Patriofelis ("father of cats") is an extinct genus of carnivorous placental mammals from the extinct subfamily Oxyaeninae within the extinct family Oxyaenidae, which lived from the early to middle Eocene, during the Bridgerian stage of NALMA. The first remains were discovered in 1869 by geologist Ferdinand Vandeveer Hayden, and the genus was named the following year by Joseph Leidy. Over the next few decades, additional specimens would be named from the same locality (or similar ones), many assigned to genera of their own or to separate species within Patriofelis. Now, only two are recognised: Patriofelis ulta, the type species, from the lower Bridger Formation and the Huerfano Formation of Colorado, and P. ferox, from the lower Bridger and lower Washakie Formations of Wyoming and Colorado, and the Clarno Formations at the John Day Fossil Beds National Monument, Oregon.
Patriofelis is the largest oxyaenid to preserve postcranial fossils (fossils which come from behind the skull). P. ulta, the smaller species, has been estimated at around , while the larger P. ferox has been estimated at . Patriofelis had a skull around the same size as of that of a lion, though shorter and broader in a way often likened to otters. A semi-aquatic lifestyle has been suggested for Patriofelis, though there are aspects of its anatomy which make this very unlikely. Its backbone was stiffened by bony structures which projected from its vertebrae, inhibiting the usual mammalian swimming method, in which the spine undulates up and down. Originally it was suggested to be a turtle specialist, though its teeth were not adapted for crushing, but rather for slicing. While some older sources reported that Patriofelis had limited arm flexibility, recent studies have shown that its forearms were capable of a great degree of pronation and supination. Though there are various hypotheses for how Patriofelis lived, the most likely is that it was a terrestrial ambush predator which hunted by grappling its prey.
Patriofelis was among the top predators of at least some localities, competing with certain crocodyliformsns and with mesonychids. During the early to middle Eocene, much of North America was covered in wet subtropical forest and marshland, and Patriofelis was present in geologic formations corresponding to both of these environments. It has been suggested that the extinction of the genus, along with other oxyaenids, corresponded with a shift in available habitat, as the end of the Eocene was marked by a period of cooling and drying which led to the replacement of dense, swampy, forests to more temperate open forests.
History
Early history Some time in 1869, geologist Ferdinand Vandeveer Hayden recovered a pair of partial mandibular rami (the structures which comprise each side of the lower jaw) from strata belonging to the Bridger Formation. These bones were later catalogued as USNM V105. In a March 1870 publication in the Proceedings of the Philadelphia Academy, palaeontologist Joseph Leidy described the jaw scientifically, making it the holotype of a new genus and species, Patriofelis ulta. He believed that P. ulta was an early member of Felidae, the cat family, hence its name.J. Leidy, (1870.) Untitled [Patriofelis ulta proposed during Proceedings of the March 8 meeting of the Academy of Natural Sciences], in Proceedings Proceedings of the Academy of Natural Sciences of Philadelphia, Volume 22, p. 9-11. Two years later, in August 1872, Othniel Charles Marsh described the remains of a "gigantic Carnivore" from the same locality, to which he assigned the name Limnofelis ferox. Marsh's specimen consisted of skull material (including a lower jaw), vertebrae,O. C. Marsh. (1872.) "Preliminary description of new Tertiary mammals. Part II." American Journal of Science 4(21):202-224 and "other less important parts of the skeleton". Marsh also described a second species, L. latidens, from an upper premolar from the same locality.
In the summer of 1879, Jacob Lawson Wortman discovered teeth and limb bones from another specimen. These remains came to the attention of Edward Drinker Cope, who, in 1880, gave them the binomial name of Protopsalis tigrinus. In 1892, Wortman suggested a second Protopsalis, P. leidyanus, based on remains in the Princeton Collection. Two years later, Wortman published a paper discussing the osteology of Patriofelis ulta. He determined that the other taxa could be rendered junior synonyms, as there was no convincing evidence in the form of diagnostic traits that they were different taxa. Several other species have been described which have been consistently synonymised, leaving only P. ferox and P. ulta as valid species.J. Alroy. (2002.) "Synonymies and reidentifications of North American fossil mammals."
Two relatively complete specimens of Patriofelis (UMNH VP 550 and CM 87673) have been discovered in the interim, both in 1953. The first of these served as the foundation of a 1957 paper on P. ulta's skull anatomy by Charles Lewis Gazin. Both were redescribed by Anne E. Kort and colleagues in 2022, and served as the basis of a description of Patriofelis' postcranial anatomy.
Classification Patriofelis is a member of the family Oxyaenidae, which is in turn part of the order Oxyaenodonta. Oxyaenids were at one point considered members of a larger order, Creodonta, alongside hyaenodonts. More recent papers have not supported the monophyly of creodonts, and the link between the two is predicated largely on the presence of a blade on the metastyle of the first molar, a trait present in many eutherians.
Description
Size
Patriofelis is the largest oxyaenid from which postcranial remains are known. Wortman suggested that its overall body size was roughly equal to that of a black bear, though was working under the assumption of felid affinities. The type species, of Patriofelis, P. ulta, was the smaller species, weighing at least . Sorkin, in his 2008 paper, suggested P. ferox was comparable to Thylacosmilus in size, which weighed within the paper. In a 2010 publication, Michael Morlo, Gregg F. Gunnell, and Doris Nagel provided general body mass estimates for both Patriofelis species of . In a 2024 paper, Mark S. Juhn and colleagues used a scaling method proposed by Van Valkenburgh in 1990, wherein the size of the largest lower molar was used in a regression analysis, to calculate the body sizes of various extinct mammalian predators, listing their mass estimates in their supplemental materials. Using this method they obtained a body mass of for P. ulta, and a larger body mass of for the larger P. ferox.
Skull
The skull of Patriofelis ulta described by Wortman was roughly the same size as that of a lion. He noted that it was fairly short, broad and low, and he likened it to the skulls of seals, though his initial restoration was inaccurate in that it depicted the skull as higher than it actually would have been. Charles Lewis Gazin noted that not only was the modification of its skull more extreme than how Wortman had depicted it, but it was superficially akin to that of an otter. The premaxilla in P. ulta was fairly reduced, forming a narrow rim of bone along the anterior (front) margin of the maxilla. The maxillae were fairly short when viewed from the side, although their nasal processes were very long, extending posteriorly (rearward) over the lacrimal bones. The nasals of P. ulta were broad anteriorly, tapering posteriorly and widening yet again around the sutures between the maxillae and frontals, after which they once again tapered. The frontals were broad anteriorly and tapered posteriorly. The parietals were long and slender, supporting a fairly well-developed sagittal crest. P. ulta's orbits (eye sockets) were not particularly large, though were positioned dorsally (high up) on the skull. The zygomatic arches, formed from the jugals and squamosals, were broad and massively built. The mastoid regions of the temporal bones were very large and prominent, and the paroccipitals were conspicuous. P. ulta's braincase was overall very small, with very thick walls. As with the rest of the skull, P. ulta's mandible (lower jaw) was heavy and robust, with short mandibular rami and mandibular symphyses. It is fairly rounded anteriorly. The condyles of the masseteric fossa were fairly low, extending quite far transversely (across) and being structured similarly to those of cats. The fossa was fairly rounded in angle, similar to that of Oxyaena.
Brain endocast The brain anatomy of P. ulta is known in part from an endocast. Its cerebellum was fairly large, with a thickly walled cavity, and was not at all covered by the cerebral lobes. As in many Palaeogene mammalian carnivores, the cerebral cavity had at least two longitudinal convolutions. It is not certain if a crucial sulcus was present.
Dentition
The dental formula of Patriofelis is uncertain, though was rendered by Gazin as . Its teeth were very robust in comparison to other oxyaenid teeth, and have been variously compared to those of felids, or to those of hyaenids. The upper incisors were fairly simple, with the medial (inner) pair being smaller than the lateral (outer) pair. The lateral pair were far larger and were somewhat caniniform. The true upper canines were very robust, with deep roots. They were oval-shaped in cross section, as opposed to being sabrelike or laterally compressed, and in the specimen studied by Gazin they had been blunted by wear. In 1900, Henry Fairfield Osborn noted that oxyaenid cheek teeth bore adaptations for slicing and for cutting into bone, such as the loss of the talonid, as in felids. True bone-crushing is unlikely, at least in Patriofelis, as P. ulta's dentition overall lacks adaptations for durophagy. P. ulta lacked a first upper premolar. The second was simple, with two roots and a broad posterior region. The third had three roots with a small anterior cuspule. The fourth had a higher anterior cusp, and a projection (also found in the preceding teeth, albeit smaller) which was equal in size to the main cusp but more bladelike in morphology. The first upper molar formed a carnassial, a specialised shearing blade, while the second, if present, was small and peglike, and difficult to make out due to its angle. The number of lower incisors apparently varied between specimens, though the one studied by Gazin had two. As with the upper jaw, the lateral lower incisor was the larger one, though was less procumbent. Both sets were somewhat blunted due to wear. The canines were robust, massively rooted, and close together, and were similarly worn. The second lower premolar of P. ulta was separated from the canine by a short diastema (gap), whereas that of P. ferox was oblique to it. It was double-rooted and broad. The third was similar but far larger and more robust. The fourth was larger still, with higher cusps and a more pronounced crest on its lingual (inner) margin. It was tilted backwards somewhat. The first lower molar was small, with a metaconid which extended upward and posterolingually (inward and rearward) from the protoconid. Similar to the first upper molar, the second lower molar formed a shearing surface and lacked a talonid.
Postcranial skeleton