PADINA AND STARCEVO: GAME, FISH AND CATTLE

A. T. Clason

CONTENTS

1. INTRODUCTION

2. THE SITES 2. 1. Padina 2.2. Starcevo

3. THE SPECIES

3. I. Mammals 3. 2. Birds 3-3· Rep ti les 3 + Fishes 3.5. Molluscs

4. DISCUSSION 4. 1. Near East 4.2. Southern Europe 4.3. The route to the middle Danube basin 4+ The Iron Gate 4.5. Padina 4.6. StarC:evo 4.7. Recapitulation

5 · ACKNOWLEDGMENTS

6. NOTES

7. LITERATURE A. T. CLASON

i. INTRODUCTION* I N I 1 ( In i971 Prof. Dr. R. W. Ehrich ), one of the exca­ f vators of the settlement of Starcevo, asked me to study the animal remains which were collected

during the excavations in I 9 32, r 969 and I 970. l Some years later Dr. B. Jovanovic2) invited me to study the faunal remains which were collected Beograd J' during the excavation of the Late Mesolithic/Early I < I Neolithic settlement of Padina in the Djerdap \ ' region (Iron Gate). Since both sites are situated on ) " " a bank of the Danube in Yugoslavia (fig. i ) , and ( both are important for our understanding of the / ) end of the Mesolithic period and the introduction --1 '....\ of agriculture in Central Europe, I decided to ) I J- _ review the results of both studies together. \. _ ,1 For several reasons it would have been pointless ...... ,_, to estimate the minimum number of individuals, the meat yield of the diffe rent species and their Fig. 1. The geographic situation of Padina and Starcevo. contribution to the daily diet of the former inhab­ itants. In the first place a large proportion of the

animal remains were not collected. In the second N place the former inhabitants used numerous bones RUNA NIA for making tools. Especially those of Starcevo had the habit of carving segments out of the long bones t of cattle and red deer to make large fish-hooks. Not to mention the spoons carved out of the meta­ podiae of cattle (Nandris, 1972). The inhabitants of Padina also used bones to fabricate tools. In the 0 5 10km third place only a small, much disturbed, part of the '------'---' settlement of Starcevo could be excavated for most of it had been destroyed. Nothing is known about Fig. 2. The settlements Padina, Lepenski Vir, Vlasac, Cuina its extensions, number of houses, etc. This also ap­ Turcului and Icoana in the Iran Gate. The Danube is the plies to a certain extent to Padina, although this site boundary between Yugoslavia and Romania. was more completely excavated. Because of these reasons only the number of bones will be given and the percentages in which 2. THE SITES they were collected. Small species are in this way certainly underrepresented. Of the large mammal 2. r. Padina species probably the same percentages have been thrown away as were collected, so that these per­ Padina is situated on a narrow strip ot ioess in the centages are at least an indication for the percen­ Iron Gate (Djerdap) on the Yugoslavian side of the tages in which the species were kept, or better Danube (fig. 2). The area of the settlement was slaughtered or killed in hunting. divided by J ovanovic ( r 97 r, r 974) in three main sectors (fig. 3), in which he could distinguish three major habitation periods:

* Tables 7-60 have been reproduced as microfiches, which are furnished in an envelope attached to the rear cover of I. The prehistoric period. Late Mesolithic (Epipalaeolith­ this volume. ic) of the lower basin of the Danube (A), Old Neolithic Padi11a and S tarcevo : GaJ11e, Fish and Catt/e

of the Djerdap (Iron Gate), a variation of the complex of Remains of the Late Mesolithic (A) were found in Starcevo-<:;:risculture (B), Recent Eneolithic of the cen­ sector I and II. The Neolithic (B) could be divided tral Danube basin (Kostolac-Cotofeni) which are restric­ into three phases B B and B . The first row of ted to isolated finds (C), Iron Age of the central and l' 2 3 lower Danube basin (the type of Basarabia-Gomolava­ houses, nr. 5 -10 in sector III, along the Danube Bosut) (D). belongs to phase B1 . The fi rst row of houses in II. The classical period is represented by the foundations of sector I, nr. 1-3, and the second row in sector III, a Roman tower, some objects and cremation graves (E). nr. r 1-1 5, belong to phase B . The third row of III. The medieval period is represented by silver and gold 2 houses in sector III, nr. r 6-z r, belongs to phase B coins of the second half of the XIVth century. 3 (fig. 4). No traces of this phase were found in sector

Fig. 3. The excavated area at Padina in the Iron Gate (after Fig. 4. Detail of sector III with the house places j-2 1. A cross­ J ovanovic, 197 1 ). The shaded area in sector III is given in section of the shaded area is given in fig. 5 (adapted from detail in fig. 4. Å Jovanovic, 1971). '

' ' ' \

17 - '

--" ... ', 1 ' ' f'\ l_ _J 19 "l_ j'

I'-- __ ,, 11 ' .,,...... " \ " > I ' ' 'I �. /':­6 L J- i_ __ _j (� 7 00 LJ.

IOm 143 A. T. CLASON

PA DINA 64.569 \J CROSS-SECTION OF SECTOR III ( Early Neolithic - Pad ina B)

HOUSE 16 FLOOR

� \J 60.798 < HOU SE 13 � FLOOR HOUSE 6 Fig. Proposed cross-section j FLOOR 5.

- - ·------of sector 111 of Padina. Early Neolithic - Padina B (data l 0 1 m provided by Jovanovic). bl..ld..I

I. The floor plan of the houses is trapezoid, with a Neolithic village consisted of three rows of houses hearth in the middle. The walls were of dry-stacked on different levels (fig. 5). According to Jovanovic stones. The backs of the houses were cut into the not all houses were inhabited at the same time. He loess plateau on which the settlement was situated. bases this assumption on the finds associated with The roofs were probably of wood. In this way the the houses, but this assumption is not corroborated by C14-dates. The C14-dates seem to suggest that what must have been the first two phases were

TA8LE 1 younger than the two younger phases. The same confusion in C14-dates is found in Lepenski Vir and 1 4 Chronology and C -dates of Padina, Lepenski Vir and Vlasac. Vlasac (table r). The fa et that the C14-dates do not Padina Lepenski Vir Vla sac confirm the archaeological stratification, can be

83 Neolithic (Starc./Cris) III not dated III possibly explained by taking into account the faet 0 5125 ± 50 8C (Starcevo) 5490 ± 10 8C that these narrow loess strips along the gorge were Sector I no houses Sector li houses no. 16-21 intensively used during the period during which they were inhabited. Owing to this organic mate­ 82 Neolithic (StarC./Cris) li li 5525 - rial could have easily become mixed, unobserved 5150 ± 80 8C 4610 ± 100 8C 5985 ± 100 8C Sector I houses no. 1-3 4680 ± 100 8C by the excavators3). Sector III houses no. 11-15 Because of the slope, could earlier material from I b4915 8 1 Older Neolithic I 5880 ± 1 00 8C - higher up have slid down over later deposits (fig. (StarC./Cris) 4670 ± 5330 8C a4965 ± 100 8C Sector III houses no. 5-10 5). The kind of excavation by horizontal layers

4620 ± 55 8C would mask this. Also, if there were sloping levels, 5115 ± 110 8C excavation by artificial strata could produce con­ (possible Proto Padina Proto Lepenski temporaneity on one level and a reversed strati­ was not found) Vir not dated graphy as well. A. Late Palaeolithic It is for this reason that the fauna! remains were 4615 ± 40 8C (charcoal) divided into only five main groups: I. Late Meso­ Sector I and li lithic; II. Late Mesolithic/Early Neolithic; III. 7500-5500 8C (human bones) Early Neolithic; IV. Early Neolithic/Early Iran 144 Age; V. Iran Age/Roman period. The majority of Padi11a a11d Stareevo: Ga111e, Fish a11d Cattle

posits, on which the ancient settlement was situat­ ed, were used by brickmakers, who destroyed most of the settlement. Starcevo was fi rst excavated in

193 l by Grbic and Fewkes, who examined two pits

( l and 2) , the contents of which, including the animal ·bones, were brought to the Peabody Mu­ seum of Harvard University and were not available to me (Ehrich, 1977).

In l 9 p the major excavation of Starcevo took place under the direction of Fewkes and Ehrich. After a lapse of nearly 40 years, Ehrich and Dr.

--- 5km Gara5anin of the National Museum in Beograd ex­ cavated three trenches on the upper flatland, some­ E::::::::3 V77?I i=-=-::lMARSH BLACK SOIL AHO what to the North and West of the old excavations. � CHERNOZfM l::L:LL'.J ALLUVIUM t..=..=JMIH ERAL BOG SOIL These were dug in public land under cultivation while five others were opened in private garden Fig. 6. The geographic situation of Starcevo (adapted from land to the South. The inhabitants of the village Barker, 1975). improved their land with the humic earth from the prehistoric sites, which makes the fauna! remains unreliable. They are partly mixed with recent the fa una! remains belongs to the Late Mesolithic bones. The stratification is not certain.

and Early Neolithic. The differences in the per­ In l 9 p four main lev els were recognized which centages of the species in both groups, seem to were then called humus, subhumus, culture and pit indicate that the material was right! y alloted to level. These levels were later renamed humus, level these phases. It seems that they also correspond I, level II and level III. The pits opened from with changes in the pottery belonging to the com­ (pit)level III and penetrated into the loess. The pits plex of the Stan:'.:evo- c;:ris-Koros culture. Group I contain mostly uncontaminated Starcevo material, can be probably roughly dated in the sixth millen­ as does (pit)level III. The bones from the pits are nium B.C. and group III at the end of the second yellow. The same applies to the bones of (pit)level half of the sixth millennium and the first half of the III. Levels II and I are partly contaminated with fifth millennium B.C. younger material. Also in 1932 two sounding trenches, l9pA-A extension and B, were opened

2.2. Starcevo some l 5 o m to the North of the main excavation (fig. 7). These trenches showed also no good strati­ Starcevo, the site that gave the name to the Star­ fication and contained some younger material cevo culture, lies upstream from Padina on the. left down to the loess. bank of the Danube in the central Danube basin In l 9 3 2 bones were collected only in so far as it near the modem town of Pancevo and the village was thought possible to identify them, all the of Starcevo. The settlement was built on an old others were discarded. None of the earth was bank of the Danube at the edge of its flood plain sieved, which is clearly demonstrated by the fish (fig. 6). Today the Danube flowsapproximately 3.5 remains which all belong to large catfish and carp km to the West, and the plain is protected against species. the river by dikes. In prehistoric times the plain In r 969 and l 970 part of the earth was sieved would have been marshy with cut-off meanders, with the result that bones of smaller fish- and bird small streams, and ponds bordered by reeds. The species were found. Since the material from r 969 plain would have been partly covered by woods and 1971 was mixed and contaminated with youn­ and copses, but there were also open spaces. The ger and even recent material, the bones that looked river probably flowed much nearer to the settle­ prehistoric were counted and partly measured, but ment than it does today. Since 1912 the loess de- not otherwise described. 145 A. T. CLASON

The bones collected in 1932 (fig. 7, 8) are di­ vided into 22 units: A-V. In table 5, 6 the number of bones of the species found in each of these units is given, together with those found in the pits, level IR. A-1931D III (pit), level II (culture), level I (subhumus), c::::::=J ". B -19 31 trench A-A extension, trench B, and the trenches of 1969/1 970. No traces of dwellings were found. The irregu­ lar pits are not considered to have served as houses. There are six C14-dates, four from animal bones, one from human bones and a sixth from sherds. This last date is unreliable. The other four indicate that Starcevo was inhabited during the first half of the fifth millennium B.C. N ""JD _ sounding l GrN-6626 StarCevo 1 bone 6610 ± 65 - 4660 BC =:J�_S oundin9 Il GrN-6627 2 bones 654 5 ± 1 05 - 459 5 BC 19 69 GrN-6628 ) sherds 761 5 ± 50 - GrN-6629 4 bones 66 15 ± 65 -4665 BC \ GrN-7 154 5 human bones 66 10 ± 100 - 4660 BC GrN-7155 6 bones 6835 ± 70 -4885 BC Uncorrected dates. Calibration would make them some 800 years earlier.

These da tes mean that the occupation of Starcevo is 2Sm '--6.d...... � - more or Jess contemporaneous with that of the Neolithic levels B1-B3 of Padina. According to Ehrich ( 1977) the Starcevo culture can be divided into early and late periods, and Starcevo itself be­ longs to the later Starcevo period.

3. THE SPECIES

Bones or shells of the following species were col­ lected : hedgehog, brown hare, beaver, fox, dog, wolf, badger, otter, pine marten, wild cat, lynx, brown bear, wild/domestic horse, wild ass, wild/ d domestic pig, roe deer, red deer, wild/dorriestic c::::,,d cattle, sheep, goat, chamois, black-throated eagle, kite, eagle, harrier, vulture, crane, great bustard, curlew, raven, rook, domestic fowl, European pond terrapin, Danube salmon, stur­ geons, pike, catfish, carp, carp fishes, bream, asp, mussel, edible snail (tables 2-6). Fig. 7. The trenches excavated at Starcevo in 1932, 1969 and 1970 (data provided by Ehrich).

3. r. Mammals hic layer (III) of Padina. The distal epiphysis was

Hedgehog - Eri11ace11s europaeus Linnaeus, r 75 8 broken off. Of the hedgehog three varieties, differ­ Of this species a left ulna was found in the Neolit- ing in the colour of the fur, are known. According Padi11a and Stareev o: Ga111e, Fish and Cattle

to Miller ( 1966 ( 191 2)) the east form cannot be dis­ Fig. 8. The pirs in the large trench excavated in 1932. Pit 4 is tinguished from the west form with certainty with­ probably a recent, partly filled with old material, ditch (data provided by Ehrich). out comparing the skull and teeth. Whether there is also a diffe rence in other skeletal parts is not men­ tioned. At present the eastern variety has a white than in Starcevo (fig. 10). In the last si te a tibia, a spot on its breast which is sharply set off from the mandibula and a humerus were found in pits, and 5 grey brown colour of the rest of the fu r. In Yugo­ bones in the excavation trenches of 1969 and 1970. slavia the east variety is living toda y (Van den The measurements of the few beaver remains from Brink, r 968). Starcevo fall into the range of measurements of The hedgehog is found on dry ground with those of Padina. bushes, hedges, etc" but not in deep woods. The In Padina both in period I as III mandibulae hedgehog makes excellent food. were found of which the lower border and the in­ cisors are missing. It might be that the lower

Brown hare - Lep11s cape11sis Linnaeus, 175 8 border was cut away to obtain the incisors which From the hare an ulna and probably a femur were could be used as implements. No part of the skul! fo und in Neolithic (III) Padina. An ulna, fe mur, and maxilla were collected. Mandibulae were found tibia, and calcaneus were found in the Mesolithic in the following numbers : (I) layers, and an undated pelvic. bone. No hare r. I. 1./r. remains were found in Starcevo. This is probably Mesolithic lower border due to the way in which the animal remains were 19 Neolithic 3 4 collected, since the environment of Starcevo seems more suitable for hares than that of Padina, al­ The length of the tooth row was 34. 5-41.0 mm. though hare remains were also found in Vlasac and According to Gaffrey (195 3) the length of the in Romanian sites from the same period on the tooth-row of recent beaver in Central Europe lies opposite bank (Bolomey, 1973). Although the hare between 32.6-40.0 mm. Bokonyi (1976) mentions prefers open country, the species can also live in fo r Obre measurements of 39.0, 39.5, 41.0 and 44.0 deciduous woods and on mountains. The hare mm. Boessneck (1963) found in Burgaschisee-Siid 'makesexcellent food, has a useful fur and its bones in Switzerland measurements which lie between can be used to make tools. 3 3 .0-37. 5 mm. For Eneolithic Vlaardingen in the Netherlands they were 32.5-38.2 mm (Clason,

Beaver - Castor fiber Linnaeus, 175 8 1967). The Yugoslavian animals seem to be rela­ The remains of beaver are more abundant in Padina tively large. 147 TABLE 2 The mammal species (wild and domesticated) collected in Padina, and the number of bones !hat were found of each of them.

III IV V VI III IV V VI % % % % % %

Mammalia

A. Domesticated

Bos taurus - domestic cattle 7 21 7 10.4 9.9 33.3

Capra!Ovis - goat/sheep 6 6 8 8.9 2.8 38.0

Sus domesticus - domestic pig 12 2 5 2 2 17.9 2.3 9.5

Canis familiaris - dog 42 179 4 7 62.6 84.8 19.0

B. Wild, insectivores, lagomorphs, rodents, carnivores

Erin aceus europaeus - hedgehog

Lepus capensis - brown hare 4

Castor fiber - beaver 37 10

Vulpes vulpes - fax 6

Canis/lupus - dog/wolf 3

Canis lupus - wolf 6 3

Martes martes - pine marten 3 2

Martes sp. - marten 3 4

Fe/is silvestris - wild ca! 4 5

Lynx /ynx - lynx 2 2

Meles meles - badger

Ursus arctos - brown bear 30 15 5

C. Wild, ungulates

Equus przewalskii - wild horse

Equus sp. 0.1 2.4

Sus scrofa - wild boar 55 50 2 5.6 13.4 4.8

Sus sp. - wild/domestic boar 23 10 6 2.3 2.6

Capreo/us capreo/us - roe deer 37 9 3.8 2.4

Cervus elaphus - red deer 763+114' 4+10' 263+50' 2' 31+2' 8 78.8 70.6 75.6

Bos/Cervus - cattle/red deer 41 12 4.2 3.2

Bos prim igenius - aurochs 28 8 4 2.8 2.1 9.7

Bos sp. - cattle 17 2 20 3 1.7 5.3 7.3

Rupicapra rupicapra - chamois 3 0.3

D. Homo sapiens - man + +

148 Sum A 67 3 211 2 21 9 5.9 30.0 33.5 25.0 31.3 39.1 Sum B 96 1 46 5 2 8.4 10.0 7.3 7.4 8.6 Sum C 968 6 372 6 41 12 85.5 60.0 59.1 75.0 61.1 52.1 Sum A+ B +c 1131 10 629 8 67 23

• antler TABLE 3 The bird, reptile, fish and molluscs species collected in Padina, and the number ol bones that were found ol each ol them.

III IV V VI

E. Aves - birds

Gavia arctica - black-throated diver

Pelecanus onocrotalus - white pelican 3

Phalacrocorax carbo - cormorant

Cygn us olor - mute swan

Cygnus sp.

Anser anser - grey-lag goose 2

Ta dorna tadorna - shell-duck

cf. Haliaeetus albicilla - eagle 7

- small bird of prey

Gypaetus barbatus

Aegypius monachus - vulture

Gyps fulvus

Corvus corax ) - raven

Corvus frugi/egus - rook

Aves inde!. - unidentified 2

Sum 16 7 4

F. Reptilia

Emys orbicularis

G. Pisces

Hucho hucho

Acipenser sp.!Huso huse

Siluris g/anis

Cyprin us carpio

Cyprinidae

Inde!.

Sum

H. Mollusca

Unio crassus

Helix sp.

Cepaea sp.

Inde!.

Sum TABLE 4

Padina I-VI. The percentages ol A, B, C, E, F, G and H.

III IV V VI III IV V VI % % % % % %

A 67 3 21 1 2 21 9 4.4 15.0 7.1 30.4 33.3 B 96 1 46 5 2 6.3 5.0 1.5 7.2 7.4 c 968 6 372 6 41 12 63.6 30.0 12.5 59.4 44.4 E 16 7 4 1.0 0.2 14.8 F 1 1 0.06 0.03 G 367 9 2275 24.1 45.0 76.5 1.4 H 5 58 0.3 5.0 1.9 1.4

Sum 1520 20 2970 8 69 27

The beaver needs light deciduous woods with Domestic dog - Canis fa 111iliaris Linnaeus undergrowth along rivers and lakes. This biotope Bones of dogs were abundant in the Mesolithic (I) was available in the vicinity of both settlements. and Neolithic (III) layers of Padina. In Starcevo The beaver has tasty meat, good fur, and the only a small amount of dog bones were found in beaver has castoreum. the Neolithic pits as well as in the other layers (fig. ro) . Fox - Vulpes vulpes Linnaeus, r 7 5 8 No complete skulls were recovered and many of In Mesolithic Padina (I) five bones and a tooth of the Padina mandibulae are broken, show carving fox were found: two mandibulae, one tooth, one and gnawing marks and have been in contact with scapula, and two calcanea. In Starcevo only one fire (fig. r r ). Also the long bones are broken. The bone of a fox was collected during the excavation broken or natura! ends are often scorched by fire. In I 970. Apparently dogs were used as food. The same was The fox can adapt itself to many biotopes, al­ found in Vlasac and Lepenski Vir (I, II). In Vlasac though it prefers dry ground and the vicinity of and the first two stages of Lepenski Vir the dog woods and light brush. was the only domestic animal (Bokonyi, r 97 3, The fox might have been hunted in the first I 975 ). place for its fu r. Since the dog was also an ap­ Of the 34 mandibulae of Padina III, 14 had been

preciated food animal in Padina, the fox might have in contact with fire of which r had a burned under­ been eaten as well. side and three were burned on the labial side. None of the fragments were calcinated, only one was

Wolf - Ca11is l11p11s Linnaeus, r 75 8 burned completely black. This seems to indicate In the layers of Mesolithic Padina (I) six bones of that the mandibulae were roasted on a fire and not wolf were found, in the Neolithic layers (III) three. cooked or devoured fresh. The point has been In Starcevo one tibia was found in a pit. No skull made, and observed among the Eskimos that an fragments were collected. uncooked bone must be split to get at the marrow,

According to Bokonyi ( r 97 5) in Late Mesolithic whereas if cooked and broken across, the marrow Vlasac the wolf was in process of being domesti­ can be sucked or fi shed out. The many gnawing cated. In Padina this could not be proved. In this marks indicate that living dogs were walking free respect Padina resembles Lepenski Vir (I, II) and through the village. In most cases only the pars the Romanian sites on the other side of the molaris was left, the pars incisiva and the vertical Danube. ramus was cut or gnawed off. The wolf is found in wooded plains and in Most dogs were slaughtered when they had al­ mountainous areas, but it can also live in open ready an adult dentition, with two exceptions. One ground with cover. of those still had milk back teeth p2p3 while M1 was The wolf might have been hunted as predator, breaking through. Of the second mandibula M3 the fur can have been used and the meat might have was breaking through. It is not yet known whether I jO been eaten. there is much difference in the seguence and age of Fig. 9. Bones found together with pottery sherds in the midden overlaying pit 6, unit K (bag nr. 383-384) in 1932 (see table 13) (photo: R. \XI. Ehrich).

\ . , _ . r ...... "" .----__,.-_,u/,

g

0 5cm 5cm

Fig. 10. Starcevo, Castorfiber, a (?), b (14/1970) - tibia, c (J bag Fig. 11. Padina, Ca11isfa 111iliaris - three mandibulae from the 379) -humerus; Ca11isja111iliaris, d, e, g (U bag 498/499), f(?) ­ Early Neolithic layer in sector III. The teeth of two lower mandibula; Fe/is si/ves/ris, h ( 1969/ 2 5 ) - mandibula (photo: mandibulae were scorched by fire (photo: CF.D., R.U. Gro­ National Museum, Belgrade). ningen). TABLE 5 Starcevo, the domesticated and wild mammal species, and the number ol bones collected ol each in 1932 (A-V) and the excavations in 1969 and 1970.

pits Mammalia A B D F G H K L Sum c E %

A. Domesticated

Bos taurus - domestic cattle 15 31 27 62 24 50 39 24 103 51 71 8 505 66.7

Capra!Ovis - goat/sheep 8 12 7 13 8 6 17 6 46 32 48 10 213 28.1

Ovis aries - sheep 4 0.5

Capra hircus - goat

Sus domesticus - domestic pig 3 2 6 4 5 7 4 31 4.1

Canis familiaris - dog 3 0.3

B. Wild, rodents, carnivores

Castor fib er - beaver 2 3

Vu/pes vulpes - lox

Canis lupus - woll

Ursus arctos - brown bear

Meles meles - badger

Lutra lutra - otter 2

Fe/is si/vestris - wild cat

C. Wild, ungulates

Equus przewa/skii - wild horse 3 2 5 1.2

Equus et. hydrun tinus

Sus scrofa - wild pig 2 3 13 16 39 2 2 8 7 7 4 103 26.1

Sus sp. - pig? 2 4 6 9 3 2 11 2 43 10.9

Cervus elaphus - red deer 3 2 1' 15+1 ' 9+1' 12+1· 9+9' 5+2' 17+1' 33+11' 8 113 28.6

Bos!Cervus - cattle/red deer 2 0.5

Capreolus capreo/us - roe deer 2 3 1+1' 9 2.2

Bos primigenius - aurochs 6 6 7 9 3 40 10.1

Bos sp. - cattle 2 16 8 24 10 13 3 79 20.0

D. Homo sapiens - man 2 5

Sum A 24 46 34 77 39 61 58 30 155 90 123 19 756

Sum B 3 3 9

Sum C 7 11 6 54 48 93 23 9 57 46 33 7 394

Sum A+ B + C 31 57 40 134 88 154 82 39 213 139 156 26 1159

A % 77.4 80.7 85.0 57.4 44.3 39.6 70.7 76.9 72.7 64.7 78.8 73.0 65.2

B % 2.2 1.1 1.2 0.4 2.1 0.7 c % 22.5 19.2 15.0 40.2 54.5 60.3 28.0 23.0 26.7 33.0 21.1 26.9 33.9

• antler trench B trenches 1969/1970 level III level li level I trench A Sum Sum 1969 69170 1970 Sum M N Sum % 0 p Q % R % s T u % V % %

58 73 131 59.5 84 117 52 253 72.9 162 69.8 117 73 75 265 83.5 148 71.8 265 46 275 586 67.7

ind 53 23 76 34.5 42 2 31 75 21.6 47 20.2 19 10 13 42 13.2 53 25.7 75 40 134+1 250 28.9

- - - - - 1 0.1 - - - - 1 - - 1 0.2 - - - - 2 0.9 1

------1 1 0.1

26 3.0 5 4 9 4.0 11 4 - 15 4.3 19 8.1 2 3 1 6 1.8 1 0.4 7 2 17

- - 1 0.1 2 2 4 1.8 - 3 - 3 0.8 4 1.7 - 1 3 4 1.2 2 0.9 1

------2 3 5

------1 ------1

------

------

------1 1

------

------

------1.6 - - - - - 4 - 4 3.4 8 11.2 - - 2 2

------1 1 0.8 - - - -

16 6 22 30.5 13 9 8 30 25.8 18 25.3 - - - - - 6 10.3 1 4 8 13 10.8

- - - - - 2 1 7 10 8.3 - 5 5 6.9 1 2 - 3 2.5 10 14.0 3 5.1

19+1' 9+1' 28+2' 38.8 18 25 18+3' 61+3' 52.5 25+1 ' 35.2 6 5 1' 11+1' - 35+9' 60.3 15+10' 6 45 66 55.0

- - - - - 1 - 1 1.3 - 2 1 3 2.5 1 1.4 - - - - - 1 1.7

- - - - - 1 9 15 12.5 1 - 1 1.3 - - - - - 3+1 ' 4.2 2 3.4 5

3 4 7 9.7 5 4 2 11 9.4 3 4.2 4 2 10 16 - 9 15.5 1 - 5 6 5.0

- 3 2.5 3 4.2 2 2 8 12 - 2 3.4 3 - 5 8 6.6 2 6 8 11.1 2 1

------1 2 - - - - - 1 - - 1 - 2 - 2 2 1

- - - - 347 88 430 865 - 118 102 220 - h38 126 83 347 232 138 87 92 317 206

------1 - 1 - 2 4 6

- - 120 - 42 30 72 - 39 46 31 116 - 71 - 12 9 18 39 58 29 12 79

- - 150 97 110 357 - 264 - 376 102 513 991 - 160 132 292 - 1 77 172 114 463 303

- - 87.2 - 73.7 77.2 75.3 - 77.9 73.2 72.8 74.9 - 76.5 - 92.0 89.6 83.6 88.7 78.0 92.2 86.2 83.8

------1.0 - 0.2 1.9 0.7 0.6

- - 12.1 - 26.2 22.7 24.6 - 22.0 26.7 27.1 25.0 - 23.4 - 8.0 9.2 16.3 10.9 21.9 7.7 11.7 15.3 A. T. CLASON

TABLE 6

Starcevo; the bird, fish and mollusc species and the numbers recovered.

B C D F H J K Sum M P Q Sum R S 1969 1970

E. Aves - birds

Domesticated

Gal/us gal/us domesticus - domestic fowl 2 2

Wild

Anas cf. clypeata - duck (shoveler) -

Anser anser - grey-lag goose 2

Anser cf. anser fabalis - goose (bean) 2 3

Cygnus cf. olor - swan (mule) 2

Cygnus cf. cygnus - swan (whooper)

Milvus sp. - kile

Aqui/a sp. - eagle

Circus sp. - harrier (1) (1)

Grus grus - crane

Otis tarda - greal buslard 2 2 5

Numenius arquata - curlew

G. Pisces - fishes

Esox lucius - pike 2

Aspius aspius - asp

Abramis brama - bream 2

Cyprinidae carp fishes

Cyprinus carpio - carp 3

Si/uris g/anis - catfish 5 3 2 5 2 17 28 16

lndet. - unidenlified

H. Mollusca - molluscs

Vi viparus viviparus

Unio crassus crassus 2 3

shedding teeth and teeth eruption between the we assume that the young dags were barn in early many races of dags, but according to Habermehl May, like wolf pups, the young dags were slaugh­ (1975) this is probably not so. All dags have their tered in August and November. At present dags adult dentition at the age of 6-7 months, M1 erupts mate the year around. I 54 between 3�-5 months, M3 between 6-7 months. If Of a left tibia the distal su ture was not yet com- Padina a11d Stareevo : Game, Fish and Cattle

pletely closed, which indicates that the dog was If the withers height is calculated with the Har­ slaughtered when it was approxii:nately 15 months court formulae,we fi nd in Starcevo for a humerus old in summer. Of a second left tibia both the distal 3 9. 8 cm. In Padina (III) for the humerus 5 8 cm and and proximal epiphyses were not yet fused with the for the two tibiae 40. 7 and 5 1.3 cm. All four heights shaft. This animal could not have been alder than fa]] within the limits given by Harcourt ( 1974) and 1 5 months. Probably it was younger. Clutton Brock (Burleigh et al" 1977) for English Of the Mesolithic dags in Padina one was slaugh­ dags in the Neolithic. tered befare the distal suture of the femur was closed, befare it was 1 8 months old. Of one tibia Badger - A1eles JJJeles Linnaeus, 175 8 the prox. epiphyses was not fused with the dia­ In the Mesolithic layer of Padina (I) a fragment - physes. pars molaris - of a mandibula of a badger was found. The back teeth were missing. In Starcevo a Mesolithic left mandibula and a right ulna were found in a pit maxilla r. mandibula r. I. (I). adult dentition adult dentition 4 The badger lives in mixed, deciduous woods ( one had been in contact with fire) with open spots, not too far away from water, both fem ur, distal su ture not yet closed I months > 8 in level and mountainous regions. tibia, prox. su ture not yet closed > I 8 months The skin is useful and the meat could have been Neolithic used as food. maxilla r. l.

Otter - Lutra lutra Linnaeus, r 7 5 8 mandibula r.* l. l.* r. Of the otter two bones were collected from pits in II 6 9 8 Starcevo: a humerus and an ulna. 32 adult dentition 7 months or alder The otter lives along rivers and in marshy areas. I p2p3M 1 erupting 3!-5 months

I M3 erupting 6-7 months The meat is good and the fur beautiful. femur, prox. suture still open 6-9 months tibia, prox. + Pine marten - Martes martes Linnaeus, r 75 8 and prox. sut. not closed Ij months ± marten - Martes sp . * has been in contact with fire. Both in Mesolithic (I) and Neolithic (III) Padina the remains of marten were found. In the Mesolith­ The bones of Padina III were broken to such an ic layers three mandibulae and in the Neolithic extent that they were hardly measurable. The length two, which could be identified as pine marten be­ of the M of the mandibula lies for Padina (III) 1 cause the foramina mentalia were placed far apart. between 19.0-23.0 mm with a mean of 21.2 mm. This characteristic according to , Gaffrey ( r 9 5 3) is Compared with M of dags in Gomolava (Vinca 1 specific for the pine marten. Three femora in and La Tene) and Roman Sirmium (Sremska Mi­ Padina I, and four femora in Padina III could not trovica) it shows that the dags of Gomolava-Vinca be identified as to species. The small size of the were smaller than those of Padina III. three mandibulae which could be measured seems to indicate that they belonged to females. Accord­ 18 19 20 21 22 23 24 25 26 mm ing to Gaffrey ( r 9 5 3) the length of the tooth row a for recent pine marten lies between 31· 5 -33.5 mm b 1 4 for <;2<;2,and between 35.5-38.0 mm for ()().For the c 2 6 2 3 2 1. beech marten between 31.3-33.0 mm for <;2<;2, and d between 34.0-36.0 mm for ()(). The lang bones e 2 2 could belong to either species, but because the mandibulae belong to the pine marten the lang bones probably do tao. Ca11isJa miliaris. M I' max. length : a Starcevo; b Gomolava - Vinea; c Padina III; d Gomolava - La Tene; e Padina V; f The pine marten is found both in deciduous and Sirmium. * Ca11is l11p11s. pine woods, the beech marten at the edge of woods A. T. CLASON

in rocky and mountainous areas. Today the last brought to the village and were used as food. species is a culture follower and lives in the vicinity Probably the bides were also used. of human settlements. Bears live in large, extended woods both on the Whether the meat would have been eaten is plains and in the mountains. questionable. The furs however would have been useful.

Wild cat - Fe/is silvestris Schreber, r 777 Remains of the wild cat were found both in Star­ cevo and Padina. One mandibula in Starcevo (fig. ro) . In Neolithic Padina (III) four bones may have belonged to the same animal. Of the distal part of a humerus, the suture was closed. In a domestic cat this happens approximately at the age of 8! months. On the distal half of a radius the suture was open. With the domestic cat it closes at ap­ proximately r r ! months, and of a femur the suture of the trochanter major was still open. In a dom­ estic cat this suture closes at approximately 8! months (Habermehl, r 97 5 ). The fourth bone is a metatarsus IV. A metatarsus II may have belonged to a second animal. It seems that the first wild cat was killed when it was approximately 8f-9 months old. The young of the wild cat are born in April or May, so that this animal was caught in winter. The biotope the wild cat prefers is wood with open spaces, or wood steppes. The meat of the animal is good for eating, and the fur would also have been useful. 0 Sem Lynx - L) 'llX b111x Linnaeus, 1758 Remains of the lynx are found in both Mesolithic (I) and Neolithic (III) Padina. In the first phase a Fig. 12. Padina, L) 'llX (J•11x - humerus, found in sector II, Late l'viesolithic (photo : C.F.D" R.U. Groningen). humerus (fig. r 2) and a metatarsus II, in the second a small part of the cranium - a I. os temporale - and a calcaneus. The distal end of the humerus was Horse - Eq1111s prze1valskii Poliakov, r 8 8 r burned black. It seems that also the lynx was used Wild ass - Eq1111s l?J 1dr1111ti1111s as food. The lynx is living in woods. During the Mesolithic and Neolithic two equid species lived in Central and Southern Europe, the

Brown bear - Ursus arctos Linnaeus, r 7 5 8 wild horse - Eq1111s prze1valskii - and a wild ass or Of the brown bear one metacarpus V came from a zebroid equid - Eq1111s l!) 1dn111ti1111s. Remains of both pit in Starcevo. Around Padina both in the Meso­ species may be present. lithic as well in the later periods bears must have In pit four (B) in Starcevo a little used incisor been hunted occasionally. All remains belonged to was found of a young animal, indicating that the adult animals. A skull fragment, fragments of animal it belonged to was young when it died. A lower jaw, loose elements of the lower jaw, as well fragment of a calcaneus and a phalanx II were as fragments of long bones (fig. r 3) with carving found in the same pit. Both bones belonged to a marks and burned patches show that the b,ears were horse4). In unit I a molar of a maxilla and an as- Parli11a a11rl Starcevo : Ga111e, Fish a11rl Ca/tie

Both horse and wild ass could have been used as fo od, the bones could have been worked into tools, and the hides into garments.

W ild bo ar - S11s scrofa Linnaeus, q 5 8

Domestic pig - S11s scrofa rlo111estic11s In both settlements remains of wild boar as well as those of domestic pig have been fo und in small numbers (fig. r4). It is not always possible to es­ tablish with certainty whether a bone belonged to a domestic or a wild animal. It seems that a fe w bones found in Padina both in Mesolithic as in Neolithic levels are so srnall that they must have belonged to domesticated animals. Probably also in the ca ve of Icoana in Romania remains of domestic animals were fo und in Mesolithic layers (Bolomey, r 973).

0 5cm

Fig. 13. Padina, Urs11s arc/os, a - mandibula, Lare Neolirhic -

Early Iron Age; b - u lna, Late Mesolithic; c - calcaneus, La te Mesolirhic (photo: C.F.D" R.U. Groningen). tragalus, both of a horse, were found. A humerus, radius, pelvis and phalanx I of a horse were col­ lected from unit P. A 1VI2 from unit Q could have belonged to an Eq1111s h] du111ti1111s, but this is not 1 certain. In unit R a r. M , a M2, an incisor of the mandibula or maxilla, the distal part of a radius, a dis tal part of a tibia of which the su ture was not yet closed (the dis tal su ture of the tibia of a horse closes at the age of 2 years; this was a young animal), an astragalus, two phalanx I belonged to a horse. In Padina I ar. calcaneus of a horse was fo und. It could not be rneasured. In the layer belonging to the Roman period a part of a right mandible of a horse was fo und with P 2 and P 3. In this period the animal was probably domesticated. h Eq1111s l!)'drtt11ti1111s is thought to have been a steppe animal, suggesting that the inhabitants of 0 5 10cm Starcevo were able to hunt these animals in the vicinity of their settlement, but the environment of Padina seems to have been a less suitable biotope Fig. 14. Starcevo, S11s, a - maxilla o (D bag 387); b - maxilla

A number of measurements of the bones from dentition of maxillae and mandibulae (table 36, 3 7).

Starcevo were compared with those of pig remains The presence of two animals ca. r month old, found in the Vinea and La Tene layers of Go­ gives a slaughter period of late spring or early molava (Hrtkovci) and layers of Roman Sirmium summer. (Sremska Mitrovica)5 . Some of the bones found in Pigs live in mixed deciduous forests with under­ levels II and I and trench A were of small sized brush and water in the vicinity. The marshy flood animals and are comparable with those of the La plain of the Danube would have been a suitable Tene and Roman layers of the other sites. They are biotope for pigs. The environment of Padina prob­ an indication that level II and I contain material ably offered a less favourable habitat for this from younger periods (tables 38-4 1). species either wild or domesticated. In Mesolithic Padina (I) r 5 canines of the lower The meat of pig is good and the bones as well as jaw of CS wild boar were found of which r r were the canine of the mandibula of the male wild boar worked into tools. From the Neolithic (III) layers have been used to make tools. The skin could have only 3 worked C fragments were collected. been made into leather and used for clothing or The age at which the animals were killed could other purposes. in some cases be established by the dentition of upper and lower jaw and by the state of fusion of Roe deer - Capreo/11s capreolus Linnaeus, r 7 5 8 the distal and proximal ends of the long bones. Roe deer remains were found in both sites in small numbers. In Mesolithic Padina a maxilla and a man­ Late Mesolithic dibula with mature dentition were collected and a single p2. Apparently also young animals were S11s scrofo caught. In Neolithic (III) Padina a mandibula with Mandibula r. I. r./I. mature dentition was collected. Roe deer remains ivl1 J\•12M3 2 2 years or alder were more numerous in Starcevo (fig. 22). pars incisiva ? Roe deer live in a large variety of biotopes from S11s sp . Maxi Ila the lowland up to the timberline in mountains but l 2 3 p p p 1Vl1 ? > 2 years favour open country with good cover. Mandibula P1P2P3i'"l 1 3 months M1J\'l2M3 1+1

Early Neolithic

S11s scrofo Mandibula MI M2M3 2+ 1 d I d 2 years or alder S11s sp . Mandibula M M M I 2 3 2 (I) 2 years or alder

Of one of the humeri the proximal suture was still open, and of three others the sutures were closed.

Of the tibiae 1 proximal suture was closed, of the distal end, 2 were found with closed and 2 with open sutures. With domestic pigs the proximal su­ tures of the humerus and tibia close at the age of 3-t years, the distal suture of the tibia fuses at the age of two years (Habermehl, 97 5 ) . A small r 0 Sem number of bones belonged to domestic pig.

In Starcevo half of the animals were killed before Fig. 1 5. Starcevo, Cenms elaph11s - skull of a stag (V bag 34 5) reaching the age of two years, according to the (photo: National M useum, Belgrade). Padi11a a11dS tadevo : Ga111e, Fish and Catt/e

Neolithic

Maxilla p2p3J'v[ l i\fandibula P2P3P4M1 M2M3 6 2 years or older

l'v!ost bones belong to the post cranial skeleton. The almost complete absence of skulls, maxillae and mandibulae is conspicuous. In Starcevo one fragment of a cranium was found but no antlers that were still attached to the skull (fig. 1 5 ). The antlers were all fragmented and

0 5 10cm

Fig. 1 6. Starcevo, Cerv11s elaph11s - natura I shed antiers (J bag 379) midden overlaying pit 6 (photo: National M useum, Bel­ grade).

e

Red deer - Cenms elaph!IS Linnaeus, 175 8 In both settlements the remains of red deer are the most numerous of the wild mammals. In Padina it was also the most important mammal species, with percentages of 78.8 (I), 70.6 (III), 75.6 (V) (table 2) . In Starcevo only maxillae and mandibulae of mature animals were found (table 36, 37). The same was true in Mesolithic Padina (I), but in the Late Neolithic (111) a maxilla was found with an imma­ ture dentition and one loose p1 or p2. The man­ dibulae had a mature dentition.

0 5cm Lare i' viesolithic

Fig. 17. Starcevo, Cenms elaph11s, a - mandibula (E bag 412); b Mandibula r. I. r./l. - radius prox. (D bag 387); c - radius

it could not be established whether or not they Roman layers of nearby Sremska Mitrovica5) - belonged to shed antlers, with the exception of Roman Sirmium - seem to be smaller than those of three shed antlers collected in the midden over­ Vinca, Gomolava and Padina. Apparently the low­ la ying pit 6 (fig. 16, 9). land conditions in the Neolithic period were more In Late Mesolithic Padina (I) four antlers were fa vourable for red deer than in the wooded moun­ still attached to a skul! fragment and there were tains around Padina. This changed later. Red deer four shed antlers. There were also 98 unworked can live in a variety of biotopes but favours open an tier fragments and l 6 worked fragments. Of woods with undercover and open spaces. They live these fragments it could not be established whether in the lowland as well as in rnountains to an altitude they belonged to shed or unshed antlers. The four of 2000 m. shed antlers show that antlers were collected. Red deer meat has a good flavour, the bones and In the Early Neolithic (III) layers one antler at­ antlers can be used for tools, the bide can be used to tached to the skul! was found, and three natura! make clothes, shoes and other objects. shed antlers. There were 3 7 unworked antler frag­

ments and IO worked ones. Aurochs - Bos pri111ige11i11s Boj anus, 182 7 Most of the long bones were broken (fig. 17). Domestic cattle - Bos ta11ms Linnaeus When we compare the rneasurements of the long In both Padina and Starcevo were remains col­ bones from Padina and Starcevo with those from lected of wild and domestic cattle but in Padina the Vinca levels of the dwelling mound of Go­ neither of the two occurred in large numbers. molava in the Vojvodina on the Sava, it appears In Starcevo domestic cattle are the most important that the lowland red deer were slightly larger than domestic animals (fig. 18, l 9, zo). A number of those hunted by the inhabitants of Padina during rneasurernents of the bones from Starcevo were the Mesolithic and Neolithic. The measurements presented in table 42-49 which can also be com­ pared with those of red deer rernains from Obre in Bosnia (Bokonyi, r976). The measurements of red deer found in Obre I (late Starcevo) and Obre II (Butmir culture) in Bosnia (Bokonyi, l 976), seem to range between the rneasurements of the gorge and the lowland, while the means of Obre I are closer to those of the gorge, and the rneans of Obre II closer to those of the lowland sites of Starcevo and Gomolava. Compared with the measurements of red deer in Neolithic Romania and Hungary (Pietschrnann, l 977), it appears that the Rornanian red deer might have been even slightly larger than those of the lowland deer of the central Danube basin. The rnaximaof both Hungarian and Romanian measure­ ments are higher than those from the gorge. Ac­ cording to Pietschmann there is a dwarfing of red deer from the South-East to the North-West, and on the Peninsulae of Greece and Italy from North to South. Other investigators have also noted the phenomenon and commented upon it. There is also a dwarfing through time which, although not ap­ 0 5 10cm parent in Obre, probably can be observed in other

later sites in Yugoslavia. The red deer remains of Fig. 18. Starcevo, Bos la11ms - horncores (U bag 498) (pboto: 160 rhe La Tene layers of Gomolava (Clason, l 979) and National J\Iuseum, Belgrade). Padi11a a11d S tareevo : Ga111e, Fish a11d Catt/e

0 5 10cm

compared with those of cattle remains found in the Fig. i9. Starcevo, Bos !r11m1s - horncores (I bag 381) (photo: Vinca and La Tene layers of Gomolava (Hrtkovci) National Tvi useum, Belgrade). and layers of Roman Sirmium (Sremska JV Iit­ rovica). Same of the bones found in levels II and I and trench A were of smal] sized animals and are the multiplying factor for bulls these figures would comparable with those of the La Tene (Clason, have been slightly higher. In Gomolava withers 1979) and Roman layers5) of the other si tes. They heights of rn2 and 12 5 cm for Vini':a level animals, are also an indication that level II and I contain and 92.0, 105.0, rn9.o and 113.0 cm for La Tene material from younger periods (tab les 5 0- 5 8, 60 ). cattle were fo und. In Starcevo domestic cattle were slaughtered at \X/ild cattle could have lived in a variety of bio­ different ages. According to the dentition of the topes, but it is thought that they preferred open mandibula ca. 50% of the animals were killed when woods and open areas with same cover. 3 years old or alder. Of the aurochs not only adult The aurochs and domestic cattle would have animals were hunted but also calves (table 37). In served as food, the bides could have been used for Padina cattle remains are found in such small num­ variety of purposes, the horns could have been bers that it is not possible to say anything about the made into drinking beakers and containers, while slaughter age. It seems however that most of the the bones were used to make tools. bones belonged to mature animals.

In Padina no complete metapodiae were found, Sheep - Ovis aries Linnaeus so that it is impossible to say much about the with­ Goat - Capra hirms Linnaeus ers height of cattle. In Starcevo there was a more or In Padina a small number of bones came from small Jess undamaged metacarpus in level III (200.0 mm) ruminants. In Starcevo after domestic cattle they and another in trench B ( 194.0 mm). This gives a were second in importance. In Padina sheep were withers height of ca. 120 cm and 116-4 cm when the certainly kept, for goat this is not certain although factors of Fock (1966) are used. In a pit a complete a humerus of phase III may be of a goat. metatarsus was found with a length of 2 2 3 .o mm, In Starcevo both sheep and goat were kept. Horn­ giving a withers height of 119. 3 cm. If we had used cores of both species were found. Sheep seems to 161 A. T. CLASON

Radius H-47 cm J'vfetacarpus 62. j, 57.j, 56.9 cm Tibia 40.2 cm J\iferatarsus 66.o, 58. 1, 56.9, 5 7.6 , 5 9-4, 56.2 cm Astragalus 74.8, 68.o, 61.2, 73.7, 70. 1 cm Calcaneus 76.9, 71.2 cm The multiplying factors of Teicherr (1975) were used.

Teichert (1975) bas mentioned that the astra­ galus and calcaneus are not very reliable for tbe calculation of withers height, because they re­ present only a smal! part of the height of the animal. A second reason for the difference in withers height may be that not all the bones are from sheep but some may also have belonged to goat. To calculate the withers height of goats, multiplying factors were workecl out for humerus, radius, ,;;�'-· femur, tibia and the metapodiae by Schramm (Von den Driesch & Boessneck:, 1974). The number of :�"; recent goats for which these factors were calculated � is small. With the exception of the factor for the metatarsus, the calculated multipl ying factors are lower for goat than for sheep, so that if we suppose that all bones were from goats, the withers height

0 5 10cm

Fig. 20. Starcevo, Bos ta11ms, a, b - mandibula (D bag 387) ; Bos prillligeni11s, c - radius dist. (N bag 417); Bos la11ms, d - cal­ caneus (D bag 387) ; e - metacarpus prox. (D bag 387) ; f, g - metatarsus prox. (D bag 387, Q bag 361); h - metacarpus dist. (D bag 387); i, k - metatarsus dist. (D bag 387) (photo : National Museum, Belgrade).

have been heavily horned (fig. 21, 22) . Although it is not always easy to separate sheep from goat bones, we get the impression that sheep were the more numerous. \Xfhen we look at the dentition of the mandibula, 46.9% of the animals were slaugh­ tered befare reaching maturity. The number of animals ca. 3-6 months old is slightly higher than those of the other immature groups, which may indicate a slaughtering of lambs in autumn. The animals seem to have been of smal! stature. The withers height calculated for a metacarpus of 0 5cm Padina V is 52. 3 cm. In Starcevo two groups of heights were found: 1 ) calculated for the long bones; 2) calculated for the calcaneus and astra­ Fig. 2 1. Starcevo, Ovis aries - skull (G bag 4 11) (photo: 162 galus. National Museum, Belgrade). Padi11a a11d Starcevo : Ga111e, Fish a11d Cattle

skul!, of a third only the tip was found (fig. 23). It is possible that some of the bones assigned to sheep/goat are from the chamois, but it does not seem very likely. According to Van den Brink ". ( r 968) the chamois has a withers height of 7 5-8 5 cm and is larger than the European wild sheep, the I \ '" moufflon, which has a withers height of 65-75 cm. ' .� ·:"'� f Also remains of R11picapra were found in other gorge settlements. Today the chamois lives in the woods of medium high mountains up to the lower meadows at an altitude between 800-2 300 m. In winter they e live only in the woods, which can be of mixed de­

ciduous woods or pine (conifer) in character.

d c

Som

Fig. 22. Starcevo, Caprco!11S rapreo!llS, a - an tier (J bag 3 79); b - radius (E bag 412); c - pelvis (E bag 412); Ovis aries, cl - horncore(V bag 346); Capra/Ovis, e - manclibula(D bag 387); f - humerus (D bag 387) ; g - radius (D bag 387); h - meta­ carpus (E bag 41 2); i - tibia (D bag 387) (photo: National !vluseum, Belgracle).

that could be calculated for the long bones would be lower. In Padina the bones of sheep/goat could also have been confused with those of the chamois of which three horncores were found. The withers height calculated for the one metacarpus of Padina is too lo>v for a chamois. The same seems to be true of the other measurements. That the Padina bones are from an ibex seems very unlikely, for these are large animals with a withers height of 105-125 cm �� and 130-150 cm 00 (Van den Brink, 1968).

0 Sem

Chamois � R11picapra mp icapra Linnaeus, r 7 5 8

Of the chamois three horncores were found m Fig. 2 3. Paclina, R11pimpra mp irapra - 3 horncores, Late Mesol­ Padina I. Two were still attached to a part of the ithic (photo: C.F.D., R.U. Groningen). A. T. CLASON

The chamois has fine tastingmeat, and the hi des, bones and horns can be used.

3.2. Birds

Both in Padina and Starcevo small numbers of bird bones representing at least 20 species \v ere collec­ ted. Most of them could have been used for food, while the long bones could have also provided raw material for tools. In a number of cases the fe athers could have been used for adornment, and other unknown purposes. (General literature : Bauer &

Glutz von Blotzheim, r 966; Heinzel, Fitter & Pars­ e low, 1972).

Black-throated diver - Gavia arctica (Linnaeus) The proximal part of an ulna came from Padina I. This was larger than those of the red-throated diver in the collection of the B.A.I. Because the black­ throated

Cormorant - Phalacrocorax carbo (Linnaeus) A coracoid of a cormorant was found in Padina I. This species breeds nowadays in Yugoslavia. Birds from Northwestern Europe also move southward c in winter. The shag - Phalacrocorax aristotelis - is smaller and almost exclusively marine. The cor­ morant feeds on fishes. 0 Sem

\Xlhite pelican - Peleca1111s 011ocrota/11S (Linnaeus) A coracoid of the white pelican was found in Fig. 24. Padina, Pderr11111s 011offotn/11s, a - epistropheus; b - )rd

Padina III. An epistropheus, third and fourth ver­ vertebrn; c - 4th vertebrn, not dated; Conms fmgileg11s, cl - tebrae of the neck, also from Padina, are undated humerus, Early Neolithic; cf. Hnlineet11s nlbicilln , e - ulna, Early eolithic (photo: R.U. Groningen). (fig. 24). These vertebrae belonged to the same in­ C.F.D" dividual. At present the species nests in Bulgaria and Ro­ mania, but during prehistory it could also have with certainty. In Starcevo three carpo-metacarpi, nested in Yugoslavia. It is found in inland water, which were fo und in level II, and unit J and K also marshes and shallow lagoons. probably belong to the mute swan. The mute swan lives along coasts and on fresh water lakes.

Mute swan - C)ig1111s olor (Gmelin)

A scapula fragment of mute swan came from \Xlhooper swan- Cjig1111s C)'glltts (Linnaeus) Padina I, an undated tibio-tarsus was not identified A tibio-tarsus found in unit B of Starcevo belongs Padi11a a11d Stareevo : Ga1J1e, Fish a11d Cattle

probably to a whooper swan. The species breeds at present in Northwestern Europe, but migrates southward during winter. The swan was probably eaten.

Grey-lag goose - A11ser a11ser (Linnaeus) From Padina I a proximal and a distal part of a fe mur of the grey-lag goose represent two different birds. In Starcevo of this species, a carpo-meta­ carpus in unit K, two ulnae in units H and S, and one undated radius were found. At present the grey-lag goose breeds in Eastern and Central Europe, nesting in rnarshes. During its migration period it looks for open spaces in the vicinity of the coasts, rivers and marshes. The en­ vironment of Starcevo would have been more suitable than that of Padina. The goose would have provided good food, fat and feathers.

Grey-lag goose or bean goose - A11ser fa balis (Latham) Two humeri and an ulna of the grey-lag goose or bean goose (Amerfabalis) were fo und in unit B and D of Starcevo. This species nests in wooded areas in Northern Europe but migrates to the south in the winter where it favours the same habitat as the grey-lag goose. The skeletons of both species are 0 Sem much alike.

Shell-duck - Tadoma tadoma (Linnaeus) · A humerus of this species came from Padina III. Fig. 25. Padina, HaliaeC'/11s a//;iti!la , a - rnd phalanx, Late Me­ The shell-duck breeds along the coasts of North­ solithic; b - tibio-tarsus, Late Mesolithic; vulture, c - hu­ western Europe, but rnigrates to the south during merus, Early Neolithic (photo: C.F.D"R.U. Groningen). the winter. It is mostly observed on plains near marshes and open water, but also in various other biotopes. probably a sea-eagle, came from levels of Padina I; an ulna from Padina III (fig. 24).

Shoveler - A11as cf. c(Jipea!a Eagles could have been hunted primarily for A humerus from unit I of Starcevo probably be­ their feathers, which could have bee11 used in longs to a shoveler. The shoveler nests not only in making arrows. extreme eutrophic inland lakes in Northwestern

Europe and the central Danube basin but also in Red or black kite - jl1ilv11s 1J1i/1;11s Linnaeus - Nli/1;11s marshy areas. Its presence in Starcevo is therefore 1J1igra11s (Boddaert) not surprising. A carpo-metacarpus of a kite was found in Starcevo in a trench excavated in 1969. An ulna comes from

Sea-eagle - Haliaei:/11s albicil/a (Linnaeus) unit S (trench A). It is not certain whether this ' A rnetacarpus, a fe mur, two tarso-metatarsi, I bone belongs to the Starcevo habitation period. phalanx I and 2 phalanx II of a large bird of prey, There is not much difference between the red and A. T. CLASON

black kite but the biotope of Starcevo, both in pre­ Raven - Conms corax Linnaeus history and later, seems to indicate that we have to A dis tal part of a tarso-metatarsus of a raven comes do with a black kite. The black kite nests in woods from the Late Mesolithic levels of Padina I and near lakes and marshes. Neolithic Padina III. The raven nests on high cliffs and in high trees. Its presence in Padina is therefore

Eagle - Aq11ila sp. not surprising. A radius of a young individual that could not be The raven is a scavenger. It was probably eaten. identified as to species was foundin Starcevo unit

F. Rook - Corv11s jmgilegus Linnaeus A humerus was foundin Neolithic Padina (III) (fig. Bird of prey - cf. Circ11s 24). The rook also needs trees for nesting, but it The ulna of a young bird found in Starcevo, unit J, feeds in more open areas than the raven. may belong to a harrier. A phalanx of a smal! bird

of prey from Padina I could not be identified. Domestic fowl - Ga//11s gall11s domestictts Although more than one bone of domestic fowl Vulture came from Starcevo, a humerus in level I, one The humerus of a vulture is from Padina III (fig. tibio-tarsus and one femur in level II, none were 25). It could not yet be identified as to species. found in level III. This agrees with the supposition that the higher levels also contain younger ma­

Crane - Gms gms (Linnaeus) terial. The domestic fowl could not have been Two femora of a crane were found in Starcevo, in found in Central Europe befare the Late Bronze unit I and M. At present the crane breeds in Nor­ Age, when it is found for the fi rst time in Southern thern Europe and in Yugoslavia, nesting on the Europe (Gejvall, r969). It was introduced from the ground in humid areas, marshes and reed fields. In South and was domesticated in India. winter it is to be found in open country along rivers. Its meat is very good and tasty. H· Reptiles

Great bustard - Otis tarda Linnaeus The remains of only one species were found. Also found in Starcevo were bones of the great

bustard. A carpo-metacarpus in unit I, and five European pond terrapin - E1101s orbiclllaris Lin­ fe mora. Two in unit J, one in unit I, one in unit K naeus, 1 75 8 and the fifth in the trench opened during the ex­ Of the European pond terrapin both in Starcevo cavation of r 970 . and in Padina fragments of the carapace have been The bustard nests on the ground in Central and fo und. The terrapin lives in stagnant or slowly Eastern Europe and is nowadays to be found on flowing water and lakes which have a rich, open open treeless plains or cornfields. Even today the vegetation. The species hibernates in the mud of bustard survives in the central Danube basin. Its the water in which it lives from September/ presence in prehistoric times points to rather open October until March. Spring and summer are the country vegetation. seasons during which they could have been caught. The bustard must have provided good food. 3 -4· Fishes Curlew - JV111m11itts arq11ata (Linnaeus) A tarso-metatarsus of a curlew was collected in In both Starcevo and Padina, but especially in pit 4 (unit B) in Starcevo. Today the southern Padina, fish must have formed a substantial part of boundary of this species is the Danube and the the diet of the fo rmer inhabitants, and fishing coasts of the Black Sea. At present it still breeds in would have been an important occupation. At pre­ Hungary, making its nest in open country in the sent five species have been identified at Padina and vicinity of water and feeding at sparsely vegetated also five at Starcevo. Only the remains of large 166 grounds which are sometimes flooded. catfish were found in both settlements. In Starcevo Padi11a a11d Starce vo: Ga111e, Fish a11d Cattle

there were only two species in the material exca­ April-July in the middle course of the Danube. It vated in 1932 and five in the trenches excavated in was commercially important until recently. Togeth­ 1970. Of these last it is not certain whether they er with the beluga it would have provided a large belong to the Starcevo Neolithic material, the more percentage of food in spring and eady summer. so since even today all have economic value. \Xfhether the inhabitants of Starcevo and Padina The fish bones of Padina were studied provi­ had the knowledge to preserve fi sh is not known. sionally by drs. D. C. Brinkhuizen to get a fi rst Probably they did. Suggested by the specialised impression of the fish species present. He will look character of the hearths in the houses of Padina at them again in future. It seems quite probable that which are thought to be designed for drying and it will then be possible to identify more species. smoking fish. The fish bones from Starcevo were identified by

Dr. J. Lepiksaar. Danube salmon - H11cho h11cho Remains of the Danube salmon were found at Sturgeons Padina I and III. This species can attain a weight of l 2-32 kg. At present this fish is fi shed by sport­ In the Danube and its tributaries six species of stur­ fishers. geon came, or still come upstream to spawn during

spring and eady summer (Terofal, l 97 l ). These Pike - Esox /11ci11s are : Two bones of pike were found in Starcevo in the trench excavated in 1970. The fish can reach a Beluga - H11so h11so - max. 8.5 m, 1300 kg length of 30-120 cm and a weight of 34 kg. It lives Common sturgeon - Acip e11ser s/11rio - iS 2 m, 'jl 6 m, 200 kg in stagnant or slowly flowing water. It is easy to

Sterlet - Acip emcr mthel/ltS - 'jl 1 m, 6-1 o kg catch and has a pleasant taste. Stellate sturgeon - Acip e11ser stella t11s - 'jl 2,2 m, 68 kg Russian sturgeon - Ac1pe11ser g11elde11s/aedti - 4 m, 160 Asp - Aspi11s aspi11s kg One bone of an asp was found in Starcevo in the Ship sturgeon - Arip e11ser 1111dive11/ris - 2 m, 40-50 kg trench excavated in l 970. The asp lives in the middle course of rivers in most of Europe. It can

It is established that in Padina at least the beluga, reach a length of l 20 cm and a weight of l 2 kg. the common sturgeon and the stedet are represen­ Even today it is an important food fish. ted. It is possible that the remains of the other

species are also present. Bream - Abra111is bra111a Two fragments of bream were found in the trench

Beluga - H11so IJ 11so excavated in l 970. The bream is living in Central The remains of this big Danube sturgeon were and Eastern Europe in slowly flowing rivers and found in Padina I and III. This species lives in the lakes. In Central Europe today it has a large com­ Black Sea and the Adriatic, entering the Danube in mercial value. May to spawn.

Today this fi sh is of enormous economic impor­ Carp - C) pri1111s ca1pio tance for its meat is good and in addition the �� One carp bone was found in pit 5A, unit D of provide caviar. Starcevo and three in the trench excavated in 1970. In Neolithic Padina (III) 3 2 carp bones were found.

Stedet - Acipe11ser mthe1111s This species can reach a length of 75 cm, and a In Padina a fe w remains of the stedet were found. weight of 19 kg. The carp lives in lakes and slowly It lives in the northern part of the Black Sea as well running rivers. It still is an important food fish, as as in the rivers. It is still an important food fish. it was in prehistory.

Sturgeon - Acipemer st11rio Carp fi shes - C)pri11idae This sturgeon lives in the sea and spawns from From both Padina (I, II and III) (fig. 26) and Star- A. T. CLASON

settlements, but only in Padina was a large number of shells preserved. These snails are best for eating in the autumn.

Unio - U11io crass11s crass11s Retzius, 1788 0 2cm Big unio shells are guite common in prehistoric settlements in Central Europe. Shells were found in Fig. 26. Padina, ske1eta1 element of a cyprinid (photo: C.F.D., both sites, but most of those available were prob­ R.U. Groningen). ably thrown away. Dr. Van der Spoel, who identi­ fied the Starcevo shells, pointed to the faet that they were damaged at the points where they could cevo bones of carp fishes still have to be further have been opened most easily, at the ends of the identified. shells. It is not known whether they were eaten. It is also possible that they were only collected for

Ca tfish - Si/11ris gla11is their attractive mother of pearl layers. Since how­ Catfish bones were found in several of the Starcevo ever no mother of pearl objects are known from pits and in the trenches excavated in 1969 and 1970. Padina or Starcevo, it seems more likely that they In Padina the majority of the fish bones collected were used as food. Another possibility is that they belongs to catfish (fig. 27). served as bait for fishing. In Starcevo 3 shells were

Catfish can reach a length of 4 meter and a collected : 2 in unit I and r in F. weight of 300 kg. The length of the fish can be

calculated from the width of the first vertebra, the Vivipams vivipams Linnaeus, l 7 5 8 atlas. In Padina 41 atlases were collectecl of which Three shells came from Starcevo. This species lives two could not be measured. From the width of the in water and empty shells were possibly washed others the following lengths have been calculated ashore. According Dr. Van der Spoel the following three points suggest this : 1) the shells were large 89.1, Il3 .8, I 14.2, 118.o, 127.0, 13o.3, I 3 2. 2, l 3 3.6, I 34.7, and of old animals; 2) in the clay in the shells, l 3 7. 3, l 38.7, 142.7, 144.1, 146.1, 147.3, 147.7, 148.4, 148.7, I, I I. I l fragments of other shells were found ; 3) only in l j 1. j 2, I j 2.6, I j)· 3 > j j -4, j 5 -4, I j 8.6, I 61. 3, 163.7, _ I 163.7, 164.8, l 68.5, I 68.9, '71.2, 175 .o, l 8 i. 3, 82.7, 194.5, one case the way in which the shell was damaged 194,5 zoo.o, 204.0. could indicate that the body might have been taken out by man. From this list we get the impression that the !argest catfish were not caught. However, although the smaller ones are also missing from the series, it 4. DISCUSSION seems guite possible that the remains of the smaller

catfish were thrown away or missed during the ex­ 4. I. Near East cavation. Catfish live in large rivers and muddy lak.es. It is by now well-known that in the Near East some They are economically important today. They can 10.000 years ago hunter-gatherers turned into agri­ be caught with baited fi sh-hooks, and large hooks culturalists. This means that they isolated small were found in Starcevo. groups of several animal species from the main herds and brought them under control. These

3. 5 . Molluscs species at first were wolf - Ca11is !11p11s, sheep - Ovis

a111 111011, goat - Capra hirc11S, later also the wild ox -

In both sites shells and bivalves of molluscs were Bos pri111ige11i11s - and the wild boar - S11s scrofa . This collected. process, which would have been gradual, we call domestication.

The edible snail - Helix sp. \Xfhat first caused this domestication is not yet 168 The edible snail must have been numerous in both clear, although it seems that the possession of a Padi11a a11d Stadevo : Ga111e, Fis/i a11d Catt/e

0 5 10cm

permanent settlement may have been one of the Fig. 27. P;!dina Si/11ris g!a11is - verrebrae and other parts of the prereguisites for its beginning. skeleton (photo: C.F.D" R.U. Groningen). In the Near East at the end of the last Ice Age the weather changed, gradually becoming warmer and more humid. The artemisia steppes disappeared population-pressure would lead in the end to stock­ and were followed by grass steppes or open woods. breeding and land cultivation to enable the people In the same period hunters and gatherers in this to keep their sedentary way of life. This prncess area exploited their surrounding ever more effi­ coulcl have been repeated again and again, and ciently. Not only did they hunt big mammals but spread from the Near East and from possible other also birds, smal! mammals, reptiles and amphibians nuclear areas in Europe, Asia and Africa, until by together with the gathering of molluscs and crus­ 2000 B.C. in Europe all the areas that were then taceans, and fishing. In favourable areas this led to suitable for cultivation, were occupied by fa rmers. a surplus of food which thus enablecl man to chose a permanent habitat. \X/ildgrains could then be har­ 4. 2. Southern Europe vested and stored away. One interpretation is that this favourable fo od In Southern Europe no permanent open-air settle­ situation led to a denser human population, which ments of hunter-gatherers are known at the end of led in the end to overpopulation, which led people the Mesolithic (Epipalaeolithic) outside the Iron to migrating to other places where the possibilities Gate area. Even though caves would have been for hunting and gathering were Jess fa vourable. In more or Jess continuously inhabited from the Late order to maintain the sedentary life to which they Palaeolithic (Epipalaeolithic) til! the Neolithic or had become accustomed they started to keep the even la ter. One of the olclest known fa rming settle­ animals, which were abundant in the area they ments in Europe is that of Lerna on the Pelopon­ came from. Also in the nucleus area the constant nesus (Gejvall, r969). As early as the seventh mil- A. T. CLASON

lennium B.C. the inhabitants of Lerna kept cattle, Vardar which would account for the location of sheep, goat, pig and dog. More to the North the Anzabegovo and Vrfoik. Yet a fourth might have inhabitants of Argissa l\fagula builthouses and also gone further up the Struma crossed the divide kept domestic animals in the seventh millennium either from the headwaters of the Struma or B.C. (Boessneck, 1961, 1962), although they did possibly the upper Timsk or Ogosta to run west­ not know pottery. Also at Nea Nicomedeia in the ward through the rugged gorges of the Nisava to plain of Macedonia the people kept cattle, small the Morava valley. ruminants, pigs and dogs, at the end of the seventh So far in Bulgaria the Starcevo culture seems rnillennium, as did the inhabitants of Knossos on limited to the northwest section, with some occur­ the Isle of Crete in the Aegean (J arman and J ar­ rences in the hill caves near Sofia, and along the man, 1968). western border. To the East the K.aranovo culture Far to the Northeast, in the Masica valley, at the characterizes the Maritsa valley which drains to the tell of K.aranovo in Bulgaria, the inhabitants of the Aegean, while the Starcevo-K.bros-c;:ris complex is earliest settlements of around ca. 5 5 oo B.C. kept definitively middle Danubian. domestic animals. In none of these settlements was hunting important. Sheep (and goat) were the most 44 The Iron Gate important domestic animals. The oldest known farming settlements in Ro­ The Iron Gate is a gorge approximately roo km mania in the lower Danube plain are as old as those long, cut into the Southern Carpathians by the of K.aranovo and belong to the c;:ris culture. The Danube. At many places steep cliffs reach to the c;:ris culture farmers kept cattle, sheep, goat, pig water, while at others, where small rivers join the and dog. In the central Danube basin lived at the Danube, small loessic terraces were formed. These same period the farmers of the Starcevo-Koros cul­ terraces may be up to twenty or thirty kms long ture, which was closely related with the c;:ris cul­ and are some hundreds of meters deep. Low moun­ ture (Tringham, 1 971). tains or high hills behind the terraces isolated them Neither on the lower Danube plain nor in the from the outside world. central Danube basin do we know of permanent Some of these terraces were recognized by Late settlements of hunter-gatherers. It seems plausible Mesolithic hunters and fi shers as good places for therefore that farmers entered those sparsely popu­ hunting and fishing and in some the conditions lated areas and settled at favourable spots. were so favourable that the hunter-fishers could build permanent villages. One of these is Padina, 4. 3. The rou te to the middle Danube basin others are Lepenski Vir and Vlasac also on the Yugoslavian right bank, and Cuina Turcului, Ve­ There is more than one possible way by which the teroni, Icoana and Ruzvrate on the Romanian left Early Neolithic people could have entered the side of the river. rniddle Danubian basin. The Danube Gorge itself The oldest habitation is found in the cave of would have been virtually impossible, but there Cuina Turcului at ca. 10650 B.C. The hunting bag may well have been a traversible passage either to consisted at least 14 marnmal species of which re­ the l orth or South of it, that linked the lower and mains of beaver and wild pig were in the rnajority, middle Danube basins. followed by chamois and ibex. Birds were also A more traditionally accepted means would be hunted and shell-fish collected. Fishing was of no by the so-called Vardar-Morava route from the importance. After a hiatus of 2 5 oo years the ca ve Aegean. However, here too the terrain is extremely was again inhabited and sorne of the same species rugged and it would have been necessary to by-pass were hunted. New species are red deer and pole cat, at least the Iron Gate of the Vardar and the gorges the hunting of wild boar increased and fishing was of the upper Morava south of Leskovac. basic (Bolomey, 1973)· A third possibility might have been via the During the following Late Mesolithic and Early Struma, then westward along the Strumitsa and leolithic periods both banks of the river were in­ 170 down the Bregalnica via the Ovce Palje to the habited. Faunal remains were collected during the Padi11a a11d Starl!e1 ;0 : Ga111e, Fish a11d Cattle

excavation of Icoana. Hunting, fishing and gather­ seems therefore more plausible that in the Iron ing of shell-fish were important. The inhabitants Gate Gorge, which vv as difficult to reach, the orig­ kept domestic dogs and the ages of the pig remains inal inhabitants of Central Europe found a refuge are suggestive of domestication : the faet that or could maintain themselves longer than in the 5 2.6% of the animals were slaughtered between 6 plains. The unexpected favourable food situation

and l l months of age, points to a type planning made it possible to build permanent houses and live that can only be effectuated with domestic animals during long periods in the same area. Contacts (Bolomey, 1973)· with the surrounding areas have however always existed and gradually the Neolithic way of life of 4.5. Padina the Starcevo farmers was also accepted in the gorge. On the Yugoslavian side among others were Padina, Vlasac and Lepenski Vir. The inhabitants 4.6. Starcevo of these settlements kept dogs and used them as fo od. In Padina it seems tbat domestic cattle, sheep, The first farmers in the central Danube basin be­ possible goat and domestic pig were also kept in longed to the Starcevo-Ki:iri:is-C,:ris complex like smal! numbers. Tbe inhabitants of Lepenski Vir those in Romania (C,:ris) and in Hungary (Koras). III, and those of Cuina Turcului III, IV and V did Starcevo was excavated in 1932, and it was then for tbe same. Hunting was important and a large the first time that finds belonging to what was later number of mammal and bird species were on tbe to be called the Starcevo culture were recognized as menu. In Padina remains of red deer were most Early Neolithic. Excavations of other sites have numerous, followed by wild boar. All species are indicated that an older and a younger phase are more or less indicative of tbe wooded surround­ distinguishable in the Starcevo culture. Starcevo ings of the settlement. Fisbing must have been very itself belonged to the younger phase. Thus the re­ important and probably provided the main source mains we are discussing in this paper are not those of food for the settlements. Fish remains increase of the earliest Neolithic domestic animals that were tremendously in Padina III, constituting 80% of all kept in Central Europe. bones collected. Although this may be the result of The remains show that domestic animals were a diffe rence in the selection of bones by the exca­ the main source of animal proteins in Starcevo. The vators, it may indicate an improvement in fishing more so since S11s sp., Bos sp. and Bos/Cervus are techniques. The species kept and hunted indicate included in the sum of group C - wild ungulates. If that the settlement of Padina was permanently in­ we look at the percentages in which the domestic habited during all the seasons of the year, and this mammals, the rodents, and carnivores and the wild is probably also true for the other sites in the ungulates were collected, then there is a marked gorge. Although few in number the faunal remains diffe rence between the pits, 6 5.2 % dornestic mam­ of period V seem to indicate that hunting also had mals, and level I, level II, level III and trench B, an important place in the Iron Age and Roman with percentages of 75 .3, 74.9, 76.5 and 78.0 and period, while fishing was no longer significant. We trench A - A extension and the trenches opened in must however keep in mind that only a small 1969/'70 with percentages of 88.7, 92.2, 86.2, 83.8 number was available for research and that not all and 8 7.2. It is tempting to explain this increase of bones were collected. the domestic animals as a sign of the increasing The main picture we get from the Iron Gate area importance of stock-breeding. 'v/e must however is that it is not a nucleus area of first domestication be careful not to accept this explanation too easily comparable with the Near East. Although Bokonyi because we have to remember that not all the bones (1975) made it acceptable that dogs were locally were collected, and we have also to keep in mind domesticated in Vlasac and the inhabitants of that the higher levels and the material from the Padina I and Icoana kept pigs, this domestication excavations of 1969/'70 are contaminated with and pig keeping did not take place before farmers younger material. If we assume that the same per­ settled in the lower and central Danube basin. It centages of the bones of the larger species were 171 A. T. CLASON

collected as were not collected, then the percen­ for the inhabitants of Padina where certainly tages more or less represent the proportions in during the spring and early summer fish and espe­ which the animals were killed and possibly kept. In cially catfish formed an important part of the food. that case the percentages of the pits and level I are As a last group we must mention the molluscs. probably representative for the Starcevo occup­ Probably, as in other settlements in Central ation, but not the percentages of the higher levels Europe, the molluscs were much more important and other trenches. This also seems to point to the as food than one can conclude from table 6. There faet that these two complexes are representative for must have been large quantities of shell of the U11io the Starcevo period. Also in some of the other Star­ and the edible snail which were not collected, or cevo settlements sheep remains were fo und in rela­ only in smal! nurnber, during the excavations. tive high numbers. If the plain had a more open vegetation than elsewhere in Europe, as the re­ 4. 7. Recapi tulation mains of horse and great bustard would seem to indicate, then the fa rmers of Starcevo could keep \,Y/e can say that it is only possible to guess at the sheep and goats, but mainly sheep in larger num­ total subsistence of the inhabitants of Padina and bers than was possible at other places. Domestic Starcevo. During Padina III there seems to have pigs were killed only in small numbers. The re­ been a major emphases on fish, while at Starcevo mains of wild boar are more numerous, and domesticated animals seem to have supplied a possibly it was easy to hunt them in the swampy larger share of protein than did hunting or fishing. flood plain of the Danube so that it was not neces­ Domesticated plants as well as gathered and cul­ sary to keep a large number of domestic pigs. \,'\/ild tivated plants seern to have been an important part boar and red

6. NOTES 4, pp. 353-366. CLASON, A. T" 1 967. A11i111al and ;;;a11 in Holland' spas!, A and B. Groningen. At that time Department of Anthropology, Brooklyn CLASON, A. T" 1979. The farmers of Gomolava in the Vinea College, New York. Current Research Associate, Peabody and La Tene period. Palaeohisloria 21, pp. 41-8 1. Museum of Harvard University. DRIESCH, A. VON DEN & J· BOESSNECK, 1974. Kritische An­ Director of the Archaeological Institute of the Yugo­ merkungen zur \'\liderristhcihenberechnung aus Liingen­ slavian Academy of Science in Beograd. massen vor- und friihgeschichtlicher Tierknochen. This was a salvage excavation that had to be carried out in Sii11gelierk1111dliche 111illeil1111gen 22, pp. 325-348. a short time. DUERST, G. u" 1930. Vergleichende Untersuchungsmethoden Pit 4 is probably a recent ditch, partly fi lled with old ma­ arn Skelett bei Siiugern. In: E. Abderhalden, Ha11db11ch terial. der biologische11 Arbeils111elhode11 1. Berlin, pp. 125 -5 30. Unpublished data. EHR!CH, R. w" 1977. Starcevo revisited. In: V. Markoric (ed.), A11cie11/ E11rope and lhe 111edilerra11ea11. \'\l arminster, pp. 5 9- 67. ELLENBERGER, w. & H. BAU�l, 194). Ha11db11rh der 11e1gleirhmdn1 7. LITERATURE A11a/0111ie der Ha11sliere, Berlin. FOC K, .J. 1966. 111etrische U11/ers11ch1111ge11 all il1e!apodim ei11ign

BARKER, G" 1975. Early Neolithjc land use in Yugoslavia. europiiische11 Ri11df1'!'(1sse11. Diss. J\'1iinchen. Proreedi11gs of the Prehistorir Soriel)' 41, pp. 8 5-104. GA FFREY, G" 195). Die Schadel der rnitteleuropiiischen

GLUTZ BLOTZHEI�!, 13.-\UER, K. �I. & u. N. VON 1966. Ha11db11ch Siiugetiere. Abha11dl1111ge11 1111d Berichle a11s de111 Staatlirhe11 der Viigel il1itteleuropas, 1- VJ. Frankfurt am J'v[ain. 1vl11Set1J11 f1ir Tierk1111de-Forsch1111gs-l11slil11t-Dresrle11 2 1, pp. BOESSNECK, .J., 1961. Haustierfunde priikeramisch-neolithi­ 5 -12). scher Zeit aus Thessalien. Zeilsrhrift fi ir Tierziicht1111g 1111d GEJVALL, N.-G" 1969. Lema, a preclassiral sile in the A1golirl. I: Ziicht1111gsbiologie 76, pp. 39-42. The fa 1111a. Princeton. BOESSNEC K, J" 1962.Die Tierreste aus der Argissa-JVIagula in HABER�!EHL, K. H" 1975. Die Alterbesli1111111111g bei rla11s- 1111rl Thessalien. In: V. MiloJcic, J. Boessneck & M. Hopf, Laborliere11. Berlin. Die de11tsche11 A11sgrab1111ge11 auf der A1gissa 111agula in HARCOURT, R. A" 1974. The dog in prehistoric and early his­ Thessalie11 I. Bonn, pp. 2 7-77. torie Britain. )011mal of archaeolog ical scie11ce 1, pp. 1 5 1-175. BOESSNECK, J" J· P . .J EQUIER & H. R. STA�lPFLT, 196). Seebe1g, HEINZEL, H" R. PITTER & .I·PA RSLOW, 1972. The birds of B11rg(ischisee-Siid. 3: Die Tierres/e (Acta Bernensia II). Kall­ Brilai11 a11d E11rope . London. H n1i.inz. JAR�IAN, "!. R. & .-N . _J AR�lAN, 1968. The faunaand econorny H.-H. T H The a11m1al of the British School JJOESSNEC K, J., �lULLER & �1. E!C ERT, 1964. Osteolog­ of Early Neolithic Knossos. ische Unterscheidungsmerkmale zwischen Schaf (Ovis of Archaeology in Alhens 6), pp. 24 1-276. aries Linne) und Ziege (Cap ra hirms Linne). Kiihn-Archiv .J OVANOVIC, B" 1971. La pre'histoire rl11 Dj erdab S11periwr. 78, pp. 5-129. JOVANov1c, B" 1974. Relativno hronoloski odnos starijeg

BOKONYl, s" 197). Die \Xlirbeltierfauna. In: D. Srejovic, Lep­ Neolita-Derdapa I Vojvodine. S1psko Arh1:olo"sko Dm.\'//lo e11ski Vir. Bergisch Gladbach, pp. 244-249. Gradski 1l111zej-S11bolira, pp. 3 1-49. "lAlTLAND, P. s" BOKONY!, s" 1974. His/ol)' of do111estic 111a111111als i11 Central and 1978. Elseviers gids /)(/I/ zoe/111a/ervissm. Eastem E11rope. Budapest. rnLLER, G. s., 1966 ( 1912). Ca!alogne of lhe 111a111111als of lll'estem

JJOKONYI, s" 1975. Vlasac: an early site of dog domestication. E11rope. ( Repri11!) . New York. In: A. T. Cia son (ed.), Archaeozoological studies. Amster­ NANDRIS, J" 1972. Bos prirnigenius and the bone spoon. B11lle­ dam, pp. 167-178. ti11 of the fllsti/11/e of Archaeolog y, pp. 63-82. w. BOKONYl, s" 1976. The vertebrate fauna. l/Yisseusrhaftliche Mil­ rrETSCH�lANN, 1977. Z11r Crosse des Rothirsrhes ( Gem11s el11p­ leil1111ge11 des Bos11isth-herzego1vi11ische11 La11des11111se11111s. IV A: h11s L.) in /lor- 1111d )i'iihgeschichllicher Zeil. Diss. Miinchen. Archiiologie, pp. 55-1 54. TE!CHERT, "!" 1969. Osteo111elrische U11/ers11ch1111ge11 z11r Bererh- BOLO�IEY, A" 1973. The present stage of knowledge of 1/)///g rie r IViderristhiJhe bei /!Ol'- 1111d ji'iihgesrhirhtlirhm mammal exploitation during the Epipalaeolithic and the Sch1ni11e11. Earliest Neolithic on the territory of Romania. In: ]. TEICHERT, "1" 1975. Osteornetrische Untersuchungen zur Be­ Matolcsi (ed.), Do111estikaliomfo rsch1111g 1111d Geschichle de r rechnung der \Xliderristhcihe bei Schafen. In: A. T. Ha11sliere . Budapest, pp. 197-20). Clason (ed.), Arrhaeozoologiral s/11dies. Amsterdam, pp. 46- BRINK, F. H. VAN DEN, 1968. Zoogdiere11gids. Amsterdam. 69. BURLEIGH, R" .J · CLUTTON-BROCK, P . .J . FELDER & G. DE G. TEROFAL, F" 1971. K wastsnoeken, steuren en verwanten. In: SIEVEKING, 1977. A further consideration of Neolithic Grzimek, Hel leve11 rier rliere11 IV, Vissen I, pp. 148-162. dogs with special reference to a skeletein from Grime's TR INGHA� !, R" 1971. H1111/ers, fishers and fa r111ers of Eas/em Graves (Norfolk), England . .fo11mal of arrhaeoloJ!/ral science Europe 6000-3000 BC. London.