Isolated cheek teeth from the late Puercan (early Paleocene) Split Lip Flats local fauna, from the head of Willow Wash, Nacimiento Formation, San Juan Basin, represent a new genus and species of hyopsodontid “condylarth”, Chacomylus sladei. The teeth are small, bunodont, and are remarkable for the enlargement of the protocone, pronounced exodaenodonty, and high degree of attritional wear caused by transverse shearing or grinding. Although C. sladei bears a superficial resemblance to the apheliscine apheliscid Gingerichia, a preliminary phylogenetic analysis confirms that it is a hyopsodontid, closely related to the Puercan hyopsodontids Valenia wilsoni and Litomylus orthronepius.
Introduction
The Nacimiento Formation of the San Juan Basin, northwestern New Mexico, contains the most diverse early Paleocene mammal assemblages in the world (Williamson 1996). Efforts to collect fossil mammals from these deposits have a long history, extending over 125 years. Rich faunas from near the base of the Nacimiento Formation form the basis for the middle and late interval zones of the Puercan (Pu2–3) North American Land Mammal Age (NALMA). A new late Puercan locale was recently discovered near the head of Willow Wash (Fig. 1), north of the classic late Puercan locales of Barrel Springs and De-na-zin Wash and southeast of late Puercan locality of Gallegos Canyon (Lucas 1984). The local fauna from this locale was termed the Split Lip Flats Local Fauna (Williamson and Weil 2002; Williamson et al. 2011). Underwater screenwashing of this locale, NMMNH localities L-4723 and 4725 (precise locality information on file at the NMMNH), resulted in the recovery of many isolated teeth representing multituberculate, eutherian, and metatherian mammals. Here we report on a new genus and species of hyopsodontid “condylarth” (Archibald 1998) from the Split Lip Flats Local Fauna. The new taxon is remarkable for its small size relative to other hyopsodontids and for its unique combination of dental characters.
Institutional abbreviations.—AMNH, American Museum of Natural History, New York, USA; KU, University of Kansas, Lawrence, USA; NMMNH, New Mexico Museum of Natural History and Science, Albuquerque, USA; UALP, University of Alberta Laboratory of Palaeontology, Edmonton, Canada; UCMP, University of California, Museum of Paleontology, Berkeley, USA.
Other abbreviations.—cc, centrocrista; co, cristid obliqua; DW, distal width; ecg, ectocingulum; encd, entocristid; end, entoconid; etx, ectoflexus; hy, hypocone; hyd, hypoconid; hyld, hypoconulid; hyxd, hypoflexid; L, length; me, metacone; mec, postmetacrista; med, metaconid; msl, metastylar lobe; mtl, metaconule; MW, mesial width; pa, paracone; pac, preparacrista; pacd, paracristid; pad, paraconid; pmlc, premetaconule crista; prc, precingulum; prl, paraconule; pprc, preprotocrista; ps, parastyle; pscg, postcingulum; psl, parastylar lobe; psplc, postparaconule crista; psprc, postprotocrista; pplc, preparaconule crista; pr, protocone; prcd, protocristid; prd, protoconid; ps, parastyle; pscg, postcingulid; psl, parastylar lobe; psmlc, postmetaconular crista; st, stylocone; sts, stylar shelf; tb, talon basin; tdb, trigonid basin; tlb, talonid basin; W, Width.
Material and methods
Tooth nomenclature (Fig. 2) follows Szalay (1969). All measurements are in mm and made to the nearest 0.05 mm using a Wild™ measuring reticule and a Leica™ MZ 6 microscope.
Systematic paleontology
Order “Condylarthra” Cope, 1881
Family Hyopsodontidae Trouessart, 1879
Genus Chacomylus nov.
Etymology: Chaco, for Chaco Canyon of the San Juan Basin, New Mexico, and Latin mylus, millstone or grinder.
Type species: Chacomylus sladei sp. nov., only known species. Nacimiento Formation, Puercan, Paleocene, Willow Wash, San Juan Basin, New Mexico.
Diagnosis.—as for the type and only species.
Stratigraphic and geographic range.—As for the type and only known species.
Chacomylus sladei sp. nov.
Figs. 3–5, Table 1.
Etymology: In recognition of Warren Slade, long-time NMMNH volunteer and his contributions to New Mexico paleontology.
Holotype: NMMNH P-41208, right M1.
Type locality: NMMNH locality L-4723, head of Willow Wash, San Juan Basin, northwestern New Mexico.
Type horizon: Fossil horizon B (Williamson 1996), Nacimiento Formation (late Puercan, Pu3).
Material—NMMNH P-34804, right P4; 38461, left P4; 44353, left M3; 51537, right partial M2; left p3?; 44345, right m2; 55397, left m3.
Diagnosis.—Similar to hyopsodontid and apheliscid “condylarths” (sensu Zack et al. 2005) in its small size and bunodont cheek teeth with a lower molar talonid that is at least about two-thirds the height of the trigonid. Similar in size to Tiuclaenus minutus, but smaller than other hyopsodontids (Table 1). Differs from all hyopsodontids (the upper dentition is unknown for Tiznatzinia, Bomburia unless Platymastus palantir is a synonym, see Williamson and Carr [2007], and Oxytomodon) by its lack of a P4 parastyle, relatively more transverse upper molars, and larger protocones. Differs in lower molar morphology from all hyopsodontids, except Litomylus orthronepius and Oxytomodon, in possessing lower molar exodaenodonty. Differs from Litomylus orthronepius in having narrower molar stylar shelves and relatively wider molars. Differs from Oxytomodon in having a more transverse orientation of the distal wall of the trigonid and in having an entocristid that closes the talonid basin lingually. Differs from all hyopsodontids in having development of wear facets that indicates greater enhancement of transverse shearing during mastication. Differs from apheliscids in having the upper molar postcingulum higher above the base of the crown than the precingulum, a relatively less developed hypocone, molar paraconid is lingually positioned, basally fused with, and closely appressed to, the metaconid, and the hypoconulid is lingual to the midline of the tooth and basally fused to the entoconid.
Description.—Chacomylus sladei is known only from isolated teeth. The isolated teeth described in this report are referred to Chacomylus sladei based on size, morphology, patterns of attritional wear, and occlusal fit. The upper dentition is represented by two P4's, an M1, a probable M2, and an M3. The lower dentition is represented by an isolated premolar, probably a p3, an m2, and an m3.
P4.—The P4 (Fig. 3A) is wider than long. The paracone is the tallest cusp and is elongated mesiodistally. A smaller swelling on the crest descending distally from the paracone likely represents the metacone. A low ectocingulum bounds the tooth buccally and widens distally. Mesially, the ectocingulum is narrow. The P4 lacks a parastyle. The protocone is distinct and separated from the paracone by a cleft. A low preprotocrista extends mesiobuccally to the mesial base of the paracone. On the distal face of the protocone, wear has breached the enamel, forming an elongate exposure of dentine ringed by enamel that extends distobuccally. The remnant of a narrow postcingulum is present lingual to this facet on the distal face of the protocone. This obliterates any evidence of a metaconule, if one was originally present. A paraconule is not developed.
M1.—The M1 (Fig. 3B) is wider than long. The paracone and metacone are subequal in size, closely spaced, and partially fused at their base. The stylar shelf is undeveloped, with only the low ectocingulum bordering the buccal margin of the crown. The parastylar lobe extends mesially and the parastyle has been beveled by wear. The paraconule and metaconule are subequal in size and positioned close to the paracone and metacone. The protocone is large and subequal in height to the para- and metacone. A hypocone is present as a distinct cusp distal to the apex of the protocone. The postmetaconule crista extends to the metastyle. The postcingulum terminates lingual to the metaconule.
Planar wear facets have beveled the mesial and distal sides of the protocone. Similar facets have also truncated the distal face of the paracone and the mesial face of the metacone, forming a V-shaped notch into the centrocrista; the original shape and orientation of the centrocrista have been obliterated. A precingulum may have been present, but this can no longer be determined because of wear to this region of the tooth.
M2.—A probable M2 (Fig. 3C), missing the metacone and distobuccal corner of the tooth, is represented by NMMNH P-51537. It is similar to the M1, but differs in that the preand postcingulum are not completely obliterated by wear. The precingulum originates lingually at a position buccal to the apex of the protocone. It is significantly lower on the protocone than the postcingulum. The postcingulum is expanded distal to the apex of the protocone, but does not support a distinct hypocone. The paracrista extends mesiobuccally from the paracone to the stylocone near the mesiobuccal corner of the tooth. A transverse groove caused by wear separates the stylocone from what may be the parastyle positioned mesial to the paracone. This groove has at least partially obliterated the stylocone and parastyle.
M3.—The M3 (Fig. 3D), represented by NMMNH P-44353, is smaller than both M1 and M2. The width of the M3 is reduced distally and the metacone is much smaller than the paracone. The distal margin of the protocone is expanded. A paraconule is present and positioned close to the paracone. A wear facet on the posterior face of the protocone that extends buccally to the metacone has largely erased evidence for a metaconule.
p3.—NMMNH P-34838 is tentatively identified as ap3 (Fig. 4A). The crown is bulbous and dominated by the protoconid; a preprotocristid descends mesially from the protoconid, terminating above the swollen base of the crown. A paraconid is not present. Distally, the tooth supports a small talonid basin that lacks distinct cuspids. A ridge extends distobuccally from the protoconid to form the buccal margin of this basin. A small talonid basin is bounded buccally, distally, and lingually by a raised ridge.
m2.—The m2 (Fig. 4D) is nearly rectangular in occlusal view. The trigonid is low so that the talonid is about twothirds its height The protoconid and metaconid are rounded and subequal in size. The paraconid is high (sensu Zack et al. 2005), lingual and closely appressed to the metaconid, with its base fused to that of the metaconid. The distal wall of the trigonid is strongly worn so that it is oriented nearly transverse. However, we suspect that the distal wall of the metaconid was originally more distally positioned than that of the protoconid and some of it has been removed by wear. The talonid is subequal in width to the trigonid. The hypoconid is the largest of the talonid cuspids followed by the entoconid and the hypoconulid. The talonid is closed lingually by an entocristid that terminates mesially at the base of the metaconid. The hypoconulid is positioned lingual to the midline of the tooth, closely coupled with the entoconid. A distinct notch separates the hypoconulid from the hypoconid. The cristid obliqua intersects the trigonid below the protocristid notch. The tooth is exodaenodont, with the buccal sides of both the trigonid and talonid inflated and overhanging the buccal surfaces of the roots. A small precingulid extends buccally from the hypoconulid and terminates below the hypoconid. A lingual cingulid is not present. A small cuspid in the hypoflexid represents a remnant of the ectocingulid. The distal face of the trigonid and the mesial face of the hypoconid, buccal to the cristid obliqua (wear facets 1 and 3, respectively; sensu Crompton and Kielan-Jaworowska 1978), are heavily worn, with the angle formed between these two planar surfaces closely matching the profile of the M1 paracone and protocone as seen in buccal view (Fig. 3G).
m3.—The single known m3 (NMMNH P-55397; Fig. 4C) closely resembles the m2, but is narrower. The talonid is narrower than the trigonid and more elongate. The hypoconulid is relatively larger and extends further distally.
Fig. 1.
Geologic map showing the the Paleocene Nacimiento Formation (Animas Formation in southern Colorado) and early Eocene San Jose Formation, San Juan Basin, New Mexico, and indicating the location of the Split Lip Flats Local Fauna.
Table 1.
Measurements (in mm) of Chacomylus sladei gen. et sp. nov.
Fig. 2.
Outlines of right lower molar (A) and left upper molar (B) of Chacomylus sladei gen. et sp. nov. showing tooth nomenclature used in the text. Abbreviations.—cc, centrocrista; co, cristid obliqua; ecg, ectocingulum; encd, entocristid; end, entoconid; etx, ectoflexus; hy, hypocone; hyd, hypoconid; hyld, hypoconulid; hyxd, hypoflexid; me, metacone; mec, postmetacrista; med, metaconid; msl, metastylar lobe; mtl, metaconule; pa, paracone; pac, preparacrista; pacd, paracristid; pad, paraconid; pmlc, premetaconule crista; prc, precingulum; prl, paraconule; pprc, preprotocrista; ps, parastyle; pscg, postcingulum; psl, parastylar lobe; psplc, postparaconule crista; psprc, postprotocrista; pplc, preparaconule crista; pr, protocone; prcd, protocristid; prd, protoconid; ps, parastyle; pscg, postcingulid; psl, parastylar lobe; psmlc, postmetaconular crista; st, stylocone; sts, stylar shelf; tb, talon basin; tdb, trigonid basin; tlb, talonid basin.
Fig. 3.
Upper dentition of the hyopsodontid “condylarth” Chacomylus sladei gen. et sp. nov., Nacimiento Formation, lower Paleocene (late Puercan, Pu3), San Juan Basin, New Mexico. A. NMMNH P-38461, left P4 in occlusal (A1, stereopair), buccal (A2), lingual (A3), mesial (A4), and distal (A5) views. B. 41208 (holotype), right M1 in occlusal (B1, stereopair), buccal (B2), lingual (B3), mesial (B4), and distal (B5) views. C. 51537, left partial M2? in occlusal (C1, stereopair), and lingual (C2) views. D. 44353, left M3 in occlusal (D1, stereopair), buccal (D2), lingual (D3), mesial (D4), and distal (D5) views.
Fig. 4.
Lower dentition of hyopsodontid “condylarth” Chacomylus sladei gen. et sp. nov., Nacimiento Formation, lower Paleocene (late Puercan, Pu3), San Juan Basin, New Mexico. A. NMMNH P-34838, left p3? in occlusal (A1, stereopair), buccal (A2), lingual (A3), mesial (A4), and distal (A5) views. B. 44345, right m2 in occlusal (B1, stereopair), buccal (B2), lilngual (B3), mesial (B4), and distal (B5) views. C. 55397, left m3 in occlusal (C1, stereopair), buccal (C2), lingual (C3), mesial (C4), and distal (C5) views.
Fig. 5.
The hyopsodontid “condylarth” Chacomylus sladei gen. et sp. nov., Nacimiento Formation, lower Paleocene (late Puercan; Pu3), San Juan Basin, New Mexico, a cast of the holotype (NMMNH P-41208). A. Electron micrograph a right M1, showing details of the 3(5) wear facet following (Hunter and Fortelius 1994) on the mesial face of the protocone (enlarged in B).
Discussion
Zack et al. (2005) discussed the difficulty in classifying many early Paleocene mammals that are small and have bunodont teeth. Many of these taxa, like Chacomylus sladei, are represented only by dental remains. In addition, several clades such as hyopsodontid “condylarths”, erinaceomorph lipotyphlans, and pentacodontid pantolestans converge in dental morphology (see Zack et al. 2005). C. sladei shows similarities in dental morphology to both hyopsodontid and apheliscid “condylarths”.
Chacomylus sladei is significantly smaller (Table 1) than any hyopsodontid (sensu Zack et al. 2005) with the exception of the early Paleocene, South American Tiuclaenus minutus (Muizon and Cifelli 2000: tables 4, 5) which is only slightly smaller. T. minutus is about 97% the size of C. sladei, based on length of m2. The upper and lower molars of C. sladei fall near or within the size range of the apheliscids Gingerichia geoteretes and Gingerichia hystrix (Zack et al. 2005). This is about 75% the size of Promioclaenus acolytus which is among the smallest of early Paleocene hyopsodontids, and is similar in size to the smallest specimens of the periptychid Oxyacodon agapetillus, previously the smallest “condylarth” documented from the Nacimiento Formation of New Mexico (Archibald et al. 1983).
Chacomylus sladei possesses a unique combination of dental characters that clearly distinguish it from other mammals. It is grossly similar to a number of approximately contemporaneous hyopsodontid and apheliscid “condylarths” (sensu Zack et al. 2005). However there are a number of significant differences. For example, the P4 lacks a parastyle, a condition that is not known for any hyopsodontid (the P4 is not known for a number of hyopsodontid taxa including Oxytomodon, Tiznatzinia, Litomylus orthronepius, and Bubogonia saskia), but it is relatively small in Choeroclaenus. Chacomylus lacks the relatively large size and inflation of the P4 found in a number of hyopsodontids and apheliscids (e.g., Choeroclaenus, Mioclaenus, Gingerichia, Phenacodaptes). In these taxa, the enlarged and swollen P4 is consistently associated with a relatively large and inflated p4. Therefore, we predict that the p4 of Chacomylus is relatively small, probably smaller than the m1. The upper molars of Chacomylus have a narrow stylar shelf consisting of an ectocingulum similar to that of other Puercan hyopsodontids such as Valenia. It is distinctly narrower than that of Litomylus orthronepius. The paracone and metacone are subequal and spaced close together so that their bases are partly conjoined. This is different than is seen in other hyopsodontid or apheliscid “condylarths” in which the bases of the paracone and metacone are usually separated by a cleft lingual and buccal to the centrocrista. Extensive wear over this area of the upper molars of Chacomylus make examination of the centrocrista impossible, however in buccal view (Fig. 3B2) the paracone and metacone are clearly merged significantly above the level of the ectocingulum. Both the M1 and partial M2 of Chacomylus show heavy wear that has obliterated much of the details of the conular region. The M1 and M2 bear large protocones that are broadly expanded lingually so that the tooth maintains a near constant mesiodistal length from a position below the protocone apex to a position at the lingual bases of the para- and metacone. Most hyopsodontids and apheliscids widen buccally, giving the teeth a more triangular shape in occlusal view, or much of the lingual expansion of the tooth is due to the linguodistal expansion of the postcingulum and hypocone rather than the protocone itself. The relative sizes of the molar protocones are similar to that of Litaletes, but the upper molars of Chacomylus are more transverse. The anterior cingulum, only visible on the partial M2 in Chacomylus, resembles that of hyopsodontids in being lower than the postcingulum rather than rising from the same level on the base of the protocone as in apheliscids. The M3 of Chacomylus is about two-thirds the size of the M1 and similar in relative size to that of the hyopsodontid Promioclaenus. The p3 of Chacomylus, if the referral of NMMNH P-44345 is correct, differs from Bomburia, Valenia and Choeroclaenus (the p3 is unknown for Tiznatzinia) in lacking a paraconid and in having a featureless talonid, lacking a medial ridge. The lower molars show a similar relative size of m3 and m2 (m3 L/m2 L of Promioclaenus lemuroides is 0.98; TEW unpublished data), but the m3 of Chacomylus is slightly larger, with a length that exceeds that of m2 (m3 L/m2 L of Chacomylus is 1.08; Table 1). The p3 is similar in shape and in size relative to the m2 as that of the hyopsodontid Litaletes, but differs in being more basally inflated, and in lacking a paraconid and talonid cusps. The lower molars of Chacomylus are strikingly similar to those of Haplaletes disceptatrix. They are similar in length and width, show similar relative height and width of the trigonids and talonids, and similar size and position of most of the cusps. In addition, the distal wall of the trigonid shows a similar nearly transverse orientation and inclination. The talonid is also similar, except that the hypoconulid and entoconulid of Chacomylus are more confluent and are not separated by a distinct notch (see Zack et al. 2005: fig. 7). The entocristid is similar in that it extends mesially to the base of the metaconid to close off the lingual margin of the talonid basin. This is in contrast to the condition seen in most hyopsodontids (Mioclaenus is an exception) in which the entocristid terminates distal to the base of the metaconid, leaving a narrow, V-shaped opening in lingual view. Lower molars are also similar in the degree of distension of buccal enamel. However, the lower molars of Chacomylus differ from those of H. disceptatrix in several important ways; as in other aphelsicids, H. disceptatrix lacks a lingually positioned paraconid perched high on the mesial edge of the metaconid, and it bears strong ectocingulids that extend completely across the buccal margins of the lower molars. The poorly known Oxytomodon perissum is also similar to Chacomylus and possesses a similar degree of exodaenodonty, but differs in the more oblique orientation of the distal wall of the trigonid, and in the shape and orientation of the entocristid which extends mesially to a position near the lingual base of the metaconid rather than directly to its distal base.
Phylogenetic analysis
Zack et al. (2005) revised the Hyopsodontidae and Apheliscidae based on the results of a phylogenetic analysis of 27 small-bodied “condylarths” from the Paleogene of western North America and Europe using a combination of 59 dental and postcranial (mostly tarsal) characters (Zack et al. 2005: appendix 3). Based on this analysis, Zack et al. (2005) identified several phylogenetically informative dental features in both the Apheliscidae and Hyopsodontidae.
As a preliminary examination of the phylogenetic relationship of Chacomylus and in order to test for its relationship with the Apheliscidae and Hyopsodontidae, we scored Chacomylus sladei for the character states provided in Zack et al. (2005: appendix 2). Chacomylus sladei was not scored for the lower premolars because of uncertainty regarding the identification and loci of the single lower premolar tentatively referred to this taxon. In addition, we included two small Puercan (early Paleocene) hyopsodontid or apheliscid taxa, Litomylus orthronepius, from the Rav W-1 locality of southern Saskatchewan and Valenia wilsoni, a hyopsodontid from New Mexico and Wyoming. The scorings of these taxa are provided in Appendix 1.
Zack et al. (2005) performed their parsimony analysis using NONA v2.0 (Goloboff 1999). We ran the same analysis (including the pentacodontid Aphronorus spp.; characters unordered; Zhelestidae was used as the outgroup taxon) using TNT (Goloboff et al. 2008), using first a new technology search (using sectorial search, rachet, drift, and tree fusing) as a driven search, finding the minimum length ten times. Maximum trees was set at 50,000. This was followed by a traditional search of the resulting starting trees from RAM using a tree bisection reconnection (TBR) swapping algorithm. This resulted in two most parsimonius trees, five fewer than obtained by Zack et al. (2005), with a minimum length of 197 (CI: 0.391; RI: 0.632), one step longer than the shortest trees obtained by Zack et al. (2005) with similar CI and RI values. The resulting consensus tree is similar to that obtained by Zack et al. (2005: fig. 8A), but Aphronorus spp. is basal to Protungulatum.
We then revised the scoring of one taxon from the original Zack et al. (2005) character-taxon matrix, finding that the apheliscid Litomylus dissentaneous has “weak” (1) distention of the enamel on buccal side of lower molars (character 8) rather than “absent” (0) exodaenodonty. The results of our analysis with the revised character-taxon matrix resulted in two trees with a minimum length of 207 and significantly higher CI (0.786) and RI (0.936) values. However, the topology of the strict consensus tree is the same.
With the addition of Chacomylus, Litomylus orthronepius, and Valenia wilsoni two most parsimonious trees were found with a minimum length of 207 (CI: 0.372; RI: 0.601). The strict consensus tree (Fig. 6) resembles that of Zack et al. (2005: fig. 8A), with Chacomylus, Litomylus orthronepius, and Valenia wilsoni occupying a position as the sister to Molinodus, near the base of the Hyopsodontidae.
This phylogenetic analysis upholds a monophyletic Hyopsodontidae and Apheliscidae as revised by Zack et al. (2005). However, we find that these nodes are poorly supported (Bremer values of 1; Fig. 6). We also conclude that Chacomylus sladei is a hyopsodontid near the poorly known Litomylus orthronepius. Two of the three characters that are synapomorphies uniting the Hyopsodontidae (Fig. 6, node B) are tarsal characters (characters 51 and 58; Appendix 2) that are unknown for Valenia wilsoni, Litomylus orthronepius, and C sladei. In addition, the single character that unites Molinodus and Valenia + (L. orthronepius + Chacomylus sladei) is unknown for L. orthronepius + Chacomylus sladei. Therefore, we consider these results to be tentative.
Fig. 6.
Strict consensus tree derived from parsimony analysis of data in Zack et al. (2005) with the addition of Chacomylus sladei, Valenia wilsoni, and Litomylus orthronepius (see text and Appendix 1). List of common synapomorphies for selected nodes (labeled A–G) are listed in Appendix 2. Numbers below nodes indicate Bremer branch support calculated from a pool of 50,000 suboptimal trees of up to 10 steps longer than the shortest trees obtained.
Dental function
Chacomylus sladei has unusual morphology and occlusal wear compared to other Puercan mammals. Puercan faunas are dominated by middle sized peryptichid, arctocyonid, and hyopsodontid “condylarths” that emphasize tip-crushing wear that results in the abrasion of enamel from cusp apices (Dewar 2003). In contrast, C. sladei exhibits little or no cusp tip wear. It is instead characterized by the conspicuous development of broad shearing surfaces. Following the terminology of Hunter and Fortelius (1994), upper and lower molars of C. sladei especially emphasize facets 2(1) and 6(3). Facet 2(1) is expanded on the lower molar to involve the entire postvallid wall and the sides of the protoconid and metaconid, and is confluent with facet 3(5). On the upper molar, facet 2(1) occupies most of the mesial sur- face of the tooth, and is confluent with 3(5). Facet 6(3) on the lower molar extends the length of the cristid obliqua superiorly and extends inferiorly to the hypoflexid. On the upper molar, facets 6(3) and 7(4) occupy the entire distal side of the paracone and mesial side of the metacone, respectively. They extend to the cusp tips but there is no wear of the cusp tips. Chacomylus sladei also exhibits development of facets 9(6) and 4(7n). These are small on the lower molars, near the tips of the entoconid and hypoconulid respectively, but they are prominent, expanded, and confluent on P-41208, an upper molar.
The presence of exodaenodonty on the lower molars is unique among hyopsodontine “condylarths” and is probably related to the strong emphasis on shearing and resultant high rate of wear experienced by this animal. Given the small size of C. sladei and that its chewing adaptations are unique among its contemporaries, we infer that it had a specialized diet compared to other Puercan “condylarths.” A similar niche was apparently adapted by the late Paleocene by some apheliscids in the northern Rocky Mountain Region, such as Gingerichia, that are convergent on some of the dental features of Chacomylus.
Conclusion
Williamson (1996) argued that the lower diversity of mammals in late Puercan (Pu3) faunas of New Mexico was largely explained by collecting biases. Screen washing techniques had been employed to recover small middle Puercan (Pu2) mammal taxa that were not recoverable from concreted Pu3 sites. This resulted in some apparent faunal differences and an apparently lower diversity for the late Puercan. The recovery of a new hyopsodontine “condylarth,” as well as other small mammals (Williamson and Weil 2002; Williamson et al. in press), from a new, late Puercan microvertebrate locality addresses some of the previous collecting bias. The absence of Chacomylus sladei from older, Pu2 microvertebrate localities, despite the more intensive microvertebrate sampling efforts, suggests that it is absent from older faunas of the Nacimiento Formation. We therefore conclude that it represents a potential Pu3 index taxon.
Chacomylus sladei exhibits a number of unique features compared to its contemporary hyopsodontids including small size, enlarged upper molar protocones, and exodaenodonty of the lower molars. The small size, lower molar exodaenodonty, and attritional wear indicating emphasis on transverse chewing, are unusual for a Puercan “condylarth” and suggest that Chacomylus sladei had a different diet than other hyopsodontines. Indeed, it may have had a specialized diet compared to other contemporary therian mammals. It ignificantly increases the known morphological, and probably ecological, diversity for late Puercan mammalian faunas.
Acknowledgements
We express gratitude to Pat Hester and Sherri Landon (Bureau of Land Management, Albuquerque, USA) for providing permitting and field assistance. We thank numerous individuals for assistance in the field and for microvertebrate recovery using underwater screening methods including Lavina Becenti, Jimmy Benally, Garrett Briggs, Chris Hughes, Warren Slade, and Will Tsosie (New Mexico Museum of Natural History & Science, Albuquerque, USA) and Kent Smith (Oklahoma State University, Tulsa, USA). We are appreciative of the constructive reviews by David Archibald (San Diego State University, San Diego, USA), Craig Scott (Royal Tyrrell Museum of Palaeontology, Drumheller, Canada), and an anonymous reviewer that led to an improved manuscript. This work was supported by the National Science Foundation (EAR 0207750 to TEW and EAR 0207732/EAR 0654096 to AW).
References
Appendices
Appendix 1
Scoring of Chacomylus sladei, Litomylus orthronepius, and Valenia wilsoni to the character states provided by Zack et al. (2005). Scoring for L. orthronepius is based on examination of UALP 15121, 15122 (holotype), 15123, 16033–16035, and 16038 and descriptions and illustrations in Johnston and Fox (1984). Scoring for V. wilsoni is based on examination of KU 9446 (holotype), AMNH 23157, 59788, 59904, and NMMNH P-15212 and descriptions in Van Valen (1978, 1988) and Muizon and Cifelli (2000).
Appendix 2
List of characters in common on the most parsimonious trees diagnosing the selected nodes on the strict consensus tree in Fig. 6. Characters are from Zack et al. (2005: appendix 2).
A. (Hyopdosodontidae + Apheliscidae)
Character 9. Lower molar trigonids much higher than talonids (0), somewhat higher than talonids (1), or trigonids and talonids subequal in height (2): 0 → 1
Character 19. Hypoconulid on m1–2 on lingual side of talonid, twinned with entoconid (0), in a median position, separate from hypoconid and entoconid (1), or shifted buccally and twinned with hypoconid (2): 1 → 0
Character 25. m3 trigonid width subequal to or slightly wider than m2 trigonid width (0), somewhat narrower than m2 trigonid width (1), or much narrower than m2 trigonid width (2): 0 → 1
B. (Hyopsodontidae)
Character 13. m2–3 paraconid distinctly separated from metaconid (0) or partially to completely fused to metaconid at base (1): 0 → 1
Character 51. Medial trochlear ridge of astragalus absent such that medial portion of trochlea faces dorsomedially (0) or present such that medial portion faces medially (1): 1 → 0
Character 58. Posteromedial projection of astragalar body absent (0) or present (1): 0 → 1
C. Molinodus suarezi + (Valenia wilsoni + [Litomylus orthronepius + chacomylus sladei])
Character 2. p4 paraconid better developed than metaconid (0), as developed as metaconid (1), or weaker than metaconid (2): 1 → 2
D. (Valenia wilsoni + [Litomylus orthronepius + Chacomylus sladei]) Character 29. P4 metacone absent (0), weakly developed (1), or well developed and well separated from paracone (2): 0 → 1
Character 34. M1–2 paracone subequal in size to metacone (0) or larger than metacone (1): 1 → 0
Character 42. M1–2 postprotocingulum absent (0) or present (1): 0 → 1
E. Litomylus orthronepius + Chacomylus sladei
Character 8. Distention of enamel on buccal side of lower molars absent (0), weak (1), or strong (2): 0 → 1
Character 43. M1–2 hypocone small to absent (0), well developed but smaller than protocone (1), or subequal in size to protocone (2): 0 → 1
F. Chacomylus sladei
Character 45. M1–2 postcingulum arises from same level on protocone as anterior cingulum (0) or arises higher on protocone than anterior cingulum (1): 0 → 1
