Arctocyonid

Arctocyonidae (from Greek ''arktos'' and ''kyôn'', "bear/dog-like") is an extinct, possibly polyphyletic family of placental mammals which lived from the late Cretaceous to the early Eocene. They were initially regarded as creodonts, though have since been reassigned to an order of their own, the Arctocyonia. Some have suggested that arctocyonids are ancestral to modern-day artiodactyls, or that they form a sister group. However, more recent phylogenetic analyses suggest that arctocyonids may represent an artificial grouping of extinct ungulates, or that they might be an assemblage of unrelated placentals related to pangolins, pantodonts, and periptychids.

Members of Arctocyonidae are characterised by long skulls, with large sagittal crests and very large canines. In the case of ''Arctocyon'' proper, the lower canines especially were large enough to require a diastema on the upper jaw to accommodate them. Arctocyonids varied considerably in size and morphology. Smaller genera, like '' Chriacus'', were about the size of a coati, while larger ones, such as ''Arctocyon'', weighed up to and stood at the shoulder. Many arctocyonids have climbing adaptations, suggesting that they were either descended from arboreal taxa, or were arboreal themselves. The North American ''Anacodon'' was more robust than other genera, and had adaptations for burrowing as well as climbing. Most genera appear to have been omnivorous, though ''Anacodon'' showed signs of an increase in herbivory.

Taxonomy

The family Arctocyonidae was named by Christoph Gottfried Andreas Giebel in 1855, as a carnivoran subfamily that included ''Arctocyon'', the amphicyonid ''Amphicyon'', and the ursid ''Agriotherium''. It was elevated to family level by Scottish zoologist Andrew Murray. At some point thereafter, arctocyonids became a family within Creodonta. In 1975, Malcolm McKenna erected a new order, Arctocyonia, to accommodate them, placing them within Ungulata. Since then, Arctocyonidae has largely been treated as a family of its own, though how that family is divided has been another matter. William Diller Matthew, in 1937, divided it into four subfamilies (Arctocyoninae, Chriacinae, Oxyclaeninae, and Triisodontinae); the latter is now regarded as a family of its own. That same year, George Gaylord Simpson suggested that arctocyonids could instead be divided into Arctocyoninae, Oxyclaeninae, and Triisodontinae. In 1978, Leigh Van Valen erected a new subfamily, Loxolophinae.

The relationship between arctocyonids and other clades has long been uncertain. Since becoming the sole representatives of their own order, they have been suggested to be either ancestral to artiodactyls or close to the clade's stem. The family's monophyly has also been called into question. In 2012, a phylogenetic analysis of '' Prolatidens waudrae'', a traditional arctocyonid, recovered it as a more basal ungulate; ''Arctocyon'', '' Landenodon'' and '' Thryptacodon'' were recovered as part of a clade sister to triisodonts and mesonychids; and the remainder of tested arctocyonids formed a polytomy basal to that clade and ''Diacodexis''. In 2015, Peter E. Kondrashov and Spencer G. Lucas recovered the family as an artificial assemblage of basal ungulates. That same year, a larger analysis by Thomas J. D. Halliday, Paul Upchurch and Anjali Goswami recovered arctocyonids as several entirely unrelated placental lineages. By their unconstrained strict consensus tree, ''Arctocyon'' is related to ''Periptychus'' and '' Protolitopterna'', Loxolophinae is related to pantolestids and pangolins. By their constrained strict consensus tree, ''Arctocyon'' and ''Loxolophus'' form a clade related to pantodonts and periptychids, and the rest of Arctocyonidae is recovered close to pangolins.

Description

Arctocyonidae, if monophyletic, was a morphologically disparate lineage. Some genera, such as ''Chriacus'', were fairly small, and bore adaptations for an arboreal or scansorial lifestyle. Others, like ''Anacodon'', were very large and robust, having adaptations for both arboreal and fossorial lifestyles. ''Arctocyon'', particularly ''A. mumak'', appears to have been the most terrestrial, though likely descended from arboreal ancestors. Most arctocyonids are fragmentary, making it difficult to determine body size. ''Chriacus'' likely weighed , and was slightly larger than a modern coati. ''Arctocyon primaevus'' may have weighed up to , and had an estimated shoulder height of . Remains of ''Mentoclaenodon'' suggest a very large body size, though exactly how large it grew to is unclear, and it may have been outsized by ''Arctocyon mumak''.

Skulls and dentition

The skulls of arctocyonids were fairly long, with a small braincase and very large sagittal and occipital crests, combined traits of herbivorous and carnivorous mammals. The zygomatic arch of ''Arctocyon'' specifically was very large, with a posterior angle anterior to the mandibular fossa, almost forming a right angle. In most genera, the incisors were small and unspecialised, though others, like '' Thryptacodon'', had modified them into grooming teeth convergent with those of lemurs. Though the canines likely had a role in feeding in certain genera, in others, like ''Anacodon'', they did not. In the case of ''Arctocyon'', the lower canines were longest, to the extent that they were accommodated by a gap (or diastema) between the upper canines and premolars. The cheek teeth were tricuspid (three-cusped) and were often bunodont. The premolars were simple, if fairly sharp, while the molars were blunt and resembled those of bears. Overall, the dental morphology of arctocyonids suggests that they were omnivorous to varying degrees.

Postcranial elements

In ''Arctocyon'', the mammillary processes of the dorsal vertebrae were robust (suggesting powerful musculature), and the caudal vertebrae appear to have been tightly interlocked, suggesting that the tail was fairly rigid. In ''Chriacus'', however, there are signs that the tail may have been prehensile. Arctocyonid limbs were fairly typical in length, with stout ulnar and fibular shafts. In ''Anacodon'', particularly, the limbs were very robust. In ''Chriacus'', the ankle joints were flexible and allowed the hind feet to rotate, enabling them to climb downward. ''Anacodon'''s ankles had very little lateral movement. Five digits were present on all limbs. The phalanges were narrow and long, though the innermost and outermost digits were slightly reduced. In the case of ''Chriacus'', the innermost digit (the hallux) was divergent. On all digits, the unguals were laterally compressed and quite clawlike. In some genera, like ''Anacodon'', they may have been used for digging.

Biology

Diet and feeding

Arctocyonid cheek teeth were bunodont, and the carniassials seen in other predatory mammal clades were essentially absent. Overall, the dental morphology of arctocyonids suggests that they were omnivorous, to varying degrees. ''Anacodon'' may have been among the least carnivorous, having flat, crenulated cheek teeth while ''Arctocyon corrugatus'' was among the most carnivorous.

See also

* Mesonychidae

References

{{Taxonbar, from=Q134853
Arctocyonia
Condylarths
Paleocene mammals
Eocene mammals
Late Cretaceous first appearances
Eocene extinctions
Prehistoric mammal families