Fish Exploitation at the Sea of Galilee (Israel) by Early Fisher

Home , Acanthobrama, Capoeta damascina, HaOn, Kinneret Farm, Mallaha, Nubia

FISH EXPLOITATION AT THE SEA OF GALILEE (ISRAEL) BY EARLY FISHER-

HUNTER-GATHERERS (23,000 B.P.):

ECOLOGICAL, ECONOMICAL AND CULTURAL IMPLICATIONS

THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

Irit Zohar

SUBMITTED TO THE SENATE OF TEL-AVIV UNIVERSITY

November, 2003

FISH EXPLOITATION AT THE SEA OF GALILEE (ISRAEL) BY EARLY FISHER-

HUNTER-GATHERERS (23,000 B.P.):

ECOLOGICAL, ECONOMICAL AND CULTURAL IMPLICATIONS

THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

Irit Zohar

SUBMITTED TO THE SENATE OF TEL-AVIV UNIVERSITY

November, 2003

This work was carried out under the supervision of

Prof. Tamar Dayan and Prof. Israel Hershkovitz

TABLE OF CONTENTS

Page

CHAPTER 1: INTRODUCTION AND STATEMENT OF PURPOSE 1

1.1 Introduction 1

1.2 Cultural setting 2

1.3 Environmental setting 4

1.4 Outline of research objectives 5

CHAPTER 2: FISH TAPHONOMY 6

2.1 Introduction 6

2.2 Naturally deposited fish 7

2.3 Culturally deposited fish 9

CHAPTER 3: SITE SELECTION AND FIELD TECHNIQUES 11

3.1. The archaeological site of Ohalo-II 11

3.2. Fish natural accumulation 13

3.3 Ethnographic study of fish procurement methods 14

CHAPTER 4: METHODS 18

4.1 Recovery bias 18

4.2 Sampling bias 18

4.3 Identification of fish remains 19

4.4 Fish osteological characteristics 20

4.5 Quantification analysis 20

4.5.1 Taxonomic composition and diversity 21

4.5.2 Body part frequency 22

4.5.3 Survival index (SI) 22

4.5.4 Fragmentation index 23

4.5.5 WMI of fragmentation 24

4.5.6 Fish exploitation index 24

4.5.7 Bone modification 25

4.5.8 Bone spatial distribution 26

Page

4.5.9 Analytic calculations 26

4.6 Osteological measurements 29

4.6.1 Body mass estimation 29

4.6.2 Vertebrae diameter 31

CHAPTER 5: FISH REMAINS RECOVERED AT OHALO-II 32

5.1. Taxonomic identification 32

5.2 Skeletal representation 35

5.2.1 Skeletal completeness in brush hut 1 37

5.2.2 Skeletal completeness in Locus-7 42

5.2.3 Skeletal completeness in Locus 8 46

5.3 MNI value 50

5.4 Bone Color 51

5.5 Fragmentation pattern 53

5.5.1 Bone fragmentation in locus 1 54

5.5.2 Bone fragmentation in locus 7 56

5.5.3 Bone fragmentation in locus 8 58

5.6 Fish remains spatial distribution 59

5.6.1 Fish spatial distribution in locus 1 59

5.6.2 Fish spatial distribution in locus 7 61

5.7 Vertebrae dimensions 62

5.8 Body mass estimation 64

5.9 Dietary value 65

5.10 Summary 66

CHAPTER 6: FISH NATURAL ACCUMULATION 69

6.1 Bones spatial distribution 69

6.2 Taxonomic identification 69

6.3 Skeletal representation 72

6.4 Bone modification 80

Page

6.5 Vertebrae dimension 80

6.6 Body size estimation 81

6.7 Summary 81

CHAPTER 7: FISH BUTCHERING METHODS 83

7.1 Butchering and utilization methods 83

7.2 Skeletal representation 86

7.3 Bone fragmentation patterns 88

7.4 Fracture typology 91

7.5 Summary 93

CHAPTER 8: OHALO-II NATURAL OR CULTURAL ACCUMULATION? 95

8.1 Taxonomic composition, richness and diversity 98

8.2 Skeletal representation 102

8.2.1 Body part representation 103

8.2.2 Skeletal completeness 107

8.3 Bone modification 108

8.4 Vertebrae dimensions 113

8.5 Fish Body size 113

8.6 Bone distribution patterns 116

8.7 Summary 116

CHAPTER 9: DISCUSSION AND CONCLUSIONS 119

9.1 Environmental setting 119

9.2 Fish exploitation 121

9.3 Fish utilization 124

9.4 Fish exploitation in the context of Epi-paleolithic broad spectrum economy 127

9.5 Summary and conclusions 128

BIBLIOGRAPHY 131

LIST OF APPENDICES

Page

APPENDIX-I : Fish remains recovered from prehistoric sites and lacustrine

environments in Israel. 152

APPENDIX-II: Levantine freshwater fish 154

II.1 Morphological and osteological characteristics 154

II.2. Cyprinidae 158

II.3. Cichlidae 174

APPENDIX-III: Cichlidae skeletal elements in a complete fish. 178

APPENDIX-IV: Cyprinidae skeletal elements in a complete fish. 180

APPENDIX-V: A. terraesanctae skeletal elements in a complete fish. 182

APPENDIX-VI: C. gariepinus skeletal elements in a complete fish. 184

APPENDIX-VII: H. nitidus skeletal elements in a complete fish. 186

APPENDIX-VIII: C. multiradiatus and A. kessleri skeletal elements

in a complete fish. 188

APPENDIX-IX: C. caninus skeletal elements in a complete fish. 190

APPENDIX-X: Frequency (NISP) of skeletal elements for loci 2, 3, and 9. 192

APPENDIX-XI: Frequency (NISP) of skeletal elements for locus 1 by taxa. 194

APPENDIX-XII: Frequency (NISP) of skeletal elements for locus 7 by taxa. 196

APPENDIX-XIII: Frequency (NISP) of skeletal elements for locus 8 by taxa. 198

APPENDIX-XIV: Skeletal elements fragmentation pattern in locus 1. 200

APPENDIX-XV: WMI of fragmentation calculated by taxa for bones from locus 1. 202

APPENDIX-XVI: Skeletal elements fragmentation pattern in locus 7 (ashes). 205

APPENDIX-XVII: WMI of fragmentation calculated by taxa for bones from locus 7. 207

APPENDIX-XVIII: Skeletal elements fragmentation pattern in locus 8. 209

APPENDIX-XIX: WMI of fragmentation calculated by taxa for bones from locus 8. 211

APPENDIX-XX: Vertebrae width dimensions mean (±SD) and range

Calculated by taxa for locus 1. 212

Page

APPENDIX-XXI: Frequency (NISP) of skeletal elements for naturally

deposited fish along the Sea of Galilee. 215

APPENDIX-XXII: Vertebrae dimensions (height, width, and length)

mean (±SD) and range calculated by taxa for naturally deposited fish. 221

LIST OF TABLES

Table 1: Frequency (NISP) and percentage of fish remains by loci at Ohalo-II.

Table 2: Morphometrics for butchered fish collected from traditional fishermen in Panama

(Central America) and southern Sinai (Egypt) by butchering method and body size.

Table 3: A list of freshwater fish from the Sea of Galilee, Jordan River, and Nile river (n=

324), that were prepared for the osteological reference collection.

Table 4: A list of Red Sea fish from Eilat and Egypt (n=71) that were prepared for the

osteological reference collection.

Table 5: Number (NISP) and percentage of bones, by anatomic regions, expected in complete

skeleton of five taxa of freshwater fish.

Table 6: Regression equations for body mass (BM) in Cyprinidae as a function of atlas and

axis dimensions.

Table 7: Regression equations for standard length (SL) in Cyprinidae as a function of atlas

and axis dimension.

Table 8: Frequency (NISP) and percentage of fish identified at Ohalo-II by family and loci.

Table 9: Frequency (NISP) and percentage of fish remains identified by genus and loci.

Table 10: NISP values of identified bones, species richness, Shannon Wiener Function, and

Brillouin Index, calculated for each locus, at Ohalo-II.

Table 11: NISP and percentage calculated by loci in four taxonomic groups.

Table 12: Ranking order of taxonomic groups identified in different loci.

Table 13: Identified NISP and number of skeletal elements identified (richness) by loci.

Table 14: NISP and ranking order of Cyprinidae and Cichlidae cranial and postcranial bones

from locus 1.

Table 15: Frequency (NISP) and percentage of skeletal elements recovered from locus 1 for

anatomic regions and taxonomic groups.

Table 16: Survival index (SI) and p (calculated by chi-square test) for fish remains from locus

Table 17: Frequency (NISP), percentage, and SI calculated for cranial and postcranial bones

in four taxa at locus 1.

Table 18: Frequency (NISP) and percentage of skeletal elements recovered for anatomic

regions and taxonomic groups in locus 7.

Table 19: Survival index (SI) and p (calculated by chi-square test) for fish remains from locus

Table 20: Frequency (NISP), percentage, and SI calculated for cranial and postcranial bones

in four taxa at locus 7.

Table 21: Frequency (NISP) and percentage of skeletal elements by anatomic regions and

taxa in locus 8.

Table 22: Survival index (SI) and p (calculated by chi-square test) for fish remains from locus

Table 23: Frequency (NISP), percentage, and SI calculated for cranial and postcranial bones

for four taxa at locus 8.

Table 24: MNI values, by taxa and loci, for the identified fish remains.

Table 25: Comparison between ranking order calculated from NISP and MNI in five loci.

Table 26: Frequency (NISP) of bone colors by loci.

Table 27: Frequency (NISP) and percentage of bones state of fragmentation.

Table 28: Bones state of fragmentation by loci.

Table 29: An example of WMI calculated for Acanthobrama sp., by fragmentation classes.

Table 30: Ranking order by taxa of the best preserved bones (>80%) recovered from locus 1.

Table 31: Comparison between WMI and SI values in locus 1 by anatomic regions and taxa.

Table 32: Comparison between WMI and SI values in locus 7 by anatomic regions and taxa.

Table 33: WMI and SI values for locus 8 by anatomic regions and taxa.

Table 34: Scheffe post hoc tests between taxonomic groups and atlas dimensions (width,

length and height) in locus 1.

Table 35: Body mass (gr) and standard length (mm) estimated by taxa from loci 1 and 7.

Table 36: Estimation of fish dietary value from predicted mean body mass (BM) and MNI.

Table 37: Fish exploitation index by loci.

Table 38: Frequency (NISP) and percentage for naturally deposited fish remains by family.

Table 39: Frequency (NISP) for naturally deposited fish by family, depositional location, and

depth.

Table 40: Frequency (NISP) for naturally deposited fish by genus, depositional location, and

depth.

Table 41: NISP, species richness and Brillouin index calculated for naturally deposited fish.

Table 42: NISP and MNI calculated for naturally deposited fish by taxa and sampling area.

Table 43: NISP, standardized NISP, and richness values calculated for naturally deposited

fish in sampling areas.

Table 44: NISP of anatomic regions in random squares by taxa and depositional depth.

Table 45: NISP of anatomic regions in recent beach surface by taxa.

Table 46: NISP of anatomic regions in recent surface of Ohalo-II by taxa.

Table 47: Survival index (SI) calculated for naturally deposited fish by taxa and anatomic

regions.

Table 48: Observed and expected percentage and SI of cranial and postcranial bones in

naturally deposited fish (random squares) for four taxa.

Table 49 : Observed and expected percentage and SI of cranial and postcranial bones in

naturally deposited fish (recent shore) for five taxa.

Table 50: Mean state of fragmentation in naturally deposited fish by location and

sedimentation.

Table 51: Frequency (NISP) of bone color recorded in naturally deposited fish.

Table 52: Acanthobrama sp. estimated body mass (gr) and standard length (mm).

Table 53: Ratio of cranial to postcranial bones in butchered fish and the ratio expected in a

complete skeleton.

Table 54: Observed and expected NISP and their survival index (SI) for anatomic regions in

fish butchered by method-1.

Table 55: Observed and expected NISP and their survival index (SI) for anatomic regions in

fish butchered by method-2.

Table 56: The most commonly absent bones relative to the butchering method applied.

Table 57: The most frequently damaged bone according to butchering method.

Table 58: WMI of fragmentation calculated for highly damaged bones of fish butchered by

method-1.

Table 59: WMI of fragmentation calculated for highly damaged bones of fish butchered by

method-2.

Table 60: Types of fractures observed on the most frequently damaged bones of fish

butchered by method-1.

Table 61: Breakages typology for the frequently damaged bones by butchering method-2.

Table 62: Comparison between fish remains recovered in the natural accumulation and at the

various structures at Ohalo-II site.

Table 63: Statistics for the correspondence analysis plot outlined in figure 52.

Table 64: Stress factors and variance explained by MDS analyses of bone fragmentation

pattern for butchered fish, natural accumulation, and loci 1 and 7.

Table 65: Acanthobrama sp. estimated body mass (gr) and standard length (mm) for naturally

deposit fish and locus 1.

LIST OF FIGURES

Figure 1: Prehistoric sites in Israel from which fish remains were recovered.

Figure 2: Map showing the maximum aerial coverage of Lake-Lisan during its highstand of

180 m BSL, location of current lakes, and Ohalo-II.

Figure 3: Ohalo-II site covered with water (on left) and its exposure at -214.0 BSL in the

summer of 2000.

Figure 4: Excavated loci at Ohalo-II where fish remains were recovered.

Figure 5: Position of 24 random squares sampled along the southern shore of the Sea of

Galilee.

Figure 6: An excavated square from the Sea of Galilee present shoreline (observe the changes

from upper sandy layer to a dark clay layer at the bottom).

Figure 7: Map of Panama and Parita Bay showing location of the studied sites.

Figure 8: Fish drying and salting seasonal camp at the mouth of Rio Santa Maria (on left),

and fish processed for long-term preservation by Francisco at Partita Bay (on

right).

Figure 9: Map of Sinai, with the location of the studied site (Nabek Oasis) indicated.

Figure 10: Fish butchering by a Bedouin family in Nabek Oasis (Sinai).

Figure 11: A generalized fish skeleton presenting selected cranial and postcranial bones.

Figure 12: Fragmentation classes used for classification of bone state of preservation.

Figure 13: Measurements performed on vertebra centrum.

Figure 14: An hypothetical histogram of vertebra width normal distribution expected from

Acanthobrama sp. and large cyprinids (Barbus sp. and Capoeta sp.).

Figure 15: Relative abundance (%) of Cichlidae and Cyprinidae by loci.

Figure 16: Rarefaction curves for species richness and loci as a function of NISP.

Figure 17: Rarefaction curve for number of skeletal elements identified and loci as a function

of NISP.

Figure 18: Number of skeletal elements identified in locus 1 by taxa.

Figure 19: Observed and expected percent of anatomic regions in Acanthobrama sp. at locus

Figure 20: Observed and expected percent of anatomic regions in Barbus sp./Capoeta sp. at

locus 1.

Figure 21: Observed and expected percent of cranial and postcranial bones in Acanthobrama

sp., large cyprinids, and cichlids at locus 1.

Figure 22: Number of skeletal elements identified for locus 7 (ashes) by taxa.

Figure 23: Observed and expected percent of anatomic regions in Barbus sp., Capoeta sp. and

small cyprinids at locus 7.

Figure 24: Observed and expected percent of anatomic regions in Cichlidae in locus 7.

Figure 25: Observed and expected percent of cranial and postcranial bones in Acanthobrama

sp., large cyprinids, and cichlids at locus 7.

Figure 26: Comparison between number of skeletal elements per taxa in locus 8.

Figure 27: Relative abundance (%) of anatomical regions in complete Cyprinidae compared

with those observed for large and small cyprinids remains in locus 8.

Figure 28: Relative abundance (%) of anatomical regions in complete Cichlidae compared

with those observed in locus 8.

Figure 29: Observed and expected percent of cranial and postcranial bones in large cyprinid

and cichlid at locus 8.

Figure 30: Bones fragmentation patterns for four taxa in locus 1.

Figure 31: Bones fragmentation patterns for four taxa in locus 7.

Figure 32: Bones fragmentation patterns for four taxa in locus 8.

Figure 33: Spatial distribution of fish remains in locus 1.

Figure 34: Spatial distribution of Acanthobrama sp., (on left), of large carps (on right top left

no.) and Cichlidae (on right the right no.) in locus 1.

Figure 35: Spatial distribution of fish remains in locus 7.

Figure 36: Atlas mean width by loci and taxa.

Figure 37: Frequency distribution (NISP) of Cyprinidae caudal vertebrae width in Locus 1.

Figure 38: Estimated standard length (mm) and body mass (gr) of Acanthobrama sp. in locus

Figure 39: Estimated body sizes of Barbus sp./ Capoeta sp. from loci 1 and 7.

Figure 40: Spatial distribution of naturally deposited fish remains.

Figure 41: Comparison between skeletal elements richness by taxa and depositional depth in

the random squares.

Figure 42: Vertebrae mean width diameter (± SD) for fish natural accumulation by taxa.

Figure 43: Flow chart for fish butchering methods observed in Parita-Bay (Panama) and south

Sinai (Egypt).

Figure 44: Fish butchered by method-1 (left) and by method-2 (right).

Figure 45: Standard length frequency distribution of 573 fish belonging to 34 species

butchered by the two different techniques.

Figure 46: Multidimensional scaling (MDS) plot of SI of bones showing a separation between

butchering methods, regardless of fish taxonomy and anatomy.

Figure 47: Typical fractures observed on the cleithrum and coracoid of fish butchered by

method-1.

Figure 48: Typical fractures observed on the cleithrum and coracoid from fish butchered by

method-2.

Figure 49: Typical fractures observed on cranial bones situated along the longitudinal

transverse cut of catfish butchered by method-2.

Figure 50: Rarefaction curves of species richness in loci 1, 7, and the natural accumulation as

a function of NISP.

Figure 51: Taxonomic groups percentage (%) in the natural accumulation and Ohalo II.

Figure 52: Correspondence analysis of taxonomic groups relative abundance (%) in the

natural accumulation and loci 1, 2, 3, 7 and 8.

Figure 53: Survival index, (SI) by anatomical regions, for loci 1, 7, and 8 and the clay

deposits of the natural accumulation.

Figure 54: Frequency of cranial and post-cranial bones in Acanthobrama sp. and small

cyprinids recovered from the natural accumulation and Ohalo-II.

Figure 55: Frequency of cranial and postcranial bones in large Cyprinidae, and Cichlidae

recovered from the natural accumulation and Ohalo-II.

Figure 56: Percentage of burned and brown bones in the natural accumulation and various

loci.

Figure 57: MDS analysis plot for fish from the natural accumulation on the left and butchered

fish on the right.

Figure 58: MDS analysis plot for bone breakage pattern in the natural accumulation (blue)

and locus 1 (black).

Figure 59: MDS analysis plot for bone breakage pattern of fish remains in the natural

accumulation (blue) and locus 7 (black).

Figure 60: Comparison between Acanthobrama sp. atlas and axis mean width (mm, ±SD)

from recent reference collection, natural accumulation, and locus 1.

Figure 61: Acanthobrama sp. estimated SL from the natural accumulation vs. locus 1.

Figure 62: Plot of fish index based on aggregated NISP by locality at Ohalo-II and the natural

accumulation.

ABSTRACT

Fishing is an important economic aspect of many societies throughout the world today and has played a significant role in the life and subsistence of many prehistoric societies. The physical environment of Israel, surrounded by the Mediterranean Sea, the Red Sea, and the

Jordan rift system, undoubtedly could have contributed to the development of fishing communities. Despite the appearance of fish remains in many prehistoric sites in Israel, most archaeofaunal researchers focused on large and small game as markers of economical and cultural changes, ignoring the fish.

The water-logged site of Ohalo-II recovered at the southern shore of the Sea of Galilee, provides, at present, the earliest evidence of fish-based economy. This is a terminal Upper-

Paleolithic/early Epi-paleolithic site dated to ca. 23,000 cal B.P., with exceptional preservation of several brush huts (loci 1, 2, 3), ashes (loci 7,9), human grave, stone installations, pits (locus 8), flint tools, stone weights, botanical and faunal remains. Until the recovery of Ohalo-II, evidence for fish exploitation in the Epi-paleolithic was recorded only from the Natufian site of Mallaha (Eynan; ca.12,000 B.P.). Analysis of fish remains recovered at Ohalo-II provides an outstanding opportunity to study fish exploitation by early hunter-gatherers (23,000 cal B.P.) during the last glacial maximum (LGM).

The goals of this study were sixfold: 1) to reconstruct the paleoecology of the Sea of

Galilee; 2) to investigate fish exploitation at the Paleo-Sea of Galilee, during the LGM; 3) to shed light on fishing and fish utilization techniques practiced by Epi-paleolithic hunter- gatherers during the LGM; 4) to estimate the role fish played in the diet of Ohalo-II inhabitants; 5) to correlate between fish remains and human activities at the site; and 6) to develop a taphonomic model that provides criteria to distinguish culturally from naturally deposited fish.

My study included 44,000 fish bones from three accumulations: archaeological (Ohalo-

II), natural and ethnographical. From Ohalo-II, I studied fish remains (19,799 bones) from

loci 1, 2, 3, 7, 8, and 9. The naturally deposited fish (5,968 bones) were recovered from three depositional layers (upper sand, median brown sand, and bottom clay) recovered from the present southern shore of the Sea of Galilee. The ethnographic study included 17,862 bones from 147 fish butchered for drying and salting by modern fishermen in Panama and South

Sinai (Egypt). The fish were butchered by two methods: along the back with the skull split or along their belly with intact skull.

Fish were identified to the lowest taxonomic level possible with a reference collection housed at Tel-Aviv University, Israel; at the Royal museum of Africa in Tervuren, Belgium; and at the Smithsonian Tropical Research Institute in Panama. Several qualitative and quantitative criteria were used: taxonomic composition, breadth, richness, and diversity; representation and completeness; bone modification and dispersion pattern; vertebrae dimension and estimated fish body size. I here present the main results of my research, according to environmental, taphonomical and cultural aspects.

Environmental aspect: Analyses of 19,799 fish remains from Ohalo-II showed the presence of 8 species from two families of freshwater fish: Cichlidae (St. Peter fish) and

Cyprinidae (Carp). The taxa identified resembled the present day fish in the Sea of Galilee.

Moreover, two of the identified taxa were endemic to the Sea of Galilee: Tristamella sp.

(Tristram's St. Peter fish) and Acanthobrama terraesanctae (Kinneret Bleak). A. terraesanctae is a primary freshwater fish and its appearance in Ohalo-II attests that despite the climatological and geological changes salinity level did not change abruptly and the

Paleo-Sea of Galilee/ Lake Lisan was a freshwater lake similar with the present lake.

Taphonomical aspect: My research demonstrated that in lacustrine sites, such as Ohalo-

II, we can not assume, a priori, that all fish remains have resulted from human activity.

Moreover a model of fish natural accumulation must be developed for each depositional area.

Comparative analysis indicated that taxonomic breadth, richness and diversity, fish index, and skeletal completeness vary between the natural and cultural accumulation. Cultural

deposits were characterized by a wider taxonomic breadth, higher species representation of

Barbus sp., Capoeta sp. Tilapia sp., and Tristamella sp.. Catfish remains (Clarias gariepinus) were absent from Ohalo-II and appeared in low frequency for the natural accumulation. Moreover A. terraesanctae remains were highly abundant in the natural accumulation and loci 1 and 7.

At the natural accumulation fish bones (NISP) increased with depositional depth, reaching their peak in the deepest clay deposits. However, species diversity (Brillouin index) decreased with depositional depth, exhibiting a high value for the upper sand. Differential preservation was observed in the different layers. For example, scales were abundant at the natural accumulation, only in the upper sand and brown layers. However, they were absent from the clay deposits. This preservation bias may explain their absence from Ohalo-II clay deposits. Otoliths did not survive in natural accumulation, and appeared in low frequencies at Ohalo-II site, mainly in locus-8. In the natural accumulation the cranial region was over- represented for most taxa, while in Ohalo-II the cranial region was under-represented for all species except for Acanthobrama sp. remains from locus 1. The branchial region was under- represented in the natural and the cultural accumulation, and could not be used as marker for fish gutting. Interestingly, bone high scatter frequency (BSF), clumped distribution, and fragmentation did not vary between the natural and cultural accumulations.

Cultural aspect: My analysis has shown that Ohalo-II inhabitants exploited Barbus sp.,

Capoeta sp., Tilapia sp., and Tristamella sp. (MNI=342). The fish dietary contribution to the inhabitants' daily diet was larger than all other faunal groups. From a small sample I estimated that large cyprinids, contributed at least 22 kg of fish for the inhabitants of locus-1, and 18 kg in locus-7.

Species diversity and the wide range of fish body sizes indicated that Ohalo-II inhabitants used non-selective fishing techniques such as weirs, baskets, and nets. Such activity could have taken place in the riverine and littoral zones during the fish breeding

season, or in the pelagic zone. However, the present data was insufficient to support determination of seasonality or fishing area and technology. The absence of Clarias sp.

(catfish) may result from survival bias, as observed at the adjacent natural accumulation.

Other possibilities were that it either indicated a lack of preferable environmental habitats, or human exploitation patterns and culinary habits.

Remains of A. terraesanctae and small cyprinids were unique for loci 1 and 7, and resembled the adjacent natural accumulation. If the inhabitants of Ohalo-II did targeted A. terraesanctae then this is the earliest evidence for small fish mass harvesting, which would have required the use of new technologies. However, this economic trend differed from the one observed in other loci examined, and appeared only in later periods. Therefore, the present data did not provide sufficient evidence to support mass exploitation of A. terraesanctae by Ohalo-II inhabitants, but rather support natural accumulation.

Another aspect of my research examined skeletal element representation and fragmentation in comparison with butchering methods by present day fishermen. My ethnographic study demonstrated that in modern-day Panama fish were butchered differently depending on their body size. Skeletal element representation and fragmentation pattern from

Panama differed from the data obtained from Ohalo-II. This might have been due to different butchering methods or due to the relatively small sample size of "large fish" obtained from loci 1, 2 ,3 and 8.

At Ohalo-II, large cyprinid and cichlid cranium region were under-represented in all excavated loci. However, the relative abundance of the cranium region varies between structures. These differences may have resulted from differential preparation and consumption methods applied according to fish taxa. The relatively high ratio of cranial remains in locus-8 may be the earliest evidence, in Israel, for fish preservation and storage.

In sum, the large number of fish remains recovered at Ohalo-II indicate that fishing activity played an important role in the inhabitants' daily life and diet, 23,000 years B.P. In

the absence of direct evidence for deep sea fishing, it is apparent that nearshore fishing was a fundamental and optimal strategy used by the inhabitants, providing a stable food resource.

The large amount of fish catch was probably processed for later consumption, and provided economic stability. From the ecological point of view, the composition of freshwater fish

23,000 years ago was found to be very similar to the present one. This clearly indicates that the Paleo Sea of Galilee/Lake Lisan was at 23,000 B.P. (and probably much earlier) already a fresh-water lake.

ACKNOWLEDGMENTS

This research was born from the fascination and pleasure of entering the world of fish, and of past human populations. The study could undoubtedly never have been accomplished without the tremendous help and support that I received from many friends, researchers, institutions and organizations in Israel and around the world, and I hope that I have remembered to thank everyone.

First, I would like to thank my supervisors, Prof. Tamar Dayan and Prof. Israel

Hershkovitz, who agreed to dive with me into this ichtyological adventure, and who supported me throughout the long and hard process. Identification of the freshwater fish would have not been possible without the enormous help and hospitality of Dr. Wim Van

Neer and the Royal Museum for Central Africa, in Tervuren, Belgium.

Dr. Dani Nadel from the Institute of Archaeology, University of Haifa, is responsible for the excavation of Ohalo-II, and I would like to express my gratitude for his enormous work and effort. I am grateful too to all the students and volunteers who spent time in the field and in the lab, digging, sieving and picking the material from the site.

From my very first step into the amazing world of fish remains, I have been constantly supported and mentored by Prof. Ehud Spanier from the Department of Maritime

Civilizations and the Leon Recanati Institute for Maritime Studies, at the University of Haifa.

My research could not have been completed without the doctoral fellowships I received from the University of Haifa and the Jacob Recanati fellowship from the Center of Maritime

Studies.

I also wish to thank all the wonderful staff of the Department of Maritime Civilizations and the Center of Maritime Studies, especially Prof. Avner Raban, Prof. Michal Artzy, Prof.

Yossi Mart, Dr. Ezra Marcus, Dr. Dorit Sivan, Ada Vulkan, Nira Karmon, and Yossi Tur-

Caspa for their help and support.

Several institutions and organizations provided me with research grants: the Maria

Rossi Ascoli Fellowship, the Irene Levi Sala CARE Archaeological Foundation, the

Smithsonian Institution-Washington, the National Center for Collaboration between Natural

Sciences and Archaeology, Weizmann Institute of Science, the Morris M. Pulver Fellowship, and the Aharon Katzir Center, Weizmann Institute of Science.

Dr. Richard Cooke and the Smithsonian Tropical Research Institute provided me with the possibility and fascinating experience of working in tropical Panama. I gratefully acknowledge assistance in the field from Conrado Tapia, Jose Tapia, Gonzalo Tapia, and the various fishing folk from Aguadulce and Parita who kindly let us study their work.

I would like to thank the Department of Zoology, the I.Meier Segals Garden for

Zoological Research, the Zoological Museum, the Institute for Nature Conservation Research, and the Department of Anatomy at Tel-Aviv University, for providing me with the facilities to conduct my research and build my reference collection.

My fish reference collection could have not been established without the help of Dr.

Menachem Goren and Prof. Avital Gasith, from Tel-Aviv University, Oren Sonin from the

Israel Department of Fisheries, Guy Ayalon and Aharon Meroz from Coral World Eilat. I would also like to thank my dear friends and colleagues Dr. Eli Geffen, Sigal Sheffer, Inbal

Ayalon, Shirley Cohen-Gross, and Dr. Wim Van Neer, who all kindly traveled with Egyptian fish in their luggage, to help me expand my reference collection.

My particular thanks go to Prof. Naama Goren-Inbar, from the Institute of Archaeology,

Hebrew-University, Jerusalem, who introduced me to the fascinating world of prehistoric communities, and supported, assisted and encouraged me throughout. Prof. Goren-Inbar and the members of my Ph.D. committee Dr. Menahem Goren, Prof. David Wool and Prof. Avner

Bdolach, provided help and support that significantly improved my research.

I am also grateful to the late Prof. Eitan Tchernov at the Department of Evolution,

Systematics and Ecology, The Hebrew University of Jerusalem, Givat Ram. Prof. Tchernov

and his wonderful group generously offered me their help, support and facilities. I am especially grateful to Miriam Belmaker who agreed to collaborate with me on the study of natural accumulation of fish.

My special thanks to Prof. Steve Weiner and Dr. Elisabetta Boaretto, from the Weizmann

Institute of Science, for their help with FTIR analysis and radiocarbon dating; to Rachel Paz and

Dr.

Sarig Gafni, from the Institute for Nature Conservation Research at Tel-Aviv University, for their help with every detail; to Josh Peabody and Naomi Paz for their editorial help; to Prof. Virginia

Butler, for her advice; and to Dr. Ruby Cerron-Carasco for contributing important data to my study.

I am also grateful to Prof. Ofer Bar-Yosef, from Harvard University, for his help, support and encouragement. My gratitude to Prof. Baruch Arensburg, Prof. Yoel Rak,

Avishag Ginsburg, and Dr. Sue Wish-Baratz from the Department of Anatomy, Tel Aviv

University, for their advice and support. Dr. Shmulik Marco from the Department of

Geophysics, Tel-Aviv University, and Alexander Tsatskin, from the Institute of Archaeology,

University of Haifa, patiently advised on the site geology. I would like to thank my friends and colleagues in Prof. Dayan's lab for their support and encouragement. Asher Pinhasov, from the Sackler School of Medicine, Tel-Aviv University, spent many days photographing fish bones. The good results are due to his patient hard work. Prof. Diane Gifford Gonzales and the Department of Anthropology at University of California Santa-Cruz provided me with their hospitality and support during the final stage of writing this dissertation.

Ultimately, this work would have never reached its final stage without the support, help and encouragement of my loving family. My father, Yuval, introduced me into the

fascinating world of fish, and together with my brother, Avi, followed and encouraged me from my first steps at the university. This work too would have never been completed without the love and support provided by my dear mother, Miriam, who sadly didn't live to see it finished.

My husband, Eli Geffen, patiently supported, encouraged, helped and assisted me in my

research in Sinai, as well as during the

archaeological and natural accumulation

excavations. He also spent many hours in

tedious and complicated mathematical

calculations of my data. Finally, all my love

to my daughter Orr, who lit up my days and

nights, and joined me in the field from her own very first steps.

ניצול דגי הכנרת על ידי אוכלוסית דייגים-ציידים-לקטים קדומה ©B.P 23,000.¨÷ מאפיינים סביבתיים¬ כלכליים¬ טכנולוגיים ותרבותיים

חיבור לשם קבלת התואר ׃דוקטור לפילוסופיה׃ מאת עירית זהר

הוגש לסנאט אוניברסיטת תל-אביב נובמבר 2003

חיבור לשם קבלת התואר ׃דוקטור לפילוסופיה׃ מאת עירית זהר

הוגש לסנאט אוניברסיטת תל-אביב נובמבר 2003

עבודה זו נעשתה בהדרכת פרופװ תמר דיין ופרופװ ישראל הרשקוביץ

׃ורדו בדגת הים®®®¢ ©בראשית א¬ 28¨ תקציר÷ אחת השאלות המענינות בהתפתחות של אוכלוסיות אנושיות הינה מתי החל ניצול של משאבים י מיים ופיתוח של טכנולוגיות דיג® בארץ ישראל יש פוטנציאל רב להתפתחות של אוכלוסיות ימיות¬ מאחר והיא מוקפת בגופי מים ©ים תיכון¬ ים סוף¬ הכנרת ועמק הירדן¨® למרות שמ רבית האתרים הפרהיסטורים ממוקמים בסמוך למקורות מים ונמצאו בהם שרידי דגים¬ מרבית ה מחקרים הפאוניסטים התמקדו בשרידי יונקים ורק מקצתם מציינים נוכחות של עצמות דגים. הדבר מפתיע מאחר ולימוד של שרידי הדגים יכול לספק עדות ישירה ועקיפה על עונת הישיבה באתר¬ על שיטות הדיג¬ על אזורי הדיג ©מים רדודים או עמוקים¬ מים מתוקים או מלוחים¨ ¬ על שיטות עיבוד הדגים לצריכה מידית ועל היכולת לשמר דגים לפרק זמן ארוך® בנוסף¬ משרידי הדגים אנו למדים על האקולוגיה של הסביבה הימית ©טמפרטורה¬ מליחות¬ זרמים¬ סד ימנטים¨ ועל תנאי האקלים שהתקיימו בעבר® הזדמנות ללמוד על תרומת הדגים לאוכלוסית צידים-לקטים קדומה נוצרה בעקבות ירידת מפלס המים בכנרת® בחופה הדרומי התגלו שרידים של אתר ארכיאולוגי המתוארך ללפני כ23,000- שנה-אוהלו II® זהו אתר מסוף הפליאולית העליון/ ראשית האפיפליאולית © Last Glacial Maximum¨ המהווה¬ נכון להיום¬ את העדות הקדומה ביותר לאוכלוסית דייגים-ציידים-לקטים ® אוהלו II הינו אתר גדול יחסית ©>1500 sqm¨ בהשוואה לאתרים אחרים בני זמנו® בנוסף התגלו באתר שרידים של בקתות © ,Loci 1 3 ,2¨¬ מוקדים ©Loci 7,9¨¬ קבורה אנושית¬ מתקנים © Locus 8¨¬ כלי אבן וצור ©תעשיה מיקרוליתית¨¬ משקולות אבן¬ שרידי חבלים¬ וממצא בוטאני ופאו ניסטי עשיר במצב שימור טוב® בעבודה זו נעשה לראשונה ניסיון לתעד את חברת הדייגים שישבה לחוף הכנרת הקדומה ©אוהל ו- (II® מחקר זה התבצע בארבעה רבדים הקשורים זה בזה÷ סביבתי¬ כלכלי¬ טכנולוגי ותרבותי ® א® רובד סביבתי ©פליאואקולוגי¨÷ זיהוי של מגוון המינים שהיו בכנרת לפני כ23,000- שנ ה עוזר לשחזור התנאים האקולוגים ששררו באגם® לאגם הכנרת קדמה ימה מלוחה- ימת הליסן ¬ אשר יש הטוענים כי המשיכה להתקיים עד לפני כ13,000- שנה® התנאים האקולוגים ששררו בימת הליסן אינם מתאימים לדגים של למים מתוקים © primary fresh water fish¨ אלא למינים אשר יכולים לשרוד במים מלוחים ובמערכת אקולוגית בילתי יציבה® לפי כך¬ מגוון המינים שזוהו באוהלו-II מהווים עדות לאופי בית הגידול ששרר באזור לפני 23,000 שנה® ב® רובד טאפונומי÷ שיחזור האופן שבו ניצלו תושבי אוהלו II דגים נעשה על ידי ניתוח ט פונומי מקיף של דגם פיזור העצמות באזורים שונים באתר והשוואתם לעצמות דגים מתמותה ט בעית. ג® רובד כלכלי÷ ממגוון ועושר מיני הדגים¬ גודלם¬ משקלם וכמותם היחסית ©NISP¨ באתר ח ושבה התרומה היחסית ©בקלוריות¨ של הדגה לסל המזון של אוכלוסית אוהלו® ד® רובד טכנולוגי÷ ±® ממגוון מיני הדגים וגודלם שוחזרו אזורי הדיג ©מים רדודים או ע מוקים¬ כנרת או נחלים¨ ושיטות הדיג® ²® ממחקר אתנוגרפי של אוכלוסיות דייגים בסיני ובפנמה אציג מודל לזיהוי שיטות לעיבוד דגים ושימורם® מחקר זה נערך על 44,000 עצמות דגים מהאתר הארכיאולוגי של אוהלו II¬ ממאסף של תמותה טבעית בחוף הכנרת וממחקר אתנוגרפי בחברות דייגים מסורתיות בנות זמננו® מאוהלו II נ דגמו 19,799 עצמות שנמצאו בששה אזורים שונים © ,loci1, 2, 3, 7, 8 9¨® מחוף הכנרת נדגמו 5,968 עצמות דגים שמתו תמותה טבעית והורבדו בשלוש שכבות ©שכב ה חולית עליונה¬ שכבת חול וחרסית אמצעית¬ ושכבת חרסית תחתונה המתוארכת ללפני 1,300 שנה¨® המחקר האתנוגרפי התבצע על 147 דגים ©17,862 עצמות¨ שנחתכו בשיטות שונות¬ על ידי אוכלוסית דייגים עכשווית בפנמה ובסיני¬ בתהליך הכנה לייבוש והמלחה® זיהוי עצמות הדגים לפי מיקומן בשלד הדג לרמת המשפחה¬ הסוג והמין¬ התבצע בעזרת אוסף אוסטיאולוגי משווה מדגים עכשויים אשר אספתי מהכנרת ומנחלים בעמק הירדן® בנוסף¬ השת משתי באוספים משווים הנמצאים בבלגיה © ,The Royal Museum of Africa Tervuren¨ ובפנמה © Smithsonian Tropical Research Institute¨ עיבוד הממצא נעשה בעזרת מספר קריטריונים ובהם÷ הרכב טאקסונומי¬ מגוון ו עושר מינים¬ שכיחות של עצמות¬ מצב השתמרות¬ דגם פיזור בשטח¬ מימדי החוליות ושחזור ג ודל הדגים® תוצאות עיקריות של מחקרי מתייחסות לפליאואקולוגיה¬ הטאפונומיה¬ הכלכלה ו החברה שהתקיימו באוהלו-II® שחזור פליאואקולוגי÷ באוהלו II זוהו שמונה מינים של דגי מים מתוקים הכוללים את÷ לבנ ון הכנרת¬ חפף ישראלי¬ בינית גדולת קשקש ובינית גדולת ראש¬ ממשפחת הקרפיוניים¬ ואמנ ון מצוי¬ אמנון הירדן¬ אמנון הגליל וטברנון ממשפחת האמנוניים® מהמינים המזוהים¬ לב נון הכנרת וטברנון הינם דגים אנדמיים לכנרת והופעתם באוהלו II מעידה על דימיון לחבר ת הדגים הקיימת בכנרת בהווה® בנוסף¬ שכיחותם הגבוהה של הקרפיוניים ©81%¨ ובעיקר של לבנון הכנרת ©15%¨ מעידה שמי הכנרת הקדומה¯אגם הליסאן היו ברמת מליחות דומה לזו כי ום ולא חלו שינויים חדים ברמת המליחות ב23,000- השנים האחרונות® שחזור טאפונומי÷ הבנת דגם ניצול הדגים באוהלו נעשתה לאחר השוואת הממצא לתמותה הטבעי ת בכינרת® תוצאות המחקר הטאפונומי מראות שבאתר חופי או טבוע קיימת אפשרות שחלק משר ידי הדגים היו תוצאה של תמותה טבעית® השוואה בין אזור התמותה הטבעית לבין אוהלו-II מראה הבדלים במגוון ועושר מיני הדגים¬ ביחס בין מיני דגים גדולים למינים קטנים © Fish index¨¬ ובדגם שימור העצמות® בחוף הכנרת מצאתי שהתמותה הטבעית מאופיינת במגוון מינ ים נמוך¬ שכיחות גבוהה של לבנון הכנרת בשכבת החרסית התחתונה ושכיחות נמוכה של קרפיו ניים ׃גדולים¢ ©חפף ובינית¨ ואמנוניים® שפמנוניים הופיעו בתמותה הטבעית בשכיחות נמ וכה ורק בשכבת פני השטח העליונה® השוואה בין שרידי הדגים שנמצאו בתמותה הטבעית לבי ן אלו מאוהלו-II מצביעה על דימיון בדגם פיזור העצמות ©מקובץ¨¬ בריכוזיותם הגבוהה בש טח ובשכיחות הנמוכה של אזור הזימים® בנוסף¬ מצאתי דימיון רב בשכיחות הגבוהה של לבנ ון הכנרת והופעה של מרבית חלקי השלד בתמותה הטבעית ובלוקוסים ± ו-· באוהלו-II® דימ יון זה מצביע ששרידי הלבנונים באוהלו הינם תוצאה של תמותה טבעית ולא פעילות אנושית® לעומת זאת חפפים¬ ביניות ©׃קרפיוניים גדולים¢¨ ואמנוניים מופיעים בשכיחות נמוכה ב תמותה הטבעית ולכן שכיחותם הגבוהה באוהלו-II מעידה על פעילות אנושית בלבד® רובד כלכלי וטכנולוגי÷ אוכלוסית אוהלו II נצלה לפחות 342 דגים ©חפפים¬ ביניום ואמנו ניים¨® דגימה הראתה שבלוקוס ± תרמו הקרפיוניים הגדולים למעלה מ²²- ק¢ג בשר ובלוקוס ³ למעלה מ18- ק¢ג® ® מגוון מיני הדגים וטווח מימדי גופם מצביע על שימוש בשיטות ד יג לא בררניות כמו רשתות¬ סלים ומלכודות® פעילות הדיג יכלה להתבצע בנחלים בתקופת ה רביה של הדגים¬ או באגם הפתוח® השרידים הרבים של לבנון הכנרת ©סרדין הכנרת¨ שנמצאו בלוקוסים ± ו-· מראים על דימיון רב לדגם שנמצא בתמותה הטבעית® אם דגים אלו כן נתפסו על ידי אוכלוסית אוהלו II¬ הר י שזוהי העדות הקדומה ביותר¬ בישראל¬ לדיג של דגים קטנים ©אורכם המקסימאלי 20 ס¢מ¨® אם תושבי אוהלו אכן פיתחו את השיטות והמיומנות לתפיסת להקות דגים¬ אזי עולה השאלה מדוע עדות לפעילות כזו אינה מופיעה בכל האתר ובנוסף¬ מדוע אינה ממשיכה בהיסטוריה ה אנושית אלא מופיעה רק אלפי שנה מאוחר יותר¿ על חשיבות הדגים למערכת הכלכלית של תושבי האתר ניתן ללמוד מדגם שכיחות חלקי השלד בא והלו-II ©למעט לוקוס ¸¨® באוהלו-II מצאתי שכיחות נמוכה של עצמות הגולגולת ©crania¨ למול ייצוג יתר של עצמות פוסט קרניאליות¬ בייחוד חוליות® תופעה זו יכולה לנבוע מת נאי שימור בקרקע או¬ משיטות חיתוך¬ בישול¬ וצריכת הדגים® שכיחות גבוהה יחסית של עצ מות גולגולת ואוטוליטים נמצאה בלוקוס ¸ ומעידה שעצמות הגולגולת יכולות להשתמר באתר® מכאן שייצוג נמוך של עצמות הגולגולת נובע משיטות בישול וצריכת הדגים® מחקר אתנוג רפי שביצעתי על שיטות חיתוך הדגים על ידי אוכלוסיות בנות זמננו בפנמה ובסיני הראה ש בפנמה גודל גוף הדג משפיע על שיטת החיתוך שלו® הייצוג הגבוה של עצמות גולגולת בלוק וס ¸ מעלה את האפשרות שזהו מתקן לשימור ואיחסון דגים® אם סברה זו נכונה אזי לפנינו העדות הקדומה ביותר¬ בישראל¬ לשימור ואיחסון דגים® פעילות כזו סיפקה יציבות כלכלי ת לתושבי אוהלו-II® לימוד של שרידי הדגים מאוהלו II מציג היסטוריה טאפונומית מורכבת המכילה מצד אחד סדי מנים אגמיים ובהם עצמות דגים קטנים ©בעיקר לבנונים¨¬ תוצר תמותה טבעית באגם הקדום¬ ומצד שני דגים גדולים¬ לפעמים בארטיקולציה¬ המעידים על פעילות דיג אינטנסיבית של תו שבי האתר® בנוסף נמצאו הבדלים באזורי פעילות שונים באתר המלמדים על הבדלים בשיטות צריכת הדגים ועיבודם¬ קרי÷ צריכה מידית למול שימור ואגירה לטווח זמן ארוך® לסיכום¬ ממצאי העבודה מראים¬ ללא ספק¬ שהאדם הפרהיסטורי בסוף תקופת הפליאולית העליו ן ידע לנצל וניצל משאבי מזון הקשורים במקןרות מים. הוא לא עשה זאת באופן מיקרי ©מז דמן¨ אלא בצורה מתוכננת ואינטנסיבית. הוא לא יכול היה לבצע את פעילות הדיג הזו ולה פיק ממנה את מירב התועלת הכלכלית ללא הבנה באקולוגיה של גוף המים ובטכניקות דיג¬ עי בוד ושימור דגים. המיקר הנוכחי מוכיח גם שהתנאים באגם הכנרת )אגם הלשון הצפוני¨ בסוף הפליאולית העליון היו דומים לאלה הקיימים בו היום. CHAPTER 1: INTRODUCTION AND STATEMENT OF PURPOSE "fish have the greatest potential for making subsistence economy more secure and more abundant" (Hayden et al., 1987) 1.1 Introduction Fishing is an important economic aspect of many societies throughout the world today and has played a significant role in the life and subsistence of many prehistoric societies (Belcher, 1998; Yesner, 1980). The physical environment of Israel, surrounded by the Mediterranean Sea, the Red Sea, and the Jordan rift system, undoubtedly could have contributed to the development of fishing communities. The antiquity of fish exploitation may be detected from the early association between site selection and aquatic habitats (Figure 1), as well as from fish remains recovered at the sites (Appendix I). Figure 1: Prehistoric sites in Israel from which fish remains were recovered.

Although fish remains appear in archaeological sites since the lower Paleolithic Mediterranean Sea

Eynan Hula Basin (Appendix I), little research was performed on Hayonim Cave Gesher Benot-Ya'aqov Haifa Kefar Ahoresh Sea of Galilee fish exploitation and fishing in prehistoric Atlit-Yam Tabun Cave Ein-Gev I-IV Neve-Yam Kebara Cave Haon I-III Ubeidiya Ohalo-II societies. Surprisingly, most archaeofaunal researchers focused on ungulates as markers of economic and cultural changes, ignoring the fish Tel-Aviv Jordan Rift V alley (Bar-Oz et al., 1999; Bar-Yosef & Belfer-Cohen,

Jerusalem 1992; Haas, 1966; Tchernov, 1979; Tchernov, Hatula Netiv Hagdod 1981; Tchernov, 1988). The scarcity of studies of

Dead Sea fish remains caused their significance to go unrecognized, and the value of fishing was

0 5 0 k m questioned. In the last few years researchers have emphasized the importance of fish to human diet and to economic stability (Stewart, 1989; Van Neer, 1989; Yesner, 1980). These studies indicate that the exploitation of marine resources was gradually developed beginning with easy to collect littoral food such as shellfish and littoral fish, and only later on when more sophisticated technology were developed, pelagic species were exploited (Kelly, 1996; Lyman, 1991a; Stewart, 1989; Yesner, 1980). Ethnographic studies demonstrated that a wide range of aquatic resources can be easily obtained all year round, regardless of technological skills (Meehan, 1982). Skillful exploitation of aquatic resources could therefore provide

1 economic stability to prehistoric hunter-gatherer populations (Hayden et al., 1987; Nicholas, 1998). We have little knowledge regarding the evolution of fishing. Until 1989 the earliest evidence for the existence of a fishing community in Israel, was Atlit-Yam, dated ca. 8,000 years B.P. (PPNC) (Galili et al., 1993; Zohar et al., 1994). Fish exploitation by the Atlit-Yam inhabitants was attributed to the diminution of the traditional food resources and the opportunities offered by the Neolithic revolution. However, in 1989 a dramatic drop of the Sea of Galilee water level exposed on its' southern shore a prehistoric site dated to 23,000 years B.P. (calibrated): Ohalo-II (Figure 1, Nadel, 2002). The recovery of Ohalo-II strengthened my conviction that the scanty information on prehistoric fishing community is due to dramatic changes in water level (Galili, 1985; Galili & Weinstein-Evron, 1985; Galili et al., 1988; Nadel, 1993b), and that fishing communities were established long before the Neolithic revolution. I strongly believe that the study of fish remains from Ohalo-II will provides crucial information regarding human diet breadth and the development of a broad- spectrum economy and fishing communities during the terminal Upper-paleolithic/ early Epi- paleolithic. 1.2 Cultural setting The major projects that led to the recognition of early Epi-paleolithic sites in the Jordan Valley began with earlier works by Bar-Yosef in Ein Gev I and II (Bar-Yosef, 1975). The term Epi-paleolithic was introduced to the Levant by Perrot (Perrot, 1966) in his report on the Natufian site of Mallaha (Eynan). At present, it is used in the Levant to refer to different Levantine cultural complexes defined by geographic distribution of certain typological features during the era of the last glacial maximum (LGM) and the end of the Pleistocene (Bar-Yosef, 1981; Bar-Yosef, 1990; Goring-Morris, 1987). Given the richness and variety of Epi-paleolithic lithic industries, they are defined by a chronological frame, and by a dominance of various types of backed bladelets which appear in lower frequencies in earlier periods (Bar-Yosef, 1981; Bar-Yosef, 1990; Bar-Yosef & Belfer-Cohen, 1992; Gilead, 1984). Ohalo-II site is exceptional compared with other early Epi-paleolithic sites in Israel, since it is a relatively large site (>1500 sqm), located on the coast of ancient lake, dated to 23,000 cal B.P. (terminal UP/early Epi-paleolithic), and posses clear architectural remains (habitation structures) from several phases of occupation (Nadel & Zaidner, 2002; Tsatskin & Nadel, 2003). A series of brush huts, a grave, several outdoors hearths, and a stone installation were recovered (Nadel, 1993a; Nadel et al., 2002). Among the finds there were three twisted fibers that were probably used as cordage (Nadel et al., 1994). All features were

2 originally dug into the Lisan formation, and where inundated in between phases of occupation (Tsatskin & Nadel, 2003). The flint assemblage is distinguished by Ouchtata, backed and pointed bladelets (Nadel & Zaidner, 2002). It also included uni-polar cores (used for the production of bladelets), Falita points, and low frequencies of retouched blades, burin and scrapers (Nadel, 1999; Nadel & Zaidner, 2002). Given the richness of studies on Epi-paleolithic industries, the number of archaeofaunal studies is surprisingly small (Bar-Oz et al., 1999; Bar-Yosef, 1990; Davis, 1974; Rabinovitch, 1998; Stiner et al., 2000; Stiner & Munro, 2002). They reflect activities of mobile hunter- gatherers with a broad-spectrum economy. Their diet included small game (hares, turtles, lizards, birds, fish etc.), ungulates (Gazella gazella and Dama mesopotamica), and marine shellfish (Bar-Oz et al., 1999; Bar-Yosef, 1990; Davis, 1974; Rabinovitch, 1998; Stiner et al., 2000; Stiner & Munro, 2002). Until the recovery of Ohalo-II, evidence for fish exploitation in the Epi-paleolithic existed only from the site of Mallaha (Eynan) of the Natufian period (ca. 12,000B.P.) (Desse, 1987; Appendix I).

3 Environmental setting Ohalo-II site is situated on the southern shore of the Sea of Galilee, which is today the largest fresh-water lake in Israel. The lake is located in the northern part of the rift valley. During the development of the rift (last 5 m.y.), the landscape changed dramatically and a series of north-south-trending axial lakes were formed (Horowitz, 1979). Sedimentary deposits demonstrate that during the late Pliocene and the Pleistocene, freshwater lakes covered parts of the northern Dead Sea rift and the Kinarot basin (Bartov et al., 2002; Begin et al., 1974; Horowitz, 1978; Horowitz, 1979; Horowitz, 1988; Hurwitz et al., 2000; Rosenthal et al., 1989). From 70,000 to ca. 15,000 BP, the saline Lake Lisan (Figure 2) occupied an area from south of today's' Dead Sea to the northern Kinarot basin (Bartov et al., 2002; Hurwitz et al., 2000). During this period, lake level fluctuated between a minimum of 500 m below sea level (BSL) and a maximum of 180 m BSL (Yechieli et al., 1993).

Figure 2: Map showing the maximum aerial coverage of Lake-Lisan during its highstand of 180 m BSL, location of current lakes, and Ohalo-II (after Hurwitz, 2000). It is assumed that the lake attained its last highstand of 164 m BSL between 26,000 and 23,000 years B.P. (Bartov et al., 2002). During its highstand, Lake Lisan covered several subbasins (Figure 2), such as the northern Kinarot basin that was connected through Yarmuk river deltas (Hurwitz et al., 2000). The salinity level of Kinarot basin during this phase, was lower than 100mg/L, but higher than the present (Hurwitz et al., 2000). Lake Lisan highstand lasted for a short period of a few thousands years, dropping and reaching 300 m BSL at ca. 15,000 B.P. (Bartov et al., 2002). Upon it's decline, two separate lakes were formed: the hypersaline terminal Dead Sea in the south, with ~340 g/ L TDS, and the Paleo Sea of Galilee in the north with ~45 g/L TDS (Hurwitz et al., 2000). Following this event, the exposed beach of Lake Lisan was settled by the early Epi-paleolithic population of Ohalo-II (Nadel, 1990). As primary freshwater fish are sensitive to changes in water salinity level (Banarescu, 1990; Banarescu & Coad, 1991), fish remains identified from Ohalo-II will attest to the ecological conditions prevailing at the Paleo-Sea of Galilee, following the separation of the two lakes. 1.4 Outline of Research Objectives The aims of this study were fivefold:

4 1 Paleoecological: To identify to the lowest taxonomic level the fish remains recovered at Ohalo-II, and to determine the changes in species richness and diversity following the changes in water salinity. 2 Taphonomical: To characterize archaeological versus natural fish bone accumulations. Through a taphonomic study, quantitative and qualitative criteria for distinguishing naturally-derived vs. culturally-derived bone accumulations will be developed. 3 Economical: To identify fish utilization for immediate or long-term consumption. This will be carried out by studying fish butchering methods in accordance with immediate and long-term consumption methods. A set of quantitative and qualitative criteria to characterize butchered fish, will be provided by an ethnographic study. In addition I will calculate the dietary contribution of fish to the inhabitants daily economy. 4 Site Organization: By studying fish bone assemblages in different structures at Ohalo-II site I will be able to better understand the type of human activities that took place at the site: fish processing, consumption, waste areas, and food storage. The ability to identify stored food is important as storage-based economy was a major step in human evolution. 5 Technological: Based on linear regressions, body size of fish captured at Ohalo-II will be reconstructed. This will enable me to determine if fish were selected by their size and what fishing methods (littoral or deep sea) were practiced by the Ohalo-II inhabitants. One of the major obstacle in the study of fish remains from coastal sites that were inundated by the sea is the possibility that the fish remains accumulated at the site resulted from lacustrine deposition and not human activity. The next chapter, therefore, presents a review of studies performed on fish taphonomy.

5 CHAPTER 2: FISH TAPHONOMY “It can not be assumed that all split and fractured bones on an archaeological site have been broken by man” (Clark, 1972) 2.1 Introduction Actualistic taphonomic research has been developed in archaeology in order to understand the processes that control formation of the archaeological record (e.g., Andrews, 1995; Binford & Bertram, 1977; Bonnichsen & Sorg, 1989; Gifford-Gonzales, 1989; Lyman, 1994). In its simplest form, taphonomy refers to the laws of burial (Efremov, 1940). There are numerous factors that influence the formation of fossil deposits such as: mode of death, substrate, decomposition, burial depth, chemical weathering, transformation and accumulation mechanisms (e.g., Bonnichsen & Sorg, 1989; Lyman, 1994). Taphonomy is primarily concerned with isolating the effects of such factors on fossil bone assemblages. A basic problem in any zooarchaeological research is to decide whether or not human agents are responsible for an observed pattern (i.e. bone distribution, fragmentation, burning signs, species diversity, etc.) (Binford, 1981; Butler, 1987). Binford (1981:26) offered a diagnostic signature that discriminates one agent or set of agents from another. The search for criteria for distinguishing taphonomic agents produced various kinds of actualistic research and analytic techniques (Binford, 1978; Bonnichsen & Sorg, 1989; Lyman, 1987; Lyman, 1991b). These included studies of bone modification by nonhuman agents (Behrensmeyer et al., 1989), mode of death (Weigelt, 1989), soft tissue decomposition, bone disarticulation (Lyman, 1994), density, preservation (Butler, 1994; Lyman, 1984; Nicholson, 1992; Robinson et al., 2003), trampling (Fiorillo, 1989), transport, sorting (Behrensmeyer et al., 1986; Behrensmeyer, 1991; Coard & Dennell, 1995), weathering, and skeletal part representation (Behrensmeyer, 1991; Marshall & Pilgram, 1991; O'Connor, 1993). Other actualistic studies concentrated on bone modification by human agents describing butchering methods (Binford, 1978; Lyman, 1987; Noe-Nygaard, 1977; Shipman et al., 1981), fracture types (Binford, 1978; Binford, 1981; Shipman & Rose, 1984), burning and cooking (Bennett, 1999; Nicholson, 1993; Shipman et al., 1984; Speth, 2000), as well as differences between killing, butchering, and consumption sites (Binford, 1978; Rabinovitch et al., 1996). Most taphonomic studies focused on large mammals and only few discussed fish remains. Fewer studies dealt with fish bone deposition resulting from noncultural agents versus human activity (Butler, 1990; Butler, 1993; Stewart, 1989). Wheeler and Jones (1989:78) argued that "natural agencies depositing fish bones rarely produce substantial concentrations of the kinds of fish preferred as human food". However, as I will demonstrate,

6 a similarity between fish natural and cultural accumulations may appear in coastal sites. As the primary goal of my research is to study fish remains from the water-logged site of Ohalo- II, it is essential to understand the depositional nature of the fish remains. Here I briefly review taphonomic studies performed on naturally and culturally deposited fish. 2.2 Naturally deposited fish Taphonomic studies of lacustrine sediments focused on fish remains as a proxy record of past fish abundance (O'connell & Tunnicliffe, 2001), as well as on detecting environmental and paleoecological conditions (Ferber & Wells, 1995; Whitefield & Elliot, 2002). Fish taphonomy differs from mammal taphonomy as a wider set of factors are affecting bones survivorship. These factors include: lake size, water temperature, water depth, pressure, salinity level, oxygen level, oxygen stability, wave activity, currents, sedimentary structure, rate of sedimentation, scavenger activity, bacterial degradation, and carcass flotation (Cutler et al., 1999; Elder & Smith, 1988; Ferber & Wells, 1995; Martin, 1999; O'connell & Tunnicliffe, 2001; Wilson & Barton, 1996). Many authors have discussed the influence of these factors on fish remains: states of articulations, bone modification, species richness and diversity. Disarticulations appear in various taphonomic modes such as: isolated scales, clumps of scales, disarticulated cranial bones (fragmented and/or complete), body fragments, and complete specimens, with and without skulls (e.g.,Mancuso, 2003). Postmortem modifications such as bone cracking, breaks, and abrasion may indicate the length of surface exposure prior to burial (Behrensmeyer, 1991; Mancuso, 2003). Elder and Smith (1988) claimed that in fish taphonomy water temperature is the most important factor in determining the carcass fate. At water temperature below 15oC, most fish carcasses will remain on the bottom until buried. Sometimes they will be consumed and disturbed by scavengers, and therefore their skeletons won't be articulated, and some cranial bones might be lost. However, in water temperature above 15oC most fish carcasses will be buoyed by bacterial decay gases and will be transported to the surface. There, the fish will further decay and pieces will fall into deepwater or drift to the beach where they will be consumed, disarticulated and scattered by scavengers and waves. In this case, only few bones might be buried in the sediments. Clay sediments function as "glue like", by trapping the fish carcasses and preventing their flotation. Therefore, fossil fish will be recovered mainly from cold or deep water, with low oxygen level, in which the fish carcasses were prevented from floating (Cutler et al., 1999; Elder & Smith, 1988; Ferber & Wells, 1995; Martin, 1999; Wilson & Barton, 1996).

7 Wilson and Barton (1996) demonstrated that skull bones become disarticulated faster than anal and dorsal fins. This pattern agrees with other studies demonstrating high survivorship of postcranial element, especially vertebrae (Butler, 1993; Butler, 1996; Elder & Smith, 1988; Ferber & Wells, 1995; Mancuso, 2003). O'conell and Tunnicliffe (2001) showed that scales suffer degradation for record longer than 500 years. They have also proved that when fish are naturally accumulated there is a taxonomic bias against less abundant fish. In an aquatic habitat with over 100 species only 20 species were present (O'connell & Tunnicliffe, 2001). In view of the diverse preservation patterns of fish natural accumulation, it is clear that similarity between naturally and culturally deposited fish may appear. Therefore, fish remains recovered from lacustrine/ coastal archaeological sites, such as Ohalo-II, should not be a-priori treated as culturally deposited and their nature must be examined.

2.3 Culturally deposited fish Until recently culturally derived fish assemblages were identified based on anomalies in skeletal element presentation. This approach predicts that three bone assemblages would be generated during fish procurement: a processing waste assemblage; a product assemblage of consumed fish, and a stored/unconsumed fish assemblage. Each assemblage is assumed to be characterized by different skeletal element representation and fragmentation patterns (Belcher, 1994; Belcher, 1998; Gifford-Gonzales et al., 1999; Hoffman et al., 2000; Stewart, 1991; Stewart & Gifford-Gonzales, 1994; Van Neer & Pieters, 1997). However, recent ethnographic and taphonomic studies demonstrate that this approach is too simplified, and that a number of factors must be considered as well. For example, researchers viewed the absence of salmon skulls as evidence for fish decapitation during processing for long term preservation (Hoffman et al., 2000; Lubinski, 1996). However, contemporary taphonomic studies demonstrate that in Salmon, bone density is lower in cranial bones and therefore their survivorship is lower compared with postcranial bones (Butler, 1994). Thus, anomalies in bone representation such as low frequencies of Salmon cranial bones at archaeological site do not necessarily testify on human activity (Butler, 1994). Ethnographic studies demonstrated a wide variety of methods applied in fish butchering, ranging from drying the whole fish to removal of all bones and drying cleaned fillet (Barrett, 1997; Belcher, 1994; Belcher, 1998; Burgess, 1965; Essuman & Diakite, 1990; Firth, 1975;

8 Hoffman et al., 2000; Locker, 2000; Michael, 1984; Stewart, 1982; Stewart, 1989; Walker, 1982a; Walker, 1982b). Although fish butchering rarely leaves distinctive cut marks on the bones (Butler, 1996; Butler & Schroeder, 1998), each method is characterized by a distinctive bone assemblage. Surprisingly, this issue has received little attention in the literature (Zohar & Cooke, 1997). Variation appears also in cooking methods applied (boiling, roasting, etc.) and their effect on bone survival. Only few studies examined the effect of cooking methods on fish bone survival (Fred et al., 2002; Lubinski, 1996; Nicholson, 1998). Nicholson (1998) demonstrated that in cod, 66% of the skeletal elements (postcrania and crania) were destroyed due to boiling. Lubinski (1996) demonstrated the effect of cooking methods and sediments acidic and alkaline conditions on Salmon bones preservation. In view of the scarcity of taphonomic studies on culturally deposited fish and the possible resemblance with natural accumulations, fish remains from lacustrine/coastal sediments should be analyzed cautiously, with a wide set of quantitative and qualitative criteria developed for each habitat (Zohar et al., 2001).

9 CHAPTER 3: SITE SELECTION AND FIELD TECHNIQUES "The stratigraphic record is certainly imperfect" (Martin, 1999) In this chapter I describe the areas selected for this study, the field methodology applied, and the sampling techniques. This study includes fish bone assemblages from three accumulations: archaeological, natural, and ethnographical. The archaeological site is of an early fishing community (23,000 cal B.P.) from the Sea of Galilee (Ohalo-II); the site for natural fish accumulation is also located along the southern shore of the Sea of Galilee; the ethnographical sites (for the study of traditional fish butchering methods for long-term preservation) are two fishing communities: one from the Pacific coast of Panama and the second from the Red Sea coast of south Sinai, Egypt. 3.1. The Archaeological Site of Ohalo-II Ohalo-II is a submerged site located at the south-western shore of the Sea of Galilee, Israel, at -212.5 meters below mean sea level (BSL) (Figure 1). This is a late Upper- Paleolithic site, dated to 23,000 years B.P. (Nadel et al., 2001) . The site was exposed as a result of a drop in water level in 1989 following several years of low precipitation, and was excavated for six seasons (1989-1991; 1999-2001; Figure 3) (Nadel, 1990; Nadel, 1991; Nadel, 2002; Nadel et al., 2002). The site size is ca. 2000 square meters, of which 400m2 were excavated. The excavation revealed the remains of six brush huts, several hearths, and a grave (Figure 4 ; Nadel, 1997; Nadel et al., 1994). These features were clearly visible on the surface, as the sediments were dark in color in comparison to the surrounding sediments (Nadel, 2002; Nadel & Werker, 1999). Figure 3: Ohalo-II site covered with water (on left) and its exposure at -214.0 BSL in the summer of 2000.

Flint, mammals, rodents, reptiles, turtles, birds, mollusks, fish, botanical, and human remains were recovered at Ohalo-II (Nadel, 2002; Nadel et al., 2002). Fish remains appeared in large quantities in most of the excavated areas. In this study I focus on a sample of 19,799 fish remains recovered from seven different localities (Table 1). These included: loci 1, 2, 3 (brush huts), Loci 7 and 9 (ashes) and locus 8, an unidentified pit (Figure 4). Table 1: Frequency (NISP) and percentage of fish remains by loci at Ohalo-II. Excavated area Identified Unidentified Total

11 NISP % NISP % NISP Locus 1 11,676 92.7 921 7.3 12,597 Locus 2 58 100.0 - . 58 Locus 3 616 65.0 332 35.0 948 Between L.7 & L.9 29 72.5 11 27.5 40 Locus 7 4,110 75.0 1,368 25.0 5,478 Locus 8 537 88.3 71 11.7 608 Locus 9 47 67.1 23 32.9 70 Total 17,037 86.2% 2,726 13.8% 19,799 Locus 1 is the largest structure recovered at the site, it is 4.5 m long in its north-south long axis, with a kidney-like shape (Figure 4). Remains of constructed wall were recovered during the excavation of locus 1 (Nadel & Werker, 1999). Locus 2 is located close to locus 1 (Figure 4), its shape is similar to that of locus 1, but is smaller in size. Locus 3 is a pear like shape, located close to loci 1 and 2 (Figure 4). All structures were of similar form (wide shallow bowls) (Nadel et al., 2001). Locus 7 is one of the largest excavated areas. It is 7 m long (north-south axis) and ca. 3 m wide (west-east axis). Its shape is not AI AG A B D I N S regular though its boundaries were clearly 70 70 - 212.28m L.11 - 212.44m visible. It comprises of series of hearths

75 L. 10 76 L.16 77 (Figure 4). The hearths were red, black or 78 L.6 L.1 L.15 79 80 80 gray in color and 5-10 cm in depth. 81 82 83 L.17 L.2 84 85 86 L.5 L.4 L.3 L.8 87 88 Figure 4: Excavated loci at Ohalo-II where 89 90 L.7 90 91 fish remains were recovered (after Nadel et L.9 - 212.28m 92 93 - 211.90m 95 al., 1994). L.12 - 212.20m Locus 8 is a small unidentified round 10 0 10 0 structure located north of locus 7 and south of locus 2. Its diameter is 40 cm and its 11 0 0 5 m 11 0 N depth is 25 cm. Locus 9 is similar to locus AL AG ABCD E FGH I JK LMN S 7, and is comprised of several hearths, each 30-40 cm in diameter. 3.2. Fish Natural Accumulation In order to gain insight into fish natural accumulation at the Sea of Galilee southern shore, an in-situ beach area that does not carry any archaeological artifacts, was selected. This was not an easy task as the sharp drop in water level, exposed many archaeological sites along the lake shore (Nadel, 1993). Therefore, I conducted a survey along the southern shore of the Sea of Galilee and consulted the site geologists and archaeologists to aid in identifying

12 potential research areas. Based on the information gathered I selected an area of natural lacustrine sediments, 150 m north of the Ohalo-II site. Excavation of the natural site took place during July 2001, when water Sea level dropped to -214 BSL. I marked an area of 100 meter in length and 50 meter in width, using an eTrek global positioning system (GPS). The GPS provided accurate data on the position of

N the excavated area, according to three dimensions (X,Y,Z). Based on this information I used Excel (Office 98) for MAC and calculated the position of 24 random squares, which were then excavated (Figure 5).

Figure 5: Position of 24 random squares sampled along the

e southern shore of the Sea of Galilee. t

0 Each square was 0.5sq.m in size, (Figure 6), and was

1 50 meter excavated to the maximum depth of 30-50cm. The maximum depth of each square varied according to its sedimentology. I divided the excavation into three layers as follows (Figure 6): upper sandy layer (3-10 cm thick), median brown layer (mixture of sand and clay : 5-10 cm thick), and dark anaerobic clay layer (10 cm thick). The bottom clay layer was radiocarbon dated to 430-620 A.D. (University of Arizona, Tucson). Figure 6: An excavated square from the Sea of Galilee present shoreline (observe the changes from upper sandy layer to a dark clay layer at the bottom).

In order to minimize collection bias against smaller elements and species (including micro-fauna), I used fine mesh (0.5 mm) screens for wet sieving. A total of 5795 fish remains from the 24 random squares, were recovered. In addition I performed a survey along the beach, in areas that were heavily vegetated, or farther away from the archaeological and natural digging area. I hand picked 72 fish remains that were scattered in these areas. I also collected (hand-picked) 101 naturally

13 accumulated fish bones from the present surface covering at the Ohalo-II site. Therefore, I recovered a total of 5968 fish remains from the natural sediments along the Sea of Galilee. 3.3 Ethnographic Study of Fish Procurement Methods In order to characterize fish remains associated with human activity, I examined fish butchering methods applied by present day traditional fishing communities. I conducted this research on fishermen from two geographical regions: Parita Bay (Panama; Figure 7) and Nabek Oasis (South Sinai-Egypt; Figure 9). Parita Bay is a small tidal embayment in the north-western corner of Panama Bay (Figure 7). Modern fishing concentrates on littoral and estuarine waters. Individuals or small groups use throw and gill nets, as well as hand-lines from dug-out canoes, with and without outboard motors. Although most modern catches are sold fresh, a few families in the coastal settlements of Aguaduce, El Rompio and Boca de

Caribbean Sea Panama City Parita still process considerable amounts of fish by

Parita bay salting and wind and sun drying, throughout the Pacific Ocean year. This process is performed either in their Rio Grande houses or in seasonal huts (Figure 8).

Figure 7: Map of Panama and Parita Bay showing Rio Pocri location of the studied sites. Agudulce El Rompeo

Cerro Mangote Parita Bay Rio Santa Maria

Bocas de Parita

Parita

N La Villa

0 5 10 15Km

14 Figure 8: Fish drying and salting seasonal camp at the mouth of Rio Santa Maria (on left), and fish processed for long-term preservation by Francisco at Partita Bay (on right).

Nabek Oasis rests on the south-eastern shore of the Gulf of Eilat in Southern Sinai (Figure 9). It is situated at the end of a wide delta stretching 5-7 km east of the Red Sea

32o 36o mountains. Several rivers including Wadi N Mediterranean Sea Jerusalem Kid feed this long delta. Ten nomadic Israe l Dead Sea Bedouin tribes are scattered throughout the South Sinai peninsula, each exploiting a different geographic area. The Muzeina tribes exploit the eastern-shore line, and Eilat Bay of Eilat South Sina i therefore practice intensive fishing year round using throw and gill nets in shallow

Nabek o Egypt water (Kobyliansky & Hershkovitz, 28 Oasis 1997). Red Sea Figure 9: Map of Sinai, with the location 0 100 200 km of the studied site (Nabek Oasis) indicated.

Some of the Muzeina families seasonally move between the seashore and the mountains, while other live in a sedentary fishing village situated in Nabek. Therefore, Nabek Oasis constitutes a permanent Bedouin's fishing village, and some seasonal huts that are scattered along the beach. These huts are exploited by different individuals or families for few fishing days, all year. Only a small amount of the catch is consumed fresh while most of it is processed by sun drying (Figure 10). Dry fish are called "Chot", and they are either scattered on the hut roofs or hanged for later consumption. The dry fish are later cooked with rice (Levi, 1987). Figure 10: Fish butchering by a Bedouin family in Nabek Oasis (Sinai).

Documentation of fish butchering methods included in Sinai 72 fish from 9 species, butchered by a single method (Method-2:longitudinal cut through the skull). In Panama, I examined 573 fish

15 from 34 species, butchered by two methods (Method-1:whole, and Method-2:longitudinal cut through the skull): 442 fish were butchered by method-1, and 131 fish were butchered by method-2. In each locality I collected the most common species (171 fish from Panama, and 47 fish from Sinai; Table 2) and carried them back to the lab. At the lab, I separated the bones (see next chapter) and examined the effect of fish butchering method on 17,862 skeletal elements, according to: absent bone, damaged bone, and typical breakage pattern. Table 2: Morphometrics for butchered fish collected from traditional fishermen in Panama (Central America) and southern Sinai (Egypt) by butchering method and body size.

Panama-Parita Bay Butchering Method Body Size (Range)

Species Count 1 2 Body Mass Total Length Standard Length Caranx caninus 29 1 28 823-5760 gr 458-790 mm 370-610 mm Haemulopsis nitidus 32 32 0 112-242 gr 200-261 mm 164-210 mm Arius kessleri 41 13 28 453-1644 gr 395-560 mm 320-478 mm Cathorops multiradiatus 29 29 0 90-208 gr 223-305 mm 184-252 mm Sciadeops troschelii 40 10 30 108-2409 gr 227-580 mm 187-496 mm Total Count 171 85 86

Sinai-Nabek Bay Butchering Body Size (Range) Species Method-2 Body Mass Total Length Standard Length Acanthurus nigrofuscus 24 52-1130 gr 137-393 mm 100-300 mm Siganus luridus 23 63-316 gr 171-255 mm 132-202 mm Total Count 47

16 CHAPTER 4: METHODS “Only a small part of what once existed was buried in the ground; only a part of what was buried has escaped the destroying hand of time; of this part all has not yet come to light again; and we all know only too well how little of what has come to light has been of service to our science” (Montelius, 1888) 4.1. Recovery Bias Recovery procedures, particularly screen size selection, have a significant effect on the results of faunal analysis in general and on fish remains in particular (Zohar & Belmaker, 2003). Researchers have demonstrated the effect of screen size selection on the differential recovery of taxa, their ordinal rank of abundance, and on skeletal representation (Butler, 1993; Butler, 1996; Cannon, 2001; Casteel, 1972; Fitch, 1967; Gobalet, 2001; Gordon, 1993; James, 1997; Zohar & Belmaker, 2003). In this study I was concerned with taxonomic representation, relative abundance, and body part representation. To minimize material lost, all sediments were sieved through 2 and 1 mm mesh size (Nadel et al., 2002). All fish remains were later analyzed in the lab under a ZEISS Stereomicroscope. 4.2. Sampling Bias A major problem that frequently arises when analyzing archaeological assemblages is sample size needed (Grayson, 1984; Leach, 1986; Reitz & Wing, 1999; Wheeler & Jones, 1989). Although every effort should be made to study all recovered bones, in the case of Ohalo-II the large amount of fish remains recovered, made this impossible. The sample size studied in the present research varied between excavated structures, following my research hypothesis. Therefore, small samples will be used only for taxonomic composition. However, samples that were thoroughly examined, will be used for taphonomic study. The differences in sample sizes taken non-randomly from different structures at Ohalo- II, raise the question regarding sample size/richness relationship. Many faunal studies have demonstrated that since rare species are inherently difficult to study in paleo-ecological communities, sample-size effects should always be evaluated before differences or similarities are assumed between two assemblages (Koch, 1987; Krebs, 1999; Zohar & Belmaker, 2003). Several methods have been developed to deal with samples that differ in size and are not random nor normally distributed. These include: log-linear regression method (Baxter, 2001; Grayson, 1984; Plog & Hegmon, 1993); Kintigh simulation method (Kintigh, 1984); and Kaufman's jackknifing method (Kaufman, 1998). These methods, however, have been widely criticized (Baxter, 2001). In the present study comparison between assemblages has been carried out using the rarefaction technique (Baxter, 2001; Grayson, 1984).

17 The rarefaction method has been used to compare between number of species found in different samples when sample sizes differ (Baxter, 2001; Grayson, 1984). Rarefaction techniques use the data from the larger sample to answer the question: "How many species would have been found in a smaller sample?" (Heck et al., 1975; Hurlbert, 1971; Krebs, 1999). I applied the rarefaction technique in order to examine the influence of sample size on species richness, and bone representation. I used a computer program for Mac: Analytic Rarefaction V.1.3, developed by Holland, (2001). 4.3. Identification of fish remains In this study fish bones were identified to the lowest taxonomic level possible. This was carried out with the help of a large osteological reference collection prepared by me. The reference collection included fish from the Sea of Galilee (See also Appendix II) the Jordan river (Table 3), and the Red Sea (Eilat & Egypt; Table 4). Several freshwater species were collected from the Nile river (Table 3). I also used the reference collection housed at the Royal Museum of Africa in Tervuren, Belgium. The latter includes skeletons of freshwater fish from Turkey, Syria and Lebanon. For identification of skeletal elements of Pacific fish from Parita Bay I used the fish osteological collection housed at the Smithsonian Tropical Research Institute (STRI) in Panama. Taxonomic identification, were based on Ben-Tuvia (1978), Golani and Darom (1997), and Goren (1974, 1983). Body mass (BM gr), total length (TL mm), standard length (SL mm), body depth (BD mm) and head length (HL mm), were taken for each fish, using digital scales, digital caliper, and an osteometric board. Following the measurements the fish were macerated with enzymes (Davis & Payne, 1992). Table 3: A list of freshwater fish from the Sea of Galilee, Jordan River, and Nile river (n= 324), that were prepared for the osteological reference collection. Species Count Species Count Acanthobrama lissneri 37 Lates niloticus 6 Acanthobrama terraesanctae 38 Liza ramada 1 Anguilla anguilla 5 Noemacheilus panthera 1 Astatotilapia flaviijosephi 10 Oreochromis aureus 42 Barbus canis 26 Salaria fluviatilis 4 Barbus longiceps 4 Sarotherodon galilaeus 39 Capoeta damascina 8 Tilapia nilotica 1 Clarias gariepinus (lazeral) 31 Tilapia zillii 44 Ctenopharyngodon idellus 1 Tristamella simonis 10 Cyprinus carpio 3 Barbus bynini 1 Garra rufa 8 Bagrus sp. 1 Hemigrammocapoeta nanus 1 Black carp 2

Table 4: A list of Red Sea fish from Eilat and Egypt (n=71) that were prepared for the osteological reference collection.

18 Species Count Species Count Abudafduf saxatilis 1 Scarus madagascarensis 1 Acanthurus nigrofuscus 24 Scarus sp. 1 Diplodus noct 2 Scomberomorus commerson 2 Epinephalus fasciatus 1 Siganus luridus 24 Hemiramphus far 2 Siganus rivulatus 7 Saurida undosquamis 3 Sparus aurata 1 Scarus ghobban 1 Synodus variegatus 1 4.4. Fish Osteological Characteristics For each species, I established a list of skeletal elements following Wheeler & Jones (1989) and Butler (1990) (Appendix III-IX). Since Cyprinidae and Cichlidae are two of the largest and most abundant freshwater fish families, I describe their biogeography and main osteological characteristics, in Appendix II. This information is important for zooarchaeological studies as there are very few osteological fish manuals. 4.5. Quantification Analysis For the purpose of this study, the term "identified" means that fish remains could be ascribed to family or genus/species level. The term "unidentified" means that the remains could be ascribed as teleost fish. As mentioned before (section 4.2) the total number of fish remains at Ohalo-II is unknown. Several counting methods were applied in the present study following Klein and Cruz- Uribe (1984) recommendations. All data are summarized using number of identified specimens (NISP). In addition, I have also calculated minimum number of individuals (MNI) for each taxonomic group, according to the most abundant and well-preserved skeletal elements (Grayson, 1984; Klein & Cruz-Uribe, 1984; Orton, 2000). NISP was used as an index for abundance of taxonomic groups in conjunction with ordinal ranking. MNI was used as general measures for fish dietary contribution, especially for estimating fish catch. In order to identify the taphonomic agents responsible for the accumulation of fossil bones, and to differentiate between cultural and natural accumulations, I examined several attributes, and calculated various quantitative and qualitative indices described blow (Zohar et al., 2001). 4.5.1. Taxonomic composition and diversity For taxonomic identification I used bones that were species specific such as (Appendix II): pharyngeal bones, pharyngeal teeth, atlas, and axis. Since only few bones were species specific I used most cranial and postcranial bones that could be identified to family level (Appendix II). Species richness was calculated as the number of species identified in each sample (Krebs, 1999). Species diversity was calculated by using the Brillouin Index (and not Shannon-Wiener function) since our samples are not random and the total number of species

HB= 1 log2( N! ) N n1!n2!n3! is unknown (Krebs, 1999). Moreover, Brillouin index (HB) is most sensitive to the abundance's of the rare species in the sample (Krebs, 1999). I calculated the Brillouin Index of diversity according to the following formula (Krebs, 1989: 362-365):

When HB is Brillouin's index, N is the total number of individuals in entire collection, n1 is the number of individuals (NISP) belonging to species 1, and n2 is the number of individuals (NISP) belonging to species 2, etc.

4.5.2. Body Part Frequency Body part frequency is commonly used to identify the agents responsible for a particular bone assemblage (Binford, 1981; Butler, 1996; Elder & Smith, 1988). It has been suggested that different agents (predators , birds, water, and humans) modify or destroy particular bones in a distinctive way. In order to document variation in skeletal element representation as a result of cultural processing, natural death, body part transport, differential preservation etc., I calculated body part frequencies by using several methods. First, according to individual skeletal elements for each taxonomic group (Appendix II VIII). Secondly, according to eight anatomic regions (following the terminology used by Wheeler and Jones, 1989). Thirdly, according to postcranial and cranial regions (following Butler, 1990:67-68; Figure 11). POSTCRA N IA L CRA N IA L Figure 11: A generalized fish skeleton second dor s al f i n first dors al fin caudal ver t ebr ae anter i or dor sal s pine presenting selected cranial and ultima te ve r tebr a e thor ac ic ver tebr ae supr a occi pita l postcranial bones (after Lyman, 1994: frontal 89). premax illa maxilla anal fin dent ar y ribs

precau dal ve rtebra e pector al fi n quadr ate anter ior a nal s pine pelv ic f in pelv ic g ir dl e pr eoper cular oper cu lar 4.5.3. Survival index (SI) To evaluate the survivorship of each bone I calculated a survival index (SI). SI is expressed as the ratio between number of observed bones (NISP) to number of expected bones

(per skeletal element). Therefore I calculated the SI as follow: SI=no. of observed bones no. of expected bones When SI=1, the observed NISP=expected NISP. SI larger than 1 imply over-representation, while SI smaller than 1 indicate under-representation at the site. I calculated the expected bone ratio for each species identified (Appendix III-IX; Table 5).

20 The expected values of each bone were calculated, for each species, from the ratio of each bone in a complete skeleton multiplied by the total NISP observed in the sites (Zohar et al., 2001). Afterwards, I compared between the expected and observed values, by using the chi-square formula: (Obs.- Exp.) 2 with df= n-1. χ2=Σ Table 5: Exp. Number (NISP) and percentage of bones, by anatomic regions, expected in complete skeleton of five taxa of freshwater fish. Anatomic region Acanthobrama sp. Barbus/Capoeta sp. Tilapia sp. Clarias sp. NISP % NISP % NISP % NISP % Crania Neurocranium 51 23.83 51 22.08 48 24.62 42 24.0 Branchial region 50 23.36 58-60 25.11 36 18.46 22 12.6 Hyoid region 16 7.48 16 6.93 16 8.21 14 8.0 Oromandibular region 20 9.35 20 8.66 20 10.26 18 10.3 Opercular series 8 3.74 8 3.46 8 4.10 6 3.4 Postcrania Appendicular skeleton 16 7.48 16 6.93 16 8.21 8 4.6 Median fins 4 1.87 7 3.03 21 10.77 2 1.1 Weberian apparatus 10 4.67 10 4.33 - - - Vertebral column 39 18.22 45 19.48 30 15.38 63 36.0 Total 214 100% 231-233 100% 195 100% 175 100% * Doesn't include the following elements: Scales, Ribs, intermuscular, pterygiophore, soft fin ray interhamel, epural, hypural, urostyle and radials. 4.5.4. Fragmentation index To evaluate the preservation status of fish bones, I calculated a fragmentation index. For terrestrial mammalian assemblages this index is used as an indicator for marrow extraction. For fish bone assemblage from archaeological sites, it may reflect resistance to post-deposition deterioration, as well as the butchering method practiced (Bullock, 1994; Zohar & Cooke, 1997). For each bone I determined a fragmentation status (Figure 12): "complete" ca. 91-100% of the bone is present; "slightly fragmented" 71-90% of the bone is present; "partially fragmented" 51-70% of the bone exist; "highly fragmented" 30-50% of the bone is present, and "heavily fragmented" ca. 25% or less of the bone remains. The bones' "fragmentation status" was used to characterize and differentiate between bones from butchered fish and bones from natural accumulation. At Ohalo-II I examined the bones' "fragmentation status" in each sample and compared it with natural and ethnographic assemblages.

21 Figure 12: Fragmentation classes 91-100% of bone remaining 71-90% used for classification of bone state of preservation. 51-70%

30-50%

<25% 4.5.5. WMI of fragmentation In order to standardize the degree of bone fragmentation I calculated their weighted mean index (WMI): [WMI=∑(Wi*Xi)/100]. The WMI expresses the mean degree of fragmentation calculated from the relative frequency (Xi) of each bone divided according to ten fragmentation categories (i.e., fragments size of: 95%; 85%;75%; etc. until 5%) . Following this calculation I calculated the correlation between the survival index (SI) and the WMI of fragmentation for each anatomic region. This calculation demonstrates whether abundant bones reflect a high occurrence of a body region, or are highly fragmented and therefore has been counted several times, and consequently may appear over-represented. 4.5.6. Fish exploitation index There are a number of methods described for detecting changes in faunal resource selection. The prey choice model suggests that human foragers will pursue resources that provide the highest energetic returns (larger animal) (Butler, 2000). Given the relationship between body size and resource ranking, changes in the contribution of high- and low-ranked resources in human diet can be estimated from the faunal remains. In the case of fish exploitation I examined changes in contribution of large taxa ( Barbus sp., Capoeta sp., and Cichlidae) vs. small taxa (Acanthobrama sp. and small Cyprinidae). The ratio between high- and low-ranked species was calculated from a fish index as follow (Butler, 2000):

This index is based on the ratio of large fish to the sum of large and small fish taxa. The Σ NISP large taxa Fish Index = larger the ratio (closer to 1), the greater the Σ NISP large taxa+Σ NISP small taxa contribution of large, higher-ranked fish (Butler, 2000). 4.5.7. Bone modification I examined all skeletal elements for cut marks, abrasion, tooth marks, and burning signs. Several methods were proposed to identify burnt bones (Nicholson, 1993; Shipman et al.,

22 1984; Stiner et al., 1995). The simplest method is observation of bone color, which was applied on Ohalo-II fish remains. A number of color/temperature scales are published, all showing a progression from dull brown through black (ca. 300oC), gray, bluish white (ca. 700oC), to white (Nicholson, 1993; Shipman et al., 1984; Stiner et al., 1995). However, since the bones from Ohalo-II were embedded in a clayish an aerobic lacustrine environment, their color is dark, probably due to chemical alterations. As a result black colored bones can not be assigned as having been burned. A similar problem had been observed by Stewart (1989) while investigating fish remains from Lake Turkana, Africa. At Lake Turkana, Stewart (1989:109) found that most bones characterized as "burnt" were actually manganese stained. Therefore, at Ohalo-II, burning signs were assigned only on bones that exhibited white and gray colors (Stiner et al., 1995). In order to address the coloring problem, the bones mineralogical composition was examined. I performed this analysis with a Fourier Transform Infra-red (FTIR) spectrometer from MIDAC Corporation (Costa Mesa, CA, USA), housed at the Weizmann Institute in Rehovot (Shahack-Gross, 2002; Weiner & Bar-Yosef, 1990). The FTIR methods belong to the spectroscopic techniques that work on the principle that under certain conditions materials are able to absorb energy. The results are displayed as a plot representing the intensity of absorption or transmission, on the y-axis, and the function of energy, on the x-axis. The relative intensity of the peaks depends on the ratio of minerals and collagen in the sample and on the absorption coefficient of each material. Changes in the bone structure (collagen and hydroxyapatite) are used to identify and differentiate between burned bones (ash), charcoal, and modern bones (Weiner & Bar-Yosef, 1990). The mineralogical identification by FTIR is performed with a commercial mineral library (Sadtler Research Laboratories, Philadelphia, PA, USA) and Prof. S. Weiner additional standards (Shahack-Gross, 2002; Weiner & Bar- Yosef, 1990). In this study the FTIR was used on a small sample of bones from various archaeological sites, including Ohalo-II, and on bones recovered at the Sea of Galilee natural accumulation. In addition, a sample of bones from the natural accumulation clay deposits was sent for radiocarbon dating at the University of Arizona, Tucson (sample no. 4385, i.d. 3469-5469). 4.5.8. Bone Spatial Distribution I measured the bones spatial distribution by calculating the standardized Morisita index of dispersion (Krebs, 1989). This index is one of the best measures for distribution because it is independent of population density and sample size (Krebs, 1989). The standardized

Morisita's index of dispersion (Ip) ranges from -1.0 to +1.0, with 95% confidence limits at

Σx2-Σx I =n i i d 2 (Σxi) -Σxi +0.5 and -0.5. Random patterns give an Ip value of zero, clumped patterns above zero, and uniform patterns below zero. Standardized Morisita's index of dispersion is calculated according to the following formula (Krebs, 1989, Box 4.5 p. 152: see further calculation for standardization): where: Id=Morisita's index of dispersion; n= sample size; ∑xi=sum of NISP 2= in quadrates count ; and ∑xi sum of NISP in squared quadrate counts. Degrees of freedom are calculated according to n-1, from the Chi squared table. I also calculated the Bone scatter frequency (BSF) as the average number of bones per sq.m. This calculation is based on Stewart's formula: NISP bones/m2 (Stewart, 1989). 4.5.9. Analytic Calculations Most of the analyses were performed with Statview V.5 , and SPSS V.6.1 for Macintosh. I used contingency tables and Chi-square tests to compare between observed and expected NISP (Zar, 1984:61-64). For testing differences between two groups (2 x 2 contingency table) Fisher's exact test was applied (Zar, 1984: 64-65). It is important to note that Chi- square test can be biased when the contingency tables contain cells where n<5. Therefore, if such frequencies occurred, the entire row was combined with a similar row, or a contingency table was calculated with probability values calculated using a randomization test (Manly, 1991). For comparison between proportions of two independent samples the following equation was used (Zar, 1984:396): P -P z= 1 2 pq pq + n1 n2 When P1= X1/n1; q1=1-P1 from one sample and P2= X2/n2 ; q2= 1-P2 from a second sample. One way ANOVA, Wilcoxon rank-test, Mann-Whitney test, and Freidman's test were applied when appropriates (Sokal & Rohlf, 1981; Zar, 1984). I have also used two multivariate procedures: correspondence analysis and multidimensional scaling (MDS). Correspondence analysis: This is a descriptive/exploratory technique designed to analyze simple two-way and multi-way tables containing some measure of correspondence between the rows and columns. The results provide information that is similar in nature to those produced by factor analysis techniques, and they allow one to explore the structure of categorical variables included in the table. The method was developed from a philosophical orientation that emphasizes the development of models that fit the data, rather than the rejection of hypotheses based on the lack of fit (Weller & Romney, 1990). Therefore, there are no statistical significance tests that are customarily applied to the results of a

24 correspondence analysis. The primary purpose of the technique is to produce a simplified (low- dimensional) representation of the information in a large frequency table (or tables with similar measures of correspondence) (Weller & Romney, 1990). In a typical correspondence analysis, a cross tabulation table of frequencies is first standardized so that the relative frequencies across all cells sum to 1.0. One way to state the goal of a typical analysis is to represent the entries in the table of relative frequencies in terms of the distances between individual rows and/or columns in a low-dimensional space. The different column values in each row of the table is regarded as coordinates in a n-dimensional space, allowing to calculate the Euclidean distances between the different row points in the n- dimensional space. The distances between the points in the n-dimensional space summarize all information about the similarities between the rows in the table. The maximum number of values that can be extracted from a two- way table is equal to the minimum of the number of columns minus 1, and the number of rows minus 1. The information about the similarities between the rows can be presented in a simple 1, 2, or 3-dimensional graph, by finding a lower-dimensional space, in which to position the row points in a manner that retains all, or almost all, of the information about the differences between the rows (Weller & Romney, 1990). This technique was used to compare between the relative abundance of taxonomic groups identified at the natural accumulation and various loci at Ohalo-II site. Multidimensional scaling: The purpose of multidimensional scaling (MDS) is to provide a visual representation of the pattern of proximities (i.e., similarities or distances) among a set of objects. MDS has no underlying assumptions about the normality or linearity of the data. MDS is an ordination procedure that compresses multidimensional space onto a simple two-dimensional representation (Borg, 1981). The fit to the two dimensional model is evaluated by a stress factor, which ideally should be lower than 0.1. For example, in a matrix of perceived similarities between various variables, MDS plots the variables on a map that places the similar variables near each other, and those that are perceived to be very different from each other are placed far away from each other. The MDS method was used to examine similarity in bone fragmentation patterns observed in different taxa of fish. It was used to separate between two butchering methods observed on recent fish, based on each bone survival index (SI). For comparing recent butchered fish with natural accumulation and the Ohalo-II site, the relative abundance of bones that are slightly fragmented (more than 75% of bone is present), were used. A similar comparison was calculated between the natural accumulation and Ohalo-II fish remains.

4.6 Osteological measurements Standard measurements on vertebrae centrum were taken, according to Morales and Rosenlund (1979). These measurements are: vertebrae maximum width, max. length and max. height (Figure 13). Measurements were carried out using electronic digital calipers with an accuracy of 0.01 mm. The vertebrae measurements taken on recent fish were used to calculate regression equation between vertebra diameter and fish body size (body mass and standard length). These measurements were also used to differentiate between taxonomic groups. Width Length Figure 13: Measurements performed on vertebra centrum. Height 4.6.1. Body mass estimation For choosing the most appropriate bones for body mass prediction, I followed Wheeler and Jones' recommendations (1989). In Cyprinidae the atlas and axis (Figure 13) suited best since they are species specific, easy to identify, their centrum is well preserved, and they were recovered in high frequencies. These diameters are highly correlated with fish body size, although atlas height display the lowest correlation in Acanthobrama sp. (Tables 6-7). The parameters of the regression equation for body mass estimation of Acanthobrama lissneri (n=27), A. terraesanctae (n=36), and Barbus canis (n=16), are presented in table 6. The parameters of the regression equation for standard length estimation appear in table 7.

26 Table 6: Regression equations for body mass (BM) in Cyprinidae as a function of atlas and axis dimensions. Variables Species Regression equation r2 F p B.M. (gr) vs. Atlas Length A. lissneri Y= -6.9 + 11.7 * X .822 115.25 <.0001 A. terraesanctae Y=-14.1 + 18.8 * X .857 203.16 <.0001 B. canis Y=-184.9 + 119.1 * X .806 66.67 <.0001

B.M. (gr) vs. Atlas Width A. lissneri Y=-2.6 + 7.3 * X .560 31.8 <.0001 A. terraesanctae Y=-17 + 17.4 * X .917 374.69 <.0001 B. canis Y=-173.6 + 106.2 * X .843 91.34 <.0001

B.M. (gr) vs. Atlas Height A. lissneri Y=-4.0 + 14.8 * X .538 29.07 <.0001 A. terraesanctae Y=-8.3 + 25.18 * X .430 25.70 <.0001 B. canis Y=-28.1 + 297.4 * X .783 50.43 <.0001

B.M. (gr) vs. Axis Length A. lissneri Y=-6.2 + 11.4 * X .866 142.18 <.0001 A. terraesanctae Y=-15.7 + 19.9 * X .796 132.80 <.0001 B. canis Y=-194.4 + 122.9 * X .823 74.38 <.0001

B.M. (mm) vs. Axis Height A. lissneri Y=-6.5 + 21.8 * X .729 54.68 <.0001 A. terraesanctae Y=-20.9 + 42.04* X .634 58.87 <.0001 B. canis Y=-224.8 + 134.8 * X .817 71.22 <.0001

B.M.. (mm) vs. Axis Width A. lissneri Y=-7.47 + 13.04 .799 91.70 <.0001 A. terraesanctae Y=-17.23 + 19.9 * X .884 258.94 <.0001 B. canis Y=-163.6 + 104.1 * X .817 71.20 <.0001

Table 7: Regression equations for standard length (SL) in Cyprinidae as a function of atlas and axis dimension. Variables Species Regression equation r2 F p S.L. (mm) vs. Atlas Length A. lissneri Y=13 + 46.4 * X; .894 211.43 <.0001 A. terraesanctae Y=37.2 + 39.3 * X .911 346.58 <.0001 B. canis Y=14.6 + 42.7 * X .994 2474.10 <.0001

S.L. (mm) vs. Atlas Width A. lissneri Y=25.6 + 32.7 * X .774 85.85 <.0001 A. terraesanctae Y=32.7 + 35.6 * X .935 489.79 <.0001 B. canis Y=20.6 + 37.3 * X .993 2313.90 <.0001

S.L. (mm) vs. Atlas Height A. lissneri Y=25.7 + 59.93 * X .551 30.70 <.0001 A. terraesanctae Y=50.3 + 51.73 * X .443 27.04 <.0001 B. canis Y=-20.8 + 107.3 * X .985 911.9 <.0001

S.L. (mm) vs. Axis Length A. lissneri Y=16.8 + 44.3 * X .894 186.43 <.0001 A. terraesanctae Y=35.8 + 40.5 * X .808 142.90 <.0001 B. canis Y=12.4 + 43.6 * X .994 2473.10 <.0001

S.L. (mm) vs. Axis Height A. lissneri Y=16.1 + 84.2 * X .725 55.42 <.0001 A. terraesanctae Y=18.5 + 92.3 * X .745 99.39 <.0001 B. canis Y=0.9 + 48.1 * X 4344.10 <.0001

S.L. (mm) vs. Axis Width A. lissneri Y = 11.7 + 50.8 * X .838 118.56 <.0001 A. terraesanctae Y = 32.1 + 40.9 * X .907 330.49 <.0001 B. canis Y = 23.1 + 37 * X .990 1545.10 <.0001 4.6.2. Vertebrae diameter Although atlas and axis are species specific in Cyprinidae, most of the vertebrae can be identified only to the family level. Therefore, I examined the possibility to differentiate between genus according to the distribution of vertebra width diameter. This method is based on the assumption that in Cyprinidae vertebrae larger than 3.5 mm in width, can represent

27 only two taxonomic groups (Barbus sp. and Capoeta sp.), in contrast to small vertebrae that can represent all taxonomic groups (see figure 14). An example of normally distributed histograms of vertebrae width from various genus of Cyprinidae is presented in figure 14. Since this graph represent different populations two peaks are observed: one representing small cyprinids, regardless of their genus/ species, and the other representing only large Barbus sp. and Capoeta sp.. Figure 14: An hypothetical histogram of vertebra width normal distribution expected from

50 Acanthobrama sp. 40 Barbus/Capoeta sp. 30

20 Frequency (%) 10

0 0.1-1 1.1-2 2.1-3 3.1-4 4.1-5 5.1-6 6.1-7 7.1-8 8.1-9 9.1-10 10.1-11 11.1-12 12.1-13 Vertebra Width (mm) Acanthobrama sp. and large cyprinids (Barbus sp. and Capoeta sp.).

28 CHAPTER 5: FISH REMAINS RECOVERED AT OHALO-II "Fish carcasses are more vulnerable to decay than other vertebrates" (Schafer, 1972:49) I identified 17,073 bones recovered from seven different structures as follow: loci 1, 2, 3 (brush huts), 7, 9 (ashes), 8 (unidentified pit), and a small sample of fish remains from the sediments located between loci 7 and 9 (Table 8). 5.1. Taxonomic identification: The distribution of fish by family and location appear in table 8. Two families of primary and secondary freshwater fish are represented: Cyprinidae and Cichlidae. Cyprinidae are highly abundant in loci 1, 2 and 9, while in locus 8 it changes and Cichlidae are more abundant. In loci 3 and 7 the two families' relative abundance is similar (Figure 15). I examined these differences with contingency table and found that they are significant (χ2= 4009.100; p<0.001, df=6), especially for loci 1,3,7 and 8 and for Cichlidae recovered from the sediments between L.7 and L.9 (Table 8). Table 8: Frequency (NISP) and percentage of fish identified at Ohalo-II by family and loci. Cyprinidae Cichlidae Total Location NISP % NISP % NISP Locus 1 10,948* 93.76 728* 6.23 11,676 Locus 2 48 82.75 10 17.24 58 Locus 3 344* 55.84 272* 44.15 616 L.7 & L.9 15 51.72 14* 48.27 29 Locus 7 2,178* 53.00 1,932* 47.00 4,110 Locus 8 249* 46.37 288* 53.63 537 Locus 9 36 76.60 11 23.40 47 Total 13,818 80.93% 3,255 19.07% 17,073

* significantly different (χ2>12.0, df=6).

100 90 Cichlidae Figure 15: Relative abundance (%) of 80 Cichlidae and Cyprinidae by loci. Cyprinidae 70 60 50 40

Frequency (%) 30 20 10 Identification of fish remains to the 0 Structure genus and species level has been

Locus 1 Locus 2 Locus 9 Locus 3 Locus 7 Locus 8 possible on 4836 bones (Table 9). I identified 8 species and observed the following patterns: 1. All taxa identified are abundant at the recent Sea of Galilee.

29 2. Two of the identified taxa are endemic to the Sea of Galilee: Tristamella sp., and Acanthobrama terraesanctae (Table 9). 3. The numbers of species identified in each structure (species richness) differ among loci (Table 10). Locus 7 exhibits the highest species richness (8). 4. Brillouin diversity index exhibits highest value for locus 7, and lowest value for locus 1 (Table 10). 5. The most common species identified at the site is the Kinneret bleak Acanthobrama terraesanctae. However, this species is abundant only in locus 1 (brush hut) and appear in lower frequencies in locus 7 (ashes). 6. Among the Cyprinidae, the large cyprinids (Barbus sp. and Capoeta sp.) are abundant in loci 2,3,7,8, and 9 (Table 9). 7. Cichlid species richness is low in all loci, except for locus 7 (Table 10). This can be attributed to the few species-specific bones. Table 9: Frequency (NISP) and percentage of fish remains identified by genus and loci. Total NISP by Excavated area Identified fish NISP % 1 2 3 7/9 7 8 9 Cichlidae Tilapia sp. 9 0.18 0 0 2 0 5 2 0 Tilapia aurea 2 0.04 0 0 0 0 2 0 0 T. zilli 9 0.18 0 0 0 0 9 0 0 S. galilaeus 3 0.06 1 0 0 0 2 0 0 Tristamella sp. 43 0.89 6 0 9 0 19 9 0 Cyprinidae Barbus sp. 178 0.38 53 1 21 0 85 18 0 B. canis 35 0.72 11 1 5 0 11 7 0 B. longiceps 87 1.8 17 8 10 0 37 10 5 C. damascina 306 6.3 78 16 25 0 131 50 6 Barbus/Capoeta 1,126 23.28 400 18 271 15 275 137 10 A. terraesanctae 3,038 62.82 2,917 0 0 0 119 1 1 Total 4,836 100% 3,483 44 343 15 695 234 22

As a correlation between sample size (NISP) and species richness exists, I examined the effect of sample size on taxonomic diversity. NISP data from loci 1, 2, 3, 7, 8, and 9 was used to calculate rarefaction analysis, with Analytic Rarefaction V.1.3 by Holland (2001; Figure 16). The rarefaction analysis demonstrates that sample size biased species richness for loci 2, 8, and 9. Therefore, these samples should be enlarged and re-analyzed in the future. However, for loci 1, 3, and 7, species richness is not biased due to sample size. Interesting to note is that although locus 1 had the largest sample size (NISP=11,676 bones), it's species richness is smaller compare to other structures with much lower NISP (Table 10, Figure 16). Table 10: NISP values of identified bones, species richness, Shannon Wiener Function, and Brillouin Index, calculated for each locus, at Ohalo-II.

30 Locus no. NISP Species Richness Shannon Wiener Brillouin Index Function 1 3,438 6 2.7 0.865 2 44 3 - 1.6 3 343 5 3.0 1.17 7 695 8 3.4 2.398 8 234 6 3.11 1.778 9 22 3 - 1.567 14 13

7 Locus 1 Locus 7 Species richness 6 Locus 2 Locus 8 5 Locus 3 Locus 9 4 0

150 300 450 600 750 900 NISP 1050 1200 1350 1500 1650 1800 1950 2100 2250 2400 2550 2700 2850 3000 3150 3300 3450 3600 3750 3900 Figure 16: Rarefaction curves for species richness and loci as a function of NISP.

Because taxonomic identification to species level was possible on a small sample, I grouped the identified bones as follows: Acanthobrama sp., small cyprinids, large cyprinids (Barbus sp. and Capoeta sp.) and cichlids. This taxonomic grouping enabled me to conduct further calculations on the fish remains recovered at Ohalo II. For each taxonomic group I calculated the relative abundance (%) and their ranking order, according to their location at the site (Tables 11-12 ). Based on the highly abundant and ranked taxonomic group I can separate the structures to those dominated by cichlid remains (loci 7 & 8) vs. structures dominated by large cyprinids (Loci 2 & 3). Locus 1 displayed a distinct pattern, since, it is the only structure that exhibits dominance of small cyprinid and Acanthobrama sp. (Tables 11-12). The relatively high abundance of small cyprinids and the appearance of Acanthobrama sp., is also observed in locus 7, although in lower frequencies. Table 11: NISP and percentage calculated by loci in four taxonomic groups. Taxonomic Locus 1 Locus 2 Locus 3 Locus 7 Locus 8 Group NISP % NISP % NISP % NISP % NISP % Acanthobrama sp. 2917 24.98 0 0.00 0 0.00 119 2.90 1 .19 Small Cyprinid 7472 63.99 4 6.90 12 1.95 1520 36.98 26 4.84 Large Cyprinid 559 4.79 44 75.86 332 53.90 539 13.11 222 41.34 Cichlid 728 6.24 10 17.24 272 44.16 1932 47.01 288 53.63

31 Total 11676 100% 58 100% 616 100% 4110 100% 537 100%

Table 12: Ranking order of taxonomic groups identified in different loci. Taxonomic Locus 1 Locus 2 Locus 3 Locus 7 Locus 8 Group Rank Rank Rank Rank Rank Small Cyprinid 1 3 3 2 3 Acanthobrama sp. 2 4 4 Large Cyprinid 4 1 1 3 2 Cichlid 3 2 2 1 1

5.2. Skeletal Representation Based on NISP values, I calculated the relative abundance of skeletal elements recovered from the different structures at Ohalo II (Appendix X- XI). I observed a high correlation between NISP and skeletal elements richness (Spearmen correlation r=.886, p=0.0476), among loci at Ohalo-II. The structures with high NISP (Loci 1 and 7) exhibit highest value for skeletal element richness. In addition, locus 8 also exhibit high values of skeletal elements richness, despite the relatively low NISP (n=537; Table 13). Since the number of skeletal elements identified highly correlate with NISP, by performing rarefaction analysis (following Holland, 2001), I examined which sample is biased. The results presented in Figure 17 clearly demonstrate that the number of skeletal elements identified from loci 2, and 9 are biased due to sample size. Sample from locus 8 exhibit relatively high skeletal element richness despite it's small NISP. Samples from loci 1, 3 and 7 are not biased (Figure 17). Table 13: Identified NISP and number of skeletal elements identified (richness) by loci. Excavated area Identified NISP No. of Skeletal elements identified Locus 1 11,676 74 Locus 2 58 8 Locus 3 616 30 Locus 7 4,110 58 Locus 8 537 46 Locus 9 47 11

32 64 60 56 52 48 44 40 36 32 28 24 Locus 1 Locus 2 20

No. of Skeletal elements 16 Locus 3 Locus 7 12 8 Locus 8 Locus 9 4 0 200 400 600 800 1000 1200 1400 1600 1800 2000 NISP Figure 17: Rarefaction curve for number of skeletal elements identified and loci as a function of NISP.

Therefore, I will address the issue of skeletal completeness for loci 1 (brush hut), 7 (ashes), and 8 (unidentified pit). Skeletal completeness analyses include: ranking order (based on NISP) of the most abundant cranial and post cranial bones, relative abundance according to anatomic location, and ratio of cranial to postcranial elements. In order to enlarge sample size, all calculations refer to four taxonomic groups: Acanthobrama sp., small cyprinids, Large cyprinids (Barbus sp. & Capoeta sp.) and Cichlids. 5.2.1. Skeletal completeness in brush hut 1: I calculated the relative abundance (NISP) of 12,597 bones from brush hut-1, according to four taxonomic groups (Appendix XI). Overall, I identified 74 skeletal elements that varied between the taxonomic groups. The most diverse group is that of small cyprinids with 63 skeletal elements. For the large cyprinids (Barbus sp./Capoeta sp.) 48 skeletal elements were identified, compare to 32 of Acanthobrama sp. and 31 of Cichlidae (Figure 18). An interesting feature is the absence of scales and otoliths (Appendix XI). A single otolith of a Barbus sp./Capoeta sp. was recovered from locus 1.

33 Cichlids

Acanthobrama sp.

Large Cyprinids

Small Cyprinids

010203040506070 No. of Skeletal elements Figure 18: Number of skeletal elements identified in locus 1 by taxa.

The low numbers of skeletal elements identified for species such as Acanthobrama sp. is not surprising since it is based on species-specific bones. This result can be better understood in view of the high diversity of skeletal elements representing small cyprinids. However, the relative low diversity of skeletal elements identified as Cichlids and large Cyprinids, is surprising.

34 Table 14: NISP and ranking order of Cyprinidae and Cichlidae cranial and postcranial bones from locus 1. Locus 1 Cyprinidae Skeletal element Acanthobrama Small Large Cichlidae NISP Rank NISP Rank NISP Rank NISP Rank Atlas 387 1 17 36 5 36 6 5th vertebrae 284 2 44 7 0 Ceratohyal 272 3 8 1 0 Pharyngeal bone 261 4 59 2 4 3rd Vert. 236 5 0 0 3 Axis 224 6 3 44 4 7 Opercle 144 7 12 0 1 Posttemporal 111 8 9 1 0 Hyomandibular 94 9 14 0 1 Os suspensorium 89 10 11 3 0 Thoracic Vert. 2 2156 1 71 2 270 1 Precaudal Vert. 4 1105 2 18 6 60 3 Precaudal/Caudal Vert. 23 597 3 114 1 138 2 Caudal Vert. 0 431 4 50 3 47 4 Scapula 0 419 5 5 0 Cranial bone-gen. 0 353 6 0 0 Tripus 0 270 7 13 0 4th Vert. 3 202 8 2 0 Fin ray 0 178 9 3 1 Angular/Articular 43 6 14 7 3 Fin Spine 0 0 5 40 5 Ultimate Vert. 0 22 1 18 7 Penultimate Vert. 1 71 6 13 8

These results can be better understood by comparing the ranking order of the most abundant bone in each taxonomic group (Appendix XI; Table 15). For Acanthobrama sp. the first and fifth vertebrae are the most abundant bones, followed by ceratohyal, pharyngeal bone and third vertebra (Table 15). In contrast, in the other taxonomic groups, the highest-ranking bones are the thoracic and precaudal vertebrae (Table 15). In small cyprinids the scapula and tripus are also abundant. In order to examine the relative representation of each anatomic region, the bones were grouped in 8 regions. Than I calculated their frequency distribution, relative abundance, and survival index, in comparison with the expected ratio calculated according to the formula described in chapter 4.5.3 (Tables 15-16; Figures 19-20). In the case of Acanthobrama sp., most anatomic regions are present. The hyoid and opercle regions are over-represented while the neurocranium is under-represented (Figure 19). I received a significant difference in body part representation for most anatomic regions, except for the oromandibular region and median fins, which resemble the expected in a complete fish (Table 16). Table 15: Frequency (NISP) and percentage of skeletal elements recovered from locus 1 for anatomic regions and taxonomic groups. Cyprinidae L.1 Anatomic region Total Acanthobrama Small Large Cichlidae

NISP % NISP % NISP % NISP % NISP %

35 Crania Neurocranium 713 6.11 186 6.38 507 6.79 7 1.25 13 1.79 Branchial region 703 6.01 425 14.57 194 2.58 80 14.3 4 0.55 Hyoid region 536 4.60 456 15.63 66 .90 8 1.43 6 0.82 Oromandibular region 338 2.89 252 8.64 48 0.64 29 5.19 9 1.24 Opercular series 187 1.60 146 5.01 35 0.47 2 0.36 4 0.55 Postcrania Appendicular skeleton 685 5.87 129 4.42 538 7.20 9 1.61 9 1.24 Median fins 635 5.4 46 1.58 555 6.92 21 3.76 51 7.01 Weberian apparatus 460 3.9 89 3.05 351 4.7 20 3.6 - - Vertebral column 7,419 63.5 1,188 40.7 5,216 69.8 383 68.5 632 86.8 Total 11,676 100% 2,917100% 7,472100% 559 100% 728 100%

Table 16: Survival index (SI) and p (calculated by chi-square test) for fish remains from locus 1. Cyprinidae L.1 Anatomic region Acanthobrama sp. Small Large Cichlidae Crania SI P SI P SI P SI P Neurocranium .268 0.001 .307 0.001 .057 0.001 .073 0.001 Branchial region .624 0.001 .103 0.001 .570 0.001 .030 0.001 Hyoid region 2.091 0.001 .128 0.001 .207 0.001 .100 0.001 Oromandibular region .924 ns .074 0.001 .599 0.01 .121 0.001 Opercular series 1.339 0.001 .135 0.001 .103 0.001 .134 0.001 Postcrania Appendicular skeleton .591 0.001 1.040 ns .232 0.001 .151 0.001 Median fins .844 ns 2.451 0.001 1.240 ns .651 0.005 Weberian apparatus .653 0.001 1.085 ns .826 ns - - Vertebral column 2.235 0.001 3.583 0.001 3.517 0.001 5.643 0.0001 SI= ( no. of observed bones ) under-representation<1< =over-representation no. of expected bones

80 Acanthobrama sp. -obs. 70 60 Acanthobrama sp.- exp. 50 40 30 20 Frequency (%) 10 0 Anatomic region Median fins Hyoid region Neurocranium Branchial region Opercular series Vertebral column Weberian apparatus Oromandibular region Appendicular skeleton Figure 19: Observed and expected percent of anatomic regions in Acanthobrama sp. at locus 1.

36 80 Barbus/Capoeta-obs. 70 60 Barbus/Capoeta-exp. 50 40 Small cyprinid-obs. 30 20 Frequency (%) 10 0 Anatomic region Median fins Hyoid region Neurocranium Branchial region Opercular series Vertebral column Weberian apparatus Oromandibular region Appendicular skeleton Figure 20: Observed and expected percent of anatomic regions in Barbus sp./Capoeta sp. at locus 1.

In large Cyprinids all cranial regions are under-represented (Figure 20). I received a significant difference for most anatomic regions, except for the median fins and the Weberian apparatus, where the observed and expected values are similar. In small cyprinids, all postcranial regions are over-represented. I received a significant difference for most anatomic regions, except for the appendicular skeleton and Weberian apparatus, where the observed and expected values are similar. Cichlid remains from locus 1 include mostly vertebrae (86.8%) that display a high SI value. However, most of their cranial and postcranial bones are absent, and significantly differ from the expected ratio (Table 16). I further examined the relative abundance of cranial and postcranial elements in locus 1, in comparison with a complete fish, and observed a significant difference for all taxonomic groups (Table 17, Figure 21; chi-square test, df=1). The under-representation of cranial elements is observed from their low SI calculated for cichlids, small cyprinids, and large cyprinids (p<0.001, χ2 test). In Acanthobrama sp. although the cranial region is under- represented, and significantly different from the expected ratio in a complete fish (z=9.3858, p<0.001) its SI is relatively high (SI=0.761) in comparison with other taxa (Table 17; Figure 21). Since Acanthobrama sp. is a small pelagic fish (T.L.<20 cm), the cranial bones' SI indicates that bone survival from locus 1 was not biased due to bone/fish size. It also indicates that the low survivorship of cranial bones from cichlids and large cyprinids should be further investigated in regard to depositional processes and human activity.

37 Table 17: Frequency (NISP), percentage, and SI calculated for cranial and postcranial bones in four taxa at locus 1. Brush Hut 1 Cranial Postcranial Total Taxonomic group NISP % SI NISP % SI NISP Acanthobrama 1,465 50.2 0.761* 1,452 49.8 1.464* 2,917 Small Cyprinid 850 11.4 0.181* 6,622 88.6 2.395* 7472 Large Cyprinid 126 22.5 0.358* 433 77.5 2.094* 559 Cichlid 36 4.9 0.078* 692 95.1 2.569* 728 Chi-square test: significantly different from the expected value p<0.001

Acanthobrama sp- L.1 Large Carp-L.1 Cichlids-L.1

100 Acanthobrama sp. -expected Large Carp-expected Cichlids-expected 90 80 70 60 50 40 Frequency (%) 30 20 10 0 Cranial Postcranial Cranial Postcranial Cranial Postcranial Figure 21: Observed and expected percent of cranial and postcranial bones in Acanthobrama sp., large cyprinids, and cichlids at locus 1.

5.2.2. Skeletal completeness in Locus 7: I calculated the relative abundance of 5478 skeletal elements recovered from locus 7 according to four taxonomic groups (Appendix XII). I identified 58 skeletal elements, with diverse representation in each taxonomic group (Figure 22). The most diverse group is that of large cyprinids (Barbus sp./ Capoeta sp.) with 40 skeletal elements. For cichlid I identified 36 skeletal elements, and 29 for small cyprinids. Only 119 bones and 11 skeletal elements were identified as Acanthobrama sp. Figure 22: Number of skeletal elements identified for locus 7 (ashes) by taxa.

Large Cyprinids

Cichlids

Small Cyprinids

Acanthobrama sp.

0 102030405038 No. skeletal elements I examined the survivorship of skeletal elements in each taxonomic group and observed that Acanthobrama sp. is represented mainly by a few pharyngeal bones, and the 3rd and 5th vertebrae. A similar pattern is observed for small cyprinids (the remains consisted mainly of vertebrae centrum). The variety of skeletal elements in large cyprinids and Cichlids is much higher and include cranial and postcranial bones (Appendix XII). I also identified 2 cichlid otoliths in locus 7. The vertebrae are the most abundant element in all taxonomic groups, while scales are totally absent (Appendix XII). To examine the relative representation and survivorship of anatomic regions, I grouped the bones into eight regions (Table 18), and with chi-square test compared the relative abundance of each region with the expected ratio in a complete fish (Table 19). I observed that the vertebral column is over-represented in all taxonomic groups (Figures 23 24), and significantly differs from the expected values (Table 19). Moreover, the neurocranium, hyoid, and opercular regions are under-represented in all taxonomic groups, and therefore significantly differ from the expected values (Tables 18-19 ). The branchial region is over- represented in large carp due to the relatively high number of pharyngeal teeth, contrary to the actual low number of most pharyngeal bones (Table 19; Appendix XII). A similar preservation bias occurred for Acanthobrama sp. (Table 19). The median fin is also over- represented due to the high number of ribs and pterygiophores. In Acanthobrama sp. and small cyprinids the median fin appear in similar value to the expected ones, while in large cyprinids and cichlids the observed value significantly differs from the expected (Table 19).

39 Table 18: Frequency (NISP) and percentage of skeletal elements recovered for anatomic regions and taxonomic groups in locus 7. Cyprinidae L.7 Anatomic region Total Acanthobrama Small Large Cichlidae NISP % NISP % NISP % NISP % NISP % Crania Neurocranium 23 0.55 1 0.84 3 0.19 3 0.55 16 0.83 Branchial region 203 4.94 21 17.65 23 1.51 149 27.64 10 0.52 Hyoid region 35 0.85 0 0.00 1 0.07 15 2.78 19 0.98 Oromandibular region 82 2.00 0 0.00 4 0.26 47 8.72 31 1.60 Opercular series 19 0.46 0 0.00 0 0.00 2 0.37 17 0.88 Postcrania Appendicular skeleton 92 2.24 1 0.84 7 0.46 13 2.41 71 3.67 Median fins 648 15.77 1 0.84 54 3.55 63 11.69 53027.43 Weberian apparatus 17 0.41 1 0.84 10 0.66 6 1.11 - - Vertebral column 2,991 72.77 94 78.99 1,418 93.29 241 44.71 1,238 64.08 Total 4,110 100% 119 100% 1,520 100% 539 100% 1,932 100%

Table 19: Survival index (SI) and p (calculated by chi-square test) for fish remains from locus 7. Cyprinidae L.7 Anatomic region Acanthobrama sp. Small Large Cichlidae Crania SI p SI p SI p SI p Neurocranium .035 0.001 .009 0.001 .025 0.001 .034 0.001 Branchial region .755 ns .060 0.001 1.101 ns .028 0.001 Hyoid region 0.0 0.005 .009 0.001 .402 0.001 .120 0.001 Oromandibular region 0.0 0.001 .030 0.001 1.007 ns .156 0.001 Opercular series 0.0 0.05 0.0 0.001 .107 0.001 .214 0.001 Postcrania Appendicular skeleton .112 0.01 .066 0.001 .348 0.001 .448 0.001 Median fins .450 ns 1.172 ns 3.857 0.001 2.547 0.001 Weberian apparatus .180 ns .152 0.001 .257 0.001 - - Vertebral column 4.334 0.025 4.789 0.001 2.295 0.001 4.165 0.0001

40 100 Barbus/Capoeta -obs. 90 80 Barbus/Capoeta -exp. 70 60 Small Cyprinid- obs. 50 40 30 Frequency (%) 20 10 0 Anatomic region Median fins Hyoid region Neurocranium Branchial region Opercular series Vertebral column Weberian apparatus Oromandibular region Appendicular skeleton Figure 23: Observed and expected percent of anatomic regions in Barbus sp., Capoeta sp. and small cyprinids at locus 7.

100 90 Cichlid-obs. 80 Cichlid-Exp. 70 60 50 40 Frequency (%) 30 20 10 0 Anatomic region Median fins Hyoid region Neurocranium Branchial region Opercular series Vertebral column Oromandibular region Appendicular skeleton Figure 24: Observed and expected percent of anatomic regions in Cichlidae in locus 7. I calculated the ratio of cranial and postcranial elements in each taxonomic group (Figure 25), and compared their abundance with the expected from a complete fish, by using chi-square test (Table 20). The cranial and postcranial regions observed frequencies significantly differ from the expected values in all taxonomic groups (Table 20). Regardless

41 of taxa, the postcranial region is over-represented while the cranial region is under- represented (Figure 25). Although a relatively high ratio of cranial region is observed for Barbus sp./Capoeta sp. (SI= 64% ; Table 20), it is still significantly different from the expected value in a complete cyprinid. Table 20: Frequency (NISP), percentage, and SI calculated for cranial and postcranial bones in four taxa at locus 7. Locus 7 Cranial Postcranial Total Taxonomic group NISP % SI NISP % SI NISP Acanthobrama 22 18.5 .280* 97 81.5 2.397* 119 Small Cyprinid 31 2.04 .032* 1,489 97.96 2.648* 1,520 Large Cyprinid 216 40.07 .636* 323 59.93 1.620* 539 Cichlid 83 4.8 .076* 1,839 95.2 1.511* 1,932 *Chi-square test: significantly different from the expected value p<0.001

Acanthobrama sp. - L.7 Large Cyprinid-L.7 Cichlids-L.7

100 Acanthobrama sp.- exp. Large Cyprinids-exp. Cichlids-exp. 90 80 70 60 50 40 Frequency(%) 30 20 10 0 Cranial Postcranial Cranial Postcranial Cranial Postcranial Figure 25: Observed and expected percent of cranial and postcranial bones in Acanthobrama sp., large cyprinids, and cichlids at locus 7. 5.2.3. Skeletal completeness in Locus 8: I calculated the percentage (NISP) of 608 bones from locus 8, according to four taxonomic groups (Appendix XIII). Despite the relatively small sample size , I identified 47 skeletal elements (Figure 26). Skeletal elements diversity vary between taxonomic groups (Figure 26). While Acanthobrama sp. is represented by a single vertebrae, small cyprinid are represented by 26 bones from 17 skeletal elements, cichlid are represented by 29 skeletal elements, and large cyprinids are the most diverse group represented by 32 skeletal elements (Figure 26).

42 Large Cyprinids

Cichlids

Small Cyprinids

Acanthobrama sp.

0 10203040 No. skeletal elements

Figure 26: Comparison between number of skeletal elements per taxa in locus 8.

I calculated the frequency distribution of the bones according to 8 anatomic regions, and observed that the vertebral column is over-represented in all taxonomic groups (Tables 21-22 ). I used chi-square test to compare the observed NISP with the expected, and received that some regions significantly differ from the expected (Table 22). Due to small sample size of small cyprinids and Acanthobrama sp.(<5) statistical analysis could not be carried out. While Acanthobrama sp. is represented by a single vertebra, in small cyprinids most of the anatomic regions are present, except for the hyoid region and appendicular skeleton. In large cyprinids, similarity with complete fish is observed for the Weberian apparatus, branchial, oromandibular, and appendicular regions (Table 22; Figure 27). Large cyprinid neurocranium bones are absent, while the hyoid region and opercular series are under- represented. In contrast, the vertebral column and median fin are over-represented (Table 22). In cichlids, although all cranial regions are present, most of them are under-represented, and significantly differ from what is expected (Table 22; Figure 28). The opercular series is the only cranial region which is over-represented. From the post-cranial region, the appendicular skeleton and median fin exhibit similarity with a complete fish (Table 22). Interestingly, six cichlid's otoliths were recovered. The presence of otoliths may indicate that intact skulls of cichlid's were deposited in locus 8. It may also indicate a different preparation method (Fred et al., 2002) in comparison with fish remains from loci 1 and 7. Table 21: Frequency (NISP) and percentage of skeletal elements by anatomic regions and taxa in locus 8. Cyprinidae L.8 Anatomic region Total Acanthobrama Small Large Cichlidae NISP % NISP % NISP % NISP % NISP % Crania

43 Neurocranium 12 2.2 0 0 2 7.7 0 0 10 3.47 Branchial region 59 10.99 0 0 3 11.54 46 20.72 10 3.47 Hyoid region 9 1.68 0 0 0 0 5 2.25 4 1.39 Oromandibular region 40 7.45 0 0 5 19.23 23 10.36 12 4.17 Opercular series 20 3.72 0 0 4 15.38 2 0.9 14 4.86 Postcrania Appendicular skeleton 33 6.15 0 0 0 0 13 5.86 20 6.94 Median fins 69 12.85 0 0 3 11.54 25 11.26 41 14.24 Weberian apparatus 7 1.3 0 0 1 3.85 6 2.7 - - Vertebral column 288 53.63 1 100 8 30.77 102 45.95 177 61.46

Total 537 100% 1 100% 26 100% 222 100% 288 100% Table 22: Survival index (SI) and p (calculated by chi-square test) for fish remains from locus 8. Cyprinidae L.8 Anatomic region Small Large Cichlidae Crania SI p SI p SI p Neurocranium .348 ns 0.0 0.001 .141 0.001 Branchial region .460 ns .825 ns .188 0.001 Hyoid region 0.00 - .325 0.01 .169 0.001 Oromandibular region 2.221 ns 1.197 ns .406 0.005 Opercular series 4.442 0.05 .260 0.05 1.185 ns Postcrania Appendicular skeleton 0.000 - .845 ns .846 ns Median fins 3.808 0.10 3.716 0.001 1.322 ns Weberian apparatus .888 Ns .624 ns - - Vertebral column 1.579 ns 2.359 0.0001 3.995 0.0001 under-representation =<1< =over-representation

44 60 Barbus/ Capoeta-obs. 50 Barbus/Capoeta-exp. 40 Small cyprinid obs. 30

20 Frequency (%) 10

0 Anatomic region Median fins Hyoid region Neurocranium Branchial region Opercular series Vertebral column Weberian apparatus Oromandibular region Appendicular skeleton Figure 27: Relative abundance (%) of anatomical regions in complete Cyprinidae compared

70 Cichlid-obs. 60 50 Cichlid-exp. 40 30 20 Frequency (%) 10 0 Anatomic region Median fins Hyoid region Neurocranium Branchial region Opercular series Vertebral column Oromandibular region Appendicular skeleton with those observed for large and small cyprinids remains in locus 8. Figure 28: Relative abundance (%) of anatomical regions in complete Cichlidae compared with those observed in locus 8.

I calculated the value and ratio of cranial and postcranial elements (Table 23, Figure 29) in cichlid, and small and large cyprinid remains from locus 8. Using the chi-square test I compared between the observed and expected values. In large cyprinids and cichlid the postcranial region is over-represented in all taxonomic groups while the cranial region is

45 under-represented (Figure 29, Table 23; p<0.001). In small cyprinids, the ratio between cranial and postcranial bones is similar to the expected value (Table 23). This result, however, might be biased due to small sample size (NISP=26).

Table 23: Frequency (NISP), percentage, and SI calculated for cranial and postcranial bones for four taxa at locus 8. Locus 8 Total Cranial Postcranial Taxonomic group NIS NISP % SI NISP % SI P Small Cyprinid 26 14 53.8 .855 12 46.2 1.247 Large Cyprinid 222 76 34.2 .543* 146 65.8 1.777* Cichlid 288 50 17.4 .276* 238 82.6 1.312* *chi-square test: significantly different from the expected value p<0.001

Figure 29: Observed and expected percent of cranial and postcranial bones in large cyprinid and cichlid at locus 8.

5.3. MNI value

Large Cyprinid-L.8 Cichlids-L.8 I calculated the

100 MNI value for the fish Large Cyprinids expected Cichlids expected 90 identified according to 80 70 family level and four 60 taxonomic groups (Table 50 24). For each group, I 40 Frequency (%) used the most common 30 20 bone identified. From 10 the total sample of 0 Cranial Postcranial Cranial Postcranial 19,799 bones I calculated a MNI value of 942 fish at Ohalo II: 817 Cyprinidae, 104 Cichlidae and 21 unidentified fish. In loci 1 and 7 the highest MNI value is of the Kinneret bleak A. terraesanctae, followed by large numbers of unidentified small cyprinids (Table 24). It is reasonable to assume that the 214 scapula of small cyprinids, recovered from locus 1 belong to the 387 A. terraesanctae. As in locus 1 it is reasonable to assume that in locus 7, the 36 fourth thoracic vertebrae belong to the 58 A. terraesanctae identified from vertebrae no. 5. In loci 3 and 8

46 the highest MNI value is of large cyprinids, followed by cichlids (Table 24). In these loci Acanthobrama sp. and small cyprinids are either absent or with MNI value of a single fish. Table 24: MNI values, by taxa and loci, for the identified fish remains. Locus 1 Locus 2 Locus 3 Locus 7 Locus 8 Identified fish Bone MNI Bone MNI Bone MNI Bone MNI Bone MNI Acanthobrama sp. Atlas 387 - - - - 5th ver. 58 3rd ver. 1 Small Cyprinid scapula 214 Pteryg. 1 Atlas 1 4th ver. 36 Dentary 1 Large Cyprinid Axis 44 Maxilla 1 Atlas 36 Atlas 27 Atlas 10 Cichlid Atlas 36 Maxilla 1 Axis 16 Axis 42 Atlas 9 unidentified Cleithrum 5 - - Vert. 1 Basihyal 13 Basihyal 1 Total MNI 686 3 54 176 23 Comparison between the ranking order of fish identified, calculated by NISP vs. MNI (Table 25) demonstrate difference, due to each method's analytical procedure (See: Grayson, 1984). For example by using MNI estimation in locus 2, a similarity appears between the taxonomic groups, despite the differences in number of identified bones (44 bones for large cyprinid and only 4 for small cyprinid). In loci 7 and 8 MNI estimates are higher for taxonomic groups with low NISP, Table 25: Comparison between ranking order calculated from NISP and MNI in five loci. Locus 1- Locus 2- Locus 3- Locus 7- Locus 8- Ranking Ranking Ranking Ranking Ranking Identified fish MNI NISP MNI NISP MNI NISP MNI NISP MNI NISP Acanthobrama sp. 1 2 - - - - 1 5 - 5 Small Cyprinid 2 1 1 3 3 3 3 2 - 1 Large Cyprinid 3 5 1 1 1 1 4 4 1 2 Cichlid 4 4 1 2 2 2 2 1 2 1 unidentified 5 3 - - 3 1 5 3 - 3

5.4. Bone Color It is generally assumed that bone color (i.e. black, gray and white) indicate on burning activity (Shipman et al., 1984; Stiner et al., 1995). Yet, burned bones do not indicate on cooking methods but rather represent the remains of consumed animals burned as fuel following their consumption (Nicholson, 1993). In order to examine burning on fish remains from Ohalo II, I recorded bone's color on 15,503 bones recovered from seven loci (Table 26). Most of the bones (70.6%) were recorded as brown, and therefore did not indicate burning. A small sample (ca. 20%) exhibited various colors ranging from dark brown and black, and I had to examine whether they should be regarded as burned, or manganese stained. By using a Fourier Transform Infra-red (FTIR) technique, I examined three samples of bones from Ohalo-II that displayed the following colors: "dark-brown", "black" and "gray- white". Their mineralogical content was compared with Prof. S. Weiner mineral library standards, demonstrating that bones with black and dark-brown colors were neither burned nor fresh. The FTIR analyses of the gray and white bones indicate that these bones were indeed burned. This means that ca. 5% of the bones found at Ohalo-II were burnt either

47 following their consumption or as food residue. Identification of heated bones due to cooking is more complicated (Koon et al., 2003), especially since a recent study failed to discriminate between boiled and buried bone (Roberts et al., 2002). Table 26: Frequency (NISP) of bone colors by loci. Locus 1 2 3 7/9 7 8 9 Total Color NISP NISP NISP NISP NISP NISP NISP NISP % Black 244 4 65 1 293 22 4 633 4.08 Gray 42 0 92 22 493 11 0 660 4.26 White 35 0 15 7 152 3 1 213 1.37 Dark brown 1124 1 227 4 1014 148 7 2525 16.29 Brown 6942 47 498 5 3078 343 40 10953 70.65 Light Brown 44 0 1 1 20 2 0 68 0.44 Orange-Brown 59 0 25 0 334 28 5 451 2.9 Total 8490 52 923 40 5384 557 57 15503 100%

Comparison between the relative abundance of gray and white bones, in each structure, reveal that most of the burned bones were recovered from loci 7, 3 and the area between Loci 7 and 9. Loci 1,2, 8 and 9 hardly exhibit any bones with burning signs (Table 26). In addition to burning signs, bones with orange-brown color were also recorded and exhibit to be heavily abraded. Their abundance is relatively high for loci 7, 8, and 9 (Table 26). These bones were probably exposed to air, in periods that sea level was low and the site was revealed. 5.5. Fragmentation pattern: Analysis of the bones fragmentation according to five categories (Table 27) demonstrates that at Ohalo-II, 23% of the bones are highly fragmented, 53% are slightly fragmented and only 14% are complete. Moreover, the bones state of fragmentation significantly differ between loci (Table 28; p=0.0001; χ2=2219.637; df=16). In locus 7 most of the bones were slightly fragmented (ca. 80%; Table 28). In locus 1 highly fragmented bones compose 26% and in loci 3 and 8 they compose 43% and 37% respectively (Table 28). In locus 1, 45% of the bones are slightly fragmented, while in loci 3, 8, and 9 the ratio is lower (ca. 30%). Table 27: Frequency (NISP) and percentage of bones state of fragmentation. Fragmentation Category (%) NISP Frequency(%) 5-29% 2,110 11.04 30-50% 2,442 12.77 51-70% 1,822 9.53 71-89% 10,088 52.77 90-100% 2,656 13.89 Total 19,118 100%

Table 28: Bones state of fragmentation by loci. Fragmentation L.1 L.3 L. 7 L. 8 L. 9 Category (%) NISP % NISP % NISP % NISP % NISP %

48 5-29% 1393 11.45 359 38.8 225 4.23 105 18.26 17 29.82 30-50% 1835 15.09 41 4.43 431 8.10 106 18.43 3 5.26 51-70% 1378 11.33 89 9.62 297 5.58 48 8.35 3 5.26 71-89% 5494 45.2 357 38.6 3981 74.7 222 38.6 18 31.58 90-100% 2062 16.95 79 8.54 390 7.33 94 16.35 16 28.07 Total 12162 100% 925 100% 5324 100% 575 100% 57 100%

Considering the bones differential state of fragmentation in each structure (Table 28), and the different NISP, I will address the bone's state of fragmentation for loci 1, 7, and 8. In each locus, the weighted mean index of fragmentation (WMI) will be examined according to four taxonomic groups: Acanthobrama sp., Barbus sp./Capoeta sp., small cyprinid, and Cichlid (Appendix XIV-XIX). 5.5.1. Bone fragmentation in locus 1: From the 12,162 bones analyzed from this locus the general pattern is of good preservation (>80% of the bone is present; Appendix XIV- XIX). However, the bones of cichlids and small cyprinids are less fragmented compared to Acanthobrama sp. and large cyprinids remain (Figure 30). A significant difference was found between the four taxonomic group (Friedman test: χ2=13.560; df=3; p=0.0036).

70 Acanthobrama sp. (n=2753) 60 Cichlids (n=728) 50 Large Cyprinid (n=505) 40 Small Cyprinid (n=7255) 30 Frequency (%) 20

0 05-29 30-50 51-70 71-89 90-100 % Fragmentation

Figure 30: Bones fragmentation patterns for four taxa in locus 1.

I calculated the weighted mean index (WMI) of fragmentation (See Table 29; Appendix XV) to examine the degree of fragmentation according to skeletal elements for each taxonomic group. For comparison between the results, I ranked the best preserved skeletal elements (>80%) with NISP larger than 30, in each taxonomic group (Table 30). Table 29: An example of WMI calculated for Acanthobrama sp., by fragmentation classes. Acanthobrama sp. Relative Abundance of Fragments by Size Skeletal element NISP 5% 15% 25% 35% 45% 55% 65% 75% 85% 95% WMI Atlas 387 0.3 0.0 0.0 0.0 0.0 0.5 0.5 3.1 66.7 28.9 87.1 5th Vertebrae 284 0.0 0.0 0.0 0.3 0.7 2.8 10.9 19.0 5.9 60.2 85.6

49 Epihyal 48 0.0 0.0 0.0 0.0 0.0 0.0 2.1 0.0 12.5 85.4 93.1 Atlas WMI = [(0.26 * 5) + (0.52 * 55) + (0.52 * 65) + (3.1 * 75)+(66.67 * 85)+(28.94 * 95))/100]=87.12%

50 Table 30: Ranking order by taxa of the best preserved bones (>80%) recovered from locus 1. Skeletal element NISP WMI Skeletal element NISP WMI Rank Small Cyprinid (%) Acanthobrama sp. (%) 1 Basapophysis 47 93.0 Epihyal 48 93.1 2 Penultimate vert. 71 90.0 Atlas 387 87.1 3 Scapula 419 89.0 5th vert. 284 86.0 4 4th vert. 202 87.0 Palatine 62 85.0 5 Caudal vert. 431 86.0 3rd vert. 238 84.0 6 6th vert. 54 85.0 Axis 224 84.0 7 Tripus 270 84.0 Ceratohyal 272 80.0 8 Thoracic vert. 2169 80.0 Cichlid Large cyprinid 1 Atlas 35 85.0 Caudal vert. 50 80.0 2 Thoracic vert. 275 81.0 3 Precaudal & Caudal vert. 413 80.0

Comparison between the ranked bones demonstrate the following (Appendix XV): 1. The number of well preserved skeletal elements differs between taxonomic groups. For small cyprinid, I identified 8 well preserved skeletal elements, for Acanthobrama sp. 7, for cichlids only 3, and for large carps only the caudal vertebrae is well preserved (Table 30). 2. Except for large cyprinids, in all other taxonomic groups, one of the first six thoracic vertebrae is well preserved (Table 30). In general, vertebrae centrum is well preserved (WMI≈80%), while the spines are usually fragmented. 3. In small cyprinids the scapula and tripus are well preserved, in contrast with their poor preservation and low representation in large cyprinids (5 scapula-WMI=39%; 13 tripus- WMI=74.2%). 4. In small cyprinids the most fragmented bones are the pharyngeal bone and ribs (WMI=29% ; 33%). The pelvis and hyomandibuar are highly fragmented in Acanthobrama sp. (WMI=45%; 32%), cichlid (WM=25%), and unidentified bones (WMI=15%). In large cyprinids the angular and pelvis are fragmented (WMI=50%; 20%). Small fragments of vertebrae, are present in all taxonomic groups (Table 30). I calculated the WMI of fragmentation to each anatomic region, and compared it with the SI (Table 31). I received no correlation between WMI and SI, demonstrating that most anatomic regions with high SI does not result from high fragmentation rate (Table 31). For example, the vertebral column is always over-represented and exhibit high WMI (>70%) testifying that most of the vertebra is present. However, an exceptional correlation is observed for Acanthobrama sp. opercular region. This is the only region that might be over- represented due to fragmentation (Table 31). Table 31: Comparison between WMI and SI values in locus 1 by anatomic regions and taxa. Cyprinidae

51 L.1 Anatomic region Acanthobrama sp. Small Large Cichlidae Crania WMI SI WMI SI WMI SI WMI SI Neurocranium 65.91 .268 50.62 .307 57.86 .057 75.00 .073 Branchial region 55.54 .624 41.09 .103 51.79 .570 67.50 .030 Hyoid region 70.86 2.091 56.82 .128 83.75 .207 50.00 .100 Oromandibular region 69.92 .924 56.67 .074 52.93 .599 43.89 .121 Opercular series 56.92 1.339 51.57 .135 65.00 .103 42.50 .134 Postcrania Appendicular skeleton 48.26 .591 79.37 1.040 46.11 .232 50.56 .151 Median fins 57.39 .844 48.28 2.451 42.62 1.240 66.76 .651 Weberian apparatus 66.57 .653 84.43 1.085 78.00 .826 - - Vertebral column 85.55 2.235 78.88 3.583 70.85 3.517 80.93 5.643 WMI vs. SI r2=0.42, p=0.06 r2=0.20, p=0.22 r2=0.02, p=0.71 r2=0.36, p=0.11

5.5.2. Bone fragmentation in locus 7: The 5324 bones analyzed from structure 7 demonstrate that most of them are well preserved (>80%; Appendix XVI). I examined the fragmentation pattern according to taxonomic groups, and encountered that small cyprinid and Acanthobrama sp. demonstrate higher preservation rate (Figure 31).

100 Acanthobrama sp .(n=119) 90 80 Cichlids (n=1932) 70 Large Cyprinid (n=402) 60 50 Small Cyprinid (n=1503) 40

Frequency (%) 30 20 10 0 05-29 30-50 51-70 71-89 90-100 % Fragmentation Figure 31: Bones fragmentation patterns for four taxa in locus 7. Skeletal elements of large cyprinid and cichlid demonstrate lower rate of preservation and increase rate of fragmentation (Figure 31). These differences were found to be statistically significant (Friedman test: p=0.0293; χ2= 9; df=3). Bone WMI of preservation for each taxonomic group found in locus 7 was calculated (Appendix XVI- XVII). The results demonstrate the follows: 1. Most of the bones from Acanthobrama sp. are well preserved (Figure 31). The best- persevered bones include the third vertebrae (WMI=93.2%; NISP=21), atlas and first dorsal pterygiophore (WMI=85%). The basioccipital bone from the neurocranium region is highly fragmented (WMI=55%; NISP=5).

52 2. In small cyprinids most of the vertebrae are well preserved (WMI>84%), except for the atlas and 5th vertebrae (WMI=75%). The first dorsal pterygiophore is heavily fragmented (WMI=38.3%), opposite from the pattern observed in Acanthobrama sp.. 3. Although many skeletal elements were identified as large cyprinids (40), most of them are highly fragmented (Figure 31). The best-preserved bone is the caudal vertebrae centrum (WMI=83.3%; NISP=42). 4. In cichlids, the vertebrae are well preserved (WMI>83%). The basioccipital, that was found to be highly fragmented for the Acanthobrama sp. is well preserved for the cichlids (WMI=86%). The pelvis and pelvic spines are fragmented, as well as many of the cranial and postcranial bones. I calculated the WMI of fragmentation for each anatomical region, and compared it with the SI (Table 32). No correlation was obtained between WMI and SI, demonstrating that fragmentation did not influence high SI (Table 32). As in locus 1, the vertebral column which is over-represented (SI>200%), exhibits low fragmentation (WMI>75%). However, an exceptional correlation is observed for the median fin region of cichlids, small and large cyprinids, exhibiting that the high SI may be due to fragmentation (Table 32).

Table 32: Comparison between WMI and SI values in locus 7 by anatomic regions and taxa. Cyprinidae L.7 Anatomic region Acanthobrama sp. Small Large Cichlidae Crania WMI SI WMI SI WMI SI WMI SI Neurocranium 55.00 .035 78.33 .01 75.00 .025 81.25 .034 Branchial region 58.33 .755 25.00 .06 46.67 1.101 64.00 .028 Hyoid region 0.00 0.00 15.00 .01 50.33 .402 65.00 .120 Oromandibular region 0.00 0.00 45.00 .03 59.04 1.007 56.29 .156 Opercular series 0.00 0.00 0 0.00 60.00 .107 46.18 .214 Postcrania Appendicular skeleton 75.00 .112 36.43 .07 39.62 .348 56.27 .448 Median fins 85.00 .450 44.44 1.17 36.90 3.857 61.08 2.547 Weberian apparatus 85.00 .180 78.00 .15 75.00 .257 - - Vertebral column 89.68 4.33 84.55 4.79 75.04 2.295 83.57 4.165 WMI vs. SI r2=0.226, p=0.34 r2=0.25, p=0.20 r2=0.121, p=0.36 r2=0.22, p=0.24

5.5.3. Bone fragmentation in locus 8: From the 575 bones analyzed, I found that fragmentation patterns vary between the bones and taxonomic groups (Appendix XVIII). Cichlid's skeletal elements are best preserved, compare to small and large cyprinids remains that are more fragmented (Figure 32). This difference is statistically significant (Friedman test: p=0.036; χ2= 13.560; df=3).

53 100 Cichlids (n=288) 90 80 Large Cyprinid (n=191) 70 Small Cyprinid (n=24) 60 50 40 Frequency (%) 30 20 10 0 05-29 30-50 51-70 71-89 90-100 % Fragmentation

Figure 32: Bones fragmentation patterns for four taxa in locus 8.

State of bone preservation in different taxa was examined by calculating the weighted mean index (WMI) of fragmentation (Appendix XIX) for each bone. The results (Appendix XVIII XIX) demonstrate the following: 1. Cichlid's vertebrae are well preserved, especially the atlas (WMI=90.6%) and thoracic vertebrae (WMI=86.06%). The pelvic spines are also well preserved (WMI=81%) compare to the highly fragmented pelvis (WMI=20; NISP=2). 2. For the large cyprinids, the thoracic vertebrae are well preserved (WMI=81%), while most cranial and post cranial bones are fragmented. I calculated the WMI of fragmentation for each anatomical region and compared it with the SI (Table 33). As in other loci, no correlation was found between the bone's fragmentation rate and SI value, except for 3 cases: the oromandibular region and opercular series of small cyprinid, and the median fins of large cyprinid. These regions do exhibit high SI due to fragmentation (Table 33). Table 33: WMI and SI values for locus 8 by anatomic regions and taxa. Cyprinidae L.8 Anatomic region Small Large Cichlidae Crania WMI SI WMI SI WMI SI Neurocranium 45.00 .35 0.00 0.00 85.00 .141 Branchial region 15.00 .46 51.36 .83 49.00 .188 Hyoid region 0.0 0.00 81.00 .33 60.00 .169 Oromandibular region 35.00 2.22 49.35 1.20 66.67 .406 Opercular series 52.50 4.44 55.00 .26 69.29 1.185 Postcrania Appendicular skeleton 0.0 0.00 44.23 .85 61.50 .846 Median fins 78.33 3.81 35.00 3.72 72.56 1.322 Weberian apparatus 75.00 .89 56.67 .62 . . Vertebral column 78.75 1.58 77.84 2.36 84.72 3.995

54 WMI vs. SI r2=0.10, p=0.48 r2=0.11, p=0.43 r2=0.30, p=0.16

5.6. Fish remains spatial distribution I calculated fish spatial distribution for loci 1 (brush hut) and 7 (ashes), according to excavation units of 0.5 m2. The bones distribution is illustrated for the total amount of fish remains recovered (including unidentified bones), according to excavation units of 1sq m. 5.6.1. Fish spatial distribution in locus 1: From 12,556 bones 42% were concentrated in squares E-79a, b, and c, and 21% were from squares E-80c,d (Figure 33). Smaller concentrations of bones were recovered from squares E-81a,b,c,d, and F-79 d. These remains present bones recovered during excavation seasons of 1989-1991. It is important to note that new material obtained from the excavation season of 2000, yield large amount of fish remains from square G-80. This material hasn't been studied yet, except for a small sample of ca. 300 bones, that is included in these calculations (Figure 33). DEFGI calculated the bones distribution

N pattern by calculating the Morisita 78 0 0.5 m 78

57 267 8 Index of dispersion, and received a clumped distribution (Id=5.6556,

7 79 9 16 5272 734 42 MU=0.9991, MC=1.0012, Ip=0.6058). Following Stewart (1989), bone scatter frequency was calculated as

80 80 398 2668 92 415 12,556/13.5m2 (54 square of 0.5m2) and received a mean value of 930.14 bones per m2. 8 89 1645 556 81 1

9 Figure 33: Spatial distribution of fish 128 155 5 remains in locus 1. 82 82

DEF G The fish distribution pattern was >50 00 >390-800 >50-119 examined according to the most >25 00 >12 0 <4 9 >15 00 abundant taxonomic groups: Acanthobrama sp., large carps, and cichlids (Figure 34). The distribution pattern of Acanthobrama sp. resembles with that observed for small cyprinids. Both taxonomic groups are concentrated in squares E-79, E-80, and E-81. High correlation (Spearman rank correlation: r= .907; p=0.001 ) was found between their spatial distribution. The distribution pattern of large cyprinids and cichlids

55 differ from the one observed for Acanthobrama sp. and small cyprinid (Figure 34). Most of them were recovered from square G 80, from excavation season of 2000. Likewise, no correlation was found between the spatial distribution of Acanthobrama sp. and small cyprinids vs. large cyprinids (Spearmen rank correlation: r= .429; p=0.1223 ) and cichlid (r=.535, p=0.053).

DEF G DE F G

0 0.5 m N 78 78 78 78 5 15 4 1 9 16 50 1

79 1483 193 2 2 62 33 7 79 79 79 0 84 86 8

80 80 80 80 65 743 4 8 3 48 8 195 23 58 11 14 6

7 344 41 4 42 51 81 81 81 81

10 84 76 3 3

9 5 30 22 3

82 82 82 29 37 0 82

DEF G DEF G

>14 00 >30 0 >40-80 >70 0 >19 0 <4 0

GHI Figure 34: Spatial distribution of Acanthobrama 01m 88 sp., (on left), of large carps (on right top left no.) N 894 and Cichlidae (on right the right no.) in locus 1. 89 1897 45 5.6.2. Fish spatial distribution in locus 7:

47 90 Spatial distribution of 5478 fish remains recovered during the excavation seasons of 1989-1991, was 1052 1055 43 91 examined. I observed that 35% of the bones were recovered from squares H-89a and H-89c, 19% 399 39 92 1000-1900 were recovered from square G-91b, and another >890

>390 7 93 19% were recovered from squares H-91c and d <50 (Figure 35). The bones distribution pattern was <10 94 calculated by Morisita Index of dispersion, and a HI G clumped distribution was observed (Id=2.2594, MU=0.9988, MC=1.0018, Ip=0.5699).

Figure 35: Spatial distribution of fish remains in locus 7. A similar pattern was found also for locus 3 (Id=2.0606, MU=0.9933, MC=1.0106, Ip=0.5584). Following Stewart (1989), I calculated the bone scatter frequency from locus 7, as 842.8 bones per sqm (5478/6.5). Moreover, a significant differences was found between the spatial distribution of each taxonomic group (Friedman test, χ2=19.89, df=3, p=0.0002). 5.7. Vertebrae dimensions Although vertebrae are the most common fish remain in archaeological sites, most of them can be identified to the family level only (Reitz & Wing, 1999; Wheeler & Jones, 1989). I compared vertebrae width dimensions (mean, minimum, maximum, standard deviation) in four loci (loci 1, 3, 7, and 8) by taxa (Appendix XX; Figure 36). The diameter of Acanthobrama sp. atlas, from locus 1 display larger dimensions than those recovered from locus 7 (Figure 36). However, atlas diameter of small cyprinids from loci 1 and 8 resemble with Acanthobrama sp. atlas dimensions from locus 1. Small cyprinid's atlas from locus 7 also display smaller sizes in comparison with their dimension in loci 1 and 8. The cichlids atlas diameter exhibit similar dimensions among all loci (Figure 36). Atlas of large cyprinids exhibit similar dimension in all loci. Their atlas is the larger among the taxonomic groups studied.

Figure 36: Atlas mean width by loci and taxa.

11 10 L.1 (n=476) L.3 (n=43) 9 L.7 (n=69) L.8 (n=19) 8

7 6 5 4 3

Atlas mean width (mm) 2 1 I performed 0 ANOVA between atlas dimensions (width,

Cichlidae length and height) and Small Carps Large Carps taxonomic groups, from Taxonomic group Acanthobrama sp. locus 1, and found a

57 similarity between Acanthobrama sp. and small cyprinids (Table 34). This result indicate that vertebrae identified as "small cyprinid" should be regarded as Acanthobrama sp. Interestingly, cichlids atlas width also exhibit similar dimension with Acanthobrama sp. and small cyprinids (Table 34). I received a significant difference between large cyprinids to Acanthobrama sp. and small cyprinids (Table 34). In locus 7 I also received a significant difference between atlas width and taxonomic groups (ANOVA: p<0.0001, F=52.246, DF=3, 50). Table 34: Scheffe post hoc tests between taxonomic groups and atlas dimensions (width, length and height) in locus 1. Taxonomic group Atlas width Atlas length Atlas height Acanthobrama sp. vs. Cichlidae 0.347 0.0001** 0.0001** Acanthobrama sp. vs. Large cyprinid 0.0001** 0.0001** 0.0001** Acanthobrama sp. vs. Small cyprinid 0.996 0.9439 0.989 Cichlidae vs. Large cyprinid 0.0001** 0.0001** 0.0001** Cichlidae vs. Small cyprinid 0.708 0.0167* 0.0001** Large cyprinid vs. Small cyprinid 0.0001** 0.0001** 0.0001** As atlas dimensions varied significantly between Cyprinidae, the possibility to use vertebrae diameter (width mm) for taxonomic separation was examined (Appendix XX; Figure 37). If vertebrae diameter in all taxa are normally distributed, than we expect to observe overlapping in the range of vertebrae diameter, between species. Figure 37 present frequency distribution of width diameter from 377 caudal vertebrae of Cyprinidae.

250 Figure 37: Frequency distribution 200 (NISP) of Cyprinidae caudal vertebrae width in Locus 1. 150

NISP 100

1-1.5 1.6-2 2.6-3 3.6-4 4.6-5 5.6-6 6.6-7 7.6-8 8.6-9 As observed, despite the 9.6-10 2.1-2.5 3.1-3.5 4.1-4.5 5.1-5.5 6.1-6.5 7.1-7.5 8.1-8.5 9.1-9.5 diameter wide range (1.3-8.1mm) Width (mm) the distribution is not normal and most of the vertebrae appear in the size category up to 3 mm. This distribution demonstrate that the small caudal vertebra belong to Acanthobrama sp., while vertebrae larger than 3mm belong to the larger taxa (Barbus/ Capoeta sp.). 5.8. Body mass estimation

58 The regression equations calculated for atlas width (see chapter 4.6.1) were used to estimate Cyprinidae body size for loci 1 and 7 (Table 35). In locus 1, Acanthobrama sp. display wide range of body size (Figure 38). Fish smaller than 84 mm in length and 10 gr in weight are absent (Figure 38). A similarity between Acanthobrama sp. reconstructed body size and small cyprinids (Table 35) was found. This implies, again, that the two groups probably represent the same taxa. The maximum SL of small cyprinids and Acanthobrama sp. is 190 mm. This is in accordance with present day fish (Goren et al., 1973). Table 35: Body mass (gr) and standard length (mm) estimated by taxa from loci 1 and 7. Body Mass (gr) Standard Length (mm) Location Taxonomic group NISP Range Mean S.D. Range Mean S.D. Locus 1 Acanthobrama sp. 371 10.6-66.6 39.02 7.3 84.5-180.8 144.4 11.5 Cyprinidae small 10 19.0-111.0 49.2 26.4 103-189.9 139.5 24.4 Locus 1 Barbus sp./ Capoeta sp. 19 139-906.0 477.0 207.0 144-461.4 285.0 85.9 Locus 7 Barbus sp./ Capoeta sp. 16 357-1014.2 525.0 164.5 235-506.2 304.0 68.0

Body sizes of large cyprinids from loci 1 and 7 is greater than those of small cyprinids and Acanthobrama sp. (Table 35). Large cyprinid from locus 1 display smaller body size compared to those recovered from locus 7 (L.1 maximum weight=906 gr; L.7 maximum weigh =1014 gr; Figure 40). I found this difference to be statistically significant for body mass (Mann-Whitney z=-2.245; p=0.025), and for standard length (Mann-Whitney z=-4.323; p<0.0001).

50 50

40 40

30 30

20 20 Frequency (%) Frequency (%) 10 10

0 0 12-17 18-22 23-27 28-32 33-37 38-42 43-47 48-52 65-85 86-105 106-125 126-145 146-165 166-185 186-205 Standard length (mm) Body mass (gr) Figure 38: Estimated standard length (mm) and body mass (gr) of Acanthobrama sp. in locus 1.

59 50 50 Locus-1 40 40 Locus-7 30 30

20 20 Frequency (%) 10 10

0 0 135-235 236-335 336-435 436-535 536-635 636-735 736-835 836-935 140-190 191-240 241-290 291-340 341-390 391-440 441-490 491-540 936-1035 Body mass (gr) Standard length (mm) Figure 39: Estimated body sizes of Barbus sp./ Capoeta sp. from loci 1 and 7.

5.9. Dietary value Based on mean body mass and MNI, I predicted the fish dietary contribution to Ohalo-II inhabitants (Table 36). If all identified Cyprinidae result from human activity, they produced a minimum of 84kg meat. Most of the caloric intake is obtained from large cyprinids (Barbus sp. & Capoeta sp.), producing at least 59 kg of food.

Table 36: Estimation of fish dietary value from predicted mean body mass (BM) and MNI. Acanthobrama sp. Small cyprinid Large cyprinid Locus MNI BM (kg) MNI BM (kg) MNI BM (kg) L.1 387 15.40 214 10.70 44 22.0 L.2 - - 1 0.05 1 0.500 L.3 - - 1 0.05 36 18.0 L.7 58 2.3 36 1.80 27 13.5 L.8 1 0.04 1 0.05 10 5.0 Total BM 17.74 12.65 59.0

To examine variation in fish exploitation (large taxa vs. small taxa) I calculated the fish index (see chapter 4.5.6) for each loci (Table 37). The fish index is similar for loci 1, 3, and 8 suggesting exploitation of large fish taxa. In locus 7 there is a small decline in large taxa and increase in small taxa exploitation. In locus 1 there is a clear decline in the importance of large taxa and sharp increase of small taxa. These variations will be discussed later in regard to changes in human exploitation strategies and will be compared with fish natural accumulation. Table 37: Fish exploitation index by loci. Locus NISP large taxa NISP small taxa Fish index

60 L.1 1287 10389 0.110 L.2 52 1 0.981 L.3 604 12 0.981 L.7 2471 1639 0.601 L.8 510 27 0.950

5.10. Summary The 19,799 fish remains analyzed from seven structures in Ohalo-II demonstrate the following patterns: 1. Two families of primary and secondary freshwater fish were identified at the site: Cyprinidae and Cichlidae. Their MNI is of 942 fish, from which 817 are cyprinid and 104 are cichlid. 2. The ratio between Cyprinidae and Cichlidae is relatively similar for loci 3 and 7, while for loci 1, 2, and 9 Cyprinidae are highly abundant. In locus 8 Cichlidae are slightly more abundant than in other loci. 3. Eight species of fish were identified, four species for each family. 4. Two of the identified taxa are endemic to the Sea of Galilee: Tristamella sp. and Acanthobrama terraesanctae. 5. Species richness and diversity vary between loci. Rarefaction analysis demonstrated that for loci 2, 8, and 9, species richness was biased due to sample size. In the un-biased samples, species richness and diversity exhibited a high value for locus 7 and low value for locus 1. 6. Locus 1 differ from other loci in its taxonomic composition dominated by Acanthobrama sp. (Kinneret Bleak) with normally distributed body size. This species appear in relatively low frequency for locus 7 and is absent from all other loci. 7. Large Cyprinidae (Barbus sp. and Capoeta sp.) are abundant for loci 3, 7, and 8. 8. Rarefaction analysis demonstrated that for loci 2 and 9, skeletal element richness is biased due to sample size. 9. In locus 1 small cyprinids exhibit widest diversity of skeletal elements, while for loci 7 and 8 large cyprinids exhibits the widest diversity. 10. In locus 1, the ratio of cranial to post cranial bones of Acanthobrama sp. resemble that of a complete fish, while in large carps and cichlid postcranial region is over-represented and the cranial region is under-represented. Over-representation of postcranial region characterize all other fish remains analyzed from various structures. 11. In loci 7 and 8 large cyprinids cranial region SI is higher than in locus 1: 63% vs. 54%.

61 12. Scales were absent from all samples analyzed, while otoliths were recovered from loci 1, 7, and 8. In locus 1 a single otolith of large cyprinid was identified. In loci 7 and 8 all otoliths belonged to cichlid's. 13. The presence of cichlid otoliths in locus 8, may imply that they were deposited with their skull intact. 14. Small number of bones (5%) exhibit burning signs. Most of them were recovered from ashes deposits in locus 7, and may have resulted from post-depositional burning. 15. Fragmentation patterns significantly differ between excavated area: bones from locus 7 (ashes) exhibit the lowest degree of fragmentation, although they were deposited in the fireplace area. Bones from loci 3, 8, and 9 are highly fragmented, although in locus 8 Cichlidae remains are relatively best preserved. In locus 1 Acanthobrama sp. and large cyprinids exhibit higher rate of fragmentation. The fragmentation pattern will be later discussed in regard to human activity or post-depositional abrasion. 16. Comparison between WMI of fragmentation and SI exhibit no correlation, implying that over-represented anatomic regions are not biased due to high fragmentation rate. 17. In Cyprinidae vertebrae width can be used to differentiate between taxonomic groups. It also demonstrate that vertebrae identified as "small cyprinid" resemble with the genus group Acanthobrama sp., and not with larger taxa. 18. Atlas diameter of Cichlidae from loci 1 and 8 exhibit relatively small dimension, which resemble with Acanthobrama sp.. 19. Large cyprinid recovered from locus 7 display larger sizes than the one from locus 1. 20. Bones spatial distribution is clumped in all excavated structures. BSF per sqm is of 930 for locus 1, and 840 for locus 7. 21. If all cyprinids recovered at Ohalo-II were utilized by Ohalo-II inhabitants, they could have provided a minimum of 84 kg of meat. Most of the caloric intake was obtained from large Cyprinidae (59 kg). The implication of these results in regard to human activity and fish exploitation strategies will be discussed later in comparison with fish remains from the natural assemblage and the ethnographic data.

62 CHAPTER 6: FISH NATURAL ACCUMULATION "Archaeologists working throughout the world are beginning to recognize the potential for natural fish remains to become incorporated into prehistoric cultural deposits" (Butler, 1990: 195) In this chapter I characterize fish natural accumulation, along the southern shore of the Sea of Galilee. For this purpose, the following criteria were used: amount of bones recovered (NISP), bone dispersal patterns, taxonomic diversity, skeletal completeness, fragmentation pattern, bone's color, and vertebra dimensions. 6.1. Bone spatial distribution patterns A total of 5795 fish remains were recovered in 24 random squares (0.5m2 each). The number of bones in the various squares varies from 8 to 2894 bone (Figure 40). The mean bone scatter frequency is 423 per 0.5 sqm. The bone spatial distribution pattern is clumped, as calculated from the Morisita index of dispersion (Id=6.5207, MU=0.9980, MC=1.0062, Ip=0.6200).

>2800 >200 >50 Figure 40: Spatial distribution of >350 >150 <50 naturally deposited fish remains.

I counted the number of bones within three sedimentation layers (taphofacies; See Figure 6): upper sand (0-5 cm depth), median brown sand with clay (5-10 cm depth), and bottom dark clay (10-15 cm depth). Sea of Galilee Most of the bones (70%) were recovered from the deep anaerobic clay deposit. The number of bones decline to 23% at the median layer and to 7% as we come closer to the sandy taphofacies. 6.2. Taxonomic identification The faunal remains recovered along the southern beach included mainly mollusks, 5968 fish remains, and a single bone of a rodent. Twenty-nine percent (1709) of the bones were identified as belonging to fish from three families of primary and secondary freshwater fish: Cyprinidae, Cichlidae, and Clariidae (Table 38). The abundance of naturally deposited bones from the three families differ significantly between the tested areas (random squares, beach, and Ohalo-II recent sediments) (Table 38; r x c randomization test χ2= 1561.314, p<0.0001, df=4).

69 Table 38: Frequency (NISP) and percentage for naturally deposited fish remains by family. Cichlidae Cyprinidae Clariidae Total Locality NISP % NISP % NISP % NISP Natural-random 157 10.0 1408 89.9 1 0.06 1566 Natural-beach 22 52.4 15 35.7 5 11.9 42 Ohalo II-natural 8 7.9 0.0 0.0 93 92.1 101 Total 187 10.9% 1423 83.3% 99 5.8% 1709

Cyprinidae are highly abundant (89.9%) in the random sample, and relatively less abundant (36%) on the recent beach (Table 38). An opposite trend is observed for Cichlidae remains which are less abundant in the random sample (10%) and highly abundant in the surface sample (52.5% and 8%; Table 38). Clariidae dominant (92.1%) the samples collected from Ohalo-II modern surface, and appear in low frequency in the recent shore sample. Surprisingly, a single bone of Clariidae was recovered from the random sample (Table 38). In the random squares, Cyprinidae and Cichlidae bones increase in number with the sediment's depth (Table 39), reaching a peak in the clay deposition. Clariidae remains appear exclusively in the sandy surface sediments (Table 39). Table 39: Frequency (NISP) for naturally deposited fish by family, depositional location, and depth. Random Squares Recent Surface Ohalo-II recent Total Family Sand Brown Clay Sand Sand Cichlidae 25 49 83 22 8 187 Cyprinidae 104 443 861 15 0 1423 Clariidae 1 0 0 5 93 99 Unidentified 209 853 3167 30 0 4259 Total 339 1345 4111 72 101 5968

Due to few species specific bones, identification of fish remains to the genus level was possible only on a small sample of 362 bones (Table 40). Acanthobrama sp. is the most abundant genus identified (66%), followed by Clarias gariepinus (27%). Interestingly, large carps (Barbus sp. and Capoeta sp.) are relatively rare (ca. 7%). Table 40: Frequency (NISP) for naturally deposited fish by genus, depositional location, and depth. Taxonomic group Ohalo-II Random squares Recent Total recent Surface Sand Sand Brown Clay Sand NISP % Clarias gariepinus 93 1 0 0 5 99 27.0 Acanthobrama sp. 0 13 64 159 2 238 66.0 Barbus sp. 0 0 5 3 1 9 2.5 Capoeta sp. 0 2 10 2 0 14 3.9 Barbus sp./Capoeta sp. 0 0 1 0 1 2 O.5 Total 93 16 80 164 9 362 100%

From the bones identified to the family and genus level, species richness and Brillouin diversity indices (HB) were calculated (Table 41). Species richness is low (2) for the sample

70 collected from Ohalo-II recent surface and high for the sample collected from the recent shore surface (5). Other samples exhibit a species richness of 4. The HB diversity indices differ significantly between all sampled areas (Table 41). It is significantly low for fish remains recovered from the recent surface of Ohalo-II, and significantly high for fish remains recovered from the recent beach surface. The random squares exhibit a general pattern: reduction in HB value with depth. Despite the similarity in their species richness, the HB diversity values are significantly different between taphofacies (Magurran, 1988). The top sandy layers display higher diversity (HB=1.26), while the clay sediments exhibits lower diversity (HB=1.09). This can be attributed to the decline in the relative abundance of Capoeta sp. and Barbus sp. remains. Table 41: NISP, species richness and Brillouin index calculated for naturally deposited fish. Locality NISP Species richness Brillouin Index (HB) Ohalo II-recent 101 2 0.38 Random squares-Sand 130 4 1.26 Random squares-Median brown 492 4 1.19 Random squares bottom clay 944 4 1.09 Recent surface 42 5 1.594

Since taxonomic identification to the genus level was possible only on a small sample of bones, in all further calculations I grouped the identified bones into five taxonomic groups, as follow: Clarias gariepinus, Acanthobrama sp., large cyprinids (Barbus sp. and Capoeta sp.), small cyprinids and cichlids (Table 42). From 1709 bones I calculated a total MNI of 105 fish, of which 56 (53%) are Acanthobrama sp. (Table 42). Table 42: NISP and MNI calculated for naturally deposited fish by taxa and sampling area. Ohalo-II Random squares Surface Total Recent Sand Brown Clay Sand NISP MNI NISP MNI NISP MNI NISP MNI NISP MNI NISP Acanthobrama sp. 0 0 13 3 64 14 159 28 2 1 238 Cichlid 8 1 25 2 49 2 83 3 22 1 187 Clarias 93 2 1 1 0 0 0 0 5 1 99 Large cyprinid 0 0 2 1 16 2 5 2 2 1 25 Small cyprinid 0 0 89 3 363 15 697 21 11 2 1160 Total 101 3 130 10 492 33 944 53 42 6 1709

Not surprisingly, there is a high correlation between NISP and MNI value calculated for each location (p=0.989). The highest MNI value of 53 fish (4 taxa) was calculated for the clay sediments, while the lowest MNI of 3 (2 taxa), was calculated for the recent sediments of Ohalo-II (Table 42). 6.3. Skeletal representation Based on NISP of 3204 bones, I calculated the relative abundance of skeletal elements recovered from natural accumulation (Appendix XXI). The most abundant skeletal elements are vertebrae (40%) and scales (24%). While vertebrae are over-represented in all samples,

71 the relative abundance of scales varies. Scales are present in the random squares surface sediments (30%), brown sediments (19%), as well as at the recent beach surface (57%). Yet, scales are absent from Ohalo-II recent surface and from the bottom clay sediments of the random squares (Appendix XXI). The absence of scales in Ohalo-II surface may be attributed to recovery bias, since this sample was hand picked. However, the clay sediments were wet sieved through a 1 mm mesh sieve, and therefore the absence of scales can not result from recovery bias, but probably from differential preservation. O'conell and Tunnicliffe (2001) described a similar pattern of scales lost in accordance to depositional age. The total number of skeletal elements (bone richness) identified in each locality varies accordingly with NISP values (Table 43). The highest bone richness was obtained from the bottom clay sediments of the random squares, while lowest number of skeletal elements (32 and 33) is obtained from the recent beach surface (Figure 41). Table 43: NISP, standardized NISP*, and richness values calculated for naturally deposited fish in sampling areas. Natural Accumulation Locality NISP Standardize NISP Skeletal element richness Random square: Surface sand 339 442.2 41 Random square: Median brown 1345 832.7 44 Random square: Bottom clay 4111 3136.1 55 Ohalo-II Recent surface 101 - 33 Recent surface 72 - 32 *Standardized NISP= NISP/sediment volume Although collecting methods could have influenced the low bone richness obtained from the recent shore, in the random squares bone richness increased with sediment depth (r=0.997; Table 43). This pattern, which does not correlated with sediments volume (r= 0.018), demonstrates that in clay deposits, except for scale loss, the general state of preservation is in favor of the bones (Figure 41). Figure 41: Comparison between skeletal elements richness by taxa and depositional depth in 50 the random squares. 45 Sandy surface 40 35 Median Brown 30 Bottom Clay 25 20 15 Another pattern 10 Skeletal element richness 5 observed concerns 0 skeletal elements richness according to Cichlids Clarias sp. Unidentified Small cyprinid

Acanthobrama sp. 72 Barbus/Capoets sp. taxonomic groups in the random squares. In all identified taphofacies, richness value is high for small cyprinids and cichlids and it is very low for large cyprinid and Clarias sp.. This ratio between bone richness and taxonomic group remains constant in all taphofacies (χ2=1.102, df=4, p=0.8884). Due to low bone richness observed for most taxa (Appendix XXI), further calculations were grouped into anatomic regions (Tables 44-46). Table 44: NISP of anatomic regions in random squares by taxa and taphofacies. Sandy surface Cyprinidae Anatomic region Acanthobrama Small Large Cichlid Clarias Unident. Total Crania Neurocranium 0 6 0 2 0 1 9 Branchial region 2 6 2 1 0 0 11 Hyoid region 1 8 0 1 0 1 11 Oromandibular region 2 5 0 3 0 1 11 Opercular series 0 4 0 3 0 1 8 Cranial general 0 0 0 0 0 8 8 Postcrania Appendicular skeleton 0 8 0 1 0 5 14 Median fins 0 4 0 7 0 2 13 Vertebral column 8 45 0 7 1 17 78 Rib 0 3 0 0 0 0 3 Scales 0 0 0 0 0 103 102 unidentified 0 0 0 0 0 70 70 Total 13 89 2 25 1 209 339 Median brown Crania Neurocranium 1 20 0 6 0 11 38 Branchial region 21 13 15 5 0 0 54 Hyoid region 1 33 0 1 0 1 36 Oromandibular region 1 45 0 1 0 0 47 Opercular series 0 36 0 1 0 9 46 Cranial general 0 0 0 0 0 19 19 Postcrania Appendicular skeleton 1 54 0 3 0 5 63 Median fins 0 18 0 13 0 16 47 Vertebral column 39 137 1 15 0 183 375 Rib 0 7 0 4 0 0 11 Scales 0 0 0 0 0 251 251 unidentified 0 0 0 0 0 358 358 Total 64 363 16 49 0 853 1345 Bottom Clay Crania Neurocranium 0 60 0 6 0 5 71 Branchial region 47 14 4 5 0 2 72 Hyoid region 12 66 0 0 0 1 79 Oromandibular region 29 54 0 4 0 3 90 Opercular series 2 74 0 3 0 16 95 Cranial general 0 0 0 0 0 11 11 Postcrania Appendicular skeleton 0 108 1 10 0 17 136 Median fins 0 19 0 20 0 14 53 Vertebral column 69 297 0 30 0 363 759 Rib 0 5 0 5 0 0 10 unidentified 0 0 0 0 0 2332 2332 Total 159 697 5 83 0 3167 4111

Table 45: NISP of anatomic regions in recent beach surface by taxa. Recent beach surface Cyprinidae

73 Anatomic region Acanthobrama Small Large Cichlid Clarias Unident. Total Crania Neurocranium 0 0 0 0 3 1 4 Branchial region 0 0 1 2 0 0 3 Hyoid region 0 1 0 1 0 0 2 Oromandibular region 0 1 0 1 1 0 3 Opercular series 0 4 0 3 1 0 8 Postcrania Appendicular skeleton 0 0 0 3 0 0 3 Median fins 1 0 1 3 0 1 6 Vertebral column 1 5 0 8 0 7 21 Rib 0 0 0 1 0 0 1 Scales 0 0 0 0 0 17 17 unidentified 0 0 0 0 0 4 4 Total 2 11 2 22 5 30 72

Table 46: NISP of anatomic regions in recent surface of Ohalo-II by taxa. Ohalo-II recent surface Anatomic region Cichlid Clarias sp. Total Crania Neurocranium 0 19 19 Branchial region 0 4 4 Hyoid region 0 7 7 Oromandibular region 0 5 5 Opercular series 0 2 2 Cranial 0 12 12 Postcrania Appendicular skeleton 4 6 10 Median fins 3 3 6 Caudal skeleton 0 16 16 Vertebral column 1 11 12 Post crania general 0 8 8 Total 8 93 101

From the bones frequency (NISP) by anatomical regions, I observed the following patterns (Tables 44-46): 1. The frequency of anatomic regions (total NISP) significantly differ between taxa and depositional area(χ2=194.639, df=32; p<0.0001). 2. The frequency (NISP) of Acanthobrama sp., small cyprinids and cichlids anatomical regions increase according to the stratigraphic depth (Table 44). They are highly abundant in the bottom clay sediments. 3. Large cyprinids are rare regardless of depositional depth or location (Tables 44-46). 4. The absence of Acanthobrama sp. neurocranial bones result probably from identification bias. The abundance of small cyprinid's neurocranium bones increase from 6 bones in the sandy sediments to 60 in the clay (Table 44). 5. Acanthobrama sp. opercular series is represented only in the clay sediments. It appears in all taphofacies for small cyprinids increasing in number with depositional depth. 6. Cichlids anatomical regions are represented in all taphofacies, except for the hyoid region that is absent from the clay sediments (Table 44).

74 7. All anatomical regions of Clariidae remains are present only in the sample collected from Ohalo-II recent surface (Table 46). In the other samples, Clariidae cranial bones and few post cranial bones are present only in the surface sediments. 8. Cichlid's branchial region is absent in all samples (Tables 44-46). When present, it is usually related to the cyprinids and found in increasing numbers in the clay deposits (Table 44). Its presentation, however, is biased toward pharyngeal teeth and bones, while in fact most of the branchial region bones are absent (see Appendix XXI). I calculated survival index (SI) for each taxon, excluding those that are represented by small sample size, such as large cyprinid. For Acanthobrama sp. (Acantho), I used exclusively the data from the random squares clay deposits, and for Clarias sp., I used only data from Ohalo-II modern deposits (Table 47). Table 47: Survival index (SI) calculated for naturally deposited fish by taxa and anatomic regions. Small cyprinid Cichlid Acantho Clarias Anatomic region Sand Brown Clay Sand Brown Clay Clay Ohalo Crania Neurocranium .305* .250* .390* .325 .497 .294* 0.00* 1.39 Branchial region .269* .143* .080* .217 .553 .326* 1.26 .341* Hyoid region 1.30 1.31 1.37* .488 .249 0.00* 1.01 .941 Oromandibular region .649 1.43* .895 1.17 .199 .470 1.95* .522 Opercular series 1.30 2.86* 3.07* 2.92 .497 .881 .336 .633 Postcrania Appendicular skeleton 1.30 2.15* 2.24* .488 .746 1.47 0.00* 1.40 Median fins 2.60* 2.27* 1.14 2.60* 3.22* 2.78* 0.00* 2.93 Weberian apparatus 0.00* 0.00* 0.00* - - - 0.00* - Vertebral column 2.60* 1.94* 2.19* 1.82 1.99* 2.35* 2.38* .806 *Significantly different from the expected in a complete fish, p<0.05 (Chi-square test). From the survival index, I observed the following patterns (Table 47): 1. In the cranial region, the neurocranium is under-represented in all samples excluding the Clarias sp. from Ohalo-II surface. This pattern is significantly different from the expected in a complete fish, and can be due to identification bias. In the case of cichlid remains from the sand and brown deposits, the sample was too small to carry out a chi- square test. 2. The branchial region is present, but under-represented, in all samples, except for Acanthobrama sp. remains from the clay deposits. This different differs significantly from the expected values in a complete fish. Sample size of cichlids remains from the sand and brown deposits was, again, too small to carry out a chi-square test. 3. The hyoid region of Acanthobrama sp. is similar with a complete skeleton, in the clay deposit. For small cyprinids a similar pattern is observed from the sand and brown deposits, and in the clay deposits it is significantly over-represented. For the cichlid

75 remains in the random squares the hyoid region is absent from the clay deposits, and represented by a single bone in other taphofacies. For Clarias sp. remains from Ohalo-II surface, the hyoid region is similar with that expected. 4. The oromandibular region is significantly over-represented for small cyprinid remains from the brown deposits, and for Acanthobrama sp. remains in the clay deposits (Table 47). For cichlid remains from the sand deposits, and small cyprinids from the clay deposits it is similar with a complete fish. For small cyprinids from the sand deposits it is slightly under-represented. 5. The opercular region is over-represented for small cyprinid remains from the brown and clay sediments, and significantly differs from the expected in a complete fish. In cichlids remains, this region is over-represented at the sand deposits and under-represented in the brown and clay sediments. The opercular series is also under-represented for Acanthobrama sp. and Clarias sp. remains (Table 47). However, these patterns do not differ from the expected in a complete fish, probably due to small sample size. 6. The Weberian apparatus is absent for cyprinid remains from all sampled areas and taphofacies. This pattern significantly differs from the expected in a complete skeleton (p<0.001). 7. The vertebral column is over-represented (SI>1) in all samples, except for Clarias sp., where it is slightly under-represented (SI<1). This pattern is significantly different from the expected in complete Cyprinidae and Cichlidae. 8. The absence of Acanthobrama sp. appendicular skeleton result probably from identification bias (see appendix II). This region is over-represented for small cyprinids from all taphofacies. In the brown and clay sediments this pattern is significantly different from the expected in a complete fish. The appendicular skeleton is under- represented for cichlid remains from the sand and brown deposit, and is over-represented in the clay deposits. Clarias sp. appendicular skeleton from Ohalo-II modern surface, is also over-represented. 9. The median fin is over-represented in all samples. It significantly differs from the expected in a complete fish for cichlid in all samples, and for small cyprinids from the sand and brown sediments. I examined the ratio of cranial and postcranial elements in the natural accumulation compared to the expected ratio in a complete fish (Tables 48-49). The SI demonstrate that the postcranial region is over-represented regardless of taxa or depositional area (Tables 48-49).

76 In the random squares, the cranial region is over-represented only for large cyprinids (Table 48). The cranial region is under-represented for all other taxa, regardless of depositional depth. The post-cranial region is under-represented for large cyprinids from the median brown layer (Table 48). The post-crania region is over-represented for all other taxa, regardless of depositional depth. The crania to postcrania ratio, in the random squares, significantly differs from the expected ratio for small cyprinids in the sand layer (χ2=16.4, p<0.0001). In the brown layer the ratio was found to be significantly different for Acanthobrama sp. (χ2=10.1, p<0.0001), small cyprinids (χ2=37.1, p<0.0001), and cichlids (χ2=5.2, p<0.05). In the bottom clay sediments the crania/ postcrania ratio was found to be significantly different for cichlid remains (χ2=15.8, p<0.0001) and small cyprinids (χ2=83.9, df=1, p<0.0001). The crania to postcrania ratio is similar with a complete fish for Acanthobrama sp. remains in the sand and clay layers, for cichlids remains from the sand layer, and for large cyprinid remains from the median brown layer (Table 48). Table 48: Observed and expected percentage and SI of cranial and postcranial bones in naturally deposited fish (random squares) for four taxa. Random Genus group Cranial Postcranial Total Squares Obs. Exp. SI Obs. Exp. SI NISP χ2 Taphofacies % % % % Sandy surface Acanthobrama sp. 38.5 66.0 .583 61.5 34.0 1.81 13ns 2.48 Cichlids 40.0 65.0 .635 60.0 35.0 .952 25ns 0.57 Small cyprinids 32.6 63.0 .517 67.4 37.0 1.82 89 * 16.4

Median Brown Acanthobrama sp. 37.5 66.0 .568 62.5 34.0 1.84 64 * 10.1 Cichlids 28.6 65.0 .454 71.4 35.0 1.13 49 * 5.21 Large cyprinids 93.7 63.0 1.50 6.2 37.0 .169 16ns 4.57 Small cyprinids 40.5 63.0 .643 59.5 37.0 1.61 363 * 37.1

Bottom clay Acanthobrama sp. 56.6 66.0 .858 43.4 34.0 1.28 159 ns 2.98 Cichlids 21.7 65.0 .344 78.3 35.0 1.24 83 * 15.8 Large cyprinids 80.0 63.0 - 20.0 37.0 - 5 - Small cyprinids 38.4 63.0 .610 61.5 37.0 1.66 697 * 83.9 Fisher exact test significant different p<0.05, df=1 In the recent beach, the crania region is under-represented for cichlids and small cyprinids, and the postcrania region is over-represented (Table 49). A similar pattern is observed for C. gariepinus from Ohalo-II modern sediments. Table 49 : Observed and expected percentage and SI of cranial and postcranial bones in naturally deposited fish (recent shore) for five taxa. Locality Taxa Crania Post-crania Total Obs. % Exp. % SI Obs. % Exp. % SI NISP Ohalo-II recent Cichlid 0.00 65.0 100.0 35.0 8 C. gariepinus 52.70 58.0 .908 47.30 42.0 1.13 93 ns

Recent Surface Acanthobrama sp. 0.00 66.0 - 100.00 34.0 - 2

77 Cichlid 31.80 65.0 .505 68.20 35.0 1.08 22 ns C. gariepinus 100.0 58.0 - 0.00 42.0 - 5 Large cyprinid 50.00 63.0 - 50.00 37.0 - 2 Small cyprinid 54.55 63.0 .866 45.45 37.0 1.23 11 ns * Fisher exact test significant different p<0.0001, df=1 6.4. Bone modification I examined bone modification pattern on 3208 bones, recovered from natural accumulation (Table 50). Most of the bones (85%) are well preserved (Table 50). Bone color was recorded demonstrating the abundance of brown (68%), and light brown (24%) colors (Table 51). About 7% of all bones displayed a dark brown color (Table 51). In Weizmann Institute FTIR analysis were performed on a sample of the “dark-brown” bones that exhibit the presence of collagen. A sample of bones from the bottom clay layer dated the bones to 430-620 AD (calibrated age). Table 50: Mean state of fragmentation in naturally deposited fish by location and taphofacies. Fragmentation Total Random square Ohalo-II Recent Recent category (%) NISP % Sand Brown Clay Sand Surface 5-20% 20 .62 2.16 .41 .50 0.00 1.43 21-40% 95 2.96 3.96 5.85 1.51 0.00 0.00 41-60% 176 5.49 3.60 8.42 4.20 7.92 1.43 61-80% 208 6.48 8.99 6.67 5.99 5.94 7.14 80-100% 2709 84.45 81.29 78.64 87.79 86.14 90.00 Total 3208 100% 100% 100% 100% 100% 100%

Table 51: Frequency (NISP) of bone color recorded in naturally deposited fish. Bone Total Ohalo-II Recent Random square Recent Color NISP % Sand Sand Brown Clay Surface Brown 4070 68.20 2 132 643 3279 14 Light Brown 1437 24.08 99 171 412 704 51 Dark brown 411 6.89 0 2 282 126 1 Orange-Brown 39 0.65 0 30 3 2 4 Black 2 0.03 0 0 1 0 1 Gray 1 0.02 0 0 1 0 0 White 8 0.13 0 4 3 0 1 Total 5968 100% 101 339 1345 4111 72

6.5. Vertebrae dimension Measurements of vertebra: width, height and length, were taken on a small sample of bones (78, 86, 79 respectively). Most vertebrae manifested small dimensions (Figure 42). The atlas is the only vertebra

4.5 exhibiting relatively wide Acanthobrama sp. 4 variability (±0.46). 3.5 Cichlids Figure 42: Vertebrae mean 3 Small cyprinid width diameter (± SD) for 2.5 fish natural accumulation by 2 taxa. 1.5

1 Vertebrae width (mm) 0.5 78

0 Atlas Axis 3rd vert.Precaudal

6.6 Body size estimation I used 19 atlas of Acanthobrama sp. for body size reconstruction (Table 52). Body mass ranged between 5-28 gr and the maximum standard length is 128 mm. The body size distribution exhibit an abundance of juvenile fish while large adult fish are absent. Table 52: Acanthobrama sp. estimated body mass (gr) and standard length (mm). Body size Count Mean Std. Range Body mass (gr) 19 14.06 6.00 5-28.00 Standard length 19 89.63 16.07 65-127.00

6.7. Summary From the 5968 fish remains recovered at the natural accumulation along the southern shore of the Sea of Galilee, I observed the following: 1. The faunal remains recovered included mainly mollusks, 5968 fish remains, and a single bone of a rodent. 2. Three families of primary and secondary freshwater fish were identified: Cyprinidae, Cichlidae, and Clariidae. 3. The number of fish remains (NISP) increases with depositional depth, reaching their peak in the deepest clay deposit, dated to 430-62- AD. 4. Clariidae remains were recovered exclusively from the sand deposits. Cichlidae and Cyprinidae remains appear in all sediments and their relative abundance increase in accordance with depositional depth. 5. The most abundant taxonomic group is of small cyprinids (probably Acanthobrama sp.) while large cyprinid remains (Barbus sp./ Capoeta sp.) are scarce in all samples. 6. The fish remains recovered from the random squares display a clumped distribution. Bone scatter frequency is of 423 bones per 0.5 sqm. 7. Species diversity (HB) displays a significantly high value (1.594) for fish remains recovered from the sandy surface of the recent beach and the random squares. Diversity decrease with depth (1.09 at the deepest layer). 8. In all samples the postcranial region is over-represented while the cranial region is under- represented, demonstrating a preservation bias toward the cranial region.

79 9. Cyprinids branchial region is present, and consists mostly of pharyngeal bone and teeth. In cichlids, this region is absent. 10. The vertebral column is over-represented for all taxa. 11. Scales were abundant in the random squares sand and brown deposit. However they were absent from the random squares clay deposits, and from Ohalo-II modern surface. As for the Ohalo sample, the absent may be due to sampling bias, due to hand picked method used. In the random squares it exhibit differential preservation according to sediment. 12. Otoliths and cyprinid's Weberian apparatus were absent from all samples. 13. Most of the bones exhibit brown and dark brown color and were well preserved . 14. FTIR analysis reveal the presence of collagen in the dark bones, dated to 1,300 years ago. 15. Acanthobrama sp. estimated body size show abundance of small juvenile fish. Fish longer than 13 cm (standard length) were not found.

80 CHAPTER 7: FISH BUTCHERING METHODS "Study of how bones are broken is geared generally towards ascertaining the taphonomic agent responsible for the fragmentation" (Lyman, 1994: 338) 7.1. Butchering and Utilization Methods There are several similarities in procurement, utilization and butchering methods observed for the fishermen in Panama and Sinai. First, fresh fish are regularly consumed at the fishing camp or at the procurement locality. This includes consumption of whole fish and disposal of the skeletal elements next to the fireplace. The second pattern observed concerns the disposal of unused fish organs. An effort is made to dispose of unused organs in water, where they will be washed away by either the river or ocean, depending on the fish camp location. In other cases these parts were consumed by surrounding animals. The third general pattern observed concerns the consistent appearance of three butchering methods (Figure 43). One method is applied to chondrichthyes (sharks), and involves the removal of straps of meat, which are sliced, while all skeletal parts are discarded. A similar method is described in the literature for whales, and large fatty osteichthyes such as Salmon (Belcher, 1993; Belcher, 1994; Burt, 1988; Butler, 1996; Stewart, 1982; Stewart, 1989). The other two methods are applied on osteichthyes in which either whole fish are opened along their belly and their entrails are discarded (Method-1; Figures 43-44) or fish that are split along the back, including the skull (Method-2; Figures 43-44). In method-2 the fish is opened from the dorsal region and the exposed entrails are discarded. The antiquity of these methods dates back at least 5,000 years, as observed from Egyptian's relief (Cutting, 1955; Forbes, 1955; Ikram, 1995; van Elsbergen, 1997).

81 Marine Fish

Chondrichthyes (Sharks) Osteichthyes (Bony fish)

Method-1 Method-2 Whole with skull intact Split dorsally along Straps of Meat Entrails, skull, ventral midline cut vertebral column and skull Fins, Vertebrae

Salt/Brine Salt/Brine Salt/Brine

Gills, entrails Dryin g Dryin gCranial bones Dryin g

Packaging Packaging Packaging Discarded Discarded Storage and distribution Storage and distribution Storage and distribution

Consumption Consumption Consumption

Figure 43: Flow chart for fish butchering methods observed in Parita-Bay (Panama) and south Sinai (Egypt).

Figure 44: Fish butchered by method-1 (left) and by method-2 (right). Regardless of the butchering method, in order to improve drying and to facilitate salt penetration, longitudinal or oblique cuts are performed on both sides of the fish (Figure 44). In some regions (e.g. Ghana, Panama, Sinai), the fish are processed with their skin and scales.

82 Following butchering, the fish are salted either by using seawater or by brining. Then the fish are sun dried either on mats, top of the huts, or hanged on a line. Our initial observations in Panama (Zohar & Cooke, 1997) on 573 fish from 34 species, prepared for salting and drying, demonstrated that the size of the fish, rather than its morphology, determines which of the two butchering methods will be applied (Figure 45; Mann-Whitney test, p<0.001). Most fish smaller than 400 g in body weight and 325 mm in length (S.L.) were butchered by method-1 (whole with intact skull). Most fish larger than this size were prepared using method-2 (split length wise, through the skull) (Zohar & Cooke, 1997). 626-650 Method-1 Method-2 601-625 576-600 551-575 Figure 45: Standard length 526-550 501-525 frequency distribution of 573 fish 476-500 451-475 belonging to 34 species butchered 426-450 401-425 by the two different techniques. 376-400 351-375 326-350 301-325 Contrary to the size Standard Length (mm) 276-300 251-275 dependent pattern observed in 226-250 201-225 Panama, the Bedouin in Nabek 176-200 151-175 Oasis, butchered all fish (72 fish 126-150 101-125 from 5 families and 9 species) by 25 20 15 10 5 0 5 10 15 20 25 method-2 (through the skull), Frequency (%) regardless to their body size or anatomy. I performed a detailed osteological study on 17,862 bones from 147 fish butchered by the fishermen in Panama (61 butchered by Method 1, and 86 by Method 2; Table 53), to examine patterns of bone loss (discarded bones that will be absent from processing and consumption sites) and damage.

83 7.2. Skeletal representation Skeletal completeness was examined in two species of fish butchered by method-1, and three species butchered by method-2 (Table 53). Bones were lost, for all species, regardless of butchering method applied. Moreover, the cranial region is always under-represented, while the post-cranial region is over-represented (Table 53). This difference is statistically significant (Fisher exact test p<.0001). Table 53: Ratio of cranial to postcranial bones in butchered fish and the ratio expected in a complete skeleton. Taxonomic Butchering Crania (%) Postcrania (%) Fisher Exact Test group Count method Obs. Exp. Obs. Exp. χ2 p H.nitidus 32 1 66.23 75.0 33.77 25.0 695.48 <.0001 C.multiradiatus 29 1 57.59 64.0 42.41 36.0 36.01 <.0001 S.troschelii 30 2 53.85 63.0 46.15 37.0 75.43 <.0001 A.kessleri 28 2 54.36 63.0 45.64 37.0 63.15 <.0001 C.caninus 28 2 65.81 75.0 34.19 25.0 85.37 <.0001

The representation of anatomical regions in butchered fish differ significantly from that expected in a complete fish (Tables 54-55), regardless of butchering method or taxonomic group. The low SI of the branchial region (ca.1-30%) indicates that this region is usually lost, regardless of taxonomic group or butchering method. This pattern can be attributed to fish gutting during the butchering process. Table 54: Observed and expected NISP and their survival index (SI) for anatomic regions in fish butchered by method-1. Anatomical region H. nitidus (n=32) C. multiradiatus (n=29) obs. exp. SI obs. exp. SI Crania Neurocranium 1241 1536 .81 1097 1247 .880 Branchial arch 13 1280 .01 84 638 .132 Hyoid region 425 512 .83 338 406 .833 Oromandibular reg. 632 640 .99 510 522 .977 Opercular series 256 256 1.00 174 174 1.000 Postcrania Appendicular reg. 445 512 .87 230 232 .991 Median fins 32 32 1.00 87 145 .600 Vertebral column 832 832 1.00 1305 1305 1.000 Total 3876 5600 .692 3825 4669 .819 χ2 test (df=7) χ2=1005.30, p<0.0001 χ2= 375.397; p<0.0001

In marine catfish butchered by method-2, the median fin SI is low (10%). This pattern can be attributed to the tendency, of fishermen ,to break fish spines upon their capture. Although the neurocranial region exhibits low SI for all fish, it is lower for fish butchered by method-2 (SI<70%). This indicates that 30% of the bones are lost due to butchering through the skull.

84 Table 55: Observed and expected NISP and their survival index (SI) for anatomic regions in fish butchered by method-2. Butchering Method-2 Anatomical region S.troscheli (n=30) A. kessleri (n=28) C. caninus (n=28) obs. exp. SI obs. exp. SI obs. exp. SI Crania Neurocranium 871 1290 .675 829 1204 .689 864 1176 .735 Branchial arch 192 660 .291 33 616 .054 10 1120 .009 Hyoid region 318 420 .757 322 392 .821 392 448 .875 Oromandibular reg. 539 540 .998 501 504 .994 558 560 .996 Opercular series 180 180 1.000 168 168 1.000 224 224 1.000 Postcrania Appendicular reg. 240 240 1.000 224 224 1.000 392 448 .875 Median fins 30 150 .200 16 140 .114 28 28 1.000 Vertebral column 1410 1410 1.000 1316 1316 1.000 672 672 1.000 Total 3640 4890 .744 3409 4564 .746 3112 4676 .665 χ2 test (df=7) χ2=295.136, p< 0.0001 χ2=542.383, p< 0.0001 χ2=877.579, p< 0.0001

In each anatomical region certain bones are lost more than others, due to the butchering method applied. For example, in large marine catfish the supraoccipital crest (from the neurocranium region) is frequently removed (64%) during butchering through the skull (Table 56). Table 56: The most commonly absent bones relative to the butchering method applied. Region Skeletal Element Method-1 Method-2 Cranium region Tripus 0.6% 21.0% Dorsal spine base 1.2% 32.6% Dorsal spine plate 2.35% 50.0% Supraoccipital crest 2.35% 64.0% Otolith 0.0% 100.0%

Hyoid region Interhyal 1.8% 18.6% Glossihyal 13.0% 28.0%

Branchial region Pharyngial plate 38.0% 32.6% Hypobranchial 45.0% 45.3% Epibranchial 77.6% 79.0% Basibranchial 62.3% 83.0% Ceratobranchial 60.0% 84.0% Pharyngobranchial 88.2% 99.0% Catfish dorsal spine base and plates are also lost due to butchering method-2, and the otoliths are lost only in fish butchered by method-2, due to the breakage of the neurocranium (Table 56). From the hyoid region the interhyal and glossihyal are also lost if fish are butchered by method-2. Although most of the branchial region bones are lost due to the gutting process, their relative abundance varies (Table 56), displaying a higher rate for the epibranchial, basibranchial, ceratobranchial and pharyngobranchial bones. 7.3. Bone fragmentation pattern I examined the relative frequency of damaged bones according to butchering method applied (Table 57). Although in both methods bones are damaged, bones of fish butchered by method 1 are less frequently damaged than bones of fish butchered by method-2.

85 Table 57: The most frequently damaged bone according to butchering method. Butchering Method-1 Butchering Method-2 Anatomic region Skeletal Element n % n % Appendicular Skeleton Cleithrum 58 34.0% 130 75.6% (Pectoral Girdle) Coracoid 34 20.0% 83 48.2% Pelvis 31 18.2% 3 1.7%

Hyoid Region Glossihyal 1 1.2% 9 10.5% Urohyal 14 16.5% 37 4.5% Epihyal 0 0.0 30 17.4%

(Skull region) Dorsal spine base 1 1.9% 30 51.7% Fused vertebrae complex 2 3.8% 58 100%

Neurocranium Basioccipital 2 2.4% 77 89.5% Ethmoid 1 1.2% 84 97.7% Exoccipital 1 0.6% 98 57.0% Frontal 1 0.6% 67 39.0% Parasphenoid 1 1.2% 82 95.3% Prefrontal 0 0.0 18 10.5% Prootic 0 0.0 80 46.5% Supraoccipital 2 2.4% 84 97.7% Vomer 1 1.2% 79 92.0%

Opercular apparatus Preopercular 5 3.0% 55 32.0% Opercle 2 1.2% 30 17.4% Interopercle 0 0.0 39 22.7%

Oromandibular Region Dentary 3 1.8% 28 16.3% Metapterygoid 4 2.4% 44 25.6% Premaxilla 2 1.2% 28 16.3% Quadrate 2 1.2% 61 35.5%

The differences in proportion of complete bones in each group are statistically significant (Wilcoxon signed rank test; p=0.0003, z=3.621). The most frequently damaged bones of fish butchered by method-1 are the cleithrum (34%), coracoid (20%) and pelvis bones (18%), from the appendicular skeleton, and the urohyal (16%) from the hyoid region (Table 57). The most frequently damaged bones of fish butchered by method-2 are (Table 57): the ethmoid (97.7%), supraoccipital (97.7%), parasphenoid (95.3), vomer (92%), basioccipital (89.5%), exoccipital (57%), dorsal spine base (51.7%) and frontal (39%), from the skull region. Other frequently damaged bones are: cleithrum (75.6%) and coracoid (48.2%) from the appendicular skeleton, preopercular (32%) and quadrate (35.5%). For each bone I calculated the WMI of fragmentation and discovered that for fish butchered by method-1 (Table 58) the few fragmented bones exhibit a relatively good state of preservation (WMI>90%). In contrast, cranial bones of large catfish butchered by method-2 exhibit low WMI of fragmentation (Table 59). The highly fragmented bones differ between taxa, according to the anatomy of the fish. Table 58: WMI of fragmentation calculated for highly damaged bones of fish butchered by method-1.

86 Species Bone WMI (%) Species Bone WMI (%) S. troschelii Basioccipital 90.00 C. multiradiatus Cleithrum 96.45 Cleithrum 95.00 Coracoid 95.87 Coracoid 97.00 Pelvis 97.42 Ethmoid 95.00 Frontal 98.00 H. nitidus Cleithrum 95.31 Parasphenoid 96.00 Coracoid 99.21 Pelvis 96.50 Pelvis 94.03 Supraoccipital 90.00 Vomer 94.00

Bones of large marine catfish (A.kessleri, S.troschelii) display a higher rate of fragmentation in comparison with C.caninus. For example, in large marine catfish the supraoccipital is highly fragmented (WMI=27-33%), and so are other neurocranial bones such as the vomer, ethmoid, parasphenoid, and basioccipital (Table 59). The cleithrum also exhibits a higher rate of fragmentation in large marine fish compared with C.caninus (Table 59). Table 59: WMI of fragmentation calculated for highly damaged bones of fish butchered by

method-2. Species Bone WMI (%) Bone WMI(%) C. caninus Basioccipital 75.71 Frontal 89.37 Cleithrum 85.02 Parasphenoid 76.96 Coracoid 91.66 Pelvis 99.46 Epihyal 91.19 Supraoccipital 88.12 Ethmoid 75.22 Vomer 88.43

A. kessleri Basioccipital 76.86 Frontal 95.27 Cleithrum 73.68 Parasphenoid 51.09 Coracoid 91.33 Supraoccipital 33.94 Epihyal 94.92 Vomer 52.49 Ethmoid 56.43

S. troschelii Basioccipital 57.09 Frontal 97.00 Cleithrum 64.97 Parasphenoid 51.85 Coracoid 91.35 Supraoccipital 27.41 Epihyal 97.34 Vomer 49.40 Ethmoid 50.94 Based on the survival index calculated, I examined the relative measure of similarity in overall breakage pattern between each of the species butchered. I used the matrix of Euclidean distance in Multidimensional scaling (MDS) to separate butchering methods (Borg, 1981). According to the MDS analysis, morphologically distinct species butchered by the same method display a similar bone breakage pattern (Figure 46). The MDS two-dimensional representation explained 99% of the variance in the original data, giving a stress factor of 0.00088.

2 Figure 46: 1.5 Method-2 Method-1 Multidimensional scaling 1 (MDS) plot of SI of bones C. caninus 0.5 showing a separation H. nitidus 0 S. troschelii between butchering C. multiradiatus -0.5 A. kessleri Dimension 2 methods, regardless of fish -1 taxonomy and anatomy. -1.5 -2 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 Dimension 1 7.4. Fractures typology Typical fractures on the most frequently damaged bones of fish butchered by method-1 were recorded (Table 60). The distal end of the left cleithrum is fractured in 55.3% of the bones (Figure 47). This is due to a transverse cut formed on the fish to facilitate salt absorption. Although less frequently, the coracoid is also damaged at its distal end due to the same transverse cut (Figure 47). The distal end of the pelvis and median area were damaged by a transverse or longitudinal cut .

Figure 47: Typical fractures observed on the cleithrum and coracoid of fish butchered by method-1. Table 60: Types of fractures observed on the most frequently damaged bones of fish butchered by method-1. Bone Breakage typology left side right side Cleithrum Breakages area distal end 55.3% 5.8% proximal end 2.3 1.2% median 1.2% 1.2% none 36.47% 91.76% cut mark 4.7% 0.0% Cut description transverse 63.5% 8.24% none 36.5% 91.76% Coracoid Breakages area distal end 23.8% 11.8% proximal end 1.2% 2.3% none 73.8% 85.9% fragmented 1.2% 0% Cut description longitudinal 1.19% 2.35% transverse 23.8% 11.76% fragmented 1.2% 0% none 73.8% 85.9% Pelvis Breakages area distal end 8.4% 9.4% median 4.8% 2.3% none 80.7% 88.2% fragmented 2.4% 0%

88 median & lateral 3.6% 0% Cut description longitudinal 2.4% 1.2% transverse 10.8% 10.6% fragmented 6.0% 0% none 80.7% 88.2%

In fish butchered by method-2, I observed a distinctive damage pattern on many bones (Table 61). On the cleithrum (left & right) the fracture is transverse, as in method-1, but its location is either median or on the proximal end (Table 61; Figure 48). The coracoid is damaged on the proximal end (Figure 48) and less frequently on the distal end (Table 61). The epihyal bone is damaged by a transverse cut located on its distal end or median. The frontal bones are damaged laterally, as a result of the longitudinal cut performed along the skull (Figure 49). The bones located along the skull midline such as the ethmoid, vomer, parasphenoid, supraoccipital and basioccipital, are cut in the middle and less frequently, laterally (Figure 49). Table 61: Breakages typology for the frequently damaged bones by butchering method-2. Bone Breakage typology left side right side Cleithrum Breakage area distal end 8.1% 9.3% proximal end 33.7% 27.9% median 32.5% 20.23% Median & lateral 3.5% 1.2% median & distal 0.0% 3.5% lateral 1.2% 0.0% cut mark 4.6% 11.6% none 16.3% 16.3% Cut description longitudinal 2.3% 1.2% transverse 81.4% 81.4% longitudinal & transverse 0.0% 1.2% none 16.3% 16.3% Coracoid Breakage area distal end 10.5% 11.6% proximal end 31.4% 34.88% median 0.0% 2.3% median & lateral 1.2% 2.3% lateral 1.2% 1.2% cut mark 3.5% 3.5% none 52.3% 44.2% Cut description longitudinal 5.8% 8.2% transverse 40.7% 45.88% longitudinal & transverse 1.2% 2.35% none 52.3% 43.5%

Epihyal Breakage area distal end 9.3% 5.8% proximal end 2.3% 0.0% median 13.9% 3.5% cut mark 2.3% 2.3% none 72.1% 88.4% Cut description transverse 28.2% 10.7% longitudinal 0.0% 1.2% none 71.8% 88.1%

89 Table 61 cont'd Bone Breakage typology left side right side Frontal Breakage area distal end 1.2% 0.0% proximal end 4.6% 1.2% lateral 20.9% 47.7% median 0.0% 1.2% cut mark 2.3% 0.0% none 70.9% 50% Cut description longitudinal 20.2% 47.7% transverse 9.5% 2.3% none 70.2% 50.0%

Figure 48: Typical fractures observed on the cleithrum and coracoid

from fish butchered by method-2.

Figure 49: Typical fractures observed on cranial bones situated along the longitudinal transverse cut of catfish butchered by method-2. 7.5. Summary From the 17,862 bones of 147 fish butchered by two methods for long-term preservation, I observed the following osteological patterns: 1. In Panama, the butchering method applied is size dependent. Large fish (>35 cm) were butchered dorsally through their skull (method-2), while small fish (<35 cm) where processed whole and gutted ventrally (method-1). 2. Detecting a fish processing local may be impossible since the gutted sections are usually thrown to the water or consumed by local animals. 3. The absence of the branchial arch region characterize gutted fish that been processed for immediate or long-term consumption. 4. Regardless of the butchering method applied, the cranial region is always under- represented while the post-cranial region is over-represented. 5. In fish butchered by method 2 (split) several cranial bones will be lost (absent from processing and consumption sites). Moreover, cranial bones that are situated along the longitudinal cut are frequently damaged and therefore their chance to survive is small.

90 6. The bone breakage typology demonstrates that several fractures appear consistently on distinctive bones and therefore can be associated with the butchering method applied. Similar types of fractures were observed on bones of fish butchered by the Bedouins' in Sinai. 7. The MDS analysis demonstrated that morphologically distinct species butchered by the same method display a similar bone breakage pattern.

CHAPTER 8: OHALO-II: NATURAL OR CULTURAL ACCUMULATION? "If we interpret the past strictly in terms of anthropocentric laws, we will never truly understand what the fossil record has to tell us" (Martin, 1998)

In this chapter I examine the depositional nature of fish remains recovered at Ohalo-II. The potential of naturally deposited fish remains to become incorporated into prehistoric cultural deposits and to provide a background noise, has been recognized in other coastal sites around the world (Butler, 1987; Butler, 1993; Stewart, 1991). Given the nature of Ohalo-II as a water-logged site and the potential of fish remains to accumulate naturally, we cannot simply assume that the fish remains resulted exclusively from cultural activity. Variation in depositional and functional processes at Ohalo-II was examined by using fish remains from different structures, and by comparison with those from the adjacent natural accumulation and with the ethnographic model (modern butchered fish). I compared the Ohalo-II fish remains with fish remains resulting a) from natural deposits and b) from cultural activity. Various quantitative and qualitative criteria were used to determine the nature of fish bone assemblages at Ohalo-II: taxonomic composition, breadth, richness, and diversity; skeletal representation, skeletal completeness, bone modification, bone dispersion pattern, and vertebra size (Table 62). Table 62: Comparison between fish remains recovered in the natural accumulation and at the various structures at Ohalo-II site. Criteria Ohalo beach natural assemblage Ohalo-II archaeological site Taxonomic composition: Taxonomic breadth Narrow: mollusks and fish. Wide: mollusks, fish, mammals, micro-fauna, birds, turtles, etc. Species representation 16% of the local fish species 30-40% of the local fish species Taxonomic richness Low:2-5 High: 5-8 (L.1=5; L.7-8) Taxonomic diversity Low 0.38- 1.594 L.1- low 0.865; L.2, 3, and 8= 1.17- (Brillouin's Index) 1.8; L.7-high 2.398 Family representation Cichlidae (6%), Cyprinidae (80%), Cichlidae (20%), Cyprinidae (80%). Clariidae (6%) Abundant taxa Acanthobrama sp., small cyprinids L.1-Acanthobrama sp., small cyprinid Other loci: Barbus sp., Capoeta sp., Cichlidae

Table 62 cont'd. Criteria Ohalo beach natural assemblage Ohalo-II archaeological site Skeletal representation: Branchial region remains Only pharyngeal bone & teeth Only pharyngeal bone & teeth. Branchial region SI Cichlid SI is 0.2 in sand to 0.55 in Cichlid SI is under-represented in all brown deposits. loci (0.028-0.188). Over-represented for Acanthobrama L.1- SI relatively high for sp. in clay SI=1.26. Under- Acanthobrama sp. (SI=0.624) and represented for all other deposits large cyprinids (SI= 0.755). SI= 0.08-0.553 L.7- over-represented for large cyprinids (SI=1.01), and present for Acanthobrama sp. (SI= 0.755). L.8- High for large cyprinids (SI=0.825). Hyoid region SI Over-represented or highly present Under-represented for all taxonomic for most taxonomic groups, except groups, regardless of loci, expect for for cichlids remains. Regardless of Acanthobrama sp. from L.1. sediment. Oromandibular region SI Over-represented for small cyprinids L.1 Acanthobrama sp. similar with from Brown layer (1.43), cichlids complete fish (0.9) other taxa under- from sand layer (1.17), represented Acanthobrama sp. from clay layer L.8 over-represented for small and (1.95). large cyprinids. Under-represented Other samples under-represented for Cichlidae. L.7 Large cyprinid similar with the expected, other taxa under- represented/ absent. Opercular series Small cyprinids over-represented L.1 under-represented in all taxa (1.3-3.07), regardless of layer except for Acanthobrama sp. (1.4). Cichlid over-represented in sand L.7 absent and highly under-re (2.9), under-represented in clay presented for all taxa. (0.88) and brown (0.5) layers. L.8 Over-represented for small Acanthobrama sp. over-represented cyprinids (4.4) and cichlids (1.2). in clay (1.95) Under-represented for large cyprinids Clarias under-represented (0.52) (0.2). Median fins Over-represented for all taxa, L.1 over-represented for small (2.5) regardless of depositional layer. and large (1.2) cyprinids. Under- represented for cichlids (0.6) and Acanthobrama sp. (0.8). L.7 Over-represented for all taxa, except for Acanthobrama sp. (0.4). L.8 Over-represented for all taxa. Weberian apparatus SI Absent L.1 similar with complete fish for small cyprinids, present but under- represented for other cyprinids. L.7 present but highly under- represented L.8 present (0.6 for large cyprinid and 0.8 for small cyprinids).

Table 62 cont'd. Criteria Ohalo beach natural assemblage Ohalo-II archaeological site Appendicular region SI Over-represented for small L.1 resemble complete fish for cyprinids, in a rising scale according Acanthobrama sp. Other taxa are to depositional depth (SI =1.3 in under-represented (SI=0.1-0.4). sand to 2.24 in clay). L.8 slightly under-represented for Over-represented for cichlids from large cyprinids and cichlids (SI=0.8). clay layer, and under-represented in sand and brown layers (SI=0.5-0.7) Clarias sp. from Ohalo surface are over-represented (SI=1.4) Vertebral column SI Over-represented for all taxa in all Over-represented for all taxa, in all samples loci. Scales Many scales from sand and brown No scales layers. Clay sediments-no scales Otoliths No otoliths Few otoliths, mainly of Cichlidae from L. 8

Crania vs. postcrania Crania is over-represented for Acanthobrama from L.1 similar to Acanthobrama sp., large Cyprinidae natural. Other taxa: cranial bones (90%) and Clariidae (55%). Under- under-represented (2-25%). represented for small cyprinids L.7 large cyprinid under-represented (30%) and cichlids (26%) but more abundant (40%)

Color +burning signs Brown (light to dark) no burning L.1-99% brown 1% burnt signs L.2-100% Brown L.3-90% Brown, 10% burnt L.7- 88% Brown, 12% burnt L.8-97.5% Brown, 2.5 % burnt L.9-98% Brown, 2 % burnt Fragmentation Most of the bones are slightly Good preservation (>80%) vary: fragmented:80-100% of the bone L.1-57% present at 85% of the bones. L.3- 42% L.8- 51% L.9-56% L.7-80% Erosion Trampling, breakage, weathering Trampling, breakage, weathering

Estimated body size Acanthobrama SL: 65-127mm L.1 Acanthobrama SL:84-185 mm Fish exploitation index Low 0.182 L.1 low 0.110 L.7 0.601 Loci 2, 3, and 8 high (>0.95)

Table 62 cont'd. Criteria Ohalo beach natural assemblage Ohalo-II archaeological site Mean bone scatter 423 bones per 0.5 sqm (range 8- L.1 930 bones per 0.5 sqm frequency (BSF) 2840 bones) L.7 842 bones per 0.5 sqm L.3 190 bones per 0.5 sqm Bone dispersion pattern Clumped Clumped

8.1. Taxonomic composition, richness and diversity The freshwater fish in the Sea of Galilee and the Jordan River system belong to seven families comprising 30 species (Appendix II). As human activity might be targeted toward a limited number of species (Stewart, 1989; Stewart, 1991; Stewart & Gifford-Gonzales, 1994; Zohar et al., 2001), I expected to find higher species richness in the natural accumulation compared to that resulting from human activity. Moreover, Stewart (1991) has demonstrated that in a natural accumulation fish smaller than 350 mm in length are absent. In contrast to my expectations, only 16% of the recent species in the Sea of Galilee were present in the natural accumulation, compared to 40% obtained at the Ohalo-II site. Moreover, the most abundant fish remains in the natural accumulation belonged to taxa smaller than 20 cm in length. These results differ from those obtained by Stewart (1991) for Lake Turkana which exhibited higher taxonomic diversity for the natural accumulation than at archaeological sites. Similar results were obtained by O'conell and Tunnicliffe (2001), who found that only 20% of the local species were present in a lacustrine natural accumulation. The taxonomic richness found in the natural accumulation (2-5) was also lower than the values obtained for locus 7 at the Ohalo-II site (5-8). Species richness at locus 1 (5) was similar to that of the natural accumulation. Since taxonomic richness may be biased due to differences in sample sizes (see chapter 5.1), I applied rarefaction analysis to the data (Figure 50). As observed in figure 50, differences in sample sizes did not bias taxonomic richness.

Locus 1

14 Locus 7 13 Natural Species no. 12 11 10 9 8 7

Species richness 6 5 4 5 40 75

110 145 180 215 250 285 320 355 390 425 460 495 530 565 600 635 670 705 740 775 810 845 880 915 950 985 NISP Figure 50: Rarefaction curves of species richness in loci 1, 7, and the natural accumulation as a function of NISP.

The Brillouin diversity index (HB) also presents low values for the natural accumulation (HB=0.38-1.594) compared to loci 2, 3, 7, and 8 at Ohalo-II (HB=1.17-2.398). However, HB value calculated for locus 1 (HB=0.865) falls within the HB values of the natural accumulation. Therefore, only loci with HB higher than 1.594 (loci 7 and 8) differed significantly from the natural accumulation (Magurran, 1988). Although the natural accumulation exhibited a narrow species taxonomic range, family representation was more diverse than that observed at Ohalo-II (Table 62). In the natural accumulation three families were present (Cyprinidae, Cichlidae, and Clariidae), compared to two at Ohalo-II (Cyprinidae and Cichlidae). When calculated by NISP this difference is statistically significant (Fisher exact test: χ2=54.338, DF=1, p<.0001). While Cyprinidae were highly abundant in both assemblages (80%), Cichlidae were rare in the natural accumulation (11%). The absence of Clariidae remains from the Ohalo-II site may be attributed either to preservation bias in clay deposits, to environmental changes, or to human activity. Clariidae remains from the natural accumulation clearly exhibited a preservation bias toward the upper sand deposits and total absence from the lower brown (median layer) and clay deposits. Since Clariidae remains are abundant in other Epi-paleolithic sites from Africa (Brewer, 1991; Van Neer, 1999; Van-Neer, 1995), there is a possibility that the inhabitants of Ohalo-II did not exploit Clariidae due to lack of technological skills or from dietary preferences (see next

chapter). The possibility that Clarias sp. is absent due to environmental changes is also discussed in depth in the next chapter. I compared between the relative abundance of the taxonomic groups in the natural accumulation and different loci from Ohalo-II (Figures 51-52; Table 63). The abundance of Acanthobrama sp. and small cyprinid remains in Locus 1 is similar with the natural accumulation (Figures 51-52; Table 63). Loci 3 and 8 differ from locus 1 and the natural accumulation due to the preponderance of large cyprinids and cichlids, and the absence of Acanthobrama sp. and small cyprinids. Interestingly, locus 7 differs from all the other loci at Ohalo-II, as well as from the natural accumulation. This can be attributed to the appearance of Acanthobrama sp. and small cyprinids, combined with higher abundance of large cyprinids and cichlids (Figures 51-52). Correspondence analysis reveals that 98% of the differences between the samples are explained by factors 1 and 2 (Table 63). Table 63: Statistics for the correspondence analysis plot outlined in figure 52. Dimension Singular Value Inertia Proportion Explained Cumulative Proportion 1 0.61217 0.37476 0.761 0.761 2 0.26518 0.07032 0.143 0.903 3 0.21704 0.0471 0.096 0.999 4 0.02241 0.0005 0.001 1 Total 0.49269 1 1

100

80 Natural Natural 70 60 Locus-1 Locus 8 50 40 30 Frequency (%) 20 10 0 80 70 Natural Natural 60 Locus 7 Locus 3 50 40 30 Frequency (%) 20 10 0 Cichlids Cichlids Acanthobrama Small cyprinid Acanthobrama Small cyprinid Large cyprinids Large cyprinids Taxonomic groups Figure 51: Taxonomic groups percentage (%) in the natural accumulation and Ohalo II.

4.00 Clariidae

2.00 Natural L7 Cichlids Small Cyprinids L1 0.00 L8 Acanthobrama RDim 2 L3 Large Cyprinids

-2.00 L2

-4.00 -2.50 -2.00 -1.50 -1.00 -0.50 0.00 0.50 1.00 RDim 1 Figure 52: Correspondence analysis of taxonomic groups relative abundance (%) in the natural accumulation and loci 1, 2, 3, 7 and 8.

101

In sum, the natural accumulation is characterized by a wide representation of family and narrow presentation of species. This finding differs from that of Lake Turkana, where a wide taxonomic range characterized natural accumulation and a narrow taxonomic range characterized results of human activity (Gifford-Gonzales et al., 1999; Stewart, 1991; Stewart, 1994). In the case of Ohalo-II we may assume that the wide species taxonomic range observed in loci 2,3,7, and 8 might be assigned to human activity; whereas the narrow taxonomic range observed at locus 1 resembles that observed for the natural accumulation. The most abundant taxa in the natural accumulation were Acanthobrama sp. and small cyprinids, which feature fish smaller than 200 mm in length. This pattern differs from that observed by Stewart (1991) for Lake Turkana, where skeletons of fish smaller than 350mm were not preserved. It also differs from the pattern observed for loci 2, 3, and 8 at Ohalo-II (Figure 52). This may indicate "background noise" reflecting natural death of fish at loci 1 and 7. Soil geomorphic analysis showed that at Ohalo-II several episodes of flooding occurred due to a rapid rise of lake levels immediately after each phase of occupation (Nadel et al., 2001; Tsatskin and Nadel, 2003). During these phases A. terraesanctae were probably deposited. However, when large cyprinids (Barbus sp. and Capoeta sp.) and cichlids appear at high frequencies, they attest to human activity (Figure 52). The preservation of these taxonomic groups in natural accumulations is relatively low, since they suffer from bone disarticulation and scattering by scavengers and wave energy, as has been observed in other lacustrine environments (Elder & Smith, 1988; Ferber & Wells, 1995; O'connell & Tunnicliffe, 2001). 8.2. Skeletal representation Skeletal representation can be affected by many factors, such as mode of death, bone density, water temperature, decomposition, sediment pH, transportation, butchering, cooking, consumption techniques, and deposition locality (Butler, 1993; Butler, 1994; Butler & Schroeder, 1998; Elder & Smith, 1988; Fred et al., 2002; Lubinski, 1996; Nicholson, 1992a; Nicholson, 1992b; Nicholson, 1996b; Stewart, 1989; Zohar et al., 2001). My ethnographic study agrees with other research demonstrating the affect of human processing and later consumption, on element representation (Belcher, 1994; Butler, 1996; Hoffman et al., 2000; Stewart, 1982; Stewart, 1989; Zohar & Cooke, 1997). Because of differences in the proportions of the carcass originally deposited in natural and cultural fish accumulations, and major differences in destruction agents, I assume that skeletal representation will differ between the assemblages. 8.2.1 Body part representation

102

In Panama and Sinai I found that the branchial region is discarded (SI= 0.009-0.29) during the preparation of fish for immediate or later consumption (See chapter 7). I expected that in natural accumulation bones from the branchial region would be more abundant compared to culturally deposited fish. Contrary to my expectations, however, I found that this region demonstrated low survivorship (SI= 0.08-0.553; Figure 53) in the natural accumulation. The branchial region was over-represented only for Acanthobrama sp. remains in the clay deposits (SI=1.26; NISP=47), and consisted exclusively of pharyngeal teeth and bones. Most of the bones from the branchial region were absent. Therefore, it appears that the branchial region can not be used as marker of human activity. The hyoid region SI is slightly under-represented in butchered fish (~80%). In the natural accumulation this region was under-represented exclusively for cichlid remains (SI=0- 0.4). However, a similar pattern was also observed for the cichlid remains from Ohalo-II (SI=~0.1; Figure 53). This region was over-represented for Acanthobrama sp. remains from locus 1 (SI=2.091), and from the natural accumulation clay deposits (SI=1.01; Figure 53). In the natural accumulation random squares sand deposits, the hyoid SI was under-represented (0.488). These results suggest that the preservation of the hyoid region may be biased due to depositional processes, in favor of clay deposits and Acanthobrama sp.. Therefore, this region too can not be used for distinguishing culturally deposited from naturally deposited assemblages (Figure 53).

6 5.5 Locus 1 Natural accumulation clay deposits Acanthobrama sp . 5 4.5 4 Small carp 3.5 3 Barbus sp./ Capoeta sp . 2.5 2 Cichlidae 1.5 Survival Index (S.I.) 1 0.5 0 6 5.5 Locus 7 Locus 8 5 4.5 4 3.5 3 2.5 2 1.5

Survival Index (S.I.) 1 0.5 0

Anatomic region Median fins Median fins Hyoid region Hyoid region Neurocranium Neurocranium Opercular series Opercular series Branchial region Branchial region Vertebral column Vertebral column Weberian apparatus Weberian apparatus Oromandibular region Oromandibular region Appendicular skeleton Appendicular skeleton Figure 53: Survival index, (SI) by anatomical regions, for loci 1, 7, and 8 and the clay deposits of the natural accumulation.

103

The oromandibular region was slightly under-represented in butchered fish (~80-90%). In the natural accumulation it was over-represented for cichlid remains from sand deposits (SI= 1.17), for small cyprinids remains from the median brown deposits (SI= 1.43), and for Acanthobrama sp. remains from the clay deposits (SI=1.95; Figure 53). In other layers the oromandibular region is under-represented. Accordingly survivorship of the oromandibular region varied according to fish species and depositional location. At Ohalo-II the oromandibular region was under-represented in most samples except for Acanthobrama sp. from locus 1, small and large cyprinids from locus 8 (SI= 2.22 and 1.19 respectively; Figure 53), and large cyprinids from locus 7. The high SI observed for Acanthobrama sp. and small cyprinid resembles the pattern observed in the natural accumulation. Since in large cyprinids the oromandibular region was absent in the natural accumulation, the high SI observed in loci 7 and 8, as well as that observed in locus 1 (0.599), may be attributed to human activity. The opercular region was undamaged in butchered fish. In fish natural accumulation this area is expected to decay relatively fast in relatively large fish, depending upon depositional processes and water temperature (Wilson & Barton, 1996). As expected, the opercular region was under-represented in a natural accumulation but not for all taxonomic groups, being over-represented for small cyprinids from all layers (SI >1.3-3.07, Figure 53), and for cichlids from the upper sandy layer (SI=2.92). This preservation pattern is not surprising and also concurs with other taphonomical studies demonstrating that in temperatures lower than 15oC fish carcasses will preserve better (Wilson & Barton, 1996). At Ohalo-II a similar pattern was observed for the Acanthobrama sp. opercular region, which was over-represented at locus 1 (SI=1.339; Figure 53). At locus 8 the opercular region was over-represented for small carps and Cichlidae (SI= 4.44 and 1.18 respectively), while it was under-represented for large cyprinids from all loci (SI =0.1-0.2), and for cichlids from loci 1 and 7 (SI=0.13 and 0.21 respectively; Figure 53). The appendicular skeleton was presented in butchered fish (80-100% present). In natural accumulation this region was over-represented for small cyprinids (SI >2.8; Figure 53). Moreover, the survival index (SI) exhibited a tendency to increase according to depositional depth, from 1.3 at the surface to 2.4 at the bottom layer (Figure 53). This region was also over-represented for cichlids from the clay layer and for Clarias sp. from the Ohalo- II surface (SI=1.4), while large cyrinid and cichlid remains from the sand and brown layers were under-represented. At Ohalo-II the appendicular skeleton was over-represented,

104

resembling the natural accumulation only for small carps from locus 1 (SI=1.4; Figure 53), while it was under-represented for all other taxa from the various structures. Not surprisingly, vertebrae are over-represented in all samples, regardless of their origin, depositional area, or taxonomic groups (Figure 53). Various studies have demonstrated that fish vertebrae centrum survivorship is not influenced by destructive cooking methods (Nicholson, 1998). Moreover, they exhibit relatively high density in comparison with other skeletal elements, which probably increases their survival chances (Butler, 1994; Falabella et al., 1994; Nicholson, 1992a; Zohar et al., 2001). Bone density of other skeletal elements was not examined since previous studies demonstrated a lack of correlation between density and survivorship (Falabella et al., 1994; Hoffman et al., 2000; Zohar et al., 2001). Several anatomic regions were absent from the natural accumulation but present in low frequencies elsewhere at Ohalo-II (Figure 53). These include cyprinid Weberian apparatus found at loci 1, 7 and 8. For small cyprinids from locus 1, the Weberian apparatus SI resembled that of a complete fish, and was slightly under-represented for Acanthobrama sp. and large cyprinids (SI= 0.6 and 0.8 respectively). In loci 7 and 8 this region was present but under-represented for all cyprinids (Table 62). Otoliths were absent from the natural accumulation, but although rare were found at Ohalo-II. These were mainly cichlid sagittal otoliths recovered at locus 8 (See Appendix XIII). Other studies have demonstrated that the exact mechanism and pathways of otolith elemental alteration and preservation are not known, but are clearly affected by cooking methods and soil pH (Fred et al., 2002). Butchering methods may also influence otolith preservation, as observed in Panama, where otoliths are absent from fish butchered by method-2 (through the skull). The absence of otoliths from the natural accumulation, regardless of depositional layer, demonstrates that environmental factors influenced their preservation. The preservation of otoliths at locus 8 attest to differential pre- and post- depositional diagenesis (soil pH , preparation methods) compared to other loci from the Ohalo-II site and the adjacent shore. Large amounts of fish scales appeared in natural accumulation in upper sand and in median brown deposits, but are absent from bottom clay deposits. This pattern of differential preservation explains the absence of scales from Ohalo-II clay deposits, and is in agreement with other taphonomical studies demonstrating that scales abundance becomes reduced with time (O'connell & Tunnicliffe, 2001).

105

The neurocranium region was of low SI in all samples, except for recent Clarias sp. (Figure 53). This is partially due to fragmentation and identification problems. Although I did observe many neurocranial fragments, I could neither identify nor count all fragments. In sum, skeletal representation of Acanthobrama sp. and small cyprinids from loci 1 and 7, resemble that observed for the clay layer of the natural accumulation. The Weberian apparatus was the only region that appeared in locus 1 and was absent from the natural accumulation. For large cyprinids and cichlids the relative abundance of most skeletal elements differed from that observed in the natural accumulation. The influence of pre- and post-depositional processes can be observed from the absence of several skeletal elements such as otoliths and scales. The relative abundance of cranial and post-cranial regions will be used for further differentiation between the assemblages. 8.2.2 Skeletal completeness As has been demonstrated, cranial bones are under-represented in butchered fish. Therefore, I expected to find over-representation of the cranial region in the natural accumulation and under-representation in the cultural accumulation. This trend was described in other taphonomic studies of fish natural and cultural accumulations (Belcher, 1994; Butler, 1993; Hoffman et al., 2000; Lubinski, 1996; Stewart, 1989; Wilson & Barton, 1996; Zohar et al., 2001). As expected, in the natural accumulation the cranial region was over-represented for Acanthobrama sp., and large cyprinids (Figures 54-55). Although it was under- represented for small cyprinids and cichlids (Table 62, Figures 54-55), it exhibited higher frequency in the natural accumulation compared to Ohalo-II (z=11.1850, p<0.05). The ratio of Acanthobrama sp. cranial remains at locus 1 was similar to that found in the adjacent natural accumulation (z=5.93362, p>0.05 NS), emphasizing that in this locus they most likely resulted from natural death. The under-representation of large cyprinids cranial bones at Ohalo-II (z=3.3785, p<0.05) may be related to human activity and consumption methods (see next chapter).

106

90 Acanthobrama sp. Small cyprinidae Natural accumulation 80 70 Locus-1 60 50 Locus-7 40

Frequency (%) 30 20 10 0 Cranial Postcranial Cranial Postcranial

Figure 54: Frequency of cranial and post-cranial bones in Acanthobrama sp. and small cyprinids recovered from the natural accumulation and Ohalo-II.

100 Large Cyprinidae Cichlidae Natural accumulation 90 80 Locus-1 70 60 Locus-7 50 40 Frequency (%) 30 Locus-8 20 10 0 Cranial Postcranial Cranial Postcranial

Figure 55: Frequency of cranial and postcranial bones in large Cyprinidae, and Cichlidae recovered from the natural accumulation and Ohalo-II.

In sum, the relative abundance of Acanthobrama sp. and small cyprinid anatomical regions identified from locus 1 resembled that of the adjacent fish natural accumulation. Moreover, the cranial region was over-represented in Acanthobrama sp. remains from both locus 1 and the natural accumulation (Figure 54). In large cyprinids and cichlids there was a significant difference between skeletal element representation in Ohalo-II compared to the natural accumulation, attesting to human activity. 8.3. Bone modification Bones that have been burned, bear cut marks, fractures, or show sign of digestion are commonly used as indicators of human activity (e.g., Colten, 1995; Lyman, 1982; Nicholson, 1993; Nicholson, 1998; Shipman et al., 1984; Van-Neer, 1995). Bone color has been widely used for identifying burned bones (Koon et al., 2003; Shipman et al., 1984; Stiner et al., 1995). In a water-logged site, such as Ohalo-II, the natural dark color of the bones impedes the ability to identify burnt bones. In the natural accumulation only a few bones from the surface

107

layer were still fresh and therefore exhibited a white color. Most of these bones exhibited a brown color in various stages (light to dark), which resembled the color variation recovered at Ohalo-II site (Figure 56). I examined the bones with a Fourier Transform Infra-red ( FTIR) spectrometer (see chapter 4.5.7) and found that despite the resemblance in bone color, in the natural accumulation the bones displayed collagen in rates similar with a fresh bone, while bones from Ohalo-II site displayed lack of collagen. In addition, dating by 14C (at the University of Arizona, Tucson) demonstrated that the bones from the adjacent natural accumulation clay deposit are relatively recent (ca. 430-620 AD). Burnt bones were absent from the natural accumulation, and rare at Ohalo-II site (Figure 56).

100 Natural accumulation Locus-9 90 80 Locus-1 Locus-7 70 60 Locus-8 Locus-3 50 40 Frequency (%) Locus-2 30 20 10 0 Burnt Brown Black/Dark brown Figure 56: Percentage of burned and brown bones in the natural accumulation and various loci.

In locus 1 only 1% of the bones were burnt compared to 10% at locus 3 and 12% at locus 7 (Figure 56). Since locus 7 comprised ashes, it is possible that the burnt fish remains were either the result of later activity, or fish discarded after consumption (waste). Recent studies have demonstrated that burned bones in archaeological sites can be the result of natural fires (e.g., forest, grass fires). Moreover, even if cultural agents were responsible for a fire that burnt the bones, there is a possibility that the fire postdated the deposited fish remains (Butler, 1996). Recently, it has been demonstrated that cooked bones can not be recognized from bone coloration but rather from morphological changes in bone structure (Koon et al., 2003). Evidence of human activity can be inferred from the bone fragmentation rate, as observed on the fish butchered in Panama (see chapter 7). However, human trampling post deposition can also damage bones deposited in the sediment. Andrews (1995) has shown that in some sites it may be difficult to distinguish between cut marks and trampling marks on bone surfaces. Distinguishing with confidence between bones that have been fractured or broken during butchering, or subfossil bones that were broken later, is difficult (Andrews,

108

1995). Recent studies have demonstrated that cut and digestion marks are rarely observed on fish bones (Butler & Schroeder, 1998; Stewart, 1989). In the natural accumulation, most of the bones were in an excellent state of preservation (80-100% of the bone is present). At Ohalo-II, the state of bone fragmentation differed among loci. At locus 7, most of the bones were well preserved (in 80% of cases, 80-100% of the bone was present), despite the relatively high rate of burn signs. At other loci the bones were more fragmented and the relative abundance of well-preserved bones varied between 42% at locus 3, to 57% at locus 1 (Table 62 ). Table 64: Stress factors and variance explained by MDS analyses of bone fragmentation pattern for butchered fish, natural accumulation, and loci 1 and 7. Comparison Stress Factor r2 Natural vs. Locus 1 0.00031 1.00000 Natural vs. Locus 7 0.00000 1.00000 Butchered fish vs. Natural 0.00000 1.00000 Butchered fish vs. Locus 1 0.11239 0.95283 Butchered fish vs. Locus 7 .01357 0.99911

To examine bone fragmentation pattern, MDS analysis was used and a comparison was made between the modern butchered fish (ethnographic data), the natural accumulation, and Ohalo-II. First, I compared between bones from the natural accumulation and butchered fish and obtained two distinctive patterns (Figure 57). The fish from the natural accumulation (blue) are scattered on the left side of the graph while the butchered fish appear on the right side. The 100% variance in the data was explained by the MDS plot (Table 64). I compared between the bone state of fragmentation from the natural accumulation and bones from loci 1 and 7. I found that in locus 1 the fragmentation pattern for both large and small Cyprinidae was similar to that observed for Acanthobrama sp. in the adjacent natural accumulation (Figure 58), indicating that these bones are well preserved. The fragmentation pattern of cichlid bones from locus 1 differed from that observed in the natural accumulation. The pattern of bone breakage for Acanthobrama sp., however, differed from that in the natural accumulation (Figure 58), indicating that Acanthobrama sp. remains from locus 1 were subjected to a higher rate of abrasion in the last 23,000 years, compared with recent natural accumulation. The 100% variance in the data was explained by the MDS plot (Table 64).

109

1.5 1.25 Acanthobrama N 1 0.75 Small cyprinid N 0.5 0.25 H. nitidus 1 C. multiradiatus 1 0 Cichlid N Dimension 2 -0.25 C.caninus 2 -0.5 S. troschelii 2 A.kessleri 2 -0.75 -1 Large cyprinid N -1.25 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 Dimension 1

Figure 57: MDS analysis plot for fish from the natural accumulation on the left and butchered fish on the right.

110

1 Large cyprinid L.1 Large cyprinid N Small cyprinid N 0.5 Acanthobrama N

0 Acanthobrama L.1 Small cyprinid L.1

-0.5 Cichlid N Dimension 2

-1 Cichlid L.1

-1.5 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5 3 3.5

Dimension 1 Figure 58: MDS analysis plot for bone breakage pattern in the natural accumulation (blue) and locus 1 (black). The fragmentation pattern observed for large cyprinids and cichlids at locus 7 differed from that observed in the natural accumulation (Figure 59). The bone fragmentation pattern of small cyprinids and Acanthobrama sp. from locus 7 was similar to that found for large cyprinid in the natural accumulation. The 100% variance in the data was explained by the MDS plot (Table 64). Figure 59: MDS analysis plot for bone breakage pattern of fish remains in the natural

Small cyprinid L.7 Small cyprinid N

0.5 Large cyprinid N Acanthobrama N Acanthobrama L.7 0 Dimension 2

-0.5 Large cyprinid L.7 Cichlid N Cichlid L.7

-1 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5 3 3.5 Dimension 1 accumulation (blue) and locus 7 (black). I also compared the fragmentation rate of fish remains from two loci at Ohalo-II (1 and 7) and modern butchered fish (Table 64). The fragmentation pattern was found to be

111

significantly different, suggesting that the Ohalo-II inhabitants applied different butchering methods. Ninety five percent of the variance in the data was explained by the MDS plot (Table 64). 8.4. Vertebrae dimension Atlas and axis dimensions (max. width) of Acanthobrama sp. from the reference collection, natural accumulation, and locus 1 from Ohalo-II, were compared (Figure 60). The vertebrae dimensions of Acanthobrama sp. from locus 1 were larger than those from the natural accumulation and the present-day population (Mann-Whitney test, p<0.0001). Since sample sizes were unequal, differences between the samples might be biased (Zar, 1984:130).

Figure 60: Comparison between 3.5 Atlas Acanthobrama sp. atlas and axis mean width 3 Axis 2.5 (mm, ±SD) from recent reference collection, 2 natural accumulation, and locus 1. 1.5 1 0.5 0 Locus 1 Natural Recent 8.5. Fish body size Acanthobrama sp. estimated body size from locus 1 was compared with those from the natural accumulation (Figure 61; Table 65), and exhibited a wider range of body mass and standard length (SL), skewing the graph toward large fish (Figure 61). In locus 1 Acanthobrama sp. maximum SL was 185mm, while in the natural accumulation it was 127mm. A similar difference was observed for the estimated body mass of a maximum 50 gr for locus 1 and 28 gr for the natural accumulation.

Table 65: Acanthobrama sp. estimated body mass (gr) and standard length (mm) for naturally deposit fish and locus 1. Natural accumulation Locus 1 Body size N Mean Std. Range N Mean Std. Range Body mass (gr) 19 14.06 6.00 5-28.0 371 34.84 4.58 12-50.0 Standard length (mm) 19 89.63 16.07 65-127.0 371 145.32 12.28 84-185.0

112

50 Figure 61: Acanthobrama Natural 40 sp. estimated SL from the Locus-1 30 natural accumulation vs.

20 locus 1.

Frequency (%) 10

65-85 86-105 106-125 126-145 146-165 166-185 Standard length (mm) The size distribution observed for A. terraesanctae in the natural accumulation was similar to that observed for present-day juveniles (Ostrovsky & Walline, 1999), while the length distribution observed for locus 1 resembled that of adult males and females (Ostrovsky & Walline, 1999). The differences in length distribution may be biased due to unequal sample size (371 vs. 19), or variation between sampling areas. It has been demonstrated that A. terraesanctae length distribution significantly differs among locations in the Sea of Galilee (Ostrovsky & Walline, 1999) . Another possibility is that the differences in body size distribution resulted from deposition in different seasons. Accumulation of adult A. terraesanctae may occur in winter, during their breeding season along the shore of the Sea of Galilee, whereas accumulation of juvenile fish may occur in spring, due to mass mortality during their first year of life (Ostrovsky & Walline, 1999). Calculation of fish index (the ratio of large fish to the sum of large fish plus small fish taxa; see chapter 4.5.6) at Ohalo-II showed a continuing exploitation of high-ranked fish (large fish) except for locus 1 (Figure 62), where there was a sharp decrease in large fish and increase in small fish. If this pattern results from changes in human economic strategies, it may suggest the occurrence of new technologies for mass harvesting of small fish. In such 1 cases the rank of small fish could be elevated relative to larger fish 0.75 because small fish captured in mass provided higher energetic 0.5 returns (relative to large fish

Fish Index 0.25 caught individually). This might provide the earliest evidence for 0

113 Locus 1 Locus 2 Locus 3 Locus 7 Locus 8 Natural

Locality

small fish mass harvesting in the Middle East (see next chapter). Figure 62: Plot of fish index based on aggregated NISP by locality at Ohalo-II and the natural accumulation.

However, this explanation might also be incorrect in light of the similarity between the low fish index obtained for locus 1 and that from the natural accumulation (Figure 62). Moreover, if fish taken with nets were targeted because they provided greater energetic return than larger fish, why was this not reflected in all structures at Ohalo-II, and in later periods (Zohar et al., 1994)? Further, if mass harvesting devices already existed, they could have been used more successfully for targeting large taxa. Given the caveats presented above, I feel that the Acanthobrama sp. remains at locus 1 are more similar to those in the natural accumulation.

8.6. Bone distribution pattern Bone distribution pattern in the natural accumulation and Ohalo-II are similar in their clumped distribution. Mean bone scatter frequency (BSF) varied between 190 to 842 bones per 0.5 sq.m in Ohalo-II and 423 per sq.m in the natural accumulation. This indicates that in the case of the Sea of Galilee, these criteria can not be used for differentiating between naturally and culturally derived accumulation. At Lake Turkana low BSF was described for both culturally and naturally-deposited assemblages (Stewart & Gifford-Gonzales, 1994), demonstrating that about 1/3 of the foraging camps could not be distinguished from "background noise". Moreover, in the case of high BSF for the naturally deposited fish they concluded that it probably resulted from continuous deposition of bones on the same lakeshore surface over a long period of time (Stewart & Gifford-Gonzales, 1994). Fish natural accumulation along the Sea of Galilee shows that they are deposited continuously on the same lakeshore surface. Since Ohalo-II is a coastal (water-logged) site it is reasonable to assume that the archaeological site is deposited among sediments that contained the "background noise" of naturally deposited fish remains (Table 62). Study of the

114

formation processes at Ohalo-II exhibit evidence of wave action, erosion and several episodes of flooding and inundation (Tsatskin & Nadel, 2003). 8.7. Summary Recent studies have demonstrated that there is a very high probability that almost every archaeological site have had suffered some degree of postdepositional disturbance (Butler, 1990; Chase et al., 1997; Schiffer, 1983; Stewart, 1991). Separating the behavioral component underlying site formation from other natural agencies is therefore essential otherwise data can be mistakenly interpreted as reflecting human behavior alone (Behrensmeyer et al., 1986; Behrensmeyer, 1991; Behrensmeyer et al., 1989; Binford, 1981; Chase et al., 1997). In the case of coastal and inundated sites such as Ohalo-II, fish natural accumulation must be ruled out. By using a large set of quantitative and qualitative criteria, I was able to investigate the nature of fish remains accumulation at the Ohalo-II site. I compared these with data from fish natural accumulation along the southern shore of the Sea of Galilee, and with data collected from butchered fish. My analyses indicated that taxonomic breadth, diversity, and composition, as well as body part representation and skeletal completeness, differ between natural and cultural settings (Table 62). My analyses also demonstrated that bone clumped distribution, high mean BSF, and erosion, cannot serve as distinguishing criteria between cultural and natural activities, as had been previously assumed (Zohar et al., 2001). The variation recorded for fish natural accumulation (i.e., Butler, 1987; Cutler et al., 1999; Elder & Smith, 1988; Ferber & Wells, 1995; Wilson & Barton, 1996) emphasize the need to characterize fish natural accumulation for each locality individually. The fish recovered from the natural accumulation consisted of three fish families (Clariidae, Cyprinidae, and Cichlidae) and a relatively low taxonomic breadth. Only 16% of the recent species from the Sea of Galilee were represented. While Acanthobrama sp. and small cyprinids were abundant, Clarias sp., Barbus sp., Capoeta sp. and Cichlidae were rare. This result differs from that observed at Lake Turkana, where taxonomic diversity in the natural accumulation resembled the living fauna in the lake, and fish with body size smaller than 350mm were absent (Gifford-Gonzales et al., 1999; Stewart, 1991). By using correspondance analysis I demonstrated that the natural accumulation differed from loci 2, 3, 7, and 8. Resemblance in taxonomic breadth and diversity was observed between locus 1 and the natural accumulation. Locus 1 exhibited a preponderance of Acanthobrama sp. and small cyprinids that resembled that found in the natural accumulation. The low fish index calculated for the natural accumulation (~0) also resembled the value obtained from locus 1.

115

The ratio between cranial and postcranial region in Acanthobrama sp. and small cyprinids also point to a similarity between locus 1 and the natural accumulation. Cultural deposits at loci 2, 3, 7, and 8 were characterized by wider taxonomic breadth (30-40%), higher diversity, and the abundance of large cyprinid (Barbus sp. and Capoeta sp.) and Cichlidae (high fish index ~1). They were also characterized by under-presentation of cranial bones (Table 62). Body size distribution of Acanthobrama sp. from locus 1 presented large fish while in the natural accumulation they were mainly of small size. Adult A. terraesanctae are pelagic fish that exploit the littoral only during their breeding season which starts around December. Small juveniles of A. terraesanctae exploit the littoral in May. The differences in body size distribution may indicate fish mass mortality in different seasons. This difference could have resulted also from the sampling methods, since the area examined for natural accumulation is relatively small. Studies have demonstrated differences in A. terraesanctae body size according to sampling areas (Ostrovsky & Walline, 1999). Another option that cannot be disregarded despite the resemblance between locus 1 and the natural accumulation, is that the Acanthobrama sp. and small cyprinid remains from locus 1 resulted from human activity (See chapter 9). Given the relationship between fish body size and resource ranking by the prey choice model (highest energetic returns= largest prey; Butler, 2000), the Ohalo-II inhabitants would have had to invent new technologies and invest more energy for harvesting small fish. Such a strategy differs from that observed for the later Paleolithic populations in Africa (Stewart, 1989), which exclusively exploited large fish that were easy to target with simple methods. Moreover, if small fish targeted with nets provided greater energetic return than larger fish taken singly, why did such a practice did not continue by all inhabitants of the site and in later periods?

116

CHAPTER 9: DISCUSSION AND CONCLUSIONS "Fish", he said, "I love you and respect you very much. But I will kill you dead before this day ends" (E. Hemingway, The Old Man and The Sea, 1952)

Little research has been performed in Israel on fish remains recovered from archaeological sites (Appendix I). Given the antiquity of association between hominid site selection and aquatic habitats, we cannot simply assume that these habitats were not exploited. However, evidence of fish systematic exploitation is still limited and further study is required. In Africa, there is evidence of fish exploitation from 40,000 years ago (Van Neer, 1986). An increase in fishing was observed in Nubia and Upper Egypt toward the end of the Upper Paleolithic (23,000 B.P.) (Gautier & Van Neer, 1989; Wendorf & Schild, 1989). My research focuses on the study of fish exploitation by an early Epi-paleolithic fishing community (23,000 cal B.P.) from the waterlogged site of Ohalo-II. By analyzing culturally and naturally fish remains assemblages, I demonstrated that the assumption that fish remains in archaeological sites result exclusively from cultural deposition is false (see Chapter 8). I also demonstrated that a general model of natural accumulation could not be developed, and that each site has to be evaluated according to its lacustrine deposits. In this chapter, I discuss the importance of the fish remains recovered at Ohalo-II for identifying environmental changes. Moreover, from the culturally deposited fish I reconstruct fishing and fish utilization methods applied 23,000 years ago. 9.1. Environmental setting Fish serve as good indicators of environmental changes since they are sensitive to fluctuations and changes in their habitat (Banarescu, 1990; Wheeler & Jones, 1976). Such changes (i.e., salinity, temperature, water level, etc.) can affect fish growth, breeding rate, and survival. In the Sea of Galilee, the history of its fauna has been affected by tectonic, climatic, salinity, and environmental changes. The last glacial maximum (LGM) was characterized by a dry cold climate, and the water level of the Paleo-Sea of Galilee was low (Horowitz, 1979; Hurwitz et al., 2000). The fish recovered at Ohalo-II provide crucial information regarding the salinity level of the Paleo-Sea of Galilee at its early formation and the evolution of endemic species. Two families of freshwater fish were identified at Ohalo-II: Cichlidae and Cyprinidae. While Cichlidae are secondary freshwater fish, Cyprinidae are primary, and therefore are intolerant to salinity changes (Banarescu, 1990; Banarescu & Coad, 1991). The taxonomic

117

composition of species identified at Ohalo-II demonstrates that most of the fish could have lived in the riverine system, except for two: Acanthobrama terraesanctae (Kinneret Bleak) and Tristamella simonis/ sacra (See Table 1 in Appendix II). These species are endemic to the Sea of Galilee and are adapted to pelagic habitats. Tristamella sp. is a secondary freshwater fish, while A. terraesanctae is a primary freshwater fish. The appearance of A. terraesanctae at Ohalo-II shows that the salinity level did not undergo any abrupt changes in the northern basin of the Jordan rift valley despite the geological and climatic changes. It also indicates that the former Lake Lisan was, in its northern boundaries, a freshwater lake with a salinity level similar to the present one. The similarity between Ohalo-II and the present day fish population at the Sea of Galilee is not restricted to taxonomic composition, but also observed in body size of A. terraesanctae (estimated standard length: 65-127 mm) (Ostrovsky & Walline, 1999). As a pelagic fish A. terraesanctae exploits the littoral zone exclusively for breeding, during winter (November-March; Appendix II). Its breeding activity is related to the rise in sea water level and exploitation of recently inundated rocky shores (Gafny et al., 1992; Gasith et al., 1996). Its occurrence at Ohalo-II attest to water level rise and breeding activity along the southern shore of the Paleo-Sea of Galilee, as at present. At Ohalo-II, Clarias sp. remains are absent. This is surprising since such remains are abundant in earlier lacustrine sites such as Erk-el-Ahmar, Ubeidiya, and Gesher Benot Ya'akov (see Appendix I). This may be explained by a preservation bias, as observed in the natural accumulation where Clarias sp. remains are absent from brown and clay layers. Another option is that environmental changes in this area during the LGM created a non- favorable habitat to this species. Since Clarias sp. is an Afrotropical fish, low water temperature during the LGM could have effected its breeding and population size. However, this explanation may be incorrect in view of the high tolerance of Clarias sp. to environmental changes (Krupp, 1987). Moreover, if water temperature was lower then than today, we would expect to observe changes in Cichlidae and Cyprinidae populations. Studies have demonstrated that reduced water temperature (lower by 2-3oC) can affect A. terraesancate hatching and growth (Ostrovsky & Walline, 1999). The abundance of A. terraesancate at locus 1 and the appearance of other taxa of Cyprinidae and Cichlidae do not support this latter hypothesis. Another option that will be discussed later is that the absence of Clarias sp. results from the fishing areas, fishing tools, and dietary preference of Ohalo-II inhabitants. 9.2. Fish exploitation

118

Recent studies (Gautier & Van Neer, 1989; Rick & Erlandson, 2000; Rick et al., 2001; Stewart, 1989; Van Neer, 1986; Wendorf & Schild, 1989; Yesner, 1980) have demonstrated that fish exploitation started earlier than previously assumed, and that nearshore fishing was a fundamental and optimal strategy, used by coastal populations not just to survive, but to flourish. They also showed that, except for deep-sea and sea-mammal fishing, there is no need to call upon increased technological efficiency to explain intensification of maritime exploitation (Rick & Erlandson, 2000; Yesner, 1980). The importance of fish and fishing as a diet component during the last glacial maximum (LGM), and evidence for long-term preservation is examined from the fish remains recovered at Ohalo-II. Ohalo-II offers, at present, the earliest appearance of fisher-hunter-gatherers during the Epi-paleolithic, which was a period of tremendous cultural and economic changes (Bar-Yosef, 1975; Bar-Yosef, 1980a; Bar-Yosef, 1980b; Bar-Yosef, 1981; Nadel & Hershkovitz, 1991; Stiner et al., 1999). Fish remains identified at Ohalo-II demonstrate that the inhabitants systematically exploited Barbus sp., Capoeta sp., and various species of Cichlidae, regardless of their body size (14-54 cm length) (the possibility that A. terraesanctae was exploited is discussed later). This pattern differs from that observed in Lake Turkana and in the Nile, where Clarias sp., and Lates sp. were selectively procured according to their ease of capture and high meat weight (Gautier & Van Neer, 1989; Stewart, 1989). Different habitats could have been exploited to capture the diverse fish identified at Ohalo-II. Barbus sp. and Capoeta sp. could have been captured either in the shallow rivers, during their breeding seasons, or from the pelagic area of the Sea of Galilee. Cichlidae could have been captured from the Sea of Galilee littoral zone (See Table 2 in Appendix II). Fishing during the breeding season might have taken place either for a long period from November until August, or for a short period during the end of winter and beginning of spring (See Table 2 in Appendix II). B. canis reproduces during the months of April-August, while B. longiceps and C. damascina breed during the winter months of December- February. Since these taxa appear together at each structure, this may suggest that they were captured during the same period, possibly the result of an overlap between breeding seasons of the different taxa (February-April). However, from the present data, I do not have enough evidence to support either season of fishing or area of capture. The absence of Clarias sp. (catfish) from Ohalo-II may also attest to specific fishing areas. In Africa, the appearance of Clarias sp. has been attributed to high water level and strong seasonal flooding, which created a favorable habitat, especially during the breeding season (Stewart, 1989). At present Clarias sp. breed from April till August in shallow water

119

in the northern parts of the Sea of Galilee (See Table 2 Appendix II). If the inhabitants of Ohalo-II had exploited the fish during their breeding season, then they could have not captured Clarias sp. which at present breed along the northern shores of the lake. Another possibility is that Ohalo-II inhabitants had some cultural taboo's forbidding consumption of Clarias sp. However, at present, the earliest evidence for fish taboos appear in Egypt in ca. 2,500 B.C., motivated by religious believes (Kreuzer, 1984). The taxonomic composition and wide range of body sizes of the fish imply that various fishing techniques were applied. They may also indicate that the fishing methods used were most probably adapted to fish movement and ecology in the littoral zones and river mouth. In Wadi Kubbaniya (20,000 B.P.), a similar trend was observed for cichlid and cyprinid exploitation, and was taken to reflect fish gathering in shallow pools after a flood episode (Wendorf & Schild, 1989). The absence of Clarias sp. from Ohalo-II may also indicate that fish were not targeted by spears or harpoons. Fishing methods may have included weirs, baskets, traps, and nets, which are perishable and therefore do not survive at archaeological sites (Stewart, 1982; von Brandt, 1972). Small pieces of burnt string found on the floor of locus 1 (Nadel et al., 1994), and the recovery of six double-notched pebbles in-situ at Ohalo-II could have been used in any of these methods (Nadel & Zaidner, 2002). Although there is no clear evidence for seafaring and deep sea fishing, the habitation structures and flint industry recovered at Ohalo-II exhibit technological skills that could have been used for such activity. Did Ohalo-II inhabitants targeted Acanthobrama terraesancate? In chapter 8, I demonstrated the similarity between A. terraesanctae remains from loci 1 and 7 and those in the adjacent natural accumulation. Due to several dissimilarities between the two accumulations (adult population at locus 1 vs. juvenile population in the natural accumulation; presence of Weberian apparatus exclusively in locus 1; higher rate of bone fragmentation in locus 1), I further examined the slight possibility that all the A. terraesanctae resulted from human activity. If A. terraesanctae were economically exploited than I would have expect to see a shift from large (high-ranked) to small (low-ranked) fish due to technological changes. Such technological changes would include mass harvesting of small fish with nets. Consequently, the economic rank of many small fish (A. terraesanctae) would be elevated relative to larger fish (Barbus sp., Capoeta sp., and Cichlidae). This explanation is unlikely for several reasons. First, it is not consistent with the increased use of large fish in other loci at Ohalo-II. Second, if small fish taken with nets were targeted because they provided greater energetic return compared to larger fish, then the data demonstrate the opposite (see table 36). The estimated contribution of 214 A. terraesanctae

120

recovered from locus 1 is 15 kg, while a small preliminary sample of 44 large cyprinids provided an estimated mean of 22 kg. Third, if technological changes appeared and small fish provided greater energetic return, why did such fishing not continue in later periods (See appendix I)? Fourth, although notched pebbles were recovered at Ohalo-II, they appear in small numbers (6) and therefore do not allow tracking of the fishing methods used. Overall, there is no distinct evidence for innovation in fishing technologies and mass harvesting of A. terraesanctae. 9.3. Fish utilization The development of fish preservation methods, and the potential of storage and long- range trade, is highly significant to the interpretation of socioeconomic patterns. However, identification of human activity is a highly subjective exercise, since the researcher reconstructs several processes from an already taphonomically biased assemblage of faunal remains. In the case of fish processing this includes: 1) immediate butchering after procurement, 2) preparation for immediate or long-term consumption (disarticulation, preservation, storage, and cooking) , 3) consumption, and 4) waste deposition area of parts removed during processing or of consumed fish (Zohar et al., 2001). In addition, recent ethnoarchaeological studies have demonstrated that there is variability in the degree to which these activities are segregated at the site, and waste is seldom deposited where activity took place (O'connell & Tunnicliffe, 2001; Simms, 1988). Fish remains recovered from different locations at Ohalo-II (loci 1,3; brush-huts,7; ashes, and 8; unidentified pit) present the opportunity to examine variability resulting from different functions. Identification of fish processing techniques requires a model of indicative criteria derived from ethnographically documented practices coupled with an understanding of other natural and cultural taphonomic agents (see chapters 2 and 7) (Barrett, 1997; Barrett et al., 1999; Hoffman et al., 2000; Zohar & Cooke, 1997; Zohar et al., 2001). At Ohalo-II interpretation of butchery and consumption patterns included the following characteristics: species diversity, population body size structure, skeletal part frequencies, and bone state of preservation. Species diversity and burning signs: At Ohalo-II the main taxa exploited in all loci were Barbus sp., Capoeta sp., Tilapia sp., and Tristamella sp. (see table 9). Loci 1 and 7 differed from other loci in the preponderance of small cyprinids and Acanthobrama sp., which most probably resulted from natural accumulation (see chapter 8). Species richness (without Acanthobrama sp. and small cyprinids) varied between loci, exhibiting highest values for locus 7. This result is not surprising since at locus 7 fireplaces were located and fish remains

121

may have resulted either from consumption after cooking, or as accumulated waste. This accords with ethnographic studies demonstrating that bone scatter frequency is higher in the hearth refuse dumps than in other areas. They also show that unburned kitchen debris was often disposed in the hearth refuse areas even though it was not thrown in the fire (Belcher, 1998; Simms, 1988). This may explain the paucity of burning signs on fish remains from Ohalo II (1-12%; see table 62) and suggests that they most probably resulted from post depositional burning. Skeletal completeness: Ethnographic studies have demonstrated the relationship between fish processing methods and skeletal completeness (see chapters 2.3, 7). Remains of fish stored for future consumption are expected to include all skeletal elements, expect those removed during processing. Remains of consumed fish are expected to be dominated by postcranial bones and to exhibit absence of cranial bones (Barrett, 1997; Barrett et al., 1999; Belcher, 1998; Bullock, 1994; Cerón-Carrasco, 1994; Gifford-Gonzales et al., 1999; Hoffman et al., 2000; Stewart, 1982; Stewart, 1991; Zohar & Cooke, 1997). For example, extreme over-representation of small fish vertebral remains recovered in an archaeological site in Portugal was reconstructed as consumed fish waste (Bicho et al., 2000). A similar pattern was observed in Lake Turkana where small- and medium- sized fish were over-represented by vertebral elements and under-represented by cranial elements (Stewart & Gifford- Gonzales, 1994). Overall however, studies have demonstrated that cooking methods such as

122

boiling will destroy the majority of fish bones (Lubinski, 1996; Nicholson, 1996a; Nicholson, 1996b). At Ohalo-II, the ratios between cranial and postcranial regions vary between loci (Table 62). At locus 1 (brush-hut) the cranial region is under-represented for all large Cyprinidae and Cichlidae. This result may imply on fish processing and consumption methods, or on a preservation bias. In locus 7 (hearths) large cyprinid cranial regions appear in relatively high frequencies (40% respectively) while cichlid cranial remains are almost absent (5%). These differences may be attributed to bone differential preservation due to preparation methods or under- presentation of small cichlid. Studies have demonstrated that in many cases cranial bones of small- to medium- sized fish are absent due to cooking and consumption methods (Belcher, 1998; Bicho et al., 2000; Butler, 1996; Stewart & Gifford-Gonzales, 1994). For example, Butler (1996) demonstrated that the inhabitants of Stillwater processed small fish (usually dried) in various methods till their bones became soft and they could be eaten whole. Belcher ( 1998) showed that in Pakistan the heads of small fish are masticated while the sharp vertebrae are removed and discarded into the trash. Overall, this latter finding reflects with the model that the fish bones were deposited in a waste area following their consumption. In locus 3 (brush hut) the postcranial region is over-represented (>88%) for all taxa and the cranial region is almost absent. Due to small sample size (see Appendix X; large cyprinids NISP=332, cichlids NISP=272) I can not regard this result as a trend for fish utilization at locus 3, but rather attribute it to sampling bias. Locus 8 (unidentified pit) differed from all other loci at Ohalo-II in the relatively high abundance of both Cyprinidae and Cichlidae cranial remains (see Table 23). Moreover, the relative abundance of cichlid otoliths in this locus is also relatively high. This abundance of cranial remains suggests that the fish in locus 8 were treated differently to those in the other loci. Since only a small sample of fish remains was analyzed from this structure, future analysis will show if this structure presents evidence for fish storage. In sum, fish remains recovered from loci 1,3,7, and 8 exhibit differences due to preparation and consumption methods. The remains recovered from locus 7 agree with that expected from a waste area. Large cyprinids and cichlids from locus 1 may also agree with the model expected from consumed fish. Fish remains from locus 3 may be biased due to small sample size and therefore can not imply any processing or consumption methods. Fish remains recovered from locus 8 may indicate different preparation and consumption methods and may also present evidence of fish storage. Identification of fish storage is important

123

since it is viewed by many researchers as a major cultural and economic breakthrough, which had profound effects on the organization of settlement and subsistence patterns. However, sample size of fish in locus 8 is small and must be enlarged before any conclusions can be drawn. 9.4. Fish exploitation in the context of Epi-paleolithic broad-spectrum economy Binford (1968) and Flannery (1969) were the first to recognize links between cultural change and expanding diet of Late Pleistocene foragers in Eurasia. Binford described dietary shift at the end of Paleolithic, accompanied by rapid diversification in hunting, food processing and food storage equipment (Binford, 1968). Flannery developed this theory with his "Broad Spectrum Revolution" hypothesis suggesting that the emergence of the Neolithic in western Asia was prefaced by local increases in dietary breadth (mainly through adding new species) in foraging societies of the late Epi-paleolithic (Flannery, 1969). Evidence of increasing dietary breadth is expected to take the form of more species in the diet, decrease of high-ranked prey and increase of low-ranked prey (Grayson & Delpech, 1998; Stiner & Munro, 2002). Recent archaeofaunal studies have demonstrated that during the Epi-paleolithic an increase in exploitation of low ranked animals (littoral shellfish, tortoise, partridges, and hares) took place (Stiner et al., 2000). However, the contribution of fish to the Epi-paleolithic economy has not been studied. At Ohalo-II, in addition to fish the inhabitants exploited birds, mountain gazelle, Dama mesopotamica, and a wide variety of plants, especially wild cereal (Nadel et al., 1994; Nadel et al., 2002; Rabinovitch, 1998). Ungulates body part representation indicate that they were butchered elsewhere and only selected parts were transported to the site (Nadel et al., 2002; Rabinovitch, 1998). Fish exploitation thus appears to have provided economic stability for Ohalo-II inhabitants, as evident from their high abundance. It could also enabled them to develop a sedentary way of life, similar to the pattern observed for fisher-hunter-gatherers in Africa (Stewart, 1989). 9.5. Summary and Conclusions The study of fish remains recovered from archaeological sites is important for understanding the role of aquatic resources in human subsistence economy. In this study, I examined 44,000 remains of fish recovered from the Ohalo-II site (23,000 B.P.), a natural accumulation along the southern shore of the Sea of Galilee, and 147 butchered fish processed for long-term preservation. My research demonstrates that in lacustrine sites, such as Ohalo-II, we can not assume, a priori, that all fish remains have resulted from human activity. Moreover a model of fish natural accumulation must be developed for each

124

depositional area. My research focused on developing criteria to distinguish fish natural from cultural accumulation, along the Sea of Galilee. Comparative analysis indicates that taxonomic breadth, richness and diversity, species representation, fish index, and skeletal completeness vary between natural and cultural accumulations. Cultural deposits are characterized by wider taxonomic breadth, higher species representation of Barbus sp., Capoeta sp. Tilapia sp., and Tristamella sp., absence of catfish remains (Clarias gariepinus), over-representation of post-cranial remains, and burning signs that appear in low frequency. Bone high scatter frequency (BSF), clumped distribution, and fragmentation did not vary between the two accumulations. My analysis has shown that Barbus sp., Capoeta sp. Tilapia sp., and Tristamella sp. remains at Ohalo-II resulted from human activity. The wide range of body sizes indicates that various fishing techniques such as weirs, baskets, and nets were probably applied. Such activity could have taken place in the riverine and littoral zones during the fish breeding season, or in the pelagic zone. However, the present data are insufficient to support determination of seasonality, fishing area and technology. The remains of Acanthobrama terraesanctae (Kinneret bleak) and small cyprinids in high frequencies in loci 1 and 7 show high resemblance to those of the adjacent natural accumulation, and attest that fish remains in Ohalo-II did not result exclusively from human activity. This conclusion requires one caveat. If the inhabitants of Ohalo-II did indeed targeted A. terraesanctae, then this is the earliest evidence for small fish mass harvesting, which would have required the use of new technologies. Evidence of such technologies was not recovered at the site, and overall, this economic trend differs from other loci examined, and was not found in later sites. Therefore, the present data do not provide sufficient evidence to support mass exploitation of A. terraesanctae by Ohalo-II inhabitants, but rather natural accumulation. The absence of Clarias sp. (catfish) remains from Ohalo-II is interesting since in Africa and Egypt this fish has been targeted due to its large size and ease of capture during its breeding season. At present Clarias sp. breed in the northern parts of the Sea of Galilee, and if a similar breeding pattern occurred 23,000 years ago, it could have not been exploited easily in the vicinity of Ohalo-II. My analysis has shown differences in completeness of skeletons recovered from various structures due to processing and consumption methods. Element representation found at locus 7 accords with that expected from a waste area of consumed fish. The burning signs, which appear in low frequencies (12%), probably result from post depositional fire. A

125

future study will examine evidence of heated bones due to cooking. Skeletal completeness in locus 8 indicates that the fish were deposited whole. The abundance of cranial elements, particularly cichlid otoliths, undoubtedly reflects cultural processing techniques. Since fish storage is viewed by many researchers as a major evolutionary breakthrough which has profound effects on the organization of settlements and subsistence patterns (Barrett, 1997; Barrett et al., 1999; Butler, 1990) this structure will be further investigated in the future. Given the geological and climates changes in the Sea of Galilee, especially in regard to its evolution from Lake Lisan, the taxa identified at Ohalo-II constitute an excellent marker of environmental conditions. I assumed that if water salinity level changed during Lake Lisan high stand (ca. 23,000 years ago), it would have affected the relative abundance of preliminary freshwater fish, especially endemic species such as Acanthobrama terraesanctae. The taxonomic groups identified at Ohalo-II are identical with the present day fish in the Sea of Galilee. Moreover, A. terraesanctae and Tristamella sp., which are endemic to the Sea of Galilee, were present, indicating that the salinity level did not undergo any abrupt changes in the Paleo-Sea of Galilee/ Lake Lisan. It also suggests that the former Lake Lisan was, in its northern boundaries, a freshwater lake with a salinity level similar to the present level. Another aspect of my research examined skeletal element representation and fragmentation in comparison with butchering methods by present day fishermen. An ethnographic study demonstrated that in modern-day Panama fish are butchered differently according to their body size. Skeletal element presentation and fragmentation patterns from Panama differ from the data obtained from Ohalo-II. This may be due to different butchering methods or due to preservation bias. Overall, the large numbers of fish remains recovered from all loci at Ohalo-II, indicate that fishing activity played an important role in the inhabitants' daily life and diet. In absent of direct evidence for deep sea fishing, nearshore fishing was probably a fundamental and optimal strategy used by the inhabitants, providing a stable subsistence economy. Fish remains have the potential to reflect issues of prehistoric cultural processing patterns and food storage. However, determination of food storage at Ohalo-II requires future enlargement of sample size.

126

BIBLIOGRAPHY Andrews P. (1995). Experiments in Taphonomy. Journal of Archaeological Science 22, 147- 153. Banarescu P. (1990a). Zoogeography of Fresh Waters. In General Distribution and Dispersal of Freshwater Animals. Wiesbaden: AULA-Varlag GmbH, pp. 1-519. Banarescu P. (1990b). Zoogeography of Fresh Waters. In Distribution and Dispersal of Freshwater Animals in North America and Eurasia. Wiesbaden: AULA-Varlag GmbH, pp. 519-1091. Banarescu P. (1990c). Zoogeography of Fresh Waters. In Distribution and Dispersal of Freshwater Animals in Africa, Pacific Areas and South America. Wiesbaden: AULA-Varlag GmbH, pp. 1099-1617. Banarescu P. & Coad B. W. (1991). Cyprinids of Eurasia. In (I. J. Winfield & J. S. Nelson, Eds.) Cyprinid Fishes: Systematic, Ecology and Exploitation. London: Chapman & Hall, pp. 127-155. Bar-Oz G., Dayan T. & Kaufman D. (1999). The Epipalaeolithic faunal sequence in Israel: A view from Neve-David. Journal of Archaeological Science 26, 67-82. Bar-Yosef O. (1975). The Epi-Paleolithic in Palestine and Sinai. In (F. Wendorf & A. E. Marks, Eds.) Problems in Prehistory: North Africa and the Levant. Dallas, pp. 363- 378. Bar-Yosef O. (1980a). The Mediterranean Levantine Epi-Paleolithic as the background to the "Neolithic revolution". In (P. Sorensen & P. Mortensen, Eds.) Origins of Agriculture and Technology: West or East Asia? Aarhus: University of Aarhus. Bar-Yosef O. (1980b). Prehistory of the Levant. Ann. Rev. Anthropol. 9, 101-133. Bar-Yosef O. (1981). The Epi-Paleolithic complexes in the Southern Levant. In (J. Cauvin & P. Sanlaville, Eds.) Prehistoire du Levant. Paris: Editions du CNRS, pp. 389-408. Bar-Yosef O. (1990). The Last Glacial Maximum in the Mediterranean Levant. In (C. Gamble & O. Soffer, Eds.) The World at 18,000 B.C., Vol.2, The Low Latitudes. London: Unwin Hyman, pp. 58-77. Bar-Yosef O. (1991). The Early Neolithic of the Levant: Recent advances. The Review of Archaeology 12, 1-18. Bar-Yosef O. & Belfer-Cohen A. (1992). From Foraging to Farming in the Mediterranean Levant. In (A. B. Gebauer & T. D. Price, Eds.) Transition to Agriculture in Prehistory: Prehistory Press, pp. 21-48.

127

Bar-Yosef O., Gopher A., Tchernov E. & Kislev M. (1991). Netiv Hagdud: An Early Neolithic village site in the Jordan valley. Journal of Field Archaeology. 18, 405- 424. Barrett J. H. (1997). Fish trade in North Orkney and Caithness: a zooarchaeological approach. Antiquity 71, 616-38. Barrett J. H., Nicholson R. A. & Cerón-Carrasco R. (1999). Archaeo-Ichtyological evidence for long-term socioeconomic trends in northern Scotland: 3500 BC to AD1500. Journal of Archaeological Science 26, 353-388. Bartov Y., Stein M., Enzel Y., Agnon A. & Reches Z. (2002). Lake levels and sequence stratigraphy of Lake Lisan, the late Pleistocene precursors of the Dead Sea. Quaternary Research 57, 9-21. Baxter M. J. (2001). Methodological issues in the study of assemblage diversity. American Antiquity 66, 715-725. Begin Z. B., Ehrlich A. & Nathan Y. (1974). Lake Lisan, The Pleistocene precursor of the Dead Sea. Geol. Surv. Isr. Bull. 3, 30p. Behrensmeyer A., Gordon K. D. & Yanagi G. T. (1986). Trampling as a cause of bone surface damage and pseudo cut marks. Nature 319, 768-771. Behrensmeyer A. K. (1991). Terrestrial vertebrate accumulation. In (P. A. Allison & D. E. Briggs, Eds.) Taphonomy: Releasing the Data Locked in the Fossil Record. New York: Plenum Press, pp. 291-337. Behrensmeyer A. K., Gordon K. D. & Yanagi G. T. (1989). Nonhuman bone modification in Miocene fossils from Pakistan. In (R. Bonnichsen & M. H. Sorg, Eds.) Bone Modification. Orono, Maine: Center for The Study of The First Americans, Institute of Quaternary Studies, University of Maine., pp. 99-120. Belcher W. (1993). Fishing strategies and butchery practices: Ethnoarchaeology of South Asian fisherfolk. Fish Remains Working Group, 7th Meeting. Belcher W. R. (1994). Butchery practices and the ethnoarchaeology of South Asian fisherfolk. In (W. Van-Neer, Ed.) Fish Exploitation in The Past: Proceedings of the 7th meeting of the ICAZ Fish Remains Working Group. Teruven: Annales du Museé Royal de l'Afrique Centralé, pp. 169-176. Belcher W. R. (1998). Fish exploitation of the Baluchistan and Indus valley traditions: An ethnoarchaeological approach to the study of fish remains. Ph.D. Thesis. University of Wisconsin-Madison.

128

Ben-Tuvia A. (1978). Fishes. In (C. Serruya, Ed.) Monographie Biologica: Lake Kinneret. London: Dr. W. Junk Publishers The Hague, pp. 407-430. Bennett J. L. (1999). Thermal alteration of buried bone. Journal of Archaeological Science 26, 1-8. Berra T. M. (2001). Freshwater Fish Distribution. San Diego: Academic Press. Bicho N. F., Hockett B., Haws J. & Belcher W. (2000). Hunter-gatherer subsistence at the end of the Pleistocene: Preliminary results from Picareiro Cave, Central Portugal. Antiquity 74, 500-506. Binford L. R. (1968). Post-Pleistocene adaptations. In (S. R. Binford & L. R. Binford, Eds.) New Perspectives in Archaeology. Chicago: Aldine Publishing, pp. 313-341. Binford L. R. (1978). Nunamit Ethnoarchaeology. New York: Academic Press. Binford L. R. (1981). Bones, Ancient Man and Modern Myths. New-York: Studies in Archaeology, Academic-Press. Binford L. R. & Bertram J. B. (1977). Bone frequencies and attritional processes. In (L. R. Binford, Ed.) For Theory Building in Archaeology. New York: Academic Press, pp. 77-153. Bonnichsen R. & Sorg H. M. Eds. (1989). Bone Modification. Orono, Maine: Center for The Study of The First Americans, Institute for Quaternary Studies, University of Maine. Borg I. Ed. (1981). Multidimensional Data Representations: When and Why? Ann Arbor: Mathesis Press. Brewer D., J, (1991). Fishing in Prehistoric Egypt: Inferences from faunal remains. In (J. R. Purdue, W. E. Klippel & B. W. Styles, Eds.) Beamers, Bobwhites, and Blue-Points. Springfield: Illinois State Museum Scientific Papers, Vol. XXIII, and the University of Tenessee, Department of Anthropology Report of Investigations No. 52, pp. 333- 340. Bullock A. E. (1994). Evidence for fish exploitation from Tudor deposits excavated at little pickle, Surrey England, in 1989. 1994. Offa 51, 267-274. Burgess G. H. O. e. a. Ed. (1965). Fish Handling and Processing HMSO. Burt J. R. (1988). Dried and smoked fishery products: Preparation and consumption. In (J. R. Burt, Ed.) Fish Smoking and Drying. London: Elsevier Applied Science, pp. 121- 160. Butler V. L. (1987). Distinguishing natural from cultural Salmonid deposits in the Pacific Northwest of North America. In (D. T. Nash & M. D. Petraglia, Eds.) Natural

129

Formation Process and the Archaeological Record. Oxford: B.A.R. 352, pp. 131- 149. Butler V. L. (1990). Distinguishing natural from cultural Salmonid deposits in the Pacific Northwest of North America. Ph.D. Thesis. Thesis. University of Washington. Butler V. L. (1993). Natural versus cultural Salmonid remains: origin of the Dalles Roadcut bones, Columbia River, Oregon, USA. Journal of Archaeological Science 20, 1-24. Butler V. L. (1994). The role of bone density in structuring prehistoric Salmon bone assemblages. Journal of Archaeological Science 21, 413-424. Butler V. L. (1996). Tui Chub taphonomy and the importance of marsh resources in the western great basin of North America. American Antiquity 61, 699-717. Butler V. L. (2000). Resource depression on the Northwest coast of North America. Antiquity 74, 649+661. Butler V. L. & Schroeder R. (1998). Do digestive processes leave diagnostic traces on fish bones? Journal of Archaeological Science 1985, 957-971. Cannon M. D. (2001). Archaeofaunal relative abundance, sample size, and statistical methods. Journal of Archaeological Science 28, 185-195. Casteel R. W. (1972). Some biases in the recovery of archaeological faunal remains. Proceeding of the Prehistoric Society. 38, 382-388. Cerón-Carrasco R. (1994). The investigation of fish remains from an Orkney farm mound. In (W. Van Neer, Ed.) Fish Exploitation in The Past: Proceedings of the 7th meeting of the ICAZ Fish Remains Working Group. Teruven: Annales du Musée Royal de l'Afrique Centrale, Sciences Zoologiques no 274., pp. 207-210. Chase P. G., McPherron S. P. & Tuffreau A. (1997). Testing the reality of a "living floor" with archaeological data. American Antiquity 62, 629-652. Clark J. D. (1972). Paleolithic butchery practices. In (P. J. Ucko, R. Tringham & G. W. Dimbleby, Eds.) Man, Settlement and Urbanism. Cambridge: Duckworth, pp. 149- 456. Coard R. & Dennell R. W. (1995). Taphonomy of some articulated skeletal remains: Transport potential in an artificial environment. Journal of Archaeological Science 22, 441-448. Colten R. (1995). Faunal exploitation during the middle to late period transition on Santa Cruz Island, California. Journal of California and Great Basin Anthropology 17, 93- 120.

130

Cutler A. H., Behrensmeyer A. K. & Chapman R. E. (1999). Environmental information in a recent bone assemblage: Roles of taphonomic processes and ecological change. Paleogeography, Paleoclimatology, Paleoecology 149, 359-372. Cutting C. L. (1955). Fish Saving: A history of fish processing from ancient to modern times. London: Leonard Hill Limited. Davis S. (1974). Animal remains from the Kebaran site of Ein Gev I, Jordan Valley, Israel. Paléorient 2, 453-462. Davis S. J. M. (1985). A preliminary report of fauna from Hatula: a Natufian-Khiamian (PPNA) site near Latroun, Israel. In (M. Lechevallier & A. Ronen, Eds.) Le site Natufian-Khiamien de Hatoula, pres de Latroun, Israel. Jerusalem: Centré de Rechérché Français de Jerusalem, pp. 71-118. Davis S. J. M. & Payne S. (1992). 101 Ways to deal with a dead hedgehog: notes on the preparation of disarticulated skeletons for zoo-archaeological use. Circaea 8, 1-6. Desse J. (1987). Mallaha: L'ichtyofauna. In (J. Bouchud, Ed.) La Fauna du Gisement Natoufien de Mallaha (Eynan) Israel. Paris: Association Paléorient, pp. 151-156. Efremov I. A. (1940). Taphonomy: a new branch of paleontology. Pan-American Geologist 74, 81-93. Elder R. L. & Smith G. R. (1988). Fish taphonomy and environmental inference in paleolimnology. Palaeogeography, Palaeoclimatology, Palaeoecology 62, 577-592. Essuman K. M. & Diakite B. (1990). L'utilisation du Balistes (Balistes carolinensis) au Ghana et au Sengal. FAO Fisheries Circular 828, 1-23. Falabella F., Vargas L. M. & Melendez R. (1994). Differential preservation and recovery of fish remains in Central Chile. In (W. Van-Neer, Ed.) Fish Exploitation in The Past: Proceedings of the 7th meeting of the ICAZ Fish Remains Working Group. Tervuren: Proceedings of the 7th meeting of ICAZ Fish Remains Working Group. Annales du Musée Royal de l'Afrique Centrale, Sciences Zoologiques no. 274., pp. 25-35. Ferber C. T. & Wells N. A. (1995). Paleolimnology and taphonomy of some fish deposits in "Fossil" and "Uinta" lakes of the Eocene Green River Formation, Utah and Wyoming. Paleogeography, Paleoclimatology, Paleoecology 117, 185-210. Fiorillo A. R. (1989). An experimental study of trampling: Implications for the fossil record. In (R. Bonnichsen & M. H. Sorg, Eds.) Bone Modification. Orono, Maine: Center for The Study of The First Americans, Institute of Quaternary Studies, University of Maine., pp. 61-72.

131

Firth R. (1975). Malay Fishermen. Their Peasant Economy. New York: The Norton Library. Fitch J. E. (1967). Fish remains recovered from a Corona del Mar, California, Indian Midden (Ora-190). California Fish and Game 53, 185-191. Flannery K. V. (1969). Origins and ecological effects of early domestication in Iran and the Near East. In (J. Ucko & G. W. Dimbleby, Eds.) The Domestication and Exploitation of Plants and Animals. London: Gerald Duckworth & Co., pp. 73-100. Forbes R. J. (1955). Studies in Ancient Technology. Leiden: E.J. Brill. Fred C., Andrus T. & Crowe D. E. (2002). Alteration of otolith aragonite: Effects of prehistoric cooking methods on otoliths chemistry. Journal of Archaeological Science 29, 291-299. Gafny S., Gasith A. & Goren M. (1992). Effect of water level fluctuation on shore spawning of Mirogrex terraesanctae (Steinitz), (Cyprinidae) in Lake Kinneret, Israel. Journal of Fish Biology 41, 863-871. Galili E. (1985). Clay exposures and archaeological finds on the sea bottom between Haifa and Atlit. M.A. Thesis. Thesis. Haifa University (in Hebrew). Galili E. & Weinstein-Evron M. (1985). Prehistory and Paleoenviroments of submerge sites along the Carmel coast of Israel. Paleorient 11, 37-51. Galili E., Weinstein-Evron M., Hershkovitz I., Gopher A., Kislev M., Lernau O., Horwitz L. K. & Lernau H. (1993). Atlit-Yam: A Prehistoric site on the sea floor off the Israeli coast. Journal of Field Archaeology 20, 133-157. Galili E., Weinstein-Evron M. & Ronen A. (1988). Holocene Sea-level changes based on submerged Archaeological sites off the Northern Carmel coast in Israel. Quaternary Research 29, 36-42. Gasith A., Goren M. & Gafny S. (1996). Ecological consequences of lowering Lake Kinneret water level: effect on breeding success of the "Kinneret Sardine". In Preservation of our world in the wake of change. Jerusalem: ISEEQS Pub., pp. 569-573. Gautier A. & Van Neer W. (1989). Animal remains from the Late Paleolithic sequence of Wadi Kubbaniya. In (A. E. Close, Ed.) The Prehistory of North Africa, vol. 2: Paleoeconomy, Environment and Stratigraphy. Dallas: Southern Methodist University Press., pp. 119-158. Gifford-Gonzales D. (1989). Ethnographic analogues for interpreting modified bones: Some cases from East Africa. In (R. Bonnichsen & M. H. Sorg, Eds.) Bone Modification. Orono, Maine: Center for The Study of The First Americans, Institute of Quaternary Studies, University of Maine., pp. 179-246.

132

Gifford-Gonzales D., Stewart M. K. & Rybczynski N. (1999). Human activities and site formation at modern lake margin foraging camps in Kenya. Journal of Anthropological Archaeology 18, 397-440. Gilead I. (1984). Is the term "Epipaleolithic" relevant to Levantine Prehistory? Current Anthropology 25, 227-229. Gobalet K. W. (2001). A critique of faunal analysis; Inconsistency among experts in blind tests. Journal of Archaeological Science 28, 377-386. Golani D. & Darom D. (1997). Handbook of The Fishes of Israel. Jerusalem: Keter Publishing House Ltd. Gordon E. A. (1993). Screen size and differential faunal recovery: A Hawaiian example. Journal of Field Archaeology 20, 453-460. Goren M. (1974). The freshwater fishes of Israel. Israel Journal of Zoology 23, 67-118. Goren M. (1983). Freshwater Fishes of Israel: Biology and Taxonomy. Tel Aviv: Hakibutz Hameuchad Publishing House Ltd. Goren M., Fishelson L. & Trewavas E. (1973). The Cyprinid fishes of Acanthobrama Heckel and related genera. Bulletin of The British Museum (Natural History) Zoology 24, 293-315. Goren M. & Ortal R. (1999). Biogeography, diversity and conservation of the inland water fish communities in Israel. Biological Conservation 89, 1-9. Goring-Morris A. N. (1987). At the Edge: Terminal Pleistocene Hunter-Gatherers in the Negev and Sinai. Oxford: B.A.R International Series. Gorring-Morris A. N., Goren Y., Horwitz L. K., Hershkovitz I., Lieberman R., Sarel J. & Bar- Yosef D. (1994-5). The 1992 season of excavation at the Pre-Pottery Neolithic B Settlement of Kfar Hahoresh. Journal of The Israel Prehistoric Society 26, 74-121. Grayson D. K. (1984). Quantitative Zooarchaeology: Topics in the analysis of archaeological faunas. New-York: Academic Press. Grayson D. K. & Delpech F. (1998). Changing diet breadth in the Early Upper Palaeolithic of Southwestern France. Journal of Archaeological Science 25, 1119-1129. Haas G. (1966). On the vertebrate fauna of the lower Pleistocene site Ubeidiya. In (M. Stekelis, Ed.) The Lower Pleistocene of the Central Jordan Valley the Excavations at 'Ubeidiya, 1960-1963. Jerusalem: The Israel Academy of Sciences and Humanities, pp. 68.

133

Hayden B., Chisholom B. & Schwartz H. P. (1987). Fishing and Foraging. Marine resources in the Upper Paleolithic of France. In (O. Soffer, Ed.) The Pleistocene Old World: Regional Perspectives. New-York: Plenum Press, pp. 279-291. Heck K. L., Jr., Van Belle G. & Simberloff D. (1975). Explicit calculation of the rarefaction diversity measurement and the determination of sufficient sample size. Ecology 56, 1459-1461. Hoffman B. W., Czederpiltz J. M. C. & Partlow M. A. (2000). Heads or Tails: The zooarchaeology of Aleut Salmon storage on Unimak Island, Alaska. Journal of Archaeological Science 27, 699-708. Holland S. M. (2001). Analytic Rarefraction: http://www.uga.edu/~strata/software/AnRareReadme.html. Horowitz A. (1978). The quaternary evolution of the Jordan Valley. In (C. Serruya, Ed.) Monographie Biologica: Lake Kinneret. London: Dr. W. Junk Publishers The Hague, pp. 33-44. Horowitz A. (1979). The Quaternary of Israel. New York: Academic Press. Horowitz A. (1988). The quaternary environments and paleogeography in Israel. In (Y. Yom- Tov & E. Tchernov, Eds.) The Zoogeography of Israel. Netherlands: Dr. W.Junks Publishers, Dordecht, pp. 35-57. Hurlbert S. H. (1971). The nonconcept of species diversity: a critique and alternative parameters. Ecology 52, 577-586. Hurwitz S., Stanislavsky E., Lyakhovsky V. & Gvirtzman H. (2000). Transient groundwater- lake interactions in a continental rift: Sea of Galilee, Israel. The Geological Society of America Bulletin 112, 1694-1702. Ikram S. (1995). Choice Cuts: Meat Production in Ancient Egypt. Leuven, Belgium: Peters Press and the Department of Oriental Studies. James S. R. (1997). Methodological issues concerning screen size recovery rates and their effects on archaeofaunal interpretations. Journal of Archaeological Science 24, 385- 397. Jelink A. J., Farrand W. R., Hass G., Horowitz A. & Goldberg P. New excavations at Tabun Cave, Mount Carmel, Israel, 1967-1972: A preliminary report., 151-183. Kaufman D. (1998). Measuring archaeological diversity: An application of the Jackknife technique. American Antiquity 63, 73-85. Kelly R. L. (1996). Foraging and fishing. In (M. Plew, Ed.) Prehistoric Hunter-Gatherer Fishing Strategies. Boise: Boise State University, pp. 208-214.

134

Kintigh K. W. (1984). Measuring archaeological diversity by comparison with simulated assemblages. American Antiquity 49, 44-54. Klein R. G. & Cruz-Uribe K. (1984). The Analysis of Animal Bones from Archaeological Sites. Chicago: The University of Chicago Press. Kobyliansky E. & Hershkovitz I. (1997). Biology of Desert Populations- South Sinai Bedouins: Growth and Development of Children in Human Isolates. Liége: ERAUL. Koch C. F. (1987). Prediction of sample size effects on the measured temporal and geographic distribution patterns of species. Paleobiology 13, 100-107. Koon H. E. C., Nicholson R. A. & Collins M. J. (2003). A practical approach to the identification of low temperature heated bone using TEM. Journal of Archaeological Science in press. Krebs C. J. (1989). Ecological Methodology. University of British Columbia: Harper Collins Publishers. Krebs C. J. (1999). Ecological Methodology. University of British Columbia: Harper Collins Publishers. Kreuzer R. (1984). Fish in region and myths of ancient Mesopotamia and Egypt. In (B. Gunda, Ed.) The fishing Culture of the world. Budapest: Akademiai Kiade, pp. 593- 618. Krupp F. (1987). Freshwater ichtyogeography of the Levant. Fauna and Zoogeography of the Middle East, pp. 229-237. Krupp F. & Schneider W. (1989). The fishes of the Jordan River drainage basin and Azraq Oasis. Fauna of Saudi Arabia 10, 348-415. Leach B. F. (1986). A method for the analysis of Pacific fish bone assemblages and an associated database management system. Journal of Archaeological Science 13, 147- 159. Lernau H. & Lernau O. (1994). The Fish Remains. In (M. Lechevallier & A. Ronen, Eds.) Le Gisement de Hatoula en Judée occidentale, Israel. Paris: Association Plaéorient, Memoires et Travaux du Centre de Recherche Francais de Jerusalem, pp. 111-121. Levi S. (1987). The Bedouin in Sinai Desert: A Pattern of Desert Society (in Hebrew). Tel- Aviv: Schocken Publishing House Ltd. Liem K. F. (1991). Functional morphology. In (M. H. A. Keenleyside, Ed.) Cichlid Fishes: Behaviour, Ecology and Evolution. London: Chapman & Hall, pp. 129-150.

135

Locker A. M. (2000). The role of stored fish in England 900-1750AD; the evidence from historical and archaeological data. Sofia, Bulgaria: Publishing Group Limited Office. Lubinski P. M. (1996). Fish heads, fish heads: An experiment on differential bone preservation in a Salmonid fish. Journal of Archaeological Science. 23, 175-181. Lyman R. L. Ed. (1982). Archaeofaunas and subsistence studies. New York: Academic Press. Lyman R. L. (1984). Bone density and differential survivorship of fossil classes. Journal of Anthropological Society 3, 259-299. Lyman R. L. (1987). Archaeofaunas and butchery studies: a taphonomic perspective. In (M. B. Schiffer, Ed.) Advances in Archaeological Method and Theory. Vol.10. New- York: Academic Press, pp. 249-337. Lyman R. L. (1991a). Prehistory of the Oregon Coast. San Diego: Academic Press. Lyman R. L. (1991b). Taphonomic problems with archaeological analyses of animal carcass utilization and transport. In (J. R. Purdue, W. E. Klippel & B. W. Styles, Eds.) Beamers, Bobwhites, and Blue-Points. Springfield: Illinois State Museum Scientific Papers, Vol. XXIII, and the University of Tenessee, Department of Anthropology Report of Investigations No. 52, pp. 125-138. Lyman R. L. (1994). Vertebrate Taphonomy. Cambridge: Cambridge University Press. Magurran A. E. (1988). Ecological Diversity and Its Measurement. Princeton: Princeton University Press. Mancuso A. C. (2003). Continental fish taphonomy: a case study in the Triassic of Argentina. Journal of South American Earth Sciences 16, 275-286. Manly B. F. J. (1991). Randomization and Monte Carlo Methods in Biology. New York: Chapman and Hall. Marshall F. & Pilgram T. (1991). Meat versus within-bone nutrients: Another look at the meaning of body part representation in archaeological sites. Journal of Archaeological Science 18, 149-163. Martin R. E. (1998). One Long Experiment: Scale and Process in Earth History. New York: Columbia University Press. Martin R. E. (1999). Taphonomy: A Process Approach. Cambridge: Cambridge University Press. Meehan B. (1982). Shell bed to shell midden. Canbarra: Australian Institute of Aboriginal Studies.

136

Michael S. M. (1984). The fishermen of India. In (B. Gunda, Ed.) The Fishing Culture of The World. Studies in Ethnology, cultural Ecology and Folklore. Budapest: Akademia Kiado, pp. 651-676. Montelius O. (1888). The Civilization of Sweden in Heathen Times. London: Macmillan. Morales A. & Rosenlund K. (1979). Fish Bone Measurements. Copenhagen: Steenstrupia. Myers G. S. (1938). Fresh-water fishes and West-Indian zoogeography. Smithsonian Report for 1937, 339-364. Myers G. S. (1951). Fresh-water fishes and East Indian zoogeography. Stanford Ichth. Bull. 4, 11-21. Nadel D. (1990). Ohalo II-A Preliminary Report. Mitekufat Haeven-Journal of the Israel Prehistoric Society 23, 48-59. Nadel D. (1991). Ohalo II-The third season. Mitekufat Haeven-Journal of the Israel Prehistoric Society 24, 158-163. Nadel D. (1993a). The organization of space in a fisher-hunter-gatherers camp at Ohalo II, Israel. In (M. Otte, Ed.) Nature et Culture. Liege, E.R.A.U.L: Colloque de Liege (13- 17 Decembre), pp. 371-386. Nadel D. (1993b). Submerged archaeological sites on the shores of Lake Kinneret. Atiqot XXII, 1-12. Nadel D. (1997). The Spatial Organization of Prehistoric Sites in the Jordan Valey: Kebaran, Natufian and Neolithic Case Studies. Unpublished Ph.D. Thesis. Thesis. Hebrew University, Mount Scopus. Nadel D. (1999). Scalene and Proto-Triangles from Ohalo II. Mitekufat Ha'even, Journal of the Israel Prehistoric Society 29, 5-16. Nadel D. Ed. (2002). Ohalo II: A 23,000 Year-Old Fisher-Hunter-Gatherers' Camp on the Shore of the Sea of Galilee. Haifa: Reuben and Edith Hecht Museum, University of Haifa. Nadel D., Belitzky S., Boaretto E., Carmi I., Heinemeier J., Werker E. & Marco S. (2001). New dates from late Pleistocene sediments in the Southern Sea of Galilee. In (H. J. Bruins, I. Carmi & E. Boaretto, Eds.) Near East Chronology: Archaeology and Environment. Proceedings of the 17th International 14C Conference. Radiocarbon. Arizona: University of Arizona, pp. 1167-1178. Nadel D., Danin A., Werker E., Schick T., Kislev M. E. & Stewart K. (1994). 19,000 Year old twisted fibres from Ohalo II. Current Anthropology 35, 451-458.

137

Nadel D. & Hershkovitz I. (1991). New Subsistence Data and Human remains from the earliest Levantine Epipalaeolithic. Current Anthropology 32, 631-635. Nadel D., Tsatskin A., Bar-Yosef Mayer D. E., Belmaker M., Boaretto E., Hershkovitz I., Kislev M. E., Rabinovich R., Simmons T., Weiss E., Zohar I., Asfur O., Emmer G., Ghraieb T., Grinberg U., Halabi H., Weissbrod L. & Zaidner Y. (2002). The Ohalo II 1999 – 2000 Seasons of excavation: A preliminary report. Mitekufat Ha'even, Journal of the Israel Prehistoric Society 32, 17-48. Nadel D. & Werker E. (1999). The oldest ever brush hut plant remains from Ohalo II, Jordan Valley, Israel (19,000 BP). Antiquity 73, 755-64. Nadel D. & Zaidner Y. (2002). Upper Pleistocene and Mid-Holocene net sinkers from the Sea of Galilee, Israel. Journal of The Israel Prehistoric Society 32, 49-71. Nicholas G. P. (1998). Wetlands and Hunter-Gatherers: A global perspective. Current Anthropology 39, 720-731. Nicholson R. A. (1992a). An assessment of the value of bone density measurements to archaeoichthyological studies. International Journal of Osteoarchaeology 2, 139- 154. Nicholson R. A. (1992b). Bone survival: The effects of sedimentary abrasion and trampling on fresh and cooked bone. International Journal of Osteoarchaeology 2, 79-90. Nicholson R. A. (1993). A morphological investigation of burnt animal bone and an evaluation of its utility in archaeology. Journal of Archaeological Science 20, 411- 428. Nicholson R. A. (1996a). Bone degradation, burial medium and species representation: debunking the myths, an experiment-based approach. Journal of Archaeological Science 23, 513-533. Nicholson R. A. (1996b). Fish bone diagenesis in different soils. Archaeofauna 5, 79-91. Nicholson R. A. (1998). Bone degradation in a compost heap. Journal of Archaeological Science 25, 393-403. Noe-Nygaard N. (1977). Butchering and marrow fracturing as a taphonomic factor in archaeological deposits. Palaeobiology 3, 218-237. O'connell J. M. & Tunnicliffe V. (2001). The use of sedimentary fish remains for interpretation of long-term fish population fluctuations. Marine Geology 174, 177- 195. O'Connor T. P. (1993). Process and terminology in mammal carcass reduction. International Journal of Osteoarchaeology 3, 63-67.

138

Orton C. (2000). Sampling in Archaeology. Cambridge: Cambridge University Press. Ostrovsky I. & Walline P. (1999). Growth and production of the dominant pelagic fish, Acanthobrama terraesanctae, in subtropical Lake Kinnerert, Israel. Journal of Fish Biology 54, 18-32. Otero O. (2001). The oldest-known Cyprinid fish of the Afro-Arabic plate, and its paleobiogeographical implications. Journal of Vertebrate Paleontology 21, 386-388. Perrot J. (1966). Le gisement Natoufien de Mallaha (Eynan) Israel. L'Anthropologie 70, 427- 483. Plog S. & Hegmon M. (1993). The sample size-richness relation: The relevance of research questions, sampling strategies, and behavioral variation. American Antiquity 58, 489- 496. Rabinovitch R. (1998). Patterns of animal exploitation and subsistence in Israel, during the Upper paleolithic and Epi-paleolithic (40,00-12,500 BP), based upon selected case studies. Ph.D. Thesis. Hebrew University. Rabinovitch R., Bar-Yosef O. & Tchernov E. (1996). How many ways to skin a gazelle- Butchery patterns from an Upper Palaeolithic site, Hayonim cave, Israel. Archaeozoologia VIII, 11-52. Reitz E. J. & Wing E. S. (1999). Zooarchaeology. Cambridge: Cambridge University Press. Rick T. C. & Erlandson J. M. (2000). Early Holocene fishing strategies on the California coast: Evidence from CA-SBA-2057. Journal of Archaeological Science 27, 621- 633. Rick T. C., Erlandson J. M. & Vellanoweth R. L. (2001). Paleocostal marine fishing on the Pacific coast of the Americas: Perspectives from Daisy Cave, California. American Antiquity 66, 595-613. Roberts S. J., Smith C. I., Millard A. & Collins M. J. (2002). The taphonomy of cooked bone: characterizing boiling and its physico-chemical effects. Archaeometry 44, 485-494. Robinson S., Nicholson R. A. & Pollard A. M. (2003). An evaluation of Nitrogen porosimetry as a technique for predicting taphonomic durability in animal bone. Journal of Archaeological Science 30, 391-403. Rojo A. L. (1991). Dictionary of Evolutionary Fish Osteology. London: CRC Press, Inc. Rosenthal Y., Katz A. & Tchernov E. (1989). The reconstruction of Quaternary freshwater lakes from the chemical and isotopic composition of gastropod shells: The Dead Sea rift, Israel. Palaeogeography, Palaeoclimatology, Palaeoecology 74, 241-253.

139

Schafer W. (1972). Ecology and Paleoecology of Marine Environments. Chicago: University of Chicago Press. Schiffer M. B. (1983). Toward the identification of formation processes. American Antiquity 48, 675-706. Shahack-Gross R. (2002). Taphonomy of Abandoned Maasai Settlements in Southern Kenya with Special Reference to Live Stock Enclosures: Implications for the Spread of Pastoralism in East Africa. Ph.D. Dissertation. Thesis. Washington University. Shipman P., Bosler W. & Davis K. L. (1981). Butchering of giant Geladas at an Acheulian site. Current Anthropology 22, 257-268. Shipman P., Foster G. & Schoeninger M. (1984). Burnt bones and teeth: An experimental study of color, morphology, crystal structure, and shrinkage. Journal of Archaeological Science 11, 307-325. Shipman P. & Rose J. (1984). Cutmarks mimics on modern and fossil bones. Current Anthropology 25, 116-117. Simms S. R. (1988). The archaeological structure of a Bedouin cap. Journal of Archaeological Science 15, 197-211. Sokal R. R. & Rohlf F. J. (1981). Biometry: The Principles and Practice of Statistics in Biological Research. New York: W.H. Freeman and Company. Speth J. D. (2000). Boiling vs. baking and roasting: A taphonomic approach to the recognition of cooking technique in small mammals. In (P. Rowley-Conwy, Ed.) Animal bones, Human Societies. Oxford: Oxbow Books, pp. 89-105. Stewart H. (1982). Indian Fishing: Early Methods on the Northwest Coast. Vancouver: University of Washington Press. Stewart K. M. (1989). Fishing Sites of North and East Africa in the Late Pleistocene and Holocene. Oxford: B.A.R. 521. Stewart K. M. (1991). Modern fishbone assemblages at Lake Turkana, Kenya: A methodology to aid in recognition of Hominid fish utilization. Journal of Archaeological Science 18, 579-603. Stewart K. M. (1994). Early hominid utilisation of fish resources and implications for seasonality and behaviour. Journal of Human Evolution 27, 229-245. Stewart K. M. & Gifford-Gonzales D. (1994). An ethnoarchaeological contribution to identifying hominid fish processing sites. Journal of Archaeological Science 21, 237- 248.

140

Stiner M., Munro N. D. & Srovell T. A. (2000). The Tortoise and the Hare: Small-game use, the broad spectrum revolution and Paleolithic demography. Current Anthropology 41, 39-59. Stiner M. C., Kuhn S. L., Weiner S. & Bar-Yosef O. (1995). Differential burning, recrystallisation, and fragmentation of archaeological bone. Journal of Archaeological Science 22, 223-237. Stiner M. C. & Munro N. D. (2002). Approaches to prehistoric diet breadth, demography, and prey ranking systems in time and space. Journal of Archaeological Methods and Theory 9. Stiner M. C., Munro N. D., Surovell T. A., Tchernov E. & Bar-Yosef O. (1999). Paleolithic population growth pulses evidence by small animal exploitation. Science 283, 190- 194. Tchernov E. (1979). Quaternary Fauna. In (A. Horowitz, Ed.) The Quaternary of Israel. New York: Academic Press, pp. 260-292. Tchernov E. (1981). The Biostratigraphy of the Middle East. In (P. Sanlaville & J. Cauvin, Eds.) Prehistoire du Levant. Paris: Edition de C.N.R.S, pp. 67-97. Tchernov E. (1988). The Biogeographical History of the Southern Levant. In (Y. Yom-Tov & E. Tchernov, Eds.) The Zoogeography of Israel. Dordrecht: Junk, pp. 401-409. Tsatskin A. & Nadel D. (2003). Formation processes at Ohalo II submerged prehistoric camp site, Israel, inferred from soil micromorphology and magnetic susceptibility studies. Geoarchaeology-An International Journal 18, 409-432. Valla F. R., Bar-Yosef O., Smith P., Tchernov E. & Desse J. (1986). Un noveau sondage sur la terrace d'El Ouad, Israel. Paleorient 12, 21-38. van Elsbergen M. J. (1997). Fischeri im Alten Agypten: Untersuchungen zu den fischfangdarstellungen in den grabern der 4 bis 6 Dynastie. Berlin: Achet Verlag. Van Neer W. (1986). Some notes on fish remains from Wadi Kubbaniya (Upper Egypt; Late Pleistocene). In (D. Brinkhuizen & A. Clason, Eds.) Fish and Archaeology. Cambridge: BAR International Series, pp. 103-113. Van Neer W. (1989). Fishing along the prehistoric Nile. In (L. Kryzaniak & M. Kobuseiwiciz, Eds.) Late Prehistory of the Nile Basin and the Sahara. Ponzan: Ponzan Archaeology Museum, pp. 49-56. Van Neer W. (1999). Contribution to the archaeozoology of the Brandberg, Namibia. Cimbebasia 15, 127-140.

141

Van Neer W. & Pieters M. (1997). Evidence for processing of flatfish at Raverside, a late Medieval coastal site in Belgium. Anthropozoologica 25, 579-586. Van-Neer W. (1995). Analysis of the fish remains. In (S. K. McIntosh, Ed.) Excavations at Jenné-Jeno, Hambarketolo, and Kaniana (Inland Niger Delta, Mali), the 1981 Season. Berkeley: University of California Press, pp. 319-347. von Brandt A. (1972). Fish catching methods of the world. London: Thanet Press. Walker B. (1982a). Scottish buildings for meat and fish preservation. Scotish Industrial History 5, 25-44. Walker B. (1982b). Scotish methods of preserving white fish. In (A. Gailey & D. Ó hÓgáin, Eds.) Gold under the Furze: Studies in Folk Tradition. Dublin: The Glendale Press, pp. 138-149. Weigelt J. (1989). Recent Vertebrae Carcasses and Their Paleobiological Implications. Chicago: University of Chicago Press. Weiner S. & Bar-Yosef O. (1990). States of preservation of bones from prehistoric sites in Near East: a survey. Journal of Archaeological Science 17, 187-196. Weller S. C. & Romney A. K. (1990). Metric Scaling: Correspondence Analysis. Park Newbury CA: Sage University. Wendorf F. & Schild R. (1989). Summary and Synthesis. In (A. E. Close, Ed.) The Prehistory of Wadi Kubbaniya. Dallas: Southern Methodist University Press, pp. 768-824. Wheeler A. & Jones A. J. K. (1976). Fish Remains. In (A. Rogerson, Ed.) Excavation on Fuller's Hill, Great Yarmouth. Norfolk: Norfolk Archaeological Unit, pp. 131-234. Wheeler A. & Jones A. K. G. (1989). Fishes. Cambridge: Cambridge University Press. Whitefield A. K. & Elliot M. (2002). Fishes as indicators of environmental and ecological changes with estuaries: A review of progress and some suggestions for the future. Journal of Fish Biology 61, 229-250. Wilson M. V. H. & Barton D. G. (1996). Seven centuries of taphonomic variation in Eocene freshwater fishes preserved in varves: paleoenvironments and temporal averaging. Paleobiology 22, 535-542. Winfield I. J. & Nelson J. S. Eds. (1991). Cyprinid Fishes: Systematic, Ecology and Exploitation. London: Chapman & Hall. Yechieli Y., Magaritz M., Levy Y., Weber U., Kafri U., Woellfel W. & Bonani G. (1993). Late Quaternary geological history of the Dea Sea area, Israel. Quaternary Research 39, 59-67.

142

Yesner D. R. (1980). Maritime hunter-gatherers: Ecology and Prehistory. Current Anthropology 21, 727-750. Zar J. H. (1984). Biostatistical Analysis. New Jersey: Prentice-Hall, Inc. Zohar I. (2002). Fish and fishing at Ohalo II. In (D. Nadel, Ed.) Ohalo II: A 23,00 Year Old Fisher-Hunter-Gatherers' Camp on the Shore of the Sea of Galilee. Haifa: Reuben and Edith Hechth Museum, University of Haifa., pp. 28-32. Zohar I. & Belmaker M. (2003). Size does matter: methodological comments on sieve size and species richness in fishbone assemblages. Journal of Archaeological Science in press. Zohar I. & Cooke R. (1997). The impact of salting and drying on fish bones: Preliminary observations on four marine species from Parita Bay, Panama. Archaeofauna 6, 59- 66. Zohar I., Dayan T., Galili E. & Spanier E. (2001). Fish processing during the early Holocene: A taphonomic case study from coastal Israel. Journal of Archaeological Science 28, 1041-1053. Zohar I., Dayan T., Spanier E., Galili E. & Lernau O. (1994). Exploitation of gray triggerfish (Balistes carolinensis) by the prehistoric inhabitants of Atlit-Yam, Israel: a preliminary report. In (W. Van Neer, Ed.) Fish Exploitation in The Past: Proceedings of the 7th meeting of the ICAZ Fish Remains Working Group. Teruven: Annales du Musée Royal de l'Afrique Centrale, Sciences Zoologiques no 274., pp. 231-237.

143 APPENDIX-I : Fish remains recovered from prehistoric sites and lacustrine environments in

Israel.

Period Site Location Fish remains Fishing tools Reference 1.9 m.y.:Early Erk-el-Ahmar Central Jordan Freshwater: Cichlidae, - (Tchernov, Pliocene Valley Clariidae, Cyprinidae 1979) , pers. obs.

1.4m.y: Early Middle Ubeidiya Central Jordan Freshwater: Clariidae - (Haas, 1966; Pleistocene (Early Valley (C.lazera),Cyprinidae Tchernov, Acheulian) (Barbus sp.), Cichlidae, 1979), pers. Cyprinodontidae obs.

0.8m.y.:Lower & Gesher North Jordan Fresh water: - pers. obs Middle Pleistocene Benot Valley Cyprinidae, Clariidae, (Middle Acheulian) Ya'akov Cichlidae

Upper-Levalois- Tabun Cave Mount Carmel Diploid fish - Haas in (Jelink Mousterian layer C et al.)

Upper-Levalois- Amud Cave Cyprinidae, Cichlidae - pers. obs. Mousterian

Early Epipaleolithic Ohalo II Sea of Galilee Freshwater: arrowheads (Zohar, 2002) (23,000 cal B.P.) Cyprinidae, Cichlidae harpoons stone weights microlith

Natufian Hayonim B 13 km east of the Unidentified fish - (Stiner & coast Munro, 2002)

Natufian Mallaha Jordan Valley Fresh water fish: stone weight (Desse, 1987), (Eynan) Cyprinidae, Clariidae, pers. obs. Cichlidae

Natufian Hilazon Western Galilee Unidentified - pers. obs. Tachtit

Natufian Hatula Jeudea Mountain Serranidae, Sparidae, - (Davis, 1985; Sciaenidae, Muglidae Lernau & Lernau, 1994)

Natufian El-Wad Eastern Sparidae, Mullidae, Points, (Valla et al., Mediterranean Muglidae, Serranidae microlithics 1986), pers. obs.

152 APPENDIX-I, cont'd.

Period Site Location Fish remains Fishing tools Reference Khiamian Hatula Jeudea Mountain Serranidae, (Davis, 1985; Percichthyidae, Lernau & Carangidae, Sparidae, Lernau, 1994) Sciaenidae, Muglidae, Scombridae

Sultanian Hatula Jeudea Mountain Serranidae, Percichthyidae, Sparidae, Sciaenidae, Muglidae, Cyprinidae, Cichlidae, Rays & Shark

PPNB Kfar Lower Galilee unidentified Arrowheads (Gorring-Morris 9200-8500 BP Hahoresh Points Sickles et al., 1994-5) blades

PPNC Atlit-Yam Eastern B. carolinenesis, stone (Galili et al., Mediterranean Serranidae,Carngidae(T. weights, 1993; Zohar et trachurus), Sparidae, spearheads al., 1994) Scianidae, arrowheadsh Muglidae,Rays&Shark arpoons

PPN Netiv- Jordan Valley Cypriniforms Khiam points (Bar-Yosef, Hagdud Sickle Blade 1991; Bar-Yosef Microliths et al., 1991)

153 APPENDIX-II: LEVANTINE FRESHWATER FISH II.1 Morphological and osteological characteristics Freshwater fish are sensitive to changes in salinity level, and are confined to drainage systems. Therefore, they can provide a relatively conservative system for examining distribution patterns that may reflect continental and environmental changes (Berra, 2001). A classification of fishes living in fresh water has been developed by Myers (Myers, 1938; Myers, 1951) based on their tolerance to salinity changes. According to this classification there are three main categories: species that usually occur in the sea and enter freshwater either rarely or frequently (peripheral); species regularly passing from sea to freshwater (Diadormous); and species restricted to freshwater (Banarescu, 1990a). The latter group comprise two sub-categories of freshwater fishes: Primary- which are strictly intolerant to salinity changes and are confined to fresh waters; and Secondary- which are somewhat strictly confined to freshwater but have some salt tolerance and whose distribution may reflect dispersal through coastal waters or across short distances of salt waters. The definition of these categories refers to ecological niches, since the first category consists of families living exclusively in freshwater, while the second category also contains a number of euryhaline species (Banarescu, 1990a; Berra, 2001). In the Levant, 65 species of primary and secondary freshwater fishes occur in freshwater habitats. Based on their distribution pattern, the Levant is divided into 13 ichthyogeographical subprovinces (Krupp, 1987; Krupp & Schneider, 1989). Comparison between the fish species that occur in each subprovince, reveals the following: 1. Freshwater species that originated from different drainage basins mainly Palaeartic, Indoasiatic and Afrotropical. Secondary freshwater fish such as the Cichlidae are of Afrotropical origin, having migrated throughout the Mediterranean Sea to the Jordan rift valley. However, primary freshwater fish such as the Cyprinidae, Cyprinodontidae, and Balitoridae are of Paleartic and Indoasiatic origin, and their distribution is dependent on freshwater riverine connections from the Euphrates throughout the Damascus basin to the Jordan River (Banarescu, 1990b; Krupp, 1987). 2. Endemic species that evolved in each subprovince, in response to ecological isolation (Krupp, 1987; Krupp & Schneider, 1989).

154 Figure 1: Ichthyogeographic 1 Tigris-Euphrates subprovinces of the Levant and adjacent areas (after (Krupp, 1987).

3 1-Ceyhan, 2-Orontes, 3- Quwaiq, 4- Tigris-Euphrates, 5- Nahr al Kabir (N), 6- 5 Nahr Marqiya, 7-Nahr al Kabir (S), 8- 2 4 Coastal rivers of Lebanon, 9-Coastal rivers 6 Orontes of Israel, 10- Litani, 11- Damascus basin, Mediterranean Sea 7 12- Jordan Rift Valley, 13- Wadi as-Sirhan

10 In the Orontes 34 species of 8

11 freshwater fish occur, of which18

Hula basin species also occur in the Tigris and Sea of Galilee Euphrates, and only 7 of whichoccur in the Sea of Galilee 9 and Jordan river system (marked as

12 * in Table 1). In the Sea of Galilee 13 and the Jordan river system seven families, comprising 30 primary and secondary freshwater species, 0 100km occur, of which 13 species are endemic to their habitat (Table 1). Therefore the appearance of freshwater fish remains (primary and secondary) and especially endemic species in lacustrine sedeiments, offer an outstanding marker of environmental conditions (Goren, 1983; Goren & Ortal, 1999; Krupp, 1987; Krupp & Schneider, 1989). The origin of the fish (Palaeartic/ Afrotropicala) also determines their breeding season (Table 2). Palearctic fish will breed during the winter while Afrotropical fish breed during the spring/ early summer. These differences can serve as markers for season of capture, as well as location. Table 1: List of recent freshwater fishes inhabiting the Sea of Galilee, Jordan Rift Valley, Dead Sea System and Coastal rivers of Israel. Jordan valley Coastal Family Species Origin North Central South rivers Cyprinidae A. hulensis Endemic-Hula + A. lissneri Endemic + + A. telavinensis Endemic + A. terraesanctae Endemic- Sea of Galilee +

155 Barbus canis Asia-Africa + B. longiceps Asia + Capoeta damascina Middle East * + + Garra rufa Asia * + G. sedomitica Endemic- Neot Hakikar + Hemigrammocapoeta nana Middle East + + Phoxinellus drusensis Asia-Europe + Ph. kervillei Asia-Europe* +

Balitoridae Noemacheilus dori Endemic Beit-Shean + N.tigris Asia-Europe* + N. jordanicus Endemic Jordan valley + N. panthera Asia-Europe* + Nun galiaeus Endemic-Hula +

Clariidae Clarias gariepinus Africa * + +

Cyprinodontidae Aphanius mento Asia* + + + A. dispar richardsoni Endemic-Dead Sea +

Cichlidae Tilapia zilli Africa + + + Oreochromis aureus Africa + + + O. nilotocus Africa + Sarotherodon galilaeus Africa + +

Astatotilapia flaviijosephi Endemic + Tritramella sacra Endemic-Sea of Galilee + T. simonis intermedia Endemic-Hula + T. simonis simonis Endemic-Sea of Galilee +

Blenniidae Salaria fluviatilis Mediterranean + +

Anguillidae Anguilla anguilla Saragaso Sea +

* Species that occur in the Oronthes, Endemic species are marked in blue

156 Table 2: Breeding seasons and areas of freshwater fish in the northern part of the Jordan rift valley (following Goren, 1983). Autumn Winter Spring Summer Breeding area Species 9 10 11 12 1 2 3 4 5 6 7 8 A. lissneri ______Rivers, on stones and algae A. telavinensis ______Coastal rivers, on stones. A. terraesanctae ______Shallow water, Sea of Galilee, on stones B. canis ______Shallow riverine water on stones. B. longiceps ______Shallow riverine water, in sand. C. damascina ______Shallow water rivers, in sand. G. rufa ______On stones H. nana ______Shallow riverine water Ph. drusensis ______Shallow riverine water, on stones. Ph. kervillei ______Shallow riverine water, on stones. N.dori __ __ Beit Shean valley N.tigris ______Sea of Galilee, on stones. N. jordanicus ______Shallow riverine water, on stones and roots. N. panthera ______Shallow riverine water, on stones. C. gariepinus ______Shallow water A. mento ______Shallow water, between vegetation. T. zilli ______Monogamy, shallow water, in hutches O. aureus ______Monogamy, shallow sandy water, in hutches. O. nilotocus ______Monogamy, shallow sandy water, in hutches. S. galilaeus ______Monogamy, shallow rocky water, in mouth. A. flaviijosephi ______Monogamy, shallow sandy water, in hutches and mouth T. sacra ______Monogamy, shallow sandy water, in hutches and mouth T. s intermedia ______Monogamy, shallow sandy water, in hutches and mouth T. s simonis ______Monogamy, shallow sandy water, in hutches and mouth

157 II.2. Cyprinidae The Cyprinidae is the largest fish family in the world, with 210 genera and more than 2,010 species (Berra, 2001). It occurs in freshwater habitats of Africa, Europe, Asia and North America (Winfield & Nelson, 1991). The earliest cyprinid fossils are from the lower- Miocene in Africa (Otero, 2001), middle Eocene in Eurasia and Oligocene age in North America (Winfield & Nelson, 1991). Cyprinids probably dominated freshwater ichthyofauna as early as the Oligocene in parts of Asia, while in Siberia, Europe and North America they appear only from the Mocene deposits (Winfield & Nelson, 1991). According to the fossils, the radiation of cyprinids into many specialized feeding types occurred in the late Miocene in the Euro-Siberian region and North America. In East and North Asia, the fossil record indicates on earlier cyprinid radiation at the end of the Palaeogene (Winfield & Nelson, 1991). In Israel there are 12 species belonging to the family Cyprinidae, inhabiting various fresh water habitats (Tables 1-3). The Cyprinidae demonstrate wide morphological diversity (Figure 2) and their body size varies from small sized species of ca. 10 cm in total length (TL) to large species of ca. 3 meter length (Berra, 2001). Despite the diverse morphological variations, there are several anatomical and osteological features that are typical for all Cyprinids (Figure 2).

Barbus canis Capoeta damascina

Acanthobrama terraesnctae Acanthobrama lis Figure 2: The most abundant Cyprinids from the Jordan rift valley and the Sea of Galilee (after Golani & Darom, 1997). These includ: a protruding mouth; presence of anterior (rostal) and posterior barbells; toothless jaws; toothless palate; enlarged, tooth bearing pharyngeal bones (modified last gill arches; Figure 3); basioccipital with an enlarged ventroposterior process enclosing the dorsal aorta and bearing a keratinized pad against which the pharyngeal teeth exert grinding force.

158 Table 3: List of Cyprinids inhabiting freshwater habitats in Israel, their reproduction season, location, and maximum total length (mm). Species Habitat Length Reproduction Season Barbus canis Sea of Galilee, Jordan rift Rivers- 400-660mm April-August Pelagic & Benthic Shallow water Barbus longiceps Sea of Galilee, Jordan rift 500-700mm December-February Rivers- Benthic Upper streams/ rivers Capoeta damascina Sea of Galilee, Jordan rift 150-350mm November-February Rivers-Benthic Upper streams/ rivers Acanthobrama terraesanctae Sea of Galilee 120-200mm November-Mars Pelagic Shallow water Acanthobrama hulensis Hulla basin (Extinct species) February-April Benthic Acanthobrama lissneri Jordan rift Rivers 100mm January-April Benthic Shallow water Acanthobrama telavinensis Coastal rivers (Hayarkon) 100mm January-April Benthic Shallow water Garra rufa Jordan rift Rivers 140mm February-August Benthic Shallow water Garra sedomitica Neot Hakikar springs 140mm Hemigrammocapoeta nana Sea of Galilee, Jordan rift 120mm Winter Rivers-Benthic Phoxinellus kervillei Sea of Galilee, Northern rivers 80-100mm February-April Shallow water Phoxinellus drusensis Ramat Hagolan rivers 90mm April-June Shallow water

Osteological characteristics The Cyprinids' pharyngeal bone and teeth constitutes one of the most characteristic bones, evolved during adaptation to their feeding habits (Figure 3-4). This bone differs between species in several characteristics: number of rows (1-3), number of teeth in each row, number of teeth on each side of the bone, teeth shape, teeth abrasion, and presence or absence of molariform teeth (Table 4).

Table 4: Characteristics of pharyngeal bone and teeth in Cyprinids species inhabiting

freshwater habitats of the southern Levant .

Species Locality No. of Number of Teeth Structure of Pharyngeal Teeth rows B. barbus Turkey 3 5 external +3+2 Pointed teeth B. capito Turkey 3 4 external +3+2 1 external molariform teeth B. esocinus Turkey 3 4 external +3+2 1 external molariform teeth B. longiceps Israel 3 4 external +3+2 1 external molariform teeth, external teeth compressed with a smooth tip, other 5 teeth smooth and rounded. B. canis Israel 3 5 external +3+2 Pointed teeth, with short hooked tips C. damascina Israel 3 4 external +3+2 Abrupted teeth, flat topped visible enamel A. terraesanctae Israel 1 5-5 external Anterior tooth short and sturdy with a pointed tip. Other 4 teeth moderately compressed and bevelled with short hooked tips. A. hulensis Israel 1 5-4 external Stout and flat topped A. lissneri Israel 1 5-5 external Smooth and hooked A. telavinensis Israel 1 5-5 external Smooth and hooked

159 A. marmid Turkey 1 5-5 external Anterior tooth compressed and bluntly pointed. Other 4 teeth bevelled with a clear cutting edge. A. mirabilis Turkey 1 5-5 external G. rufa Israel 3 4 external +3+2

Figure 3 : Typical skull of Cyprinidae and location of pharyngeal bone (after Winfield & Nelson, 1991)

The first characteristic differentiating between species is the number of teeth rows (Table 4). Most species have three row of teeth (Figure 4), except for Acanthobrama sp., which exhibit a single row (Figure 5). The second characteristic is the number of teeth in a row, which usually exhibits symmetry between left and right pharyngeal bones. A. hulensis is the only species that exhibits asymmetry (Goren et al., 1973). The third characteristic is teeth shape and abrasion, which vary between species, due to different feeding habits (Figures 4-7). In bottom feeders such as A. hulensis the teeth are stout and flat (Figure 5), while in A. terraesanctae which is a pelagic feeder the teeth are beveled with short hooked tip (Figure 5). In molluscivore fish, such as Barbus longiceps, "molariform" teeth are present (Figures 4, 6),while in B.canis the teeth are pointed (Figure 4). The molariform teeth are species specific and differ according to the crown contour and surface modification (Figure 6). In Capoeta damascina the pharyngeal teeth are highly worn and the enamel can be observed (Figure 7).

Figure 4: Pharyngeal bone and teeth of Barbus canis (left) and B. longiceps (right).

160 Figure 5: Pharyngeal bone and teeth of A. terraesanctae(left), A. hulensis (middle; after Goren

et al., 1973), and A. issneri (right).

Figure 6: An example of cyprinid's molariform pharyngeal teeth (Dorsal and lateral view).

161 Figure 7: Pharyngeal bone and teeth of Capoeta damascina (left) and Gara ruffa (right).

Unlike the diverse variation in pharyngeal bone and teeth, Cyprinidae jaws display a relatively conservative toothless structure (Figures 8-9). This includes the dentary (Figure 8), articular (Figure 8), maxilla (Figure 9), and premaxilla (Figure 9), which are species specific.

Figure 8: Morphological characteristics of dentary (left; ventral & dorsal view) and of articular (right) bones from the oromandibular region of A. terraesanctae. Figure 9: Morphological characteristics of maxilla (left; ventral & dorsal view) and

premaxilla (right) bones from the oromandibular region of A. terraesanctae.

Fish adaptation to different feeding habits influences many of the skull and branchial region bones. Therefore, the morphological characteristics of bones from the hyoid region, especially the urohyal (Figure 10) are species specific. These include the hyomandibular bone (Figure 11), and the ceratohyal and epihyal (Figure 12) bones.

Figure 10: Morphological characteristics of A. terraesnctae urohyal.

162 Figure 11: Morphological characteristics of hyomandibular bone of A terraesanctae. (ventral & dorsal view).

Fig ure 12: Mo rph olo gic al characteristics of epihyal (left) and ceratohyal (right) of A terraesanctae. (ventral & dorsal view).

Another osteological feature that characterises the cyprinids is the Weberian apparatus, that was first described by Weber in 1820. In Cyprinidae (Figure 13) this set of bones is connected to the gas bladder (Rojo, 1991). These bones originated from the first four vertebrae and include the ribs (Figure 14), tripus (Figure 15), basapophysis (Figure 15), spina neuralis of 3rd vertebrae (Figure 16), and os suspensorium of 4th vertebrae (Figure 16). Figure 13: General structure of Weberian apparatus in Cyprinid's (after Winfield & Nelson, 1991).

163

Figure 14: B. canis rib proximal end, dorsal view.

Figure 15: Morphological characteristic of tripus (left) and basapophysis (right) from the Weberian apparatus of A. terraesanctae.

164 Figure 16: Morphological characteristic of spina neuralis 3rd vertebrae (top) and os suspensorium 4th vertebrae (bottom), from the Weberian apparatus of A. terraesanctae. According to the thoracic vertebrae morphology, mainly the fusion or separation of vertebrae 2 and 3, Cyprinids can be identified to genus and species level (Table 5). Acanthobrama sp. are characterised by separated vertebrae (Figure 17) while in all other genera,the second and third thoracic vertebrae are fused (Figures 18-19). The Weberian apparatus also influences the morphology of other thoracic vertebrae, especially the atlas, and vertebrae number 4 and 5. These vertebrae can be identified to the genus level, and in some cases they are species specific (Figures 20-21).

Table 5: Weberian apparatus characteristics according to the structure of axis and third thoracic vertebrae in various Cyprinidae inhabiting the Middle East region. Species 2nd & 3rd vertebrae Species 2nd & 3rd vertebrae B. barbus fused A. terraesanctae separated B. capito fused A. hulensis separated B. esocinus fused A. lissneri separated B. canis fused A. telavinensis separated B. longiceps fused A. marmid separated C. damascina fused A. mirabilis separated G. rufa fused

Figure-17: Morphological characteristics of Axis and 3rd thoracic vertebrae in A.terraesanctae (left) and A.lissneri (right).

165

Figure 18: Axis orphological characteristics in Barbus canis (left) and B. longiceps (right).

Figure 19: Axis morphological characteristics of in C. damascina (left), and G. ruffa (right).

Barbus canis Barbus longiceps

Capoeta damascina Gara ruffa

Acanthobrama terraesanctae Acanthobrama lissneri

Figure 20: Morphological characteristic of Cyprinidae atlas, by species.

166 Figure 21: Morphological characteristics of A. terraesanctae 4th vertebrae (left) and 5th vertebrae (right: ventral and dorsal view).

Although vertebrae are usually the most common bones found in archaeological sites, most of them are not species specific (Wheeler & Jones, 1989). In the case of Cyprinidae all vertebrae can be identified to family level. For example, in figure 22 , I demonstrate the characteristic features of the cyprinids thoracic vertebrae. This can be compared to the typical structure of cyprinid precaudal and caudal vertebrae (Figure 23), and to penultimate and ultimate vertebrae (Figure 24). Figure 22: Morphological characteristics of A. terraesanctae thoracic vertebrae (ventral and dorsal view).

167

Figure 23: Morphological characteristics of A. terraesanctae precaudal (left) and caudal (right) vertebrae.

Figure 24: Morphological characteristics of A. terraesanctae penultimate (left) and ultimate (right) vertebrae.

In Cyprinidae most skeletal elements are family specific and several cranial and post cranial bones are species specific. In figures 25-38 some of the cranial bones are displayed.

Figure 25: A.terraesanctae frontal bone on the left , and prefrontal bone on the right .

168

Figure 26: A.terraesanctae ethmoid bone, on the left and B. canis ectethmoid bone, on the right.

Figure 27:

A.terraesanctae vomer on the left and parasphenoid on the right.

Figure 28: A.terraesanctae exoccipital on the left, and basioccipital bone on the right (Dorsal &

Lateral view).

Figure 29: A. terraesanctae supraoccipital bone.

169 Figure 30: A. terraesanctae pterotic bone (dorsal & ventral view).

Figure 31: B. canis Prootic on the left and sphenotic bone on the right.

Figure 32: B. canis epiotic bone on the left and alisphenoid bone on the right.

170

Figure 33: A. terraesanctae parietal bone on the left and B. canis orbito sphenoid bone on the right. Figure 34: A. terraesanctae posttemporal bone.

Figure 35: B. canis representative bones from the opercle apparatus: subopercle on the left and interopercle on the right.

Figure 36: A. terraesanctae representative bones

from the opercle apparatus: Opercle, on top, and

preopercle, on bottom.

171 Figure 37: A. terraesanctae representative bones of mandibular arch: palatine bone on left, and

quadrate on right. Figure 38: B. canis representative bones of mandibular arch: metapterygoid on left and entopterygoid bone on the right. From the postcranial region, several bones are species specific. These include the supracleithrum (Figure 39), cleithrum (Figure 40), and coracoid (Figure 42). The first dorsal pterygiophore is species specific (Figure 44). However, other pterygiophores can be identified only to family level (Figure 44).

Figure 39: B. canis supracleithrum, on right, and A. terraesanctae postcleithrum, on left. Figure 40: A. terraesanctae cleithrum (Ventral & Dorsal view).

172

Figure 41: A. terraesanctae scapula (Ventral & Dorsal view) on the left and mesocoracoid bone, on the right.

Figure 42: A. terraesanctae coracoid bone (dorsal & ventral view).

Figure 43: A. terraesanctae pelvis bone (dorsal & ventral view).

173

Figure 44: A. terraesanctae 1st dorsal pterygiophore on left and median pterygiophore on right.

II.3. Cichlidae The Cichlidae is probably the third largest family of bony fishes, after the Cyprinidae and Gobiidae. The exact number of genera and species is not precisely known but has been estimated at 200 genera and 2000 species (Berra, 2001). The cichlids are widely distributed in central America, South America, West Indies, Madagascar, coastal India, Sri Lanka, and the Middle East (Banarescu, 1990a; Berra, 2001). Several researchers have explained their distribution pattern as a result of dispersal from the Tethys Sea (Banarescu, 1990a). The cichlids exhibit tolerance to salinity changes and therfore many species are found in brackish water, while others travel along the coast line between river systems (Banarescu, 1990a; Berra, 2001). In Israel out of the eight species of cichlids (Table1), four are considered to be of Afro- origin, while the other four are endemic species (Berra, 2001). Cichlid body shape and color are diverse, and consequently they have became one of the most important groups of aquarium animals. Most cichlids are relatively small (10-30cm TL), except for few African species that can reach 1 meter in TL (Berra, 2001). Some cichlids have a compressed, disc- shaped body, while others are elongated and streamlined (Figure 45). The cichlids are easily identified in having only one nostril on each side instead of the usual two, and their lateral line is interrupted (i.e., it consists of an anterior, longer upper section and a posterior, shorter lower section of pored scales; Figure 45).

174

Sarotherodon galilaeus Oreochromis aureus

Tristamella simonis Tilapia zilli Figure 45: The most abundant cichlids from the Sea of Galilee (after Golani and Darom).

Cichlid feeding habits are diverse and include: zooplankton, scales, plants, and anthropods eaters, piscivore hunters, molluscs crushers and rock scrapers (Liem, 1991). The differences in feeding habits are clearly reflected in the shape of several cranial mechanical units, especially the neurocranium, suspensory apparatus, jaw apparatus (Figure 46) and pharyngeal jaw apparatus (Liem, 1991). The pharyngeal jaw apparatus (Figure 47) is composed of the upper pharyngeal jaw ( pharyngobranchials 2 and 3 and the toothplate of the fourth) and lower pharyngeal jaw (left and right 5th ceratobranchials). Therefore, variations between cichlids are found in the size and shape of the pharyngeal bone apparatus. In mollusc-crusher cichlids the pharyngeal jaws have round molariform teeth. Figure 46: O. aureus dentary on the left, and maxilla on the right.

175

Figure 47: O. aureus lower pharyngeal bone, on left, and upper pharyngeal bones, on right.

Other osteological characteristics of cichlids include the appearance of fin spines. Most cichlids have 7-25 dorsal fin spines and 3-9 anal fin spines that are easy to identify to family level. The first dorsal fin spine caries distinctive characteristics, to family level (Figure 48). Moreover, the cichlids have ca. 30 vertebrae, which can be differentiated according to their location at the vertebral column, to atlas, axis, thoracic, precaudal, caudal, and ultimate vertebrae (Figures 49-50).

Figure 48: O. aureus

first dorsal fin spine.

Figure 49: O. aureus atlas on left, and axis vertebrae on the right.

Figure 50: O. aureus thoracic vertebrae on the left, compare to its ultimate vertebrae, on the right.

176

A comparison between cichlid and cyprinid osteological characteristics demonstrates distinct differences between the skeletal elements of each family (Figure 51). For example, in cichlids the supraoccipital crest is more pronounced than in cyprinids (Figure 51).

Figure 51: Cichlids skull on the left, compare to Cyprinids skull on the right (after Liem, 1991; Winfield & Nelson, 1991).

177

APPENDIX-III Cichlidae skeletal elements in a complete fish*

Anatomic region/ Skeletal element- Cranial Expected no.

Olfactory Region (n=6) Ethmoid 1 Vomer 1 Prefrontal 2 Nasal 2 Orbital Region (n=16) Frontal 2 Lacrimal 2 Alisphenoid 2 Total orbital Series 10 Otic Region (n=23) Epiotic 2 Exoccipital 2 Opisthotic 2 Pareital 2 Posttemporal 2 Prootic 2 Pterotic 2 Sphenotic 2 Supraoccipital 1 Otolith 6 Basicranial Region (n=3) Basioccipital 1 Basisphenoid 1 Parasphenoid 1 Oromandibular region (n=20) Premaxilla 2 Maxilla 2 Dentary 2 Articular&Angular 4 Palatine 2 Ectopterygoid 2 Metapterygoid 2 Entopterygoid 2 Quadrate 2 Hyoid region (n=16) Hyomandibular 2 Symplectic 2 Interhyal 2 Epihyal 2 Ceratohyal 2 Hypohyal 4 Urohyal 1 Glossihyal 1

178

APPENDIX-III cont'd

Anatomic region/ Expected no. Skeletal element-Cranial Opercular Series (n=8) Peropercular 2 Opercular 2 Interopercular 2 Subopercular 2 Branchial Arch (n=36) Epibranchial 8 Ceratobranchial 10 Hypobranchial 6 Basibranchial 6 Pharyngobranch 4 Lower pharyngeal bone 2 PostCranial Appendicular skeleton (n=16) Pelvis 2 Pelvic Spines 2 Cleithrum 2 Coracoid 2 D.Postcleithrum 2 V.Postcleithrum 2 Scapula 2 Supracleithrum 2 Median Fin (n=21) 1st anal pterygiophore 1 Fin spines 20 Vertebrae n=30 Atlas 1 Axis 1 Thoracic vertebrae 14 Precaudal & caudal vertebrae 12 Penultimate vertebrae 1 Ultimate vetebrae 1 *Doesn't include the following elements:Scales, Ribs, intermuscular, pterygiophore, soft fin ray interhamel, epural, hypural, urostyle and radials.

179

APPENDIX-IV Cyprinidae skeletal elements in a complete fish*.

Anatomic region/ Skeletal element- Cranial Expected no. Olfactory Region(n=8) Ethmoid (=mesethmoid) 1 Ectethmoid 2 Vomer 1 Prefrontal 2 Nasal 2 Orbital Region (n=15) Frontal 2 Lacrimal 2 Alisphenoid 2 Orbito sphenoid 1 Total orbital Series 8 Otic Region (n=25) Epiotic 2 Exoccipital 2 Opisthotic 2 Pareital 2 Posttemporal 2 Prootic 2 Pterotic 2 Sphenotic 2 Scale Bone 2 Supraoccipital 1 Otoliths no. 6 Basicranial Region (n=3) Basioccipital 1 Basisphenoid 1 Parasphenoid 1 Oromandibular region (n=20) Premaxilla 2 Maxilla 2 Dentary 2 Articular&Angular 4 Palatine 2 Ectopterygoid 2 Metapterygoid) 2 Entopterygoid 2 Quadrate 2 Hyoid region (n=16) Hyomandibular 2 Symplectic 2 Interhyal 2 Epihyal 2 Ceratohyal 2 Hypohyal 4 Urohyal 1 Glossihyal 1

180

APPENDIX-IVcont'd

Anatomic region/ Expected no. Skeletal element-Cranial Branchial Arch (n=58-60) Epibranchial 6 Ceratobranchial 8 Hypobranchial 8 Basibranchial 8 Pharyngobranch 6 pharyngial bone 2 Branchial teeth 18-20 Opercular Series (n=8) Peropercular 2 Opercular 2 Interopercular 2 suboperclar 2 Postcranial Appendicular skeleton (n=16) Pelvis 2 Cleithrum 2 Coracoid 2 Mesocoracoid 2 D.Postcleithrum 2 V.Postcleithrum 2 Scapula 2 Supracleithrum 2 Median Fin (n=7) 1st dorsal pterygiophore 1 3rd internural 3 Fin spines 2 Spina haemalis penultimate v. 1 Vertebrae (n=45) Atlas 1 Axis (v.2&3 fused) 1 Fourth vertebrae 1 Fifth vertebrae 1 Thaoracic vertebrae 17 Precaudal & caudal vertebrae 21 Penultimate vertebrae 1 Ultimate vertebrae 1 Weberian apparatus (n= 10) Tripus 2 Spina neuralis 2nd vert. 2 Os suspensorioum 4th vert. 2 Basapophysis 3rd vert. 2 Spinaneuralis 3rd vert. 2

181

APPENDIX-VAcanthobrama terraesanctae skeletal elements in a complete fish*.

Anatomic region/ Expected no. Skeletal element-Cranial Olfactory Region(n=8) Ethmoid (=mesethmoid) 1 Ectethmoid 2 Vomer 1 Prefrontal 2 Nasal 2 Orbital Region (n=15) Frontal 2 Lacrimal 2 Alisphenoid 2 Orbito sphenoid 1 Total orbital Series 8 Otic Region (n=25) Epiotic 2 Exoccipital 2 Opisthotic 2 Pareital 2 Posttemporal 2 Prootic 2 Pterotic 2 Sphenotic 2 Scale Bone 2 Supraoccipital 1 Otolith 6 Basicranial Region (n=3) Basioccipital 1 Basisphenoid 1 Parasphenoid 1 Oromandibular region (n=20) Premaxilla 2 Maxilla 2 Dentary 2 Articular&Angular 4 Palatine 2 Ectopterygoid 2 Metapterygoid) 2 Entopterygoid 2 Quadrate 2 Hyoid region (n=16) Hyomandibular 2 Symplectic 2 Interhyal 2 Epihyal 2 Ceratohyal 2 Hypohyal 4 Urohyal 1 Glossihyal 1 APPENDIX-V cont'd

Anatomic region/ Expected no. Skeletal element-Cranial Branchial Arch (n=50) Epibranchial 6 Ceratobranchial 8 Hypobranchial 8 Basibranchial 8

182

Pharyngobranch 6 pharyngial bone 2 branchial teeth 10 Opercular Series (n=8) Peropercular 2 Opercular 2 Interopercular 2 suboperclar 2 Postcranial

Appendicular skeleton (n=16) Expected no. Pelvis 2 Cleithrum 2 Coracoid 2 Mesocoracoid 2 D.Postcleithrum 2 V.Postcleithrum 2 Scapula 2 Supracleithrum 2 Median fin (n=4) 1st dorsal pterygiophore 1 3rd internural 2 Spina haemalis penultimate vert. 1 Vertebrae n=39 Atlas 1 Axis 1 Third vertebrae 1 Fourth vertebrae 1 Fifth vertebrae 1 Thaoracic vertebrae 12 Precaudal vertebrae 4 Caudal vertebrae 16 Penultimate vertebrae 1 Ultimate vertebrae 1 Weberian apparatus (n=10) Tripus 2 Spina neuralis 2nd vert. 2 Os suspensorioum 4th vert. 2 Basapophysis 3rd vert. 2 Spinaneuralis 3rd vert. 2

183

APPENDIX-VI Clarias gariepinus skeletal elements in a complete fish*.

Anatomic region/ Expected no. Skeletal element-Cranial Olfactory Region (n=6) Ethmoid 1 Nasal 2 Prefrontal 2 Supraethmoid 1 Orbital Region (n=10) Frontal 2 Lachrymal 2 Alisphenoid 2 Total orbital Series 4 Otic Region (n=21) Epiotic 2 Exoccipital 2 Post temporal 2 Prootic 2 Pterotic 2 Sphenotic 2 Supraoccipital 1 Scale bones 2 Otolith 6 Basicranial Region (n=5) Basioccipital 1 Parasphenoid 1 Fused vertebra complex 1 Tripus 2 Oromandibular region (n=18) Premaxilla 2 Maxilla 2 Dentary 2 Articular&Angular 4 Palatine 2 Pal. teeth patch 2 Metapterygoid 2 Quadrate 2 Hyoid region (n=14) Hyomandibular 2 Interhyal 2 Epihyal 2 Ceratohyal 2 L&U Hypohyal 4 Urohyal 1 Glossihyal 1 Opercular Series (n=6) Peropercular 2 Opercular 2 Interopercular 2

184

APPENDIX-VI cont'd

Anatomic region/ Expected no. Skeletal element-Cranial Branchial Arch (n=22) Epibranchial 6 Ceratobranchial (8) 8 Hypobranchial (4) 4 Basibranchial (2) 2 Pharyngobranch(2) 2 Postcranial Appendicular skeleton (n=8) Pelvis 2 Cleithrum 2 Scapula 2 Coracoid 2 Vertebrae (n=63) Atlas 1 Axis 1 Thoracic vert. 14 Precaudal vert. 22 Caudal vert. 24 Ultimate vert. 1 Median fin (n=2) Pectoral spines 2

185

APPENDIX-VII Pomadasys nitidus skeletal elements in a complete fish*.

Anatomic region/ Expected no. Skeletal element-Cranial Olfactory Region (n=6) Ethmoid 1 Vomer 1 Prefrontal 2 Nasal 2 Orbital Region (n=16) Frontal 2 Lacrimal 2 Alisphenoid 2 Total orbital Series 10 Otic Region (n=23) Epiotic 2 Exoccipital 2 Opisthotic 2 Pareital 2 Posttemporal 2 Prootic 2 Pterotic 2 Sphenotic 2 Supraoccipital 1 Otolith 6 Basicranial Region (n=3) Basioccipital 1 Basisphenoid 1 Parasphenoid 1 Oromandibular region (n=20) Premaxilla 2 Maxilla 2 Dentary 2 Articular&Angular 4 Palatine 2 Ectopterygoid 2 Metapterygoid 2 Entopterygoid 2 Quadrate 2 Hyoid region (n=16) Hyomandibular 2 Symplectic 2 Interhyal 2 Epihyal 2 Ceratohyal 2 Hypohyal 4 Urohyal 1 Glossihyal 1

186

APPENDIX-VII cont'd

Anatomic region/ Expected no. Skeletal element-Cranial Opercular Series (n=8) Peropercular 2 Opercular 2 Interopercular 2 subopercular 2 Branchial Arch (n=40) pharyngial plate 4 Epibranchial 6 Ceratobranchial 8 Hypobranchial 8 Basibranchial 8 Pharyngobranch 6 Postcranial Appendicular skeleton (n=16) Pelvis 2 Pelvic Spines 2 Cleithrum 2 Coracoid 2 D.Postcleithrum 2 V.Postcleithrum 2 Scapula 2 Supracleithrum 2 Vertebrae n=? Median Fin (n=?) Total pterygiophgore 1st anal pterygio. 1

187

APPENDIX-VIII Cathorops multiradiatus and Arius kessleri skeletal elements in a complete fish*.

Anatomic region/ Expected no. Skeletal element-Cranial Olfactory Region (n=6) Ethmoid 1 Nasal 2 Prefrontal 2 Supraethmoid 1 Orbital Region (n=10) Frontal 2 Lachrymal 2 Alisphenoid 2 Total orbital Series 4 Otic Region (n=21) Epiotic 2 Exoccipital 2 Post temporal 2 Prootic 2 Pterotic 2 Sphenotic 2 Supraoccipital 1 Scale bones 2 Otoliths no. 6 Basicranial Region (n=5) Basioccipital 1 Parasphenoid 1 Fused vertebra com. 1 Tripus 2 Oromandibular region (n=18) Premaxilla 2 Maxilla 2 Dentary 2 Articular&Angular 4 Palatine 2 Palatine teeth patch 2 Metapterygoid 2 Quadrate 2 Hyoid region (n=14) Hyomandibular 2 Interhyal 2 Epihyal 2 Ceratohyal 2 L&U Hypohyal 4 Urohyal 1 Glossihyal 1 Opercular Series (n=6) Peropercular 2 Opercular 2 Interopercular 2

188

APPENDIX-VIII cont'd

Anatomic region/ Expected no. Skeletal element-Cranial Branchial Arch (n=22) Epibranchial 6 Ceratobranchial 8 Hypobranchial 4 Basibranchial 2 Pharyngobranch 2 Appendicular skeleton (n=8) Pelvis 2 Cleithrum 2 Scapula 2 Coracoid 2 Median fin 1st dorsal spine 1 Dorsal spine base 1 Dorsal spine plate 1 Pectoral spines 2 Vertebra total no.

189

APPENDIX-IX Caranx caninus skeletal elements in a complete fish*.

Anatomic region/ Expected no. Skeletal element-Cranial Olfactory Region (n=6) Ethmoid 1 Vomer 1 Prefrontal 2 Nasal 2 Orbital Region (n=14) Frontal 2 Lacrimal 2 Alisphenoid 2 Total orbital Series 8 Otic Region (n=19) Epiotic 2 Exoccipital 2 Opisthotic 2 Pareital 2 Posttemporal 2 Prootic 2 Pterotic 2 Scale Bone 2 Sphenotic 2 Supraoccipital 1 Otolith 6 Basicranial Region (n=3) Basioccipital 1 Basisphenoid 1 Parasphenoid 1 Oromandibular region (n=20) Premaxilla 2 Maxilla 2 Dentary 2 Articular&Angular 4 Palatine 2 Ectopterygoid 2 Metapterygoid 2 Entopterygoid 2 Quadrate 2 Hyoid region (n=16) Hyomandibular 2 Symplectic 2 Interhyal 2 Epihyal 2 Ceratohyal 2 Hypohyal 4 Urohyal 1 Glossihyal 1

190

APPENDIX-IX cont'd

Anatomic region/ Expected no. Skeletal element-Cranial Opercular Series (n=8) Peropercular 2 Opercular 2 Interopercular 2 Subopercular 2 Branchial Arch (n=40) pharyngial plate 4 Epibranchial 6 Ceratobranchial 8 Hypobranchial 8 Basibranchial 8 Pharyngobranchial 6 Postcranial Appendicular skeleton (n=16) Pelvis 2 Pelvic Spines 2 Cleithrum 2 Coracoid 2 D.Postcleithrum 2 V.Postcleithrum 2 Scapula 2 Supracleithrum 2 Vertebrae median Fin (n=?) Total pterygiophgore 1st anal pterygiophore 1

191

APPENDIX-X: Frequency (NISP) of skeletal elements for loci 2, 3, and 9.

Brush Hut 2: Fish remains Cyprinid Anatomic region/ Total Small Large Cichlids skeletal element-Cranial NISP NISP NISP NISP Oromandibular region Maxilla 3 0 2 1 Branchial arch Pharyngeal bone 3 0 3 0 Pharyngeal molariformteeth 3 0 3 0 Pharyngeal teeth 20 0 20 0 Postcranial Appendicular skeleton Scapula 2 0 1 1 Median fin Fin ray 8 0 8 0 Fin Spine 6 0 0 6 Pterygiophore 5 1 3 1 Rib 3 0 3 0 Total 53 1 43 9

Locus 9: Fish remains Cyprinidae Anatomic region/ Total Acanthobrama Small Large Cichlids Unident. Skeletal elements NISP NISP NISP NISP NISP NISP Cranial: Branchial arch Pharyngeal molariformteeth 2 0 0 2 0 0 Pharyngeal teeth 11 0 2 9 0 0 Postcrania/ Vertebral column Axis 1 0 0 0 1 0 Third Thoracic Vert. 1 1 0 0 0 0 Fourth Thoracic vert. 1 0 0 1 0 0 Fifth Thoracic vertebrae 1 0 1 0 0 0 Thoracic Vert. 10 0 6 0 4 0 Precaudal/Caudal Vert. 13 0 3 5 5 0 Caudal Vert. 5 0 1 4 0 0 Penultimate Caudal Vert. 1 0 1 0 0 0 Ultimate vert. 1 0 0 0 1 0 Vertebrae 3 0 0 0 0 3 Vertebrae fragment 20 0 0 0 0 20 Total 70 1 14 21 11 23

192

APPENDIX-X cont'd

Brush Hut-3: Fish remains Cyprinidae Anatomic region/ Total Small Large Cichlids uniden. Skeletal element- Cranial NISP NISP NISP NISP NISP Neurocranium Basioccipital 7 0 0 7 0 Oromandibular region Angular/Articular 6 0 6 0 0 Dentary 1 0 0 1 0 Maxilla 1 0 0 1 0 Hyoid Region Hypohyal 4 1 3 0 0 Branchial arch Pharyngeal bone 4 0 3 1 0 Molariformteeth 6 0 6 0 0 Pharyngeal teeth 20 0 20 0 0 Postcranial Scapula 1 0 1 0 0 Supracleithrum 1 0 0 1 0 Pelvis 1 0 0 1 0 Vertebral Column Atlas 43 1 36 6 0 Axis 20 0 4 16 0 4th vert 2 0 2 0 0 5th vert 9 0 9 0 0 Thoracic vert. 211 1 78 132 0 Caudal vert. 76 5 57 14 0 Penultimate vert. 16 0 9 7 0 Ultimate vert. 4 0 4 0 0 Precaudal vert. 7 0 2 5 0 Precaudal/Caudal vert. 149 1 71 75 2 Vertebrae 2 0 0 0 2 Vertebrae fragment 329 0 1 0 328 Others 1st Dorsal Pterygio 1 0 1 0 0 Dorsal Pterygiophore 1 0 1 0 0 Fin Spine 5 0 0 5 0 Pterygiophore 5 2 3 0 0 Rib 12 0 12 0 0 Weberian apparatus Weberian apparatus 1 0 1 0 0 Basapophysis 1 0 1 0 0 Os suspensorium 2 1 1 0 0 Total 948 12 332 272 332

193

APPENDIX-XI: Frequency (NISP) of skeletal elements for locus 1 by taxa (ranking order is marked for the most abundant bones in each group).

Brush Hut-1: Fish remains Cyprinidae Anatomic region/ Total Acanthobrama Small Large Cichlidae Uniden. Skeletal element-Cranial NISP NISP NISP NISP NISP NISP Olfactory region Vomer 4 0 1 0 3 0 Prefrontal 3 0 1 2 0 0 Orbital region Frontal 39 25 13 1 0 0 Lacrymal 2 0 0 0 2 0 Otic region Epiotic 3 0 3 0 0 0 Exoccipital 2 0 0 0 2 0 Parietal 53 41 12 0 0 0 Posttemporal 121 111 (8) 9 1 0 0 Pterotic 39 5 33 1 0 0 Prootic 2 0 2 0 0 0 Sphenotic 1 0 1 0 0 0 Supraoccipital 4 0 2 0 2 0 Otolith 1 0 0 1 0 0 Basicranial region Basioccipital 72 4 63 1 4 0 Parasphenoid 12 0 12 0 0 0

Cranial bone-gen 369 0 353 (6) 0 0 16 Neurocranium-gen 3 0 2 0 0 1 Oromandibular region Premaxilla 19 12 5 2 0 0 Maxilla 4 0 3 1 0 0 Dentary 96 75 14 5 2 0 Angular/Articular 66 43 6 14 (7) 3 0 Palatine 67 62 2 3 0 0 Entopterygoid 1 0 0 1 0 0 Quadrate 85 60 18 3 4 0 Hyoid region Hyomandibular 110 94 (9) 14 0 1 1 Hypohyal 4 0 4 0 0 0 Interhyal 1 0 1 0 0 0 Epihyal 86 48 35 3 0 0 Ceratohyal 281 272 (3) 8 1 0 0 Urohyal 51 42 3 1 5 0 Hyoid app. (general) 7 0 1 3 0 3 Opercular series Opercle 157 144 (7) 12 0 1 0 Opercle app. 2 2 0 0 0 0 Preopercle 27 0 23 1 3 0 Interopercle 1 0 0 1 0 0

194

APPENDIX-XI cont'd

Brush Hut-1: Fish remains Cyprinidae Anatomic region/ Total Acanthobrama Small Large Cichlidae Uniden. Skeletal element-Cranial NISP NISP NISP NISP NISP NISP Branchial arch Ceratobranchial 1 0 1 0 0 0 Pharyngeal bone 326 261 (4) 59 2 4 0 Molariformteeth 7 0 0 7 0 0 Pharyngeal teeth 340 164 105 71 0 0 Branchial region-gen. 31 0 29 0 0 2 Postcranial Appendicular skeleton Pelvis 102 79 18 2 3 0 Cleithrum 52 1 43 0 3 5 Coracoid 96 44 52 0 0 0 Mesocoracoid 13 5 6 2 0 0 Scapula 424 0 419 (5) 5 0 0 Supracleithrum 3 0 0 0 3 0 Vertebral column Atlas 476 387 (1) 17 36 (5) 36 (6) 0 Atlas/Axis 43 24 13 1 5 0 Axis 278 224 (6) 3 44 (4) 7 0 3rd vert.. 239 236 (5) 0 0 3 0 4th vert. 207 3 202 (8) 2 0 0 5th vert. 335 284 (2) 44 7 0 0 6th vert. 54 0 54 0 0 0 Thoracic vert. 2500 2 2156 (1) 71 (2) 270 (1) 1 Precaudal vert. 1187 4 1105 (2) 18 (6) 60 (3) 0 Precaudal/Caudal vert. 872 23 597 (3) 114 (1) 138 (2) 0 Caudal vert. 528 0 431 (4) 50 (3) 47 (4) 0 Penultimate vert. 91 1 71 6 13 (8) 0 Ultimate vert. 41 0 22 1 18 (7) 0 Vertebrae gen. 253 0 7 1 28 217 Vertebrae fragment 1169 0 456 31 7 675 Tail vert.-Epural 14 0 14 0 0 0 Tail vert.elements 25 0 24 1 0 0 Median Fin Ribs 63 0 55 7 1 0 Fin ray 182 0 178 (9) 3 1 0 Fin Spine 45 0 0 5 40 (5) 0 1st Dorsal Pterygio 56 46 8 2 0 0 Dorsal Pterygiophore 135 0 133 2 0 0 Pterygiophore 154 0 143 2 9 0 Weberian apparatus Tripus 283 0 270 (7) 13 0 0 Basapophysis 47 0 47 0 0 0 Basapophysis 3rd vert. 1 0 0 1 0 0 Weberian apparatus 3 0 0 3 0 0 Spina neuralis 3rd vert. 23 0 23 0 0 0 Os suspensorium 103 89 (10) 11 3 0 0 Total 12597 2917 7472 559 728 921

195

APPENDIX-XII: Frequency (NISP) of skeletal elements for locus 7 by taxa. Locus-7: Fish remains Cyprinidae Anatomic regions/ Total Acanthobrama Small Large Cichlid uniden. Skeletal elements- Cranial NISP NISP NISP NISP NISP NISP Olfactory Region Mesoethmoid 1 0 1 0 0 0 Vomer 1 0 0 0 1 0 Orbital Region Frontal 1 0 0 1 0 0 Otic Region Posttemporal 6 0 1 1 4 0 Otolith 2 0 0 0 2 0 Basicranial Region Basioccipital 11 1 1 0 9 0 Parasphenoid 1 0 0 1 0 0 Oromandibular Region Premaxilla 6 0 0 0 6 0 Maxilla 9 0 1 4 4 0 Angular/Articular 21 0 0 16 5 0 Dentary 22 0 2 19 1 0 Palatine 1 0 0 0 1 0 Entopterygoid 1 0 0 1 0 0 Quadrate 22 0 1 7 14 0 Hyoid Region Hyomandibular 6 0 0 2 4 0 Epihyal 9 0 0 4 5 0 Ceratohyal 6 0 0 1 5 0 Glossihyal 26 0 0 0 0 26 Urohyal 14 0 1 8 5 0 Opercular Series Opercle 14 0 0 1 13 0 Preopercle 6 0 0 1 4 1 Subopercle 1 0 0 0 0 1 Branchial arch Branchial region 11 0 0 3 0 8 Pharyngeal bone 46 21 6 9 10 0 Molariformteeth 12 0 0 12 0 0 Pharyngeal teeth 142 0 17 125 0 0

196

APPENDIX-XII cont'd

Locus-7: Fish remains Cyprinidae Anatomic regions/ Total Acanthobrama Small Large Cichlid uniden. Skeletal elements- postcranial NISP NISP NISP NISP NISP NISP Appendicular Skeleton Cleithrum 11 0 4 3 4 0 Coracoid 1 1 0 0 0 0 Dorsal Postcleithrum 1 0 0 0 1 0 Ventral Postcleithrum 13 0 0 0 13 0 Mesocoracoid 1 0 0 1 0 0 Supracleithrum 19 0 1 0 18 0 Scapula 8 0 1 3 4 0 Pelvis 10 0 1 6 3 0 Pelvic Spine 28 0 0 0 28 0 Weberian Apparatus Os suspensorium 8 1 3 4 0 0 Spina neuralis 2 0 1 1 0 0 Basapophysis 7 0 6 1 0 0 Vertebrae Atlas 69 1 2 27 39 0 Axis 64 5 7 10 42 0 3rd vert. 30 28 0 0 2 0 4th vert. 41 1 36 4 0 0 5th vert. 75 58 10 7 0 0 6th vert. 4 0 0 4 0 0 Thoracic vert. 1644 0 1033 54 553 4 Precaudal vert. 74 0 12 6 55 1 Precaudal/Caudal vert. 1998 0 270 77 436 1215 Caudal vert. 154 0 39 42 65 8 Penultimate vert. 30 1 5 3 21 0 Ultimate vert. 23 0 0 3 20 0 Tail vert.-elements 2 0 2 0 0 0 Vertebrae 95 0 0 0 1 94 Vertebrae fragment 20 0 2 4 4 10 Median Fin Fin ray 8 0 4 4 0 0 Fin Spine 423 0 0 9 414 0 1st Dorsal Pterygiophore 15 1 3 4 7 0 Rib 137 0 15 30 92 0 Dorsal Pterygiophore 10 0 0 10 0 0 Pterygiophore 55 0 32 6 17 0 Total 5478 119 1520 539 1932 1368

197

APPENDIX-XIII: Frequency (NISP) of skeletal elements for locus 8 by taxa. Cyprinidae Locus 8: Fish remains

Anatomic regions/ Total Acanthobrama Small Large Cichlids unident. Skeletal element-Cranial NISP NISP NISP NISP NISP NISP Olfactory region Vomer 4 0 0 0 4 0 Otic region Supraoccipital 1 0 1 0 0 0 Pterotic 1 0 1 0 0 0 Otolith 6 0 0 0 6 0 Oromandibular region Angular/Articular 9 0 1 4 4 0 Dentary 13 0 2 10 1 0 Maxilla 12 0 2 7 3 0 Premaxilla 4 0 0 0 4 0 Quadrate 2 0 0 2 0 0 Hyoid region Hyomandibular 2 0 0 1 1 0 Epihyal 5 0 0 4 1 0 Ceratohyal 2 0 0 0 2 0 Glossihyal 1 0 0 0 0 1 Opercular series Opercle 13 0 0 0 13 0 Preopercle 6 0 4 1 1 0 Interopercle 1 0 0 1 0 0 Branchial arch Pharyngeal bone 14 0 1 4 9 0 Molariformteeth 6 0 0 6 0 0 Pharyngeal teeth 39 0 2 36 1 0

198

APPENDIX-XIII cont'd

Cyprinidae Locus 8: Fish remains

Anatomic regions/ Total Acanthobrama Small Large Cichlids unident. Skeletal element-Postcranial NISP NISP NISP NISP NISP NISP Appendicular skeleton Cleithrum 3 0 0 2 1 0 Mesocoracoid 2 0 0 2 0 0 Pelvis 8 0 0 6 2 0 Pelvic Spine 5 0 0 0 5 0 Scapula 4 0 0 3 1 0 Supracleithrum 9 0 0 0 9 0 Ventral Postcleithrum 2 0 0 0 2 0 Median Fin 1st Dorsal Pterygio 5 0 1 2 2 0 Dorsal Pterygiophore 5 0 1 4 0 0 Pterygiophore 6 0 1 1 4 0 Fin ray 8 0 0 8 0 0 Fin Spine 38 0 0 3 35 0 Rib 7 0 0 7 0 0 Vertebrae Atlas 19 0 0 10 9 0 Axis 11 0 0 4 7 0 3rd vert. 1 1 0 0 0 0 4th vert. 1 0 0 1 0 0 5th vert. 1 0 0 1 0 0 Thoracic vert. 141 0 1 36 104 0 Precaudal vert. 10 0 1 0 9 0 Precaudal/Caudal vert. 47 0 3 18 26 0 Caudal vert. 46 0 2 25 19 0 Penultimate vert. 3 0 1 2 0 0 Ultimate vert. 3 0 0 2 1 0 Vertebrae 2 0 0 0 2 0 Vertebrae fragment 73 0 0 3 0 70 Weberian apparatus Basapophysis 1 0 0 1 0 0 Os suspensorium 6 0 1 5 0 0 Total 608 1 26 222 288 71

199

APPENDIX-XIV: Skeletal elements fragmentation pattern in locus 1.

Brush Hut-1 Fragmentation (%) Total Skeletal element 5-29 30-50 51-70 71-89 90-100 NISP Angular/Articular 15.2 30.3 9.1 6.1 39.4 66 Atlas 1.5 .8 5.2 64.1 28.4 476 Axis 1.1 7.5 4.3 66.9 20.1 278 3rd vert. .8 1.7 15.8 41.5 40.2 241 4th vert. .5 2.9 8.2 28.5 59.9 207 5th vert. .6 2.1 21.2 23.3 52.8 335 6th vert. 0.0 0.0 0.0 98.2 1.9 54 Basapophysis 0.0 2.1 0.0 6.4 91.5 47 Basapophysis 3rd vert. 0.0 0.0 0.0 100.0 0.0 1 Basioccipital 1.4 29.2 45.8 20.8 2.8 72 Branchial region 25.8 0.0 0.0 0.0 74.2 31 Caudal Vert. 1.1 1.5 5.3 66.1 25.9 528 Ceratobranchial 0.0 100.0 0.0 0.0 0.0 1 Ceratohyal 2.1 22.1 8.5 12.5 54.8 281 Cleithrum 46.1 50.0 3.9 0.0 0.0 52 Coracoid 12.5 62.5 19.8 4.2 1.0 96 Cranial bone .8 99.2 0.0 0.0 0.0 369 Dentary 8.3 46.9 30.2 7.3 7.3 96 Dorsal Pterygiophore 37.0 28.1 28.9 1.5 4.4 135 Entopterygoid 0.0 0.0 100.0 0.0 0.0 1 Epihyal 1.2 9.3 19.8 16.3 53.5 86 Epiotic 0.0 0.0 0.0 0.0 100.0 3 Exoccipital 0.0 0.0 50.0 0.0 50.0 2 Fin ray 16.9 38.0 5.6 36.6 2.8 71 Fin Spine 13.3 31.1 13.3 28.9 13.3 45 1st Dorsal Pterygiophore 26.8 33.9 16.1 14.3 8.9 56 Frontal 7.7 53.9 25.6 7.7 5.1 39 Hyoid app. 0.0 28.6 14.3 14.3 42.9 7 Hyomandibular 65.5 26.4 4.5 2.7 .9 110 Hypohyal 25.0 0.0 0.0 0.0 75.0 4 Interhyal 0.0 0.0 0.0 0.0 100.0 1 Interopercle 0.0 0.0 100.0 0.0 0.0 1 Lacrymal 0.0 0.0 0.0 100.0 0.0 2 Maxilla 0.0 50.0 0.0 50.0 0.0 4 Mesocoracoid 0.0 15.4 7.7 7.7 69.2 13 Neurocranium 33.3 0.0 0.0 66.7 0.0 3 Opercle 22.3 31.2 17.2 26.1 3.2 157 Opercle app. 0.0 0.0 100.0 0.0 0.0 2 Os suspensorium 4th vert. 7.8 37.9 21.4 17.5 15.5 103 Otolith 0.0 0.0 0.0 0.0 100.0 1 Palatine 0.0 6.0 22.4 20.9 50.8 67 Parasphenoid 10.0 60.0 20.0 10.0 0.0 10

200

APPENDIX-XIV cont'd

Brush Hut-1 Fragmentation (%) Total Skeletal element 5-29 30-50 51-70 71-89 90-100 NISP Parietal 18.9 30.2 7.5 13.2 30.2 53 Pelvis 34.3 42.2 15.7 5.9 2.0 102 Penultimate Vert. 0.0 1.1 1.1 47.2 50.5 91 Pharyngeal bone 26.1 40.0 14.8 17.5 1.5 325 Pharyngeal teeth 45.3 54.7 0.0 0.0 0.0 53 Posttemporal .8 36.4 34.7 10.7 17.4 121 Precaudal vert. .5 1.5 10.3 82.9 4.8 1160 Precaudal/Caudal Vert. 1.6 2.9 8.6 69.5 17.4 872 Prefrontal 66.7 0.0 0.0 33.3 0.0 3 Premaxilla 47.4 42.1 10.5 0.0 0.0 19 Preopercle 25.9 40.7 29.6 3.7 0.0 27 Prootic 0.0 0.0 0.0 50.0 50.0 2 Pterotic 5.1 66.7 15.4 10.3 2.6 39 Pterygiophore 33.1 37.7 11.0 14.3 3.9 154 Quadrate 10.6 14.1 35.3 20.0 20.0 85 Rib 52.4 38.1 4.8 4.8 0.0 63 Scapula 1.2 10.8 1.4 .9 85.6 424 Sphenotic 0.0 0.0 100.0 0.0 0.0 1 Spina neuralis 0.0 13.0 8.7 0.0 78.3 23 Supracleithrum 0.0 0.0 66.7 0.0 33.3 3 Supraoccipital 0.0 25.0 25.0 25.0 25.0 4 Tail vert.-Epural 0.0 0.0 0.0 0.0 100.0 14 Tail vertebrae-elements 0.0 0.0 0.0 64.0 36.0 25 Thoracic Vert. 1.2 3.7 12.0 78.8 4.4 2541 Tripus .7 3.9 10.2 61.1 24.0 283 Ultimate vert. 2.4 0.0 9.8 73.2 14.6 41 Urohyal 5.9 37.2 17.6 13.7 25.5 51 Vertebrae 0.0 0.0 32.0 66.4 1.6 253 Vertebrae fragment 65.4 26.2 8.4 0.0 0.0 1169 Vomer 25.0 0.0 25.0 0.0 50.0 4 Weberian apparatus 0.0 0.0 33.3 0.0 66.7 3 Total 11.4 15.1 11.3 45.2 16.9 12162

201

APPENDIX-XV: WMI of fragmentation calculated by taxa for bones from locus 1.

L.1 Small Cyprinid (n=7255) Skeletal element WMI Bone WMI Angular/Articular 48.33 Parasphenoid 57.00 Atlas 75.59 Parietal 70.83 Axis 51.67 Pelvis 38.33 Basapophysis 93.30 Penultimate Vert. 90.92 Basioccipital 66.27 Pharyngeal bone 29.14 Branchial region 80.52 Pharyngeal teeth 24.26 Caudal Vert. 86.11 Posttemporal 60.56 Ceratobranchial 45.00 Precaudal vert. 83.31 Ceratohyal 63.75 Precaudal/Caudal Vert. 84.13 Cleithrum 38.26 Prefrontal 85.00 Coracoid 47.31 Premaxilla 35.00 Cranial bone 44.83 Preopercle 47.17 Dentary 52.14 Prootic 90.00 Dorsal Pterygiophore 49.81 Pterotic 58.33 Epihyal 66.14 Pterygiophore 48.78 Epiotic 95.00 Quadrate 65.56 5th Vertebrae 78.64 Rib 33.36 Fin ray 58.88 Scapula 89.37 1st Dorsal Pterygiophore 27.50 6th Vertebrae 85.00 4th vert. 87.57 Sphenotic 65.00 Frontal 51.92 Spina neuralis 86.30 Hyoid app. 95.00 Supraoccipital 90.00 Hyomandibular 24.29 Tail vert.-Epural 95.00 Hypohyal 72.50 Tail vert. elements 87.50 Interhyal 95.00 Thoracic Vert. 80.37 Maxilla 68.33 Tripus 83.96 Mesocoracoid 76.67 Ultimate Vert. 81.36 Neurocranium 55.00 Urohyal 35.00 Opercle 60.00 Vertebrae 77.86 Os suspensorium 4th vert. 54.09 Vertebrae fragment 40.35 Palatine 70.00 Vomer 95.00

L.1 Unidentified Fish (n=905) Skeletal element WMI Bone WMI Branchial region 15.00 Neurocranium 85.00 Cleithrum 25.00 Thoracic Vert. 65.00 Cranial bone 55.00 Vertebrae 76.49 Hyoid app. 45.00 Vertebrae fragment 21.75 Hyomandibular 15.00

202

APPENDIX-XV cont'd

L.1 Acanthobrama sp. (n=2753) Skeletal element WMI Bone WMI Angular/Articular 75.93 Opercle app. 65.00 Atlas 87.12 Os suspensorium 4th vert. 66.57 Axis 84.20 Palatine 85.16 Basioccipital 80.00 Parietal 62.07 Ceratohyal 80.51 Pelvis 44.75 Cleithrum 45.00 Penultimate Vert. 85.00 Coracoid 50.00 Pharyngeal bone 55.534 Dentary 58.33 Posttemporal 70.31 Epihyal 93.12 Precaudal vert. 87.50 5th Vertebrae 85.63 Precaudal/Caudal Vert. 85.00 1st Dorsal Pterygiophore 57.39 Premaxilla 33.33 4th Vertebrae 91.67 Pterotic 41.00 Frontal 55.40 Quadrate 71.67 Hyomandibular 31.91 3rd Vertebrae 84.41 Mesocoracoid 89.00 Thoracic Vert. 82.50 Opercle 56.81 Urohyal 70.00

L.1 Barbus/Capoeta (n=505)

Skeletal element WMI Bone WMI Angular/Articular 50.00 Pelvis 20.00 Atlas 70.00 Penultimate Vert. 78.33 Axis 77.27 Pharyngeal bone 35.00 Basapophysis 3rd Vert. 85.00 Pharyngeal teeth 53.08 Basioccipital 65.00 Posttemporal 55.00 Caudal Vert. 79.80 Precaudal vert. 78.75 Ceratohyal 25.00 Precaudal/Caudal Vert. 74.30 Dentary 61.00 Prefrontal 15.00 Dorsal Pterygiophore 35.00 Premaxilla 30.00 Entopterygoid 65.00 Preopercle 55.00 Epihyal 91.67 Pterotic 95.00 5th Vetebrae 62.14 Pterygiophore 50.00 Fin ray 65.00 Quadrate 48.33 Fin Spine 25.00 Rib 37.86 1st Dorsal Pterygiophore 70.00 Scapula 39.00 4th Vertebrae 95.00 Tail vert. elements 95.00 Frontal 65.00 Thoracic Vert. 72.64 Hyoid app. 91.67 Tripus 74.23 Interopercle 75.00 Ultimate vert. 85.00 Maxilla 45.00 Urohyal 95.00 Mesocoracoid 90.00 Vertebrae 65.00 Os suspensorium 4th vert. 85.00 Vertebrae fragment 25.32 Otolith 95.00 Weberian apparatus 85.00 Palatine 71.67

APPENDIX-XV cont'd

L.1 Cichlid (n=728) Skeletal element WMI Bone WMI Angular/Articular 61.67 Precaudal vert. 80.00 Atlas 85.28 Precaudal/Caudal Vert. 80.65

203

Axis 85.00 Preopercle 41.67 Basioccipital 77.50 Pterygiophore 56.11 Caudal Vert. 82.23 Quadrate 45.00 Cleithrum 31.67 Rib 85.00 Dentary 15.00 Supracleithrum 75.00 Exoccipital 85.00 Supraoccipital 65.00 Fin ray 95.00 3rd Vertebrae 85.00 Fin Spine 68.00 Thoracic Vert. 81.29 Hyomandibular 25.00 Ultimate vert. 83.89 Lacrymal 85.00 Urohyal 55.00 Opercle 45.00 Vertebrae 79.64 Pelvis 45.00 Vertebrae fragment 36.43 Penultimate Vert. 83.46 Vomer 65.00 Pharyngeal bone 67.50

204

APPENDIX-XVI: Skeletal elements fragmentation pattern in locus 7 (ashes).

Locus 7 Fragmentation (%) Total Skeletal element 05-29 30-50 51-70 71-89 90-100 NISP Angular/Articular 33.33 19.05 0.00 14.29 33.33 21 Atlas 0.00 7.25 15.94 47.83 28.99 69 Axis 3.12 4.69 15.62 53.12 23.44 64 3rd vert. 0.00 0.00 0.00 23.33 76.67 30 4th vert. 0.00 0.00 4.88 92.68 2.44 41 5h vert. 0.00 5.33 4.00 53.33 37.33 75 6h vert. 0.00 0.00 0.00 100.00 0.00 4 Basapophysis 0.00 28.57 0.00 0.00 71.43 7 Basioccipital 0.00 9.09 0.00 72.73 18.18 11 Branchial region 0.00 0.00 0.00 0.00 100.00 11 Caudal Vert. 0.00 0.00 3.25 94.81 1.95 154 Ceratohyal 0.00 33.33 16.67 0.00 50.00 6 Cleithrum 45.45 54.55 0.00 0.00 0.00 11 Coracoid 0.00 0.00 100.00 0.00 0.00 1 Dentary 4.55 27.27 40.91 22.73 4.55 22 Dorsal Postcleithrum 0.00 0.00 100.00 0.00 0.00 1 Dorsal Pterygiophore 40.00 40.00 10.00 0.00 10.00 10 Entopterygoid 0.00 100.00 0.00 0.00 0.00 1 Epihyal 0.00 0.00 33.33 55.56 11.11 9 Fin ray 12.50 50.00 12.50 25.00 0.00 8 Fin Spine 17.02 39.95 9.46 23.40 10.17 423 1st Dorsal Pterygiophore 20.00 33.33 20.00 26.67 0.00 15 Frontal 0.00 100.00 0.00 0.00 0.00 1 Glossihyal 0.00 0.00 0.00 0.00 100.00 26 Hyomandibular 33.33 50.00 16.67 0.00 0.00 6 Maxilla 55.56 22.22 11.11 11.11 0.00 9 Mesocoracoid 0.00 0.00 0.00 0.00 100.00 1 Mesoethmoid 0.00 0.00 0.00 0.00 100.00 1 Opercle 28.57 42.86 7.14 21.43 0.00 14 Os suspensorium 4th vert. 0.00 12.50 25.00 50.00 12.50 8 Otolith 0.00 0.00 0.00 0.00 100.00 2 Palatine 0.00 0.00 0.00 100.00 0.00 1 Parasphenoid 0.00 0.00 0.00 100.00 0.00 1 Pelvic Spine 3.57 32.14 28.57 28.57 7.14 28 Pelvis 40.00 60.00 0.00 0.00 0.00 10 Penultimate vert. 0.00 0.00 3.33 56.67 40.00 30 Pharyngeal bone 36.96 21.74 23.91 2.17 15.22 46 Posttemporal 0.00 33.33 0.00 66.67 0.00 6 Precaudal vert. 0.00 2.70 8.11 79.73 9.46 74 Precaudal/Caudal vert. .15 .50 3.35 95.65 .35 1998

205

APPENDIX-XVI cont'd

Locus 7 Fragmentation (%) Total

Skeletal element 05-29 30-50 51-70 71-89 90-100 NISP Premaxilla 0.00 83.33 0.00 0.00 16.67 6 Preopercle 33.33 50.00 16.67 0.00 0.00 6 Pterygiophore 30.91 41.82 20.00 7.27 0.00 55 Quadrate 18.18 22.73 50.00 4.55 4.55 22 Rib 23.36 48.91 8.03 18.25 1.46 137 Scapula 25.00 25.00 0.00 50.00 0.00 8 Spina neuralis 0.00 50.00 0.00 50.00 0.00 2 Subopercle 0.00 0.00 100.00 0.00 0.00 1 Supracleithrum 15.79 52.63 21.05 5.26 5.26 19 Tail vertebrae-elements 0.00 0.00 0.00 50.00 50.00 2 Thoracic vert. .91 1.58 3.41 85.52 8.58 1644 Ultimate vert. 0.00 0.00 8.70 34.78 56.52 23 Urohyal 50.00 14.29 28.57 7.14 0.00 14 Ventral Postcleithrum 30.77 46.15 23.08 0.00 0.00 13 Vertebrae unident. 0.00 0.00 4.21 95.79 0.00 95 Vertebrae fragment 40.00 60.00 0.00 0.00 0.00 20 Vomer 0.00 100.00 0.00 0.00 0.00 1 Total Count 225 431 297 3981 390 5324

206

APPENDIX-XVII:WMI of fragmentation calculated by taxa for bones from locus 7.

L.7 Acanthobrama sp. (n=119) Skeletal element WMI NISP Skeletal element WMI NISP Atlas 85.00 119 1st Dorsal Pterygiophore 85.00 58 Axis 93.00 1 Os suspensorium 4h vert.. 85.00 1 4h vert. 85.00 1 Penultimate vert. 85.00 1 5h vert. 87.93 1 Pharyngeal bone 58.33 1 Basioccipital 55.00 5 3rd vert. 93.21 21 Coracoid 75.00 1

L.7 Unidentified (n=1368) Skeletal element WMI NISP Skeletal element WMI NISP Branchial region 95.00 8 Precaudal vert. 85.00 1 Glossihyal 95.00 26 Precaudal/Caudal vert. 84.72 1215 Preopercle 25.00 1 Thoracic vert. 82.50 4 Subopercle 65.00 1 Vertebrae 84.47 94 Caudal vert. 85.00 8 Vertebrae fragment 42.00 10

L.7 Small cyprinid (n=1503) Skeletal element WMI NISP Skeletal element WMI NISP Atlas 75.00 2 Pharyngeal bone 25.00 6 Axis 83.57 7 Posttemporal 55.00 1 4h vert. 84.44 36 Precaudal vert. 85.00 12 5h vert. 74.00 10 Precaudal/Caudal Vert. 84.30 270 Basapophysis 78.33 6 Pterygiophore 42.19 32 Basioccipital 85.00 1 Quadrate 45.00 1 Caudal Vert. 85.00 39 Rib 48.33 15 Cleithrum 35.00 4 Scapula 15.00 1 Dentary 60.00 2 Spina neuralis 55.00 1 Fin ray 52.50 4 Supracleithrum 55.00 1 1st Dorsal Pterygiophore 38.33 3 Tail vertebrae-elements 90.00 2 Maxilla 15.00 1 Thoracic vert. 84.79 1033 Mesoethmoid 95.00 1 Urohyal 15.00 1 Os suspensorium 4h vert. 85.00 3 Vertebrae fragment 25.00 2 Pelvis 45.00 1 Penultimate vert. 93.00 5

207

APPENDIX-XVII cont'd

L.7 Large cyprinids (n=402) L.7 Cichlid (n=1932) skeletal element WMI NISP Skeletal element WMI NISP Angular/Articular 66.88 16 Angular/Articular 33.00 5 Atlas 76.11 27 Atlas 86.03 39 Axis 62.00 10 Axis 82.14 42 4h vert. 80.00 4 3rd vert. 85.00 2 5h vert. 69.29 7 Basioccipital 86.11 9 6h vert. 85.00 4 Caudal Vert. 84.54 65 Basapophysis 95.00 1 Ceratohyal 75.00 5 Branchial region 95.00 3 Cleithrum 30.00 4 Caudal Vert. 83.33 42 Dentary 75.00 1 Ceratohyal 95.00 1 Dorsal Postcleithrum 75.00 1 Cleithrum 31.67 3 Epihyal 79.00 5 Dentary 65.53 19 Fin Spine 61.18 414 Dorsal Pterygiophore 41.00 10 1st Dorsal Pterygiophore 62.14 7 Entopterygoid 55.00 1 Hyomandibular 35.00 4 Epihyal 87.50 4 Maxilla 55.00 4 Fin ray 65.00 4 Opercle 48.08 13 Fin Spine 31.67 9 Otolith 95.00 2 1st Dorsal Pterygiophore 55.00 4 Palatine 85.00 1 Frontal 55.00 1 Pelvic Spine 66.43 28 Hyomandibular 45.00 2 Pelvis 18.33 3 Maxilla 22.50 4 Penultimate vert. 87.38 21 Mesocoracoid 95.00 1 Pharyngeal bone 64.00 10 Opercle 75.00 1 Posttemporal 72.50 4 Os suspensorium 4th vert. 67.50 4 Precaudal vert. 83.73 55 Parasphenoid 85.00 1 Precaudal/Caudal vert. 83.23 436 Pelvis 35.00 6 Premaxilla 53.33 6 Penultimate vert. 91.67 3 Preopercle 40.00 4 Pharyngeal bone 30.56 9 Pterygiophore 59.12 17 Posttemporal 85.00 1 Quadrate 62.86 14 Precaudal vert. 81.67 6 Rib 60.87 92 Precaudal/Caudal Vert. 76.82 77 Scapula 85.00 4 Preopercle 45.00 1 Supracleithrum 53.89 18 Pterygiophore 25.00 6 Thoracic vert. 83.57 553 Quadrate 45.00 7 Ultimate vert. 90.00 20 Rib 33.33 30 Urohyal 65.00 5 Scapula 38.33 3 Ventral Postcleithrum 44.23 13 Spina neuralis 85.00 1 Vertebrae 65.00 1 Thoracic Vert. 69.63 54 Vertebrae fragment 45.00 4 Ultimate vert. 81.67 3 Vomer 45.00 1 Urohyal 27.50 8 Vertebrae fragment 20.00 4

208

APPENDIX-XVIII: Skeletal elements fragmentation pattern in locus 8.

Locus 8 Fragmentation (%) Total

Skeletal element 5-29 30-50 51-70 71-89 90-100 NISP Angular/Articular 11.11 44.44 0.00 11.11 33.33 9 Atlas 5.26 5.26 5.26 42.11 42.11 19 Axis 27.27 0.00 0.00 54.55 18.18 11 3rd vert. 0.00 0.00 0.00 100.00 0.00 1 4h vert. 0.00 0.00 0.00 100.00 0.00 1 5h vert. 0.00 0.00 0.00 100.00 0.00 1 Basapophysis 0.00 0.00 100.00 0.00 0.00 1 Basihyal 0.00 0.00 0.00 0.00 100.00 1 Caudal Vert. 0.00 13.04 6.52 41.30 39.13 46 Ceratohyal 0.00 0.00 100.00 0.00 0.00 2 Cleithrum 0.00 100.00 0.00 0.00 0.00 3 Dentary 0.00 61.54 23.08 7.69 7.69 13 Dorsal Pterygiophore 20.00 20.00 40.00 0.00 20.00 5 Epihyal 0.00 0.00 0.00 0.00 100.00 5 Fin ray 100.00 0.00 0.00 0.00 0.00 1 Fin Spine 5.26 31.58 7.89 39.47 15.79 38 1st Dorsal Pterygiophore 20.00 40.00 0.00 0.00 40.00 5 Hyomandibular 100.00 0.00 0.00 0.00 0.00 2 Interopercle 0.00 100.00 0.00 0.00 0.00 1 Maxilla 33.33 50.00 16.67 0.00 0.00 12 Mesocoracoid 0.00 0.00 0.00 0.00 100.00 2 Opercle 7.69 15.38 23.08 46.15 7.69 13 Os suspensorium 4th vert. 16.67 33.33 16.67 16.67 16.67 6 Otolith 0.00 0.00 0.00 0.00 100.00 6 Pelvic Spine 0.00 20.00 0.00 60.00 20.00 5 Pelvis 87.50 0.00 12.50 0.00 0.00 8 Penultimate Caudal Vert. 0.00 0.00 33.33 33.33 33.33 3 Pharyngeal bone 21.43 64.29 7.14 0.00 7.14 14 Molariformteeth 0.00 100.00 0.00 0.00 0.00 4 Pharyngeal teeth 0.00 100.00 0.00 0.00 0.00 15 Precaudal vert. 0.00 10.00 10.00 60.00 20.00 10 Precaudal/Caudal Vert. 4.26 6.38 4.26 82.98 2.13 47 Premaxilla 0.00 25.00 0.00 25.00 50.00 4 Preopercle 0.00 66.67 33.33 0.00 0.00 6 Pterotic 0.00 100.00 0.00 0.00 0.00 1 Pterygiophore 16.67 16.67 16.67 16.67 33.33 6 Quadrate 100.00 0.00 0.00 0.00 0.00 2 Rib 57.14 28.57 14.29 0.00 0.00 7 Scapula 25.00 50.00 0.00 0.00 25.00 4 Supracleithrum 11.11 44.44 33.33 11.11 0.00 9 Supraoccipital 0.00 100.00 0.00 0.00 0.00 1

209

APPENDIX-XVIII cont'd

Locus 8 Fragmentation Total

Skeletal element 5-29 30-50 51-70 71-89 90-100 NISP Thoracic vert. 0.00 .71 6.38 75.89 17.02 141 Ultimate vert. 0.00 0.00 0.00 66.67 33.33 3 Ventral Postcleithrum 0.00 0.00 50.00 50.00 0.00 2 Vertebrae 0.00 0.00 100.00 0.00 0.00 2 Vertebrae fragment 90.41 9.59 0.00 0.00 0.00 73 Vomer 0.00 25.00 50.00 0.00 25.00 4 Total Count 105 106 48 222 94 575

210

APPENDIX-XIX: WMI of fragmentation calculated by taxa for bones from locus 8.

L. 8 Cichlid (n=288) Skeletal element WMI NISP Skeletal element WMI NISP Angular/Articular 67.50 4 Pelvis 20.00 2 Atlas 90.56 9 Pharyngeal bone 48.33 9 Axis 85.00 7 Pharyngeal teeth 55.00 1 Caudal vert. 80.26 19 Precaudal vert. 79.44 9 Ceratohyal 65.00 2 Precaudal/Caudal Vert. 83.85 26 Cleithrum 55.00 1 Premaxilla 75.00 4 Dentary 95.00 1 Preopercle 45.00 1 Epihyal 95.00 1 Pterygiophore 62.50 4 Fin Spine 72.43 35 Scapula 95.00 1 1st Dorsal Pterygiophore 95.00 2 Supracleithrum 53.89 9 Hyomandibular 15.00 1 Thoracic vert. 86.06 104 Maxilla 45.00 3 Ultimate vert. 85.00 1 Opercle 71.15 13 Ventral Postcleithrum 75.00 2 Otolith 95.00 6 Vertebrae 65.00 2 Pelvic Spine 81.00 5 Vomer 70.00 4

L.8 small cyprinid (n= 24) Skeletal element WMI NISP Skeletal element WMI NISP Angular/Articular 25.00 1 Pharyngeal bone 15.00 1 Caudal vert. 95.00 2 Precaudal vert. 95.00 1 Dentary 55.00 2 Precaudal/Caudal vert. 55.00 3 Dorsal Pterygiophore 95.00 1 Preopercle 52.50 4 1st Dorsal Pterygiophore 45.00 1 Pterotic 45.00 1 Maxilla 20.00 2 Pterygiophore 95.00 1 Os suspensorium 4th vert. 75.00 1 Supraoccipital 45.00 1 Penultimate Caudal Vert. 95.00 1 Thoracic vert. 85.00 1

L.8 Large Cyprinid (n=191) Skeletal element WMI NISP Skeletal element WMI NISP 4h vert. 85.00 1 Pharyngeal teeth 55.00 14 Hyomandibular 25.00 1 Precaudal/Caudal vert. 77.78 18 Interopercle 55.00 1 Preopercle 55.00 1 Maxilla 40.71 7 Pterygiophore 55.00 1 Mesocoracoid 95.00 2 Quadrate 20.00 2 Os suspensorium 4th vert. 55.00 5 Rib 29.29 7 Pelvis 31.67 6 Scapula 38.33 3 Penultimate Caudal Vert. 70.00 2 Thoracic vert. 81.39 36 Pharyngeal bone 35.00 4 Ultimate vert. 90.00 2 Molariformteeth 55.00 4 Vertebrae fragment 45.00 3

211

APPENDIX-XX: Vertebrae width dimensions mean (±SD) and range Calculated by taxa for locus 1.

Locus 1 Vertebrae Width (mm) Genus Group Vertebrae Count Mean Std. Dev. Min. Max. Acanthobrama sp. Atlas 371 3.06 .29 1.60 4.02 Axis 196 2.84 .30 1.86 3.71 3rd vert. 205 2.54 .28 1.37 3.35 4th vert. 3 2.66 .87 1.67 3.28 5th vert. 277 2.90 .37 1.50 4.00 Thoracic vert. 17 2.70 .32 1.80 3.16

Small Carps Atlas 10 2.94 .61 2.05 4.23 Axis 1 2.60 . 2.60 2.60 Caudal Vert. 372 2.17 .52 1.31 8.10 4th vert. 177 2.83 .37 1.59 4.33 5th vert. 27 2.75 .44 1.75 3.40 Thoracic vert. 1828 2.16 .31 .97 6.75 Precaudal vert. 69 2.35 .61 1.13 5.50 Precaudal/Caudal vert. 531 2.12 .54 .87 4.70 Penultimate vert. 68 2.25 .55 1.78 5.81 Ultimate vert. 17 2.20 .21 1.96 2.58

Large Carps Atlas 19 6.91 2.28 3.20 11.61 Axis 22 4.24 2.90 1.05 10.61 Caudal Vert. 44 6.14 1.08 3.82 9.02 5th vert. 4 6.54 .47 6.18 7.22 4th vert. 2 5.86 .53 5.48 6.23 Penultimate vert. 5 5.32 .68 4.54 6.08 Precaudal vert. 2 5.46 1.70 4.25 6.66 Precaudal/Caudal vert. 96 6.34 1.15 3.72 9.44 Thoracic vert. 56 6.02 1.06 4.04 9.39 Ultimate vert. 1 3.90 3.90 3.90

Cichlidae Atlas 25 3.30 .65 1.74 4.50 Axis 5 3.02 .84 1.74 3.86 3rd vert. 2 2.45 .04 2.42 2.48 Thoracic vert. 214 2.79 .82 .88 5.10 Precaudal vert. 48 2.56 .55 1.77 5.05 Precaudal/Caudal vert. 103 2.89 .75 1.42 6.44 Caudal vert. 37 2.50 .59 1.09 3.98 Penultimate vert. 11 2.43 .55 1.80 3.50 Ultimate vert. 12 2.56 .37 1.93 3.17

212

APPENDIX-XX, cont'd.

Locus 3 Vertebrae Width (mm) Genus Group Vertebrae Count Mean Std. Dev. Min Max. Cichlids Atlas 6 4.09 .44 3.66 4.63 Axis 16 3.48 .74 2.50 4.82 Caudal vert. 14 2.82 .40 2.17 3.64 Penultimate vert. 7 2.74 .55 1.58 3.29 Precaudal vert. 4 3.21 .68 2.60 3.84 Precaudal/Caudal vert. 18 3.08 .48 2.14 4.19 Thoracic vert. 120 2.99 .47 2.03 4.33

Large Carps Atlas 23 7.37 1.93 4.55 14.42 Axis 3 5.68 1.42 4.08 6.77 5th vert. 7 6.20 1.52 4.82 8.90 Thoracic vert. 60 5.67 .83 4.14 7.68 Precaudal vert. 2 5.71 .58 5.30 6.12 Precaudal/Caudal vert. 63 6.29 .88 4.39 9.25 Caudal vert. 53 6.02 .88 4.45 8.87 Penultimate vert. 9 5.73 1.26 4.44 8.33 Ultimate vert. 4 5.11 .40 4.64 5.62

Small Carps Caudal vert. 4 4.04 1.18 2.56 5.43 Locus-8 Acanthobrama sp. 3rd vert. 1 2.15 2.15 2.15

Cichlids Atlas 9 3.16 .70 2.22 4.00 Axis 6 3.42 .21 3.11 3.70 Caudal vert. 18 3.02 .47 1.95 3.64 Thoracic vert. 60 3.39 .60 1.93 4.70 Precaudal vert. 8 3.04 .52 2.13 3.63 Precaudal/Caudal vert. 8 2.62 .40 2.18 3.33 Ultimate vert. 1 2.89 2.89 2.89

Large Carps Atlas 6 7.67 .79 7.00 8.90 Axis 2 7.17 .81 6.60 7.74 4th vert. 1 7.68 7.68 7.68 5th vert. 1 7.14 7.14 7.14 Thoracic vert. 34 6.46 1.53 3.05 8.70 Penultimate vert. 2 7.79 3.73 5.15 10.42 Precaudal/Caudal vert. 17 6.89 1.04 5.25 9.15 Caudal vert. 24 6.04 .94 4.35 7.76 Ultimate vert. 2 5.75 .73 5.24 6.27

Small Carps Caudal vert. 2 2.79 .84 2.19 3.38 Thoracic vert. 1 2.56 2.56 2.56 Precaudal/Caudal vert. 3 3.14 .43 2.84 3.64 Penultimate vert. 1 3.71 3.71 3.71

213

APPENDIX-XX, cont'd. Locus-7 Vertebrae Width (mm) Genus Group Vertebrae Count Mean Std. Dev. Min Max. Acanthobrama sp. Atlas 1 1.57 1.57 1.57 Axis 5 1.41 .19 1.11 1.60 3rd vert. 20 1.49 .22 1.26 2.26 5th vert. 51 1.55 .20 1.00 2.00 Penultimate vert. 1 1.62 1.62 1.62

Cichlids Atlas 36 3.30 .70 1.90 4.70 Axis 37 2.90 .76 1.20 4.53 3rd vert. 1 4.00 4.00 4.00 Thoracic vert. 483 3.13 .85 .80 5.18 Precaudal vert. 46 2.50 .78 1.15 4.56 Caudal vert. 53 2.64 .43 1.65 3.85 Penultimate vert. 19 2.47 .60 .97 3.18 Ultimate vert. 19 2.77 .35 2.00 3.40 Large Carps Atlas 16 7.43 1.80 5.60 12.80 Axis 6 6.23 2.54 1.34 8.32 4th vert. 4 7.92 1.12 7.20 9.60 5th vert. 5 3.96 2.96 1.60 8.66 6th vert. 4 4.82 .36 4.55 5.35 Thoracic vert. 37 5.57 .72 4.00 8.08 Precaudal vert. 2 6.17 .18 6.04 6.29 Caudal vert. 22 5.21 .96 3.94 8.30 Penultimate vert. 3 5.99 .67 5.38 6.70 Ultimate vert. 1 4.56 4.56 4.56

Small Carps Atlas 1 1.38 1.38 1.38 Axis 5 1.37 .32 .92 1.80 4th vert. 27 1.49 .19 1.15 1.90 5th vert. 8 2.82 2.44 1.24 7.00 Thoracic vert. 97 1.35 .34 .90 2.98 Precaudal vert. 9 1.18 .19 1.00 1.52 Precaudal/Caudal vert. 22 1.56 .75 1.20 4.90 Caudal vert. 33 1.84 .62 1.19 3.47 Penultimate vert. 3 2.23 .81 1.39 3.00

214

APPENDIX-XXI: Frequency (NISP) of skeletal elements for naturally deposited fish along the Sea of Galilee.

Random squares-Sand Cyprinids Skeletal element Acanthobrama Small Large Cichlid Clarias Unident. Total Angular/Articular 1 2 0 2 0 0 4 Atlas 3 0 0 0 0 0 3 Atlas/Axis 1 0 0 0 0 0 1 Axis 3 0 0 0 0 0 3 Basapophysis 0 1 0 0 0 0 1 Basioccipital 0 3 0 2 0 0 5 Caudal vert. 0 3 0 0 0 0 3 Ceratohyal 0 7 0 0 0 0 7 Cleithrum 0 1 0 0 0 2 3 Coracoid 0 1 0 0 0 1 2 Dentary 0 1 0 0 0 0 1 Epihyal 0 1 0 0 0 0 1 Frontal 0 2 0 0 0 0 2 Hyomandibular 0 0 0 1 0 1 2 Maxilla 0 2 0 0 0 0 2 Neurocranium 0 0 0 0 0 1 1 1st Dorsal Pterygio. 0 2 0 1 0 0 3 Opercle 0 4 0 2 0 0 6 Opercle app. 0 0 0 0 0 1 1 Pelvic Spine 0 0 0 1 0 0 1 Pelvis 0 4 0 0 0 1 5 Cranial bone 0 0 0 0 0 5 5 Pharyngeal bone 2 0 0 1 0 0 3 Pharyngeal teeth 0 6 2 0 0 0 8 Preopercle 0 0 0 1 0 0 1 Pterotic 0 1 0 0 0 0 1 Pterygiophore 0 2 0 2 0 0 4 Quadrate 1 0 0 1 0 1 3 Rib 0 3 0 0 0 0 3 Scapula 0 2 0 0 0 1 3 Fin ray 0 0 0 0 0 2 2 Fin Spine 0 0 0 4 0 0 4 Scale 0 0 0 0 0 102 102 Tail vertebrae-element 0 0 0 0 1 0 1 3rd vert. 1 0 0 0 0 0 1 4th vert. 0 1 0 0 0 0 1 5th vert. 0 1 0 0 0 0 1 Thoracic vert. 0 15 0 3 0 0 18 Precaudal vert. 0 0 0 2 0 0 2 Precaudal/Caudal vert. 0 21 0 1 0 3 25 Penultimate vert. 0 2 0 0 0 0 2 Ultimate vert. 0 1 0 1 0 0 2 Vertebrae fragment 0 0 0 0 0 1 1 unidentified 0 0 0 0 0 73 73 Total 13 89 2 25 1 208 337 APPENDIX-XXI cont'd. Random-median brown Cyprinids Skeletal element Acanthobrama sp. Small Large Cichlids Unident. Total Angular/Articular 0 5 0 0 0 5 Atlas 14 0 0 1 0 15 Atlas/Axis 0 2 0 0 0 2 Axis 13 0 0 0 0 13 Basapophysis 0 4 1 0 0 5 Basioccipital 0 11 0 0 0 11

215

Caudal Vert. 0 4 0 0 0 4 Ceratohyal 0 12 0 0 0 12 Cleithrum 0 13 0 1 5 19 Coracoid 0 8 0 0 0 8 Dentary 0 21 0 0 0 21 Entopterygoid 0 1 0 0 0 1 Epihyal 0 5 0 0 0 5 Exoccipital 0 0 0 1 0 1 Frontal 0 6 0 0 0 6 Hyomandibular 0 16 0 1 1 18 Interopercle 0 0 0 1 0 1 Maxilla 0 6 0 0 0 6 Neurocranium 0 0 0 2 9 11 Opercle 0 31 0 0 2 33 Opercle app. 0 0 0 0 4 4 Os suspensorium 4th vert. 0 5 0 0 0 5 Palatine 1 0 0 0 0 1 Parasphenoid 0 0 0 0 1 1 Pelvis 1 21 0 0 0 22 Penultimate vert. 0 2 0 0 0 2 Pharyngeal bone 15 0 0 5 0 20 Molariformteeth 0 0 4 0 0 4 Pharyngeal teeth 6 13 11 0 0 30 Posttemporal 0 0 0 3 0 3 Precaudal vert. 0 1 0 1 5 7 Precaudal/Caudal Vert. 0 2 0 3 66 71 Premaxilla 0 0 0 1 0 1 Preopercle 0 5 0 0 3 8 Pterotic 0 1 0 0 0 1 Pterygiophore 0 8 0 2 0 10 Quadrate 0 12 0 0 0 12 Rib 0 7 0 4 0 11 Scapula 0 12 0 0 0 12 Spina neuralis 0 1 0 0 0 1 Supracleithrum 0 0 0 2 0 2 Supraoccipital 0 1 0 0 0 1 Fin ray 0 0 0 0 16 16 Fin Spine 0 0 0 9 0 9 1st Dorsal Pterygiophore 0 10 0 2 0 12 Urohyal 1 0 0 0 0 1 Vomer 1 0 0 0 0 1 unidentified 0 0 0 0 358 358 Scale 0 0 0 0 252 252 Cranial bone 0 0 0 0 19 19 APPENDIX-XXI cont'd Random-median brown Cyprinids Skeletal element Acanthobrama sp. Small Large Cichlids Unident. Total Mesethmoid 0 1 0 0 0 1 3rd vert. 12 0 0 0 0 12 4th vert. 0 9 0 0 0 9 5th vert. 0 10 0 0 0 10 Thoracic vert. 0 93 0 10 0 103 Ultimate vert. 0 3 0 0 0 3 Vertebrae 0 1 0 0 112 113 Total 64 363 16 49 853 1345

Random squares Clay Cyprinids Skeletal element Acanthobrama Small Large Cichlids Unident. Total Angular/Articular 5 10 0 2 0 17 Atlas 15 0 0 3 0 18 Atlas/Axis 0 14 0 0 0 14

216

Axis 28 0 0 2 0 30 Basioccipital 0 21 0 2 0 23 Ceratohyal 0 27 0 0 0 27 Cleithrum 0 30 1 1 14 46 Coracoid 0 18 0 3 0 21 Dentary 5 33 0 0 0 38 Epihyal 0 17 0 0 0 17 Exoccipital 0 3 0 0 0 3 Frontal 0 9 0 0 0 9 Hyomandibular 11 19 0 0 1 31 Hypohyal 0 1 0 0 0 1 Maxilla 10 2 0 1 2 15 Neurocranium 0 0 0 1 4 5 Opercle 2 31 0 3 5 41 Opercle app. 0 9 0 0 3 12 Os suspensorium 4th vert. 0 7 0 0 0 7 Palatine 4 2 0 0 0 6 Parasphenoid 0 6 0 2 1 9 Parietal 0 3 0 0 0 3 Pelvis 0 32 0 2 3 37 Pharyngeal bone 47 6 1 5 2 61 Pharyngeal teeth 0 8 3 0 0 11 Posttemporal 0 5 0 0 0 5 Preopercle 0 18 0 0 8 26 Pterotic 0 10 0 0 0 10 Pterygiophore 0 7 0 2 0 9 Quadrate 5 7 0 1 1 14 Rib 0 5 0 5 0 10 Scapula 0 27 0 1 0 28 Spina neuralis 0 2 0 0 0 2 Subopercle 0 16 0 0 0 16 Supracleithrum 0 0 0 1 0 1 Supraoccipital 0 0 0 1 0 1 Epiotic 0 1 0 0 0 1 Postcleithrum 0 1 0 0 0 1 Tripus 0 3 0 0 0 3

217

APPENDIX-XXI cont'd Random squares Clay Cyprinids Skeletal element Acanthobrama Small Large Cichlids Unident. Total Ultimate vert. 0 6 0 1 0 7 Urohyal 1 2 0 0 0 3 Ventral Postcleithrum 0 0 0 2 0 2 Vomer 0 2 0 0 0 2 Fin ray 0 2 0 0 14 16 Fin Spine 0 0 0 18 0 18 1st Dorsal Pterygiophore 0 10 0 0 0 10 3rd vert. 22 0 0 0 0 22 4th vert. 0 22 0 0 0 22 5th Vert. 4 5 0 0 0 9 Thoracic vert. 0 213 0 9 0 222 Precaudal vert. 0 0 0 3 0 3 Precaudal/Caudal vert. 0 0 0 2 52 54 Caudal vert. 0 20 0 3 0 23 Penultimate vert. 0 3 0 1 0 4 Tail vert.-Epural 0 0 0 0 1 1 Tail vertebrae-elements 0 0 0 0 1 1 Vertebrae 0 1 0 6 309 316 Vertebrae fragment 0 1 0 0 0 1 unidentified 0 0 0 0 2332 2332 Total 159 697 5 83 3167 4111

Recent Beach surface Cyprinid Skeletal element Acanthobrama Small Large Cichlids Clarias Unident. Total Atlas 0 0 0 1 0 0 1 Axis 1 0 0 0 0 0 1 Caudal vert. 0 0 0 1 0 0 1 Cleithrum 0 0 0 1 0 0 1 Dentary 0 0 0 0 1 0 1 Frontal 0 0 0 0 2 0 2 Hyomandibular 0 1 0 0 0 0 1 Neurocranium 0 0 0 0 0 1 1 Opercle 0 3 0 1 1 0 5 Opercle app. 0 1 0 0 0 0 1 Palatine 0 1 0 0 0 0 1 Pelvic Spine 0 0 0 1 0 0 1 Pharyngeal bone 0 0 0 2 0 0 2 Molariformteeth 0 0 1 0 0 0 1 Premaxilla 0 0 0 1 0 0 1 Preopercle 0 0 0 2 0 0 2 Pterygiophore 0 0 0 1 0 0 1 Rib 0 0 0 1 0 0 1 Sphenotic 0 0 0 0 1 0 1 Urohyal 0 0 0 1 0 0 1 Ventral Postcleithrum 0 0 0 1 0 0 1 1st Dorsal Pterygiophore 1 0 1 1 0 0 3 Fin ray 0 0 0 0 0 1 1 Fin Spine 0 0 0 1 0 0 1 Scale 0 0 0 0 0 17 17

218

APPENDIX-XXI cont'd Recent Beach surface Cyprinid Skeletal element Acanthobrama Small Large Cichlids Clarias Unident. Total 4th vert. 0 1 0 0 0 0 1 5th vert. 0 1 0 0 0 0 1 Thoracic vert. 0 2 0 2 0 0 4 Precaudal vert. 0 0 0 3 0 0 3 Precaudal/Caudal vert. 0 1 0 0 0 1 2 Tail vert.-elements 0 0 0 1 0 0 1 Vertebrae 0 0 0 0 0 6 6 unidentified 0 0 0 0 0 4 4 Total 2 11 2 22 5 30 72

219

APPENDIX-XXI cont'd

Ohalo-II recent surface Skeletal element Cichlids Clarias Total Basioccipital 0 1 1 Branchial region 0 4 4 Ceratohyal 0 1 1 Coracoid 0 4 4 Dentary 0 1 1 Epihyal 0 1 1 Fin ray 0 1 1 Fin Spine 1 2 3 Frontal 0 2 2 Hyomandibular 0 2 2 Hypohyal 0 2 2 Interopercle 0 2 2 Neurocranium 0 1 1 Parasphenoid 0 1 1 Pelvis 2 0 2 Prefrontal 0 3 3 Pterotic 0 2 2 Quadrate 0 2 2 Scapula 2 2 4 Sphenotic 0 2 2 Supraoccipital 0 1 1 Urohyal 0 1 1 Vomer 0 1 1 Cranial bone 0 12 12 Supraethmoid 0 1 1 Nasal 0 2 2 Radials 2 0 2 Epiotic 0 2 2 Metapterygoid 0 2 2 Thoracic vert. 1 10 11 Precaudal vert. 0 4 4 Caudal vert. 0 12 12 Vertebral complex 0 1 1 unidentified 0 8 8 Total 8 93 101

220

APPENDIX-XXII: Vertebrae dimensions (height, width, and length) mean (±SD) and range calculated by taxa for naturally deposited fish.

Vertebrae Height (mm) Genus group Vertebrae NISP Mean Std. Dev. Minimum Maximum Acanthobrama sp. Atlas 21 .91 .11 .72 1.14 Axis 23 .78 .06 .67 .93 Third vert. 11 1.24 .13 .98 1.42

Small cyprinids Fourth vert. 1 1.04 1.04 1.04 Fifth vert. 1 1.42 1.42 1.42 Caudal vert., 2 3.29 .37 3.03 3.55 Precaudal/Caudal vert. 9 1.99 .33 1.48 2.38 Thoracic vert 4 1.75 .24 1.51 2.08 Vertebrae 1 2.30 . 2.30 2.30

Cichlids Atlas 3 1.98 .56 1.33 2.34 Penultimate vert 1 2.22 2.22 2.22 Precaudal/Caudal vert. 1 3.20 3.20 3.20 Thoracic vert. 1 3.53 3.53 3.53

Unidentified Vertebrae 7 1.49 .33 1.11 2.06

Vertebrae Width (mm) Genus group Vertebrae NISP Mean Std. Dev. Minimum Maximum Acanthobrama sp. Atlas 19 1.73 .38 1.15 2.62 Axis 19 1.50 .14 1.30 1.80 Third vert. 11 1.40 .19 1.13 1.70

Small cyprinids Fourth vert. 1 1.35 1.35 1.35 Fifth vert. 1 1.58 1.58 1.58 Thoracic vert 3 1.40 .41 1.13 1.87 Precaudal/Caudal Vert 8 1.75 .25 1.44 2.16 Caudal vert. 2 3.37 .47 3.04 3.70 Ultimate vert. 2 1.29 .02 1.27 1.30

Cichlids Atlas 3 1.81 .39 1.40 2.17 Precaudal/Caudal Vert 1 3.14 3.14 3.14 Penultimatevert. 1 2.70 2.70 2.70 unidentified Vertebrae 6 1.46 .12 1.35 1.68

221

APPENDIX-XXIII cont'd.

Vertebrae Length (mm) Genus group Vertebrae NISP Mean Std. Dev. Minimum Maximum Length (mm) Total 79 1.57 .47 1.00 3.45 Acanthobrama sp. Atlas, 19 1.53 .22 1.23 2.10 Axis 21 1.42 .13 1.22 1.80 Third vert. 10 1.44 .14 1.22 1.70

Small cyprinids Fifth vert. 1 1.22 1.22 1.22 Thoracic vert 4 1.34 .18 1.19 1.60 Precaudal/Caudal vert. 8 1.49 .22 1.18 1.83 Caudal vert., 2 3.34 .16 3.23 3.45 Ultimate vert. 2 1.17 .05 1.13 1.20

Cichlids Atlas 3 2.03 .51 1.50 2.52 Thoracic vert. 1 3.02 3.02 3.02 Precaudal/Caudal vert. 1 3.15 3.15 3.15 Penultimate vert. 1 2.53 2.53 2.53 unidentified Vertebrae 5 1.28 .22 1.00 1.62

222