n^o ,50 ^ FIELDIANA GEOtOGYUBRARy Geology NEW SERIES, NO. 50 Notoungulata and Litopterna of the Early Miocene Chucal Fauna, Northern Chile Darin A. Croft John J. Flynn Andre R. Wyss CO c\# CD July 30, 2004 Publication 1528 PUBLISHED BY FIELD MUSEUM OF NATURAL HISTORY Information for Contributors to Fieldiana Fieldiana is a peer-reviewed monographic series published by the Field Museum. The series publishes the research of staff members and our research associates. Submissions from nonaffiliated authors may be considered as space permits; however, nonaffiliated authors will be required to pay full page charges. The page charge currently is $65 per page. This figure is subject to change. All authors are encouraged to provide fiinding to support the production of their works. Submission procedures: A submission procedures document is available from the scientific editor of the journal and on the Museum's web site (by 2004). Manuscripts submitted for review should be prepared using standard electronic software and submitted to the scientific editor in three photocopies. The scientific editor also requires the names, addresses, telephone numbers, and e-mail addresses of at least five suggested reviewers. All elements of the work, including the illustrations, must be completed before the manuscript is submitted. References should be styled according to a recent Fieldiana publication, which may also be consulted for order of elements (title page, table of contents, list of illustrations, abstract, and the like). Length: As a monographic series, Fieldiana publishes mid-length works. We do not publish short, joumal-article-length works. As a general rule, we are unable to accommodate manuscripts of less than 75-100 manuscript pages. Authors seeking an exemption for a manuscript of fewer pages should consult the scientific editor. Fieldiana also is unable to publish extremely long works without substantial support from the author. The publication of long works should be arranged well in advance with the scientific editor. Please bear in mind: Changes on page proofs are very expensive. Author-generated changes on page proofs can only be made if the author agrees in advance to pay for them. UNIVERSITY OF ILLINOIS LIBRARY AT U^SANA-CHAMPAIGN 6E0L0GY FIELDIANA GEOLOGY LIBRARY Geology NEW SERIES, NO. 50 Notoungulata and Litopterna of the Early Miocene Chucal Fauna, Northern Chile Darin A. Croft Department of Anatomy Case Western Reserve University School of Medicine 10900 Euclid Avenue Cleveland, Ohio 44106 U.S.A. Department of Geology Field Museum of Natural History 1400 South Lake Shore Drive Chicago, Illinois 60605-2496 U.S. A. John J. Flynn Department of Geology Field Museum of Natural History 1400 South Lake Shore Drive Chicago, Illinois 60605-2496 U.S.A. Andre R. Wyss Department of Geological Sciences University of California— Santa Barbara Santa Barbara, California 93106 U.S.A. Published July 30, 2004 Publication 1528 PUBLISHED BY FIELD MUSEUM OF NATURAL HISTORY © 2004 Field Museum of Natural History ISSN 0096-265 1 PRINTED IN THE UNITED STATES OF AMERICA Table of Contents Abstract 1 Introduction 1 Geographic and Geologic Setting 2 Materials and Methods 2 Abbreviations 4 Systematic Paleontology 4 Notoungulata 4 Toxodontidae 4 Nesodon imbricatus 5 Adinotherium sp 8 Mesotheriidae 11 Altitypotherium paucidens 12 Altitypotherium chucalensis 20 Eotypotheriuum chico 25 Hegetotheriidae 31 Hegetotherium cf. H. mirabile 33 Litoptema 36 Macraucheniidae 36 Theosodon sp. indet 38 Phylogeny of Mesotheriines 40 Age and Faunae Comparisons 42 Discussion and Conclusions 47 Acknowledgments 49 Literature Cited 49 9. Upper dentition of holotype of Altity- potherium chucalensis, SGOPV 4100, a partial skull with portions of all pre- molars and complete molars 21 10. Portions of lower dentition oi Altity- potherium chucalensis, SGOPV 4100 .... 22 1 1 . Holotype of Eotypotherium chico, SGOPV 5157, a partial right maxilla and zygomatic arch with P3-M3 27 12. Upper dentitions referred to Eotypo- therium chico 28 13. Partial left mandible with m2-3, SGOPV 5158, referred to Eotypothe- rium chico 29 14. Comparison of the degree of fusion of the tibia and fibula among selected Santacrucian typotheres 32 15. Isolated upper cheek teeth referred to Hegetotherium cf. H. mirabile 34 16. Lower dentitions referred to Hegeto- therium cf. H. mirabile 35 17. Selected postcranial specimens referred to Theosodon sp. indet 39 18. Strict consensus trees of mesotheriine relationships 43 List of Tables List of Illustrations 1 . Geography and geology of the Chucal Formation 3 2. Partial skull of Nesodon imbricatus, SGOPV 4053 7 3. Bivariate plot of lower molar dimen- sions for Adinotherium and Nesodon 8 4. Partial left ?i2 of Adinotherium, SGOPV 4064 9 5. Relative sizes of mesotheriines and toxodontids from Chucal, as illustrated by the size of the ulna 10 6. Morphology of the proximal radius in mesotheriines and Adinotherium 11 7. Holotype of Altitypotherium paucidens, SGOPV 4038, a partial rostrum with left II, left P4-M3 and right P4-M1 .... 13 8. Lower dentitions referred to Altitypo- therium paucidens 14 1. Measurements for Chucal specimens of Nesodon imbricatus 6 2. Measurements for specimens of Altity- potherium paucidens 12 3. Development of suborbital fossa and size of infraorbital foramen in selected mesotheriines 19 4. Measurements for specimens of Altity- potherium chucalensis 20 5. Measurements for specimens of Eoty- potherium chico 26 6. Measurements for Chucal specimens of Hegetotherium mirabile 37 7. List of characters and character states used in phylogenetic analysis of me- sotheriine relationships 41 8. Character- taxon matrix for phylogenet- ic analysis of mesotheriine relation- ships 42 9. Ungulates of middle-latitude early to middle Miocene faunas of northern Chile and Bolivia 45 m 10. Ungulates of selected low-, middle-, 12. Revised taxonomic list of the Chucal and high-latitude Miocene faunas of Fauna 47 South America 46 13. Body mass estimates of mesotheri- 1 1 . Ungulate faunal resemblance indices ines 48 (Simpson coefficients) among selected low-, intermediate-, and high-latitude Miocene faunas of South America 47 IV Notoungulata and Litopterna of the Early Miocene Chucal Fauna, Northern Chile Darin A. Croft John J. Flynn Andre R. Wyss Abstract This study describes the notoungulates and Htopterns (endemic South American ungulates) from the Chucal Fauna of northern Chile. Eight species are known to date, including seven notoungulates and one litoptern. The bulk of the toxodontid notoungulate material, including a well-preserved partial skull, is referred to Nesodon imbricatus. A few postcranial elements and a single partial tooth are referred to Adinotherium sp. indet. Poorly preserved dental ma- terial may represent a third (indeterminate) toxodontid. Fragmentary dental and postcranial remains are referred to Hegetotherium cf. H. mirabile (Hegetotheriidae). Theosodon sp. indet. (Litopterna: Macraucheniidae) is recorded on the basis of a variety of diagnostic limb bones, but dental remains of this taxon are not yet recorded at Chucal. Three new species of meso- theriid notoungulates are described: Altitypotherium paucidens, Altitypotherium chucalensis, and Eotypotherium chico. The largest, Altitypotherium paucidens, differs from all previously described mesotheriids in the loss of an upper premolar (P3). Altitypotherium chucalensis is represented by an associated skull and mandibles plus postcranial remains. Eotypotherium chi- co, based on gnathic remains, is the smallest mesotheriid known. Bracketing radioisotopic dates and biostratigraphic correlation argue for referral of the Chucal Fauna to the Santacrucian South American Land Mammal "Age" (late early Miocene). The forms from Chucal are the only mesotheriids known for this temporal interval and also mark the earliest known occurrence of mesotheriines. A phylogenetic analysis places the three taxa from Chucal as early diverging members of the Mesotheriinae, with Eotypotherium chico representing the outgroup to all re- maining members of the clade. In the presence and abundance of mesotheriines and in the absence of interatheriids, the Chucal Fauna more closely resembles geographically proximate (but younger) faunas in Bolivia and the Precordillera of northernmost Chile than temporally correlative faunas from southern South America. This pattern indicates significant faunal pro- vinciality in South America during the late early Miocene and suggests that the intermediate latitudes might have been a center of diversification for mesotheriines and potentially other groups of mammals. Introduction Study of South American fossil mammals has traditionally focused on faunas from the high lat- itudes, especially Patagonian Argentina (Patter- son &. Pascual, 1972; Simpson, 1980). The dis- covery of faunas elsewhere in South America (e.g., Bolivia, Brazil, Chile, Colombia) has con- tributed to more complete understanding of the evolution of South American mammals but has also revealed unanticipated complexities (see Flynn «fe Wyss, 1998; Flynn, 2002). These extra- Patagonian faunas demonstrate, for one, that lat- itudinal provinciality characterized South Amer- ican mammal faunas for much of the Cenozoic, necessitating sampling of a wide geographic range for any given time slice to gain a fuller understanding of mammal evolution on the con- tinent (Pascual et al., 1985; Wyss et al., 1994; Madden et al., 1997). Our research team has been working to un- cover and develop fossil mammal faunas of var- FIELDIANA: GEOLOGY, N.S., NO. 50, JULY 30, 2004, PP. 1-52 ious ages throughout the length of Chile. The Chucal Fauna represents the northeasternmost of these fossiliferous localities. The first fossil ver- tebrate specimen from this fauna, discovered in 1992 during the course of geologic mapping northeast of Salar de Surire (Fig. 1), pertains to Nesodon, an early Miocene toxodontid notoun- gulate (Charrier, Mufioz, Wyss, et al., 1994). To determine whether the Chucal Formation would produce other identifiable fossil mammal re- mains, a return trip to the area was undertaken in July 1998. This expedition yielded a variety of well-preserved specimens, including a diver- sity of endemic ungulates, as well as the oldest chinchilline rodent (Flynn, Croft, et al., 2002). A second expedition to the area in 2001 produced many additional specimens and at least five spe- cies not recorded previously (Charrier et al., 2002). The present study provides descriptions and taxonomic analyses of the most diverse and abundant component of the Chucal Fauna, of its notoungulates, as well as of its single litoptern. Prior to discovery of the Chucal Fauna, only isolated Cenozoic mammal specimens had been reported from northern Chile. The first, a meso- theriid mandible (MLP 86-VIII-lO-l), was col- lected by D. Pacci in 1969 in the area of Caragua (approx. 18° 25 'S, 69°35'E) west of the town of Belen. The second, a nearly complete mesotheriid skull and mandibles with associated postcrania (SGOPV 4004), was discovered in 1989 (Salinas et al., 1991), probably from the same locality as Pacci's specimen. Both specimens were collected from the Huaylas Formation, a unit demonstrably younger (middle late Miocene, definitely <16 Ma and most likely ~ 10.8-1 1.7 Ma; Flynn et al., in press) than the Santacrucian-aged horizons de- scribed here. The two mesotheriid specimens from Caragua (and a third discovered in the same area by our team in 2001) have been referred to a new mesotheriine taxon (Flynn et al., in press). Geographic and Geologic Setting Charrier et al., 2000, 2002; Womer et al., 2000; Chavez, 2001; Bond & Garcia, 2002; Flynn, Croft, et al., 2002), including initial paleontological re- ports, geochronologic analysis, and studies of the syntectonic deposition of the fluviolacustrine sedi- ments of the Chucal Formation (Fig. 1). Fossilif- erous strata of the Chucal Formation are well ex- posed on both limbs of a major N-S-oriented growth anticline (the Chucal Anticline), which doc- uments marked thickness and facies variation on either flank and at least four progressive erosional unconformities (U1-U4 in Charrier et al., 2002; Fig. 1; see also Riquelme & Herail, 1997, and Ri- quelme, 1998). As detailed in Charrier et al. (2002), the Chucal Formation at Cerro Chucal is thickest (—600 m) on its west-dipping (35-40°) western flank (informal Members W1-W4) and is thinner on its steeply east-dipping (generally >50°) eastern flank (informal Members E1-E3), where the syntectonic erosional unconformities are most pronounced. Mammal fossils are most common in the floodplain facies (and some lacustrine-influ- enced intervals) in the lower parts of Members W3 and W4 and the fluviatile facies of Members El and E3 (Charrier et al., 2002). Unconformities separate the Chucal Formation from the underlying Lupica Formation (early Miocene, dated at 21.7 ± 0.8 Ma at Chucal and at >18.5 Ma elsewhere in the region) and the overlying Quebrada Macusa Formation (the base of which is dated at 17.5 ± 0.4 Ma at Cerro Chu- cal; Muiioz, 1991; Riquelme, 1998; Bond & Garcia, 2002; Charrier et al., 2002). The Chucal Formation, therefore, was deposited and deformed very rapidly (possibly over considerably less than 4-5 million years); it is no older than 22.5 Ma (possibly 18.5 Ma) and no younger than about 17 Ma. Thus, based on both the biostratigraphic oc- currences and radioisotopic constraints, we con- sider the mammalian fossils from the Chucal For- mation in this region to represent a single short- duration assemblage, the Chucal Fauna (Charrier et al., 2002; Flynn et al., 2002). Material described herein is derived from the <600-m-thick Chucal Formation near Cerro Chu- cal, located on the Chilean Altiplano northwest of the Salar de Surire (18°43'S, 69°10'W). Several earlier studies have detailed the geology of this re- gion (e.g., Mufioz, 1991; Charrier, Mufioz, & Pal- ma-Heldt, 1994; Charrier, Mufioz, Wyss, et al., 1994; Riquelme & Herail, 1997; Riquelme, 1998; Materials and Methods Upper tooth loci are indicated by uppercase let- ters (e.g., II, P2, Ml) and lower tooth loci by lowercase letters (e.g., il, p2, ml); deciduous teeth are indicated by a "D/d" preceding the tooth position. Terminology for toxodontid molar mor- phology follows Madden (1990). All measure- FIELDIANA: GEOLOGY CROFT ET AL.: CHUCAL UNGULATES merits were taken to the nearest O.I mm using digital calipers. We recognize that many groups (e.g., Nesodon- tinae, Hegetotheriinae), as currently conceived, are likely paraphyletic. Because these names are familiar to South American paleomammalogists, we continue to employ them in their traditional sense until phylogenetic definitions and/or alter- nate names are proposed. Taxonomic names that are potentially invalid because of synonymy (e.g., Mesotherium "angus- tirostmm") are indicated through the use of quo- tation marks. To our knowledge, no comprehensive analysis of the stratigraphic and geographic occurrences of the taxa considered in the present study has been published previously. The distribution cited for each taxon therefore represents a preliminary ef- fort to detail the geologic formations in which that taxon is known to occur based on firsthand ob- servations and summary compilations of Bonde- sio et al. (1980), Marshall et al. (1983), Savage and Russell (1983), Mones (1986), Madden (1990), and Marshall and Sempere (1991). The Friasian SALMA remains controversial (Flynn & Swisher, 1995; Madden et al., 1997; Flynn, Croft, et ai., 2002). We distinguish be- tween Friasian in its traditional sense (Friasian sensu lato; subsuming the Mayoan, Colloncuran, and Friasian sensu stricto SALMAs) and the term's more restrictive definition (Friasian sensu stricto; including only the temporal interval char- acterized by the type assemblage in southern Chile, and correlative faunas), following Flynn, Croft, et al. (2002). Abbreviations AMNH, American Museum of Natural History, New York; FMNH, Field Museum of Natural His- tory, Chicago; GB, Servicio Geologico de Bolivia (housed at the Museo Nacional de Historia Nat- ural, La Paz); MACN, Museo Argentino de Cien- cias Naturales, "Bernardino Rivadavia," Buenos Aires; MLP, Museo de La Plata, Argentina; MNHN, Museum National d'Histoire Naturelle, Paris; PU, Princeton University Collection (now housed at Yale University, New Haven, Connect- icut); SGOPV, vertebrate paleontology collec- tions, Museo Nacional de Historia Natural, San- tiago; Ma, megannum; mm, millimeter; cm, cen- timeter; S.I., sensu lato; s.s., sensu stricto; SAL- MA, South American Land Mammal "Age"; C\, consistency index; RI, retention index; RC, re- scaled consistency index. Metacarpals and meta- tarsals are abbreviated MC and MT, respectively. Systematic Paleontology Mammalia Linnaeus, 1758 Notoungulata Roth, 1903 Toxodontia Owen, 1853 Toxodontidae Gervais 1847 Nesodontinae Murray, 1866 Diagnosis — The Nesodontinae, as traditionally conceived (e.g., Nesodon, Adinotherium, Proadi- notherium, etc.), is likely a paraphyletic group of toxodontids lacking derived character states pre- sent in other toxodontid subclades (Madden, 1990; Nasif et al., 2000). Specimens from Chucal are referred to the Toxodontidae based on large size; presence of very hypsodont cheek teeth; cheek teeth with protocone and hypocone united early in wear; upper molars with distinct and per- sistent first crista; lingual origin of the crochet in upper molars; triangular cross section of i 1 ; hyp- selodont, procumbent, and tusklike 13; and re- duced enamel along the lingual portions of the lower molars (Madden, 1990; Cifelli, 1993; Nasif et al., 2000). Nesodontines differ from other tox- odontids in absence of derived mandibular sym- physis morphology (Madden, 1990); unreduced sagittal crest (Nasif et al., 2000); lack of hypse- lodont molars (although M3 roots form only in old individuals; Madden, 1990); absence of sim- plified, straight ectoloph (i.e., lacking distinct par- astyle, paracone, and metacone) on P2-4; pres- ence of lingual folds on upper molars that become completely isolated as fossettes; and presence of two lingual folds plus an accessory fossettid in the lower molars. Nesodon Owen, 1846 Type Species — Nesodon imbricatus. Included Species — The type, Nesodon con- spurcatus, and Nesodon cornutus. Diagnosis — A nesodontine differing from Proadinotherium in larger size; greater degree of hypsodonty; presence of II with trapezoidal cross section (triangular in Proadinotherium); absence of cingulae in II; presence of simple and rounded FIELDIANA: GEOLOGY F3 fossette on upper molars; reduction of lingual enamel band on ml; presence of flattened and sharply angled posterior external face of ml tal- onid; and presence of deep, wide lingual groove on molars (groove is shallow in Proadinothe- hiim). Differs from Adinotherium in larger size (approximately twice as large in linear dimen- sions), absence of angled posteroventral border of mandible, presence of II with trapezoidal cross section (kidney shaped in Adinotherium), pres- ence of P2-4 lingual precingulae, presence of rel- atively larger lower premolar row (p2-p4 length >50% ml-m3 length in Nesodon, <50% in Adi- notherium), absence of anterior and entohypocon- ulid folds on p3, relatively lower-crowned p4, re- duction of lingual enamel band on ml, presence of flattened and sharply angled posterior external face of ml talonid, and distinct calcaneal mor- phology (see Scott, 1912). Differs from nearly contemporaneous Palyeidodon (likely a basal tox- odontine and not a nesodontine, fide Madden, 1990; Cifelli, 1993; Nasif et al., 2000) in unre- duced sagittal crest, presence of more developed upper premolar parastyles, presence of upper pre- molar lingual precingulum, presence of parastyle sulcus on Ml, presence of accessory fossette on upper premolars, presence of F3 accessory fos- sette in upper molars, isolation of primary lingual fold as fossette in later wear stages in M2 and M3, presence of more persistent F3 fold/fossette in upper molars, and presence of accessory fos- settid in lower molars (Madden, 1990; Cifelli, 1993; Nasif et al., 2000). Distribution — Santa Cruz Formation, Patagon- ia, Argentina, late early Miocene age, Santacru- cian SALMA (Scott, 1912; Madden, 1990); Chu- cal Formation, northern Chile, late early Miocene age, Santacrucian SALMA; Cura-Mallfn Forma- tion, south-central Chile, late early Miocene age, Santacrucian SALMA (Croft, Radic, et al., 2003); unnamed formation (possibly equivalent to Rio Zeballos Formation), southern Chile, late early Miocene age, Santacrucian SALMA (Flynn, No- vacek, et al., 2002). Comments — Nesodon is best known from the rich late early Miocene Santa Cruz deposits along the Atlantic coast of Patagonia (Scott, 1912). Re- cent studies suggest that Nesodon may serve as a useful index taxon for the Santacrucian SALMA (Croft, Radic, et al., 2003). Reports of Nesodon from other time periods and from areas outside of Chile and Patagonia are poorly substantiated and are therefore not included in the above list (Flynn, Croft, et al., 2002; Croft, Radic, et al., 2003). Bond and Garcia (2002) questionably referred three poorly preserved and fragmentary toxodon- tid teeth from the upper part of the Chucal For- mation to Palyeidodon, a taxon resembling Ne- sodon that occurs in slightly younger deposits (early middle Miocene Friasian s.s. and Collon- curan SALMAs; Madden, 1990). They further noted that these teeth (all pertaining to a single individual) possess a mix of features normally found in those two taxa as well as some distinct features (autapomorphies) that indicate it might represent a new taxon. The dental similarities of Nesodon and Palyeidodon make them difficult to distinguish without sufficiently complete and well-preserved specimens. Moreover, some of the characters used by Bond and Garcia (2002) to dif- ferentiate the two are of ambiguous utility: the absence of a posteroexternal fossette in M 1 (char- acteristic of Palyeidodon) also occurs in some specimens oi Nesodon (e.g., FMNH P10379), the Ml lingual fold becomes isolated with wear in both Nesodon and Palyeidodon (although this is not the case for M2 and M3; Madden, 1990), and it is unknown whether the isolated Ml they de- scribed would eventually have formed roots (as in Nesodon) or would have maintained open roots (as in Palyeidodon) later in life. Additionally, the specimens described by Bond and Garcia (2002) differ in several respects from both Nesodon and Palyeidodon: the parastylar region is much more pronounced and is deflected lingually instead of labially (as noted by Bond and Garcia), the ec- toloph is strongly convex instead of straight or slightly concave (also noted by Bond and Garcia), and the teeth (especially Ml) are significantly larger than those of both Nesodon (Scott, 1912) and Palyeidodon (Madden, 1990). It may there- fore represent a new toxodontid, but one that cur- rently is too poorly represented to be named. Pending description of more diagnostic material, we consider these teeth to pertain to an indeter- minate toxodontid. Nesodon imbricatus Owen, 1846 (Figures 2, 4, Table 1) Referred Specimens from the Chucal For- mation — SGOPV 5073, partial skull with broken right ?I2, right P2-M3, isolated left M2; SGOPV 4043, fragmentary and poorly preserved skull and right mandible; SGOPV 4027, left postorbital pro- cess; SGOPV 4054, incisor/canine and other tooth fragments (unprepared); SGOPV 4066, left M3; CROFT ET AL.: CHUCAL UNGULATES Table 1 . Measurements for Chucal specimens of Nesodon imbricatus. P2 P3 P4 Ml M2 Upper dentition L W W W W W M3 W SGOPV 5073 (R) SGOPV 5073 (L) SGOPV 4066 (L) Lower dentition m2 19.8 25.2 21.6 28.3 31.4 27.6 ni3 45.1 44.0 34.5 34.4 W W 58.6 46.4 31.3 (26.6) SGOPV 4105 (R) 39.8 17.2 SGOPV 4099 (L) (75) (22) SGOPV 5036, right mandibular fragment with portions of ?p3-4; SGOPV 4105, right m2; SGOPV 4099, left mandibular fragment with m3; SGOPV 5118, right ?dp; SGOPV 4018 partial left glenoid; SGOPV 5225, left humerus (see Charrier, Murioz, Wyss, et al., 1994); SGOPV 4044, prox- imal left ulna fragment; SGOPV 4040, fragmen- tary large postcranial bones, right metacarpal II, and right magnum; SGOPV 4015, fragmentary postcranial bones (associated with SGOPV 5036). Age and Distribution — Santa Cruz Formation, Patagonia, Argentina, late early Miocene age, Santacrucian SALMA; Chucal Formation, north- ern Chile, late early Miocene age, Santacrucian SALMA. Diagnosis — Differs from Nesodon conspurca- tus in larger size. Differs from N. cornutus in low- er occiput and absence of dermal horn. Description — The most complete specimen of Nesodon imbricatus in the Chucal collections is SGOPV 5073, a partial skull with much of the right toothrow preserved. Although the bone is not especially well preserved, most of the teeth are in excellent condition. The skull is partially prepared, and the occlusal surfaces of right P2- M3 have been exposed, as has the entire length of an associated left M2 (almost certainly from the same individual but displaced from its alveo- lus). Other teeth are present as "float" in the block containing the skull but remain unprepared. Dental measurements for this and other specimens referred to N. imbricatus are presented in Table 1 . The premolars of SGOPV 5073 are all roughly quadrangular in outline. The anterior half of P2 is not preserved, portions of the anterior and lingual faces of P3 are missing, but P4 is essentially com- plete. The ectoloph of each tooth includes a par- astyle and parastyle sulcus. These structures are most pronounced in P2 and less so in more pos- terior premolars. All three premolars show mod- erate wear, and the only occlusal feature remain- ing is the central fossa. The fossa is a very small oval in P2 and is larger and more elongate an- teroposteriorly in P3 and P4. A noteworthy feature of the premolar series is the variation in the level of the occlusal surfaces of the teeth. The occlusal surface of P4 is at the same level as the molars, while that of P2 is more dorsally situated, approximately 1.75 cm nearer the base of the tooth. The third premolar bridges this gap in occlusal surfaces; most of P3 is at the same level as P2, but the posterior quarter has been worn such that it forms an extremely steep dentine slope (ranging from 45° to 85° at the lin- gual and labial margins, respectively) that blends into the anterior face of Ml. As in many other notoungulates, enamel is absent along the labial portions of both the posterior face of P4 and the anterior face of Ml in this area, resulting in the confluence of the two dentine surfaces. The first and second upper molars are both trap- ezoidal. The ectolophs are straight with little ev- idence of a parastyle or parastyle sulcus. Only the central fossa is present on the occlusal surface of Ml. In M2, the bifurcate primary lingual enamel fold has become isolated as a large, bifurcate, cen- tral fossa. The F3 fold is present as a small fos- sette located almost directly posterior to the point of bifurcation of the central fossa. The third upper molar is nearly triangular in shape and displays less wear than the two anterior molars. Both the primary lingual enamel fold and the F3 fold are present; the former has not yet become isolated by wear, while the latter recently has. Specimen SGOPV 5118 is apparently an un- erupted deciduous right lower premolar Its lack of wear, small size, and lack of roots suggest it was from a very young individual. The anterior FIELDIANA: GEOLOGY Fig. 2. Partial skull of Nesodon imbricatus, SGOPV 4053, including right ?I2 and P2-M3. A. Right lateral view. B. Occlusal view. Scale bars equal 5 cm. and posterior faces converge slightly toward the base (in lingual view), indicating that the tooth would not have increased in size with further de- velopment. It is similar in overall form to the an- terior tooth of MLP 12-245, a partial right man- dible with ?dp4-ml from Santa Cruz referred to Nesodon imbricatus (pers. observ.). Relatively few postcranial bones from Chucal are referable to Nesodon; these include the left humerus found in 1992 (SGOPV 5225), a partial left glenoid (SGOPV 4018), a proximal left ulnar fragment (SGOPV 4044), and some fragmentary large long bones associated with a right MC II and right magnum. Comments — The size difference between Ne- sodon imbricatus and N. conspurcatus is pro- nounced, implying that the validity of the two species is well founded (Croft, Radic, et al., 2003); in this regard, specimens from Chucal compare well with N. imbricatus (Fig. 3). The sta- tus of N. cornutus is less certain; the species is represented only by the holotype (a nearly com- plete skull with heavily damaged teeth) and is dis- tinguished from other species by skull proportions CROFT ET AL.: CHUCAL UNGULATES 3.0- ■ ■ ^ ^ 2.8- A ^ A - A 2.6- 2.4- D • ■ 2.2- o n • A A - OoO o o 8 A A A 1 ' ' 1 ' ' 1 2.0- m1 Adinotherium O N. imbricatus # N. conspurcatus # m2 m3 D A ■ A ■ A 1.8- 1 r 1 1 ' 1 2.5 2.8 3.1 3.4 In(Length) 3.7 4.0 4.3 Fig. 3. Bivariate plot of log-transformed lower molar dimensions for various specimens of Adinotherium, Nesodon imbricatus, and Nesodon conspurcatus. Chucal specimens SGOPV 4105 (m2) and SGOPV 4099 (m3) are denoted by an asterisk (*). and the possible presence of a dermal horn. It is similar in size to smaller specimens of A^. imbri- catus (Scott, 1912). Given the rarity of this taxon, the lack of diagnostic dental characters, the lack of appropriate cranial material for comparison, and the overall large size of the species of Neso- don from Chucal, there is no basis for referring it to A^. cornutus (if indeed this species is valid). Adinotherium Ameghino, 1887 Type Species — Adinotherium ovinum. Included Species — The type, Adinotherium ka- raikense, Adinotherium robustum, Adinotherium splendidum, and Adinotherium nitidum. Diagnosis — A nesodontine differing from Proadinotherium in greater degree of hypsodonty; presence of II with kidney-shaped cross section and lingual groove (triangular cross section in Proadinotherium); absence of cingulae on II; and presence of deep, wide lingual groove on molars (groove is shallow in Proadinotherium). Differs from Nesodon in smaller size (40-50% smaller in tooth dimensions), presence of II with kidney- shaped cross section and lingual groove (trape- zoidal cross section in Nesodon), presence of an- gled posteroventral border of mandible (rounded in Nesodon), absence of P2-4 lingual precingulae, presence of relatively shorter lower premolar row (p2-p4 length <50% ml-m3 length in Adino- therium, >50% in Nesodon), presence of anterior and entohypoconulid folds on p3, presence of rel- atively higher-crowned p4, larger lingual enamel band on m 1 , absence of flattened and sharply an- gled posterior external face of ml talonid, and distinctive calcaneal morphology (see Scott, 1912). Distribution — Santa Cruz Formation, Patagon- ia, Argentina, late early Miocene age, Santacru- cian SALMA (Scott, 1912; Madden, 1990); Chu- cal Formation, northern Chile, late early Miocene age, Santacrucian SALMA; unnamed formation (possibly equivalent to Rio Zeballos Formation), southern Chile, late early Miocene age, Santacru- cian SALMA (Flynn, Novacek, et al., 2002). FIELDIANA: GEOLOGY Fig. 4. Partial left ?i2 of Adinotherium, SGOPV 4064, in labial view. Scale bar = 5 mm. Comments — Although Scott (1912) noted dif- ferences in the vertebral column, scapula, and an- kle of Adinotherium and Nesodon, morphological characters currently used to distinguish the two taxa are all craniodental (Madden, 1 990; Nasif et al., 2000). The limb bones of toxodontids are fair- ly easy to distinguish from those of other endemic South American ungulate clades (e.g., litopterns, astrapotheres, pyrotheres) and other contempora- neous notoungulates (e.g., mesotheriids, hegetoth- eriids). Given the significant metric differences between Adinotherium and Nesodon, postcranial specimens of the two Santacrucian toxodontids can be discriminated on the basis of size alone (Scott, 1912; Croft, Radic, et al., 2003). Since only a single tooth of Adinotherium (a partial in- cisor. Fig. 4) has been collected at Chucal, refer- ence of postcranial elements to this taxon is based on their similar morphology — but much smaller size — compared to Nesodon (Fig. 5). Adinotherium sp. indet. (Figures 4-6) Referred Specimens from Chucal — SGOPV 4064, partial ?left i2, partial right articular con- dyle, various ?mandibular fragments (Fig. 4); SGOPV 4012, left unciform, proximal left MC IV, left MC V, dorsal half of right magnum, many other partial hand/foot bones and bone fragments; SGOPV 5035, phalanx, many small bone frag- ments; SGOPV 4065, partial femoral head, prox- imal ?MT II, many limb bone fragments; SGOPV 4084, proximal portions of left ulna and radius, ?elbow sesamoid (Figs. 5, 6). Description — Only a single dental specimen clearly attributable to Adinotherium has been col- lected from Chucal: SGOPV 4064, a partial tooth, probably left i2 (Fig. 4). The tooth preserves only the base of its crown and part of the root; none of the occlusal surface is present. The root mea- sures 5.6 X 6.5 mm. Enamel is restricted primar- ily to the front of the tooth but extends slightly onto the mesial surface toward the crown. The labial enamel is much thicker than the mesial. The labial enamel forms an inverted triangle in ante- rior view, with a rounded apex and a base formed by the broken surface of the tooth. In size and morphology, the specimen closely resembles the base of left i2 of FMNH PI 3097, a specimen of Adinotherium from Santa Cruz, Argentina; it is much smaller than FMNH UC1330, a specimen of Nesodon from Santa Cruz. The tooth is asso- ciated with various mandibular fragments, includ- ing a partial right articular condyle that compares well with Adinotherium. A second craniodental specimen, SGOPV 401 1, may pertain to Adinotherium. It appears to be a partial maxilla preserving the bases of two teeth, a fragment of a third, and part of the alve- olus of a fourth. Based on the portions of teeth that are present and the size of the alveolus, the most notable aspect of SGOPV 4011 is the dra- matic mesial-to-distal increase in tooth size; this resembles the condition seen in the premolar se- ries of Adinotherium and does not closely resem- ble the condition present in any other specimens from Chucal. The fragmentary nature of the spec- imen precludes a precise identification, however. Despite the paucity of dental remains, several well-preserved postcranial specimens (primarily forelimb elements) increase confidence in the rec- ognition of Adinotherium at Chucal. Specimen SGOPV 4084 includes proximal fragments of the left ulna and radius (Figs. 5, 6). The radius is much more robust than that of the larger mesoth- eriids at Chucal (e.g., A. chucalensis, SGOPV 4042) and differs most conspicuously in having a very pronounced interosseous crest (Fig. 6). Sim- ilarly, the fragmentary ulna, while preserving only the distal portion of the sigmoid notch, differs from the corresponding bone of mesotheriids in lacking a well-demarcated fossa distal to the cor- onoid process (Fig 5.). Except for its much small- CROFT ET AL.: CHUCAL UNGULATES ^^v^^l f r "^ ■ ^^^m- T ^-^JS- 1 ^rm ..uIlxka^^^^^^^^H Wk te^^^V^IS. ^|d[^H r ' !^'^^-^ 1 i 1^1 Fig. 5. Relative sizes of mesotheriines and toxodontids from Chucal, as illustrated by the size of the ulna. A. SGOPV 4044, proximal left ulna fragment of Nesodon imbricatus. B. SGOPV 4084, proximal portion of left ulna of Adinotherium. C. SGOPV 4042, left ulna of A. chucalensis (Mesotheriinae, described in text). All specimens are in medial view. Scale bars — 2 cm. er size, it resembles a partial ulna from Chucal referred to Nesodon (SGOPV 4044). A small, rounded bone with two closely appressed facets found near the proximal ulna and radius may rep- resent an elbow sesamoid bone of Adinotherium (as noted for specimens of similar morphology by Scott, 1912). The other specimens attributed to Adinotherium also resemble the corresponding el- 10 FIELDIANA: GEOLOGY Fig. 6. Morphology of the proximal radius in mesotheriines and Adinotherium. A. FMNH PI 4482, proximal left radius of "Typotheriopsis sp." from the late Miocene Araucano Fm. of Catamarca, Argentina. B. SGOPV 4042, proximal right radius (pictured as left) of A. chucalensis (Mesotheriinae, described in text). C. SGOPV 4084, proximal left radius of Adinotherium. All specimens are in medial view. Scale bars - 2 cm. ements of Nesodon while differing in morphology from those of mesotheriids (and other notoungu- lates). Comments — As discussed below (Litoptema) appendicular bones of the major South American ungulate clades are easily distinguished (Cifelli, 1983, 1985, 1993). Those from Chucal here re- ferred to the Toxodontidae (viz., Nesodon and Ad- inotherium) are similar in morphology but differ markedly in size. This difference is well illustrat- ed by SGOPV 4044, a partial left proximal ulna of Nesodon, versus SGOPV 4084, a partial left proximal ulna oi Adinotherium (Fig. 5); the depth of the ulna just distal to the trochlear notch of the former is approximately 8 cm, while the corre- sponding measurement of the latter is less than 3 cm. The specimens attributed to Adinotherium above are all fully ossified and therefore do not pertain to immature individuals of Nesodon. Typotheria Zittel, 1892 Mesotheriidae Alston, 1876 Mesotheriinae Simpson, 1945 Mesotheriids are traditionally divided into two subgroups: the Trachytheriinae (comprising De- seadan and earlier forms, most assigned to Trach- ytherus) and the Mesotheriinae (including Friasian s.l. and later forms) characterized by the loss of I2-P2/i3-p3 and the presence of persistently tri- lobed upper molars, among other features (Fran- cis, 1965; Villarroel, 1974a; Cerdeiio &. Montal- vo, 2001; Reguero & Castro, 2002; Croft, Flynn, & Wyss, 2003; Flynn et al., in press). Mesothe- riids are absent from faunas of Colhuehuapian age and had not been recorded in faunas of Santacru- cian age prior to their discovery at Chucal (Flynn, Croft, et al., 2002). Mesotheriines are in great need of systematic revision (Pascual & Bondesio, 1985; Cerdeiio & Montalvo, 2001; Flynn, Croft, et al., 2002; Croft, Flynn, & Wyss, 2003; Flynn et al., in press), mak- ing taxonomic assignments of newly discovered forms challenging. Based on Francis's (1965) re- vision of Argentine taxa and the subsequent de- scription of additional forms from Bolivia (Vil- larroel, 1974a,b), Flynn et al. (in press) recog- nized seven potentially valid mesotheriine "gen- era" and described a new taxon from the middle Miocene Huaylas Formation of northern Chile. A character that has traditionally been impor- tant in mesotheriine taxonomy is imbrication: the degree to which the anteroextemal comer of an upper cheek tooth overlaps the posteroexternal CROFT ET AL.: CHUCAL UNGULATES 11 Table 2. Measurements for specimens of Altitypotherium paucidens. 11 9 P4 Ml M2 M3 Upper dentition L W L W L W L W L W SGOPV 4038 (R) 6.9 7.7 (12.3) 8.5 SGOPV 4038 (L) (14.4) (6.7) (6.9) (7.6) 12.5 8.6 12.8 8.8 13.9 8.4 SGOPV 4024 (L) 14.0 6.8 SGOPV 5041 (L) 7.0 8.3 SGOPV 5072 (L) 14.1 8.3 SGOPV 5102 (R) (12.5) (8.6) il i2 p4 ml m2 ni3 Lower dentition L W L W L W L W L W L W SGOPV 4024 (R) 8.1 5.1 (10.5) 6.4 — (6.4) 16.6 5.6 SGOPV 4024 (L) — — — — 8.3 5.6 10.7 6.5 13.0 6.6 16.2 5.8 SGOPV 5046 (R) — — (11.3) 6.6 16.1 6.4 SGOPV 5046 (L) 7.3 5.0 (10.1) (6.7) — — 16.1 6.3 SGOPV 4101 (R) 7.0 4.9 5.7 3.8 8.4 4.9 10.5 6.3 12.0 6.5 15.5 5.9 SGOPV 4101 (L) 7.1 4.9 5.0 3.8 SGOPV 5142 (R) 7.1 4.7 SGOPV 5088 (R) 5.5 4.0 SGOPV 5045 (L) 6.0 4.0 ' SGOPV 5125 (L) 7.5 5.3 SGOPV 5091 (R) 10.8 6.3 SGOPV 5051 (L) (11.6) (7.0) (15.9) (5.9) corner of the tooth preceding it (Francis, 1965). In his review of mesotheriine evolution, Villarroel (1974a) noted that imbrication includes two com- ponents: "chevauchement," the degree of mesio- distal overlap between teeth, and "debordement," the amount of labiolingual overlap. He stated that the former shows little variation through the SAL- MA sequence and that the latter is more useful for distinguishing among mesotheriine species. Following Villarroel's observation, we measure imbrication ("debordement") as the width from M2 parastyle to Ml hypocone divided by the width of Ml posterior face (see also Flynn et al., in press). Mesotheriine remains are among the most com- mon fossils in the Chucal Formation; approxi- mately one-third of identified specimens are re- ferable to this group. Mesotheriids are also abun- dant in most Tertiary mammal faunas from Boliv- ia (e.g., Salla, Achiri, Nazareno), suggesting that intermediate latitudes (and possibly higher eleva- tions) may have been a center of diversification for the group (Croft, Flynn, & Wyss, 2003). The mesotheriines from Chucal clearly pertain to at least three distinct species. A phylogenetic anal- ysis of currently recognized mesotheriines and the Chucal taxa (discussed below) suggests that all three species from Chucal represent new, early di- verging mesotheriines. Two of these appear to be mutual nearest relatives, and thus share a common anterior half of the binomial to reflect this provi- sional hypothesis of relationship. Altitypotherium paucidens, species novum (Figures 7, 8, Table 2) HoLOTYPE — SGOPV 4038, partial rostrum with left II, left P4-M3 and right P4-M1 (Fig. 7). Hypodigm— SGOPV 5041, left ?P4; SGOPV 5072, left M3; SGOPV 5102, partial right M3; SGOPV 4024, partial left and right mandibles and isolated left II (left mandible has partial il-2 and complete p4-m3, right mandible has partial il-2, ml-2, and complete p4, m3; Fig. 8 A); SGOPV 5046, partial left and right mandibles and frag- mentary incisors (left mandible bears complete p4-ml, m3, and partial m2, right has base of ml, complete m2-3); SGOPV 4101, partial right man- dible and symphysis with full dentition plus left il-2 (Fig. 8B); SGOPV 5142 right il; SGOPV 5045, left i2; SGOPV 5088, right il; SGOPV 5125, left p4; SGOPV 5091, right ml; SGOPV 5051, partial left mandible with m2-3. Type Locality— Loc. C-ALT-98-12B; Chucal Formation (Member W4 of Charrier et al., 2002); west flank of Chucal Anticline. Age and Distribution — Chucal Formation, 12 FIELDIANA: GEOLOGY Fig. 7. Holotype of Altitypotherium paucidens, SGOPV 4038, a partial rostrum with left II. left P4-M3 and right P4-M 1 . A. Left lateral (above, pictured as right) and occlusal (below) views. B. Line drawings of occlusal morphology (toothrows not in natural position). Scale bars = I cm. CROFT ET AL.: CHUCAL UNGULATES 13 Fig. 8. Lower dentitions referred to Altitypotherium paucidens. A. SGOPV 4024, partial left mandible with fragmentary il-2 and complete p4-m3 in occlusal view. B. SGOPV 4101, partial right mandible and symphysis with full dentition plus left il-2 in lateral (above) and occlusal (below) views. Scale bars = 1 cm. northern Chile, late early Miocene age, Santacru- cian SALMA. Etymology — Aid, in reference to the Altipla- no, the high plateau of northern Chile and neigh- boring Bolivia and Peru, where the referred spec- imens were collected, and typotherium, in refer- ence to the familiar root commonly used in me- sotheriine taxon names; pauci, Latin for "few," and dens, Latin for "tooth," in reference to the lack of P3, a characteristic feature of this species. Diagnosis — A mesotheriine that differs from all other mesotheriines in the absence of P3. Differs from Microtypotherium in presence of deep, narrow postpalatal notch that extends to the level of M3 (does not reach M3 in Microtypo- therium); presence of subtriangular II that is 14 FIELDIANA: GEOLOGY pointed distally, with two lingual sulci (a single mesial sulcus is present in Microtypotherium); presence of enlarged M3 middle lobe (not sur- rounded by other two lobes); presence of one la- bial sulcus on premolar ectolophs (two sulci pre- sent in Microtypothehum); and absence of short- ened m2 (length/width between 1.60 and 2.30 in Altitypotherium, <1.6 in Microtypothehum). Differs from Eutypotherium in the root of zy- gomatic arch being directed perpendicular to Ml (posteriorly directed from the level of M 1 or M2 in Eutypotherium); presence of deep, narrow post- palatal notch that extends to the level of M3 (wide and shallow in Eutypotherium); presence of sub- triangular II that is pointed distally, with two lin- gual sulci (wide and rounded with lingual sulcus in Eutypotherium); presence of rounded il with little or no lingual sulcus (subtriangular with smooth sulcus in Eutypotherium); presence of one labial sulcus on premolar ectolophs (two sulci present in Eutypotherium); absence of extensive lingual exposure of Ml middle lobe; and smaller size (based on dental measurements, approximate- ly 20% smaller than Eutypotherium tehmannnit- schei). Differs from Plesiotypotherium in posteriorly diverging upper and lower diastemata (parallel in Plesiotypotherium); absence of well-developed suborbital fossa; presence of deep, narrow post- palatal notch that extends to the level of M3 (wide and shallow in Plesiotypotherium); subtriangular II that is pointed distally, with two lingual sulci (wide and rounded with lingual sulcus in Plesi- otypotherium); presence of rounded i 1 with little or no lingual sulcus (subtriangular with smooth sulcus in Plesiotypotherium); presence of one la- bial sulcus on premolar ectolophs (two sulci pres- ent in Plesiotypotherium); presence of Ml middle lobe with lingually convergent anterior and pos- terior sides (subparallel in Plesiotypotherium); ab- sence of extensive lingual exposure of M 1 middle lobe; and smaller size (based on dental measure- ments, approximately 30% smaller than Plesioty- potherium achirense). Differs from Caragua New Taxon (Flynn et al., in press) in presence of posteriorly divergent up- per and lower diastemata (parallel in Caragua New Taxon); lesser degree of molar imbrication (<1.25 m Altitypotherium, between 1.25 and 1.50 in Caragua New Taxon); presence of subtriangular II that is pointed distally, with two lingual sulci (wide and rounded with lingual sulcus in Caragua New Taxon); presence of short, subtriangular P4 without central fossette (P4 bilobed with short and poorly defined lingual plication in Caragua New Taxon); presence of enlarged M3 middle lobe (not surrounded by other two lobes); relatively smaller il (il/i2 <1.50 in Altitypotherium, between 1.50 and 2.00 in Caragua New Taxon); presence of rounded il with little or no lingual sulcus (sub- triangular with smooth sulcus in Caragua New Taxon); presence of one labial sulcus on premolar ectolophs (two sulci present in Caragua New Tax- on); presence of Ml middle lobe with lingually convergent anterior and posterior sides (subpar- allel in Caragua New Taxon); absence of exten- sive lingual exposure of Ml middle lobe; and smaller size (based on dental measurements, ap- proximately 20% smaller than Caragua New Tax- on). Differs from Typotheriopsis in presence of pos- terolaterally divergent upper and lower diastemata (parallel in Typotheriopsis); absence of well-de- veloped suborbital fossa; presence of deep, nar- row postpalatal notch that extends to the level of M3 (does not reach M3 in Typotheriopsis); pres- ence of subtriangular II that is pointed distally, with two lingual sulci (wide and rounded with lin- gual sulcus in Typotheriopsis); presence of en- larged M3 middle lobe (not surrounded by other two lobes); relatively smaller il (il/i2 <1.50 in Altitypotherium, between 1.50 and 3.00 in 7\'- potheriopsis); presence of rounded il with little or no lingual sulcus (subtriangular with smooth sulcus in Typotheriopsis); presence of M 1 middle lobe with lingually convergent anterior and pos- terior sides (subparallel in Typotheriopsis); ab- sence of extensive lingual exposure of Ml middle lobe; and smaller size (based on dental measure- ments, approximately 30% smaller than Typo- theriopsis chasicoensis). Differs from Pseudotypotherium in presence of posterolaterally divergent upper and lower diaste- mata (parallel in Pseudotypotherium); absence of well-developed suborbital fossa; presence of deep, narrow postpalatal notch that extends to the level of M3 (does not reach M3 in Pseudotypotherium); lesser degree of molar imbrication (< 1 .25 in Al- titypotherium, >1.50 in Pseudotypotherium); presence of short, subtriangular P4 without central fossette (P4 is bilobed in Pseudotypotherium); rel- atively smaller il (il/i2 <1.50 in Altitypotherium, between 2.00 and 3.00 in Pseudotypotherium); presence of rounded il with little or no lingual sulcus (trapezoidal with well-demarcated lingual sulcus in Pseudotypotherium); presence of one la- bial sulcus on premolar ectolophs (two sulci pre- sent in Pseudotypotherium); absence of elongate CROFT ET AL.: CHUCAL UNGULATES 15 P4 (length/width < 1 .50 in Altitypotherium, > 1 .50 in Pseudotypotherium); absence of elongate Ml (length/width <1.75 m Altitypotherium, ^1.75 in Pseudotypotherium); presence of Ml middle lobe with lingually convergent anterior and posterior sides (subparallel in Pseudotypotherium); absence of extensive lingual exposure of Ml middle lobe; absence of elongate m2 (length/width between 1.60 and 2.30 in Altitypotherium, >2.30 in Pseu- dotypotherium); and smaller size (based on dental measurements, approximately is 25% smaller than Pseudotypotherium exiguum). Differs from Mesotherium in posterolaterally divergent upper and lower diastemata (parallel to strongly convergent in Mesotherium); presence of poorly defined suborbital fossa (absent in Meso- therium); lesser degree of molar imbrication (<1.25 in Altitypotherium, >1.50 in Mesothe- rium); presence of subtriangular II that is pointed distally, with two lingual sulci (wide and rounded with lingual sulcus or wide and compressed la- biolingually with two lingual sulci in Mesothe- rium); presence of short, subtriangular P4 without central fossette (P4 is persistently bilobed in Me- sotherium); relatively smaller il (il/i2 <1.50 in Altitypotherium, >3.00 in Mesotherium); presence of rounded il with little or no lingual sulcus (trap- ezoidal with well-demarcated lingual sulcus or el- liptical with smooth sulcus in Mesotherium); pres- ence of one labial sulcus on premolar ectolophs (two sulci present in Mesotherium); absence of elongate P4 (length/width <1.50 in Altitypothe- rium, > 1 .50 in Mesotherium); absence of elongate Ml (length/width <1.75 in Altitypotherium, >1.75 in Mesotherium); presence of Ml middle lobe with lingually convergent anterior and pos- terior sides (lingually divergent in Mesotherium); absence of extensive lingual exposure of Ml mid- dle lobe; absence of elongate m2 (length/width between 1 .60 and 2.30 in Altitypotherium, >2.30 in Mesotherium); and smaller size (based on den- tal measurements, approximately 20% smaller than Mesotherium maendrum and half the size of Mesotherium pachygnathum). Differs from A. chucalensis (described below) in absence of P3; presence of elongate Ml (length/width >1.3 in A. paucidens, <1.3 in A. chucalensis); middle lobe of M3 unreduced, not surrounded by anterior and posterior lobes; pres- ence of narrower lower molars (widths <6.7 mm in A. paucidens, >:7.3 mm in A. chucalensis); presence of elongate m3 ( 1 5-20% longer than m2 in A. paucidens, similar in length to m2 in A. chu- calensis); and slightly larger size (based on dental measurements, A. paucidens is approximately 10% larger than A. chucalensis). Description — The holotype of A. paucidens consists of a moderately well-preserved rostrum (Fig. 7). The nasals have been crushed, and much of the premaxilla is missing, but the left side pre- serves the zygomatic arch, palate, and all the up- per teeth. The inferior surface of the zygomatic arch is slightly damaged but preserves a well-developed surface for the origin of the masseter. A shallow suborbital fossa is present on the dorsal surface of the arch. This fossa is similar in its degree of development to that of the holotype of Eutypo- therium superans (MACN 11079), though it is oriented differently; in A. paucidens, the primary axis of the fossa is anteroposterior, whereas in E. superans it is transverse. In mesotheriines in which the suborbital fossa is developed to an even greater degree, it is also oriented transversely (e.g., Typotheriopsis silveyrai, MLP 36-XI-10-2). The suborbital fossa is demarcated posteriorly by a thick strut of bone oriented perpendicular to the toothrow. A large, elliptical infraorbital foramen is present, its major axis oriented dorsoventrally; it measures 10.5 X 5.5 mm. Very little of right II is preserved in SGOPV 4038, but left II is well preserved except for its occlusal surface; in cross section, the major axis of the tooth is set at an angle of approximately 45° to the sagittal plane. Much of the premaxilla is missing from the left side of the specimen, re- vealing nearly the entire dorsoventral extent of 1 1 . It is a strongly curved tooth (approximating a 90° arc) and lacks any evidence of a root. The entire surface of the tooth is covered by enamel, with the exception of the damaged occlusal surface. An isolated 1 1 referred to this species (SGOPV 4024) better demonstrates the occlusal morphology of the upper incisor; it is roughly subtriangular in cross section and pointed distally, with two smooth sulci present on the lingual face. The oc- clusal surface is worn such that it is strongly con- cave. A large incisive foramen is present on the an- terior surface of the palate just posterior to the pair of incisors. It measures approximately 9.5 X 7.3 mm and is divided longitudinally by a septum that is recessed relative to the surface of the pal- ate. Large (nearly 25 mm long), posteriorly di- vergent diastemata separate the pair of incisors from the cheek teeth. The palate is relatively flat anterior to the cheek teeth but is highly arched between them. 16 FIELDIANA: GEOLOGY The lone upper premolar (P4) is subtriangular in outline and has a single sulcus located near the anterior edge of its labial surface. The occlusal surface is featureless, and there is no evidence of a lingual groove or plication, as occurs in many mesotheriines. The shallower implantation angle of this tooth is reflected in the anteroposteriorly elongate wear on the occlusal surface. The three upper molars are of typical mesotheriine form. Each is rectangular with a pronounced parastyle and three lobes extending lingually from the ec- toloph (formed by two labially directed invagi- nations of enamel from the lingual margins of the teeth). The first upper molar in A. paucidens is proportionately more elongate than the corre- sponding tooth in A. chucalensis, and the three lobes increase in size posteriorly. The anterior lobe is directed posterolingually, while the pos- terior lobe is roughly perpendicular to the ecto- loph; the middle lobe therefore approximates a right triangle, with the hypotenuse forming its an- terior face. The presence of a strongly triangular Ml middle lobe is unique among described me- sotheriines, with the exception of Microtypothe- rium choquecotense (Villarroel, 1974b) and the other taxa from Chucal. However, one of the spec- imens from Nazareno, Bolivia (GB Naz-014) fig- ured by Oiso (1991, p. 658, plate 2) and assigned to ''' Plesiotypotherium sp." also appears to pos- sess this feature. Other mesotheriine specimens (e.g. GB Naz-001, GB Naz-013, GB Naz-047) from that site figured by Oiso (1991) lack a tri- angular Ml middle lobe. It is unknown whether this represents individual or interspecific variation among the specimens from Nazareno. Based on a phylogenetic analysis of mesotheriines (below), the presence of a triangular Ml middle lobe is judged ancestral for the group. The posterior lobe of Ml bears a slight lingual plication in the posterior end of its lingual face. This plication is increasingly prominent on M2 and M3. A similar condition is present in the ho- lotype of Plesiotypotherium achirense (although not noted in the original description; Villarroel 1974a, p. 267, fig. 9), in the holotype of Pseu- dotypotherium carhuense (MLP 37-III-7-1), and also in a juvenile specimen of Trachytherus spe- gazzinianus (MACN 3264) figured by Patterson (1934b, fig. 24). The increased prominence of this feature in more posterior cheek teeth, coupled with its absence in a more mature specimen of Trachytherus (e.g., FMNH PI 3281), suggests that this feature may be widespread among meso- theriids but is lost with wear. The second upper molar of A. paucidens differs from Ml in having a slightly more rectangular middle lobe, although the anterior and posterior faces still converge lingually. The last upper mo- lar has a smaller anterior lobe, similar to that seen in Ml, and a small, anterolingually directed pos- terior lobe. The middle lobe therefore has sub- equal anterior and posterior faces. The enamel surface of this lobe does not reach the lingual face of the tooth but rather has a thick covering of cementum that fills the portion of the lingual face between the anterior and posterior lobes. The mid- dle lobe of M3 is proportionately larger than that of A. chucalensis, and thus (unlike in A. chuca- lensis) this lobe remains unenclosed by the ante- rior and posterior lobes. A small but distinct in- vagination of enamel occurs on the lingual face of the posterior lobe; greater wear would presum- ably obliterate this structure. The posterior face of M3 bears a slight indentation near its center, iso- lating a posteriorly directed extension of the ec- toloph known as the "fourth lobe" (Francis, 1965). The degree of cheek tooth imbrication in A. paucidens is 1.16. The enamel is thinnest along the anterior and posterior faces of the cheek teeth, causing the occlusal surfaces of these teeth to join with wear. Almost no enamel is visible on the posterior face of M3. The cheek teeth are covered by moderate amounts of cementum. This cemen- tum fills the lingual invaginations of enamel on the molars (that demarcate the three lobes) and is especially thick along the lingual face of M3 (much thicker than the enamel of the middle lobe). The hypselodont nature of the cheek teeth is visible in left M3 and right M 1 of the holotype. The dimensions of M 1 of the holotype of A. paucidens are similar to those of Microtypothe- rium choquecotense (see below), the smallest me- sotheriine known until now (Villarroel, 1974b). The second and third upper molars are signifi- cantly shorter \x\ A. paucidens than in Microty- potherium, resulting in a shorter molar row over- all. In size, A. paucidens closely matches most specimens assigned to "" Plesiotypotherium sp." by Oiso (1991). Several partial lower dentitions are referred to A. paucidens based on their relatively longer and narrower cheek teeth (as compared to the holo- type of y4. chucalensis, SGOPV 4100, an associ- ated skull and mandibles). The best preserved of these is SGOPV 4101, a partial mandible and symphysis with complete right dentition plus left il-2 (Fig. 8B). The symphysis, as in other me- CROFT ET AL.: CHUCAL UNGULATES 17 sotheriines, is solid and completely fused; there is no evidence of a suture. Its labial surface is smoothly rounded, joining the external surfaces of the mandibles. The lingual (dorsal) surface is strongly excavated, forming a long, concave spout flanked by the second lower incisors. This spout appears to be proportionately narrower and deeper than that of E. lehmannnitschei. The right dentary is approximately 27 mm high at the level of p4 and 35 mm high at the posterior end of m3. A wide, shallow, slightly roughened groove runs along the lingual surface of the right dentary. At the posterior margin of m3, this groove is posi- tioned about midway between the superior and in- ferior surfaces of the ramus; its anterior end lies near the base of the horizontal ramus at the level of p4. No mandibular foramen is observed, likely because of incomplete preservation of the speci- men. A mental foramen occurs on the external surface of the right dentary, just anterior to p4, at a level slightly higher than the dorsoventral mid- point of the horizontal ramus. It lies almost di- rectly lateral to the base of 12. The first lower incisor has a gently convex la- bial face and more strongly convex lingual face. Enamel covers the external surface of the tooth but is completely absent from the occlusal sur- face. The occlusal surface is gently concave, with the points of highest relief being the two labial corners of the tooth. In mesial view, the tooth is gently curved and procumbent, nearly aligned with the anterior surface of the mandibular sym- physis. The second lower incisor is significantly smaller than the first, though not as proportion- ately reduced as in later-diverging mesotheriines. It is peglike in overall form. While the long axis of il runs nearly perpendicular to the sagittal plane, that of i2 is parasagittal. The most pro- nounced wear occurs on a surface approximately perpendicular to the tooth's vertical axis. Addi- tional wear occurs on the posterior face of the tooth. The left 12 is visible in its entirety; it is approximately 25 mm long and has an open base. A small amount of cementum covers portions of all the incisors. The diastema separating 12 from p4 on the right side is incompletely preserved but was likely approximately 15 mm long. The lower cheek teeth resemble those of other mesotheriines. The sole lower premolar has a small trigonid and a larger talonid. The dentine of the occlusal surface is fully confluent between the trigonid and talonid; no fossettes or sulci are pre- sent. The occlusal surface of the talonid is not worn completely perpendicular to the vertical axis of the tooth but rather slopes to a low point at the posterior end of the labial side. The first and sec- ond lower molars are roughly the shape of a figure eight; the trigonid and the talonid are both round- ed and are separated by a deep labial sulcus ori- ented slightly posterolingually. The second molar is slightly larger than the first. In both molars a very slight sulcus is present on the posterior end of the lingual face. The trigonid of m3 is similar in size and shape to that of m2, but the talonid is elongate with a pointed posterior end, typical of mesotheriines. A moderately pronounced sulcus is present on the lingual face of the talonid at ap- proximately the midpoint, and a little-pronounced sulcus is present about halfway between the mid- point and the posterior edge of the tooth. The pos- terior end of m3 is visible in its entirety; the tooth is 31 mm high, and its "root" is completely open. Occlusal relief is highest on the molars where the deep labial sulcus divides the trigonid from the talonid (being formed of more resistant enamel) and the dentine surface of the talonid is the low- est. All cheek teeth are covered by a moderate amount of cementum. Comments — The most unusual characteristic of A. paucidens is its single upper premolar (all other mesotheriines having two). Although large cracks are present in the rostrum immediately anterior of this tooth in the holotype, evidence for the lack of an alveolus for an additional premolar is per- suasive; the specimen was recovered as three sep- arate pieces (subsequently glued together), per- mitting a thorough examination of the region im- mediately anterior of the preserved cheek teeth. The absence of a second premolar is clearly not a preservational artifact. The single premolar in A. paucidens is assumed to be P4, but the possi- bility of another homology cannot be excluded. In lateral view, the premolar is not implanted parallel to the anterior face of Ml, as P4 is in other me- sotheriines, but rather approaches it at an angle of ~15°; a small portion of the maxilla (lacking an alveolus) is present between the bases of these teeth. This space suggests that the single large premolar may represent an enlarged P3 that pres- ently fills all or most of the space formerly oc- cupied by both P3 and P4 (no concomitant reduc- tion in the number of lower cheek teeth seems to have occurred). While the loss of a cheek tooth in this position would be an unusual occurrence among mammals, mesotheriines are already pe- culiar in having three deciduous premolars that are replaced by only two permanent premolars. These permanent premolars are conventionally re- 18 FIELDIANA: GEOLOGY Table 3. Development of suborbital fossa and size of infraorbital foramen in selected mesotheriines. Relative size = cross-sectional area of infraorbital foramen (estimated as an ellipse )/M 1 occlusal area (estimated as a rectangle). Holotype specimens are denoted with an asterisk (*) next to the specimen number Suborbital Infraorbital Ml size Relative Taxon Specimen fossa foramen (mm) (mm) size Alitypothehum paucidens SGOPV 4038* shallow 10.5 X 5.5 12.5 X 8.6 0.42 Eutypotherium chico SGOPV5157* shallow 4.0 X 3.5 9.5 X 6.4 0.18 Eutypotherium lehmannnitschei MLP 12-1701* shallow 7.5 X 5.4 14.6 X 10.0 0.22 Eutypotherium superans MACN 1 1079* shallow 5.5 X 4.1 14.1 X 10.0 0.13 Caragua New Taxon SGOPV 4004* shallow 6.5 X 4.9 16.5 X 10.0 0.15 Typotheriopsis silveyrai MLP 36-XI- 10-2* deep 6.6 X 5.7 16.9 X 11.4 0.15 Typotheriopsis sp. FMNH P14477 deep 7.5 X 6.2 22.5 X 14.5 0.11 IPseudotypotherium sp. MACN 8010 deep 14.2 X 12.4 23.8 X 11.0 0.53 IPseudorypotherium sp. MACN 1111 deep 12.2 X 7.7 19.2 X 10.5 0.37 Mesotherium cristatum MACN 2036 absent 3.9 X 3.0 21.8 X 9.8 0.04 Mesotherium pachygnathum MACN 1665 absent? 9.4 X 5.8 24.4 X 11.4 0.15 Mesotherium "augustirostrum" MACN 6040* absent 5.1 X 3.5 23.6 X 11.1 0.05 garded as P3-4, but in the absence of ontogenetic data with which to definitively establish homolo- gy, they could be P2-3 or even P2 and P4. Re- gardless of their identity, at some time during me- sotheriine evolution, two permanent premolars came to fill the positions previously occupied by three permanent premolars (as suggested by the presence of three deciduous precursors). Perhaps the single premolar in A. paucidens represents the end stage of this process: a single, enlarged per- manent premolar filling the position once occu- pied by three. No examples of the deciduous up- per dentition of A. paucidens are currently known to establish the actual number of deciduous pre- molars in this taxon. Previous workers have noted a close associa- tion between the development of the suborbital fossa and the size of the infraorbital foramen. Francis (1965:12) stated that "la presencia de fos- eta sub-orbitaria se asocia siempre a un agujero infra-orbitario grande; la ausencia a uno pe- queiio." He hypothesized that the presence of a gland (located in the suborbital fossa) that emp- tied its contents via the infraorbital foramen might explain the observed association between the two. In their analysis of mesotheriine relationships, Cerdeno and Montalvo (2001) included the sub- orbital fossa and infraorbital foramen as a single character (#6) with three states: fossa absent and foramen very small (0), fossa hinted at and fora- men small (1), and fossa developed and foramen large (2). A more detailed examination of distri- butions of the fossa and foramen suggests that the relationship between the development of the sub- orbital fossa and the size of the infraorbital fora- men might not be so straightforward (Table 3). Relative to body size, the infraorbital foramen in A. paucidens is among the largest seen in me- sotheriines, yet the suborbital fossa is only mod- erately developed. Conversely, Typotheriopsis has a relatively small infraorbital foramen, but it ex- hibits a well-developed suborbital fossa. A gen- eral pattern does seem to exist, however, in that the infraorbital foramen is relatively smallest in Mesotherium (which completely lacks a suborbital fossa), largest in Pseudotypotherium (which has a well-developed suborbital fossa), and intermedi- ate in most other taxa. Since Mesotherium in- cludes the largest mesotheriines, however, the rel- atively small size of its infraorbital foramen may result merely from allometric relationships. The lack of an obvious taxonomic pattern in these characters may be the result of intraspecific vari- ability in one or both of these structures and/or subjective assessment of the degree of develop- ment of the suborbital fossa (which is difficult to quantify). Since the association between changes in the infraorbital foramen and the suborbital fos- sa (character #6) is unclear, only the development of the suborbital fossa has been included in the phylogenetic analysis below. Two of the mesotheriines from Chucal, Altity- potherium paucidens and Altitypotherium chuca- lensis (below), form a clade diagnosed by the shared presence of a single labial sulcus on the premolar ectolophs (character #27, state #1) and the presence of upper first incisors that are sub- triangular, pointed distally, with two lingual sulci (character #16, state #4; see phylogenetic analysis below). The two species differ slightly in size — by about 10% based on dental measurements — but both are among the smallest mesotheriines CROFT FT AL.: CHUCAL UNGULATES 19 Table 4. Measurements for specimens of Altitypotherium chucalensis. II P3 P4 Ml M2 Upper dentition W W W W W M3 W SGOPV4100(R) (4.9) (5.9) 6.3 7.3 SGOPV41(X)(L) (4.5) (6.0) 5.9 — SGOPV 5148 (R) 12.4 6.6 SGOPV 5172 (R) (12.5) 6.8 SGOPV 5148 (L) 11.9 6.5 SGOPV 4033 (R) 5.5 5.8 6.8 7.6 SGOPV 4033 (L) — — 6.4 — SGOPV 5146 (R) 6.1 5.3 SGOPV 5147 (L) 5.4 6.1 SGOPV 5056 (R) 5.1 5.4 5.9 7.3 SGOPV 51 13 (L) 5.9 6.9 SGOPV 5098 (R) 10.4 8.6 12.9 8.5 13.1 7.6 10.4 8.9 — 8.4 13.3 7.7 — — — (8.8) — (7.3) SGOPV 5098 (R) 12.3 8.5 il i2 p4 ml m2 m3 Lower dentition L W L W L W L W L W L W SGOPV 4042 (R) 6.9 4.6 SGOPV 4042 (L) (6.8) (4.3) 5.5 3.7 SGOPV 41 10 (R) 6.7 4.3 5.0 3.4 SGOPV 41 10 (L) 6.8 4.4 5.0 3.4 SGOPV 5179 (R) 6.8 — SGOPV 5109 (R) 6.2 4.0 SGOPV 51 10 (L) 6.0 4.0 SGOPV 5178 (R) 5.2 3.4 SGOPV 5084 (R) SGOPV 4100 (R) SGOPV 4100 (L) SGOPV 4103 (R) SGOPV 5048 (R) SGOPV 5170 (L) SGOPV 5089 (R) SGOPV 5090 (R) SGOPV 5085 (L) SGOPV 5126 (R) SGOPV 5128 (R) SGOPV 5101 (L) SGOPV 5052 (L) SGOPV 5127 (L) SGOPV 4032 (R) SGOPV 5083 (R) SGOPV 51 19 (L) SGOPV 4107 (R) 7.0 7.2 4.7 4.3 10.3 6.2 10.0 6.1 (5.7) (6.5) (4.2) 10.0 5.9 10.7 5.7 6.9 4.5 9.7 6.1 10.9 6.1 13.8 5.7 7.2 4.5 9.7 6.0 11.4 6.1 (6.1) (4.0) 9.5 9.6 10.4 5.6 6.0 6.3 10.8 11.5 12.1 10.8 11.7 11.6 11.0 11.1 11.5 5.7 6.1 5.8 6.0 5.9 5.9 6.1 5.8 6.0 15.8 (13.5) 16.2 (15.5) 5.4 (5.5) 5.4 5.5 known. Because the presence of these shared, de- rived characters suggests these two species share a unique common ancestry, we elect to provide them with the same first half of a bipartite species name (praenomen, sensu Griffiths, 1976). Altitypotherium chucalensis, species novum (Figures 9-11, Table 4) HoLOTYPE— SGOPV 4100, partial skull and mandibles lacking upper incisors, portions of right P3 and left P3-4, and portions of lower incisors (Figs. 9, 10). Hypodigm— SGOPV 5148, right II; SGOPV 5172, right II; SGOPV 5143 left II; SGOPV 4033, fragmentary skull with portions of many teeth, right P3-4 well preserved, left Ml-3 par- tially preserved; SGOPV 5146, right P3; SGOPV 5147, left P3; SGOPV 5056, partial right maxilla with P3-4; SGOPV 5113, partial left P4; SGOPV 5098, partial right M2; SGOPV 4042, partial skel- eton including right mandible and symphysis (with left i2, portions of left and right il, right p4-ml and bases of m2-3) right femur, partial left 20 FIELDIANA: GEOLOGY Fig. 9. Upper dentition of holotype oi Altitypotherium chucalensis, SGOPV 4100, a partial skull with portions of all premolars and complete molars. A. Line drawing of occlusal morphology of right P3-M3. B. Skull in right lateral (above) and occlusal (below) views. Scale bars = 1 cm. innominate including acetabulum, proximal and distal portions of left femur, distal half of left tib- ia, partial ?left fibula, left ulna and radius (lacking distal portion of each), proximal portion of right radius, atlas and axis (articulated) plus various other vertebrae; SGOPV 4110, partial left man- dible and symphysis with il-2 on both sides, left p4-ml, partial left m2, and partial left Ml; SGOPV 5179, right il; SGOPV 5109 right il; SGOPV 5110, left il; SGOPV 5178, right i2; SGOPV 5084. partial right mandible with base of p4, ml -2; SGOPV 4103, partial right mandible with partial p4, ml-2; SGOPV 5048, partial right ml; SGOPV 5170, left ml; SGOPV 5089, right m2; SGOPV 5090, right m2; SGOPV 5085, par- tial left mandible with base of m2; SGOPV 5126, partial right m2; SGOPV 5128, partial right m2; SGOPV 5101, left m2; SGOPV 5052, left m2; CROFT ET AL.: CHUCAL UNGULATES 21 Fig. 10. Portions of lower dentition of Altitypotherium chucalensis, SGOPV 4100. A. Posterior left mandible and ascending ramus in lateral view (pictured as right). B. Partial right mandible with p4-m3 in occlusal view (pictured as left). Scale bars = 1 cm. SGOPV 5127, partial left m2; SGOPV 4032, right m3; SGOPV 5083, partial right mandible with base of m3; SGOPV 51 19, left m3; SGOPV 4107, partial right m3. Questionably Referred Specimens — SGOPV 5138, partial right p4; SGOPV 5173, partial right p4. Type Locality— Loc. C-ALT-01-19; Chucal Formation (Member W4 of Charrier et al., 2002); west flank of Chucal Anticline. Age and Distribution — Chucal Formation, northern Chile, late early Miocene age, Santacru- cian SALMA. Etymology — Alti, in reference to the Altipla- no, the high plateau of northern Chile and neigh- boring Bolivia and Peru, where the referred spec- imens were collected and typotherium, in refer- ence to the familiar root commonly used in me- sotheriine taxon names; chucalensis, in reference to Chucal, the area in which the holotype and all referred specimens were discovered. Diagnosis — A mesotheriine differing from all other described mesotheriines in having an un- elongate Ml (length/width <1.30 in A. chucalen- sis, >1.30 in all other mesotheriines). Differs from Microtypotherium in presence of deep, narrow postpalatal notch that extends to the level of M3 (does not reach M3 in Microtypo- therium); presence of subtriangular II that is pointed distally, with two lingual sulci (a single mesial sulcus is present in Microtypotherium); presence of one labial sulcus on premolar ecto- 22 FIELDIANA: GEOLOGY lophs (two sulci present in Microtypotherium); and absence of shortened m2 (length/width be- tween 1.60 and 2.30 in Altitypotherium, <1.6 in Microtypotherium). Differs from Eutypotherium in the root of zy- gomatic arch being directed perpendicular to Ml (posteriorly directed from the level of M 1 or M2 in Eutypotherium); presence of deep, narrow post- palatal notch that extends to the level of M3 (wide and shallow in Eutypotherium); presence of sub- triangular 1 1 that is pointed distally, with two lin- gual sulci (wide and rounded with lingual sulcus in Eutypotherium); presence of rounded il with little or no lingual sulcus (subtriangular with smooth sulcus in Eutypotherium); presence of one labial sulcus on premolar ectolophs (two sulci present in Eutypotherium); absence of extensive lingual exposure of Ml middle lobe; and smaller size (based on dental measurements, approximate- ly 20% smaller than Eutypotherium lehmannnit- schei). Differs from Plesiotypotherium in posteriorly diverging upper and lower diastemata (parallel in Plesiotypotherium); absence of well-developed suborbital fossa; presence of deep, narrow post- palatal notch that extends to the level of M3 (wide and shallow in Plesiotypotherium); subtriangular II that is pointed distally, with two lingual sulci (wide and rounded with lingual sulcus in Plesi- otypotherium); absence of enlarged middle lobe of M3 (surrounded by other lobes in A. chucalensis); presence of rounded il with little or no lingual sulcus (subtriangular with smooth sulcus in Ple- siotypotherium); presence of one labial sulcus on premolar ectolophs (two sulci present in Plesioty- potherium); presence of M 1 middle lobe with lin- gually convergent anterior and posterior sides (subparallel in Plesiotypotherium); absence of ex- tensive lingual exposure of Ml middle lobe; and smaller size (based on dental measurements, 30% smaller than Plesiotypotherium achirense). Differs from Caragua New Taxon (Flynn et al., in press) in presence of posteriorly divergent up- per and lower diastemata (parallel in Caragua New Taxon); lesser degree of molar imbrication (<1.25 m Altitypotherium, between 1.25 and 1.50 in Caragua New Taxon); presence of subtriangular II that is pointed distally, with two lingual sulci (wide and rounded with lingual sulcus in Caragua New Taxon); presence of short, subtriangular P4 without central fossette (P4 bilobed with short and poorly defined lingual plication in Caragua New Taxon); relatively smaller il (il/i2 <1.50 in Alti- typotherium, between 1.50 and 2.00 in Caragua New Taxon); presence of rounded il with little or no lingual sulcus (subtriangular with smooth sul- cus in Caragua New Taxon); presence of one la- bial sulcus on premolar ectolophs (two sulci pre- sent in Caragua New Taxon); presence of Ml middle lobe with lingually convergent anterior and posterior sides (subparallel in Caragua New Taxon); absence of extensive lingual exposure of M 1 middle lobe; and smaller size (based on dental measurements, approximately 20% smaller than Caragua New Taxon). Differs from Typotheriopsis in posterolateral ly divergent upper and lower diastemata (parallel in Typotheriopsis); absence of well-developed sub- orbital fossa; presence of deep, narrow postpalatal notch that extends to the level of M3 (does not reach M3 in Typotheriopsis); presence of sub- triangular II that is pointed distally, with two lin- gual sulci (wide and rounded with lingual sulcus in Typotheriopsis); relatively smaller il (il/i2 <1.50 in Altitypotherium, between 1.50 and 3.00 in Typotheriopsis); presence of rounded il with little or no lingual sulcus (subtriangular with smooth sulcus in Typotheriopsis); presence of M 1 middle lobe with lingually convergent anterior and posterior sides (subparallel in Typotheriop- sis); absence of extensive lingual exposure of Ml middle lobe; and smaller size (based on dental measurements, approximately 30% smaller than Typotheriopsis chasicoensis). Differs from Pseudotypotherium in presence of posterolaterally divergent upper and lower diaste- mata (parallel in Pseudotypotherium); absence of well-developed suborbital fossa; presence of deep, narrow postpalatal notch that extends to the level of M3 (does not reach M3 in Pseudotypotherium); lesser degree of molar imbrication (<1.25 in Al- titypotherium, >1.50 in Pseudotypotherium); presence of short, subtriangular P4 without central fossette (P4 is bilobed in Pseudotypotherium); rel- atively smaller il (il/i2 <1.50 in Altitypotherium, between 2.00 and 3.00 in Pseudotypotherium); presence of rounded il with little or no lingual sulcus (trapezoidal with well demarcated lingual sulcus in Pseudotypotherium); presence of one la- bial sulcus on premolar ectolophs (two sulci pre- sent in Pseudotypotherium); absence of elongate P4 (length/width <1.50 \n Altitypotherium, >1.50 in Pseudotypotherium); absence of elongate Ml (length/width <1.75 in Altitypotherium, >1.75 in Pseudotypotherium); presence of M 1 middle lobe with lingually convergent anterior and posterior sides (subparallel in Pseudotypotherium); absence of extensive lingual exposure of Ml middle lobe; CROFT ET AL.: CHUCAL UNGULATES 23 absence of elongate m2 (length/width between 1 .60 and 2.30 in Altitypotherium, >2.30 in Pseu- dotypotherium); and smaller size (based on dental measurements, approximately 25% smaller than Pseudotypotherium exiguum). Differs from Mesotherium in posterolaterally divergent upper and lower diastemata (parallel to strongly convergent in Mesotherium); presence of poorly defined suborbital fossa (absent in Meso- therium); lesser degree of molar imbrication (<1.25 in Altitypotherium, >1.50 in Mesotheri- um); presence of subtriangular II that is pointed distally, with two lingual sulci (wide and rounded with lingual sulcus or wide and compressed la- biolingually with two lingual sulci in Mesotheri- um); presence of short, subtriangular P4 without central fossette (P4 is persistently bilobed in Me- sotherium); absence of enlarged middle lobe of M3 (surrounded by other lobes in A. chucalensis); relatively smaller il (il/i2 <1.50 in Altitypothe- rium, >3.00 in Mesotherium); presence of round- ed il with little or no lingual sulcus (trapezoidal with well-demarcated lingual sulcus or elliptical with smooth sulcus in Mesotherium); presence of one labial sulcus on premolar ectolophs (two sulci present in Mesotherium); absence of elongate P4 (length/width <1.50 in Altitypotherium, >1.50 in Mesotherium); absence of elongate Ml (length/ width <1.75 in Altitypotherium, >1.75 in Meso- therium); presence of Ml middle lobe with lin- gual ly convergent anterior and posterior sides (lingually divergent in Mesotherium); absence of extensive lingual exposure of M 1 middle lobe; ab- sence of elongate m2 (length/width between 1 .60 and 2.30 in Altitypotherium, >2.30 in Mesotheri- um); and smaller size (based on dental measure- ments, approximately 20% smaller than Mesothe- rium maendrum and about half the size of Meso- therium pachygnathum). Differs from A. paucidens in presence of P3, absence of elongate Ml (length/width <1.3 in A. chucalensis, > 1 .3 in A. paucidens), presence of M3 with middle lobe surrounded by other two lobes, presence of wider lower molars (widths ^7.3 mm in A. chucalensis; ^6.7 mm in A. pau- cidens), absence of elongate m3 (similar in length to m2 in A. chucalensis, 15-20% longer than m2 in A. paucidens), and slightly smaller size (based on dental measurements, A. chucalensis is ap- proximately 10% smaller than A. paucidens). Description — A. chucalensis is the only me- sotheriine species from Chucal for which the ho- lotype includes associated upper and lower den- titions (SGOPV 4100, Figs. 9, 10). Additionally, much of the palatal and basicranial anatomy of this taxon is preserved. The palate is arched, but not to the degree exhibited by A. paucidens. The postpalatal notches are narrow and deep, extend- ing to the posterior edge of M3. The notch is fully prepared on the right side, demonstrating a round- ed morphology more resembling a foramen than a notch. The general structure of the ear region of mesotheriines has been described elsewhere (Pat- terson, 1934a, 1936) and will not be repeated here, although this specimen should prove useful for future comparative studies within the Meso- theriidae and other typotheres. The most obvious difference in the dentition between this taxon and A. paucidens is the pres- ence of two premolars in A. chucalensis (inter- preted as P3 and P4, but see discussion above). This is a primitive character state that A. chuca- lensis shares with all previously described me- sotheriines. The possibility that these teeth rep- resent deciduous premolars instead of permanent ones can be ruled out; the labial faces of the two teeth on the holotype are exposed to their bases (approximately 25 mm in P4), confirming their hypselodont condition and the lack of overlying replacement teeth above. They also do not dem- onstrate the occlusal morphology typical of me- sotheriine deciduous premolars teeth (see Francis, 1965), instead exhibiting the typically adult mor- phology of featureless occlusal surfaces that are subtriangular in shape. In the holotype and one of the referred speci- mens (SGOPV 4033, a fragmentary palate), P3 and P4 are closely appressed to each other and to Ml. In another specimen (SGOPV 5056, a partial right maxilla with P3-4), there is a small (~2 mm) diastema between them, although both are implanted at approximately the same angle. A sin- gle sulcus is present near the posterior end of the labial face of P3, whereas P4 exhibits a sulcus situated near the anterior end of the labial face as in A. paucidens (and Typotheriopsis). The P3 is smaller than P4, and neither exhibits any sign of a fossette or lingual plication. In SGOPV 5056, the entire posterior face of P4 is visible, demon- strating the near total lack of enamel on this sur- face. On the anterior face of this tooth, enamel is absent only from a small strip along the labial edge. The same is true for P3. A. chucalensis is further distinguished from A. paucidens by the absence of an elongate Ml, an ancestral character state shared only with Trach- ytherus among mesotheriids. Unlike Trachyther- us, however, the middle lobe of Ml in A. chuca- 24 FIELDIANA: GEOLOGY lensis appears to persist throughout the tooth's wear and does not become isolated as a fossette (based on examinations of Hngual faces and bases of isolated teeth). Like Microtypotherium and the other Chucal mesotheriines, the anterior and pos- terior sides of the middle lobe of M 1 in A. chu- calensis are lingually convergent, making the lobe triangular. This middle lobe is relatively smaller than in any of these other taxa, completely sur- rounded by the subequal anterior and posterior lobes. The second upper molar resembles M 1 , ex- cept that the parastyle is more pronounced and the middle lobe slightly larger. Imbrication of the up- per molars is slight, as in all the other mesother- iines from Chucal and a few additional taxa (see Table 7). The third upper molar most closely re- sembles that of Microtypotherium, with a small middle lobe surrounded by the other two. No lin- gual sulcus is evident on the posterior lobe. A fourth lobe is present in M3 but is slightly shorter and broader than that of A. paucidens. Enamel is absent from the posterior face. All upper cheek teeth are covered by a moderate amount of ce- mentum, which is thickest lingually as in A. pau- cidens. The degree of cheek tooth imbrication in A. chucalensis is 1.22. The anterior lower dentition of A. chucalensis is best exhibited by SGOPV 4110, a partial left mandible with left and right il-2, left p4-ml, par- tial left m2, and partial left Ml. These teeth are very similar to the corresponding teeth of A. pau- cidens. The symphyseal region is unknown. The cheek teeth and the posterior portion of the mandible are preserved in the holotype (Fig. 10). The horizontal ramus is less robust than that of A. paucidens; it measures approximately 18.5 mm deep between p4 and ml and 30 mm deep at the posterior face of m3. On the left side, it appears that a mental foramen is present but incompletely preserved; its position is similar to that of A. pau- cidens. The anterior edge of the ascending ramus creates an angle of slightly less than 90° with the horizontal ramus and ends in a broad coronoid process. At its vertical midpoint, the ascending ramus measures approximately 35 mm from its anterior to its posterior edge. The coronoid pro- cess rises approximately 13.5 mm above the level of the articular condyle and is nearly 20 mm long. The articular condyle is broad, smooth, and —14 mm wide. The angle of the mandible is large, but the bone comprising much of it is quite thin. A mandibular foramen is present along the lingual side of the dentary, where the horizontal ramus joins the ascending ramus. It measures approxi- mately 7X4 mm. The posterior part of the in- ferior edge of the angle is inflected medially. Slightly anterior to this, a pronounced shelf is pre- sent on the lateral surface. The mandible's lateral surface is fairly smooth, but the medial surface of the angle and ascending ramus exhibits a series of well-pronounced ridges, suggesting that a well- developed medial pterygoid muscle inserted in this area. The lower cheek teeth are similar to those of A. paucidens, except for their relatively smaller size. Additionally, the talonid of m3 is propor- tionately shorter and narrower A partial skeleton and other postcranial speci- mens are referred to A. chucalensis (SGOPV 4042). These will be described in detail and com- pared to other mesotheriid postcrania elsewhere. Comments — Although roughly similar in size to A. paucidens, A. chucalensis is clearly distin- guished from the former species by the suite of character states presented in the diagnosis. Eotypotherium chico, species novum (Figures 11-13, Table 5) Holotype — SGOPV 5157, partial right maxilla and zygomatic arch with P3-M3 (Fig. 11). Hypodigm— SGOPV 5108, partial right II; SGOPV 5074, partial left maxilla with P3-M1 and partial M2 (Fig. 12 A); SGOPV 5156, partial palate with right P3-M3 (M2 damaged labially; Fig. 12B); SGOPV 5124, little-worn left maxil- lary fragment with Ml and partial P4, M2; SGOPV 5132, left Ml/2; SGOPV 5175 right Ml/ 2; SGOPV 5095, partial left Ml/2; SGOPV 5134, partial left Ml/2; SGOPV 5133, left M3; SGOPV 4071, symphysis with portions of left and right il-2; SGOPV 5050, right ml; SGOPV 5129, right mandibular fragment with partial ml; SGOPV 5159, partial right ml/2; SGOPV 5160, partial right ml/2; SGOPV 5158, partial left mandible with m2-3 (Fig. 13); SGOPV 5086, partial right mandible with bases of m2-3; SGOPV 5093, right m3. Type Locality— Loc. C-ALT-01-12; Chucal Formation (Member E3 of Charrier et al., 2002); east flank of Chucal Anticline. Age and Distribution — Chucal Formation, northern Chile, late eariy Miocene age, Santacru- cian SALMA. Etymology — Eo, Greek for "dawn," in refer- ence to the relative antiquity of this species with respect to previously known mesotheriines, and CROFT ET AL.: CHUCAL UNGULATES 25 Table 5. Measurements for specimens of Eotypotherium chico. II P3 P4 Ml M2 M3 Ml/2 Upper dentition L W L W L W L W L W L W L W SGOPV 5157 (R) 4.4 4.0 5.2 5.3 9.5 6.4 9.6 6.5 9.8 5.8 SGOPV 5108 (R) (6.7) 4.0 SGOPV 5074 (L) 4.2 4.8 5.8 5.6 8.2 5.9 — (6.6) SGOPV 5156 (R) 3.1 3.7 4.9 5.3 8.0 (6.2) (9.5) — 9.6 5.4 SGOPV 5124 (L) — — 8.3 3.9 — — SGOPV 5132 (L) (8.5) 6.2 SGOPV 5175 (R) 9.4 5.5 SGOPV 5095 (L) (9.6) (6.7) SGOPV 5134 (L) (7.5) (5.3) SGOPV 5133 (L) 11.5 6.4 il i2 m il m2 m3 ml/2 Lower dentition L W L W L W L W L W L W SGOPV 4071 (R) 4.3 3.0 3.5 (1.5) SGOPV 4071 (L) 4.4 2.7 3.2 1.8 SGOPV 5050 (R) 9.6 6.0 SGOPV 5129 (R) 8.7 4.9 SGOPV 5159 (R) 8.9 4.6 SGOPV 5160 (R) 8.5 4.7 SGOPV 5158 (L) 8.9 4.7 11.0 3.8 SGOPV 5086 (R) (9.0) 4.7 7.3 3.4 SGOPV 5093 (R) 8.0 3.6 typotherium, a common suffix for mesotheriines; chico, Spanish for "little," referring to the small size of the species. Diagnosis — A mesotheriine differing from Al- titypotherium paucidens (above) in absence of well-developed zygomatic plate; presence of wide postpalatal notch not reaching the level of M3 (narrow and reaching the level of M3 in A. pau- cidens); presence of subtriangular II that is point- ed distally, with mesial sulcus (with two lingual sulci in A. paucidens); presence of short, subtrian- gular P4 with central fossette (fossette absent in A. paucidens); absence of enlarged middle lobe of M3 (surrounded by other lobes in E. chico); pres- ence of subquadrangular P3 (P3 absent in A. pau- cidens); absence of single labial sulcus on pre- molar ectolophs; and smaller size (based on dental measurements, approximately 25% smaller). Differs from Altitypothehum chucalensis (above) in absence of well-developed zygomatic plate; presence of wide postpalatal notch not reaching the level of M3 (narrow and reaching the level of M3 in A. chucalensis); presence of sub- triangular II that is pointed distally, with mesial sulcus (with two lingual sulci in A. chucalensis); presence of short, subtriangular P4 with central fossette (fossette absent in A. chucalensis); pres- ence of subquadrangular P3 (subtriangular in A. chucalensis); absence of single labial sulcus on premolar ectolophs; presence of elongate Ml (<1.30 in A. chucalensis, >1.30 in Eotypothe- rium); and smaller size (based on dental measure- ments, approximately 20% smaller). Differs from Microtypotherium in absence of well-developed zygomatic plate; presence of wide postpalatal notch not reaching the level of M3 (narrow in Microtypotherium); presence of short, subtriangular P4 with central fossette (fossette ab- sent in Microtypotherium); presence of subquad- rangular P3 (subtriangular in Microtypotherium); absence of elongate Ml (length/width <1.30 in Eotypotherium, between 1 .30 and 1 .75 in Micro- typotherium); absence of shortened m2 (length/ width between 1.60 and 2.30 in Eotypotherium, <1.60 in Microtypotherium); and smaller size (based on dental measurements, approximately 30% smaller than Microtypotherium choqueco- tense). Differs from Eutypotherium in root of zygo- matic arch perpendicular to Ml (posteriorly di- rected from the level of Ml or M2 in Eutypo- therium); absence of well-developed zygomatic plate; presence of subtriangular 1 1 that is pointed distally, with mesial sulcus (wide, rounded, with lingual sulcus in Eutypotherium); presence of short, subtriangular P4 with central fossette (fos- 26 FIELDIANA: GEOLOGY Fig. 11. Holotype of Eotypotherium chico, SGOPV 5157, a partial right maxilla and zygomatic arch with P3-M3. A. Labial (above) and occlusal (below) views. B. Line drawing of occlusal morphology. Scale bars - 1 cm. sette absent in Eutypotherium, lingual plication sometimes present); presence of il that is rounded in cross section with little or no lingual sulcus (subtriangular with smooth sulcus in Eutypothe- rium); presence of subquadrangular P3 (subtrian- gular in Eutypotherium); absence of elongate Ml (length/width <1.30 in Eotypotherium, between 1.30 and 1.75 in Eutypotherium); absence of ex- tensive lingual exposure of Ml middle lobe; and much smaller size (based on dental measure- ments, approximately 40% smaller than Eutypo- therium lehmannnitschei). Differs from Plesiotypotherium in absence of well-developed suborbital fossa and large infra- orbital foramen; absence of well-developed zy- gomatic plate; presence of subtriangular II that is pointed distally, with mesial sulcus (wide, round- ed, with lingual sulcus in Plesiotypotherium); presence of short, subtriangular P4 with central fossette (fossette absent in Plesiotypotherium, lin- CROFT ET AL.: CHUCAL UNGULATES 27 Fig. 12. Upper dentitions referred to Eotypotherium chico. A. SGOPV 5074, partial left maxilla (pictured as right) with P3-M1 and partial M2 in occlusal view. B. SGOPV 5156. a partial palate with right P3-M3 in labial (above) and occlusal (below) views. C. Line drawing of occlusal morphology. Scale bars = 1 cm. 28 FIELDIANA: GEOLOGY Fig. 13. Partial left mandible with m2-3 in occlusal view, SGOPV 5158. referred to Eotypotherium chico. Scale bar = 1 cm. gual plication sometimes present); presence of il that is rounded in cross section with little or no lingual sulcus (subtriangular with smooth sulcus in Plesiotypotherium); presence of subquadran- gular P3 (subtriangular in Plesiotypotherium); ab- sence of elongate Ml (length/width ^1.30 in Eotypotherium, between 1.30 and 1.75 in Plesi- otypotherium); presence of Ml middle lobe with lingually convergent anterior and posterior sides (subparallel in Plesiotypotherium); absence of ex- tensive lingual exposure of Ml middle lobe; and much smaller size (based on dental measure- ments, approximately half the size of Plesioty- potherium achirense). Differs from Typotheriopsis in absence of well- developed suborbital fossa and large infraorbital foramen; absence of well-developed zygomatic plate; presence of wide postpalatal notch not reaching the level of M3 (narrow in Typotheriop- sis); presence of subtriangular 11 that is pointed distally, with mesial sulcus (wide, rounded, with lingual sulcus in Typotheriopsis); presence of short, subtriangular P4 with central fossette (fos- sette absent in Typotheriopsis); relatively smaller il (il/i2 <1.50 in Eotypotherium, between 1.50 and 3.00 in Typotheriopsis); presence of i 1 that is rounded in cross section with little or no lingual sulcus (subtriangular with smooth sulcus in 7\'- potheriopsis); presence of subquadrangular P3 (subtriangular in Typotheriopsis); absence of sin- gle labial sulcus on premolar ectolophs; absence of elongate M 1 (length/width ^ 1 .30 in Eotypoth- erium, between 1.30 and 1.75 in Typotheriopsis); presence of Ml middle lobe with lingually con- vergent anterior and posterior sides (subparallel in Typotheriopsis); absence of extensive lingual ex- posure of Ml middle lobe; and much smaller size (based on dental measurements, approximately half the size of Typotheriopsis chasicoensis). Differs from Caragua New Taxon in presence of wide postpalatal notch not reaching the level of M3 (narrow and reaching the level of M3 in Caragua New Taxon); lesser degree of molar im- brication (<1.25 in Eotypotherium, between 1.25 and 1.50 in Caragua New Taxon); presence of subtriangular II that is pointed distally, with me- sial sulcus (wide, rounded, with lingual sulcus in Caragua New Taxon); presence of short, subtrian- gular P4 with central fossette (fossette absent in Caragua New Taxon, lingual plication present); relatively smaller il (il/i2 <1.50 in Eotypothe- rium, between 1 .50 and 2.00 in Caragua New Tax- on); presence of i 1 that is rounded in cross section with little or no lingual sulcus (subtriangular with smooth sulcus in Caragua New Taxon); presence of subquadrangular P3 (subtriangular in Caragua New Taxon); absence of elongate Ml (length/ width < 1 .30 in Eotypotherium, between 1 .30 and 1.75 in Caragua New Taxon); presence of Ml middle lobe with lingually convergent anterior and posterior sides (subparallel in Caragua New Taxon); absence of extensive lingual exposure of Ml middle lobe; absence of two lingual sulci on m3 talonid; and much smaller size (based on den- tal measurements, approximately 40% smaller than Caragua New Taxon). Differs from Pseudotypotherium in absence of well-developed suborbital fossa and large infra- orbital foramen; absence of well-developed zy- gomatic plate; lesser degree of molar imbrication (<1.25 in Eotypotherium, >1.50 in Pseudotypo- therium); presence of short, subtriangular P4 with central fossette (fossette absent and tooth bilobed in Pseudotypotherium); presence of M3 with mid- CROFT ET AL.: CHUCAL UNGULATES 29 die lobe surrounded by the other two (not sur- rounded in Pseudotypotherium); relatively smaller il (il/i2 <1.50 in Eotypotherium, between 2.00 and 3.00 in Pseudotypotherium); presence of il that is rounded in cross section with little or no lingual sulcus (trapezoidal with well-demarcated lingual sulcus in Pseudotypotherium); presence of subquadrangular P3 (subtriangular in Pseudoty- potherium); absence of elongate P4 (length/width <1.50 in Eotypotherium, ^1.50 in Pseudotypo- therium); absence of elongate Ml (length/width ^1.30 in Eotypotherium, ^1.75 in Pseudotypo- therium); presence of Ml middle lobe with lin- gually convergent anterior and posterior sides (subparallel in Pseudotypotherium); absence of extensive lingual exposure of Ml middle lobe; ab- sence of elongate m2 (length/width between 1 .60 and 2.30 in Eotypotherium, >2.30 in Pseudoty- potherium); and much smaller size (based on den- tal measurements, approximately 40% smaller than Pseudotypotherium exiguum). Differs from Mesotherium in presence of poor- ly defined suborbital fossa (absent in Mesothe- rium); absence of well-developed zygomatic plate; presence of wide postpalatal notch not reaching the level of M3 (narrow in Mesotherium, sometimes extending to the level of M3); lesser degree of molar imbrication (<1.25 in Eotypo- therium, >1.50 in Mesotherium); presence of sub- triangular II that is pointed distally, with mesial sulcus (wide, rounded, with lingual sulcus or wide and compressed labiolingually with two lingual sulci in Mesotherium); presence of short, subtrian- gular P4 with central fossette (fossette absent and tooth bilobed in Mesotherium); presence of M3 with middle lobe surrounded by the other two (not surrounded in Mesotherium); relatively smaller il (il/i2 <1.50 in Eotypotherium, >3.00 in Meso- therium); presence of il that is rounded in cross section with little or no lingual sulcus (trapezoidal with well-demarcated lingual sulcus or elliptical with smooth sulcus in Mesotherium); presence of subquadrangular P3 (subtriangular in Mesothe- rium); absence of elongate P4 (length/width <1.50 in Eotypotherium, ^1.50 in Mesotherium); absence of elongate Ml (length/width <1.30 in Eotypotherium, ^1.75 in Mesotherium); presence of Ml middle lobe with lingually convergent an- terior and posterior sides (lingually divergent in Mesotherium); absence of extensive lingual ex- posure of Ml middle lobe; absence of elongate m2 (length/width between 1.60 and 2.30 in Eoty- potherium, >2.30 in Mesotherium); and much smaller size (based on dental measurements, ap- proximately 40% smaller than Mesotherium maendrum, 60% smaller than Mesotherium pach- ygnathum). Description — This mesotheriine is much smaller than the other two known from Chucal; indeed, it represents the smallest mesotheriid known. In addition to its small size, it is distin- guished from all other mesotheriines by the reten- tion of several primitive character states lost in other taxa (see phylogenetic analysis below). The holotype (SGOPV 5157; Fig. 11) consists of a partial right maxilla and zygomatic arch bear- ing P3-M3. The superior surface of the zygomatic arch is well preserved, and the suborbital fossa is present but poorly developed, much less so than in Eotypotherium. A second specimen referred to Eotypotherium (SGOPV 5074; Fig. 12A) also pre- serves this region and confirms the lack of a well- developed fossa. The zygomatic plate is also poorly developed, more resembling the condition seen in Trachytherus than other mesotheriines. The infraorbital foramen, although small in ab- solute size (4.0 X 3.5 mm), is similar to that of most other mesotheriines when scaled to body size (Table 3). No upper incisors are preserved in the holotype (only II would be predicted to be present), but their morphology can be discerned from an iso- lated partial II referred to this taxon based on its small size (SGOPV 5108). Like II of Altitypo- therium, that of Eotypotherium is subtriangular and pointed distally in cross section. However, un- like Altitypotherium, a single mesial sulcus is pre- sent on II instead of two lingual ones. In this respect, the tooth resembles that of Trachytherus and Microtypotherium. Unlike that of Trachyth- erus (and like those in all other mesotheriines), II is completely surrounded by enamel, save for its occlusal surface. As in A. chucalensis, Eotypotherium possesses two premolars, presumably P3 and P4. These are both subquadrangular (as opposed to subtriangu- lar, as in most other mesotheriines), and each pos- sesses a single labial sulcus. Although the occlu- sal surfaces of the premolars in the holotype are featureless, another specimen referred to Eoty- potherium (SGOPV 5074, a partial left maxilla with P3-M1 and partial M2) demonstrates that these likely vary in occlusal morphology with wear in this species. In SGOPV 5074, a small enamel lake is present in the center of P3. Pre- sumably, this is the remnant of a lingual sulcus that has been isolated with wear and that would probably have disappeared altogether with addi- 30 FIELDIANA: GEOLOGY tional wear. The presence of such an internal sul- cus (or an enamel lake) in P3 is unknown among mesotheriines, although it has been reported in DP3 (Francis, 1965). In P4 the sulcus has simi- larly been isolated by wear but differs from that of P3 in being "Y" shaped; the base of the "Y" is directed lingually and the two arms anterola- bially and posterolabially, isolating a very small middle lobe adjacent to the ectoloph. The mor- phology of P3-4 in SGOPV 5156 (a third maxilla referred to Eotypotherium; Fig. 12B) is similar to that exhibited by SGOPV 5074, except that the fossettes are less pronounced, likely because of greater wear. A lingual P4 sulcus (or plication) is absent in Typotheriopsis and Microtypotherium choquecotense, variably present in Eutypotherium (potentially interspecific variation) and Plesioty- potherium (potentially intraspecific variation), and consistently present in Caragua New Taxon, Pseu- dotypotherium, and Mesotherium (Francis, 1965; Cerdeilo and Montalvo, 2001; Flynn et al., in press). In none of these taxa, however, is the sul- cus isolated with wear or shaped like a "Y" The presence of isolated fossettes in the premolars in Eotypotherium most closely resembles the condi- tion seen in Trachytherus, suggesting that it is a retained primitive feature (see phylogenetic anal- ysis below). The fourth upper premolar is im- planted at —10° relative to Ml in the holotype, while P3 is at an even greater angle, closer to 25°. The first upper molar in the holotype has a tri- angular middle lobe, more reminiscent of that in A. paucidens than A. chucalensis. A triangular lobe is also present in Ml of SGOPV 5156, while SGOPV 5074 exhibits a more rectangular middle lobe, more typical of later mesotheriines. The morphology of the middle lobe of Ml therefore seems somewhat variable in this taxon. This tooth is also relatively elongate, unlike the condition seen in A. chucalensis. The remaining molars re- semble those of A. paucidens, save for their much smaller size; the degree of imbrication (1.12) is similar to that exhibited by other mesotheriines from Chucal. The lower anterior dentition is represented by SGOPV 4071, a symphysis bearing portions of left and right il-2. As in other mesotheriines, the symphysis is completely fused and relatively thick; even in this diminutive taxon, the symphy- sis is approximately 8.5 mm thick at the midline. The dorsal surface of the symphysis is strongly concave and spoutlike, as in A. paucidens. The spout is approximately 5 mm wide. Both il and i2 are small and rounded; i 1 lacks a lingual sulcus. The lower incisors are procumbent and implanted at an angle, as in other mesotheriines. Little of the occlusal morphology is preserved, save for the an- terior portion of left i 1 . No p4 referable to Eoty- potherium is known. The lower molars referred to this taxon closely resemble those of the other me- sotheriines from Chucal, except for their smaller size. Specimen SGOPV 5158 preserves a small portion of the dentary, which is approximately 28 mm deep at the posterior face of m3 (Fig. 13). The corresponding measure from SGOPV 5086, likely from a young individual, is approximately 20 mm. Comments — The most distinctive aspect of Eotypotherium is its small size; it is approximate- ly two-thirds the size of A. paucidens and Micro- typotherium choquecotense, making it the small- est, securely identified mesotheriine known. Vil- larroel (1978) described a specimen (GB 223) from near Quehua that he assigned to Microty- potherium cf. M. choquecotense; despite this as- signment, he noted that it was smaller than M. choquecotense (presumably referring to the holo- type) and therefore was the smallest mesotheriine known. The dental measurements provided by Villarroel (1978) indicate that this specimen is roughly similar in size to, or perhaps a bit larger than, Eotypotherium; the lengths for il and ml are close to those measured for Eotypotherium, but the widths for these and other teeth are great- er. Similarly, m2 is slightly longer and m3 signif- icantly (50%) longer than that of Eotypotherium. It is therefore unlikely that GB 223 represents a taxon smaller than Eotypotherium. The only pub- lished specimen representing a mesotheriine pos- sibly smaller than Eotypotherium was collected by Oiso (1991) in Nazareno, Bolivia, and included without comment under '' Plesiotypotherium sp." (Oiso, 1991, table 2, GB Naz-060). If this isolated Nazareno specimen (about half the size of Eoty- potherium) represents an adult individual (the de- ciduous premolars of mesotheriines resemble adult permanent molars and have historically caused confusion; Francis, 1965), it undoubtedly represents a species distinct from the remaining Nazareno specimens and would represent the smallest mesotheriine known. Hegetotheria Simpson, 1945 Hegetotheriidae Ameghino, 1894 Hegetotheriinae Ameghino, 1894 Diagnosis — Hegetotheriids are distinguished from other notoungulates by the unique combi- CROFT ET AL.: CHUCAL UNGULATES 31 B Fig. 14. Comparison of the degree of fusion of the tibia and fibula among selected Santacrucian typotheres. A. Pachyrukhos moyani (Hegetotheriidae: Pachyrukhinae), AMNH 9242, left tibiofibula, pictured as right. B. Hegeto- therium mirabile (Hegetotheriidae: Hegetotheriinae), PU 15298, right tibiofibula. C. Protypotherium attenuatum (In- teratheriidae: Interatheriinae), PU 15341, right tibia and fibula. All specimens in anterior view. Scale bars = 1 cm. Individual illustrations are modified from Sinclair (1909). nation of the following derived character states: presence of hypselodont first incisors and cheek teeth; presence of flat wall on lingual faces of up- per and lower molars; presence of deep, labial sul- cus on ml and m2; presence of cementum; ab- sence of fossettes in upper and lower cheek teeth; and proximal and distal fusion of tibia and fibula (Cifelli, 1993; Croft, 2000). Among hegetothe- riids, hegetotheriines share a uniquely configured distal fusion of the tibia and fibula (see below) and lack many derived character states typical of pachyrukhines. Derived character states of the dentition diagnostic of the Pachyrukhinae (and lacking in the Hegetotheriinae) include the pres- ence of a sharply defined third lobe on m3 (Cer- deno and Bond, 1998), significant reduction or ab- sence of I3-C/i3-c, and presence of i 1 much larger than i2. Comments — The hegetotheriid material from the Chucal Formation includes isolated and/or po- tentially associated dental and postcranial speci- mens that fall within the size range typical of he- getotheriids. Currently, there are no size or mor- phological differences suggesting that more than one hegetotheriid taxon occurs in the Chucal Fau- na. Assignment to the Pachyrukhinae is precluded by the morphology of m3: the Chucal taxon has a bilobed or weakly trilobed m3, whereas in all known pachyrukhines this tooth is sharply tri- lobed. Moreover, the Chucal taxon is larger than any described pachyrukhine, and occlusal outlines of the upper molars are dissimilar. In addition to dental morphology, recovered postcranial elements support referral of this ma- terial to the Hegetotheriinae. In most hegetothe- riids for which postcrania are known, the tibia and fibula are fused at both the proximal and the distal ends. In typical pachyrukhines (e.g., Pachyrukhos, Paedotherium, Tremacyllus), the distal fusion is long, comprising approximately 40% the length of the two elements. The fibula bows medially at the proximal point of fusion but then bows laterally again at the distal articulation with the ankle (Fig. 14A). In overall structure, the condition resembles that seen in modern leporids (Sinclair, 1909). In hegetotheriines (and, potentially, the pachyru- khine Propachyrucos), the distal fusion is quite short, approximately 15% the length of the tibia and fibula. Additionally, the fibula parallels the tibia throughout its length, resulting in a much broader articulation distally — approximately equal in breadth to the fused proximal ends of the two bones (Fig. 14B). A similar configuration occurs in interatheriids, but it differs from that of hege- 32 FIELDIANA: GEOLOGY totheriines in lacking complete distal fusion of the tibia and fibula (Fig. 14C). Two partial hegeto- theriid tibiofibula specimens have been recovered from the Chucal Formation, both with the hege- totheriine configuration of a short, broad distal fu- sion. Hegetotherium Ameghino, 1887 Type Species — Hegetotherium mirabile. Included Species — The type, H. convexum, H. minum, H. anceps, and H. novum. Diagnosis — A hegetotheriine differing from Ethegotherium in absence of median labial sulcus/ groove between paracone and metacone on upper molars, presence of imbrication in lower molars, presence of lower canine root labial to toothrow, and presence of labial groove on m3 talonid. Dif- fers from Prohegetotherium in absence of median labial sulcus between paracone and metacone on upper molars, presence of P2/p2 notably smaller than P3/p3, presence of small diastemata between i2 and p3, and reduction or absence of lingual groove on m3 talonid. Differs from Hemihege- totherium (including Pseudohegetotherium; see below) in absence of thick cementum on lingual faces of upper and lower molars, absence of mo- lars with convex lingual faces and posterior faces that are narrow in comparison to the rest of the occlusal surface, variable presence of small but well-demarcated parastyle and parastyle sulcus on Ml-2, and presence of notch on posterior face of M3. Distribution — Sarmiento Formation, Patagon- ia, Argentina, early Miocene age, Colhuehuapian SALMA (Bordas, 1939); Santa Cruz Formation, Patagonia, Argentina, late early Miocene age, Santacrucian SALMA (Sinclair, 1909); Chucal Formation, northern Chile, late early Miocene age, Santacrucian SALMA (present study); un- named formation (possibly equivalent to Rio Ze- ballos Formation), southern Chile, late early Mio- cene age, Santacrucian SALMA (Flynn, Novacek, et al., 2002); CoUon-Cura Formation, Patagonia, middle Miocene age, Colloncuran SALMA (Bon- desio et al., 1980); Nazareno Formation, southern Bolivia, ?middle Miocene age, ?Colloncuran SALMA (Oiso, 1991). Comments — The hegetotheriid material collect- ed from Chucal in 1998 was tentatively referred to cf. Pseudohegetotherium torresi (Flynn et al., 1999; Flynn, Croft, et al., 2002). Additional he- getotheriine specimens collected in 2001 have clarified the morphology of the Chucal taxon (es- pecially characters of the upper dentition), sug- gesting it more likely pertains to Hegetotherium. Bond and Lopez (1997) concluded that al- though P. torresi (the only species of Pseudohe- getotherium) is a valid taxon, it should perhaps be included within Hemihegetotherium. In the above diagnosis and discussion below, Hemihegetothe- rium subsumes '^Pseudohegetotherium'" torresi. The earlier members of the Hegetotheriinae (i.e., Ethegotherium, Prohegetotherium, and He- getotherium) have upper molars that are roughly rectangular in outline; the lingual and labial faces are approximately straight and parallel (Croft, 2000). In contrast, the upper molars of later forms {Hemihegetotherium) have more convex lingual faces and posterior faces that are narrower than the rest of the occlusal surface. Additionally, the cementum is thicker in these later hegetotheriines, and molar paracones and metacones are less pro- nounced than in earlier members of the group (Bond and Lopez, 1997). Although hegetotheriid upper molars from the Chucal Formation exhibit relatively thick cementum, this feature is not de- veloped to the degree seen in later hegetotheri- ines. Moreover, the cheek teeth are roughly rect- angular in outline, with a posterior face that is about as wide as the rest of the occlusal surface. The paracones and metacones are variably pro- nounced on specimens from Chucal, but most ex- hibit a small yet well-demarcated parastyle and parastyle sulcus. The latter features are absent in later hegetotheriines. The single hegetotheriine M3 known from Chucal also exhibits a prominent notch on the distal face. A report of Hegetotherium from the middle Miocene Quebrada Honda Fauna of Bolivia (Mar- shall & Sempere, 1991) appears to be in error, based on the taxon's large size and sharply tri- lobed m3 (pers. observ.); this fauna is thus ex- cluded from the cited distribution of Hegetothe- Hegetotherium sp. cf. H. mirabile (Figures 15, 16, Table 6) Referred Specimens— SGOPV 5047, left II; SGOPV 5039, left maxillary fragment with bases of Pl-2; SGOPV 5120, right P3; SGOPV 5040, right P4 (Fig. 15 A); SGOPV 5058, left ?M1 (Fig. 15B); SGOPV 5121, left ?M1 (Fig. 15C); SGOPV 5122, partial left ?M2 (lacking occlusal surface); SGOPV 5123, left M3 (Fig. 15D); SGOPV 4055, CROFT ET AL.: CHUCAL UNGULATES 33 Fig. 15. Isolated upper cheek teeth referred to He- getotherium cf. H. mirabile, in occlusal view. A. SGOPV 5040, right P4. B. SGOPV 5058, left ?M1 (pic- tured as right). C. SGOPV 5121, left ?M1 (pictured as right). D. SGOPV 5123g, left M3 (pictured as right). Scale bars = 5 mm. partial left mandible with root of c, pl-2 (Fig. 16A); SGOPV 5096, partial right mandible with root of c, pl-2; SGOPV 4104, partial left man- dible with p4-m3 (Fig. 16B); SGOPV 5042, left ml with labial fragment of mandible; SGOPV 5049, partial right mandible with ml-2; SGOPV 5082, partial left mandible with bases of ml-2; SGOPV 5222, partial right mandible with m2; SGOPV 4023, partial right mandible with talonid of ml, complete m2-3 (Fig. 16C); SGOPV 5038, partial right mandible with base of m3; SGOPV 5059, left distal tibiofibula fragment lacking epiphysis; SGOPV 5070, partial right leg and pes including distal tibiofibula (lacking epiphyses), calcaneus (epiphysis reattached), cuboid, ?meso- cuneiform, metatarsal IV, partial metatarsal ?III, proximal phalanx, distal phalanx; SGOPV 5224, left calcaneus. Diagnosis — Differs from Hegetotherium con- vexum in presence of M3 ^ M2 in length, absence of highly convex lingual face of M3, and presence of pronounced posteroexternal angle of M3. Dif- fers from H. minum in deeper mandibular ramus. Differs from H. anceps in presence of M3 ^ M2 in length. Differs from H. novum in presence of convex internal face on P2, presence of relatively wider cheek teeth, and presence of M3 ^ M2 in length. Age and Distribution — Santa Cruz Formation, Patagonia, Argentina, late early Miocene age, Santacrucian SALMA; Chucal Formation, north- ern Chile, late early Miocene age, Santacrucian SALMA. Description — The hegetotheriine specimens from Chucal are fragmentary, the most complete specimen being SGOPV 4104, a partial right man- dible with p4-m3 (Fig. 16B). A variety of other mandibular fragments with partial lower denti- tions are known, but the upper dentition is known exclusively from isolated teeth. Specimen SGOPV 5047 is a left upper first in- cisor with an occlusal surface measuring approx- imately 7.8 X 2.8 mm. It is rootless and was im- planted obliquely, as in other hegetotheriids. At least as judged from the length of the occlusal surface, this specimen is small for Hegetotherium mirabile; the corresponding tooth in PU 15542 (Sinclair, 1909, Plate I) measures approximately 11.5 X 3.0 mm, and in FMNH P13194 it mea- sures approximately 10.0 X 2.9 mm. Since the occlusal surface of the Chucal specimen is set at approximately the same angle to the rest of the tooth as in other specimens of H. mirabile ex- amined, this difference in surface dimensions may 34 FIELDIANA: GEOLOGY Fig. 16. Lower dentitions referred to Hegetotherium cf. H. mirabile. in occlusal view. A. SGOPV 4055, partial left mandible with root of c, pl-2. B. SGOPV 4104, partial left mandible with p4-m3. C. SGOPV 4023, partial right mandible with talonid of ml and complete m2-3 (viewed as left). Scale bars equal 5 rmn. result from differences in wear. Alternatively, the Chucal specimen might be from a smaller than average individual (as in SGOPV 5123; Fig. 15D). Most examples of the upper dentition from Chucal compare favorably with specimens of H. mirabile from Patagonia both in size and in mor- phology (Fig. 15). Although somewhat small, the upper third molar (SGOPV 5123) has a pro- nounced notch in the posterior face of the tooth, CROFT ET AL.: CHUCAL UNGULATES 35 characteristic of Hegetotherium (Fig. 15D). Of the two posterior projections created by this notch, the labial one is slightly longer. This feature is variable within Hegetotherium from Patagonia; the labial projection ranges from slightly longer to significantly longer than the lingual (Sinclair, 1909). Several Chucal specimens from the same locality (SGOPV 5121, left ?M1; SGOPV 5122, partial left ?M2; SGOPV 5123, left M3) may rep- resent the left upper molar series from a single individual. However, since this locality has pro- duced an abundance of isolated teeth from a va- riety of taxa, the basis for such an association is not compelling. The lower dentition of the Chucal hegetotheriid is much better represented than the upper denti- tion; seven well-preserved partial mandibles have been collected, although none preserves the inci- sors or p3. The specimens resemble H. mirabile in morphology and are comparable in size. As in M3, the morphology of m3 is variable among H. mirabile specimens both from Patagonia and from Chucal. In some specimens, the labial talonid groove on m3 is very shallow; in others, it is pro- nounced, lending the tooth a nearly trilobed ap- pearance. This feature is also variably expressed in the three specimens from Chucal; in SGOPV 4023 (Fig. 16C) and SGOPV 5038, the groove is very weakly present, whereas in SGOPV 4104 (Fig. 16B), the talonid groove is deep, and a third lobe is present (though it is not developed to the extent seen in pachyrukhines). Comments — In his description of the typotheres of the Santa Cruz Beds, Sinclair (1909) referred all hegetotheriines in the Princeton collections to Hegetotherium mirabile. He considered five other taxa to be junior synonyms of H. mirabile (H. strigatum, H. cuneatum, H. costatum, ''Selathe- rium'' pachymorphum, and ''Selatherium'" remis- sum) and listed three others as of questionable sta- tus (//. convexum, H. minum, and H. anceps). All three questionable taxa were named by Ameghino based on Santacrucian specimens. The holotype of H. convexum is a partial left maxilla with M2- 3 figured by Ameghino (1891, fig. 30). If the fig- ure and the description provided are accurate, the specimen would not appear to pertain to H. mi- rabile (see diagnosis above) and may be distinct. H. minum is said to be distinguished from H. mi- rabile primarily by its shallower mandibular ra- mus (Ameghino, 1894), but the variability of this character (especially with wear) is uncertain. H. anceps differs from H. mirabile in having an M3 that is shorter than M2 (as in H. convexum) but is not described as differing from H. mirabile in M3 shape (Ameghino, 1891). H. anceps may rep- resent a species distinct from both H. mirabile and H. convexum, or may be synonymous with the latter. H. novum was named by Bordas (1939) based on a partial skull recovered from Colhue- huapian strata in the Valle del Rio Chubut and shares with both H. convexum and H. anceps the shortened M3 in relation to M2. The other distin- guishing characters of H. novum described by Bordas (1939) suggest this taxon may be distinct from H. mirabile and other recognized species of Hegetotherium. Owing to the incompleteness of Hegetotherium specimens from Chucal and the lack of a compre- hensive revision of Hegetotherium taxonomy, a definitive identification of the material cannot now be made. Nevertheless, no compelling evi- dence argues against reference of these specimens to H. mirabile. Upper cheek teeth from Chucal (Table 6) are wider than those of H. novum, and SGOPV 5039 confirms that P2 of the Chucal form lacks the concave lingual face (characteristic of H. novum). Two mandibular fragments from Chu- cal (SGOPV 5104 and SGOPV 5049) illustrate that the horizontal ramus below ml is deeper in the Chucal taxon than in H. minum (15.5 mm and 15.4 mm in the Chucal specimens versus 14.0 mm in H. minum). The single hegetotheriine M3 col- lected from Chucal (SGOPV 5123) is a small tooth, shorter than any of the other hegetotheriine molars known from the area (Table 6). Although this may imply that the Chucal Hegetotherium has an M3 shorter than M2 (as in H. convexum and H. anceps), the lack of an associated M2 clouds this interpretation; the possibility that this M3 merely pertains to a smaller individual (whose M2 is not preserved) cannot be excluded. Moreover, even though the Chucal M3 is smaller than the other upper molars from Chucal, it is larger than M3 in the holotypes of both H. convexum and H. anceps. The presence of a pronounced postero- external angle and the lack of a greatly concave lingual face in this M3 differs from the holotype of H. convexum and more closely resembles H. mirabile. Litopterna Ameghino, 1889 Macraucheniidae Gervais, 1855 Cramaucheniinae Ameghino, 1902 Diagnosis — The postcrania (especially the tar- sals) of litopterns are highly diagnostic and read- 36 FIELDIANA: GEOLOGY Table 6. Measurements for Chucal specimens of Hegetotherium mirabile. Upper dentition Position W SGOPV 5120 (R) P3 4.9 3.7 SGOPV 5040 (R) P4 6.8 4.1 SGOPV 5058 (L) ?M1 8.1 (4.9) SGOPV 5121 (L) ?M1 7.4 4.7 SGOPV 5122 (L) ?M2 (7.2) (4.6) SGOPV 5123 (L) M3 6.0 3.8 Lower dentition Pl p2 p3 p4 ml m2 W W W W W W m3 W SGOPV 4055 (L) 3.2 (1.7) 5.2 3.3 SGOPV 5096 (R) (2.8) (1.7) 5 2.8 SGOPV 4104 (L) SGOPV 5042 (L) SGOPV 5049 (R) SGOPV 5082 (L) SGOPV 5222 (R) SGOPV 4023 (R) SGOPV 5038 (R) 6.5 4.1 7.4 3.9 7.1 3.9 3.6 7.3 3.9 (7.0) 4.4 7.3 4.3 (7.1) (3.7) 7.5 4 7.0 4.2 8.5 (7.7) 3.5 (3.1) ily distinguishable from those of other South American ungulates by presence of cuneiform- magnum contact in the manus (and loss of lunate- unciform contact), presence of elongate neck of calcaneus with strongly oblique calcaneocuboid facet, sustentacular facet of calcaneus anteropos- teriorly concave, presence of deeply grooved as- tragalar surface of the tibia; absence of medial malleolus of tibia, and a variety of astragalar fea- tures (see Cifelli, 1983, 1985). Macraucheniids are distinguished from adianthids postcranially by relatively large size; absence of fenestrate olec- ranon fossa; absence of pronounced tubercle on dorsolateral ischium, posterior to acetabulum; presence of well-developed third trochanter on fe- mur; and absence of fusion between proximal tib- ia and fibula (Cifelli, 1991). Macraucheniids are distinguished from proterotheriids postcranially by relatively large size; absence of fenestrate olecranon fossa; presence of tridactyl pes with metatarsals II, III, and IV subequal in size (Shock- ey, 1999); absence of tall greater trochanter of fe- mur (not extending far above level of femoral head); and presence of astragalus with broad body, foreshortened neck, and flattened head (Ci- felli & Guerrero Diaz, 1989). Among macrauch- eniids, cramaucheniines are distinguished from macraucheniines by generally smaller size, lack of elongate rostrum, unreduced nasals, absence of dorsally positioned nasal aperture, absence of dor- sally projecting maxillae, unshortened zygomatic arch, absence of closed orbit, absence of imbri- cation in il-pl, absence of fossettes in p3-m3 trigonid basin, absence of cementum, lack of ra- dioulnar fusion, and presence of elongate calca- neus (Soria, 1981). Comments — Although a variety of cranial and dental characters are diagnostic of Litopterna and various subgroups within that clade, no cranio- dental specimens definitely pertaining to the group have been collected from the Chucal For- mation. In contrast, well-pre.served diagnostic postcranial bones are known (see below). No met- ric or morphological evidence suggests that the litoptern postcrania from Chucal pertain to more than one species. The size and morphology of the two calcaneae and the humerus collected at Chu- cal closely resemble other members of the Ma- craucheniidae. Three associated litoptern metatar- sals (MT II-IV) are subequal in size, supporting this identification. Theosodon Ameghino, 1887 Type Species — Theosodon lydekkeri. Included Species — The type, ?r. frenguelli (Colhuehuapian), T. lallemanti, T. fontanae, T. garretorum, T. gracilis, T. patagonicus, T. karai- kensis, T. pozzii (Kraglievich & Parodi, 1931), and T. hystatus (Chasicoan). Postcranial Diagnosis — A cramaucheniine that differs from Cramauchenia in larger size (ap- proximately 20-30% larger in linear dimensions); articular facets of astragalus and calcaneus rela- tively larger than in Cramauchenia (Soria, 1981). CROFT ET AL.: CHUCAL UNGULATES 37 Differs from Coniopternium (including Cali- phrium and Notodiaphorus sensu Cifelli & Soria, 1983) in larger size (approximately 20-25% larg- er in linear dimensions) and in having broader cal- caneus with larger peroneal facet, more robust calcaneal tuber, and tendon grooves (Cifelli & So- ria, 1983). Differs from Pternoconius in larger size (approximately 20-25% larger in linear di- mensions); postcranial elements of Pternoconius are unknown. Differs from Polymorphis in much larger size (approximately twice as large as Po- lymorphis in linear dimensions); we are unaware of any postcrania referable to Polymorphis. Distribution — Early Miocene of Chubut, Ar- gentina, Colhuehuapian SALMA (Simpson, 1935; Soria, 1981); late early Miocene of Patagonia, Ar- gentina, Santacrucian SALMA; ?early middle Miocene of Rio Frias/Cisnes, Chile, Friasian s.s. SALMA (Kraglievich, 1930); middle Miocene of La Venta, Colombia, Laventan SALMA (Cifelli & Guerrero, 1997); ?early late Miocene of Buenos Aires Province, Argentina, Chasicoan SALMA (Cabrera & Kraglievich, 1931). Discussion — Cifelli and Guerrero (1997) sug- gested that Chasicoan "Ci/Z/m/a" laevis (Cabrera & Kraglievich, 1931) and Mayoan "Phoenixau- chenia'" tehuelche (Kraglievich, 1930) may also pertain to Theosodon. Theosodon sp. indet. (Figure 17) ; Referred Specimens — SGOPV 4013, articulat- ed proximal left ulna and radius (Fig. 17 A); SGOPV 4019, partial distal radius; SGOPV 4037, associated limb bones including left and right tib- iae (lacking epiphyses except for the right distal epiphysis), articulated right distal humerus and proximal ulna, right proximal humerus fragment, ?radius fragment, proximal left MT IV, proximal right MC II and III, partial shafts of two other large limb bones (Fig. 17B); SGOPV 4014, partial right calcaneus (Fig. 17C); SGOPV 5054, partial left calcaneus. Description — The litoptern postcrania from Chucal are among the best preserved limb bones of the fauna, but no elements of the pectoral or pelvic girdle are preserved. Most of the Chucal litoptern postcranial elements pertain to a single partial skeleton, SGOPV 4037. The proximal hu- merus from this specimen preserves little mor- phology, save for most of the articular surface of the head. The distal humerus is complete and was found in articulation with a partial proximal ulna. It measures approximately 4.5 cm across the con- dyles, the lateral being broader than the medial. The lateral epicondyle is much larger than the me- dial, though a distinct process projects from the posterior surface of the medial condyle. The olec- ranon fossa is very deep and the supratrochlear fossa shallow; these two structures are not joined by a foramen. Little of the ulna is preserved in SGOPV 4037, except for the articular surface and surrounding area. A pronounced intercondylar ridge bisects the sigmoid notch, separating the larger medial sur- face from the smaller lateral one. It lateral view, the notch approximates a semicircle. A large ru- gose area for attachment of the elbow flexors is present distal to the articular surface, and two small facets mark the proximal articulation of the radius. The olecranon process is incompletely pre- served in SGOPV 4037 but measures more than 6.5 cm in length in SGOPV 4013. The latter spec- imen preserves most of the ulna except its distal end and measures approximately 33 cm in length; the corresponding bone in T. garretorum mea- sures 37.5 cm (Scott, 1910). A similar portion of the radius is preserved in SGOPV 4013. The proximal facet is oblong and gently concave and is reflected for a short distance onto the medial surface. A prominent tuberosity is present just dis- tal to the extension of the articular surface. The ulnar and radial shafts are subequal in length and diameter. Of the manus, only the proximal portions of MC II and III are preserved. These bones are sub- equal in size; the shaft of MC II is approximately 1 8 mm in its greatest diameter. The second meta- carpal has a roughly triangular proximal articular surface for the trapezoid, the latter of which is incompletely preserved in SGOPV 4037. A prom- inent facet is present on the lateral surface for contact with the magnum; it is oriented at an angle of approximately 45° to the long axis of the bone. The medial surface of MC II overlaps the proxi- mal articular surface of MC III, creating one facet that is positioned distally and another that is po- sitioned laterally. Very little of MC III is pre- served, but part of the surface articulating with the distally oriented MC II facet is present. Of the hind limb, the tibia and portions of the pes have been recovered. Specimen SGOPV 4037 includes both tibiae, mostly lacking epiph- yses. The right tibia, which preserves the distal epiphyses, is approximately 24 cm long. Scott (1910) listed the length of the corresponding 38 FIELDIANA: GEOLOGY ViS ■jj^^^^^^HH r^ ^^^Hp^fii- ' ^:0. a MT ^^^^^^^^^^^^^1 ft |p^^;;a|^| imi^^H B^^^^^H^^^^H^I w m i i^^i ^^^^^^^^^^^^^^^^^^^k " .jtjT '_'r?- ^^t ^^^^^L.'^^ ^.^. m F*-^^^^^^^^^^! ^ J ^^H B ■ Fig. 17. Selected postcranial specimens referred to Theosodoii sp. indet. A. SGOPV 4013, proximal left ulna in anterolateral view. B. SGOPV 4037, left tibia, lacking epiphyses, in anterior view. C. SGOPV 4014, partial right calcaneus in superior view. D. SGOPV 5054, partial left calcaneus in superior view. Scale bars = 5 cm in A and B, 2 cm in C and D. CROFT ET AL.: CHUCAL UNGULATES 39 bone in T. garretorum as 27 cm. The astragalar surface of the tibia has two deep grooves for ar- ticulation with the astragalar trochlea and a small lateral facet for articulation with the fibula. The posterior portion of the intercondylar ridge ex- tends much farther distally than the anterior por- tion. Two partial calcanea are represented, SGOPV 4014 and SGOPV 5054; the former includes the poster two-thirds of the element, while the latter preserves the anterior half (Fig. 17C, D). The tu- ber is long, narrow, and dorsoventrally broad. A narrow, strongly convex facet is present for artic- ulation with the fibula. Medial to this, a large, triangular facet is present for articulation with a portion of the ectal facet of the astragalus. This triangular facet is continuous distally with a smaller, oval facet that articulates with the re- mainder of the ectal facet. The astragalar facet on the sustentaculum is also oval, and a large non- articular surface is interposed between this and the other astragalar facets of the calcaneus. The distal surface of the calcaneus is almost saddle shaped and evidently articulated only with the cu- boid; no distal articulation with the astragalus is indicated. The fourth metatarsal (SGOPV 4037) is similar in size to MC II described above. Its proximal articular surface, for articulation with the middle cuneiform, is gently convex and has a prominent plantar hook. A prominent, oblong facet is present on the dorsal side of the medial surface for artic- ulation with MT III. A second facet for articula- tion with MT III is only partly present along the medial surface of the plantar hook. Two small fac- ets are present on the lateral surface for articula- tion with MT V. Comments — As discussed by Cifelli and Guer- rero (1997, p. 298), taxonomy within the Ma- craucheniidae is in a state of "general disrepair." Theosodon has been reported from the Colhue- huapian (Simpson, 1935; Soria, 1981) through the Chasicoan SALMA (Pascual, 1966) — a temporal range of some 10 million years (Flynn & Swisher, 1995) — but many of these reports are based on fragmentary remains of questionable taxonomic utility. The Santacrucian members of Theosodon likely form a clade, but the differences among some of the species are unclear (Cifelli & Guer- rero, 1997). In size and morphology, the Chucal litoptern resembles the various species of Theosodon from the Santa Cruz Formation of Patagonia described by Scott (1910). Owing to the relatively small number of specimens from the Chucal Formation and the lack of well-differentiated species within Theosodon, the taxonomic reliability of these fea- tures currently cannot be assessed. Accordingly, a more tightly circumscribed identification of the Chucal taxon cannot be made at this time. Phylogeny of Mesotheriines To assess the taxonomic affinities of the me- sotheriids from Chucal, we performed a phylo- genetic analysis of mesotheriines based primarily on an emended version of the data matrix of Flynn et al. (in press; see also Cerdeiio & Mon- talvo, 2001). A single character was added to this matrix (#29: Ml proportions), and an additional character state was added to character #16 (state #4: II subtriangular, pointed distally, with two tal- onid sulci) and to character #26 (state #2: P3 ab- sent). The list of characters used in this analysis is presented in Table 7, and the complete data ma- trix is given in Table 8. Two analyses were per- formed, both using Archaeohyracidae as the out- group (following Cerdefio & Montalvo, 2001). In the first analysis, all characters were considered unordered. In the second, selected characters were ordered (indicated in Table 7) on the basis of de- velopmental and metric criteria (e.g., it was as- sumed that a taxon could not proceed from "small" to "large" without passing through "me- dium"). For both analyses, trees were generated using the heuristic search option in PAUP 3.11 for the Macintosh computer. OTUs with multiple character states were considered polymorphic. The first analysis (using unordered characters) re- sulted in nine equally most parsimonious trees, each with a length of 93 steps. The second anal- ysis (using the ordered characters) resulted in 81 equally most parsimonious trees, each with a length of 96 steps. The strict consensus trees for the two analyses have identical topologies at the base and near the crown of the tree, but the first analysis (with unordered character states) exhibits greater resolution in the middle (Fig. 18). We fa- vor the first analysis, as it makes fewer assump- tions regarding character evolution and is better resolved. The relationships in the resulting consensus trees are nearly identical to those of Flynn et al. (in press). Aside from the greater resolution in the unordered analysis, the only other difference is that both analyses here suggest that E. lehmann- 40 FIELDIANA: GEOLOGY Table 7. List of characters and character states used in phylogenetic analysis of mesotheriine relationships. Characters with multiple derived character states were ordered in analysis #2 unless denoted by an asterisk (*). 1.* Rostrum: elongated and low (0); long and tall (1); short, rodent like (2) 2. Lengthening of premaxillae: absent (0); poorly defined (1); very pronounced (2) 3.* Anterior rostral notch (premaxillae): very smoothly concave (0); forming obtuse angle (1); forming acute angle (2); tall, wide, "U" shaped (3) 4. Diastemata (upper and lower): absent or little de- veloped (0); posterolaterally divergent (1); par- allel or gently convergent (2); very convergent, convex (3) 5.* Postorbital process: short (0); long and transverse (1); long, directed posteriorly (2) 6. Suborbital fossa: absent (0); poorly defined (1); well developed (2) 1* Root of zygomatic arch: posteriorly directed from the level of Ml or M2 (0); perpendicular to Ml (1); perpendicular to Ml with biconcave edge (2) 8. Zygomatic plate: absent or hardly developed (0); little developed (1); very developed (2) 9.* Postpalatal notch: narrow, deep, and removed from M3 (0); wide and removed from M3 (1); deep to the level of M3 (2) 10. Occipital notch: absent (0); present (1) 11. Paraoccipital process: short (0), long (1) 12. Dentition: mesodont (0); protohyposodont (1); euhypsodont (2) 13. Upper molar imbrication (imbrication = width from M2 parastyle to M 1 hypocone/width of M 1 posterior face): <1.25 (0); between 1.25 and 1.50 (1);>1.50(2) 14. II : with continuous enamel (0); with labial enam- el only (1) 15. II: obliquely implanted and not procumbent (0); obliquely implanted and procumbent (1); more transverse implantaton and procumbent (2) 16.* II: oval (0); subtriangular, pointed distally, with mesial sulcus (1); wide, rounded, with lingual sulcus (2); wide and compressed labiolingually with two lingual sulci (3); subtriangular, pointed distally, with two lingual sulci (4) 17. I2-I3-C-Pl/i3-c-pl: developed (0); reduced, at times absent in the adult (1); always absent (2) 18. P2: developed (0), reduced (1); absent (2) 19.* P4: subtriangular, short, with central fossette (0); subtriangular, short, without central fossette (1); bilobed, with short and poorly defined lingual plication (2); bilobed, with well-developed and patent lingual plication (3) 20. M1-M2: subtriangular with fossette (0); with an- terior and posterior lobes convergent, enclosing a middle lobe that disappears with wear (I); with persistent middle lobe (2) 21.* M3: subtriangular, with fossette (0); trilobed with little differentiated middle lobe (1); with short middle lobe, surrounded by the other two lobes (2); with the middle lobe less enclosed but still shorter than the others (3) Tabi.f. 7. Continued. 22. 23.^ 24. 25. 26.* 27. 28. 29. 30. 31. 32.* 33. Ratio of il/i2: <1.50 (0); between 1.50 and 2.00 (1); between 2.00 and 3.00 (2), >3.00 (3) Section of il: subcylindrical (0); rounded with little or no lingual sulcus (1); subtriangular with smooth sulcus (2); trapezoidal with well-demar- cated lingual sulcus (3); elliptical with smooth sulcus (4) p3: present (0); absent (1) Lower molars: with fossettes (0); without fos- settes (1) P3: subquadrangular (0); subtriangular (1); ab- sent (2) Premolar ectolophs: with two labial sulci (0); with one labial sulcus ( 1 ) P4 proportions (length/width): <1.50 (0); >1.50 (1) Ml proportions (length/width): <1.30 (0); be- tween 1.30 and 1.75 (1); > 1.75 (2) Anterior and posterior sides of M 1 middle lobe: lingually convergent (0); subparallel (1); lingual- ly divergent (2) Lingual exposure of Ml middle lobe: little or none (0); extensive (1) m2 proportions (length/width); between 1 .60 and 2.30(0); >2.30(1); > 1.60 (2) Two pronounced lingual sulci on m3 talonid: ab- sent (0); present (1) nitschei and E. superans share a common ancestor exclusive of other mesotheriines; in Flynn et al. (in press), these two taxa form a polytomy with a later diverging clade of mesotheriines (node 5 of Fig. 18). In the present analyses, this paring of Eutypotherium species is weakly supported by the loss of a single character state, having the root of the zygomatic arch perpendicular to Ml (charac- ter #7, state 1 >0; variable in T. chasicoensis, lost independently in M. hystatum). The names Mesotheriinae and Mesotheriidae have never before been defined phylogenetically (sensu de Quieroz & Gauthier, 1990). The phy- togeny established here facilitates this undertaking whereby names are explicitly tied to given clades. Consistent with previous usage, we define Me- sotheriidae as all notoungulates more closely re- lated to Mesotherium than to Hegetotherium, Ar- chaeohyrax, or Oldfieldthomasia (essentially equivalent to node 1 of Fig. 1 8 plus its stem), and we define Mesotheriinae as all mesotheriids more closely related to Mesotherium than to Trachy- therus (node 2 of Fig. 1 8 plus its stem). Following this definition of the name Meso- theriinae, the phylogenetic analyses place the three Chucal mesotheriids at the base of the Mesotheriinae, with Eotypotherium as its earliest diverging member. Eotypotherium is a member of CROFT ET AL.: CHUCAL UNGULATES 41 Table 8. Character-taxon matrix for phylogenetic analysis of mesotheriine relationships. 1 11111 11112 22222 22223 333 12345 67890 12345 67890 12345 67890 123 Archaeohyracidae 00000 00000 1 20111 00000 1 01011 00000 00000 00007 701 Trachytheriinae 10101 11101 1 22222 10100 1 20211 00000 000 Eutypotherium lehmannnitschei 21?1? 10211 72001 10010 100 Eutypotheriuin superans 2121? 10211 72001 22212 37770 10010 170 Typotheriopsis chasicoensis 21722 20201 1 21201 12001 1 12071 22212 21211 2 27777 11011 100 Typotheriopsis silveyrai 21222 72212 11011 170 Pseudotypotherium subinsigne 21721 21201 1 71217 12201 72222 3 77732 22311 3 32711 10121 110 Pseudotypotheriiim exiguum 27771 722 7 7 10121 110 Mesotherium hystatuin 21231 70201 12271 72232 37711 1012? 110 Mesotherium cristatuin 22332 02221 12202 22232 33311 10122 110 Mesotherium pachy gnat hum 21722 01221 12202 22232 33711 10122 110 Mesotherium maendrum 22232 01221 12202 32232 33411 10122 110 Plesiotypotherium achirense 21122 21211 12001 22212 2 12212 30211 10011 100 Microtypotherium choquecotense 2??1? 71201 72001 27771 10010 020 Caragua New Taxon 21127 11221 12101 22222 21211 10011 101 A Ititypotherium paucidens 27717 11277 72001 42212 30111 21010 000 A Ititypotherium chucalensis 77717 7 7 727 72007 42212 20111 11000 000 Eotypotherium chico 7 7777 11117 72007 12202 20111 00010 000 the Mesotheriinae based on its possession of nu- merous characters typifying the clade (node 2 of Fig. 18), including the presence of a poorly de- fined suborbital fossa and small infraorbital fora- men (character #6, state 1 ; DELTRAN optimiza- tion only, lost in most Mesotherium species), root of zygomatic arch perpendicular to M 1 (character #7, state 1 ; lost in Eutypotherium clade and M. hystatum, variably present in T. chasicoensis). hypselodont (euhypsodont) dentition (character #12, state 2), I2-I3-C-Pl/i3-c-pl always absent (character #17, state 2), P2 absent (character #18, state 2), M1-M2 with persistent middle lobe (character #21, state 2), M3 with short middle lobe surrounded by the other two lobes (character #21, state 2), p3 absent (character #24, state 1), lower molars without fossettes (character #25, state 1), and Ml proportions (length/width) >1.30 (character #29, state 1). A. paucidens and A. chucalensis are united by the unique presence of a subtriangular 1 1 that is pointed distally, with two lingual sulci (character #16, state 4) and the presence of a single labial sulcus on premolar ectolophs (character #27, state 1). Within the Mesotheriinae, Altitypotherium forms a polytomy (node 3 of Fig. 18) with Mi- crotypotherium and a clade including all other lat- er diverging mesotheriines (node 4 of Fig. 18). Node 3 is diagnosed by the presence of a short. rodentlike rostrum (character #1, state 2; un- known in Eotypotheriwn and A. chucalensis); presence of posterolaterally divergent diastemata (character #4, state 1; unknown in Eotypothe- rium); presence of a well-developed zygomatic plate (character #8, state 2); presence of short, subtriangular P4, without a central fossette (char- acter #19, state 1); and presence of subtriangular P3 (character # 26, state 1; absent in A. pauci- dens). Altitypotherium and Microtypotherium are ex- cluded from membership in the clade stemming from node 4 (Fig. 18) by the lack of the following derived character states: wide, rounded II with a lingual sulcus (character #16, state 2; unknown in T. silveyrai and Pseudotypotherium); il with sub- triangular section and smooth lingual sulcus (character #23, state 2; unknown in Microtypo- therium, P. exiguum, M. hystatum, and M. pach- ygnathum); and extensive lingual exposure of Ml middle lobe (character #31, state 1). Age and Faunal Comparisons Although the preliminary report on the Chucal Fauna suggested that it might be as old as San- tacrucian (late early Miocene) or as young as 42 FIELDIANA: GEOLOGY Archaeohyracidae Trachytheriinae Eotypothehum chico Altitypotherium chucalensis Altitypothehum paucidens Microtypotherium choquecotense Eutypotherium lehmannnitschei Eutypotherium superans Caragua New Taxon Plesiotypotherium achirense Typotheriopsis chasicoensis Typothehopsis silveyrai Pseudotypotherium subinsigne Pseudotypothehum exiguum Mesotherium hystatum Mesotherium cristatum Mesotherium maendrum Mesotherium pachygnathum Archaeohyracidae Trachytheriinae Eotypothehum chico Altitypothehum chucalensis Altitypothehum paucidens Microtypothehum choquecotense Eutypothehum lehmannnitschei Eutypothenum superans Typothehopsis chasicoensis Typothehopsis silveyrai Caragua New Taxon Plesiotypotherium achirense Pseudotypothehum subinsigne Pseudotypothehum exiguum Mesothenum hystatum Mesothehum chstatum Mesothenum maendrum Mesothehum pachygnathum Fig. 18. Strict consensus trees of mesotheriine relationships based on the phylogenetic analyses in the present study. A. Strict consensus of nine trees resulting from the analysis with all characters unordered; length = 82, CI = 0.79, RI = 0.83, RC = 0.66. B. Strict consensus of 81 trees resulting from the analysis with some characters ordered (see Table 7); length = 88, CI = 0.74. RI = 0.81, RC = 0.60. In both trees, node 1 is Mesotheriidae, node 2 is Mesotheriinae, and nodes 3-5 are unnamed clades discussed in the text. CROFT ET AL.: CHUCAL UNGULATES 43 Chasicoan (late Miocene) (Flynn, Croft, et al., 2002), new dates from a subsequent report (Bond & Garcia, 2002) have restricted the age the fauna to within a 4- to S-miUion-year-long interval (Ta- ble 9; see also Charrier et al., 2002). The over- lying Quebrada Macusa Formation has been dated at 17.4 ± 0.4 Ma (Bond & Garcia, 2002), and the underlying Lupica Formation has been dated at 21.7 ± 0.8 Ma at Chucal (Riquelme, 1998) and at 1 8.79 ± 0. 1 1 Ma in the surrounding area (Wor- ner et al., 2000). Based on the radioisotopic dates from the Chucal region and the most recent re- vision of SALMAs (Flynn & Swisher, 1995), this interval overlaps with or slightly pre-dates the Santacrucian SALMA (-16.3-17.5 Ma). There is a remote possibility, however, that it coincides with the Colhuehuapian SALMA (estimated to be 719-21? Ma since the Colhuehuapian is uncon- strained within the 24.5-1 7. 5-Ma gap between the dated Deseadan and Santacrucian SALMAs), as no such faunas have been dated. Paleomagnetic correlations for Chucal are not yet available, but samples for a preliminary analysis are currently being analyzed. These uncertainties notwithstanding, based on biochronologic evidence, the Chucal Fauna al- most certainly pertains to the Santacrucian SAL- MA; Nesodon likely is an indicator taxon of the Santacrucian SALMA (Croft, Radic, et al., 2003), as is Neoreomys (A. Kramarz, pers. comm.; Table 12), and Hegetotherium mirabile is recorded only in Santacrucian deposits (see above). Taking into account the geochronologic constraints on the Chucal Fauna, the Santacrucian SALMA may ex- tend slightly older than previously believed, at least in intermediate latitudes. This is not entirely unexpected, as most high-latitude strata bearing Santacrucian assemblages are underlain by either an unconformity and/or marine strata, thus pre- cluding determination of the complete duration of this SALMA; the youngest dated assemblages that definitively pre-date the Santacrucian are —25.5- Ma Deseadan sites near Salla, Bolivia (Kay et al., 1998) and the -20.1 -Ma? Colhuehuapian Rio Las Leilas assemblage from central Chile (Flynn et al., 1995). The greater resemblance of the Chucal Fauna to lower Santacrucian levels compared to upper ones (Tables 10 and 11) lends additional support to the relative antiquity of the Chucal Fauna. Paleomagnetic analyses curtently under way may further constrain the age of the Chucal Fauna and clarify the onset of the Santacrucian SALMA. Despite several shared taxa, there are profound differences between the ungulate fauna of Chucal and those from typical Santacrucian localities in Argentina and southern Chile (Table 10). Simpson coefficients, comparing Chucal to these faunas, range only from 43-57 (Table 11). These differ- ences arise primarily because mesotheriids — con- stituting one-third of the ungulate "genera" at Chucal — have never before been recorded in San- tacrucian-aged strata. This lack of Santacrucian mesotheriids elsewhere represents but a small por- tion of an even larger middle Cenozoic gap in the fossil record of mesotheriids in higher latitudes: none has been recorded there between the Desea- dan and Mayoan SALMAs (early to middle Mio- cene), a span of some 13 million years (Flynn & Swisher, 1995; Croft, Flynn, & Wyss, 2003). In contrast, mesotheriines are characteristic compo- nents of Bolivian Miocene faunas (Table 10) and a very large fraction of isolated Miocene verte- brate specimens collected in northern Chile and Bolivia are referable to mesotheriines (Marshall & Sempere, 1991; Flynn et al., in press). Another conspicuous difference between the Chucal Fauna and typical Santacrucian faunas is the complete absence of interatheriid notoungu- lates in northern Chile. In this respect, Chucal re- sembles other middle-latitude Miocene faunas of Bolivia (i.e., Nazareno, Quebrada Honda, and Achiri), in which interatheriids also are rare and/ or absent (Table 10). Although it is certainly pos- sible that interatheriids were present at Chucal but remain unsampled, the lack of specimens refer- able to this clade from among more than 300 identified specimens at least speaks to the rarity of this group, if they existed in this region at that time; it is difficult to envision a taphonomic bias that would preserve other closely related and sim- ilarly sized typothere notoungulates (i.e., hege- totheriines and mesotheriines) to the exclusion of interatheriids. The absence of interatheriids is par- ticularly striking given their abundance not only in the Santa Cruz Formation of Argentina (Sin- clair, 1909) and the Santacrucian-aged Pampa Castillo Fauna of southern Chile (Flynn, Novacek, et al., 2002) but also in nearly all Oligocene through early Miocene faunas of South America. The close geographic proximity of the faunas in which interatheriids are rare or absent (Bolivian and Chilean Altiplano) suggests that some bio- geographic and/or ecological factor served to lim- it or exclude interatheriids from at least the west- ern portion of these intermediate latitudes (east- ern, noncordilleran faunas of this age are un- known) during much of the Miocene. In 44 FIELDIANA: GEOLOGY u 60 c D c cs :s .1 ^ « 08 O 1 = d c ._ s s d. d. a 3 o C C m u <;> ^ cs es c« -c -s: s .E .S c c ^ o o Ji><^ t-^ •a -o •2 -S o o ^ ^ X X ^ ^ o o FF 5: 5; cs ■V e o X cs ■o cs u .o 3 o ■^ cs cs 5 cs -^ «"• v« oe ^® +1 o c ■^ cs cs cs 5 " — ^ C/5 •^ ^ ^• (^ . o -o OQ CROFT ET AL.: CHUCAL UNGULATES 45 4> r a> K5 lis I cr ~ Si .2 So c^ «5 ^^ ^ S c U g g ^ ~ ?.2 ® es ti, o a «i OS .5 ex) !n C8 a> c/3 OS es ON (U 3 ;^ u « C !« 0) -a n u a; ^ ii -^ ill CO i2 XI n 3 C/3 -M ^ e ,.o i- U C 09 ^^ ■a 2 T3 IB E « S o ■a "o C U C5 •2 « ^ a s 3 5 B a s a U A) ^ ^ ^ ^ t C ^^-i:-!: a; 1 ^0,11, s s 5 i s: -a >: o S S ^ c S J2 i^ a Id'^ « ^ "i s^ t« c ^ t« g erium sp. ni rium . 5 S a.s .s '4) -s: •..„ ■^ 2 ium ium eriu sodon ; inother inother oth ius 'the o -c -C -C Q erat erat otyp a O 00 3) ^ ^ 1^1 1:::: t2 !< C 5" ^ '^ '^ 1 ^ 1 ^^a. a,ua: r- ^ o lile (4 ovace pers. o. . s a.s t/) C c Si s Flynn, (2002 11 II s ^ g a c •a f c^- '"" '^ 'S K Si C o c o ^ ^ H « "3 II si i < O r- HZ "5 « s 5 .a a 'C ^ C "^ -5 ■c 5 !: ^ o ^ f^ ty Is .to •S c§-S S il 5 .a 'C 'c i: s ^ -s: ^ ■^ a -J :: h^ !^ l« 5 a a ^ ^ M-s -s: -s: -2 c S -2 S 3 a ? 2. -s; -s: .5 (£ ^ . o c 5-- t/5 D. a t" -I a "^ S s s ^ ^ ■5 o P e^ .£ o o a a 2 2 to ■s? < 46 FIELDIANA: GEOLOGY Table 11. Ungulate faunal resemblance indices (Simpson coefficients) among selected low-, intermedi- ate-, and high-latitude Miocene faunas of South Amer- ica. Simpson coefficients were calculated as (# of shared genera/# of genera in the smaller fauna)* 100. Table 12. Revised taxonomic list of the Chucal fau- na. Pampa Lower La Chu- Cas- Santa Venta cal tillo Cruz Chucal 14 Pampa Castillo 11 57 — Lower Santa Cruz 8 57 100 — Upper Santa Cruz 43 89 100 Patagonia, interatheriids declined dramatically in both diversity and abundance after the Santacru- cian (M. Reguero, pers. comm.), although they evidently remained a dominant component of low-latitude faunas (i.e., La Venta, Colombia) through at least the middle Miocene (Stirton, 1953; Kay et al., 1997). Many of the nonungulate elements of the Chu- cal Fauna are equally unusual in comparison to contemporaneous Patagonian faunas (e.g., very low rodent diversity, low abundance of nonchin- chilline rodents, low glyptodontid abundance, ab- sence, and/or low abundance of sloths; Table 12), further supporting the existence of a significant biogeographic and/or ecological barrier between the two regions during the middle Cenozoic. Discussion and Conclusions The revised faunal list for Chucal (Table 12) now includes 16 mammal species. The present list represents a two-thirds increase in the number of species relative to the preliminary faunal list (based on a single field season) presented in Flynn, Croft, et al. (2002); this is due to the ad- dition of taxa recognized following a second field season (September 2001) and newly identified taxa represented only by postcrania. Also, an in- determinate, possibly new toxodontid was de- scribed by Bond and Garcia (2002; see discussion above). Although the nonungulates from Chucal have yet to be described, preliminary analyses in- dicate that, for the most part, they pertain to taxa recorded in other Santacrucian faunas, consistent with the ungulate-based biochronologic age esti- mate and with the geochronologic constraints of bracketing high-precision radioisotopic dates. As is true for the ungulates, however, several species Mammalia Notoungulata Toxodontidae Toxodontidae ?new taxon (Palyeidodon [?] sp. of Bond and Garcia. 2002) Nesodontinae Nesodon imbricatus Adinotherium sp. indet. Mesotheriidae Mesotheriinae Altitypotherium paucidens (Undescribed Tax- on A, aff. Microtvpotherium sp. of Flynn, Croft, et al., 2002) Altitypotherium chucalensis (Undescribed Taxon B, aff. Microtypotherium sp. of Flynn, Croft, et al.. 2002) Eotypotherium chico (Undescribed Taxon C, aff. Typotheriopsis sp., Plesiotypotherium sp., or Eutypotherium sp. of Flynn, Croft, et al., 2002) ' Hegetotheriidae Hegetotheriinae Hegetotherium cf. H. mirabile (cf. Pseudohe- getotherium torresi of Flynn, Croft, et al., 2002) Litopterna Macraucheniidae Cramaucheniinae Theosodon sp. indet. Rodentia Chinchillidae Chinchillinae Undescribed taxon Dasyproctidae Neoreomys sp. nov. Octodontoidea Undescribed taxon Xenarthra Cingulata Dasypodidae Euphractini sp. indet. Glyptodontidae sp. indet. Peltephilidae Peltephilus sp. Mammalia incertae sedis Tiny Undescribed Taxon 1 Tiny Undescribed Taxon 2 from Chucal are unique compared to other San- tacrucian faunas, including the earliest known chinchilline chinchillid (Flynn, Croft, et al., 2002) and a new basal octodontoid rodent. The presence of these distinctive taxa (both ungulates and other mammals) at Chucal demonstrates that significant provinciality had begun to develop among South American faunas by at least early Miocene time, possibly attributable to Andean tectonics or other paleoenvironmental factors. CROFT ET AL.: CHUCAL UNGULATES 47 Table 13. Body masses of mesotheriines estimated by regression of Ml length on body mass for extant ungulates (based on the equation: ln[mass in g] = 3.09*ln[Ml length in mm] +1.21 [Damuth, 1990]) and caviomorph rodents (based on the equation ln[mass in g] = 3.09*ln[Ml length in mm] + 3.08 [Croft, 2000]). Holotypes are indicated by an asterisk (*) following the specimen number. Ungulate Rodent body body Ml mass mass length estimate estimate Taxon Specimen Age (mm) (kg) (kg) Eotypothehum chico SGOPV 5157* late early Miocene 9.5 3.52 22.85 Altitypotherium paucidens SGOPV 4038* late early Miocene 12.5 8.22 53.34 Altitypotherium chucalensis SGOPV 4100* late early Miocene 10.4 4.66 30.22 Microtypothehum choquecotense GB 002* middle Miocene 11.6 6.53 42.35 Plesiotypotherium achirense MNHN ACH 26* middle Miocene 17.0 21.26 137.95 Eutypotherium lehmannnitschei MLP 12-1701* late middle Miocene 14.6 13.28 86.19 Eutypotherium superans MACN 11079* late middle Miocene 14.1 11.93 77.40 Caragua New Taxon SGOPV 4004* late middle Miocene 16.5 19.39 125.79 Typotheriopsis chasicoensis MLP 12-1666* early late Miocene 17.3 22.44 145.61 Typotheriopsis silveyrai MLP 36-XI-10-2* middle late Miocene 16.9 20.88 135.46 IPseudotypotherium sp. MACN 8010 late Miocene-early Pliocene 23.8 60.13 390.17 IPseudotypotherium sp. MACN 1111 late Miocene-early Pliocene 19.2 30.97 200.92 Mesothehum cristatum MACN 2036 late Pliocene-early Pleistocene 21.8 45.85 297.48 Mesotherium pachygnathum MACN 1665 late Pliocene-early Pleistocene 24.4 64.94 421.368 Mesotherium maendrum MACN 2648* late Pliocene-early Pleistocene 16.1 17.97 116.61 Of the eight species of ungulates present at Chucal, seven (all but Theosodon) have hypselo- dont (rootless, ever-growing) or very hypsodont (very high-crowned) cheek teeth. In modern mammals, hypsodonty is highly correlated with dietary preference and can be used to distinguish animals feeding on vegetation in open habitats (which have more hypsodont teeth) from those feeding in closed habitats (Janis, 1988, 1995; Wil- liams & Kay, 2001; Janis et al., 2002). This high proportion of hypsodont taxa suggests that the mammals of the Chucal Fauna lived in a relatively dry, open environment, especially as compared to roughly contemporaneous faunas from elsewhere in South America (Table 11). Two of the three rodent species at Chucal for which cheek teeth are known are also very hypsodont or hypselodont, a proportion congruent with that of the ungulates. The single hypselodont rodent taxon (the chin- chilline) dominates the rodent fauna numerically, with only a single specimen representing each of the other two taxa. The phylogenetic analysis presented here sug- gests that the smallest mesotheriines {Eotypo- thehum, Altitypotherium, and Microtypotherium; Table 13) represent early diverging members of the clade. Although the size of the most recent common ancestor (IVIRCA) of mesotheriines and trachytheriines cannot be reliably estimated with- out a more detailed knowledge of the topology within Trachytheriinae, the presence of four small species near the base of Mesotheriinae (the small- est being the nearest outgroup to all the remaining mesotheriines) suggests that the initial diversifi- cation of mesotheriines took place at diminutive body size. The latest-occurring mesotheriines (Mesotherium spp.) are also the largest, but there does not appear to be a progressive or stepwise increase in body size throughout the group's evo- lution (Table 13); most mesotheriines are similar in size from the late middle Miocene (approx. 12 Ma) until the beginning of the Pliocene (approx. 5 Ma). Of the two regression equations used to estimate mesotheriine body masses, the ungulate regression appears to be more appropriate for these notoungulates, even though mesotheriines more closely resemble extant caviomorph rodents than artiodactyls or perissodactyls; mass estimates based on rodents seem unreasonably high. Perhaps the most striking aspect of the Chucal Fauna, especially given its age, is the abundance and diversity of mesotheriine notoungulates. In 48 FIELDIANA: GEOLOGY terms of abundance, Chucal is similar to other early to middle Miocene faunas of Bolivia, where mesotheriines constitute a significant proportion of identified specimens (e.g., Cerdas; MacFadden et al., 1995). This is in stark contrast to both high latitudes (Argentina) and the lower latitudes (e.g.. La Venta, Colombia; Kay et al., 1997) of South America, where early to middle Miocene mesotheriines are uncommon or absent. In terms of mesotheriine diversity, the Chucal Fauna is unique, even compared to Bolivian faunas also containing mesotheriines; no other South Ameri- can locality records three contemporaneous spe- cies of this clade. Bracketing dates indicate the age of the Chucal mesotheriines is at least 17.5 Ma (or ~ 17-18 Ma taking into account the un- certainty of the date from levels overlying the fau- na), pre-dating the earliest Patagonian mesotheri- ines (Eutypotherium of the Mayoan SALMA) by more than 5 million years. Although slightly younger than the Chucal specimens, other Boli- vian taxa also significantly pre-date Eutypother- ium; both Microtypotherium cf. M. choquecotense and Plesiotypotherium minor have been collected from the Quehua Formation near Cerdas, Bolivia, (16.5-15.5 Ma; MacFadden et al., 1995). Consis- tent with the conclusion reached by Reguero and Castro (2001) based on trachytheriine relation- ships, the presence of the earliest members of the Mesotheriinae in northern Chile and Bolivia, in- cluding early diverging species, suggests that the intermediate latitudes of South America served as a center of diversification for this clade of no- toungulates (Croft, Flynn, & Wyss, 2003). Only later did this clade disperse to higher latitudes, not becoming common there until some 10 million years after the group's origin. Although the ear- liest taxa occur in the modem high-altitude Alti- plano or Puna, the potential role of elevation in this diversification is uncertain; paleoelevations at these latitudes certainly were lower than modem elevations, and definitively low-elevation faunas of this age are unknown from the midlatitudes. Acknowledgments This study was greatly facilitated by access to specimens at the Museo de La Plata (courtesy of Rosendo Pascual and Marcelo Reguero) and the Museo Argentino de Ciencias Naturales "Bernar- dino Rivadavia" (courtesy of Jose Bonaparte and Alejandro Kramarz). The Chucal specimens dis- cussed in this paper were skillfully prepared by Lorie Barber, Matt Brown, Akiko Shinya, Wendy Taylor, and Connie Van Beek. We thank Ralph Hitz, Rick Madden, Emesto Rodrigo Paz, Mar- celo Reguero, and Bmce Shockey for helpful dis- cussions on notoungulates and South American paleofaunas; Marcelo Reguero and Bruce Shock- ey also helped improve this work by providing critical reviews. Most of the photographs were ex- ecuted by Mark Widhalm, and line drawings were created by Marlene Donnelly. Our collaborators on this project, Reynaldo Charrier and Gerard Herail, have been unstinting in their continuing support, and we gratefully acknowledge help in the field from Gabriel Carrasco, Andres Charrier, the Flynn family (Alison, Peter, Rachel), Claude Herail, and Sergio Villagran. A crew from the Quiborax Mine at Salar de Surire deserves special thanks for pulling both of our mired field vehicles out of a small streambed some distance from our camp. Funding for this study was provided by the National Geographic Society (Grant 5371-94), with additional support from our home institu- tions, U.S. NSF grant DEB 9317943, and FON- DECYT Chile. Our paleontological investigations in Chile have been undertaken in collaboration with the Museo Nacional de Historia Natural (Dra. Maria Eliana Ramirez, Directora; Daniel Frasinetti, Jefe, Paleontologia) and the Consejo de Monumentos Nacionales (Sr. Angel Cabezas, Se- cretaria Ejecutivo), Santiago. Literature Cited Ameghino, p. 1891. Observaciones cn'ticas sobre los mamfferos eocenos de la Patagonia austral. Revista Argentina de Historia Natural, 1: 328-380. . 1894. Enumeration synoptique des espfeces mammiferes fossiles des formations eocenes de Pata- gonie. Boletin de la Academia Nacional de Ciencias, Cordoba, 13: 259-452. Bond, M., and M. Garci'a. 2002. Nuevos restos de tox- odonte (Mamalia, Notoungulata) en estratos de la For- macion Chucal, Mioceno, Altiplano de Arica, norte de Chile. Revista Geologica de Chile, 29(1): 81-91. Bond, M., and G. Lopez. 1997. Los Hegetotheriinae (Notoungulata, Hegetotheriidae) de la Formacion Ar- royo Chasic6 (Mioceno superior), Provincia de Buen- os Aires, Argentina. Ameghiniana, 34(4): 533. BoNDESio, R, J. Rabassa, R. Pascual, M. G. Vucetich, AND G. J. SciLLATO Yane. 1980. La Formacion Colldn Cura de Pilcaniyeu Viejo y sus alrededores (Rio Ne- gro, Republica Argentina). Su antigUedad y las con- diciones ambientales segiin su distribucion, su lito- genesis y sus vertebrados. Actas del Segundo Congre- so Argentino de Paleontologi'a y Bioestratigrafia y CROFT ET AL.: CHUCAL UNGULATES 49 Primer Congreso Latinoamerica de Paleontologia, Tomo 3: 85-99. BoRDAS, A. F. 1939. Diagnosis sobre algunos mamiferos de las capas con Colpodon del Valle del Ri'o Chubut (Repiiblica Argentina). Physis, 14: 413-433. Cabrera, A., and L. Kraglievich. 1931. Diagnosis previas de los ungulados fosiles del Arroyo Chasico. Notas Preliminares del Museo de La Plata, Tomo 1: 107-113. Cerdeno, E., and M. Bond. 1998. Taxonomic revision and phylogeny of Paedothehum and Tremacyllus (Pa- chyrukhinae, Hegetotheriidae, Notoungulata) from the late Miocene to Pleistocene of Argentina. Journal of Vertebrate Paleontology, 18(4): 799-811. Cerdeno, E., and C. I. Montalvo. 2001. Los Meso- theriinae (Mesotheriidae, Notoungulata) del Mioceno Superior de La Pampa, Argentina. Revista Espailola de Paleontologia, 16(1): 63-75. Charrier, R., a. Chavez, S. Elgueta, G. Herail, J. J. Flynn, D. a. Croft. A. Wyss, and M. Garci'a. 2002. Rapid tectonic and paleogeographic evolution: The Chucal Anticline, Altiplano of Arica, northern Chile. 5th International Symposium of Andean Geodynam- ics, Universite Paul Sabatier, Toulouse, France, pp. 137-140. Charrier, R., G. Herail, J. J. Flynn, R. Riquelme, M. Garcia, D. A. Croft, and A. Wyss. 2000. El cordon Chapiquifia-Belen en el borde occidental del Altiplano Chileno: Significado paleogeografico y contexto tec- tonico regional. IX Congreso Geologico Chileno, Puerto Varas, Actas 1: 763-767. Charrier, R., N. Munoz, and S. Palma-Heldt. 1994. Edad y contenido paleofloristico de la Formacion Chucal y condiciones paleoclimaticas para el Oligo- ceno Tardio-Mioceno Inferior en el Altiplano de Ari- ca, Chile. VII Congreso Geologico Chileno, Concep- cion, pp. 434-437. Charrier, R., N. Munoz, A. R. Wyss, J. J. Flynn, and G. Herail. 1994. Hallazgo de un humero de Toxo- donte (Mammalia) en la Formacion Chucal (Oligo- ceno Tardio — Mioceno Inferior), en el Altiplano de Arica, Chile. VII Congreso Geologico Chileno, Con- cepcion, Actas 1: 430-433. Chavez, A. 2001. Evolucion paleoambiental y sedimen- tacion sintectonica durante el Mioceno en el sector del cerro Chucal (Cordillera Occidental, Region de Tara- paca: 69° 10' 0-18°45' S), y su relacion con la estruc- turacion del Altiplano. Thesis, Universidad de Chile, Santiago, 73 pp. CiFELLi, R. L. 1983. Eutherian tarsals from the Late Pa- leocene of Brazil. American Museum Novitates, 2761: 1-31. . 1985. South American ungulate evolution and extinction, pp. 249-266. In Stehli, F G., and S. D. Webb, eds.. The Great American Biotic Interchange. Plenum Press, New York. . 1991. A new adianthid litoptern (Mammalia) from the Miocene of Chile. Revista Chilena de His- toria Natural, 64: 119-125. . 1993. The phylogeny of the native South Amer- ican ungulates, pp. 195-216. In Szalay, F S., M. J. Novacek, and M. C. McKenna, eds.. Mammal Phy- logeny: Placentals. Springer- Verlag, New York. CiFELLi, R. L., and J. Guerrero. 1997. Litopterns, pp. 289-302. In Kay, R. F, R. H. Madden, R. L. Cifelli, and J. J. Flynn, eds.. Vertebrate Paleontology in the Neotropics: The Miocene Fauna of La Venta, Colom- bia. Smithsonian Institution Press, Washington, D.C. Cifelli, R. L., and J. Guerrero Diaz. 1989. New re- mains of Prothoatherium colombianus (Litopterna, Mammalia) from the Miocene of Colombia. Journal of Vertebrate Paleontology, 9(2): 222-231. Cifelli, R. L., and M. F Soria. 1983. Notes on Desea- dan Macraucheniidae. Ameghiniana, 20: 141-153. Croft, D. A. 2000. Archaeohyracidae (Mammalia, No- toungulata) from the Tinguiririca Fauna, central Chile, and the evolution and paleoecology of South Ameri- can mammalian herbivores. PhD diss.. University of Chicago, Chicago, Illinois, 311 pp. Croft, D. A., J. J. Flynn, and A. R. Wyss. 2003. Di- versification of mesotheriids (Mammalia: Notoungu- lata: Typotheria) in the middle latitudes of South America. Journal of Vertebrate Paleontology, 23(Sup- plement 3): 43A. Croft, D. A., J. P. Radic, E. Zurita, R. Charrier, J. J. Flynn, and A. R. Wyss. 2003. A Miocene toxodontid (Notoungulata) from the sedimentary series of the Cura-Mallin Formation, Lonquimay, Chile). Revista Geologica de Chile, 30(2): 285-298. Damuth, J. 1990. Problems in estimating body masses of archaic ungulates using dental measurements, pp. 229-253. In Damuth, J., and B. J. MacFadden, eds., Body Size in Mammalian Paleobiology: Estimation and Biological Implications. Cambridge University Press, Cambridge, United Kingdom. Flynn, J.J. 1990. Cenozoic Andean paleoenvironments and tectonic history: Evidence from fossil mammals. 5th International Symposium of Andean Geodynam- ics, Universite Paul Sabatier, Toulouse, France, pp. 215-218. Flynn, J. J., D. A. Croft, R. Charrier, G. Herail, and A. R. Wyss. 2002. The first Cenozoic mammal fauna from the Chilean Altiplano. Journal of Vertebrate Pa- leontology, 22(1): 200-206. Flynn, J. J., D. A. Croft, R. Charrier, and A. R. Wyss. In press. A new specimen and taxonomic re- vision of mesotheriids (Mammalia, Notoungulata, Ty- potheria) from northern Chile: Implications for age and of strata on the western flank of the Altiplano. Journal of South American Earth Sciences. Flynn, J. J., M. J. Novacek, H. E. Dodson, D. Fras- sinetti, C. McKenna, M. A. Norell, K. E. Sears, C. C. Swisher, and A. R. Wyss. 2002. A new fossil, mammal assemblage from the southern Chilean An- des: Implications for geology, geochronology, and tec- tonics. Journal of South American Earth Sciences, 15(3): 285-302. Flynn, J. J., and C. C. Swisher, III. 1995. Cenozoic South American Land Mammal Ages: Correlation to global geochronologies, pp. 317-333. In Berggren, W. A., D. V. Kent, M.-P Aubry, and J. Hardenbol, eds., Geochronology, Time Scales, and Global Stratigraphic Correlation. SEPM (Society of Sedimentary Geology) Special Publication No. 54. Flynn, J. J., and A. R. Wyss. 1990. Recent advances in 50 FIELDIANA: GEOLOGY South American mammalian paleontology. Trends in Ecology and Evolution, 13(1 1): 449-454. Flynn, J. J., A. R. Wyss, R. Charrier, and C. C. Swish- er, III. 1995. An Early Miocene anthropoid skull from the Chilean Andes. Nature, 373: 603-607. Francis, J. C. 1965. Los generos de la subfamilia Me- sotheriinae (Typotheria, Notoungulata) de la Repiib- lica Argentina. Boletin del Laboratorio de Paleonto- logfa de Vertebrados, 1(1): 7-31. Griffiths, G. C. D. 1976. The future of Linnaean no- menclature. Systematic Biology, 25: 168-173. Herail, G., J. Oller, R Baby. M. Bonhomme, and R SoLER. 1996. Strike-slip faulting, thrusting and related basins in the Cenozoic evolution of the southern branch of the Bolivian Orocline. Tectonophysics, 259: 201-212. Janis, C. M. 1988. An estimation of tooth volume and hypsodonty indices in ungulate mammals, and the cor- relation of these factors with dietary preference, pp. 371-391. In Russell, D. E., J. R Santoro, and D. Sig- ogneau-Russell, eds., Teeth Revisited: Proceedings of the Vllth International Congress of Dental Morphol- ogy. Memoirs de Musee d'Histoire Naturelle, Paris, Serie C, Volume 53. . 1995. Correlations between craniodental mor- phology and feeding behavior in ungulates: Recipro- cal illumination between living and fossil taxa, pp. 76-98. In Thomason, J. J., ed.. Functional Morphol- ogy in Vertebrate Paleontology. Cambridge University Press, New York. Jams, C. M., J. Damuth, and J. M. Theodor. 2002. The origins and evolution of the North American grassland biome: The story from the hoofed mammals. Palaeo- geography, Palaeoclimatology, Palaeoecology, 177(1- 2): 183-198. Kay, R. F, B. J. MacFadden, R. H. Madden, H. San- DEMAN, AND F. Anaya. 1998. Revised age of the Salla beds, Bolivia, and its bearing on the age of the De- seadan South American Land Mammal "Age." Jour- nal of Vertebrate Paleontology, 18(1): 189-199. Kay, R. F, R. H. Madden, R. L. Cifelli, and J. J. Flynn. 1997. Vertebrate Paleontology in the Neotrop- ics: The Miocene Fauna of La Venta, Colombia. Smithsonian Institution Press, Washington, D.C., 592 pp. Kraglievich, L. 1930. La Formacion Friaseana del Rio Fn'as, etc., y su fauna de mamiferos. Physis, 10: 127- 166. -. 1934. La antigiiedad pliocena de las faunas de Monte Hermoso y Chapadmalal, deducidas de su comparacion con las que le precedieron y sucedieron. Montevideo: El Siglo Ilustrado, 938: 1-136. MacFadden, B. J., F Anaya, H. Perez, C. W. Naeser, P. K. Zeitler, and K. E. Campbell, Jr. 1990. Late Cenozoic paleomagnetism and chronology of Andean basins of Bolivia: Evidence for possible oroclinal bending. Journal of Geology, 98: 541-555. MacFadden, B. J., F Anaya, and C. C. Swisher, III. 1995. Neogene paleomagnetism and oroclinal bending of the central Andes of Bolivia. Journal of Geophys- ical Research, 100(B5): 8153-8167. Madden, R. H. 1990. Miocene Toxodontidae (Notoun- gulata, Mammalia) from Colombia, Ecuador and Chile. PhD diss., Duke University, Durham, North Carolina, 407 pp. Madden, R. H., J. Guerrero, R. F Kay, J. J. Flynn, C. C. Swisher. Ill, and A. H. Walton. 1997. The Lav- entan Stage and Age, pp. 355-381. In Kay, R. F. R. H. Madden. R. L. Cifelli, and J. J. Flynn, eds.. Ver- tebrate Paleontology in the Neotropics: The Miocene Fauna of La Venta, Colombia. Smithsonian Institution Press, Washington, D.C. Marshall. L. G.. R. Hoffstetter. and R. Pascual. 1983. Mammals and stratigraphy: Geochronology of the mammal-bearing Tertiary of South America. Pa- laeovertebrata, M6moire Extraordinaire, 1983: 1-93. Marshall, L. G., and T Sempere. 1991. The Eocene to Pleistocene vertebrates of Bolivia and their strati- graphic context: A review, pp. 631-652. In Sudrez- Soruco, R., ed., F6siles y Facies de Bolivia — Vol. I Vertebrados. Yecimientes Petrolificos Fiscales Boli- vianos (YPFB), Santa Cruz, Bolivia. Marshall, L. G., C. C. Swisher, III, A. Lavenu, R. Hoffstetter, and G. H. Curtis. 1992. Geochronol- ogy of the mammal-bearing late Cenozoic on the northern Altiplano, Bolivia. Journal of South Ameri- can Earth Sciences, 5(1): 1-19. MoNES, A. 1986. Palaeovertebrata Sudamericana. Catdl- ogo Sistematico de los Vertebrados F6siles de Amer- ica del Sur. Parte I. Lista Preliminar y Bibliografia. Courier Forschungsinstitut Senckenberg, 82: 1-625. MuNoz, N. 1991. Marco geologico y estratigrafico de un sistema fluviolacustre Paleogeno, Altiplano de Arica, Norte de Chile. VI Congreso Geologico Chileno, Ac- tas 1: 201-204. Nasif, N. L., S. Musalem, and E. Cerdeno. 2000. A new toxodont from the Late Miocene of Catamarca, Argentina, and a phylogenetic analysis of the Toxo- dontidae. Journal of Vertebrate Paleontology, 20(3): 591-600. Oiso, Y. 1991. New land mammal locality of middle Miocene (Colloncuran) age from Nazareno, southern Bolivia, pp. 653-672. In Suarez-Soruco, R., ed., F6s- iles y Facies de Bolivia — Vol. I Vertebrados. Pascual, R. 1966. Litoptema, pp. 161-170. In Borello, A. v., ed., Paleontografia Bonaerense, Vol. 4. Prov- incia de Buenos Aires, Buenos Aires. Pascual, R., and P. Bondesio. 1985. Mamiferos terres- tres del Mioceno medio-tardio de las cuencas de los rios Colorado y Negro (Argentina): Evoluci6n am- biental. Ameghiniana, 22: 133-145. Pascual, R., M. G. Vucetich, G. J. Scillato Yane, and M. Bond. 1985. Main pathways of mammal diversi- fication in South America, pp. 219-247. In Webb, F. G. S. a. S. D., ed.. The Great American Biotic Inter- change. Plenum Press, New York. Patterson, B. 1934a. The auditory region of an upper Pliocene typotherid. Geological Series, Field Museum of Natural History, 6(5): 83-89. . 1934b. Trachytherus, a typotherid from the De- seado beds of Patagonia. Geological Series, Field Mu- seum of Natural History, 6(8): 1 19-139. 1936. The internal structure of the ear in some notoungulates. Geological Series, Field Museum of Natural History, 6(15): 199-227. Patterson, B., and R. Pascual. 1972. The fossil mam- CROFT ET AL.: CHUCAL UNGULATES 51 mal fauna of South America, pp. 247-309. In Keast, A., F. C. Erk, and B. Glass, eds.. Evolution, Mammals, and Southern Continents. State University of New York Press, Albany. DE QuEiROZ, K., AND J. Gauthier. 1990. Phylogeny as a central principle in taxonomy: Phylogenetic defini- tions of taxon names. Systematic Zoology, 39(4): 307-322. Reguero, M. a., and p. V. Castro. 2002. Un Nuevo Trachytherinae (Mammalia, tNotoungulata, Mesoth- eriidae) del Deseadense (Oligoceno tardio) de Cabeza Blanca, Chubut, Argentina. Importancia bioestrati- grafica y filogenetica del genero Trachytherus. VIII Congreso Argentino de Paleontologia y Bioestratigra- fia, Correientes, Resumenes, p. 103. RiQUELME, R. 1998. Evolucion tectosedimentaria post- oligocenica del borde occidental del Altiplano, entre Tignamar y Salar de Surire, I Region, Chile. Thesis, Universidad de Chile, Santiago, 123 pp. RiQUELME, R., AND G. Herail. 1997. Puesta en evidencia de discordancias progresivas en el Cenozoico Superior del Altiplano de Arica: Implicancias en la interpreta- cion de la Cordillera Occidental. VIII Congreso Geo- logico Chileno, Antofagasta, Actas 1: 231-235. Salinas, P., C. Villarroel, L. Marshall, P. Sepulveda, AND N. Munoz. 1991. Typotheriopsis sp. (Notoungu- lata, Mesotheriidae), mamifero del Mioceco Superior en las cercanias de Belen, Arica, norte de Chile. VI Congreso Geologico Chileno, Vifia del Mar, Actas 1: 314-317. Savage, D. E., and D. E. Russell. 1983. Mammalian Paleofaunas of the World. Addison-Wesley, Reading, Massachusetts, 432 pp. Scott, W. B. 1910. Mammalia of the Santa Cruz beds. Vol. 7, Paleontology. Part 1: Litopterna of the Santa Cruz Beds, pp. 1-156. In Scott, W. B., ed.. Reports of the Princeton University Expedition to Patagonia (1896-1899). Princeton University, E. Schweizer- bart'sche Verlashandlung, Stuttgart. . 1912. Mammalia of the Santa Cruz beds. Vol. 6, Paleontology. Part 2: Toxodonta of the Santa Cruz Beds, pp. 1 1 1-238. In Scott, W. B., ed.. Reports of the Princeton University Expedition to Patagonia (1896-1899). Princeton University, E. Schweizer- bart'sche Verlashandlung, Stuttgart. Shockey, B. J. 1999. Postcranial osteology and func- tional morphology of the Litopterna of Salla, Bolivia (late Oligocene). Journal of Vertebrate Paleontology, 19(2): 383-390. Simpson, G. G. 1935. Early and middle Tertiary geology of the Gaiman Region, Chubut, Argentina. American Museum Novitates, 775: 1-29. . 1945. The principles of classification and a clas- sification of mammals. Bulletin of the American Mu- seum of Natural History, 85: 1-350. . 1980. Splendid Isolation, the Curious History of South American Mammals. Yale University Press, New Haven, Connecticut. Sinclair, W. J. 1909. Mammalia of the Santa Cruz beds. Vol. 6, Paleontology. Part 1 : Typotheria of the Santa Cruz Beds, pp. 1-110. In Scott, W. B., ed.. Reports of the Princeton University Expeditions to Patagonia, 1896-1899. Princeton University, Princeton, New Jer- sey. SoRiA, M. F. 1981. Los Litopterna del Colhuehuapense (Oligoceno tardio) de la Argentina. Revista del Museo Argentino de Ciencias Naturales "Bernardino Riva- davia," 3(1): 1-54. Stirton, R. a. 1953. A new genus of interatheres from the Miocene of Columbia. University of California Publications in Geological Sciences, 29(6): 265-348. Villarroel, C. 1974a. Les Mesotherines (Notoungula- taf. Mammalia) du Pliocene de Bolivie. Leurs rapports avec ceux D'Argentine. Annales de Paleontologie, 60(2): 245-281. . 1974b. Un Mesotheriine nouveau (Notoungu- lata. Mammalia) dans le Miocene superieur de Boli- vie. Comptes Rendus de la Academie des Sciences, Paris, Serie D, 279: 551-554. 1978. Edades y correlaciones de algunas uni- dades litoestratigraficas del Altiplano boliviano y es- tudia de algunos representantes mesotheriinos. Revista de la Academia Nacional de Ciencias de Bolivia, 1(1): 159-170. Williams, S. H., and R. E Kay. 2001. A comparative test of adaptive explanations for hypsodonty in un- gulates and rodents. Journal of Mammalian Evolution, 8(3): 207-229. Worner, G., K. Hammerschmidt, E Henjes-Kunst, J. Lezaun, and H. WiLKE. 2000. Geochronology (^°Ar/ ^'Ar, K-Ar and He-exposure ages) of Cenozoic mag- matic rocks from Northern Chile (18-22°S): Implica- tions for magmatism and tectonic evolution of the central Andes. Revista Geologica de Chile, 27(2): 205-240. 52 FIELDIANA: GEOLOGY UNIVERSITY OF ILLINOIS-URBANA 3 0112 060380208 & Field Museum of Natural History 1400 South Lake Shore Drive Chicago, Illinois 60605-2496 Telephone: (312) 665-7055