Hesperolagomys niobrarensis, sp. nov.

Holotype. UNSM 123305, dentary with i-m2 and alveolus for m3, from Garner Bridge South locality, Nebraska, in Crookston Bridge Member, Valentine Formation, from deposits of late late Barstovian age (Ba2c).

Hypodigm. Type; topotypes: (all UNSM numbers) 123306, partial dentary with p4-m3, alveolus for p3; 123307, partial maxilla with P3-P4, alveolus for P2. From Valentine Railway Quarries, by locality: West Valentine Quarry: 122500, partial maxilla with P3-P4; 122509, 122719, partial maxillae with P4; 123348, edentulous partial maxillary; 122996, upper anterior incisors; 122516-122526, 122543-122545, 122830, P3s; 122630, 122634-122641, 122643-122662, 122673, 122745-122751, P4s; 122700-122701, 122703-122710, 122714, 122718, 122774, 122801-122808, M1s; 122809-122814, 122817, 122819-122820, 122823, M2s; 122512, partial dentary with p3-m3; 122501, partial dentary with p4-m3 and alveolus for p3; 122502, partial dentary with p4-m2 and alveolus for m3; 122503-122504, 122510, and 122513-122514, partial dentaries with m1-m2; 122515, partial dentary with m1; 122511, partial dentary with m2; 122993, partial right lower incisor; 122838, 122859, p3s; 122878, 122881, 122891-122907, 122943, 122945, 122951, p4s; 122908, 122940-122942, 122944, 122946, 122952-122953, 122986, m1s; 122983-122985, 122987-122992, m2s. Stewart Quarry: 123021-123030, P3s; 123052-123068, 123081, P4s; 123082-123088, 123103, M1s; 123097, 123099, M2s; 123104, partial upper cheek tooth; 123003, partial dentary with p4; 123140-123145, p4s; 123162-123163, m1; 123177, m2. Valentine Quarry 'B': 123186, partial maxilla with P4; 123300, p4.

Referred Specimen from Nebraska. UNSM 45206, m1, from Myers Farm.

Etymology. niobrarensis; after 'Niobrara', for the abundant Miocene fossiliferous deposits exposed along the Niobrara River and tributaries in northern Nebraska deposits, from which specimens of this species were first recognized.

Diagnosis. Larger than other species of Hesperolagomys. Talonid moderately wide compared to trigonid in lower molariform teeth, intermediate between that of H. fluviatilis and H. galbreathi (talonid width/trigonid width of p4 ≈1.41). Additionally distinguished from H. fluviatilis by presence of anterolingual groove on p3. Also differs from H. galbreathi in possessing a weaker (but cement-filled) anterolingual groove on p3 and more asymmetric talonid on lower molariform teeth.

Geologic Age and Distribution. Late late Barstovian (informally designated Ba2c) of Nebraska (see discussion in Biochronology Context section below).

Description

Skull

Four partial maxillae of H. niobrarensis provide new information on skull structure. The specimens show an overall similarity to Oreolagus and Ochotona in features of the anterior zygoma and maxilla but differ notably in having rooted upper cheek teeth that occupy a relatively greater occlusal area. The specimen from Garner Bridge South and two of the three from the Valentine Railway Quarries (Figure 3.1, 3.2, 3.3) each have a small premolar foramen located medial to and just posterior of P3 (the relevant portion of the palate is not preserved in the other Valentine Railway Quarries specimen); the presence of this character indicates affinity with the Ochotonidae (Dawson 2008; McKenna 1982; but see also discussion below). As seen in ventral view, the anterior zygomatic root extends from a point anterior to P3 to a line between P4 and M1, leaving the shafts of M1 and M2 free of the zygoma (Figure 3.1-3.4). The preserved portion of the anterior zygoma has a shallow concavity on the lateral surface nearly identical to that in Oreolagus, and much shallower and less well defined than the zygomasserteric fossa in Ochotona princeps.

Upper Dentition

Upper Incisors. Two anterior upper incisors (one shown in Figure 3.5) are referred to H. niobrarensis, primarily because they closely resemble specimens of H. galbreathi described by Clark et al. (1964) in both size and morphology, with a longitudinal groove on the anterior surface displaced slightly medially of center. Both incisors are slightly anteroposteriorly longer (Table 1A) than those of H. galbreathi in Clark et al. (1964), and both lack the posterior portion and are slightly damaged at the wear facet. The teeth are similar in size and morphology to those of Ochotona princeps but the groove on the anterior surface is slightly shallower and less angular, and the tooth is more compressed laterally. Both Hesperolagomys and Ochotona anterior upper incisors have elongate wear facets and sharp tips compared with those of Sylvilagus and Lepus, which have shorter facets and less acuminate tips.

P2. P2 is not known, but an alveolus for this tooth suggests that it had a single (lingual) root and was much smaller than the other upper cheek teeth (Figure 3.3; Table 1A).

P3. P3 in anatomical position in UNSM 123307 (Figure 3.1) and its alveolus in UNSM 123348 (Fig. 3.3) allow confident identification of this tooth in Hesperolagomys for the first time. 26 P3s are known for H. niobrarensis, allowing documentation of variation and wear stages (Figure 4). Two buccal roots are present; the anterior one is shifted lingually compared with its position in more posterior cheek teeth, such that the tooth is nonmolariform and anteroposteriorly shorter buccally than lingually. The disposition of stria and overall gracile structure of P3 is unique to Hesperolagomys. In anterior view (see Figure 4.1-4.3; 4.5), two cement-filled striae are present: an anterolingual stria close to and paralleling the curvature of the lingual margin, and a more centrally-located anterior stria, also paralleling the outer curvature of the tooth and extending nearly to the base of the crown. A third and shorter stria is often but not always cement-filled, and merges with the anterior stria (Figure 4.3). In lingual view (Figure 4.2), two cement-filled striae (the anterolingual stria and a more posteriorly-located lingual hypostria) extend to near the base of the crown (more easily seen in isolated specimens). In occlusal view, the anterior stria corresponds to the anterior limb of the crescentic valley, and the cement-filled anterolingual stria and lingual hypostria separate the lingual edge of P3 into three subequal rounded lingual lobes (see Figure 2.1, 2.2 for a schematic view of these relationships). The cement-filled crescentic valley is separated from the anterolingual stria by only a narrow bridge of dentine and enamel. The crescent forms the anterior edge of the central part of the tooth.

Isolated teeth demonstrate a slight curvature in the anteroposterior plane, concave to posterior (illustrated in Figure 4.2). In the youngest individuals (Figure 4.1) the crown base is still forming, and the enamel-dentine junction extends beyond perpendicular relative to the slightly worn occlusal surface. Three distinct cement-filled striae are visible in anterior view (e.g., Figure 4.1 and 4.2). The lingual most is the anterolingual stria, separated from the anterior stria by the precone column; in these specimens, the buccal stria is completely separated from the cement of the anterior stria (but see Figure 4.3 for a slight connection). The anterolingual and anterior striae nearly merge towards the crown apex. In occlusal view, the precone is topographically lower than the rest of the lingual quarter of the tooth, but is worn and connected to the lingual cusps by a thin portion of dentine. A small rounded area of the crescent is worn through to dentine, with the anterior portion topographically lower and unworn. The anterior limb of the crescentic valley projects nearly directly anteriorly into the anterior stria from the posterobuccally-directed posterior limb. The buccal valley is large and primarily unworn, separating the crescent from a small peninsula of exposed dentine on the posterobuccal corner, with a large anterobuccal area topographically lower and unworn.

In a somewhat older individual, (Figure 4.2) crown formation is complete, occlusal width is greater, and areas of dentine exposed at the posterobuccal corner and in the crescent are expanded. The precone is worn to the same topographic level as other lingual cusps. In the illustrated specimen, cement of the buccal and anterior striae are separated, but a depression connects the striae. With increasing wear stage (Figure. 4.3-4.4), transverse width at the occlusal surface increases while length stays relatively constant. The crescent, posterobuccal cusp, and precone become transversely elongated, and greater areas of dentine are exposed. The precone becomes more buccally-directed, and cement-filled valleys are diminished.

The most worn well-preserved P3 of H. niobrarensis (122517: Figure 4.5) has a crown height of approximately one half that of the least worn specimen and is approximately twice as wide at the occlusal surface as the least worn examples. Both lingual striae persist nearly to crown base, although the lingual hypostria is relatively shallow and lacks cement at the occlusal surface. The crescentic valley is present in P3 even in highly worn specimens. Several additional specimens lack the buccal edge and posterobuccal root; this area between the buccal roots appears to be the thinnest part of the crown and prone to breakage at advanced stages of wear.

Most notable morphologic variation in P3 appears to be related to stage of wear; however, the separation or connection of cement in the buccal and anterior striae varies independent of wear. Mean size of P3 of H. niobrarensis is: length, 1.41 mm; width, 2.94 mm (Table 1A).

P4. P4s of H. niobrarensis at similar stages of wear strongly resemble the two P4s included in the original hypodigm of H. galbreathi, differing primarily in their larger size. P4 is the largest and tallest crowned upper cheek tooth of Hesperolagomys, characterized by an anteroposteriorly-oriented buccal margin, anteroloph wider than posteroloph, and a persistent (extending to or nearly to crown base) cement-filled lingual hypostria that is shallower than that of M1 and M2 (compare Figure 5, Figure 6, and Figure 7). The tooth is strongly curved buccolingually, but is nearly straight in the anteroposterior plane (Figure 5.3). The crown is incompletely formed, exhibiting incipient hypselodonty (wear occurring before completion of the crown base). A broad range of wear stages is represented among the more than 60 specimens available, demonstrating dramatic changes in occlusal morphology and width through wear.

A nearly unworn P4 (Figure 5.1) is characterized by a very tall crown, a large open crown base in which enamel formation is incomplete, absence of wear at the buccal margin of the occlusal surface, extensive cement filling large valleys, a topographically tall and barely worn crescent with a thin line of dentine exposed, and a narrow occlusal width. The buccal and crescentic valleys have confluent cement, and the buccal cusps are low. At this stage, the lingual hypostria is anteroposteriorly elongate and buccolingually narrow at the occlusal surface.

P4s exhibiting progressively more wear (e.g., Figure 5.2-5.3) have completely formed crowns and less cement in valleys. The occlusal surface widens through wear, crown elements are buccolingually stretched, and the lingual hypostria becomes absolutely wider. Further wear (e.g., Figure 5.4-5.5) produces further buccal reduction in cement and buccolingual stretching of elements at the occlusal surface. The lingual hypostria curves posteriorly towards the crown base; as the occlusal surface intersects the crown closer to the base, the lingual portion of the anteroloph expands at the expense of the posteroloph, and the anterior wall of the tooth becomes more curved. The lingual hypostria terminates in a slightly rounded point buccally. At its maximum (Figure 5.5) the lingual hypostria extends across no more than one third of the occlusal width of the tooth.

The most worn well-preserved P4 is UNSM 122643 (Figure 5.6), which has only a small portion of enamel remaining on the posterior wall of the tooth, shows extreme buccolingual stretching of the occlusal surface (width nearly 2.5 times that of the youngest individual), with only the extreme buccal portions of both limbs of the crescentic valley retaining enamel, and the lingual hypostria very shallow but still containing some cement.

Mean size of P4 of H. niobrarensis is: length, 1.63 mm; width, 3.43 mm (Table 1A). The composite 'complete' specimen (Figure 5.7) has an estimated crown height of 11.69 mm and an estimated hypsodonty index of 7.17.

M1. M1 is smaller and slightly shorter-crowned than P4, such that it experiences substantial but slightly less dramatic changes in occlusal size and shape with wear (Figure 6.1-6.5). Other differences from P4 are: (1) an expanded lingual hypostria, which extends relatively and absolutely farther buccally across the occlusal surface to at least one third its width, and terminates very near the crescentic valley in a squared-off groove, (2) buccal cusps are more prominent in M1 (such that cement is more rapidly lost from the buccal valleys, and may never have been confluent as in P4), and (3) lingual portion of the crescentic valley retains cement even through the latest stages of wear. In the most worn M2s, (Figure 6.5), the crescentic valley can become separated into lingual and buccal portions, separated by dentine. Mean size of M1 of H. niobrarensis is: length, 1.42 mm; width, 3.03 mm (Table 1A).

M2. M2s of Hesperolagomys are not currently known in anatomical position in maxillary fragments with other cheek teeth; however, eight upper cheek teeth of H. niobrarensis are distinct from previously described upper cheek teeth of Hesperolagomys and are here identified as M2s. In comparison with M1, in M2 the posterobuccal portion of the crown is reduced, such that the medial edge of the crown tapers anteroposteriorly, and the anteroloph is much wider than the posteroloph. The most noticeable characteristic of the occlusal surface is the expanded hypostria, which extends across approximately 80% the width of the occlusal surface and is separated from the crescentic valley by a very thin strip of dentine. Crown elements buccal to the crescentic valley are much reduced in comparison with their size in P4 and M1. The lingual root is more strongly posteriorly curved than P4 and M1.

Three M2s are illustrated in Figure 6.6-6.8 and illustrate variation. Based on available specimens, M2 appears to have less dramatic changes to occlusal morphology with wear than more anterior cheek teeth. The wear series reveals that though wear M2 increases in transverse width at the occlusal surface, and the lingual hypostria extends over progressively more of the occlusal surface as the crescentic valley is reduced. Figure 6.6 and 6.7 are M2s at a similar stage of wear, but showing some differences not associated with ontogeny: variation in anteroloph width compared with posteroloph width (although anteroloph always wider, the extent to which this is the case varies), the buccal termination of the lingual hypostria is either squared off (Figure 6.7) or may expand into a bifurcation (as in Figure 6.6 and 6.8). Mean size of M2 of H. niobrarensis is: length, 1.42 mm; width, 2.72 mm (Table 1A).

Dentary

The holotype of H. niobrarensis (UNSM 123305: Figure 7.1) is the most complete dentary available, lacking only the dorsal portion of the ascending ramus and m3. It closely resembles previously described jaws of H. galbreathi overall shape (Clark et al. 1964). The number and relative size of mental foramina are consistent with those in the type specimen of H. galbreathi in Clark et al. (1964), but their position differs somewhat. The largest mental foramen is on the lateral surface ventral to the border of p3-p4, slightly posterior of its position in the type specimen of H. galbreathi (in line with and beneath p3). A depression with two smaller foramina is beneath the talonid of m1, very similar to the condition of the type specimen of H. galbreathi (below m1-m2). The smallest and anterior-most foramen is anterior to p3, positioned more anterodorsally than the anterior-most foramen in H. galbreathi and just ventral to p3.

The posterior portion of the horizontal ramus is more completely preserved than for other known species of Hesperolagomys. The ventral border is concave beneath the anterior margin of the masseteric fossa. The preserved portion of the masseteric fossa resembles that of Ochotona. The incisor capsule forms a bulge on the medial wall, and a small portion is damaged, revealing the incisor extending posteriorly to approximately the middle of m1. The roots of p4, partially visible through a small area of damage to the lateral wall of the dentary, are displaced laterally from the incisor.

Additional dentaries of H. niobrarensis (Figure 7.2-7.4; Table 2) closely resemble the holotype in morphology but vary somewhat in size, in the size and position of the anterior-most mental foramina, and in the relative depth of the jaw. The anterior-most foramen is more ventrally located in the three additional figured specimens than in the type. The type specimen is shallower at m1 than other figured specimens, but is not anomalous compared with values for all referred specimens (Table 2).

Lower Dentition

Lower Incisor. The lower incisor in UNSM 123305 (Figure 7.1) is slightly larger than but otherwise similar in cross-sectional morphology to type material of H. galbreathi (Table 1B; Clark et al. 1964, table 2). The wear facet has a rounded triangular shape, in which the medial edge comes to a sharp point whereas the lateral edge is more rounded, similar to that of Ochotona princeps. An isolated left lower incisor, UNSM 122993, lacks both ends but cross-sectional morphology and size is nearly identical to that of the type. Mean size of H. niobrarensis lower incisor is: length, 1.44 mm; width, 1.46 mm (Table 1B).

p3. The p3 in the type specimen has a shallower cement-filled anterolingual groove than in the type specimen of H. galbreathi and additionally differs in having a small anterior-facing projection in the thin enamel of the anterobuccal wall of the talonid (Figure 7.1). Three other specimens exhibit wear-related and individual differences from the holotype. The least worn (Figure 8.1) is similar in occlusal pattern to p3 in the type specimen and likewise has a minor inflection of the anterobuccal talonid wall. The remaining p3s (Figure 7.3, Figure 8.2) are shorter-crowned, and in both the anterolingual wall is slightly concave and bears traces of cement (not visible in figures). The enamel is thicker at the occlusal surface, and the trigonid is more anteriorly oriented in both worn specimens. Mean size of H. niobrarensis p3 is: length, 1.34 mm; width, 1.48 mm (Table 1B).

Lower Molariform Teeth. Teeth in the holotype, in referred dentaries, and isolated lower molariform teeth are like those of the type specimen of H. galbreathi in having trigonids transversely wider and anteroposteriorly shorter than talonids, and in the buccolingually narrow anterior projections of talonids (Clark et al. 1964). Lower molariform teeth can be distinguished in the majority of cases by their relative size and anteroposterior curvature, with p4 the largest and straightest tooth (with no anteroposterior curvature), m1 the smallest and intermediate in curvature, and m2 intermediate in most dimensions and with the strongest curvature. In early stages of wear, all three teeth have small posterolophids isolated from the main body of the talonid by cement (Figure 8.3, 8.9, and 8.13). Clark et al. (1964) noted the presence of completely isolated posterolophids on p4 and m1 of a referred partial jaw of a young individual of H. galbreathi (UCMP 29626; originally tentatively identified by Hall 1930, figure 29, as Sylvilagus? sp.).

p4. An extensive sample of p4s illustrates how the occlusal pattern of this tooth varies with increasing wear, particularly in the shape of the lingual edge of the anterior projection of the talonid. In earliest wear stages (Figure 8.3), the anterior projection of the talonid has a nearly straight lingual edge merging smoothly into the main body of the talonid; the buccal edge is concave and crenulated. A relatively large posterolophid is completely isolated at the occlusal surface from the main body of the talonid, and is a short ovoid shape (appearing buccolingually wider than in H. galbreathi as figured in Hall (1930, figure 29). In slightly more worn teeth (Figure 8.4) the posterolophids are buccally connected to the main body, but are still separated lingually by cement. In specimens at slightly more advanced stages of wear (Figure 8.5-8.6) posterolophids have been obliterated, filling out the tooth in the posterolingual corner. In these specimens, the lingual wall of the anterior projection of the talonid has become slightly concave; the buccal edge of the anterior project of the talonid shows little change with wear and remains concave and is often crenulated. The most worn p4s (Figure 8.7-8.8) show a slight reduction of the posterolingual corner of the talonid; the posterior talonid wall is more curved, and the lingual wall of the anterior projection is more concave, such that the talonid is more buccolingually symmetrical than in earlier wear. Mean size of p4 of H. niobrarensis is: length, 1.90 mm; trigonid width, 2.15 mm; talonid width, 1.52 mm (Table 1B). The ratio of trigonid width to talonid width in p4 is on average ≈1.41. The tallest-crowned p4 has a crown height of 5.18 mm, resulting in a hypsodonty index of 2.96.

m1. The m1 of H. niobrarensis is more curved anteroposteriorly than p4 and has a shorter talonid, similar to that of the talonid of p4 in early stages of wear (before the posterolophid is fully incorporated into the main body). No completely unworn m1 is known but a slightly worn example (Figure 8.9) has a minor cement-filled posterolingual groove marking the point of attachment of a small posterolophid. The anterior projection of the talonid varies with wear (Figure 8.10-8.12) but perhaps less so than in p4, the lingual wall of the anterior projection of the talonid being nearly always straight or slightly convex. Mean size of Valentine Railway Quarries + Garner Bridge South combined sample of m1 of H. niobrarensis is: length, 1.69 mm; trigonid width, 2.01 mm; talonid width, 1.37 mm (Table 1B).

UNSM 45206 from Myers Farm, Nebraska, is an m1 of Hesperolagomys referred to H. niobrarensis. This specimen is in an extremely early stage of wear, as evidenced by a very small posterolophid completely isolated from the talonid and topographically well below the occlusal surface. Occlusal morphology closely resembles little worn m1s of H. niobrarensis (e.g., Figure 8.10). Measured dimensions of this specimen are well within the range of values of the Valentine Railway Quarries and Garner Bridge South combined sample.

m2. Second lower molars of H. niobrarensis can be distinguished from p4 and m1 by their strong anteroposterior curvature, and, in occlusal view, by nearly straight posterior talonid walls with squared posterobuccal corners. An m2 in an early stage of wear (Figure 8.13) has a very small, completely isolated posterolophid. Little variation in the talonid shape is notable through wear, although the most worn m2 (Figure 8.18) has a more concave lingual wall of the anterior projection than other specimens (Figure 8.13-8.17). Mean size of Valentine Railway Quarries and Garner Bridge South combined sample for m2 of H. niobrarensis is: length, 1.80 mm; trigonid width, 1.94 mm; talonid width, 1.42 mm (Table 1B).

m3. The lower third molar is present in only two dentaries (Figure 7.2, 7.3). This tooth is small and relatively simple, consisting of an oval column of enamel filled by dentine; its long axis is oriented buccolingually and the crown and root curve to mediad and posterior. Mean size of m3 of H. niobrarensis is: length, 0.61 mm; width, 1.05 mm (Table 1B).

Remarks

The three lower molariform teeth (p4, m1, and m2) of H. niobrarensis show subtle but consistent differences in size and morphology that allow them to be distinguished from each other in most cases. The p4 is larger in average length, trigonid width, and talonid width, and the talonid has a rounded posterior wall. The m1 is on average smaller in all three measured dimensions, and has an anteroposteriorly short talonid with an only slightly rounded posterior wall. The m2 is intermediate in length and trigonid width between p4 and m1, and has a somewhat wider talonid in which the posterior wall forms a nearly straight transverse surface and a squared posterobuccal corner. Lower molariform teeth also differ in the degree of curvature in the anteroposterior plane: p4 is relatively straight, m1 curves slightly posteriorly, and m2 is most strongly curved; however, in practice isolated specimens of m1 and m2 may be difficult to distinguish on the basis of anteroposterior curvature alone.

Although Voorhies (1990a) included Hesperolagomys fluviatilis in a list of mammalian species from the Valentine Railway Quarries, morphological and size differences reported here support the distinction of this Hesperolagomys sample as a new species related to but more derived than H. fluviatilis. Research to be reported elsewhere suggests that H. niobrarensis is likely more widespread in other Nebraska faunas of latest Barstovian age.

Hesperolagomys galbreathi
Clark, Dawson, and Wood 1964

Synonymies. Hesperolagomys cf. H. galbreathi Korth 1998: p. 336; figure 13C; Russellagus cf. R. vonhofi Korth 1998: p. 336-338.

Holotype. MCZ 17890, partial dentary with broken incisor, p3-m2, and alveolus of m3 (Clark et al. 1964) from Fish Lake Valley, Nevada (UCMP Fish Lake Valley 1 locality).

Distribution. Medial Clarendonian (Cl2) of Nevada (Clark et al. 1964), medial to late Clarendonian (Cl2-Cl3) of Nebraska (Korth 1998; this paper), and Clarendonian of northwestern Utah (Tedrow and Robison 1999).

Emended Diagnosis. Differs from other species of Hesperolagomys in possessing lower molariform teeth with relatively wide and buccolingually symmetrical talonids (p4 trigonid width/talonid width≈1.35), and anteroposteriorly short p3 with a prominent cement-filled anterolingual groove. Additionally distinct from H. niobrarensis in smaller size (cheek tooth lengths average 6-25% less).

Hypodigm

Original Hypodigm. (Revised from Clark et al. 1964) Type; MCZ 17891, partial dentary with p3-m2; MCZ 17892, partial dentary with talonid of p4 to m2; UCMP 29626, partial dentary with p4-m1; MCZ 17893-19894, upper incisors; MCZ 17895-17899, isolated upper cheek teeth; MCZ 7651-7652, isolated lower cheek teeth; UCMP 29833, dP4.

Previously Undescribed Topotypes. (All UCMP numbers) 149484, partial maxillary with P3-P4; 149485, upper anterior incisor; 149486, P3; 149487-149488, P4; 149489, M2; 149480, fragmentary dentary with p4-m2; 75269, lower incisor; 149481, p4; 149482, m1; 149483, m2.

Referred Specimens

Nebraska (all UNSM numbers, by locality): Bluejay Quarry: 123346, m1. Poison Ivy Quarry: 123350, P4. Pratt Quarry: 101709, P3; 101776, P3. Tiensvold Ranch: 123311, partial maxillary with P3-M1; 123312, partial dentary with incisor, p3-p4; 123313, partial dentary with p3-m3.

Utah: IMNH 15871, M2; IMNH 15869, ?p4; IMNH 15870, m1; from Salt Lake Formation (Tedrow and Robison 1999).

Description

Skull

Two partial maxillaries with adult cheek teeth (Figure 9.1, 9.2) strongly resemble those described for H. niobrarensis in most characters. Both possess a distinct premolar foramen medial to P3-P4; in the topotypic maxillary (Figure 9.1) it is in line with the posterior wall of P3, whereas in the Nebraska specimen (Figure 9.2) it lies between P3 and P4.

Upper dentition

Anterior Incisor. A topotypic specimen (Figure 9.3; Table 3) lacks the tip but is otherwise very similar in both size and shape to those described by Clark et al. (1964), with a longitudinal groove on the anterior surface displaced slightly medially of center. The tooth is slightly anteroposteriorly longer than those of H. niobrarensis (1.04 mm vs. 1.11 mm).

P2. P2 is unknown, but an alveolus for the tooth is preserved in UNSM 123311 and is consistent with the single (lingual) root and small size as observed for P2 of H. niobrarensis. The dimensions of the alveolus suggest that the tooth was approximately as long as but wider than that of H. niobrarensis (Figure 9.2; Table 3).

P3. Two P3s are known from the type locality, one in a maxilla (Figure 9.1) and one isolated (Figure 9.4); morphology in both is consistent with specimens of H. niobrarensis at similar stages of wear. The P3 in UMPC 149484 (Figure 9.1) is lightly worn, appearing very similar to the H. niobrarensis specimen illustrated in Figure 4.2 in morphology and in stage of wear. The cement of the anterior stria is broad near the occlusal surface and is confluent with a shallower cement-filled buccal stria.

The isolated topotypic P3 is slightly more worn than P3 in the maxilla; occlusal features appear buccolingually stretched, and the precone curves more buccad in occlusal view. Cement in the anterior stria extends further buccad near the occlusal surface in this specimen than in the P3 in the maxilla. P3 in the referred Tiensvold Ranch locality maxilla (Figure 9.2) is nearly identical in size and morphology, and likely represents a similar stage of wear as the isolated topotype.

Two incomplete specimens from Pratt Quarry, UNSM 101709 (identified by Korth 1998, as P2 of Hesperolagomys) and 101776 (identified by Korth 1998, figure 13C, as P3 of Russellagus) are both identified instead as P3s of Hesperolagomys galbreathi based on morphology and size (Table 3) consistent with specimens described above. P2 of H. galbreathi, judging from the alveolus in the maxilla described above, is much smaller than either of the Pratt Quarry specimens; P3 of Russellagus is larger and with a distinct morphology (Bair 2006).

P4. Two P4s of different wear stages were included in the original hypodigm of H. galbreathi; previously unpublished specimens are consistent with most aspects of morphology with specimens of H. niobrarensis at similar stages of wear, and indistinguishable in size from published specimens of H. galbreathi. P4 of H. galbreathi appears to differ somewhat from P4 of H. niobrarensis in the nature of the lingual hypostria. In all known specimens of H. galbreathi, the lingual hypostria nearly contacts the lingual wall of the crescentic valley, terminates buccally in a broad (anteroposteriorly long) squared-off groove, and is oriented nearly buccolingually. In contrast, the lingual hypostria in P4 of H. niobrarensis is separated from the crescentic valley by a wider bridge of dentine, terminates in a slightly rounded point buccally, and is oriented more posteriorly (compare Figures 9.1, 9.2, 9.5, and 9.6 with 5.2-5.5). As evidenced by specimens at various stages of wear, this difference remains present throughout the stages of wear known for both taxa.

The least worn unpublished specimen, with crown base incompletely formed (Figure 9.5), is morphologically nearly identical to MCZ 17896 as described and figured by Clark et al. (1964, figure 3C, D). The enamel border to the buccal valley is more prominent than that seen in P4 in the maxilla (Figure 9.1), such that cement of the buccal valley is separate from cement in the anterobuccal portion of the crescentic valley. A composite 'complete' P4 reconstructed from two topotypes (Figure 9.7), although not fully representing the unworn crown, gives an estimated crown height of 9.61 mm and an estimated minimum hypsodonty index of 6.49.

UNSM 123350, from Poison Ivy Quarry, Nebraska, is a P4 referred to H. galbreathi. This specimen is nearly unworn, with an occlusal pattern very similar to that in Figure 4.2 (H. niobrarensis); it is somewhat larger in size than other known P4s of H. galbreathi (Table 3).

M1. In size and shape, M1 in UNSM 123311 (Figure 9.2, Table 3) closely resembles MCZ 17898, described as a probable M1 by Clark et al. (1964, figure 3A, B), as well as M1s of H. niobrarensis, confirming its identity as such. M1 length in the referred specimen is 1.21 mm, less than for previously known values of P4 (Table 3). Clark et al. (1964, table 2) report the length of MCZ 17898 as 1.3 mm, overlapping with length of P4 in UNSM 123311 but less than the value reported for P4s in the original hypodigm (length of 1.5 mm for MCZ 17895 and 17896; Clark et al. 1964, table 2).

As for M1s of H. niobrarensis, M1s of H. galbreathi can consistently be distinguished from P4s by: (1) greater occlusal extent of the lingual hypostria (covering at least one third of the occlusal surface and nearly meeting the crescentic valley); (2) anteroloph and posteroloph subequal in width; (3) buccal edge of crown angled slightly posteriorly (posterobuccal cusp shifted lingually in comparison with anterobuccal cusp); and (4) slight curvature of the tooth in the anteroposterior plane, concave to posterior.

M2. UCMP 149489 (Figure 9.8) is consistent in morphology with M2s described for H. niobrarensis and is here identified as an M2. Although the crown is damaged on the posterobuccal edge, the anterobuccal root is preserved and allows estimation of occlusal width. M2 length is 1.24 mm, less than that of P4 and close to that of M1 (Table 3).

A more complete M2 of H. galbreathi is represented by IMHN 15971, an isolated tooth from the Clarendonian of northwestern Utah, tentatively identified as an M1 or M2 of Hesperolagomys cf. H. galbreathi by Tedrow and Robison (1999). This specimen is very similar in shape and size to UCMP 149489, possessing a "deep and persistent internal hypostria which nearly reaches the crescentic valley" (Tedrow and Robison 1999, figure 3d-e, pg. 483) and a reduced posterobuccal cusp; length is 1.31 mm (Tedrow and Robison 1999, table 2).

Deciduous Upper Premolar. A single upper deciduous P4 is known, UCMP 29633 (see description in Clark et al. 1964; Wood 1936; and Hall 1930, figure 1). In occlusal view dP4 shares features with permanent upper dentition of Hesperolagomys, including a prominent 'J'-shaped crescentic valley, lingual hypostria extending buccolingually across approximately one fourth of the occlusal surface and nearly meeting the crescentic valley, deep buccal valley, and poorly developed buccal cusps. However, it is very small and has a much shorter crown compared with permanent upper teeth.

Dentary

The holotypic jaw was fully described and illustrated by Clark et al. (1964). Three previously undescribed specimens, one from the type locality and two from the Tiensvold Ranch locality, closely resemble the holotype in size and require only brief additional comments regarding the disposition and relative size of major mental foramina and the posterior extent of the incisor.

The preserved portion of a topotypic jaw fragment (UCMP 149480: Figure 10.1) contains a small circular depression with a single mental foramen ventral to the trigonid of m1 and close to the ventral margin of the jaw, slightly more anteriorly placed than the depression with two small foramina in the type specimen (below m1-m2) but interpreted as a corresponding structure. The incisor is missing but the capsule indicates that it extended posteriorly to beneath m1, as in the type specimen. An additional dentary fragment (UCMP 149481: Figure 10.2) possess a larger foramen in line with the center of p4, also slightly more anteriad than the corresponding large foramen in the type specimen, which lies beneath p3.

In both specimens from the Tiensvold Ranch locality (Figure 11.1, 11.2), the roots of p3-m1 are displaced laterally from the incisor, and the incisor capsule makes a prominent bulge on the medial side of the jaw, as in the type specimen. The more complete specimen (UNSM 122313: Figure 11.1) has all three major mental foramina preserved: the largest is the middle foramen located beneath the talonid of p3, a smaller foramen is in line with the talonid of m1 and closer to the ventral border of the dentary, and the smallest is anterior to p3 and more dorsally positioned than the posterior foramina. The medial wall of the incisor capsule is broken to reveal the cavity, which extends posteriorly to beneath the talonid of m1. Except for the slightly more anterior placement and single opening of the most posterior mental foramen, this specimen matches the holotype. UNSM 122312 (Figure 11.2) is less complete and preserves only the anterior-most foramen located anterior to p3; the incisor is exposed and extends posteriorly to beneath the position of the talonid of m1.

Lower Dentition

Lower Incisor. A lower left incisor lacks both ends, but is similar in cross-sectional morphology and dimensions to type material (Table 3; Clark et al. 1964, table 2) and to lower incisors described for H. niobrarensis. The incisor in UNSM 122312 (Figure 11.2, 11.3) is better preserved and is nearly identical to UCMP 75269 in size and cross-sectional shape.

p3. The p3s in the Tiensvold Ranch jaws (Figure 11.1, 11.2) resemble that in the type specimen (Clark et al. 1964, figure 4B) in overall shape and size (Table 3), and in possessing a deep, cement-filled buccal fold separating the trigonid and talonid, and a shallower, wide cement-filled anterolingual 'groove'. Both teeth taper markedly from the crown apex to base. The less worn specimen (Figure 11.2) additionally has a small cement-filled enamel lake at the connection between trigonid and talonid, a feature unique to this specimen, in which the trigonid is also anteroposteriorly shorter and more buccally directed than in the type specimen. The p3 in UNSM 123313 lacks the buccal portion of the crown at the occlusal surface and is slightly shorter-crowned (older) than 123312, and has neither the enamel islet nor the crenulation of the anterior wall of the talonid present in the latter. The trigonid is slightly more anteriorly directed than in 123312, resembling the type specimen in this respect. The p3 of H. galbreathi does not expand anteriorly towards crown base as much as in p3 of H. niobrarensis, so maximum length is less in H. galbreathi.

p4-m2. Teeth in the topotypic dentary fragment, in referred dentaries, and isolated lower molariform teeth are like those of the type specimen and in specimens described for H. niobrarensis in having trigonids transversely wider and anteroposteriorly shorter than talonids, and in the buccolingually narrow anterior projections of talonids (Clark et al. 1964). The lower molariform teeth can be distinguished on the basis of relative size and anteroposterior curvature, as noted for H. niobrarensis. In contrast to lower molariform teeth of H. niobrarensis, those of H. galbreathi have relatively wide and buccolingually more symmetrical talonids. Some wear-related variation in the shape of the anterior projection of the talonid is apparent between the type specimen, topotypes, and referred material.

Lower cheek teeth of the type specimen have buccolingually narrow and nearly symmetrical anterior projections of the talonid, with the main body of the talonid connected to the anterior projection lingually and buccally by concave enamel surfaces (Clark et al. 1964, figure 4). Clark et al. (1964) also referred a partial jaw to Hesperolagomys (UCMP 29626; originally tentatively identified by Hall 1930, figure 29, as Sylvilagus? sp.) and identified it as a young individual as evidenced by completely isolated posterolophids on p4 and m1. This specimen has anterior projections of the talonids of p4 and m1 that are more buccolingually asymmetrical than in the more mature type specimen and have straight anterolingual talonid walls. In the topotypic dentary (UCMP 149480: Figure 10.1), interpreted here as intermediate between the above two specimens in ontogenetic age, the m1 talonid is buccolingually narrow and nearly symmetrical with concave enamel walls on both lingual and buccal sides of the anterior projection. In contrast, the buccal enamel wall connecting the projection to the main body of the talonid in m2 is straight, while in p4 the anterobuccal wall is convex and the anterolingual wall is concave; in both teeth the talonid is asymmetrical, with an expanded anterobuccal corner in comparison with the anterolingual corner. Lower cheek teeth of UCMP 149481 (Figure 10.2) and UNSM 123313 (Figure 11.1), and isolated lower cheek teeth UCMP 149482 (Figure 10.3) and 149483 (Figure 10.4) have wider, asymmetrical anterior projections of the talonid and more closely resemble p4 and m2 of UCMP 149480 than the narrower, symmetrical anterior projection of m1 of that specimen or the lower molariform teeth of the type specimen.

The p4 in UNSM 123312 (Figure 11.2) from the Tiensvold Ranch locality is in an early stage of wear as evidenced by a posterolophid in the process of merging with the main body of the talonid and has a distinctly different talonid morphology. The lingual side of the posterolophid is still separated by cement, but the buccal side is marked only by a minor enamel groove. Rather than a coherent single projection, the anterior portion of the talonid is composed of two rounded anteriorly-directed points separated by a small central groove; the anterobuccal enamel wall is sharply convex and attaches near the center of the talonid, whereas the anterolingual wall is nearly straight and aligned with the main body of the talonid. UNSM 123346 from Bluejay Quarry (Nebraska) is an m1 referred to H. galbreathi; this specimen is slightly smaller in measured dimensions than other m1s (Table 3), but has postmortem enamel loss, which may account for the discrepancy since it closely resembles topotypic specimens in occlusal morphology (Figure 10.1 and 10.4).

m3. The m3 is present only in UNSM 123313 (Figure 11.1; Table 3) and does not differ notably from that of H. niobrarensis in size or morphology.

REMARKS

Specimens referred to Hesperolagomys cf. H. galbreathi from the Black Butte Fauna, Oregon (Shotwell, 1970), do not belong to this taxon; they are referable to Russellagus vonhofi and will be discussed in a future publication on Russellagus. Other referred specimens of Clarendonian age from Utah and Nebraska are consistent with H. galbreathi.

Hesperolagomys fluviatilis Storer, 1970

Synonymies. Russellagus vonhofi Storer 1970, 1975 (in part): p. 116-121; figure 83M, N; 85B; Hesperolagomys fluviatilis Voorhies 1990b (in part): p. A79, 323-324, 400, 487; Russellagus vonhofi Voorhies 1990b (in part) p. A79-80, 324, 487.

Holotype. ROM 7320, left m1, from Kleinfelder Farm locality, Wood Mountain Formation, south-central Saskatchewan, early late Barstovian (Ba2a) (Storer 1970, 1975).

Emended Diagnosis. Differs from other species of Hesperolagomys in: (1) p3 relatively anteroposteriorly long and lacking a cement-filled anterolingual groove; (2) lower molariform teeth with relatively narrow talonids (p4 trigonid width/talonid width≈1.50); (3) smaller than both H. niobrarensis (cheek tooth length 10-20% less) and H. galbreathi (cheek tooth length 1-10% less) in most measured dimensions. Also differs from H. galbreathi in possessing more asymmetric talonids on lower molariform teeth.

Geologic age and distribution. Early late Barstovian (Ba2a) of Saskatchewan (Storer 1970, 1975); medial late Barstovian (Ba2b) of Nebraska (Voorhies 1990a, 1990b; this paper).

Hypodigm

Original Hypodigm (revised from Storer 1975). Type; (all ROM numbers) 7385, 7386, P3s; 7342, 7399, 7412, P4s; 7345, 7352, M1; 7343, 7346, 7348, 7350-7351, 7353, 7354, M2s; 7302, 7304, 7305, 7306, p3s; 7314, 7319, 7321, 7322, 7324, 7338, 7416, p4s; 7313, 7315, 7318, 7320, 7321, 7326, 7330, 7331, 7336, 7337, 7418, m1s; 7310, 7311, 7312, 7315, 7317, 7325, 7327, 7333, 7335, m2s; 7323, 7328, lower molariform teeth; 7303, 7329, trigonids of lower molariform teeth; 7340, dP4.

Specimens Formerly Included in Russellagus vonhofi Hypodigm. ROM 7385, 7386, P3s; 7399, P4; 7349, M2;

Previously Undescribed Topotypes. (ROM numbers): 52635, P4; 52637, M1; 52636, M2; 52642, partial dentary with p4-m2 and alveolus for m3; 52638, p4; 52641, m1; 52639, m2. (UNSM numbers): 123357, m1; 123355, m2.

Referred Specimens from Nebraska (all UNSM numbers, by locality).

Achilles Quarry: 83031, P4; 83029, m1; 83032, m2.

Egelhoff Quarry: 90001, M1; 90000, m2.

Immense Journey: 123328, P4; 123331 and 123332, M2.

Lost Duckling Quarry: 83141, m2.

Miller Creek: 123314, fragmentary maxillary with P4 and M1.

Norden Bridge Quarry: 83738, P3.

Description

Skull

Cranial material for this taxon is fragmentary and consistent with characters described for other taxa.

Upper Dentition

Several upper cheek teeth in the original descriptions of H. fluviatilis and Russellagus vonhofi of Storer (1975) are here reinterpreted in light of new information on upper cheek tooth morphology of H. niobrarensis and H. galbreathi, as described above. Upper cheek teeth of H. fluviatilis closely resemble those of H. niobrarensis in morphology but are slightly smaller (Table 4).

P3. Two teeth originally described as P2s of Russellagus (ROM 7385 and 7386) by Storer (1975) are here identified as P3s of H. fluviatilis on the basis of a nearly perfect match in occlusal morphology with P3 of H. galbreathi and H. niobrarensis (compare P3s in Figure 4, Figure 9.1, 9.2, 9.4, and Figure 12.1 in this paper with figure 83F of Storer 1975). ROM 7386 (Figure 12.1) represents a moderately young individual, while ROM 7385 (Storer 1975, figure 83F) is an older individual with a crown approximately half as tall, and with crown elements buccolingually 'stretched' compared to the younger tooth. Mean P3 length for H. fluviatilis is 1.14 mm; mean width is 2.78 mm (Table 4A).

UNSM 83738 from Norden Bridge Quarry, initially referred to R. vonhofi by Voorhies (1990b) on the basis of its close resemblance to specimens considered to be Russellagus P2s by Storer (1975), is here identified as a P3 of Hesperolagomys; this specimen was not available for measurement at the time of this study but is referred to H. fluviatilis as the size of dimensions of other Hesperolagomys specimens support the presence of this taxon at Norden Bridge Quarry (see below).

P4. The numerous specimens now available for H. galbreathi and H. niobrarensis allow upper cheek teeth of H. fluviatilis to be unambiguously distinguished. P4 of H. fluviatilis is more similar in morphology to that of H. niobrarensis in the rounded point of the buccal edge of the lingual hypostria, contrasting with the squared-off edge of the feature in H. galbreathi (Figure 12.2 through 12.5). Changes in occlusal morphology through wear differ little from that documented for the tooth in the other species. Two teeth identified as probable P4s and figured as paratypes of H. fluviatilis by Storer (1975, figures 81L-M and 82C-E) are here identified as M1s as discussed below. ROM 7351 and 7354, also listed as P4s by Storer (1975) are identified here as M2s. Three upper cheek teeth (ROM 7399, 7342, 7412), determined here to be P4s of H. fluviatilis, were assigned to Russellagus vonhofi by Storer (1975). These, along with four newly described specimens, demonstrate occlusal changes associated with wear in this tooth.

A moderate wear stage is shown by ROM 7399 with nearly complete crown base (Storer 1975, figures. 83M-N and 85B; Figure 12.2, this paper), in which the cement-filled posterobuccal and crescentic valleys are reduced in size, and the lingual hypostria has closed basally. A slightly older individual is represented by ROM 7412 (Figure 12.3), in which crown formation is complete, and the crown is approximately half as tall as in ROM 7399; the occlusal pattern appears stretched buccolingually and cement has nearly been obliterated from the posterobuccal valley. ROM 52635 (Figure 12.4) is slightly more worn but has a similar occlusal pattern. The oldest available P4 is ROM 7342 (Figure 12.5) in which buccolingual 'stretching' has proceeded still further, and the crescentic valley has become separated into two portions still containing cement: an anterobuccal portion approximately the same size as the buccal valley, and a lingual 'v'-shaped portion. Mean values for dimensions of P4 of H. fluviatilis from Kleinfelder Farm: length, 1.44 mm; width, 3.31 mm (Table 4A).

Three P4s from additional Ba2b Nebraska localities are referred to H. fluviatilis on the basis of their very similar morphology and occlusal dimensions. UNSM 83031 from Achilles Quarry appears to represent a very early stage of wear of P4 (comparable with Figure 12.2), with only a narrow sliver of dentine exposed in the central crescent, prominent cement in the buccal valleys, and an incompletely formed crown base. UNSM 123314 from Miller Creek and UNSM 123328 from Immense Journey Quarry are slightly more worn. Occlusal dimensions of these referred specimens (Table 5) are within the range for P4s in the hypodigm of H. fluviatilis.

M1. Three isolated teeth closely resemble the morphology of M1 described for H. niobrarensis and H. galbreathi and are identified as M1s of H. fluviatilis. One of these, ROM 7352, was identified tentatively by Storer (1975, figures 81M and 82E) as P4 of H. fluviatilis. In this specimen, the lingual hypostria covers nearly half of the occlusal surface, much greater than that observed for P4, and only a thin portion of dentine separates the crescentic valley from the lingual hypostria. ROM 52637 (Figure 12.6) is buccolingually wider with a shorter crown and represents a slightly older wear stage. Mean values for dimensions of M1 of H. fluviatilis from Kleinfelder Farm are: length, 1.28 mm; width, 2.23 mm (Table 4A).

An M1 (in UNSM 123314) from Miller Creek, Nebraska, supports referral of this specimen (a maxilla fragment) to H. fluviatilis. The M1 is lacking the buccal cusps and the buccal portion of the crescentic valley due to minor damage at the occlusal surface, but is similar to other M1 specimens in size (compare Table 4 and Table 5) and lingual occlusal features, closely resembling the specimen in Figure 12.6 in these features and in stage of wear.

An upper cheek tooth from Egelhoff Quarry (Nebraska), UNSM 90001, is tentatively referred to H. fluviatilis as an M1. This specimen shows a strong resemblance to other M1s (such as Figure 12.6) in occlusal morphology but it is anomalously small in occlusal dimensions (Table 5); some damage to the enamel at the anterolingual corner of the tooth is evident, and may account for the size difference.

M2. M2s of H. fluviatilis fall within the range of morphological variation expressed in the Valentine Railway Quarries H. niobrarensis sample (Figure 6.6-6.8); they differ significantly only in size, having a shorter length (Table 4). ROM 7346 (Figure 12.7) was originally tentatively identified by Storer (1975, figure 81P) as a M1 or M2, and is more confidently identified here as a little-worn M2. In this specimen, the posterior limb of the crescentic valley terminates in a cingulum, offset topographically below the rest of the crown by approximately 0.5 mm. The lingual hypostria is anteroposteriorly long and extends across between one half and one third the buccolingual extent of the occlusal surface. ROM 52636 (Figure 12.8) is slightly more worn than 7346, still retaining the posterior cingulum but the lingual hypostria extends over approximately one half the width of the occlusal surface and is anteroposteriorly narrower. ROM 7348 (Storer 1975, figure 81K, M) is very similar to 52636 but has less enamel exposed in the posterior part of the crescentic valley. ROM 7353 (Figure 12.9) and ROM 7343 (not figured) are next oldest in the wear series, with the cingulum no longer present (the posterior limb of the crescentic valley now is narrowly separated from the posterior enamel wall of the tooth and topographically at the same level). The lingual hypostria dominates the occlusal surface, extending more than one half the width of the tooth. Mean values of dimensions for M2 of H. fluviatilis from Kleinfelder Farm are: length, 1.21 mm; width, 2.20 mm (Table 4A).

Two M2s from Immense Journey Quarry (Nebraska), UNSM 123331 and 123332, are referred to H. fluviatilis. UNSM 123331 is more worn than other H. fluviatilis specimens, most closely resembling a worn H. niobrarensis specimen (Figure 6.8) in occlusal morphology. UNSM 123332 is lacking the buccal-most portion, but is very similar to the H. fluviatilis specimen in Figure 12.9.

Deciduous Upper Premolar. I agree with Storer's interpretation of ROM 7340 as a dP4 of H. fluviatilis (1975, figure 81H, I). The tooth is much smaller than P4 (Table 3) and has a very short crown with the lingual edge broadening dorsolingually. The occlusal pattern is nearly identical to that of UCMP 29633 (dP4 of H. galbreathi, Hall, 1930, figure 1). Except for its short crown and small size it could be confused with a P4 in very early wear stages (e.g., UNSM 83031, Figure 12.2).

Dentary

A previously undescribed lower jaw of H. fluviatilis from the type locality (ROM 52642: Figure 13) has similar proportions to other known dentaries but is quite small as compared to those of H. niobrarensis and slightly smaller than H. galbreathi (Table 2). Two large mental foramina nearly equal in size are present, one beneath p3 and the other beneath m1. A smaller foramen anterior to p3 is poorly preserved but is positioned similarly to that in specimens of H. galbreathi. The incisor capsule extends posteriorly to beneath m1. The ventral border of the horizontal ramus is fairly straight between the level of the ascending ramus and p4.

Lower Dentition

p3. Three p3s from Kleinfelder Farm (Figure 14.1-14.3) are very similar in shape to those of H. galbreathi but lack the characteristic anterolingual groove. All specimens are relatively low crowned and with thick enamel at occlusal surface; both features are interpreted as representing advanced wear stages as noted in p3 for the other Hesperolagomys species. Mean size of p3 of H. fluviatilis is: length, 1.48 mm; width, 1.37 mm (Table 4B).

Lower Molariform Teeth. As in other Hesperolagomys species, lower molariform teeth can be differentiated by the length of the talonid (shorter in m1 than in p4 and m2), the nature of the posterior talonid wall (rounded in p4 and m1, transversely oriented and straight in m2), and in curvature in the anteroposterior plane (p4 is straight, m1 moderately curved, and m2 more strongly curved). p4-m2 in the dentary (ROM 52642: Figure 13) have short crowns and lack posterolophids, indicating they are at least somewhat worn. Each has a similarly-shaped talonid, with a wide, buccolingually asymmetrical anterior projection characterized by a concave buccal wall and lingual wall varying from slightly concave (p4 to m1), to straight or slightly convex (m2).

No p4s in early stages of wear (with posterolophids) are known. Most p4s of H. fluviatilis have talonids with a concave lingual wall and a slightly concave to straight buccal wall (e.g., Figure 14.5-14.8); the specimen with the tallest crown (ROM 52638: Figure 14.4) has a more concave lingual wall than other specimens. No clear trends in changing morphology of the lingual wall of the anterior projection are apparent within the wear stages represented. Mean dimensions for p4 of H. fluviatilis is: length, 1.62 mm; trigonid width, 1.84 mm; talonid width, 1.24 mm (Table 4B). Talonids are relatively narrow, such that the ratio of trigonid width to talonid width is ≈1.50.

A tall-crowned, little-worn m1 (Figure 14.9) has a small posterolophid separate from the talonid. The anterior projection is divided into two small anteriorly facing projections, similar to but smaller than in H. galbreathi at a similar stage of wear (Figure 10.3); the lingual wall of the anterior projection is straight. A slightly more worn specimen (Figure 14.10) lacks a posterolophid but also has a relatively tall crown and a straight lingual wall. Other specimens, exhibiting more advanced stages of wear, (e.g., Figure 14.11-14.13) have a slightly concave lingual wall. Mean size of m1 of H. fluviatilis from Kleinfelder Farm is: length, 1.66 mm; trigonid width, 1.84 mm; talonid width, 1.26 mm (Table 4B).

UNSM 83029 from Achilles Quarry, Nebraska, is tentatively referred to H. fluviatilis as a p4 or m1; it is very strongly worn, with little crown remaining and is slightly damaged on the posterolingual edge of the talonid. It is similar in occlusal morphology to known specimens and has strong roots as observed in other lower (permanent) molariform teeth, but is anomalously small (Table 5).

No specimens of m2 with a posterolophid are known. A relatively unworn specimen (Figure 14.14) has crown base still forming and a straight lingual margin, as in tall-crowned m1s (e.g., Figure 14.15-14.16). Teeth in more advanced stages of wear (e.g., Figure 14.17), have a similarly shaped, slightly concave lingual margin. The most worn m2 (Figure 14.18) also has a concave lingual margin, but the main body of the talonid extends farther anteriorly on the lingual side than observed in other specimens, such that the talonid length on the lingual side is nearly two times that on the buccal side, and the talonid appears more asymmetrical than in other specimens. Mean size of m2 of H. fluviatilis from Kleinfelder Farm is: length, 1.56 mm; trigonid width, 1.66 mm; talonid width, 1.19 mm (Table 4B).

Three specimens from Nebraska localities (Table 5) are referred to H. fluviatilis as m2s: UNSM 83032 from Achilles Quarry, UNSM 90000 from Egelhoff Quarry, and UNSM 83141 from Lost Duckling Quarry. UNSM 83032 and UNSM 90000 are very tall-crowned, representing an early stage of wear, closely resembling that figured in Figure 14.15. UNSM 83141 is highly worn with little crown remaining; its occlusal morphology is similar to that of Figure 14.18.

Remarks

Storer (1975:111) recognized that the "shape and strength" of the anterior projection of the talonid of p4-m2 was an important characteristic differentiating H. galbreathi and H. fluviatilis, with that of H. fluviatilis being "always stronger" than that of H. galbreathi. Voorhies (1990b) also noted a consistent morphologic difference between Hesperolagomys lower molariform teeth of Clarendonian age (H. galbreathi) and those of Barstovian age (referred by Voorhies 1990b, to H. fluviatilis). He emphasized the buccolingual symmetry of the talonid and the narrow protrusion "bounded both lingually and buccally by a concave enamel wall" in H. galbreathi, contrasting with the wider anterior projection and asymmetry of the talonid of H. fluviatilis created by the straight or convex anterolingual wall, and suggested that "...the compressed, pointed anterior talonid extensions in H. galbreathi are actually more, not less, prominent than those of H. fluviatilis" (Voorhies 1990b:A323)

Based on observations of the more numerous specimens now available for comparison, I find that the talonid of lower molariform teeth of H. galbreathi is distinct from that of other species in two respects: (1) the anterior projection is narrower and more clearly separated from the main body of the talonid, as both anterolingual and anterobuccal walls of the projection attach medial to the outer walls of the main body; and (2) the main body of the talonid is buccolingually more symmetrical. In H. fluviatilis and H. niobrarensis, it is more difficult to demarcate the anterior projection from the main body of the talonid because the anterolingual wall attaches more laterally, confluent with the lingual wall of the talonid; additionally, the main body of the talonid is asymmetrical, extending more anteriad on the lingual side than on the buccal side. Specimens described here illustrate the degree to which variation in the shape of the talonid can be attributed to stage of wear. In all three species, the anterolingual wall of the talonid is straight in very early wear, becoming much more concave in later wear. In earlier wear stages of p4 of H. galbreathi, the anterior projection appears more asymmetrical than in later wear, as the lingual side is straight and does not mirror the buccal side, but the talonid is never as anterolingually expanded as it is in H. fluviatilis.

Hesperolagomys sp.

Synonymies. Oreolagus colteri Barnosky 1986 (in part); Hesperolagomys sp. (Kraatz and Barnosky 2004).

Referred specimens from Wyoming. (all UWBM numbers) 62702, P3; 62704 and 62720, M1s; 62698, M2.

Geologic Age and Distribution. Medial late and/or late late Barstovian (Ba2b-c) of Wyoming.

Description. UWBM 62698 (Barnosky 1986:plate IIIB), the type specimen of O. colteri, is strikingly similar to M2s of Hesperolagomys and is here reinterpreted as such. The Cunningham Hill specimen is nearly identical in occlusal morphology to ROM 7343 (Storer 1975, figure 81O) and to ROM 7353 (Figure 11.9, this paper). Length reported for this specimen (1.2 mm) is less than for specimens included in H. niobrarensis but overlaps H. fluviatilis in size (see Table 1 and Table 4).

Two specimens originally placed in the hypodigm of Oreolagus colteri by Barnosky (1986:plate IIIA, D, and G) as P4s or M1s are here interpreted as M1s of Hesperolagomys. Both are incomplete buccally, but clearly have wide lingual hypostriae nearly meeting the crescentic valley, a characteristic of Hesperolagomys. Occlusal lengths of these specimens (1.5 and 1.4 mm) reported by Barnosky (1986:table 2) fall within the range for H. niobrarensis but are larger than those known for specimens of H. fluviatilis from Saskatchewan and Nebraska

UWBM 62702 from the Wyoming Colter Formation is here identified as a partial P3 of Hesperolagomys. Although the tooth lacks the buccal margin at the crown apex, an accurate measure of length was possible (1.13 mm), which places it within the size range of both Barstovian species, H. fluviatilis and H. niobrarensis.

Remarks. Confusion over the identity of ochotonid remains from Cunningham Hill highlights the similarity of occlusal features of some upper molariform teeth of Hesperolagomys and Oreolagus. Hesperolagomys is rare among lagomorphs in possessing both a deep lingual hypostria and persistent crescentic valley in (rooted) upper molariform teeth. P4-M1 of Oreolagus also always possess deep lingual hypostriae, but only in some species is a crescentic valley variably present (Dawson 1965), or early wear stages known only for some species. Incomplete preservation of buccal roots in the Cunningham Hill specimens apparently let to their misidentification as Oreolagus. As shown above, however, upper molariform teeth of Hesperolagomys can also be distinguished from those of Oreolagus by the presence of persistent buccal cusps in the former and lack of enamel-covered buccal cusps in the latter. Since the few available specimens from Cunningham Hill overlap with both H. fluviatilis and H. niobrarensis in size, no definite referral to species level is attempted here.