The Lower Paleolithic sites of Schöningen (Germany) have played a major role in discussions of the origin of active and specialized hunting of large mammals. The presence of a kit of complete wooden hunting weapons (nine throwing spears, a thrusting lance, and two throwing sticks; Thieme, 2007; Conard et al., 2020), along with the Schöningen mammalian skeletal assemblage, have yielded clear indications of the exploitation of large herbivores for meat and marrow, bone for tool production, and potentially skins (Voormolen, 2008; Julien et al., 2015; van Kolfschoten et al., 2015a, van Kolfschoten et al., 2015b; Starkovich and Conard, 2015; Turner et al., 2017; Hutson et al., 2018). In the present study, we describe new evidence for the exploitation of bears from Schöningen and compare this evidence to other published faunal assemblages with evidence of bear exploitation from open-air archaeological sites in Northwestern Europe dating to the Lower and Middle Paleolithic. We compare Schöningen only to open-air sites in this study, as cave sites are known to primarily yield bear skeletal remains from natural deaths during hibernation. We discuss the mode of exploitation and compare the sites based on three proxies that are regularly used as supporting evidence for the exploitation of bears: the relative occurrence of bears within an archaeological assemblage, the mortality profile of the population, and the presence of the anthropogenic marks that indicate their exploitation. The exploitation of bears, especially cave bears, has been an ongoing debate for over a century and is relevant not only in the context of hominin diets but also for the use of skins. Tracing the origins of hide exploitation can contribute to the understanding of survival strategies in the cold and harsh conditions of Northwestern Europe during the Middle Pleistocene.

During the Palaeolithic occupation of Europe, bears were an omnipresent part of the Pleistocene fauna, both during warm and cold conditions, either with members of the cave bear or the brown bear lineage (Stiner, 1999). As omnivores and frequent cave dwellers, they shared their environments with humans, and interactions between humans and bears would likely have occurred. The nature of these interactions has been the subject of scientific debate for at least a century, with paradigms having shifted multiple times. In the first half of the 20th century, the discussion of bear hunting during the Paleolithic revolved around the ‘cave bear cult’ theory initiated by Bächler (1921, 1923, 1940). The accumulation of large masses of cave bear remains, sometimes found in peculiar constellations in cave sites where stone artifacts were also discovered, led to the interpretation that cave bears were actively hunted and worshiped as part of a cult. Even at the time, the theory was already contested and the accumulations were explained as a taphonomic phenomenon (Soergel, 1940), leading some authors to be skeptical of cave bear hunting under any circumstances (e.g., Koby, 1953). Over the past three decades, substantial new evidence of bear hunting during the Paleolithic was found and research on the relationships between bears and humans was reinitiated (Pacher, 1997, 2000; Tillet and Binford, 2002).

The clearest examples of cave- and brown-bear exploitation have been discovered from Upper Paleolithic sites, and among them are some true ‘smoking guns’ of bear hunting. From an Early Gravettian layer at Hohle Fels in Germany, dated to 29,000 BP, archaeologists found a thoracic vertebra with an imbedded flint fragment (Conard et al., 2001; Münzel and Conard, 2004). A large number of bear remains from the same horizon displays butchery marks, evidence of marrow fracturing, and burning, documenting the full exploitation of bears by humans for their meat, skins, marrow, and bones (Münzel and Conard, 2004; Kitagawa et al., 2012). Wojtal et al. (2015) describe additional evidence at open-air Gravettian sites from Poland and the Czech Republic with cutmarked bones from both cave bears and brown bears. Although the remains of bears at Gravettian open-air sites are always recovered in low numbers, the anthropogenic marks show that bear hunting in this time was a common phenomenon. In the late Upper Paleolithic site of Grotte du Bichon, Switzerland, a female brown bear skeleton was found with an imbedded flint fragment in the third cervical vertebra. It was discovered together with a skeleton of a man hypothesized to have died from a hunting accident (Morel, 1993; Chauvière, 2008). Damage patterns and imbedded flint particles on a brown bear rib from the Late Epigravettian site of the Cornafessa rock shelter in Italy have been experimentally shown to result from bow-and-arrow hunting (Duches et al., 2018).

Numerous cave sites show clear evidence of bear exploitation, even during the Middle Paleolithic. Romandini et al. (2018) published a comprehensive overview of Middle Paleolithic sites with evidence of cave bear exploitation, adding detailed taphonomic descriptions of new material from Rio Secco Cave and Fumane Cave (Italy). Their study highlights that Neanderthals not only exploited bears for their skins but also used them as a food source, demonstrated by filleting marks, burned bones, and potential human tooth marks. At the well-known Neanderthal site of Krapina (Croatia), at least a part of the adult-dominated bear assemblage shows cutmarks indicating human exploitation of cave bears (Miracle, 2007, 2008). The Eemian site of Lehringen (Germany), famous for the discovery of both a yew spear associated with a straight-tusked elephant skeleton and stone tools (Adam, 1951; Thieme and Veil, 1985), yielded a distal femur fragment of a brown bear with cutmarks (Wenzel, 1998). Exceptional amounts of bear remains with clear anthropogenic modifications (n = 2496) were discovered at the open-air site of Biache-Saint-Vaast (France, MIS 7; Auguste, 1988, 1995, 2003). A reappraisal of the faunal material excavated in the 19th and early 20th century from Taubach (Germany, MIS 5) revealed cutmarks on a large number of bear bones, which Bratlund (1999) interpreted as the result of active hunting.

During the Middle Paleolithic, bear bones were also used as a raw material for tool production: in Scladina Cave, six long bone fragments were used as retouchers, of which four could be refitted as part of a femur shaft of a cave bear (Abrams et al., 2014). Similarly, in Biache-Saint-Vaast (Auguste, 2002, 2003), as well as in Fumane Cave and Rio Secco Cave (Romandini et al., 2018), archaeologists discovered bone retouchers made from ursid bones.

Evidence for bear exploitation in the Lower Paleolithic is much rarer: at the approximately 600 ka site of Isernia La Pineta (Molise, Italy), cutmarks related to skinning and scraping marks have been found on skeletal elements of bears (Thun Hohenstein et al., 2005). The earliest documented evidence in Northwestern Europe is an approximately 500 ka, cutmarked zygomatic arch from Boxgrove (UK) (Parfitt, 1999; Smith, 2012). A recent taphonomic study (Brasser, 2017, 2020) from a sample of the skeletal material from Bilzingsleben (Germany, ca. 400 ka) described several cutmarked bear metatarsals and skull fragments. Three associated metatarsals from Grays Thurrock (UK; MIS 9) show cutmarks on the dorsal side (Schreve, 1997; Schreve and Currant, 2003), but no detailed taphonomic description of the rather substantial bear assemblage (n = 104) has yet been published. The relationship between the artifacts found in the quarries at Grays Thurrock and the bear remains with cutmarks is unclear, as they originate from old excavations and stratigraphic and spatial data of the finds are scarce. Given the assigned age of the bone assemblage to MIS 9, it is unclear if they were associated with Lower or Middle Paleolithic industries. The new material from Schöningen described in this study is therefore a noteworthy addition to the scarce evidence of early bear exploitation during the Lower and early Middle Paleolithic.

The Lower Paleolithic open-air site complex of Schöningen (Lower Saxony, Germany; Fig. 1A) is world-famous for the well-preserved organic archaeological finds including a series of wooden spears from the late Middle Pleistocene, dating to around 320–300 ka (Thieme, 1997; Conard et al., 2015; Serangeli et al., 2015a). The large mammal skeletal assemblages from Schöningen, currently comprising over 20,000 specimens, have yielded a wealth of data on the exploitation of large mammals by late Middle Pleistocene hominins (Voormolen, 2008; van Kolfschoten, 2014; Julien et al., 2015; Starkovich and Conard, 2015; van Kolfschoten et al., 2015a, van Kolfschoten et al., 2015b; Turner et al., 2017; Hutson et al., 2018). Among this large assemblage, a small subsample (<1%) consists of finds from species that belong to Carnivora (e.g., Serangeli et al., 2015b; Hutson et al., 2021; Verheijen et al., 2022). Most carnivore species are represented by only a few skeletal elements at Schöningen; however, bear remains from the lowermost levels of sites of Schöningen (Schö) 12 II-1 and Schö 12 B are relatively numerous (n = 34). Among these bear remains are two cutmarked specimens that are the focus of this article.

During large-scale mining operations east of the town of Schöningen, archaeological and geological monitoring revealed a long sequence of Quaternary deposits covering three glacial cycles (Elsterian, Saalian, and Weichselian glaciations) and four intercalated interglacial cycles (Holstein, Reinsdorf, Schöningen, and Eemian interglacial). In 1992, the first Lower Paleolithic archaeological and faunal remains were discovered in the quarry sections within Middle Pleistocene lacustrine sediments. The first rescue excavations at the sites of Schö 12 A, Schö 12 B, and Schö 12 C yielded a rich Middle Pleistocene mammalian fauna, associated with Lower Paleolithic flint tools and several potential wooden hafting tools (Thieme et al., 1993; Thieme and Maier, 1995; Thieme, 2007). At the site of Schö 12 B, the largest archaeological assemblage was recovered from a single find horizon (findlayer 1 or Fundschicht [FS] 1). Rescue excavations of a 150 m2 surface yielded approximately 1000 finds (Serangeli et al., 2015a), of which 678 belong to large mammals (Voormolen, 1997). Although some aspects of the faunal assemblage from site 12 B, FS 1 have been published (Thieme et al., 1993; van Kolfschoten, 1993; 1995; van Zijderveld and Kirkels, 1996), a thorough taphonomic analysis of the entire assemblage was only reported in an unpublished doctoral essay by Voormolen (1997). Approximately 80% of the material originates from the 150 m2 excavation area, whereas about 20% of the bones were collected from the spoil heaps belonging to the FS 1 that were already removed by the mining company. Owing to time pressure during the rescue excavations, contextual information such as vertical distribution of the finds and their stratigraphic context was only partially recorded (Thieme et al., 1993). The assemblage contains a rich interglacial large mammal fauna, including several bear remains previously assigned to Ursus spelaeus and Ursus thibetanus (van Kolfschoten, 1995; van Zijderveld and Kirkels 1996).

In 2008 and 2009, rescue excavations at the site of Schö 12 II were executed during the mining of a ten-hectare area within the Schöningen lignite quarry (DB-Pfeiler), formerly separating the quarry in a northern and a southern part (Conard et al., 2015; Serangeli et al., 2015a). The excavations were done in a series of plateaus (0–9), sequentially numbered toward the north in the order of their excavation (Fig. 1B). The majority of the faunal remains from the 2008–2009 excavations at Schöningen site 12 II were published by Julien et al. (2015). The analysis focused on the occurrence of potential bone tools from site 12 II-4 and their geoarchaeological context. The study highlighted the clearly archaeological nature of the assemblage, with the occurrence of bone retouchers, a horse metapodial used as a soft hammer, and several bones and an elephant tusk fragment with worn/polished ends that may relate to hominin use (Julien et al., 2015). In addition, the assemblage contained faunal remains with hominin butchery marks (cutmarks, marrow-fractured long bones; Julien et al., 2015), found together with flint artifacts (Serangeli et al., 2015a).

The stratigraphy at Schö 12 II incorporates five sedimentary cycles, interpreted as lake-level shallowing cycles. These shallowing cycles, characterized by calcareous marl grading up to organic mud deposits (Stahlschmidt et al., 2015) are correlated with the well-known archaeological site complex of Schö 13 II, located approximately 800 m from the site of Schö 12 II (Serangeli et al., 2015a). Schö 12 B and plateau 0–1 of the site of Schö 12 II are geographically adjacent (Fig. 1B). Schö 12 II-1 and Schö 12 B, FS 1 are both situated at the base of sedimentary channel II (Mania, 1995). Therefore, we interpret the faunal assemblages found in both sites as being broadly contemporaneous. This base is ascribed to the interglacial optimum of the Reinsdorf Interglacial based on the pollen assemblages that are indicative of a forested area around the lake dominated by Alnus (alder; Urban, 1995; Urban and Bigga, 2015) and faunal species corresponding to temperate, wooded environments (Thieme et al., 1993; van Kolfschoten, 1993; van Zijderveld and Kirkels, 1996). Thermoluminescence (TL) dating of heated flint from the underlying channel I deposits has established a maximum date of the channel II deposits of approximately 321 ± 16 ka (Richter and Krbetschek, 2015). Direct Uranium-series (U-series) dating of the peat deposits from channel II provided a date of approximately 300 ka (Sierralta et al., 2012). Based on relative dating (biostratigraphy) combined with the U-series dating (Sierralta et al., 2012) and TL-dating (Richter and Thieme, 2012; Richter and Krbetschek, 2015), the channel II Reinsdorf Interglacial deposits in Schöningen are currently correlated with MIS 9 (Richter and Krbetschek, 2015; Urban and Bigga, 2015).