Lycosuchidae
Lycosuchidae is a family of therocephalians known from fossils from what is now the Beaufort Group of South Africa and that lived during the Middle to Late Permian between roughly 265 to 259.2 million years ago. It currently includes only two genera each with a single species, Lycosuchus, represented by L. vanderrieti, and Simorhinella, represented by S. baini, both named by paleontologist Robert Broom in 1903 and 1915, respectively. Both species are large predators characterised by their size, reduced tooth counts with large, almost "sabre toothed" canine teeth, and relatively short, broad and low snouts.
Lycosuchids were once thought to be defined by having two simultaneously functional pairs of canines, so-called "double canines", instead of a single pair of like in all other predatory therapsids. However, it has since been recognised that these actually represent overlap between an older pair and their alternated replacements in separate tooth sockets, and that fossils of lycosuchids with "double canines" in fact preserve teeth at different stages of growth and replacement. This is the same method of canine replacement used by other predatory therapsids, though the pattern appears to be unusual in lycosuchids as the alternating canines occur together more often compared to other predatory therapsids like other therocephalians and gorgonopsians.
Lycosuchids are among the earliest known therocephalians and are also thought to be the most basal. The Russian genus Gorynychus, containing two species, may also belong to the family, although this result is not typically recovered. Lycosuchids are only known from the upper Tapinocephalus and lower Endothiodon Assemblage Zones of the Karoo Basin, surviving the Capitanian mass extinction event between them that wiped out many other therapsid groups and ended the Middle Permian, but apparently only persisting as a "dead clade walking" that went extinct as ecosystems fully recovered, potentially competing with recently evolved large gorgonopsians.
Morphology and biology
Lycosuchids are characterised by several physical features of their skulls, including a low and wide snout that is proportionately short, typically half or less of the skull's whole length. The number of teeth are also distinctive, as lycosuchids only ever possess five or less incisors on each side of the upper jaw, a pair of canines, and very few postcanines behind them, typically only two or three behind each. In the lower jaws, there are only three incisors, a canine, and around five postcanines in each mandible.While lycosuchids have shorter, wider and more robustly built snouts than early therocephalians in the family Scylacosauridae, they are nonetheless broadly similar to them, and in some ways also superficially resemble gorgonopsians. The large, almost "sabre-tooth" like canines are especially striking similarities. This anatomy suggests lycosuchids were capable of relatively strong bites with their large canines and incisors, and were resistant to the twisting of prey in their powerful jaws. All lycosuchids, including dubious genera and indeterminate specimens, are very similar superficially and share all of these features with little variation. The only accepted valid genera, Lycosuchus and Simorhinella, are mostly told apart by details of palate and the jaw joint.
Lycosuchids are large therocephalians, with the largest known Lycosuchus skull measuring almost long and the even larger skull of Simorhinella measuring long. The largest lycosuchid is an indeterminate specimen previously named as the holotype of "Scymnosaurus major" labelled SAM-PK-9005. It is a poorly preserved and incomplete snout measuring long, but proportionately it exceeds the snout dimensions of Simorhinella. SAM-PK-9005 is both the largest known lycosuchid and also one of the largest therocephalians known altogether.
Little is known about the postcranial skeleton of lycosuchids, and there is little overlap in the available material to make broad statements about their anatomy. Nonetheless, two scapulae from different specimens both have a distinctive bony protuberance above the glenoid, thought to be an enlarged attachment for the part of the triceps muscle. The ulna, known in both Simorhinella and SAM-PK-9005, is robust with a short olecranon process at the elbow, a feature probably related to their large size. A complete skeleton of a large therocephalian that was previously described as a specimen of "Cynariognathus platyrhinus" in 1967 was alluded by palaeontologist Christian Kammerer to be a lycosuchid in a 2023 paper on scylacosaurid therocephalians, though it has yet to be compared in detail with other lycosuchid remains.
"Double Canines"
Historically, "double canines" in the upper jaw were regarded as the most defining trait of lycosuchids, and sometimes even the sole trait to distinguish them. Unlike other predatory therapsids, which only ever have one functioning pair of upper canines at a time, lycosuchids were thought to have two distinct pairs, each with its own tooth socket one immediately behind the other, that were both simultaneously functional and independently replaced. Notably, these other therapsids also have two positions for each canine and the active tooth alternated between the alveoli every time it was replaced. The distinction made for lycosuchids was that two distinct sets teeth supposedly occupied both positions at the same time, rather than a single pair that alternated between them. This was thought to be a primitive characteristic retained by lycosuchids from earlier sphenacodont ancestors such as Dimetrodon, which were also thought to have two distinct sets of canines.It has since been determined that the "double canines" of lycosuchids do in fact represent this same alternating pattern of replacement, as the ages of the canines are always staggered and specimens with only a single erupted pair exist. Nonetheless, the pattern of canine replacement in lycosuchids appears to be unique compared to other predatory therapsids. Although lycosuchids are relatively rare as fossils, over half of the known specimens preserve "double canines". This is much more frequent than other, more common groups of therapsid, and suggests that there was more overlap between the alternating pairs during the lifetime of lycosuchids than in other therapsids.
In 2014 Fernando Abdala and colleagues proposed this could be explained by a more rapid rate of replacement, such that they overlapped more often and that old teeth did not shed until after its replacement had already grown to a similar size. A team led by Luisa Pusch in 2020 proposed an alternate hypothesis based on CT-scans of the jaws of Lycosuchus. They found that in addition to the old canine remaining in place as its replacement developed, its own direct replacement in the same socket was already developing inside the jaw bone. They proposed instead that complete tooth replacement in lycosuchids was protracted, with the older canines remaining set in the jaws long after the alternate pair had erupted and grew to a similar size to the point that their own direct replacements had started to develop before finally being shed.
Though not truly representing two functional pairs of canines as originally thought, the frequency of overlap between the alternating old and replacement canines suggests they were both still functional together, at least to some degree, during the animals' lifetimes. This is unlike other predatory therapsids, where the old and worn canine is shed to be functionally replaced by its alternate. How two concurrent large canines in each upper jaw functioned is not clear, and it has been argued that such "double canines" acting as a single unit would be less efficient at both puncturing and tearing flesh due to their bulk and by obscuring the serrations of the other tooth.
Taxonomy and classification
The concept for a family of "double canined" early therocephalians was first put forward by Broom in 1908, when he proposed that the "double canined" early therocephalians recognised at the time formed a separate evolutionary unit from other early therocephalians, then recognised as the "Pristerognathidae". Broom would continue to be a proponent for this division, but for many decades he did not attribute a name to such a group, even after Lycosuchidae was already made available. The family Lycosuchidae was first established by Baron Franz Nopcsa in 1923, although the name was often misattributed to other authors by later researchers until the end of the 20th century when his precedent was recognised. Other early uses of the name Lycosuchidae were by Samuel W. Williston in his 1925 publication The Osteology of the Reptiles, and a similar concept was used by Sidney H. Haughton and A. S. Brink in 1954 catalogue of fossil "reptiles" from the Karoo, though neither of them attributed any authorship to Lycosuchidae.The validity of Lycosuchidae was not always supported by researchers, particularly when the "double canines" were the only proposed criterion. This scepticism culminated in a paper by Juri van den Heever in 1980, where he argued that lycosuchids were an unnatural, artificial collection of "pristerognathids" that had died during the process of alternating canine replacement typical of predatory therapsids, and argued it should be invalidated. When not recognised as forming their own family, lycosuchid genera were typically subsumed under the family Scylacosauridae, which was more commonly identified as Pristerognathidae by most researchers during the 20th century. Van den Heever would later reconsider his view on lycosuchids and recognise them as a distinct lineage after all, establishing the more thorough anatomical criteria for which the family is diagnosed today in his PhD thesis and a subsequent publication. Therein, he also established the taxonomically higher group Lycosuchia containing the family, intended to be the equivalent sister group to Scylacosauria, though this higher taxon has not seen use since then.
Two genera are considered valid today, Lycosuchus and Simorhinella, each with a single species. Both genera were named by Robert Broom, a palaeontologist who worked on and named many Karoo fossils, in the early 20th century, but Simorhinella was not actually identified as a lycosuchid until almost a century after its discovery. Lycosuchus is known from five skulls and mandibles of varying completeness, including the nearly complete holotype. Simorhinella was named from the skull of a very young specimen and was initially identified as a "scaloposaurid". It was only recognised as a lycosuchid in 2014 following the discovery of a skull from an adult individual. All accepted lycosuchid fossils come from the Beaufort Group of the Karoo Basin in South Africa from rocks of the Abrahamskraal and Teekloof Formations, corresponding to the upper Tapinocephalus Assemblage Zone and lower Endothiodon AZ.
An alternative but functionally identical name for the family, Trochosuchidae, was established by Alfred Romer in 1956, apparently unaware of the pre-existing use of Lycosuchidae by other authors. Romer named the family after Trochosuchus, a now dubious genus of lycosuchid. Curiously, Romer would erect a family for the lycosuchid genera for a second time in 1966, this time as Trochosauridae after the lycosuchid Trochosaurus. Romer likely did this because in 1966 he felt that Trochosuchus was distinct from other lycosuchids and instead assigned it to another family, the Alopecodontidae. This is despite the fact that Romer had previously considered the genera Trochosuchus and Trochosaurus synonymous while under Trochosuchidae. Although Lycosuchidae has priority over either name, some authors perpetuated the use of Trochosuchidae and Trochosauridae, including some who used the former well after Romer proposed replacing it with Trochosauridae.