550 5 Fi GEOIDGYLIBRMW n.s. no.49 J r lELDTANA
Geology NEW SERIES, NO. 49
Large Archaeohyracids (Typotheria, Notoungulata) from Central Chile and Patagonia, Including a Revision of Archaeotypotherium
Darin A. Croft Mariano Bond
John J. Flynn Marcelo Reguero Andre R. Wyss
I GO CM
November 26, 2003 Publication 1527
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. The page charge currently is $65 per page. This figure is subject to change. All authors are encouraged to provide funding 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).
The decision of the scientific editor to accept or reject a work is final.
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 We do not publish short, joumal-article-length works.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.
UMiVERSirr OF iLLINOIS LIBRARY AT URBANA-CHAMPAIGN GEOLOGY FIELDIANA
Geology NEW SERIES, NO. 49
Large Archaeohyracids (Typotheria, Notoungulata) from Central Chile and Patagonia, Including a Revision of Archaeotypotherium
Darin A. Croft Mariano Bond John J. Flynn Marcelo Reguero Andre R. Wyss
' Contributors affiliations are listed on page v.
Accepted May 2, 2003 Published November 26, 2003 Publication 1527
PUBLISHED BY FIELD MUSEUM OF NATURAL HISTORY © 2003 Field Museum of Natural History ISSN 0096-2651 PRINTED IN THE UNITED STATES OF AMERICA ^.^ ijrc^t^ UO V GEOLOGY LIBRARY
Table of Contents 1 1 . Casts of skull and deciduous dentition of A. tinguiriricanese, SGOPV 2900 and SGOPV 3080 18 12. Casts of of A. Abstract 1 specimens pattersoni: maxilla SGOPV Introduction 1 right (holotype), 2918, Taxonomic Notes 3 and right mandible, SGOPV 2917 23 13. of Abbreviations 3 Bivariate plots of lower molars 25 Tooth Wear AND Dental Measurements .... 5 Pseudhyrax 14. Mandibles from Fauna re- Systematic Paleontology 8 Tinguiririca ferred to Pseudhyrax 26 A rchaeotypotherium propheticus 9 15. Pseudhyrax cf. P. eutrachytheroides, Archaeotypotherium tinguiriricaense 14 SGOPV 2877 27 A rchaeotypotherium pattersoni 22 16. Pseudhyrax eutrachytheroides speci- Pseudhyrax eutrachytheroides 25 mens from the Mustersan of Chubut, Pseudhyrax strangulatus 30 Argentina: occlusal views 28 Pseudhyrax sp. indet 30 17. Pseudhyrax sp. indet.: left mandibular Phylogenetic Relationships 31 fragment, SGOPV 2901 30 Conclusions 35 18. Strict consensus tree representing the Acknowledgments 36 phylogenetic relationships among ar- Literature Cited 36 chaeohyracids 33
List of Illustrations List of Tables
1 . of Holotype Archaeohyrax patagoni- 1. Measurements of teeth of Archaeohy- CM5, MACN A52-617 2 rax sp. nov. from the Deseadan SAL- 2. Bivariate of and lower first plots upper MA deposits at Salla, Bolivia 4 molars of of specimens Archaeohyrax 2. Measurements of molar row lengths for nov 7 sp. specimens of Archaeohyrax sp. nov. 3. Bivariate plots of upper and lower from the Deseadan SALMA of Salla, third molars of of Archaeo- specimens Bolivia, and Pseudhyrax species from hyrax sp. nov 7 the Mustersan SALMA of Chubut, Ar- 4. of Holotype Archaeotypotherium pro- gentina 8 pheticus, MLP 52-XI-4-168a, occlusal 3. Dental measurements of A rc/iaeofy- 9 view potherium propheticus specimens 12 5. Casts of species synonymized under A. 4. Dental measurements of Archaeoty- 10 propheticus potherium tinguiriricaense specimens .... 19 6. Specimens included in hypodigm of A. 5. Dental measurements of Archaeoty- propheticus 11 potherium pattersoni specimens 24 7. Cast of holotype of Archaeotypother- 6. Dental measurements of Pseudhyrax ium tinguiriricaense, SGOPV 2823 15 specimens 28 8. Cast of A. tinguiriricaense: palate, oc- 7. Identification of archaeohyracid speci- clusal view, SGOPV 2851 16 mens 29 9. Lower dentition of A. tinguiriricaense, 8. Matrix of characters used in phyloge- SGOPV 3067 and SGOPV 3052 16 netic analysis of archaeohyracid rela- 10. Cast of A. tinguiriricaense: mandibles, tionships 32 SGOPV 3034 17 9. Currently recognized hyracid species .... 32
ui
Contributors
Darin A. Croft Marcelo Reguero Department of Organismal Biology Departamento Cientifico de Paleontologia and Anatomy de Vertebrados The University of Chicago Museo de La Plata 1027 East 57th Street Paseo del Bosque s/n Chicago, Illinois 60637 1900 La Plata U.S.A. Argentina Mariano Bond Andre R. Wyss Departamento Cientifico de Paleontologia Department of Geological Sciences de Vertebrados University of California, Santa Barbara Museo de La Plata Santa Barbara, California 93106 Paseo del Bosque s/n U.S.A. 1900 La Plata Argentina
John J. Flynn Department of Geology Field Museum of Natural History 1400 South Lake Shore Drive Chicago, Illinois 60605-2496 U.S.A.
Large Archaeohyracids (Typotheria, Notoungulata) from Central Chile and Patagonia, Including a Revision of Archaeotypotherium
Darin A. Croft Mariano Bond John J. Flynn Marcelo Reguero Andre R. Wyss
Abstract
The Tinguiririca Fauna of the Andean Main Range of central Chile is remarkable for its abundant and diverse archaeohyracids. This study recognizes four relatively large-bodied spe- cies from the Tinguiririca Fauna, two of which are new. Together with two previously described small-bodied forms, the total of six species makes the archaeohyracid assemblage from Tin- guiririca the most diverse known, representing the co-occurrence of at least 40% of all ar- chaeohyracid species in a single fauna. The two new archaeohyracids are referred to Archaeo- typotherium {A. tinguiriricaense and A. pattersoni), for which a revised diagnosis also is pre- sented. The revision synonymizes the Argentine taxa Archaeohyrax propheticus, Archaeoty- potherium transitum, and Archaeohyrax {''''Bryanpattersonia") nesodontoides under Archaeotypotherium propheticus (new combination). Four specimens from central Chile, in- cluding three mandibular fragments and one partial upper dental series, are referred to the two existing species of Pseudhyrax. The upper dental series is one of the best examples known for the taxon, and is referred to Pseudhyrax cf. P. eutrachytheroides. Based on a metric study of Pseudhyrax specimens from Argentina, one of the Chilean mandibles is referred to Pseudhyrax eutrachytheroides, one to Pseudhyrax strangulatus, and a third to Pseudhyrax sp. indet. Anal- ysis of a large sample of Archaeohyrax specimens from Salla, Bolivia, provides the basis for interpreting wear-related metric variation in archaeohyracid tooth dimensions. It demonstrates that most cheek teeth decrease in length and increase in width through increasing wear, although upper and lower third molars are exceptions. Owing to the dramatic metric and morphologic differences between worn and unworn archaeohyracid teeth, care should be taken when inter- preting the systematic significance of metric differences among specimens of different wear states. A preliminary phylogenetic analysis suggests that taxa traditionally included in the Archaeohyracidae do not form a monophyletic group exclusive of Hegetotheriidae and that a comprehensive review of the names associated with major clades of typothere notoungulates is needed.
Introduction ognized as such (Wyss et al., 1990, 1994). In 1988, a small prospecting team from our research discovered the first Despite the rich Cenozoic record of South group (see Novacek, 2(X)2) of a diverse fossil in cen- American fossil mammals, significant gaps punc- specimens assemblage tuate the sequence of South American Land Mam- tral Chile, later known as the Tinguiririca Fauna mal "Ages" (SALMAs; Flynn & Swisher, 1995). (Novacek et al., 1989; Wyss et al., 1990, 1994, Until recently, faunas between the well-known 1996; Flynn & Wyss, 1990, 1999; Charrier et al., Mustersan (late Eocene) and Deseadan (mid- to 1990, 1996; Wyss & Flynn, 1991; Flynn et al., late Oligocene) SALMAs were not known or rec- 1991, 2003; Wyss, Flynn, et al., 1992; Wyss, Nor-
FIELDIANA: GEOLOGY, N.S., NO. 49, NOVEMBER 26, 2003, PP. 1-38 Fig. 1. MACN A52-617, holotype oi Archaeohyrax patagonicus, left side of skull, viewed as right (adapted from ^^ Ameghino, 1897). Scale bar 1 cm.
ell, et al., 1992; Wyss, Flynn, et al., 1993; Wyss, 31.5-37.5 Ma or more (spanning the Eocene-Ol- Norell, et al., 1993). Subsequent studies have re- igocene transition). Various lines of evidence sug- vealed that the Tinguiririca Fauna helps fill this gest that it might be only of very short duration long and important late Eocene-early Oligocene (possibly less than 2 m.y. and entirely earliest Ol- gap in the SALMA sequence and provides im- igocene, —31-33 Ma; Flynn et al., in press.). Ho- portant evidence regarding the response of South rizons unconformably bracketing the correlative American mammal faunas to the worldwide cool- Tinguirirican "APS" faunas in the Gran Barranca ing event and environmental transformation of Argentina have yielded radioisotopic dates and known as the Eocene-Oligocene Transition (Flynn paleomagnetic stratigraphies, including a K-Ar & Wyss, 1998; Flynn et al., in press). Addition- date of 28.8 ± 0.9 Ma (Marshall et al., 1986) ally, it has clarified the interpretation of several above and 36-38.5 Ma below (Kay et al., 1999; other faunas from Argentina that were previously see also Flynn & Swisher, 1995 and Flynn et al., regarded as Mustersan in age or mixtures of sep- 2003). arate Casamayoran, Mustersan, and Deseadan The abundance and diversity of its archaeohy- faunas (Wyss et al., 1994; Bond et al., 1996; racid assemblage are among the most distinctive Bond, Lopez, et al., 1997; Bond, Reguero, et al., features of the Tinguiririca Fauna (Wyss, Norell, 1997; Reguero, 1998; Hitz et al., 2000; Flynn et et al., 1993; Wyss et al., 1994; Croft, 1998, 2000; al., 2003; Reguero et al., 2003). Reguero et al., 2003; Flynn et al., 2003). At least We have presented robust evidence elsewhere six archaeohyracid species occur at Tinguiririca that the Tinguiririca Fauna from Chile represents (40% of all currently recognized archaeohyracid a new SALMA (e.g., Wyss et al., 1990, 1994; species), the greatest diversity of archaeohyracids Wyss, Flynn, et al., 1993). This biochronologic known from a single fauna. These six taxa include interval was informally referred to as New SAL- two new species of Protarchaeohyrax (Reguero MA ("Tinguirirican") in the SALMA chronology et al., 2003), two new species of Archaeotypo- of Flynn and Swisher (1995) and is named and therium (present study), and two previously de- defined formally as a new SALMA, the Tingui- scribed species of Pseudhyrax (Simpson, 1967). rirican, in Flynn et al. (2003). The stratotype se- Based on the new material from the Tinguiririca quence for this SALMA assemblage is a compos- Fauna and detailed comparative studies of Argen- ite section representing the basal Abanico tine specimens, revised diagnoses of both Ar- (=Coya-Machali) Formation near Termas del Fla- chaeotypotherium and Pseudhyrax are presented co, in the upper Rio Tinguiririca valley, Chile here. (Charrier et al., 1996; Flynn et al., 2003), with The Archaeohyracidae was named by Ameghi- correlative assemblages in Argentina, including no in 1897 and placed in the Hyracoidea, an oth- those from the "Astraponoteen plus superieur" erwise Old World group. He based Archaeohyrax, ("APS") of the Gran Barranca, Chubut (see Bond and his taxonomic placement of the group, on the et al., 1996; Bond, Reguero, et al., 1997; Reguero, type specimen of Archaeohyrax patagonicus 1998; Kay et al., 1999; Hitz et al., 2000). Levels (MACN A52-617), an excellent skull and man- in the stratotype producing the Tinguiririca Fauna dibles from the Deseadan SALMA of Chubut, Ar- have yielded high-precision '"'Ar/''^Ar radioisoto- gentina (Fig. 1). Although Ameghino believed the pic dates, as have underlying nonfossiliferous Archaeohyracidae to be allied with hyraxes, most
beds, indicating that the fauna is at least 31.5 Ma other early workers (e.g., Sinclair, 1909; Loomis, in age, spanning a range potentially as large as 1914) recognized that archaeohyracids (and other
FIELDIANA: GEOLOGY typotheres) were members of the endemic South study strives to clarify some of the taxonomy and American group Notoungulata. George Gaylord nomenclature of the Archaeohyracidae as a basis Simpson's views on the taxonomic position of the for future comprehensive studies on the phylog- the course of Archaeohyracidae changed during eny, biogeography, biochronology, and paleobi- In his landmark classification of mam- his studies. ology of the group. mals (Simpson, 1945), he placed the Archaeohy- racidae in the Toxodontia (contra Sinclair, 1909), but later allied them with the hegetotheriids in the Taxonomic Notes Hegetotheria (Simpson, 1967). Although some studies have suggested that hegetotheriids are more related to mesotheriids than to closely Most recent phylogenetic analyses suggest that archaeohyracids (Wyss, Norell, et al., 1993; Croft, the Archaeohyracidae (as that name is currently 1998), most recent phylogenetic studies have sup- employed) represents a paraphyletic assemblage ported the close relationship between archaeohyr- of taxa closely related to hegetotheriids (Cifelli, acids and hegetotheriids (Cifelli, 1993; Hitz, 1993; Hitz, 1995; Croft, 1998, 2000; Reguero, 1995; Reguero, 1998; Croft, 2000). 1999; see also the phylogenetic analysis below). For many years MACN A52-617 (the holotype However, the relationships among archaeohyra- of Archaeohyrax patagonicus; Fig. 1) was the cids and hegetotheriids have yet to be rigorously only archaeohyracid skull known. Until this study, tested. Because the term Archaeohyracidae is fa- it remained the only one figured and described.' miliar to South American paleomammalogists, As exemplified by MACN A52-617, taxa cur- and because no one has a rently included within the Archaeohyracidae are proposed phylogenetic definition and/or alternative name for the least in- characterized by a complete (3/3, 1/1, 4/4, 3/3) clusive clade these taxa, we use the and essentially closed dentition (lacking signifi- encompassing terms and in cant diastemata), with enlarged and procumbent Archaeohyracidae archaeohyracids first incisors and an evenly graded cheek tooth their traditional sense for the purposes of this pub- series. The skull is low and broad across the orbits lication.
and occiput but narrow across the snout. Archaeo- Taxonomic names that are potentially invalid
hyracids are small- to medium-sized mammals; due to synonymy (e.g., "Archaeohyrax concen- the maximum length of the skull of MACN A52- tricus'') are indicated through the use of quotation 617 is approximately 15 cm, and Tinguirirican ar- marks. chaeohyracids are estimated to have had body masses between 250 g and 4 kg (Croft, 2000; Rynn et al., 2003). No postcranial bones definite- Abbreviations ly attributable to the family Archaeohyracidae have yet been described, although currently un- are in the prepared specimens present Tinguiririca AMNH, American Museum of Natural History, Fauna. New York; MLP, Museo de La Plata, Argentina; Until recently, archaeohyracids have been the SGOPV, vertebrate paleontology collections, Mu- subject of very few studies. This has been due seo Nacional de Historia Natural, Santiago; primarily to the scarcity of archaeohyracid mate- MACN, Museo Argentino de Ciencias Naturales, rial, even in many well-sampled faunas. The dis- "Bernardino Rivadavia," Buenos Aires; APS, covery of the Tinguiririca Fauna has renewed in- "Astraponoteen plus superieur"; CV, coefficient terest in this group of notoungulates (Wyss, No- of variation; FA, first occurrence; HI, hypsodonty rell, et al., 1993; Wyss et al., 1994; Reguero, index (height of unworn tooth/length of tooth); K, 1993, 1998; Croft, 1998, 2000; Reguero «fe Lopez, potassium; Ar, argon; Ma, megannum; m.y., mil- 1999; Reguero et al., 2003), as has continued col- lions of mm, millimeter; cm, centimeter; lecting in the archaeohyracid-rich Deseadan Salla years; g, South American Beds of Bolivia (MacFadden et al., 1985; Shock- gram; kg, kilogram; SALMA, Land Mammal tooth loci are in- ey, 1997; Reguero & Cifelli, 1997). The present "Age." Upper dicated by uppercase letters (e.g., II, P2, Ml) and lower tooth loci lowercase letters il, ' by (e.g., p2, Other skulls are known from both the Tinguiririca deciduous teeth are indicated a "d/D" Fauna (Tinguirirican SALMA) and Salla, Bolivia (De- ml); by seadan SALMA). preceding the tooth position.
CROFT ET AL.: LARGE ARCHAEOHYRACIDS FROM CHILE AND PATAGONIA r~ ^^n
II +1 I \ cr c fs _: «"
°° "" 2! — Tt
II +1 I 7 rj 00 fS —
II +1 I Id 00 >S — w • rn — "* 00 r- o I? — as — o\ II +1 I I 06 as r~ ON 00 " ° u-i — <0 II +1 I I vd — >s c vo _^- S; ^ 00 w-1 >ri-d^~ II +M 7 t~ as 10 ^ ^ o ^ '- ^ • II +1 i I — ''O fS — * 00 ON O . — „ t^ — 00 II +1 I I O^ c — • s^ =- • II +1 I 7 ®^ c o -_• •^ On ^CN U-, ii: NO I + 1 I r • "O 00 >0 On ^ 5 +1 »i « CS ^ isolated), or heavy wear (central fossa absent; fea- within the Salla beds^). This has important impli- tureless occlusal surface). Although division of cations for the use of these sorts of metric values tooth wear into a greater number of stages would alone to distinguish archaeohyracid species (see have been desirable, the morphology of the teeth recommendations below). was such that this could not have been done in an The data also demonstrate that as teeth undergo objective way without a more detailed study of wear, they generally become shorter (mesiodistal- the dentition of this undescribed species (currently ly) and wider (buccolingually). The relationship underway by one of us [M.R.] and R. Cifelli). between tooth wear and tooth shape is most easily Moreover, these three categories can be general- seen in bivariate plots of tooth length and width ized to other archaeohyracids without having to for upper and lower first molars (Fig. 2). The Ml- worry about differences in hypsodonty and rela- 2 and ml -2 become shorter and broader with in- tive depth of accessory fossettes/fossettids (which creasing wear. Accordingly, in comparing speci- could potentially be used to further subdivide mens of unequal wear, if the longer one is more wear stages). In addition to individual tooth worn, it is more likely that these differences re- lengths, molar row lengths were recorded for the flect taxonomic rather than individual (species or Salla Archaeohyrax specimens, as well as for a ontogenetic) variation; the reverse is true when complementary but smaller sample of Pseudhyrax the longer tooth is less worn, since it can be in- specimens from Chubut, Argentina.^ Upper tooth ferred that it would shorten with additional wear. lengths were measured parallel to, and widths per- Upper and lower third molars do not follow this pendicular to, the ectoloph. Lower tooth lengths trend (Fig. 3); upper third molars increase in were measured parallel to, and widths perpendic- length with wear, and whereas the plot for lower ular to, the long axis of the tooth. It should be third molars suggests a similar trend may exist, noted that since each tooth is scored for wear in- the large amount of overlap among wear catego- dividually, some mandibular fragments include ries precludes useful differentiation. teeth of all three wear states. Graphical and sta- Finally, molar row lengths (M/ml-3) tend to tistical procedures were performed using Micro- be less variable than individual tooth lengths (Ta- soft Excel 4.0 or StatView 4.1 for the Macintosh ble 2). Archaeohyrax specimens from Salla and computer. specimens of Pseudhyrax from Chubut both have Table 1 lists univariate statistics for all the mo- small molar row length CVs (<10), with nearly lars and the last two premolars. The coefficient of all specimens (28 out of 30 Archaeohyrax speci- = variation (CV) was calculated as CV 100 X mens, all Pseudhyrax specimens) varying within (standard deviation/mean), following Gingerich 10% of the mean. Two factors might contribute to (1974). Values for the entire sample, along with this lower variability. One possibility is that mea- wear stage, are given by tooth position. suring tooth-row lengths tends to "average out" Several generalizations can be drawn from the wear-related size differences between teeth. Spec- data. Disregarding wear, the coefficients of vari- imens with the molars in different wear stages ation for each tooth position are very high for a (due to different eruption times) are common, single population from a single locality. In most while specimens having either all unworn or en- cases, CVs decrease when each wear category is tirely heavily worn molars are rare. Thus, com- considered separately. Therefore, when compar- parisons between the extremes of the wear spec- ing metric values from different specimens to de- trum are much less common for molar rows than termine their taxonomic affinity, it is important to for individual teeth. Also, as noted previously, be- compare specimens displaying similar wear. Even cause third molars tend to lengthen with wear, this when comparing similarly worn teeth, however, may offset some of the shortening associated with significant variation persists. A tooth 20% larger wear in the more anterior molars. Regardless of or smaller than the mean is not unusual, and two the precise cause, this study documents that molar teeth may vary by as much as 40% in their di- row lengths are less variable than individual tooth mensions (although the large variation in Ar- lengths and should be used whenever possible in chaeohyrax sp. nov. may result in part from pool- taxonomic studies. This is especially true for more ing specimens from different stratigraphic levels hypsodont archaeohyracids, as suggested by the ~ ^ Individual tooth lengths were also recorded for the Testing for the effects of stratigraphic level on size Pseudhyrax specimens but are not included in the pre- was precluded by the lack of stratigraphic data for the sent study due to the small sample size. majority of specimens measured. FIELDIANA: GEOLOGY /.a- E E, -» •D Table 2. Measurements of molar row lengths for specimens of Archaeohyrax sp. nov. (Salla, Bolivia; Deseadan SALMA) and Pseudhyrax species (Chubut, Argentina; Mustersan SALMA). Measurements are made to the nearest = 0.1 mm. Coefficient of variation (CV) 100 X (standard deviation/mean). Chubut Salla Archaeohyrax Pseudhyrax = Ml-3 (n 11) ml-3 (n = 19) ml-3 (n = 5) Mean ± SD 25.0 ± 1.3 26.0 ± 1.8 30.8 ± 0.8 Range 22.3-26.8 23.2-29.7 30.0-31.8 Range (% of mean) 89%- 108% 89%- 114% 97%-103% CV 5.1 6.9 2.6 15% if it is the more worn; this should be reversed located to Archaeohyrax, Eomorphippus,'^ and for isolated M3s (i.e., the difference should be Bryanpattersonia as more properly included in greater if the larger specimen is the more worn, Archaeotypotherium. These specimens/species all less if it is the less worn). Individual lower third appear to be conspecific, and the name Archaeo- molars of different wear states should differ by at typotherium propheticus has chronologic priority. least 30%. Given the temporal congruence of the specimens referred to Archaeotypotherium (Wyss et al., 1994; Flynn et al., 2003; see also Reguero, 1993, 1998; Bond et al., 1996; Kay et al., 1999; Hitz et Systematic Paleontology al., 2000), this synonymy is not surprising. Included Species—The type, Archaeotypo- therium tinguiriricaense (new species, below), Mammalia Linnaeus, 1758 and Archaeotypotherium pattersoni (new species, Notoungulata Roth, 1903 below). Typotheria Zittel, 1892 Diagnosis—A member of the Hegetotheria that Hegetotheria Simpson, 1945 differs from Hegetotheriidae in the absence of Archaeohyracidae Ameghino, 1897 hypselodont (ever-growing) cheek teeth, absence of hypselodont II, absence of straight lingual face 1903 Archaeotypotherium Roth, on lower molars, presence of fossettes/fossettids at some stages of wear in upper and lower molars, Archaeohyrax Ameghino, 1897 (partim): 435. and presence of significant change in cheek tooth Archaeohyrax Ameghino, 1901 (partim): 361- occlusal shape with wear. Differs from Archaeo- 362. hyrax in its lower-crowned cheek teeth (newly Eomorphippus Ameghino, 1901 (partim): 373. erupted teeth are rooted in Archaeotypotherium, = Bryanpattersonia Simpson, 1967 (partim): 115. unrooted in Archaeohyrax; HI 1.75-2.25 in Ar- chaeotypotherium, HI > 2.5 in Archaeohyrax), less Type Species—Archaeotypotherium transitum relatively larger II, triangular upper molars, more elongate central fossae on upper and lower Roth, 1903, here regarded as a junior synonym of molars, lower more Archaeotypotherium propheticus (Ameghino, larger premolars, persistent fossettes/fossettids on and lower 1897). accessory upper molars, M3 without lobe formed me- Comments—Archaeotypotherium transitum posterior by m3 shorter, and labial sul- was recognized by Roth (1903) on the basis of a tastyle, proportionately cus on m3 talonid less Differs from maxillary fragment from Canadon Blanco, in pronounced. Protarchaeohyrax by more persistent accessory Chubut, Argentina. (For a discussion of this enig- matic fossettes/fossettids on upper and lower molars, locality, see Wyss et al., 1994; Flynn et al., absence of lingual sulcus on uppers molars, M3 2003.) Although this taxon was ignored by most without posterior lobe formed by metastyle, and subsequent authors, we recognize it here as dis- tinct from all other described archaeohyracids, " Roth misidentified several specimens of Archaeoty- consistent with Roth's view. we con- Moreover, potherium as Eurystomus stehlini, a junior synonym of sider the various specimens/species previously al- Eomorphippus obscurus (Simpson, 1967). FIELDIANA: GEOLOGY Fig. 4. MACN A52-618, holotype of Archaeotypotherium propheticus, mandibular symphysis and portion of right horizontal ramus with right il-2, c, p2-4 and left i2-3, c, pi, occlusal view (from Ameghino, 1897). Scale bar = 1 cm. much larger size. Differs from Pseudhyrax by Archaeotypotherium propheticus (Ameghino, = higher-crowned cheek teeth (HI 1.3-1.6 in 1897), new combination 1 1 less Pseudhyrax), relatively larger , pronounced (Figures 4-6) paracone and parastyle ridges on upper molars, larger lower premolars, and proportionately lon- Archaeohyrax propheticus Ameghino, 1897: 435. ger talonids on lower molars. Differs from the Archaeohyrax nesodontoides Ameghino, 1901: new archaeohyracid taxon from Antofagasta de 361; Simpson, 1967: 113. La Sierra (Lopez, 1997; Reguero «fe Lopez, 1999) Archaeohyrax concentricus Ameghino, 1901: by its much larger size, higher-crowned cheek 361-362. teeth (HI of Antofagasta form similar to that of Eomorphippus rutilatus Ameghino, 1901: 373. Pseudhyrax), longer molar talonids, and more per- Archaeotypotherium transitum Roth, 1903: 22- sistent talonid fossettids. Differs from Eohyrax by 23. much higher-crowned cheek teeth (HI < 1.25 in Bryanpattersonia sp. Simpson, 1967: 115. Eohyrax) and much larger size. Age and Distribution—Earliest Oligocene Holotype of Archaeohyrax propheticus— Tinguirirican SALMA (Flynn et al., 2003) of Pa- MACN A52-618, mandibular symphysis and por- tagonia and east-central Chile: Abanico (=Coya- tion of right horizontal ramus with right il-2, c, Machali) Formation, Chile; Cailadon Blanco, pre- p2-4 and left 12-3, c, pi (Fig. 4). sumably the Sarmiento Formation, Chubut, Ar- Lectotype of Archaeohyrax nesodontoides— gentina (Roth, 1903); "Astraponoteen plus super- MACN 10905a, a right ml measuring 12.2 by 6.3 ieur" (APS) horizon of the Sarmiento Formation mm (Fig. 5A). at the Gran Barranca south of Lake Colhue Huapi, Paralectotypes of Archaeohyrax nesodonto- Chubut, Argentina. Horizons bearing the Chilean ides—MACN 10905 (9 isolated lower cheek teeth Tinguiririca Fauna have yielded two high-preci- [catalogued b to j] and 7 isolated upper cheek sion '*°Ar/^^Ar dates near 31.5 Ma (the range of teeth [catalogued k to q]; selected specimens il- the means and associated errors spanning —31-32 lustrated in Fig. 5A). Ma), and they are underlain by a basalt unit whose Lectotype of Archaeohyrax concentricus— upper part also was dated at about 3 1 .5 ± 1 .0 Ma. MACN A52-625, incomplete facial region of Available radioisotopic dates from horizons skull with left I1-M2 and right I1-M3 (Fig. 5B). bracketing Tinguirirican "APS" levels in the Paralectotype of Archaeohyrax concentri- Gran Barranca of Argentina (Kay et al., 1999) are cus—MAC^ A52-629, left p4. generally congruent with the geochronologic in- Lectotype of Eomorphippus rutilatus— formation from Chile. MACN 10915, right upper molar. CROFT ET AL.: LARGE ARCHAEOHYRACIDS FROM CHILE AND PATAGONIA ^^Vi^^^^^^ (Fig. 6B); AMNH 28943, neariy unworn isolated left M3; AMNH 28955, damaged left P2-M3; MLP 89-XII-21-1 (Roth's Nos. 1414), left Ml, left ml, and right ml; MLP 69-III-31-2, right upper molar; MLP 12-1516, left mandible with ml-3. Localities—Only one of the type specimens of "Archaeohyrax concentricus" (MACN A52-625) is accompanied by a locality label. This label reads "Coluapi Pyrotherium,'' indicating a ?De- seadan SALMA locality—precisely which one is uncertain—near or at the south end of Gran Bar- ranca, Chubut. Similarly, only a single specimen assigned to ''Archaeohyrax nesodontoides" (MACN A- 10911, perhaps part of the hypodigm for that taxon) has geographic provenience data. The accompanying label reads: "Colhuapi Noto- stylops (parte sup.)," the "parte sup." indicating a level in the Gran Barranca distinct from that where Casamayoran SALMA fossil vertebrates were collected. It appears to cortespond to the portion of profile M of Simpson that Bond et al. (1996) cortelated with the "Astraponoteen le plus superieur" of Ameghino (1901). It is not improb- able, then, that all of the specimens referted to "Archaeohyrax nesodontoides^' come from the Gran Bartanca. Ameghino's placement of "A. ne- sodontoides'" {Archaeotypotherium propheticus of this study) and "A. sulcidens'' {Protarchaeohyrax gracilis of Reguero et al., in press) in the "Capas con Astraponotus" (Mustersan SALMA; Amegh- ino 1901, 1902) appears to be incortect, taking into account the recent stratigraphic data collected Fig. 6. Specimens included in hypodigm of Ar- from the Gran Bartanca by M. Bond and one of A. chaeotypotherium propheticus. MACN A52-630, the authors (M.R.) and, more recently, by joint M2 and m3. B. MACN A52-628, right right partial right Argentine-U.S. teams led by R. Kay. Moreover, mandible with ml-3. Scale bar = 1 cm. at typical Mustersan localities (La Gran Hondon- ada, Cerro del Humo, and Laguna del Mate), HOLOTYPE OF ArCHAEOTYPOTHERIUM TRANSI- Pseudhyrax is the only archaeohyracid present. " TUM—MLP 12- 1524a, left maxilla bearing Ml-3 The type of Archaeotypotherium transituni''' 12- (M3 partially erupted; Fig. 5C). (MLP 1524a) and MLP 12-1516 come from Hypodigm—Holotype of A. propheticus; all Cafiad6n Blanco, Chubut. The former has a label specimens included in junior synonyms above; that reads "T.i.C.B.", indicating "Terciario infe- MACN A52-630, right M2 and right m3 (Fig. 7); rior de Cafiadon Blanco" ("Formacion terciaria MACN A52-628, partial right mandible with ml inferior, Cafiadon Blanco [Territorio del Chu- FiG. 5. Epoxy casts of species synonymized under Archaeotypotherium propheticus. A. MACN 10905, selected syntypes of Archaeohyrax nesodontoides including (from left to right) right ml (lectotype, MACN 10905a); left ml or m2 with little wear (paralectotype, MACN 10905g or 10905h); right P4 (paralectotype, MACN 10905p). B. MACN A52-625, lectotype of Archaeohyrax concentricus, incomplete facial region of skull with left I1-M2 and right II- M3. C. MLP 12- 1524a, holotype of Archaeotypotherium transitum, left maxilla with Ml-3 (M3 partially erupted). Scale bar = 1 cm. CROFT ET AL.: LARGE ARCHAEOHYRACIDS FROM CHILE AND PATAGONIA 11 00 00 ^ ON ON NO "n p r-; (^ ro ^" .-H en ^ir; in p -st ;i^ o o\ 00 o s CT\ \0 ^ (N On O (^ O ^' O •S s gS ^ si a> o E a. ^ O >= c ^ O Ji *-• rv U ' S3 M S^ -S o 2 I c S hJ -J J J hJ E.y 3 ° «^ << o ^ On E ^ (N Tf . ^ 1 1 the with crowned , clearly major cropping tooth, c ? o o the relatively much higher-crowned posterior pre- >> >s >s >. >. > > > > > molars and molars is a curious feature of this tax- cd cQ cQ c^ c^ 1) 4) U 1) 1) as in HM MM MM MM MM A. c on; it is implanted obliquely Archaeohyrax, £-2 hegetotheriids, and mesotheriids. 12 is slightly vm< O longer in A. propheticus than in Archaeohyrax, its anterior portion projecting forward well beyond vo ui ^ o o o o C are offset from each other short diastemata. 1^ by differs from that of uuuuuZZ Z Z Z The first premolar Archaeo- <<< << hyrax in its more convex external face and by the presence of a shallow anterolingual notch. The es- CROFT ET AL.: LARGE ARCHAEOHYRACIDS FROM CHILE AND PATAGONIA 13 sential characters of the remaining upper cheek convincingly shown by MACN A52-628 and teeth have been given in the diagnosis of Ar- A52-630 (Fig. 6), both old individuals in which chaeotypotherium. Detachment of the facial por- the roots are either formed (ml -2) or forming tion of the right maxilla of MACN A52-625 re- (m3). A talonid fossettid persists on m2 in both veals a progressive, steplike increase in the specimens, a trigonid fossettid on m3 of A52-630, heights of the premolar crowns, and very clearly and both fossettids on m3 of A52-628. shows that Archaeotypotherium is less hypsodont The facial region and the mandible show no than Archaeohyrax. The cheek teeth are at an ear- important differences from Archaeohyrax, al- lier stage of wear than in the type of A. patago- though the muzzle of A. propheticus is slightly nicus, yet they are notably lower-crowned. On broader than in A. patagonicus. P2-4 the anteroexternal fossette is more persistent than the posteroexternal fossette, whereas on the molars the reverse is the case. The internal fos- Archaeotypotherium tinguiriricaense sp. nov. sette is elongate, and is bifid anteriorly, in P3- (Figures 7-11) M2. The incomplete M3 of MACN A52-625, the HoLOTYPE—SGOPV 2823, rostrum with left least worn of the available teeth, reveals a few and right C1-M2 (canines barely erupted; Fig. 7). details of crown pattern (Fig. 5B). A shallow Hypodigm—Holotype of A. tinguiriricaense; notch still persists between the protocone and the SGOPV 2851, palate with highly worn left 11-3, slender lingual extremity of the hypocone. The left and right P2-M3 (Fig. 8); SGOPV 2853, right metaloph appears to have been directly transverse maxilla with transversely crushed C1-M2 (P3 in orientation. There is a very small posterior cin- with damaged ectoloph); SGOPV 2900, slightly gulum, interrupted labially and joined lingually to distorted skull with crushed dentition (left and the tip of the hypocone. The posteroexternal fos- right II, right P2-M3, left P2-3, Ml-2; Fig. sette is a long slit that curves posteriorly toward 1 1 A); SGOPV 2909, left maxillary fragment with the midline, the lingual portion being very shal- ?Ml-2; SGOPV 3043, partial left and right man- low. The internal fossette is a mere slit labial to dibles (fused) with left il-3, pl-m3 and right il- the protocone. Between this fossette and the pos- p3 (Fig. 10); SGOPV 3052, partial left mandible teroexternal one is a wide, abraded area. To judge with p3-m3 (Fig. 9B); SGOPV 3060, left ?M1; from an unworn Ml of the Mustersan-aged SGOPV 3067, skull and mandible with mildly Pseudhyrax eutrachytheroides, this area, when worn dentitions, partially prepared (Fig. 9A, man- newly erupted, consisted of a posterior crista dible only); SGOPV 3080, palate with left and joined to a bulbous cusp or crochet attached to right PI, dP2-4, Ml (Fig. IIB); SGOPV 3260, the metaloph. partial right mandible with ml-3; SGOPV 3261, Lower Dentition—The lower incisors, canines, partial left mandible with p4-m2. and pi -3 are known only from the type of A. Type Locality—Locality C-89-21b, purple- propheticus, a very mature individual in which brown volcaniclastic sediment of the Abanico the dentition is deeply worn (Fig. 4). The teeth at Formation (Charrier et al., 1996; Flynn et al., this stage show no appreciable differences in 2003); 34°59'S, 70°26'W, approximately 1 km structure from those oi Archaeohyrax. A compar- north of pass identified by its elevation, 2738 m ison of a p4 oi Archaeotypotherium (MACN A52- (Anonymous, 1985), known locally as Portezuelo 629) and a mandible of Archaeohyrax patagoni- El Fierro (Charrier et al., 1996), south of the Rio cus (MACN A52-624) at similar stages of wear Tinguiririca, northeast of Cerro Alto del Padre, demonstrates that the talonid fossettid and the lin- and about 3 km south of Termas del Flaco, Chile. gual groove in the trigonid are more persistent in Age and Distribution—Earliest Oligocene Archaeotypotherium than in Archaeohyrax, and Tinguirirican SALMA (Flynn et al., 2003) of east- that the lingual groove posterior to the entoconid central Chile (see above). is less persistent. They also demonstrate that the Diagnosis—As for Archaeotypotherium. Dif- lingual groove between the trigonid and talonid is fers from A. propheticus in its smaller size less persistent than the one posterior to the ento- (approx. 20% smaller), less persistent fossettes on conid in Archaeotypotherium, whereas in Ar- upper molars, and absence of trigonid and talonid chaeohyrax the reverse is the case. The greater fossettids on lower molars. Differs from A. pat- persistence of the trigonid and talonid fossettids tersoni (new species, below) in smaller size in the lower molars of Archaeotypotherium is (approx. 20% smaller), shallower mandible, and 14 FIELDIANA: GEOLOGY Fig. 7. Epoxy cast of SGOPV 2823, holotype of Archaeotypothehum tinguiriricaense, palate with left and right C1-M2 (canines barely erupted), in right lateral view (above) and occlusal view (middle and below). Specimen = illustrates relatively unworn dental morphology of Archaeotypotherium tinguiriricaense. Scale bar 2 cm. proportionately smaller and more triangular pre- common archaeohyracid in the Tinguiririca Fau- molars. na; numerous specimens are referable to it, in- Etymology—In reference to the Rio Tinguirir- cluding two skulls (only the second and third pub- ica and the Tinguiririca Fauna of Chile, the type lished for an archaeohyracid), a palate bearing a locality and fauna for the species. deciduous dentition, and several other upper and Description—A. tinguiriricaense is the most lower dentitions. It is slightly larger than the com- CROFT ET AL.: LARGE ARCHAEOHYRACIDS FROM CHILE AND PATAGONIA 15 Fig. 8. Epoxy cast of SGOPV 2851, palate with left 11-3, left and right P2-M3, in occlusal view. Specimen = illustrates heavily worn upper dental morphology of Archaeotypotherium tinguiriricaense. Scale bar 2 cm. Fig. 9. Lower dentition of Archaeotypotherium tinguiriricaense. A. Portion of SGOPV 3067, little worn left mandible showing p2-m2, occlusal view. B. Epoxy cast of SGOPV 3052, partial left mandible with p3-m3, occlusal view. Scale bars - 1 cm. 16 FIELDIANA: GEOLOGY Fig. 10. Epoxy cast of SGOPV 3043, partial left and right mandibles (fused) of Archaeotypotherium tinguiriri- caense with left il-3, pl-m3, and right il-p3. A. Left lateral view (shown as right lateral). B. Occlusal view. Scale bar - 1 cm. mon Archaeohyrax sp. nov. from the younger fau- aration, the relatively unworn upper dentition is na at Salla, Bolivia (Deseadan SALMA) but much best represented by the holotype specimen less hypsodont. Dental measurements are present- SGOPV 2823 (Fig. 7), a rostrum with both left ed in Table 4. A rostrum with upper dentition and right C1-M2. No incisors are preserved. The (SGOPV 2823) is designated the holotype for this upper canines are visible within their alveoli and new taxon because it preserves many diagnostic are in the process of erupting. M3, if present, is features, and archaeohyracid taxa typically are completely unerupted. The lateral surface of the most easily distinguished on the basis of the upper maxilla is present on the left side of the rostrum, dentition. A skull was not designated the holotype but is not preserved on the right. because in one (SGOPV 2900), the teeth are The canine is small and bladelike (approxi- crushed, and in the other (SGOPV 3067), some mately 2.8 mm in length) and appears to be sep- adult teeth are unerupted and deciduous teeth are arated from PI by a small diastema. The posterior present. edge of PI abuts the anterior edge of P2. It is a Upper Dentition—In the current state of prep- simple tooth, consisting primarily of a single. CROFT ET AL.: LARGE ARCHAEOHYRACIDS FROM CHILE AND PATAGONIA 17 Fig. 1 1. Epoxy casts of skull and deciduous dentition of Archaeotypotherium tinguiriricaense. A. SGOPV 2900, slightly distorted skull with crushed teeth (left and right II, right P2-M3, left P2-3, Ml -2) in right lateral view (above) and occlusal view (below). B. SGOPV 3080, palate with left and right PI, dP2-4, and Ml in occlusal view. Scale bar = 2 cm. rounded cusp occupying nearly the entire anterior semble the molars in shape and complexity. An half of the tooth. A small, sloping shelf is located anterolabially projecting parastyle is present on all distolingual to this main cusp. premolars, but the parastyle does not extend as far The other premolars (P2-P4) more closely re- apically as does the adjacent paracone. Strong la- FIELDIANA: GEOLOGY Table 4. Measurements to the nearest 0.1 mm for specimens of Archaeotypotherium tinguiriricaense. Specimens are in order of as arranged increasing wear, determined by various morphologic indicators (e.g., presence or absence of fossettes, height of cheek teeth). Measurements in parentheses are estimated. Upper dentition stage, the internal cingulum is slightly below the fossette with the internal fossa before wear had occlusal surface of the lingual portion of the taken place. tooth. The presence of a hypocone and the distal Although the incisors are not present in lengthening of the tooth create a more trapezoidal SGOPV 2823, they are at least partially preserved tooth shape in which the labial and lingual sides in SGOPV 3067 (the remaining teeth are unpre- form roughly parallel edges and the segment pared), and SGOPV 2851 (Fig. 8), a specimen along the distal edge of the tooth is perpendicular with highly worn teeth. In SGOPV 3067, the left to both. first incisor is recently erupted and displays little Five fossae/fossettes are present on the occlusal or no wear. It appears to be much less robust than surface of Ml. Most anterior is a small fossette in Archaeohyrax, but, owing to the unprepared resulting from the joining of the antecrochet with lingual surface of the tooth, this is uncertain. The the first crista (i.e., the anterior portion of the con- occlusal edge of the tooth is approximately 6.5 joined first and second cristae). Posterior to this, mm in length. II is about twice as large as 12, and situated approximately halfway along the labial 12 is slightly larger than 13. 13 is in the process edge of the tooth, is the posterior external fossette. of erupting, and 12 may be broken away from its Between these two fossettes, but located more lin- root, precluding a more detailed description of is a small fossette formed their In 285 1 all teeth have gually, anterolingual by morphology. SGOPV , the "pinching off" of the internal fossa by the undergone significant wear and all three incisors well-developed crochet; it is not a true external have been worn nearly to their roots. They are fossette, as the first and second cristae become separated by small diastemata (each approximate- confluent with wear, preventing the fossette from ly 1 mm long) and all have less than 2 mm of reaching the ectoloph. The posterior portion of the enamel present labially. internal fossa is situated lingually, just lateral to Measurements of upper cheek teeth for speci- the small internal cingulum. The internal fossa is mens of Archaeotypotherium tinguiriricaense are connected to the anterolingual fossette by the listed in Table 4. It should be noted that the mea- elongate remnant of the internal fossa. Finally, a surements of teeth of different specimens vary in mediolaterally elongate posterior fossette is pre- the orderly fashion predicted from the aforemen- sent between the posterior cingulum and the me- tioned studies of changes with wear in other ar- taloph. chaeohyracid taxa. The second upper molar is partially erupted; Lower Dentition—The relatively unworn lower only the anterior half of the tooth has undergone dentition of Archaeotypotherium tinguiriricaense any wear and the posteriormost portion of the is best exemplified by SGOPV 3067 (Fig. 9A). tooth is nearly flush with the surface of the palate. Both mandibles are preserved, but the posterior Two cingula are clearly demonstrated in this portion of the right mandible remains unprepared. tooth: a small, somewhat papillate posterior cin- The two mandibles are joined at a completely gulum and a slightly shorter, gently rounded, and fused symphysis that is approximately 21 mm in more robust internal cingulum. The hypocone length; in lateral view, the symphysis is oriented arises from between these two structures and is at an angle of about 30° below the level of the the lowest of the cusps. Its apex is flat and is mandibular tooth row. already confluent with the rest of the metaloph. The lower incisors are small and procumbent, The metastyle lies opposite the hypocone, along their size increasing from mesial to distal. The the labial margin of the tooth. It is significantly first two incisors are very small and are the only larger than the hypocone and has a distinct, point- teeth of the mandible exhibiting significant wear. ed apex. It is confluent with the rest of the ecto- The third incisor exhibits no wear and is much loph, but only near its base. The second crista is larger than the first two. Because SGOPV 3067 well developed and is nearly as high as the worn represents a young individual of A. tinguiriri- anterior portion of the tooth. The crochet, how- caense (most teeth are unworn and m3 is unerupt- ever, is not nearly as tall, but extends anteriorly ed), comparisons with specimens of older individ- from the metaloph to contact the second crista. As uals of this taxon (e.g., SGOPV 3043, described on Ml, an anterior external fossette is present below) suggest that the first two incisors in along the lingual margin of the ectoloph, just an- SGOPV 3067 are deciduous. The identification of terior to the combined first and second cristae. the third incisor is ambiguous; it may be either a Two bands of enamel extend lingually from the late-erupting deciduous tooth or an early-erupting fossette, suggesting a possible connection of the permanent tooth. Based on its similar size and 20 FIELDIANA: GEOLOGY structure to i3 in other specimens, we favor the structure is much more evident in the molars, es- latter alternative, pending additional specimens pecially with increasing wear. that might suggest otherwise. Tooth replacement The molars of SGOPV 3067 have undergone in notoungulates has yet to be thoroughly inves- little wear and there is still much relief between tigated, and the presence of 13 in this specimen the metaconid, with its posterior crest, and the re- would suggest that the third incisor may not have mainder of the occlusal surface. The ml trigonid a deciduous precursor in archaeohyracids.^ is a long, thin structure with two grooves on its In SGOPV 3067, the greatest occlusal dimen- internal face. The first of these grooves is located sion is perpendicular to the sagittal plane in dil, anteromedially on the trigonid, dividing the para- 45° to this plane in di2, and nearly parallel to this conid from the metaconid. The second groove is plane in 13. A slight differentiation of two cusps located on the lingual side of the posterior meta- is present on the lingual surface of 13, dividing conid crest; it divides the crest into two unequal the tip of the lingual face into a large, rounded portions, the more posterior one being about twice posterior cusp and a smaller anterior cuspule. This as long as the anterior On m2, the paraconid is division is even more pronounced on the canines; enlarged lingually, creating a trigonid basin much the two cusps show greater separation and are larger than on ml (though the larger basin may more similar in size than they are in 13. The right be due, in part, to increased wear). The more an- canine has recently erupted, but on the left side terior of the lingual trigonid grooves is essentially the tip of the already fully erupted tooth is par- absent in m2, but a small protuberance is still ev- tially broken off. ident on the anteromedial side of the trigonid, just The teeth of the premolar series grade from ca- anterior to the more posterior groove. As on the niniform to molariform. The first premolar is ca- premolars, a lingual groove is present on the ml niniform, consisting of a simple pointed crest. The talonid, creating a bifid posterolingual talonid second premolar is significantly longer than pi face. The internal talonid groove is short in ml, and shows a differentiated trigonid and talonid. extending perhaps a quarter of the distance from The trigonid of p2 consists of a single sigmoid the occlusal surface to the alveolus. On m2, how- crest with its highest point located at the meta- ever, the groove is much deeper and extends be- conid. The short segment of the trigonid crest pos- low the level of the alveolus. The m3 is unerupt- terior to the metaconid hooks around the lingual ed. side of the talonid, which is low, narrow, and c- The morphology of a heavily worn lower den- shaped. The highest point of the talonid is at its tition is shown by SGOPV 3043 (Fig. 10) and posterior end, where a cuspule lies just medial to 3052 (Fig. 9B). SGOPV 3052 is slightly more the most anteriorly projecting part of p3. The third worn, but other than small wear-related differenc- premolar is similar in morphology to p2, but both es, the two specimens are almost identical in size the trigonid and talonid are expanded buccolin- and morphology (Table 4). gually. The metaconid and its posterior crest are SGOPV 3043 includes the anterior portion of still the highest points on the occlusal surface, but the right mandible (with il-p3) and a complete the trigonid has expanded to form a small basin left mandible and associated dentition. The first instead of merely a crest. Lingual expansion of incisor on each side is small, peglike, and oriented the paraconid, which is much better developed at approximately 45° to the long axis of the man- than on p2 (best seen in lingual view), creates this dible (in lateral view). In occlusal view, the sur- basin. A similar expansion of the entoconid oc- face of il is nearly perpendicular to the sagittal curs in the talonid, creating a talonid basin. The plane, whereas 12 is nearly parallel to it. Both il structure of p4 is very similar to that of p3, except and 12 are larger and more robust than their de- the talonid is proportionately longer In both p3 ciduous counterparts in SGOPV 3067. The third and p4 a very slight vertical groove is present on incisor and the canine possess a high anterior por- the posterolingual portion of the talonid, separat- tion and low posterior portion, similar to the con- ing the entoconid from the hypoconulid. This figuration of the trigonid and talonid in many gen- eralized mammal lower molars. The size of each tooth increases progressively from il-c. In both 5 Sinclair (1909) noted that Pl/pl do not have decid- SGOPV 3043 and 3067, the unusual spoutlike na- uous precursors in Santacrucian SALMA interatheriids, ture of the is so the lack of deciduous precursors for some teeth is not symphysis evident, appearing unprecedented in notoungulates. "pinched in" on both sides. A mental foramen is CROFT ET AL.: LARGE ARCHAEOHYRACIDS FROM CHILE AND PATAGONIA 21 present within the lateral depressions on the sym- Type Locality—Locality C-89-37, otherwise physis, almost directly beneath the canine. as for A. tinguiriricaense. The premolars are all heavily worn, and the pi Age and Distribution—Earliest Oligocene occlusal surface level is approximately equal in Tinguirirican SALMA (Flynn et al., 2003) of east- height to that of the posterior portion of the ca- central Chile (see above). nine. No occlusal morphology is retained on pi Diagnosis—As for Archaeotypotherium. Dif- or p2, while on p3-p4 and ml only a single fossa fers from Archaeotypotherium propheticus in less is present separating the trigonid from the talonid. persistent fossettes on upper molars, smaller and Enamel height on the labial surface of p3 is ap- more quadrangular upper premolars, and absence proximately 4.2 mm. Both the m2-3 talonid and of trigonid and talonid. Differs from A. tinguirir- trigonid are slightly damaged, but most of the icaense in larger size (approx. 20% larger), pro- morphology remains discernible. In both SGOPV portionately larger premolars, deeper mandible, 3043 and 3052 the hypoconulid is directed pos- and thicker labial enamel on upper molars. teriorly. Although not evident in SGOPV 3043, Etymology—After Bryan "Pat" Patterson, in SGOPV 3052 appears to demonstrate that on m3, honor of his extensive contributions to South in addition to the fossa between the trigonid and American paleontology in general and archaeo- talonid, a fossettid forms between the hypoconid hyracids in particular.^ and hypoconulid. Description—A maxilla with upper dentition is Discussion—Owing to the large number of designated the holotype for A. pattersoni because well-preserved specimens of this taxon from the it preserves many diagnostic features, and ar- Tinguiririca Fauna, A. tinguiriricaense will likely chaeohyracid taxa typically are most readily dis- prove to be one of the best-known archaeohyra- tinguished using characters of the upper dentition. cids. Since many of these specimens have yet to A well-preserved mandible with moderately worn be fully prepared from the exceptionally hard vol- p3-m3 (SGOPV 2917; Fig. 12B) probably rep- caniclastic matrix, however, a complete descrip- resents the lower dentition of A. pattersoni. This tion of the skull and preserved/available parts of specimen matches the holotype maxilla very well the appendicular skeleton is not yet possible. in size, degree of hypsodonty, and state of wear. These specimens will be thoroughly described The two specimens were found near each other in when preparation has been completed. Of note, the field and exhibit similar preservation, and it is however, is the well-preserved basicranium of quite possible that they represent the same indi- SGOPV 2900, which illustrates two characters vidual. Patterson (1936, unpublished manuscript) recog- Upper Dentition—Specimen SGOPV 2918 nized as linking Archaeohyrax and hegetotheriids: (Fig. 12A) consists of a right maxilla with mod- the absence of a hamular process on the pterygoid erately worn P2-M3; most of the teeth have in- and an anteriorly located carotid foramen (see completely preserved labial surfaces. A very phylogenetic analysis below). small portion of PI is also present within its al- A. tinguiriricaense includes both of the speci- veolus. The specimen is quite high-crowned, with mens (SGOPV 2823 and SGOPV 2900) previous- a level of hypsodonty comparable to, or perhaps ly referred to "archaeohyracid new taxon A" and slightly exceeding, that of A. tinguiriricaense. Be- one of the specimens (SGOPV 285 1 ) referred to cause the teeth have undergone moderate wear, "archaeohyracid new taxon D" in Wyss et al., little occlusal morphology is evident, except the 1994 (see discussion below). central fossa. Because the first and second upper molars are broken labially, we are unable to dis- cern the pattern of external fossettes (if present). Archaeotypotherium pattersoni However, as M3 exhibits no external fossettes and (Figure 12) * Most of Patterson's research on archaeohyracids ex- HOLOTYPE—SGOPV 2918 12A), (Fig. right ists only in the form of an unpublished manuscript. maxilla with P2-M3, labial portions of teeth in- Simpson (1967) alluded to this work-in-progress and completely preserved. chose to honor Patterson's research by erecting a new taxon, Because the publica- Hypodigm—Holotype of A. pattersoni; SGOPV Bryanpattersonia. present tion formally invalidates the name Bryanpattersonia 2917 (Fig. 12B), right mandible with p2-m3 (based on the principle of priority), it is especially ap- from the same individual as SGOPV (likely propriate that one of the new species in this group be 2918). named for Pat in its place. 22 FIELDIANA: GEOLOGY Fig. 12. Epoxy casts of specimens of Archaeotypotherium pattersoni. A. SGOPV 2918 (holotype), right maxilla with P2-M3, slightly damaged labially, occlusal view. B. SGOPV 2917, right mandible with p2-m3 (likely same = individual as SGOPV 2918), occlusal view. Scale bar 1 cm. is the least worn of the molars, it is likely that groove; this has not yet been isolated through these fossettes (if indeed present in more anterior wear in m3. The lingual sides of the teeth are molars) disappeared early in wear. The premolars nearly flat, although a short groove is present on all are trapezoidal in outline, as are M2 and M3. the posterointernal corner of the talonid, separat- Well-defined parastyle and paracone grooves are ing the entoconid from the hypoconulid; it has present on all cheek teeth and are the only pro- been obliterated by wear in ml but is still present nounced features of the labial surface. Dental as a wide sulcus in m3. measurements are presented in Table 5. The mandibular ramus is substantially deeper Lower Dentition—A single fossettid is located than in A. tinguiriricaense (SGOPV 3043, 3067). between the trigonid and the talonid on p4-m2, In A. pattersoni the mandible is 18.6 mm deep resulting from an isolated fragment of the internal just anterior to m 1 and perpendicular to the lower CROFT ET AL.: LARGE ARCHAEOHYRACIDS FROM CHILE AND PATAGONIA 23 Table 5. Measurements to the nearest 0.1 mm for specimens oi Archaeotypotherium pattersoni. Measurements in parentheses are estimated. Upper dentition P2 P3 P4 Ml M2 M3 Specimen Lower dentition p2 p3 p4 ml in2 Included Species—The type species, Pseudhy- ; rax strangulatus (Ameghino, 1901), and the in- determinate material described here. Diagnosis—Simpson (1967, p. 109) diagnosed Pseudhyrax as "Closely similar to Eohyrax and with that but more 1.8- probably intergrading genus, g:5 progressive, more hypsodont, lower molars tend- "c 1.7 ing to develop a second closed talonid fossette and a closed trigonid fossette." As emended here, Pseudhyrax is a member of the Hegetotheria that differs from Hegetotheriidae in absence of hyp- selodont (ever-growing) cheek teeth, absence of 1 1 face hypselodont , absence of straight lingual on lower molars, presence of fossettes/fossettids at some stages of wear in upper and lower molars, and presence of significant change in cheek tooth shape with wear. Differs from Archaeotypother- ium, Protarchaeohyrax, Archaeohyrax and the new archaeohyracid taxon from Antofagasta de La Sierra (Lopez, 1997; Reguero & Lopez, 1999) by lower degree of hypsodonty, greater persis- tence of accessory fossettes with wear, presence of internal sulcus on endoloph, more pronounced ectoloph ridges, and more quadrangular cheek teeth. Larger than Protarchaeohyrax and the new archaeohyracid taxon from Antofagasta de La Si- erra (Lopez, 1997; Reguero & Lopez, 1999). Pro- toloph less transverse than the new archaeohyra- cid taxon from Antofagasta de La Sierra (Lopez, 1997; Reguero & Lopez, 1999). Age and Distribution—Latest Eocene Muster- san of Patagonia and earliest Oligocene Tinguirir- ican SALMA (Flynn et al., in press) of Patagonia and east-central Chile (see above). Comments—Simpson (1967) recognized two species of Pseudhyrax: P. eutrachytheroides and P. strangulatus. These two are differentiated by size only, and Simpson (1967), while still recog- nizing the specific status of P. strangulatus, ex- pressed doubt as to whether its smaller size was due to taxonomic distinction or merely intraspe- cific variation. In order to examine size variation within the Argentine sample of Pseudhyrax, molar lengths and widths from a large number of specimens (N = 23) were recorded and examined on a bivariate plot (Fig. 13). This plot reveals the following points. First, a gap in the size distribution appears to be present, and is most pronounced for m3. This distribution suggests that Pseudhyrax speci- mens with m3 longer than 11.5 mm and wider than 5.0 mm should be referred to P. eutrachy- theroides, while those with m3 shorter than 9.2 mm and narrower than 5.0 mm should be referred Fig. 14. Mandibles from the Tinguiririca Fauna referred to Pseudhyrax. A, Epoxy cast of SGOPV 2887, Pseud- hyrax strangulatus, right partial mandible with m2-3. B, SGOPV 2985, Pseudhyrax eutrachytheroides, right man- - dibular fragment with partial m2, m3, photo of epoxy cast (left) and illustration (right). Scale bar 5 mm. archaeohyracid lower dentitions, one or both of teeth are moderately worn, except for M3, which the Tinguiririca Fauna lower dentition specimens is only slightly worn. identified as Pseudhyrax might instead pertain to The two incisors are enlarged and spatulate. another archaeohyracid species from Tinguiriri- They curve posteriorly and are obliquely implant- ca, better known from upper teeth. The other ed, meeting at the tips. A strong enamel face is possibihty, endorsed here, is that some (or at present on the anterior face of each tooth, but least one) of the well-known upper dentitions enamel appears to be absent from the back, thus from Tinguiririca is/are actually referable to creating a long, self-sharpening edge. Pseudhyrax (viz., SGOPV 2877, "archaeohy- The premolars in SGOPV 2877 show a rela- racid new taxon B" of Wyss et al., 1994; Table tively high degree of molarization. In lateral view, 7). the metacone, paracone, and parastyle ridges are SGOPV 2877 is a well-preserved upper dental easily distinguishable, the latter two being conflu- series, including left and right II and left P3-M3 ent at the base of the tooth. In occlusal view, the (Fig. 15; Table 6). Although a small fragment of ridges of the paracone and parastyle are more pro- P2 is preserved, the morphologies of the posterior nounced than that on the metacone, but no par- incisors, canine, and anteriormost premolar (if astylar "spur" is present. Each premolar is sub- they were present) are unknown. No bone of the quadrangular and longer than wide; the lingual palate or maxilla is present, so the presence of and labial sides are essentially parallel, and the teeth cannot be discerned from alveoli. The cheek anterior and posterior faces are nearly so. The teeth are relatively low-crowned (at least as com- only feature on the tooth distinguishing the two pared with other contemporaneous archaeohyra- internal cusps is a slight sulcus on the lingual cids), similar to that exhibited by Pseudhyrax. The face. Both premolars are worn and have nearly 26 FIELDIANA: GEOLOGY B Fig. 15. SGOPV 2877, Pseudhyrax cf. P. eutrachytheroides, left and right II, left P3-M3. A. Occlusal view. B. = Reconstruction of SGOPV 2877 with M3 in life position. Scale bar 1 cm. featureless occlusal surfaces, save for an elongate fossette appears as an elongate structure, roughly central fossa oriented anterolabially. parallel to the median external fossette and its The first two molars also display moderate isthmus. wear. As in the premolars, the amount of wear on The third molar is separated from M2 by a siz- each tooth increases posterolingually, resulting in able gap (5 mm), presumably the result of post- the anterolabial comer of each tooth being the mortem dislocation. As expected with the eruption highest part of the crown. The ectoloph is at least pattern, M3 shows the least amount of wear and half again as high as the endoloph in these teeth, preserves the greatest occlusal detail among the and the front edge of each tooth is higher than the molars; the lesser degree of wear is especially ev- posterior edge of the tooth directly in front of it. ident in the shape of the tooth (subtriangular), the The molars possess a better-developed parastylar high protocone (it is flush with the occlusal sur- spur than do the premolars, but this structure face in worn teeth), and the shape of the ectoloph overlaps little, if any, with the proximate anterior in lateral aspect (the ectoloph is tapered at both tooth. Enamel is absent along the labial portion of the base and the occlusal surface in M3, whereas the anterior and posterior faces of Ml -2, permit- it only tapers at the base in worn teeth like Ml- ting a confluence of the dentine of the occlusal 2). It possesses a pronounced metastylar ridge that surfaces just medial to the ectoloph. becomes more prominent toward the base of the Both Ml and M2 retain a central fossa that is tooth. confluent with the median external fossette. In M 1 As in M2, a more posteriorly located external this fossa is concave anteromedially. The central fossette is evident on M3, in addition to the cen- fossa is larger in M2, owing to the lesser amount tral fossa and associated median external fossette. of wear. Additionally, whereas the external fos- This posterior external fossette has less occlusal sette is a small circular structure abutting the cen- exposure than does the corresponding fossette in tral fossa in Ml, in M2 it is located further labi- M2 (owing to a less oblique angle of wear) and ally and is joined to the central fossa by a long, it appears as an oval structure that is approxi- thin isthmus. Only M2 retains a posterior external mately the same size as the posterior end of the fossette. Due to the high angle of wear along the central fossa. Other very small, isolated enamel internal face of the ectoloph, the posterior external lakes (microfossettes) are located on the surface CROFT ET AL.: LARGE ARCHAEOHYRACIDS FROM CHILE AND PATAGONIA 27 of M3. Two microfossettes form a transversely oriented pair in the anteroexternal region of the tooth; two others form an obliquely oriented pair in the posteroexternal region of the tooth; and a single microfossette is located posteriorly and in- ternal to the posterior external fossette. SGOPV 2877 has been described as one of the more unusual typotheres known from the Tingui- ririca Fauna, and was designated "archaeohyracid new taxon B" by Wyss et al. (1994) and as Ty- potheria incertae sedis by Croft (2000). However, further comparisons among a wider series of spec- imens and a better understanding of wear-related change in archaeohyracid teeth suggest that the atypical morphology of SGOPV 2877 may result from unusual preservation (e.g., the lack of bone and the absence of some teeth), a relative scarcity of appropriate comparative material, and unusual wear-related shape changes in the teeth relative to other taxa. SGOPV 2877 can be excluded from all other typotheres except the Archaeohyracidae based on the following characteristics: differs from Cam- panorcidae in presence of more hypsodont teeth, rooted 1 1 un- presence of more deeply , absence of dulating ectoloph on upper molars; differs from Archaeopithecidae in presence of spatulate first incisors, premolars with enlarged hypocone, ab- Fig. 16. Occlusal views of P. eutrachytheroides sence of sulcus on more specimens from the Mustersan of Chubut, Argentina. A. lingual upper molars, MLP 61-IV-9-1, slightly worn left maxilla with P1-M3 pronounced parastyle ridge on molars, and large from del as 67-11- Laguna Mate, shown right. B. MLP size; differs from Oldfieldthomasiidae in more 27-359, heavily worn palate with left and right II -Ml, hypsodont cheek teeth and enlarged incisors; dif- right M2 and partial M3 from La Gran Hondonada. fers from in size and lack of Scale bar = 1 cm. Notopithecinae large bilobed incisors; lacks hypsodont, bilobed upper cheek teeth of Interatheriinae; lacks hypsodont, trilobed upper cheek teeth of Mesotheriidae; and Table 6. Measurements to the nearest 0.1 mm for specimens of Pseudhyrax (P. eutrachytheroides: SGOPV 2985 and tentatively SGOPV 2877; P. strangulatus: SGOPV 2887; Pseudhyrax sp. indet.: SGOPV 2901). Measurements in parentheses are estimated. Upper dentition P2 P3 P4 Ml M2 M3 Specimen lacks hypselodont, simplified upper cheek teeth of Hegetotheriidae. Among the Archaeohyracidae, the following characters suggest affinities with Pseudhyrax: a a ?> hypsodonty greater than that exhibited by Eohy- rax, but less than that in Archaeotypotherium, ^ 5) Protarchaeohyrax, and Archaeohyrax; upper 3 C c c g a 3 U ^ ^> cheek teeth with prominent paracone and para- S R -S S <4i s 00.^ 5 '> style ridges on molar ectolophs; accessory fos- a s 00 s § § ^ a •C 'CO, settes on molars even after moderate wear; <;> ^ . ^ "^ 'u ^ * present -s: -c '<-' -s; a. CQ ^ ^ .« of central oriented anter- o o ^ o o o presence elongate fossa, g^g olabially; and presence of a slight sulcus on en- ^ i;j -s: 1^ ^ S3 c c doloph dividing protocone and hypocone. as - 53- -Q_ a a a tjtj'4>'«U^!jJjP The total absence of the maxillary bones (and i^ !., to t»5 l« lO l". 1.. > =n 1«. Otf a associated alveoli) does not permit the presence or absence of I2-P1 to be ascertained. Although it has been speculated that a significant diastema could have been present (Wyss et al., 1994; Croft, 2000), the presence of a closed dentition is equal- ly plausible. Nothing in the morphology of the preserved dentition precludes referral to Pseudhy- — o ?^ pn o ?N O O ?a decrease and the teeth to become shorter (antero- "^ -c « ^ Such s -c .« > posteriorly), wider, and more quadrangular. <^ ^ ^ > a hypothetical specimen would likely be very sim- ilar to SGOPV 2877. Unfortunately, few Pseudhy- rax specimens containing the upper cheek tooth series are known, and so a direct comparison be- tween SGOPV 2877 and a Pseudhyrax specimen of comparable wear state is not possible. MLP 67- >4> .a a -" 11-27-359 (Fig. 16B) represents Pseudhyrax pal- < Q fo o !2 !2 c^3, rS ."2 a reasonable 2 3 t" that these represent [« o o o o 53 (U specimens may CQ cd c^ c^ Ci. V C §" S:? wear series for Pseudhyrax. 1§ j= x: ^3 s^ h o o o o o o If the dentition of SGOPV 2877 does u u upper « ca it solves the of x: -c represent Pseudhyrax, "problem" O O CJ p lack of candidates for the lower dentition C3 rt csn,«cacac3oa,.t: the good of SGOPV 2877 (see further discussion below); there do exist good candidates, they just were never recognized as such. Similarly, the recogni- also ro — r- r-~ O — r~ 00 Tt tion of SGOPV 2877 as Pseudhyrax provides (N in t-- 00 o o — — IT) 00 vo 00 00 00 00 ON 0\ o^ cs a^ CTn o of no remains referable CROFT ET AL.: LARGE ARCHAEOHYRACIDS FROM CHILE AND PATAGONIA 29 sembles specimens of Pseudhyrax, but the "bi- zarre anteroposteriorly shortened talonid on m2" and "its posteroexternal termination which forms a tight, nearly 90° corner" make this taxon "un- mistakably distinct from any archaeohyracid known." A similar shortening of the talonid is described in the new archaeohyracid taxon from Antofagasta de La Sierra (Lopez, 1997; Reguero & Lopez, 1999), an archaeohyracid allied to Pseudhyrax (Reguero & Lopez, 1999, in prep.) from probable Mustersan-aged strata from north- west Argentina (Lopez, 1997). However, the de- gree to which the m2 talonid is compressed and Fig. 17. SGOPV 2901, Pseudhyrax sp. indet., left the abrupt angle of the posteroexternal corner of mandibular fragment with partial ml-m3 in oblique la- = SGOPV 2901 make the specimen distinct from bial view (left) and occlusal view (right). Scale bar 1 cm. any other notoungulate currently known. Wyss et al. (1994) referred SGOPV 2901 to "archaeohyracid new taxon C" and suggested, would have been present, but was unrecognized. based on size and morphology, that it might rep- As P. eutrachytheroides is the better known of the resent a portion of the lower dentition of what two species of Pseudhyrax, and because SGOPV they called "archaeohyracid new taxon B" 2877 is similar in size and morphology to Argen- (SGOPV 2877, discussed above). The plausibility tine specimens referred to P. eutrachytheroides, of SGOPV 2877 and SGOPV 2901 representing SGOPV 2877 is tentatively referred to that taxon. the same taxon was also supported by Croft (2000). If the conclusions reached above regard- ing the taxonomic affinities of SGOPV 2877 are Pseudhyrax strangulatus (Ameghino, 1901) correct (viz., that SGOPV 2877 is referable to (Figure 14A) Pseudhyrax eutrachytheroides), the potential for the referral of SGOPV 2901 to that taxon warrants HoLOTYPE—MACN 10774, partial right man- consideration. dible with p4-m2. Indeed, certain attributes of SGOPV 2901 sup- Referred Specimen—SGOPV 2887, right man- port the interpretation that the very unusual fea- dibular fragment with m2-3 (Fig. 14A). tures of this specimen may be the result of incom- Type Locality—Unknown. plete preservation and post-mortem deformation, Age and Distribution—As for Pseudhyrax. rather than peculiar attributes of a new taxon. Diagnosis—Differs from P. eutrachytheroides First, although the lingual portions of the talonids in its smaller size. of both ml and m2 are preserved, only the talonid Comments—SGOPV 2887, a right mandibular of m2 exhibits the unusual, 90° posteroexternal fragment with m2-3, is morphologically similar termination. In other notoungulates, the morphol- to relatively unworn specimens of Pseudhyrax ogies of the talonids on m 1 and m2 are generally from Argentina. Based on its size, SGOPV 2887 very similar to each other. This suggests that one is referred to P. strangulatus (see Fig. 13 and Ta- of the two talonids in SGOPV 2901 may have ble 6). been deformed, probably the talonid of m2. Sec- ond, a slight ridge is present on the lateral surface of the mandible in SGOPV 2901, oriented nearly Pseudhyrax sp. indet. perpendicular to the long axis of the mandible. (Figure 17) The ridge terminates at the base of the talonid of m2 and looks as if it might have resulted from a Referred Specimen—SGOPV 2901, left man- small anteroposterior compression fold in that dibular fragment with partial ml-m3 (Fig. 17), portion of the mandible. If this deformation took from the Tinguirirican-aged Tinguiririca Fauna of place plastically, it might have altered the typical Chile (see above for age discussion). rounded talonid morphology (exhibited by ml) Comments—SGOPV 2901 is a peculiar speci- into the peculiar, sharply angled morphology ex- men. As noted by Wyss et al. (1994: 18), it re- hibited by m2. Third, the talonid of ml presents 30 FIELDIANA: GEOLOGY a small fossettid directly posterior to the central 2. Cementum: absent (0); present (1). fossa, a condition characteristic of Pseudhyrax; no 3. Relative size of II: mesiodistal length < other characters of ml, the trigonid of m2, or m3 50% premaxilla length (0); mesiodistal preclude assignment to that taxon. Finally, the di- length ^ 50% premaxilla length (1). that 2901 mensions of at least m3 suggest SGOPV 4. Enamel on 1 1 : present on anterior and pos- falls within the range exhibited by Argentine and terior faces (0); present only on anterior face Tinguirirican Pseudhyrax material (Fig. 13; Ta- (1). ble 6). 5. 12: present (0); significantly reduced (peg- into consideration these Taking observations, like and < 50% size of M 1 ) or absent ( 1 ). is a reasonable that SGOPV 2901 there possibility 6. Upper dentition: closed (0); with diastemata How- represents Pseudhyrax eutrachytheroides. separating anterior teeth (1). if the unusual structure of the talonid is not ever, 7. Upper molars: with lingual sulcus (0); with- due to deformation, the specimen surely repre- out lingual sulcus (1). sents a new allied to the archaeohyracid, perhaps 8. Accessory fossettes on upper molars after new taxon from de archaeohyracid Antofagasta isolation of central fossa: present (0); absent La Sierra & 1999). (Lopez, 1997; Reguero Lopez, (1). Given the abundance of founded and/or poorly 9.* Shape of lingual notch separating upper mo- names that exist for so synonymous already many lar protocone and hypocone; broad-based, we refrain from proposing a new notoungulates, "u-shaped" (0); sharp-based, "v-shaped" taxon based on SGOPV 2901, at least until ad- (1); no notch present (2). ditional come to light that further clar- specimens 10. M3: without posterior lobe (0); with poste- ify the specimen's affinities. SGOPV 2901 is rior lobe formed by metastyle (1). therefore referred to Pseudhyrax sp. provisionally 11. Lower incisors: implanted subvertically (0); indet. markedly procumbent (1). 12. Relative size of p3 and p4: smaller than an- terior molars (0); similar in size to anterior molars (1). Phylogenetic Relationships 13. Lower molar central trigonid fossettid: ab- sent (0); present (1). of To determine the phylogenetic relationships 14. Lower molar lingual fossettid between tri- the new Tinguirirican archaeohyracids, a prelim- gonid and talonid: absent (0); present (1). inary analysis of the five currently recognized ar- 15. Lingual side of lower molars: with sulcus was undertaken. Two chaeohyracid "genera" (0); flat after advanced wear (1); present as well-known basal were chosen as out- typotheres salient, straight wall (2). groups, Oldfieldthomasia (Oldfieldthomasiidae) 16. Lower molar shape: varies little with wear and A Acropithecus (Archaeopithecidae). repre- (0); develops lingual talonid extension with sentative was also hegetotheriid (Hegetotherium) heavy wear (1). included in the since recent analysis, many phy- 17.* Shape of lingual notch separating m3 ento- have that logenetic analyses suggested hegetoth- conid and hypoconulid: narrow and deep eriids are nested within the series of taxa tradi- (0); wide and shallow (1); notch absent (2). included with the tionally Archaeohyracidae (Ci- 18. Small fossettid between m3 entoconid and felli, 1993; Hitz, 1995; Croft, 1998; Reguero, hypoconulid: absent (0); present (1). The taxa were scored 1999; Croft, 2000). eight 19. Labial surface of m3 talonid: relatively for 22 craniodental characters, most modified smooth (0); with sulcus between hypoconid from the more detailed analyses in Croft (2000). and hypoconulid (1); with pronounced notch The complete list of characters follows. Charac- (2). ters with derived character states are or- multiple 20. Carotid foramen: located posterior to audi- dered unless designated by an asterisk. The char- tory bulla (0); shifted anteriorly, medial to acter-taxon matrix is presented in Table 8. auditory bulla (0). in bulla: absent 2 1 . Vertical auditory (0); 1. Height of cheek tooth crowns: brachydont, septum HI < 1.0 (0); moderately hypsodont, 1.0 < present (1). 22. Hamular of (0); HI < 1.75 (1); very hypsodont, HI > 1.75 process pterygoid: present (2); rootless, HI undefined (3). absent (1). CROFT ET AL.: LARGE ARCHAEOHYRACIDS FROM CHILE AND PATAGONIA 31 Table 8. Matrix of characters used in phylogenetic analysis of archaeohyacid relationships. Oldfieldthomasia ium [ACCTRAN optimization] or independently derived in Pseudhyrax, Archaeotypotherium, and Acropithecus Archaeohyrax [DELTRAN optimization]); No. Eohyrax 19(1): labial surface of m3 talonid with sulcus be- tween hypoconid and hypoconulid (lost in Prot- Pseudhyrax archaeohyrax). Eohyrax is located at the base of the tree in a Archaeotypotherium polytomy with the two outgroup taxa and the Protarchaeohyrax clade of remaining ingroup taxa. Although this topology does not indicate that Eohyrax and other Archaeohyrax archaeohyracids share a most recent common an- cestor exclusive of the the Hegetotherium (MRCA) outgroups, presence of a single derived character in Eohyrax Fig. 18. Strict consensus tree representing the phy- and all archaeohyracids for which the character logenetic relationships among archaeohyracids. The out- could be coded [No. 4(1): lack of enamel on the groups for the analysis are Oldfieldthomasia (Oldfield- thomasiidae) and Acropithecus (Archaeopithecidae). posterior face of II] suggests more detailed anal- = Rescaled Consistency Index (CI) 0.78; Consistency yses will find greater Eohyrax = = support grouping Index (RCI) 0.59; Retention Index (RI) 0.76; = and other archaeohyracids. Based on their distri- Length 36 steps. bution, the derived character states present in Ac- ropithecus appear to have been acquired indepen- dently in some archaeohyracids and/or Hegeto- and phies uniting Protarchaeohyrax Hegetother- therium. ium to the exclusion of Archaeohyrax were iden- A notable feature of this analysis is the strong tified.) congruence between the branching order of the The taxa forming the focus of the present study, cladogram and the relative stratigraphic positions Archaeotypotherium and Pseudhyrax, represent of the taxa (cf. Fig. 18 and Table 9); in no instance successive outgroups to the Protarchaeohyrax- does the proximal outgroup to a clade have a first Archaeohyrax-Hegetotherium triad. Five synapo- appearance later than any of the ingroup taxa. morphies Hnk Archaeotypotherium with these Since many of the derived character states in this three taxa, making it the most robustly supported analysis are likely correlated with the consump- clade in the analysis: No. 1(2—3): HI of cheek tion of more abrasive vegetation (e.g., higher HI teeth > 1.75; No. 6(1): upper dentition with dia- values, simplified occlusal surfaces in the cheek stemata separating anterior teeth; No. 7(1): upper teeth, increased emphasis on the anterior dentition molars without lingual sulcus (reversed in Protar- and cheek teeth), the tight correlation between chaeohyrax); No. 15(1 ~2): lingual side of lower cladogenesis and morphologic evolution suggests molars flat; No. 17(1): wide and shallow lingual that dietary selection strongly influenced archaeo- notch separating m3 entoconid and hypoconulid hyracid evolution. The most robustly supported (the notch is absent in Hegetotherium). The three clade in the analysis {Archaeotypotherium, Pro- basicranial characters (Nos. 20-22) and two in- tarchaeohyrax, Archaeohyrax, Hegetotherium) re- cisor characters (3 and 1 1 ) represent derived cords its first appearance in the earliest Oligocene states shared by at least these taxa (and possibly SALMA (Flynn et al., 2(X)3); this more basal archaeohyracids as well, although the Tinguirirican an initial coin- lack of skulls and anterior dentitions for those ear- implies divergence approximately cident with, and possibly causally related to, the lier diverging taxa precludes more definitive iden- climatic deterioration (see be- tification of the node[s] to which these synapo- Eocene-Oligocene low). From a phylogenetic perspective, most (if morphies pertain). not of the characters used to Three synapomorphies diagnose Pseudhyrax all) conventionally (enlarged and hypselo- and later diverging archaeohyracids (plus Hege- diagnose hegetotheriids dont cheek teeth, lack totherium) to the exclusion of Eohyrax (the ear- II, hypselodont complete of dental fossettes and Cifelli, 1993; liest diverging archaeohyracid): No. 13(1) pres- fossettids; can be as ence of lower molar central trigonid fossettid (lost Reguero, 1998; Croft, 2(X)0) interpreted of derived character states in Hegetotherium); No. 18(1): presence of small elaborations already fossettid between m3 entoconid and hypoconulid present in the later diverging archaeohyracids. It (either lost in Protarchaeohyrax and Hegetother- is therefore not surprising that Archaeohyracidae CROFT ET AL.: LARGE ARCHAEOHYRACIDS FROM CHILE AND PATAGONIA 33 (in the classical sense of that name) is now shown totherium) than to Mesotherium, Interatherium, to be paraphyletic. Acropithecus, or Oldfieldthomasia. From the perspective of phylogenetic taxonomy Perhaps the more difficult question is which (de Quieroz & Gauthier, 1990), however, whereby names to attach to the major subclades of typoth- names are defined by linking them to specified eres about whose existence we are confident. The clades, names refer to monophyletic groups (by names most commonly associated with the groups definition). Thus, if one were to fashion a phylo- discussed herein are the "family" names Ar- least inclusive clade in- genetic definition for the chaeohyracidae and Hegetotheriidae. If conserv- and Ar- cluding Eohyrax Archaeohyrax (e.g., ing both of these long-used names is considered = of those two taxa chaeohyracidae MRCA plus desirable, Archaeohyracidae could be defined all its would re- descendants), Archaeohyracidae such that it refers to all taxa previously referred fer to a clade to have as one of its happening to both of these "families" collectively. As noted members a of group (viz., Hegetotheriidae) above, however, having Hegetotheriidae nested taxonomic rank least as far as "equivalent" (at within another clade bearing a name with the traditional taxonomic suffixes are concerned). same suffix (Archaeohyracidae) could potentially Whether or not this "clash of suffixes" is viewed cause confusion. This problem might be circum- as the significant, present phylogenetic analysis vented by amending Hegetotheriidae to Hegetoth- to the need for a review of points comprehensive eriinae (and changing the included groups Hege- the names associated with major clades of typo- totheriinae and Pachyrukhinae to Hegetotherini there notoungulates. and Pachyrukhini, respectively), but this tactic Based on the topology obtained in this analysis, would not preserve the two "family" names (the the species classically referred to as archaeohy- original intention of such a scheme) and would racids and hegetotheriids (with the possible ex- likely cause more confusion than clarification. Al- ception of Eohyrax) share an exclusive common ternatively, the name Hegetotheriidae could be de- ancestry, and thus form an evolutionary entity fined such that it applies to hegetotheriids plus the worthy of naming. A phylogenetic definition of a taxa traditionally termed archaeohyracids, but name for this clade can be fashioned any number again, this arrangement would fail to preserve the of ways, each with pros and cons. For example, two "family" names. In either case, these names in naming this clade, the clade itself might be de- would dramatically differ from their traditional scribed (in the definition statement) with reference conceptions. to the MRCA of its earliest diverging member Rather than employ a traditional "family" {Eohyrax) and some other constituent (e.g., Ar- name to refer to the clade including both archaeo- chaeohyrax, Hegetotherium), plus all of its de- hyracids and hegetotheres, an alternative would scendants. Which taxon is specified in the second be to redefine a name traditionally used to encom- position of this definition statement is relatively pass a similar group. One obvious choice would unimportant, as any number of options would en- be Simpson's Hegetotheria (Simpson, 1945). This compass all species presently termed archaeohyr- suborder of Notoungulata originally included only acids or hegetotheriids and would exclude other the Hegetotheriidae, but was later modified to also notoungulates. include the Archaeohyracidae (Simpson, 1967). It One potential disadvantage of phrasing a defi- would thus seem reasonable to tie either a node- nition in such a node-based fashion is that it based or a stem-based definition to this name to would not accommodate the discovery of earlier both and diverging members of the clade. A stem-based def- encompass hegetotheriids archaeohyra- concordant with of the inition (sensu de Queiroz & Gauthier, 1990), how- cids, Simpson's conception Another would be ever, would overcome this problem. Mesotheri- group. option Hegetotheroidea, Romer as both ar- idae has been advocated as the proximal outgroup proposed by (1966) including and Since Romer's to archaeohyracids plus hegetotheriids (e.g., Ci- chaeohyracids hegetotheriids. of the association between archaeo- felli, 1993; Croft, 2000), but the interrelationships recognition and was a of typotheres have yet to be thoroughly examined. hyracids hegetotheriids proposed year Pending a comprehensive examination of the phy- prior to Simpson's revision of Hegetotheria, use logenetic relationships among typotheres, a stem- of the name Hegetotheroidea would be an appro- based definition for archaeohyracids plus hege- priate tribute. Similar (but novel) names based on totheriids could be formulated as all typotheres the earliest diverging members of the group (e.g., more closely related to Archaeohyrax (or Hege- Archaeohyracia, Archaeohyracoidea) or a combi- 34 FIELDIANA: GEOLOGY nation of the members (e.g., Archaeohegetotheria, took place over a relatively short period of time— could also be used. Hegetoarchaeohyracoidea) the entire stratigraphic range of the group may be of Some combination these higher-level and/or as little as 15 million years, depending on the age traditional "family" names could be used if there of the beginning of the Casamayoran SALMA were sufficient justification to attach names to (wherein the first archaeohyracids are recorded). multiple components of the preferred phylogeny. With the exception of the Archaeopithecidae (a The precise phrasing of such definitions may be small group of basal notoungulates known exclu- tailored to insure nomenclatural stability in the sively from the Casamayoran) and Campanorci- face of potential future changes in topology. dae (a monotypic family, also from the Casama- None of the potential combinations of defini- yoran), this is likely the shortest range of any of tions and names discussed above is ideal; each the 14 traditional notoungulate "families." represents a compromise. As the present phylo- Although the reason for such a rapid diversifi- genetic study is not a comprehensive examination cation of archaeohyracids is unknown, its tem- of relationships among archaeohyracids and he- poral proximity to the Eocene/Oligocene bound- getotheriids, we do not propose new definitions of ary suggests the group might have been "pre- names at this time. We raise these issues to illus- adapted" to exploit, or at least were able to re- trate potential options for doing so, since we our- spond extremely rapidly to, the changing habitats selves have grappled with the question of how associated with the Eocene-Oligocene transition best to refer to the clades in question. We defer and early Oligocene "climatic deterioration" the naming of clades to a more comprehensive (Wolfe, 1971; Prothero & Berggren, 1992; Proth- cladistic analysis. ero, 1994). Alternatively, these changes may have led to greater habitat fragmentation and increased isolation of archaeohyracid populations, with al- lochthony yielding higher levels of speciation and Conclusions cladogenesis. Current evidence suggests that the global events that marked the Eocene-Oligocene One of the most striking aspects of the Tingui- transition also resulted in cooler, drier climates ririca Fauna of Chile is its high diversity of ar- and more open habitats in South America (e.g., chaeohyracids. In the preliminary description of MacFadden, 1985; Pascual & Ortiz Jaureguizar, the fauna, Wyss et al. (1994) argued that no fewer 1990; Wyss et al., 1993, 1994, 1996; Pascual et than five and perhaps as many as nine archaeo- al., 1996; Flynn & Wyss, 1998, 1999; Kay et al., hyracid taxa might be represented. The present 1999; Croft, 2001; Flynn et al., 2003). As modem study (combined with the findings of Reguero et mammalian herbivores that graze in open habitats al., 2(X)3) recognizes at least six archaeohyracid tend to exhibit high levels of hypsodonty (Janis taxa from the Tinguiririca Fauna, and notes the 1984, 1988, 1990, 1995), the environmental possibility of a seventh. The revised identifica- changes associated with the Eocene-Oligocene tions of archaeohyracid specimens considered in transition would likely have favored more hyp- Wyss et al. (1994) are listed in Table 7. sodont herbivores. Indeed, most clades of early This remarkably high diversity represents the Oligocene Tinguirirican notoungulates exhibit peak for the Archaeohyracidae. Besides Tinguirir- sharply increased hypsodonty as compared with ica, no single fauna is known to have had more their late Eocene Mustersan relatives (e.g., inter- than three contemporaneous archaeohyracid spe- atheriids, archaeohyracids, notohippids), suggest- cies, and among South American Land Mammal ing across-clade responses to these changing hab- Ages, the Tinguirirican records the occurrence of itats (Flynn et al., 2003). Since archaeohyracids eight of the 15 currently recognized species, al- were among the first hypsodont members of the most triple that of any other SALMA (no others Notoungulata (as illustrated by the late Eocene have more than three species. Table 9; see also Casamayoran Eohyrax; Simpson, 1967), the Flynn et al., 2(X)3, Table 2). With increasingly group might have diversified so rapidly during the precise age constraints for middle Cenozoic South Tinguirirican as a result of already having initi- American fossil localities (Flynn & Swisher, ated hypsodonty and thus being "better posi- to the 1995; Madden et al., 1997; Kay et al., 1999; tioned" to quickly respond evolutionarily Flynn et al., 2003), it has become apparent that dramatically changing climate and habitats. This the archaeohyracid radiation that resulted in this advantage seems to have been short-lived, how- peak diversity during the Tinguirirican SALMA ever, as archaeohyracids are last known from the CROFT ET AL.: LARGE ARCHAEOHYRACIDS FROM CHILE AND PATAGONL\ 35 ium. Boletfn del Instituto 18: mostly late Oligocene Deseadan SALMA. Filling Geografico Argentino, 406-521. a niche that presumably was similar to that of the 1901. Notices preliminaires sur des bngules hypsodont archaeohyracids, other small notoun- nouveaux des terrains cretaces de Patagonie. Boletfn with cheek gulates hypselodont (ever-growing) de la Academia Nacional de Ciencias de Cordoba, 16: teeth (e.g., hegetotheriids, interatheriids, mesoth- 350-426. the late eriids) diversified during Oligocene (De- . 1902. Notices preliminaires sur des mamiferes seadan SALMA) and into the early Miocene. nouveaux des terrains cretaces de Patagonie. Boletin de la Academia Nacional de Ciencias de Cordoba, 17: Nevertheless, Tinguirirican SALMA assemblages 5-73. document a short, previously unrecognized burst Bond, M., G. Lopez, and M. Reguero. 1996. "Astra- of exceptional evolutionary success for Archaeo- ponoteen plus superieure" of Ameghino: Another in- the first domi- hyracidae, notoungulate "family" terval in the Paleogene record of South America. Jour- nated by hypsodont species. nal of Vertebrate Paleontology, 16(Suppl. to No. 3): 23A. . 1997a. Rocas Bayas, una localidad fosilifera Paleogena de la Provincia de Rio Negro, Repiiblica Argentina. Ameghiniana, 34(4): 533. Acknowledgments Bond, M., M. Reguero, G. Lopez, A. A. Carlini, F. GoiN, R. H. Madden, M. G. Vucetich, and R. F. Kay. 1997b. The "Astraponoteen plus superieur" (Paleo- This publication was critically reviewed and gene) in Patagonia. Ameghiniana, 34(4): 533. improved by the insightful comments of Guiller- Charrier, R., A. R. Wyss, J. J. Flynn, C. C. Swisher, Bruce and an re- mo Lopez, Shockey, anonymous M. A. Norell, F. Zapatta, M. C. McKenna, and M. viewer. We thank Daniel Frassinetti and the Mu- J. Novacek. 1996. New evidence for late Mesozoic- Cenozoic evolution of the Chilean Andes in the seo Nacional de Historia Natural (Santiago), and early upper Tinguiririca Valley (35°S), central Chile. Jour- the Consejo de Monumentos Naturales de Chile, nal of South American Earth Sciences, 9(5/6): 393- for their long-term support of our Andean work. 422. Charrier shared his ex- Reynaldo freely geological Charrier, R., A. R. Wyss, M. A. Norell, J. J. Flynn, pertise in the area. He, along with Gabriel Car- M. J. Novacek, M. C. McKenna, C. C. Swisher III, rasco and numerous others, provided invaluable D. Frassinetti, and P. Salinas. 1990. Hallazgo de mamiferos fosiles del Terciario Inferior en el sector de assistance in the field. Barry Albright, Jose Bon- Termas del Flaco, Cordillera Principal, Chile Central: aparte, Bruce MacFadden, and Rosendo Pascual implicaciones paleontologicas, estratigraficas y tecton- allowed access to under their care. specimens icas. Actas Segundo Simposio sobre le Terciario de Richard Cifelli graciously provided access to Bry- Chile, Concepcion, pp. 73-84. an Patterson's unpublished manuscript on Desea- Cifelli, R. L. 1993. The phylogeny of the native South American 195-216. In F. M. dan archaeohyracids. Mark Widhalm (FMNH) ex- ungulates, pp. Szalay, S., J. Novacek, and M. C. McKenna, eds., Mammal Phy- ecuted many of the photographs, and Marlene logeny: Placentals. Springer-Verlag, New York. Donnelly created the line drawings. This work Croft, D. A. 1998. Experiments in herbivory: Evolution would not have been without the skill possible in the Archaeohyracidae (Mammalia: Notoungulata). and dedication of Andrew Lehman, Robert Ma- Journal of Vertebrate Paleontology, 18(Suppl. to No. sek, William Simpson, and the late Steve Mc- 3): 36A. . Notoun- CarroU, who prepared the exceptionally challeng- 2000. Archaeohyracidae (Mammalia, gulata) from the Tinguiririca Fauna, central Chile, and ing material. Work was supported by NSF grants the evolution and paleoecology of South American DEB-9020213, 9317943, and 9318126 (J.J.F and mammalian herbivores. Ph.D. diss.. University of Chi- a John Simon Memorial A.R.W); Guggenheim cago, 311 pp. Foundation and the Hinds Fund fellowship (J.J.F); . 2001. Cenozoic environmental change in South and NSF Biodiversity Training Grant (GRT- American as indicated by mammalian body size dis- tributions and 7: 9355032) from the University of Chicago, and the (cenograms). Diversity Distributions, 271-287. Paleobiological Fund (D.A.C.). DE Queiroz, K., and J. Gauthier. 1990. Phylogeny as a central principle in taxonomy: Phylogenetic defini- tions of taxon names. Systematic Biology, 39: 307- 322. J. and A. R. Literature Cited Flynn, J., M. A. Norell, C. C. Swisher, Wyss. 1991. Pre-Deseadan, post-Mustersan mammals of Vertebrate Ameghino. F. 1897. Les Mammiferes cretaces de from central Chile: An update. Journal to No. 29A. I'Argentine. Deuxieme contribution a la connaissance Paleontology, ll(Suppl. 3): de la faune mammalogique des couches a Pyrother- Flynn, J. J., and C. C. Swisher III. 1995. Cenozoic 36 FIELDIANA: GEOLOGY South American Land Mammal Ages: Correlation to gonia. Runford Press, Concord, New Hampshire, 232 global geochronologies, pp. 317-333. In Berggren, W. pp. D. V. Kent, M.-P. and J. Hardenbol, eds., A., Aubry, Lopez, G. M. 1997. Paleogene faunal assemblage from Time Scales, and Global Geochronology, Stratigraphic Antofagasta de la Sierra (Catamarca Province, Argen- Correlation. SEPM of (Society Sedimentary Geology) tina). Palaeovertebrata, 26: 61-81. Special Publication No. 54. MacFadden, B. J. 1985. Drifting continents, mammals, J. AND A. Wyss. 1990. New Flynn, J., early Oligocene and time scales: Current developments in South from the Andean Cordillera, Chile. Journal marsupials America. Journal of Vertebrate Paleontology, 5(2): of Vertebrate Paleontology, 10(Suppl. to No. 3): 22A. 169-174. J. and A. R. Wyss. 1998. Recent advances in Flynn, J., MacFadden, B. J., K. E. Campbell, Jr., R. L. Cifelli, South American mammalian Trends in paleontology. O. SiLES, N. M. Johnson, C. W Naeser, and P. K. and Evolution, 1 1 ): 449-454. Ecology 13( Zeitler. 1985. Magnetic polarity stratigraphy and mammalian fauna of the Deseadan . 1999. New marsupials from the Eocene-Oli- (Late Oligocene- gocene transition of the Andean Main Range, Chile. Early Miocene) Salla Beds of northern Bolivia. Jour- Journal of Vertebrate Paleontology, 19(3): 533-549. nal of Geology, 93(3): 223-250. R. J. R. F. J. J. C. Flynn, J. J., A. R. Wyss, D. A. Croft, and R. Char- Madden, H., Guerrero, Kay, Flynn, Swisher A. H. 1997. RiER. (2003). The Tinguiririca Fauna, Chile: Biochro- C. III, and Walton. The Lav- entan In R. R. nology, paleoecology, biogeography, and a new ear- Stage and Age, pp. 355-381. Kay, F, liest Oligocene South American Land Mammal H. Madden, R. L. Cifelli, and J. J. Flynn, eds.. Ver- "Age." Palaeogeography, Palaeoclimatology, Pa- tebrate Paleontology in the Neotropics: The Miocene of Institution laeoecology, 195(3-4): 229-259. Fauna La Venta, Colombia. Smithsonian Press, Washington, D.C. Gingerich, p. D. 1974. Size variability of the teeth in L. R. L. R. E. and G. living mammals and the diagnosis of closely related Marshall, G., Cifelli, Drake, H. Curtis. 1986. Vertebrate sympatric fossil species. Journal of Paleontology, paleontology, geology, and of the de and Scar- 48(5): 895-903. geochronology Tapera L6pez ritt Pocket, Chubut Province, Argentina. Journal of HiTZ, R. 1995. Typothere (Notoungulata) phylogeny and Paleontology, 60(4): 920-951. proposed taxonomic revisions. Journal of Vertebrate M. J. 2002. Time Traveler: In Search of Di- Paleontology, 15(Suppl. to No. 3): 34A. Novacek, nosaurs and Ancient Mammals from Montana to Mon- HiTZ, R., M. Reguero, A. R. Wyss, and J. J. Flynn. golia. Farrar, Straus and Giroux, New York, 368 pp. 2000. New interatheriines (Interatheriidae, Notoun- Novacek, M. J., A. Wyss, D. Frassinetti, and P. Sa- gulata) from the Paleogene of Central Chile and linas. 1989. A new ?Eocene mammal fauna from the Southern Argentina. Fieldiana: Geology, New Series, 42: 1-26. Andean Main Range. Journal of Vertebrate Paleontol- ogy, 9(Suppl. to No. 3): 34A. Jams, C. M. 1984. The use of fossil ungulate commu- Pascual, R., and E. O. Jaureguizar. 1990. Evolving nities as indicators of climate and environment, pp. climates and mammal faunas in Cenozoic South 85-104. In Brenchley, P., ed.. Fossils and Climate. America. Journal of Human Evolution, 19: 23-60. John Wiley & Sons Ltd., Chichester, U.K. Pascual, R., E. Ortiz Jaureguizar, and J. L. Prado. . 1988. An estimation of tooth volume and hyp- 1996. Land mammals: Paradigm for Cenozoic South sodonty indices in ungulate mammals, and the corre- American geobiotic evolution. Miinchner Geowissen- lation of these factors with dietary preference, pp. schaftliche Abhandlungen A, 30: 265-319. 371-391. In Russell, D. E., J. R Santoro, and D. Sig- structure of the ear in ogneau-Russell, eds., Teeth Revisited: Proceedings of Patterson, B. 1936. The internal the Vllth International Congress of Dental Morphol- some notoungulates. Geological Series, Field Museum 6: 199-227. ogy. Memoires du Museum d'Histoire Naturelle de of Natural History, Paris, serie C. Prothero, D. R. 1994. The late Eocene-Oligocene ex- tinctions. Annual Review of Earth and Sci- . 1990. Correlation of cranial and dental vari- Planetary 22: 145-165. ables with dietary preference in mammals: A com- ences, parison of macropodids and ungulates. Memoirs of the Prothero, D. R., and W. A. Berggren. 1992. Eocene- Queensland Museum, 28(1): 349-366. Oligocene Climatic and Biotic Evolution. Princeton Press, Princeton, New 568 . 1995. Correlations between craniodental mor- University Jersey, pp. phology and feeding behavior in ungulates: Recipro- Reguero, M. A. 1993. Los Typotheria y Hegetotheria cal illumination between living and fossil taxa, pp. (Mammalia, Notoungulata) eocenos de la localidad 76-98. In Thomason, J. J., ed.. Functional Morphol- Canad6n Blanco, Chubut. Ameghiniana, 30(3): 336. ogy in Vertebrate Paleontology. Cambridge University . 1998. El problema de las relaciones sistemdticas Press, New York. y filogeneticas de los Typotheria y Hegetotheria Kay, R. F, R. H. Madden, M. G. Vucetich, A. A. Car- (Mammalia, tNotoungulata): Analisis de los taxones LiNi, M. M. Mazzoni, G. H. Re, M. Heizler, and H. de Patagonia de la Edad-mamifero Deseadense (Oli- Sandeman. 1999. Revised geochronology of the Cas- goceno). Ph.D. diss., Universidad de Buenos Aires, amayoran South American Land Mammal Age: Cli- Buenos Aires. matic and biotic of the Na- implications. Proceedings Reguero, M. A., and R. L. Cifelli. 1997. Deseadan tional of Sciences 13235-13240. Academy 96(23): Archaeohyracidae from Salla, Bolivia. Ameghiniana, LooMis, F. B. 1914. The Deseado Formation of Pata- 34(3): 539. CROFT ET AL.: LARGE ARCHAEOHYRACIDS FROM CHILE AND PATAGONIA 37 Reguero, M. a., D. a. Croft, J. J. Flynn, and A. R. Wyss, A. R., and J. J. Flynn. 1991. Phylogenetic and Wyss. (2003). Small archaeohyracids (Typotheria, biostratigraphic implications of a new early Oligocene Notoungulata) from Chubut Province, Argentina, and marsupial fauna from the Tinguiririca Valley, Chile. central Chile: Implications for trans-Andean temporal PaleoBios, 13(Suppl. to No. 50): 10. correlation. Fieldiana: n.s. 48: 1-17. Geology, Wyss, A. R., J. J. Flynn, M. A. Norell, C. C. Swisher Reguero, M., and G. Lopez. 1999. Un nuevo Archaeo- in, R. Charrier, M. J. Novacek, and M. C. McKen- hyracidae (Notoungulata, Hegetotheria) del Paleogeno na. 1993. South America's earliest rodent and recog- de Antofagasta de la Sierra, provincia de Catamarca, nition of a new interval of mammalian evolution. Na- Argentina. Ameghiniana, 36(1): 107. ture, 365: 434-437. Vertebrate RoMER, A. S. 1966. Paleontology. University Wyss, A. R., J. J. Flynn, M. A. Norell, C. C. Swisher of 468 Chicago Press, pp. 111, M. J. Novacek, M. C. McKenna, and R. Char- Roth, S. 1903. Noticias preliminares sobre nuevos rier. 1994. Paleogene mammals from the Andes of mamiferos fosiles del Cretaceo superior y Terciario central Chile: A preliminary taxonomic, biostrati- inferior de la Patagonia. Revista, Museo de La Plata, graphic, and geochronologic assessment. American 11: 133-158. Museum Novitates, 3098: 1-31. Shockey, B. J. 1997. Toxodontia of Salla, Bolivia (late Wyss, A. R., J. J. Flynn, C. C. Swisher III, R. Char- Oligocene): Taxonomy, systematics, and functional rier, AND M. A. Norell. 1992. Fossil mammals from morphology. Ph.D. diss.. University of Florida, the central Chilean Andes: A new interval in the Gainesville, 275 pp. South American land mammal succession, and impli- cations for events Simpson, G. G. 1945. The principles of classification and Eocene-Oligocene boundary and a classification of mammals. Bulletin of the American Andean tectonics. Paleontological Society, Special Museum of Natural History, 85: 1-350. Publication, 6: 318. J. J. . 1967. The beginning of the age of mammals in Wyss, A. R., M. A. Norell, and Flynn. 1993. An South America. Part IL Bulletin of the American Mu- exceptional archaeohyracid fauna from the Tinguirir- seum of Natural History, 137: 1-260. ica River valley of central Chile. Journal of Vertebrate to No. 64A. Sinclair, W. J. 1909. Typotheria of the Santa Cruz beds, Paleontology, 13(Suppl. 3): pp. 1-110. In Scott, W. B., ed.. Reports of the Prince- Wyss, A. R., M. A. Norell, J. J. Flynn, M. J. Novacek, ton University Expeditions to Patagonia, 1896-1899. R. Charrier, M. C. McKenna, C. C. Swisher, D. Princeton University, Princeton, NJ. Frasinetti, and M. Jin. 1990. A new early Tertiary central Chile: for Wolfe, J. A. 1971. Tertiary climatic fluctuations and mammal fauna from Implications and tectonics. Journal of Verte- methods of analysis of Tertiary floras. Palaeogeogra- Andean stratigraphy 518-522. phy, Palaeoclimatology, Palaeoecology, 9: 27-57. brate Paleontology, 10(4): Wyss, A. R., R. Charrier, and J. J. Flynn. 1996. Fossil Wyss, A. R., M. A. Norell, M. J. Novacek, and J. J. mammals as a tool in Andean stratigraphy: Dwindling Flynn. 1992. New ?early Tertiary localities from the evidence of Late Cretaceous volcanism in the South Chilean Andes. Journal of Vertebrate Paleontology, Central Main Range. PaleoBios, 17(2-4): 13-27. 12(Suppl. to No. 3): 61 A. 38 FIELDIANA: GEOLOGY UNIVERSriY OF ILLINOI8-URBANA 3 0112 060367569 74 74 Field Museum of Natural History 1400 South Lake Shore Drive Chicago, Illinois 60605-2496 Telephone: (312) 665-7055