INTRODUCTION
The cave bear was a prominent member of the Late Pleistocene megafauna in Europe and western Asia and became extinct around 28–25 cal ka BP (Pacher and Stuart Reference Pacher and Stuart2009; Baca et al. Reference Baca, Popović, Stefaniak, Marciszak, Urbanowski, Nadachowski and Mackiewicz2016). The majority of remains were found in caves where this mammal probably died during winter hibernation. Dozens of caves containing bones and teeth of cave bear are known from the Alps from the valleys up to the high mountains. Genetic research has shown that four different subspecies (or species) were present in this mountain range (Hofreiter et al. Reference Hofreiter, Capelli, Krings, Waits, Conard, Münzel, Rabeder, Nagel, Paunovic, Jambresic, Meyer, Weiss and Pääbo2002, Reference Hofreiter, Rabeder, Jaenicke-Després, Withalm, Nagel, Paunovic, Jambresic and Pääbo2004). Ursus ladinicus (Rabeder et al. Reference Rabeder, Hofreiter, Nagel and Withalm2004) was one of them and its small body size, slender extremities, and distinct dentition (especially the higher evolution level of lower m2) have been attributed to adaptations to a life upon the mountains (Rabeder et al. Reference Rabeder, Hofreiter, Nagel and Withalm2004, Reference Rabeder, Debeljak, Hofreiter and Withalm2008).
The chronology of cave bears is almost exclusively based on radiocarbon (14C) dates of bones produced over several decades by a variety of laboratories. Pacher and Stuart (Reference Pacher and Stuart2009) reassessed the available data and concluded that cave bears were present in the Alps and its foreland during Marine Isotope Stage (MIS) 3—also known as Middle Würmian in the Alpine Quaternary stratigraphy (Preusser Reference Preusser2004; Heiri et al. Reference Heiri, Koinig, Spötl, Barrett, Brauer, Drescher-Schneider, Gaar, Ivy-Ochs, Kerschner, Luetscher, Moran, Nicolussi, Preusser, Schmidt, Schoeneich, Schwörer, Sprafke, Terhorst and Tinner2014)—but several dates were close to or older than the limit of 14C dating. Very little chronological information is currently available prior to about 50 ka because speleothems (which could be dated using U/Th far beyond 50 ka) are rarely interlayered with clastic sediments containing cave bear remains. In the past, U/Th dating of bone material from alpine caves was attempted using alpha-spectrometry (Leitner-Wild and Steffan Reference Leitner-Wild and Steffan1993; Leitner-Wild et al. Reference Leitner-Wild, Rabeder and Steffan1994), yielded large uncertainties and is regarded as problematic as bones are known to be open systems to U (e.g., Pike et al. Reference Pike, Hedges and van Calsteren2002; Sambridge et al. Reference Sambridge, Grün and Eggins2012).
Several of the caves occupied by the cave bear in the Alps during the last glacial period are located close to the modern (natural) treeline. The complex internal stratigraphy of the cave sediments and the precision of the 14C dates do not allow to conclusively answer the question whether the cave bear used individual alpine sites continuously throughout MIS 3 (an interval of large climate shifts) or only during some intervals, e.g. major interstadials. Starting about 40 ka ago, the data indicate a gradual disappearance of this animal from these alpine sites and about 5 ka later also from lower-elevation caves (Rabeder and Frischauf Reference Rabeder and Frischauf2016), suggesting a general climate control.
A special case in this context is Conturines Cave in the Dolomites of northern Italy, which is the highest site used by cave bears in the Alps, topping other alpine caves by several hundred meters. It ranks as the world’s highest-elevation cave containing remains of the cave bear. Located in a barren landscape today, the presence of this herbivore cave bear population raised serious questions about the paleovegetation and paleoclimate as initial 14C dates suggested an MIS 3 age of this fauna (Rabeder Reference Rabeder1991; Hofreiter et al. Reference Hofreiter, Rabeder, Jaenicke-Després, Withalm, Nagel, Paunovic, Jambresic and Pääbo2004). Here, we report new dates obtained by two different laboratories, which indicate that these remains are in fact older than the 14C dating limit and that they therefore have no implications for the paleoenvironment and paleoclimate during MIS 3.
SETTING
Conturines Cave (Ander dles Cunturines in the local Ladin language) is located 4 km east of St. Kassian (San Ciascian, San Cassiano) in the Dolomites of northern Italy. The entrance opens at an elevation of 2775 m asl at the upper end of an east-facing cirque (Figure 1). The cave is about 160 m long and comprises a single ascending conduit, on average about 8 m in diameter. Breakdown blocks dominate the character of this cave and a bird guano deposit is present behind the entrance. The inner part also contains large dripstone formations including a several-meter-thick flowstone. These carbonate deposits are inactive and show strong signs of corrosion. No active speleothem deposition occurs in the cave reflecting the high elevation and the lack of a soil-covered catchment of the cave’s drip water.
(A) Oblique view towards northwest of the Conturines massif and location of the cave entrance (Google Earth Pro image); (B) site map; (C) simplified map of Conturines Cave and the Skull Chamber where the cave bear bones were excavated.
Remains of Ursus ladinicus were found at the end of the ascending passage some 110 m behind the entrance (Figure 1). In this so-called Schädelhalle (Skull Chamber) up to 0.5 m of unconsolidated sediments rest on a thick flowstone. A thin and discontinuous layer of unfossiliferous dolomitic sand forms the base of this sediment package. This sand is unconformably overlain by the main fossil-bearing layer, consisting of yellowish-grey dolomite sand and numerous angular dolomite blocks. Intact and fragmented bones as well as teeth were dispersed in this layer. These remains have been redeposited by running water and are mixed with the coarse clastic sediment (Rabeder Reference Rabeder1991). Later on, small cave streams partly eroded this fossiliferous layer, exhumed some of the bones and skulls, and sand devoid of fossil remains was locally deposited in these channels.
In addition to cave bear bones and teeth (showing a high percentage of juvenile individuals), two bone fragments (a juvenile mandible and a maxillary fragment from the same individual) of a cave lion were found (Rabeder Reference Rabeder1991).
The fact that the Skull Chamber lies about 58 m higher than the entrance results in a slight thermal anomaly of the interior of the cave of about +1.5°C instead of about –1.0°C as expected from the mean annual air temperature outside the cave at this altitude. The cave has undergone some minor changes since the time the cave bear used this site for hibernation (including partial collapse of the ceiling in the hall just behind the Skull Chamber). However, the basic geometry of the cave—the lack of an upper entrance and hence the positive thermal anomaly in the interior—very likely has not changed since then and apparently allowed this animal to survive the harsh winter in the high mountains.
SAMPLES AND METHODS
Twenty-five randomly selected small cave bear bones (metapodials and phalanges) and one pelvis fragment (Table 1) excavated from the coarse-grained fossiliferous layer in the Skull Chamber were prescreened for their whole bone nitrogen content. Collagen extraction was performed on 11 samples showing values between 3.3 and 4.1% N. Collagen was extracted from cleaned, crushed bone samples with an acid-base-acid treatment followed by gelatinization and ultrafiltration (Brock et al. Reference Brock, Higham, Ditchfield and Ramsey2010) using a Vivaspin® filter cleaning method introduced by Bronk Ramsey et al. (Reference Bronk Ramsey, Higham, Bowles and Hedges2004a). The collagen was then freeze-dried. The dried samples were weighed into precombusted quartz tubes with an excess of copper oxide (CuO), sealed under vacuum and combusted to carbon dioxide (CO2). The CO2 was converted to graphite on an iron catalyst using the zinc reduction method (Slota et al. Reference Slota, Jull, Linick and Toolin1987). The 14C/12C and 13C/12C ratios were measured by accelerator mass spectrometry (AMS) at 14CHRONO, Queen’s University Belfast. The sample 14C/12C ratio was background-corrected using measurements on collagen extracted from the Latton mammoth bone (Lewis et al. Reference Lewis, Maddy, Buckingham, Coope, Field, Keen, Pike, Roe, Scaife and Scott2006) and normalized to the HOXII standard (SRM 4990C; National Institute of Standards and Technology). The 14C/12C ratios were corrected for isotope fractionation using the AMS measured δ13C which accounts for both natural and machine fractionation. The 14C age and one standard deviation were calculated using the Libby half-life of 5568 yr following the methods of Stuiver and Polach (Reference Stuiver and Polach1977). Stable isotopes (δ13C and δ15N), % carbon and % nitrogen were measured on a Delta V Advantage with a Flash elemental analyzer. Stable isotope standards IA-R041 L-Alanine, IAEA-N-2 Ammonium Sulphate and IAEA-CH-Sucrose were analyzed with the unknown samples to provide a two-point calibration. An internal fishbone collagen sample was also analyzed). Reproducibility for ultrafiltered bone collagen is 0.22‰ for δ13C and 0.15‰ for δ15N.
N content of whole bones and atomic C/N ratio, radiocarbon and stable C and N isotope data of extracted collagen from Conturines Cave bear bones.
Aliquots of five of these bone samples were also analyzed in the Oxford Radiocarbon Accelerator Unit Research Laboratory for Archaeology using the ultrafiltration method outlined in Brock et al. (Reference Brock, Higham, Ditchfield and Ramsey2010), and the ultra-filter cleaning method introduced by Bronk Ramsey et al. (Reference Bronk Ramsey, Higham, Bowles and Hedges2004a). The samples were combusted in a CHN analyzer yielding C/N ratios in the range 3.13–3.20. As part of the pretreatment process, carbon isotope measurements were measured on an Sercon mass spectrometer against an alanine standard with a precision of±0.3‰ relative to VPDB. The samples were graphitized using the method outlined in Dee and Bronk Ramsey (Reference Dee and Bronk Ramsey2000) and measured in an HVEE Tandetron AMS as described in Bronk Ramsey et al. (Reference Bronk Ramsey, Higham and Leach2004b). In this study, all of the measurements made at Oxford were indistinguishable from background.
RESULTS
The 25 prescreened bone samples showed N-contents between 2.0 and 4.1% N (one sample yielded 1.2%), suggesting good preservation of collagen. This was confirmed by measurements of the atomic C/N ratio on the extracted collagen (3.1–3.2; Table 1), which falls within the acceptable range (2.9–3.6) for the bones (DeNiro Reference DeNiro1985; van Klinken Reference van Klinken1999).
Radiocarbon dates obtained on 11 samples with the highest N values yielded infinite ages in the Belfast laboratory. This was confirmed by analyses carried out in the Oxford laboratory (Table 1). δ13C and δ15N values range from –23.2 to –21.5‰ and from –0.6 to +2.9‰ (Table 1).
DISCUSSION
The new 14C data consistently show that cave bear hibernated and died in the interior of Conturines Cave prior to ca. 50 ka. The new data are regarded as reliable given the good state of collagen preservation (facilitated by the cold environment) and the independent confirmation by two laboratories. Random sampling for dating is regarded as a reasonable approach in such a setting which lacks a well-stratified sedimentary succession and suggests local reworking of the bones.
Hofreiter et al. (Reference Hofreiter, Capelli, Krings, Waits, Conard, Münzel, Rabeder, Nagel, Paunovic, Jambresic, Meyer, Weiss and Pääbo2002, Reference Hofreiter, Rabeder, Jaenicke-Després, Withalm, Nagel, Paunovic, Jambresic and Pääbo2004) reported a single AMS date of 44.3±0.9 ka (determined by Beta Analytics) from Conturines Cave. Unfortunately, no analytical details were published. Given the progress in preparation and analysis of old bone material in the past two decades this finite date should not be regarded as a contradiction to the new data set. In their review of 14C dating of Pleistocene bone material van der Plicht and Palstra (Reference van der Plicht and Palstra2016:250) stress that “fossil bones with reported ages older than 45,000 BP must be considered with great care.” The new dates are consistent with earlier attempts to constrain the age of the Conturines samples using U-series dating. These dates (87±5 ka BP and 108+8/–7 ka BP), however, were only briefly mentioned but never properly published (Rosendahl et al. Reference Rosendahl, Döppes and Kempe2007; Döppes and Rosendahl Reference Döppes and Rosendahl2009) and it is difficult to judge their validity given that bones are known to readily exchange U with the surrounding environment (e.g., Pike et al. Reference Pike, Hedges and van Calsteren2002, Reference Pike, Eggins, Grün, Hedges and Jocobi2005).
The infinite 14C dates are minimum age estimates and it is currently not possible to convert them into minimum calendar ages.
The former presence of cave bears at this altitude is surprising given the almost complete lack of vegetation near the cave site. The modern treeline is located at about 2000–2100 m asl in the Dolomites, i.e. about 700–800 m below the cave (Peer Reference Peer1980). No other high-elevation cave bear sites are currently known from the Dolomites and it is instructive to take a look at the large database obtained from caves in the Northern Calcareous Alps of Austria. Chronological data from this mountain range—biased by the 14C method (i.e. to less than about 50 ka)—indicate that cave bears used these mountain caves during MIS 3 (e.g., Hofreiter et al. Reference Hofreiter, Rabeder, Jaenicke-Després, Withalm, Nagel, Paunovic, Jambresic and Pääbo2004; Döppes et al. Reference Döppes, Stiller and Rabeder2011; Rabeder and Frischauf Reference Rabeder and Frischauf2016). These sites are on average 500–1400 m lower in elevation and partly close to or below the modern treeline. These bear remains therefore indicate climatically benign and certainly ice-free conditions during MIS 3 in Europe, a time interval characterized by high-frequency, high-amplitude shifts in temperature (e.g., Genty et al. Reference Genty, Blamart, Ouahdi, Gilmour, Baker, Jouzel and Van-Exter2003; Wohlfart et al. Reference Wohlfarth, Veres, Ampel, Lacourse, Blaauw, Preusser, Andrieu-Ponel, Kéravis, Lallier-Vergès, Björck, Davies, de Beaulieu, Risberg, Hormes, Kasper, Possnert, Reille, Thouveny and Zander2008; Heiri et al. Reference Heiri, Koinig, Spötl, Barrett, Brauer, Drescher-Schneider, Gaar, Ivy-Ochs, Kerschner, Luetscher, Moran, Nicolussi, Preusser, Schmidt, Schoeneich, Schwörer, Sprafke, Terhorst and Tinner2014; Fankhauser et al. Reference Fankhauser, McDermott and Fleitmann2016). Closed forests had disappeared from the Alps and their northern foreland already at the end of MIS 5a (Drescher-Schneider Reference Drescher-Schneider2000; Müller et al. Reference Müller, Pross and Bibus2003). Available proxy records from the Alps covering parts of MIS 3 indicate an Arctic tundra vegetation even in the lowlands (Bortenschlager and Bortenschlager Reference Bortenschlager and Bortenschlager1978) interrupted by intervals of milder conditions during some larger interstadials (Jost-Stauffer et al. Reference Jost-Stauffer, Coope and Schlüchter2005; Drescher-Schneider et al. Reference Drescher-Schneider, Jacquat and Schoch2007; Tütken et al. Reference Tütken, Furrer and Vennemann2007; Starnberger et al. Reference Starnberger, Drescher-Schneider, Reitner, Rodnight, Reimer and Spötl2013). Even during the pronounced Greenland Interstadial 14, whose maximum lasted from 54.2 to 51.7 ka, glaciers in the Eastern Alps were most likely larger than during the maximum of the Little Ice Age, as shown by speleothem data (Spötl et al. Reference Spötl, Mangini and Richards2006). During intervening stadials (including Heinrich events) temperatures dropped drastically and loess sections in the foreland close to the Eastern Alps show evidence of permafrost (Terhorst et al. Reference Terhorst, Thiel, Peticzka, Sprafke, Frechen, Fladerer, Roetzel and Neugebauer-Maresch2011; Thiel et al. Reference Thiel, Terhorst, Jaburová, Baylaert, Murray, Fladerer, Damm, Frechen and Ottner2011; Nigst et al. Reference Nigst, Haesaerts, Damblon, Frank-Fellner, Mallol, Viola, Götzinger, Niven, Trnka and Hublin2014). The level of precision of the available 14C data from the Northern Calcareous Alps does not permit to unequivocally relate alpine cave bear data to the climate history e.g. known from Greenland ice cores and alpine speleothems (e.g., Svensson et al. Reference Svensson, Andersen, Bigler, Clausen, Dahl-Jensen, Davies, Johnsen, Muscheler, Parrenin, Rasmussen, Röthlisberger, Seierstad, Steffensen and Vinther2008; Moseley et al. Reference Moseley, Spötl, Cheng, Svennson, Brandstätter and Edwards2014). It is likely, however, that (high) alpine sites were occupied only during pronounced interstadials and it is tempting to speculate that the high abundance of cave bear remains in at least some of these sites reflects high mortality during hibernation at the end of these milder intervals. In contrast, data from low-lying cave sites such as Tischoferhöhle in Tyrol (594 m asl) suggest that these warmer caves were also used by the cave bear during stadials, e.g. the two stadials preceding and following Greenland Interstadial 8 (Spötl et al. Reference Spötl, Reimer, Rabeder and Scholz2014).
The new 14C data from Conturines Cave strongly suggest that this extreme site was used by Ursus ladinicus prior to MIS 3, possibly during the Last Interglacial or, less likely, during the First or Second Early Würmian Interstadial (MIS 5c and 5a, respectively). During these times the treeline was likely up to a few hundred meters higher (Last Interglacial) respectively lower (Early Würmian Interstadials) than today, but still lower than the cave site. Rather than claiming an extreme treeline rise up to the elevation of the cave (e.g., Rabeder Reference Rabeder1991) we attribute the survival of this population to the adaptation of the small Ursus ladinicus to steep mountainous environments at the very edge of their ecological niche. This is supported by the stable isotope composition of its bones, which yielded record-low δ13C values and among the lowest δ15N values of all European cave bear samples (Bocherens Reference Bocherens2015; Krajcarz et al. Reference Krajcarz, Pacher, Krajcarz, Laughlan, Rabeder, Sabol, Wojtal and Bocherens2016). These studies show that δ13C of collagen scales inversely with altitude and our data (Table 1) confirm the published data of Conturines bone samples (Fernández-Mosquera et al. Reference Fernández-Mosquera, Vila-Taboada and Grandal-d’Anglade2001; Horacek et al. Reference Horacek, Frischauf, Pacher and Rabeder2012). These two previous studies also reported δ15N values between +0.3 and +2.1‰, which are somewhat higher than those of our samples (–0.6 to +0.4‰, one value of +2.9‰, see Table 1). The data do support, however, the previously recognized deviation of the Conturines samples from a general trend of decreasing cave bear δ15N values with increasing altitude. This was attributed to different altitudes of feeding and hibernating of Ursus ladinicus at this extreme site in the Dolomites (Bocherens Reference Bocherens2015; Krajcarz et al. Reference Krajcarz, Pacher, Krajcarz, Laughlan, Rabeder, Sabol, Wojtal and Bocherens2016).
CONCLUDING REMARKS
The new 14C analyses help to resolve a long-standing conundrum showing that these cave bear remains carry no implications for the MIS 3 paleoclimate in the Alps. In the absence of speleothems associated with these remains further refinement of these infinite 14C dates will be challenging.
This study in conjunction with others performed on old bone material (e.g. Plicht and Palstra Reference van der Plicht and Palstra2016) emphasize that previous age assignments and interpretations of alpine cave bear dates close to the upper limit of 14C dating require a reassessment. Examples include two dates from Brieglersberghöhle (Totes Gebirge, 1960 m asl; Rabeder et al. Reference Rabeder, Hofreiter and Wild2005), which yielded 50,700+4200/–2700 and 48,700+3200/–2300, an age of 48,740±800 BP from Bärenfalle (Tennengebirge, 2100 m asl; Frischauf et al. Reference Frischauf, Krutter and Rabeder2015) and an age of 50,800+4300/–2800 BP from Gauerblickhöhle (Sulzfluh, 2305 m asl). The latter sample was subsequently reanalyzed following ultrafiltration and the resulting age was >52,600 BP (Büchel et al. Reference Büchel, Laughlan and Rabeder2014). Ages beyond some 45 ka BP are extremely sensitive to contamination by modern carbon and issues related to the blank become very important.
ACKNOWLEDGMENTS
The ongoing research on Conturines Cave is carried out with permission from the Naturpark Fanes-Sennes-Prags and we are particularly grateful to G Nagler for this continuous support. Part of the work received funding from the Autonomous Province of Bozen-Südtirol. We appreciate the thoughtful comments by M Robu and an anonymous reviewer.
Open access
