Ursus deningeri subsp. hercynicus Rode 1935

Ursus deningeri hercynicus Rode, 1935 REFERRED MATERIAL. — The material of U. deningeri, the most abundant carnivore species from TW is represented by almost all skeletal elements (NISP 641, MNI 26; left/right bones are given in brackets): 4 cranium fr., 5 maxilla fr. (4/1), 15 mandibles (8/7), 19 I1 (10/9), 26 I2 (8/18), 30 I3 (14/16), 19 C1 (11/8), P1 (0/1), 15 P4 (9/6), 30 M1 (17/13), 44 M2 (22/22), 15 i1 (7/8), 26 i2 (15/11), 15 i3 (6/9), 25 c1 (12/13), 3 p1 (3/0), 13 p4 (5/8), 43 m 1 (17/26), 34 m 2 (16/18), 27 m 3 (15/12), scapula (1/0), humerus (0/1), 7 radii (3/4), 5 ulnae (3/2), 6 tibiae (3/3), 2 fibulae (2/0), sternum (1), 4 patellae (2/2), 4 naviculares (4/0), 6 calcanei (0/6), 4 tali (0/4), 2 capitatums (1/1), 2 cuboides (0/2), 3 ectocuneiformes (1/2), 4 pisiformes (3/1), 2 scapholunares (0/2), 3 mc 1 (1/2), 6 mc 2 (1/5), 9 mc 3 (3/6), 7 mc 4 (3/4), 9 mc 5 (6/3), 7 mt 1 (5/2), 5 mt 2 (3/2), 6 mt 3 (3/3), 13 mt 4 (5/8), 7 mt 5 (5/2), 2 mtpd, 5 atlas (5), 8 thoracic (8), cervical (1), lumbar (1), 2 caudale (2), 44 ph 1, 17 ph 2, 20 ph 3, and 2 sesamoides (Appendix 1). Some bones hold the traces of chewing, gnawing and biting, and documented the interaction with other carnivores. The main factors responsible for the accumulation of bear remains on the site were probably carnivores, accidental falls, and water transport. EMENDED DIAGNOSIS. — Smaller than the nominotypical subspecies Ursus deningeri deningeri (von Reichenau, 1904), often present P3, I3 with a calyx, i1 and i2 with larger and broader crowns and enlarged distoconid, P4 and p4 with morphologically more complicated crowns, M1 with a stronger developed parastyle and lingual cingulum, M2 with a narrower talon, m1 with a longer talonid and triple entoconid, m2 with a broader and longer talonid, and m3 with a shorter trigonid (Rode 1935; Schütt 1968; Baryshnikov 2007). DESCRIPTION Upper incisors Among 39 I2s from TW, 29 teeth belong to the most ancestral morphotype d with an asymmetrical, narrow crown without a lingual edge nor fossa lunaris (Fig. 3). The other 10 I1-I2s represent the more advanced morphotype d/p,which is broader, with a weakly developed lingual part, but still without any cusplets on the mesial cingulum. The morphotype d predominates in the sites dated on MIS 19-13; it is characteristic for all I2s from Kozi Grzbiet, Mosbach 2, and Stránska skálá (Table 1). The I2 from TW is metrically comparable to those from the sites dated on MIS 19-15, but with more robust crowns (Fig. 3). The B/L index for TW is 81.5 (75.6-85.6, n = 10), and it is higher than MIS 19-15 group of U. deningeri (B/L = 77.9, 77.1-78.7, n = 9) (Table 1). The bear crowns of younger populations (MIS 11-9) are larger and broader (B/L = 82.7, 77.8-91.9, n = 15), the ratio of the morphotypes p to the d is higher, and the calyx size increasing (Table 1). All I3s from the group 1 (MIS 19-13) represent the most ancestral morphotype 0, without calyx (Table 1). Only 2 I3s from Hundsheim hold a very delicate structure resembling a minute calyx. Among 26 I3s from TW, seven specimens represent the more evolved morphotype 1, with a minute to small calyx located on the buccal side, above the apex (Fig. 4). The rest 21 specimens of I3 from TW were assigned to the morphotype 0 (Fig. 4). The I3 from TW has a developed lingual edge and fossa lunaris that occurs as a clearly visible pit on the mesial side of the tooth. The cingulum is weak and even absent in some of the specimens (Fig. 4). In all I3s from TW, the mesial and distal edges delimit a somewhat recessed field with the lingual cingulum. The I3s from TW are larger than those from MIS 19-13, but smaller than I3s from MIS 11-9. A general tendency to increasing in size and broadening of all three upper incisors is observed (Table 1). Lower incisors (i1-i3) and canines (C1/c1) Among 14 i1s from TW, eight specimens represent the most ancestral morph. A, simply build and small, while five others were assigned to the morph. C (Table 2). The latter have larger and broader crowns, with an enlarged distoconid, and thickened mesial and mesio-buccal margins. The moph. A is the only occurring in the group 1 (MIS 19-13). The i1 from TW is large and robust, and its size exceeds the dimensions of the i1 from most of the sites the material from which was used for comparison (Table 2). The relatively large crown of the i2 is divided into two (in most specimens) or three (in a few teeth) parts, and its apex is oriented vertically and slightly distally, being oval-shaped in occlusal view (Fig. 5; Table 2). Among 25 i2s from TW, 10 represent the most ancestral morph. d, with a simply build crown, divided into two parts (Table 2). The other 10 i2s from TW belong to the morph. d/s, with a weakly developed mesioconid, smooth distal surface, and without any cusplets. Finally, two specimens were assigned to the tricuspid morph. s, with a large mesioconid and a double distoconid (Fig. 5; Table 2). The i2 from TW represents a higher evolutionary level as compared to the group 1 (MIS 19-13). Among six i2s from Kozi Grzbiet, five specimens represent the morph. d, while only a single – d/s. In Hundsheim (n = 5) and Stránska skálá (n = 4), the morph. d/s is represented only by a single tooth. The i2 from the group 1 (MIS 19-13) is also slightly smaller than that from TW, which in turn is smaller than i2 from the group 2 (MIS 11-9). In the group 2, the number of more evolved morphotypes (d/s and s) is higher than that in TW (Table 2). Among 15 specimens of i3 from TW, nine were assigned to the morph. B, a double-crowned, well-developed distoconid distinct from the protoconid, and a well-developed sulcus medialis, while six others belong to B/C (Fig. 6). The i3 from the group 1 (MIS 19-13) is smaller and evolutionarily more ancestral, although a high variability is inherent to this group (Table 2). Among 37 canines from TW, majority of the specimens is more or less worn, and therefore it is not possible to take their precise measurements. Among them, large and robust male canines are easily distinguishable from the smaller and narrower female canines. In addition, the lower canines are moderately curved distally and weakly flattened bucco-lingually, while the upper canines are straighter. All canines are characterised by an elongated and massive root, oval-shaped in cross-section. The crown is quite short (as compared to the root), blunt, with a poorly developed, very thin inner enamel crest running through its entire length. P4 Among 15 P4s from TW, eight specimens represent the ancestral morph. A, with a simple morphology, without any cusplets such as protoloph and metaconule (Fig. 7; Table 3). Seven other P4s from TW are more evolved, with a distinct metaloph or a small metaconule (Fig. 7; Table 3). Comparison of P4s from TW with those from other Middle Pleistocene bear palaeopopulations showed that they do not differ metrically (Table 3), but the TW material represents a slightly higher evolutionary level. Among the P4s from the group 1 (MIS 19-13), teeth with a small metaloph or metaconule can be found, but the morph. A predominates in the sample. Among nine P 4 specimens from Kozi Grzbiet, only a single tooth represents A/B, while the others – A (Table 3). p4 The morphology of p4 from TW is highly variable, and metrically most of them are quite small (Fig. 8; Table 4). Four specimens from TW belong to the ovoid morphotype B2, with a prominent protoconid, moderate metaconid edge running distally, conical paraconid located mesio-lingually and a small but distinct hypoconid shifted disto-lingually. Two other p4s from TW represent the morphotype D2, with a well-developed paraconid and metaconid of variable size, and a small cusplet between them. One specimen was recognised as representing the morphotype D3, with a small entoconid (Fig. 8; Table 4). Two p4s from TW belong to C2, similar to B2, but with a small metaconid and a strong hypoconid shifted mesio-lingually. Three others were assigned to C3, with a small entoconid located between the paraconid and the hypoconid (Fig. 8; Table 4). Ancestral morphotypes of p4 (B1 and C1) are the most common in the group 1 (MIS 19-13), where also a very ancestral p4 is recorded (Table 4). These single-cusped teeth without para- and metaconid closely resemble the morph. b1, typical for arctoid bears. Metrically, p4s from TW do not differ from those in bear palaeopopulations of both groups, also the width to the length ratio (B/L) does not show any particular differences. Only the material from Biśnik Cave shows slightly higher dimensions (Table 4). The p4 from TW is already not so ancestral as those from the group 1 (MIS 19-13), but it cannot by any means be considered to be advanced. The morphodynamic index of P4/p4 shows that the TW material is grouped alightly above the populations from the group 1 (MIS 19-13) (Fig. 9). Simultaneously, the TW index is clearly lower than that of the group 2 (MIS 11-9) and U. deningeri from TW holds an intermediate position between both groups (Fig. 9). M1 Morphologically, the M1 from TW shows the predominance of ancestral features, closely resembling those from the group 1 (MIS 19-13), although with a number of more progressive characteristics (Fig. 10; Table 5). Among 21 M1s from TW, nine teeth have a smooth internal slope of the paracone, (morph. 0), while the next 12 specimens have a moderately thick edge running from the paracone apex to its base, and accompanying with one or two smaller thin pillars reaching the half of its height (morph. 1) (Table 5). Internal slope of the metacone is smooth in 10 of 24 teeth (morph. 0), while 14 others represent the morph. 1 (Fig. 10; Table 5). The small metastyle is shallow and elongated, without any distal edge and is strongly associated with the distal cingulum. The mesocone is undivided. The shallow and small metastyle is weakly developed, with a smooth internal surface. The protocone is a low rectangular cusp, on the internal wall of