The Origins of Transhumant Pastoralism in Temperate SE Europe - Arnold Greenfield

The Origins of Transhumant
Pastorialism in Temperate South

Eastern Europe:
A Zooarchaeological Perspective
from the Central Balkans
Elizabeth R. Arnold
Haskel J. Greenfield
BAR International Series XXXX

2006
BOOK JUSTIFICATION
This book is designed as a test case for the identification of transhumant pastoralism using zooarchaeological
techniques that have been proposed in the literature. Most studies that try to identify transhumance do not use
zooarchaeological methods and hence are open to a variety of interpretations. By moving the study out of an arid
environment and into a temperate environment, we remove the important variable of environment as a constraint. Most
studies have tried to identify transhumance as deriving from the Early Neolithic (and beginning of animal
domestication), where it is difficult to see differences because of similarities between hunter-gatherer and early pastoral
behaviour. In our study, we try to get beyond these issues by working in an area without any expectation of
transhumance associated with the earliest farming cultures and testing for its later appearance as part of a package of
larger changes associated with the secondary products revolution. Little is still known about the evolution of this
important and ancient form of land use and domestic animal management, especially in temperate environmental zones.
Yet, zooarchaeological data can be used to answer basic questions concerning the origin and nature of early
transhumant pastoralism. Such research has yielded data suggesting that transhumant pastoralism may have initially
appeared in the central Balkans early in the Post Neolithic (Eneolithic ca. 3300 B.C., calibrated radiocarbon dating).
This is the earliest dates for the appearance of transhumant pastoralism in the temperate zones of Europe.
ABSTRACT
This book addresses the issue of the temporal origins of transhumant pastoralism in temperate southeastern Europe
(northern half of the Balkan Peninsula). In this region, there is little of the environmental imperative frequently cited to
account for the origins of transhumance, in contrast to the Mediterranean littoral. The climate does not force the
migration of animals from the lowlands in the summer into the highlands, and back into the lowlands because of
insufficient graze and harsh temperatures. Yet, this form of land use and animal exploitation pattern has a very long
history in southern Europe, extending at least to the Roman period. However, little is known about its origins and
development.
In recent years, several hypotheses have been suggested to explain when and why transhumant pastoralism with
domestic animals appeared across the southern Mediterranean. Each hypothesis proposes a different point in time when
transhumance would appear, ranging from the appearance of the earliest domestic animals (advent of the Early
Neolithic), to the appearance of secondary product exploitation (advent of the Post Neolithic), and to the appearance of
complex societies (advent of the Iron Age). Previous attempts to test these hypotheses has indicated that transhumance
with domestic stock did not appear in the temperate zone until the advent of the Eneolithic (c. 3300 BC), long after the
beginning of animal domestication in the Early Neolithic (c. 6100 BC).
The hypotheses are tested by examining the tooth remains from three domestic animal taxa (Ovis/Capra, Bos taurus and
Sus scrofa) from archaeological sites in the central part of the northern Balkans (also known as the Central Balkans).
Data from eleven sites in the region, with statistically sufficient samples and spanning the period from the Early
Neolithic through to the Early Iron Age, were tabulated to test the hypotheses.
The primary technique involved the creation of harvest profiles from mandibular tooth wear and eruption data of
domestic animal to examine age of death, the associated season of death and exploitation strategies. Season of death and
the seasonality of culling practices were also examined for each taxon through additional graphical evaluation of the
data and were supplemented by the secondary technique of cementum analysis of modern and archaeological
mandibular Ovis aries and Capra hircus teeth. The specific hypothesis used in this investigation was that transhumant
pastoralism would appear at the temporal point where complementary culling patterns between highland and lowland
sites in the region appear. Based on other sources of data, such a pattern was expected to appear at the advent of the
Post Neolithic.
Several overriding taphonomic issues affecting sample size greatly hampered the creation of the traditional harvest
profiles and limited the capability to evaluate the original hypotheses by these means. However, the harvest profile data
does lend itself to provide further support for the secondary products revolution model, which is hypothesized to occur
at the advent of the Post Neolithic (Eneolithic-Iron Age) in the region. The seasonality data, both graphically and from
the cementum analysis, show complementary profiles between highland and lowland areas in the Post Neolithic,
lending support to the original hypotheses of transhumant movement. At the same time, problematic divisions of age
groups of Ovis/Capra are revealed as potentially masking herd movements that might otherwise be revealed in the
traditional harvest profiles.
It is clear from the analysis that strong changes in not only culling patterns, but also land use take place during the
Eneolithic and Bronze Age of the region. These are interpreted in light of two regional developments the advent of
transhumant pastoralism and the beginning of the Secondary Products Revolution.
iii
CHAPTER SUMMARY
Chapter 1 presents the theoretical background to the research problem and outlines the methodology, techniques and
data that are utilized in the research.
Chapter 2 provides a definition and discussion of transhumant pastoralism and relevant environmental and ecological
parameters.
Chapter 3 examines previous research on the origins of transhumant pastoralism in Europe focusing on both the
Mediterranean and the northern temperate region of the Balkans. The research of Geddes, Halstead and Greenfield are
the focus.
Chapter 4 introduces the characteristics of the regional environment including topography, climate and vegetation in
order to consider the viability of transhumant pastoralism within the northern Balkans.
Chapter 5 summarizes the culture history of southeastern Europe from the Neolithic through the Early Iron Age.
Aspects of settlement, fauna, and evidence for sedentism and mobility are examined.
Chapter 6 describes the methodology to be utilized and details the two chosen techniques, tooth wear and eruption and
cementum analysis.
Chapter 7 describes the data examined in this investigation. In the chapter, each site is described, including site
location, environment and nature of deposits. The mandibular and loose teeth remains are quantified by species and by
major period. It also describes the thin sectioning data, both the modern comparative collection and the archaeological
sample.
Chapter 8 presents the results of the data analysis. The first part focuses on the tooth wear and eruption data, examining
both production strategies and implications for the transhumant movement of herds. The second part focuses on the thin
sectioning results.
Chapter 9 presents the final conclusions regarding the data and discusses their implication in terms of the origins of
transhumant pastoralism in the northern Balkans.
FOREWORD
This work is an example of an idea that had a long gestation period and that has gone through almost 20 years of data
collection. It was one of those tasks that required the collection of data from a long range of time periods and over a
vast geographic scale. As a result, it has its origins in the PhD thesis research of Haskel Greenfield and it became
complete with the MA thesis research of Elizabeth Arnold. In 1977, Greenfield began to work in Serbia under the
supervision of H. Arthur Bankoff (Brooklyn College). This initial field season of survey and test excavations in the
central Balkans led to their return and the beginning of systematic excavation at the Post Neolithic site of Novaka
uprija. The faunal data from this and other excavations by Bankoff were offered to Greenfield for analysis as part of
his developing PhD thesis on Post Neolithic subsistence practices. In preparation for the identification of
zooarchaeological specimens from the central Balkans, Greenfield received training in Budapest by the late Sandor
Bknyi. The experiences at Novaka uprija eventually led to the analyses of many other collections in the region
from both the Neolithic and Post Neolithic, many of which were reported in Greenfield (1986) and subsequent
publications. Greenfields thesis and early publications were primarily concerned with testing Andrew Sherratts
hypothesis on the advent of secondary products exploitation in Europe.
Early on in the analysis of the specimens for the testing of the secondary products hypothesis, Greenfield noticed an
apparent discrepancy between the patterns for highland and lowland Post Neolithic harvest profiles of ovicaprines and
cattle, which he interpreted as evidence for the advent of transhumant pastoralism (1988, 1991, 1999a, 2001a).
However, he felt that he only had sufficient data to point out the possible pattern rather than to comprehensively test this
hypothesis. One gap in his early research was the paucity of highland (alpine or subalpine assemblages). In order to
rectify this gap, he joined Blagoje Govedaricas research project which was in the process of excavating Post Neolithic
settlements on the Glasinac Plateau (east Bosnia). The project eventually focused on the site at Kadica Brdo which was
excavated from 1986-1990. The excavations, as many others, were terminated by the Yugoslavian civil wars and have
never resumed. Only preliminary reports of the excavations were ever published. However, sufficient fauna were
recovered and analyzed to finally provide a test for transhumance from the highland zone for the EIA.
One of the major problems that Greenfield encountered in his initial analyses was that he included all fragments in his
ageing of specimens in order to maintain high sample sizes. The detailed tooth eruption and wear data were never
separately analysed because of problems with the early computer recording of the information. In the 1990s, with the
appearance of easily accessible computer spreadsheets, Greenfield was able to recover all of his old data from the
region and make them comparable for reanalysis. This included the tooth wear and eruption data.
During the later years of Greenfields research in the region, new and exciting techniques in zooarchaeology were
beginning to appear and spread throughout the discipline i.e. tooth cementum analysis. Influenced by its potential for
the study of transhumance, Greenfield began to collect what were then considered to be suitable specimens for such
analysis. These were brought back to the University of Manitoba for eventual study, and there they lay until Elizabeth
Arnold entered the picture.
Arnold worked with Greenfield on a variety of data sets from the region, upgrading them and putting them all into
compatible computer formats (Excel spreadsheets). From this database, the tooth wear and eruption data were extracted
by Arnold for a more complete analysis. Some of the archaeological faunal assemblages described herein were stored at
the University of Manitoba. These were sorted and all the teeth were separated for analysis by Arnold. Under the
guidance of Dr. Ariane Burke, Arnold gained the knowledge and experience of thin sectioning and cementum analysis.
A sample of the domestic Ovis/Capra material was selected for testing of the transhumance hypotheses using these
techniques. Comparative modern Ovis/Capra material was collected on monthly trips to local abattoirs and small-scale
farms over the course of a year and prepared in a similar manner. This work is a result of these combined efforts and
author order does not imply relative effort in completing this monograph.
vii
ACKNOWLEDGEMENTS
There are too many people to properly thank for access to data used in this work. They include all of Greenfields
various collaborators through the years in ex-Yugoslavia. In particular, we would like to thank the directors (or
codirectors) of the various projects or people who arranged access to the data from each of the sites used in the analyses
presented within this book, including Florin Draovean (Foeni-Sala), eljko Je (Petnica), Mirjana Vukmanovi and
Petar Popovi (Livade), H. Arthur Bankoff (Novaka uprija), Blagoje Govedarica (Kadica Brdo), Milenko
Bogdanovi (Ljuljaci), Mihalis Fotiadis (Megalo Nisi Galanis), Ljubomir Bukvi (Opovo) and the late Svetozar
Stankovi (Blagotin, Stragari-ljivik), and the late Professors Milutin Garaanin and Dragoslav Srejovi (Vina-Belo
Brdo). Other individuals also played vital roles in the data collection from the region, including Vesna Jeremenko, Tina
Jongsma, Dimitrije Madas, Guilmine Eygun, Igor and Andrej Starovi and the staff of the Istraivaka Stanica Petnica,
Alexandr Radoman, Bojana Vojkovi, Zev Greenfield, and the late Vladimir Lekovi. Without the help of each and
every one of the above, it would never have been possible to have accumulated such a wealth of comparative
information from the region.
Thanks to Clayton Robins (Manitoba Sheep Association) and Sharon Peddler (Manitoba Goat Association) for
providing contacts and direction; to Monica Griffiths, for supplying all the modern goat specimens as well as valued
information, and for continued interest; to Lee Perreault and the staff at Prairie Abattoir and Jim and Doris Holmes and
staff from Carmen Meats; and Randy and Solange Eros for having the interest and taking the time to provide sheep
specimens.
Thanks must go to our colleagues and graduate students at the University of Manitoba who helped at various stages in
the preparation of specimens, organization of data, and presentation of results. In particular, we would like to thank Val
McKinley and Ariane Burke for Elizabeth Arnolds training and their assistance in the University of Manitobas thin
sectioning laboratory. Vals support went far beyond simple technical advice. Special thanks must also be accorded to
Tina Jongsma who aided in the field collection and analysis of many of the specimens described here and has allowed
them to be incorporated into our analysis. In addition, we would like to Chris Meiklejohn and Karin Wittenberg for their
involvement, patience and advice during the various revisions of this work and Dennis Murphy (Statistical Assistance
Center, University of Manitoba) for his help with statistical issues. Figures 6.3, 8.4, 8.24, and 8.25 are reproduced with
the permission of Sebastian Payne.
Most of all, special thanks must be extended to our families for their constant support. None of this could have been
accomplished without them. They continuously sacrificed in order to ensure that this work would see the light of day.
This work is dedicated to the memory of Rita Fecher, Haskels mother, who passed away on June 13, 2003. This
research could not have been undertaken without her inspiration for Haskel to persevere in the face of insurmountable
odds during his studies and especially during the long periods of field work that were required to collect the body of
data. She would have been proud to have seen its completion before her untimely passing.
ix
TABLE OF CONTENTS
CHAPTER AND SECTION TITLES
Book Justification
Abstract
Chapter Summary
Foreword
Acknowledgments
Table of Contents
I. Chapter and section titles
II. List of Figures
III. List of Tables
IV. List of Appendices
i
iii
v
vii
xi
xi
xi
xv
xvii
xix
CHAPTER 1: INTRODUCTION
I. Theoretical background to research problem

II. Research hypotheses
III. Method and Technique
IV. Data
A. The region
B. Temporal span
C. Data
V. Conclusions
1
1
2
3
3
3
3
5
CHAPTER 2: TRANSHUMANT PASTORALISM: SOME ETHNOGRAPHIC CONSIDERATIONS

I. Introduction to pastoralism
II. Types of pastoralism
A. Nomadic pastoralism
B. Semi-nomadic pastoralism
C. Transhumant pastoralism
III. Aspects of modern transhumant pastoralism in the northern Balkans
IV. Conclusion
CHAPTER 3: PREVIOUS RESEARCH ON ORIGINS OF TRANSHUMANCE
PASTORALISM IN PREHISTORIC SOUTHERN EUROPE
I. Introduction
II. Theories on the origins of transhumant pastoralism
A. Political instability
B. Regional symbiosis
C. A by-product of initial animal domestication in an ecologically varied habitat
D. Secondary Products Revolution
III. Previous research on the origins of transhumant pastoralism in Mediterranean Europe
A. Mesolithic/Neolithic continuity
B. Complex societies
IV. Previous research in temperate southeastern Europe
VI. Differences between models comparing apples and oranges
VII. Conclusions
xi
7
7
7
7
7
8
8
11
13
13
13
13
13
14
14
14
15
15
16
17
18
CHAPTER 4: REGIONAL ENVIRONMENT
19
I. Introduction
II. Landforms
A. Topography
B. Rivers
III. Climate
A. Temperature
B. Precipitation
IV. Environmental Regions
A. Highland areas
B. Lowland areas
VI. Regional environment and climate in prehistory
A. Environment and climate in the Balkans during the Neolithic
B. Anthropogenic environmental changes during the Neolithic
C. Environment and climate in the Balkans during the Post Neolithic
D. Anthropogenic environmental changes during the Post Neolithic
VII. Site locations in relation to major environmental regions
VIII. Conclusion
CHAPTER 5: EVIDENCE FOR SETTLEMENT, SEDENTISM AND
MOBILITY IN CENTRAL BALKAN CULTURE HISTORY
19
19
19
20
20
20
21
22
22
22
23
23
23
23
24
24
24
25
I. Introduction
II. Early Neolithic
A. Chronology
B. Cultures
C. Settlement
D. Fauna
E. Evidence for sedentism/mobility
III. Middle Neolithic
A. Chronology
B. Cultures
C. Settlement
D. Fauna
IV. Late Neolithic
A. Chronology
B. Cultures
C. Settlement
D. Fauna
V. Eneolithic (Chalcolithic)
A. Chronology
B. Cultures
C. Settlement
D. Fauna
VI. Early Bronze Age
A. Chronology
B. Cultures
C. Settlement
D. Fauna
VII. Middle and Late Bronze Age
A. Chronology
B. Cultures
C. Settlement
D. Fauna
25
25
25
25
25
26
26
27
27
27
27
27
27
28
28
28
28
28
28
29
29
29
29
29
29
30
30
30
30
30
31
31
31
31
31
31
xii

VIII. Early Iron Age
A. Chronology
B. Cultures
C. Settlement
D. Fauna
IX. Conclusions
31
32
32
32
32
32
32
32
CHAPTER 6: METHODOLOGY
33
I. Introduction
II. Tooth wear and eruption
III. Establishing absolute age using tooth eruption and wear
IV. Production strategies and construction of harvest profiles
V. Cementum analysis
VI. The control sample
VII. The fossil sample
VIII. Age determination with cementum analysis
IX. Season of death determination with cementum analysis
X. Conclusion
CHAPTER 7: DESCRIPTION OF SITES, TIME PERIODS, AND NATURE OF SAMPLES
I. Introduction
II. Blagotin
A. Site description
B. Mandibular and loose tooth remains
III. Foeni-Sala
A. Site description
IV. Kadica Brdo
A. Site description
V. Livade
A. Site description
VI. Ljuljaci
A. Site description
VII. Megalo Nisi Galanis
A. Site description
VIII. Novaka uprija
A. Site description
IX. Opovo
A. Site description
X. Petnica
A. Site description
XI. Selevac
A.Site description
XII. Stragari-ljivik
A. Site description
33
34
35
35
38
38
39
39
39
39
41
41
41
41
41
43
43
44
44
44
46
46
46
48
48
48
50
50
50
50
53
53
53
56
56
56
56
56
56
59
59
62
62
62
62
xiii
XII. Vina
A. Site description
XIII. The modern tooth wear and eruption control sample
XIV. Cementum analysis data: archaeological and modern
A. The modern control sample
B. Archaeological cementum analysis sample
XV. Conclusions
CHAPTER 8: THE IDENTIFICATION OF TRASHUMANT PASTORALISM
THROUGH HARVEST PROFILE AND CEMENTUM ANALYSES
I. Introduction
II. Construction of harvest profiles first stage of analysis
A. Sus scrofa dom.
B. Ovis/Capra
C. Bos taurus
III. Regional scale of analysis second stage of analysis
A. Transhumant movement of pigs
B. Transhumant movement of ovicaprines
C. Transhumant movement of cattle
IV. Comparison of tooth wear and eruption ageing methods third stage of analysis
V. Cementum analysis fourth stage of analysis
A. The modern comparative sample
B. The archaeological sample
VI. Conclusions
CHAPTER 9 CONCLUSIONS
I.
II.
III.
IV.
V.
62
62
64
64
64
65
65
65
71
71
71
71
80
92
102
103
103
104
104
116
116
116
117
119
Introduction
Methodological concerns
The Secondary Products Revolution
Evidence for transhumant pastoralism
Conclusions
119
119
119
120
122
REFERENCES CITED
123
xiv
LIST OF FIGURES
1.1
The Central Balkans and site locations
4.2
Map showing climatic divide between Mediterranean

and temperate central Europe
21
6.3
Proportional allocation example
36
8.4
8.5
8.6
8.7
8.8
8.9
8.10
8.11
8.12
8.13
8.14
8.15
8.16
8.17
8.18
8.19
8.20
8.21
8.22
8.23
8.24
8.25
8.26
8.27
8.28
8.29
8.30
8.31
8.32
8.33
8.34
8.35
8.36
8.37
8.38
8.39
8.40
8.41
8.42
8.43
8.44
8.45
8.46
8.47
8.48
8.49
8.50
8.51
8.52
8.53
Paynes production models

Harvest profile (Sus scrofa) Middle Neolithic Petnica
Harvest profile (Sus scrofa) Late Neolithic Petnica
Harvest profile (Sus scrofa) Late Neolithic Vina
Harvest profile (Sus scrofa) Late Neolithic Opovo
Harvest profile (Sus scrofa) Middle/Late Neolithic Selevac
Harvest profile (Sus scrofa) Eneolithic Petnica
Harvest profile (Sus scrofa) Early Bronze Age Novaka uprija
Harvest profile (Sus scrofa) Early/Middle Bronze Age Ljuljaci
Harvest profile (Sus scrofa) Middle Bronze Age Vina
Harvest profile (Sus scrofa) Late Bronze Age Livade
Harvest profile (Sus scrofa) Early Iron Age Kadica Brdo
Harvest profile (Ovis/Capra) Early Neolithic Foeni-Sala
Harvest profile (Ovis/Capra) Early Neolithic Blagotin
Harvest profile (Ovis/Capra) Early Neolithic
Harvest profile (Ovis/Capra) Middle Neolithic Stragari
Harvest profile (Ovis/Capra) Late Neolithic Vina
Harvest profile (Ovis/Capra) Middle/Late Neolithic Selevac
Harvest profile (Ovis/Capra) Late Neolithic Petnica
Harvest profile (Ovis/Capra) Late Neolithic
Paynes Milk Production
Paynes Wool Production
Harvest profile (Ovis/Capra) Eneolithic Novaka uprija
Harvest profile (Ovis/Capra) Eneolithic Petnica
Harvest profile (Ovis/Capra) Early Bronze Age Novaka uprija
Harvest profile (Ovis/Capra) Middle Bronze Age Vina
Harvest profile (Ovis/Capra) Late Bronze Age Novaka uprija
Harvest profile (Ovis/Capra) Late Bronze Age Livade
Harvest profile (Ovis/Capra) Late Bronze Age Petnica
Harvest profile (Ovis/Capra) Late Bronze Age
Harvest profile (Ovis/Capra) Early Iron Age Kadica Brdo
Harvest profile (Ovis/Capra) Final Neolithic Megalo Nisi Galanis
Harvest profile (Ovis/Capra) Final Neolithic/Early Bronze Age Megalo Nisi Galanis
Harvest profile (Bos taurus) Early Neolithic Foeni-Sala
Harvest profile (Bos taurus) Early Neolithic Blagotin
Harvest profile Early Neolithic Bos vs. Ovis/Capra
Harvest profile (Bos taurus) Middle Neolithic Stragari
Harvest profile (Bos taurus) Middle Neolithic Petnica
Harvest profile (Bos taurus) Middle Neolithic
Harvest profile (Bos taurus) Late Neolithic Opovo
Harvest profile (Bos taurus) Late Neolithic Vina
Harvest profile (Bos taurus) Late Neolithic Petnica
Harvest profile (Bos taurus) Middle/Late Neolithic Selevac
Harvest profile (Bos taurus) Late Neolithic
Harvest profile (Bos taurus) Eneolithic Blagotin
Harvest profile (Bos taurus) Early/Middle Bronze Age Ljuljaci
Harvest profile (Bos taurus) Middle Bronze Age Vina
Harvest profile (Bos taurus) Late Bronze Age Livade
Harvest profile (Bos taurus) Early Iron Age Kadica Brdo
Percentage of age groups (Sus scrofa) Neolithic
73
75
75
76
76
77
77
78
78
79
79
80
82
82
83
83
84
84
85
85
86
87
88
89
89
94
94
95
95
96
96
97
97
98
98
99
99
100
100
101
105
105
106
106
107
107
108
108
109
109
xv
8.54
8.55
8.56
8.57
8.58
8.59
8.60
8.61
8.62
8.63
8.64
8.65
8.66
8.67
Percentage of age groups (Sus scrofa) Post Neolithic

Percentage of age groups (Ovis/Capra) Neolithic
Percentage of age groups (Ovis/Capra) Post Neolithic
Percentage of age groups (Bos taurus) Neolithic
Percentage of age groups (Bos taurus) Post Neolithic
Modern comparative thin sectioning sample (Sheep #1)
Modern comparative thin sectioning sample (Goat 14 b)
Modern comparative thin sectioning sample (Goat #13 b)
Archaeological thin sectioning sample (Kadica Brdo sample #2)
Archaeological thin sectioning sample (Vina sample #1)
Archaeological thin sectioning sample (Vina sample #10)
xvi
110
110
111
111
112
113
113
113
114
114
114
115
115
115
LIST OF TABLES
6.1
6.2
6.3
Sheep/Goat mandibular wear stages (MWS) and suggested ages

Cattle mandibular wear stages (MWS) and suggested ages
Pig mandibular wear stages (MWS) and suggested ages
36
36
37
7.4
7.5
7.6
7.7
7.8
7.9
7.10
7.11
7.12
7.13
7.14
7.15
7.16
7.17
7.18
7.19
7.20
7.21
7.22
Stage distribution of Ovis aries mandibles and loose teeth from Early Neolithic Blagotin
Stage distribution of Ovis/Capra mandibles and loose teeth from Early Neolithic Blagotin
Stage distribution of Ovis/Capra mandibles and loose teeth from Early Iron Age Blagotin
Stage distribution of Bos taurus mandibles and loose teeth from Early Neolithic Blagotin
Stage distribution of Bos taurus mandibles and loose teeth from Eneolithic Blagotin
Stage distribution of Ovis aries mandibles and loose teeth from Early Neolithic Foeni-Sala
Stage distribution of Ovis/Capra mandibles and loose teeth from Early Neolithic Foeni-Sala
Stage distribution of Ovis/Capra mandibles and loose teeth from Early Iron Age Foeni-Sala
Stage distribution of Bos taurus mandibles and loose teeth from Early Neolithic Foeni-Sala
Stage distribution of Ovis aries mandibles and loose teeth from Early Iron Age Kadica Brdo
Stage distribution of Ovis/Capra mandibles and loose teeth from Early Iron Age Kadica Brdo
Stage distribution of Bos taurus mandibles and loose teeth from Early Iron Age Kadica Brdo
Stage distribution of Sus scrofa mandibles and loose teeth from Early Iron Age Kadica Brdo
Stage distribution of Ovis/Capra mandibles from Late Bronze Age Livade
Stage distribution of Bos taurus mandibles and loose teeth from Late Bronze Age Livade
Stage distribution of Sus scrofa mandibles and loose teeth from Late Bronze Age Livade
Stage distribution of Bos taurus mandibles and loose teeth from Early/Middle Bronze Age Ljuljaci
Stage distribution of Sus scrofa mandibles and loose teeth from Early/Middle Bronze Age Ljuljaci
Stage distribution of Ovis aries mandibles and loose teeth from Late Neolithic/Final Neolithic
Megalo Nisi Galanis
Stage distribution of Ovis/Capra mandibles and loose teeth from Late Neolithic/
Final Neolithic Megalo Nisi Galanis
Stage distribution of Ovis aries mandibles and loose teeth from Final Neolithic Megalo Nisi Galanis
Stage distribution of Ovis/Capra mandibles and loose teeth from Final Neolithic Megalo Nisi Galanis
Stage distribution of Ovis aries mandibles and loose teeth from Final Neolithic/
Early Bronze Age Megalo Nisi Galanis
Stage distribution of Ovis/Capra mandibles and loose teeth from Final Neolithic/
Early Bronze Age Megalo Nisi Galanis
Stage distribution of Ovis/Capra mandibles and loose teeth from Eneolithic Novaka uprija
Stage distribution of Ovis/Capra mandibles and loose teeth from Early Bronze Age Novaka uprija
Stage distribution of Ovis/Capra mandibles and loose teeth from Late Bronze Age Novaka uprija
Stage distribution of Sus scrofa mandibles and loose teeth from Early Bronze Age Novaka uprija
Stage distribution of Bos taurus mandibles and loose teeth from Late Neolithic Opovo
Stage distribution of Sus scrofa mandibles and loose teeth from Late Neolithic Opovo
Stage distribution of Ovis/Capra mandibles and loose teeth from Late Neolithic Petnica
Stage distribution of Ovis/Capra mandibles and loose teeth from Eneolithic Petnica
Stage distribution of Ovis/Capra mandibles and loose teeth from Late Bronze Age Petnica
Stage distribution of Bos taurus mandibles and loose teeth from Middle Neolithic Petnica
Stage distribution of Bos taurus mandibles and loose teeth from Late Neolithic Petnica
Stage distribution of Sus scrofa mandibles and loose teeth from Middle Neolithic Petnica
Stage distribution of Sus scrofa mandibles and loose teeth from Late Neolithic Petnica
Stage distribution of Sus scrofa mandibles and loose teeth from Eneolithic Petnica
Stage distribution of Ovis/Capra mandibles and loose teeth from Middle Neolithic Stragari
Stage distribution of Bos taurus mandibles and loose teeth from Middle Neolithic Stragari
Stage distribution of Ovis/Capra mandibles and loose teeth from Late Neolithic Vina
Stage distribution of Ovis/Capra mandibles and loose teeth from Middle Bronze Age Vina
Stage distribution of Bos taurus mandibles and loose teeth from Late Neolithic Vina
Stage distribution of Bos taurus mandibles and loose teeth from Middle Bronze Age Vina
Stage distribution of Sus scrofa mandibles and loose teeth from Late Neolithic Vina
Stage distribution of Sus scrofa mandibles and loose teeth from Middle Bronze Age Vina
Modern Ovis/Capra comparative cementum analysis Summary of readings
Archaeological Ovis/Capra cementum analysis Summary of readings
42
42
42
43
43
45
45
45
46
47
47
47
48
49
49
49
51
51
51
7.23
7.24
7.25
7.26
7.27
7.28
7.29
7.30
7.31
7.32
7.33
7.34
7.35
7.36
7.37
7.38
7.39
7.40
7.41
7.42
7.43
7.44
7.45
7.46
7.47
7.48
7.49
7.50
7.51
xvii
52
52
52
54
54
54
55
55
57
57
57
58
58
60
60
60
61
61
61
63
63
63
65
66
66
66
67
68
69
8.52
8.53
8.54
Summary of sieving and weathering

Expectations for movement in transhumant pattern
Summary of strata and weathering
xviii
72
73
91
LIST OF APPENDICES
A. SUMMARY OF AGEABLE TOOTH WEAR AND ERUPTION DATA BY SITE, TAXON AND PERIOD.
133
B. STAGE DISTRIBUTION DATA OF SMALL SAMPLES:

Table B1. Stage distribution of Ovis/Capra mandibles and loose teeth from Eneolithic Blagotin.
Table B2. Stage distribution of Bos taurus mandibles and loose teeth from Early Iron Age Blagotin.
Table B3. Stage distribution of Sus scrofa mandibles and loose teeth from Early Neolithic Blagotin.
Table B4. Stage distribution of Sus scrofa mandibles and loose teeth from Early Iron Age Blagotin.
Table B5. Stage distribution of Ovis aries mandibles and loose teeth from Early Iron Age Foeni-Sala.
Table B6. Stage distribution of Bos taurus mandibles from Early Iron Age Foeni-Sala.
Table B7. Stage distribution of Sus scrofa mandibles and loose teeth from Early Neolithic Foeni-Sala.
Table B8. Stage distribution of Capra hircus mandibles and loose teeth from Early Iron Age Kadica Brdo.
Table B9. Stage distribution of Ovis/Capra mandibles and loose teeth from Early Bronze Age Ljuljaci.
Table B10. Stage distribution of Ovis/Capra mandibles and loose teeth from Early/
Middle Bronze Age Ljuljaci.
Table B11. Stage distribution of Ovis/Capra mandibles and loose teeth from Middle Bronze Age Ljuljaci.
Table B12. Stage distribution of Bos taurus mandibles and loose teeth from Early Bronze Age Ljuljaci.
Table B13. Stage distribution of Sus scrofa mandibles and loose teeth from Early Bronze Age Ljuljaci.
Table B14. Stage distribution of Sus scrofa mandibles and loose teeth from Middle Bronze Age Ljuljaci.
Table B15. Stage distribution of Bos taurus mandibles and loose teeth from Final Neolithic Megalo
Nisi Galanis.
Table B16. Stage distribution of Sus scrofa mandibles and loose teeth from Late Neolithic/
Final Neolithic Megalo Nisi Galanis.
Table B17. Stage distribution of Sus scrofa mandibles and loose teeth from
Final Neolithic Megalo Nisi Galanis.
Table B18. Stage distribution of Bos taurus mandibles and loose teeth from
Eneolithic Novaka uprija.
Early Bronze Age Novaka uprija.
Late Bronze Age Novaka uprija.
Eneolithic Novaka uprija.
Late Bronze Age Novaka uprija.
Table B23. Stage distribution of Ovis/Capra mandibles and loose teeth from Late Neolithic Opovo.
Table B24. Stage distribution of Ovis/Capra mandibles and loose teeth from Middle Neolithic Petnica.
Table B25. Stage distribution of Ovis/Capra mandibles and loose teeth from
Late Neolithic/Eneolithic Petnica.
Table B26. Stage distribution of Ovis/Capra mandibles and loose teeth from
Late Bronze Age/Early Iron Age Petnica.
Table B27. Stage distribution of Bos taurus mandibles and loose teeth from Eneolithic Petnica.
Table B28. Stage distribution of Bos taurus mandibles and loose teeth from Late Bronze Age Petnica.
Table B30. Stage distribution of Sus scrofa mandibles and loose teeth from Late Bronze Age Petnica.
Table B32. Stage distribution of Ovis aries mandibles and loose teeth from Middle Neolithic Stragari.
Table B33. Stage distribution of Sus scrofa mandibles and loose teeth from Middle Neolithic Stragari.
Table B34. Stage distribution of Ovis/Capra mandibles and loose teeth from Eneolithic Vina.
Table B35. Stage distribution of Sus scrofa mandibles and loose teeth from Eneolithic Vina.
xix
177
177
177
178
178
178
179
179
180
180
180
181
181
181
182
182
182
183
183
183
184
184
184
185
185
185
186
186
186
187
187
187
188
188
C. COMPARISON OF THE KNOWN AGE OF DEATH FROM MODERN SPECIMENS

WITH GRANT (1975) AND PAYNES (1973) TOOTH ERUPTION AND WEAR SEQUENCES RECORDING METHODS.
189
D. DETAILS OF THE DENTAL CEMENTUM ANALYSIS FROM ARCHAEOLOGICAL AND MODERN SPECIMENS.
197
E. STATISTICAL ANALYSIS DATA
199
Table E1. Statistical analysis data - Late Neolithic Sus scrofa.

Table E2. Statistical analysis data - Bronze Age Sus scrofa.
Table E3. Statistical analysis data - comparison of major periods (Sus scrofa).
Table E4. Statistical analysis data - Early Neolithic Ovis/Capra.
Table E5. Statistical analysis data - Late Neolithic Ovis/Capra.
Table E6. Statistical analysis data - Eneolithic Ovis/Capra.
Table E7. Statistical analysis data - Early/Middle Bronze Age Ovis/Capra.
Table E8. Statistical analysis data - Late Bronze Age Ovis/Capra.
Table E9. Statistical analysis data - Early Neolithic Bos taurus.
Table E10. Statistical analysis data - Early Neolithic Bos vs. Ovis/Capra.
Table E11. Statistical analysis data - Middle Neolithic Bos taurus.
Table E12. Statistical analysis data - Late Neolithic Bos taurus.
xx
CHAPTER 1
INTRODUCTION
I. Theoretical background to research problem
Europe. Greenfield (1986, 1988, 1991, 1999a, 2001b)

hypothesized that transhumant pastoralism became a
major element of the subsistence strategies in temperate
southeastern Europe (northern Balkans) only at the
beginning of the Post Neolithic, that is, the Eneolithic and
Bronze Age (ca. 3300 BC). At this time, there were
significant changes in land use, such as the colonization
of agriculturally marginal highlands. He proposed that the
appearance of transhumant pastoralism was part of the
shifts in settlement patterns, mortuary practices,
architectural and artifactual remains, and economy
(including the appearance of secondary animal
productsfollowing Sherratt 1980, 1982) that occurred
at this time.
Transhumant pastoralism is an economic activity

involving the seasonal movement of domestic herds
between altitudinally differentiated and complementary
pastures (Geddes 1983; Khazanov 1984). It is still an
important form of land use in many parts of the world, in
particular in mountainous regions in Africa (e.g. EvansPritchard 1940; Dahl and Hjort 1975), Europe (southern
EuropeBartosiewicz and Greenfield 1999; Greenfield
2001b; NorwayPaine 1994), South America (Argentina
Glatzer 1982; the AndesLynch 1971, 1980, 1983),
Near East (Barth 1961; Bates 1974; Irons 1975; Sweet
1965), and Asia (PakistanEhlers and Kreutzmann
2000; and TibetEkvall 1968). With a stronger
understanding of its origins, we can better understand its
development and effects on the shaping of a cultures
social organization (Bartosiewicz and Greenfield 1999).
Although it has been argued that transhumant

pastoralism, as a strategy of domestic animal
management and land use, has had a long history of
development in the arid regions of the Mediterranean
littoral (e.g. Barker 1973; 1975; Chapman 1981; 1982;
Chapman and Mller 1990; Geddes 1982; Higgs et al.
1967; Hesse 1982; Hole 1978; Hole, Flannery and Neely
1969; Wheeler Pires-Ferreira 1975), little is known about
its evolutionary development in the temperate
environmental zones found in the mountainous interior of
the European sub-continent. Transhumant pastoralism
can be historically documented in the Balkans at least as
far back as the Roman era (e.g. in Dalmatia
Antonijevi 1982; Dedijer 1916; Cviji 1918; Sterud
1978). Few studies dealing with earlier periods in the
temperate zone have convincingly established its
presence or absence in the temperate zone. This
investigation seeks to delimit the origins of transhumant
pastoralism in Europe, specifically within the temperate
regions of southeastern Europe (northern Balkans). The
results of this study will provide an important
comparative example and determine if the ancestry of
transhumant pastoralism lies deeper in the pasts than
suggested by historical reconstructions.
Historically, transhumant pastoralism has been a

significant part of the economy in southern Europe,
including the Mediterranean littoral and the Balkan
Peninsula (Bartosiewicz and Greenfield 1999; Chang and
Koster 1986; Cherry 1988; Nisbet et al. 1991; Whickham
1985). Archaeologically, many researchers assume and
some explicitly argue (Geddes 1983; Greenfield 1986a,
1988, 1999a, 2001a; Halstead 1981, 1996; Harding 2000;
Sherratt 1981, 1983a; Sterud 1978) that it was also an
important element of the economy in prehistoric times.
However, the exact temporal origins of transhumant
pastoralism in southern Europe are still debated within
the archaeological literature (Bartosiewicz and Greenfield
1999), and generate intense discussions.
Over the past twenty years, several research projects have
been directed at elucidating these temporal origins. Most
scholars have focused on the Mediterranean littoral
environment. Geddes (1983) and others (e.g. Miracle and
Forenbaher 2005) proposed that transhumant pastoralism
appeared at the Mesolithic-Neolithic junction because they
argued that prehistoric hunters and gatherers and early
subsistence farmers would have easily incorporated
domestic migratory stock into their seasonal exploitation of
land and resources. Halstead (1981, 1991, 1996) suggested
that transhumance only appears with the emergence of Late
Bronze Age and Classical complex societies in the
Mediterranean. Lewthwaite (1981, 1984) and Walker (1983)
argued that transhumance appeared even later, during the
Classical or Medieval periods, with the need for the
movement of specialized pastoralism for urban markets.
II. Research hypotheses

Temporally, three major temporal moments have been
hypothesized to be the points in time when transhumant
pastoralism might have appeared. They are the following:
1.
In contrast, few projects have focused on the temperate

European environment, immediately to the north of the
mountainous divide between Mediterranean and central
Early NeolithicTranshumant pastoralism has long

been proposed to have existed from the beginning of
animal domestication in Europe. It is thought to be
either a continuation of the migratory patterns of
Mesolithic indigenous hunter-gatherers who adopted
domestic animals into their repertoire (Geddes 1982;
Sterud 1978) or appeared as a natural consequence of
pastoralism during the Neolithic periods (Barker 1973;
INTRODUCTION
1975; Chapman 1981; 1982; Chapman and Mller
1990; Hesse 1982; Hole 1978; Nisbet et al. 1991;
Miracle and Forenbaher 2005; Wheeler Pires-Ferreira
1975).
It is with the shift to longer distance transhumance, where

the animals are away from the village for a lengthy period
of the year, that transhumance may become
archaeologically identifiable.
2.
Early Post NeolithicThe Secondary Products

Revolution model posits that the diffusion of new
production technologies and domestic breeds from the
East at the onset of the Post Neolithic (Eneolithic or
Early Bronze Age) enabled domestic animal
exploitation patterns to shift from primary (meat, hide
and bone) to secondary products (wool, milk and
traction). Herds increased in size to more effectively
produce secondary products. To avoid placing strains
upon local economies (to produce and store winter
fodder), herds were moved to unoccupied highland
pastures for the summer and returned to the lowlands
with the onset of inclement weather (Sherratt 1981,
1983a).
3.
Iron Age (or later)Transhumant pastoralism is

historically known from vast areas of the
Mediterranean littoral by classical times (e.g. Dalmatia
and Greece Classical era (Antonijevi 1982; Halstead
1987, 1991; Nisbet et al. 1991; Sterud 1978). The
large scale long distance specialized transhumant
adaptations characteristic of the area (Antonijevi
1982; Sterud 1978) and elsewhere in southern Europe
(Chang and Tourtelotte 1993; Halstead 1981, 1987;
Lewthwaite 1981, 1984; Walker 1983) appear to be
very late (early historical periods) phenomenon.
Transhumant pastoralism has been hypothesized to be
a function of the appearance of large urban markets
and productive specialization that appear in Classical
or Medieval times (seen also in the Near East at an
earlier timeLees and Bates 1974). As a result, it can
be hypothesized to be a result of the formation of early
complex societies in the Iron Age or with the
introduction of the Roman era.
Long versus short distance transhumant migration would

have different archaeological signatures because of the
differences in the scale of movement. Long distance
movements by specialists require the construction of pens
and huts in both highland and lowland pastures (e.g.
Chang and Tourtellotte 1991; Creighton and Segui 1998).
This additional architectural investment results in
identifiable archaeological signatures. In contrast, short
distance village based herders rarely require the
construction of such temporary structures, except where
they are very distant from their home bases (e.g. among
lowland herders who have moved into summer highland
pastures). These differences are borne out by the
ethnoarchaeological studies (e.g. Barker 1991; Baker
1999; Chang and Koster 1986; Nandris 1985, 1991) and
we maintain should be apparent in the zooarchaeological
data, as well.
On the basis of ethnographic analogy, several scholars
have argued over the past twenty years that transhumance
pastoralism should be identifiable on the basis of the
presence or absence of different age classes of domestic
animals (e.g. Arnold and Greenfield 2003; Geddes 1983;
Grant 1991; Greenfield 1999a; Nisbet et al. 1991).
Transhumant pastoralism involves the movement of herds
as part of a regional subsistence strategy. As a result,
there should be a complementary pattern in season of
death of these animals between highland and lowland
sites. The transhumant movement of domestic stock is
predictable in a mountainous temperate environmental
zone. The herd will move into highland pastures in the
early spring, soon after lambing/calving occurs and
returns to the lowlands during the autumn. In a
subsistence economy, the age groups that are slaughtered
in the highlands and the lowlands will be different.
Therefore, animal remains should be particularly relevant
to answering the question.
The hypotheses can be aptly summarized to propose that

transhumant pastoralism appears at the beginning of the
Neolithic, the beginning of the Post Neolithic, or in later
periods with the emergence of complex urban societies.
Therefore, the appearance of transhumant pastoralism in

a region may be identified through the initiation of
complementary culling practices between highland and
lowland settlements. For example, if transhumant
pastoralism appears in temperate southeastern Europe at
the advent of the Post Neolithic, then a complementary
seasonal culling pattern will begin to be apparent
between from this period onwards between highland and
lowland sites. The previous periods should have little or
no evidence of complementarity. This hypothesis
presupposes that herds were largely absent from parts of
the region for part of the year. The null hypothesis is,
therefore, that transhumant pastoralism was not
practiced. The null hypothesis can be accepted if herds
were resident year-round in each region. This would be
demonstrated by a random culling pattern between
highland and lowland sites.
III. Method and Technique

Can transhumant pastoralism be identified from faunal
remains? Several decades ago, Fleming (1971) argued
against the identification of pastoralism using archaeological
material. Most scholars in the intervening years, in contrast,
agree that zooarchaeological (animal remains from
archaeological contexts) material can be a suitable base for
testing hypotheses about animal exploitation strategies (e.g.
Barker 1985; Davis 1987; Hesse and Wapnish 1986;
OConnor 2000; Reitz and Wing 1999).
This economic adaptation probably arose as a result of
short term movement of animals between highland and
lowland pastures by shepherds embedded within villages.
2
THE ORIGINS OF TRANSHUMANT PASTORALISM IN TEMPERATE SOUTHEASTERN EUROPE

Two techniques within zooarchaeology will be utilized in
this investigationtooth wear and eruption and dental
cementum analysis. These two sources of data are used
here because they are appropriate for testing the above
hypothesis concerning herd movements.
often juxtaposed in close proximity. This is the kind of

environment in which one would expect transhumant
pastoralism to exist. Historically, transhumant
pastoralism is part of the regional subsistence system
(e.g. Cviji 1918; Dedijer 1916). Hence, it is logical to
expect that this practice extended back in time. Also, it is
one of the few regions which have experienced
transhumant pastoralism in Europe that have easily
available archived zooarchaeological data from a variety
of environments (including highland and lowland).
First, harvest profiles are created from the analysis of the

tooth wear and eruption data from the mandibles of
domestic animals, specifically, sheep (Ovis aries), goat
(Capra hircus), cow (Bos taurus) and pig (Sus scrofa
dom.). This technique provides information on both age
and season of death information for each species. If
transhumant pastoralism was present, then the tooth
eruption and wear patterns from archaeological teeth of
the youngest age classes will show complementary
slaughtered age groups between highland and lowland
regions. Those ages that are slaughtered in highland sites
will be missing from the lowland sites and vice versa.
The null hypothesis is therefore, if transhumant
pastoralism was not practiced, all age groups will be
present in both areas.
B. Temporal span
Data from the beginning of the Early Neolithic through
the Early Iron Age of the central Balkans will be
evaluated. This time range was selected because all
previous research indicates that the origins of
transhumant pastoralism in the region occurred within
this time span. Therefore, the time range has to extend
beyond the point where transhumant pastoralism occurs
in order to be identified. It is necessary to include periods
where it probably did not yet exist in order to pin point
the appearance, rather than simply the existence, of
transhumant pastoralism.
Second, the results of the tooth wear and eruption

analysis are supplemented by cementum analysis of
mandibular Ovis/Capra teeth to provide additional
seasonality estimates. If transhumant pastoralism was
present in temperate southeastern Europe during the Post
Neolithic, then the season of death information will show
complementary season of occupation of highland and
lowland sites. The null hypothesis is that if transhumant
pastoralism was not practiced, then the data will not show
a complementary pattern of seasonal culling between
highland and lowland sites.
C. Data
There are two sources of data in this investigationancient
and modern. The ancient data from the region are used to
directly test the above hypotheses. The modern data were
collected and used as a control for the application of both
techniques (above) used to test the hypotheses.
The ancient data can be divided into two types:
computerized database and limited archaeological tooth
material. Greenfield collected a large database on the
zooarachaeology of the region between 1977 and 1994.
These data are computerized and archived at the
University of Manitoba.
While not new, techniques such as tooth wear and

eruption and cementum analyses have not previously
been applied to this problem at the same time. The
mandibular remains of four species of domestic animals
(Ovis aries, Capra hircus, Bos taurus and Sus scrofa) are
examined for evidence of transhumant movement. The
first three species are believed to have engaged in such
movements, while the latter species is included as a
control, as pigs are not commonly herded in a
transhumant manner.
The identification stage of analysis, including species

determinations from all of the sites to be presented here,
have already been completed and presented elsewhere
(e.g. Greenfield 1986a, 1994, 1997, 2005b, in press a, b;
Greenfield and Fowler 2002, 2005). These data were the
basis for his previous analyses (Greenfield 1986a,
1988a, 1999a, 2001a) to identify the advent of
transhumant pastoralism. Greenfields research lumped
together the cranial and post-cranial data. However, the
cranial material, specifically the mandibles and teeth,
were not given the separate attention they deserve.
Many researchers (e.g. Payne 1973; Grant 1975, 1982;
Crabtree 1982) would argue that mandibular data
provide the most accurate age information for
archaeological material since it is less affected by
differential preservation and recovery. Mandibular tooth
eruption and wear data will provide better control over
both age and season of death information for each
species.
IV. Data
A. The region
In order to fill the lacunae in our knowledge concerning
the emergence of transhumant pastoralism, the above
hypotheses will be tested by re-examining the evidence
from one regionthe central Balkans (central region of
the northern Balkans in southeastern EuropeFigure
1.1). This region was chosen because it was hypothesized
by Greenfield to be the first region in temperate Europe
(north of the Mediterranean littoral) to experience the
advent of transhumant pastoralism. The region is
appropriate for this type of analysis because of the nature
of its topography. Both highlands and lowlands exist,
3
HUNGARY
Da
nu
be
Mures
ROMANIA
Vojvodina
Foeni-Salas
Ti
sa
Drava
Danube
BOSNIA
AND
HERZEGOVINA
Sava
Transylvania Alps
Opovo
v
Vinca
Danube
Belgrade
Novacka Cuprija
Petnica
Ljuljaci
Kadica
Brdo
Stragari
Blagotin
BULGARIA
a
av
or
YUGOSLAVIA
Montenegro
Livade
Balkan
Mts.
Kosovo
Adriatic Sea
THE FORMER YUGOSLAV REPUBLIC OF MACEDONIA

Republic boundary
Autonomous province boundary
ALBANIA
River
Modern city
Megalo Nisi
Galanis
Archaeological site
Scale 1: 3,550,000
50km
Pindos Mts.
GREECE

This study presents the analysis of the mandibular tooth
wear and eruption data of domestic animals, specifically,
sheep (Ovis aries), goat (Capra hircus), cow (Bos taurus)
and pig (Sus scrofa dom.), from the central Balkan
assemblages previously analyzed by Greenfield. It also
includes new and unpublished data from previously
unreported sites. It is expected that the specific focus on
this material will provide more concise data than has been
provided by previous analysis. It will also expand the
regional coverage of the study of transhumant pastoralism
by including a larger sample of sites. Greenfield also
brought back from his fieldwork selected samples of
animal bone remains from many of the sites, including
mandibular tooth remains. A portion of these remains
were sectioned for cementum analysis in order to
determine seasonality of slaughter, and by implication,
season of occupation of sites. It is hoped that this would
allow for determination of the presence of transhumant
herds at sites. The University of Manitoba thin sectioning
laboratory provided access to the equipment and training
necessary for this investigation.
The proposed results of this research should reveal the

temporal origins of transhumant pastoralism in a
temperate environment, specifically in the northern
Balkans. This will have a significant impact on the
models for cultural development in this region. It is not
until the origins of transhumance are established that its
role in the evolution of complex systems of land use in
temperate climate zones of southeastern Europe can be
fully understood.
In this book, both tooth wear and eruption and cementum
analysis will be used to test for the advent of transhumant
pastoralism in the central Balkans. Both tooth wear and
eruption and cementum analysis have been critically
evaluated for their utility in age and season of death
determination within the archaeological literature. Each
has problems which can affect the overall analysis. For
example, tooth wear and eruption can be subject to error
due to variations in diet, nutrition and health of the
animals (Monks 1981: 189). These same problems may
be relevant to cementum analysis, as the same factors
may affect the deposition of cementum (Lieberman
1993a, b). Cementum analysis has also been criticized as
being too subjective (A. Stutz, pers. comm., 2000). It is
hoped that through the utilization of these techniques in
conjunction with each other, some of the difficulties that
limit their use may be overcome. If the application of this
methodology proves successful, it should end the debate
as to whether transhumant pastoralism can be revealed
through archaeological remains. The application of these
zooarchaeological techniques to the question of pastoral
movements in prehistory promises to yield valuable
information, given appropriate samples. If the application
of this methodology is successful, it may be extrapolated
to increase understanding of other cultures with a
transhumant pastoral basis. As a result, its methodology
can be applied to other areas of the world, to cultures that
have a similar economic basis. Understanding the nature
of pastoral patterns in an area will have significant
implications for the understanding of behavioral and
social patterns of the culture. This discussion continues in
the next chapter with a summary of previous approaches
to the archaeological investigation of transhumant
pastoralism in Europe, including both the Mediterranean
and the northern temperate region of the Balkans.
The modern data consists of comparative data from

modern livestock breeders in Manitoba, including Ovis
aries and Capra hircus mandibles with teeth. These data
were collected by Arnold. Tooth wear and eruption data
were recorded on all mandibles and a sample of teeth was
selected for cementum analysis. A modern control sample
is necessary to establish the time of formation of
incremental growth structures in the cementum in order
to apply this information to the study of an archaeological
sample (Burke and Castenet 1995). It is hoped that the
cementum analysis will provide additional seasonal
information to the question of the origins of transhumant
pastoralism.
V. Conclusions
Transhumant pastoralism is hypothesized to become a
major element of the subsistence strategies of cultures in
southeastern Europe during the Post Neolithic. Its
adoption is proposed as the essential ingredient for the
significant shifts in economic organization, settlement
patterns, and social and political organization that occur
during this time (Greenfield 1988, 1999a, 2001b).
CHAPTER 2
TRANSHUMANT PASTORALISM: SOME ETHNOGRAPHIC CONSIDERATIONS
outbreaks, localized drought, the number of people
involved in the herd movement and relations with nonpastoral groups. The mobility of livestock does not
always correlate with human mobility. The proportion of
the human population traveling with the transhumant
herds can vary culturally, but also in response to a variety
of political, economic and ecological circumstances
(Niamir-Fuller and Turner 1999: 22). The species
composition, the age and sex structure of herds are
determined by several factors. Major domestic species
used by pastoralists include sheep and goats, cattle,
camel, horses, reindeer, llamas, and alpacas. The biology
of the species and geographic conditions will be the
primary determinate. These same factors will affect
whether animals are utilized for primary products (such
as meat), or for secondary products (such as wool, milk,
or traction). Additionally, it is important to realize that
economic, social, political and cultural factors are also
influential (Khazanov 1984: 27).
I. Introduction to pastoralism
One of the problems in investigating the origins of any
type of pastoralism is the lack of any clear or agreed upon
definition of pastoralism, generally, and transhumant
pastoralism specifically (Bartosiewicz and Greenfield
1999; Khazanov 1984; Nisbet et al. 1991). Pastoralism is
both a land use strategy and a system of animal
production (Krader 1959: 499). It is a distinctive form of
human subsistence economy in which domestic animals
play a predominant, but not an exclusive role in the
shaping of the economic and cultural lives of the people
who depend on them (Galaty and Johnson 1990).
There is a wide spectrum in the forms of pastoralism
(Ehlers and Kreutzmann 2000: 13). This is due to a range
of factors and can include the quantitative and qualitative
characteristics of the herds, the extent and range of
mobility, degree of inclusion of agricultural products,
environment and ecological aspects of the region and the
extent of ties with an external market (Logashova 1982:
53). In this chapter, we will discuss the various forms of
pastoralism that are suspected to have relevance to the
archaeological record in the region.
A. Nomadic pastoralism
Nomadic pastoralism (also known as pure pastoralism)
has been characterized by the absence of agriculture,
even in a supplementary capacity (Khazanov 1984: 19).
However, some researchers (e.g. Whittaker 1988)
maintain that nomadic pastoralism never existed in a pure
form, in other words with the total absence of agriculture.
There is always a spectrum in the relative importance of
one towards the other (Whittaker 1988: 1).
II. Types of pastoralism

Ethnographic research on pastoral societies provides
some general elements of a pastoral economy and society.
In general, pastoralism is grouped into two basic types,
nomadic pastoralism and semi-nomadic pastoralism. On a
very basic level, both types of pastoralism can be defined
as the movement of domestic herds between seasonally
complementary pastures. Two basic variables are used to
distinguish the types of pastoralism: the degree of
mobility of the human community accompanying their
herds (Ehlers and Kreutzman 2000; Rafiullah 1966), and
degree of reliance upon animal and agricultural products
(Khazanov 1984: 19). Further distinctions arise from
consideration of the spatial (geographical and altitudinal)
relationship between herd pastures (Chang 1993: 709).
The choice of the variable to be emphasized often leads
to confusion and contradictions in the type of pastoralism
being discussed.
Nomadic pastoralism involves the movement of people

and animals within a large and defined geographic area
according to a set schedule. This is an economic
adaptation where the major economic orientation of the
culture relies upon domestic stock (Barth 1961). In
nomadic pastoralism, all or the majority of the human
population migrates with their domestic herds year-round
within a system of pastures and do not have a fixed (i.e.
sedentary) settlement (Chang 1993: 709). These
pastoralists are highly mobile, move over vast areas, and
are characterized by the absence of, or minimal
investment in agriculture. This is the form of pastoralism
has been long practiced by the Basseri tribe of South
Persia (Barth 1961). Their migratory pattern involves
movement between arid steppes and mountainous
environments in order to utilize extensive but localized
pasturelands for their herds.
The mobility of pastoral groups is not limitless.

Constraints, such as geographic knowledge, social
contacts, and access to resources (such as pasture), often
result in pastoralists revisiting encampment points on an
annual or semi-annual basis. The mobility of livestock
has both spatial and temporal elements. A variety of
factors will affect the distance, timing and location of the
movement of livestock. These can include the location
and seasonal changes of water and pasture, disease
B. Semi-nomadic pastoralism
The difference between nomadic and semi-nomadic
pastoralism is the degree of mobility of the human
population (and settlement). In semi-nomadic
7
TRANSHUMANT PASTORALISM
from movement of the entire community, to no movement
of the community but rather solely movement of animals
under the supervision of professional shepherds (Rafiullah
1966: 6). It is possible that the same groups in a given
society (or sub-society) are occupied with both agriculture
and pastoralism, or conversely, there are specialized
groups that devote themselves primarily, or even
exclusively, to pastoralism, in conjunction with groups
which are primarily occupied with agriculture (Khazanov
1984: 20). Transhumant pastoralism may also involve
specialized groups within a community. The majority of
the population may be involved in agriculture or other
economic activities, such as hunting or gathering, while a
specific group within the community is focused on the
maintenance of herds as a specialized occupation. The
existence of variation is also supported by recent
ethnographic investigations, which have shown that in
modern transhumant economies, shepherds are wage
labourers hired by livestock owners. This is in contrast to
modern nomadic pastoral groups, where relatives manage
their personal resources. This further distinction of
transhumance from other forms of pastoralism in recent
ethnographic literature stems from a focus on the
relationship of the shepherds to the animals (Ehlers and
Kreutzmann 2000: 16).
pastoralism, the herd moves between fixed locations

(winter vs. summer pastures), but the human population
has less residential flexibility and a lower degree of
mobility over space and through time in comparison to
nomadic pastoralism (Chang 1993: 709). Semi-nomadic
pastoralism is characterized by an economy where
pastoralism is the predominant activity but includes
varying emphasis on agriculture as a supplementary
activity. All elements of a semi-nomadic pastoral
existence, including species composition, routes and
timing of migration, will be affected by even limited
investment in agriculture. Within semi-nomadic pastoral
groups, two main alternatives are observed. It may be that
the entire population in a given society is involved in
both agriculture and pastoralism. Alternatively, there are
specialized groups within the society that devote
themselves primarily, or even exclusively, to pastoralism,
alongside groups that are primarily occupied with
agriculture (Khazanov 1984: 19-20).
C. Transhumant pastoralism
There are two main types of movement in all types of
pastoralism: vertical and horizontal. Length of
transhumant migrations vary highly depending upon
topography, climate, political conditions, and other
variables. These periodical movements may involve
journeys of several hundred kilometers or only a few
kilometers (Rafiullah 1966: 5; Walker 1983: 37).
Horizontal (or lateral) pastoralism occurs in large flat
basins and steppes and involves the movement between
spatially differentiated pastures within the same
altitudinal zone. In this situation, herds are moved
laterally within the same altitude zone in order to exploit
seasonally available resources. Vertical (or transhumant)
pastoralism occurs between altitudinally differentiated
pastures, typically located in mountainous and highland
zones. In this instance, herds are moved up and down
mountains in order to exploit seasonally available
resources at different altitudes. In transhumance, the
seasonal migration of domestic herds occurs between
summer pastures in the mountains and winter pastures in
the lowlands (Ehlers and Kreutzmann 2000: 16). Both
types of movement are apparent for Mediterranean and
temperate southern Europe. Vertical pastoralism is found
under conditions of topographic variability (e.g. in the
mountainous areas of the Alps and Dinarics). Lateral
pastoralism is found under conditions of little
topographic variability (e.g. on the steppes of Eastern
Europe and in the Hungarian/Pannonian PlainMatley
1970; Szabadfalvi 1968; Vincze 1980).
Historically, transhumant pastoralism has been, and is, a

significant part of the economy in the Mediterranean and
the Balkans. Even today, seasonal migrations of herds
and herdsmen takes place on a fairly large scale in
different parts of the Balkans and neighboring countries
(i.e. Albania, Bulgaria, Greece, Romania, and the various
countries of the former Yugoslavia) (Bartosiewicz and
Greenfield 1999; Rafiullah 1966). Many researchers
(Halstead 1981; Geddes 1983; Greenfield 1986a, 1991,
1999a, 2001a) agree that it was also an important element
of the economy in prehistoric times (Harding 2000: 138).
As a result, this investigation will focus on a specific type
of semi-nomadic pastoralism, that is, transhumant
pastoralism, rather than nomadic pastoralism.
III. Aspects of modern transhumant pastoralism in
the northern Balkans
Descriptions of the traditional way of life of transhumant
shepherds are far less abundant in the ethnographic and
historical literature of the northern than southern (along the
Mediterranean littoral) Balkans. The dividing line is the
Dinaric Mountain. In the more arid Mediterranean littoral
(including the Adriatic coast) or alpine environments, the
advantages of pursuing transhumant strategies are more
obvious than in temperate climatic zones. In the arid zones,
stock is moved from lowland to highland pastures to
escape the devastating summer heat and to secure adequate
water and pasture. This would be the case in the southern
Balkans or along the Mediterranean littoral. Similarly,
alpine communities would move stock back up to the
mountains so that their products may be more efficiently
exploited. The herds of both would be moved down the
mountains during the autumn to avoid the freezing
Transhumant pastoralism is part of a more broadly based

economic system, which incorporates crop cultivation and
transhumance in a single economic scheme (Geddes 1983:
51). The practitioners of this type of pastoralism, oftentermed mixed or specialized pastoralists (Cherry 1988: 7),
are often semi-sedentary (Geddes 1983: 51). In this
situation, they would be semi-nomadic pastoralists. The
degree of mobility for the human populations can vary
8

temperatures and precipitation in the mountains, while
taking advantage of the abundant winter grass growth in
the lowlands (Bates 1973; Hole 1978; Rhoades and
Thompson 1975; Sterud 1978).
villages. Highland shepherds practice vertical

transhumance as a means to avoid variations in climate
and in order to economically exploit two environmental
zones that would otherwise be unable to support livestock
(Nandris 1975; Sterud 1978; Zdanovski 1954). This is the
environment that presents the greatest difficulties in terms
of herd survival. Highland winters are severe and over
wintering creates great challenges for the herd and the
herder. It is also the environment most subject to
overgrazing (and consequent erosion), and the need for
shorter stays (Matley 1968; Zdanovski 1954).
In the temperate environment of the northern Balkans, the

benefit of doing so is not as apparent. Pastoralists in nonalpine temperate environmental zones may herd their
animals up into the mountains during the summer to
avoid the greater heat and humidity in the lowlands and
to take advantage of delayed plant growth, and to access
the potentially a higher quality summer graze in the
highlands. However, the need to do so is not as great.
Temperature extremes in the lowland are not sufficiently
extreme during the summer to drive lowland-based
livestock into the mountains in search of grazing and
water. Sufficient water and graze are available year-round
in most low and mid-altitude pastures in temperate
climatic zones. In the northern Balkans, sufficient
microenvironments exist in and close to the lowlands,
such as marshes, streams, hills and plains, to allow for
stock to be safely herded throughout the year in the
lowlands. As a result, ecologically, there are fewer
incentives for pastoralists from low and mid-altitude
settlements in temperate regions to practice
transhumance.
Under conditions of vertical transhumance, two different

strategies of settlement are found: normal and inverse. In
normal transhumance, a permanent village base is
maintained in the lowlands. In inverse transhumance, the
permanent base is maintained in the mountains. The
former is more common in the Mediterranean zone, while
the latter is more common in the temperate zone
(AlpineSterud 1978: 389). In this way, herders are able
to spend a substantial amount of time close to home. A
portion of the village population tends to be absent
seasonally from the village since they (often men and
boys) accompany the herds in their seasonal migrations.
Studies of traditional pastoralism in temperate southeast
Europe may be generalized into five basic types of herd
movements and village settlement (cf. Matley 1970;
Nandris 1985, 1991; Vincze 1980):
Lowland based herds do not have to be moved between

altitudinally differentiated and seasonally complementary
pastures in the search for graze and water. Instead, they
are moved laterally. Even today (and in the ethnohistoric
past), most livestock in the temperate lowlands are (were)
grazed locally, rather than moved between distant
highland and lowland pastures (Gal and Gunst 1972;
Halpern 1967; Szabadfalvi 1968; Vincze 1980). Forests
(and clearings), marshes and swamplands served as
pastures in the winter and early spring. In winter, the
snow was thawed by the high temperatures of the
decaying, steaming boggy soil enabling animals to find
forage. This was such a favorable environment for winter
that highland herders often sought them out (Szabadfalvi
1968: 145). This system was used for cattle, sheep, goats
and horses, whereas pigs were often left to roam freely
through nearby forests (Halpern 1967; Halpern and
Kerewsky-Halpern 1986; Szabadfalvi 1968). In the
summer, lowland based herds are often brought to nearby
hills or to non-cultivated steppe environments (if hill
country is lacking). In drier environments, the
geographical range of annual migration is wider in order
to access graze and water resources since they are often
spread out farther (Gal and Gunst 1972; Szabadfalvi
1968; Vincze 1980).
1.
2.
3.
4.
5.
The highland and lowland zones present different

opportunities (and trials) for lowland and highland based
temperate zone pastoralists (Sterud 1978; as well as
Mediterranean pastoralistsKoster and Koster 1976).
While the lowland based herder is not necessarily driven
to a transhumant lifestyle in the temperate zone, the same
situation does not exist for herders based in highland
Village is located proximate to both lowland and

highlands. There will be localized herding from a
village base, without overnight stays. There are no
pens or huts away from the village.
Village is in or surrounded by mountains. There will
be localized herding on summer pasture near the
village, and wintering in the village. There are pens
and huts in the summer pastures.
Village is in the lowlands. Herds will be moved to
summer pastures in the mountains, but wintered in
the village. There are pens and huts in the summer
pastures.
Village is in the lowlands. Spring pasturing will take
place near the village, during which arable land is
manured. Summer pasturing will occur in the
mountains. Herds will be wintered outside of the
village, often near the edge of forests or in swamps.
There may be pens and huts in both the highlands
and lowlands.
Herds are not village based. In this situation, herds
will be moved by professional transhumant
shepherds from lowland winter to highland summer
pastures.
In terms of seasonality of movement, the following

model can be postulated. Winter settlement is in the
lowlands and the herd moves into the mountains in late
April or early May. The herds return to the lowlands in
late September or early October (Popovi 1971; Ryder
1999: 191; Sterud 1978: 391-392; Szabadfalvi 1968;
9
and Koster 1994; Grsan-Salzman 1984; Vinak 1988).
Lowland based herders often retain seasonal huts or
lodgings in the highlands (Lockwood 1975; Vincze 1975:
397-8; Vinak 1999).
Vinak 1983, 1988; Zdanovski 1954: 121). The exact

time of the movements depends largely on environmental
conditions such as the melting of mountain snows in
spring and the onset of bad weather in the autumn.
Originally, entire families made the seasonal migration,
while today it is largely the work of men (Dedijer 1916;
Grsan-Salzman 1984; Ryder 1999: 191).
Traditionally, transhumant pastoralists were integrated

into the communities through which they passed, by
networks of exchange and lineage (marital, filial
relationships, etc.Antonijevi 1982; Chalkea 1999;
Lockwood 1975; Vincze 1983). These ensured access to
pastures and supplies, and entailed various social
obligations. But, it also meant that herders were never
isolated from the mainstream of village life in the
surrounding countryside.
In terms of specialization of herds, the production of milk

(stored as cheese) is the major focus of modern
transhumant herds (Ryder 1999: 190). This is followed
by meat and wool. In the 17-18th century, 65% of the
revenues in Bosnia came from milk, whereas only 1013% came from wool (malcelj 1954: 203). In modern
times with the widespread appearance of synthetic
materials and importation of cotton, the exploitation of
wool has further declined throughout the region. Shifts in
meat consumption have also affected the nature of milk
exploitation. For example, only goats are milked in
Dalmatia and most of their milk is used for making
cheese and for shepherds food. In contrast, sheep (ewes)
are not milked. Instead, their milk is given to the lambs in
order that the lambs can gain weight faster. The weight of
the lamb affects its sale pricethe heavier the lamb, the
more it will yield in the market place and the more the
herder will reap (Vinak 1988). This is a reflection of
the shift in sheep exploitation away from wool toward
meat. The same is not true for Romania, where sheep
were milked as well. However, sheep are milked less in
order to allow lambs to rapidly gain weight (GrsanSalzman 1984). Among herds in Bosnia, ewes lactate
from 5.5-7.0 months. They tend to finish lactating and the
lambs are weaned by the end of September (Dracun 1954:
257). Historically, however, sheep and goats were
exploited for their full combination of potential
productsmilk, meat, hide and wool (Gal and Gunst
1972: 10, 38; Seri 1956).
During 1985 (and subsequent visits in 1987-88),

Greenfield was able to visit and interview shepherds in
the area of the Glasinac Plateau (E. Bosnia) near the site
of Kadica Brdo about contemporary patterns of animal
husbandry in the area. In combination with ethnographic
and ethnohistoric studies of pastoralism in the region and
from nearby areas (e.g. Antonijevi 1982; Halpern 1967;
Lockwood 1975), is apparent that modern practices are as
complex as are the hypotheses (and implications) of the
prehistoric economy. There are two patterns that are
apparent: long distance and local movements. These are
similar to those found even among lowland villages
(Vincze 1974, 1980).
Most domestic animal production in the highlands
involves sheep and goats, with some cattle (see Plate 1).
Locally, few farms raise pigs, but this may reflect the
strong Moslem influence exerted over the region until the
latter half of the last century. Cattle are grazed locally and
brought home every night in order to milk them. Cattle
are stalled and fodder fed through the winter. Almost all
transhumance is conducted with ovicaprines (sheep and
goats).
Cattle exploitation depended upon whether the animals

were part of migratory herds or was stabled locally. It
was only in locally stabled herds that cattle were
intensively milked and used for draught (Gal and Gunst
1972: 10). With the shifts in transport and opening of
new markets, cattle production undergoes profound
changes. While previously, cattle were bred with an eye
to their exploitation for draught potential, meat and milk
production become the principal aims (Gal and Gunst
1972: 30). With the appearance of the tractor as a normal
part of local agricultural industries, the draught potential
of cattle further declined (Mileusni 1987). The same is
not true in Hungary with respect to sheep meat. There
was an aversion to sheep meat, while wool remained the
major product for herders (Gal and Gunst 1972: 38). It is
apparent that market conditions and local tastes affect
which products are exploited. Transhumant pastoralists
are summer visitors to the mountains of temperate SE
Europe. In the past, they walked (or rode horses or mules)
to and from pastures, moving with their flocks. Presently,
many use vehicles to truck their animals and supplies
across the region (e.g. Baker 1999; Chalkea 1999; Chang
Local farmers tend to move their ovicaprine flocks

around the region in order to take advantage of seasonally
available pasture and/or fodder. Locally moved herds are
small in size and there is enough natural pasture and
fodder to support the animal population. Some fodder
crops are also cultivated to supplement winter supplies,
mostly for the stalled cattle. Of the few sheep kept in the
highlands during the winter, many of these animals are
eaten through the winter as fodder supplies dwindle.
Transhumant pastoralists bring flocks into the mountain
basins during the summer and return to the lowlands (via
the Drina) for the winter. Local herders will often buy
transhumant stock in the lowlands at the end of winter
(when they are cheapest), then bring them to highlands
(e.g. Glasinac) during the spring. They are fattened over
the summer and slaughtered during the fall/early winter.
This avoids the costs involved in supporting large
resident herds throughout the winter within the basin and,
at the same time, satisfies local demand for meat without
causing undue pressure on local graze. An interesting
10

subjective observation offered by several informants is
that the quality of meat (as measured by taste) from
animals who annually migrate is considered to be poorer
than that of locally bred stock. This may be the result of
the increased toughness in the meat of animals who are
more active (as would occur in transhumant populations),
thereby lowering taste value (cf. Jochim 1976).
IV. Conclusion
From the above discussion, it is obvious that a common
element of all pastoral societies is the mobility of
domestic herd animals. This movement of livestock has
been noted historically as a method to reduce uncertainty
and risk in the environment. The mobility of pastoral
herds functions to: 1) take advantage of spatially and
temporally variable ecosystems, 2) take advantage of
unpredictable resources, 3) utilize pastures distant from
settlements, 4) make use of seasonal pastures, and 5)
provide fodder for livestock at minimal labour and
economic cost (Niamir-Fuller and Turner 1999: 21).
The evolution of the large scale long distance

transhumant adaptations characteristic of the area
(Antonijevi 1982; Seri 1956; Sterud 1978) seem to be a
function of the appearance of large urban markets and
productive specialization (Halpern 1999). Toward the end
of the 18th century, the traditional pattern of agriculture
is disrupted with the development of new modes of
transportation. With the creation of extensive canals in
the Hungarian lowlands, large quantities of grain could be
moved inexpensively from producers to consumers. This
trend is accelerated with the appearance of railroads in
the 19th century (Gal and Gunst 1972: 8, 18). Prior to
this, cattle could be moved to distant markets on the hoof,
but it was not economic to move animals such as pigs.
Pigs, sheep, and horses were raised and mostly consumed
locally. Sheep and pigs were raised as food animals and
in order to provide clothing, while horses were raised for
riding and draught (Gal and Gunst 1972: 8). From this
point in time, we begin to see large scale movement of
pigs to distant markets (Halpern 1999) and a shift in
emphasis from cattle to sheep production, especially on
the large estates. The large estates turn to sheep
production as a consequence of increases in wool prices
and the need to raise cash to pay land taxes. There are
fewer changes in the small-scale peasant land holdings
(Gal and Gunst 1972: 8).
Additionally, the movement of domestic herds also

protects against problems such as disease outbreaks and
droughts (Niamir-Fuller and Turner 1999: 22). These
functions have been noted for arid environments, where
the majority of pastoral economies exist. But, they are
rarely taken into consideration in more temperate
environments, such as will be investigated in this work.
In sum, transhumant pastoralism is a risk reducing
economic strategy for increasing ones survival in an
inherently unstable environment. Environments are never
stable, changing between different political, social,
economic, and/or climatic conditions. Transhumance
reduces risk by allowing people to shift their focus from
immobile (agricultural) to mobile (animals) resources and
utilize the same landscape in different ways (Sterud 1978:
383). It is for this reason that transhumance has survived
as a widespread practice throughout the Mediterranean
and temperate zone of southern Europe (Bartosiewicz and
Greenfield 1999; Nisbet et al. 1991).
11
Plate 1: Photograph of traditional pastoralists on the Glasinac Plateau,

eastern Bosnia (1988). Photograph by Haskel J. Greenfield.
12
CHAPTER 3
PREVIOUS RESEARCH ON ORIGINS OF TRANSHUMANT PASTORALISM
IN PREHISTORIC SOUTHERN EUROPE
patterns and material culture then become evidence of
similar shifts toward a more pastoral economy. For
example, the shift to a more ephemeral settlement pattern
in the Eneolithic is frequently interpreted by prehistorians
as evidence of the migrations of the steppe peoples (also
known as Kurganse.g. Bukvi 1979; Brukner,
Jovanovi, and Tasi 1974). It is assumed that the local
economies became more pastorally oriented and
transhumance arrived with nomads from the eastern
steppes. However, there are frequent confusions between
nomadic and transhumant pastoralism (cf. Milisauskas
1978) and between specialized and mixed agro-pastoral
economies (cf. Lees and Bates 1974). Additionally, the
evidence for political instability or increased pastoralism
in the economy is not assured. In the prehistoric
archaeological record of temperate southeast Europe,
there is no unambiguous evidence for early migrations,
warfare and the kinds of state-sponsored politicaleconomic instability so characteristic of the past
millennium (Bankoff and Greenfield 1984; Greenfield
1986a; Sterud 1978).
I. Introduction
Transhumant pastoralism has long been a significant part
of the economy in the Mediterranean and the Balkans. It
has been well documented from Classical through
Medieval times and into the modern era (e.g.
Bartosiewicz and Greenfield 1999; Cherry 1988;
Hodkinson 1988; Nisbet et al. 1991; Whickham 1985).
However, there are strong disagreements and much
confusion between researchers when the data are pushed
farther back in time. In this chapter, we will review and
discuss the various archaeological theories that have been
proposed for the origins of transhumant pastoralism,
review some of the studies that have attempted to identify
it prehistorically, and present a series of hypotheses that
can be used for testing its presence with
zooarchaeological data.
II. Theories on the origins of transhumant pastoralism
There are a plethora of theories that have been proposed
to account for the origins of transhumant pastoralism. It is
possible to group them into four major categories. Most
of them are focused on the exploitation of agriculturally
marginal environments. Some have applicability to
temperate climates, such as southeast Europe. Each of the
major theories is summarized and their applicability to
the archaeological record of the central Balkans evaluated
next.
B. Regional symbiosis
The development of symbiotic relationships between
groups and/or regions (cf. Barth 1956) has been another
widely cited hypothesis. More complex adaptations to an
environment are made possible through the specialization
of productive relations. Goods and services, lacking in
one group or region, are provided by another.
Transhumance provides the modus operandi by which
goods are moved between highland and lowland areas
(Lees and Bates 1974; Bates and Lees 1977; Chapman
1981; Moreno Garca 1997, 1999).
A. Political instability
Political instability has been the most popular perspective
among prehistorians working in the Balkans (and Eastern
Europe, in general). In this view, the agriculturally
marginal highland zones were colonized by transhumant
pastoralists as an adaptation to political instability. Herders
use their migratory patterns to flee from invading armies,
tax collectors, etc. The unstable political conditions
existing in the Balkans (during the past millennium) and
consequent detrimental effects upon agro-pastoral systems
of food production are frequently cited to explain why
transhumant pastoralism has been an important adaptation
in the region (e.g. Brice 1974; Navy 1945).
This concept has been applied to two different periods in

the Balkans. In earlier periods, such as the beginning of
the Post Neolithic, the development of larger scale
metallurgical extraction and the need to control both
mineral/metal sources and routes of movement have been
hypothesized to create an impetus for the colonization of
agriculturally marginal highland marginal zones
(Greenfield 2001b; Sherratt 1976). There is no evidence
for the sponsorship of settlement in the highlands by
lowland groups to warrant support for this model. In the
Balkans and other parts of the Mediterranean, this
theorem is frequently linked with the emergence of the
productive specialization of large-scale complex
societies. These begin to emerge in the Late Bronze Age
of the Aegean and become paramount in Classical
antiquity (Cherry 1988; Fotiadis 1980; Halstead 1981,
1996; Lewthwaite 1981, 1984; Walker 1983).
Balkan prehistorians have frequently seized upon the

historical association often existing in the Balkans
between the decline in agricultural, increased investment
in pastoralism, appearance of long-distance transhumance
and political instability to extrapolate similar conditions
back into the past (e.g. Garaanin 1972, 1973, 1983;
Gimbutas 1965, 1973). Shifts in prehistoric settlement
13
PREVIOUS RESEARCH ON ORIGINS OF TRANSHUMANT PASTORALISM IN PREHISTORIC SOUTHERN EUROPE

and production strategies (including new domestic
breeds) from the Near East at the onset of the Post
Neolithic (Eneolithic or Early Bronze Age). Domestic
animal exploitation patterns shift from a focus upon
primary products (meat, hide and bone) to secondary
products (wool, milk and traction). It is assumed that herd
sizes increase to more effectively produce secondary
products. At the same time, human populations were also
increasing across the landscape. This would have resulted
in the twin dilemma of increased human and domestic
animal population pressure upon resources and
consequent habitat degradation. The Secondary Products
Revolution model posits that the agriculturally marginal
highlands were colonized at this time in order to avoid
placing strains upon local economies (to produce and
store winter fodder). Herds were moved to highland
pastures for the summer and returned to the lowlands
with the onset of inclement weather.
C. A by-product of initial animal domestication in an

ecologically varied habitat
In this model, hunter-gatherers (i.e. Mesolithic) exploited
both high- and lowlands in the search for seasonally
available resources. As agriculture was introduced into
the region, hunter-gatherers increasingly include
domestic livestock into their subsistence system
becoming herder-hunter-gatherers. Movement patterns
continued as in earlier pre-agricultural periods in the
seasonal round of hunting and gathering, with the
addition of domestic stock. The need to find adequate
seasonal pasture for the livestock begins to dominate and
alter the annual migratory route, rather than the available
wild resources for hunting and gathering. As wild game
dwindled, transhumance was transformed from an
initially advantageous agro-pastoral practice into an
ecologically determined necessity in the maintenance of
protein availability (Chapman 1982; Geddes 1982; Higgs
et al. 1967; Jacobsen 1984; Miracle and Forenbaher
2005). Along the Mediterranean littoral, domestic stock
must be moved out of the lowlands during the summer to
avoid the searing heat and out of the highlands in the
winter to avoid freezing temperatures and precipitation
(cf. Hole, Flannery and Neely 1969). The continued
spread of agricultural fields in the lowlands would
eventually limit the available pasture. As time passed,
herders would be forced to exploit seasonally available
upland pastures to avoid over-exploiting the limited
lowland pastures (cf. Geddes 1982). If protein availability
was to be maintained, this model assumes that
colonization and transhumance are synonymous and byproducts of initial animal domestication. But it does not
deal with the possibility that a long period of time may
have existed during which transhumance was not a factor
in local pastoral strategies (which would also be the
refuting hypothesis in this case).
Several aspects of this model are difficult to test

archaeologically, such as the assumption of population
pressure (Bankoff and Greenfield 1984; Greenfield 1984,
1986a) and of increased herd size (Hesse 1982). There is
no logical reason that the posited shift in production
strategies required growth in the size of individual herds
since modern agropastoralists in the region typically
operate on a small scale. Additionally, the timing of
arrival of some of these new technologies is in dispute
(Chapman 1982). Further, analysis of zooarchaeological
data from the Balkans and other regions do not show the
appearance of specialized secondary products economy at
this juncture, but rather more diversified production
strategies including components of both secondary and
primary products (Arnold and Greenfield 2003;
Greenfield 1986a, 1988, 1989, 1991, 1999a, 2001a).
III. Previous research on the origins of transhumant
pastoralism in Mediterranean Europe
Archaeologically, transhumant pastoralism is often

assumed to be a by-product of initial animal
domestication and does not deal with the possibility that a
long period of time may have existed during which
transhumance was not a factor in local pastoral strategies
(which would also be the refuting hypothesis in this
case). In most parts of southern Europe, there is very poor
evidence for indigenous development or adoption of
agricultural or pastoral lifestyles. This argues against the
notion that transhumant pastoralism evolved out of a
Mesolithic subsistence base. Additionally, in cases where
there is evidence for the entry of Early Neolithic cultures
into a region, there is a dearth of evidence for
transhumance (Halstead 1981, 1996; Greenfield 1993).
There is extensive documentary evidence that

transhumant pastoralism was well established
throughout much of the Classical and Medieval
Mediterranean (e.g. Grassl 1985; Gyni 1951;
Hodkinson 1988; Skydsgarrd 1974; Wickham 1983).
How far back in time can it be pushed? Most early
studies of the origins and nature of pre-modern
transhumant pastoralism in Europe are relatively facile
attempts. Usually they are simply visual associations,
wherein modern transhumant routes are compared to
archaeological distributions. If they are coincident, then
transhumance is used as an explanation for the
archaeological patterns (e.g. Bosch-Gimpera 1944;
Brice 1974; Higgs 1976; Higgs et al. 1967; Sterud
1978). As Chapman (1979) long ago noted for Spain,
the distributions of archaeological material do not
closely coincide with transhumant routes and
particularly termini of the migration routes where
transhumant pastoralists would be expected to spend
much of their time. Relatively little time is spent along
the migration routes, in contrast to the amount of time
D. Secondary Products Revolution

In this model, colonization and transhumance appear as
part of the package deal associated with the Secondary
Products Revolution (Sherratt 1981, 1983a). This model
posits the diffusion of new domestic animal exploitation
14

In the Early Neolithic, there is increased agricultural
activity, including the appearance of cultivated cereals
and legumes. Further, there is an increased ovicaprine
frequency in the zooarchaeological samples. Wild
resources continue to be an important element of the
subsistence economy with wild boar as frequent as the
sheep/goat at all sites. Other wild animals, such as red
deer, roe deer, aurochs and ibex, accompany the domestic
animals in upland sites. Geddes (1983) highlights the
evidence for Early Neolithic seasonal occupations
concomitant with transhumant movements between
upland and lowland sites. Seasonal evidence from
lowland sites specifically indicates an occupation during
the winter and the spring lambing season. Additionally,
the proximity of arable soil suitable for the cultivation of
cereals and legumes suggests occupation for a greater
part of the year. Year-round occupation of the Upper
Aude valley (higher altitudes) did occur in historic
periods. However, it required the cultivation of
supplementary fodder crops and the sheltering of
domestic stock from the rigorous winter cold. As a result,
so it is believed this type of perennial occupation of
upland areas was not occurring in the Early Neolithic
periods (Geddes 1983).
spent in the pastures at the termini. Another pattern of

research is to associate types of tombs (e.g. tumuli) or
cave burials with transhumant pastoralists. This,
however, is an unwarranted assumption since burial
type may reflect shared heritage or class, more than an
economic type such as transhumant pastoralism (Walker
1987; Govedarica 2005).
Recently, a few studies have taken a much more
sophisticated approach. A summary of the previous
research conducted on transhumance in Europe both
within southern Balkans, with its Mediterranean
environment, and the temperate environment of the
Northern Balkans is necessary. Some of the more
intensive ones are described below.
A. Mesolithic/Neolithic continuity
Several studies propose that transhumance appears as a
by-product of initial animal domestication in a varied
ecological habitat (Geddes, 1983, 1985; Voytek 1991).
The most widely cited is Geddes (1983, 1985) research in
the Aude Valley of southern France, wherein he proposes
that transhumant pastoralism appears at the MesolithicNeolithic junction. This region falls into the
Mediterranean climatic regime and provides classic
geographic conditions, such as varied relief, and a
diversity of microenvironments favourable to the
adoption of transhumance as a dominant economic
strategy. He argues that prehistoric hunters and gatherers
and early subsistence farmers would have easily
incorporated domestic migratory stock into their seasonal
exploitation of land and resources. Transhumance has
played an important role in modern and historic times in
this area.
Together, the sites of Grotte Gazel (lowland), Abri Jean

Cros (middle altitude) and Abri de Dourgne (upland) are
argued to demonstrate the appearance of transhumance of
domestic herds during the Early Neolithic. Herds were
moved seasonally into middle and high altitude
environments for the exploitation of pastures. In addition,
these areas continued to be exploited for boar, red and roe
deer, aurochs, ibex and, undoubtedly, wild plant
resources. Increasing dependence on domestic stock
including ovicaprines, cattle and pigs is evident
throughout the period (Geddes 1983).
In the Mesolithic of the Aude valley, the subsistence

economy is largely dominated by wild species including
wild boar (Sus scrofa scrofa), red deer (Cervus elaphus),
and the Pyrenean ibex (Capra pyrenaica). The latter
becomes more common as altitude increases. Other
archaeological material and macrobotanical evidence
suggest that these sites were occupied on a seasonal basis.
Geddes (1983, 1985) proposes that the faunal,
geographical, and archaeological evidence indicates that
ovicaprine transhumance began at this time with the
diverse exploitation of the lowland, middle altitude and
mountain environments as small domestic herds of sheep
were incorporated into the seasonal round. These domestic
animals make up a small percentage of the faunal remains
of these sites in both upland and lowland regions (Geddes
1983, 1985). Recently, doubt has been cast upon Geddes
claim for domesticated sheep in late Mesolithic
depositional contexts. The paucity of supposedly
domesticated remains, uncertain stratigraphic association
of the bones and the difficulty of distinguishing from wild
and domestic forms have undermined his claim that
domesticated sheep were adopted by late Mesolithic
hunters and gatherers (Binder 2000).
B. Complex societies
Halstead (1981, 1985, 1991, 1996) proposes that
transhumant pastoralism only appears with the Late
Bronze Age and Classical complex societies in Greece
(and the Mediterranean littoral) and is associated with
large-scale economic specialization of transhumance. He
argues that the extensive use of upland pastures in a
transhumant manner would not have occurred on a large
scale until the development of extensive agriculture. It is
only when flocks become too large to be fed year-round
in either the lowlands or uplands and large scale
clearance is occurring in both areas that transhumant
pastoralism becomes an element of the subsistence and
economic system. This requires that extensive blocks of
fallow land suitable for grazing be maintained in lowland
regions, as well as an urban market or comparable outlet
for pastoral produce. Halstead argues that these
preconditions begin to exist only from the latter half of
the second millennium BC and in subsequent periods.
Even at this late date, the evidence for extensive
agriculture and sheep herding suggests that these
practices may only have been characteristic of land use
15

Sterud (1978). In this work, he argued for local
transhumance by the Early Neolithic occupants of sites in
Bosnia (e.g. Obre). While he cites the faunal data
collected by Bknyi (1974b) in support of the presence
of transhumance, he does not bring to bear any evidence
for seasonality or complementarity between highland and
lowland assemblages. This is largely because the data are
absent. It is unlikely that localized and smaller-scale
forms of transhumance (especially where there would
only be localized movement of herds around the site) did
not exist at this time in the mountain valleys of Bosnia.
This kind of transhumance may have always existed, but
it is impossible to demonstrate its presence or absence
since it would leave behind minimal evidence (Chang and
Koster 1986: 104).
under the direct control of the late Bronze Age palaces

(Halstead 1991).
A fundamental difference between the data and models
offered by Geddes and Halstead relates to the nature of
transhumance. Geddes envisions that transhumance is
conducted as part of a groups annual subsistence cycle.
In other words, early agricultural groups would employ
transhumance as part of their diversified approach to
subsistence. In contrast, Halstead views the origins of
subsistence from the perspective of the appearance of
productive specialization. Implicit in his analyses is the
view that transhumant pastoralists are specialists in a
larger economic system, and that this economic
adaptation would not appear until the evolution of largescale complex societies. This is a view echoed by Cherry
(1988). Hence, the nature and scale of transhumance
would be different. In essence, they are discussing vastly
different situations and they are both probably correct
with respect to the types of transhumance. Presently in
the Mediterranean, transhumance is practiced largely by
specialists (Chalkea 1999; Moreno Garca 1999; Perez
and Saez 1990).
Greenfields research attempts to determine the origins of

transhumant pastoralism in the neighboring region
Serbia (central Balkans) (Greenfield 1984, 1986, 1988,
1989, 1991, 1999a, 2001b). He evaluates the issue
through the analysis of zooarchaeological remains from
highland and lowland sites, ranging from the Neolithic to
the Early Iron Age. He also considers other
archaeological and ecological evidence for the adoption
of transhumance in the Post Neolithic of the region.
Other recent research on the origins and nature of

transhumance in Mediterranean Europe fall into these two
categories. Most have argued for the appearance of
transhumance as part of the productive specialization
characteristic of later complex societies (Classical and
Medieval). For example, Lewthwaite (1981) proposes that
transhumance in Corsica and Sardinia appeared during the
Medieval period as part of a tactical adaptation to difficult
economic and political circumstance. Wickham (1985)
demonstrated that, transhumance is not an ecologically
determined activity, but is embedded into the social and
economic context of the medieval societies of Spain.
Ecological considerations highlight the advantages of the

adoption of transhumance in the Balkans during the Post
Neolithic (c. 3300 BC). This is reflected in the
archaeological material from eastern and central Europe
which indicates minimal exploitation of agriculturally
marginal areas of the highlands during the Neolithic
(6500-3300 BC) (Greenfield 1999a, 2001b). These areas
are not extensively utilized or colonized until the
Eneolithic and Bronze Age (3300-1000 BC).
Colonization of these areas includes the widespread
adoption of transhumant pastoralism as part of the
overall domestic animal management strategy
(Greenfield 1989: 37).
IV. Previous research in temperate southeastern

Europe
Archaeological evidence from the Post Neolithic also

shows extensive reorganization of settlement patterns that
includes more closely spaced and smaller sites with
horizontally displaced occupations, lower artifact
densities and insubstantial structures. These sites tend to
not occupy the same loci as Late Neolithic sites, and there
is little evidence for large and internally well-organized
communities. Greenfield and others interpret this
evidence for population redistribution. There is a move
from substantial nucleation of settlements, with the
regional population concentrated in a few large
settlements to a pattern of an increased number of smaller
and less intensively occupied residential localities. The
introduction of the wagon and the plough into this region
at this time may well have contributed to this
redistribution, offering increased efficiency in
transportation and cultivation respectively (Bankoff and
Greenfield 1984; Greenfield 1984, 1986, 1999a; Sherratt
1981, 1983).
Studies such as Geddes (1983), Flannery (1965) and

others emphasize the element of ecological necessity to
move stock from the hot and dry lowlands in the summer
to the cool and moist highlands in the winter. This
concept is frequently used in order to explain the
development of transhumant pastoralism in arid
environments, such as the Mediterranean. However, such
an ecological necessity is less obvious beyond the
Mediterranean littoral. The climate north of the
mountains surrounding the Mediterranean is more
temperate, with a reverse seasonal relationshipwet
summers in both highland and lowland. This allows stock
to graze year-round in the lowlands. Hence, the
discussion shifts from a focus on ecological necessity to
the inclusion of cultural choices.
The earliest attempt at explicitly incorporating
transhumant pastoralism into discussions of prehistoric
subsistence in temperate southeastern Europe was by
16

appears as a byproduct of the exploitation of appearance
of productive specialization. In most cases, this assertion
is made without bringing to bear the necessary
zooarchaeological data (e.g. Clark 1991; Lewthwaite
1984).
Transhumant pastoralism would have been an efficient

response to the new challenges occurring at this time.
Human populations were increasing and bringing larger
areas under cultivation in the low and mid-altitudes
resulting in less available and predictable pasture.
Additionally, the paleo-environmental data indicate that
significant change in the physical, vegetative and micromammalian environments took place at this time. The net
result of these shifts was a decrease in the upper altitudinal
limit for cultivation, mostly affecting highland
communities. The adoption of transhumance at this time
encourages exploitation of minimally utilized highland
zones, less suited for agriculture. It is likely that members
of the lowland and mid-altitude communities would have
been responsible for the colonization of these areas in order
to ensure continued access to these new and vital grazing
resources for their stock. The highland settlements would
also remain dependent upon lower altitude settlements and
pastures to buffer the seasonal variability in resource in the
highlands (Greenfield 1999a, 2001b).
Greenfields analysis of the zooarchaeological remains

from Late Neolithic and early Post Neolithic (Eneolithic
and Bronze Age) sites lumped together the cranial and
post-cranial data. An abundance of recent research has
indicated that mandibular tooth remains can be a more
precise means for reconstructing age of death and culling
patterns (Crabtree 1982; Grant 1975; Payne 1973; and
many others cited in Lyman 1994 and Reitz and Wing
1999). This study will test the validity of Greenfields
results by extending the temporal range to include sites
from the Early Neolithic through to the Early Iron Age and
take advantage of the increased precision from a focus on
the teeth.
VI. Differences between models comparing apples
and oranges
Greenfields analyses (1984, 1986, 1988, 1989, 1991,

1999a, 2001a) utilized ternary (three-pole) graphs to
illustrate the respective percentages of the three age
cohort variables (very immature, sub-adult and adult). It
was clear from the analysis that there was a distinct shift
in the percentage of slaughtered age cohorts at the advent
of the Post Neolithichighland sites are beginning to
slaughter individuals at younger ages than lowland sites.
As a result, Greenfield interprets these data to indicate
that transhumant strategies have appeared at the onset of
the Post Neolithic.
While the proposals of the various scholars appear to be

greatly at odds, this is not really the case. They are
essentially arguing differing explanations for different
times, places and adaptations. Geographically, Geddes
(1982) and Halstead (1991) argue for the appearance of
transhumant pastoralism in an essentially Mediterranean
European environment (southern France and Greece,
respectively). Within a Mediterranean environment, the
benefits of pursuing such practices, escaping summer
heat and the provision of adequate water and pasture are
obvious (Greenfield 1999a: 15). In contrast, Greenfield
(1986, 1988, 1991, 1999a, 2001b) discusses its origins
in a temperate European environment (northern
Balkans). Within the temperate environment, the
benefits of transhumant pastoralism are less obvious as
there are few ecological incentives to the movement of
herds to highland areas. The temperature extremes
found in the Mediterranean environment do not exist in
temperate lowland areas and a variety of microenvironments exist in the northern Balkans which
provide sufficient water and graze in the lowland areas
year round (Greenfield 1999a, 2001b). All the
researchers agree that transhumant pastoralism was an
important part of the pre-modern economy of these
regions.
As the advent of transhumance is hypothesized to occur

at roughly the same time as this shift to secondary
products, this diversified economic focus complicated the
results in Greenfields test for the origins of transhumant
pastoralism in southeastern Europe. Consideration of the
use of secondary products of these herds is a factor which
cannot be ignored in this region. Once it was recognized
that the type of exploitation strategy will complicate
identification of transhumance in a region, a new series of
hypotheses was generated (see Greenfield 1999a).
Greenfield (1999a) offers a series of sub-hypotheses to
come up with a means for identifying transhumance
amidst the complications of a diversified animal
exploitation. His analysis supports Sub-Hypothesis 2C,
which states that if herds (cattle and ovicaprines) were
being exploited for both primary and secondary products
during the Post Neolithic, then mortality profiles would
show less clear cut differences between highland and
lowland sites (Greenfield 1999a: 24).
Both Geddes and Greenfield propose that the adoption

of transhumant pastoralism is simply a change in local
adaptation. In contrast, Halstead (1981, 1991),
Lewthwaite (1981, 1984) and others suggest that
transhumant pastoralism only appears with the Late
Bronze Age and/or Classical complex societies in the
Mediterranean and is associated with large-scale
economic specialization of transhumance. Halstead
(1981: 334) and Walker (1983) further maintain that
extensive use of upland pastures in a transhumant
manner would not have occurred on a large scale until
Similar to Greenfield, several researchers have attempted

to examine the relationship between the advent of
transhumance and secondary products exploitation in the
Mediterranean region. These studies invariably rely upon
and note the changes in settlement pattern and
subsistence accompanying the advent of the Post
Neolithic. They argue that transhumant pastoralism
17

the development of extensive agriculture. They
explicitly link this type of economic strategy with the
use of secondary animal products. Several researchers,
notably Lees and Bates (1974) and Cribb (1990) also
argue in support of a later date for specialized
pastoralism in the Near East, suggesting that large scale
mobility of people and their domestic herds had to occur
at a time when specialization in either a pastoral or
agricultural production had became a possibility.
VII. Conclusions
Many researchers (Halstead 1981; Geddes 1983;
Greenfield 1986a, 1999a, 2001b) agree that transhumant
pastoralism was an important element of the economy in
prehistoric times. This realization has become widely
accepted in the literature (e.g. Harding 2000: 138;
Milisauskas 2000). However, the exact temporal origins of
transhumant pastoralism in Europe are still debated.
Geddes (1983), focusing on the Mediterranean
environment, proposes that transhumant pastoralism
appears at the Mesolithic-Neolithic junction. Halstead
(1981, 1991) suggests that transhumant pastoralism
appears with the Late Bronze Age and Classical complex
societies in the Mediterranean (Harding 2000: 138). In
contrast, Greenfield (1986, 1988, 1991, 1999a, 2001a)
focuses on the temperate European environment, and
tested the hypothesis that transhumant pastoralism
becomes a major element of the subsistence strategies in
southeastern Europe only at the beginning of the Post
Neolithic, that is, the Eneolithic and Bronze Age, along
with the advent of secondary products exploitation. Most
importantly, each of the recent approaches incorporates a
regional approach to the question. This is necessary since
the comparison of upland and lowland assemblages is the
basic ingredient necessary for the identification of
productive complementarity, as implied by transhumant
pastoralism under mixed farming conditions.
Once again, it appears that the various models are not

really discussing the same situation. In fact, it may seem
as if they are investigating apples and oranges.
Greenfield and Geddes are focusing upon the evolution
of transhumant pastoralism as part of a diversified
subsistence strategy, while Halstead, Lewthwaite, Lees
and Bates, Walker, and Cribb are concerned with the
evolution of specialized transhumant pastoralism as part
of complex societies. It appears that Halstead is arguing
that transhumance would not be visible in Greece until
it became a specialized economic adaptation. From the
above summary, it should be clear that the presence or
absence of transhumant pastoralism in prehistoric
periods has not been firmly established (Greenfield
1999a).
18
CHAPTER 4
REGIONAL ENVIRONMENT
I. Introduction
valleys often exhibit very different combinations of

regional environmental variables, yet retain the general
pattern of environmental diversity within the area as a
whole (Greenfield 1991).
An important factor in the consideration of the origins of

transhumant pastoralism in any geographic area is the
regional environment. Several questions need to be
addressed including: Is transhumance a viable and
productive economic strategy for that area? Would
aspects of the regional environment affect or encourage
its adoption as a dominant economic strategy? To provide
answers, the area of interest must be defined and
described. In this chapter the regional environment,
including topography, climate and vegetation, is
described in order to consider the viability of transhumant
pastoralism within the study area.
A transhumance strategy within this area would allow for

the productive exploitation of the variety of resource
areas. Prehistoric peoples had a wealth of environmental
and ecological knowledge of a variety of plants and
animals, which enabled successful adaptation to a variety
of conditions (Flannery 1965). The adoption of
transhumant pastoralism would have offered a viable
strategy in order to take full advantage of the variety of
micro-environments created by the topographic and
climatic variations within the northern Balkan area.
Additionally, cultural factors may have played a
significant role in the adoption of transhumant
pastoralism. As agricultural practices required more
arable land, the movement of domestic herds may have
been a practical response to changes in settlement
patterns and restriction of pasture for domestic stock in
lowland areas.
The area of interest for this investigation is commonly

referred to as the central Balkans (Figure 1.1) (and
sometimes, northern Balkans). It is somewhat of an
arbitrary geographic region in that it has few defining
topographic, geographic or political borders. In terms of
modern political units, it includes Serbia and the
Vojvodina, eastern Bosnia, northwestern Bulgaria and
southwestern Romania (the Banat and Oltenia).
Topographically, its boundaries to the east are defined by
the mountain ranges of eastern and central Serbia. The
northeast division follows along the topographic break
between the lowlands of the Banat and southern edge of
the Transylvanian plateau. The boundary to the south is
defined by the mountains dividing Serbia (including
Kosovo) from Macedonia. The Dinaric Alps divide the
interior from the Mediterranean littoral along the Croatian
coast in the southwest. To the west, the Bosna and Sava
rivers form the western boundary of the region while the
Danube marks the northwest limit. The Fruka Gora
defines the northeastern limit. There are no obvious
geographical features to define the northern boundary
(Ehrich 1965; Greenfield 1986a).
II. Landforms
A. Topography
The central Balkans is characterized by extreme
topographic complexity and there are no rigid
topographic boundaries that clearly define the extent of
the region. A diversity of environmental conditions exists
in both the highland and the lowland areas. Some
highland areas are heavily forested, while others lack
heavy forest cover. Lowland areas may be marsh-like,
while others are well-drained (Greenfield 1986a).
The topography of the central Balkans displays great

variation within a small area. Generally, the area can be
divided into highland and lowland areas. Highland
regions include the rolling hills and low mountains of
central Serbia (umadija), the high mountain areas of
eastern Bosnia, and the Iron Gates. The plains of
Pannonia, the Dacian Plain and the Morava valley are
lowland areas. Local climate is greatly influenced by this
variation in landforms, and can be described as
transitional between the arid Mediterranean climate and
the more temperate climate of central Europe (Figure
4.2.) (Greenfield 1986a, 1991; Pounds 1969).
The word Balkan is of Turkish origin and denotes

a chain of mountains (Navy 1944: 1), and is an
accurate descriptive word for the region. Several major
mountain chains dominate and divide the Balkan
Peninsula into the major vegetation and climatic divisions
that characterize the region. Perhaps the most important,
and the one from which the area derives its name, is the
Balkan range (or the Stara Planina in Slavonic). This
mountain range is an extension of the Carpathian system
which forms an inverted S as it curves through
southeastern Europe. Most of the Carpathian mountain
range curve is in Romania, and encircles the
Transylvanian basin (Gottmann 1969). The Carpathian
range is interrupted by the Danube at the Iron Gates.
Climate and plant and animal communities take on

different characteristics not only in a general N-S
gradient but also with increasing altitude. Neighbouring
South of this division is the Balkan mountain range which

curves to the east and south of the Lower Danube. The
northern slope of the Balkan range descends steeply to the
19
low and flat Danube Plain, while the southern slope is a
complicated system of highlands and ranges that occupies
almost the entire Balkan Peninsula (Gottmann 1969).
the Morava which flows northwards from Kosovo and

southern Serbia, and the Sava and the Drava which flow
northwards and eastwards, respectively, from the Dinaric
and Julian Alps. The Sava also has several important
tributaries including the Drina, Bosna and Una. Other
important rivers in the region include the Neretva, Vardar
and Martisa. Only the Danube and the Vardar are capable
of supporting major water transport, as the others are too
shallow, rapid or rocky to allow for safe passage of even
small boats. While the river system in the Balkans is
extensive, there are few large lakes in the region (Danta
and Hill 1996; MacDonald 1973).
To the northwest is a system of mountains ranges and

plateaus that are an extension of the Italian Alps. These
ranges, often called the Dinaric Alps, extend from
Slovenia to Macedonia (Markovi 1968; Navy 1944;
Greenfield 1986a), with their highest point overlooking
the Adriatic Sea. Steep cliffs overlook small plains in the
lowland areas along the coast. Inland, there are an
alternating series of highlands and lowlands, with valleys
and basins where altitudes range from roughly 300 to 650
meters asl, which descend towards the Pannonian Plain to
the north. Extensive plateaus are found above 1300 and
even 2000 meters asl (Gottmann 1969).
III. Climate
There are two major and two minor climatic zones found
within the Balkan Peninsula (Figure 4.2). They are:
A third mountain range, which adds to the balkanization

or division of the region into sectors, is the Rhodopian
Massif. It extends into southeastern Serbia, Macedonia
and Bulgaria. This range runs roughly from east to west
to the Black Sea (Danta and Hall 1996; Markovi 1968;
Navy 1944).
The Balkan Peninsula is a region of complex tectonic
activity. There are numerous fault, fold and joint systems
that influence the relief of the region. Steep rugged slopes
are the dominant features of the landscape, especially
along the waterways. Additionally, the predominance of
limestone, especially in the western half of the peninsula,
has a great effect on landform development in the area.
Limestone is porous and easily eroded. This causes
abundant subsurface solution weathering which creates the
karst topography characteristic of this region. Features of
this type of landscape include underground caves,
travertine terraces, sinkholes, dolines and poljes. Dolines
are closed hollows, larger than sinkholes. These can
combine to form poljes, large flat areas that can extend for
several kilometers and are often the only viable agricultural
land in mountainous areas (Danta and Hall 1996).
1.
The Mediterranean (and Mediterranean Transitional)

Zone, which includes the regions of the southern
Balkans (Dalmatia, Macedonia, Greece, southern
Bulgaria and the southwestern Black Sea Coast);
2.
The Continental (temperate) Zone this includes all

countries north of the regions detailed above. This is
a temperate climatic zone. This zone can be divided
into two sub-zones:
a. The Humid Continental Warm Summer Zone,
which includes the northern Balkans (Serbia,
northwestern Bulgaria and southern Romania);
b. The Humid Continental Cool Summer Zone,
which is found to the north, closer to central
Europe (Pound 1969).
The climate varies throughout the region as a function of

altitude and proximity to the two neighbouring climatic
zones. In general, the climatic pattern is continental with
very cold winters and very hot summers (Figure 2.2)
(Halpern 1967; MacDonald 1973; Navy 1944). The area
of investigation lies within climatic zone 2a. It can be
described as a southern temperate climate (similar to that
of central Europe) that retains features of the more arid
Mediterranean climate zone to the south.
B. Rivers
Various rivers cross cut the major mountain ranges of the
region and play an important role in placing boundaries
between areas. The Sava River in the west and the
Danube in the northeast have been the traditional
geographic limit between the Balkans and central Europe
(by those who exclude Romania). At the same time, the
major river systems and their tributaries are essential in
communication and movement through the mountainous,
forested and marshy areas of the region from prehistory
to the present (Ehrich 1965; Greenfield 1986a).
A. Temperature
Throughout the region, July is the warmest month of the
year, while January is the coldest. Temperatures vary in the
different regions due to variability in altitude and relief. In
the Vojvodina, July temperatures can reach as high as
+35C, in combination with high humidity. North of the
Sava River January temperatures can fall to as low as 15C.
In the interior of the Balkan Peninsula winter temperatures
are slightly warmer, ranging from 15 to 10C. The
difference between winter minima and summer maxima
over the area is generally between +50 to +55C. The most
extreme temperature range is found in the Banat (eastern
Vojvodina) where July maxima may reach 42C and
February minima may reach 12C (Furlan 1977; Halpern
1967; MacDonald 1973; Navy 1944).
North of the Dinaric Mountain divide, all of the rivers

and streams flow into the Danube, which empties into the
Black Sea. Along the way, it runs parallel to sections of
the borders between Serbia, Romania and Bulgaria.
Major tributaries of the Danube systems include the Tisza
and Maros which flow southwards from the Carpathians,
20
Figure 4.1 Map showing climatic divide between Mediterranean and temperate central Europe
Again,
as
with
temperature,
the
various
microenvironments vary in the amount of precipitation
that they receive. The annual mean in the umadija is
between 600 and 1200 mm, the Banat receives somewhat
less (600-700 mm) and the Iron Gates region receives the
highest levels, between 800 and 1200 mm (Furlan 1977).
B. Precipitation
The region is characterized by a continental rainfall
pattern, where the highest levels of precipitation fall in
the summer (June) and the lowest levels fall in the winter
(February). There is little variation in precipitation levels
throughout the region over the course of the year and
seasonal variations are not as extreme as in the
Mediterranean areas further south. The majority of
precipitation (nearly one-half) falls between April and
July and the number of days with rain is high, often more
than 150 days. This pattern generally ensures adequate
water throughout the agricultural growing season,
although in some areas, a high evapo-transpiration rate
exceeds precipitation (Barker 1975; Furlan 1977; Halpern
1967; Navy 1944).
As temperatures begin to drop in the late fall and early

winter, precipitation begins to fall in the form of snow.
While the earliest frost can occur as early as the end of
September, the average date is October 21st. Snow
accumulations occur from January to March. The last frost
generally occurs around April 13th, although it can occur
during the first week of May. The amount and timing of
frost and snow is largely a function of altitude. In the
uplands, snow accumulations can often remain for long
21
Dacian Plain and the Morava River valley. Grasslands
and mixed forests predominate in the plains. In contrast
to highland areas, the soils in the plains and river valleys
are generally deep and fertile. Consequently, agriculture
is limited to these lowland areas in much of the region
(Danta and Hill 1996; Halpern 1967).
periods and would have had significant effect on settlement

and occupation, as movement would have been inhibited
(Furlan 1977; Halpern 1967; MacDonald 1973; Navy 1944).
IV. Environmental Regions
The extensive mountain ranges have created several
micro-regions within the Balkan Peninsula. These can
generally be divided into highland and lowland regions,
each with distinguishing environmental conditions.
Pannonia, as the plains north of the Danube-Sava line are

called, is also known as the Great Hungarian Plain. It is
located north of the Danube and Sava Rivers and is
surrounded by the Carpathian, Dinaric and Alpine
Mountains. As part of the Middle Danube Basin, the
Pannonian plain is characterized by the extensive river
system including the Sava, Drava, Tisza, Krs, Mure
and Timi Rivers. Pannonia is an area of low, flat relief
and lies between 100 and 200 m above sea level
extending into Croatia, Serbia, Romania and Hungary
(Greenfield 1986a). The floodplain of the rivers is wide
and is bordered by low terraces and plains. The lowland
areas experience frequent and devastating flooding. There
are extensive permanent and seasonal marshes that are
impassable and unusable for agriculture (Barker 1975;
MacDonald 1973; Nandris 1970). The annual
precipitation is 600-700 mm and increases slightly as one
moves westward (Furlan 1977). Soils are predominantly
loess deposits, although small areas of alluvial and brown
forest soils are found. Today, some lowland areas are
used for agriculture. Additionally, some mixed oak
forests are found in the higher elevations. On the upper
terraces, pine and spruce dominate the vegetation (Navy
1944).
A. Highland areas
Highland regions include the rolling hills and low
mountains of Central Serbia, also known as umadija,
and the high mountain areas of eastern Bosnia, eastern
Serbia, and the Iron Gates.
The vegetation of highland areas is characterized by
coniferous forests at the higher elevations, while scrub
forests cover the coastal areas and the more southerly
portions of the peninsula. Throughout the more
mountainous regions of the area, vegetation is sparser
than in the rolling hills and lowlands due to tree-line
limits, the limestone environment and its associated karst
topographic features. Additionally, the soils that cover the
limestone are often acidic (Danta and Hill 1996).
umadija is the northern-most extension of the Dinaric
Alps. This is an area of high altitude (1000 m asl) that
decreases as it descends into the foothills and terraces of
the Pannonian Plain (200-300 m asl - Grubi 1980). In
general, the average summer temperature in this region
ranges from +20-25C, while the average winter
temperature ranges from 0 to 2.5C.
The Morava valley is an extensive and important river

system that is part of the Middle Danube Drainage Basin.
It is split into a southern (the Juna) and a western branch
(the Zapadna) and drains a large area south of the
Danube. North of Ni, the western and southern branches
converge to form the Lower Morava that empties into the
Danube near Golubac. While the river meanders within a
zone only three or four kilometers wide, the river plain is
extremely wide (over 16 km) and flat. Fertile areas are
found near the river, as flooding is common. The rest of
the river plain is relatively dry (Barker 1975; Chapman
1981; Navy 1944).
The Balkan Mountains of eastern Serbia and their

extension across the Danube in Romania are an area of
jagged, broken relief. They are highest in the south,
generally more than 2000 m asl, decreasing to roughly
1000 m asl in the north. In the north, while still
mountainous, the relief is gentler and broken up by various
rivers and streams. Settlement in the northern part of this
region appears to be restricted to a series of small isolated
basins that provide the only areas suitable for agriculture.
The southern basins are larger and more contiguous (Navy
1944; MacDonald 1973; Stoianovi 1966).
Abundant forests covered the hills and mountains

bordering the Morava valley in prehistoric times. These
were the focus of extensive clearing in prehistoric and
historic periods, and so little remains today. However, the
soils are very fertile and today are under cultivation. The
main crops include corn, sunflowers, wheat, alfalfa,
vines, vegetable, and fruit orchards. A continental climate
characterizes the area. Rainfall (roughly 640 mm
annually) is evenly distributed throughout the year, and in
combination with the water retentive soil, farming is
possible without the use of irrigation (Halpern 1967;
Markovi 1968).
The Iron Gates is a series of gorges that connect the

Pannonian (or Hungarian) Plain and the Dacian basin. It
is a corridor formed by the Danube cutting through the
Carpathian and Balkan mountain range. While the
mountains on either side of the river are steeply sloped,
small basins where the river widens offers areas
appropriate for settlement (MacDonald 1973; Navy
1944).
B. Lowland areas
The Dacian basin is found where the Danube emerges

from the Iron Gates, east of the arc of the Carpathians.
The lowland areas include the plains of Pannonia, the

22

6000 and 2500 BP. As agriculture and animal husbandry
practices developed throughout this period, the
increased proportion of cattle and sheep/goat would
have caused progressive disturbances of the vegetation
by humans. While this was a gradual process there
would be long-term environmental effects resulting
from continuous animal and human exploitation of
vegetation with ruthless forest clearance for cultivation
and the collection of fodder for domestic animals. These
interpretations are supported by the faunal data that
show the increase in the proportion of domestic species
within assemblages (Halstead 1981). Additionally,
pollen diagrams indicate extensive manipulation of the
ecosystem throughout the Neolithic period (Nandris
1976). Even if one takes a highly conservative view of
the magnitude of the climatic changes at the Atlantic to
Sub-Boreal transition, when these are combined with
the human actions of deforestation and cultivation over
the course of the Neolithic, they undoubtedly
contributed to ecological changes during the Bronze
Age (Greenfield 1986a).
Here, the Danube changes to a north-south orientation

and the relief of the area is much more subdued. It falls
into the eastern Balkan region (Greenfield 1986a; Grubi
1980; Navy 1944).
VI. Regional environment and climate in prehistory
A. Environment and climate in the Balkans during the
Neolithic
The Neolithic period is characterized by the spread of
agriculture throughout Europe and corresponds with the
warm, moist Atlantic phase (ca. 6200-3300 BC),
dominated by a warm, maritime climate. This period was
distinguished by warm, wet winters and cool, wet
summers (Butzer 1971). The climatic optimum is reached
at the very beginning of the Atlantic period followed by a
decrease in temperatures thereafter. During the climatic
maximum, the average July temperature for Hungary
ranged from a high of about +19C to lows of +16C
toward the end of the period. Annual precipitation levels
reached a peak at about 5700 BC, then decrease towards
the end of the Atlantic period (Kordos 1978a).
C. Environment and climate in the Balkans during the

Post Neolithic
The abundance of moisture during the Atlantic enabled

maximum deciduous forest growth and spread, roughly
200-300 m higher in altitude than today. This is evident
from the dramatic increases in arboreal pollen throughout
the Lower Danube drainage. Pine and spruce species
dominated the upper altitudes, while oak, hazel, birch and
spruce were common in other areas, dependent on soil
type. Loess plateaus were dominated by steppe-grasses.
Dry grassland conditions appear and spread during this
period as well (Kordos 1978a).
The transition from Atlantic to Sub-Boreal times (33003000 BC) is characterized by a climatic trend of gradual
cooling). There is a shift to a more continental climate
and is characterized by increased dryness as compared
with that of the preceding Atlantic period (Butzer 1971;
Frenzel 1966). Climate was significantly cooler with an
increase in frequency of temperature oscillations. Winter
temperatures decease while summer temperatures
increase slightly. While average July temperatures for
Hungary at the beginning of the sub-Boreal demonstrated
highs of roughly +16C, these decreased to lows of
roughly +15C towards the end of the period. Annual
precipitation levels are at their lowest levels of the postGlacial period at the beginning of the Sub-Boreal. Both
the vertebrate remains (Kordos 1978a, 1978b) and
palynological evidence from the northern Balkans
support these interpretations (Frenzel 1966; Willis and
Bennett 1994; Zolyomi 1980).
While minor cold fluctuations occurred throughout the

Atlantic period, changes in temperature were slight.
However, the last five hundred years of the period is
characterized by a much sharper oscillation toward a
colder climate. Changes towards the end of the period
include a reduction of the altitudinal extent of the forest
as spring and summer temperatures are cooler and frosts
arrive earlier in the fall (Lamb 1977). Winters also
become cooler. Elm and ivy species decline in the pollen
spectra and spruce forests are replaced alpine meadows.
The height of this cold period was reached at c. 3200 BC.
This was followed by rapid improvement (stabilization)
towards a more pronounced continental climate (Frenzel
1966).
Fossil molluscan evidence from eastern Moravia dated to

the end of the Atlantic period show that species
characteristic of forest-steppes or open forests increased
steadily while those mollusc species adapted to lakes,
ponds and other moist environments decreased at this
time. This data presents evidence of a worsening of the
water budget, with reductions in the availability of water
(Frenzel 1966).
B. Anthropogenic environmental changes during the

Neolithic
It is also important to consider human action on the
environment. While agriculture moved into temperate
southeastern Europe at roughly 5900 BC, the human
effects on the environment were not immediately
apparent. Recent palaeoecological research (Willis and
Bennett 1994) indicates that the first indication of
anthropogenic change on the landscape occurs between
Changes in forest composition occurred as palynological

evidence indicates a decline in elm and ivy pollen spectra
(Frenzel 1966) and in increased dominance of oak and
beech (Butzer 1971). Several small climatic fluctuations
take place, with marked rainfall fluctuations (Butzer
1971; Frenzel 1966; Lamb 1977).
23
D. Anthropogenic environmental changes during the Post
Neolithic
environmental units. The sites are described in greater

detail in Chapter 7.
Sedimentary analysis has been conducted in the

Mediterranean in order to provide data on processes of
deposition and erosion in the past. Geomorphological and
pedological changes, particularly in the formation of soils
at the end of the Neolithic and in the Early Bronze Age,
are evident. These are seen as mainly human-induced due
to reduced vegetation cover and impeded drainage.
Through a combination of deforestation and intensive
agriculture caused erosion from the hills and deposition
in the plain. This resulted in a reduction of areas suitable
for cereals in the lowlands and resulted in the
abandonment of tell sites, which had been occupied for
significant periods of time. These settlement shifts were
the result of the reduction of environmental possibilities.
Erosional processes in these areas caused shifts in
settlement with an increase in settlements in the valley
bottoms (Nandris 1977). Unfortunately, such research has
not had parallels in the central Balkans. But, settlement
does not shift to the valley bottoms. In fact, it spreads to
include the highlands during the Post Neolithic.
Kadica Brdo is located on the Glasinac Plateau amidst the

mountains of eastern Bosnia. The sites of Blagotin,
Ljuljaci, Novaka uprija, Petnica, and Stragari-ljivik
are found nestled among the rolling hills of the umadija
region of central Serbia. Vina-Belo Brdo is found on the
bank of the Danube at the southern edge of Pannonia,
with the hills of umadija to its back. Foeni-Sala and
Opovo are located within the lowlands of Pannonia.
Livade is found on the bank of the Danube at the edge of
the Dacian basin with its back to the mountains of eastern
Serbia (Figure 1.1) (Greenfield 1986a).
When the sites are classified according to elevation and
environment, they can be placed into three groups.
Lowland sites include Foeni-Sala, Livade, Novaka
uprija, Opovo, Stragari-ljivik and Vina-Belo Brdo.
Mid-altitude sites are Blagotin, Ljuljaci and Petnica. The
Greek Macedonian site of Megalo Nisi Galanis (included
in the analysis) would fall into this category. The sample
of highland sites is limited to Kadica Brdo.
It is not suggested that the climatic shifts and effect upon

the environment of the region occurring at the
Atlantic/sub-Boreal boundary were severe enough to
have been the sole cause of the significant cultural
changes seen at the Late Neolithic/Post Neolithic
boundary. There were undoubtedly other factors, both
social and economic, at work. However, the climatic
shifts may have added some impetus to changes already
taking place. It has been suggested that the adoption of
transhumant pastoralism as an economic and land use
strategy was an adaptation to these new ecological
constraints (Greenfield 1986a).
VIII. Conclusion
The nature of the environment makes pastoralism a
possible economic strategy in the region. At the same
time, it does not negate the pursuit of an agricultural
lifestyle. There often exists a correlation between
pastoral forms of production and a range of ecological
factors. While these factors may serve to effectively
limit agricultural production, this is not true in all areas
and does not lead directly to the adoption of pastoralism
(Bonte 1981; Dhavalikar 1989). Particularly in
temperate environments, ecological conditions may not
directly limit agriculture. Rather it is the combination of
climatic, topographic and vegetative factors that present
pastoralism as an equally viable economic strategy that
may be pursed. Whether it is pursued or not lies in the
realm of the decision-makers rather than the
environment.
VII. Site locations in relation to major environmental

regions
Using the above information, it is possible to allocate the
archaeological sites used in this analysis to different
24
CHAPTER 5
EVIDENCE FOR SETTLEMENT, SEDENTISM AND MOBILITY
IN CENTRAL BALKAN CULTURE HISTORY
I. Introduction
II. Early Neolithic
The hypotheses to be investigated here propose that

the origin of transhumance took place at one of three
time periods, the advent of the Early Neolithic, the
Late Neolithic/Post Neolithic junction, or with the
advent of productive specialization (in this area, during
the EIA or later periods). To properly investigate the
other periods of interest, it is necessary to examine
data from the periods preceding and following each of
these temporal divisions. In order to ensure that the
reader has sufficient background to evaluate the
information from the various periods under
discussions, the culture history of the central Balkans
from the Neolithic through the Early Iron Age is
summarized below.
A. Chronology
The Early Neolithic began in the northern half of the
Balkans about 5900 BC, as agriculture spreads north from
the southern Balkans. The period ends roughly at 4900 BC,
based on radiocarbon dates (Ehrich and Bankoff 1990;
Manson 1990), although more recent citations indicate
slightly older dates 6500-5200 BC (Tringham 2000). For
the purposes of this research, we will use the older system
of dates since the later periods have not yet been reassessed
in this respect.
B. Cultures
The cultures of the northern Balkans are referred to
collectively as the Karanovo I-Kremikovci-StarevoKrs-Anza-Cri culture group. Each represents a
geographical variation on the same larger material culture
theme Karanovo I in southern Bulgaria, Kremikovci in
western Bulgaria and southern ex-Yugoslavia
(Macedonia), Starevo in eastern ex-Yugoslavia (Serbia
and Bosnia) and Krs in SE Hungary and northernmost
Serbia (Vojvodina), and Cri in Romania (Tringham
1971, 2000).
The manufacture and decoration of pottery has

traditionally been the focus of culture historical research
in the Balkans and has been used to distinguish the
different cultural groups, as well as to mark distinctions
between periods. A great deal of research has been done
detailing the characteristics and changes through time of
pottery (e.g. Dumitrescu 1983; Garaanin 1983; Bailey
2000). However, as the focus of this investigation is the
economic activities and subsistence practices, the
discussion will not dwell on the typological history of
pottery. This chapter provides a brief summary of the
major periods of interest, and the general economic and
culture historical characteristics of each.
The abundance of different cultural names can generate a

misleading picture of the cultural landscape. The different
names correspond to variations in settlement type,
location, and economy. There is little variation in
material culture (Tringham 1971: 78). Others suggest that
the distinct names reflect adaptations to the microenvironments of the Balkan Peninsula (Kaiser 1984: 46),
or are the basis of distinct regional pottery styles (Barker
1985: 90). While these are all valid interpretations,
archaeological nationalism can certainly not be ignored as
a major factor (Jongsma 1997: 55; Tringham 1971).
It is widely recognized, particularly within the Neolithic

periods, that there are considerable differences between
the traditional relative chronology and the radiocarbon
dates. These differences are often as great as one
thousand years for the periods of concern (Garaanin
1983: 84). Attempting to reconcile the two systems is an
exercise in frustration. It is beyond the scope of this
investigation to attempt to resolve these issues. As such,
the relative chronological ages of the period will be
focused on, as the data have been analyzed within this
context. It would be preferable to provide calibrated
(BC) radiocarbon dates, as these have been corrected for
the fluctuations in the formation of radioactive carbon
(Cauver 2000). Unfortunately, uncalibrated dates are the
most common form of reporting from the region, often
without the necessary information to allow calibration.
As such, dates will be given in both uncalibrated (bc)
and calibrated (BC) radiocarbon years, where possible.
C. Settlement
Early Neolithic settlements are generally located on the
terraces surrounding river valleys or on the edge of
plateaus. These areas offered dry dwelling places on
abandoned levee/channel systems and much arable land,
which provided a buffer against drought years, enabling
populations to survive and flourish. These environments
were rich in natural resources, such as wild deer and boar,
fish and shellfish, water birds and a wide range of plants,
25
EVIDENCE FOR SETTLEMENT, SEDENTISM AND MOBILITY IN CENTRAL BALKAN CULTURE HISTORY
that could be exploited (Barker 1975; Sherratt 1983b; van
Andel and Runnels 1995).
these sites, wild resources may dominate the assemblage

(Bknyi 1974a; Greenfield 1993, 1994, in press a;
Whittle 1996).
These early farming communities were located on the

brown forest soils and alluvial deposits of the rivers. It is
clear that a range of soils for easy cultivation and
proximal grazing were chosen. Settlements in the flat
plains to the north of the Serbian hill country were
commonly found on the well-drained loess deposits of the
Danube basin (Barker 1975; Chapman 1981; Sherratt
1983b; Tringham 1971; van Andel and Runnels 1995).

The Early Neolithic period has been assumed to be
associated with greater sedentism (than in the preceding
Mesolithic), with some researchers going so far as to say
that the production of food demanded permanent
settlements (Kalicz 1970). Researchers have utilized
many aspects of settlements such as depth of deposits,
architectural remains, site size or organization and
reliance on domesticates in an effort to substantiate
claims of sedentism for the Early Neolithic. Although
increased permanence and sedentism is listed as a
defining difference between the Neolithic and the earlier
periods, it is a characteristic that researchers admit is
difficult to prove. In the past it has been argued that tell
sites are evidence of increased permanent settlement
during the Neolithic (Kotsakis 1999; Whittle 1985).
Evidence from tell sites, such as Achilleion, indicate
periods of abandonment and relocations within the
mound (Gimbutas et al. 1989). In central Bulgaria, sites,
such as Karanovo, demonstrate depth of deposits is over
12 meters, although this frequently cited figure is
misleading given that it includes later occupations at the
site. In the more central part of the Balkans (from
northeastern Bulgaria through SW Romania and the
southern part of the Hungarian plain), most settlements
were relatively briefly occupied, with only thin
occupation layers. Recent research (Greenfield 2000;
Greenfield and Draovean 1994; Greenfield and Jongsma
in press a, b, c; Whittle 1996) has indicated that Starevo
sites have largely been misinterpreted. It is clear that
settlements were not permanent sedentary affairs and
probably experienced seasonal or annual abandonment
(Greenfield and Jongsma in press a, b, c; Jongsma and
Greenfield 2001). There is no evidence for permanent
sedentism in the Early Neolithic of the central Balkans
and it should not be considered a defining characteristic
of the earliest phase.
There is little evidence of multi-level occupations. Most

sites have thin layers, often from a single occupation
(Whittle 1996). Early Neolithic sites are small, usually
less than two hectares in size.
Settlements are not differentiated into functionally
distinct areas and there is no evidence of settlement
hierarchy, social ranking or occupational specialization
(Greenfield 1993; Tringham 2000). Evidence of spatial
arrangement from large-scale horizontal excavations of
Early Neolithic sites has revealed that settlements were
organized around a large central pit house structure
surrounded by a ring of smaller pit houses with an open
plaza between them (Greenfield 2000; Greenfield and
Jongsma 2003, in press a and b: Jongsma 1997; Jongsma
and Greenfield 2001).
D. Fauna
Domestication was introduced into the southern Balkans
from the Near East (ca. 6500 BC). In this region, the
similar climatic conditions resulted in no significant
changes in domesticated species or exploitation systems.
However, as agriculture spreads northwards across the
climatic divide into the northern Balkans during the Early
Neolithic, the existing domesticated plants (wheat and
barley) and animals (sheep and goat) required adaptation
to the temperate environment. Additionally, there was the
need for the domestication of new species already
indigenous to the area, such as cattle and pigs (Bknyi
1974a; Greenfield 1993; Tringham 1971), although recent
genetic research indicates that European cattle were
probably descended from Near Eastern ancestors
(Edwards et al. 2004).
Increased investment in architecture within a settlement

can also be used as an indicator of increased sedentism
(Whittle 1996). Semi-subterranean houses, or pit houses,
are a common type of architecture during the Early
Neolithic of the central Balkans. These structures are
characterized by very little modification of the living
area. They were often large and irregular in shape,
contained no internal structural features of postholes, and
roofs were simple wooden affairs (Jongsma 1997;
Jongsma and Greenfield 2001). These structural elements
suggest structural impermanence and would have been
appropriate for a lifestyle of considerable mobility
(Whittle 1996: 52).
In the overall subsistence economy, hunting and animal

husbandry were of roughly equal importance. When sites
are examined individually, hunting dominated over
husbandry in some sites, and the opposite occurred in
others. There is a tendency for the proportion of domestic
animals on any given site to decrease as one goes from
south to north. The common domestic animals included
sheep and goat, cattle and pig. Sheep dominated the fauna
assemblages over cattle, goats and pigs. Only small
numbers of goat and pig remains were found. Wild
species included red deer, aurochs and wild pig. In
settlements near appropriate water sources, fish,
waterfowl and other marine resources were exploited. In
The evidence for permanence rests on the assumption that

increased investment in architecture and the constraining
needs of subsistence activities will result in increased
26

sedentism. However, the current reinterpretation of tell
sites as discussed above and the extensive investigation
into architectural styles in this area during the Early
Neolithic does not support these assumptions (Jongsma
1997; Jongsma and Greenfield 2001). Additionally, it
should be realized that the adoption of animal husbandry
and plant cultivation does not inevitably lead to
sedentism (Whittle 1996). Greenfield and Jongsma (in
press a, b, and c) believe that most flat Early Neolithic
sites in the northern Balkans were occupied for short time
periods, for at most two or three years. These cannot be
considered permanent settlements, since a degree of
mobility was still an important element of the cultures.
that dominates the central Balkans is the Vina A and B

culture (c. 4500-4100 BC) (Chapman 1981; Greenfield
1986a; Tringham 1971), also known as the VinaTordo culture (Garaanin 1983).
C. Settlement
In the Middle Neolithic, more sites appear and these are
more widely spread across the landscape and encompass
a wider range of sizes. Sites are located in an increasing
number of environments, but are still restricted to the
lowlands or upland riverine basins (e.g. Obre I).
Sites began to expand in size, with some reaching over
50 hectares. However, it is likely that this large area of
occupation was the result of a series of shifting smaller
occupations (Tringham and Krsti 1990).
III. Middle Neolithic

A. Chronology
D. Fauna
The Middle Neolithic period of the northern Balkans

covers the period from 4900-4100 BC (Chapman 1981;
Greenfield 1986a). The chronology for this period may
appear confusing, as there are two major chronological
systems in use. Miloji (1949) divides the culture into
four major periods, A through D, each with additional
sub-phases. Also widely used is the system by Garaanin
(1983). It divides the culture into only two major periods
(with subdivisions), the earlier Vina-Tordo (I and II)
and later Vina-Plonik (I, IIA and IIB). These two
systems are easily reconcilable as they are nearly coterminus with Vina-Tordo I = Vina A, Vina-Tordo
II = Vina B, Vina-Plonik I = Vina C, Vina-Plonik
II = Vina D (Chapman 1981; Greenfield 1991: 163).
The Early Vina sites were agricultural communities,

but had less dependence on wild resources than in
previous periods (Lekovi 1991; Tringham 1971). Most
characteristic about this period was the increase in the
importance of domestic cattle. In the more heavily
forested areas, the primary domestic species was cattle,
followed in decreasing order of frequency by pigs and
then by sheep/goat. In the drier more open ecological
zones, sheep/goat predominated, with cattle in second
place, and pigs a distant third (Greenfield 1986a). There
is evidence for increased local domestication as remains
of aurochs increased as well as the transitional forms
between domestic and wild cattle. The same pattern of
local domestication is seen in pig. Wild pig remains
increased as well as the occurrence of transitional
forms. In contrast, species such as red deer decreased in
importance (Bknyi 1974a).
However, in Chapmans (1981: 31) extensive exploration

of the Vina culture, a new division of the culture into
early and late phases is suggested. This new interpretation
is based on radiocarbon dates and artifactual evidence
that indicate only a part of the total length of the Vina
culture is represented by the type-site of Vina Belo
Brdo. Therefore, Chapman suggests an early phase,
contemporary with the Vina A, B, and C phases at the
type site, (c. 4500-3950 bc), and a late phase, (c. 39503300 bc), represented by the Vina D phase at the type
site and later occupations at other Vina sites, such as
Divostin. All this being said, the system to be used in this
investigation is the most well known and most widely
cited in other publications, the Miloji and Garaanin
subdivisions of the Vina culture. In this analysis, the
Middle Neolithic period is represented by the Vina A
and B cultures (c. 4500-4100 BC) (Ehrich and Bankoff
1990; Greenfield 1986a).

Sedentism was a gradual process during the Neolithic
period. It is clear that the degree of sedentism of
settlements increases throughout the Neolithic period.
Researchers do not make solid claims for sedentism
until later phases of the Neolithic (Kaiser and Voytek
1983). As in the Early Neolithic, the evidence indicates
that the Early Vina sites were not permanently (yearround) occupied, but, in contrast, were more often reoccupied. This caused an increase in the formation of
tell sites (Lekovi 1991; Tringham 1971). However, flat
sites without significant occupational development
continue to exist, highlighting the lack of permanent
sedentism and further demonstrating the range of
variation throughout the region (Whittle 1996).
B. Cultures
In northern Bulgaria and southern Romania, the Middle
Neolithic is represented by the Vdastra I culture in the
west (Oltenia) and the Dudetri culture in the east
(Muntenia). In southern Bulgaria, the Middle Neolithic
is characterized by the Veselinovo culture. The culture
There is also evidence for increases in population based

upon regional survey, in particular the Selevac valley
(Chapman 1990). Further evidence comes from the
growth of architectural units over the course of time at
various sites. Houses increase in number and become
27
analyzed) separate into lowland and mid-altitude
clusters. This is due to differences in the pig and cattle
ratio between these areas. Cattle dominate in all sites
but pigs are more common in the lowlands than in the
uplands. The major domestic species were all exploited
primarily for meat, although goat gives some indication
of also being exploited for secondary products such as
milk (Greenfield 1991).
more solidly built and longer lasting (Tringham and

Krsti 1990; Whittle 1996).
IV. Late Neolithic
A. Chronology
The Late Neolithic period of the northern Balkans
covers the period from 4100-3300 BC (Chapman 1982;
Ehrich and Bankoff 1990).
Additionally, there is evidence for a significant increase

in the hunting of wild species in the Late Neolithic
settlements. Faunal remains from red deer, wild pig,
aurochs, and fish in appropriate areas, constitute a high
proportion of the faunal assemblage (Bknyi 1974a, b;
Greenfield 1986a, 1991; Russell 1998; Tringham 1971).
The same type of lowland/highland division seen in the
domestic species is seen in the wild species. There is a
lower domestic:wild ratio in the lowland sites than in
the upland sites. It has been suggested that because the
mid-altitude areas were better suited for agriculture with
well-drained soils and easily cleared land wild animals
were pushed out of this environment. In contrast, the
lowlands, surrounded by wetlands in the plains,
specialized in the exploitation of microenvironments,
and incorporated both wild and domestic species into
the subsistence system (Greenfield 1991). Fishing
continued to play an important role in the subsistence
economy (Greenfield 1986a: 27).
B. Cultures
The Late Neolithic is represented by the Vina-Plonik
(I and II) culture. As discussed above, this is equivalent
to the Vina C (c. 3700-3400 bc; 4100-3900 BC) and
Vina D (c. 3400-3000 bc; 3900-3300 BC) phases
(Chapman 1981; Ehrich and Bankoff 1990; Garaanin
1983).
C. Settlement
Vina-Plonik settlements are often located in the same
areas as the Middle Neolithic Vina-Tordo settlements,
on the lower terraces of the rivers. However, shifts in
settlement patterns are evident as Vina-Plonik
settlements become organized into a two level site size
hierarchy. Numerous small sites surround larger sites,
and they more dispersed across the landscape. It is
during this period that permanent and highly organized
agricultural villages are established (Chapman 1981;
Kaiser and Voytek 1983; Tringham 1971; Tringham and
Krsti 1990).

The archaeological evidence, specifically the
architectural evidence, points to increases in settlement
permanency. The Middle Neolithic Vina-Tordo
settlements were overlain in later periods by thick
habitation layers of the Late Neolithic cultures
(Chapman 1981; Tringham 1971). There were increases
in building durability, and more rigid demarcation of
space including a focus on specialized areas for
economic and production activities (Bailey 2000). Sites
are larger and there is greater longevity of settlement.
This may be taken as an indication of increases in
population size and a high degree of agricultural
productivity (Chapman 1981). Many of the Late
Neolithic sites show relatively deeper (commonly over
four metres in depth), and stratigraphically
superimposed deposits and structures. This resulted in
better feature preservation and higher artifact and
feature density over large areas (Greenfield 1986).
A significant change in houses is also apparent as

buildings became more substantial. At this time, houses
were built of thick clay on a framework of wooden
posts. A central row of posts in the interior of the house
supported a gabled roof. Increased investment in houses
is evident with evidence for the external decoration
including painting, incised decoration and even clay
representations of cattle heads. Many houses had thick
clay floors over a foundation of horizontal logs
(Chapman 1981; McPherron and Srejovi 1988;
Tringham 1971).
D. Fauna
Trends in animal exploitation patterns established in
earlier periods continue through the Late Neolithic.
Domestic animals continue to dominate faunal
assemblages and the significant species (cattle, sheep,
goat and pig) remained unchanged. The movement away
from ovicaprine exploitation to a focus on cattle
continues (Bailey 2000; Bknyi 1974a, b; Greenfield
1986a, 1991; Tringham 1971). At this time, cattle
typically constitute 53-76% of the major domestic
fauna. What is most interesting to note is that in this
period the faunal assemblages (when quantitatively
In addition to this architectural evidence, the faunal

remains from Late Neolithic sites in the central Balkans
can be used to argue for sedentary lifestyles without
marked occupational seasonality. The age distributions
of several of the domestic species include all age groups
implying year-round occupation, as do the antler data.
Furthermore, comparisons of lowland and mid-altitude
sites did not find any shifts in seasonality between them
(Greenfield 1991).
28

In addition, settlement size appears to increase, but this is a
very problematic assertion given the fact that sites became
characterized by laterally-displaced deposits. Probably, the
size of actual occupied area and population within
settlements declined in this period since sites were
characterized by occupation by a few family homesteads
instead of a large population agglomeration. This pattern
becomes even more prevalent in the next period (Bankoff
and Greenfield 1984; Greenfield 1986a, 2001a).
V. Eneolithic (Chalcolithic)
A. Chronology
The Eneolithic period, also known as the Chalcolithic or
Copper Age, was the period during which metallurgy
begins to spread through the region. Stone tools,
however, continue to be the dominant technology, since
copper is a relatively soft metal and was mostly used for
decoration. The Eneolithic period of the northern Balkans
covers the period from 3300-2500 BC (Ehrich and
Bankoff 1991; Garaanin 1983; Greenfield 1986a,
1999b). During this period there were significant changes
in settlement patterns and subsistence that show a marked
break from the Neolithic. These new patterns show
continuity into the following Bronze Age (Greenfield
1986a, 2001a; Champion et al. 1984; Harding 2000;
Milisauskas 2000).
While settlement patterns in the Late Neolithic were

characterized by the continuous localized use of a small
area, there was a change to a geographically more
extensive approach in the Eneolithic. Linked with this
new system were the increasing importance of domestic
animals and the use of plough agriculture. As settlements
expanded onto soils that were inappropriate for sustained
crop agriculture, more land was cleared, which in turn left
increasing amounts of land fallow to feed more domestic
animals (Champion et al. 1984; Sherratt 1981, 1983a).
B. Cultures
The Balkan Eneolithic is characterized by geographically
wide-spread cultural complexes that are composed of a
series of regional variants. The main Eneolithic cultures
of interest in this investigation include the Baden group
to the north, which consists of a series of regional and
temporal (Baden, Kostolac, and Vuedol) variants, and
the Bubanj-Hum group to the south (I-II local southern
Serbian variant). The Baden culture has been subdivided
with an initial Baden subculture, followed by the
Kostalac group. In central Serbia, the temporal variants
(Baden and Kostolac) often cannot be separated and are
described as the Baden-Kostalac variant. The Vuedol
group appears after the Kostolac group in the Vojvodina
(Garaanin 1983).
D. Fauna
The changes apparent with the Secondary Products
Revolution (Sherratt 1981, 1983a) begin during the
Eneolithic. This included the dramatic increase in the use
of domestic animals for their secondary products, such as
cattle for traction with the plough and the cart, the
exploitation of sheep for wool, and the milking of
domestic cattle, sheep and goats. The data signal an
increase in the scale of domestic animal keeping over the
previous periods (Greenfield 1986a, 1988, 1989, 2004;
Sherratt 1981, 1983a). As in earlier periods, the animal
segment of the economy was dominated by the raising of
domestic stock, primarily sheep, goats and cattle
(Bknyi 1974a; Greenfield 1986a, 1988, 1989, 2004).
There is also the sense that there are more animals out on
the landscape than previously (Sherratt 1981). The
variation in fauna across the region is an indication that
agricultural systems were closely adapted to their specific
regions and environments (Champion et al. 1984).
C. Settlement
During the Eneolithic, tell sites are common in the plains
to the north of the Danube-Sava line, but flat sites with
laterally displaced deposits are common in the hill
country to the south. However many tells were
abandoned during the Eneolithic. In some areas, a small
number were fortified. Fortified settlements are often
located on naturally dominating, fortified positions or in
places suitable for defense (Garaanin 1983: 149). The
development of metallurgy and the new importance of
metal sources created a new sense of territoriality and the
need to secure access to these resources. As a result,
clashes often occurred and introduced the need for
fortification of sites (Garaanin 1983; Greenfield 2001a).

It is obvious that the Late Neolithic/Post Neolithic
boundary represents an important transition period in the
central Balkans. Prehistoric societies underwent dramatic
reorganization as well as a demographic redistribution.
The Late Neolithic pattern is one of high intra-settlement
population density and size, with regional population
heavily concentrated in large settlements. In contrast, the
Post Neolithic pattern is one of low intra-settlement
population density and size accompanied by population
dispersion into a larger number of more closely spaced
residential localities. This shift appears not be associated
with a population decline, but rather is simply
redistribution over the landscape. The largest settlements
have broken up and their population dispersed while the
smaller settlements continued to exist, but usually in
In the Eneolithic, there was an expansion of the

geographic range of site locations. Settlement expands to
include the agriculturally-marginal highlands and the
spaces between previously settled areas (Champion et al.
1984: 160; Greenfield and Bankoff 1984; Greenfield
1986a, 2001a).
29
different locations (Bankoff and Greenfield 1984;
Greenfield 1986a).
defensible positions such as on hilltops. While these sites

are small, they are considered to be regional centres and
were permanently occupied. Each river valley area is
dominated by one or two such sites. Lower level sites
were small and undifferentiated villages or hamlets. River
valleys may contain several of these sites. However, this
pattern of settlement hierarchies did not occur uniformly
throughout the region. Some areas have a clearer
dominance hierarchy than others (Greenfield 2001a).
The evidence for sedentism and mobility within the

Eneolithic is difficult to interpret. Two explanations (not
necessarily incompatible) have been offered. Sherratt
(1981: 292) suggested that settlements were occupied for
shorter amounts of time and there was a more rapid
turnover of cultivated areas, as compared with earlier
periods. Greenfield (1986a: 32) suggested that part of the
population began to move around the landscape rather
than seasonal abandonment of settlements. Domestic
stock was moved to between lowland and highland areas
to provide greater protection from winter conditions and
better quantity and quality of graze.
Another feature of Bronze Age settlements is that they

are essentially shallow, usually with deposits of only 50
cm or less (excluding pits or disturbed plow zones).
These Bronze Age settlements show insubstantial wattle
and daub architecture, rarely more than one vertically
definable occupation level, features that are easily
disturbed by ancient and modern plowing and
horizontally displaced stratigraphy. Additionally, these
sites are spatially extensive and have low surface and
subsurface artifact densities (Bankoff and Greenfield
VI. Early Bronze Age

A. Chronology
The Bronze Age period begins in 2500 BC, ends at 1000
BC in the central Balkans and is commonly divided into
early, middle and late phases. The Early Bronze Age
extends from 2700-2000 BC; 2000-1700/1600 bc
(Garaanin 1983; Greenfield 1986a, 2001a; Harding
2000).
While statements on community organization in the Early

Bronze Age are limited by the lack of extensive
excavation, two dominant patterns emerge. Most Early
Bronze Age sites demonstrate a haphazard layout of
structures (Greenfield 2001a).
B. Cultures
D. Fauna
Cultural groups in the area, while mutually related, had
their own regional limits and differences, based mainly
on pottery types. These groups include Vinkovci in Srem
and Slavonia, Slatina in central Serbia, and Bubanj-Hum
III in the valley of the Southern (Juna) Morava River of
southern Serbia (Garaanin 1983).
The changes described above for the Eneolithic, in terms

of the Secondary Products Revolution, come to dominate
the domestic animal assemblages (Greenfield 1986a,
1988, 1989, 2004). In the Bronze Age, farmers practiced
a mixed subsistence strategy with cattle, sheep, goat and
pig as the main domestic animals, exploiting them for
both their primary and secondary products. But there is
no evidence for specialization. Some variation in species
importance occurs in response to local environmental
conditions. Several changes were evident in domestic
livestock frequency. Cattle decrease in frequency while
still remaining the dominant species, with a
corresponding increase in sheep, goat and to a lesser
extent pig (Bknyi 1974a; Greenfield 1986a).
C. Settlement
At this time, the major changes in settlement patterns that
begin in the Eneolithic become dominant throughout the
region. The data from the Eneolithic are suggestive, but too
sparse at present to arrive at more definitive conclusions.
In contrast to earlier periods, the Early Bronze Age pattern
is one of low intra-settlement population density and size
accompanied by population dispersion into a larger number
of more closely spaced residential localities (or settlement
sites). Population increases and is accompanied by a
redistribution of the population over the landscape. The
settlements of the Bronze Age are located in a wider range
of environments, utilizing most of the major environmental
zones of the Balkan Peninsula. This shift may be the result
of the greater diversification of subsistence that
necessitated a change in land use. Settlements were no
longer restricted to the rich alluvial soils along water
sources, but now were found in all areas, highland and
lowland (Greenfield 1986a, 2001a).
The domestic horse also appears in the region during this

period. The appearance of this species is one of the
features that most sharply distinguish Bronze Age animal
husbandry from preceding periods. It is the first new
species to appear in the region since the Early Neolithic
(Bknyi 1974a: 33). Horse remains appear to be
associated with elite settlements during this period
(Greenfield 1986a, 2006).
Wild species are still present in faunal assemblages and
evidently continued to be part of the subsistence
economy. But their importance was greatly reduced. Fish
continue to be an important component of the wild fauna,
particularly in lowland sites (Bknyi 1974a; Greenfield
1986a, 2001a).
In certain areas, such as the hills surrounding the

Pannonian plain, settlement hierarchies emerge. Upper
level sites are often fortified and located in naturally
30

Age continue into the later phases. There is low intrasettlement population density and size accompanied by
population dispersion into a larger number of more
closely spaced residential localities. The redistribution of
the population over the landscape results in few large
settlements and many smaller settlements. A distribution
pattern of several sites per valley, dominated by a single
regional center, is common (Greenfield 1986a; Harding
2000).

Significant shifts in subsistence and settlement patterns
begin during the Eneolithic and continue during the Early
Bronze Age (Greenfield 1986a). Most importantly is the
increase in geographic range of settlement to include
highland environments. This pattern began in the
Eneolithic. It has been suggested that these important
changes were the result of developments in agriculture
and husbandry. Sherratt (1981, 1983a) argued that
increased specialization in the secondary products of the
main domestic species, specifically the introduction of
the cart, would have resulted in increased ease of
movement for people at this time. A complementary
suggestion has been the adoption of transhumant
pastoralism as the dominant economic strategy
(Greenfield 1986a, 1988, 1999a). Transhumance has
historically been a significant part of the economy in the
Mediterranean and the Balkans and probably was also an
important element of the economy in prehistoric times
(Geddes 1983, Greenfield 1986a, Halstead 1981; Harding
2000). Each of these hypotheses suggests greater mobility
of populations across the landscape. The settlement data
in particular supports this idea.
Bronze Age settlements show shallow occupation

horizons and have insubstantial wattle and daub
architecture, rarely more than one vertically definable
occupation level. These sites are spatially extensive and
have low surface and subsurface artifact densities
(Bankoff and Greenfield 1984; Greenfield 1986a,
2001a).
D. Fauna
It is important to stress that there is little variation in
animal husbandry across the region. The relative
importance of each species in an assemblage seems to
vary in response to local environmental conditions
(Bknyi 1974a; Greenfield 1986a; Harding 2000: 134).
However, it is possible to highlight some trends within
the period. Cattle are the dominant species, although it
has declined somewhat in importance in comparison to
the previous period. Sheep/goat increases in importance
at the expense of cattle and pigs, although pigs increase in
frequency somewhat as well (Bknyi 1974a: 34;
Greenfield 1986a).
VII. Middle and Late Bronze Age

A. Chronology
The data from the Middle and Late Bronze Ages of the
region are extremely similar. As a result, they will be
discussed together. Within the central Balkans, the
Middle Bronze Age extends from 1900-1400 BC (15001300 bc), while the Late Bronze Age extends from 14001000 BC (1300-1200 bc) (Garaanin 1983).
In contrast to the variety seen in the exploitation of

domestic animals, the hunting of wild species was barely
incidental and much more homogeneous throughout the
Bronze Age. Aurochs were less frequent than red deer
and generally less than other wild species. This indicates
two features of the later Bronze Age animal exploitation
patterns. First, aurochs were becoming rarer throughout
the landscape. Additionally, this indicates the decline of
the domestication process. Animal husbandry had
advanced to the point where it can support itself not only
by increasing the stock of domestic animals without
further domestication and providing the population with
meat but also supplying other foodstuffs (milk), giving
draught power and material for clothing (Bknyi
1974a: 34).
B. Cultures
As the Bronze Age progressed, closely linked cultural
groups evolved that shared many common features. This
makes the delimitation of their exact territories rather
difficult as they often overlapped. As a result, the Middle
and Late Bronze Age are often grouped together in
publications (e.g. Garaanin 1983; Harding 2000). Major
cultural groups often extend from the first into the second
sub-period and include the Vatin and Dubovac-uto Brdo,
found in southern Pannonia (the Vojvodina) and Danubian
Serbia, and the Verbicioara ceramic groups found in
Romania (Dumitrescu 1983; Garaanin 1983). These
cultural groups all appear to be contemporary and fall
within the Middle-Late Bronze Ages of the central Balkans
(Reinecke Br A2/B1 to D ca. 1600-1200 bc). The end
of the Bronze Age in the region is usually associated with
the Halstatt A (c. 1200-1000 bc) (following Reinecke
Garaanin 1983; Greenfield 1986a).

Whether it was due to the increased use of the cart (Piggott
1982), or the adoption of transhumant pastoralism
(Greenfield 1999a), the evidence seems to suggest that
there was an increase in mobility of people across the
landscape during the Bronze Age periods. Smaller site size
and lack of significant architectural structures suggests a
degree of impermanence in the settlements.
C. Settlement
The settlement patterns established in the Early Bronze
31
exploited for both primary and secondary products. Horse
remains became a small, but common part of faunal
assemblages. The most noted difference between faunal
assemblages of this period and earlier periods is the
almost complete lack of any bones of wild species, except
in the highlands. Fishing may have continued in certain
areas, but was relatively unimportant away from the
riverine environments (Bknyi 1974a; Garaanin 1983;
Greenfield 2005b).
VIII. Early Iron Age

A. Chronology
The chronology of the Early Iron Age in the central
Balkans is once again complex. Before World War II,
Reineckes system for central Europe was extended to the
Balkans. The Late Bronze and Early Iron Age of the
region was divided along the same lines as in Central
Europe. The application of this system to the Balkan area
was problematic since it assumed that the Iron Age began
with Halstatt A.

The dominant settlement pattern of the Early Iron Age is
a continuation of the patterns established in the Early
Bronze Age. The occurrence of smaller more dispersed
settlements seems to suggest an increased mobility of the
population across the landscape. These observations have
been interpreted as facilitating the growing importance
of pastoralism in the economy (Raish 1992: 15).
Subsequently, Halstatt A has been recognized to be part

of the Late Bronze Age for the central Balkans (14001000 BC). It is subdivided into two sub-periods (Halstatt
A1-A2). Halstatt B is the transitional period between the
Late Bronze Age and Early Iron Age (1000-800 BC) and
subdivided into through sub-periods (Halstatt B1-B3).
Finally, the true or classic Early Iron Age of the central
Balkans is characterized by Halstatt C and D (Garaanin
IX. Conclusions
The extensive temporal range of this study, from the Early
Neolithic to the Early Iron Age, necessitated a brief
summary of relevant aspects of central Balkan culture
history. Over the course of this vast time frame, local
cultures experienced the initial colonization by early
farmers in the Early Neolithic through to the evolution and
dominance of the region by chiefdoms in the Early Iron
Age. Other significant cultural change occurs throughout
this time, specifically developments in population
distribution and stability, intra- and inter-regional
interaction, agriculture, transport, the development of
pottery manufacture and the use of metals.
B. Cultures
The Early Iron Age in southeastern Europe is represented
by the Halstatt C and D (800 - 500 BC) cultures
(Garaanin 1983; Greenfield 1986a). While traditionally,
Halstatt material culture covers a vast area of central
Europe, from Budapest into Germany (Wells 1984), it
clearly extends based on similarities in material culture
into the central Balkans (Garaanin 1983).
C. Settlement
The first major animal domesticates (sheep/goat, cattle

and pig) appear during the Early Neolithic and continue
to be the dominant fauna through all periods, including
the Early Iron Age. During the Post Neolithic, the
exploitation of cattle and ovicaprines shift to include their
secondary products. Most importantly for this study are
the indications for the appearance of transhumant
pastoralism in the region at the advent of the Post
Neolithic. These include dramatic shifts in human
population distribution over the landscape to include the
agriculturally marginal highlands and the exploitation of
animals for their secondary products. It is this linkage
between the shifting importance of the various domestic
species over time and the potentially interrelated changes
in settlement patterns that hints at the appearance of
transhumance. This hypothesis will be tested in
subsequent chapters.
During this period, there is a continuation of the trend of

the abandonment of tell sites and settlement dispersion
into smaller communities. These sites were commonly
located on hilltops or other places well placed for defense
(Garaanin 1983: 591) and were often fortified (Raish
1992: 15). The pattern of highland settlement continues
with the location of sites along inter-valley routes of
movement, probably to control the movement of goods,
animals, and people (Greenfield 1999a).
D. Fauna
It appears that the patterns of animal husbandry and
exploitation that were common in the Bronze Age
continue into the Early Iron Age. There is the continued
dominance of cattle over the other common domestic
species of sheep/goat and pig and these species were
32
CHAPTER 6
METHODOLOGY
I. Introduction
uses of the domestic species. Finally, epiphyseal data

are often biased because of differential attrition between
early and late fusing bone ends. As a result, early fusing
bone elements tend to be more resistant to attrition. This
leads to a bias against certain age classes especially
against the very young age classes. Their bones are not
fused and are very subject to attrition (Greenfield
1986a, 1991; Lyman 1994; Munson 2000). The
disadvantages of epiphyseal fusion for ageing can be
overcome by the use of tooth eruption and wear data.
The general approach or methodology chosen for this

investigation is zooarchaeology (also known as faunal
analysis). Zooarchaeology is the study of animal bones
from archaeological sites, in other words, from cultural as
opposed to natural deposits (which would be
palaeontologyOlsen 1971; Olsen and Olsen 1981;
Reitz and Wing 1999). The primary purpose of
zooarchaeology is to investigate the interactions between
humans and animals. Among hunter-gathers, these data
form the basis of subsistence systems. In food producing,
it forms the basis for the regional economy.
Two techniques of analysis within Zooarchaeology were

selected for this study: tooth eruption and wear and
cementum analysis. Tooth eruption and wear is the
second most commonly used technique in the ageing of
individual animals from zooarchaeological collections
(Payne 1973; Reitz and Wing 1999). Cementum analysis
is a technique that is often separated from primary faunal
identification because it requires specialized equipment
(Reitz and Wing 1999), is significantly more time
consuming, and is destructive of specimens.
Nevertheless, its application has proven useful for the
determination of both age and season of death in
mammals (Burke and Castenet 1995; Lieberman and
Meadow 1992).
Zooarchaeology is particularly suited to the investigation

of transhumant pastoralism because it utilizes the remains
of the animals of interest (Geddes 1983; Greenfield
1999a). While the proportion of the human population
that was involved in the movement of the livestock is still
debated, the majority of animals were moved around the
landscape in a mixed pastoral economy (Khazanov 1984).
As a result, the remains of the animals should yield a
clear pattern of complementary movement between
highland and lowland sites (Greenfield 1986a, 1991,
1999a, 2001b).
II. Tooth wear and eruption
In order to test the hypotheses presented in earlier

chapters, the techniques chosen for this investigation
must establish both age of death and season of death of
the archaeological faunal remains of the major domestic
species (cattle, goats, pigs, and sheep). There are a variety
of techniques for the determination of age of death in
zooarchaeological samples. These include epiphyseal
fusion and closure of cranial sutures, tooth growth and
replacement sequences, tooth wear, incremental
structures, and antler and horn development (Davis 1987;
Reitz and Wing 1999; Wilson et al. 1982).
Tooth wear is one of the oldest techniques for age

determination, and is particularly applicable to large
herbivores (Davis 1987). It is both easy to apply and
provides accurate age determination when tied into a
known age sequence (Lowe 1967). The sequence and
timing of the eruption of the teeth in the mandible for
domestic animals has been established for a number of
species (Silver 1969; Habermehl 1961; Getty 1975). It
is widely utilized today (Hambleton 1999: Munson
2000).
The major technique of age determination in

zooarchaeology has traditionally been epiphyseal fusion
of bones (Silver 1969). There are several problems
associated with this method. First, the chronological age
of fusion is not known for most wild species. Second,
the epiphyseal age classes are sometimes too coarse to
establish meaningful mortality profiles (Klein and CruzUribe 1984: 43). It is this second characteristic which is
of great concern for this investigation. Additionally, the
fusion of epiphyses ceases relatively early in an
animals life (Payne 1973: 283) when an animal reaches
maturity. As a result, it is difficult to distinguish a prime
adult (young but full-grown) from an aged one, based
purely on the epiphyseal data (Crabtree 1982: 242;
Klein and Cruz-Uribe 1984: 43). This is a necessary
analytical step if we are to reconstruct the economic
The eruption of deciduous and permanent teeth in

animals allows for ageing in much the same manner as
the state of epiphyseal fusion of the post-cranial skeleton.
The dentition of animals offers several advantages over
epiphyseal fusion. Teeth can monitor age more or less
continuously throughout the life of an individual, and
allows for ageing beyond the range of epiphyseal fusion
(Reitz and Wing 1999). Analysis of tooth wear and
eruption enables a clear distinction between prime adults
and senile adults (Klein and Cruz-Uribe 1984). Further,
the mandibles and mandibular teeth are less affected by
preservation bias (taphonomy) (Payne 1973; Munson
2000). This greater durability of teeth makes them more
abundant in samples.
33
METHODOLOGY
There are three main disadvantages associated with tooth
wear and eruption. The primary disadvantage is that the
nature of the animals environment, specifically, the
foods eaten, and the amount of grit consumed when
eating can affect the degree of wear. As well, the general
nutritional health of the animal can affect wear stages of
teeth (Reitz and Wing 1999). While tooth eruption and
wear stages can provide a good estimation of
physiological age of an individual, there are problems in
attempts to assign an absolute age to wear stages
(Hambleton 1999). Finally, while cow and pig can easily
be identified and distinguished morphologically from
each other, it is more difficult to distinguish sheep and
goat mandibles, particularly with fragmentary material
(Boessneck et al. 1963; Ewbanks 1964). Recent
techniques have been proposed for the separation of the
taxa among young animals (Payne 1985) and have
recently been extended to older animals as well (Halstead
et al. 2002).
Absolute age for mandibles in the database was

established with the following procedure. Each
mandibular tooth row is assigned a score based on the
tooth eruption and wear pattern of the three molar teeth
(Grant 1975; Payne 1973). The score for the mandible is
then converted to an absolute age (Tables 6.1-6.3). For
mandibles that are missing one of the three molar teeth, it
is still possible to assign a mandibular score. The wear
stage of the missing tooth can be predicted with reference
to complete molar rows in which the wear stages of the
teeth are the same as those that are present in the
incomplete mandibles (Grant 1982; Payne 1973). It is
possible to establish a range of mandibular stages, and so,
a range of absolute ages, utilizing the charts discussed
above (Tables 6.1-6.3).
Loose teeth are useful especially in those sites where
there are few or no mandibles. Normally, one does not
like to use them because of the possibility of inflating the
count of Minimum Number of Individuals. But in many
cases in this investigation, it is necessary to use them due
to the problem of small sample size. The only loose teeth
included in the examination were those molar teeth that
were specifically identified to number, i.e. first, second or
third molar teeth. Lower loose teeth will be dealt with as
if they were mandibles with two missing teeth.
There are two main systems for the recording of tooth

eruption and wear data. Both systems utilize a series of
diagrams showing successive tooth wear stages, where
individual wear or eruption stage of each tooth is
recorded by reference to the diagrams (Grant 1982; Payne
1973). The first system to be established is by Payne
(1973) and focuses on sheep and goat remains. The
second system is from Grant (1975) and is applicable to
sheep, goats, cattle and pigs. Halstead (1985) adapted
Paynes technique for use with cattle as well. Both
systems identify tooth wear stages for permanent
mandibular molars and the deciduous and permanent
fourth premolar. The wear stages of each molar tooth are
combined to provide a mandibular wear stage. The
systems differ in terms of number of wear stages, how the
tooth wear stages are defined and the relation to absolute
age. As both systems have been used to record tooth wear
and eruption for the sites considered here, data were
converted to the same format to make them comparable
using the methodology detailed by Hambleton (1999).
The system ultimately utilized to describe the tooth
eruption and wear from the various assemblages is
Paynes because his results allow for easier recognition of
herd age structure and kill-off patterns. Additionally,
Grants method provides neither indication of absolute
age nor the relative duration of different wear stages
(Hambleton 1999).
Sometimes, the mandibular score does not correspond

with a single age stage. This normally happens when
some teeth are missing in the tooth row. Under these
conditions, the tooth wear for the missing tooth must be
estimated (based on the other teeth) and the numerical
score calculated. In these cases, the numerical score will
cover a range of stages, instead falling into a single stage
(e.g. stages B-C in the Payne system). The data will be
recorded in the following manner. The raw count
represents mandibles and loose teeth that have been
identified as belonging to each stage on the basis of the
defined criteria as defined above (Payne 1973; Grant
1975; Hambleton 1999). Following the method proposed
by Payne (1973) the corrected count represents specimens
that cannot be placed more closely than within a group of
stages. These specimens are proportionally allocated to
the age stages based on the raw count data. This is
explained by Paynes (1973) example (Figure 6.3).
The modern goat comparative collection provides a
control of the assigned ages of sheep/goat given by Payne
(1973). The suggested ages listed by Payne for sheep/goat
will be compared to the known ages of the collected
specimens. While it would appear more logical to assume
that the earlier breeds would be more applicable to
archaeological populations, the research does not bear out
that assumption.
III. Establishing absolute age using tooth eruption and

wear
In some investigations, it is adequate to assign individual
bones to age classes, such as juvenile, immature and adult
as was done by Greenfield (1986a) previously. But where
assessment of the economic significance of domestic
species is required, establishing the absolute age of the
species is highly desirable (Bullock and Rackham 1982).
For the purposes of this investigation, the assignment of
absolute ages is a necessary step and will be assigned to
each species as above.
While it has been argued that the 19th century data for the
relationship between tooth eruption and wear and age
would
be
most
applicable
to
prehistoric
zooarchaeological data, Payne (1985) suggests that
modern 20th century eruption timetables are more
34

applicable to archaeological populations. Additionally,
research on the dental eruption and wear of Soay sheep, a
feral breed believed to be the closest living analogy to
prehistoric breeds, also agrees with Silvers modern
estimates and Paynes suggested absolute ages
(Hambleton 1999). Therefore, the use of a modern
collected sample for this type of testing seems both
appropriate and applicable. The suggested ages for cattle
and pig cannot be tested in this way, as there is no
adequate comparative modern mandible collection of
known age available to the researchers. The suggested
absolute ages from Hambleton (1999) for these breeds
will be accepted as valid.
pastoralism on a regional basis, the sample of sites needed

to be extended to provide a baseline from which to test the
hypotheses. As this is the first zooarchaeological approach
to this question in this region, we were seeking to lay the
essential foundation and to highlight areas for necessary
future work. Where we see the effects of small sample size
on the analysis, this immediately highlights where larger
faunal assemblages are required and where future
fieldwork needs to be directed.
The chi square statistic was used in all cases where there
are two or more profiles per period in order to compare
the degree of similarity between harvest profiles from
different sites. The youngest age group was eliminated
from consideration due to the taphonomic issues and the
age groups E through to I were combined. The analysis
was performed using StatXact 4 software. The statistical
approach and software was specifically chosen to deal
with the issue of small sample size. A significance level
(p value) of less than 0.05 was considered to be
statistically significant. The software provided both a
usual p value that is the original chi square test and is best
for large samples. Additionally, an exact p value was
calculated, which corrects for small sample size (i.e. cells
less than 5) and provides a more accurate calculation
(Dennis Murphy, personal communication 2001).
IV. Production strategies and construction of harvest

profiles
Harvest profiles of faunal remains are widely utilized by
archaeologists to elucidate the exploitation strategies of a
society. Herders will differentially slaughter age and sex
groups according to the type of products they wish to
produce. The harvest profile is the distribution of age at
death for a species (Hesse 1982; Greenfield 1988, 1991;
Payne 1973). Production strategies, whether the domestic
animals were utilized for primary products, such as meat,
or whether they were utilized for secondary products,
such as milk, traction, wool or hair, will be reflected in
the harvest profile derived from the data. The use of
harvest profiles for the examination of exploitation
strategies has several advantages over other methods.
First, the species frequencies in faunal samples are
subject to taphonomic biases. Second, it is nearly
impossible to reconstruct the original herd structure.
Finally, the goal for archaeologists is to determine the
ways the animals were managed.
Through the course of analysis, it became apparent that

the traditional harvest profiles constructed in the initial
stages of analysis provided less than concrete evidence of
transhumant movement. As such, a reanalysis of the data
was undertaken. They are graphed differently. A second
set of mortality profiles, more sensitive to seasonality
issues, was constructed. Bar graphs were found to be
more useful to illustrate the pattern. New profiles were
established for all three domestic species (sheep/goat,
cattle and pig) for each site and for each major period. All
staged mandibles are counted and the percentage of the
total sample calculated. However, rather than being
sequentially subtracted from 100 and the resulting values
graphed the percentage that is graphed as percentage of
the total assemblage in a bar graph format. This form of
presentation of the data is more sensitive to seasonality
issues in the transhumant movement of herds. One can
evaluate where (highland or lowland) particular age
groups were predominantly slaughtered based on the
original hypothesized timing of transhumant movements
in the Post Neolithic.
The initial harvest profiles for this investigation are

established for all three domestic taxa (sheep/goat, cattle
and pig) for each site and for each major period using the
methodology proposed by Hambleton (1999), following
Payne (1973). Cumulative frequency graphs were used to
visually depict the patterns. All staged mandibles are
counted and the percentage of the total sample calculated.
These percentages are then sequentially subtracted from
100 and the resulting values graphed. Loose teeth are
included where they provide a sufficiently narrow age
stage or are necessary to provide a sufficient sample size.
Minimal sample size for inclusion was ten elements per
period per site. Shennan (1988) maintains that samples of
mandibles less than 15-20 should be considered too small
to provide accurate harvest profiles. For more accurate
statistical analyses, the minimum sample size should be
forty mandibles. If we had been restricted to this number, a
regional approach would have been impossible and some
of the major periods of interest would not have any
samples. This would have resulted in the inclusion of less
than 20% of the samples. While larger samples would be
more desirable, they were not available from many sites.
Since our goal was to test for the origins of transhumant
V. Cementum analysis
Incremental growth structures have been observed in a
variety of organisms. These may be present in
mineralized tissues, such as bone, molluscs, teeth,
otoliths, fish spines and antler pedicles. Correlations
between growth lines in teeth have been recognized for
more than 30 years as the most reliable means of
establishing age and season of death of individual
specimens (Lieberman 1993a, b; Lieberman and Meadow
1994; Pike-Tay 1991).
35
METHODOLOGY
If there are 3As, 6 Bs, And 1C, one of the 3Abs is allocated to A and 2 to B according to the ratio 3A:6B;
similarly the 14BCs are allocated 12 to B and 2 to C according to the ratio 6B:1C, giving 4As, 20Bs and 3Cs.
ABCs are now allocated not according to the original 3A:6B:1C ratio, but according to the new ratio
4A:20B:3C (Payne 1973: 296).
Figure 6.3. Proportional allocation example (from Payne 1973).
Table 6.1. Sheep/Goat mandibular wear stages (MWS) and suggested ages
Payne MWS
A
B
C
D
E
F
G
H
I
Grant MWS
1-2
3-7
8-18
19-28
29-33
34-37
38-41
42-44
45+
Suggested Age (Payne 1973)

0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
Table 6.2. Cattle mandibular wear stages (MWS) and suggested ages
Payne MWS
A
B
C
D
E
F
G
H
I
Grant MWS
1-3
4-6
7-16
17-30
31-36
37-40
41-43
44-45
46+
Suggested Age (Halstead 1985)

0-1 months
1-8 months
8-18 months
18-30 months
30-36 months
young adult
adult
old adult
senile
36

Table 6.3. Pig mandibular wear stages (MWS) and suggested ages
Payne MWS
Grant MWS
A
B
C
D
E
F
G-I
0-1
2-8
9-17
18-32
33-42
43-46
46+
Suggested Age (Higham 1967; Bull and

Payne 1982)
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
adult
Teeth are composed of three distinct tissues, enamel,

dentine and cementum. Enamel covers the crown of the
tooth, while cementum covers the roots of a tooth.
Dentine forms the main structure of the tooth and
underlies both the enamel and the cementum (Hillson
1986). While incremental structures have been identified
in all three of the dental tissues and previous study has
focused on both the dentine and the cementum, it is the
analysis of cementum that has proven more effective in
determining both age and season of death information
(Lieberman and Meadow 1994; Pike-Tay 1991).
Therefore, it is the incremental structures in cementum of
domestic animal teeth that are of interest in this
investigation.
(1979) provide a comprehensive summary. It is generally

recognized that the formation of incremental structures in
cementum are governed to a considerable degree by an
internal rhythm related to the animals growth and
metabolism. However, external factors can still exert
some effects on cementum formation. These can include
environment and climate (Burke 1995; Grue and Jensen
1979), dietary restrictions, hormonal changes, reduced
food intake (Saxon and Higham 1969) and reduced
nutrition and biomechanical stress (Lieberman 1993a, b).
It is likely that it is the result of a complex of factors and
the need for further investigation as to their influence is
required.
Several techniques are available for the observation of
growth structures in cementum. These include polarized
light, ordinary light, stained, histological thin sections,
transmitted or scanning electron microscopy and
microradiography (Burke 1995). The chosen technique
for this investigation is the examination of undecalcified,
ground thin sections of teeth to be viewed using a
polarizing light microscope with a rotating stage under
both transmitted and polarized light. This is one of the
few techniques that are applicable to both modern and
archaeological teeth.
Acellular cementum is the most reliable for indications of

age at and season of death. It is deposited towards the
cervical end of the tooth root and is composed of
seasonally deposited bands that appear either opaque or
translucent in polarized light (Lieberman and Meadow
1994). Dental cementum has been considered extremely
reliable, since it is deposited through the entirety of an
animal's life and is rarely reabsorbed (Gordon 1993).
Seasonal differences in the deposition of cementum are
generally manifest and described as growth zones (which
represent the period of active deposition of cementum),
annuli (which occur as a result of growth slowdown) and
Line of Arrested Growth (LAG) that occur as a result of
growth cessation. Together, these structures may be
referred to as a Growth Line Group and is the equivalent
of a year in time (Burke 1995; Grue and Jensen 1979;
Klevezal 1996).
The wider, growth zones of cementum appear light in

colour and opaque under reflected light microscopy and
transmitted polarized light. In contrast, these same
structures will appear dark under transmitted white (or
natural) light. The narrower annuli in cementum appear
dark in color and translucent under reflected light
microscopy as well as with transmitted polarized light. In
contrast, these increments will appear bright under
transmitted white (natural) light (Burke 1995; Pike-Tay
1995).
Cementum is deposited around the roots of teeth by

cementoblasts. The cementum is deposited on the surface
of the dentine, within and around the Sharpeys fibers
that protrude from the periodontal ligament (Lieberman
and Meadow 1992). Therefore, the earliest layers of
cementum are closest to the dentin, while those formed
later are closer to the outside part of the root (Burke
1995; Grue and Jensen 1979; Klevezal 1996). For a full
discussion of the histology and structure of the various
types of cementum see Grue and Jensen (1979).
Acellular cementum is deposited at the cementum-enamel

junction, or the cervical region of the tooth. This makes
this area of the tooth ideal for the counting of the
incremental structures in the cementum. In this region, the
lines are most distinct and even. Additionally, in ungulates
it has been shown that layers in cementum are often
disturbed in the apical area of the root, and layers here are
less distinct. As a result, the coronal half of the root is
preferred for interpretation (Grue and Jensen 1979).
The causes of the annual growth cycle of cementum are

still largely debated within the literature. Grue and Jensen
37
METHODOLOGY
mandible, both with Paynes system and with Grants
system. Then, the (M1) of the left side of the mandible
was extracted for thin sectioning. Only mandibular M1
was selected for sectioning for a variety of reasons.
VI. The control sample

The purpose of a modern control sample is to establish
the time of formation of incremental growth structures in
the cementum in order to apply this information to the
study of an archaeological sample (Burke and Castenet
1995). A sample of sheep and goat skulls were collected
from August 2000 to March 2001 from local Manitoba
abattoirs. Approximately six sheep per month were
collected from Carmen Meats in Carmen, Manitoba
(Canada). The goats were collected from Prairie Abattoir
in Portage la Prairie, Manitoba. The abattoir slaughtered
two animals per week, which were collected each month.
The utilization of these local animals is considered
applicable to archaeological material in the Balkans
because the animals are obtained from an environment
with roughly similar patterns of seasonality. The animals
of the comparative collection are goats raised on smallscale farms in Manitoba. They are domestic farm animals
raised with seasonal differences both in food availability
and seasonal environment.
1.
2.
3.
4.
The establishment of this control sample is for the

purpose of interpretation of archaeological teeth.
Mandibular teeth are easier to section due to their
simpler root forms.
Incisors and canines are less frequent than cheek
teeth in archaeological samples.
Incisors and canines are more difficult to identify to
the species level in archaeological samples than are
cheek teeth (Pike-Tay 1991).
A total of fourteen goats were included in the sample

collected monthly from August to March of the same
year. As these were animals slaughtered for meat
consumption, they were generally young animals (less
that a year and half). Due to the young age of the goats,
the comparative sample was supplemented by five sheep
specimens (over the age of 18 months) in order to bulk
out the older age groups of the sample.
All the goats were born in the month of April, with a

mean birth date estimated by the producer to be April
15th. The goats range in age from six months to a year
and a half. As the pattern of slaughter was two animals
per week and the specimens were collected at roughly the
middle of each month throughout the collection period,
date of death was estimated over a period of two months.
This changed through the time of collection when
collection was at the end of each month.
The modern thin section sample of goats will provide an

interesting methodological contribution by testing
whether the incremental growth visible in the teeth of the
goats agrees or does not agree with the eruption/wear
stages and/or with the actual ages. It will act as a threeway comparison between tooth eruption and wear,
incremental structures and true ages (Ariane Burke,
personal communication, 2001).
It was possible to obtain several older sheep, from a

private producer who raises sheep for wool. This small
sample included one sheep specimen that was four and a
half years old (54 months) and two animals that were a
year and a half-year-old at death. These were also
included in the final sample in order to expand the age
range covered by the sample. Additionally, two older
sheep with no age information that were collected from
the abattoir were included. As a result, the final modern
comparative collection consists of both sheep and goat.
There are forty-three animals, twenty-nine goats and
fourteen sheep, in the sample that compares tooth
eruption and wear. The thin sectioning sample consists of
nineteen animals, fourteen goats and five sheep.
VII. The fossil sample

As cementum is the least hard of the dental tissues, it can
easily be stripped from the tooth as the result of postmortem damage. As a result, only lower molar 1 (M1)
still encased in the mandible are selected for the
archaeological sample (Lieberman and Meadow 1994).
A sample of twenty left mandibular M1 teeth in total - ten
teeth from a highland site and ten teeth from a lowland
site. The sites chosen were Vina (Belo-Brdo, Serbia), a
lowland site that includes material from the Late
Neolithic, Eneolithic and Middle Bronze Age, and Kadica
Brdo (Kneina, Bosnia) a highland site that includes
material from the Early Iron Age.
It was not detrimental to the analysis to include both

sheep and goats in the sample. They are often lumped
together in zooarchaeological investigations due to
difficulties in reliably distinguishing the two species
(Boessneck et al. 1963; Payne 1973). The teeth of sheep
and goat are very similar in structure and size (Hillson
1986). Other researchers have utilized the application of
goat thin sections to Gazella gazella (Lieberman 1993b)
based on the close phylogenetic relationship (family
Bovidae).
The teeth selected for sectioning must first be removed

from the mandible. This was accomplished through the
use of a small handsaw and a high-speed rotary saw. Two
cuts were made on either side of M1 and the mandible
was broken along these lines using a pair of cutters and/or
a screwdriver. The modern teeth were then soaked in a
degreasing agent (in a 50:50 solution with water) in order
to remove as much of the fat and grease from the teeth.
Teeth were soaked for five days, and then allowed two
full days to air dry. This degreasing step was omitted for
the archaeological teeth. All teeth were measured
Mandibles were extracted from the skull and defleshed.

The tooth wear and eruption was recorded for each
38

according to guidelines in Von den Driesch (1976; see
Appendix A and B below for these data).
scratches and polish the surface. Polishing was finished

by applying a small amount of 0.05 micron powder polish
to the specimen and rubbing with a chamois cloth. The
slide was then accurately labeled with permanent ink.
The tooth must be encased in resin (Buehler Epoxide) in

order to fill any voids in the sample and protect fragile
materials. Samples were placed into vessels, or boats,
suitable to the size of the tooth. The appropriate amount
of both hardener and resin was calculated, weighted and
then mixed together (100 parts of resin to 36 parts of
hardener by weight). The mixture was poured into the
boat to fully cover the sample. The vessel was placed into
the vacuum chamber for 15 to 20 minutes, removed and
allowed to harden for 24 hours.
VIII. Age determination with cementum analysis

In most species, cementum deposition begins shortly
before or just at the time of eruption (Grue and Jensen
1979). As a result, by adding the number of seasons of
cementum growth to the age at which the tooth erupted
the age of death is established. However, to be most
accurate, one must determine age, one must establish the
time of the formation of the first increment line, from
which counting should begin (Klevezal 1996).
In order for the sample to be mounted on a slide for thin

sectioning, an initial cut through the sample was
necessary. The cut was made longitudinally, parallel to
the cementum surface. An Isomet saw with a wafering
blade was used. The block was then polished on a
polishing/grinding machine as evenly as possible using
first a 320 micron abrasive paper and, then with a 400
micron abrasive paper prior to mounting on glass slides.
The eruption time for M1 in sheep and goats will be taken

to be three months, following Silvers (1969) eruption
timetable of primitive modern breeds. Because the timing
of the eruption is before the beginning of the first winter,
some amount of cementum is present on the root before
the first winter. As a result, the first cementum layer
visible corresponds to the first year of an animals life. In
these circumstances, there is a rule to be accepted in age
determination by cementum layers. It is stated as follows:
Beuhler Epoxide was then used to glue specimens to

slides. A small amount of resin was spread on the slide in
a T shape, roughly the same length as the specimen to
be glued. Beginning at the top of the T, one edge of the
specimen was placed on the slide. The block was then
quickly lowered. The sample was moved around slightly
to allow the resin/hardener to spread under the entire
sample and release any air bubbles that may be trapped
underneath the specimen. The sample was then
positioned in the centre of the slide. Slides were left to
dry, approximately 24 hours or more.
When we see the first (innermost)

incremental line close to the dentin, it should
be considered to be formed in the first winter
of a given tooth function. When the first
incremental line is separated form the dentin
by a wide cementum band, then this line
should be considered to be a line of the
second winter of the tooth function (Klevezal
1996: 122).
Next, the slide was mounted to the vacuum chuck of the

thin sectioning machine that cuts away the majority of the
specimen block. Using 320 and 400 micron abrasive
papers, the freshly cut side was then hand thinned as
evenly as possible on the grinding/polishing wheel. This
was necessary to allow light to pass through the sample
and provide a clear representation of growth increments.
During this process the slide was checked periodically
under the microscope to estimate how much more
thinning was required to allow for optimal visibility of
the growth increments.
IX. Season of death determination with cementum

analysis
Researchers have correlated growth layers in cementum
with seasonal growth of many mammalian species. As a
result, it is possible to deduce the season of death for
archaeological samples. Under transmitted polarized light,
the bands that appear opaque are usually deposited in most
animals during the season of reduced growth; in contrast
the bands that appear translucent are most often deposited
during the growth season. The establishment of a control
sample will determine in which season these different
bands of acellular cementum are formed, from which the
season of death of an individual can be determined through
an examination of the nature of the outermost band (Grue
and Jensen 1979; Lieberman 1993a, b).
Once the structures were clearly visible and readable under

the microscope, a thin layer of resin/hardener was applied.
This prevents the specimen from drying out and lifting
from the slide. The appropriate amount the Epoxide
resin/hardener is mixed and a thin layer spread on the slide
and left to harden for at least twenty four hours.
X. Conclusion
The final stage of the thin sectioning process was to
remove as much of this final layer of resin without
exposing and further abrading the specimen. The 320
micron abrasive paper was used to carefully remove some
of the top coat of resin. The 400 micron and finally the
600 micron abrasive papers were used in turn to remove
The tooth wear and eruption data has only been analyzed
in a coarse manner (combined with the post-cranial
material) in previous research. As a potentially
informative source of information for this question, it
deserves more significant and precise attention. By
39
METHODOLOGY
analyzing it separately, a new dimension of the age-ofdeath of domestic animals can be estimated, and
information on production strategies and seasonality
inferred. While providing additional age estimation, the
cementum analysis will also supply season of death
information, an important element in the consideration of
the transhumant movement of herds.
The two zooarchaeological techniques chosen here utilize

the most relevant source of information on the
transhumant movement of domestic herds, that is, the
remains of the animals of interest. It is hoped that the
combination of both techniques will provide a complete
picture of the economic movement of these animals and
may minimize some of the problems of interpretation
associated with each technique.
40
CHAPTER 7
DESCRIPTION OF SITES, TIME PERIODS, AND NATURE OF SAMPLES
I. Introduction
leaving forty-five mandibles and three loose teeth in the

final analysis (Appendix A). Table 7.4 summarizes the
age stage distributions of Ovis aries remains from the
Early Neolithic. Both the Eneolithic and Early Iron Age
samples of Ovis aries consisted of only one mandible.
Each was identifiable to an age stage.
Primary zooarchaeological analysis including species

identification from all sites has been completed by
Greenfield and have been described in detail elsewhere
(1984, 1986a, b, 1994, 1996, 2005b, in press b, n.d. a, b,
c, and d; Greenfield and Fowler 2003, 2005). In this
chapter, the samples will be described in terms of the site,
surrounding environment, period(s) of occupation, and
the size and nature of the sample selected for this
investigation. Only identified domestic taxa mandibles
and loose teeth will be considered in the present analysis.
The sites are discussed in alphabetical order, with the
remains in each site discussed by time period. Several
samples were too small for the types of analyses for
consideration here and these are detailed in Appendix B.
A description of the remains and the final sample size for
each species and period within each site is summarized
below.
A total of 299 domestic Ovis/Capra mandibular remains

(mandibles n=233 and loose teeth n=66) were recovered
from Early Neolithic deposits at Blagotin. Most mandibles
(n=191) and many loose teeth (n=19) were unidentifiable to
age, leaving forty-two mandibles and forty-seven loose teeth
in the final analysis (Appendix A). Table 7.5 summarizes the
age stage distributions of Ovis/Capra remains from the Early
Neolithic.
As the Eneolithic sample of Ovis aries was too small for
effective analysis, it was lumped into the Ovis/Capra
category to increase sample size. The Ovis/Capra
sample probably is composed of only Ovis aries since
no goat remains were found. As a result, nine
Ovis/Capra mandibles and loose teeth were recovered
from Eneolithic levels of Blagotin (three mandibles, one
of Ovis aries and six loose teeth). Three mandibles were
unidentifiable to age, leaving six loose teeth in the final
analysis (Appendix A). This sample is too small to be
included in the analysis. The age stage distributions of
Ovis/Capra remains from the Eneolithic are
summarized in Appendix B (Table B1).
II. Blagotin
A. Site description
Blagotin, a site in central Serbia (Figure 1.1), is located at
the base of the Blagotin Mountain (250 m asl) at the
headwaters of the Blagotin stream. It is a mid-altitude site
set amidst rolling hills and low mountains, and
surrounded by mixed oak forests and rich agricultural
areas. The site is on a gentle slope above the stream and
extends onto the flatter terrace above the slope. The slope
has contributed to severe erosion in some areas of the site
(Greenfield 2000; Greenfield and Jongsma in press b, c).
The Early Iron Age sample of Ovis/Capra included a

single Ovis aries mandible from this period as well.
Eighteen domestic Ovis/Capra mandibular elements were
recovered from Early Iron Age levels at Blagotin (9
mandibles and 9 loose teeth). Five of the mandibles and
two of the loose teeth were unidentifiable to age, leaving
four mandibles and seven loose teeth in the final analysis
(Appendix A). Table 7.6 summarizes the age stage
distributions of Ovis/Capra remains from the Early Iron
Age.
The site has three periods of occupation, Early Neolithic,

Eneolithic and Early Iron Age. It is relatively small, c.
100x80 m during the Early Neolithic, 100x50 m in the
Eneolithic, and 200x200 m during the Early Iron Age.
The deposits range up to a depth of 3 m at the center of
the site, with vertically superimposed remains. There was
little laterally displaced stratigraphy at the site.
Approximately 50% of the site was wet-sieved and
floated. Over 35,000 animal bone fragments were
recovered and analyzed from the 1989-1994 excavations
at the site (Greenfield and Jongsma n.d.).
A total of 281 Bos taurus mandibles and loose teeth were

recovered from Early Neolithic deposits at Blagotin (232
mandibles and 49 loose teeth). There were 194 mandibles
and nine loose teeth unidentifiable to age, leaving thirtyeight mandibles and forty loose teeth in the final analysis
(n=78 Appendix A). Table 7.7 summarizes the age stage
distributions of Bos taurus remains for the Early
Neolithic.

Eighty-one mandibles and four loose teeth from
domestic Ovis aries were recovered from Early
Neolithic levels at Blagotin. Thirty-six mandibles and
one loose tooth were unidentifiable to an absolute age
stage (due to absence of or damage to appropriate teeth),
Nineteen Bos taurus mandibular elements were recovered

from Eneolithic levels at Blagotin (8 mandibles and 11
loose teeth). Five mandibles and two loose teeth were
41

Table 7.4. Stage distribution of Ovis aries mandibles and
loose teeth from Early Neolithic Blagotin
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
6
3
19
4
1
2
2
0
0
37
Corrected Count
%
16
9
51
11
3
5
5
0
0
100
No.
6
3
21.8
6.4
1.9
2.9
6
0
0
48
%
12
7
45
13
5
6
12
0
0
100
Table 7.5. Stage distribution of Ovis/Capra mandibles and

Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
4
1
10
7
1
1
1
3
0
28
Corrected Count
%
14
4
35
25
4
4
4
10
0
100
No.
4
1.1
13.9
26
6.6
11.6
9
16.8
0
89
%
4
1
16
30
7
13
10
19
0
100

loose teeth from Early Iron Age Blagotin
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
0
1
2
0
0
1
0
0
4
42
Corrected Count
%
0
0
25
50
0
0
25
0
0
100
No.
0
0
1.6
4.4
0
0
5
0
0
11
%
0
0
15
40
0
0
45
0
0
100
Table 7.7. Stage distribution of Bos taurus mandibles and

Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
Young adult
Adult
old adult
Senile
No.
9
3
12
15
2
4
0
1
1
47
Corrected Count
%
20
6
26
32
4
8
0
2
2
100
No.
9
3
12.4
22.9
3.9
12.8
0
7.2
6.3
77.5
%
12
4
16
29
5
17
0
9
8
100

loose teeth from Eneolithic Blagotin
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
Young adult
Adult
old adult
Senile
No.
0
0
2
3
2
1
0
0
0
8
Corrected Count
%
0
0
25
38
25
12
0
0
0
100
unidentifiable to age, leaving three mandibles and nine

loose teeth in the final sample (n=12 Appendix A).
Table 7.8 summarizes the age stage distributions of Bos
taurus remains from the Eneolithic.
No.
0
0
2
4.8
3.2
2
0
0
0
12
%
0
0
17
40
26
17
0
0
0
100
unidentifiable to age, leaving no Eneolithic Sus sample.

In the Early Iron Age levels only one Sus scrofa mandible
was found. It is identifiable to age and the final sample
size is one. This sample was too small to be included in
the analysis. The age stage distributions are detailed in
Appendix B (Tables B3 and B4).
Twenty-two Bos taurus mandibular elements were

recovered from Early Iron Age deposits at Blagotin (15
mandibles and 7 loose teeth). Thirteen mandibles and two
loose teeth were unidentifiable to age, leaving two
mandibles and five loose teeth in the final analysis (n=7
Appendix A). This sample was too small to be included in
the analysis. The age stage distributions are detailed in
Appendix B (Table B2).
III. Foeni-Sala
A. Site description
Foeni-Sala is a lowland site located in the western edge

of the Romanian Banat (Figure 1.1) near the village of
Foeni (approximately 3 km away from the Serbian
border). The site is in the midst of the flat alluvial plain of
the Banat. It is located on the bank of the Timiat, a
tributary of the Timi River. This is an area distinguished
by low altitude (c. 80 m asl) and low relief (Greenfield
n.d. d; Greenfield and Draovean 1994; Greenfield and
Jongsma in press a).
All of the domestic pig remains were too few to be

included in the analysis. Nine mandibles were recovered
from Early Neolithic levels at Blagotin. Seven were
unidentifiable to age, leaving two mandibles in the final
analysis (Appendix A). One Sus scrofa mandible was
recovered from the Eneolithic levels at Blagotin. It was
43

Two features of this area, a high water table and the gentle
slope of the land, combine to contribute to frequent flooding
of the rivers and the formation of swamps. The area was
drained in the 19th century. Forests existed along the river
edge, but these are no longer in existence. Most of the area is
presently used for cultivation, but the soil tends to be very
heavy. It was likely not extensively cultivated before the
advent of the plough (Greenfield and Draovean 1994).
Thirteen Ovis/Capra mandibles and loose teeth were

recovered from Early Iron Age levels at Foeni-Sala (11
mandibles and 2 loose teeth). Four mandibles were
unidentifiable to age, leaving seven mandibles and two
loose teeth in the final analysis (n=9 Appendix A).
Table 7.11 summarizes the age stage distributions of
Ovis/Capra remains from the Early Iron Age.
Forty-two Bos taurus mandibular remains (mandibles
n=34 and loose teeth n=8) were recovered from Early
Neolithic deposits at Foeni-Sala. Twenty-four mandibles
and one loose tooth were unidentifiable to age, leaving
ten mandibles and seven loose teeth in the final analysis
(n = 17 Appendix A). Table 7.12 summarizes the age
stage distributions of Bos taurus remains from the Early
Neolithic.
There are at least six periods of occupation at the site, the

Early Neolithic, Eneolithic, Middle Bronze Age, Early Iron
Age, Dacian (later Roman), and Medieval (12th century)
periods. For the most part, only the Early Neolithic and
Early Iron Age periods will be considered here. Few or no
relevant bone (mandibles and teeth) remains could be
assigned to the Eneolithic or Middle Bronze Age deposits.
The site is found on a low natural mound, part of an
ancient alluvial terrace. It is small, roughly half a hectare
and consists of a single, thin Early Neolithic occupation
horizon and discrete features from the Early Neolithic and
other periods (Greenfield and Draovean 1994; Greenfield
and Jongsma in press a). The sample was almost entirely
dry and wet sieved (c. 85%). Over 10,000 fragments were
recovered and analyzed from the 1992-94 excavations at
the site (Greenfield n.d. d).
Eleven Bos taurus mandibular remains (mandibles n=6

and loose teeth n=5) were recovered from Early Iron Age
levels at Foeni-Sala. Four mandibles and two loose teeth
were unidentifiable to age, leaving two mandibles and
three loose teeth in the final analysis (n=5 Appendix A).
These data were too small to be included in the analysis.
Appendix B (Table B6) summarizes the age stage
distributions of Bos taurus remains from the Early Iron
Age.

Eight Sus scrofa mandibles and loose teeth were
recovered from Early Neolithic levels at Foeni-Sala.
Two were identified as belonging to wild species (Sus
scrofa fer.) and were eliminated from the sample. The
total domestic sample size was six (4 mandibles and 2
loose teeth). Three mandibles and one loose tooth were
unidentifiable to age, leaving one mandible and one loose
tooth in the final analysis (n=2 Appendix A). The Sus
scrofa sample from Middle Bronze Age Foeni-Sala
includes a single mandible from a domestic pig. It was
unidentifiable to age, leaving no Middle Bronze Age
sample. There are no Early Iron Age Sus scrofa remains
recovered from Foeni-Sala. The age stage distribution
data for the Early Neolithic are described in Appendix B
(Table B7).
Twelve Ovis aries mandibles were recovered from Early

Neolithic levels at Foeni-Sala. Two mandibles were
indeterminate to a specific age, leaving ten mandibles in
the final analysis (Appendix A). Table 7.9 summarizes the
age stage distributions of Ovis aries remains from the
Early Neolithic.
Four Capra hircus mandibles were recovered from Early
Neolithic deposits of Foeni-Sala. Two mandibles were
unidentifiable to age, making two the final sample size. In
the Early Iron Age levels, there was only one Capra
hircus mandible. As both the Ovis aries and the Capra
hircus samples were small, they were lumped into the
category Ovis/Capra for analysis (below).
IV. Kadica Brdo
Ninety-three Ovis/Capra mandibular remains (mandibles

n=76 and loose teeth n=17) were recovered from
Early Neolithic levels at Foeni-Sala. Forty mandibles
and three loose teeth were unidentifiable to age, leaving
thirty-six mandibles and fourteen loose teeth in the final
sample (n=50 Appendix A). Table 7.10 summarizes the
age stage distributions of Ovis/Capra remains from the
Early Neolithic.
A. Site description
Kadica Brdo is located on a hilltop overlooking the
western end of the Glasinac plateau in eastern Bosnia
(Figure 1.1). It overlooks and is close (5 km) to the town
of Kneina. Glasinac is the largest highland plateau in
southeast Europe at this elevation. It is found at
elevations mostly above 1000 m asl, with surrounding
mountain heights extending up to 2000 m asl. Extensive
sub-alpine grasslands and coniferous forests cover the
plateau. Today, the area is known mainly for grazing and
forestry as it is not an environment suitable for extensive
agriculture. Pockets of small-scale agriculture are located
in depressions within the basin (Greenfield 2005b).
Five Ovis aries mandibles were recovered from Early

Iron Age levels at Foeni-Sala. All were identifiable to
age using tooth wear and eruption, making the final
sample size five (Appendix A). The age stage distribution
data are summarized in Appendix B (Table B5). Since
these data were too small to be used on their own in the
analysis, this sample was combined with the Ovis/Capra
remains for the analysis (below).
44

loose teeth from Early Neolithic Foeni-Sala
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
1
7
1
0
0
0
0
0
9
Corrected Count %
%
0
11
78
11
0
0
0
0
0
100
0
1
7.9
1.3
0
0
0
0
0
10.2
0
10
80
10
0
0
0
0
0
100

Stage
A
B
C
D
E
F
G
H
I
Suggested Age
Raw Count
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
1
2
12
3
0
2
3
0
0
23
Corrected Count
%
4
9
52
13
0
9
13
0
0
100
No.
1.5
2.1
16.5
12.6
0
6
11.7
0
0
50.4
%
3
4
33
25
0
12
23
0
0
100

loose teeth from Early Iron Age Foeni-Sala
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
0
4
1
0
0
1
1
0
7
Corrected Count
%
0
0
58
14
0
0
14
14
0
100
45
No.
0
0
4
1
0
0
2.5
1.5
0
9
%
0
0
44
11
0
0
28
17
0
100

Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
young adult
adult
old adult
senile
No.
3
1
3
2
0
2
0
0
1
12
The site at Kadica Brdo was occupied during the

Eneolithic, Early Iron Age, and Late Iron Age. Only the
faunal remains from the Early Iron Age levels derive
from secure temporal contexts. As a result, only these
will be considered in this investigation.
Corrected Count
%
25
8
25
17
0
17
0
0
8
100
No.
3
1
3
2
0
5
0
0
3
17
%
18
6
18
12
0
28
0
0
18
100
remains from the Early Iron Age.

Sixty-six Bos taurus mandibular remains (mandible
Iron Age levels at Kadica Brdo. Eighteen mandibles and
one loose tooth were undeterminable to age, leaving
eighteen mandibles and twenty-nine loose teeth in the
final analysis (n=47 Appendix A). Table 7.15
summarizes the age stage distributions of Bos taurus
remains from the Early Iron Age.
Kadica Brdo is a small site (100x50 m), surrounded by a

thick fortification wall, which preserved the internal
stratigraphy at the site. Only one trench was dry sieved,
but selected samples across the site were also floated.
Approximately 10% of the spatial contexts were
systematically recovered. Over 20,000 fragments were
recovered and analyzed from the 1980, and 1987-89
excavations at the site (Greenfield 2005b).
Seventy-two Sus scrofa mandibles and loose teeth were

recovered from Early Iron Age levels at Kadica Brdo (65
mandibles and 7 loose teeth). Twenty-three mandibles
were unidentifiable to age, leaving forty-two mandibles
and seven loose teeth in the final analysis (n=49
Appendix A). Table 7.16 summarizes the age stage
distributions of Sus scrofa remains from the Early Iron
Age.

Fifty-seven Ovis aries mandibular remains (mandibles
Iron Age deposits at Kadica Brdo. Ten mandibles were
unidentifiable to age, leaving thirty-nine mandibles and
eight loose teeth in the final analysis (n=47 Appendix
A). Table 7.13 summarizes the age stage distributions of
Ovis aries remains from the Early Iron Age.
V. Livade
A. Site description
Livade is a lowland site located on the right bank of the
Danube in northeastern Serbia at the western edge of the
Dacian Plain (Figure 1.1). On the left bank of the river,
the broad Dacian plain begins. The site is located in the
wide alluvial plain that borders the Danube, just beyond
the eastern edge of the Iron Gates, on the eastern side of
the mountains of eastern Serbia. The mountains of East
Serbia are less than 5 km distant. It is near the modern
village of Mala Vrbica, on the eastern edge of the area of
investigation.
Eleven Capra hircus mandibular remains (mandibles n=10 and loose teeth n=1) were recovered from Early
Iron Age deposits at Kadica Brdo. Three mandibles
were unidentifiable to age, leaving seven mandibles and
one loose tooth in the final analysis (n=8). The age stage
distribution data are summarized in Appendix B (Table
B8). Since these data were too small to be used on their
own in the analysis, this sample was combined with the
Ovis/Capra remains for the analysis (below).There were
a total of 254 mandibular remains (mandible n=153
and loose teeth n=101) of Ovis/Capra recovered from
Early Iron Age levels at Kadica Brdo. Forty-two
mandibles and four loose teeth were undeterminable to
age, leaving 111 mandibles and 97 loose teeth in the
summarizes the age stage distributions of Ovis/Capra
The site contains ceramic and other artifactual remains

with strong cultural affinities to the other sites in the study
region. The ceramic assemblage and figurines place Livade
in the Dubovac-uto Brdo ceramic group. Some ceramic
elements also reveal affiliation with the Verbicioara
ceramic group across the river in Romanian Oltenia.
46

loose teeth from Early Iron Age Kadica Brdo
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
4
5
20
5
1
0
1
0
0
36
Corrected Count
%
11
14
55
14
3
0
3
0
0
100
No.
4
5
25.6
9
1.4
0
2
0
0
47
%
8
10
54
20
4
0
4
0
0
100

Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
1
1
15
10
5
16
17
3
3
71
Corrected Count
%
1
1
22
14
7
23
24
4
4
100
No.
1
1
25.5
40.5
19.2
69.8
39.8
5.6
5.6
208
%
1
1
12
19
9
34
18
3
3
100

Stage
Suggested Age
Raw Count
Corrected Count
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
Young adult
Adult
old adult
No.
4
0
3
8
0
2
1
1
6
%
16
0
12
32
0
8
4
4
24
No.
4
0
3
14.5
0
7.1
3.8
3.3
11.2
%
8
0
6
32
0
15
8
7
24
25
100
46.9
100
Senile
47

Table 7.16. Stage distribution of Sus scrofa mandibles and
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
adult
old adult
senile
No.
2
11
4
7
6
0
0
0
1
31
Corrected Count
%
6
36
13
23
19
0
0
0
3
100
As a result, Livade is considered to be a Late Bronze Age

site. Chronologically, the Dubovac-uto Brdo horizon falls
within the Middle-Late Bronze ages of the central Balkans
(Reinecke Br A2/B1 to D ca. 1600-1200 bc)
(Greenfield 1986a).
No.
2
11.7
10
17.7
6.5
0
0
0
1
48.9
%
4
25
20
36
13
0
0
0
2
100
and one loose tooth in the final analysis (n=27

distributions of Sus scrofa remains from the Late
Bronze Age.
VI. Ljuljaci
The size of Livade is difficult to determine because much

of the site was buried beneath a thick sandy horizon
(from river flooding). It extends for a distance of up to
100 m from the riverbank, and was investigated over a
length of 200 m along the river. The sandy horizon
preserved the site so that there was little mixing of
remains in the cultural horizon (contrary to contemporary
sites that are usually disturbed by plowing). None of the
sediments from the site was sieved or floated. Over 5,000
fragments were recovered and analyzed from the 1980-81
excavations at the site (Greenfield 1986a).
A. Site description
Ljuljaci is found on the top of the hill, Milica Brdo, in the
village of Ljuljaci, about 20 km west of the city of
Kragujevac, in Serbia (Figure 1.1). It is located on the
western half of a natural plateau overlooking the
surrounding valleys at approximately 300 m asl and is
considered to be a mid-altitude site. Three sides of the
plateau are steep, leaving the eastern side as the only
approach to the site. It is considered to be a fortified position
due to its less accessible location and the possible palisade
and ditch along one side. A small stream, the Glavica, winds
itself around the base of the plateau on the western, northern
and eastern sides. The plateau area was subject to
deforestation and cultivation during its period of occupation.
The surrounding environment is one of rolling hills and low
mountains covered by mixed oak forests and some
agricultural cultivation (Greenfield 1986a, 2006).

Sixteen domestic Ovis/Capra mandibles (n=15) and loose
teeth (n=1) were recovered from Late Bronze Age levels at
Livade. This included a single mandible of Ovis aries. Four
mandibles and one loose tooth were unidentifiable to age,
leaving eleven mandibles in the final analysis (Appendix A).
Ovis/Capra remains from the Late Bronze Age.
Ljuljaci includes material from the late Early Bronze Age

and the Middle Bronze Age. The stratigraphy of the site is
relatively simple consisting of three layers. Dates are based
on ceramic groups, and cross-dating with other sites.
Twenty-one
Bos
taurus
mandibular
remains
(mandibles n=14 and loose teeth n=7) were
recovered from Late Bronze Age levels at Livade. Four
mandibles and one loose tooth were undeterminable to
age, leaving ten mandibles and six loose teeth in the
summarizes the age stage distributions of Bos taurus
remains from the Late Bronze Age.
1.
2.
3.
Forty-two Sus scrofa mandibular elements (mandibles

n=41 and loose teeth n=1) were recovered from Late
Bronze Age levels at Livade. Fifteen mandibles were
undeterminable to age, leaving twenty-six mandibles
Ljuljaci I Early Bronze Age (ca. 1950 bc,

uncalibrated)
Ljuljaci II late Early Bronze Age- early Middle
Bronze Age (1730-1690 bc)
Ljuljaci III Middle Bronze Age (ca. 1600-1550 bc)
(Greenfield 1986a: 125).
The site is relatively small, 100x50 m. In the early

periods (Ljuljaci I and II), the settlement remained at
about the same size. It expanded in size in the final period
48

Table 7.17. Stage distribution of Ovis/Capra mandibles from Late Bronze Age Livade
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
Raw Count
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
0
0
1
0
2
3
2
0
8
Corrected Count
%
0
0
0
12
0
25
38
25
0
100
No.
0
0
0
1
0
2.8
5.01
2.19
0
11
%
0
0
0
9
0
25
46
20
0
100
Table 7.18. Stage distribution of Bos taurus mandibles from Late Bronze Age levels of Livade
Stage
Suggested Age
Raw Count
Corrected Count
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
Young adult
Adult
Old adult
No.
0
0
1
1
0
0
0
1
2
%
0
0
20
20
0
0
0
20
40
No.
0
0
1
3.86
0
1.3
1.58
3.19
5.01
%
0
0
6
24
0
8
10
21
31
100
15.94
100
Senile
Table 7.19. Stage distribution of Sus scrofa mandibles and loose teeth from Late Bronze Age Livade
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
Raw Count
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
Adult
Old adult
Senile
No.
0
1
2
7
7
1
0
0
0
18
49
Corrected Count
%
0
6
10
39
39
6
0
0
0
100
No.
0
2
4
9.5
10.38
1.12
0
0
0
27
%
0
8
15
35
38
4
0
0
0
100

(Ljuljaci III). The stratigraphy was relatively thin, with a
maximum depth of c. 1 m beneath the surface (except for
the occasional pit). None of the site was sieved or floated.
Over 5,000 fragments were recovered and analyzed from
the 1975-77 excavations at the site (Greenfield 1986a).
VII. Megalo Nisi Galanis

A. Site description
Megalo Nisi Galanis is in the southern Ptolemais basin,
located in Greek Macedonia near the city of Kozani. The
Ptolemais basin is the southernmost extension of the
Pelagonian plain stretching from Greek Macedonia into the
Former Yugoslav Republic of Macedonia (Figure 1.1).
The plateau is surrounded by mountainous height. The site
is considered a mid-altitude site (c. 500 m asl) in terms of
regional long distance transhumant pastoralists. It is a
small site, with thick deposits, ranging from a depth of 23m, covering an area of c. 100 x 100 m. The entire
excavated deposit was dry and wet sieved. The site is
located on a paleo-lake drained in the 19th century, and
falls within the Mediterranean climatic zone. Surrounding
heights are currently used for grazing, the flats of the
plateau for agriculture. The excavation uncovered material
from the Late Neolithic, Final Neolithic (=Eneolithic of
central Balkans), and Early Bronze Age. Underlying
Middle and Final Neolithic deposits exist but are not yet
analyzed (Greenfield and Fowler 2003, 2005).

Five Ovis/Capra mandibles and loose teeth were
recovered from Early Bronze Age levels at Ljuljaci (4
mandibles and 1 loose tooth). Two mandibles were
unidentifiable to age, leaving two mandibles and one
loose tooth in the final analysis (n=3 Appendix A).
Since these data were too few to be used on their own,
they were combined with the rest of the Bronze Age
material and designated Early/Middle Bronze Age.
There were a total of ten Ovis/Capra mandibles and
loose teeth from the site (9 mandibles and 1 loose
tooth); including one mandible identified as Ovis aries.
Only one mandible was unidentifiable to age, that of the
Ovis aries specimen. Therefore, only eight mandibles
and one loose tooth were included in the final analysis
(n=9). Lumping of the data still results in too small a
sample size for inclusion in the analysis. Age stage
distribution data are summarized in Appendix B (Tables
B9-B11).
Though out of the geographical area of study, Megalo

Nisi Galanis is included for two main reasons; 1) it is a
large and systematically collected and analyzed faunal
sample from western Macedonia spanning the critical
Late Neolithic-Post Neolithic temporal divide that is the
focus of this investigation; 2) its inclusion in the sample
provides a connection with the southern Balkans. Over
20,000 fragments were recovered and analyzed from the
1980-1986 excavations at the site (Greenfield and Fowler
2003, 2005).
Seven Bos taurus mandibles were recovered from the

Early Bronze Age levels at Ljuljaci. Five mandibles
were unidentifiable to age, leaving two mandibles in the
final analysis (Appendix A). Since these data were too
few to be used by themselves, all of the Bronze Age
material from the site was combined and designated
Early/Middle Bronze Age. Age stage distribution data for
the Early Bronze Age are summarized in Appendix B
(Table B12). To increase sample size, the Middle Bronze
Age assemblage was combined with Early Bronze Age
material. There were a total of twenty-four Bos taurus
mandibles (n=16) and teeth (n=8). Eleven mandibles
five mandibles and seven loose teeth in the final
analysis (n=12 Appendix A). Table 7.20 summarizes
the age stage distributions of Bos taurus remains from
the Early/Middle Bronze Age.

Three Ovis aries mandibles were recovered from Late
Neolithic/Final Neolithic levels at Megalo Nisi Galanis.
All were identifiable to age, and all were included in the
final analysis (n=3). Table 7.22 summarizes the age stage
distributions of Ovis aries remains from the Late
Neolithic/Final Neolithic.
Fourteen domestic Ovis/Capra mandibles were recovered
from Late Neolithic/Final Neolithic levels at Megalo Nisi
Galanis. Five were unidentifiable to age, leaving nine
mandibles in the final analysis (n=9 Appendix A). Table
7.23 summarizes the age stage distributions of Ovis/Capra
remains from the Late Neolithic/Final Neolithic.
There were five ageable remains recovered from Early

Bronze Age deposits and six ageable remains recovered
from Middle Bronze Age deposits and age stage
distributions are detailed in Appendix B (Tables B13
and B14). To increase sample size, all Bronze Age
material was combined for Sus scrofa. In total, there
were forty-eight domestic Sus scrofa mandibles (n=46)
and loose teeth (n=2) from the Bronze Age deposits.
Twenty-eight of the mandibles were unidentifiable to
age, leaving eighteen mandibles and two loose teeth in
the final sample (n=20 Appendix A). Table 7.21
summarizes the age stage distributions of Sus scrofa
remains from the Early/Middle Bronze Age.
Nine Ovis aries mandibles and loose teeth were

recovered from Final Neolithic levels at Megalo Nisi
Galanis (7 mandibles and 2 loose teeth). Four mandibles
could not be assigned an age, leaving three mandibles and
two loose teeth in the final analysis (n=5). Table 7.24
summarizes the age stage distributions of Ovis aries
remains from the Final Neolithic.
50

Table 7.20. Stage distribution of Bos taurus mandibles from Early/ Middle Bronze Age Ljuljaci
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
Raw Count
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
Young adult
Adult
Old adult
Senile
No.
0
2
2
2
0
0
0
0
1
7
Corrected Count
%
0
28
28
28
0
0
0
0
15
99
No.
0
2
2
2
0
0.9
2.6
0
2.9
12.4
%
0
16
16
16
0
7
22
0
23
100

loose teeth from Early/Middle Bronze Age Ljuljaci
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
Adult
No.
0
1
1
5
5
1
0
0
%
0
8
8
38
38
8
0
0
No.
0
1
1.2
7.3
9
1.5
0
0
%
0
5
6
37
45
7
0
0
0
13
0
100
0
20
0
100
Old adult
Senile
Corrected Count
Table 7.22. Stage distribution of Ovis aries mandibles and loose teeth from
Late Neolithic/Final Neolithic Megalo Nisi Galanis
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
Raw Count
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
0
1
0
0
0
0
0
0
1
Corrected Count
%
0
0
100
0
0
0
0
0
0
100
51
No.
0
0
2
0.5
0.5
0
0
0
0
3
%
0
0
66
17
17
0
0
0
0
100

Table 7.23. Stage distribution of Ovis/Capra mandibles and loose teeth from
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
0
2
0
0
0
0
0
0
2
Corrected Count
%
0
0
100
0
0
0
0
0
0
100
No.
0
0
3
0.5
1
1.9
1.3
0.6
0.6
8.9
%
0
0
33
6
11
21
15
7
7
100

loose teeth from Final Neolithic Megalo Nisi Galanis
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
1
0
2
0
0
0
0
0
0
3
Corrected Count
%
33
0
67
0
0
0
0
0
0
100
No.
1
0
4
0
0
0
0
0
0
5
%
20
0
80
0
0
0
0
0
0
100

Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
1
0
5
3
2
1
1
0
0
13
Corrected Count
%
8
0
38
23
15
8
8
0
0
100
52
No.
1
0
8.8
11.7
8.6
5.8
8.1
0
0
44
%
2
0
21
26
20
13
18
0
0
100

was too small to be included in the analysis. The age stage
distributions are detailed in Appendix B (Table B17). There
were no Sus species remains for the Final Neolithic/Early
Bronze Age period.
Fifty-seven Ovis/Capra mandibles and loose teeth were

recovered from Final Neolithic levels at Megalo Nisi
Galanis (24 mandibles and 33 loose teeth). Twelve
mandibles and one loose tooth were unidentifiable to age,
leaving twelve mandibles and thirty-two loose teeth in the
remains from the Final Neolithic.
VIII. Novaka uprija

A. Site description
Novaka uprija is a lowland site that is located roughly
5 km north of the city of Smederevska Palanka in central
Serbia (Figure 1.1). It is located on the western edge of a
low plateau east of a small tributary of the Jasenica River.
The site is located in a non-obtrusive position, offering no
significant defensive advantage. In close proximity are
suitable agricultural land and access to water and wood
supplies (Greenfield 1986a).
Six Ovis aries mandibles and loose teeth were recovered

from Final Neolithic/Early Bronze Age levels at Megalo
Nisi Galanis (5 mandibles and 1 loose tooth). One
mandible was unidentifiable to age, leaving four
mandibles and one loose tooth in the final analysis (n=5
distributions of Ovis aries remains from the Final
Neolithic/Early Bronze Age. As these samples were too
small for effective analysis, they were combined with the
Ovis/Capra remains.
The site is divided into four areas, of which only the data
from the central area is relevant and utilized. This area is
approximately four hectares in area. The plateau on
which the site is situated is generally flat. It increases in
width and rises slightly to the east (Greenfield 1986a).
Eighteen Ovis/Capra mandibles and loose teeth were

recovered from Final Neolithic/Early Bronze Age levels at
Megalo Nisi Galanis (8 mandibles and 10 loose teeth).
Four mandibles and one loose tooth were unidentifiable to
age, leaving four mandibles and nine loose teeth in the
final analysis (n=13 Appendix A). Table 7.27 summarizes
the age stage distributions of Ovis/Capra remains from the
Final Neolithic/Early Bronze Age.
Novaka uprija is a lowland site that includes material

from the Eneolithic through to the Roman period,
although there is no evidence for occupational continuity
over this entire time span. The most intensive occupation
took place during the Early Bronze Age (Greenfield
1986a).
There were insufficient Late Neolithic/Final Neolithic

remains of Bos taurus from Megalo Nisi Galanis to
include in the analysis. Only two mandibles were
recovered, both unidentifiable to age (Appendix A).
The prehistoric settlements at Novaka uprija represents

the remains of a series of small villages that laterally
moved up and down the length of the terrace overlooking
the stream. Each settlement probably represented no more
than 100x100 m in size. Because the site has moved
laterally over time, the total length of the site is almost 1
km. The deposits were relatively thin since there was
little vertical deposition. Most of the material above the
pit levels were disturbed and mixed by modern and
ancient plowing. The soils from the entire excavation
were sieved and/or floated. Over 5,000 bone fragments
were recovered and analyzed from the 1977 and 1980
excavations at the site (Greenfield 1986a, 1986b).
Ten Bos taurus mandibular remains (mandibles n=4 and

loose teeth n=6) were recovered from Final Neolithic
levels at Megalo Nisi Galanis. Two mandibles were
unidentifiable to age, leaving two mandibles and six loose
teeth in the final analysis (n=8 Appendix A). This sample
was too small to be included in the analysis. The age stage
distributions are detailed in Appendix B (Table B15)
Two Bos taurus mandibles and loose teeth were
recovered from Final Neolithic/Early Bronze Age levels
at Megalo Nisi Galanis (1 mandible and 1 loose tooth),
both are unidentifiable to age. As a result, there was no
sample included in the analysis (Appendix A).

Ten domestic Ovis/Capra loose teeth were recovered
from Eneolithic levels at Novaka uprija. All the teeth
were assigned an absolute age and were included in the
Eneolithic.
Two Sus scrofa mandibles were recovered from Late

Neolithic/Final Neolithic levels at Megalo Nisi Galanis.
Only one was identifiable to age. As a result, the sample
size is one. This sample is too small for inclusion in the
analysis. Stage distribution data is included in Appendix
B (Table B16).
Thirty-six Ovis/Capra mandibular remains (mandibles n=12 and loose teeth n=24) were recovered from Early
Bronze Age levels at Novaka uprija. Four mandibles
were unidentifiable to age, leaving eight mandibles and
twenty-four loose teeth in the final analysis (n=32
Nine Sus scrofa mandibular remains (mandibles - n=6 and

loose teeth n=2) were recovered from Final Neolithic
levels at Megalo Nisi Galanis. Two mandibles were
unidentifiable to age, leaving four mandibles and two loose
teeth in the final analysis (n=6 Appendix A). This sample
53

Table 7.26. Stage distribution of Ovis aries mandibles and loose teeth from
Final Neolithic/Early Bronze Age Megalo Nisi Galanis
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
1
2
2
0
0
0
0
0
5
%
0
20
40
40
0
0
0
0
0
100
Table 27. Stage distribution of Ovis/Capra mandibles and loose teeth from
Final Neolithic/Early Bronze Age Megalo Nisi Galanis
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
1
2
3
0
0
0
0
0
6
Corrected Count
%
0
17
33
50
0
0
0
0
0
100
No.
0
1
2
4.7
0.6
1.2
1
1.5
1
13
%
0
8
15
35
5
9
8
12
8
100

loose teeth from Eneolithic Novaka uprija
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
Raw Count
No.
0
0
0
1
0
3
0
0
1
5
54
%
0
0
0
20
0
60
0
0
20
100
Corrected Count
No.
0
0
0
1.5
0
6.5
0
0
2
10
%
0
0
0
15
0
65
0
0
20
100

loose teeth from Early Bronze Age Novaka uprija
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
Raw Count
No.
0
1
1
8
1
3
1
0
2
17
%
0
6
6
46
6
18
6
0
12
100
Corrected Count
No.
0
1.5
1.5
10.9
1.7
5
7.8
0
3.6
32
%
0
5
5
34
5
16
24
0
11
100

loose teeth from Late Bronze Age Novaka uprija
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
Raw Count
No.
0
0
1
1
1
1
0
0
0
4
distributions of Ovis/Capra remains from the Early Bronze

Age.
%
0
0
25
25
25
25
0
0
0
100
Corrected Count
No.
%
0
0
0
0
1
6
1.6
10
3.6
23
3.6
23
4.5
28
1.6
10
0
0
15.9
100
Four Bos taurus remains (mandibles n=1 and loose

teeth n=3) were recovered from Late Bronze Age levels
at Novaka uprija. All were assigned an absolute age
and the final sample size was four. These samples were
too small to be included in the analysis. The age stage
distributions are detailed in Appendix B (Tables B18B20).
Seventeen Ovis/Capra mandibular remains (mandibles

Bronze Age deposits at Novaka uprija. One mandible
was unidentifiable to age, leaving three mandibles and
thirteen loose teeth in the final analysis (n=16 Appendix
A). Table 7.30 summarizes the age stage distributions of
Ovis/Capra remains from the Late Bronze Age.
Nine Sus scrofa remains (mandibles n=5 and loose

teeth n=4) were recovered from Eneolithic levels at
Novaka uprija. One mandible was unidentifiable to
age, leaving four mandibles and four loose teeth in the
final analysis (n=8 Appendix A). This sample was too
small to be included in the analysis. The age stage
distributions are detailed in Appendix B (Table B21).
An insufficient sample size of Bos taurus remains exists

for Novaka uprija for any period. Three Bos taurus
mandibles were recovered from Eneolithic levels at
Novaka uprija. One mandible was unidentifiable to
age, leaving two mandibles in the final analysis
(Appendix A).
Sixteen Sus scrofa remains (mandibles n=11 and loose

teeth n=5) were recovered from Early Bronze Age levels
of Novaka uprija. Five mandibles and one loose tooth
were unidentifiable to age, leaving six mandibles and four
loose teeth in the final analysis (n=10 Appendix A). Table
7.31 summarizes the age stage distributions of Sus scrofa
remains from the Early Bronze Age.
Nine Bos taurus mandibular remains (mandibles n=5

and loose teeth n=4) were recovered from Early Bronze
Age levels at Novaka uprija. Three mandibles were
unidentifiable to age, leaving two mandibles and four
55

Four Sus scrofa remains (mandibles n=3 and loose
teeth n=1) were recovered from Late Bronze Age
deposits at Novaka uprija. One mandible was
unidentifiable to age, leaving two mandibles and one
This sample was too small to be included in the
analysis. The age stage distributions are detailed in

Neolithic Opovo. Nine mandibles were undeterminable to
age, leaving ten mandibles and one loose tooth in the
final sample (n=11 Appendix A). Table 7.33
remains from the Late Neolithic deposits at Opovo.
IX. Opovo
A. Site description
A. Site description
Petnica is located in western Serbia, c. 5 km from the city

of Valjevo. It is at the edge of the village of Petnica
(Figure 1.1). The site is located on a slope overlooking a
depression at the end of a valley. The site is situated at
the base of a steep escarpment. Above (to the south) the
site is a series of plateaus and ridges, beyond which the
mountains of western Serbia rise. The highlands above
and the slopes around the site are still heavily forested
and present a dissected landscape of flat and sloping
terrain interspersed with steeper inclines. Below the site,
a small stream flows down to the Kolubara River (which
is a tributary of the Sava). The valley is a rich agricultural
area, but forests cover the ridges and mountains to the
south of the site (Greenfield 1986a).
X. Petnica
Opovo is a Late Neolithic settlement in the Banat, a

region of northern Serbia located between the Danube
and Romania (Figure 1.1). Opovo is situated on a
slightly elevated location and was surrounded on three
sides by an extensive series of annually flooded lowland
marshes. The area around the site is not well suited to
either crop cultivation or the herding of domestic
animals. Drainage is poor due to the high water table.
However, a wide range of wild resources including fish,
birds, shellfish, amphibians, reptiles, wild pig, roe and
red deer was available in the aquatic and semi-aquatic
environments of the surrounding habitat (Greenfield
1986a).
There were five major occupation horizons at the site.

From the earliest to the latest, Middle Neolithic Vina B,
Late Neolithic Vina C, Late Neolithic Vina D,
Eneolithic Baden-Kostalac, and late Bronze-Early Iron
Age Halstatt A, B and C (Greenfield 1986a). Most of the
Halstatt occupation is Halstatt A and therefore, most is
considered to be Late Bronze Age.
Ceramic analysis indicated the presence of two phases of

the Late Neolithic Vina-Plonik period (Tringham et al.
1985). The data analyzed here derive from the earlier
Yugoslav excavations at the site and the phases were not
separable stratigraphically at the time. As a result, the
faunal remains were analyzed as a single temporal unit.
None of the site was sieved or floated. Over 2,000
fragments were recovered from excavation of three
trenches during the 1979-1980 excavations at the site.
These were analyzed in 1982 (Greenfield 1986a).
Subsequent excavations produced a large sample of
remains that were unavailable for comparative analysis
due to the revision of phasing (Russell 1998; Tringham et
al. 1985, 1992).
Petnica is a small site, occupying roughly one hectare.

The stratigraphy extends to depth of 2-3 m across much
of the site. Approximately 20% of the site was sieved or
floated (Greenfield 1986a). Over 20,000 fragments were
recovered and analyzed from the 1980-1986 excavations
at the site (Greenfield 1991, 1999b, n.d. c).

Nine Ovis/Capra mandibular remains (mandibles n=6
and loose teeth n=3) were recovered from Middle
Neolithic levels at Petnica. Two mandibles were
unidentifiable to age, leaving four mandibles and three
loose teeth in the final analysis (n=7 Appendix A). This
sample was too small to be included in the analysis. The
age stage distributions are detailed in Appendix B (Table
B24).
Six Ovis/Capra remains (mandibles n=5 and loose teeth

n=1) were recovered from Late Neolithic levels at
Opovo. All were identifiable to age (Appendix A). This
sample was too small to be included in the analysis. The
age stage distributions are detailed in Appendix B (Table
B23).
Nineteen Bos taurus mandibular remains (mandibles
Neolithic levels at Opovo. Five mandibles were
unidentifiable to age, leaving seven mandibles and seven
taurus remains from the Late Neolithic.
Twenty-one Ovis/Capra remains (mandibles n=7 and

loose teeth n=14) were recovered from Late Neolithic
levels at Petnica. Two mandibles were unidentifiable to
age, leaving five mandibles and fourteen loose teeth in
the final analysis (n=19 Appendix A). Table 7.34
remains from the Late Neolithic.
Twenty Sus scrofa mandibular remains (mandibles

56

Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
Adult
Old adult
Senile
No.
0
0
0
3
4
0
0
0
0
7
Corrected Count
%
0
0
0
43
57
0
0
0
0
100
No.
0
1
1
4
4
0
0
0
0
10
%
0
10
10
40
40
0
0
0
0
100

loose teeth from Late Neolithic Opovo
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
Young adult
Adult
Old adult
Senile
No.
0
0
0
6
0
1
0
1
0
8
Corrected Count
%
0
0
0
75
0
12
0
12
0
99
No.
0
0.5
0.5
7
0
4.01
0
1.99
0
14
%
0
4
4
50
0
28
0
14
0
100

Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
Adult
Old adult
Senile
No.
0
2
2
3
2
0
0
0
0
9
Corrected Count
%
0
22
22
33
22
0
0
0
0
99
57
No.
0
2
2.4
3.6
3
0
0
0
0
11
%
0
18
22
33
27
0
0
0
0
100

loose teeth from Late Neolithic Petnica
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
0
2
2
1
2
1
1
0
9
Corrected Count
%
0
0
22
22
11
22
11
11
0
99
No.
0
0
3.5
4.2
1.9
4.8
3.2
1.5
0
19.1
%
0
0
18
22
10
25
17
8
0
100

loose teeth from Eneolithic Petnica
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
0
4
2
1
1
1
0
0
9
Corrected Count
%
0
0
45
22
11
11
11
0
0
100
Seven domestic Ovis/Capra mandibles and loose teeth

were recovered from Late Neolithic/Eneolithic levels at
Petnica (3 mandibles and 4 loose teeth); including one
loose tooth of domestic Ovis aries. Only the Ovis aries
loose tooth was undeterminable to age, leaving three
mandibles and three loose teeth in the final analysis (n=6
Appendix A). This sample was too small to be included
in the analysis. The age stage distributions are detailed in
No.
0
0
6.5
3.3
1.1
1.1
1
0
0
13
%
0
0
50
26
8
8
8
0
0
100
As such, the final analysis included four mandibles and

nine loose teeth (n=13 Appendix A). Table 7.36
remains from the Late Bronze Age.
Eleven Ovis/Capra mandibular remains (mandibles - n=4
and loose teeth n=7) were recovered from Late Bronze
Age/Early Iron Age deposits at Petnica. One mandible
three mandibles and six loose teeth in the final analysis
(n=9 Appendix A). This sample was too small to be
included in the analysis. The age stage distributions are
detailed in Appendix B (Table B26).
Thirteen Ovis/Capra mandibles and loose teeth were

recovered from Eneolithic levels at Petnica (9 mandibles
and 4 loose teeth). All elements were assigned an age.
The final sample size is thirteen (Appendix A). Table 7.35
remains from the Eneolithic.
Thirty-four Bos taurus mandibular remains (mandibles

n=27 and loose teeth n=7) were recovered from Middle
Neolithic deposits at Petnica. Twelve mandibles were
unidentifiable to age, leaving fifteen mandibles and seven
loose teeth in the final sample (n=22 Appendix A).
Fourteen Ovis/Capra remains (mandibles n=5 and loose

teeth n=9) were recovered from Late Bronze Age levels
at Petnica. Only one mandible was indeterminable to age.
58

taurus remains from the Middle Neolithic.
Twelve Sus scrofa mandibular remains (mandibles
n=10 and loose teeth n=2) were recovered from
Eneolithic levels at Petnica. Two mandibles were
unidentifiable to age, leaving eight mandibles and two
Table 7.41 summarizes the age stage distributions of Sus
scrofa remains from the Eneolithic.
Thirty-six Bos taurus mandibular remains (mandibles

Neolithic levels at Petnica. Fifteen mandibles and two loose
teeth were unidentifiable to age, leaving four mandibles and
fifteen loose teeth in the final sample (n=19 Appendix A).
taurus remains from the Late Neolithic.
Seven Sus scrofa mandibular remains (mandibles n=5

Age levels at Petnica. Two mandibles were unidentifiable
to age, leaving three mandibles and two loose teeth in the
All subsequent periods (Late Neolithic through the Early

Iron Age) from Petnica have insufficient sample size for
inclusion in the analysis. Four Bos taurus mandibular
remains (mandibles n=2 and loose teeth n=2) were
recovered from Late Neolithic/Eneolithic levels. Both
mandibles were unidentifiable to age, leaving two loose
teeth in the final analysis (Appendix A). It can be noted
that both the loose teeth were aged to cover mandibular
stages G and I (adult-senile).
Seven Sus scrofa mandibular remains (mandibles n=6

Age/Early Iron Age levels at Petnica. Two mandibles
were unidentifiable to age, leaving four mandibles and
one loose tooth in the final analysis (n=5 Appendix A).
This sample was too small to be included in the analysis.
The age stage distributions are detailed in Appendix B
(Table B31).
Twelve Bos taurus mandibular remains (mandibles n=6

and loose teeth n=6) were recovered from Eneolithic
deposits. Three mandibles were unidentifiable to age,
leaving three mandibles and six loose teeth in the final
analysis (n=9 Appendix A). Twelve Bos taurus
mandibular remains (mandibles n=2 and loose teeth
n=10) were recovered from Late Bronze Age deposits.
Two mandibles and one loose tooth were unidentifiable
to age, leaving nine loose teeth in the final analysis (n=9).
XI. Selevac
Site description
Selevac (also known as Selevac-Staro Selo) is a
settlement dated to the Middle to Late Neolithic in central
Serbia, near the modern village of Selevac. It is nestled in
the hills to the west of the Morava River, close to its
confluence with the Danube. It is on the south-facing
slope of the hill above a large stream (the Vrbica) which
eventually drains into the Morava River. The site is close
to well-drained and fertile soils, a variety of forest and
riverine resources, and high quality clay sources. The site
is located on the nearest high ground above the Morava
Rivers floodplain (150 m asl). It should be considered a
lowland occupation in the context of this study.
Eight Bos taurus mandibles and loose teeth were

recovered from the Late Bronze Age/Early Iron Age
levels at Petnica (4 mandibles and 4 loose teeth). Two
mandibles were unidentifiable to age, leaving two
mandibles and four loose teeth in the final analysis (n=6).
All these samples were too small to be included in the
analysis. The age stage distributions are detailed in
Appendix B (Tables B27-B29)).
Thirty-two Sus scrofa mandibular remains (mandibles
n=31 and loose teeth n=1) were recovered from Middle
Neolithic deposits at Petnica. Nineteen mandibles were
unidentifiable to age, leaving twelve mandibles and one
scrofa remains from the Middle Neolithic.
The site covers a relatively large area (53 ha), with a

maximum dimension of 780x1080 m (Tringham and
Krsti 1990). Excavations took place over three field
seasons (1976-1978). The deposits ranged in thickness
from c. 1.5 to 3 m in thickness. The site has two main
periods of occupation, an earlier phase known as VinaTordo and a later phase referred to as Vina-Plonik.
Four stratigraphic-architectural phases were identified
during excavation (I-IV). These corresponded with the
latter half of the Vina-Tordo (I-II) and to the earlier
half of the Vina-Plonik (III-IV) culture phasing.
Between these two cultures lies the intermediate Gradac
phase of the Vina culture, which is present at the site.
The Gradac or transitional phase was originally
considered to be part of the Late Neolithic, but was
absent from the type-site and identified at the Gradac site
Twenty-four Sus scrofa mandibular remains (mandibles

Neolithic deposits at Petnica. Eleven mandibles were
unidentifiable to age, leaving eleven mandibles and two
scrofa remains from the Late Neolithic.
Two Sus scrofa mandibles were recovered from Late
Neolithic/Eneolithic deposits at Petnica. Both were
unidentifiable to age. There was no Late
Neolithic/Eneolithic Sus scrofa sample for Petnica.
59

loose teeth from Late Bronze Age Petnica
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
1
1
0
0
3
2
0
0
7
Corrected Count
%
0
14
14
0
0
43
29
0
0
100
No.
0
1
2
0.5
0.9
5.2
3.4
0
0
13
%
0
8
15
4
7
40
26
0
0
100

loose teeth from Middle Neolithic Petnica
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
young adult
adult
old adult
senile
No.
2
0
1
4
0
1
0
0
2
10
Corrected Count
%
20
0
10
40
0
10
0
0
20
100
No.
2
0
1
8.2
0
3.8
0
0
6.9
21.9
%
9
0
4
38
0
17
0
0
32
100

loose teeth from Late Neolithic Petnica
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
Raw Count
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
young adult
adult
old adult
senile
No.
1
0
0
3
1
0
0
0
1
6
Corrected Count
%
17
0
0
50
17
0
0
0
17
101
60
No.
1
1
1
5.8
5.5
0.8
1.6
0
2.3
19
%
5
5
5
31
30
4
8
0
12
100

Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
adult
old adult
senile
No.
1
5
2
2
0
0
0
0
2
12
Corrected Count
%
8
41
17
17
0
0
0
0
17
100
No.
1
5
2
3
0
0
0
0
2
13
%
8
39
15
23
0
0
0
0
15
100
Table 7.40. Stage distribution of Sus scrofa mandibles and loose teeth from Late Neolithic Petnica
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
adult
old adult
senile
No.
0
2
3
2
0
0
0
0
1
8
Corrected Count
%
0
25
38
25
0
0
0
0
12
100
No.
0
2.4
3.6
6
0
0
0
0
1
13
%
0
18
28
46
0
0
0
0
8
100
Table 7.41. Stage distribution of Sus scrofa mandibles and loose teeth from Eneolithic Petnica
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
Raw Count
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
adult
old adult
senile
No.
0
2
0
1
0
0
0
0
1
4
61
Corrected Count
%
0
50
0
25
0
0
0
0
25
100
No.
0
4
0
4
0.5
0.5
0
0
1
10
%
0
40
0
40
5
5
0
0
10
100

near Ni in southern Serbia. Some later material was
found (Roman and Medieval), but is present in very
minor deposits. The Neolithic occupation horizons have
been dated to 4390-3650 bc (5020-4400 BC, calibrated)
(Tringham and Krsti 1990). Through an examination of
the frequency of data by architectural-stratigraphic phase
from the site, it would seem that most of the relevant data
are from the later Neolithic Vina-Plonik phase (Legge
1990; Tringham and D. Krsti 1990).
Mandibular and loose tooth remains

Four domestic Ovis aries mandibles were recovered from
Middle Neolithic levels at Stragari. All were identifiable
to an age and included in the final analysis. Appendix B
(Table B32) summarizes the age stage distributions of
Ovis aries remains. Only one domestic Capra hircus
mandible was recovered from Stragari. It was
unidentifiable to age. Due to these limited samples sizes,
the Ovis aries and Capra hircus remains were lumped
into the category Ovis/Capra to provide an adequate
sample size.

Selevac is the only site in the sample that was not
collected by Greenfield. But the nature of the analysis
and publication makes it useful for comparison to the
other samples and for inclusion in this study. Recovery of
remains from the site was systematic, and included both
dry (0.7 mm) and wet sieving, over the entire area of
excavation. Preservation of bones was excellent. The
zooarchaeological analysis is based on the 7,035 bones
identified to genus or species. Harvest profiles for the
major domestics (sheep/goat, cattle and pigs) were
generated by Legge (1990). In our reanalysis, all
mandibles from all Middle and Late phases at Selevac are
combined and considered as a single Middle/Late
Neolithic assemblage.
Twenty-four Ovis/Capra remains (mandibles n=9 and

loose teeth n=15) were recovered from Middle
Neolithic levels at Stragari. Two mandibles were
unidentifiable to age, leaving seven mandibles and fifteen
Ovis/Capra remains from the Middle Neolithic.
Forty-eight Bos taurus mandibular remains (mandibles n=31 and loose teeth n=17) were recovered from
Middle Neolithic levels at Stragari. Eighteen mandibles
thirteen mandibles and sixteen loose teeth in the final
analysis (n=29 Appendix A). Table 7.43 summarizes the
age stage distributions of Bos taurus remains from the
Middle Neolithic.
Legge (1990) provides some information on quantities of

remains that are relevant to this study. The majority of
Ovis/Capra remains from Selevac are identified as sheep.
There are few goats in the sample. As a result, there was no
attempt to separate the mandibles into species and they are
analysed as Ovis/Capra. There are 110 ageable Ovis/Capra
mandibles, of which 71 were included in Legges analysis.
There are 78 cattle mandibles, partial mandibles and loose
permanent third molars, of which only 41 mandibles were
used in the final analysis. No specific information on
sample size was provided for pigs (Legge 1990).
Nine Sus scrofa mandibles were recovered from the

Middle Neolithic levels at Stragari. Two were
unidentifiable to age, leaving seven mandibles in the
XII. Vina
XII. Stragari-ljivik
A. Site description
A. Site description
The site of Vina-Belo Brdo is a lowland site located on

the right bank of the Danube, 15 km southeast of the city
of Belgrade, in Serbia (Figure 1.1). It is positioned in an
optimal position to monitor and control movement along
the major waterways of the region. It is near the
confluence of the Danube with the Morava, the Tisza, the
Timi and the Sava rivers. On the opposite bank of the
Danube are found the flat plains of Pannonia.
Stragari-ljivik is a lowland site located on a tributary of

the Western Morava at the edge of the village of Stragari,
near the city of Kruevac, in central Serbia (Figure 1.1).
The site is located in the floodplain of a deeply incised
stream, surrounded by low rolling hills. Presently, it is a
rich agricultural area.
Stragari is a Middle Neolithic site, with two levels of
occupation from the Vina A and B cultures. It is a small
site, extending c. 200x200 m, with deposits extending for
about 1 m beneath the plough zone. None of the
sediments from the site were sieved or floated. Over
10,000 fragments were recovered and analyzed from the
1987-1989 excavations at the site (Greenfield n.d. b).
The soils surrounding the site are high quality fertile soils
including river alluvium and chernozems. These were
highly cultivable with the available Neolithic technology
and continue to support modern grasslands and cultivated
crops. Sporadic areas of deciduous forests are also found
within the area (Barker 1975; Chapman 1981).
62

loose teeth from Middle Neolithic Stragari
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
Raw Count
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
1
3
5
1
7
0
0
0
17
Corrected Count
%
0
6
18
29
6
41
0
0
0
100
No.
0
1
3.9
8.3
1.4
7.3
0
0
0
21.9
%
0
5
18
38
6
33
0
0
0
100

loose teeth from Middle Neolithic Stragari
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
Young adult
Adult
old adult
Senile
No.
2
0
6
2
0
0
1
0
1
12
Corrected Count
%
17
0
50
17
0
0
8
0
8
100
No.
2
0
7.5
4.3
0
0
9.2
0
6
29
%
7
0
26
15
0
0
32
0
20
100

loose teeth from Late Neolithic Vina
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
Raw Count
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
1
6
5
1
3
5
3
2
26
Corrected Count
%
0
4
23
19
4
12
19
12
7
100
63
No.
0
1.1
8.0
5.8
1.5
5.6
6.6
3.2
2.1
34
%
0
3
25
17
4
16
20
9
6
100

Vina is a lowland site that extends throughout the entire
Neolithic period, as well as having Eneolithic, Middle
Bronze Age, and Medieval levels. This is the type-site for
the Middle and Late Neolithic Vina culture and is a
yardstick for the Balkan Neolithic (Chapman 1981).
The faunal data described here derive from the Late
Neolithic, Eneolithic and Middle Bronze Age levels.

taurus remains from the Middle Bronze Age.
Forty-four Sus scrofa mandibular remains were recovered
from Late Neolithic deposits at Vina. Eleven mandibles
were undeterminable to age, leaving thirty-four
mandibles in the final analysis (Appendix A). Table 7.48
remains from the Late Neolithic.
The site of Vina is a large artificial mound that extends

for several hundred meters along the riverbank. It is
estimated that over one-third of the site has been eroded
by the river. The site is believed to have also extended
several hundred metres back from the river as well. It is
similar in construction to a tell site with the stratigraphy
extending greater than nine metres (Chapman 1981).
In the Eneolithic layers, there were only two domestic

Sus scrofa mandibles. Only one was determinable with an
age. Age stage distribution is detailed in Appendix B
(Table B35).
Thirty Sus scrofa mandibular remains (mandibles n=29
and loose teeth n=1) were recovered from Middle
Bronze Age deposits at Vina. Ten mandibles were
unidentifiable to age, leaving nineteen mandibles and one
scrofa remains from the Middle Bronze Age.
The fauna reported here derive from the 1982 excavations

at the site. They were hand-collected, but over 10,000
fragments were recovered and analyzed (Greenfield n.d.
a).
XIII. The modern tooth wear and eruption control

sample
Fifty-one domestic Ovis/Capra mandibles were recovered

from Late Neolithic levels at Vina. Seventeen were
unidentifiable to age, leaving thirty-two mandibles in the
Late Neolithic.
The modern control sample included forty-three animals:

twenty-nine goats and fourteen sheep. All animals were
produced locally in Manitoba by small-scale producers.
They are domestic farm animals raised with seasonal
differences both in food availability and seasonal
environment. General age data was able to be estimated
for all animals based on information provided by the
producer and extrapolation backwards from the date of
death. All the goats were born in the month of April, with
a mean birth date estimated by the producer to be April
15th. The goats range in age from 6 months to a year and
a half. As the pattern of slaughter was two animals per
week and the specimens were collected at roughly the
middle of each month throughout the collection period,
date of death was estimated over a period of two months.
Further specific age data was provided by ear tag
information where available. Unfortunately, all animals
did not have ear tags.
The total sample of Ovis/Capra from Eneolithic deposits

was a single mandible. The final sample size is one and
age stage distribution data is summarized in Appendix B
(Table B34).
One domestic Ovis aries mandible and one domestic
Capra hircus mandible from Middle Bronze Age levels
were lumped into the Ovis/Capra heading. As a result,
there was a total of twenty-three Ovis/Capra mandibular
remains (mandibles n=13 and loose teeth n=10) from
the Middle Bronze Age deposits at Vina. All were
included in the final analysis (Appendix A). Table 7.45
remains from the Middle Bronze Age.
Nineteen Bos taurus mandibular remains (mandibles
Neolithic deposits at Vina (eleven mandibles and eight
loose teeth). Five mandibles were undeterminable to age,
leaving six mandibles and eight loose teeth in the final
analysis (n=14 Appendix A). Table 7.46 summarizes the
age stage distributions of Bos taurus remains from the
Late Neolithic. There were no Bos taurus mandibular
remains from the Eneolithic levels.
Tooth eruption and wear was recorded for all mandibles

using the Payne (1973) and Grant (1975) methods. This
data will be utilized to test the agreement between tooth
wear and eruption and the assignment of absolute age.
The modern tooth wear and eruption data are presented in
Appendix C.
Twenty-four Bos taurus mandibular remains (mandibles n=16 and loose teeth n=8) were recovered from Middle
Bronze Age deposits at Vina. Eight mandibles were
unidentifiable to age, leaving eight mandibles and eight
Data for the cementum analysis included both a modern

control sample of sheep and goats and an archaeological
sample. A control sample of the same species of known
age and known season of death was required in order to
accurately determine the timing of the deposition of the
XIV. Cementum analysis data: archaeological and

modern
64

different cementum growth layers in order to accurately
interpret the archaeological teeth. At the same time, the
comparative sample was utilized to test the agreement
between the two most common methods of recording
tooth wear and eruption (Payne 1973; Grant 1975) and
the assignment of absolute age to the wear stages.
Additionally, the modern thin section sample of goats
provided an interesting methodological contribution by
testing whether the incremental growth visible in the
teeth of the goats agrees or did not agree with the
eruption/wear stages and/or with the actual ages. It acted
as a three-way comparison between tooth eruption and
wear, incremental structures and true ages (Ariane Burke,
personal communication, 2001). Details on the modern
and archaeological samples for cementum analysis are
provided in Appendix D.
transhumance is hypothesized to be present. Finally, there

was the spatial issue. The sample needed to cover both
highland and lowland areas. As such, two sites were
selected, Kadica Brdo, an Early Iron Age highland site,
and Vina, a lowland site. All the teeth from Vina were
from Late Neolithic deposits. Ten teeth were sectioned
from Kadica Brdo and nine teeth were sectioned from
Vina. Details of the data from Kadica Brdo and Vina
appear in Table 7.51.
XV. Conclusions
Zooarchaeological remains were available for analysis
from seventeen sites. Ten sites from the central Balkan
region had sufficient data to be utilized in this research.
One site, Megalo Nisi Galanis (in Greek Macedonia) in
the southern Balkan region, is also included for
comparison. The sites range in time from the Early
Neolithic to the Early Iron Age. Only if the mandibular
and loose tooth samples from sites included ten or more
elements per species and per period, were they considered
in the final analysis. Samples with a sample size lower
than ten were considered too small to provide accurate
harvest profiles. These were eliminated from the analysis,
but are described in Appendix B in the event that more
data becomes available.
A. The modern control sample

The modern control cementum (thin sectioning) sample
consisted of fourteen goats and five sheep. All animals
had their left mandibular M1 sectioned using the
methodology discussed in Chapter 6. Where possible,
two slides per tooth were produced. A description of the
cementum observations appears in Table 7.50.
B. Archaeological cementum analysis sample
The central Balkan region can be divided into lowland,

mid-altitude and highland areas. Lowland sites to be
examined include Foeni-Sala, Livade, Novaka uprija,
Opovo, Stragari and Vina. Mid-altitude sites are
Blagotin, Ljuljaci and Petnica. The only highland site is
Kadica Brdo.
The ancient sample for cementum analysis had to meet

several criteria. First, there was the simple issue of
availability. The faunal material had to be available at the
University of Manitoba. Second, there was consideration
of the nature of the sample. As cementum is the least hard
of the dental tissues, it can easily be stripped from the
tooth as the result of post-mortem damage. Only teeth
still encased in the mandible should be selected
(Lieberman and Meadow 1994). Third, there was a
consideration of the temporal issue. The archaeological
sample would ideally cover the periods within which
A large modern sample of sheep and goats from the

province of Manitoba were selected as a control on the
relationship between absolute age and tooth eruption and
wear. Samples of the modern goat and the ancient sheep
and goat remains were selected for cementum analysis.

loose teeth from Middle Bronze Age Vina
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
Raw Count
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
0
3
5
1
1
1
1
1
13
Corrected Count
%
0
0
27
46
9
0
9
9
0
100
65
No.
0
0
3
6.7
2
3.6
3.6
2.9
1
22.8
%
0
0
13
29
9
16
16
13
4
100

Table 7.46. Stage distribution of Bos taurus mandibles
and loose teeth from Late Neolithic Vina
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
young adult
adult
old adult
senile
No.
0
1
1
1
0
0
0
0
2
5
Corrected Count
%
0
20
20
20
0
0
0
0
40
100
No.
0
1.5
1.2
6
0.25
0.25
0.25
0.25
4
13.7
%
0
10
9
44
2
2
2
2
29
100

Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
young adult
adult
old adult
senile
No.
1
1
4
1
0
0
0
1
1
9
Corrected Count
%
11
11
45
11
0
0
0
11
11
100
No.
1
1
4
1
0
0
0
5.5
3.5
16
%
6
6
25
6
0
0
0
34
23
100

loose teeth from Late Neolithic Vina
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
Adult
old adult
Senile
No.
0
5
8
9
1
0
0
0
0
23
Corrected Count
%
0
23
34
39
4
0
0
0
0
100
66
No.
0.0
6.1
11.3
13.3
2.3
0.0
0.0
0.0
0.0
33.0
%
0
18
33
42
7
0
0
0
0
100

Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
Adult
old adult
Senile
No.
1
1
5
6
2
1
0
0
0
16
Corrected Count
%
6
6
31
38
13
6
0
0
0
100
67
No.
1.0
1.0
5.5
8.8
2.8
1.0
0.0
0.0
0.0
20.0
%
5
5
27
44
14
5
0
0
0
100
68
Goat #11
Goat #11
Goat #12
Goat #12
Goat #13
Goat #13
Goat #14
Goat #14
Goat #15
Goat #8
Goat #10
Goat #7
Goat #7
Goat #10
Goat #6
1 year, 5-6 months
Goat #6
Goat #8
1 year, 4-5 months
Goat #4
9-10 months
9-10 months
9-10 months
8 months
8 months
8 months
8 months
6-7 months
6-7 months
6-7 months
6-7 months
1 year, 5-6 months
1 year, 5-6 months
1 year, 5-6 months
1 year, 5-6 months
1 year, 5-6 months
1 year, 4-5 months
Goat #2
Age
1 year, 4-5 months
Slide
Goat #1
Sample
Jan/Feb 2001
Jan/Feb 2001
Jan/Feb 2001
December 2000
December 2000
December 2000
December 2000
Oct/Nov 2000
Oct/Nov 2000
Oct/Nov 2000
Oct/Nov 2000
Sept/Oct 2000
Sept/Oct 2000
Sept/Oct 2000
Sept/Oct 2000
Sept/Oct 2000
Sept/Oct 2000
Aug/Sept 2000
Aug/Sept 2000
Aug/Sept 2000
Date of Death
growth zone
indeterminable
growth zone
growth zone
growth zone
undeterminable
undeterminable
undeterminable
undeterminable
undeterminable
no annulus/zone
final
growth zone?
Nature of final
increment
0 GLG
0 GLG
reading taken at end of root
bright outer increment = annulus

forming?
unreadable
unreadable
root area too thick for accurate reading
slide too thick
Comments
growth
zone/annulus final
growth zone
tumor tooth
2 annuli are
undeterminable
formed low on the
root
0 GLG
undeterminable
undeterminable
no GLG
indeterminable
no GLG
no GLG
no GLG
undeterminable
1GLG
1GLG
undeterminable
1GLG
1GLG
1GLG
undeterminable
1GLG (partial)
indeterminable
# of increments
(GLGs)
Table 7.50. Modern Ovis/Capra comparative cementum analysis Summary of readings
Vinca
Vinca
Vinca
Vinca
Vinca
Vinca
Vinca
Vinca
V #3
V #4
V #5
V #6
V #7
V #8
V #9
V #10
Goat #17
Kadica Brdo
Kadica Brdo
Kadica Brdo
Kadica Brdo
Kadica Brdo
Kadica Brdo
Kadica Brdo
Kadica Brdo
Kadica Brdo
Kadica Brdo
Vinca
Vinca
Goat #17
KB #1
KB #2
KB #3
KB #4
KB #5
KB #6
KB #7
KB #8
KB #9
KB #10
V #1
V #2
Goat #16
Site
Goat #16
Sectioning
Code
Slide
Sample
Jan/Feb 2001
Jan/Feb 2001
Jan/Feb 2001
Jan/Feb 2001
Date of Death
undeterminable
undeterminable
0 GLG
0 GLG
# of increments
(GLGs)
growth zone
growth zone
Nature of final
increment
unreadable
unreadable
69
Lowland
Lowland
Lowland
Lowland
Lowland
Lowland
Lowland
Lowland
Highland
Highland
Highland
Highland
Highland
Highland
Highland
Highland
Highland
Highland
Lowland
Lowland
Location
5
6
6
5
5
5
5
5
4
24
23
12
1
16
24
3
25
13
6
6
Level
Ovis aries
Ovis aries
Capra hircus
Ovis aries
Ovis aries
Capra hircus
Ovis aries
Ovis/Capra
Ovis aries
Ovis/Capra
Ovis aries
Ovis aries
Ovis aries
Ovis aries
Ovis aries
Capra hircus
Possible goat
Ovis aries
Capra hircus
Ovis aries
Taxon
8-10 years
4-6 years
6 months - 4 years
4-6 years
2-3 years
6-8 years
1-2 years
3-10 years
3-4 years
2-3 years
1-2 years
4-6 years
6-8 years
4-6 years
2-3 years
6-12 months
3-4 years
6-12 months
6-8 years
4-6 years
unreadable
2 GLG
1 GLG
6 GLG
unreadable
5 GLG
unreadable
1 GLG
unreadable
0 GLG
9 GLG
Tooth completely
broken during
extraction
unreadable
unreadable
unreadable
4 GLG
unreadable
unreadable
1 GLG
8 GLG
Absolute Age (from tooth # of increments

wear and eruption)
counted
indeterminate
possible annulus final
growth zone
indeterminate
growth zone
growth zone
indeterminate
growth zone
Nature of Final
Increment
right side bright outer increment =

annulus forming?
number of secondary GLGs
Comments
Table 7.51. Archaeological Ovis/Capra cementum analysis Summary of readings
9-10 months
9-10 months
9-10 months
9-10 months
Age
CHAPTER 8
THE IDENTIFICATION OF TRANSHUMANT PASTORALISM
THROUGH HARVEST PROFILE AND CEMENTUM ANALYSES
I. Introduction
taphonomic issues of differential preservation and

excavation practices, specifically the extent of sieving at
sites (Table 8.52). Differential preservation is necessary
to consider because the remains of the very young age
classes are very fragile. It can be assumed that a very
high proportion of those originally present in
archaeological contexts would be lost through pre- and
post depositional processes and excavation procedures
(Munson 2000: 391; Cribb 1985). The presence/absence
of this age group is too affected by the taphonomic issues
mentioned above to be considered reliable in making
conclusions on the existence of transhumance. Munson
(2000) goes as far as to exclude the consideration of
neonates and notes only presence/absence. As such, any
indications of transhumance must be established by
looking at the other age groups involved in the movement
of herds, specifically the 2-6 months and the 6-12 month
group in sheep/goat and the 1-8 month and 8-18 month
group in cattle. The expected patterns for the lowland,
mid-altitude and high altitude sites for each species are
summarized in Table 8.53. Excavation practices have a
similar attritional effect. Hand collection (as opposed to
sieving) favours the gathering of the remains of larger
(older) animals over small (younger) animals. These
issues will be considered during the analyses below.
This chapter will undertake several types of analyses. The

first section deals with the analysis of the tooth wear and
eruption data discussed in Chapter 7. This first stage of
the analysis will present the interpretation of the data by
taxon within each period and site. Harvest profiles of
each taxon/site/period are constructed and analyzed
utilizing the methods discussed in Chapter 6. The
analysis will consider the harvest patterns exhibited by
each species over the time period covered in this
investigation, that is, from the Early Neolithic to the
Early Iron Age. Interpretations on animal exploitation
strategies and the possibility of transhumance are
examined for each taxon/site/period individually. The
second stage of analysis in this chapter will synthesize the
harvest profile data on a regional scale to test for the
appearance of transhumance at a moment in time. The
focus will be on changes over time. The third stage of
analysis will evaluate the tooth wear and eruption
recording techniques against the modern comparative
sample to determine their validity. Finally, the results of
the cementum (thin-sectioning) analysis are discussed.
The analysis was limited by the paucity of high altitude
sites gathered for this investigation, of which there is only
one (Kadica Brdo). The reason is that the high altitude
environments are not colonized until the Eneolithic
(beginning of the Post Neolithic). Hence, it is difficult to
find directly comparable sites from both the Neolithic and
Post Neolithic. The mid-altitude sites are hypothesized to
be a relevant substitute in that they should show harvest
profiles most similar to the high altitude sites in the
transhumant movement of herds. This is because the
over-wintering of herds in these mid-altitude regions
would be difficult without sufficient shelter for the herds
and collected fodder. As such, the environment of the
mid-altitude sites in the region, as would that of the high
altitude sites, would force the movement of domestic
herds out of this altitude during the colder months of the
year (i.e. transhumance). This is in contrast to the
lowland sites, where the variety of microenvironments in
these areas can provide sufficient graze and water for
herds year round. As discussed earlier, there is no
environmental factor that forces the transhumant
movement of lowland herds.
II. Construction of harvest profiles first stage of

analysis
A. Sus scrofa dom.
Domestic pig can be considered as a control for any
analysis of the transhumant movement of herds
(Greenfield 1988, 1999a). It is expected that the
exploitation patterns of pig should not change over time
as these animals were unlikely to be subjected to the
transhumant movement characteristic of either cattle or
sheep/goat that is hypothesized to occur. This is due to
the fact that pigs are less suited to the types of movement
that is required of transhumant herds. Some researchers
have noted that pigs are capable of long distance
movements, such as the driving of herds from Serbia to
Vienna in the 19th century AD, (Halpern 1999). This
tends to be a one time movement under highly developed
market conditions and should not be equated with the
regular movements expected in transhumance under the
pre-market conditions of prehistory.
The absence of the youngest age classes for all the

domestic species is a problem throughout this
investigation. The hypotheses are stated in such a way as
to place significant emphasis on the presence or absence
of this age class. However, one must consider the
Given previous analyses, it is expected that domestic pigs

were exploited predominantly for their primary products,
as there are no secondary products that are readily
available from pigs (Greenfield 1988, 1989). As a result,
71
72
Vina
Stragari
Petnica
Novaka
uprija
Opovo
Ljuljaci
Megalo Nisi
Galanis
Livade
Kadica Brdo
Foeni-Sala
Blagotin
Site
10
20
50
100
100
85
10
0-2 month
11-25 26-50 51-75 76-100
1-10
Age class
(Ovis/Capra)
%
Sieved
2-6
month
11
23
22
25
32
39
6-12
month
0-1 month
Age class
(Bos taurus)
Table 8.52. Summary of sieving and

presence of youngest age
1-8
month
14
8-18
month
0-2 month
Age class
(Sus scrofa)
11
11
2-7
month
17
12
10
7-14
month
Table 8.53. Expectations for movement in transhumant pattern

Ovis/Capra
Lowland
Mid-altitude
Highland
Bos taurus
Lowland
Mid-altitude
Highland
Age class
0-2
Present
Absent
Absent
2-6 month
Absent
Present
Present
6-12 month
Present
Absent
Absent
Age Class
0-1
Present
Absent
Absent
1-8 months
Absent
Present
Present
8-18 months
Present
Absent
Absent
Figure 8.4. Paynes meat production mortality profile (from Payne 1973).
73

the harvest profiles constructed for pigs are expected to
show patterns similar to Paynes (1973) meat production
kill off pattern (Figure 8.4).
may be due to taphonomic issues discussed above. The

presence of the earliest age groups (2-7 through to 7-14
months) indicates that herds were present in some
proximity to the site during all seasons.
As predicted above, the harvest profiles for all sites and

periods demonstrate a pattern most similar to Paynes
meat production.
The harvest profile from the site of Selevac (Legge 1990)

was included for consideration and shows a significant
difference from other Late Neolithic sites in the central
Balkans (Figure 8.9). This difference is the result of the
fact that the report (Legge 1990) combined both wild and
domestic forms of pigs when the harvest profiles were
constructed. As only domestic forms were considered from
the other sites, the data from Selevac are not comparable
and cannot be considered valid for this analysis.
Additionally, and most importantly for this investigation,

the harvest patterns do not change over time. All the
profiles can be characterized as showing a high mortality
of animals between the ages of 2-7 months and 27-36
months. Virtually all animals are slaughtered before
reaching adulthood. In the following sections, the
remains from each period are separately considered.
4. Eneolithic
Only one site contained sufficient Eneolithic material to
construct a harvest profile Petnica (Figure 8.10). The
profile indicates the absence of the youngest age class (02 months). There is high mortality of animals between the
ages of 2-7 months, and again between 14-21 months.
The intervening age group is missing. Subsequently,
there is a low rate of mortality until 27-36 months. Then,
there is no mortality until the senile age group, which is a
single senile individual. The Eneolithic profile indicates
continuity of exploitation from the previous Late
Neolithic. While the youngest age group (0-2 months) is
absent, this may be due to the taphonomic issues
discussed above. The presence of the earliest age groups
(2-7 through to 7-14 months) indicates that herds were
present in some proximity to the site during all seasons.
1. Early Neolithic
None of the Early Neolithic assemblages had a large
enough sample of pig remains to construct a harvest
profile. This is due to the dearth of pig remains in such
assemblages. This is typical of most Early Neolithic sites
in the region (e.g. Bknyi 1974a; Greenfield 1993, in
press a).
2. Middle Neolithic
In the Middle Neolithic, only the deposits from Petnica
had a sufficient sample size to reconstruct a harvest
profile. The harvest profile (Figure 8.5) shows a rapid
mortality of the youngest age classes (0-21 months). The
adult end of the profile shows the absence of subadults
and adults, except for the presence of two senile
individuals. The profile basically conforms to the
expected exploitation pattern for meat. The presence of
the earliest age groups (0-2 months through to 7-14
5. Early/Middle Bronze Age

The Early and Middle Bronze Age material will be
discussed together since one of the sites (Ljuljaci)
overlaps both periods. There are three sites with Early
and/or Middle Bronze Age materials (Vina, Ljuljaci and
Novaka uprija). The profile from Novaka uprija
(Figure 8.11) indicates a very high mortality of animals
between the ages of 2-7 and 21-27 months. There is a
total dearth of remains from the 0-2 and >27 month
classes.
3. Late Neolithic
Three Late Neolithic assemblages had sufficient pig
remains to construct a harvest profile (Petnica, Vina and
Opovo). Each of the sites shows essentially the same
pattern, regardless of whether the sites are in lowland or
mid-altitude locations. Chi square analysis indicates that
the harvest profiles for each of the three sites show no
statistically significant difference (Appendix E, Table
E1). All are missing the youngest age class (0-2 months).
There is a very high mortality of animals between the
ages of 2-7 and 27-36 months. The Late Neolithic Petnica
deposits (Figure 8.6) continue to show a presence of
small quantities of senile individuals (n=1). The sites of
Vina (Figure 8.7) and Opovo (Figure 8.8) show an
absence of senile individuals. In these two sites, all
animals are slaughtered before the age of 21-27 months.
In the mid-altitude site of Ljuljaci (Figure 8.12), the

harvest pattern remains the essentially the same. The only
difference is that a few animals are being slaughtered at a
slightly older age (up to 36 as opposed to 27 months at
Novaka uprija).
The same general pattern exists in the Middle Bronze
Age lowland site of Vina (Figure 8.13). Just as in
Ljuljaci and Novaka uprija, there is a very high
mortality in the 2-7 to the 21-27 month age classes.
Vina, however, is more similar to Ljuljaci in that the age
classes extend to the 27-36 month range. It differs from
both of them by the presence of the youngest age classes
(0-2 months).
The Late Neolithic pattern is a continuation of the Middle

Neolithic pattern. The profiles indicate exploitation of
pigs for meat. The presence of very young individuals in
the sample confirms the hypothesis that no domestic
animals are moving between highland and lowland sites.
While the youngest age group (0-2 months) is absent, this
Statistical analysis indicates that the harvest profiles for

each of the three sites show no statistically significant
74
100
% Age Survival
80
60
40
20
0
0 months
0-2 months
A
2-7 months
B
7-14 months 14-21 months 21-27 months 27-36 months

C
D
E
F
adult
G
old adult
H
senile
I
Mandibular Stage and Absolute Age
Figure 8.5. Harvest Profile (Sus scrofa) Middle Neolithic Petnica
100
% Age Survival
80
60
40
20
0
0 months
0-2 months
A
2-7 months
B

C
D
E
F
adult
G
Figure 8.6. Harvest Profile (Sus scrofa) Late Neolithic Petnica
75
old adult
H
senile
I
100
60
40
20
0
0 months
0-2 months
A
2-7 months
B

C
D
E
F
adult
G
old adult
H
senile
I
Figure 8.7. Harvest Profile (Sus scrofa) Late Neolithic Vina
100
80
% Age Survival
% Age Survival
80
60
40
20
0
0 months
0-2 months
A
2-7 months 7-14 months

B
C
14-21
months
D
21-27
months
E
27-36
months
F
adult
G
Figure 8.8. Harvest profile (Sus scrofa) Late Neolithic Opovo
76
old adult
H
senile
I
100
% of age survival
80
60
40
20
senile
I
old adult
H
adult
G
27-36 months
F
21-27 months
E
14-21 months
D
7-14 months
C
2-7 months
B
0-2 months
A

Selevac (Middle/Late Neolithic)
Petnica (Middle Neolithic)
Opovo (Late Neolithic)
Vina (Late Neolithic)
Figure 8.9. Harvest profile (Sus scrofa) Middle/Late Neolithic Selevac
100
% Age Survival
80
60
40
20
0
0 months
0-2 months
A
2-7 months
B

C
D
E
F
adult
G
Figure 8.10. Harvest Profile (Sus scrofa) Eneolithic Petnica
77
old adult
H
senile
I
100
% Age Survival
80
60
40
20
0
0 months
0-2 months
A
2-7 months
B

C
D
E
F
adult
G
old adult
H
senile
I
Figure 8.11. Harvest Profile (Sus scrofa) Early Bronze Age Novaka uprija
100
% Age Survival
80
60
40
20
0
0 months
0-2 months
A

B
C
14-21
months
D
21-27
months
E
27-36
months
F
adult
G
old adult
H
Figure 8.12. Harvest profile (Sus scrofa) Early/Middle Bronze Age Ljuljaci
78
senile
I
100
% Age Survival
80
60
40
20
0
0 months
0-2 months
A

B
C
14-21
months
D
21-27
months
E
27-36
months
F
adult
G
old adult
H
senile
I
Figure 8.13. Harvest profile (Sus scrofa) Middle Bronze Age Vina
100
% Age Survival
80
60
40
20
0
0 months
0-2 months
A

B
C
14-21
months
D
21-27
months
E
27-36
months
F
adult
G
Figure 8.14. Harvest profile (Sus scrofa) Late Bronze Age Livade
79
old adult
H
senile
I
100
% Age Survival
80
60
40
20
0
0 months
0-2 months
A
2-7 months
B

C
D
E
F
adult
G
old adult
H
senile
I
Figure 8.15. Harvest profile (Sus scrofa) Early Iron Age Kadica Brdo
difference (Appendix E, Table E2). This result indicates
continuity of exploitation in pigs over time. The presence
of young individuals in lowland and mid-altitude sites
also supports the hypothesis that pigs are not moving in a
transhumant fashion. While the youngest age group (0-2
months) is absent, this may be due to taphonomic issues
discussed above. The presence of the earliest age groups
(2-7 through to 7-14 months) indicates that herds were
present in some proximity to the site during all seasons.

8. Implications of pig exploitation patterns
The exploitation pattern of pigs does not change
significantly over the entire time period considered in this
investigation. Statistical analysis of sites by major periods
indicates no statistically significant changes (Appendix E,
Table E3). The exploitation of pigs for their primary
products continues to be the dominant feature of all the
harvest profiles.
6. Late Bronze Age

During the Late Bronze Age, there is only a single sample
with sufficient pig remains, the lowland site of Livade
(Figure 8.14). This profile follows the trend of the
preceding periods. The absence of the 0-2 month class in
a lowland site is likely a result of differential destruction
of younger age classes at the site (Cribb 1985; Greenfield
1986a; Munson 2000). The presence of the earliest age
groups (2-7 through to 7-14 months) indicates that herds
were present in some proximity to the site during all
seasons.
Most importantly for the investigation into the origins of

transhumant pastoralism, there is no difference in the
exploitation patterns of domestic pig between highland
and lowland sites either before and after the Late
Neolithic/Post Neolithic juncture or at any other time.
B. Ovis/Capra
Separate harvest profiles for Ovis aries and Capra hircus
were not possible to produce. Only three sites had
sufficient Ovis aries remains to produce separate harvest
profiles and no sites had sufficient remains of Capra
hircus. Therefore, the remains of both taxa were combined
in order to achieve a number of sites with sufficient
remains for analysis. The drawback of this approach is that
it is difficult to separate out the pattern of sheep versus
goat. This is somewhat negated by examining the
frequency distribution of sheep and goat in each site. In
every case, sheep remains far outnumber those of goats
(often at a 10:1 ratio). Goats tend to be an insignificant part
of total identified assemblages. In addition, most
previously identified goat remains were from adult
7. Early Iron Age

The Early Iron Age highland site of Kadica Brdo (Figure
8.15) yields a pattern broadly similar to those of the other
sites. The highest mortality rate is between 2-7 months
and 21-27 months. In addition, there are both very old
and very young individuals at the site. The presence of
the 0-2 month age class would indicate that animals are
born at or near the site. This would indicate the lack of
seasonal transhumant movement of pigs at the site. The
presence of the earliest age groups (0-2 through to 7-14
80

individuals (Greenfield 1986a, 1986b, 1988, 1991, 1994,
1996). Therefore, it is proposed that the perceived pattern
probably represents the remains of sheep, with a slight
influence by goats on the older age classes.
production model and from the Early Neolithic assemblages

is in the absence of the oldest age groups (4-6 to 8-10 years).
Otherwise, this Middle Neolithic site follows the same
general pattern established in the Early Neolithic.
1. Early Neolithic
During the Early Neolithic period, both the lowland site
of Foeni-Sala and the mid-altitude site of Blagotin have
sufficient sample sizes to produce harvest profiles for
Ovis/Capra. At Foeni-Sala (Figure 8.16), there is some,
but very little mortality of the youngest age classes (0-2
to 2-6 months). This is followed by very high mortality of
the age groups 2-6 months to 1-2 years. Then there is a
plateau in the mortality rate between 1-2 and 2-3 year age
classes. This is followed by a rapid mortality of the age
groups from 1-2 years to 4-6 years, but at a decreased rate
than during the previous rapid mortality phase. This
pattern is essentially repeated at Blagotin (Figure 8.17),
with two major differences: the absence of the plateau of
the mortality rate between 1-2 and 2-3 year age classes
and the oldest age group (8-10 years). The general Early
Neolithic mortality pattern for Ovis/Capra is shown in
Figure 18. Statistical analysis indicates that there are no
statistically significant differences between the profiles of
Foeni-Sala and Blagotin (Appendix E, Table E4).
There is a lack of evidence for transhumance during the

Middle Neolithic. Stragari is a lowland site. The 0-2
month age class would be expected to be present. The
absence of the 0-2 month age class would normally imply
that the animals are not being born in and around the site.
The absence of this age class may be because this is an
unsieved sample, but such remains are present in other
unsieved or partially sieved samples. The presence of the
2-6 and 6-12 month age classes would indicate that the
herds were around the site for the majority of the year.
The continuity of occupation and exploitation from these
age classes would imply that there is no evidence for
transhumance during this period.
3. Late Neolithic
Two Late Neolithic assemblages have sufficient
Ovis/Capra remains to construct a harvest profile (midaltitude Petnica and lowland Vina). The lowland site of
Selevac (Legge 1990) is also included in the analysis. The
harvest profile from Vina (Figure 8.20) indicates an
absence of the youngest age group (0-2 months) followed
by a very low mortality of the 2-6 month age groups.
Selevac (Figure 8.21) shows a higher rate of mortality in
the 2-6 month group (than at Vina) followed by a
relatively rapid mortality through the 6-12 months and 1-2
years age groups. There is a rapid, but continued high
mortality rate, in comparison to the earlier groups, of the
older age groups (2-3 through 8-10 years). At the midaltitude site of Petnica (Figure 8.22), a slightly different
pattern emerges. Both the 0-2 months and the 2-6 months
are absent (whereas they were present in all earlier and
contemporary assemblages). Following this, there is a
rather steady rate of mortality from the 6-12 month group
through to the 6-8 year stage. The oldest individuals are
also absent. The harvest profiles for Ovis/Capra (Figure
8.23) from the Late Neolithic sites show no statistically
significant differences between the profiles (Appendix E,
Table E5).
The harvest profiles from both sites indicate the

exploitation of the herd for primary products, that is, meat
production. It most closely resembles Paynes Model A
kill-off pattern (see Figure 8.4).
The Early Neolithic sites do not show any evidence for a
transhumant pattern. The presence of the 0-2, 2-6, and 612 month age classes in both sites implies a year round
availability of the herds. This suggests a non-transhumant
movement and continuous culling of the animals and
residential stability throughout the year. There is no
complementarity between the sites and transhumance is
not occurring between mid-altitude Blagotin and lowland
Foeni-Sala.
2. Middle Neolithic
There is only one site that provides information on the
Middle Neolithic harvest profile patterns of Ovis/Capra the lowland site of Stragari (Figure 8.19). The harvest
profile of Stragari shows the same pattern as those from
Early Neolithic. The 0-2 month class is missing, and there
is very low mortality of the youngest age groups (2-6
months), followed by a rapid mortality from 6-12 months
to 1-2 years. This is followed by a rapid mortality, but at a
decreased rate in the older age groups (1-4 years). What is
unusual about this profile in relation to the Early Neolithic
is the complete lack of the oldest age groups (3-4 through
to 8-10 years). However, these age stages are not sensitive
to the hypotheses being tested.
The harvest profiles of all the Late Neolithic sites most

closely resemble the herd exploitation model for primary
products, that is, meat production, during this period
(Paynes 1973 Model A kill-off patternsee Figure 8.4).
None of the harvest profiles appear to fully fit the
expected pattern for Neolithic ovicaprine exploitation
with its absence of the earliest age groups. These sites
stand in contrast to the profiles created for the other sites
in the sample for the Neolithic period. It cannot be argued
that the absence of these youngest age groups was the
result of differential recovery practices as this absence is
most pronounced at Petnica, where approximately 20% of
the sample was sieved. Also, there is no evidence for
additional taphonomic issues that might plague Petnica in
contrast to other sites from this and later time periods
(Greenfield 1986a, 1991).
The harvest profile most closely resembles the herd

exploitation model for primary products, that is, meat
production, during this period (Paynes 1973 Model A killoff patternsee Figure 8.4). Where it differs from the
81
100
60
40
20
0
0 months
0-2 months
A

B
C
1-2 years
D
2-3 years
E
3-4 years
F
4-6 years
G
6-8 years
H
8-10 years
I
Mandibular Stage and Absolute Ages
Figure 8.16. Harvest Profile (Ovis/Capra) Early Neolithic Foeni-Sala
100
80
% Age Survival
% Age Survival
80
60
40
20
0
0 months
0-2 months
A

B
C
1-2 years
D
2-3 years
E
3-4 years
F
4-6 years
G
6-8 years
H
Mandibular Stages and Ages
Figure 8.17. Harvest Profile (Ovis/Capra) Early Neolithic Blagotin
82
8-10 years
I
100
60
40
20
0
Suggested
Age
Stage
0 months
0-2 months
A
2-6 months
B
6-12 months
C
1-2 years
D
2-3 years
E
3-4 years
F
4-6 years
G
6-8 years
H
6-8 years
H
8-10 years
I
Figure 8.18. Harvest Profile (Ovis/Capra) Early Neolithic
100
80
% Age Survival
% Age Survival
80
60
40
20
0
0 months
0-2 months
A
2-6 months
B
6-12 months
C
1-2 years
D
2-3 years
E
3-4 years
F
4-6 years
G
Figure 8.19. Harvest Profile (Ovis/Capra) Middle Neolithic Stragari
83
100
% Age Survival
80
60
40
20
0
0 months
0-2 months
A

B
C
1-2 years
D
2-3 years
E
3-4 years
F
4-6 years
G
6-8 years
H
8-10 years
I
Figure 8.20. Harvest Profile (Ovis/Capra) Late Neolithic Vina
100
% of age survival
80
60
40
20
0
0
0-1 months
A

B
C
18-30
months
D
30-36
months
E
young adult
F
adult
G
old adult
H

Selevac
Opovo
Petnica
Vina
Figure 8.21. Harvest Profile (Bos taurus) Middle/Late Neolithic Selevac

84
senile
I
100
% Age Survival
80
60
40
20
0
0 months
0-2 months
A
2-6 months
B
6-12 months
C
1-2 years
D
2-3 years
E
3-4 years
F
4-6 years
G
6-8 years
H
8-10 years
I
Figure 8.22. Harvest profile (Ovis/Capra) Late Neolithic Petnica
Vinca
Petnica
100
% Age Survival
80
60
40
20
0
0 months
0-2 months
A
2-6 months
B
6-12 months
C
1-2 years
D
2-3 years
E
3-4 years
F
4-6 years
G
Figure 8.23. Harvest profile (Ovis/Capra) Late Neolithic
85
6-8 years
H
8-10 years
I
8.24 Paynes (1973) Milk Production
86
8.25 Paynes (1973) Wool Production
87
100
% Age Survival
80
60
40
20
0
0 months
0-2 months
A
2-6 months
B
6-12 months
C
1-2 years
D
2-3 years
E
3-4 years
F
4-6 years
G
6-8 years
H
8-10 years
I
Figure 8.26. Harvest profile (Ovis/Capra) Eneolithic Novaka uprija

While the absence of the youngest age groups (especially
in the lowland site) may be seen as an indication of
transhumant movement in the Late Neolithic, this is not
believed to be the case here. If transhumance was present,
the second youngest age group (2-6 months) would be
expected to be present in mid-altitude sites. The absence
of both the 0-2 and 2-6 month class in the mid-altitude
sites (e.g. Petnica) would argue that these age classes
were not being exploited. The absence of the youngest
age groups from Late Neolithic Petnica, a mid-altitude
site, would argue against the appearance of transhumance
during the Late Neolithic.
value is given as 0.0110, the exact p value is 0.048. If one

were to consider only the usual p value, one might
assume that the results are extremely significant.
However, as it is the exact p value that corrects for small
sample sizes (as is the case here), its values indicate that
the difference is not as strong as it is implied by the other
value. If transhumance was occurring, these differences
would be expected between mid altitude Petnica and
lowland Novaka uprija at this time. However, the
small sample size mitigates our ability to confidently
identify its presence.
Neither of these patterns fit any of Paynes proposed
specialized herd production models (Figures 8.4, 8.25
and 8.26). Both, however, conform to Greenfields
(1988) expectations for a herd based on long-term
stability and animal exploitation with a mixed subsistence
economy (both primary and secondary products) and an
emphasis on secondary products.
4. Eneolithic
There are two sites from the Eneolithic with sufficient
samples to construct harvest profiles - lowland Novaka
uprija and mid-altitude Petnica. At Novaka uprija
(Figure 8.26), there is a complete absence of the early
age groups (0-12 months). There is low mortality of the
1-2 year group, none in the 2-3 year group and extremely
high mortality of 3-4 year animals. Then, the mortality
rate plateaus until 8-10 years. At Petnica (Figure 8.27),
there is an absence of the youngest age groups (0-2 and 26 months) followed by a very high mortality of the 6-12
month age group. Subsequently, there is a decreased rate
of mortality at 1-2 years, which slightly levels off at 2-3
years. It continues at this rate until the 4-6 year class.
There is a complete absence of the two oldest age groups
(6-8 and 8-10 years).
During the Eneolithic, clear indications for transhumance

are absent. If transhumance was to appear during the
Eneolithic, the youngest age group, then the 0-2 month
class should be present at lowland sites. Can their
absence be the result of differential attrition? Sieving is
an unlikely cause since Novaka uprija was completely
sieved. However, the assemblage was not deeply buried
and was subject to higher rates of weathering than many
of the other contemporary assemblages (Table 8.54). The
absence of the expected 0-2 month class at Novaka
uprija cannot be used to monitor the presence or
absence of the transhumant movement of herds.
Chi square analysis indicates that there is a marginally

statistically significant difference between these two sites
in this period (Appendix A, Table E6). While the usual p
88
100
% Age Survival
80
60
40
20
0
0 months
0-2 months
A
2-6 months
B
6-12 months
C
1-2 years
D
2-3 years
E
3-4 years
F
4-6 years
G
6-8 years
H
8-10 years
I
6-8 years
H
8-10 years
I
Figure 8.27. Harvest profile (Ovis/Capra) Eneolithic Petnica
100
% Age Survival
80
60
40
20
0
0 months
0-2 months
A
2-6 months
B
6-12 months
C
1-2 years
D
2-3 years
E
3-4 years
F
4-6 years
G
Figure 8.28. Harvest profile (Ovis/Capra) Early Bronze Age Novaka uprija
89

As a mid-altitude site, Petnica is expected to show
mortality profiles similar to that of a highland site. This
would require the absence of the youngest age class (0-2
months) and the presence of the second youngest age
class (2-6 months), which would indicate a seasonal
presence in the highlands between May and August.
However, both of these age classes are missing from the
profile. As a result, it can be concluded that there is no
evidence for transhumance, since Eneolithic Petnica does
not conform to the expectation of a highland site involved
in the transhumant movement of herds.
of assemblage attrition, given the lack of sieving at

Vina (Table 8.49). The slightly older age classes
should be less affected (cf. Munson 2000). The absence
of the 2-6 month class and presence of the 6-12 month
class at Vina are in accordance with the expectations
for a lowland site involved in the transhumant
movement of herds. The problem is that the presence of
the 2-6 month classes at Novaka uprija is not
accordance with the expectations for a lowland site
involved in the transhumant movement of herds. One
site may be interpreted to be part of a transhumant
system and the other may not. A major problem with
this period is that there is an absence of mid- and high
altitude site locations with sufficient data with which to
compare them.
The mid-altitude site of Petnica is missing the expected 2-6

month class for a transhumant pattern and the lowland site
of Novaka uprija is missing its 6-12 month age group.
6. Late Bronze Age

The situation is different for the Late Bronze Age. Three
sites yielded sufficient data for this periodmid-altitude
Petnica and lowland Novaka uprija and Livade.
Novaka uprija (Figure 8.30) shows an absence of very
young age groups (0-6 months). There is a medium
mortality rate of age classes 6-12 months and 1-2 years
followed by a higher mortality rate through the 2-6 years
age classes. After this, the rate declines (6-8 years). There
are no remains from oldest age group (8-10 years)
sample. At Livade (Figure 8.31), there is complete
absence of age groups 0-12 months and a steep decline in
the profile indicating a rapid mortality of subadult/adults
(2-3 through to 6-8 years). There is an absence of the
oldest age group (8-10 years). Petnica (Figure 8.32)
shows a different pattern. Even though the youngest age
groups (0-2 months) are absent, the next age group (2-6
months) is present at the site. There is a pattern of
gradual mortality from 2-6 months through to 2-3 years
followed by a more rapid mortality of the age groups 34 years and 4-6 years. There is an absence of the oldest
age groups (6-8 and 8-10 years).
In addition, the 6-12 month class is present in Petnica. The

animals in this age class would be expected to have
returned to their lowland pastures for the autumn and
winter months. As a result, it is less likely that transhumant
pastoralism is taking place at this time.
As with pigs, the Early and Middle Bronze Ages are
considered together. There is a great deal of cultural
continuity between these periods. Only two sites yielded
samples, both of which are from the lowlands - Early
Bronze Age Novaka uprija and Middle Bronze Age
Vina. The Novaka uprija (Figure 8.28) mortality
profile shows the absence of the youngest age group (0-2
months) followed by little mortality of the 2-6 and 6-12
months age groups. This is followed by rapid mortality of
age groups 1-2 years and finally a rapid mortality, but at a
decreased rate, of the older age groups (2-10 years) in
comparison to the earlier groups. The harvest profile from
Vina (Figure 8.29) shows an absence of the youngest
age groups (0-2 and 2-6 months). There is continued and
rapid mortality of the age groups beginning with 6-12
months and continuing through to 8-10 years, with slight
changes along the way. The profiles of Early Bronze Age
Novaka uprija and Middle Bronze Age Vina are very
similar. There are no statistically significant differences
between the profiles of these sites (Appendix E, Table
E7).
It would be expected that the lowland sites of Novaka

uprija and Livade would be similar to each other and
significantly different from the mid-altitude site of
Petnica (Figure 8.33). However, chi square analysis of
the profiles from all three sites from this period
indicates that there is no statistically significant
difference between them (Appendix E, Table E8). It is
probable that the sample sizes are not sufficiently large
to reflect the expected differences using statistics.
Neither of these patterns fit any of Paynes proposed

specialized herd production models (Figures 8.4, 8.24 and
8.25). Both, however, conform to Greenfields (1988)
expectations for a herd based on long-term stability and
animal exploitation with a mixed subsistence economy
(both primary and secondary products) and an emphasis
on secondary products.
None of the sites fit any of Paynes proposed specialized

herd production models (Figures 8.4, 8.24 and 8.25).
Both conform to Greenfields (1988) expectations for a
herd based on long-term stability and animal exploitation
with a mixed subsistence economy (both primary and
secondary products) and an emphasis on secondary
products.
The evidence for transhumance from this period is

somewhat mixed. Since both samples come from
lowland sites, they should have evidence of the 0-2 and
6-12 month, and absence of 2-6 month age classes if
transhumance is occurring. The 0-2 age class is missing
in both cases. The absence of the expected 0-2 month
class is not surprising in these sites. It may be a function
The evidence for transhumance in this period is

ambiguous. At Petnica, the evidence is mixed. On the one
hand, the presence of the 2-6 month class and the absence
90

Table 7.54. Summary of strata and weathering
Site
Superimposed strata
Weathering
Blagotin
Foeni-Sala
Kadica Brdo
Kozani
Livade
Ljuljaci
Novaka uprija
Opovo
Petnica
Stragari
Vina
Deep
Shallow
Deep
Deep
Shallow
Shallow
Shallow
Deep
Deep
Deep
Deep
Low
Medium
Low
Low
Very high, except in deep pits
Medium
Medium
Low
Low
Low
Low
of the younger class would indicate the presence of herds

during the warmer half of the year. This is in accordance
with expectations since Petnica is a mid-altitude site (and
not a pleasant area to over winter with ovicaprine herds).
On the other hand, the presence of the 6-12 month age
class at Petnica, which would be during the colder half of
the year, is not in accordance with the stated hypotheses
that herds would be moved out of the hill country during
the winter. At the same time, this pattern would indicate
that Petnica is a sedentary year-round occupation.
gaps in the youngest age groups (0 to 1-2 years). The

harvest pattern at Kadica Brdo does not fit any of Paynes
(1973) proposed models for exploitation strategies
(Figures 8.4, 8.24 and 8.25), but conforms to Greenfields
(1988) expectations for a herd based on long-term
stability and animal exploitation with an emphasis on
secondary products.
The remains from Kadica Brdo are not indicative of
transhumance. The harvest profile shows the presence of
all age classes, including the youngest, which implies a
year round availability of the herds. This, in turn,
suggests a non-transhumant movement and continuous
culling of the animals and residential stability throughout
the year. It may be that at this later period, the economy
of highland areas was developed enough to support year
round residence of domestic herds through the production
and storage of winter fodder and appropriate shelter
available for the animals. This does not preclude the
possibility that the bulk of the herd may have been moved
in a transhumant fashion, while a part may have been
kept year-round in proximity to the settlement.
The data from the two lowland sites (Livade and Novaka
uprija) implies that the occupants are participating in a
transhumant economy. The 2-6 month age class is absent
at both lowland sites, which is in accordance with
expectations for a transhumant economyherds are
absent from the lowlands during the summer months.
Additionally, the presence of the 6-12 month class at
Livade, which is contrary to expectations for a lowland
site, and can be traced to the taphonomic history of the
assemblage, with its high rate of weathering (Table 8.54).
Other taphonomic issues arise when considering the
youngest age class (0-2 months), complicating any
interpretations of transhumance. At all of these sites, the
Late Bronze Age levels were subjected to a great deal of
weathering. The youngest age classes are most likely to
disappear under such conditions. The degree of sieving
may also have an effect upon assemblage diversity,
having a greater effect upon some assemblages than
other, but clearly influencing the presence of the youngest
age classes (Table 8.52).
8. The Southern Balkans

Megalo Nisi Galanis was included in the analysis in an
effort to make a comparison with the southern Balkans.
The data from each period are considered separately.
Only ovicaprine remains were abundant enough to be
included in the analysis.
a. Final Neolithic
The harvest profile from Megalo Nisi Galanis shows low
mortality of the 0-2 month and absence of the 2-6 month
age groups. There is a rapid mortality of the 6-12 month
through to the 4-6 year age groups. There are no
indications of the presence of the 6-8 years or 8-10 year
age groups.
7. Early Iron Age

The harvest profile from the high altitude Early Iron Age
site of Kadica Brdo (Figure 8.34) shows a very low
mortality in the youngest age groups (0-6 months),
followed by rapid mortality of age groups 6-12 months
and 1-2 year, followed by a slowing in the 2-3 year, and a
final very rapid mortality rate between 3-4 and 4-6 years.
There is a very low rate of older age groups (6-8 and 8-10
years) in comparison to the earlier groups. There are no
The mortality profile of the Final Neolithic (Figure 8.35)

of Megalo Nisi Galanis most closely resembles the herd
exploitation model for primary products, that is meat
production, during this period (Paynes 1973 Model A
91

kill-off patternsee Figure 8.4).
1-8 month age class followed by rapid mortality of the 830 months age groups. The 30-36 month age group is
missing from the profile. There is again a rapid mortality
rate of the young adult age group. Adult and old adult age
classes are also missing. This site was totally sieved, so
the sample is believed to be well representative of the
youngest age groups that may have been omitted in
unsieved sites. This pattern is essentially repeated at
Blagotin, with two major differences: both the 30-36
months age group and the old adult age group are present
and there is an absence of senile individuals.
The Final Neolithic period at Megalo Nisi Galanis is

temporally contemporaneous to the Eneolithic period of
the central Balkans. Therefore, one would expect to see a
harvest profile similar to those seen in the Eneolithic
from further north. The harvest profile shows the
presence of the youngest age group (0-2 months), the
absence of the next age group (2-6 months), and a high
mortality rate of the subsequent age groups. This appears
to fit exactly with the expectations of a lowland site
involved in a transhumant movement of herds. However,
Megalo Nisi Galanis is a site located on a mid-altitude
plateau. As a result, one would have expected the
opposite pattern.

exploitation of the herd for primary products, that is, meat
production. It most closely resembles Paynes Model A
kill-off pattern (see Figure 8.4).
b. Final Neolithic/Early Bronze Age

The harvest profile of the Final Neolithic/Early Bronze
Age of Megalo Nisi Galanis (Figure 8.36) shows the
absence of the youngest age group (0-2 months),
followed by a somewhat rapid mortality of the 2-6 month
groups and the 6-12 months group. A much more rapid
mortality rate is seen for the 1-2 year group, followed by
a gradual mortality of the remaining groups.
The Early Neolithic harvest profiles imply a year round

availability of the herds which suggests non-transhumant
movement and continuous culling of the animals and
residential stability throughout the year. The profile from
both Foeni-Sala and Blagotin shows the presence of the
youngest age classes (0-1, 1-8 and 8-18 months). The
harvest profiles from the two sites have no statistically
significant differences (Appendix E, Table E9).
Additionally, there are no statistically significant
differences between an average Bos taurus pattern and an
average Ovis/Capra pattern (Appendix E, Table E10) for
the same period (Figure 8.39). The similarity between
both profiles is remarkable, given that one is lowland and
the other is mid-altitude. It would be correct to state that
there is no complementarity between them and that
transhumance is not occurring between highland and
lowland areas.
The mortality profile of the Final Neolithic/Early Bronze

Age remains from Megalo Nisi Galanis most closely
resemble the herd exploitation model for primary
products, that is meat production, during this period
(Paynes 1973 Model A kill-off patternsee Figure 4).
Culling of the youngest age groups is later than in the
Final Neolithic within Megalo Nisi Galanis.
The harvest profile from both the Final Neolithic and
Final Neolithic/Early Bronze Age deposits at Megalo
Nisi Galanis do not fit the pattern hypothesized for a midaltitude site involved in transhumance. This is a region in
which transhumance is expected as an ecological
necessity (cf. Geddes 1983). The environmental
differences between the two regions (Mediterranean and
temperate Europe) leads to different expectations about
the nature of transhumanceit is expected due to
climatic forces in the Mediterranean, while it is not in the
lowlands of the central Balkans. The limited evidence
that we present here supports the expectation that the two
regions would have different patterns of transhumance. In
order to have an effective comparison between the
patterns of the central and southern Balkans, a larger
sample of sites, including those from both high and low
altitude locations, is necessary.
2. Middle Neolithic
Two sites provide information on the Middle Neolithic
harvest profile patterns of Bos taurus. These are the
lowland site of Stragari (Figure 8.40) and mid-altitude
Petnica (Figure 8.39). The harvest profile of Stragari
shows the same general pattern as those from Early
Neolithic. There is low mortality of the youngest age
group (0-1 month) followed by an absence of the 1-8
month age group. A rapid mortality rate of the 8-30
months groups is seen. The 30-36 months and young
adult groups are both missing from the profile. A rapid
mortality rate of the adult age group is seen and an
absence of the old adult group. The harvest profile of
Petnica shows similar patterns. But there are several
differences: there is a low mortality from 8-18 months
and rapid mortality of the young adult group. The adult
group is also missing from the profile. While both sites
have adequate sample sizes, Stragari (n=29) and Petnica
(n=22), both are missing important age classes for this
investigation (Figure 8.40). Significantly, while both
sites include the remains of the 0-1 month age class, both
are missing the 1-8 months age class.
C. Bos taurus
1. Early Neolithic
In the Early Neolithic, there are two sites with sufficient
sample sizes to produce harvest profiles for Bos taurus the lowland site of Foeni-Sala (Figure 8.37) and the
mid-altitude site of Blagotin (Figure 8.38). The harvest
profile shows a rapid mortality of the youngest age
groups (0-1 month), a slight reduction in mortality of the

exploitation of the herd for primary products, which is for
92

meat, hide and bone production. It most closely
resembles Paynes Model A kill-off pattern (see Figure
8.4).
Petnica, Vina, Opovo and Selevac show the presence of

the youngest age classes (0-1, 1-8 and 8-18 months). This
implies a year round availability of the herds, which in
turn, suggests a non-transhumant movement and
continuous culling of the animals and residential stability
throughout the year. It would be correct to state that there
is no complementarity between them and that
transhumance is not occurring between highland and
lowland areas. This pattern conforms to the pattern
established in the Early Neolithic. There are no
statistically significant differences between the harvest
profiles of the four Late Neolithic sites (Appendix E,
Table E12).
The harvest profiles would seem to imply that the

occupants of both sites are moving their herds in a
transhumant fashion. This would be the pattern for
lowland sites. While Stragari can be considered a lowland
site, Petnica is a clear mid-altitude site. This contradiction
is not a function of small sample size, since both have
adequate samples (above). The statistical analysis
indicates that there are some statistically significant
differences between the profiles for this period (Appendix
E, Table E11). However, this difference should only be
considered to be a moderate (as the usual p value is
greater than 0.05 and the exact value is only slightly less
than 0.05). Even though the sample size is high for this
investigation, the age groups are not appropriately
representative. However, since a significant number of
age groups are missing from both profiles, it may be
suggested that there is a problem with how representative
is the sample, rather than indication of transhumance.
4. Eneolithic
In the Eneolithic period, there is only one site with
sufficient dataBlagotin. The harvest profile from
Blagotin (Figure 8.48) shows a complete absence of the
youngest age groups (0-8 months) followed by a steep
mortality rate for the 8-18 month group through to the
young adult group. There is the complete absence of the
oldest age groups, adult, old adult and senile.
3. Late Neolithic
There are four sites with data from the Late Neolithic.
The lowland sites of Opovo (Figure 8.43) and Vina
(Figure 8.44), the mid-altitude site of Petnica (Figure
8.45) and Selevac (Figure 8.46) analyzed by Legge
(1990) all have sufficient sample sizes of Bos taurus for
the Late Neolithic.
The pattern does not fit any of Paynes proposed

specialized herd production models (Figures 8.4, 8.24 and
8.25). It does, however, conform to Greenfields (1988)
expectations for a herd based on long-term stability and
animal exploitation with a mixed subsistence economy
(both primary and secondary products) and an emphasis
on secondary products.
The site of Opovo shows a low rate of mortality for the

youngest age groups (0-1 through to 8-18 months). This
is followed by a steep mortality of the 18-30 months
group. The 30-36 month age class is missing from the
sample. Finally, there is a rapid rate of mortality for the
young adult age class, although more gradual than the
previous classes. The adult and senile age classes are
missing from the sample. At Vina, there is a more
rapid mortality rate of the youngest age classes and a
very low mortality rate of the 30-36 month group
through to the adult age group. The senile age group is
missing at Vina. The mid-altitude site of Petnica
indicates the same harvest profile pattern. There is a
gradual rate of mortality for the youngest age groups (01 months through to 8-18 months). The steep mortality
of the 18-30 months seen Opovo and Vina, extends to
the 30-36 months group. This is followed by the low
rate of mortality for the oldest age groups. The harvest
profile from Selevac shows slightly greater mortality of
the youngest age groups, and then follows a similar
pattern through the older age groups.
Blagotin is a mid-altitude site and would be expected to

conform to the highland pattern within a transhumant
strategy. There would be an expected absence of the
youngest age classes (0-1 month). This is the pattern
that is seen. However, as discussed above, this age class
is unreliable for indications of transhumance based on
taphonomic issues. One should then look at the expected
presence of the next age group (1-8 months) that would
be expected to dominate the harvest profile. The
complete absence of this age class would indicate that
the harvest profile of Blagotin does not show results that
conform to the expectations of a highland site involved
in the transhumant movement of herds. A problem with
this period is that there is an absence of lowland site
locations with sufficient data with which to compare
them.
Two sites yielded sufficient data to construct harvest
profiles for the Early/Middle Bronze Age - Ljuljaci and
Vina. The mid-altitude site of Ljuljaci (Figure 8.49)
combines the data from both the Early/Middle Bronze
Age period. The profile shows an absence of the youngest
age class (0-1 months). This is followed by a rapid
mortality rate of the 1-8 month group through to the 1830 month group. The age class 30-36 months is missing
from the profile. This is followed by a low mortality rate
of young adults and a more rapid mortality of adult age
groups. The old adult group is missing from the profile.
The harvest profiles of all four Late Neolithic sites

(Figure 8.47) most closely resemble the herd exploitation
model for primary products, that is, meat production,
during this period (Paynes 1973 Model A kill-off
patternsee Figure 8.4).
In terms of transhumance, the harvest profiles from the
93
100
80
% Age Survival
60
40
20
0
0 months
0-2 months
A

B
C
1-2 years
D
2-3 years
E
3-4 years
F
4-6 years
G
6-8 years
H
8-10 years
I
Figure 8.27. Harvest profile (Ovis/Capra) Middle Bronze Age Vina
100
% Age Survival
80
60
40
20
0
0 months
0-2 months
A
2-6 months
B
6-12 months
C
1-2 years
D
2-3 years
E
3-4 years
F
4-6 years
G
6-8 years
H
Figure 8.30. Harvest profile (Ovis/Capra) Late Bronze Age Novaka uprija
94
8-10 years
I
100
% Age Survival
80
60
40
20
0
0 months
0-2 months
A
2-6 months
B
6-12 months
C
1-2 years
D
2-3 years
E
3-4 years
F
4-6 years
G
6-8 years
H
8-10 years
I
Figure 8.31. Harvest profile (Ovis/Capra) Late Bronze Age Livade
100
% Age Survival
80
60
40
20
0
0 months
0-2 months
A
2-6 months
B
6-12 months
C
1-2 years
D
2-3 years
E
3-4 years
F
4-6 years
G
Figure 8.32. Harvest profile (Ovis/Capra) Late Bronze Age Petnica
95
6-8 years
H
8-10 years
I
Petnica (mid-altitude)
Novacka Cuprija (lowland)
Livade (lowland)
100
% Age Survival
80
60
40
20
0
0 months
0-2 months
A
2-6 months
B
6-12 months
C
1-2 years
D
2-3 years
E
3-4 years
F
4-6 years
G
6-8 years
H
8-10 years
I
Figure 8.33. Harvest profile (Ovis/Capra) Late Bronze Age
100
% Age Survival
80
60
40
20
0
0 months
0-2 months
A
2-6 months
B
6-12 months
C
1-2 years
D
2-3 years
E
3-4 years
F
4-6 years
G
6-8 years
H
Figure 8.34. Harvest profile (Ovis/Capra) Early Iron Age Kadica Brdo
96
8-10 years
I
100
% Age Survival
80
60
40
20
0
0 months
0-2 months
A
2-6 months
B
6-12 months
C
1-2 years
D
2-3 years
E
3-4 years
F
4-6 years
G
6-8 years
H
8-10 years
I
Figure 8.35. Harvest profile (Ovis/Capra) Final Neolithic Megilo Nisi Galanis
100
% Age Survival
80
60
40
20
0
0 months
0-2 months
A
2-6 months
B
6-12 months
C
1-2 years
D
2-3 years
E
3-4 years
F
4-6 years
G
6-8 years
H
8-10 years
I
Figure 8.36. Harvest profile (Ovis/Capra) Final Neolithic/Early Bronze Age Megilo Nisi Galanis
97
100
% Age Survival
80
60
40
20
0
0 months
0-1 month
A

B
C
18-30
months
D
30-36
months
E
young adult
F
adult
G
old adult
H
senile
I
Mandibular Stages and Absolute Ages
Figure 8.37. Harvest profile (Bos taurus) Early Neolithic Foeni-Sala
100
% Age Survival
80
60
40
20
0
0 months
0-1 month
A

B
C
18-30
months
D
30-36
months
E
young adult
F
adult
G
Mandibular Stages and Absolute Ages
Figure 8.38. Harvest profile (Bos taurus) Early Neolithic Blagotin

98
old adult
H
senile
I
Bos taurus Early Neolithic Pattern
Ovis/Capra Early Neolithic Pattern
100
% Age Survival
80
60
40
20
0
A
old adult
H
senile
I
Mandibular Stage
Figure 8.39. Harvest profile Early Neolithic Bos vs. Ovis/Capra
100
% Age Survival
80
60
40
20
0
0 months
0-1 month
A
1-8 months
B
8-18 months 18-30 months 30-36 months young adult

C
D
E
F
adult
G
Figure 8.40. Harvest profile (Bos taurus) Middle Neolithic Stragari
99
100
% Age Survival
80
60
40
20
0
0 months
0-1 month
A
1-8 months
B

C
D
E
F
adult
G
old adult
H
senile
I
Figure 8.41. Harvest profile (Bos taurus) Middle Neolithic Petnica
Stragari
Petnica
100
% Age Survival
80
60
40
20
0
0 months
0-1 month
A
1-8 months
B

C
D
E
F
adult
G
Figure 8.42. Harvest profile (Bos taurus) Middle Neolithic
100
old adult
H
senile
I
100
% Age Survival
80
60
40
20
0
0 months
0-1 month
A

B
C
18-30
months
D
30-36
months
E
young adult
F
adult
G
old adult
H
senile
I
Figure 8.43. Harvest profile (Bos taurus) Late Neolithic Opovo

The Middle Bronze Age mortality profile of Bos taurus
remains from Vina (Figure 8.50) indicates some interesting
patterns. There is a gradual mortality of the youngest age
groups (0-1 month and 1-8 months). This is followed by a
rapid mortality of the 8-18 month group and then a low
mortality of the 18-30 month group. There is a complete
absence of age stages 18-30 months through to adult.
pattern of a highland site, with the absence of the 0-1

month group and the occurrence of rapid mortality of
the 1-8 months age group. The lowland site of Vina
does not fit the transhumant pattern for lowland sites.
As expected the youngest age group (0-1 month) is
present, however the next age group (1-8 months) is
also present contrary to expectations for transhumance.
The Ljuljaci harvest profile does not fit any of Paynes

proposed specialized herd production models (Figures 8.4,
8.24 and 8.25). It does, however, conform to Greenfields
(1988) expectations for a herd based on long-term stability
and animal exploitation with a mixed subsistence economy
(both primary and secondary products) and an emphasis on
secondary products. The unusual pattern at Middle Bronze
Age Vina indicates that there is no culling of the
subadult/adult age groups. This pattern strongly indicates a
traction profile. It seems likely that the subadult/adult age
classes were kept in the lowland sites as agents of traction
for agricultural cultivation.
6. Late Bronze Age

Only a single site, the lowland site of Livade (Figure
8.51) has sufficient sample size of Bos taurus to construct
a harvest profile for the Late Bronze Age period. The
harvest profile indicates an absence of the youngest age
classes (0-1 and 1-8 months). There is a rapid mortality of
the 18-30 month age group and the 30-36 month group is
absent from the profile. This is followed by a low
mortality rate of the young adult and adult age classes
and rapid mortality of the old adult and senile age classes.
The data from Middle Bronze Age Vina indicates that

there is no culling of the subadult/adult age groups. This
pattern strongly indicates a traction profile, which is very
similar to the wool profile for sheep (Figures 8.25;
Greenfield 1988). It seems likely that the subadult/adult
age classes were kept in the lowland sites as agents of
traction for agricultural cultivation.
As Ljuljaci is a mid-altitude site, the harvest profile is
expected to show a highland pattern in a regional
transhumant economy. Ljuljaci shows the expected
101
The harvest profile from Livade does not fit any of

Paynes proposed models for kill of patterns indicating a
single exploitation strategy (Figures 8.4, 8.24 and 8.25).
Instead, it appears to follow Greenfields (1988) pattern
for mixed exploitation strategy of both primary and
secondary products, with a greater emphasis upon the
latter, and herd stability.
Assuming a transhumant movement of the domestic herds,
the absence of the 1-8 month group is expected in the
lowland site of Livade. The presence of the next age group
8-18 months also fits with the expectations of a transhumant
movement. As such, it seems that the data implies that

animals are being moved in a transhumant fashion.
analysis can take place. We suggest that transhumance can

only be identified during the next step of the analysis,
when sites are compared on a regional scale. It is the shifts
in age group representations in the harvest scale of all sites
across time that transhumance becomes apparent for the
first time. In this section, the data to support this assertion
are presented.
7. Early Iron Age

The high altitude site of Kadica Brdo (Figure 8.50)
provides the only harvest profile for the Early Iron Age.
There is a very low mortality of the youngest age group (01 months), and complete absence of the 1-8 month age
group. This is followed by a low mortality rate of the 8-18
and rapid mortality of the 18-30 month groups. The 30-36
month group is missing from the profile. There is a rapid
mortality rate of the young adult group through to senile,
although at a reduced rate from the previous age groups.
The harvest pattern at Kadica Brdo does not fit any of
Paynes (1973) proposed models for exploitation
strategies (Figures 8.4, 8.24 and 8.25), but conforms to
Greenfields (1988) expectations for a herd based on
long-term stability and animal exploitation with an
emphasis on secondary products.
The Bos taurus remains from the Early Iron Age highland
site of Kadica Brdo yield a problematic harvest profile. It
is one of the larger sample sizes of remains included in
this study, yet there is a significant gap in the youngest
age group (1-8 months). This is exactly the age group that
is expected to be present if there were indications of
transhumance since young animals would likely be present
in the warm weather highland pastures. Therefore, the
absence of this age group is startling, to the say least, and
difficult to explain as a function of simply transhumance.
An alternative explanation comes from the fact that the
youngest age categories are not exploited when a strategy
of herd stability is pursued. Such a pattern is found in very
unpredictable climates and marginal environments, such as
existed in the high mountains of eastern Bosnia (where
Kadica Brdo is located). In such environments, herds often
tend to pursue a herd management style that minimizes
risk. Part of this strategy is to keep as many animals alive
as possible. The consequence would be minimal harvesting
of veal (as represented by the immature age groups
Arnold and Greenfield 2003; Redding 1984). This is
supported by the absence or near absence of individuals in
most of the immature age groups and the high rate of
mortality among adults. This is probably the best way to
interpret the harvest profile.
Several problems exist with the use of harvest profiles for

the investigation of transhumance. One of the major
problems is that the age classes of the standard techniques
for the recording of tooth eruption and wear do not match
those necessary for identifying the transhumant movement
of herds. This lack of precision limits the elucidation of
transhumance from these traditional zooarchaeological
approaches. Additionally, within this investigation,
hypotheses for expectations of transhumance were
constructed with the assumption of a complete absence of
specific age groups, an extreme expectation.
This problem can be seen in the timing of movement and
birth in the northern Balkans. The time of lambing/calving
for all the major transhumant domestic species (cattle and
sheep/goat) is February. As the herds would not be moved
immediately after lambing/calving, the herd would be
found in the lowland areas from February to April.
Animals found in lowland areas (or archaeological sites)
would be expected to be between birth and 3 months old,
plus the >1 year old animals from previous seasons. Herds
are moved into the highland areas from roughly May to
September (based on ethnographic observations). Animals
expected in the highland areas from that seasons birthing
are between four and eight months old. The domestic herds
return to the lowland areas in the autumn (roughly
October) when the youngest cohort is at least 8 months
old (Greenfield 1999a: 19).
In the transhumant movement of both domestic cattle and
ovicaprine herds, it is expected that the youngest age group
(class A, which is 0-2 months in sheep/goat, 0-1 months in
cattle) would be found only in the lowlands and would
indicate a late winter/spring occupation. Unfortunately, the
paucity of this age group, due to the taphonomic issues
discussed above, limits the identification and
interpretations for a transhumant movement of herds.
Utilization of the next age group for sheep, goat and cattle
for the elucidation of transhumance is also problematic. As
the sheep and goat herds move into the highlands from
roughly May to September, the next age stage (class B, 2-6
months in sheep/goat, 1-8 months in cattle) overlaps the
hypothesized timing of movement into the highland areas.
A two months old individual would be found only in
lowland areas, while a six months old individual would be
expected to be found in the highlands. The same overlap is
present with the return movement to the lowlands in the
autumn. The division of age grades is simply not subtle
enough. It would be ideal to be able to split the ovicaprine
2-6 month class and the cattle 1-8 month age class into
finer categories to enable finer distinctions. However, there
is as yet no valid analytical justification for this.
III. Regional scale of analysis second stage of analysis

In the above section, where each site or group of sites is
examined period by period, we subjected the reader to a
litany of problems in the identification of transhumance
with zooarchaeological data. The essential flaw in such an
approach is that transhumance is a regional economic
strategy and therefore must be searched for on a regional
scale. Trying to identify whether one or another site is part
of a transhumant pattern is an exercise in frustration, only
occasionally yielding data in accordance with expectations.
It is a less productive, but necessary first step in any
analytical approach to transhumance. The data must be
examined in such a manner before the next step in the
102

The lack of precision in the age class structure in
combination with the assumption of a complete absence of
specific age groups (Table 8.55) in the original statements
of the hypotheses limits the search for transhumant
movement using traditionally structure harvest profiles
(Payne 1973). As a result, the original hypotheses were
reformulated to expect an increased percentage of the
relevant age groups in each region. For example, one
should expect to see a greater percentage of the 0-2 month
age group for sheep and goat, and 0-1 month age group for
cattle in the lowlands, in comparison to mid-altitude and
highland areas. A greater percentage of the 2-6 months
sheep and goat age group and the 1-8 months cattle group
were expected in mid-altitude and highland sites in
comparison to the lowland sites. Finally, the proportion of
the 6-12 months sheep and goat age group and the 8-18
month cattle age group is expected to be higher in lowland
sites in comparison to mid-altitude and highland sites. This
reformulation of hypotheses is a means of dealing with the
problem of the overlap between the age classes and the
transhumant movement of domestic herds.
transhumance was occurring, one expects the youngest age

group (0-2 months) to show the highest percentage of the
assemblage in the lowland sites. This hypothesis is difficult
to evaluate with the Sus scrofa data, as no lowland sites
have produced any remains from this age group. The 2-7
months age group would be expected to account for a
higher percentage of the assemblage in the mid-altitude
and highland sites. In the Neolithic period, a random
pattern is evident (Figure 8.53). Early Neolithic Blagotin, a
lowland site, shows the highest percentage, Late Neolithic
Vina and Opovo, both lowland sites, show an equal
percentage as Late Neolithic Petnica, a mid-altitude site.
Finally, the 7-14 months age group should dominate the
lowland sites. While Late Neolithic Vina does show the
highest percentage, it is only minimally larger than Petnica,
a mid-altitude site. There are no consistent patterns of
presence/absence or increased percentage in lowland, midaltitude or highland sites that might indicate a transhumant
movement of domestic pig in the Neolithic.
The pattern in the Post Neolithic is somewhat different.
Again the youngest age group cannot be considered due to
its absence. Interestingly, the first expectation of the
hypothesis is shown in the Post Neolithic (Figure 8.54).
The 2-7 months age group accounts for a higher percentage
of the assemblage in the mid-altitude and highland sites.
However, this age group is missing from the Early and
Middle Bronze Age levels of Ljuljaci, limiting a full
regional expression of the pattern. In considering the final
expectation of the hypothesis, one expects the 7-14 months
age group should dominate the lowland sites. This is
clearly not the pattern with the Post Neolithic pig data,
where there is no consistent pattern of increased percentage
of this age group in lowland or mid-altitude sites that might
indicate a transhumant movement of domestic pig in this
period (Figure 8.53). As a result, the hypothesis for
transhumant movement of pigs cannot be supported. There
are no indications of the transhumant movement of
domestic pig in the Post Neolithic.
For all domestic animals (Sus scrofa, Ovis/Capra and Bos

taurus), the paucity of the youngest age group (0-2 months
for pig, sheep and goat, and 0-1 month in cattle) is
problematic. Taphonomic issues, specifically the extent of
sieving and weathering at the sites, have greatly affected
the recovery of the 0-2 month age group throughout this
investigation. This makes it difficult to evaluate the
reformulated hypotheses with this age group. Therefore,
the focus is on the other age groups.
The high-altitude site of Kadica Brdo is not considered
with the reformulated hypotheses due to several
observations discussed above. It may be that by this later
period, highland areas have developed enough to support
year round residence of domestic herds through the
production and storage of winter fodder and appropriate
shelter available for the animals. Alternatively, the
unpredictable climate and marginal environment of the
high mountains of eastern Bosnia (where Kadica Brdo is
located) may have contributed to a herd management style
that minimizes risk.
B. Transhumant movement of ovicaprines

The Ovis/Capra data present some interesting
observations. No remains from the two youngest age
groups are present in mid-altitude or highland sites from
Neolithic sites. One can only evaluate the 2-6 months and
6-12 months age groups for their agreement with the
hypothesis. The 2-6 months age group would be expected
to account for a higher percentage of the assemblage in the
mid-altitude and highland sites. In the Neolithic, it is the
lowland site of Foeni-Sala that shows the highest
percentage (Figure 8.55). Mid-altitude Blagotin and
lowland Vina show an equal percentage. The 6-12 months
age group should dominate the lowland sites. A random
patterning of age groups between lowland and mid-altitude
sites is evident and implies no transhumant movement of
domestic sheep and goats in the Neolithic.
A. Transhumant movement of pigs

Seasonal statements about pigs are based upon the
assumption of a single birthing period during the year. This
is because pigs were not kept ethnographically in stalls or
fed throughout the year in this region. Instead, they were
left to forage in the forests, with a pig herder. Whenever
pigs were needed for food, the herder brought back one or
more (Halpern 1999). As a result, it is unlikely that the
modern conditions for multiple births throughout the year
would have occurred in the past.
Domestic pig is considered a control for the transhumant
movement of herds as these animals were not subjected to
the transhumant movement characteristic of either cattle or
sheep/goat. With the restructured hypotheses above, if
In the Post Neolithic (Figure 8.56), again the 0-2 month

age group is eliminated from consideration due to lack of
103

availability of data. The 2-6 month age group depictions fit
with the reformulated hypotheses, with the mid-altitude
site showing a greater percentage of 2-6 months animals
than the lowland site. Again, taphonomic issues play a role
here with only one mid-altitude site and a single lowland
site with sufficient remains.
in a transhumant pattern. Blagotin again dominates, with

Middle Neolithic Stragari also high. However, the other
sites display a random patterning of age groups between
lowland and mid-altitude sites and imply no transhumant
movement of domestic cattle in the Neolithic.
The bar graph regional analysis of the Post Neolithic
cattle data conforms to the expectations of transhumant
movement (Figure 8.58). Paucity of remains from the 0-1
month age group eliminates this group from evaluation.
In the 1-8 months age group, one observes a greater
percentage of individuals are slaughtered in the midaltitude site of Ljuljaci in comparison to the lowland site
of Vina. The reverse pattern is present in the 8-18
months age group between the sites of Livade and Vina.
A greater percentage of 8-18 month individuals are
slaughtered in the lowlands than the mid-altitude sites.
The original hypotheses state that the youngest
individuals would be culled from the herds in the
highland and mid-altitude pastures. This would be the
infants and the juveniles (0-8 months old).
Interestingly, the 6-12 month age group shows exactly the

opposite of the reformulated hypotheses. The mid-altitude
sites show more 6-12 months individuals than are present
at the lowland sites. Petnica displays significantly more
individuals (50% of its total assemblage). An explanation
for these observations is the divergence between the
division of the age stages established from the tooth
eruption and wear and the hypothesized timing of the
transhumant movement. The original hypothesis states that
the transhumant movement of domestic stock is predictable
and identifiable. The herd will move into highland pastures
in the early spring, soon after lambing/calving occurs and
returns to the lowlands during the autumn. In the northern
Balkan area, the time of lambing/calving for sheep/goat
and cattle is February. As the herds would not be moved
immediately after lambing/calving, the herd would be
found in the lowland areas from February to April. It is
hypothesized that the herds would then be moved into the
highland areas from roughly May to September. The
domestic herds return to the lowland areas in the autumn
(roughly October). The specific hypotheses state that the
youngest individuals would be culled from the herds in the
highland and mid-altitude pastures. This age category
would be the infants and the juveniles 0-8 months old
(Greenfield 1999a). If these animals were being
slaughtered in the mid and high-altitude sites towards the
end of that period, i.e. at 7-8 months, just before the herds
move back into the lowlands, they would appear in the
harvest profiles in the 6-12 month age group. This is what
is indicated by the Post Neolithic data. Consequently,
while the individual site tooth wear and eruption age
grades are somewhat problematic for seasonality and
elucidation of the temporal aspects of a transhumant
movement, a reconsideration of the data does argue for
support for the transhumant movement of sheep, goat and
cattle herds occurring in the Post Neolithic.
The data from the lowland site of Livade are problematic,

in that they do not fit the pattern. While it would be easy
to dismiss this anomaly as a limitation imposed by small
sample size (n=16), the mid-altitude site of Ljuljaci has a
similar sample size (n=15). It is likely that the problem
arises again from the division of age groups at 8 months.
One must question that when an 8 month old individual is
coded for tooth wear and eruption, are they coded into the
end of class B, 1-8 months, or the beginning of class C, 818 months? However, the pattern emerging in the data
suggests a transhumant movement. In comparison, the
Neolithic bar graphs display only a random patterning of
age groups between lowland and mid-altitude sites,
implying no transhumant movements.
IV. Comparison of tooth wear and eruption ageing
methods third stage of analysis
The modern ovicaprine data, while collected as the control
sample to establish the timing of the formation of the
cementum increments to accurately investigate the nature
of the increments visible in the archaeological sample, also
provides a methodological contribution. It is possible to
test the agreement of the incremental cementum layers in
the teeth with the eruption/wear stages and with the actual
ages. It will act as a three-way comparison between tooth
eruption and wear, incremental structures and true ages
(Ariane Burke, personal communication, 2001).
Additionally, the agreement of the two recording systems
used for tooth eruption and wear can be evaluated (Grant
1975; Payne 1973).
C. Transhumant movement of cattle

Better representation of the youngest age group (0-1
month) of Bos taurus from the Neolithic allows a full
evaluation of the hypotheses (Figure 8.57). One expects a
greater percentage of the 0-1 month age group in the
lowland sites, in comparison to mid-altitude and highland
sites if transhumance was occurring. This expectation is
not met, with mid-altitude Blagotin showing the greatest
percentage of the assemblage. Other sites show a random
pattern. The 1-8 months age group would be expected to
account for a higher percentage of the assemblage in the
mid-altitude and highland sites. Again, Blagotin shows
the highest percentage of the assemblage, but the
differences between the sites are not large. Finally, the 818 months age group should dominate the lowland sites
The two tooth wear and eruption recording methods utilize

similar diagrams for their representation of the different
stages. Grant (1975) has many more stages and the
diagrams are more representational. In contrast, Payne
(1973) utilizes less stages and more stylistic representations.
104
100
% Age Survival
80
60
40
20
0
0 months
0-1 month
A
1-8 months
B

C
D
E
F
adult
G
old adult
H
senile
I
Figure 8.44. Harvest profile (Bos taurus) Late Neolithic Vina
100
% Age Survival
80
60
40
20
0
0 months
0-1 month
A
1-8 months
B

C
D
E
F
adult
G
Figure 8.45. Harvest profile (Bos taurus) Late Neolithic Petnica
105
old adult
H
senile
I
100
% of age survival
80
60
40
20
0
0
0-1 months
A

B
C
18-30
months
D
30-36
months
E
young adult
F
adult
G
old adult
H
senile
I

Selevac
Opovo
Petnica
Vina
Figure 8.46. Harvest Profile (Bos taurus) Middle/Late Neolithic Selevac
Petnica
Vinca
Opovo
100
% Age Survival
80
60
40
20
0
0 months
0-1 month
A
1-8 months
B

C
D
E
F
adult
G
Figure 8.47. Harvest profile (Bos taurus) Late Neolithic
106
old adult
H
senile
I
100
% Age Survival
80
60
40
20
0
0 months
0-1 month
A
1-8 months
B

C
D
E
F
adult
G
old adult
H
senile
I
Figure 8.48. Harvest profile (Bos taurus) Eneolithic Blagotin
100
% Age Survival
80
60
40
20
0
Suggested
Age
Stage
0 months
0-1 month
A

B
C
18-30
months
D
30-36
months
E
young adult
F
adult
G
old adult
H
Figure 8.49. Harvest profile (Bos taurus) Early/Middle Bronze Age Ljuljaci
107
senile
I
100
% Age Survival
80
60
40
20
0
0 months
0-1 month
A

B
C
18-30
months
D
30-36
months
E
young adult
F
adult
G
old adult
H
senile
I
Figure 8.50. Harvest profile (Bos taurus) Middle Bronze Age Vina
100
% Age Survival
80
60
40
20
0
0 months
0-1 month
A

B
C
18-30
months
D
30-36
months
E
young adult
F
adult
G
Mandibular Stage and Age
Figure 8.51. Harvest profile (Bos taurus) Late Bronze Age Livade
108
old adult
H
senile
I
100
% Age Survival
80
60
40
20
0
0 months
0-1 month
A
1-8 months
B

C
D
E
F
adult
G
old adult
H
senile
I
Figure 8.52. Harvest profile (Bos taurus) Early Iron Age Kadica Brdo
60
50
% of assemblage
40
Blagotin Early Neolithic Lowland
Petnica Middle Neolithic Mid-altitude
Petnica Late Neolithic Mid-altitude
Vina Late Neolithic Lowland
Opovo Late Neolithic Lowland
30
20
10
0
0-2
months
2-7
months
7-14
months
14-21
months
21-27
months
27-36
months
adult
old adult
senile
Age groups
Figure 8.53. Percentage of Age Groups (Sus scrofa) Neolithic
109
45
40
35
% of assemblage
30
Petnica Eneolithic Mid-altitude
Novaka uprija Early Bronze Age Lowland
Ljuljaci Early Bronze Age Mid-altitude
Ljuljaci Middle Bronze Age Mid-altitude
Livade Late Bronze Age Lowland
Kadica Brdo Early Iron Age Highland
25
20
15
10
0
0-2
months
2-7
months
7-14
months
14-21
months
21-27
months
27-36
months
adult
old adult
senile
Age groups
Figure 8.54. Percentage of age groups (Sus scrofa) Post-Neolithic
30
Foeni-Sala Early Neolithic Lowland
25
Blagotin Early Neolithic Mid-altitude
% of assemblage
20

15
10
0
0-2
months
2-6
months
6-12
months
1-2 years 2-3 years 3-4 years 4-6 years 6-8 years
8-10
years
Age groups
Figure 8.55. Percentage of age groups (Ovis/Capra) Neolithic
110
70
Novaka uprija Eneolithic

Lowland
Novaka uprija EBA Lowland
60
Vina MBA Lowland

Novaka uprija LBA Lowland
50
% of assemblage
Livade LBA Lowland

Petnica Eneolithic Mid-altitude
40
Petnica LBA Mid-altitude

Kadica Brdo EIA Highland
30
20
10
0
0-2
months
2-6
months
6-12
months
1-2 years 2-3 years 3-4 years 4-6 years 6-8 years
8-10
years
Age groups
Figure 8.56. Percentage of age groups (Ovis/Capra) Post Neolithic

25
% of assemblage
20
Foeni-Sala Early Neolithic Lowland

Blagotin Early Neolithic Mid-altitude
Stragari Middle Neolithic Lowland
Petnica Middle Neolithic Mid-altitude
Opovo Late Neolithic Lowland
15
10
0
0-1 month
1-8
months
8-18
months
18-30
months
30-36
months
young
adult
adult
old adult
senile
Age groups
Figure 8.57. Percentage of age groups (Bos taurus) Neolithic
111
40
Vina MBA Lowland
35
Livade LBA Lowland
30
% of assemblage
Ljuljaci E/MBA Mid-altitude

25
Kadica Brdo EIA Highland
20
15
10
0
8-18
months
18-30
months
30-36
months
young
adult
adult
old adult
senile
Age groups
Figure 8.58. Percentage of Age Groups (Bos taurus)

The comparative collection demonstrates excellent
agreement between the two methods (see Appendix C).
Within this sample,
Paynes mandibular wear stage C is

Grant mandibular wear stages 9-13;
Paynes mandibular wear stage D is
Grants mandibular wear stage 21-26;
Paynes mandibular wear stage E is
Grants mandibular wear stage 33; and
Paynes mandibular wear stage G is
Grants stage 41.
increments that correspond well with absolute age.

However, this observation must be cautiously applied as
only one sheep in the collection demonstrates this pattern
(Sheep #1 Figure 8.59). This may be due to the fact that
the modern comparative collection consists predominantly
of younger (<2 years) animals.
equivalent to
equivalent to
In those samples where only one growth layer group is

expected, it is often indistinct or unobservable. This may
be due to the fact that the layers are indistinct until the
individual is old enough to have formed successive layers.
In long living animals such as ungulates, that form the first
growth layer of cementum during their first winter, the
incremental line of the first growth layer is less distinct that
the incremental lines of growth layers that are formed later
(Klevezal 1996). Klevezal further notes that in one-yearold animals there is considerable frequency of indistinct
first incremental lines. This pattern is less frequent in older
animals. These observations seem to indicate that the first
incremental line is not very distinct in the period of its
formation, but becomes more distinct with age. However,
it may not become as distinct as the incremental lines
formed during the later years (Klevezal 1996: 41). The
cause of this poor contrast of the first incremental line is
explained by the period of its formation. Seasonal rhythms
are less distinct in young than older animals. As a result,
the annual growth ring formed during the winter is not as
distinct in younger animals (Klevezal 1996). The increase
in the distinctiveness of this incremental line over time is
equivalent to
equivalent to
Additionally, the modern comparative samples were used

as a test for Hambletons method (1999) for converting
the results of different analyses of Ovis/Capra
mandibular tooth wear into a similar format.
Absolute age data for the modern comparative material
was provided by ear tag information where available
(not all animals had ear tags) and by the producer. The
absolute ages that are linked to the various tooth
eruption and wear stages by various other researchers
(e.g. Payne 1973; Grant 1985) are in close agreement
with the known ages from our modern comparative
sample (Appendix C).
It can also be noted that sheep and goats form annual
112
Figure 8.59. Modern comparative thin sectioning sample (Sheep #1)
Figure 8.60. Modern comparative thin sectioning sample (Goat 14 b)
Figure 8.61. Modern comparative thin sectioning sample (Goat #13 b)
113
Figure 8.62. Archaeological thin sectioning sample (Kadica Brdo sample #2)
Figure 8.63 .Archaeological thin sectioning sample (Kadica Brdo sample #4)
114
Figure 8.66. Archaeological thin sectioning sample (Vina sample #1)
Figure 8.67. Archaeological thin sectioning sample (Vina sample #10)
115

readable samples indicated a cold season slaughter.
explained by the secondary deposition of minerals

(Klevezal 1996: 41) that is, secondary calcification in the
already formed cementum over time (Klevezal 1996: 82).
At the lowland site of Vina, nine teeth were sectioned.

Five were unreadable, however. Of the remaining four
teeth, the number of increments counted generally agrees
with the age estimates obtained from tooth wear and
eruption. Some slight variation was found, such as with
Vina Sample #1, which has 9 Growth Line Groups
(counted by the author), whereas the tooth wear and
eruption indicates an age of between six and eight years.
However, the independent reading by Burke indicates 6
growth line groups, which is exactly in line with what
would be expected from the tooth wear and eruption.
V. Cementum analysis fourth stage of analysis

In this section, we will discuss the cementum analysis of
both the modern and ancient specimens.
A. The modern comparative sample
The number of growth line groups that can be counted in
the cementum of each tooth and the nature of the final
increment are recorded for each slide. The readings are
summarized in Table 7.50. The following observations can
tentatively be made on the comparative collection in order
to aid in interpretation of the archaeological sample.
The determination of seasonality estimates was

problematic for the site of Vina. The nature of the final
increment was determinable for only three of the readable
slides (Vina Samples #1, 6 and 10 Figures 8.66 and
8.67). Based on the observations established with the
modern comparative sample, these were interpreted by
the author as the forming of an annulus with a bright
outer increment. This would indicate a cold season death.
Burkes readings agreed with only one of these
determinations. Two of Burkes determinations indicated
the final zone is a growth zone. This would indicate a
warm season death. This problem highlights one of the
major problems in cementum studies subjectivity of
readings and variation between observers. As a result, it
is difficult to conclusively determine the seasonality of
death for Vina samples. If Arnolds determinations are
accepted, this would indicate that the animals died during
the warm months of the year. If Burkes determinations
are correct, this would indicate that the animals were
present at the site during both cold and warm seasons.
It appears that the annulus forms between November and

February. This is suggested by the bright outer increment
seen on teeth from animals slaughtered during this period.
It is visible in both the sheep and the goat samples, but is
most clearly shown by Goat sample no. 14 (slide b
Figure 8.60). Therefore, as the goats were born in April
and the sheep in May, it can be expected that a growth
zone will be apparently from April through to November.
Several readings had to be taken at the end of the root. This
has been demonstrated to be not an ideal place for
readings, and is demonstrated by the occurrence of two
annuli on the root of an 8-month-old goat (Figure 8.61).
By determining the month of the year when annual
growth rings are formed, it becomes possible to interpret
the archaeological sample. The analysis from the modern
comparative sample shows that the annuli form during
the cold months of the year (November-February).
It must be realized that the conclusions concerning the

relationship between season of death and cementum
annuli are based on an extremely small sample of modern
comparative specimens. All of the animals in the modern
comparative collection are young, and therefore, as
discussed above, do not show their increments as they are
formed. Determination of timing for the formation of
layers is based on the single older individual in the
sample (4.5 years). Therefore, while the modern
comparative includes nineteen animals, the observations
regarding increment formation are based on a single
animal. This was due to the fact that the comparative
sample was collected in Manitoba where meat is the
dominant focus of production (and the animals are almost
always slaughtered before two years of age).
B. The archaeological sample

The determination of the number of increments and the
nature of the outer increment were performed both by the
one of the authors (E. Arnold) and by Dr. Ariane Burke.
Burke examined the slides in a blind test, so as the initial
readings by Arnold were not known, in an effort to
reduce bias. The results of both are reported here.
In the Early Iron Age highland site of Kadica Brdo, the
archaeological thin sectioning sample consisted of ten
teeth. Of these, the slides of four teeth were unreadable. Of
the six remaining teeth the number of increments counted
(by both readers) agrees with the age estimates obtained
from tooth wear and eruption. The nature of the final
increment was only determinable on four of the readable
slides (Kadica Brdo Samples #2, 4, 6 and 10 Figures
8.62-65). Each of the animals was slaughtered during what
is determined to be a growth zone. Kadica Brdo Sample #4
provided the best reading with the highest degree of
confidence from both readers and indicates the final zone
is a growth zone. This would indicate that it and the rest
were slaughtered during the warm months. None of the
Additional problems exist with regards to the application

of the modern comparative to the archaeological sample.
As the birth month in the central Balkan region is known to
be February, and the birth of the modern comparative
sample is known to be April/May, this implies a degree of
incompatibility between the comparative and the
archaeological sample. While the modern comparative can
be used to establish that the increments formed by sheep
and goat correspond well with absolute age, this is not the
116

main focus of our research with the archaeological sample.
It was hoped that the thin sectioning of the archaeological
teeth would provide seasonality information, since the
primary technique of tooth wear and eruption is limited in
its ability to establish this information. As the timing of the
birth of these animals is variable (February-May), it can
reasonably be assumed that the timing of the formation of
the increments will also vary. The ideal would be the
collection of a modern sample from the central Balkan
region a task for future research.
hypothesesnon-transhumant movement of pigs.

Furthermore, the harvest profiles of these animals indicate
their exploitation for primary products only and
demonstrate no change over the temporal period of this
investigation. As pigs have no secondary products for
which they can be exploited, their harvest profiles are not
complicated by issues of the Secondary Products
Revolution, as are the other domestic species.
For Ovis/Capra and Bos taurus, the data are more complex
and interpretation is not as simple. During the Neolithic
period, the exploitation of these domestic animals was
solely (primarily) for primary products. Additionally, the
harvest profiles indicate that the herds were resident yearround in the region, indicating the non-transhumant
movement of herds. In the Post Neolithic, the harvest
profiles provide limited evidence of the transhumant
movements of domestic herds. There is no evidence for
transhumance in the Eneolithic for either taxon, but it is
during the Early/Middle Bronze Age that we see the first
hints in the harvest profiles. For Ovis/Capra, only the
harvest profile from the lowland site of Vina suggests
some evidence of transhumance in the Early/Middle
Bronze Age. It is only during the Late Bronze Age that
there is any indication for transhumance of cattle (at the
site of Livade).
One may be tempted to see that the cementum analysis

yields complementary seasonality of culling in the
archaeological sample between the high and low altitude
sites. The readable slides from the lowland site of Vina
show an annulus as the final increment, while all the
readable slides from the highland site of Kadica Brdo show
a growth zone as the final increment. As such, while the
timing of the formation of these increments may be only
tentatively established based on the comparative collection,
there is a complementary pattern of seasonality between
the highland and lowland sites. Since some of the
observations (between Arnold and Burke) do not agree,
one cannot use this result with extreme confidence.
A final problem to consider, when attempting to draw
any conclusions from this limited archaeological thin
sectioning sample, is the period of the sites. The sample
from the lowland site of Vina is from the Late Neolithic.
As transhumance is not hypothesized to be occurring at
this time, the herd should be resident in this lowland site
for the entire year. One would expect to find seasonality
evidence for occupation year-round. Therefore, if one
accepts either researchers readings, the sample follows
expectations. The highland site of Kadica Brdo, if involved
in the transhumant movement of herds, would be expected
to show only a summer occupation, which it does. As such,
while the sample from Kadica Brdo appears to conform to
the hypotheses put forth in this investigation, it is severely
limited by the lack of complementary lowland sites from
the same or earlier Post Neolithic periods. Additionally,
since Kadica Brdo is an EIA site (and the highlands had
been colonized beginning with the Eneolithic), it is
possible that colonization of the highlands has increased
sufficiently for year-round settlement and herd
management in these areas by this time. However, if this
was indeed the case, one must query as to the lack of any
teeth showing a winter occupation. Overall, it would seem
that the data from Kadica Brdo indicates the absence of
Ovis/Capra herds from the site during the cold half of the
year, implying the presence of transhumance.
In the second stage, the tooth eruption and wear data are
reconfigured into another form of harvest profilesbar
graphs depicting the percentage of the assemblage
represented by each age group. These provide more
tangible evidence for transhumance through the increased
presence of specific age groups in the lowland and
highland regions coincident with the hypothesized
movement of transhumant herds of both cattle and
ovicaprines. The pig data again provide a control for nontranshumant movement of domestic herds. The random
patterning of the data between lowland and mid-altitude
sites in the Neolithic period implies no transhumant
movement of any domestic species in the Neolithic. While
pig continues to show this random pattern, complementary
patterns of seasonality of culling become evident in the
Post Neolithic sheep/goat and cattle data and provide
strong indications for the origins of transhumant
pastoralism at this time.
The third stage of analysis is a test of the validity of some
of the assumptions underlying the methods for ageing of
domestic animals with tooth eruption and wear. Most tooth
eruption and wear studies utilize 19th century age data for
domestic stock. Recent suggestions have indicated that
modern stock data may be more appropriate. As a result, a
modern sample of sheep and goat with known age and
season of death were collected from a modern industrial
slaughterhouse in Manitoba. The tooth eruption and wear
sequence data were evaluated against the classic studies of
Payne (1973) and Grant (1975) and found to be in close
correspondence. The linking of absolute ages with these
systems was also found to be valid. We conclude that it is
possible to utilize the established modern relationship
VI. Conclusions
In this chapter, the data undergoes four stages of analyses
in the search for evidence of transhumant pastoralism in
the central Balkans. In the first stage, the tooth eruption
and wear data from each site and time period were used to
create harvest profiles for each taxon. The pig harvest
profile data conforms exceedingly well to the proposed
117

between tooth eruption and wear and absolute age in sheep
and goats. This relationship can be applied to the
archaeological past.
death. In the archaeological sample, the seasonality data is

limited and at times contradictory. However, none of the
cementum results directly contradict any of the
expectations of a transhumant movement. The strongest
evidence for transhumance comes from the Early Iron Age
site of Kadica Brdo in the highlands of eastern Bosnia. All
of the tooth sections indicate the presence of Ovis/Capra at
the site only in warmer half of the year, in accordance with
our expectations of transhumant movement of domestic
herds.
The fourth stage of analysis examined cementum

increments in sheep and goat teeth, both modern and
archaeological. While the modern sample was limited in
size, it can be stated with confidence that both sheep and
goat develop increments in their dental cementum, which
correspond well with their absolute age and season of
118
CHAPTER 9
CONCLUSIONS
I. Introduction
The analysis was hampered by the absence of comparable

numbers of sites in each period from both highland and
lowland environments. The absence of highland sites
during the Neolithic is unavoidable since this
environment was not widely colonized or exploited
during this period. The data were more abundant for the
Post Neolithic. Also the distribution of highland sites
within the Post Neolithic was also uneven. In particular,
the lack of highland sites from the earliest periods of the
Post Neolithic, specifically the Eneolithic and Bronze
Age, was most detrimental to the analysis. Mid-altitude
sites were originally hypothesized to be an appropriate
substitute, but the data does not support this assumption.
It made the data more difficult to interpret at times. Only
one site (Kadica Brdo) was considered to be a true
highland site, and it derived from the latest time period
under consideration (Early Iron Age). As a result, this
problem made it more difficult to determine which
temporal moment hypothesis (e.g. early Post Neolithic or
advent of complex societies) was valid.
In the introductory chapter, three temporal moments were

hypothesized as the points in time when transhumant
pastoralism could be expected to appear in the central
Balkans. These hypotheses were tested with
zooarchaeological data from the region extending from
the Early Neolithic to the Early Iron Age. The
expectation was that if transhumant pastoralism was
present, then a complementary culling pattern for
domestic stock (sheep, goat and cattle) would be seen
between highland and lowland sites. The null hypothesis
stated that if transhumant pastoralism was not present,
there will be a random culling pattern between highland
and lowland sites.
In this chapter, a number of issues are discussed. First,
some methodological concerns are addressed. Second, the
results of the analysis are summarized. Third, the
implications for understanding the origins and evolution
of transhumant pastoralism in a temperate European
environment (i.e. central Balkans) are discussed.
Fourth, the evidence presented by the traditional harvest

profile data was limited by the coarseness of the
technique for coding tooth wear and eruption. The
problem with the method is that the age groups overlap
the months when the animals would be in a highland or a
lowland site. The technique doesnt allow the
identification of a 6 month from an 8 month individual,
the former of which would be expected to be pasturing in
the highland and the latter in the lowlands. The bar graph
data circumvented this problem to some extent, but by no
means resolves the issue. In reality, zooarchaeologists
must develop finer resolution techniques for ageing large
bulk samples of animal remains. Finer techniques, such
as cementum analysis, can do this, but are too expensive
and time consuming to apply to large samples. We are by
no means advocating the elimination of the tooth eruption
and wear technique as it continues to be broadly
employed due to its simplicity, low cost and widely
applicability to a variety of taxa. However, it is not
sensitive enough for more than a broadly based
discussion of transhumant pastoralism.
II. Methodological concerns

Several overriding methodological issues hampered this
research. Most cogent among them were sample size,
assemblage
taphonomy,
comparability
between
assemblages, and problems with coding for tooth wear
and eruption.
Sample size was a major limitation of the investigation.
Several sites had such small sample sizes that they could
not be included in the analysis at all. Of the remaining
sites, several had sample sizes for a particular taxon from
a single period that were too small. This did not allow for
the examination of all taxa from all sites in all periods
among the available data. It is hoped that additional data
will become available in the future. For this reason, we
provide the raw data from even the small samples. Future
researchers can incorporate these data into any new
analyses (Appendices B).
Taphonomic issues, notably the extent of sieving and
differential preservation of the various age classes under
consideration, were major complicating factors in not
only the creation of harvest profiles, but also the
interpretation of the data. This problem was particularly
relevant with respect to the frequencies in the youngest
age groups. This was the age group that was deemed most
essential for identifying the presence or absence of
transhumance and on which the hypotheses were
dependent.
III. The Secondary Products Revolution

The temporal origins of transhumant pastoralism have
been the primary focus of this research. However, one of
the temporal moment hypotheses proposed for the origins
of transhumance coincides in time with the Secondary
Products Revolution model (Sherratt 1981, 1983a). In the
Secondary Products Revolution model, domestic animals
(such as cattle, sheep and goats) were initially exploited
for their primary products (meat, bone, and hide). It was
only after several millennia that these animals began to be
119
CONCLUSIONS
exploited for their secondary products, such as milk, wool
or hide, and as agents of traction. It is clear from a
temporal comparison of data from the Near East and
Europe that the origins of secondary animal products
must be sought in the Near East. It is during the
Eneolithic of the central Balkans that it begins to spread
to southeastern Europe and from there to the rest of
Europe (Greenfield 1988, 1989, 1991, 2005; Sherratt
1981, 1983a).
proposed by transhumance from those that would result

from adoption of a secondary or mixed animal products
exploitation strategy (Greenfield 1999a, 2005a).
Nonetheless, it is possible to see differences in the
exploitation of age groups between highland and lowland
sites during the Post Neolithic and not the Neolithic,
supporting the hypothesis for the advent of transhumant
pastoralism during the later period (Greenfield 1988,
1989, 1999a).
The Eneolithic begins a period of major population

growth and expansion of territorial settlement that
continues through the Bronze and Iron Age. This is due
to the benefits of the adoption of secondary products
along with the appearance of the cart, plough, and
domestic horses during this period. The plough and the
use of animals as agents of traction allowed for the
intensification of agriculture with the cultivation of a
range of poorer quality soils. Additionally, the cart and
draught livestock (cattle and horses) enabled easier
transport of goods and reduced the difficulties
experienced previously in moving goods, people, and
livestock across the landscape. These developments
enabled not only an expansion of areas under cultivation,
but also expanded the range of settlement locations and
settlement types.
IV. Evidence for transhumant pastoralism

It is clear that there is no evidence for transhumant
pastoralism during the Neolithic of the central Balkans.
The evidence for the transhumant movement of domestic
herds at the Post Neolithic juncture is mixed. The results
from each taxon will be discussed in turn.
The Sus scrofa dom. harvest profiles (based on tooth
eruption and wear) conform to expectations for their nontranshumant
herding
strategy.
There
is
no
complementarity between harvest profiles between
highland and lowland sites, nor any perceptible change in
culling patterns over time. It can be concluded that
domestic pigs were not subjected to a transhumant herd
management strategy during any of the periods examined
in this study.
The vast previously uncolonised highlands of the region

were settled during this period. In these new areas, which
were probably considered agriculturally marginal in
earlier periods, new economic means were developed.
First and foremost in terms of the animal segment of the
economy, the development of new textiles made from
wool provided a valuable commodity for trade, where
there had previously been none. These areas were also
well suited to the larger herds that were now possible
with an increasing emphasis on the production of milk
and wool. The growth of herds enabled the development
of a pastoral sector of society (Sherratt 1981, 1983a).
While this model has been criticized (e.g. Chapman
1982), it has become widely accepted throughout the
literature in recent years (Greenfield 2005; Harding 2000;
Milisauskas 2000).
The results from the Ovis/Capra data (tooth eruption and

wear and cementum analysis) indicate that there was a
shift in their herd management strategy over time. The
harvest profiles from the Neolithic imply a year round
availability of the herds around sites, whether low or midaltitude. This suggests a non-transhumant movement and
continuous culling of the animals and residential stability
throughout the year. There is no complementarity
between mid-altitude and lowland areas in culling and, as
a result, transhumance is probably not occurring. It is
clear from the Ovis/Capra data that the Early Neolithic
temporal moment hypothesis is not supported.
In contrast, changes in harvest profiles are seen during
the Post Neolithic periods. The Ovis/Capra data from the
Middle Bronze Age site of Vina conforms to the
expectations of the hypotheses for a lowland site that
animals should be present in the lowlands during the
colder half of the year, at the very least. The harvest
profile from Vina shows an absence of the youngest age
groups (0-2 and 2-6 months). There is continuous and
rapid mortality of the age groups beginning with 6-12
months and continuing through to 8-10 years, with slight
changes along the way. As a lowland site, it would be
expected to have evidence of the 0-2 and 6-12 month,
with an absence of 2-6 month age groups, if
transhumance is occurring. While, the 0-2 month age
group is missing, this group is most susceptible to
taphonomic issues and differential recovery. The absence
of the 2-6 month class and presence of the 6-12 month
class at Vina are in accordance with the expectations for
a lowland site involved in the transhumant movement of
It is necessary to recognize the effects of the advent of

secondary product exploitation upon the harvest profiles.
As Greenfield (1988, 1989, 1991, 2005a) has
demonstrated, the advent of secondary products use
clearly begins during the Eneolithic and becomes widely
accepted during the Bronze Age. At the same time, there
is no evidence for specialized secondary product
exploitation in the region during any of the prehistoric
periods. Domestic sheep, goat, and cattle exploitation
strategies shift from a focus on primary products during
the Neolithic to a more diversified pattern, including both
primary and secondary products during the Post
Neolithic. The effect of the incorporation of secondary
animal products into subsistence regions is to distort the
harvest profiles, as they shift toward older animals. This
makes it somewhat difficult to distinguish the changes
120

herds. As a result, the zooarchaeological assemblage from
Middle Bronze Age Vina provides some suggestion for
the presence of transhumant pastoralism.
Therefore, herders based in mid-altitude would not be

forced to undertake the arduous task of moving most or
all of herds.
In contrast, at Early Bronze Age Novaka uprija, also a

lowland site, the 2-6 month age group is present. This
group is not expected to be there if the herds from the site
are being moved in a transhumant fashion to the highland
areas for the warmer months of the year. On the one
hand, the conflicting data from Vina and Novaka
uprija can be seen as not supporting the hypothesis
that the 2-6 month age group of domestic stock involved
in a transhumant economy would be absent from the
lowlands during the warm months of the year. On the
other, it can be interpreted to support the hypothesis, in
the following way. We have always maintained that the
environmental constraints that exist in the Mediterranean
(hot, arid) are not the same as in temperate (cool, moist)
climates, such as the central Balkans. In the
Mediterranean, Near East and similar semi-arid climatic
regions, domestic stock tend to be moved out of the
lowlands during the warmer half of the year in order to
avoid the devastating heat and aridity and to obtain
sufficient pasture (which will only exist in the highlands
during the summer). In temperate climates, in contrast,
sufficient microenvironments exist in the lowlands
throughout the year. Stock does not need to be moved out
of the lowlands in order to find graze. As a result, it
would not be surprising and in fact should be expected
that the harvest profiles and other seasonality data from
lowland zooarchaeological assemblages in a temperate
environment would present either no evidence or mixed
evidence for transhumance. It is only in the highlands of
the temperate zone that transhumance becomes more of
an ecological necessity and would be expected to appear
archaeologically. It is also difficult to determine from the
data if only a portion of the domestic herd is moved
between highland and lowland pasture (Greenfield 1988,
1991, 1999a, 2001b).
The ecological constraints on herders were expected to be

greatest in highland areas. The cementum analysis from
the highland site of Kadica Brdo demonstrates clear
evidence for only warm season culling of herds (even
though the harvest profile is equivocal). This is clearly in
accordance with expectations that the highland sites will
show more dramatic indications of transhumance than
found in lowland sites.
Therefore, it is not surprising that the evidence for
transhumance from the Early/Middle Bronze Age is
somewhat mixed. One site may be interpreted to be part
of a transhumant system, while the other may not. As
there is no evidence for transhumant pastoralism in the
Eneolithic periods, this data suggests that transhumance
begins to develop in the Early/Middle Bronze Age in a
patchy manner. A major problem with this period is that
there is an absence of more high altitude sites with
sufficient data with which to compare them.
Just as with pigs, sheep/goat, the profiles for Bos taurus
from the Neolithic periods implies a year round
availability of the herds. This, in turn, suggests a nontranshumant movement and continuous culling of the
animals and residential stability throughout the year.
There is no complementarity between highland and
lowland areas. Cattle transhumance is not occurring
during the Neolithic. This observation once again negates
the Early Neolithic temporal moment hypothesis as with
the Ovis/Capra data.
A shift in the harvest profiles of Bos taurus is seen in the
Post Neolithic, but later than with Ovis/Capra. The
harvest profile for Late Bronze Age Livade suggests the
transhumant movement of herds. The absence of the 1-8
month group is expected in the lowland site assuming a
transhumant movement of the domestic herds. The
presence of the next age group (8-18 months) also fits
with the expectations of a transhumant movement. As
such, it seems that the data imply that animals are being
moved in a transhumant fashion. Again, it is apparent that
the lowland sites show mixed evidence for transhumance,
as noted above.
Additionally, the bar graphs indicate that the divisions

between the tooth wear and eruption age groups in sheep
and goats is not fine enough to elucidate the temporal
aspects of a transhumant movement. With the youngest
individuals potentially culled from the herds in the
highland and mid-altitude pastures at 8 months, just
before the herds move back into the lowlands, these
groups appear in the harvest profiles in the 6-12 month
age group and subsequently do not support the original
hypotheses. Ideally, a seasonal transhumant movement of
domestic ovicaprine herds could be revealed, if there was
some means to divide the 6-12 month age class into two
smaller groups (i.e. 6-8 months and 9-12 months). At the
present time, there is no adequate justification for such a
split. The mid-altitude sites were originally hypothesized
to be adequate substitutes for highland sites. However,
the harvest profiles from these sites show patterns more
in line with lowland sites (i.e. not dramatic differences).
It may be that the ecological constraints found in midaltitude sites are also not as strong as in highland sites.
There is a problem with consideration of the Bos taurus

data from the EIA highland site of Kadica Brdo. There is
no clear evidence in the cattle harvest profile for
transhumance. This is in accordance with the sheep/goat
harvest profile from the site. If the data were taken at face
value, one can conclude that herds were resident yearround in the environs of the site. However, the cementum
analysis of sheep/goat remains from the site indicates
otherwise. It indicates much clearer support for the
transhumance hypothesis. Tooth cementum is a much
more sensitive indicator of seasonality than tooth eruption
and wear. This highlights the difficulty of elucidating
121
CONCLUSIONS
transhumance from solely harvest profile data.
Post Neolithic. The temporal hypothesis that specified that

transhumance appeared with the earliest domestic animals,
during the Early Neolithic, can be rejected. There is no
evidence to support this hypothesis in a temperate
environment. The construction of traditional harvest
profiles provides only tentative indications of transhumant
pastoralism. Stronger evidence for the seasonality of
movement occurs by reanalyzing the tooth wear and
eruption data into various formats and by combining these
results with cementum analysis. This indicates a
complementary seasonal movement between lowland and
highland sites beginning in the Post Neolithic.
Again, as with the sheep/goat data, the bar graph analysis

for cattle from Kadica Brdo is hampered by the division
of age groups at 8 months. Is an 8 month old individual
being coded for tooth wear and eruption coded into the
end of class B (1-8 months) or the beginning of class C
(8-18 months)? Further analysis of the tooth eruption and
wear data in a different format, and/or the application of
additional zooarchaeological analytical techniques (e.g.
cementum, isotopic analysis) are necessary. This explains
the reason for the incongruity between the harvest
profiles and cementum data at this site.
There is no evidence that all domestic stock were moved

en masse from the beginning of the Eneolithic. Early
transhumance movements probably included only
portions of herds. As transhumance develops and herd
sizes grew, larger pastoral populations began to move
across the landscape. Additionally, it would appear that
transhumance appears at slightly different times for
different taxa. The evidence for Ovis/Capra occurs
during the Early/Middle Bronze Age, while that for Bos
taurus only appears during the Late Bronze Age.
Obviously, the harvest profile data do not demonstrate

concrete evidence for the transhumant movement of
domestic herds. However, the lack of evidence with
harvest profiles is likely the result of the methodological
problems discussed above. The bar graph analysis of the
tooth wear and eruption data offers more support for the
identification of transhumant movement of domestic
herds. In the Post Neolithic bar graphs, there is an
increased presence of complementary age groups
between lowland and highland sites. This is in accordance
with and supports the hypothesis that the transhumant
herds of both cattle and ovicaprines began during the Post
Neolithic. The cementum analysis offers further support
for the appearance of transhumant movement of herds
during the Post Neolithic.
Our initial hypotheses expected transhumance early in the

Post Neolithic (during the Eneolithic or Early Bronze
Age) with the initial colonization of the agriculturally
marginal highlands. Due to the lack of appropriate
Eneolithic sites for analysis in this study, the data does
not allow us to evaluate this element of the hypothesis
i.e. that the advent of transhumance occurred during the
Eneolithic. But, transhumance is definitely present by the
beginning of the Bronze Age.
As a result, it can be conclusively stated that the Early

Neolithic hypothesis for the advent of transhumance in
temperate southeast Europe is not supported. There is
little evidence for the presence of transhumance during
the earliest part of the Post Neolithic (i.e. Eneolithic).
But, this may be reflection of the small database from this
period. It is impossible, however, to conclusively
determine (given the limitations in the current database)
when during the Bronze Age transhumant pastoralism
would have appeared. There are hints that it is present
already from the Early, Middle and Late Bronze Age
assemblages, implying that it appears at the beginning (or
during the Eneolithic) and evolves throughout the Bronze
Age. By the Early Iron Age, transhumant patterns of
movement are present and archaeologically visible.
Pastoralism, in general, and transhumant pastoralism,

specifically, is visible in the archaeological record. The
techniques utilized in this investigation offer a means for
identifying prehistoric transhumant movement through
space and time. With further refinement of these
techniques, the addition of other techniques (such as
stable isotope analysis), and through the collection of
larger and regionally representative zooarchaeological
samples, both modern and archaeological, it becomes
possible to further elucidate pastoral movements in
prehistory.
V. Conclusions
Strong evidence has been presented for the advent of
transhumant pastoralism in the central Balkans during the
122
REFERENCES CITED
Antonijevi, Dragoslav 1982 Obredi i Obicaji Balkanckih
Stoara. Beograd: Srpska Akademija Nauka i
Umetnosti.
Arnold, Elizabeth R. and Haskel J. Greenfield 2003 A
zooarchaeological perspective on the origins of
vertical transhumant pastoralism and the
colonization of marginal habitats in temperate
southeastern Europe. In Colonisation, Migration
and Marginal Areas: A Zooarchaeological
Approach, edited by Mariana Mondini, Sebastin
Muoz, and Stephen Winkler. Oxford: Oxbow
Press, pp. 96-117.
Bailey, D. W. 1994 Reading prehistoric figurines as
individuals. World Archaeology 25: 321-331.
Bailey, D. W. 2000 Balkan Prehistory: Exclusion,
Incorporation and Identity. New York: Routledge.
Baker, Frederik 1999 The ethnoarchaeology of
transhumance of southern Abruzzi of central Italy
an interdisciplinary approacy. In Transhumant
Pastoralism in Southeastern Europe: Recent
Perspectives from Archaeology, History and
Ethnology, edited by Lszl Bartosiewicz and
Haskel J. Greenfield. Budapest: Archaeolingua, pp.
99-110.
Bankoff, H. Arthur and Haskel J. Greenfield 1984
Decision-making and culture change in Yugoslav
Bronze Age. Balcanica 15: 7-31.
Barker, Graeme 1973 Cultural and economic change in
the prehistory of central Italy. In The Explanation of
Culture Change, edited by Renfrew, Colin, pp. 359370. London: Methuen.
Barker, Graeme 1975 Early Neolithic land use in
Yugoslavia. Proceedings of the Prehistoric Society
41: 85-104.
Barker, Graeme 1985 Prehistoric Farming in Europe.
Cambridge: Cambridge University Press.
Barker, Graeme 1991 Archaeological survey and
ethnoarchaeology in the Cicolano Mountains,
Central Italy. Preliminary results. In Archeologia
Della Pastorizia Nell'Europa Meridionale, edited by
R. Nisbet, R. Maggi, and G. Barker. Rivista di Studi
Liguri 56 (1-4), pp. 109-121.
Barth, F. 1961 Nomads of South Persia: The Basseri
Tribe of the Khamseh Confederacy. Illinois:
Waveland Press Inc.
Bartosiewicz, Lszl and Haskel J. Greenfield (eds.)
1999 Transhumant Pastoralism in Southern Europe:
Recent Perspectives from Archaeology, History and
Ethnology. Budapest: Archaeolingua.
Bates, Daniel 1973 Nomads and Farmers: A Study of the
Yoruk of Southeastern Turkey. Papers of the
Museum of Anthropology University of Michigan,
Ann Arbor.
Bates, Daniel and Lees, S. 1977 The role of exchange in
productive specialization. American Anthropologist
79: 824-841.
Binder, D. 2000 Mesolithic and Neolithic interaction in

southern France and northern Italy: new data and
current hypotheses. In Europes First Farmers,
edited by T. Douglas Price. Cambridge: Cambridge
University Press, pp. 117-143.
Boessneck, J. 1962 Die Tierreste aus der Argissa-Magula
vom prakeramischen Neolithikum bis zur mittleren
Bronzezeit. In Die deutsche ausgrabungen auf der
Argissa-Magula in Thesselien, edited by V.
Mikoji, J. Boesseneck and M. Hopf. Bonn:
Habelt, pp. 27-99.
Boessneck, J. 1969 Osteological differences between
sheep (Ovis aries Linne) and goats (Capra hircus
Linne). In Science in Archaeology, edited by D.
Brothwell and E. Higgs. London: Thames and
Hudson, pp. 331-359.
Bknyi, S. 1974a The History of Domestic Mammals in
Central and Eastern Europe. Budapest: Akademiai
Kiado.
Bknyi, S. 1974b The Vertebrate Fauna. In Obre I and
II: Neolithic sites in Bosnia, edited by M. Gimbutas.
Wissenschaftliche Mitteilungen des BosnischHerzegowinischen Landesmuseum (Arheologie) III,
part B. Sarajevo, pp. 55-154.
Bonte, P. 1981 Ecological and economic factors in the
determination of pastoral specialization. Journal of
Asian and African Studies 16 (1-2): 33-49.
Bosch-Gimpera, P. 1944 El poblamiento antiguo y la
formacion de los pueblos de Espaa. Mexico.
Brice, William C. 1974 Nomadism in Thrace its nature
and origins. Thracia (Sofia) 2: 99-101.
Brukner, Bogdan, Borislav Jovanovi, and Nikola Tasi
1974 Praistorija Vojvodine. Belgrade: Savez
Arheolokih Drutva.
Bukvi, Ljubomir 1979 Results of the researches of the
mound near Jabuka - A contribution to the study of
the culture of graves under Tumuli. Archaeological
Iugoslavica 1914-18.
Bull, G. and S. Payne 1982 Tooth eruption and
epiphyseal fusion in pigs and wild boar. In Ageing
and Sexing Animal Bones from Archaeological
Sites, edited by R. Wilson, C. Grigson and S. Payne.
British Archaeological Reports, British Series 109,
pp. 55-72. Oxford: BAR.
Bullock, D. and J. Rackham 1982 Epiphyseal fusion and
tooth eruption of feral goats from Moffatdate,
Dumfries and Galloway, Scotland. In Ageing and
Sexing Animal Bones from Archaeological Sites
edited by R. Wilson, C. Grigson and S. Payne.
British Archaeological Reports, British Series 109.
Oxford: BAR, pp. 73-80.
Burke, A. 1995 Prey Movements and Settlement Patterns
during the Upper Paleolithic in Southwestern
France.
British
Archaeological
Reports,
International Series 619. Oxford: BAR.
Burke, A. and J. Castenet 1995 Histological observations
of cementum growth in horse teeth and their
123
REFERENCES CITED
application
to
Archaeology.
Journal
of
Archaeological Science 22: 479-493.
Butzer, K. W. 1971 Environment and Archeology: An
Ecological Approach to Prehistory. 2nd ed.
Chicago: Aldine-Atherton.
Cauver, J. 2000 The Birth of the Gods and the Origins of
Agriculture. Cambridge: Cambridge University
Press.
Chalkea, E. 1999 Ecological adaptations of the
Sarakatsani
(Epirus-Zagori-Greece).
In
Transhumant Pastoralism in Southeastern Europe:
181-188.
Champion, T. and C. Gamble, S. Shennan and A. Whittle
1984 Prehistoric Europe. New York: Academic
Press.
Chang, Claudia 1993 Pastoral transhumance in the
southern Balkans as a social ideology:
ethnoarchaeological research in northern Greece.
American Anthropologist 95: 687-703.
Chang, Claudia 1999 The ethnoarchaeology of pastoral
sites in the Grevena region of northern Greece. In
133-144.
Chang, Claudia and Hal A. Koster 1986 Beyond bones:
toward an archaeology of pastoralism. In Advances
in Archaeological Method and Theory, vol. 9, edited
by M. L. Schiffer. New York: Academic Press, pp.
97-148.
Chang, Claudia and Hal A. Koster (eds.) 1994
Pastoralists at the Periphery. Tuscon: The
University of Arizona Press.
Chang, Claudia and Perry A. Tourtellotte 1993
Ethnoarchaeological
survey
of
pastoral
transhumance sites in the Grevena region, Greece.
Journal of Field Archaeology 20: 249-264.
Chapman, J. C. 1981 The Vina Culture of South-East
Europe: Studies in Chronology, Economy and
Society.
British
Archaeological
Reports,
International Series 117 (Part 1 and 2). Oxford:
BAR.
Chapman, J. C. 1982 The Secondary Products
Revolution and the limitations of the Neolithic.
Bulletin of the Institute of Archaeology 19: 107-122.
Chapman, J. C. and J. Mller 1990 Early farmers in the
Mediterranean basin: the Dalmatian evidence.
Antiquity 64: 127-134.
Chapman, Robert 1979 Transhumance and Megalithic
tombs in Iberia. Antiquity 53: 15-152.
Cherry, John 1988 Pastoralism and the role of animals in
the pre- and protohistoric economies of the Aegean.
In Pastoral Economies in Classical Antiquity, edited
by C. R. Whittaker. The Cambridge Philological
Society, Supplementary Volume no. 14, pp. 6-34.
Cambridge: The University Press.
Clark, Gillian 1991 The contribution of faunal analysis to

the study of prehistoric and historical pastoralism in
Italy. In Archeologia Della Pastorizia Nell'Europa
Meridionale, edited by R. Nisbet, R. Maggi, and G.
Barker. Rivista di Studi Liguri 56 (1-4), pp. 73-80.
Crabtree, P. J. 1982 Patterns of Anglo-Saxon Animal
Economy: An Analysis of the Animal Bone Remains
from the Early Saxon Site of West Stow, Suffolk. Ph.
D. Thesis, University of Pennsylvania.
Creighton, Oliver H. and Joan R. Segui 1998 The
ethnoarchaeology of abandonment and postabandonment behaviour in pastoral sites: evidence
from Famorca, Alacant province, Spain. Journal of
Mediterranean Archaeology 11: 31-52.
Cribb, R. 1990 Nomads in Archaeology. Cambridge: The
University Press.
Cviji, J. 1918. La Peninsule Balkanique: geographie
humaine. Paris: A. Colin.
Dahl, G. and A. Hjort 1976 Having Herds: Pastoral Herd
Growth and Household Economy. Stockholm
Studies in Social Anthropology Department of
Social Anthropology, University of Stockholm,
Stockholm.
Danta, D. and D. Hall 1995 Contemporary Balkan
questions: the geographic and historic context. In
Reconstructing the Balkans: A Geography of
Southeast Europe, edited by D. Danta and D. Hall.
New York: John Wiley & Sons Ltd, pp. 15-32.
Davis, S. 1987 The Archaeology of Animals. New Haven:
Yale University Press.
Dedijer, J. 1916 La transhumance dans les pays
dinariques. Annales de Gographie 25 (137): 347365.
Dhavalikar, R. 1989 Farming to Pastoralism: Effects of
Climatic Change in the Deccan. In The Walking
Larder. Patterns of Domestication, Pastoralism and
Predation, edited by J. Clutton-Brock. London:
Unwin Hyman, pp. 156-168.
Dracun, Vladimir 1954 Neke vanije rase ovaca, njihov
istoriski razvitak njihova budunost u svetlu
dananjih ekonomskih odnosa (Some more
important sheep breeds, their historical development
and future in the light of present-day economic
relations). Veterinaria (Sarajevo) 3: 248-263.
Dumitrescu, V. 1983 The prehistory of Romania from the
earliest times to 1000 BC. Cambridge Ancient
History. Cambridge: The University Press, vol. 3,
part 1 (2nd edition), pp. 1-74.
Edwards, Ceiridwen J., David E. MacHugh, Keith M.
Dobney, Louise Martin, Nerissa Russell, Liora K.
Horwitz, Susan K. McIntosh, Kevin C. MacDonald,
Daniel Helmer, Anne Tresset, Jean-Denis Vigne,
and Daniel G. Bradley 2004 Ancient DNA analysis
of 101 cattle remains: limits and prospects. Journal
of Archaeological Science 31: 695-710.
Ehlers, E. and H. Kreutzmann 2000 High mountain
ecology and economy: potential and constraints. In
High Mountain Pastoralism in Northern Pakistan
edited by E. Ehlers and H. Kreutzmann. Stuttgart: F.
Steiner, pp. 9-36.
124

Getty, R. 1975 Sisson and Grossmans The Anatomy of
Domestic Animals. Philadelphia: Saunders.
Gimbutas, Marija 1965 Bronze Age Cultures in Central
and Eastern Europe. The Hague: Mouton.
Gimbutas, Marija 1973 The beginning of the Bronze Age
in Europe and the Indo-Europeans: 3500-2500 B.C.
The Journal of Indo-European Studies 1 (2): 163214.
Gimbutas, Marija, Shan Winn and Daniel Shimabuku
1989 Achilleion: A Neolithic Settlement in Thessaly,
Greece, 6400-5600 BC. Monumenta Archaeologica
14. Los Angeles: University of California.
Glatzer, B. 1982 Processes of nomadization in West
Afghanistan. Contemporary Nomadic and Pastoral
Peoples: Asia and the North. Studies in Third World
Societies 18: 61-86.
Gordon, B. C. 1993 Archaeological tooth and bone
seasonal increments: the need for standardized terms
and techniques. Archaeozoologia 5 (2): 9-16.
Gottmann, J. 1969 A Geography of Europe, 4th edition.
New York: Holt, Rinehart and Winston Inc.
Govedarica, Blagoje 2004 Zeptertrger: Herrscher der
Steppen. Die frhen Ockergrber des lteren
neolithikums im karpatenbalkanischen Gebiet und
in Steppenraum Sdost- und Osteuropas.
Balkankommission der Heidelberger Akademie der
Wissenschaften 6.Mainz: Philipp von Zabern.
Grant, A. 1975 The use of tooth wear as a guide to the
age of domestic animals. In Excavations at
Portchester Castle, edited by Barry Cunliffe, vol. 2.
London: Society of Antiquaries, pp. 245-279.
Grant, A. 1982 The use of uooth wear as a guide to the
age of domestic ungulates. In Ageing and Sexing
Animal Bones from Archaeological Sites, edited by
R. Wilson, C. Grigson and S. Payne. British
Archaeological Reports, British Series 109. Oxford:
BAR, pp. 91-108.
Grant, A. 1991 Identifying and understanding
pastoralism
and
transhumance:
an
archaeozoological approach. In Archeologia Della
Pastorizia Nell'Europa Meridionale, edited by R.
Nisbet, R. Maggi, and G. Barker. Rivista di Studi
Liguri 56 (1-4), pp. 13-20.
Grassl, Herbert 1985 Hirtenkultur in Griechenland.
Verffentlichungen des Verbandes sterreichischer
Geschichtsvereine 25: 77-85.
Greenfield, Haskel J. 1984 A model of faunal
exploitation for the Bronze Age of the central
Balkans. MASCA Journal 3 (2): 53-55.
Greenfield, Haskel J. 1986a The Paleoeconomy of the
Central Balkans (Serbia). A Zooarchaeological
Perspective on the Late Neolithic and Bronze Age
(ca. 4500-1000 B.C). British Archaeological
Reports, International Series 304. Oxford: BAR.
Greenfield, Haskel J. 1986b Summary report on the
vertebrate fauna from Novaka uprija, Serbia
(Eneolithic-Late Bronze Age). Zbornik Radova
Narodnog Muzeja (Belgrade, Yugoslavia; National
Museum) 12: 63-74.
Ehrich, R. 1965 Geographical and chronological patterns

in east-central Europe. In Chronologies in Old
World Archaeology, edited by R. Ehrich. Chicago:
University of Chicago Press, pp. 403-458.
Ehrich, Robert W. and H. Arthur Bankoff 1990 Absolute
chronology of the Neolithic in southeastern Europe.
In Chronologies in Old World Archaeology, edited
by Ehrich, R., 3rd edition. Chicago: University of
Chicago Press, pp. 375-394.
Ekvall, R. B. 1968 Fields on the Hoof: Nexus of Tibetan
Nomadic Pastoralism. New York: Holt, Rinehart
and Winston.
Evans-Pritchard, E. E. 1940 The Nuer: A Description of
Modes of Livelihood and Political Institutions of a
Nilotic People. Oxford: Clarendon Press.
Ewbanks, J. M. 1964 Sheep in the Iron Age: a method of
study. Proceedings of the Prehistoric Society 30:
423-426.
Flannery, K. V. 1965 The ecology of early food
production in Mesopotamia. Science 147 (3663):
1247-1256.
Fleming, A. 1971 Territorial patterns in Bronze Age
Wessex. Proceedings of the Prehistoric Society 27:
138-66.
Fotiades, M. 1980 Transhumance: was it indeed practiced
in the prehistoric Mediterranean? American Journal
of Archaeology 84 (2): 207.
Frenzel, B. 1966 Climatic change in the Atlantic/subBoreal transition. In World Climate 8000-0 BC,
edited by J. S. Sawyer. London: Royal
Meterological Society, pp. 99-123.
Furlan, D. 1977 The climate of Southeast Europe. In
Climates of Central and Southern Europe, edited by
C. C. Wallen. Amsterdam: World Survey of
Climatology, (Elsevier Scientific) 6: 185-235.
Gal, Lszlo and Pter Gunst 1972 Livestock husbandry
in Hungary from 1848 to World War I.
Agraitrtneti Szemle 14 (Supplement): 7-48.
Galaty, J. C. and D. L. Johnson 1990 Introduction
pastoral systems in global perspective. In The World
of Pastoralism: Herding Systems in Comparative
Perspective, edited by John G. Galaty, Douglas L.
Johnson. New York: Guilford Press, pp. 1-67.
Garaanin, Milutin 1972 The Bronze Age of Serbia.
Belgrade: National Museum.
Garaanin, Milutin 1973 Praistorija na Tlu SR Srbije, 2nd
edition. Belgrade: Srbska Knjuevna Zadruga.
Garaanin, Milutin 1983 The Stone Age in the Central
Balkans; The Bronze Age in the Central Balkans;
The Early Iron Age in the Central and East Balkan
Area, c. 1000-750 BC. Cambridge Ancient History
Cambridge: The University Press, Vol. 3, part 1, pp.
75-135; 163-186; 582-618.
Geddes, D. S. 1983 Neolithic transhumance in the
Mediterranean Pyrenees. World Archaeology 15:
52-66.
Geddes, D. S. 1985 Mesolithic domestic sheep in West
Mediterranean Europe. Journal of Archaeological
Science 12: 25-48.
125
REFERENCES CITED
Greenfield, Haskel J. 1988 The origins of milk and wool
production in the Old World. Current Anthropology
29 (4): 573-748.
Greenfield, Haskel J. 1989 Zooarchaeology and aspects
of the Secondary Products Revolution: A Central
Balkan perspective. Zooarchaeologia 3 (1-2): 191200.
Greenfield, Haskel J. 1991 Fauna from the Late Neolithic
of the Central Balkans: issues in subsistence and
land use. Journal of Field Archaeology 18: 161-186.
Greenfield, Haskel J. 1993 Zooarchaeology, taphonomy, and
the origins of food production in the Central Balkans.
In Culture and environment: A fragile co-existence.
Proceedings of the 24th Chacmool Conference, edited
by Ross W. Jamieson, Sylvia Abonyi and Neil A.
Mirau. Archaeological Association, University of
Calgary, Alberta, pp. 111-117
Greenfield, Haskel J. 1994 Faunal remains from the Early
Neolithic Starevo Settlement at Bukovaka esma,
Starinar: Journal of the Archaeological Institute
(Belgrade, Yugo.) 43-44, pp. 103-114.
Greenfield, Haskel J. 1996 Sarina Medja, Vrbica, and
Veina Mala: the zooarchaeology of three Late
Bronze Age/Early Iron Age Transition localities
near Jagodina, Serbia. Starinar: Journal of the
Archaeological Institute 45-46:133-142.
Greenfield, Haskel J. 1999a The advent of transhumant
pastoralism in temperate Southeast Europe: a
zooarchaeological perspective from the Central
Balkans. In Transhumant Pastoralism in
Southeastern Europe: Recent Perspectives from
Archaeology, History and Ethnology, edited by
Lszl Bartosiewicz and Haskel J. Greenfield.
Budapest: Archaeolingua, pp 15-36.
Greenfield, Haskel J. 1999b The origins of metallurgy:
distinguishing stone from metal cut marks on bones
from archaeological sites. Journal of Archaeological
Science 26 (7): 797-808.
Greenfield, Haskel J. 2000 The application of integrated
surface and subsurface reconnaissance techniques to
later prehistoric SE European sites: Blagotin Serbia.
Geoarchaeology 15: 167-201.
Greenfield, Haskel J. 2001a The Early Bronze Age of
Central Europe. In The Encyclopedia of Prehistory:
Human Area Relations Files, edited by Peter
Peregrine
and
Melvin
Ember.
Kluwer
Academic/Plenum Publishers, pp. 124-131.
Greenfield, Haskel J. 2001b Transhumant pastoralism
and the colonization of the highlands in temperate
southeastern Europe. In Untrampled ground untrammelled views: Human exploitation of and
settlement patterns on new landscapes (Proceedings
of the 31st (1998) Annual Chacmool Conference),
edited by Susan Tupakka, Jason Gillespie, and
Christy de Mille. University of Calgary, Department
of Archaeology, Alberta, pp. 471-492.
Greenfield, Haskel J. 2005a A reconsideration of the
secondary products revolution: 20 years of research
in the central Balkans. In The Zooarachaeology of
Milk and Fats (Proceedings of the 9th ICAZ
Conference, Durham 2002), edited by Jacqui

Mulville and Alan Outram. Oxford: Oxbow Press,
pp. 14-31.
Greenfield, Haskel J. 2005b The fauna and systematic
recovery systems from the Iron Age excavations at
Kadica Brdo, Yugoslavia. Godinjak (Academy of
Sciences and Arts, Center for Balkan Research,
Sarajevo) XXXIV/32: 107-150.
Greenfield, Haskel J. 2006 The social and economic
context for domestic horse origins in southeastern
Europe: a view from Ljuljaci in the central Balkans.
In Horses and Humans: The Evolution of EquineHuman Relationships, edited by Sandra L. Olsen,
Susan Grant, Alice M. Choyke, and Lszl
Bartosiewicz. British Archaeological Reports,
International Series. Oxford: Archaeopress.
Greenfield, Haskel J. in press a Subsistence systems
during the Early Neolithic of the Central Balkans: a
faunal taphonomic perspective. In The Iron Gates in
Prehistory, edited by Clive Bonsall, Vasile
Boronean and Ivana Radovanovi. Oxford:
Archaeopress.
Greenfield, Haskel J. in press b Re-analysis of the
vertebrate fauna from Hajduka Vodenica in the
Danubian Iron Gates of Yugoslavia: subsistence and
taphonomy from the Early Neolithic and Mesolithic.
In The Iron Gates in Prehistory, edited by Clive
Bonsall, Vasile Boronean and Ivana Radovanovi.
Oxford: Archaeopress.
Greenfield, Haske J.n.d. a The vertebrate fauna from
Vina: 1982 excavations of Late Neolithic,
Eneolithic and Middle Bronze Age deposits.
Manuscript on file at the University of Manitoba.
Greenfield, Haskel J. n.d. b The vertebrate faunal remains
from the Middle Neolithic site at Stragari,
Yugoslavia. Manuscript on file at the University of
Manitoba.
Greenfield, Haskel J. n.d. c The Paleoeconomy of
Prehistoric Petnica, Yugoslavia: Fauna and Flora
(1980-1986). Manuscript on file at the University of
Manitoba.
Greenfield, Haskel J. n.d. d Zooarchaeological remains
from Foeni-Sala: an Early Neolithic Starevo-Cri
Settlement in southwestern Romania. Manuscript on
file at the University of Manitoba.
Greenfield, Haskel J. and Florin Draovean 1994 An
Early Neolithic Starevo-Cri settlement in the
Romanian Banat: preliminary report on the 1992
excavations at Foeni-Sala. Annale Banatului:
Journal of the Museum of the Banat 3: 45-85
Greenfield, Haskel J. and Kent D. Fowler 2003 Megalo
Nisi Galanis and the Secondary Products Revolution
in Macedonia. Zooarchaeology in Greece: Recent
Advances. British School at Athens 9: 142-151.
Greenfield, Haskel J. and Kent D. Fowler 2005 The
Secondary Products Revolution in Macedonia: The
Faunal Remains from Megalo Nisi Galanis, A Late
Neolithic-Early Bronze Age Site in Greek Macedonia.
British Archaeological Reports, International Series
no. 1414. Oxford: John and Erica Hedges Ltd..
126

Greenfield, Haskel and Tina L. Jongsma in press a
Sedentary pastoral gatherers in the Early Neolithic
architectural, botanical, and zoological evidence for
mobile economies from Foeni-Salas, SW Romania.
In Unsettling the Past, edited by Douglas Bailey and
Alisdair Whittle. UCLA, Institute of Archaeology.
Greenfield, Haskel and Tina L. Jongsma in press b Early
Neolithic Blagotin: a summary of recent research. In
Unsettling the Past, edited by Douglas Bailey and
Alisdair Whittle. UCLA, Institute of Archaeology.
Greenfield, Haskel J. and Tina L. Jongsma in press c The
spatial organization of Early Neolithic settlements in
temperate southeastern Europe: a view from
Blagotin, Serbia. In An Odyssey of Space:
Proceedings of the 34th (2001) Chacmool
conference, Calgary, Alberta, edited by Christine
Cluney. November 2001. Submitted January 1, 2002
and accepted for publication.
Greenfield, Haskel J. and Svetozar Stankovi n.d
Preliminary report of the excavations of Blagotin
1989-1992. Unpublished manuscript.
Grigson, C. 1982 Sex and age determination of some
bones and teeth of domestic cattle: review of the
literature. In Ageing and Sexing Animal Bones from
Archaeological Sites, edited by Wilson, Grigson and
Payne. British Archaeological Reports, British
Series 109, pp. 7-23. Oxford: BAR.
Grubi, A. (ed.) 1980 Yugoslavia: An outline of geology
of Yugoslavia, Excursions 201a-202c. Paris: 26th
International Geological Congress and Belgrade:
National Committee for Mineral Resources of
Yugoslavia.
Grue, H. and B. Jensen 1979 Review of the formation of
incremental lines in tooth cementum of terrestrial
animals. Danish Review of Game Biology 10 (3): 348
Grsan-Salzman, Aye 1984 Shepherds of Transylvania.
Natural History 1984 (7): 42-53.
Gyni, M. 1951 La transhumance des Vlaques
Balkaniques au Moyen Age. Byzantinoslavica
(Prague) 12: 29-42.
Habermehl, K. H. 1961 Die Alterbestimmung bie
Haustieren, Pelztieren und beim jagdbaren
Wildtieren. Berlin and Hamburg: Paul Parer.
Hapern, J. M. 1967 A Serbian Village. New York:
Columbia University Press, 2nd edition.
Halpern, J. M. 1999 The ecological transformation of a
resettled area, pig herders to settled farmers in
Central Serbia (umadija, Yugoslavia) during the
19th and 20th centuries. In Transhumant Pastoralism
in Southeastern Europe: Recent Perspectives from
Archaeology, History and Ethnology, edited by
Lszl Bartosiewicz and Haskel J. Greenfield.
Budapest: Archaeolingua, pp.79-98.
Halpern, J. M. and Barbara Kerewsky-Halpern 1986 A
Serbian village in historical perspective. Prospect
Heights: Waveland Press.
Halstead, P. 1981 Counting sheep in Neolithic and
Bronze Age Greece. In Patterns of the Past: Studies
in Honour of David Clarke, edited by I. Hodder, G.
Isaac and N. Hammond. Cambridge: Cambridge

University Press, pp. 307-339.
Halstead, P. 1985 A study of mandibular teeth from
Romano-British contexts at Maxey. In The Fenland
Project: archaeology and environment in the Lower
Welland Valley, volume 1, edited by Francis Pryor,
Charles French, David Crowther, David Gurney,
Gavin Simpson and Maisie Taylor. East Anglican
Archaeology Report No. 27. The Fenland Project
Committee,
Cambridgeshire
Archaeological
Committee, pp. 219-224.
Halstead, P. 1987 Traditional and ancient rural economy
in Mediterranean Europe: plus a change? Journal
of Hellenic Studies 107: 77-87.
Halstead, P. 1991 Present to past in the Pindhos:
diversification and specialization in mountain
economies. In Archeologia Della Pastorizia
Nell'Europa Meridionale, edited by R. Nisbet, R.
Maggi, and G. Barker. Rivista di Studi Liguri 56 (14), pp. 61-80.
Halstead, P. 1996 Pastoralism or household herding?
Problems of scale and specialization in early Greek
animal husbandry. World Archaeology 28: 20-42.
Halstead, P., P. Collins and V. Isaakidou 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
Hambleton, E. 1999 Animal Husbandry Regimes in Iron
Age Britain: A Comparative Study of Faunal
Assemblages from British Iron Age Sites. British
Archaeological Reports, British Series 282. Oxford:
BAR.
Harding, A. F. 2000 European Societies in the Bronze
Age. Cambridge: The University Press.
Hesse, B. 1982 Slaughter patterns and domestication: the
beginnings of pastoralism in western Iran. Man 17:
403-417.
Hesse, Brian and Paula Wapnish 1986 Animal Bone
Archaeology: From Objectives to Analysis.
Washington, D.C: Taraxacum Press.
Higgs, E. S. 1976 The history of European agriculture:
the uplands. Philosphical Transactions of the Royal
Society, London, series B, CCLXXV: 159-173.
Higgs, E. S., C. Vita-Finzi, R. Harris, and A. E. Fagg
1967. The climate, environment and industries of
Stone Age Greece. Part III. Proceedings of the
Prehistoric Society 33: 1-29.
Higham, C. F. W. 1967 Appendix: stock rearing as a
cultural factor in Prehistoric Europe. Proceedings of
the Prehistoric Society 33: 84-106.
Hillson, Simon 1986 Teeth. Cambridge: The University
Press.
Hodkinson, S. 1988 Animal husbandry in the Greek
polis. In Pastoral Economies in Classical Antiquity,
edited by C. R. Whittaker. The Cambridge
Philological Society, Supplementary Volume no. 14,
pp. 35-74.
Hole, Frank 1978 Pastoral Nomadism in Western Iran. In
Explorations in Ethnoarchaeology, edited by Gould,
127
REFERENCES CITED
Richard, pp. 127-168. Albuquerque, NM: University
of New Mexico Press.
Hole, Frank, Flannery, Kent V., and Neely, J. A. 1969
Prehistory and Human Ecology of the Deh Luran
Plain: An Early Village Sequence from Khuzistan,
Iran. Memoirs no. 1, Museum of Anthropology,
University of Michigan, Ann Arbor, MI.
Horvath, Ferenc 1989 A survey on the development of
Neolithic settlement pattern and house types in the
Tisza region. Neolithic of Southeastern Europe and
its Near Eastern Connections. Varia Archaeologica
Hungarica 2. Budapest: Institute of Archaeology,
Academy of Sciences, pp. 85-103.
Irons, William G. 1975 The Yomut Turkmen of Iran.
Anthropological Papers 58, University of Michigan,
Ann Arbor.
Jacobsen, T. W. 1984 Seasonal pastoralism in southern
Greece: a consideration of the ecology of Neolithic
Urfinirnis pottery. In Pots and Potters: Current
Approaches in Ceramic Archaeology, edited by
Prudence M. Rice. Institute of Archaeology,
University of California, Los Angeles. Monograph
24, pp. 27-42.
Jochim, Michael 1976. Hunter-Gatherer, Subsistence and
Settlement A Predictive Model. New York;
Academic Press.
Jongsma, T. L. 1997 Distinguishing Pits from Pit Houses
through Daub Analysis: the Nature and Location of
Early Neolithic Starevo-Cri Houses at FoeniSala, Romania. M.A. Thesis, University of
Manitoba.
Jongsma, Tina and Haskel J. Greenfield 2001
Architectural technology and the spread of early
agricultural societies in temperate southeastern
Europe In Untrampled ground - untrammelled
views: Human exploitation of and settlement
patterns on new landscapes. Proceedings of the 31st
(1998) Annual Chacmool Conference, edited by
Susan Tupakka, Jason Gillespie, and Christy de
Mille. University of Calgary, Department of
Archaeology, Alberta, pp. 181-200.
Kaiser, T. 1984 Vina Ceramics: Economic and
Technological Aspects Of Late Neolithic Pottery
Production in Southeast Europe. Ph.D. Dissertation,
University of California, Berkley.
Kaiser, Timothy and Barbara Voytek 1983 Sedentism and
economic change in the Balkan Neolithic. Journal
of Anthropological Archaeology 2: 323-353.
Kalicz, N. 1970 Clay Gods: The Neolithic Period and the
Copper Age in Hungary. Budapest: Athenaeum
Printing House.
Khazanov, A. M. 1984 Nomads and the Outside World.
Cambridge: Cambridge University Press.
Klein, R. G. and K. Cruz-Uribe 1984 The Analysis of
Animal Bones from Archaeological Sites. Chicago:
University of Chicago Press.
Klevezal, G. A. 1996 Recording Structures of Mammals.
Determination of Age and Reconstruction of Life
History, Academy of Sciences USSR. Translated by
M. V. Mina and A. V. Oreshkin. Rotterdam: A. A.

Balkema.
Klevezal, G. A. and S. E. Kleinenberg 1967 Age
Determination of Mammals from Annual Layers in
Teeth and Bones. Academy of Sciences USSR.
Translated 1969 Department of the Interior and
National Science Foundation, US Department of
Commerce, Clearinghouse for Federal Scientific and
Technical Information, Springfield, IL.
Kordos, L. 1978a Changes in the Holocene climate of
Hungary reflected by the Vole-Thermometer
Method. Kulonnyomat A Foldrajzi Kozlemenyek (13): 222-229.
Kordos, L. 1978b A sketch of the vertebrate
biostratigraphy of the Hungarian Holocene.
Kulonnyomat A Foldrajzi Kozlemenyek (1-3): 144160.
Koster, H. J. and J. B. Koster 1976 Competition of
symbiosis? pastoral adaptive strategies in the
Southern Argolid, Greece. In Regional Variation in
Modern Greece and Cyprus, edited by M. Dimen
and E. Friedl. Annals of the New York Academy of
Sciences 268: 275-285.
Kotsakis, Kostas 1999 What tells can tell: social space
and settlement in the Greek Neolithic, In Neolithic
Society in Greece. Sheffield Studies in Aegean
Archeology 2, edited by Paul Halstead. Sheffield:
Sheffield Academic Press, pp. 66-76.
Krader, L. 1959 The ecology of nomadic pastoralism.
International Social Science Journal 11 (4): 499510.
Lamb, H. H. 1977 Climatic History and the Future.
Princeton: Princeton University Press.
Lees, S. H. and D. G. Bates 1974 The origins of
specialized nomadic pastoralism: a systemic model.
American Antiquity 39 (2): 187-193.
Legge, A. J. 1990 Animals, economy and environment.
In Selevac: A Neolithic Village in Yugoslavia, edited
by Ruth Tringham and Duan Krsti. Los Angeles:
UCLA Institute of Archaeology, pp. 215-242.
Legge, A. J. and E. Dorrington 1985 Harlow Temple: the
animal bones. In The Romano-British Temple at
Harlow, Essex, edited by E. France and M. Gobel.
Harlow: West Essex Archaeological Group, pp.
122-133.
Lekovi, Vladimir 1991 The Vinanization of the
Starevo Culture. In Vina and Its World:
International Symposium on The Danubian Region
from 6000 to 3000 BC, edited by D. Srejovi.
Belgrade: Serbian Academy of Science and Arts,
pp. 67-73.
Lewthwaite, J. 1981 Plains tales from the hills:
transhumance in Mediterranean Archaeology. In
Economic Archaeology: Towards an Integration of
Ecological and Social Approaches, edited by A.
Sheridan and G. Bailey. British Archaeological
Reports, International Series 96 (Oxford), pp. 57-66.
Lewthwaite, J. 1984 Pasture, Padrone: the social
dimensions of pastoralism in Prenuragic Sardinia, In
The Denya Conference of Prehistory, edited by
128

Milisauskas, Sarunas 1978 European Prehistory New
York: Academic Press.
Milisauskas, Sarunas 2000 European Prehistory, 2nd
edition. New York: Kluwer/Plenum.
Miracle, Preston T. and Stao Forenbaher 2005 Neolithic
and Bronze-Age Herders of Pupicina Cave, Croatia.
Monks, G. G. 1981 Seasonality studies. In Advances in
Archaeological Method and Theory, vol. 4, edited
by M. L. Schiffer. London: Academic Press, pp.
177-240.
Moreno Garca, M. 1997 The zooarchaeological evidence
for transhumance in Medieval Spain. Medieval
Europe Brugge 9: 45-54.
Moreno Garca, M. 1999 Ethnographic observations of
transhumant husbandry practices in Spain and their
applicability to the archaeological sample. In
Haskel J. Greenfield. Budapest: Archaeolingua, pp
159-180.
Munson, P. J. 2000 Age-correlated differential
destruction of bones and its effect on archaeological
mortality profiles of domestic sheep and goats.
Journal of Archaeological Science 27: 391-407.
Nandris, J. G. 1970 Groundwater as a factor in the first
temperate Neolithic settlement in the Krs region.
Zbornik Narodni Muzej Beograd (Belgrade) 6: 5971.
Nandris, J. G. 1977 The perspective of long-term change
in South-East Europe. In A Historical Geography of
the Balkans, edited by F. W. Carter. New York:
Academic Press, pp. 25-57.
Nandris, J. G. 1985. The Stina and the Katun:
foundations of a research design in European
Highland
Zone
Ethnoarchaeology.
World
Archaeology 17: 256-268.
Nandris, John 1991 The Balkan dimension of Highland
Zone in pastoralism. In Archeologia Della
Pastorizia Nell'Europa Meridionale, edited by R.
Nisbet, R. Maggi, and G. Barker. Rivista di Studi
Liguri 56 (1-4), pp. 99-107.
Navy 1944 Yugoslavia. Geographical Handbook Series,
Intelligence Division, Great Britain, volume 1.
Navy 1945 Yugoslavia. Geographical Handbook Series,
Intelligence Division, Great Britain, volume 3.
Niamir-Fuller, M. N. and Matthew D. Turner 1999 A
review of recent literature on pastoralism and
transhumance in Africa. In Managing Mobility in
African Rangelands: The Legitimization of
Transhumance, edited by Maryam Niamir-Fuller.
Rome: Food and Agriculture Organization of the
United Nations, pp. 18-46.
Nisbet, R., R. Maggi, and G. Barker (eds.) 1991
Archeologia
Della
Pastorizia
Nell'Europa
Meridionale. Rivista di Studi Liguri 56-57 (1-4).
OConnor, T. 2000 The archaeology of animal bones.
Phoenix Mill: Sutton Publishing Limited.
Olsen, S. J. 1971 Zooarchaeology: Animal Bones in
Waldern, W., Chapman, R., Lewthwaite, J., and

Kennard, R. British Archaeological Reports,
International Series 229 (Oxford), pp. 251-68.
Lieberman, D. E. 1993a Life history variables preserved
in dental cementum microstructure. Science 261
(27): 1162-1164.
Lieberman, D. E. 1993b Mobility and Strain: The Biology
of Cementogenesis and its Application to the
Evolution of Hunter-Gatherer Seasonal Mobility
During the Late Quaternary in the Southern Levant.
Ph.D. Thesis. Harvard University.
Lieberman, D. E. 1997 The biological basis for seasonal
increments in dental cementum and their application
to
archaeological
research.
Journal
of
Archaeological Science 21: 525-539.
Lieberman, D. E. and R. Meadow 1994 The biology of
cementum increments (with an archaeological
application). Mammal Review 22 (2): 57-77.
Lockwood, William 1975 European Moslems: Economy
and Ethnicity in Western Bosnia. Academic Press:
New York.
Logashova, B. R. 1982 Transformation of the social
organization of Iranian Tribes. Contemporary
Nomadic and Pastoral Peoples: Asia and the North.
Studies of Third World Societies 18: 53-60.
Lowe, V. P. W. 1967 Teeth as indicators of age with
special reference to Red Deer (Cervus elaphus) of
known age from Rhum. Journal of Zoology 152:
137-153.
Lyman, R. L. 1994 Vertebrate Taphonomy. Cambridge:
Cambridge University Press.
Lynch, T. F. 1971 Preceramic transhumance in the
Callejon de Huaylas, Peru. American Antiquity 36
(2):139-147
Lynch, T. F. 1980 (ed.) Guitarrero Cave: Early Man in
the Andes. Toronto: Academic Press.
Lynch, T. F. 1983 Camelid pastoralism and the
emergence of Tiwanaku Civilization in the SouthCentral Andes. World Archaeology 15: 1-14.
MacDonald, G. D. 1973 Area Handbook for Yugoslavia.
Washington, DC: US Government Printing Office.
Manson, Joni L. 1990 A Renanalysis of Starevo Culture
Ceramics: Implications for the Neolithic
Development in the Balkans. Unpublished Ph. D.
Thesis, Southern Illinois University at Carbondale.
Markovi, J. 1968 Fizina Geografija Jugoslavije.
Belgrade.
Matley, I. M. 1968 Transhumance in Bosnia and
Herzegovinia. The Geographical Review 58: 231261.
Matley, I. M. 1970 Traditional pastoral life in Romania.
Professional Geographer 22: 311-316.
McPherron, Alan and Dragoslav Srejovi, eds., 1988
Divostin and the Neolithic of Central Serbia.
University
of
Pittsburgh,
Department
of
Anthropology, Ethnology Monographs no. 10.
Mileusni, Zlatko 1987 Narodna privreda teanjskog
kraja (The folk economy in the region of Teanj).
Glasnik Zemaljskog muzeja Bosne I Hercegovine,
Etnologija, Nova serija, 41/42: 25-67.
129
REFERENCES CITED
Archaeology and their Interpretation. A McCaleb
Module in Anthropology from the series AddisonWesley Modular Publications, pp. 1-30.
Olsen, S. J. and S. Olsen 1981 A comment on
nomenclature in faunal studies. American Antiquity
46: 192-194.
Paine, R. 1998 Herds of the Tundra: A Portrait of Saami
Reindeer
Pastoralism.
Washington,
DC:
Smithsonian Institution Press.
Payne, S. 1973 Kill-off patterns in sheep and goats: the
mandibles from Avan Kale. Anatolian Studies 23:
281-303.
Payne, S. 1985 Morphological distinctions between the
mandibular teeth of young sheep, Ovis, and goats,
Capra. Journal of Archaeological Science 12: 139147.
Perez, Manuel Ruiz and Adelina Valero Saez 1990
Transhumance with cows as a rational land use
option in the Gredos Mountains (Central Spain).
Human Ecology 18 (2): 187-202.
Piggott, S. 1982 The Earliest Wheeled Transport: From
the Atlantic Coast to the Caspian Sea. Ithaca, New
York: Cornell University Press.
Pike-Tay, A. 1991 Red Deer Hunting in the Upper
Paleolithic of South-West France: A Study in
Seasonality. British Archaeological Reports,
International Series 569. Oxford: BAR.
Popovi, C. D. 1971 Les migrations des troupeaux en
Bosnie
et
Herzgovine.
Wissenschaftliche
Mitteilungen
des
Bosnisch-Herzegowinischen
Landesmuseums (Volkskunde: 1, Part B). Sarajevo,
pp. 99-116.
Pounds, N. 1969 Eastern Europe. Chicago: University of
Chicago Press.
Rafiullah, S. M. 1966 The Geography of Transhumance.
Aligarh: Aligarh Muslim University.
Raish, C. 1992 Domestic Animals and Stability in PreState Farming Societies. British Archaeological
Reports, International Series 579. Oxford: BAR.
Redding, R. 1984 Theoretical determinants of a herder's
decisions modeling variation in the sheep/goat ratio.
In Animals and Archaeology 3: Early Herders and
their Flocks, edited by J. Clutton-Brock and C.
Grigson, British Archaeological Reports S202.
Oxford, British Archaeological Reports, pp.
223-242.
Reitz, E. and E. Wing 1999 Zooarchaeology. Cambridge:
The University Press.
Rhoades, R. E. and S. I. Thompson 1975 Adaptive
strategies in alpine environments: beyond ecological
particularism. American Ethnologist 2: 535-551.
Russell, Nerissa 1998 Cattle as wealth in Neolithic
Europe: Where's the beef? In The Archaeology of
Value: Essays on Prestige and the Processes of
Valuation, edited by D. W. Bailey. British
Archaeological Reports, International Series no.
730. Oxford: British Archaeological Reports, ed. pp.
42-54.
Ryder, M. L. 1999 Did Vlach shepherds spread sheepmilking customs through South-East Europe? In

189-196.
Saxon, A. and C. Higham 1969 A new research method
for economic prehistorians. American Antiquity 34
(3): 303-311.
Seri, Haim 1956 Stoarska proizvodnja Bosne i
Hercegovine za vrijeme bosanske samostalnosti
(Animal production in Bosnia and Herzegovina
during the period of independence). Veterinaria
(Sarajevo) 5: 141-144.
Shennan, S. 1988 Quantifying Archaeology. Edinburgh:
Sherratt, A. 1981 Plough and Pastoralism: Aspects of the
Secondary Products Revolution. In Patterns of the
Past: Studies in Honour of David Clarke, edited by
Ian Hodder, G. Isaac and N. Hammond. Cambridge:
The University Press, pp. 261-306.
Sherratt, A. 1983a The secondary exploitation of animals
in the Old World. World Archaeology 15 (1): 90104.
Sherratt, A. 1983b The Development of Neolithic and
Copper Age Settlement in the Great Hungarian
Plain: Part II: Site survey and settlement dynamics.
Oxford Journal of Archaeology 2: 13-42.
Silver, I. A. 1969 The ageing of domestic animals. In
Science in Archaeology, edited by D. Brothwell and
E. S. Higgs. London: Thames and Hudson, pp. 283302.
Skydsgaard, Jens Erik 1974 Transhumance in ancient
Italy. Analecta Romana Instituti Danici 7: 7-47.
malcelj, Ivan 1954 Osnove ovarske politike u svijetu u
kod nas (The bases of sheep policy abroad and in
this country). Veterinaria (Sarajevo) 3: 203-215.
Sterud, Eugene 1978 Prehistoric populations of the
Dinaric Alps. In Social Archaeology: Beyond
Subsistence and Dating, edited by Charles. L.
Redman, M. J. Berman, E. V. Curtin, W. T.
Langhorne, Jr., N. M. Versaggi, and J. C. Wanser.
New York: Academic Press, pp. 381-408.
Stoianovi, T. 1967 A Study in Balkan Civilization. New
York: Knopf.
Sweet, Louise E. 1965 Camel pastoralism in north Arabia
and the minimal camping unit. In Man, Culture, and
Animals: The Role of Animals in Human Ecological
Adjustments, edited by A. Leeds and A.P. Vayda,
pp. 129-152. American Association for the
Advancement of Science Publication 78.
Washington, DC: American Anthropological
Association.
Szabadfalvi, Jzsef 1968 Nomadic winter system on the
Great Hungarian Plain. Acta Ethnographic
Academiae Scientarum Hungaricae 17: 139-167.
Tringham, R. 1971 Hunters, Fishers and Farmers of
Eastern Europe 6000-3000 BC. London:
Hutchinson University Library.
Tringham, R. 2000 Southeastern Europe in the transition
to agriculture in Europe: bridge, buffer or mosaic. In
130

Europes First Farmers, edited by T. Douglas Price.
Cambridge: The University Press, pp. 19-56.
Tringham, R. and D. Krsti 1990 Relative and absolute
chronology. In Selevac: A Neolithic Village in
Yugoslavia, edited by Ruth Tringham and Duan
Krsti. Los Angeles: UCLA Institute of
Archaeology, pp. 45-56.
Tringham, Ruth, Bogdan Brukner, and Barbara Voytek
1985 The Opovo Project: a Study of Socioeconomic
Change in the Balkan Neolithic Journal of Field
Archaeology 12: 425-444.
Tringham, Ruth, Bogdan Brukner, Timothy Kaiser,
Ksenija Borojevi, Ljubomir Bukvi, Petar Steli,
Nerissa Russell, Mirjana Stevanovi, and Barbara
Voytek 1992 Excavations at Opovo, 1985--1987:
Socioeconomic Change in the Balkan Neolithic
Van Andel, T. H. and C. N. Runnels 1995 The earliest
farmers in Europe. Antiquity 69: 481-500.
Vincze, Lajos 1974 Organization of work in herding
teams of the Great Hungarian Plain. Ethnology 13
(2): 159-170.
Vincze, Lajos 1980 Peasant animal husbandry: a dialectic
model of techno-environmental integration in agropastoral societies. Ethnology 19 (4): 387-404.
Vincze, Lajos 1983 Peasant herding associations in
Hungary and Romania. In The Keeping of Animals:
Adaptation and Social Relations in Livestock
Producing Communities, edited by Riva BerleantSchiller and Eugenia Shanklin. Allanheld, Osmun
Publishers; Totowa, NJ, pp. 37-55.
Vinak, Tomo 1983 Transhumatno stoarstvo na
Velebitu (Transhumant shepherds on Velebit).
Etnoloki Pregled 18: 101-105.
Vinak, Tomo 1988 The Leaja Family 1987: The Last
Transhumance in Bukovica, North Dalmatia.
Pamphlet accompanying video produced for the
Audiovisual Laboratory, Centre of Scientific
Research of the Slovene Academy of Sceinces and
Arts, Ljubljana, Yugoslavia.
Von den Driesch, A. 1976 Bone Measurements for
Archaeologists. Harvard University, Peabody
Museum, Bulletin no. 1. Cambridge.
Voytek, B. 1991 Resource use and the inference of a

pastoral economy. In Archeologia della Pastorizia
nell'Europa Meridionale, edited by R. Maggi,
Nisbet, and G. Barker. Rivista di Studi Liguri, A.
LVII (1991) Vol. 1. Bordighera: Istituto
Internazionale di Studi Liguri, pp. 47-58.
Walker, M. J. 1983 Laying a mega-myth: dolmens and
drovers in Prehistoric Spain. World Archaeology 15:
37-50.
Wells, P. 1984 Farms, Villages and Cities. Commerce
and Urban Origins in Late Prehistoric Europe.
Ithaca: Cornell University Press.
Wheeler Pires-Ferreira, J. 1975 Tepe Tula'i: faunal
remains from an early camp site in Khuzistan, Iran.
Paleorient 3: 275-280.
Whittaker, C. R. 1988 Pastoral Economies in Classical
Antiquity. Cambridge: Cambridge Philosophical
Society.
Whittle, A. 1985 Neolithic Europe: A Survey. Cambridge:
Whittle, A. 1996 Europe in the Neolithic: The Creation
of New Worlds. Cambridge: The University Press.
Wickham, Chris 1985 Pastoralism and underdevelopment
in the Early Middle Ages. Settimane di Studio del
Centro Italiano di Studi Sullalto Medioevo
(Spoleto) 31: 401-455.
Willis, K. J. and K. D. Bennett 1994 The Neolithic
Transition fact or fiction? Paleoecological
evidence from the Balkans. Holocene 4 (3): 326330.
Wilson, R., C. Grigson, and S. Payne (eds.) 1982 Ageing
and Sexing Animal Bones from Archaeological
Sites. British Archaeological Reports, British Series
109. Oxford: BAR.
Zdanovski, N. 1954 Stoarska kretanja (shephard
migrations). Radovi Poljoprivredno umarskog
Fakultet, Univerziteta u Sarajevo 3 (4-5): 117-143.
Zolyomi, B. 1980 Landwirtschaftliche Kultur und
Wandlung der Vegetation im Holozan am Balaton.
Phytocoenologia 7: 121-126.
131
TABLES
Table 1. Sheep/Goat mandibular wear stages (MWS) and suggested

ages
Payne MWS
A
B
C
D
E
F
G
H
I
Grant MWS
1-2
3-7
8-18
19-28
29-33
34-37
38-41
42-44
45+
Suggested Age (Payne 1973)

0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
Table 2. Cattle mandibular wear stages (MWS) and suggested ages

Payne MWS
A
B
C
D
E
F
G
H
I
Grant MWS
1-3
4-6
7-16
17-30
31-36
37-40
41-43
44-45
46+
Suggested Age (Halstead 1985)

0-1 months
1-8 months
8-18 months
18-30 months
30-36 months
young adult
adult
old adult
senile
Table 3. Pig mandibular wear stages (MWS) and suggested ages

Payne MWS
Grant MWS
A
B
C
D
E
F
G-I
0-1
2-8
9-17
18-32
33-42
43-46
46+
Suggested Age (Higham 1967;

Bull and Payne 1982))
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
adult
114
Table 4. Stage distribution of Ovis aries mandibles and loose teeth from
Early Neolithic Blagotin
Stage
Suggested Age
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
Raw
Count
No.
6
3
19
4
1
2
2
0
0
37
Corrected Count
%
16
9
51
11
3
5
5
0
0
100
No.
6
3
21.8
6.4
1.9
2.9
6
0
0
48
%
12
7
45
13
5
6
12
0
0
100

Stage
Suggested Age
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
Raw
Count
No.
4
1
10
7
1
1
1
3
0
28
115
Corrected Count
%
14
4
35
25
4
4
4
10
0
100
No.
4
1.1
13.9
26
6.6
11.6
9
16.8
0
89
%
4
1
16
30
7
13
10
19
0
100

Early Iron Age Blagotin
Stage
Suggested Age
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
Raw
Count
No.
0
0
1
2
0
0
1
0
0
4
Corrected Count
%
0
0
25
50
0
0
25
0
0
100
No.
0
0
1.6
4.4
0
0
5
0
0
11
%
0
0
15
40
0
0
45
0
0
100
Table 7. Stage distribution of Bos taurus mandibles and loose teeth from
Stage
Suggested Age
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
Young adult
Adult
old adult
Senile
Raw
Count
No.
9
3
12
15
2
4
0
1
1
116
Corrected Count
%
20
6
26
32
4
8
0
2
2
No.
9
3
12.4
22.9
3.9
12.8
0
7.2
6.3
%
12
4
16
29
5
17
0
9
8
47
100
77.5
100
Eneolithic Blagotin
Stage
Suggested Age
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
Young adult
Adult
old adult
Senile
Raw
Count
No.
0
0
2
3
2
1
0
0
0
8
Corrected Count
%
0
0
25
38
25
12
0
0
0
100
No.
0
0
2
4.8
3.2
2
0
0
0
12
%
0
0
17
40
26
17
0
0
0
100
Table 9. Stage distribution of Ovis aries mandibles and loose teeth from Early
Neolithic Foeni-Sala
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
Raw
Count
No.
0
1
7
1
0
0
0
0
0
9
117
Corrected %
Count
%
0
11
78
11
0
0
0
0
0
100
0
1
7.9
1.3
0
0
0
0
0
10.2
0
10
80
10
0
0
0
0
0
100
Early Neolithic Foeni-Sala
Stage
A
B
C
D
E
F
G
H
I
Suggested Age Raw

Count
No.
0-2 months
1
2-6 months
2
6-12 months
12
1-2 years
3
2-3 years
0
3-4 years
2
4-6 years
3
6-8 years
0
8-10 years
0
23
Corrected Count
%
4
9
52
13
0
9
13
0
0
100
No.
1.5
2.1
16.5
12.6
0
6
11.7
0
0
50.4
%
3
4
33
25
0
12
23
0
0
100
Early Iron Age Foeni-Sala
Stage
A
B
C
D
E
F
G
H
I
Suggested Age Raw

Count
No.
0-2 months
0
2-6 months
0
6-12 months
4
1-2 years
1
2-3 years
0
3-4 years
0
4-6 years
1
6-8 years
1
8-10 years
0
7
118
Corrected Count
%
0
0
58
14
0
0
14
14
0
100
No.
0
0
4
1
0
0
2.5
1.5
0
9
%
0
0
44
11
0
0
28
17
0
100
Table 12. Stage distribution of Bos taurus mandibles and loose teeth from Early
Neolithic Foeni-Sala
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
young adult
adult
old adult
senile
Raw
Count
No.
3
1
3
2
0
2
0
0
1
12
Corrected Count
%
25
8
25
17
0
17
0
0
8
100
No.
3
1
3
2
0
5
0
0
3
17
%
18
6
18
12
0
28
0
0
18
100
Table 13. Stage distribution of Ovis aries mandibles and loose teeth from Early Iron
Age Kadica Brdo
Stage
Suggested Age
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
Raw
Count
No.
4
5
20
5
1
0
1
0
0
36
119
Corrected Count
%
11
14
55
14
3
0
3
0
0
100
No.
4
5
25.6
9
1.4
0
2
0
0
47
%
8
10
54
20
4
0
4
0
0
100
Early Iron Age Kadica Brdo
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
Raw
Count
No.
1
1
15
10
5
16
17
3
3
71
Corrected Count
%
1
1
22
14
7
23
24
4
4
100
No.
1
1
25.5
40.5
19.2
69.8
39.8
5.6
5.6
208
%
1
1
12
19
9
34
18
3
3
100
Table 15. Stage distribution of Bos taurus mandibles and loose teeth from Early
Iron Age Kadica Brdo
Stage
Suggested Age
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
Young adult
Adult
old adult
Senile
Raw
Count
No.
4
0
3
8
0
2
1
1
6
25
120
Corrected Count
%
16
0
12
32
0
8
4
4
24
100
No.
4
0
3
14.5
0
7.1
3.8
3.3
11.2
46.9
%
8
0
6
32
0
15
8
7
24
100
Table 16. Stage distribution of Sus scrofa mandibles and loose teeth from
Early Iron Age Kadica Brdo
Stage
A
B
C
D
E
F
G
H
I
Suggested Age Raw

Count
No.
0-2 months
2
2-7 months
11
7-14 months
4
14-21 months 7
21-27 months 6
27-36 months 0
adult
0
old adult
0
senile
1
31
Corrected Count
%
6
36
13
23
19
0
0
0
3
100
No.
2
11.7
10
17.7
6.5
0
0
0
1
48.9
%
4
25
20
36
13
0
0
0
2
100
Table 17. Stage distribution of Ovis/Capra mandibles from Late Bronze Age
Livade
Stage
Suggested Age
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
Raw
Count
No.
0
0
0
1
0
2
3
2
0
8
121
Corrected Count
%
0
0
0
12
0
25
38
25
0
100
No.
0
0
0
1
0
2.8
5.01
2.19
0
11
%
0
0
0
9
0
25
46
20
0
100
Table 18. Stage distribution of Bos taurus mandibles from Late Bronze Age
levels of Livade
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
Young adult
Adult
Old adult
Senile
Raw
Count
No.
0
0
1
1
0
0
0
1
2
5
Corrected Count
%
0
0
20
20
0
0
0
20
40
100
No.
0
0
1
3.86
0
1.3
1.58
3.19
5.01
15.94
%
0
0
6
24
0
8
10
21
31
100
Table 19. Stage distribution of Sus scrofa mandibles and loose teeth from Late Bronze
Age Livade
Stage
Suggested Age
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
Adult
Old adult
Senile
Raw
Count
No.
0
1
2
7
7
1
0
0
0
18
122
Corrected Count
%
0
6
10
39
39
6
0
0
0
100
No.
0
2
4
9.5
10.38
1.12
0
0
0
27
%
0
8
15
35
38
4
0
0
0
100
Table 20. Stage distribution of Bos taurus mandibles from Early/ Middle Bronze
Age Ljuljaci
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
Young adult
Adult
Old adult
Senile
Raw
Count
No.
0
2
2
2
0
0
0
0
1
7
Corrected Count
%
0
28
28
28
0
0
0
0
15
99
No.
0
2
2
2
0
0.9
2.6
0
2.9
12.4
%
0
16
16
16
0
7
22
0
23
100
Early/Middle Bronze Age Ljuljaci
Stage
Suggested Age
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
Adult
Old adult
Senile
Raw
Count
No.
0
1
1
5
5
1
0
0
0
13
123
Corrected Count
%
0
8
8
38
38
8
0
0
0
100
No.
0
1
1.2
7.3
9
1.5
0
0
0
20
%
0
5
6
37
45
7
0
0
0
100
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
Raw
Count
No.
0
0
1
0
0
0
0
0
0
1
Corrected Count
%
0
0
100
0
0
0
0
0
0
100
No.
0
0
2
0.5
0.5
0
0
0
0
3
%
0
0
66
17
17
0
0
0
0
100
Table 23. Stage distribution of Ovis/Capra mandibles and loose teeth

from Late Neolithic/Final Neolithic Megalo Nisi Galanis
Stage
A
B
C
D
E
F
G
H
I
Suggested
Age
Raw
Count
No.
0-2 months 0
2-6 months 0
6-12 months 2
1-2 years
0
2-3 years
0
3-4 years
0
4-6 years
0
6-8 years
0
8-10 years 0
124
Corrected Count
%
0
0
100
0
0
0
0
0
0
No.
0
0
3
0.5
1
1.9
1.3
0.6
0.6
%
0
0
33
6
11
21
15
7
7
100
8.9
100
Final Neolithic Megalo Nisi Galanis
Stage
Suggested Age
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
Raw
Count
No.
1
0
2
0
0
0
0
0
0
3
Corrected Count
%
33
0
67
0
0
0
0
0
0
100
No.
1
0
4
0
0
0
0
0
0
5
%
20
0
80
0
0
0
0
0
0
100

from Final Neolithic Megalo Nisi Galanis
Stage
A
B
C
D
E
F
G
H
I
Suggested
Age
Raw
Count
No.
0-2 months 1
2-6 months 0
6-12 months 5
1-2 years
3
2-3 years
2
3-4 years
1
4-6 years
1
6-8 years
0
8-10 years 0
125
Corrected Count
%
8
0
38
23
15
8
8
0
0
No.
1
0
8.8
11.7
8.6
5.8
8.1
0
0
%
2
0
21
26
20
13
18
0
0
13
100
44
100
Table 26. Stage distribution of Ovis aries mandibles

and loose teeth from Final Neolithic/Early Bronze Age
Megalo Nisi Galanis
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
Raw
Count
No.
0
1
2
2
0
0
0
0
0
5
%
0
20
40
40
0
0
0
0
0
100

from Final Neolithic/Early Bronze Age Megalo Nisi Galanis
Stage
A
B
C
D
E
F
G
H
I
Suggested
Age
Raw
Count
No.
0-2 months 0
2-6 months 1
6-12 months 2
1-2 years
3
2-3 years
0
3-4 years
0
4-6 years
0
6-8 years
0
8-10 years 0
6
126
Corrected Count
%
0
17
33
50
0
0
0
0
0
100
No.
0
1
2
4.7
0.6
1.2
1
1.5
1
13
%
0
8
15
35
5
9
8
12
8
100
Eneolithic Novaka uprija
Stage
Suggested Age
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
Raw Count
No.
0
0
0
1
0
3
0
0
1
5
%
0
0
0
20
0
60
0
0
20
100
Corrected Count
No.
%
0
0
0
0
0
0
1.5
15
0
0
6.5
65
0
0
0
0
2
20
10
100
Table 29. Stage distribution of Ovis/Capra mandibles and loose teeth from Early
Bronze Age Novaka uprija
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
Raw Count
No.
0
1
1
8
1
3
1
0
2
17
127
%
0
6
6
46
6
18
6
0
12
100
Corrected Count
No.
%
0
0
1.5
5
1.5
5
10.9
34
1.7
5
5
16
7.8
24
0
0
3.6
11
32
100
Table 30. Stage distribution of Ovis/Capra mandibles and loose teeth from Late
Bronze Age Novaka uprija
Stage
Suggested Age
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
Raw Count
No.
0
0
1
1
1
1
0
0
0
4
%
0
0
25
25
25
25
0
0
0
100
Corrected Count
No.
%
0
0
0
0
1
6
1.6
10
3.6
23
3.6
23
4.5
28
1.6
10
0
0
15.9
100
Early Bronze Age Novaka uprija
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
Adult
Old adult
Senile
Raw
Count
No.
0
0
0
3
4
0
0
0
0
7
128
Corrected Count
%
0
0
0
43
57
0
0
0
0
100
No.
0
1
1
4
4
0
0
0
0
10
%
0
10
10
40
40
0
0
0
0
100
Late Neolithic Opovo
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
Young adult
Adult
Old adult
Senile
Raw
Count
No.
0
0
0
6
0
1
0
1
0
8
Corrected Count
%
0
0
0
75
0
12
0
12
0
99
No.
0
0.5
0.5
7
0
4.01
0
1.99
0
14
%
0
4
4
50
0
28
0
14
0
100
Stage
Suggested Age
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
Adult
Old adult
Senile
Raw
Count
No.
0
2
2
3
2
0
0
0
0
9
129
Corrected Count
%
0
22
22
33
22
0
0
0
0
99
No.
0
2
2.4
3.6
3
0
0
0
0
11
%
0
18
22
33
27
0
0
0
0
100
Late Neolithic Petnica
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
Raw
Count
No.
0
0
2
2
1
2
1
1
0
9
Corrected Count
%
0
0
22
22
11
22
11
11
0
99
No.
0
0
3.5
4.2
1.9
4.8
3.2
1.5
0
19.1
%
0
0
18
22
10
25
17
8
0
100
Eneolithic Petnica
Stage
Suggested Age
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
Raw
Count
No.
0
0
4
2
1
1
1
0
0
9
130
Corrected Count
%
0
0
45
22
11
11
11
0
0
100
No.
0
0
6.5
3.3
1.1
1.1
1
0
0
13
%
0
0
50
26
8
8
8
0
0
100
Late Bronze Age Petnica
Stage
Suggested Age
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
Raw
Count
No.
0
1
1
0
0
3
2
0
0
7
Corrected Count
%
0
14
14
0
0
43
29
0
0
100
No.
0
1
2
0.5
0.9
5.2
3.4
0
0
13
%
0
8
15
4
7
40
26
0
0
100
Middle Neolithic Petnica
Stage
Suggested Age
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
young adult
adult
old adult
senile
Raw
Count
No.
2
0
1
4
0
1
0
0
2
10
131
Corrected Count
%
20
0
10
40
0
10
0
0
20
100
No.
2
0
1
8.2
0
3.8
0
0
6.9
21.9
%
9
0
4
38
0
17
0
0
32
100
Table 38. Stage distribution of Bos taurus mandibles and loose teeth from Late
Neolithic Petnica
Stage
Suggested Age
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
young adult
adult
old adult
senile
Raw
Count
No.
1
0
0
3
1
0
0
0
1
6
Corrected Count
%
17
0
0
50
17
0
0
0
17
101
No.
1
1
1
5.8
5.5
0.8
1.6
0
2.3
19
%
5
5
5
31
30
4
8
0
12
100
Table 39. Stage distribution of Sus scrofa mandibles and loose teeth from Middle
Neolithic Petnica
Stage
Suggested Age
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
adult
old adult
senile
Raw
Count
No.
1
5
2
2
0
0
0
0
2
12
132
Corrected Count
%
8
41
17
17
0
0
0
0
17
100
No.
1
5
2
3
0
0
0
0
2
13
%
8
39
15
23
0
0
0
0
15
100
Table 40. Stage distribution of Sus scrofa mandibles and loose teeth from Late
Neolithic Petnica
Stage
A
B
C
D
E
F
G
H
I
Suggested Age Raw

Count
No.
0-2 months
0
2-7 months
2
7-14 months
3
14-21 months 2
21-27 months 0
27-36 months 0
adult
0
old adult
0
senile
1
8
Corrected Count
%
0
25
38
25
0
0
0
0
12
100
No.
0
2.4
3.6
6
0
0
0
0
1
13
%
0
18
28
46
0
0
0
0
8
100
Eneolithic Petnica
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
adult
old adult
senile
Raw
Count
No.
0
2
0
1
0
0
0
0
1
4
133
Corrected Count
%
0
50
0
25
0
0
0
0
25
100
No.
0
4
0
4
0.5
0.5
0
0
1
10
%
0
40
0
40
5
5
0
0
10
100
Middle Neolithic Stragari
Stage
Suggested Age
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
Raw Count
No.
0
1
3
5
1
7
0
0
0
17
%
0
6
18
29
6
41
0
0
0
100
Corrected Count
No.
%
0
0
1
5
3.9
18
8.3
38
1.4
6
7.3
33
0
0
0
0
0
0
21.9
100
Stage
Suggested Age
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
Young adult
Adult
old adult
Senile
Raw
Count
No.
2
0
6
2
0
0
1
0
1
12
134
Corrected Count
%
17
0
50
17
0
0
8
0
8
100
No.
2
0
7.5
4.3
0
0
9.2
0
6
29
%
7
0
26
15
0
0
32
0
20
100
Late Neolithic Vina
Stage
A
B
C
D
E
F
G
H
I
Suggested Age Raw

Count
No.
0-2 months
0
2-6 months
1
6-12 months 6
1-2 years
5
2-3 years
1
3-4 years
3
4-6 years
5
6-8 years
3
8-10 years
2
26
Corrected Count
%
0
4
23
19
4
12
19
12
7
100
No.
0
1.1
8.0
5.8
1.5
5.6
6.6
3.2
2.1
34
%
0
3
25
17
4
16
20
9
6
100
Middle Bronze Age Vina
Stage
A
B
C
D
E
F
G
H
I
Suggested Age Raw

Count
No.
0-2 months
0
2-6 months
0
6-12 months 3
1-2 years
5
2-3 years
1
3-4 years
1
4-6 years
1
6-8 years
1
8-10 years
1
13
135
Corrected Count
%
0
0
27
46
9
0
9
9
0
100
No.
0
0
3
6.7
2
3.6
3.6
2.9
1
22.8
%
0
0
13
29
9
16
16
13
4
100
Late Neolithic Vina
Stage
A
B
C
D
E
F
G
H
I
Suggested Age Raw

Count
No.
0-1 month
0
1-8 months
1
8-18 months
1
18-30 months 1
30-36 months 0
young adult
0
adult
0
old adult
0
senile
2
5
Corrected Count
%
0
20
20
20
0
0
0
0
40
100
No.
0
1.5
1.2
6
0.25
0.25
0.25
0.25
4
13.7
%
0
10
9
44
2
2
2
2
29
100
Stage
A
B
C
D
E
F
G
H
I
Suggested Age Raw

Count
No.
0-1 month
1
1-8 months
1
8-18 months
4
18-30 months 1
30-36 months 0
young adult
0
adult
0
old adult
1
senile
1
9
136
Corrected Count
%
11
11
45
11
0
0
0
11
11
100
No.
1
1
4
1
0
0
0
5.5
3.5
16
%
6
6
25
6
0
0
0
34
23
100
Late Neolithic Vina
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
Adult
old adult
Senile
Raw
Count
No.
0
5
8
9
1
0
0
0
0
23
Corrected Count
%
0
23
34
39
4
0
0
0
0
100
No.
0.0
6.1
11.3
13.3
2.3
0.0
0.0
0.0
0.0
33.0
%
0
18
33
42
7
0
0
0
0
100
Stage
Suggested Age
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
Adult
old adult
Senile
Raw
Count
No.
1
1
5
6
2
1
0
0
0
16
137
Corrected Count
%
6
6
31
38
13
6
0
0
0
100
No.
1.0
1.0
5.5
8.8
2.8
1.0
0.0
0.0
0.0
20.0
%
5
5
27
44
14
5
0
0
0
100
Table 50. Modern Ovis/Capra comparative cementum analysis Summary of readings

Sample
Slide Age
Goat #1
1 year, 4-5
months
1 year, 4-5
months
Goat #2
Goat #4
Goat #6
Goat #6
Goat #7
Goat #7
Goat #8
Goat #8
Goat #10
Goat #10
Goat #11
Goat #11
a
b
a
b
1 year, 4-5
months
1 year, 5-6
months
1 year, 5-6
months
1 year, 5-6
months
1 year, 5-6
months
1 year, 5-6
months
1 year, 5-6
months
6-7 months
6-7 months
6-7 months
6-7 months
Date of Death # of increments Nature of final Comments

(GLGs)
increment
Aug/Sept 2000 indeterminable growth zone?
Aug/Sept 2000 1GLG (partial) no
annulus/zone
final
Aug/Sept 2000 undeterminable
slide too thick
Sept/Oct 2000 1GLG
undeterminable reading taken at end of root
Sept/Oct 2000 1GLG
Sept/Oct 2000 1GLG
Sept/Oct 2000 undeterminable

Sept/Oct 2000 1GLG
root area too thick for accurate

reading
Sept/Oct 2000 1GLG
Oct/Nov 2000
Oct/Nov 2000
Oct/Nov 2000
Oct/Nov 2000
undeterminable
no GLG
growth zone
no GLG
growth zone
no GLG
growth zone
138
unreadable
Goat #12 a
8 months
December
2000
December
2000
December
2000
December
2000
Goat #12 b
8 months
Goat #13 a
8 months
Goat #13 b
8 months
Goat #14 a
Goat #14 b
9-10 months
9-10 months
Jan/Feb 2001
Jan/Feb 2001
Goat #15
Goat #16 a
9-10 months
9-10 months
Jan/Feb 2001
Jan/Feb 2001
Goat #16 b
Goat #17 a
Goat #17 b
9-10 months
9-10 months
9-10 months
Jan/Feb 2001
Jan/Feb 2001
Jan/Feb 2001
indeterminable indeterminable
no GLG
growth zone
undeterminable
bright outer increment =

annulus forming?
unreadable
2 annuli are
formed low on
the root
0 GLG
0 GLG
growth
zone/annulus
final
0 GLG
growth zone
tumor tooth
0 GLG
growth zone
right side - bright outer
increment = annulus forming?
0 GLG
growth zone
number of secondary GLGs
undeterminable
unreadable
undeterminable
unreadable
139
Table 51. Archaeological Ovis/Capra cementum analysis Summary of readings

Sectioning Site
Code
Location Level Taxon
Absolute Age
# of increments Nature of Final
(from tooth wear counted
Increment
and eruption)
KB #1
Highland
3-4 years
Highland
24 Ovis/Capra 2-3 years
2 GLG
growth zone
Highland
23 Ovis aries
1-2 years
1 GLG
indeterminate
Highland
12 Ovis aries
4-6 years
6 GLG
growth zone
Highland
Ovis aries
6-8 years
unreadable
Highland
16 Ovis aries
4-6 years
5 GLG
Highland
24 Ovis aries
2-3 years
unreadable
Highland
6-12 months
1 GLG
3-4 years
unreadable
Highland
Capra
hircus
25 Possible
goat
13 Ovis aries
6-12 months
0 GLG
growth zone
V #1
Kadica
Brdo
Kadica
Brdo
Kadica
Brdo
Kadica
Brdo
Kadica
Brdo
Kadica
Brdo
Kadica
Brdo
Kadica
Brdo
Kadica
Brdo
Kadica
Brdo
Vinca
Lowland
6-8 years
9 GLG
possible annulus
final
V #2
Vinca
Lowland
4-6 years
Tooth
KB #2
KB #3
KB #4
KB #5
KB #6
KB #7
KB #8
KB #9
KB #10
Highland
Ovis aries
Capra
hircus
Ovis aries
unreadable
140
growth zone
indeterminate
V #3
V #4
V #5
Vinca
Vinca
Vinca
Lowland
Lowland
Lowland
5
6
6
V #6
Vinca
Lowland
V #7
V #8
Vinca
Vinca
Lowland
Lowland
5
5
V #9
V #10
Vinca
Vinca
Lowland
Lowland
5
5
Ovis aries
Ovis aries
Capra
hircus
Ovis aries
Ovis aries
Capra
hircus
Ovis aries
Ovis/Capra
completely
broken during
extraction
8-10 years
unreadable
4-6 years
unreadable
6 months - 4 years unreadable
4-6 years
4 GLG
2-3 years
6-8 years
unreadable
unreadable
1-2 years
3-10 years
1 GLG
8 GLG
141
possible annulus
final
indeterminate
possible annulus
final
Table 52. Summary of sieving and weathering

Site
%
Sieved
0
Blagotin
FoeniSala
Kadica
Brdo
Megalo
Nisi
Galanis
Livade
Age class
(Ovis/Capra)
1-10 11-25 26-50 51-75 76-100
50
85
10
Age class
(Bos taurus)
Age class
(Sus scrofa)
2-6 6-12
1-8 8-18
2-7
0-2 month month month 0-1 month month month 0-2 month month
no
x
x
x
x
x
x
no data
data
no
x
x
x
x
x
x
no data
data
7-14
month
no
data
no
data
x
0
no
data
0
no
data
x
no data
0
no
data
x
no
data
x
no data
x
0
no
data
no data
0
Ljuljaci
Novaka
uprija
x
0
no
data
x
no
data
x
x
x
Opovo
Petnica
Stragari
Vina
x
no
data
x
x
x
0
x
0
0
x
x
0
x
x
x
0
x
100
100
20
0
0
0
no data
0
0
0
142
0
x
x
x
x
x
0
x
x
x
x
x
Table 53. Expectations for movement in transhumant

pattern
Age class
0-2
Lowland
Present
Mid-altitude Absent
Highland
Absent
2-6 month
Absent
Present
Present
6-12 month
Present
Absent
Absent
Age Class
0-1
Lowland
Present
Mid-altitude Absent
Highland
Absent
1-8 months
Absent
Present
Present
8-18 months
Present
Absent
Absent
Ovis/Capra
Bos taurus
Table 54. Summary of strata and weathering

Site
Superimposed strata
Weathering
Blagotin
Foeni-Sala
Kadica Brdo
Kozani
Livade
Ljuljaci
Novaka uprija
Opovo
Petnica
Deep
Shallow
Deep
Deep
Shallow
Shallow
Shallow
Deep
Deep
Low weather
Medium weathering
Low weather
Low weather
Very high weathering, except in deep pits
Medium weathering
Medium weathering
Low weather
Low weather
143
Stragari
Vina
Deep
Deep
Low weather
Low weather
144
APPENDIX A2
SUMMARY OF AGEABLE TOOTH WEAR AND ERUPTION DATA BY SITE,
TAXON AND PERIOD.
Site
Blagotin
Foeni-Sala
Kadica Brdo
Livade
Ljuljaci
Megalo Nisi
Galanis
Species
Ovis aries
Ovis/Capra
Ovis/Capra
Ovis/Capra
Bos taurus
Bos taurus
Bos taurus
Sus scrofa
Sus scrofa
Sus scrofa
Ovis aries
Ovis aries
Ovis/Capra
Ovis/Capra
Bos taurus
Bos taurus
Sus scrofa
Sus scrofa
Ovis aries
Ovis/Capra
Bos taurus
Sus scrofa
Ovis/Capra
Bos taurus
Sus scrofa
Ovis/Capra
Ovis/Capra
Ovis/Capra
Bos taurus
Bos taurus
Bos taurus
Sus scrofa
Sus scrofa
Sus scrofa
Ovis/Capra
Ovis/Capra
Ovis/Capra
Bos taurus
Sus scrofa
Sus scrofa
Period
Early Neolithic
Early Neolithic
Eneolithic
Early Iron Age
Early Neolithic
Eneolithic
Early Iron Age
Early Neolithic
Eneolithic
Early Iron Age
Early Neolithic
Early Iron Age
Early Neolithic
Early Iron Age
Early Neolithic
Early Iron Age
Early Neolithic
Middle Bronze Age
Early Iron Age
Early Iron Age
Early Iron Age
Early Iron Age
Late Bronze Age
Late Bronze Age
Late Bronze Age
Early Bronze Age
Early/Middle Bronze Age
Middle Bronze Age
Early Bronze Age
Middle Bronze Age
Early Bronze Age
Middle Bronze Age
Late Neolithic/Final Neolithic
Final Neolithic
Final Neolithic/Early Bronze Age
Final Neolithic
Late Neolithic/Final Neolithic
Final Neolithic
173
No. of
mandibles
45
42
0
4
38
3
2
2
0
1
10
5
36
7
10
2
1
1
39
111
18
42
11
10
26
2
8
1
2
5
1
4
18
5
9
12
4
2
1
4
No. of
Loose teeth
3
47
6
7
40
9
5
0
0
0
0
0
14
2
7
3
1
0
8
97
29
7
0
6
1
1
1
0
0
7
0
1
2
1
0
32
9
6
0
2
Total
48
89
6
11
78
12
7
2
0
1
10
5
50
9
17
5
2
1
47
208
47
49
11
16
27
3
9
1
2
12
1
5
20
6
9
44
13
8
1
6
APPENDIX A
Site
Novaka
uprija
Opovo
Petnica
Selevac
Stragari
Vina
Species
Ovis/Capra
Ovis/Capra
Ovis/Capra
Bos taurus
Bos taurus
Bos taurus
Sus scrofa
Sus scrofa
Sus scrofa
Ovis/Capra
Bos taurus
Sus scrofa
Ovis/Capra
Ovis/Capra
Ovis/Capra
Ovis/Capra
Ovis/Capra
Ovis/Capra
Bos taurus
Bos taurus
Bos taurus
Bos taurus
Bos taurus
Bos taurus
Sus scrofa
Sus scrofa
Sus scrofa
Sus scrofa
Sus scrofa
Ovis/Capra
Bos taurus
Sus scrofa
Ovis/Capra
Bos taurus
Sus scrofa
Ovis/Capra
Ovis/Capra
Ovis/Capra
Bos taurus
Bos taurus
Sus scrofa
Sus scrofa
Sus scrofa
Period
Eneolithic
Early Bronze Age
Late Bronze Age
Eneolithic
Early Bronze Age
Late Bronze Age
Eneolithic
Early Bronze Age
Late Bronze Age
Late Neolithic
Late Neolithic
Late Neolithic
Middle Neolithic
Late Neolithic
Late Neolithic/Eneolithic
Eneolithic
Late Bronze Age
Late Bronze Age/Early Iron Age
Middle Neolithic
Late Neolithic
Late Neolithic/Eneolithic
Eneolithic
Late Bronze Age
Middle Neolithic
Late Neolithic
Eneolithic
Late Bronze Age
Middle/Late Neolithic
Middle Neolithic
Middle Neolithic
Middle Neolithic
Late Neolithic
Eneolithic
Middle Bronze Age
Late Neolithic
Middle Bronze Age
Late Neolithic
Eneolithic
Middle Bronze Age
174
No. of
mandibles
0
8
3
2
2
1
4
6
2
5
7
10
4
5
3
9
4
3
15
4
0
3
0
2
12
11
8
3
4
110
78
unavailable
7
13
7
32
1
13
6
8
34
1
19
No. of
Loose teeth
10
24
13
0
4
3
4
4
1
1
7
1
3
14
3
4
9
6
7
15
2
6
9
4
1
2
2
2
1
0
0
unavailable
15
16
0
0
0
10
8
8
0
0
1
Total
10
32
16
2
6
4
8
10
3
6
14
11
7
19
6
13
13
9
22
19
2
9
9
6
13
13
10
5
5
110
78
22
29
7
32
1
23
14
16
34
1
20
APPENDIX B
STAGE DISTRIBUTION DATA OF SMALL SAMPLES.
Table B1. Stage distribution of Ovis/Capra mandibles
and loose teeth from Eneolithic Blagotin
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
0
2
0
0
1
0
0
0
3
Corrected Count
%
0
0
66
0
0
33
0
0
0
99
No.
0
0
2
1
1
1
0.5
0.5
0
6
%
0
0
33
17
17
17
8
8
0
100
Table B2. Stage distribution of Bos taurus mandibles and

Stage
A
B
C
D
E
F
G
H
I
Suggested Age
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
Young adult
Adult
old adult
Senile
Raw Count
No.
0
0
0
1
0
1
0
1
0
3
%
0
0
0
33
0
33
0
33
0
99
Corrected Count
No.
%
0
0
0
0
0
0
2.7
39
0
0
1.7
24
0
0
2.6
37
0
0
7
100
Table B3. Stage distribution of Sus scrofa mandibles and loose teeth from Early Neolithic Blagotin
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
adult
old adult
senile
No.
0
1
0
0
1
0
0
0
0
2
175
%
0
50
0
0
50
0
0
0
0
100
APPENDIX B
Table B4. Stage distribution of Sus scrofa mandibles and
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
adult
old adult
senile
No.
0
0
0
0
0
0
0
0
0
0
Corrected Count
%
0
0
0
0
0
0
0
0
0
0
No.
0
0
0.5
0.5
0
0
0
0
0
1
%
0
0
50
50
0
0
0
0
0
100
Table B5. Stage distribution of Ovis aries mandibles and loose teeth from Early Iron Age Foeni-Sala
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
0
4
1
0
0
0
0
0
5
%
0
0
80
20
0
0
0
0
0
100
Table B6. Stage distribution of Bos taurus mandibles

from Early Iron Age Foeni-Sala
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
young adult
adult
old adult
senile
No.
0
0
0
0
3
1
1
0
0
5
176
%
0
0
0
0
60
20
20
0
0
100

Table B7. Stage distribution of Sus scrofa mandibles and loose teeth from Early Neolithic Foeni-Sala
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
adult
old adult
senile
No.
0
0
0
1
1
0
0
0
0
2
%
0
0
0
50
50
0
0
0
0
100
Table B8. Stage distribution of Capra hircus mandibles and

Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
0
2
1
0
1
0
1
0
5
Corrected Count
%
0
0
40
20
0
20
0
20
0
100
No.
0
0
2
1.3
0
3.4
0
1.3
0
8
%
0
0
25
16
0
43
0
16
0
100
Table B9. Stage distribution of Ovis/Capra mandibles and

loose teeth from Early Bronze Age Ljuljaci
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
0
1
1
0
0
0
0
0
2
177
Corrected Count
%
0
0
50
50
0
0
0
0
0
100
No.
0
0
1
1
0
0
0.5
0.5
0
3
%
0
0
33
33
0
0
17
17
0
100
APPENDIX B

loose teeth from Early/Middle Bronze Age Ljuljaci
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
0
1
0
0
3
0
0
0
4
Corrected Count
%
0
0
25
0
0
75
0
0
0
100
No.
0
0
1
0
0
5
1
1.5
0.5
9
%
0
0
11
0
0
55
11
17
6
100
Table B11. Stage distribution of Ovis/Capra mandibles

and loose teeth from Middle Bronze Age Ljuljaci
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
0
0
0
1
0
0
0
0
1
%
0
0
0
0
100
0
0
0
0
100

Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
Young adult
Adult
Old adult
Senile
No.
0
1
0
0
0
0
0
0
0
1
178
Corrected Count
%
0
100
0
0
0
0
0
0
0
100
No.
0
1
0
0
0
0
0.3
0.3
0.3
1.9
%
0
52
0
0
0
0
16
16
16
100

Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
Adult
Old adult
Senile
No.
0
0
1
1
0
0
0
0
0
2
Corrected Count
%
0
0
50
50
0
0
0
0
0
100
No.
0
0
1
2
0
0
0.7
0.7
0.7
5.1
%
0
0
20
40
0
0
13
13
13
99

loose teeth from Middle Bronze Age Ljuljaci
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
Adult
Old adult
Senile
No.
0
0
1
1
1
0
0
0
0
3
Corrected Count
%
0
0
33
33
33
0
0
0
0
99
No.
0
0
1
2.5
2.5
0
0
0
0
6
%
0
0
16
42
42
0
0
0
0
100

Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
Young adult
Adult
old adult
Senile
No.
0
0
0
4
0
0
0
0
1
5
179
Corrected Count
%
0
0
0
80
0
0
0
0
20
100
No.
0
0
0
7
0
0
0
0
1
8
%
0
0
0
88
0
0
0
0
12
100
APPENDIX B
Table B16. Stage distribution of Sus scrofa mandibles and loose teeth from Late Neolithic/Final Neolithic
Megalo Nisi Galanis
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
adult
old adult
senile
No.
0
1
0
0
0
0
0
0
0
1
%
0
100
0
0
0
0
0
0
0
100

Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
Adult
Old adult
Senile
No.
0
2
0
2
0
0
0
0
0
4
Corrected Count
%
0
50
0
50
0
0
0
0
0
100
No.
0
2
0
4
0
0
0
0
0
6
%
0
33
0
66
0
0
0
0
0
99

Stage
Suggested Age
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
young adult
adult
old adult
senile
Raw Count
No.
0
0
0
0
1
0
0
0
1
2
180
%
0
0
0
0
50
0
0
0
50
100

Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
Young adult
Adult
Old adult
Senile
No.
0
0
0
2
0
0
0
0
1
3
Corrected Count
%
0
0
0
66
0
0
0
0
33
99
No.
0
0
0
2
0
0
0
0
4
6
%
0
0
0
40
0
0
0
0
60
100

loose teeth from Late Bronze Age Novaka uprija.
Stage
A
B
C
D
E
F
G
H
I
Suggested Age
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
Young adult
Adult
Old adult
Senile
Raw Count
No.
0
0
0
0
0
0
0
0
3
3
%
0
0
0
0
0
0
0
0
100
100
Corrected Count
No.
%
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
4
100
4
100

Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
Adult
Old adult
Senile
No.
0
1
0
3
1
0
0
0
0
5
181
Corrected Count
%
0
20
0
60
20
0
0
0
0
100
No.
0
1
0
5.5
1.5
0
0
0
0
8
%
0
12
0
69
19
0
0
0
0
100
APPENDIX B

loose teeth from Late Bronze Age Novaka uprija
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
adult
old adult
senile
No.
0
0
0
2
1
0
0
0
0
3
%
0
0
0
66
33
0
0
0
0
99

Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
0
2
1
0
0
1
0
0
4
Corrected Count
%
0
0
50
25
0
0
25
0
0
100
No.
0
0
2
1.3
0
0
2.7
0
0
6
%
0
0
33
22
0
0
45
0
0
100

Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
1
3
0
1
0
0
0
0
5
182
Corrected Count
%
0
20
60
0
20
0
0
0
0
100
No.
0
1
3
0
1.5
0.75
0.75
0
0
7
%
0
14
43
0
21
11
11
0
0
100

loose teeth from Late Neolithic/Eneolithic Petnica
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
0
1
1
0
1
0
0
0
3
Corrected Count
%
0
0
33.3
33.3
0
33.3
0
0
0
99.9
No.
0
0
2
2
0
2
0
0
0
6
%
0
0
33.3
33.3
0
33.3
0
0
0
99.9

loose teeth from Late Bronze Age/Early Iron Age Petnica
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
0
4
0
1
0
0
0
0
5
Corrected Count
%
0
0
80
0
20
0
0
0
0
100
No.
0
0
5
0
1.4
0.7
1.4
0.5
0
9
%
0
0
56
0
15
8
15
6
0
100
Table B27. Stage distribution of Bos taurus mandibles and loose teeth from Eneolithic Petnica
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
young adult
adult
old adult
senile
No.
0
0
0
1
1
1
0
1
0
4
183
Corrected Count
%
0
0
0
25
25
25
0
25
0
100
No.
0
0.5
0.5
2.5
2.5
1
0
2
0
9
%
0
6
6
28
28
10
0
22
0
100
APPENDIX B
Table B28. Stage distribution of Bos taurus mandibles and loose teeth from Late Bronze Age Petnica
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
young adult
adult
old adult
senile
No.
0
0
0
0
1
0
0
0
1
2
Corrected Count
%
0
0
0
0
50
0
0
0
50
100
No.
0
0
0.5
0.5
2.7
0.8
0.8
0
3.8
9.1
%
0
0
5
5
30
9
9
0
42
100
Table B29. Stage distribution of Bos taurus mandibles and loose teeth from Late Bronze Age/Early Iron Age Petnica
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-1 month
1-8 months
8-18 months
18-30 months
30-36 months
young adult
adult
old adult
senile
No.
0
0
0
0
0
0
0
0
1
1
Corrected Count
%
0
0
0
0
0
0
0
0
100
100
No.
0
0.5
0.5
1
1
0
0
0
3
6
%
0
8
8
17
17
0
0
0
50
100
Table B30. Stage distribution of Sus scrofa mandibles and loose teeth from Late Bronze Age Petnica
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
adult
old adult
senile
No.
0
0
1
2
0
0
0
0
0
3
184
Corrected Count
%
0
0
33.3
66.6
0
0
0
0
0
99.9
No.
0
0
2.3
2.6
0
0
0
0
0
4.9
%
0
0
47
53
0
0
0
0
0
100

Table B31. Stage distribution of Sus scrofa mandibles and loose teeth from Late Bronze Age/Early Iron Age Petnica
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
adult
old adult
Senile
No.
0
0
2
0
0
0
0
0
0
2
Corrected Count
%
0
0
100
0
0
0
0
0
0
100
No.
0
0
4
0.5
0.5
0
0
0
0
5
%
0
0
80
10
10
0
0
0
0
100
Table B32. Stage distribution of Ovis aries mandibles and loose teeth from Middle Neolithic Stragari
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
1
2
0
0
0
0
0
0
3
Corrected Count
%
0
33
66
0
0
0
0
0
0
99
No.
0
1
3
0
0
0
0
0
0
4
%
0
25
75
0
0
0
0
0
0
100
Table B33. Stage distribution of Sus scrofa mandibles and loose teeth from Middle Neolithic Stragari
Stage
Suggested Age
Raw Count
A
B
C
0-2 months
2-7 months
7-14 months
No.
0
1
2
%
0
25
50
No.
0
1
3.3
%
0
14
48
D
E
F
G
H
I
14-21 months
21-27 months
27-36 months
adult
old adult
senile
1
0
0
0
0
0
4
25
0
0
0
0
0
100
2.6
0
0
0
0
0
6.9
38
0
0
0
0
0
100
185
Corrected Count
APPENDIX B
Table B34. Stage distribution of Ovis/Capra mandibles and loose teeth from Eneolithic Vina
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-6 months
6-12 months
1-2 years
2-3 years
3-4 years
4-6 years
6-8 years
8-10 years
No.
0
0
0
1
0
0
0
0
0
1
%
0
0
0
0
100
0
0
0
0
100

loose teeth from Eneolithic Vina
Stage
Suggested Age
Raw Count
A
B
C
D
E
F
G
H
I
0-2 months
2-7 months
7-14 months
14-21 months
21-27 months
27-36 months
adult
old adult
senile
No.
0
0
0
1
0
0
0
0
0
1
186
%
0
0
0
100
0
0
0
0
0
100
Appendix C. Comparison of the known age of death from modern specimens with Grant (1978) and Paynes (1973) tooth eruption and wear sequences recording methods.
Recording
method
Species Specimen #
Date of
Death
Age at
Death
m3/P4 code
Grant
Goat
Goat #3
Aug/Sept 1 year, 42000

5 months
P4/ E
Payne
Goat
Goat #3

5 months
P4/ E
Grant
Goat
Goat #4

5 months
m3/ m
Payne
Goat
Goat #4

5 months
Grant
Goat
Goat #1

5 months
Payne
Goat
Goat #1

5 months
Grant
Goat
Goat #2

5 months
P4/ E
Payne
Goat
Goat #2

5 months
P4/ E
Grant
Sheep
Sheep #8
Sept. 20,
2000
5 months
m3/ g
Payne
Sheep
Sheep #8
Sept. 20,
2000
5 months
Grant
Sheep
Sheep #7
Sept. 20,
2000
5 months
Payne
Sheep
Sheep #7
Sept. 20,
2000
5 months
m3/P4
description
M1 Code
M1
description
M2 Code
M2
description
M3 Code
Numerical
Value
(Grant
M3 description
1975)
m3/ g
1-2 years
1-2 years
1-2 years
6-12
months
6-12
months
24
1-2 years
24
24
m3/ n
Absolute
Age
24
Age
Class
(Payne
1973)
11
Recording
method
Species Specimen #
Date of
Death
Age at
Death
m3/P4 code
Grant
Sheep
Sheep #4
Nov. 21,
2000
Payne
Sheep
Sheep #4
Nov. 21,
2000
Grant
Goat
Sept/Oct 1 year, 52000

6 months
P4/ E
Payne
Goat
Sept/Oct 1 year, 56 months

2000
P4/ E
Grant
Goat

6 months
m3/ m
Payne
Goat

6 months
Grant
Goat

6 months
P4/ E
Payne
Goat

6 months
P4/ E
Grant
Goat

6 months
m3/ n
Payne
Goat

6 months
Grant
Goat

6 months
Payne
Goat

2000
Grant
Goat

6 months
m3/ n
m3/P4
description
M1 Code
M1
description
M2 Code
M2
description
M3 Code
Numerical
Value
(Grant
M3 description
1975)
lingual half of
Unrecordable tooth missing
lingual half of
m3/ n
1-2 years
1-2 years
1-2 years
1-2 years
Not
recordable
1-2 years
23
m3/ n
24
1-2 years
21
23
Absolute
Age
26
Age
Class
(Payne
1973)
23
Recording
method
Species Specimen #
Date of
Death
Age at
Death
m3/P4 code
m3/P4
description
M1 Code
M1
description
Payne
Goat

6 months
Grant
Goat

6 months
Payne
Goat

6 months
Grant
Goat
Goat #5

6 months
P4/ E
lingual half of
Payne
Goat
Goat #5

2000
P4/ E
lingual half of
Grant
Goat
Goat #8

6 months
P4/ 1/2
Payne
Goat
Goat #8

6 months
P4/ 1/2
Grant
Goat
Goat #6

6 months
m3/ m
Payne
Goat
Goat #6

6 months
Grant
Goat
Goat #7

6 months
Payne
Goat
Goat #7

6 months
Grant
Goat
Goat #9

6 months
Payne
Goat
Goat #9

2000
M2 Code
M2
description
M3 Code
Numerical
Value
(Grant
M3 description
1975)
m3/ n
1-2 years
1-2 years
1-2 years
1-2 years
1-2 years
1-2 years
21
m3/ l
21
C
m3/ n
1-2 years
22
Not
recordable
Absolute
Age
24
Age
Class
(Payne
1973)
22
Recording
method
Species Specimen #
Date of
Death
Age at
Death
m3/P4 code
m3/P4
description
M1 Code
M1
description
Grant
Sheep
Sheep #6
Oct. 13,
2000
P4
posterior cusp
unrecordable
damaged
Payne
Sheep
Sheep #6
Oct. 13,
2000
posterior cusp
P4
damaged
unrecordable
Grant
Sheep
Sheep #10
Oct. 13,
2000
6 months
Payne
Sheep
Sheep #10
Oct. 13,
2000
6 months
Grant
Sheep
Sheep #9
Oct. 13,
2000
6 months
Payne
Sheep
Sheep #9
Oct. 13,
2000
6 months
Grant
Sheep
Sheep #13
Oct. 13,
2000
6 months
Payne
Sheep
Sheep #13
Oct. 13,
2000
6 months
Grant
Sheep
Sheep #11
Oct. 13,
2000
6 months
Payne
Sheep
Sheep #11
Oct. 13,
2000
6 months
Grant
Goat
Goat #10
Oct/Nov
2000
6-7
months
P4/ 1/2
Unrecordable
posterior cusp
missing
Payne
Goat
Goat #10
Oct/Nov
2000
6-7
months
P4/ 1/2
Unrecordable
posterior cusp
missing
Grant
Goat
Goat #11
Oct/Nov
2000
6-7
months
P4/ E
m3/ g
M2 Code
g
M2
description
M3 Code
c
Numerical
Value
(Grant
M3 description
1975)
6-12
months
6-12
months
6-12
months
6-12
months
1-2 years
11
Not
recordable
10
2-3 years
12
m3/ g
10
m3/ g
Absolute
Age
33
m3/ g
Age
Class
(Payne
1973)
22
Recording
method
Species Specimen #
Date of
Death
Age at
Death
m3/P4 code
Payne
Goat
Goat #11
Oct/Nov
2000
6-7
months
P4/ E
Grant
Sheep
Sheep #2
Nov. 26,
2000
1 year, 6
months
m3/ m
Payne
Sheep
Sheep #2
Nov. 26,
2000
1 year, 6
months
Grant
Sheep
Sheep #3
Nov. 26,
2000
1 year, 6
months
Payne
Sheep
Sheep #3
Nov. 26,
2000
1 year, 6
months
Grant
Sheep
Sheep #1
Nov. 26, 4 years,

2000
6 months
Payne
Sheep
Sheep #1
Nov. 26, 4 years,

2000
6 months
M1 Code
M1
description
M2 Code
M2
description
M3 Code
m3/ h
m3/ n
(abnormal
wear - anterior
cusp very
worn)
Goat
Dec. 2000
Payne
Goat
Dec. 2000
Grant
Goat
Dec. 2000
P4/ U
Payne
Goat
Dec. 2000
P4/ U
Grant
Goat
Dec. 2000 8 months
m3/ n
1-2 years
1-2 years
1-2 years
4-6 years
1-2 years
1-2 years
25
25
41
Absolute
Age
26
P4/ j
Age
Class
(Payne
1973)
26
Grant
Goat #13
m3/P4
description
Numerical
Value
(Grant
M3 description
1975)
24
Recording
method
Species Specimen #
Date of
Death
Age at
Death
Payne
Goat
Goat #13
Dec. 2000 8 months
Grant
Goat
Goat #12
Dec. 2000 8 months
Payne
Goat
Goat #12
Dec. 2000 8 months
Grant
Goat
Dec. 2000
Payne
Goat
Dec. 2000
Grant
Sheep
Nov. 20,
2000
Payne
Sheep
Nov. 20,
2000
Grant
Sheep
Nov. 20,
2000
Payne
Sheep
Nov. 20,
2000
Grant
Sheep
Nov. 20,
2000
Payne
Sheep
Nov. 20,
2000
Grant
Goat
Goat #14
Jan/Feb
2001
9-10
months
Payne
Goat
Goat #14
Jan/Feb
2001
9-10
months
m3/P4 code
m3/P4
description
M1 Code
M1
description
M2 Code
M2
description
M3 Code
Numerical
Value
(Grant
M3 description
1975)
m3/ n
1-2 years
1-2 years
6-12
months
6-12
months
6-12
months
1-2 years
13
P4/ 1/2
12
m3/ h
1-2 years
13
m3/ j
21
m3/ j
Absolute
Age
23
m3/ m
Age
Class
(Payne
1973)
25
Recording
method
Species Specimen #
Date of
Death
Age at
Death
m3/P4 code
m3/ n (very
worn)
Grant
Goat
Goat #15
Jan/Feb
2001
9-10
months
Payne
Goat
Goat #15
Jan/Feb
2001
9-10
months
Grant
Goat
Goat #16
Jan/Feb
2001
9-10
months
Payne
Goat
Goat #16
Jan/Feb
2001
9-10
months
Grant
Goat
Goat #17
Jan/Feb
2001
9-10
months
Payne
Goat
Goat #17
Jan/Feb
2001
9-10
months
m3/P4
description
M1 Code
M1
description
M2 Code
d
M2
description
M3 Code
C
Numerical
Value
(Grant
M3 description
1975)
M3/ n
lingual half of
tooth
Unrecordable
damaged
lingual half of
tooth
damaged
Unrecordable
E
E
1-2 years
1-2 years
1-2 years
23
Absolute
Age
23
P4/ U
Age
Class
(Payne
1973)
Not
recordable
Appendix D: Details of the dental cementum analysis from archaeological and modern specimens.
Specimen
Number
Goat #1
Goat #2
Goat #4
Goat #5
Goat #6
Goat #7
Goat #8
Goat #9
Goat #10
Goat #11
Goat #12
Goat #13
Goat #14
Goat #15
Goat #16
Goat #17
Sheep #1
Sheep #2
Sheep #3
Sheep #4
Sheep #6
Vina #1
Vina #2
Vina #3
Vina #4
Vina #5
Vina #6
Vina #7
Vina #8
Vina #9
Vina #10
Kadica Brdo #1
Kadica Brdo #2
Kadica Brdo #3
Kadica Brdo #4
Kadica Brdo #5
Kadica Brdo #6
Kadica Brdo #7
Kadica Brdo #8
Kadica Brdo #9
Kadica Brdo #10
Type
GL (mm) Tht (mm) Cht (mm) OL (mm) OW (mm) Wt (g)
Modern
Modern
Modern
Modern
26.54
26.54
14.76
8.33
3.94
Modern
24.17
24.17
14.57
8.39
3.58
Modern
24.1
24.1
15.64
7.97
4.17
Modern
24.13
24.13
14.53
8.09
3.56
Modern
23.72
23.72
15.47
7.81
4.05
Modern
Modern
Modern
35.12
35.12
21.6
14.05
8.47
3.51
Modern
33.59
33.59
22.56
13.69
8.13
3.29
Modern
34.71
34.71
22.49
15.17
8.87
4.33
Modern
32.68
32.68
23.34
14.9
7.65
3.67
Modern
32.74
32.74
22.38
14.33
7.87
3.29
Modern
34.01
34.01
23.6
15.4
8.51
4.22
Modern
28.93
28.93
9.99
12.01
7.98
2.22
Modern
38.83
38.83
23.51
15.36
7.7
4.18
Modern
39.09
39.09
23.89
15.4
8.3
4.7
Modern
35.71
35.71
22.93
15.5
8.27
3.71
Modern
39.35
39.35
23.6
15.54
8.22
4.13
Archaeological
23.32
23.32
6.70
10.73
7.67
1.57
Archaeological Eliminated from sample, tooth completely broken during extraction
Archaeological
19.61
19.61
6.01
9.59
6.81
1.28
Archaeological
26.48
26.48
15.63
11.36
7.08
1.88
Archaeological
23.41
23.41
11.85
11.03
6.91
1.87
Archaeological
27.89
27.89
10.85
11.26
7.03
2.15
Archaeological
26.20
26.20
11.55
11.23
7.03
1.96
Archaeological
19.60
19.60
5.80
9.89
6.24
1.33
Archaeological
30.12
30.12
17.92
13.21
6.91
2.67
Archaeological
18.59
18.59
7.33
10.41
6.05
1.37
Archaeological
26.73
26.73
15.81
11.90
7.34
2.28
Archaeological
26.19
26.19
11.77
11.46
7.40
2.04
Archaeological
31.40
31.40
22.61
12.64
6.75
2.94
Archaeological
20.81
20.81
11.88
11.62
6.67
1.54
Archaeological
20.38
20.38
6.34
10.50
7.61
1.71
Archaeological
21.72
21.72
6.98
10.31
7.31
1.36
Archaeological
31.04
31.04
16.40
12.15
7.40
2.47
Archaeological
30.43
30.43
21.80
13.63
7.12
2.71
Archaeological
26.93
26.93
12.50
11.92
7.74
2.32
Archaeological
32.42
32.42
21.37
14.01
6.58
2.74
Definitions:
GL = greatest length
Tht = tooth height, or greatest length if broken
CH = crown height, cortch to highest part of the crown of the tooth
OL = overall length
OW = overall width
Wt = dry weight of the tooth after processing and drying

Notes:
The first group of modern samples had no measurements taken.
The omission was not realized until the teeth had already been embedded and cut.
APPENDIX E
STATISTICAL ANALYSIS
Table E1. Statistical analysis data - Late Neolithic Sus scrofa
Suggested Age
2-7 months
7-14 months
14-21 months
21 months - senile
Late Neolithic Vina
2
4
6
1
6
11
13
2
Late Neolithic
Opovo
2
2
4
3
X2 = 4.431 with 6 degrees of freedom

Exact p-value = 0.6486
Usual p value = 0.6185
Table E2. Statistical analysis data - Bronze Age Sus scrofa

Suggested Age
Early Bronze Age Novaka

uprija
2-7 months
7-14 months
14-21 months
21 months-senile
Early/Middle Bronze Age Ljuljaci Middle Bronze Age

Vina
1
1
4
4
1
1
7
11
1
6
9
3

Exact p value = 0.1499
Table E3. Statistical analysis data - comparison of major periods (Sus scrofa)
Suggested Age
2-7 months
7-14 months
14-21 months
21months- senile
Eneolithic Petnica
2
4
6
1
4
0
4
2

197
Early Bronze Age

Novaka uprija
1
1
4
4
Table E4. Statistical analysis data - Early Neolithic Ovis/Capra

Suggested Age
2-6 months
6-12 months
1-2 years
2-10 years
2
16
13
18
1
14
26
44

Table E5. Statistical analysis data - Late Neolithic Ovis/Capra
Suggested Age
Late Neolithic Vina
2-6 months
6-12 months
1-2 years
2-10 years
1
8
6
19
0
4
4
11

Table E6. Statistical analysis data - Eneolithic Ovis/Capra

Suggested Age
Eneolithic Petnica
2-6 months
6-12 months
1-2 years
2-10 years
Eneolithic Novaka uprija

0
6
3
3
0
0
1
8

Table E7. Statistical analysis data - Early/Middle Bronze Age Ovis/Capra

Suggested Age
Early Bronze Age Novaka uprija
2-6 months
6-12 months
1-2 years
2-10 years
1
1
11
18

0
3
7
13

198
Table E8. Statistical analysis data - Late Bronze Age Ovis/C

Suggested Age
2-6 months
6-12 months
1-2 years
2-10 years
Late Bronze Age Petnica
Late Bronze Age Novaka uprija Late Bronze Age Livade
1
2
0
9
0
1
2
13
0
0
1
10

Table E9. Statistical analysis data - Early Neolithic Bos taurus

Suggested Age
1-8 months
8-18 months
18-30 months
30 months-senile

1
3
2
8
3
12
23
30

Table E10. Statistical analysis data - Early Neolithic Bos vs. Ovis/Capra
Suggested Age
1-8 months
8-18 months
18-30 months
30 months-senile
Bos taurus
Ovis/Capra
4
15
25
38
3
30
39
62

Table E11. Statistical analysis data - Middle Neolithic Bos taurus
Suggested Age
1-8 months
8-18 months
18-30 months
30 months-senile
Middle Neolithic Petnica

0
7
4
15
0
1
8
11

199
Table E12. Statistical analysis data - Late Neolithic Bos taurus

Suggested Age
1-8 months
8-18 months
18-30 months
30 months-senile
1
1
6
10
0
0
7
6

200
Late Neolithic Vina

1
1
6
5

The Origins of Transhumant Pastoralism in Temperate SE Europe - Arnold Greenfield

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Origins of Transhumant Pastoralism in Temperate SE Europe - Arnold Greenfield

Uploaded by

Copyright:

Available Formats

The Origins of Transhumant

Pastorialism in Temperate South

BAR International Series XXXX

I. Theoretical background to research problem

CHAPTER 2: TRANSHUMANT PASTORALISM: SOME ETHNOGRAPHIC CONSIDERATIONS

CHAPTER 4: REGIONAL ENVIRONMENT

E. Evidence for sedentism/mobility

The Central Balkans and site locations

Map showing climatic divide between Mediterranean

Proportional allocation example

Paynes production models

Percentage of age groups (Sus scrofa) Post Neolithic

Sheep/Goat mandibular wear stages (MWS) and suggested ages

Summary of sieving and weathering

B. STAGE DISTRIBUTION DATA OF SMALL SAMPLES:

C. COMPARISON OF THE KNOWN AGE OF DEATH FROM MODERN SPECIMENS

E. STATISTICAL ANALYSIS DATA

Table E1. Statistical analysis data - Late Neolithic Sus scrofa.

Europe. Greenfield (1986, 1988, 1991, 1999a, 2001b)

Transhumant pastoralism is an economic activity

Although it has been argued that transhumant

Historically, transhumant pastoralism has been a

II. Research hypotheses

In contrast, few projects have focused on the temperate

Early NeolithicTranshumant pastoralism has long

It is with the shift to longer distance transhumance, where

Early Post NeolithicThe Secondary Products

Iron Age (or later)Transhumant pastoralism is

Long versus short distance transhumant migration would

The hypotheses can be aptly summarized to propose that

Therefore, the appearance of transhumant pastoralism in

III. Method and Technique

THE ORIGINS OF TRANSHUMANT PASTORALISM IN TEMPERATE SOUTHEASTERN EUROPE

often juxtaposed in close proximity. This is the kind of

First, harvest profiles are created from the analysis of the

Second, the results of the tooth wear and eruption

While not new, techniques such as tooth wear and

The identification stage of analysis, including species

THE FORMER YUGOSLAV REPUBLIC OF MACEDONIA

THE ORIGINS OF TRANSHUMANT PASTORALISM IN TEMPERATE SOUTHEASTERN EUROPE

The proposed results of this research should reveal the

The modern data consists of comparative data from

II. Types of pastoralism

Nomadic pastoralism involves the movement of people

The mobility of pastoral groups is not limitless.

pastoralism, the herd moves between fixed locations

Historically, transhumant pastoralism has been, and is, a

Transhumant pastoralism is part of a more broadly based

THE ORIGINS OF TRANSHUMANT PASTORALISM IN TEMPERATE SOUTHEASTERN EUROPE

villages. Highland shepherds practice vertical

In the temperate environment of the northern Balkans, the

Under conditions of vertical transhumance, two different

Lowland based herds do not have to be moved between

The highland and lowland zones present different

Village is located proximate to both lowland and

In terms of seasonality of movement, the following

Vinak 1983, 1988; Zdanovski 1954: 121). The exact

Traditionally, transhumant pastoralists were integrated

In terms of specialization of herds, the production of milk

During 1985 (and subsequent visits in 1987-88),

Cattle exploitation depended upon whether the animals

Local farmers tend to move their ovicaprine flocks

THE ORIGINS OF TRANSHUMANT PASTORALISM IN TEMPERATE SOUTHEASTERN EUROPE

The evolution of the large scale long distance