Beavers and flying squirrels (Rodentia: Castoridae, Pteromyini) from the Late Pliocene of Hambach 11 C, Germany Author Laere, Gaëlle Van Department of Palaeobiology, Swedish Museum of Natural History, P. O. Box 50007, 104 05 Stockholm (Sweden) gaelle.vanlaere@gmail.com Author Mörs, Thomas Department of Palaeobiology, Swedish Museum of Natural History, P. O. Box 50007, 104 05 Stockholm (Sweden) and Bolin Centre for Climate Research, Stockholm University, Stockholm (Sweden) thomas.moers@nrm.se text Geodiversitas 2023 2023-04-20 45 7 223 241 journal article 10.5252/geodiversitas2023v45a7 1638-9395 7867111 urn:lsid:zoobank.org:pub:1B8E4EDF-D6E0-4118-A8C9-CFC7589EC5C4 Trogontherium minus Newton, 1890 ( Figs 4-6 ) MATERIAL EXAMINED . — Left mandibular fragment with incisor to m2 ( IPB-HaR-5642 ), right maxillary fragment with P4 to M2 ( IPB-HaR-5643 ). Isolated teeth: left dP4 ( IPB-HaR-5650 ), left P4 ( IPB-HaR-5645 ), right P4 ( IPB-HaR-5649 ), right M1/2 ( IPB- HaR-5652 ), left M1/2 ( IPB-HaR-5646 and IPB-HaR-5647 ) , right M3 ( IPB-HaR-5644 and IPB-HaR-5651 ) , left M3 ( IPB- HaR-5648 ) and incisor fragment ( IPB-HaR-5654 ). MEASUREMENTS . — See Table 2. DESCRIPTIONS i – HaR-5642 ( Fig. 4 ) The specimen exhibits the only complete incisor presented here despite several fragments also being found. The rounded enamel band is finely wrinkled but does not show longitudinal grooves. The wear facet is slightly curved and runs from the tip of the incisor to the lower jaw symphysis. m1/2 – HaR- 5642 m 1 and HaR- 5642 m 2 ( Fig. 4 ) The hypostriid reaches the base of the exposed tooth. No cement is in the striids. The anterior half of the tooth is higher than the posterior half. The paraflexid is curving anteriorly. The hypoflexid and mesoflexid seem to be “face to face” and meet at the middle of the tooth. The mesoflexid is horizontal and the hypoflexid seems to be oriented posteriorly. The metaflexid is oriented anteriorly. The rectangular shaped m1 is slightly worn. The tooth is dipping anteriorly ( Fig. 4A, C ). The labial side is shorter than the lingual. The entire height of the striids is exposed. The hypostriid closes at the end of the exposed tooth. There is no cement in the striids. The hypoflexid is curving posteriorly. The metaflexid is horizontal and reaches the middle of the tooth. The mesoflexid is short, slightly curving anteriorly. The paraflexid is curving anteriorly, almost reaching the anterior wall. The m2 is slightly worn. There is no cement in the striids. The tooth morphology identical with the m1. p4 – HaR-5642 ( Fig. 4 ) The p4 is erupting. It has not reached the occlusal surface yet and is therefore completely unworn. There is no cement in the striids. The hypostriid reaches the base of the exposed tooth. The anterior half is higher than the posterior half. The paraflexid is curving anteriorly. The hypoflexid and mesoflexid seem to be “face to face” and meet at mid-length of the tooth. The mesoflexid is horizontal to the width and the hypoflexid seems to be oriented posteriorly. The metaflexid is oriented anteriorly. I – HaR-5654 ( Fig. 5A ) The figured incisor fragment shows the characteristic crenulations and the longitudinal lines. The other incisor fragments from this site which are not figured, also show these features although not as clearly as in this one. All fragments show a convex surface. dP4 – HaR-5650 ( Fig. 5B ) This medium worn tooth is very low crowned. The shape is close to triangular, although quite rounded. The anterior and posterior sides are rounded and the anterior one is wider. The parastria and the mesostria span the whole preserved height of the tooth, the metastria is almost worn off. The hypostria is longer than the others but does not reach the base of the tooth crown. The paraflexus is facing the hypoflexus although the hypoflexus ends slightly more anteriorly. The mesoflexus is open on both sides of the tooth and creates an indentation on the labial side. The metaflexus is curving anteriorly before dipping posteriorly (with a smaller angle). FIG . 4. — Trogontherium minus Newton, 1890 from Hambach 11C, IPB-HaR-5642, left jaw with i to m2: A , labial view; B , occlusal view; C , lingual view. Scale bar: 1 cm. P4 – HaR-5643, HaR-5645 and HaR-5649 ( Fig. 5 C-G) The three teeth are heavily worn with the hypoflexus closing. HaR-5643 is preserved in a bone fragment with two other teeth ( Fig. 5 C-E). Shape is triangular, the anterior side is rounded and the labial side shorter than the lingual side. The hypostria is the only remaining stria. There is no cement in the fossettes nor the hypoflexus. The enamel of HaR-5649 ( Fig. 5G ) is wrinkled, which is more strongly visible on its antero-lingual side. The hypoflexus meets the parafossette. It is oriented anteriorly on HaR-5645 ( Fig. 5F ) and rather straight on HaR-5643 and HaR-5649. On this tooth the fossettes are also rather straight whereas they are more curved on HaR-5645. Both metafossette and mesofossette are parallel to the anterior wall except on HaR-5643 where they are tilting postero-labially. They are touching the outer walls of the tooth. The fossettes and the hypoflexus are more or less parallel to each other’s. The enamel of the anterior wall, on the posterior side of the fossettes and of the hypoflexus are thicker. On HaR-5645 and HaR-5649, the posterior root is broken off ( Fig. 5F 1 ; G 1 ). FIG . 5. — Trogontherium minus Newton, 1890 from Hambach 11C: A , IPB-HaR-5654, labial view of an incisor fragment; B , IPB-HaR-5650, left dp4 labial, occlusal, and lingual views; C , IPB-HaR-5643, upper right jaw fragment with teeth P4 to M2 labial view; D , IPB-HaR-5643, occlusal view; E , IPB-HaR-5643, lingual view; F , IPB-HaR-5645, left P4 labial, occlusal, and lingual views; G , IPB-HaR-5649, right P4 labial, occlusal, and lingual views. All teeth are figured as left teeth, an underlined letter indicates a right tooth. Scale bar: 2 mm. FIG . 6. — Trogontherium minus Newton, 1890 from Hambach 11C: A , IPB-HaR-5646, left M1/2 labial, occlusal, and lingual views; B , IPB-HaR-5647, left M1/2 labial, occlusal, and lingual views; C , IPB-HaR-5652, right M1/2 labial, occlusal, and lingual views; D , IPB-HaR-5644, upper right M3 occlusal, labial and lingual views; E , IPB-HaR-5648, left M3 labial, occlusal, and lingual views; F , IPB-HaR-5651,right M3 labial, occlusal, and lingual views.All teeth are figured as left teeth, an underlined letter indicates a right tooth. Scale bar: 2 mm. M 1/2 – HaR-5643, HaR-5646, HaR-5647 and HaR-5652 ( Figs 5 C-E; 6A-C) HaR-5643 ( M 1 and M 2, Fig. 5 C-E), HaR-5646 ( Fig. 6A ) and 5647 ( Fig. 6B ) are heavily worn. On HaR-5643 and HaR-5647, the hypoflexus is not fully closed yet and the hypostria is still visible on HaR-5643 although shorter on the M 2 than on the M . HaR-5652 ( Fig. 6C ) is only slightly worn and is the only specimen still retaining striae. The parastria is really short and the mesostria a little longer. The metastria is not observable. On this specimen the hypostria is quite long, although not reaching the base of the tooth. Due to the early wear stage, this tooth still has a relatively triangular shape with a rounded anterior side, wider than the posterior side. The postero-buccal face, below the occlusal surface is damaged. The shape of the other four teeth is close to square with the anterior side rounded and the labial side shorter than the lingual. On HaR-5646 and 5647, the parafossette and the metafossette are small, although smaller on HaR-5647. They are also really small on the M 1 and M 2 on HaR-5643. The hypofossette touches the parafossette. The fossettes are more or less parallel to the anterior side. On HaR-5643, the fossettes are titling postero-lingually and more or less parallel to the hypoflexus. On these teeth, the hypoflexus is going anteriorly and meeting the parafossette. The parafossette and metafossette of the M 2 on HaR-5643 are bigger than in the M 1 and hence, the hypoflexus is shorter. The metafossette of HaR-5647 is a round island ( Fig. 6B 2 ). On HaR-5652, only the metafossette is closed and it has a curving shape, pointing toward the anterior side. The hypoflexus is facing the paraflexus and the fold anterior to the hypoflexus is not as wide as the posterior fold. The mesoflexus is curving posteriorly, reaching the labial wall. There is no cement in the fossettes of all the specimens. On HaR-5646, two roots are preserved. An anterior one, wide, and a small, postero-lingual one ( Fig. 6A 1 ). On HaR-5647, only one root is preserved, the postero-lingual one is broken off. On HaR-5652, no roots are formed but the base of the crown is visible on the anterior side. The roots of the M 2 on HaR-5643 are more exposed than for the M 1 and a postero-labial small root seem to have been broken off. M 3 – HaR-5644, HaR-5648 and HaR-5651 ( Fig. 6 D-F) The shape of the specimens is elongated and triangular. The anterior side is rounded, and the posterior side is narrower than the anterior. HaR-5644 ( Fig. 6D ) and 5648 ( Fig. 6E ) are heavily worn with only the hypoflexus not closed. In these two teeth, the height of the hypostria, which is the only stria remaining, indicates that it is closing. HaR-5651 ( Fig. 6F ) is slightly to medium worn. Its parastria and mesostria are visible, although quite short. The metastria is worn off. The hypostria does not span the whole height of the tooth but is clearly visible and about twice the length of the parastria. The parafossette and metafossette of HaR-5644 and 5648 are rather small and about the same size. There is an additional posterior fossette of approximately the same size ( Fig. 6D 1 , E 1 ). The metafossette and the additional posterior fossette are oriented towards the postero-lingual side. On HaR- 5644, the hypoflexus is oriented towards the anterior side. On HaR-5648, the hypofossette meets the anterior side of the parafossette. The occlusal surface of HaR- 5651 is different from the other two teeth due to a difference in wear. On this tooth, the paraflexus is facing the hypoflexus and the fold anterior to the hypoflexus is not as wide as the posterior fold. The mesoflexus is open on the whole width of the tooth and forms a small indentation on the labial side. The metaflexus is closed but the metafossette is slightly open on the labial side indicating that the metaflexus was probably spanning the whole width of the occlusal surface. There is a small hole posteriorly, potentially indicating the presence of a posterior fossette. On HaR-5644, three roots are preserved. The anterior-most one is the widest and is curving like the anterior wall of the tooth. The antero-lingual one is small, erupting from the wider, most anterior root. The third one is located posteriorly. On HaR-5648, two roots are present. The anterior one is wide and following the anterior curve of the tooth. The posterior one is oval. On HaR-5651, no roots are preserved but the base of the crown is visible on the anterior side. REMARKS The right maxillary fragment HaR-5643 and the potentially corresponding M 3, HaR-5644, have been matched based on size, position, and degree of wear. The similarity in size and wear stages of the two upper tooth rows of Trogontherium (specimens HaR-5643 to HaR- 5648) indicate that they might have belonged to the same individual. The lack of longitudinal lines on the incisor of HaR-5642 is probably due to the fact that it is a juvenile specimen and therefore smaller and less developed ( Hugueney et al. 1989 ). The attribution of our material to Trogontherium is based on several characteristics. These are the presence of elongated, cone shaped upper M 3s, P4 substantially bigger than the molars and a wrinkled enamel on the incisors ( Mayhew 1978 ) as shown in Fig. 6G . The taxonomy of smaller species of Trogontherium is very debated but we can safely attribute our material to Trogontherium minus based on stratigraphic age and on size. A comparison with data from different species found in the literature confirms this ( Fig. 7 ). The P 4 in our material are larger than most T. minus which could be a result of wear or of the tendency of the Trogontherium lineage to increase in size ( Stefen 2011 ). We also notice that while the length of our specimens matches the range for T. minus , they are wider and reach the lower width range of Trogontherium cuvieri that is visible in the figure 3 of Fostowicz-Frelik 2008 . This is coherent with the age and the assumption that T. minus and T. cuvieri are two representatives in the same evolutionary lineage proposed by Mayhew (1978) who also stated that only one species of Trogontherium was present at a time. However, Hugueney et al. (1989) have shown that T. cuvieri and T. minus coexisted during the Late Pliocene/Early Pleistocene. Some argue against the lineage hypothesis based on a too great size disparity ( Fostowicz-Frelik 2008 ). However, our specimens show that the size disparity might not be as big as previously thought when compared to the data in the literature ( Fostowicz-Frelik 2008 , Heinrich 1998 ). This is mainly observed in the P4s as the width of the M 1/2 is closer to others found in the literature ( Hugueney et al. 1989 , Newton 1890 ). The measurements used for comparison are the maximum occlusal surface and can vary with wear, therefore one of our specimens is much less wide than the rest ( Fig. 7 ). The shape and appearance of the incisor are comparable to that of T. cuvieri described by Newton (1902) , although much smaller, further confirming the attribution to T. minus . This species is rare ( Hugueney et al. 1989 ) and less well known than T. cuvieri ( Stefen 2011 ) , which makes it hard to compare our material. As stated by Mörs et al. (1998) , there are only a few previously reported occurrences of Trogontherium in Germany from the Miocene and Pliocene. T. minus was previously found in deposits from Hambach of the same age ( Mörs et al. 1998 ). It was also reported from the Slovakian MN 16a locality of Hajnačka I ( Sabol et al. 2006 ) and from the MN 16 of Podpusk-Lebyazh’e (Western Siberia) where the most complete material was found ( Vislobokova 1996 ). Trogontherium aff. minus was reported by Pascari (2021) from the Pliocene of Musaitu in the Taraclia district ( Moldova ), but the published measurements are inconsistent and were therefore not used here for comparison. Apoltsev & Rekovets (2015) place T. minus in the sub-genus Euroxenomys . However, we follow Hugueney & Duranthon (2012) and Mörs & Tomida (2018) who only place the Miocene Euroxenomys minutus and Euroxenomys nanus in the genus Euroxenomys . Trogontherium is an extinct genus of beavers, found from the Late Miocene ( Mayhew 1978 ) to the Late Pleistocene ( Yang et al. 2021 ). T. minus seems to have been restricted to Europe ( Yang et al. 2021 ). Unlike Castor , Trogontherium is subhypsodont ( Hugueney 1999 ; Stefen 2011 ), and seems to have a more terrestrial lifestyle ( Fostowicz-Frelik 2008 ). FIG . 7. — Size comparison of our specimens with different species. Trogontherium minus Newton, 1890 from Czernielewski 2022 , Hugueney et al. 1989 and Newton 1890 . Trogontherium cuvieri Owen, 1846 from Fostowicz-Frelik 2008 and Langeveld 2013 . Trogontherium ( Euroxenomys ) minutum (von Meyer, 1838) from Stefen & Rummel 2003 . The lower jaw HaR-5642 belongs to a juvenile individual, according to its erupting P4 and the early wear stage of the two preserved molars. Tentative age determination was done based on the timing of eruption of the p4. Different possible ages have been found in the literature based on studies of recent Castor . According to Mayhew (1978) the premolars erupt between six to ten months after birth and according ( Heinrich & Maul (2020) p4s come into wear around the end of the first year. Hugueney & Escuillié (1996) consider individuals of Steneofiber with erupting p4s to be yearlings which correspond to individuals one year old or in their second year. According to these authors, it seems that our individual was around one year old, although probably a little less. However, these estimates are to be taken with caution as we do not know the lifespan nor the postnatal development of fossil species ( Stefen & Rummel 2003 ).