Rengert Elburg Landesamt für Archäologie Sachsen, Dresden Man-animal relationships in the Early Neolithic of Dresden (Saxony, Germany) Elburg, R., 1999 - Man-animal relationships in the Early Neolithic of Dresden (Saxony, Germany) - in: Reumer, J.W.F. & De Vos, J. (eds.) - ELEPHANTS HAVE A SNORKEL! PAPERS IN HONOUR OF PAUL Y. SONDAAR - DEINSEA 7: 169-186 [ISSN 0923-9308]. Published 10 December 1999 A small sample of animal bones dating to the Early Neolithic Linear Ceramic Culture from DresdenMockritz (Saxony, Germany) has been investigated. It stands out for containing a wide range of species showing evidence of the use of very diverse environments. The remarkably high amount of wild animals as compared to domesticates is discussed and put in a wider context. It is shown that regional models explaining differences in animal exploitation patterns in the Early Neolithic are not tenable in the light of new evidence and a much more local mode of adaptation to different circumstances is suggested. Drawing evidence from the taphonomy of the complexes under study, an assessment of the representatives of the sample is made and its value to archeological interpretation indicated. An antler artefact, a so called T-axe, associated with the faunal remains, is discussed and thought to be an indicator of changing patterns of subsistence in the Neolithic of the Dresden area. Correspondence: Rengert Elburg, Landesamt für Archäologie Sachsen, Japanisches Palais, D-01097 Dresden, Germany; email: elburg@gmx.net Keywords: Subsistence, Early Neolithic, faunal remains, archeological assemblages INTRODUCTION Although the Early Neolithic Linear Pottery Culture is one of the most widely researched prehistoric cultures in central Europe, certain aspects and regions are badly underilluminated. The Linear Pottery Culture (or LBK for short from the German ‘Linearbandkeramik’) dates from c. 5500 to 4900 cal. yBC and is known from numerous sites stretching from the Ukraine in the east to the Paris Basin in the west, and the Great Hungarian Plain in the south up to the mouth of the river Oder. This culture represents the first farmers in Middle-Europe, colonising the lightly workable and fertile loess-soils as they spread through the region from the Southeast. This preference for mineral soils hampers the study of faunal remains from LBK-sites badly. Most excavations only yield very small amounts of bone, as bone-preservation in these by now mostly deeply decalcified soils is extremely poor. During the last decade or so, this gap in our knowledge is being gradually filled, but only from a few dozen sites fully published faunal data are available (Fig. 1). Recent overviews are given by Benecke (1994) and Döhle (1994). Central in the region occupied by the LBK lies the so called ‘Dresdener Elbtalweitung’, a climatic favourable region on the upper course of the river Elbe between the cities of Pirna and Meissen in the province of Saxony, Germany. From this area, measuring c. 400 square kilometres, numerous Early Neolithic sites are known, but up to now only a few have been published, mostly as preliminary reports (e.g. Elburg 1994, Kurz 1994, Brestrich & Elburg 1996, Pratsch in press). In 169 ELEPHANTS HAVE A SNORKEL! this study a small complex of animal bones, found during archeological excavations in the south of the city of Dresden in 1996, will be presented and its significance discussed. MATERIAL AND METHODS During archeological excavations in advance of extensive construction activities a section of c. 2 ha from a much larger LBK settlement was uncovered. The site lies among gently undulating loess hills near the confluence of two small brooks, the ‘Kaitzbach’ and ‘Nöthnitzbach’, in the south of the metropolitan area of Dresden. It is known by its site- DEINSEA 7, 1999 code DD-27 from the Saxony Archeological Service and is at the moment virtually unpublished (Brestrich & Elburg 1996). A f t e r excavation over 30 plans of, for this period typical, longhouses could be reconstructed (Fig. 2), with traces of about a dozen more. One of these buildings (‘House 1’, at the north-eastern end of the excavation) belongs with its length of 50 metres to the longest known. The site spans the whole of the LBK period, with exception of the earliest phase, which is known from a nearby excavation. Some further features could be dated to later prehistoric cultures. Figure 1 Sites with faunal assemblages refe rred to in the text and used as references. Circles: Linear Pottery Culture (LBK), squares: other Neolithic cultures.White fill: low or very low proportion of wild animals (<10 %), light grey: 10-20 % hunted animals, dark fill: over 20 % of bones from wild animals. Numbers: 1-Dresden Mockri t z , 2-Dresden-Cotta, 3-Eilsleben, 4-Barleben, 5-Magdeburg-Prester, 6-Gatersleben, 7-Grossörner, 8-Hohlstedt, 9-Halle-Trotha, 10-Tröbsdorf, 11-Erfurt-Nordhäuserstrasse, 12-Rosdorf, 13-Müdderheim, 14- B ruchenbrücken, 15-Frankfurt-Niedereschbach, 16-Schernau, 17- K raichtal-Gochsheim, 18-Hüde, 1 9-Vaihingen-Enz, 20-Ammerbuch (also Middle Neolithic), 21-Colmar, 2 2-Wettolsheim, 23-Rouffach, 24-Ensisheim, 25-Hilzingen, 26-Singen, 2 7-Twann, 28-Seeberg-Burgäschisee-Süd, 29-Egolzwil, 30-Hienheim, 31-Regensburg-Pürkelgut, 32-Straubing-Lerchenhaid, 33-Künzing-Unternberg, 34-Strögen, 35-Pulkau, 36-Mikulov, 37-Neckenmerkt, 38-Békásmegyer-Vörös Csillag, 39-Györ-Pápai Vám, 40-Neszmély, 41-Bina-Nové Zámky, 42-Stúrovo, 43-Pomáz-Zdravylák, 44-Chotébudice, 45-Roztoky, 46-Klicany, 47-Bylany, 4 8-Nové Dvory, 49-Gniechowice, 50-Kujavia (seve ral sites), 51-Zalecino, 52-Zuków, 53-Flemsdorf, 54-Zollchow. 170 ELBURG: Neolithic animal bones recovered but even here most pieces are badly weathered. Among the features containing faunal remains a cluster of pits near one of the houses (‘House 8’, Figs. 2 and 3) stood out for providing a quantity of better-preserved material, some of it clearly worked. To gain an impression of the composition of the site’s faunal assemblage this area was selected for further analysis. Figure 2 Generalised plan of the excavation DD-27 in Dresden-Mockritz. Dashed line marks area enlarged in Figure 3. Bone preservation, on the site as a whole, was poor to very poor. Only in the northeastern part, towards the limits of the excavation, and with it, of the settlement, as well as in a few isolated pits, some bone could be The features under study consist of an isolated pit (feature 1410) and three pit complexes (complex 1430, 1432 and 1434) (Fig. 3). Two of these complexes, 1430 and 1432, flank House 8 in the classical way for daub pits. From these kind of features loam was dug up to construct the wattle-and-daub walls of the longhouses. By their position as well as from the taphonomy of the bones (see below), and the occurrence of sherds of the same pot in the fill they can be designated as ‘decapitated’ parts of the same pit system. The top part of this complex was machined away, as it was invisible in the covering chernozem-type of humic soil. Although no reliable chronology for pottery exists in the area, the decoration of the ceramics (Fig. 4, findnumbers 826 and 835) clearly belongs to the last phase of the LBK, being of the so-called sarka-type (Vencl 1961). The isolated pit of unknown function (feature 1410) contained pottery of a similar type. For pit 1410 contemporaneity with complex 1430, and so with 1432, can be proven by the zoological contents (see below). In contrast, the pottery from complex 1434 is decorated in a clearly different style (Fig. 4, find-number 877), normally ascribed to the middle phases of the LBK-period. The material can thus be separated in two groups, an older and a younger one. The amount of time separating the two clusters is unknown as no radiometric dates are available yet, but at least 14C will date the later of the groups in the near future. Apart from ceramics and bones, finds consist of fragments of stone axes, burnt loam, grindstones and querns as well as worked quartzite and flints. Of special interest among these are the fine-grained 171 ELEPHANTS HAVE A SNORKEL! Figure 3 Reconstructed houses 3 and 8 with positions of the features studied. quartzite, as these originate from two different tertiary deposits in northern Bohemia and demonstrate close contacts between the two regions during the early Neolithic (Elburg in prep.). DEINSEA 7, 1999 Available for analysis were over 700 fragments of bone weighing a little more than 7 kg. Apart from a few isolated teeth and some carpals, tarsals and phalanges, no bone was complete. All bones were laid out to see if fits between recently or prehistorically broken fragments could be found. After this, the material was checked for anatomical fits, which in one case could be established (see below). Subsequently, identification was carried out using the very limited reference collection of the Archeological Service in Dresden as well as some bones from other excavations and recent specimens from a private collection. Every identifiable fragment was weighed and data on age, fragmentation, burning etc. were scored and entered in a database. Unidentifiable pieces were counted and weighed together. No attempt was made to identify fragments of ribs and vertebrae, as this would have taken a disproportionately high amount of time in relation to the gain of information. Where possible, measurements were taken following Von den Driesch (1976). The distribution of the material over the different complexes can be seen in Table 1. No listing by skeletal elements is given, as the assemblage is too small for any meaningful conclusions to be drawn on diff e r e n t i a l representation of different parts of the body. Figure 4 Decorated pottery from Complex 1432 (nu m b e rs 826 and 835) and complex 1434 (number 877). [Illustration: D. M a c kowsky] 172 ELBURG: Neolithic animal bones Ta ble 1 Weight (w) and number (n) of identified bone listed by animal species and complex. All relevant measurements are given in the Appendix. DEPOSITION AND TAPHONOMY Between the four complexes clear diff e r e n c e s and similarities in taphonomy are discernible, as becomes clear in Figure 5. The two pits flanking the house (complexes 1430 and 1432) are virtually indistinguishable in the proportion of identified bone, in weight as well as in numbers. This confirms the interpretation that both were originally parts of the same larger unit. The pattern in pit 1410 is very similar with a high percentage of unidentified specimens but a markedly higher proportion of identifiable material by weight. On average, unidentified specimens in the complexes 1430 and 1432 are with 1.7 g nearly twice as heavy as those in pit 1410 were the average lies at only 0.9 g. A logical explanation for this distinction is not obvious, but indicates at least some differences in the taphonomical processes at work. The contrast between these features in the immediate vicinity of a house, and complex 1434, which cannot be assigned to a specific building (Figs. 2 and 3), is striking. Whereas the average weight of a specimen in the complexes belonging to house 8 is only about 4 g, in complex 1434 it is an astonishing 32 grams. Differential excavation techniques are not of Figure 5 Percentages of bone identified by weight (narr ow bar) and number (wide bar). 173 ELEPHANTS HAVE A SNORKEL! significant importance as all features were excavated by the same team under the same circumstances approximately at the same time. The most likely explanation of this dichotomy is that in the complexes 1410, 1430 and 1432 refuse resulting from activities in and around the house was disposed of, while the contents of complex 1434 represent deliberately dumped material. This interpretation is supported by the fact that the bones in complex 1434 are in general less weathered and show less gnawing marks than those from the other pits. Some more information about the processes at work in the formation of the assemblage can be gathered from anatomical fits. In one case, six bones belonging to the same individual, coming from two different complexes, 1410 and 1430, could be refitted (Fig. 6). The tibia bears heavy chewing marks from dogs while the metapode was used as a kind of ad-hoc implement: a larg e splinter of the diaphyse was rammed into the proximal end and used for an unknown purpose. After the splinter had broken off, the ‘tool’ was discarded, and must have been buried rapidly as it is in very good condition. On the other bones of the leg no traces of use or damage by animals could be recorded. This finding has two important implications. DEINSEA 7 1999 Firstly, complex 1430 and pit 1410 must have been open at the same time, a proof of contemporaneity seldom encountered with this clarity in the archeological record. Secondly, the filling of the pits has been relatively rapid and the finds contained represent few discrete events. Other material confirms the latter: the concentrated occurrence of Alces phalanges in complex 1432 and the presence in pit 1410 of both the left and right scapula of the same wild boar. This conclusion underlines the sample-like character of faunal assemblages from archeological excavations. RESULTS Bos Of the two genera of large bovids known from the Neolithic of central Europe, Bison and B o s, only Bos seems to be represented. In most archeozoological literature separation between the wild and domesticated varieties is made on the specific level: Bos primigeni us BOJANUS and Bos taurus L. The validity of d i fferentiation on this level in the early history of domestication remains unclear, but the terminology will be maintained in this text for comparability. With 34.3% of the number of identified specimens (NISP), by weight Figure 6 Articulating elements from right hind leg of Cervus elaphus. Note splinter used as an implement inserted in metatarsal, causing the bone to crack. 174 ELBURG: Neolithic animal bones even 66.3%, Bos is by far the most abundant genus in the material. This is usually the case in LBK assemblages, where on some sites percentages of over 80% of (domestic) cattle are reached. Although there is ample literature on the distinction of B. taurus and B. pri migenius, separation of the two species remains ambiguous in many cases. Roughly, three size-classes can be distinguished within the genus in the Early Neolithic (e.g., Bökönyi 1962, Döhle 1994). The group with the smallest individuals represents mostly female domesticated cattle whereas aurochs bulls cluster in the upper range. For the wide range of intermediate-sized animals, the species can often hardly be established. In this light it is remarkable that in most publications nearly all bones are assigned to either B. primigenius or B. tauru s (one exception: Stork 1993). In general, the motto seems: if it is not certainly aurochs (i.e. extremely large), it must be domesticated. This probably leads to a serious underestimation of the amount of wild cattle in earlier Neolithic contexts, especially in those regions, like central Germany, where the variation among B. primigenius in the Neolithic is insufficiently known. To do some justice to this complex matter, the material classified as cattle was split into four groups (Table 1 and Appendix): certainly domesticated animals (Bos tauru s), certain aurochs (B. primigenius), ‘Bos primigenius ?’ for robust bones not clearly assignable to one of the two species and a restgoup. The latter contains most juveniles and all specimens that are not notably robust or gracile and where no measurements could be taken. For better comparison with other publications, this group can, for the reasons mentioned above, be lumped with the domesticated animals. Bos taurus This group consists of 20 pieces (11.8% of the total number, by weight 22.5%), mainly coming from complex 1434. The metric values (Appendix) match those already published for the region well (Benecke in press, Döhle 1994, Müller 1964a), although it must be said that some of the very large measurements for domestic cattle in Müller (1964a) should probably be reinterpreted as B. primigenius (see below). Because of the very small sample, nothing can be said about the age structure, especially as most juveniles would fall into the ‘B o s sp.‘ category. Bos primigenius Only one piece could definitely be identified as aurochs: a proximal part of a left radius. Although even larger specimens are classified by Müller (1964a) as belonging to domestic cattle, comparison with other data (Bökönyi 1962, Döhle 1994) makes clear that it lies in the range of large cows/small bulls of B. pri m i g e n i u s. Bos primigenius ? For seven specimens from complex 1434 classification as either B. primigenius or B. t a u ru s seems, in spite of their good condition, nearly impossible. Although contributing only 4.1% to the NISP, they account for 21.7% of the total weight; this already indicates their massiveness. Of these pieces, five could be measured (Appendix), the other two specimens, a heavily damaged proximal part of a tibia and a proximal fragment of a radius can only be classified as very robust. It is remarkable that all parts of the limbs come from the left-hand side, so that at least two individuals are represented. The proximal radius fragment is very similar to the specimen classified as B. primigenius and could so be tentatively determined as aurochs, giving evidence of a second individual. Compared to other material from the early Neolithic in Dresden, the tibia appears extremely massive. An estimate of the original width would be 120 mm, quite precisely the delimitation most archeozoologists maintain for distinction between domesticated cattle and aurochs. As said before, the minimum values set for aurochs probably cause a large number of wild cattle to go unrecognised. 175 ELEPHANTS HAVE A SNORKEL! The measurable specimens do not fare any better, as can be illustrated by the example of the astragalus. Astragali, because of their high compactness, are among the most frequently found bones in archaeological excavations. Therefore, a large body of data stands at disposal for comparison. The most widely used measurement for classification is the greatest lateral length, which is 80 mm for the piece under study. The diagram in Benecke (1994, fig. 57), where this measurement is depicted for finds from the whole middle European prehistory, is highly instructive. The upper limit for B. tauru s, with only a slight overlap with B. primigenius, is 80 mm. A determination of this piece as aurochs is thus highly probable. For the two distal parts of radii with a Bd/BFd of 96/87 and (90)/(83) similar doubts about their classification exist. Other authors are confronted with the same question, as Döhle (1994: 46) puts it: ‘Again, problems arise for the classification of specimens in the size class with a distal width between 90 and 97 mm’ (translation author). Based on his comparisons (Döhle 1994: 46 and fig. 15), he is inclined to classify specimens with a Bd of 90 mm and greater as B. primigenius. A decision to class the entire group of material discussed here as aurochs, considering the high proportion of wild animals and the presence of certain pri migenius in the assemblage, seems justified. Capra/Ovis As no wild species of sheep and goat were present in the indigenous fauna of the region and no other small bovids are known from the Neolithic, this group is clearly the least problematic to distinguish. Problems arise, however, when differentiating between Capra hircus and Ovis ammon. As the two species occupy quite different ecological niches, knowledge about the proportions in the material can give some insight into the landscapes used by prehistoric herdsmen. Due to poor conservation, only in two cases a tentative attribution to one of the species can be made, both seem to belong to C a p r a. The presence 176 DEINSEA 7, 1999 of Ovis in the material is nevertheless highly probable since sheep generally outnumber goats 2:1 in the early Neolithic of central Germany (Müller 1964a). With 22 elements Capra/Ovis constitute the third largest group among the finds, contributing 12.9% to the total number of identified specimens, but only 2.5% by weight. In the case of two fragments of smaller horncores, no decision could be made between (small) Bos and C a p r a / O v i s, these pieces are listed in Table 1 as ‘Bovid’. The few measurements that could be taken (appendix) seem to fit well into the known range (Müller 1964a, Döhle 1994). Sus scro fa The same problems that arise for the distinction between wild and domestic varieties of Bos hold true for Sus scro f a, although less so, as pig reacts faster on domestication with marked size reduction (Benecke 1994). Classification of adult specimens is therefore quite straightforward; the main difficulty lies with juvenile animals, which are normally quite abundant in archaeological material, Dresden-Mockritz being no exception. The same procedure as for Bos is therefore followed with classification in three groups: wild, domesticated and indeterminate. An additional group ‘probably wild’ can be omitted as no specimens are on the robust side of the intermediate range (Appendix). With 14.2% of the NISP, 8.1% by weight, the share of pig lies within the normal range for central Germany and Bohemia. S. scrofa f. domestica Astonishingly, only two specimens of Sus could be definitely classified as domesticated, although most pieces from the indeterminate group probably originate from domesticated animals. Both the humerus and the tibia fit well in the known variation of LBK pigs in the region (Müller 1964a, Döhle 1994). S. scrofa f. fera Six elements can beyond any doubt be clas- ELBURG: Neolithic animal bones sified as wild boar. Additional to the elements listed in the appendix, a juvenile calcaneus was recorded with a greatest width of 27 mm and a length without the epiphysis of 92 mm. It clearly belonged to a very larg e , wild animal. The measurements (Appendix) can hardly be compared with data from the region, as only very limited numbers of wild boar are known. From the Neolithic settlement in Twann, Switzerland, larger series have been published (Becker & Johansson 1981), which can be used as reference, although caution should be taken in comparing measurements between two different regions. The two scapulae, a left and a right from the same animal, fit in very well, as does the astragalus. The m3 seems to be very long compared to the values from Twann, where the average length is 43.6 ± 2.0 mm (n=42) with the largest piece measuring 49.4 mm and an average width 18.7 ± 0.8 mm (n=40). The proportions of the ulna don’t match this series either: although the BPC lies within one standard deviation, the DPA falls below the variation. This could well be an indication of morphological differences between the two populations, as the dimensions of the piece fall clearly outside the range of central German domestic pigs. S. scrofa, indeterminate A total of 16 specimens was assigned to this category, most of which are either very fragmented or juvenile, so no decision could be made for either of the two varieties. The high proportion of young animals in this group already indicates they were probably domestic. Eight skull fragments, representing at least two juvenile animals coming from complex 1434, constitute half of the material. Cervus elaphus For most Neolithic assemblages a high proportion of wild animals is nearly always synonymous with a high proportion of deer. So it is no surprise that in Mockritz, with nearly half the bones coming from hunted species, C e rvus elaphus is with 31 identified elements the second most abundant animal. Among this material are two pieces of antler, the basal part of a shed beam and a piece worked into an axe (see below), that are left out of the total count of hunted animals. This sets the relative abundance of C e rv u s at 17.2% by number and 17% by weight. Even in a very limited sample like this one, an overrepresentation of elements from the hind legs, especially the tibia, seems evident. These bones make up more than half of the material, whereas the front legs are represented by one part of a humerus only. Partially, this is caused by the presence of six elements from the same limb (see above), but there are another eight fragments of tibiae, two of which come from the older complex 1434. The measurements that could be taken fall within the normal range of the region (Appendix). Alces alces The elk is rarely found in LBK context. In the larger region of central Germany and Bohemia only two specimens have been published: a worked tine in a collection of bone and antler tools from the LBK settlement of Nerkewitz (near Jena, Thuringia) without further documentation (Müller 1964a, 1964b) and an unknown element from Chotebudice (Peske 1989). In southern Germany, elk is known from StraubingLerchenhaid, where one second phalanx was found (Ziegler 1989). It is therefore remarkable that Alces is represented in the faunal spectrum of at least two excavations in the Dresden area. One juvenile metatarsal (Fig. 7) comes from an excavation in the west of the city, Dresden-Cotta, and was identified as ‘probably elk’ ( E l b u rg 1994), but this find is for unknown reasons not included in the definitive study of the faunal material from this site (Benecke in press). To fill this gap, the piece will be briefly described here. The specimen with find-number DD-04/40 comes from a ‘settlement pit’, feature 55, of unknown function (Pratsch in press), contai- 177 ELEPHANTS HAVE A SNORKEL! DEINSEA 7, 1999 Figure 7 Left metatarsal of Alces alces used as a polishing bone from Dresden-Cotta. ning a mixed inventory of LBK and somewhat later Stroked Ware pottery. The distal epiphyses are missing and this end of the shaft still has a very porous surface. On the plantar side, towards the proximal end, some cut marks are visible which are caused by butchering. The lateral side shows polishing and scratching at mid-shaft and is slightly hollowed out due to use as a polishing bone. The measurements are: greatest length (without the epiphyses): 342 mm, proximal width: 48 mm, proximal depth 49.5, smallest depth of the diaphysis: 31 mm. The smallest width of the diaphysis could not be taken because of the wear in this area. These dimensions point to a normal sized animal. The cut marks indicate that the bone comes from a hunted animal and was not collected as a surface find. In Dresden-Mockritz, Alces is represented by at least six phalanges (Fig. 8), three firsts and three seconds, all coming from complexes 1430/1432. They account for 3.6% of the NISP, 2.2% of the total weight. At first sight, the presence of only phalanges seems remark- Figure 8 First and second phalanges of Alces alces from Dresden-Mockri t z . 178 ELBURG: Neolithic animal bones able. Although in this case probably caused by the small number of bones in the sample, actual overrepresentation of elk phalanges is a well known phenomenon in Neolithic material (Stampfli 1992). The presence of Alces in the assemblage causes some doubt in a few cases as to the attribution of bone fragments, mostly phalanges, of larger cervids. These cases are listed in Table 1 as ‘Cervid’. Only one measurable element, a first phalanx (Appendix) comes into this group because of lacking reference material for elk. In comparison with data from Switzerland (Becker & Johansson 1981), central Germany (Benecke in press, Döhle 1994) and south-west Germany (Brink 1990, Ziegler 1989, Stork 1993), this phalanx outranges published measurements of C e rv u s. For a definite conclusion that it is Alces, far too little Neolithic material from the region is published for the range of variation of this species to be clear. C a p reolus capreolus The smallest cervid accounts for 5.9% of the NISP, but makes with 0.8% hardly any contribution to the total weight. Only two out of the ten bones present could be measured (Appendix). Both come from medium sized animals. Roe deer is present in all LBK assemblages with a somewhat higher (>10%) share of wild animals and in most other too. In most cases Capreolus is, like in Mockritz, clearly less abundant than C e rv u s and of minor importance for the supply of meat in Neolithic settlements. Equus sp. In many of the larger faunal assemblages from the LBK a low percentage of bones from equids is present. The taxonomy of Early Neolithic horses is very unclear. Most specimens are described as Equus feru s or Equus sp., a few are assigned to ‘Equus (A s i n u s) hydruntinus’ (Peske 1989). The discussion if horses are already domesticated in this period still continues (for a recent overview: Benecke 1994), although this seems highly unlikely. In Mockritz an equid of un- known species is represented by one left astragalus, wearing heavy gnawing marks from dogs. Because of this damage the only two measurements that could be taken are the greatest height: (54.5) mm and the length of the medial part of the Trochlea tali: (54.5). These dimensions point to a small animal, even for Neolithic standards. Erinaceus sp. From pit 1410 comes a mandible fragment, without dentition, of a hedgehog. The seemingly only other find of this animal in the LBK comes from Roztoky (Peske 1989) where a bone of Erinaceus is listed. In the lakeside dwellings of Switzerland, where these animals are found much more frequent, they are represented by a disproportionally high amount of mandibulae, most of which wear traces of polishing (Becker & Johansson 1981, Stampfli 1992). The small areas preserved on the specimen from Mockritz show a slight lustre too. It seems these bones were often used as a kind of decorative pendant or amulet in the Neolithic (Stampfli 1992). Lepus euro p a e u s Among the faunal remains there is the crown of one lower incisor of the European hare. The combination of employed excavation techniques and poor bone preservation cause the small numbers of Lepus found in Neolithic assemblages. As with horse, the presence of hare is a good indicator for open landscapes. Castor fiber Beaver is frequently found among animal remains dating to the Early Neolithic. Its occurrence in the assemblage does not surprise, as even today a few of these animals survive in the region. Although the one distal part of a right humerus with a weight of 8.5 g does only contribute 0.15% to the total weight (0.59% of the NISP), it is another illustration of the use of a very wide range of environments. The three recordable measurements, a distal width of (27) mm, width of 179 ELEPHANTS HAVE A SNORKEL! the Tro c h l e a of (22) mm and a smallest width of the diaphysis of 11.5 mm, seem to indicate a relatively small animal. Ave s In all, five bone fragments of larger birds were found. Identification, even to the generic level, proved impossible as all epiphyses are missing and the reference collection used contained very few bird bones. The only statement that can be made is that the birds have been relatively large, about the size of geese. Bird bones are rare in LBK contexts, but this is probably again a bias caused by differential conservation and employed excavation techniques. Wo r ked bone and antler Some of the bones, like the proximal part of a metatarsus of a deer already described, show clear traces of use. Few were shaped into real tools, like the spatula or point in Figure 9. Remains of at least two more pointed implements were recognised. They all correspond with known types from the period. A very uncommon piece of equipment for the LBK, however, is a so-called T-axe, manufactured from an antler of red deer (Fig.10). To make this axe, the beam was cut at both sides of the trez tine by scraping away the cortex with a flint implement after which the spongiosa was broken. In the same manner, the trez tine itself was removed. The shafting hole was made by drilling into the spongiosa in the remainder of the tine and picking and scraping away the cortex at the opposite side. Figure 9 Bone implement from pit 1410. [Illustration: D. Mackowsky] 180 DEINSEA 7, 1999 Polishing one side of the beam then created a working edge. The cutting edge is splintered due to heavy blows delivered during use. Literally hundreds of these axes have been found in Europe (e.g. Clason 1983, Van der Waals 1972, Werning 1983), but they remain a very problematic class of artefacts. There seem roughly two cultural zones where this type of implement occurs. In the northern lowlands they are characteristic for the so called Ertebølle-Ellerbek culture (TroelsSmith 1967), which represents a mixture of Neolithic and Mesolithic life-styles. In central Europe on the other hand, they belong to the Middle Neolithic, specifically to the Lengyel horizon (Berlekamp 1969). In earlier contexts they are extremely rare in central Europe and up to now only one, found in Eilsleben, could be dated to the LBK proper (Kaufmann 1986). In general, they are ascribed to cultural groups that show a high proportion of hunted animals in the faunal assemblages (Pratsch 1994). For two finds from earlier contexts the share of wild animals in faunal assemblages associated with them can be established. Uerpmann (1977) published the animal remains from a pit dating to the earlier Middle Neolithic (Stroked Ware/Rössen) near Ammerbuch containing two T-axes (Albert & Schröter 1973). For the only specimen that dates to the LBK (Kaufmann 1986) from Eilsleben, the data can be found in the publication of the faunal remains from this site (Döhle 1994). For both features the amount of identifiable bone is quite low but nevertheless worthwhile to look at: in Ammerbruch only two of 25 bones classifiable to the specific level come from domesticates. In Eilsleben, of 35 identified specimens, six bones come from wild animals and one third molar probably belongs to B. primi genius. In Ammerbuch the very low amount of domesticated animals doesn’t surprise, as high proportions of wild animals are common in this region (Stork 1993), but it confirms the connection between this type of implement and high percentages of hunted species. In Eilsleben this correlation seems to be absent: ELBURG: Neolithic animal bones Figure 10 T-axe made from antler of Cervus elaphus. [Illustration: D. Mackowsky] the percentage of wild animals from the same pit as the T-axe (feature 25/84) is only 17-20%, but as the site as a whole has only 9.8% of hunted animals, this percentage can be called somewhat elevated in spite of the small sample. DISCUSSION The most striking features of the assemblage are the low percentage of domesticated animals and the wide range of species represented. At least 13 species are present, including some that are rarely encountered in LBK assemblages (like elk and hedgehog) or only seldom found (horse, hare and birds). This is especially surprising, as the contents of the pits do not represent a slow accumulation of bones, but probably a short series of discrete events, as is shown by depositional and taphonomical evidence. The composition gives evidence of the use of very diverse environments: woodland (deer and pig), open woodland (cattle, roe deer and sheep/goat), plains (horse and hare) as well as wetlands with beaver, elk and probably birds. This type of exploitation correspondents well with the behavioural pattern of ‘tethered mobility’ proposed by Whittle (1996) for the Early Neolithic. The large contribution of wild animals to the meat supply confirms the findings from Dresden-Cotta (Benecke in press), and overturns the established picture for central Germany and Bohemia. Until now, a contri- 181 ELEPHANTS HAVE A SNORKEL! bution to the faunal assemblages by domesticates of more than 90% was thought to be typical (Döhle 1993), ignoring the Bohemian site of Nové Dvory (Peske 1986) as this material comes from one single pit. On this basis two different patterns of animal exploitation (Döhle 1993) were formulated, referring to the work of Sielman (1972). In this model Bohemia, central Germany and larg e areas of the Rhine basin belong to one ecological area, ‘ecosphere A’, characterised by moderate to low precipitation, moderate temperatures and a high aridity index (Sielman 1972). Other regions, having a climate judged less favourable to Neolithic agriculture are lumped together as ‘ecosphere B’. According to Döhle (1993), ‘ecosphere A’ would correspond with low percentages of hunted animals combined with high values of pig coupled with low amounts of sheep/goat among the domesticates in the faunal remains from the LBK and succeeding cultures. If in Dresden-Mockritz all bones from Sus of indeterminate status and Bos sp. (Table 1) are counted as domesticates, the ratio cattle:sheep/goat:pig comes to 56:24:20, closely resembling the ratio 52:23:25 for the LBK from Dresden-Cotta (Benecke in press). In regions with high percentages of wild animals, like southern Germany (‘ecosphere B’), this ratio varies from 63:11:26 (at the Danube) to 43:17:40 (the region of Lake Constance). For central Germany Benecke (1994) gives 52:34:14 and for Bohemia 72:17:11, both regions with low amounts of wild animals. Strikingly, best agreement with the Dresden region can be found in the Alsace where the proportions are given as 54:23:21 (Döhle 1994), but this again is an area with only c. 8% of non-domesticates. The thesis that there are large supra-regional patterns to be found seems untenable in the light of the evidence from Saxony and other regions like around the river Main (Fig. 1, numbers 14 and 15) (Uerpmann 1997), that belong to Sielmanns ‘ecosphere A’. Probably a much more localised pattern of landscape and animal exploitation is at work, like sug- 182 DEINSEA 7, 1999 gested by Pucher (1987). It is remarkable that both Dresden-Cotta and Dresden-Mockritz are relatively low-lying sites with rising hinterland near a formerly marshy terrain (Ebert et al. 1934). These would be good sites for a settlement if a high diversity of biotopes in the immediate surroundings was sought for. The choice of such sites is more reminiscent of foragers than of strict agriculturists and cattle breeders, as always envisaged for the carriers of the LBK. Comparing the two groups of material from Dresden-Mockritz to each other seems premature on the basis of the small samples, but there seems to be a marked rise in the percentage of wild animals. In the older complex 1434, wild animals account for at least 34% when the very robust B os bones are counted as B. primigenius. In the younger group consisting of 1410, 1430 and 1432, this share rises to a remarkable 55%. In this light the presence of the T-axe could indicate a transition to the Middle Neolithic with its broader subsistence basis and wider range of exploited environments. ACKNOWLEDGEMENTS I would like to express my thanks to Dieter Mackowsky for splendidly drawing the finds and to Andreas Kinne for making his private collection of bones available for reference. I am especially grateful to Paul van der Kroft and Edy Mulié from I/O-Graph in A m s t e r d a m for extensive help with the digital processing of the figures and long discussions on all aspects of archeology. REFERENCES Albert, S. & Schröter, P., 1973 - Zu einer neolithischen Grube mit Tüllengeweihäxten bei AmmerbruchReusten (Kr. Tübingen, Baden-Württemberg) Archäologisches Korrespondenzblatt 3: 289-292 Becker, C. & Johansson, F. 1981 - Tierknochenfunde. Zweiter Bericht - Die neolithischen Ufersiedlungen van Twann 11, Bern Benecke, N., 1994 - Archäozoologische Studien zur Entwicklung der Haustierhaltung - Akademie Verlag, ELBURG: Neolithic animal bones Berlin. Benecke, N., in press - Die Tierreste aus bandkerami- Kurz, S., 1994 - Archäologische Untersuchungen im Gewerbegebied Dresden-Nickern 1. Eine Bestands- schen Siedlungen von Dresden-Cotta - in: Pratsch, A. übersicht - Archäologie aktuell im Freistaat Sachsen - Die linien- und stichbandkeramische Siedlung 2: 23-30 in Dresden-Cotta. Eine frühneolithische Siedlung im Dresdener Elbkessel - Beier & Beran, WilkauHaßlau Berlekamp, H.D., 1969 - Die Einflüsse des Müller, H.-H., 1964a - Die Haustiere der mitteldeutschen Bandkeramiker - Schriften der Sektion für Ur und Frühgeschichte Bd. 17, Berlin Müller, H.-H., 1964b - Bandkeramische Knochengeräte donauländischen Kreises der jüngeren Steinzeit auf zoologisch betrachtet - Varia Archaeologica, Schriften das Odermündungsgebiet - Ethnografisch- der Sektion für Ur und Frühgeschichte Bd. 16, Berlin Archäologische Zeitschrift 10: 213-216 Bökönyi, S., 1962 - Zur Naturgeschichte des Ures in Ungarn und das Problem der Domestikation des Hausrindes - Acta Archaeologica Academiae Scientiarum Hungaricae 14: 175-214 Brestrich, W. & Elburg, R., 1996 - Zwischen den Bächen. Die bandkeramische Siedlung von Dresden- Peske, L., 1986 - The results of osteological analyses in: Pavlu, I. et al. - Theses on the Neolithic site of Bylany - Památky Archeologické 77: 404-406 Peske, L., 1989 - Animal bones from Bylany - Bylany seminar 1987 collected papers, Prague: 265-271 Pratsch, A., in press - Die linien- und stichbandkeramische Siedlung in Dresden-Cotta. Eine frühneolithische Mockritz (DD-27) - Archäologie aktuell im Freistaat Siedlung im Dresdener Elbkessel - Beier & Beran, Sachsen 4: 9-13 Wi l k a u - H a ß l a u Brink, J.S., 1990 - Middle Neolithic animal remains from Pratsch, S., 1994 - Die geweihartefakte des mesolithisch- „Bajuwarenkaserne“, district Regensburg - in: neolithischen Fundplatzes von Friesack 4, Kr. Schibler, J. et al. - Festschrift für Hans R. Stampfli, Havelland - Veröffentlichungen des Branden- Basel: 31-37 Clason, A.T., 1983 - Spoolde. Worked and unworked antlers and bone tools from Spoolde, De Gaste, the b u rgischen Landesmuseums für Ur- und Frühgeschichte 28, Potsdam Pucher, E., 1987 - Viehwirtschaft und Jagd zur Zeit der IJsselmeerpolders and adjacent areas - Paleohistoria ältesten Linearbandkeramik von Neckenmarkt 25: 77- 130 ( B u rgenland) und Strögen (Niederösterreich) - Döhle, H.-J., 1993 - Haustierhaltung und Jagd in der Wissenschaftliche Mitteilungen der Linienbandkeramik-ein Überblick - Zeitschrift für Anthropologischen Gesellschaft in Wien 117: 141- Archäologie 27: 105-124 155 Döhle, H.-J., 1994 - Die linienband- keramische Tierknochen von Eilsleben, Bördekreis - Sielman, B., 1972 - Die frühneolithische Besiedlung Mitteleuropas - Fundamenta A/3 - Die Anfänge des Ve r ö ffentlichungen des Landesamtes für archäologi Neolithikums vom Orient bis Nordeuropa. Va sche 47, Halle (Saale) Westliches Mitteleuropa, Köln: 1-65 Ebert, H., Grahmann, R., Pietzsch, K., 1934 Erläuterungen zur geologischen Karte von Sachsen Stampfli, H.R., 1992 - Die Tierknochen aus den jungsteinzeitlichen Siedlungen Nr. 66 Blatt Dresden - Landesvermessungsamt Egolzwil 3 und Egolzwil 4 - Archäologische Sachsen, Dresden Schriften Luzern 1.1992 Elburg, R., 1994 - Dorf auf dem Lößlehm. Eine Siedlung der ersten Bauern in Dresden-Cotta - Archäologie aktuell im Freistaat Sachsen 2: 39-46 Elburg, R., in prep. - Rohmaterialbenutzung und v e r s o rgung in der bandkeramischen Silexindustrie der Dresdener Elbtalweitung. Kaufmann, D., 1986 - Ausgrabungen im linienbandkeramischen Erdwerk von Eilsleben, Kr. Wanzleben, in den Jahren 1980 bis 1984. 5. Vorbericht - Zeitschrift für Archäologie 20: 237-251 Stork, M., 1993 - Tierknochenfunde aus neolitischen Gruben in der Gemeinde Ammerbuch, Kr. Tübingen Zeitschrift für Archäologie 27: 91-104 Troels-Smith, J., 1967 - The Ertebølleculture and its Background - Paleohistoria 12: 508-528 Uerpmann, H.-P., 1977 - Betrachtungen zur Wirtschaftsform neolithischer Gruppen in Südwestdeutschland - Fundberichte aus BadenW ü r t t e m b e rg 3: 144-161 Uerpmann, H.-P., 1997 - Die Tierknochenfunde - in: 183 ELEPHANTS HAVE A SNORKEL! Lüning, J. (ed.) - Ein Siedlungsplatz der Ältesten Bandkeramik in Bruchenbrücken, Stadt F r i e d b e rg/Hessen - Universitäts-forschungen zur prähistorischen Archäologie 39, Bonn: 333-348 Van der Waals, J.D., 1972 - Die durchlochten Rössener Keile und das frühe Neolithikum in Belgien und in den Niederlanden - Fundamenta A/3 - Die Anfänge des Neolithikums vom Orient bis Nordeuropa. Va Westliches Mitteleuropa, Köln: 153184 Vencl, S., 1961 - Studie o sáreckém typu - Sbornik DEINSEA 7, 1999 of animal bones from archaeological sites - Peabody Museum Bulletin 1 Werning, J., 1983 - Die Geweihartefakte der neolithischen Moorsiedlung Hüde I am Dümmer, Kreis Grafschaft Diepholz - Neue Ausgrabungen und Forschungen in Niedersachsen 16: 27-187 Whittle, A., 1996 - Europe in the Neolithic: the creation of new worlds - Cambridge University Press Ziegler, R., 1989 - Neolithische Tierreste aus Straubing-Lerchenhaid (Niederbayern) - Bericht der Bayerischen Bodendenkmalpflege 26/27: 9-32 národního muzea v Praze 15 (3): 93-140 Von den Driesch, A., 1976 - A guide to the measurement 184 R e c e i ved 2 February 1999 ELBURG: Neolithic animal bones APPENDIX Measurements of skeletal elements used in this study. Elements of Cervus elaphus m a rked with * belong to the hind leg depicted in Figure 6. 185 ELEPHANTS HAVE A SNORKEL! DEINSEA 7, 1999 APPENDIX (continued) DEINSEA - ANNUAL OF THE NATURAL HISTORY MUSEUM ROTTERDAM P. O . B o x 2 3 4 5 2 , N L - 3 0 0 1 K L R o t t e r d a m T h e N e t h e r l a n d s 186