Elymys, Emry and Korth, 1989

Kelly, Thomas S. & Murphey, Paul C., 2016, Mammals from the earliest Uintan (middle Eocene) Turtle Bluff Member, Bridger Formation, southwestern Wyoming, USA, Part 1: Primates and Rodentia, Palaeontologia Electronica 7 (8), pp. 1-55 : 41-47

Elymys ? emryi new species

Figure 12.1 -16, Table 8

zoobank.org/ 92BA7EA2-3D41-4D82-91A2-BB2B56AC29FF

Holotype. RM1 , SDSNH 110466 .

Referred specimens. From SDSNH Locality 5844: M1, SDSNH 110463, 110464, 110465; M2, SDSNH 100444; M3, SDSNH 110448; m1, SDSNH 110455, 110456; m2, SDSNH 110457; m3, SDSNH 110461. From SDSNH Locality 5843: m2, SDSNH 110405.

Type Locality. SDSNH Locality 5844, 105 m above the base of the Turtle Bluff Member, Bridger Formation , Wyoming .

Distribution and Age. Type locality and SDSNH Locality 5843, which is at the same stratigraphic level on strike and just west of SDSNH Locality

5844, Turtle Bluff Member, Bridger Formation, Uinta County, Wyoming. Earliest Uintan (Ui1a).

Diagnosis. Differs from Elymys complexus Emry and Korth, 1989 , by the following: 1) significantly larger size; 2) molars slightly higher crowned and slightly more lophate; 3) M1-2 usually with short to moderate mesoloph; 4) m1-2 commonly with short mesolophid; 5) m2-3 with metalophulid I; and 5) m1-2 usually with distinct hypoconulid.

Etymology. Patronym for Robert J. Emry of the Smithsonian Institution in honor of his many contributions to vertebrate paleontology, including his detailed analysis of the myomorph rodent Elymys .

Description. The M1-2 and all lower molars are longer than wide and low crowned. Four teeth are identified as M1, three of which are complete and one has the anterolabial portion of paracone broken away. The M1 talon and trigon vary from being equal in width to slightly narrower anteriorly. Two of the M1s have very small, but distinct, anterior appression facets, indicating that a small P4 was presumably present in the species. The M1 anterior cingulum is strong, extending lingually from a distinct parastyle to terminate near the anterolingual base of the protocone. Two cuspules are present along the anterior cingulum, the more lingually positioned one being the larger. The primary cusps (paracone, metacone, protocone, and hypocone) are robust and about equal in height. The preprotocrista (anterior arm of the protocone) extends anterolabially from the anterolabial corner of the protocone to a distinct protoconule and then continues to connect with a cuspate parastyle (= anterocone of cricetids) at the anterolabial corner of the tooth. A short paralophule (or protolophule I) is present that extends from the protoconule to connect with the anterolingual edge of the paracone. The endoloph is complete, connecting a small mesoconid anteriorly to the protocone and posteriorly to the hypocone. A distinct mesostyle is present between the paracone and metacone. Two M1s have a mesoloph that extends labially from the mesocone to about half the way to the mesostyle, whereas the other two have a shorter mesoloph that only extends about one third or one quarter of the way. The metaloph is complete and moderate height, connecting the metacone and hypocone. The posterior cingulum (or posteroloph) is robust, extending from the posterolabial corner of the hypocone to a distinct hypoconule and then continues to the posterolabial base of the metacone.

The one tooth identified as M2 has the enamel somewhat abraded. It is very similar in size and occlusal morphology to the referred M1s, but exhibits certain differences. The preprotocrista extends similarly from the paracone to a distinct protoconule, but then continues more anteriorly than in the M1 to connect with the anterior cingulum rather than with the parastyle. The anterocone is a distinct cuspule positioned slightly nearer the lingual terminus of the anterior cingulum and the accessory cingulum cuspule seen in the M1s is lacking. The anterior cingulum is slightly wider and oriented more vertically, whereas in the M1s it obliquely inclined. The mesocone is slightly larger.

The M3 has the talonid expanded transversely, resulting in a trapezoidal occlusal outline. The anterior cingulum is robust, extending from the anterolabial corner of the paracone to the anterolingual corner of the tooth, where it is separated from the protocone by valley. A robust anterocone is present on the anterior cingulum near its lingual terminus. The protocone and paracone are the largest primary cusps and about equal in size. The preprotocrista (anterior arm of the protocone) extends anterolabially to a weak protoconule and then continues labially connecting to the anterolingual corner of the paracone, resulting in a complete protoloph. The hypocone is distinct and connected to the posterior cingulum (posteroloph), which extends labially in a gentle arc to an incipient entoconid. A short accessory crest extends anterolingually from the entoconid into the central valley towards the mesocone. The mesocone is distinct and connected by a very short, anterolingually directed crest to the anterolabial base of the protocone and a short, posterolingually directed crest to the anterolabial base of the hypocone, resulting in a complete, but low, endoloph. A short mesoloph extends from the mesocone into the central valley. The protocone and hypocone are separated by a deep valley lingual to the endoloph.

The m1 is distinctly narrower anteriorly than posteriorly. It has a distinct anteroconid that is separated from the protoconid and metaconid by a valley. The protoconid is large with a metalophulid II (= posterior arm of the protoconid) that extends lingually in an arc from the protoconid to the posterolabial base of the metaconid. The metaconid is conical and slightly smaller than the protoconid. The entoconid and hypoconid are robust, about equal in size and larger than the metaconid and protoconid. The mesoconid is distinct, connected to the protoconid anteriorly by a relatively tall cristid and posteriorly to the hypoconid by an equally tall cristid, resulting in a complete ectolophid. A short mesolophid is present in one m1, whereas the other has no mesolophid. A strong mesostylid is present between the metaconid and entoconid. The hypolophid (entolophulid) is low, but complete, extending from the entoconid to either the lingual base of the hypoconid or the hypoconulid. The posterior cingulid is robust, extending from the posterolingual edge of the hypoconid to a strong hypoconulid and then continues lingually to terminate at the posterior base of the entoconid.

The m2 has a rectangular occlusal outline. The anterior cingulid extends lingually from a small parastyle positioned near the anterolabial edge of the metaconid to a distinct anteroconid and then continues to the anterolabial corner of the tooth, where it is separated from the protoconid by a valley. The primary cusps (metaconid, entoconid, protoconid, and hypoconid) are robust and about equal in size. The preprotocristid (anterior arm of the protoconid) extends from the anterolingual corner of the protoconid to the anteroconid, and a short cristid extends from the anteroconid to the anterolabial base of the metaconid, resulting in a complete metalophulid I. The metalophulid II is complete, extending lingually from the posterolingual corner of the protoconid to the posterolabial base of the metaconid. The mesoconid is connected anteriorly to the protoconid by a relatively tall cristid and posteriorly by a relatively tall cristid, resulting in a complete ectolophid. A short mesolophid extends lingually from the mesoconid into the central valley. The hypolophid is relatively tall and complete, connecting the entoconid to the hypoconid. A distinct mesostylid is present between the metaconid and entoconid. The posterior cingulid extends posterolingually from the hypoconid to a distinct hypoconulid and then continues to join the posterolabial base of the entoconid.

The m3 has an elongate talonid that is considerably narrower than the trigonid. The anterior cingulid is robust, extending labially from an incipient parastylid (slight swelling) at the anterolingual corner of the tooth to a distinct anteroconid, and then continues to the anterolabial corner of the tooth, where it is separated from the protoconid by a valley. The primary cusps (metaconid, entoconid, protoconid, and hypoconid) are conical and robust. The preprotocristid (anterior arm of the protoconid) extends lingually from the anterolingual corner of the protoconid to the anterolabial edge of the metaconid, resulting in a complete metalophulid I. A distinct cristid extends anteriorly from the metalophulid I to the anteroconid. The postprotocristid (posterior arm of the protoconid) extends lingually in an arc to join with the posterolabial base of the metaconid, resulting in a complete metalophulid II. The mesoconid is distinct, connected anteriorly to the protoconid, and posteriorly to the hypoconid by relatively tall cristids, resulting in a complete ectolophid. A short mesolophid extends lingually from the mesoconid. A distinct mesostylid is present between the metaconid and entoconid. The entolophulid extends labially from the entoconid to either the labial base of the hypoconid or to the posterior cingulid near its origin from the hypoconid, resulting in a complete metalophid. The posterior cingulid extends lingually from the posterolingual corner of the hypoconid to the posterior base of the entoconid.

Remarks. Emry and Korth (1989) described Elymys complexus from the early Bridgerian Elderberry Canyon Local Fauna, Sheep Pass Formation, Nevada, and questionably assigned it to the Zapodidae Coues, 1875 , based on the presence of a small, peg-like P4. Subsequently, Emry (2007) provided a detailed account of the species based on a much larger sample, wherein he considered Elymys to be a definitive basal myomorph possibly representing the ancestral morphotype that lead to both cricetids and dipodoids.

Rodrigues et al. (2010) also recognized Elymys as the earliest known myomorph and regarded it as a basal dipodoid, but noted that it shares a suite of characters with the first cricetids. For example, in early representatives of the basal cricetid Pappocricetodon Tong, 1992 , of Asia (e.g., P. antiquus Wang and Dawson, 1994 ; P. kazakstanicus Emry et al., 1998 , and P. neimongolensis Qian, 2012 ) a small P4 is present, the anterior arm of the M1 protocone extends anterolabially to join the protoconule and then commonly continues as an anterolophule to terminate at a small cuspate anterocone (= parastyle), the M2 has a squared occlusal outline, the m1 trigonid is significantly narrower than the talonid, and the m2-3 have a small, but distinct, transversely elongate anteroconid. Although compared to more derived cricetids, the M1 anterocone of these early members of Pappocricetodon is significantly smaller and does not extend nearly as far forward. The TBM teeth share the above characters of these basal cricetids, but differ by having a shorter, weaker M1 mure that extends less anterolabially, a slightly more distinct (more cuspate) M1 anterocone (= parastyle), a taller, better developed M1 anterolophule, more weakly developed upper molar mesolophs (not mesostylar), a stronger m1 posterior arm of the protoconid, a stronger metalophulid I and II on m2- 3, distinct molar hypoconules and hypoconulids, and lower molars lacking ectomesolophids. Although difficult to distinguish basal cricetids from basal dipodoids ( Rodrigues et al. 2010), the above differences appear to eliminate the TBM teeth as representing a basal cricetid immigrant from Asia.

Based on a small sample of specimens from the late Uintan through early Duchesnean portions of the Sespe Formation of southern California, Wilson (1935a, 1935b) described three species of Simimys ( S. simplex , S. vetus , and S. murinus ). Lillegraven and Wilson (1975) provided an analysis of a large sample of Simimys from California and concluded that Wilson's (1935a) three species actually represent a single normally variable species, making S. vetus and S. murinus junior synonyms of S. simplex . One additional species of Simimys , S. landeri from the late Duchesnean portion of the Sespe Formation, was described by Kelly (1992), and is significantly larger than S. simplex .

When examining the TBM teeth, it was clear they represent a new species, but their generic allocation was questionable because they appear to be intermediate morphologically between Elymys and Simimys . It could be argued that they represent a transitional morphotype that rises to the level of generic separation, but erecting a new genus based on a small sample of isolated teeth cannot be justified. Because the new species exhibits more similarity to Elymys than Simimys , we questionably referred them to the former genus.

Elymys ? emryi shares the following characters with E. complexus : 1) brachydont; 2) presumably a small P4, based on small anterior appression facets on half of the M1s; 3) M1-2 longer than wide; 4) M1 anterior arm of the protocone (preprotocrista) extending anterolabially from the protocone to a protoconule and continuing as an anterolophule to connect with a cuspate parastyle (= anterocone of cricetids) on the anterolabial terminus of the anterior cingulum; 5) M3 talonid notably wide transversely and long posteriorly; 6) M3 protoloph low, but complete; 7) m1 trigonid notably narrower than talonid; 8) m2 protoconid connected to anteroconid by a short cristid; 9) m3 trigonid distinctly wider than talonid; and 10) endolophs and ectolophids complete on the upper and lower molars, respectively. Elymys ? emryi also shares the following characters with Simimys simplex and S. landeri , some of which are also shared with E. complexus (see above): 1) M1-2 longer than wide; 2) the M1 anterior arm of the protocone extends anterolabially to the protoconule and then continues as an anterolophule to connect with a distinct parastyle on the labial terminus of the anterior cingulid; 3) m2-3 with a metalophulid I; and 4) M1-2 and m1-2 with mesolophs and mesolophids, respectively. Elymys ? emryi differs from Simimys simplex and S. landeri by the following: 1) cheek teeth significantly more brachydont and less lophodont; 2) mesolophs and mesolophids much weaker and shorter; 3) M2 anterolophule lacking (no connecting lophule between protoconule and parastyle); 4) M3 lacking any connections with anterior cingulum; and 5) molar accessory lophs and lophids lacking.

Wilson (1935a) regarded Simimys as a possible cricetid based on similarities in occlusal morphology to later cricetids. Since then, numerous investigators have debated the phylogenetic relationship of Simimys to other myomorph rodents (e.g., Wood, 1937, 1974; Stehlin and Schaub, 1951; Lindsay, 1968, 1977; Wilson, 1975; Emry, 1981; Vianey-Liaud, 1985; Wang and Meng, 1986; Kelly, 1992; Korth, 1994; Emry and Korth, 1989; Walsh, 1997; Emry, 2007; Rodrigues et al., 2010). Vianey-Liaud (1985) suggested that dipodoids first immigrated to North America from Asia in the middle Eocene (Uintan), as represented by Simimys . Dawson et al. (1990) described Armintomys from the earliest Bridgerian and questionably assigned it to Dipodoidea Fisher de Waldheim, 1817, but Wang and Dawson (1994) later noted that it lacks the derived condition of the neurovascular canal seen in Dipodoidea (including Simimys ). Walsh (1997) described additional material of Simimys from southern California and amended the diagnosis of the genus. Based on the presence of a single-rooted, peg-like P 4 in some early representative specimens, a generalized myodont occlusal morphology and a hystricomorphous zygomasseteric structure, Walsh (1997) regarded Simimys along with Elymys and Nonomys Emry and Dawson, 1973 , as possible basal dipodoids and the closest sister taxa to the Zapodidae . Walsh (1997) further regarded Myomorpha Brandt, 1855, to be derived from Sciuravidae with Eomyidae , Armintomys , and Pauromys as the closest successive sister taxa, respectively, to Dipodoidea (including Elymys and Simimys ) and Muroidea Illiger, 1811. Emry (2007) provided the following scenario in his reinterpretation for the origin of Myomorpha: 1) myomorphs originated in North America from Sciuravidae at least as early as the beginning of the middle Eocene (earliest Bridgerian) with subsequent immigration of an Elymys -like myomorph to Asia during the early middle Eocene, from which dipodoids ( Zapodidae ) and muroids ( Cricetidae Fisher von Waldheim, 1817 ) evolved in Asia; 2) an immigration of eumyines derived from something close to Asian Eucricetodon Thaler, 1966 , back from Asia to North America near the end of the Eocene (Chadronian); and 3) followed by an immigration from Asia to North America of zapodid dipodoids in the Miocene. Emry (2007) also considered Simimys as more likely to have a North American origin, where it was derived from an Elymys -like ancestor.

The increased size, slightly higher crowns, slightly more lophate occlusal pattern, development of short mesolophs and mesolophids, and presence of a metalophulid I on m2-3 of E.? emryi indicate that it is slightly more derived than E. complexus . These characters also suggest that Simimys could have been easily derived from E.? emryi or a very similar ancestral morphotype, further supporting Emry's (2007) proposal of a North American origin for Simimys .

Rodentia , family undetermined

Figure 6.2-3 , Tables 2, 9

Referred specimen. From UCM Locality 92189, Lm1 or 2, UCM 95700.

Description. UCM 95700 is unworn with an anteroposteriorly elongate occlusal outline, wherein its length is significantly longer than its width (ap/tra = 1.305 and ap/trp = 1.283). The trigonid and talonid widths are nearly equal. The anterior cingulid is complete and relatively tall, extending from the anterolingual corner of the protoconid to the anterolabial corner of the metaconid. The metalophulid II is tall and complete, extending posterolingually from the posterolingual corner of the protoconid and, at about its center, bends back anterolingually to join the anterolabial corner of the metaconid, resulting in a distinct inflection and relatively wide trigonid basin. The metastylid crest is a relatively strong ridge positioned along the posterolingual border of the metaconid and separated from the metaconid apex by a distinct notch. The primary cusps (protoconid, metaconid, entoconid, and hypoconid) are robust and nearly equal in size. The trigonid and talonid basins are deep. A ridge is present on the center of the labial surface of the entoconid that extends labially from its apex to its base (incipient hypolophid). The ectolophid is tall, straight, and complete, connecting a small, but distinct, oval shaped mesoconid anteriorly to the protoconid and posteriorly to the hypoconid. The posterior cingulid (posterolophid) is tall, robust and extends from the posterolingual edge of the hypoconid to a strong, centrally positioned hypoconulid and then continues lingually to terminate near the posterolingual corner of the tooth, where it is separated from the entoconid by a relatively wide, deep valley. The enamel is moderately crenulated.

Remarks. UCM 95700 was identified in the UCM catalog as Aplodontidae , genus and species undetermined. When first examining UCM 95700, the question arose; could this tooth possibly represent a cylindrodontid that lacks the typical complete hypolophid as seen in some examples of Bridgerian Mysops (e.g., m1 of M. parvus plicatus or m2 of M. minimus ; see also Wilson, 1938b: p. 216)? Moreover, it is similar in length to that of m1-2 of Bridgerian Mysops . UCM 95700 differs from the m1-2s of Bridgerian Mysops and cf. Pareumys sp. from the TBM in a number of characters. Its occlusal outline is significantly more elongate, with the tra and trp narrower relative to the ap (ap/tra ratio = 16-17% narrower and ap/trp ratio = 16-18% narrower), and the trigonid and talonid are nearly equal in width. The entoconid is very robust, relatively larger and wider, and not anteroposteriorly compressed. The anterior cingulid is taller and more complete. The metalophulid II (posterior arm of the protoconid) is relatively taller, more completely connected to the posterolabial wall of the metaconid, and inflected posteriorly at about its midline. The hypoconulid is taller and more distinct, positioned near the midline of the posterior cingulid (= posterolophid). The ectolophid is taller, straighter, and more complete, whereas those of Mysops and cf. Pareumys sp. are more convex (inflected), especially posteriorly. The metastylar crest is more distinct. The enamel is crenulated (although not heavily). UCM 95700 also exhibits some occlusal similarity to the m1-2 of certain species of the early Eocene (Wasatchian) cylindrodont Tuscahomys Dawson and Beard, 2007 , which have a distinct hypoconulid and mesoconid, and commonly exhibit an incomplete or short hypolophid (e.g., T. ctenodactylops [ Korth, 1984] and T. walshi Anemone et al., 2012 ). However, UCM 95700 can be easily distinguished from these species by the following ( Korth, 1984; Dawson and Beard, 2007; Beard and Dawson, 2009; Anemone et al., 2012; Rose et al., 2012): 1) a significantly taller crown with deeper talonid and trigonid basins; 2) a relatively wider trigonid, more nearly equal in width to the talonid; 3) a much taller metalophulid II, connecting higher on the metaconid wall, and with a prominent posterior flexure at its midline; 4) a more distinct metastylid crest; 5) a much weaker hypolophid, barely extending labially into the talonid basin, and not directed towards or connected to the ectolophid or hypoconid; and 6) crenulated enamel.

Stock (1935) described three species of Eohaplomys from the late Uintan portion of the Sespe Formation of southern California and regarded his new genus as the then known earliest representative of the Aplodontidae . However, Rensberger (1975) provided convincing evidence that the dental similarities of Eohaplomys and prosciurine aplodontids (e.g., Spurimus Black, 1971 , and Haplomys Miller and Gidley, 1918 ) represent convergence. Subsequently, Korth (1988) recognized that Eohaplomys along with Mytonomys Wood, 1956 , actually represent ailuravine ischryomyids closely related the middle Eocene Ailuravus Rütimeyer, 1891 , of Eurasia (see also Weitzel, 1949; Michaux, 1968; Wood, 1976; Hartenberger, 1995; Escarguel, 1999; Bernard et al., 2012).

In an attempt to further determine the familial status of UCM 95700, its dental characters were compared to those of the aplodontids Spurimus and Haplomys , along with the ailuravines Eohaplomys and Ailuravus , which exhibit some convergent occlusal similarity to early aplodontids ( Table 9). Based on this comparison, UCM 95700 appears to have an occlusal morphology that is slightly more similar to the ailuravines Eohaplomys and Ailuravus ( Ischyromyidae ) than to Haplomys and the basal aplodontid Spurimus ( Aplodontidae ). Comparison of UCM 95700 to the aplodontid Prosciurus Matthew, 1903 , which has many of the same dental character states as Haplomys ( Rensberger, 1975) , further indicates that it does not represent a basal aplodontid. UCM 95700 differs from late Uintan Eohaplomys by having a less lophate lower molar with the hypolophid incomplete, and with slightly less development of the metastylid crest and a weaker mesoconid. UCM 95700 is also significantly smaller than Eohaplomys . Moreover, no upper molars have been recovered from the TBM with the distinctive morphology seen in those of Eohaplomys or Ailuravus .

To summarize, UCM 95700 appears to represent a previously unknown rodent whose familial and generic allocations are uncertain. It could represent a cylindrodont or possibly a basal ailuravine.

Figures

FIGURE 6. Thisbemys and Rodentia, family undetermined from TBM. 1, Thisbemys corrugatus, Rm1, UCM 70671. 2-3, Rodentia, family undetermined, Rm1 or 2, UCM 95700. 1-2, occlusal views. 3, oblique labial view. Scale bar equals 1 mm.

FIGURE 12. Elymys? emryi new species from TBM, macrophotographs with corresponding line drawings of specimens: 1 and 9, LM1, SDSNH 110463; 2 and 10, RM1, SDSNH 110465; 3 and 11, RM1 (holotype), SDSNH 110466; 4 and 12, RM2, SDSNH 110444; 5 and 13, LM3, SDSNH 110448; 6 and 14, Rm1, SDSNH 110456; 7 and 15, Lm2, SDSNH 110457; 8 and 16, Lm3, SDSNH 110461. All occlusal views. Scale bar equals 1 mm.