Oreodonts: Extinct large burrowing mammals of the Oligocene

By: Dr. Kent A. Sundell, Casper College, Wyoming


Oreodonts Merycoidodontoidea are the most commonly collected fossil from the White River Formation of central North America. They locally comprise 25% of the fauna. They have previously been thought of as sheep-like ruminants, migrating in large herds across the early developing grasslands of North America, presumably protected by sheer numbers and musk ox-like defensive techniques from sabertooth cats, Hyaenadons, and large dog predators. Recent extensive collections of multiple (more than one associated individual) oreodont specimens (34 multiples) indicate they were burrowing denning mammals using their distinctive large canines for defensive purposes, fending off predators at the mouth of their burrows. Of eight observed specimens with unborn fetuses, three had 4 fetuses, two had 3+ fetuses, two had 2 fetuses, and one had 1 fetus. At least six specimens were composed of subadults (siblings?) with 3-5 individuals in calcareous enriched nodules (dens) that varied from 24"x18"x6" up to 60"x24"x10". Three of the subadult specimens are fully articulated (not scavenged). Approximately two thirds of all multiple specimens are disarticulated (predated) with mostly skulls, feet and miscellaneous broken skeletal elements in den-size (24"x36"x 8") nodules. The occurrence of these multiple oreodont specimens in close proximity, both predated and not predated is the primary physical proof of the burrowing and denning nature of oreodonts. Multiple fetuses (2-4 commonly), digging feet with four claws, large heads relative to bodies, small stocky bodies, large genetic diversity, and even the diagnostically large canines are additional supportive evidence of the oreodonts burrowing lifestyle. The White River Formation was composed primarily of soft yet firm silty and ash-rich mudstones a perfect digging and denning medium. In summary, oreodonts dug many burrows and dens into the soft firmly consolidated sediments, they raised small litters of offspring , occupied the dens most of the year, and defended themselves from predators by backing into the den and exhibiting the characteristic large canines. When carnivores were successful in penetrating the dens they commonly feasted on the juveniles and subadults, leaving the leftover skulls, feet, and broken bones on the floor of the den, which later collapsed and formed the core of the nodules that we find the specimens in today. Oreodont dens can be distinguished from carnivore dens because they contain only oreodont material (carnivore dens have fragments of many (5+) species) and commonly of similar size subadults and juveniles (siblings). Oreodonts successfully developed and occupied an ecological niche most similar to the modern warthog that probably provided habitat and shelter for many other animals during the middle Tertiary of North America.


Oreodonts Merycoidodontoidea (Leidy, 1852) are the most commonly collected vertebrate fossil from the White River Formation of central North America (Figures 1-12). They locally comprise more than 25% of the fauna (K. Sundell, this volume). They have previously been thought of as sheep-like ruminants, migrating in large herds across the early developing grasslands of North America, presumably protected by sheer numbers and musk ox-like defensive techniques from sabertooth cats, Hyaenadons, and large dog predators. However given their general physiology (body form) it would appear difficult for these squatty, big headed, and short legged animals to out run or defend themselves on open ground from the predators of the day, that were larger, more agile, faster, and hungry. Oreodonts should have been easy prey to Hoplophoneus, Hyaenadon, and Daphoenus like modern domestic sheep to coyotes and wolves. So how did they become the most common herbivorous mammal and dominate that position for more than 15 million years during the late Eocene, Oligocene, and Miocene in North America?.


Based on twenty three years of personal field collecting in the White River Formation in Wyoming and discovering hundreds of oreodont skeletons, I suggest that the generally accepted palaeoecology (lifestyle) of oreodonts is wrong. They have generally been viewed as living in large herds (O'Harra, 1920) that roamed the plains of South Dakota, Nebraska, Wyoming, and surrounding states during the Oligocene and Miocene. I believe oreodonts were relatively large burrowing and denning mammals that spent most of their nonconsuming time (when not eating or seeking mates) underground, especially females and young. This would make their lifestyles more similar to modern rodents (like prairie dogs) than to large ungulates (sheep or horses). The results of this change are profound, given that oreodonts may comprise more than 25 % of all vertebrate fossils found in the Oligocene age "oreodon beds". I must emphasize that oreodonts are all extinct and no ancestors live today so the following comparisons are of lifestyle and physiology with modern animals only. Oreodonts and their palaeoecology were unique.


The following list of observations that support the hypothesis can be divided into three main categories, including anatomy of the oreodont, close association of multiple skulls and skeletons, and paleoecology inferences.

Anatomy - The following anatomical features are primarily suggestive that oreodonts oreodonts could and did use their feet for digging and their bodies were reasonably built for living in and defending a burrow entrance.

Canine teeth - One of the most diagnostic and visually impressive features of an oreodont skull is the pressence of large canines on the upper (Figure 2) and a corresponding large premolar on the lower jaw (Romer, 1966) that gives a viscous carnivorous frontal appearance to the oreodont skull. Whereas the rear dention shows broad flattened shearing teeth typical of a ruminating herbivore. These large frontal teeth would be a good defensive weapon against most predators of the day if attacked from the front only. Although somewhat reduced in size in females and later forms (Leptauchenia) the oreodonts as a group generally retained the large canines throughout their evolutionary history. The slightly reduced size of canines in females may suggest that males were needed to help defend the den and family, possibly making them even more socially ordered, improving the survivability of the young.

Skull size and body proportion - The skull of an oreodont is proportionately larger, bulkier, and broader than other similar size ungulates (i.e. horse) compared to their skeletons. Their legs are short and stout with an overall pig-like form. The larger skull is needed for defensive display and combined with short legs to minimize the height of the burrow needed (Figure 1). Evolutionarily, they kept the short stocky shape and even evolved stockier and more grotesque forms (Promerycochoerus), unlike typical running ungulates that evolved longer legs to run faster. Obviously, oreodonts were not primarily running from their predators.

Feet - The earliest oreodonts (agriochoerids) had four toes, a small opposable thumb, and re-evolved definite claws. The claws may have been for digging or tree climbing. The feet of most common oreodonts during the Oligocene and Miocene have four toes, each with a wider tipped claw (hoof?). Evolutionarily, they did not reduce the number of toes like horses in order to run faster. Oreodonts needed the toes and claws to dig with and more toes and wider claws allow faster digging in soft sediments. The shape and position of the foot was very cup-like (Figure 3) with the two central metatarsals being longer, stronger, and dorsally raised slightly above the lateral metatarsals. This shape could also improve digging capabilities.

Sesamoid bones - Oreodont feet have many sesamoid bones to support joints between the phalanges (Figure 4). The middle toes have three between the proximal and medial phalanges and at least one between the medial and distal phalanges. The smaller lateral toes have smaller and fewer sesamoid bones. Generally more sesamoid bones indicate hard usage of the paws such as digging. Oreodonts have more sesamoid bones than most modern den digging carnivores (dogs) today.

Associated skeletons - The following comments are based on thirty four multiple (more than one individual together) oreodont specimens that I have collected.

Fetuses - Of eight observed female specimens with unborn fetuses, three had 4 fetuses, two had 3+ fetuses, two had 2 fetuses, and one had 1 fetus (Figures 5-9). Most recognized fetuses had teeth developing. One specimen (Figure 7) has two fetuses, but is at a much earlier stage of fetal development with no teeth visible and skulls approximately three quarters of an inch long.
Multiple skeletons
Small juveniles - At least six specimens are composed of subadults (siblings?) with 3-5 individuals in calcareous enriched nodules (dens) that varied from 24"x18"x6" up to 50"x24"x8". Three of the subadult specimens are fully articulated (not scavenged).
Families in burrows - One specimen has seven+ individual oreodonts in a single nodule (den) measuring 60" x 24" x 10". This specimen has two adults (6" skulls), three subadults (4" skulls), and 2+ juveniles (3" skulls). A second family specimen has two adults and at least two small juvenile skeletons in a 50" x18" x 6" nodule. A third family specimen (Figure 8) shows 1 adult, 1 juvenile, and 4 fetuses.
Predation in burrows. Approximately two thirds of all multiple specimens are disarticulated (predated) with mostly skulls, feet and miscellaneous broken skeletal elements in den-size (24"x36"x 8") nodules. Figure 10 is a good example of one such predated burrow filling. The nodule (den) measures 25" x 14" x 8" before preparation. Upon preparation the specimen includes three subadult oreodont upper and lower skulls (3.5" long) with erupting teeth (Figures 11 &12) all approximately the same size and maturity, presumably siblings. Three articulated legs and feet, three strands of vertebrae, a pelvis, clumps of ribs, miscellaneous scattered limb bones. They have obviously been predated with most of the skeletons now gone, leaving mainly skulls and feet (least edible parts). If these subadults were above ground when killed: 1) How did the predator (Hyaenadon?) kill all three? 2) Eat most of all three? 3) and then stack the skulls back together in a pile? The best explanation is that a predator entered the burrow, which was not defended by the parents at the time. Predator was able to kill all three, because they couldn't get away except past the predator. The predator(s) ate their fill of most meaty parts and left the remains in the burrow, where they couldn't be easily scattered as they would have been above ground. It is not a predator's den, because it contains only remnants of equal size subadult oreodonts. The burrows of predator dens have many sizes of bones from different individuals and commonly contain several different species (5+). This form of predated sibling multiples is not unique and more than half of the multiple specimens are of this type. The occurrence of these multiple oreodont specimens in close proximity, both predated and not predated is the primary physical proof of the burrowing and denning nature of oreodonts.


Sediments of the White River Formation
The most fossiliferous portions of the White River Formation are composed primarily of silty and ash-rich mudstones (Evanoff, 1993), a perfect digging and denning medium. The high clay content of these sediments would have made the recently deposited sediments firm for support of burrows, yet soft and easy to dig. The mudstones were deposited as periodic overbank deposits during periods of high stream flow and flooding of the adjacent floodplains. The associated coarse grained conglomerate and sandstone stream channels only contain rare fossils, usually large herbivores (titanotheres and rhinos) or rounded and transported reworked bone. The proliferation, success, and evolution of oreodonts as a group may be directly related to the widespread deposition of near perfect digging and denning medium, the White River Formation and latter somewhat sandier, but still ash and clay-rich Miocene formations throughout the northwestern great plains of North America.

Articulated skeletons in burrows (nodules) Most (95%) of the articulated oreodont skeletons in my collection area (Wyoming) are found in calcite enriched erosionally resistant nodules within the mudstones of the White River Formation. The calcite enrichment causes permineralization (preservation) of the bone and forms a harder concretionary form around the skeletons. The skeletons are generally fully encased in the nodules. The nodules are surrounded by softer (easy digging) mudstones with little observed textural differences between the nodules and surrounding matrix. Sometimes minor portions of the skeletons (toes, rib tips, sagittal crest on skull) will extend 1 inch outside of the nodule. I believe the nodules were formed by post depositional groundwater bringing calcium in solution and depositing it in and near more permeable and porous zones of the White River Formation. Vertical clastic dikes enriched in calcite are common and were a conduit of calcium rich waters that permeated laterally throughout the White River Formation. Sometimes these calcite enriched zones are laterally extensive (possibly soil horizons) and form nodular ledges. Often they appear as layers of isolated nodules. Unfortunately not all nodules have fossils and probably less than 0.1 % do have fossils as a nucleus for the nodule. However, certain intervals contain nodules in classic burrow shapes and forms (C. Sundell, this volume) and appear to be nearly replacing the actual burrow of rodents and smaller animals. Unfortunately, polished cross sections and detailed thin sections across the nodule boundaries have not been done to verify if the nodules are exact calcite filling of the disrupted burrow fillings. The average size and shape of these nodules that contain oreodont skeletons 3'length x 2'width x8"height is suggestive of a flattened, collapsed burrow. Some do appear to have inclined entrances (smaller nodules leading to den area), but are rarely exposed in the correct present day erosional perspective to see the entrance.

Genetic diversity
This burrowing hypothesis does help explain the tremendous genetic diversity of the oreodonts relative to other similar size and distribution ungulates during the middle Tertiary in North America. The oldest oreodonts (agriochoerids) are found in the middle to late Eocene (Diplobunops protoreodon) and the earliest forms re-evolved claws from a hoofed ancestor to be used for climbing and/or digging. At least 17 genera of oreodonts are found in the Oligocene. This compares with 3 genera of horses and 5 genera of camels over the same time and area. Even within a species (Merycoidodon culbertsoni) tremendous genetic size and shape variation occurs. The size and shape variation commonly resembles that of modern dogs. I believe this high variation and great genetic diversity is due to the burrowing nature of oreodonts. They were not migrating in large herds across the plains, keeping a good mixed gene pool. Instead they were tied closely to the dens, which restricted their movements and allowed for greater genetic diversity over the same geographic areas as contemporaneous horses and camels. Local variation may have been controlled by lithification and changes in the surficial sediments, requiring more strength to dig dens in one area verse another.

A mega-burrowing community
Given that oreodonts were digging dens and burrows large enough to shelter them and their families and that oreodonts are the most common vertebrate fossil from the Oligocene of North America the palaeoecology was very unique and dramatically different from anything today. It was truly a burrowing community. More than 95% of the vertebrate fauna during the Oligocene could have fit inside an oreodont's old abandoned burrow, all except titanotheres, rhinoceros', entelodonts, and large tortoises. So even if other animals didn't dig burrows they could have utilized the oreodonts homes for shelter and cover from elements and predators.
I envision a partially forested (near streams) generally brushy environment in Wyoming during the Oligocene as the regional forests of the Eocene disappeared and before the vast savannah grasslands of the Miocene took over. This would be particularly well suited for oreodonts that were peccory to boar size and shape, feeding on the shrubs and browse of the day and frantically scurrying back to the den at the first sight, smell, or sound of a predator. Possibly social communities similar to prairie dogs existed that would alert nearby neighbors of approaching predators with whistles or grunts. Females and children entered first and the male would be the last guard slowly backing into the den as the predator approaches. Upon reaching the mouth of the burrow the predator is greeted by a large head with 4 long sharp teeth. Have you ever tried to pull a badger out of his den? This scenario explains why predators could not control oreodont populations, which were probably controlled by food source.


I believe the strongest hard evidence (proof) herein presented for the burrowing nature of oreodonts are the occurrence of multiple juvenile predated specimens in single nodules (burrows) and the pressence of multiple fetuses (3-4) in female oreodonts. Any individual characteristic as I have described might be argued away, but taken in sum I believe they provide a plausible explanation as to why there were so many oreodonts only in North America during the middle Tertiary. It was truly a unique environment and the paleoecology of many of the smaller animals may have to be reexamined in light of the tremendous impact oreodonts may have had on the availability of subterranean shelter. Their extinction may lie in the sediments that made them great. As the supply of clay rich ash diminished in late Miocene and Pliocene the climate became progressively cooler, grasses replaced shrubs, and the association of the right feed and soft digging stratum probably became more scarce.


Evanoff, E., 1993, Stratigraphy of the White River Formation, Converse County, Wyoming: [Ph.D. Thesis] University of Colorado, Boulder, 298 p.

Leidy, J., 1852, Extinct Mammalia and Chelonia from Nebraska Territory; in Owen, D. D., Report of a geological survey of Wisconsin, Iowa, and Minnesota and incidentally of a portion of Nebraska Territory: Lippincott, Grambo & Co., Philadelphia, p. 539- 572.

O'Harra, C.C., 1920, The White River Badlands: South Dakota School of Mines, Bulletin 13, 181 p.

Romer, A. S., 1966, Vertebrate Paleontology: University of Chicago Press, 468 p.