Systematics of Paleogene Micromomyidae (Euarchonta, Primates) from North America

Journal of Human Evolution 65 (2013) 109e142

Contents lists available at SciVerse ScienceDirect

Journal of Human Evolution journal homepage: www.elsevier.com/locate/jhevol

Systematics of Paleogene Micromomyidae (Euarchonta, Primates) from North America Stephen G.B. Chester a, *, Jonathan I. Bloch b a b

Department of Anthropology, Yale University, P. O. Box 208277, New Haven, CT 06520, USA Florida Museum of Natural History, University of Florida, P. O. Box 117800, Gainesville, FL 32611-7800, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Received 21 September 2012 Accepted 23 April 2013 Available online 12 July 2013

New specimens of micromomyid plesiadapiforms recovered from the late Paleocene and early Eocene of the Clarks Fork and Powder River Basins, Wyoming, include previously unknown tooth positions of Chalicomomys antelucanus, the earliest record and first substantial Paleocene sample of Tinimomys graybulliensis, and additional specimens of early Eocene T. graybulliensis, forming the largest known sample (n ¼ 84, MNI ¼ 14) of a micromomyid species from a single fossil locality. These specimens and newly documented intraspecific variability, coupled with the first detailed descriptions of the dentition of Dryomomys szalayi, allow for a systematic revision of the family. Cladistic analysis of the 11 known micromomyid species using 28 morphological characters produced three most-parsimonious cladograms. Results suggest that several Tiffanian taxa previously classified in the genus Micromomys (excluding the type species Micromomys silvercouleei) are more primitive and are referred to a new genus Foxomomys (Foxomomys fremdi, Foxomomys vossae, and Foxomomys gunnelli). Two other Paleocene and early Eocene species previously classified in Micromomys are instead found to share a special relationship with Dryomomys (Dryomomys millennius and Dryomomys willwoodensis) based primarily on the relative size and shape of the premolars. Results further suggest that early Eocene Chalicomomys (monotypic: Chalicomomys antelucanus) is the sister taxon to a clade that includes Dryomomys and Tinimomys, which diverged from each other by the late Tiffanian. The shape of P4 and the relative size of P3 have distinct patterns of change through the evolution of the group. Additionally, there is a gradual reduction of P2, with Foxomomys having a double-rooted P2, Micromomys, Chalicomomys, and Dryomomys having a single-rooted P2, and Tinimomys lacking a P2. Body size increases from more primitive micromomyids (Foxomomys and Chalicomomys) to more derived genera (Dryomomys and Tinimomys), and size also increases from the older and/or more primitive species within the Dryomomys and Tinimomys lineages. Ó 2013 Elsevier Ltd. All rights reserved.

Keywords: Euarchontan Micromomyid Plesiadapiform Primate origins Late Paleocene Early Eocene

Introduction Micromomyid plesiadapiforms are diminutive euarchontan mammals known from the late Paleocene (middle Tiffanian North American Land Mammal Age (NALMA)) to the early Eocene (early Wasatchian NALMA) of western North America (Szalay, 1973, 1974; Bown and Rose, 1976; Krause, 1978; Bown, 1979; Rose, 1981; Rose and Bown, 1982; Fox, 1984; Gingerich, 1987; Beard, 1989, 1993a; Beard and Houde, 1989; Gunnell, 1989; Rose et al., 1993; Robinson, 1994; Wilf et al., 1998; Bloch, 2001; Strait, 2001; Bloch et al., 2007; Secord, 2008; Chester and Beard, 2012; Fig. 1, Supplementary Table 1). More specifically, micromomyids have been found in

* Corresponding author. E-mail addresses: [email protected] flmnh.ufl.edu (J.I. Bloch).

(S.G.B.

Chester),

0047-2484/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jhevol.2013.04.006

jbloch@

three NALMAs including the Tiffanian (faunal zones Ti-3eTi-5b) and succeeding Clarkforkian (Cf-1eCf-3) of the late Paleocene, and the Wasatchian (Wa-1eWa-2) of the early Eocene. Outside North America, only an undescribed dentary from the early Eocene Wutu fauna, Shandong Province, China (Tong and Wang, 1998) has tentatively been referred to Micromomyidae. Micromomyidae is generally considered a monophyletic group (e.g., Beard and Houde, 1989; Silcox, 2001) and these ‘plesiadapiforms’ (also referred to as ‘archaic primates’) are recognized in part by their large, specialized fourth premolars, and very small body size (Silcox and Gunnell, 2008). The first known micromomyids, Micromomys and Tinimomys, were originally thought to be closely related to archaic primates similar to Plesiolestes, Palaechthon, and Palenochtha, and were therefore referred to the Paromomyidae (sensu lato; Szalay, 1973, 1974). Subsequent studies essentially agreed with these proposed relationships, but classified Micromomys and Tinimomys in the family Microsyopidae (Bown and Rose,

110

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

Institutional abbreviations CM

Carnegie Museum of Natural History, Pittsburgh, Pennsylvania, U.S.A. UALVP University of Alberta Laboratory for Vertebrate Paleontology, Edmonton, Alberta, Canada UCM University of Colorado Museum, Boulder, Colorado, U.S.A. UCMP University of California Museum of Paleontology, Berkeley, California, U.S.A. UM University of Michigan Museum of Paleontology, Ann Arbor, Michigan, U.S.A. USNM National Museum of Natural History, Smithsonian Institution, Washington, D.C., U.S.A. USGS United States Geological Survey, Denver, Colorado, U.S.A. UW Geological Museum, University of Wyoming, Laramie, Wyoming, U.S.A. YPM-PU Princeton collection at Yale Peabody Museum, New Haven, Connecticut, U.S.A.

1976; Krause, 1978; Fox, 1984; Gunnell, 1989; but see Gingerich, 1976; Szalay and Delson, 1979). Fox (1984) suggested that the oldest known micromomyid, Micromomys fremdi, was a microsyopid most closely related to Purgatorius and Palenochtha minor. Additional evidence supported classification of micromomyids among the most primitive plesiadapiforms including the oldest and most primitive taxon Purgatorius (e.g., Beard and Houde, 1989; Van Valen, 1994; Silcox, 2001). Recent observations that tarsals of Purgatorius are uniquely similar to those of micromomyids also suggest that micromomyids are among the most basal primates (Chester et al., 2012).

Figure 1. Map of fossil localities where micromomyid plesiadapiforms have been found. (A) UADW-2, Paskapoo Formation, Alberta (Foxomomys fremdi); (B) UAR2a and UAR2, Roche Percee local fauna, Ravenscrag Formation, Saskatchewan (F. vossae); (C) Polecat Bench region, Bighorn Basin, Wyoming: Silver Coulee beds, Fort Union Formation: Princeton Quarry (Micromomys silvercouleei), Y2K Quarry (Dryomomys millennius), and Schaff Quarry (F. gunnelli); and Sand Coulee localities: Fort Union and Willwood Formations (D. willwoodensis, Chalicomomys antelucanus, Tinimomys graybulliensis, D. szalayi); (D) No Water Creek area, Willwood Formation, Bighorn Basin, Wyoming (T. graybulliensis); (E) Powder River Basin localities, Wasatch Formation, Wyoming (C. antelucanus, T. graybulliensis, ‘Myrmekomomys loomisi’); (F) Big Multi Quarry, Washakie Basin, Wyoming (D. dulcifer and T. tribos).

Along with other plesiadapiforms, micromomyids have traditionally been recognized as being more closely related to crown clade Primates than to other euarchontan mammals (i.e., Scandentia and Dermoptera). In contrast, Beard (1989, 1990, 1993a,b) suggested that micromomyid and paromomyid plesiadapiforms share unique postcranial features associated with mitten-gliding in extant flying lemurs and classified these extinct families in Dermoptera. More recent discoveries of partial skeletons of paromomyids and micromomyids, coupled with new phylogenetic analyses using morphological data from the dentition, cranium, and postcranium, do not support this hypothesis and instead suggest that they are primitive stem primates (Bloch et al., 2007; Silcox et al., 2010; Fig. 2) that did not glide (Boyer and Bloch, 2008). Micromomyids have a reduced dental formula (2.1.3.3/1.1.2-3.3) suggesting that in this respect they are not as primitive as some other plesiadapiforms, such as Purgatorius janisae (lower dental formula: 3.1.4.3). Nevertheless, with complete dentitions, partial crania, and most of the postcranium recovered, micromomyids are the best known of the most primitive stem primates (Fig. 2). Micromomyidae is represented by 11 species that span approximately five million years. A clear understanding of relationships within Micromomyidae allows for a better understanding of what might be considered primitive not only for the clade, but for early primates in general. As primitive euarchontans (and likely stem primates), micromomyids share a relatively recent common ancestry with euprimates (crown clade primates), were arboreal in some ways like the earliest euprimates, overlap temporally with the first documented euprimates, were fairly similar in size to some of the earliest known euprimates (e.g., Altanius, Altiatlasius, Teilhardina), and had similar dental characteristics such as lowcrowned molars with bunodont cusps. A revised phylogeny and better understanding of the ecological significance of the distinctive morphological features of micromomyids may also provide a clearer view of the ecological changes characterizing the evolution of the earliest euprimates.

Figure 2. Hypotheses of evolutionary relationships of micromomyids and other eutherian mammals. (A) Simplified resulting single-most-parsimonious cladogram modified from Bloch et al. (2007) with micromomyids as the most basal primates other than Purgatorius. (B) Simplified resulting strict consensus cladogram modified from Silcox et al. (2010) with a clade consisting of micromomyids and microsyopids as the most basal primates other than Purgatorius. Note that these analyses support the supraordinal group Euarchonta with Sundatheria (Scandentia and Dermoptera) as the sister taxon to Primates, and plesiadapiforms as stem primates. Micromomyid image modified from Bloch et al. (2007).

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

Eleven species of micromomyids have been classified in five genera (Micromomys, Tinimomys, Chalicomomys, Myrmekomomys, Dryomomys). Species-level alpha taxonomy of micromomyids is complicated by the fact that four of the 12 proposed species are only represented by their respective holotype, and three of the 12 species proposed are only each represented by two specimens. It has been suggested that Micromomyidae suffers from oversplitting, absent an understanding of intraspecific morphological variability (e.g., Rose and Bown, 1996). Here we report 54 new specimens of early Eocene Tinimomys graybulliensis from the Clarks Fork Basin, Wyoming, which add to the largest sample of any micromomyid species from one fossil locality (84 dental specimens, MNI ¼ 14). Newly documented morphological variability allows for more accurate diagnoses of micromomyid taxa. Additional descriptions of new micromomyid specimens from the Clarks Fork and Powder River Basins, Wyoming, including the first detailed dental descriptions of Dryomomys szalayi and a cladistic analysis of dental and gnathic characteristics that includes all recognized species of micromomyids, allows for a systematic revision of this family. Materials and methods All species and most specimens previously attributed to Micromomyidae were analyzed in this study (see Appendix A). New specimens including a premolar of Dryomomys sp., five specimens of Chalicomomys antelucanus, and 14 Clarkforkian specimens of T. graybulliensis are described. Fifty-four new early Eocene (Wa-1) specimens of T. graybulliensis are reported from UM locality SC-4. These specimens were previously recovered by etching freshwater limestone blocks from the Fort Union and Willwood Formations of the Clarks Fork Basin, Park County, Wyoming (see Bloch, 2001; Bloch and Boyer, 2001). Eight new specimens of T. graybulliensis from the Wasatch Formation of the Powder River Basin, Wyoming, are also described.

111

Specimens were micro-CT scanned at 6e10 microns on a Scanco Medical mCT 35 machine at Yale University. Three-dimensional digital reconstructions were created and measured using Avizo 6 software (http://www.vsg3d.com/avizo). In the rare case that a specimen such as a holotype could not be acquired on loan for CT scanning, either a cast was scanned, or the specimen or a cast was measured using digital calipers under a microscope. Measurements include maximum mesiodistal length (L), maximum buccolingual width (W), and P4 crown height (H) (see Fig. 3; Appendix A). Digital photographs were taken using a Visionary Digital (Palmyra, Virginia) setup with Canon 40D and 5D cameras at the Florida Museum of Natural History. To assess relationships among species of micromomyids, a matrix including 13 taxa and 28 dental and gnathic characters (see Appendix B) was assembled in MacClade 4.06 (Maddison and Maddison, 2003). Cranial and postcranial characters were not included because dentally associated cranial and postcranial elements are only known for two species (T. graybulliensis and D. szalayi). Cladistic analysis included the 11 species of micromomyids recognized in this study with Ptilocercus lowii and P. janisae as outgroups. P. lowii is the most basally divergent extant treeshrew based on morphological and molecular evidence (e.g., Sargis, 2001, 2002a,b; Olson et al., 2005; Roberts et al., 2011) and was used to root the analysis. Purgatorius was chosen because it is clearly outside the ingroup and it is the most primitive plesiadapiform. The species P. janisae was selected because it is the best-known species of Purgatorius and can be scored for 25 of 28 characters in this analysis. Although Purgatorius coracis (Fox and Scott, 2011) may represent the most primitive species of Purgatorius known, only 12 of 28 characters in this analysis can be scored for this species, and these 12 characters would be scored exactly as those for P. janisae. Similarly, Purgatorius ceratops is only represented by one isolated lower molar, can only be scored for four characters, and is not considered here. Purgatorius unio was also considered (with

Figure 3. Dental and gnathic measurements of micromomyid specimens recorded using Avizo 6 software on three-dimensional micro-CT scan reconstructions or using digital calipers under a microscope (following in part Rose, 1975; Bloch and Gingerich, 1998). Micro-CT scan images of Dryomomys szalayi (UM 41870) dentition captured using Avizo 6: (A) I2eM3 in occlusal view; (B) I1 in occlusal (left) and buccal (right) views; (C) I1eM3 in occlusal view; (D) I1eM3 and portion of dentary in buccal view; and (E) P4 in lingual view. Length (L) measures maximum mesiodistal dimensions and width (W) measures maximum crown dimensions perpendicular to length. Height (H) of incisors measures from the base of the enameledentin junction to the apex of the crown, and P4 H measures lingual side between mesial and distal roots to the apex of the crown (as micromomyid P4s can be moderately exodaenodont on buccal side). Mandibular depth (MD) measures below distal root of P4.

112

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

Purgatorius titusi viewed as a junior synonym of P. unio) because it may be more primitive than P. janisae. Again, fewer characters could be scored (17 of 28) than for P. janisae, and only one character in this analysis (P4 paraconid) could potentially be scored differently for P. unio. An analysis was run with Ptilocercus and P. unio as outgroup taxa, and the tree topology remained the same. Species of Purgatorius other than P. janisae are not included in the analysis to minimize the amount of missing data present in the matrix. A branch and bound search was conducted in PAUP* 4.0b10 for Macintosh PPC (Swofford, 2003). All characters were treated as unordered, no characters were weighted, and multistate characters were interpreted as polymorphic. Most of the characters used in this cladistic analysis were created for this study, and several characters were modified from Silcox (2001). Tooth homology and lower dental formula of the Micromomyidae Krause (1978) documented the presence of five alveoli between I1eP4 in Foxomomys vossae (lower dental formula of 2.1.3.3; Ti-4). He suggested the presence of a double-rooted P3, a singlerooted P2, a single-rooted C1, and a single-rooted I2. Based on more complete material of Foxomomys fremdi (Ti-3), and by comparison to Purgatorius (Pu-2-3; Clemens, 1974), the lower dental formula of F. fremdi and by inference F. vossae, was interpreted by Fox (1984) to have been 1.1.3.3, with a double-rooted P2. This interpretation is supported through the comparison of the mesial part of the dentary of P. janisae (Pu-2-3; Clemens, 2004). Recently published Foxomomys gunnelli (Ti-5a; Secord, 2008) has an identical inferred dental formula to both F. vossae and F. fremdi, also with an apparent double-rooted P2 in separate alveoli (five alveoli between I1eP4).

Beard and Houde (1989) cited the double-rooted condition of P2 in F. fremdi as a character that can be used to distinguish the subtribe Micromomyina (i.e., Foxomomys and Micromomys silvercouleei) from the subtribe Tinimomyina (Tinimomys and Chalicomomys). However, Secord (2008) noted that M. silvercouleei appears to have a single-rooted P2 like that of Chalicomomys. Here, using micro-CT scans, we show that M. silvercouleei, the type species of Micromomys known from Princeton Quarry (Ti-5a), and C. antelucanus (Wa-1; Beard and Houde, 1989) have only four alveoli between I1eP4 (Fig. 4). The two distal-most alveoli in M. silvercouleei are interpreted here as those for a double-rooted P3 as in the holotype of C. antelucanus. Mesially, the next alveolus would have contained a single-rooted P2. The mesial-most postincisor alveolus appeared to have contained a slightly procumbent, single-rooted canine in both taxa. Secord (2008) further suggested that the P2 alveolus in the holotype C. antelucanus may be relatively smaller than that of M. silvercouleei and may represent a more progressive condition. Though this observation is correct for the holotype of C. antelucanus, the P2 alveolus in a new dentary of C. antelucanus (UM 512221; Fig. 4) is larger than that of the holotype, and as large as that of M. silvercouleei. D. szalayi (Cf-3) is similar to M. silvercouleei and C. antelucanus in the number, size, and position of alveoli between I1eP4 (Fig. 3D). Thus, we interpret the lower dental formula of D. szalayi to also be 1.1.3.3. In contrast, T. graybulliensis (Cf-1 through Wa-2) has a double-rooted P3, a single-rooted canine, and has lost its P2 completely (lower dental formula: 1.1.2.3). We suggest that there is a gradual reduction in P2 size from a distinct double-rooted tooth in F. fremdi, F. vossae, and F. gunnelli, to a reduced single-rooted crown in M. silvercouleei, C. antelucanus, and D. szalayi, to complete loss of the tooth in T. graybulliensis. This interpretation lends further support to the hypothesized homology of mesial-most post-incisor

Figure 4. Comparison of lower dental formula between Micromomys silvercouleei (AeC) and Chalicomomys antelucanus (DeF). Micro-CT slice of M. silvercouleei, YPM-PU 17676, R dentary P4, M2, sectioned along sagittal plane and portrayed in buccal view (A), and three dimensional scan images of YPM-PU 17676 were captured using Avizo 6 in buccal view (B), and occlusal view (C). C. antelucanus, USNM 512221, L dentary M1e2, sectioned along sagittal plane and portrayed in buccal view (D), and three dimensional scan images of USNM 512221 were captured using Avizo 6 in buccal view (E), and occlusal view (F). White lines (C, F) indicate location of CT slice shown in A and D, respectively. Scale bar ¼ 1 mm.

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

tooth position as ‘C1’ rather than ‘P2’ in Tinimomys. Neither Dryomomys willwoodensis nor Dryomomys millennius are preserved well enough to reconstruct the dental formula mesial to the double-rooted P3. Systematic paleontology Order: Primates Linnaeus, 1758 Family: Micromomyidae Szalay, 1974 Type genus: Micromomys Szalay, 1973 Included genera: Micromomys, Foxomomys, Dryomomys, Chalicomomys, Tinimomys. Diagnosis: Differs from all other plesiadapiforms in having P4 with a trenchant paracristid. Further differs from all other plesiadapiforms except some species of picromomyids, uintasoricine microsyopids, and toliapinids in being very small (M1 length ¼ w1.00e1.25 mm). Further differs from Purgatorius in having one lower incisor (instead of three), lacking P1, and having a relatively larger P4 than M1. Further differs from palaechthonids in lacking a pronounced postprotocingulum on upper molars. Further differs from picromomyids in having a smaller P4 talonid with a hypoflexid and more lingually oriented cristid obliqua, and lower molars with less mesiodistally compressed trigonids compared to overall length of the crown. Further differs from microsyopids in lacking a lanceolate I1, lacking closely approximated entoconid and hypoconulid on lower molars, and further differs from uintasoricine microsyopids in having a double-rooted P3 with the mesial and distal roots clearly separated, lower molars with distinct paraconids, P4 with a metacone, and unreduced upper and lower third molars. Further differs from toliapinids in having lower molars with a weaker postmetacristid. Discussion: Szalay (1974) erected Micromomyini and placed M. silvercouleei and T. graybulliensis in this tribe within Paromomyidae. Gunnell (1989) later elevated Micromomyini to the subfamily Micromomyinae within Microsyopidae and provided a formal diagnosis. Beard (1989) elevated Micromomyidae to a family level in his doctoral dissertation, but did not provide a formal diagnosis for the new rank there or in subsequent publications (e.g., Beard, 1993a,b; Rose et al., 1993). Therefore, the differential diagnosis presented here is the first diagnosis for the family Micromomyidae. Foxomomys gen. nov. (Micromomys in part Krause, 1978; Fox, 1984; Secord, 2008) Type species: F. fremdi (Fox, 1984) Included species: F. fremdi (Fox, 1984), F. vossae (Krause, 1978), and F. gunnelli (Secord, 2008). Known distribution: Late Paleocene of North America (middlelate Tiffanian NALMA of Saskatchewan (Ti-3) and Alberta (Ti-4), Canada; late Tiffanian (Ti-5a) of the Bighorn Basin, Wyoming). Diagnosis: Differs from all other micromomyids in having double-rooted P2, narrower P4 relative to length with a cristid obliqua that climbs the postvallid for a short distance, narrower lower molars relative to length with a protoconid that is considerably taller than the metaconid (rather than subequal in height), smaller P3 relative to size of upper molars, P4 with a more distobuccally positioned metacone relative to the paracone on a more obliquely oriented metacrista, more transversely elongate, triangular-shaped, upper molars with a deeper ectoflexus, and more acute molar cusps. Further differs from all micromomyids except Chalicomomys in having a relatively more transversely elongate M2 compared with M1. Etymology: Named for Richard C. Fox, who described the type species, and in recognition of his contributions to our understanding of plesiadapiforms and other Paleogene mammals. Suffix

113

omomys by analogy with Micromomys and to suggest primate affinities of the genus. Discussion: The genus Foxomomys is erected to reflect differences between the oldest micromomyid species, F. fremdi, and other micromomyids, including the type species of Micromomys, M. silvercouleei. Two poorly known Tiffanian species, F. vossae and F. gunnelli, are tentatively assigned to this genus. The only characters that are clearly unique to all three species of Foxomomys among micromomyids are also present in Purgatorius, and therefore appear to be plesiomorphic. Consequently, like the genus Purgatorius, it is possible that Foxomomys may represent a paraphyletic grouping. However, unlike the previous state of the genus Micromomys, there is no evidence that Foxomomys is polyphyletic, and current evidence suggests that M. silvercouleei is more closely related to other micromomyids than to F. fremdi, F. vossae, and F. gunnelli (see below). F. fremdi differs from Purgatorius and other micromomyids in having a relatively higher protoconid than metaconid on its lower molars. F. vossae appears similar to F. fremdi in this regard, although it is only known from M1, and molars are not known for F. gunnelli. Additionally, F. fremdi and F. gunnelli are unique compared with Purgatorius and other micromomyids in the position of the mesial mental foramen under the distal root of P2 (not known for F. vossae). F. vossae and F. gunnelli remain very poorly known, and additional specimens are needed to evaluate whether these characters are synapomorphies for a Foxomomys clade. This grouping is in agreement with the previous suggestion that it may be appropriate to place F. fremdi and F. gunnelli in a separate genus because M. silvercouleei, C. antelucanus, and D. millennius appear more derived (Secord, 2008). F. fremdi (Fox, 1984) (Fig. 5AeE) Holotype: UA 21010, L partial dentary I1 alveolus, crowns of C1e M3. Known distribution: Middle Tiffanian (Ti-3) UADW-2, Paskapoo Formation, Alberta, Canada. Emended diagnosis: Differs from other species of Foxomomys in being slightly larger. Further differs from F. vossae in having P4 with a uniformly sloping mesiodorsal margin in buccal view and one or two (rather than three poorly defined) talonid cusps. Further differs from F. gunnelli in having a diastema between C1 and P2, and a narrower P4 relative to length. Discussion: Scott (2008) reported several new specimens of F. fremdi from the type locality. He described what appears to be the first upper canine known for the species, which is larger than P2, unlike the condition present in D. szalayi. Scott also described a new specimen that has a slightly larger and more transverse P3e4 than the only previously known specimen for which these tooth loci are well preserved. F. vossae (Krause, 1978) (Fig. 5FeH) Holotype: UA 9273, L partial dentary P4eM1. Known distribution: Late Tiffanian (Ti-4) UAR2, Ravenscrag Formation, Saskatchewan, Canada. Emended diagnosis: Differs from other species of Foxomomys in having P4 with a more distinctive break in slope along the mesiodorsal margin in buccal view, and three poorly defined talonid cusps rather than one or two. Further differs from F. fremdi in being slightly smaller. Further differs from F. gunnelli in having a narrower P4 relative to length. F. gunnelli (Secord, 2008) (Fig. 5IeK) Holotype: UM 77528, L partial dentary P4. Known distribution: Late Tiffanian (Ti-5a), Schaff Quarry, northern Bighorn Basin, Wyoming. Emended diagnosis: Differs from other species of Foxomomys in having a smaller and wider P4 relative to length. Further differs from F. fremdi in lacking a diastema between C1 and P2. Further differs from F. vossae in having P4 and a uniformly sloping

114

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

Figure 5. Micro-CT scan images of Foxomomys captured using Avizo 6: F. fremdi, UA 21015, L maxilla P2eM3 in buccal (A) and occlusal (B) views. F. fremdi, UA 21011, R dentary C1, P3e M3 (reversed) in buccal (C), lingual (D), and occlusal (E) views. F. vossae, UA 9151, L P4 in buccal (F), lingual (G), and occlusal (H) views. F. gunnelli, UM 77528, L dentary P4, in buccal (I), lingual (J), and occlusal (K) views. Scale bar ¼ 1 mm.

mesiodorsal margin in buccal view and one (rather than three poorly defined) talonid cusp(s). Micromomys Szalay, 1973 Type species: M. silvercouleei Szalay, 1973 Diagnosis: As for the type species. M. silvercouleei Szalay, 1973 (Figs. 4AeC, 6) Holotype: YPM-PU 17676, R partial dentary P4 and M2, and alveoli for I1, C1, P2, P3, M1, M3. Known distribution: Late Paleocene of North America (late Tiffanian (Ti-5a), Princeton Quarry, northern Bighorn Basin, Wyoming). Emended diagnosis: Differs from all other micromomyids in having P4 with mesiodistally short trigonid that is relatively taller than the talonid. Further differs from Chalicomomys and Dryomomys in having a dentary with the mesial mental foramen below the single-rooted P2 (rather than below mesial root of P3). Further differs from Dryomomys in having less enlarged P4 relative to M2, relatively larger M2 paraconid, and mediolaterally narrower mandibular corpus. Further differs from Foxomomys in having single-rooted P2, wider P4 relative to length with a cristid obliqua that does not climb the postvallid, and rounder cusps on M2 with

protoconid and metacone subequal in height. Further differs from Tinimomys in retaining P2, less bunodont molar cusps, and a dentary that is more uniform in depth. Discussion: M. silvercouleei was the first species of micromomyid described. This species was and still is only represented by the holotype, a dentary preserving two broken teeth (P4 and M2). P4 morphology appears uniquely derived, however, the broken and distorted nature of the teeth makes it somewhat difficult to assess whether there are any synapomorphies uniting M. silvercouleei with any other species of micromomyid. An interpretation of the original, undistorted morphology of these teeth has been illustrated (Szalay, 1973; Szalay and Delson, 1979), but no attempts have been made to physically prepare these teeth in fear that the holotype would be further damaged. In order to gain confidence about the true morphology of these teeth, the holotype was micro-CT scanned, and the teeth were digitally disassembled along breaks and reconstructed using Avizo 6 software (Fig. 6). Results suggest that certain natural breaks are filled with matrix, making P4 appear taller than it actually was. However, M. silvercouleei can still be diagnosed from all other micromomyids in having P4 with a mesiodistally short trigonid that is relatively taller than the talonid. Micro-CT scans also

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

115

Figure 6. Micro-CT scan images of the holotype and only known specimen of Micromomys silvercouleei, YPM-PU 17676, R dentary P4, M2, captured using Avizo 6: P4 was digitally separated into three parts along natural breaks (A), and reconstructed (B, distal; C, lingual; D, buccal; E, occlusal views) providing a more accurate view of P4 morphology than the current condition of the specimen (F, distal; G, lingual, H, buccal; I, occlusal views). YPM-PU 17676 in buccal (J) view. M2 was also digitally separated into three parts along natural breaks (K) and reconstructed (L, buccal, M, lingual; N, occlusal views) providing a more accurate view of M2 morphology than the current condition of the specimen (O, buccal, P, lingual; Q, occlusal views). All digital reconstructions were created using Avizo 6 software. Scale bar ¼ 1 mm.

allowed us to confidently assess the lower dental formula of M. silvercouleei, which is 1.1.3.3, with a single-rooted P2 (see Fig. 4AeC). Chalicomomys Beard and Houde, 1989 Type species: C. antelucanus Beard and Houde, 1989 Diagnosis: As for the type species. C. antelucanus Beard and Houde, 1989 (Figs. 4DeF, 7, 8) Holotype: USNM 425589, L partial dentary P3 (broken), P4eM2, and alveoli for I1, C1, P2, and M3. Referred specimens: UCM locality 84126: UCM 54585, R partial dentary P4eM1, alveoli for M2 (Fig. 8DeF); UM locality SC-123: UM 76682, L partial dentary M2e3 (Fig. 8AeC); UM locality SC-4: USNM 512221, L dentary M1e2, and alveoli for I1, C1, P2e4, and M3 (Fig. 7KeM); USNM 512241, R M2 (Fig. 7I,J); USNM 512258, R P4 (Fig. 7E,F); USNM 512261, R partial maxilla M1 and alveoli for M2 (Fig. 7G,H); USNM 516587, L partial maxilla P2eM3 (Fig. 7C,D).

Locality and horizon: UCM 54585 was collected at Urruty’s Crossing (UCM locality 84126) in the Wasatch Formation, Powder River Basin, Wyoming (see Robinson, 1994). UM 76682 was collected at UM SC-123 in the Clarks Fork Basin, Wyoming (see Gunnell, 1989). All USNM specimens were recovered from early Eocene faunal zone Wa-1, Willwood Formation, Clarks Fork Basin, Park County, Wyoming. These specimens were etched from a freshwater limestone block that was collected near UM fossil locality SC-4 (see Beard and Houde, 1989 and references therein). Known distribution: Early Eocene of North America (Wasatchian (Wa-1) of the Bighorn Basin, Wyoming, and early Wasatchian of the Powder River Basin, Wyoming). Emended diagnosis: Differs from all micromomyids other than Foxomomys in having a relatively more transversely elongate M2 compared with M1. Differs from Foxomomys in having single-rooted P2, wider P4 relative to length, protoconid and metaconid subequal in

116

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

Figure 7. Micro-CT scan images of new specimens of Chalicomomys antelucanus (CeM) and the only previously known maxillary specimen (AeB) captured using Avizo 6: USNM 425588, R maxilla P3e4, M2e3, in buccal (A) and occlusal (B) views. USNM 516587, L maxilla P2eM3, in buccal (C) and occlusal (D) views. USNM 512258, R P4, in buccal (E) and occlusal (F) views. USNM 512241, R M2, in buccal (I) and occlusal (J) views. USNM 512261, R maxilla M1, in buccal (G) and occlusal (H) views. USNM 512221, L dentary M1e2, in buccal (K), lingual (L), and occlusal (M) views. Scale bar ¼ 1 mm.

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

117

Figure 8. Micro-CT scan images of Chalicomomys antelucanus captured using Avizo 6: UM 76682, L dentary M2e3, in buccal (A) lingual (B), and occlusal (C) views, and UCM 54585, R dentary P4eM1, in buccal (D) lingual (E), and occlusal (F) views. Scale bar ¼ 1 mm.

height on lower molars, relatively larger P3 compared with M1, and less pronounced ectoflexus and parastyle on upper molars. Further differs from F. fremdi in lacking diastemata between I1 and C1, and C1 and P2. Further differs from Micromomys in having a less tall P4 relative to length with a mesiodistally longer trigonid relative to width, and a mesial mental foramen under mesial root of P3 (rather than under P2). Further differs from Dryomomys in having relatively smaller P3e4 compared with lower molars, mediolaterally narrower mandibular corpus, relatively smaller P2e4 compared with upper molars, and P3 with smaller protocone situated on less inflated protocone lobe. Further differs from Tinimomys in retaining P2, having a P4 with a smaller talonid, lacking lingually continuous cingula with pericones and hypocones on P4eM2, having a less developed P3 protocone lobe, having more pronounced ectoflexus on upper molars, and having a dentary that is fairly uniform in depth. Description and comparison: Detailed descriptions of previously known tooth positions of C. antelucanus were published by Beard and Houde (1989). Here, based on five new specimens from the type locality, we describe new tooth positions (P2, M1) and document variation in previously described tooth positions. A summary of dental measurements of C. antelucanus specimens from UM SC-4 can be found in Table 1. The upper dentition of C. antelucanus was previously known only from USNM 425588 (Fig. 7A,B), a right partial maxilla with P3e 4 , M2e3 (Beard and Houde, 1989). USNM 516587, a left partial maxilla with P2eM3 (Fig. 7C,D), is the most complete maxillary specimen of C. antelucanus known to date. USNM 516587 is broken directly mesial to P2, so the presence of a diastema between P2 and the tooth mesial to it (presumably C1) cannot be evaluated. P2 is buccolingually compressed and double-rooted like all other micromomyids for which this tooth position is known. The crown is dominated by a paracone with a postparacrista that descends from the apex of the paracone and becomes less distinct near the base of

a small metastylar cusp. A smaller, yet distinct parastylar cuspule is present on the mesial portion of the crown and a faint lingual cingulum is present on the distal end of the tooth. The P2 of C. antelucanus is very similar to that of Tinimomys tribos (see Chester and Beard, 2012). T. graybulliensis (see Gunnell, 1989; Rose Table 1 Summary of dental measurements (in millimeters) for Chalicomomys antelucanus from SC-4. Tooth

N

OR

x

SD

CV

P3L P3W P4L P4W P4H M1L M1W M2L M2W M3L M3W MD P2L P2W P3L P3W P4L P4W M1L M1W M2L M2W M3L M3W

1 e 3 3 3 2 2 3 2 e e 2 1 1 2 2 3 3 2 2 3 3 2 2

0.91 e 1.29e1.35 0.90e0.91 0.94e1.08 1.01e1.07 0.78e0.79 0.98e1.06 0.81e0.83 e e 1.75e1.95 0.80 0.47 1.13e1.15 0.96e1.00 1.20e1.30 1.34e1.54 0.95e0.98 1.47e1.59 0.93e0.97 1.40e1.60 0.79e0.88 1.55e1.58

e e 1.33 0.91 0.99 1.04 0.79 1.01 0.82 e e 1.85 e e 1.14 0.98 1.25 1.47 0.97 1.53 0.96 1.52 0.84 1.57

e e 0.03 0.01 0.08 0.04 0.01 0.04 0.01 e e 0.14 e e 0.01 0.03 0.05 0.12 0.02 0.08 0.02 0.11 0.06 0.02

e e 2.42 0.64 7.89 4.08 0.90 4.11 1.72 e e 7.64 e e 1.24 2.89 4.04 7.84 2.20 5.55 2.41 6.96 7.62 1.36

L, length; W, width; H, height; MD, mandibular depth below distal root of P4; N, sample size; OR, observed range; x, mean; SD, standard deviation; CV, coefficient of variation.

118

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

et al., 1993), D. millennius (see Secord, 2008), and D. szalayi (see Fig. 3), differ from C. antelucanus and T. tribos in lacking a distinct parastylar cusp. D. millennius and D. szalayi further differ in being relatively larger and more bulbous (especially D. szalayi, which also has a more pronounced lingual cingulum that extends to the mesial aspect of the tooth). The only published P2 of F. fremdi (UA 21015; Fox, 1984) is too incomplete for comparisons, although a new complete specimen has been reported (Scott, 2008). The P3 and P4s of C. antelucanus reported here are very similar to those of the only previously known specimen (USNM 425588; Fig. 7A,B). The parastylar and metastylar regions of P3 are slightly more expanded in USNM 516587 (Fig. 7C,D), although not to the degree of that of Dryomomys dulcifer, D. millennius, or D. szalayi. Similarly, the protocone lobe of P3 of USNM 516587, although worn, is not quite as narrow as it appears in the only previously known specimen. The P4s of USNM 516587 and 512258 (Fig. 7E,F) are extremely similar to the previously known specimen. USNM 516587 also has a median crest between the pre- and postprotocristae that connects to the lingual base of the paracone, but differs from the other known specimens in having a second median crest that connects the postprotocrista to the midpoint between the lingual base of the paracone and metacone. This forms two sets of ‘wings’ that resemble the cristae of the molar conules, which occurs in other micromomyids and appears to be variable. The M1 of C. antelucanus was not previously known, and this tooth position is described here based on two specimens (USNM 516587, Fig. 7C,D; USNM 512261, Fig. 7G,H). M1 is smaller than P4, and slightly longer and narrower (buccolingually) than M2, giving it a less transverse appearance. The paracone is slightly larger than the metacone, and the postparacrista and premetacrista are positioned in a mesiodistally-oriented line between these cusps. The preparacrista and postmetacrista curve mesiobuccally and distobuccally, respectively, to join a continuous ectocingulum. The ectoflexus of M1 is not as pronounced as that of M2 in Chalicomomys, which is common in micromomyids. The M1 ectoflexus is not as pronounced as that of F. fremdi, but is similar to that of T. tribos in being more pronounced than that of T. graybulliensis or D. szalayi, which suggests that Chalicomomys is primitive in this regard. A parastyle is present and is about the same size as that of M2, but not as pronounced as that of M3 in Chalicomomys. The paraconule and metaconule are small, and have respective cristae. The conules are connected to the protocone by the pre- and postprotocristae, which outline the trigon basin. The area of the trigon basin increases from M1 to M3. The protocone is larger, yet approximately the height of the other two trigon cusps, and it is oriented mesiobuccally. Preand postcingula are present, but they do not connect to form a lingual cingulum as in Tinimomys. No hypocone or pericone is present. Although there appears to be a break in slope where a postprotocingulum would be present, this structure is not distinct. The two new M2s of C. antelucanus reported here (USNM 516587, Fig. 7C,D; USNM 512241, Fig. 7I,J) are very similar to the M2 described by Beard and Houde (1989). The diminutive cuspule buccal to metacone on the ectocingulum that was documented on the M2 of USNM 425588 (Beard and Houde, 1989) is not as distinct in these specimens, although the stylar region is similarly quite developed with a pronounced ectoflexus. Overall, M2 of C. antelucanus is quite similar to that of D. dulcifer (Chester and Beard, 2012). The new specimen of M3 (USNM 516587) is very similar to the previously described specimen, although it differs slightly in having a more pronounced metacone and metaconule. A new dentary of C. antelucanus (USNM 512221, Fig. 7KeM) preserves the first and second molars and the mesial alveoli better than any previously known specimen. This specimen confirms the interpretation of Beard and Houde (1989) that C. antelucanus had a

dental formula of 1.1.3.3 (with a single-rooted I1, C1, and P2, and double-rooted teeth distal to P2), and lacks diastemata between I1 and C1 and between C1 and P2. The P2 alveolus is larger than that of the holotype. Other than an extremely worn and partially broken M2 (USNM 425587), the crowns of M1 and M2 of C. antelucanus were previously only known from the holotype. USNM 512221 has M1 and M2 that are generally similar to those of the holotype, although the paraconid is slightly more lingually located on M2. Discussion: Nine specimens referred to C. antelucanus have been recovered from the type locality SC-4. A left dentary with M2e3 (UM 76682; Fig. 8AeC) from a different Wa-1 locality in the Clarks Fork Basin, SC-123, was originally attributed to ‘Micromomys’ willwoodensis (Gunnell, 1989), and was later referred to C. antelucanus when Chalicomomys was first described (Beard and Houde, 1989). Beard and Houde (1989) referred UM 76682 to C. antelucanus because its size, molar morphology (unknown for D. willwoodensis), and provenance (Wa-1) may be in better agreement with C. antelucanus. These authors suggested that lower molars with a metaconid and protoconid of subequal height was diagnostic of C. antelucanus, but this condition can also be found in M. silvercouleei, as well as D. szalayi, D. dulcifer, and D. millennius. Referral of UM 76682 is complicated by the fact that it only preserves M2, which is not particularly diagnostic, and M3, which is not otherwise known for C. antelucanus. Therefore, the best comparison is that of mandibular depth. The depth of the dentary below the distal alveolus of M2 in UM 76682 is 2.07 mm, whereas it is 2.56 mm in the holotype of D. willwoodensis and 2.42 mm in D. szalayi. The same measurement was taken for three specimens of C. antelucanus (USNM 425586, 1.96 mm; USNM 425587, 1.87 mm; USNM 512221, 1.91 mm), which are all more similar to that of UM 76682 than to D. willwoodensis. This suggests that UM 76682 belongs to C. antelucanus, and represents the only M3 known for this species. The M3 is similar to that of Dryomomys in having a hypoconulid that is more developed than that of F. fremdi, yet not as expanded and broad as in Tinimomys. A partial dentary with P4eM1 (UCM 54585; Fig. 8DeF) from the early Wasatchian of the Powder River Basin was attributed to D. willwoodensis (Robinson, 1994). The absolute size of P4, as well as the depth and mediolateral width of the dentary, is smaller than that of D. willwoodensis and more similar in size to that of C. antelucanus. Lower molars are not known for D. willwoodensis, although the size of the alveoli suggests that the relative size of P4 to M1 is quite large as in D. szalayi. It should also be noted that the M1 paraconid in UCM 54585 is somewhat reduced, which is a condition similar to that found in some species of Dryomomys. Overall, the relative size of P4 to M1 and absolute size of the dentary in UCM 54585 appears to be most similar to that of C. antelucanus and it is tentatively assigned to this species. Validity of the genus Chalicomomys: Beard and Houde (1989) suggested that Chalicomomys is useful as a separate genus because it serves as a morphologic intermediate between Micromomys sensu lato (i.e., Foxomomys and M. silvercouleei) and Tinimomys. Rose and Bown (1996) claimed that characters originally used by Beard and Houde (1989) to distinguish Chalicomomys from Micromomys do not validate generic separation. Secord (2008) also considered Chalicomomys a junior synonym of Micromomys and demonstrated that the original diagnosis of Chalicomomys suffers from using the most plesiomorphic and best-known species of ‘Micromomys,’ F. fremdi, instead of the type species, M. silvercouleei, for all comparisons. Our current understanding of the relationship between C. antelucanus and M. silvercouleei is complicated by the fragmentary nature of the only known specimen of M. silvercouleei. There are clear differences between the two taxa and similarities are best interpreted as plesiomorphic. Also, there is a three million year difference in age between

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

M. silvercouleei and C. antelucanus with no known intermediates. As such, we suggest that recognizing the genus Chalicomomys as a junior synonym of Micromomys is unneccesary especially given the arbitrary nature of ranks above the species level. Dryomomys Bloch et al., 2007 Type species: D. szalayi Bloch et al., 2007 Included species: D. szalayi (Bloch et al., 2007), D. willwoodensis (Rose and Bown, 1982), D. millennius (Secord, 2008), D. dulcifer (Chester and Beard, 2012). Known distribution: Late Paleocene to early Eocene of North America (Tiffanian (Ti-5b) to early Wasatchian (Wa-2) of the Bighorn Basin, Wyoming, and middle Clarkforkian (Cf-2) of Big Multi Quarry, Washakie Basin, Wyoming). Emended diagnosis: Differs from all other micromomyids in having large and inflated P2e4 and P3e4 relative to molars, and a mandibular condyle that is higher relative to the toothrow. Further differs from all other micromomyids except T. graybulliensis in having lingually expanded P3 with strong protocone and small distolingual basin bounded by the preprotocrista and postprotocingulum. Further differs from Foxomomys and M. silvercouleei in having a mesial mental foramen under the mesial root of P3. Further differs from Foxomomys and Chalicomomys in having a taller P4 relative to length and similarly sized M1 and M2 (rather than a relatively more transversely elongate M2 compared with M1). Further differs from Foxomomys in having single-rooted P2, and a longer and broader M3 hypoconulid (although not to the extent as that of Tinimomys). Further differs from Tinimomys in having a P2, having a taller P4 relative to length, and lacking lingually continuous cingula with no distinct pericone or hypocone on P4eM2. Further differs from T. graybulliensis in having slightly narrower I1 with less developed mediocone and no lateroconule, and in lacking P3 metacone. Discussion: Two species previously allocated to other micromomyid genera are attributed to Dryomomys here. ‘Micromomys’ willwoodensis and ‘M.’ millennius are referred to Dryomomys primarily based on the relatively large size of their known premolar tooth positions. D. willwoodensis was first attributed to Micromomys when micromomyids were poorly known (Rose and Bown, 1982). D. willwoodensis was then referred to Chalicomomys because its P4 was viewed as more similar to that of C. antelucanus than to P4s of known species of Micromomys, and like C. antelucanus, it was also only known from the early Eocene (Beard and Houde, 1989). As recognized by Rose and Bown (1982), the P4 of D. willwoodensis was considerably larger than that of any other micromomyid known. The only P4 of similar size that has been described since is that of D. szalayi (Bloch et al., 2007). D. millennius was attributed to Micromomys by Secord (2008), but shares many premolar characteristics that were originally considered autapomorphies of D. szalayi by Bloch et al. (2007). Although it is possible that more than one lineage of micromomyids independently evolved relatively large premolars, a Dryomomys clade is supported in the phylogenetic analysis presented below. The seemingly progressive increase in premolar size from the late Tiffanian D. millennius to D. dulcifer to D. szalayi to D. willwoodensis in the early Wasatchian may well illustrate the evolution of premolar specializations in this group of micromomyids (Figs. 9 and 10). The premolar specializations of Dryomomys are similar to those of certain early fossil euprimates, especially those of omomyids. The relatively pronounced lingual expansion of P3e4 found in D. szalayi is similar to that of omomyids such as Absarokius abotti. D. szalayi is also similar to some omomyids such as Uintanius ameghini in having relatively large upper and lower premolars (see Szalay, 1976; Szalay and Delson, 1979). Though similarities of premolar specializations between Dryomomys and these omomyids

119

Figure 9. Micro-CT scan images of Dryomomys P3e4 captured using Avizo 6. D. millennius, UM 110140, L P3e4 cropped from L maxilla P2eM1, in occlusal (A) view. D. dulcifer, CM 71796, L P3 (cast), and CM 72149, L P4 (cast), in occlusal (B) view. D. szalayi, UM 41870, R P3e4 (reversed) cropped from palate in occlusal (C) view. Dryomomys sp., UM 39849, R P4 (reversed) in buccal (D) and occlusal (E) views. Scale bar ¼ 1 mm.

are clearly convergent, this observation may indicate that these small early euarchontans were eating and or processing comparable food in similar ways. D. szalayi Bloch et al. 2007 (Figs. 3, 9C, 10GeI, 11e15GeI) Holotype: UM 41870, partial cranium with L and R I1eM3 (R C1 absent), L dentary I1eM3 (M2 absent), R dentary I1eM3, and partial postcranial skeleton. Known distribution: Late Clarkforkian (Cf-3), UM locality SC327, lower Willwood Formation, Clarks Fork Basin, Wyoming. Emended diagnosis: Differs from all other species of Dryomomys in having relatively larger P2e4 compared with upper molars, wider P3 relative to length with larger protocone, more pronounced protocone lobe, and larger distolingual basin, and P4 that is wider compared with the molars with a more lingually dipping protocone lobe. Also differs from all species of Dryomomys in having a narrower P3 relative to length with more developed talonid basin, M1 that lacks a distinct paraconid, and M2 with relatively small paraconid. Further differs from D. millennius in having larger P4 relative to size of lower molars, M1e2 with more centrally situated paraconids, and P4 with the apex of the protocone more closely situated to that of the paracone and metacone. Further differs from D. willwoodensis in having P4 with broader talonid basin and more ventrally distended distobuccal lobe. Description: There are clearly two single-rooted incisors in the premaxillae of D. szalayi (Fig. 11). Distal to that, the mesial-most tooth in the maxilla is double-rooted, followed by P2e4 and M1e3 (Fig. 12). We interpret the mesial-most tooth in the maxilla as a

120

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

Figure 10. Micro-CT scan images of Dryomomys lower teeth captured using Avizo 6: D. millennius, UM 109659, R dentary P3eM2, in buccal (A), lingual (B), and occlusal (C) views. D. dulcifer, CM 71657, R P3 (cast), and CM 69990, L P4 (cast, reversed), in buccal (D), lingual (E), and occlusal (F) views. D. szalayi UM 41870, cropped portion of R dentary I1eM3 illustrating P2eM3 in buccal (G), lingual (H), and occlusal (I) views. D. willwoodensis, YPM-PU 17732, L dentary P4 (reversed), in buccal (J), lingual (K), and occlusal (L) views. Scale bar ¼ 1 mm.

canine, with P1 absent and suggest that the upper dental formula of D. szalayi is 2.1.3.3. The crowns of I1e2 (Fig. 11) are complete in the right and left premaxillae. The crown of I1 is vaguely ‘mitten-shaped’ (although, in lateral perspective, it appears ‘hook-shaped’) with a subdivision of the distal end into a strong anterocone and a weak mediocone, the latter almost completely defined by a shallow furrow separating the two cusps, and the presence of a short internal crest running proximally from its apex. The ‘thumb’ of the mittenshaped crown has a single, small (but distinct) cusp. The crown of I2 is shorter than that of I1, laterally compressed, and slightly recurved at its apex. It has a single cusp (anterocone) at its apex, with a weak medial crest, but no mediocone. There is a swelling at the base of the tooth, although there are no cusps present. The crown of C1 (Fig. 12) is broken, missing most of the distal half of the crown. It is smaller than P2, double-rooted, longer than wide with a mesially situated apical cusp, and lacks a lingual cingulum. There is a swelling at the base of the crown, giving the impression that it may have been similar in that respect to the morphology of P2. The mesial root is considerably smaller in diameter than the distal root. P2 (Fig. 12) has two roots and is separated from C1 by a short diastema. The crown is longer than wide with a mesially situated apical cusp and a strong lingual cingulum that extends to an incipiently basined distal heel. In comparable morphology, the mesial bulge is generally similar to that of the canine. The P3 (Fig. 12) has three roots. In occlusal view, the crown of P3 is generally triangular in outline, with the buccal margin of the crown slightly convex, and the mesio- and distolingual margins strongly concave at the midline. The crown of P3 is dominated by a large cusp (paracone) that is situated on the central part of the buccal margin. P3 has a strong postparacrista, but lacks any evidence of a metacone. The postparacrista descends in a straight line distally from the apex of the paracone to the distal margin of the crown, where it swings sharply around the distobuccal corner of the crown and continues mesially along the buccal margin as a cingulum for about one-third of the length. The mesial face of the

paracone is mostly smooth, although a faint trace of a preparacrista descends from the apex of the paracone mesially for a very short distance. Though there is no obvious parastyle, a strong cingulum is present on the mesiobuccal corner of the tooth that extends lingually to join the protocone. The crown of P3 is lingually expanded with a strong protocone and a small distolingual basin bounded by the preprotocrista and postprotocingulum. The P4 (Fig. 12) has three roots, is semimolariform, is distinctly longer buccally than lingually, and is concave at the midline, at the level of the protocone, on the mesial and distal margins of the crown (much more so on the mesial margin). The crown of P4 is the widest in the toothrow. The buccal aspect of the crown of P4 is nearly equal in length to that of P3, and considerably longer than that of M1. The crown of P4 has three well-developed buccal cusps. The paracone is the largest cusp and is situated slightly distal to the center of the buccal side of the tooth. The metacone is more worn but clearly smaller than the paracone, and is connate with the paracone for most of its height. A shallow valley separates the paracone and metacone buccally and lingually. The parastyle is located on a large, mesially-projecting parastylar lobe on the buccal end of the tooth. The parastyle would very likely have been a distinct cusp, perhaps even larger than the metacone, although this observation is mostly by inference as much of the mesial portion of the parastylar lobe is strongly worn. The parastyle is separated from the paracone by a well-developed deep valley that is defined by two crests: the first descends the parastyle lingually and is joined to the protocone by a strong preprotocrista, and the second descends the worn parastyle buccally to join the mesial portion of an ectocingulum before it is interrupted by the base of the paracone. A straight line can be drawn from a faint preparacrista across to a short postparacrista that likely connects to the closely appressed metacone by a very short premetacrista, although this is difficult to see due to wear especially on the metacone. The postmetacrista curves distobuccally to connect with the distal ectocingulum, which is also interrupted by the base of the paracone. The protocone is situated very close to the paracone and metacone, and is lingual and slightly distal to the paracone, resulting in a very small trigon basin. The

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

Figure 11. Dryomomys szalayi (holotype: UM 41870) L premaxilla I1e2 in buccal (A), lingual (B), and occlusal (C) views.

postprotocrista extends posterobuccally to about half the distance between the protocone and metacone, at which point it dips down around the distal margin of the crown to the interstitial facet between P4 and M1. While no distinct metaconule is present, a small

121

crest extends to the base of the paracone from this point forming part of the margin of the trigon basin, similar to the morphology of a premetaconule crista. Likewise, a small median crest, similar to the morphology of a postparaconule crista (although no paraconule is present), extends from the preprotocrista to the base of the paracone and forms part of the margin of the trigon basin. The preprotocrista extends from the protocone to a crest connected to the parastyle. A large lingually sloping protocone lobe extends from the apex of the protocone to the lingual margin of the crown. The lobe is narrow and oval shaped, lacking a flat lingual margin. Welldeveloped pre- and postcingula are present, but not lingually continuous. Both cingula climb from the base of the protocone, with the precingulum ending part way up and the postcingulum extending all the way to the apex. These crests form two incipient basins, or ‘furrows’ (Secord, 2008) with the mesial longer and shallower than the distal one. The upper molars of D. szalayi are generally similar to those of other micromomyids (Fig. 12). The lingual half of the M1 crown is narrower than the buccal half and shifted slightly distally relative to the midline. The M1 trigon cusps are all fairly similar in height and size. The largest cusp is the protocone, which faces mesially. The second largest cusp is the paracone and the smallest trigon cusp is the metacone. Like that of P4, the metacone is the most worn cusp on M1. The parastyle is weakly developed. Like that of P4, the preand postparacristae and premetacrista are worn but appear to form a straight line across the tooth mesiodistally. The postmetacrista curves distobuccally and connects with the ectocingulum. The ectocingulum is buccally continuous, and the ectoflexus is not well developed. The paraconule, metaconule, and respective cristae are well developed, although are short with the conules closely appressed to the paracone and metacone. The pre- and postprotocristae connect the protocone to conules and surround a moderately expanded trigon basin. Pre- and postcingula are well defined, but not lingually continuous, instead they briefly climb the protocone before terminating. There is no obvious development of a pericone or hypocone. The crown of M2 of D. szalayi is very similar to that of M1 in size and morphology (Fig. 12). M2 differs from M1 in being slightly narrower mesiodistally on the buccal side, resulting in a more quadrate tooth. The ectocingulum on M2 is slightly interrupted by the buccal base of the paracone, and the ectoflexus is more pronounced than that of M1. There is less development of a hypocone swelling on M2 relative to that of M1. The M3 of D. szalayi has a relatively smaller metacone with no postmetacrista and a relatively larger parastyle than M1 and M2 (Fig. 12). The ectocingulum is interrupted by the paracone and forms a minor buccal extension where a mesostyle would be located. We interpret the lower dental formula of D. szalayi to be 1.1.3.3 (see discussion on tooth homology above). The crown of I1 is mostly preserved on the left side (the tip is missing) and completely preserved on the right side (Figs. 13 and 14). The crown of I1 is procumbent, long, highly compressed mediolaterally, and culminates in a sharp, tapering tip. A thin but distinct dorsal crest (margocristid; Gingerich, 1976) extends from the tip to the base of the crown, with no margoconid present. A second crest runs from the lingual aspect of the tip lower on the lingual surface of the crown, and then swings up to nearly join the margocristid at the base of the crown. The two crests define a very shallow lingual surface with a slight bulge running down its midline marking the break in slope between the dorsal and lingual surfaces of the crown. The single-rooted C1 of D. szalayi is positioned directly distal to I1, with a very small diastema separating the two alveoli (Figs. 13 and 14). The crown of C1 is elongate and projects mesially away from the root, covering the diastema, to the base of I1. The crown

122

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

Figure 12. Dryomomys szalayi (holotype: UM 41870) R maxilla P2eM3 in buccal (A) and occlusal (B) views. L maxilla C1eM3 in buccal (C) and occlusal (D) views.

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

has a large mesially placed protoconid followed by a low bulbous talonid heel. The single-rooted P2 of D. szalayi is positioned directly distal to C1, also with a very small diastema separating the two alveoli (Figs. 13 and 14). The morphology of P2 is generally similar to that of the C1 in being elongate and projecting mesially away from the root. It differs from the C1 in being smaller and having a shorter (about two-thirds the length) mesial projection (protoconid) that does not cover the entire mesial diastema. It also has a low talonid heel, but with a more distinctly developed cusp than that of C1. The P3 of D. szalayi is double-rooted and the crown is about twice the size of P2 (Figs. 13 and 14). The crown of P3 is oval in occlusal view and is approximately twice as long as it is wide. The trigonid is mesially inflated, lacks a paraconid, is dominated by a protoconid, and a faint crest (paracristid or preprotocristid) descends from the apex of the protoconid mesially before curving lingually toward the base. The crown of P3 has a wide, short talonid heel with a shallow basin and no clearly defined cusps. The crown of P4 is considerably larger and taller than any other in the toothrow distal to I1 (Figs. 13 and 14). The shape of P4 is oval in occlusal view, with the mesial portion of the tooth somewhat narrower than the distal portion of the tooth. The crown of P4 is fairly exodaenodont with the base moderately distended distobuccally. The trigonid is wide, formed by a tall protoconid, and there is no paraconid or metaconid present. A paracristid descends the protoconid mesially for about one-third of the length of the trigonid, then turns mesiolingually and continues down most of the height of the crown nearly vertically before terminating about at the level of the talonid. In buccal view, there is a break of slope along the mesiodorsal margin at the point along the paracristid where it curves lingually. Also at this point, another smaller crest splits off the paracristid and descends mesially merging with the faint crest on the mesiobuccal margin of the trigonid. The paracristid defines the mesial margin of a distinct mesiolingual excavation that is distally bounded by a rib descending from the apex of the protoconid. A second, slightly smaller lingual excavation is also present distal to the apex and underlying rib of the protoconid. The postvallid face is almost transverse across the crown. The cristid obliqua and postcristid are worn, but it seems as though two poorly defined cusps were present, separated by a shallow valley. The hypoconid appears to have been larger and more distinct than the entoconid. The talonid basin is small and wider than long. The cristid obliqua is angled slightly lingually and connects to the postvallid slightly lingual to the midline of the wall. The hypoflexid is deep and occupies the buccal half of the talonid. The crown of M1 of D. szalayi is square in outline and generally similar to that of some other micromomyids in being low-crowned with fairly bulbous cusps (Figs. 13 and 14). The trigonid is about twice the height of the talonid with a distinct protoconid and metaconid. The protoconid and metaconid are of similar size with the metaconid only slightly larger, and the two cusps are closely appressed, separated by a small but distinct valley. A small paraconid is present on the midline of the trigonid, directly mesial to the valley. The paraconid is situated on a very short paracristid that connects to the base of the protoconid. In occlusal view the cusp is lophate, whereas in lingual view it is cuspate (but very small). A very faint precingulid is present on the mesiobuccal wall of the trigonid. The talonid has a fairly deep basin, surrounded on all sides by crests. Two main cusps dominate the talonid, the entoconid and a somewhat larger hypoconid located at the distolingual and distobuccal extremes of the basin respectively. The cristid obliqua is a straight crest that extends mesiolingually from the hypoconid to the base of the postvallid where it ends below the protoconid without extending up the face of the postvallid. A poorly-defined, lophate hypoconulid is located lingual to the midline of the tooth,

123

slightly closer to the entoconid than the hypoconid. A faint postcingulid is present on the distal aspect of the talonid between the hypoconulid and the hypoconid that is not continuous with the precingulid (i.e., no evidence of an ectocingulid). The crown of M2 of D. szalayi is similar to that of M1 in being square in outline (although slightly shorter and wider), with lowcrowned and fairly bulbous cusps (Fig. 13). The trigonid of M2 is not as tall as that of M1, about 1.5 times the height of the talonid with a distinct protoconid, metaconid, and paraconid. As for M1, the protoconid and metaconid are of similar size with the metaconid slightly larger. The two cusps are less closely appressed in M2 than for M1, separated by a larger and distinct valley. A moderately sized paraconid, larger than that of M1, is present on the midline of the trigonid of M2, directly mesial to the valley but closer to the other trigonid cusps than that of M1. The paraconid is situated on a short paracristid that connects to the protoconid. In occlusal and lingual views, the paraconid is cuspate and projects somewhat mesially in M2, rather than lophate as in M1. A faint but slightly more pronounced precingulid is present on M2 on the buccal wall of the trigonid. As for M1 the talonid has a fairly deep basin, surrounded on all sides by crests. Two main cusps, the entoconid and a somewhat larger hypoconid located at the distolingual and distobuccal extremes of the basin, respectively, dominate the talonid. The talonid basin of M2 is slightly wider than that of M1, with the entoconid and hypoconid more widely spaced. As for M1, the cristid obliqua of M2 is a straight crest that extends mesiolingually from the hypoconid to the base of the postvallid where it ends below the protoconid without extending up the face of the postvallid. Unlike that of M1, a faint swelling in the position of a mesoconid is present on the cristid obliqua of M2 just before it meets the postvallid. As in M1, a poorly defined, lophate hypoconulid is located lingual to the midline of M2, slightly closer to the entoconid than the hypoconid. Also like M1, a faint postcingulid is present on the distal aspect of the talonid between the hypoconulid and the hypoconid of M2 that is not continuous with the precingulid (i.e., no evidence of an ectocingulid). The crown of M3 of D. szalayi is similar to that of other micromomyids for which it is known in being elongate in outline, with low-crowned and fairly bulbous cusps (Figs. 13 and 14). The trigonid of M3 is similar to that of M2 in being about 1.5 times the height of the mesial aspect of the talonid with a distinct protoconid, metaconid, and paraconid. The protoconid and metaconid on the M3 are of similar size with the protoconid slightly larger, unlike that of M1e2 in which the metaconid is slightly larger. The two cusps are less closely appressed in M3 than for M1 (similar to that of M2), and are separated by a larger and distinct valley. A distinct paraconid, larger than that of M1e2, is present slightly lingual to the midline of the trigonid of M3, mesial to the valley but closer to the other trigonid cusps than that of M1 (similar to that of M2). The paraconid of M3 is connected to the protoconid by a somewhat longer paracristid than that of M1e2. In occlusal and lingual views, the paraconid is cuspate and projects somewhat mesially in M3 (similar to that of M2), rather than reduced and lophate as in M1. A more pronounced precingulid is present on the buccal wall of the trigonid of M3 than that of the other lower molars. The talonid of M3 has a fairly deep, elongate basin that is dominated by a distally expanded hypoconulid. The pronounced hypoconulid extends dorsally almost to the height of the trigonid. The entoconid and a larger hypoconid are located at the lingual and buccal extremes of the M3 talonid basin, respectively. The talonid basin of M3 is similar in width to that of M1, with the entoconid and hypoconid similarly spaced. The entoconid of M3 is shifted more distally relative to the hypoconid than that of M2, and to a greater extent M1 in which the cusps are nearly level with each other. As for M1e2, the cristid obliqua of M3 is a straight crest that extends mesiolingually from

124

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

Figure 13. Dryomomys szalayi (holotype: UM 41870) R dentary I1eM3 in lingual (A), buccal (B), and occlusal (C) views.

the hypoconid to the base of the postvallid where it ends below the protoconid without extending up the face of the postvallid with no development of a mesoconid (which is present on M2 but not M1). Also like the other lower molars, a faint postcingulid is present on the distal aspect of the talonid between the hypoconulid and the hypoconid of M3 that is not continuous with the precingulid (i.e., no evidence of an ectocingulid across the face of the protoconid). The dentary is fairly uniform in depth from P4eM3, with two distinct mental foramina below the mesial roots of P4 and P3, respectively (Figs. 13 and 14). A short, unfused mandibular symphysis quickly divides into two distal extensions. The dorsal portion extends to the C1 root and the ventral portion as far as the root of P2. The coronoid process is tall and vertically oriented. The mandibular condyle is situated high on the ascending ramus, well above the height of the cheek teeth. The articular surface of the condyle is smoothly convex, wide, and hemicylindrical. The angular process is relatively elongate and pointed, with a flat dorsal surface and slight medial inflection. Comparison and discussion: The I1 of D. szalayi is similar to that of T. graybulliensis in general shape and proportions, although it is somewhat narrower (Fig. 15). It differs in having a less developed mediocone, mostly defined by an internal crest, giving D. szalayi a narrower distal end than that of T. graybulliensis. It further differs from that of T. graybulliensis in having only a single cusp at its base. Following the description and terminology of Rose et al. (1993), this cusp may be the homolog of the lateroconule in the I1 of

T. graybulliensis, with the posterocone completely absent, although this cusp has previously been considered to be the posterocone (Bloch et al., 2007; Chester and Beard, 2012). Recently, isolated upper central incisors of micromomyids were described from Big Multi Quarry, Washakie Basin, Wyoming (the type locality of T. tribos and D. dulcifer; Chester and Beard, 2012). These I1s have a faint mediocone and no lateroconule, making them more similar to that of D. szalayi than to that of T. graybulliensis. This seems to suggest that they should be attributed to D. dulcifer. However, the dentition of T. tribos appears to be more plesiomorphic than T. graybulliensis in several ways, and T. tribos is more abundant than D. dulcifer at Big Multi Quarry, so it is also possible that these isolated plesiomorphic I1s could belong to T. tribos (Chester and Beard, 2012). Rose et al. (1993) suggested that while micromomyids were generally similar to plesiadapids, carpolestids, saxonellids, and paromomyids in having multicusped upper incisors, the relatively simple I1 of T. graybulliensis was most similar among plesiadapiforms to isolated incisors from Gidley Quarry that were tentatively assigned to Palaechthon alticuspis. This could suggest a close relationship among micromomyids and palaechthonids, or that this relatively simple I1 morphology is plesiomorphic for plesiadapiforms (Rose et al., 1993). Bloch et al. (2002) described the deciduous dentition of the paromomyid plesiadapiform, Acidomomys hebeticus, including a dI1 that was quite similar to the isolated incisors from Gidley Quarry. Given the similarity, Bloch et al. (2002) suggested that these isolated upper incisors belong to the paromomyid

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

125

Figure 14. Dryomomys szalayi (holotype: UM 41870) L dentary I1eM3 (M2 absent) in lingual (A), buccal (B), and occlusal (C) views.

Paromomys despressidens (not to P. alticuspis), which is also known from Gidley Quarry. These authors also suggested that such simple I1 morphology may have been present in the ancestor that gave rise to all later plesiadapiforms, however, they noted that the I1 of the oldest and possibly most primitive plesiadapiform, Purgatorius, was still unknown (Bloch et al., 2002). Two years later, an isolated right I1 was attributed to P. janisae from the earliest Paleocene of Montana (Clemens, 2004). This tooth is similar to that of D. szalayi in having only a poorly developed mediocone and narrower distal crown. The I1 of P. janisae differs from that of D. szalayi, however, in having a small posterocone in addition to the lateroconule, making it more similar in this respect to T. graybulliensis. The relatively simple morphology of I1 documented in D. szalayi has also been documented in primitive members of other plesiadapiform groups, such as Pandemonium dis (Van Valen, 1994) and Chronolestes simul (Beard and Wang, 1995). Additionally, an isolated incisiform tooth (likely I1) recently attributed to the palaechthonid Torrejonia wilsoni lacks complex multicusped morphology (Silcox and Williamson, 2012). These observations suggest that this simple incisor morphology is primitive and that multicusped upper incisors evolved more than once among plesiadapiforms (Beard and Wang, 1995; Bloch et al., 2007). The fact that the simple I1 morphology of Dryomomys is so similar to that of other primitive

plesiadapiforms such as Chronolestes suggests that more complex incisors evolved within Micromomyidae, from a more simple tooth like that of D. szalayi to a more complex tooth like that of T. graybulliensis. It is also possible that the simple morphology of the I1 of D. szalayi is more plesiomorphic than that of the I1 that was attributed to P. janisae. The I2 of D. szalayi differs from that of T. graybulliensis (Rose et al., 1993) in having a less well-developed medial crest, again making the crown narrower at the distal end (Fig. 15). Like T. graybulliensis, the I2 of D. szalayi is generally similar to that of other plesiadapiforms, such as the paromomyid Ignacius and the plesiadapid Nannodectes, in being simple and compressed laterally (Rose et al., 1993). It appears that the C1 of T. graybulliensis (only known from alveoli; Rose et al., 1993) would have been similar to that of D. szalayi in crown size and proportions of roots. D. millennius also had a similarly sized, double-rooted C1 (only known from alveoli; Secord, 2008). The diastema between C1eP2 is relatively shorter in T. graybulliensis than that of D. millennius and D. szalayi. Many plesiadapiforms have a single-rooted C1 and several derived members of various families (e.g., Carpolestidae and Plesiadapidae) have lost this tooth completely (Silcox, 2001). All micromomyids have a double-rooted C1, as in many microsyopids, and the paromomyid Phenacolemur jepseni (Simpson, 1955). The presence of a

126

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

Figure 15. Micro-CT scan images of upper incisors of micromomyids Tinimomys graybulliensis and Dryomomys szalayi captured using Avizo 6. Wasatchian T. graybulliensis, USNM 461201, palate with L I1e2, R P2eM3, in lingual (A), buccal (B), and occlusal (C) views. New Clarkforkian T. graybulliensis, UM 39927, L premaxilla I1e2, in lingual (D), buccal (E), and occlusal (F) views. Dryomomys szalayi, UM 41870, R premaxilla I1e2 (reversed), in buccal (G), lingual (H), and occlusal (I) views. Scale bar ¼ 1 mm.

double-rooted C1 in these fossil taxa and in extant P. lowii and dermopterans suggests that this condition was present in the ancestral euarchontan. The crown of P2 of D. szalayi is distinctly different from that of T. graybulliensis (Gunnell, 1989; Rose et al., 1993) in having a stronger lingual cingulum that extends onto a distal, almost basined, heel (in the area of the metastylar cusp of F. fremdi, although most of the P2 is not preserved in that specimen). D. millennius has a moderate cingulum on P2, although it is not as strong as that of D. szalayi. In general, the crown of P2 is somewhat larger relative to M1 in D. szalayi compared with that of D. millennius, and substantially larger than that of T. graybulliensis and F. fremdi.

D. szalayi differs from all other micromomyids in having a larger P3 relative to the size of its upper molars. D. dulcifer and D. millennius most closely approximate this condition (Fig. 9), and F. fremdi is least similar due to its extremely small P3. The P3 of D. szalayi has more pronounced parastylar, metastylar, and protocone lobes than any other micromomyid, giving the tooth a more inflated look, although D. dulcifer, D. millennius, and T. graybulliensis approach this condition. The P3 of C. antelucanus and F. fremdi lack the development of these lobes, although C. antelucanus is unique in possessing a distinct parastyle (Beard and Houde, 1989), and F. fremdi is unique in possessing a metastyle. The P3 of D. szalayi is similar to that of all other micromomyids for which this tooth is known in lacking a metacone, with the exception of

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

T. graybulliensis, which has a variably developed metacone. The P3 of D. szalayi possesses a distinct protocone with a moderately developed basin, which is most similar to that of D. dulcifer and D. millennius compared with all other micromomyid species. The P4 of D. szalayi is larger and more lingually sloping relative to M1 than that of any other micromomyid. Though the position and relative size of the buccal cusps of D. szalayi are generally similar to other micromomyids for which P4 is known, they differ in some critical ways. The position of the metacone and the orientation of the metacrista in D. szalayi is similar to nearly that of all other micromomyids for which P4 is known, with the metacone located directly distal to the paracone before the postmetacrista curves distobuccally. In contrast, the P4 metacone of F. fremdi is located more distobuccally on a more gradually curving metacrista forming a blade-like structure, similar to that of Purgatorius (e.g., Lillegraven et al., 1979; Fox and Scott, 2011) and the presumed primitive condition. The parastylar lobe of D. szalayi is similar to that of Tinimomys in being larger and more transversely extended than other micromomyids. The valley separating the parastyle from the paracone is more pronounced in D. szalayi than in any other known micromomyid, although most closely approximating the condition of D. dulcifer and D. millennius (Fig. 9). The mesial and distal aspects of the ectocingulum in D. szalayi are similar to those of D. dulcifer and T. graybulliensis in being more developed than those of D. millennius, F. fremdi, and C. antelucanus. The P4 protocone of D. szalayi is most similar among micromomyids to that of D. dulcifer and D. millennius, and to a lesser extreme C. antelucanus, in being situated very close to the paracone and metacone and having a transversely extended protocone lobe. In this way, it is distinctly different from F. fremdi, which has a more lingually situated protocone and transversely shorter lobe. Comparison with that of Tinimomys is complicated by the fact that the protocone is usually worn, and by the presence of a hypocone, which is absent in D. szalayi. However, although the protocone lobe of Tinimomys is fairly large, the protocone cusp is more lingually positioned with a larger trigon basin than that of D. szalayi. As for all other micromomyids except Tinimomys, the P4 pre- and postcingula of D. szalayi are not lingually continuous. D. szalayi is also similar to other micromomyids in having a rounded lingual margin of the protocone lobe, in contrast to the squared-off condition in T. graybulliensis. The crowns of M1e3 of D. szalayi are similar to those of T. graybulliensis in having a shallow ectoflexus and parastyle. In contrast, the ectoflexus and parastyle are more pronounced in F. fremdi and to a lesser extent C. antelucanus and T. tribos. D. szalayi differs from F. fremdi and is more similar to other micromomyids in having less transverse molars with similar buccal and lingual lengths, resulting in more quadrate teeth. The relative molar size of D. szalayi is close to that of T. graybulliensis and T. tribos in having similarly sized M1 and M2, whereas M2 is considerably more elongate than M1 in F. fremdi and C. antelucanus (no other species of micromomyid preserves both tooth positions). The upper molars of T. graybulliensis and T. tribos are distinctly different from all other micromomyids, including D. szalayi, in having a continuous lingual cingulum with a well-developed hypocone and pericone (except M3, which lacks a hypocone and pericone in all known micromomyid taxa). The condition in D. szalayi, in which the upper molars have well-developed pre- and postcingula (with only incipient development of cuspules in M1e2) but lack a lingually continuous cingulum, is most similar to that of the M2 of D. dulcifer and M1 of D. millennius. In contrast, C. antelucanus, and to a greater extent F. fremdi, have less well-developed pre- and postcingula with no development of cuspules on the upper molars. The M2 of D. szalayi is similar to other micromomyids in lacking the variably expressed cuspule buccal to the metacone that has been documented in that of C. antelucanus

127

(Beard and Houde, 1989). The M1 of D. szalayi is similar to that of other micromomyids in having a rounded lingual margin with a stronger postcingulum than precingulum. In contrast, the M1 of D. millennius is more squared-off with a flatter lingual margin. The crown of I1 in D. szalayi is very similar to that of T. graybulliensis. It differs in that the I1 crown of T. graybulliensis has a dorsoventrally deeper base and a more expanded dorsal surface, forming a small dorsal shelf running for most of its length. The I1 morphology of micromomyids is most similar among plesiadapiforms to that of Saxonella (Rose et al., 1993), Phenacolemur, and Picromomys (Silcox, 2001; Silcox et al., 2002), although other plesiadapiform taxa that have an elongate I1 and lack a margoconid are also broadly similar (e.g., Carpolestes; Bloch and Gingerich, 1998), contrasting with the lanceolate form of other plesiadapiforms (e.g., microsyopids). D. szalayi is the third micromomyid (see also T. graybulliensis and F. fremdi) for which the crown of C1 is known, although a single alveolus present in several other species (M. silvercouleei, F. vossae, F. gunnelli, C. antelucanus) indicates that this tooth is consistently single-rooted within Micromomyidae. F. fremdi is distinct in having a longer and more mesially extended C1 than that of T. graybulliensis and D. szalayi to a lesser extreme. Though the crown of C1 is somewhat longer in D. szalayi than that of T. graybulliensis, they are similar in having a more horizontally oriented crown, as opposed to that of F. fremdi, which is oriented at almost 45 degrees to the toothrow. The crown of C1 in D. szalayi has a low bulbous talonid heel that lacks the distinct talonid cusp present on that of T. graybulliensis and to a lesser extreme F. fremdi. The C1 of D. szalayi is similar to that of T. graybulliensis, M. silvercouleei, C. antelucanus, and F. gunnelli in having a shorter diastema between I1 and C1 than that of F. fremdi. Other than D. szalayi, the only micromomyid for which the crown of P2 has been recovered is F. fremdi (Fox, 1984). The crown of P2 in D. szalayi is smaller than that of F. fremdi relative to lower molars, and more mesially projecting. As such, the morphology of the crown of P2 in D. szalayi is very similar to that of its canine, whereas the P2 crown morphology in F. fremdi is more similar to that of its own P3. The P2 of D. szalayi is similar to that of M. silvercouleei and C. antelucanus in being single-rooted (based on alveoli; see Fig. 4). In contrast F. fremdi, F. gunnelli (based on alveoli), and probably F. vossae (based on alveoli) had double-rooted P2 and T. graybulliensis has lost P2 entirely. The P3 of D. szalayi is most similar to that of D. dulcifer and D. millennius among micromomyids in being larger and higher crowned relative to the lower molars. The P3 of D. szalayi has a more mesially extended trigonid than that of F. fremdi, perhaps most similar to the condition in D. millennius (the protoconid is broken in the only specimen for which P3 is preserved; Fig. 10), although it differs from that of T. graybulliensis in lacking a cusp mesial to the protoconid. Though comparison to C. antelucanus is difficult as the only known specimen for which the P3 is preserved is heavily damaged, with the entire buccal side missing (Beard and Houde, 1989), it seems that this tooth would also have been mesially extended, with no clear evidence of a mesial cusp. The P3 of D. szalayi differs from that of T. graybulliensis and especially D. dulcifer and D. millennius in being narrower and less inflated, more like that of F. fremdi. It also has a slightly more developed talonid basin than all other micromomyids for which P3 is known, although it most closely resembles D. dulcifer and D. millennius in this respect. The P4 of D. szalayi is larger relative to the lower molars than that of all other micromomyids except possibly D. willwoodensis (Fig. 10; this specimen only preserves the crown of P4 and molar alveoli). Much of this size difference is due to the wider and more inflated trigonid of D. szalayi, which is most similar to that of D. dulcifer, D. willwoodensis, and D. millennius in this regard, and that of C. antelucanus and T. graybulliensis to a lesser extent. The

128

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

mesial aspect of the trigonid of D. szalayi is similar to that of D. millennius in that it does not narrow as much mesially as in other micromomyids. The distobuccal aspect of the trigonid of D. szalayi is similar to that of D. dulcifer and M. silvercouleei, and to a lesser extent F. vossae, F. fremdi, and F. gunnelli, in being less extended buccally than that of D. willwoodensis, D. millennius, C. antelucanus, and T. graybulliensis. However, this feature is variable and influenced by the breadth of the mesial aspect of the trigonid and the nature of the hypoflexid. For example, D. szalayi differs from D. willwoodensis in having the floor of the hypoflexid extend to the buccal margin of the tooth, giving the impression that it has a less pronouced distobuccal lobe despite the fact that it is still quite wide. As for all other micromomyids, except T. graybulliensis, which has a variably present, small paraconid (Szalay, 1974; Bown and Rose, 1976), there is no development of a paraconid on the P4 of D. szalayi. The P4 of D. szalayi is similar to that of D. dulcifer and C. antelucanus in being moderately exodaenodont with its base more distended distobuccally than that of D. willwoodensis and less distended than that of T. graybulliensis and Tiffanian micromomyids such as D. millennius. The P4 paracristid of D. szalayi is similar to that of D. dulcifer, D. millennius, and C. antelucanus in that it is not as distinct and does not extend as far down to the base of the tooth as it does in D. willwoodensis, but is more distinct than other micromomyids. As for all other micromomyids except F. fremdi and F. gunnelli, when oriented in buccal view, the P4 paracristid of D. szalayi has a break in slope along the mesiodorsal margin where it curves lingually. However, like D. dulcifer, D. szalayi differs from the remaining species in having a shorter mesiodorsal margin with a shallow slope before the paracristid curves lingually, superficially almost approaching the condition of that of F. fremdi and F. gunnelli in which the mesial margin appears smoothly convex with no distinct break in slope. The cristid obliqua of D. szalayi is similar to that of D. dulcifer, D. millennius, D. willwoodensis, C. antelucanus, M. silvercouleei, and T. graybulliensis in not climbing the postvallid and differs from that of F. fremdi, F. vossae, and F. gunnelli in which it climbs the postvallid for a short distance. The P4 talonid of D. szalayi is shorter than wide, and shorter relative to the length of the tooth, than that of any other micromomyid. The talonid basin is small like that of all other micromomyids other than T. graybulliensis. The P4 of D. szalayi has two talonid cusps, a hypoconid and an entoconid, similar to that of D. dulcifer, D. millennius, M. silvercouleei, T. graybulliensis, C. antelucanus, and F. fremdi (although presence of entoconid is variable; Fox, 1984). F. gunnelli differs from D. szalayi in only having a poorly defined hypoconid on the talonid of P4, and F. vossae and possibly D. willwoodensis differ in having three faint P4 talonid cusps. As for all other known micromomyids, the P4 talonid cusps of D. szalayi are not as pronounced as those of T. graybulliensis. The crowns of M1e2 of D. szalayi are similar to those of all other micromomyids in being generally square in outline and being lowcrowned with fairly bulbous cusps (Fig. 10). More specifically, like other micromomyids, the lower molar cusps of D. szalayi are not as bulbous as those of T. graybulliensis, nor are they as high-crowned as those of F. fremdi. The relative height of the trigonid to talonid in M1e 2 (twice the height in M1 and slightly less in M2) in D. szalayi is similar for all micromomyids for which these tooth positions are known. As in other micromomyids, D. szalayi has a M1 with a distinct protoconid and metaconid closely appressed, yet separated by a small but distinct valley, and a M2 with a wider valley making the protoconid and metaconid less closely appressed. However, the crowns of M1e2 in D. szalayi differ from those of other micromomyids in having a slightly larger metaconid than protoconid. In contrast, the protoconid is considerably larger than the metaconid in that of F. fremdi, and somewhat larger in D. dulcifer, D. millennius, M. silvercouleei (only known molar is M2), T. graybulliensis, and C. antelucanus,

although this difference is considerably stronger in M1 than M2. The crowns of M1e2 of D. szalayi are similar to all other micromomyids in that the paraconid is the smallest trigonid cusp. However, the M1 of D. szalayi differs from that of all other micromomyids in having a considerably more reduced paraconid that forms a small loph on a very short paracristid in occlusal view. The M1 paraconid is also more centrally located, directly mesial to the valley separating the metaconid and protoconid, than that of other micromomyids including C. antelucanus. In contrast, the paraconid of M1 is larger, more cuspate, connects to a longer paracristid, and is more lingually situated (mesial to the metaconid) in F. fremdi, F. vossae, D. millennius, and T. graybulliensis. The M1 paraconid of D. dulcifer is similar to that of D. szalayi in being centrally located, however, it is much more distinct (Chester and Beard, 2012). The paraconid of M2 in D. szalayi is also more reduced than that of other micromomyids, but less so than that of M1. The paraconid of M2 in D. szalayi is similar to that of D. dulcifer and C. antelucanus in the central location of this cusp relative to the midline, but differs from that of F. fremdi, M. silvercouleei, D. millennius, and T. graybulliensis in which the cusp is more lingually positioned. Though the crowns of M1e3 of D. szalayi are similar to all other micromomyids in lacking a well-defined ectocingulid, they differ somewhat in having less defined pre- and postcingula on the mesio- and distobuccal margins. The talonid morphology of M1e2 of D. szalayi is generally similar to that of all other micromomyids. Prior to the discovery of the holotype of D. szalayi (UM 41870), M3 was only unambiguously known for T. graybulliensis and F. fremdi. In these taxa the talonid of M3 is markedly different, with the hypoconulid considerably expanded and distally projecting in T. graybulliensis, whereas the hypoconulid of F. fremdi is large and extended distally, but is more cuspate. The hypoconulid of M3 in D. szalayi is somewhat intermediate in that it is similar to that of T. graybulliensis in being expanded, but it is narrower and tapers distally like that of F. fremdi. More recently, M3 has been recovered for T. tribos and D. dulcifer (Chester and Beard, 2012), and the morphology of these teeth is most similar to that of T. graybulliensis and D. szalayi, respectively. The M3 of D. szalayi is also quite similar overall to the M3 attributed to C. antelucanus above, especially in the slight expansion of the hypoconulid, but it differs in having a shorter trigonid relative to M3 length, with a more centrally situated paraconid (UM 76682; Fig. 8AeC). As for most other micromomyids, the dentary of D. szalayi is more uniform in depth than that of T. graybulliensis and T. tribos in which the mandibular depth increases distally from the tip of the mandibular symphysis, with its greatest depth below P4 (Beard and Houde, 1989; Chester and Beard, 2012). The position of the mental foramina are below the mesial roots of P3 and P4, respectively, and is similar in this way to the condition in D. millennius, C. antelucanus, and Tinimomys, but differs from that of F. fremdi and F. gunnelli in which the mesial-most alveolus is farther forward, under the distal root of P2, and from that of M. silvercouleei in which it is under a single-rooted P2. The presumed primitive condition is illustrated by that of Purgatorius in which the mental foramina are beneath and slightly mesial to the mesial roots of P2 and P4. The back of the dentary of D. szalayi is very similar to that of other micromomyids for which it is known, although the condyle appears to be higher relative to the toothrow than that of T. graybulliensis and C. antelucanus, in which it is about the same height as P4. This may be related to a relatively more herbivorous diet in D. szalayi than these other taxa (see Rose, 1975), although this conclusion remains to be tested with more rigorous quantitative analyses of morphology. The articular surface of the condyle of D. szalayi is similar to that of T. graybulliensis and C. antelucanus in being smoothly convex and hemicylindrical, but is clearly relatively wider mediolaterally than for either of those taxa. The angular process of D. szalayi is similar to that of T. graybulliensis, and contrasts with the

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

condition in C. antelucanus (a simple pointed structure) in having a dorsally oriented concavity with a slight medial inflection, thought to be related to the insertion of the medial pterygoid muscle (Beard and Houde, 1989). The short, unfused symphysis of D. szalayi is similar in morphology to that of T. graybulliensis (Figs. 13 and 14). D. willwoodensis (Rose and Bown, 1982) (new combination) (Fig. 10JeL) Holotype: YPM-PU 17732, L partial dentary P4 and alveoli for M1e2. Known distribution: Early Wasatchian (Wa-2), UM locality SC-2, Willwood Formation, Bighorn Basin, Wyoming. Emended diagnosis: Differs from all other species of Dryomomys in having P4 with more distinct paracristid that extends farther down toward the base of the crown, narrower talonid basin relative to length with three poorly defined talonid cusps (rather than two more defined cusps), less ventrally distended distobuccal lobe, and an invagination at the base of the hypoflexid at the distobuccal margin of the crown. Further differs from D. dulcifer and D. szalayi in having a wider P4 relative to length. Further differs from D. millennius in having relatively larger P4. Discussion: D. willwoodensis is the only species of Dryomomys known from the early Eocene (Wa-2). Rose and Bown (1982) noted that the molars of D. willwoodensis were approximately the same size as those of other micromomyids (based on alveoli; the crowns were not recovered), yet P4 was approximately 25% larger than that of M. silvercouleei and M. vossae. This observation appears to be correct and such a relatively large P4 is one of the characteristics unique to Dryomomys. However, before other species of Dryomomys were described, D. willwoodensis was referred to Chalicomomys because C. antelucanus is also known from the early Eocene (Wa-1) and is somewhat similar in P4 morphology (Beard and Houde, 1989). Although no other species of Dryomomys are known from the earliest Eocene, and it is possible that P4 size may have increased convergently in more than one group of micromomyids, the size and morphology of P4 (and mandibular depth) of D. willwoodensis is most similar to those of Dryomomys species such as D. szalayi. The phylogenetic analysis presented below supports the placement of D. willwoodensis in a Dryomomys clade. D. millennius (Secord, 2008) (new combination) (Figs. 9A, 10AeC) Holotype: UM 109659, R partial dentary P3eM2 and partial alveolus for I1. Known distribution: Late Tiffanian (Ti-5b), Y2K Quarry, northern Bighorn Basin, Wyoming. Emended diagnosis: Differs from all other species of Dryomomys in having a wider P4 relative to length and a smaller P4 relative to lower molars, with more ventrally distended distobuccal lobe and more distinct distobuccal cingulum. Further differs from D. szalayi and D. dulcifer in having smaller and less inflated P3 with smaller protocone, and smaller P4 relative to upper molars with a less lingually expanded protocone lobe. Further differs from D. szalayi in having a wider P3 relative to length, M1e2 with more cuspate and more lingually situated paraconids, and more pronounced distobuccal cingula, and M1 with more squared-off lingual margin. Discussion: Secord (2008) tentatively attributed UM 110140, a maxilla with P2eM1 and alveoli for C1, to ‘M.’ millennius, although he noted that it could also belong to M. silvercouleei. This specimen was collected at Y2K Quarry, which lies 85 m above Princeton Quarry where the only known specimen of M. silvercouleei was discovered. The P4 of D. millennius is considerably wider than that of M. silvercouleei, and the P3 and P4 of D. millennius are quite wide like the P3 and P4 of UM 110140 (Secord, 2008). These characteristics are consistent with the generally wide upper premolars known for D. dulcifer and D. szalayi, and support the attribution of UM 110140 to D. millennius.

129

D. dulcifer Chester and Beard, 2012 (Figs. 9B, 10D-F) Holotype: CM 72292, L partial dentary M1e3. Known distribution: Middle Clarkforkian (Cf-2), Big Multi Quarry, Washakie Basin, Wyoming. Emended diagnosis: Differs from all other species of Dryomomys in having a narrower P4 relative to length. Differs from D. szalayi in having relatively longer and narrower P3e4, smaller P3 relative to known upper molar tooth positions, P3 with less pronounced protocone lobe and smaller protocone, M2 with more pronounced ectoflexus and shorter lingual margin, wider P3 relative to length, and M1 with distinct paraconid. Further differs from D. millennius in having relatively larger, more inflated P3, and P4 with more lingually expanded protocone lobe and apex of protocone more closely situated to that of paracone and metacone. Discussion: Chester and Beard (2012) suggested that Plesiadapis cookei might have occurred earlier at Big Multi Quarry in the Washakie Basin than at Cf-2 localities in the Bighorn Basin. Regardless, Big Multi Quarry is Cf-2 by definition (P. cookei range zone; Gingerich, 2003) based on the presence of P. cookei and absence of Copecion (56.2e56.5 Ma (millions of years ago); Secord et al., 2006). Dryomomys sp. (Fig. 9D,E) Referred specimen: UM 39849, R P4 (broken). Locality and horizon: Middle Clarkforkian (Cf-2) SC-19, Bighorn Basin, Wyoming. Description and comparison: The isolated P4 is broken with the entire mesial margin and parastylar lobe missing. Nevertheless, this tooth preserves much of the crown and is noticeably smaller (1.65 mm width) than that of D. szalayi and D. dulcifer. It differs from that of Tinimomys in lacking a lingually continuous cingulum and hypocone, and differs from Chalicomomys in having a more elongate protocone lobe and a relatively smaller metacone with a less well-defined postmetacrista. UM 39849 has an elongate protocone lobe with pronounced mesial and distal furrows most similar to that of D. szalayi, D. dulcifer, and D. millennius. Although the protocone lobe may be somewhat more elongate than that of D. millennius, this P4 is similar to that of D. millennius in having a protocone that is not as closely appressed to the paracone and metacone as it is in D. szalayi and D. dulcifer. Discussion: UM 39849 differs from the P4s of D. dulcifer and D. szalayi in being smaller and having a protocone that is more lingually situated on the crown. This morphology is similar to that of D. millennius, F. fremdi, and C. antelucanus, and is likely plesiomorphic. Although this specimen may represent morphological variability in P4 of D. szalayi, it appears most similar to that of D. millennius overall. Given the isolated and incomplete nature of the specimen, it is only attributed to the genus Dryomomys here. Tinimomys Szalay, 1974 Type species: T. graybulliensis (Szalay, 1974) Included species: T. graybulliensis (Szalay, 1974), T. tribos (Chester and Beard, 2012). Known distribution: Late Paleocene to early Eocene of North America (early Clarkforkian (Cf-1) to early Wasatchian (Wa-2) of the Bighorn Basin, Wyoming; middle Clarkforkian (Cf-2) of Big Multi Quarry, Washakie Basin, Wyoming; and early Wasatchian of the Powder River Basin, Wyoming). Emended diagnosis: Differs from all other micromomyids in having pre- and postcingula strongly confluent lingually on P4eM3 with variably developed hypocone and pericone on P4eM2. Further differs from all other micromomyids in lacking P2, having large P4 talonid basin with pronounced cusps, relatively low-crowned molars with more bulbous cusps, M3 with a more expanded hypoconulid, and dentary that reaches its greatest depth below P4. Discussion: The diagnosis for the genus Tinimomys has been emended to include the recently described species, T. tribos (Chester

130

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

and Beard, 2012). T. tribos retains several plesiomorphic dental features that have been further modified in T. graybulliensis. Tinimomys can be further differentiated from other micromomyids in having a P4 paraconid, although this feature is variably expressed. T. graybulliensis Szalay, 1974 (Figs. 15AeF, 16e18) Holotype: YPM-PU 17899, L partial maxilla P4eM2. Referred specimens: UCM locality 84117 (early Wasatchian): UCM 62749, R partial dentary P4eM3 (Fig. 17EeG). UCM locality 89079 (early Wasatchian): UCM 76918, L partial maxilla P3e4 (Fig. 17A,B); UCM 76919, R partial dentary I1, M1. UCM locality 91056 (early Wasatchian): UCM 65916, R partial dentary I1, P4. UCM locality 98074 (early Wasatchian): UCM 86800, R partial dentary M1e 2. UCM locality 2004086 (early Wasatchian): UCM 101486, L partial maxilla P4eM3 (Fig. 17C,D). UCM locality 2010009 (early Wasatchian): UCM 62762, R partial dentary P4eM3 (Fig. 18A,B); UCM 70619, L partial dentary M3 (Fig. 17HeK). UM locality SC-179 (Cf-1): UM 36199, R M2 (Fig. 16AeD). UM locality SC-117 (Cf-2): UM 39877, R M2; UM 39878, R partial dentary M3, alveoli for M2; UM 39879, R M3; UM 39880, R M1; UM 39881, L partial maxilla M2; UM 39883, L partial dentary I1eM1, M3 (C1 broken) (Fig. 16EeG), alveoli for M2; UM 39927, L partial premaxilla I1e2 (Fig. 15DeF); UM 39928, R M2; UM 39929, L I1; UM 39931, L partial dentary M3 (M3 broken); UM 39932, R partial dentary M2; UM 39933, R partial dentary P3eM3, I1 (I1 broken) (Fig. 16HeJ). UM locality SC-62 (Cf-2): UM 82683, R partial dentary I1, P3eM1, alveoli for C1, M2e3. UM locality SC-4 (Wa-1): USNM 512211, L partial maxilla P4eM3 (P4, M3 broken); USNM 512212, R partial maxilla P3eM1, alveoli for M2e3; USNM 512213, R partial maxilla P3eM2, alveoli for P2, M3; USNM 512214, R partial maxilla P4eM2 (M1 broken); USNM 512215, R partial maxilla M1e2, alveoli for M3; USNM 512216, L partial maxilla M1e3; USNM 512217, L partial maxilla M2e3; USNM 512218, L partial maxilla M1e2; USNM 512219, L partial dentary P4eM1, alveoli for M2e3; USNM 512220, L partial dentary P3eM2, alveoli for C1, M3; USNM 512225, L partial dentary P3e4, R partial dentary P3e4 (broken), L partial maxilla P2e M2 (P2 broken); USNM 512231, R partial dentary P4eM1, alveoli for

M2e3; USNM 512232, R partial maxilla P3eM2, alveoli for P2, M3; USNM 512233, L partial maxilla P2eM2, alveoli for M3; USNM 512234, L partial maxilla M1 (M1 broken), alveoli for M2; USNM 512235, R partial maxilla M1e2, alveoli for M3; USNM 512236, L partial maxilla P4eM1, alveoli for M2; USNM 512237, L partial maxilla P2e4 (P4 broken), alveoli for C1; USNM 512238, R partial dentary P4eM1, alveoli M2e3; USNM 512242, L partial dentary M3; USNM 512248, R P4; USNM 512249, L P4 (broken); USNM 512251, R P4; USNM 512252, R M3; USNM 512253, R M1 or M2 (broken); USNM 512254, L M3; USNM 512255, R M3; USNM 512256, L M3 (broken); USNM 512257, L M3; USNM 512259, R M1 (fragment); USNM 512260, L M1; USNM 512263, R M2; USNM 512266, L M3; USNM 512267, L M3; USNM 512268, R M1; USNM 512269, R M1; USNM 512271, L M3; USNM 512273, R P4; USNM 512274, R M3; USNM 512275, L M3; USNM 512277, R M1; USNM 512280, L P4; USNM 512281, L M3; USNM 512282, L M3; USNM 512283, L M2; USNM 512285 L M2, USNM 512681 L P3; L M2 (broken); USNM 512682, R P3; USNM 512692, R maxilla P3, alveoli for C1eP2; USNM 512772, L M2 (broken); USNM 512778, R I1; USNM 512786, L M3; USNM 512804, L P4; USNM 516544, L dentary M3. Locality and horizon: UCM specimens were collected from the early Wasatchian (Wa-1 or Wa-2) of the Wasatch Formation, Powder River Basin, Wyoming. UM and USNM specimens were recovered from freshwater limestones from the late Paleocene (Cf1, Cf-2) of the Fort Union and (Cf-2, Cf-3) Willwood Formations, and early Eocene (Wa-1, Wa-2) of the Willwood Formation in the Clarks Fork Basin, Wyoming. All USNM specimens reported here were recovered from a freshwater limestone block that was collected near the early Eocene (Wa-1) UM fossil locality SC-4 (see Beard and Houde, 1989 and references therein). Emended diagnosis: Differs from T. tribos in being slightly larger in most tooth dimensions, lacking diastema between P2 and P3, having P3 with metacone (although variably expressed) with more enlarged protocone lobe, having P4 parastyle that is positioned higher relative to the protocone resulting in a shorter preparacrista, and having more bulbous molar cusps.

Figure 16. Micro-CT scan images of new specimens of Tinimomys graybulliensis from the Clarkforkian of the Clarks Fork Basin, Wyoming, captured using Avizo 6. UM 36199, R M2, in buccal (A), occlusal (B), lingual oblique (C), and lingual (D) views. UM 39883, L dentary I1eM1, M3, in buccal (E), lingual (F), and occlusal (G) views. UM 39933, R dentary P3eM3 in buccal (H), lingual (I), and occlusal (J) views. Scale bar ¼ 1 mm.

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

131

Figure 17. Micro-CT scan images of new specimens of Tinimomys graybulliensis from the Wasatchian of the Powder River Basin, Wyoming, captured using Avizo 6. UCM 76918, L maxilla P3e4, in buccal (A) and occlusal (B) views. UCM 101486, L maxilla P4eM3, in buccal (C) and occlusal (D) views. UCM 62749, R dentary I1, P4eM3, in buccal (E), lingual (F), and occlusal (G) views. UCM 70619, L dentary M3, in buccal (H), occlusal (I), oblique (J), and lingual (K) views. Scale bar ¼ 1 mm.

Description and comparison: Fourteen late Paleocene specimens of T. graybulliensis previously recovered from Clarkforkian limestones (Bloch, 2001) are reported here. Most of these specimens are isolated molars, although several nearly complete dentaries and upper incisors are also preserved. These teeth are quite similar to previously known specimens of T. graybulliensis, as are

the 54 new specimens of T. graybulliensis from the early Wasatchian that are also reported here. I1 and I2 of T. graybulliensis were previously only known from one specimen, USNM 461201 (Fig. 15A-C), from the early Wasatchian (Rose et al., 1993). A partial premaxilla with I1 and I2 (UM 39927; Fig. 15D-F) and an isolated I1 (UM 39929) of

132

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

Figure 18. Micro-CT scan images of a new specimen of Tinimomys graybulliensis (AeB) and specimens previously attributed to Myrmekomomys loomisi (CeF) from the Wasatchian of the Powder River Basin, Wyoming, captured using Avizo 6. UCM 62762, R dentary P4eM3, in buccal (A) and occlusal (B) views. UCM 48544, R dentary M1e3, in buccal (C) view, and M3 cropped in occlusal (D) view. UCM 52038, R dentary M1e2, in buccal (E) and occlusal (F) views. Scale bar ¼ 1 mm.

T. graybulliensis from the Clarkforkian (Cf-2) are morphologically very similar to the early Wasatchian specimen. Both I1s are too worn to assess how developed the posterocone or lateroconule were, yet the development of the anterocone and mediocone in UM 39927 is similar to that of USNM 461201 in having a relatively larger mediocone than that of D. szalayi (Fig. 15G-I) and isolated teeth from Big Multi Quarry (see Chester and Beard, 2012). An isolated I1 attributed to T. graybulliensis from the early Wasatchian (USNM 512778) is less worn than the Clarkforkian specimens, and shows a distinct lateroconule and a bulge in the place of the posterocone. Although this specimen is fairly narrow compared with what has been documented previously for T. graybulliensis, it still clearly represents the species and differs from D. szalayi in having a more pronounced mediocone and lateroconule. UM 36199 (Fig. 16AeD) is a right M2 that represents the earliest occurrence of Tinimomys (Cf-1). This tooth is larger than average for T. graybulliensis, but within the size range and morphological variability known for this species. The lack of apical wear on the crown, and the lack of roots and some borders of the crown, suggest that this tooth may have still been developing in the crypt. Three of the four more complete dentaries of T. graybulliensis from the Clarkforkian (Cf-2 and Cf-3) preserve enough of the lower jaw to measure mandibular depth below P4 (see Fig. 16EeJ). These specimens have slightly deeper dentaries than all measured Wasatchian specimens of T. graybulliensis from the Bighorn Basin. Clarkforkian specimens also have a P3 that is relatively small compared with M1 (yet they are within the lower end of the size range for Eocene T. graybulliensis). Eight new specimens of T. graybulliensis from the early Wasatchian of the Powder River Basin, Wyoming, are within the size range and morphological variability documented in the large sample of early Eocene (Wa-1) T. graybulliensis from UM locality SC-

4, Bighorn Basin, Wyoming. UCM 101486 is the most complete maxillary specimen of T. graybulliensis from the Powder River Basin, and it has continuous lingual cingula on P4eM3 with a hypocone and or pericone on P4eM2, which is diagnostic for Tinimomys (Fig. 17C,D). UCM 76918 is another maxilla of T. graybulliensis with P3e4 (Fig. 17A,B). Although this specimen lacks a P3 metacone like T. tribos and some specimens of T. graybulliensis, it is most similar to T. graybulliensis in overall tooth dimensions, in having a wide P3 protocone lobe, and in having a short and less vertical preparacrista on P4. Three partial dentaries, (UCM 62749, Fig. 17EeG; UCM 65916; UCM 76919) preserve at least one premolar or molar, as well as a partial I1 that is elongate and mediolaterally compressed as in T. graybulliensis. Two dentaries with P4eM3 (UCM 62749, Fig. 17Ee G; UCM 62762, Fig. 18A,B) represent the most complete dentaries of T. graybulliensis from the Powder River Basin and are extremely similar to specimens of this species from SC-4, although some have slightly greater mandibular depth under P4. UCM 70619 is a partial dentary of T. graybulliensis with an erupting M3 (Fig. 17HeK). The crowns of M1 and M2 have already erupted, but were broken away and are only represented by roots at present. The dentary is broken slightly mesial to the mesial root of M1, although the broken distal root of dP4 can be differentiated from the crown of P4 that was still developing in the crypt (Fig. 17J). The crown of M3 is unworn and similar to other specimens of T. graybulliensis in having three distinct trigonid cusps, but differs in having a slightly longer and broader hypoconulid (approaching the lobate condition of some paromomyids). Discussion: The late Paleocene sample of T. graybulliensis reported here includes the oldest known occurrence of this species (earliest Clarkforkian, Cf-1). This specimen, an isolated M2 (UM 36199), was recovered from a limestone from UM locality SC-179 in the Bighorn Basin, which until recently was the lowest documented occurrence of rodents on Polecat Bench, and possibly the oldest occurrence of rodents in North America (Bloch, 2001; although see Dawson and Beard, 1996; Secord, 2008). New Clarkforkian specimens of T. graybulliensis are very similar to those previously known from the early Eocene, although they appear to differ in having greater dentary depth below P4. Tinimomys is unique among micromomyids in having a dentary that gradually increases in depth until it reaches its maximum depth below P4, whereas other micromomyids have a dentary that is more uniform in depth (Beard and Houde, 1989). The new dentaries of T. graybulliensis from the Clarkforkian are even more exaggerated in mandibular depth below P4, although sample size is small (n ¼ 3). It is unclear whether this morphological difference relates to diet, but the crowns of the teeth preserved in these specimens do not appear to differ from those of Wasatchian specimens of T. graybulliensis in any appreciable way. New specimens of T. graybulliensis from the Wasatchian of the Powder River Basin are also very similar to those previously known from the Wasatchian of the Bighorn Basin. One notable difference is the slightly longer and considerably broader hypoconulid of M3 in UCM 70619 (Fig. 17I). We attribute this difference to variability given that this tooth position represents one of the more variable regions of the toothrow (Gingerich, 1974), and that this specimen is found co-occurring with another specimen of T. graybulliensis (UCM 62762; Fig. 18A,B) at locality UCM 2010009. Even with over one hundred micromomyid plesiadapiform specimens recovered from freshwater limestones, UCM 70619 is the first known micromomyid specimen with erupting teeth. UCM 70619 demonstrates that M1 and M2 erupt before M3, and that M3 erupts before P4. This specimen is between stages 2 and 3 of Bloch et al. (2002; see fig. 6) and suggests that micromomyids have a similar dental eruption sequence to that of paromomyid plesiadapiforms.

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

Validity of Myrmekomomys loomisi: Robinson (1994) proposed a new genus and species, M. loomisi, based on two specimens from the early Eocene of the Powder River Basin, Wyoming (Fig. 18CeF). He also reported three specimens of T. graybulliensis in the Powder River Basin, and one of these specimens was collected at the fossil locality where the holotype of M. loomisi was recovered. Robinson (1994) suggested that Myrmekomomys differs from Micromomys in having more lingually placed trigonids on the lower molars, and differs from all known micromomyids in having a more elongate M3 hypoconulid and higher crowned lower molars. He also noted that M. loomisi is similar to T. graybulliensis in having a trigonid that is lingually situated, and claimed that M. loomisi differs from T. graybulliensis in having relatively greater molar trigonid relief and a taller talonid, M1 larger than M2, M1e2 hypoconulids that are more developed, and an M3 hypoconulid that is relatively narrow, elongate, and tall (Robinson, 1994). However, Rose and Bown (1996) argued that characters used to distinguish Myrmekomomys from Tinimomys did not validate generic separation. New specimens of micromomyids reported here from the Powder River Basin, as well as the large Wa-1 sample of T. graybulliensis from UM locality SC-4 in the Bighorn Basin, allow us to assess the diagnosis of M. loomisi. All micromomyid specimens from the Powder River Basin that preserve lower molars have similarly oriented trigonids with lingually situated paraconids (except UCM 54585 that is attributed to C. antelucanus above). Robinson (1994) claimed that Tinimomys has relatively lowcrowned molars (M1 trigonid 1.5 times taller than talonid) with much less relief than those of Myrmekomomys (M1 trigonid two times taller than talonid). However, the only illustrated specimen that was considered to be T. graybulliensis is UCM 48620, a partial dentary with P4 and an extremely worn M1 (Robinson, 1994). Comparisons with less worn T. graybulliensis molars from the new Powder River Basin sample and the UM locality SC-4 sample show that the specimens attributed to M. loomisi are within the range of variation of molar height for T. graybulliensis (Fig. 18). Robinson (1994) also noted that M1 is larger than M2 in his diagnosis for M. loomisi, although this is also virtually always true for T. graybulliensis. The development of the hypoconulids on M1e2 of the two specimens previously attributed to M. loomisi also fall within the variability of this feature in the new Powder River Basin and SC-4 samples of T. graybulliensis (Fig. 18). The only feature present in ‘M. loomisi’ that is outside the variability documented for T. graybulliensis is the expression of the M3 hypoconulid. As noted by Robinson (1994), the M3 hypoconulid in the holotype of M. loomisi (UCM 48544; Fig. 18D) is narrower, elongate, and tall compared with the more broad, low-crowned M3 hypoconulid of T. graybulliensis. Specimens of T. graybulliensis (such as UCM 62749, Fig. 17EeG; 62762, Fig. 18A,B) and the second specimen originally referred to Myrmekomomys in the sample from the Powder River Basin (UCM 52038; Fig. 18E,F), have somewhat wider molars than those of the holotype of M. loomisi. The presence of such a narrow hypoconulid in the holotype of M. loomisi may relate to the general narrowness of the lower molars of this particular individual. The erupting M3 in UCM 70619 (Fig. 17I) has an especially broad hypoconulid that falls outside of the morphological variability known for T. graybulliensis. It may not be a coincidence that this is also the only feature that differs from the large samples of T. graybulliensis in the specimens originally attributed to M. loomisi. Again, expression of the M3 hypoconulid may be especially variable because it is a feature on a tooth locus that has been demonstrated to be quite variable in size (Gingerich, 1974). Additionally, all of the new specimens from the Powder River Basin, including upper teeth, appear to belong to T. graybulliensis and lack unique morphological features that one might expect if another distinct species was present. Dental measurements of all new

133

specimens from the Powder River Basin fall within the range documented in the SC-4 T. graybulliensis sample (Table 2). We therefore consider Myrmekomomys a junior synonym of Tinimomys, and consider the proposed species M. loomisi as a variant of T. graybulliensis. Stratigraphic record of tooth size in T. graybulliensis: The Clarks Fork Basin preserves a continuous late Paleocene and early Eocene stratigraphic section, with well-documented meter levels for UM fossil localities where specimens of T. graybulliensis have been collected (e.g., Gingerich, 2003; UM stratigraphic database). Results from plotting the area of molars (M1, M2, and M2) of T. graybulliensis against meter levels suggest that there was no significant shift in body size in this species from the early Clarkforkian to the early Wasatchian in the Bighorn Basin, Wyoming (Fig. 19). It might be expected for T. graybulliensis to exhibit relatively small body size or ‘dwarfing’ in Wa-0, the principal faunal zone of the global warming event known as the Paleocene Eocene Thermal Maximum (PETM), as this has been documented in many herbivorous and some carnivorous mammals during this interval (e.g., Gingerich, 1989, 2003; Heinrich et al., 2008; Chester et al., 2010; Secord et al., 2012). However, this cannot be evaluated because the two isolated lower molars previously attributed to T. graybulliensis from Wa-0 (Strait, 2001) represent the microsyopid Niptomomys (Fet and Strait, 2006; SGBC, Personal observation), and therefore no micromomyids are known from the PETM. It is surprising that T. graybulliensis has not yet been documented during the PETM given that this species is found in the bracketing faunal zones, Cf-3 and Wa-1, and because many specimens of small-bodied mammals including the plesiadapiform Niptomomys and euprimate Teilhardina have been discovered in PETM strata (e.g., Strait, 2001; Rose et al., 2011). T. tribos Chester and Beard, 2012 Holotype: CM 70062, R maxilla P4eM2. Known distribution: Middle Clarkforkian (Cf-2) of Big Multi Quarry, Washakie Basin, Wyoming. Diagnosis: Differs from T. graybulliensis in being slightly smaller in most tooth dimensions, having a short diastema between P2 and P3, lacking a metacone and having a mesiodistally narrower Table 2 Summary of dental measurements (in millimeters) for Tinimomys graybulliensis from SC-4. Tooth

N

OR

x

SD

CV

P3L P3W P4L P4W P4H M1L M1W M2L M2W M3L M3W MD P2L P2W P3L P3W P4L P4W M1L M1W M2L M2W M3L M3W

9 9 24 24 20 20 20 11 13 15 15 13 2 2 11 11 16 15 20 15 17 17 9 9

0.89e1.06 0.49e0.69 1.20e1.46 0.78e0.99 0.87e1.06 1.05e1.22 0.84e1.04 1.02e1.14 0.82e1.03 1.30e1.55 0.79e0.95 2.32e2.76 0.83e0.88 0.45e0.51 0.96e1.15 0.95e1.21 1.24e1.49 1.46e1.95 1.02e1.13 1.58e1.73 0.95e1.12 1.55e1.75 0.71e0.96 1.34e1.65

0.97 0.61 1.36 0.91 0.95 1.12 0.91 1.07 0.93 1.43 0.84 2.58 0.86 0.48 1.07 1.10 1.35 1.74 1.07 1.68 1.03 1.64 0.85 1.49

0.06 0.07 0.07 0.06 0.05 0.04 0.04 0.04 0.06 0.06 0.05 0.12 0.04 0.04 0.06 0.09 0.06 0.13 0.03 0.04 0.05 0.06 0.08 0.10

5.83 12.06 5.32 6.35 5.43 3.35 4.85 3.39 5.95 4.41 6.07 4.47 4.14 8.84 5.65 8.34 4.48 7.56 2.46 2.26 4.53 3.70 8.89 6.53

L, length; W, width; H, height; MD, mandibular depth below distal root of P4; N, sample size; OR, observed range; x, mean; SD, standard deviation; CV, coefficient of variation.

134

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

protocone lobe on P3, having less inflated paracone and metacone on P4eM2, and having a parastyle on P4 that is positioned lower relative to the paracone, resulting in a longer, more nearly vertical preparacrista. Micromomyid body mass estimation In the description of the first known micromomyid, M. silvercouleei, Szalay (1973) noted that the dentary and teeth of this taxon were smaller than those of the smallest extant primate, Microcebus murinus. It has since become well established that a distinctive characteristic of micromomyid plesiadapiforms is their very small size. Size likely played a major role in the physiology and ecology of these animals, so it is important to have accurate estimates of body mass for the species discussed here. If possible, it is best to estimate body mass for a fossil species based on equations derived from various anatomical regions since proportions of dental, cranial, and postcranial elements vary across taxa (Gingerich and Gunnell, 2005). Also, although teeth are commonly used for estimating body mass due to their abundance in the fossil record, previous studies have shown that postcranial measurements perform somewhat better than dental measurements for estimating body mass in various groups of primates (e.g., Delson et al., 2000). Body mass was first estimated based on the size of the lower first molar, which does not vary considerably compared with the size of other tooth positions (Gingerich, 1974). Choosing M1 allowed body mass to be calculated for eight of 11 micromomyid species. The ‘prosimian grade’ and ‘all primate’ regressions of Conroy (1987), and the ‘small mammals’ and ‘primate’ regressions of Legendre (1986) were used (see Table 3). Body mass was also estimated for the two micromomyid species for which dentally-associated postcrania are known (T. graybulliensis and D. szalayi). Humeral and femoral length and midshaft diameter regressions of Gingerich (1990) were used for T. graybulliensis and D. szalayi as an independent source to evaluate the resulting values of the four regression equations for M1. Isolated humeri previously attributed to T. graybulliensis and C. antelucanus (Beard, 1989) are not considered here because they differ significantly in morphology from dentally-associated humeri of T. graybulliensis and D. szalayi, and likely do not belong to Micromomyidae. The results of all four M1 regression equations suggest that micromomyid plesiadapiforms were approximately 4e40 g. The ‘prosimian grade’ and ‘all primate’ regression equations of Conroy (1987) produced fairly similar intermediate values (9e16 g), whereas the ‘small mammals’ and ‘primate’ regression equations of Legendre (1986) produced considerably smaller (4e7 g) and larger (24e40 g) values, respectively. Previously estimated weights of T. graybulliensis (81 g and 86 g; Conroy, 1987) are too high because the incorrect logged value is listed for M1 area (the same value is correctly listed for the larger euprimate Teilhardina americana). Results calculated in this study suggest a general trend with more primitive micromomyids (Foxomomys and Chalicomomys) being smaller than more derived taxa such as Dryomomys and Tinimomys (see cladistic results below). Also, within these more derived genera, the older and more primitive species such as D. millennius and T. tribos tend to be smaller than the younger and more derived species, such as D. szalayi and T. graybulliensis. However, it should be noted that these apparent trends are based on very small sample sizes of M1 for several species of micromomyids (Table 3). Figure 19. Stratigraphic record of tooth size of Tinimomys graybulliensis from the early Clarkforkian to the early Wasatchian in the Clarks Fork Basin, Wyoming. M1 (A), M2 (B), and M2 (C) are tooth positions that are known not to vary considerably in size. These tooth loci were chosen to illustrate that tooth size of T. graybulliensis in the latest Paleocene and early Eocene falls within the range of variation documented in the large sample of early Eocene T. graybulliensis from UM locality SC-4 (Wa-1; 1570 m). Mean is represented by black squares. Error bars show two standard deviations from the mean.

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

135

Table 3 Summary of body mass estimates (g) for species of micromomyid plesiadapiforms based on regressions of average lower first molar area. Taxon

N

M1L x

M1L SD

M1W x

M1W SD

Prosimian Reg. (g)

All Primate Reg. (g)

Small Mammal Reg. (g)

Primates Reg. (g)

F. fremdi F. vossae F. gunnelli M. silvercouleei C. antelucanus D. millennius D. dulcifer D. szalayi D. willwoodensis T. tribos T. graybulliensis

2 1 e e 3 1 4 1 e 5 33

1.13 1.08 e e 1.04 1.11 1.06 1.11 e 1.03 1.13

e e e e 0.03 e 0.09 e e 0.04 0.04

0.78 0.76 e e 0.80 0.90 0.96 0.96 e 0.90 0.93

e e e e 0.02 e 0.06 e e 0.03 0.05

12 11 e e 11 14 15 16 e 13 16

10 9 e e 9 13 13 14 e 11 14

5 4 e e 4 6 6 7 e 5 6

28 24 e e 25 36 37 40 e 31 39

Average first molar area was calculated by multiplying the first lower molar maximum length (M1L) and width (M1W). Standard deviation (SD) was calculated for M1L and M1W, however, this does not account for error of the regression coefficients. Prosimian Reg., Prosimian regression of Conroy (1987); All Primate Reg., All-primates regression of Conroy (1987); Small Mammal Reg., Small mammals (1.85); (1) intermediate (1.40e 1.85); (2) wide (1.55); (1) intermediate (1.40e 1.55); (2) wide (1.4); (1) wide (2 mm2); (1) M1 or M2 area ( 1.65 and or P4/M2 > 1.90). 25. P4 metacone: (0) distobuccal to paracone; (1) buccal to paracone. 26. P4 protocone: (0) lingually situated; (1) buccally situated. 27. P4eM2 lingual cingulum: (0) absent; (1) present. 28. M1 to M2 width: (0) M2 relatively elongate; (1) similar in width.

Appendix D. Supplementary material Supplementary material associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.jhevol. 2013.04.006.

142

S.G.B. Chester, J.I. Bloch / Journal of Human Evolution 65 (2013) 109e142

Gingerich, P.D., 1989. New earliest Wasatchian mammalian fauna from the Eocene of northwestern Wyoming: composition and diversity in a rarely sampled highfloodplain assemblage. Univ. Mich. Pap. Paleontol. 28, 1e97. Gingerich, P.D., 1990. Prediction of body mass in mammalian species from long bone lengths and diameters. Contrib. Mus. Paleontol. Univ. Mich. 28, 79e92. Gingerich, P.D., 2003. Mammalian responses to climate change at the Paleocenee Eocene boundary: Polecat Bench record in the northern Bighorn Basin, Wyoming. In: Wing, S.L., Gingerich, P.D., Schmitz, B., Thomas, E. (Eds.), Causes and Consequences of Globally Warm Climates in the Early Paleogene. Geol. Soc. Am. Spec. Pap., vol. 369. Boulder, pp. 463e478. Gingerich, P.D., Gunnell, G.F., 2005. Brain of Plesiadapis cookei (Mammalia, Proprimates): surface morphology and encephalization compared to those of Primates and Dermoptera. Contrib. Mus. Paleontol. Univ. Mich. 31, 185e195. Gingerich, P.D., Smith, B.H., Rosenberg, K., 1982. Allometric scaling in the dentition of primates and prediction of body weight from tooth size in fossils. Am. J. Phys. Anthropol. 58, 81e100. Gunnell, G.F., 1989. Evolutionary history of Microsyopoidea (Mammalia, ?Primates) and the relationship between plesiadapiforms and primates. Univ. Mich. Pap. Paleontol. 27, 1e157. Heinrich, R.E., Strait, S.G., Houde, P., 2008. Earliest Eocene Miacidae (Mammalia: Carnivora) from northwestern Wyoming. J. Paleontol. 82, 154e162. Kay, R.F., 1984. On the use of anatomical features to infer foraging behavior in extinct primates. In: Rodman, P.S., Cant, J.G.H. (Eds.), Adaptations for Foraging in Nonhuman Primates: Contributions to an Organismal Biology of Prosimians, Monkeys and Apes. Columbia University Press, New York, pp. 21e53. Krause, D.W., 1978. Paleocene primates from western Canada. Can. J. Earth Sci. 15, 1250e1271. Legendre, S., 1986. Analysis of mammalian communities from the late Eocene and Oligocene of southern France. Palaeovertebrata 16, 191e212. Lillegraven, J.A., Kielan-Jaworowska, Z., Clemens, W.A. (Eds.), 1979. Mesozoic Mammals: the First Two-thirds of Mammalian History. University of California Press, Berkeley. Linnaeus, C., 1758. Systema naturae: per regna tria naturae, secundum classis, ordines, genera, species cum characteribus, differentiis, synonymis, locis: edito decimal. In: Reformata, vol. 1. Laurentii Salvii, Stockholm. Maddison, D.R., Maddison, W.P., 2003. MacClade 4: Analysis of Phylogeny and Character Evolution, Version 4.0.6. Sinauer, Sunderland. Olson, L.E., Sargis, E.J., Martin, R.D., 2005. Intraordinal phylogenetics of treeshrews (Mammalia: Scandentia) based on evidence from the mitochondrial 12S rRNA gene. Mol. Phylogenet. Evol. 35, 656e673. Roberts, T.E., Lanier, H.C., Sargis, E.J., Olson, L.E., 2011. Molecular phylogeny of treeshrews (Mammalia: Scandentia) and the timescale of diversification in Southeast Asia. Mol. Phylogenet. Evol. 60, 358e372. Robinson, P., 1994. Myrmekomomys, a new genus of Micromomyine (Mammalia, ? Microsyopidae) from the lower Eocene rocks of the Powder River Basin, Wyoming. Contrib. Geol. Univ. Wyom. 30, 85e90. Rose, K.D., 1975. The Carpolestidae, early Tertiary primates from North America. Bull. Mus. Comp. Zool. 147, 1e74. Rose, K.D., 1981. The Clarkforkian land-mammal age and mammalian faunal composition across the PaleoceneeEocene boundary. Univ. Mich. Pap. Paleontol. 26, 1e197. Rose, K.D., Bown, T.M., 1982. New Plesiadapiform primates from the Eocene of Wyoming and Montana. J. Vert. Paleontol. 2, 63e69. Rose, K.D., Bown, T.M., 1996. A new plesiadapiform (Mammalia: Plesiadapiformes) from the early Eocene of the Bighorn Basin, Wyoming. Ann. Carnegie Mus. 65, 305e321. Rose, K.D., Beard, K.C., Houde, P., 1993. Exceptional new dentitions of the diminutive plesiadapiforms Tinimomys and Niptomomys (Mammalia), with comments on the upper incisors of Plesiadapiformes. Ann. Carnegie Mus. 62, 351e361. Rose, K.D., Chester, S.G.B., Dunn, R.H., Boyer, D.M., Bloch, J.I., 2011. New fossils of the oldest North American euprimate Teilhardina brandti (Omomyidae) from the PaleoceneeEocene Thermal Maximum. Am. J. Phys. Anthropol. 146, 281e305.

Sargis, E.J., 2001. A preliminary qualitative analysis of the axial skeleton of tupaiids (Mammalia, Scandentia): functional morphology and phylogenetic implications. J. Zool. 253, 473e483. Sargis, E.J., 2002a. Functional morphology of the forelimb of tupaiids (Mammalia, Scandentia) and its phylogenetic implications. J. Morphol. 253, 10e42. Sargis, E.J., 2002b. Functional morphology of the hindlimb of tupaiids (Mammalia, Scandentia) and its phylogenetic implications. J. Morphol. 254, 149e185. Scott, C.S., 2008. Late Paleocene mammals from near Red Deer, Alberta, and a phylogenetic analysis of the earliest Lipotyphla (Mammalia, Insectivora). Ph.D. Dissertation, University of Alberta. Secord, R., 2008. The Tiffanian land-mammal age (middle and late Paleocene) in the Northern Bighorn Basin, Wyoming. Univ. Mich. Pap. Paleontol. 35, 1e192. Secord, R., Gingerich, P.D., Smith, M.E., Clyde, W.C., Wilf, P., Singer, B.S., 2006. Geochronology and mammalian biostratigraphy of middle and upper Paleocene continental strata, Bighorn Basin, Wyoming. Am. J. Sci. 306, 211e245. Secord, R., Bloch, J.I., Chester, S.G.B., Boyer, D.M., Wood, A.R., Wing, S.L., Kraus, M.J., McInerney, F.A., Krigbaum, J., 2012. Evolution of the earliest horses driven by climate change in the PaleoceneeEocene thermal maximum. Science 335, 959e962. Silcox, M.T., 2001. A phylogenetic analysis of plesiadapiformes and their relationship to Euprimates and other archontans. Ph.D. Dissertation, Johns Hopkins University School of Medicine. Silcox, M.T., Gunnell, G.F., 2008. Plesiadapiformes. In: Janis, C.M., Gunnell, G.F., Uhen, M.D. (Eds.), Evolution of Tertiary Mammals of North America. Small Mammals, Xenarthrans, and Marine Mammals, vol. 2. Cambridge University Press, Cambridge, pp. 207e238. Silcox, M.T., Williamson, T.E., 2012. New discoveries of early Paleocene (Torrejonian) primates from the Nacimiento Formation, San Juan Basin, New Mexico. J. Hum. Evol. 63, 805e833. Silcox, M.T., Rose, K.D., Walsh, S., 2002. New specimens of picromomyids (Plesiadapiformes, ?Primates) with description of a new species of Alveojunctus. Ann. Carnegie Mus. 71, 1e11. Silcox, M.T., Bloch, J.I., Boyer, D.M., Houde, P., 2010. Cranial anatomy of Paleocene and Eocene Labidolemur kayi (Mammalia: Apatotheria), and the relationships of the Apatemyidae to other mammals. Zool. J. Linn. Soc. Lond. 160, 773e825. Simpson, G.G., 1955. The Phenacolemuridae, new family of early primates. Bull. Am. Mus. Nat. Hist. 105, 411e442. Strait, S.G., 2001. New Wa-0 mammalian fauna from castle gardens in the southeastern Bighorn Basin. In: Gingerich, P.D. (Ed.), PaleoceneeEocene Stratigraphy and Biotic Change in the Bighorn and Clarks Fork Basins, Wyoming. Univ. Mich. Pap. Paleontol. 33, 127e143. Swofford, D.L., 2003. PAUP*: Phylogenetic Analysis Using Parsimony (*and Other Methods). Version 4. Sinauer, Sunderland. Szalay, F.S., 1973. New Paleocene primates and a diagnosis of the new suborder Paromomyiformes. Folia Primatol. 19, 73e87. Szalay, F.S., 1974. A new species and genus of early Eocene primate from North America. Folia Primatol. 22, 243e250. Szalay, F.S., 1976. Systematics of the Omomyidae (Tarsiiformes, Primates): taxonomy, phylogeny, and adaptations. Bull. Am. Mus. Nat. Hist. 156, 157e450. Szalay, F.S., Delson, E., 1979. Evolutionary History of the Primates. Academic Press, New York. Tong, Y., Wang, J., 1998. A preliminary report on the early Eocene mammals of the Wutu fauna, Shandong Province, China. In: Beard, K.C., Dawson, M.R. (Eds.), Dawn of the Age of Mammals in Asia. Bull. Carnegie Mus. Nat. Hist. 34, 186e193. Van Valen, L.M., 1994. The Origin of the Plesiadapid Primates and the Nature of Purgatorius. Univ. Chicago Evol. Monographs 15, 1e179. Wilf, P., Beard, K.C., Davies-Vollum, K.S., Norejko, J.W., 1998. Portrait of a late Paleocene (early Clarkforkian) terrestrial ecosystem: big multi quarry and associated strata, Washakie Basin, southwestern Wyoming. Palaios 13, 514e532.