Middle Palaeolithic Hunting Economies in the Rhineland

International Journal of Osteoarchaeology Int. J. Osteoarchaeol. 10: 286–309 (2000)

NICHOLAS J. CONARD* AND TIMOTHY J. PRINDIVILLE Abteilung A8ltere Urgeschichte und Quarta¨ro¨ kologie, Institut fu¨r Ur- und Fru¨ hgeschichte und Archa¨ ologie des Mittelalters, Universita¨t Tu¨ bingen, Schloss Hohentu¨ bingen, Tu¨ bingen, Germany

ABSTRACT Until the 1980s, archaeologists routinely assumed that faunal material found in association with Middle Palaeolithic artifacts represented the remains of animals hunted by Neanderthals. Over the last two decades, most researchers have come to the conclusion that diverse anthropogenic and non-anthropogenic agents can lead to the co-occurrence of lithic artifacts and faunal remains. With this taphonomic problem in mind, this paper considers the evidence for Middle Palaeolithic hunting in the Rhineland. Special attention will be paid to the recent excavations at Ariendorf, Hummerich, Schweinskopf, To¨ nchesberg and Wannen in the Neuwied Basin, and Wallertheim in Rheinhessen. The analysis of the data from the faunal assemblages themselves form the basis of this paper, but emphasis will also be placed on the taphonomic contexts, the nature of the accompanying lithic assemblages, and the topographic and stratigraphic positions of the sites, in order to address the question of how hominids responded to changing environmental conditions during the Middle Palaeolithic. This paper discusses the diversity in patterns of prey selection, skeletal part transport, seasonality and occupation intensity documented at key sites in the region, and addresses their implications for interpreting Middle Palaeolithic behaviour. Copyright © 2000 John Wiley & Sons, Ltd.

Key words: Middle Palaeolithic; prey selection; skeletal part transport; seasonality; occupation intensity

Introduction depth review of these criteria lies beyond the scope of this paper, but reviews of these issues Over the last two decades, establishing past can be found in several recent publications (e.g. subsistence practices among Palaeolithic ho- Marean, 1998; Klein et al., 1999; Marean & minids has increasingly been viewed as a key Assefa, 1999). means of addressing major trends in human The current debate over the provisioning ca- evolution. Given the relative abundance of fau- pabilities of hominids during the Middle Palaeo- nal remains over other indicators of Palaeolithic lithic of Eurasia and the Middle Stone Age of subsistence practices, the advent of effective Southern Africa has offered a wide spectrum of hunting has come to be seen as a defining interpretations concerning hominid behaviour. characteristic of human behaviour, relative to While Binford (1984) has presented Middle other primates and, especially, relative to earlier Palaeolithic and Middle Stone Age hominids as forms of hominids. In order to investigate Mid- scavengers, Stiner (1991, 1994) concludes that dle Palaeolithic subsistence, researchers have the Middle Palaeolithic people of west-central Italy practised both scavenging and hunting, attempted to define criteria for the identifica- with the former dominating the Pontinian as- tion of hunting and scavenging, and to draft semblages before ca. 55 ka, and the latter there- hypotheses about hominid behaviour. An in- after. Klein’s (1976, 1995, 1998) faunal analyses * Correspondence to: Abteilung A8ltere Urgeschichte und Quar- led him to the conclusion that Middle Stone ta¨ro¨ kologie, Institut fu¨r Ur- und Fru¨hgeschichte und Archa¨ologie des Mittelalters, Universita¨t Tu¨bingen, Schloss Hohentu¨bingen, Age people could hunt with some game-specific 72070, Tu¨bingen, Germany. limitations. He argues further that Later Stone Copyright © 2000 John Wiley & Sons, Ltd. Received 19 December 1999 Accepted 27 February 2000 Hunting Economies of the Rhineland 287

Age faunal assemblages provide evidence for Data from studies of relative skeletal abun- more complex and intensive exploitation of ter- dance, bone modification and mortality patterns restrial and marine faunal resources, and that provide three lines of evidence used to investi- these skills, combined with other cultural devel- gate both the subsistence strategies of hominids opments, led to the expansion of modern hu- and the activities of other taphonomic agents in man populations out of Africa and into Eurasia the past. These data may be augmented with around 50 ka. On the basis of faunal analyses information on seasonality, environmental set- from Combe Grenal in France, Chase (1984, ting and associated lithic technology. Inconsis- 1988, 1989) has argued in favour of effective tencies in the recording and interpretation of hunting, which did not essentially differ from faunal data, small sample sizes, highly complex hunting practices documented in the faunal as- taphonomic histories, and the anticipated high semblages of Upper Palaeolithic sites. degree of variability in hominid behaviour can Hypotheses in favour of scavenging have re- make results of these analyses difficult to assess. cently received significant critique (Marean, While Middle Palaeolithic faunal assemblages 1998; Marean & Kim, 1998; Mussi, 1999), and remain difficult to interpret, discussions of sub- many new data further support the interpreta- sistence during this period have started to focus tion of frequent, active hunting by both Nean- less on the simplistic dichotomies between derthals and their contemporaries outside of hunting and scavenging and the differences be- Europe (e.g. Conard, 1992, 1997; Farizy & tween modern and pre-modern humans. Rather, David, 1992; Gaudzinski, 1995a,b, 1996; many researchers are now examining the dy- Conard et al., 1998; Marean, 1998; Milo, 1998). namics of past subsistence and settlement sys- Support for the hypothesis that hominids were tems, in which multiple variables in subsistence strategies are addressed to gain insight into the active hunters during the Middle Palaeolithic lifeways of Palaeolithic hominids. With these can also be found in the analysis of stone and introductory remarks in mind, this paper will organic tool technology. Some Middle Palaeo- consider some of the key Middle Palaeolithic lithic lithic assemblages have provided remains faunal assemblages from the Rhineland. After of mastic used to haft stone tools (Mania & summarizing many of the noteworthy character- Toepfer, 1973; Boe¨da et al., 1996), as well as istics of these assemblages, the paper will con- use-wear and damage patterns consistent with sider the importance of the geographic settings hafted tools and projectile technology (e.g. of sites, prey selection, mortality profiles, and Shea, 1989, 1997; Meignen et al., 1998; Plisson seasonality data. Finally, the paper will remark & Beyries, 1998). Furthermore, the remains of upon issues related to changes in Middle organic technology, such as the wooden spears Palaeolithic technology and subsistence prac- from Clacton-on-Sea (Oakley et al., 1977) and tices, the intensity of occupation and mobility Lehringen (Jacob-Friesen, 1956; Thieme and of Middle Palaeolithic hominids. Veil, 1985), as well as the discovery of several wooden handles and eight throwing-spears, in association with the remains of 20 horse individuals from primary Middle Pleistocene con- Middle Palaeolithic faunal texts at Scho¨ ningen (Thieme et al., 1993; assemblages from the Rhineland Thieme, 1995, 1997; Van Kolfschoten, personal communication), demonstrate that Lower and Following Bosinski (1963, 1982) and Conard & Middle Palaeolithic hominids were competent Fischer (2000), we divide the Middle Palaeo- hunters. Studies of the physiological characteris- lithic of Germany into two phases, the Early tics of the human digestive tract further indicate Middle Palaeolithic, which corresponds to the that the habitual consumption of carrion is a last two glacial cycles of the Middle Pleis- highly improbable hominid adaptation, ill suited tocene, and the Late Middle Palaeolithic, which to human dietary needs (Uerpmann, personal includes Eemian and Weichselian/Wu¨rmian as- communication). semblages. Relative to the Lower Palaeolithic,

Copyright © 2000 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. 10: 286–309 (2000) 288 N.J. Conard and T.J. Prindiville the Middle Palaeolithic is seen as a period of al., 1986). The classic loess profile at Ariendorf technological diversity and variability. Prepared is located on an upper terrace on the right bank cores are known from the last three glacial of Rhine between Koblenz and Bonn (Turner, cycles (Bosinski, 1967) and Mousterian assem- 1997). Wallertheim lies in an old brickyard in blages are well known from the penultimate the Wiesbach drainage in Rheinhessen. Since glaciation of Germany and neighbouring coun- Schmidtgen’s excavation in the 1920s, the site ties. Although their great diversity warns against has been the best known Middle Palaeolithic generalizations, handaxes are present in both site in the region (Schmidtgen & Wagner, the Early and Late Middle Palaeolithic. The 1929). All six of these localities have been ‘Jungacheule´en’ is seen as characteristic of the studied intensely using interdisciplinary meth- Early Middle Palaeolithic, while the bifacial ‘Mi- ods. Many papers and several monographs are coquien’or‘Keilmessergruppe’ is more common in available on these localities, and abundant geo- the Late Middle Palaeolithic. Eemian assem- logical and ecological information, as well as a blages are often too small to allow a technolog- nearly exhaustive array of luminescence and ical or typological classification, whereas the 39Ar/40Ar dates have been published (e.g. van Wu¨rmian/Weichselian assemblages show a di- den Van den Bogaard et al., 1989; Turner, 1990, versity that lies beyond those known from tradi- 1997; Van den Bogaard & Schminke, 1990; tional areas of study, such as southwestern Frechen, 1991, 1994; Zo¨ ller et al., 1991; Conard, France. Laminar assemblages are known from 1992; Gaudzinski, 1995a; Conard et al., 1995b; the early phases of the last glaciation, while Boenigk & Frechen, 1997; Zo¨ ller, 1997; Frechen diverse scraper assemblages, as well as handaxe & Justus, 1998). and Keilmesser sites, do not appear to correspond With the exception of the original research at to narrow temporal zones within the Late Wallertheim, all of the find horizons discussed Middle Palaeolithic (see Richter, 1997 for an- here were systematically investigated using other interpretation). The complexity of the modern excavation methods, which were well archaeological record makes the development of suited for sites with faunal material. All identifi- a universally applicable taxonomic system for able faunal remains and fragments greater than 5 the lithic assemblages of the German Middle cm in size were piece-plotted. Smaller finds Palaeolithic appear unlikely in the immediate were collected by layer and quarter metre, and, future. A more detailed discussion of the in most cases, systematic water screening was chronostratigraphic and taxonomic problems as- conducted to recover samples of small bone sociated with the classification of the region’s fragments and microfauna. Even in the case of Middle Palaeolithic industries can be found in a Schmidtgen’s original excavation at Wallert- recent paper by Conard & Fischer (2000). Al- heim, particular attention was paid to the recov- though the distinction between the Early and ery and analysis of all classes of mammalian Late Middle Palaeolithic is well established, and faunal material (Schmidtgen & Wagner, 1929; generally accepted in Germany, differences in Gaudzinski, 1995b). While taphonomic and in- the subsistence patterns during these periods terpretive problems commonly associated with remain to be demonstrated. Palaeolithic assemblages remain, a systematic Turning specifically to the Rhineland, this selection bias during excavation can be ruled paper will examine many new Middle Palae- out for all of the find horizons discussed below. olithic find horizons that have been excavated Thorough analysis of these faunal assemblages during the 1980s and 1990s. The localities that included the identification of shaft fragments, form the basis of this paper include Ariendorf, refitting of bones and bone fragments, and at- Plaidter Hummerich, Schweinskopf-Karmelen- tributing bones to specific individuals. This pa- berg, To¨ nchesberg, Wallertheim and Wannen. per draws heavily on the work of many With the exception of Ariendorf and colleagues who have excavated and prepared Wallertheim, the other find horizons are lo- publications on faunal assemblages in the cated on the elevated, loess filled craters of the Rhineland in recent years (Scha¨fer, 1990; volcanic mountains of the East Eifel (Bosinski et Turner, 1990, 1998; Conard, 1992; Justus, 1992;

Gaudzinski, 1995a,b; Kro¨ ger, 1995; Street, woolly rhinoceros (Coelodonta antiquitatis,50/2) 1995; Conard, 1997; Conard et al., 1998). These and mammoth (Mammuthus primigenius, 152/1), data provide the basis for a reconstruction of with fewer remains of red deer (Cervus elaphus, Middle Palaeolithic subsistence in the region. 77/2), large bovid (cf. Bison priscus,10/2) and Turner (1990) has published a palaeontological wolf (Canis lupus), 37 lithic artifacts including overview of the Middle and Upper Pleistocene unmodified flakes, cores and core fragments fauna of the region, but with the exception of (Turner, 1998). Here, as elsewhere in this paper, Conard’s (1997) report at the 1995 International the numbers in parentheses correspond to values Conference of Archaezoologists (ICAZ) meet- for NISP/MNI. While the horizon had previ- ing in Konstanz, no attempt has been made to ously been correlated with the penultimate synthesize the archaeological significance of the glaciation, Turner (1998) now argues on the faunal data from the Middle Palaeolithic of the basis of a wide variety of biostratigraphic data Rhineland. In the following section, several of and absolute dates that a correlation with the the most important faunal assemblages of the third-to-last glaciation is more likely. This early region will be considered. Saalian correlation makes the horizon the oldest under consideration in this review. An open, cold steppe environment has been Assemblages from the Early Middle reconstructed for the time of occupation. Palaeolithic Whereas only eight fragments belonging to one refitted rhinoceros humerus carry impact scars, 17 of the 482 recorded bones have been modi- Ariendorf, archaeological horizon 2 fied by carnivores. No cut-marked bones could be identified. Several remains from two rhi- The loess deposits of this Saalian-aged find noceros individuals and lithics were found spa- horizon, located upon a terrace about 100 m tially associated with an irregularly shaped above the Rhine River (Figure 1), have pre- pit-like feature, whose origin is unclear. Turner served the remains of horse (Equus sp. 47/3), rejects Bosinski’s (Bosinski et al., 1983a) sugges- tion that this feature represents the remains of a hut (Turner, 1998, pp. 136–137). Mammoth remains in this layer are most numerous, but most of the specimens are molars and molar fragments with smaller quantities of tusk, ribs, vertebrae and several long bone fragments. The majority of the postcranial elements derive from a single young individual. The horse remains from this layer represent a minimum of three individuals of different ages. Head and limb elements predominate, and some foot bones were found in articulation. Although few in number, remains of bison are represented only by limb elements. Of the 77 finds attributable to red deer, 69 are fragments of antler. The remaining bone fragments represent the heads and limbs of a minimum of at least two individuals. A total of 12 shed antler bases were recovered from this layer. These finds may Figure 1. Map of the Rhineland with sites mentioned in the not indicate the presence of deer at the site, but text. 1=Schweinskopf-Karmelenberg, 2=In den Wannen, could have been collected by hominids. The 3=Ariendorf, 4=To¨ nchesberg, 5=Wallertheim, 6=Plaidter Hummerich, 7=Kartstein, 8=Metternich, 9=cave sites in bases of two, non-shed red deer antler suggest a the Lahntal. winter season of occupation (Turner, 1998).

Nearly all species from this horizon carry diverse faunal remains from this site indicate an some traces of carnivore chewing. The presence open, cold steppe environment. Of the 644 of carnivore-induced damage could indicate that identified bones from layer IV, horse (Equus sp.) hominids played a subordinate role in the for- remains dominate the assemblage (330/11), fol- mation of the find horizon. Turner, however, lowed by relatively high amounts of woolly remarks that the general lack of cut-marked rhino (C. antiquitatis, 133/4), reindeer (Rangifer bone may well be a product of the poor preser- tarandus,60/6), and red deer (C. elaphus,93/4). vation of the bone surfaces. The remains of Mammoth (M. primigenius,4/1) is represented by elephant suggest that the entire animal was four fragments of tooth and bone. Only a small present at the site. Based on comparisons with number of carnivore remains (wolf, C. lupus,1/1; recent elephant mortality patterns and their re- arctic fox, Alopex lagopus,5/3) were recovered lationship to landscape position and individual from layer IV. Horses of all ages are present in age, Turner (1998, pp. 172–173) discusses the the assemblage, and the abundance of skeletal possibility that the remains of the young adult elements from horse and rhinoceros suggest that male elephant in horizon 2 represent a kill site. entire animals were present at the site and may The pattern of skeletal abundance from other have been hunted there (Scha¨fer, 1990, p. 99). species in this archaeological occurrence appear The other species are represented predomi- to have been chosen, and suggest that these nantly by adult individuals. Anthropogenic and remains were selectively transported to the site. non-anthropogenic modifications to bone at If anthropogenic origin of the pit-like feature is Schweinskopf are not always quantified by rejected, and the high number of species is Scha¨fer. He describes carnivore modification of considered, the finds recovered from Ariendorf this assemblage as rare (Scha¨fer, 1990, p. 109) 2 speak in favour of a locality that was used on and documents 11 bones with traces of carni- multiple occasions by both carnivores and vore chewing. In contrast, Scha¨fer describes hominids. several concentrations of broken long bone fragments bearing impact scars that appear to represent the remains of bone cracking for mar- Schweinskopf-Karmelenberg, layers IV and V row extraction by hominids. A spring occupation at Schweinskopf IV is suggested by one These Middle Palaeolithic find horizons are cranial fragment of red deer from which the located within the Saalian loess deposits, atop antler had been recently shed. the Schweinskopf scoria cone (Bosinski et al., Only two stones recovered from layer V can 1986; Scha¨fer, 1987, 1990). The locality is adja- be securely categorized as artifacts, and neither cent to the prominent Karmelenberg mountain of these can be directly associated with the that stands approximately 300 m above the faunal remains (Scha¨fer, 1990, pp. 55–56). For floor of the Rhine Valley. Systematic excava- layer V, this paper focuses on the fauna from tions have recovered the remains of animals in the 89/50-95/54 area of excavation, which association with stone artifacts (Scha¨fer, 1987, Scha¨fer (1990, p. 72) considers to be a represen- 1990). While other find horizons have been tative sample of the deposit. The faunal sample examined, the main find horizon IV, and the from layer V (n=568) is smaller than that of lower lying find horizon V, are noteworthy layer IV, and contains less identifiable material, when considering subsistence practices during making interpretation of this occurrence more the penultimate glaciation. difficult. Once again, remains of horse predomi- In layer IV, local quartz cobbles that were nate (Equus sp., 60/3) among the remains of red brought to and reduced at the site make up 95% deer (C. elaphus,25/1), reindeer (R. tarandus, of the 657 artifacts. Other raw materials with 13/1), large bovids (cf. Bison priscus,25/2) and better chipping properties were often imported woolly rhinoceros (3/2). Limb bones of red into the site as complete tools (Scha¨fer, 1990, p. deer, reindeer and bison are more abundant 137). Three bifacially worked scrapers from this than other skeletal elements. Scha¨fer (1990, pp. horizon are of typological significance. The 67–70) emphasizes the highly fragmented

Copyright © 2000 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. 10: 286–309 (2000) Hunting Economies of the Rhineland 291 nature of the appendicular elements of these to the site from a kill site outside the area of species. In contrast, the remains of a minimum excavation (Justus, 1992, pp. 102–104). Bone of three sub-adult horses are represented by a weathering and other taphonomic processes predominance of limb bones, several of which may help to explain the scarcity of modified are complete. No remains of carnivores were faunal remains and less dense faunal elements. recovered from this layer and traces of carnivore Traces of carnivore chewing are present on five modification are generally rare on all taxa horse, four rhinoceros and at least three red (Scha¨fer, 1990, pp. 68–69). Impact scars and deer bones, while one rhinoceros and six horse bones broken in a fresh state are numerous bones were broken in a fresh state. (Scha¨fer, 1990, p. 69), indicating on-site extrac- Only three quartz artifacts were recovered in tion of marrow by hominids. Scha¨fer adds that association with the 1116 bones and bone frag- the remains of red deer reflect anthropogenic ments in layer V, (Justus, 1992, pp. 105–107). transport of selected body parts to the site from As was the case in layer IV, the most numerous off-site kills, although this may have been the Equus case with other species. species in layer V are horse ( sp., 133/7), rhinoceros (94/6) and red deer (61/3), with less numerous remains of mammoth (Mammuthus sp., In den Wannen, layers IV and V 2/1), large bovid (Bos/Bison,2/1), chamois (Rupi- capra sp., 3/1) and lion (Panthera leo speleaus,4/1). Located in a volcanic crater above the surround- Justus (1992, p. 122) argues for two overlapping ing landscape (Figure 1), five of the ten distinct patterns among horse remains: one in which the geological deposits at this site preserve a large extremities and cranial remains predominate, number of faunal remains found in association and one in which whole carcasses are present. with stone tools (Justus & Urmersbach, 1987). As a result of the relatively good preservation in All but the uppermost of these layers date to the this layer, the observed pattern of skeletal ele- penultimate glaciation (Frechen & Justus, 1998). ments cannot be attributable to differential bone A cool, steppe-like environment has been recon- preservation. According to Justus (1992), a por- structed for the time in which layers IV and V tion of the horse remains were transported by were deposited during the late Saalian. In layer hominids to the site, while others represent the IV, numerous remains of woolly rhinoceros (C. remains of animals killed at or near the site by antiquitatis, 157/5), horse (Equus sp., 83/5) and carnivores. Only six horse bones preserve unam- red deer (C. elaphus,31/3) have been recovered, biguous characteristics of having been broken in in association with 25 flakes and flake fragments a fresh state by hominids, while four bones and three retouched tools, all made of local preserve damage attributable to carnivores (Jus- quartz (Justus, 1992). Less numerous remains of tus, 1992, p. 207). The axial and cranial ele- reindeer (R. tarandus,3/1), badger (Meles meles, ments of red deer are under-represented leading 3 1), wolf (C. lupis,21), large bovid (Bos Bison, / / / Justus to suggest that limb bones may have been 1 1) and mammoth (Mammuthus sp., 1 1) were / / selectively transported to the site. Conversely, also recovered. The rhinoceros remains include faunal elements from at least one juvenile and the remains of rhinoceros suggest that entire one adult individual; however, the age of the animals were present at the site and presumably other individuals could not be determined. The died there (Justus, 1992, pp. 122–123). Only skeletal abundance of a relatively complete adult two rhinoceros bones with traces of carnivore individual indicates that the entire animal was chewing could be identified; no information on present at the site. Justus (1992, p. 103) inter- anthropogenic modification is given. It seems prets this feature as a possible kill site, and that the faunal remains are attributable to the argues that this animal died on site. The remains activities of both carnivores and hominids. A of the juvenile rhinoceros, horse and red deer winter occupation at Wannen V is suggested by show a bias towards limb and head parts, sug- one fragment of unshed antler of red deer (Jus- gesting the selective transport of these elements tus, 1992, p. 140).

To¨nchesberg, find horizon 1A, upper lava loess (1992) has suggested that hominids were active in the formation of this find horizon, and that The Saalian deposits from To¨ nchesberg 1A they hunted or had early access to the carcasses were recovered from the fill of a scoria cone ca. of cervids, equids and bovids from To¨ nchesberg 100 m above the Nette River, a tertiary tribu- 1A. tary of the Rhine (Figure 1). Although the loess and scoria deposits are likely to have been the Assemblages from the Late Middle product of solufluction, the finds contained therein offer environmental and cultural infor- Palaeolithic mation about the use of the site. The general depositional environment is similar to that of Wallertheim, find horizon A Saalian horizons at Schweinskopf and Wannen (Conard, 1992). An open vegetation character- The open-air site Wallertheim lies in the fluvial ized by cool temperatures has been recon- deposits of the Wiesbach, a tertiary drainage of structed for the site at the time of deposition. the Rhine (Figure 1). Sedimentological (Becze- Three lithic artifacts were recovered from the Dea´k & Langohr, 1997) and biostratigraphical main excavation in association with a fauna, (Damblon, 1997; Mania, 1997; Turner, 1997; which included the remains of horse (E. ferus, Uerpmann & Dechert, 1997; Van Kolfschoten & Thomassen, 1997) evidence, as well as thermo- 49/2), red deer (C. elaphus, 101/4), large bovid luminscence determinations (Zo¨ ller, 1997) and (Bos/Bison,38/2), and less abundant remains of earlier comparisions with the Milankovitch reindeer (R. tarandus,8/1), woolly rhinoceros (C. curve (Schmidtgen & Wagner, 1929) and local antiquitatis,3/1), wild ass (E. hydruntinus,1/1), and terrace stratigraphy indicate interglacial condi- lion (Felis leo,21) (Table 2). A smaller excava- / tions and suggest a correlation with the Eemian. tion of the same stratum produced remains of Carbonized botanical remains were preserved horse, red deer, woolly rhinoceros and reindeer, in the fine-grained sediments of this layer. Pre- along with a quartz core, while surface collec- liminary results from analyses of these remains tions of this find horizon produced additional document the presence of wild cherry or black- faunal remains, as well as four cores, two scrap- thorn (Primus sp.) and poplar (Poplus). The pres- ers and one flake. ence of these thermophilic species reflects not Horse, red deer and large bovid represent only interglacial conditions, but also the vegeta- 89% of bone finds from the main excavation. tion cover in the area adjacent to the site. Ninety percent of 736 bone fragments of these Analysis of the molluscan fauna from layer A species derive from the limbs. The participation indicates a generally open, dry and warm of non-anthropogenic agents in the formational meadow to steppe-like landscape, with signifi- history of the deposit is documented by six cant bush and tree cover in the lower lying bones with definite traces of carnivore chewing. areas near the Wiesbach (Mania, 1997). The Although no cut-marked bones could be identi- occurrence of fallow deer (Dama dama) indicates fied, 512 of the bone fragments had been bro- the presence of wooded areas adjacent to the ken in a fresh state, of which four show clear site, as this species is known to forage in small signs of anthropogenic fractures from dynamic groups and prefer niches with lush bush and loading (Binford, 1981; Shipman, 1981; John- tree cover (Van den Brink, 1968). Data from a son, 1985). Analysis of cementum annuli of two detailed analysis of the sediment in layer A red deer teeth by Pike-Tay indicates occupa- allow for the reconstruction of a rather stable, tions in late fall or early winter and fall, whereas vegetated surface on the floodplain of the Wies- the results from a horse tooth sectioned by bach. This low-lying position in the landscape Burke indicate spring. Based on these data and received a contribution of non-calcareous allu- comparisons with a palaeontological collection vial sediments and calcareous sediments from from a similar Saalian-aged setting of the the surrounding slopes, which led to poor bone To¨ nchesberg 1A lower lava loess, Conard preservation (Becze-Dea´k & Langohr, 1997).

The analysis of the more than 6700 lithic generally poor bone preservation in the layer. artifacts recovered from this horizon document Four bovid hind limb bones exhibit impact frac- the opportunistic, primary reduction of the red tures and 18 appear to have been broken in a rhyolite and tuffaceous rhyolite that dominate fresh state. One tibia from D. dama exhibits the assemblage. These two raw materials were impact scars, and nine other bones of this spe- most likely procured from sources approxi- cies were broken in a fresh state. Only six bones mately 3 km from the site, and brought on site from this assemblage show traces of carnivore in one or more blocks and reduced within the chewing and of these, two could be attributed area of excavation. Fifteen tools of diverse lithic to E. ferus. These data indicate that the remains raw materials were recovered from the assem- of bovids and cervids are of anthropogenic blage. Superimposed among these artifacts, 382 origin, while those of horse cannot unambigu- bones and bone fragments could be docu- ously be attributed to hominids. Analysis of mented. The remains of fallow deer (23/5), cementum annuli from two specimens (D. dama large bovid (Bos/Bison,46/4) and horse (E. ferus, and Bos/Bison) suggests occupation during the 13/1) are the most numerous (Table 2); no summer (Pike-Tay, 1997). specimens of these taxa indicate sub-adult individuals in the assemblage. In addition, three bones could be attributed to wolf (C. lupis) and Wallertheim, the Schmidtgen excavations, one to beaver (Castor fiber). The remains of ‘Fundschicht’ fallow deer and large bovid show traces of human activity. Thirty-three small fragments of From 1925 to 1928, Schmidtgen & Wagner burned bone (3–19 mm in size), most likely (1929) conducted archaeological and geological attributable to fallow deer, were found in a 60 investigations at Wallertheim. Several find-bear- cm diameter, semi-circular concentration, which ing strata were identified at the site. The richest has been interpreted as the remains of a hearth. of these is usually referred to as the ‘Fundschicht’, The dark brown to white colour of the burned and from it, 542 stone artifacts and more than remains can be correlated with Shipman et al. 12500 bones and bone fragments were recov- (1984), stages II–IV, suggesting a fire which ered (Gaudzinski, 1995a). Stratigraphically, this burned at temperatures between 285–940°C. early Weichselian layer correlates most closely Not only the high temperature of the fire, but with find horizon C from the recent excavation also the small size of the fragments are indica- (Conard et al., 1995b). The lithic assemblage tive of a long burning fire (Shipman et al., 1984). consists mostly of unmodified flakes of local raw The skeletal abundance of large bovid re- materials, which appear to have been chipped mains indicates that appendicular elements and at the site. Some unifacially and bifacially teeth occur more often in the assemblage than retouched scrapers and points were also ribs, cranial elements or vertebrae. Although recovered. this may be the result of selective transport, it Although discrepancies in the exact prove- may also be a product of preferential preserva- nance of some finds do exist, the large mam- tion of animal remains with high bone densities malian fauna, in which the remains of large (cf. Brain, 1981; Lyman, 1984). The relative low bovid (66% of the assemblage) outnumber those abundance of high density bovid teeth suggests of horse, indicate an open, steppe-like environ- that differential weathering of whole skeletons ment, though indicators of more wooded areas cannot explain the pattern of skeletal elements are present in both molluscan and macrofaunal at the site. A similar pattern can be seen among (Sus scrofa) samples (Gaudzinski, 1995a, pp. the remains of fallow deer from this horizon, 280–283). although, in this case, upper limb bones are A re-analysis of the faunal material conducted particularly well represented. Only one cut- by Gaudzinski confirms Schmidtgen’s interpre- marked bone, determinable only to size-class 3 tation of the numerous bison remains (861/52), (see Table 2), could be identified in this find which she identified as Bison priscus, as resulting horizon. This is probably a product of the from hunting by Middle Palaeolithic hominids

(Gaudzinski, 1995a,b). In contrast to Schmidt- Wiesbach floodplain. As a result of the rela- gen, however, Gaudzinski does not attribute tively rapid burial, bone preservation in this horse remains (Equus germanicus and Equus przewal- deposit is better than that of find horizon A. skii, 228/13) to hominid activities. Other spe- Data from the more than 2300 stone artifacts cies, including rhinoceros (Dicerorhinus cf. found in this layer indicate that both the pri- hemitoechus,1/1), wolf (C. lupis,1/1), hyena (Cro- mary production of flakes and blades, as well as cuta sp., 1/1) and lion (Panthera leo spelaea,1/1), the rejuvenation and recycling of previously play a subordinate role in the interpretation of made tools, took place here (Conard & Adler, the assemblage. Among the bovid remains, 1997). Thirty-three retouched tools and a vari- lower limbs and head parts are over-represented, ety of points were recovered from the find and the mortality profile is prime-dominated. horizon. The debitage from this horizon has a The bison remains preserve 65 impact scars, a strong laminar component, and numerous feature not observed among other species in this backed and double backed forms are present. layer. Cut-marked bones are rare (Gaudzinski, Although the remains of large bovid (Bos/Bison, 1995a, p. 372), although this may be a product 21/3) occur in this layer, the remains of equids of poor surface preservation of bone. So far, it (E. ferus, 100/3), which belong to at least one has not been possible to obtain results from the adult and two sub-adult individuals, predomi- analysis of cementum annuli for the large bovid nate (Table 2) and exhibit the most modifica- teeth, although Gaudzinski suggests that the tions. In addition, remains attributable to wild presence of several young animals could indi- ass (E. hydruntinus,7/2), wolf (C. lupis,1/1), cate a summer occupation (Gaudzinski, 1995a, beaver (C. fiber,1/1) and lion (25/1) were identi- p. 298). This find horizon provides the fied. Cut-marked horse bones (n=14) include strongest evidence for multiple episodes of the ventral surfaces of the pelvis and several monospecific hunting during the Middle Palaeo- lumbar vertebrae, as well as the medial surface lithic of the Rhineland, and Gaudzinski (1996) of the scapula. This concentration of cut-marks has addressed this issue in a pan-European is atypical for filleting or disarticulating context. (Binford, 1981, p. 106), and warrants further investigation. Two distal humerus fragments from E. hydruntinus are also cut-marked. Six im- Wallertheim, find horizon D pact scars were identified on shaft fragments of E. ferus, and one on a humerus fragment of E. Information on palaeoenvironments at the time hydruntinus. The cementum analysis of one horse of deposition of find horizon D indicates inter- (E. ferus) tooth indicates a summer death (Burke, mediate conditions after the Eemian Interglacial 1997). that probably correlate with oxygen isotope Skeletal abundance data of horse indicate that stage 5c (Conard et al., 1995b). Evidence from both axial and appendicular elements are molluscs found in this stratum suggest rather present, with an unusually low frequency of warm, dry summers and cool, but not extremely metapodia and phalanges. This cannot be ex- cold winters (Mania, 1997). The wooded areas plained by differential preservation as these ele- associated with find horizon A gave way to ments are relatively durable, but may instead increased amounts of open grassland, with only suggest selective removal of these parts from the occasional bushes and trees adjacent to the area of excavation by carnivores or hominids. low-lying site on the floodplain. No molluscan Alternatively, the horses may have been killed forms typical for true forest settings were found, elsewhere by hominids, who then carried se- although the presence of some hydrophilic lected parts to the site, as Conard (1992) has forms indicate the presence of standing water suggested for early Weichselian finds from near the site. Analysis of the sediments of this To¨ nchesberg 2B. It is clear that hominids were layer do not indicate a stable, vegetated surface, processing portions of horses at the site, and the but rather one on which primarily clayey, non- apparent selection of equine over other species calcareous fluvial sediments aggraded on the in this archaeological occurrence may be a

Copyright © 2000 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. 10: 286–309 (2000) Hunting Economies of the Rhineland 295 result in part of the increased amount of grass- ence of diverse archaeological and palaeonto- land that is attractive to horses. logical signatures in the Upper Pleistocene sediments of the Wiesbach floodplain.

Wallertheim, find horizon E Wallertheim, find horizon F This find horizon is dominated by the remains from at least two well-preserved large bovids This deposit preserves the remains of at least 15 (Bos primigenius, 361/2), which were deposited on E. ferus (643/15) individuals, and at least one E. the marshy banks of the Wiesbach under rela- hydruntinus (27/1, Table 2). Horse remains repre- tively warm conditions that probably corre- sent more than 90% of the identified faunal spond to OIS 5c. Additional, less numerous remains by count and weight (Prindiville, 1998). species identified in this layer include: horse Other, less numerous species include large (44/5), wild ass (2/1), wolf (32/1), red deer bovid (Bos/Bison,21/2), red deer (C. elaphus,1/1), (3/1), fallow deer (1/1) and lion (1/1) (Table 2). fallow deer (D. Dama,2/1), roe deer (C. capreolus, The presence of charcoal from oak (Quercus sp.) 3/1), rhinoceros (Rhinocerotidae, ind., 1/1), lion and black poplar (Poplus nigra), as well as un- (1/1), bear (Ursus sp., 30/1) and wolf (1/1). In burned leaves, tentatively identified by addition to other environmental indicators, the Damblon (1997) as black poplar suggest warm presence of charcoal from species including moist conditions, which are consistent with the spruce (Picea), pine (Pinus) and birch (Betula) geographic setting of the site, and the results suggests that this deposit formed under cooler from other ecological analyses (Mania, 1997; conditions, with more open grassland than Van Kolfschoten & Thomassen, 1997). The fau- those of find horizon E (Damblon, 1997). nal remains from this horizon were found in Ninety-nine stone artifacts of numerous raw association with 17 stone artifacts, including materials and diverse forms were recovered from four tools, made on diverse raw materials. No this layer. Despite the fact that one horse bone lithic refits could be found, and the horizon was used for retouching lithic artifacts, the lithic provides little evidence for on-site knapping. assemblage itself provides relatively little evi- These observations, along with the faunal stud- dence for primary on-site stone knapping. Thus, ies, suggest that the artifacts probably represent like Wallertheim E, this lithic assemblage ap- a background lithic assemblage, which accumu- pears, at least in part, to represent a background lated independently of the bovine fauna accumulation, which formed independently of (Conard, 1998). Many of the bovid remains the majority of the faunal remains (Conard, were found in articulation, and the skeletal rep- 1998). The mortality profile indicates a predom- resentation of these adult individuals does not inance of sub-adult and senile individuals, which indicate that particular elements were selected, Levine (1983) describes as attritional. While at but rather that the animals had died at the least one adult individual is represented by all location where they were excavated. No clear skeletal elements, other individuals are only rep- traces of carnivore or anthropogenic modifica- resented by limb bones. Ten impact scars and tion could be found. Analysis of the cementum nine cut-marked horse bones indicate that ho- annuli from four bovid teeth indicates deaths in minids butchered and extracted marrow from late summer and late winter to early spring several of the horses, but traces of carnivore (Pike-Tay, 1997). The presence of complete chewing (n=26) are the most abundant type of bovid carcasses and the apparent lack of modifi- modification. Seasonality data for this layer cations suggest that the large bovid remains in derive from the incremental analysis of two this layer were deposited in the absence of horse teeth (E. ferus), which indicate both sum- hominid activity (Conard, 1999). In the light of mer and winter deaths (Burke, 1997). The pres- the assemblage excavated earlier by Schmidt- ence of several very young individuals gen, the seemingly natural bovid-dominated demonstrates mortality during the warmer faunal assemblage in layer E points to the pres- months, while studies of tooth eruption indicate

Copyright © 2000 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. 10: 286–309 (2000) 296 N.J. Conard and T.J. Prindiville deaths during multiple seasons (Prindiville, in concentrations that could be refitted. Al- 1998). As was the case at Wannen V, a combi- though several limb bones from bovids were not nation of anthropogenic and non-anthropogenic broken, all of the major limb bones of equids agents was responsible for the accumulation of and cervids have been cracked by hominids for equids in F. extracting marrow. Pike-Tay’s examination of a single red deer tooth suggests that the occupation occurred in the late fall or early winter To¨nchesberg, find horizon 2B (Conard, 1992). To¨ nchesberg 2B offers evidence for off-site hunting of cervids, equids and, A wide variety of environmental indicators re- possibly, bovids, followed by the transport of covered from this find horizon allows the recon- high utility meat and marrow-bearing body struction of an open habitat at the time of parts to the site for further processing and occupation, which has been dated to one of the consumption (Conard, 1992). first cooling episodes after the Eemian interglacial (Bittmann, 1990; Van Kolfschoten, 1990; Conard, 1992). Several dozen burnt faunal re- Plaidter Hummerich, ‘Hauptfundschicht’, D1 mains and seven burnt lithic finds were recovered ina4m2 concentration, and have been Finds from Plaidter Hummerich, whose main interpreted as the remains of a hearth. The find horizon dates to the early part of the last lithic assemblage includes 557 chipped artifacts, glaciation (Bosinski et al., 1983b; Kro¨ ger, 1987, and is dominated by locally available quartz, 1995; Street, 1995) were recovered from a vol- with lesser amounts of quartzites and siliceous canic crater high above the surrounding land- slates from the region, and flints from 100 km scape. Of all the Palaeolithic sites in the East or more distance. Noteworthy, is the abundance Eifel, Plaidter Hummerich has been described as of laminar debitage and diverse backed blades the most exposed, offering the least protection and several backed points, including several from the elements (Bosinski et al., 1986). Open, with damaged tips. If the 574 fragments of steppe-like conditions have been reconstructed antler are counted, the most abundant species in for the time of occupation (Street, 1995). The this archaeological occurrence is red deer. How- lithic assemblage from layer D1 contains more ever, for the purpose of examining subsistence than 500 pieces, and is dominated by artifacts and hunting, the shed antlers can be excluded, made of local quartz. Local materials appear to leaving 32 faunal remains from three individuals. have been knapped at the site, whereas several Remains of horse (Equus sp., 27/2) and large artifacts on exotic raw materials were brought bovid (Bos/Bison,56/4) are also relatively numer- into the site as completed tools. As is the case at ous, with smaller amounts of rhinoceros (D. other sites in the East Eifel, flints from the hemitoechus,9/1), fallow deer (D. Dama,4/2), wild Meuse Valley over 100 km away are present in ass (E. hydruntinus,1/1), lynx (Lynx lynx,(5/1), the assemblage. Occasional bifacially worked fox (Vulpes vulpes,21/1) and hyena (Hyaenidae sp., tools are present among several retouched flakes 1/1) (Table 1). The relative frequency of skele- (Kro¨ ger, 1987, 1995). The broad spectrum of tal elements of large bovid represented in the taxa identified in the assemblage (Turner, 1990; assemblage deviates slightly from those of horse Street, 1995) is dominated by red deer (C. and red deer. Horse and red deer share a similar elaphus, NISP=310), horse (Equus sp., NISP= body part distribution, in which upper limb 164) and large bovid (Bos/Bison, NISP=180). bones predominate, and both axial and cranial This pattern is similar to that of other Early elements are rare or completely absent. While Weichselian sites in the Rhineland. Although 15 bones exhibiting traces of carnivore chewing the skeletal distributions of the main species are could be observed, none of the bones preserve limb-biased, and the assemblage could be clear cut-marks. The majority of long bone viewed as the product of selective transport by fragments from equids and cervids carry impact hominids, interpretations of this layer must be fractures, and several of these bones were found treated with caution. Information on seasonality

Copyright © 2000 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. 10: 286–309 (2000) Hunting Economies of the Rhineland 297 Weight 8 1313 498 5 340 13 510 1 n.a. 546 3 307 20 603 976 18 645 211 19 029 394 5397

n rhinoceros, elephant, = 83 360 Weight 2472 1208 103 449

n Weight horse, aurochs, bison; 4 =

n 77 938 135 6558 37 2832 2184 122 094 Weight 9 509 9 509 141 110 1 4 4 677 510 498 1 3 29 200 21 160 2 8 24 170

n Weight 3 905 2 94 1 22 1 205 9 422 4 53 45 43 19 n.a. 4 n.a. 72 96 17 48283 606 8659 59 673 27 67 3022 4976 9 12 1887 749 812 3 572 69 186 74 1077 191 4440 56 620 21 13 48 108 2218 35 479 13 188 11 639 921 1B 2B 2C 2D Total n 114 10 562 730 73 498 79 5938 16 n.a. 3 n.a. 19 72 46 257 7162 9512 2122 1914 6343 boar, fallow deer, red deer, reindeer, wild ass; 3 Weight = 3 1 8 1 1 5 2 38 1638 56 10 000 1 13 99 11 831 540 191 n 209 408 109 447 Weight 4636 beaver, fox, roe deer; 2 = 2 3 150 12 9 254 101 1 11 27 186 4 180 40 3633 49 2394 58 1653 320 60 154 7272 749 15 855 1A (lower lava loess)n 1A (upper lava loess) . ¨ nchesberg, faunal remains from find horizons 1A, 1B, 2B, 2C and 2D indet. Bison / Total, indeterminateTotal 94 2636 Indeterminate, size classIndeterminate, 2 size class 3 34 57 976 36 2066 2 141 4 277 477 12 275 Indeterminate, size class 4 1 525 2 42 Small Caprid Table 1. To Counts of identifiedExamples and of unidentified indeterminate remains size with classes: approximate 1 weight in grams. Total, identifed Indeterminate Indeterminate, size class 1 Bos Cervus elaphus Dama dama Equus hydruntinus Equus ferrus Coelodonta antiquitatis Diceerorhinus hemitoechus Rangifer tarandus Canis lupus Vulpes vulpes Felis leo Lynx lynx Aves Hyaenidae indet mammoth.

Copyright © 2000 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. 10: 286–309 (2000) 298 N.J. Conard and T.J. Prindiville and modifications to bone is not yet available. the same location on an isolated hilltop, and Despite the occasional presence of articulated been buried with numerous lithic artifacts with- bones, the preservation in horizon D1 is gener- out active predation, hominid hunting of these ally poor and portions of the deposit appear to animals appears likely. The season of death of have been post-depositionally reworked in mul- the animals, based on tooth eruption, indicates tiple episodes (Street, 1995). The under-repre- that several distinct episodes are preserved here. sentation of vertebrae, ribs, cranial elements and cancellous tissue as well as cut-marked bone may be the result of bone weathering (Street, Discussion 1995). While Middle Palaeolithic hominids are responsible for a portion of the large mammalian fauna, the difficult taphonomic setting Geographic setting and site preservation has hindered several analyst’s attempts to isolate this component of the assemblage. The comparison of faunal assemblages in the Rhineland is impeded by variable bone preservation. This problem is seen most acutely in find To¨nchesberg 1B horizons such as To¨ nchesberg 2A, the main Middle Palaeolithic horizon at Metternich The youngest Middle Palaeolithic faunal assem- (Conard et al., 1995a), and the horizons at the blage in the region is that of To¨ nchesberg 1B. locality of Rheindahlen (Thieme, 1983; Thissen, This find horizon lies at the base of the Middle 1986), where abundant lithic artifacts have been Weichselian loess, and has been dated both recovered, but faunal material is absent. While chronostratigraphically and with TL to ca. 65 ka useful in addressing a number of other ques- (Conard, 1992). The sedimentological setting tions, these sites provide little information to- and the faunal assemblage are indicative of an ward addressing the issues of Middle Palaeo- open grassland environment. The lithic assem- lithic hunting and subsistence. In some cases, blage includes 120 chipped artifacts of diverse such as To¨ nchesberg 1A, 1B and Hummerich raw materials, of which quartz is the most nu- D1, faunal assemblages have been recovered merous. The faunal remains are dominated by from reworked contexts that hinder their inter- horse (83/5), with lesser amounts of red deer (C. pretation. Furthermore, the conspicuous absence elaphus,17/1), reindeer (R. tarandus,9/1), woolly or low frequency of cut-marks and other surface rhinoceros (C. antiquitatis,3/1) and wild ass (E. modifications in the assemblages from nearly all hydruntinus,2/1). The sediments are probably find horizons suggests that imperfect preserva- not in an in situ context, but could not have tion has led to the loss of information on been moved far from their original positions. butchery practices. While the last two decades The assemblage contains the remains of five of fieldwork have provided the information nec- young horses, all of which would still have been essary to open a dialogue on Middle Palaeo- members of family herds at the times of their lithic subsistence systems, faunal analysts must deaths. Mandibles and limb elements are abun- still view the current data as a fragmentary, dant, while most of the axial skeletal remains are fossil remnant of a much more complicated poorly represented, possibly a result in part of whole. The several dozen known find horizons taphonomic attrition of elements with low bone located in sedimentary basins represent a frac- density. The remains of other species are too tion of the total number of sites from this few in number to allow meaningful conclusions period, spanning three glacial cycles and ca. to be drawn. In comparison with the faunal 250000 years. Thus, we are looking at only the assemblages from To¨ nchesberg 2B, the bones in smallest pieces of the original picture, which has horizon 1B had not been broken as frequently largely been destroyed by a variety of tapho- for marrow. Poor surface preservation prevents nomic processes. The placement of the known the identification of cut-marks. As it is improba- sites is entirely shaped by the positioning of ble that five young horses would have died in brickyards and lava, pumice and gravel quarries,

Copyright © 2000 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. 10: 286–309 (2000) Hunting Economies of the Rhineland 299 which expose the deeply buried Middle and cance of the absence of Weichselian-aged find Upper Pleistocene sediments containing sites. horizons at Ariendorf should not be over-inter- Despite the selection of sites from mining preted. The absence of Early Middle Palaeo- areas, the preceding survey of Middle Palaeo- lithic find horizons at Wallertheim is not lithic occurrences from the Rhineland reveals surprising, as pre-Eemian deposits are generally the exploitation of a wide variety of geographic lacking at the site. The general lack of low-lying and topographic settings ranging from elevated sites dating to the Early Middle Palaeolithic volcanic craters in the East Eifel (Schweinskopf, probably reflects the rarity of low-lying Saalian Wannen, To¨ nchesberg, Hummerich and oth- exposures that could contain such find hori- ers), to sites on small tributaries of the Rhine zons. Both Metternich (unpublished fieldwork, (Wallertheim), as well as sites within the Mosel Conard) and Rheindahlen (Thieme, 1983), how- Valley (Metternich) and the Rhine Valley ever, preserve pre- and post-Eemian finds, albeit (Ariendorf). Cave sites in this region, such as lacking fauna, suggesting a level of continuity in Kartstein (Rademacher, 1911) and sites in the the placement of sites. neighbouring Lahntal (Terberger, 1993), are Concerning the location of sites within the rare, and offer limited information on subsis- scoria cones of the East Eifel, Scha¨fer (1990, pp. tence because of their early discovery, excava- 108–111, 117–118) has suggested that these tion and selective recovery methods. Some positions may have been sought out as unique trends are visible in the stratigraphic and sedi- biotopes offering standing water, different vege- mentary contexts of Middle Palaeolithic sites in tation and shelter, relative to the surrounding the region of study. Sites from the Early Middle landscape. If this had been the case, these previ- Palaeolithic are almost exclusively limited to ously attractive settings would have ceased to be loess deposits, while Late Middle Palaeolithic a focus of human activity, as the basins were sites, and, especially, those from the early filled with sediment during the Late Pleistocene glacial, are often found in humic deposits. The (Conard, 1992). This shift is clearly docu- only securely dated Eemian find horizon is mented by the near absence of Upper Palaeo- Wallerthiem A, and the lack of other sites of lithic finds from the scoria cones of the East similar age is almost certainly related to the Eifel. After more than 100000 years of relatively frequent erosion of interglacial sediments in the intense Middle Palaeolithic occupation of these region. The humic deposits of the later Early mountain top localities, this change reflects a Weichselian, which probably correlate to cold dramatic shift in land-use. dry phases of OIS 5a, do not contain finds (e.g. Conard et al., 1995b). Finally, aside from the equid assemblage at the base of the Middle Prey selection and transport Weichselian loess at To¨ nchesberg 1B, no sites mentioned above date to either OIS 4 or the The predominance of a small number of species Middle Weichselian. This is particularly surpris- within a faunal assemblage can reflect prey ing, as in other regions of central and western selection by hominids, a feature that has been Europe, this period represents a phase of inten- used as a criterion to differentiate Middle and sive occupation. Upper Palaeolithic hunting economies (Mellars, A temporal shift in the positioning of sites 1989). For example, reindeer-dominated Mag- during the Early and Late Middle Palaeolithic is dalenian assemblages have been interpreted as a not readily detectable. Finds from the scoria reflection of a highly organized hunting econ- cones of the East Eifel are equally numerous omy (Enloe, 1991). However, as Mellars (1989) before and after the last interglacial. The new points out, these specialized faunal assemblages interpretation of the Ariendorf stratigraphy are dependent upon a variety of other extrinsic (Turner, 1998) does not allow for a Weichselian and intrinsic variables (location, season, group- find horizon, but the broad temporal spacing of size), and their interpretation must be treated horizons 1–3 indicates that this locality was with caution. Chase (1989) notes that special- used over multiple glacial cycles. The signifi- ized hunted assemblages in the Late Upper

Palaeolithic, although impressive, are relatively always present, but never truly dominate an infrequent, may represent only one part of a assemblage. So far, the rich assemblage of bovid larger subsistence strategy, and can be found in remains from Schmidtgen’s excavations at the Middle Palaeolithic as well. The osteometric Wallertheim presents the best example of spe- analysis of faunal remains from over 20 Eu- cialized hunting in the Middle Palaeolithic of ropean Middle and Upper Palaeolithic sites led the Rhineland (Gaudzinski, 1995a,b). However, Weinstock-Arenovitz (1997, pp. 187–188) to the diversity of site types and prey species the conclusion that a predator–prey relation- documented at Wallertheim over a period of ship between hominids and reindeer has existed several tens of thousands of years should not be since oxygen isotope stage 6, and one may underestimated (Conard, 1998; Conard et al., assume that this relationship also existed with 1998). other species. This hypothesis is supported by Recognizable patterns in the transport of ani- the increasing number of open-air Middle mal resources can also be seen among the sites Palaeolithic sites dated to the last glacial cycle, discussed above. These patterns in skeletal in which the remains of large bovids predomi- abundance of species at times correlate with the nate (Gaudzinski, 1996). It seems likely that the position of sites in the landscape. For example, economic strategies behind the monospecific the body part distributions of taxa from the tops assemblages at sites like Wallertheim (Schmidt- of scoria cones, like To¨ nchesberg 2B and 1A, gen & Wagner, 1929; Gaudzinski, 1995a), Le Wannen IV and V and Schweinskopf V, suggest Borde (Jaubert et al., 1990) and Mauran (Farizy et that horse, red deer, large bovids and, perhaps, al., 1994) may have been different from those of juvenile rhinoceros were acquired off-site, and the Late Upper Palaeolithic. These differences that their limbs were selectively transported to may not only reflect changes in hominid cogni- the sites. Some of the faunal constituents of tive abilities, but also changes in the environ- lower-lying sites, like Wallertheim E, F and the ment, hunting practices and social behaviour. Schmidtgen excavations, suggest that these ani- During the Middle Palaeolithic of the mals died at or near where their skeletons were Rhineland, several differences in prey selection excavated. These sites represent positions in the exist between sites. Find horizons containing landscape where hominids hunted animals, or, elephants and rhinoceros are more common at times, scavenged from fresh carcasses. In before the Eemian Interglacial than after. While contrast, Wallertheim A and D represent Late mammoth is virtually absent from Late Middle Middle Palaeolithic sites to which skeletal por- Palaeolithic faunal assemblages, both woolly tions from multiple kills were repeatedly trans- rhinoceros (Coelodonta antiquitatis) and narrow- ported. At the Early Middle Palaeolithic sites, nosed rhinoceros (Dicerorhinus hemitoecus) are oc- including Schweinskopf IV, skeletal part data casionally found in Late Middle Palaeolithic indicate that horses and even rhinoceros were assemblages (Turner, 1990). Bovids, cervids, and hunted at or very near the site. The size of equids are usually the key game species quarry also appears to have influenced skeletal throughout the Middle Palaeolithic and are frequencies. The elephant remains from Arien- present in varying frequency, at times to the dorf 2, and the rhinoceros remains from Wan- near exclusion of other species. Reindeer, while nen IV, document the presence of entire present in Weichselian horizons including animals, suggesting that these animals died or To¨ nchesberg 1B, is more frequent in the sites were killed on site. with Saalian loess deposits (Turner, 1990). The In summary, two distinct patterns of body fluctuating abundances of bovids, cervids and part representation can be identified: equids is illustrated in the comparisons of find horizons from To¨ nchesberg and Wallertheim, 1. Sites in which most skeletal elements of presented in Tables 1 and 2. Here, one sees that animals are present. In the Rhineland, this the abundance of both bovids and equids, based pattern is particularly common with larger on NISP and weight, occasionally represent well mammalian species, and is unknown among over 75% of assemblages. Cervids are nearly game species smaller than large equids.

n 1879 123 597 1937 19 841 = Weight 43 517 6160 1601 315 30 1953 30 1953 21 1857 642 65 966

n Weight

n red deer, horse, aurochs, bison; 4 = 9911 721 60 178 1206 48715 3816 143 438 Weight 625 32 38 3 8 3169 2 2 8 27

n 122 625 77 733 55 1043 542 1242 Weight 3 197 4 23222 296 2 71 3993 2 2 94 1 25 87 1524 1 27 1 32 811 3 159 38 1593 1 9 31 34 162 1218553541 2 5 1 2 11 54 28 2 45 222 41 304 40 201 9 59 74 511 15 358 4 52 8 127 31 54 2746 21 1202 361 53 637 61 3620 100 3340 44 1507 643 10 94 1 3 1 9 1 3 295458 3807 11 198 326 503 3304 268 4590 440 163 7391 177206 6607 3084 453 157 55 588 2370 766 179 3790 308 Cn D E F Total 4121083 58 4 27 288 158 281 657 boar, fallow deer, reindeer, wild ass; 3 2903 5198 9220 2223 3628 1135 16 713 1862 38409 892 49 250 = Weight 16 9 3 340 6 1 25 123 8 64 1 2 1 4 2 6 60 391 31 321 186 n 225 beaver, fox; 2 96 62 512 144 = 1274 1618 Weight 12 7 45 5 1 39 3 95 99 23 192 31 152 49 330 46 257 139 6267 13 382 3224 546 12 123 157 180 86 350 AB n . 2 80 1 95 4 832

ind , , ind. sp. sp. indet. Bison / Counts of identifiedExamples and unidentified of remains indeterminatemammoth. with size-classes: weight 1 in grams. Total, identifed Indeterminate, size-class 3 Indeterminate, size-class 4Total, indeterminateTotal 287 1950 100 7 297 Indeterminate Indeterminate, size-class 1 Indeterminate, size-class 2 Table 2. Wallertheim, faunal remains from six find horizons (A–F) Bos Bos primigenius Capreolus capreolus Cervus elaphus Dama dama Cervidae Sus scrofa Equus ferus Equus hydruntinus Equus Rhinocerotidae Castor fiber Ursus Canis lupus Vulpes vulpes Panthera leo Aves Crocuta crocuta

2. Sites in which limb bones of higher utility ity pattern should be expected or seen as a predominate and have frequently been bro- universal signature for hominid hunting. For ken open to extract marrow. This type of example, the assemblage from To¨ nchesberg 2B distribution is particularly prevalent among can be characterized as containing prime age cervids, bovids and equids, and is docu- adults that were hunted in multiple episodes mented in both the Early and Late Middle from the surrounding populations of bovids, Palaeolithic. cervids and equids. In contrast, the assemblage from To¨ nchesberg 1B is dominated by very Find horizons with over-representation of heads young horses, apparently taken from family and distal limb bones are uncommon. This may herds. The occurrence of old and young animals be explained by the thorough collection proce- in an assemblage may at times be the result of dures used during excavation, and by the heavy deliberate hunting, and not necessarily the re- emphasis on studying and refitting all classes of sult of hominid scavenging or carnivore kills. remains, including limb bone fragments (cf. Just as modern human and non-human hunters Marean & Kim, 1998). do not exclusively restrict their hunting to prime age animals, there is little reason to ex- Mortality profiles pect this to be the single signature of hunting for all prey species among Palaeolithic hunters. For many species at sites from the Rhineland, age profiles are of limited use because the samples are often too small to produce meaningful Seasonality results. However, when a sufficient number of individuals are present, several recognizable pat- The seasonality data from several sites in the terns can be observed. Remains of horses from Rhineland indicate occupation at different times all age groups are present in the Saalian deposits of the year, which is to be expected in any at Schweinskopf IV (Scha¨fer, 1990), suggesting region in which year-round occupation per- that no single age group was selected over sisted. On the basis of cementum annuli studies another. Prindiville (1998) describes the horse of bovids, cervids and equids (Burke, 1997; mortality profile at Wallertheim F as ‘attritional’, Pike-Tay, 1997), the two well-preserved camps as both very young and very old individuals of Wallertheim find horizons A and D on the predominate, whereas Justus (1992) suggests Wiesbach floodplain both represent summer oc- that two different patterns among horse remains cupations. Studies of equid tooth eruption are present at Wannen V. In contrast to this, (Prindiville, 1998) and cementum annuli (Burke, the majority of bovids from Schmidtgen’s exca- 1997) indicate multiple seasons of death at vations at Wallertheim are attributable to Wallertheim F. Provided these determinations ‘prime-aged’ adults (Gaudzinski, 1995a). can be seen as indicators of hominid presence, These sites demonstrate variable, and in some they suggest more complex seasonal use of the cases, overlapping mortality patterns, rather Weisbach floodplain (Conard et al., 1998). At than the dominance of a single idealized age Wallertheim E, cementum annuli studies indi- profile, as has been presented in other Palaeo- cate bovid deaths in both the warm and cold lithic contexts (cf. Klein, 1982; Levine, 1983; parts of the year (Pike-Tay, 1997). To¨ nchesberg Stiner, 1990). Specific animal behaviour and its has also yielded a complex pattern of occupa- relationship to an animal’s age and sex must also tion. Pike-Tay (1997) has identified late autumn be considered when interpreting mortality pro- and early winter as the season of death of red files. One cannot necessarily assume that the deer at To¨ nchesberg 1A and 2B. Burke’s (1997) hunting of prime-aged animals was the goal of cementum annuli determination on a single Palaeolithic peoples. As Turner (1998, p. 172) horse tooth indicated a spring death at the has pointed out, for some species, including upper lava loess of To¨ nchesberg 1A. elephant, it may have been easier to hunt young A purely palaeontological accumulation at and old animals. Furthermore, no single mortal- To¨ nchesberg 1A offers an example of one

Copyright © 2000 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. 10: 286–309 (2000) Hunting Economies of the Rhineland 303 assemblage that formed independently of ho- others have discussed, the level of surface minids. In contrast to the archaeological hori- preservation at sites in the Rhineland often zons, this palaeontological accumulation yielded leaves little possibility for identifying cut-marks. winter and late winter seasonality, based on The levels of carnivore-chewed bone, while cementum annuli for horse and red deer, respec- marginally higher, are still far lower than those tively (Conard, 1992). This period of nutritional of assemblages published from other regions. stress is when a high proportion of deaths The most common indications of anthropogenic independent of hominid influence could be modification to bones from the Middle Palaeo- expected. lithic of the Rhineland are impact fractures, Unfortunately, many find horizons in the which result from dynamic loading of compact region have yet to yield reliable seasonality bone (Johnson, 1985). These modifications are data. However, annual cementum studies may more commonly preserved, but of only marginal provide the best prospects for gaining additional consequence, as the act of bone cracking does data, which would be useful in confirming or not require that hominids have hunted the ani- refuting the following patterns: mal in question. In the Rhineland, the main evidence for hunt- 1. The Wiesbach floodplain in Wallertheim ing consists of repeated cases in which high served as an attractive summer camp to utility body parts of predominantly prime-aged which hominids transported high quality prey have been transported to sites for process- body parts of individual bovids, cervids and ing and consumption. While neither of the equids. these observations rigorously prove that hunting 2. Red deer found in archaeological contexts at was, in every case, the mode of acquisition, they To¨ nchesberg represent late autumn or early do indicate repeated access to high utility body winter kills. parts. These observations suggest hunting or, at 3. Predominantly non-anthropogenic faunas in- a minimum, early access scavenging as a mode clude examples of deaths both during the of procurement. Given the well-documented cold season, as well as from multiple existence of wooden hunting spears in the seasons. Lower and Middle Palaeolithic of Germany, More data on the seasonal use of sites in the and the presence of stone points among the region is critical for any attempts to reconstruct Middle Palaeolithic artifact assemblages of the the Middle Palaeolithic settlement systems of Rhineland, hunting is the most likely explana- the Rhineland. tion for the composition of these faunal assemblages. In addition, evidence for both seasonal patterns of prey exploitation and the curation Hunting versus scavenging and maintenance of hunting tool kits (Conard, 1997; Conard & Adler, 1997) suggest a system- Two commonly cited criteria for scavenging atic and controlled use of the landscape in included an over-representation of low utility which hunting was carefully planned and exe- skeletal parts, particularly heads and feet, and cuted to meet the needs of Middle Palaeolithic the predominance of attritional mortality pro- groups. files characterized by an abundance of very Nonetheless, the likelihood is high that scav- young and very old individuals (Binford, 1981, enging augmented the more dominant mode of 1984, 1985; Stiner, 1991, 1994; Marean, 1998; food procurement via hunting. As O’Connell et Marean & Kim, 1998). Rarely, if ever, can both al. (1988) have demonstrated, scavenging plays of these criteria be met with faunal assemblages a detectable role among the food procurement of the Rhineland. strategies of contemporary hunters and gather- The generally low frequency of cut-marked ers. There is no need to view scavenging as bone at sites in the Rhineland could also be either fundamentally undesirable relative to seen as a further argument for scavenging. How- hunting, or as indicative of lesser cognitive skill ever, as Conard (1992), Turner (1998) and when comparing hominid populations. Barring

Copyright © 2000 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. 10: 286–309 (2000) 304 N.J. Conard and T.J. Prindiville behavioural taboos, hominids from all time peri- Middle Palaeolithic sites in the Rhineland repre- ods would have welcomed the windfall of scav- sent either: enging a freshly dead animal when the 1. remains of large camps which supported circumstance arose. larger groups of people for relatively long periods of time; Lithic technology and subsistence 2. or sites which were visited by smaller groups on several occasions. In the Rhineland, most lithic assemblages asso- Problems of demonstrating contemporaneity of ciated with faunal remains contain high percent- finds in a single geological horizon can hinder ages of unmodified flakes, low percentages of interpretation in these contexts (Conard & formal tools and no clear correlation between Adler, 1997; Conard et al., 1998). The lithic lithic types and particular hunting or butchering assemblages associated with the faunal remains activities. As a result of the abundance of non- from these sites are often small, and there is flint raw materials, microwear studies have little evidence for structures during the Middle rarely been attempted, effectively eliminating Palaeolithic of the Rhineland. Such features, this line of inquiry. The only lithic assemblage when present, usually represent the remains of types which clearly stand out among the find hearths as at To¨ nchesberg 2B and Wallertheim horizons discussed above are the laminar assem- A. The only other candidates for features are blages from To¨ nchesberg 2B (Conard, 1990) possible structures made of lava and antler at and Wallertheim D (Conard & Adler, 1997). To¨ nchesberg 2B and the thus far unpublished Both of these sites provide evidence for lithic concentrations of silicified limestone manuports retooling and the curation of tools at camps to in several of the find horizons at Wallertheim. which high-utility skeletal parts of equids, Had these sites been intensely used, more fea- cervids and, to a lesser extent, bovids were tures could be expected. On balance, these data transported. Conspicuous among the lithic arti- are suggestive of relatively brief occupations facts of both sites are backed and double backed that may reflect the presence of several family blades, bladelets and points (Conard & Adler, groups or smaller segments of these Middle 1997). Several points from To¨ nchesberg 2B and Palaeolithic populations. Wallertheim D preserve damage to their tips, While not always the case, the density of which can be seen as an indication that they faunal and lithic materials at these sites is gener- were used as weapons. At Wallertheim D, burins ally low, and this may be a by-product of are also well represented. The lithic assemblages settlement in unconstrained space and small from the other sites discussed above are either areas of excavation that rarely cover more than too small to be clearly characterized, or are a few hundred square metres. These excavations dominated by tools and debitage of quartz are often too small to identify the spatial limits or other materials with irregular knapping of occupations (cf. O’Connell, 1987). As dis- properties. cussed above, some horizons such as Wallertheim A and D appear to result from Mobility and intensity of occupation occupations at a single time of the year, but even these horizons could easily reflect multiple The duration and intensity of Palaeolithic occu- periods of occupation and be taken to suggest a pations can be approximated by examining the subsistence strategy in which mobility was high abundance of faunal and lithic materials and by and occupation intensity, relatively low. The archaeological features. Although the majority Rhineland has yet to produce indications of true of species discussed above are represented by base camps reflecting several months of occupa- fewer than five individuals per sample, these tion in which dozens of prey animals and a wide remains still represent hundreds or even thou- array of archaeological features are found sands of kilograms of edible and otherwise use- (Conard, 1998). At present, the duration of ful materials. The faunal evidence suggests that occupation of most find horizons could be best

Copyright © 2000 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. 10: 286–309 (2000) Hunting Economies of the Rhineland 305 measured in hours, days and, in some cases, at To¨ nchesberg and Wallertheim were made weeks; nowhere in the Rhineland are sites in- possible through the financial support of dicative of long-term occupation known. Com- the Deutscher Akademischer Austauschdienst pelling arguments for long-term occupations (DAAD), Deutsche Forschungsgemeinschaft first appear in the Upper Palaeolithic, with the (SFB 275), the Landesamt fu¨r Boden- appearance of semi-permanent occupations at denkmalpflege in Mainz, the University of Con- the Magdalenian sites of Andernach and necticut and the Leakey Foundation. The staffs Go¨ nnersdorf. The settlement intensity and pop- of the Ro¨ misch-Germanisches Zentralmuseum ulation density in the Rhineland during the in Mainz and the Department of Early Prehis- Middle Palaeolithic was probably sparse when tory and Quaternary Ecology of the Univeristy compared with archaeologically richer regions, of Tu¨bingen provided much technical support including northern and southwestern France or for the fieldwork and laboratory analyses re- the parts of Levant during periods of favourable ported here. environmental conditions. References Conclusions Becze-Dea´k J, Langohr P. 1997. Geopedological This review of the current data from Middle study of the Wallertheim site: soil system dynam- Palaeolithic excavations in the Rhineland repre- ics approach. In Reports for the Second Wallertheim sents an attempt to synthesize the most impor- Workshop, Conard NJ, Kandel AW (eds). Institut tant data relevant to subsistence, settlement and fu¨r Ur- und Fru¨hgeschichte: Tu¨bingen; 14–20. Binford L. 1981. Bones: Ancient Men and Modern Myths. hunting economies in the region. Better tempo- Academic Press: New York. ral resolution and environmental data are still Binford L. 1984. The Faunal Remains of Klasies River needed to help identify more meaningful pat- Mouth. Academic Press: London. terns of subsistence. Many of the observations Binford L. 1985. Human ancestors: changing views and generalizations presented here can be of their behavior. Journal of Anthropological Archaeol- viewed as working hypotheses that could help ogy 4: 292–327. to enhance critical studies of Middle Palaeo- Bittmann F. 1990. Vegetationsgeschichtliche Unter- lithic subsistence and settlement. If the pace of suchungen. In Rheingeschichte zwischen Mosel und field and laboratory research in the region re- Maas, Schirmer W (ed.). Deutsche Quarta¨rvereini- mains high, the coming decades should allow gung: Hannover; 53–55. for a refinement of the ideas and hypotheses Boe¨da E, Connan J, Dessort D, Mehesen S, Mercier N, Valladas. H. 1996. Bitumen as hafting material presented here. on Middle Palaeolithic artefacts. Nature 380: 336– 338. Boenigk W, Frechen M. 1997. Stratigraphie des Acknowledgements Pleistoza¨nprofils Ariendorf am Mittelrhein. Jahrbuch des Ro¨misch-Germanishen Zentralmuseums 44: Our thanks go to G. Bosinski, S. Gaudzinski, A. 26–36. Justus, K. Kro¨ ger, J. Scha¨fer and, especially, M. Bosinski G. 1963. Eine mittelpala¨olithische Formen- Street and E. Turner for countless discussions gruppe und das Problem ihrer geochronologischen over the years. We are grateful to D. Adler, A. Einordnung. Eiszeitalter und Gegenwart 14: 124–140. Kandel, L. Niven, M. Prindiville, H.-P. Uerp- Bosinski G. 1967. Die mittelpala¨olithischen Funde im west- mann, and an anonymous reviewer for their lichen Mitteleuropa. Fundamenta A4. Bo¨ hlau Verlag: Cologne. useful comments on an earlier draft of this Bosinski G. 1982. The transition Lower/Middle paper. We also thank A. Burke for inviting us to Palaeolithic in northwest Germany. In The Transi- participate in the 1998 ICAZ symposium, and tion from Lower to Middle Palaeolithic and the Origin of particularly for her patience as we prepared this Modern Man, British Archaeological Reports Inter- paper during three field seasons far from national Series 151, Ronen A (ed.). Oxford; 165– our home institute. Excavation and analyses 175.

Bosinski G, Brunnacker K, Turner E. 1983a. Ein Gibier et Moyens de Subsistance du Paleólithique aux Siedlungsbefund des fru¨hen Mittelpalaölithikums Paleóindiens des Grandes Plaines. Colloque Interna- von Ariendorf, Kr. Neuwied. Archaölogisches Korre- tional 1995, Toulouse, Brugal J-P, David F, Enloe spondenzblatt 13(2): 157–169. JG, Jaubert J (eds). APDCA: Antibes; 301–318. Bosinski G, Kro¨ ger K, Scha¨fer J, Turner. E. 1986. Conard NJ, Adler DS. 1997. Lithic reduction and Altsteinzeitliche Siedlungspla¨tze auf den Osteifel- hominid behavior in the Middle Palaeolithic of the Vulkanen. Archaölogisches Korrespondenzblatt 15(1): Rhineland. Journal of Anthropological Research 53: 97–130. 147–176. Bosinski G, Kulemeyer J, Turner E. 1983b. Ein mit- Conard NJ, Bosinski G, Adler. D. 1995a. Koblenz- telpalaölithischer Fundplatz auf dem Vulkan Metternich. In The Palaeolithic and Mesolithic of the Hummerich bei Plaidt, Kreis Mayen-Koblenz. Ar- Rhineland. Quaternary Fieldtrips in Central Europe chaölogisches Korrespondezblatt 13: 415–428. 15, vol. 2 (14) (Series editor W. Schirmer), Bosin- Brain CK. 1981. The Hunters or the Hunted? An Introduc- ski G, Street M, Baales M (eds). INQUA-Congress tion to African Cave Taphonomy. University of Berlin: Munich; 882–885. Chicago Press: Chicago. Conard NJ, Fischer B. 2000. Are there recognizable Burke A. 1997. Incremental analysis of equid teeth cultural entities in the German Middle Paleolithic? from Wallertheim, Germany. In Reports for the Sec- In Toward Modern Humans: The Yabrudian and Mico- ond Wallertheim Workshop, Conard NJ, Kandel AW quian 400–50 K-Years Ago, Ronen A, Weinstein- (eds). Institut fu¨r Ur- und Fru¨hgeschichte: Tu¨bin- Evron M (eds). British Archaeological Reports gen; 21–22. International Series 850, Archaeopress: Oxford; Chase P. 1984. The Hunters of Combe Grenal: Approaches 7–24. to Middle Paleolithic Subsistence in Europe. British Ar- Conard NJ, Preuss J, Langohr R, Haesaerts P, van chaeological Reports International Series 286. Kolfschoten T, Becze-Dea´k J, Rebholz A. 1995b. Oxford. New geological research at the Middle Paleolithic Chase P. 1988. Scavenging and hunting in the Mid- locality of Wallertheim in Rheinhessen. Archaölo- dle Palaeolithic: the evidence from Europe. In gisches Korrespondenzblatt 25: 1–27. Upper Pleistocene Prehistory of Western Eurasia, Dibble Conard NJ, Prindiville TJ, Adler DS. 1998. Refitting H, Montet-White A (eds). University Museum: bones and stones as a means of reconstructing University of Pennsylvania, PA; 225–232. Middle Paleolithic subsistence in the Rhineland. In Chase P. 1989. How different was Middle Palaeo- Ećonomie Pre´historique: Les Comportements de Subsistance lithic subsistence? A zooarchaeological perspective au Paleólithique. XVIIIe Rencontres Internationals on the Middle to Upper Palaeolithic transition. In d’Archeólogie et d’Histoire d’Antibes. 23-25 Octo- The Human Revolution: Behavioural and Biological Per- bre, 1997, Brugal J-P, Meignen L, Patou-Mathis M spectives on the Origin of Modern Humans, Mellars P, (eds). E´ ditions APDCA: Sophia Antipolis; 273– Stringer C (eds). Princeton University Press: 290. Princeton; 321–337. Damblon F. 1997. Plant remains from the Conard NJ. 1990. Laminar lithic assemblages from Wallertheim find horizons and their significance the Middle Paleolithic from the last interglacial in for environmental reconstruction. In Reports for the northwestern Europe. Journal of Anthropological Re- Second Wallertheim Workshop, Conard NJ, Kandel search 46: 243–262. AW (eds). Institut fu¨r Ur- und Fru¨hgeschichte: Conard NJ. 1992. To¨ nchesberg and its position in Tu¨bingen; 45. the prehistory of Northern Europe. Monographien Enloe J. 1991. Subsistence organization in the Upper Ro¨ misch-Germanisch Zentralmusuem 20, Bonn. Palaeolithic: carcass refitting and food sharing at Conard NJ. 1997. Middle Paleolithic subsistence in Pincevent, PhD Dissertation. University of New the central Rhine Valley. Anthropozoologica 25, 26: Mexico. 329–336. Farizy C, David F. 1992. Subsistence and behavioral Conard NJ. 1998. Middle Palaeolithic settlement in patterns of some Middle Palaeolithic groups. In the Rhineland. In Middle Palaeolithic and Middle Stone The Middle Palaeolithic: Adaptation, Behavior and Vari- Age Settlement Systems. U.I.S.P.P-XIII Congre´s, ability, Dibble H, Mellars P (eds). University Mu- Conard NJ, Wendorf F (eds). ABACO: Forlı`; 255– seum: University of Pennsylvania, Philadelphia; 268. 87–96. Conard NJ. 1999. Bovid remains from the 1991– Farizy C, David F, Jaubert J. 1994. Hommes et Bisons du 1994 excavation in Wallertheim and their bearing Paleólithique moyen a` Mauran (Haute Garonne). CNRS on the interpretation of the locality. In Le Bison: E´ ditions: Paris.

Frechen M. 1991. Thermolumineszenz-Datierungen Neanderthals and Modern Humans in Western Asia, an Lo¨ ssen des Mittelrheingebiets. Sondervero¨f- Akazawa T, Aoki K, Bar-Yosef O (eds). Plenum fentlichungen, Geologisches Institut der Universita¨t zu Ko¨ln Press: New York; 509–521. 79: 1–137. Klein R, Cruz-Uribe K, Milo R. 1999. Skeletal part Frechen M. 1994. Thermolumineszenz-Datierungen representation in archeofaunas: comments on ‘Ex- an Lo¨ ssen des To¨ nchesberges aus der Osteifel. plaining ‘Klasies Pattern’: kua ethnoarcheology, Eiszeitalter und Gegenwart 44: 79–93. the Die Kelders Middle Stone Age archeofauna, Frechen M, Justus A. 1998. Zur Geologie der long bone fragmentation and carnivore ravaging’ Wannen-Vulkangruppe in der Osteifel. GeoArcheo- by Bartram and Marean. Journal of Archaeological Rhein 2: 213–240. Science 26: 1225–1234. Gaudzinski S. 1995a. Wisentja¨ger in Wallertheim: Kro¨ ger K. 1987. Die Fortsetzung der Arbeiten auf Zur Taphonomie einer Mittelpalaölithischen Frei- dem Plaidter Hummerich. Archaölogisches Korrespon- landstation in Rheinhessen. Jahrbuch des Ro¨misch- denzblatt 17: 13–30. Germanischen Zentralmuseums 39: 245–423. Kro¨ ger K. 1995. Plaidter Hummerich: Ein Fundplatz Gaudzinski S. 1995b. Wallertheim revisited: a re- aus der Zeit des Neandertalers im Krater eines analysis of the fauna from the Middle Palaeolithic erloschenen Vulkans, PhD Dissertation. University site of Wallertheim (Rheinhessen/Germany). Jour- of Cologne. nal of Archaeological Science 22: 51–66. Levine M. 1983. Mortality profiles and the interpre- Gaudzinski S. 1996. On bovid assemblages and their tation of horse population structure. In Hunter- consequences for the knowledge of subsistence Gatherer Economy in Prehistory: A European Perspective, patterns in the Middle Palaeolithic. Proceedings of the Bailey G (ed.). Cambridge University Press: Cam- Prehistoric Society 62: 19–39. bridge; 23–46. Jacob-Friesen KH. 1956. Eiszeitliche Elefantenja¨ger Lyman RL. 1984. Bone density and survivorship of in der Lu¨neburger Heide. Jahrbuch des Ro¨misch- fossil classes. Journal of Anthropological Archaeology 3: Germanischen Zentralmuseums 3: 1–22. 259–299. Mania D. 1997. Die Molluskenfauna von Jaubert J, Lorblanchet M, Laville H, Slott-Moller R, Wallertheim. In Reports for the Second Wallertheim Turq A, Brugal J-P. 1990. Les Chasseurs d’Aurochs de Workshop, Conard NJ, Kandel AW (eds). Institut La Borde: une Site du Paleólithique moyen (Livernon, Lot). fu¨r Ur- und Fru¨hgeschichte: Tu¨bingen; 56–68. E´ ditions de la Maison de Sciences de l’Homme: Mania D, Toepfer V. 1973. Ko¨nigsaue: Gliederung, Paris. O8kologie und mittelpalaölithische Funde der letzten Eiszeit. Johnson E. 1985. Current developments in bone Vero¨ ffentlichungen des Landesmuseums fu¨r technology. In Advances in Archaeological Method and Vorgeschichte in Halle, 43. Deutscher Verlag fu¨r Theory, vol. 8, Schiffer M (ed.). Academic Press: Wissenschaft: Berlin. London; 157–235. Marean C. 1998. A critique of the evidence for Justus A. 1992. Der Mittelpalaölithische Fundplatz 8 scavenging by Neandertals and early modern hu- ‘In den Wannen’ (Ostlicher-Wannen-Vulkan- mans: New data from Kobeh Cave (Zagros Moun- Komplex), PhD Dissertation. University of tains, Iran) and Die Kelders Cave 1 Layer 10 Cologne. (South Africa). Journal of Human Evolution 35: 111– Justus A, Urmersbach K-H. 1987. Mittelpalaölithis- 136. che Funde vom Vulkan ‘Wannen’ bei Ochtendung, Marean C, Assefa. Z. 1999. Zoological evidence for Kreis Mayen-Koblenz. Archaölogisches Korrespon- the faunal exploitation behavior of Neandertals denzblatt 17: 409–417. and early modern humans. Evolutionary Anthropology Klein R. 1976. The mammalian fauna of the Klasies 8(1): 22–37. River Mouth sites, Southern Cape Province, South Marean C, Kim SY. 1998. Mousterian large-mammal Africa. South African Archaeological Bulletin 31:75– remains from Kobeh Cave: Behavioral implications 98. for Neanderthals and early modern humans. Current Klein R. 1982. Age (mortality) profiles as a means of Anthropology 39: 79–113. distinguishing hunted species from scavenged ones Meignen L, Beyries S, Speth J, Bar-Yosef O. 1998. in archaeological sites. Palaeobiology 8(2): 151– Acquisition, traıˆtment des matie`res animales et 158. fonction du site au Paleólithique moyen dans le Klein R. 1995. Anatomy, behavior and modern hu- grotte de Kebara (Israe¨l): approche interdisi- man origins. Journal of World Prehistory 9: 176–198. plinaire. In Ećonomie Pre´historique: Les Comportements de Klein R. 1998. Why anatomially modern people did Subsistance au Paleólithique. XVIIIe Rencontres Inter- not disperse from Africa 100000 years ago. In nationals d’Archeólogie et d’Histoire d’Antibes.

23–25 Octobre, 1997, Brugal J-P, Meignen L, and Qafzeh caves, Israel. In The Human Revolution: Patou-Mathis M (eds). E´ ditions APDCA: Sophia Behavioural and Biological Perspectives on the Origins of Antipolis; 273–290. Modern Humans, Mellars P, Stringer C (eds). Prince- Mellars P. 1989. Major issues in the emergence of ton University Press: Princeton; 611–625. modern humans. Current Anthropology 30: 285–349. Shea J. 1997. Middle Palaeolithic spear point tech- Milo RG. 1998. Evidence for hominid predation at nology. In Projectile Technology, Knecht H (ed.). Klasies River Mouth, South Africa, and its implica- Plenum Press: New York; 79–106. tions for the behaviour of early modern humans. Shipman P. 1981. The Life History of a Fossil: An Journal of Archaeological Science 25: 99–133. Introduction to Taphonomy and Paleoecology. Harvard Mussi M. 1999. The Neanderthals in Italy: a tale of University Press, Harvard: Cambridge. many caves. In The Middle Palaeolithic Occupation of Shipman P, Foster G, Schoeninger M. 1984. Burnt Europe, Roebroeks W, Gamble C (eds). University bones and teeth: an experimental study of color, of Leiden Press: Leiden; 49–80. morphology, crystal structure and shrinkage. Jour- Oakley KP, Andrews P, Keeley LH, Clark. JD. 1977. nal of Archaeological Science 11: 307–325. Stiner M. 1990. The use of mortality patterns in A reprisal of the Clacton spearpoint. Proceedings of archaeological studies of hominid predatory adap- the Prehistoric Society 43: 13–30. tations. Journal of Anthropological Archaeology 9: 305– O’Connell J. 1987. Alyawara site structure and its 351. archaeological implications. American Antiquity 52: Stiner M. 1991. Food procurement and transport by 74–108. human and non-human predators. Journal of O’Connell J, Hawkes K, Blurton Jones. N. 1988. Archaeological Science 18: 455–482. Hazda scavenging: implications for Plio-Pleis- Stiner M. 1994. Honor Among Thieves: A Zooarchaeologi- tocene hominid subsistence. Current Anthropology cal Perspective on Neandertal Ecology. Princeton Uni- 29(2): 356–363. versity Press: Princeton. Pike-Tay A. 1997. Incremental analysis of bovid and Street M. 1995. Plaidter Hummerich. In The Palaeo- cervid teeth from Wallertheim, Germany. In Re- lithic and Mesolithic of the Rhineland. Quaternary Field- ports for the Second Wallertheim Workshop, Conard NJ, trips in Central Europe 15, vol. 2 (14) (Series editor Kandel AW (eds). Institut fu¨r Ur- und Fru¨hges- W. Schirmer), Bosinski G, Street M, Baales M chichte: Tu¨bingen; 69. (eds). INQUA-Congress Berlin: Munich; 898– Plisson H, Beyries S. 1998. Pointes ou outils triangu- 901. laires? Donneés fonctionnelles dan le Mouste´rien Terberger K. 1993. Das Lahntal-Palaölithikum. Ma- levantin. Paleórient 24: 5–24. terialien zur Vor- und Fru¨hgeschichte von Hessen. Prindiville TJ. 1998. Taphonomie und Archaölogie Selbstverlag des Landesdenkmalamtes fu¨r Denk- der Pferdereste aus Wallertheim, Fundschicht F malpflege. Wiesbaden. (Grabung Conard 1991–1994), MA Thesis. Uni- Thieme H. 1983. Der Palaölithische Fundplatz versity of Tu¨bingen. Rheindahlen, PhD Dissertation. University of Rademacher E. 1911. Die Fauna des Kartsteins. Cologne. Pra¨historische Zeitschrift 3: 225–229. Thieme H. 1995. Ein altpalaölithischer Lagerplatz Richter J. 1997. Sesselfelsgrotte III: der G-Schichten- aus der Zeit des Urmenschen von Scho¨ ningen 13 Komplex der Sesselfelsgrotte: zum Versta¨ndnis des Mico- II (Reinsdorf-Interglazial). In Archaölogische Aus- quien. Saarbru¨cker Druckerei und Verlag: Saar- grabungen im Braunkohlentagebau Scho¨ningen, Landkreis bru¨cken. Helmstedt, Thieme H, Maier R (eds). Verlag Hahn’- Scha¨fer J. 1987. Der altsteinzeitliche Fundplatz sche Buchhandlung: Hannover; 95–107. Schweinskopf am Karmelenberg, Gem. Bassen- Thieme H. 1997. Lower Palaeolithic hunting spears Nature heim, Kreis Mayen-Koblenz: Vorbericht. Archaölo- from Germany. 385: 807–810. Thieme H, Mania D, Urban B, von Kolfschoten T. gisches Korrespondenzblatt 17: 1–12. 1993. Scho¨ ningen (Nordharzvorland): eine alt- Scha¨fer J. 1990. Der altsteinzeitliche Fundplatz auf palaölithische Fundstelle aus dem mittleren Eiszei- dem Vulkan Schweinskopf-Karmelenberg, PhD talter. Archaölogisches Korrespondenzblatt 23: Dissertation. University of Cologne. 147–163. Schmidtgen O, Wagner W. 1929. Eine altpalaölithis- Thieme H, Veil S. 1985. Neue Untersuchungen zum che Fundstelle bei Wallertheim in Rheinhessen. eemzeitlichen Elefanten-Jadgplatz Lehringen, Notizblatt des Vereins fu¨r Erdkunde und der Hessischen Ldkr. Verden. Die Kunde 36: 11–58. Landesanstalt zu Darmstadt 11: 3–41. Thissen J. 1986. Ein weiterer Fundplatz der West- Shea J. 1989. A functional study of the lithic indus- wandfundschicht (B1) von Rheindahlen. Archaölo- tries associated with hominid fossils in the Kebara gisches Korrespondenzblatt 16: 111–121. Copyright © 2000 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. 10: 286–309 (2000) Hunting Economies of the Rhineland 309

Turner E. 1990. Middle and Late Pleistocene macro- In DEUQUA-Fu¨hrer 1, Schirmer W (ed.). Deutsche faunas of the Neuwied Basin region (Rhineland– Quarta¨rvereinigung: Hannover; 166–190. Palatine) of West Germany. Jahrbuch des Ro¨misch- Van den Brink FH. 1968. Saügertiere Europas. Verlag Germanischen Zentralmuseum Mainz 37: 135–403. Paul Parey: Hamburg. Turner E. 1997. Large vertebrates from the Middle Van Kolfschoten T. 1990. Die Kleinsaüger von Palaeolithic levels at Wallertheim, Rheinhessen. In To¨ nchesberg 2. In Rheingeschichte zwischen Mosel und Reports for the Second Wallertheim Workshop, Conard Maas, Schirmer W (ed.). Deutsche Quartaërverein- NJ, Kandel AW (eds). Institut fu¨r Ur- und gung: Hannover; 52. Fru¨hgeschichte: Tu¨bingen; 87–93. Van Kolfschoten T, Thomassen H. 1997. The Turner E. 1998. Ariendorf: quaternary deposits and smaller mammal fauna from Wallertheim in a Palaeolithic excavations in the Karl Schneider larger Central European context. In Reports for the gravel pit. Jahrbuch des Ro¨misch-Germanischen Zentral- Second Wallertheim Workshop, Conard NJ, Kandel museums Mainz 44: 3–191. AW (eds). Institut fu¨r Ur- und Fru¨hgeschichte: Uerpmann H-P, Dechert B. 1997. Die Pferdereste Tu¨bingen; 100. aus Wallertheim im Vergleich mit anderen Fund- Weinstock-Arenovitz J. 1997. Late Pleistocene rein- stellen: Anregungen fu¨r das Versta¨ndnis der deer populations in Western and Central Europe: Gro¨ ssenvera¨nderugen der Equiden im Pleistoza¨n. an osteometrical study of Rangifer tarandus, PhD In Reports for the Second Wallertheim Workshop, Conard Dissertation. University of Tu¨bingen. NJ, Kandel AW (eds). Institut fu¨r Ur- und Zo¨ ller L, Conard NJ, Hahn. J. 1991. Thermolu- Fru¨hgeschichte: Tu¨bingen; 94–96. minescence dating of Middle Palaeolithic open air Van den Bogaard P, Hall CM, Schmincke H-U, York sites in the middle Rhine Valley/Germany. Natur- D. 1989. Precise single-grain 40Ar/39Ar dating of wissenschaften 78: 408–410. a cold to warm climate transition in Central Eu- Zo¨ ller L. 1997. Thermoluminscence dating of loess rope. Nature 342: 523–525. and loess-derivatives from the Wallertheim profile. Van den Bogaard P, Schminke H-U. 1990. Die In Reports for the Second Wallertheim Workshop, Conard Entwicklungsgeschichte des Mittelrheinraumes und NJ, Kandel AW (eds). Institut fu¨r Ur- und die Eruptionsgeschichte des Osteifel-Vulkanfeldes. Fru¨hgeschichte: Tu¨bingen; 101.