007).we grouped the occupations into 4 phases: Roman (3rd-4th cent.), Late Antiquity (6th-7th cent.), Omayyad (8th-9th cent.), Almoravid 9th-11th century), and included a larger Islamic period (8th-11th cent.) where it was not possible to refine the chronology further. The first and eighth stratigraphic horizons did not contain fauna.

The largest numbers of animal bones are from the Islamic period layers. The fact that animal bones are less frequent in Roman and Late Antiquity layers suggests a less intense occupation of this part of the site during Roman times. However, since faunal remains generally represent waste, it is likely that they were removed from time to time from the inhabited areas during the Roman period, practice that decreased with the decline of the Roman city in Late Antiquity and subsequent periods.

Results and Discussion The assemblage studied here comprises domesticated and wild animals commonly found in domestic environments of the periods under study: cattle, caprines, pigs, red deer, rabbits, a few carnivores, and birds. We observed and

The final period (9th-11th cent. AD) also has higher frequencies of animal bones, a pattern that may support 99

Figure 5. Graphic with measurements of the third inferior molar of Sus sp. Comparing elements from wild boar and domesticated form (pig).

Figure 4. Histograms showing the height of throclea of the distal humerus of Sus sp., from Conimbriga in the different periods and from Mesolithic Muge shellmiddens (Detry, 2007) to compare with wild boar measurements.

pattern might be explained by the small sample size or else prior human selection, for example, if the carcass was butchered elsewhere in the city and only the meatbearing parts taken to the site. There is no evidence for the consumption of animals younger than 3-4 years as few unfused bones were recovered. Animals may have been kept into maturity for their secondary products, e.g. milk and power for transport of goods and/or ploughing, and only subsequently killed for their meat. Indeed, some bones exhibit pathological conditions associated with traction.

the argument that the area was cleaned prior to the Moslem occupation. For all parts of the skeleton by period and species see tables 2 to 6. Bone recovery Overall the bones are well preserved. However, those of larger animals may be over-represented in relation to smaller animals, e.g. rabbit and birds, due to a recovery bias (no screening was carried out; the bones were handrecovered). Sixty-four to 84% of the bone assemblage was not identifiable to species. Macrofaunal remains were more abundant than Mesofauna, a pattern that is consistent with the list of species identified. The high number of unidentifiable remains due to bone fragmentation, which is normal for bone assemblages due to natural factors such as erosion and the effects of soil, weather and anthropic action over time.

b) Sheep/Goat (Ovis aries/Capra hircus) Caprines are the second most frequent group in most of the periods except Late Antiquity, in which they are the most frequent group. Distal parts of the limbs are absent in the Roman Period; likewise, distal parts of the limbs are absent in Late Antiquity levels as well as the bones of the head. These patterns support the argument that the animals were butchered elsewhere and that the faunal assemblage represents food refuse, i.e., carcasses were brought into the settlement already butchered and/or otherwise prepared. The age at death indicates the consumption of young animals, although in later contexts some animals were kept to an older age, presumably to exploit their milk and wool. However, the sample of fused bones is too small to draw firm conclusions.

Species present a) Cattle (Bos taurus) Cattle were the most common bones throughout the various periods of occupation, except in Late Antiquity. According to Castaños (1991) aurochs became extinct in Iberia during the Bronze Age. This is consistent with the zooarchaeological patterning in Portuguese sites, including Conimbriga, in which none of the Bos bones were of a size identifiable as aurochs, therefore all the Bos remains most probably belong to domestic cattle.

c) Suids (Sus scrofa) Suids are relatively abundant, particularly in the Islamic levels. The consumption of wild boar during the Islamic period is suggested by two humeri of wild boar size (see fig. 4) and a single lower third molar that is shaped like that of the wild species (fig. 5). This suggests that although the consumption of domestic pig is generally banned in Islam, wild boar consumption was sometimes permitted (Simmoons, 1961).

Bearing in mind a potential recovery bias, cattle appear to have been of great importance in the Conimbriga diet and local economy. In most contexts, all parts of the skeleton of cattle were present. The Late Antiquity levels are an exception; only the meat-bearing parts were present. This 100

Figure 6. Percentages of pig in several sites with Islamic occupation in Portugal. Data from: Paderne e Portela 3 Pereira, 2010; Mesas do Castelinho – Cardoso (1994); Silves – Davis et al., 2008; Álcácer do Sal – Valente, 1996; Lisboa – Moreno-Garcia & Gabriel, 2001; Torres Vedras – Gabriel, 2003; Alcáçova de Santarém – Davis, 2006; Conimbriga – this work). The presence of pig also suggests that the eating habits of local people were not significantly altered by the Moslem practices of their occupiers. It can be inferred that Moslem practices were not strictly imposed in this region, possibly because the Islamic occupation was both shorter and had less influence on the indigenous social and economic practices of central Portugal. Fig. 6 summarises the percentage of suids found in various other archaeological bone assemblages in Portugal. Suids are more frequent in the north, which corroborates the hypothesis that Islamic culture had less influence or was less strict in the higher latitudes of the Iberian Peninsula.

At some archaeological sites from the medieval period onwards, red deer is less frequent, probably due to overhunting and habitat destruction (Cardoso, 2002; Davis, 2006). Curiously, at Conimbriga the pattern is reversed: deer occur in higher frequencies in the Medieval Period, especially during the Almoravid occupation. Again, this pattern may indicate that the site was occupied by groups of high social or military status at this time. High-status hunting related objects are also known from the archaeological record of LateRoman/Early-Medieval times in Conimbriga (Correia 2003, 39). In fact, Cardoso’s (1994) faunal analysis of samples from the Almohad castle of Mesas do Castelinho (Almodovar, in southern Portugal) found he bones of red deer to be remarkably abundant, which was attributed to the hunting activity by military garrison. In the Almohad castle example, the high number of wild species was explained by the population being military rather than agriculturalist or pastoralist.

d) Red deer (Cervus elaphus) This species is consistently present throughout the sequence at Conimbriga, although in small numbers. The pattern suggests that big game hunting continued but was uncommon. Possibly hunting was associated with higher status (see Mackinon 1999-2000 regarding the link between hunting and status). Complex hunting methods, using horses and dogs, may have been used as such scenes are prominent in illustrations on late 3rd cent. mosaics found in Conimbriga (Oleiro 1992, 104-109).

Davis (2006) when analysing the remains from “Alcáçova de Santarém” proposed that deer populations began to decrease during the Roman period of Santarém, which can be explained by the gradual deforestation of Portugal, as indicated by the palaeobotanical evidence (Mateus 1992). In Conimbriga the reverse pattern is 101

g) Bear (Ursus arctos) Only one proximal phalanx of bear was found (see fig. 7). It exhibited cutmarks, a feature that usually indicates fur removal, particularly on this meatless part of the skeleton. Bear is commonly valued for its fur. Until the end of the middle ages, bear-hunting in Portugal was an activity reserved for the aristocracy and the king (Cardoso, 2001). Similar examples of bear hunting occurred in the Medieval Palmela castle (13th-14th cent.) (Cardoso & Fernandes, 2012) and Santarém (Davis, 2006).Various parts of bear skeleton were found in the Late Medieval levels of Leiria castle (Cardoso, 2001) and Roman levels of Monte Molião (Lagos, Algarve) (Detry & Arruda, 2013). h) Other Carnivores Domestic dog (Canis familiaris) is represented by a single metatarsal, recovered from the Islamic period. The dog is notably underrepresented given that it is commonly kept by humans. However, its absence from the contexts examined here is understandable because dogs are rarely used as food and the contexts examined here appear to contain primarily food refuse.

Figure 7. Proximal phalanx of a bear (Ursus arctos) found in the roman amphitheatre of Conimbriga. evident, red deer increases during the medieval occupation, although its proportion is always under 10%. In the Islamic period in Silves, in southern Portugal, red deer is also scarce (Davis, 2008; Antunes, 1997).

Otter (Luttra lutra) is represented by a single proximal ulna. Badger (Meles meles) is represented by two pelvis fragments, one proximal radius and one proximal femur. Both otter and badger were found in Silo 2 (Almoravid period). These remains are probably intrusive since both animals commonly burrow into archaeological sites.

e) Roe deer (Capreolus capreolus) At Conimbriga roe deer is represented by only 11 bone fragments which probably belong to a single individual. The fact that several parts of the skeleton (cranium, posterior and anterior members) are present might mean the animal was brought complete to the site, rather than having been butchered elsewhere.

i) Birds Bird bones are commonly underrepresented in archaeological assemblages because their bones are fragile and rarely survive deposition. In addition, their small size can influence their representation in the sample, particularly in cases, such as the present study, where the archaeological sediments were not sieved.

Roe deer is rare in Portugal today and now inhabits exclusively the mountainous areas of the north, most likely due to human predation. The comparatively low numbers of roe deer found in archaeological sites, compared to red deer, may be explained by the fact that it is more difficult to hunt, than red deer, due to its speed. Its density could also be rare at this time or take refuge in areas of difficult access, such as the mountain range near Conimbriga.

Six bird bones were identified – all are Galliformes (land fowl) and Anatidae (water fowl). One complete femur of red partridge (Alectoris rufa) was found in Islamic period contexts, confirming the hunting of small game during this period.

f) Rabbit (Oryctolagus cuniculus)

Three of the bones were identified as Gallus domesticus: one distal ulna at the Silo 1 (8th-9th cent.), one complete tibiotarsus from the same period, and one distal tibiotarsus from Late Antiquity. This species has been found at several Portuguese Roman sites including, among others, Castro Marim, Quinta das Longas and Alcáçova de Santarém. Anatidae, were represented by two distal humeri, one from the Islamic period and the other, probably duck, from the Roman layers.

Although rabbit is probably underrepresented, due to the lack of sieving during excavation, its presence indicates that small game hunting was a part of this economy. There was no signs of intrusions of this species such as burrows.

102

Another aspect is that until the beginning of the Middle Ages, the south of Portugal was densely forested providing the perfect environment for species like red deer and wild boar (Cardoso, 1994). Perhaps central Portugal was less densely forested and the larger towns, such as Conimbriga, depended to a greater extent upon domesticated livestock for their meat supply. Zooarchaeological studies of urban sites like Conimbriga are still all too scarce.

A musical instrument made with a griffon vulture (Gyps fulvus) ulna, probably Roman, with unknown provenance in Conimbriga (published by Moreno-García & Pimenta 2008) provided an example of the use of an animal for purposes other than food. The economy of Conimbriga compared to other sites In the majority of Portuguese Roman and Iron Age sites, the most common species of medium-large mammals represented are sheep and goats (4th-5th cent. AD Torre de Palma, Mackinon, 1999-2000; 2nd – 4th cent. Quinta do Marim, Antunes & Chauviré, 1992; 4th cent. BC Quinta das Longas, - Cardoso & Detry, 2005; Castro Marim 1st cent. BC - 1st cent. AD – Davis, 2007). The Conimbriga sheep and goat assemblages are small compared with these other sites, suggesting that beef was probably of higher importance in this economy (see Cardoso & Detry, 2005, for calculations on amount of meat).

Conclusions The faunal assemblage from Conimbriga is little different from those found in most Portuguese Roman and Medieval faunal assemblages. It is dominated by four main species of domesticated livestock – cattle, pig, sheep and goat (as reported by Cardoso 1995 report on Conimbriga). These animals undoubtedly provided both primary food products (meat, fat, skin, bones etc) as well as secondary products including as milk, wool, dung and in the case of cattle, muscle power.

In Conimbriga, wild animals represent approximately 30% of the number of identified specimens (NISP). Rabbit appears to be the main hunted species, assuming that rabbits were hunted and not tamed or kept in pens.

Cattle dominate the fauna throughout the occupation until Late Antiquity when smaller animals like the caprines and rabbit become more important.

Red deer and wild boar probably provided the people of Conimbriga with about 5% of their meat intake. This small percentage suggests low levels of big game hunting, an activity probably reserved for the privileged classes, as Mackinon (1999-2000) suggested at Torre de Palma.

The presence of pig in the Moslem period suggests that the Koranic prohibition on the consumption of this animal was not strictly enforced here. It is also quite likely that many of the inhabitants of Conimbriga were Christian. We concluded that Islam had less influence on the food and husbandry practises of the indigenous people of central Portugal than indigenous communities in the south of Portugal. We also concluded that big game hunting was only of minor importance at Conimbriga.

At Castanheira do Ribatejo, near Lisbon, there is also residual hunting of red deer and rabbits (Cardoso, 2009). In the other opposite Red deer constituted c. 28% of the bones found at the 1st to 4th c. Roman occupation of Pessegueiro (Cardoso, 1993), a pattern that is thought to be evidence of hunting by visiting sailors.

References Alarcão, J. (2004) – Conimbriga, 20 anos depois. In Correia, V. H. (ed) Perspectivas sobre Conimbriga. Conimbriga, 97-116. Antunes, M.T. (1997) – Arqueozoologia medieval em Silves. Setúbal Arqueológica. Setúbal. 11-12, p. 269-277. Antunes, M. T.; Mourer-Chauviré, C. (1992) - The Roman site (2nd to 5th centuries A.D.) at Quinta do Marim near Olhão (Algarve, Portugal): vertebrate faunas. Setúbal Arqueológica. Setúbal. 9/10, p. 375-382. Cardoso, J. L. (1992) - Um camelídeo de Conimbriga. Conimbriga. Coimbra. 31, p. 181-187. Cardoso, J. L. (1993) – Contribuição para o conhecimento dos grandes mamíferos do Plistocénico Superior de Portugal. Oeiras: Câmara Municipal de Oeiras. Cardoso, J. L. (1994) – A fauna de mamíferos da época muçulmana das Mesas do Castelinho (Almodôvar). Materiais das campanhas de 1989--1992. Arqueologia Medieval. Porto. 3, p. 201-220. Cardoso, J. L. (1995) – Os mamíferos no quotidiano romano. Algumas reflexões a propósito dos restos de

At Mesas do Castelinho (southern Portugal) which was adjacent to dense forests containing abundant game, big game is also very abundant in the remains recovered (Cardoso, 1994). Castro Marim (Davis, 2007) and Monte Molião (Detry & Arruda, 2013) red deer bones are less common than those of domesticated species, but very close to Bos taurus in numbers which appears to represent an intermediate situation. Quinta das Longas also has abundant red deer and rabbit, although most (63%) of the faunal remains are domesticated species (Cardoso & Detry, 2005). The faunal patterning suggests that, at Roman sites located in the more rural areas (e.g. Mesas do Castelinho), people depended more on hunting and less on pastoralism. And in more urban sites such as Conimbriga and Castanheira do Ribatejo, domesticated species were the main source of meat and secondary products.

103

Conimbriga. Estudos Arqueológicos de Oeiras. Oeiras. 5, p. 299-313. Cardoso, J. L. (2001) - Sobre a presença do urso em Portugal, a propósito de uma peça do castelo de Leiria. Torre de Menagem do castelo de Leiria. Leiria. Câmara Municipal de Leiria: 40- 55. Cardoso, J. L. (2002) - Arqueofaunas: balanço da sua investigação em Portugal. Arqueologia e História. Lisboa. 54, p. 281-298. Cardoso, J.L. (2009) – Estudo Arqueozoológico sumário dos restos recuperados nas escavações. In A Villa Romana da sub-serra de Castanheira do Ribatejo (Vila Franca de Xira). Trabalhos Arqueológicos efectuados no âmbito de uma obra da EPAL:Vila Franca de Xira: EPAL. Cardoso, J. L.; Detry, C. (2005) – A lixeira baixoimperial da villa da Quinta das Longas (Elvas): análise arqueozoológica e significado económico-social. Revista Portuguesa de Arqueologia. Lisboa. 8:1, p. 369-386. Cardoso, J. L. & Fernandes, I. (2012) – A economia alimentar dos muçulmanos e cristãos do castelo de Palmela: um contributo. Arqueologia Medieval, 12: 211233. Castaños, P. (1991) - Animales domésticos y salvajes en Extremadura. Origen y evolución. Revista de Estudios Extremeños.Badajoz. 47, p. 9-67. Correia, V. H., (1994) Relatório das escavações arqueológicas no anfiteatro de Conimbriga em 1993 (Conimbriga, Arquivo do MMC, policopiado). Correia, V. H. (2003) – A deer antler medallion from Conimbriga (Portugal). Instrumentum, 17 , p. 39. Davis, S. (2006) – Faunal remains from Alcáçova de Santarém (Portugal). Lisboa: IPA. Davis, S. J. M. (2007) – The mammals and birds from the Iron Age and Roman periods of Castro Marim, Algarve, Portugal.Lisboa: IPA (Trabalhos do CIPA.107). Davis, S.J.M. (2008) – Zooarchaeological evidence for Moslem and Christian improvements of sheep and cattle in Portugal. Journal of Archaeological Science. London. 35: 4, p. 991-1010. De Man, A., 2006: Conimbriga, do Baixo-Império à Idade Média (Lisboa, Ed. Sílabo). De Man, A. e Soares, A. M. M., 2007: “A datação pelo radiocarbono de contextos pós-romanos de Conimbriga”. Revista Portuguesa de Arqueologia 10-2, 285-294. Detry, C. & Arruda, A. M. (2012) – A fauna da Idade do Ferro e Época romana de Monte Molião (Lagos, Algarve): continuidades e rupturas na dieta alimentar. Revista Portuguesa de Arqueologia, 15: 215-227. Gabriel, S. (2003) - Estudo dos restos faunísticos do silo 1 dos Paços do concelho de Torres Vedras. Lisboa: Instituto Português de Arqueologia (Trabalhos do CIPA: 48) Mackinnon, M. (1999/2000) - O papel dos animais na economia rural da Lusitânia romana: zooarqueologia de Torre de Palma. A Cidade. Revista Cultural de Portalegre. Lisboa. 13/14, p. 129-140. Mateus, J. (1992) – Holocene and present-day ecosystems of the Carvalhal region, southwest Portugal. Utrecht: Rijksuniversiteit. PhD thesis.

Moreno-García, M. & Gabriel, S. (2001) – Faunal remains from 3 islamic contexts at Núcleo Arqueológico da Rua dos Correeiros, Lisbon. Trabalhos do CIPA, 20: 30pp. Moreno-García, M. & Pimenta, C. (2008) – Arqueozoologia cultural: O aerofone de Conímbriga. Revista Portuguesa de Arqueologia. Lisboa. 7:2, p. 407425. Oleiro, J. M. B. (1992) : Conimbriga. Casa dos repuxos (Lisboa, IPM, Corpus dos Mosaicos Romanos de Portugal I). Reis, Maria Pilar; De Man, Adriaan e Correia, Virgílio Hipólito (2011) - "Conimbriga". In Remolá Vallverdú, Josep Anton e Acero Pérez, Jesús (eds.) La Gestion de los residuos urbanos en Hispania (Mérida, Instituto de Arqueologia/CSIC, Anejos de Archivo Español de Arqueología, LX), 181-202. Simmoons, F. J. (1961) - Eat not this flesh: Food avoidances in the Old World. Madison: University of Wisconsin Press.

104

Bos sp.

Cranium Cranium Maxillary Mandible Teeth Axial Atlas Axis Ribs Vertebrae Anterior Scapula Humerus Radius Ulna Metacarpal Posterior Pelvis Femur Patela Tibia Os maleolus Calcaneum Astragalus Metatarsal Metapodial Phalanx I Phalanx II Phalanx III Long Bones Undetermined TOTAL

Ovis/Capr a

Sus sp.

C. elap.

Artiod.

Equid

O. cunicu.

2 1 1 2

Meso

10

2

Und.

1 1 5

1 1

2

Macro

8 3 1

4 1

2 1

4 3 1 3

2 2

2

1

12 20

10 8

1

1

1

5 1 6 1 4 1

1

3 1

4

1

1

1 3 2

1

3

3

36

19

19

2

4

1

5

23 20 98

35 2 58

Table 2. Number of indentified specimens by parts of the skeleton in the Roman period -3rd-4th century from the Conimbriga Amphitheatre.

105

0

Bos sp.

Ovis/Capra

Sus sp.

C. elap.

Artiod.

O. cunicu.

Macro

Cranium Cranium

Meso

Und.

1

Maxillary Mandible

1

3

4

13

Teeth Axial Atlas

1

Axis Ribs Vertebrae

1

6

2

3

Sacrum Anterior Scapula Humerus

2

1

1

Radius

2

1

1

Ulna

1

1

1

Metacarpal

1

Posterior Pelvis

1

Femur

1

1

3

Patela Tibia

1

1

Fibula

1

Os maleolus Calcaneum

1

1

Astragalus Metatarsal

1

2

Metapodial Phalanx I

1

Phalanx II Phalanx III Long Bones

11

14

Undetermined

5

86

TOTAL

4

10

4

1

5

7

30

17

100

Table 3. Number of indentified specimens by parts of the skeleton in the Late antiquity - 6th-7th century from the Conimbriga Amphitheatre.

106

Cranium Cranium

Bos sp.

Ovis/Capr a

1

2

Maxillary

Sus sp.

C. elap.

Artiod.

O. cunicu.

2

Mandible

2

Teeth

6

9

6

Macro

Meso

9

6

Und.

1 9

5

19

14

27

15

2

Axial Atlas

2

Axis

1

1 1

Ribs Vertebrae

2

Sacrum

1

Anterior Scapula

1

Humerus

3

Radius

1

1

1

Ulna

1

1

4

Scaphoid

2

Metacarpal

1

Posterior Pelvis

1

2

3 1

1

8

2

1

1

2

1

4

2

5

2

Long Bones

15

20

Undetermin ed TOTAL

26

1

165

113

64

165

Femur

4

1

Tibia

1

1

Calcaneum

3

Astragalus

4

Metatarsal

2

1

2

Phalanx I

4

1

1

Phalanx II

1

Phalanx III

1

Metapodial

4 1

1 1

41

1

7

1

2 1

17

21

9

7

31

Table 4. Number of indentified specimens by parts of the skeleton in the Omiads period – 8th – 9th century from the Conimbriga Amphitheatre.

107

Cranium Cranium

Bos sp.

Ovis/Cap ra

1

4

Maxillary

Sus sp.

C. elap.

Artiod.

Equid

O. cunicu.

Macro

Meso

Und.

16

7

23

1

1

1

5

18

13

55

74

54

39

Mandible

4

2

4

Teeth

11

14

15

Axial Atlas

2

1

Axis

2

1

3 1 1

Ribs Vertebrae

2

Sacrum

5

Anterior Scapula

1

2

2

1

Humerus

3

2

1

1

2

Radius

6

2

1

1

2

Ulna

1

Scaphoid

2

Piramidal

1

Metacarpal

5

Posterior Pelvis

4

Femur

4

2

3

4

8

5

Long Bones

97

103

Undetermine d TOTAL

55

12

539

307

262

562

1

1

4

1

1

2

3

1

4 3

9 2

Patela

5

Tibia

2

Calcaneum

7

Astragalus

5

Metatarsal

2

3 2

6

Phalanx II

5

Phalanx III

3

77

3

3

4 1

1

Metapodial Phalanx I

1

2

1

2 2

4

1

4

4

4

1

2

39

42

24

13

2

36

Table 5. Number of indentified specimens by parts of the skeleton in the Almoravids period 9th-11th century from the Conimbriga Amphitheatre.

108

Bos sp.

Cranium Cranium

Ovis/Ca pra

Sus sp.

1

1

Maxillary

C. elap.

C. cap.

Artiod.

Equid

O. cunicu.

2

1

Mandible

2

Teeth

9

8

11

3

6

2

Macro

Meso

Und.

7

6

17

1

1

1

7

11

Ribs

43

37

Vertebrae

40

17

Axial Atlas Axis

3

Sacrum

2

Anterior Scapula

4

8

8

1

2

5

2

Humerus

2

3

5

2

1

4

1

1

Radius

4

5

7

3

2

1

2

1

Ulna

1

1

2

Metacarpal

5

3

3

Pelvis

4

4

4

Femur

1

2

3

1

1

1

2

Posterior 1

5

5

4

4

3

1

Long Bones

36

54

Undetermin ed TOTAL

70

1

241

221

137

258

1

1

2

Patela

1

Tibia

3

Calcaneum

5

1

4

Astragalus

3

1

2

Cuneiforme

1

Metatarsal

4

5

1

1

2

1

Phalanx I

4

2

1

1

Phalanx II

4

1

Phalanx III

2

1

Metapodial

61

54

58

1

14

6

1

1

1

1

1

3

1 2 3

11

11

16

Table 6. Number of indentified specimens by parts of the skeleton in the Islamic period 7th-11th century from the Conimbriga Amphitheatre.

109

Firemarks

Cutmarks

Chopped

Rodents

Carnivores

Bos

0

11

1

4

1

Ovis/Capra

0

2

0

1

0

Sus sp.

0

5

0

2

1

C. elaphus

0

0

0

0

0

Equus sp.

0

0

0

1

0

O. cuniculus

0

0

0

0

0

Artiodactyla

0

0

0

1

0

Macrofauna

0

3

1

2

1

Mesofauna

0

2

0

2

0

Undetermined

1

0

0

0

0

Bos

0

2

0

1

1

Ovis/Capra

0

1

0

0

0

Sus sp.

0

1

0

0

0

C. elaphus

0

0

0

0

0

Bos

0

11

1

4

1

Ovis/Capra

0

2

0

1

0

Sus sp.

0

5

0

2

1

C. elaphus

0

0

0

0

0

Equus sp.

0

0

0

1

0

O. cuniculus

0

0

0

0

0

Artiodactyla

0

0

0

1

0

Macrofauna

0

3

1

2

1

Mesofauna

0

2

0

2

0

Undetermined

1

0

0

0

0

Bos

0

2

0

1

1

Ovis/Capra

0

1

0

0

0

Sus sp.

0

1

0

0

0

C. elaphus

0

0

0

0

0

3rd /4th cent. – Roman period

6th/7th cent. - Late Antiquity

3rd /4th cent. – Roman period

6th/7th cent. - Late Antiquity

Table 7. Taphonomic information of anthropic and animal origin in the elements observed for this study.

110

Zooarchaeological perspective of the Islamic sites in Algarve Current State of Knowledge Vera Pereira1 1

PhD Student, Universidade de Coimbra, CEAUCP; [email protected]

Abstract With still a long way to go, comparing with other studies and analysis of archaeological sites, a lot has changed in the last decade in Portugal. Zooarchaeology is now seen as an essential tool to understand human behavior and its relation with animals. We aim the summary of work done and its methodology, considering taxonomic, taphonomic and osteological approaches to animal remains, focusing the Algarve and the Islamic period as the perfect paradigm. Keywords: Zooarchaeology; Islamic; Algarve. country we now know as Portugal suffered many phases, through advances and declines, with the Christians gaining territory of the Muslims from the north to the south. As a result, the region of Algarve experienced Muslim influence longer, ending in 1248/49 with the military triumphs of D. Afonso III and D. Paio Peres Correia, which reflects archaeologically with an enormous amount of sites within this time frame.

Introduction Zooarchaeology in Portugal presents itself still as a shy producer of data and useful knowledge. Although, the study of faunal assemblages recovered from archaeological contexts increased in the last decade, when compared with other approaches of archaeological sites and its goods, as pottery or lithics, faunal studies still have a long way to go, as a useful aid on reconstructing the past and ancient populations.

At the moment, the “Endovélico”1 data base encloses 352 archaeological sites or goods of the Islamic period in the Algarve, marked with white dots in Figure 1 (bearing in mind that a higher number comes out every day due to new finds).

It seems very clear that we can distinguish two phases regarding zooarchaeological approaches, within the Portuguese territory: - In the 1990s and early 2000s the first studies of fauna emerge, mainly centered on mammal and/or molluscs remains and its taxonomic identification as the central goal, with constricted or non existing taphonomic approaches, as well as limited osteometric concerns and non-existing studies of other classes of animals, accomplishing straightforward reviews based on social, economic and religious points of view.

If we acknowledge that of these 352 entries, only 13 faunal assemblages were studied, regarding 12 archaeological sites, a lot is yet to be done concerning animal remains. Nonetheless, it is a good framework, with new studies carried out in the past few years, leading to a better view of the Algarve, its past and inhabitants during the Muslim period.

- With the establishment of the Archaeosciences laboratory of DGPC (Direção-Geral do Património Cultural) and the creation of the osteological reference collection, in 2000, a wider range of species and animals classes are examined. It develops a greater concern with taphonomic matters: human, animal and/or geochemical modifications on bones and its significance. Osteometry is perceived as a method to distinguish close species, as well as domestic ratios, sexual dimorphism, breed enhancements and others. New approaches on remains and its functional and symbolic uses emerge.

Zooarchaeological studies The faunal assemblages were obtained from the excavations of 12 Muslim sites, which included mainly rural settlements and key fortifications, from the 8th to the 13th centuries, as seen in Table 1. The animal remains are mostly from rubbish pits and silos from housing areas suggesting that, probably, derive from leftovers of meals that were thrown away to these disposal areas. Moreover, a general outline of the Muslim region of the Algarve can be made as 9 of the 16 municipalities are characterized here with, at least, one site.

Muslim Algarve Islamic authority in Iberia and the “Reconquista” (from 711 with the Muslim invasion trough the Strait of Gibraltar, until 1492 with the end of the Nasrids of Granada) are notoriously recognized as a crucial part of Spanish and Portuguese history. The establishment of the

1

Public data base of DGPC (Direção-Geral do Património Cultural) with all the archaeological sites or archaeological finds, of all chronologies, identified in Portugal.

111

Figure 1. Identified Muslim sites of the Algarve - Portugal.

Archaeological Site

Category

Context

Chronology

Albufeira

_

Silo 1

13th cent.

Alcaria d'Arge

Rural settlement

House 1/Silos

12-13th cent.

Alcariais de Odeleite

Rural settlement

Residential space

10-13th cent.

Aljezur - Barranco da Alcaria

Rural settlement

Silo

8-12th cent.

Castelo das Relíquias

Rural settlement

_

9-10th cent.

Castelo Paderne

Fortification

Rubbish pit

12-13th cent.

Castelo Velho de Alcoutim

Fortification

Fireplace

9-12th cent.

Portela 3

Rural settlement

Rubbish pit

9-13th cent.

Ribat da Arrifana

Ribat (small fortification)

_

12th cent.

Salir

Fortification

Silo 8

12-13th cent.

Silves - Arrabalde

Eastern suburbs

Rubbish pit

12-13th cent.

Silves - Castelo

Fortification

Residential space

8-12th cent.

Tavira - Parque das Festas

_

Silo/Rubbish pit

Islamic

Table 1. Faunal assemblages. compare results, it appears very clear that the methodologies used are quite varied, bearing in mind that different researchers applied them with the intention of answering their individual queries.

Results and discussion After all the data is gathered trough the available bibliography, with the mission of summarize and 112

Archaeological Site Albufeira

Counts

Mammals

Birds

Fishes

Molluscs

Amphibians

Reptiles

Total

References Callapez, P. 2001

NISP

1250

1250

%

100%

100%

NISP

1381

176

48

%

84,88%

10,82%

2,95%

NISP

151

6

9

166

%

91%

3,6%

5,4%

100%

Aljezur Barranco da Alcaria

NISP

*

Castelo das Relíquias

NISP

210

58

268

%

78,2%

21,8%

100%

NISP

86

73

159

%

53,8%

46,2%

100%

NISP

77,5

12

4

93,5

%

82,8%

12,9%

4,3%

100%

NISP

785

101

30

%

1

11%

3,3%

NISP

788

12

18

%

96,3%

1,5%

2,2%

NISP

397

79

205

959

22

1662

%

23,86%

4,75%

12,33%

57,69%

1,32%

100%

NISP

264

10

2

276

%

95,7%

3,6%

0,1%

99,4%

Silves Arrabalde oriental

NISP

2982

152

88

3

3225

%

92,5%

4,7%

2,7%

0,1%

100%

Tavira Parque das Festas

NISP

629

629

%

100%

100%

Alcaria d'Arge

Alcariais de Odeleite

Castelo Velho de Alcoutim Castelo de Paderne Portela 3

Ribat da Arrifana Salir

Silves Castelo

*

21

1

1627

1,29%

0,06%

100%

*

MorenoGarcía, M. et al. 2008 Pereira, V. 2012 Silvério, S. 2001

%

*

6

922

0,7%

100%

*

818 100%

Catarino, H. 1997/1998 Catarino, H. 1997/1998 Pereira, V. 2009/2010 Pereira, V. 2012 Antunes, M. 2007; Callapez, P. 2007 Martins, S. 2012 Antunes, M. 1991 Davis, S. et al. 2008 Covaneiro, J. & Cavaco, S. 2012

Table 2. Number of recorded bones and percentages, by animal Classes and Phylum. * Present at the site but without available data. Adding an extra intricacy to this task, the majority of the papers have no taphonomic or osteometric data, among others, giving mainly a one sided view of the assemblages. On the other hand, it appears that some of the animal samples were separated by Class, and only part of it was studied (e.g. only the mammals or only the molluscs were considered).

still have a general glimpse of other animals present among the samples. It appears that birds and fishes are an important complement to these inhabitants diet, in contrast with the scarce fragments of amphibians and reptiles (which can also be intrusions). It seems essential to question why mammals stand out. On one hand, it is well known that people favour these animals for their meat input as well as their labour force, making them an indispensable resource. But, on the other hand, some recovery issues, where only the good, big and well preserved bones are recovered, without sieving the soil, or even preservation problems due to the Ph of the

Mammals clearly stand out on the majority of the assemblages, comprising large portions of the total remains, when compared with other classes of animals, as seen in Figure 2 and Table 2. Nevertheless, we can 113

Site

Dog

Bear

Cat

Horse

Pig

Deer

Cow

Alcaria d'Arge Alcariais de Odeleite Castelo das Relíquias C. V. de Alcoutim Paderne

135

50

18

1

1

1

1

6

8

21

Portela 3

21

1

39

3

2

17

Ribat da Arrifana Silves

1

2

Silves Arrabald e Tavira P. F. Salir

26

2 1

Hare

340

Sheep/ Goat 174

11

38

19

1

1

15

169

23

210

1

21

46

17

86

6

1

49 1

57

12 51

1

Rabbit

Others

Total

632

32

1381

79

2

151

12

1

21

6

78

175

394

9

93

32

786

260

360

2

30

133

788

5

239

14

2

37

515

2038

21

232

13

8

13

163

13

405

3

18

2

263

58

264 4

2981 629

1

397

Table 3. Number of Identified Specimens (NISP) – mammal bones. soil, may be the causes of other remains absence.

lower than other animals, it does not seem surprising that such high numbers show up, as more individuals would be needed to add up the same amount of meat supply. Unfortunately, it is extremely difficult to distinguish if these animals were domestic or wild and no additional information about it was given in the bibliography.

In any case, the majority of records appear to focus on mammal remains, with the concern of classification to species level. Furthermore, birds and molluscs families or sometimes species are often identified; leaving fishes, amphibians and reptiles only identified considering the Class they belong to, under sizing the available data. With these concerns in mind, it seems reasonable to scrutinize mammals a little further.

Bovine cattle captures third place on the charts as a great meat supply, milk and working force, especially on the rural settlements, a particular feature that will be discussed later on.

If we look carefully at Table 3, it is very clear that, although the assemblages differ a lot in size and Number of Identified Specimens (NISP), some mammals unquestionably stand out.

Considering other mammals, it appears that hunting was used as a helpful resource to support provisions but with just a few species preferred, as deer and hare as the most frequent ones. Cervus elaphus (red deer) looks certainly appreciated since it is present in all the studied sites; with the exception of the interior Castle of Silves, it appears in the suburbs of the same city (Silves – Arrabalde).

Sheep and goats (grouped together by reason of its skeletal similarity and consequent complexity to separate from one another) take the lead with really high percentages, as 90,5% in Silves – Castelo (# 239) or 80,5% in Castelo das Relíquias – Alcoutim (# 169), showing a real preference by the Muslims on the consumption of these animals, a statement still accurate nowadays within the Islam followers.

An additional peculiarity acknowledged here is that all of the three bones matching Ursus arctos (brown bear), from three dissimilar archaeological sites, are metapodials. When skinning bears with the intention of using or selling its fur, it is quite common to cut the anterior and posterior limbs and leave them attached to it. For now, we can assume that brown bear was present nearby and, that at least, its fur was being used. Its presence in these locations is not necessary due to consumption patterns but it is equally possible.

Rabbits are also very numerous, with a lot of bones/fragments in high percentages in the samples studied. If we take into account that its meat supply is

114

100 90 80 70 60 50 40 30 20 10 0

Defensive Residential space

Figure 2. NISP % - Defensive vs. Residential contexts. Another interesting viewpoint to examine these remains is from the religious perspective. The Quran forbids dogs as companions as they are perceived as an impure animal, as well as swines and its meat consumption. Were these populations really converted to Islam? Were these individuals following the Quran as they would if they were closer to the core of the empire?

preference for felines as companions, probably due to its additional skill to catch small animals. If we look at the same data, with the NISP percentages, but considering its residential or defensive role – Figure 2, would it give us any distinction in animal consumptions or species preferences? Although more numbers and study cases would be needed to answer this question, some ideas can be brought into light.

Although the Quran forbids all swines, it is somehow common to recognize wild boar as less harmful and eaten it in times of need. It is not unusual to find some swine remains in archaeological excavations of the Muslim period, but it is rather difficult to distinguish if those remnants are from wild or domestic species. Nevertheless, insignificant samples of Sus bones were identified in the majority of the sites, with the exception of Alcaria d’Arge with 50 NISP, and in four of the assemblages no pig bones were found. Since these remains are so scarce, there are some doubts whether it is food supply of Muslim populations or of Christian neighbours, if Muslims experienced hunger and turned to pork consumption or if they simple did not follow Quran as they should.

It appears that dogs are more abundant in residential areas, agreeing with the thought that they could wander around for scraps of food. Also bovines show better consistence in residential sites, were higher number of inhabitants could have their household to sustain and larger areas where the big cattle could graze. On the other hand, it seems that caprines and lagomorphs are quite equally found on both context types. Conclusions

Furthermore, in Islam, domestic dogs are also seen as polluted animals and dismissed as companions, in contrast with cats, the favoured animals. Again, it is very difficult to know if the domestic dog remains found among the samples were from Muslim or Christian owners, or just around the area to eat the scraps and leftovers it could find. Anyway, the general numbers indicate that more remains of cats than dogs were found in the archaeological sites. Apparently, it shows a slight

Faunal studies are still insufficient and shy producers of data in Portugal. Although hundreds of Islamic archaeological sites are identified in the Algarve, only 13 bone assemblages were studied, recovered from the excavations of 12 sites. It is possible to recognize a common absence or insufficiency of taphonomic, osteometric, statistic or other approaches. The lack of general methodological 115

guidelines shows different approaches of the archaeological sites and, in particular, the faunal assemblages. In addition, the majority of the assemblages are not entirely studied, leaving the data undersized and with incomplete knowledge of these populations and their lifestyle. All of these issues contribute to extremely difficult comparisons between fauna and its assemblages.

Cardoso, J. L., in Catarino, H. (1997/1998) - O Algarve Oriental durante a ocupação islâmica: Povoamento rural e recintos fortificados. In Al-Ulya Nº 6 – Revista do Arquivo Histórico Municipal de Loulé, pp. 745-747. Covaneiro, J.; Cavaco, S. (2012) – Comer em Tavira. Análise dos restos faunísticos do sítio do Parque de Festas (Tavira). In JIA 2011, Actas das IV Jornadas de Jovens em Investigação Arqueológica, Gambelas: Universidade do Algarve, pp. 269-276. Davis, S.; Gonçalves, M. ; Gabriel, S. (2008) - Animal remains from a Moslem period (12th/13th century AD) lixeira (garbage dump) in Silves, Algarve, Portugal. Revista Portuguesa de Arqueologia, volume 11, número 1, pp. 183-258. Martins, S. (2012) Estudo faunístico do Silo 8 do Castelo de Salir (Loulé). Contribuição para o conhecimento da dieta alimentar islâmica. In JIA 2011, Actas das IV Jornadas de Jovens em Investigação Arqueológica, Gambelas: Universidade do Algarve, pp. 277-282. Moreno-García, M.; Pimenta, C.; Roselló, E.; Morales, A.; Gonçalves, D. (2008) Um retrato faunístico dos vertebrados de Alcaria de Arge (Portimão). Xelb 8 – Actas do 5º Encontro de Arqueologia do Algarve, Vol. I Comunicações e Conferências, pp. 275-306. Pereira, V. (2009/2010) Comunidades islâmicas e medievais-cristãs do Castelo de Paderne: continuidade e mudança. Perspectiva zooarqueológica. In Promontória 7/8 - Revista do Departamento de História, Arqueologia e Património da Universidade do Algarve, pp. 177-190. Pereira, V. (2011) Estudo zooarqueológico de comunidades islâmicas do Algarve. Dissertação de Mestrado em Teoria e Métodos da Arqueologia, não publicada, Faro: Universidade do Algarve. Pereira, V. (2012) Alcariais de Odeleite – Perspectiva zooarqueológica. In Actas do V Encontro de Arqueologia do Sudoeste Peninsular. Almodôvar, pp. 821-830. Silvério, S. (2001) Silos Islâmicos de Alcaria – Aljezur (Séculos VIII-XII), Aljezur: Assoc. Defesa do Património Histórico e Arqueológico de Aljezur, Câmara Municipal de Aljezur, pp. 49-50.

All these assumptions could change when other counting’s as Minimum Number of Individuals (MNI) or Minimum Number of Elements (MNE) are made. Unfortunately there were not enough data on the published papers that could assist this task and we can only perceive this article as a first glance on these assemblages. A lot more can be done when other perspectives approach the same animal bones. In spite of this, the assemblages studied are generally from disposal areas, such as rubbish pits and silos, were the leftovers of meals would be set out. Mammals come out as prime resource, specifically ovine, caprine and bovine cattle as domestic livestock, as well as leporids, whether they are domestic or hunted. The recollection of molluscs, fishing and hunting or breeding of birds complemented the food supplies of these households. Although difficult to distinguish, it appears that in residential areas there are higher frequencies of bovine cattle, as well as dogs. On the other hand, for sheep, goats and rabbits no significant change was identified. Finally, overall, these populations appear to follow the Quran and its austere rules, as low frequencies of dogs and pigs were recognized. Acknowledgements Research supported by FCT PhD fellowship SFRH/BD/77256/2011, with QREN/POPH, European Social Fund and National MEC funds. References Antunes, M. (1991) - Restos de animais no Castelo de Silves (séculos VIII-X) Contribuição para o conhecimento da alimentação em contexto islâmico. In Estudos Orientais 2 – O legado cultural de judeus e mouros, pp. 41-74. Antunes, M. (2007) - Ribāt da Arrifana – Estudo arqueozoológico. In Ribāt da Arrifana – Cultura material e espiritualidade, Aljezur, pp. 83-86. Callapez, P. (2001) - Invertebrados do arqueossítio Silo 1 (Albufeira): aspectos da malacofauna e do consumo de moluscos no Algarve muçulmano, Albufeira: Câmara Municipal. Callapez, P. (2007) - Fauna malacológica do ribāt da Arrifana – Análise preliminar. In Ribāt da Arrifana – Cultura material e espiritualidade, Aljezur, pp. 87-90. 116

At table with the nuns: the mammals of 17th century Santa-Clara-a-Velha Monastery (Coimbra, Portugal) Cleia Detry1, Lígia Inês Gambini2 & Artur Corte-Real2 1 2

UNIARQ - Centro de Arqueologia da Universidade de Lisboa, Portugal Mosteiro de Santa-Clara-a-Velha, Coimbra, Portugal

Abstract The Santa Clara-a-Velha monastery is located on the left bank of the river Mondego, opposite the city of Coimbra. Founded in 1286 by D. Mor Dias, it was occupied by the nuns for almost 400 years until its abandonment in 1677. In 1995 a major refurbishment and archaeology project began for the purpose of restoring the monastery and excavating the church interior and cloisters. In this paper we present the results of a faunal analysis, focusing on mammal bones that were recovered during the excavation of rubbish dumps that date from the first half of the 17th century, the final period that the nuns occupied Santa Clara-a-Velha. Faunal remains are scarce for the modern period thus the assemblage from Santa Clara-a-Velha provides unique new information about the diets of people from those times. Mammals, mostly sheep, pork, cow and rabbit, dominate the bone assemblage. Their patterns of age-at-death indicate the slaughter and consumption of young animals and even piglets. The parts of the skeleton found in the middens indicate that only the more meaty joints were brought into the monastery. These results suggest that the residents were people of high-status, who enjoyed a privileged lifestyle that is clearly distinct from that of the rest of the population, particularly the large amount of meat and the high quality of secondary animal products, which suggest the consumption of large quantities of sweets. Keywords: Santa Clara-a-Velha Monastery; Modern Period; Mammal; Zooarchaeology. seasons. Area 41, the first survey was conducted in 1991 by IPPAR (Corte-Real et al. 2000); Area 46 was excavated by CIPA (Archaeociences Laboratory - from the former Portuguese Institute of Archaeology - IPA) in 2006; and Room B was excavated by the Archaeology company - GAAIA - in 2008 (Fig. 2).

Introduction The Santa Clara-a-Velha Monastery is located on the left bank of the Mondego River, opposite to the city of Coimbra (Fig. 1). Founded in 1283, the monastery has great symbolic meaning for the residents of Coimbra because - Queen Isabel - was herself one of the founders of Santa Clara- and housed her tomb till the evacuation of the site. Built in relatively isolated place and on shallow ground the buildings of the monastery were badly affected by flooding from the rising waters of the Mondego River. Flooding was an ongoing problem until finally, in the 17th century it forced the nuns to relocate. The restoration of the monastery and its transformation into a museum began in 1995 under the direction of the Portuguese Institute of Architectural Heritage (IPPAR), now owned by the Regional Directorate of Culture (DRCC), and was coordinated by the archaeologist Artur Corte-Real. Extensive and detailed excavations were conducted inside the church and cloister where several rubbish dumps were identified. The faunal remains that are the subject of this paper were recovered from the kitchenmiddens.

Different recovery methods were applied during each of the excavations and, while there are some differences in the bone assemblages (see below), the overall results are similar. The proportions of the major species of birds and mammals proved to be comparable for all three assemblages. Likewise, there was homogeneity in the ages of death, parts of the skeleton and distribution of cut marks of the three assemblages (see Detry & MorenoGarcia 2008, 2010 and Moreno-Garcia & Detry 2010). Methodology The analysis began with morphological and taxonomic identifications of the osteological materials, followed by the quantification of all faunal finds. Vertebrate taxonomic identifications were made with the aid of the reference collection of the Laboratory of Arqueociências of the DGPC (Direcção Geral do Património Cultural General Directorate for Cultural Heritage). In the case of mammals, when it was possible to identify to species, namely the fragments of long bones, skull, ribs and vertebrae, we used general size categories (Macrofauna and Mesofauna) or recorded as unidentified. To distinguish between sheep and goat we used both osteometric parameters as well as the morphological characteristics discussed by Boessneck (1969),

Faunal remains of the modern period are unfortunately scarce, demonstrating the high importance of the present analysis in understanding the diets of these times. Historic documents report on some aspects of the diet; the zooarchaeological results provide unique evidence with which to confirm or question the historic record. The three assemblages examined here were deposited at the same time, in the first half of the 17 th century. These assemblages were recovered by different teams in three separate excavations, each carried out in different field 117

Figure 1. Location of the Santa Clara-a-Velha Monastery. particularly for the calcaneus, astragalus and distal humerus. However, the vast majority of bone of these two species are indistinguishable and therefore classified as an “artificial taxon” Ovis / Capra.

of the cloister. These deposits accumulated in the final phase of occupation of the monastery when the dump areas were submerged due to flooding. Areas 41 and 46 thought to have been beside the bedrooms. The trash in Room B, by contrast, appears to have originated in the cloister. Given that the largest cloister was already submerged the rubbish appears to have been thrown from an upper floor that no longer exists.

The age of death of the main species were calculated whenever possible. In long bones we recorded the existence of non-fused and fused epiphyses or shafts. In addition we registered also the eruption of the permanent dentition and the various stages of tooth wear. In the case of pig and cattle we applied the scale of the teeth defined by Grant (1982) and for caprines we used the wear stages set by Payne (1973, 1987).

During excavation, the material recovered from Area 41 was screened mesh. In A46 all the sediment recovered in the excavation was water sieved in 8 mm and 4 mm mesh, at the laboratory of CIPA. The materials in Room B were recovered by the company of archeology GAAIA and were fully screened with 1mm mesh in the field.

Taphonomic data were also recorded, concerning cutmarks or fire marks of anthropogenic origin and in relation to the status of conservation of the sample. Measurements were taken with a digital caliber Mitutoyo CD-15DC with accuracy of 0.1 mm. We obtained the measures defined by von den Driesch (1976) and in the case of ovicaprids was also used HTC (height trochlea) and BT (width trochlea) humeral defined by Davis (1996).

The reduced number of microfaunal elements in Area 41 differently from Room B and Area 46, enables us to realize that the type of recovery affected the collection of smaller elements. Nevertheless all the sets have a good level of recovery of smaller bones. Taxonomic identification

Results and Discussion We examined about 28,000 skeletal remains from the three areas (Table 1). 70% to 85% of the sample, depending on the area analyzed, could not be identified to a taxonomic level. Most mammals were from Room B. Some of these remains were classified as indeterminate (550 - Area 41, 911- Area 46, 4604- RoomB), and ranked

Deposition and Recovery Areas 41 and 46, in the west wing of the cloister, appear to have been produced from by outer side of the cloister; while in Room B they were produced from the south wall 118

Figure 3. Proximal femur of Gallus domesticus with puncture marks produced by a carnivore. In cattle the %NMI decreases in relation to the %NRD, which shows that cattle carcasses were cut into smaller pieces in order to be handled or cooked. The hind part was also the most common both in cattle and pork, probably these parts would be purchased or received has rent instead of handling the complete animals. We identified 196 lagomorphs in Area 41, of which 133 were rabbit and 14 hare. It is not possible to discern if the rabbit is domestic or wild. The hare is the only species that we can confirm as hunted.

Figure 2. Location of the areas excavated in the Santa Clara-a-Velha monastery. by size class, of macrofauna (806 - Area 41, 231- Area 46, 2082-RoomB) and mesofauna ( 4084-Area 41, 756Area 46, 8152-RoomB). No horse or deer were identified which lead us to conclude that all the macrofauna were cattle. Most fragments classified as mesofauna probably belong to caprines (goat or sheep) since they are dominant in the sample. The vast majority of macrofauna and mesofauna remains were vertebrae and ribs, anatomical elements that are difficult to identify.

As for Area 46, around 87 rabbit bones were identified and 12 lagomorphs remains. Only one bone was firmly identified as hare. In Room B and Area 46 the number of rabbits increased dramatically, a pattern that is the result of improved recovery methods that allowed the collection of smaller bones and bone fragments. The rabbit found in this sample was comparable with wild rabbit in size. However, it is likely that wild rabbits were caught and kept in captivity rather than hunted; according to Callou (2003) the only truly domesticated rabbit appears in the eighteenth and nineteenth centuries. Prior to the eighteenth century wild rabbits were caught and held captive.

Historic documents from the Santa Clara-a-Velha report that the monastery was allowed to keep up to about 900 sheep a year (Oliveira 1972). In various cases sheep and fowl were provided to the monastery in the form of rents of the lands they owned. Thus most of the meat consumed by these nuns came from animals raised on their properties, a system that requires organization and a high degree of wealth.

Finally, a domestic cat was found in Area 41 (NISP-31; MNI-2). This species was clearly not for consumption but commensal. The presence of cats at such monasteries was allowed by the rules in order to keep mice away. (“Nullum animal infra claustrum, exceptis murilegis, teneatur” – “Dentro da clausura não se tenha nenhum animal, a não ser os que caçam ratos”.”Bullarium Franciscanum III, RgIs,FVlF7.). The presence of cat may also explain the bite mark on many of the other bones, which appear as relatively isolated punctures, corresponding to the patterns made by the canines of cats (Fig. 3).

Cattle and swine constitute 11% to 13%, and 9% to 22% of the sample identified (in NISP-Numbers of identified specimens), respectively. The calves and piglets would complement and bring variety to the diet. The rent records make reference to pork but the origin of the cattle is not documented. The cattle remains are, however, significantly less frequent and seemingly dominated by the most valuable parts of the carcass, suggesting that they were not kept on monastery lands but otherwise purchased or provided from elsewhere.

119

Bos taurus (Cattle) Ovis/Capra (Sheep/Goat) Ovis aries (Sheep) Sus sp. (Pig/Wild boar) Oryctolagus cuniculus (Rabbit) Lepus sp. (Hare) Lagomorpha (Rabbit/Hare) Felis catus (Cat) Total Determined Macrofauna Mesofauna Undetermined Total Non Determinaded TOTAL

Area41 NISP % 265 3

Area 46 NISP % 50 2

Room B NISP % 256 1

Area 41 MNI % 5 5

Area 46 MNI % 10 17

Room B MNI % 9 10

1495

19

211

9

1440

8

53

48

13

22

41

47

151

2

10