Paleoecological Implications of the Craniodental and Premaxilla Morphologies of Two Rhinoceroroids (Perissodactyla) from Badlands National Park, South Dakota

Alfred J. Mead1 and William P. Wall2

1University of Nebraska State Museum, Division of Vertebrate Paleontology, Lincoln, NE 68588-0514
2Department of Biology, Georgia College and State University, Milledgeville, GA 31061

Abstract—Analysis of the craniodental and premaxilla morphologies suggests differing feeding strategies for two late Eocene/Oligocene rhinocerotoids, Hyracodon and Subhyracodon, from Badlands National Park, South Dakota. The relatively longer upper and lower premolar rows, wider central incisors, relatively wider and more stout premaxillae, complete anterior dentition, and strong cingula indicates that Hyracodon was morphologically similar to modern browsers and mixed feeders. The proportionately shorter premolar rows, enlarged lateral incisors, relatively narrower and more delicate premaxillae, and less well developed cingula in Subhyracodon illustrates similarities to modern grazers and mixed feeders. Hyracodon was likely a browser of nuts, twigs, and tougher vegetation growing on the distal reaches of a bushland floodplain. Subhyracodon appears to have been a selective-mixed feeder utilizing the more high-fiber vegetation of wooded habitats proximal to the Oligocene water courses.


Athorough understanding of the paleoautecology of extinct mammalian species enables researchers to generate more accurate interpretations of paleoenvironments. Among the ecological attributes, dietary preference is possibly the most informative. Various cranial morphological traits (e.g. premaxilla shape) have been used to determine dietary preferences of extinct ungulate species (Gordon and Illius, 1988; Solounias and Saunders, 1988; Solounias et al., 1988; Solounias and Moelleken, 1993a; Dompierre and Churcher, 1996). Tooth microwear analysis (Solounias and Moelleken, 1992a, b, 1993b, 1994), tooth wear facet analysis (Janis, 1990a), and enamel isotope composition (Wang et al., 1994) have also been used to discern dietary preference. Additionally, craniodental indices have been employed to compare and contrast ungulate feeding strategies (Janis, 1988, 1990b, c, 1995; Janis and Ehrhardt, 1988). These craniodental indices include the hypsodonty index (HI), relative muzzle width (RMW), relative incisor width (RIW), and relative lengths of the upper and lower premolar series (RLPM).

Hyracodon (Hyracodontidae) and Subhyracodon (Rhinocerotidae) were two temporally sympatric rhinocerotoids that inhabited the area of Badlands National Park (BADL), South Dakota, during the latest Eocene through the late-middle Oligocene (Emry et al., 1987). Their fossilized remains are commonly found in the same White River Group horizons within BADL, but usually in differing lithofacies. This apparent dichotomous facies distribution has led researchers to suggest that Hyracodon was an open plains dweller and Subhyracodon was a denizen of the riparian strip (Matthew, 1901; Clarke et al., 1967). The purpose of this study is: 1) to utilize craniodental indices and premaxilla shapes as indicators of dietary preferences in Hyracodon and Subhyracodon;
and 2) to speculate on modern analogues for these two Oligocene rhinocerotoids.

Materials and Methods

The Hyracodon and Subhyracodon material examined in this study was collected from BADL and is housed in the Georgia College & State University Vertebrate Paleontology (GCVP) and South Dakota School of Mines (SDSM) collections. Only molars that were fully erupted and exhibited light wear were used for the calculation of HI values (HI = H/W). The heights and widths of m3's were taken at the protoconid on the labial side of the lower molars. The heights and widths of M3's were measured on the labial side at the paracone. Height was measured as the distance from the occlusal surface to the dentine/enamel junction, and the width measured as the maximum tooth width normal to the trend of the cheek tooth row. For the calculation of the relative muzzle width (RMW = MW/PW), palatal width (PW) and muzzle width (MW) were measured between the M2 protocones and at the maxilla/premaxilla suture respectively. The lower incisors were measured at their widest point above the alveolar rim for the calculation of the relative incisor width (RIW = i1/i2 or i1/i3). To determine the relative length of the premolar series (RLPM = PM/M), upper and lower premolar (PM) and molar series (M) were measured parallel to the labial side of the cheek tooth row at bone height. Additional RLPM measurements were obtained from the extant rhinoceroses Ceratotherium simum, Diceros bicornis, Rhinoceros unicornis, and the Miocene rhinoceroses Aphelops and Teleoceras in the University of Nebraska State Museum (UNSM) collections. The Student's t-test was used to analyze the comparative data. Photographs of the extant Dicerorhinus sumatrensis (Groves and Kurt, 1972) were examined to determine its RLPM values. Published HI and RMW values (Janis, 1988; Janis and Ehrhardt, 1988) for extant perissodactyls with known dietary preferences and habitat usage were used for comparative purposes. Habitat usage and dietary preferences of extant ungulates comes from Nowak (1991) unless otherwise stated.

The premaxillae morphologies of Hyracodon and Subhyracodon were qualitatively compared to a variety of modern browsers (Tapirus terrestris, T. bairdii, D. bicornis, Odocoileus virginianus, O. hemionus, Giraffa camelopardalis), grazers (Equus caballus, E. burchelli, E. grevyi, C. simum), and mixed feeders (grasses represent between 10% and 90% of the diet; R. unicornis, Tayassu tajacu, Antilocapra americana, Cervus elaphus, Boselaphus tragocamelus) in the Georgia College & State University Mammal (GCM) and UNSM collections.


Both Hyracodon and Subhyracodon are brachydont, lophodont rhinocerotoids exhibiting some degree of upper premolar molarization. Hyracodon retains complete anterior dentition and strong cingula on the cheek teeth. Subhyracodon exhibits a reduced number of incisors, absence of canines, and less well developed cingula on the cheek teeth. HI values (Table 1) for both m3's and M3's are nearly identical between Hyracodon and Subhyracodon, and when compared to HI values for extant perissodactyls, fall between those of the browsing T. terrestris and the mixed feeding R. unicornis.

A discernible difference exists between the lower RLPM values for Hyracodon and Subhyracodon denoting a proportionately longer lower premolar row in Hyracodon. The rela
tive length of the upper premolar series is significantly (a = .10) longer in Hyracodon. Hyracodon and the Miocene rhinocerotid Aphelops exhibit nearly identical RLPM values while the Subhyracodon RLPM values are intermediate between those of Aphelops and Teleoceras.

The extant D. bicornis exhibits an upper RLPM ratio intermediate between those of Hyracodon and Subhyracodon, and a lower RLPM value nearly equal to that of Subhyracodon. The R. unicornis lower RLPM values are nearer those of Hyracodon while the upper RLPM values are more similar to those of Subhyracodon. Both RLPM ratios of C. simum are nearly equal to those of Subhyracodon. Photographs of the upper cheek tooth rows of Dicerorhinus sumatrensis demonstrates that the upper premolar series is longer than the molar series, similar to the upper RLPM condition of Hyracodon.

The central incisors of Hyracodon are nearly the same size as the lateral incisors while the central incisors in Subhyracodon are much smaller than the lateral incisors. The RMW of study specimens exhibiting non-distorted MW and PW is not significantly different for Hyracodon and Subhyracodon. Several extant perissodactyls exhibit RMW values similar to those of Hyracodon and Subhyracodon. The RMW value of 0.840 for Hyracodon is within 0.2 of Equus grevyi (0.822), E. hemionus (0.838), Ceratotherium simum (0.842), and Diceros sumatrensis (0.847). The average RMW of 0.910 for Subhyracodon is very similar to that of E. kiang (0.900) and within 0.025 of E. przewalskii (0.885). The relatively widest muzzles in extant perissodactyls belong to the mixed feeding R. unicornis and the browsing R. sondaicus. The browsing perissodactyls Diceros bicornis, T. indicus, and T. terrestris exhibit the relatively narrowest muzzles. However, D. bicornis lacks upper incisors and exhibits greatly reduced premaxillae.

The premaxillae of Hyracodon (Figure 1A) are rounded and stout, whereas those of Subhyracodon (Figure 1B) are more pointed and delicate. Qualitative comparisons to the artiodactyl premaxillae morphologies figured by Gordon and Illius (1988) and Solounias et al. (1988) indicate that the shape of the Hyracodon premaxillae more closely resembles the mixed feeding artiodactyls (e.g. Cervus elaphus, Fig. 1N) while the premaxillae of Subhyracodon (Figure 1B) more closely resembles those of the browsing artiodactyls (e.g. G. camelopardalis, Figure 1D). The same results are obtained when the quantitative method of premaxillae analysis (Dompierre and Churcher, 1996) is applied. In terms of robustness, the Hyracodon premaxillae is similar to that exhibited by Tayassu tajacu (Figure 1M).

A comparison of the premaxillae of Hyracodon (Figure 1A) with modern perissodactyls demonstrates a striking similarity to Tapirus terrestris (Figure 1C). None of the extant perissodactyls examined provides a reasonable modern analogue for the premaxilla shape of Subhyracodon. Unlike Hyracodon or Subhyracodon, the premaxillae of extant grazers (Figures 1G, H, I, and J) are more bulbous with broader incisor arcades. The only mixed feeding perissodactyl figured (Rhinoceros unicornis, Figure 1K) exhibits much wider premaxillae than either Hyracodon or Subhyracodon. The
Table 1—Calculated craniodental indices. ( ) = number of individuals sampled.

Oligocene rhinocerotids

Hyracodon Subhyracodon dental formula 3/3,1/1,4/3,3/3 2/2,0/0,4/4-3,3/3

Hypsodonty index (HI = H/W)

m3 light wear 0.983 (10) 0.938 (5)

M3 light wear 0.826 (13) 0.730 (5)

Relative length of premolars (RLPM = PM/M)

lower premolar series 0.869 (25) 0.783 (14) upper premolar series 1.050 (21) 0.859 (11) Relative incisor width (RIW = i1/i3 or i1/i2)

1.030 (8) 0.471 (10)

0.840 (14) 0.910 (4)

Relative muzzle width (RMW = MW/PW)

Miocene rhinocerotids

Aphelops Teleoceras RLPM = PM/M

lower premolar series 0.820 (3) 0.488 (9)

upper premolar series 1.050 (3) 0.671 (10)

Extant rhinocerotids

D. bicornis R. unicornis C. simum


lower premolar series 0.788 (1) 0.957 (1) 0.762 (2)

upper premolar series 0.901 (1) 0.737 (1) 0.819 (2)

major feeding groups for extant ungulates. When compared to grazers, browsers exhibit a low HI, long premolar row, and medium width muzzle. Grazers have a higher HI, shorter premolar row, and a relatively broader muzzle. Mixed feeders exhibit a moderate HI, short premolar row, and narrow muzzle (narrowest of the three groups). Tests of these dietary divisions on extant ungulates show that grazers and browsers are nearly always correctly identified.

The HI values for Hyracodon and Subhyracodon suggest that the cheek teeth were optimally suited for browsing. Hyracodon possessed minimally enlarged central lower incisors and Subhyracodon exhibited enlarged lateral lower incisors. Modern grazers tend to possess subequal lower incisors that are relatively broad, browsers possess enlarged central incisors, and intermediate feeders possess significantly wider lateral incisors than browsers (Janis and Ehrhardt, 1988). Direct analogy would suggest browsing habits for Hyracodon and mixed feeding habits for Subhyracodon. However, enlarged lateral lower incisors are a defining characteristic of the Rhinocerotidae and may not reflect a feeding adaptation in Subhyracodon.

The RMW values of Hyracodon are most similar to the extant E. grevyi, E. hemionus, C. simum, and Dicerorhinus sumatrensis. Equus grevyi is a grazer that consumes fibrous grasses that are inedible to cattle and other ungulates in sub-desert grasslands. Equus hemionus inhabits the flat deserts of Asia subsisting on grass and low succulent plants. Ceratotherium simum inhabits the open forests and plains of Africa consuming a variety of grasses (Groves, 1972) and D. sumatrensis inhabits hilly, humid forests, eating fruits, leaves, twigs, and bark (Groves and Kurt, 1972). The RMW of Subhyracodon is most similar to the extant E. kiang and extinct E. przewalskii. Equus kiang inhabits the Tibetan Plateau and grazes on grasses and low succulent plants. Equus przewalskii inhabited the plains and hills of eastern Europe and grazed on a variety of grasses.

The proportionately longer upper premolar row in Hyracodon is similar to the conditions found in the browsing Miocene rhinocerotid Aphelops (Prothero et al., 1989) and the extant browsing Dicerorhinus sumatrensis. The proportionately shorter premolar row in Subhyracodon is similar to the mixed-feeding R. unicornis and grazing C. simum. Both the upper and lower RLPM values of the browsing Diceros bicornis are intermediate between the values for Hyracodon and Subhyracodon. When compared to the standard craniodental morphologies determined by Janis (1995), Hyracodon most closely resembles a modern browsing ungulate while Subhyracodon favors the mixed feeding group. Well developed cingula on the cheek teeth, as exhibited by Hyracodon, are generally recognized as an indication of an herbivore utilizing thorny or rough vegetation. Less well developed cingula on the cheek teeth, as in Subhyracodon, suggests an organism not optimally suited to utilize such vegetation.

The premaxillae morphology of Hyracodon is similar to that exhibited by Cervus elaphus, Tayassu tajacu, and Tapirus terrestris. Cervus elaphus is both a grazer and a browser,
Figure 1— Premaxillae of (A) Hyracodon, (B) Subhyracodon, extant browsers [(C) Tapirus terrestris, (D) Giraffa camelopardalis, (E) Odocoileus virginianus, (F) O. hemionus], grazers [(G) Equus burchelli, (H) E. caballus, (I) E. grevyi, (J) Ceratotherium simum], and mixed feeders [(K) Rhinoceros unicornis, (L) Antilocapra americana, (M) Tayassu tajacu, (N) Cervus elaphus]. Scale bars = 4 cm.
premaxillae of C. simum (Figure 1J) and D. bicornis (not figured) are greatly reduced. C. simum, a grazer, possesses a broad, squarish lip that functions as a cropping mechanism. D. bicornis, a browser, has a narrow prehensile lip (hook lip) used to pull browse items into the mouth. Both Hyracodon and Subhyracodon lack the elevated and retracted nasals, expanded nasal incisions, and rostral muscle scars indicative of a prehensile lip or proboscis (Wall, 1980).


Paleosol analyses suggest that the paleo-vegetation of the BADL region was transitional from middle Eocene tropical forests, to early Oligocene open woodlands, to late Oligocene open bushland (Retallack, 1983). Some of the paleosols in the Pinnacles area of BADL preserve root traces suggestive of low desert scrub (Retallack, 1983). Although direct botanical evidence is sparse, wooded areas with succulent vegetation most likely existed in the riparian strips, while dry scrub and tougher vegetation was more abundant distal to the Oligocene stream courses.

Janis (1995) used the combination of HI, RLPM, and RMW to determine the general characteristics of the three utilizing fresh grasses in the spring and early summer, and browsing forbs, woody plants, shrubs and conifers in the late summer, fall, and winter. Tayassu tajacu is a browsing artiodactyl of the southwestern United States that grubs for fruits, berries, tubers, bulbs, and rhizomes. Tapirus terrestris is a browser of leaves, buds, twigs, and fruits of low-growing terrestrial plants. The premaxillae morphology of Subhyracodon is most similar to G. camelopardalis, a select browser that consumes mainly leafy vegetation from acacia, mimosa, and wild apricot trees.

Speculation concerning modern analogs for Hyracodon and Subhyracodon requires comparisons to both perissodactyls and artiodactyls. Taking into consideration the craniodental indices, incisor arcade structure, apparent habitat usage, and appendicular morphology, Tayassu tajacu of Arizona and Texas may be the most appropriate modern analog for Hyracodon. Although T. tajacu is an artiodactyl, it lacks the rumen digestive system found in the more diverse ruminants. Tayassu tajacu inhabits desert scrub and arid woodlands, escaping danger with quick bursts of speed. Biomechanical analysis of locomotor morphology suggests that Hyracodon was functionally similar to modern wild pigs (Wall and Hickerson, 1995). The robust snout of T. tajacu is used to grub for food. Cactus fruit, berries, and bulbs are the primary dietary components. A grubbing nature for Hyracodon could help to explain the presence of the robust premaxillae and nasals. Subhyracodon is enigmatic in comparison to extant ungulates. The craniodental indices suggest an organism more suited as a mixed feeder/grazer, yet the premaxilla morphology, which is very similar to G. camelopardalis, indicates a select browser. Although the premaxillae morphologies differ, the South American Tapirus terrestris may provide a legitimate ecological analogue for Subhyracodon.


The analysis of craniodental and premaxilla morphologies suggests differing feeding habits for the Oligocene rhinocerotoids Hyracodon and Subhyracodon. The relatively longer upper and lower premolar rows, wider central incisors, relatively wider and more stout premaxillae, complete anterior dentition, and well developed cingula suggests that Hyracodon was morphologically similar to modern browsers and mixed feeders. The proportionately shorter premolar rows, enlarged lateral incisors, narrower and more delicate premaxillae, and less well developed cingula in Subhyracodon suggests an herbivore more morphologically suited as a grazer and mixed feeder. The distinctive premaxillae shape exhibited by Subhyracodon is suggestive of a selective mixed feeding perissodactyl. This evidence suggests that Hyracodon was a browser of the nuts, fruits, twigs, and tougher vegetation growing on the distal reaches of bushland floodplains present in the region of BADL during the Oligocene. Subhyracodon was likely a mixed feeder utilizing the more high-fiber vegetation and succulent browse in the wooded habitats proximal to Oligocene water courses.


The senior author thanks W. Wall for suggesting this thesis project and providing continued advise and support until its completion, and D. Parmley for advice, encouragement, and many stimulating conversations. We acknowledge R. Benton of Badlands National Park for her assistance in the park, M. Voorhies and P. Freeman of the Nebraska State Museum, and P. Bjork of the South Dakota School of Mines for the use of specimens in their care. This manuscript has benefited greatly from critical reviews by M. Voorhies, B. Bailey, G. Corner, and D. Terry. We thank three anonymous reviewers for their useful comments. We thank D. Terry for his enlightening conversations dealing with the sedimentology of The White River Group. The senior author thanks H. Mead for her critical reviews, patience, and encouragement. Finally we thank Mr. Vince Santucci for his enthusiasm and support for paleontological research in the National Parks. This research was partially supported by grants from the Georgia College & State University Faculty Research Fund.


Clark, J., J. R. Beerbower, and K. K. Kietzke. 1967. Oligocene sedimentation, stratigraphy, paleoecology and paleoclimatology in the Big Badlands of South Dakota. Fieldiana: Geology Memoirs, 5:1-158.

Dompierre, H. and C. S. Churcher. 1996. Premaxillary shape as an indicator of the diet of seven extinct late Cenozoic New World camels. Journal of Vertebrate Paleontology, 16(1):141-148.

Emry, R. J., L. S. Russell, and P. R. Bjork. 1987. The Chadronian, Orellan, and Whitneyan North American Land Mammal Ages, p. 118-152. In M. O. Woodburne (ed.), Cenozoic Mammals of North America. University of California Press, Berkeley.

Gordon, I. J. and A. W. Illius. 1988. Incisor arcade structure and diet selection in ruminants. Functional Ecology, 2:15-22.

Groves, C. P. 1972. Ceratotherium simum. Mammalian Species, 8:1-6.

———, and F. Kurt. 1972. Dicerorhinus sumatrensis. Mammalian Species, 21:1-6.

Janis, C. M. 1988. An estimation of tooth volume and hypsodonty indices in ungulate mammals, p.371-391. In D. E. Russell, J. P. Santoro, and D. Sigogneau-Russel (eds.) Teeth revisited: Proceedings of the VIIth International Congress of Dental Morphology, Memoires du Musum d'Histoire naturelle, de Paris, serie C, 53.

———. 1990a. The correlation between diet and dental wear in herbivorous mammals, and its relationship to the determination of diets of extinct species, p. 241-259. In A. J. Boucot (ed.), Evolutionary paleobiology of behavior and coevolution. Elsevier, New York.

———. 1990b. Correlation of cranial and dental variables with body size in ungulates and macropodoids, p. 255-299. In J. Damuth and B. J. MacFadden (eds.), Body size in mammalian paleobiology: estimation and biological implications. Cambridge University Press.

———. 1990c. Correlation of cranial and dental variables with dietary preferences in mammals: a comparison of macropodoids and ungulates. Memoirs of the Queensland Museum, 28(1):349-366.

———. 1995. Correlations between craniodental morphology and feeding behavior in ungulates: reciprocal illumination between living and fossil taxa, p. 76-98. In J. Thomason (ed.), Functional Morphology in Vertebrate Paleontology. Cambridge University Press, New York.

———, and D. Ehrhardt. 1988. Correlation of relative muzzle width and relative incisor width with dietary preference in ungulates. Zoological Journal of the Linnean Society, 92:267-284.

Matthew, W. D. 1901. Fossil mammals of the Tertiary of northeastern Colorado. Memoirs of the American Museum of Natural History, 1(7):353-447.

Nowak, R. M. 1991. Walker's Mammals of the world. Fifth edition. The John Hopkins University Press. Baltimore, Maryland.

Prothero, D. R., C. Guerin, and E. Manning. 1989. The history of the Rhinocerotoidea, p. 321-340. In D. R. Prothero and R. M. Schoch (eds.), The Evolution of Perissodactyls. Oxford University Press, New York.

Retallack, G. J. 1983. A paleopedological approach to the interpretation of terrestrial sedimentary rocks: The mid-Tertiary fossil soils of Badlands National Park, South Dakota. Geologic Society of America Bulletin, 94:823-840.

Solounias N. and S. M. C. Moelleken. 1992a. Tooth microwear analysis of Eotragus sansanensis (Mammalia: Ruminantia), one of the oldest known bovids. Journal of Vertebrate Paleontology, 12(1):113-121.

———. 1992b. Dietary adaptations of two goat ancestors and evolutionary considerations. Geobios, 25(6):797-809.

———. 1993a. Dietary adaptation of some extinct ruminants determined by premaxillary shape. Journal of Mammalogy, 74(4):1059-1071.

———. 1993b. Tooth microwear and premaxillary shape of an archaic antelope. Lethaia, 26:261-268.

———. 1994. Differences in diet between two archaic ruminants species from Sansan, France. Historical Biology, 7:203-220.

Solounias N. and B. D. Saunders. 1988. Dietary adaptations and paleoecology of the late Miocene ruminants from Pikermi and Samos in Greece. Palaeogeography, Palaeoclimatology, Palaeoecology, 65:149-172.

Solounias N., M. Teaford, and A. Walker. 1988. Interpreting the diet of extinct ruminants: the case of a non-browsing giraffid. Paleobiology, 14(3):287-300.

Wall, W. P. 1980. Cranial evidence for a proboscis in Cadurcodon and a review of snout structure in the family Amynodontidae (Perissodactyla, Rhinocerotoidea). Journal of Paleontology, 54:968-977.

———, and W. Hickerson. 1995. A biomechanical analysis of locomotion in the Oligocene rhinocerotoid, Hyracodon, p. 19-26. In V. L. Santuci and L. McClelland (eds.), National Park Service Paleontological Research. Technical Report NPS/NRPO/NRTR-95/16.

Wang, Y., T. E. Cerling, and B. J. MacFadden. 1994. Fossil horses and carbon isotopes: new evidence for Cenozoic dietary, habitat, and ecosystem changes in North America. Palaeogeography, Palaeoclimatology, Palaeoecology, 107:269-279.